Split (1)

train · 93.8k rows

file stringlengths 27 28	text stringlengths 791 1.41M
PMC005xxxxxx/PMC5120729.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101665956 44235 Trends Cancer Trends Cancer Trends in cancer 2405-8033 27891533 5120729 10.1016/j.trecan.2016.06.007 NIHMS829254 Article Screening for Cancer in Persons Living with HIV Infection Goedert James J. M.D. 1 Hosgood H. Dean Ph.D. 2 Biggar Robert J. M.D. 3 Strickler Howard D. M.D. 2 Rabkin Charles S. M.D., M.Sc. 1 1 Infections and Immunoepidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 2 Department of Epidemiology and Population Health, Albert Einstein College of Medicine, Bronx, New York 3 Private practice, Bethesda, Maryland Correspondence: Dr. Rabkin, 9609 Medical Center Drive MSC 9767, Bethesda MD 20892. [email protected] 14 11 2016 8 2016 01 8 2017 2 8 416428 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Survival with human immunodeficiency virus (HIV) infection has greatly improved due to effective antiretroviral therapy (ART). As infectious complications have declined, malignancy now accounts for over one-third of deaths among people living with HIV (PLWH). Based on practices in the general population, cancer screening of PLWH can decrease both morbidity and mortality. In this article, we review and consider directed approaches for colorectal, breast, cervical and lung cancer screening. Furthermore, routine physical examinations may detect lymphomas and skin, anal and oral cancers. Comprehensive cancer prevention in PLWH should also include ART adherence, vaccination against oncogenic viruses, treatment of hepatitis viruses and smoking cessation. Cancer screening for PLWH warrants further research on safety and efficacy as well as targeted efforts to increase adherence. Malignancies among people living with HIV Human immunodeficiency virus (HIV) infection destroys CD4+ mononuclear cells, and can lead to severe immune deficiency and the acquired immunodeficiency syndrome (AIDS), including a high risk of the AIDS-defining malignancies – Kaposi sarcoma (KS), non-Hodgkin lymphoma (NHL), and cervical cancer. Some non-AIDS-defining cancers also occur excessively in people living with HIV (PLWH) relative to the general population, especially cancers related to smoking tobacco (e.g., lung cancer) and oncogenic virus co-infections (e.g., Hodgkin lymphoma [HL], anal cancer, and liver cancer). It is not clear whether the cancer risks result from immune dysfunction, higher oncogenic exposures, or both; regardless, these cancers are more common in PLWH and must be of concern to the physician. PLWH are surviving longer through the increasing use and earlier initiation of well tolerated, effective antiretroviral therapy (ART). The median age of PLWH in the United States (USA) today is above 50 years, placing them at growing risk of the age-related malignancies that are common in the general population. Several cohort studies of PLWH have reported that diagnoses of non-AIDS-defining cancers now outnumber AIDS-defining cancers.1,2 For the USA general population, consensus guidelines currently recommend screening for four cancers – cervix, colorectal, breast and lung.3 The survival benefits and adverse consequences of screening for these cancers have not been firmly established for PLWH, but benefits and risks may be assumed similar to those in the general population. Cancer screening can and should be part of routine clinical care for PLWH.4 While detailed regimens of HIV care are beyond the scope of this review, we advocate that PLWH regularly undergo careful, repeated physical examinations and testing for hepatitis B and C viruses (HBV, HCV) to facilitate primary or secondary cancer prevention. In addition, the following PLWH-specific factors are particularly relevant. First, control of HIV replication with an effective and well tolerated ART regimen is the highest priority for managing an individual newly found to be HIV-infected. Thus, efficacy of a prescribed ART regimen should be determined by quantifying HIV viremia and CD4 count. ART toxicities and side effects should be assessed with clinical chemistry, hematology and a symptom checklist, in order to maximize adherence and (early) retention in care. Cancer screening should then be considered once a stable regimen has been established. Second, while ART has reduced the epidemic rates and individual risks of KS and NHL, these cancers continue to occur at much higher rates in PLWH than in the general population.5 Continuing high rates of KS and NHL are mostly related to late presentation; unavailability, intolerance of, or poor adherence to ART; or residual immune perturbations. Thus, detection of KS or NHL in PLWH should signal re-evaluation for treatment failure. Third, many PLWH come from ‘risk-taking’ population groups that have heightened risk of cancer due to certain sexual activities, sex partner selection, injection drug use, and cigarette smoking.6,7 PLWH include a high proportion of men who have sex with men (MSM) who may be be exposed to oral or anal hr-HPV by sexual contact with numerous casual partners.8,9 In addition, the high prevalence of recreational drug use in PLWH may involve exposure to HBV/HCV-contaminated blood and needles or may include sex-for-drugs exchanges that broaden their exposure within a high-risk network. PLWH are also more likely to smoke and less likely to quit smoking than the general adult population.10 Therefore, excess cancers could occur even in PLWH who are immunocompetent, because tobacco carcinogens and oncogenic infections, particularly high-risk human papillomavirus (hrHPV) and HBV/HCV, can lead to cancer regardless of immune status. Fourth, certain communities may have inadequate facilities and providers for safe and effective detection, diagnosis, and treatment of malignancies. As discussed below, this limitation is particularly problematic for cervical cancer, which is a major cause of morbidity and mortality for women in developing countries, and for anal cancer. Appropriate resources for definitive diagnosis and treatment must be assured prior to screening for cancer in any population. The cancers of major importance for PLWH are summarized in the Key Table. While the screening approach outlined in the Key Table is generally applicable to clinically stable PLWH on effective ART, screening practices should be tailored to the individual patient's clinical and exposure history. Major open issues and screening approaches that are currently under consideration are highlighted in the Outstanding Questions Box. Site-Specific Screening Approach Malignancies associated with severe immune deficiency Lymphomas The incidence of AIDS-related non-Hodgkin lymphoma in the USA is diminished by ART, but in the early 2000s NHL was still nearly 7-fold higher in PLWH than in the general population.5 Moreover, it continues to be the most common malignancy in PLWH, arising largely from inadequate HIV treatment.11 Hodgkin lymphoma also remains important because ART is relatively less potent in reducing AIDS-associated incidence of this malignancy.12 Earlier diagnosis and treatment of these disorders would potentially improve patient outcomes. While there are no definitive hematologic or radiographic screening tests for lymphoma, physical examination and chest X-ray performed as part of routine care may detect lymphadenopathy that raises suspicion of malignancy. Because of the high risk of lymphoma, biopsy should be considered for a lymph node that is enlarged to at least 2 cm in diameter, persistent for at least one month, not readily explained by another condition, and especially if accompanied by systemic signs such as fever, weight loss, drenching night sweats, or elevated serum lactate dehydrogenase (LDH). Abdominal, pelvic, hepatic, splenic, or gastrointestinal lymphoma may be undetectable on physical exam and chest X-ray, but additional screening for those internal lesions is not warranted. Kaposi sarcoma KS is caused by infection with human herpesvirus 8 (HHV8, also known as KS-associated herpesvirus) and presents most often as red or violaceous lesions on the skin or in the oral cavity or conjunctivae. Because KS incidence is very high in PLWH,5 especially among sub-Saharan Africans and MSM, two populations with a high likelihood of HHV8 infection,13 the skin, mouth, and eyes should be carefully examined for potential lesions during routine clinical care. As with other AIDS-defining opportunistic diseases, appearance of a KS-suspicious lesion should trigger evaluation of ART failure. Biopsy for histopathologic confirmation of KS is advised if a suspicious lesion is progressing or disseminating, as KS-specific chemotherapy may be required in addition to effective ART.14 Testing for anti-HHV8 antibodies or viremia is not appropriate, as licensed assays for this infection have not been developed and clinical care would not be altered by knowledge of HHV8 status. Cancers associated with oncogenic human papillomavirus co-infection Several HPV vaccines are now widely used in many countries, which may be expected to greatly reduce the incidence of HPV-related cancers in years to come. However, many PLWH are beyond the currently recommended upper age for receiving HPV vaccine, generally 26 years.15 Thus, screening and treatment is still important for cervical cancer prevention for current PLWH and will remain the mainstay for the foreseeable future in populations with limited access to HPV vaccination. Cervical Cancer Invasive cervical cancer (ICC) is the fourth most common cancer in women worldwide,16 and many regions with the highest rates of ICC and ICC-related mortality also have the highest rates of HIV.17 The high risk of ICC in PLWH has important implications for screening practices. For example, a recent multi-cohort study of >13,000 women in the USA found that ICC risk increased with worsening immunosuppression, such that PLWH with CD4+ T-cell counts <200/ul had eight-fold greater ICC incidence than HIV-uninfected women in these same cohorts.18 Importantly, 94% of cases were in women who lacked a recent Pap test (cervical cytology), had an abnormal Pap but no follow-up colposcopy, or had evidence of pre-cancer but no treatment – all of which strongly indict inadequate screening and follow-up as a factor in these cases. Conversely, a cohort study of PLWH who had routine serial Pap testing and treatment reported very low ICC incidence, suggesting that proper surveillance and follow-up can reduce ICC morbidity and mortality in HIV-infected women.19 Indeed, recent data from the USA suggest a modest decrease in ICC incidence among HIV-infected women,20 which likely reflects improving care and management. U.S. Public Health Service (USPHS) guidelines on cervical cancer screening for HIV-infected women (available at http://aidsinfo.nih.gov/contentfiles/lvguidelines/adult_oi.pdf) have recently been updated to incorporate co-testing for hr-HPV as well as longer intervals between screening visits conditional on the test results.21-24 As shown in Table 1, PLWH ≥30 years of age with a normal Pap result who co-test negative for hr-HPV should receive follow-up screening in three years. In contrast, if typing is conducted and HPV16 or HPV18 is detected, immediate colposcopy is warranted; women with other hr-HPV types but a normal Pap should be screened again in one year. Because HPV prevalence is very high and likely to clear spontaneously in young sexually active women, hr-HPV testing is not recommended in women <30 years of age. Regardless of age, a three-year screening interval is recommended for women who have had three consecutive normal Pap tests. While primary HPV screening (e.g., instead of Pap tests) was recently approved for use in the general population in the USA, this approach is not currently recommended for PLWH, nor is it recommended that PLWH discontinue cervical cancer screening after age 65, as in the general population. Despite negative results of HPV-cotesting, many HIV-infected women without clinically significant cervical disease are inappropriately referred to colposcopy, which is considered an unnecessary harm by USPHS standards. Thus, continued improvement in screening methods is warranted. Resource Limited Settings Cervical cancer risk is increased among populations with inadequate access to health services; as individuals diagnosed with HIV become engaged in appropriate care, integration of cervical cancer screening is a potential approach to reduce morbidity and mortality.25 However, Pap-based screening is impractical in most resource-limited settings, because of high costs, the paucity of trained clinical and laboratory personnel, and the requirement for multiple clinical visits, sometimes with lengthy intervals awaiting laboratory results. Instead, the World Health Organization (WHO) recommends a “screen-and-treat” approach that allows for treatment soon (if not immediately) after screening with hr-HPV testing, followed by visual inspection after acetic acid application (VIA; available at http://www.who.int/reproductivehealth/publications/cancers/screening_and_treatment_of_precancerous_lesions/en/). If the HPV test is positive, VIA results are used to determine the extent of the lesions, the immediacy of treatment, and the optimal therapy (e.g., cryosurgery vs. Loop Electrosurgical Excision Procedure).26 If HPV testing is unavailable, “screen-and-treat” can be based solely on VIA, although this approach is considered inferior to HPV screening. Women in the general population who have a normal VIA (or normal Pap) can be rescreened in 3-5 years, and those who are hr-HPV negative in 5 years. Although data in PLWH are limited, the main difference in recommendations for PLWH under the current WHO guidelines is that they should be re-screened 3 years following a negative hr-HPV test or normal VIA or Pap. Anal Cance The incidence of anal cancer is elevated in PLWH relative to the general population,6,27,28 particularly in HIV-infected MSM (who account for 83% of all cases in PLWH).7,29-31 For several reasons, however, anal cancer screening is not currently routine even in HIV-infected MSM. Most importantly, it has not yet been established that treatment of high-grade anal intraepithelial neoplasia (hgAIN) reduces the risk of invasive anal cancer.32 To address this issue, a large multi-institutional randomized clinical trial, the ANCHOR (Anal Cancer/HSIL Outcomes Research) study, is currently enrolling 5000 PLWH with biopsy-confirmed hgAIN to compare treatment versus close observation for at least 5 years. Pending these results, a range of screening practices and treatment approaches have been adopted, with no uniform national or international recommendations for routine anal cancer screening.33 Most proposed strategies follow a model similar to that used for ICC, namely, cytology and/or hr-HPV testing, followed by high resolution anoscopy with biopsy if indicated.33 More than 70% of anal cancers in the general population are positive for HPV16,34-37 but anal cancers among PLWH contain a broader distribution of hrHPV types. Furthermore, even in the absence of lesions, more than two-thirds of HIV-infected MSM have anal hrHPV.38 Thus, anal HPV screening will likely need to be combined with other methods, and cost effectiveness will need to be evaluated. Digital ano-rectal examination (DARE) is currently being studied as a screening approach, including self-screening.39,40 Farther on the horizon, serologic screening for HPV16 E6 antibodies might prove useful in some populations but has not yet been assessed in PLWH.41 It is widely agreed that anal cancer screening should not be done without the availability of referral for high resolution anoscopy, and subsequent treatment if indicated. Overall, our approach is to consider anal Pap testing or DARE only for HIV-infected MSM, assuming that the resources are in place to conduct high resolution anoscopy and follow-up treatment if indicated. We note that this type of care will not be available in most resource-limited areas, or even in many settings in developed countries. Oral and Oropharyngeal Cancer Squamous cell cancers at these sites have several commonalities, including three-fold higher incidence in PLWH compared to the general population, four-fold higher risk with older age (>50 years versus <40 years), and higher risks with tobacco use or lower CD4 count.6,7,42 However, differences between cancers of the oral cavity and cancers of the oropharynx are important, especially for prevention. HPV, especially HPV16, is a major contributor to cancer of the tonsil and other oropharyngeal sites, whereas this virus is seldom found in cancers of the oral cavity. Among PLWH, the incidence of HPV-related cancer has been increasing, while the incidence of HPV-unrelated cancer has been decreasing.42 Detection of HPV16 E6 antibodies may be predictive of oropharyngeal cancer, but its utility in PLWH is unknown.43 Oropharyngeal cancers are not readily detected by visual inspection, and detection of these by oral cytology has been disappointing.44 Unlike cancers of the oropharynx, most cancers of the oral cavity are visible and therefore amenable to detection by inspection. To be protective, screening must detect and enable effective treatment of early cancers or truly precancerous lesions. Several “oral potentially malignant disorders” (OPMDs) have been recognized and associated with high risk of oral cancer.45 These lesions include erythroplakia, leukoplakia (not to be confused with the Epstein-Barr virus-related oral hairy leukoplakia that occurs in immunocompromised PLWH), mixed red and white lesions, submucous fibrosis, and subtypes of lichen planus.46,47 Further, while expert opinion suggests screening for these lesions in the general population using standard visual and tactile exam (VTE)44 followed by surgical biopsy, this approach is associated with potential morbidity.48 Moreover, both a meta-analysis and a subsequent, large cohort study questioned the clinical utility of VTE and surgical biopsy49,50 Adjunctive visualization methods such as VELscope and Vizilite have shown good sensitivity but poor specificity.49 While results with oral cytology in those with visible OPMD have been encouraging (e.g., to triage who requires biopsy), appropriate studies in the general clinic population are required, and there is a paucity of relevant data in HIV-infected individuals. Based on sparse data, the prevalence of OPMD is probably low among PLWH in highly industrialized countries.51,52,53 There is insufficient evidence to recommend intensive screening approaches, particularly since randomized trials have not been conducted to assess the efficacy of surgery in preventing cancer morbidity or mortality in any population.54 However, a community-based trial of oral examinations in India, where incidence is high in the general population, reported a 24% reduction in oral cancer mortality among tobacco and alcohol users.55 In a study in US Medicare beneficiaries, oral cavity cancer was reported to be diagnosed at an earlier stage among people with, versus those without, a prior finding of leukoplakia.56 Although the benefits and risks are uncertain, careful examination of the mouth with biopsy of suspicious lesions as part of routine HIV care may afford detection of cancer at a curable stage. Cancer associated with chronic hepatitis virus co-infection Liver cancer Nearly all cases of liver cancer are hepatocellular carcinoma (HCC), which generally arises in cirrhosis that is the end result of chronic active infection with HBV or HCV.57 The disease is highly lethal unless detected very early. Screening with liver ultrasonography for single, small HCC lesions has been endorsed by a panel of expert hepatologists,58 but this approach is very controversial.58,59 The most recent Cochrane review identified only three randomized clinical trials that were all deemed to be at high risk of bias,60 and the relevance of findings from HBV carriers in Asia to HCV carriers in the USA and other developed countries has been questioned.59 Among PLWH, liver cancer risk is elevated more than 3-fold compared to the demographically adjusted general population.61 Thus, compared to the general population, the potential benefits of screening for liver cancer may be modestly higher for cirrhotic PLWH who have chronic HBV or HCV infection. However, there are no data on the sensitivity, specificity, positive/negative predictive values, or adverse consequences of screening for HCC in PLWH.62 HCC may be discovered in HBV/HCV co-infected PLWH during evaluation of cirrhosis,63 but data are insufficient to justify screening for potentially treatable HCC. Instead, ascertainment of HBV and HCV status in all PLWH at initial presentation, followed by primary prevention through HBV vaccination and treatment of chronic HCV or HBV has low risk and high efficacy. Other major cancers Lung cancer Lung cancer is the most important non-AIDS-defining cancer for PLWH. Due to its high incidence and high case-fatality rate, lung cancer accounts for about 30% of all cancer deaths and 10% of all non-HIV-associated deaths in PLWH.64,65 As in the general population, the disease is often locally advanced or disseminated at diagnosis. Even after adjustment for reported tobacco use, PLWH are estimated to have at least two-fold higher risk than the general population.66-68 Obviously, strong efforts to encourage smoking cessation are warranted. With post-diagnosis survival similar in PLWH compared to the general population, early diagnosis and treatment may be beneficial. For the general population, based largely on results from the National Lung Screening Trial (NLST),69 the United States Preventive Services Task Force (USPSTF) recommends annual low-dose computed tomography (LD-CT) for tobacco smokers ages 55-80 years with a history of at least 30 pack-years of smoking and who currently smoke or have quit within the past 15 years, provided they would be able to tolerate lung cancer surgery if indicated.70 Whether these recommendations are appropriate for PLWH is unclear. A particular concern, which might argue against screening, is that false positive findings (e.g., due to calcified tuberculous- or fungal-related lung nodules) could in theory be more frequent in PLWH and lead to more unnecessary invasive procedures for definitive diagnosis, though there is a paucity of relevant data. Conversely, PLWH diagnosed with lung cancer tend to be younger and have lower pack-years of exposure, which would suggest that screening should be done with modified criteria. 71 In a study of 442 French PLWH of median age 49.8 years who were screened with one LD-CT, there were no serious adverse events; lung cancer was found in 10 (2%), including 5 who were outside USPSTF guidelines but had early, resectable cancer.72 Clinicians should maintain a high index of suspicion for lung cancer in PLWH who have a history of cigarette smoking. Criteria for LD-CT screening should follow USPSTF guidelines for the general population, pending additional studies in younger and moderate-smoker PLWH to assess sensitivity, specificity, and complication rates from follow-up procedures. Colorectal cancer Colorectal cancer is the second leading cause of cancer death in the general population; and screening by colonoscopy or repeated fecal occult blood testing, with appropriate follow-up, markedly reduces mortality.73 While colorectal cancer incidence in PLWH is not elevated relative to the general population,5 aging PLWH are still at risk and require appropriate screening. However, two studies have reported that colorectal cancer screening rates in PLWH are 20% lower than in the general population.74,75 Increasing PLWH participation in colorectal cancer screening would be expected to improve their survival. Breast cancer As for colorectal cancer, the risk of female breast cancer for PLWH is similar to that observed in the general population.5 Compared to no screening, mammographic screening to detect early stage disease has been estimated to reduce breast cancer mortality by 15-35% and to reduce total mortality by 1%.76 On the assumption that screening efficacy would match efficacy in the general population, biennial mammography to detect early breast cancer should be encouraged for female PLWH ages 50-74 years who are clinically stable.77 As is true for women in general, a screening mammogram can be considered on a case-by-case basis for younger female PLWH deemed to be at exceptionally high risk by virtue of breast biopsy history or breast cancer in first-degree relatives. Unfortunately, survey data suggest that female PLWH have a substantial deficit in the use of screening mammography.74 Prostate cancer Potential screening modalities for prostate cancer include serologic testing for prostate specific antigen (PSA) and digital rectal examination, although these modalities either individually or combined have low sensitivity, specificity and positive predictive value. Prostate cancer is a common “incidental finding,” and there are no accepted tests or criteria to determine which asymptomatic cancers found by screening warrant aggressive treatment to avert future health problems. The value of screening in the general population is debated, with the relatively small survival benefit possibly outweighed by substantial morbidities associated with overdiagnosis and overtreatment of non-life-threatening cancers.78,79 Importantly, PLWH do not have increased prostate cancer incidence,5 but prostate cancers in PLWH do tend to present at more advanced stage and to be associated with 70% higher cancer-specific mortality.80,81 For PLWH who have extended life expectancy on ART, considerations for or against prostate cancer screening are the same as in the general population. American Cancer Society (ACS) guidelines support PSA screening for African American men and those with a strong family history of prostate cancer at a young age. Melanoma and non-melanoma skin cancers Skin examination during routine HIV care can detect not only KS but also malignant melanoma and non-melanoma skin cancers. The incidence of advanced-stage melanoma and melanoma-specific mortality are both approximately doubled in PLWH.80,81 Common basal and squamous cell skin cancers, as well as rare adnexal and skin appendage neoplasms, are increased nearly 3-fold, but few of these cancers are lethal.82,83 Particular attention should be paid to PLWH with fair complexion or history of excessive sun exposure, as these characteristics may be common in some AIDS-risk groups and are major risk factors for melanoma and other skin cancers. Suspicious skin lesions, especially those fitting the ABCDE mnemonic for suspicion of melanoma (Asymmetrical, Border irregular, Color irregular, Diameter >6mm, Enlarging) should be excised, yielding diagnosis and potential cure. Pathologic confirmation of melanoma should trigger comprehensive clinical staging. Concluding Remarks In the 20 years since the introduction of ART, the burden of diseases, causes of death, and health priorities for PLWH have changed dramatically.84 Non-AIDS malignancies have emerged as major causes of morbidity and mortality due to prolonged infection by oncogenic viruses, accumulating tobacco exposures, and population aging, although two AIDS-defining cancers, NHL and KS, remain the most common in PLWH. Accordingly, effective ART, vaccination against HPV and HBV, treatment of chronic HBV and HCV, and cessation of smoking are primary measures of cancer prevention. Screening and treatment of curable cancer represents potentially beneficial, secondary prevention. The Outstanding Questions Box highlights some of the pressing issues to address toward further reducing cancer morbidity and mortality in PLWH. Screening programs in the general population are proven to reduce mortality from cervical cancer, colorectal cancer, breast cancer, and lung cancer. Participation in these programs may be expected to reduce cancer mortality for PLWH as well, but empirical data are needed. Quantification of screening sensitivity, specificity, and adverse event rates in so-called “bridging studies” of PLWH could energize implementation and verify favorable risk-benefit ratios for this population. Modified procedures to screen for cervical neoplasia in PLWH have already been recommended (Table 1), and modification of lung cancer screening (for younger ages and lighter smoking histories) may also be found appropriate. A major concern is that PLWH, disproportionally poorly educated and socio-economically of low status, too often fail to get the cancer screening recommended for the general population or fail to have adequate follow-up of positive findings.85,86 This public health problem warrants the attention of national and community health services. Furthermore, in resource limited settings, cancer screening remains a challenging issue, not only in its implementation but also in the inability to respond to positive findings. PLWH require clinical monitoring for ART drug toxicities and efficacy against HIV. We advocate routine physical examination by their clinical care providers, because it affords the possibility to detect KS and other skin malignancies, lymphomas, oral cancers, and anal cancer. While the efficacy of physical examination to detect cancer, as well as the adverse event rates from follow-up diagnostic procedures and treatment, are largely unknown, cancer rates and thus potential benefits are higher for PLWH than the general population. If physical examination is safe and even modestly effective for cancer screening, it could be rapidly applied to improving health care of PLWH, particularly in resource-limited settings. Supplementary Material Material Materiala The authors thank their many colleagues for sharing their experience and their critiques of this review, as well as the Reviewers for their constructive comments. Glossary Acquired immune deficiency syndrome (AIDS) Terminal stage of human immunodeficiency virus infection, manifesting as a life-threatening opportunistic infection or malignancy. Anal Cancer/HSIL Outcomes Research study (ANCHOR) A research study, sponsored by U.S. National Cancer Institute AIDS Malignancy Consortium, to evaluate whether detection and treatment of high-grade intraepithelial neoplasia in the anal canal effectively reduces anal cancer development with acceptably low side effects for PLWH. Results are expected in 2022. Antiretroviral therapy (ART) A combination of two or three licensed medications that block different stages of the human immunodeficiency virus lifecycle, allowing partial recovery of immunity but not eradication of the infection. Bridging study A supplemental research study designed to provide data allowing extrapolation of prior findings to a different population. Hepatitis B virus (HBV) A partially double-stranded DNA virus of humans that is highly infectious and primarily transmitted by direct contact with blood or other body fluids. HBV primarily targets and replicates in hepatocytes. Chronic HBV infection, manifest as persistence of HBV antigens and genome in the blood, ensues in approximately 90% of infected neonates and 10% of people infected as adults. HBV-related hepatic inflammation can progress to fibrosis, cirrhosis, and hepatocellular carcinoma. Hepatitis C virus (HCV) A single-stranded RNA virus of humans that is highly infectious but almost exclusively transmitted by direct contact with blood. Chronic HCV infection and hepatic inflammation persist in the majority of infected people, progressing slowly but unpredictably to fibrosis, cirrhosis, and hepatocellular carcinoma. High-grade anal intraepithelial neoplasia (hg-AIN) A potentially, but as yet unproven, pre-cancerous abnormality of the squamous epithelium of the anal canal. Human herpesvirus 8 (HHV8) A large double-stranded DNA virus of humans that is the primary cause of Kaposi sarcoma and that is also known as the Kaposi sarcoma-associated herpesvirus (KSHV). HHV8 is primarily transmitted by saliva, persists lifelong in lymphocytes, and seldom manifests as Kaposi sarcoma or another disease, except in people with AIDS or another deficiency in cell-mediated immunity. Human immunodeficiency virus (HIV) The human lentivirus that targets, replicates in, and progressively destroys T lymphocytes, resulting in profound deficiency in cell-mediated immunity that ultimately manifests as AIDS. HIV is primarily transmitted by sexual intercourse, blood injection or transfusion, during childbirth, or by breastfeeding. HIV's pathogenesis is greatly reduced by adherence to an effective ART regimen. Hodgkin lymphoma (HL) Malignancy of B lymphocytes with peak incidence rates at ages 15-35 and after age 55. Typically presents as one or more enlarged lymph nodes (especially cervical or supraclavicular) plus fevers, weight loss and night sweats. The disease often responds favorably to intensive, combination chemotherapy. Human papilloma virus (HPV) A small, highly diverse DNA virus usually transmitted by direct contact. Of the many HPV types that infect mucosae, four (HPV-16, -18, -31, and -45) cause nearly all cervical cancers. These are classified as high-risk HPV (hr-HPV). Licensed vaccines safely and effectively prevent nearly all hr-HPV infections. High-grade squamous intraepithelial lesion (HSIL) A pre-cancerous abnormality of the squamous epithelium of the cervix. HSIL can be detected by routine Pap testing (cervical cytology) and effectively treated by Loop Electrosurgical Excision Procedure or cryosurgery. Invasive cervical cancer (ICC) Malignancy of the epithelium of the cervix that is caused by long-term, persistent infection with high-risk HPV. Some individuals progress to pre-cancerous abnormalities (low-grade and high-grade squamous intraepithelial lesion) that may be detected by Pap testing (cervical cytology). Kaposi sarcoma (KS) A malignancy of mesenchymal (connective) tissue that probably originates in lymphatic endothelial cells. KS appears most often as red, purple, or darker patches or papules, especially on the skin but also on mucous membranes. KS is primarily caused by human herpesvirus 8 infection, and risk is greatly increased with HIV-, transplant-, or medication-induced deficiency of cell-mediated immunity. Lactate dehydrogenase (LDH) An enzyme in all cells that catalyzes the conversion of pyruvate to lactate and back again, contingent on availability of oxygen. Increased levels of LDH in serum occur with a range of conditions including myocardial infarction, hemolysis, and several malignancies including lymphomas. Non-Hodgkin lymphoma (NHL) A highly diverse group of approximately 80 malignancies of lymphocytes. Most NHLs associated with AIDS are diffuse large B-cell or Burkitt lymphomas. AIDS NHL is usually highly aggressive, often lethal, and frequently arises in the central nervous system and other extranodal sites. These tumors may respond to intensive, combination chemotherapy, although survival benefit is variable. National Lung Screening Trial (NLST) Clinical trial at 33 medical centers in the USA in which 53,454 current or recent heavy smokers, ages 55-74 years, were randomized to three annual examinations with low-dose computed tomography (LD-CT) versus conventional chest X-ray. As lung cancer was more often detected at a treatable stage, participants who were randomized to LD-CT had 20% lower mortality due to lung cancer and 8% lower mortality due to all causes. Oral potentially malignant disorders (OPMD) A group of potentially, but as yet unproven, pre-cancerous abnormalities of the oral mucosa including erythroplakia, leukoplakia, mixed red and white lesions, submucous fibrosis, and subtypes of lichen planus. Major risk factors are tobacco smoking, use of smokeless tobacco, and heavy alcohol consumption. United States Preventive Services Task Force (USPSTF) An independent panel of experts in primary care and prevention that systematically reviews the evidence of effectiveness and develops recommendations for clinical preventive services. The USPSTF is funded, staffed, and appointed by the U.S. Department of Health and Human Services. Table 1 Recommendations for Cervical Cancer Screening for HIV-Infected Women* HIV-Infected Women Aged <30 Years: If younger than age 21, known to be HIV-infected or newly diagnosed with HIV, and sexually active, screen within 1 year of onset of sexual activity regardless of mode of HIV infection. HIV-infected women aged 21-29 should have a Pap test following initial diagnosis. Pap test should be done at baseline and every 12 months. Some experts recommend a Pap test at 6 months after baseline test. If results of 3 consecutive Pap tests are normal, follow-up Pap tests can be performed every 3 years. Co-testing (Pap test and HPV test) is not recommended. HIV-Infected Women Aged ≥30 Years: Pap Testing Only: Pap test should be done at baseline and every 12 months. Some experts recommend a Pap test at 6 months after baseline test. If results 3 consecutive Pap tests are normal, follow-up Pap tests can be performed every 3 years. Or Pap Test and HPV Co-Testing: Pap test and HPV co-testing should be done at baseline. If result of the Pap test is normal and HPV co-testing is negative, follow-up co-testing can be performed every 3 years. If the result of the Pap test is normal but HPV co-testing is positive, follow-up co-testing should be performed in one year. If the one-year follow-up Pap test is abnormal or HPV co-testing is positive, referral to colposcopy is recommended. Or Pap Test and HPV 16 or HPV 16/18 Specified in Co-Testing: Pap test and HPV 16 or 16/18 co-testing should be done at baseline. If result of the Pap test is normal and HPV 16 or 16/18 co-testing is negative, follow-up co-testing can be performed every 3 years. If initial test or follow-up test is positive for HPV 16 or 16/18, referral to colposcopy is recommended. * Adapted from guidelines of the United States Public Health Service, Centers for Disease Control and Prevention, National Institutes of Health, and American Congress of Obstetrics and Gynecology. Key Table Malignancies potentially amenable to screening for early diagnosis and intervention in persons living with HIV (PLWH) Malignancy Annual No. PLWH Cases in the USA* Method Benefit Risks Common Treatment Authors' Approach Notes Associated with severe immune deficiency Lymphomas, non-Hodgkin (NHL) and Hodgkin (HL) 1970 Careful lymph node exam, chest X-ray Potential early diagnosis at lower stage Minimal Cytotoxic chemotherapy Biopsy node if accompanied by other signs or symptoms that are highly suspicious of malignancy Annual cases NHL: 1650 HL: 320 Kaposi Sarcoma (KS) 910 Careful exam of skin and conjunctivae HIV/AIDS prognostic marker Minimal Consider chemotherapy; indicative of antiretroviral treatment failure, which should be investigated. Biopsy lesion at follow-up if progression or dissemination is suspected 88% occur in men who have sex with men (MSM) Associated with oncogenic human papillomavirus (HPV) co-infection Cervical Cancer 80 Pap test +/- high-risk (hr) HPV testing Lower mortality with routine Pap (53-80%↓) Anxiety, pain, bleeding with colposcopy and biopsy For precancer: ablation or excision. Age 30+ yrs: Pap test alone, or Pap plus hr-HPV co-testing. Age <30 yrs: Pap test alone. See Table 1. PLWH have high rates of invasive cancer in developing countries and of precancer in the USA. Anal Cancer 760 Digital anal rectal exam (DARE); anal Pap test Unproven Anxiety, pain, bleeding with high-resolution anoscopy and biopsy. For high-grade anal intraepithelial neoplasia (hg-AIN): ablation, cautery, infrared coagulation, laser, or cryotherapy. Only in MSM and only if necessary diagnostic and treatment resources available: DARE or Pap test; if suspicious for malignancy, high-resolution anoscopy, possible biopsy. HPV testing not recommended. 83% occur in MSM; uncertain whether hg-AIN is a cancer precursor. Oral and Oropharyn-geal Cancer 280 Careful visual and tactile oral exam Unproven Anxiety, pain, bleeding with biopsy of suspect lesion Excision, radiotherapy, chemotherapy Biopsy lesion(s) if highly suspicious of malignancy Suspect lesions: erythroplakia, leukoplakia, mixed red and white lesions, submucous fibrosis, lichen planus subtypes Associated with chronic hepatitis virus co-infection Liver Cancer 390 Liver sonography if cirrhosis is suspected Inconsistent evidence for hepatitis B virus (HBV), no data for hepatitis C virus (HCV) and other etiologies Anxiety, pain, bleeding, possible bile peritonitis or pneumothorax with biopsy of suspect lesion Surgery, chemotherapy Serologic screening for treatable (HCV, HBV) or preventable (HBV) infections; no cancer screening Cancer risk reduced by prevention or effective treatment of HCV and HBV Other Lung Cancer 840 Low-dose computed tomography (LD-CT) Lower mortality in current and recent smokers of 30+ pack-yrs(20%↓ lung cancer deaths, 8%↓ overall) Anxiety, pain, bleeding, possible pneumothorax with biopsy of suspect lesion Surgery, radiotherapy, chemotherapy Annual LD-CT as per US Preventative Services Task Force (USPSTF) criteria;assess efficacy in younger, moderate-smokers USPSTF: Current or recent smoker, age 55-80 yrs, 30+ pack year smoking history, able to tolerate lung cancer surgery Colorectal Cancer 360 Colono-scopy or annual fecal occult blood test 18,800 fewer all-cause deaths/yr in USA Colonoscopy-related gut perforation, bleeding Polyp excision, cancer surgery, chemotherapy Screen age 50-75 yrs as per USPSTF Colonoscopy repeat interval: q10 yrs if normal, q3 yrs if polyps found Breast Cancer 180 Mammo-graphy Lower mortality (15-35%↓ breast cancer deaths, 1% overall) Anxiety, pain Surgery, hormonal therapy, chemotherapy Biennial screen age 50-74 yrs as per USPSTF; age <50 or 75+ yrs, consult physician Prostate Cancer 570 Prostate specific antigen (PSA) Inconsistent evidence of minimally reduced mortality (0.1%↓) Anxiety, pain, bleeding with biopsy of suspect lesion Surgery, radiotherapy, hormone therapy If life expectancy 10+ years, consider biennial PSA screening only for high/very high risk as per American Cancer Society (ACS) criteria (differs from USPSTF) ACS criteria: High risk = African-American or prostate cancer age<65 in a brother or father; Very high risk = prostate cancer age<65 in >1 brothers/father From Robbins HA, et al.21 i http://aidsinfo.nih.gov/contentfiles/lvguidelines/adult_oi.pdf ii www.who.int/reproductivehealth/publications/cancers/screening_and_treatment_of_precancerous_lesions/en/ 1 Long JL Engels EA Moore RD Gebo KA Incidence and outcomes of malignancy in the HAART era in an urban cohort of HIV-infected individuals Aids 2008 22 489 96 18301061 2 Patel P Hanson DL Sullivan PS Incidence of types of cancer among HIV-infected persons compared with the general population in the United States, 1992-2003 Ann Intern Med 2008 148 728 36 18490686 3 Sabatino SA White MC Thompson TD Klabunde CN Cancer screening test use - United States, 2013 MMWR Morbidity and mortality weekly report 2015 64 464 8 25950253 4 Mani D Aboulafia DM Screening guidelines for non-AIDS defining cancers in HIV-infected individuals Curr Opin Oncol 2013 25 518 25 23942295 5 Engels EA Biggar RJ Hall HI Cancer risk in people infected with human immunodeficiency virus in the United States Int J Cancer 2008 123 187 94 18435450 6 Grulich AE van Leeuwen MT Falster MO Vajdic CM Incidence of cancers in people with HIV/AIDS compared with immunosuppressed transplant recipients: a meta-analysis Lancet 2007 370 59 67 17617273 7 Chaturvedi AK Madeleine MM Biggar RJ Engels EA Risk of human papillomavirus-associated cancers among persons with AIDS J Natl Cancer Inst 2009 101 1120 30 19648510 8 Goedert JJ Sarngadharan MG Biggar RJ Determinants of retrovirus (HTLV-III) antibody and immunodeficiency conditions in homosexual men Lancet 1984 2 711 6 6148471 9 Poynten IM Machalek D Templeton D Comparison of age-specific patterns of sexual behaviour and anal HPV prevalence in homosexual men with patterns in women Sexually transmitted infections 2016 92 228 31 26306829 10 Mdodo R Frazier EL Dube SR Cigarette smoking prevalence among adults with HIV compared with the general adult population in the United States: cross-sectional surveys Ann Intern Med 2015 162 335 44 25732274 11 Achenbach CJ Buchanan AL Cole SR HIV viremia and incidence of non-Hodgkin lymphoma in patients successfully treated with antiretroviral therapy Clin Infect Dis 2014 58 1599 606 24523217 12 Goedert JJ Bower M Impact of highly effective antiretroviral therapy on the risk for Hodgkin lymphoma among people with human immunodeficiency virus infection Curr Opin Oncol 2012 24 531 6 22729154 13 Rohner E Wyss N Heg Z HIV and human herpesvirus 8 co-infection across the globe: Systematic review and meta-analysis Int J Cancer 2016 138 45 54 26175054 14 Bower M Dalla Pria A Coyle C Prospective stage-stratified approach to AIDS-related Kaposi's sarcoma J Clin Oncol 2014 32 409 14 24378415 15 Markowitz LE Dunne EF Saraiya M Human papillomavirus vaccination: recommendations of the Advisory Committee on Immunization Practices (ACIP) MMWR Recommendations and reports : Morbidity and mortality weekly report Recommendations and reports / Centers for Disease Control 2014 63 1 30 16 Ferlay J Soerjomataram I Dikshit R Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012 Int J Cancer 2015 136 E359 86 25220842 17 McDonald AC Tergas AI Kuhn L Denny L Wright TC Jr Distribution of Human Papillomavirus Genotypes among HIV-Positive and HIV-Negative Women in Cape Town, South Africa Front Oncol 2014 4 48 24672770 18 Abraham AG D'Souza G Jing Y Invasive cervical cancer risk among HIV-infected women: a North American multicohort collaboration prospective study J Acquir Immune Defic Syndr 2013 62 405 13 23254153 19 Massad LS Seaberg EC Watts DH Long-term incidence of cervical cancer in women with human immunodeficiency virus Cancer 2009 115 524 30 19127538 20 Robbins HA Shiels MS Pfeiffer RM Engels EA Epidemiologic contributions to recent cancer trends among HIV-infected people in the United States Aids 2014 28 881 90 24300545 21 D'Souza G Burk RD Palefsky JM Massad LS Strickler HD Cervical human papillomavirus testing to triage borderline abnormal pap tests in HIV-coinfected women Aids 2014 28 1696 8 25232904 22 Massad LS D'Souza G Tian F Negative predictive value of pap testing: implications for screening intervals for women with human immunodeficiency virus Obstetrics and gynecology 2012 120 791 7 22996096 23 Keller MJ Burk RD Massad LS Cervical Precancer Risk in HIV-Infected Women Who Test Positive for Oncogenic Human Papillomavirus Despite a Normal Pap Test Clin Infect Dis 2015 61 1573 81 26187020 24 Keller MJ Burk RD Xie X Risk of cervical precancer and cancer among HIV-infected women with normal cervical cytology and no evidence of oncogenic HPV infection Jama 2012 308 362 9 22820789 25 Huchko MJ Maloba M Nakalembe M Cohen CR The time has come to make cervical cancer prevention an essential part of comprehensive sexual and reproductive health services for HIV-positive women in low-income countries Journal of the International AIDS Society 2015 18 20282 26643456 26 WHO Guidelines Approved by the Guidelines Review Committee WHO Guidelines for Screening and Treatment of Precancerous Lesions for Cervical Cancer Prevention Geneva World Health Organization 2013 27 Frisch M Biggar RJ Goedert JJ Human papillomavirus-associated cancers in patients with human immunodeficiency virus infection and acquired immunodeficiency syndrome J Natl Cancer Inst 2000 92 1500 10 10995805 28 Palefsky JM Gillison ML Strickler HD Chapter 16: HPV vaccines in immunocompromised women and men Vaccine 2006 24 3 S3/140 6 29 Robbins HA Pfeiffer RM Shiels MS Li J Hall HI Engels EA Excess cancers among HIV-infected people in the United States J Natl Cancer Inst 2015 107 30 Silverberg MJ Lau B Justice AC Risk of anal cancer in HIV-infected and HIV-uninfected individuals in North America Clin Infect Dis 2012 54 1026 34 22291097 31 Piketty C Selinger-Leneman H Bouvier AM Incidence of HIV-related anal cancer remains increased despite long-term combined antiretroviral treatment: results from the french hospital database on HIV J Clin Oncol 2012 30 4360 6 23091098 32 Palefsky JM Screening to prevent anal cancer: Current thinking and future directions Cancer cytopathology 2015 123 509 10 26237741 33 Patel J Salit IE Berry MJ Environmental scan of anal cancer screening practices: worldwide survey results Cancer medicine 2014 3 1052 61 24740973 34 Abramowitz L Jacquard AC Jaroud F Human papillomavirus genotype distribution in anal cancer in France: the EDiTH V study Int J Cancer 2011 129 433 9 20839262 35 De Vuyst H Clifford GM Nascimento MC Madeleine MM Franceschi S Prevalence and type distribution of human papillomavirus in carcinoma and intraepithelial neoplasia of the vulva, vagina and anus: a meta-analysis Int J Cancer 2009 124 1626 36 19115209 36 Steinau M Unger ER Hernandez BY Human papillomavirus prevalence in invasive anal cancers in the United States before vaccine introduction Journal of lower genital tract disease 2013 17 397 403 23609590 37 Hoots BE Palefsky JM Pimenta JM Smith JS Human papillomavirus type distribution in anal cancer and anal intraepithelial lesions Int J Cancer 2009 124 2375 83 19189402 38 Wiley DJ Li X Hsu H Factors affecting the prevalence of strongly and weakly carcinogenic and lower-risk human papillomaviruses in anal specimens in a cohort of men who have sex with men (MSM) PloS one 2013 8 e79492 24278140 39 Ong JJ Temple-Smith M Chen M Walker S Grulich A Fairley CK Exploring anal self-examination as a means of screening for anal cancer in HIV positive men who have sex with men: a qualitative study BMC public health 2014 14 1257 25496368 40 Ong J Chen M Temple-Smith M The inside story. Physicians' views on digital ano-rectal examination for anal cancer screening of HIV positive men who have sex with men Journal of medical screening 2013 20 188 91 24307004 41 Kreimer AR Brennan P Lang Kuhs KA Human papillomavirus antibodies and future risk of anogenital cancer: a nested case-control study in the European prospective investigation into cancer and nutrition study J Clin Oncol 2015 33 877 84 25667279 42 Beachler DC Abraham AG Silverberg MJ Incidence and risk factors of HPV-related and HPV-unrelated Head and Neck Squamous Cell Carcinoma in HIV-infected individuals Oral oncology 2014 50 1169 76 25301563 43 Kreimer AR Johansson M Waterboer T Evaluation of human papillomavirus antibodies and risk of subsequent head and neck cancer J Clin Oncol 2013 31 2708 15 23775966 44 Lingen MW Brush-based cytology screening in the tonsils and cervix: there is a difference! Cancer prevention research (Philadelphia, Pa) 2011 4 1350 2 45 Edwards PC The natural history of oral epithelial dysplasia: perspective on Dost et al Oral surgery, oral medicine, oral pathology and oral radiology 2014 117 263 6 46 Panwar A Lindau R Wieland A Management for premalignant lesions of the oral cavity Expert review of anticancer therapy 2014 14 349 57 24559323 47 Warnakulasuriya S Johnson NW van der Waal I Nomenclature and classification of potentially malignant disorders of the oral mucosa Journal of oral pathology & medicine : official publication of the International Association of Oral Pathologists and the American Academy of Oral Pathology 2007 36 575 80 48 Mehrotra R Gupta DK Exciting new advances in oral cancer diagnosis: avenues to early detection Head & neck oncology 2011 3 33 21798030 49 Epstein JB Guneri P Boyacioglu H Abt E The limitations of the clinical oral examination in detecting dysplastic oral lesions and oral squamous cell carcinoma Journal of the American Dental Association (1939) 2012 143 1332 42 23204089 50 Goodson ML Sloan P Robinson CM Cocks K Thomson PJ Oral precursor lesions and malignant transformation--who, where, what, and when? The British journal of oral & maxillofacial surgery 2015 53 831 5 26388071 51 Ranganathan K Umadevi M Saraswathi TR Kumarasamy N Solomon S Johnson N Oral lesions and conditions associated with human immunodeficiency virus infection in 1000 South Indian patients Annals of the Academy of Medicine, Singapore 2004 33 37 42 52 Patton LL Current strategies for prevention of oral manifestations of human immunodeficiency virus Oral surgery, oral medicine, oral pathology and oral radiology 2016 121 29 38 53 Petersen PE Policy for prevention of oral manifestations in HIV/AIDS: the approach of the WHO Global Oral Health Program Advances in dental research 2006 19 17 20 16672544 54 Balasundaram I Payne KF Al-Hadad I Alibhai M Thomas S Bhandari R Is there any benefit in surgery for potentially malignant disorders of the oral cavity? Journal of oral pathology & medicine : official publication of the International Association of Oral Pathologists and the American Academy of Oral Pathology 2014 43 239 44 55 Sankaranarayanan R Ramadas K Thara S Long term effect of visual screening on oral cancer incidence and mortality in a randomized trial in Kerala, India Oral Oncol 2013 49 314 21 23265945 56 Yanik EL Katki HA Silverberg MJ Manos MM Engels EA Chaturvedi AK Leukoplakia, Oral Cavity Cancer Risk, and Cancer Survival in the U.S. Elderly Cancer Prev Res (Phila) 2015 8 857 63 26159805 57 Ben Ari Z Weitzman E Safran M Oncogenic viruses and hepatocellular carcinoma Clinics in liver disease 2015 19 341 60 25921667 58 Sherman M Bruix J Porayko M Tran T Screening for hepatocellular carcinoma: the rationale for the American Association for the Study of Liver Diseases recommendations Hepatology 2012 56 793 6 22689409 59 Lederle FA Pocha C Screening for liver cancer: the rush to judgment Ann Intern Med 2012 156 387 9 22393134 60 Aghoram R Cai P Dickinson JA Alpha-foetoprotein and/or liver ultrasonography for screening of hepatocellular carcinoma in patients with chronic hepatitis B The Cochrane database of systematic reviews 2012 9 Cd002799 61 Shiels MS Pfeiffer RM Gail MH Cancer burden in the HIV-infected population in the United States J Natl Cancer Inst 2011 103 753 62 21483021 62 El-Serag HB Kramer JR Chen GJ Duan Z Richardson PA Davila JA Effectiveness of AFP and ultrasound tests on hepatocellular carcinoma mortality in HCV-infected patients in the USA Gut 2011 60 992 7 21257990 63 Goedert JJ Eyster ME Lederman MM End-stage liver disease in persons with hemophilia and transfusion-associated infections Blood 2002 100 1584 9 12176875 64 Lambert AA Merlo CA Kirk GD Human immunodeficiency virus-associated lung malignancies Clinics in chest medicine 2013 34 255 72 23702175 65 Winstone TA Man SF Hull M Montaner JS Sin DD Epidemic of lung cancer in patients with HIV infection Chest 2013 143 305 14 23381313 66 Engels EA Brock MV Chen J Hooker CM Gillison M Moore RD Elevated incidence of lung cancer among HIV-infected individuals J Clin Oncol 2006 24 1383 8 16549832 67 Shiels MS Cole SR Kirk GD Poole C A meta-analysis of the incidence of non-AIDS cancers in HIV-infected individuals J Acquir Immune Defic Syndr 2009 52 611 22 19770804 68 Kirk GD Merlo C OD P HIV infection is associated with an increased risk for lung cancer, independent of smoking Clin Infect Dis 2007 45 103 10 17554710 69 Aberle DR Adams AM Berg CD Reduced lung-cancer mortality with low-dose computed tomographic screening N Engl J Med 2011 365 395 409 21714641 70 Church TR Black WC Aberle DR Results of initial low-dose computed tomographic screening for lung cancer N Engl J Med 2013 368 1980 91 23697514 71 Goedert JJ HIV-associated lung cancer: ambiguities and challenges Aids 2012 26 1031 3 22552477 72 Makinson A Eymard-Duvernay S Raffi F Feasibility and efficacy of early lung cancer diagnosis with chest computed tomography in HIV-infected smokers Aids 2016 30 573 82 26829006 73 Zauber AG Winawer SJ O'Brien MJ Colonoscopic polypectomy and long-term prevention of colorectal-cancer deaths N Engl J Med 2012 366 687 96 22356322 74 Momplaisir F Mounzer K Long JA Preventive cancer screening practices in HIV-positive patients AIDS care 2014 26 87 94 23742681 75 Keller SC Momplaisir F Lo Re V Colorectal cancer incidence and screening in US Medicaid patients with and without HIV infection AIDS care 2014 26 716 22 24188387 76 Massat NJ Dibden A Parmar D Cuzick J Sasieni PD Duffy SW Impact of Screening on Breast Cancer Mortality: The UK Program 20 Years On Cancer Epidemiol Biomarkers Prev 2016 25 455 62 26646362 77 Screening for breast cancer: U.S. Preventive Services Task Force recommendation statement Ann Intern Med 2009 151 716 26 w-236 19920272 78 Schroder FH Hugosson J Roobol MJ Screening and prostate-cancer mortality in a randomized European study N Engl J Med 2009 360 1320 8 19297566 79 Andriole GL Crawford ED Grubb RL 3rd Mortality results from a randomized prostate-cancer screening trial N Engl J Med 2009 360 1310 9 19297565 80 Coghill AE Shiels MS Suneja G Engels EA Elevated Cancer-Specific Mortality Among HIV-Infected Patients in the United States J Clin Oncol 2015 33 2376 83 26077242 81 Shiels MS Copeland G Goodman MT Cancer stage at diagnosis in patients infected with the human immunodeficiency virus and transplant recipients Cancer 2015 121 2063 71 25739496 82 Zhao H Shu G Wang S The risk of non-melanoma skin cancer in HIV-infected patients: new data and meta-analysis International journal of STD & AIDS 2015 83 Shiels MS Engels EA Increased risk of histologically defined cancer subtypes in human immunodeficiency virus-infected individuals: clues for possible immunosuppression-related or infectious etiology Cancer 2012 118 4869 76 22359254 84 Vandenhende MA Roussillon C Henard S Cancer-Related Causes of Death among HIV-Infected Patients in France in 2010: Evolution since 2000 PLoS One 2015 10 e0129550 26083524 85 Liss DT Baker DW Understanding current racial/ethnic disparities in colorectal cancer screening in the United States: the contribution of socioeconomic status and access to care American journal of preventive medicine 2014 46 228 36 24512861 86 Beer L Mattson CL Bradley H Skarbinski J Understanding Cross-Sectional Racial, Ethnic, and Gender Disparities in Antiretroviral Use and Viral Suppression Among HIV Patients in the United States Medicine 2016 95 e3171 27043679
PMC005xxxxxx/PMC5120762.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101470926 34502 J Comput Graph Stat J Comput Graph Stat Journal of computational and graphical statistics : a joint publication of American Statistical Association, Institute of Mathematical Statistics, Interface Foundation of North America 1061-8600 1537-2715 27891045 5120762 10.1080/10618600.2015.1043010 NIHMS729835 Article Reinforced Angle-based Multicategory Support Vector Machines Zhang Chong 1 Liu Yufeng 234* Wang Junhui 5 Zhu Hongtu 4 1 Department of Statistics and Actuarial Science, University of Waterloo 2 Department of Statistics and Operations Research, University of North Carolina at Chapel Hill 3 Department of Genetics, University of North Carolina at Chapel Hill 4 Department of Biostatistics, University of North Carolina at Chapel Hill 5 Department of Mathematics, City University of Hong Kong * [email protected] 21 10 2015 5 8 2016 2016 23 11 2016 25 3 806825 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. The Support Vector Machine (SVM) is a very popular classification tool with many successful applications. It was originally designed for binary problems with desirable theoretical properties. Although there exist various Multicategory SVM (MSVM) extensions in the literature, some challenges remain. In particular, most existing MSVMs make use of k classification functions for a k-class problem, and the corresponding optimization problems are typically handled by existing quadratic programming solvers. In this paper, we propose a new group of MSVMs, namely the Reinforced Angle-based MSVMs (RAMSVMs), using an angle-based prediction rule with k − 1 functions directly. We prove that RAMSVMs can enjoy Fisher consistency. Moreover, we show that the RAMSVM can be implemented using the very efficient coordinate descent algorithm on its dual problem. Numerical experiments demonstrate that our method is highly competitive in terms of computational speed, as well as classification prediction performance. Supplemental materials for the article are available online. Coordinate Descent Algorithm Fisher Consistency Multicategory Classification Quadratic Programming Reproducing Kernel Hilbert Space 1 Introduction Classification is a standard supervised learning technique to handle problems with categorical response variables. Among various existing classifiers, the Support Vector Machine (SVM, Boser et al., 1992; Cortes and Vapnik, 1995) is a popular method originated from the machine learning community. It is a typical example of large-margin classifiers, that use a single classification function for prediction in binary problems. In particular, the binary SVM searches for a hyperplane in the feature space that maximally separates the two classes. It has been shown to be very useful and has achieved excellent performance on many applications in various disciplines. The corresponding theoretical properties, such as Fisher consistency and asymptotic convergence rates, are also well established. In particular, a classifier being Fisher consistent means that it can achieve the optimal prediction performance asymptotically, if the underlying functional space is rich enough. More details about Fisher consistency will be provided in Section 3. Cristianini and Shawe-Taylor (2000) and Hastie et al. (2009), among others, provide comprehensive reviews for existing classification methods. In practice, it is prevalent to have more than two classes in the data. Despite the success on binary classification, it remains challenging to adapt SVMs to multicategory classification problems. To handle a multicategory problem with k classes using SVMs, there are two common approaches in the literature. The first approach is to train a sequence of binary SVMs and combine the results for multicategory classification. Examples include one-versus-one and one-versus-rest methods (Hastie and Tibshirani, 1998; Allwein et al., 2001). Although this approach is simple in concept and implementation, it has certain drawbacks. For example, the one-versus-one method can have a tie-in-vote problem, and the one-versus-rest method can suffer from inconsistency when there is no dominating class (Lee et al., 2004; Liu, 2007). The second approach is to consider all k classes in one optimization problem simultaneously. The common method is to use k classification functions to represent the k classes in the corresponding optimization, with a prediction rule based on which classification function is the maximum. Many existing simultaneous Multicategory SVM (MSVM) classifiers have been proposed in this framework. See, for example, Vapnik (1998), Weston and Watkins (1999), Crammer and Singer (2001), Lee et al. (2004), Liu and Shen (2006), Liu and Yuan (2011), and Guermeur and Monfrini (2011). In this paper, we focus our discussion on simultaneous MSVM methods. Among these MSVMs, Vapnik (1998), Weston and Watkins (1999), Crammer and Singer (2001) are not always Fisher consistent, whereas the MSVM proposed by Lee et al. (2004) is. Recently, Liu and Yuan (2011) proposed a new family of Fisher consistent MSVMs, namely, the Reinforced MSVMs (RMSVMs), which includes the method in Lee et al. (2004) as a special case. In particular, RMSVM uses a convex combination of the loss in Lee et al. (2004) and the naive SVM hinge loss to form a new group of Fisher consistent hinge loss functions for multicategory problems. To avoid confusion, we would like to point out that the reinforced MSVM in this paper refers to the MSVM using a convex combination of loss functions as in Liu and Yuan (2011), and is different from the reinforcement learning in the machine learning literature (Kaelbling et al., 1996; Barto, 1998). For binary SVMs, one uses a single function for classification. Analogously, for a k-class multicategory problem, it should suffice to use k − 1 classification functions. Therefore, using k classification functions in the regular MSVMs can be redundant. To circumvent this difficulty, the existing MSVMs use different optimization formulations, which can be grouped into two main categories. Classifiers in the first group impose an explicit sum-to-zero constraint on the k classification functions, in order to reduce the function space and to ensure some theoretical properties such as Fisher consistency. Examples in this group include the MSVMs in Lee et al. (2004), Liu and Yuan (2011), and Guermeur and Monfrini (2011). For the second group, the corresponding optimization problem is based on pairwise differences among the k classification functions. One can verify that without an explicit sum-to-zero constraint, the obtained classification functions sum to zero automatically with an appropriate regularization term (Guermeur, 2012; Zhang and Liu, 2013). This can be regarded as an alternative way to reduce k functions to k − 1 by an implicit sum-to-zero property. The MSVM methods in Vapnik (1998), Weston and Watkins (1999), Crammer and Singer (2001), and Liu and Shen (2006) are examples in this group. Despite differences in the optimization formulation, the corresponding dual problems of all the aforementioned MSVMs involve linear equality constraints, and are typically solved via existing Quadratic Programming (QP) solvers. See more discussion in Section 4.4. To facilitate comparison and implementation, Guermeur (2012) proposed a unified family of MSVMs, which includes the methods in Vapnik (1998), Weston and Watkins (1999), Crammer and Singer (2001), Lee et al. (2004), and Guermeur and Monfrini (2011) as special cases. Guermeur (2012) then studied the corresponding optimization problems of these MSVMs under this unified framework, and proposed to solve the dual problems using QP solvers. A corresponding powerful package “MSVMpack” was provided by aLauer and Guermeur (2011). For Liu and Yuan (2011), the authors also proposed to solve the dual problem by QP solvers. See Section 4.4 for more details. Recently, Zhang and Liu (2014) proposed the multicategory angle-based large-margin classification framework, which uses only k − 1 classification functions, and implicitly transfers the sum-to-zero constraint onto the newly defined functional margins. As a result, the computational speed of the angle-based classifiers can be faster than the regular methods with the explicit or implicit sum-to-zero constraint. In particular, the angle-based classifiers consider a centered simplex with k vertices in a k − 1 dimensional space. Each vertex represents one class. The classification function naturally defines k angles with respect to the k vertex vectors, and the corresponding prediction rule is based on which angle is the smallest. Details of the least-angle prediction rule can be found in Section 2. Zhang and Liu (2014) introduced a new set of functional margins in the angle-based framework, and showed that the new functional margins sum to zero without an explicit constraint. Consequently, the angle-based classifiers can be more efficient than the regular multicategory classification methods. Although the angle-based classification does not require the sum-to-zero constraint, the direct generalization of SVM in the angle-based classification structure is not Fisher consistent (Zhang and Liu, 2014). Therefore, the naive angle-based MSVM can be asymptotically suboptimal, and it is desirable to develop an MSVM classifier in the angle-based framework that enjoys Fisher consistency. To this end, we propose the Reinforced Angle-based Multicategory Support Vector Machine (RAMSVM). The loss function we employ for RAMSVM is a convex combination of two MSVM losses, and in Section 5 we show that such combination can lead to a stable classifier whose performance is often close to the optimum. In particular, we show through numerical examples that our proposed RAMSVM tends to have stable and competitive performance for many different cases. Our contribution in this paper is two fold. First, we modify existing MSVM losses in the angle-based classification framework, and introduce our RAMSVM loss family as a convex combination. We show that with a proper choice of the convex combination parameter, the new RAMSVM classifier enjoys Fisher consistency. Second, we show that the corresponding optimization can be reduced to minimizing a quadratic objective function with box constraints only. This can then be solved using the very efficient coordinate descent method (Fan et al., 2008; Friedman et al., 2010). We show in Section 5 that our new RAMSVM can enjoy a very fast computational speed. Moreover, RAMSVM is also highly competitive in terms of classification accuracy. The rest of this article is organized as follows. In Section 2, we briefly review some existing MSVM methods, and introduce our RAMSVM classifier. In Section 3, we study Fisher consistency of the RAMSVM family. In Section 4, we develop the coordinate descent algorithm for solving the RAMSVM optimization problem. We also compare the dual problem of RAMSVM with those of the existing MSVM approaches. Numerical studies with simulated and real data sets are presented in Section 5. Some discussions are provided in Section 6. All proofs are collected in the appendix. 2 Methodology For a multicategory classification problem with k classes, let (x1, y1), … , (xn, yn) be the observed training data points. Here xi denotes a p-dimensional predictor vector, and yi ∈ {1, …, k} is the corresponding label. The regular simultaneous multicategory large-margin classifiers use a k-dimensional classification function f(x)=(f1(x),⋯,fk(x))T, and the prediction rule is y^(x)=argmaxj∈{1,⋯,k}fj(x). The corresponding optimization can typically be written as (1) minf∈F1n=∑i=1nV(f(xi),yi)+12λJ(f), where F denotes the function class, J (·) is the penalty term that controls the complexity of F to prevent overfitting, and λ is a tuning parameter that balances the loss and penalty terms. Here V (f (x), y) measures the loss of using f (x) as the classification function for (x, y). Different loss functions correspond to different classifiers. In the simultaneous MSVM literature, the following loss functions are commonly used as extensions from binary SVMs to MSVMs: MSVM1 (Naive hinge loss) V (f (x), y) = [1 − fy (x)]+; MSVM2 (Vapnik, 1998; Weston and Watkins, 1999) V(f(x),y)=[1−fy(x)]+; MSVM3 (Crammer and Singer, 2001; Liu and Shen, 2006) V(f(x),y)=∑j≠y[1−(fy(x)−fj(x))]+; MSVM4 (Lee et al., 2004) V(f(x),y)=∑j≠y[1−minj(fy(x)−fj(x))]+; MSVM5 (Liu and Yuan, 2011) V(f(x),y)=∑j≠y[1+fj(x)]+, where [u]+ = max(u, 0) and γ ∈ [0, 1] in MSVM5 is the convex combination parameter. As discussed in Section 1, MSVMs 1, 4 and 5 employ an explicit sum-to-zero constraint ∑j=1kfj=0. Besides MSVMs 2-5, the MSVM method in Guermeur and Monfrini (2011) (MSVM6) cannot be formulated in the framework of (1). In particular, the primal optimization problem of MSVM6 can be written as minf,ξ(∣∣Mξ∣∣22+λ∣∣PH(f)∣∣H2), (2) s.t.{−fj(xi)≥1k−1−ξij(i=1,⋯,n,j≠yi),∑j=1kfj=0.} Here M is a matrix of rank (k − 1)n, ξ is a vector of length (k − 1)n with its (in + j)th element ξij, the slack variable corresponding to the ith subject and jth class with j /= yi, and ∣∣PH(f)∣∣H2 is the squared norm of PH(f) in a reproducing kernel Hilbert space H. Notice that each element of f involves an intercept, and PH(f ) represents the projection of f onto the kernel space H. See Guermeur and Monfrini (2011) for details. Notice that Guermeur (2012) proposed a generic model of MSVMs, which cover MSVMs 2-4 and 6 as special cases. We discuss the corresponding dual problems of MSVMs 2-6 in Secion 4.4 and the appendix. Although binary SVM is known to be Fisher consistent, MSVMs 1-3 are not. To overcome this challenge, Lee et al. (2004) proposed MSVM4, which can be shown to be Fisher consistent. Furthermore, Liu and Yuan (2011) proposed the Reinforced MSVM (RMSVM, MSVM5) method, which uses a convex combination of the naive hinge loss and the loss in Lee et al. (2004) as a new class of loss functions. Note that with γ = 0, RMSVM includes the MSVM of Lee et al. (2004) as a special case. Liu and Yuan (2011) showed that RMSVM is Fisher consistent when γ ∈ [0, 0.5]. As mentioned in Section 1, it can be inefficient to train a multicategory classifier with k classification functions, which are reduced to k − 1 by explicit or implicit sum-to-zero constraints. To overcome this difficulty, Zhang and Liu (2014) proposed the multicategory angle-based large-margin classification technique. The details of the angle-based classification are as follows. Consider a centered simplex with k vertices in a (k − 1)-dimensional space. Let W be a collection of vectors {Wj ; j = 1, … , k}, where Wj={(k−1)−1∕21ifj=1,−1+k(k−1)3∕21+kk−1ej−1,if2≤j≤k.} Here ej ∈ ℝk−1 is a vector of 0’s except its jth element is 1, and 1 ∈ ℝk−1 is a vector of 1. It can be verified that each vector Wj has Euclidean norm 1, and the angles between any pair (Wi, Wj ) are equal. Consequently, the vectors in W form a k-vertex simplex. For an observation x, we map it into ℝk−1 by the classification function vector f(x)=(f1(x),⋯,fk−1(x))∈ℝk−1. Note that f (x) defines k angles with respect to Wj (j = 1, …, k), namely, ∠(f (x), Wj) (j = 1, …, k). The label prediction for x is then based on which angle is the smallest. In other words, y^(x)=argminj∈{1,⋯,k}∠(f(x),Wj). One can verify that the least angle prediction rule is equivalent to y^(x)=argmaxj∈{1,⋯,k}(〈Wj,f(x)〉), where 〈⋅,⋅〉 denotes the dot product. With this prediction rule, Zhang and Liu (2014) proposed the following optimization for the angle-based classification (3) minf∈F1n∑i=1nl(〈f(xi),Wyi〉)+λ2J(f), where £(·) is any binary large-margin classification loss function. Here the dot products 〈f (x), Wj) (j = 1, … , k〉 can be regarded as new functional margins of (x, y). Note that Lange and Wu (2008), Wu and Lange (2010) and Wu and Wu (2012) also used the simplex structure for multicategory classification with a different classification rule and the E-insensitive loss, and Hill and Doucet (2007) and Saberian and Vasconcelos (2011) studied MSVM and multicategory boosting in the simplex based structure as well. The angle-based classification framework uses k − 1 classification functions directly, and it can be computed more efficiently. However, when using the regular hinge loss £(u) = [1 − u]+ in (3), the corresponding angle-based SVM is not Fisher consistent. Therefore, it is desirable to have a generalization of the binary SVM classifier in the angle-based classification framework which enjoys Fisher consistency. Motivated by the convex combination idea in Liu and Yuan (2011), we propose the following Reinforced Angle-based Multicategory SVM (RAMSVM) classifier with the following optimization (4) minf∈F1n∑i=1n{(1−γ)∑j≠yi[1+〈f(xi),Wj〉]++γ[(k−1)−〈f(xi),Wyi〉]+}+λ2J(f), where γ ∈ [0, 1] is the convex combination parameter. Note that the first part in the loss term of (4) can be regarded as a modified MSVM loss of Lee et al. (2004) in the angle-based classification framework. For many other existing MSVMs, we can generalize them into the angle-based classification framework accordingly. For example, the naive angle-based MSVM method studied in Zhang and Liu (2014) can be regarded as an extension of MSVM1 in the angle-based classification framework. The motivation of using such a combination of loss functions in (4) is based on the prediction rule of the angle-based classification. Since the least angle prediction rule y^(x)=argminj∈{1,⋯,k}∠(f(x),Wj) is equivalent to y^(x)=argmaxj∈{1,⋯,k}(〈Wj,f(x)〉), we need to have (f (x), Wy ) to be the maximum among all the k dot products (f (x), Wj) (j = 1, … , k). To that end, the second part in the loss term of (4) encourages (f (x), Wy) to be large. On the other hand, observe that ∑j=1k〈f(x),Wj〉=0 for all x, which means that the angle-based classification framework transfers the explicit sum-to-zero constraint onto the dot products. Hence, as the first part in the loss term of (4) encourages 〈f(x),Wj〉(j≠y) to be small, it implicitly encourages 〈f(x),Wj〉 to be large. As we will see in Section 5, encouraging 〈f(x),Wj〉 to be large explicitly and implicitly has their advantages respectively, and combining the two loss terms yields a classifier that is consistent, stable and highly competitive in accuracy. Since most existing MSVMs can be cast into QP problems with linear constraints, they are typically implemented using existing QP solvers. For RAMSVM, we show in Section 4 that its dual problem is a QP with box constraints only, hence can be solved using the very efficient coordinate descent method. Compared to existing MSVM implementations using QP solvers, our RAMSVM can often enjoy a faster computational speed. We demonstrate the computational advantage of RAMSVM using numerical examples in Section 5. In the next section, we introduce the details of Fisher consistency, and show that RAMSVM is Fisher consistent when γ ∈ [0, 1/2]. 3 Fisher Consistency For a classification method, Fisher consistency implies that the classifier can achieve the best classification performance asymptotically, when the underlying functional space is rich enough. In other words, under some conditions, the classifier can yield the Bayes classification rule with infinitely many training data points. In the literature, Lin (2004) explored the Fisher consistency of binary margin based classifiers, and a systematic study on Fisher consistency of regular multicategory large margin classification methods using k functions was provided by Zhang et al. (2014). To introduce Fisher consistency, we need some extra notation. Let Pj (x) = Pr(Y = j\|X = x) (j = 1, …, k) be the class conditional probabilities. For a given x, one can verify that y∗(x)=argmaxj∈{1,⋯,k}Pj(x) attains the smallest expected classification error rate (5) Pr(y∗(x)≠Y∣x) which can be regarded as a 0 − 1 loss function E(I(y∗≠Y)∣x). Here one often refers to y∗(x)=argmaxj∈{1,⋯,k}Pj(x) as the Bayes classifier. However, in practice, because the indicator function I(·) is discontinuous, the empirical minimizer of (5) is hard to find. To overcome this difficulty, in the large-margin classification literature, one can employ different surrogate loss functions, which correspond to different classification methods. Fisher consistency requires that prediction based on the conditional minimizer of a surrogate loss function is identical to y∗(x)=argmaxj∈{1,⋯,k}Pj(x). In particular, for angle-based classifiers, Fisher consistency requires that argmaxj∈{1,⋯,k}〈f∗(x),Wj〉=argmaxj∈{1,⋯,k}Pj(x), and when y∗(x)=argmaxj∈{1,⋯,k}Pj(x) is unique, so is argmaxj∈{1,⋯,k}〈f∗(x),Wj〉. Here f ∗(x) is the minimizer of the conditional loss E[V (f (X), Y)\|X = x]. In other words, Fisher consistency ensures that, if one uses f ∗ as the classification function, then the predicted class has the largest class conditional probability, hence the corresponding error rate is minimized. For the RAMSVM classifier (4), we study its Fisher consistency in the following theorem. Theorem 1 For multicategory classification problems with k > 2, the RAMSVM loss function (4) is Fisher consistent when γ ∈ [0, 0.5], and is not Fisher consistent when γ ∈ (0.5, 1]. From Theorem 1, we can conclude that the proposed RAMSVM provides a large family of consistent MSVM classifiers with γ ∈ [0, 1/2]. For γ > 1/2, the Fisher consistency cannot be guaranteed. In Section 5, we study the effect of different γ on the classification accuracy. Interestingly, we observed that the Fisher consistent RAMSVM with γ = 1/2 can provide a stable classifier with competitive classification accuracy. In particular, the numerical results show that RAMSVM with γ = 0 or γ = 1 can be suboptimal in certain cases, whereas γ = 1/2 gives a stable classifier whose performance is close to optimal in many situations. This demonstrates the advantage of the convex combination of MSVM loss functions in the angle-based classification structure. In the MSVM literature, despite the fact that MSVMs 1-3 have shown to deliver reasonable performance for many problems with small or moderate sample sizes, they are not always Fisher consistent. An inconsistent classifier can yield suboptimal prediction results for certain problems. In contrast, the proposed RAMSVM is stable and competitive for finite sample applications, efficient for computation, and Fisher consistent. In the next section, we develop an efficient algorithm to solve (4). In particular, we show that RAMSVM can be solved using the coordinate descent method, and within each step of the coordinate-wise optimization procedure, the update value can be calculated explicitly. This greatly boosts the computational speed. 4 Algorithm In this section, we show how to solve the optimization problem (4). We demonstrate that with the intercepts penalized in linear learning (Fan et al., 2008) or kernel learning, RAMSVM can be solved using the coordinate descent method (Friedman et al., 2010) and consequently enjoys a fast computational speed. First, we focus on linear learning with the commonly used L2 penalty. Then we develop the algorithm for kernel learning and weighted learning problems. Lastly, we discuss the difference of the dual problems among RAMSVM and the existing MSVM methods. 4.1 Linear Learning For linear learning, we assume fq (x) = xTβq (q = 1, … , k−1), where βq (q = 1, …, k−1) are the parameters of interest. The penalty term J (f ) in (4) can be written as J=∑q=1k−1βqTβq. Note that we include the intercepts in x to simplify notation. As a result, the obtained function margins would be slightly different from those margins obtained without regularization on the intercepts. In particular, consider the original space X and the augmented space X′={(1,xT)T:x∈X}. The original binary SVMs aim to maximize the smallest margins within X. On the other hand, if the intercept is penalized, one can verify that it is equivalent to maximizing the smallest margins within the augmented space X′. However, the difference between these two types of margins is often negligible in binary problems (Fan et al., 2008). For MSVMs, the two types of function margins (i.e., with or without penalty on the intercepts) also differ slightly. Our numerical experience suggests that including the intercepts in the penalty term can still yield good classification performance. We solve (4) in its dual form. Using new slack variables ξi,j (i = 1, … , n, j = 1, … , k), (4) with linear learning can be written as minβq,ξi,jnλ2∑q=1k−1βqTβq+∑i=1n[γξi,yi+(1−γ)∑j≠yiξi,j],s.t.ξi,j≥0(i=1),⋯,n,j=1,⋯,k),ξi,yi+〈f(xi),Wyi〉−(k−1)≥0(i=1,⋯,n),ξi,j+〈f(xi),Wj〉−1≥(i=1,⋯,n,j≠yi). Define the corresponding Lagrangian function L as L=nλ2∑q=1k−1βqTβq+∑i=1n[γξi,yi]+(1−γ)∑i≠yiξi,j−∑i=1nαi,yi[ξi,yi+〈f(xi),Wyi〉−(k−1)]−∑i=1n∑j≠yiαi,j[ξi,j−〈f(xi),Wj〉−1], where αi,j and τi,j (i = 1, … , n, j = 1, … , k) are the Lagrangian multipliers. One can verify that L can be rewritten as L=nλ2∑q=1k−1βqTβq+∑i=1n∑j=1k[Ai,j−τi,j−αi,j]ξi,j+∑i=1nαi,yi(k−1)+∑i=1n∑j≠yiαi,j−∑i=1nαi,yi〈f(xi),Wyi〉+∑i=1n∑j≠yiαi,j〈f(xi),Wj〉, where Ai,j = [γI(j = yi) + (1 − γ)I(j /= yi)]. After taking partial derivative of L with respect to βq (q = 1, … , k − 1) and ξi,j (i = 1, … , n, j = 1, … , k), we have ∂L∂ξi,j=Ai,j−τi,j−αi,j=0(i=1,⋯,n,j=1,⋯,k), ∂L∂βq=nλβq−∑i=1nαi,yiWyi,qxi+∑i=1n∑j≠yiαi,jWj,qxi=0(q=1,⋯,k−1), where Wj,q represents the qth element of Wj . Hence (6) βq=1nλ[∑i=1nαi,yiWyi,qxi−∑i=1n∑j≠yiαi,jWj,qxi]. Plug (6) in L, and one can obtain that, after simplification, (7) L=−nλ2∑q=1k−1βqTβq+∑i=1nαi,yi(k−1)+∑i=1n∑j≠yiαi,j, where βq is given in (6). Because maximizing L with respect to αi,j is equivalent to minimizing the negative of L, the dual form of (4) can be expressed as (8) minαi,j(i=1,⋯,n,j=1,⋯,k)12nλ∑q=1k−1[∑i=1nαi,yiWyi,qxi−∑i=1n∑j≠yiαi,jWj,qxi]T[∑i=1nαi,yiWyi,qxi−∑i=1n∑j≠yiαi,jWj,qxi]−∑i=1nαi,yi(k−1)−∑i=1n∑j≠yiαi,j,s.t.0≤αi,j≤Ai,j(i=1,⋯n,j=1,⋯,k). In (7), one can verify that L is strictly concave with respect to βq . Note that βq ’s are linear functions of αi,j ’s, hence L is a quadratic term of {αi,j (i = 1, … , n, j = 1, … , k)}, and −L in (8) is strictly convex with respect to each αi,j . Moreover, the constraints in (8) are box constraints. Therefore, the dual optimization (8) and can be solved by the well known coordinate descent method. Furthermore, because the object function is quadratic, within each step of coordinate-wise optimization the next update value can be calculated explicitly. This greatly boosts the computational speed. Compared to the regular MSVMs that involve linear equality constraints in the QP step (Lee et al., 2004; Liu and Yuan, 2011; Lauer and Guermeur, 2011), our proposed RAMSVM has box constraints only, hence can enjoy a faster computational speed. In Section 5, we show that RAMSVM often outperforms the MSVMs 2-6 in terms of computational speed. We point out that for RAMSVM using linear learning, the number of dual variables is O(nk). Hence, the implementation can be very fast for problems with high dimensional predictors and low sample sizes. In the optimization literature, many authors have considered the general problem of quadratic programming with box constraints only (see, for example, Floudas and Gounaris, 2009, for a review). Moreover, there exist many packages that are aimed for such optimization problems (for example, Bochkanov and Bystritsky, 2013). For our RAMSVM, we provide an R package “ramsvm”, which is developed with focus on RAMSVM optimization. In Section 5, we show that our package can enjoy a fast computational speed. 4.2 Kernel Learning Next, we briefly discuss the case with kernel learning. We show that with the regular squared norm regularization and the intercepts penalized, the optimization can also be solved using the coordinate descent method. To begin with, denote by K the corresponding kernel function, and by K=(K(xi,xi′))i=1,⋯,n,i′=1,⋯,n the gram matrix. Define Ai,j as in the linear case. If the penalty we choose is the squared norm penalty in the corresponding kernel space (see, for example, Shawe-Taylor and Cristianini, 2004, for details), then by the Representer Theorem (Kimeldorf and Wahba, 1971), we have fq(x)=θq,0+∑i=1nθq,iK(xi,x) and J(f)=∑q=1k−1θqTKθq. Here θ=(θq,1,⋯,θq,n)T is the kernel product coefficient vector, for q = 1, … , k − 1. We introduce the slack variables ξi,j as in the linear case. If we impose penalization on the intercepts θq,0 for q = 1, … , k − 1 as well, (4) is equivalent to (9) minθq,θq,0,ξi,jnλ2∑q=1k−1θqTKθq+nλ2∑q=1k−1θq2,0+∑i=1n[γξi,yi+(1−γ)∑j≠yiξi,j],s.t.ξi,j≥0(i=1,⋯,n,j=1,⋯,k),ξi,yi+〈f(xi),wyi〉−(k−1)≥0(i=1,⋯,n),ξi,j+〈f(xi),wj〉−1≥0(i=1,⋯,n,j≠yi). Next, we introduce the Lagrangian multipliers τi,j and αi,j , take partial derivative with respect to θq , θq,0 and ξi,j and set to zero, as in the linear case. Without loss of generality, assume that the gram matrix K is invertible. We have that (10) θq=1nλK−1[∑i=1nαi,yiWyi,qKi−∑i=1n∑j≠yiαi,jWj,qKi], (11) θq,0=1nλ[∑i=1nαi,yiWyi,q−∑i=1n∑j≠yiαi,jWj,q], where Ki is the ith column of K. After plugging (10) and (11) in (9), (4) can be shown to be equivalent to (12) minαi,j(i=1,⋯,n,j=1,⋯,k)12nλ∑q=1k−1[∑i=1nαi,yiWyi,qKi−∑i=1n∑j≠yiαi,jWj,qKi]TK−1[∑i=1nαi,yiWyi,qKi−∑i=1n∑j≠yiαi,jWj,qKi]+12nλ∑q=1k−1[∑i=1nαi,yiWyi,q−∑i=1n∑j≠yiαi,jWj,q]2−∑i=1nαi,yi(k−1)−∑i=1n∑j≠yiαi,j,s.t.0≤αi,j≤Ai,j(i=1,⋯,n,j=1,⋯,k). Note that K−1Kj is the jth column of the identity matrix, and KTK−1Kj = K(xi, xj ). Therefore, we do not need to calculate the inverse matrix K−1 in the optimization. One can verify that (12) can be solved in an analogous manner as (8). 4.3 Weighted Learning So far we have focused on the optimization problem with equal weights of loss on different classes. In practice, it is prevalent to have one class whose size is significantly larger than that of another class. For example, in cancer research, the number of patients with a rare cancer can be very limited, while the number of healthy samples is large. In this case, standard classification without adjusting for the unbalanced sample sizes can lead to a suboptimal result. To overcome this difficulty, it has been proposed to use different weights of loss for different classes (Qiao and Liu, 2009). This weighted learning technique can also be applied to practical problems with possible biased sampling. See Zhang et al. (2013) for more discussions. Therefore, it is desirable to study the extension from standard classification to weighted learning. Next, we use linear learning as an example and demonstrate how to solve the corresponding optimization problems with given weights on different observations. Let the weight of loss for the ith observation be wi, and assume that wi > 0 (i = 1, … , n). In weighted learning, the optimization (4) becomes minf∈F1n∑i=1nwi{(1−γ)∑j≠yi[1+〈f(xi,Wj)〉]++γ[(k−1)−〈f(xi),Wyi〉]+}+λ2J(f), which, after some calculation, can be rewritten as (13) minαi,j(i=1,⋯,n,j=1,⋯,k)12nλ∑q=1k−1[∑i=1nαi,yiWyi,qxi−∑i=1n∑j≠yiαi,jWj,qxi]T[∑i=1nαi,yiWyi,qxi−∑i=1n∑j≠yiαi,jWj,qxi]−∑i=1nαi,yi(k−1)−∑i=1n∑j≠yiαi,j,s.t.0≤αi,j≤A‒i,j(i=1,⋯,n,j=1,⋯,k), where A‒i,j=wi[γI(j=yi)+(1−γ)I(j≠yi)]. Note that the difference between (8) and (13) is in the definition of Ai,j and A‒i,j. Because A‒i,j≥0, the optimization (13) can be solved in a similar manner as (8). 4.4 Comparison with Dual Problems in MSVMs 2-6 In this section, we provide a brief comparison between the dual problems of MSVMs 2-6 and our RAMSVM method. We first discuss the similarities among the dual problems of these classifiers, then illustrate the key difference between RAMSVM and other existing MSVMs. As a result, RAMSVM can be solved using the coordinate descent algorithm, while existing MSVMs typically rely on QP solvers. Numerical analysis in Section 5 suggests that the coordinate descent algorithm can be faster than QP solvers for solving the dual problems of MSVMs. In the literature, Guermeur (2012) proposed a general form of primal optimization for MSVM methods, and MSVMs 2-4 and 6 are included as special cases in this framework. See Problem 1 and Table 1 in Guermeur (2012) for details. Guermeur (2012) then derived the corresponding dual problems for soft and hard margin MSVMs. For MSVM6, Liu and Yuan (2011) derived its dual problem in their (3.8). For better illustration, we include the dual problems of Guermeur (2012) and Liu and Yuan (2011) in (A.1), (A.2), and (A.3) in the appendix. The objective functions of these dual problems are all quadratic functions with respect to the dual variables. This is similar to the RAMSVM case, such as (8) and (12). Notice that the number of dual variables αi,j in Guermeur (2012) is nk with αi,yi = 0. Hence the number of effective dual variables is n(k − 1). For our RAMSVM method, the number of effective dual variable is generally nk, which is slightly larger than n(k − 1). This is because in RAMSVM we use the convex combination of loss functions as in (4), hence we introduce k more slack variables. Nevertheless, as we will see in Section 5, the combined loss function can provide stable classification performance under various settings, which is highly desirable. Notice that for γ = 0, our RAMSVM method also uses n(k − 1) effective dual variables, because αi,yi = 0 by the box constraints. For RAMSVM, our numerical experience suggests that the difference between n(k − 1) or nk dual variables is small in terms of computational speed. The key difference between MSVMs 2-6 and RAMSVM is that the former methods have equality constraints in the corresponding dual problems. See (A.1), (A.2), and (A.3) in the appendix. Hence, the coordinate descent algorithm cannot be directly implemented. Instead, existing QP solvers are typically employed to solve the corresponding optimization. Compared to these dual formations, our RAMSVM is free of equality constraints, and can be solved using the coordinate descent algorithm. More importantly, because the objective function is quadratic, each update value in the coordinate-wise minimization can be calculated explicitly, without any further loops such as those in Newton-Raphson methods (Friedman et al., 2010). This helps to alleviate the computational burden, and boosts the speed greatly. In the next section, we demonstrate that solving RAMSVM with coordinate descent algorithm is efficient to compute. 5 Numerical Results In this section, we examine the numerical performance of RAMSVM. As we will see, the algorithms developed in Section 4 can provide an efficient way to solve the corresponding optimization problem. In particular, in Section 5.1, we compare the performance of RAMSVM with MSVMs 2-6 via three simulated examples, and in Section 5.2, we study six real world data sets. For MSVMs 2-4 and 6, we use the MSVMpack package developed by Lauer and Guermeur (2011). For RAMSVM, we apply the same stopping criterion as in the MSVMpack package, which is to stop the iteration when the ratio of the dual objective function is larger than 98% of the primal value. For RMSVM, the R code is publicly available, and we follow the suggestion in Liu and Yuan (2011) to use γ = 0.5. For RAMSVM, we demonstrate the effect of γ on the classification accuracy using simulated examples. We show that RAMSVM can often enjoy a faster computational speed, and the Fisher consistent RAMSVM with γ = 0.5 yields a stable and competitive classifier. All numerical analyses are done with R (R Core Team, 2015) on a computer with an Intel(R) Core(TM) i7-4770 processor at 3.4GHz and 16GB of memory. The core code of the coordinate descent algorithm for RAMSVM is written in C. An R package “ramsvm”, which contains the code to perform RAMSVM classification, is provided in the Supplementary Materials. Notice that both MSVMpack and our ramsvm implement parallel computing to train the classifiers. In particular, the algorithm splits the data into small chunks, and the dual variables for each chunk are updated using parallel threads. Hence, for problems with large data sets, parallel computing can significantly reduce the computational time. However, as Lauer and Guermeur (2011) mentioned, sometimes the algorithm for one chunk of data needs output from previous steps. If the computation in the previous steps is not finished before the next update step, the corresponding thread would be idle for a while, and this can reduce the corresponding efficiency. How to determine the chunk size can be an issue, and is discussed in the ramsvm reference manual. Moreover, for our ramsvm package, we observe that for small data sets (hundreds of observations), using parallel computing in R can indeed be slower. This is because creating threads in R can take some time, and is not efficient for small problems. Hence, in our ramsvm package, we provide the option to use parallel computing or not. For numerical results in the following sections, we report the smaller computational time of RAMSVM and MSVMpack with or without parallel computing. 5.1 Simulated Examples We consider three simulated examples in this section. The first example is constructed such that linear learning suffices, and we apply linear learning as well as the Gaussian kernel learning. For the second example, we design the marginal distribution of X such that linear classification boundaries cannot separate the classes well. Therefore, we use the second order polynomial kernel and the Gaussian kernel for this example. For Gaussian kernel learning, the kernel parameter σ is chosen to be the median of all the pairwise distances between one category and others. The third example is a simulated data set from the UCI machine learning website (Bache and Lichman, 2015). We apply linear and second order polynomial kernel learning for this example. We let the convex combination parameter γ vary in {0, 0.1, 0.2, … , 0.9, 1} and study the effect of γ on the classification accuracy for RAMSVM. To select the best tuning parameter λ, we use a grid search. In particular, we train the classifiers on a training data set. The classifier that has the smallest prediction error rate on an independent tuning data set is then selected, and we record the prediction error rate on a separate testing data set. The size of the testing data is 106 for the first and second examples. We repeat this procedure 50 times and report the average prediction error rate on the testing data set. To compare the computational speed among different methods, in each replication we record the total time of training the classifiers for 50 different tuning parameters, selecting the best parameter λ and predicting on the testing data set. We report the average time used for one replication as a measurement of computational speed. For RAMSVM, we only report the time for γ = 1/2, since the differences in terms of computational time for various γ values are small. Example 1 This is an eight class example with equal probabilities Pr(Y = j) (j = 1, … , 8). The marginal distribution of X for each class is normal, and the mean vectors for different classes form a simplex in ℝ7. We choose the covariance matrices of the normal distributions so that the corresponding Bayes classification error is 5%. We use 150 observations for training and another 150 for tuning. Example 2 We generate four classes with Pr(Y = j) = 1/4 (j = 1, … , 4) on ℝ2. For each class, the predictor vector X \| Y = j follows a mixed normal distribution. In particular, for class j, X follows 0.5N ((cos(jπ/4), sin(jπ/4))T, Σ) + 0.5N ((cos(jπ/4 + π), sin(jπ/4 + π))T, Σ). Here Σ is chosen such that the Bayes error is 5%. Both the training and tuning data are of size 150. Example 3 This is the three-class Waveform Database Generator (Version 1) data set from the UCI machine learning website. The number of observations for the three classes are 1657, 1647 and 1696, respectively. There are 21 predictors. For each replicate, we divide the data into six parts of roughly the same size. Four parts are used as the training data, one as the tuning, and the rest as the testing data set. The effect of γ on classification error rates for RAMSVM is illustrated in Figure 1. For linear learning, the classification error rate decreases as γ increases from 0 to 1/2. When γ ranges in [1/2, 1], the change in classification accuracy is small. However, for kernel learning, the pattern is reversed according to Figure 1. In particular, the classification error rates increase as γ increases from 1/2 to 1, and when γ is in [0, 1/2], the classification accuracy does not change much. The results suggest that although RAMSVM with γ > 1/2 is not Fisher consistent, in linear learning where the functional class F is relatively simple, encouraging 〈f,Wy〉 to be large explicitly (the second part of loss in (4)) may work better than doing so implicitly (the first part of loss). However, for kernel learning problems where F is more flexible, Fisher consistency can become more relevant and the classification accuracy of RAMSVM with γ > 1/2 may suffer from being inconsistent. From Figure 1, one can conclude that the classification accuracy of RAMSVM with γ = 1/2 is close to the optimum for all situations, hence choosing γ = 1/2 provides a Fisher consistent and stable classifier. We recommend using γ = 1/2 for all classification problems. We report the comparison among RAMSVM and MSVMs 2-6 in Table 1. In particular, we calculate the p-values for testing the null (alternative) hypotheses that the classification error of RAMSVM is larger than or equal to (smaller than) that of the other methods, using the two sample proportion test. Based on the p-values, our RAMSVM with γ = 1/2 can work better than all the considered existing MSVMs in Example 1 using linear learning, and in Example 2 using Gaussian learning. For other cases, one can see that RAMSVM is still very competitive. In terms of computational speed, compared to MSVMpack, RAMSVM has computational advantages on linear and polynomial learning, which is also demonstrated in Section 5.2. For Gaussian kernel learning, the speed of RAMSVM is comparable with respect to MSVMpack. 5.2 Real Data Analysis In this section, we test the performance of RAMSVM using six real data sets, among which five data sets can be found on the UCI machine learning repository website. The Glioblastoma data can be found in Verhaak et al. (2010). A summary of these data sets is provided in Table 2. The predictors of these real data sets are standardized. For the Vertebral, Optical, and Glioblastoma data sets, we perform a 5-fold cross validation to select the best tuning parameters. We report the average prediction error rates among 50 replicates. As a measurement of computational speed, we report the average time for solving the optimization problems with the entire training data sets throughout the 50 replicates. The Gaussian kernel parameters are chosen analogously as in Section 5.1. For the Gas, Isolet, and Pendigits data sets, because their sizes are relatively large, we choose a small proportion of observations to select the tuning parameters. Then we report the computational time of optimization with the entire training data set as a measurement of computational speed. Moreover, we report the prediction error rates. Notice that we only perform one replicate for the Gas, Isolet, and Pendigits data sets to assess the corresponding classification accuracy and computational speed, due to the large sample sizes of these problems. The real data results are reported in Table 3. One can verify that RAMSVM can enjoy a very efficient computational speed, especially with linear learning. This is consistent with the findings in the simulation study. For Gaussian kernel learning, the computational speed of RAMSVM and MSVMpack is comparable. In terms of classification accuracy, RAMSVM with γ = 1/2 is a stable classifier, and the corresponding prediction error rates are close to the optimum. Overall, RAMSVM is a very competitive classifier. Notice that the code for MSVM5 does not converge for Gas, Isolet, and Pendigits data sets after 48 hours. 6 Discussion In this paper, we propose the RAMSVM as a new angle-based MSVM method. We show that the RAMSVM has two advantages. First, it is free of the sum-to-zero constraint, hence the corresponding optimization procedure can be more efficient. In particular, we develop a new algorithm to train the classifier using coordinate descent method, which does not rely on existing QP solvers. Second, the RAMSVM overcomes the difficulty of inconsistency, compared to the existing angle-based MSVM method. Numerical comparisons between the RAMSVM and some existing MSVMs demonstrate the usefulness of our method. Although one can treat γ as an additional tuning parameter, it requires more computational time. We recommend using the RAMSVM with γ = 1/2, which yields a Fisher consistent classifier whose performance is stable and often close to the optimum. Supplementary Material SuppFile Acknowledgments The authors would like to thank the Editor, the Associate Editor, and two reviewers for their constructive comments and suggestions. Zhang and Liu were supported in part by NSF grant DMS-1407241, NIH/NCI grant R01 CA-149569, and NIH/NCI P01 CA-142538. Zhu was supported by NSF grants DMS-1407655 and SES-1357666 and NIH grants RR025747-01, MH086633, and EB005149-01. Figure 1 The left panel displays the effect of different γ values on the classification performance of RAMSVM for simulated Example 1. The standard errors of the classification error rates in Example 1 range from 0.001 to 0.012. The middle panel reports the pattern for Example 2, and the corresponding standard errors range from 0.001 to 0.010. The right panel shows the effect of different γ values on classification error rates for Example 3. The corresponding standard errors range from 0.003 to 0.023. Table 1 Prediction error rates and computational time in seconds for the simulated examples. MSVM Method Ex 1 Linear Ex 2 Poly Ex 3 Linear Error Time p-value Error Time p-value Error Time p-value MSVM2 14.67 18 0.000 15.53 54 1.000 14.48 306 0.419 MSVM3 15.21 21 0.000 14.96 48 1.000 14.41 299 0.533 MSVM4 22.47 20 0.000 21.79 64 0.000 14.56 338 0.297 MSVM5 11.34 151 0.000 17.76 452 0.000 - - - MSVM6 14.98 27 0.000 16.14 77 0.000 14.76 422 0.088 RAMSVM 9.80 13 - 15.71 23 - 14.43 115 - MSVM Method Ex 1 Gauss Ex 2 Gauss Ex 3 Poly Error Time p-value Error Time p-value Error Time p-value MSVM2 8.64 13 1.000 11.62 17 0.000 14.64 599 0.109 MSVM3 9.16 11 0.000 12.88 21 0.000 14.91 478 0.010 MSVM4 11.71 15 0.000 15.19 23 0.000 14.88 705 0.013 MSVM5 14.09 277 0.000 15.82 298 0.000 - - - MSVM6 11.57 15 0.000 13.57 30 0.000 14.29 818 0.581 RAMSVM 8.78 14 - 11.34 21 - 14.34 355 - Poly: Second order polynomial kernel learning. Gauss: Gaussian kernel learning. The standard errors of the error rates range from 0.6% to 1.7%. The standard errors of the computational time range from 1 to 32 seconds. Note that MSVM5 cannot be computed for Example 3 due to its large n. Table 2 Summary of the real data sets used in Section 5.2. Data # classes # predictors # training obs # testing obs Gas 6 129 11592 2318 Glioblastoma 4 16548 296 60 Isolet 26 617 6238 1559 Optical 10 64 3823 1797 Pendigits 10 16 7494 3498 Vertebral 3 6 258 52 Table 3 Prediction error rates and computational time for the real examples. Data & Kernel Method Error Time Gas Linear MSVM2 1.57 01:33:24 MSVM3 1.72 32:19:17 MSVM4 1.92 48:00:00* MSVM5 - - MSVM6 1.78 48:00:00* RAMSVM 1.51 00:47:42 Glioblastoma Linear MSVM2 24.21 00:00:02 MSVM3 23.17 00:00:02 MSVM4 24.28 00:00:02 MSVM5 23.80 00:00:21 MSVM6 23.92 00:00:02 RAMSVM 22.95 00:00:02 Isolet Linear MSVM2 9.07 44:26:09 MSVM3 8.93 29:16:35 MSVM4 9.41 48:00:00* MSVM5 - - MSVM6 9.39 48:00:00* RAMSVM 8.92 07:26:04 Optical Gaussian MSVM2 2.63 00:00:15 MSVM3 2.58 00:00:18 MSVM4 2.74 00:00:31 MSVM5 - - MSVM6 2.68 00:01:44 RAMSVM 2.56 00:00:16 Pendigits Gaussian MSVM2 5.13 00:03:19 MSVM3 6.60 00:00:29 MSVM4 4.93 00:01:20 MSVM5 - - MSVM6 5.11 00:05:17 RAMSVM 5.06 00:00:52 Vertebral Gaussian MSVM2 18.26 00:00:01 MSVM3 18.41 00:00:01 MSVM4 18.63 00:00:01 MSVM5 18.20 00:00:12 MSVM6 18.39 00:00:01 RAMSVM 17.92 00:00:01 For Vertebral, Optical, and Glioblastoma, the standard errors of the error rates range from 0.05% to 0.24%, and the standard errors of the computational time range from 0.001 to 2 seconds. Note that MSVM5 cannot be computed for Gas, Isolet, and Pendigits, due to the large n. * Here means the MSVMpack algorithm did not converge according to the stopping rule within 48 hours, and the iteration was manually stopped. The models at 48 hours were used to assess the performance. See Lauer and Guermeur (2011) for more details. Supplementary Materials R-package for RAMSVM: The supplemental files for this article include the R-package “ramsvm”, which contains R code to perform the RAMSVM method described in the article. Please refer to the file “ramsvm.pdf” for detailed information on how to use the package. (ramsvm 1.0.tar.gz) Appendix: The supplemental files for this article include the appendix to the paper, which provides detailed proof of Theorem 1, and a review on the dual problems of MSVM methods in Guermeur (2012) and Liu and Yuan (2011). References Allwein EL Schapire RE Singer Y Reducing Multiclass to Binary: a Unifying Approach for Margin Classifiers Journal of Machine Learning Research 2001 1 113 141 Bache K Lichman M UCI Machine Learning Repository 2015 Barto AG Reinforcement Learning: An Introduction 1998 MIT press Bochkanov S Bystritsky V Alglib 2013 Boser BE Guyon IM Vapnik VN A Training Algorithm for Optimal Margin Classifiers Proceedings of the fifth annual workshop on Computational learning theory 1992 144 152 ACM New York, NY, USA COLT ’92 Cortes C Vapnik V Support Vector Networks Machine Learning 1995 20 273 297 Crammer K Singer Y On the Algorithmic Implementation of Multiclass Kernel-based Vector Machines Journal of Machine Learning Research 2001 2 265 292 Cristianini N Shawe-Taylor J An Introduction to Support Vector Machines 2000 Cambridge University Press Fan R-E Chang K-W Hsieh C-J Wang X-R Lin C-J LIBLINEAR: A Library for Large Linear Classification Journal of Machine Learning Research 2008 9 1871 1874 Floudas CA Gounaris CE A Review of Recent Advances in Global Optimization Journal of Global Optimization 2009 45 1 3 38 Friedman JH Hastie TJ Tibshirani RJ Regularized Paths for Generalized Linear Models via Coordinate Descent Journal of Statistical Software 2010 33 1 Guermeur Y A Generic Model of Multi-Class Support Vector Machine International Journal of Intelligent Information and Database Systems 2012 6 6 555 577 Guermeur Y Monfrini E A Quadratic Loss Multi-Class SVM for which a Radius-Margin Bound Applies Informatica 2011 22 1 73 96 Hastie TJ Tibshirani RJ Classification by Pairwise Coupling Annals of Statistics 1998 26 2 451 471 Hastie TJ Tibshirani RJ Friedman J The Elements of Statistical Learning 2009 Springer Hill SI Doucet A A Framework for Kernel-based Multi-category Classification Journal of Artificial Intelligence Research 2007 30 1 525 564 Kaelbling LP Littman ML Moore AW Reinforcement Learning: A Survey Journal of Artifcial Intelligence Research 1996 4 237 285 Kimeldorf G Wahba G Some Results on Tchebycheffian Spline Functions Journal of Mathematical Analysis and Applications 1971 33 82 95 Lange K Wu T An MM Algorithm for Multicategory Vertex Discriminant Analysis Journal of Computational and Graphical Statistics 2008 17 3 527 544 Lauer F Guermeur Y MSVMpack: A Multi-class Support Vector Machine Package Journal of Machine Learning Research 2011 12 2293 2296 Lee Y Lin Y Wahba G Multicategory Support Vector Machines, Theory, and Application to the Classification of Microarray Data and Satellite Radiance Data Journal of the American Statistical Association 2004 99 67 81 Lin Y A Note on Margin-based Loss Functions in Classification Statistics and Probability Letters 2004 68 73 82 Liu Y Fisher Consistency of Multicategory Support Vector Machines Eleventh International Conference on Artificial Intelligence and Statistics 2007 289 296 Liu Y Shen X Multicategory ψ-learning Journal of the American Statistical Association 2006 101 500 509 Liu Y Yuan M Reinforced Multicategory Support Vector Machines Journal of Computational and Graphical Statistics 2011 20 4 901 919 Qiao X Liu Y Adaptive Weighted Learning for Unbalanced Multicategory Classification Biometrics 2009 65 159 168 18363773 R Core Team R: A Language and Environment for Statistical Computing 2015 R Foundation for Statistical Computing Vienna, Austria Saberian MJ Vasconcelos N Multiclass Boosting: Theory and Algorithms Advances in Neural Information Processing Systems 2011 2124 2132 Shawe-Taylor J Cristianini N Kernel Methods for Pattern Analysis 2004 Cambridge University Press Vapnik V Statistical Learning Theory 1998 Wiley Verhaak RG Hoadley KA Purdom E Wang V Qi Y Wilkerson MD Miller CR Ding L Golub T Mesirov JP Alexe G Lawrence M O’Kelly M Tamayo P Weir BA Gabriel S Winckler W Gupta S Jakkula L Feiler HS Hodgson JG James CD Sarkaria JN Brennan C Kahn A Spellman PT Wilson RK Speed TP Gray JW Meyerson M Getz G Perou CM Hayes DN Cancer Genome Atlas Research Network Integrated Genomic Analysis Identifies Clinically Relevant Subtypes of Glioblastoma Characterized by Abnormalities in PDGFRA, IDH1, EGFR, and NF1 Cancer Cell 2010 17 1 98 110 20129251 Weston J Watkins C Support Vector Machines for Multi-class Pattern Recognition ESANN 1999 99 219 224 Wu T Lange K Multicategory Vertex Discriminant Analysis for High-Dimensional Data Annals of Applied Statistics 2010 4 4 1698 1721 Wu T Wu Y Nonlinear Vertex Discriminant Analysis with Reproducing Kernels Statistical Analysis and Data Mining 2012 5 167 176 23205165 Zhang C Liu Y Multicategory Large-margin Unified Machines Journal of Machine Learning Research 2013 14 1349 1386 24415909 Zhang C Liu Y Multicategory Angle-based Large-margin Classification Biometrika 2014 101 3 625 640 26538663 Zhang C Liu Y Wu Z On the Effect and Remedies of Shrinkage on Classification Probability Estimation The American Statistician 2013 67 3 134 142 Zhang Z Chen C Dai G Li W-J Yeung D-Y Multicategory Large Margin Classification Methods: Hinge Losses vs. Coherence Functions Artificial Intelligence 2014 215 55 78
PMC005xxxxxx/PMC5120857.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101186374 31761 Nat Struct Mol Biol Nat. Struct. Mol. Biol. Nature structural & molecular biology 1545-9993 1545-9985 19234465 5120857 10.1038/nsmb.1568 NIHMS107472 Article PRMT5-mediated methylation of histone H4R3 recruits DNMT3A, coupling histone and DNA methylation in gene silencing Zhao Quan 12 Rank Gerhard 1 Tan Yuen T 1 Li Haitao 3 Moritz Robert L 4 Simpson Richard J 4 Cerruti Loretta 1 Curtis David J 1 Patel Dinshaw J 3 Allis C David 5 Cunningham John M 6 Jane Stephen M 17 1 Rotary Bone Marrow Research Laboratories, Melbourne Health Research Directorate, c/o Royal Melbourne Hospital Post Office, Grattan Street, Parkville, VIC 3050, Australia. 2 Molecular Immunology and Cancer Research Center, The State Key Laboratory of Pharmaceutical Biotechnology, School of Life Sciences, Nanjing University, 22 Hankou Road, Nanjing 210093, China. 3 Structural Biology Program, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, New York 10021, USA. 4 Joint Protein Structure Laboratory, Ludwig Institute for Cancer Research and Walter and Eliza Hall Institute for Medical Research, 1G Royal Parade, Parkville, VIC 3050, Australia. 5 Laboratory of Chromatin Biology, The Rockefeller University, 1230 York Avenue, New York, New York 10021, USA. 6 Department of Pediatrics and Institute of Molecular Pediatric Sciences, University of Chicago, 5839 South Maryland Avenue, Chicago, Illinois 60637, USA. 7 Department of Medicine, University of Melbourne, Grattan Street, Parkville, VIC 3050, Australia. Correspondence should be addressed to S.M.J. ([email protected]) or Q.Z. ([email protected]). 18 2 2015 22 2 2009 3 2009 23 11 2016 16 3 304311 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Mammalian gene silencing is established through methylation of histones and DNA, although the order in which these modifications occur remains contentious. Using the human β-globin locus as a model, we demonstrate that symmetric methylation of histone H4 arginine 3 (H4R3me2s) by the protein arginine methyltransferase PRMT5 is required for subsequent DNA methylation. H4R3me2s serves as a direct binding target for the DNA methyltransferase DNMT3A, which interacts through the ADD domain containing the PHD motif. Loss of the H4R3me2s mark through short hairpin RNA–mediated knockdown of PRMT5 leads to reduced DNMT3A binding, loss of DNA methylation and gene activation. In primary erythroid progenitors from adult bone marrow, H4R3me2s marks the inactive methylated globin genes coincident with localization of PRMT5. Our findings define DNMT3A as both a reader and a writer of repressive epigenetic marks, thereby directly linking histone and DNA methylation in gene silencing. Covalent modification of DNA and histone molecules, the core components of chromatin, provides a heritable mechanism for regulating gene expression1,2. These histone marks function cooperatively to establish distinct repressive or active chromatin states that extend the information potential of the genetic code. Integral to this process are effector molecules, which interpret specific modifications to influence downstream events through recruitment or stabilization of chromatin-template machinery3. Although histone modifications and DNA methylation have also been shown to function cooperatively, in many settings the order in which these epigenetic marks are established remains unclear. In mammals, methylation of DNA is largely confined to position five of the cytosine ring in CpG dinucleotides and is most commonly associated with a repressed chromatin state and inhibition of gene expression4,5. Although some overlap exists6, two general classes of cytosine DNA methyltransferases are known: the de novo methyltransferases, DNMT3A and DNMT3B, which are responsible for modifying unmethylated CpG sites, and the maintenance methyltransferase DNMT1, which copies pre-existing methylation patterns onto the new DNA strand during replication7. The precise sequence of events linking histone modifications and DNA methylation varies in different organisms, and at different gene loci, which suggests that it is context dependent. Evidence that DNA methylation can influence the histone modification pattern has been obtained in several model systems. Transgenes methylated in vitro before stable genomic integration associate with deacetylated histones, whereas an identical but unmethylated integrated transgene is enriched for acetylated his-tones8,9. Methyl-CpG–enriched regions may directly recruit histone methyltransferases10, or are bound by methyl-CpG–binding proteins, which in turn recruit repressor complexes containing histone deacetylases4 and histone methyltransferases11,12. Conversely, studies in fungi, plants and mammals have suggested that trimethylation of H3K9 (ref. 13), H3K27 (ref. 14) and H4K20 (ref. 15) is a prerequisite for subsequent DNA methylation. The functional link between these processes appears to be attributable to a physical association between the histone lysine methylation system and DNA methyltransferases, as has been established in the context of H3K9 and H3K27, where the relevant lysine histone methyltransferases, SUV39h1 and EZH2, have been shown to interact directly with DNMT1 and with DNMT3A and DNMT3B (refs. 16–18). Histone arginine methylation has also been implicated in gene silencing, but so far no links between arginine methyltransferases and DNA methyltransferases have been established. Using the human β-globin locus as a model, we now report evidence of a link between arginine methylation of histones and DNA methylation and gene silencing. Central to this link is the protein arginine methyltransferase PRMT5, which symmetrically dimethylates H4R3. We show that this histone modification serves as a direct target for binding of DNMT3A, thus providing a new mechanism by which a repressive histone modification and DNA methylation are coordinated. RESULTS PRMT5 induces the H4R3me2s mark on the γ-globin gene The β-globin locus has served as a paradigm for analyzing the role of epigenetic modifications in the regulation of tissue and developmentally specific gene expression. In both humans and primates, the genes encoding fetal (γ) globin are progressively silenced after birth, displaying methylation of a cluster of CpG dinucleotides in the proximal promoters and 5’ untranslated regions in adult bone marrow19. To identify proteins involved in silencing of genes encoding γ-globin (γ-genes), we analyzed by mass spectrometry immunoprecipitates from an erythroid cell line (K562) expressing Flag-tagged NF-E4, a proximal γ-promoter–binding transcription factor implicated in both activation and repression of the γ-genes20. We identified a number of proteins specific to the Flag-NF-E4–expressing cells that were not detected in Flag immunoprecipitates from cells transfected with the control vector expressing the Flag epitope alone (Fig. 1a). The protein methyltransferase PRMT5 (ref. 21) was of particular interest, as the highly related protein PRMT1 has been shown to have a key role in activation of the adult β-globin gene22. PRMT5 has been implicated in gene silencing, which raises the possibility that the PRMT factors could have coordinate roles in regulating the fetal-to-adult switch in globin gene subtype. We confirmed the interaction between NF-E4 and PRMT5 by co-immunoprecipitation of endogenous proteins from untransfected cells (Fig. 1b). This interaction was direct, as demonstrated by glutathione S-transferase (GST) pulldown, and required the residues between 49 and 100 in the N-terminal half of NF-E4 (Supplementary Fig. 1a online). The two proteins were localized at the proximal γ-promoter (P-1, which contains the NF-E4 binding site) by chromatin immunoprecipitation (ChIP) using antibodies to the endogenous proteins (Fig. 1c). Binding of PRMT5 and NF-E4 to far upstream regions of the γ-promoter was not detected, but weak binding was detected to the promoter region adjacent to P-1 (P-2), and to regions downstream of the transcriptional start site (TSS) extending to exon 2 (P+1 to P+3) (Supplementary Fig. 1b). Analysis of the other proteins identified by mass spectrometry will be described elsewhere. PRMT5 is a type II arginine methyltransferase that has been linked to gene silencing through establishment of repressive histone marks, including symmetrical dimethylation of H4R3, H2AR3 (refs. 21,23) and H3R8 (ref. 24). To identify the histone substrates of PRMT5 in K562 cells, we derived a cell line expressing Flag-tagged PRMT5 (PRMT5-f) to facilitate immunoprecipitation of the active enzyme. The Flag-immunoprecipitate from this line was subjected to a standard radioactive histone methyltransferase activity assay using free histones, which demonstrated radiolabeling of histone H4 (Fig. 1d, free histones). To ensure that methylation of this substrate was specific for PRMT5 methyltransferase activity, we also derived a line containing a Flag-tagged mutant form of PRMT5 in which five amino acids in the S-adenosyl-l-methionine binding motif had been deleted (PRMT5Δ-f)25. Methylation of histone H4 induced with the Flag-immunoprecipitate from this line was markedly reduced, and the weak signal we observed presumably reflects a small amount of endogenous PRMT5 dimerized to PRMT5Δ-f (ref. 25) (Fig. 1d, free histones). Comparable amounts of the tagged proteins were contained in the precipitates used in both assays (Fig. 1d, immunoblot: α-Flag). No methyltransferase activity was observed with the Flag-precipitate from untransfected cells (data not shown). In this setting, methylation of the other known substrates of PRMT5 (histones H2A and H3) was not detected. This substrate specificity was identical when nucleosomes purified from K562 cells were used as the source of histones (Fig. 1d, K562 nucleosomes). To confirm that PRMT5 induced the H4R3me2s modification in this assay, we incubated the 3H-methylated histones with an antibody against H4R3me2s or with normal rabbit IgG before immunoprecipitation. Substantial levels of radioactivity were detected in the precipitate generated with the H4R3me2s antibody from the PRMT5-f–labeled histones, but not from the PRMT5D-f reaction (Fig. 1e). Consistent with this finding, the repressive H4R3me2s mark was enriched at the γ-promoter (and binding extended downstream of the TSS) in cells expressing PRMT5-f but not PRMT5Δ-f, despite both proteins binding robustly to the proximal promoter site (Fig. 1f and Supplementary Fig. 1b). PRMT5 mediates transcriptional silencing of the γ-gene To ascertain whether perturbations to the levels or enzymatic activity of PRMT5 affect γ-gene expression, we performed quantitative real-time PCR (Q-RT-PCR) analyses on cells expressing PRMT5-f and cells expressing PRMT5Δ-f (Fig. 2a). Expression of the two tagged proteins was comparable (immunoblot: α-Flag) and equivalent to the level of endogenous PRMT5 in the vector control cells (immunoblot: α-PRMT5). Increased levels of PRMT5 resulted in a twofold to threefold downregulation of γ-gene expression, whereas enforced expression of PRMT5Δ-f induced a two-fold increase in γ-gene transcripts compared with the vector control (Fig. 2a, left). To examine the effect of reduced PRMT5 expression, we used stably expressed short hairpin RNAs (shRNAs) (PRMT5-kd). Cells transfected with an expression vector containing a scrambled sequence served as the control (Scr). Western blots confirmed that PRMT5 protein levels were reduced by more than 90% in the PRMT5-kd cells compared with the scrambled control, but no effect was observed on the control proteins tubulin, PRMT1 and GATA-1 (Fig. 2b, right). The knockdown of PRMT5 led to a sixfold induction of γ-gene expression compared with the scrambled vector (Fig. 2b, left). This effect was specific, as other markers of erythroid differentiation in K562 cells (GATA-1, glycophorin A and v-myb) remained unchanged in the PRMT5-kd cells compared with the Scr control (Fig. 2b, right; Supplementary Fig. 2 online). Consistent with the γ-gene expression data, the H4R3me2s repressive mark was markedly reduced at the γ-promoter in the PRMT5-kd cells, and RNA polymerase II localization was increased fivefold compared with the Scr control (Fig. 2c). We attempted to determine the role of NF-E4 in PRMT5 regulation of γ-gene expression by generating a knockdown of this factor. Despite the use of several different shRNA constructs, we were unable to modulate the levels of the endogenous protein. PRMT5 and DNMT3A function cooperatively in gene silencing Methylation of four CpG dinucleotides immediately flanking the transcriptional start site is essential for γ-gene silencing, and two of these CpGs reside within the PRMT5 binding region19,26,27. In view of this, we examined whether DNA methylation of the γ-gene was altered in PRMT5-f, PRMT5Δ-f, PRMT5-kd or scrambled control (Scr) cells using bisulfite DNA sequencing (Fig. 3a). Consistent with the observation that globin gene expression is not detectable in a large proportion of uninduced K562 cells, we observed methylation of the four CpG dinucleotides immediately flanking the γ-globin transcriptional start site in 21% of clones derived from scrambled control cells. This frequency was increased in three of the four sites in the PRMT5-f cells, with 35% of the clones showing methylation. In contrast, methylation of all CpG dinucleotides was abolished in clones derived from the PRMT5-kd cells. It was also abolished in all clones derived from the PRMT5Δ-f cells, which suggests that the enzymatic activity of PRMT5, and not just its physical occupation of the γ-globin promoter, is essential for the epigenetic modification of DNA in this setting. No change in the high levels of DNA methylation (95%) at the six CpGs flanking the b-gene transcriptional start site (−415 to +110) was observed in either the PRMT5-kd, PRMT5Δ-f or scrambled control cells (Supplementary Fig. 3a online). To determine the mechanism by which PRMT5 influences DNA methylation, we initially examined whether the protein methyltransferase is associated directly with a DNA methyltransferase. We analyzed Flag-antibody immunoprecipitates from PRMT5-f cells by immunoblot with antibodies to DNMT1, DNMT3A and DNMT3B, and demonstrated that DNMT3A was co-immunoprecipitated with PRMT5 (Fig. 3b). We confirmed this using antibodies to the endogenous proteins with extract from untransfected cells (Fig. 3c). The identification of DNMT3A as a PRMT5-interacting protein suggested that it might be responsible for DNA methylation in the context of PRMT5-induced fetal globin gene silencing. To address this, we examined the binding of DNMT3A to γ-promoter in PRMT5-f and PRMT5Δ-f cells by ChIP (Fig. 3d). Consistent with our methylation studies, binding of DNMT3A to the γ-gene was markedly reduced in the PRMT5Δ-f cells compared with cells expressing PRMT5-f. DNMT3A binding was also markedly reduced in the PRMT5-kd cells compared with the scrambled control cells (Fig. 3e). To assess the role of DNMT3A in γ-gene silencing, we used an shRNA approach to knock down the expression of the protein in K562 cells (DNMT3A-kd). A reduction in DNMT3A levels to 30% of the scrambled control (Fig. 3f, right) induced a fivefold increase in γ-gene expression levels (Fig. 3f, left). This was accompanied by a twofold reduction in CpG methylation at the γ-gene (Fig. 3g), but no change at the β-gene (Supplementary Fig. 3b). DNMT3A binds specifically to histone H4R3me2s The importance of PRMT5 enzymatic activity for the recruitment of DNMT3A to the γ-gene suggests that the methyltransferase domain of PRMT5 is involved in the interaction between the two proteins. To address this, we used a GST pulldown assay to analyze the binding of 35S-radiolabeled in vitro transcribed and translated DNMT3A to GST-PRMT5 and GST-PRMT5Δ (Fig. 4a). Surprisingly, no difference was observed between the wild-type and mutant proteins, which implies that the enzymatic function of PRMT5 may lead to DNMT3A recruitment via an alternate mechanism. One possibility is that the PRMT5-induced H4R3me2s modification could directly recruit DNMT3A. To examine this, we performed a peptide pulldown assay using COOH-terminal biotin-tagged 20-mer N-terminal peptides of histone H4 with the Arg3 residue unmethylated, symmetrically methylated or asymmetrically methylated. We confirmed symmetric methylation by western blot with the H4R3me2s antibody (Fig. 4b, immunoblot: α-H4R3me2s). We incubated equivalent amounts of each peptide (Fig. 4b, Coomassie stain) coupled to streptavidin beads with nuclear extract from K562 cells, washed the beads and analyzed the eluate by immunoblot with an antibody to DNMT3A (Fig. 4b, immunoblot: α-DNMT3A). Binding of DNMT3A was observed with the H4R3me2s peptide, but not with the unmethylated or asymmetrically methylated peptides. The DNMT3A protein contains a PWWP domain implicated in DNA and chromatin binding, an ATRXDNMT3-DNMT3L (ADD) domain that contains a plant homeo-domain (PHD) zinc finger motif that may mediate interactions to other proteins (including histones), and a C-terminal catalytic domain3,28–30. We demonstrated that the interaction between DNMT3A and H4R3me2s was direct and specific using pulldown assays with the three peptides and radiolabeled in vitro transcribed and translated DNMT3A (Fig. 4c). We then mapped the regions of DNMT3A required for binding to the H4R3me2s peptide. No binding was observed with the N-terminal third of the protein (amino acids 1–354). A larger N-terminal fragment (1–587) containing the GATA and PHD domains of ADD, but lacking an adjacent C-terminal helix, bound to levels comparable to those of the full-length protein (Fig. 4c). These findings suggested that binding of DNMT3A to the H4R3me2s peptide was mediated through a region between residues 354 and 587, which incorporates most of the ADD domain. We therefore repeated the peptide pulldown assays using the full-length ADD (479–610) and PWWP (281–424) domains generated as GST fusion proteins (Fig. 4d). Specific binding of the ADD domain was observed with the H4R3me2s peptide, but not with the unmodified or H4R3me2a peptides. No binding was observed with the PWWP domain. As an additional control, we included an unmodified N-terminal histone H3 peptide, as the ADD domain of DNMT3L has been shown to bind to histone H3 (ref. 31). Binding of the ADD domain of DNMT3A was also observed with this peptide. The significance of DNMT3A binding to histone H3 is unclear in this context, as loss of the H4R3me2s mark was sufficient to abolish DNMT3A binding to the γ-promoter in PRMT5-kd cells (Fig. 3e). We compared these findings to studies examining the binding of DNMT3A truncation mutants to PRMT5 (Supplementary Fig. 4a online). In contrast to the H4R3me2s peptide, only very weak binding to PRMT5 was observed with a mutant containing the ADD domain (351–912). Binding to levels comparable to those of the full-length protein was seen with the N-terminal fragment (1–354). Analysis of the isolated PWWP and ADD domains confirmed these findings (Supplementary Fig. 4b). Role of PRMT5 in developmental globin gene silencing To determine whether the H4R3me2s mark was evident at the human γ-globin promoter in a developmentally specific pattern in primary human cells, we isolated erythroid progenitors from cord blood and adult bone marrow. As expected, the expression of the genes encoding γ-globin was higher in cord blood compared with bone marrow using Q-RT-PCR (Supplementary Fig. 5 online). ChIP analysis demonstrated a fourfold increase in the H4R3me2s mark at the γ-promoter in adult bone marrow erythroid progenitors compared with progenitors derived from cord blood (Fig. 5a). This was accompanied by a reduction in RNA polymerase II localized to the γ-promoter in the adult cells. We then examined the distribution of PRMT5 and NF-E4 binding, and the H4R3me2s mark across the β-globin locus in bone marrow erythroid progenitors using ChIP analyses with anti-PRMT5, anti-NF-E4 and anti-H4R3me2s antibodies (Fig. 5b). We used primer pairs spanning regions of the locus control region (LCR) hypersensitive sites (HS1–HS4), the e-globin, γ-globin and β-globin promoters, and the intergenic region between the Gγ-genes and Aγ-genes. We detected high levels of PRMT5, NF-E4 and H4R3me2s at the transcriptionally inactive γ-promoters and e-promoters, but not at the b-promoter in these adult cells. We also observed increased levels at HS3 of the LCR, but not at the other hypersensitive sites or in the intergenic region. To further examine the role of PRMT5 in developmental silencing of the γ-genes, we determined the cellular localization of the protein by immunofluorescence in cord blood and bone marrow erythroid progenitors (Fig. 5c). PRMT5 has previously been shown to trans-locate from the nucleus to the cytoplasm at the time of extensive epigenetic reprogramming of mouse germ cells32. We demonstrated that the protein was predominantly nuclear in the bone marrow progenitors, whereas it was primarily localized in the cytoplasm in the cord blood progenitors. These findings suggest an additional mechanism by which PRMT5 may play a developmentally specific role in regulating gene expression at the human β-globin locus. DISCUSSION In addition to their direct role in influencing chromatin compaction, an emerging paradigm suggests that post-translational modifications of histones also serve as direct targets of effector molecules. These effectors (or “readers”) induce downstream functional consequences, including (i) remodeling or stabilization of the chromatin structure, (ii) introduction of further post-translational modifications through interactions with “writers” and (iii) other gene regulatory effects33. Our studies provide an important addition to this paradigm by defining a direct link between the repressive histone modification H4R3me2s and DNA methylation. We show that DNMT3A functions as both a reader and a writer in this context, by binding directly to the H4R3me2s mark and inducing DNA methylation and gene silencing (Fig. 6). Central to this is PRMT5, which establishes the histone modification and also interacts directly with DNMT3A. As with other effector molecules, DNMT3A targets the post-translationally modified histone through the ADD domain, which contains the PHD zinc finger region. PHD motifs have been shown in numerous factors to bind to methyllysine marks, including H3K4me2 and H3K4me3 (refs. 34–38), H3K9me3 (ref. 39), and H3K36me3 (ref. 40). Recently, the PHD domain of RAG2 was shown to bind coordinately to H3K4me3 and H3R2me2s41. Readers that recognize methylated arginine residues in histones have not been identified previously, and consequently no structure of a reader bound to this mark is available3. Tudor domain–containing proteins have been shown to recognize arginine-glycine–rich motifs in non-histone proteins in a methylarginine-dependent manner42, and although DNMT3A contains a Tudor-like motif, this region of the protein appears not to have a role in the H4R3me2s interaction29,43. Notably, it is involved in the protein-protein interaction between DNMT3A and PRMT5. Previous studies in the chicken β-globin locus have demonstrated that asymmetric methylation of H4R3 by the type I arginine methyltransferase PRMT1 is essential for the establishment and maintenance of a wide range of active chromatin modifications22. Taken together with our findings, this suggests that arginine methyltransferases may play coordinated and contrasting roles in the developmental regulation of the β-globin locus. In addition to increased levels of the H4R3me2s mark at the γ-promoter, we also observed enrichment of this mark at HS3, the site proposed to preferentially activate the human embryonic and fetal globin genes44. It is conceivable that coordinated repressive marks established synchronously at the promoters and HS3 could interfere with the LCR–globin gene interaction45,46, thereby inducing substantial changes in locus conformation in addition to the more localized modifications of chromatin structure. Alternatively, the complex could initiate at the LCR and spread across the locus leading to gene repression, much like the recent demonstration of epigenetic modifiers acting at a distance in the β-globin cluster in the setting of activation of β-gene expression47. The silencing of the genes encoding γ-globin after birth heralds the onset of b-thalassemia and sickle cell disease, which has prompted the search for therapies that will prevent or reverse this process. DNA methyltransferase inhibitors have been used for this purpose, with some effect48. An alternate approach that we are pursuing could involve the use of specific inhibitors of PRMT5 methyltransferase activity to block the H4R3me2s modification and subsequent recruitment of DNMT3A and DNA methylation. Our results provide an important extension to the established links between repressive histone modifications and DNA methylation. Interactions between lysine methyltransferases and DNA methyltransferases16–18,49, methyl-CpG–binding proteins12 and methyl-CpG–enriched regions10 have all been reported. Recently, the heterochromatin protein 1 (HP1) has been shown to translate methylation information from histone to DNA, to cement gene repression50. In this model, the G9a enzyme induces the repressive H3K9me2 mark at euchromatic sites, which is read by HP1. The bound HP1 functions as an adaptor by targeting DNMT1 enzymatic activity to these sites, thereby enhancing cytosine methylation. Further stabilization is achieved through a direct interaction between the G9a histone methyltransferase and DNMT1 (ref. 49). Our findings exhibit some parallels with this model—in particular, the dual role of PRMT5 in establishing the repressive mark and binding to DNMT3A, analogous to G9a. However, the recruitment of DNMT3A to H4R3me2s is direct, requiring no adaptor protein for the interaction. As such, it resembles the scenario in Arabidopsis thaliana, in which the DNA methyltransferase CMT3 is recruited directly to sites of repressive histone modifications51. Notably, other DNA methyltransferase subunits (for example, DNMT3L) fail to bind histone peptides that harbor activating histone marks such as H3K4me31. Thus, histone modifications, whether activating or repressing in nature, can influence DNA methylation signatures in positive or negative ways. METHODS Mass spectrometry We resolved Flag immunoprecipitates from K562 cells expressing NF-E4-Flag on a 4–20% (w/v) gradient SDS-PAGE gel and stained with SimplyBlue Safestain (Invitrogen). We excised protein bands of interest from preparative one-dimensional gels and subjected them to electrospray–ion trap (ESI-IT) tandem mass spectrometry (MS/MS) (LCQ-Deca, Finnigan). Cell culture and immunofluorescence K562 cells were grown as described previously20. We isolated CD34+ cells from human cord blood using a MiniMACS magnetic cell sorting system (Miltenyi Biotec) and cultured them in Iscove's modified Dulbecco's medium (IMDM) supplemented with 15% (v/v) fetal calf serum (FCS), SCF (100 ng ml−1), erythropoietin (5 U ml−1), insulin-like growth factor-1 (IGF-1, 40 ng ml−1) and dexamethasone (1 μM) to induce erythroid differentiation. We cultured CD34+ cells isolated from fresh adult bone marrow (as above) in IMDM supplemented with 15% (v/v) FCS, SCF (100 ng ml−1), IL-3 (10 ng ml−1) and Flt-3 ligand (500 ng ml−1) for 7 d, followed by erythropoietin (5 U ml−1) alone for 5 d to induce erythroid differentiation. Cell surface marker analysis with CD71 and glycophorin A indicated that cultured cord blood and bone marrow cells were greater than 90% erythroid lineage. For immunofluorescence, cord blood and bone marrow cells were mounted on polylysine slides and permeabilized with 0.1% (v/v) Triton X-100. We incubated slides with mouse monoclonal anti-PRMT5 antibody overnight at 4 °C, washed and incubated with Texas Red–conjugated horse anti-mouse secondary antibody (Vector Laboratories) for 1 h at room temperature (25 °C). We then washed the slides and counterstained them with 4’,6-diamidino-2-phenylindole (DAPI) for 3 min before imaging with a Zeiss Axioplan microscope (Zeiss). Protein interaction studies Immunoprecipitation, immunoblotting and GST pulldown assays were performed as described previously52. We used the following antibodies in the immunoprecipitations: Flag (Sigma-Aldrich), PRMT5 (Abcam) and DNMT1, DNMT3A, DNMT3B, tubulin and GATA-1 (Santa Cruz Biotechnology). Peptide pulldown assays were performed as previously described31. Briefly, we coupled 2 μg of COOH-terminal biotin-tagged 20-mer N-terminal peptides of histone H4 with the Arg3 residue unmethylated, symmetrically methylated or asymmetrically methylated to streptavidin beads and incubated with K562 cellular extract prepared with high salt extraction (420 mM NaCl)27. We eluted specifically bound protein from stringently washed beads and visualized by western blot with anti-DNMT3A or anti-GST antibody after SDS-PAGE. For peptide pulldown assays with GST proteins, the molar ratio of peptide to GST protein was 10:1. The peptides were synthesized with a polyethylene glycol 4 (PEG4) linker between the histone sequence and the biotin moiety, and the N terminus was acetylated. ChIP analysis ChIP assays were performed as described previously20. We immunoprecipitated chromatin fractions from K562 cells with specific antibodies. No antibody and normal rabbit IgG served as the controls. Sequences of the primers for the globin locus are listed in Supplementary Table 1 online. We used the following antibodies: H4R3me2s and PRMT5 (Abcam), Flag (Sigma-Aldrich), and DNMT3A and RNA polymerase II (Santa Cruz Biotechnology). We calculated the relative enrichment using a method described previously53. We calculated the percentage of ChIP DNA relative to the input DNA. In all assays it ranged from 0.35% to 0.71%. The most substantial enrichment was assigned the value 1, and all other conditions were normalized to this value. For the NF-E4 and PRMT5 ChIP, 32 cycles were used, and for the H4R3me2s ChIP, 35 cycles were used. We performed each experiment at least twice independently. In vitro methyltransferase assays Beads from the immunoprecipitation assays from K562 cells transfected with PRMT5-f or PRMT5Δ-f were used as the enzyme source in in vitro methyltransferase assays as described previously54, with slight modifications. Briefly, we incubated the beads with 10 μg of purified histone H2A, H2B, H3 and H4 (Roche), or purified nucleosomes55, and 2 mCi of S-adenosyl-l-methyl-3H-methionine (3H-SAM, Amersham) as the methyl donor, in a mixture of 20 μl of HMTase buffer (25 mM NaCl, 25 mM Tris, pH 8.8) for 2 h at 30 °C. Proteins were resolved on a 14% (w/v) SDS-PAGE gel, stained with Coomassie blue and then dried and subjected to autoradiography. Bisulfite sequence analysis Bisulfite sequence analysis was performed as described previously56. Primers to amplify the bisulfite-treated γ-promoter are provided in Supplementary Table 2 online. We performed PCR with HiFi Taq polymerase (Roche) as follows: 30 cycles, 94 °C for 20 s, 55 °C for 20 s and 68 °C for 35 s. We cloned PCR products into pCRII (Invitrogen) and then sequenced nucleotides using the Big-Dye Termination method (Applied Biosystems). The significance of the differences between cell lines was calculated using the Fisher's exact test. RNA interference and retroviral infection The small interfering RNA target sequences for PRMT5 and DNMT3A were inserted into the pSUPER.retro.-neo+gfp retroviral vector according to the manufacturer's recommendations (OligoEngine). The oligo sequences are provided in Supplementary Table 3 online. Retrovirus production by 293T cells and infection of K562 cells were performed as described20. Transduced cells were selected for green fluorescent protein expression by fluorescence-activated cell sorting (FACS). Q-RT-PCR Total RNA was isolated from cells with Trizol reagent (Invitrogen). Complementary DNA was generated using the reverse transcription system (Promega). Q-RT-PCR primers are provided in Supplementary Table 4 online. The identities of the amplified bands were confirmed by sequencing. We performed Q-RT-PCR in a Rotorgene 2000 (Corbett Research) in a final volume of 20 μl. Reaction mixtures contained 1× reaction buffer, 2.5 mM MgCl2, 0.05 mM deoxynucleotides (Roche), 0.1 μM gene-specific primers, 1 U Taq polymerase (Fisher Biotech), a 1:10,000 dilution of SYBR Green I (Molecular Probes) and 2 μl of sample or standard. Cycling conditions were 94 °C for 15 s, 57 °C for 30 s and 72 °C for 30 s. Standard curves for hypoxanthine guanine phosphoribosyltransferase (HPRT), γ-globin and β-globin were generated from bone marrow and cord blood cells. The relative quantity of the transcripts was calculated for all individual cell lines. Each reaction was done in duplicate. Supplementary Material 1 ACKNOWLEDGMENTS We thank S. Pestka (University of Medicine and Dentistry of New Jersey) for the PRMT5D plasmid, R. Gaynor (University of Texas Southwestern) for the PRMT5δ plasmid and members of the Jane and Cunningham laboratories for helpful discussions. This work was supported by grants from The National Health and Medical Research Council of Australia, the US National Institutes of Health (PO1 HL53749-03 and RO1 HL69232-01) (S.M.J.), The Roche Foundation for anemia research (RoFAR) (S.M.J.), The Cooley's Anemia Foundation (Q.Z.), The Natural Science Foundation of China #30670422 (Q.Z.), Cancer Centre Support CORE Grant P30 CA 21765, the American Lebanese Syrian Associated Charities (ALSAC) and the Assisi Foundation of Memphis (J.M.C.). Figure 1 PRMT5 symmetrically dimethylates histone H4R3 on the gene encoding γ-globin. (a) SimplyBlue Safestain of an SDS-PAGE gel of α-Flag antibody immunoprecipitates from K562 cells transfected with NF-E4-Flag, or vector alone (control) before analysis by mass spectrometry. The bands corresponding to PRMT5, NF-E4, polyubiquitinated NF-E4 (ref. 52) and the known PRMT5 partner proteins nucleolin and pICln are shown. Hash marks indicate bands that correspond to common background proteins including keratin, tubulins and ribosomal proteins. Asterisks indicate immunoglobulin chains. The PRMT5 peptide sequences identified are shown below. (b) Co-immunoprecipitation of endogenous PRMT5 and NF-E4 from K562 cells. (c) Interaction of the endogenous PRMT5 and NF-E4 with the γ-promoter by ChIP assays. (d) In vitro methyltransferase assays of Flag immunoprecipitates from K562 cells expressing PRMT5-f and K562 cells expressing PRMT5Δ-f. Autoradiographs (upper panels) and Coomassiestained gels (lower panels) are shown for each. (e) The HMTase assays with free histones detailed in d were immunoprecipitated with antibodies to histone H4R3me2s or normal rabbit IgG, and the radioactivity was quantitated in the respective precipitates. CPM, counts per minute. (f) H4R3me2s enrichment at the γ-promoter was measured by ChIP in K562 cells expressing PRMT5-f or PRMT5Δ-f. Error bars show s.d. Figure 2 PRMT5 mediates transcriptional silencing of the γ-gene. (a) Extracts from PRMT5-f, PRMT5Δ-f or vector control K562 cells were analyzed by western blot with anti-Flag or PRMT5 antibodies (right panel). RNA from these cells was analyzed by Q-RT-PCR with primers specific for the gene encoding γ-globin, and the signals were normalized against HPRT mRNA levels (left panel). (b) K562 cells expressing either an shRNA to PRMT5 (PRMT5-kd) or a scrambled sequence (Scr) were analyzed by western blot (with antibodies indicated, right panel) and Q-RT-PCR (left panel) as in a. Error bars show s.d. (c) H4R3me2s enrichment at the γ-promoter was measured by ChIP in PRMT5-kd and Scr cells. Error bars show s.d. Figure 3 PRMT5 and DNMT3A function cooperatively in gene silencing. (a) Effect of perturbed PRMT5 expression on DNA methylation at the human γ-gene. Each row shows the methylation status of individual CpG dinucleotides derived from sequence analysis of at least 40 individual cloned PCR products of the γ-genes following bisulfite modification from PRMT5-f, PRMT5Δ-f, PRMT-kd and the scrambled control (Scr) K562 cells. The differences between the three lines and the Scr were highly significant (P< o 0.01). The numbers on the left represent the positions of the CpG dinucleotides relative to the transcriptional start site of the γ-gene. The results are quantitated in the bar graph. ND, not detectable. (b) DNMT3A, but not DNMT1 or DNMT3B, co-immunoprecipitates with PRMT5-f from K562 cells. High salt extraction (420 mM NaCl) was used for the cellular extract preparation. (c) Co-immunoprecipitation of endogenous DNMT3A and PRMT5 from K562 cells. High salt extraction (420 mM NaCl) was used for the cellular extract preparation. (d) DNMT3A binding at the γ-promoter was measured by ChIP in PRMT5-f and PRMT5Δ-f cells. Error bars show s.d. (e) DNMT3A binding at the γ-promoter was measured by ChIP in PRMT-kd and Scr cells as above. (f) K562 cells expressing either an shRNA to DNMT3A (DNMT3A-kd) or a scrambled sequence (Scr) were analyzed by western blot, with indicated antibodies (right panel), and Q-RT-PCR with primers specific for the γ-globin genes, with the signal normalized against HPRT mRNA levels (left panel). (g) Effect of DNMT3A knockdown on DNA methylation at the human γ-genes as detailed in a. The difference between the two lines was significant (P < 0.02). Figure 4 DNMT3A binds specifically to histone H4 carrying the R3me2s modification. (a) Binding of 35S-labeled in vitro transcribed and translated (IVTT) DNMT3A to purified GST, GST-PRMT5 and GSTPRMT5D. Top panel, autoradiograph; bottom panel, Coomassie. Input represents 30% of the in vitro translated DNMT3A used in the assay. (b) Binding of DNMT3A to N-terminal peptides of histone H4 with the Arg3 residue unmethylated, symmetrically methylated or asymmetrically methylated. Specifically bound protein was visualized by western blot with anti-DNMT3A antibody after SDS-PAGE. Input represents 10% of the cellular extract used in the assay. The H4R3me2s modification of the synthesized peptide was confirmed by immunoblot. Coomassie staining shows equivalent loading of the three peptides on a 20% (w/v) SDS-PAGE gel. (c) Peptide pulldown assays as described in b, with 35S-labeled fragments of DNMT3A as shown in the schematic. Numbers refer to amino acids. The 1–354 construct contains the PWWP module but lacks its adjacent C-terminal helical motif. The 1–587 construct contains the GATA and PHD domains of ADD, but lacks an adjacent C-terminal helix. (d) Peptide pulldown assays as described in b, with purified GST, GST-DNMT3A (281–424, containing the PWWP domain) and GST-DNMT3A (479–610, containing the ADD domain). Specifically bound protein was visualized by western blot with anti-GST antibody after SDS-PAGE. Input represents 5% of the GST fusion proteins used in the assay. The H4R3me2s modification of the synthesized peptide was confirmed by immunoblot. Coomassie staining shows equivalent loading of the four peptides. Figure 5 Role of PRMT5 in developmental globin gene silencing. (a) H4R3me2s and RNA polymerase II enrichment at the γ-promoter was measured by ChIP in chromatin fractions from erythroid progenitors from cord blood and adult bone marrow. (b) Localization of PRMT5, NF-E4 and H4R3me2s across the β-globin locus measured by ChIP in chromatin fractions from erythroid progenitors from adult bone marrow. The precipitated DNA was amplified with primers specific for the indicated regions of the β-globin locus. HS, hypersensitive site; Pro, promoter; G/Aγ, intergenic region between Gγ-globin and Aγ-globin genes. Error bars show s.d. (c) Cellular localization of PRMT5 in erythroid progenitors from cord blood and adult bone marrow shown by immunofluorescence with anti-PRMT5 antibody and DAPI nuclear counterstaining. Scale bar, 10 μm. Figure 6 Model of PRMT5-induced silencing of gene expression. Symmetric dimethylation of H4R3 by PRMT5 induces direct binding of DNMT3A, resulting in methylation of adjacent CpG dinucleotides and gene silencing. AUTHOR CONTRIBUTIONS Q.Z., G.R., L.C. and Y.T.T. performed the experiments; R.L.M. and R.J.S. performed and analyzed the mass spectrometry; Q.Z., G.R. and S.M.J. designed and analyzed the research; D.J.C., C.D.A., H.L., D.J.P. and J.M.C. provided ideas, reagents and discussion; Q.Z., G.R. and S.M.J. wrote the manuscript. Supplementary information is available on the Nature Structural & Molecular Biology website. 1 Jenuwein T Allis CD Translating the histone code. Science 2001 293 1074 1080 11498575 2 Felsenfeld G Groudine M Controlling the double helix. Nature 2003 421 448 453 12540921 3 Taverna SD Li H Ruthenburg AJ Allis CD Patel DJ How chromatin-binding modules interpret histone modifications: lessons from professional pocket pickers. Nat. Struct. Mol. Biol 2007 14 1025 1040 17984965 4 Bird AP Wolffe AP Methylation-induced repression–belts, braces, and chromatin. Cell 1999 99 451 454 10589672 5 Jones PA Baylin SB The fundamental role of epigenetic events in cancer. Nat. Rev. Genet 2002 3 415 428 12042769 6 Freitag M Selker EU Controlling DNA methylation: many roads to one modification. Curr. Opin. Genet. Dev 2005 15 191 199 15797202 7 Klose RJ Bird AP Genomic DNA methylation: the mark and its mediators. Trends Biochem. Sci 2006 31 89 97 16403636 8 Eden S Hashimshony T Keshet I Cedar H Thorne AW DNA methylation models histone acetylation. Nature 1998 394 842 9732866 9 Schubeler D Genomic targeting of methylated DNA: influence of methylation on transcription, replication, chromatin structure, and histone acetylation. Mol. Cell. Biol 2000 20 9103 9112 11094062 10 Johnson LM The SRA methyl-cytosine-binding domain links DNA and histone methylation. Curr. Biol 2007 17 379 384 17239600 11 Fuks F The methyl-CpG-binding protein MeCP2 links DNA methylation to histone methylation. J. Biol. Chem 2003 278 4035 4040 12427740 12 Sarraf SA Stancheva I Methyl-CpG binding protein MBD1 couples histone H3 methylation at lysine 9 by SETDB1 to DNA replication and chromatin assembly. Mol. Cell 2004 15 595 605 15327775 13 Tamaru H Trimethylated lysine 9 of histone H3 is a mark for DNA methylation in Neurospora crassa. Nat. Genet 2003 34 75 79 12679815 14 Fuks F DNA methylation and histone modifications: teaming up to silence genes. Curr. Opin. Genet. Dev 2005 15 490 495 16098738 15 Schotta G A silencing pathway to induce H3–K9 and H4–K20 trimethylation at constitutive heterochromatin. Genes Dev 2004 18 1251 1262 15145825 16 Lehnertz B Suv39h-mediated histone H3 lysine 9 methylation directs DNA methylation to major satellite repeats at pericentric heterochromatin. Curr. Biol 2003 13 1192 1200 12867029 17 Fuks F Hurd PJ Deplus R Kouzarides T The DNA methyltransferases associate with HP1 and the SUV39H1 histone methyltransferase. Nucleic Acids Res 2003 31 2305 2312 12711675 18 Vire E The Polycomb group protein EZH2 directly controls DNA methylation. Nature 2006 439 871 874 16357870 19 van der Ploeg LH Flavell RA DNA methylation in the human gamma-delta-beta-globin locus in erythroid and non-erythroid cells. Cell 1980 19 947 958 6247075 20 Zhao Q Repression of human gamma-globin gene expression by a short isoform of the NF-E4 protein is associated with loss of NF-E2 and RNA polymerase II recruitment to the promoter. Blood 2006 107 2138 2145 16263792 21 Pollack BP The human homologue of the yeast proteins Skb1 and Hsl7p interacts with Jak kinases and contains protein methyltransferase activity. J. Biol. Chem 1999 274 31531 31542 10531356 22 Huang S Litt M Felsenfeld G Methylation of histone H4 by arginine methyltransferase PRMT1 is essential in vivo for many subsequent histone modifications. Genes Dev 2005 19 1885 1893 16103216 23 Fabbrizio E Negative regulation of transcription by the type II arginine methyltransferase PRMT5. EMBO Rep 2002 3 641 645 12101096 24 Pal S Vishwanath SN Erdjument-Bromage H Tempst P Sif S Human SWI/ SNF-associated PRMT5 methylates histone H3 arginine 8 and negatively regulates expression of ST7 and NM23 tumor suppressor genes. Mol. Cell. Biol 2004 24 9630 9645 15485929 25 Branscombe TL PRMT5 (Janus kinase-binding protein 1) catalyzes the formation of symmetric dimethylarginine residues in proteins. J. Biol. Chem 2001 276 32971 32976 11413150 26 Goren A Fine tuning of globin gene expression by DNA methylation. PLoS ONE 2006 1 e46 17183675 27 Jane SM Ney PA Vanin EF Gumucio DL Nienhuis AW Identification of a stage selector element in the human gamma-globin gene promoter that fosters preferential interaction with the 5′ HS2 enhancer when in competition with the beta-promoter. EMBO J 1992 11 2961 2969 1639067 28 Cheng X Blumenthal RM Mammalian DNA methyltransferases: a structural perspective. Structure 2008 16 341 350 18334209 29 Maurer-Stroh S The Tudor domain ‘Royal Family’: Tudor, plant Agenet, Chromo, PWWP and MBT domains. Trends Biochem. Sci 2003 28 69 74 12575993 30 Bienz M The PHD finger, a nuclear protein-interaction domain. Trends Biochem. Sci 2006 31 35 40 16297627 31 Ooi SK DNMT3L connects unmethylated lysine 4 of histone H3 to de novo methylation of DNA. Nature 2007 448 714 717 17687327 32 Ancelin K Blimp1 associates with Prmt5 and directs histone arginine methylation in mouse germ cells. Nat. Cell Biol 2006 8 623 630 16699504 33 Ruthenburg AJ Allis CD Wysocka J Methylation of lysine 4 on histone H3: intricacy of writing and reading a single epigenetic mark. Mol. Cell 2007 25 15 30 17218268 34 Shi X ING2 PHD domain links histone H3 lysine 4 methylation to active gene repression. Nature 2006 442 96 99 16728974 35 Pena PV Molecular mechanism of histone H3K4me3 recognition by plant homeodomain of ING2. Nature 2006 442 100 103 16728977 36 Wysocka J A PHD finger of NURF couples histone H3 lysine 4 trimethylation with chromatin remodelling. Nature 2006 442 86 90 16728976 37 Li H Molecular basis for site-specific read-out of histone H3K4me3 by the BPTF PHD finger of NURF. Nature 2006 442 91 95 16728978 38 Vermeulen M Selective anchoring of TFIID to nucleosomes by trimethylation of histone H3 lysine 4. Cell 2007 131 58 69 17884155 39 Iwase S The X-linked mental retardation gene SMCX/JARID1C defines a family of histone H3 lysine 4 demethylases. Cell 2007 128 1077 1088 17320160 40 Shi X Proteome-wide analysis in Saccharomyces cerevisiae identifies several PHD fingers as novel direct and selective binding modules of histone H3 methylated at either lysine 4 or lysine 36. J. Biol. Chem 2007 282 2450 2455 17142463 41 Ramon-Maiques S The plant homeodomain finger of RAG2 recognizes histone H3 methylated at both lysine-4 and arginine-2. Proc. Natl. Acad. Sci. USA 2007 104 18993 18998 18025461 42 Cote J Richard S Tudor domains bind symmetrical dimethylated arginines. J. Biol. Chem 2005 280 28476 28483 15955813 43 Slater LM Allen MD Bycroft M Structural variation in PWWP domains. J. Mol. Biol 2003 330 571 576 12842472 44 Fraser P Pruzina S Antoniou M Grosveld F Each hypersensitive site of the human beta-globin locus control region confers a different developmental pattern of expression on the globin genes. Genes Dev 1993 7 106 113 8422981 45 Carter D Chakalova L Osborne CS Dai YF Fraser P Long-range chromatin regulatory interactions in vivo. Nat. Genet 2002 32 623 626 12426570 46 Tolhuis B Palstra RJ Splinter E Grosveld F de Laat W Looping and interaction between hypersensitive sites in the active beta-globin locus. Mol. Cell 2002 10 1453 1465 12504019 47 Demers C Activator-mediated recruitment of the MLL2 methyltransferase complex to the beta-globin locus. Mol. Cell 2007 27 573 584 17707229 48 Saunthararajah Y Effects of 5-aza-2¢-deoxycytidine on fetal hemoglobin levels, red cell adhesion, and hematopoietic differentiation in patients with sickle cell disease. Blood 2003 102 3865 3870 12907443 49 Estève PO Direct interaction between DNMT1 and G9a coordinates DNA and histone methylation during replication. Genes Dev 2006 20 3089 3103 17085482 50 Smallwood A Esteve PO Pradhan S Carey M Functional cooperation between HP1 and DNMT1 mediates gene silencing. Genes Dev 2007 21 1169 1178 17470536 51 Lindroth AM Dual histone H3 methylation marks at lysines 9 and 27 required for interaction with CHROMOMETHYLASE3. EMBO J 2004 23 4286 4296 15457214 52 Zhao Q Cumming H Cerruti L Cunningham JM Jane SM Site-specific acetylation of the fetal globin activator NF-E4 prevents its ubiquitination and regulates its interaction with the histone deacetylase, HDAC1. J. Biol. Chem 2004 279 41477 41486 15273251 53 Wysocka J WDR5 associates with histone H3 methylated at K4 and is essential for H3 K4 methylation and vertebrate development. Cell 2005 121 859 872 15960974 54 Rea S Regulation of chromatin structure by site-specific histone H3 methyltransferases. Nature 2000 406 593 599 10949293 55 Brand M Rampalli S Chaturvedi CP Dilworth FJ Analysis of epigenetic modifications of chromatin at specific gene loci by native chromatin immunoprecipitation of nucleosomes isolated using hydroxyapatite chromatography. Nat. Protoc 2008 3 398 409 18323811 56 Lavelle D Vaitkus K Hankewych M Singh M DeSimone J Effect of 5-aza-2′-deoxycytidine (Dacogen) on covalent histone modifications of chromatin associated with the epsilon-, gamma-, and beta-globin promoters in Papio anubis. Exp. Hematol 2006 34 339 347 16543068
PMC005xxxxxx/PMC5120863.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 2985117R 4816 J Immunol J. Immunol. Journal of immunology (Baltimore, Md. : 1950) 0022-1767 1550-6606 27742829 5120863 10.4049/jimmunol.1501859 NIHMS818722 Article Endothelial Plasmalemma Vesicle Associated Protein regulates the homeostasis of splenic immature B cell and B1 B cells Elgueta Raul \|\| Tse Dan ‡#\| Deharvengt Sophie J. ‡ Luciano Marcus R. ‡ Carriere Catherine §¶#\| Noelle Randolph J. ¶ Stan Radu V. †‡# Department of Microbiology and Immunology, Geisel School of Medicine at Dartmouth, Lebanon, NH 03756 † Department of Biochemistry and Cell Biology, Geisel School of Medicine at Dartmouth, Lebanon, NH 03756 ‡ Department of Pathology, Geisel School of Medicine at Dartmouth, Lebanon, NH 03756 § Department of Medicine, Geisel School of Medicine at Dartmouth, Lebanon, NH 03756 ¶ Norris Cotton Cancer Center, Geisel School of Medicine at Dartmouth and Dartmouth-Hitchcock Medical Center, Lebanon, NH 03756 \|\| Department of Immune Regulation and Intervention, MRC Centre for Transplantation, King’s College London, Guy’s Hospital, London, SE1 9RT, UK # Corresponding Authors: Radu V. Stan: Department of Biochemistry and Cell Biology, Geisel School of Medicine at Dartmouth, One Medical Center Drive, Lebanon, NH 03756. Tel: (603) 650-8781 Fax: (603) 650-6166. [email protected]. Randolph J Noelle: Department of Microbiology and Immunology, Geisel School of Medicine at Dartmouth, One Medical Center Drive, Lebanon, NH 03756. Tel: (603) 653-9908. [email protected] #\| present address ImmuNext, Lebanon, NH 03756 27 9 2016 14 10 2016 15 11 2016 15 5 2017 197 10 39703981 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Plasmalemma vesicle associated protein (Plvap) is an endothelial protein with roles in endothelial diaphragm formation and maintenance of basal vascular permeability. At the same time Plvap has roles in immunity by facilitating leukocyte diapedesis at inflammatory sites and controlling peripheral lymph node morphogenesis and the entry of soluble antigens into lymph node conduits. Based on its postulated role in diapedesis, we have investigated the role of Plvap in hematopoiesis and show that deletion of Plvap results in a dramatic decrease of IgM+IgDlo B cells in both the spleen and peritoneal cavity. Tissue specific deletion of Plvap demonstrates that the defect is B cell extrinsic, as B cell and pan hematopoietic Plvap deletion has no effect on IgM+IgDlo B cell numbers. Endothelial specific deletion of Plvap in the embryo or at adult stage recapitulates the full Plvap knockout phenotype whereas endothelial specific reconstitution of Plvap under the Chd5 promoter rescues the IgM+IgDlo B cell phenotype. Taken together, these results show that Plvap expression in endothelial cells is important in the maintenance of IgM+ B cells in the spleen and peritoneal cavity. Introduction The innate immune response is the host’s first and most rapid response to infection with a pathogen, whereas the adaptive immune response involves a complex process including activation, expansion and differentiation of pathogen-specific B and T cells. The development of adaptive immunity requires several days to weeks to generate a long-standing effector and memory immune response (1, 2). A key transition from innate to adaptive immunity is mediated by the marginal zone (MZ) B and B-1 cells as they produce the first set of low-affinity antibodies against the pathogen (3). MZ B and B-1 cells are localized in marginal sinus and peritoneal cavity, respectively, where they are favored as the first cells to sample antigens in the blood and gut. Moreover, MZ and B-1 B cells are well characterized as having a low activation threshold and their BCRs recognize a wide range of microbial antigens (4). Both B cell subsets significantly contribute to levels of serum IgM, and the production of natural antibodies. Natural antibodies in many cases can be specific to pathogen-encoded molecules and be critical in the rapid neutralization of both viruses and bacteria (5). MZ B cells arise from bone marrow precursors through transitional B cells, which colonize the periarteriolar lymphoid sheath (5). The differentiation of transitional B cells to MZ B cells is driven by a weak BCR activity through a dependent pathway Bruton’s tyrosine kinase (6–8). This and the interaction of NOTCH expressed on transitional B cells with the ligand, Delta-like 1, on endothelial cells induce the differentiation to MZ B cells (9). The homing of MZ B cells is dependent on circulating sphingosine-1- phosphate (S1P) binding to S1P1 and S1P3 receptors expressed in the endothelial cells of blood vessels of MZ (10, 11). After migration, MZ B cells are retained by the interaction of αLβ2 and α4β1 with ICAM1 and VCAM1, respectively (12). In contrast, B-1 cells are competently produced before birth and throughout the first couple weeks after birth. The precursors for B-1 cells have been discovered in the splanchnopleura region, yolk sac and intra-embryonic hemogenic endothelium, fetal liver but they are absent from adult bone marrow (13–16). B-1 cells constantly circulate to and from the peritoneal space across the omentum in a process that involves CXCL13, which is likely produced by macrophages (17). Collectively, these findings show that B cell progenitors migration is highly regulated by molecules expressed on endothelial cells. However, it is not known whether molecules expressed on endothelial cells are involved in B cell differentiation and trafficking. Plasmalemma vesicle associated protein (Plvap) is a vertebrate gene (18, 19) whose product, Plvap, is a heparin-binding (20) homodimeric, single span type II membrane glycoprotein (21–23) critical for the formation of the stomatal diaphragms of caveolae, transendothelial channels and vesiculo-vacuolar organelles as well as the diaphragms of fenestrae in both mice (24–27) and humans (28). Microscopic (19, 22) and genetic (26, 28, 29) lines of investigation led to the conclusion that Plvap is specifically expressed in the endothelial cells of blood vessel capillaries and venules in select vascular beds and in the heart endocardium and absent from lymphatic endothelial cells. This pattern of expression was fully supported by a large body of literature obtained with two endothelial specific monoclonal antibodies that bind Plvap, such as MECA-32 (30, 31) in the mouse and PAL-E (32–34) in humans (reviewed in (35, 36)). However, recently Plvap expression was also demonstrated in the sinus lymphatic endothelial cells of peripheral lymph nodes (PLN) while confirming its absence from peripheral lymphatics elsewhere ((27) and immgen.org). At the whole organism level, endothelial diaphragms formed by Plvap (24, 37, 38) are critical for maintaining basal permeability of fenestrated blood vessels, their absence resulting in disrupted blood homeostasis and reduced survival (26, 28). In PLN, Plvap positive diaphragms in the sinus lymphatic endothelial cells control the entry of soluble antigens and lymphocytes in PLN parenchyma (27). The expression of Plvap and diaphragm formation is increased in activated endothelial states associated with inflammation (39, 40) and physiological and pathological angiogenesis (35, 41, 42) where it has active roles. Plvap is required for cancer progression (43) and diapedesis of leukocytes into inflammation sites in vivo (31). In vitro Plvap knock-down and antibody-mediated blockade experiments, suggest that endothelial Plvap is important for the transcellular transmigration but not for adhesion and rolling of lymphoblasts with no effect on neutrophils transmigration (31). Plvap is thought to control the transcellular migration of lymph-borne lymphocytes into PLN parenchyma (27). Deletion of Plvap results in defective PLN morphogenesis with mild decreases in the T cell compartment (both CD4 and CD8 T cells), hyperplastic B cell follicles and increases in both PLN B and T cells activation. Intriguingly, Plvap deletion increases the entry of adoptively transferred lymph borne splenocytes (both B and T cells) whereas its ligation with MECA-32 antibody inhibits the recruitment of these subsets (27). The mechanism of how Plvap mediates transendothelial migration of immune cells in currently unclear. Here, we have examined whether Plvap plays a role in the development and homeostasis of hematopoietic lineages taking advantage of recently created genetic models of Plvap gain and loss of function and endothelial specific reconstitution (26). Our studies show that deletion of Plvap results in an intense reduction of IgM+ B cells in both spleen and peritoneal cavity. Tissue specific deletion of Plvap demonstrates that the defect is B cell extrinsic, as B cell and pan hematopoietic Plvap deletion has no effect on IgM+ B cell numbers. Endothelial specific deletion of Plvap recapitulates full Plvap knockout, while endothelial specific reconstitution of Plvap rescues the IgM+ B cell phenotype. Taken together these results demonstrate that Plvap expression on endothelial cells is key in the maintenance of IgM+ B cells into spleen and peritoneal cavity. Material and Methods Conditional deletion of Plvap Homozygous PlvaploxP (PlvapL/L) mice were generated by knock-in using homologuous recombination in mice, as already described (26). PlvapL/L mice express Plvap at normal levels and have no overt phenotype (26). The PlvapL/L mice were bred to mice expressing the cre recombinase under the control of different promoters allowing to generate compound mice where Plvap was deleted in: a) germline (label Plvap−/−, genotype Plvap−/−;CMV-cretg/+) using CMV-cretg/+ transgenic mice, (JAX strain BALB/c-Tg(CMV-cre)1Cgn/J). b) endothelial and hematopoietic cells in the embryo using Ins-VEC-cre transgenic mice (44) (label PlvapECKO-VEC, genotype PlvapL/L;Ins-VEC-cretg/+) c) endothelial and hematopoietic cells in the embryo using Tie2/Tek-cretg/+ mice (JAX strain B6.Cg-Tg(Tek-cre)12Flv/J) (label PlvapECKO-Tie2, genotype PlvapL/L;Tek-cretg/+) d) B cell specific deletion of Plvap in the embryo using CD19-creKI/+ knock-in mice, (JAX strain B6.Cg-Cd19tm1(cre)Cgn/J) (45) (label CD19CrexPlvapL/L, genotype PlvapL/L;CD19-creki/+) e) hematopoietic cells in the embryo using Vav1-cre+/− transgenic mice (46), for deletion of Plvap in all the hematopoietic cell lineages but not in endothelium (44, 46); f) for inducible deletion of Plvap in the endothelial cells of the adult mice we used end-SCL-Cre-ERTtg/+ transgenic mice to generate mice with the genotype: PlvapL/L; end-SCL-CreERT tg/+ (label PlvapiECKO). Deletion of Plvap was achieved by dosing 4 weeks old mice (both males and females) by gavage with 7 doses of 4mg tamoxifen spaced at 48 hours. Experiments were carried out two weeks after the last tamoxifen dose administration. Control animals for germline deletion were sex and age matched WT or CMV-cretg/+ (labeled WT) and Plvap+/− or Plvap+/−;CMV-cretg/+ (labeled Plvap+/−) littermates. Control animals for tissue specific deletions were sex and age matched WT or PlvapL/L littermates. VEC-Plvap-HAtg/+ transgenic mice (26) that express Plvap-HA fusion protein specifically in the endothelial cells under the control of cadherin 5 (VE Cadherin) promoter, were used to reconstitute Plvap in endothelial cells in the context of the Plvap−/− (label PV1ECRC) as described (26). The same strategy was used to reconstitute Plvap in the context of PlvapiECKO (label PlvapiECRC, genotype PlvapL/L; end-SCL-CreERT tg/+; VEC-Plvap-HAtg/+). All animals were maintained in a pathogen-free facility at Dartmouth College. All procedures were approved by the local IACUC. Antibodies, Staining and analysis by flow cytometry The following antibodies and staining reagents were used: IgG1 (clone A85-1), IgG2a/b (clone R2-40), CD138 (clone 281-2), IgM (clone 11–41), CD24 (clone M1/69), CD4 (clone RM4-5), CD21/35 (clone 7E9), CD8 (clone 53-6.7), CD25 (clone 3C7), CD62L (clone MEL-14), CD69 (clone H1.2F3), CD44 (clone lM7), CD11c (clone HL3), CD80 (clone 16-10A1), CD86 (clone GL1), CD11b (clone M1/70), c-Kit (CD117, clone 2B8), streptavidin-PerCP, from BD Pharmingen; Plvap (clone MECA-32) was from Abcam; CD38 (clone 90), B220 (clone 6B2), CD23 (clone B3B4), and IgD (clone 11-26c), FceRI (clone MAR-1) from eBioscience; peanut agglutinin from Vector Laboratories. MECA-32 mAb (anti mouse Plvap rat IgG2a) secreting hybridoma was obtained from the Developmental Studies Hybridoma Bank (DSHB, U. Iowa) was produced in serum free conditions by BioXCell (Lebanon, NH). The antibody was labeled with AlexaFluor (AF) fluorochromes using the protein labeling kits for AF488, and AF647 (Invitrogen), as per manufacturer’s instructions. Flow cytometry was performed on either a refurbished FACSCAN or a FACSCalibur running CellQuest software (BD), or a FACSCanto running FACSDiva software within the Norris Cotton Cancer Center DartLab Immune Monitoring Facility. The data analysis performed using FlowJo (Tree Star, Inc.). Isolation of cells from spleen, peripheral lymph nodes and Peyer’s patches To analyze B, T cells and monocytes, single-cell suspensions of lymphocytes were prepared from spleens by mechanical disruption in HBSS followed by passing the cells through a 70μm cell strainer. In experiments where dendritic cells were profiled an enzymatic digestion step (45min, 37°C) was included using DNAse (10 mg/mL Roche) and Liberase (12.5 mg/mL, Roche) before mechanical disruption. Cells were collected by centrifugation (5min, 500xg, 4°C) and red blood cells were lysed (2min, 37°C) using RBC Lysis buffer (BioLegend). Total number of cells and cell viability were determined using either a hemacytometer and trypan blue or a Guava system (Millipore). Cells (105–106) were stained with antibody cocktails, as noted. Analysis was performed after co-staining with the cocktail of antibodies by flow-cytometry. For isolation of peritoneal cells, the peritoneal cavity was flushed with 5 mL warm (37°C) PBS, 2% BSA, 2mM EDTA, 0.02% sodium azide and 10 U/mL Heparin. Analysis of Plvap expression in splenocytes of naïve and LPS treated mice Female wild type C57Bl6/J mice were either treated with 50 μg of LPS (E. coli 055:B5, Sigma) in PBS or an equal volume of PBS alone (control mice), i.p. 12 hrs before spleens were harvested, enzymatically dissociated, as noted above, followed by antibody staining of splenocytes, and analyzed by flow cytometry. Immunofluorescence and Laser Confocal microscopy Tissues were snap-frozen in optimal cutting temperature medium and sectioned to 8 μm. Sections were collected on charged slides (Surgipath) fixed (−20°C, 10min) with cold methanol, rinsed (3×2min, RT) in PBS, encircled with hydrophobic barriers (PapPen), blocked (30min, RT) with 10% rat serum in PBS containing 10μg/ml mouse Fc block, incubated (1h, RT, in dark) with various fluorescently labeled primary antibody cocktails in blocking buffer, rinsed (3×5min, RT, in dark) again in PBS, stained (10min, RT, in dark) with 300 nM 40,6-diamidino-2-phenylindole dihydrochloride (DAPI, D1306; Life Technologies) and washed (3× 5 min, RT) in PBS. Labeled sections were mounted under #1.5 coverslips using a polymerizing mounting medium (Fluoromount G, Southern Biotech). The antibodies used were: rat anti-mouse Plvap– AF568 (clone MECA-32), rat anti-mouse CD169 -FITC (clone MOMA-1, AbD Serotec) and rat anti-mouse/human B220-AF647 (clone 6B2). Labeled sections were analyzed using a Zeiss LSM510 Meta confocal microscope equipped with appropriate lasers (405 nm, 488 nm, 532nm, 633nm) and filters, all within the Norris Cotton Cancer Center microscopy facility. The acquired images were processed for brightness and contrast and analyzed using ImageJ (http://imagej.nih.gov/ij/), and the figures were mounted using Adobe Photoshop and Adobe Illustrator CS6. Statistics Results are expressed as mean plus or minus SEM. Two-tailed Student t test with unequal variance was used to evaluate the statistical significance of the data. Results Deletion of Plvap in mice results in reduced numbers of splenic IgM+ B cells Plvap is an endothelial protein that is involved in the diapedesis of leukocytes at sites of inflammatory challenge and PLN sinuses. Diapedesis is a process that is central to the development and homeostasis of hematopoietic lineages. To understand the role of Plvap in these processes, we used recently generated Plvap−/− mice (26) to characterize the function of this molecule on the numbers and subset composition of leukocytes in the blood, spleen, PLN and Peyer’s patches (PP). Plvap−/−, Plvap+/− and wild type (WT) littermate control mice were profiled by flow cytometry to determine whether there are modifications in terms of leukocyte subset numbers, frequency or function. No differences were found in cellular composition of peripheral blood, LN and PP with respect with percentages, number and viability of granulocytes, T- and B-lymphocytes, NK cells and monocytes (Fig. 1A, supplemental Fig. S1 and data no shown). In the spleen, there was a drastic reduction in the percentage and absolute number of IgM+ IgD− B cells in Plvap−/− mice (Fig. 1B and 1C), whereas IgD+ B cells were not affected (Fig. 1B). In the spleen, there are several prominent B cell subsets that are represented (47). To evaluate whether the reduction of this population is due to a reduction of a specific sub-population of splenic B cells, we stained with a panel of antibodies that identify those subpopulations (47). Neither the percentage of follicular (CD21/35int IgMint) or marginal zone B cells (CD21/35+ IgM+) is affected in Plvap−/− mice (Fig. 1D). Interestingly, CD21/35lo IgM+ B cells are reduced in the Plvap−/− mice compared with the WT or Plvap+/− mice (Fig. 1D and E). Lastly, CD21/35lo IgM+ B cells can be subdivided in Transitional 1 (T1) or Transitional 2 (T1) B-lymphocytes using the expression of CD23 and HSA (Figure 1F). Our results show that the proportion of T1 or T2 B cells is not affected in the Plvap−/− mice (Figure 1F and G), indicating that both, T1 and T2 B cells are reduced. Taken together, these results show that amongst the splenic B cell populations, in Plvap-deficient mice there is a selective reduction in tansitional IgM+ B cells. Similarly, no difference in the proportion of transitional B cells in Plvap−/− mice (data not shown) was found when expression of CD93 was used for the analysis of transitional B cells (48). Thus, irrespective of the markers used to analyze transitional B cells, we obtain the same results. IgM+ B cells are reduced in the peritoneum of Plvap−/− mice Based on the above observations showing a decrease of IgM+ B cells in the spleen of Plvap−/− mice, we hypothesized that the same reduction would also be found in peritoneal B-1 B cells. Profiling of the resident leukocytes in the peritoneum obtained by peritoneal lavage, demonstrated a drastic reduction in the total number of cells in the peritoneum of Plvap−/− mice compared to WT or Plvap+/− littermates (Fig. 2A). The reduction in total viable leukocyte numbers in the peritoneal cavity was accompanied by a low percentage and absolute number of IgD+ B cells and the absence of IgM+ IgDlow B cells (Fig. 2B–D). Taken together these results suggest that Plvap plays a role in the recruitment or retention of B cells into the peritoneal cavity and/or their survival. Due to the drastic reduction of cells in the peritoneum cavity of Plvap−/− mice, we also analyzed the percentage of mast cells and monocytes in this compartment. Our results show that mast cells (cKit+ FceRI+ cells, supplemental Fig. S2A) and monocyte (CD11b+ Gr-1+ cells, data not shown) percentages were also reduced in Plvap−/− mice compare with control mice, indicating that Plvap is an important molecule in migration to, or retention into the peritoneum. Plvap deletion within the hematopoietic compartment does not impact on the frequency of B cells in the spleen and peritoneum To understand the underlying mechanisms responsible for altered B cell frequencies in the spleen and peritoneum, we sought to determine whether these effects are B cell intrinsic or extrinsic. First, we inquired whether Plvap was expressed on hematopoietic cells and, second, established the impact on B cell frequencies of genetic B-cell specific and pan-hematopoietic deletion of Plvap. LPS-activated or resting B cells or dendritic cells did not express Plvap, as detected with antibody staining and using Plvap−/− splenocytes as controls. Furthermore, CD4+ and CD8+ T cells were also negative for Plvap expression (Fig. 3A, B and C). Examination of Plvap expression by confocal microscopy revealed that expression was limited to the splenic blood vessels in the MZ area (determined by MOMA-1+ macrophage localization) (Figure 3D). Taken together, these data suggest that Plvap is not expressed within the hematopoietic subsets studied, in agreement with recent data published by the Immunological Genetic Consortium (immgen.org) (49). While expression analysis conclusively established the lack of hematopoietic expression of Plvap, genetic deletion of Plvap in hematopoietic cells and subsets was used to confirm these findings. PlvapL/L mice were interbred with CD19-Cre (45) or Vav1-Cre mice (46) to obtain compound mice lacking Plvap in the B cells (CD19CrexPlvapL/L) or all hematopoietic cells (VavCrexPlvapL/L) (26), respectively. VavCrexPlvapL/L mice have over 97% Plvap deletion in the hematopoietic compartment (26). B cells isolated from CD19CrexPlvapL/L mice using magnetic separation showed >95% deletion of Plvap (data not shown). Immune profiling of the peritoneal lavage showed that there was no effect on B frequencies or phenotype in the spleen and peritoneum in either CD19CrexPlvapL/L (Fig. 4A – B) or Vav1CrexPlvapL/L (Fig. 4C–D) mice. The combination of microscopy, flow cytometry, and genetic data clearly demonstrate that Plvap is not expressed in the hematopoietic compartment and that the maintenance of IgM+ B cells in both spleen and peritoneal cavity is due to Plvap expression outside the hematopoietic compartment. Plvap deletion on endothelial cells reduces IgM+ B cells in spleen and peritoneum Plvap is a molecule specifically expressed in endothelial cells of blood vessels (26, 29) and PLN sinus (27), suggesting that endothelial cells might impact on the presence of IgM+ B cells in spleen and peritoneum. For this purpose, we made use of Tie2CrexPlvapL/L mice (26) where Plvap is efficiently (>95%) deleted in the endothelial cells and the hematopoietic compartment at embryonic stage. The results obtained in Tie2CrexPlvapL/L mice phenocopied those obtained in Plvap−/− mice. There was a reduction in IgM+ IgD− B cell percentage in the spleen from Tie2CrexPlvapL/L compared with control littermates (Fig. 5A–C). Additionally, there also was a striking reduction in the percentage of IgD+ and IgMpos IgD− B cells in the peritoneal cavity of Tie2CrexPlvapL/L compared to control mice (Figure 5D–F). These data suggest that the expression of Plvap on endothelial cells regulates the numbers of IgM+ B cells in the spleen and peritoneal cavity. Endothelial reconstitution of Plvap rescues the Plvap−/− phenotype and restores B cell frequencies in the spleen and peritoneal cavity In order to demonstrate that Plvap−/− phenotype is not due to a distortion of other genetic loci close to the Plvap locus, we transgenically complemented the Plvap-deficiency by the expression of Plvap-HA in endothelial cells of Plvap−/− mice (26). Previously, we have generated mouse lines (VEC-Plvap-HA+/tg) ta (26) expressing Plvap under the control of the VE Cadherin promoter and 5′ intronic enhancer (50). We used VEC-PlvapHA+/tg and Plvap+/− mice, to generate compound Plvap−/−; VEC-PlvapHA+/tg (PlvapECRC) mice, which display between 30% and 50% reconstitution of native endothelial Plvap levels (26). As shown in Fig. 6, IgM+ IgDlow B cells are recovered to normal levels in both spleen (Fig. 6A–C) and peritoneum (Fig. 6D–F) of PlvapECRC mice compare to Plvap−/− mice. Moreover, IgD+ B cells were also recovered in the peritoneum of PlvapECRC mice (Fig. 6D–F). Together all these results suggest that the expression of Plvap in endothelial cells is necessary to maintain normal levels of B cells in the spleen and peritoneum compartment. Plvap deletion in blood vessel endothelial cells does not affect the development of B cells in the bone marrow To determine whether Plvap deletion leads to a defect in B cell development in bone marrow, we profiled B cells progenitors in the marrow of PlvapL/L; end-SCL-CreERT tg/+ compound mice (PlvapiECKO) and mice with endothelial reconstitution of Plvap in the context of the PlvapiECKO (labeled PlvapiECRC mice) (see Methods). The promoter driving the expression of CreERT fusion protein consists of SV40 virus minimal promoter and the endothelial enhancer of the Stem Cell Leukemia gene (SCL) (51). This compound promoter confers activity in endothelial cells from select vascular beds, bone marrow included, as previously shown (51, 52). To minimize the effects on the peritoneal cavity the tamoxifen was administered by gavage. As shown in Fig. 7A and 7B, IgM+ IgDlow B cells are reduced in both spleen (Fig. 7A–B) and peritoneum (Fig. 7C–D) of PlvapiECKO mice compared to control (tamoxifen treated PlvapL/L) mice, whereas endothelial reconstitution in PlvapiECRC mice reverses the effect. In addition, IgD+ B cells were also reduced in the peritoneum of PlvapiECKO mice (Fig. 7C–D), indicating that endothelial deletion in adult mice has a similar phenotype as that obtained in Plvap−/− mice with respect to B-1 and MZ B cells. (Fig. 1). Using these models, we analyzed the proportion of immature B cells (B220+IgMhiIgD+), mature B cells (B220+IgMintIgD+) and pre-B cells (B220+IgM+IgD+) from bone marrow (53, 54). However, no changes were detected in B cell subsets in the bone marrow of PlvapiECKO mice as compared to controls or PlvapiECRC mice (Fig. 7E–G). These results together suggest that the defect observed in the splenic immature B cells and peritoneum B cells are not due to a defect in the B cell development in the bone marrow. Discussion The development of effective immune responses is dependent on endothelial cell-mediated leukocyte migration into sites of inflammation. The results presented here are the first to determine the critical role of Plvap in the maintenance of IgM+ B cells in spleen and peritoneum. We show that the lack of Plvap results in a reduction of transitional splenic IgM+ B cells as well as B-1 B cells in the peritoneum. The tissue specific deletion of Plvap clearly demonstrate that the role of Plvap is extrinsic to the B cell compartment, as B cell and pan hematopoietic Plvap deletion has no effect on the IgM+ B cells frequencies in either the spleen or peritoneal cavity. In contrast, endothelial specific deletion of Plvap recapitulates full Plvap knockout phenotype and endothelial specific reconstitution of Plvap in the context of germline Plvap knockout rescues the IgM+ B cell phenotype. Taken together, these results conclusively demonstrate that Plvap expression in endothelial cells is key in maintenance of IgM+ B cells in spleen and peritoneal cavity. The reduced number of immature B cells in spleen and B-1 cells in the peritoneum of Plvap deficient mice could also explain the previously reported marked reduction of IgM and IgA titers in Plvap deficient mice (26), further suggesting that Plvap may regulate the abundance of B cells that produce natural antibodies. Previous reports have shown Plvap expression in spleen ((19, 55) and immgen.org), especially on the endothelial cells of the marginal sinus of the spleen (55). Our observations confirm these findings by showing Plvap expression on vessels in the marginal zone of spleen (identified by co-localization of MOMA-1+ macrophages) as well as capillaries in the red pulp. Thus Plvap is expressed at the site where it could regulate IgM+ B cells trafficking or retention in this histological site. While human Plvap expression has been reported in circulating human lymphocytes and monocytes (i.e. PBMCs) by intracellular staining (31), we could not find Plvap expression by antibody staining in a variety of circulating or parenchymal murine immune subsets. Our data is in accordance with RNAseq data on Plvap expression, published by immgen.org (49, 56). Additionally, we found that neither B cells nor dendritic cells express Plvap in basal or under inflammatory (LPS stimulation) conditions. Furthermore, when we abrogate Plvap expression on B cells or in the hematopoietic compartment, there is no effect on the abundance of IgM+ B cells in the spleen or peritoneum. These findings suggest that if Plvap has a B cell trafficking/retention function in the spleen or peritoneum, this role is not B cell or hematopoietic cell intrinsic. In contrast, when we used a mouse model where the lack of Plvap was specific to endothelial cells, the defect in B cell migration/retention was indistinguishable from that observed in the global Plvap deficient mice. Furthermore, the transgenic overexpression of Plvap expression on endothelial cells in Plvap deficient mice rescued the phenotype. Together, these results demonstrate that Plvap expression in endothelial cells is key in the B cell trafficking/retention in the spleen and peritoneum. Considering the data in the literature, there are several mechanisms by which Plvap may lower the number of B-1 and MZ B cells: Plvap might hamper the generation of the B-1 B cell hematopoietic progenitors in the vascular plexus of the amniotic sac and hemogenic endothelium (16, 57, 58) where Plvap is expressed very early (18, 59). Preliminary data show that the number of B-1 B cell progenitors in the amniotic sac (Yoshimoto, Yoder, Stan unpublished) at E9 in the WT, PlvapL/L and Plvap−/− is similar. In addition, B cells precursors in the bone marrow are also similar in the Plvap deficient mice (Figure 7), which makes Plvap involvement in B cell development unlikely. Plvap controls B-1 and MZ B cell recruitment and/or retention in the peritoneum and spleen, respectively. Integrins, chemokines and other adhesion molecules have been shown to be involved in B cell migration (12, 60–62). Different types of adhesion molecules regulate distinct events in lymphocyte extravasation. We know from these studies that integrins play a role in leukocyte tethering, whereas chemokines play a role in the rolling process (63). Immature B cells express αLβ2 and α4β1, which retain B cells in the splenic marginal sinus, where they interact with their ligand, ICAM1 and VCAM1 expressed in endothelial cells (12). In addition, an increased gradient of CXCL13 in the periarterial lymphocyte sheath induces the migration of B cells from marginal zone to the follicles (64). In contrast, S1P1 expressed on B cells is able to induce the migration of B cells to the sinus of marginal zone and peritoneum (10, 11, 65). Conceivably S1P1 expressed on B cells could interact with Plvap, through vimentin, to permit the extravasation of B cell progenitors to peritoneum and the splenic marginal sinus. However, additional studies are necessary to test this hypothesis. Plvap may also be an integral mediator of inflammation-induced migration. It has been previously observed that the blocking of Plvap in vivo reduced the number of monocytes found in the peritoneum after the induction of peritonitis (31). Our results show that Plvap has a novel role in the progression of leukocyte migration to peritoneum not just in inflammation but also in the steady state. The global deletion or specific obliteration of Plvap in endothelial cells reduced the migration not just of B cells but also others leukocytes by approximately 80% in steady state. In addition, when we restore the expression of Plvap exclusively in endothelial cells, we observed the rescue of B cells number in steady state, suggesting that Plvap expression in endothelial cells is the master regulator of immune cell migration to peritoneum. Plvap−/− vessels leak plasma components leading to hypoproteinemia and formation of ascites, a condition known to reduce viability of resident peritoneal macrophages and other immune cell subsets. Deletion of Plvap may induce conditions depleting the peritoneal progenitors of B-1 and MZ B cells. Experiments involving Plvap deletion at adult stages should shed light on which mechanism(s) are involved. In conclusion, the expression of the glycoprotein Plvap on endothelial cells is a key regulator of B cells in the peritoneum, as well as, immature IgM+ B cells in the spleen. It also has a physiological role in facilitating the production of natural IgM and IgA antibodies by peritoneal and splenic IgM+ B cells. Collectively, our findings point to a novel immunologic significance for Plvap in innate humoral immunity. Future work will determine the precise mechanism by which Plvap regulates B cell subsets in peritoneum and spleen. Supplementary Material 1 The authors wish to thank Drs. T. Graf (CRG Barcelona), N. Speck (U. Penn), M. Chen (Harvard U.), Patricia A. Ernst (U. Colorado Denver) and M. Yoshimoto and M. Yoder (U. Indiana) for reagents and suggestions. Funding This work was supported by the National Institutes of Health grants GM120592, CA175592, CA172983, CA023108 and S10OD010330. Abbreviations Plvap Plasmalemma Vesicle Associated Protein MZ marginal zone PLN peripheral lymph nodes PP Peyer’s patches WT wild type S1P sphingosine-1- phosphate Figure 1 IgM+IgDlo B cells are reduced in the Plvap deficient mice. Splenocytes from wild type, heterozygous or Plvap−/− mice were analyzed for the proportion of different B cells subpopulation. A) Representative histogram of B220 expression on splenocytes. The number in the corner represents the percentage of B220+ cells. B) Representative contour plot of IgD and IgM expression in B220+ B cells. The number in the upper right corner represents the percentage of IgD+ B220+ cells, whereas the number down in the corner shows the percentage of IgM+IgD− B220+ B cells. C) Quantification of percentage (top graph) and absolute number (bottom graph) IgM+IgD− B220+ B cells in spleen from wild type, heterozygous or Plvap−/− mice. D) Representative contour plot of CD21/35 and IgM expression on B220+ B cells. The number in each corner represents the percentage of different gates which B cells were divided. E) Quantification of CD21/35loIgM+ B cells in spleen from wild type, heterozygous or Plvap−/− mice. F) Contour plots of CD23 and HSA expression on CD21/35loIgM+ B cells. G) Quantification of the percentage of T2 and T1 CD21/35loIgM+ B cells. =p<0.05. n=4 independent experiments with at least 3 Plvap−/− mice/experiment. Figure 2 B cells are absent in the peritoneum of Plvap−/− mice. A) Quantification of total number of cells in the peritoneum from wild type or Plvap−/− mice is shown in the graph. B) A representative contour plot of IgM and IgD expression on peritoneal B cells from wild type, heterozygous or Plvap−/− mice is shown. The number in each corner represents the percentage of IgD+/IgMlo (upper corner) and IgD−/IgM+ (lower corner) B cells. C) Percentage and D) Absolute number of IgD+/IgMlo (left graph) and IgM+/IgD− (right graph) B cells in peritoneum lavage from wild type, heterozygous or Plvap−/− mice. =p<0.05, *=p<0.01. n=4 independent experiments with at least 3 Plvap−/− mice/experiment. Figure 3 Plvap expression on T cells, dendritic cells and B cells. A) Gate on CD4 and CD8 T cells (Left panel), Plvap expression on CD8 (middle panel) and CD4 T cells (right panel). B) Plvap expression on dendritic cells (left panel, gated in CD11c) from mice immunized with (black line) or without (grey line) 50 ug of LPS 12 hrs before. As a control of activation of dendritic cells, CD80 expression has been shown in the right panel (Black line: immunized with LPS; grey line: non-treated). C) Plvap expression on B cells from mice immunized with (right panel) or without (left panel) 50 ug of LPS 12 hrs before. B cells were gated on B220 and CD38 double positive. D) Confocal micrographs of spleen frozen sections labeled with anti-Plvap (red), anti-B220 (blue) and CD169 (green). Figure 4 Plvap expression on the hematopoietic compartment does not affect the localization of B cells in the spleen and peritoneum compartment. A and D) A representative contour plot of IgM and IgD expression on spleen (top panel) and peritoneal (bottom panel) B cells from CD19CrexPlvapL/L (A) or VavCrexPlvapL/L (D) with correspondingly littermate controls is shown. The number at the top and bottom in each plot represents the percentage of IgD+/IgMlo and IgM+/IgD− B cells respectively. B, C, E and F) Quantification of IgD+/IgMlo (right graph) and IgM+/IgD− (left graph) B cells from spleen (B and E) and peritoneum lavage (C and F) from CD19 Cre x PlvapL/L (B and C) or Vav Cre x PlvapL/L (E and F) mice. n=3 independent experiments with at least 3 mice/experiment. Figure 5 Localization of B1 B cells in spleen and peritoneum is reduced in the lack of Plvap expression on endothelial cells. A) A representative contour plot of IgM and IgD expression on spleen B cells from Tie2CrexPlvapL/L (right panel) and littermate control mice (PlvapL/L, left panel) is shown. The number at the top and bottom in each plot represents the percentage of IgD+/IgMlo and IgM+/IgD− B cells respectively. B) Percentage of IgD+/IgMlo (left graph) and IgM+/IgD− (right graph) B cells in spleen from Tie2CrexPlvapL/L and littermate control mice. C) Absolute number of IgD+/IgMlo (left graph) and IgM+/IgD− (right graph) B cells in spleen from Tie2CrexPlvapL/L and littermate control mice. D) A representative contour plot of IgM and IgD expression on peritoneal B cells from Tie2CrexPlvapL/L (right panel) and littermate control mice (left panel) is shown. The number at the top and bottom in each plot represents the percentage of IgD+/IgMlo and IgM+/IgD− B cells respectively. E) Percentage of IgD+/IgMlo (left graph) and IgM+/IgD− (right graph) B cells from peritoneum lavage of Tie2CrexPlvapL/L and littermate control mice. F) Absolute number of IgD+/IgMlo (left graph) and IgM+/IgD− (right graph) B cells from peritoneum lavage of Tie2CrexPlvapL/L and littermate control mice. =p<0.05, =p<0.01. n=3 independent experiments with at least 3 Plvap−/− mice/experiment. Figure 6 Localization of B1 B cells in spleen and peritoneum is recovered with Plvap expression on endothelial cells. A and D) A representative contour plot of IgM and IgD expression on spleen (A) and peritoneal (D) B cells from Plvap ECRC (right panel), Plvap−/− (middle panel) and wild type (left panel) mice. The number at the top and bottom in each plot represents the percentage of IgD+/IgMlo and IgM+/IgD− B cells respectively. (B and E) Percentage of IgD+/IgMlo (left graph) and IgM+/IgD− (right graph) B cells from spleen (B) and peritoneum lavage (E) of Plvap ECRC (grey bar), Plvap−/− (white bar) and wild type (black bar) mice. (C and F) Absolute numbers of IgD+/IgMlo (left graph) and IgM+/IgD− (right graph) B cells from spleen (C) and peritoneum lavage (F) of PlvapECRC (grey bar), Plvap−/− (white bar) and wild type (black bar) mice. =p<0.01. n=3 independent experiments with at least 3 mice/group. Figure 7 Plvap does not affect the development of the B cells. Defect in the number of B cells is not due to a development issue. A) Percentage of IgM+/IgD− (left graph) and IgD+/IgMlo (right graph) B cells in spleen from PlvapL/L, PlvapiECKO and PlvapiECRC mice. B) Absolute number of IgM+/IgD− (left graph) and IgD+/IgMlo (right graph) B cells in spleen from PlvapL/L, ERTCrexPlvapL/L and ERTCrexPlvapECRC mice. C) Percentage of IgM+/IgD− (left graph) and IgD+/IgMlo (right graph) B cells from peritoneum lavage of PlvapL/L, PlvapiECKO and PlvapiECRC mice. D) Absolute number of IgM+/IgD− (left graph) and IgD+/IgMlo (right graph) B cells from peritoneum lavage of PlvapL/L, PlvapiECKO and PlvapiECRC mice. E) A representative contour plot of immature (B220+IgMhiIgD+), mature B cells (B220+IgMintIgD+) and pre-B cell (B220+IgM+IgD+) from bone marrow of PlvapL/L, PlvapiECKO and PlvapiECRCmice. The number in each plot represents the percentage of each subset. F) Percentage and G) absolute number of immature, mature and pre-B cell from bone marrow of PlvapL/L, PlvapiECKO and PlvapiECRC mice. =p<0.05, *=p<0.01. n=2 independent experiments with at least 6 mice/group/experiment. Online Supplemental Material Supplemental figure 1 shows that there is not difference in the ratio or activation markers on T cells from Plvap−/− mice. A) CD4 and CD8 T cells ratio from spleen were analyzed, a representative figure is shown for wt mice (left panel), Plvap+/− mice (middle panel) and Plvap−/− mice (right panel). Activation markers were analyzed in gated CD4 T cells (B) or CD8 T cells (C). Histogram of Plvap−/− (green), heterozygous (red) and wild type (blue) T cells are shown. Supplemental figure 2 shows the percentage of mast cells in the peritoneum cavity. Disclosure The authors of this manuscript have no conflicts of interest to disclose as described by the Journal of Immunology. 1 Pepper M Jenkins MK 2011 Origins of CD4(+) effector and central memory T cells Nat Immunol 12 467 471 21739668 2 Elgueta R V de Vries C Noelle RJ 2010 The immortality of humoral immunity Immunol Rev 236 139 150 20636814 3 Martin F Kearney JF 2001 B1 cells: similarities and differences with other B cell subsets Curr Opin Immunol 13 195 201 11228413 4 Baumgarth N 2011 The double life of a B-1 cell: self-reactivity selects for protective effector functions Nat Rev Immunol 11 34 46 21151033 5 Cerutti A Cols M Puga I 2013 Marginal zone B cells: virtues of innate-like antibody-producing lymphocytes Nat Rev Immunol 13 118 132 23348416 6 Wen L Brill-Dashoff J Shinton SA Asano M Hardy RR Hayakawa K 2005 Evidence of marginal-zone B cell-positive selection in spleen Immunity 23 297 308 16169502 7 Martin F Kearney JF 2000 Positive selection from newly formed to marginal zone B cells depends on the rate of clonal production, CD19, and btk Immunity 12 39 49 10661404 8 Cariappa A Tang M Parng C Nebelitskiy E Carroll M Georgopoulos K Pillai S 2001 The follicular versus marginal zone B lymphocyte cell fate decision is regulated by Aiolos, Btk, and CD21 Immunity 14 603 615 11371362 9 Hampel F Ehrenberg S Hojer C Draeseke A Marschall-Schroter G Kuhn R Mack B Gires O Vahl CJ Schmidt-Supprian M Strobl LJ Zimber-Strobl U 2011 CD19-independent instruction of murine marginal zone B-cell development by constitutive Notch2 signaling Blood 118 6321 6331 21795747 10 Cinamon G Matloubian M Lesneski MJ Xu Y Low C Lu T Proia RL Cyster JG 2004 Sphingosine 1-phosphate receptor 1 promotes B cell localization in the splenic marginal zone Nat Immunol 5 713 720 15184895 11 Kunisawa J Gohda M Kurashima Y Ishikawa I Higuchi M Kiyono H 2008 Sphingosine 1-phosphate-dependent trafficking of peritoneal B cells requires functional NFkappaB-inducing kinase in stromal cells Blood 111 4646 4652 18292290 12 Lu TT Cyster JG 2002 Integrin-mediated long-term B cell retention in the splenic marginal zone Science 297 409 412 12130787 13 Lalor PA Herzenberg LA Adams S Stall AM 1989 Feedback regulation of murine Ly-1 B cell development Eur J Immunol 19 507 513 2785046 14 Herzenberg LA 2000 B-1 cells: the lineage question revisited Immunol Rev 175 9 22 10933587 15 Godin IE Garcia-Porrero JA Coutinho A Dieterlen-Lievre F Marcos MA 1993 Para-aortic splanchnopleura from early mouse embryos contains B1a cell progenitors Nature 364 67 70 8316299 16 Yoshimoto M Montecino-Rodriguez E Ferkowicz MJ Porayette P Shelley WC Conway SJ Dorshkind K Yoder MC 2011 Embryonic day 9 yolk sac and intra-embryonic hemogenic endothelium independently generate a B-1 and marginal zone progenitor lacking B-2 potential Proc Natl Acad Sci U S A 108 1468 1473 21209332 17 Ansel KM Harris RB Cyster JG 2002 CXCL13 is required for B1 cell homing, natural antibody production, and body cavity immunity Immunity 16 67 76 11825566 18 Stan RV Arden KC Palade GE 2001 cDNA and protein sequence, genomic organization, and analysis of cis regulatory elements of mouse and human PLVAP genes Genomics 72 304 313 11401446 19 Stan RV Kubitza M Palade GE 1999 PV-1 is a component of the fenestral and stomatal diaphragms in fenestrated endothelia Proc Natl Acad Sci U S A 96 13203 13207 10557298 20 Hnasko R McFarland M Ben-Jonathan N 2002 Distribution and characterization of plasmalemma vesicle protein-1 in rat endocrine glands J Endocrinol 175 649 661 12475376 21 Stan RV 2004 Multiple PV1 dimers reside in the same stomatal or fenestral diaphragm Am J Physiol Heart Circ Physiol 286 H1347 1353 14630628 22 Stan RV Ghitescu L Jacobson BS Palade GE 1999 Isolation, cloning, and localization of rat PV-1, a novel endothelial caveolar protein J Cell Biol 145 1189 1198 10366592 23 Stan RV Roberts WG Predescu D Ihida K Saucan L Ghitescu L Palade GE 1997 Immunoisolation and partial characterization of endothelial plasmalemmal vesicles (caveolae) Mol Biol Cell 8 595 605 9247641 24 Stan RV Tkachenko E Niesman IR 2004 PV1 is a key structural component for the formation of the stomatal and fenestral diaphragms Mol Biol Cell 15 3615 3630 15155804 25 Herrnberger L Ebner K Junglas B Tamm ER 2012 The role of plasmalemma vesicle-associated protein (PLVAP) in endothelial cells of Schlemm’s canal and ocular capillaries Exp Eye Res 105 27 33 23063469 26 Stan RV Tse D Deharvengt SJ Smits NC Xu Y Luciano MR McGarry CL Buitendijk M Nemani KV Elgueta R Kobayashi T Shipman SL Moodie KL Daghlian CP Ernst PA Lee HK Suriawinata AA Schned AR Longnecker DS Fiering SN Noelle RJ Gimi B Shworak NW Carriere C 2012 The diaphragms of fenestrated endothelia: gatekeepers of vascular permeability and blood composition Dev Cell 23 1203 1218 23237953 27 Rantakari P Auvinen K Jappinen N Kapraali M Valtonen J Karikoski M Gerke H Iftakhar EKI Keuschnigg J Umemoto E Tohya K Miyasaka M Elima K Jalkanen S Salmi M 2015 The endothelial protein PLVAP in lymphatics controls the entry of lymphocytes and antigens into lymph nodes Nat Immunol 16 386 396 25665101 28 Elkadri A Thoeni C Deharvengt SJ Murchie R Guo C Stavropoulos JD Marshall CR Wales P Bandsma R Cutz E Roifman CM Chitayat D Avitzur Y Stan RV Muise AM 2015 Mutations in Plasmalemma Vesicle Associated Protein Result in Sieving Protein-Losing Enteropathy Characterized by Hypoproteinemia, Hypoalbuminemia, and Hypertriglyceridemia Cell Mol Gastroenterol Hepatol 1 381 394 e387 26207260 29 Herrnberger L Seitz R Kuespert S Bosl MR Fuchshofer R Tamm ER 2012 Lack of endothelial diaphragms in fenestrae and caveolae of mutant Plvap-deficient mice Histochem Cell Biol 138 709 724 22782339 30 Hallmann R Mayer DN Berg EL Broermann R Butcher EC 1995 Novel mouse endothelial cell surface marker is suppressed during differentiation of the blood brain barrier Dev Dyn 202 325 332 7626790 31 Keuschnigg J Henttinen T Auvinen K Karikoski M Salmi M Jalkanen S 2009 The prototype endothelial marker PAL-E is a leukocyte trafficking molecule Blood 114 478 484 19420356 32 Schlingemann RO Dingjan GM Emeis JJ Blok J Warnaar SO Ruiter DJ 1985 Monoclonal antibody PAL-E specific for endothelium Lab Invest 52 71 76 3880842 33 Keuschnigg J Tvorogov D Elima K Salmi M Alitalo K Salminen T Jalkanen S 2012 PV-1 is recognized by the PAL-E antibody and forms complexes with NRP-1 Blood 120 232 235 22627768 34 Niemela H Elima K Henttinen T Irjala H Salmi M Jalkanen S 2005 Molecular identification of PAL-E, a widely used endothelial-cell marker Blood 106 3405 3409 16099878 35 Stan RV 2007 Endothelial stomatal and fenestral diaphragms in normal vessels and angiogenesis J Cell Mol Med 11 621 643 17760829 36 Tse D Stan RV 2010 Morphological heterogeneity of endothelium Semin Thromb Hemost 36 236 245 20490976 37 Ioannidou S Deinhardt K Miotla J Bradley J Cheung E Samuelsson S Ng YS Shima DT 2006 An in vitro assay reveals a role for the diaphragm protein PV-1 in endothelial fenestra morphogenesis Proc Natl Acad Sci U S A 103 16770 16775 17075074 38 Tkachenko E Tse D Sideleva O Deharvengt SJ Luciano MR Xu Y McGarry CL Chidlow J Pilch PF Sessa WC Toomre DK Stan RV 2012 Caveolae, fenestrae and transendothelial channels retain PV1 on the surface of endothelial cells PLoS One 7 e32655 22403691 39 Ichimura K Stan RV Kurihara H Sakai T 2008 Glomerular endothelial cells form diaphragms during development and pathologic conditions J Am Soc Nephrol 19 1463 1471 18480313 40 Yamamoto I Horita S Takahashi T Tanabe K Fuchinoue S Teraoka S Hattori M Yamaguchi Y 2007 Glomerular expression of plasmalemmal vesicle-associated protein-1 in patients with transplant glomerulopathy Am J Transplant 7 1954 1960 17617859 41 Madden SL Cook BP Nacht M Weber WD Callahan MR Jiang Y Dufault MR Zhang X Zhang W Walter-Yohrling J Rouleau C Akmaev VR Wang CJ Cao X St Martin TB Roberts BL Teicher BA Klinger KW Stan RV Lucey B Carson-Walter EB Laterra J Walter KA 2004 Vascular gene expression in nonneoplastic and malignant brain Am J Pathol 165 601 608 15277233 42 Strickland LA Jubb AM Hongo JA Zhong F Burwick J Fu L Frantz GD Koeppen H 2005 Plasmalemmal vesicle-associated protein (PLVAP) is expressed by tumour endothelium and is upregulated by vascular endothelial growth factor-A (VEGF) J Pathol 206 466 475 15971170 43 Deharvengt SJ Tse D Sideleva O McGarry C Gunn JR Longnecker DS Carriere C Stan RV 2012 PV1 down-regulation via shRNA inhibits the growth of pancreatic adenocarcinoma xenografts J Cell Mol Med 16 2690 2700 22568538 44 Chen MJ Yokomizo T Zeigler BM Dzierzak E Speck NA 2009 Runx1 is required for the endothelial to haematopoietic cell transition but not thereafter Nature 457 887 891 19129762 45 Rickert RC Roes J Rajewsky K 1997 B lymphocyte-specific, Cre-mediated mutagenesis in mice Nucleic Acids Res 25 1317 1318 9092650 46 Stadtfeld M Graf T 2005 Assessing the role of hematopoietic plasticity for endothelial and hepatocyte development by non-invasive lineage tracing Development 132 203 213 15576407 47 Verma S Alexander CM Carlson MJ Tygrett LT Waldschmidt TJ 2008 B-cell studies in chronic ethanol mice Methods Mol Biol 447 295 323 18369926 48 Allman D Pillai S 2008 Peripheral B cell subsets Curr Opin Immunol 20 149 157 18434123 49 Heng TS Painter MW C. Immunological Genome Project 2008 The Immunological Genome Project: networks of gene expression in immune cells Nat Immunol 9 1091 1094 18800157 50 Hisatsune H Matsumura K Ogawa M Uemura A Kondo N Yamashita JK Katsuta H Nishikawa S Chiba T 2005 High level of endothelial cell-specific gene expression by a combination of the 5′ flanking region and the 5′ half of the first intron of the VE-cadherin gene Blood 105 4657 4663 15746076 51 Gothert JR Gustin SE van Eekelen JA Schmidt U Hall MA Jane SM Green AR Gottgens B Izon DJ Begley CG 2004 Genetically tagging endothelial cells in vivo: bone marrow-derived cells do not contribute to tumor endothelium Blood 104 1769 1777 15187022 52 Gothert JR Gustin SE Hall MA Green AR Gottgens B Izon DJ Begley CG 2005 In vivo fate-tracing studies using the Scl stem cell enhancer: embryonic hematopoietic stem cells significantly contribute to adult hematopoiesis Blood 105 2724 2732 15598809 53 Elgueta R Marks E Nowak E Menezes S Benson M Raman VS Ortiz C O’Connell S Hess H Lord GM Noelle R 2015 CCR6-dependent positioning of memory B cells is essential for their ability to mount a recall response to antigen J Immunol 194 505 513 25505290 54 Sanderson RD Lalor P Bernfield M 1989 B lymphocytes express and lose syndecan at specific stages of differentiation Cell Regul 1 27 35 2519615 55 Umemoto E Otani K Ikeno T Verjan Garcia N Hayasaka H Bai Z Jang MH Tanaka T Nagasawa T Ueda K Miyasaka M 2012 Constitutive plasmacytoid dendritic cell migration to the splenic white pulp is cooperatively regulated by CCR7- and CXCR4-mediated signaling J Immunol 189 191 199 22634622 56 TIGP 2016 The Immunological Genome Project http://www.immgen.org/databrowser/index.html The Immunological Genome Project Last accessed 9/1/2016 57 Pieper K Grimbacher B Eibel H 2013 B-cell biology and development J Allergy Clin Immunol 131 959 971 23465663 58 Yoshimoto M 2015 The first wave of B lymphopoiesis develops independently of stem cells in the murine embryo Ann N Y Acad Sci 59 Takase H Matsumoto K Yamadera R Kubota Y Otsu A Suzuki R Ishitobi H Mochizuki H Kojima T Takano S Uchida K Takahashi S Ema M 2012 Genome-wide identification of endothelial cell-enriched genes in the mouse embryo Blood 120 914 923 22535667 60 Legler DF Loetscher M Roos RS Clark-Lewis I Baggiolini M Moser B 1998 B cell-attracting chemokine 1, a human CXC chemokine expressed in lymphoid tissues, selectively attracts B lymphocytes via BLR1/CXCR5 J Exp Med 187 655 660 9463416 61 Kunisawa J Kurashima Y Gohda M Higuchi M Ishikawa I Miura F Ogahara I Kiyono H 2007 Sphingosine 1-phosphate regulates peritoneal B-cell trafficking for subsequent intestinal IgA production Blood 109 3749 3756 17234743 62 Green JA Suzuki K Cho B Willison LD Palmer D Allen CD Schmidt TH Xu Y Proia RL Coughlin SR Cyster JG 2011 The sphingosine 1-phosphate receptor S1P(2) maintains the homeostasis of germinal center B cells and promotes niche confinement Nat Immunol 12 672 680 21642988 63 Bono MR Elgueta R Sauma D Pino K Osorio F Michea P Fierro A Rosemblatt M 2007 The essential role of chemokines in the selective regulation of lymphocyte homing Cytokine Growth Factor Rev 18 33 43 17324605 64 Bowman EP Campbell JJ Soler D Dong Z Manlongat N Picarella D Hardy RR Butcher EC 2000 Developmental switches in chemokine response profiles during B cell differentiation and maturation J Exp Med 191 1303 1318 10770798 65 Kunisawa J Kurashima Y Higuchi M Gohda M Ishikawa I Ogahara I Kim N Shimizu M Kiyono H 2007 Sphingosine 1-phosphate dependence in the regulation of lymphocyte trafficking to the gut epithelium J Exp Med 204 2335 2348 17875673
PMC005xxxxxx/PMC5120878.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101465514 34788 Mol Plant Mol Plant Molecular plant 1674-2052 1752-9867 25578268 5120878 10.1016/j.molp.2014.12.002 NIHMS654301 Article The application of synthetic biology to elucidation of plant mono-, sesqui- and di-terpenoid metabolism Kitaoka Naoki a1 Lu Xuan a Yang Bing b Peters Reuben J. a2 a Department of Biochemistry, Biophysics & Molecular Biology, Iowa State University, Ames, IA 50011, U.S.A b Department of Genetics, Developmental & Cell Biology, Iowa State University, Ames, IA 50011, U.S.A 2 To whom correspondence should be addressed: [email protected], tel. 1-515-294-8580, fax 1-515-294-0453 1 Current address: Tohoku University, Sendai 980-8577, JAPAN 18 11 2016 11 12 2014 1 2015 23 11 2016 8 1 616 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Plants synthesize a huge variety of terpenoid natural products, including photosynthetic pigments, signaling molecules and defensive substances. These are often produced as complex mixtures, presumably shaped by selective pressure over evolutionary timescales, some of which have been found to have pharmaceutical and other industrial uses. Elucidation of the relevant biosynthetic pathways can provide increased access (e.g., via molecular breeding or metabolic engineering), and enable reverse genetic approaches towards understanding the physiological role of these natural products in plants as well. While such information can be obtained via a variety of approaches, this review describes the emerging use of synthetic biology to recombinantly reconstitute plant terpenoid biosynthetic pathways in heterologous host organisms as a functional discovery tool, with a particular focus on incorporation of the historically problematic cytochrome P450 mono-oxygenases. Also falling under the synthetic biology rubric and discussed here is the nascent application of genome-editing tools to probe physiological function. metabolic engineering cytochromes P450 terpene synthase reverse genetics Introduction Plants are a prolific source of natural products, with their complex chemical mixtures reflecting adaptation to previous and current selective pressures over evolutionary timescales. From origins in central metabolism, often via initial derivation from hormone biosynthetic pathways, natural products have diversified into the astounding array of extant compounds, only a fraction of which are currently known. The exact physiological roles played by these small molecules are then presumably similarly diverse. However, our understanding of physiological function lags well behind even our efforts to catalog chemical diversity, almost invariably relying on the biological activity exhibited in laboratory tests by isolated compounds. Nevertheless, plant natural products have been utilized for millennia as pharmaceuticals, flavors and fragrances, along with other industrial uses. Not surprisingly, investigation of plant natural products biosynthesis is largely focused on those compounds with economic significance (i.e., industrial use) or of obvious physiological importance (e.g., hormones). These studies have progressed from biochemical assays with cell-free extracts and isolation of the relevant enzymes to molecular biology based recombinant expression and analysis of plants genetically modified to over or under express the encoding genes. However, outside of hormones, the biosynthesis of very few plant natural products has been fully elucidated. Identification of the relevant biosynthetic pathway enables access to the resulting natural product, via metabolic engineering in the native plant or in microbial hosts. The use of microbes as production platforms provided one the early examples of the use of the term synthetic biology due in large part to the substantial efforts required to optimize production levels. Such application of synthetic biology has been extensively reviewed elsewhere (Keasling, 2012; Keasling et al., 2012). Here is reviewed the use of this synthetic biology approach to investigate plant metabolism (i.e., pathways/ networks), as well as the use of synthetic biology-enabled genome-editing tools to investigate physiological function. Terpenoid metabolism More specifically, this review focuses on terpenoid metabolism, which has been extensively diversified in plants, to the point that these represent the largest class of natural products with over 45,000 known examples (Renault et al., 2014). Terpenoids are derived from 5-carbon building blocks, namely isopentenyl diphosphate (IPP) and dimethyl allyl diphosphate (DMAPP). These building blocks are produced by two distinct pathways, the mevalonate (MVA) dependent pathway that operates in the cytosol of plants, and the 2-C-methyl-D-erythritol-4-phosphate (MEP) dependent pathway found in the plastids (Vranova et al., 2013). DMAPP and IPP are coupled together in a head-to-tail fashion to form the precursor to the 10-carbon monoterpenoids, in turn this is coupled to another molecule of IPP to form the precursor to the 15-carbon sesquiterpenoids, with coupling of yet another molecule of IPP yielding the precursor to the 20-carbon diterpenoids. These precursors take the form of allylic diphosphate containing acyclic chains, generally with internal double bonds of trans configuration, although the use of analogous cisoid precursors also has been reported (Bohlmann and Gershenzon, 2009). Formation of these elongated precursors is catalyzed by isoprenyl diphosphate synthases. Two of the transoid C15 farnesyl diphosphate can be condensed together in a head-to-head manner to form the squalene precursor to the triterpenoids, and two of the C20 geranylgeranyl diphosphate (GGPP) similarly condensed to form the phytoene precursor to the tetraterpenoids, although neither of these types of terpenoids will be further discussed here as at least the triterpenoids have been recently reviewed (Moses et al., 2013). The allylic diphosphate ester bond of the acyclic prenyl diphosphate precursors for mono-, sesqui- and di-terpenoids is lysed, with the resulting carbocation used to drive cyclization and/or rearrangement of the hydrocarbon chain in reactions catalyzed by (class I) terpene synthases (TPSs). In the case of diterpenoids, this can be preceded by an initial (bi)cyclization of the general precursor GGPP, typically forming copalyl diphosphate (CPP), in reactions catalyzed by CPP synthases (CPSs). In the course of the catalyzed reaction TPSs and CPSs can add water to produce oxygenated terpenes. However, the resulting compounds are most often olefins (Gao et al., 2012). Given the hydrophobic nature of the hydrocarbon terpene products of TPSs, the addition of oxygen is almost invariably catalyzed by endoplasmic reticulum membrane-associated (microsomal) cytochromes P450 (CYPs). These heme-thiolate mono-oxygenases often catalyze the insertion of oxygen into carbon-hydrogen bonds to produce hydroxyl groups, although they can mediate more complex reactions as well (Guengerich and Munro, 2013). In any case, the addition of oxygen both increases polarity/solubility and imparts hydrogen-bonding capacity, enabling specific binding interactions, as well as providing functional groups for further modification. The extent of terpenoid natural product diversity in plants perhaps can be best appreciated by noting the number of TPSs found in the known genomes, with all spermatophytes containing at least 40 such genes, with over 100 observed in some species (Chen et al., 2011). Although all of genes these may not encode functional TPSs, and there is some overlap in catalyzed product outcome (and/or outright genetic redundancy), this is balanced by the propensity for these enzymes to yield a range of products. In addition, it has been noted that the CYPs form the largest family of plant metabolic enzymes, with over 200 found in all spermatophytes (Nelson and Werck-Reichhart, 2011), and many of these CYPs participate in terpenoid metabolism (Hamberger and Bak, 2013). Thus, even without consideration of the other classes of enzymes that participate in terpenoid biosynthesis, it can be appreciated that terpenoid metabolism forms a complex network in any given plant! However, very few pathways are known, and there is no example of a completely elucidated network, even at just the level of the TPSs – e.g., not even for the model plant Arabidopsis thaliana. Accordingly, it also has not been possible to fully determine the full range of physiological roles played by the various terpenoid natural products made by any one species. Our understanding of plant terpenoid metabolism has been limited by a number of factors. These include the perceived recalcitrance of eukaryotic enzymes to functional recombinant expression in the usual hosts – particularly for membrane-associated CYPs – as well as the difficulty of identifying and obtaining the correct intermediate/substrate for the enzyme in question. In addition, the typical lack of functional gene clustering in plants dictates a general need to laboriously individually identify each enzyme, unlike the situation in microbes where such clustering often translates to identification of any relevant enzyme leading to the complete biosynthetic pathway, which is mediated by the enzymes encoded by the neighboring genes. The difficulty in identification of plant biosynthetic enzymes is increased by the relative lack of plant sequence information – e.g., high-quality genomes and/or transcriptomes – although the amount of such information is currently rapidly increasing. Elucidation of menthol biosynthesis: A monoterpenoid case study With regards to pathway elucidation it is perhaps most instructive to examine the case of the monoterpene menthol, whose biosynthesis was worked out over the course of the stellar career of Prof. Rodney Croteau at Washington State University (Figure 1). Enabled by the localization of methol production to protruding glandular trichomes that could be physically isolated (Gershenzon et al., 1992), this work began with enzymatic assays with plant cell-free extracts, progressed to molecular genetic identification of the relevant biosynthetic enzymes (for both on- and off-pathway reactions), leading to genetic/metabolic engineering in peppermint (Mentha x piperita) plants (Croteau et al., 2005). The full elucidation of the relevant pathway and (relatively few) branch points enabled a systems biology approach to understanding of biochemical regulation (Rios-Estepa et al., 2008), and comprehensive mathematical modeling guided evaluation of the pathway (Rios-Estepa et al., 2010). Notably, in part based on these advances, mint plants have been engineered for higher and more reliable yield of menthol, as demonstrated in multi-year field trials (Lange et al., 2011). As part of this work, the Croteau group was able to easily recombinantly express the relevant mint TPS, 4S-limonene synthase, in Escherichia coli (Colby et al., 1993), with particularly strong activity observed upon truncation of the N-terminal plastid targeting sequence (Williams et al., 1998). However, functional recombinant expression for identification of the initial acting peppermint CYP, limonene-3-hydroxylase (CYP71D13), was carried out using a cumbersome insect cell culture system (Lupien et al., 1999), in large part due to the presumed similarity in membrane composition and presence of the requisite CYP reductase (CPR), which is not present in bacteria. Later work then demonstrated that it was possible to also functionally express N-terminally modified versions of CYP71D13 in Saccharomyces cerevisiae (yeast) or E. coli, with activity observed upon reconstitution with a plant CPR (Haudenschild et al., 2000). Identification of these enzymes offered the possibility for pathway reconstitution in a heterologous host (micro)organism. Such metabolic engineering was attempted in E. coli, with some production of limonene observed, but flux through even the first CYP step was not observed (Carter et al., 2003). Notably, feeding studies indicated that this lack of flux was due to insufficient production of the upstream terpenoids (and possibly also the volatile nature of limonene), rather than lack of CYP activity per se. Nevertheless, functional expression of eukaryotic microsomal CYPs in E. coli is traditionally considered problematic, and most studies use yeast if not insect cells or, in the case of plant CYPs, plant expression systems. Increasing flux: Production of artemisinic acid Even as this work was taking place, the issue of flux towards terpenoid metabolism was being addressed by the group of Prof. Jay Keasling at the Univ. of California at Berkeley. Although the complete biosynthetic process underlying production of the important antimalarial sesquiterpene artemisinin was and still remains to this day unknown (Brown, 2010), the Keasling group had received substantial funding to metabolically engineer microbial production of the intermediate artemisinic acid, from which artemisinin can be efficiently made (Paddon and Keasling, 2014). Focusing on production of the olefin intermediate, amorphadiene, for which the relevant TPS had already been identified, the Keasling group imported the entire MEV pathway from S. cerevisiae into E. coli (Figure 2), dramatically improving terpenoid production levels (Martin et al., 2003). Notably, this work also featured one of the first examples of the use of a synthetic gene, recoded to optimize codon usage for expression in E. coli. Later reported work led to even more dramatic increases in yield, with production levels reported to be over 27 g/L of amorphadiene, and application of the term synthetic biology to this work (Keasling, 2012). The development of such highly engineered microbial production systems obviously provides sufficient substrate for subsequently acting enzymes, either via isolation and re-feeding or additional incorporation of the enzyme into the metabolically engineered microorganism. Of particular interest here, the latter approach – extended metabolic engineering, albeit carried out in yeast rather than E. coli – was used for functional identification of the CYP71AV1 relevant to artemisinin biosynthesis, which is capable of catalyzing the multiple reactions necessary to convert amorphadiene to artemisinic acid (Ro et al., 2006). Indeed, yeast was chosen as the microbial host platform for further development by Amyris, the company founded by Keasling and co-workers to commercialize such engineering of high-level terpene production. In addition to extensive modifications of the endogenous metabolic network of yeast, it is of interest here to note that commercial production of artemisinic acid (reported to be 25 g/L) also was dependent on incorporation of the additional alcohol and aldehyde dehydrogenases involved in the plant biosynthetic pathway (Paddon et al., 2013), which were identified from work carried out in the group of Dr. Pat Covello at the National Research Council of Canada, Saskatoon (Teoh et al., 2009; Teoh et al., 2006). Engineering yeast to elucidate CYP roles in terpenoid metabolism The use of metabolic engineering of yeast for functional characterization of CYPs involved in plant terpenoid biosynthesis was pioneered by Dr. Dae-Kyun Ro. First, working with Prof. Joerg Bohlmann at the Univ. of British Columbia, where this approach was used to at least confirm the functional identification of CYP724B1 as a promiscuous diterpene oxidase involved in diterpene resin acid biosynthesis in loblolly pine, which is capable of catalyzing the multiple reactions necessary to convert a methyl to carboxylic acid (Ro et al., 2005). He then went on to the Keasling group, and was a key player in their identification of amorphadiene oxidase (Ro et al., 2006). Now a professor at the Univ. of Calgary, his group has used this approach to functionally identify the CYPs involved in sesquiterpene lactone biosynthesis from plant species in the Asteraceae family, again multiply-reactive, methyl to carboxylic acid converting oxidases that also are members of the CYP71AV sub-family (Nguyen et al., 2010), and subsequently acting hydroxylases from the CYP71BL sub-family, with presumably spontaneous formation of the lactone ring occurring upon hydroxylation at appropriately spaced locations (Ikezawa et al., 2011). Prof. Bohlmann’s group also has applied this approach to functional identification of the spruce CYP720B4 involved in diterpene resin acid biosynthesis (Hamberger et al., 2011), and the sandalwood CYP71F sub-family members involved in production of the essential oil sesquiterpene components santalols and bergamotol (Diaz-Chavez et al., 2013). Emerging alternative expression systems Part of the rationale for the use of yeast in such work is the ability to functionally express unmodified plant microsomal CYPs. Another system attracting use for this reason is transient transformation in tobacco, Nicotiana benthamiana (Bach et al., 2014a). In addition, a particularly promising system for investigation of terpenoid biosynthesis is the moss Physcomitrella patens (Bach et al., 2014b). However, there do not appear to be any CYPs from terpenoid biosynthesis identified using these systems as of yet. Moreover, at least with the tobacco system production of terpenoids can be plagued by interference from endogenous metabolism – e.g., the introduction of artemisinic acid biosynthesis led to glycosylated compound (van Herpen et al., 2010) – which presumably would complicate analysis of novel enzymatic activity. Data deluge: Increasing availability of plant sequence information The use of metabolic engineering to investigate biosynthetic pathways is of particular interest when genes encoding enzymes of potential relevance are readily available. Such information is being provided by next-generation sequencing based approaches, such as the RNA-Seq methodology that has been applied over the last few years by several consortia to medicinal plants (Giddings et al., 2011; Gongora-Castillo et al., 2012; Marques et al., 2013; Xiao et al., 2013; Zerbe et al., 2013), with explicit incorporation of the use of metabolic engineering to at least functionally characterize the encoded enzymes by the PhytoMedSyn group in Canada (Facchini et al., 2012). Even beyond these medicinal plant transcriptomic sequence based investigations, which were largely targeted at natural products of pharmaceutical interest, with the availability of plant genome sequences we now have essentially complete gene inventories that can be used to probe the encoded natural product metabolic networks. Of particular interest here, rice (Oryza sativa) provides a long-standing example that couples a high-quality genome with extensive cDNA sequence information (International Rice Genome Sequencing, 2005; Kikuchi et al., 2003). Rice as a model system for investigating diterpenoid metabolism Intrigued by the presence of several families of diterpene phytoalexins in rice, we have undertaken a functional genomics approach towards investigating the relevant biosynthetic network. Although a diterpene phytoalexin derived from casbene, which is formed by direct TPS catalyzed cyclization of GGPP, has been recently reported (Inoue et al., 2013), the remainder require initial cyclization by a CPS, with subsequent cyclization and/or rearrangement catalyzed by TPSs. These fall into the TPS-e sub-family related to the ent-kaurene synthase (KS) required for gibberellin hormone biosynthesis (Chen et al., 2011), which we refer to as KS-like (KSLs). The resulting phytoalexins then fall into the labdane-related diterpenoid (LRD) super-family (Peters, 2010). Since only the LRD phytoalexins were originally known, we focused on biochemical characterization of the corresponding rice diterpene synthases – i.e., the OsCPSs and OsKSLs. Our work was carried out at the same time as similar studies by groups led by Profs. Hisakazu Yamane at the Univ. of Tokyo and Tomonobu Toyomasu at Yamagata Univ., all of which relied on biochemical in vitro assays and the use of authentic standards for the expected diterpene olefin intermediates, almost all of which were provided by Prof. Robert M. Coates from the Univ. of Illinois. The results of these studies have been extensively reviewed elsewhere (Peters, 2006; Schmelz et al., 2014; Toyomasu, 2008; Yamane, 2013), and here we simply note that all the functional OsCPSs and OsKSLs have been characterized, with rice found to encode a diterpene metabolic network even more complex than originally expected (Figure 3). The complex nature of rice diterpene metabolism was evident to some degree simply from the number of OsKLS, which exceeded that required for the known families of phytoalexins – each of which is defined on the basis of their derivation from a common hydrocarbon backbone formed by the combined activity of a CPS and KSL. Specifically, seven functional OsKS(L) are found in the rice genome, although only five were expected (this includes the OsKS required for gibberellin biosynthesis; see Figure 3). In our own work, we found an OsKSL whose product did not match any of the olefin intermediates for the known LRDs, forcing an almost year-long search for corresponding authentic standard to verify its production of syn-stemodene (Morrone et al., 2006). Development of a modular metabolic engineering system in E. coli Both daunted by this experience and inspired by the complexity of rice diterpene biosynthesis, we developed a modular metabolic engineering system to co-express any pairing of a CPS and KS(L), along with the necessary upstream GGPP synthase in E. coli (Cyr et al., 2007). In order to produce enough of the resulting diterpene(s) for de novo structural analysis (e.g., by NMR) we further extended this system with either increased flux to terpenoids via either overexpression of key enzymes from the endogenous MEP pathway, or importation of the yeast MEV pathway using vectors kindly shared with us by Prof. Keasling (Morrone et al., 2010b). This system has proven utility for functional characterization of novel diterpene synthases from various plants (Gao et al., 2009; Jackson et al., 2014; Wu et al., 2012; Zhou et al., 2012), as well as microbes (Hershey et al., 2014; Lu et al., submitted; Morrone et al., 2009; Xu et al., submitted), along with mutational analysis (Criswell et al., 2012; Morrone et al., 2008; Potter et al., 2014). Even beyond such de novo characterization of functionally novel diterpene synthases, this plug-and-play approach enables examination of enzymatic substrate specificity, which revealed further complexity/range for rice diterpene metabolism. One of the OsKSLs uniquely reacts with both stereoisomers of CPP (ent and syn) produced by rice, with both resulting products found in planta. Moreover, two other OsKSLs that act on syn-CPP in rice also will react with CPP of normal stereochemistry (Morrone et al., 2011). This promiscuity may be related to the evolution of alternative stereoisomers of CPP (i.e., other than the ent-CPP required for gibberellin biosynthesis) in the cereals, as the wheat ortholog of the rice syn-CPP producing CPS produces normal CPP instead (Wu et al., 2012), and several wheat KSLs similarly react with both the endogenous (normal) CPP as well as syn-CPP, which does not seem to be produced by wheat (Zhou et al., 2012). Accordingly, the ability to easily interrogate diterpene synthase activity via this modular metabolic engineering system offers insights not only into existing metabolism, but also (through examination of related plants) the evolution of diterpene biosynthesis as well. Investigation of rice CYPs in E. coli: Enabled by gene synthesis Fortuitously, in the course of characterizing the rice CPSs and KSLs, we noted that some of the genes encoding sequentially acting CPS and KSL were close together in the rice genome, one pair on chromosome 4 and others grouped together on chromosome 2, which seemed to define diterpenoid biosynthesis gene clusters (Prisic et al., 2004; Wilderman et al., 2004). Indeed, in a report on rice gibberellin metabolism, these regions had been previously noted to also contain genes encoding a number of CYPs, as well as short-chain alcohol dehydrogenases (Sakamoto et al., 2004). Many of these genes were found to be co-regulated, specifically inducible by the fungal cell wall component chitin, as well (Okada et al., 2007). Subsequently, genetic (RNAi) evidence was reported indicating a role for the CYPs from the chromosome 4 cluster in momilactone diterpenoid metabolism (Shimura et al., 2007), consistent with the role of the co-clustered OsCPS4 and OsKSL4 [Figure 3 (Wilderman et al., 2004)]. This information was particularly useful in enabling focused investigation of CYPs – i.e., rather than investigating all >350 CYP found in the rice genome, we then simply targeted those from the diterpenoid biosynthesis clusters, along with the few others that exhibited inducible transcription (15 total). Based on our ability to functionally express the CYP kaurene oxidase required for gibberellin biosynthesis from A. thaliana (CYP701A3) in E. coli (Morrone et al., 2010a), we attempted to incorporate rice CYPs into our metabolic engineering system. In an initial study, we were able to do so using an N-terminally modified version of the native gene for one of these, CYP76M7, which acts on the ent-cassadiene product of the co-clustered OsCPS2 and OsKSL7, carrying out C11α-hydroxylation that is consistent with a role in phytocassane biosynthesis (Swaminathan et al., 2009). However, no activity was observed for the other CYPs with this approach. Knowing that at least one of the two closely related CYPs found in the cluster on chromosome 4 (i.e., CYP99A2 and 3) played a role in momilactone biosynthesis (Shimura et al., 2007), we turned to use of synthetic genes optimized for expression in E. coli, along with N-terminal modification. This led to functional expression and successful incorporation into our metabolic engineering system of at least one of these, CYP99A3, which acts on the syn-pimaradiene product of the co-clustered OsCPS4 and OsKSL4, carrying out conversion of C19 from a methyl to carboxylic acid (Wang et al., 2011). Such use of synthetic gene constructs has since led to functional characterization of the ability of a number of rice CYPS to act on the diterpene olefin products of the OsCPSs and OsKSLs (Wang et al., 2012a; Wang et al., 2012b; Wu et al., 2011), with the further optimization of this synthetic biology approach reported in the accompanying paper enabling functional expression of even previously recalcitrant CYPs (Kitaoka et al., submitted). Based on the observed activity with olefins, it is clear that several CYPs will react with the same diterpene olefin intermediate (Figure 4). Thus, it is clear that some CYPs will act downstream (i.e., sequentially). This can be verified in part by incorporation of multiple CYPs into the metabolic engineering system, which we have demonstrated for at least two consecutively acting CYPs from rice oryzalexin metabolism (Figure 4C). However, such analysis leaves the order of reaction undefined. Defining pathway then requires isolation and refeeding experiments, which we have used not only in rice (Wu et al., 2013), but also in our studies of diterpenoid metabolism from other plants as well (Zi and Peters, 2013). Notably, use of this synthetic biology approach also enables facile access to such putative biosynthetic intermediates as well as functional recombinant expression of multiple plant CYPs. Other groups have independently demonstrated the ability to functionally incorporate plant microsomal CYPs for the purpose of reconstituting terpenoid biosynthetic pathways in E. coli, most notably for the production of early intermediates from the biosynthetic pathway leading to the anti-cancer drug taxol (Ajikumar et al., 2010). Although this study was not directed at identification of novel CYPs, instead building on previous work from the Croteau group (Jennewein et al., 2004), it emphasizes the continuing exploration of synthetic biology approaches in E. coli for production of high-value terpenoid natural products. Investigation of physiological function via synthetic biology Given the paucity of information, investigation of the physiological role played by most natural products seems well worth pursuit. With molecular identification of biosynthetic pathways one approach might involve transplantation into heterologous host plants, with subsequent investigation of physiological effect (i.e., phenotype). Such an approach might be useful for investigation of evolution as well. For example, the change in CPP stereochemistry (and that of the derived diterpenoid natural products) between rice and wheat might be probed by expression of the CPS producing the alternative stereoisomer and subsequent analysis of the array of resulting diterpenoids (to probe metabolic promiscuity) as well as physiological effect. Particularly given that the biosynthesis of natural products is often embedded in more complex metabolic networks, definitive investigations rely on reverse genetic analysis. For example, this can be used to verify enzymatic roles in biosynthetic pathways assigned on the basis of biochemical activity observed with recombinantly expressed genes. Arguably more importantly, this further enables investigation of the physiological function of the resulting natural products, which is particularly critical given the complex mixtures of compounds generally produced by plants that confound functions ascribed solely on the basis of in vitro analyses of natural product biological activity. For example, despite being the first rice diterpenoids suggested to act as phytoalexins against the devastating fungal blast pathogen Magneportha oryzae (Cartwright et al., 1977), the momilactones seem to be more important as allelochemicals that are constitutively secreted from the roots and suppress the growth of other plant species (Xu et al., 2012). This is consistent with the original isolation of the momilactones as plant growth inhibitors (Kato et al., 1973), and more recent work on rice allelopathy (Kato-Noguchi and Peters, 2013), although it should be noted that some evidence has been presented indicating a role for the momilactones (or related OsCPS4-dependent diterpenoids) as phytoalexins (Toyomasu et al., 2013). Although RNAi can be used to knock-down expression of the relevant genes, this approach often leads to incomplete suppression of natural product biosynthesis and can be somewhat non-selective (i.e., affecting more than the targeted gene), confounding interpretation of studies with the resulting plants. However, identification of gene knockout mutants has relied on laborious screening of large-scale mutagenesis projects (generally insertional), and only for Arabidopsis have such efforts led to a (almost) complete set of mutants. Even in rice there are many genes for which no mutant is yet available, and those that are available are found in a variety of different genetic backgrounds, complicating comparative analysis. Accordingly, the recent development of targetable genome-editing tools, particularly the modular nature of the recently developed transcription activator-like nucleases (TALENs) and clustered regularly-interspaced short palindromic repeats (CRISPR)/Cas9 based technologies (Christian et al., 2010; Jinek et al., 2012; Li et al., 2011), enables interrogation of physiological functional in native biological context. Notably, judicious construction of a construct targeting any given gene (or multiple genes or even gene cluster) allows both construction of all related mutants in same genetic background (e.g., cultivar), and identical mutants in different backgrounds (depending on conservation of the targeted sequence, perhaps even in closely related species), enabling relatively sophisticated analyses. Such genome editing has been successfully carried out with regeneration of whole plants that stably carry the modified genes over multiple generations (Feng et al., 2013; Miao et al., 2013; Zhang et al., 2014; Zhou et al., 2014), and we anticipate that such an approach will be applied to investigation of the physiological roles of plant natural products in the near future. CONCLUSIONS Traditional approaches to plant natural products generally involved isolation on the basis of targeted biological activity with subsequent investigation of the underlying biosynthetic pathway via purification of the relevant enzymes from the native producing plant. With the rapidly increasing availability of plant sequence information, both genomic and transcriptomic, it is possible to more systematically study both biosynthesis and physiological function. Here we have reviewed the application of synthetic biology approaches to such investigation of plant terpenoid natural products. Key developments for study of plant terpenoid biosynthesis are derived from the work of Prof. Keasling on production of artemisinin via semi-synthesis from artemisinic acid, demonstrating the use of synthetic biology for commercial purposes. As reviewed here, this has further led to the use of synthetic biology as a functional discovery tool for elucidation of plant terpenoid metabolism more generally, including investigation of biosynthetic networks and their evolution. In addition, we suggest the potential for application of synthetic biology-enabled genome-editing tools to systematic reverse genetic studies of physiological function. FUNDING Our work on rice diterpenoid natural products has been funded by the USDA-NIFA (most recently grant 2014-67013-21720 to R.J.P. and B.Y.), while our work on diterpene synthases more generally is funded by the NIH (grant GM076324 to R.J.P.). We apologize to colleagues whose work was not cited due to space constraints. No conflict of interest declared. Figure 1 Menthol biosynthetic pathway from glandular trichomes of peppermint The inset depicts a glandular trichome on the leaf surface, and schematic of the secretory cell isolation process. The isoprenyl diphosphate precursors IPP and DMAPP are produced via the plastidial MEP pathway, with subsequent more specific enzymatic steps individually shown. DMAPP is condensed with IPP to form geranyl diphosphate (GPP) by a GPP synthase (GPS), which is then cyclized by a TPS, limonene synthase (LS). The resulting (−)-limonene is transformed to (−)-t rans-isopiperitenol by limonene-3-hydroxylase (L3OH), with the 3α-hydroxyl group oxidized to the ketone of (−)-isopiperitenone by isopiperitenol dehydrogenase (iPD). The endo-cyclic double bond is stereospecifically reduced to (+)-cis-isopulegone by isopiperitenone reductase (iPR), with subsequent isomerization of the exo-cyclic double-bond to (+)-pulegone by isopulegone isomerase (iPI). Pulegone is a branchpoint metabolite, which can either be transformed to the undesirable by-product (+)-menthofuran by menthofuran synthase (MFS) or reduced to the intermediate (+)-menthone by pulegone reductase (PR), which is then reduced to the desired (+)-menthol end-product by menthone reductase (MR). Figure 2 Schematic for production of artemisinic acid by E. coli Genes for the MVA pathway, to produce IPP from acetyl-CoA via mevalonate, were imported from S. cerevisiae, along with those encoding an isopentenyl diphosphate isomerase and farnesyl diphosphate synthase. The resulting farnesy diphosphate (FPP) is then converted to artemisinic acid using enzymatic genes from the native producing plant Artemisia annua. First FPP is cyclized by a TPS, amorphadiene synthase (AS), and then amorphadiene is hydroxylated by CYP71AV1 (in conjunction with a CPR), which can further oxidize the resulting artemisinol to arteminsinal and then artemisinic acid, although this is much more efficiently catalyzed by artemisinol and artemisinal dehydrogenases (ADH and ALDH, respectively). Isolation of the artemisinic acid “end” product is followed by chemical conversion to the anti-malarial drug artemisinin. Figure 3 Metabolic map of the rice diterpene synthases Shown are the functional diterpene synthases, both CPSs and KS(L)s, from rice (Oryza sativa), alongside the reaction catalyzed by each of these OsCPS and OsKS(L) (thicker arrows indicate the reactions/enzymes involved in gibberellin (GA) phytohormone biosynthesis rather than more specialized secondary metabolism). Also shown are the derived natural products, where known. Notably, the metabolic fate of several of the resulting diterpene olefins, which can be found in planta, is not yet known (as indicated by “???”). Figure 4 Characterized CYPs from rice diterpenoid biosynthesis Depicted are reactions catalyzed by the noted rice CYPs with endogenous diterpene olefins. A) CYP71Z6 catalyzes C2α- and then C3α-hydroxylation of ent-isokaurene. These reactions may be relevant to the production of oryzadione and related diterpenoids (Kitaoka et al., submitted). B) With ent-cassadiene alternative reactions are catalyzed by CYP701A8 [C3α-hydroxylation (Wang et al., 2012b)], CYP76M7 & 8 [both catalyze C11α-hydroxylation (Wang et al., 2012a)], or CYP71Z7 [C2α-hydroxylation followed by further oxidation to a C2-keto and additional C3α-hydroxylation (Kitaoka et al., submitted)]. The C11α-hydroxylation reaction catalyzed by CYP76M7 & 8 is relevant to phytocassane biosynthesis (Wang et al., 2012a), while those catalyzed by CYP71Z7 and (potentially) CYP701A8 may be relevant as well. C) While CYP76M5, 6 & 8 all catalyze C7β-hydroxylation of ent-sandaracopimaradiene, it seems likely that this olefin first undergoes C3α-hydroxylation catalyzed by CYP701A8, with subsequent hydroxylation at C7β catalyzed by CYP76M8 forming oryzalexin D and alternative C9β-hydroxylation catalyzed by CYP76M6 to form oryzalexin E (Wu et al., 2013). D) Both CYP99A2 & 3 can convert C19 of syn-pimaradiene to a carboxylic acid, although this olefin can be alternatively hydroxylated at C3β by CYP701A8 or C7β by CYP76M8. The C19 oxidation catalyzed by CYP99A2 & 3 is relevant to momilactone biosynthesis (Shimura et al., 2007), while those catalyzed by CYP701A8 and CYP76M8 may be relevant as well (Kitaoka et al., submitted). Summary Synthetic biology offers a means to explore plant metabolism, such as the smaller terpenoid natural products reviewed here. This includes both investigation of biosynthetic pathways via reconstruction in heterologous host (micro)organisms and the nascent application of genome-editing tools to enable reverse genetic studies of physiological function. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. Ajikumar PK Xiao WH Tyo KE Wang Y Simeon F Leonard E Mucha O Phon TH Pfeifer B Stephanopoulos G 2010 Isoprenoid pathway optimization for Taxol precursor overproduction in Escherichia coli Science 330 70 74 20929806 Bach SS Bassard JE Andersen-Ranberg J Moldrup ME Simonsen HT Hamberger B 2014a High-throughput testing of terpenoid biosynthesis candidate genes using transient expression in Nicotiana benthamiana Methods Mol Biol 1153 245 255 24777803 Bach SS King BC Zhan X Simonsen HT Hamberger B 2014b Heterologous stable expression of terpenoid biosynthetic genes using the moss Physcomitrella patens Methods Mol Biol 1153 257 271 24777804 Bohlmann J Gershenzon J 2009 Old substrates for new enzymes of terpenoid biosynthesis Proc Natl Acad Sci U S A 106 10402 10403 19553206 Brown GD 2010 The biosynthesis of artemisinin (Qinghaosu) and the phytochemistry of Artemisia annua L. (Qinghao) Molecules 15 7603 7698 21030913 Carter OA Peters RJ Croteau R 2003 Monoterpene biosynthesis pathway construction in Escherichia coli Phytochemistry 64 425 433 12943759 Cartwright DW Langcake P Pryce RJ Leworthy DP Ride JP 1977 Chemical activation of host defence mechanisms as a basis for crop protection Nature 267 511 513 Chen F Tholl D Bohlmann J Pichersky E 2011 The family of terpene synthases in plants: a mid-size family of genes for specialized metabolism that is highly diversified throughout the kingdom Plant J 66 212 229 21443633 Christian M Cermak T Doyle EL Schmidt C Zhang F Hummel A Bogdanove AJ Voytas DF 2010 Targeting DNA double-strand breaks with TAL effector nucleases Genetics 186 757 761 20660643 Colby SM Alonso WR Katahira EJ McGarvey DJ Croteau R 1993 4S-limonene synthase from the oil glands of spearmint (Mentha spicata). cDNA isolation, characterization, and bacterial expression of the catalytically active monoterpene cyclase J Biol Chem 268 23016 23024 8226816 Criswell J Potter K Shephard F Beale MB Peters RJ 2012 A single residue change leads to a hydroxylated product from the class II diterpene cyclization catalyzed by abietadiene synthase Org Lett 14 5828 5831 23167845 Croteau RB Davis EM Ringer KL Wildung MR 2005 (−)-Menthol biosynthesis and molecular genetics Die Naturwissenschaften 92 562 577 16292524 Cyr A Wilderman PR Determan M Peters RJ 2007 A Modular Approach for Facile Biosynthesis of Labdane-Related Diterpenes J Am Chem Soc 129 6684 6685 17480080 Diaz-Chavez ML Moniodis J Madilao LL Jancsik S Keeling CI Barbour EL Ghisalberti EL Plummer JA Jones CG Bohlmann J 2013 Biosynthesis of Sandalwood Oil: Santalum album CYP76F cytochromes P450 produce santalols and bergamotol PLoS One 8 e75053 24324844 Facchini PJ Bohlmann J Covello PS De Luca V Mahadevan R Page JE Ro DK Sensen CW Storms R Martin VJ 2012 Synthetic biosystems for the production of high-value plant metabolites Trends in biotechnology 30 127 131 22209518 Feng Z Zhang B Ding W Liu X Yang DL Wei P Cao F Zhu S Zhang F Mao Y 2013 Efficient genome editing in plants using a CRISPR/Cas system Cell research 23 1229 1232 23958582 Gao W Hillwig ML Huang L Cui G Wang X Kong J Yang B Peters RJ 2009 A functional genomics approach to tanshinone biosynthesis provides stereochemical insights Org Lett 11 5170 5173 19905026 Gao Y Honzatko RB Peters RJ 2012 Terpenoid synthase structures: a so far incomplete view of complex catalysis Nat Prod Rep 29 1153 1175 22907771 Gershenzon J McCaskill D Rajaonarivory JIM Mihaliak C Karp R Croteau R 1992 Isolation of secretory cells from plant glandular trichomes and their use in biosynthetic studies of monoterpenes and other glanular products Anal Biochem 200 130 138 1595887 Giddings LA Liscombe DK Hamilton JP Childs KL DellaPenna D Buell CR O’Connor SE 2011 A stereoselective hydroxylation step of alkaloid biosynthesis by a unique cytochrome P450 in Catharanthus roseus J Biol Chem 286 16751 16757 21454651 Gongora-Castillo E Fedewa G Yeo Y Chappell J DellaPenna D Buell CR 2012 Genomic approaches for interrogating the biochemistry of medicinal plant species Methods in enzymology 517 139 159 23084937 Guengerich FP Munro AW 2013 Unusual cytochrome p450 enzymes and reactions J Biol Chem 288 17065 17073 23632016 Hamberger B Bak S 2013 Plant P450s as versatile drivers for evolution of species-specific chemical diversity Philosophical transactions of the Royal Society of London Series B, Biological sciences 368 20120426 23297350 Hamberger B Ohnishi T Seguin A Bohlmann J 2011 Evolution of diterpene metabolism: Sitka spruce CYP720B4 catalyzes multiple oxidations in resin acid biosynthesis of conifer defense against insects Plant Physiology 157 1677 1695 21994349 Haudenschild C Schalk M Karp R Croteau R 2000 Functional expression of regiospecific cytochrome P450 limonene hydroxylases from mint (Mentha spp.) in Escherichia coli and Saccharomyces cervisiae Arch Biochem Biophys 279 127 136 Hershey DM Lu X Zi J Peters RJ 2014 Functional conservation of the capacity for ent-kaurene biosynthesis and an associated operon in certain rhizobia J Bact 196 100 106 24142247 Ikezawa N Gopfert JC Nguyen DT Kim SU O’Maille PE Spring O Ro DK 2011 Lettuce costunolide synthase (CYP71BL2) and its homolog (CYP71BL1) from sunflower catalyze distinct regio- and stereoselective hydroxylations in sesquiterpene lactone metabolism J Biol Chem 286 21601 21611 21515683 Inoue Y Sakai M Yao Q Tanimoto Y Toshima H Hasegawa M 2013 Identification of a novel casbane-type diterpene phytoalexin, ent-10-oxodepressin, from rice leaves Biosci Biotechnol Biochem 77 760 765 23563545 International Rice Genome Sequencing Project 2005 The map-based sequence of the rice genome Nature 436 793 800 16100779 Jackson AJ Hershey DM Chesnut T Xu M Peters RJ 2014 Biochemical characterization of the castor bean ent-kaurene synthase(-like) family supports quantum chemical view of diterpene cyclization Phytochemistry in press Jennewein S Long RM Williams RM Croteau RB 2004 Cytochrome p450 taxadiene 5alpha-hydroxylase, a mechanistically unusual monooxygenase catalyzing the first oxygenation step of taxol biosynthesis Chem Biol 11 379 387 15123267 Jinek M Chylinski K Fonfara I Hauer M Doudna JA Charpentier E 2012 A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity Science 337 816 821 22745249 Kato T Kabuto C Sasaki N Tsunagawa M Aizawa H Fujita K Kato Y Kitahara Y Takahashi N 1973 Momilactones, growth inhibitors from rice, Oryza sativa L Tetrahedron Lett 14 3861 3864 Kato-Noguchi H Peters RJ 2013 The role of momilactones in rice allelopathy J Chem Ecol 39 175 185 23385366 Keasling JD 2012 Synthetic biology and the development of tools for metabolic engineering Metabolic engineering 14 189 195 22314049 Keasling JD Mendoza A Baran PS 2012 Synthesis: A constructive debate Nature 492 188 189 23235869 Kikuchi S Satoh K Nagata T Kawagashira N Doi K Kishimoto N Yazaki J Ishikawa M Yamada H Ooka H 2003 Collection, mapping, and annotation of over 28,000 cDNA clones from japonica rice Science 301 376 379 12869764 Kitaoka N Wu Y Xu M Peters RJ submitted Optimization of a synthetic biology approach enables discovery of novel cytochrome P450 activity in rice diterpenoid biosynthesis Mol Plant Lange BM Mahmoud SS Wildung MR Turner GW Davis EM Lange I Baker RC Boydston RA Croteau RB 2011 Improving peppermint essential oil yield and composition by metabolic engineering Proc Natl Acad Sci U S A 108 16944 16949 21963983 Li T Huang S Jiang WZ Wright D Spalding MH Weeks DP Yang B 2011 TAL nucleases (TALNs): hybrid proteins composed of TAL effectors and FokI DNA-cleavage domain Nucleic Acids Res 39 359 372 20699274 Lu X Hershey DM Wang L Bogdanove AJ Peters RJ submitted An ent-kaurene derived diterpenoid virulence factor from Xanthomonas oryzae pv. oryzicola New Phytol Lupien S Karp F Wildung M Croteau R 1999 Regiospecific cytochrome P450 limonene hydroxylases from mint (Mentha) species. cDNA isolation, characterization, and functional expression of (−)-4S-limonene-3-hydroxylase and (−)-4S-limonene-6-hydroxylase Arch Biochem Biophys 368 181 192 10415126 Marques JV Kim KW Lee C Costa MA May GD Crow JA Davin LB Lewis NG 2013 Next generation sequencing in predicting gene function in podophyllotoxin biosynthesis J Biol Chem 288 466 479 23161544 Martin VJJ Pitera DJ Withers ST Newman JD Keasling JD 2003 Engineering a mevalonate pathway in Escherichia coli for production of terpenoids Nature biotechnology 21 796 802 Miao J Guo D Zhang J Huang Q Qin G Zhang X Wan J Gu H Qu LJ 2013 Targeted mutagenesis in rice using CRISPR-Cas system Cell research 23 1233 1236 23999856 Morrone D Chambers J Lowry L Kim G Anterola A Bender K Peters RJ 2009 Gibberellin biosynthesis in bacteria: Separate ent-copalyl diphosphate and ent-kaurene synthases in Bradyrhizobium japonicum FEBS Lett 583 475 480 19121310 Morrone D Chen X Coates RM Peters RJ 2010a Characterization of the kaurene oxidase CYP701A3, a multifunctional cytochrome P450 from gibberellin biosynthesis Biochem J 431 337 344 20698828 Morrone D Hillwig ML Mead ME Lowry L Fulton DB Peters RJ 2011 Evident and latent plasticity across the rice diterpene synthase family with potential implications for the evolution of diterpenoid metabolism in the cereals Biochem J 435 589 595 21323642 Morrone D Jin Y Xu M Choi S-Y Coates RM Peters RJ 2006 An unexpected diterpene cyclase from rice: Functional identification of a stemodene synthase Arch Biochem Biophys 448 133 140 16256063 Morrone D Lowry L Determan MK Hershey DM Xu M Peters RJ 2010b Increasing diterpene yield with a modular metabolic engineering system in E. coli: comparison of MEV and MEP isoprenoid precursor pathway engineering Appl Microbiol Biotechnol 85 1893 1906 19777230 Morrone D Xu M Fulton DB Determan MK Peters RJ 2008 Increasing complexity of a diterpene synthase reaction with a single residue switch J Am Chem Soc 130 5400 5401 18366162 Moses T Pollier J Thevelein JM Goossens A 2013 Bioengineering of plant (tri)terpenoids: from metabolic engineering of plants to synthetic biology in vivo and in vitro New Phytol 200 27 43 23668256 Nelson D Werck-Reichhart D 2011 A P450-centric view of plant evolution Plant J 66 194 211 21443632 Nguyen DT Gopfert JC Ikezawa N Macnevin G Kathiresan M Conrad J Spring O Ro DK 2010 Biochemical conservation and evolution of germacrene A oxidase in asteraceae J Biol Chem 285 16588 16598 20351109 Okada A Shimizu T Okada K Kuzuyama T Koga J Shibuya N Nojiri H Yamane H 2007 Elicitor induced activation of the methylerythritol phosphate pathway towards phytoalexin biosynthesis in rice Plant Mol Biol 65 177 187 17634747 Paddon CJ Keasling JD 2014 Semi-synthetic artemisinin: a model for the use of synthetic biology in pharmaceutical development Nature reviews Microbiology 12 355 367 24686413 Paddon CJ Westfall PJ Pitera DJ Benjamin K Fisher K McPhee D Leavell MD Tai A Main A Eng D 2013 High-level semi-synthetic production of the potent antimalarial artemisinin Nature 496 528 532 23575629 Peters RJ 2006 Uncovering the complex metabolic network underlying diterpenoid phytoalexin biosynthesis in rice and other cereal crop plants Phytochemistry 67 2307 2317 16956633 Peters RJ 2010 Two rings in them all: The labdane-related diterpenoids Nat Prod Rep 27 1521 1530 20890488 Potter K Criswell J Peters RJ 2014 Mechanistic analysis of the ent-copalyl diphosphate synthases required for plant gibberellin hormone biosynthesis leads to novel product chemistry Angew Chem in press Prisic S Xu M Wilderman PR Peters RJ 2004 Rice contains two disparate ent-copalyl diphosphate synthases with distinct metabolic functions Plant Physiol 136 4228 4236 15542489 Renault H Bassard JE Hamberger B Werck-Reichhart D 2014 Cytochrome P450-mediated metabolic engineering: current progress and future challenges Current opinion in plant biology 19 27 34 24709279 Rios-Estepa R Lange I Lee JM Lange BM 2010 Mathematical modeling-guided evaluation of biochemical, developmental, environmental, and genotypic determinants of essential oil composition and yield in peppermint leaves Plant Physiol 152 2105 2119 20147490 Rios-Estepa R Turner GW Lee JM Croteau RB Lange BM 2008 A systems biology approach identifies the biochemical mechanisms regulating monoterpenoid essential oil composition in peppermint Proc Natl Acad Sci U S A 105 2818 2823 18287058 Ro DK Paradise E Ouellet M Fisher KJ Newman KL Ndungu JM Ho KA Eachus RA Ham TS Kirby J 2006 Production of the antimalarial drug precursor artemisinic acid in engineered yeast Nature 440 940 943 16612385 Ro DK Arimura G Lau SY Piers E Bohlmann J 2005 Loblolly pine abietadienol/abietadienal oxidase PtAO (CYP720B1) is a multifunctional, multisubstrate cytochrome P450 monooxygenase Proc Natl Acad Sci U S A 102 8060 8065 15911762 Sakamoto T Miura K Itoh H Tatsumi T Ueguchi-Tanaka M Ishiyama K Kobayashi M Agrawal GK Takeda S Abe K 2004 An overview of gibberellin metabolism enzyme genes and their related mutants in rice Plant Physiol 134 1642 1653 15075394 Schmelz EA Huffaker A Sims JW Christensen SA Lu X Okada K Peters RJ 2014 Biosynthesis, elicitation and roles of monocot terpenoid phytoalexins Plant J 79 659 678 24450747 Shimura K Okada A Okada K Jikumaru Y Ko K-W Toyomasu T Sassa T Hasegawa M Kodama O Shibuya N 2007 Identification of a biosynthetic gene cluster in rice for momilactones J Biol Chem 282 34013 34018 17872948 Swaminathan S Morrone D Wang Q Fulton DB Peters RJ 2009 CYP76M7 is an ent-cassadiene C11α-hydroxylase defining a second multifunctional diterpenoid biosynthetic gene cluster in rice Plant Cell 21 3315 3325 19825834 Teoh KH Polichuk DR Reed DW Covello PS 2009 Molecular cloning of an aldehyde dehydrogenase implicated in artemisinin biosynthesis in Artemisia annua Botany 87 635 642 Teoh KH Polichuk DR Reed DW Nowak G Covello PS 2006 Artemisia annua L. (Asteraceae) trichome-specific cDNAs reveal CYP71AV1, a cytochrome P450 with a key role in the biosynthesis of the antimalarial sesquiterpene lactone aremisinin FEBS Lett 580 1411 1416 16458889 Toyomasu T 2008 Recent Advances Regarding Diterpene Cyclase Genes in Higher Plants and Fungi Biosci Biotechnol Biochem 72 1168 1175 18460786 Toyomasu T Usui M Sugawara C Otomo K Hirose Y Miyao A Hirochika H Okada K Shimizu T Koga J 2013 Reverse-genetic approach to verify physiological roles of rice phytoalexins: characterization of a knockdown mutant of OsCPS4phytoalexin biosynthetic gene in rice Physiologia plantarum van Herpen TW Cankar K Nogueira M Bosch D Bouwmeester HJ Beekwilder J 2010 Nicotiana benthamiana as a production platform for artemisinin precursors PLoS One 5 e14222 21151979 Vranova E Coman D Gruissem W 2013 Network analysis of the MVA and MEP pathways for isoprenoid synthesis Annu Rev Plant Biol 64 665 700 23451776 Wang Q Hillwig ML Okada K Yamazaki K Wu Y Swaminathan S Yamane H Peters RJ 2012a Characterization of CYP76M5-8 indicates metabolic plasticity within a plant biosynthetic gene cluster J Biol Chem 287 6159 6168 22215681 Wang Q Hillwig ML Peters RJ 2011 CYP99A3: Functional identification of a diterpene oxidase from the momilactone biosynthetic gene cluster in rice Plant J 65 87 95 21175892 Wang Q Hillwig ML Wu Y Peters RJ 2012b CYP701A8: A rice ent-kaurene oxidase paralog diverted to more specialized diterpenoid metabolism Plant Physiol 158 1418 1425 22247270 Wilderman PR Xu M Jin Y Coates RM Peters RJ 2004 Identification of syn-pimara-7,15-diene synthase reveals functional clustering of terpene synthases involved in rice phytoalexin/allelochemical biosynthesis Plant Physiol 135 2098 2105 15299118 Williams DC McGarvey DJ Katahira EJ Croteau R 1998 Truncation of limonene synthase preprotein provides a fully active ‘pseudomature’ form of this monoterpene cyclase and reveals the function of the amino-terminal arginine pair Biochemistry 37 12213 12220 9724535 Wu Y Hillwig ML Wang Q Peters RJ 2011 Parsing a multifunctional biosynthetic gene cluster from rice: Biochemical characterization of CYP71Z6 & 7 FEBS Lett 585 3446 3451 21985968 Wu Y Wang Q Hillwig ML Peters RJ 2013 Picking sides: Distinct roles for CYP76M6 and -8 in rice oryzalexin biosynthesis Biochem J 454 209 216 23795884 Wu Y Zhou K Toyomasu T Sugawara C Oku M Abe S Usui M Mitsuhashi W Chono M Chandler PM 2012 Functional characterization of wheat copalyl diphosphate synthases elucidates the early evolution of labdane-related diterpenoid metabolism in the cereals Phytochemistry 84 40 46 23009878 Xiao M Zhang Y Chen X Lee EJ Barber CJ Chakrabarty R Desgagne-Penix I Haslam TM Kim YB Liu E 2013 Transcriptome analysis based on next-generation sequencing of non-model plants producing specialized metabolites of biotechnological interest Journal of biotechnology 166 122 134 23602801 Xu M Galhano R Wiemann P Bueno E Tiernan M Wu W Chung IM Gershenzon J Tudzynski B Sesma A 2012 Genetic evidence for natural product-mediated plant-plant allelopathy in rice (Oryza sativa) New Phytol 193 570 575 22150231 Xu M Hillwig ML Lane AL Tiernan MS Moore BS Peters RJ submitted Characterization of an orphan diterpenoid biosynthetic operon from Salinispora arenicola J Nat Prod Yamane H 2013 Biosynthesis of phytoalexins and regulatory mechanisms of it in rice Biosci Biotechnol Biochem 77 1141 1148 23748776 Zerbe P Hamberger B Yuen MM Chiang A Sandhu HK Madilao LL Nguyen A Hamberger B Bach SS Bohlmann J 2013 Gene discovery of modular diterpene metabolism in nonmodel systems Plant Physiol 162 1073 1091 23613273 Zhang H Zhang J Wei P Zhang B Gou F Feng Z Mao Y Yang L Zhang H Xu N 2014 The CRISPR/Cas9 system produces specific and homozygous targeted gene editing in rice in one generation Plant biotechnology journal 12 797 807 24854982 Zhou H Liu B Weeks DP Spalding MH Yang B 2014 Large chromosomal deletions and heritable small genetic changes induced by CRISPR/Cas9 in rice Nucleic Acids Res 10.1093/nar/gku806 Zhou K Xu M Tiernan MS Xie Q Toyomasu T Sugawara C Oku M Usui M Mitsuhashi W Chono M 2012 Functional characterization of wheat ent-kaurene(-like) synthases indicates continuing evolution of labdane-related diterpenoid metabolism in the cereals Phytochemistry 84 47 55 23009879 Zi J Peters RJ 2013 Characterization of CYP76AH4 clarifies phenolic diterpenoid biosynthesis in the Lamiaceae Org Biomol Chem 11 7650 7652 24108414
PMC005xxxxxx/PMC5120992.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 2984726R 1011 Biochem J Biochem. J. The Biochemical journal 0264-6021 1470-8728 27462123 5120992 10.1042/BCJ20160466 NIHMS830823 Article CD133-positive dermal papilla-derived Wnt ligands regulate postnatal hair growth Zhou Linli 1 Yang Kun 1 Carpenter April 2 Lang Richard A. 2 Andl Thomas 3 Zhang Yuhang 1 1 Division of Pharmaceutical Sciences, College of Pharmacy, University of Cincinnati, Cincinnati, OH 45267, USA 2 Division of Pediatric Ophthalmology and Division of Developmental Biology, Cincinnati Children’s Hospital Medical Center, Cincinnati, OH 45229, USA 3 Burnett School of Biological Sciences, University of Central Florida, Orlando, FL 32816, USA Correspondence: Yuhang Zhang ([email protected]) 18 11 2016 26 7 2016 1 10 2016 01 10 2017 473 19 32913305 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Active Wnt/β-catenin signaling in the dermal papilla (DP) is required for postnatal hair cycling. In addition, maintenance of the hair-inducing ability of DP cells in vitro requires external addition of Wnt molecules. However, whether DP cells are a critical source of Wnt ligands and induce both autocrine and paracrine signaling cascades to promote adult hair follicle growth and regeneration remains elusive. To address this question, we generated an animal model that allows inducible ablation of Wntless (Wls), a transmembrane Wnt exporter protein, in CD133-positive (CD133+) DP cells. CD133+ cells have been shown to be a specific subpopulation of cells in the DP, which possesses the hair-inducing capability. Here, we show that ablation of Wls expression in CD133+ DP cells results in a shortened period of postnatal hair growth. Mutant hair follicles were unable to enter full anagen (hair growth stage) and progressed toward a rapid regression. Notably, reduced size of the DP and decreased expression of anagen DP marker, versican, were observed in hair follicles when CD133+ DP cells lost Wls expression. Further analysis showed that Wls-deficient CD133+ DP cells led to reduced proliferation and differentiation in matrix keratinocytes and melanocytes that are needed for the generation of the hair follicle structure and a pigmented hair shaft. These findings clearly demonstrate that Wnt ligands produced by CD133+ DP cells play an important role in postnatal hair growth by maintaining the inductivity of DP cells and mediating the signaling cross-talk between the mesenchyme and the epithelial compartment. Introduction The mature hair follicle is a regenerating mini-organ that primarily comprises of epithelial and dermal compartments [1]. The cells in these two compartments interact with each other through molecular signals that collectively drive a repetitive process of hair follicle reconstitution [2], comprising cyclical periods of growth (anagen), regression (catagen) and rest (telogen) [3]. In mice, the size and shape of murine pelage hairs and their cycling properties are affected by the number of cells in the DP [4]. It has been conclusively shown that DP cells of growing hair follicles possess inductive properties that regulate the behavior of epidermal keratinocytes, resulting in the regeneration of hair follicle structure and fiber [5,6]. Unfortunately, reciprocal signaling between the epithelial compartment and the DP is complicated and still remains to be defined. The canonical Wnt/β-catenin signaling pathway is broadly active in hair follicles and modulates the epithelial–mesenchymal interactions (EMIs) that control hair follicle regenerative cycles [7]. This pathway becomes active when the secreted Wnt ligands, which are glycoproteins, bind to Frizzled (Fzd) receptors and form a complex with Lrp5/6 coreceptors [8]. In the absence of Wnt ligands (off-state), cytoplasmic β-catenin, a dual function protein involved in both cell–cell adhesion and transcriptional regulation, is degraded upon phosphorylation by the Apc/Axin/Gsk-3β complex and subsequent ubiquitination [9]. When Wnt ligands are present (on-state), they form a complex with Fzd receptors and Lrp5/6 coreceptors, triggering inactivation of the Apc/Axin/Gsk-3β complex. Consequently, disheveled is activated to displace Gsk-3β from the Apc/Axin/Gsk-3β complex, so that cytoplasmic β-catenin can no longer be targeted for degradation. Subsequently, β-catenin accumulates in the cytoplasm and translocates to the nucleus where it forms complexes with members of the Lef1/Tcf transcription factor family. This Lef1/Tcf–β-catenin complex displaces transcriptional repressors from target gene promoters and recruits additional coactivators to activate gene expression of Wnt signaling target genes [10]. The on/off of Wnt/β-catenin signaling is not only controlled by Wnt ligands, but also by different Wnt inhibitors, including members of the Dickkopf family and Wnt inhibitory factor 1 [11]. Many genetic experiments have clearly indicated that co-ordinated Wnt/β-catenin activity in both epithelial and mesenchymal compartments of the mature hair follicle is indispensable for hair follicle regenerative cycling. The source and nature of Wnt ligands remain intriguing questions. Epidermal Wnt ligands have been shown to be necessary for hair follicle morphogenesis and postnatal hair cycling [12–15]. Simlarly, the DP could be a source of Wnt ligands for postnatal hair cycling as well. Wnt5a, Wnt10a and Wnt11 mRNAs are expressed in the upper dermis or dermal condensate at mouse embryonic day E14.5 [16]. In one study, Wnt5a expression was identified in the DP during the adult hair growth cycle [16]. Deletion of Wnt5a in DP cells caused abnormal hair follicle differentiation [17]. Another important point to consider is that DP cells require external addition of Wnt protein to maintain their hair-inducing ability in vitro, which may also be the case in vivo [18,19]. These observations suggest an important role for DP-derived Wnt ligands in postnatal hair growth. However, despite these advances, the key question still remains on whether DP-derived Wnt ligands are required for active Wnt/β-catenin signaling in both epithelial and mesenchymal compartments in the adult hair cycle. To address this question, we have designed a novel genetic approach to block the export of Wnt ligands from CD133+ DP cells by inducible deletion of a Wnt-exporting protein, Wntless (Wls) [20,21]. Wls, also referred to as ‘evenness interrupted’ in Drosophila and as MOM-3/MIG-14 in Caenorhabditis elegans, is a seven-pass transmembrane protein [22]. Because Wnt ligands need first to be exported from the producing cells before they can function on their target cells, the ablation of Wls will result in the accumulation of Wnt ligands inside Wnt-producing cells and consequently the absence of Wnt signaling in Wnt target cells [22–24]. Multiple reports have confirmed that Wls is indeed responsible for the surface delivery and secretion of Wnt proteins in hair follicles [12–15,25]. CD133+ DP cells are the population in the DP that has trichogenic ability and is capable of regulating DP size [26,27]. However, whether CD133+ DP cells produce Wnt proteins to maintain the hair-inducing capacaity and interact with keratiocytes to build hair follicles in vivo remains unknown. Here, we report that ablation of Wls expression in CD133+ DP cells delays hair growth and causes premature hair follicle regression. Further analysis shows that Wls deficiency in CD133+ DP cells not only affects the proliferation and differentiation of matrix keratinocytes, but also modulates the biological properties of DP cells. Taken together, our data highlight that Wnt ligands generated by CD133+ DP cells are critical for the postnatal hair regenerative cycle by co-ordinately regulating both the epithelial compartment and the mesenchymal niche. Methods Mice CD133-CreERT2 (Prom1C-L) mice and Wlsfl/fl mice were generated as previously reported [20,28]. To generate CD133-CreERT2; Wlsfl/fl mice, CD133-CreERT2 mice were crossed with Wlsfl/fl mice for several generations. TOPGAL mice were obtained from the Jackson Laboratory (stock number 004623). All mice were housed in the Laboratory Animal Services Facility of the University of Cincinnati under an artificial 12/12 light–dark cycle and were allowed free access to normal mouse feedings and water. The Institutional Animal Care and Use Committee of the University of Cincinnati approved all experimental procedures involving mice. All animals were cared daily, 7 days per week, by three full-time veterinarians, six veterinary technicians plus animal care staff. Mice were genotyped by polymerase chain reaction (PCR) analysis of genomic DNA extracted from mouse tail biopsies. Wls alleles were detected using the following primers: P1, CTTCCCTGCTTCTTTAAGCGTC; P2, AGGCTTCGAACGTAACTGACC; P4, CTCAGAACTCCCTTCTTGAAGC. P2 and P4 primers were used for the detection of 411-bp wild-type allele and 556-bp floxed allele. P1 and P4 primers were used for 1625-bp wild-type allele and 410-bp recombined allele after successful Cre deletion. CD133 alleles were genotyped using the following primers: forward primer: CAGGCTGTTAGCTTGGGTTC; reverse primer 1: AGGCAAATTTTGGTGTACGG; reverse primer 2: TAGCGTGGTCATGAAGCAAC. Wild-type CD133 allele was genotyped using forward primer with reverse primer 2. Insertion of CreERT2 transgene in CD133 locus was genotyped using forward primer with reverse primer 1. TOPGAL transgene (LacZ) was genotyped using the following primers: forward primer: ATCCTCTGCATGGTCAGGTC and reverse primer: CGTGGCCTGATTCATTCC. The PCR protocol used was 94°C for 3 min followed by 35 cycles of 94°C for 30 s, 62°C for 30 s and 72°C for 40 s, and a final extension at 72°C for 10 min. Whole-mount X-gal staining Skin biopsies of a size of 0.5 × 1 cm from the mid-dorsal region of TOPGAL mice were collected at postnatal day 25 (P25), P28, P32, P35, P40 and P45 for X-gal whole-mount staining. Briefly, skin tissues were fixed in 4% paraformaldehyde (PFA) at room temperature for 30 min, washed with phosphate-buffered saline (PBS) and immersed in X-gal staining buffer [1 mg/ml X-gal (5Primer, Gaithersburg, MD), 5 mM potassium ferricyanide (Sigma-Aldrich, St. Louis, MO), 5 mM potassium ferrocyanide (Sigma-Aldrich, St. Louis, MO) in PBS] at room temperature for 48 h. After staining, tissues were fixed in 4% PFA with 10% sucrose overnight at 4°C and then embedded in tissue embedding medium (OCT) for sectioning. Frozen sections were counterstained with nuclear red and examined under a Nikon Eclipse 80i microscope. Images were analyzed using Adobe Photoshop. Transgene induction To induce Cre activity and recombination of Wls floxed alleles, mice were administered with tamoxifen (TAM) (Sigma-Aldrich, St. Louis, MO) in corn oil (10 mg/ml) by intraperitoneal (IP) injection at 1 mg/g body weight starting from P21 to P27. Skin samples were collected at P28, P30, P32, P35, P40 and P45. To ablate Wls expression during mid-anagen, mice were administered with TAM at P30 by IP injection for 5 or 7 consecutive days based on the date when skin biopsies were harvested. Skin biopsies were collected at P35, P40 and P50. Skin biopsies Stages of normal hair cycle were determined according to the classification system published previously [1]. Mid-dorsal skin biopsies were taken from CO2 inhalation-killed mice, the standard American Veterinary Medical Association method. Death was confirmed by the lack of heartbeat. Cervical dislocation was performed to ensure mice would not recover. Two hours before sacrifice, all mice were given one IP injection of bromodeoxyuridine (BrdU) based on their body weight (50 µg/g of body weight). Hair on the back of mice was carefully shaved using an electric clipper before harvesting skin biopsies. Collected skin tissues were then fixed and processed for paraffin and frozen sectioning. Histology and immunostaining Histology and immunohistochemistry were performed as described previously [12]. For histology, paraffin sections were stained with hematoxylin and eosin (H&E) using standard protocols. For immunostaining, paraffin sections were deparaffinized, rehydrated and then demasked in citrate buffer (pH 6.0) using the microwave heating method. After washing with PBS, sections were blocked in 10% bovine serum albumin in PBS and subsequently incubated at 4°C overnight with each primary antibody. Slides were then washed with PBS for three times and incubated with the corresponding biotin-conjugated secondary antibodies (Vector Laboratory, Burlingame, CA) at room temperature for 1 h. For immunofluorescence staining, slides were incubated with either fluorescein or Texas red-conjugated streptavidin for 1 h. Microscopy was performed using a Nikon Eclipse 80i fluorescence microscope and images were analyzed using ImageJ (NIH). The following primary antibodies were used: anti-Wntless (1:500, a gift from Dr Richard Lang, Cincinnati Children’s Hospital Medical Center, Cincinnati, OH), anti-β-catenin (Invitrogen, 15B8, 1:1000), anti-BrdU (Abcam, BU1/75, 1:25), anti-Ki67 (Imgenex, 1:50), anti-Lef1 (Cell Signaling, 1:100), anti-versican (Millipore, 1:200), anti-Sox9 (Millipore, 1:100), anti-GATA3 (Sant Cruz, 1:50), anti-K15 (Vector Laboratory, 1:50), anti-AE13 (1:25, a gift from Dr Tung-Tien Sun, NYU Langone Medical Center, New York, NY), anti-AE15 (1:25, a gift from Dr Tung-Tien Sun, NYU Langone Medical Center, New York, NY) and anti-microphthalmia-associated transcription factor (MITF; Thermo Fisher, 1:200). Stained slides were examined under a Nikon Eclipse 80i fluorescence microscope. Isolation of CD133+ DP cells Isolation of CD133+ DP cells was performed as previously described with modification [21]. Briefly, TAM in corn oil (10 mg/ml) was administered to adult mice by IP injection at 1 mg/g body weight for 7 consecutive days from P21 to P27 to induce Cre expression and Wls ablation. Mid-dorsal back skins were harvested at P30 and floated in 0.1% dispase (Thermo Fisher, Waltham, MA) at 4°C for 16 h to separate the epidermis and dermis. The epidermis was then discarded, and the dermis was treated with 0.5% collagenase IV (Thermo Fisher, Waltham, MA). Dissociated dermal cells were filtered with cell strainer, collected by centrifugation and followed by resuspension in 100 µl of culture medium and incubation with APC-conjugated anti-CD133 antibodies (eBioscience, San Diego, CA, 1:50) for 30 min at 4°C. Cell sorting was performed using a MoFlo high-speed sorter (Dako Cytomation, Carpinteria, CA). Western blotting Immunoblotting for Wnt5a expression (Cell Signaling, 1:500) using isolated CD133+ DP cells was performed as described recently [29]. To verify equal loading of samples, membranes were stripped and reprobed with monoclonal anti-β-actin antibody (Invitrogen, 1:5000), followed by an HRP-conjugated goat anti-mouse IgG (Cell Signaling, 1:2000). The intensities of Wnt5a and β-actin bands were determined by scanning X-ray films and measuring using ImageJ software (NIH). The relative expression level of Wnt5a was calculated by dividing the value of Wnt5a by the net loading control of β-actin. Hair follicle number counting Stages of hair follicles were determined according to previously published classifications [3], which describe the key morphological features of murine hair follicles during the hair cycle comprising of hair follicle growth anagen, catagen and telogen stages. The number of hair follicles was counted manually by examining H&E-stained skin samples (×10 objective). At least 10 fields for each skin biopsy were examined for hair follicle counting. A minimum of three CD133-CreERT2; Wlsfl/fl mutant mice and three control littermates were counted for each time point. Alkaline phosphatase staining Alkaline phosphatase (AP) staining was performed using a VECTOR Red AP Substrate Kit (Vector Laboratory, Burlingame, CA) according to the manufacturer’s instruction. Briefly, frozen sections were washed with PBS for 5 min twice and incubated with the substrate solution for 30 min in the dark. Slides were then washed in 100 mM Tris–HCl buffer (pH 8.5) for 5 min, rinsed in water and mounted with a VECTASHIELD Mounting Medium with DAPI for observation. DP cell number counting DP cells in each hair follicle were counted on skin biopsy sections after AP staining and DAPI counterstaining under a Nikon Eclipse 80i fluorescence microscope using the high power 40× objective. For counting, a minimum of 10 hair follicles was randomly picked from at least two pieces of mid-dorsal skin tissues of each mouse. For each genotype, three mice were used with a total of 30 hair follicles counted. Statistical analysis All graphs were generated using Microsoft Excel (2016). Statistical analysis of difference was carried out by Student’s t-test using a GraphPad Prism 5.01 software package (GraphPad Software, Inc., San Diego, CA) and represented as mean ± SEM, with P < 0.05 considered statistically significant. Result Wls is expressed in DP cells during anagen stage except onset To delineate the potential contribution of the dermal papilla (DP) to the production of Wnt ligands, we evaluated the expression of Wls in murine hair follicles at successive stages of the postnatal hair growth cycle. As shown in Figure 1A, at P21, Wls expression could not be detected in DP cells when hair follicles were at the beginning of the anagen stage (surrounded by white circle). In contrast, Wls was expressed at a low level in the hair follicle epithelium (indicated by red triangle). By P28, Wls expression was significantly up-regulated in the DP while maintained at a high level in transit-amplifying keratinocytes (Figure 1B). From P32 to P40 (Figure 1C–E), high level of Wls was readily detected in the DP, matrix keratinocytes surrounding the DP (indicated by red triangle) and inner root sheath cells (indicated by yellow arrows). However, there was a lack of Wls expression in the area above the DP (denoted by orange stars) where normally hair shaft precursor cells reside. Hair shaft precursor cells are Lef1-positive but Ki67-negative as shown in Figure 1O–Q. When hair follicles enter the regression stage at P45 (Figure 1F), Wls expression could no longer be detected in the DP. These observations show that Wls is dynamically expressed in DP cells during the anagen phase, with the exception of transition stage from telogen to anagen [14]. We first utilized TOPGAL Wnt reporter mice to assay for Wnt/β-catenin signaling activity in postnatal skin, in which three consensus Lef/Tcf-binding sites upstream of a minimal c-fos promoter drive the expression of lacZ transgene [30]. As shown in Figure 1G, lacZ expression was undetectable in the DP of early anagen hair follicles at P25, possibly due to the lack of sensitivity of reporter gene at this stage. At P28 (Figure 1H), there was weak blue staining in the DP, while higher lacZ expression was found in the hair matrix (indicated by red triangle). Consistent with Wls expression, TOPGAL transgene was expressed very strongly in the DP and hair matrix keratinocytes surrounding the DP (Figure 1I–K). At P45, weak lacZ expression could still be seen in the DP and the epithelial strand (Figure 1L). Since the sensitivity of TOPGAL reporter is affected by the number and placement of Lef/Tcf sites and the integration site of the transgene, we needed additional Wnt activity indicators to confirm Wnt/β-catenin signaling in the skin. Activation of canonical Wnt signaling by Wnt ligands results in the stabilization of β-catenin, which associates with members of the Lef1/Tcf family of DNA-binding proteins to initiate target gene transcription [10]. Lef1 has been demonstrated to be a Wnt target gene; therefore, it could be used as an indicator for cells potentially responding to Wnt ligands and being able to activate Wnt/β-catenin signaling. Thus, we evaluated the expression of Lef1 in the hair follicle from anagen onset to telogen. As shown in Figure 1M, at P25, Lef1 expression was exclusively identified in the DP during early anagen. From P28 to P40 (Figure 1N–Q), expression of Lef1 was seen in both the DP and matrix and precortex keratinocytes that encircle the DP of anagen hair follicles (indicated by red triangles). At P45, when hair follicles entered the regression stage, Lef1 expression could still be detected in the DP and differentiating keratinocytes (Figure 1R). By telogen, no Lef1 expression was detected in the DP and the hair follicle epithelium (data not shown). These observations show that the expression of Wls overlaps with active Wnt/β-catenin signaling in matrix keratinocytes and DP cells during hair growth anagen. Wls in CD133+ DP cells is required for hair follicle anagen To determine whether CD133+ DP cells produce Wnt ligands for postnatal hair growth, we generated CD133-CreERT2; Wlsfl/fl mice, which allow for the inducible ablation of Wls in CD133+ DP cells (Figure 2A). The CD133-CreERT2 mouse line expresses a recombined CreER fusion protein consisting of Cre recombinase and a mutated ligand-binding domain of the human estrogen receptor in CD133+ cells [28]. The activation of Cre recombinase requires administration of TAM or 4-hydroxytamoxifen (4-OHT) [28,31]. It has been confirmed that Cre recombinase encoded by the CreER transgene in CD133-CreERT2 mice is uniquely expressed in CD133+ DP cells in skin [27]. The Wls allele in the Wlsfl/fl mouse is flanked by two loxP sites, which, when recombined, will result in a null allele [20]. The administration of TAM at different time points to CD133-CreERT2; Wlsfl/fl mice induces efficient ablation of Wls in CD133+ DP cells during different stages of the hair growth cycle. Control littermates were either of genotypes CD133-CreERT2; Wlsfl/+ or CD133-CreERT2; Wls+/+. The ablation of Wls in CD133-CreERT2; Wlsfl/fl mice and littermate controls was induced by IP injection of TAM from P21 for 7 consecutive days until P27 (Figure 2B). Mid-dorsal skin biopsies were harvest at P28, P30, P32, P35, P40 and P45 for analysis. Efficient Wls ablation was confirmed by PCR analysis of isolated genomic DNA from induced CD133-CreERT2; Wlsfl/fl mice (lanes 1 and 4 in the top panel of DNA agarose gel) showing efficient Wls recombination (Figure 2C). As reported before [32], we have observed that administration of TAM to mice led to a delay of anagen onset by 2–3 days. Therefore, timing of the hair cycle stages in our study differs slightly from those in untreated mice [3]. At P28, when hair follicles entered early anagen, hair follicles in CD133-CreERT2; Wlsfl/fl mice (Figure 2E) and controls (Figure 2D) did not show any obvious morphological difference. Starting from P30, hair follicles in control mice continued to develop normally (Figure 2F). On the contrary, the growth of hair follicles in CD133-CreERT2; Wlsfl/fl mice appeared slightly delayed (Figure 2G), suggesting that Wnt ligands from CD133+ DP cells may start to play an important role during this stage. Consistent with this observation, hair follicles in CD133-CreERT2; Wlsfl/fl (Figure 2I) grew slower than those in control (Figure 2H) at P32. As shown in Figure 2J, while hair follicles in controls entered anagen IV at P35, hair follicles in CD133-CreERT2; Wlsfl/fl skin failed to show the same advancement (Figure 2K). At P40, control hair follicles were still present in later anagen stages (Figure 2L), whereas mutant hair follicles had already regressed into catagen (Figure 2M). At P45, hair follicles in CD133-CreERT2; Wlsfl/fl skin were in telogen (Figure 2O), whereas control hair follicles just entered the catagen stage (Figure 2N). We counted the numbers of hair follicles that entered different stages of the hair cycle in P38–40 skin. As shown in Figure 2P, an average of 85% of hair follicles appeared in anagen stage in control mice, whereas an average of 60% of hair follicles had already progressed to catagen stage in CD133-CreERT2; Wlsfl/fl mice. To determine whether Wls ablation would cause any accumulation of Wnt protein, we isolated CD133+ DP cells from P30 CD133-CreERT2; Wlsfl/fl and control skin tissues after tamoxifen induction by FACS for Wnt5a immunoblotting. Wnt5a has been repeatedly shown as a major DP signature protein in mature hair follicles [16,17]. As shown in Figure 2Q, Wnt5a was expressed in CD133+ DP cells isolated from both CD133-CreERT2; Wlsfl/fl and control hair follicles. The intensity of Wnt5a expression bands was normalized to β-actin expression and compared between CD133-CreERT2; Wlsfl/fl and control samples. As shown in Figure 2R, there was no indication of Wnt5a protein accumulation in CD133+ Wls-deficient DP cells. Delayed hair growth and rapid regression in CD133-CreERT2; Wlsfl/fl mice are associated with inhibition of matrix cell differentiation To determine whether Wls deletion in CD133+ DP cells had any effects on the hair follicle structure, the expression of Sox9 (outer root sheath) and Gata3 (inner root sheath) were analyzed by immunofluorescence staining [33]. At P32, expression levels of Gata3 (Figure 3B) and Sox9 (Figure 3F) were much lower in mutant than those of control hair follicles (Figure 3A,E). By P35, decreased Gata3 (Figure 3D) and Sox9 (Figure 3H) expression were still apparent in CD133-CreERT2; Wlsfl/fl hair follicles (Figure 3C,G). However, expression of K15, a marker for hair follicle stem cells (HFSCs) in the bulge and secondary hair germ, was readily seen in mutant hair follicles, and the level of expression was comparable between CD133-CreERT2; Wlsfl/fl and control hair follicles at P32 and P35 (Figure 3I–L), suggesting that delayed hair growth and rapid regression in CD133-CreERT2; Wlsfl/fl was not caused by major changes in the stem cell compartment. Wnt/β-catenin signaling plays a central role in regulating hair follicle matrix cell proliferation and differentiation [34]. To determine whether the lack of hair growth in mutant mice was caused by inhibition of matrix cell proliferation and differentiation due to the blockage of Wnt export from CD133+ DP cells, we evaluated the expression of Ki67 and Lef1. Ki67 is a cell proliferation marker for matrix keratinocytes that surround the DP (transit-amplifying cells), and Lef1 is required for the differentiation of hair shaft progenitor cells [35]. At P32 and P35, there was intense labeling of matrix cells in mutant hair follicles for Ki67 (Figure 3N,P). The number of Ki67-positive cells was slightly reduced in matrix cells of CD133-CreERT2; Wlsfl/fl hair follicles and possibly due to the more advanced anagen stages in control hair follicles (Figure 3M,O). However, there was an obvious lack of Lef1+ hair shaft precursor cells in CD133-CreERT2; Wlsfl/fl hair follicles at P32 (Figure 3R) when compared with the high number of Lef1+ cells in control hair follicles (Figure 3Q). At P35, Lef1+ cells appeared in CD133-CreERT2; Wlsfl/fl hair follicles (Figure 3T), albeit at lower numbers as compared with controls (Figure 3S). These data show that Wnt ligands from CD133+ DP cells are required for hair matrix differentiation but not the maintenance of hair follicle stem cells. The biological activities of hair follicle keratinocytes and melanocytes are co-ordinated in order to produce mature pigmented hairs. To determine whether Wls deficiency in CD133+ DP cells affected melanocytes in the hair follicles, we evaluated the expression of Mitf by immunofluorescence staining [36]. As shown in Figure 3V, X, the numbers of Mitf-positive melanocytes were significantly decreased in CD133-CreERT2; Wlsfl/fl hair follicles when compared with control hair follicles (Figure 3U,W). Wls ablation in CD133+ DP cells leads to a reduced DP compartment and loss of versican expression In vitro DP cell cultures require the addition of exogenous Wnt ligands to maintain hair inductivity [18,19]. To ask whether Wnt ligands from CD133+ DP cells regulate the DP in an autocrine manner, we first examined the expression of AP, which is a specific marker for the DP [37]. As shown in Figure 4B, at P32, the size of the DP in CD133-CreERT2; Wlsfl/fl mice was smaller than that in control mice (Figure 4A). Similarly, at P35, the size of mutant DP compartment (Figure 4D) was smaller than that of control DPs (Figure 4C). By manually counting, mutant DPs contained a mean of 9 DP cells at P32, whereas control DP had a mean of 19 DP cells in each hair follicle (Figure 4E). At P35, there was a mean of 11 DP cells in mutant hair follicles, compared with 21 in controls (Figure 4F; n = 30 hair follicles from control mice and HFs from mutant mice). We also examined the expression of versican by immunostaining. Versican is a marker for DP cells of anagen hair follicles [38]. At P32, expression of versican was easily detectable in control DPs (Figure 4G), but completely absent from CD133-CreERT2; Wlsfl/fl DP (Figure 4H). There was a weak versican expression in mutant DP at P35 (Figure 4J), although appearing much weaker than control DPs (Figure 4I). Thus, Wnt ligands from CD133+ DP cells contribute to the proper maintenance of DPs and the expression of anagen DP markers. Ablation of Wls in CD133+ DP cells during mid-anagen leads to shortened hair follicle growth stage To exclude the possibility that delayed growth and accelerated regression in CD133-CreERT2; Wlsfl/fl hair follicles were due to the loss of Wls at anagen onset, we designed an alternative strategy to induce Wls ablation in CD133+ DP cells from P30 after anagen onset (Figure 5A). The Cre activity induction lasted for a maximum of 7 days depending on when skin biopsies were collected. There was no significant difference in hair follicle histology between CD133-CreERT2; Wlsfl/fl and control littermates at P35 (data now shown). However, at P40, 10 days after initial induction of Wls loss, control mice showed darker skin color because of hair growth (Figure 5B). Skin of CD133-CreERT2; Wlsfl/fl mice appeared lighter, indicating that the growth of hair follicles was blocked. Histological analysis confirmed this observation. As shown in Figure 5C, hair follicles in control littermates had entered full anagen, whereas a majority of hair follicles in mutant skin either were arrested at early anagen stages or regressed prematurely and entered catagen (Figure 5D). Overall, no hair follicles in CD133-CreERT2; Wlsfl/fl mice appeared in full anagen, whereas a majority of hair follicles in control littermates were in anagen to early catagen stages (Figure 5E). Similar to our previous set of experiments where we induced Wls ablation starting in early anagen, Lef1 expression was also remarkably reduced in CD133-CreERT2; Wlsfl/fl hair follicles at P35 (Figure 6B) and P40 (Figure 6D). Interestingly, while the expression of versican was maintained in mutant hair follicles at P35 (Figure 6F), it was clearly lost at P40 (Figure 6H). This was also the case for Ki67 expression. While mutant hair follicles exhibited strong Ki67 expression at P35 (Figure 6J), the level of Ki67 expression was significantly decreased at P40 (Figure 6L) when compared with control hair follicles (Figure 6K). The findings show that Wnt ligands from CD133+ DP cells are important for postnatal hair growth by regulating events in both the epithelial and mesenchymal compartments. Discussion Identification of the intricacies of EMIs during the hair follicle growth cycle is a prerequisite to improve our abilities to artificially generate hair follicles for clinical applications. Wnt signaling is a critical player in EMIs [39]. To better understand how Wnt signals regulate cell–cell interactions between epithelial and mesenchymal compartments during the hair growth cycle, more detailed studies on the source and nature of Wnt ligands that are responsible for triggering Wnt/β-catenin signaling are required. Although the requirements of Wnt ligands from the hair follicle epithelium have been well reported [12–14,25], there is a surprising lack of progress in the understanding of the role of Wnt ligands secreted from the DP. Here, we have started to explore the role of Wnt ligands expressed in the DP and how they influence postnatal hair growth. Our study clearly demonstrates that Wnt ligands generated by CD133+ DP cells are important for postnatal hair regenerative cycling by exerting influences over both the hair follicle epithelium and the DP. For the purpose of the study, we have introduced the Wls conditional knockout mouse (Wlsfl/fl) as our genetic tool since Wls is essential for Wnt secretion from Wnt-producing cells [24]. Although Wls might also have other functions, it has been clearly demonstrated in several reports that ablation of Wls blocks the release of Wnt proteins and results in the blockade of the activation of Wnt/β-catenin signaling in Wnt target cells [22,23]. We and several other groups have recently demonstrated that Wls ablation in the hair follicle epithelium blocks postnatal hair cycling [12–14,25]. Owing to the pivotal importance of the DP in hair follicle biology, key components of the Wnt signaling pathway have been investigated in the DP. However, the focus so far has generally been on β-catenin with the notable exception of Wnt5a [17]. Therefore, the genetic approach to ablate Wls expression in DP cells may offer an opportunity for us not only to further explore Wnt-mediated dynamic interactions between the epithelial compartment and the DP, but also to identify the nature and identity of Wnt ligands produced in the DP during the postnatal hair cycle. An unique CD133+ cell population exists in the DP during the anagen stage [21]. It has been convincingly demonstrated that CD133+ DP cells are a subpopulation in the DP that has the ability to induce hair follicles when mixed with keratinocytes [26]. This observation clearly suggests that CD133+ DP cells may possess the ability to produce and secret Wnt ligands required for adult hair follicle growth. Our model, CD133-CreERT2; Wlsfl/fl, is the first to actually show that Wnt ligands originating from CD133+ DP cells are essential for adult hair follicle growth in vivo. Because CD133+ DP cells are actually not present at the anagen onset, we focused our attention on the role of Wnt ligands in regulating hair growth and how they affect the activity of the DP and adjacent keratinocytes, mainly matrix and hair shaft precursor cell populations. It is poorly understood how these epithelial cell populations are actually regulated by the DP and how Wnts originating in the DP are used to regulate proliferation and differentiation of matrix cell populations. Surprisingly, we found that Wls ablation in CD133+ DP cells inhibits the proliferation and differentiation of matrix cells and progeny cells derived from them. At P32, there was a lack of Lef1 expression in matrix cells and hair shaft precursor cells that eventually differentiate into inner root sheath and hair shaft cells. Furthermore, the number of mature melanocytes in the hair matrix was also reduced upon Wls ablation in CD133+ DP cells. This phenotype can be best explained taking into consideration that the behavior of keratinocytes, melanocytes and DP cells is co-ordinated and synchronized during hair growth [40]. Consequently, changes in one cell population should at some level result in changes in the other cell population during anagen progression. Although it is unclear whether these interactions between melanocytes, keratinocytes and fibroblasts solely and directly rely on Wnt ligands from the DP or are indirectly influenced by the lack of unknown molecules from the DP that are affected by Wls deficiency in CD133+ DP cells, it is important to note that Wnt/β-catenin signaling is highest in the bulb region of the hair follicle. Based on the canonical model of Wnt signaling, this activity is induced by actual Wnt ligands. Our data strongly suggest that one possible source of these Wnt ligands at this stage is the DP, although there clearly exists the possibility of Wnt redundancy between the epithelium and the DP. Several Wnt ligands have, at least in in vitro systems, exhibited the ability to maintain DP cell function and DP inductive capacity [19]. In vivo, Wnt ligands from DP cells may function as autocrine factors that induce nuclear localization of β-catenin within DP cells, which can be detected in many DP cells during anagen. The Corin+ DP cell population requires β-catenin to maintain proliferation of matrix keratinocytes at the base of the follicle, and loss of β-catenin in these Corin+ DP cells also results in catagen induction and prevents subsequent anagen induction [32]. However, these results only indicate that DP cells have to receive Wnt signals to initiate the canonical Wnt/β-catenin signaling pathway. The origin, timing and exact nature of this Wnt signal has not been fully addressed in detail. In the present study, we have shown that CD133+ DP cells produce Wnt5a, a major DP Wnt, during early anagen. However, it remains to be seen whether this DP subpopulation generates other Wnt signals as well. Regardless, it will be important to identify Wnt ligands that are produced in the DP for hair follicle growth and regeneration in order to improve approaches to treat hair growth disorders such as alopecia. Data from our study suggest that Wnt ligands generated by CD133+ DP cells function within the DP potentially in an autocrine fashion. Support for this interpretation stems from a lack of versican expression, a Wnt signaling target gene [41], in the DPs of P32 hair follicles upon Wls ablation in CD133+ DP cells. Versican is a specific marker for anagen DP [42], and its reduction indicates that DP cells have lost biological properties that are needed to maintain hair inductivity. Deficiencies in the DP functionality may also contribute to the abnormal hair phenotypes in CD133-CreERT2; Wlsfl/fl mice. These results suggest that Wnt ligands generated by DP cells during anagen stages possibly contribute to active Wnt/β-catenin signaling in both epithelial and mesenchymal compartments of hair follicles to promote hair growth. Functional redundancy of Wnt ligands generated by hair follicle keratinocytes and DP cells needs always to be taken into consideration. This could well be the case for CD133-CreERT2; Wlsfl/fl hair follicle as we saw the weak appearance of Lef1 expression in hair matrix and precursor cells from P35 onwards. Wnt ligands made by hair follicle matrix cells and inner root sheath cells could eventually substitute for the missing DP-derived Wnts. On the other hand, since Wnt ligands, as activators, could travel for a short distance in the hair follicle, it is possible that Wnt ligands from the hair follicle epithelium could eventually diffuse from the epithelium to the DP and subsequently activate Wnt/β-catenin signaling in DP cells. This compensatory mechanism could be one explanation as to why the DP maintained weak versican expression at P35. In addition, there remains the possibility that there are other DP populations that have the ability to generate Wnt ligands when hair follicles enter different stages of the hair growth cycle and substitute the functional activity of CD133+ DP cells. To address this issue, we may need to develop and use additional genetic tools to target CD133− DP cell populations. The lack of changes in anagen onset in CD133-CreERT2; Wlsfl/fl hair follicles is not a surprise as endogenous CD133 expression and the resulting Cre activity is absent at anagen onset but initiated shortly after. Therefore, this particular cell population is not available to interact with HFSCs during anagen onset. This limitation of our CD133-CreERT2-based model means that the role of DP-derived Wnts at anagen onset before the CD133+ cell population emerges cannot be studied in the model we present here. Further development of alternative models in combination with our data will help to refine our understanding of the role of Wnts in DP biology. Regardless, it is also unlikely that the bulge stem cell population would be fundamentally affected by the lack of Wnt ligands from the DP after anagen onset. This is supported by our observation that HFSCs and secondary hair germ were not disturbed to an obvious degree in CD133-CreERT2; Wlsfl/fl mutant hair follicles, although we cannot exclude the possibility that intrinsic properties of HFSCs were actually affected contributing to later phenotypes. Therefore, in future studies, we will address the long-term effects of Wls ablation in CD133+ DP cells during the first hair cycle and observe whether the second hair cycle or induced hair growth by depilation is affected, which will definitely answer the question as to whether the activation of HFSCs requires Wnt ligands from CD133+ DP cells. In summary, our study demonstrates that CD133+ DP cells generate Wnt ligands that are important for postnatal hair regenerative cycling. Without Wls, and therefore without Wnts, from CD133+ DP cells, hair follicles are not able to sustain a normal anagen. This finding carries important implications as it provides a new insight into the biology of EMIs and DP cell biology that may lead to the discovery and development of new targets for treating hair growth disorders. We thank Dr Richard J. Gilbertson and the St Jude Children’s Research Hospital for CD133-CreERT2 (Prom1C-L) mice. Funding This work was supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS) R03 AR062788 (Y.Z.). Abbreviations 4-OHT 4-hydroxytamoxifen AP alkaline phosphatase APC Adenomatous polyposis coli Axin axis inhibition protein BrdU bromodeoxyuridine CD133+ CD133-positive DP dermal papilla EMIs epithelial–mesenchymal interactions Fzd Frizzled GSK-3β Glycogen synthase kinase 3 beta H&E hematoxylin and eosin HFSCs hair follicle stem cells IP intraperitoneal Mitf microphthalmia-associated transcription factor P25 postnatal day 25 PBS phosphate-buffered saline PFA paraformaldehyde TAM tamoxifen Wls Wntless Figure 1 Wls is expressed in the DP during hair growth anagen except onset Mid-dorsal skin biopsies of wild-type C57BL/6 (B6) mice were collected at P21, P25, P28, P32, P35, P40 and P45, and processed for paraffin sections. Sections were analyzed for Wls (A–F) and LEF1 (M–R) expression by immunofluorescence staining at each time point as indicated. (G–L) Mid-dorsal skin biopsies of TOPGAL mice were collected at indicated age for X-gal whole-mount staining. The DP was circled by either white or black dash line in each hair follicle. Red triangles indicate either Wls-positive or Lef1-positive cells outside the DP in their respective pictures. Orange stars denote keratinocytes that are Lef1-positive but Wls-negative. Yellow arrows indicate inner root sheath cells in the hair follicle. For every indicated mouse age, at least three mice were analyzed. Scale bar: 100 µm. Figure 2 Ablation of Wls in CD133+ DP cells delays hair growth and induces premature hair regression (A) Illustration of the CD133-CreERT2; Wlsfl/fl transgenic mouse model, which allows specific Wls ablation in CD133+ DP cells upon TAM or 4-OHT administration. (B) Time scheme for TAM administration during early anagen stage and skin biopsy collection. CD133-CreERT2; Wlsfl/fl mice and control littermates were administered of tamoxifen daily by IP injection for 7 days starting from P21 to P27. Mid-dorsal skin biopsies were harvested for examination at P28, P30, P32, P35, P40 and P45. (C) Recombination of floxed Wls alleles in CD133+ DP cells was confirmed by PCR analysis. Upper DNA agarose gel picture: genomic DNA extracted from skin biopsies was PCR genotyped using P1/P4 primer set to identify wild-type Wls allele and recombined Wls allele. Lower DNA agarose gel picture: genomic DNA extracted from skin biopsies was PCR genotyped using P2/P4 primer set to identify wild-type Wls allele and floxed Wls allele. Lanes labeled with same number in upper and lower DNA gel pictures were genotyping results of genomic DNA extract from same mouse. Genotype of CD133-CreERT2 for each mouse is labeled above each lane of top DNA agarose gel picture. ‘+’ means mouse carried CD133-CreERT2 transgene, while ‘−’ means mouse did not carry CD133-CreERT2. (D–O) H&E-stained skin sections from CD133-CreERT2; Wlsfl/fl (D, F, H, J, L and N) and control littermates (E, G, I, K, M and O) as labeled. Scale bar: 100 µm (n = 3). (P) Comparison of hair follicle numbers that were present at different hair cycle stages between control and CD133-CreERT2; Wlsfl/fl mice from P38 to P40. A minimum of three skin biopsies from three pairs of mutant mice and control littermates were manually counted. Two-tailed paired Student’s t-test was employed. *P < 0.05. (Q) Generation of Wnt5a protein in CD133+ DP cells was confirmed by Western blot analysis. Two pairs of CD133-CreERT2; Wlsfl/fl mice (MU1, 2) and control littermates (CT1, 2) are shown. (R) Intensity of Wnt5a band was shown by scanning X-ray film and normalized to β-actin band. Three pairs of mice were analyzed. Figure 3 Wls deficiency in CD133+ DP cells inhibits matrix cells differentiation Paraffin sections of mid-dorsal skin biopsies of P32 or P35 old CD133-CreERT2; Wlsfl/fl mice and control littermates were analyzed by immunofluorescent staining to assess the expression of following markers: Gata3 for inner root sheath (control: A and C; mutant: B and D); Sox9 for outer root sheath (control: E and G; mutant: F and H); K15 for hair follicle stem cells and secondary hair germ cells (control: I and K; mutant: J and L); Ki67 for proliferating matrix cells (control: M and O; mutant: N and P); Lef1 for proliferating matrix cells and hair shaft precursor cells (control: Q and S; mutant: R and T); MITF for mature melanocytes (control: U and W; mutant: V and X). Sections were nuclear counterstained with DAPI (blue). Images shown are representative of at least three replicates at each indicated age. Scale bars: 200 µm. Figure 4 Wls deficiency in CD133+ DP cells affects DP compartments (A–D) Frozen sections of mid-dorsal skin biopsies of P32 or P35 old CD133-CreERT2; Wlsfl/fl mice (B and D) and control littermates (A and C) were immunostained using a VECTOR Red AP Substrate Kit to reveal the DP compartment (n = 3). (E and F) AP+ DP cells were manually counted and compared between CD133-CreERT2; Wlsfl/fl mice and control littermates at P32 (E) and P35 (F). A minimum of 30 hair follicles was counted from three mice of same genotypes (n = 3). (G–J) Expression of versican in the DP was evaluated by immunofluorescent staining (green color inside the white circle) (n = 3). Sections were co-stained with an antibody against cell proliferative marker BrdU (red) (control: G and I; mutant: H and J). Sections were nuclear counterstained with DAPI (blue). Scale bar = 200 µm. Figure 5 Ablation of Wls in CD133+ DP cells during mid-anagen induces premature hair regression (A) Time scheme for TAM administration during mid-anagen and sample collection. Briefly, CD133-CreERT2; Wlsfl/fl mice and control littermates were administered with TAM by IP injection at P30 for 5 or 7 days based on the date of skin biopsy collection. (B) Representative pictures of skin color of CD133-CreERT2; Wlsfl/fl mouse and control littermate at P40 after shaving (n = 3). (C and D) H&E-stained skin biopsies of CD133-CreERT2; Wlsfl/fl mouse (D) and control littermate (C) at P40 (n = 3). Scale bar: 100 µm. (E) Comparison of hair follicle numbers that were present at different hair cycle stages between CD133-CreERT2; Wlsfl/fl mice and control mice at P40 (n = 3). A minimum of three skin biopsies from three pairs of mutant mice and control littermates were manually counted. Two-tailed paired Student’s t-test was employed. Figure 6 Wls deficiency in CD133+ DP cells during mid-anagen affects both epithelial compartment and the DP Paraffin sections of skin biopsies from P35 and P40 CD133-CreERT2; Wlsfl/fl mice and control littermates (n = 3) after Wls ablation in CD133+ DP cells during mid-anagen stage (starting from P30) were analyzed for the expression of following markers: Lef1 (control: A and C; mutant B and D); Ki67 (control: E and G; mutant: F and H) and versican/BrdU (control: I and K; mutant: J and L). Sections were nuclear counterstained with DAPI (blue). Scale bar = 200 µm. Author Contribution L.Z. and K.Y. performed the experiments. A.C. and R.L. generated and provided the animal model. Y.Z. and T.A. contributed to the analysis and interpretation of the data. Y.Z. supervised the study and wrote the manuscript with feedback from all authors. Competing Interests The Authors declare that there are no competing interests associated with the manuscript. References 1 Alonso L Fuchs E The hair cycle J. Cell Sci 2006 119 391 393 16443746 2 Botchkarev VA Kishimoto J Molecular control of epithelial–mesenchymal interactions during hair follicle cycling J. Invest. Dermatol. Symp. Proc 2003 8 46 55 3 Müller-Röver S Kerstin F Paus R Handjiski B van der Veen C Eichmüller S A comprehensive guide for the accurate classification of murine hair follicles in distinct hair cycle stages J. Invest. Dermatol 2001 117 3 15 11442744 4 Chi W Wu E Morgan BA Dermal papilla cell number specifies hair size, shape and cycling and its reduction causes follicular decline Development 2013 140 1676 1683 23487317 5 Cohen J The transplantation of individual rat and guinea-pig whisker papillae J. Embryol. Exp. Morphol 1961 9 117 127 PMID: 13694386 13694386 6 Jahoda CA Reynolds AJ Oliver RF Induction of hair growth in ear wounds by cultured dermal papilla cells J. Invest. Dermatol 1993 101 584 590 8409527 7 Millar SE Molecular mechanisms regulating hair follicle development J. Invest. Dermatol 2002 118 216 225 11841536 8 Millar SE WNTs: multiple genes, multiple functions J. Invest. Dermatol 2003 120 7 8 12535190 9 Willert K Nusse R β-catenin: a key mediator of Wnt signaling Curr. Opin. Genet. Dev 1998 8 95 102 9529612 10 Cadigan KM Waterman ML TCF/LEFs and Wnt signaling in the nucleus Cold Spring Harb. Perspect. Biol 2012 4 a007906 23024173 11 Cruciat C-M Niehrs C Secreted and transmembrane wnt inhibitors and activators Cold Spring Harb. Perspect. Biol 2013 5 a015081 23085770 12 Choi YS Zhang Y Xu M Yang Y Ito M Distinct functions for Wnt/β-catenin in hair follicle stem cell proliferation and survival and interfollicular epidermal homeostasis Cell Stem Cell 2013 13 720 733 24315444 13 Fu J Hsu W Epidermal Wnt controls hair follicle induction by orchestrating dynamic signaling crosstalk between the epidermis and dermis J. Invest. Dermatol 2013 133 890 898 23190887 14 Myung PS Takeo M Ito M Atit RP Epithelial Wnt ligand secretion is required for adult hair follicle growth and regeneration J. Invest. Dermatol 2013 133 31 41 22810306 15 Chen D Jarrell A Guo C Lang R Atit R Dermal β-catenin activity in response to epidermal Wnt ligands is required for fibroblast proliferation and hair follicle initiation Development 2012 139 1522 1533 22434869 16 Reddy S Andl T Bagasra A Lu MM Epstein DJ Morrisey EE Characterization of Wnt gene expression in developing and postnatal hair follicles and identification of Wnt5a as a target of Sonic hedgehog in hair follicle morphogenesis Mech. Dev 2001 107 69 82 11520664 17 Hu B Lefort K Qiu W Nguyen B-C Rajaram RD Castillo E Control of hair follicle cell fate by underlying mesenchyme through a CSL–Wnt5a–FoxN1 regulatory axis Genes Dev 2010 24 1519 1532 20634318 18 Kishimoto J Burgeson RE Morgan BA Wnt signaling maintains the hair-inducing activity of the dermal papilla Genes Dev 2000 14 1181 1185 PMCID: PMC316619 10817753 19 Shimizu H Morgan BA Wnt signaling through the β-catenin pathway is sufficient to maintain, but not restore, anagen-phase characteristics of dermal papilla cells J. Invest. Dermatol 2004 122 239 245 15009701 20 Carpenter AC Rao S Wells JM Campbell K Lang RA Generation of mice with a conditional null allele for Wntless Genesis 2010 48 554 558 20614471 21 Ito Y Hamazaki TS Ohnuma K Tamaki K Asashima M Okochi H Isolation of murine hair-inducing cells using the cell surface marker prominin-1/CD133 J. Invest. Dermatol 2007 127 1052 1060 17185982 22 Bänziger C Soldini D Schütt C Zipperlen P Hausmann G Basler K Wntless, a conserved membrane protein dedicated to the secretion of Wnt proteins from signaling cells Cell 2006 125 509 522 16678095 23 Bartscherer K Pelte N Ingelfinger D Boutros M Secretion of Wnt ligands requires Evi, a conserved transmembrane protein Cell 2006 125 523 533 16678096 24 Ching W Nusse R A dedicated Wnt secretion factor Cell 2006 125 432 433 16678089 25 Huang S Zhu X Liu Y Tao Y Feng G He L Wls is expressed in the epidermis and regulates embryonic hair follicle induction in mice PLoS ONE 2012 7 e45904 23029304 26 Driskell RR Juneja VR Connelly JT Kretzschmar K Tan DW-M Watt FM Clonal growth of dermal papilla cells in hydrogels reveals intrinsic differences between Sox2-positive and -negative cells in vitro and in vivo J. Invest. Dermatol 2012 132 1084 1093 22189784 27 Kaushal GS Rognoni E Lichtenberger BM Driskell RR Kretzschmar K Hoste E Fate of prominin-1 expressing dermal papilla cells during homeostasis, wound healing and Wnt activation J. Invest. Dermatol 2015 135 2926 2934 26288357 28 Zhu L Gibson P Currle DS Tong Y Richardson RJ Bayazitov IT Prominin 1 marks intestinal stem cells that are susceptible to neoplastic transformation Nature 2009 457 603 607 19092805 29 Zhou L Yang K Randall Wickett R Zhang Y Dermal fibroblasts induce cell cycle arrest and block epithelial-mesenchymal transition to inhibit the early stage melanoma development Cancer Med 2016 5 1566 1579 27061029 30 DasGupta R Fuchs E Multiple roles for activated LEF/TCF transcription complexes during hair follicle development and differentiation Development 1999 126 4557 4568 PMID: 10498690 10498690 31 Metzger D Clifford J Chiba H Chambon P Conditional site-specific recombination in mammalian cells using a ligand-dependent chimeric Cre recombinase Proc. Natl Acad. Sci. USA 1995 92 6991 6995 7624356 32 Enshell-Seijffers D Lindon C Kashiwagi M Morgan BA β-catenin activity in the dermal papilla regulates morphogenesis and regeneration of hair Dev. Cell 2010 18 633 642 20412777 33 Lynch MH O’Guin WM Hardy C Mak L Sun TT Acidic and basic hair/nail (“hard”) keratins: their colocalization in upper cortical and cuticle cells of the human hair follicle and their relationship to "soft” keratins J. Cell Biol 1986 103 2593 2606 2432071 34 Rishikaysh P Dev K Diaz D Qureshi WM Filip S Mokry J Signaling involved in hair follicle morphogenesis and development Int. J. Mol. Sci 2014 15 1647 1670 24451143 35 Merrill BJ Gat U DasGupta R Fuchs E Tcf3 and Lef1 regulate lineage differentiation of multipotent stem cells in skin Genes Dev 2001 15 1688 1705 11445543 36 Levy C Khaled M Fisher DE MITF: master regulator of melanocyte development and melanoma oncogene Trends Mol. Med 2006 12 406 414 16899407 37 Iida M Ihara S Matsuzaki T Hair cycle-dependent changes of alkaline phosphatase activity in the mesenchyme and epithelium in mouse vibrissal follicles Dev. Growth Differ 2007 49 185 195 17394597 38 Kishimoto J Ehama R Wu L Jiang S Jiang N Burgeson RE Selective activation of the versican promoter by epithelial-mesenchymal interactions during hair follicle development Proc. Natl Acad. Sci. USA 1999 96 7336 7341 10377415 39 van Genderen C Okamura RM Farinas I Quo RG Parslow TG Bruhn L Development of several organs that require inductive epithelial-mesenchymal interactions is impaired in LEF-1-deficient mice Genes Dev 1994 8 2691 2703 7958926 40 Rabbani P Takeo M Chou W Myung P Bosenberg M Chin L Coordinated activation of Wnt in epithelial and melanocyte stem cells initiates pigmented hair regeneration Cell 2011 145 941 955 21663796 41 Rahmani M Read JT Carthy JM McDonald PC Wong BW Esfandiarei M Regulation of the versican promoter by the β-catenin-T-cell factor complex in vascular smooth muscle cells J. Biol. Chem 2005 280 13019 13028 15668231 42 du Cros DL LeBaron RG Couchman JR Association of versican with dermal matrices and its potential role in hair follicle development and cycling J. Invest. Dermatol 1995 105 426 431 7665924
PMC005xxxxxx/PMC5121000.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101120028 22411 Dev Cell Dev. Cell Developmental cell 1534-5807 1878-1551 27720608 5121000 10.1016/j.devcel.2016.09.005 NIHMS821735 Article Acetylation of VGLL4 Regulates Hippo-YAP Signaling and Postnatal Cardiac Growth Lin Zhiqiang 1* Guo Haidong 12 Cao Yuan 13 Zohrabian Sylvia 1 Zhou Pingzhu 1 Ma Qing 1 VanDusen Nathan 1 Guo Yuxuan 1 Zhang Jin 1 Stevens Sean M. 1 Liang Feng 4 Quan Qimin 4 van Gorp Pim R. 5 Li Amy 6 dos Remedios Cristobal 6 He Aibin 7 Bezzerides Vassilios J. 1 Pu William T. 189* 1 Department of Cardiology, Boston Children's Hospital, 300 Longwood Avenue, Boston, MA 02115, USA 2 Department of Anatomy, School of Basic Medicine, Shanghai University of Traditional Chinese Medicine, Shanghai 201203, China 3 Peking University, Fifth School of Clinical Medicine, Beijing 100730, China 4 Rowland Institute at Harvard, Harvard University, Cambridge, MA 02142, USA 5 Department of Cardiology, Leiden University Medical Center, 2300 RC Leiden, the Netherlands 6 Department of Anatomy & Histology, Bosch Institute, University of Sydney, Sydney, NSW 2006, Australia 7 Institute of Molecular Medicine, Peking University, PKU-Tsinghua U Joint Center for Life Sciences, Beijing 100871, China 8 Harvard Stem Cell Institute, Harvard University, Cambridge, MA 02138, USA 9 Lead Contact * Correspondence: [email protected] (Z.L.), [email protected] (W.T.P.) 27 10 2016 6 10 2016 21 11 2016 21 11 2017 39 4 466479 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. SUMMARY Binding of the transcriptional co-activator YAP with the transcription factor TEAD stimulates growth of the heart and other organs. YAP overexpression potently stimulates fetal cardiomyocyte (CM) proliferation, but YAP's mitogenic potency declines post-natally. While investigating factors that limit YAP's postnatal mitogenic activity, we found that the CM-enriched TEAD1 binding protein VGLL4 inhibits CM proliferation by inhibiting TEAD1-YAP interaction and by targeting TEAD1 for degradation. Importantly, VGLL4 acetylation at lysine 225 negatively regulated its binding to TEAD1. This developmentally regulated acetylation event critically governs postnatal heart growth, since overexpression of an acetylation-refractory VGLL4 mutant enhanced TEAD1 degradation, limited neonatal CM proliferation, and caused CM necrosis. Our study defines an acetylation-mediated, VGLL4-dependent switch that regulates TEAD stability and YAP-TEAD activity. These insights may improve targeted modulation of TEAD-YAP activity in applications from cardiac regeneration to cancer. INTRODUCTION Control of organ growth is fundamental to animal development and organ homeostasis. Unrestrained activity of growth-promoting pathways leads to cancer, whereas targeted activation of these pathways may be a strategy for organ regeneration (Pan, 2010; Lin and Pu, 2015). One area with great need for advances in regenerative medicine is heart disease. Heart failure is the world's leading cause of death, and its prevalence is expected to further increase as the population ages (Heidenreich et al., 2013). Cardiomyocyte (CM) loss is a central pathogenic mechanism in heart failure, but limited endogenous regenerative capacity in the adult heart has precluded development of therapeutic approaches to efficiently replace these lost CMs (Lin and Pu, 2014). Unlike adult CMs, fetal CMs robustly proliferate to match the rapid growth of the embryo. This transition from CM proliferation to cell-cycle exit occurs in the first week of life in the mouse (Walsh et al., 2010). The mechanisms that regulate this neonatal cell-cycle exit are poorly understood, and greater insights would inform efforts to enhance cardiac regeneration. The transcriptional co-activator YAP (Yes-associated protein) is a key driver of organ growth (Pan, 2010). YAP binds to TEA domain (TEAD)-containing transcription factors (TEAD1–TEAD4) to activate transcription of cell-cycle and cell-survival genes and thereby promotes organ growth. The potent growth-promoting activity of the YAP-TEAD complex is closely regulated through incompletely understood signaling pathways. Among the most studied regulatory mechanisms is the Hippo kinase cascade, which phosphorylates YAP, leading to its nuclear exclusion (Huang et al., 2005). Both YAP and its regulation by the Hippo kinase cascade have been shown to be essential for normal heart development (Heallen et al., 2011; von Gise et al., 2012; Xin et al., 2011). YAP was necessary for fetal CM proliferation, and its activation through overexpression or Hippo inhibition was sufficient to drive massive fetal cardiac overgrowth. YAP activation likewise stimulated neonatal as well as adult CM proliferation, but the level of CM cell-cycle activity achieved diminished with postnatal age (Lin et al., 2014; Xin et al., 2013; Heallen et al., 2013). These data show that regulation of YAP activity is crucial for normal cardiac growth control. Moreover, they suggest that unknown mechanisms suppress YAP mitogenic activity in the postnatal heart. In addition to the Hippo kinase pathway, Hippo-independent YAP regulatory mechanisms also exist. For example, α-catenin, a cellular adhesion molecule, binds YAP under high cell density conditions, promoting its cytoplasmic sequestration by limiting its dephosphorylation (Schlegelmilch et al., 2011; Li et al., 2014). Recently, a new level of YAP-TEAD regulation was described. In Drosophila, the orthologs of YAP and TEAD are named Yorkie (Yki) and Scalloped (Sd), respectively. The gene Tgi was discovered in Drosophila screens for Yki-Sd antagonists (Koontz et al., 2013). TGI protein contains two TEAD-binding regions, named Tondu (TDU) domains, and competes with YKI for SD binding. By reducing YKI-SD activity and the transcription of YKI-SD target genes, TGI inhibited growth. In mammals, there are four TDU domain-containing proteins, vestigial-like 1 to 4 (VGLL1–VGLL4), with VGLL4 being the most closely related to TGI. Massive liver overgrowth driven by YAP was suppressed by VGLL4 (Koontz et al., 2013), indicating that VGLL4 is a potent inhibitor of YAP in mammalian cells. Profiling of VGLL4 across mouse tissues showed that it was most highly expressed in heart (Chen et al., 2004), suggesting that VGLL4 potentially suppresses postnatal cardiac YAP activity. Here we studied VGLL4 function in regulating cardiac YAP-TEAD activity and neonatal cardiac growth and function. We found that VGLL4 regulates both TEAD stability and its interaction with YAP. Moreover, VGLL4 acetylation at a key residue within the TDU domain regulates its binding to TEAD, revealing a novel YAP-TEAD regulatory mechanism. Acetylation of VGLL4 in neonatal heart was essential to limit its activity and thereby permit normal heart growth and function. RESULTS The Major Cardiac Interaction Partner of TEAD1 Changes with Postnatal Age During heart development, the Hippo-YAP pathway regulates CM proliferation (Heallen et al., 2011; von Gise et al., 2012; Xin et al., 2011). YAP and TEAD1 are terminal effectors of the Hippo-YAP pathway. VGLL4, a TEAD1-binding protein that antagonizes overexpressed YAP in the liver (Koontz et al., 2013), was previously reported to have cardiac-restricted RNA expression (Chen et al., 2004). To better understand YAP, TEAD1, and VGLL4 function in regulating organ growth, we measured their expression in several adult mouse tissues. YAP was widely expressed, as we demonstrated previously (von Gise et al., 2012). We detected robust VGLL4 expression in the heart, with lower levels also present in the brain, liver, and lung (Figure 1A). TEAD1 protein was abundant in the lung, less expressed in the heart, and undetectable in the other organs examined (Figure 1A). Focusing on the postnatal heart, we measured expression of these proteins at several different ages. Interestingly, VGLL4 expression increased from low levels in the newborn heart to high levels in the adult heart (Figure 1B). TEAD1 and YAP levels were anti-correlated with VGLL4 and decreased with age (Figure 1B). These changes in cardiac protein level did not correlate with changes in their corresponding transcripts (Figure 1C), indicating that the expression of these proteins is regulated post-transcriptionally. To determine whether these proteins were expressed in CMs or non-CMs in adult heart, we dissociated hearts and isolated purified cell populations. VGLL4 and TEAD1 were mainly expressed in CMs rather than non-CMs, whereas YAP was predominantly expressed in non-CMs (Figure 1D). Based on these protein expression changes, we hypothesized that TEAD1's primary interaction partner changes from YAP to VGLL4 between newborn and adult heart. To test this hypothesis, we generated a Tead1 knockin allele, Tead1fb, which contains the FLAG and AviTag (Bio) (He et al., 2012) epitope tags fused to the Tead1 C terminus (Tead1fb; Figures S1A and S1B). The Escherichia coli enzyme BirA specifically recognizes and biotinylates the Bio tag (de Boer et al., 2003), permitting high-affinity pull-down on immobilized streptavidin (SA). Tead1fb/fb; Rosa26BirA/BirA mice survived normally to weaning, had no overt phenotype (Figures S1C and S1D), and expressed biotinylated TEAD1fb (Figures S1E and S1F). We then precipitated TEAD1fb and its interacting proteins on SA from postnatal day 1 (P1), P8, and adult (P50) hearts (Figures 1E–1G). This revealed that TEAD1 and VGLL4 strongly interacted in the adult but not the neonatal (P1 or P8) heart (Figures 1E–1G). TEAD1 and YAP interaction showed the opposite pattern, with strong interaction detected in the neonatal heart and weaker interaction in the adult heart (Figures 1F and 1G). Precocious VGLL4 Overexpression Did Not Suppress Neonatal Cardiac Growth To further test the hypothesis that VGLL4 limits CM proliferation by reducing TEAD1-YAP interaction, we overexpressed VGLL4 in the newborn heart using adeno-associated virus serotype 9 (AAV9), an efficient cardiac gene delivery vector (Lin et al., 2014). We generated AAV9.VGLL4-GFP (AAV9.VGLL4) and AAV9.GFP, which express VGLL4-GFP fusion protein or GFP, respectively, from the cardiomyocyte-specific chicken cardiac troponin T (cTNT) promoter (Figure 2A), and injected them into P1 wild-type pups. Hearts were analyzed 7 days later. Immunoblots confirmed cardiac VGLL4-GFP expression (Figure 2A). Unexpectedly, AAV9.VGLL4 did not significantly change heart function or size compared with untreated (Ctrl) or AAV9.GFP-treated hearts (Figures 2B–2D). Staining for phospho-histone H3 (pH3), an M-phase cell-cycle marker, suggested that CM cell-cycle activity was not significantly changed by AAV9.VGLL4 compared with AAV-GFP (Figures 2E and 2F). Consistent with this observation, the expression of cell-cycle genes Aurka, Cdc20, and Ccna2 did not differ significantly between groups (Figure 2G). We investigated the effect of AAV9.VGLL4 on YAP-TEAD1 interaction. In adult heart, where TEAD1 robustly interacted with endogenous VGLL4, we confirmed that VGLL4-GFP bound TEAD1 (Figure S2A). This result indicated that the GFP tag did not disrupt VGLL4-TEAD1 interaction. However, we did not detect VGLL4-GFP binding to TEAD1 in the neonatal heart (Figure S2B), suggesting that the lack of robust interaction between VGLL4 and TEAD1 in neonatal heart was not solely due to lower neonatal VGLL4 expression. Rather, the finding that forced expression of VGLL4-GFP in neonatal heart was not sufficient to drive its interaction with TEAD1 led us to hypothesize that additional factors regulate this interaction. These additional factors would limit interaction of overexpressed VGLL4-GFP and TEAD1 and thereby could account for the lack of phenotype in AAV9.VGLL4-treated heart. VGLL4 Activity Is Regulated by Its TDU Domain Acetylation Post-translational modification is one potential regulatory mechanism that might govern VGLL4-TEAD1 interaction in the neonatal heart. Adopting a candidate strategy, we first investigated VGLL4 acetylation. Histone acetyltransferases such as p300 or CBP acetylate lysine resides of non-histone proteins, including transcription factors, in addition to histones (Chan and La Thangue, 2001). In co-immunoprecipitation (coIP) as-says, VGLL4 robustly interacted with p300, whereas its interaction with CBP was considerably weaker (Figure 3A). Moreover, p300 but not CBP heavily acetylated VGLL4 (Figure 3A). To identify VGLL4 acetylation sites, we co-expressed VGLL4-GFP fusion protein and p300 in HEK293T cells. Immunoprecipitated VGLL4-GFP was then analyzed by mass spectrometry. Several acetylation sites were identified, with the highest fraction of acetylated residues occurring at lysine 225 (K255; Figure 3B). To confirm this observation, we mutated VGLL4 K225 to arginine (VGLL4[R]), which is structurally similar to lysine but cannot be acetylated. In p300 co-transfected cells, compared with wild-type VGLL4, VGLL4[R] acetylation was significantly reduced (Figure 3C, ratio of KAc to total VGLL4 in lane 5 versus 3), consistent with VGLL4 being acetylated predominantly but not solely at K225. Vestigial-like family members interact with TEAD through their Tondu (TDU) domains (Koontz et al., 2013), and VGLL4 contains two TDU domains. K225 is located in the first TDU domain of human VGLL4 and is conserved among vertebrate VGLL4 proteins but not in TDU domains from other proteins (Figure 3D). Based on its location within the TEAD1-binding domain of VGLL4, we hypothesized that acetylation of K225 modulates VGLL4-TEAD interaction. To test this hypothesis, we synthesized peptides corresponding to the VGLL4 TDU domains, with or without K225 acetylation (Figure S3A). These V5 epitope-tagged VGLL4 peptides were co-incubated with recombinant, His-tagged TEAD1 (residues 211–427), which contains the YAP and VGLL4 interaction domains (Jiao et al., 2014) (Figures S3B and S3C). The interaction between VGLL4 peptide and His-TEAD1 [211–427] was measured by immunoprecipitation followed by western blotting (Figure 3E). The synthetic non-acetylated VGLL4-TDU domain peptide bound TEAD1 in a dose-dependent manner. In contrast, the acetylated VGLL4-TDU peptide did not detectably interact with TEAD1. To quantify this result, we used a recently developed nanoscale photonic interaction assay (Yang et al., 2014). His-TEAD1[211–427] was immobilized on a nanobeam sensor and then incubated with increasing concentrations of the VGLL4-TDU peptide. The non-acetylated VGLL4-TDU peptide induced a concentration-dependent resonance shift of the sensor (Figure 3F), indicative of binding to TEAD1. Fitting the curve to the Langmuir equation yielded a VGLL4-TEAD1 interaction affinity of 3.1 ± 1.3 nM. In contrast, the acetylated VGLL4-TDU peptide did not induce a resonance shift up to a peptide concentration of 1 μg/mL (Figure 3F). Together these results indicate that VGLL4 acetylation at K225 strongly impedes its binding to TEAD1. To study the effect of VGLL4 acetylation on VGLL4-TEAD1 interaction in a cellular context, we co-expressed VGLL4-GFP or VGLL4[R]-GFP with Tead1fb in 293T cells and evaluated their interaction by coIP. Even though TEAD1 level was lower in VGLL4[R]- than in VGLL4-expressing cells (see following section), TEAD1 co-precipitated significantly more VGLL4[R] (Figure 3G), suggesting that acetylation at K225 reduces VGLL4-TEAD1 interaction in vivo. Consistent with this result, VGLL4[R] inhibited TEAD1-YAP interaction with greater potency than wild-type VGLL4 (Figure 3G). We measured the functional effect of VGLL4 and its acetylation on TEAD1-YAP transcriptional activity using 8xGTIIC-Luci (Dupont et al., 2011), a luciferase reporter driven by a multimerized TEAD1 binding site. TEAD1 alone weakly stimulated reporter activity, and this was inhibited by both VGLL4 and VGLL4[R] (Figure 3H). Consistent with its broad role as a transcriptional co-activator, p300 strongly stimulated TEAD1 transcriptional activity. VGLL4 partially blocked this stimulation, as expected based on its antagonism of TEAD1-YAP interaction (Figure 3G). Compared with VGLL4, VGLL4[R] more potently blocked p300 stimulation (Figure 3H), in agreement with the more potent disruption of TEAD1-YAP interaction by VGLL4[R] (Figure 3G). To determine whether VGLL4 acetylation affects VGLL4 and TEAD1 interaction in CMs, we used the proximity ligation assay (PLA) (Söderberg et al., 2006) to study the in situ interaction between VGLL4 and TEAD1 in cultured neonatal rat ventricular cardiomyocytes (NRVMs), with or without the overexpression of p300. NRVMs stained with TEAD1 or VGLL4 antibodies individually showed that TEAD1 was localized in the nucleus, while VGLL4 was located in both cytoplasm and nucleus (Figure S3D). The PLA showed in situ TEAD1-VGLL4 interaction primarily in the nucleus (Figure 3I). p300 overexpression significantly reduced TEAD1-VGLL4 interaction (p < 0.001; Figures 3I and 3J). These data indicate that acetylation of VGLL4 decreases its interaction with TEAD1 in CMs. To determine whether VGLL4 is acetylated endogenously in CMs in vivo, we developed an antibody that recognized VGLL4 acetylated at K225 but not VGLL4 lacking this modification (Figures S3E and S3F). We used the antibody to assess acetylation of the corresponding residue of VGLL4 in murine heart (murine K216 corresponds to K225 of human VGLL4). Immunoblotting of mouse hearts showed that mVGLL4-K216Ac and total VGLL4 level increased 3- and 6-fold, respectively, between P6 and P60 (Figure 3K), indicating that the ratio of mVGLL4-K216Ac to total VGLL4 decreases with age. Consistent with the decrease in the fraction of acetylated VGLL4, p300 expression in the mouse heart also decreased with age (Figure S3G). To assess the potential relevance of developmental changes in VGLL4, VGLL4 K225Ac, TEAD1, YAP, and p300 to the human heart, we examined their expression in normal human myocardium at different postnatal ages (Figure 3L). TEAD1 protein levels declined with age, as we observed in mouse. Unlike mouse, YAP expression and VGLL4 expression were relatively constant in 2- to 18-year-old hearts. However, VGLL4-K225Ac strongly decreased with postnatal age, as did expression of p300, paralleling our observations in mouse. These results suggest that developmentally regulated protein expression and VGLL4-K225 acetylation contribute to regulation of YAP-TEAD activity in the human heart. Whereas VGLL4 expression may contribute to stage-specific regulation of VGLL4-TEAD interaction, VGLL4 acetylation appears to predominate in humans. We conclude that K225 acetylation of VGLL4 within its TDU domain inhibits its binding to TEAD1 in the heart. Mutation of VGLL4 K225 to arginine abrogates this inhibitory acetylation and promotes TEAD1-VGLL4 interaction. VGLL4 Suppresses YAP Activity Partially by Promoting TEAD1 Degradation While investigating VGLL4-TEAD1 interaction, we noticed that VGLL4 overexpression reduced TEAD1 protein level (e.g., Figure 3G). To expand on this observation, we studied the effect of VGLL4 on different levels of transfected TEAD1. Consistently, VGLL4 reduced the steady-state level of TEAD1 (Figure 4A), leading to the hypothesis that VGLL4 regulates TEAD1's degradation rate. To test this hypothesis, we expressed TEAD1 fused to Dendra2, a fluorescent protein that converts from green to red fluorescence upon transient illumination with 405-nm light (Zhang et al., 2007) (Figures 4B and 4C). This allows the degradation of photoconverted protein to be monitored in real time, independent of ongoing protein synthesis. To allow rapid expression of VGLL4, we used an inducible expression system in which addition of doxycycline to the culture medium rapidly induced VGLL4 expression (Figures S4A and S4B). In control cells that did not express VGLL4, doxycycline treatment did not significantly affect steady-state TEAD1 levels over a 10-hr period (Figure 4D). In contrast, in cells that expressed VGLL4 upon doxycycline addition, steady-state TEAD1 levels declined by approximately 50% over the same period (p < 0.05; Figures 4D and 4E). VGLL4 did not reduce the mRNA level of Tead1-Dendra2 (Figure S4C), indicating that the effect of VGLL4 was post-transcriptional. We then combined doxycycline-induced VGLL4 expression with live cell imaging of TEAD1-Dendra2 to specifically probe the effect of VGLL4 on TEAD1 stability. Four hours after doxycycline treatment, TEAD1-Dendra2 was pulse-labeled by photoconversion. In the absence of VGLL4, TEAD1-Dendra2 protein fluorescence intensity dropped by 10% during the 3-hr imaging process, reflecting the basal degradation rate of TEAD1. In contrast, in the presence of VGLL4 the fluorescence intensity of photoconverted TEAD1-Dendra2 declined 30% over the same period (Figure 4F), indicating that VGLL4 accelerates TEAD1-Dendra2 degradation. VGLL4 did not affect the stability of Dendra2 itself (Figure 4F), indicating that the loss of TEAD1-Dendra2 was specific to the TEAD1 component. These data demonstrate that VGLL4 strongly influences TEAD1 stability. Major pathways for protein degradation are the proteasome and cysteine, serine, and threonine peptidases. To determine whether these candidate pathways are involved in VGLL4-mediated TEAD1 degradation, we studied the effect of MG132 (a universal proteasome inhibitor [Lee et al., 1998]), leupeptin (an inhibitor of cysteine, serine, and threonine peptidases [Umezawa, 1976]), and E64 (an inhibitor specific to cysteine proteases [Barrett et al., 1982]) on VGLL4-mediated TEAD1 degradation. In 293T cells co-transfected with TEAD1 and VGLL4, leupeptin, and E64, but not MG132, reduced VGLL4-mediated reduction of TEAD1 steady-state levels (Figure S4D), suggesting that VGLL4 triggers TEAD1 degradation through cysteine proteases and not the proteosome. We used pulse-labeled TEAD1-Dendra2 and live cell imaging to confirm that E64 reduced the destabilizing effect of VGLL4 on TEAD1-Dendra2 (Figure 4F). Together, our data indicate that VGLL4 is sufficient to stimulate TEAD1 degradation through a cysteine protease-dependent pathway. Previously, VGLL4 was shown to antagonize YAP-TEAD1 transcriptional activity by competitively binding to TEAD, displacing YAP (Koontz et al., 2013). Our data suggest an additional mechanism, in which VGLL4 binding to TEAD1 promotes its degradation and thereby reduces the amount available to interact with YAP. To test this hypothesis, we used the TEAD1 luciferase reporter 8xGTIIC-Luci (Dupont et al., 2011) to monitor TEAD1-YAP transcriptional activity. 293T cells were co-transfected with 8xGTIIC-Luci, doxycycline-inducible VGLL4, and YAP expression constructs. One day after transfection, luciferase reporter activity was measured 0–8 hr after doxycycline treatment. In the control group lacking doxycycline-inducible VGLL4, reporter activity was relatively stable after the addition of doxycycline (Figure 4G). In contrast, in cells with doxycycline-induced VGLL4, luciferase activity declined significantly over this time period (p < 0.05, Figure 4G). To probe the role of protein degradation in the effect of VGLL4, we treated cells with E64. This significantly increased reporter activity 8 hr after doxycycline treatment (p < 0.05; Figure 4G), suggesting that VGLL4 stimulation of TEAD1 degradation contributes to the decrease in transcriptional activity observed after VGLL4 induction. However, E64 did not completely abrogate the reduction of TEAD1 transcriptional activity caused by VGLL4 induction. In part this reflects incomplete E64 protection of TEAD1 from VGLL4-mediated degradation (Figure S4D), as well as the additional inhibitory effect of VGLL4 on YAP recruitment to TEAD1 (Figure 4H). We conclude that VGLL4 is sufficient to stimulate TEAD1 degradation through cysteine-dependent proteases, and that VGLL4 antagonizes TEAD1-YAP transcriptional activity by stimulating TEAD1 degradation in addition to disrupting TEAD1-YAP interaction. Precocious Overexpression of VGLL4[R] in Neonatal Heart Leads to Heart Failure VGLL4 overexpressed in the neonatal heart did not interact with TEAD1 and did not significantly affect neonatal heart growth or function (Figure 2). However, VGLL4-K225 acetylation in the neonatal heart reduced VGLL4 effect on TEAD1-YAP and thus may have masked VGLL4's biological activity. Because the VGLL4[R] mutant is refractory to inhibition by K225 acetylation, we hypothesized that this single amino acid substitution would reveal the cardiac activity of overexpressed VGLL4. To test this hypothesis, we introduced the mutant protein into the neonatal heart by developing and administering AAV9.VGLL4[R]. AAV9.GFP and AAV9.VGLL4 were used as negative controls. TEAD1-interacting proteins were detected by coIP. Consistent with our prior results, we did not detect significant interaction between TEAD1 and VGLL4 (Figure 5A, lane 6 versus lane 5). Accordingly, AAV9.VGLL4 did not affect TEAD1 level or TEAD1-YAP interaction. In contrast, VGLL4[R] did interact with TEAD1 (Figure 5A, lane 7 versus lanes 5 and 6). Consistent with this interaction, TEAD1 level was reduced by VGLL4[R] (Figure 5A, lane 7 versus lanes 5 and 6), and the balance between TEAD1's interaction partners was altered, with greater binding to VGLL4[R] and reduced binding to YAP. Together, these results support our model that VGLL4 acetylation at K225 governs its interaction with TEAD1 in the neonatal heart. By blocking K225 acetylation, the K225R mutation reveals the potent effect of VGLL4 on neonatal heart function. Next, we addressed the role of p300 in VGLL4 K225R acetylation in the neonatal heart. p300 level was not affected by overexpression of VGLL4 or VGLL4[R] (Figure 5B, lanes 2 and 3 versus lane 1). VGLL4 and VGLL4[R] both co-immunoprecipitated with p300 (Figure 5B, lanes 6 and 7 versus lane 5) in neonatal heart, while this interaction was not detected in adult heart (Figure S5A). Co-precipitated VGLL4 was acetylated, whereas VGLL4[R] was not detectably acetylated (Figure 5B, lane 6 versus lane 7). This result validated that the K225R mutation reduced VGLL4 acetylation, and suggested that p300 mediates VGLL4 acetylation in vivo. Given that VGLL4 K225R mutation increases TEAD1-VGLL4 interaction in neonatal heart at the expense of TEAD1-YAP, we next examined the biological effect of this single amino acid substitution. We delivered AAV9.VGLL4[R] to P1 mice. Litter-mates treated with AAV9.VGLL4 (wild-type) and AAV9.GFP-treated were used as negative controls. At P8, all three groups had similar heart and body weights (Figure S5B), and, as we observed previously, heart function was no different between AAV9.VGLL4 and AAV-GFP groups (Figure 5C). However, AAV9.VGLL4[R] induced severe myocardial dysfunction and myocardial wall thinning (Figures 5C and 5D). The AAV9.VGLL4[R]-treated mice failed to grow normally, and 30% died prior to a planned necropsy date at P12 (Figures S5B and S5C). Those that survived to P12 had striking ventricular and atrial enlargement that was not observed in either negative control group (Figure 5E). These mice had lower body weight and higher heart weight than the other two groups (Figures 5F, S5B, and S5C). Staining of heart sections with picrosirius red showed that AAV9.VGLL4[R] hearts had extensive fibrosis (Figures 5G and 5H). Interestingly, AAV9.VGLL4 hearts also had mildly but significantly increased fibrosis compared with AAV9.GFP. AAV9. Gene expression measurements by qRT-PCR showed that VGLL4[R] induced cardiac upregulation of Nppa and downregulation of Myh6, changes frequently observed in heart failure (Figures 5I and 5J). Together, these results show that K225R acetylation regulates VGLL4 activity in the neonatal heart. Blocking K225R uncovered potent VGLL4 activity to disrupt TEAD1-YAP interaction and neonatal heart development and function. Activation of VGLL4 in the Neonatal Heart Suppresses Cardiomyocyte Proliferation by Disrupting the YAP-TEAD1 Complex The YAP-TEAD complex promotes CM proliferation (von Gise et al., 2012; Xin et al., 2011; Heallen et al., 2011; Lin et al., 2014), and loss of this activity caused heart failure at least in part due to reduced CM number (Del Re et al., 2013). In addition, YAP-TEAD has been implicated in regulating CM survival (Del Re et al., 2013). To understand the cellular mechanisms underlying VGLL4[R]-induced heart failure, we investigated processes that might alter CM survival (apoptosis or necrosis) or production (proliferation). VGLL4[R] did not significantly induce CM apoptosis, as measured by TUNEL staining (Figure S6A). CMs undergoing necrosis have plasma membranes that are abnormally permeable to macromolecules such as antibodies. Therefore, uptake of injected anti-myosin antibody by CMs has been used as an assay of CM necrosis (Nakayama et al., 2007). To determine whether AAV9.VGLL4[R] caused CM necrosis, we injected mouse anti-myosin antibody (MF20) into mouse pups at P7, and hearts were analyzed at P8. In both negative control groups, MF20+ CMs were rarely observed, whereas MF20+ CMs were readily observed in the AAV9.VGLL4[R] group (Figures 6A and 6B). To determine whether VGLL4[R] reduced CM proliferation, we performed quantitative pH3 staining. VGLL4[R] strongly decreased the fraction of pH3+ CMs compared with VGLL4 or GFP (Figure 6C). Because CM multinucleation or polyploidization can dissociate CM M-phase activity from CM number, we undertook a clonal analysis to more directly probe the effect of VGLL4 on CM proliferation. As we described previously (Lin et al., 2014), by pulse-labeling a low fraction of CMs and later counting the number of CMs in individual labeled clusters, one can assess the extent of productive CM cell-cycle activity. Use of the multi-color Confetti reporter mouse (Snippert et al., 2010), in which Cre stochastically activates expression of one of four fluorescent proteins, further enhances this strategy by allowing one to distinguish chance labeling of two neighboring cells (potentially yielding multichromatic clusters) from expansion of a single CM (monochromatic clusters). AAV9.cTNT::Cre (AAV9.Cre) was injected into the P1 Brainbow pups at a low dose to irreversibly label a small fraction of CMs with one of the four different fluorescent proteins: CFP, RFP, nuclear GFP (nGFP), and YFP. For technical reasons, we only considered RFP and YFP readouts. In pilot experiments, we determined the optimal dose of AAV9.Cre to achieve the desired CM labeling rate (Figures S6B–S6D). We delivered this dose of AAV9.Cre to P1 Confetti mouse pups. At the same time, we delivered AAV9.VGLL4fb or AAV9.VGLL4 [R]fb, in which the GFP tag has been replaced by the FLAG-Bio tag, at doses which should transduce greater than 90% of cardiomyocytes. In P8 hearts, we determined the frequency of bichromatic and monochromatic cell clusters, where a cluster was defined as two or more labeled, adjacent cells (Figure 6E). As expected, bichromatic clusters, representing independent labeling of neighboring cells, occurred at similar frequency in AAV9.Cre, AAV9.Cre+AAV9.VGLL4fb, and AAV9.Cre+AAV9.VGLL4[R]fb groups (Figure 6F). In contrast, monochromatic clusters, representing productive CM proliferation, were less common in the VGLL4 [R] group (Figure 6F), supporting our hypothesis that VGLL4[R] overexpression reduces neonatal CM proliferation. Consistent with these measurements of CM proliferation, expression of cell-cycle genes Aurka and Cdc20, as well as the canonical TEAD-YAP target gene Ctgf, was reduced in P12 hearts treated with AAV9.VGLL4[R], compared with AAV9.GFP (Figure 6G). Both reduction of CM cell-cycle activity and CM survival would reduce CM number, leading to increased workload for remaining CMs. Stressed CMs undergo compensatory hypertrophy, and indeed we observed that AAV9.VGLL4[R]-treated CMs are larger than CMs in the GFP-negative control group (Figures 6H and 6I). In combination with increased cardiac fibrosis, these changes may account for the increased weight of AAV9.VGLL4[R]-treated hearts (Figures 5F and S5C). We conclude that blocking VGLL4 K225 acetylation reveals the potent effect of VGLL4 on CM proliferation and survival in the neonatal heart, at least in part through disruption of TEAD1-YAP interaction and destabilization of TEAD1. DISCUSSION Our work identifies VGLL4 as an important negative regulator of cardiac growth driven by YAP-TEAD1, defines VGLL4 acetylation as a novel mechanism to regulate Hippo-YAP signaling and organ growth, and reveals VGLL4 regulation of TEAD1 stability as an additional mechanism whereby VGLL4 modulates YAP-TEAD activity. Although our studies focused on the effect of VGLL4 on YAP-TEAD activity, it is possible that VGLL4 has other important biological functions in CMs. For instance, the Hippo-YAP pathway is closely intertwined with Wnt/β-catenin signaling (Heallen et al., 2011; Varelas et al., 2010), and in future studies it will be invaluable to assess the impact of VGLL4 on this pathway. Developmental Regulation of TEAD-Interacting Proteins The Hippo-YAP pathway controls the growth of mitotic organs, such as liver (Dong et al., 2007), intestine (Camargo et al., 2007), skin (Schlegelmilch et al., 2011), and fetal heart (Heallen et al., 2011; von Gise et al., 2012; Xin et al., 2011). In the adult heart, YAP activation induced limited CM proliferation and was insufficient to cause cardiomegaly (Lin et al., 2014). We investigated mechanisms that restrain YAP activity in the adult heart, and our data identify VGLL4 as a potential negative regulatory factor. Among the organs that we studied, VGLL4 was most highly expressed in the heart. As proliferative neonatal CMs transitioned to become postmitotic adult CMs, VGLL4 protein levels increased markedly, whereas YAP and TEAD1 protein levels declined. Interestingly, these protein level changes were post-transcriptionally mediated through uncharacterized mechanisms. Consistent with these changes in protein expression, the main interaction partner of TEAD1 switched from YAP in the neonatal heart to VGLL4 in the adult heart. This developmentally regulated switch in interaction partners is crucial for normal heart maturation, since precocious formation of TEAD1-VGLL4 complex in the neonatal heart caused cardiac hypoplasia, CM necrosis, and lethal heart failure. We are now developing essential reagents to test whether blocking VGLL4 activity in the adult heart will boost YAP-TEAD1 mitogenic activity. Regulation of YAP-TEAD1 Activity through VGLL4 Acetylation Many studies have examined the regulatory pathways that converge on YAP to regulate its cellular localization and transcriptional activity, often through YAP post-translational modifications that include phosphorylation, acetylation, and methylation (Pan, 2010; Hata et al., 2012; Oudhoff et al., 2013). As a transcriptional co-activator, YAP transcriptional activity depends upon its binding to a partner DNA-binding transcription factor. On a genome-wide scale, TEAD is the major transcription factor partner of YAP (Zanconato et al., 2015; Galli et al., 2015), and we previously showed that TEAD1-YAP interaction is essential for fetal heart growth (von Gise et al., 2012). TEAD1 likely also has additional roles in the regulation of muscle gene expression (Yoshida, 2008). Nevertheless, little attention has been directed at mechanisms that control TEAD1 activity and stability. We found that VGLL4-TEAD1 interaction is developmentally regulated. Although this is partially explained by developmentally regulated changes in protein expression, the lack of interaction between overexpressed VGLL4 and TEAD1 in the neonatal heart pointed to additional regulatory mechanisms. We discovered a second mechanism, VGLL4 acetylation at K225, which normally impedes VGLL4-TEAD interaction in the neonatal heart (Figure 6J). Abrogation of acetylation at this residue in the VGLL4 [R] mutant precociously impaired VGLL4-TEAD1 interaction, thereby reducing YAP-TEAD1 mitogenic and pro-survival activity (Figure 6J). Thus, our data show that this post-translational modification is a critical regulatory switch that is essential for neonatal CM proliferation and survival. One of the major acetyltransferases that acetylates VGLL4 is p300. We readily detected interaction between endogenous p300 and VGLL4 in the neonatal but not adult heart, indicating that this interaction is also developmentally regulated. Decline of p300 levels, alteration of its primary interaction partners, or destabilization of the p300-VGLL4 complex (e.g., due to changes in the composition of each protein's interacting partner complexes) may all contribute to this developmentally reduced interaction. Decreased p300-VGLL4 interaction between neonatal and adult CMs correlated with a decrease in the fraction of VGLL4 that is acetylated. However, VGLL4 acetylation was still present in the adult heart, and indeed the absolute level of acetylated VGLL4 was higher. This suggests that a different acetyltransferase may acetylate VGLL4 in adult CMs. Alternatively p300 may still acetylate VGLL4, albeit with reduced efficiency that reflects the decline of p300-VGLL4 interaction detectable by coIP. Understanding the mechanisms that regulate VGLL4 acetylation may lead to therapeutic strategies to enhance VGLL4 acetylation and thereby mitigate its inhibitory effects on YAP-TEAD activity in mature CMs. Interestingly, VGLL4 appeared to be regulated differently in human and mouse hearts. Whereas mouse VGLL4 increased markedly with postnatal age, its expression in human heart was relatively constant. On the other hand, VGLL4-K225Ac declined with age in human heart. In mouse it increases, but total VGLL4 increases even more, so that the proportion of VGLL4-K225Ac decreases. This suggests that developmental regulation of VGLL4 activity may be achieved in different species by varying combinations of regulated expression or acetylation. Precise YAP-TEAD regulation is often achieved through signaling pathways that either add or remove post-translational modifications. For instance, the Hippo kinase cascade phosphorylates YAP to trigger its cytoplasmic sequestration (Huang et al., 2005). This is counterbalanced by YAP dephosphorylation by the protein phosphatase PP2A, a process regulated by YAP interaction with α-catenin (Schlegelmilch et al., 2011). Lysine acetylation is also a reversible post-translational modification that can be removed by deacetylases. Thus in future work it will be interesting to consider whether deacetylases counterbalance VGLL4 acetylation. VGLL4 Regulates TEAD1 Stability Previous work suggested that VGLL4 suppresses YAP activity by competing for TEAD binding (Zhang et al., 2014; Koontz et al., 2013). Our data support this mechanism, and uncovered an additional previously unrecognized mechanism by which VGLL4 regulates YAP-TEAD1 activity. We found that VGLL4 interaction with TEAD1 accelerated TEAD1 protein degradation through cysteine peptidases (Figure 6J). Cysteine peptidase inhibition by E64 only partially rescued YAP-TEAD1 transcriptional activity, suggesting that VGLL4 regulates YAP-TEAD1 activity by both causing TEAD1 degradation and inhibiting YAP-TEAD1 interaction. Future work is required to define the mechanism by which VGLL4 enhances TEAD1 degradation, and the specific cysteine peptidases that degrade TEAD1. VGLL4 Affects Postnatal Cardiac Growth and Maturation by Suppressing CM Proliferation and Inducing Necrosis We found that VGLL4 acetylation at K225 blocks its activity in the neonatal heart. Overriding VGLL4 acetylation by mutating this residue to arginine unmasked the potent effect of VGLL4 gain of function in the neonatal heart, resulting in destabilization of the TEAD1-YAP complex and reduced CM proliferation and YAP target gene expression. Moreover, VGLL4[R] induced CM necrosis but not apoptosis. As a result of reduced CM proliferation and increased necrosis, AAV9.VGLL4[R]-transduced pups developed heart failure. These data identify an additional, previously unreported role of YAP-TEAD1 in CMs to suppress necrosis. This function of YAP may cross cell types, as YAP deficiency predisposed hepatocytes to undergo necrosis after bile duct ligation (Bai et al., 2012). CM necrosis is an important mechanism of CM loss following experimental myocardial infarction and genetically induced CM calcium overload (Kajstura et al., 1998; Nakayama et al., 2007). It will be interesting to dissect the mechanisms by which the balance between VGLL4-TEAD and YAP-TEAD governs CM necrosis. Conclusions This study identified novel mechanisms that regulate YAP-TEAD activity. Normal deployment of these mechanisms is crucial for neonatal cardiac growth and maturation. Our study suggests that manipulation of the balance between VGLL4, TEAD, and YAP activity, and/or suppressing VGLL4-mediated TEAD1 degradation, may be of use for therapeutic cardiac regeneration or repair. Alternatively, augmenting the inhibitory action of VGLL4 or reducing its acetylation may be useful strategies for the control of oncogenic growth driven by excessive YAP-TEAD activity. EXPERIMENTAL PROCEDURES For detailed methods, please refer to Supplemental Experimental Procedures. Animal Experiments Animal procedures were approved by the Boston Children's Hospital Animal Care and Use Committee. Tead1fb knockin mice (MMRRC:037514-JAX) were generated by targeting the C terminus of Tead1 in murine embryonic stem cells to introduce FLAG and Bio epitope tags. Human Myocardium Human left ventricular myocardium from unused donor hearts was snap-frozen within 2 hr of organ harvest, under protocols approved by the Institutional Review Boards of the University of Sydney and St. Vincent's Hospital. AAV9 Vector Generation and Delivery Vgll4 or Vgll4[R] were cloned into ITR-containing AAV plasmid (Penn Vector Core P1967) harboring the chicken cardiac TNT promoter (Prasad et al., 2011). AAV9 was packaged as described by Lin et al. (2014) and delivered by subcutaneous injection. Proteins and Peptides Murine TEAD1 residues 211–427, containing the YAP and VGLL4 binding domain, was fused to a polyhistidine tag, expressed in bacteria, and purified on Ni-nitrilotriacetic acid agarose (Qiagen). Peptides VGLL4-TDU-V5 and VGLL4-TDU-KAc-V5 (Figure S3) were chemically synthesized. VGLL4-TEAD1 binding affinity was determined using photonic crystal nanobeam sensors (Yang et al., 2014). Cell or tissue-soluble protein extracts for coIP were immunoprecipitated using protein A or streptavidin Dynabeads (Life Technologies). Acetylation was analyzed using pan-acetylated lysine antibody (Cell Signaling Technology, 9441S) or mass spectrometry (Harvard Medical School Taplin Biological Mass Spectrometry Facility). Murine VGLL4-K216Ac specific antibody was raised in rabbits and isolated by affinity purification. Co-immunoprecipitation and In Vitro VGLL4 Acetylation Analysis Cell or tissue-soluble protein extracts for coIP were prepared in lysis buffer (20 mM Tris-HCl [pH 8], 137 mM NaCl, 10% glycerol, 1% Triton X-100, and 2 mM EDTA). Protease inhibitor cocktail (Roche) was added to the lysis buffer immediately before use. The protein solution was diluted with 1 volume of immunoprecipitation buffer (lysis buffer without glycerol) and pre-cleared with protein A Dynabeads (Life Technologies, 10008D). Antibody or immunoglobulin G was added to the pre-cleared extract, and antibody bound protein complexes were pulled down with pre-equilibrated protein A Dynabeads. After three washes, the immunoprecipitated proteins were eluted with 1× SDS loading buffer. For analysis of VGLL4 acetylation, 293T cells were co-transfected with VGLL4-GFP and other indicated plasmids and cultured for 48 hr. Two hours before harvest, cells were treated with 5 μM trichostatin A (TSA; Cayman Chemical, CAS 58880-19-6). GFP antibody was used to pull down VGLL4-GFP in the presence of 5 μM TSA. Acetylation was analyzed using pan-acetylated lysine antibody (Cell Signaling Technology, 9441S) or mass spectrometry (Harvard Medical School Taplin Biological Mass Spectrometry Facility). Cardiomyocyte Isolation and Culture Neonatal rat ventricle cardiomyocytes (NRVMs) were isolated from 2-day-old Wistar rats (Charles River) using the Neomyts cardiomyocyte dissociation kit (Cellutron). Isolated cardiomyocytes were initially cultured for 24 hr in the presence of 10% fetal bovine serum. Cardiomyocytes were then cultured on fibronectin-coated glasses in 1% horse serum medium and cultured for 24 hr before 4% paraformaldehyde (PFA) fixation. Duolink In Situ PLA PFA (4%)-fixed NRVMs were permeabilized with PBS containing 0.1% Triton X-100. Duolink PLA was carried out using the Sigma Duolink In Situ Red Starter Kit Mouse/Rabbit (Duo92101-1KT). To visualize cardiomyocytes, we co-stained cells with goat-originated cardiac troponin I antibody (Abcam). In brief, NRVMs were incubated at 4°C overnight with primary antibodies against VGLL4, TEAD1, and cardiac troponin I. On the second day, PLA was carried out following the kit protocol. Alexa 488-conjugated donkey anti-goat secondary antibody was added at the signal amplification step. Statistics Values are expressed as mean ± SEM. Student's t test or ANOVA with Tukey's honest significant difference post hoc test was used to test for statistical significance involving two or more than two groups, respectively. Source Data Original immunoblot data have been deposited at Mendeley Data with the following digital object identifiers, corresponding to Figures 1, 2, 3, 4, and 5 and Figures S1–S6, respectively: 10.17632/36rcz2ndks.1, 10.17632/8mg7zn6k9n.1, 10.17632/s9x6x6shp3.1, 10.17632/xn4xhbfms9.1, 10.17632/yxmwb5r5j7.1, and 10.17632/jfdxk7nnhs.1. Supplementary Material 1 2 ACKNOWLEDGMENTS We thank Ross Tomaino from Taplin Biological Mass Spectrometry Facility for mass spectroscopy analysis of VGLL4 post-translational modifications. W.T.P. was supported by NIH HL116461, U01 HL100401, and UM1 HL098166, by an AHA Established Investigator Award (12EIA8440003), and by charitable contributions from Gail Federici Smith and Dr. and Mrs. Edwin A. Boger. Z.L. was supported by an American Heart Association Scientist Development Grant. Figure 1 Developmental Changes in VGLL4-TEAD1 and YAP-TEAD1 Interaction in the Mouse Heart (A) Immunoblot (IB) of protein extracts from adult mouse brain (B), heart (H), kidney (K), liver (Li), and lung (Lu). (B) Immunoblot (IB) of heart protein extracts from mice with the indicated postnatal (P) age in days. GAPDH internal control belonging to respective immunoblots is shown. (C) qRT-PCR measurement of Vgll4, Tead1, and Yap mRNA level in 3-day-old (P3) and 90-day-old (P90) hearts. p < 0.05, n = 3. (D) VGLL4, TEAD1, and YAP expression in CMs and non-CMs. Adult hearts were dissociated by collagenase perfusion and then separated into CM and non-CM fractions. Protein extracts were immunoblotted with the indicated antibodies. (E) Age-dependent association of VGLL4 and TEAD1 in mouse heart. Tead1fb/+;R26BirA/+ heart extract was incubated with immobilized streptavidin (SA). Co-precipitated VGLL4 and TEAD1 were measured by immunoblotting. Tead1+/+;R26BirA/+ heart extract was used as a negative control. (F) Age-dependent association of YAP and TEAD1 in mouse heart. TEAD1 was precipitated from protein from P1, P8, or P50 mouse heart as in (D). Co-precipitated proteins were detected by immunoblotting. (G) Relative YAP or VGLL4 coIP with TEAD1, determined by quantification of (E). Precipitated proteins were normalized to TEAD1fb. p < 0.05, n = 3. All error bars represent the SEM. Figure 2 VGLL4 Overexpression Did Not Suppress Neonatal Cardiac Growth P1 pups were injected subcutaneously with AAV9.GFP or AAV9.VGLL4-GFP. Control (Ctrl) mice were untreated. Hearts were analyzed at P8. (A) AAV9 expression constructs. AAV9.GFP and AAV9.VGLL4-GFP incorporate the cardiac troponin T (cTNT) promoter to drive selective CM expression. Heart protein immunoblots probed with GFP antibody demonstrated VGLL4-GFP fusion protein expression (asterisk). (B) Echocardiographic assessment of neonatal heart function. FS%, fractional shortening. n = 4. (C) Whole-mount images of hearts showing lack of substantial differences between groups. Scale bar, 2 mm. (D) Heart to body weight ratio was not significantly different (NS) between groups. n = 3. (E) Representative pH3 staining results. Scale bar, 50 μm. (F) Quantitation of pH3+ CMs. n = 3. NS, not significant. (G) Cell-cycle gene expression from P8 ventricular myocardium after treatment with AAV9.GFP or AAV9.VGLL4-GFP. Gene expression was measured by qRT-PCR and normalized to the AAV9.GFP group. n = 4. NS, not significant. All error bars represent the SEM. Figure 3 VGLL4 TDU Domain Acetylation Decreased VGLL4-TEAD1 Interaction (A) p300 bound and acetylated VGLL4. HEK293T cells were transfected with the indicated GFP and histone acetyltransferase (HAT; HA-tagged) expression plasmids. Proteins that co-precipitated with GFP were detected by immunoblotting (IB). K-Ac Ab, acetylated lysine-specific antibody. (B) VGLL4-K225 is the major VGLL4 acetylation site. VGLL4-GFP was overexpressed in HEK293T cells in the presence of p300, immunoprecipitated with GFP antibody, and analyzed by mass spectrometry. The area of the red circles is proportional to the fraction of peptides detected that contain the acetyl lysine residue indicated by the corresponding number. T1 and T2 represent the two TDU domains of VGLL4. (C) VGLL4 K225R mutation decreased VGLL4 acetylation. Wild-type (WT) or K255R mutated (R) VGLL4-GFP were co-expressed in HEK293T with p300, as indicated. VGLL4-GFP acetylation was detected by immunoprecipitation and western blot. Acetylation of VGLL4 K225R was normalized to wild-type VGLL4. p < 0.05, n = 4. (D) Alignment of TDU domains from different proteins (top group) or from the first TDU domain of VGLL4 from different species (bottom group). The residue aligned with K225 of human VGLL4 is highlighted in red and indicated by an arrow. This residue is conserved in vertebrate VGLL4 but is not conserved across TDU domains. (E and F) VGLL4 K225 acetylation decreased VGLL4-TEAD1 interaction in vitro. Interaction between recombinant His-TEAD1[211–427] and synthetic, un-acetylated, or K225-acetylated VGLL4 TDU domain peptides was detected by immunoprecipitation and western blot (E) or by a nanoscale photonic interaction assay (F). (G) VGLL4[R] increased VGLL4-TEAD1 and decreased YAP-TEAD1 interaction in cultured cells. TEAD1fb and VGLL4-GFP expression plasmids were co-transfected into 293T cells. TEAD1 CoIP was carried out using FLAG antibody. The ratio between VGLL4 and TEAD1 in the immunoprecipitate was quantified by densitometry. p < 0.05, n = 3. (H) p300 effect on YAP-TEAD1 transcriptional activity. 293T cells were co-transfected with the indicated plasmids plus pRL-TK. Left: 24 hr after transfection, cells were collected for luciferase activity measurement. Firefly luciferase activity was normalized to Renilla luciferase. p < 0.05, n = 4. Right: western blot showing the expression of p300, Vgll4, and TEAD1fb in transduced cells. (I and J) Effect of VGLL4 acetylation on VGLL4-TEAD1 interaction in NRVM. The PLA was used to detect endogenous VGLL4-TEAD1 interaction in cultured NRVMs. Representative image (I) and quantification of TEAD1-VGLL4 interaction events in the nucleus (J) are shown. Each red dot was counted as an interaction event. Scale bar, 20 μm. *p < 0.0001, Wilcoxon. (K) Endogenous levels of mVGLL4 (murine VGLL4) and mVGLL4-K216Ac (which corresponds to human K225Ac) in P6 and P60 heart. Hearts were lysed with denaturing buffer containing 2% SDS, and 100 μg of total protein was immunoblotted for total VGLL4 or VGLL4-K216Ac. Fold change of protein levels between P60 and P6 was determined by densitometry. p < 0.05. (L) Endogenous levels of VGLL4 and VGLL4-K225Ac in human left ventricular myocardium, obtained from unused transplant donor hearts of the indicated ages. All error bars represent the SEM. Figure 4 VGLL4 Overexpression Decreased TEAD1 Stability (A) VGLL4 overexpression decreased TEAD1 protein level. Different doses of TEAD1 plasmids (indicated in μg) were co-transfected with 1.6 μg of VGLL4-GFP plasmid. Cells were collected for western blot 24 hr after transfection. (B and C) Generation and validation of TEAD1-Dendra2 construct. Tead1-Dendra2 plasmid was transfected into 293T cells. Western blot confirmed expression of TEAD1-Dendra2 fusion protein (B). TEAD1-Dendra2 merge fusion protein was green before illumination with 405 nm light. After 30 s of illumination, a fraction of TEAD1-Dendra2 exhibited red fluorescence (C). (D and E) Doxycycline (Dox)-inducible expression of VGLL4 caused TEAD1-Dendra2 degradation. pTEAD1-Dendra2 and pEF1a-rtTA were co-transfected into 293T cells along with pTetO empty vector (upper panel) or pTetO:HA-VGLL4 (lower panel). Twenty-four hours after transfection, Dox was added. Cells were analyzed at the indicated time points (D). Quantification of TEAD1-Dendra2 protein level is shown in (E). p < 0.05, n = 3. (F) Time-lapse imaging of TEAD1-Dendra2 or Dendra2 proteins. Indicated plasmids were transfected into 293T cells. Twenty-four hours later, cells were treated with Dox. Four hours later, Dendra2 was photoconverted with 405-nm light. Relative red fluorescence intensity (RFI) was monitored for 3 hr by taking one image per minute. RFI was normalized to the value immediately after photoconversion. Plot shows average RFI signal over 10 min. n = 10. Experiment is representative of three independent repeats. (G) Dual luciferase assay of YAP-TEAD1 transcriptional activity. 293T cells were co-transfected with YAP[S127A], EF1a:rtTA, 8xGIITC-luciferase, pRL-TK internal control, and either TetO empty vector or TetO-VGLL4 as indicated. E64 was added as indicated. Twenty-four hours after transfection, cells were treated with Dox for the indicated number of hours, when cell extracts were analyzed for Firefly and Renilla luciferase activity. Relative luciferase activity was the ratio of Firefly to Renilla luciferase, normalized to empty vector at time 0. p < 0.05, n = 4. (H) Model of VGLL4 regulation of YAP-TEAD1 activity. In the absence of VGLL4, YAP binds to TEAD1 to activate target gene expression. VGLL4 overexpression suppressed YAP-TEAD1 activity by both inhibiting TEAD1 transcriptional activity (i) and promoting TEAD1 degradation (ii). All error bars represent the SEM. Figure 5 Abrogation of VGLL4-K225 Acetylation Unmasked Disruptive Effects of VGLL4 on YAP-TEAD Interaction and Neonatal Heart Maturation P1 pups were treated with AAV9.VGLL4, AAV9.VGLL4[R] (containing the K225R mutation), or AAV.GFP. Hearts were examined at P8 or P12, as indicated. (A) Assay of cardiac TEAD1-interacting proteins. TEAD1 and its associated proteins were immunoprecipitated, and indicated proteins were detected by western blotting. Asterisk indicates the VGLL4-GFP band. (B) Endogenous p300 interacts with and acetylates VGLL4 in the neonatal heart. AAV9.GFP, AAV9.VGLL4-GFP, or AAV9.VGLL4[R]-GFP were administered to P1 mouse pups. p300 was immunoprecipitated from P8 heart extracts and probed with indicated antibodies. K-Ac Ab, acetyl lysine-specific antibody. Asterisk indicates acetylated VGLL4-GFP. Arrowhead indicates the VGLL4-GFP band, which runs just above the immunoglobulin heavy chain. (C and D) Echocardiographic measurement of left ventricular (LV) systolic function (fractional shortening, FS) (C) and diastolic LV wall thickness (D) at P8. p < 0.05 compared with GFP control, n = 4. (E) Whole-mount (upper panels) and H&E-stained short-axis sections of AAV-transduced hearts at P12. Scale bars, 2 mm. (F) Heart to body weight ratio of AAV-transduced hearts at P8 or P12. p < 0.05, n = 4. (G and H) Cardiac fibrosis was visualized by pirosirius red/fast green staining. Representative images (G) and quantification (H). Scale bar, 1 mm. p < 0.05, n = 3. (I and J) qRT-PCR measurement of heart failure marker gene transcripts Myh6 and Nppa. Levels were normalized to GAPDH and expressed relative to the AAV9.GFP control group. p < 0.05, n = 4. All error bars represent the SEM. Figure 6 Acetylation-Deficient VGLL4[R] Decreased Cardiomyocyte Proliferation and Survival AAV9.GFP, AAV9.VGLL4, or AAV9.VGLL4[R] were delivered to P1 pups, and hearts were examined at P8. (A and B) Measurement of CM necrosis. Rosa26mTmG (membrane localized RFP) P1 pups were treated with AAV9. Anti-myosin antibody MF20 was injected into mice at P7. At P8, mice were collected and intracellular MF20 antibody was detected by immunofluorescent staining. Representative images (A) and quantification (B). Scale bar, 50 μm. (C and D) Measurement of CM proliferation using pH3 immunofluorescence staining. (C) Representative images with boxed regions magnified in insets. Scale bar, 50 μm. (D) Quantification of pH3+ CMs. p < 0.05, n = 3. (E and F) Clonal assay for CM proliferation. Confetti P1 pups were treated with mosaic dose of AAV9.Cre to label individual CMs, in addition to treatment with VGLL4 or control AAVs at the usual dose, which transduces nearly all CMs. Hearts were examined at P8. (E) Representative images with boxed regions magnified in insets. Scale bar, 50 μm. (F) quantification of clusters of adjacent, labeled CMs containing one color (monochromatic, potentially arising from proliferation) or two colors (bichromatic, arising from adjacent labeling events). Scale bar, 50 μm. p < 0.05, n = 4. (G) qRT-PCR measurement of relative levels of YAP-TEAD canonical targets gene transcripts Aurka, Cdc20, and Ctgf in P12 heart. p < 0.05, n = 3. (H and I) Measurement of cardiomyocyte cross-sectional area. CMs were outlined by wheat germ agglutinin staining. Representative images (H) and quantification of cross-sectional area (I). p < 0.05, n = 3. (J) Model of VGLL4 regulation of heart growth. In normal newborn heart, predominant YAP-TEAD1 stimulates CM proliferation. The interaction between VGLL4 and TEAD1 is blunted by p300-mediated VGLL4 acetylation. Inhibition of VGLL4 acetylation, as in the K225R mutant, suppresses cardiac growth by both inhibiting YAP-TEAD1 interaction and decreasing TEAD1 stability. All error bars represent the SEM. Highlights TEAD-YAP stimulates organ growth, but its regulation is incompletely understood p300-mediated acetylation of VGLL4 K225 regulates VGLL4 antagonism of TEAD-YAP VGLL4-TEAD1 interaction promotes TEAD1 degradation VGLL4-K225Ac regulates TEAD1 activity to allow normal heart development AUTHOR CONTRIBUTIONS Z.L. designed and performed the study and wrote the manuscript. W.T.P. analyzed the data and co-wrote the manuscript. H.G., S.Z., Y.C., N.V.D., Y.G., Q.M., P.Z., J.Z., A.H., P.R.v.G., and S.S. generated reagents and acquired data. F.L. and Q.Q. performed nanoscale photonic interaction assays. A.L. and C.d.R. provided human heart samples. V.J.B. analyzed the Dendra2 time-lapse imaging data. SUPPLEMENTAL INFORMATION Supplemental Information includes Supplemental Experimental Procedures and six figures and can be found with this article online at http://dx.doi.org/10.1016/j.devcel.2016.09.005. REFERENCES Bai H Zhang N Xu Y Chen Q Khan M Potter JJ Nayar SK Cornish T Alpini G Yes-associated protein regulates the hepatic response after bile duct ligation. Hepatology 2012 56 1097 1107 22886419 Barrett AJ Kembhavi AA Brown MA Kirschke H Knight CG Tamai M Hanada K L-trans-Epoxysuccinyl-leucylamido(4-guanidino) butane (E-64) and its analogues as inhibitors of cysteine proteinases including cathepsins B, H and L. Biochem. J 1982 201 189 198 7044372 Camargo FD Gokhale S Johnnidis JB Fu D Bell GW Jaenisch R Brummelkamp TR YAP1 increases organ size and expands un-differentiated progenitor cells. Curr. Biol 2007 17 2054 2060 17980593 Chan HM La Thangue NB p300/CBP proteins: HATs for transcriptional bridges and scaffolds. J. Cell Sci 2001 114 2363 2373 11559745 Chen HH Mullett SJ Stewart AF Vgl-4, a novel member of the vestigial-like family of transcription cofactors, regulates alpha1-adrenergic activation of gene expression in cardiac myocytes. J. Biol. Chem 2004 279 30800 30806 15140898 de Boer E Rodriguez P Bonte E Krijgsveld J Katsantoni E Heck A Grosveld F Strouboulis J Efficient biotinylation and single-step purification of tagged transcription factors in mammalian cells and transgenic mice. Proc. Natl. Acad. Sci. USA 2003 100 7480 7485 12802011 Del Re DP Yang Y Nakano N Cho J Zhai P Yamamoto T Zhang N Yabuta N Nojima H Yes-associated protein isoform 1 (Yap1) promotes cardiomyocyte survival and growth to protect against myocardial ischemic injury. J. Biol. Chem 2013 288 3977 3988 23275380 Dong J Feldmann G Huang J Wu S Zhang N Comerford SA Gayyed MF Anders RA Maitra A Pan D Elucidation of a universal size-control mechanism in Drosophila and mammals. Cell 2007 130 1120 1133 17889654 Dupont S Morsut L Aragona M Enzo E Giulitti S Cordenonsi M Zanconato F Le Digabel J Forcato M Role of YAP/TAZ in mechanotransduction. Nature 2011 474 179 183 21654799 Galli GG Carrara M Yuan W-C Valdes-Quezada C Gurung B Pepe-Mooney B Zhang T Geeven G Gray NS YAP drives growth by controlling transcriptional pause release from dynamic enhancers. Mol. Cell 2015 60 328 337 26439301 Hata S Hirayama J Kajiho H Nakagawa K Hata Y Katada T Furutani-Seiki M Nishina H A novel acetylation cycle of transcription co-activator Yes-associated protein that is downstream of Hippo pathway is triggered in response to SN2 alkylating agents. J. Biol. Chem 2012 287 22089 22098 22544757 He A Shen X Ma Q Cao J von Gise A Zhou P Wang G Marquez VE Orkin SH Pu WT PRC2 directly methylates GATA4 and represses its transcriptional activity. Genes Dev 2012 26 37 42 22215809 Heallen T Zhang M Wang J Bonilla-Claudio M Klysik E Johnson RL Martin JF Hippo pathway inhibits Wnt signaling to restrain cardiomyocyte proliferation and heart size. Science 2011 332 458 461 21512031 Heallen T Morikawa Y Leach J Tao G Willerson JT Johnson RL Martin JF Hippo signaling impedes adult heart regeneration. Development 2013 140 4683 4690 24255096 Heidenreich PA Albert NM Allen LA Bluemke DA Butler J Fonarow GC Ikonomidis JS Khavjou O Konstam MA Forecasting the impact of heart failure in the United States: a policy statement from the American Heart Association. Circ. Heart Fail 2013 6 606 619 23616602 Huang J Wu S Barrera J Matthews K Pan D The Hippo signaling pathway coordinately regulates cell proliferation and apoptosis by inactivating Yorkie, the Drosophila homolog of YAP. Cell 2005 122 421 434 16096061 Jiao S Wang H Shi Z Dong A Zhang W Song X He F Wang Y Zhang Z A peptide mimicking VGLL4 function acts as a YAP antagonist therapy against gastric cancer. Cancer Cell 2014 25 166 180 24525233 Kajstura J Leri A Finato N Di Loreto C Beltrami CA Anversa P Myocyte proliferation in end-stage cardiac failure in humans. Proc. Natl. Acad. Sci. USA 1998 95 8801 8805 9671759 Koontz LM Liu-Chittenden Y Yin F Zheng Y Yu J Huang B Chen Q Wu S Pan D The Hippo effector Yorkie controls normal tissue growth by antagonizing scalloped-mediated default repression. Dev. Cell 2013 25 388 401 23725764 Lee Y Shioi T Kasahara H Jobe SM Wiese RJ Markham BE Izumo S The cardiac tissue-restricted homeobox protein Csx/ Nkx2.5 physically associates with the zinc finger protein GATA4 and cooperatively activates atrial natriuretic factor gene expression. Mol. Cell Biol 1998 18 3120 3129 9584153 Li J Gao E Vite A Yi R Gomez L Goossens S van Roy F Radice G Alpha-catenins control cardiomyocyte proliferation by regulating yap activity. Circ. Res 2014 116 70 79 25305307 Lin Z Pu WT Strategies for cardiac regeneration and repair. Sci. Transl. Med 2014 6 239rv1 Lin Z Pu WT Releasing YAP from an a-catenin trap increases cardiomyocyte proliferation. Circ. Res 2015 116 9 11 25552687 Lin Z von Gise A Zhou P Gu F Ma Q Jiang J Yau AL Buck JN Gouin KA Cardiac-specific YAP activation improves cardiac function and survival in an experimental murine MI model. Circ. Res 2014 115 354 363 24833660 Nakayama H Chen X Baines CP Klevitsky R Zhang X Zhang H Jaleel N Chua BHL Hewett TE Ca2+- and mitochondrial-dependent cardiomyocyte necrosis as a primary mediator of heart failure. J. Clin. Invest 2007 117 2431 2444 17694179 Oudhoff MJ Freeman SA Couzens AL Antignano F Kuznetsova E Min PH Northrop JP Lehnertz B Barsyte-Lovejoy D Control of the hippo pathway by Set7-dependent methylation of Yap. Dev. Cell 2013 26 188 194 23850191 Pan D The hippo signaling pathway in development and cancer. Dev. Cell 2010 19 491 505 20951342 Prasad K-MR Xu Y Yang Z Acton T French A Robust cardiomyocyte-specific gene expression following systemic injection of AAV: in vivo gene delivery follows a Poisson distribution. Gene Ther 2011 18 43 52 20703310 Schlegelmilch K Mohseni M Kirak O Pruszak J Rodriguez JR Zhou D Kreger BT Vasioukhin V Avruch J Yap1 acts downstream of alpha-catenin to control epidermal proliferation. Cell 2011 144 782 795 21376238 Snippert HJ van der Flier LG Sato T van Es JH van den Born M Kroon-Veenboer C Barker N Klein AM van Rheenen J Intestinal crypt homeostasis results from neutral competition between symmetrically dividing Lgr5 stem cells. Cell 2010 143 134 144 20887898 Söderberg O Gullberg M Jarvius M Ridderstråle K Leuchowius K-J Jarvius J Wester K Hydbring P Bahram F Direct observation of individual endogenous protein complexes in situ by proximity ligation. Nat. Methods 2006 3 995 17072308 Umezawa H Structures and activities of protease inhibitors of microbial origin. Methods Enzymol 1976 45 678 695 1012021 Varelas X Miller BW Sopko R Song S Gregorieff A Fellouse FA Sakuma R Pawson T Hunziker W The Hippo pathway regulates Wnt/beta-catenin signaling. Dev. Cell 2010 18 579 591 20412773 von Gise A Lin Z Schlegelmilch K Honor LB Pan GM Buck JN Ma Q Ishiwata T Zhou B YAP1, the nuclear target of Hippo signaling, stimulates heart growth through cardiomyocyte proliferation but not hypertrophy. Proc. Natl. Acad. Sci. USA 2012 109 2394 2399 22308401 Walsh S Ponten A Fleischmann BK Jovinge S Cardiomyocyte cell cycle control and growth estimation in vivo—an analysis based on cardiomyocyte nuclei. Cardiovasc. Res 2010 86 365 373 20071355 Xin M Kim Y Sutherland LB Qi X McAnally J Schwartz RJ Richardson JA Bassel-Duby R Olson EN Regulation of insulin-like growth factor signaling by Yap governs cardiomyocyte proliferation and embryonic heart size. Sci. Signal 2011 4 ra70 22028467 Xin M Kim Y Sutherland LB Murakami M Qi X McAnally J Porrello ER Mahmoud AI Tan W Hippo pathway effector Yap promotes cardiac regeneration. Proc. Natl. Acad. Sci. USA 2013 110 13839 13844 23918388 Yang D Kita S Liang F Wang C Tian H Ji Y Lončar M Quan Q High sensitivity and high Q-factor nanoslotted parallel quadrabeam photonic crystal cavity for real-time and label-free sensing. Appl. Phys. Lett 2014 105 063118 Yoshida T MCAT elements and the TEF-1 family of transcription factors in muscle development and disease. Arterioscler. Thromb. Vasc. Biol 2008 28 8 17 17962623 Zanconato F Forcato M Battilana G Azzolin L Quaranta E Bodega B Rosato A Bicciato S Cordenonsi M Piccolo S Genome-wide association between YAP/TAZ/TEAD and AP-1 at enhancers drives oncogenic growth. Nat. Cell Biol 2015 17 1218 1227 26258633 Zhang ZD Paccanaro A Fu Y Weissman S Weng Z Chang J Snyder M Gerstein MB Statistical analysis of the genomic distribution and correlation of regulatory elements in the ENCODE regions. Genome Res 2007 17 787 797 17567997 Zhang H von Gise A Liu Q Hu T Tian X He L Pu W Huang X He L Cai CL Yap1 is required for endothelial to mesenchymal transition of the atrioventricular cushion. J. Biol. Chem 2014 289 18681 18692 24831012
PMC005xxxxxx/PMC5121003.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 8100316 1138 Biomaterials Biomaterials Biomaterials 0142-9612 1878-5905 27760399 5121003 10.1016/j.biomaterials.2016.10.019 NIHMS823610 Article Shifts in macrophage phenotype at the biomaterial interface via IL-4 eluting coatings are associated with improved implant integration Hachim Daniel ab LoPresti Samuel T. ab Yates Cecelia C. ac Brown Bryan N. abd* a McGowan Institute for Regenerative Medicine, University of Pittsburgh. 450 Technology Drive, Suite 300. Pittsburgh, PA 15219. b Department of Bioengineering, Swanson School of Engineering, University of Pittsburgh. 302 Benedum Hall, 3700 O'Hara Street, Pittsburgh, PA 15260. c Department of Health Promotion and Development, School of Nursing, University of Pittsburgh. 440 Victoria Building, 3500 Victoria Street, Pittsburgh, PA 15213. d Department of Obstetrics, Gynecology and Reproductive Sciences, University of Pittsburgh. 300 Halket Street, Pittsburgh, PA 15213. * Corresponding author: Bryan N. Brown, Ph.D. Assistant professor. McGowan Institute for Regenerative Medicine, University of Pittsburgh. 450 Technology Drive, Suite 300. Pittsburgh, PA 15219. [email protected], Phone: 412-624-5273 26 10 2016 11 10 2016 1 2017 01 1 2018 112 95107 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. The present study tests the hypothesis that transient, early-stage shifts in macrophage polarization at the tissue-implant interface from a pro-inflammatory (M1) to an anti-inflammatory/regulatory (M2) phenotype mitigates the host inflammatory reaction against a non-degradable polypropylene mesh material and improves implant integration downstream. To address this hypothesis, a nanometer-thickness coating capable of releasing IL-4 (an M2 polarizing cytokine) from an implant surface at early stages of the host response has been developed. Results of XPS, ATR-FTIR and Alcian blue staining confirmed the presence of a uniform, conformal coating consisting of chitosan and dermatan sulfate. Immunolabeling showed uniform loading of IL-4 throughout the surface of the implant. ELISA assays revealed that the amount and release time of IL-4 from coated implants were tunable based upon the number of coating bilayers and that release followed a power law dependence profile. In-vitro macrophage culture assays showed that implants coated with IL-4 promoted polarization to an M2 phenotype, demonstrating maintenance of IL-4 bioactivity following processing and sterilization. Finally, in-vivo studies showed that mice with IL-4 coated implants had increased percentages of M2 macrophages and decreased percentages of M1 macrophages at the tissue-implant interface during early stages of the host response. These changes were correlated with diminished formation of fibrotic capsule surrounding the implant and improved tissue integration downstream. The results of this study demonstrate a versatile cytokine delivery system for shifting early-stage macrophage polarization at the tissue-implant interface of a non-degradable material and suggest that modulation of the innate immune reaction at early stages of the host response may represent a preferred strategy for promoting biomaterial integration and success. Layer by layer coating IL-4 surgical mesh polypropylene macrophage foreign body reaction Introduction The host response to implanted materials begins immediately upon introduction of the material into the host tissue and encompasses multiple overlapping phases including injury, protein adsorption, acute inflammation, chronic inflammation, foreign body reaction, granulation tissue formation and eventual encapsulation [1]. It is well recognized that the early interactions which occur at the material-tissue interface represent the initiating events which drive subsequent paracrine and autocrine processes of the host response and subsequent tissue remodeling with significant implications for downstream performance. Recently, macrophage-implant interactions in particular have received considerable attention as a primary determinant of the outcome of biomaterials placement [2-7]. A spectrum of macrophage phenotypes contained between two extremes has been identified, ranging from pro-inflammatory (M1) to anti-inflammatory/regulatory (M2) phenotypes, with significant implications in disease, tissue remodeling following injury, and biomaterial performance [3, 8-12]. The findings of studies in multiple tissue and organ systems have now demonstrated that materials which elicit improved or regenerative remodeling outcomes are often associated with a shift from an initially M1 to a more M2-like profile during the early stages of the inflammatory response which follows implantation [13-18]. However, many of these studies have been performed as retrospective analyses of macrophage phenotypes associated with successful biomaterials. As a consequence, recent studies have now begun to evaluate improvements in regenerative outcomes following purposeful modulation of macrophage polarization induced by cytokine delivery from degradable materials [19-21]. These studies have demonstrated that outcomes are improved when M2-polarizing cytokines (IL-4, IL-10) are delivered and that, conversely, outcomes are negatively affected by the delivery of M1-polarizing cytokines (IFN-γ) [19, 21, 22]. A recent study, however, showed that sequential delivery of an M1 (IFN-γ) and then an M2 (IL-4) polarizing cytokine enhanced the vascularity and subsequent healing response associated with the implantation of degradable bone scaffolds [21], demonstrating the importance of both the M1 and M2 responses in the remodeling process. Others have demonstrated that the host macrophage response is also important to the performance of permanent implants. For example, it has been demonstrated that wear particle-induced polarization of macrophages towards an M1 phenotype is associated with periprosthetic osteolysis, possibly resulting in the need for surgical revision following total joint replacement [23-26]. Subsequent studies demonstrated that methods including the delivery of inhibitors of macrophage recruitment, inhibitors of M1 polarization, and promoters of M2 polarization have the ability to mitigate wear particle induced osteolysis, potentially improving the long-term performance of total joint replacements [22, 27-29]. The present study sought to examine the effects of surface-localized cytokine delivery in the early macrophage response upon the integration of a non-degradable polypropylene mesh material commonly utilized in the repair of pelvic organ prolapse. These materials have recently been shown to elicit a predominantly M1 type response which is associated with the potential for tissue degradation and downstream complications when unresolved [30, 31]. Macrophage modulation only in early-stages of the host response, but also localized to the tissue-implant interface was sought in the present study, representing an advantage over strategies promoting non-local and/or extended shifts in the host response as it limits the potential for adverse long-term interactions and exacerbation of conditions which may exist at distant sites. Also, nanometer thickness of this coating was desired to preserve the architecture and pore space of the mesh, which is commonly thought to be important for adequate tissue ingrowth and mechanical performance in clinical settings [32, 33]. To accomplish these goals, a nanometer thickness multilayered coating able to control the release of IL-4 from polypropylene mesh in a localized and temporal manner was developed. This coating is based on the layer by layer (LbL) technique [34, 35], consisting of an alternate cyclic deposition of multiple polyelectrolyte layers mediated by opposite electrostatic charges on the surface of a charged substrate. This method has previously been shown to produce a tunable, uniform and conformal coating of nanometer thickness for controlled release of proteins [27, 34-40]. Therefore, the number and sequence of layers can be easily modified in order to provide the desired amount and release time of IL-4. Methods Materials A polypropylene mesh, Gynemesh® PS (Ethicon, Somerville, NJ) was used. Maleic anhydride, chondroitin sulfate B, chitosan (low molecular weight, deacetylation degree 85%), chondroitinase ABC, chitosanase, bovine serum albumin (BSA) and histologic staining materials were purchased from Sigma Aldrich (St. Louis, MO). Murine IL-4, anti-murine IL-4 antibody, murine IL-4 ELISA detection kit were purchased from Peprotech (Rocky Hill, NJ). Mouse arginase-1 antibody (rabbit), anti-rabbit Alexa-fluor 488 (donkey) and anti-rabbit Alexa-fluor 594 (donkey) were purchased from Abcam (Cambridge, MA). Mouse iNOS antibody (rabbit) was purchased from Santa Cruz (Dallas, TX). Mouse F4/80 antibody (rat) was purchased from AbD Serotec/Bio Rad (Raleigh, NC). Anti-rat Alexa-fluor 488 (donkey) and anti-rabbit Alexa-fluor 546 (donkey) were purchased from Thermo Fisher (Pittsburgh, PA). Plasma treatment of polypropylene meshes Polypropylene (PP) meshes were cleaned using a 1:1 acetone:isopropanol mixture and then air dried prior to irradiation with 15 seconds of argon plasma at 600W, an argon gas flow of 35 mL/min and a steady state pressure of 250 mTorr (50 mTorr initial pressure) using an Ion 40 Gas Plasma System (PVA Tepla America, Inc). An adapted radio frequency glow discharge (RFGD) based on a previously developed microwave plasma procedure was used to obtain a negatively charged surface [41]. Briefly, maleic anhydride powder (1.5 gr) was placed into a glass plate inside of the machine chamber. 1 cm2 pieces of PP mesh were then placed around the plate to a distance of 8.5 cm. After an initial pressure of 50 mTorr was reached, 30 seconds of maleic anhydride plasma treatment was performed at 600W, an argon gas flow of 35 mL/min and a steady state pressure of 250 mTorr. Finally, in order to remove the physisorbed maleic anhydride and to hydrolyze the anhydrides and produce carboxylic acid groups (negatively charged at physiological pH), PP meshes were rinsed for 30 minutes with milli-Q water and then boiled for 20 minutes in fresh milli-Q water. Layer by Layer coating of charged polypropylene meshes In order to deposit a conformal coating of nanometer thickness onto the surface of negatively charged PP meshes, a Layer by Layer (LbL) procedure was performed (Figure 1). Chitosan was chosen as polycation and dermatan sulfate (chondroitin sulfate B) as polyanion. Chitosan was dissolved in 0.5 % acetic acid and dermatan sulfate in milli-Q water. Both polyelectrolites were prepared at a concentration of 2 mg/mL. First, meshes were dipped in chitosan for 10 minutes at room temperature, then meshes were washed 3 times (10, 20 and 30 seconds) in milli-Q water and air dried (pressurized clean air). Next, meshes were dipped in a dermatan sulfate solution for 10 minutes at room temperature. Meshes were washed again in milli-Q water and air-dried. This cycle was repeated until a core coating of 10 bilayers (PP−[CH/DS]10) was achieved. After coating, meshes were lyophilized and stored at 4°C. IL-4 Loading of Coated PP Meshes Prior to IL-4 loading onto the meshes, an IL-4 (1.5 μg/mL) - dermatan sulfate (2 mg/mL) mixture was made and incubated overnight at 4°C in order to complex IL-4 into the polyanion. Then, polypropylene meshes with a 10-bilayer core coating were further coated with 20, 40 and 60 bilayers containing IL-4 (PP−[CH/DS]10[CH/DSIL-4]x, where x stands for number of bilayers, and DSIL-4 stands for dermatan sulfate bound IL-4). After coating, IL-4 loaded meshes were lyophilized and stored at −20°C. Coated (no IL-4) meshes were used as controls, using the same numbers of bilayers used for IL-4 loaded meshes. All mesh materials were then terminally sterilized using ethylene oxide. Coating characterization An Alcian blue staining was performed on whole samples to stain the GAG components and reveal the coating. A 1% Alcian blue solution was made in 3% acetic acid and adjusted to pH 2.5. Coated meshes and controls were re-hydrated in distilled water and then immersed into the Alcian blue solution for 30 minutes at RT. Then meshes were washed in running tap water for 5 minutes and rinsed 5 minutes in distilled water. Images were taken using a standard optical camera. Additionally, elemental composition of the coated meshes was performed using an X-ray photoelectron spectroscopy (XPS), using an ESCALAB 250Xi, Thermo Scientific (Pittsburgh, PA). To identify the elements in the coating/surface of the meshes, an initial survey of 10 scans was obtained and for detailed elemental information, spectra of 25 scans were obtained for Carbon, Oxygen, Nitrogen and Sulfur. Spectra data was analyzed using Avantage software, Thermo Scientific. Finally, meshes were analyzed under Fourier transform infrared spectroscopy with attenuated total reflectance (ATR-FTIR) using a Bruker Vertex 70 (Billerica, MA) equipped with a germanium ATR crystal at a resolution of 1 cm−1, 2 mm of aperture, 32 scans and processed by OPUS software to adjust the baseline, to smooth spectra and to remove H2O and CO2 peaks due to environmental noise. IL-4 loading and release assays Immunolabeling was used to qualitatively demonstrate the loading and distribution of IL-4 throughout the coating. IL-4 loaded, coated (no IL-4) and pristine meshes were immersed in a 1% BSA solution to block non-specific adsorption of antibodies (1h, RT). Washing was performed in between each step by dipping the meshes 4 times in 0.05% Tween 20. Then meshes were immersed and incubated in a solution of anti-murine IL-4 (from rabbit) as primary antibody (1:100 in 0.1% BSA, 2 hours, RT). Meshes were then immersed in a solution of anti-rabbit-Alexa Fluor 546 as a secondary antibody (1:100 in 0.1% BSA, 30 min, RT). Mesh fluorescence was observed under confocal microscopy (Leica DMI4000 B, Buffalo Grove, IL), in which an excitation/emission of 480/520 nm was used to observe the mesh autofluorescence (green) and 561/572 nm to observe the specific fluorescence due to the loaded IL-4 (red). Loading efficiency and release assays were performed following manufacturer instructions of the Peprotech IL-4 ELISA kit. First, 1 cm2 pieces of IL-4 loaded (20, 40 and 60 bilayers) and coated (no IL-4) meshes were immersed into 400 μL of a solution 0.05 units/mL chondroitinase ABC and 0.05 units/mL chitosanase in 1X PBS. Incubation was performed to multiple time points at 37°C, after which 400 μL of solution were aliquoted and stored at −80°C until the end of the experiment. After collection, replacement with fresh solution was performed to continue the release assay. To perform the ELISA assays, 100 μL aliquots were used from each sample (N = 9) at each time point. To determine release profile kinetics; correlation and curve fitting analyses were performed using the data from cumulative release versus time, until the first time point where the release reaches a plateau, which corresponds to the total release. To corroborate power law dependence, besides direct curve fitting tests, a linear trend was corroborated using a LOG (cumulative release) versus LOG (time) curve. In-vitro macrophage culture assay An in-vitro macrophage culture assay was performed in order to demonstrate preservation of bioactivity of IL-4 released from the coated meshes. Bone-marrow mononuclear cells were obtained from murine bone marrow as previously described [42], then these cells were seeded in plates and differentiated to macrophages with DMEM, 10% FBS, 10% L929 supernatant, 1% HEPES, 2% MEM NEAA, 0.1% β-2-mercaptoethanol (Sigma Aldrich, St. Louis, MO) for 7 days. 5 × 105 cells were plated into 24-well plates with α-MEM, 10% FBS, 0.05 units/mL of both chondroitinase ABC and chitosanase. Macrophages were exposed to 1 cm2 pieces of pristine, coated (no IL-4) and IL-4 loaded (40 bilayers) meshes. Immunolabeling isotype (rabbit IgG) and soluble IL-4 (20 ng/mL) were used as negative and positive controls, respectively. Cells were incubated at 37°C and 5% CO2 for 72 hours. After incubation, cells were fixed with 2% PFA and then blocked with 2 % horse serum, 1% BSA, 0.1% triton X-100, 0.1% tween-20 for 1 hour at RT. Immunolabeling was performed using anti arginase-1 as primary antibody (1:200, overnight at 4°C) and Alexa Fluor-488 (1:300, 1 hour at RT) as secondary. A 500 nM DAPI solution was used stain nuclei. Images were taken in an array of 3 × 3 images per each well using a Carl Zeiss Observer.Z1 microscope and then the intensity of arginase-1 staining was analyzed using Cell Profiler Image Analysis Software (Broad Institute, Cambridge, MA) using the same number of cells for all tested conditions. In-vivo mouse mesh implantation An implantation model using C57BL/6J female mice, 8 – 10 weeks old was used, following proper housing and treatment procedures approved by the Institutional Animal Care and Use Committee (IACUC) of the University of Pittsburgh and the National Institutes of Health Guide for the Care and Use of Laboratory Animals. A power analysis was performed to determine that 7 animals per group was required to maintain a statistical power of 80 %. Briefly, a midline incision was made and a subcutaneous pocket was created in the abdomen of each mouse in order to implant a 1 cm2 piece of pristine, coated (no IL-4) and IL-4 loaded (40 bilayers) meshes. PCL sutures were used to close the incision, then 0.5 mg/kg of Baytril and 0.2 mg/kg of Buprenex were administered for 3 days as antibiotic and analgesic, respectively. Buprenorphine (Buprenex), an opioid analgesic, has been studied and shown not to exert any effects nor alterations in the immunological response, both acutely and chronically administered [43, 44]. After 7, 14 or 90 days, mice were euthanized and skin/mesh/muscle tissue complex were harvested and fixed for 72 hours in neutral buffered formalin. Finally, fixed tissues were paraffin embedded and cross-sections of 7 μm were used for histological studies. Immunolabeling of histological sections Paraffin embedded tissue sections were deparaffinized and hydrated in a series of xylene/alcohol/water. Incubation with proteinase K (1X) for 10 minutes at 37°C was performed to retrieve antigens. After 3 washes in water, samples were incubated at 37°C in 50 mM of CuSO4 in 10 mM NH4Ac buffer pH = 5, to reduce tissue background fluorescence. Slides were washed twice in TBST (25 mM Tris buffer + 0.1% tween 20). Then, a 5% donkey serum + 2% BSA + 0.1% tween 20 + 0.1% triton X-100 solution was used as blocking agent (2 hours, RT). To immunolabel M2 macrophages, arginase-1 (1:100) and F4/80 (1:50) primary antibodies were used (overnight at 4°C), followed by anti-rabbit Alexa Fluor 594 (1:200) and anti-rat Alexa Fluor 488 (1:100) secondary antibodies (40 min at RT) in blocking buffer. To immunolabel M1 macrophages, iNOS (1:100) and F4/80 (1:50) primary antibodies were used (overnight at 4°C), followed by anti-rabbit Alexa Fluor 594 (1:100) and anti-rat Alexa Fluor 488 (1:100) secondary antibodies (40 min at RT) in blocking buffer. Vectashield with DAPI mounting media (Vector laboratories, Burlingame, CA) was used to stain nuclei and mount. Images of centered single fibers (3 different single fibers per sample, N = 8 each group) were taken on a Nikon Eclipse E600 microscope equipped with epi-fluorescence at 40X and cell counts were analyzed using ImageJ (version 1.51a, NIH). Image analysis algorithms were used to quantify the results obtained by imaging of histological tissue sections. First, a custom-designed algorithm (Wolfram Mathematica, version 10.0. Champaign, IL) was used to quantify both arginase-1 and iNOS expression at 7 and 14 days by means of arginase-1/DAPI and iNOS/DAPI pixel ratio versus the distance from the surface of single centered mesh fiber (3 different single fibers, N = 8 each group) images taken from histological tissue sections per sample. Next image analysis was performed using ImageJ (version 1.51a, NIH) in order to quantify the number of pro-inflammatory (iNOS, M1) and regulatory (arginase-1, M2) macrophages (F4/80) surrounding single mesh fibers in each group. Histological stainings Paraffin embedded tissue cross-sections were used for H&E, Masson's trichrome and Picro Sirius Red staining. H&E and Masson's trichrome stained tissue sections were imaged on a Nikon Eclipse E600 microscope (Tokyo, Japan) at 10X and 20X, respectively. Picro Sirius Red stained tissue sections were imaged at 20X on a Nikon Eclipse TE2000-E (Tokyo, Japan), equipped with circularly polarized light. ImageJ (version 1.51a, NIH) equipped with a color deconvolution plug-in (version 1.5) was used to quantify the area and mean thickness (calculated as the mean of apical, basal, and lateral measurements taken perpendicular to the surface of the mesh fiber, Figure S1) of capsule surrounding mesh fibers at 90 days (3 different single fibers per sample, N = 8 each group) in images taken from histological tissue sections stained with Masson's Trichrome. A custom-designed algorithm (Mathworks MathLab, version R2015a. Natick, MA) was used to evaluate quantitatively the distribution of collagen fiber sizes surrounding mesh fibers at 90 days (3 different single fibers per sample, N = 8 each group) in images taken from histological tissue sections stained with Picro Sirius Red. Statistical Analysis Comparisons of means were performed by either one-way or two-way analysis of variance (ANOVA), using at least p < 0.05 as statistical significance criteria followed by Tukey's test to compare groups and Sidak's test to compare time points. Shapiro-Wilk was used to test normality. All statistical tests were performed on GraphPad Prism V7 (La Jolla California, USA). Results and Discussion Surgical mesh plasma irradiation, LbL coating and characterization An adapted radio frequency glow discharge (RFGD) method [41] was used to form a consistent and durable negative charge on the surface of polypropylene (PP) mesh in order to facilitate the desired LbL coating. The presence of a negatively charged surface was confirmed by the appearance of two peaks at 284 eV (C-C) and 288 eV (O-C=O) on the carbon spectrum and a peak at 532 eV on the oxygen spectrum, while pristine mesh only had a peak at 284 eV (C-C) when evaluated by X-ray photoelectron spectroscopy (XPS) (Figure 2a). RFGD treated meshes were then LbL coated using chitosan as a polycation and dermatan sulfate as a polyanion. Chitosan was chosen for its known biocompatibility, antimicrobial activity, and as activated macrophages highly express chitinase-like proteins (chitin and chitosan degrading enzymes) [45-47]. Dermatan sulfate (also known as chondroitin sulfate B) was chosen for its key role in extracellular matrix (ECM) regulation and its described ability to enhance IL-4 bioactivity in-vivo [48]. Thus, a chitosan-dermatan sulfate LbL complex has the potential to provide enhanced release and bioactivity of IL-4 in the context of macrophage mediated host-implant interactions. A coating of 10 bilayers was performed as core coating prior to IL-4 loading. Alcian blue staining was used to visualize the chitosan and dermatan sulfate components of the coating. Blue coloration and absence of precipitates along the mesh surface suggested the presence of a conformal and uniform coating on LbL coated meshes (Figure 2b). Electron microscopy (Figure 3) was used to confirm the conformal nature of the coating and showed no apparent changes in surface topography, porosity and thickness between LbL coated, RFGD treated and pristine meshes. The presence of chitosan in the LbL coating was corroborated by the appearance of two peaks at 399 eV (C-N) and 401 eV (O-C-N) in the nitrogen spectrum and the presence of dermatan sulfate by the appearance of a peak at 168 eV (C-S-O) in the sulfur spectrum when evaluated by XPS (Figure 2a), in addition to the presence of peaks at 288 eV (O-C=O) and 286 eV (C-O) in the carbon spectrum, confirming the presence of both polyelectrolyte chains. These measurements were performed at different points on the surface of the PP mesh and spectra were identical throughout the mesh surface. These findings were consistent with ATR-FTIR measurements (Figure S2). IL-4 loading, release and bioactivity assessments Mesh coated with a 10-bilayer core coating was then coated with 20, 40 and 60 additional bilayers containing IL-4. IL-4 was pre-incubated with dermatan sulfate prior to LbL coating, promoting the loading of the cytokine due to the high affinity of IL-4 (net positive charge, given its isoelectric point of 9.17) for sulfated glycosaminoglycans (negatively charged). Confocal microscopy demonstrated positive IL-4 labeling distributed throughout the entire surface of IL-4 loaded meshes in contrast to the absence of positive labeling on coated (no IL-4) mesh and pristine mesh (Figure 4a). ELISA assays were performed to quantify IL-4 release over time. Results showed that both the amount of IL-4 and the length of release are dependent on the number of bilayers containing IL-4 in the LbL coating (Figure 4b). In particular, the in vitro release of IL-4 was observed up to 14, 22 and 30 days for coatings of 20, 40 and 60 bilayers, respectively. The release profile for all IL-4 loaded meshes followed a power law dependence, regardless of the number of coating bilayers (Figure 4c). These findings are consistent with other studies done on LbL films as a platform to study protein release [36, 37, 40]. Based upon these results, meshes coated with 40 bilayers containing IL-4 were selected for further in-vitro and in-vivo assays, since the coating released about 90% of IL-4 only at early stages of the host response (up to 14 days). All further assays included coated (40 additional bilayers with no IL-4) and pristine mesh groups as control groups. In order to show that IL-4 bioactivity remained after the coating procedure and terminal sterilization by ethylene oxide, an in-vitro macrophage polarization assay was performed using mouse bone marrow-derived macrophages. Macrophages exposed to IL-4 loaded meshes for 72 hours were fixed and immunolabeled against arginase-1, an M2 macrophage specific marker. Image analysis (CellProfiler, Broad Institute, Cambridge, MA) of arginase-1 positive cells revealed that the IL-4 released from the IL-4 loaded mesh remained bioactive and able to polarize macrophages towards an M2 phenotype (Figure 5). No significant increase of arginase-1 was observed for coated mesh compared to pristine mesh. Of note, the pattern of arginase-1 expression following exposure to IL-4 coated meshes was similar to the IL-4 positive control (20 ng/mL) despite the lower levels of IL-4 (2.25 ng/mL) released from the mesh surface at 72h (Figure 5d), suggesting that the coating components may enhance IL-4 bioactivity or that IL-4 is protected by the coating and released gradually. Studies on macrophage polarization and the early-stage host response against implanted mesh A mouse implantation model was used to test the ability of IL-4 loaded mesh to promote an early shift (7 and 14 days) in the polarization of macrophages towards an M2 phenotype in-vivo and to examine the effects of such shifts in macrophage polarization upon downstream tissue remodeling (90 days). 1 cm2 of pristine mesh, coated (no IL-4) mesh and IL-4 loaded mesh (40B) were implanted into a subcutaneous pocket in the abdomen of 8-10 week old female C57BL/6J mice. Mesh and surrounding tissue (muscle and skin) were then harvested at 7 and 14 days post-implantation and used to study macrophage polarization. Sham surgeries (no mesh implantation) were also performed. In sham animals, a normal wound healing process observed (Figure 6, top panel) and was characterized by a transient inflammatory response including significant immune cell infiltration at 7 days which was largely resolved by 14 days post-inflammation with restoration of normal tissue architecture resembling healthy tissue controls. The histologic appearance in mice implanted with mesh was also characterized by the presence of inflammatory cell infiltration in the surgical site at 7 days; however, this reaction was not resolved at 14 days and was largely localized to the area surrounding mesh fibers, regardless of mesh type (Figure 6, bottom panel), thereafter. The presence of foreign body giant cells was noted beginning at 14 days post implantation and at the 90 day time point, regardless of mesh type. While the number and distribution of foreign body giant cells was qualitatively similar across all groups, no attempt was made in the present study to quantify the number of foreign body giant cells. Immunolabeling of F4/80 (pan macrophage marker), arginase-1 (an M2 marker) and inducible nitric oxide synthase (iNOS, an M1 marker) was performed to assess the number, location, and phenotypic profiles of the macrophages within the site of implantation at 7 and 14 days (Figure 7 a, c). In all mesh groups, the number of both arginase-1 and iNOS positive cells was observed to peak within the first 50 μm from the mesh surface (Figure 7 b, d). Therefore, this distance was considered as the tissue-biomaterial interface, where the most important interactions of the biomaterial with the surrounding tissue occur and determine the implant success in the long term. Total cell infiltration around single mesh fibers was assessed by DAPI staining, revealing no differences between groups at 7 or 14 days (Figure 8a). However, the small increases in the number of cells within the remodeling site were observed from 7 to 14 days in the pristine and IL-4 loaded mesh implantation groups. Analysis of F4/80+ macrophage populations revealed a significantly higher presence of F4/80+ cells as a percentage of the total cell population in mice implanted with pristine mesh, compared to both coated (no IL-4) and IL-4 loaded mesh groups at 7 days (Figure 8b). At 14 days, the percentage of F4/80+ cells in the pristine mesh group was significantly reduced and was similar to levels similar to those found in both coated (no IL-4) and IL-4 loaded meshes. The percentage of F4/80+ cells in the implantation site of IL-4 loaded meshes were also significantly decreased compared to 7 days, but these decreases were smaller than those observed for the pristine mesh group. There were no differences in the percentage of F4/80+ cells between coated (no IL-4) and IL-4 loaded mesh groups at 7 or 14 days. These results suggest that the coating may have had an inhibitory effect upon the recruitment of macrophages into the implantation site at early time points. Additional co-labeling was performed for arginase-1 and iNOS to assess the M1/M2 polarization profile of the cells within the implantation site. Results at 7 days post-implantation revealed that mice implanted with IL-4 loaded mesh were associated with an increased percentage of Arg-1+ macrophages (F4/80+) as compared to both coated mesh and pristine mesh groups (Figure 8c). The difference in arginase-1 labeling in the IL-4 loaded mesh group was greatest in the first 40 to 50 μm from the mesh surface (Figure 7b) as compared to both coated and pristine mesh, suggesting that the effects of IL-4 released from the LbL coating are limited to distances up to 50 μm from the surface of the implanted mesh. Coated mesh did not elicit an increase in Arg-1+ macrophages as compared to pristine mesh (Figure 8c). These results are consistent with the in-vitro findings showing significant increases in M2 macrophage polarization only in the IL-4 loaded mesh group (Figure 5). Results at 7 days post-implantation also showed a reduction of iNOS+ macrophages in mice implanted with IL-4 loaded meshes compared to mice implanted with pristine meshes (Figure 8d). Mice implanted with coated mesh also showed a reduction in iNOS+ macrophages compared to the pristine mesh implanted group; however, no significant differences were observed between the coated mesh and IL-4 loaded groups (Figure 8d). These results suggest that the coating material may have impacted the polarization of macrophages towards an M1 profile. Differences in iNOS labeling were observed to peak at 25 μm from the mesh surface of the pristine mesh implanted group at 7 days (Figure 7d), again suggesting that the effects of the coating were limited to the first 50 μm from the mesh surface. Results at 14 days post-implantation revealed a decrease in Arg-1+ macrophages in mice implanted with IL-4 mesh as compared to the 7 day time point; however, the percentage of Arg-1+ macrophages was still higher than both coated (no IL-4) and pristine mesh groups (Figure 8c). The percentage of iNOS+ macrophages at 14 days was found to decline in mice implanted with pristine mesh as compared to 7 days; however, there were no significant differences observed between any groups at the 14 day time point (Figure 8d). When the effects of IL-4 coating upon the percentage of Arg-1+ macrophages at 7 and 14 days were compared (Figure 8c) it can be appreciated that arginase-1 expression in the IL-4 coated group at 14 days returned to levels similar to those observed for pristine mesh at both 7 and 14 days. This suggests that the length of IL-4 release from the LbL coated meshes occurs mostly at early stages of the host response (< 14 days) and that its effects on macrophage polarization in-vivo are declining by 14 days. While increases in the M2 macrophage population can likely be attributed to the release of IL-4 from loaded mesh as demonstrated in vitro, there are two potential rationales for the observed reduction in the number of iNOS+ macrophages in the IL-4 loaded mesh group. First, iNOS expression may be reduced as a consequence of the polarization of the macrophages at the tissue implant interface towards an M2 phenotype, given the known competitive nature of pathways leading to iNOS and arginase-1 expression in mice [49-51]. Second, decreased percentage of iNOS+ macrophages may be due to the effects of the coating components and/or surface plasma treatment upon macrophage polarization. This second mechanism is supported by the significant reduction in the percentage of iNOS+ macrophages (Figure 8d) but also a reduction in F4/80+ cells (Figure 8b) observed in the coated mesh group at 7 days. Therefore, the coating components and/or modified mesh surface themselves appear to have effects in the reduction of M1 macrophages but not in promoting M2 macrophage polarization. Thus, the observed results are likely a combination of mechanisms driving the reduction of M1 macrophages with IL-4 mediated increases in the M2 population. We noted that some Arg-1+ and iNOS+ cells did not express F4/80. This suggests cells other than macrophages may produce Arg-1 and iNOS in the area of implantation, or that a population of macrophages which express other markers such as CD11b or CD68 but not F4/80 are present within the remodeling site. Supplemental Figure 3 shows the percentage of total cells expressing Arg-1 and iNOS co-labeled with F4/80, and Supplemental Figure 4 shows the percentage of total cells expressing only Arg-1 or iNOS. Downstream effects in the host response upon macrophage polarization promoted by implanted meshes Mesh and the surrounding tissue complex were harvested at 90 days post-implantation to evaluate the effects of mesh coating and IL-4 loading upon long-term tissue remodeling outcomes. Image analysis of Masson's trichrome stained histological sections was performed to identify and quantify capsule formation. Results revealed capsule formation around mesh fibers for all groups (Figure 9a); however, IL-4 loaded mesh elicited reduced capsule area and thickness compared to the prominent and dense capsules surrounding fibers of both coated and pristine meshes (Figure 9 b, c). Subsequent analysis of collagen fiber distribution in picrosirius red stained sections was performed using a custom-designed algorithm (Mathworks MatLab R2015a). Circularly polarized light microscopy was able to reveal the relative thickness of the collagen fibers as a function of the color hue from thin green fibers to increasingly thick yellow, orange and red fibers [52]. Results revealed that mice implanted with IL-4 loaded meshes had reduced content of both thick orange and thicker red collagen fibers, compared to both pristine and coated meshes (Figure 10). A concurrent increase in thin yellow and thinner green collagen fibers was found for IL-4 loaded mesh compared to both pristine and coated mesh. While we note that the tissue remodeling process is incomplete at the 90 day time point, these outcomes indicate a change in the quantity and type of the collagen fibers composing the fibrotic capsule. These findings may be relevant to potential improvements in mechanical performance of the implanted mesh in-vivo, and are the subject of future studies. The results of the present study demonstrate that the release of IL-4 from LbL coated mesh promotes shifts in early-stage macrophage polarization that are associated with positive long-term effects such as minimized capsule formation and improved tissue quality and composition compared to coated and pristine meshes. These results also suggest that long-term positive outcomes are due to an early-stage increase in the proportion of M2 macrophages, rather than a decrease in the presence of M1 macrophages or total F4/80+ macrophages, given that coated meshes were also capable of significantly decreasing the proportion of M1 macrophages and F4/80+ macrophages (Figure 7, 8) as compared with pristine mesh, but were not associated with improved tissue remodeling outcomes (Figure 9, 10). The present study also clearly demonstrates that it is possible to transiently shift the early phases of the host response to implants which otherwise elicit a chronic pro-inflammatory response with significant impact upon the tissue remodeling outcome downstream while leaving key implant characteristics such as material properties and porosity intact. It has been previously suggested that excessive long-term polarization towards either an M1 or an M2 phenotypes may have negative effects on remodeling outcomes [3, 13, 53]. Additionally, an increasing number studies have described pathologies associated with an imbalance and long-term presence of M1 or M2 macrophages, including but not limited to cancer, diabetes and atherosclerosis [12, 54, 55]. Therefore, localized and temporal delivery of bioactive agents represents an advantage over strategies promoting systemic and or permanent shifts in the host response as it limits the potential for adverse long-term interactions and exacerbation of conditions which may exist at distant sites. Similarly, promoting transient shifts in macrophage polarization in the early host response represents an improved approach as compared to strategies which seek to evade the host immune response. Previous studies using surface modification of biomaterials and coatings to escape the innate immune system have shown only modest improvements at early stages of the host response against biomaterials and few improvements in long-term performance [56-60]. Finally, the present delivery system represents an advantage over previous delivery approaches, given that significant effects on macrophage polarization are observed at lower, controllable and safer doses (picograms to nanograms), compared to the high doses (nanograms to micrograms) of IL-4 used in previous studies [19, 21, 22]. The systemic release of larger amounts IL-4 may lead to effects upon distal tissues and/or exacerbated and contradictory outcomes associated with a fibrotic process or potential enhancement of the foreign body reaction [61, 62]. At present, the stability of the coating and release profile of IL-4 in vivo is unknown. Alcian blue labeling, which as used to identify the coating in vitro, was used to attempt to identify the coating in histologic sections. The presence of the coating was indistinguishable at any time point which we evaluated in the present study. The lack of Alcian Blue staining around the fibers, however, does not necessarily indicate that the coating is degraded. As the coating is on the nanometer scale, it may not be possible to distinguish the staining using routine histologic methods. In-vivo, both the chitosan and dermatan sulfate components of the coating are mostly degraded by macrophages and other cells participating in the host response. Layer by layer films have shown multiple release mechanism, depending the nature of the polyelectrolytes composing the films, being surface degradation the most predominant mechanism that gradually releases the entrapped bioactive agents by degradation of the most external layer films, followed by more internal layers [63, 64]. Therefore, IL-4 is most likely released by gradual surface degradation of the coating multilayers, mainly triggered by macrophages and other cells of the immune system with enzymatic capacity. However, release by diffusion of IL-4 may also occur to a lesser extent. Additional follow up studies should be performed to specifically address in-vivo release profiles. There were several limitations to the present study. First, while the present study demonstrated positive results from the use of IL-4 as a macrophage polarizing molecule, it should be noted that multiple IL-4 isoforms do exist with potentially different functional effects [65, 66]. The IL-4δ2 isoform of IL-4, for example, has been shown to promote the expression of multiple pro-inflammatory cytokines and the chemotaxis of T and B cells [66]. Thus, careful consideration of the exact formulation and potential adverse effects of the IL-4 containing coating is necessary. Further, it is well known that these pathways and others, including IL-4, are regulated differently in humans and mice [65, 67, 68]. Thus, additional testing must be performed to demonstrate potential relevance to human conditions. The LbL method described in the present manuscript is amenable to the use of other macrophage polarization inducing molecules and alternative molecules can likely be substituted for IL-4 if needed for further development. Second, the present study the primary metric of macrophage polarization was the Arg/iNOS percentage relative to F4/80+ macrophages. While the results of the present study clearly demonstrate a shift in macrophage phenotype at the host-material interface, surface marker labeling is not sufficient to fully characterize the phenotype of the macrophage population. Thus, further investigation is needed to clearly define the resultant phenotype and the potential paracrine mechanisms by which the observed shifts in macrophage phenotype affect the remodeling process. Increases in total cellularity were observed between the 7 and 14 day time points. While these increases were moderate, they were often also accompanied by a decrease in the F4/80+ cell population, suggesting a replacement of the macrophages by other cells. These changes could be associated with recruitment of fibroblasts and other cells involved in late phases of the host response. Studies in both the biomaterials and wound healing literature have shown complex and reciprocal cell-cell crosstalk between macrophages and fibroblasts during the processes of healing and fibrosis [69-72]. In the context of the foreign body reaction, these interactions have been shown to result in increased expression of MCP-1 (a potent recruiter of macrophages and a mitogen for fibroblasts), the inhibition of macrophage secretion of cytokines, and the promotion of fibroblast production of cytokines, among others [71]. Thus, the potential effects of IL-4 upon fibroblasts and subsequent effects upon macrophages and other immune cells are important and should also be considered. Further, it has been shown that macrophages can also drive fibroblasts to form multinucleated cells in-vitro, resembling some characteristics of the foreign body giant cells [72]. The exact implications of such findings and their role in the present study are unknown, but warrant further examination to determine the specific effects of IL-4 coatings upon both macrophages and other cells participating in the remodeling process. Third, in the present study, the evaluation of macrophage phenotype was performed in the area of single fibers in order to simplify quantification of macrophage numbers as well as downstream histologic outcomes around a complex shape. However, we note that recent published studies have demonstrated that density of the mesh is a local driver of the host macrophage response, where increased local mesh density is associated with an increased local pro-inflammatory response [30, 31]. While investigation of the differences in the host response between fibers and knots was beyond the scope of the present study, it will be important to evaluate the effects of IL-4 coated mesh knots in order to illustrate the effects in the host response of the released IL-4 and the coating in the presence of high density and complex geometry of the implant. Fourth, the host inflammatory response is also known to be variable by tissue location and context. Of particular relevance to the present study, the host response to polypropylene mesh in the vagina has been reported to be significantly different than that in the abdominal wall musculature [30, 73]. Thus, testing of the present strategy in more relevant model systems must be performed to demonstrate utility in these applications. Additionally, the present study used only one concentration of IL-4 and only investigated 40 bilayers of IL-4 loaded coating. Further studies will be necessary to determine the ideal number of coating layers and optimal time of release. Ideally, the smallest dose of IL-4 and shortest time of release should be identified to reduce the potential for systemic effects of IL-4 release. Finally, the mechanism by which shifts in macrophage phenotype drives improved implant integration have yet to be identified. While it is logical that the presence of a chronically activated M1 macrophage population at the tissue-implant interface would lead to tissue degradation or encapsulation in the long term, it is unclear how M2 macrophages affect the tissue remodeling process and whether M1 macrophages play any essential role in the early remodeling phase (i.e. phagocytosis, activation of local cell populations, production of chemotactic signals). It is likely that both M1 and M2 macrophages play important roles in tissue remodeling, and that the timing of the phenotypic switch will prove an essential factor in determining the degree of success. Conclusions The presence of a uniform and conformal coating composed of both chitosan and dermatan sulfate is demonstrated. This coating can be loaded with IL-4 in a uniform manner through the entire surface of the mesh, and the amount and length of release can be tuned by simply changing the number and sequence of coating bilayers. The released IL-4 from LbL coated meshes is bioactive and can promote macrophage polarization towards an M2 phenotype both in-vitro and in-vivo. In addition, the effects of the local released IL-4 from LbL coated meshes on macrophage polarization extend up to 50 μm of distance from the mesh surface only at early stages of the host response against biomaterials. At long term, a decreased fibrotic capsule formation surrounding mesh fibers and an improved quality of collagen fibers composing the capsule observed only in mice implanted with IL-4 loaded meshes indicating improved resolution of the foreign body reaction and subsequent tissue remodeling process. Finally, these results are strong evidence to support our hypothesis that early-stage macrophage polarization at the tissue-implant interface towards an M2 phenotype would mitigate the foreign body reaction and hence promote better integration of non-degradable biomaterials into the host tissue in the long term. While the present study focused only upon polypropylene mesh commonly used for soft tissue reconstruction, the methods and findings presented can likely be extended to include other material types and applications. Supplementary Material Acknowledgments Thanks to Rahul Rege and Yuta Umeda (undergraduate students, Bioengineering Department, University of Pittsburgh) for helping with the coating process. Immunolabeling custom-designed algorithm (Wolfram Mathematica, V10.0) was developed by William Barone, Ph.D., Department of Bioengineering, University of Pittsburgh. Picro Sirius Red custom-designed algorithm (MathWorks MathLab, VR2015a) was developed by Christopher Carruthers, Ph.D., Department of Bioengineering, University of Pittsburgh. Thanks to Dr. Pamela Moalli (Women's Research Institute, University of Pittsburgh) for the guidance throughout the course of the study. This work was performed, in part, at the Nanoscale Fabrication and Characterization Facility, a laboratory of the Gertrude E. and John M. Petersen Institute of NanoScience and Engineering; and the Center for Biologic Imaging, both housed at the University of Pittsburgh. Funding: This work was supported by the National Institutes of Health [grant numbers K12HD043441 and R21GM107882]. Figure 1 Schematic of layer by layer coating procedure performed on polypropylene surgical meshes. Figure 2 (a) X-ray photoelectron spectroscopy spectra (XPS) of LbL coated (green), RFGD treated (orange) and pristine (blue) mesh. (b) Images of alcian blue stained 1 cm2 pieces of pristine (i), RFGD treated (ii) and LbL coated meshes (iii). Figure 3 Scanning electron microscopy images at 40X (a-d) and 150X (e-h) of pristine (a, e), RFGD treated (b, f), LbL coated (c, g) and IL-4 loaded [40B] (d, h) meshes. Scale bars represent 200 μm. Figure 4 (a) Confocal microscopy images of IL-4 immunolabeled (red) polypropylene fibers (green) of pristine (i), coated [no IL-4] (ii) and IL-4 loaded [40B] (iii) mesh. Scale bars represent 100 μm. (b) Cumulative release of IL-4 (nanograms) versus time (days) from 1 cm2 pieces of IL-4 loaded mesh (20, 40 and 60 bilayers). Coated (no IL-4) mesh was used as a control. Power law dependence curves are y = 0.363x0.262 (r2 = 0.995), y = 0.718x0.437 (r2 = 0.997) and y = 1.078x0.412 (r2 = 0.983) for 20, 40 and 60 bilayers, respectively. (c) Log - log linear fittings of IL-4 cumulative release versus time. Linear equations are y = 0.242x - 0.440 (r2 = 0.995), y = 0.347x - 0.144 (r2 = 0.997), y = 0.412x + 0.033 (r2 = 0.983) for 20, 40 and 60 bilayers, respectively. Points represent the mean ± SEM. Figure 5 (a) CellProfiler image analysis from arginase-1 immunolabeled murine macrophages in an in-vitro culture exposed to 1 cm2 pieces of pristine (yellow), coated [no IL-4] (green) and IL-4 loaded [40B] (blue) mesh. Isotype (gray) and IL-4 [20 ng/mL] (red) were used as negative and positive controls, respectively. (b) Number of arginase-1 positive macrophages determined from the CellProfiler analysis. Bars represent the mean ± SEM. Statistical significance as () p < 0.01 and () p < 0.001, using one-way ANOVA with Tukey's test. (ns) Non-significant. (c) Arginase-1 immunolabelled bone marrow-derived macrophage cultures exposed to 1 cm2 pieces of pristine, coated (no IL-4) and IL-4 loaded (40B) meshes for 72 hrs. IL-4 (20 ng/mL) was used as positive control. Scale bars represent 100 μm. (d) Concentration of IL-4 released by IL-4 loaded meshes (40 bilayers) for 72 hours. Bar represents the mean ± SEM. Figure 6 H&E stained tissue sections at 10X from mice implanted with a 1 cm2 piece of pristine, coated (no IL-4) and IL-4 loaded (40B) mesh at 7 days (top panel) and 14 days (bottom panel). Healthy and SHAM (no mesh surgery) were used as controls. Scale bars represent 200 μm. Figure 7 Fluorescence microscopy images of (a) Arginase-1 (red) and F4/80 (green) co-immunolabeling, and (c) iNOS (red) and F4/80 (green) co-immunolabeling at a single mesh fiber of tissue cross sections of mice implanted with a 1 cm2 piece of pristine, coated (no IL-4) and IL-4 loaded (40B) mesh for 7 days (top panel) and 14 days (bottom panel). DAPI was used to stain cell nuclei. Scale bars represent 50 μm. (b) Arginase-1/DAPI pixel ratio and (d) iNOS/DAPI pixel ratio versus distance of arginase-1 and iNOS immunolabeled tissue sections at 7 days, respectively. Points represent the mean ± SEM (N = 8). Figure 8 Image analysis of (a) total cells (DAPI) and (b) F4/80+ cells as percentages of total cells (DAPI) surrounding single mesh fibers of tissue cross sections of mice implanted with a 1 cm2 piece of pristine, coated (no IL-4) and IL-4 loaded (40B) mesh for 7 days and 14 days. Image analysis of (c) Arg-1+ F4/80+ cells and (d) iNOS+ F4/80+ cells as percentages of total F4/80+ cells surrounding single mesh fibers of tissue cross sections of mice implanted with a 1 cm2 piece of pristine, coated (no IL-4) and IL-4 loaded (40B) mesh for 7 days and 14 days. Bars and points represent the mean ± SEM (N = 8). Statistical significance as () p < 0.05, () p < 0.01, () p < 0.001 and (*) p < 0.0001, using two-way ANOVA with Tukey's (groups) and Sidak's (days) tests. All other differences are non-significant. Figure 9 (a) Masson's Trichrome stained tissue sections of mice implanted with a 1 cm2 piece of pristine, coated (no IL-4) and IL-4 loaded (40B) mesh at 90 days. Arrowheads indicate the capsule surrounding single mesh fibers. Scale bars represent 200 μm. (b) Image analysis of capsule deposition (area %) and (c) Mean thickness surrounding mesh fibers (3 images of a single fiber at 20X per sample, N = 8 samples). Bars represent the mean ± SEM. Statistical significance as () p < 0.01 and (**) p < 0.0001, using two-way ANOVA with Tukey's test. All other differences are non-significant. Figure 10 (a) Picro Sirius Red stained tissue sections (20X) of mice implanted with a 1 cm2 piece of pristine, coated (no IL-4) and IL-4 loaded (40B) mesh at 90 days. Arrowheads indicate the capsule surrounding single mesh fibers. Scale bars represent 100 μm. (b) Image analysis of collagen capsule quality, surrounding mesh fibers (3 images of a single fiber at 20X per sample, N = 8). Bars represent the mean ± SEM. Statistical significance as () p < 0.01, (*) p < 0.001 and (**) p < 0.0001, using two-way ANOVA with Tukey's test. All other differences are non-significant. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. References 1 Anderson JM Rodriguez A Chang DT Foreign body reaction to biomaterials Seminars in immunology 2008 20 2 86 100 18162407 2 Brown BN Badylak SF Expanded applications, shifting paradigms and an improved understanding of host-biomaterial interactions Acta Biomater 2013 9 2 4948 55 23099303 3 Brown BN Ratner BD Goodman SB Amar S Badylak SF Macrophage polarization: an opportunity for improved outcomes in biomaterials and regenerative medicine Biomaterials 2012 33 15 3792 802 22386919 4 Miron RJ Bosshardt DD OsteoMacs: Key players around bone biomaterials Biomaterials 2016 82 1 19 26735169 5 Barbeck M Motta A Migliaresi C Sader R Kirkpatrick CJ Ghanaati S Heterogeneity of biomaterial-induced multinucleated giant cells: Possible importance for the regeneration process? J Biomed Mater Res A 2016 104 2 413 8 26422451 6 Anderson JM Jones JA Phenotypic dichotomies in the foreign body reaction Biomaterials 2007 28 34 5114 20 17706278 7 Bryers JD Giachelli CM Ratner BD Engineering biomaterials to integrate and heal: the biocompatibility paradigm shifts Biotechnology and bioengineering 2012 109 8 1898 911 22592568 8 Xue J Schmidt SV Sander J Draffehn A Krebs W Quester I De Nardo D Gohel TD Emde M Schmidleithner L Ganesan H Nino-Castro A Mallmann MR Labzin L Theis H Kraut M Beyer M Latz E Freeman TC Ulas T Schultze JL Transcriptome-based network analysis reveals a spectrum model of human macrophage activation Immunity 2014 40 2 274 88 24530056 9 Mosser DM Edwards JP Exploring the full spectrum of macrophage activation Nature reviews. Immunology 2008 8 12 958 69 10 Brown BN Sicari BM Badylak SF Rethinking regenerative medicine: a macrophage-centered approach Frontiers in immunology 2014 5 510 25408693 11 Murray PJ Wynn TA Protective and pathogenic functions of macrophage subsets Nature reviews. Immunology 2011 11 11 723 37 12 Mills CD M1 and M2 Macrophages: Oracles of Health and Disease Critical reviews in immunology 2012 32 6 463 88 23428224 13 Brown BN Londono R Tottey S Zhang L Kukla KA Wolf MT Daly KA Reing JE Badylak SF Macrophage phenotype as a predictor of constructive remodeling following the implantation of biologically derived surgical mesh materials Acta Biomater 2012 8 3 978 87 22166681 14 Brown BN Valentin JE Stewart-Akers AM McCabe GP Badylak SF Macrophage phenotype and remodeling outcomes in response to biologic scaffolds with and without a cellular component Biomaterials 2009 30 8 1482 91 19121538 15 Fearing BV Van Dyke ME In vitro response of macrophage polarization to a keratin biomaterial Acta Biomater 2014 10 7 3136 44 24726958 16 Guo R Merkel AR Sterling JA Davidson JM Guelcher SA Substrate modulus of 3D-printed scaffolds regulates the regenerative response in subcutaneous implants through the macrophage phenotype and Wnt signaling Biomaterials 2015 73 85 95 26406449 17 Madden LR Mortisen DJ Sussman EM Dupras SK Fugate JA Cuy JL Hauch KD Laflamme MA Murry CE Ratner BD Proangiogenic scaffolds as functional templates for cardiac tissue engineering Proc Natl Acad Sci U S A 2010 107 34 15211 6 20696917 18 Sussman EM Halpin MC Muster J Moon RT Ratner BD Porous implants modulate healing and induce shifts in local macrophage polarization in the foreign body reaction Ann Biomed Eng 2014 42 7 1508 16 24248559 19 Mokarram N Merchant A Mukhatyar V Patel G Bellamkonda RV Effect of modulating macrophage phenotype on peripheral nerve repair Biomaterials 2012 33 34 8793 801 22979988 20 Reeves AR Spiller KL Freytes DO Vunjak-Novakovic G Kaplan DL Controlled release of cytokines using silk-biomaterials for macrophage polarization Biomaterials 2015 73 272 83 26421484 21 Spiller KL Nassiri S Witherel CE Anfang RR Ng J Nakazawa KR Yu T Vunjak-Novakovic G Sequential delivery of immunomodulatory cytokines to facilitate the M1-to-M2 transition of macrophages and enhance vascularization of bone scaffolds Biomaterials 2015 37 194 207 25453950 22 Pajarinen J Tamaki Y Antonios JK Lin TH Sato T Yao Z Takagi M Konttinen YT Goodman SB Modulation of mouse macrophage polarization in vitro using IL-4 delivery by osmotic pumps J Biomed Mater Res A 2015 103 4 1339 45 25044942 23 Goodman SB Wear particles, periprosthetic osteolysis and the immune system Biomaterials 2007 28 34 5044 8 17645943 24 Jamsen E Kouri VP Olkkonen J Cor A Goodman SB Konttinen YT Pajarinen J Characterization of macrophage polarizing cytokines in the aseptic loosening of total hip replacements Journal of orthopaedic research : official publication of the Orthopaedic Research Society 2014 32 9 1241 6 24897980 25 Lin TH Kao S Sato T Pajarinen J Zhang R Loi F Goodman SB Yao Z Exposure of polyethylene particles induces interferon-gamma expression in a natural killer T lymphocyte and dendritic cell coculture system in vitro: a preliminary study J Biomed Mater Res A 2015 103 1 71 5 24616165 26 Pajarinen J Jamsen E Konttinen YT Goodman SB Innate immune reactions in septic and aseptic osteolysis around hip implants Journal of long-term effects of medical implants 2014 24 4 283 96 25747031 27 Keeney M Waters H Barcay K Jiang X Yao Z Pajarinen J Egashira K Goodman SB Yang F Mutant MCP-1 protein delivery from layer-by-layer coatings on orthopedic implants to modulate inflammatory response Biomaterials 2013 34 38 10287 95 24075408 28 Lin TH Pajarinen J Sato T Loi F Fan C Cordova LA Nabeshima A Gibon E Zhang R Yao Z Goodman SB NF-kappaB decoy oligodeoxynucleotide mitigates wear particle-associated bone loss in the murine continuous infusion model Acta Biomater 2016 29 Sato T Pajarinen J Behn A Jiang X Lin TH Loi F Yao Z Egashira K Yang F Goodman SB The Effect of Local IL-4 Delivery or CCL2 Blockade on Implant Fixation and Bone Structural Properties in a Mouse Model of Wear Particle Induced Osteolysis J Biomed Mater Res A 2016 30 Brown BN Mani D Nolfi AL Liang R Abramowitch S Moalli PA Characterization of the host inflammatory response following implantation of prolapse mesh in rhesus macaque Am J Obstet Gynecol 2015 31 Nolfi AL Brown BN Liang R Palcsey SL Bonidie MJ Abramowitch SD Moalli PA Host response to synthetic mesh in women with mesh complications Am J Obstet Gynecol 2016 32 Orenstein SB Saberski ER Kreutzer DL Novitsky YW Comparative analysis of histopathologic effects of synthetic meshes based on material, weight, and pore size in mice J Surg Res 2012 176 2 423 9 22099590 33 Klinge U Klosterhalfen B Birkenhauer V Junge K Conze J Schumpelick V Impact of polymer pore size on the interface scar formation in a rat model J Surg Res 2002 103 2 208 14 11922736 34 Borges J Mano JF Molecular interactions driving the layer-by-layer assembly of multilayers Chem Rev 2014 114 18 8883 942 25138984 35 Ai H Jones SA Lvov YM Biomedical applications of electrostatic layer-by-layer nano-assembly of polymers, enzymes, and nanoparticles Cell Biochem Biophys 2003 39 1 23 43 12835527 36 Macdonald M Rodriguez NM Smith R Hammond PT Release of a model protein from biodegradable self assembled films for surface delivery applications J Control Release 2008 131 3 228 34 18721835 37 Macdonald ML Rodriguez NM Shah NJ Hammond PT Characterization of tunable FGF-2 releasing polyelectrolyte multilayers Biomacromolecules 2010 11 8 2053 9 20690713 38 Shah NJ Hyder MN Moskowitz JS Quadir MA Morton SW Seeherman HJ Padera RF Spector M Hammond PT Surface-mediated bone tissue morphogenesis from tunable nanolayered implant coatings Sci Transl Med 2013 5 191 191ra83 39 Shah NJ Hyder MN Quadir MA Dorval Courchesne NM Seeherman HJ Nevins M Spector M Hammond PT Adaptive growth factor delivery from a polyelectrolyte coating promotes synergistic bone tissue repair and reconstruction Proc Natl Acad Sci U S A 2014 111 35 12847 52 25136093 40 Shah NJ Macdonald ML Beben YM Padera RF Samuel RE Hammond PT Tunable dual growth factor delivery from polyelectrolyte multilayer films Biomaterials 2011 32 26 6183 93 21645919 41 Aumsuwan N Ye SH Wagner WR Urban MW Covalent attachment of multilayers on poly(tetrafluoroethylene) surfaces Langmuir 2011 27 17 11106 10 21800880 42 Sicari BM Dziki JL Siu BF Medberry CJ Dearth CL Badylak SF The promotion of a constructive macrophage phenotype by solubilized extracellular matrix Biomaterials 2014 35 30 8605 12 25043569 43 Sacerdote P Opioids and the immune system Palliat Med 2006 20 Suppl 1 s9 15 16764216 44 Martucci C Panerai AE Sacerdote P Chronic fentanyl or buprenorphine infusion in the mouse: similar analgesic profile but different effects on immune responses Pain 2004 110 1-2 385 92 15275790 45 Shukla SK Mishra AK Arotiba OA Mamba BB Chitosan-based nanomaterials: a state-of-theart review Int J Biol Macromol 2013 59 46 58 23608103 46 Kong M Chen XG Xing K Park HJ Antimicrobial properties of chitosan and mode of action: a state of the art review Int J Food Microbiol 2010 144 1 51 63 20951455 47 Lee CG Da Silva CA Dela Cruz CS Ahangari F Ma B Kang MJ He CH Takyar S Elias JA Role of chitin and chitinase/chitinase-like proteins in inflammation, tissue remodeling, and injury Annu Rev Physiol 2011 73 479 501 21054166 48 den Dekker E Grefte S Huijs T ten Dam GB Versteeg EM van den Berk LC Bladergroen BA van Kuppevelt TH Figdor CG Torensma R Monocyte cell surface glycosaminoglycans positively modulate IL-4-induced differentiation toward dendritic cells J Immunol 2008 180 6 3680 8 18322173 49 Porta C Riboldi E Ippolito A Sica A Molecular and epigenetic basis of macrophage polarized activation Seminars in immunology 2015 27 4 237 48 26561250 50 Tugal D Liao X Jain MK Transcriptional control of macrophage polarization Arterioscler Thromb Vasc Biol 2013 33 6 1135 44 23640482 51 Lawrence T Natoli G Transcriptional regulation of macrophage polarization: enabling diversity with identity Nature reviews. Immunology 2011 11 11 750 61 52 Nadkarni SK Pierce MC Park BH de Boer JF Whittaker P Bouma BE Bressner JE Halpern E Houser SL Tearney GJ Measurement of collagen and smooth muscle cell content in atherosclerotic plaques using polarization-sensitive optical coherence tomography J Am Coll Cardiol 2007 49 13 1474 81 17397678 53 Jetten N Roumans N Gijbels MJ Romano A Post MJ de Winther MP van der Hulst RR Xanthoulea S Wound administration of M2-polarized macrophages does not improve murine cutaneous healing responses PLoS One 2014 9 7 e102994 25068282 54 Wynn TA Chawla A Pollard JW Macrophage biology in development, homeostasis and disease Nature 2013 496 7446 445 55 23619691 55 Liu YC Zou XB Chai YF Yao YM Macrophage polarization in inflammatory diseases Int J Biol Sci 2014 10 5 520 9 24910531 56 Udpa N Iyer SR Rajoria R Breyer KE Valentine H Singh B McDonough SP Brown BN Bonassar LJ Gao Y Effects of chitosan coatings on polypropylene mesh for implantation in a rat abdominal wall model Tissue Eng Part A 2013 19 23-24 2713 23 23859182 57 Meyers SR Grinstaff MW Biocompatible and bioactive surface modifications for prolonged in vivo efficacy Chem Rev 2012 112 3 1615 32 22007787 58 van Bilsen PH Popa ER Brouwer LA Vincent J Taylor CE de Leij LF Hendriks M van Luyn MJ Ongoing foreign body reaction to subcutaneous implanted (heparin) modified Dacron in rats J Biomed Mater Res A 2004 68 3 423 7 14762921 59 Liu L Chen G Chao T Ratner BD Sage EH Jiang S Reduced foreign body reaction to implanted biomaterials by surface treatment with oriented osteopontin J Biomater Sci Polym Ed 2008 19 6 821 35 18534099 60 Khandwekar AP Patil DP Hardikar AA Shouche YS Doble M In vivo modulation of foreign body response on polyurethane by surface entrapment technique J Biomed Mater Res A 2010 95 2 413 23 20648535 61 McNally AK Anderson JM Interleukin-4 induces foreign body giant cells from human monocytes/macrophages. Differential lymphokine regulation of macrophage fusion leads to morphological variants of multinucleated giant cells Am J Pathol 1995 147 5 1487 99 7485411 62 Jones JA McNally AK Chang DT Qin LA Meyerson H Colton E Kwon IL Matsuda T Anderson JM Matrix metalloproteinases and their inhibitors in the foreign body reaction on biomaterials J Biomed Mater Res A 2008 84 1 158 66 17607751 63 Smith RC Riollano M Leung A Hammond PT Layer-by-layer platform technology for small-molecule delivery Angew Chem Int Ed Engl 2009 48 47 8974 7 19847838 64 Choi D Hong J Layer-by-layer assembly of multilayer films for controlled drug release Arch Pharm Res 2014 37 1 79 87 24277693 65 Luzina IG Keegan AD Heller NM Rook GA Shea-Donohue T Atamas SP Regulation of inflammation by interleukin-4: a review of “alternatives” Journal of leukocyte biology 2012 92 4 753 64 22782966 66 Luzina IG Lockatell V Todd NW Highsmith K Keegan AD Hasday JD Atamas SP Alternatively spliced variants of interleukin-4 promote inflammation differentially Journal of leukocyte biology 2011 89 5 763 70 21285395 67 Murray PJ Wynn TA Obstacles and opportunities for understanding macrophage polarization Journal of leukocyte biology 2011 89 4 557 63 21248152 68 Rath M Muller I Kropf P Closs EI Munder M Metabolism via Arginase or Nitric Oxide Synthase: Two Competing Arginine Pathways in Macrophages Frontiers in immunology 2014 5 532 25386178 69 Wynn TA Vannella KM Macrophages in Tissue Repair, Regeneration, and Fibrosis Immunity 2016 44 3 450 62 26982353 70 Barron L Wynn TA Fibrosis is regulated by Th2 and Th17 responses and by dynamic interactions between fibroblasts and macrophages Am J Physiol Gastrointest Liver Physiol 2011 300 5 G723 8 21292997 71 Holt DJ Chamberlain LM Grainger DW Cell-cell signaling in co-cultures of macrophages and fibroblasts Biomaterials 2010 31 36 9382 94 20932568 72 Holt DJ Grainger DW Multinucleated giant cells from fibroblast cultures Biomaterials 2011 32 16 3977 87 21397323 73 Pierce LM Rao A Baumann SS Glassberg JE Kuehl TJ Muir TW Long-term histologic response to synthetic and biologic graft materials implanted in the vagina and abdomen of a rabbit model Am J Obstet Gynecol 2009 200 5 546 e1 8 19286140
PMC005xxxxxx/PMC5121008.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 8100316 1138 Biomaterials Biomaterials Biomaterials 0142-9612 1878-5905 27816000 5121008 10.1016/j.biomaterials.2016.10.044 NIHMS827598 Article Protease-Degradable Microgels for Protein Delivery for Vascularization Foster Greg A. 1 Headen Devon M. 1 González-García Cristina 12 Salmerón-Sánchez Manuel 12 Shirwan Haval 3 García Andrés J. 1* 1 Woodruff School of Mechanical Engineering and Petit Institute for Bioengineering and Bioscience, Georgia Institute of Technology, Atlanta, GA, U.S.A. 2 School of Engineering, Division of Biomedical Engineering, University of Glasgow, Glasgow, Scotland, U.K. 3 Department of Microbiology and Immunology, University of Louisville, Louisville, KY, U.S.A. Corresponding author: A.J. García Woodruff School of Mechanical Engineering 315 Ferst Dr NW, Atlanta, GA 30332, USA, [email protected] 4 11 2016 28 10 2016 1 2017 01 1 2018 113 170175 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Degradable hydrogels to deliver bioactive proteins represent an emerging platform for promoting tissue repair and vascularization in various applications. However, implanting these biomaterials requires invasive surgery, which is associated with complications such as inflammation, scarring, and infection. To address these shortcomings, we applied microfluidics-based polymerization to engineer injectable poly(ethylene glycol) microgels of defined size and crosslinked with a protease degradable peptide to allow for triggered release of proteins. The release rate of proteins covalently tethered within the microgel network was tuned by modifying the ratio of degradable to non-degradable crosslinkers, and the released proteins retained full bioactivity. Microgels injected into the dorsum of mice were maintained in the subcutaneous space and degraded within 2 weeks in response to local proteases. Furthermore, controlled release of VEGF from degradable microgels promoted increased vascularization compared to empty microgels or bolus injection of VEGF. Collectively, this study motivates the use of microgels as a viable method for controlled protein delivery in regenerative medicine applications. VEGF microfluidics biomaterials hydrogels protein delivery 1. Introduction Synthetic hydrogel microparticles (microgels) have broad biomedical applications including cell encapsulation and transplantation [1-8], wound healing [9], imaging tools [10], and protein and drug delivery [11-15]. Microgels offer additional advantages to the attributes of bulk hydrogels for cell and protein delivery, including delivery via catheters or injection via small diameter needles, which minimizes complications associated with surgery (e.g. trauma, infection, scarring), and preserves native tissue structure without in situ gelling considerations that often limit biomedical applications of bulk hydrogels. Furthermore, when appropriately sized, microgels conform to the geometry of the application site, which facilitates uniform distribution of biomolecules to target sites. Importantly, microgels with different characteristics (e.g., different proteins, release rates) can be synthesized in separate batches and simple co-delivery of the microgels in the desired ratios will result in a “mosaic” formulation resulting in complex or multi-component materials. Of various synthesis routes available to generate synthetic microgels, microfluidics-based polymerization is particularly well-suited for preparing microgels containing proteins and cells because of the aqueous, cytocompatible nature and precise control over particle size of this continuous process [7]. Microgels for protein delivery rely on passive diffusion of the protein through a non-degradable microgel network, and therefore the release kinetics are solely dictated by protein size and microgel mesh size [16]. This inability to modulate protein delivery rate severely hinders the application of microgels to regenerative medicine, immunoengineering, and cancer therapy. We present a strategy to engineer synthetic microgels with protease-degradable crosslinks and tunable protein release kinetics. Furthermore, we demonstrate that these protease-degradable microgels promote in vivo vascularization by controlled release of vascular endothelial growth factor (VEGF) and complete degradation of microgels that allows for tissue ingrowth and remodeling. 2. Materials and Methods 2.1 Microfluidic device fabrication PDMS microfluidic flow focusing devices were fabricated using soft lithography from silicon and SU8 masters. Devices were plasma treated and then bonded directly to glass slides. Microfluidic devices were then heated to 110 °C for 30 minutes to improve PDMS-glass sealing. Prior to use, devices were infused with Aquapel™ for 30 seconds and then purged with nitrogen to render surfaces hydrophobic. 2.2 4-arm poly(ethylene glycol)-maleimide (PEG-4MAL) microgel generation PEG-4MAL (20 kDa, Laysan Bio) was dissolved in phosphate-buffered saline (PBS) at 5% (w/v) then filtered through a 40 μm cell strainer (Corning). For experiments involving the injection of microgels in vivo, microgels were functionalized with GRGDSPC (RGD, Genescript). PEG-4MAL was reacted with 2.0 mM RGD for 30 minutes at 37 °C to create RGD-functionalized macromer. For all other experiments, RGD was not used in the formation of microgels. Crosslinker solutions (DTT (Sigma) or GCRDVPMSMRGGDRCG (VPM, Genescript) or combinations of both) were prepared at predetermined molar concentrations and then adjusted to a pH of 4.5 to slow down gelation kinetics in order to prevent the device from clogging. PEG-4MAL and crosslinker were then infused into the flow-focusing microfluidic device to form polymer droplets. Droplets were formed within an oil solution consisting of light mineral oil (Sigma) mixed with 2% SPAN80 (Sigma) and then collected into a 15 mL conical tube (Falcon). After formation, microgels were washed in PBS five times by centrifugation to remove mineral oil and surfactant. 2.3 Microgel degradation Two hundred microgels were loaded into each well of a 96-well plate. Collagenase or PBS was then added to each well and microgels were incubated at 37 °C for 20 hours. After incubation with protease or PBS, images of each well were acquired using a fluorescent microscope and the total number of microgels in the well was quantified. 2.4 Protein release kinetics Prior to microgel formation, PEG-4MAL was reacted with AlexaFluor488-labeled IgG (rat anti-mouse, Thermo Fisher), AlexaFluor555-labeled IgG (rat anti-mouse, Thermo Fisher), or VEGF165 homodimer (Thermo Fisher) pre-labeled with NHS-Dylight488 (Thermo Fisher). All proteins were reacted with PEG-4MAL at 20 μg/mL for 30 minutes at 37 °C protected from light. After washing, 100 μL of 200 μm diameter microgels were added to transwells with 8 μm pore sizes in a 48 well plate (Corning) then treated with 3.9 or 39 units/mL of type 1 collagenase in 500 μL of PBS (Worthington). The microgels were then maintained in an incubator at 37 °C with gentle shaking. At indicated time points, the supernatant was sampled and analyzed on a plate reader (Biotek). Images of the microgels were acquired on an inverted microscope (Nikon TE 300) with a fluorescent camera (Hamamatsu Orca ER II). 2.5 VEGF bioactivity assay We have previously shown that PEGylation of VEGF homodimer primarily results in a VEGF molecule conjugated to two PEG-4MAL macromers [17]. To confirm PEGylated VEGF maintained bioactivity, human umbilical vein endothelial cells (HUVEC, Lonza) were grown in endothelial growth media (EGM-2, Lonza) and synchronized in growth factor free basal media (EBM-2, Lonza) with 1% fetal bovine serum overnight followed by addition of VEGF, PEG-4MAL conjugated VEGF, PEG-MAL only, or EGM-2 for 24 h. Cell metabolism was assayed by the CellTiter 96 MTS Aqueous Cell Proliferation Assay (Promega). To confirm VEGF released from microgels maintained bioactivity, microgels containing VEGF were incubated in MMP-2 (50 nM) (R&D Systems) for 30 minutes at 37°C followed by addition of TIMP-1 (50 nM). HUVEC were then exposed to released VEGF (100 ng/mL) or soluble VEGF (100 ng/mL) for 24 h and cell metabolism was assayed. 2.6 Microgel vascularization To track microgel retention in vivo, RGD was conjugated with Dylight750 for IVIS imaging or Dylight555 for microscopy imaging then tethered to PEG-4MAL macromer. Under protocols approved by Georgia Tech’s Institutional Animal Care and Use Committee, C57BL/6J mice (Jackson Labs) were anesthetized with 2.5% isoflurane during microgel injection and image acquisition. Backs of mice were shaved, dilapidated with Nair™, and sterilized with 70% ethanol. A 1 mL syringe with a 23 gauge 0.5” needle was loaded with 100 μL of microgels. The tip of needle was then inserted into the subcutaneous space of the dorsum and microgels were slowly injected, taking care not to disturb the native tissue structures. A total of 16 mice received two of the following microgel formulation chosen for this study: VPM + VEGF, VPM – VEGF, DTT + VEGF, VPM + sVEGF. Experimental groups were designed such that 4 samples were used for each group. IVIS Spectrum CT (Perkin Elmer) imaging system was used to track microgel position and persistence over time. At 14 day, following injection, functional vasculature was labeled by perfusing anesthetized mice with 1.0 mg/mL Dylight649-labeled tomato lectin (Vector Labs) via tail vein injection. To wash out excess fluorescent lectin, mice were perfused with saline solution. Mice were then euthanized with CO2 and the regions of the skin where microgels were injected were excised. Microgels and vasculature were imaged using a confocal microscope (Nikon Ti-E with Perfect Focus System and C2-Plus Confocal System) and analyzed with ImageJ software. 3. Results 3.1 Generation of microgels using microfluidics To engineer synthetic microgels, we used a PEG-4MAL macromer, which is crosslinked into a network via a Michael-type addition reaction with thiols. The PEG-4MAL platform outperforms other PEG-based polymers in generating structurally defined hydrogels with stoichiometric incorporation of ligands and improved crosslinking efficiency [17]. In addition, PEG-4MAL exhibits minimal local and systemic inflammation and toxicity and is rapidly excreted in the urine [18], important criteria for in vivo applications. We designed a microfluidic flow focusing device to produce droplets of PEG-4MAL and crosslinker (Fig. 1a). Three independent flow inlets (PEG-4MAL, crosslinker, and mineral oil containing SPAN80 surfactant) were used to produce droplets. After the PEG-4MAL and crosslinker flow streams merge, the solution is focused into an oil-covered droplet where it crosslinks and is then collected at the outlet (Video S1, Supporting Information). Microgels of defined diameters with homogeneous size distribution can be simply produced by changing the inlet flow rates and nozzle size (Fig. 1b,c). To generate protease-degradable microgels, we used the crosslinking peptide GCRDVPMSMRGGDRCG (VPM), which is rapidly cleaved by matrix metalloproteinase (MMP)-1 and MMP-2 proteases [19]. To confirm protease-dependent degradation, we first reacted PEG-4MAL with AlexaFluor488-labeled IgG (AF488-IgG) to form an AF488-IgG-functionalized macromer and then generated 200 μm diameter microgels crosslinked with VPM or dithiothreitol (DTT). Microgels were incubated in type 1 collagenase at 39 units/mL or 3.9 units/mL in buffer solution and imaged on a fluorescence microscope (Fig. 1d). After 20 hours, 100% of the VPM-crosslinked microgels in 39 units/mL collagenase degraded whereas only 25% of VPM-crosslinked microgels in 3.9 units/mL collagenase degraded (Fig. 1e). The DTT-crosslinked microgels did not degrade in the presence of the protease. The degradation of VPM-crosslinked microgels was dependent on the concentration of collagenase used (Fig. S1, Supporting Information). These results show that VPM-crosslinked microgels are degradable by proteases in a concentration-dependent fashion. 3.2 Tuning microgel release kinetics In order to engineer microgels with tunable sensitivity to proteases, and therefore tunable release kinetics, we prepared microgels that were crosslinked with varying VPM to DTT molar ratios. Fluorescent IgG-loaded microgels were incubated in the presence of proteases, the supernatant was sampled at indicated intervals, and fluorescent signal was measured. As expected, protein released most quickly from microgels that were crosslinked exclusively with protease-sensitive VPM crosslinker, while microgels crosslinked with protease-insensitive DTT did not release significant protein after one hour in protease, suggesting that the ~20% protein release observed resulted from passive diffusion of untethered protein (Fig. 2a). By varying the crosslinker ratio of MMP-sensitive VPM to protease-insensitive DTT, the degradation rate of capsules in collagenase can be controlled, and therefore the release rate of encapsulated protein can be controlled. Protein release in all groups had plateaued after 3 days, suggesting that remaining crosslinks were protease-insensitive. Importantly, microgels incubated in the absence of protease did not undergo degradation for any crosslinker formulation tested, as evidenced by minimal protein release over 80 hours (Fig. 2b). These results demonstrate that the degradation and release rate of proteins encapsulated within microgels can be engineered. Furthermore, mixtures of different microgels, in this case protease-degradable and non-degradable, can be prepared in order to co-deliver bioactive molecules with different release rates (Video S2, Fig. S2, Supporting Information). 3.3 In vitro VEGF release kinetics An important application for hydrogels is delivery of VEGF protein to promote vascularization [20]. We generated VEGF-containing microgels that degrade in the presence of protease to explore the application of the microgel delivery platform in regenerative medicine. Consistent with our previous observations [18], covalently tethering of VEGF to PEG-4MAL through its free cysteine does not affect its bioactivity (Fig. S3, Supporting Information). To measure the release kinetics, VEGF was labeled with a fluorescent dye and then reacted with PEG-4MAL. The microfluidic focusing device was then used to generate 200 μm diameter VPM- or DTT-crosslinked microgels (Fig. 3a). VEGF was released for VPM-crosslinked microgels incubated in protease, whereas minimal VEGF release occurred over 3 days from DTT-crosslinked microgels incubated in protease as well as from VPM-crosslinked microgels incubated in saline (Fig. 3b). In a separate set of experiments, unlabeled VEGF was released from VPM crosslinked microgels by incubating microgels in MMP-2. Upon complete microgel degradation, VEGF was incubated in TIMP-1 in order to inhibit MMP-2 activity. VEGF released from the microgels exhibited similar bioactivity levels compared to soluble VEGF (Fig 3c). These results confirm protease-dependent release of bioactive VEGF from VPM-crosslinked microgels, consistent with observations seen in IgG-loaded microgels. Notably, the release kinetics were similar to that observed for AF488-IgG-containing microgels, indicating release kinetics independent of protein size. 3.4 Microgel injection and vascularization in vivo To examine the ability of VEGF-releasing microgels to promote vascularization in vivo, we injected different microgel formulations in the subcutaneous space in the dorsum of mice and measured retention of VEGF and functional vascular ingrowth at the site of injection (Fig. 4). Microgel suspensions were simply injected using a standard tuberculin syringe with no incision required, and none of the time constraints due to crosslinking kinetics that accompany injectable bulk gels. The simplicity of microgel injection lends itself to clinical applications where more complex schemes could produce heterogeneous material properties (due to inadequate mixing) or more trauma at the implant site (due to incision), especially in non-expert hands. In order to track microgels and promote tissue repair at the injection site, microgels were covalently functionalized with 2.0 mM of the fluorescently labeled adhesive peptide GRGDSPC (‘RGD’) by co-incubation of labeled RGD and PEG-4MAL before microgel generation. Using an intravital imaging system (IVIS), we monitored the position and fluorescence intensity of injected microgels over 14 days in vivo (Fig. 4a). The fluorescent signal decreased exponentially over time for the VPM-crosslinked microgels (half-life = 1.3 days). In contrast, the signal from DTT-crosslinked gels decreased initially but stabilized to ~80% of the original signal after day 1 and remained relatively unchanged thereafter. (Fig. 4b). This result indicates that VPM-crosslinked microgels degrade in a well-defined pattern in vivo, while DTT crosslinked microgels do not degrade significantly over 14 days. At day 14, the circulatory system of the mice was perfused with fluorescently labeled lectin to stain functional vasculature. Confocal images of explanted skin regions where microgels were injected show significant increases in the number of blood vessels for VEGF-encapsulated VPM-crosslinked microgels compared to empty (no VEGF) VPM-crosslinked microgels and to DTT-crosslinked gels loaded with VEGF (Fig. 4c,d). Importantly, a bolus of soluble VEGF (sVEGF) co-delivered with empty VPM-crosslinked microgels resulted in minimal vascularization, indicating that controlled release of VEGF at the treatment site is required to achieve robust vascularization. Degradation of VPM-crosslinked microgels promoted host tissue ingrowth, while DTT-crosslinked microgels were still present at the injection site and prevented tissue ingrowth and remodeling (Fig. 4c). These results demonstrate that controlled VEGF delivery from protease-degradable microgels promotes robust vascularization and tissue remodeling in vivo. 4. Discussion Implantation of devices via surgical incisions is often associated with complications such as inflammation, infection, trauma, and scarring. Injectable biomaterials mitigate these complications and are thus an attractive means for cell and drug delivery applications. Here, we present an injectable microgel platform to deliver VEGF and promote vascularization. Mice that received degradable microgels releasing VEGF exhibited increased vessel formation compared to mice that received empty microgels or a bolus injection of VEGF. We attribute this increase in vascularization to 1) sustained release of VEGF from the microgels, and 2) host cell binding to RGD within microgels that provide a scaffold for tissue ingrowth and vascularization. Previous reports have shown that sustained release of VEGF combined with a cell-adhesive biomaterial scaffold are critical driving factors for improved vascularization [21-23]. However, the need for invasive surgery to implant bulk, pre-formed devices limits the translation of these tools to the clinic. In situ gelation offers a solution for noninvasive delivery of biomaterials, though the choice of polymer and crosslinker is often limited by physiological conditions that affect gelling such as temperature, pH, and the presence of ions [24]. Our results support a microgel-based drug delivery system as an effective method for delivering VEGF and potentially other bioactive molecules. In a previous report, we utilized a microfluidic device to generate non-degradable PEG-4MAL microgels crosslinked in a DTT/oil emulsion [6]. This microfluidic device, however, could not produce microgels crosslinked with peptides due to the limited solubility of these peptides in oil. We therefore designed a unique microfluidic device to generate protease-degradable microgels by reacting PEG-4MAL with VPM. This new microfluidic device brings together streams of PEG-4MAL macromer and VPM and the mixture is pinched off into a droplet by an oil stream. The new device incorporates a serpentine channel which is necessary for allowing sufficient crosslinking of the microgels before collection into a conical tube. We demonstrate precise control over microgel size using the PDMS based microfluidic flow focusing device. For subcutaneous injections, relatively large diameter particles (~200 μm) and needles did not significantly damage native structures. For more delicate procedures which would require a smaller diameter needle, smaller diameter microgels (<50 μm) would be preferable. We demonstrate protease-dependent degradation of these microgels to release the incorporated IgG or VEGF. We also show control over the in vitro degradation rate of the microgels via tuning the ratio of protease-degradable (VPM) to non-degradable (DTT) crosslinker. In addition to the VPM peptide used in this study, other protease-cleavable peptides with faster or slower degradation rates could be employed as gel crosslinkers in order to more precisely control the degradation and protein release kinetics [19]. In this way, the release kinetics of proteins from the microgels are dependent on local cellular demand rather than release initiated by hydrolysis or other chemical stimuli. Microgels labeled with a near infrared dye were used to track microgel degradation in vivo. VPM-crosslinked microgels exhibited an exponential decay in fluorescence signal to low levels, which is attributed to microgel degradation. DTT-crosslinked microgels exhibited a small decrease in signal initially that stabilized and remained relatively constant for the duration of the observation period. Although we cannot rule out in vivo degradation of DTT-crosslinked microgels by collagenases or other mechanisms, the signal loss for these microgels could be related to release of unincorporated dye or oxidation of the fluorochrome. A key advantage of the 4-arm PEG-maleimide material is the ability to easily conjugate molecules presenting free thiols for efficient tethering of biomolecules within the microgel. We show that ~80% of fluorescent IgG or VEGF was maintained within the microgels in the absence of collagenase, with the remaining 20% of untethered proteins released within 4 hours. Importantly, the bioactivity of the VEGF tethered within the microgels remained similar to non-PEGylated VEGF. In vivo experiments also suggest VEGF remains bioactive over the course of the study. We posit that cell-controlled degradation of the microgels and sustained release of VEGF improves vascularization compared to materials that do not allow cellular based remodeling. Non-degradable microgels containing VEGF exhibited vascularization around the perimeter of the injection site, however, vessels were unable to penetrate within areas occupied by the microgels. Furthermore, without the sustained release of VEGF, vessels did not form even in the presence of a degradable microgel scaffold. 5. Conclusion Synthetic microgels offer significant advantages as protein delivery vehicles, including minimally invasive, injectable delivery, control of protein release rates, control of particle size, and the ability to deliver multiple proteins with independently controllable release rates simultaneously. We demonstrate that protease-degradable microgels promote in vivo vascularization by controlled release of the vasculogenic protein VEGF and complete degradation of microgels that allows for tissue ingrowth and remodeling. Supplementary Material 1 2 3 Acknowledgements This work was funded by the U.S. National Institutes of Health (R21 EB020107) and the Juvenile Diabetes Research Foundation (2-SRA-2014-287-Q-R). C.G.G. was funded by the IOF-Marie Curie Post-Doctoral Fellowship (331655). M.S.S. is funded by ERC (306990). Fig. 1 Generation of protease degradable microgels using flow focusing microfluidics (a) Image of microfludic flow focusing device with 200 μm nozzle. (b) Image of microgels generated using a 200 μm (left) or 50 μm (right) nozzle. (c) Coefficient of variation of diameter for microgels generated from 200 μm or 50 μm nozzles. (d) Images of microgels crosslinked with DTT or VPM in the presence of collagenase or PBS. (e) Percent of DTT or VPM crosslinked microgels remaining after 20 hour incubation with type 1 collagenase or PBS (n = 3 independent experiments). Significance compared to PBS control was determined using two-way ANOVA with Dunnett’s post-test, p<0.01, p<0.005. Fig. 2 Controlled degradation of microgels and release of protein Release kinetics of fluorescent IgG from microgels crosslinked with different molar ratios of VPM to DTT. (a) Microgels were treated with type 1 collagenase (3.9 units/mL) or (b) PBS (n = 3 independent experiments). Fig. 3 Release kinetics and bioactivity of VEGF Prior to formation into a gel, PEG-4MAL was functionalized with fluorescently labeled VEGF (10 μg/mL). (a) Representative image of fluorescent VEGF bound within the microgels. (b) Percent of VEGF released over time in the presence of type 1 collagenase (3.9 units/mL) or PBS (n = 4 independent experiments). (c) Endothelial cell metabolic assay for soluble VEGF (100 ng/mL) or VEGF released from protease-degradable microgels treated with MMP-2 (50 nM) and TIMP-1 (50nM) (n = 3 independent experiments). A total dose of 15 ng was used for both soluble and released VEGF conditions. Fig. 4 Protease degradable VEGF microgels promote vascularization in vivo Microgels functionalized with a fluorescent RGD molecule were injected into the subcutaneous space on the back of C57BL/6 mice. (a) Images of microgel fluorescence at day 0 and day 14. (b) Percent of signal remaining over time compared to day 0. Data was fit with a simply decay model. (c) Representative fluorescent images of skin explants perfused with lectin to label vasculature. Fluorescent images of microgels in skin explants (bottom left). Dotted white line indicates the area represented in the large image. (d) Number of lectin labeled vessels per field of view (n = 4 mice per group). Significance was determined using one-way ANOVA with Tukey post-test, **p<0.005. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. References [1] Allazetta S Hausherr TC Lutolf MP Microfluidic Synthesis of Cell-Type-Specific Artificial Extracellular Matrix Hydrogels Biomacromolecules 2013 14 1122 1131 doi:10.1021/bm4000162 23439131 [2] Siltanen C Yaghoobi M Haque A You J Lowen J Soleimani M Microfluidic fabrication of bioactive microgels for rapid formation and enhanced differentiation of stem cell spheroids Acta Biomaterialia 2016 34 125 132 doi:10.1016/j.actbio.2016.01.012 26774761 [3] Chan HF Zhang Y Leong KW Efficient One-Step Production of Microencapsulated Hepatocyte Spheroids with Enhanced Functions Small 2016 12 2720 2730 doi:10.1002/smll.201502932 27038291 [4] Cha C Oh J Kim K Qiu Y Joh M Shin SR Microfluidics-Assisted Fabrication of Gelatin-Silica Core–Shell Microgels for Injectable Tissue Constructs Biomacromolecules 2014 15 283 290 doi:10.1021/bm401533y 24344625 [5] Rossow T Heyman JA Ehrlicher AJ Langhoff A Weitz DA Haag R Controlled Synthesis of Cell-Laden Microgels by Radical-Free Gelation in Droplet Microfluidics J. Am. Chem. Soc 2012 134 4983 4989 doi:10.1021/ja300460p 22356466 [6] Headen DM Aubry G Lu H Garcia AJ Microfluidic-Based Generation of Size-Controlled, Biofunctionalized Synthetic Polymer Microgels for Cell Encapsulation Adv. Mater 2014 26 3003 3008 doi:10.1002/adma.201304880 24615922 [7] Liu AL Garcia AJ Methods for Generating Hydrogel Particles for Protein Delivery Ann Biomed Eng 2016 44 1946 1958 doi:10.1007/s10439-016-1637-z 27160672 [8] Vegas AJ Veiseh O rtler MGU Millman JR Pagliuca FW Bader AR Long-term glycemic control using polymer-encapsulated human stem cell–derived beta cells in immune-competent mice Nature Medicine 2016 1 10 doi:10.1038/nm.4030 [9] Griffin DR Weaver WM Scumpia PO Di Carlo D Segura T Accelerated wound healing by injectable microporous gel scaffolds assembled from annealed building blocks Nat Mater 2015 14 737 744 doi:10.1038/nmat4294 26030305 [10] Zhang QM Wang W Su Y-Q Hensen EJM Serpe MJ Biological Imaging and Sensing with Multiresponsive Microgels Chem. Mater 2016 28 259 265 doi:10.1021/acs.chemmater.5b04028 [11] Zhang X Lü S Gao C Chen C Zhang X Liu M Highly stable and degradable multifunctional microgel for self-regulated insulin delivery under physiological conditions Nanoscale 2013 5 6498 9 doi:10.1039/c3nr00835e 23752741 [12] Lai W-F Susha AS Rogach AL Multicompartment Microgel Beads for Co-Delivery of Multiple Drugs at Individual Release Rates ACS Appl. Mater. Interfaces 2016 8 871 880 doi:10.1021/acsami.5b10274 26720613 [13] Stukel J Thompson S Simon L Willits R Polyethlyene glycol microgels to deliver bioactive nerve growth factor J. Biomed. Mater. Res 2014 103 604 613 doi:10.1002/jbm.a.35209 [14] Tibbitt MW Han BW Kloxin AM Anseth KS Student award for outstanding research winner in the Ph.D. category for the 9th World Biomaterials Congress, Chengdu, China, June 1-5, 2012 Synthesis and application of photodegradable microspheres for spatiotemporal control of protein delivery J. Biomed. Mater. Res 2012 100A 1647 1654 doi:10.1002/jbm.a.34107 [15] Sivakumaran D Maitland D Hoare T Injectable microgel-hydrogel composites for prolonged small-molecule drug delivery Biomacromolecules 2011 12 4112 4120 doi:10.1021/bm201170h 22007750 [16] Wang K Lin S Nune KC Misra RDK Chitosan-gelatin-based microgel for sustained drug delivery Journal of Biomaterials Science, Polymer Edition 2016 27 441 453 doi:10.1080/09205063.2016.1143673 26775820 [17] Phelps EA Templeman KL Thulé PM Garcia AJ Engineered VEGF-releasing PEG-MAL hydrogel for pancreatic islet vascularization Drug Deliv Transl Res 2015 5 125 136 doi:10.1007/s13346-013-0142-2 25787738 [18] Phelps EA Headen DM Taylor WR Thulé PM Garcia AJ Vasculogenic bio-synthetic hydrogel for enhancement of pancreatic islet engraftment and function in type 1 diabetes Biomaterials 2013 34 4602 4611 doi:10.1016/j.biomaterials.2013.03.012 23541111 [19] Patterson J Hubbell JA Enhanced proteolytic degradation of molecularly engineered PEG hydrogels in response to MMP-1 and MMP-2 Biomaterials 2010 31 7836 7845 doi:10.1016/j.biomaterials.2010.06.061 20667588 [20] Briquez PS Clegg LE Martino MM Mac Gabhann F Hubbell JA Design principles for therapeutic angiogenic materials Nature Reviews Materials 2016 1 15006 doi:10.1038/natrevmats.2015.6 [21] Phelps EA Landázuri N Thulé PM Taylor WR Garcia AJ Bioartificial matrices for therapeutic vascularization Proc Natl Acad Sci U S A 2010 107 3323 3328 doi:10.1073/pnas.0905447107 20080569 [22] Seliktar D Zisch AH Lutolf MP Wrana JL Hubbell JA MMP-2 sensitive, VEGF-bearing bioactive hydrogels for promotion of vascular healing J. Biomed. Mater. Res 2004 68 704 716 doi:10.1002/jbm.a.20091 [23] Zhu S Segura T Cell-Demanded VEGF Release via Nanocapsules Elicits Different Receptor Activation Dynamics and Enhanced Angiogenesis Ann Biomed Eng 2016 44 1983 1992 doi:10.1007/s10439-016-1581-y 26940611 [24] Hoare TR Kohane DS Hydrogels in drug delivery: Progress and challenges Polymers with Aligned Carbon Nanotubes: Active Composite Materials 2008 49 1993 2007 doi:10.1016/j.polymer.2008.01.027
PMC005xxxxxx/PMC5121020.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 8100316 1138 Biomaterials Biomaterials Biomaterials 0142-9612 1878-5905 27770633 5121020 10.1016/j.biomaterials.2016.10.013 NIHMS824533 Article POLYMER FIBER-BASED MODELS OF CONNECTIVE TISSUE REPAIR AND HEALING Lee Nancy M. 1 Erisken Cevat 1 Iskratsch Thomas 2 Sheetz Michael 2 Levine William N. 3 Lu Helen H. 14 1 Biomaterials and Interface Tissue Engineering Laboratory, Department of Biomedical Engineering, Columbia University, New York, NY 10027 2 Department of Biological Sciences, Columbia University, 713 Fairchild Center, MC 2408, New York, NY 10027 3 Department of Orthopaedic Surgery, Columbia University Medical Center, New York, NY 10032 4 To whom all correspondence should be addressed: Helen H. Lu, Ph.D., Department of Biomedical Engineering, Columbia University, 351 Engineering Terrace Building, MC 8904, 1210 Amsterdam Avenue, New York, NY 10027, 212-854-4071 (office); 212-854-8725 (fax), [email protected] 23 10 2016 12 10 2016 1 2017 01 1 2018 112 303312 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Physiologically relevant models of wound healing are essential for understanding the biology of connective tissue repair and healing. They can also be used to identify key cellular processes and matrix characteristics essential for the design of soft tissue grafts. Modeling the various stages of repair post tendon injury, polymer meshes of varying fiber diameter (nano-1 (390 nm) < nano-2 (740 nm) < micro (1420 nm)) were produced. Alignment was also introduced in the nano-2 group to model matrix undergoing biological healing rather than scar formation. The response of human tendon fibroblasts on these model substrates were evaluated over time as a function of fiber diameter and alignment. It was observed that the repair models of unaligned nanoscale fibers enhanced cell growth and collagen synthesis, while these outcomes were significantly reduced in the mature repair model consisting of unaligned micron-sized fibers. Organization of paxillin and actin on unaligned meshes was enhanced on micro-compared to nano-sized fibers, while the expression and activity of RhoA and Rac1 were greater on nanofibers. In contrast, aligned nanofibers promoted early cell organization, while reducing excessive cell growth and collagen production in the long term. These results show that the early-stage repair model of unaligned nanoscale fibers elicits a response characteristic of the proliferative phase of wound repair, while the more mature model consisting of unaligned micron-sized fibers is more representative of the remodeling phase by supporting cell organization while suppressing growth and biosynthesis. Interestingly, introduction of fiber alignment in the nanofiber model alters fibroblast response from repair to healing, implicating matrix alignment as a critical design factor for circumventing scar formation and promoting biological healing of soft tissue injuries. wound repair model fiber diameter alignment tendon Introduction The repair of injured connective tissues such as ligaments or tendons begins with an inflammatory response followed by a proliferative phase, in which the predominate cell types are fibroblasts, macrophages, and mast cells. A disorganized collagen fiber matrix is first laid down[1], and the collagen fibers deposited during this period display a fiber diameter smaller than that of native tissue[2,3] (Figure 1: Stage 1 and 2). The next phase of wound repair is the remodeling phase, which is characterized by both increasing collagen fiber diameter and organization (Figure 1: Stage 3 and Repair), augmenting the tensile strength of the repairing tissue[3,4]. As a consequence of how both collagen fiber diameter and organization during wound repair deviate from those of the healthy extracellular matrix (ECM), regeneration or true biological healing is not achieved and the repaired tissue is often compositionally and mechanically inferior to native matrix[1,5,6]. This results in suboptimal performance of the repaired tissue[7] and increases the potential for reinjury. The disparity that exists between normal and scar/repaired tissue underscores the need for physiologically relevant matrix models, which can be used to investigate the cellular and matrix processes that differ between wound repair and biological healing. Such models will also be critical for the design of functional grafts to enhance connective tissue regeneration. It has been reported that the response of a variety of cell types including fibroblasts[8-10], osteoblasts[11] and mesenchymal stem cells (MSCs)[12,13] can be modulated by the underlying substrate topography, namely fiber diameter and alignment. Specifically, fibroblast organization[8], matrix production[9], and migration[10] were enhanced by culture on aligned compared to unaligned nanofiber matrices. Investigating the effects of fiber diameter using unaligned polylactide-co-glycolide (PLGA) fibers, Bashur et al. observed diminished spreading and lower aspect ratio for NIH 3T3 mouse fibroblasts cultured on 140 nm and 760 nm fibers compared to 3.6 μm fibers[14]. Stem cells grown on unaligned polyurethane meshes measured greater cell density on submicron (0.28, 0.82 μm) meshes compared to 2.3 μm fibers[15]. More recently, Erisken et al. observed a diameter-dependent response on tendon fibroblasts cultured on aligned PLGA meshes, whereby nanofibers regulated cell proliferation and production, while the microfibers upregulated the expression of fibroblastic markers[16]. These findings underscore the complex relationship between the sub-cellular geometry and global cell function. Cells respond to their 3D environment through cytoskeletal actin stress fibers[17] and focal adhesions which activate downstream signaling pathways[18]. This outside-in signaling is known to affect critical cell functions including adhesion, migration, morphology, proliferation, gene expression, and differentiation[18-22]. Accordingly, interactions between cells and their local environment play a significant role in both tissue homeostasis and wound repair[23]. As wounds are heterogeneous and dynamic, the repair process is complex and influenced by factors such as extracellular matrix components and growth factors[24,25], that can lead to scar tissue formation instead of tissue regeneration or biological healing. Model systems that recapitulate the dynamic characteristic of the extracellular matrix during repair or healing can be used to investigate the biological mechanisms guiding these distinct processes. Such in vitro wound repair models will allow for the controlled examination of the tissue response to the environmental factor or change, isolating the effects from other cells or tissues inherent in in vivo models[26]. Furthermore, comparisons between models of tissue repair for which the outcome is scar tissue, and models of tissue regeneration or biological healing, which results in recapitulation of native tissue architecture and function, will be instrumental to the design of biomaterials and grafts that will promote tissue regeneration. To this end, the focus of this study is to develop polymer fiber-based models of tendon repair and healing and to examine the response of human tendon fibroblasts to these substrates. Specifically in this work, tissue repair refers to the events characteristic of the formation of scar or physiologically inferior tissue, while healing or regeneration refers to the restoration of native tissue properties. Given that collagen fiber diameter increases during tissue healing[2-4], it is hypothesized that nanometer diameter fibers will promote cell proliferation and matrix deposition, similar to the proliferative phase of wound repair, modeled by Stages 1 and 2 in Figure 1. This is in comparison to the more mature stage of repair, as represented by Stage 3. Furthermore, as it is established that collagen fibers become more aligned from an initial unorganized tissue during wound repair[3,4], cell response elicited by nanofiber organization is also examined here. By keeping fiber diameter constant, these groups will allow for the decoupling of effects based on fiber alignment and diameter. It is anticipated that cell attachment and spreading, as well as the expression of fibroblastic markers will be promoted on aligned fibers, while unaligned fibers will enhance proliferation and matrix deposition. The elucidation of these wound repair and regeneration interactions is anticipated to reveal parameters critical to the design of functional scaffolds that will promote biological healing instead of scar tissue formation post injury. Materials and Methods Mesh Fabrication and Characterization Unaligned polylactic-co-glycolic acid (PLGA) meshes of three fiber diameter ranges were fabricated by electrospinning[25;26] according to conditions outlined in Table 1. Briefly, granules of PLGA (85:15, DL, Mw ≈ 123.6 kDa, Lakshore Biomaterials, Birmingham, AL) were premixed in a solution of acetone (Ace, Sigma-Aldrich, St. Louis, MO) and N,N-dimethylformamide (DMF, Sigma-Aldrich), followed by the addition of 14% v/v ethanol (Decon, King of Prussia, PA). The polymer solution (32%, 43% and 50% w/v) was loaded into a 5-mL syringe attached with a blunt-tip needle, and dispensed using a syringe pump (Harvard Apparatus, Holliston, MA). A voltage of 8-10 kV was applied to the needle, and fibers were collected on a grounded stationary plate placed at a distance of 13-15 cm from the needle tip. To fabricate aligned nanofibers, PLGA was premixed with DMF, followed by the addition of 0.5 mL ethanol, yielding a 54% w/v solution. The polymer solution was loaded into a 5-mL syringe with an 18.5 gauge blunt-tip needle, and electrospun at a voltage of 8-10 kV and a flow rate of 1 mL/hr. Fibers were collected on a grounded rotating mandrel (2500 rpm) placed at a 12 cm distance from the needle tip. All polymer solutions were electrospun to achieve a thickness of 0.10 ± 0.02 mm. The fiber diameter, alignment, and morphology of as-fabricated matrices were analyzed by scanning electron microscopy (SEM, 1-2.5 kV, 10 μA Hitachi 4700, Tokyo, Japan). Prior to SEM imaging, samples were coated (Cressington 108auto, Watford, England) with a layer of gold-palladium (10 seconds, 2 nm). Fiber diameter, measured using ImageJ (National Institutes of Health, Bethesda MD), is reported as an average of measurements taken from three independent regions (n=6 images/group), with 20 measurements (10-15 fibers) per image. Fiber alignment (n=6) was quantified using the method of Costa et al.[27], in which SEM images were analyzed using circular statistics software customized for evaluating fiber alignment (Fiber3). Parameters include: (1) the mean vector angle (MA), which represents the average fiber alignment in the matrix (\|θ\| ≤ 90°), with 0° representing a horizontal orientatio n; (2) the mean vector length (MVL, 0 ≤ r ≤ 1), in which 0 indicates a random and 1 an aligned morphology; and (3) angular deviation (AD) which characterizes the dispersion of the non-Gaussian angle distribution of the nanofibers (0° ≤ θ ≤ 40.5°), whereby 0° represents aligned and 40.5° a random distribution. Cells and Cell Culture Human rotator cuff fibroblasts were isolated from explant culture of tissues obtained from patients (male aged 51- 65 years, female 76 years, institutional review board exempt) who underwent rotator cuff repair surgery. Briefly, tissue samples were rinsed with phosphate buffered saline (PBS, Sigma-Aldrich), plated in tissue culture dishes, and maintained in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS, Atlanta Biologicals, Lawrenceville, GA), 1% non-essential amino acids, 1% penicillin/streptomycin, 0.1% amphotericin B and 0.1% gentamicin/sulfate (fully supplemented media). The cells from the first and second migrations were discarded, and the tissue was re-plated in fresh media. Only cells obtained from the third migration were used in this study as this method has been shown to yield a relatively homogenous fibroblast population[28]. Prior to seeding, cells from all donors were pooled. All media and supplements were purchased from Mediatech (Herndon, VA) unless otherwise noted. Cell Seeding on Scaffolds Prior to seeding, the polymer scaffolds were secured using custom-made clamps, sterilized with UV radiation (356nm), and preincubated (37°C, 5% CO2) in DMEM supplemented with 20% FBS overnight to promote cell attachment. Fibroblasts (passage 4-5) were seeded at density of 30,000 cells/cm2, and were cultured for up to four weeks in fully supplemented DMEM containing 50 μg/mL ascorbic acid[29] (Sigma-Aldrich). Cells were maintained at 37°C, 5% CO 2 and media was exchanged every 2-3 days. The effect of fiber diameter and alignment on cell morphology, attachment, proliferation, gene expression and matrix production were determined over four weeks of culture. Cell Viability and Adhesion Cell viability (n=3) was visualized using Live/Dead staining (Molecular Probes, Eugene, OR) following the manufacturer's protocol. Samples were imaged using confocal microscopy (Olympus Fluoview FV100, Center Valley, PA) at excitation wavelengths of 488 nm and 568 nm. For alignment analysis, cell viability images (n=3, two regions/sample) were evaluated using circular statistics software, as described in the procedure for evaluating fiber alignment. Immunohistochemistry (n=3) was performed to visualize paxillin focal adhesions and the actin cytoskeleton. Briefly, samples were rinsed in PBS, fixed for 10 minutes at room temperature in a 4% paraformaldehyde solution (pH 7.2), followed by a 10 minute permeabilization in 0.1% Triton-X (Sigma-Aldrich). Blocking was performed by immersion in a 1% BSA and 5% goat serum (Jackson Laboratories, Bar Harbor, ME) in immunostaining buffer solution (20mM TRIS, 155mM NaCl, 2nM EGTA, 2mM MgCl2, pH 7.4) for 1 hour. Samples were incubated with paxillin primary antibody (1:500 dilution, BD Biosciences, Franklin Lakes, NJ) for 1 hour at room temperature, and then with secondary antibody (Alexa Fluor 555, 1:300 dilution, Invitrogen), along with Alexa Fluor 647 conjugated phallodin (1:100 dilution, Invitrogen) for 1 hour at room temperature. The samples were subsequently incubated with DAPI nuclear counterstain for 30 minutes, then mounted and imaged with confocal microscopy. All antibody dilutions were performed in immunostaining buffer containing 1% BSA. Quantitative analysis of Rac1 activation (n=6) was performed using the G-LISA Rac activation assay (Cytoskeleton, Denver CO), according to the manufacturer's specifications. Briefly, after one day of culture scaffolds were rinsed with PBS and 100μl of lysis buffer was added to the samples. The lysates were collected, snap frozen, and stored at −80°C until analysis. Rac activity was d etermined using a 1:250 dilution of the primary antibody and 1:200 dilution of the secondary antibody. Horseradish peroxidase reagent was used to detect samples and luminescence was quantified (Tecan, Research Triangle Park, NC) using an integration time of 10 ms and gain of 100. Results were normalized to the total cell number. Cell Proliferation Total DNA content (n=5) was measured using the PicoGreen double stranded DNA assay (Invitrogen, Carlsbad, CA). At each time point, the samples were homogenized in 0.1% Triton-X and subjected to ultrsonication at 5W (Microson XL-2000, Qsonica, Newton, CT) for 15 seconds. Fluorescence was measured with a Tecan microplate reader at excitation and emission wavelengths of 485 and 535 nm respectively. Cell number was determined using a conversion factor of 8pg DNA/cell[30]. Cell Matrix Production Collagen production (n=5) was quantified with a modified hydroxyproline assay[30]. Briefly, the samples were first desiccated (CentriVap, Labconco, Kansas City, MO) and digested for 18 hours at 65°C in a solution of 20 μL/mL papain (Sigma-Aldrich), buffered in 0.1 M sodium acetate, 10 mM cysteine HCl, and 50 M ethylenediaminetetraacetate. A 250 μl aliquot of the digest was concentrated by desiccation and samples were subsequently hydrolyzed in 10μl of 10N NaOH and autoclaved for 25 min. The hydrolysate was then oxidized by a buffered chloramine-T reagent for 25 min before the addition of Ehrlich's reagent. Sample absorbance was measured at 550 nm (Tecan), and collagen content was obtained by interpolation along a standard curve generated using bovine collagen I (Sigma-Aldrich). Matrix and cell distribution (n=2) was also visualized using hematoxylin and eosin (H&E, Sigma-Aldrich) and Picrosirius red stains[31]. Briefly, samples were fixed overnight in a 10% neutral buffered formalin containing 1% cetylpyridinium chloride (Sigma-Aldrich) solution, and subsequently stored in 0.01 M cacodylic acid (Sigma-Aldrich) until sectioning. Scaffolds were embedded in 5% poly vinyl alcohol (Sigma-Aldrich) frozen sectioning medium, and cut in cross-section to 10 μm thickness using a cryostat (Bright Instrument Company, Cambridgeshire, England). Gene Expression The expression of the fibroblastic markers type I and III collagen, intergrins α2 and β1, and the Rho GTPases RhoA, Rac1, and Cdc42 were measured at days 7 and 14 using quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR, n=5), with custom-design primers (Table 2). Total RNA was isolated via Trizol extraction (Invitrogen), and then reverse-transcribed into cDNA using the SuperScript III First-Strand Synthesis System (Invitrogen). The cDNA product was amplified and quantified through real-time PCR using SYBR Green Supermix (Invitrogen). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) served as the house-keeping gene. All reactions were run for 50 cycles using the iCycler iQ Real-Time PCR Detection System (BioRad, Hercules, CA). Normalized expression levels were calculated based on the difference between threshold cycles of the gene of interest and GAPDH. Statistical Analyses Results are reported as mean ± standard deviation, with n equal to the number of replicates per group. One-way analysis of variance (ANOVA) was performed to determine the effects of fiber diameter or organization on scaffold structural and mechanical properties. Two-way ANOVA was used to determine fiber diameter and/or organization and temporal effects on cell proliferation, matrix production and gene expression. The Tukey-Kramer post hoc test was used for all pair-wise comparisons, and significance was attained at p < 0.05. Statistical analyses were performed with JMPIN (4.0.4, SAS Institute, Inc., Cary, NC). Results Polymeric Mesh Characterization Electrospinning conditions were optimized to achieve fiber diameters which averaged 390 ±140 nm (Nano-1), 740 ± 160 nm (Nano-2), and 1420 ± 370 nm (Micro) for unaligned fiber matrices and 650 ± 170nm for the aligned matrix (Figure 1 and Table 1). Fiber diameters of the unaligned matrices were statistically different from each other (p<0.05). In contrast, no difference in fiber diameter was found between the aligned and unaligned Nano-2 meshes. Alignment analysis results [mean vector angle (MA), mean vector length (MVL) and angular deviation (AD)] are summarized in Table 1. While the MVL and AD of the Nano-2 group differed from the other unaligned groups, overall the values for these parameters were indicative of a disorganized fiber morphology. Analysis confirmed the organization of nanofibers in the Nano-2 aligned group, and as expected, all unaligned groups were significantly different from the aligned group in MA, MVL and AD. Effect of diameter on cell attachment and organization Cells remained viable on all scaffolds tested over the four-week culture period. Focal adhesions (pink) and actin filaments (green) were visualized using immunohistochemistry. Cell adhesion and spreading were enhanced as fiber diameter increases (Figure 2). Cell polarization was evident after one hour of culture in the Micro group, while cells on nanofiber matrices still appeared rounded. Furthermore, after one day of culture, cells in the Micro group appeared more elongated and organized relative to the nanofiber scaffolds (Nano-1 and Nano-2) despite the fact that all three substrates are unaligned. Differences in cell alignment were also observed during the first two weeks of culture, and in general, cells became more organized over time. Specifically, at day 1, the MA of the Nano-1 and Nano-2 groups was significantly larger than the Micro group, with a lower MA indicative of an organized orientation. At days 7 and 14, cell alignment parameters for Nano-2 (MA, AD) and Nano-1 (MVL, AD) groups continued to indicate a greater degree of randomly oriented cells compared to the Micro group. Over time a significant increase in MVL and concurrent decrease in MA and AD were observed for all three unaligned groups, and by day 28 the only difference between groups was in AD between Nano-1 and Micro. Effect of fiber diameter on cell growth and phenotype Fibroblasts proliferated in all groups through the four-week culture period. While a temporal increase in cell number was observed in all groups (Figure 3), at day 28, cell number on the nanofiber matrices were greater than that on the Micro group (p<0.05). Hematoxylin staining revealed that cell distribution was uniform through the depth of the scaffold for all groups. A significant increase in collagen production over time was observed for the Nano-1 and Nano-2 groups (Figure 3). Furthermore at day 28, collagen per scaffold was significantly greater for the Nano-2 group compared to the Micro group. Picrosirius red staining at day 28 (Figure 3) confirms the synthesis of a collagen-rich matrix. A more intense staining was observed in the Nano-1 and Nano-2 groups compared to the Micro group. The expression of collagens I and III, and integrins α2 and β1 were examined at days 7 and 14. After one week of culture, an upregulation of collagens I and III was observed in the Micro group compared to both nanofiber matrices. There was a significant decrease in collagen I and III expression for both Nano-1 and Micro groups over time. The ratio of collagen III/I expression increased over time for the Nano-1 group. At day 7, the expression of α2 was significantly downregulated on the nanofiber matrices compared to the Micro group, while β1 expression was upregulated in Nano-2 compared to Nano-1. The expression of integrins increased significantly from day 7 to 14 for all groups. Effect of fiber alignment on cell attachment and organization Cell adhesion was promoted in the Aligned group as indicated by more intense positive staining for paxillin (Figure 4). This is particularly evident at 30 minutes and 1 hour post-seeding. After 1 hour, an aligned actin fiber organization was evident on the aligned matrix. In contrast, no such organization of cell adhesions or actin fibers was noted for the unaligned Nano-2 group. Cell alignment analysis revealed a significantly larger MA and AD, and smaller MVL on unaligned fibers during the first two weeks of culture. With time, cells on the unaligned fibers became significantly more aligned, and by day 28, there were no significant differences in alignment parameters between the Aligned and unaligned Nano-2 groups. Effect of fiber alignment on cell growth and phenotype Cell proliferation was noted over time (Figure 5) and at day 28, the cell number was significantly greater on the unaligned compared to the Aligned Nano-2 mesh. Cell penetration, as observed with hematoxylin staining, was more extensive on unaligned fibers compared to aligned. Collagen deposition increased with culture time. Hematoxylin & Eosin as well as Picrosirius red staining at day 28 confirms a greater degree of matrix deposition on unaligned fibers. At day 14, the expression of collagen III was significantly upregulated on unaligned nanofibers (Figure 5). Collagen I expression decreased over time for both groups, while collagen III decreased between day 7 and 14 for the aligned group. Effect of fiber diameter and alignment on Rho GTPase expression and activity After 1 day of culture, RhoA was upregulated on the Nano-1 group compared to all other unaligned matrices (p<0.05), and Rac1 was upregulated on Nano-1 compared to Nano-2 (p<0.05), while no difference in neither Cdc42 expression nor Rac1 activity as a function of diameter was observed (Figure 6). Comparing results based on fiber alignment, Cdc42 was upregulated on aligned nanofibers, while active Rac1 was significantly lower than that of the unaligned Nano-2 group. Discussion Models to study wound repair and healing are important for investigating the biology of tissue repair and regeneration, yielding critical insights that can guide the design and evaluation of therapies that will support and promote tissue regeneration. This study investigates the response of human tendon fibroblasts as a function of matrix fiber diameter and alignment, using PLGA fiber mesh models designed to represent the various stages of tendon repair post injury (Figure 1). PLGA is used as a prototypical polymer as it is a well-studied biodegradable material with highly tunable properties. Additionally, as the polymer degrades via bulk erosion, mesh fiber diameter and architecture are maintained, and mechanical properties remain within the range of the native tendon after four weeks of in vitro culture[8]. Observations from this study indicate that both matrix fiber diameter and alignment regulate fibroblast adhesion, spreading, and organization. In turn, downstream response, including proliferation, gene expression, and matrix deposition are also modulated by these matrix parameters. A distinct cell response is observed with changes in model fiber diameter and alignment, with matrix alignment as an important design factor for biological healing of soft tissue injuries. The unaligned mesh models used in this study are designed to simulate a wound environment, with a disorganized fiber matrix but increasing fiber diameter as wound healing progresses. It is observed here that culture on substrates mimetic of early injury states stimulates cells to proliferate and deposit matrix, which is characteristic of the initial fibroblast response displayed during wound healing[1]. In contrast, fibers with a larger, micron-sized, diameter, representing a more mature wound environment, promoted better cell organization, which in turn facilitates wound closure. Additionally, upregulation of integrin expression observed on the microfibers is suggestive of matrix remodeling or contraction[32], characteristic of the later stages of wound repair[3,4]. As such, these disorganized matrices fail to support the maintenance of the fibroblast phenotype and better represents the scar tissue. In contrast, an aligned morphology which more closely resemble the matrix of healthy tendons during development, inherently promotes cell organization, spreading, and maintenance of the fibroblastic phenotype. At the cellular level, the results of this study are suggestive of the link between sub-cellular geometry and global cell function. Cells respond to their environment through cytoskeletal components that activate downstream signaling cascades[33]. The interactions between cells and their substrata are mediated by cytoskeletal components such as paxillin, an adapter protein which binds a variety of signaling molecules, and is localized to focal adhesions[34], and actin cytoskeletal fibers. Both focal adhesion formation as well as cytoskeletal organization is enhanced by culture on unaligned micron-sized fibers, where the tendon fibroblasts appear to be more elongated and organized relative to those on the unaligned nanofiber models. Moreover, paxillin is known to affect the activity of molecules such as the Rho family of GTPases[32,35], which in turn, control a variety of signal transducing pathways regulating responses such as cytoskeletal reorganization, transcription, and cell migration[36]. To further explore this expression of RhoA, Rac1, and, Cdc42, proteins from the Rho family of GTPases were evaluated. As GTPases alternate between active (GTP-bound) and inactive (GDP-bound) states, the amount of active Rac1 produced by the cells was also assessed. Critical to the tissue repair process is cell migration[32]. Rac and Cdc42 stimulate formation of lamellipodia and filopodia[37,38] respectively, at the front of migrating cells. Furthermore, Cdc42 mediates cell polarization, which is also required for movement[32,39]. Rho acts at the rear of the cells generating contractile forces[32,37], while also playing a role in maintaining cell adhesion[40]. Thus the cooperative action of these proteins promotes cell motility. Nur-E-Kamal et al. reported that unaligned Ultra-Web fibers with an average diameter of 180 nm promotes the activation of Rac over 2D substrates[41]. Jaiswal et al. observed that the fiber diameter of poly methyl methacrylate unaligned fibers can modulate the activation patterns mitogen-activated protein kinases (MAPKs)[42]. Results indicated that as fiber diameter decreases from 2.41 μm to 882 nm p38 activity increases, while a further decrease in diameter to 605 nm caused a decrease in activity. A similar but opposite trend was noted for p38 activity, highlight the varying effect that fiber diameter can have on cell phenotype response. The observations of this study, where the expression of both RhoA and Rac1 are upregulated on the unaligned nanofibers, coupled with enhanced activity of Rac1 by fibroblasts on unaligned compared to aligned nanofibers, are indicative of enhanced cell motility on the unaligned nanofibers. As greater cell motility is commonly associated with wound contraction and tissue repair[32], cells on the unaligned nanofibers are likely progressing toward scar tissue formation rather than biological healing. Given that fiber diameter of the Nano-2 and Nano-2 aligned group are statistically similar, it is interesting that by altering fiber alignment, active Rac 1 level is reduced while the cells become more polarized, suggesting that fibroblast response on the nanofiber substrates maybe be changed from repair to healing by controlling matrix alignment. Given that the organization of focal adhesions and actin fibers were enhanced on aligned fibers compared to unaligned immediately post seeding, it is not surprising that downstream cell responses are in turn affected, with both much lower cell growth and collagen deposition observed on aligned nanofibers, indicative of a normal healing response instead of the commonly observed proliferative phase of wound repair. The observed response of fibroblast on the aligned nanofiber also confirms the findings of Erisken et al., which explored human tendon fibroblast response to aligned PLGA nanofibers and microfibers[16]. Interestingly, once alignment is maintained on the fiber mesh, increasing the fiber diameter from nano- to micron-size further reduces cell proliferation and biosynthesis, and most importantly, promotes the expression of fibroblastic markers such as tenomodulin[16]. Collectively, these observations suggest that compared to fiber diameter, fiber alignment is a critical early matrix characteristic essential in directing tendon fibroblasts towards a physiologically relevant adhesion and organization. Thereafter, any increase in matrix fiber diameter will promote the maintenance of the tendon fibroblast phenotype, with alignment and fiber diameter collectively driving the cells towards biological healing. These observations suggest that aligned nanofiber meshes rather than microfibers are an optimal matrix for tendon regeneration. Aligned nanofibers promote a balance between cell growth and biosynthesis required for the remodeling of a biodegradable polymer matrix, and the need to direct cell adhesion and organization early in the healing process. It is likely the neotissue with physiological alignment will mature and increase in fiber diameter, which will ensure the maintenance of the fibroblast phenotype and progress towards biological healing. Collectively, the results of this study demonstrate the promise that synthetic fiber substrates have for modeling cell response to tendon injury and healing. While the unorganized fiber matrices of varied diameters can be used as a model of the different stages of tissue injury, aligned fibers can also be used to study regenerated or healthy tissue states. It is noted that the model systems presented here focus on fiber diameter and alignment. Many other factors can contribute to the complex cellular events and cell-matrix interactions which lead to scar-less healing. For example, the age of the patient that fibroblasts are harvested from can affect cell response to substrates. Specifically, no differences in cell proliferation were observed based on fiber alignment in a study which explored ACL fibroblast, which are typically harvested from younger patients, response to polyurethane fibers 500-800 nm[9] in diameter. Additionally, while an increase in col III/I ratio is often associated with scar tissue formation[43], it was observed here that the expression of col I and III were upregulated on the unaligned microfiber group, but there were no differences in col III/I ratio based on fiber diameter. It is likely that other parameters, either matrix or cell-related, play a more important role in driving the collagen type ratio. Another limitation of these models is the lack of biological factors and immune cells typically involved in wound healing. As such, the relative simplicity of the fiber-based models allows the distillation of key matrix parameter, and their interplay in directing fibroblast response. Future studies will focus on enhancing the physiological relevance of the fiber-based model, further exploration of downstream effects as well as in vivo validation of these systems. Conclusion Modeling the events that occur during tissue repair and healing, this study explored the effect of fiber diameter of unaligned meshes and fiber organization on human tendon fibroblast response. The tissue repair model consisting of unaligned fibers with nanometer diameters promoted cell proliferation and matrix deposition synthesis as well as the expression and activity of RhoA and Rac1, characteristic of the initial, proliferative phase of wound repair. The mature repair model represented by unaligned micron-sized fibers supported cell organization and adhesion, while suppressing growth and biosynthesis, indicative of the remodeling phase of tissue repair. Finally, imparting alignment to nanofibers can guide cell response from repair to healing, potentially serving as a critical component to promoting the biological healing of soft tissues. Funding Sources: This study was supported by NIH-NIAMS (5R01-AR055280, AR056459 HHL), a CTICE graduate research fellowship (NML). Figure 1 The various stages of connective tissue repair are represented along with tissue healing are represented in the top panel. Scanning electron micrographs of the matrices used to model the corresponding stages of wound repair and healing are shown directly below. Figure 2 Effect of Fiber Diameter on Cell Attachment, Viability, and Alignment (A) Immunohistochemistry was used to visualize cytoskeletal components at 30 minutes, 1 hour, and overnight, in which cell adhesions, spreading, and elongation increases with increasing fiber diameter and over time. (B) Cell viability is maintained and proliferation is noted on all groups. (C) Cell alignment is greatest on the Micro matrix during the first two weeks, and increases with time for all groups. Significant difference between: # groups, * consecutive timepoints (n=5, p < 0.05). Figure 3 Effect of Fiber Diameter on Proliferation and Matrix Deposition (A) Significant cell proliferation is noted for all groups over time (cell number/wet weight, n=5). (B) Collagen on nanofiber matrices is significantly greater compared to the Micro group (collagen/wet weight and collagen/wet weight/cell, n=5). (C) Collagen is deposited predominately at the substrate surface, with greater staining on Nano-1 and Nano-2 groups (H&E and picrosirius red, n=2). (D) The expression of col I and III, ratio col III/I, and integrins α2 and β1 were evaluated after 7 and 14 days of culture. Col I and III are upregulated on the Micro matrix compared to both Nano groups on day 7. The expression of collagens decreases over time. Integrin α2 is upregulated on Micro compared to both Nano groups at day 7, while β1 expression is upregulated on Nano-2 compared to Nano-1 (n=5). Significant difference between: # groups, * consecutive timepoints (p < 0.05). Figure 4 Effect of Fiber Alignment on Cell Attachment, Viability, and Alignment (A) Immunohistochemistry was used to visualize cytoskeletal components at 30 minutes, 1 hour, and overnight, in which cell adhesions, spreading, and elongation is enhanced on aligned matrices at all timepoints. (B) Cell remain viable and proliferation is noted on both substrates. (C) Cells are significantly more organized on the aligned matrix (MA, MVL, AD) at Day 1, 7, 14, while cell alignment increases with time for the Nano-2, unaligned, group. Significant difference between: # groups, * consecutive timepoints (n=5, p < 0.05). Figure 5 Effect of Fiber Alignment on Proliferation and Matrix Deposition (A-B) Cell number and collagen is significantly greater on the unaligned matrices at Day 28 (cell number/wet weight, collagen/wet weight, collagen/wet weight/cell, n=5). (C) H&E histology shows greater cell penetration through the matrix depth on the unaligned group compared to the aligned (H&E and picrosirius red, n=2). (D) Gene expression of col I and III, ratio col III/I, and integrins α2 and β1 were evaluated after 7 and 14 days of culture. Integrin α2 expression is upregulated on the aligned matrix on day 7, and col III is upregulated on Nano-2 at day 14 (n=5). Significant difference between: # groups, * consecutive timepoints (p < 0.05). Figure 6 Rho GTPase Signaling Response to Matrices The expression of RhoA, Rac1, and Cdc42 were evaluated at day 1, based on fiber diameter and alignment (n=5). (A) Evaluating the effects of fiber diameter, RhoA expression is upregulated on Nano-1 compared to all other groups and Rac1 expression is upregulated on Nano-1 compared to Nano-2. (B) Based on fiber alignment, Cdc42 is significantly upregulated on aligned matrices. (C-D) Active Rac1 was quantified after one day of culture (n=5). There were no differences based on fiber diameter, while significantly greater levels of active Rac1 was detected on the Nano-2 group compared to aligned. Significant difference between: # groups, * consecutive timepoints (n=5, p < 0.05). Table 1 Electrospinning Conditions and Characterization of Fiber Alignment Group Fiber Diameter (nm) Polymer (%wt/vol) DMF:Ace Flow Rate (ml/hr) Needle Gauge Mean Vector Angle (n=6) Angular Deviation (n=6) Mean Vector Length (n=6) Nano-1 390 ± 140^ 32% 70:30 0.35 26.5 49.55 ± 7.89A 37.49 ± 0.50A,U 0.14 ± 0.023A,U Nano-2 740 ± 160* 43% 80:20 0.35 26.5 48.39 ± 2.06A 31.96 ± 1.04^ 0.38 ± 0.040^ Micro 1420 ± 370^ 50% 80:20 1 18 40.96 ± 16.80A 37.97 ± 1.76A,U 0.12 ± 0.08A,U Nano-2 Aligned 650 ± 170* 54% 100:0 1 18 0.53 ± 0.72^ 8.54 ± 0.84^ 0.96 ± 0.0089^ Fiber diameter values are an average of measurements taken from 3 independent regions imaged at 1 and 2.5K magnifications, with 20 measurements taken per image. No statistical differences in fiber diameter were noted between the Nano-2 (unaligned) and aligned groups. Alignment parameters of as-fabricated matrices are provided, in which a Mean Vector Angle of 0° indicates a horizontal orientation, an Angular Deviation of 0° is aligned and 40.5° is random, and Mean Vector Length of 1 is aligned and 0 is random. Significant difference with: A aligned U Nano-2 * Nano-1 and Micro ^ all groups (n=6, p < 0.05). Table 2 Gene primer sequences for qRT-PCR Gene Sense Antisense Blast Product Size (bp) GAPDH 5’-GGCGATGCTGGCGCTGAGTA-3’ 5’-ATCCACAGTCTTCTGGGTGG-3’ 306 Collagen I 5’-TGGTCCACTTGCTTGAAGAC-3’ 5’-ACAGATTTGGGAAGGAGTGG-3’ 118 Collagen III 5’-GGCTACTTCTCGCTCTGCTT-3’ 5’-CATATTTGGCATGGTTCTGG-3’ 130 α2 5’-CAGAATTTGGAACGGGACTT-3’ 5’-CAGGTAGGTCTGCTGGTTCA-3’ 333 β1 5’-GAGGAATACAGCCTGTGGGT-3’ 5’-ATTGCAGGATTCAGGGTTTC-3’ 121 RhoA 5’-GGGAGCTAGCCAAGATGAAG-3’ 5’-GGTCTTTGCTGAACACTCCA-3’ 55 Rac1 5’-CCATGGCTAAGGAGATTGGT-3’ 5’-GTCTTGAGGCCTCGCTGT-3’ 52 Cdc42 5’-TGGTGTCGGCATCATACTAAA-3 5’-TGTCTCACACGAGTGCATGT-3’ 98 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. All authors have no conflicts of interest 1 Frank CB Hart DA Shrive NG Molecular biology and biomechanics of normal and healing ligaments—a review. Osteoarthr Cartilage 1999 7 1 130 40 2 Frank C McDonald D Bray D Bray R Rangayyan R Chimich D Shrive N Collagen fibril diameters in the healing adult rabbit medial collateral ligament. Connect Tissue Res 1992 27 4 251 63 PM:1576825 1576825 3 Frank C McDonald D Shrive N Collagen fibril diameters in the rabbit medial collateral ligament scar: a longer term assessment. Connect Tissue Res 1997 36 3 261 9 PM:9512894 9512894 4 Doillon CJ Dunn MG Bender E Silver FH Collagen fiber formation in repair tissue: development of strength and toughness. Coll Relat Res 1985 5 6 481 92 PM:3833451 3833451 5 Lin TW Cardenas L Soslowsky LJ Biomechanics of tendon injury and repair. J Biomech 2004 37 6 865 77 PM:15111074 15111074 6 Hildebrand KA Frank CB Scar formation and ligament healing. Can J Surg 1998 41 6 425 9854530 7 Gimbel JA Van Kleunen JP Mehta S Perry SM Williams GR Soslowsky LJ Supraspinatus tendon organizational and mechanical properties in a chronic rotator cuff tear animal model. J Biomech 2004 37 5 739 49 PM:15047003 15047003 8 Moffat KL Kwei AS Spalazzi JP Doty SB Levine WN Lu HH Novel nanofiber-based scaffold for rotator cuff repair and augmentation. Tissue Eng Pt A 2009 15 1 115 26 PM:18788982 9 Lee CH Shin HJ Cho IH Kang YM Kim IA Park KD Shin JW Nanofiber alignment and direction of mechanical strain affect the ECM production of human ACL fibroblast. Biomaterials 2005 26 11 1261 70 PM:15475056 15475056 10 Xie J Li X Lipner J Manning CN Schwartz AG Thomopoulos S Xia Y “Aligned-to-random” nanofiber scaffolds for mimicking the structure of the tendon-to-bone insertion site. Nanoscale 2010 2 6 923 6 20648290 11 Meng ZX Wang YS Ma C Zheng W Li L Zheng YF Electrospinning of PLGA/gelatin randomly-oriented and aligned nanofibers as potential scaffold in tissue engineering. Mater Sci Eng C 2010 30 8 1204 10 12 Subramony SD Dargis BR Castillo M Azeloglu EU Tracey MS Su A Lu HH The guidance of stem cell differentiation by substrate alignment and mechanical stimulation. Biomaterials 2013 34 8 1942 53 PM:23245926 PMCID:3689925 23245926 13 Baker BM Mauck RL The effect of nanofiber alignment on the maturation of engineered meniscus constructs. Biomaterials 2007 28 11 1967 77 PM:17250888 17250888 14 Bashur CA Dahlgren LA Goldstein AS Effect of fiber diameter and orientation on fibroblast morphology and proliferation on electrospun poly(D,L-lactic-co-glycolic acid) meshes. Biomaterials 2006 27 33 5681 8 PM:16914196 16914196 15 Bashur CA Shaffer RD Dahlgren LA Guelcher SA Goldstein AS Effect of fiber diameter and alignment of electrospun polyurethane meshes on mesenchymal progenitor cells. Tissue Eng Pt A 2009 15 9 2435 45 PM:19292650 16 Erisken C Zhang X Moffat KL Levine WN Lu HH Scaffold fiber diameter regulates human tendon fibroblast growth and differentiation. Tissue Eng Pt A 2013 19 3-4 519 28 PM:23150905 17 Mathur A Moore SW Sheetz MP Hone J The role of feature curvature in contact guidance. Acta Biomater 2012 8 7 2595 601 PM:22426288 22426288 18 Giancotti FG Integrin signaling: specificity and control of cell survival and cell cycle progression. Curr Opin Cell Biol 1997 9 5 691 700 9330873 19 Giancotti FG Ruoslahti E Integrin signaling. Science 1999 285 5430 1028 32 PM:10446041 10446041 20 Cary LA Han DC Guan JL Integrin-mediated signal transduction pathways. Histol Histopathol 1999 14 3 1001 9 PM:10425567 10425567 21 Ozdemir T Xu LC Siedlecki C Brown JL Substrate curvature sensing through Myosin IIa upregulates early osteogenesis. Integrative Biology 2013 5 11 1407 16 24104522 22 Higgins AM Banik BL Brown JL Geometry sensing through POR1 regulates Rac1 activity controlling early osteoblast differentiation in response to nanofiber diameter. Integrative Biology 2015 7 2 229 36 25539497 23 Grinnell F Fibroblast biology in three-dimensional collagen matrices. Trends Cell Biol 2003 13 5 264 9 PM:12742170 12742170 24 Gurtner GC Werner S Barrandon Y Longaker MT Wound repair and regeneration. Nature 2008 453 7193 314 21 PM:18480812 18480812 25 Metcalfe AD Ferguson MW Tissue engineering of replacement skin: the crossroads of biomaterials, wound healing, embryonic development, stem cells and regeneration. J R Soc Interface 2007 4 14 413 37 PM:17251138 17251138 26 Gottrup F Agren MS Karlsmark T Models for use in wound healing research: a survey focusing on in vitro and in vivo adult soft tissue. Wound Repair Regen 2000 8 2 83 96 PM:10810034 10810034 27 Costa KD Lee EJ Holmes JW Creating alignment and anisotropy in engineered heart tissue: role of boundary conditions in a model three-dimensional culture system. Tissue Eng 2003 9 4 567 77 PM:13678436 13678436 28 Lu HH Cooper JA Jr. Manuel S Freeman JW Attawia MA Ko FK Laurencin CT Anterior cruciate ligament regeneration using braided biodegradable scaffolds: in vitro optimization studies. Biomaterials 2005 26 23 4805 16 PM:15763260 15763260 29 Pinnell SR Regulation of collagen biosynthesis by ascorbic acid: a review. The Yale journal of biology and medicine 1985 58 6 553 3008449 30 Kim YJ Sah RL Doong JY Grodzinsky AJ Fluorometric assay of DNA in cartilage explants using Hoechst 33258. Anal Biochem 1988 174 1 168 76 PM:2464289 2464289 31 Wang IE Mitroo S Chen FH Lu HH Doty SB Age-dependent changes in matrix composition and organization at the ligament-to-bone insertion. J Orthop Res 2006 24 8 1745 55 PM:16779829 16779829 32 Ridley AJ Schwartz MA Burridge K Firtel RA Ginsberg MH Borisy G Parsons JT Horwitz AR Cell migration: integrating signals from front to back. Science 2003 302 5651 1704 9 PM:14657486 14657486 33 Vogel V Sheetz M Local force and geometry sensing regulate cell functions. Nat Rev Mol Cell Biol 2006 7 4 265 75 PM:16607289 16607289 34 Cary LA Guan JL Focal adhesion kinase in integrin-mediated signaling. Front Biosci 1999 4 D102 D113 PM:9889179 9889179 35 Turner CE Paxillin and focal adhesion signalling. Nature cell biology 2000 2 12 E231 E236 11146675 36 Etienne-Manneville S Hall A Rho GTPases in cell biology. Nature 2002 420 6916 629 35 PM:12478284 12478284 37 Jaffe AB Hall A Rho GTPases: biochemistry and biology. Annu Rev Cell Dev Biol 2005 21 247 69 PM:16212495 16212495 38 Ridley AJ Rho GTPases and cell migration. J Cell Sci 2001 114 Pt 15 2713 22 PM:11683406 11683406 39 Nobes CD Hall A Rho GTPases control polarity, protrusion, and adhesion during cell movement. J Cell Biol 1999 144 6 1235 44 PM:10087266 10087266 40 Ridley AJ Hall A The small GTP-binding protein rho regulates the assembly of focal adhesions and actin stress fibers in response to growth factors. Cell 1992 70 3 389 99 PM:1643657 1643657 41 Nur-E-Kamal Ahmed I Kamal J Schindler M Meiners S Three dimensional nanofibrillar surfaces induce activation of Rac. Biochem Biophys Res Commun 2005 331 2 428 34 PM:15850777 15850777 42 Jaiswal D Brown JL Nanofiber diameter-dependent MAPK activity in osteoblasts. J Biomed Mater Res A 2012 100 11 2921 8 PM:22700490 22700490 43 Vunjak-Novakovic G Altman G Horan R Kaplan DL Tissue engineering of ligaments. Annu Rev Biomed Eng 2004 6 131 56 PM:15255765 15255765
PMC005xxxxxx/PMC5121022.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 0370500 441 Am J Ophthalmol Am. J. Ophthalmol. American journal of ophthalmology 0002-9394 1879-1891 27644590 5121022 10.1016/j.ajo.2016.09.014 NIHMS817321 Article Persistently Vitreous Culture-Positive Exogenous Fungal Endophthalmitis Leung Ella H. Kuriyan Ajay E. Flynn Harry W. Jr. Relhan Nidhi Huang Laura C. Miller Darlene Department of Ophthalmology, University of Miami Miller School of Medicine/Bascom Palmer Eye Institute, Miami, FL 33136 Corresponding author: Harry W. Flynn, Jr, 900 NW 17th St, Miami, FL 33136, Phone: 305-326-6118, Fax: 305-326-6417, [email protected] * Dr. Leung is currently at the Cullen Eye Institute, Baylor College of Medicine, Houston, TX, and Dr. Kuriyan is at the Flaum Eye Institute, University of Rochester Medical Center, Rochester, NY. 26 9 2016 16 9 2016 12 2016 01 12 2017 172 4550 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Purpose To report the clinical settings, microbiological isolates, and best corrected visual acuities (BCVA) of patients with persistently culture-positive exogenous fungal endophthlamitis. Design Retrospective, consecutive, case series. Methods Setting Tertiary referral center. Patient population 16 eyes of 16 patients with at least two consecutive positive vitreous cultures between 1981-2015. Interventions Intravitreal antifungal injections, pars plana vitrectomy (PPV) Main Outcome Measure Clinical settings, microbiologic isolates, BCVA Results The most common clinical settings were after cataract surgery (9/16, 56%), glaucoma surgery (4/16, 25%), and trauma (2/16, 13%). The most common single fungal isolate was Candida (4/16, 25%), but 75% of all isolates were molds. Treatment for presumed bacterial endophthalmitis was given initially in 14 patients (88%). All patients underwent a vitrectomy during the course of their treatment, and all received intravitreal or systemic antifungal therapy. The mean initial BCVA was 1.76 ± 0.9 logMAR (Snellen equivalent ≈20/1200), and the mean final BCVA was 1.84 ± 1.2 logMAR (≈20/1400, P=0.83). The 9 patients (56%) who had IOL and capsular bag removals had better final BCVAs than those who did not (P=0.011). The BCVAs were similar in eyes with yeast and mold (P=0.37). The visual acuity at the last follow-up was ≥20/40 in 13% (2/16), ≥20/400 in 50% (8/16), and no light perception in 25% (4/16). Conclusions Candida was the single most common isolate, but the majority of isolates were molds. Eyes managed with PPV and removal of the IOL and capsular bag had better visual outcomes. Persistently culture-positive fungal endophthalmitis was associated with poor final visual acuities. Introduction Exogenous fungal endophthalmitis can occur from direct inoculation after surgery or trauma or extension from infectious keratitis. The incidence of fungal endophthalmitis is approximately 1-4% after trauma, 0.002-0.005% after cataract surgery, and less than 0.5% after infectious keratitis.1-5 After trabeculectomies, the 5 year cumulative incidence of culture-proven endophthalmitis is approximately 0.5%; however, only 1% is fungal.6, 7 Reports of fungal endophthalmitis after intravitreal injections are rare and primarily associated with contaminated triamcinolone.8 The symptoms of fungal endophthalmitis can occur acutely or insidiously, with chronic inflammation and infection leading to significant damage to intraocular structures. The purpose of the study is to identify the clinical settings, fungal isolates, and best corrected visual acuities (BCVA) of patients with persistently vitreous culture-positive exogenous fungal endophthalmitis. Methods The retrospective, consecutive, case series was approved by the Institutional Review Board of the University of Miami Miller School of Medicine and was compliant with the Health Insurance Portability and Accountability Act of 1996. The research adhered to the tenets of the Declaration of Helsinki. The inclusion criteria were patients with the same exogenous fungal organism identified on at least 2 consecutive vitreous cultures obtained on separate days at the Bascom Palmer Eye Institute from January 1, 1981 to December 31, 2015. Patients with polymicrobial cultures, endogenous fungal endophthalmitis, and incomplete medical records were excluded. The vitreous samples were plated on Sabouraud agar, incubated at 35° Celsius for 72 hours, then examined daily for fungal colonies for 2 weeks. Positive samples were stained with Giemsa or calcofluor white. Fungal identification and antifungal susceptibilities were determined by sending the samples to the Fungus Testing Laboratory in San Antonio, Texas. The fungal sensitivities were not repeated for the second set of positive vitreous cultures. The treatment regimen was chosen by the individual physician based on the patient's clinical course. There were no predefined protocols for the timing and types of treatments. The indications for retreatment were clinically persistent or worsening endophthalmitis, as evidenced by increasing fibrin or hypopyon, persistent fungal infiltrates, worsening visual acuity, and/or persistent intraocular inflammation. Statistical calculations were performed using the Statistical Package for the Social Sciences software (SPSS Inc, Chicago, Illinois, USA), with a P value less than 0.05 being considered statistically significant. Snellen visual acuity was converted to its logarithm of minimal angle of resolution (logMAR) equivalent, with counting fingers being assigned a value of 1.9, hand motion 2.3, light perception 2.7, and no light perception 3.0.9 The best corrected visual acuities (BCVA) are presented as the mean logMAR ± standard deviation, followed by the approximate Snellen chart equivalent. The visual acuities were analyzed using student's t-test and one-way analysis of variance with Tukey post-hoc analyses. Results Of the 165 patients with positive fungal vitreous cultures over the 35 year study period, eighteen patients had two positive fungal cultures (11%). Two cases from the 1980s were unavailable for review; therefore, sixteen eyes in 16 patients were included in the study. The mean age was 68.6 ± 18 years old; ten patients were males (63%), and half were right eyes (8/16). The most common past medical histories were hypertension (7/16, 44%) and diabetes mellitus (4/16, 25%). Excluding the 4 diabetic patients, none were systemically immunocompromised. Four patients (25%) were on topical steroids on presentation. The most common past ocular histories were glaucoma (5/16, 31%), corneal transplants (2/16, 13%), and retinal detachment (1/16, 6%). The mean follow-up period was 42 months (range: 7-288 months). Table 1 summarizes the patient demographics. The clinical settings included cataract surgery (9/16, 56%), glaucoma surgery (4/16, 25%), trauma (2/16, 13%), and corneal ulcer (1/16, 6%). Excluding the two patients whose previous surgical dates were unknown, the mean time from the predisposing event to the initial onset of symptoms and treatment was 2.0 ± 1.6 months (Figure 1). The fungal isolates were Candida (4/16, 25%, including C. albicans (2), C. parapsilosis (2)), Fusarium (3/16, 19%), Acremonium strictum (2/16, 13%), Curvularia (2/16, 13%), Paecilomyces lilacinus (2/16, 13%), Aspergillus nigricans (1/16, 6%), Phialophora richardsiae (1/16, 6%), and Helicomyces (1/16, 6%). Fourteen patients (88%) were initially treated for presumed bacterial endophthalmitis with intravitreal injections of vancomycin and ceftazidime; two patients received concurrent intravitreal steroids before the diagnoses of fungal endophthalmitis were made (13%). Only two patients (13%) who presented with fungal infiltrates in the anterior chamber and vitreous cavity received intravitreal antifungal therapy as part of their initial therapy. A mean of 4.4 ± 2.8 antifungal injections and 1.1 ± 0.6 antibacterial injections were administered. Upon identification of fungal isolates, five patients (31%) received intravitreal antifungal injections while 9 (56%) had PPV with antifungal injections. One patient resolved his infection without intravitreal antifungal therapy; he had oral voriconazole and a PPV with removal of the IOL and capsular bag. The initial intravitreal antifungal therapies were amphotericin B (12/16, 75%), voriconazole (4/16, 25%), and miconazole (3/16, 19%). Twelve patients (75%) were placed on systemic antifungal treatment in order to augment local therapy; the medications included oral voriconazole (3/16, 19%), oral diflucan (3/16, 19%), oral ketoconazole (2/16, 13%), oral natamycin (2/16, 13%), intravenous voriconazole (1/16, 6%), and oral fluconazole (1/16, 6%). All patients underwent a pars plana vitrectomy at least once during the course of their treatment. The 9 patients (53%) whose IOLs and capsular bags were removed had significantly better final visual acuities than those who did not (1.1 ± 1.4 logMAR, ≈20/260, vs. 2.5 ± 0.8 logMAR, light perception, P=0.011) (Figure 2). Of the 7 patients whose IOLs and capsular bags were not removed, four (57%) proceeded to enucleation or alcohol ablation. The fungal sensitivities were available in 9 patients (56%). The minimum inhibitory concentrations (MIC) for all isolates ranged from 0.06-2.1bμg/ml for voriconazole, 0.2-8 μg/mL for amphotericin B, and 0.25-6bμg/ml for fluconazole. Molds had higher median MICs compared to yeast. The sensitivities are summarized in Table 2. In the current study, the mean visual acuity on presentation was 1.76 ± 0.9 logMAR (Snellen equivalent ≈20/1200), which was similar to the mean vison at the last examination (1.84 ± 1.2 logMAR, ≈20/1400, P=0.83). Two patients (13%) achieved BCVAs of 20/40 or better at the last examination, and half achieved ≥20/400 (8/16). The final visual acuity was similar in mold and yeast (P=0.37). The pre-infection BCVA was known in 10 patients; four patients (40%) achieved a final visual acuity within 1 line of their pre-endophthalmitis BCVA. Patients who underwent initial vitrectomies had worse final vision than those who had initial vitreous tap and injects (2.2 ± 1.0 logMAR, ≈20/3000, vs. 1.1 ± 1.0 logMAR, ≈ 20/220, P=0.040). The two patients who received early intravitreal steroids had worse initial and final visual BCVAs than patients who did not (final BCVA: no light perception vs. 1.6 ± 1.1 logMAR, ≈20/800, respectively, P=0.0028). There was no difference in the final BCVAs based on the type of antifungal therapy (P=0.31), early antifungal therapy (P=0.14), or the adjunctive use of systemic antifungal therapy (P=0.63). Table 3 summarizes the mean best corrected visual acuities at the initial and last follow-up examinations. Discussion Since fungal endophthalmitis is rare, many patients are initially treated for suspected bacterial infections. All but 2 patients with fungal endophthalmitis in the current series were initially treated for bacterial endophthalmitis. Delays in the identification and treatment of fungal infections can result in significant intraocular inflammation and tissue damage. Untreated, Aspergillus can cause diminution of the a-waves and b-waves on electroretinogram within 72 hours of infection in animal models.10 Furthermore, active persistent infection can be difficult to differentiate from residual non-infectious inflammation; in these cases, repeat vitreous cultures can help guide further treatments. The most common clinical setting for persistent exogenous fungal endophthalmitis was after cataract surgery, which is one of the most frequently performed anterior segment surgeries. Glaucoma surgery was the second most common clinical setting. Artificial shunts and glaucoma drainage implants may be associated with bleb leakage, tube erosions, and infections.11, 12 The most common fungal isolate in the current case series was Candida (4/16, 25%); however, mold was the most common type of fungal isolate (Fusarium, Acremonium, Curvularia, Paecilomyces, Aspergillus, Phialophora, and Helicomyces). Aspergillus comprised 6% (1/16). In contrast, the most commonly reported cause of exogenous fungal endophthalmitis in the literature was Aspergillus (54-74%).13, 14 The present study was conducted in an academic facility in South Florida, with referrals from tropical regions; therefore, the fungal isolates and the prevalence of persistent fungal infections in the current study may not reflect the organisms present in other regions of the world. Antifungal sensitivities were obtained in 56% (9/16) of the persistent fungal endophthalmitis patients. The MICs of isolates to voriconazole and amphotericin B were similar between persistent cases and all cases of fungal endophthalmitis.14, 15 Intravitreal voriconazole causes less retinal toxicity than amphotericin B, but the concentrations of voriconazole remain above the MIC for most fungal organisms for only about 8 hours. In contrast, amphotericin B has a half-life of 1.8 days in aphakic, vitrectomized eyes and 7-15 days in phakic eyes.15, 16 Repeat injections of antifungal therapy may be necessary to adequately treat fungal endophthalmitis. The pathogenesis of persistent exogenous fungal endophthalmitis is unclear but may occur through antifungal resistance, poor drug penetrance, intraocular fungal sequestration, and/or host immunosuppression.14-17 Resistance to azoles and amphotericin B have been reported due to genetic mutations that increase azole efflux, upregulate P450 drug metabolism, alter the ergosterol biosynthesis pathway, overexpress sterole binding sites, and cause chromosomal non-disjunction.17 In addition, ketoconazole has poor oral absorption, and intravenous amphotericin B does not reach therapeutic levels in the vitreous.14 Spores from molds may also sequester inside the eye and are more resistant to azole antifungals.17 Certain species of Candida, Aspergillus, and Fusarium can form biofilms, which can withstand up to 1000 times the concentration of antifungals required to treat non-encased, planktonic fungus.18, 19 Methods to overcome antifungal resistance include injecting intravitreal medications, increasing drug concentrations, using lipid formulations of amphotericin B, combining antifungal therapies, and surgically excising fungal infiltrates.17 Pars plana vitrectomy with removal of the intraocular lens and capsular bag may eliminate the nidus for infection. Irrigation during vitrectomy may remove inflammatory and infectious debris.20 In two case series with a total of 18 patients with fungal endophthalmitis, the infections resolved after PPV with intravitreal antifungal therapy and IOL explantations.20, 21 All patients in the current series underwent a PPV during the course of their treatment. One patient (6%) resolved his infection without intravitreal antifungal therapy by undergoing a PPV with removal of the IOL and capsular bag and oral voriconazole. The overall final visual acuity was poor in patients with persistent fungal endophthalmitis and was similar between eyes with mold or yeast. Only 13% of patients (2/16) achieved 20/40 or better, half were ≥20/400, and 25% of eyes (4/16) were enucleated. A study that included both persistent and non-persistent cases of fungal endophthalmitis found similar visual acuity outcomes; approximately 54% achieved 20/400 or better, and 24% were enucleated.22 The 9 patients (57%) who underwent IOL/capsular bag removals had better final visual acuities than those who did not (20/260 vs. light perception, P=0.011), which may have been partly due to improved clearance of infections after the vitrectomies and IOL/capsular bag removals. Patients who underwent initial PPV and initial intravitreal steroids had worse final visual outcomes, likely reflecting selection biases. Patients with worse clinical presentations were more likely to undergo PPVs than tap and injects. The two patients who received steroids presented with light perception vision and were eventually enucleated. Steroids decrease inflammation, but when administered without concurrent antifungal therapy, in vitro studies have demonstrated decreased effectiveness of macrophages and monocytes in suppressing fungal growth.19 When dexamethasone is given with amphotericin B, however, several case series reported faster clearance of endophthalmitis and not worse clinical outcomes.19, 23 The use of intravitreal steroids in fungal endophthalmitis remains controversial. The limitations of the current study include the retrospective nature, relatively small number of patients, selection bias, and difficulty in obtaining positive fungal cultures. The ability to identify fungal growth with positive cultures has been reported to be between 25-70%.24 Although not used in the current study, polymerase chain reaction could increase early detection of fungal infections.22 Furthermore, it was not possible to demonstrate a difference in clinical outcomes between patients treated with voriconazole compared to amphotericin B in the current study. In conclusion, persistent vitreous culture-positive exogenous fungal endophthalmitis is uncommon. The most common clinical setting was cataract surgery, and the most common isolate was mold. Multiple intravitreal antifungal injections may be necessary to resolve the fungal endophthalmitis. Patients who underwent pars plana vitrectomies with removal of the IOLs/capsular bags had better final vision. Regardless of the etiology, visual acuity outcomes at the last follow-up were generally poor. Supplementary Material 1 2 Funding/Support: The research was supported in part by the National Eye Institute Center Core Grant (P30EY014801), a Department of Defense Grant (W81XWH-09-1-0675, Washington, D.C.), and an unrestricted grant from the Research to Prevent Blindness, Inc., New York, NY to the Department of Ophthalmology, University of Miami Miller School of Medicine. The funding organizations had no roles in the design or conduct of the research. Other Acknowledgements: None. Figure 1 Time to Presentation Box-plot demonstrating the months between the predisposing event to initial presentation with endophthalmitis. The graph excludes 2 patients whose surgical dates were unknown. Figure 2 Endophthalmitis resolution without intravitreal antifungal An 81 year old male presented with pain and blurriness in the right eye 2 months after cataract surgery. The initial vision was 20/200 in the right eye. The patient was found on examination to have white plaques on the capsular bag that was visible on slit lamp microscopy (Top Left) and retro-illumination (Top Right). The patient underwent a PPV/intravitreal vancomycin/intravitreal ceftazidime in the right eye, and the vision improved to 20/50. Three months later, however, the patient developed a recurrent hypopyon and vitritis (Bottom Left). He underwent a PPV/IOL removal/intravitreal voriconazole/intravitreal triamcinolone/vitreous cultures; he was also on oral voriconazole. The endophthalmitis resolved, and at the last examination 21 months later, the vision had improved to 20/30 in the right eye (Bottom Right). Table 1 Demographics and Clinical Course Summary of the patient demographics, clinical course, and final visual acuities in patients with persistently culture-positive fungal endophthalmitis. Fungal Endophthalmitis Mean Age 68.8 ± 18 years Mean Follow-Up 42.0 ± 72 months Mean Number of Antibiotics 1.1 Mean Number of Antifungals 2.7 Mean Number of Treatments 4.4 Mean time to 1st treatment 2.0 ± 1.6 months Mean Initial Visual Acuity (logMAR, Snellen equivalent) 1.76 ± 0.9 ≈ 20/1200 Mean Final Visual Acuity (logMAR, Snellen equivalent) 1.84 ± 1.2 ≈ 20/1400 Table 2 Fungal Minimum Inhibitory Concentrations Comparison of the Minimum Inhibitory Concentrations (μg/ml) for different antifungal therapies in patients with persistent culture-positive endophthalmitis.* Fungal Isolate (Number of isolates= 9) Amphotericin B (μg/ml) Voriconazole (μg/ml) Fluconazole (μg/ml) Acremonium (1) 1 0.5 0.5 Aspergillus (1) 8 0.25 64 Candida (1) 0.06 0.02 0.25 Curvularia (1) 0.5 0.06 4 Dematicus (1) 0.2 - - Fusarium (2) 1.0 2.1 64 Paecilomyces (1) 16 0.25 64 Phialophora (1) 0.5 - - * Included are 3 eyes whose specimens were previously re-cultured to determine the antifungal sensitivities.12 Table 3 Mean Visual Acuities Comparison of mean initial and final best corrected visual acuities based on clinical setting of patients with persistent culture-positive fungal endophthalmitis. Clinical Setting (Number of patients) Initial Visual Acuity Final Visual Acuity logMAR Snellen Equivalent logMAR Snellen Equivalent Cataract Surgery (9) 1.4 ± 0.9 20/470 1.4 ± 1.1 20/500 Glaucoma Surgery (4) 1.6 ± 0.9 20/730 2.3 ± 1.2 Hand Motion Trauma (2) 2.3 ± 0.6 Hand Motion 2.9 ± 0.2 Light Perception Corneal Surgery (1) 1.3 ± 0 20/400 1.3 ± 0 20/400 Financial Disclosures: No financial disclosures. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. 1 Du DT Wagoner A Barone SB Incidence of endophthalmitis after corneal transplant or cataract surgery in a medicare population Ophthalmology 2014 121 1 290 298 23993357 2 Behlau I Martin KV Martin JN Infectious endophthalmitis in Boston keratoprosthesis: incidence and prevention Acta Ophthalmol 2014 92 7 e546 555 24460594 3 Ahmed Y Schimel AM Pathengay A Endophthalmitis following open-globe injuries Eye (Lond) 2012 26 2 212 217 22134598 4 Gupta A Srinivasan R Gulnar D Risk factors for post-traumatic endophthalmitis in patients with positive intraocular cultures Eur J Ophthalmol 2007 17 4 642 647 17671943 5 Henry CR Flynn HW Jr Miller D Infectious keratitis progressing to endophthalmitis: a 15-year study of microbiology, associated factors, and clinical outcomes Ophthalmology 2012 119 12 2443 2449 22858123 6 Vaziri K Kishor K Schwartz SG Incidence of bleb-associated endophthalmitis in the United States Clin Ophthalmol 2015 9 317 322 25709395 7 Busbee BG Recchia FM Kaiser R Bleb-associated endophthalmitis: clinical characteristics and visual outcomes Ophthalmology 2004 111 8 1495 1503 15288977 8 Mikosz CA Smith RM Kim M Fungal endophthalmitis associated with compounded products Emerg Infect Dis 2014 20 2 248 256 24447640 9 Lange C Feltgen N Junker B Resolving the clinical acuity categories “hand motion” and “counting fingers” using the Freiburg Visual Acuity Test (FrACT) Graefes Arch Clin Exp Ophthalmol 2009 247 1 137 142 18766368 10 Harrison JM Glickman RD Ballentine CS Retinal function assessed by ERG before and after induction of ocular aspergillosis and treatment by the anti-fungal, micafungin, in rabbits Doc Ophthalmol 2005 110 1 37 55 16249956 11 Levinson JD Giangiacomo AL Beck AD Glaucoma drainage devices: risk of exposure and infection Am J Ophthalmol 2015 160 3 516 521 26032191 12 Kim EA Law SK Coleman AL Long-Term Bleb-Related Infections After Trabeculectomy: Incidence, Risk Factors, and Influence of Bleb Revision Am J Ophthalmol 2015 159 6 1082 1091 25748577 13 Kim DY Moon HI Joe SG Recent Clinical Manifestation and Prognosis of Fungal Endophthalmitis: A 7-Year Experience at a Tertiary Referral Center in Korea J Korean Med Sci 2015 30 7 960 964 26130961 14 Silva RA Sridhar J Miller D Exogenous fungal endophthalmitis: an analysis of isolates and susceptibilities to antifungal agents over a 20-year period (1990-2010) Am J Ophthalmol 2015 159 2 257 64 25449001 15 Shen YC Wang MY Wang CY Clearance of intravitreal voriconazole Invest Ophthalmol Vis Sci 2007 48 5 2238 2241 17460285 16 Wingard LB Jr Zuravleff JJ Doft BH Intraocular distribution of intravitreally administered amphotericin B in normal and vitrectomized eyes Invest Ophthalmol Vis Sci 1989 30 10 2184 2189 2793359 17 Kontoyiannis DP Lewis RE Antifungal drug resistance of pathogenic fungi Lancet 2002 359 9312 1135 1144 11943280 18 Nett JE Andes DR Fungal Biofilms: In Vivo Models for Discovery of Anti-Biofilm Drugs Microbiol Spectr 2015 3 3 19 Majji AB Jalali S Das T Role of intravitreal dexamethasone in exogenous fungal endophthalmitis Eye (Lond) 1999 13 5 660 665 10696322 20 Vinekar A Dogra MR Avadhani K Management of recurrent postoperative fungal endophthalmitis Indian J Ophthalmol 2014 62 2 136 140 24618484 21 Mithal K Pathengay A Bawdekar A Filamentous fungal endophthalmitis: results of combination therapy with intravitreal amphotericin B and voriconazole Clin Ophthalmol 2015 9 649 655 25926714 22 Wykoff CC Flynn HW Jr Miller D Exogenous fungal endophthalmitis: microbiology and clinical outcomes Ophthalmology 2008 115 9 1501 1507 18486220 23 Coats ML Peyman GA Intravitreal corticosteroids in the treatment of exogenous fungal endophthalmitis Retina 1992 12 1 46 51 1565871 24 Liu K Fang F Li H Reliability of vitreous histological detection of pathogenic fungi in the diagnosis of fungal endophthalmitis Eye (Lond) 2015 29 3 424 427 25397787
PMC005xxxxxx/PMC5121042.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 0406041 4668 J Comp Neurol J. Comp. Neurol. The Journal of comparative neurology 0021-9967 1096-9861 27213991 5121042 10.1002/cne.24044 NIHMS789132 Article PATTERN OF DISTRIBUTION OF SEROTONERGIC FIBERS TO THE AMYGDALA AND EXTENDED AMYGDALA IN THE RAT Linley Stephanie B. 12 Olucha-Bordonau Francisco 3 Vertes Robert P. 2* 1 Department of Psychology, Florida Atlantic University, Boca Raton, FL 33431 2 Center for Complex Systems and Brain Sciences, Florida Atlantic University, Boca Raton, FL 33431 3 Departamento de Medicina, Facultad de Ciencias de la Salud, Universitat Jaume I, Castellón, Spain Address for correspondence: Dr. Robert P. Vertes, Center for Complex Systems and Brain Sciences, Florida Atlantic University, Boca Raton, FL 33431, phone: 561-297-2362, fax: 561-297-2363, [email protected] 25 5 2016 19 6 2016 1 1 2017 01 1 2018 525 1 116139 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. As well recognized, serotonergic (5-HT) fibers distribute widely throughout the forebrain, including the amygdala. Although a few reports have examined the 5-HT innervation of select nuclei of the amygdala in the rat, no previous report has described overall 5-HT projections to the amygdala in the rat. Using immunostaining for the serotonin transporter, SERT, we describe the complete pattern of distribution of 5-HT fibers to the amygdala (proper) and to the extended amygdala in the rat. Based on its ontogenetic origins, the amygdalar was subdivided into two major parts: pallial and subpallial components, with the pallial component further divided into superficial and deep nuclei (Olucha-Bordonau et al., 2015). SERT+ fibers were shown to distribute moderately to densely to the deep and cortical pallial nuclei, but, by contrast, lightly to the subpallial nuclei. Specifically, (1) of the deep pallial nuclei, the lateral, basolateral and basomedial nuclei contained a very dense concentration of 5-HT fibers; (2) of the cortical pallial nuclei, the anterior cortical and amygdala-cortical transition zone, rostrally, and the posteromedial and posterolateral nuclei, caudally, contained a moderate concentration of 5-HT fibers; and (3) of the subpallial nuclei, the anterior nuclei and the rostral part of the medial (Me) nuclei contained a moderate concentration of 5-HT fibers, whereas caudal regions of Me as well as the central nuclei and the intercalated nuclei contained a sparse/light concentration of 5-HT fibers. Regarding the extended amygdala (primarily the bed nucleus of stria terminalis, BST), on the whole, the BST contained moderate numbers of 5-HT fibers, spread fairly uniformly throughout BST. The findings were discussed with respect to a critical serotonergic influence on the amygdala, particularly on the basal complex, and extended amygdala in the control of states of fear and anxiety. Graphical Abstact Transverse section mid-rostrocaudally through the amygdala showing the pattern of distribution of serotonin-immunoreactive (SERT-positive) fibers to nuclei of the amygdala. As depicted, SERT+ fibers terminate heavily in the anterior and posterior basolateral nucleus of amygdala (BLAa and BLAp), which contrasts with the sparse distribution to the central nucleus of amygdala, dorsomedial to BLA. fear anxiety stress basolateral complex of amygdala central nucleus of amygdala pallial amygdala subpallial amygdala bed nucleus of stria terminalis 5-HT1A receptors 5-HT2C receptors The neurochemical and neuroanatomical substrates for affective behavior have been well studied. At the forefront, both serotonin, chemically, and the amygdala, anatomically, are key substrates for emotional behavior (LeDoux, 2000, 2003; Shinnick-Gallagher et al., 2003; Whalen and Phelps, 2009; Asan et al., 2013; Bauer, 2015). The amygdala (and extended amygdala) has been shown to be critically involved in affective behaviors including fear, stress, anxiety and reward, and is central to the processing of emotionally relevant information (LeDoux, 2000, 2003; Shinnick-Gallagher et al., 2003; Phelps, 2006; Whalen and Phelps, 2009; Duvarci and Pare, 2014; Olucha-Bordonau et al., 2015). Serotonin also serves a well-documented role in affective processes (Lowry et al., 2005, 2008; Hayes and Greenshaw, 2011; Asan et al., 2013; Bauer, 2015). Dysfunction of either system has been associated with a host of psychiatric disorders including depression, chronic anxiety, and post-traumatic stress disorder (PTSD) (Walker et al., 2003; Kalia, 2005; Nemeroff et al., 2006; Shin et al., 2006; Holmes, 2008; Shin and Liberzon, 2010; Hale et al., 2012; Fox and Lowry, 2013) While earlier studies in the rat, using various techniques, identified 5-HT fibers within select nuclei of the amygdala, no previous report has comprehensively examined the overall distribution of 5-HT fibers to the amygdala, and the 5-HT innervation of the ‘extended amygdala’ has been largely unexplored (Parent et al., 1981; Steinbush, 1981). Previous studies in the rat have mainly focused on 5-HT input to the central and basolateral (BLA) nuclei of the amygdala, showing that 5-HT fibers distribute densely to BLA and by comparison lightly to Ce (Asan et al., 2005; Muller et al., 2007; Smith and Porrino, 2008). Interestingly, this is essentially the reverse of the pattern shown for (non-human) primates wherein Ce is densely populated, BLA relatively lightly so, with 5-HT axons (Sadikot and Parent, 1990; Freedman and Shi, 2001; Bauman and Amaral, 2005; O’Rourke and Fudge, 2006; Smith and Porrino, 2008). As both the amygdala and serotonergic systems critically participate in several affective functions, notably fear and anxiety, it would be important to determine the precise pattern of termination of 5-HT fibers to the amygdala and BST -- as they influence these functions. Using immunohistochemical procedures for the detection of the serotonin transporter protein (SERT), we examined the distribution of 5-HT fibers to the amygdala (proper) and to BST, with attention to differential innervation of subnuclei. In brief, we showed serotonergic fibers distribute strongly, but heterogeneously, throughout the amygdala and BST. Overall, 5-HT fibers terminate densely in the deep and cortical pallial nuclei, particularly pronounced to the basolateral complex. By contrast, the central and medial nuclei of the subpallial amygdala receive a modest serotonergic innervation. These differences in the pattern and density of 5-HT afferents to the amygdala would appear to reflect a differential serotonergic influence on discrete regions of the amygdala in various affective behaviors. MATERIALS AND METHODS Ten (5 male, 5 female) naïve Sprague Dawley rats (Harlan, Indianapolis, IN) weighing 275–300 g on arrival were housed in pairs on a 12:12 light cycle for 7 days, during which food and water were given ad libitum. These experiments were approved by the Florida Atlantic University Institutional Animal Care and Use Committee and conform to all federal regulations and National Institutes of Health guidelines for the care and use of laboratory animals. Rats were deeply anaesthetized with an intraperitoneal injection of sodium pentobarbital (Nembutal, 75 mg/kg). Rats were perfused transcardially with 30–50 ml of ice cold heparinized 0.1 M phosphate buffer saline (PBS) followed by 200–300 ml of chilled 4% paraformaldehyde in 0.1 M phosphate buffer (PB) at a pH = 7.4. The brains were removed and postfixed for 24–48 hrs in 4% paraformaldehyde in 0.1M PB. Brains were then placed in a 30% sucrose solution for another 48 hours. Following sucrose cryoprotection, 50 μm sections were cut on a freezing sliding microtome. Sections were collected in a six well plate using 0.1 M PB as a storage solution, so that every sixth section was represented throughout the brain for each series of sections. Sections were stored in 0.1 M PB at 4°C until the tissue was prepared for immunohistochemistry. SERT immunohistochemistry Immunohistochemical analysis to detect serotonergic axons was done with an antiserum to SERT using an avidin biotin protein complex protocol as was described previously (Vertes et al., 2010; Linley et al., 2013). Free-floating sections were treated with 1% sodium borohydride (NaBH4) in 0.1 M PB to remove excess aldehydes. Following copious 0.1 M PB washes, sections were stored for one hour in 0.5% bovine serum albumin (BSA) in 0.1 M tris buffered saline (TBS; pH = 7.6) containing 0.25% Triton X-100. Sections were then incubated in the primary polyclonal antibody, rabbit anti-SERT (ImmunoStar, cat. # 24330, RRID: AB_572209; Hudson, WI). The SERT antibody was raised to a synthetic peptide corresponding to amino acids 579–599 of rat SERT coupled to keyhole limpet hemocyanin (KLH). The primary antibody was placed in a diluent of 0.1% BSA TBS containing 0.25% Triton X-100 at a concentration of 1:5,000 at room temperature for 24–48 hours. Following further washes, sections were placed in a secondary antibody, biotinylated goat anti-rabbit immunoglobulin (Vector Laboratories, cat. # BA-1000, RRID: AB_2313606, Burlingame, CA) in diluent at a 1:500 concentration for two hours. This was followed by another series of PB washes. Sections were then incubated for two hours in a tertiary antibody: biotinylated horse anti-goat immunoglobulin (Vector Laboratories, cat. # BA9500, RRID: AB_2313580, Burlingame, CA) in diluent at a 1:500 concentration. After washing the tissue in 0.1M PB, sections were incubated for one hour in an avidin biotin complex (ABC) using the VECTASTAIN Elite ABC-Peroxidase kit (Vector Laboratories, cat. # PK-7100, RRID: AB_2336827, Burlingame, CA) in a diluent at a 1:200 concentration. Following final 0.1 M PB washes, serotonin fibers expressing the serotonin transporter protein were visualized with the chromogen: 0.022% 3,3′ diaminobenzidine (DAB) (Aldrich, Milwaukee, WI) and 0.003% hydrogen peroxide in TBS for approximately 4–6 minutes. Sections were stored in 0.1M PB at 4°C until mounted onto chrome-alum gelatin-coated slides, dehydrated using graded methanols and coverslipped with Permount. The specificity of the primary and secondary antibodies have been verified in previous studies (Vertes et al., 2010; Linley et al., 2013). Using the present antiserum to SERT, immunostained sections through the pons and mesencephalon displayed identical patterns of cell and fiber SERT+ labeling as shown previously for these regions of the upper brainstem (Sur et al., 1996; Vertes and Crane, 1997; Yamamoto et al., 1998). Additionally, sections reacted without the primary or secondary antibodies did not show immunoreactivity (data are not shown). Nissl staining To reference the cytoarchitectonic borders of nuclei of the amygdala and BST, sections throughout the amygdala and extended amygdala were visualized using a cresyl violet stain. Briefly, sections were mounted on chrome-alum–coated slides, rinsed 2 × 10 min in xylene, rehydrated in graded methanols, and immersed in a 0.1% cresyl violet (Acros Organics, Thermo Fisher Scientific, Waltham, MA) in an acetic acid buffer for 15–20 minutes. Slides were then rinsed briefly with deionized water, dehydrated in graded methanols (to 100% methanol), cleared with xylene, and coverslipped with Permount. Photomicroscopy Lightfield photomicrographs at 100× magnification were taken for visualization of SERT+ fibers throughout the rostrocaudal extent of the amygdala (proper) and the extended amygdala. Photomicrographs were captured using a Q Imaging (QICAM) camera mounted onto a Nikon Eclipse E600 microscope. Digital images were captured and reconstructed using Nikon Elements and then imported into Adobe Illustrator and Photoshop CC 2014 (Mountain View, CA) to adjust brightness and contrast and to outline borders of nuclei of the amygdala/BST. Representative Nissl stained sections throughout the extent of the amygdala/BST were captured and imported into Adobe Illustrator to map the subdivisions of nuclear groups. A note on nomenclature for the amygdala and extended amygdala The nomenclature and nuclear demarcations used for nuclei of the amygdala and extended amygdala are those of Olucha-Bordonau et al. (2015), as follows. The amygdaloid complex has recently been subdivided according to its ontogenetic origin into two major divisions: pallial and subpallial components, with the pallial component further divided into superficial and deep nuclei (Medina et al., 2004; García-López et al., 2008; Martínez-García et al., 2008; Bupesh et al., 2011; Hertel et al., 2012; Olucha-Bordonau et al., 2015). This demarcation is analogous to a cortical/subcortical differentiation -- with the cortical component showing a distinct lamination, and with a few exceptions the subcortical sector exhibiting no layering. Further, the pallial amygdala, like the cortex, contains a greater proportion of glutamatergic to GABAergic cells whereas the subpallial amygdala, like subcortical telencephalic structures, contains significantly more GABAergic relative to glutamatergic cells. The superficial division of the pallial amygdala (or the cortical pallial amygdala) consists of the anterior (ACo), posterolateral (PLCo) and posteromedial (PMCo) cortical nuclei as well as transition zones between the ACo and piriform cortex; that is, rostrally, the cortical-amygdala transition zone (CxA) and caudally the amygdalo-piriform transition area (APir). The deep (or nuclear) division of the pallial amygdala consists of the oft-referred to ‘basolateral complex’ of the amygdala comprised of the lateral (La), basolateral (BLA) and basomedial (BMA) nuclei -- as well as the amygdalo-hippocampal area (AHi). The subpallial division of the amygdala primarily consists of the medial (Me) and central (Ce) and anterior nuclei, and the intercalated cell masses. The subpallial region of the ‘extended amygdala’ (EA) consists of two components: (1) the bed nucleus of the stria terminalis (BST) which includes the intra-amygdala area (STIA), the supracapuslar EA (STS) and the BST (proper) divisions; and (2) the sublenticular extended amygdala (SLEA) composed of the sublenticular part of the substantia innominata (SI) and the interstitial nucleus of the posterior limb of the anterior commissure (IPAC), lying dorsolateral to the anterior amygdaloid area. (Olucha-Bordonau et al., 2015). The septal sector of BST (or BST proper) surrounds the anterior commissure and is composed of several subnuclei -- with those of the anterior division associated with the central nucleus of amygdala and those of the posterior division associated with the medial nucleus of amygdala (Dong et al., 2001; Alheid, 2003; Olucha-Bordonau et al., 2015). RESULTS We describe the pattern of distribution of serotonergic fibers to the amygdala (proper) as well as that to the extended amygdala (or bed nucleus of the stria terminalis complex), as illustrated by a representative case (case 14). The findings of non-illustrated cases directly correspond to those of the illustrated case. Amygdala Figures 1–11 depict the pattern of distribution of 5-HT fibers rostrocaudally throughout the amygdala, and consist of a series of Nissl-stained sections (Figs. 1A–11A) showing the locations of nuclei of the amygdala and corresponding SERT-immunostained sections (Figs. 1B–11B) depicting patterns of labeling at these levels of the amygdala. Figure 1B shows the pattern of distribution 5-HT fibers at a caudal level of the extended amygdala (EA) as it merges with the amygdala proper. As depicted (Fig. 1B), the interstitial nucleus of the posterior limb of the anterior commissure (IPAC), surrounding the anterior commissure (ac), was moderately labeled. This contrasts with considerably lighter labeling of IPAC at a more anterior level (see Fig. 14). At the rostral pole of the amygdala (Fig. 2B), the anterior dorsal (AAD) and anterior ventral (AAV) nuclei as well as the rostral extent of the basomedial nucleus (BMA) were moderately to densely labeled. SERT+ fibers spread moderately throughout cortical pallial nuclei, distributing non-homogenously to the anterior cortical nucleus (ACo) and the cortical amygdala transition zone (CxA), but evenly within the nucleus of the lateral olfactory tract (LOT). Regarding ACo and CxA, labeling was dense in superficial layers (layer 1) and tapered in deeper layers. This same pattern continued laterally from CxA to the piriform cortex. With the emergence of the central nucleus (Ce) at this level (Fig. 2B), the capsular division of Ce (CeC) was lightly labeled; the medial division (CeM) moderately labeled. Further caudally in the anterior amygdala (Fig. 3B), 5-HT fibers distributed heavily to the anterior basolateral (BLAa) and anterior basomedial (BMAa) nuclei, and significantly but less densely to AAD and to the anterodorsal part of the medial nucleus (MeAD), lying ventral to AAD. By contrast, labeling was less pronounced (or moderate) in CeM and in the amygdalostriatal transition area (ASt), lateral to CeC. The pattern of labeling of ACo and CxA was similar to that of rostral levels, with a decrease in density from superficial to deep layers, and was stronger in ACo and CxA than in the medially adjacent bed nucleus of the accessory olfactory tract (BAOT). As seen rostrally, the CeC was lightly labeled, whereas the intercalated nuclei (I) were lightly to moderately labeled. Progressing posteriorly (Fig. 4B), marked differences began to emerge in the relative density of labeling across nuclei of the amygdala. More precisely, the very dense labeling of the lateral nucleus (La) and the anterior (BLAa) and posterior (BLAp) divisions of the basolateral nucleus (BLA), strongly contrasted with the light labeling of the medially adjacent central nucleus and ASt. This trend would continue further caudally. The ventral basolateral nucleus (BLAv) and BMAa contained slightly fewer labeled fibers than did BLA, whereas AAD, the dorsal (MeAD) and ventral (MeAV) medial nuclei and the cortical nuclei (CxA, ACo, and BAOT) were less densely labeled than the basal groups. As rostrally, layer 1 of cortical areas ACo and CxA was heavily labeled; layers 2/3 moderately labeled. At the mid-rostrocaudal amygdala (Fig. 5B), the basal group continued to be strongly labeled which, as rostrally, differed significantly from the sparse to light labeling of Ce. Labeling was slightly heavier in CeM than in CeC or in the lateral division of Ce (CeL). Of the basal group, labeling was densest in BLA (BLAa and BLAp) followed by La, BMAa and BLAv which were similarly labeled. With the exception of the intercalated group which was lightly labeled, remaining nuclei at this level were moderately labeled. This includes the ASt, MeAD, MeAV, ACo, CxA, and the intra-amygdalar component of BST (STIA). As rostrally, labeling gradually weakened from superficial to deep layers of ACo and CxA. Further caudally (Fig. 6B), labeled fibers were densely concentrated in La and bordering regions of BLAa and BLAp, but tapered slightly in ventral parts of the basal complex; that is, in the anterior (BMAa) and posterior (BMAp) divisions of BMA and in BLAv. All divisions of the Ce were sparsely labeled, with a few more fibers in CeM than in the other divisions of Ce. The ASt component of Ce was moderately labeled. Labeling thinned from that of anterior levels in MeAD and MeAV such at this level both divisions of Me were lightly labeled. The ACo and the posterolateral cortical nucleus (PLCo) (which now emerged), as well as the STIA were moderately labeled; the intercalated group lightly labeled. At a mid-posterior level (Fig. 7B), labeled fibers distributed moderately to densely to both superficial (cortical) and deep pallial nuclei; that is, to ACo, PLCo, La, BLAa, BLAp, BLAv, BMAa and BMAp. Labeling was heaviest in La, the anterior and posterior basolateral nuclei, and BMAp. By contrast with the dense labeling of pallial structures, the major subpallial nuclei were lightly labeled. These included the posterodorsal (MePD) and posteroventral (MePV) divisions of Me, CeL, CeC, and the intercalated group. As rostrally, although light, labeling was more pronounced in CeM than in CeL or CeC. Other subpallial nuclei, ASt and STIA, were moderately labeled. Proceeding caudally (Fig. 8B), labeled fibers continued to densely innervate pallial structures, particularly BLA. Labeling was very pronounced in BLAa and BLAp as well in the dorsolateral part of La (LaDL). By comparison, the ventrolateral (LaVL) and ventromedial (LaVM) divisions of La were moderately labeled, as were BLAv, BMAp and PLCo. For subpallial structures, ASt and STIA were moderately labeled, whereas MePD and MePV, all divisions of Ce, and the intercalated nuclei were lightly labeled. Further posteriorly (Fig. 9B), labeling remained dense in the basal complex, heaviest in BLAa and BLAp, slightly less so in LaDL, and strong but somewhat weaker still in BLAv, BMAp and medial (LaVM) and lateral (LaVL) divisions of La. As with the anterior cortical nuclei, labeling was denser in superficial than deep layers of PLCo. Subpallial structures, including STIA, MePD and MePV, were lightly to moderately labeled. Similar to the central and medial nuclei (see above), labeling gradual thinned in STIA from rostral to caudal levels. At the posterior amygdala (Fig. 10B), BLAp was densely (or massively) labeled -- clearly differentiating it from surrounding structures. Labeling was also pronounced in other deep pallial structures including La (LaDL, LaVM, LaVL), BMAp, and within the anterolateral amygdalohippocampal transition area (AHiAL). Of cortical pallial structures, the posteromedial cortical nucleus (PMCo), the PLCo, and the amygdalopiriform transition area (APir) were moderately to densely labeled. Unlike the laminar differences in labeling of PMCo and PLCo, labeled fibers spread quite uniformly throughout APir. At the caudal aspect of the amygdala (Fig. 11B), deep and cortical pallial nuclei contained moderate to dense concentrations of labeled fibers, with labeling heaviest in medial aspects of BLAp and BMAp. While less pronounced, La was also strongly labeled with a dorsoventral gradient such that LaDL was more heavily labeled than either LaVM or LaVL. The medial (AHiPM) and lateral (AHiPL) nuclei of the amygdalohippocampal transition area were moderately labeled; AHiPM slightly more heavily than AHiPL. PMCo and APir, contained moderate numbers of labeled fibers, with labeling heaviest in layer 1 of PMCo. Bed nucleus of the stria terminalis (BST) As indicated (see Introduction), the subpallial extended amygdala consists of two components, BST and the sublenticular extended amygdala (SLEA). The BST contains three subdivisions, the main one being BST proper. The BST (proper), which at rostral levels straddles the anterior commissure (ac), is composed of several subnuclei which, as will be described, contain differing concentrations of 5-HT fibers. Figures 12–15 depict the pattern of distribution of 5-HT fibers rostrocaudally throughout BST, and consist of a series of Nissl-stained sections (Figs. 12A–15A) showing the locations of nuclei of BST and corresponding SERT immuno-stained sections (Figs. 12B–15B) depicting the patterns of labeling at these levels of BST. At the rostral pole of BST (Fig. 12B), labeled fibers spread quite homogenously throughout BST distributing moderately to three subnuclei, the anterior medial (STMA), the lateral intermediate (STLI) and the lateral ventral (STLV) nuclei, and sparsely to the lateral dorsal nucleus (STLD). A similar pattern was observed further caudally (Fig. 13B). Specifically, with exception of STLD which, as rostrally, was sparsely labeled, each of the subnuclei surrounding the anterior commissure were moderately labeled. The dorsal subnuclei, including STMA, STLI, and lateral posterior (STLP) nuclei, were slightly more densely labeled than the ventral group which included the medial ventral nucleus (STMV), fusiform nucleus (Fu) and the parastrial nucleus (PS). The pattern of labeling further caudally in BST (Figs. 14B, 15B) largely mirrored that at rostral levels. As shown (Fig. 14B), the STLD was lightly labeled, whereas neighboring regions, medially and ventrally, were moderately labeled. They included STMA, STMV, STLV, STLP, STLJ, PS, and Fu. Of these nuclei, labeling was heaviest in FU and lightest in STMA. At the caudal extent of BST (Fig. 15B) labeling was stronger ventrally/ventrolaterally than dorsomedially. The medal part of the medial posterior nucleus (STMPM) was lightly labeled, whereas adjacent nuclei dorsally and ventrolaterally were moderately labeled. This included the dorsal subgroup (STD), the intermediate (STMPI) and lateral parts (STMPL) of the medial posterior nucleus, the STLP, the lateral intermediate nucleus (STLI) and the bed nucleus of the anterior commissure (BAC). DISCUSSION The present report describes the pattern of distribution of serotonergic (5-HT) fibers to the amygdala (proper) and to the extended amygdala – or the bed nucleus of the stria terminalis complex (BST). Amygdala Overall, serotonergic fibers distribute densely but differentially throughout nuclei of the amygdala. SERT+ fibers heavily innervate most of the deep and cortical pallial nuclei of amygdala which contrasts with minimal innervation of subpallial nuclei. Of the deep pallial structures, the nuclei of the basal complex, consisting of the lateral, basolateral and basomedial nuclei (and their subregions) contained the densest concentration of 5-HT fibers, particularly posterior aspects of the basolateral complex. In general, cortical pallial structures contained fewer SERT+ fibers than the deep pallial nuclei. Of the subpallial amygdalar nuclei, there is a rostrocaudal gradient, whereby SERT+ fibers distribute more heavily to anterior than to posterior divisions of subpallial structures. 5-HT fibers terminate moderately in the anterior dorsal and anterior ventral nuclei of the anterior amygdala, in the amygdalostriatal transition area and in the anterodorsal and anteroventral parts of the medial nucleus; lightly in the posterodorsal and posteroventral sectors of Me, and in the intercalated nuclei, and generally sparsely in the central nucleus. Bed nucleus of the stria terminalis (BST) The subpallial extended amygdala consists of two components: (1) BST including the intra-amygdalar (STIA) and BST (proper) divisions; and (2) the sublenticular extended amygdala consisting of SI and IPAC (Olucha-Bordonau et al., 2015). The BST proper contained a moderate collection of SERT+ fibers, which on the whole was of lesser magnitude than those to the amygdala. With the exception of a light innervation of lateral dorsal division of BST (STLD) and the medial posterior nucleus (STMPM), 5-HT fibers terminate moderately within structures of BST. These included: STMA, STMV, STLI, STLV, STLP, Fu and PS, rostrally, and the STD, STMPI, STMPL, STLI, and the lateral STLP, caudally. The anterior pole of IPAC contained moderate numbers of 5-HT fibers, as did the STIA, lying ventral to Ce and co-extensive with it. Comparison with previous examinations of serotonergic innervation of the amygdala As discussed, no previous report has described the overall pattern of distribution of 5-HT fibers to the amygdala (or to BST) in the rat. Nonetheless, some studies have examined the 5-HT innervation of select nuclei of the amygdala, concentrating on the basal group and the central nucleus (Sur et al., 1996; Commons et al., 2003; Muller et al., 2007; Smith and Porrino, 2008; Bonn et al., 2013). With minor differences among reports, the common finding was that 5-HT fibers distribute densely to the basolateral nucleus, strongly but less heavily to the lateral nucleus, and sparsely to the central nucleus. The 5-HT innervation of the basomedial and medial nuclei was intermediate to that of BLA/La and Ce. The present results are generally consistent with these findings with the exception that we observed: (1) denser labeling of the posterior than the anterior BLA; (2) stronger labeling of the medial than lateral or capsular divisions of Ce; and (3) light labeling of the posterodorsal and posteroventral divisions of the medial nucleus. Correspondence between 5-HT innervation of the amygdala/BST and midbrain raphe projections to the amygdala/BST The serotonergic input to the forebrain originates almost entirely from midbrain raphe nuclei; namely, the dorsal (DR) and median raphe (MR) nuclei and the B9 group (Vertes and Martin, 1988; Vertes, 1991; Vertes and Crane, 1997; Vertes et al., 1999; Morin and Meyer-Bernstein, 1999; Vertes and Linley, 2007, 2008; Muzerelle et al., 2015). Early reports in the rat (Vertes, 1991; Vertes et. al., 1999) and hamster (Morin and Meyer-Bernstein, 1999), using anterograde tracers, described pronounced DR projections (mainly from the rostral DR) to the amygdala -- with minor projections from MR. In accord with present findings, the DR was shown to strongly target the lateral, basolateral and cortical nuclei of amygdala and lightly the medial nucleus (Vertes, 1991; Morin and Meyer-Bernstein, 1999; Muller et al., 2007; Bonn et al., 2013). Unlike, the present demonstration of sparse 5-HT labeling of the central nucleus, Bienkowski and Rinaman (2013) described moderate DR-Ce projections. Some of the DR input to Ce, however, undoubtedly originated from non-serotonergic DR neurons. Similar to the amygdala (proper), midbrain raphe projections to BST predominately arise from DR (or the rostral DR) and minimally from MR (Vertes, 1991; Vertes et al., 1999; Morin and Meyer-Bernstein, 1999; Vertes and Linley, 2008; Muzerelle et al., 2015). Consistent with present findings, DR fibers were shown to distribute moderately, and quite uniformly, throughout BST, with a preference for the anterior ventrolateral sector of BST – which includes the fusiform, parastrial, and medial and lateral ventral nuclei of BST (Vertes, 1991; Morin and Meyer-Bernstein, 1999; Vertes and Linley, 2008; Bienkowski and Rinaman, 2013; Muzerelle et al., 2015). Functional role of serotonergic input to the amygdala -- with a focus on the basolateral nucleus As discussed, both the amygdala and the serotonergic system serve direct, and likely complementary, roles in emotional behavior, prominently including fear and anxiety (Davis, 2000). For instance, SSRIs have been described as the “gold standard” in the treatment of anxiety disorders (Inoue et al., 2011; Burghardt and Bauer, 2013), possibly in large part involving their actions on the amygdala. The detailed circuitry of the amygdala responsible for the acquisition and expression of fear/anxiety has been well described (see Duvarci and Pare, 2014; Herry and Johansen, 2014; Tovote et al. 2015). The primary cell groups of the amygdala involved in fear/anxiety are the basal group (La, BLA and BMA), the central nucleus (or CeL and CeM) and the intercalated nuclei. For the most part, information flows unidirectionally from La to BLA to CeL and then to CeM, such that La is the principal afferent node and CeM the output site in fear/anxiety. In general, associations between neutral sensory stimuli and aversive events (tone-foot shock pairings) are made in La, transferred with modifications through BLA/BMA and the intercalated nuclei to CeL and then to CeM to affect behavior. Accordingly, the BLA (and BMA) is pivotally positioned to modify information transferred from La to CeM. This coupled with the present demonstration that the BLA receives a dense 5-HT input, suggests that BLA may be a primary (amygdalar) target for 5-HT actions on fear/anxiety (Amano et al., 2011). Consistent with this, serotonin has been shown to exert significant modulatory effects at the BLA in anxiety-like behaviors (Asan et al., 2013; Burghardt and Bauer, 2013; Bauer, 2015). Anxiety-producing stimuli/conditions release serotonin to the amygdala/BLA (Hale et al., 2012; Fox and Lowry, 2013) which is initially (or acutely) anxiogenic, but with time (or chronically) becomes anxiolytic (Burghardt et al., 2004; Burghardt and Bauer, 2013). In a parallel manner, various serotonin reuptake inhibitors (SSRIs) initially elicit fear/anxiety-like behaviors (Burghardt et al., 2004, 2007; Grillion et al., 2007; Ravinder et al., 2011), but with long term use become anxiolytic (Zhang et al., 2000; Li et al., 2001; Burghardt et al., 2004; Hashimoto et al., 2009; Deschaux et al., 2011; Inoue et al., 2011). While the effects of SSRIs on fear/anxiety are not limited to the amygdala or BLA, the BLA appears to be a principal site for the early anxiogenic and later anxiolytic effects of SSRIs (Inoue et al. 2004, 2011; Kitaichi et al. 2014). In addition, the acute anxiogenic actions of SSRIs reportedly involve 5-HT2C receptors (mainly of BLA), whereas the chronic anxiolytic effects involve 5-HT1A receptors of BLA (Burghardt et al., 2007; Christianson et al., 2010; Vicente and Zangrossi, 2012, 2014; de Andrade Strauss et al., 2013). For instance, regarding 5-HT2C receptors, Vicente and Zangrossi (2012) showed that: (1) intra-BLA injections of the 5-HT2C receptor agonist, MK-212, and the 5-HT2C receptor antagonist, SB-242084, enhanced or supressed anxiety-like behaviors, respectively, on the elevated T-maze; and (2) SB-242084 blocked the (acute) anxiogenic effect elicited by the systemic administration of the antidepressants, imipramine or fluoxetine. Vicente and Zangrossi (2014) subsequently confirmed that MK-212 injected into BLA was anxiogenic, and further showed that anxiogenic actions were attenuated with chronic treatment with imipramine or fluoxetine. Taken together the findings support the view that the acute anxiogenic, as well as the chronic anxiolytic, effects of antidepressants are, in part, mediated through 5-HT2C receptors of BLA. Specifically, a 5-HT2C mediated activation of BLA neurons is anxiogenic, whereas the gradual (chronic) desensitization of 5-HT2C receptors of these cells contributes to anxiolysis. By contrast with the anxiogenic effects of 5-HT2C receptors, several studies have shown that 5-HT1A receptors produce anxiolytic actions at the BLA (Zangrossi et al., 1999; Li et al., 2012; de Andrade Strauss et al., 2013; Vicente and Zangrossi, 2014). For example, de Andrade Strauss et al. (2013) showed that BLA injections of the 5HT1A receptor agonist, 8-OH-DPAT, suppressed anxiety-like behavior on three tests of anxiety – with effects blocked by the prior administration of the 5-HT1A antagonist, WAY-100635. Moreover, Vicente and Zangrossi (2014) reported that the anxiogenic actions produced by intra-BLA injections of the 5-HT2C receptor agonist, MK-212, were blocked with chronic antidepressant administration, but were re-instated with intra-BLA injections of the 5-HT1A antagonist, WAY-100635. Accordingly, the authors (Vincente and Zangrossi, 2014) concluded that “both a reduction in 5-HT2C-R- and a facilitation of 5-HT1A-R mediated neurotransmission in the BLA are involved in the anxiolytic effect of antidepressant drugs”. Consistent with this, Li et al. (2012), using recombinant adenovirus-induced alterations in 5-HT receptor expression in the amygdala, demonstrated that decreases in 5-HT1A, or increases in the 5-HT2C, receptor expression produced anxiogenic actions on two tests of anxiety-like behavior in mice. While the precise mechanism(s) whereby 5-HT1A and 5-HT2C receptors of BLA cells affect anxiety-like states remains to be fully determined, the process understandably involves interneurons and principal cells. With some exceptions (see below) increases in BLA pyramidal cell activity appears to be anxiogenic, decreases anxiolytic. With respect to 5-HT1A receptors, an early report in anesthetized rats (Stein et al., 2000) showed that the 5-HT1A agonist, 8-OH-DPAT, suppressed the activity of most pyramidal cells (PCs) of BLA. In accord with this, Cheng et al. (1998) had shown that serotonin, acting via 5-HT1A receptors, suppressed the activity of principal BLA cells and reduced a depolarization-evoked influx of calcium into these cells. Whereas the effects of serotonin on 5-HT1A-receptor-containing cells of BLA are fairly straightforward, specific 5-HT actions on 5-HT2C receptor cells of BLA are not fully resolved. For instance, Rainnie (1999) demonstrated in the slice preparation that the application of 5-HT or 5-HT2 agonists, to BLA activated GABAergic interneurons -- with the consequent inhibition of pyramidal cells. While this finding appears inconsistent with a proposed role for 5-HT2C receptors in anxiogenesis (i.e., the activation, not the suppression, of principal BLA activity is thought to be anxiogenic) it should be noted that: (1) a non-specific 5-HT2 agonist was used (α-methyl-5-HT); results may differ with a more specific 5-HT2c agonist; and (2) high concentrations or the prolonged application of 5-HT suppressed the excitatory action of 5-HT on BLA interneurons -- and hence their inhibitory effect on PCs. Further, in contrast to these results, Stein et al., (2000) showed, in intact rats, that the 5-HT2C agonist, DOI, increased the activity of virtually all principal BLA neurons, while Reznikov et al. (2008) reported that restraint stress produced elevated levels of c-fos expression in the vast majority of PCs of BLA. Taken together, the foregoing findings would indicate that 5-HT2/2C receptor cell activation excites both interneurons and principal cells of BLA, but as noted by Campbell and Merchant (2003), the likely net effect would be an “augmentation of BLA output which could underlie the anxiogenic responses of 5-HT2C agonists at behaviourally relevant doses.” Whereas a reduction of principal cell activity of BLA is generally thought to contribute to anxiolysis, Tye et al. (2011) recently identified a subset of BLA cells whose activation suppressed anxiety-like behavior in mice. Specifically, using optogenetic techniques, they demonstrated that stimulation of a select population of BLA cells projecting to CeL produced anxiolytic actions, presumably via a BLA-induced activation of GABAergic CeL cells and the consequent inhibition of CeM output neurons. Interestingly, however, the overall (or non-specific) activation of BLA neurons was found to be anxiogenic which is consistent with previously-discussed reports (Vicente and Zangrossi, 2012; Zangrossi and Graeff, 2014). Finally, based on the pronounced 5-HT input to the basal nuclei of the amygdala, we focused on the role of 5-HT afferents to the BLA complex in fear/anxiety, but it is certainly well recognized that the amygdala participates in wide range of emotional behaviors, many of which have been linked to particular subnuclei. For instance, Ferguson et al. (2001) showed that oxytocin acts on the medial nucleus of the amygdala to facilitate social recognition in mice. Possibly directly related to this, it has recently been found that selective serotonin reuptake inhibitors (SSRIs) taken during the second or third trimester, for the treatment for depression, increases the risk for autism spectrum disorders (Boukhris et al., 2015). Although we showed 5-HT fibers distribute moderately to the medial nucleus, it is nevertheless possible that a SSRI-induced developmental disruption (or reorganization) of serotonergic projections to the medial nucleus could contribute to the social deficits associated with autism spectrum disorders. Functional role of serotonergic input to the BST Similar to the amygdala proper, the BST also plays a prominent role in affective behaviors, prominently stress and anxiety (Davis et al., 2010). In general, serotonin exerts anxiolytic effects at the BST (Levita et al., 2004; Gomes et al., 2011, 2012). This partly involves a 5-HT-mediated suppression of excitatory glutamatergic and corticotrophin releasing factor (CRF) inputs to the BST as well as on intrinsic CRF-mediated excitation at BST. For instance, CRF infused into the anterolateral BST elicits anxiety-like behaviors including increased startle and reduced exploratory behavior in the elevated plus maze (Lee and Davis, 1997; Sahuque et al., 2006). Rainnie and colleagues (Hammack et al. 2009; Daniel and Rainnie, 2015) recently proposed a model for the interactive roles of the DR and BST in stress/anxiety. Specifically, stressors produce an increase in the release of CRF to both the DR and BST which initially is anxiogenic but with time becomes anxiolytic. In effect, at low doses CRF binds to high affinity CRF1 receptors of the DR to inhibit the firing of 5-HT DR cells, whereas at higher doses CRF also binds to CRF2 receptors to increase the discharge of DR neurons – and thereby releases 5-HT to the BST in the suppression of anxiety (Hammack et al., 2003). Three types of neurons have been identified in the lateral BST of rodents: type I–III cells. Type I and III neurons discharge at regular rates, with the threshold to activation higher for type III than type I cells, whereas type II cells display a bursting pattern of discharge which has been associated with the release of various peptides from the cells (Egli and Winder, 2003; Hammack et al., 2007). Type I neurons constitute 29%, type II neurons 55%, and type III neurons 16% of the population of lateral BST neurons (Hammack et al., 2007). The BST contains a rich array of 5-HT receptors which differentially populate the three cell types – and thus determine their roles in stress/anxiety (Daniel and Rainnie, 2015). The main 5-HT receptor types are 5-HT1A, 5-HT1B, 5-HT2C, and 5-HT7 receptors – and most abundantly 5-HT1A receptors. The 5-HT1A receptor is expressed on all three types of BST cells (Hazra et al., 2012). Since 5-HT1A receptor activation hyperpolarizes neurons, the predominant (or net) effect of 5-HT applied to BST is a hyperpolarization of BST neurons (Guo et al., 2009; Hazra et al., 2012) – and anxiolytic actions at BST (Levita et al., 2004; Gomes et al., 2011, 2012). In contrast to 5-HT1A receptors, 5-HT depolarizes 5-HT2C and 5-HT7 receptor-containing neurons of the BST. 5-HT2C receptors are almost entirely located on type III cells, which are mainly CRF neurons, and reportedly form a feed forward loop with the DR in the elicitation of anxiety-like behaviors. Specifically, a DR release of 5-HT to the BST activates type III cells which, in turn, releases CRF to the DR, drives DR neurons, and thus maintains the cycle. However, the release of 5-HT to BST also activates 5-HT7 receptor containing BST neurons, present on type I and type II neurons which suppress the activity of type III CRF neurons to dampen anxiety-like states. In sum, the actions of 5-HT at the BST are complex with ultimate effects on stress/anxiety dependent on the relative recruitment of various types of 5-HT receptors -- favoring 5-HT1A receptors and anxiolysis. Grant sponsor: NIMH, grant number: MH099590 Abbreviations 5-HT serotonin AAA anterior amygdaloid area AAD anterior amygdaloid area, dorsal part AAV anterior amygdaloid area, ventral part ac anterior commissure AC nucleus accumbens ACo anterior cortical nucleus of amygdala AHi amygdalo-hippocampal area AHiAL amygdalohippocampal transition area, anterolateral part AHiPL amygdalohippocampal transition area, posterolateral part AHiPM amygdalohippocampal transition area, posteromedial division APir amygdalopiriform transition area ASt amygdalostriatal transition area BAC bed nucleus of anterior commissure BAOT bed nucleus of accessory olfactory tract BLA basolateral amygdala BLAa basolateral nucleus of amygdala, anterior part BLAp basolateral nucleus of amygdala, posterior part BLAv basolateral nucleus of amygdala, ventral part BMA basomedial amygdala BMAa basomedial nucleus of amygdala, anterior part BMAp basomedial nucleus of amygdala, posterior part BST bed nucleus of the stria terminalis Ce central nucleus of amygdala CeC central nucleus of amygdala, capsular division CeL central nucleus of amygdala, lateral division CeM central nucleus of amygdala, medial division CRF corticotrophin releasing factor CxA cortical amygdala transition zone DEn dorsal endopiriform nucleus DR dorsal raphe nucleus EA extended amygdala Fu fusiform nucleus of BST fx fornix HDB nucleus of horizontal limb of diagonal band I intercalated nuclei of amygdala IPAC interstitial nucleus of posterior limb of anterior commissure La lateral nucleus of amygdala LaDL lateral nucleus of amygdala, dorsolateral part LaVL lateral nucleus of amygdala, ventrolateral part LaVM lateral nucleus of amygdala, ventromedial part LGP globus pallidus, lateral division lot lateral olfactory tract LOT nucleus of lateral olfactory tract LPO lateral preoptic area LSS lateral stripe of striatum LSV lateral septal nucleus, ventral part Me medial nucleus of amygdala MeAD medial nucleus of amygdala, anterodorsal part MeAV medial nucleus of amygdala, anteroventral part MePD medial nucleus of amygdala, posterodorsal part MePV medial nucleus of amygdala, posteroventral part mPFC medial prefrontal cortex MR median raphe nucleus OT olfactory tubercle Pir piriform cortex PLCo posterolateral cortical nucleus of amygdala PMCo posteromedial cortical nucleus of amygdala PS parastrial nucleus of BST Pu putamen PVA paraventricular nucleus of thalamus, anterior part SERT serotonin transporter protein SI substantia innominata SIB substantia innominata, basal division SLN supralemniscal nucleus sm stria medullaris SO supraoptic nucleus SLEA sublenticular extended amygdala STD dorsal division of BST STIA intra-amygdala division of BST STLD lateral dorsal division of BST STLI lateral intermediate division of BST STLJ lateral juxtacapsular division of BST STLP lateral posterior division of BST STLV lateral ventral division of BST STMA medial anterior division of BST STMPI medial posterior intermediate division of BST STPML medial posterior lateral division of BST STMPM medial posterior medial division of BST STMV medial ventral division of BST STS supracapsular extended amygdala VEn ventral endopiriform nucleus VP ventral pallidum Figure 1 A: Low magnification Nissl-stained transverse section at the transition between the amygdala (proper) and the extended amygdala showing the locations of nuclei at this level. B: Pattern of distribution of 5-HT fibers to nuclei of the amygdala at the same level – or to sublenticular region of the amygdala. See list for abbreviations. Scale bar for A = 1.2 mm; for B = 500 μm. Figure 2 A: Low magnification Nissl-stained transverse section through the anterior amygdala showing the locations of nuclei at the same level. B: Pattern of the distribution of 5-HT fibers to nuclei of the amygdala at this level (B). See list for abbreviations. Scale bar for A = 1.3 mm; for B = 500 μm. Figure 3 A: Low magnification Nissl-stained transverse section through the anterior amygdala showing the locations of nuclei at this level. B: Pattern of distribution of 5-HT fibers to nuclei of the amygdala at the same level. See list for abbreviations. Scale bar for A = 800 μm; for B = 500 μm. Figure 4 A: Low magnification Nissl-stained transverse section through the anterior amygdala showing the locations of nuclei at this level. B: Pattern of distribution of 5-HT fibers to nuclei of the amygdala at the same level. Note the dense labeling of nuclei of the basolateral complex (La, BMA, BLA) and the light labeling of divisions of the central nucleus (Ce). See list for abbreviations. Scale bar for A = 1 mm; for B = 500 μm. Figure 5 A: Low magnification Nissl-stained transverse section through the mid-portion of the amygdala showing the locations of nuclei at this level. B: Pattern of distribution of 5-HT fibers to nuclei of the amygdala at the same level. Note the dense labeling of nuclei of the basolateral complex (La, BMA, BLA), particularly BLA, and the sparse labeling of divisions of the central nucleus (Ce). See list for abbreviations. Scale bar for A = 1 mm; for B = 500 μm. Figure 6 A: Low magnification Nissl-stained transverse section through the mid-portion of the amygdala showing the locations of nuclei at this level. B: Pattern of distribution of 5-HT fibers to nuclei of the amygdala at the same level. Note the dense labeling of nuclei of the basolateral complex (La, BMA, BLA), particularly BLA, and the sparse labeling of divisions of the central nucleus (Ce). See list for abbreviations. Scale bar for A = 1 mm; for B = 500 μm. Figure 7 A: Low magnification Nissl-stained transverse section through the mid-portion of the amygdala showing the locations of nuclei at this level. B: Pattern of distribution of 5-HT fibers to nuclei of the amygdala at the same level. Note the dense labeling of nuclei of the basolateral complex (La, BMA, BLA), particularly BLA, and the sparse to light labeling of divisions of the central (Ce) and medial nucleus (Me). See list for abbreviations. Scale bar for A = 835 μm; for B = 500 μm. Figure 8 A: Low magnification Nissl-stained transverse section through the posterior amygdala showing the locations of nuclei at this level. B: Pattern of distribution of 5-HT fibers to nuclei of the amygdala at the same level. Note the dense labeling of nuclei of the basolateral complex (La, BMA, BLA), particularly BLA, and the sparse to light labeling of divisions of the central (Ce) and medial nuclei (Me). See list for abbreviations. Scale bar for A = 700 μm; for B = 500 μm. Figure 9 A: Low magnification Nissl-stained transverse section through the posterior amygdala showing the locations of nuclei at this level. B: Pattern of distribution of 5-HT fibers to nuclei of the amygdala at the same level. Note the dense labeling of nuclei of the basolateral complex (La, BMA, BLA), particularly BLA, and the light labeling of the medial nucleus. See list for abbreviations. Scale bar for A = 670 μm; for B = 500 μm. Figure 10 A: Low magnification Nissl-stained transverse section through the posterior amygdala showing the locations of nuclei at this level. B: Pattern of distribution of 5-HT fibers to nuclei of the amygdala at the same level. Note the dense labeling of the posterior part of the basolateral nucleus, setting it apart from surrounding structures. See list for abbreviations. Scale bar for A = 1 mm; for B = 500 μm. Figure 11 A: Low magnification Nissl-stained transverse section through the posterior amygdala showing the locations of nuclei at this level. B: Pattern of distribution of 5-HT fibers to nuclei of the amygdala at the same level. Note the dense labeling of the posterior part of the basomedial and basolateral nuclei. See list for abbreviations. Scale bar for A = 750 μm; for B = 500 μm. Figure 12 A: Low magnification Nissl-stained transverse section through the anterior part of the extended amygdala showing the locations of nuclei at this level. B: Pattern of distribution of 5-HT fibers to nuclei of the extended amygdala at the same level. See list for abbreviations. Scale bar for A = 380 μm; for B = 250 μm. Figure 13 A: Low magnification Nissl-stained transverse section through the anterior part of the extended amygdala showing the locations of nuclei at this level. B: Pattern of distribution of 5-HT fibers to nuclei of the extended amygdala at the same level. See list for abbreviations. Scale bar for A = 360 μm; for B = 250 μm. Figure 14 A: Low magnification Nissl-stained transverse section through the posterior part of the extended amygdala showing the locations of nuclei at this level. B: Pattern of distribution of 5-HT fibers to nuclei of the extended amygdala at the same level. See list for abbreviations. Scale bar for A = 375 μm; for B = 250 μm. Figure 15 A: Low magnification Nissl-stained transverse section through the posterior part of the extended amygdala showing the locations of nuclei at this level. B: Pattern of distribution of 5-HT fibers to nuclei of the extended amygdala at the same level. See list for abbreviations. Scale bar for A = 275 μm; B = 250 μm. CONFLICT OF INTEREST STATEMENT The authors declare no conflict of interest. ROLE OF THE AUTHORS SBL collected the data, SBL and FO-B served major roles in the preparation of the figures, and RPV served a major role in writing the manuscript. All authors, however, were involved in all phases of the planning and execution of the research. LITERATURE CITED Alheid GF Extended amygdala and basal forebrain 2003 Ann NY Acad Sci 985 185 205 12724159 Amano T Duvarci S Popa D Paré D 2011 The fear circuit revisited: contributions of the basal amygdala nuclei to conditioned fear J Neurosci 31 15481 15489 22031894 Asan E Yilmazer-Hanke DM Eliava M Hantsch M Lesch KP Schmitt A 2005 The corticotropin-releasing factor (CRF)-system and monoaminergic afferents in the central amygdala: investigations in different mouse strains and comparison with the rat Neuroscience 131 953 967 15749348 Asan E Steinke M Lesch KP 2013 Serotonergic innervation of the amygdala: targets, receptors, and implications for stress and anxiety Histochem Cell Biol 139 785 813 23494464 Bauer EP 2015 Serotonin in fear conditioning processes Behav Brain Res 277 68 77 25078294 Bauman MD Amaral DG 2005 The distribution of serotonergic fibers in the macaque monkey amygdala: an immunohistochemical study using antisera to 5-hydroxytryptamine Neuroscience 136 193 203 16182456 Bienkowski MS Rinaman L 2013 Common and distinct neural inputs to the medial central nucleus of the amygdala and anterior ventrolateral bed nucleus of stria terminalis in rats Brain Struct Funct 218 187 208 22362201 Bonn M Schmitt A Lesch KP Van Bockstaele EJ Asan E 2013 Serotonergic innervation and serotonin receptor expression of NPY-producing neurons in the rat lateral and basolateral amygdaloid nuclei Brain Struct Funct 218 421 435 22527118 Boukhris T Sheehy O Mottron L Bérard A 2016 Antidepressant use during pregnancy and the risk of autism spectrum disorder in children JAMA Pediatr 170 117 124 26660917 Bupesh M Abellan A Medina L 2011 Genetic and experimental evidence supports the continuum of the central extended amygdala and a mutiple embryonic origin of its principal neurons J Comp Neurol 519 3507 3531 21800302 Burghardt NS Bauer EP 2013 Acute and chronic effects of selective serotonin reuptake inhibitor treatment on fear conditioning: implications for underlying fear circuits Neuroscience 247 253 272 23732229 Burghardt NS Sullivan GM McEwen BS Gorman JM LeDoux JE 2004 The selective serotonin reuptake inhibitor citalopram increases fear after acute treatment but reduces fear with chronic treatment: a comparison with tianeptine Biol Psychiatry 55 1171 1178 15184036 Burghardt NS Bush DE McEwen BS LeDoux JE 2007 Acute selective serotonin reuptake inhibitors increase conditioned fear expression: blockade with a 5-HT(2C) receptor antagonist Biol Psychiatry 62 1111 1118 17524369 Campbell BM Merchant KM 2003 Serotonin 2C receptors within the basolateral amygdala induce acute fear-like responses in an open-field environment Brain Res 993 1 9 14642825 Cheng LL Wang SJ Gean PW 1998 Serotonin depresses excitatory synaptic transmission and depolarization-evoked Ca2+ influx in rat basolateral amygdala via 5-HT1A receptors Eur J Neurosci 10 2163 2172 9753102 Christianson JP Ragole T Amat J Greenwood BN Strong PV Paul ED Fleshner M Watkins LR Maier SF 2010 5-hydroxytryptamine 2C receptors in the basolateral amygdala are involved in the expression of anxiety after uncontrollable traumatic stress Biol Psychiatry 67 339 345 19914601 Commons KG Connolley KR Valentino RJ 2003 A neurochemically distinct dorsal raphe-limbic circuit with a potential role in affective disorders Neuropsychopharmacology 28 206 215 12589373 Daniel SE Rainnie DG 2015 Stress modulation of opposing circuits in the bed nucleus of the stria terminalis Neuropsychopharmacology in press Davis M 2000 Role of the amygdala in conditioned and unconditioned fear and anxiety Aggleton JP The amygdala 2 Oxford, UK Oxford University Press 213 287 Davis M Walker DL Miles L Grillon C 2010 Phasic vs sustained fear in rats and humans: role of the extended amygdala in fear vs anxiety Neuropsychopharmacology 35 105 135 19693004 de Andrade Strauss CV Vicente MA Zangrossi H 2013 Activation of 5-HT1A receptors in the rat basolateral amygdala induces both anxiolytic and antipanic-like effects Behav Brain Res 246 103 110 23499701 Deschaux O Spennato G Moreau JL Garcia R 2011 Chronic treatment with fluoxetine prevents the return of extinguished auditory-cued conditioned fear Psychopharmacology (Berl) 215 231 237 21181120 Dong HW Petrovich GD Swanson LW 2001 Topography of projections from amygdala to bed nuclei of the stria terminalis Brain Res Rev 38 192 246 11750933 Duvarci S Pare D 2014 Amygdala microcircuits controlling learned fear Neuron 82 966 980 24908482 Egli RE Winder DG 2003 Dorsal and ventral distribution of excitable and synaptic properties of neurons of the bed nucleus of the stria terminalis J Neurophysiol 90 405 414 12649311 Ferguson JN Aldag JM Insel TR Young LJ 2001 Oxytocin in the medial amygdala is essential for social recognition in the mouse J Neurosci 21 8278 8285 11588199 Fox JH Lowry CA 2013 Corticotropin-releasing factor-related peptides, serotonergic systems, and emotional behavior Front Neurosci 7 169 24065880 Freedman LJ Shi C 2001 Monoaminergic innervation of the macaque extended amygdala Neuroscience 104 1067 1084 11457591 García-López M Abellán A Legaz I Rubenstein JL Puelles L Medina L 2008 Histogenetic compartments of the mouse centromedial and extended amygdala based on gene expression patterns during development J Comp Neurol 506 46 74 17990271 Gomes FV Resstel LB Guimarães FS 2011 The anxiolytic-like effects of cannabidiol injected into the bed nucleus of the stria terminalis are mediated by 5-HT1A receptors Psychopharmacology (Berl) 213 465 473 20945065 Gomes FV Reis DG Alves FH Corrêa FM Guimarães FS Resstel LB 2012 Cannabidiol injected into the bed nucleus of the stria terminalis reduces the expression of contextual fear conditioning via 5-HT1A receptors J Psychopharmacol 2012 26 104 113 Grillon C Levenson J Pine DS 2007 A single dose of the selective serotonin reuptake inhibitor citalopram exacerbates anxiety in humans: a fear-potentiated startle study Neuropsychopharmacology 32 225 231 16971899 Guo JD Hammack SE Hazra R Levita L Rainnie DG 2009 Bi-directional modulation of bed nucleus of stria terminalis neurons by 5-HT: molecular expression and functional properties of excitatory 5-HT receptor subtypes Neuroscience 164 1776 1793 19778589 Hale MW Shekhar A Lowry CA 2012 Stress-related serotonergic systems: implications for symptomatology of anxiety and affective disorders Cell Mol Neurobiol 32 695 708 22484834 Hammack SE Schmid MJ LoPresti ML Der-Avakian A Pellymounter MA Foster AC Watkins LR Maier SF 2003 Corticotropin releasing hormone type 2 receptors in the dorsal raphe nucleus mediate the behavioral consequences of uncontrollable stress J Neurosci 23 1019 1025 12574432 Hammack SE Mania I Rainnie DG 2007 Differential expression of intrinsic membrane currents in defined cell types of the anterolateral bed nucleus of the stria terminalis J Neurophysiol 98 638 656 17537902 Hammack SE Guo JD Hazra R Dabrowska J Myers KM Rainnie DG 2009 The response of neurons in the bed nucleus of the stria terminalis to serotonin: implications for anxiety Prog Neuropsychopharmacol Biol Psychiatry 33 1309 1320 19467288 Hashimoto S Inoue T Muraki I Koyama T 2009 Effects of acute citalopram on the expression of conditioned freezing in naive versus chronic citalopram-treated rats Prog Neuropsychopharmacol Biol Psychiatry 33 113 117 19010374 Hayes DJ Greenshaw AJ 2011 5-HT receptors and reward-related behaviour: a review Neurosci Biobehav Rev 35 1419 1449 21402098 Hazra R Guo JD Dabrowska J Rainnie DG 2012 Differential distribution of serotonin receptor subtypes in BNST(ALG) neurons: modulation by unpredictable shock stress Neuroscience 225 9 21 22922122 Herry C Johansen JP 2014 Encoding of fear learning and memory in distributed neuronal circuits Nat Neurosci 17 1644 1654 25413091 Hertel N Redies C Medina L 2012 Cadherin expression delineates the divisions of the postnatal and adult mouse amygdala J Comp Neurol 520 3982 4012 22592879 Holmes A 2008 Genetic variation in cortico-amygdala serotonin function and risk for stress-related disease Neurosci Biobehav Rev 32 1293 1314 18439676 Inoue T Li XB Abekawa T Kitaichi Y Izumi T Nakagawa S Koyama T 2004 Selective serotonin reuptake inhibitor reduces conditioned fear through its effect in the amygdala Eur J Pharmacol 497 311 316 15336949 Inoue T Kitaichi Y Koyama T 2011 SSRIs and conditioned fear Prog Neuropsychopharmacol Biol Psychiatry 35 1810 1819 21925230 Kalia M 2005 Neurobiological basis of depression: an update Metabolism 54 5 Suppl 1 24 27 Kitaichi Y Inoue T Nakagawa S Omiya Y Song N An Y Chen C Kusumi I Koyama T 2014 Local infusion of citalopram into the basolateral amygdala decreased conditioned fear of rats through increasing extracellular serotonin levels Prog Neuropsychopharmacol Biol Psychiatry 54 216 222 24928686 LeDoux JE 2000 Emotion circuits in the brain Annu Rev Neurosci 23 155 184 10845062 LeDoux J 2003 The emotional brain, fear, and the amygdala Cell Mol Neurobiol 23 727 738 14514027 Lee Y Davis M 1997 Role of the hippocampus, the bed nucleus of the stria terminalis, and the amygdala in the excitatory effect of corticotropin-releasing hormone on the acoustic startle reflex J Neurosci 17 6434 6446 9236251 Levita L Hammack SE Mania I Li XY Davis M Rainnie DG 2004 5-hydroxytryptamine1A-like receptor activation in the bed nucleus of the stria terminalis: electrophysiological and behavioral studies Neuroscience 128 583 596 15381287 Li Q Luo T Jiang X Wang J 2012 Anxiolytic effects of 5-HT1A receptors and anxiogenic effects of 5-HT2C receptors in the amygdala of mice Neuropharmacology 62 474 484 21925519 Li XB Inoue T Hashimoto S Koyama T 2001 Effect of chronic administration of flesinoxan and fluvoxamine on freezing behavior induced by conditioned fear Eur J Pharmacol 425 43 50 11672573 Linley SB Hoover WB Vertes RP 2013 Pattern of distribution of serotonergic fibers to the orbitomedial and insular cortex in the rat J Chem Neuroanat 48–49 29 45 Lowry CA Johnson PL Hay-Schmidt A Mikkelsen J Shekhar A 2005 Modulation of anxiety circuits by serotonergic systems Stress 8 233 246 16423712 Lowry CA Hale MW Evans AK Heerkens J Staub DR Gasser PJ Shekhar A 2008 Serotonergic systems, anxiety, and affective disorder: focus on the dorsomedial part of the dorsal raphe nucleus Ann NY Acad Sci 1148 86 94 19120094 Martínez-García F Novejarque A Lanuza E 2008 Two interconnected functional systems in the amygdala of amniote vertebrates Brain Res Bull 75 206 213 18331872 Medina L Legaz I Gonzalez G De Castro F Rubenstein JL Puelles L 2004 Expression of Dbx1, neurogenin 2, semaphoring 5A, cadherin 8, and Emx1 distinguish ventral and lateral pallial histogenetic divisions in the developing mouse claustroamygdaloid complex J Comp Neurol 474 504 523 15174069 Morin LP Meyer-Bernstein EL 1999 The ascending serotonergic system in the hamster: comparison with projections of the dorsal and median raphe nuclei Neuroscience 91 81 105 10336062 Muller JF Mascagni F McDonald AJ 2007 Serotonin-immunoreactive axon terminals innervate pyramidal cells and interneurons in the rat basolateral amygdala J Comp Neurol 505 314 335 17879281 Muzerelle A Scotto-Lomassese S Bernard JF Soiza-Reilly M Gaspar P 2015 Conditional anterograde tracing reveals distinct targeting of individual serotonin cell groups (B5–B9) to the forebrain and brainstem Brain Struct Funct in press Nemeroff CB Bremner JD Foa EB Mayberg HS North CS Stein MB 2006 Posttraumatic stress disorder: a state-of-the-science review J Psychiatr Res 40 1 21 16242154 Olucha-Bordonau FE Fortes-Marco L Otero-García M Lanuza E Martínez-García F 2015 Amgydala: Structure and Function Paxinos G The rat nervous system 4 San Diego Academic Press 441 490 O’Rourke H Fudge JL 2006 Distribution of serotonin transporter labeled fibers in amygdaloid subregions: implications for mood disorders Biol Psychiatry 60 479 490 16414028 Parent A Descarries L Beaudet A 1981 Organization of ascending serotonin systems in the adult rat brain. A radioautographic study after intraventricular administration of [3H]5-hydroxytryptamine Neuroscience 6 115 138 6164006 Phelps EA 2006 Emotion and cognition: insights from studies of the human amygdala Annu Rev Psychol 57 27 53 16318588 Rainnie DG 1999 Serotonergic modulation of neurotransmission in the rat basolateral amygdala J Neurophysiol 82 69 85 10400936 Ravinder S Pillai AG Chattarji S 2011 Cellular correlates of enhanced anxiety caused by acute treatment with the selective serotonin reuptake inhibitor fluoxetine in rats Front Behav Neurosci 5 88 22232580 Reznikov LR Reagan LP Fadel JR 2008 Activation of phenotypically distinct neuronal subpopulations in the anterior subdivision of the rat basolateral amygdala following acute and repeated stress J Comp Neurol 508 458 472 18335544 Sadikot AF Parent A 1990 The monoaminergic innervation of the amygdala in the squirrel monkey: an immunohistochemical study Neuroscience 36 431 447 1977101 Sahuque LL Kullberg EF Mcgeehan AJ Kinder JR Hicks MP Blanton MG Janak PH Olive MF 2006 Anxiogenic and aversive effects of corticotropin-releasing factor (CRF) in the bed nucleus of the stria terminalis in the rat: role of CRF receptor subtypes Psychopharmacology (Berl) 186 122 132 16568282 Shin LM Rauch SL Pitman RK 2006 Amygdala, medial prefrontal cortex, and hippocampal function in PTSD Ann NY Acad Sci 1071 67 79 16891563 Shin LM Liberzon I 2010 The neurocircuitry of fear, stress, and anxiety disorders Neuropsychopharmacology 35 169 191 19625997 Shinnick-Gallagher P McKernan MG Xie J Zinebi F 2003 L-type voltage-gated calcium channels are involved in the in vivo and in vitro expression of fear conditioning Ann NY Acad Sci 985 135 149 12724155 Smith HR Porrino LJ 2008 The comparative distributions of the monoamine transporters in the rodent, monkey, and human amygdala Brain Struct Funct 213 73 91 18283492 Stein C Davidowa H Albrecht D 2000 5-HT(1A) receptor-mediated inhibition and 5-HT(2) as well as 5-HT(3) receptor-mediated excitation in different subdivisions of the rat amygdala Synapse 38 328 337 11020236 Steinbusch HW 1981 Distribution of serotonin-immunoreactivity in the central nervous system of the rat-cell bodies and terminals Neuroscience 6 557 618 7017455 Sur C Betz H Schloss P 1996 Immunocytochemical detection of the serotonin transporter in rat brain Neuroscience 73 217 231 8783244 Tovote P Fadok JP Lüthi A 2015 Neuronal circuits for fear and anxiety Nat Rev Neurosci 16 317 331 25991441 Tye KM Prakash R Kim SY Fenno LE Grosenick L Zarabi H Thompson KR Gradinaru V Ramakrishnan C Deisseroth K 2011 Amygdala circuitry mediating reversible and bidirectional control of anxiety Nature 471 358 362 21389985 Vertes RP 1991 A PHA L analysis of ascending projections of the dorsal raphe nucleus in the rat J Comp Neurol 313 643 668 1783685 Vertes RP Martin GF 1988 Autoradiographic analysis of ascending projections from the pontine and mesencephalic reticular formation and the median raphe nucleus in the rat J Comp Neurol 275 511 541 3192756 Vertes RP Crane AM 1997 Distribution, quantification, and morphological characteristics of serotonin-immunoreactive cells of the supralemniscal nucleus (B9) and pontomesencephalic reticular formation in the rat J Comp Neurol 378 411 424 9034900 Vertes RP Linley SB 2007 Comparisons of projections of the dorsal and median raphe nuclei, with some functional considerations Takai K Interdisciplinary conference on tryptophan and related substances: chemistry, biology, and medicine International Congress Series, 1304 Oxford, UK Elsevier 98 120 Vertes RP Linley SB 2008 Efferent and afferent connections of the dorsal and median raphe nuclei in the rat Monti JM Pandi-Perumal SR Jacobs BL Nutt D Serotonin and sleep: molecular, functional and clinical aspects Basel, Switzerland Birkhauser-Verlag 69 102 Vertes RP Fortin WJ Crane AM 1999 Projections of the median raphe nucleus in the rat J Comp Neurol 407 555 582 10235645 Vertes RP Linley SB Hoover WB 2010 Pattern of distribution of serotonergic fibers to the thalamus of the rat Brain Struct Funct 215 1 28 20390296 Vicente MA Zangrossi H 2012 Serotonin-2C receptors in the basolateral nucleus of the amygdala mediate the anxiogenic effect of acute imipramine and fluoxetine administration Int J Neuropsychopharmacol 15 389 400 21733232 Vicente MA Zangrossi H 2014 Involvement of 5-HT2C and 5-HT1A receptors of the basolateral nucleus of the amygdala in the anxiolytic effect of chronic antidepressant treatment Neuropharmacology 79 127 135 24275045 Walker DL Toufexis DJ Davis M 2003 Role of the bed nucleus of the stria terminalis versus the amygdala in fear, stress, and anxiety Eur J Pharmacol 463 199 216 12600711 Whalen PJ Phelps EA 2009 The human amygdala New York Guilford Press Yamamoto H Fujimiya M Shirai Y Nakashita M Oyasu M Saito N 1998 Immunohistochemical localization of serotonin transporter in normal and colchicine treated rat brain Neurosci Res 32 305 312 9950057 Zangrossi H Graeff FG 2014 Serotonin in anxiety and panic: contributions of the elevated T-maze Neurosci Biobehav Rev 3 397 406 Zangrossi H Viana MB Graeff FG 1999 Anxiolytic effect of intra-amygdala injection of midazolam and 8-hydroxy-2-(di-n-propylamino)tetralin in the elevated T-maze Eur J Pharmacol 369 267 270 10225362 Zhang Y Raap DK Garcia F Serres F Ma Q Battaglia G Van de Kar LD 2000 Long-term fluoxetine produces behavioral anxiolytic effects without inhibiting neuroendocrine responses to conditioned stress in rats Brain Res 855 58 66 10650130
PMC005xxxxxx/PMC5121055.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 8100316 1138 Biomaterials Biomaterials Biomaterials 0142-9612 1878-5905 27815996 5121055 10.1016/j.biomaterials.2016.10.046 NIHMS827606 Article Human Airway Organoid Engineering as a Step toward Lung Regeneration and Disease Modeling Tan Qi Choi Kyoung Moo Sicard Delphine Tschumperlin Daniel J. Department of Physiology and Biomedical Engineering, Mayo Clinic, Rochester, MN, USA 4 11 2016 28 10 2016 1 2017 01 1 2018 113 118132 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Organoids represent both a potentially powerful tool for the study cell-cell interactions within tissue-like environments, and a platform for tissue regenerative approaches. The development of lung tissue-like organoids from human adult-derived cells has not previously been reported. Here we combined human adult primary bronchial epithelial cells, lung fibroblasts, and lung microvascular endothelial cells in supportive 3D culture conditions to generate airway organoids. We demonstrate that randomly-seeded mixed cell populations undergo rapid condensation and self-organization into discrete epithelial and endothelial structures that are mechanically robust and stable during long term culture. After condensation airway organoids generate invasive multicellular tubular structures that recapitulate limited aspects of branching morphogenesis, and require actomyosin-mediated force generation and YAP/TAZ activation. Despite the proximal source of primary epithelium used in the airway organoids, discrete areas of both proximal and distal epithelial markers were observed over time in culture, demonstrating remarkable epithelial plasticity within the context of organoid cultures. Airway organoids also exhibited complex multicellular responses to a prototypical fibrogenic stimulus (TGF-β1) in culture, and limited capacity to undergo continued maturation and engraftment after ectopic implantation under the murine kidney capsule. These results demonstrate that the airway organoid system developed here represents a novel tool for the study of disease-relevant cell-cell interactions, and establishes this platform as a first step toward cell-based therapy for chronic lung diseases based on de novo engineering of implantable airway tissues. 3D culture Self-organization de novo lung regeneration YAP Organoid implantation Pulmonary fibrosis modeling 1. Introduction Respiratory diseases such as chronic obstructive pulmonary disease and pulmonary fibrosis represent a large and growing public health burden [1, 2], are associated with substantial morbidity and mortality, and currently lack curative therapies. At their end-stage, such diseases require lung transplantation for therapy, but the supply of donor organs is extremely limited, and lung transplant outcomes remain suboptimal [3]. Regenerative approaches offer potential long-term hope for addressing both the epidemic of chronic lung diseases, and the shortage of donor organs, but critical hurdles remain to be overcome [4]. While recent studies have made great progress delineating the mechanisms of lung development and developing methods to drive iPS cells toward mature lung lineages [5–7], relatively less progress has been made in designing strategies by which these advances might be translated into tissue repair, and ultimately advanced toward human studies. Current approaches to engraft dissociated cells in the lung show promise, but have thus far been limited to the setting of severe infections [8] or radiation-induced preconditioning [9]. A major alternative emphasis has been on the generation of decellularized and recellularized lung scaffolds as an engineered organ replacement [10–13]. Relatively less attention has been devoted to the de novo generation of complex three-dimensional lung-like tissues in culture suitable for eventual translational applications. A potential additional benefit from developing such complex engineered lung tissues is for disease modeling. Chronic lung diseases are distinguished by specific tissue remodeling processes and complex cell-cell interactions that are not easily recapitulated in typical cell culture systems. Therefore, we sought to develop an airway organoid culture system combining multiple lung cell types as both a step toward eventual regenerative approaches, and as a system to study disease-relevant cell-cell interactions and complex tissue remodeling processes. To generate highly organized 3D tissues that mimic organ structure and function, tissue engineers have attempted to recapitulate the in vivo organogenesis process by manipulating critical aspects of the cell culture environment. During embryonic development, the lungs and other internal organs first emerge as organ buds composed of epithelial and mesenchymal progenitors. Through repeated rounds of outgrowth and branching primitive organ buds grow into mature organs [14]. The reciprocal epithelial-mesenchymal interactions critical to organogenesis during embryonic development can be recapitulated in three dimensional co-culture systems to guide formation of similar tissue-like structures in vitro [15]. Recently, complex structures termed organoids [16] have been generated for brain [17], liver [18], pancreas, and lung [19] using combinations of induced pluripotent stem cells, inductive soluble factors, and supportive three dimensional culture conditions. Alternatively, resident progenitor cells from adult tissues can be cultured in supportive 3D systems, and can also generate organoids. Typical examples include LGR5+ cells from intestine and liver [20], and in the field of lung biology, the generation of tracheospheres [21] and alveospheres [22, 23] from airway and alveolar epithelial progenitors. While organoids have shown promise in transplantation models in the colon [24] and liver [25], similar advances have not been reported using adult-derived lung progenitors. Similarly, although organoids have potential for disease modeling and drug screening, tractable human lung cell-based organoid systems have not been reported. Here we combined adult human primary bronchial epithelial cells, lung fibroblasts, and lung microvascular endothelial cells in 3D culture conditions to generate airway organoids. By combining epithelial differentiation conditions with a multicellular aggregation culture system, we generated self-assembling bioengineered airway organoids that are amenable to ectopic transplantation and study of cell-cell interactions crucial to tissue biology. This system represents a novel tool for studying disease-relevant cellular and molecular function, and an important step toward cell-based therapy for chronic lung diseases based on de novo engineered airway tissues. 2. Material and methods 2.1 2D cell culture and labeling Human bronchial epithelial cells (NHBE, 8 cell lines used) were purchased from Lonza and cultured in bronchial epithelial growth medium (BEGM, Lonza) with 1% Antibiotic-Antimycotic. Human microvascular lung endothelial cells (HMVEC-L, 3 cell lines used) were also purchased from Lonza and cultured in Endothelial Cell Growth Medium (EBM-2MV, Lonza) with 1% Antibiotic-Antimycotic. Human lung fibroblasts (generously provided by Carol Feghali-Bostwick) previously isolated by explant culture from donor lungs rejected for transplantation under a protocol approved by the University of Pittsburgh Institutional Review Board, were culture in DMEM with 10% FBS, 1% Antibiotic-Antimycotic. The cells were regularly maintained in humidified 5% CO2 at 37 °C. For cell labeling, cells were incubated with pre-warmed CellTracker™ Working Solution (1µM from 1000× DMSO dissolved stock solution) for 30 mintues. 2.2 3D cell culture and generation of human airway organoids Culture plates were coated with 40% Matrigel (Corning) combined with PneumaCult-ALI Maintenance Medium (Stemcell Technologies). The former is an extracellular matrix preparation derived from a gelatinous protein mixture secreted by Engelbreth-Holm-Swarm (EHS) mouse sarcoma cells, composed of approximately 60% laminin, 30% collagen IV, and 8% entactin. Entactin is a bridging molecule that interacts with laminin and collagen IV, and contributes to structural organization. Although Matrigel has inherent long term limitations for clinical use, it is a valuable proof of concept matrix system for supporting organoid morphogenesis [26]. PneumaCult-ALI is a media formulation optimized to induce primary human bronchial epithelial cell mucociliary differentiation under air-liquid interface (ALI) culture conditions. After 45 minutes incubation at 37°C for gelation of the thick Matrigel layer, a single cell suspension including NHBE, HMVEC-L and HLF cells was combined with 5% Matrigel and PneumaCult-ALI Maintenance Medium, and was seeded on top of the Matrigel layer. The ratio of each type of cell was human NHBE:HMVEC-L:HLF = 10:7:2 following a previously described approach for generation of liver organoids [18]. The cells were fed with 5% Matrigel in PneumaCult- ALI Maintenance Medium every other day and observed by light microscopy. Cells labelled with Fluorescent CellTracker™ were observed in a Cytation 5 Cell Imaging Multi-Mode Reader for 24 hour time-lapse fluorescence microscopy to visualize organoid compaction and formation. 2.3 Organoid actomyosin and YAP analysis Day 2 organoids were used for treatment. ON-TARGETplus Human YAP1 (10413) siRNA SMARTpool (30 pmol, Dharmacon) in Opti-Mem medium (Life Technologies) was transfected using Lipofectamine RNAi Max (Life Technologies) followed by incubation for 24 hours at 37 °C, with media than changed to the normal differentiation medium for another 48 hours. Non-muscle myosin inhibitor blebbistatin (10µmol, Sigma, B0560), F-actin inhibitor cytochalasin D (10µmol, Sigma, C8273) and Rho kinase inhibitor Y-27632 (100µmol, Stemcell Technologies, 72304) were used to treat day 2 organoids for 72 hours. Photos were taking by Nikon TMS Microscope or Cytation 5 Cell Imaging system. Buds or invasive tubular structures were counted manually and average diameters of organoids were counted in 3 organoids per group using ImageJ. 2.4 Atomic force microscopy Organoids were embedded in O.C.T (Optimal Cutting Temperature, n°4583) and stored at −80°C. 10 µm thickness organoid slices were cut by cryosection (Leica) at −21°C and mounted on poly-L-lysine coated glass slides. To avoid drying, tissue slices were maintained in PBS. Measurements were performed using a BioScope Catalyst AFM (Bruker) mounted on an inverted microscope equipped with epifluorescence (Olympus) using a spherical tip (Novascan) with a radius of 2.5 µm and a spring constant of ~98 pN/nm. Force curves were acquired with MIRO 2.0 (NanoScope 9.1, Bruker) at room temperature in PBS. The indentation was estimated at ~250 nm for an applied force of ~24 nN. The preparation of organoid slice sample and the AFM measurements were performed on the same day. Force curves were analyzed by NanoScope Analysis (Bruker). The extend curve was fitted to determine the Young’s modulus using the Hertz model assuming Poisson’s ratio of 0.4 [27]. 2.5 Gene expression analysis (RNA isolation, cDNA synthesis, RT-PCR) Total RNA was isolated with RNeasy Plus Mini kit and cDNA was synthesized with SuperScript™ IV Reverse Transcriptase. Gene expression levels were quantified by qRT-PCR on the Lightcycler 96 Real-Time PCR System (Roche) according to the manufacturer’s instruction. qRT-PCR was performed by incubating at 95°C for 10 min and then cycling 40 times at 95°C for 10 s, 60°C for 10s, 72°C for 10s. Ct values within each experiment were normalized against GAPDH. Primers are listed in Supplemental Table 1. 2.6 ELISA for MUC5AC Total protein was collected from organoids and Human MUC5AC ELISA Quantitation Kit (LifeSpan BioSciences) was used to assess protein expression of MUC5AC according to the manufacturer’s instructions. One measurement was made per single organoid. 2.7 Tissue processing, histological tissue assessment, immunostaining and confocal microscopy Organoids were collected and embedding in O.C.T. 10 µm cryosections were placed on Poly-L-Lysine Coated Slides (Fisher Scientific) for immunostaining or standard histological staining with the BBC Histo·Perfect H&E Staining kit. For immunostaining, tissues were fixed in 4% paraformaldehyde for 20 minutes and permeabilized using 0.25% Triton X-100 for 15 minutes. Antigen retrieval was applied with PH=6.0 citrate buffer when necessary. Organoids were treated with 5% goat serum in 1% BSA/PBS (block solution) for 1 hour. Following two additional PBS washes, samples were incubated overnight with primary antibodies at 4°C. Specimens were washed twice in PBS and incubated with the corresponding secondary antibodies at 1:500 dilution for 1 hour (Alexa Fluor488, 546, 555; Invitrogen). Samples were mounted with mounting media containing nuclear counterstain DAPI (Vector Lab). Images were acquired using the LSM780 inverted laser scanning confocal microscope (Carl Zeiss). Images were processed using ZEN microscope and imaging software, Photoshop or ImageJ. Cell quantification of immune-positive cells and DAPI was performed using the Cytation 5 Cell Imaging Reader software GEN 2.09 or ImageJ using the cell counter plugin. Antibodies are listed in Supplemental Table 2. 2.9 Statistical Analysis Data are presented as mean ± SEM. Statistical analyses were performed using GraphPad Prism 5 (GraphPad Software). The significance of the differences between two mean values was determined using an unpaired t test. P values less than 0.05 were accepted as significant. 3. Results 3.1 Development of an airway organoid from adult lung cells To develop airway organoids from adult lung cells we combined methods optimized for growth and differentiation of human airway epithelial cells [21] with mesenchymal cell-guided condensation on a supportive Matrigel layer [28]. Specifically, we combined adult primary human bronchial epithelial cells (NHBE) at a ratio of 10:7:2 with adult primary human microvascular lung endothelial cells (HMVEC-L) and adult primary human lung fibroblasts (HLF) atop a 3D Matrigel-coating in each well of a 96 well culture plate (Fig 1A). The cells rapidly condensed atop the Matrigel layer and formed 3D cell aggregates within approximately 24 hours, continuing to condense up to 72 hours. When seeding fewer than 1×105 epithelial cells per well, multiple spheroids formed in each well, sometimes with large variance in organoid size, and with increasing organoid numbers and smaller organoids observed as the epithelial cell density was decreased (Fig 1B and S1). For seeding densities of 1×105 epithelial cells and higher, a single organoid formed in each well, with consistent formation noted across multiple NHBE donor lines (Fig S1). In order to minimize variation in organoid formation associated with low seeding density, and potential complications introduced by diffusion limitations associated with high seeding densities and larger organoids, we chose to focus all subsequent studies on organoids formed from 1 × 105 adult primary human bronchial epithelial cells (total cell number of ~1.9×105 cells/well). At this seeding density, multicellular aggregates remained stable in overall size after 3 days, with diameters in the range of ~1000–2000 µm (Fig 1C). To identify the critical cell types for this dynamic and directed condensation process, we examined single cell type behaviors and various combinations of the three cell lineages (Fig 1D–F). Omitting fibroblasts from the co-culture led to a failure in cell aggregate condensation, though smaller spherical colonies did ultimately appear (Fig 1D), recapitulating previous descriptions of tracheosphere formation [21]. In contrast, omission of endothelial cells did not prevent organoid formation in the 3D culture system. Fibroblasts alone (Fig 1E) or in combination (Fig 1F) with either or both other cell types were sufficient for aggregation, confirming that the presence of fibroblasts is critical for the self-condensation process. These observations are consistent with previous reports that fibroblast-dependent paracrine support through the activation of FGF and BMP pathways and fibroblast-dependent cytoskeletal contractions are essential for cell aggregate condensation [28]. Because of our ultimate interest in regenerative and disease-modeling applications, we continued to focus on three cell co-cultures that include epithelial, endothelial and mesenchymal resident populations of the mature lung in all subsequent work. We termed our organ bud-like 3D structures “airway organoids” because the epithelial cell type used originated in human bronchial airways. Comparison of epithelial markers expressed in primary NHBE cells in monolayer culture confirmed the dominant proximal basal phenotype (P63+) of this starting cell population (Fig 1G and Fig S2). As a preliminary evaluation of the self-organizing capability and mechanical stability of airway organoids, we labelled the cells individually with different CellTracker™ probes in 2D culture conditions before seeding them randomly on Matrigel-coated plates for organoid culture. Fluorescence time-lapse images were then obtained, allowing us to visualize self-organization of endothelial cell (Cyan) and fibroblast (Green) clusters among comparatively homogenously distributed NHBE cells (Red) (Fig. 1A). Using the approach detailed above, successful organoids formed (diameter larger than 500µm) at a rate of ~80% (64/80), and airway organoids were manipulable with spatulas or forceps, suggestion potential for surgical handling and implantation. Atomic force microscopy (AFM) micro-indentation was performed to evaluate the local elastic modulus within airway organoids (expressed as Young’s modulus) guided by light and fluorescence microscopy (Fig 1H). The relatively high modulus observed within the aggregates was consistent with our manual manipulation observations, and suggested that nascent airway organoids form physically robust aggregates through a mechanically dynamic process that includes considerable mechanical heterogeneity during the process of condensation and self-organization (Fig 1I). 3.2 Epithelial, mesenchymal, and endothelial cells re-organize within the airway organoid To further study the self-organization of lung epithelium, endothelium and mesenchyme within airway organoids, we performed immunostaining and confocal microscopy at various time points after cell seeding. At early time points, we found that airway organoids possessed broadly distributed mesenchymal cells, identified by vimentin immunostaining, and clusters of epithelial cells positive for E-Cadherin (CDH1), airway epithelial basal marker P63 [5], and luminal marker KRT8 [21] (day 7, Fig 2A, 2B, 2L). Epithelial cell clusters self-assembled into tubular structures of varying shape and size. The expression of basal airway stem cell [5] transcripts (P63+, KRT5+, KRT14+) diminished in 3D airway organoid culture compared with 2D cultures of NHBE cells (Fig 2C), consistent with a transition away from the predominant basal stem cell phenotype expressed by these primary cells in 2D culture. A gradual basal to luminal shift in epithelial phenotype was confirmed by the increased observation of the luminal airway epithelial marker (KRT8+) in the airway organoids (Fig 2C). Similar to the observed epithelial self-organization, CD31+ endothelial cells were also visualized to form cell clusters along with tube like-structures in airway organoids (Fig 2D, 2E). These endothelial structures were closely associated with fibroblasts, but not epithelial cells. CD31+ endothelial cells decreased dramatically in abundance over time (Fig 2D, 2E) and after 14 days CD31+ cells were rarely observed. This finding was confirmed by the gradual loss of CD31 transcripts over time in airway organoid co-cultures (Fig 2F). Such a finding was not surprising given our use of culture media optimized for NHBE cells. These findings thus demonstrate the capacity of human microvascular endothelial cells to self-organize in airway organoid culture, but clearly indicate the need to optimize long term culture to promote endothelial 3D stabilization and maturation of vascular structures. Proliferations in the airway organoids, assessed and quantified by Ki67 expression, was most abundant at the earliest time point analyzed (day 3, Figure 2H) and decreased over time (Fig 2H–K, N), consistent with the relative stability in airway organoid size over time. The vast majority of Ki67 expression was not in cells expressing epithelial lineage markers (CDH1+), but rather in cells expressing the mesenchymal lineage (VIM+, CD31−), suggesting that only limited epithelial proliferation and modest fibroblast proliferation was present during long-term culture of airway organoids. The apparent low cell turn over in airway organoids mimics adult lung and airway homeostasis [29], but also suggests that more proliferative progenitor populations, or more stimulative culture conditions, may be required to enhance organoid growth. Taken together the results documented in Figure 2 demonstrate that the airway organoids formed from adult human bronchial epithelial, fibroblast and endothelial cell populations undergo rapid self-organization over time into highly dense epithelial-lined structures surrounded by mesenchyme, with a gradual loss of vascular endothelium. 3.3 Proximal airway epithelial differentiation in airway organoids NHBE cells cultured under defined conditions at air-liquid interface are capable of undergoing maturation and differentiation to a pseudo-stratified, mucociliary epithelium [30]. To evaluate the capacity of NHBE cells to undergo such differentiation in 3D airway organoid culture, we analyzed epithelial protein and transcript markers by immunostaining and quantitative PCR respectively. Airway organoids demonstrated gradual increases in expression of the key secretory product of airway club cells, the protein uteroglobin [31], denoted SCGB1A1 or Club Cell-Specific 10 kD Protein (CC10, Fig. 3A–C). Club cells are responsible for detoxifying harmful substances inhaled into the lungs, including engulfment of airborne toxins and breakdown via their cytochrome P-450 enzymes [32]. Consistent with this function, we found that mRNA expression of cytochrome P-450 enzymes increased significantly in the 3D culture (Fig 3D). Secretory club cells are mitotically active and also act as progenitors for epithelial repair, as they are able to differentiate into ciliated cells or non-ciliated cells to regenerate the bronchiolar epithelium [31]. CC10 protein (Fig 3A–C) and transcript levels (Fig 3D) increased dramatically over time in the airway organoid 3D culture compared with NHBE 2D culture (Fig 3D). Goblet cells are specialized secretory cells found in the respiratory epithelium of the trachea, bronchi, and larger bronchioles, and are responsible for mucus secretion and maintenance [33]. Goblet cells are denoted by secretory vesicles containing the highly charged glycoprotein MUC5AC. We observed increasing organization of MUC5AC+ immunostaining inside apparent lumens within airway organoids over time (Fig 3A–3C), consistent with epithelial organization and goblet cell differentiation. Transcript levels for MUC5AC were also significantly increased in prolonged airway organoid culture, consistent with epithelial secretory differentiation over time in airway organoids (Fig 3D). Moreover, robust MUC5AC protein expression was detected by ELISA, peaking at day 7 but maintained out to day 21 in airway organoids (Fig 3E). In contrast to abundant signs of secretory differentiation, comparatively few cells expressed the ciliated cell marker [34] FOXJ1 (Fig 3F, 3G), suggesting that the 3D culture environment employed here does not adequately promote terminal differentiation of all airway epithelial cell types. Recent work has highlighted a prominent role for Notch signaling in controlling airway epithelial differentiation [35]. Notch signaling inhibition with DAPT treatment increased the mRNA expression level of FOXJ1 (Fig 3J) as well as FOXJ1 immunostaining (Fig 3H, 3I). FOXJ1+ cells were accompanied by intense and polarized co-staining for acetylated-alpha tubulin (Fig 3H–I), consistent with mature ciliogenesis [36]. DAPT treatment also dramatically reduced transcript levels of CC10 while increasing transcript levels of MUC5AC (Fig 3J). In combination, these data demonstrate that the current airway organoid culture conditions support gradual club and goblet cell epithelial differentiation over time, and that the cells remain capable of responding to additional cues to undergo enhanced ciliated cell formation, recapitulating many of the maturation and differentiation processes that play out in gold-standard ALI culture of NHBE cells. 3.4 Expression of distal lung epithelial lineage markers in airway organoids In addition to epithelial clustering within compacted organoids, we also observed invasive tubular structures emanating from airway organoids with prolonged culture, as shown in a representative H&E stained section (Fig 4A) and brightfield images of live organoid cultures (Fig S3A, S3B). In contrast, we did not observe any invasive tubular structures forming from long-term NHBE mono-cultures (day21, Figure S3C) which instead remained uniformly spherical in shape. Organoid invasive structures invariably demonstrated close apposition of E-cadherin and vimentin immunostaining, suggesting close epithelial-mesenchymal cell interactions (Fig 4B). E-cadherin positive epithelial cells were typically tightly surrounded by vimentin positive cells (Fig 4C), recapitulating epithelial-mesenchymal orientation during airway branching morphogenesis [14, 37]. Some epithelial-fibroblast invasive structures extended far from the originating organoid, with a transition to a more flattened and extended epithelial morphology (Fig 4D) suggesting the possibility that aspects of distal lung development might also be evoked within airway organoid culture. The distal epithelium of the lung is lined by Type I and type II alveolar epithelial cells (AECI, AECII). The AECI marker Aquaporin 5 (AQP5) [38–40] was strongly expressed within tubular structures near the periphery of the main airway organoid body, and within the thin epithelial lining of invasive projections (Fig 4E, 4F), where it co-stained with E-cadherin (Fig 4G). Podoplanin (PDPN), another marker of AECs highly expressed in lung parenchyma [41], but also in lung fibroblasts [42], was much more broadly distributed in the fibroblast-populated area of human airway organoids, but not typically in AQP5+ or E-cadherin+ cells (Fig 4E, 4F, Fig S3D). In multiple instances we observed co-staining of the more proximal epithelial marker CC10 with the distal epithelial markers AQP5 and HOPX (Fig 4I–L), implying considerable complexity and the presence of immature or mixed-lineage expression within areas of organoid-derived tubular structures, consistent with noted epithelial cell type diversity in distal lung development [43]. In other locations, relatively strong and exclusive expression of distal epithelial markers were observed within tubular structures (Fig 4F, L), consistent with cells undergoing further distal lineage-specific differentiation. Additional evidence for AECI lineage expression within airway organoids was obtained by qPCR, which demonstrated prominent upregulation of AECI markers AQP5, PDPN and HOPX [5, 23] over time (Fig 4M). Within organoids, but distinct from invasive structures, we also observed large epithelial clusters that were relatively devoid of interacting fibroblasts (Fig 4H), and stained positively for CC10, the AECII marker SPC, or both [31, 44, 45] (Fig 4H), suggesting further regional complexity in epithelial lineage marker expression within airway organoids, and the potential emergence of unique mixed lineage cells distinct from the original basal bronchial epithelial cell population. Taken together, our data demonstrates that the epithelium within airway organoids displays remarkable plasticity, and in a limited and surprising fashion recapitulates aspects of both proximal airway and distal alveolar cell self-assembly and differentiation. 3.5 Invasive tubular structures require actomyosin contraction and localized YAP activation The omission of fibroblasts in our organoid culture led not only to formation of smaller aggregates, but also to a failure of invasive tubular structures to form (Fig 1D, Fig S3C). These observations suggested to us that both organoid formation and invasion requires mechanical contributions, particularly those generated by contractile fibroblasts. In agreement with this concept, inhibition of myosin II with blebbistatin both attenuated organoid compaction and reduced the number and extent of invasive bud formation at the periphery of airway organoids (Fig. 5A–C). Similar findings were observed with the Rho kinase inhibitor Y27632 and the inhibitor of actin polymerization cytochalasin D (Fig. 5C), consistent with an essential role for actomyosin-mediated contractile forces in organoid formation as previously observed [25, 28], and also for tubular invasion as noted here. Interestingly, we observed that the overall levels of cell proliferation, as quantified by Ki67 immunopositive cells, were significantly higher in organoids treated with blebbistatin, cytochalasin D and Y-27632 (Fig 5H–J, S3J, S3K) consistent with proliferative effects of actomyosin inhibition in other contexts [46, 47], and suggestive of enhanced proliferation also contributing to greater organoid size under these conditions. The mechanoresponsive transcriptional regulator YAP has been widely implicated in tissue morphogenesis [48, 49], and plays important roles in both lung epithelium [50–52] and mesenchyme [53–55]. Therefore, we hypothesized that YAP activation may play key roles in the formation of organoids and invasive tubular structures. Knockdown of YAP by siRNA did not significantly reduce organoid compaction (Fig. 5C). However, immunostaining for YAP demonstrated robust nuclear localization at the periphery of organoids associated with invasive tubular structures (Fig. 5D, 5F). Knockdown of YAP by siRNA (Fig S4D) significantly diminished invasive bud formation (Fig. 5C), demonstrating an essential role for YAP in formation and invasion of these structures. Notably, YAP knockdown disrupted the organization of epithelial and mesenchymal cells, as well as deposition of an organized fibronectin matrix (Fig 5M), suggesting profound reliance on this mechanoresponsive pathway for tissue organization and morphogenesis (S4A, S4E). Consistent with the mechanoresponsive nature of this process, blebbistatin (Fig 5E, 5G and 5N), cytochalasin D (Fig S4B, S4F) and Y-27632 (Fig S4C, S4G) disrupted the localized pattern of YAP expression, as well as dramatically disorganizing epithelial-mesenchymal distribution and patterns of fibronectin deposition within airway organoids. While overall cell proliferation was not significantly altered with YAP siRNA knockdown (Fig 5J, 5I, S4I), the localized pattern of proliferation typically observed in invasive tubular structures at the organoid edge (Fig 5H) was lost with YAP knockdown (Fig S4I). Similar disorganization of proliferation patterns was observed with blebbistatin (Fig 5I), cytochalasin D (Fig S4J) and Y-27632 treatment (S4K), consistent with prior observations of mechanical patterning of proliferation via YAP [56]. Based on previous observations that YAP plays a critical role in airway epithelial patterning, we also examined expression of basal and luminal epithelial markers after treatment to disrupt actomyosin or knockdown YAP by siRNA. Real-time PCR analysis demonstrated that basal cell markers KRT5 and KRT14 were significantly reduced (Fig 5L) and airway club cell differentiation marker CC10 were up-regulated with YAP siRNA or actomyosin disruption (Fig 5K). We also observed YAP nuclear localization and expression of CC10 were exclusive of one another in the organoid (Fig S4H, S4L). These observations are consistent with prior in vivo findings that YAP expression controls progenitor cell fate in the lungs and promotes a basal cell expression program [52, 57], demonstrating that the organoid system recapitulates aspects of YAP-mediated epithelial fate determination observed in mouse models of lung development. 3.6 Modeling a fibrogenic stimulus response in human airway organoids Currently, in vitro studies of lung fibrosis typically focus on a single cell types, especially fibroblasts[58] or lung epithelial cells[59]. The close proximity of primary lung epithelial cells and mesenchymal cells within airway organoids prompted us to investigate whether our model could be useful for studying the orchestrated responses of multiple cell types to a fibrogenic stimulus. Therefore we employed a pathophysiologically relevant pro-fibrotic stimulus, TGF-β1[60]. Not only is TGF-β1 highly relevant in the context of pulmonary fibrosis [61], it is implicated in lung development [62], airway and alveolar remodeling [63], and is known to evoke profound responses in primary human epithelial, endothelial and mesenchymal cell types [60]. To evaluate the response to TGF-β1, we treated airway organoids (AOs) with 5 ng/mL TGF-β1 for 96 hours and then analyzed gene expression by qPCR and organoid morphology by immunofluorescence. We found that TGF-β1 treatment inhibited/reversed tubular invasion of the matrix at the periphery of airway organoids (Fig 6A), consistent with known inhibitory effects of TGF-β1 on lung branching and epithelial differentiation in embryonic and fetal lung in vivo [64]. Immunofluorescence imaging demonstrated increasing αSMA expression (Fig 6B) with TGF-β1 treatment, and qPCR confirmed significant induction of several fibrogenic genes in airway organoids, including FN(1.5-fold), ACTA2 (2.3-fold), CTGF(1.6-fold) as shown in Figure 6C. To address which cell types contribute to increased αSMA, we labeled either fibroblasts (HLF) or epithelial cells (NHBE) with CellTracker™ Red CMTPX prior to organoid formation, then treated with TGF-β1 as above and used immunofluorescence imaging to visualize αSMA+ cells. Co-staining of αSMA was observed with both CellTracker+ fibroblasts (Fig S5A) and epithelial cells (Fig S5B), indicating the both fibroblasts and epithelial cells contribute to the enhanced αSMA expression under TGF-β1 treatment, echoing prior observations of TGF-β1 responses in primary bronchial epithelial cells[65]. Interestingly, TGF-β1 had the opposite effect on expression of proximal lung epithelial cell markers CC10 and FOXJ1 and the distal lung epithelial cell marker AQP5, all of which decreased dramatically after treatment (Fig 6D). Meanwhile, the expression of airway epithelial basal cell markers (P63, KRT5) were increased by TGF-β1 treatment (Fig 6E). These observations highlight the robust multicellular effects of TGF-β1 in airway organoids, with simultaneous increases in fibrogenic gene expression and decreases in expression of epithelial differentiation markers, and support the potential for this model to be used in elucidating cell-cell interactions that regulate tissue level responses to fibrogenic stimuli. 3.7 Ectopic transplantation of human airway organoids To evaluate whether airway organoids can be transplanted in vivo and are capable of surviving and vascularizing in vivo, we transplanted day 7 airway organoids, embedded within Matrigel, under the kidney capsule of NSG mice. Kidneys were harvested after 1 week or 6 weeks for visual and microscopic examination. Organoids were easily visible after 1 week in the kidney capsule (Fig S6A). At 1 week after implantation there was prominent proliferation in the vicinity of the implanted organoid, particularly at the interface between host tissue and airway organoid grafts. However, co-staining with human specific E-cadherin (Fig 7B) and vimentin (Fig 7C), along with an antibody specific to human mitochondria (Fig 7A), demonstrated that the vast majority of proliferation was not within airway organoid cells, but rather within host tissue. Human specific CD31+ endothelial cells were observed within airway organoids (Fig 7E–H), with some in close proximity to epithelial tubular structures (Fig 7E–F), while others were distributed outside epithelial structures surrounded by fibroblasts (Fig 7G–H). By week 6 after implantation organoids regressed in size. Under microscopic visualization the remaining human cells appeared to be fewer in number (Fig 7D), and although human specific CD31+ endothelial cells were observed, they were less abundant than at 1 week, and restricted to the edge of the implanted organoid (Fig S6B, S6C). In contrast, abundant host vasculature, as indicated by anti-CD31 antibody that reacts with both mouse and human (green) was found to have broadly invaded the organoid area (Fig S6B, S6C). Thus, while we did not observe any direct human to mouse vasculature connection, these observations are consistent with a robust capability of airway organoids to recruit host vasculature. Further immunostaining with lung epithelial differentiation markers demonstrated enriched and relatively distinct expression of proximal secretory airway epithelial cell marker CC10 (Fig 7K, L, S6D) or distal alveolar epithelial markers AQP5 (Fig 7K, L, S6E) and SPC (Fig 7I, J), consistent with the remarkable epithelial plasticity observed in airway organoids in vitro, but with less evidence of the mixed-lineage cells observed in vitro. In combination with the relative paucity of proliferation observed at 1 week, and the general lack of ectopic tissue growth from week 1 to 6, our results suggest that ectopic transplantation of airway organoids in vivo prompted a shift toward cell lineage commitment and differentiation toward mature, non-proliferating states. Thus our data demonstrate that the bronchial epithelial, mesenchymal and vascular cells in the airway organoid can survive and undergo considerable maturation following engraftment in vivo. However, under the present conditions adult human airway organoids did not undergo robust growth and expansion when ectopically transplanted under the kidney capsule, suggesting a somewhat limited regenerative potential of the current system. 4. Discussion We report the development of airway organoids from a mixed population of human adult lung epithelial, mesenchymal and endothelial cells. We observed a remarkable capacity for self-assembly, morphogenesis, and differentiation within airway organoids that mimicked many aspects of lung tissue formation and maturation. Lung organoids have recently been reported from human iPS cells [19], and smaller organoids have been formed from defined lung progenitors alone or in combination with supportive niche cells [21–23]. Our study differs from these efforts in two key ways. First, we used an unsorted population of human adult-derived cells. This is advantageous from a tissue sourcing perspective as all of the cells were commercially available, but clearly limiting in delineating mechanisms of differentiation, and in the observed limited proliferative potential of human adult cell-derived airway organoids. Second, we used organ bud culture conditions [25] which allowed a mixed population of cells to rapidly aggregate and self-condense, organize and engage in reciprocal cell-cell interactions. Such a method is ideal for engineering organ buds from a large mixed cell population, and offers the potential to create relatively large and transplantable tissue-like aggregates, but is clearly less amenable to mechanistic dissection of cellular lineage and fate choices during organoid culture. Interestingly, we observed expression of both proximal and distal epithelial markers within airway organoids. Our source for lung epithelium was commercially available normal human bronchial epithelial cells with a predominant basal cell phenotype. Lung epithelial cells in general display tremendous plasticity in regenerative contexts [66]. Recent studies have shown that adult airway epithelial cells preserve the potential, when needed, to proliferate, migrate and differentiate into several cell types during the severe lung injury [8, 67, 68]. Our study shows that an unsorted population of NHBE cells maintains surprising potential to express hallmarks of both proximal and distal epithelial lineages found in the adult lung. Recent studies have shown that the fate and multilineage potential of epithelial stem cells can change depending on whether a stem cell exists within its resident niche and responds to normal tissue homeostasis, whether it is mobilized to repair a wound, or whether it is taken from its niche and undergoes de novo tissue morphogenesis after transplantation [69]. Additional efforts will be needed to determine if the remarkable epithelial plasticity observed here reflects the unique niche(s) provided by organoid culture, or instead reflects the mixed starting epithelial population and the possible presence of alveolar (or other) lung progenitors with distal lineage potential. In addition, further investigation of additional niche cells specific to the proximal airways, such as airway smooth muscle cells [70], may offer additional insight into the cell-cell interactions that regulate airway epithelial differentiation within our organoid system. Whatever the outcome of these future studies, the results provided here demonstrate a remarkable potential for an adult human lung epithelial population to self-organize and mature toward both airway and alveolar lineages, suggesting the presence of important cell-cell interactions within airway organoids that merit further investigation. One striking observation from our airway organoids was the formation of invasive tubular structures that recapitulated the basic architecture of epithelial and mesenchyme juxtaposition observed during lung development. Multicellular tubular structures are essential elements in formation of complex organs and tissues, and their formation here by budding from the central organoid mass mimics, though clearly in a limited fashion, a basic aspect of organ development. Despite the formation of these invasive structures, we note that distinct highly-organized and integrated proximal and distal structures mimicking mature airway and alveoli were not commonly observed in the organoids. Rather, while the organoids display remarkable self-organization capacity, the patterns of cellular differentiation and morphogenesis suggest a relatively stochastic process, as is apparent in detailed IF images (Fig 3–4). This is perhaps due to the organoid formation process we have adopted, in which compaction of the mixed cell population generates local variations in epithelial, endothelial and fibroblast cell densities, superimposed on the cellular variability inherent to these cell populations. Thus while the organoids display robust self-organizing capacity and potential for differentiation in vitro, they remain far from mature lung tissue. Our observations are similar to other organoids studies [25, 71, 72] that show that in vivo incubation is essential to further maturation and organization of tissue like functionality in organoids beyond the relatively primitive embryonic tissue-like states that form in vitro. Excitingly, we were able to demonstrate that molecular control over organoid invasion and differentiation appears to be governed by YAP, consistent with its important role in airway branching, morphogenesis and differentiation in the developing lung [50–52, 54, 55, 57], suggesting that airway organoid formation recapitulates key signaling aspects of lung development. Better understanding the morphogenetic function of YAP could facilitate advances in generating complex tissue engineered structures such as organoids, and given YAP’s mechanoresponsive nature [53, 56], our findings suggests that exogenous control of mechanical forces to better mimic the developmental environment may be necessary to promote enhanced organoid growth and morphogenesis [73]. A major goal going forward will be to focus on cell signaling and genetic factors that guide branching morphogenesis of airway organoids, as well as regulatory cues from ECM, mesenchyme and vascular endothelial cells that influence cellular differentiation and organoid morphogenesis. We also demonstrated the capacity for ectopic transplantation of human adult airway organoids into immuno-deficient mice, with the presence of both recruited host vascular cells and lineage-specific differentiation of graft epithelial cells. While the observed in vivo differentiation potential was supportive of possible future application of this approach in regenerating functional lung tissue, several barriers remain. Notably, engrafted airway organoids did not grow over time, suggesting limited proliferative potential of the adult-derived cells used in our approach. More proliferative responses have been previously reported when using selected resident lung epithelial progenitor subpopulations [67, 74] or immortalized airway derived cells [75]; thus, sorting and utilization of resident progenitor populations may provide improvements in organoid growth [76]. Alternatively, fetal-derived lung cells may offer growth advantages innate to their developmental state, and have been shown previously to undergo organoid-like organization and growth [77, 78]. Ultimately, human iPS-derived cells driven toward lung lineages may provide the clinically-relevant cell source needed for patient-specific applications [79, 80]. Moving from ectopic transplantation of airway organoids to lung-specific application will require further development of tissue implantation strategies. Recent work has demonstrated successful attachment and vascular recruitment into biocompatible hydrogels on the pleural surface of the lung [81], suggesting one possible path forward. Further efforts to improve vascularization may also be critical in supporting organoid engraftment and growth. A recent study demonstrated that co-seeding of endothelial and perivascular cells is essential to successful revascularization of decellularized lung tissue [82]. Sorting and seeding of mesenchymal cell subpopulations that support vascularization in organoids may be one means by which to enhance organoid growth and engraftment potential. Finally, inclusion of additional matrix and soluble cues [13] will likely be essential to further optimize cell organization, growth and differentiation, and in vivo integration of airway organoid grafts with host tissue. A more immediate opportunity for use of airway organoids is apparent in the realm of disease-modeling. There is a clear rationale and drive to develop new 3D multi-cell culture models that overcome limitations of traditional 2D culture systems, allowing cells to grow and respond in environments that mimic native tissue and promote close interactions between cells, ECM, and soluble cues[83]. In this study, we demonstrated that the fibrogenic factor TGF-β1 not only enhances expression of mesenchymal genes implicated in fibrosis, but simultaneously and dramatically attenuates expression of both proximal and distal lung epithelial differentiation markers. The airway organoid method developed here provides a tractable method by which cellular and molecular responses to disease-relevant perturbations can be studied within a 3D tissue-like environment approximating the native lung. Because mixed human cell populations are used, extension of this method to human primary disease-derived cells is straight-forward, and will allow comparison of normal and pathological cell population interactions under baseline and disease modeling conditions. The resulting platform has the potential to become a valuable tool for assessing tissue-, cell- and molecular-level responses, and for evaluating novel therapies and patient-specific interventions. 5. Conclusion Our results demonstrate that multicellular airway organoids derived from adult human cells are able to self-organize and mature toward lung tissue-like structures. Fibroblasts are essential for the self-condensation of aggregates into mechanically stable airway organoids, and likely play a continuing role in reciprocal interactions with endothelial and epithelial cells. YAP was necessary for organization and invasion of tubular structures, and controlled aspects of epithelial differentiation, mimicking key aspects of in vivo lung development. Despite the proximal source of primary epithelium used in the airway organoids, both proximal and distal epithelial markers were expressed over time both in vitro and in vivo, demonstrating remarkable epithelial plasticity within organoid cultures. Successful ectopic engraftment of airway organoids represented an important initial step toward future application in lung regenerative approaches, though many challenges remain. Finally, airway organoids assembled from adult human primary cells represent a new and potentially powerful tool within which to study physiologic and pathologic cell-cell interactions. Supplementary Material 1 2 This work was supported by NIH HL092961, the Caerus Foundation, as well as a Postdoctoral Fellowship Award in Regenerative Medicine and Science from the Mayo Clinic Center for Regenerative Medicine. We thank Dr. Gary C. Sieck and Dr. Y.S. Prakash for providing access to cryosection facility and other equipment support. We thank Dr. Andrew Haak for image analysis support, and Dr. Giovanni Ligresti for helpful discussion. We thank Yunhua Fang and Kristin J. Mantz for their technical support. Figure 1 Airway organoid formation from adult human lung cells. Schematic overview of methods for forming and growing airway organoids in vitro and transplanting into NSG mice in vivo (A). Early formation of a cell aggregate visualized at day 3 by confocal imaging using Cell Tracker labeled cells (Green = HLF, Red = NHBE, Cyan = HMVEC-L) is shown in the inset. Correlation between the number of NHBE cells input and organoid sizes (B). Light microscopy images show the self-condensation of mixed NHBE, HMVEC-L and HLF cells and formation of airway organoids over time (C) NHBE cells alone form small spheres in 3D culture at 21days (D). HLFs only in 3D culture at 21 days (E). Mixed NHBE and HLF 3D culture at 21 days (F). Epithelial lineage marker quantification for NHBE cells (mean ± SEM from 4 individuals) in monolayer culture demonstrate a dominant basal cell (P63+) phenotype prior to organoid culture (G). The highly heterogeneous mechanical properties of airway organoids (day 7) were measured by AFM (I) and correlated with light microscopy and fluorescence images (Green = HLF, H). Values in graph represent mean ± SEM; n=4 (B, G). Scale bars, 200µm (C–F). Figure 2 Self-organization of cells within airway organoids. P63 (red, A), KRT8 (red, B) and E-cadherin (CDH1, red, L and green, H, I, J, K) positive epithelial cells clustered together and were also visualized in close contact with vimentin positive mesenchymal cells in airway organoids at day 7 (A, B, L). CD31 positive cells also clustered together with close interactions with vimentin positive (green) cells in the airway at day 7 (D) and day 14 (E). Relative gene expression of P63, KRT5, KRT14, KRT8, vimentin, CD31 and CDH1 in the airway organoids at different time points was compared to NHBE (C, M), HMVEC-L (F) or HLF (G) cells in traditional 2D monocultures. Proliferation was assessed by ki67 (red) in the airway organoids at various time points (H, I, J, K) and quantified (N). Values in graphs represent mean ± SEM; n=3, P < 0.05; P < 0.01; t-test. Scale bars, 100µm (A–J), 50µm (K). Figure 3 Proximal airway epithelial differentiation in airway organoids. Immunostaining of airway markers CC10 (red) and MUC5AC (green) at day 7, day 14, and day 21 (A–C). Relative gene expression of CC10, CYP2E1, CYP3A5 and MUC5AC in the airway organoids at different time points as compared to NHBE 2D cell culture (D), n=3. ELISA measurements of human MUC5AC production per organoid at each time point, n=5 (E). Airway organoids were stained with ciliated markers acetylated alpha-tubulin (red) and FOXJ1 (green) with low and high power magnification in the airway organoids under control (F, G) and DAPT treatment conditions (H, I). Comparison of relative gene expression of FOXJ1, CC10 and MUC5AC in the airway organoids between control and DAPT treated conditions (J), n=3. Values in graphs represent mean ± SEM; P < 0.05; *P < 0.01; t-test. Scale bars, 100µm (A, B, C, F, H), 50µm (G, I). Figure 4 Expression of distal lung epithelial lineage markers in airway organoids. H&E staining of airway organoid section (A). Co-staining of E-cadherin (red) and Vimentin (green) reveals diverse architecture of epithelial and mesenchymal compartments within airway organoids at day 21 in vitro with low- and high-power imaging (B–D). AECI markers AQP5 (red) and PDPN (green) were seen throughout the airway organoids with low- and high- power magnification (E, F), as well as with E-cadherin (green) in the airway organoids (G). Co-staining AQP5 (red) with CC10 (green) in the airway organoids with low- and high- power magnification (I, J). AECII marker SPC (red) co-staining with Vimentin (green) the airway organoids (H). Co-staining CC10 (red) with HOPX (green) in the airway organoids with low- and high- power magnification (K, L). Relative gene expression of SPC, AQP5, HOPX, and PDPN in airway organoids at different time points compared to NHBE 2D cell culture (M). Values in graphs represent mean ± SEM; n=3, P < 0.05; *P < 0.01; t-test. Scale bars, 200µm (A, B), 100µm (E, G, H, I, K), 50µm (C, D, F, J, L). Figure 5 Invasive tubular structures require localized YAP and actomyosin. Tubular outgrowths from airway organoids invade surrounding matrix shown in the control (A), loss of such structures with blebbistatin treatment (B). Quantification the number of invasive tubular structures (number of buds) and the average diameter of airway organoids in different treatment groups comparing with control (C). Co-staining of YAP/TAZ (red) and E-cadherin (CDH1, green) reveals high expressing YAP/TAZ cells within invasive epithelial tubular structures of airway organoids in the control (D) and their relative loss and disorganization in blebbistatin treatment group (E). Co-staining of YAP/TAZ (red) and Vimentin (green) in the control (F) and blebbistatin treatment group (G). Co-staining of Ki67 (red) and Vimentin (green) in the control (H) and blebbistatin treatment group (I), along with quantification of Ki67 postitive cells (J). Co-staining of Fibronectin (red) and Vimentin (green) in the control (M) and blebbistatin treatment group (N). Relative gene expression of KRT5, KRT14 and CC10 in airway organoids with and without Si-YAP, blebbistatin and cytochalasin D treatment (K, L). Values in graphs represent mean ± SEM; n=3, P < 0.05; *P < 0.01; t-test. Scale bars, 200µm (A–B), 100µm (D, E, F, G, H, I, M, N). Figure 6 TGF-β1 responses in human airway organoids. Tubular outgrowths from airway organoids invade surrounding matrix without TGF-β1 treatment, absence of such outgrowths with TGF-β1 treatment (A). αSMA (green) staining of airway organoids with and without TGF-β1 treatment (B). Relative gene expression of ACTA2, CTGF, FN (C), CC10, FOXJ1, AQP5 (D) and P63, KRT5, KRT14 (E) in airway organoids with and without TGF-β1 treatment. Values in graphs represent mean ± SEM; n=4, P < 0.05; **P < 0.01; t-test. Scale bars, 100µm (A), 50µm (B). Figure 7 Ectopic transplantation of human airway organoids. Human mitochondria (green, A, D), human E-Cadherin (green, B) human Vimentin (green, C), and proliferating cells (Ki67, red, A–D) were assessed by immunostaining at the week 1(A–C) and week 6 (D). Human endothelial cells were assessed by human CD31 (red) co-staining with human E-cadherin (green, E, F) and human Vimentin (green, G, H) at week 1. SPC (red) were used for additional epithelial staining with low- and high- power magnification (I, J), as was double staining for AQP5 (red) and CC10 (green) with low- and high- power magnification (K, L). Scale bars, 200µm (B, C), 100µm (A, D, E, G, I, K), 50µm (F, H, J, L). n=4. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. References 1 Ferkol T Schraufnagel D The global burden of respiratory disease Ann Am Thorac Soc 2014 11 404 406 24673696 2 Rabe KF Hurd S Anzueto A Barnes PJ Buist SA Calverley P Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: GOLD executive summary Am J Respir Crit Care Med 2007 176 532 555 17507545 3 Arcasoy SM Kotloff RM Lung transplantation N Engl J Med 1999 340 1081 1091 10194239 4 Lau AN Goodwin M Kim CF Weiss DJ Stem cells and regenerative medicine in lung biology and diseases Mol Ther 2012 20 1116 1130 22395528 5 Kotton DN Morrisey EE Lung regeneration: mechanisms, applications and emerging stem cell populations Nat Med 2014 20 822 832 25100528 6 Hogan BL Barkauskas CE Chapman HA Epstein JA Jain R Hsia CC Repair and regeneration of the respiratory system: complexity, plasticity, and mechanisms of lung stem cell function Cell Stem Cell 2014 15 123 138 25105578 7 Huang SX Green MD de Carvalho AT Mumau M Chen YW D'Souza SL The in vitro generation of lung and airway progenitor cells from human pluripotent stem cells Nat Protoc 2015 10 413 425 25654758 8 Zuo W Zhang T Wu DZ Guan SP Liew AA Yamamoto Y p63(+)Krt5(+) distal airway stem cells are essential for lung regeneration Nature 2015 517 616 620 25383540 9 Rosen C Shezen E Aronovich A Klionsky YZ Yaakov Y Assayag M Preconditioning allows engraftment of mouse and human embryonic lung cells, enabling lung repair in mice Nat Med 2015 21 869 879 26168294 10 Ott HC Clippinger B Conrad C Schuetz C Pomerantseva I Ikonomou L Regeneration and orthotopic transplantation of a bioartificial lung Nat Med 2010 16 927 933 20628374 11 Price AP England KA Matson AM Blazar BR Panoskaltsis-Mortari A Development of a decellularized lung bioreactor system for bioengineering the lung: the matrix reloaded Tissue Eng Part A 2010 16 2581 2591 20297903 12 Petersen TH Calle EA Zhao L Lee EJ Gui L Raredon MB Tissue-engineered lungs for in vivo implantation Science 2010 329 538 541 20576850 13 Prakash YS Tschumperlin DJ Stenmark KR Coming to terms with tissue engineering and regenerative medicine in the lung Am J Physiol Lung Cell Mol Physiol 2015 309 L625 L638 26254424 14 Zorn AM Wells JM Vertebrate endoderm development and organ formation Annu Rev Cell Dev Biol 2009 25 221 251 19575677 15 Sasai Y Next-generation regenerative medicine: organogenesis from stem cells in 3D culture Cell Stem Cell 2013 12 520 530 23642363 16 Lancaster MA Knoblich JA Organogenesis in a dish: modeling development and disease using organoid technologies Science 2014 345 1247125 25035496 17 Hattori N Cerebral organoids model human brain development and microcephaly Mov Disord 2014 29 185 24375826 18 Takebe T Zhang RR Koike H Kimura M Yoshizawa E Enomura M Generation of a vascularized and functional human liver from an iPSC-derived organ bud transplant Nat Protoc 2014 9 396 409 24457331 19 Dye BR Hill DR Ferguson MA Tsai YH Nagy MS Dyal R In vitro generation of human pluripotent stem cell derived lung organoids Elife 2015 4 20 Huch M Gehart H van Boxtel R Hamer K Blokzijl F Verstegen MM Long-term culture of genome-stable bipotent stem cells from adult human liver Cell 2015 160 299 312 25533785 21 Rock JR Onaitis MW Rawlins EL Lu Y Clark CP Xue Y Basal cells as stem cells of the mouse trachea and human airway epithelium Proc Natl Acad Sci U S A 2009 106 12771 12775 19625615 22 Lee JH Bhang DH Beede A Huang TL Stripp BR Bloch KD Lung stem cell differentiation in mice directed by endothelial cells via a BMP4-NFATc1-thrombospondin-1 axis Cell 2014 156 440 455 24485453 23 Barkauskas CE Cronce MJ Rackley CR Bowie EJ Keene DR Stripp BR Type 2 alveolar cells are stem cells in adult lung J Clin Invest 2013 123 3025 3036 23921127 24 Yui S Nakamura T Sato T Nemoto Y Mizutani T Zheng X Functional engraftment of colon epithelium expanded in vitro from a single adult Lgr5(+) stem cell Nat Med 2012 18 618 623 22406745 25 Takebe T Sekine K Enomura M Koike H Kimura M Ogaeri T Vascularized and functional human liver from an iPSC-derived organ bud transplant Nature 2013 499 481 484 23823721 26 Greggio C De Franceschi F Figueiredo-Larsen M Gobaa S Ranga A Semb H Artificial three-dimensional niches deconstruct pancreas development in vitro Development 2013 140 4452 4462 24130330 27 Thundat T Warmack RJ Chen GY Allison DP Thermal and Ambient-Induced Deflections of Scanning Force Microscope Cantilevers Appl Phys Lett 1994 64 2894 2896 28 Takebe T Enomura M Yoshizawa E Kimura M Koike H Ueno Y Vascularized and Complex Organ Buds from Diverse Tissues via Mesenchymal Cell-Driven Condensation Cell Stem Cell 2015 16 556 565 25891906 29 Wansleeben C Barkauskas CE Rock JR Hogan BL Stem cells of the adult lung: their development and role in homeostasis, regeneration, and disease Wiley Interdiscip Rev Dev Biol 2013 2 131 148 23799633 30 Fulcher ML Gabriel S Burns KA Yankaskas JR Randell SH Well-differentiated human airway epithelial cell cultures Methods Mol Med 2005 107 183 206 15492373 31 Rawlins EL Okubo T Xue Y Brass DM Auten RL Hasegawa H The role of Scgb1a1+ Clara cells in the long-term maintenance and repair of lung airway, but not alveolar, epithelium Cell Stem Cell 2009 4 525 534 19497281 32 Fanucchi MV Murphy ME Buckpitt AR Philpot RM Plopper CG Pulmonary cytochrome P450 monooxygenase and Clara cell differentiation in mice Am J Respir Cell Mol Biol 1997 17 302 314 9308917 33 Rogers DF Airway goblet cells: responsive and adaptable front-line defenders Eur Respir J 1994 7 1690 1706 7995400 34 Rawlins EL Hogan BL Ciliated epithelial cell lifespan in the mouse trachea and lung Am J Physiol Lung Cell Mol Physiol 2008 295 L231 L234 18487354 35 Rock JR Gao X Xue Y Randell SH Kong YY Hogan BL Notch-dependent differentiation of adult airway basal stem cells Cell Stem Cell 2011 8 639 648 21624809 36 Ishikawa H Marshall WF Ciliogenesis: building the cell's antenna Nat Rev Mol Cell Biol 2011 12 222 234 21427764 37 Herriges M Morrisey EE Lung development: orchestrating the generation and regeneration of a complex organ Development 2014 141 502 513 24449833 38 McElroy MC Kasper M The use of alveolar epithelial type I cell-selective markers to investigate lung injury and repair Eur Respir J 2004 24 664 673 15459148 39 Flodby P Borok Z Banfalvi A Zhou B Gao D Minoo P Directed expression of Cre in alveolar epithelial type 1 cells Am J Respir Cell Mol Biol 2010 43 173 178 19767448 40 Kreda SM Gynn MC Fenstermacher DA Boucher RC Gabriel SE Expression and localization of epithelial aquaporins in the adult human lung Am J Respir Cell Mol Biol 2001 24 224 234 11245621 41 Ramirez MI Millien G Hinds A Cao Y Seldin DC Williams MC T1alpha, a lung type I cell differentiation gene, is required for normal lung cell proliferation and alveolus formation at birth Dev Biol 2003 256 61 72 12654292 42 Hoshino A Ishii G Ito T Aoyagi K Ohtaki Y Nagai K Podoplanin-positive fibroblasts enhance lung adenocarcinoma tumor formation: podoplanin in fibroblast functions for tumor progression Cancer Res 2011 71 4769 4779 21610106 43 Treutlein B Brownfield DG Wu AR Neff NF Mantalas GL Espinoza FH Reconstructing lineage hierarchies of the distal lung epithelium using single-cell RNA-seq Nature 2014 509 371 375 24739965 44 Kim CF Jackson EL Woolfenden AE Lawrence S Babar I Vogel S Identification of bronchioalveolar stem cells in normal lung and lung cancer Cell 2005 121 823 835 15960971 45 Kim CF Paving the road for lung stem cell biology: bronchioalveolar stem cells and other putative distal lung stem cells Am J Physiol Lung Cell Mol Physiol 2007 293 L1092 L1098 17693488 46 Suprynowicz FA Upadhyay G Krawczyk E Kramer SC Hebert JD Liu X Conditionally reprogrammed cells represent a stem-like state of adult epithelial cells Proc Natl Acad Sci U S A 2012 109 20035 20040 23169653 47 Mih JD Marinkovic A Liu F Sharif AS Tschumperlin DJ Matrix stiffness reverses the effect of actomyosin tension on cell proliferation J Cell Sci 2012 125 5974 5983 23097048 48 Porazinski S Wang H Asaoka Y Behrndt M Miyamoto T Morita H YAP is essential for tissue tension to ensure vertebrate 3D body shape Nature 2015 521 217 221 25778702 49 Low BC Pan CQ Shivashankar GV Bershadsky A Sudol M Sheetz M YAP/TAZ as mechanosensors and mechanotransducers in regulating organ size and tumor growth FEBS Lett 2014 588 2663 2670 24747426 50 Zhao R Fallon TR Saladi SV Pardo-Saganta A Villoria J Mou H Yap tunes airway epithelial size and architecture by regulating the identity, maintenance, and self-renewal of stem cells Dev Cell 2014 30 151 165 25043474 51 Mahoney JE Mori M Szymaniak AD Varelas X Cardoso WV The hippo pathway effector Yap controls patterning and differentiation of airway epithelial progenitors Dev Cell 2014 30 137 150 25043473 52 Szymaniak AD Mahoney JE Cardoso WV Varelas X Crumbs3-Mediated Polarity Directs Airway Epithelial Cell Fate through the Hippo Pathway Effector Yap Dev Cell 2015 34 283 296 26235047 53 Dupont S Morsut L Aragona M Enzo E Giulitti S Cordenonsi M Role of YAP/TAZ in mechanotransduction Nature 2011 474 179 183 21654799 54 Liu F Lagares D Choi KM Stopfer L Marinkovic A Vrbanac V Mechanosignaling through YAP and TAZ drives fibroblast activation and fibrosis Am J Physiol Lung Cell Mol Physiol 2015 308 L344 L357 25502501 55 Bertero T Cottrill KA Lu Y Haeger CM Dieffenbach P Annis S Matrix Remodeling Promotes Pulmonary Hypertension through Feedback Mechanoactivation of the YAP/TAZ-miR-130/301 Circuit Cell Rep 2015 13 1016 1032 26565914 56 Aragona M Panciera T Manfrin A Giulitti S Michielin F Elvassore N A mechanical checkpoint controls multicellular growth through YAP/TAZ regulation by actin-processing factors Cell 2013 154 1047 1059 23954413 57 Lange AW Sridharan A Xu Y Stripp BR Perl AK Whitsett JA Hippo/Yap signaling controls epithelial progenitor cell proliferation and differentiation in the embryonic and adult lung J Mol Cell Biol 2015 7 35 47 25480985 58 Kendall RT Feghali-Bostwick CA Fibroblasts in fibrosis: novel roles and mediators Front Pharmacol 2014 5 123 24904424 59 Blackwell TS Tager AM Borok Z Moore BB Schwartz DA Anstrom KJ Future directions in idiopathic pulmonary fibrosis research. An NHLBI workshop report Am J Respir Crit Care Med 2014 189 214 222 24160862 60 Leask A Abraham DJ TGF-beta signaling and the fibrotic response Faseb J 2004 18 816 827 15117886 61 Wynn TA Integrating mechanisms of pulmonary fibrosis J Exp Med 2011 208 1339 1350 21727191 62 Bartram U Speer CP The role of transforming growth factor beta in lung development and disease Chest 2004 125 754 765 14769761 63 Lee CG Kang HR Homer RJ Chupp G Elias JA Transgenic modeling of transforming growth factor-beta(1): role of apoptosis in fibrosis and alveolar remodeling Proc Am Thorac Soc 2006 3 418 423 16799085 64 Zeng X Gray M Stahlman MT Whitsett JA TGF-beta1 perturbs vascular development and inhibits epithelial differentiation in fetal lung in vivo Dev Dyn 2001 221 289 301 11458389 65 Hackett TL Warner SM Stefanowicz D Shaheen F Pechkovsky DV Murray LA Induction of epithelial-mesenchymal transition in primary airway epithelial cells from patients with asthma by transforming growth factor-beta1 Am J Respir Crit Care Med 2009 180 122 133 19406982 66 Blanpain C Fuchs E Stem cell plasticity. Plasticity of epithelial stem cells in tissue regeneration Science 2014 344 1242281 24926024 67 Vaughan AE Brumwell AN Xi Y Gotts JE Brownfield DG Treutlein B Lineage-negative progenitors mobilize to regenerate lung epithelium after major injury Nature 2015 517 621 625 25533958 68 Kumar PA Hu Y Yamamoto Y Hoe NB Wei TS Mu D Distal airway stem cells yield alveoli in vitro and during lung regeneration following H1N1 influenza infection Cell 2011 147 525 538 22036562 69 Donati G Watt FM Stem cell heterogeneity and plasticity in epithelia Cell Stem Cell 2015 16 465 476 25957902 70 Volckaert T Dill E Campbell A Tiozzo C Majka S Bellusci S Parabronchial smooth muscle constitutes an airway epithelial stem cell niche in the mouse lung after injury J Clin Invest 2011 121 4409 4419 21985786 71 Watson CL Mahe MM Munera J Howell JC Sundaram N Poling HM An in vivo model of human small intestine using pluripotent stem cells Nat Med 2014 20 1310 1314 25326803 72 Xinaris C Benedetti V Rizzo P Abbate M Corna D Azzollini N In vivo maturation of functional renal organoids formed from embryonic cell suspensions J Am Soc Nephrol 2012 23 1857 1868 23085631 73 Nelson CM Gleghorn JP Sculpting organs: mechanical regulation of tissue development Annu Rev Biomed Eng 2012 14 129 154 22524386 74 Chapman HA Li X Alexander JP Brumwell A Lorizio W Tan K Integrin alpha6beta4 identifies an adult distal lung epithelial population with regenerative potential in mice J Clin Invest 2011 121 2855 2862 21701069 75 Franzdottir SR Axelsson IT Arason AJ Baldursson O Gudjonsson T Magnusson MK Airway branching morphogenesis in three dimensional culture Respir Res 2010 11 162 21108827 76 Ruiz EJ Oeztuerk-Winder F Ventura JJ A paracrine network regulates the cross-talk between human lung stem cells and the stroma Nat Commun 2014 5 3175 24430801 77 Mondrinos MJ Koutzaki S Lelkes PI Finck CM A tissue-engineered model of fetal distal lung tissue Am J Physiol Lung Cell Mol Physiol 2007 293 L639 L650 17526596 78 Mondrinos MJ Jones PL Finck CM Lelkes PI Engineering de novo assembly of fetal pulmonary organoids Tissue Eng Part A 2014 20 2892 2907 24825442 79 Ghaedi M Calle EA Mendez JJ Gard AL Balestrini J Booth A Human iPS cell-derived alveolar epithelium repopulates lung extracellular matrix J Clin Invest 2013 123 4950 4962 24135142 80 Huang SX Islam MN O'Neill J Hu Z Yang YG Chen YW Efficient generation of lung and airway epithelial cells from human pluripotent stem cells Nat Biotechnol 2014 32 84 91 24291815 81 Mammoto T Mammoto A Implantation of fibrin gel on mouse lung to study lung-specific angiogenesis J Vis Exp 2014 82 Ren X Moser PT Gilpin SE Okamoto T Wu T Tapias LF Engineering pulmonary vasculature in decellularized rat and human lungs Nat Biotechnol 2015 33 1097 1102 26368048 83 Baker BM Chen CS Deconstructing the third dimension: how 3D culture microenvironments alter cellular cues J Cell Sci 2012 125 3015 3024 22797912
PMC005xxxxxx/PMC5121069.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 7505689 5980 Mycopathologia Mycopathologia Mycopathologia 0301-486X 1573-0832 27502502 5121069 10.1007/s11046-016-0048-x NIHMS809168 Article RISK FACTORS FOR CRYPTOCOCCAL MENINGITIS — A SINGLE UNITED STATES CENTER EXPERIENCE Henao-Martínez Andrés F. 1 Gross Lilyana 2 Mcnair Bryan 2 McCollister Bruce 1 DeSanto Kristen 3 Montoya Jose G. 4 Shapiro Leland 12345 Beckham J. David 1 1 Department of Medicine, Division of Infectious Diseases, University of Colorado Denver, Denver, CO, USA 2 Department of Biostatistics and Informatics, Colorado School of Public Health, Denver, CO, USA 3 Health Sciences Library, University of Colorado Denver, Denver, CO, USA 4 Division of Infectious Diseases and Geographic Medicine, Stanford University Medical Center, Denver, CO, USA 5 Division of Infectious Diseases, Denver Veterans Affairs Medical Center, Denver, CO, USA Corresponding author: Andrés F. Henao-Martínez, MD, University of Colorado Denver 12700 E. 19th Avenue, Mail Stop B168. Aurora, CO 80045; [email protected] Tel.: +1 (720)-848-0820; fax: +1 (720)-848-0192 11 8 2016 8 8 2016 12 2016 01 12 2017 181 11-12 807814 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Cryptococcal meningitis carries a high mortality. Further understanding of immune suppression factors associated with neuroinvasive infection will improve risk stratification and enhance early diagnosis and treatment with antifungal therapy. The aim of the study was to corroborate established or find novel clinical predictors for cryptococcal meningitis. We performed a matched case-control study of Cryptococcus infection in immunocompromised patients with or without cryptococcal meningitis. All patients with a diagnosis of cryptococcal disease were collected at University of Colorado Hospital between 2000 and 2015 (n=51). Thirty patients were diagnosed with cryptococcal meningitis. We built a logistic regression model for risk factors associated with cryptococcal meningitis. The single predictor univariate model found that a positive blood culture, positive serum cryptococcal antigen, current malignancy, and headaches were significantly associated with cryptococcal meningitis (p= 0.02). In the adjusted multivariate model, central nervous system disease was significantly associated with a diagnosis of HIV infection (OR: 24.45, 95% CI: 1.62 – 350.37; p=0.022) and a positive serum cryptococcal antigen test (OR: 42.92, 95% CI: 3.26 – 555.55; p=0.0055). In patients with HIV infection or a positive serum cryptococcal antigen, the pre-test probability of neuroinvasive Cryptococcus infection is increased and an aggressive diagnostic evaluation should be conducted to exclude infection and consider empiric therapy. Cryptococcal antigens cryptococcal meningitis Cryptococcus neoformans Fungemia HIV risk factors Introduction It is estimated that cryptococcal infection causes approximately 3,400 hospitalizations every year in the U.S.A. alone [1]. Cryptococcal central nervous system (CNS) infection carries a high mortality of approximately 20 % in immunosuppressed patients [2]. Cryptococcus neoformans/C. gattii species complex are the main agents of cryptococcosis. Although infections with C. gattii have been described in the U.S.A [3], they are relatively rare in our region (Mountain West). It is recognized that initial pulmonary exposure through inhalation of spores precedes CNS infection [4, 5]. Dissemination following pulmonary infection is associated with higher mortality [2]. HIV infection with or without AIDS, solid organ transplantation, systemic lupus erythematosus (SLE), malignancy, sarcoidosis, and cirrhosis are immunosuppressive settings known to increase the risk for Cryptococcus dissemination and neuroinvasion [6–10]. Likewise, the presence of fever, headaches, altered mental status (AMS), weight loss, high dose steroid use or evidence of pleural effusion or pulmonary parenchymal infiltrates favor dissemination of Cryptococcus infection in patients with pulmonary disease [11, 6, 12, 13]. Other host factors associated with extra-pulmonary Cryptococcosis are male gender, fungemia, smoking, diabetes and Hispanic ethnicity [14–18]. Many of these predictors of cryptococcal meningitis have not been confirmed. Moreover, characterization of additional clinical features of neuroinvasive disease may permit better prediction of CNS dissemination. Finally, few cohort reports have taken place in the U.S.A. The aim of the study was to perform a matched case-control study of cryptococcal infected immunocompromised patients with or without cryptococcal meningitis infection to corroborate established or find novel clinical predictors for CNS cryptococcal infection. Methods Ethics statement The present project is in health insurance portability and accountability act (HIPAA) compliance according to the Colorado Multiple Institutional Review Board (COMIRB) at University of Colorado Denver. Analyzes of clinical data have been performed under an approved protocol (COMIRB Protocol 15-1340). Patients and data collection All patients with culture that grew Cryptococcus or with positive serum antigens detected were collected at University of Colorado Hospital through the microbiology laboratory between January 2000 and April 2015. Respective medical reports were accessed to collect clinical and laboratory variables for all patients. The following data was collected: demographics (gender, race, age, and occupation); symptoms (constitutional, headaches, altered mental status, respiratory abnormalities, fever, and others); medical history (smoking, lung disease, diabetes mellitus, malignancy, sarcoidosis, cirrhosis, HIV infection, solid organ transplant, use of calcineurin inhibitors or steroids and prednisone dose); HIV (time since diagnosis, history of HIV antiretroviral drug resistance, antiretroviral therapy, CD4 count, and viral load); transplant (type and time since transplant); absence of presence of cryptococcal meningitis, laboratory results (complete cell count, comprehensive metabolic panel, baseline renal function, lumbar puncture opening pressure, serum cryptococcal antigen, cerebrospinal fluid (CSF) cryptococcal antigen, CSF culture, blood culture, CSF cell count, CSF glucose and CSF protein), and outcomes of cryptococcal infection: immune reconstitution syndrome (IRS), treatment regimen, death and attributable death, cognitive deficits, use of ventriculoperitoneal shunts (VPS), new onset cryptococcal infection, and relapse. Definitions Occupation was recorded as written in the history and physical (H&P) report. An occupation was labeled outdoor if it was performed primarily in the open air. Symptoms and past medical history were recorded as written in the initial H&Ps (by medicine residents, attending physicians and/or sub-specialties consults notes). Abstracted constitutional symptoms included weakness, weight loss, fatigue, fever, myalgias, night sweats, and malaise. Recorded respiratory symptoms included cough, shortness of breath, congestion, sore throat and chest pain. Other recorded symptoms included rash, pruritus, and gastrointestinal (GI) complaints (diarrhea, flank pain, hematochezia, nausea, vomiting). Fever was defined in the initial H&P as temperature >37.7 °C. Smoking history was considered current or any former use of tobacco. Lung diseases included chronic obstructive pulmonary disease (COPD), pulmonary embolism, pneumonia, bronchiectasis, obstructive sleep apnea and lung neoplasm. Malignancy included current hematologic or solid organ neoplasms. Prednisone dose was calculated in milligrams and for those on non-prednisone steroids an equivalence converter was used to calculate the corresponding prednisone dose. HIV resistance was defined as written in the HIV provider history note or by the presence of any major mutations on a standard HIV genotype. Cryptococcus infection was identified through Immuno-Mycologics Inc. (IMMY, OK) serum and CSF cryptococcal antigen tests (CrAg® LFA —Cryptococcal Antigen Lateral Flow Assay) using semi-quantitative enzyme immunoassay. Confirmation was done through regular fungal culture. These tests unfortunately cannot distinguish the species or the genotype of the isolate. Blood cultures were processed using the BD BACTEC 9240 automated culturing system. CD4 count, viral load, and laboratory data were obtained at the time of diagnosis with the cryptococcal infection. IRS was defined per previous published guided criteria [19]. Cryptococcal meningitis was defined as a positive cryptococcal CSF antigen study or positive CSF culture or a positive blood cryptococcal culture with endophthalmitis or known history of cryptococcal meningitis. Standard treatment for cryptococcal meningitis was defined as at least 14 days of amphotericin B plus flucytosine. Cryptococcus attributable death was defined as mortality with Cryptococcus infection considered to be the direct cause of death. Residual cognitive deficits were speech or gait abnormalities documented in follow-up assessments more than three months after the episode of cryptococcal meningitis. New onset infection was defined as first episode of cryptococcal infection. Relapse was any episode of recurrence of infection following clinical and microbiological pathogen control. Statistical analysis Statistical analyzes were performed using SAS 9.2 (SAS Institute, Cary, NC, USA). The medians for continuous variables with Inter Quartile Range (IQR) were calculated. For categorical variables, the frequencies and percentages were calculated. We used data from 51 patients, (20 of whom did not contract CNS disease, 30 of whom did contract CNS disease) and one with absent cryptococcal meningitis outcome data. Using previous research [2, 6, 20, 9, 17], we ranked the primary predictor variables by order of clinical importance: HIV infection, use of steroids, transplant, cirrhosis, sarcoidosis, and type 2 diabetes (in order of ranking). There were several combinations of these predictors for which there were no data. To avoid over-parameterization, sarcoidosis, and type 2 diabetes were removed from the list. We created a single predictor model for the remaining variables. Selecting variables having a significance at the 0.02 alpha level, we fit a logistic regression model. We used backward elimination to create our final model that best predicted cryptococcal meningitis in the cohort of 50 patients with cryptococcal infection. Results Total cohort with Cryptococcus infection A total of 51 patients with Cryptococcus infection were obtained (Table 1). The median age was 53.6 years. Patient were predominantly male and Caucasian, and about half had an outdoor occupation. Constitutional symptoms and fever were present in 33 (66%) and 11 (23%) of the patients respectively. The most prevalent comorbidities were smoking (52%), HIV infection (46%), steroid use (29%), malignancy (28%), transplant (18%) and diabetes (16%). Among patients with HIV, the mean time since diagnosis was 5.9 years with only 12% of patients receiving highly active antiretroviral therapy (HAART). Of those on HAART at diagnosis, therapy was started within the year prior to Cryptococcus infection. In the HIV infected group (n=23), prior to cryptococcal infection the median CD4 count and viral loads were 56 cells/µL and 81×103 copies/ml respectively. Significantly abnormal laboratory data included anemia (median hemoglobin of 11.9 g/dl), lymphopenia (median lymphocyte count of 0.9 × 109/L), and hypoalbuminemia (median albumin of 3.1 g/dl). IRS developed as a complication in 10% of the patients. CNS disease (cryptococcal meningitis) In the 30 patients with cryptococcal meningitis, the need for VPS and the presence of cognitive deficits were seen in 8% and 40%, respectively. There was an overall 33% crude mortality with an 18% Cryptococcus attributable mortality. Patients with cryptococcal meningitis had higher mortality attributed to cryptococcal infection (27% vs. 5%). Eighty eight percent of infections were new onset and 14% represented relapses. Among patients with cryptococcal meningitis, there was an increase in opening pressure (median of 24 cm H2O). The positivity rate for serum cryptococcal antigen, CSF cryptococcal antigen, blood culture and CSF culture was 91%, 83%, 54% and 72% respectively. CSF studies revealed a mildly lymphocytic predominant pleocytosis (CSF white blood cell count (WBC) median of 57 × 106/L), hypoglycorrhachia (CSF glucose median of 39.5 mg/dl) and elevated CSF protein (median 89 mg/dl). Predictors of CNS disease A single predictor univariate model for all variables found: significant at the 0.05 level: CSF glucose (OR: 0.93, 95% CI: 0.87 – 0.99; p=0.022), and lung disease (OR: 0.21, 95% CI: 0.046 – 0.93; p=0.04). Significant at the 0.02 level: blood culture positive for Cryptococcus (OR: 15.17, 95% CI: 1.72 – 133.53; p=0.014), serum cryptococcal antigen (OR: 18.33, 95% CI: 3.15 – 106.70; p=0.001), malignancy (OR: 0.15, 95% CI: 0.04 – 0.61; p=0.007), and headaches (OR: 26.92, 95% CI: 3.16 – 229.32; p=0.003) (Table 2). In the single predictor models using primary predictors (described in previous studies), only HIV was found to be statistically significant (OR: 3.66, 95% CI: 1.01 – 13.22; p=0.0476). In addition to HIV infection, we saw a significant association between serum cryptococcal antigen, and CNS Disease (Table 3). The odds ratio of developing cryptococcal meningitis in patients who have HIV was 24.45, 95% CI= (1.6252, 350.3704), p-value= 0.0223 and 42.92, 95% CI= (3.2616, 555.5556), p-value= 0.0055 in patients with positive serum cryptococcal antigens. Discussion In this mixed population in a low cryptococcal meningitis prevalence U.S.A setting —0.4–1.3 per 100,000 people [21] — the presence of cryptococcal serum antigen showed the highest sensitivity for detecting cryptococcal meningitis. Based on these findings, a positive serum test should prompt immediate evaluation of CSF and initiation of empirical cryptococcal meningitis therapies in high-risk patients. This study also supported the male predominance documented with cryptococcal infections. As revealed in other studies, HIV infection, steroids use, and malignancy were also common predisposing factors for cryptococcal meningitis. The HIV infected patients in our cohort were characterized by profound CD4 depletion, high viral loads, and the majority of patients were not taking antiretroviral therapy. The significant presence of immunosuppression contributed to the development of this opportunistic infection. In the entire cohort, lymphopenia was a predominant laboratory abnormality along with markers of chronic illness including hypoalbuminemia and anemia. The mortality in our series of about 30% is in accordance with other studies [22–24]. The observed high rate of sequelae —up to 40% of cognitive deficit — is an uncommonly reported concerning finding. Despite meningeal and central invasion, a weak immune response was echoed by the disproportionately low CSF white count observed. Finally, we found additional clinical factors such as headaches, lung disease and hypoglycorrhachia as potential markers for cryptococcal meningitis. The pathogenesis of cryptococcal meningitis is complex and the mechanisms of neuroinvasion remain elusive. One recognized enhancer of neuroinvasion is impaired cell-mediated immunity (CMI). HIV infection, associated with CMI depression, is one of the most predominant conditions driving the risk for cryptococcal meningitis [1]. Within this vulnerable population with CMI dysfunction, we lack clinical indicators to predict accurately an increased risk of cryptococcal meningitis. Clinical indicators can be divided as (1) preceding modulators increasing the risk of cryptococcal meningitis (e.g. male gender, smoking, steroid use, fungemia, and increased pulmonary tissue invasiveness) or (2) consequence of an already established cryptococcal meningitis pathophysiology (e.g. fever, weight loss, headaches, cryptococcemia, and AMS). Positive Cryptococcus blood culture or a positive cryptococcal serum antigen test are well-recognized markers for cryptococcal meningitis. In HIV infected persons, serum antigen testing has been recommended as a screening test in some settings [25, 26]. We confirmed the presence of some of the already described factors in a low prevalence setting in the U.S.A. Cryptococcosis is estimated to have an overall annual age-adjusted mortality of 0.07 per 100,000 population in 2010 in the U.S.A. [21]. Mortality is highly associated with HIV infection, cirrhosis, malignancy and autoimmune disorders. Globally, the burden of disease is even more worrisome. Cryptococcal meningitis is estimated to affect approximately 1 million persons each year and caused approximately 625,000 deaths [27]. Factors associated with increased mortality include AMS and bacterial co-infections among HIV-negative patients. Other mortality associates include older age, syncope, pneumonia, respiratory failure, ICU admission, fluconazole based therapy, high intracranial pressure and CSF fungal burden (among HIV+ infected patients) [28–30, 24, 2]. This study has multiple limitations. Its retrospective nature limit us to prove the CNS invasion predictors described. Selection bias is also a risk. Other potential factors could have been missed due to the relatively low number of patients studied. This retrospective review of cases of cryptococcal infection at a single institution showed a strong predictor risk rate of cryptococcal meningitis with the diagnosis of HIV/AIDS and a positive cryptococcal serum antigen. These findings highlight the importance of having a high clinical index of suspicion in patients with immunosuppression as well as to have a low threshold to screen patients with a serum cryptococcal antigen even in low prevalence settings. Prospective larger studies are needed to enhance the characterization of clinical predictors of cryptococcal meningitis among vulnerable immunosuppressive populations in the U.S.A. Cryptococcal meningitis can have devastating consequences. Patients are at risk for mortality and disabling sequela. In patients with known immunosuppressive conditions, the presence of an HIV diagnosis, headaches, active malignancy, lung disease, hypoglycorrhachia or a positive serum cryptococcal antigen should alert the possibility of cryptococcal meningitis and trigger aggressive diagnostic workups to rule out disease and to consider empiric antifungal therapy. No funding agencies had any role in the preparation, review, or approval of this manuscript. We thank Daniela Garcia for her help reviewing the manuscript. The views expressed in this article are those of the authors and do not necessarily represent the views of the University of Colorado Denver or Stanford University Medical Center. J.D.B receives research support from the University of Colorado Neurosciences Institute and NINDS 1R01NS097729. Table 1 Clinical characteristics in cases of Cryptococcus infection Patient characteristics Number Total Count (%) Count (%), Median (IQR) Non-CNS (n=20) CNS (n=30) Demographics Gender (Male) 50 41 (82%) 14 (70%) 27 (90%) Race (White) 48 32 (67%) 13 (68%) 19 (66%) Age 51 53.6 (21.6) 60.2 (15.9) 48.2 (25) Occupation (outdoor) 26 14 (54%) 6 (67%) 8 (47%) Symptoms Constitutional 50 33 (66%) 11 (55%) 22 (76%) Headaches 50 18 (36%) 1 (5%) 17 (59%) AMS 50 10 (20%) 1 (5%) 9(31%) Respiratory 50 22 (45%) 14 (70%) 8 (28%) Other Symptoms 50 15 (30%) 3 (15%) 12 (41%) Fever 47 11 (23%) 2 (11%) 9 (32%) Past medical history Smoking (Former and current) 48 25 (52%) 9 (47%) 16 (55%) Lung disease 50 10 (20%) 7 (35%) 3 (10%) DM2 50 8 (16%) 4 (20%) 4 (13%) Malignancy 50 14 (28%) 10 (50%) 4 (13%) Sarcoidosis 50 1 (2%) 0 (0%) 1 (3%) Cirrhosis 50 5 (10%) 2 (10%) 3 (10%) HIV 50 23 (46%) 5 (26%) 17 (57%) Transplant 49 9 (18%) 4 (21%) 5 (17%) Steroid 49 14 (29%) 4 (21%) 10 (33%) Prednisone dose (mg) 13 12.5 (10) 20 (40) 11.25 (7.5) CNI 49 7 (14%) 3 (16%) 4 (13%) HIV and Transplant Time since HIV Diagnosis (y) 20 5.9 (17.7) 5.1 (6.6) 9 (18.5) HIV resistance 6 3 (50%) 1 (100%) 3 (60%) HAART 22 6 (12%) 1 (5%) 5 (17%) Time since HAART (y) 12 0 (0.3) 0 (4.1) 0 (0.4) Viral load (103 copies/ml) 19 81 (211) 39 (187) 83 (137) Type of SOT (Kidney) 10 6 (60%) 2 (40%) 4 (80%) Time since Transplant (y) 9 1.1 (1.3) 0.9 (1.2) 1.1 (3.2) CD4 (cells/µL) 24 56 (117) 14 (92) 60 (123) Laboratory data WBC (4.0–11.1 × 109/L) 50 6.1 (4.5) 5.9 (4.3) 6.5 (4.5) Hemoglobin (14.3–18.1 g/dl) 50 11.9 (4.5) 12.2 (3.9) 11.2 (3.3) Platelets (150–400 × 109/L) 50 193 (165) 216.5 (98) 165 (179) Lymphocytes (1.0–4.8 × 109/L) 48 0.9 (1) 0.9 (0.8) 0.8 (1.8) Monocytes (0.2–0.9 × 109/L) 48 0.4 (0.5) 0.5 (0.4) 0.4 (0.6) Neutrophil (1.8–6.6 × 109/L) 48 3.6 (3.3) 4.2 (3.1) 3.6 (3.5) Eosinophil (0.0–0.4 × 109/L) 48 0 (0.1) 0 (0.1) 0 (0.1) Na (133–145 mmol/L) 49 135 (6) 137 (5) 135 (6) Creatinine (0.7–1.3 mg/dl) 49 1 (0.7) 0.9 (0.4) 1.2 (0.8) Baseline Creatinine 49 1 (0.3) 0.9 (0.1) 1.1 (0.6) Corrected Ca (8.6–10.3 mg/dl) 47 9.4 (0.8) 9.3 (0.7) 9.5 (0.8) Albumin (3.5–5.7 g/dl) 47 3.1 (1) 2.8 (1.4) 3.2 (0.8) Alk. Phosphatase (39–117 U/L) 46 68.5 (35) 68.5 (30) 68 (39) AST (12–39 U/L) 46 25.5 (17) 26 (11) 25 (25.5) ALT (7–52 U/L) 46 24 (21) 22 (11) 24.5 (25) Total Bilirubin (0.1–1.3 mg/dl) 46 0.7 (0.7) 0.7 (0.6) 0.7 (0.8) Opening Pressure (<20 cm H2O) 16 24 (23) 13.5 (0) 24 (23) Serum cryptococcal antigen 39 26 (67%) 6 (35%) 20 (91%) CSF cryptococcal antigen 24 20 (67%) NA 20 (83%) Blood culture 41 16 (39%) 1 (7%) 14 (54%) CSF Culture 25 18 (58%) NA 18 (72%) CSF WBC (0–5 × 106/L) 31 35 (130) 1 (3) 57 (150) CSF PMN (0–2 × 106/L) 31 1.5 (32.5) 0 (0) 5.9 (42.8) CSF LYM (0–4 × 106/L) 31 10.7 (31.9) 0.7 (2.4) 20.6 (32.4) CSF MONO (0–2 × 106/L) 31 1.8 (34.5) 0.1 (0.6) 7.6 (36) CSF glucose (40–70 mg/dl) 30 42.5 (21) 62 (30) 39.5 (20.5) CSF protein (15–45 mg/dl) 31 66 (84) 40 (10) 89 (82) Outcomes IRS 49 5 (10%) 2 (11%) 3 (10%) Treatment (AF) 47 24 (51%) 1 (6%) 23 (79%) Death 49 16 (33%) 7 (37%) 9 (30%) Attributable death 49 9 (18%) 1 (5%) 8 (27%) Cognitive deficits 42 10 (24%) 0 (0%) 10 (42%) VPS 43 2 (5%) 0 (0%) 2 (8%) New onset 50 44 (88%) 20 (100%) 24 (80%) Relapse 36 5 (14%) 0 (0%) 5 (21%) * CNS: Central nervous system; AMS: altered mental status; DM2, diabetes mellitus type 2, CNI: calcineurin inhibitors; y: years; HAART: highly active antiretroviral therapy, SOT: solid organ transplant; CSF: cerebrospinal fluid; IRS: immune reconstitution syndrome; AF: Amphotericin B plus Flucytosine; VPS: ventriculoperitoneal shunt Table 2 Single predictor models of risk factors for cryptococcal meningitis. Characteristic Odds Ratio (95% Confidence Intervals) p-value HIV infection 3.6616 (1.0143, 13.2183) 0.0476 Headaches 26.9168 (3.1593, 229.3184) 0.003 Corticosteroid use 1.8750 (0.4744, 7.4105) 0.3622 CSF glucose 0.9292 (0.8728, 0.9894) 0.022 Lung disease 0.2063 (0.0458, 0.9301) 0.040 Blood culture 15.1666 (1.7227, 133.5252) 0.014 Malignancy 0.1538 (0.03910, 0.6053) 0.007 Transplant 0.7500 (0.1672, 3.3646) 0.7015 SCA 18.3333 (3.1499, 106.7027) 0.001 Cirrhosis 1 (0.1445, 6.9212) 1.0000 SCA: Serum cryptococcal antigen. Table 3 Adjusted model parameter estimates for cryptococcal meningitis. Characteristic Odds Ratio (95% Confidence Intervals) p-value HIV infection 24.4500 (1.6252, 350.3704) 0.0223 Corticosteroid use 8.6655 (0.5558, 135.1351) 0.1193 Transplant 2.4284 (0.1795, 33.8983) 0.4984 Cirrhosis 14.1844 (0.5029, 400) 0.1156 SCA 42.9185 (3.2616, 555.5556) 0.0055 SCA: Serum cryptococcal antigen. Conflicts of Interest: No conflict of interest were reported by the authors. References 1 Pyrgos V Seitz AE Steiner CA Prevots DR Williamson PR Epidemiology of cryptococcal meningitis in the US: 1997–2009 PLoS ONE 2013 8 2 e56269 23457543 2 Brizendine KD Baddley JW Pappas PG Predictors of mortality and differences in clinical features among patients with Cryptococcosis according to immune status PLoS ONE 2013 8 3 e60431 23555970 3 Bartlett KH Cheng PY Duncan C Galanis E Hoang L Kidd S A decade of experience: Cryptococcus gattii in British Columbia Mycopathologia 2012 173 5–6 311 319 21960040 4 Kronstad JW Attarian R Cadieux B Choi J D'Souza CA Griffiths EJ Expanding fungal pathogenesis: Cryptococcus breaks out of the opportunistic box Nat Rev Microbiol 2011 9 3 193 203 21326274 5 Baker RD The primary pulmonary lymph node complex of crytptococcosis Am J Clin Pathol 1976 65 1 83 92 1246992 6 Baddley JW Perfect JR Oster RA Larsen RA Pankey GA Henderson H Pulmonary cryptococcosis in patients without HIV infection: factors associated with disseminated disease Eur J Clin Microbiol Infect Dis 2008 27 10 937 943 18449582 7 Wang LR Barber CE Johnson AS Barnabe C Invasive fungal disease in systemic lupus erythematosus: a systematic review of disease characteristics, risk factors, and prognosis Semin Arthritis Rheum 2014 44 3 325 330 25129259 8 Cancelli I Merlino G Serafini A Valente M Gigli GL Sarcoidosis as risk factor for cryptococcal meningitis in an apparently immunocompetent patient Neurol Sci 2008 29 1 33 35 18379738 9 Lin YY Shiau S Fang CT Risk factors for invasive Cryptococcus neoformans diseases: a case-control study PLoS One 2015 10 3 e0119090 25747471 10 Henao-Martinez AF Beckham JD Cryptococcosis in solid organ transplant recipients Current opinion in infectious diseases 2015 28 4 300 307 26098495 11 Hung MS Tsai YH Lee CH Yang CT Pulmonary cryptococcosis: Clinical, radiographical and serological markers of dissemination Respirology (Carlton, Vic) 2008 13 2 247 251 12 Goldstein E Rambo ON Cryptococcal Infection Following Steroid Therapy Ann Intern Med 1962 56 1 114 120 13899889 13 Olson PE Earhart KC Rossetti RJ Newton JA Wallace MR Smoking and risk of cryptococcosis in patients with AIDS JAMA 1997 277 8 629 630 9039878 14 Sorvillo F Beall G Turner PA Beer VL Kovacs AA Kerndt PR Incidence and factors associated with extrapulmonary cryptococcosis among persons with HIV infection in Los Angeles County AIDS 1997 11 5 673 679 9108950 15 Khan ZU Smoking, melanization, and cryptococcosis: is there a connection? J Clin Microbiol 2006 44 3 1207 16517935 16 Osawa R Alexander BD Lortholary O Dromer F Forrest GN Lyon GM Identifying predictors of central nervous system disease in solid organ transplant recipients with cryptococcosis Transplantation 2010 89 1 69 74 20061921 17 Chuang YM Ku SC Liaw SJ Wu SC Ho YC Yu CJ Disseminated Cryptococcus neoformans var. grubii infections in intensive care units Epidemiol Infect 2010 138 7 1036 1043 19796452 18 Lin KH Chen CM Chen TL Kuo SC Kao CC Jeng YC Diabetes Mellitus is Associated with Acquisition and Increased Mortality in HIV-uninfected Patients with Cryptococcosis: A Population-Based Study J Infect 2016 19 Singh N Perfect JR Immune reconstitution syndrome associated with opportunistic mycoses Lancet Infect Dis 2007 7 6 395 401 17521592 20 Wu G Vilchez RA Eidelman B Fung J Kormos R Kusne S Cryptococcal meningitis: an analysis among 5,521 consecutive organ transplant recipients Transpl Infect Dis 2002 4 4 183 188 12535260 21 Barragan NC Sorvillo F Kuo T Cryptococcosis-related deaths and associated medical conditions in the United States, 2000–2010 Mycoses 2014 57 12 741 746 25135153 22 Lee YC Wang JT Sun HY Chen YC Comparisons of clinical features and mortality of cryptococcal meningitis between patients with and without human immunodeficiency virus infection J Microbiol Immunol Infect 2011 44 5 338 345 21524972 23 Bitar D Lortholary O Le Strat Y Nicolau J Coignard B Tattevin P Population-based analysis of invasive fungal infections, France, 2001–2010 Emerg Infect Dis 2014 20 7 1149 1155 24960557 24 Jarvis JN Bicanic T Loyse A Namarika D Jackson A Nussbaum JC Determinants of mortality in a combined cohort of 501 patients with HIV-associated Cryptococcal meningitis: implications for improving outcomes Clin Infect Dis 2014 58 5 736 745 24319084 25 Jarvis JN Lawn SD Vogt M Bangani N Wood R Harrison TS Screening for cryptococcal antigenemia in patients accessing an antiretroviral treatment program in South Africa Clin Infect Dis 2009 48 7 856 862 19222372 26 Longley N Jarvis JN Meintjes G Boulle A Cross A Kelly N Cryptococcal Antigen Screening in Patients Initiating ART in South Africa: A Prospective Cohort Study Clin Infect Dis 2015 27 Park BJ Wannemuehler KA Marston BJ Govender N Pappas PG Chiller TM Estimation of the current global burden of cryptococcal meningitis among persons living with HIV/AIDS AIDS 2009 23 4 525 530 19182676 28 Pongmekin P Chongtrakool P Santanirand P Kiertiburanakul S Clinical characteristics and mortality risk factors of cryptococcal infection among HIV-negative patients J Med Assoc Thai 2014 97 1 36 43 24701727 29 Sajadi MM Roddy KM Chan-Tack KM Forrest GN Risk factors for mortality from primary cryptococcosis in patients with HIV Postgrad Med 2009 121 2 107 113 19332968 30 Jean SS Fang CT Shau WY Chen YC Chang SC Hsueh PR Cryptococcaemia: clinical features and prognostic factors QJM 2002 95 8 511 518 12145390
PMC005xxxxxx/PMC5121090.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 8100316 1138 Biomaterials Biomaterials Biomaterials 0142-9612 1878-5905 27770631 5121090 10.1016/j.biomaterials.2016.10.010 NIHMS824537 Article Click chemistry improved wet adhesion strength of mussel-inspired citrate-based antimicrobial bioadhesives Guo Jinshan 1# Kim Gloria B. 1# Shan Dingying 1 Kim Jimin P. 1 Hu Jianqing 2 Wang Wei 3 Hamad Fawzi G. 4 Qian Guoying 3 Rizk Elias B. 5 Yang Jian 1* 1 Department of Biomedical Engineering, Materials Research Institute, The Huck Institutes of The Life Sciences, The Pennsylvania State University, University Park, PA 16802, USA 2 School of Chemistry and Chemical Engineering, South China University of Technology, Guangzhou 510640, China 3 Zhejiang Provincial Top Key Discipline of Bioengineering, College of Biological and Environmental Sciences, Zhejiang Wanli University, Ningbo 315100, China 4 Department of Materials Science and Engineering, The Pennsylvania State University, University Park, PA 16802, USA 5 Department of Neurosurgery, College of Medicine, The Pennsylvania State University, Hershey, 17033, USA * Corresponding author: Jian Yang, W340 Millennium Science Complex, University Park, PA 16802. Tel.: (+1) 814-865-1278; [email protected] # These authors contribute equally to this work. 8 11 2016 12 10 2016 1 2017 01 1 2018 112 275286 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. For the first time, a convenient copper-catalyzed azide-alkyne cycloaddition (CuAAC, click chemistry) was successfully introduced into injectable citrate-based mussel-inspired bioadhesives (iCMBAs, iCs) to improve both cohesive and wet adhesive strengths and elongate the degradation time, providing numerous advantages in surgical applications. The major challenge to developing such an adhesive was the mutual inhibition effect between the oxidant used for crosslinking catechol groups and the Cu(II) reductant used for CuAAC, which was successfully minimized by adding a biocompatible buffering agent typically used in cell culture, 4-(2-hydroxyethyl) -1-piperazineethanesulfonic acid (HEPES), as a copper chelating agent. Among the investigated formulations, the highest adhesion strength achieved (223.11 ± 15.94 kPa) was around 13 times higher than that of a commercially available fibrin glue (15.4 ± 2.8 kPa). In addition, dual-crosslinked (i.e. click crosslinking and mussel-inspired crosslinking) iCMBAs still preserved considerable antibacterial and antifungal capabilities that are beneficial for the bioadhesives used as hemostatic adhesives or sealants for wound management. click chemistry bioadhesives mussel citric acid antimicrobial 1. Introduction Approximately 114 million surgical and procedure-based wounds occur every year worldwide and it is expected that the global wound closure market would reach $14 billion by 2018 [1]. Currently, approximately 60% of wound closure procedures are still performed with sutures, staples, and other mechanical methods [2]. However, these methods often require special surgical and/or delivery tools and follow-up visits are needed if nondegradable sutures are used, and sometimes liquid or air leakage is inevitable. Therefore, bioadhesives have obtained much attention as a cost-effective technology for wound closure. In the past three decades, bioadhesives have transformed clinical surgical practices in both external and internal lacerations by promoting tissue healing while preventing blood loss [3–7]. However, most current adhesives (i.e. a biologically-derived fibrin glue [4, 7, 8]) cannot provide sufficient mechanical support especially when used in wet environments during surgical procedures. Inspired by the adhesion strategy of marine mussels, a new family of biomimetic, mussel-inspired adhesives has become an area of intense research. Mussel-inspired polymers are synthesized from catechol-containing amino acids such as L-3,4-dihydroxy -phenylalanine (L-DOPA), typically derived from various mussel adhesion proteins, known to contribute to the strong wet adhesion strength of marine mussels to non-specific surfaces [5, 9–13]. Among those polymers, injectable, citrate-based, mussel-inspired bioadhesives (iCs) [11] and antimicrobial iCs [12] developed in our group have been acknowledged for their cost-effective and convenient syntheses along with vastly improved wet adhesion strength as compared to that of fibrin glue. However, mussel-inspired polymers including iCs commonly suffer from insufficient cohesive strength under wet conditions, as polymers can easily detach from adhered surfaces by deformation or stretching [10–13]. Strong wet mechanical strength is particularly vital for in vivo applications, hence our goal was to further chemically modify iCs to optimize this property. The catechol groups in previously reported formulations of iCs participate in the formation of crosslinked polymer networks, reducing the number of available catechol groups that can chemically react with wet tissues [10–13]. Moreover, the rapid degradation rate of iCs remains a significant challenge as it may lead to undesirable early structural collapse and failure of wound closure. In our previous study, click chemistry was introduced into a citrate-based, polyester elastomer to yield a mechanically robust, surface-clickable, and biodegradable poly (1, 8-octanediol citrate) (POC-click) [14–18]. Propelled by the successful incorporation of click chemistry into POC-click materials, click chemistry was introduced in order to synthesize more mechanically robust iCs. The rationale behind modifying iCs with click chemistry are as follows: 1) the triazole rings formed by click chemistry could imitate amide bonds and serve as cohesive strength-improving moieties in polyester-based iCs; 2) click crosslinking could spare more catechol groups from network formation to participate in adhesion, thus further improving adhesion strength; 3) to slow down the initial degradation rate of iCs and sustain mechanical integrity in the early tissue regeneration stage by using click chemistry; 4) to achieve first-slow-then-fast degradation without prolonging total degradation time via click modification (similar to POC-click) [14]; and 5) to enable convenient bioconjugation after or concurrent with crosslinking. One major concern in applying click chemistry for biomedical applications is that the copper catalysts often used in click reactions can also catalyze the production of reactive oxygen species (ROS), which is harmful to cells [19]. In addition, copper-free thermal click reactions used to produce the aforementioned POC-click materials typically occur at 100°C so they are not suitable for crosslinking bioadhesive polymers due to the low use temperature of bioadhesives, usually around the body temperature of 37°C. An alternative to copper-free click chemistry is strain-promoted azide-alkyne cycloaddition (SPAAC), used for forming hydrogels at room or body temperature [20–22], though strain-promoted triple-bond containing reagents are not readily available with this method. With its high reaction rates and nearly quantitative yields, CuAAC has still been widely used for synthesizing hydrogel-based biomaterials and encapsulating cells [23–29]. To reduce the cytotoxicity of copper catalyst systems, water-soluble copper-chelating ligands/agents have been used, as they include copper-chelating azides [19], bis(L-histidine) [30], tris-(hydroxypropyltriazolylmethyl)amine [31], bipyridine [32], and bis[(tert -butyltriazoyl)methyl]-[(2-carboxymethyltriazoyl)-methyl]-amine (BTTAA) [33]. All of these ligands accelerate the CuAAC reaction and act as sacrificial reductants, helping to protect cells and bioactive molecules from the damage by ROS produced by the Cu-catalyzed reduction of oxygen [19, 34]. However, the toxicity and low water-solubility of some of these copper-chelating ligands/agents still remain problematic. Here, we present a dual crosslinking iC system as a new strategy to address the challenges related to the low cohesive strength and rapid degradation of existing mussel-inspired adhesives. In detail, azide- or alkyne-functionalized iCs were synthesized by adding pentaerythritol triazide (3N3) or trimethylolethane dipropiolate (2PL) monomers to the side groups of iCs to produce iC-3N3 or iC-2PL, respectively (Scheme 1). Alkyne-functionalized gelatin (Gelatin-Al, GL) (Scheme S1) was also synthesized and introduced into the iC-X (X = click) system to further improve the mechanical strength and biocompatibility of the system. The crosslinking of iC-X system can be initiated by the use of either an oxidant (ie. sodium (meta) periodate (PI) for mussel-inspired crosslinking) or a copper catalyzed click crosslinking, or both (dual crosslinking, Scheme 2). To resolve the foreseen mutual inhibition effect between PI (oxidant) and Cu(II) reductant (sodium L-ascorbate, NaLAc, to reduce Cu(II) into Cu(I)), a zwitterionic organic chemical buffering agent, HEPES, was introduced into the system as it is widely used in cell culture as a buffering agent. The two adjacent tertiary amine structures in HEPES as found similarly in the chemical structures of previously used copper-chelating agents, can chelate copper [19, 30–35]. With the use of HEPES, the mutual inhibition effect between PI and NaLAc was successfully minimized and the cytotoxicity caused by the copper catalyst was also significantly reduced. The dual-crosslinked iC-X-PI bioadhesives showed significantly enhanced tensile (cohesive) and adhesive strength, as well as prolonged degradation, making mussel-inspired adhesive polymers well suited for many surgical adhesion applications. It is also worthwhile to note that the iC-X-PI bioadhesives exhibited intriguing antimicrobial properties due to the presence of inherently anti-microbial citrate and PI [12]. 2. Experimental Section 1-Ethyl-3-(3-dimethylaminopropyl)-carbodiimide (EDC) and pentaerythritol tribromide were purchased from AK Scientific, Inc. (Union City, CA, USA). Propiolic acid was purchased from Carbosynth Ltd. (Compton, UK). All other chemicals were purchased from Sigma-Aldrich and were used without further purification. 2.1 Characterization Synthesized pre-polymers and their monomer components were characterized by 1H-NMR spectra in DMSO-d6 and CDCl3 respectively using a 300 MHz Bruker DPX-300 FT-NMR spectrometer, as well as by Fourier transform infrared (FTIR) spectra using a Nicolet 6700 FTIR spectrometer, cast in acetone on KBr slices. UV-vis spectra were also recorded on a UV-2450 spectrometer (Shimadzu, Japan). 2.2 Monomer and Polymer Syntheses The synthesis of click chemistry enhanced iCs (clickable iC) involves three steps: synthesis of the 3N3 monomer, synthesis of the 2PL monomer, and then synthesis of functionalized pre-polymers by polycondensation in a facile, one-pot reaction of citric acid (CA), poly(ethylene glycol) (PEG) diol, a catechol-containing compound such as dopamine or L-DOPA, along with either the 3N3 monomer or the 2PL monomer to achieve the azide (iC-3N3) pre-polymer or the alkyne (iC-2PL) pre-polymer respectively. 2.2.1 Pentaerythritol triazide (3N3) monomer synthesis To synthesize the 3N3 monomer, we modified the synthesis process of 2, 2-Bis(azidomethyl)propane-1,3-diol (diazido-diol monomer, DAzD), which contains similar azide group functionality. [14, 36–38]. Briefly, pentaerythritol tribromide (32.48 g, 0.10 mol) and sodium azide (NaN3) (26.87 g, 0.42mol) were heated at 120°C overnight in a round-bottom flask with 120 mL of dimethylformamide (DMF), stirring. Following, we removed the solvent at 80–120°C under vacuum, re-dissolved the crude product in acetone, filtered out the salt, and then removed acetone by rotary evaporation. We then extracted our product by dissolving in dichloromethane (DCM, 200 mL) and saturated saline (50 mL×3), so that the organic phase was separated and dried by anhydrous magnesium sulfate (MgSO4). After filtration and rotatory evaporation, we obtained the final product (19.9 g, 94% yield). 1H NMR (Figure S1, 300 MHz; CDCl3; δ, ppm) of 3N3 monomer: 3.38 (3, s, -CH2-N3), 3.54 (1, s, -CH2-OH). FTIR of 3N3 monomer (Fig. 1A, cast on KBr, cm−1): 2103 (strong, -N3). 2.2.2 Trimethylolethane dipropiolate (2PL) monomer synthesis We adapted the synthesis of 2PL monomer from literature [39], in which 1,1,1-tris(hydroxymethyl)propane (13.4 g, 0.1 mol) and propiolic acid (14.01 g, 0.2 mol) was dissolved in 120 mL dry toluene and placed in a dry 250 mL round-bottom flask equipped with a Dean-Stark trap and a magnetic stir bar, adding four drops of sulfuric acid (H2SO4) as catalyst and heating to 125°C to reflux for 24 hours. We performed purification by removing toluene, re-dissolving the crude product in ethyl acetate (150 mL), washing with 5% sodium bicarbonate (NaHCO3) solution (30 mL×2) and shortly after with a brine solution (30 mL×2). Thereafter, we dried our product over anhydrous MgSO4, filtered, and removed the remaining solvent by rotary evaporation, so that the final product was obtained as slightly yellow oil (21.9 g, 92% yield). 1H NMR (Figure S1, 300 MHz; CDCl3; δ, ppm) of 2PL monomer: 0.91 (br, -CH2-CH3), 1.48 (br, -CH2-CH3), 2.94 (s, -OOC≡CH), 3.55 (br, -CH2-OH), 4.18 (br, -CH2-OOC≡CH). FTIR of 2PL monomer (Fig. 1B, cast film on KBr, cm−1): 2118 (strong, -OOC≡CH), 1709 (-CH2-OOC≡CH). 2.3 Clickable iCMBA pre-polymer synthesis The general synthesis of iC pre-polymers is adapted from our previous work [6, 11, 12], to which we introduced clickable functionalities in this work. For example, iC-P4-3N3 denotes a pre-polymer, in which P4 refers to the molecular weight of PEG (PEG-400) and 0.2 indicates the feed ratio of 3N3 monomer to PEG used in the formulation. Since the feeding ratio of dopamine to PEG used in all of our formulations was 0.3, it was omitted in the formula names for convenience. Briefly, we heated citric acid (CA) and PEG400 with a monomer ratio of 1.2:1 in a round-bottom flask at 160°C to obtain a clear mixture, into which we added dopamine (0.3 molar ratio to PEG) under N2 at a reduced temperature of 140°C. At a further reduced temperature of 120°C, we next added the 3N3 monomer (0.2 molar ratio to PEG), reacting for ~36 hours under vacuum and quenched with deionized (DI) water. The dissolved pre-polymer was purified by dialysis against water (molecular weight cut-off (MWCO) of 1 kDa), and then freeze-dried to obtain the purified pre-polymer iC-P400D0.3-3N3-2. FTIR of iC-P4-3N3 (Fig. 1A, by casting pre-polymer solution in acetone on KBr slice, cm−1): 2103 (-N3) and 1723 (COO-), 1634 (CONH-). 1H NMR of iC-P4-3N3 (Fig. S1A, 300 MHz; DMSO-d6; δ, ppm): 2.69–3.03 (m, -CH2- from CA), 3.30–3.78 (br, -(CH2)2-O- from PEG and -CH2- from 3N3 monomer), 4.10, 4.16 (br, -O-CH2-CH2-OOC- from the terminals of PEG that connected to ester groups), 6.26, 6.49, 6.65, 8.83 (m, protons from dopamine). The iC-P4-2PL (pre-polymer with alkyne groups) was synthesized likewise, using 2PL instead of 3N3 in equal molar ratios. FTIR of iC-P4-2PL (Fig. 1B, by casting pre-polymer solution in acetone on KBr slice, cm−1): 2123 (-OOC≡CH) and 1730 (COO-), 1628 (CONH-). 1H NMR of iC-P4-2PL (Fig. S1B, 300 MHz; DMSO-d6; δ, ppm): 0.79 (br, -CH2CH3 from 2PL monomer), 2.65–2.96 (m, -CH2- from CA, -OOC ≡CH from 2PL monomer), 3.36, 3.53 (br, -(CH2)2-O- from PEG and -CH2- from 2PL), 4.10, 4.16 (br, -O-CH2-CH2-OOC- from the terminals of PEG that connected to ester groups), 6.26, 6.49, 6.65, 8.72. 8.85 (m, protons from dopamine). In order to increase the crosslinking density of iCs, PEG400 and PEG400/PEG200 mixture (mol/mol = 1/1) were also used in this paper, and the compositions of the obtained pre-polymers, iC-P2/4-3N3 and iC-P2/4-2PL, are listed in Table 1. Normal iCMBA pre-polymer iC-P4 was also synthesized as control. The successful incorporation of dopamine into iCMBA and clickable iCMBA pre-polymers was also proven by 1H-NMR (Fig. 1A and B) and UV-vis spectra (Fig. 1D), and the dopamine contents in these pre-polymers were determined by UV-vis spectra (Fig. 1D and Table 1). 2.4 Alkyne functional gelatin (Gelatin-Al, GL) synthesis To prepare GL as outlined in Scheme S1, we first synthesized monopropargyl succinate (Al-COOH) through a ring opening reaction of succinic anhydride (20 g, 0.2 mol) using propargyl alcohol (11.64 g, 0.2 mol) and a catalyst, 4-dimethylaminopyridine (DMAP) (0.98 g, 0.008 mol, 4mol% to alcohol) (Scheme S1A) in 120 mL acetone in a 250 mL round-bottom flask. After 8 hours of refluxing at 80°C, acetone was removed under vacuum, and the crude product was re-dissolved in EtAc (200 mL) and washed with brine (30 mL×3). The organic phase was dried over anhydrous MgSO4, filtered, and the solvent was removed by rotary evaporation. After conducting recrystallization in EtAc/hexane (v/v = 1/1) three times, the final product was obtained as a slightly yellow crystal (30.0 g, 96% yield). 1H NMR (Fig. 1C, 300 MHz; CDCl3; δ, ppm) of monopropargyl succinate: 2.51 (s, -C≡CH), 2.66 (s, HOOC-CH2-CH2-COO-), 4.68 (s, -CH2-C≡CH), 9.46 (br, HOOC-CH2-). Next, gelatin-Al was synthesized by reacting Al-COOH and gelatin in dimethyl sulfoxide (DMSO) in the presence of N-hydroxysuccinimide (NHS) and EDC (Scheme S1B). This required an initial activation of Al-COOH (1.56 g, 10 mmol) by NHS (1.16 g, 10.1 mmol) and DMAP (0.122 g, 1 mmol) dissolved in 15 mL of DMSO. Next, EDC (1.94 g, 10.1 mmol) was added at 0°C; the reaction mixture was allowed to warm to room temperature and to be reacted for another 24 hours. Then, gelatin (5 g, from bovine skin, Sigma) dissolved in 40 mL DMSO was added to the NHS activated Al-COOH solution, and the reaction mixture was stirred for another 24 hours. After dialysis (MWCO: 1000Da) against DI water for 3 days following by freeze-drying, GL was obtained as a pale yellow powder (10.8 g, yield: 94%). The 1H NMR and FTIR spectra of GL and unmodified gelatin are shown in Fig. 1C and S2. 2.5 Crosslinking of Clickable iCs and Setting Time Measurement Three different routes were conducted on the iC-3N3/2PL pre-polymer (equal-weight mixture of iC-3N3 and iC-2PL): oxidation of catechol groups by sodium (meta) periodate (PI); CuAAC (click reaction); and dual crosslinking by PI and CuAAC. 2.5.1 iC-3N3/2PL mixture crosslinking by PI (iC-BD-PI, BD means blending) The crosslinking of iC-3N3/2PL equal-weight mixture by PI adapted from our previous work [6, 11, 12]. For 2 g of 50 wt% iC-3N3/2PL pre-polymer solution, 2 mL of 8wt% PI solution was used. The gel times of iC-P4-3N3/2PL mixture and iC-P2/4-3N3/2PL mixture cross-linked by 8wt% PI are listed in Table 2. 2.5.2 Crosslinking of iC-3N3/2PL mixture wtih CuAAC (iC-X) Since an iC-3N3/2PL mixture contains both azide and alkyne groups, it can be crosslinked through CuAAC. The obtained crosslinked gel is named iC-X for convenience. CuSO4 and sodium L-ascorbate (NaLAc) catalyst system that can generate Cu+ ions in situ were used for the CuAAC. For example, for 2 g of 50 wt% iC-3N3/2PL pre-polymer solution, a 2 mL solution containing 5 mM CuSO4 and 50 mM NaLAc was used as a copper catalyst. For varied concentrations of CuSO4 and ratios of NaLAc/CuSO4, their corresponding gel times were recorded as listed in Fig. 2A. Considering that the use of oxidant, such as PI in the dual crosslinking (PI and click) process, will affect the reduction of Cu2+ to Cu+ by the NaLAc. To reduce the side effects of the oxidant to click reaction, HEPES, a biocompatible buffer agent often used in cell culture, was used as a copper ion chelating agent to protect and stabilize the reduced Cu+ ions. The effect of HEPES to the gel times was investigated as shown in Fig. 2B. The optimized HEPES concentration was determined to be 20 mM, which generated the fastest crosslinking in our investigated formulations for only using CuSO4-NaLAc. 2.5.3 Dual crosslinking of iC-3N3/2PL mixture with PI and CuAAC (iC-X-PI) The dual crosslinking of iC-3N3/2PL mixture was conducted at the presence of both copper catalyst and PI. For example, 2 g of 50 wt% iC-3N3/2PL pre-polymer solution, click catalyst and oxidant solution with a total volume of 2 mL were added separately and combined in the final solution with 5 mM CuSO4, 50 mM NaLAc, 20 mM HEPES, and 4wt% PI. The effect of HEPES to the gel times of iC-X-PI was investigated at room temperature (Fig. 2C). The gel times of iC-X-PIs cross-linked at 37°C are shown in Fig. 2D. 2.6 Rheological Tests Rheological tests were undertaken by using an AR-G2 rheometer (TA Instruments, UK) in a parallel plate configuration, by employing sandblasted stainless steel 40 mm diameter plates throughout and by using a Peltier plate for temperature control. In a typical rheological test for gelling kinetics, 800 μL of iC-P4-BD (equal-weight blend 50wt% solution in DI water of iC-P4-3N3 and iC-P4-2PL) was mixed with 800 μL 8wt% PI solution in DI water (or copper catalyst, or the mixture of copper catalyst and PI, the final volume was 800 μL) and the mixture was applied to the lower plate, which was preheated to 25°C. The upper plate was immediately brought down to a plate separation of 0.5 mm and the measurement was taken. A low frequency of 1 Hz and 1% strain was applied to minimize interference with the gelation process and to keep the measurement within the linear viscoelastic region. The gelation kinetics was measured in a time sweep by the monitoring of the change of storage (G′) and loss (G″) moduli as a function of time. All measurements were repeated in triplicates. 2.7 Properties of Cross-linked Clickable iCs To investigate the properties of clickable iCs crosslinked by PI, CuAAC, or both, we evaluated mechanical properties according to ASTM D412A (Instron, Norwood, MA) [6, 11, 12]. The iC samples at both dry and wet states (swollen in water for 4 hours) were cut into strips (25 mm × 6 mm × 1.5 mm, length × width × thickness), placed in the mechanical tester, and pulled to failure at a rate of 500 mm/min. The adhesion strengths of different clickable iC formulations were measured using the lap shear strength test according to a modified ASTM D1002-05 [6, 11, 12]. Briefly, porcine-derived, acellular small intestine submucosa (SIS) material strips (40×4 mm) (OASIS®, HealthPoint Ltd. TX) were adhered to the terminals of albumin strips (40×100 mm) using superglue (3M Scotch). The clickable iC formulations (20 μL in total volume) were in-situ mixed and applied onto the two SIS terminals of two albumin strips, which were pre-soaked in PBS (pH 7.4) to mimic the wet environment of clinical settings. The two strips were pressed against each other to form a bond and the attached strips were placed in a highly humid chamber for 2 hours. The lap shear strength of bonded strip specimens was subsequently measured using Instron mechanical tester (Norwood, MA) fitted with a 10 N load cell at a crosshead speed of 1.3 mm/min. Detailed methods for evaluating sol contents, swelling ratios, and degradation profiles have been described previously [6, 11, 12]. Degradation study was conducted by incubating dried crosslinked samples in PBS (pH 7.4) at 37°C. Mass loss was recorded at preset time points. 2.8 Cyto-compatibility Evaluations of Clickable iC Pre-polymers and Cross-Linked Clickable iC Hydrogels The biocompatibility evaluations of clickable iC pre-polymers, GL, and cross-linked clickable iC hydrogels were conducted similar to our previous work using Human-derived mesenchymal stem cells (hMSCs, ATCC® PCS-500-012TM) as cell model. Cells at passages 5–10 were used for cytotoxicity study in this work. [6, 11, 12], in which sol contents and degradation products of clickable iC hydrogel, crosslinked by PI, CuAAC, or both were evaluated for cytotoxicity by respectively setting iC-P4 PI 8 wt% or commercially available PLGA (LA/GA=50/50, Mw~60KDa, purchased from Polyscitech) as control [11]. To evaluate the cytotoxicity of pre-polymers, 10, 1, and 0.1 mg/mL pre-polymer solutions were prepared by directly dissolving pre-polymers in complete Dulbecco’s modified eagle’s medium (DMEM, growth media, MG). Poly(ethylene glycol) diacrylate (molecular weight ~700Da) was used as control. The sol content solutions or degradation products of different formulations were obtained by incubating equal mass (0.5 g) hydrogel specimens in 5 mLs of PBS (pH 7.4, for sol content) at 37°C for 24 hours (sol) or in 5mLs of 0.1 M NaOH solution (for degradation product) till full degradation (degradation), respectively. Subsequently, three different dilutions of degraded products or leachable parts were prepared: 1×, 10× and 100× (1× was the solution of degradation products or leached parts with no dilution; 10× and 100× indicate 10-fold and 100-fold dilutions of 1× solution diluted with PBS, respectively). 200 μL of hMSC suspension in MG medium (5×104 cells/mL) was dispensed in each well of a 96-well plate and incubated for 24 hours. Then, 20 μL of sol content/degradation product solutions at various concentrations were added into the culture media in the plate and the cells were incubated for another 24 hours before their numbers were quantified with a MTT assay. All the solutions above were adjusted to neutral pH and filtered with sterile 0.2 μm filters prior to use for cell culture. We also studied cell attachment and proliferation of hMSC cells on crosslinked iC-P2/4-X-PI 4wt% films using the process described previously [6, 11, 12]. 2.9 Antimicrobial Performance of Cross-Linked Clickable iC Hydrogels The antimicrobial performance, including short-term and long-run antibacterial and antifungal performance, of crosslinked clickable iC hydrogels was conducted using Staphylococcus aureus (S. aureus) and Escherichia coli (E. coli) as respective positive and negative bacteria models, and by using Candida albicans (C. albicans) as a fungi model. 2.9.1 Bacterial incubation Staphylococcus aureus (S. aureus, ATCC® 6538™) and Escherichia coli (E. coli, ATCC® 25922™) were purchased from ATCC (American Type Culture Collection) and used per safety protocols. Tryptic soy agar (Cat. #: C7121) and tryptic soy broth (Cat. #: C7141) for culturing S. aureus were purchased from Criterion. Luria broth base (LB broth, Cat. #: 12795-027) and select agar (Cat. #: 30391-023) for culturing E. coli were purchased from Invitrogen. S. aureus and E. coli were cultured at 37°C in sterilized tryptic soy broth and LB broth, respectively with a shaking speed of 150 rpm in a rotary shaker for 24 hours and the obtained bacteria suspensions were diluted into desired concentrations before use. 2.9.1.1 Anti-bacterial performance of crosslinked clickable iC hydrogels We measured the kinetics of bacterial inhibition according to our method described previously [12]. Briefly, S. aureus and E. coli were cultured onto crosslinked clickable iC hydrogels using iC-P2/4-BD-PI 4wt%, iC-P2/4-X, and iC-P2/4-X-PI 4wt% as the representative experimental groups. As controls, we used iC-P4-PI 4wt% and PEGDA/2-Hydroxyethyl methacrylate (HEMA) hydrogels (1:1 w/w for PEGDA to HEMA, with PEGDA ~ 700 Da) [40, 41], as well as commercially available Hydrofera Blue bacteriostatic foam dressing [42]. Briefly, 0.2 g of test sample was immersed in 20 mL of a germ-containing broth solution with a bacterial concentration of 100× diluted from optical density (OD) at 600 nm around 0.07. The mixture was then incubated at 37°C with shaking speed of 150 rpm. 200 μL of the mixture was taken out at each pre-set time-point, and the OD value of the medium at 600 nm was recorded by a microreader (TECAN, infinite M200 PRO). We calculated inhibition ratios of hydrogels or Hydrofera Blue according to equation (1): (1) Inhibitionratio(%)=100-100×At-A0Acon-A0 where Ao is the starting optical density (bacteria in media prior to incubation), which is incubated at 37°C for the predetermined time points to obtain Acon (pure medium control) or At (medium containing hydrogel). 2.9.1.2 Long-term anti-bacterial evaluation of crosslinked clickable iC hydrogels We also simulated long-term antibacterial capability of our hydrogels by evaluating degradation products and periodical release solutions against S. au and E. coli based on the bacterial inhibition ratio test, described previously [12]. 2.9.2 Anti-fungal study Fungi (Candida albicans, C. albicans) purchased from ATCC (ATCC® 10231™) were used per safety protocols. Yeast malt (YM) medium broth (Lot #: 1964C030) and agar (Lot #: 1964C030) for fungi culture were purchased from Amresco and Acumedia, respectively. Tween 20 added to the YM broth medium at a final concentration of 0.5wt% was sterilized before use. C. albicans was maintained on YM agar plates in all cases. For experiments, C. albicans cells were scraped from YM agar plates and dispersed in a Tween 20 (0.5 wt%)-containing YM broth. The cells were counted with a hemocytometer and were diluted to obtaine a final fungi concentration of 0.5–1×107 cells/mL [12]. The measurement of the exact amount of fungi used a colony growth assay on YM agar plates, which will be described in detail below. 2.9.2.1 Anti-fungal study performance of cross-linked clickable iC hydrogels Following incubation of C. albicans [12, 43], we evaluated direct exposure antifungal effects of cross-linked clickable iC hydrogels based on iC-P2/4-BD-PI 4wt%, iC-P2/4-X, and iC-P2/4-X-PI 4wt% as the representative experimental groups and iC-P4-PI 4wt%, PEGDA/HEMA hydrogels and Hydrofera Blue as controls [12]. 2.9.2.2 Long-term anti-fungal evaluation of crosslinked clickable iC hydrogels We also simulated long-term antifungal capability of our hydrogels by evaluating degradation products and periodical release solutions against C. albicans based on the bacterial inhibition ratio test, described previously [12]. 3. Results and Discussion 3.1 Synthesis and characterization of clickable iC pre-polymers and GL Challenges to mussel-inspired polymers include fast oxidation of catechol groups on L-DOPA or its derivatives (i.e. dopamine) if left unprotected under neutral or basic conditions [2, 44, 45]. Thus our approach based on iCs conveniently provides L-DOPA or dopamine with a slightly acidic environment mainly imparted by citric acid, enabling facile and low-cost one-pot polycondensation reactions without further protection of these groups [11, 12]. At the same time, our iC platform introduces versatile clickable functionalities by the same polycondensation method. In order to maintain the elasticity of modified iCs and to take into consideration the large reactivity difference between long-chain PEG and short-chain azide/alkyne functionalized diols used in POC-click development [14], azide- or alkyne-functionalized monohydroxyl compounds (Scheme 1A, B) were introduced concurrently with dopamine, as functional monomers, into the side chains of iCs to form azide- or alkyne-functionalized iC, iC-3N3 or iC-2PL pre-polymers, respectively (Scheme 1C). The FTIR spectra of representative iC-3N3 and iC-2PL pre-polymers, iC pre-polymers, and corresponding 3N3 and 2PL monomers are shown in Fig. 1A and 1B, respectively. The successful introduction of 3N3 monomers into the iC-3N3 pre-polymer was confirmed by the characteristic infrared absorption peak of azide at 2100 cm−1, which matches the azide peak of 3N3 monomers (Fig. 1A). Likewise, the appearance of characteristic infrared peak of alkyne (around 2123 cm−1, Fig. 1B) and the 1H NMR peak of the protons –CH2CH3 (around 0.93 ppm, Fig. S1 right) from 2PL monomer confirmed the incorporation of alkyne groups into iC-2PL. Similar multiple peaks at 6.4–6.7 ppm in the 1H NMR spectra of both clickable iCs and normal iC were assigned to the protons of phenyl group from dopamine [11, 12], indicating that click functionalization of iC did not affect the function of catechol groups (Fig. S1). The dopamine contents of clickable iCs and iC pre-polymers were determined by measuring the UV-Vis absorbance at 280 nm, then comparing to the dopamine standard curve (Table 1). To maintain the molar ratios of dopamine: PEG at 0.2: 1, the amount of dopamine was reduced accordingly for the iCs made from PEG with high molecular weights (Fig. 1D and Table 1). The UV-Vis absorbance spectra of all clickable iC and iC pre-polymers are shown in Fig. 1D. GL was also synthesized as illustrated in Scheme S1. The modification of gelatin with alkyne groups was confirmed by proton peaks from -CH2-C≡CH in the 1H NMR spectrum of GL (a in Fig. 1C). There was no significant difference between the FTIR spectra of GL and gelatin (Fig. S2). 3.2 Crosslinking of clickable iC pre-polymers by CuAAC, PI, or both Possessing both catechol groups and click functionalities, the mixture of iC-3N3 and iC-2PL pre-polymers can undergo mussel-inspired intermolecular crosslinking via oxidation of catechol groups by the use of an oxidant (such as PI) as shown in our previous work [11, 12], or CuAAC, or both. For the CuAAC, CuSO4-NaLAc catalyst system is the most commonly used catalyst for the aqueous system. It can form Cu(I) in situ upon mixing CuSO4 (a stable Cu(II) salt) and NaLAc (a reductant) [23, 25, 29]. First, we assessed the cytotoxicity of CuSO4, NaLAc and CuSO4-NaLAc systems (CuSO4: NaLAc = 1:10, 1:5, or 1:2) by using the MTT (methylthiazolyl-diphenyl-tetrazolium bromide) assay against human-derived mesenchymal stem cells (hMSCs) (Fig. S3). As shown in Fig. S3, both Cu2+ and NaLAc treated hMSCs showed cell viability higher than 85% when the concentrations were lower than 1000 μmol/L. For the CuSO4-NaLAc system, cytotoxicity was also within the acceptable range when the concentrations of copper ion were lower than 1000 μmol/L, especially at the molar ratios of CuSO4 to NaLAc of 1:5 and 1:2. For the equal-weight mixture of iC-3N3 and iC-2PL (clickable iC) crosslinked by CuAAC using different concentrations of CuSO4 and different CuSO4 to NaLAc ratios, the gel times are shown in Fig. 2A. Higher CuSO4 concentrations and lower CuSO4/NaLAc ratios led to faster crosslinking, even below a minute for certain formulations. Both CuSO4-NaLAc and PI were used to enable dual crosslinking of the clickable iC system. When both CuSO4-NaLAc and PI were used in the system, the gel time, determined by a vial tilting test, was much longer than that of the same system crosslinked by either CuAAC or PI [11] (Fig. 2C). This phenomenon suggests that there should be a mutual inhibition effect between PI (oxidant) and NaLAc (reductant). We thus hypothesized that a copper chelating agent that can protect reduced Cu(I) ions from the oxidant (PI) used in the system may resolve this problem. A zwitterionic organic chemical buffering agent, HEPES, was selected, mainly because it is widely used in cell culture and contains two adjacent tertiary amine structures that are also found in other copper-chelating agents [19, 30–35]. Our hypothesis was confirmed by the gel time test (Fig. 2C). After the addition of HEPES, the gel times of clickable iC systems crosslinked by CuSO4-NaLAc and PI combined were reduced (Fig. 2C). The effect of HEPES on the gel time of the same system crosslinked solely by CuAAC was also studied. Among the investigated formulations, 20 mM of HEPES was found to generate the quickest crosslinking for all the three CuSO4/NaLAc ratios used (Fig. 2B). Based on this, we determined the optimized concentration of HEPES for crosslinking to be 20 mM, a relatively safe dosage found to be cytocompatible with hMSCs (Fig. S3) and typically below dosages used in some DMEM formulations. The gel times of dual-crosslinked clickable iCs crosslinked by CuSO4-NaLAc-PI-HEPES at 37 °C were below 5 minutes for all the formulations tested (Fig. 2D). When the concentrations of HEPES and CuSO4-NaLAc were fixed, increasing the concentration of PI within the range of 0 to 4 wt% led to prolonged gelation times (Fig. 2D). Gel time slightly decreased when the concentration of PI was increased to 8 wt%. This decrease might be due to the relatively increased catechol group-based crosslinking over the click crosslinking (Fig. 2D). The representative gel times of 1) iCs or clickable iCs, 2) crosslinked by either PI or CuAAC, or both, and 3) with or without HEPES, are listed in Table 2. 3.3 Rheological characterization The viscoelastic profiles of representative clickable iCs crosslinked by PI, CuAAC or both, including clickable iCs crosslinked by 8 wt% PI, or CuSO4-NaLAc-HEPES (5-50-20 mM), or CuSO4-NaLAc -HEPES (5-50-20 mM) and 4 wt% PI, were evaluated by rheological studies (Fig. 3). The gel times, as characterized by the crossover point of the rheological storage (G′) and loss (G″) moduli, were in agreement with the gel times observed by vial tilting tests (Fig. 2D and Table 2). 3.4 Properties of crosslinked clickable iCs Next, we tabulated the mechanical properties of PI, CuAAC, or dual-crosslinked clickable iCs in both dry and fully hydrated states (Fig. 4A, S4, Table S1). The dry mechanical properties (Fig. 4A and S4) reveal that the click (CuAAC) crosslinked iC films (iC-P4-X, iC-P2/4-X, and GL-iC-P2/4-X) exhibited comparable tensile strength and lower moduli, but possessed higher elongation at break compared to those of the corresponding iC films crosslinked by PI (iC-P4-PI 8wt%, iC-P2/4-PI 8wt%, and GL-iC-P2/4-PI 8wt%). Contrarily, dual-crosslinked (with PI and CuAAC) iC films (iC-P4-X-PI 4wt%, iC-P2/4-X-PI 4wt%, and GL-iC-P2/4-X-PI 4wt%) possessed both much higher tensile strength and moduli than the corresponding films crosslinked solely by either PI or CuAAC. The material with a higher elongation at break is more elastic and it can better resemble the elastic nature of soft tissues and withstand large amplitude and frequency of relative movement between adhered tissue surfaces, which all render the material suitable for bioadhesion applications. It is worthwhile to note that the PI concentration used for dual-crosslinked films was 4 wt%, which is lower than the concentration used for crosslinking films with only PI (8 wt%). It clearly demonstrates the improvement of mechanical strength achieved by introducing click reactions to iCs. The addition of GL further improved the mechanical strength (particularly moduli) of dual-crosslinked clickable iCs (Fig. 4A and S4). The stress-strain curves of PI, CuAAC or dual-crosslinked clickable iCs are shown in Fig. 4A. All of the crosslinked iCs exhibited elastic behavior with their elongation at break higher than 150%. CuAAC crosslinked clickable iCs possessed the highest elongation at break as some of them had elongation at break higher than 500% (Fig. 4A and S4). The mechanical strength of iCs decreased after they were hydrated and swollen and it is due to their hydrophilicity derived from hydrophilic PEG (Table S1). The degradation studies of crosslinked clickable iCs demonstrate that the rate of degradation was indeed inversely proportional to the degree of crosslinking, and that the inclusion of gelatin did not prolong the degradation time (Fig. 4B). Dual-crosslinked clickable iCs possessed prolonged degradation times as compared to corresponding clickable iCs crosslinked solely by either PI or CuAAC. Complete degradation could be realized in 35 days in all formulas tested. It is a remarkable feature as the degradation profiles of dual-crosslinked clickable iCs are well suited for many surgical adhesive applications. Sol contents and swelling ratios of different clickable iCs are shown in Fig. 4C and 4D. Although CuAAC enabled fast crosslinking of clickable iCs in less than 2 minutes (Fig. 2 and 3), both the sol contents and swelling ratios of hydrogels crosslinked only by CuAAC were much higher than those of the corresponding hydrogels crosslinked either by PI or dual crosslinking mechanism. It indicates that click crosslinked clickable iCs (iC-X) possessed low crosslinking densities, whereas dual-crosslinked clickable iCs (iC-X-PI) possessed much lower sol contents (all < 5%) than iCs crosslinked by PI (~17.5%) (Fig. 4C). The swelling ratios of iC-X-PI were also lower than those of normal iCs crosslinked by PI (Fig. 4D). More pendent carboxyl groups on iC’s side groups consumed by hydrophobic clickable monohydroxyl compounds are also considered to contribute to the low sol contents and low swelling ratios of dual-crosslinked clickable iCs. 3.5 Adhesion strength The various formulations of clickable iC showed wet tissue lap shear strengths ranging from 41.21 ± 4.48 kPa (for iC-P4-X) to 223.11 ± 15.94 kPa (for GL-iC-P2/4-X-PI 4 wt%) (Fig. 5). The introduction of click chemistry and the application of dual crosslinking mechanism vastly improved the wet adhesion strength of iCs (from ~50 kPa for iC-P4-PI 8 wt% to ~223 kPa for GL-iC-P2/4-X-PI 4 wt%). The inclusion of gelatin also helped increase the adhesion strength of iCs as shown in Fig. 5. Most importantly, the highest adhesion strength achieved (223.11 ± 15.94 kPa) during the test was approximately 13 times higher than that of commercially available fibrin at 15.4 ± 2.8 kPa, which is widely considered the gold standard [11]. 3.6 In vitro evaluations of cytotoxicity and proliferation The cytocompatibility of 1) clickable iC pre-polymers and GL, 2) sol contents and 3) degradation products of different crosslinked clickable iCs were evaluated using hMSCs. In-vitro cytotoxicity of clickable iC pre-polymers tested was comparable to those of the controls, commercially available PEGDA (700 Da) and gelatin (from cold fish skin) (Fig. 6A). The sol contents and degradation products of different crosslinked clickable iCs formulations tested did not induce any significant cytotoxicity against hMSCs (Fig. 6B and 6C). The above results, supported by qualitative cell morphology (Fig. 6D) suggested that clickable iCs and their different crosslinked bioadhesive formulations are indeed cytocompatible. 3.7 Anti-bacterial performance of crosslinked clickable iCs The antibacterial performance of clickable iCs crosslinked by either PI, CuAAC, or both was assessed against representative Gram-positive and Gram-negative bacteria, S. au than in E. coli, using commercially available Hydrofera Blue bacteriostatic foam dressing and PEGDA/HEMA hydrogels as controls. The bacterial inhibition ratios tested by exposing bacteria-containing broths directly to different crosslinked clickable iCs and other samples are shown in Fig. 7A and 7B. Inhibition was greater in S. au than in E. coli for all tested normal and clickable iC formulations. Except the iC-X hydrogels, all other formulations tested, including iC-P4-PI 4wt%, iC-P2/4-BD-PI 4wt%, and iC-P2/4-X-PI 4wt%, exhibited comparable or even better bacterial inhibition effect than Hydrofera Blue against either S. aureus or E. coli, during the bacterial incubation periods of 24 hours. The bacterial inhibition ratios of iCs and clickable iCs crosslinked by PI against S. au or E. coli were still higher than 90% or 60%, respectively, after 24 hours. Even though the bacterial inhibition ratios of iC-X-PIs dropped from ~100% after 13 hours of incubation against S. aureus or after 8 hours of incubation against E. coli, they remained at ~70% against S. au or ~50% against E. coli even after 24 hours of incubation. On the other hand, nearly no bacterial inhibition effect of PEGDA/HEMA hydrogels was observed. 3.8 Anti-fungal performance of crosslinked clickable iCs The antifungal performance was evaluated on the iC-X series and normal iCs by exposing them to a broth containing Candida albicans (C. albicans), fungi commonly seen in diabetic foot infections. Hydrofera Blue and PEGDA/HEMA hydrogels served controls, respectively. Based on the results shown in Fig. 7C, it can be seen that crosslinked iCs, including the CuAAC crosslinked one, all exhibited better fungal inhibition effect than Hydrofera Blue (~ 25% fungal survival). Especially, iCs crosslinked only by PI, had nearly no fungi grown after 3-hour exposure. Although a bit weaker, dual-crosslinked clickable iCs still exhibited considerable fungal inhibition: only ~ 10% fungal survival was observed. PEGDA/HEMA hydrogels also did not show any obvious fungal inhibition effect. 3.9 Long-term evaluations of antibacterial and antifungal properties in crosslinked clickable iCs Lastly, we evaluated long-term antibacterial and antifungal properties of our hydrogels by incubating bacteria or fungi in 1x, 10x, and 100x dilutions of degradation products or 4, 8, and 12-day release solutions of crosslinked clickable iCs. Hydrofera Blue and PEGDA/HEMA hydrogels were set as controls. Survival ratios, measured after 24 h for bacteria and after 6 h for fungi, showed that iC and clickable iCs, crosslinked by either PI or CuAAC, or both, all exhibited great long-term inhibition, particularly against the Gram-positive S. au and fungi. Compared to our crosslinked iCs and clickable iC hydrogels, the Hydrofera Blue foam demonstrated weaker long-term antibacterial and antifungal properties because the small molecular antimicrobial agents (i.e. crystal violet and methylene blue) that are simply mixed in the foam tend to be released rapidly in the early days of incubation in the broths [46]. Due to the antimicrobial property of citric acid that was used to synthesize iCs and clickable iC pre-polymers, crosslinked iCs and clickable iC hydrogels still possessed certain antimicrobial properties even on day 8 and 14 after small molecular PI were mostly released from the hydrogels [42]. 4. Conclusions In conclusion, copper-catalyzed azide-alkyne cycloaddition (CuAAC, click chemistry) was successfully introduced into injectable citrate-based mussel-inspired bioadhesives (iCs). The biggest challenge in this system, the mutual inhibition effect between sodium (meta) periodate (PI, as oxidant used for catechol group crosslinking) and the Cu(II) reductant sodium L-ascorbate, (used to reduce Cu(II) into Cu(I) to form CuAAC), was successfully minimized by utilizing a biocompatible buffering agent often used in cell culture, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), as a copper chelating agent. The introduction of click chemistry into iCs as a secondary crosslinking mechanism greatly improved the cohesive (tensile) strength of iCs. Furthermore, the application of click chemistry freed additional catechol groups that may contribute to adhesion in a greater extent instead of mainly participating in polymer crosslinking, leading to the design of bioadhesives with vastly improved wet adhesion strength to bio-surfaces. Dual-crosslinked (PI and CuAAC) clickable iCs presented intriguing antibacterial and antifungal abilities, which is beneficial for the application of bioadhesives in hemostasis and wound/tissue closure applications. Partial click functionalities were preserved even after crosslinking, imparting our polymers with further potential to conjugate bioactive molecules and growth factors, as the conjugation of collagen mimetic peptide p15 in our previous paper [14], further improving the biocompatibility and bioactivity of iCs. Supplementary Material supplement This work was supported in part by a National Cancer Institute (NCI) Award CA182670, and a National Heart, Lung, and Blood Institute (NHLBI) Award HL118498. Fig. 1 FTIR spectra of iC-3N3 (A) and iC-2PL (B) pre-polymers; 1H-NMR spectra of alkyne functionalized gelatin (Gelatin-Al, GL, C); and UV-vis spectra (D) of iC-3N3 and iC-2PL pre-polymers. Fig. 2 Gel times of iC-P4-3N3/2PL equal-weight mixture crosslinked by different formulations of CuSO4-sodium L-ascorbate (NaLAc) (A); the effect of HEPES to the gel times of iC-P4-3N3/2PL equal-weight mixture crosslinked by CuSO4-NaLAc (B) and dual-crosslinked by CuSO4-NaLAc and PI (C); gel times of dual-crosslinked iC-X-PI at 37°C (D, X = click). Fig. 3 Rheological analysis on the gelation of iC-P4-3N3/2PL (1:1 weight ratio) solution (50 wt%) crosslinked by oxidant (8 wt% PI, black color), CuAAC (CuSO4-NaLAc-HEPES, 5-50-20 mM, blue color), or dual-crosslinking (by CuSO4-NaLAc -HEPES (5-50-20 mM) and 4 wt% PI together, red color) at 25°C. Fig. 4 Evaluation of mechanical properties, showing stress-strain curves (A), degradation profiles (B), sol contents (leachable fractions) (C), and swelling ratios (D) of clickable iC mixture crosslinked by either sodium periodate (PI) or copper-catalyst, or both. Fig. 5 Adhesion strength of fibrin glue, and normal iC and clickable iCs crosslinked by sodium periodate (PI), CuSO4-NaLAc-HEPES or both, onto wet porcine small intestine submucosa, determined by lap shear strength test. Fig. 6 Cytotoxicity evaluation of clickable iC family: MTT assay for hMSCs cultured with: iC-click pre-polymers and Gelatin-Al (GL) (A), leachable part (sol content) (B) and degradation product (C) of clickable iCs crosslinked through different routes for 24 hours. Live/Dead assay for hMSCs seeded on dual-crosslinked iC-X-PI casted glass slides 1, 4, and 7 days post cell seeding (D). Fig. 7 Antibacterial and antifungal performance of clickable iCs (iC-X) hydrogels: Inhibition ratios kinetics curves of crosslinked iC-X series, iC-P4-PI 4wt%, PEGDA/HEMA (w/w=1/1, as negative control), and Hydrofera Blue (as positive control) against S. aureus (A) and E. coli. (B); Fungal survival ratios after direct exposure to crosslinked iC-click series, iC-P4-PI 4wt%, PEGDA/HEMA, and Hydrofera Blue for 3 hrs (C) (**p<0.01). Long-term antimicrobial performance of iC-X series: Antifungal and antibacterial performance of degradation products at different dilutions (1×, 10×, and 100×) (D) and periodical release solutions (E) of crosslinked hydrogels: iC-X series, iC-P4-PI 4wt%, PEGDA/HEMA (w/w=1/1, as negative control), and Hydrofera Blue (as positive control). Scheme 1 Synthesis schemes of triazide (3N3) (A) monomer, dipropiolate (2PL) (B) monomer, and clickable injectable citrate-based mussel-inspired bioadhesive pre-polymers (iC-3N3 and iC-2PL) (C). Scheme 2 Dual crosslinking (oxidant and click (CUAAC)) of clickable iC (mixture of iC-3N3 and iC-2PL). Table 1 Pre-polymer identification, feeding ratios, and dopamine contents Pre-polymer name MW of PEG (Da) Feeding ratio of CA: PEG: dopamine: 3N3 or 2PL monomer Dopamine content in pre-polymer (mmol/g) iC-P4 400 1.1 : 1 : 0.3 : 0 0.570 iC-P4-3N3 400 1.2 : 1 : 0.3 : 0.2 0.554 iC-P4-2PL 400 1.2 : 1 : 0.3 : 0.2 0.452 iC-P2/4-3N3 (PEG 200 and 400, mol/mol=1/1) 200 and 400 1.2 : 1 : 0.3 : 0.2 0.874 iC-P2/4 -2PL 200 and 400 1.2 : 1 : 0.3 : 0.2 0.547 Table 2 Gel times of different iCs and clickable iCs crosslinked by various concentrations of sodium periodate (PI) or copper catalyst, or both at room temperature (pre-polymer concentration used was 50 wt%). Pre-polymer name a PI b concentration (wt%) PI to pre-polymer ratio (wt%) CuSO4 (mM) NaLAc c (mM) HEPES d (mM) Measured gel time (s) iC-P4 8 8 0 0 0 163±9 iC-P4-3N3/2PL 8 8 0 0 0 155±8 iC-P4-3N3/2PL 0 0 5 50 0 142±6 iC-P4-3N3/2PL 0 0 5 50 20 61±5 iC-P4-3N3/2PL 4 4 5 50 0 663±6 iC-P4-3N3/2PL 4 4 5 50 20 479±7 iC-P4-3N3/2PL 2 2 5 50 0 529±4 iC-P4-3N3/2PL 2 2 5 50 20 359±6 iC-P2/4-3N3/2PL 4 4 5 50 20 368±5 a For clickable pre-polymer mixtures, for example, iC-P4-3N3/2PL, the weight ratio between iC-P4-3N3 and iC-P4-2PL was 1/1; b PI: sodium periodate; c NaLAc: sodium L-ascorbate; d HEPES: 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. 1 Kazemzadeh-Narbat M Annabi N Khademhosseini A Surgical sealants and high strength adhesives Mater Today 2015 18 176 7 2 Khanlari S Dubé MA Bioadhesives: A review Macromol React Eng 2013 7 573 87 3 Giano MC Ibrahim Z Medina SH Sarhane KA Christensen JM Yamada Y Injectable bioadhesive hydrogels with innate antibacterial properties Nat Commun 2014 5 4095 24958189 4 Duartea AP Coelhoa JF Bordadob JC Cidadec MT Gil MH Surgical adhesives: Systematic review of the main types and development forecast Prog Polym Sci 2012 37 1031 50 5 Mehdizadeh M Yang J Design strategies and applications of tissue bioadhesives Macromol Biosci 2013 13 271 88 23225776 6 Xie D Guo J Mehdizadeh MR Tran RT Chen R Sun D Development of injectable citrate-based bioadhesive bone implants J Mater Chem B 2015 3 387 98 7 Burks SS Spotnitz W Safety and usability of hemostats, sealants, and adhesives AORN J 2014 100 160 76 25080417 8 Zhang H Bré LP Zhao T Zheng Y Newland B Wang W Mussel-inspired hyperbranched poly(amino ester) polymers as strong wet tissue adhesive Biomaterials 2014 35 711 9 24140046 9 Barrett DG Bushnell GG Messersmith PB Mechanically robust, negative-swelling, mussel-inspired tissue adhesives Adv Healthcare Mater 2013 2 745 55 10 Zhang H Zhao T Newland B Duffy P Annaidh AN Ocearnhaill ED Wang W On-demand and negative-thermo-swelling tissue adhesive based on highly branched ambivalent PEG-catechol copolymers J Mater Chem B 2015 3 6420 8 11 Mehdizadeh M Weng H Gyawali D Tang L Yang J Injectable citrate-based mussel-inspired tissue bioadhesives with high wet strength for sutureless wound closure Biomaterials 2012 33 7972 83 22902057 12 Guo J Wang W Hu J Xie D Gerhard E Nisic M Synthesis and characterization of anti-bacterial and anti-fungal citrate-based mussel-inspired bioadhesives Biomaterials 2016 85 204 17 26874283 13 Zhang H Bré LP Zhao T Newland B Costa MD Wang W A biomimetic hyperbranched poly(amino ester)-based nanocomposite as a tunable bone adhesive for sternal closure J Mater Chem B 2014 2 4067 71 14 Guo J Xie Z Tran RT Xie D Jin D Bai X Yang J Click chemistry plays a dual role in biodegradable polymer design Adv Mater 2014 26 1906 11 24375469 15 Tang J Guo J Li Z Yang C Chen J Li S A fast degradable citrate-based bone scaffold promotes spinal fusion J Mater Chem B 2015 3 5569 76 16 Sun D Chen Y Tran RT Xu S Xie D Jia C Citric acid-based hydroxyapatite composite scaffolds enhance calvarial regeneration Sci Rep 2014 4 6912 25372769 17 Guo Y Tran RT Xie D Wang Y Nguyen DY Gerhard E Citrate-based biphasic scaffolds for the repair of large segmental bone defects J Biomed Mater Res A 2015 103 772 81 24829094 18 Hu J Guo J Xie Z Shan D Gerhard E Qian G Yang J Fluorescence imaging enabled poly(lactide-co-glycolide) Acta Biomater 2016 29 307 19 26463014 19 Uttamapinant C Tangpeerachaikul A Grecian S Clarke S Singh U Slade P Fast, cell-compatible click chemistry with copper-chelating azides for biomolecular labeling Angew Chem Inter Ed 2012 51 5852 6 20 Burdick JA Bioengineering: Cellular control in two clicks Nature 2009 460 469 70 19626106 21 DeForest CA Anseth KS Cytocompatible click-based hydrogels with dynamically-tunable properties through orthogonal photoconjugation and photocleavage reactions Nat Chem 2011 3 925 31 22109271 22 DeForest CA Polizzotti BD Anseth KS Sequential click reactions for synthesizing and patterning three-dimensional cell microenvironments Nat Mater 2009 8 659 64 19543279 23 Guo J Wei Y Zhou D Cai P Jing X Chen XS Huang Y Chemosynthesis of poly (ε-lysine)-analogous polymers by microwave-assisted click polymerization Biomacromolecules 2011 12 737 46 21302898 24 Yigit S Sanyal R Sanyal A Fabrication and functionalization of hydrogels through “click” Cchemistry Chem Asian J 2011 6 2648 59 21954074 25 Jiang Y Chen J Deng C Suuronen EJ Zhong Z Click hydrogels, microgels and nanogels: emerging platforms for drug delivery and tissue engineering Biomaterials 2014 35 4969 85 24674460 26 Adzima BJ Tao Y Kloxin CJ DeForest CA Anseth KS Bowman CN Spatial and temporal control of the alkyne–azide cycloaddition by photoinitiated Cu(II) reduction Nat Chem 2011 3 256 9 21336334 27 Öberg K Hed Y Rahmn IJ Kelly J Lowenhielm P Malkoch M Dual-purpose PEG scaffolds for the preparation of soft and biofunctional hydrogels: the convergence between CuAAC and thiol–ene reactions Chem Commun 2013 49 6938 40 28 Hu X Li D Zhou F Gao C Biological hydrogel synthesized from hyaluronic acid, gelatin and chondroitin sulfate by click chemistry Acta Biomater 2011 7 1618 26 21145437 29 Guo J Zhou D Hu J Chen X Jing X Huang Y Emulsion click microspheres: morphology/shape control by surface cross-linking and a porogen RSC Adv 2014 4 23685 9 30 Kennedy DC McKay CS Legault MCB Danielson DC Blake JA Pegoraro AFA Cellular consequences of copper complexes used to catalyze bioorthogonal click reactions J Am Chem Soc 2011 133 17993 8001 21970470 31 Hong V Presolski SI Ma C Finn MG Analysis and optimization of copper-catalyzed azide-alkyne cycloaddition for bioconjugation Angew Chem Int Ed 2009 48 9879 83 32 Guo J Meng F Jing X Huang Y Combination of anti-biofouling and ion-interaction by click chemistry for endotoxin selective removal from protein solution Adv Healthcare Mater 2013 2 784 9 33 Besanceney-Webler C Jiang H Zheng T Feng L del Amo DS Wang W Increasing the efficacy of bioorthogonal click reactions for bioconjugation: A comparative study Angew Chem Int Ed 2011 50 8051 6 34 Hong V Steinmetz NF Manchester M Finn MG Labeling live cells by copper-catalyzed alkyne-azide click chemistry Bioconjugate Chem 2010 21 1912 6 35 Mash HE Chin YP Sigg L Hari R Xue H Complexation of copper by zwitterionic aminosulfonic (good) buffers Anal Chem 2003 75 671 7 12585500 36 Zhang X Mei H Hu C Zhong Z Zhuo R Amphiphilic triblock copolycarbonates with poly(glycerol carbonate) as hydrophilic blocks Macromolecules 2009 42 1010 6 37 Xu J Prifti F Song J A versatile monomer for preparing well-defined functional polycarbonates and poly(ester-carbonates) Macromolecules 2011 44 2660 7 21686053 38 Hu J Peng K Guo J Shan D Kim GB Li Q Click crosslinking-improved waterborne polymers for enviornment-friendly coatings and adhesives ACS Appl Mater Interfaces 2016 27 17499 510 39 Tsotsalas M Liu J Tettmann B Grosjean S Shahnas A Wang Z Fabrication of highly uniform gel coatings by the conversion of surface-anchored metal-organic frameworks J Am Chem Soc 2014 136 8 11 24328287 40 Gyawali D Nair P Zhang Y Tran RT Zhang C Samchukov M Citric acid-derived in situ crosslinkable biodegradable polymers for cell delivery Biomaterials 2010 31 9092 105 20800893 41 Zhang X Soontornworajit B Zhang Z Chen N Wang Y Enhanced loading and controlled release of antibiotics using nucleic acids as an antibiotic-binding effector in hydrogels Biomacromolecules 2012 13 2202 10 22658064 42 Su LC Xie Z Zhang Y Nguyen KT Yang J Study on the antimicrobial properties of citrate-based biodegradable polymers Front Bioeng Biotechnol 2014 2 23 25023605 43 Hudson SP Langer R Fink GR Kohane DS Injectable in situ cross-linking hydrogels for local antifungal therapy Biomaterials 2010 31 1444 52 19942285 44 Deming TJ Synthetic polypeptides for biomedical applications Prog Polym Sci 2007 32 858 75 45 Faure E Falentin-Daudré C Jérôme C Lyskawa J Fournier D Woisel P Detrembleura C Catechols as versatile platforms in polymer chemistry Prog Polym Sci 2013 38 236 70 46 Data from online: http://www.hollisterwoundcare.com/files/pdfs/ifus/HydroferaBlueIFU.pdf
PMC005xxxxxx/PMC5121103.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 8711562 6325 Oncogene Oncogene Oncogene 0950-9232 1476-5594 27212032 5121103 10.1038/onc.2016.186 NIHMS778107 Article Functional Interaction of Histone Deacetylase 5 (HDAC5) and Lysine-specific Demethylase 1 (LSD1) Promotes Breast Cancer Progression Cao Chunyu 123 Vasilatos Shauna N. 13 Bhargava Rohit 134 Fine Jeffrey L. 14 Oesterreich Steffi 123 Davidson Nancy E. 123 Huang Yi 123 1 University of Pittsburgh Cancer Institute, University of Pittsburgh, Pittsburgh, PA 2 Department of Pharmacology & Chemical Biology, University of Pittsburgh, Pittsburgh, PA 3 Women’s Cancer Research Center, University of Pittsburgh, Pittsburgh, PA 4 Department of Pathology, University of Pittsburgh, Pittsburgh, PA To whom correspondence should be addressed: Yi Huang, M.D., Ph.D., Magee Womens Research Institute, Room 406, 204 Craft Ave, Pittsburgh, PA 15213. Phone:412-641-3589; Fax:412-641-2458; [email protected] 17 4 2016 23 5 2016 05 1 2017 24 11 2016 36 1 133145 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. We have previously demonstrated that crosstalk between lysine-specific demethylase 1 (LSD1) and histone deacetylases (HDACs) facilitates breast cancer proliferation. However, the underlying mechanisms are largely unknown. Here we report that expression of HDAC5 and LSD1 proteins were positively correlated in human breast cancer cell lines and tissue specimens of primary breast tumors. Protein expression of HDAC5 and LSD1 was significantly increased in primary breast cancer specimens in comparison with matched normal adjacent tissues. Using HDAC5 deletion mutants and co-immunoprecipitation studies, we showed that HDAC5 physically interacted with LSD1 complex through its domain containing nuclear localization sequence and phosphorylation sites. While the in vitro acetylation assays revealed that HDAC5 decreased LSD1 protein acetylation, siRNA-mediated HDAC5 knockdown did not alter the acetylation level of LSD1 in MDA-MB-231 cells. Overexpression of HDAC5 stabilized LSD1 protein and decreased the nuclear level of H3K4me1/me2 in MDA-MB-231 cells, whereas loss of HDAC5 by siRNA diminished LSD1 protein stability and demethylation activity. We further demonstrated that HDAC5 promoted the protein stability of USP28, a bona fide deubiquitinase of LSD1. Overexpression of USP28 largely reversed HDAC5-KD induced LSD1 protein degradation, suggesting a role of HDAC5 as a positive regulator of LSD1 through upregulation of USP28 protein. Depletion of HDAC5 by shRNA hindered cellular proliferation, induced G1 cell cycle arrest, and attenuated migration and colony formation of breast cancer cells. A rescue study showed that increased growth of MDA-MB-231 cells by HDAC5 overexpression was reversed by concurrent LSD1 depletion, indicating that tumor-promoting activity of HDAC5 is an LSD1 dependent function. Moreover, overexpression of HDAC5 accelerated cellular proliferation and promoted acridine mutagen ICR191 induced transformation of MCF10A cells. Taken together, these results suggest that HDAC5 is critical in regulating LSD1 protein stability through posttranslational modification, and the HDAC5-LSD1 axis plays an important role in promoting breast cancer development and progression. breast cancer HDAC5 LSD1 USP28 Jade-2 epigenetic crosstalk tumorigenesis Introduction LSD1 is the first identified FAD-dependent histone demethylase that has been typically found in association with a transcriptional repressor complex that includes CoREST, HDAC1/2, BHC80, and others (1–4). A role for elevated expression of LSD1 has been implicated in tumorigenesis in various cancers including breast cancer (3, 5–9). Studies from our and other laboratories consistently showed that inhibition of LSD1 hindered proliferation of breast cancer cells (6, 8, 10). Lim et al. reported that LSD1 is highly expressed in ER-negative breast cancers (6). A recent study found that LSD1 is significantly overexpressed in high grade DCIS or IDC versus low/intermediate DCIS (11). These studies point to a tumor promoting role for LSD1 in breast cancer. We were among the first to report the use of small molecule compounds and preclinical treatment strategies that have promise to work through this target in cancer (8, 9, 12). The development of novel LSD1 inhibitors is progressing rapidly. For example, a new generation of (bis)urea/(bis)thiourea LSD1 inhibitors displayed improved potency against LSD1 in cancer cells (13). A newly reported GSK-LSD1 inhibitor exhibited interesting cell type specific inhibition against small cell lung cancer cells in preclinical models. (14). However, how LSD1 is upregulated in breast cancer and the precise role of LSD1 in breast cancer development are still unclear. Our most recent work showed that siRNA-mediated inhibition of HDAC5 led to a significant increase of H3K4me2, a known substrate of LSD1, suggesting a potential role of HDAC5 in regulating LSD1 activity (10). However, little is known about the precise role of HDAC5 and mechanisms underlying its regulation on LSD1 activity in breast cancer. HDAC5 is an important member of class IIa HDAC isozymes with important functions in transcriptional regulation, cell proliferation, cell cycle progression, and cellular developmental activities (15, 16). HDAC5 has been shown to play important roles in many diseases including cancer (17, 18). In this study, we addressed the following clinically relevant issues that have been understudied: (1) Is elevation of LSD1 expression associated with HDAC5 overexpression during breast cancer development? (2) How is LSD1 regulated by HDAC5 in breast cancer? (3) What is the role of the HDAC5-LSD1 axis in breast cancer initiation, proliferation and metastasis? To answer these questions, we delineated the mechanisms underlying the functional link between LSD1 and HDAC5 in chromatin remodeling and demonstrated that these two important chromatin modifiers closely cooperate to mediate proliferation, cell cycle and metastasis of breast cancer cells. Results 1. HDAC5 and LSD1 proteins are coordinately expressed in human breast cancer To study the potential association of HDAC5 and LSD1 in breast cancer, we first examined mRNA levels of HDAC5 and LSD1 in human immortalized normal mammary epithelial MCF10A cells, fully malignant MCF10A–CA1a cells transformed from MCF10A cells with transfection of HRAS (19), and several human breast cancer cell lines. qPCR studies showed that there was no clear association of mRNA expression between HDAC5 and LSD1 in breast cancer cell lines (Figure 1a). The Oncomine-TCGA database showed moderate change of the mRNA level of LSD1 and HDAC5 in IBC (Supplementary Figure 1a and 1b). mRNA levels of both HDAC5 and LSD1 are altered in approximately 6% of breast cancer patients (www.cbioportal.org) without an apparent association with specific subtypes (Supplementary Figure 1c and 1d). However, protein expression of both HDAC5 and LSD1 was significantly elevated in malignant breast cell lines compared with MCF10A (Figure 1b), and protein levels of HDAC5 and LSD1 were positively correlated (Figure 1c). The correlation of HDAC5 and LSD1 protein expression was further validated in 50 primary breast cancers using immunohistochemical staining with validated antibodies (Supplementary Figure 2a and 2b). Chi-square analysis showed a statistically significant correlation between HDAC5 and LSD1 protein expression in these tumors (Figure 1d). Furthermore, IHC analysis showed that breast cancer tissues (n=18) expressed significantly higher level of HDAC5 and LSD1 than matched normal adjacent tissues (n=18) (Figure 1e). The mean H-score for HDAC5 staining in stage 3 breast tumors (n=25) was statistically significantly higher than stage 2 counterparts (n=25). The mean H-score of LSD1 staining for stage 3 tumors was also higher than that of stage 2 tumors with a p-value of 0.07 (Figure 1f). These results were further validated with independent manual H score evaluations by two breast cancer pathologists with moderate interobserver concordance (Supplementary Figure 3a and 3b). Taken together, these findings suggest that HDAC5 and LSD1 proteins are coordinately overexpressed in breast cancer cell lines and tissue specimens. 2. Physical interaction of LSD1 and HDAC5 in breast cancer cells To address whether LSD1 and HDAC5 physically interact, a co-immunoprecipitation study was carried out in MDA-MB-231 and MCF10A–CA1a cells transiently transfected with pcDNA3.1 or pcDNA3.1-FLAG-HDAC5 plasmids. After immunoprecipitation (IP) with LSD1 antibody, we found that both endogenous and exogenous HDAC5 proteins were co-immunoprecipitated with LSD1 protein (Figure 2a). The interaction between native LSD1 and HDAC5 was further validated in additional breast cancer cell lines (Figure 2b). A similar result was obtained in the reciprocal immunoprecipitation using anti-FLAG antibody to confirm that LSD1 was co-immunoprecipitated with FLAG-HDAC5 (Figure 2c). To precisely map the HDAC5 domain(s) responsible for interaction with LSD1, we expressed a series of HDAC5 deletion mutants engineered in pcDNA3.1-FLAG plasmids in MDA-MB-231 cells (Figure 2d). Immunoprecipitation assays of cells transfected with full length HDAC5 cDNA confirmed the HDAC5-LSD1 interaction. Deletion of an N-terminal myocyte enhancer factor-2 (MEF2) binding domain (HDAC5-Δ1) alone had no impact on HDAC5-LSD1 interaction. However, removal of both the MEF2 domain and nuclear localization sequence (NLS) (HDAC5-Δ2) completely abolished HDAC5-LSD1 interaction. Further deletion of an N-terminal HDAC and nuclear export sequence (NES) (HDAC5-Δ3) and MEF2 domain (HDAC5-Δ4) did not adversely alter LSD1 binding with HDAC5 fragments (Figure 2e). Immunofluorescence studies showed nuclear localization of full length HDAC5, HDAC5-Δ1, HDAC5-Δ3 and HDAC5-Δ4. Depletion of the NLS-containing domain (HDAC5-Δ2) completely blocked HDAC5 nuclear translocation (Figure 2f). In vitro pull-down assays by using His-tag recombinant LSD1 protein incubating with HDAC5 full length or deletion mutants validated that HDAC5 domain containing NLS element is essential for interaction with LSD1 (Supplementary Figure 4). 3. HDAC5 promotes LSD1 protein stability and activity Next, we examined whether the mRNA or protein levels of HDAC5 and LSD1 were affected by their interaction with each other. Overexpression of HDAC5 in MDA-MB-231 cells failed to alter LSD1 mRNA expression, but led to a significant increase of LSD1 protein expression (Figure 3a and 3b). HDAC5 knockdown by siRNA attenuated LSD1 protein expression without affecting its mRNA level (Figure 3c and 3d). The effect of LSD1 on HDAC5 expression was subsequently assessed using our previously established MDA-MB-231-LSD1-KD cells (10). Depletion of LSD1 exerted no effect on HDAC5 mRNA or protein levels (Figure 3e and 3f). Simultaneous overexpression of pcDNA3.1-HDAC5 with HDAC5 siRNA significantly reversed the decrease of LSD1 (Supplementary Figure 5a). These results suggest that HDAC5 functions as an upstream regulator that governs LSD1 protein stability via posttranslational regulation. Quantitative immunoblots showed that levels of H3K4me1/2 and AcH3K9, the substrates for LSD1 and HDAC5 respectively, were downregulated by HDAC5 overexpression, whereas loss of HDAC5 exerted the opposite effect (Figure 3g; Supplementary Figure 5b), suggesting a critical role of HDAC5 in governing chromatin modifying activity of LSD1. Cycloheximide (CHX) chase assay showed that overexpression of HDAC5 significantly extended LSD1 protein half-life, whereas depletion of HDAC5 by siRNA decreased LSD1 protein half-life in MDA-MB-231 cells (Figure 3h and 3i; Supplementary Figure 5c). To determine whether other recognized LSD1 cofactors or HDACs exert similar effects on LSD1 protein stability, MDA-MB-231 cells were treated with siRNA against several LSD1 complex co-factors (CoREST, HDAC1 and HDAC2) or other class II HDAC isozymes (HDAC 4, 6, 7, 9, 10) respectively. Transfection with siRNA probes effectively knocked down mRNA expression of target genes without affecting their protein levels (Figure 3j; Supplementary Figure 6a). To confirm the qPCR results, quantitative immunoblotting was performed and showed depletion of CoREST led to insignificant change of LSD1 protein stability (Supplementary Figure 6b and 6c). Together, these results strengthen the conclusion that HDAC5 functions as a positive regulator of LSD1 protein in breast cancer cells. 4. HDAC5 regulates LSD1 protein stability through modulation of the LSD1 associated ubiquitination system Protein ubiquitination assays indicated that HDAC5 overexpression significantly attenuated LSD1 polyubiquitination (Figure 4a), whereas depletion of HDAC5 by siRNA facilitated LSD1 polyubiquitination (Supplementary Figure 7a). Recently, Jade-2 and USP28 were identified as specific E3 ubiquitin ligase and deubiquitinase for LSD1 respectively (20, 21). Our study showing that increase of LSD1 protein expression by Jade-2 siRNA and decrease of LSD1 protein expression by USP28 siRNA in MDA-MB-231 cells confirmed the roles of Jade-2/USP28 as LSD1 ubiquitin ligase/deubiquitinase in breast cancer cells (Figure 4b; Supplementary Figure 7b). qPCR studies demonstrated that mRNA level of either Jade-2 or USP28 was not altered by HDAC5 knockdown or overexpression (Figure 4c). The regulation of HDAC5 on protein expression of Jade-2 or USP28 was subsequently assessed. Due to the lack of highly specific antibody against Jade-2, plasmids expressing Jade-2-FLAG fusion protein were transfected into cells as an alternative approach. MDA-MB-231 and MCF10A–CA1a cells expressing Jade-2-FLAG protein were simultaneously treated with HDAC5 siRNA to evaluate the effect of HDAC5 on Jade-2 protein expression. Immunoblot showed that depletion of HDAC5 did not change the protein level of Jade-2 (Figure 4d). However, overexpression of HDAC5 led to significant increase of USP28 protein expression in both cell lines (Figure 4e). In vitro pull-down assay using His-tag recombinant LSD1 protein incubated with USP28-FLAG protein indicated a direct interaction of HDAC5 and USP28 (Supplementary Figure 4), and HDAC5 overexpression significantly attenuated USP28 polyubiquitination (Supplementary Figure 7c). To understand whether HDAC5 may stabilize LSD1 protein through upregulation of USP28 protein stability, a rescue study was carried out in MDA-MB-231 and MCF10A–CA1a cells using concurrent transfection of HDAC5 siRNA and USP28 expression plasmids, and showed that overexpression of USP28 completely blocked the destabilization of LSD1 by HDAC5 depletion (Figure 4f, Supplementary Figure 7d). In an additional rescue experiment, overexpression of HDAC5 failed to promote LSD1 protein expression when cells were simultaneously treated with USP28 by siRNA (Supplementary Figure 7e). All these data support the notion that HDAC5 stabilizes LSD1 protein by enhancing protein expression of its deubiquitinase. To examine whether interaction of HDAC5 with LSD1/USP28 complex deacetylates LSD1 or USP28, in vitro protein acetylation assays was first carried out by incubating GST-tagged recombinant HDAC5 protein with cellular pull-down of LSD1-FLAG or USP28-FLAG by IP, and immunoprecipitates of IgG was incubated with recombinant HDAC5 protein as negative control of assays (Figure 5a). Bulk histone was used as control substrate (Supplementary Figure 8). Quantitative immunoblots using antibody against pan-acetylated lysine showed that HDAC5 reduced acetylation level of LSD1 without altering the acetylation status of USP28 (Figure 5a and 5b). Next, the in vivo effect of HDAC5 depletion on LSD1 acetylation was investigated in MDA-MB-231 cells transfected with scramble or HDAC5 siRNAs. After immunoprecipitation with LSD1 antibody or IgG (negative control), immunoblotting was performed and the results showed that expression levels of both total LSD1 protein and acetylated LSD1 protein were decreased by HDAC5 depletion (Figure 5c). Quantitative immunoblots indicated that the relative acetylation level of LSD1 was not statistically altered by HDAC5 siRNA in MDA-MB-231 cells (Figure 5d). AcetylH3K9 was used as control of substrate and its expression was increased by HDAC5 siRNA (Figure 5c). These results suggest that inhibition of HDAC5 alone is not sufficient enough to increase LSD1 acetylation in breast cancer cells. 5. Inhibition of HDAC5 reactivates expression of LSD1 target genes In cancer cells, amplified LSD1 expression is frequently associated with abnormal suppression of key tumor suppressor genes (TSGs) (3, 22). We next examined whether expression of LSD1 target TSGs could be reactivated following HDAC5 inhibition. Loss of expression of CDK inhibitor p21 and epithelial marker claudin-7 (CLDN7) has been reported to be associated with an aggressive phenotype of breast cancer (23, 24). The transcription activity of p21 and CLDN7 has been found to be suppressed by enhanced activity of LSD1 in breast cancer (6, 25). Transfection of HDAC5 siRNA resulted in significantly increased mRNA expression of p21 and CLDN7 in MDA-MB-231 cells (Figure 5e). Quantitative chromatin immunoprecipitation (qChIP) assays revealed that depletion of HDAC5 decreased occupancy of both HDAC5 and LSD1, and increased enrichment of H3K4me2 and acetylH3K9 at the promoters of both genes (Figure 5f). These data suggest that transcriptional de-repression of these genes lies largely in the cooperation between HDAC5 and LSD1 at key active histone marks. 6. Inhibition of HDAC5-LSD1 axis hinders breast cancer proliferation and metastasis To explore the functional role of the HDAC5-LSD1 axis in regulating breast cancer development, stable knockdown of HDAC5 mRNA (HDAC5-KD) was generated in MDA-MB-231 and MCF10A–CA1a cells by infection with shRNA lentiviral particles. Similar to the effect of transient inhibition of HDAC5 by siRNA, stable knockdown of HDAC5 expression significantly reduced LSD1 protein expression in two independent HDAC5-KD clones (Figure 6a). Loss of HDAC5 in both clones hindered cell proliferation and colony formation in soft agar (Figure 6b and 6c). Flow cytometry analysis showed that inhibition of HDAC5 resulted in a greater fraction of cells accumulated at G1 phase and reduction of the S-phase cell fraction (Figure 6d; Supplementary Figure 9). Moreover, loss of HDAC5 attenuated motility and invasion of MDA-MB-231 cells in a Boyden chamber assay (Figure 6e). A rescue experiment indicated that HDAC5 overexpression promoted growth of MDA-MB-231-Scramble cells, but failed to alter the growth of MDA-MB-231-LSD1-KD cells (Figure 6f). An additional rescue study revealed that LSD1 overexpression rescued growth inhibition by HDAC5 depletion in MDA-MB-231-HDAC5-KD cells (Figure 6g). Taken together, these results demonstrate that tumor promoting activity of HDAC5 is dependent on LSD1 activity in breast cancer cells. 7. Overexpression of HDAC5 promotes mutagen-induced tumorigenic development in MCF10A cells To address whether enhanced interaction between HDAC5 and LSD1 is a critical epigenetic alteration driving tumorigenic transformation of breast cancer, we generated two MCF-10A cell lines overexpressing HDAC5 (MCF10A–HDAC5). Stable overexpression of HDAC5 in MCF10A cells increased LSD1 protein level and promoted cell proliferation of both clones (Figure 7a and 7b), indicating a growth-promoting role for HDAC5 in MCF10A cells. Inhibition of LSD1 by shRNA significantly hindered MCF10A growth and reversed the growth promotion mediated by HDAC5 overexpression, suggesting that HDAC5 promotes MCF10A growth in an LSD1 dependent manner (Figure 7c; Supplementary Figure 10). To evaluate if MCF10A–HDAC5 cells have altered susceptibility to tumorigenesis, MCF10A–Vector and MCF-10A–HDAC5 cells were cultured for 7 months in medium containing 500ng/ml ICR191. ICR191 generates genomic instability and genetic variability, and has been successfully used to induce epithelial cell transformation in several models including MCF-10A (26, 27). MCF10A–HDAC5 cells were subsequently tested for the capacity of anchorage-independent growth in soft agar for 4 weeks. Soft agar colony formation study demonstrated that ICR191 treatment improved the ability of MCF10A cells to form growing colonies, and overexpression of HDAC5 significantly promoted ICR191-induced colony formation in MCF10A cells (Figure 7d). To determine the role of LSD1 in HDAC5 enhanced tumorigenic transformation induced by ICR191, scramble control and LSD1 shRNA lentivirus particles were infected into MCF10A–Vector or MCD10A–HDAC5 cells which had been treated with ICR191 for 7 months, and soft agar growth assays showed that loss of LSD1 in MCF10A–HDAC5 cells significantly abolished cellular ability in colony formation (Figure 7e). A model illustrating the role of HDAC5-LSD1 axis in breast cancer development is proposed based on the above findings (Figure 7f). DISCUSSION High levels of HDAC5 have been found to be associated with poor survival in multiple cancer types (28, 29). LSD1 overexpression has been reported to be a poor prognostic factor in basal-like breast cancer, a subtype with aggressive clinical characteristics (6, 30). In this study, IHC analysis showed that breast cancers expressed higher levels of HDAC5 compared to the matched normal adjacent breast tissue. Importantly, our study found a positive correlation between HDAC5 and LSD1 proteins in breast tumor cell lines and patient tissue specimens. Increased expression of HDAC5 and LSD1 is correlated with higher stage of breast cancer in our exploratory study. These findings suggest that the coordinated overexpression of HDAC5 and LSD1 may serve as potential novel prognostic markers as well as possible therapeutic targets for breast cancer. More robust studies will be necessary to understand the precise role of elevated protein expression levels of HDAC5 and LSD1 in the risk stratification of breast cancer patients. LSD1 protein stability is controlled by several posttranslational modifications such as ubiquitination and methylation (20, 21, 31). However, the precise mechanism of how LSD1 protein stability is regulated is still not understood. A previous study reported that stable depletion of CoREST facilitated LSD1 degradation in HeLa cells (32). However, siRNA-mediated knockdown of CoREST alone in breast cancer cells failed to destabilize LSD1 protein, suggesting additional layers of control of LSD1 protein stability are required in breast cancer. In this study, we observed for the first time that LSD1 protein stability is promoted by HDAC5. We further found that the HDAC5 domain containing NLS is essential for LSD1-HDAC5 interaction. The NLS element provides docking sites for 14–3-3 chaperone binding and has been shown to be critical for HDAC5 import into the nucleus and the regulation of its repressor activity (17, 33). Although an in vitro assay demonstrated that HDAC5 reduced LSD1 acetylation, HDAC5 siRNA treatment in breast cancer cells failed to alter acetylation of LSD1 protein. Our in vivo results suggest that LSD1 acetylation is likely regulated by a large complex that may involve additional protein deacetylases or cofactors. Further studies are needed to identify the regulatory complex and clarify the precise role of HDAC5 in regulation of LSD1 acetylation in breast cancer cells. Our studies revealed that HDAC5 regulates LSD1 via enhancement of the protein stability of deubiquitinase USP28. High expression of USP28 has been found to promote the progression of breast and colon cancers (20, 34). Importantly, USP28 has been reported to deubiquitinate important tumor growth regulators such as c-Myc and TP53BP1 that are involved in MYC proto-oncogene stability and DNA damage response checkpoint regulation respectively (35, 36). Our pilot microarray study revealed that inhibition of the HDAC5-LSD1 axis down-regulates c-Myc expression (data not shown). Sen et al. recently reported that HDAC5 is a key component in the temporal regulation of p53-mediated transactivation (37). All of these findings imply an interaction of HDAC5/LSD1 axis and USP28-associated ubiquitin proteasome system in regulating downstream targets involved in tumor development. USP28 has been well-characterized for its role in promoting tumorigenesis, and thus is a potential candidate target in cancer therapy. Given the current inability to use drugs to directly target USP28–driven cancer proliferation, our study suggests a novel alternative approach of targeting USP28 stability by development of HDAC5-specific inhibitors in cancer. Our findings provide supportive evidence showing that HDAC5 control of cell proliferation is largely dependent on LSD1 stabilization. Furthermore, in this study, we showed that non-transformed MCF10A cells overexpressing HDAC5 significantly promoted ICR191-induced transformation of MCF10A cells. The overexpressed HDAC5 is consistently associated with upregulated LSD1 protein expression over the entire course of transformation induction. These data indicate that enhanced crosstalk between HDAC5 and LSD1 may represent a critical mechanism contributing to breast tumorigenesis. HDAC inhibitors (HDACi) hold great promise for cancer therapy. Despite the promising clinical results produced by HDACi in treatment of hematological cancers such as T cell lymphoma, no apparent clinical evidence indicates that HDAC inhibitors work effectively as a monotherapy against solid tumors including breast tumors (38–41). From a clinical perspective, our novel findings have significance for design and development of novel combination strategies targeting HDAC5-LSD1 axis as an alternative approach for improvement of therapeutic efficacy of HDAC inhibitors in breast cancer. As summarized in Figure 7f, we show for the first time that LSD1 protein stability is promoted by HDAC5 through the LSD1 associated ubiquitin-proteasome system, confirming that the regulation of LSD1 by HDAC5 is a posttranslational event. Our novel findings also provide supportive evidence that an orchestrated interaction between HDAC5 and LSD1 is a critical epigenetic mechanism to suppress transcriptional activities of important tumor suppressor genes that may contribute to breast cancer development. Materials and methods Reagents and cell culture conditions MDA-MB-231, MDA-MB-468, MCF-7, T47D, HCC-202 and SK-BR-3 cell lines were obtained from the ATCC/NCI Breast Cancer SPORE program. MCF10A–parental and MCF10A–CA1a cells were gifts from Dr. Saraswati Sukumar (Johns Hopkins University). Cells were cultured in growth medium as described previously (10, 42). Tissue Microarrays (TMAs) and immunohistochemistry TMAs (US Biomax, Rockville, MD) were stained using LSD1 or HDAC5 antibodies. Standard staining procedure for paraffin sections was used for IHC according to manufacturer’s recommendations (Vector Labs, Inc., Burlingame, CA). Monoclonal antibodies were used for detection of LSD1 (1:800; Cell Signaling, Danvers, MA) and HDAC5 (1:100; Santa Cruz). The staining was visualized using diaminobenzidine, and quantitated using IHC Profiler, an ImageJ plugin (43). H-scores were calculated as previously described (44). The manual scoring of H-scores was also carried out by two breast cancer pathologists. Plasmid construction and stable transfection Plasmids pcDNA3.1(+)-FLAG, pcDNA3.1(+)-FLAG-HDAC5, and pDZ-FLAG-USP28 were purchased from Addgene (Cambridge, MA). pReceiver-FLAG-LSD1 was obtained from Gene Copoeia (Rockville, MD). A FLAG-tagged ORF cDNA clone for Jade-2 was purchased from GenScript (Piscataway, NJ). pcDNA3-HA-ubiquitin was obtained from Dr. Yong Wan (University of Pittsburgh). HDAC5 deletion mutants were engineered into pcDNA3.1(+)-FLAG-HDAC5 by PCR with primers shown in Table S1. HDAC5-Δ2 was constructed by digesting full length plasmids with SacII from amino acid 61 to 489. Stable transfection was carried out using Lipofectamine 3000™ transfection reagent (Life Technologies, Grand Island, NY), and colonies were selected with 800 µg/ml G418. siRNA and shRNA treatment and stable cell line generation Pre-designed siRNA and non-targeting scramble siRNA (Santa Cruz) were transfected into cells following the manufacturer’s protocol. Cells were harvested 48 h post-transfection for further analysis. Scramble control, LSD1-specific or HDAC5-specific shRNA lentiviral particles (Santa Cruz) were infected into cells according to manufacturer’s protocol. Cells were treated with 10µg/ml puromycin 72 h after infection. Single colonies were analyzed for expression of LSD1 or HDAC5 via immunoblots. RNA extraction and qPCR Total RNA extraction and cDNA synthesis used the methods described previously (10). Quantitative real-time PCR was performed on the StepOne real-time PCR system (Life Technologies). All of the TaqMan® Gene Expression Assays were predesigned and obtained from Life Technologies. Western blotting Western blotting was performed as previously described (12, 45, 46). Antibodies used in this study were shown in Table S2. Membranes were scanned with Li-Cor BioScience Odyssey Infrared Imaging System (Lincoln, NE). Crystal violet, and cell invasion assays Crystal violet proliferation assays were performed as described in our previous study (47). The invasive capability of breast carcinoma cells was tested with Millipore QCM™ 24-Well invasion assay Kit (Merck KGaA, Germany) according to manufacturer’s protocol. Soft agar colony formation assay 1.2 % Bacto-agar (BD Biosciences, Franklin Lakes, NJ) was autoclaved and mixed with growth medium to produce 0.6 % agar. The mixture was quickly plated and solidified for 45 min. Cells were suspended in 0.6 ml 2x growth medium and mixed gently with 0.6 ml 0.8 % agar /medium. 1 ml of cells with 0.4% agar/medium mixture was added onto plate for solidification. Colony formation was examined using stereo microscopy and analyzed (CellSens Dimension, Olympus). Flow cytometry analysis Cells were harvested and fixed with 70% ethanol. The cell pellet was then treated with 1% TritonX-100. Cells were subsequently resuspended in 50 µg/ml propidium iodide (Sigma) containing RNaseI (Roche, Indianapolis, IN) followed by analysis on the LSR II XW4400 workstation (BD Biosciences). Immunofluorescence 48 h after transfection, cells were fixed with 4% paraformaldehyde and incubated with primary antibodies (1:250) overnight at 4 °C. After washing, cells were incubated with fluorescence-labeled secondary antibody (1:100). After washing, coverslips were placed on a glass slide using UltraCruz™ mounting medium (Santa Cruz) before fluorescence microscope examination. Immunoprecipitation, ubiquitination and protein half-life assays The cell lysate was obtained by using immunoprecipitation lysis buffer as described previously (48). LSD1 or IgG antibodies were added to cell lysate. Protein G-plus agarose beads (Santa Cruz) or Flag-M2 affinity gel were collected and subjected to immunoblotting. HA-Ubiquitin, pcDNA3.1-Flag-HDAC5 or empty vector plasmids were co-transfected into cells for 38 h. Cells were then treated with 10 µM MG-132 for 10 h and harvested for immunoprecipitation assay with protein G-plus agarose beads. For half-life studies 48 h after transfection with pcDNA3.1-HDAC5 or HDAC5-siRNA, cells were treated with 100 µg/ml cycloheximide and then harvested at indicated times for immunoblotting. Protein acetylation assay The immunoprecipitates of FLAG-M2 agarose from MDA-MB-231 cells overexpressing FLAG-tag USP28 or FLAG-tag LSD1 were used as substrates for protein deacetylation assay. Pull-down of IgG was used as negative control. 0.25 µg of recombinant human GST-tagged HDAC5 protein (Creative BioMart, NY, NY) was mixed with 30 µl immunoprecipitates or 1.5 µg bulk histone at 37°C for 6h in a buffer containing 40 mM Tris-HCl (pH 8.0), 2.5 mM MgCl2, 50 mM NaCl, 2 mM KCl, 0.5mM DTT, 1mM EDTA and protease inhibitor. The reactions were then subjected to immunoblots with anti-acetyl lysine antibody (EMD Millipore, Billerica, MA). FLAG-tagged USP28 or LSD1 and bulk histone were probed with anti-FLAG antibody or H3 antibody as loading control. Inactive HDAC5-GST protein was used as negative control by heating recombinant protein at 95°C for 5 min. In vivo protein acetylation assay was performed using cell lysate of MDA-MB-231 cell transfected with scramble and HDAC5 siRNAs. LSD1 or IgG antibodies were added to cell lysate. Protein G-plus agarose beads (Santa Cruz) were collected and subjected to immunoblotting with anti-acetyl lysine or LSD1 antibodies. Chromatin immunoprecipitation Chromatin immunoprecipitation assay was performed as described previously (12). Primary antibodies against HDAC5, LSD1, H3K4me2 and acetyl-H3K9 were used as indicated for immunoprecipitation of protein–DNA complexes. PCR primer sets used for amplification of precipitated fragments were shown in Table S1. Input DNA was used for normalization. Statistical analysis Data were represented as the mean ± standard deviation of the mean (s.d.) of three independent experiments. The quantitative variables were analyzed by two tailed Student's t-test. Chi-square study was used to assess the correlation between HDAC5 and LSD1 protein expression by using median H-scores as the cutoff for high vs low protein expression. P-value<0.05 was considered statistically significant for all tests. Statistical analyses were performed using GraphPad Prism 6 (GraphPad Software, Inc., La Jolla, CA). Supplementary Material 1 7 8 9 13 10 11 12 2 3 4 5 6 This work is supported by US Army Breast Cancer Research Program (W81XWH-14-1-0237 to YH; W81XWH-14-1-0238 to NED), Breast Cancer Research Foundation (to NED and SO), and UPCI Genomics Core Facility supported by NCI P30CA047904. Figure 1 Correlated overexpression of HDAC5 and LSD1 protein in breast cancer. (a) The levels of mRNA expression of HDAC5 and LSD1 in breast cancer cell lines versus MCF10A cells (set as fold 1) using real-time qPCR with β-actin as an internal control. (b) Immunoblots with anti-HDAC5 and LSD1 antibodies in indicated cell lines. β-actin protein was blotted as a loading control. (c) Histograms represent the mean protein levels of HDAC5 or LSD1 in three determinations relative to β-actin ± s.d. as determined by quantitative immunoblots. (d) 50 primary human invasive breast tumor samples were immunostained with antibodies against HDAC5 or LSD1. Chi-square study was performed by using median H-scores as the cutoff for high vs low protein expression. (e) Representative HDAC5 and LSD1 staining (200x) in invasive breast carcinoma and adjacent normal tissue specimens from one representative patient. H-scores represent average staining intensity in breast tumors (n=18) versus adjacent normal breast tissue (n=18). (f) Representative HDAC5 and LSD1 staining (200x) in stage 2 and 3 invasive breast carcinoma specimens. H-scores represent average staining intensity in stage 3 breast tumors (n=25) versus stage 2 breast tumors (n=25). p<0.05, p<0.01, * p<0.001, Student’s t-test. Figure 2 HDAC5 and LSD1 physically interact in breast cancer cells. (a) MDA-MB-231 or MCF10A–CA1a cells were transfected with control vector pcDNA3.1 or pcDNA3.1-HDAC5 plasmids. Immunoprecipitation (IP) was performed with anti-LSD1 antibody followed by immunoblotting (IB) with anti-LSD1, anti-FLAG or anti-HDAC5 antibodies, respectively. (b) Whole cell lysates were immunoprecipitated with anti-LSD1 antibody followed by IB with anti-HDAC5 and LSD1 antibodies in indicated breast cancer cell lines. IgG was used as negative control. (c) MDA-MB-231 cells were transfected with control vector pcDNA3.1 or pcDNA3.1-HDAC5-FLAG plasmids, and IP was performed with anti-FLAG followed by IB with anti-LSD1 and anti-FLAG antibodies, respectively. (d) Schematic representation of full length and deletion mutants of HDAC5-FLAG constructs. (e) FLAG-tagged full-length or deletion mutants of HDAC5 were expressed in MDA-MB-231 cells. Extracts were immunoprecipitated with anti-FLAG antibody, and bound LSD1 was examined by IB using anti-LSD1 antibody. IB with anti-FLAG was used to detect the levels of FLAG-tagged HDAC5 full-length or deletion mutants in IP and input samples (10%). (f) MDA-MB-231 cells were transfected with plasmids expressing FLAG-tagged full-length or deletion mutants of HDAC5 proteins. Immunofluorescence study was performed using anti-FLAG antibody. DAPI was used as a control for nuclear staining. All the experiments were performed three times with similar results. Figure 3 HDAC5 stabilizes LSD1 protein in breast cancer cells. (a) MDA-MB-231 cells were transfected with control vector pcDNA3.1 or pcDNA3.1-HDAC5 for 48 h. mRNA expression of HDAC5 and LSD1 was measured by quantitative real-time PCR with β-actin as an internal control. (b) MDA-MB-231 cells were transfected with control vector pcDNA3.1 or pcDNA3.1-HDAC5 plasmids for 48 h. Effect of HDAC5 overexpression on LSD1 protein expression in MDA-MB-231 cells was evaluated by immunoblots with anti-LSD1 and anti-HDAC5 antibodies. (c) MDA-MB-231 cells were transfected with scramble siRNA or HDAC5 siRNA for 48 h. Effect of HDAC5 knockdown on LSD1 mRNA expression was examined by quantitative real-time PCR with β-actin as internal control. (d) Effect of HDAC5 siRNA on LSD1 protein expression in MDA-MB-231 cells. (e) Effect of depletion of LSD1 on mRNA expression of HDAC5 in MDA-MB-231-Scramble or MDA-MB-231-LSD1-KD cells. (f) Effect of LSD1-KD on protein expression of HDAC5 in MDA-MB-231-scramble or MDA-MB-231-LSD1-KD cells. (g) MDA-MB-231 cells were transfected with control vector pcDNA3.1, pcDNA3.1-HDAC5, scramble siRNA or HDAC5 siRNA for 48 h and analyzed by immunoblots for nuclear expression of indicated histone marks. PCNA was used as loading control. (h) Effect of HDAC5 overexpression or siRNA on LSD1 protein half-life in cycloheximide chase study. (i) Measurement of LSD1 half-life using Calcusyn program. (j) Effect of siRNA knockdown of LSD1 cofactors or class II HDACs on LSD1 protein level. All the experiments were performed three times. Bars represent the mean of three independent experiments ± s.d. * p<0.05, p<0.01, * p<0.001, Student’s t-test. Figure 4 HDAC5 regulates LSD1 by altering USP28 stability. (a) MDA-MB-231 cells transfected with pcDNA3.1-FLAG, pcDNA3.1-FLAG-HDAC5 or pcDNA3-HA-ubiquitin plasmids were treated with or without proteasome inhibitor 10µM MG132 for 10 h followed by immunoprecipitation (IP) using LSD1 antibody and immunoblots with anti-HA, LSD1 or HDAC5 antibodies. (b) Effect of siRNA of Jade-2, USP28 and HDAC5 on LSD1 protein expression in MDA-MB-231 cells. Results represent the mean of three independent experiments ± s.d. *** p<0.001, Student’s t-test. (c) MDA-MB-231 cells were transfected with scramble siRNA, HDAC5-siRNA, control vector pcDNA3.1, or pcDNA3.1-HDAC5 plasmids for 48 h. mRNA expression of Jade-2 and USP28 was measured by quantitative PCR. β-actin was used as an internal control. (d) MDA-MB-231 or MCF10A–CA1a cells were simultaneously transfected with pcDNA3.1-FLAG-Jade-2 and HDAC5 siRNA for 48 h and subjected to immunoblots with anti-HDAC5 or Jade-2 antibodies. β-actin was used as loading control to normalize target protein levels. (e) After MDA-MB-231 or MCF10A–CA1a cells were transfected with control vector pcDNA3.1 or pcDNA3.1-HDAC5 plasmids for 48 h, immunoblotting was performed for expression of HDAC5 and USP28. (f) MDA-MB-231 or MCF10A–CA1a cells were transfected with scramble or HDAC5 siRNA alone, or in combination with pDZ-USP28 for 48 h. Whole cell lysates were analyzed for protein levels of HDAC5, USP28 and LSD1. β-actin was used as loading control to normalize target protein levels. The experiments were performed three times with similar results. Figure 5 Effect of HDAC5 on protein acetylation of LSD1/USP28 and transcription of LSD1 target genes. (a) The immunoprecipitates of FLAG-M2 agarose from MDA-MB-231 cells overexpressing FLAG-tagged USP28 or FLAG-tagged LSD1 were used as substrates for protein deacetylation assay. IgG was used as negative control. Active or heat inactivated recombinant human GST-tagged HDAC5 protein were mixed with immunoprecipitates and incubated at 37°C for 6 h as described in “Materials and Methods”. The reactions were then subjected to immunoblots with anti-acetyl lysine antibody. FLAG-tagged USP28 or LSD1 proteins were probed with anti-FLAG antibody. HDAC5-GST protein was probed with anti-HDAC5 antibody. (b) Histograms represent the means of levels of acetyl-LSD1, acetyl-USP28 and acetyl-histone determined by quantitative immunoblotting using infrared immunoblotting detection and analysis. (c) MDA-MB-231 cell transfected with scramble or HDAC5 siRNAs for 48 h. LSD1 or IgG antibodies were added to cell lysate. Immunoprecipitation (IP) was performed with anti-LSD1 antibody followed by immunoblotting with anti-acetyl lysine and anti-LSD1 antibodies, respectively. Effect of HDAC5 siRNA on Acetyl-H3K9 protein expression in MDA-MB-231 cells was examined by immunoblotting with anti-acetyl-H3K9 antibody. (d) Histograms represent the means of relative levels of acetyl-LSD1 determined by quantitative immunoblotting using infrared immunoblotting detection and analysis. (e) mRNA expression of indicated genes in MDA-MB-231 cells transfected with scramble siRNA or HDAC5 siRNA. Data are means ± s.d. of three independent experiments. (f) Quantitative ChIP analysis was used to determine the occupancy by acetyl-H3K9, H3K4me2, LSD1, and HDAC5 at promoters of p21 or CLDN7 in MDA-MB-231 cells transfected with scramble or HDAC5 siRNA. p<0.05, p<0.01, ** p<0.001, Student’s t-test. Figure 6 HDAC5-LSD1 axis is implicated in breast cancer progression. (a) Depletion of HDAC5 by shRNA lentivirus infection downregulated LSD1 protein expression in MDA-MB-231 and MCF10A–CA1a cells. (b) Scramble shRNA and HDAC5-KD cells were analyzed for growth and viability by crystal violet assays. (c) Soft agar colony formation for HDAC5-KD and scramble control of MDA-MB-231 and MCF10A–CA1a cells. (d) Scramble shRNA and HDAC5-KD cells were harvested and stained for DNA with propidium iodide for flow cytometric analysis. The fractions corresponding to G1, S and G2/M phases of the cell cycle are indicated. (e) Boyden Chamber transwell migration assays for cell invasion for MDA-MB-231-Scramble or MDA-MB-231-HDAC5-KD-1 cells. (f) MDA-MB-231-Scramble or MDA-MB-231-LSD1-KD cells were transfected with control vector pcDNA3.1 or pcDNA3.1-HDAC5 for 5 days and crystal violet assays for growth were carried out. (g) MDA-MB-231-Scramble or MDA-MB-231-HDAC5-KD cells were transfected with empty or pReceiver-LSD1 expression plasmids for 5 days and crystal violet assays for growth were carried out. Bars represent the means of three independent experiments ± s.d. * p<0.05, p<0.01, * p<0.001, Student’s t-test. Figure 7 Effect of HDAC5 on growth and mutagen-induced tumorigenic transformation in MCF10A cells. (a) pcDNA3.1 or pcDNA3.1-HDAC5 transfected MCF10A cells (clone 1 and 2) were analyzed for protein levels of HDAC5 and LSD1 by immunoblots with anti-HDAC5 and anti-LSD1 antibodies. (b) Crystal violet assay for growth of MCF10-A stably transfected with control vector or pcDNA3.1-HDAC5 plasmids. (c) MCF10A–Vector-1 or MCF10A–HDAC5-1 cells were infected with scramble or LSD1 shRNA lentivirus particles for 5 days followed by crystal violet assays for growth. (d) MCF10A cells transfected with pcDNA3.1 or pcDNA3.1-HDAC5 plasmids were treated with DMSO or 500 ng/ml ICR191 for 7 months followed by soft agar colony formation assays. (e) After treatment with 500 ng/ml ICR191 for 7 months, MCF10A–HDAC5 cells were infected with scramble control or LSD1 shRNA lentivirus particles and soft agar colony formation assay was carried out. (f) Proposed model of the role of HDAC5-LSD1 axis in breast cancer development. Bars represent the means of three independent experiments ± s.d. p<0.01, * p<0.001, Student’s t-test. Conflicts of Interest The authors declare no conflict of interest. References 1 Shi Y Lan F Matson C Mulligan P Whetstine JR Cole PA Histone demethylation mediated by the nuclear amine oxidase homolog LSD1 Cell 2004 119 7 941 953 15620353 2 Lee MG Wynder C Cooch N Shiekhattar R An essential role for CoREST in nucleosomal histone 3 lysine 4 demethylation Nature 2005 437 7057 432 435 16079794 3 Huang Y Marton LJ Woster PM Casero RA Polyamine analogues targeting epigenetic gene regulation Essays Biochem 2009 46 95 110 20095972 4 Lee M Wynder C Cooch N Shiekhattar R An essential role for CoREST in nucleosomal histone 3 lysine 4 demethylation Nature 2005 437 7057 432 435 16079794 5 Garcia-Bassets I Kwon YS Telese F Prefontaine GG Hutt KR Cheng CS Histone methylation-dependent mechanisms impose ligand dependency for gene activation by nuclear receptors Cell 2007 128 3 505 518 17289570 6 Lim S Janzer A Becker A Zimmer A Schule R Buettner R Lysine-specific demethylase 1 (LSD1) is highly expressed in ER-negative breast cancers and a biomarker predicting aggressive biology Carcinogenesis 2010 31 3 512 520 20042638 7 Metzger E Wissmann M Yin N Muller J Schneider R Peters A LSD1 demethylates repressive histone marks to promote androgen-receptor-dependent transcription Nature 2005 437 7057 436 439 16079795 8 Zhu Q Huang Y Marton LJ Woster PM Davidson NE Casero RA Jr Polyamine analogs modulate gene expression by inhibiting lysine-specific demethylase 1 (LSD1) and altering chromatin structure in human breast cancer cells Amino Acids 2012 42 2–3 887 898 21805138 9 Huang Y Greene E Murray Stewart T Goodwin AC Baylin SB Woster PM Inhibition of lysine-specific demethylase 1 by polyamine analogues results in reexpression of aberrantly silenced genes Proc Natl Acad Sci U S A 2007 104 19 8023 8028 17463086 10 Vasilatos SN Katz TA Oesterreich S Wan Y Davidson NE Huang Y Crosstalk between lysine-specific demethylase 1 (LSD1) and histone deacetylases mediates antineoplastic efficacy of HDAC inhibitors in human breast cancer cells Carcinogenesis 2013 34 6 1196 1207 23354309 11 Serce N Gnatzy A Steiner S Lorenzen H Kirfel J Buettner R Elevated expression of LSD1 (Lysine-specific demethylase 1) during tumour progression from pre-invasive to invasive ductal carcinoma of the breast BMC clinical pathology 2012 12 13 22920283 12 Huang Y Stewart TM Wu Y Baylin SB Marton LJ Perkins B Novel oligoamine analogues inhibit lysine-specific demethylase 1 and induce reexpression of epigenetically silenced genes Clin Cancer Res 2009 15 23 7217 7228 19934284 13 Nowotarski SL Pachaiyappan B Holshouser SL Kutz CJ Li Y Huang Y Structure-activity study for (bis)ureidopropyl- and (bis)thioureidopropyldiamine LSD1 inhibitors with 3-5-3 and 3-6-3 carbon backbone architectures Bioorg Med Chem 2015 23 7 1601 1612 25725609 14 Mohammad HP Smitheman KN Kamat CD Soong D Federowicz KE Van Aller GS A DNA Hypomethylation Signature Predicts Antitumor Activity of LSD1 Inhibitors in SCLC Cancer Cell 2015 28 1 57 69 26175415 15 Yang XJ Gregoire S Class II histone deacetylases: from sequence to function, regulation, and clinical implication Mol Cell Biol 2005 25 8 2873 2884 15798178 16 Martin M Kettmann R Dequiedt F Class IIa histone deacetylases: regulating the regulators Oncogene 2007 26 37 5450 5467 17694086 17 McKinsey TA Zhang CL Lu J Olson EN Signal-dependent nuclear export of a histone deacetylase regulates muscle differentiation Nature 2000 408 6808 106 111 11081517 18 Martin M Kettmann R Dequiedt F Class IIa histone deacetylases: conducting development and differentiation The International journal of developmental biology 2009 53 2–3 291 301 19412888 19 Santner SJ Dawson PJ Tait L Soule HD Eliason J Mohamed AN Malignant MCF10CA1 cell lines derived from premalignant human breast epithelial MCF10AT cells Breast Cancer Res Treat 2001 65 2 101 110 11261825 20 Wu Y Wang Y Yang XH Kang T Zhao Y Wang C The deubiquitinase USP28 stabilizes LSD1 and confers stem-cell-like traits to breast cancer cells Cell reports 2013 5 1 224 236 24075993 21 Han X Gui B Xiong C Zhao L Liang J Sun L Destabilizing LSD1 by Jade-2 promotes neurogenesis: an antibraking system in neural development Mol Cell 2014 55 3 482 494 25018020 22 Huang Y Nayak S Jankowitz R Davidson NE Oesterreich S Epigenetics in breast cancer: what's new? Breast Cancer Res 2011 13 6 225 22078060 23 Abukhdeir AM Park BH P21 and p27: roles in carcinogenesis and drug resistance Expert Rev Mol Med 2008 10 e19 18590585 24 Kominsky SL Argani P Korz D Evron E Raman V Garrett E Loss of the tight junction protein claudin-7 correlates with histological grade in both ductal carcinoma in situ and invasive ductal carcinoma of the breast Oncogene 2003 22 13 2021 2033 12673207 25 Lin T Ponn A Hu X Law BK Lu J Requirement of the histone demethylase LSD1 in Snai1-mediated transcriptional repression during epithelial-mesenchymal transition Oncogene 2010 29 35 4896 4904 20562920 26 Chen WD Eshleman JR Aminoshariae MR Ma AH Veloso N Markowitz SD Cytotoxicity and mutagenicity of frameshift-inducing agent ICR191 in mismatch repair-deficient colon cancer cells J Natl Cancer Inst 2000 92 6 480 485 10716966 27 Zientek-Targosz H Kunnev D Hawthorn L Venkov M Matsui S Cheney RT Transformation of MCF-10A cells by random mutagenesis with frameshift mutagen ICR191: a model for identifying candidate breast-tumor suppressors Mol Cancer 2008 7 51 18534021 28 Milde T Oehme I Korshunov A Kopp-Schneider A Remke M Northcott P HDAC5 and HDAC9 in medulloblastoma: novel markers for risk stratification and role in tumor cell growth Clin Cancer Res 2010 16 12 3240 3252 20413433 29 He P Liang J Shao T Guo Y Hou Y Li Y HDAC5 promotes colorectal cancer cell proliferation by up-regulating DLL4 expression Int J Clin Exp Med 2015 8 4 6510 6516 26131280 30 Nagasawa S Sedukhina AS Nakagawa Y Maeda I Kubota M Ohnuma S LSD1 overexpression is associated with poor prognosis in basal-like breast cancer, and sensitivity to PARP inhibition PloS one 2015 10 2 e0118002 25679396 31 Piao L Suzuki T Dohmae N Nakamura Y Hamamoto R SUV39H2 methylates and stabilizes LSD1 by inhibiting polyubiquitination in human cancer cells Oncotarget 2015 6 19 16939 16950 26183527 32 Shi YJ Matson C Lan F Iwase S Baba T Shi Y Regulation of LSD1 histone demethylase activity by its associated factors Mol Cell 2005 19 6 857 864 16140033 33 Greco TM Yu F Guise AJ Cristea IM Nuclear import of histone deacetylase 5 by requisite nuclear localization signal phosphorylation Molecular & cellular proteomics : MCP 2011 10 2 M110004317 34 Diefenbacher ME Popov N Blake SM Schulein-Volk C Nye E Spencer-Dene B The deubiquitinase USP28 controls intestinal homeostasis and promotes colorectal cancer J Clin Invest 2014 124 8 3407 3418 24960159 35 Popov N Wanzel M Madiredjo M Zhang D Beijersbergen R Bernards R The ubiquitin-specific protease USP28 is required for MYC stability Nat Cell Biol 2007 9 7 765 774 17558397 36 Zhang D Zaugg K Mak TW Elledge SJ A role for the deubiquitinating enzyme USP28 in control of the DNA-damage response Cell 2006 126 3 529 542 16901786 37 Sen N Kumari R Singh MI Das S HDAC5, a key component in temporal regulation of p53-mediated transactivation in response to genotoxic stress Mol Cell 2013 52 3 406 420 24120667 38 Traynor AM Dubey S Eickhoff JC Kolesar JM Schell K Huie MS Vorinostat (NSC# 701852) in patients with relapsed non-small cell lung cancer: a Wisconsin Oncology Network phase II study J Thorac Oncol 2009 4 4 522 526 19347984 39 Luu TH Morgan RJ Leong L Lim D McNamara M Portnow J A phase II trial of vorinostat (suberoylanilide hydroxamic acid) in metastatic breast cancer: a California Cancer Consortium study Clin Cancer Res 2008 14 21 7138 7142 18981013 40 Modesitt SC Sill M Hoffman JS Bender DP A phase II study of vorinostat in the treatment of persistent or recurrent epithelial ovarian or primary peritoneal carcinoma: a Gynecologic Oncology Group study Gynecol Oncol 2008 109 2 182 186 18295319 41 Blumenschein GR Jr Kies MS Papadimitrakopoulou VA Lu C Kumar AJ Ricker JL Phase II trial of the histone deacetylase inhibitor vorinostat (Zolinza, suberoylanilide hydroxamic acid, SAHA) in patients with recurrent and/or metastatic head and neck cancer Invest New Drugs 2008 26 1 81 87 17960324 42 Shaw PG Chaerkady R Wang T Vasilatos S Huang Y Van Houten B Integrated proteomic and metabolic analysis of breast cancer progression PloS one 2013 8 9 e76220 24086712 43 Varghese F Bukhari AB Malhotra R De A IHC Profiler: an open source plugin for the quantitative evaluation and automated scoring of immunohistochemistry images of human tissue samples PloS one 2014 9 5 e96801 24802416 44 Ishibashi H Suzuki T Suzuki S Moriya T Kaneko C Takizawa T Sex steroid hormone receptors in human thymoma J Clin Endocrinol Metab 2003 88 5 2309 2317 12727990 45 Huang Y Hager ER Phillips DL Dunn VR Hacker A Frydman B A novel polyamine analog inhibits growth and induces apoptosis in human breast cancer cells Clin Cancer Res 2003 9 7 2769 2777 12855657 46 Huang Y Keen J Pledgie A Marton L Zhu T Sukumar S Polyamine analogues down-regulate estrogen receptor alpha expression in human breast cancer cells J Biol Chem 2006 281 28 19055 19063 16679312 47 Katz TA Vasilatos SN Harrington E Oesterreich S Davidson NE Huang Y Inhibition of histone demethylase, LSD2 (KDM1B), attenuates DNA methylation and increases sensitivity to DNMT inhibitor-induced apoptosis in breast cancer cells Breast Cancer Res Treat 2014 146 1 99 108 24924415 48 Huang Y Vasilatos SN Boric L Shaw PG Davidson NE Inhibitors of histone demethylation and histone deacetylation cooperate in regulating gene expression and inhibiting growth in human breast cancer cells Breast Cancer Res Treat 2012 131 3 777 789 21452019
PMC005xxxxxx/PMC5121929.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 9005353 1149 Cytokine Cytokine Cytokine 1043-4666 1096-0023 27701021 5121929 10.1016/j.cyto.2016.08.028 NIHMS820597 Article Endothelial Adhesion Molecules and Multiple Organ Failure in Patients With Severe Sepsis Amalakuhan Bravein 1 Habib Sheila A. 1 Mangat Mandeep 1 Reyes Luis F. 12 Rodriguez Alejandro H. 6 Hinojosa Cecilia A. 1 Soni Nilam J. 12 Gilley Ryan P. 1 Bustamante Carlos A. 3 Anzueto Antonio 12 Levine Stephanie M. 12 Peters Jay I. 12 Aliberti Stefano 4 Sibila Oriol 5 Chalmers James D. 7 Torres Antoni 8 Waterer Grant W. 9 Martin-Loeches Ignacio 10 Bordon Jose 11 Blanquer Jose 12 Sanz Francisco 13 Marcos Pedro J. 14 Rello Jordi 15 Ramirez Julio 16 Solé-Violán Jordi 17 Luna Carlos M. 18 Feldman Charles 19 Witzenrath Martin 20 Wunderink Richard G. 21 Stolz Daiana 22 Wiemken Tim L. 16 Shindo Yuichiro 23 Dela Cruz Charles S. 24 Orihuela Carlos J. 125 Restrepo Marcos I. 12 1 University of Texas Health Science Center San Antonio, San Antonio, TX, USA 2 South Texas Veterans Health Care System, Barranquilla, Colombia 3 Camino Distrital Universitario Adelita de Char, Barranquilla, Colombia 4 University of Milan Bicocca, Clinica Pneumologica, Monza, Italy 5 Servei de Pneumologia, Departament de Medicina, Hospital Santa Creu i Sant Pau, Universitat Autònoma de Barcelona, Barcelona, España 6 Critical Care Department, Joan XXIII University Hospital and Pere Virgili Health Institute, CIBERES, Tarragona, Spain 7 University of Dundee, Dundee, UK 8 Hospital Clinic, Universitat de Barcelona, Servei de Pneumologia, Barcelona, Spain 9 School of Medicine and Pharmacology, Royal Perth Hospital Unit, University of Western Australia, Perth, Australia 10 St. James’s Hospital, Trinity Centre for Health Sciences, CIBERES, Dublin, Ireland 11 Department of Medicine, Section of Infectious Diseases, Providence Hospital, DC, USA 12 Unidad Cuidados Intensivos Respiratorios, Hospital Clínic Universitari, Valencia, España 13 Pulmonology Department, Consorci Hospital General Universitari de Valencia, Valencia, Spain 14 Servicio de Neumología, Instituto de investigación Biomédica de A Coruña (INIBIC), Complejo Hospitalario Universitario de A Coruña (CHUAC), Sergas. Universidade da Coruña (UDC). A Coruña, Spain 15 Critical Care Department, Hospital Universitario Vall d’Hebron, CIBERES, Barcelona, Spain 16 Division of Infectious Diseases, University of Louisville, Louisville, KY, USA 17 Intensive Care Unit, Hospital Universitario Dr. Negrín, CIBERES, Las Palmas de Gran Canaria, Spain 18 Division of Pulmonary Medicine, Department of Medicine, Hospital de Clinicas, Division of Pulmonology, Universidad de Buenos Aires, Buenos Aires, Argentina 19 Department of Internal Medicine, Charlotte Maxeke Johannesburg Academic Hospital, Johannesburg, South Africa; Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa 20 Department of Infectious Diseases and Pulmonary Medicine, Charité-Universitätsmedizin Berlin and SFB-TR84 “Innate Immunity of the Lung”, Berlin, Germany 21 Division of Pulmonary and Critical Care Medicine, Department of Medicine, Feinberg School of Medicine, Northwestern University, Chicago, IL 22 Clinic of Pulmonary Medicine and Respiratory Cell Research, University Hospital Basel, Basel, Switzerland 23 Institute for Advanced Research and Department of Respiratory Medicine, Nagoya University Graduate School of Medicine, Nagoya, Japan 24 Section of Pulmonary, Critical Care and Sleep Medicine, Yale University School of Medicine, New Haven, Connecticut 25 Department of Microbiology, The University of Alabama at Birmingham, Birmingham, Alabama, USA Corresponding author: Marcos I. Restrepo, MD, MSc; South Texas Veterans Health Care System ALMD - 7400 Merton Minter Boulevard - San Antonio Texas, 78229; Phone: (210)-617-5300 ext. 15413 - Fax: (210) 567-4423; [email protected] 26 10 2016 1 10 2016 12 2016 01 12 2017 88 267273 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Objective To determine if serum levels of endothelial adhesion molecules were associated with the development of multiple organ failure (MOF) and in-hospital mortality in adult patients with severe sepsis. Design This study was a secondary data analysis of a prospective cohort study. Setting Patients were admitted to two tertiary intensive care units in San Antonio, TX, between 2007 and 2012. Patients Patients with severe sepsis at the time of intensive care unit (ICU) admission were enrolled. Inclusion criteria were consistent with previously published criteria for severe sepsis or septic shock in adults. Exclusion criteria included immunosuppressive medications or conditions. Interventions None. Measurements Baseline serum levels of the following endothelial cell adhesion molecules were measured within the first 72 hours of ICU admission: Intracellular Adhesion Molecule 1 (ICAM-1), Vascular Cell Adhesion Molecule-1 (VCAM-1), and Vascular Endothelial Growth Factor (VEGF). The primary and secondary outcomes were development of MOF (≥2 organ dysfunction) and in-hospital mortality, respectively. Main results Forty-eight patients were enrolled in this study, of which 29 (60%) developed MOF. Patients that developed MOF had higher levels of VCAM-1 (p=0.01) and ICAM-1 (p=0.01), but not VEGF (p=0.70) compared with patients without MOF (single organ failure only). The area under the curve (AUC) to predict MOF according to VCAM-1, ICAM-1 and VEGF was 0.71, 0.73, and 0.54, respectively. Only increased VCAM-1 levels were associated with in-hospital mortality (p=0.03). These associations were maintained even after adjusting for APACHE and SOFA scores using logistic regression. Conclusions High levels of serum ICAM-1 was associated with the development of MOF. High levels of VCAM-1 was associated with both MOF and in-hospital mortality. Biomarkers Sepsis Shock Mortality Multiple Organ Failure Intracellular Adhesion Molecule-1 Vascular Cell Adhesion Molecule-1 and Vascular Endothelial Growth Factor INTRODUCTION Mortality rates for patients with severe sepsis and septic shock range from 40–60%, costing the United States health care system approximately $17 billion annually [1]. To expedite the initiation of effective treatments and thereby reduce mortality and associated costs, better methods to identify patients with severe sepsis are needed Hemodynamic parameters and laboratory tests, including lactate levels are currently used to predict multiple organ failure (MOF) in patients with sepsis; however information regarding how to detect early organ failure is limited [2, 3]. Clinical deterioration and death result from a complex interaction between inflammation and coagulation that leads to organ dysfunction [4]. Vascular endothelial damage precedes organ dysfunction and plays an important role by increasing vascular permeability, promoting activation of the coagulation cascade and compromising regional perfusion in vital organs (e.g. kidneys, liver, gut, etc.) [5]. Different biomarkers have been proposed to assess vascular endothelial damage in patients with sepsis and the development of MOF and mortality [6]. Cell adhesion molecules (CAMs) have emerged as potential biomarkers that may be used to detect early endothelial injury in septic patients [7]. Vascular Endothelial Growth Factor (VEGF), Intracellular Adhesion Molecule 1 (ICAM-1), and Vascular Cell Adhesion Molecule-1 (VCAM-1) are a group of transmembrane CAM proteins that are responsible for the cell adhesion process. These CAMs allow cells to interact with the extracellular matrix, the cytoskeleton, and other cells within the vascular endothelium, as well as other cells within the circulation [8]. ICAM-1 and VCAM-1 are present in the cell membrane of both leukocytes and the cells lining the vascular endothelium, allowing inflammatory cells to transmigrate into nearby tissues [8]. These adhesion molecules are expressed in very large quantities in patients with an uncontrolled inflammatory state such as sepsis [9]. As a result, some of these CAMs also leak into circulation, and especially after the vascular endothelial injury that occurs during sepsis, making them measurable. The value of VEGF for prediction of clinical outcomes in patients with sepsis is at present still controversial[10, 11], and limited data are available in adult patients with sepsis regarding ICAM-1 and VCAM-1 levels [12]. Therefore, more studies are needed to evaluate the role of CAMs in predicting the outcomes of patients with sepsis. Our hypothesis was that higher levels of CAMs are related to a higher incidence of MOF and in-hospital mortality. Therefore, the aim of this study was to determine the association of levels of CAMs with the development of MOF and in-hospital mortality in adult patients with severe sepsis or septic shock. MATERIAL AND METHODS Study Design This study was a secondary analysis of the data derived from a cohort of patients admitted to the intensive care unit (ICU) with severe sepsis or septic shock at two hospitals (South Texas Veterans Health Care System and University Hospital, San Antonio, TX), between 2007 and 2012, as previously descried [13]. This study was approved by the local institutional review board (HSC20070713H), and is posted on www.clinicaltrials.gov (NCT00708799). All participants signed a consent form before entry into the study. Inclusion criteria for enrollment included age >18 years, written informed consent obtained from the patient/patient’s legal representative, and criteria for severe sepsis or septic shock [14]. Diagnosis of severe sepsis was established according to the International Sepsis Definition in which patients with Systemic Inflammatory Response Syndrome (SIRS) with suspected or proven source of infection develop at least one organ dysfunction [15]. Septic shock was defined as the need for vasopressor support in septic patients with hypotension non-responsive to fluid resuscitation [16]. Patients who developed MOF before blood collection were excluded. Immunosuppressed patients were excluded if they fulfilled the following definitions: 1) Chemotherapy within the last one month, 2) Leukemia/lymphoma not in remission, 3) Solid organ or bone marrow stem cell transplant, 4) HIV infection with CD4+ T lymphocyte count <200 cells/mm3, or 5) Chronic steroid use, defined as > 10 mg/day of prednisone. Enrollment and Follow-Up All patients were screened for eligibility at the time of admission to the ICU and followed daily until hospital discharge. Both the ‘Acute Physiology and Chronic Health Evaluation (APACHE) II’ and the ‘Sepsis-related Organ Failure Assessment (SOFA) scores were obtained for each patient during the first 24-hours of admission to the ICU [17]. Clinical Outcomes The primary outcome of this study was development of MOF and the secondary outcome was in-hospital mortality. Multiple organ failure was defined as two or more organ failures during the ICU hospitalization among patients with severe sepsis or septic shock. Organ failure criteria are shown in Table 1. All the patients had their blood collected before the onset of MOF, and within 96 hours post ICU admission. Biomarkers and Assays Venous blood was drawn from patients within the first 72 hours of ICU admission. Serum ICAM-1, VCAM-1 and VEGF were measured using a commercially available Human Inflammation Panel kit from Luminex Technology that was analyzed at Myriad Rules Based Medicine Inc. (Austin, Texas). Statistical Analysis Categorical variables were compared between groups using Fisher’s exact test. Continuous variables were evaluated using non-parametric analysis with the Mann-Whitney U Test. Values are expressed as median (IQR). Statistical significance was defined as p-value ≤ 0.05. A receiver operating characteristic (ROC) curve was developed to assess the accuracy of ICAM-1, VCAM-1 and VEGF to predict outcomes. Multivariate analysis was performed using multiple logistic regressions to evaluate the relation of serum levels of ICAM-1, VCAM-1 and VEGF with the proposed outcomes after adjusting the analysis with the APACHE II and SOFA scores. All statistical analyses were performed with IBM SPSS, Statistics for Macintosh, version 23.0 (Armonk, NY: IBM Corp). RESULTS Patient characteristics Forty-eight patients were included in the study cohort, of which 29 (60%) developed MOF. There were no significant differences in baseline characteristics among severe septic patients who developed MOF compared with those that did not develop MOF (Table 2). Severe septic patients with MOF had higher median APACHE II scores on admission compared with those without MOF (21[IQR 15,28] vs. 19[14,21]; p=0.04). Outcomes Patients with severe sepsis who developed MOF had higher levels of VCAM-1 (median 1,690 ng/mL [IQR 1,065, 3,590] vs. 1,090 ng/mL [789, 1,410]; p=0.01) and ICAM-1 (median 362 ng/mL [IQR 270, 449] vs. median 258 ng/mL [IQR199, 341]; p=0.01), compared to patients with severe sepsis alone respectively (Figure 1). This association remained significant even after controlling for APACHE II and SOFA scores in the logistic regression analysis (odds ratio (OR)=1.2, 95%CI: 1.1–1.3, p=0.01). No significant association between VEGF and the development of MOF was found (p=0.70). Figure 2 shows the number of cases of MOF and its relation with serum levels of VCAM-1, ICAM-1 and VEGF. The area under the ROC curve (AUC) for MOF was 0.73 for VCAM-1, 0.70 for ICAM-1 and 0.53 for VEGF (Figure 3). Only higher VCAM-1 levels showed a statistically significant association with in-hospital mortality (median 2,210 ng/mL [IQR 1,500, 3,432] vs. 1,240 ng/mL [806, 1,882]; p=0.03). This association was maintained even after controlling for APACHE II and SOFA scores in the logistic regression analysis (OR=1.06, 95%CI: 1.02 –1.11, p=0.02). The AUC for in-hospital mortality for CAMs are shown in Figure 3. DISCUSSION The main finding of our study was uncovering the major role of endothelial adhesion molecules in the clinical evolution of critically ill patients. Whilst ICAM-1 played a role in the development of MOF alone, VCAM-1 showed a significant association not only with MOF but also with the clinical outcome of hospital mortality. Previous studies have shown that VCAM-1 is usually expressed at low levels in the membranes of leukocytes, macrophages, and vascular endothelial cells [5]. Infection increases transcription of VCAM-1, which is expressed on the surface of vascular endothelial cells [5]. Leukocytes activated by inflammatory mediators bind to the VCAM-1 endothelial surface receptors and translocate into local tissues to combat infection [5]. During states of diffuse and uncontrolled inflammation, as in sepsis, VCAM-1 is expressed in large quantities. The ensuing leukocyte adhesion/transmigration and the associated inflammatory cascade have been linked to vascular endothelial damage, capillary leakage, and organ dysfunction [5]. Several in vitro studies have demonstrated this phenomenon, but few clinical studies have evaluated the significance of elevated VCAM-1 levels during sepsis. One study in neonates demonstrated that higher levels of VCAM-1 were associated with more severe forms of sepsis and MOF [18]. These findings are consistent with our experimental results and the notion that elevated serum levels of VCAM-1 at the onset of organ dysfunction predicts the development of MOF in adult patients with severe sepsis and septic shock. However, there is a paucity of data evaluating the association of VCAM-1 levels with the clinical outcome of mortality. Our study is among the first to assess in-hospital mortality, finding that in-hospital mortality was in fact associated with higher VCAM-1 levels. During inflammatory states, ICAM-1 has the same role as VCAM-1, a vascular endothelial surface receptor that allows for leukocytes and other inflammatory cells to bind and translocate into local tissues [5]. Thus, ICAM-1 plays a similar role to VCAM-1 in capillary leakage and organ dysfunction when the normally localized inflammatory cascade becomes wider spread [5]. Considering these similarities to VCAM-1, previous studies have shown that higher serum levels of ICAM-1 are also associated with severe sepsis and MOF in neonates [12]. Furthermore, multiple mouse models have shown that ICAM-1 knockout mice with severe forms of sepsis have lower mortality rates [5, 19]. Our finding that higher serum levels of ICAM-1 predict development of MOF in patients with severe sepsis and septic shock is a novel finding in the adult population. In contrast to studies with mouse models, ICAM-1 levels were not associated with increased in-hospital mortality in our sample of adult patients with severe sepsis and septic shock. A non-significant statistical association may have been due to the small sample size of our study. Alternatively, the pathological role of VCAM-1 may be more pronounced during sepsis compared to ICAM-1. This may explain why VCAM-1 levels were found to correlate more highly with vascular endothelial damage and poor clinical outcomes. However, this speculation requires further testing to confirm. VEGF also plays an integral role in the integrity of the vascular endothelium. VEGF has been shown to induce the expression of both VCAM-1 and ICAM-1, and to contribute to vascular endothelial damage and capillary leakage due to up-regulation during sepsis[20]. Unlike VCAM-1 and ICAM-1, VEGF has a multitude of other roles. In adults with sepsis, high VEGF levels have been shown to increase vascular permeability by binding to the tyrosine-protein kinase receptor-1 (FLT-1), altering the configuration of endothelial actin filaments, and increasing fenestrations in the endothelium [21]. Consequently, the vascular endothelium is predisposed to damage, ultimately resulting in organ dysfunction [21]. FLT-1 receptor antagonists/anti-VEGF receptor antibodies have been shown to decrease mortality in septic mice[21]. VEGF also has a procoagulant effect, causing microthrombi in the peripheral vasculature [21]. Yet despite these diverse functions that are active during sepsis, an association for high VEGF levels on mortality in human clinical studies have yielded inconsistent results [10, 11]. Along such lines, our study demonstrated that VEGF levels were not associated with MOF or mortality. This was consistent with the study by Karlsson at al [11] that involved 215 patients with severe sepsis and septic shock. Of note, Van Der Flier et al. in 2005 showed a positive association of VEGF levels and mortality in a sample size of only 18 septic adult patients [10]. Baseline patient characteristics and differences in sample size are potential explanations for the discrepancy in these findings, and as such further investigation is required. Also of note, Jiang et al [22] found that the VEGF-to-platelet ratio was predictive of 28-day mortality in patients with sepsis in China, illustrating the need for further investigation. Our results are novel and supported by a recent systematic review that summarize the data on multiple endothelial-derived molecules including the ones reported in our study [23]. The key difference between our study and those cited in this review were the comparison/control groups which in fact strengthens the clinical utility of our results. Among the 19 studies that reported data on sICAM-1 the comparison groups included mainly healthy controls, patients with trauma, postoperative patients, patients with other forms of shock, and non-septic ICU patients [23]. Similarly, in the 12 studies that evaluated sVCAM-1 in sepsis, they evaluated patients in the emergency department, postoperative patients, critically ill trauma patients and patients with sepsis not necessarily with organ dysfunction [23]. Moreover, the four studies that examined sVEGF in septic patients compared non-septic critically ill patients to healthy controls [23]. Our study compared the levels of ICAM-1, VCAM-1 and VEGF in septic patients with one organ dysfunction to those with MOF. It is clear from previous studies that levels of these markers are elevated in comparison to healthy controls or non-septic patients, but what is not clear is whether they are elevated in comparison to those with single organ dysfunction. This is clinically useful because our study illustrates that ICAM-1 and VCAM-1 are even more sensitive of a predictor (and more specific) for MOF and mortality than previously thought. They are not only significantly elevated in septic patients compared to health controls, but even significantly elevated when comparing to patients with MOF compared to those with single organ damage, with threshold values determined in this study to be significantly predictive of in-hospital and 30-day mortality. Said in another way, the results from our study predict further organ dysfunction even when the process has already started. Furthermore, of the 12 studies that evaluated sVCAM-1, only 1 found that it was statistically predictive of in-hospital mortality [23]. Our study was the second study to do this. The methodology preferred to address organ dysfunction is the use of the SOFA score, which is usually assessed at the time of presentation and at 24 hours after enrollment [24]. Skibsted S et al showed a weak correlation of sICAM-1 (r=0.15) and sVCAM-1 (r=0.35) with the SOFA score, which although they reported a statistically significant p-value for this correlation, is not likely clinically significant. Considering this, there are still much unanswered questions regarding sICAM-1 and sVCAM-1 when it comes to their correlation with SOFA score and ultimately mortality. Although this study is far from definitive and needs further study and validation, our study serves to add evidence that VCAM-1 levels are indeed able to predict mortality. The only biomarker that showed a significant association with the SOFA score was soluble fms-like tyrosine kinase (sFIT-1; r=0.58). Similar results were reported by Shapiro NI et al [25] in which only sFIT-1 had the strongest association with SOFA score, but not sICAM-1 and sVCAM-1. In addition, only one study addressed the correlation of sVCAM-1 and showed a modest correlation with SOFA and APACHE II scores [25]. Therefore, a progression from one to two or more organ dysfunction may be determined by the use of sICAM-1 and sVCAM-1 as we suggest in our study. To put this in prospective, in our study, of those patients that had single organ dysfunction (the inclusion criteria to be included in our study in the first place), 60% went on to develop MOF, and our results indicate that sICAM-1 and sVCAM-1 could predict this. Furthermore, of those 29 (60%) that developed MOF, sVCAM-1 levels could predict in-hospital. Our study has potential limitations including the small sample size, however these patients with severe sepsis and septic shock were well characterized. Other limitations included the absence of a control group and that serial follow-up of the different biomarkers were not performed. Generalizations based on these results require further exploration in larger cohorts. Several other mechanisms may also be associated with the development of MOF beyond the endothelial adhesion molecules. Finally, it is important to clarify that all the patients had a blood collection before the MOF developed and within 96 hours of ICU admission, leading to a gap of 72 hours were potential therapeutic strategies could be used. CONCLUSIONS High serum levels of VCAM-1 and ICAM-1 at the onset of acute organ dysfunction are associated with the development of MOF, with high VCAM-1 levels also being associated with higher in-hospital mortality. These biomarkers have the potential to assist in early recognition and initiation of appropriate therapies to ultimately improve clinical outcomes. Further studies are needed to investigate the role of VCAM-1, ICAM-1, and VEGF as prognostic markers in patients with sepsis. Figure 1 Boxplots representing serum levels of VCAM-1 (Panel A), ICAM-1 (Panel B) and VEGF (Panel C) according to multiple organ failure (MOF) and in-hospital mortality Figure 2 Distribution of VCAM-1, ICAM-1 and VEGF levels by the number of subjects according to the presence or absence of multiple organ failure (MOF). Figure 3 ROC curves of VCAM-1 (Panel A), ICAM-1 (Panel B) and VEGF (Panel C) predict multiple organ failure (MOF) and in-hospital mortality. Table 1 Organ failure criteria Organ/system Criteria Cardiovascular Mean arterial pressure <65 mmHg or requirement for vasopressors or inotropes after appropriate volume resuscitation (30 mL/Kg) Respiratory PaO2/FiO2 <300 or requirement for mechanical ventilation or non-invasive ventilation Renal Patients who receive renal replacement therapy or serum creatinine ≥354 umol/L (4 mg/dL) or urine output <0.3 mL/kg/h Hematologic Prothrombin time and partial thromboplastin time >1.5 – 3 times the normal range or platelet count <100.000/mm3 Hepatic Alanine aminotransferase and aspartate aminotransferase > 100 Units/L or total plasma bilirubin >1 mg/dL (0.17 mmol/L) Central nervous system Glasgow coma scale score <12 points in absence of sedation Table 2 Baseline characteristics of patients with severe sepsis stratified according to the presence of multiple organ failure (MOF) during ICU hospitalization Characteristic No MOF (n=19) MOF (n=29) p Value Demographic Male 19 (100) 27 (93) 0.60 Age, mean (IQR) 59 (51, 65) 58 (45, 79) 0.60 Comorbid conditions, n (%) Obesity 7 (37) 12 (41) 0.50 Active cancer 2 (10) 0 (0) 0.20 Prior cancer 3 (16) 5 (17) 0.60 Cardiovascular disease 7 (37) 5 (17) 0.10 Chronic heart failure 3 (16) 1 (3) 0.20 COPD 3 (16) 3 (10) 0.40 Chronic kidney disease 3 (16) 2 (7) 0.30 Depression 2 (10) 7 (24) 0.20 Diabetes mellitus 10 (53) 13 (45) 0.40 HIV infection 1 (5) 0 (0) 0.70 Hyperlipidemia 4 (17) 1 (3) 0.60 Leukemia 1 (5) 0 (0) 0.40 Liver disease 1 (5) 2 (7) 0.60 Tobacco use 6 (32) 8 (28) 0.50 Alcohol use 4 (21) 7 (24) 0.50 Asthma 1 (5) 3 (10) 0.50 Source of infection, n (%) Pulmonary 7 (37) 8 (28) 0.50 Urinary tract 6 (32) 10 (34) 0.30 Gastrointestinal 2 (10) 4 (14) 0.40 Skin 3 (16) 5 (17) 0.30 Endocarditis 0 (0) 1 (3) 0.90 IQR, interquartile range Highlights VEGF, ICAM-1, and VCAM-1 are a group of key vascular endothelial proteins Elevated ICAM-1 levels predict multi-organ failure (MOF) during sepsis Elevated VCAM-1 levels predict MOF and in-hospital mortality during sepsis Conflict of interest: All authors have no conflicts of interest. Dr. Restrepo’s time is partially protected by Award Number K23HL096054 from the National Heart, Lung, and Blood Institute. Dr. Rodriguez’s time is protected by grant M-BAE 15/00063 from Carlos III Health Institute from Spain. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Heart, Lung, And Blood Institute or the National Institutes of Health or the Department of Veterans Affairs. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. 1 Martin GS Sepsis, severe sepsis and septic shock: changes in incidence, pathogens and outcomes Expert Rev Anti Infect Ther 10 6 2012 701 6 22734959 2 De Backer D Donadello K Sakr Y Ospina-Tascon G Salgado D Scolletta S Vincent JL Microcirculatory alterations in patients with severe sepsis: impact of time of assessment and relationship with outcome Crit Care Med 41 3 2013 791 9 23318492 3 Wang B Chen G Cao Y Xue J Li J Wu Y Correlation of lactate/albumin ratio level to organ failure and mortality in severe sepsis and septic shock J Crit Care 30 2 2015 271 5 25537574 4 de Montmollin E Annane D Year in review 2010: Critical Care--Multiple organ dysfunction and sepsis Crit Care 15 6 2011 236 22146601 5 van Griensven M Probst C Muller K Hoevel P Pape HC Leukocyte-endothelial interactions via ICAM-1 are detrimental in polymicrobial sepsis Shock 25 3 2006 254 9 16552357 6 De Backer D Orbegozo Cortes D Donadello K Vincent JL Pathophysiology of microcirculatory dysfunction and the pathogenesis of septic shock Virulence 5 1 2014 73 9 24067428 7 Steeber DA Tedder TF Adhesion molecule cascades direct lymphocyte recirculation and leukocyte migration during inflammation Immunol Res 22 2–3 2000 299 317 11339364 8 Kerr JR Cell adhesion molecules in the pathogenesis of and host defence against microbial infection Mol Pathol 52 4 1999 220 30 10694943 9 Parent C Eichacker PQ Neutrophil and endothelial cell interactions in sepsis. The role of adhesion molecules Infect Dis Clin North Am 13 2 1999 427 47 x 10340176 10 van der Flier M van Leeuwen HJ van Kessel KP Kimpen JL Hoepelman AI Geelen SP Plasma vascular endothelial growth factor in severe sepsis Shock 23 1 2005 35 8 15614129 11 Karlsson S Pettila V Tenhunen J Lund V Hovilehto S Ruokonen E G. Finnsepsis Study Vascular endothelial growth factor in severe sepsis and septic shock Anesth Analg 106 6 2008 1820 6 18499616 12 Figueras-Aloy J Gomez-Lopez L Rodriguez-Miguelez JM Salvia-Roiges MD Jordan-Garcia I Ferrer-Codina I Carbonell-Estrany X Jimenez-Gonzalez R Serum soluble ICAM-1, VCAM-1, L-selectin, and P-selectin levels as markers of infection and their relation to clinical severity in neonatal sepsis Am J Perinatol 24 6 2007 331 8 17564956 13 Hsu CH Reyes LF Orihuela CJ Buitrago R Anzueto A Soni NJ Levine S Peters J Hinojosa CA Aliberti S Sibila O Rodriguez A Chalmers JD Martin-Loeches I Bordon J Blanquer J Sanz F Marcos PJ Rello J Sole-Violan J Restrepo MI Chromogranin A levels and mortality in patients with severe sepsis Biomarkers 20 3 2015 171 6 26154393 14 Dellinger RP Levy MM Rhodes A Annane D Gerlach H Opal SM Sevransky JE Sprung CL Douglas IS Jaeschke R Osborn TM Nunnally ME Townsend SR Reinhart K Kleinpell RM Angus DC Deutschman CS Machado FR Rubenfeld GD Webb SA Beale RJ Vincent JL Moreno R Surviving Sepsis Campaign Guidelines Committee including the Pediatric, Surviving sepsis campaign: international guidelines for management of severe sepsis and septic shock: 2012 Crit Care Med 41 2 2013 580 637 23353941 15 Bone RC Balk RA Cerra FB Dellinger RP Fein AM Knaus WA Schein RM Sibbald WJ Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine Chest 101 6 1992 1644 55 1303622 16 Levy MM Fink MP Marshall JC Abraham E Angus D Cook D Cohen J Opal SM Vincent JL Ramsay G Sccm/Esicm/Accp/Ats/Sis, 2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference Crit Care Med 31 4 2003 1250 6 12682500 17 Knaus WA Draper EA Wagner DP Zimmerman JE APACHE II: a severity of disease classification system Crit Care Med 13 10 1985 818 29 3928249 18 Whalen MJ Doughty LA Carlos TM Wisniewski SR Kochanek PM Carcillo JA Intercellular adhesion molecule-1 and vascular cell adhesion molecule-1 are increased in the plasma of children with sepsis-induced multiple organ failure Crit Care Med 28 7 2000 2600 7 10921602 19 Hildebrand F Pape HC Harwood P Muller K Hoevel P Putz C Siemann A Krettek C van Griensven M Role of adhesion molecule ICAM in the pathogenesis of polymicrobial sepsis Exp Toxicol Pathol 56 4–5 2005 281 90 15816357 20 Kim I Moon SO Kim SH Kim HJ Koh YS Koh GY Vascular endothelial growth factor expression of intercellular adhesion molecule 1 (ICAM-1), vascular cell adhesion molecule 1 (VCAM-1), and E-selectin through nuclear factor-kappa B activation in endothelial cells J Biol Chem 276 10 2001 7614 20 11108718 21 Yano K Liaw PC Mullington JM Shih SC Okada H Bodyak N Kang PM Toltl L Belikoff B Buras J Simms BT Mizgerd JP Carmeliet P Karumanchi SA Aird WC Vascular endothelial growth factor is an important determinant of sepsis morbidity and mortality J Exp Med 203 6 2006 1447 58 16702604 22 Jiang W Ouyang W Chen C Zhu G Huang L Zeng H Significance of the ratio of plasma vascular endothelial growth factor level to platelet count in the prognosis of patients with sepsis Zhonghua Wei Zhong Bing Ji Jiu Yi Xue 26 7 2014 484 8 25027426 23 Xing K Murthy S Liles WC Singh JM Clinical utility of biomarkers of endothelial activation in sepsis--a systematic review Crit Care 16 1 2012 R7 22248019 24 Skibsted S Jones AE Puskarich MA Arnold R Sherwin R Trzeciak S Schuetz P Aird WC Shapiro NI Biomarkers of endothelial cell activation in early sepsis Shock 39 5 2013 427 32 23524845 25 Shapiro NI Schuetz P Yano K Sorasaki M Parikh SM Jones AE Trzeciak S Ngo L Aird WC The association of endothelial cell signaling, severity of illness, and organ dysfunction in sepsis Crit Care 14 5 2010 R182 20942957
PMC005xxxxxx/PMC5122317.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 0370623 1028 Biochemistry Biochemistry Biochemistry 0006-2960 1520-4995 27304983 5122317 10.1021/acs.biochem.6b00258 NIHMS829334 Article Tolerance of a knotted near infrared fluorescent protein to random circular permutation Pandey Naresh 13 Kuypers Brianna E. 24 Nassif Barbara 1 Thomas Emily E. 13 Alnahhas Razan N. 13 Segatori Laura 145 Silberg Jonathan J. 15* 1 Department of Biosciences, Rice University, Houston, Texas 77005, United States 2 Systems, Synthetic, and Physical Biology Graduate Program, Rice University, Houston, Texas 77005, United States 3 Biochemistry and Cell Biology Graduate Program, Rice University, Houston, Texas 77005, United States 4 Department of Chemical and Biomolecular Engineering, Rice University, Houston, Texas 77005, United States 5 Department of Bioengineering, Rice University, Houston, Texas 77005, United States * To whom correspondence should be addressed: Jonathan J. Silberg, Biosciences Department, 6100 Main Street, Houston TX 77005; Tel: 713-348-3849; [email protected] 12 11 2016 29 6 2016 12 7 2016 24 11 2016 55 27 37633773 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Bacteriophytochrome photoreceptors (BphP) are knotted proteins that have been developed as near-infrared fluorescent protein (iRFP) reporters of gene expression. To explore how rearrangements in the peptides that interlace into the knot within the BphP photosensory core affect folding, we subjected iRFP to random circular permutation using an improved transposase mutagenesis strategy and screened for variants that fluoresce. We identified twenty seven circularly permuted iRFP that display biliverdin-dependent fluorescence in Escherichia coli. The variants with the brightest whole cell fluorescence initiated translation at residues near the domain linker and knot tails, although fluorescent variants were discovered that initiated translation within the PAS and GAF domains. Circularly permuted iRFP retained sufficient cofactor affinity to fluoresce in tissue culture without the addition of biliverdin, and one variant displayed enhanced fluorescence when expressed in bacteria and tissue culture. This variant displayed a similar quantum yield as iRFP, but exhibited increased resistance to chemical denaturation, suggesting that the observed signal increase arose from more efficient protein maturation. These results show how the contact order of a knotted BphP can be altered without disrupting chromophore binding and fluorescence, an important step towards the creation of near-infrared biosensors with expanded chemical-sensing functions for in vivo imaging. biliverdin circular permutation knot near-infrared fluorescent protein phytochrome protein engineering INTRODUCTION The first description of a knotted protein was controversial because of the entropic penalty that is associated with interlacing a polypeptide chain.1,2 Since that time, approximately one thousand knotted structures have been deposited in the protein data bank,3 representing approximately 1% of the total structures deposited, suggesting that this structural feature is present in diverse macromolecules that support cellular reactions. A recent experimental study provided evidence that small knotted proteins (≤160 residues) spontaneously fold within the context of a cell free translation system without forming misfolded species,4 although knot formation in these proteins was found to occur post-translationally and was rate limiting for folding. At times, knots may also provide extra mechanical stability to proteins.5 However, the lower free energy relative to the unfolding transition is thought to come at the expense of folding rates, since the introduction of a knot into a polypeptide decreased the folding rate by over an order of magnitude.6 While this protein design study demonstrated that knotted proteins can arise from an unknotted topology by changing protein contact order (CO), our understanding of knotted protein tolerance to other sequence changes that alter CO remains limited. In nature, changes in CO arise as genes experience duplication, fusion, and fission.7 Support for this idea has come from the discovery of circularly permuted proteins that share the same topology but differ in the location of their NH2 and COOH termini.8 Circular permutation has also been used in the lab for protein design,9 where it has yielded proteins with improved activity,10,11 altered ligand binding,12 and enhanced stability.13 Additionally, circularly permuted variants of natural proteins have been leveraged to construct biosensors and molecular switches through domain insertion,14,15 as well as artificial zymogens.16,17 An artificial zymogen has been successfully designed using a knotted protein. This zymogen was built by connecting the termini of a knotted fluorescent protein using a small protease-cleavable linker that inhibits folding and cofactor-binding that are both required for maturation into a protein that displays fluoescence.17 When exposed to a caspase that cleaves the linker, the resulting protein fragments matured into a fluorescent protein, consistent with that observed in fragmentation studies.18 While this study showed how to inactivate a knotted protein using circular permutation,17 there have been no systematic studies examining the structural tolerance of a knotted protein to random circular permutation. Bacteriophytochrome photoreceptors containing a knot are well suited for studies of permutation tolerance because structural conservation can be easily measured in mutants by analyzing fluorescence. BphP use light-dependent conformational changes in their knotted photosensory core to alter the activity of their signaling domains.19 Structural studies have shown that this photosensory core is made up of a pair of domains, the Per/ARNT/Sim (PAS) and cGMP phosphodiesterase/adenylylcyclase/FhlA (GAF) domains (Figure 1a), which interlace into the knot and bind linear tetrapyrrole cofactors such as biliverdin.20,21 When this cofactor experiences photochemical isomerization, it interconverts the PAS-GAF photosensory core between two different conformational states that differ in their activation of adjacent signaling domains.22 Because the PAS-GAF core interacts with far red light that is well suited for deep tissue imaging, photosensory cores have been mutated and screened to create variants with enhanced near-infrared fluorescence, such as IFP and iRFP.23–25 However, these and other BphP have not been subjected to random circular permutation, and it remains unclear which of the possible circularly permuted BphP are best suited for constructing near-infrared biosensors or light-dependent switches using domain insertion. EXPERIMENTAL PROCEDURES Permuteposon design Permuteposon P4 was built by mutating a previously described permuteposon to introduce a RBS into the R2R1 transposase recognition sequence.26 In addition, a tac promoter containing a lacI operator was inserted after the kanamycin-resistance cassette (KanR) but before the R2R1 sequence. In contrast to the previously described permuteposon, which constitutively expresses permuted proteins with an 18 amino acid peptide scar at their NH2-terminus,26 P4 allows for inducible expression of permuted proteins and only adds a two amino acid scar to protein termini. Library construction DNA encoding the iRFP gene was PCR amplified from pBAD/His-B-iRFP24 using primers that introduce NotI restriction sites and 37 base pairs between the first and last codon of iRFP and the adjacent NotI site. These extra base pairs and the NotI restriction sites will ultimately encode a peptide linker having fifteen residues (GGSGGSAAAGGSGGS) that links the original NH2 and COOH-termini of iRFP in each permuted variant. This long linker was used in our design because the iRFP residues that become connected through permutation are separated by ~35Å.20 The product of this reaction was digested with NotI, size selected using agarose gel electrophoresis, purified using Zymoclean Gel DNA Recovery Kit (Zymo Research), and circularized by incubating with T4 DNA ligase for 16 hours at 16°C. The P4 permutepsoon (100 ng) was inserted into the circularized iRFP (60 ng) by incubating these DNA in a 20 µL reaction containing 1 µL (0.22 µg/µL) of HyperMu MuA transposase (ThermoFisher Scientific) and buffer for 16 hours at 37°C. After terminating the reaction at 70°C for 10 min, DNA was purified using DNA Clean & Concentrator (Zymo Research), electroporated into MegaX DH10B electrocompetent E. coli (Invitrogen), and grown on Luria Broth (LB)-agar plates containing 25 µg/mL kanamycin at 37°C. Lawns of cells (>50,000 cfu) were obtained after overnight growth, cells were harvested by scraping, and total plasmid DNA was purified using a Qiagen DNA Miniprep kit. The ensemble of DNA was linearized using NotI, and iRFP-transposon hybrids (2.9 kb) were separated from larger DNA, which are thought to represent iRFP containing multiple inserted permuteposons. The iRFP-transposon hybrids were circularized by ligation for 16 hours at 16°C to create the final library. Screening E. coli XL1 Blue transformed with the final library were grown on LB-agar plates containing 50 µg/mL kanamycin at 37°C. Single colonies from these plates were used to inoculate 200 µL cultures in 96-well deep well plates containing 25 µg/mL kanamycin. Deep well plates were grown at 37°C while shaking at 250 rpm. After 18 hours of growth, stationary phase cultures were diluted by adding 600 µL LB containing 25 µg/mL kanamycin. After growing for 1 hour at 37°C, 200 µL LB containing isopropyl β-D-1-thiogalactopyranoside (IPTG) and BV (Frontier Scientific) were added to final concentrations of 0.5 mM and 80 µM, respectively. Deep well plates were grown for additional 4 hours in the dark at either 23 or 37°C. A fraction (200 µL) of each culture was transferred to clear polystyrene 96-well V-bottom plates (Rainin), cells were pelleted by centrifuging at 3,000×g for 5 minutes, supernatant containing excess BV was removed, and whole cell absorbance (600 nm) and fluorescence (λex = 690 nm; λem = 700–725 nm) was measured using a Tecan M1000 plate reader. Measurements were performed using 5 nm bandwidths in top mode with 1 nm step size. For each variant, the emission intensities measured from 705 to 720 nm were averaged, and this average value was normalized to absorbance in each well to obtain the relative near-infrared fluorescence signals of each variant. As negative controls, each plate contained four cultures of cells transformed with a vector lacking the iRFP gene. Cultures that yielded absorbance-normalized fluorescence signals that were ≥5σ higher than the signal obtained from cells lacking the iRFP gene were sequenced (n = 81) and given names that correspond to the iRFP residue encoded by the first codon in the permuted open reading frame. Vector construction All of the permuted iRFP were PCR amplified and cloned into pBAD/His-B (Invitrogen) vector using standard cloning to create plasmids that use the araBAD promoter to express each permuted iRFP with a NH2-terminal (His)6 tag. The (His)6 tag was fused to each variant through a 27 residue polypeptide having the sequence GMASMTGGQQMGRDLYDDDDKDPSSRS. To create mammalian vectors that express permuted iRFP, select variants were PCR amplified using primers that introduce BglII and SalI restriction sites and an HA tag to C-terminus. PCR amplified variants were cloned into pEGFP-C1 vector (Clontech) using standard cloning. All vectors were sequence verified. Whole cell fluorescence measurements Circularly permuted iRFP yielding significant fluorescence over background were transformed into E. coli XL1 Blue using heat shock, and grown on LB-agar plates containing 50 µg/mL kanamycin at 37°C. After overnight growth, three or more colonies derived from each vector were used to inoculate 3 mL LB cultures in 15 mL Falcon tubes containing 50 µg/mL kanamycin. After 16 hour growth at 37°C and 250 rpm, 1 mL of cells were harvested by centrifugation, resuspended in 1 mL LB medium, and used to inoculate a fresh 4 mL LB culture containing 0.5 mM IPTG, 80 µM BV, and 50 µg/mL kanamycin. Cells were grown for 5 hours at the indicated temperatures while shaking at 250 rpm in the dark, washed with 25% glycerol (1 mL), and resuspended in 25% glycerol (1 mL) to a similar density. Aliquots (200 µL) of each resuspended sample were transferred to four wells in transparent flat-bottom 96-well plates (Corning), and whole cell absorbance (600 nm) and fluorescence (λex = 690 nm, λem = 700–800 nm) was acquired using a Tecan M1000 plate reader. Emission data was normalized to absorbance in each well, and data reported represent the average of measurements performed on samples derived from three or more colonies. A vector that expresses full-length iRFP (pBAD/His-B-iRFP) and circularized P4 lacking the iRFP gene were used as positive and negative controls, respectively. BV dependence Arabinose-inducible vectors encoding circular permuted iRFP were transformed into E. coli XL1 Blue, and individual colonies were used to inoculate LB cultures containing 100 µg/mL ampicillin. After 16 hour incubation at 37°C and 250 rpm, cells (2 mL) were harvested by centrifugation and used to inoculate fresh 8 mL LB cultures containing 1 mM arabinose, 100 µg/mL ampicillin, and varying final concentrations of BV (0, 5, 10, 20, 40, and 80 µM). Aliquots of these cultures (1 mL) were added into 96-well deep well plates and incubated for 5 hours at 37°C while shaking at 250 rpm in the dark. Fractions (200 µL) of each culture were transferred to three wells within clear polystyrene 96-well v bottom plates (Rainin), cells were pelleted by centrifuging plates at 3,000 × g for 5 minutes, supernatant containing excess BV was removed, and whole cell absorbance (600 nm) and fluorescence (λex = 690 nm; λem = 700–800 nm) was measured using a Tecan M1000 plate reader. Control experiments analyzing the signal from cells expressing iRFP with varying levels of BV yielded a lower coefficient of variance when fluorescence was measured in cells pelleted in V-bottom plates compared with cells resuspended in liquid medium. Translation initiation calculations The effect of circular permutation on protein expression from the P4 permuteposon was analyzed using a thermodynamic model for translation initiation.27 The rate of translation initiation at the start codon within P4 was calculated by using sequence that encompasses 47 base pairs before the intended start codon and 96 base pairs following that codon. Western immunoblots E. coli XL1 Blue transformed with pBAD-derived vectors were grown as described for whole cell fluorescence measurements, with the exception that growth medium contained 100 µg/mL ampicillin and 1 mM arabinose. After resuspending cultures to identical optical densities, cells from each culture (10 µL) were analyzed using sodium dodecyl sulfate (SDS)-PAGE under reducing conditions using NuPAGE 12% Bis-Tris Gels (Life Technologies) and MOPS running buffer. Total protein was transferred to a Protran nitrocellulose membrane (Whatman) using a TE 22 Mini Tank Transfer Unit (GE Healthcare) at 200 mA for 45 minutes. Membranes were washed with 20 mL TBST buffer (100 mM Tris pH 7.5, 150 mM NaCl, and 0.1% Tween 20) for 5 minutes, and blocked for 1 hour with 30 mL 10% dry milk in TBST. Membranes were incubated for 1 hour with a His-tag polyclonal antibody (Qiagen) diluted 1:1000 in 20 mL TBST, incubated for 1 hour in 20 mL TBST containing a secondary goat anti-rabbit IgG conjugated to peroxidase conjugate (Calbiochem) at a dilution of 1:1000. A visual signal was generated using the ECL western blotting substrate (GE Healthcare) and imaged using autoradiographic film (BioExcell). HeLa cells transiently transfected with different plasmids were collected by trypsinization and lysed in Complete lysis-M buffer containing a protease inhibitor cocktail (Roche Applied Science) for 8 minutes with intermittent vortexing. Lysed cells were then centrifuged for 5 min at 15,000 rpm at 4°C, and the supernatant was collected. Protein concentrations were determined by Bradford assay. Total protein (5 µg) from each sample was separated by 12% SDS-PAGE and transferred to PVDF membrane (Bio-Rad). Membranes were blocked using 5% dry milk in TBST overnight at 4°C. Primary antibody incubations were performed for 3 hours at 23°C using rabbit anti-HA (Santa Cruz Biotechnology, 1:8,000), chicken anti-GFP (AnaSpec; 1:8,000), and mouse anti-α-Tubulin (Sigma Aldrich, 1:10,000) diluted in 1% milk in TBST. Secondary antibody incubations were conducted for 1 hour at room-temperature using horseradish peroxidase-conjugated goat anti-rabbit, goat anti-chicken and goat anti-mouse (Santa Cruz Biotechnology; 1:12,000). Chemiluminescent visualization was done using Luminata Forte Western HRP substrate (Millipore) on an LAS4000 imager (GE Healthcare). Quantification of all protein bands was performed using ImageJ. Protein purification E. coli JW2509-2 transformed with a plasmid that constitutively expresses Synechocystis PCC 6803 hemeoxygenase (pSR34-Bvd) and arabinose-inducible vectors expressing different iRFP variants were grown on LB-agar plates containing 100 µg/mL streptomycin and 100 µg/mL ampicillin at 37°C. Single colonies were used to inoculate fresh LB cultures (50 mL) containing 100 µg/mL streptomycin and 100 µg/mL ampicillin. After overnight growth, 1 mL culture was used to inoculate fresh 50 ml LB containing 100 µg/mL streptomycin and 100 µg/mL ampicillin. After 3.5 hours of growth at 37°C while shaking at 250 rpm, cells (10 mL) were harvested by centrifugation and used to inoculate 1 L LB containing the same antibiotics. When these cultures reached an OD ~0.5, arabinose was added to a final concentration of 1 mM to induce protein production. Cells were grown for 19 hours at 37°C while shaking at 250 rpm in dark. Cells harvested by centrifugation (4,000×g for 10 minutes at 4°C) were resuspended in lysis buffer (50 mM phosphate buffer pH 7.0, 300 mM NaCl, 10 mM Imidazole, 1 mM MgCl2, 0.5 mg/mL lysozyme, and 0.04 mg/mL DNAse I) and frozen at −80°C. Cells were thawed, centrifuged at 45,000×g for 1 hour at 4°C, the soluble fraction was loaded onto a Ni-NTA column (Qiagen), and washed with NTA wash buffer (50 mM phosphate buffer pH 7.0, 300 mM NaCl, and 10 mM Imidazole). Permuted iRFP were eluted using NTA elution buffer (50 mM phosphate buffer pH 7.0, 300 mM NaCl, and 250 mM Imidazole). Fractions were analyzed using SDS-PAGE, and those containing permuted iRFP were pooled. Ammonium sulfate was added to a final concentration of 40%, protein was loaded onto a HiTrap Phenyl HP (GE Healthcare) that had been equilibrated in a similar buffer using an AKTA protein purification system (GE Healthcare), and permuted iRFP were eluted using a linear ammonium sulfate gradient from 40 to 0% in 20 mM Tris buffer. Protein appearing homogeneous by SDS-PAGE was pooled, dialyzed against phosphate buffered saline (PBS) pH 7.5 (137 mM NaCl, 2.7 mM KCl, 4.3mM Na2HPO4, and 1.47 mM KH2PO4), and stored at −80°C. All purification steps were performed under green or no light conditions. Gel filtration chromatography A Superdex 200 10/300 GL (GE Healthcare) run using PBS and a flow rate of 0.5 ml/min at 4°C was used to assess the size of each permuted iRFP. A standard curve generated using the elution volumes of five standards (β-amylase, alcohol dehydrogenase, bovine serum albumin, carbonic anhydrase, and cytochrome C) was used to estimate the weight of each protein. Native gel electrophoresis Purified proteins (5 µg) were loaded on native gel and run using a non-denaturing running buffer (25 mM Tris, 192 mM glycine) with 30 mA for 45 minutes. The native gel was composed of a separating gel (10% acrylamide pH 8.8) and a stacking gel (3.9% acrylamide pH 6.8). Quantum yield determination Spectral measurements used purified iRFP variants diluted in PBS to a maximum absorbance <0.1 and Nile blue dye in ethanol with an absorbance of 0.1. For each sample, absorbance spectra (205 to 750 nm) were collected using a Cary 50 UV-visible spectrophotometer. Fluorescence emission (669 to 800 nm) arising from excitation at 659 nm was acquired using a Tecan M1000 plate reader. The quantum yield was calculated using the equation [ϕF(x)=(As/Ax)(Fx/Fs)(nx/ns)2·ϕF(S)], where ϕF is fluorescence quantum yield of the standard (s) and unknown (x), A is absorbance at the excitation wavelength for each sample, F is area of the emission spectrum for each sample, and n is refractive index of the solvents.28 Extinction coefficients were determined by calculating the ratio of absorbance at 694 nm to 391 nm and multiplying this value with 39,900, which represents the extinction coefficient of the free BV.22 Equilibrium unfolding A 6 M stock of GdnHCl was mixed with PBS to make solutions containing a range of GdnHCl concentrations (0, 0.5, 1, 1.5, 2, 2.25, 2.5, 2.75, 3, 3.25, 3.5, 4 and 4.5 M). Aliquots of each protein were mixed with 325 µL of each GndHCl solution, incubated at room temperature for 2 hours in dark, and transferred into transparent flat-bottom 96-well plate (Corning) where fluorescence (λex = 690 nm and λem = 700 to 800 nm) was measured. The fluorescence signal at 715 nm was used for comparison. Mammalian tissue culture and flow cytometry HeLa cells (American Type Culture Collection) were cultured in DMEM (Dulbecco modified eagle medium), Lonza, supplemented with 10% fetal bovine serum (Sigma) and 1% penicillin–streptomycin–glutamine (Hyclone) and maintained at 37 °C and 5% CO2. Cells were plated in 12-well plates at 8×104 cells per well. 24 hours after seeding, cells were transiently transfected using JetPrime (Polyplus transfection) according to the manufacturer’s protocol with 0.5 ug of plasmid expressing the iRFP variant along with 0.05 ug of pcDNA4-GFP plasmid as transfection control per well. The culture media of transfected cells was replaced with fresh media 16 hours post transfection. Cells were harvested for analysis at 48 hours post transfection by trypsinization (TrypLE, GIBCO Invitrogen). Cells were analyzed with a FACSCanto II flow cytometer (BD, San Jose, CA) to measure fluorescence intensity of iRFP variants in the APC-Cy7 (Allophycocyanin conjugated with cyanine dye) channel (633 nm laser, 780/60 nm emission filter). At least 10,000 cells were recorded in each sample for analysis. For transient transfection experiments, the APC-Cy7 signal changes were monitored within GFP-positive cells (488 nm laser, 530/30 emission filter) to monitor changes within transfected cells. The reported output signal was calculated by normalizing APC-Cy7 intensity by GFP intensity in order to eliminate differences arising from transfection efficiencies. RESULTS Mapping knotted protein tolerance to permutation As a model system for studying BphP tolerance to random circular permutation, we chose iRFP,24 a mutant of the photosensory core from Rhodopseudomonas palustris BphP2 whose native structure displays enhanced near-infrared fluorescence upon biliverdin (BV) binding. iRFP was targeted because it displays enhanced stability and fluorescence compared with the other bacteriophytochromes.24 We hypothesized that these properties would maximize: (i) the reliable detection of folded variants and (ii) the fraction of variants that retain an iRFP-like structure upon permutation. Previous studies have shown that stability increases the fraction of folded variants in libraries encoding proteins subjected to random mutation and fission.29,30 A modified PERMUTE protocol was used to construct a library of vectors that express circularly permuted iRFP.26 With PERMUTE, an artificial transposon containing all of the attributes of a vector (called a permuteposon) is randomly inserted into a second vector harboring the gene of interest, and a multistep procedure is used to convert the product of this reaction into a library. To allow for regulated protein expression in PERMUTE libraries, we synthesized a new transposon, permuteposon P4 (Figure S1a), and we randomly inserted P4 into a circular iRFP gene (Figure 1b). Restriction digestion of the library yielded a DNA pattern consistent with random insertion (Figure S1b). Our vector library was transformed into Escherichia coli, and individual colonies were screened at either 23 or 37°C for near-infrared fluorescence. Two temperatures were used for screening to increase the likelihood of identifying marginally-stable variants. Screening a total of 1760 colonies identified 38 circularly permuted iRFP (cp-iRFP) with emission >5σ over that observed with the parental strain transformed with a vector lacking an iRFP gene. Assuming that each cp-iRFP had an equal probability of occurring in our library, this level of screening was predicted to sample 58% of the possible cp-iRFP.31 Sequencing vectors from these colonies revealed 27 unique in frame cp-iRFP that retained near-infrared fluorescence with signals ≥5σ of cells lacking an iRFP (Table S1). To visualize how the structures of these permuted proteins relate to one another, the first iRFP residue that is translated in each cp-iRFP was mapped onto the structure of the Deinococcus radiodurans BphP photosensory core (Figure 1c),20 a BphP that displays 39% identity with iRFP.24 This analysis revealed the highest density of variants arising from backbone fission proximal to the domain linker and near the termini of the protein. To determine how this tolerance relates to iRFP folding, we also evaluated the contact order of fluorescent cp-iRFP. Previous studies have shown that CO is inversely correlated with protein folding rates,32 suggesting that increases in CO may add an additional burden to achieving a knotted protein topology. Fluorescent variants had CO values that are both higher and lower than native iRFP. However, the highest density of permuted variants had new termini that clustered in a region with CO values that are lower than iRFP. Comparison of circularly permuted iRFP in bacteria Distinct sets of cp-iRFP were discovered at the different temperatures where screening was performed (23 and 37°C). This finding suggested that some cp-iRFP displayed temperature-dependent fluorescence. To test this hypothesis, we compared the emission and excitation spectra of each cp-iRFP with those of native iRFP. Initial spectroscopic measurements at 23°C revealed that all of the cp-iRFP displayed similar peak excitation and emission as native iRFP, albeit with a range of peak intensities (Table S1). At 23°C, five cp-iRFP yielded whole cell emission that were greater than that observed with iRFP, three cp-iRFP yielded intensities that were comparable to that of iRFP, and the remaining cp-iRFP showed emission significantly lower than that of iRFP (Figure 2a). When measurements were performed at 37°C, the majority of the cp-iRFP displayed a signal that was lower than iRFP; only three variants yielded a signal that matched or exceeded iRFP. The iRFP fluorescence signal increased ~2.5-fold as temperature was increased from 23 to 37°C (Figure 2b). To determine whether any of our cp-iRFP display enhanced signals as temperature increases, we calculated the ratio of fluorescence emission for each variant expressed in cells grown at these different temperatures. This analysis revealed that 52% of the cp-iRFP displayed higher emission intensity at 37°C compared with 23°C like iRFP (ratio >1), 37% had similar fluorescence ratios as iRFP (ratio ≥ 2.5), and 11% displayed decreased emission as temperature was increased (ratio <1). Among the cp-iRFP with ratios ≤1, 80% were identified when screening was performed at 23°C. The variability in fluorescence intensity across our different cp-iRFP could have arisen because permutation differentially affected protein expression, folding, cofactor binding, stability, degradation, or quantum yield. In a previous study, we found that backbone fission increases the concentration of BV required for fluorescence in IFP, a homolog of iRFP.18 Because permutation involves backbone fission, this finding suggested that permutation could have altered permuted protein fluorescence by decreasing BV affinity to levels that decreased the fraction of cofactor bound under the conditions of the measurements. To test this hypothesis, we evaluated whether near-infrared fluorescence varied with the amount of BV added to the cells that express each cp-iRFP (Figure S2). Under the conditions of the fluorescence measurements conducted in this study, a majority of the unmodified iRFP was expected to have BV bound, since the BV concentration required for half-maximal fluorescence in cells expressing the unmodified iRFP (11 µM) was ~7-fold lower than the concentration used in our screening protocol. A majority of our cp-iRFP (85%) required a similar (or lower) amount of BV for half-maximal fluorescence, 4 to 15 µM. The remaining variants, which arose from backbone fission in the GAF domain, required between 18 and 25 µM BV for half-maximal fluorescence. These values are lower than the concentration (80 µM) used in our whole cell analysis of fluorescence that yielded signal differences. These findings suggest that the large variability in whole cell fluorescence among our variants, which varied by as much as 98-fold, was not due to differences in the BV-binding affinity and attachment. To determine if the cp-iRFP arising from backbone fission within the N-terminal tail (e.g., cp-iRFP-11 and cp-iRFP-12) require the residues from this tail for maturation into a fluorescent protein, we created a truncated iRFP that encodes residues 12–316 but lacks residues 1–11. This truncated iRFP displayed similar fluorescence as cp-iRFP-12 (data not shown), suggesting that the initial residues in iRFP are not required for chromophore binding or stability. Translation initiation variability Previous studies have shown that the genetic context of a ribosomal binding site (RBS) can alter the strength of translation initiation in bacteria by more than an order of magnitude,27 and vectors in our library express cp-iRFP with gene sequences that change the context of the RBS in our permuteposon. To examine whether permuted gene sequences influence protein expression from this RBS, we used a thermodynamic model for translation initiation to estimate how the strength of our RBS site varies in vectors expressing different cp-iRFP.27 This model uses the sequence and the genetic context of an RBS to quantify the strength of the mRNA-ribosome interaction that is required for translation initiation and competing off-pathway RNA folding reactions that inhibit formation of a productive mRNA-ribosome complex. Experimental validation of this model through forward engineering of RBS sites revealed that the predictions are accurate within a factor of 2.3-fold over a range of translation initiative values spanning five orders of magnitude.27 We found that the calculated translation initiation rates for all possible cp-iRFP varied by three orders of magnitude when expressed using the RBS in permuteposon P4 (Figure 3a). To test whether the fluorescent signals obtained with our original clones correlate with translation initiation rates, we compared the calculated rates for each variant to the whole cell fluorescence signals (Figure 3b). This analysis revealed significant correlations between calculated translation rates and fluorescence intensity. These findings suggest that translation initiation variability contributed to the dispersion in signals in whole cell measurements. To minimize differences in protein synthesis that arose from changes in RBS genetic context, we created vectors that expressed each of our cp-iRFP with a NH2-terminal His tag connected through a 27 residue polypeptide. When expressed from these vectors, cp-iRFP with NH2 and COOH-termini proximal to the domain linker in unmodified iRFP displayed the highest whole cell fluorescence (Figure S3). The fluorescence of each variant was also compared in the presence and absence of the affinity tag. A majority of the cp-iRFP (85%) displayed an increased signal in the presence of the His tag, indicating that this expression strategy yielded higher levels of each fluorescent BV-bound variant, consistent with predictions from thermodynamic calculations. To better understand the signal variability observed when proteins are expressed under conditions that yield more consistent calculated translation initiation rates, we analyzed the levels of a series of selected cp-iRFP (Figure S4). Specifically, we focused on the subset of cp-iRFP that displayed enhanced fluorescence at 37°C compared with 23°C. Immunoblot analysis revealed that the levels of these cp-iRFP varied up to 10-fold from iRFP (Figure 4a). When whole cell fluorescence was normalized to protein expression in matched samples, a majority of the cp-iRFP displayed a signal that was similar to that measured using cells expressing native iRFP, with the exception of cp-iRFP-198, which displayed a signal 25-fold lower than iRFP, and cp-iRFP-12 and cp-iRFP-102, which exhibited average signals that were 1.9- and 4.4-fold higher than iRFP, respectively. While the signal from cp-iRFP-12 was significantly higher than iRFP, the signal from cp-iRFP-102 displayed a large error and was not significantly different from iRFP. In vitro analysis of circularly permuted iRFP The variation in protein-normalized fluorescence signals suggested that permutation affected the fraction of cp-iRFP that accumulated in a fluorescence-competent conformation and/or quantum yield. To directly evaluate the effects of permutation on spectroscopic properties, we expressed, purified, and characterized four cp-iRFP. Two circularly permuted iRFP were purified that displayed improved signals (cp-iRFP-12 and cp-iRFP-102), and two were purified that exhibited iRFP-like signals (cp-iRFP-129 and cp-iRFP-133). Proteins were purified using a combination of affinity and reverse phase chromatography. During purification, we found that the reverse phase column yielded two protein populations, one that eluted as a monodisperse peak at ~5% (NH4)2SO4 and one that eluted when the column was washed with buffer lacking (NH4)2SO4. In the case of iRFP, <50% of the total recombinant iRFP eluted at 5% ammonium sulfate, while the remaining iRFP eluted in the absence of ammonium sulfate. All of our biochemical characterization was performed using the protein from the former peak, since it contained holophytochrome as previous observed.25 All of the cp-iRFP had three absorbance peaks with maxima at similar wavelengths (280, 391, and 693 nm) to those observed with iRFP (Figure S5).24 In addition, all four cp-iRFP displayed maximal emission wavelengths and quantum yields that were not significantly different from those obtained with iRFP. These findings suggest that permutation does not disrupt the iRFP structure sufficiently to alter the chromophore environment. We next sought to determine how permutation affects protein structure, since the introduction of NH2- and COOH-termini within the iRFP dimerization interface could disrupt oligomerization. cp-iRFP-129 and cp-iRFP-133 have termini that map onto the dimer interface in the BphP crystal structure (Figure 5a), while cp-iRFP-12 and cp-iRFP-102 have termini that map to sites distal from the interface. To test whether these cp-iRFP have structures that differ from iRFP, we compared their electrophoretic mobility using native PAGE (Figure 5b). This analysis revealed that cp-iRFP-102, cp-iRFP-129 and cp-iRFP-133 migrated farther than iRFP. Two of these variants have termini that map onto the dimer interface. To test whether the mobilities of these cp-iRFP were altered because oligomerization was disrupted, analytical gel filtration was used to evaluate their oligomeric state (Figure 5c). As previously observed, iRFP eluted as a monodisperse peak with an apparent MW (72,443 Da) that was 1.92-fold greater than the calculated MW of the iRFP polypeptide. All of the cp-iRFP eluted at volumes consistent with oligomerization. However, the cp-iRFP-129 displayed an apparent molecular weight that was greater than iRFP (91,201 Da). This cp-iRFP also displayed the largest change in electrophoretic mobility with native polyacrylamide gel electrophoresis (PAGE). Taken together, these findings suggest that circular permutation affects the radius of gyration but does not disrupt oligomerization. Since permutation has been shown to alter the stability of other proteins,13 we compared the equilibrium unfolding of cp-iRFP with iRFP, which was previous shown to be irreversible.33 Unfolding was monitored by measuring fluorescence as a function of the chemical denaturant guanidine (GdnHCl). With iRFP, GdnHCl-induced equilibrium unfolding yielded a single cooperative transition with a midpoint at 2.85 M guanidine. All of the variants were destabilized compared with iRFP with the exception of cp-iRFP-12 (Figure 5d), which displayed enhanced stabilization. Fluorescence in mammalian cells The rate of protein folding varies between prokaryotes and eukaryotes, with bacterial proteins folding post-translationally due to faster translation rates and eukaryotic proteins folding cotranslationally.34,35 To evaluate whether the nature of the expression system (i.e., prokaryotic or eukaryotic) affects the folding and cofactor binding of cp-iRFP, which are both required for maturation into a fluorescent protein, we assessed the fluorescence signal of seven variants in mammalian cells. Five of these cp-iRFP (11, 12, 101, 102, and 129) were selected because they display the highest fluorescence signals among the iRFP variants discovered in E. coli. The remaining two variants initiate translation within the domain linker (cp-iRFP-133) and within the GAF domain (cp-iRFP-198). The signals from these permuted iRFP were evaluated using HeLa cells that had been transiently transfected with a plasmid encoding each HA-tagged cp-iRFP (or unmodifiied iRFP) and a plasmid constitutively expressing green fluorescent protein (GFP) as a transfection control. Cellular fluorescence was measured using flow cytometry 48-hours post-transfection (Figure S6a). Relative cp-iRFP fluorescence intensities were calculated by normalizing the absolute iRFP fluorescence by the absolute GFP fluorescence to account for differences due to transfection efficiency (Figure 4b). Comparison of the relative near-infrared fluorescence values revealed that cp-iRFP-12 displayed near-infrared fluorescence that was 18% higher than native iRFP. Additionally, one variant displayed a similar intensity as iRFP (cp-iRFP-11), four cp-iRFP showed fluorescence approximately 3-fold lower than iRFP (101, 102, 129 and 133) and one variant presented near undetectable fluorescence (cp-iRFP-198). To compare the expression levels of cp-iRFP, we evaluated the levels of HA-tagged native and permuted iRFP using Western immunoblots. Experiments were performed using total protein from HeLa cells transfected as described for cytometry analyses (Figure S6b). The intensity of all of the protein bands detected using an HA-specific antibody were normalized to the intensity of an α-tubulin loading control, since our goal was to normalize the cellular fluorescence signal by the total protein expressed. Immunoblot detection of GFP was used to confirm transfection. The relative iRFP fluorescence values obtained by cytometry were normalized to the relative intensity of HA bands, which had been normalized to the loading control (Figure S6c). cp-iRFP-101, cp-iRFP-102 and cp-iRFP-133 presented modest increases (22 to 50%) in protein-normalized fluorescence compared to iRFP, while cp-iRFP-11 and cp-iRFP-12 exhibited modest (8 to 31%) decreases in signal. cp-iRFP-129 presented only 28% of the signal obtained with iRFP, and variant 198 displayed fluorescence that was ~2% of the iRFP signal. While a subset of the variants displayed a similar protein-normalized signal in tissue culture and bacterial cells, several variants displayed weaker signals in tissue culture, whose underlying causes are not known. Biophysical implications The results from this study show that knotted BphP are tolerant to a wide range of sequence rearrangements that alter contact order. High-throughput screening revealed that 9% of all possible cp-iRFP display detectable near-infrared fluorescence. Since our screening was predicted to sample only 58% of all possible cp-iRFP, this number likely represents a lower bound on the fraction of permuted iRFP that fold and fluoresce. Our findings also suggest that many of these cp-iRFP display BphP-like structures, because retention of fluorescence requires a bound cofactor within a similar structural environment as iRFP. However, it is not clear which of these cp-iRFP acquire a knot. Our in vitro biochemical studies revealed parent-like stability and spectral properties, suggesting that the BV chromophore environment is not disrupted by permutation, while native gel analysis suggested that some variants may display subtle structural differences from iRFP. Some members of the BphP family lack protein knots, suggesting that our cp-iRFP might not require a knot for fluorescence. Studies of phytochromes having sensory modules that are made up of tandem GAF domains have revealed structures that lack the knot found in PAS-GAF modules.36 Sequence comparisons have also shown that the cyanobacteriochromes AnPixJ and TePixJ lack the loop that serves as the knot lasso,37,38 although the structures of these proteins have not yet been reported. In addition, a recent study revealed that the BphP knot can be removed through truncation without disrupting cofactor binding, provided that a new cysteine residue is incorporated into the GAF domain to covalently attach the cofactor.39 Structural studies are needed to determine how the topology of our cp-iRFP relate to iRFP, and if some of our these variants must thread a peptide through the loop that is larger than the peptide in iRFP to mature into folded and fluorescent proteins. All of our cp-iRFP displayed excitation and emission maxima that are similar to iRFP. However, at each temperature, we observed a large variation in fluorescence intensities even though each variant was transcribed and translated using the same promoter and RBS. Circular permutation has been shown to alter protein folding,40 ligand binding,12 quantum yield,41 and expression.42 Our results suggest that variability in fluorescence signals do not arise from changes in BV binding affinity. For each variant, the BV concentration required for half-maximal fluorescence was uniformly lower than the concentration used in our assays. Variability in protein expression, on the other hand, is thought to make larger contributions to variability in cellular fluorescence. Thermodynamic calculations analyzing the effects of different cp-iRFP genes on the RBS in our permuteposon revealed a correlation between fluorescence and calculated translation initiation rates. In addition, expression of each cp-iRFP with the same peptide tag at the NH2 terminus, which is thought to improve the consistency of translation initiation across the different variants,27 differentially altered the relative signals obtained with each cp-iRFP from variants expressed without tags. These observations emphasize the importance of considering protein levels when using whole cell assays to screen for circularly permuted proteins with improved fluorescence. Among the fluorescent variants discovered in our screen, cp-iRFP-12 displayed significantly higher protein-normalized fluorescence compared with iRFP in bacteria. Biochemical characterization of this variant did not reveal any differences in spectral properties compared to the parental iRFP, implicating improved maturation as the underlying cause of the enhanced signal. Our biochemical results also suggest that only a fraction of the iRFP translated in E. coli matures into a fluorescent protein. When we subjected recombinant iRFP to reverse phase chromatography, it eluted at two different salt concentrations, which differed in their fluorescent signals as previously observed.25 cp-iRFP-12 also eluted at two different salt concentrations, although the results from whole cell fluorescence measurements suggest that a greater fraction of this variant fully matures into a fluorescent protein. Mapping cp-iRFP-12 onto the D. radiodurans BphP structure suggests a possible mechanism for enhanced maturation in E. coli.20 Because this permuted protein arises from fission within the NH2-terminal tail, near the BV attachment site at residue fifteen, it has fewer residues to thread through the loop that serves as the iRFP knot lasso. The effects of permutation on protein maturation could be evaluated in future studies by analyzing protein folding and fluorescence within prokaryotic and eukaryotic cell free transcription and translation systems. A recent study demonstrated that this approach can be used to study cofactor binding to a protein using the heme-binding myoglobin.43 A comparison of cofactor incorporation into myoglobin orthologs with varying stability revealed correlation between the levels of holo-myoglobin and the protein stability in the absence of cofactor.43 This trend is consistent with our finding that cp-iRFP-12 displays the highest stability among the proteins characterized using in vitro studies, although our analysis was performed with a holophytochrome rather than an apoprotein. Applications of circularly permuted iRFP The permuted proteins discovered have implications for protein design. Circularly permuted proteins are increasingly being used to construct protein switches through domain insertion, where an ensemble of permuted protein variants are randomly inserted at different locations within the backbone of a second protein to create new proteins.44 Domain insertion studies have shown that a subset of the variants created in such libraries display switching behavior, where ligand binding within one domain modulates the structure and function of the second domain.14,15 In fact, high signal-to-noise switches that couple the ligand-binding to the fluorescence of different GFP family members have been created by inserting circularly permuted fluorescent proteins into periplasmic binding proteins.45 The cp-iRFP discovered should be useful for creating similar biosensors, albeit with emission in the near-infrared region which is ideally suited for tissue imaging.24 All of the cp-iRFP displayed parent-like BV affinity, which was sufficient to fluoresce in tissue culture without the addition of biliverdin. This finding suggests that biosensors created using these variants will exhibit sufficient BV affinity to fluoresce in mammalian cells without requiring the addition of exogenous BV. The extent to which each of our cp-iRFP will fold, bind BV, and fluoresce upon insertion into other proteins will require further studies. If knot formation is required for cp-iRFP fluorescence, then it is possible that some cp-iRFP could experience folding challenges upon insertion into other domains. By using a circular iRFP gene for library construction, rather than a vector-encoded iRFP, we were able to create a library that encodes different cp-iRFP in a single step. This modified approach is quicker than the original protocol described for PERMUTE because it avoids manipulations required after the initial MuA reaction.26 The modified protocol described herein will simplify future studies that seek to create vector libraries that express circularly permuted variants of other proteins. Our results also showed that PERMUTE can be used with a permuteposon that contains a hybrid transposase and ribosomal binding site that only amends two extra amino acids to protein termini.46 The simplified PERMUTE protocol should also be compatible with other permuteposons that have been developed for PERMUTE, including those that amend peptides of varying size to the NH2-termini of permuted proteins.26 Beyond simplifying the construction of permuted protein libraries, this approach has the advantage that it creates permuted proteins with well-defined sequence diversity and avoids random deletions that arise with other methods.47 Applications of deep sequencing to PERMUTE libraries will be useful in future studies to improve our understanding of library diversity and sampling of individual variants, similar to that described for libraries created using random mutation.48 Supplementary Material Supplemental This work was supported by the National Science Foundation [MCB-1150138] (to J.J.S.), the Robert A Welch Foundation [C-1614 and C-1824] (to J.J.S. and L.S.), the John S. Dunn Collaborative Research Award (to J.J.S.), and National Science Foundation (CBET-1336053 and CBET-1254318) (to L.S.). Figure 1 iRFP tolerance to circular permutation. (A) In iRFP, a knot is formed when the NH2 terminus of the PAS domain passes through a loop created by the GAF domain. (B) Random insertion of a permuteposon into a circular iRFP gene using the transposase MuA yields a library of cp-iRFP vectors. (C) Relationship between domain structure, contact order, and the location of the NH2 termini created by permutation. The fission sites in fluorescent cp-iRFP are mapped (with lines) onto BphP (PBD = 1ZTU) to relate their location to the NH2 tail (cyan), PAS domain (blue), domain linker (grey), loop from the GAF domain (black) and remainder of the GAF domain (orange). Each line represents the first residue in the cp-iRFP. Figure 2 Effect of temperature on fluorescence. (A) The fluorescence of cells expressing each cp-iRFP was measured at 23°C and 37°C, normalized to cell density, and compared to the fluorescence of cells expressing parental iRFP and lacking iRFP (−ctrl). All values are reported as the fraction of the signal obtained with the variant that yielded the highest signal at each temperature. All cp-iRFP except variants 59, 182, 198 and 291 had significantly higher signal than background at 37°C, and all variants except 59, 122, 129, 182 and 198 had signals that were significantly higher than background 23°C (two-tailed t test; p<0.05). (B) The ratio of the fluorescence signals acquired at 37°C to the signals observed at 23°C. Blue and red symbols represent the variants identified in screens at 23°C and 37°C, respectively. White symbols represent cp-iRFP discovered at both temperatures, and the dashed line represents the value obtained for iRFP. Error bars represent ±1σ calculated using three or more independent measurements. Figure 3 Thermodynamic analysis of protein expression. (A) Calculated translation rates of all cp-iRFP (line) are compared with the calculated rates for fluorescent variants (●). (B) The relative whole cell fluorescent signals of cp-iRFP at 23°C and 37°C are compared with calculated rates. Calculation of the Spearman's rank correlation using our data acquired at 23°C (RSR = 0.403) and 37°C (RSR = 0.384) yielded values that were significant (two-tailed t test, p < 0.05). Figure 4 Protein-normalized whole cell fluorescence. (A) The whole cell fluorescence of His-tagged cp-iRFP at 37°C in E. coli was normalized to cell density, and immunoblot analysis was used to quantify the levels of each cp-iRFP. The ratio of fluorescence to the level of each protein was calculated using data acquired from matched samples. Error bars represent ±1σ calculated using three or more independent measurements. cp-iRFP-12 displayed protein-normalized fluorescence that was significantly higher than iRFP (two tailed t test; p<0.05), while the cp-iRFP-102 signal was not significantly higher. (B) Fluorescence of circularly permuted iRFP in tissue culture. HeLa cells were transiently transfected with vectors that express different cp-iRFP and a vector that expresses GFP, green and red fluorescence was measured using flow cytometry, and the near-infrared fluorescence signal for each cp-iRFP was normalized to the GFP transfection control. Cells were grown in the medium that had not been supplemented with BV. Error bars represent ±1σ calculated from two independent measurements. Variant 12 displayed a GFP-normalized signal that was significantly higher than iRFP (two tailed t test; p<0.05). Figure 5 Effect of circular permutation on iRFP structure. (A) Protein termini created by permutation mapped onto the structure of the D. radiodurans BphP chromophore-binding domain. A comparison of the subunits (red and purple) with the residues at the termini of variants 12 and 102 (green) and 129 and 133 (yellow) reveal that only the latter pair of cp-iRFP create termini proximal to the dimer interface. (B) Native PAGE shows how the migration of unmodified iRFP (N) relates to each cp-iRFP. (C) Gel filtration analysis reveals how the elution profiles of iRFP and the four cp-iRFP compare with the elution of standards (dashed lines) having molecular weights of 12.4, 29, 66, 150, and 200 kDa. (D) The effect of GdnHCl concentrations on cp-iRFP fluorescence (●) is compared with unmodified iRFP (○). The concentration required for the unfolding midpoint is indicated for each variant. For cp-iRFP-12, GdnHCl concentrations >1 M and <3.25 displayed signals that were significantly higher than iRFP (two-tailed t-test; P<0.05). Error bars represent ±1σ calculated using three or more measurements. SUPPORTING INFORMATION AVAILABLE Detailed experimental procedures, supplementary figures, and a supplementary table. This material is available free of charge via the Internet at http://pubs.acs.org. REFERENCES 1 Liang C Mislow K Topological Features of Protein Structures: Knots and Links J. Am. Chem. Soc 1995 117 4201 4213 2 Mansfield ML Are there knots in proteins? Nat. Struct. Biol 1994 1 213 214 7656045 3 Jamroz M Niemyska W Rawdon EJ Stasiak A Millett KC Sułkowski P Sulkowska JI KnotProt: a database of proteins with knots and slipknots Nucleic Acids Res 2015 43 D306 D314 25361973 4 Mallam AL Jackson SE Knot formation in newly translated proteins is spontaneous and accelerated by chaperonins Nat. Chem. Biol 2012 8 147 153 5 Sułkowska JI Sulkowski P Szymczak P Cieplak M Stabilizing effect of knots on proteins Proc Natl Acad Sci USA 2008 105 19714 19719 19064918 6 King NP Jacobitz AW Sawaya MR Goldschmidt L Yeates TO Structure and folding of a designed knotted protein Proc Natl Acad Sci USA 2010 107 20732 20737 21068371 7 Weiner J Bornberg-Bauer E Evolution of circular permutations in multidomain proteins Mol. Biol. Evol 2006 23 734 743 16431849 8 Lindqvist Y Schneider G Circular permutations of natural protein sequences: structural evidence Curr. Opin. Struct. Biol 1997 7 422 427 9204286 9 Peisajovich SG Rockah L Tawfik DS Evolution of new protein topologies through multistep gene rearrangements Nat. Genet 2006 38 168 174 16415885 10 Yu Y Lutz S Improved triglyceride transesterification by circular permuted Candida antarctica lipase B Biotechnol. Bioeng 2010 105 44 50 19609971 11 Daugherty AB Govindarajan S Lutz S Improved biocatalysts from a synthetic circular permutation library of the flavin-dependent oxidoreductase old yellow enzyme J. Am. Chem. Soc 2013 135 14425 14432 23987134 12 Guntas G Kanwar M Ostermeier M Circular permutation in the Ω-loop of TEM-1 β-lactamase results in improved activity and altered substrate specificity PLoS ONE 2012 7 e35998 22536452 13 Whitehead TA Bergeron LM Clark DS Tying up the loose ends: circular permutation decreases the proteolytic susceptibility of recombinant proteins Protein Eng. Des. Sel 2009 22 607 613 19622546 14 Baird GS Zacharias DA Tsien RY Circular permutation and receptor insertion within green fluorescent proteins Proc. Natl. Acad. Sci. U.S.A 1999 96 11241 11246 10500161 15 Guntas G Mansell TJ Kim JR Ostermeier M Directed evolution of protein switches and their application to the creation of ligand-binding proteins Proc. Natl. Acad. Sci. U.S.A 2005 102 11224 11229 16061816 16 Plainkum P Fuchs SM Wiyakrutta S Raines RT Creation of a zymogen Nat. Struct. Biol 2003 10 115 119 12496934 17 To T-L Piggott BJ Makhijani K Yu D Jan YN Shu X Rationally designed fluorogenic protease reporter visualizes spatiotemporal dynamics of apoptosis in vivo Proc Natl Acad Sci USA 2015 112 3338 3343 25733847 18 Pandey N Nobles CL Zechiedrich L Maresso AW Silberg JJ Combining random gene fission and rational gene fusion to discover near-infrared fluorescent protein fragments that report on protein-protein interactions ACS Synth Biol 2015 4 615 624 25265085 19 Rockwell NC Su Y-S Lagarias JC Phytochrome structure and signaling mechanisms Annu Rev Plant Biol 2006 57 837 858 16669784 20 Wagner JR Brunzelle JS Forest KT Vierstra RD A light-sensing knot revealed by the structure of the chromophore-binding domain of phytochrome Nature 2005 438 325 331 16292304 21 Wagner JR Zhang J Brunzelle JS Vierstra RD Forest KT High resolution structure of Deinococcus bacteriophytochrome yields new insights into phytochrome architecture and evolution J. Biol. Chem 2007 282 12298 12309 17322301 22 Quail PH Phytochrome photosensory signalling networks Nat. Rev. Mol. Cell Biol 2002 3 85 93 11836510 23 Shu X Royant A Lin MZ Aguilera TA Lev-Ram V Steinbach PA Tsien RY Mammalian expression of infrared fluorescent proteins engineered from a bacterial phytochrome Science (New York, N.Y.) 2009 324 804 807 24 Filonov GS Piatkevich KD Ting L-M Zhang J Kim K Verkhusha VV Bright and stable near-infrared fluorescent protein for in vivo imaging Nat. Biotechnol 2011 29 757 761 21765402 25 Auldridge ME Satyshur KA Anstrom DM Forest KT Structure-guided engineering enhances a phytochrome-based infrared fluorescent protein Journal of Biological Chemistry 2012 287 7000 7009 22210774 26 Mehta MM Liu S Silberg JJ A transposase strategy for creating libraries of circularly permuted proteins Nucleic Acids Res 2012 40 e71 e71 22319214 27 Salis HM Mirsky EA Voigt CA Automated design of synthetic ribosome binding sites to control protein expression Nat. Biotechnol 2009 27 946 950 19801975 28 Würth C Grabolle M Pauli J Spieles M Resch-Genger U Relative and absolute determination of fluorescence quantum yields of transparent samples Nat Protoc 2013 8 1535 1550 23868072 29 Bloom JD Silberg JJ Wilke CO Drummond DA Adami C Arnold FH Thermodynamic prediction of protein neutrality Proc. Natl. Acad. Sci. U.S.A 2005 102 606 611 15644440 30 Segall-Shapiro TH Nguyen PQ Santos, Dos ED Subedi S Judd J Suh J Silberg JJ Mesophilic and hyperthermophilic adenylate kinases differ in their tolerance to random fragmentation J. Mol. Biol 2011 406 135 148 21145325 31 Bosley AD Ostermeier M Mathematical expressions useful in the construction, description and evaluation of protein libraries Biomol. Eng 2005 22 57 61 15857784 32 Plaxco KW Simons KT Baker D Contact order, transition state placement and the refolding rates of single domain proteins J. Mol. Biol 1998 277 985 994 9545386 33 Stepanenko OV Bublikov GS Stepanenko OV Shcherbakova DM Verkhusha VV Turoverov KK Kuznetsova IM A knot in the protein structure - probing the near-infrared fluorescent protein iRFP designed from a bacterial phytochrome FEBS J 2014 281 2284 2298 24628916 34 Netzer WJ Hartl FU Recombination of protein domains facilitated by co-translational folding in eukaryotes Nature 1997 388 343 349 9237751 35 Agashe VR Guha S Chang H-C Genevaux P Hayer-Hartl M Stemp M Georgopoulos C Hartl FU Barral JM Function of trigger factor and DnaK in multidomain protein folding: increase in yield at the expense of folding speed Cell 2004 117 199 209 15084258 36 Anders K Daminelli-Widany G Mroginski MA Stetten, von D Essen L-O Structure of the cyanobacterial phytochrome 2 photosensor implies a tryptophan switch for phytochrome signaling Journal of Biological Chemistry 2013 288 35714 35725 24174528 37 Narikawa R Fukushima Y Ishizuka T Itoh S Ikeuchi M A novel photoactive GAF domain of cyanobacteriochrome AnPixJ that shows reversible green/red photoconversion J. Mol. Biol 2008 380 844 855 18571200 38 Ishizuka T Narikawa R Kohchi T Katayama M Ikeuchi M Cyanobacteriochrome TePixJ of Thermosynechococcus elongatus harbors phycoviolobilin as a chromophore Plant Cell Physiol 2007 48 1385 1390 17715149 39 Rumyantsev KA Shcherbakova DM Zakharova NI Emelyanov AV Turoverov KK Verkhusha VV Minimal domain of bacterial phytochrome required for chromophore binding and fluorescence Sci Rep 2015 5 18348 26679720 40 Lindberg M Tångrot J Oliveberg M Complete change of the protein folding transition state upon circular permutation Nat. Struct. Biol 2002 9 818 822 12368899 41 Topell S Hennecke J Glockshuber R Circularly permuted variants of the green fluorescent protein FEBS Lett 1999 457 283 289 10471794 42 Jones AM Mehta MM Thomas EE Atkinson JT Segall-Shapiro TH Liu S Silberg JJ The structure of a thermophilic kinase shapes fitness upon random circular permutation ACS Synth Biol 2016 43 Samuel PP Smith LP Phillips GN Olson JS Apoglobin Stability Is the Major Factor Governing both Cell-free and in Vivo Expression of Holomyoglobin Journal of Biological Chemistry 2015 290 23479 23495 26205820 44 Ostermeier M Designing switchable enzymes Curr. Opin. Struct. Biol 2009 19 442 448 19473830 45 Marvin JS Schreiter ER Echevarría IM Looger LL A genetically encoded, high-signal-to-noise maltose sensor Proteins 2011 79 3025 3036 21989929 46 Segall-Shapiro TH Meyer AJ Ellington AD Sontag ED Voigt CA A “resource allocator” for transcription based on a highly fragmented T7 RNA polymerase Molecular Systems Biology 2014 10 742 25080493 47 Graf R Schachman HK Random circular permutation of genes and expressed polypeptide chains: application of the method to the catalytic chains of aspartate transcarbamoylase Proc. Natl. Acad. Sci. U.S.A 1996 93 11591 11596 8876180 48 Firnberg E Ostermeier M PFunkel: efficient, expansive, user-defined mutagenesis PLoS ONE 2012 7 e52031 23284860
PMC005xxxxxx/PMC5122464.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 9607835 20545 Mol Psychiatry Mol. Psychiatry Molecular psychiatry 1359-4184 1476-5578 27217145 5122464 10.1038/mp.2016.80 NIHMS765523 Article Conditional ablation of neuroligin-1 in CA1 pyramidal neurons blocks LTP by a cell-autonomous NMDA receptor-independent mechanism Jiang Man 123 Polepalli Jai 14 Chen Lulu Y. 23 Zhang Bo 23 Südhof Thomas C. 23 Malenka Robert C. 4 2 Department of Molecular and Cellular Physiology, Stanford University School of Medicine, Stanford, CA 94305 3 Howard Hughes Medical Institute, Stanford University School of Medicine, Stanford, CA 94305 4 Nancy Pritzker Laboratory, Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, CA 94305 Correspondence and requests for materials should be addressed to: R.C. Malenka ([email protected]) or T.C. Südhof ([email protected]) 1 These authors contributed equally to this work. 22 9 2016 24 5 2016 3 2017 22 2 2017 22 3 375383 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Neuroligins are postsynaptic cell-adhesion molecules implicated in autism and other neuropsychiatric disorders. Despite extensive work, the role of neuroligins in synapse function and plasticity, especially NMDA receptor (NMDAR)-dependent LTP, remains unclear. To establish which synaptic functions unequivocally require neuroligins, we analyzed single and triple conditional knockout (cKO) mice for all three major neuroligin isoforms (NL1-NL3). We inactivated neuroligins by stereotactic viral expression of Cre-recombinase in hippocampal CA1 region pyramidal neurons at postnatal day 0 (P0) or day 21 (P21), and measured synaptic function, synaptic plasticity, and spine numbers in acute hippocampal slices 2–3 weeks later. Surprisingly, we find that ablation of neuroligins in newborn or juvenile mice only modestly impaired basal synaptic function in hippocampus, and caused no alteration in postsynaptic spine numbers. However, triple cKO of NL1-NL3 or single cKO of NL1 impaired NMDAR-mediated excitatory postsynaptic currents (NMDAR EPSCs), and abolished NMDAR-dependent LTP. Strikingly, the NL1 cKO also abolished LTP elicited by activation of L-type Ca2+-channels during blockade of NMDARs. These findings demonstrate that neuroligins are generally not essential for synapse formation in CA1 pyramidal neurons but shape synaptic properties and that NL1 specifically is required for LTP induced by postsynaptic Ca2+-elevations, a function which may contribute to the pathophysiological role of neuroligins in brain disorders. neuroligin synapse NMDA receptor long-term potentiation (LTP) INTRODUCTION Neuroligins are postsynaptic cell-adhesion molecules that interact with presynaptic neurexins to shape synaptic properties in robust and complex ways1–3. Interest in elucidating the specific functions of neuroligins has been stimulated by their genetic association with neuropsychiatric diseases, most notably autism spectrum disorders (ASD)1,4–6. Despite decades of work, however, controversies remain about specific functions of neuroligins (reviewed in 7). The inconsistencies in results from experiments that study neuroligins have legitimate biological causes (e.g. studying different synapses in different preparations; performing neuroligin manipulations at different ages), but may also partly be due to the methods used. Knockdown approaches using shRNAs or microRNAs as well as overexpression approaches provide valuable information but suffer from inherent limitations. By definition, knockdown approaches do not eliminate all of the targeted proteins and often have off-target effects. Overexpression of proteins may cause targeting of proteins inappropriately such that the observed effects do not necessarily reflect the functions of the endogenous protein. Similarly, constitutive knockouts (KOs) of neuroligins provide information about their absolute necessity but are potentially compromised by developmental compensation, which may hinder functional analyses. Finally, many studies on neuroligins used dissociated cultured neurons or slice cultures, preparations that are generated from developing brains and exhibit on-going synaptogenesis with altered target specificity and thus may not precisely mimic properties of synapses in the intact mammalian brain. To address these issues, we generated single and triple conditional knockout (cKO) lines of the three major neuroligins found in mouse brain8–10 (NL1, NL2, and NL3). NL1 is exclusively localized at excitatory synapses11; NL2 is selectively localized at inhibitory and cholinergic synapses12,13; while NL3 is localized at both excitatory and inhibitory synapses1,14. We did not examine NL4 because it is found at very low levels in mouse brain, and predominantly localizes to glycinergic synapses15. We chose to genetically delete neuroligins in vivo at two different developmental stages in hippocampal CA1 pyramidal neurons by stereotactically injecting viruses expressing Cre-recombinase. By deleting one or more neuroligins at P0 and analyzing acute slices ~3 weeks later, we could define the role of neuroligins in synapse formation and synapse maturation since P0-P14 is a time of robust synaptogenesis in CA116,17. By performing the same experiments after deleting one or more neuroligins at P21, a time at which synapse formation is largely complete, we could assess the role of neuroligins in mature synapses. These experimental procedures provide, arguably, the most biologically relevant and rigorous test of the role of neuroligins in synapse function since conditional genetic deletion in vivo minimizes the possibility of compensatory adaptations while performing the deletion in the intact brain. Our results demonstrate that neuroligins perform an essential role in long-term synaptic plasticity in the hippocampus and additionally contribute to shaping synapse properties and strength, but are expendable for synapse formation and maintenance both in developing and in mature neurons. MATERIALS AND METHODS Mouse lines All experiments were approved by the Administrative Panel on Laboratory Animal Care at Stanford University. All experiments used homozygous NL1, NL2, and NL3 single and triple cKO mice8–10 that were maintained on a mixed CD1/C57BL6 background except for the NL3 cKO line (pure C57BL6 background) and NL2 cKO line (pure CD1 background). Stereotactic injections of lentiviruses Lentiviruses expressing eGFP-tagged Cre-recombinase driven by a ubiquitin promoter were procured from the Stanford Gene Vector and Virus Core and stereotactically injected into the CA1 region of the hippocampus as described18,19. For P0 injections (coordinates: ~1.0 mm anterior to interaural; ~0.7 mm lateral to midline; depth of ~1.15 mm from the surface of the skull, with 0.5 μl lentivirus injected at 0.8 μl/min with a microinjection pump), pups were anesthetized on ice for 5 min and immobilized with tape on a mold that maintained the top surface of the skull horizontally. Procedures for virus injections at P21 were as previously described20. For vGAT quantification experiments 1.5 μl of lentivirus was injected unilaterally. Electrophysiology Mice were analyzed at P18-25 (P0 injections) or at P35-42 (P21 injections). Transverse hippocampal slices from dorsal hippocampus were cut20,21 in a solution containing (in mM): 228 sucrose, 26 NaHCO3, 11 glucose, 2.5 KCl, 1 NaH2PO4, 7 MgSO4 and 0.5 CaCl2, and recovered in artificial cerebrospinal fluid (ACSF) containing (in mM): 119 NaCl, 26 NaHCO3, 11 glucose, 2.5 KCl, 1 NaH2PO4, 1.3 MgSO4 and 2.5 CaCl2. Cre-recombinase expressing cells were identified by eGFP epifluorescence; uninfected cells (eGFP negative) from the same slices were used as controls. The identity of recorded infected versus uninfected cells was confirmed by the eGFP signal in the nucleus which was pulled into the recording electrode after the recording. Whole-cell recordings were made using an internal solution containing (in mM): 140 CsMeSO4, 8 CsCl, 10 HEPES, 0.25 EGTA, 2 Mg2ATP, 0.3 Na3GTP, 0.1 spermine, 7 phosphocreatine (pH 7.25–7.3; osmolarity 294–298). For morphological reconstruction experiments, the internal solution additionally contained 0.2% biocytin. AMPAR-mediated mEPSCs were recorded with tetrodotoxin (TTX, 1 μM), D-APV (50 μM), and picrotoxin (PTX, 50 μM) in the ACSF. For GABAAR-mediated mIPSCs, CsMeSO4 was replaced by 140 mM CsCl in the internal soultion and TTX, CNQX (20 μM), and D-APV (50 μM) were added to the ACSF. A theta glass pipette filled with ACSF was placed in the stratum radiatum to evoke EPSCs in CA1 pyramidal cells. PTX was included in extracellular ACSF in all experiments examining excitatory synaptic transmission. Stimulation pulses were delivered every 10 s. NMDAR/AMPAR ratios were calculated as the averaged NMDAR-mediated EPSC (20 trials, measured at 50 ms after the onset of EPSCs at +40 mV) divided by the averaged AMPAR-mediated EPSC (20 trials, measured as the peak amplitude of EPSCs at −70 mV). Two pulses at different intervals (20, 50,100, 200, 500 ms) were delivered to calculate paired pulse ratios (PPR). To induce NMDAR-dependent LTP, after collecting a 5–10 min baseline 2 trains of high frequency stimulation (100 Hz, 1s) separated by 20 s were delivered while clamping the postsynaptic cell at 0 mV20. To induce NMDAR-independent LTP22, twenty postsynaptic depolarization pulses (80 mV, 1 s separated by 6 s) were applied to cells clamped at −70 mV. 5 μM Bay K8644 (an L-type calcium channel activator) and 50 μM APV were included in the ACSF. All drugs were obtained from Tocris except tetanus-toxin light chain (List Biological Laboratories). Immunohistochemistry Immunohistochemical experiments were performed as previously described10. Briefly, triple NL123 floxed mice (~P21) were anesthetized with isoflurane, perfused with phosphate buffered saline (PBS) followed by 4% paraformaldehyde (PFA) in 0.1 M PBS via a perfusion pump (2 ml/min). Perfused brains were post-fixed in 4% PFA for 2 h at room temperature (RT) then cryoprotected in 30% sucrose (in 1× PBS) for 24 h at 4°C. Coronal brain sections (30 μm) were cryo-sectioned at −20°C (Leica CM1050). Sections were serially washed with PBS and incubated in blocking solution (0.3 % Triton X-100 and 5 % goat serum in PBS) for 1 h at RT, and incubated for 24 h at 4 °C with primary antibodies diluted in PBS (anti-vGAT, 1:500, rabbit, Synaptic Systems). Sections were washed 4 times (15 min each time) in PBS, treated with secondary antibodies (1:1000, Alexa 555, Invitrogen) at 4°C overnight, and washed 4 times (15 min each time) again in PBS. Sections were then mounted on superfrost slides and covered with mounting media (Vectashield, Vector Labs). For triple NL123 cKO group, only slices with robust virus infection were selected for imaging. Single plane images (at 1024 x 1024 resolution) from hippocampal CA1 region were acquired using a Nikon confocal microscope (A1Rsi) with a 60x oil objective (PlanApo, NA1.4). All acquisition parameters were kept constant among different conditions within experiments. Image backgrounds were normalized, and immunoreactive elements were analyzed with Nikon analysis software. The density of vGAT positive puncta in striatum pyramidale (40 μm X 40 μm) and striatum radiatum (40 μm X 60 μm) were automatically analyzed using the same setting for both control and cKO images. For spine imaging experiments, slices were fixed after electrophysiology experiments with 4% PFA in PBS for 2 h, washed 3 times (15 min each time) with PBS, and sequentially incubated at room temperature with PBS containing 10% goat serum/0.5% Triton X-100 for 1 h, and with PBS containing streptavidin (Alexa Fluor 594 conjugate, 1:1000, Invitrogen) for 2 h. Slices were washed 3 times (15 min each time) in PBS, mounted, and analyzed by 3D-imaging of dendrites using the same confocal microscope. For every labeled neuron, 5–8 regions of interest (10 μm diameter) from different secondary/tertiary dendrites were selected for manual spine counting on anonymized images. RESULTS Conditional deletion of neuroligins at P0 reduces inhibitory but not excitatory synapse formation To define the requisite roles of NL1-3 in synapse formation, maturation and function, we used mice containing cKO alleles of NL1, NL2 and NL3 that were generated using standard approaches8–10 (Fig. 1A). Lentiviruses expressing eGFP-tagged cre-recombinase were stereotactically injected into the CA1 region of the hippocampus at P0, and whole-cell recordings from infected and uninfected cells were made from acute slices prepared at P18-25 (Fig. 1B). Using modest titers of lentivirus, we obtained sparse infectivity (Fig. 1C) that allowed recordings in the same slice from control and neuroligin-deficient neurons. The triple NL123 cKO at PO caused ~50% reduction in mean mIPSC frequency with a clear right-shift in the cumulative distribution of mIPSC inter-event intervals and only a minor decrease in mIPSC amplitudes (Fig. 1D). Accordingly, immunostaining for GABAergic synapses showed a ~30% decreased number of vGAT puncta in the CA1 pyramidal cell layer and stratum radiatum in NL123 cKO mice (Fig 1E). To assess the relative contributions of NL2 and NL3, which are both expressed at inhibitory synapses13,14, we repeated this analysis in single NL2 and NL3 cKOs. The NL2 cKO phenocopied the NL123 cKO (Fig. S1A). Surprisingly, NL3 cKO at P0 caused a modest increase in mIPSC frequency but no change in mIPSC amplitudes (Fig. S1B). In marked contrast to the effects on mIPSCs, the triple NL123 cKO had no detectable effect on the frequency or amplitude of AMPA receptor (AMPAR)-mediated mEPSCs (Fig. 1F). Similarly, the triple NL123 cKO at P0 had no effect on the density of dendritic spines on secondary and tertiary dendritic branches of CA1 pyramidal neurons (Fig. 1G). Thus, neuroligins are not required for the formation of excitatory synapses or function of AMPARs on CA1 pyramidal neurons during early postnatal development, but are required for the development of normal inhibitory synaptic function. Neuroligins play a modest role in mature inhibitory and excitatory synapse function To evaluate whether neuroligins play a requisite role in the maintenance of function at mature inhibitory and excitatory synapses on CA1 pyramidal neurons, we injected Cre-expressing lentivirus at P21 into triple NL123 cKO mice and prepared acute slices 2–3 weeks later (Fig 2A). Unexpectedly, this manipulation caused only very small decreases in both mIPSC frequency and mIPSC amplitudes (Fig. 2B). To assess the relative contributions of NL2 and NL3 in the maintenance of inhibitory synapse function, we repeated this analysis in single NL2 and NL3 cKOs. The NL2 cKO showed a small decrease in the amplitudes of mIPSCs without any change in their frequency (Fig. S2A) while NL3 cKO at P21 caused no changes in either mIPSC amplitude or frequency (Fig. S2B) thus specifying a redundant role for NL2 and NL3 in inhibitory synapse maintenance and a role for NL2 alone in the functioning of these synapses in adult mice. The genetic deletion of NL123 or NL1 alone at P21 had no detectable effects on mEPSC frequency or amplitude (Fig. 2C, S3A). Morphological analysis revealed that spine density was also not altered by NL123 KO at P21 (Fig. 2D) Thus, neuroligins are not necessary for maintaining a normal complement of excitatory synapses on CA1 pyramidal cells and play only a minor role in maintaining inhibitory synaptic function. NL1 cKO at P0 and P21 reduces NMDAR-mediated synaptic transmission Previous work using constitutive NL1 KO mice or NL1 KD approaches demonstrated a critical role for NL1 in maintaining a normal complement of NMDARs, as measured by changes in NMDAR EPSCs7, 23–27. Consistent with these findings, the triple NL123 cKO at P0 caused ~40% decrease in the ratio of NMDAR- to AMPAR EPSCs (NMDAR:AMPAR ratio; Fig. 3A, B). This decrease was entirely due to the loss of NL1 as the NL1 cKO phenocopied the decrease in the NMDAR:AMPAR ratio, whereas the NL3 cKO did not (Fig. 3B). This finding was further validated by a ~35% decrease in the input-output relationship for NMDAR EPSCs without any change in the input-output relationship for AMPAR EPSCs in the NL1 cKO (Fig. S3B). The lack of an effect in the NL3 cKO is consistent with previous results obtained in NL3 constitutive KO mice28, demonstrating that NL3 cannot compensate for the loss of NL1 in CA1 pyramidal cells. To examine whether the triple NL123 cKO influences presynaptic function via trans-synaptic interactions, we measured paired-pulse ratios (PPRs) of AMPAR EPSCs, which inversely correlate with presynaptic release probability29. The NL123 cKO had no effect on PPRs examined at multiple inter-stimulus intervals (Fig. 3C). Essentially identical results were obtained when the triple NL123 cKO and the single NL1 and NL3 cKOs were analyzed after injections of Cre lentivirus at P21 (Fig. 3D, E). Neither NL1 nor NL3 cKO at P21 showed any changes in PPRs, confirming that deletion of neuroligins has no effect on presynaptic release probability (Fig. S4). Thus, at excitatory synapses on CA1 pyramidal cells NL1 is required for the development and maintenance of normal NMDAR-mediated synaptic transmission. NL1 cKO blocks long-term potentiation (LTP) Constitutive KO of NL1 impairs LTP in CA1 pyramidal neurons30, 31, yet microRNA mediated KD of NL1 in mature hippocampus had no effect on LTP at these same CA1 synapses27,30. To address whether neuroligins are in fact required for this classic form of synaptic plasticity, we genetically deleted NL123 at P0 (Fig. 4A–D) and P21 (Fig. 4E–H). Strikingly, we found that at both ages, LTP was, on average, completely blocked. This block was entirely due to the loss of NL1 as the single NL1 but not NL3 cKO at P21 phenocopied the block of LTP (Fig. 4I–P). Given that the genetic deletions of neuroligins, which blocked LTP, also caused a decrease in NMDAR EPSCs, an obvious possibility is that the block of LTP is due to the ~40% decrease in NMDAR-mediated synaptic transmission. To address this possibility, we first identified a concentration of D-APV (2 μM) that reduced isolated NMDAR-EPSCs and NMDAR charge transfer during high frequency stimulation by ~50% (Fig. S5A, B). At this concentration of D-APV, LTP was not blocked, although its magnitude was reduced (Fig. 5A–D). Thus, this experiment suggested that the block in LTP was independent of the effect of the neuroligin ablation on NMDARs, but is not conclusive given the partial effect. To more definitively explore the importance of the decrease in NMDAR EPSCs in contributing to the block of LTP caused by neuroligin KO, we used a method that induces LTP in an NMDAR-independent manner. In this protocol, NMDARs are blocked pharmacologically by high concentrations of AP5 (50 μM). Potentiation of AMPAR-EPSCs is then elicited by repetitive activation of voltage-gated Ca2+-channels (VGCCs), primarily L-type Ca2+-channels22,32,334 (Fig. S6). VGCC LTP was prevented by loading cells with tetanus toxin (TeTx) (Fig. 5E–H), which cleaves synaptobrevin-2 in a manner similar to botulinum toxin that blocks NMDAR-dependent LTP when loaded in CA1 pyramidal cells34. These results suggest that similar to NMDAR-dependent LTP, VGCC LTP also requires SNARE-mediated exocytosis, which is presumably used for the delivery of AMPARs to the plasma membrane35. Importantly, the NL1 cKO also completely blocked VGCC LTP (Fig. 5I–L). Together, these results strongly suggest that NL1 plays a critical role in LTP in CA1 pyramidal cells independent of its influence on NMDARs. DISCUSSION Neuroligins are postsynaptic cell-adhesion molecules that are ubiquitously present at excitatory and inhibitory synapses, are genetically associated with several neuropsychiatric disorders, and appear to play complex roles in synapse function1–6. Despite decades of work, the specific roles of neuroligins in synapse formation and synaptic function remain uncertain, even controversial (see ref. 7 for a summary). Both methodological differences and biologically relevant factors likely contribute to the variability in results obtained with molecular manipulations of neuroligins. Knockdown of neuroligins in culture and in vivo using shRNAs or microRNAs has proved useful but suffers from possible off-target effects and incomplete loss of targeted proteins. Constitutive knockout of neuroligins eliminates the targeted proteins but allows compensatory developmental adaptations. Thus, using a conditional KO approach currently provides the optimal manipulation by allowing complete genetic deletion of the targeted proteins in individual cells in a temporally controlled fashion. For proteins such as neuroligins, which may play distinct roles at different developmental stages, the ability to control the time point at which they are genetically eliminated is particularly important. However, even with conditional genetic deletion approaches, because it takes some time for the gene to be deleted and endogenous proteins to degrade, compensatory adaptations may still occur. Using single and triple cKOs of NL1, NL2, and NL3, we examined their requisite roles in excitatory and inhibitory synapse formation and function in hippocampal CA1 pyramidal neurons. We chose to study these specific synapses because they are, arguably, the most extensively studied synapses in the mammalian brain. By expressing cre-recombinase in vivo at two different ages (P0 and P21), we were able to address the role of neuroligins in both synapse formation and mature synaptic function. Examining sparsely infected neurons enabled us to assess neuroligin functions under conditions where mutant neurons are surrounded by, and competing with, adjacent wild-type neurons. In this approach, our goal was to define, in the most rigorous fashion possible, the requisite synaptic roles of neuroligins. Consistent with previous analyses of constitutive triple NL123 KO mice36, cKO of NL123 did not affect excitatory synapse formation or maintenance as assessed by measurements of dendritic spine density and mEPSCs. Also in line with previous studies8,12,37–40, we found that the triple NL123 cKO at P0 caused a robust decrease in mIPSC frequency and VGAT puncta density, likely due to the loss of NL2, the deletion of which alone largely phenocopied the triple NL123 cKO. It was surprising that NL3 cKO at P0 caused a small increase in mIPSC frequency. This may occur because NL3 modestly impairs inhibitory synapse formation or function when present alone at these inhibitory synapses. Alternatively, in the absence of NL3, NL2 may be able to stimulate inhibitory synapse formation and function more robustly. It is also conceivable that despite the use of cKO lines, the cre-recombinase mediated deletion allows sufficient time for compensatory adaptations to occur. Furthermore, since the genetic backgrounds of the NL3 cKO line and NL2 cKO line are different than the other cKO lines, it is formally possible that the synaptic effects of neuroligin deletion are influenced by the background genetic strain in which the deletion is performed. Despite these caveats, the in vivo conditional KO approach we have taken here clearly allows a much more rigorous test of the requisite function of neuroligins than previous attempts using KD or constitutive KO approaches. In contrast to previous studies in which neuroligins were deleted in other brain regions, NL123 cKO at P21 in the hippocampus had only modest effects on inhibitory synaptic transmission. However, consistent with previous work using neuroligin KD and constitutive KO approaches23–25, cKO of NL123 at both P0 and P21 caused a large decrease in the ratio of NMDAR- to AMPAR EPSCs combined with a lack of effect on AMPAR-mediated mEPSCs. This decrease in synaptic NMDAR number and/or function was entirely due to loss of NL1 since cKO of NL1, but not cKO of NL3, phenocopied these synaptic changes. Arguably, our most interesting finding is that the triple NL123 cKO, due to deletion of NL1, profoundly impairs LTP in CA1 pyramidal cells at both P0 and, importantly, at P21 after synapses are fully mature. Although impairments of LTP following neuroligin KD or KO have been reported23,25,27,30,31, previous work had produced conflicting conclusions, possibly because of the approaches used. Our data establish that NL1 is essential for LTP independent of development. The simplest explanation for the apparent difference in our results from previous ones using a knockdown of NL127 is that the NL1 knockdown in mature CA1 pyramidal cells did not reduce NL1 levels sufficiently to detect its critical role in LTP. The fact that NL1 is essential for maintaining NMDAR responses24 raises the obvious possibility that the loss of LTP in NL1-deficient pyramidal neurons is a secondary effect of smaller NMDAR-mediated synaptic currents. Surprisingly, our data show that this is not the case. First, pharmacologically reducing NMDAR-mediated transmission to a degree greater than the reduction caused by NL1 deletion reduced the magnitude of LTP to a degree much less than the complete block caused by NL1 cKO (Fig. 5A–D). Second, activation of voltage-dependent Ca2+-channels to bypass NMDARs elicited a robust LTP in control neurons that was eliminated by cKO of NL1 (Fig. 5I–L). This form of LTP is incompletely understood22,32,33, but our results show that it is completely blocked by tetanus-toxin mediated cleavage of the SNARE-protein synaptobrevin-2 similar to NMDAR-dependent LTP34, suggesting a similar cellular mechanism induced by a different calcium source. The apparently independent effects of the NL1 cKO on NMDAR-mediated synaptic transmission and LTP demonstrate that NL1 is a critical component of the molecular machinery that underlies synaptic plasticity in CA1 pyramidal cells. It seems likely that NL1 enables LTP by recruiting crucial cytoplasmic proteins to synaptic junctions such as PSD95, which directly binds to NL141. In this manner, NL1 constitutes a central component of the postsynaptic machinery. In summary, using a rigorous conditional KO approach, our data provide evidence for the following conclusions (summarized in Fig. S7). First, neuroligins are not absolutely necessary for the formation or maintenance of excitatory synapses on CA1 pyramidal cell dendritic spines, even if the mutant neuron is surrounded by wild-type neurons. Second, NL1 is essential for normal NMDAR-mediated synaptic transmission, presumably by influencing the numbers of synaptic NMDARs. The same decrease in NMDAR-mediated EPSCs was seen in slices of the basolateral amygdala following NL1 KD23, suggesting that this function of NL1 applies to many different cell types. Third, neuroligins make only a modest contribution to excitatory synaptic strength, but a much more significant contribution to inhibitory synaptic strength8,9. Fourth, at least in CA1 pyramidal neurons, NL1 plays a mandatory role in LTP, and this function of NL1 is independent of its influence on NMDARs. Although much work remains to be done to elucidate the molecular mechanisms underlying the ubiquitous and cell-type specific roles of neuroligins in excitatory and inhibitory synaptic function, these results provide a framework that can form the basis for specialized future studies. Supplementary Material 1 We thank members of the Malenka and Südhof labs for many helpful conversations. This study was supported by a grant from NIH (P50MH086403) and the Simons Foundation Autism Research Initiative. Figure 1 Sparse deletion of Neuroligins from hippocampal CA1 pyramidal neurons during postnatal development produces impairments in inhibitory but not fast excitatory synaptic transmission (A) Schematic diagram of the NL cKO alleles. (B) Experimental design for injections at P0, followed by experimental analyses at postnatal P18–25. (C) Representative images of the hippocampus from a P0 injected mouse (left) and of sparsely infected CA1-region neurons (right; blue, DAPI; green, GFP). In all experiments, non-infected neurons served as controls for adjacent infected neurons in the same slices. (D) Analysis of mIPSCs in triple NL123 cKO neurons. Left, representative traces; middle, cumulative distribution of the mIPSC inter-event interval (inset: data points from individual cells and means of mIPSC frequency); right, cumulative distribution of mIPSC amplitudes (inset: data points from individual cells and mean mIPSC amplitudes). (E) vGAT staining in stratum pyramidale and stratum radiatum of hippocampal CA1 region. Left, low resolution images showing without (top, control) and with (bottom, cKO) robust infection of Ub-eGFP-Cre lentivirus, scale bar: 200 μm; Middle, high resolution images from the white box in control and cKO slices used for analysis, scale bar: 50 μm; Group data showing that the density of inhibitory synapses in stratum pyramidale and stratum radiatum of CA1 region was decreased in triple NL123 cKO. (F) Same as D, but for mEPSCs in triple NL123 cKO neurons. (G) NL123 deletion in newborn mice does not change the spine density of CA1 pyramidal cells. Left, representative images of biocytin-labeled CA1 pyramidal neurons after patch-clamp recording, with a lower and higher magnification images shown side by side (calibration bars: 50 μm and 2 μm, respectively); middle, cumulative distribution of spine density (control: 173 dendrites/22 neurons; cKO: 77 dendrites/10 NL123 cKO pyramidal neurons); right, summary graph of mean spine densities with data points from individual neurons. Data in summary graphs are means ± SEM; statistical comparisons were performed with the Kolmogorov-Smirnov test (cumulative distributions) or student’s t-test (, p<0.05; , p<0.01; *, p<0.001; non-significant comparisons are not labeled). Numbers indicate number of cells/mice examined. Figure 2 Sparse deletion of Neuroligins from mature hippocampal CA1 pyramidal neurons in juvenile mice causes similar but less severe phenotypes as deletions in developing neurons (A) Experimental design for injections at P21, followed by experimental analyses from P35 onwards. (B) Analysis of mIPSCs in triple NL123 cKO neurons. Left, representative traces; middle, cumulative distribution of the mIPSC inter-event interval (inset: data points from individual cells and means of mIPSC frequency); right, cumulative distribution of mIPSC amplitudes (inset: data points from individual cells and mean mIPSC amplitudes). (C) Same as B, but for mEPSCs. (D) NL123 deletion in juvenile mice also does not change the spine density of CA1 pyramidal cells. Left, representative images of biocytin-stained CA1 pyramidal neurons after patch-clamp recording, with a lower and higher magnification images shown side by side (calibration bars: 50 μm and 2 μm, respectively); middle, cumulative distribution of spine density (control: 100 dendrites/13 neurons; cKO: 45 dendrites/6 NL123 cKO pyramidal neurons); right, summary graph of mean spine densities with data points from individual neurons. Data in summary graphs are means ± SEM; statistical comparisons were performed with the Kolmogorov-Smirnov test (cumulative distributions) or student’s t-test (, p<0.05; , p<0.01; , p<0.001; non-significant comparisons are not labeled). Numbers indicate number of cells/mice examined. Figure 3 Single NL1 but not NL3 deletion significantly reduces NMDAR-mediated synaptic transmission in developing and juvenile CA1 pyramidal neurons (A) Experimental design for conditional NL deletions in developing mice using P0 injections. (B) Measurements of the ratio of NMDAR- to AMPAR-mediated EPSCs in CA1 pyramidal neurons in triple NL123 cKO (left) and NL1 (middle) and NL3 single cKO mice (right). AMPAR-mediated EPSCs were quantified as peak EPSC amplitude monitored at -70 mV; NMDAR-mediated EPSCs were quantified as the EPSC amplitude at 50 ms after presynaptic stimulation monitored at +40 mV. Top, representative traces; bottom, summary plots of mean NMDAR/AMPAR ratios and the values from individual neurons (C) Paired pulse ratio (PPR) of EPSCs was not changed in triple NL123 cKO neurons. Top, representative traces; bottom, summary plot of mean PPRs as a function of the inter-stimulus interval (n = 16 control neurons/5 mice and 14 cKO neurons/5 mice, respectively). (D) Experimental design for conditional NL deletions in juvenile mice using P21 injections. (E) Same as B, but for juvenile mice. (F) Same as C, but for juvenile mice (n = 10 control neurons/3 mice and 10 cKO neurons/3 mice, respectively). Data in summary graphs are means ± SEM; statistical comparisons were performed with student’s t-test (, p<0.05; , p<0.01; , p<0.001; non-significant comparisons are not labeled). Numbers indicate number of cells/mice examined. Figure 4 NL123 triple and NL1 single deletion but not NL3 deletion abolishes NMDAR-dependent LTP in CA1 pyramidal neurons (A) Representative LTP experiments in a control (left) and NL123 cKO neuron (right) at P18 after lentiviral deletion of Neuroligins at P0. Top, sample traces; bottom, plots of the EPSC amplitude as a function of time before (yellow background) and after the LTP induction stimulus (2 trains of 100 Hz for 1 s with the cell depolarized to 0mV, separated by 20 s). (B) Summary plot of LTP shows LTP was completely blocked in triple NL123 cKO following P0 lentiviral injection. (C) Cumulative distribution of normalized LTP magnitud at 35-40 min after LTP induction. (D) Summary graph of the mean LTP magnitude at 35-40 min after LTP induction. (E–H), same as (A–D), except for the triple NL123 cKO was induced by lentiviral infection of CA1 pyramidal neurons in juvenile mice at P21. (I–L), same as (E–H), except for the NL1 single cKO at P21. (M–P), same as (E–H), except for the NL3 single cKO at P21. Data in summary graphs are means ± SEM; statistical comparisons were performed with the Kolmogorov-Smirnov test (cumulative distributions) or student’s t-test (, p<0.05; , p<0.01; , p<0.001; non-significant comparisons are not labeled). Numbers indicate number of cells/mice examined. Figure 5 NL1 cKO abolishes NMDAR-independent LTP in juvenile mice (A) Representative LTP experiments in wild-type mice showing that partial inhibition of NMDARs with AP5 (2 μM) impairs, but does not block, LTP induced by a standard induction protocol (2 trains of 100 Hz for 1 s separated by 20 s). Top, sample traces; bottom, plots of the EPSC amplitude as a function of time before (yellow background) and after the LTP induction stimulus. (B) Summary plot of LTP in wild-type mice showing that this low dose of AP5 impaired, but did not block, LTP induced by a standard induction protocol. (C) Cumulative distribution of normalized LTP magnitude at 35–40 min after LTP induction for B. (D) Summary graph of the mean LTP magnitude at 35–40 min after LTP induction for B. (E) Representative LTP experiments in wild-type mice using an NMDAR-independent induction protocol and demonstrating that similar to NMDAR-dependent LTP, NMDAR-independent LTP is also inhibited by postsynaptic tetanus-toxin (TeTx) light chain (right, 100 nM tetanus-toxin in the pipette solution). LTP was induced by 20 postsynaptic depolarizations (80 mV, 1 s separated by 6 s) in the presence of 50 μM AP5 and 5 μM Bay K 8644 Top, sample traces; bottom, plots of the EPSC amplitude as a function of time before (yellow background) and after the LTP induction stimulus. (F) Summary plot of LTP induced by L-type Ca2+-channel mediated Ca2+-influx under continuous NMDAR-inhibition in control cells from wild-type mice, and its inhibition by tetanus toxin light chain. (G) Cumulative distribution of normalized LTP magnitude at 35–40 min after LTP induction for F. (H) Summary graph of the mean LTP magnitude at 35–40 min after LTP induction for F. (I–L) Same as E–H, but comparing control neurons to NL1 cKO neurons produced by stereotactic lentiviral injection at P21. Data in summary graphs are means ± SEM; statistical comparisons were performed with the Kolmogorov-Smirnov test (cumulative distributions) or student’s t-test (, p<0.05; , p<0.01; *, p<0.001; non-significant comparisons are not labeled). Numbers indicate number of cells/mice examined. CONFLICT OF INTEREST The authors declare no conflict of interest. 1 Sudhof TC Neuroligins and neurexins link synaptic function to cognitive disease Nature 2008 455 7215 903 911 18923512 2 Hu X Luo J-h Xu J The interplay between synaptic activity and neuroligin function in the CNS BioMed Research International 2015 2015 13 3 Bang M Owczarek S A matter of balance: Role of neurexin and neuroligin at the synapse Neurochemical Research 2013 38 6 1174 1189 23559421 4 Huguet G Ey E Bourgeron T The genetic landscapes of autism spectrum disorders Annu Rev Genomics Hum Genet 2013 14 191 213 23875794 5 Jamain S Quach H Betancur C Rastam M Colineaux C Gillberg IC Mutations of the X-linked genes encoding neuroligins NLGN3 and NLGN4 are associated with autism Nat Genet 2003 34 1 27 29 12669065 6 Feng J Schroer R Yan J Song W Yang C Bockholt A High frequency of neurexin 1beta signal peptide structural variants in patients with autism Neurosci Lett 2006 409 1 10 13 17034946 7 Soler-Llavina GJ Fuccillo MV Ko J Sudhof TC Malenka RC The neurexin ligands, neuroligins and leucine-rich repeat transmembrane proteins, perform convergent and divergent synaptic functions in vivo Proc Natl Acad Sci USA 2011 108 40 16502 16509 21953696 8 Liang J Xu W Hsu YT Yee AX Chen L Sudhof TC Conditional neuroligin-2 knockout in adult medial prefrontal cortex links chronic changes in synaptic inhibition to cognitive impairments Mol Psychiatry 2015 20 7 850 859 25824299 9 Zhang B Chen L Liu X Maxeiner S Lee SJ Gokce O Neuroligins sculpt cerebellar purkinje-cell circuits by differential control of distinct classes of synapses Neuron 2015 87 4 781 796 26291161 10 Rothwell PE Fuccillo MV Maxeiner S Hayton SJ Gokce O Lim BK Autism-associated neuroligin-3 mutations commonly impair striatal circuits to boost repetitive behaviors Cell 2014 158 1 198 212 24995986 11 Song J-Y Ichtchenko K Südhof TC Brose N Neuroligin 1 is a postsynaptic cell-adhesion molecule of excitatory synapses Proc Natl Acad Sci USA 1999 96 3 1100 1105 9927700 12 Poulopoulos A Aramuni G Meyer G Soykan T Hoon M Papadopoulos T Neuroligin 2 drives postsynaptic assembly at perisomatic inhibitory synapses through gephyrin and collybistin Neuron 2009 63 5 628 642 19755106 13 Takacs VT Freund TF Nyiri G Neuroligin 2 is expressed in synapses established by cholinergic cells in the mouse brain PLoS One 2013 8 9 e72450 24039767 14 Budreck EC Scheiffele P Neuroligin-3 is a neuronal adhesion protein at GABAergic and glutamatergic synapses Eur J Neurosci 2007 26 7 1738 1748 17897391 15 Hoon M Soykan T Falkenburger B Hammer M Patrizi A Schmidt K-F Neuroligin-4 is localized to glycinergic postsynapses and regulates inhibition in the retina Proc Natl Acad Sci USA 2011 108 7 3053 3058 21282647 16 Cohen AS Lin DD Coulter DA Protracted postnatal development of inhibitory synaptic transmission in rat hippocampal area CA1 neurons J Neurophysiol 2000 84 5 2465 2476 11067989 17 Paolicelli RC Bolasco G Pagani F Maggi L Scianni M Panzanelli P Synaptic pruning by microglia is necessary for normal brain development Science 2011 333 6048 1456 1458 21778362 18 Xu W Morishita W Buckmaster PS Pang ZP Malenka RC Sudhof TC Distinct neuronal coding schemes in memory revealed by selective erasure of fast synchronous synaptic transmission Neuron 2012 73 5 990 1001 22405208 19 Kaeser PS Deng L Wang Y Dulubova I Liu X Rizo J RIM proteins tether Ca2+ channels to presynaptic active zones via a direct PDZ-domain interaction Cell 2011 144 2 282 295 21241895 20 Ahmad M Polepalli JS Goswami D Yang X Kaeser-Woo YJ Sudhof TC Postsynaptic complexin controls AMPA receptor exocytosis during LTP Neuron 2012 73 2 260 267 22284181 21 Skrede KK Westgaard RH The transverse hippocampal slice: a well-defined cortical structure maintained in vitro Brain Res 1971 35 2 589 593 5135556 22 Kato HK Watabe AM Manabe T Non-Hebbian synaptic plasticity induced by repetitive postsynaptic action potentials J Neurosci 2009 29 36 11153 11160 19741122 23 Kim J Jung SY Lee YK Park S Choi JS Lee CJ Neuroligin-1 is required for normal expression of LTP and associative fear memory in the amygdala of adult animals Proc Natl Acad Sci USA 2008 105 26 9087 9092 18579781 24 Chubykin AA Atasoy D Etherton MR Brose N Kavalali ET Gibson JR Activity-dependent validation of excitatory versus inhibitory synapses by neuroligin-1 versus neuroligin-2 Neuron 2007 54 6 919 931 17582332 25 Jung SY Kim J Kwon OB Jung JH An K Jeong AY Input-specific synaptic plasticity in the amygdala is regulated by neuroligin-1 via postsynaptic NMDA receptors Proc Natl Acad Sci USA 2010 107 10 4710 4715 20176955 26 Shipman SL Schnell E Hirai T Chen BS Roche KW Nicoll RA Functional dependence of neuroligin on a new non-PDZ intracellular domain Nat Neurosci 2011 14 6 718 726 21532576 27 Shipman SL Nicoll RA A subtype-specific function for the extracellular domain of neuroligin 1 in hippocampal LTP Neuron 2012 76 2 309 316 23083734 28 Etherton M Foldy C Sharma M Tabuchi K Liu X Shamloo M Autism-linked neuroligin-3 R451C mutation differentially alters hippocampal and cortical synaptic function Proc Natl Acad Sci U S A 2011 108 33 13764 13769 21808020 29 Zucker RS Regehr WG Short-term synaptic plasticity Annu Rev Physiol 2002 64 355 405 11826273 30 Blundell J Blaiss CA Etherton MR Espinosa F Tabuchi K Walz C Neuroligin-1 deletion results in impaired spatial memory and increased repetitive behavior J Neurosci 2010 30 6 2115 2129 20147539 31 Budreck EC Kwon O-H Jung JH Baudouin S Thommen A Kim H-S Neuroligin-1 controls synaptic abundance of NMDA-type glutamate receptors through extracellular coupling Proc Natl Acad Sci U S A 2011 110 2 725 730 32 Wyllie DJ Nicoll RA A role for protein kinases and phosphatases in the Ca2+-induced enhancement of hippocampal AMPA receptor-mediated synaptic responses Neuron 1994 13 3 635 643 7917294 33 Kullmann DM Perkel DJ Manabe T Nicoll RA Ca2+ entry via postsynaptic voltage-sensitive Ca2+ channels can transiently potentiate excitatory synaptic transmission in the hippocampus Neuron 1992 9 6 1175 1183 1361129 34 Lledo P-M Zhang X Sudhof TC Malenka RC Nicoll RA Postsynaptic membrane fusion and long-term potentiation Science 1998 279 399 403 9430593 35 Jurado S Goswami D Zhang Y Molina AJ Sudhof TC Malenka RC LTP requires a unique postsynaptic SNARE fusion machinery Neuron 2013 77 3 542 558 23395379 36 Varoqueaux F Aramuni G Rawson RL Mohrmann R Missler M Gottmann K Neuroligins determine synapse maturation and function Neuron 2006 51 6 741 754 16982420 37 Fu Z Vicini S Neuroligin-2 accelerates GABAergic synapse maturation in cerebellar granule cells Mol Cell Neurosci 2009 42 1 45 55 19463950 38 Hines RM Wu L Hines DJ Steenland H Mansour S Dahlhaus R Synaptic imbalance, stereotypies, and impaired social interactions in mice with altered neuroligin 2 expression J Neurosci 2008 28 24 6055 6067 18550748 39 Jedlicka P Hoon M Papadopoulos T Vlachos A Winkels R Poulopoulos A Increased dentate gyrus excitability in neuroligin-2-deficient mice in vivo Cereb Cortex 2011 21 2 357 367 20530218 40 Gibson JR Huber KM Sudhof TC Neuroligin-2 deletion selectively decreases inhibitory synaptic transmission originating from fast-spiking but not from somatostatin-positive interneurons J Neurosci 2009 29 44 13883 13897 19889999 41 Irie M Hata Y Takeuchi M Ichtchenko K Toyoda A Hiaro K Binding of neuroligins to PSD-95 Science 1997 277 5331 1511 1515 9278515
PMC005xxxxxx/PMC5122468.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 0236217 6077 Neuropharmacology Neuropharmacology Neuropharmacology 0028-3908 1873-7064 27235163 5122468 10.1016/j.neuropharm.2016.05.008 NIHMS794116 Article Corticotropin-Releasing Factor (CRF) in Ventromedial Prefrontal Cortex Mediates Avoidance of a Traumatic Stress-Paired Context Schreiber Allyson L. 1* Lu Yi-Ling 2* Baynes Brittni B. 1 Richardson Heather N. 3 Gilpin Nicholas W. 14 1 Department of Physiology, Louisiana State University Health Science Center, New Orleans, LA 70112 2 Neuroscience and Behavior Program, University of Massachusetts, Amherst, MA, 01003 3 Department of Psychological and Brain Sciences, University of Massachusetts, Amherst, MA, 01003 4 Neuroscience Center of Excellence, Louisiana State University Health Science Center, New Orleans, LA, 70112 Please send all correspondence to: Nicholas W. Gilpin, Ph.D., Department of Physiology, Room 7205, 1901 Perdido Street, New Orleans, LA 70112, Email: [email protected], Phone: 1.504.548.6192 * These authors contributed equally to this manuscript 11 6 2016 24 5 2016 2 2017 01 2 2018 113 Pt A 323330 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Post-traumatic stress disorder affects 7.7 million Americans. One diagnostic criterion is avoidance of stimuli related to the traumatic stress. Using a predator odor stress conditioned place avoidance (CPA) model, rats can be divided into groups based on stress reactivity, measured by avoidance of the odor-paired context. Avoider rats, which show high stress reactivity, exhibit avoidance of stress-paired context and escalated alcohol drinking. Here, we examined the role of corticotropin-releasing factor (CRF), a neuropeptide that promotes anxiety-like behavior, in mediating post-stress avoidance and escalated alcohol drinking. CRF is expressed in the medial prefrontal cortex (mPFC). The dorsal and ventral sub-regions of mPFC (dmPFC and vmPFC) have opposing roles in stress reactivity and alcohol drinking/seeking. We hypothesized that CRF-CRFR1 signaling in the vmPFC contributes to stress-induced avoidance and escalated alcohol self-administration. In Experiment 1, adult male Wistar rats were exposed to predator odor stress in a CPA paradigm, indexed for avoidance of odor-paired context, and brains processed for CRF-immunoreactive cell density in vmPFC and dmPFC. In Experiment 2, rats were tested for avoidance of a context repeatedly paired with intra-vmPFC CRF infusions. In Experiment 3, rats were stressed, indexed, and tested for the effects of intra-vmPFC CRFR1 antagonism on avoidance and alcohol self-administration. Post-stress, Avoiders exhibited higher CRF cell density in vmPFC, but not dmPFC. Intra-vmPFC CRF infusion produced conditioned place avoidance. Intra-vmPFC CRFR1 antagonism reversed avoidance of odor-paired context, but did not alter post-stress alcohol self-administration. These findings suggest that vmPFC CRF-CRFR1 signaling mediates avoidance of stimuli paired with traumatic stress. Post-traumatic stress disorder (PTSD) Avoidance Alcohol Self-Administration Medial Prefrontal Cortex (mPFC) Corticotropin-releasing factor (CRF) Anxiety 1. Introduction Traumatic stress disorders (e.g. post-traumatic stress disorder; PTSD) affect 7.7 million Americans (National Institutes of Health, 2015). Major diagnostic criteria for PTSD include re-experiencing the traumatic event, increased arousal, negative affective state, and persistent avoidance of stimuli that are related to the traumatic stress (American Psychiatric Association, 2013). Additionally, approximately 22–43% of PTSD patients meet diagnostic criteria for alcohol use disorder (AUD; Blanco et al., 2013). Persistent avoidance of stimuli that are related or perceived to be related to the traumatic stress is a major diagnostic criterion for PTSD in humans (American Psychiatric Association, 2013), and avoidance of trauma-related stimuli predicts subsequent onset of PTSD symptoms (Shin et al., 2015). Our lab uses a predator odor model of PTSD in which rats are divided into groups based on their stress reactivity, as indexed by avoidance of a predator odor-paired chamber (i.e., Avoiders and Non-Avoiders; Edwards et al., 2013). Using this model, our lab has previously shown that predator odor stress produces persistent avoidance of the predator odor-paired stimuli in some but not all animals (i.e., Avoiders), and that these animals escalate alcohol self-administration after stress (Edwards et al., 2013). Interestingly, these post-stress differences do not appear to be caused by differences in learning and memory. Non-Avoiders show no difference in appetitive conditioning and are able to acquire operant conditioning at the same rate as Avoiders prior to stress exposure (Edwards et al, 2013). Furthermore, these animals do not differ in baseline alcohol self-administration, anxiety-like behavior (Edwards et al., 2013), or nociception (Itoga et al., in press) prior to stress. Corticotropin-releasing factor (CRF) mediates affective regulation and arousal after stress (Heinrichs & Koob, 2004). In addition to activating the hypothalamic-pituitary-adrenal (HPA) axis, CRF acts as a neuromodulator in extra-hypothalamic areas of the brain including the prefrontal cortex (PFC) and the extended amygdala (Koob & Heinrichs, 1999). Brain CRF is dysregulated in psychiatric disorders including anxiety disorders, depression, and drug use disorders (George & Koob, 2010). Central administration of CRF increases anxiety-like behaviors and autonomic stress responses (Dunn & Berridge, 1990; Fisher, 1989), and produces conditioned place avoidance (Cador et al., 1992). CRF1 receptors (CRFR1) mediate anxiety-like behavior and hyperarousal in stressed rats (Heinrichs and Koob 2004; Roltsch et al., 2014). In addition, CRFR1s mediate binge-like ethanol consumption in mice (Lowery-Gionta et al., 2012) and post-stress and post-dependent escalations in alcohol drinking in mice and rats (Lowery et al., 2007; Sommer et al., 2008). Prior work from our lab showed that systemic antagonism of CRFR1s reduces hyperalgesia, hyperarousal, and alcohol self-administration in stressed rats (Roltsch et al., 2014). The medial PFC (mPFC) has a major role in cognitive and emotional regulation, and the mPFC sends highly organized efferent projections to sub-cortical regions important for affective and motivated behaviors. The mPFC exerts top-down regulation of subcortical sites that coordinate stress responses, including the extended amygdala, the paraventricular nucleus of the hypothalamus, and the dorsal raphe nucleus (Sotres-Bayon & Quirk, 2010; George & Koob 2010). Medial PFC projections to basolateral and central amygdala (CeA) are important for fear conditioning (Sotres-Bayon & Quirk, 2010) and are disrupted during alcohol withdrawal (George et al., 2012). The dmPFC and vmPFC have opposing roles in various behaviors. For example, the dmPFC promotes fear learning and fear expression through its connections with the basolateral amygdala (BLA), and the vmPFC promotes fear extinction through connections with the intercalated cells, CeA, and/or basomedial amygdala (Sotres-Bayon & Quirk, 2010; Adhikari et al., 2015). Similarly, the dmPFC inhibit hypothalamic stress responses (Radley et al., 2006), whereas the vmPFC facilitates hypothalamic stress responses (Radley et al 2006). Past work from our lab showed that Avoider rats have increased activation, as measured by ERK phosphorylation, of the vmPFC, but not the dmPFC, following a stress reminder (Edwards et al., 2013). Furthermore, human imaging studies demonstrate that lesions of the vmPFC significantly decrease PTSD symptoms, suggesting a critical role for vmPFC in development of PTSD following traumatic stress (Koenigs & Grafman, 2008). Recent studies demonstrate a role for CRF in the mPFC in mediating anxiety-like behavior (Miguel et al., 2014; Pentkowski et al., 2013). Specifically, low CRF doses infused into the mPFC increase anxiety-like behavior, while higher doses reduce anxiety-like behavior (Ohata & Shibasaki, 2011; Jaferi & Bhatnagar, 2007; Pentkowski et al., 2013). It is also worth noting that a recent study reports a role for CRF-CRFR1-PKA signaling in mPFC in acute stress-induced executive dysfunction (Uribe-Mariño et al., 2016). These studies examined the mPFC as a whole, but here we tested the effects of stress on CRF in vmPFC and dmPFC, due largely to their different afferents, efferents, and assigned functions (Heidbreder & Groenewegen, 2003). The purpose of the current studies was to determine the effect of predator odor stress on CRF cell counts in mPFC sub-regions, to assess the aversive properties of CRF in vmPFC, and to determine the role of vmPFC CRF-CRFR1 signaling in avoidance of predator odor-paired stimuli and post-stress alcohol self-administration. We hypothesized that 1) predator odor stress would increase the density of CRF-immunoreactive (CRF-ir) cells in vmPFC of Avoiders, 2) CRF infusions into vmPFC would produce conditioned place avoidance (CPA) in rats, 3) CRFR1 antagonism in vmPFC would abolish avoidance of a predator odor-paired context, and finally that 4) Avoiders would exhibit increased alcohol self-administration that would be reversed by CRFR1 antagonism in the vmPFC. 2. Methods and Materials 2.1 General Methods 2.1.1 Animals Specific-pathogen free adult male Wistar rats (N=152) (Charles River) weighing 225–250g at the time of arrival were housed in groups of two in a humidity- and temperature-controlled (22°C) vivarium on a 12 h light/12 h dark cycle (lights off at 8 a.m.). Animals had ad libitum access to food and water throughout experiments. Rats were acclimated for one week before start of experiments and were handled daily prior to initiation of surgical and experimental protocols. Behavioral tests occurred during the dark period. All procedures were approved by the Institutional Animal Care and Use Committee of the Louisiana State University Health Sciences Center and were in accordance with the National Institute of Health Guidelines. 2.1.2 Predator Odor Conditioned Place Avoidance To index rats for avoidance, animals underwent a 4-day CPA procedure, as previously described (Edwards et al., 2013). Briefly, on the first day, rats were allowed 5 min to explore two conditioning chambers with distinct tactile (circles vs. grid rod floor) and visual (stripes vs. circles) cues. Sessions were videotaped and scored by an observer. Rats were assigned to predator odor stress and non-stress groups that were counterbalanced for magnitude of baseline preference for one chamber versus the other. For rats in the stress group, an unbiased and counterbalanced procedure was used to determine which chamber would be paired with predator odor exposure in individual rats. On the second day, rats were placed in one chamber without odor (neutral environment) for 15 min. On the third day, rats were placed in the opposite context for 15 min with a urine-soaked sponge (bobcat urine) placed under the floor of the chamber (predator odor environment), or no odor for control animals. On the fourth day, rats were again allowed to explore the two conditioning chambers in a 5 min video-recorded post-test. Avoidance was calculated as a difference score between post-conditioning time spent in odor-paired context and pre-conditioning time spent in odor-paired context. Rats that displayed a >10 sec decrease in time spent in odor-paired chamber were classified as ‘Avoiders’. Rats that displayed a <10 sec decrease in time spent in the predator odor-paired chamber were classified as ‘Non-Avoiders’. The 10s cut-off for Avoiders and Non-Avoiders is consistent with previous publications from our lab (e.g., Edwards et al., 2013; Itoga et al., in press; Whitaker & Gilpin, 2015). By not using a median split, our criterion for separating those two groups remains identical across studies. This 10-s cut-off has been used to show differences between groups in alcohol drinking (Edwards et al., 2013), nociception (Itoga et al., in press), and the corticosterone response to stress (Whitaker & Gilpin, 2015). 2.1.3 Immunohistochemical Staining and Counting of CRF cells in mPFC sub-regions Rats (n=32) were deeply anesthetized with isoflurane, injected with chloral hydrate (35% 2ml), and then intracardially perfused with 4% paraformaldehyde/0.1M borate buffer, pH 9.5 as previously described (Richardson et al., 2006). Brains were post-fixed in the same fixative for 4 h at 4°C and submerged in the 20% sucrose/0.1M phosphate buffer, pH 7.4 for 48–72 hours before being snap-frozen in isopentane (2-methylbutane, Fisher Scientific) on dry ice. Brains were stored at −80°C, then coronally sectioned at 35 µm on a freezing microtome. Sections were stored in cryoprotectant (30% ethylene glycol, 30% sucrose and 1% polyvinyl pyrrolidone in 0.1M PBS) at −20°C until immunolabeling. Brain sections containing mPFC (from 3.72 mm to 2.52 mm relative to bregma) were sorted and labeled by rabbit anti-h/rCRF antiserum (1:5000, generously provided by Dr. Wylie Vale, Salk Institute) and biotin-conjugated goat anti-rabbit antiserum (1:200, Vector Laboratories) using free-floating immunohistochemistry procedure (Karanikas et al., 2013). CRF labeling signal was developed with 3,3’-diaminobenzidine (DAB) and nickel (Vector Laboratories), and sections were mounted on slides for microscopic analysis. Slides containing prefrontal sections were digitally scanned at high resolution (20×) under bright field illumination by Aperio ePathology (Leica Biosystems). Four coronal sections (spaced 350 µm apart) containing CRF immunoreactive (CRF-ir) cells and fibers were used to quantify CRF cell number in mPFC (Fig 2A). The dmPFC contained anterior cingulate cortex (AC) and dorsal prelimbic cortex (PrL); the vmPFC contained ventral PrL and infralimbic cortex (IL) (Figure 2A). Areas of interest were traced and measured using ImageScope software (Leica Biosystems), and all CRF-ir cells were counted within traced regions. Experimenters blind to the treatment group counted CRF-ir cells under 10× magnification of the slide images. The criteria for identifying CRF-ir cells were a clearly defined border of the soma and evidence of extended neurites. 2.1.4 CRF Conditioned Place Avoidance Rats underwent a conditioned place avoidance procedure similar to the one described by Cador et al., 1992. Briefly, on day 1, rats underwent a 5 min video-recorded pre-test to explore two conditioning chambers with distinct tactile (circles vs. grid rod floor) and visual (stripes vs. circles) cues. Rat assignment to CRF dose groups was counterbalanced for context preference. Within dose groups, CRF infusions were paired with context for each rat in an unbiased counterbalanced design. On days 2, 4, and 6, rats were infused with sterile saline and immediately placed in one chamber for 15 minutes (neutral chamber). On days 3, 5, and 7, rats were infused with CRF (0, 0.05, 0.25, or 0.5 µg/0.5µl per side) and immediately placed in the opposite chamber for 15 minutes (CRF chamber). It is important to note that each individual rat was repeatedly infused with the same CRF dose across conditioning days (i.e., between-subjects dose-response). On day 8, rats underwent a 5 min video-recorded post-test to explore the two conditioning chambers. Avoidance was calculated as a difference score between post-conditioning time spent in CRF-paired context and pre-conditioning time spent in CRF-paired context. 2.1.5 Operant Self-Administration Rats were trained to orally self-administer 10% w/v ethanol or water in a concurrent, two-lever, free-choice contingency that did not incorporate a sweet fading procedure, as previously described (Roltsch et al., 2014). Rats were first given a single 24-hr period of access to 10% w/v ethanol vs. water in the home cage to prevent neophobia upon presentation of ethanol in operant boxes. Rats were then given a single 15-h operant session to learn to press a single lever for water (right lever; FR1) in the presence of ad libitum food on floor of operant chamber. Rats were then allowed one 3-h two-lever operant session for 10% w/v ethanol (right lever; FR1) vs. water (left lever; FR1), one 2-h session, then one 1-h session, followed by daily 30-min session for ~15 days. All operant sessions after the initial 15-h session occurred in the absence of food. Rats with post-surgery baselines <10 alcohol presses in a 30-min session after vehicle infusion were excluded from further analysis. 2.1.6 Surgical Procedures On the day of surgery, rats were anesthetized with isoflourane and mounted in a stereotaxic frame (Stoelting). An incision was made, small bilateral holes were drilled into the skull, and stainless steel guide cannulae (26 gauge) lowered such that tips were 1mm dorsal to vmPFC (AP +2.8, ML ± 3.1, DV −3.9 from the skull at an angle of 30°). Guide cannulae were secured to the skull with metal screws and dental cement, the incision closed, and dummy cannulae inserted. Rats were allowed to recover and monitored daily for 1 week following surgery before initiating experimental procedures. Before sacrifice, animals were infused with Evans Blue to verify cannula placement (Figure3B and 3D; Paxinos and Watson, 2005). Animals whose surgeries did not hit the target region bilaterally were excluded from analysis. 2.1.7 Infusion Procedure Animals underwent sham infusions on days preceding infusion to acclimate them to the procedure. A Harvard instruments microinfusion pump was used for all infusions. Infusions were delivered via polyethylene tubing (PE20) connected to a 10 µl Hamilton syringe. Infusions were administered bilaterally at a rate of 0.2 µl/min over the course of 2.5 min via injectors (33ga, stainless steel) that extended 1 mm beyond the tip of the guide cannulae. Injection cannulae were left in the guide cannulae for an additional 1 min to allow for diffusion. 2.1.8 Drugs The CRFR1 antagonist R121919 (generously supplied by Neurocrine, Inc.) was solubilized first in 1M HCl (10% final volume) then diluted into 2-hydroxypropyl-cyclodextrin (HBC; Sigma-Aldrich, 20% wt/vol final concentration in distilled water) and back-titrated with NaOH to pH 4.5. CRF (Sigma Aldrich, St. Louis, OM) was solubilized in sterile saline. 2.2 Experimental Protocols 2.2.1 Experiment 1. Effect of predator odor stress on CRF-ir cells in vmPFC and dmPFC Rats were exposed to predator odor stress, indexed for avoidance, and separated into Control (n=10), Non-Avoider (n=12), and Avoider (n=9). Nine days post-stress, rats were deeply anesthetized with isoflurane, then injected with chloral hydrate (35% 2ml), and perfused intracardially. Brains were processed for CRF-ir cell density by immunohistochemical labeling as described in General Methods. 2.2.2 Experiment 2. Intra-vmPFC CRF-induced conditioned place avoidance Rats (N=32) were implanted with bilateral cannulae targeted 1 mm dorsal to vmPFC. After one week of recovery, rats underwent the CRF CPA procedure described in General Methods. 2.2.3 Experiment 3. Effect of intra-vmPFC CRFR1 antagonism on post-stress avoidance and alcohol drinking Rats were tested for baseline ethanol self-administration. Upon stabilization of operant responding, rats were implanted with bilateral cannulae targeted 1 mm dorsal to vmPFC. After one week of recovery, rats were re-tested for baseline self-administration with vehicle infusions, then exposed to predator odor stress, indexed for avoidance, and separated into Control (n=12), Non-Avoiders (n=15), and Avoiders (n=19). On post-stress days 2, 5, 8, and 11, rats were allowed to self-administer alcohol for 30 minutes, as described above. On post-stress day 12, rats were again tested for avoidance of the predator-odor paired chamber. Five minutes prior to each test, rats received intra-vmPFC infusion of vehicle (20% HBC, 0.5 µl) or R121919 (0.25 µg/0.5 µl); each rat received the same treatment (vehicle or R121919) throughout the experiment. 2.3 Statistical Analysis Data are shown as mean ± SEM with the number of animals in each experiment indicated in Figure 1. In Experiment 1, CRF-ir cell density was analyzed with a mixed-design ANOVA where stress history was a between factor and bregma was a within-subjects factor. Also, overall mean CRF-ir cell number was analyzed with stress as the single between-subjects factor. In Experiment 2, avoidance of CRF-paired chamber was analyzed with a one-way ANOVA where drug dose was the treatment factor, and with trend analysis to determine whether the linear (change in dependent variable as a function of change in the independent variable) and/or quadratic (rate of change in the dependent variable as a function of change in the independent variable) components of responding were significantly affected by CRF dose. CRFR1 antagonist effects on avoidance were analyzed with two-way ANOVA where stress history and R121919 drug dose were the treatment factors. Differences in ethanol self-administration over time were analyzed with a three-way RM ANOVA where stress history and drug dose were the between-subject factors and time was the within-subjects factor. Post hoc analysis with the Tukey’s HSD Test was used when appropriate. Statistical significance was set as p≤0.05. 3. Results 3.1 Predator odor stress increases CRF cell density in the vmPFC but not dmPFC of Avoiders In Experiment 1, Avoiders exhibited higher density of CRF-ir cells relative to Non-Avoiders and unstressed Controls 9 days post-stress in the vmPFC (F[2,28]=5.698, p=0.008), but not dmPFC (F[2,28]=1.550, p=0.230) (Figure 2D, 2B). CRF-ir cell number in vmPFC and dmPFC was positively correlated with avoidance of the predator odor-paired chamber 24 h post-stress (dmPFC, R2=0.18, p=0.05; vmPFC, R2=0.35, p=0.004) (Figure 2C, 2E). Because there was no significant change in CRF immunoreactivity in the dmPFC of stressed rats, behavioral pharmacology studies were only conducted in the vmPFC. 3.2 CRF infusions into vmPFC produce conditioned place avoidance In Experiment 2, rats underwent a CPA test using intra-vmPFC CRF infusion as the unconditioned stimulus. Compared to vehicle-infused rats, CRF-infused rats spent less time in the drug-paired chamber (F[3,28]=2.862; p=0.054; Figure 3A). Quadratic trend analysis revealed that the amount of time spent avoiding the CRF-paired chamber was related to dose of CRF, with lower doses producing more avoidance than higher doses (p=0.054). 3.3 CRFR1 antagonism in vmPFC reduces avoidance of stress-paired context In Experiment 3, rats underwent a CPA procedure 1 d post-stress for indexing of avoidance behavior and again 12 d post-stress with or without antagonism of vmPFC CRFR1. A 1-way ANOVA of Day 1 avoidance data showed that Avoiders spent significantly more time avoiding the predator odor paired chamber compared to Non-Avoiders and Controls (Controls −5.94 ± 11.36 s; Non-Avoiders 25.24 ± 9.5 s; Avoiders −39.25 ± 6.21 s compared to pre-conditioning baseline) (F[2,43]=16.45; p<0.01). A separate, 2-way ANOVA of Day 12 avoidance data revealed a main effect of stress on the amount of time spent avoiding the predator odor paired chamber. On Day 12 post-stress, Avoiders spent significantly less time than Non-Avoiders in the predator odor-paired chamber (F[1,24]=18.06; p<0.01) (Figure 3C), as would be expected based on the model and past reports from our lab (Whitaker & Gilpin, 2015). Based on the a priori hypothesis that CRFR1 antagonism would reduce avoidance, we analyzed day 12 (i.e., after R121919 treatment) avoidance scores in Avoider rats using a two-samples t-test where R121919 dose was the between-subjects factor. This analysis revealed that R121919 significantly reduced avoidance in Avoiders relative to vehicle-treated controls (t(17)=2.09; p=0.026) (Fig 3C). A separate two-samples t-test indicated that R121919 did not affect avoidance behavior in Non-Avoiders 12 days post-stress (t(10)=0.4994; p=0.523). 3.4 CRFR1 antagonism does not affect alcohol self-administration In Experiment 3, rats were trained to self-administer 10% ethanol. The pre-surgery baseline was 42.08±1.76 presses per 30 min session. After surgery, mean ethanol responding per 30-min session was 27.92±1.62 presses. This decrease in ethanol self-administration was due either to surgery, the introduction of pre-operant sham infusions, or both. Once post-surgery operant responding stabilized, rats were stressed, then were allowed to self-administer alcohol on days 2, 5, 8, and 11 post-stress after infusion of either vehicle or R121919 into the vmPFC (Figure 4). When operant response data were analyzed as percent of baseline for each rat, (Figure 4), there was a trend for Avoiders to respond more for alcohol than the other two groups after predator odor exposure (F[2,40]=2.99; p=0.06). A 3-way RM ANOVA revealed no significant increases in raw operant alcohol lever presses across days after stress. There was also no significant main effect of R121919 treatment on operant alcohol self-administration (F[1,40]=0.08; p=0.77), nor was there a significant stress × drug interaction effect (F[2,40]=1.28; p=0.29). 4. Discussion In this study, we tested the role of CRF in the vmPFC in mediating predator odor stress-induced avoidance and escalation of alcohol drinking. We demonstrated that 1) after predator odor stress, Avoiders have higher CRF cell density in vmPFC, but not dmPFC, that is positively correlated with avoidance of the predator odor-paired chamber 24 h after stress, 2) CRF infused into the vmPFC produces conditioned avoidance of a CRF-paired context, 3) CRFR1 antagonism in vmPFC reduces avoidance of a stress-paired context, and 4) CRFR1 antagonism in vmPFC does not reduce post-stress alcohol drinking. The present study demonstrates a role for CRF in the vmPFC in stress-induced avoidance behavior. After exposure to predator odor stress, Avoiders exhibited higher CRF cell density 9 days post-stress. We assume that the difference CRF-ir cell counts are not pre-existing, but are instead caused by exposure to the predator odor. It is unlikely that random assignment of animals into stress and controls groups would select for animals with different numbers of CRF-ir cells in vmPFC. Moreover, the spread of the data in Avoiders and Non-Avoiders extends outside the spread of data in non-stressed Control animals. Increases in number of CRF-ir cells within the vmPFC were correlated with more time spent avoiding the predator odor-paired chamber 24 hours post-stress, suggesting that more avoidance is related to higher CRF cell density in the vmPFC. We did not measure CRF release into the synapse, but it is notable that chronic CRFR1 antagonism in the vmPFC reduced avoidance behavior in Avoiders (Figure 3C). Overall, this suggests that avoidance of a stress-paired context is mediated, at least in part, by CRF signaling in vmPFC. This result seems to agree with previously published work from our group showing that exposure to a traumatic stress reminder increases ERK phosphorylation, a marker and mediator of neuronal activity downstream of CRFR1 signaling (Meng et al., 2011), in vmPFC of Avoiders compared to Non-Avoiders (Edwards et al., 2013). Here, we used a conditioned place avoidance paradigm to show that CRF in vmPFC is aversive. We demonstrate that low CRF doses (0.05 µg/side) produce conditioned avoidance of a CRF-paired chamber compared to vehicle, while higher doses of CRF (0.25 and 0.5 µg/side) produced less avoidance. According to the literature, different doses of CRF in the mPFC have different effects on anxiety-like behavior in rodents. Specifically, low doses (0.02 – 0.05 µg/side) produce anxiety-like behavior (Jaferi & Bhatnagar, 2007; Ohata & Shibasaki, 2011), while high doses (0.2 – 1.0 µg/side) of CRF infused into the mPFC are anxiolytic (Ohata & Shibasaki, 2011; Pentkowski et al., 2013). In our study, the lowest CRF dose (0.05 µg/side), which matches anxiogenic CRF doses in the literature, produced the strongest avoidance of the CRF-paired chamber. Both the anxiogenic and anxiolytic effects of vmPFC CRF are thought to be mediated by activation of CRFR1 on glutamatergic pyramidal cells, because there are no CRFR2 in the vmPFC (Chalmers et al., 1995). It is possible that low and high CRF doses initiate different CRFR1 signaling cascades to produce opposite behavioral responses: for example, it has been suggested that the anxiogenic actions of low CRF doses may be mediated by protein kinase A signaling (Miguel et al., 2014), whereas the anxiolytic-like actions of high CRF doses may be mediated by protein kinase C signaling (Tan et al., 2004). Interestingly, CRFR1 antagonism in vmPFC reduced avoidance behavior in Avoiders, suggesting that physiological levels of CRF in vmPFC promote avoidance, and perhaps anxiety-like behavior. Further studies will determine the cellular mechanisms whereby CRF-CRFR1 signaling in vmPFC produces aversion. The vmPFC projects to sub-cortical brain regions that control anxiety, fear, and drug addiction. The vmPFC facilitates hypothalamic stress responses and may regulate long-lasting anxiety through its indirect projections to the paraventricular nucleus of the hypothalamus and direct projections to bed nucleus of the stria terminalis, respectively (Radley et al., 2006; Motzkin et al., 2015). The vmPFC also sends projections to the basomedial amygdala and/or the intercalated cells of the amygdala to mediate fear extinction (Adhikari et al., 2015; Keifer et al., 2015; Sotres-Bayon and Quirk, 2010). In humans, damage to the vmPFC leads to disinhibition of the amygdala, which has been hypothesized to contribute to PTSD symptoms (Motzkin et al., 2015). These studies all support a role for dysregulation of the vmPFC in traumatic stress and anxiety disorders. We also tested whether CRF-CRFR1 signaling in the vmPFC mediates escalation of alcohol drinking after stress. After stress, there was a trend toward stress group differences in alcohol self-administration (p=0.06). Previously, we demonstrated that Avoider rats self-administer more alcohol than Non-Avoiders and unstressed Controls (Edwards et al., 2013). The current study may have failed to reach significance due to the stress caused by the drug infusion procedure 5 minutes before alcohol self-administration, despite the inclusion of post-surgery sham and vehicle infusions to acclimate animals to the infusion procedure. CRFR1 antagonism in vmPFC did not affect alcohol drinking. Although our data do not demonstrate a role for vmPFC CRF-CRFR1 signaling in post-stress alcohol drinking, the vmPFC is important for extinction and relapse of alcohol-seeking behavior. For example, the vmPFC inhibits drug-seeking following extinction (Van den Oever et al., 2010; Peters et al., 2008), and enhancement of mGluR5 signaling in the infralimbic cortex facilitates extinction of alcohol responding in rats (Gass et al., 2014). Pfarr et al. (2015) identified a specific vmPFC neuronal ensemble critical for inhibiting alcohol-seeking behavior after cue-induced reinstatement, whereas complete vmPFC inactivation does not affect extinction or cue-induced reinstatement of alcohol-seeking behavior (Willcocks & McNally, 2012; Pfarr et al., 2015). Our data suggest that CRF-CRFR1 signaling in vmPFC is not a critical mediator of non-extinguished alcohol self-administration after stress. Overall, this study shows that stress alters CRF cell density in the vmPFC, but not dmPFC, of some but not all stressed animals. We also show that CRF in vmPFC is aversive and that CRF-CRFR1 signaling in vmPFC mediates avoidance of a predator odor-paired chamber but not escalation of alcohol self-administration after traumatic stress. Funding This work was supported by NIH grants AA023696 (ALS), AA018400 (NWG), AA023305 (NWG), and AA021013 (HNR). NWG is a consultant for Glauser Life Sciences. We thank Dimitri Grigoriadis of Neurocrine for the generous donation of R121919 and Dr. Wylie Vale at the Salk Institute for the generous donation of rabit anti-h/r CRF antiserum. We also thank Lynn Bengston, Aditi Dave, and Frank Jackson for their technical assistance with CRF immunohistochemistry. Figure 1 Timelines for Experiments 1–3 and Corresponding Figures illustrates the experimental timeline for Experiments 1–4 described below. Explanation of groups, treatments, experimental timelines, and analyses are contained in the Materials and Methods section. Experiments 1 and 3 used conditioned place aversion (CPA) paradigm to index the animals as Controls, Non-Avoiders, or Avoiders. The purpose of Experiment 1 was to generate tissue for ex vivo processing and analysis (Figure 2). Experiments 2 and 3 were behavioral experiments in which animals received brain site-specific pharmacology (Figures 3 and 4). Figure 2 Predator Odor Stress Increases CRF+ Cell Density in vmPFC of Avoiders A. Representative images of examined sections (B1–B4) and CRF-ir cells in the mPFC. B. CRF-ir cell density in the dmPFC in 4 sections was not significantly changed after predator odor stress. Insert: Average CRF-ir cell density in the dmPFC. C. dmPFC CRF-ir cell density positively correlated with avoidance 1 day post-stress (R2=0.18, p=0.05). C. Avoiders had significantly higher CRF cell density in 4 sections than Non-Avoiders and Controls in the cmPFC 9 days post stress (p=0.008). Insert: Average CRF-ir cell density across 4 sections in the vmPFC. E. CRF cell density positively correlated with avoidance in the vmPFC (R2=0.35, p<0.01). Data are represented as Mean ± SEM. Control, white; Non-Avoiders, grey; Avoiders, black. Figure 3 CRF in vmPFC mediates avoidance bahavior A. Intra-vmPFC CRF infusions produced avoidance of a CRF-paired context 1 day post-containing (p=0.054). Quadratic trend analysis revealed time avoiding the chamber was related to the dose of CRF (p=0.054) with lower doses producing more avoidance than higher doses. B. cannula placements for Experment 2 (White Dots – Misses; Black Dots – Hits; Images modified from Paxinos and Watson, 2005). Insert: Representative image of histological verification by Evan's Blue Dye. C. Avoiders spend significantly less time than Non-Avoiders in the predator-odor paired chambar relative to the pre-conditioning baseline 12 days post-stress (p<0.05). Antagonism of CRF1R receptors reduced the amount of time spent avoiding the predator odor-paired chamber in Avoider (p<0.05). D. Cannula placements for Experiment 3. (White Dotes – Misses; Black Dots – Hits; Images modified from Paxinos and Watson, 2005). Data are represented as Mean ± SEM. Non-Avoiders, grey; Avoiders, black. Figure 4 Post-stress alcohol self administration does not differ across groups Average percent of baseline self-administration per group across drinking days post-stress. After stress, there were no significant differences in alcohol self-administration between stress groups (p=0.06). Antagonism of CRF1Rs with R121919 in the vmPRC did not have a significant effect on alcohol self-administration (p=0.77). Data are represented as Mean ± SEM. Control, white color; Non-Avoider, grey color; Avoider, black color Disclosure All other authors report no biomedical financial interests of potential conflicts of interest. References 1 American Psychiatric Association Diagnostic and statistical manual of mental disorders: DSM-5 2013 Washington, D.C American Psychiatric Association 2 Adhikari A Lerner TN Finkelstein J Pak S Jennings JH Davidson TJ Ferenczi E Gunaydin LA Mirzabekov JJ Ye L Kim SY Lei A Deisseroth K Basomedial amygdala mediates top-down control of anxiety and fear Nature 2015 527 7577 179 185 26536109 3 Blanco C Xu Y Brady K Pérez-Fuentes G Okuda M Wang S Comorbidity of posttraumatic stress disorder with alcohol dependence among US adults: Results from National Epidemiological Survey on Alcohol and Related Conditions Drug Alcohol Depend 2013 132 630 638 23702490 4 Cador M Ahmed SH Koob GF Le Moal M Stinus L Corticotropin-releasing factor induced a place aversion independent of its neuroendocrine role Brain Research 1992 597 304 309 1473001 5 Chalmers DT Lovenberg TW De Souza EB Localization of Novel Cortioctropin-Releasing Factor Receptor (CRF2) mRNA Expression to Specific Subcortical Nuclei in Rat Brain: Comparison with CRF1 Receptor mRNA Expression J Neurosci 1995 15 10 6340 6350 7472399 6 Dunn AJ Berridge CW Physiological and behavioral responses to corticotropin-releasing factor administration: is CRF a mediator of anxiety of stress responess? Brain Research Reviews 1990 15 71 100 1980834 7 Edwards S Baynes BB Carmichael CY Zamora-Martinez ER Barrus M Koob GF Gilpin NW Traumatic stress reactivity promotes excessive alcohol drinking and alters the balance of prefrontal cortex-amygdala activity Transl Psychiatry 2013 3 e296 23982628 8 Fisher LA Corticotropin-releasing factor: endocrine and autonomic integration of responses to stress TIPS Reviews 1989 10 189 193 9 Gass JT Trantham-Davidson H Kassab AS Glan WB Jr Olive MF Chander LJ Enhancement of Extinction Learning Attenuates Ethanol-Seeking Behavior and Alters Plasticity in Prefrontal Cortex J Neurosci 2014 34 22 7532 7574 10 George O Koob GF Individual differences in prefrontal cortex function and the transition from drug use to drug dependence Neurosci Biobehav Rev 2010 35 2 232 247 20493211 11 Heidbreder CA Greonewegen HJ The medial prefrontal cortex in the rat: evidence for a dorso-ventral distinction based upon functional and anatomic characteristics Neurosci Biobehav Rev 2003 27 555 579 14599436 12 Heinrichs SC Koob GF Corticotropin-Releasing Factor in Brain: A Role in Activation, Arousal, and Affect Regulation J Pharmacol Exp Ther 2004 311 2 427 440 15297468 13 Itoga CA Roltsch-Hellard EA Whitaker AM Lu YL Schreiber AL Baynes BB Baiamonte BA Richardson HN Gilpin NW Traumatic Stress Promotes Hyperalgesia via Corticotropin-Releasing Factor-1 Receptor (CRFR1) Signaling in Central Amygdala Neuropsychopharmacology 2016 In press 14 Jaferia A Bhatnagar S Corticotropin-releasing hormone receptors in the medial prefrtonal cortex regulate hypothalamic-pituitary-adrenal activity and anxiety-related behavior regardless of prior stress experience Brain Research 2007 1186 212 223 18001698 15 Karanikas CA Lu YL Richarson HN Adolescent drinking targets corticotropin-releasing factor peptide-labeled cells in the central amygdala of male and female rats Neuroscience 2013 249 98 105 23628776 16 Keifer OP Hurt RC Ressler KJ Marvar PJ The Physiology of Fear: Reconceptualizing the Role of the Central Amygdala in Fear Learning Physiology 2015 30 389 401 26328883 17 Koenigs M Huey ED Raymont V Cheon B Solomon J Waserman EM Grafman J Focal brain damage protects against post-traumatic stress disorder in combat veterans Nature Neuroscience 2008 11 232 237 18157125 18 Koob GF Heinrichs SC A role for corticotropin releasing factor and urocortin in behavioral responses to stressors Brain Res 1999 848 141 152 10612706 19 Lowery EG Sparrow AM Breese GR Knapp DJ Thiele TE The CRFR-1 receptor antagonist, CP-154,526, attenuates stress-induced increase in ethanol consumption by BALB/cJ mice Alcohol Clin Exp Res 2007 32 240 248 18162074 20 Lowery-Gionta EG Navarro M Li C Pleil KE Rinker JA Cox BR Sprow GM Kash TL Thiele TE Corticotropin Releasing Factor Signaling in the Central Amygdala is Recruited during Binge-like Ethanol Consumption in C57BL/6J Mice J Neurosci 2012 32 10 3405 3413 22399763 21 Meng QY Chen XN Tong DL Zhou JN Stress and glucocorticoids regulated corticotropin releasing factor in prefrontal cortex Mol Cell Endocrinol 2011 342 54 63 21664419 22 Miguel TT Gomes KS Nunes-de-Souza RL Tonic modulation of anxiety-like behavior by corticotropin-releasing factor (CRF) type 1 receptor (CRFR) within the medial prefrontal cortex (mPFC) in male mice: Role of protein kinase A (PKA) Hormones and Behavior 2014 66 247 256 24848364 23 Motzkin JC Philippi CL Wolf RC Baskaya MK Koenigs M Ventromedial Prefrontal Cortex is Critical for the Regulation of Amygdala Activity in Humans Biol Psychiatry 2015 77 276 284 24673881 24 Motzkin JC Philippi CL Oler JA Kalin NH Baskaya MK Koenigs M Ventromedial prefrontal cortex alters resting blood flow to the bed nucleus of stria terminalis Cortex 2015 64 281 288 25569763 25 http://www.nimh.nih.gov 26 Ohata H Shibasaki T Microinjection of Different Doses of Corticotropin-Releasing Factor into the Medial Prefrontal Cortex Produces Effects Opposing Anxiety-related Behavior in Rats J Nippon Med Sch 2011 78 5 286 292 22041875 27 Ovari J Leri F Inactivation of ventromedial prefrontal cortex mimics re-emergence of heroin seeking caused by heroin reconditioning Neurosci Lett 2008 444 1 52 55 18706971 28 Paxinos G Watson C The Rat Brain in Stereotaxic Coordinates 2005 San Diego, CA, USA Elsevier Academic Press 29 Pentkowski NS Tovote P Zavala AR Litvin Y Blanchard DC Spiess J Blanchard RJ Cortagine infused into the medial prefrontal cortex attenuates predator-induced defensive behaviors and Fos protein production in selective nuclei of the amygdala in male CD1 mice Hormones and Behavior 2013 64 519 526 23845323 30 Peters J LaLumiere RT Kalivas PW Infralimbic prefrontal cortex is responsible for inhibiting cocaine seeking in extinguished rats J Neurosci 2008 28 23 6046 6053 18524910 31 Pfarr S Meinhardt MW Klee ML Hansson AC Vengeliene V Schönig K Bartsch D Hope BT Spanagel R Sommer WH Losing Control: Excessive Alcohol Seeking after Selective Inactivation of Cue-Responsive Neurons in the Infralimbic Cortex J Neurosci 2015 35 30 10750 10761 26224858 32 Radley JJ Arias CM Sawchenko PE Regional Differentiation of the Medial Prefrontal Cortex in Regulating Adaptive Responses to Acute Emotional Stress J Neurosci 2006 26 50 12967 12976 17167086 33 Richardson HN Zorilla EP Mandyam CD Rivier CL Exposure to repetitive verus varied stress during prenatal development generates two distinct anxiogenic and neuroendocrine profiles in adulthood Endocrinology 2006 147 5 2506 2517 16455779 34 Roltsch EA Bayes BB Mayeux JP Whitaker AM Baiamonte BA Gilpin NW Predator odor stress alters corticotropin-releasing factor-1 receptor (CRF1R)-dependent behaviors in rats Neuropharmacology 2014 79 83 89 24269607 35 Shin KM Chang HY Cho SM Kim NH Kim KA Chung YK Avoidance symptoms and delayed verbal memory are associated with post-traumatic stress symptoms in female victims of sexual violence J Affect Disord 2015 184 145 148 26093033 36 Sommer WH Rimondini R Hansson AC Hipskind PA Gehlert DR Barr CS Heilig MA Upregulation of Voluntary Alcohol Intake, Behavioral Sensitivity to Stress, and Amygdala Crhr1 Expression Following a History of Dependence Biol Pyschiatry 2008 63 139 145 37 Sotres-Bayon F Quirk GJ Prefrontal control of fear: more than just extinction Current Opinion in Neurobiology 2010 20 231 235 20303254 38 Tan H Zhong P Yan Z Corticotropin-releasing factor and acute stress prolongs serotonergic regulation of GABA transmission in prefrontal cortical pyramidal neurons J Neurosci 2004 24 21 5000 5008 15163692 39 Uribe-Mariño A Gassen NC Wiesbeck MF Balsevich G Santarelli S Solfrank B Dournes C Fries GR Masna M Labermeier C Wang XD Hafner K Schmid B Rein T Chen A Deussing JM Schimdt MV Prefrontal cortex corticotropin-releasing hormone receptor 1 conveys acute stress-induced executive dysfunction Biol Psychiatry 2016 http://dx.doi.org/10.1016/j.biopsych.2016.03.2016 40 Van den Oever MC Spijker S Smit AB De Vries TJ Prefrontal cortex plasticity mechanisms in drug seeking and relapse Neurosci Biobehav Rev 2010 35 276 284 19932711 41 Whitaker AM Gilpin NW Blunted hypothalamo-pituitary adrenal axis response to predator odor predict high stress reactivity Physiology and Behavior 2015 147 16 22 25824191 42 Willcocks AL McNally GP The role of medial prefrontal cortex in extinction and reinstatement of alcohol-seeking in rats Eur J Neurosci 2013 37 2 259 268 23106416
PMC005xxxxxx/PMC5122475.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 9508565 8783 Oral Dis Oral Dis Oral diseases 1354-523X 1601-0825 27219464 5122475 10.1111/odi.12509 NIHMS790093 Article A practical guide to the oral microbiome and its relation to health and disease Krishnan K 12 Chen T 1 Paster BJ 13* 1 The Forsyth Institute, Department of Microbiology, Cambridge, Massachusetts, USA 2 New England BioLabs, Ipswich, Massachusetts, USA 3 Department of Oral Medicine, Infection & Immunity, Harvard School of Dental Medicine, Boston, Massachusetts, USA * Corrspondence: Bruce J. Paster, The Forsyth Institute, 245 First St., Cambridge, Massachusetts 02142, USA. Tel:+1 617 892 8288, Fax: +1 617 892 8432, [email protected] 29 5 2016 04 7 2016 4 2017 01 4 2018 23 3 276286 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. The oral microbiome is incredibly complex with the average adult harboring about 50 to 100 billion bacteria in the oral cavity, which represent about 200 predominant bacterial species. Collectively, there are approximately 700 predominant taxa of which less than 1/3 still have not yet been grown in vitro. Compared to other body sites, the oral microbiome is unique and readily accessible. There is extensive literature available describing the oral microbiome and discussing the roles that bacteria may play in oral health and disease. However, the purpose of this review is not to rehash these detailed studies but rather to educate the reader with understanding the essence of the oral microbiome, namely that there are abundant bacteria in numbers and types, that there are molecular methods to rapidly determine bacterial associations, that there is site-specificity for colonization of the host, that there are specific associations with oral health and disease, that oral bacteria may serve as biomarkers for non-oral diseases, and that oral microbial profiles may have potential use to assess disease risk. oral microbiome 16S rDNA bacterial associations Next Generation Sequencing Introduction Although there have been wildly diverse estimates of the total number of oral bacterial taxa, it is generally accepted that, collectively speaking, there are 687 predominant species in the oral cavity (www.homd.org) (Dewhirst et al., 2010). These estimations are based on years of traditional identification of bacteria from cultural and phenotypic characterization studies, but mostly from identification of bacteria from culture-independent molecular studies using 16S rRNA gene comparative analyses (Paster et al., 2001, Aas et al., 2005, Dewhirst et al., 2010, Paster et al., 2006, Aas et al., 2008). As of May 2016, 31% of oral bacterial taxa have not been grown in vitro (www.homd.org). These not-yet-cultivated taxa are typically referred to as phylotypes, or colloquially as “uncultivables.” About 400 to 500 oral taxa have been detected in the subgingival crevice alone (Aas et al., 2005, Paster et al., 2001). The remaining taxa are distributed on the many oral habitats including different areas on the tongue, tooth surface, buccal mucosa, tonsils, soft and hard palate, and lip vestibule (Aas et al., 2005, Human Microbiome Project, 2012a, Human Microbiome Project, 2012c, Segata et al., 2012). The salivary microbiome would essentially be comprised of a mixture of microbes sloughed off from all sites. Although there is considerable overlap of species detected in all oral sites, such as certain species of Streptococcus, Gemella, Granulicatella, Neisseria, and Prevotella, there is often site-specificity. For example, species of Rothia typically colonize the tongue or tooth surfaces, Simonsiella colonizes only the hard palate, Streptococcus salivarius mainly colonizes the tongue and treponemes are typically restricted to the subgingival crevice (Aas et al., 2005, Kazor et al., 2003, Mager et al., 2003, Segata et al., 2012). It is well known that specific bacterial taxa that colonize the oral cavity are associated with oral health and oral diseases or afflictions, such as dental caries, periodontal diseases, endodontic lesions, dry socket, halitosis, and odontogenic infections (Aas et al., 2008, Aas et al., 2005, Dewhirst et al., 2010, Johansson et al. 2016; Mager et al., 2003, Paster et al., 2001, Paster et al., 2006, Socransky et al., 1998, Socransky & Haffajee, 2005). Furthermore, oral bacteria may be linked or serve as biomarkers for certain systemic diseases, such as pancreatic cancer (Farrell et al., 2012), diabetes type II (Demmer et al, 2015), pediatric Crohn’s Disease (Docktor et al., 2012), heart disease (Leishman et al., 2010), and low weight, preterm birth (Shira Davenport, 2010). However, it is yet to be established if there is a causal relationship between the oral microbiome and these systemic disorders. The Human Microbiome Project (HMP) Our ability to study the human microbiome has been greatly improved by the advances made in sequencing technologies and recent developments in bioinformatics. These advances have led to a plethora of genomic and metagenomic studies investigating the role of microbes in several different ecosystems (Gilbert & Dupont, 2011). Established in 2008, the HMP aimed to determine the microbiomes from 242 healthy human subjects from sites including the oral cavity (7 sites), nasal cavity, skin, gastrointestinal tract and urogenital tract. The data obtained from sequencing was used for taxonomic assignment and is also available through the HMP Data Analysis and Coordination Center data browser. This enables the advance of research relating to the human microbiome by acting as a community resource that is widely accessible. The establishment of such an effort led to the development of a variety of new protocols including methods for laboratory and sequence processing, and analysis of 16S rDNA and whole genome shotgun sequences and profiles of the microbiome (Human Microbiome Project, 2012a). Results from the HMP analyses indicated that repertoire and abundance of microbiota found on individuals varies greatly depending on multiple factors, with ethnic/racial background having one of the strongest associations to microbes with clinical metadata (Human Microbiome Project, 2012c). Such studies provided insights into what constitutes the normal microbiota of each organ or mucosa in the body, enabling a better understanding of how they impact human health. As of April 2016, over 1,300 reference strains isolated from the human body were sequenced and the data publicly available for researchers. The Human Oral Microbe Database (HOMD) The purpose of HOMD (www.homd.org) is to provide the scientific community with comprehensive information on the approximately 700 predominant bacterial species that inhabit the human oral cavity (Dewhirst et al., 2010); www.homd.org. This curated 16S rDNA database provides a provisional naming scheme for currently unnamed species or phylotypes. The HOMD also links sequence data of 701 oral taxa with phenotypic, phylogenetic, clinical, and bibliographic information. A phylogenetic tree of 118 of the most predominant and other key taxa, i.e., identified in 16S rRNA cloning studies, is shown in Figure 1. Note that there are many well-known oral species, e.g., including species Prevotella, Porphyromonas, Treponema, Tannerella, Fusobacterium, and Streptococcus, as well as perhaps lessor known species, e.g., phylotypes of members of the phyla SR1, GN02 and TM7, Fretibacterium, Solobacterium, and Abiotrophia. Many are associated with oral health and disease and will be discussed in more detail below. Phylogenetic trees of members of each bacterial phylum or family can be downloaded from the HOMD website. As part of HOMD, HMP and other sequencing projects, genome sequences are available for approximately 400 oral bacterial taxa, which represent 58% of the known oral species. BLAST tools are available to rapidly determine oral bacterial identification from 16S rDNA sequences. Easy to use tools for viewing all publically available oral bacterial genomes are also offered on the HOMD site. CORE, another phylogenetically-curated 16S rDNA database of the oral microbiome, is also available and can be used identify bacterial taxa from large next generation sequence (NGS) 16S rDNA datasets (Griffen et al., 2011). Bacterial-bacterial and host-bacterial interactions It has long been known that oral bacteria exhibit specificity for their respective colonization sites and to each other, directed by adhesin-receptor binding (Kolenbrander, 2000). Thus, adhesins on bacterial cells bind to receptors on epithelial cells or to other bacteria, including pili, auto-transporters, and extracellular matrix-binding proteins (Nobbs et al., 2011). Some receptors are derived from salivary components, such as proline- or serine-rich-proteins, that undergo conformational change when they are adsorbed onto surfaces such as the tooth surface. Consequently, bacteria do not simply bind or randomly pile on to oral surfaces or other bacteria—there is a specific interaction with a strong affinity. Such specificity can be readily seen in situ by using Combinatorial Labeling and Spectral Imaging FISH (CLASI-FISH) which utilizes multiple taxa-specific, fluorescently labeled probes. (Mark Welch et al., 2016, Valm et al., 2011). Figure 2 illustrates a specific spatial organization of bacterial taxa within dental plaque and that the bacteria do not randomly aggregate. Based on these CLASI-FISH data, the authors were able to propose a model for plaque microbiome development integrating known metabolic, adherence, and environmental information. Thus, we can deduce functional traits of the specific members of the consortium, e.g., anaerobic species are at the center with facultatively anaerobic or aerobic species are at the edge. Tools to define oral microbiome Culture techniques Historically, bacterial taxa were identified using culture-dependent methodologies such as microscopy, biochemical and other phenotypic tests, growth conditions, sugar utilization, and antibiotic sensitivity. Since 31% of the known oral taxa still cannot be grown in vitro, bacterial culture is still important in microbiology (Vartoukian et al., 2010). However for diagnostic purposes, except for antibiotic sensitivity, culture-dependent methods are labor-intensive, costly, and not as comprehensive as the molecular DNA-based technologies, which circumvent the need for culture. Gel-based technologies High-throughput analysis of microbial communities has been possible due to several culture-independent methodologies. Early on, community-fingerprinting techniques such as denaturing gradient gel electrophoresis (DGGE) and temperature gradient gel electrophoresis (TGGE) separated DNA of the same size, which in turn could be sequenced for identification purposes (Anderson & Cairney, 2004, Deng et al., 2008, Nishigaki et al., 2000). Restriction fragment length polymorphism (RFLP) was used to digest homologous DNA sequences and variation in the resulting fragment length was then used as a tool for genome mapping (Deng et al., 2008). This enabled a macro-level analysis where large shifts or variations in the population of a particular microbial community could be identified. These community-profiling methods were based on PCR where specific primers were used to amplify regions of interest and then subjected to analysis. DNA Microarrays One of the first methods to rapidly assess specific bacterial associations in oral health and disease was DNA-DNA checkerboard hybridization on a solid membrane (Socransky et al., 1994). In the original method, 30 whole genomic probes were hybridized to 45 DNA samples bound to a membrane for up to 1,350 simultaneous hybridizations. Another version of checkerboard hybridization was a reverse-capture protocol that utilized labeled 16S rDNA PCR products that were hybridized to 16S rDNA taxa-specific oligonucleotide probes bound to the membrane (Paster et al., 1998, Socransky et al., 1994). DNA microarrays used signals from hybridization of DNA fragments to hundreds or thousands of complementary probes arrayed on a glass slide for expression profiling (Schena et al., 1995). This was modified to identify microbial populations (Bodrossy & Sessitsch, 2004, Zhou, 2003), such as the PhyloChip to screen for 16S rDNA (Wilson et al., 2002) and GeoChip for functional analysis (He et al., 2007). The Human Oral Microbe Identification Microarray(HOMIM), another reverse-capture protocol, was developed using 379 species-level probes to identify ~290 oral bacterial species and has been used in several disease related and oral microbiome characterization studies (Colombo et al., 2009, Colombo et al., 2012, Olson et al., 2011, Lif Holgerson et al., 2011, Duran-Pinedo et al., 2011, Torlakovic et al., 2012, Belstrom et al., 2015). An array platform with broad range, higher taxa probes can help to identify/estimate a community population composition at a family or phylum level even if species level specificity is absent (Hamady & Knight, 2009). 16S rRNA gene sequencing The 16S rRNA has been used as an evolutionary clock (Woese, 1987) for the identification and classification of pure cultures of bacteria as well as for estimation of bacterial diversity in environmental samples (Rajendhran & Gunasekaran, 2011). Comparative analyses of 16S rDNA sequences have been the primary basis of defining the microbiome from all environments, including the oral cavity. With pure bacterial cultures, PCR amplicons (approximately 1,500 bp) of 16S rRNA genes were simply sequenced using Sanger sequencing (Sanger et al., 1977). Phylogenetic identity of bacterial taxa, whether they are cultivable or not-yet-cultivated, in a mixed population, e.g., plaque, was determined using what has been referred to as the 16S rRNA approach (Paster et al., 2006). Briefly, DNA isolated from any given environment is amplified using universally conserved PCR primers for 16S rRNA genes. The resultant amplicons were cloned into Escherichia coli, and the 16S rDNA inserts were sequenced to determine species identity (Hugenholtz & Pace, 1996, Paster et al., 2001). Typically, a >98.5% identity defines a species/phylotype. Consequently, 16S rDNA sequences from an isolate or cloned insert with <98.5% similarity to previously defined phylotypes would be considered representatives of new species (Dewhirst et al., 2010). Next Generation Sequencing (NGS) Platforms In the last decade, next generation sequencing methods have revolutionized the study of microbial diversity, which allow for large-scale sequencing projects to be completed in a few days or sometimes hours. The main technologies for next-generation sequencing are as follows. 454 pyrosequencing. This method clonally amplified fragmented DNA on beads within an emulsion (Margulies et al., 2005). The sequencer was able to generate over 250 bp long reads and about 400,000 reads per run. Applied Biosystem (Life Technologies) ○ Sequencing by Oligo Ligation Detection (SOLiD). This technique was similar to 454 pyrosequencing in that fragmented DNA was amplified on agarose beads. However, this technique utilized the incorporation of a ligase and universal oligonucleotides, which resulted in millions of reads. ○ Personal Genome Machine, or Ion Torrent. Newer technology with a similar emulsion PCR for amplification technique but with an underlying semiconductor technology (Nakano et al., 2003). Illumina ○ Early instruments utilized a sequencing-by-synthesis platform where DNA fragments were clonally amplified on a flow cell and binding of complementary fluorescently labeled dNTPs is detected. Millions of 35 bp reads could be produced in one run and, depending on the instrument, multiplexing of sample across lanes was possible. ○ In the past few years, Illumina has emerged as the market leader with a suite of instruments like HiSeq, HiSeq X, NextSeq 500 and MiSeq with varying abilities for sequencing length and number of reads. The MiSeq can generate up to 2×300bps reads and HiseqX can produce ~600Gb of data. The MiSeq is generally used for 16S rDNA profiling. Pacific Biosciences (PacBio), single molecule real-time (SMRT) technology. This instrument is sensitive enough to detect a single fluorescently labeled nucleotide (Korlach et al., 2008, Levene et al., 2003, Lundquist et al., 2008) and is purportedly able to generate ~10,000 bp reads. The PacBio platform is often used to determine whole genomic sequences, without the need for a reference genome. Oxford Nanopore, MinION technology. One of the most recent technologies, released in May 2015, enables sequencing of single DNA molecules (Mikheyev & Tin, 2014, Quick et al., 2014). As with the PacBio, MinION would allow for de novo sequencing of whole genomes. All NGS analysis requires extensive bioinformatic capabilities and involved data quality control, filtering for good quality reads, aligning and mapping to good reference genomes, removing chimeras, normalizations across samples and populations for meaningful interpretations. 16S rDNA profiling The 16S rRNA approach was refined further using NGS methodologies to rapidly sequence hypervariable regions of the 16S rRNA genes (Caporaso et al., 2011). This involves amplification of DNA samples using universally conserved PCR primers of 16S rDNA and sequencing of the amplified regions to produce millions of reads enabling multiplexing of several samples in one run. The length of sequencing reads varies depending upon the primers used, but many studies utilize about 500 base-reads for a typical sequencing run, which allows for microbial community identification (Liu et al., 2007, Liu et al., 2008, Wang et al., 2007, Mougeot et al., 2016). At the present time, using an Illumina platform (described below), 500 to 600 base-reads is the size limit for sequencing. This technique does rely on PCR amplification and care should be taken on which region of the 16S rRNA gene is used in analysis for accurate classification of the population (Yang et al., 2016), but nevertheless is a valuable tool to identify species in a population. A common bioinformatics tool for analysis has been Quantitative Insights Into Microbial Ecology (QIIME), which picks Operational Taxonomic Units (OTUs) and assigns taxonomic identities based on comparisons to sequences from a reference database (Caporaso et al., 2010). Typically, especially with only 500 or base pair reads, these analyses identify taxa at the genus level with some species level identification. Human Oral Microbe Identification using Next Generation Sequencing (HOMINGS), http://homings.forsyth.org, is the new HOMIM, which utilizes standard NGS methodologies (Caporaso et al., 2010), and is capable of species level identification for most of the prevalent oral bacterial taxa. This is achieved by an in silico search for specific ‘probe’ sequences, called ProbeSeq for HOMINGS, that targets approximately 600 species. ProbeSeq is an iterative process in which for those sequences that are not identified, the search is repeated with 129 genus-level probes that will identify those species at the genus level (Gomes et al., 2015). The advantages of HOMINGS are that it is computationally efficient, rapid, reproducible, and can identify the majority of the oral microbiome at the species level. There is good correlation between HOMINGS and HOMIM (Mougeot et al., 2016). HOMINGS has been used in several recent studies demonstrating bacterial associations with endodontic lesions (Gomes et al., 2015), salivary microbiomes in caries and periodontitis (Belstrom et al., 2016b), temporal differences in salivary microbiomes (Belstrom et al., 2016a), and in biofilm models in response to sucrose induced dysbiosis (Rudney et al., 2015). Most recently, a multi-stage algorithm for 16S rDNA NGS reads was developed for species-level identification. Although this method requires more computing power, it is able to maximize the percentage of reads classified at the species level (Al-Hebshi et al., 2015). In that paper, this technique was used to determine the oral microbiome in subjects with oral cancer. Whole-Genome Shotgun Metagenome Sequencing The entire DNA (genome) of a single microbial culture or a complex microbial population can now be sequenced to great depth allowing us to generate reference genomes (de novo assembly) as a resource for future studies or identify the composition of microbial community respectively (mapping back to a reference genome). This culture and PCR independent technique allows parallel sequencing and identification of several organisms. Long read lengths enable a more accurate assembly of the genomes present in the population, however the huge volume of data generated still poses a bioinformatic challenge (Grice & Segre, 2012). Fortunately, at least for the oral microbiome, there are complete genomes for about 400 bacterial species that facilitate assembly. At great depth of short read sequencing, metagenomic analysis also allows quantification of copy number and allelic variants of genes within the microbial population (Vincent et al., 2016). Microbial Metatranscriptomic Sequencing Both 16S rDNA and metagenome sequencing allow us to determine ‘who is there’ however metatranscriptomic analysis would tell ‘what they are doing’. The metatranscriptome represents the RNA encoded by the microbial population; this functional analysis is performed by enriching for the mRNA, converting it into cDNA and sequencing the fragments. The reads are mapped back to reference genomes for gene expressions profiling within the microbial communities. Clinical samples can be sequenced to identify changes in gene expression between disease and normal states to identify key pathways upregulated in disease and expression patterns of potential pathogenic factors and microbial diversity (Yost et al., 2015, Duran-Pinedo et al., 2014, Benitez-Paez et al., 2014) (Wade, 2011). Bacterial associations in health and disease Unlike many human diseases, oral bacterial diseases, such as caries and periodontitis, are not caused by a single species, but by a consortium of species that are likely living harmlessly in very low numbers (often below the limit of detection) in the oral cavity. In essence, oral bacterial diseases are opportunistic infections and thus disease occurs under the proper circumstances and conditions, e.g., diet, host immune response, complicating systemic or genetic disorders, pH, poor oral hygiene, life style, and even bad luck. In using the molecular techniques described above, bacterial associations have been determined in their relationships to health status. Such studies help to determine the role of specific species in oral health and disease, including extraoral sites in systemic diseases. However, note that these associations do not necessarily identify actual etiological agents, hence often the designation “putative” pathogens or biomarkers of disease. Also of note is that, in all cases, bacterial associations are usually more complex than previously believed. Oral Health The oral microbiome changes during the life of an individual from bacterial acquisition at birth (Berkowitz & Jones, 1985, Asikainen & Chen, 1999) to bacterial colonization of the elderly (Preza et al., 2008). Species of Streptococcus are usually the first pioneering microorganisms to colonize the oral cavity with Streptococcus salivarius found mostly on the tongue dorsum and in saliva, Streptococcus mitis on the buccal mucosa and Streptococcus sanguinis on the teeth (Gibbons & Houte, 1975, Socransky & Manganiello, 1971, Smith et al., 1993). The growth and metabolism of these pioneer species change local environmental conditions such as local redox potential, pH, coaggregation, and availability of nutrients, thereby enabling more fastidious organisms to colonize after them (Marsh, 2000). Over time, other microbial communities take over including Prevotella melaninogenica, Fusobacterium nucleatum, Veillonella, Neisseria and nonpigmented Prevotella (Kononen et al., 1992). With the development of teeth, an increase in the presence of genera such as Leptotrichia and Campylobacter is observed and along with colonization by additional species such as Prevotella denticola and members of the Fusobacterium and Selenomonas genera (Kononen et al., 1994). The eruption of teeth creates a new habitat, the gingival crevice, which is nourished by the gingival crevicular fluid (GCF). Along with saliva, GCF is critical for the maintenance of the integrity of the gingival crevices and contains antimicrobial peptides, immunoglobulins and a range of other active proteins that enable it to influence the ecology of the oral cavity. Moreover, it also contains nutrients that support the resident microflora (Marsh, 2000). This continual succession of microbes is eventually replaced with a stable homeostasis of microbial communities that is referred to as the climax community whereby different bacteria interact to establish an ecosystem where each community contributes in some form (Marsh, 2000). The “keystone pathogen” hypothesis suggests that specific low-abundance pathogens can influence periodontal disease by altering the “healthy” microflora to a disease state (Hajishengallis et al., 2012). However, you still have to know health before you know disease. Depending upon the oral site and individuals, many health-associated species have been identified. By using the 16S rRNA approach, Aas et al (Aas et al., 2005) analyzed sites from five clinically healthy subjects. Sites included tongue dorsum, lateral sides of tongue, buccal mucosa, hard and soft palate, palate, supragingival and subgingival plaque, maxillary anterior vestibule, and tonsils. Species typically associated with periodontitis and caries were not detected. Using NGS, Zaura et al 2009 performed a similar analysis of multiple oral sites. In both of these studies, species and phylotypes of Streptococcus, Granulicatella, Neisseria, Haemophilus, Corynebacterium, Rothia, Actinomyces, Prevotella, Capnocytophaga, Porphyromonas and Fusobacterium were common. They concluded that most oral taxa found in unrelated healthy individuals was similar, and supported the concept of a healthy core microbiome. From the HMP 16S rRNA gene data of 200 subjects,(Segata et al., 2012) found similar taxa, but there was significant subject-to-subject variation. Periodontitis Putative pathogens have long been implicated in periodontal disease including Porphyromonas gingivalis, Tannerella forsythia, Aggregatibacter actinomycetemcomitans, and species of Treponema and Prevotella. However, using checkerboard hybridization, Socransky et al. (Socransky et al., 1998) determined that oral diseases were better defined as based on a combination of species, or complexes, rather than a single specific etiologic agent. The authors defined five complexes of which the “red complex” was the most pathogenic one. This complex contained P. gingivalis, T. forsythia, and Treponema denticola and depended on earlier colonization of the pocket by the orange complex (Socransky & Haffajee, 2005, Socransky et al., 1998). The literature is quite extensive in regards to those species that are associated with periodontal disease (Socransky & Haffajee, 2005, Teles et al., 2013). Several studies suggested that there may be additional red complex species (Kumar et al., 2003, Paster et al., 2001) that are associated with chronic periodontitis. Recently, there was systematic review (Perez-Chaparro et al., 2014) of 1,450 bacterial association studies of subgingival plaque, of which 41 studies qualified for analysis. Consequently, based on these analyses, they concluded that there were 17 additional disease-associated species or phylotypes and these are listed in Table 1 along with their Human Oral Taxon (HOT) designations (www.homd.org). Refractory periodontal disease Some subjects who have destructive periodontal disease do not respond to conventional therapy and continue to lose periodontal attachment. This has often been termed as refractory periodontal disease (Adams, 1992). It has been suggested that subjects with refractory disease may be mildly immunocompromised making them more susceptible to periodontal disease. Early studies of refractory periodontal disease demonstrated a lack of typical periodontal pathogens (Magnusson et al., 1991). In contrast, using newer molecular techniques such as HOMIM, Colombo et al. 2009 were able to demonstrate that refractory periodontitis differed from treatable periodontitis by having a higher frequency of putative periodontal pathogens as listed above and in Table 1. However, they also found additional species that are not commonly detected in treatable periodontal disease, including P. alactolyticus, Brevundimonas diminuta, Shuttleworthia satelles, D. invisus, Granulicatella adiacens, Veillonella atypica, and Mycoplasma salivarium. A more recent study implicated Actinetobacter baumannii, an important nosocomial pathogenthat is notoriously antibiotic resistant, as a risk factor for refractory periodontitis (Richards et al., 2015). Aggressive periodontitis This periodontal disease was previously referred to as localized juvenile periodontitis or generalized juvenile periodontitis. As the name implies, this is an aggressive form of periodontitis that typically affects only incisors and first molars in teenagers or young adults. There is usually a lack of gingival inflammation, even with deep probing depths (Albandar, 2014). Aggregatibacter (Actinobacillus) actinomycetemcomitans has long been considered the etiologic agent of aggressive periodontitis, however more recent evidence has shown that the subgingival plaque microbiome of aggressive periodontitis resembles that of chronic periodontitis (Kononen & Muller, 2014). Shaddox et al (2012) showed that A. actinomycetemcomitans, Filifactor alocis, Tannerella spp. Solobacterium moorei, Parvimonas micra, and Capnocytophaga spp were most abundant in aggressive periodontitis. A recent study showed that a consortium of A. actinomycetemcomitans, Streptococcus parasanguinis, and Filifactor alocis may serve as a biomarker of disease, i.e., predict bone loss before it occurs (Fine et al., 2013). Caries One of the most prevalent human bacterial infections is dental caries, which leads to tooth decay and potentially tooth loss. Streptococcus mutans has long been considered at the etiological agent of caries inasmuch as it not only produces lactic acid, but also thrives in the low pH environment. However, 10 to 20% of subjects with caries do not have detectable levels of S. mutans, so clearly other acid producing bacterial taxa must be involved. Molecular methods such as the 16S rRNA approach or microarrays (Aas et al., 2008) have demonstrated that in carious lesions with S. mutans, additional species belonging to the genera Atopobium, Propionibacterium, and Lactobacillus, were present at significantly higher levels. In those subjects with no detectable levels of S. mutans, Lactobacillus spp., Bifidobacterium dentium, and low-pH non-S. mutans streptococci were predominant. Based on these results, it was suggested that bacterial species other than S. mutans, e.g., species of the genera Veillonella, Lactobacillus, Scardovia, and Propionibacterium, low-pH non-S. mutans streptococci, Actinomyces spp., and Atopobium spp., may play an important role in caries progression. NGS analyses of the microbiome of populations with a low and high prevalence of caries found that adolescents in Romania, who had limited access to care, were colonized with S. mutans and S. sobrinus. In contrast, those adolescents in Sweden, who had very good care, were colonized only infrequently with S. mutans and S. sobrinus., but were colonized more with species of Actinomyces, Selenomonas, Prevotella, and Capnocytophaga (Johansson et al. 2016). The oral microbiomes do differ between primary and secondary dentitions as well as in root surface caries. In the primary dentition (Becker et al., 2002), S. mutans was typically detected at high levels. Other disease associated species included Actinomyces gerencseriae, Scardovia wiggsiae, Veillonella, S. salivarius, S. constellatus, S. parasanguinis, and Lactobacillus fermentum. Root surface caries differ from primary and secondary caries in that root surfaces do not have enamel. Preza et al (Preza et al., 2008) demonstrated that S. mutans was also associated with root surface caries, but that the predominant taxa included Actinomyces spp., Lactobacillus, Enterococcus faecalis, Mitsuokella sp. HOT131, Atopobium and Olsenella spp., Prevotella multisaccharivorax, P. alactolyticus, and Propionibacterium acidifaciens. Of note, detectable levels of E. faecalis and P. alactolyticus are typically found only in endodontic lesions and not in dental plaque. Odontogenic infections These pus-laden infections typically originate within a tooth or surrounding structures resulting in swellings of the head, face and neck (Flynn et al., 2012). . Using the 16S rRNA approach, several predominant species that had been previously associated with odontogenic infections were detected (Flynn et al., 2012). These species included Fusobacterium spp, Parvimonas micra, Porphyromonas endodontalis, and Prevotella oris. However, they also detected newly-associated species including Dialister pneumosintes, D. invisus, and Eubacterium brachy, as well as several phylotypes. An interesting finding in this study was that species of Streptococcus were not detected. Endodontic lesions In primary infections, predominant taxa detected include species of Peptostreptococcus, P. micra, Filifactor alocis, P. alactolytcus, species of Dialister, Fusobacterium nucleatum, Treponema denticola, Porphyromonas endodontalis, P. gingivalis, T. forsythia, Prevotella baroniae, Prevotella intermedia, Prevotella nigrescens and Bacteroidaceae [G-1] HOT272 (Siqueira & Rocas, 2009). Enterococcus faecalis was detected, but as lower levels. However in retreatment cases, the predominant taxa include Enterococcus species such as E. faecalis, Parvimonas micra, Filifactor alocis, P. alactolytcus, Streptococcus constellatus and Streptococcus anginosus, and Propionibacterium propionicum. The microbiomes of endodontic-periodontal lesions had possessed similar profiles including E. faecalis, P. micra, Mogibacterium timidum, F. alocis, and Fretibacterium fastidiosum (Gomes et al., 2015). Oral bacteria as biomarkers for non-oral diseases Oral bacteria have been linked to a number of systemic diseases including bacterial endocarditis (Berbari et al., 1997), ischemic stroke (Joshipura et al., 2003), cardiovascular disease (Beck & Offenbacher, 2005; Teles & Wang, 2011); pancreatic cancer (Farrell et al 2012), pediatric Crohn’s Disease (Docktor et al., 2012), and pneumonia (Awano et al., 2008). Periodontal disease has been shown to predispose individuals to cardiovascular disease through its ability to induce chronic inflammation (Valtonen, 1991, Syrjanen, 1990). Similarly, the presence of several anaerobic oral bacterial species has been shown to predispose to bacterial pneumonia including Actinobacillus actinomycetemcomitans and Streptococcus constellatus (Venkataramani et al., 1994, Shinzato & Saito, 1994). In Alzheimer’s disease, inflammation, a key feature of the disease (Olsen & Singhrao, 2015), could be caused in part by peripheral infections, such as periodontal disease. Periodontal pathogens such as Aggregatibacter actinomycetemcomitans and Prevotella intermedia are capable of eliciting systemic inflammation, which results in the release of pro-inflammatory cytokines that traverse the blood-brain barrier. An intriguing suggestion has been that oral bacteria may play a role in nitric oxide (NO) homeostasis, which is important in renal and cardiovascular health (Hezel & Weitzberg, 2015). Dietary nitrates can be reduced to nitrites by oral bacteria and nitrite, absorbed in the blood, is further reduced to NO by a variety of mechanisms. NO then acts on vascular smooth muscle to stimulate vasodilation. Important Considerations and Conclusions Researchers are now routinely identifying bacterial composition, and high-throughput sequencing of the microbiome will now progress into functional studies encompassing genomics, transcriptomics and metabolomics of both host and pathogens. Such analysis could provide insights into activity of the microbes, their relationship to hosts, and possible causative mechanisms. The second phase of the NIH Human Microbiome Project will study the host-microbiome relationship in longitudinal studies (Integrative, 2014). These data will guide researchers to develop new therapies that target key mechanisms. With such huge datasets, we will likely identify ‘community signatures’ of certain diseases. It can be envisioned that chairside or bedside, point-of-care diagnostics could be developed that target key bacterial taxa. Consequently, these potential biomarkers of disease, the proverbial “canary of the coal mine in human disease” could be used to warn dentists or physicians of disease yet to come or to assess risk of disease. Thus, for the dentist, oral medicine specialist, or periodontist, these warning “danger” microbial profiles would allow for early treatment to combat disease in the preclinical stages. The oral microbiome could be used further to monitor health status after treatment, i.e., is the treatment working to establish a more healthy microbial profile? Regardless of specific application, microbial analysis is in an exciting phase of research with huge prospects for the clinic. Acknowledgements Supported in part by National Institute of Dental and Craniofacial Research/National Institute of Health grant, R01 DE021565. We thank Rebecca Misra for her help in preparation of the manuscript. Figure 1 Figure 2 Table 1 Newly identified putative periodontal pathogens (from Perez-Chaparro et al 2014) Bacterial taxa Anaeroglobus geminatus HOT 121 cultivable Archaea spp. cultivable Bacteroidales [G-2] sp. oral taxon 274 unnamed Desulfobulbus sp. oral taxon 041 phylotype Eubacterium [XI] [G-5] saphenum HOT 759 cultivable Filifactor alocis HOT 539 cultivable Fretibacterium fastidiosum HOT 363 cultivable Fretibacterium sp. oral taxon 360 phylotype Fretibacterium sp. oral taxon 362 phylotype Mogibacterium timidum HOT 042 cultivable Peptostreptococcus stomatis HOT 112 cultivable Porphyromonas endodontalis HOT 273 cultivable Selenomonas sputigena HOT 151 cultivable TM7 [G-5] sp. oral taxon 356 phylotype Treponema lecithinolyticum HOT 653 cultivable Treponema medium HOT 667 cultivable Treponema vincentii HOT 029 cultivable Conflicts of interest: none to declare References Aas JA Griffen AL Dardis SR Lee AM Olsen I Dewhirst FE Leys EJ Paster BJ Bacteria of dental caries in primary and permanent teeth in children and young adults J Clin Microbiol 2008 46 1407 17 18216213 Aas JA Paster BJ Stokes LN Olsen I Dewhirst FE Defining the normal bacterial flora of the oral cavity J Clin Microbiol 2005 43 5721 32 16272510 Adams DF Diagnosis and treatment of refractory periodontitis Curr Opin Dent 1992 2 33 8 1520935 Al-Hebshi NN Nasher AT Idris AM Chen T Robust species taxonomy assignment algorithm for 16S rRNA NGS reads: application to oral carcinoma samples J Oral Microbiol 2015 7 28934 26426306 Albandar JM Aggressive periodontitis: case definition and diagnostic criteria Periodontol 2000 2014 65 13 26 24738584 Anderson IC Cairney JW Diversity and ecology of soil fungal communities: increased understanding through the application of molecular techniques Environ Microbiol 2004 6 769 79 15250879 Asikainen S Chen C Oral ecology and person-to-person transmission of Actinobacillus actinomycetemcomitans and Porphyromonas gingivalis Periodontol 2000 1999 20 65 81 10522223 Awano S Ansai T Takata Y Soh I Akifusa S Hamasaki T Yoshida A Sonoki K Fujisawa K Takehara T Oral health and mortality risk from pneumonia in the elderly J Dent Res 2008 87 334 9 18362314 Beck JD Offenbacher S Systemic effects of periodontitis: epidemiology of periodontal disease and cardiovascular disease J Periodontol 2005 76 2089 100 16277581 Becker MR Paster BJ Leys EJ Moeschberger ML Kenyon SG Galvin JL Boches SK Dewhirst FE Griffen AL Molecular analysis of bacterial species associated with childhood caries J Clin Microbiol 2002 40 1001 9 11880430 Belstrom D Fiehn NE Nielsen CH Klepac-Ceraj V Paster BJ Twetman S Holmstrup P Differentiation of salivary bacterial profiles of subjects with periodontitis and dental caries J Oral Microbiol 2015 7 27429 25854857 Belstrom D Holmstrup P Bardow A Kokaras A Fiehn NE Paster BJ Temporal Stability of the Salivary Microbiota in Oral Health PLoS One 2016a 11 e0147472 26799067 Belstrom D Paster BJ Fiehn NE Bardow A Holmstrup P Salivary bacterial fingerprints of established oral disease revealed by the Human Oral Microbe Identification using Next Generation Sequencing (HOMINGS) technique J Oral Microbiol 2016b 8 30170 26782357 Benitez-Paez A Belda-Ferre P Simon-Soro A Mira A Microbiota diversity and gene expression dynamics in human oral biofilms BMC Genomics 2014 15 311 24767457 Berbari EF Cockerill FR 3rd Steckelberg JM Infective endocarditis due to unusual or fastidious microorganisms Mayo Clin Proc 1997 72 532 42 9179137 Berkowitz RJ Jones P Mouth-to-mouth transmission of the bacterium Streptococcus mutans between mother and child Arch Oral Biol 1985 30 377 9 3857909 Bodrossy L Sessitsch A Oligonucleotide microarrays in microbial diagnostics Curr Opin Microbiol 2004 7 245 54 15196491 Caporaso JG Kuczynski J Stombaugh J Bittinger K Bushman FD Costello EK Fierer N Pena AG Goodrich JK Gordon JI Huttley GA Kelley ST Knights D Koenig JE Ley RE Lozupone CA McDonald D Muegge BD Pirrung M Reeder J Sevinsky JR Turnbaugh PJ Walters WA Widmann J Yatsunenko T Zaneveld J Knight R QIIME allows analysis of high-throughput community sequencing data Nat Methods 2010 7 335 6 20383131 Caporaso JG Lauber CL Walters WA Berg-Lyons D Lozupone CA Turnbaugh PJ Fierer N Knight R Global patterns of 16S rRNA diversity at a depth of millions of sequences per sample Proc Natl Acad Sci U S A 2011 108 Suppl 1 4516 22 20534432 Colombo AP Bennet S Cotton SL Goodson JM Kent R Haffajee AD Socransky SS Hasturk H Van Dyke TE Dewhirst FE Paster BJ Impact of periodontal therapy on the subgingival microbiota of severe periodontitis: comparison between good responders and individuals with refractory periodontitis using the human oral microbe identification microarray J Periodontol 2012 83 1279 87 22324467 Colombo AP Boches SK Cotton SL Goodson JM Kent R Haffajee AD Socransky SS Hasturk H Van Dyke TE Dewhirst F Paster BJ Comparisons of subgingival microbial profiles of refractory periodontitis, severe periodontitis, and periodontal health using the human oral microbe identification microarray J Periodontol 2009 80 1421 32 19722792 Demmer RT Jacobs DR Jr Singh R Zuk A Rosenbaum M Papapanou PN Desvarieux M Periodontal bacteria and prediabetes prevalence in origins: the oral infections, glucose intolerance, and insulin resistance study J Dent Res 2015 94 201S 11S 26082387 Deng W Xi D Mao H Wanapat M The use of molecular techniques based on ribosomal RNA and DNA for rumen microbial ecosystem studies: a review Mol Biol Rep 2008 35 265 74 17484038 Dewhirst FE Chen T Izard J Paster BJ Tanner AC Yu WH Lakshmanan A Wade WG The human oral microbiome J Bacteriol 2010 192 5002 17 20656903 Docktor MJ Paster BJ Abramowicz S Ingram J Wang YE Correll M Jiang H Cotton SL Kokaras AS Bousvaros A Alterations in diversity of the oral microbiome in pediatric inflammatory bowel disease Inflamm Bowel Dis 2012 18 935 42 21987382 Duran-Pinedo AE Chen T Teles R Starr JR Wang X Krishnan K Frias-Lopez J Community-wide transcriptome of the oral microbiome in subjects with and without periodontitis ISME J 2014 8 1659 72 24599074 Duran-Pinedo AE Paster B Teles R Frias-Lopez J Correlation network analysis applied to complex biofilm communities PLoS One 2011 6 e28438 22163302 Farrell JJ Zhang L Zhou H Chia D Elashoff D Akin D Paster BJ Joshipura K Wong DT Variations of oral microbiota are associated with pancreatic diseases including pancreatic cancer Gut 2012 61 582 8 21994333 Fine DH Markowitz K Fairlie K Tischio-Bereski D Ferrendiz J Furgang D Paster BJ Dewhirst FE A consortium of Aggregatibacter actinomycetemcomitans, Streptococcus parasanguinis, and Filifactor alocis is present in sites prior to bone loss in a longitudinal study of localized aggressive periodontitis J Clin Microbiol 2013 51 2850 61 23784124 Flynn TR Paster BJ Stokes LN Susarla SM Shanti RM Molecular methods for diagnosis of odontogenic infections J Oral Maxillofac Surg 2012 70 1854 9 22326175 Gibbons RJ Houte JV Bacterial adherence in oral microbial ecology Annu Rev Microbiol 1975 29 19 44 1180512 Gilbert JA Dupont CL Microbial metagenomics: beyond the genome Ann Rev Mar Sci 2011 3 347 71 Gomes BP Berber VB Kokaras AS Chen T Paster BJ Microbiomes of Endodontic-Periodontal Lesions before and after Chemomechanical Preparation J Endod 2015 41 1975 84 26521147 Grice EA Segre JA The human microbiome: our second genome Annu Rev Genomics Hum Genet 2012 13 151 70 22703178 Griffen AL Beall CJ Firestone ND Gross EL Difranco JM Hardman JH Vriesendorp B Faust RA Janies DA Leys EJ CORE: a phylogenetically-curated 16S rDNA database of the core oral microbiome PLoS One 2011 6 e19051 21544197 Hajishengallis G Darveau RP Curtis MA The keystone-pathogen hypothesis Nat Rev Microbiol 2012 10 717 25 22941505 Hamady M Knight R Microbial community profiling for human microbiome projects: Tools, techniques, and challenges Genome Res 2009 19 1141 52 19383763 He Z Gentry TJ Schadt CW Wu L Liebich J Chong SC Huang Z Wu W Gu B Jardine P Criddle C Zhou J GeoChip: a comprehensive microarray for investigating biogeochemical, ecological and environmental processes ISME J 2007 1 67 77 18043615 Hezel MP Weitzberg E The oral microbiome and nitric oxide homoeostasis Oral Dis 2015 21 7 16 23837897 Howe K Bateman A Durbin R QuickTree: building huge Neighbour-Joining trees of protein sequences Bioinformatics 2002 18 1546 7 12424131 Hugenholtz P Pace NR Identifying microbial diversity in the natural environment: a molecular phylogenetic approach Trends Biotechnol 1996 14 190 7 8663938 Human Microbiome Project C A framework for human microbiome research Nature 2012a 486 215 21 22699610 Human Microbiome Project C Structure, function and diversity of the healthy human microbiome Nature 2012c 486 207 14 22699609 Integrative HMPRNC The Integrative Human Microbiome Project: dynamic analysis of microbiome-host omics profiles during periods of human health and disease Cell Host Microbe 2014 16 276 89 25211071 Johansson I Witkowska E Kaveh B Lif Holgerson P Tanner AC The microbiome in populations with a low and high prevalence of caries J Dent Res 2016 95 80 6 26442950 Joshipura KJ Hung HC Rimm EB Willett WC Ascherio A Periodontal disease, tooth loss, and incidence of ischemic stroke Stroke 2003 34 47 52 12511749 Kazor CE Mitchell PM Lee AM Stokes LN Loesche WJ Dewhirst FE Paster BJ Diversity of bacterial populations on the tongue dorsa of patients with halitosis and healthy patients J Clin Microbiol 2003 41 558 63 12574246 Kolenbrander PE Oral microbial communities: biofilms, interactions, and genetic systems Annu Rev Microbiol 2000 54 413 37 11018133 Kononen E Asikainen S Jousimies-Somer H The early colonization of gram-negative anaerobic bacteria in edentulous infants Oral Microbiol Immunol 1992 7 28 31 1528621 Kononen E Asikainen S Saarela M Karjalainen J Jousimies-Somer H The oral gram-negative anaerobic microflora in young children: longitudinal changes from edentulous to dentate mouth Oral Microbiol Immunol 1994 9 136 41 7936718 Kononen E Muller HP Microbiology of aggressive periodontitis Periodontol 2000 2014 65 46 78 24738586 Korlach J Marks PJ Cicero RL Gray JJ Murphy DL Roitman DB Pham TT Otto GA Foquet M Turner SW Selective aluminum passivation for targeted immobilization of single DNA polymerase molecules in zero-mode waveguide nanostructures Proc Natl Acad Sci U S A 2008 105 1176 81 18216253 Kumar PS Griffen AL Barton JA Paster BJ Moeschberger ML Leys EJ New bacterial species associated with chronic periodontitis J Dent Res 2003 82 338 44 12709498 Leishman SJ Do HL Ford PJ Cardiovascular disease and the role of oral bacteria J Oral Microbiol 2010 2 Levene MJ Korlach J Turner SW Foquet M Craighead HG Webb WW Zero-mode waveguides for single-molecule analysis at high concentrations Science 2003 299 682 6 12560545 Lif Holgerson P Harnevik L Hernell O Tanner AC Johansson I Mode of birth delivery affects oral microbiota in infants J Dent Res 2011 90 1183 8 21828355 Liu Z DeSantis TZ Andersen GL Knight R Accurate taxonomy assignments from 16S rRNA sequences produced by highly parallel pyrosequencers Nucleic Acids Res 2008 36 e120 18723574 Liu Z Lozupone C Hamady M Bushman FD Knight R Short pyrosequencing reads suffice for accurate microbial community analysis Nucleic Acids Res 2007 35 e120 17881377 Lundquist PM Zhong CF Zhao P Tomaney AB Peluso PS Dixon J Bettman B Lacroix Y Kwo DP McCullough E Maxham M Hester K McNitt P Grey DM Henriquez C Foquet M Turner SW Zaccarin D Parallel confocal detection of single molecules in real time Opt Lett 2008 33 1026 8 18451975 Mager DL Ximenez-Fyvie LA Haffajee AD Socransky SS Distribution of selected bacterial species on intraoral surfaces J Clin Periodontol 2003 30 644 54 12834503 Magnusson I Marks RG Clark WB Walker CB Low SB McArthur WP Clinical, microbiological and immunological characteristics of subjects with "refractory" periodontal disease J Clin Periodontol 1991 18 291 9 2066442 Margulies M Egholm M Altman WE Attiya S Bader JS Bemben LA Berka J Braverman MS Chen YJ Chen Z Dewell SB Du L Fierro JM Gomes XV Godwin BC He W Helgesen S Ho CH Irzyk GP Jando SC Alenquer ML Jarvie TP Jirage KB Kim JB Knight JR Lanza JR Leamon JH Lefkowitz SM Lei M Li J Lohman KL Lu H Makhijani VB McDade KE McKenna MP Myers EW Nickerson E Nobile JR Plant R Puc BP Ronan MT Roth GT Sarkis GJ Simons JF Simpson JW Srinivasan M Tartaro KR Tomasz A Vogt KA Volkmer GA Wang SH Wang Y Weiner MP Yu P Begley RF Rothberg JM Genome sequencing in microfabricated high-density picolitre reactors Nature 2005 437 376 80 16056220 Mark Welch JL Rossetti BJ Rieken CW Dewhirst FE Borisy GG Biogeography of a human oral microbiome at the micron scale Proc Natl Acad Sci U S A 2016 113 E791 800 26811460 Marsh PD Kuramitsu HK Ellen RP Oral ecology and its impact on oral microbial diversity Oral Bacterial Ecology: The Molecular Basis 2000 Horizon Scientific Press Wymondham Mikheyev AS Tin MM A first look at the Oxford Nanopore MinION sequencer Mol Ecol Resour 2014 14 1097 102 25187008 Mougeot JL Stevens CB Cotton SL Morton DS Krishnan K Brennan MT Lockhart PB Paster BJ Bahrani Mougeot FK Concordance of HOMIM and HOMINGS technologies in the microbiome analysis of clinical samples J Oral Microbiol 2016 8 30379 27065347 Nakano M Komatsu J Matsuura S Takashima K Katsura S Mizuno A Single-molecule PCR using water-in-oil emulsion J Biotechnol 2003 102 117 24 12697388 Nishigaki K Naimuddin M Hamano K Genome profiling: a realistic solution for genotype-based identification of species J Biochem 2000 128 107 12 10876164 Nobbs AH Jenkinson HF Jakubovics NS Stick to your gums: mechanisms of oral microbial adherence J Dent Res 2011 90 1271 8 21335541 Olsen I Singhrao SK Can oral infection be a risk factor for Alzheimer’s disease? J Oral Microbiol 2015 7 29143 26385886 Olson JC Cuff CF Lukomski S Lukomska E Canizales Y Wu B Crout RJ Thomas JG McNeil DW Weyant RJ Marazita ML Paster BJ Elliott T Use of 16S ribosomal RNA gene analyses to characterize the bacterial signature associated with poor oral health in West Virginia BMC Oral Health 2011 11 7 21362199 Paster BJ Bartoszyk IM Dewhirst FE Identification of oral streptococci using PCR-based, reverse-capture, checkerboard hybridization Methods in Cell Science 1998 20 223 231 Paster BJ Boches SK Galvin JL Ericson RE Lau CN Levanos VA Sahasrabudhe A Dewhirst FE Bacterial diversity in human subgingival plaque J Bacteriol 2001 183 3770 83 11371542 Paster BJ Olsen I Aas JA Dewhirst FE The breadth of bacterial diversity in the human periodontal pocket and other oral sites Periodontol 2000 2006 42 80 7 16930307 Perez-Chaparro PJ Goncalves C Figueiredo LC Faveri M Lobao E Tamashiro N Duarte P Feres M Newly identified pathogens associated with periodontitis: a systematic review J Dent Res 2014 93 846 58 25074492 Preza D Olsen I Aas JA Willumsen T Grinde B Paster BJ Bacterial profiles of root caries in elderly patients J Clin Microbiol 2008 46 2015 21 18385433 Quick J Quinlan AR Loman NJ A reference bacterial genome dataset generated on the MinION portable single-molecule nanopore sequencer Gigascience 2014 3 22 25386338 Rajendhran J Gunasekaran P Microbial phylogeny and diversity: small subunit ribosomal RNA sequence analysis and beyond Microbiol Res 2011 166 99 110 20223646 Richards AM Abu Kwaik Y Lamont RJ Code blue: Acinetobacter baumannii, a nosocomial pathogen with a role in the oral cavity Mol Oral Microbiol 2015 30 2 15 25052812 Rudney JD Jagtap PD Reilly CS Chen R Markowski TW Higgins L Johnson JE Griffin TJ Protein relative abundance patterns associated with sucrose-induced dysbiosis are conserved across taxonomically diverse oral microcosm biofilm models of dental caries Microbiome 2015 3 69 26684897 Sanger F Nicklen S Coulson AR DNA sequencing with chain-terminating inhibitors Proc Natl Acad Sci U S A 1977 74 5463 7 271968 Schena M Shalon D Davis RW Brown PO Quantitative monitoring of gene expression patterns with a complementary DNA microarray Science 1995 270 467 70 7569999 Segata N Haake SK Mannon P Lemon KP Waldron L Gevers D Huttenhower C Izard J Composition of the adult digestive tract bacterial microbiome based on seven mouth surfaces, tonsils, throat and stool samples Genome Biol 2012 13 R42 22698087 Shaddox LM Huang H Lin T Hou W Harrison PL Aukhil I Walker CB Klepac-Ceraj V Paster BJ Microbiological characterization in children with aggressive periodontitis J Dent Res 2012 91 927 33 22863892 Shinzato T Saito A A mechanism of pathogenicity of “Streptococcus milleri group” in pulmonary infection: synergy with an anaerobe J Med Microbiol 1994 40 118 23 8107060 Shira Davenport E Preterm low birthweight and the role of oral bacteria J Oral Microbiol 2010 2 Siqueira JF Jr. Rocas IN Diversity of endodontic microbiota revisited J Dent Res 2009 88 969 81 19828883 Smith DJ Anderson JM King WF van Houte J Taubman MA Oral streptococcal colonization of infants Oral Microbiol Immunol 1993 8 1 4 8510978 Socransky SS Haffajee AD Periodontal microbial ecology Periodontol 2000 2005 38 135 87 15853940 Socransky SS Haffajee AD Cugini MA Smith C Kent RL Jr. Microbial complexes in subgingival plaque J Clin Periodontol 1998 25 134 44 9495612 Socransky SS Manganiello SD The oral microbiota of man from birth to senility J Periodontol 1971 42 485 96 4998039 Socransky SS Smith C Martin L Paster BJ Dewhirst FE Levin AE “Checkerboard” DNA-DNA hybridization Biotechniques 1994 17 788 92 7833043 Syrjanen J Vascular diseases and oral infections J Clin Periodontol 1990 17 497 500 2202746 Teles R Wang C-Y Mechanisms involved in the association between peridontal diseases and cardiovascular disease Oral Diseases 2011 17 450 461 21223455 Teles R Teles F Frias-Lopez J Paster B Haffajee A Lessons learned and unlearned in periodontal microbiology Periodontol 2000 2013 62 95 162 23574465 Torlakovic L Klepac-Ceraj V Ogaard B Cotton SL Paster BJ Olsen I Microbial community succession on developing lesions on human enamel J Oral Microbiol 2012 4 Valm AM Mark Welch JL Rieken CW Hasegawa Y Sogin ML Oldenbourg R Dewhirst FE Borisy GG Systems-level analysis of microbial community organization through combinatorial labeling and spectral imaging Proc Natl Acad Sci U S A 2011 108 4152 7 21325608 Valtonen VV Infection as a risk factor for infarction and atherosclerosis Ann Med 1991 23 539 43 1756023 Vartoukian SR Palmer RM Wade WG Cultivation of a Synergistetes strain representing a previously uncultivated lineage Environ Microbiol 2010 12 916 28 20074237 Venkataramani A Santo-Domingo NE Main DM Actinobacillus actinomycetemcomitans pneumonia with possible septic embolization Chest 1994 105 645 6 8306797 Vincent AT Derome N Boyle B Culley AI Charette SJ Next-generation sequencing (NGS) in the microbiological world: How to make the most of your money J Microbiol Methods 2016 Wade WG Has the use of molecular methods for the characterization of the human oral microbiome changed our understanding of the role of bacteria in the pathogenesis of periodontal disease? J Clin Periodontol 2011 38 Suppl 11 7 16 21323699 Wang Q Garrity GM Tiedje JM Cole JR Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy Appl Environ Microbiol 2007 73 5261 7 17586664 Wilson KH Wilson WJ Radosevich JL DeSantis TZ Viswanathan VS Kuczmarski TA Andersen GL High-density microarray of small-subunit ribosomal DNA probes Appl Environ Microbiol 2002 68 2535 41 11976131 Woese CR Bacterial evolution Microbiol Rev 1987 51 221 71 2439888 Yang B Wang Y Qian PY Sensitivity and correlation of hypervariable regions in 16S rRNA genes in phylogenetic analysis BMC Bioinformatics 2016 17 135 27000765 Yost S Duran-Pinedo AE Teles R Krishnan K Frias-Lopez J Functional signatures of oral dysbiosis during periodontitis progression revealed by microbial metatranscriptome analysis Genome Med 2015 7 27 25918553 Zhou J Microarrays for bacterial detection and microbial community analysis Curr Opin Microbiol 2003 6 288 94 12831906
PMC005xxxxxx/PMC5123593.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101581657 40080 RSC Adv RSC Adv RSC advances 2046-2069 27895899 5123593 10.1039/c6ra24642g NIHMS830374 Article Oxidatively Degradable Poly(thioketal urethane)/Ceramic Composite Bone Cements with Bone-Like Strength McEnery Madison A.P. a Lu Sichang b Gupta Mukesh K. a Zienkiewicz Katarzyna J. b Wenke Joseph C. c Kalpakci Kerem N. d Shimko Daniel d Duvall Craig L. a Guelcher Scott A. abe a Department of Biomedical Engineering, Vanderbilt University, Nashville, TN, USA b Department of Chemical and Biomolecular Engineering, Vanderbilt University, Nashville, TN, USA c U.S. Army Institute of Surgical Research, San Antonio, TX, USA d Medtronic Spinal and Biologics, Memphis, TN, USA e Center for Bone Biology, Vanderbilt University Medical Center, Nashville, TN 18 11 2016 8 11 2016 2016 08 11 2017 6 111 109414109424 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Synthetic bone cements are commonly used in orthopaedic procedures to aid in bone regeneration following trauma or disease. Polymeric cements like PMMA provide the mechanical strength necessary for orthopaedic applications, but they are not resorbable and do not integrate with host bone. Ceramic cements have a chemical composition similar to that of bone, but their brittle mechanical properties limit their use in weight-bearing applications. In this study, we designed oxidatively degradable, polymeric bone cements with mechanical properties suitable for bone tissue engineering applications. We synthesized a novel thioketal (TK) diol, which was crosslinked with a lysine triisocyanate (LTI) prepolymer to create hydrolytically stable poly(thioketal urethane)s (PTKUR) that degrade in the oxidative environment associated with bone defects. PTKUR films were hydrolytically stable for up to 6 months, but degraded rapidly (<1 week) under simulated oxidative conditions in vitro. When combined with ceramic micro- or nanoparticles, PTKUR cements exhibited working times comparable to calcium phosphate cements and strengths exceeding those of trabecular bone. PTKUR/ceramic composite cements supported appositional bone growth and integrated with host bone near the bone-cement interface at 6 and 12 weeks post-implantation in rabbit femoral condyle plug defects. Histological evidence of osteoclast-mediated resorption of the cements was observed at 6 and 12 weeks. These findings demonstrate that a PTKUR bone cement with bone-like strength can be selectively resorbed by cells involved in bone remodeling, and thus represent an important initial step toward the development of resorbable bone cements for weight-bearing applications. 1. Introduction Injectable and settable bone cements restore function to bone damaged by trauma or disease in a number of orthopaedic procedures, such as vertebroplasty, repair of tibial plateau fractures, and screw augmentation. Poly(methyl methacrylate) (PMMA) bone cements exhibit mechanical properties exceeding those of trabecular bone, and therefore provide mechanical stability to damaged bone.1 However, PMMA cements are non-resorbable and do not integrate with host bone. Ceramic bone cements are osteoconductive and integrate with host bone, but their brittle mechanical properties preclude their use in weight-bearing applications.2 Thus, composites of ceramics with resorbable polymers have emerged as an alternative approach that combines the ductile mechanical properties of polymers with the osteoconductivity of ceramics to provide mechanical stability and integration with host bone.3 Poly(ester urethane)s (PEUR) have been investigated as injectable bone grafts due to their injectability, settability, tunable mechanical properties, and resorption to breakdown products easily cleared from the body. PEUR grafts set within clinically relevant working times and attain strengths in the range of 10 – 80 MPa.4, 5 Lysine-derived PEUR composites incorporating ceramic particles or allograft bone set in situ with no surgical complications and support bone remodeling in sheep, rats, and rabbits.4–8 Previous work has shown lysine triisocyanate (LTI)-derived PEURs undergo autocatalytic hydrolytic degradation in which the acidic breakdown products accelerate resorption.9, 10 A degradation mechanism that allows for a more controlled and predictable degradation rate is desired to ensure that the graft degrades at a rate complementary to bone formation and remodeling. Bone remodeling is commonly achieved by creeping substitution, a process by which osteoclasts resorb residual graft and osteoblasts deposit new mineralized matrix near the graft-bone interface.11–13 The normal endogenous bone healing cascade involves an initial hematoma formation that induces the immune response accompanied by a release of pro-inflammatory factors.14 The inflammatory phase is followed by a soft callus formation that is rapidly replaced by woven mineralized bone.15 As a result of the inflammatory response, reactive oxygen species (ROS) are generated by infiltrating cells at the defect site.16, 17 Mature osteoclasts at sites of active bone remodeling are also associated with an increase in ROS.17–19 These findings suggest that hydrolytically stable biomaterials that degrade in response to cell-secreted ROS may be a useful new approach for the design of cell-degradable bone cements. Thioketals (TK), the sulfur analogs of ketals, degrade in response to cell-secreted ROS to thiol decomposition products with low cytotoxicity.20, 21 Poly(thioketal urethane) (PTKUR) foams synthesized from a TK macrodiol (1000 g mol−1) have been reported to support ROS-mediated degradation and healing in cutaneous wounds.20 However, macrodiol-based PTKURs cannot achieve bone-like strength or the number of degradable units afforded by the new single TK-containing crosslinker. Furthermore, TK-based biomaterials have not been previously investigated in bone. In this study, a novel low molecular weight thioketal (TK) diol was synthesized and utilized to formulate PTKUR bone cements that are hydrolytically stable but degradable by cell-secreted ROS. The TK diol was reacted with LTI to form a moldable and settable PTKUR cement with bone-like strength using a low-toxicity iron (III) acetylacetonate gelling catalyst. To enhance the osteoconductivity of the PTKUR, it was combined with two different types of ceramics: (1) 85% β-tricalcium phosphate (β-TCP)/15% hydroxyapatite (HA) ceramic mini-granules (MASTERGRAFT®, MG), or (2) nanocrystalline hydroxyapatite (nHA) particles.22, 23 The reactivity, rheological properties, mechanical properties, degradation rate, and cell proliferation response of the cements were assessed in vitro. The biocompatibility and remodeling of PTKUR/ceramic composite cements were investigated in a rabbit femoral condyle plug defect model to assess material resorption and integration with the host bone. 2. Materials and Methods 2.1 Materials Thioglycolic acid, 2,2-dimethoxypropane, bismuth chloride, lithium aluminum hydride, ε-caprolactone, nanocrystalline hydroxyapatite (nHA, <200 nm), and anhydrous solvents were purchased from Sigma-Aldrich (St. Louis, MO). The ε-caprolactone was treated with magnesium sulfate, and nHA was dried under vacuum at 80°C for at least 24 hours prior to use. Acros Organics iron (III) acetylacetonate (FeAA) was purchased from Fisher Scientific and used as received. Lysine triisocyanate (LTI) was purchased from Jinan Haohua Industry Co., LTD (Jinan, China) and carbon-treated in methyl-tert-butyl ether 3 times for 24 hours at 70°C to remove impurities. MasterGraft (MG) particles supplied by Medtronic (Memphis, TN) were ground to 100–300 μm diameter particles using a mortar and pestle and filtered between 100 and 300 μm sieves. The resulting microparticles were washed in 95% acetone, triple rinsed with water, and dried under vacuum. MC3T3 cells were supplied by ATCC (Manassas, VA). Gibco™ α-MEM medium, penicillin/streptomycin (P/S) and a Pierce™ bicinchoninic (BCA) Protein Assay kit were purchased from Thermo Scientific™ (Waltham, MA). Sterile phosphate buffered saline (PBS) and 0.25% trypsin were purchased from Corning Cellgro (Manassas, VA) and fetal bovine serum (FBS) from HyClone (Pittsburgh, PA). Reagents for cell fixation including glutaraldehyde and osmium tetroxide were purchased from Fisher Scientific and Sigma Aldrich, respectively. 2.2 Thioketal diol synthesis The schematic for thioketal diol synthesis is illustrated in Figure 1A. Bismuth (III) chloride was added to a dry boiling flask that was subsequently dried with a hot air gun under vacuum for about 5 minutes to ensure completely dry catalyst conditions. The flask was then purged with nitrogen and left under a positive pressure with nitrogen for the remainder of the reaction. Anhydrous acetonitrile was charged to the flask to dissolve the catalyst. 2,2-dimethoxypropane and thioglycolic acid were added to the flask, and the reaction was allowed to proceed for 24 hours while stirring at room temperature. The carboxyl-terminated intermediate was filtered with a Buchner funnel, rotary evaporated (Buchi Rotovap R-200, 35 °C), and dried under vacuum overnight. The carboxyl groups were then reduced to produce a hydroxyl-terminated TK. A 3-neck boiling flask was fitted to a 10 °C condenser capped with a 1-way glass stop-cock, a constant pressure dropping funnel, and a rubber stopper. The reactor was heated with a heat gun under vacuum for about 5 minutes to ensure completely dry reaction conditions. The reactor was then placed in an ice bath, purged with dry nitrogen, and maintained under positive pressure with nitrogen throughout the functionalization. Lithium aluminum hydride (LiAlH4) was added to the 3-neck boiling flask and dissolved in diethyl ether. Using anhydrous techniques, anhydrous tetrahydrofuran was added to the boiling flask containing the carboxyl-terminated TK. The resulting solution was then transferred to the dropping funnel and added to the LiAlH4 solution dropwise at 0 °C. After all of the TK solution was added, the ice bath was replaced with an oil bath and the reaction mixture was refluxed at 52°C for 6–8 hours. Unreacted LiAlH4 was quenched by adding DI water dropwise followed by 1M sodium hydroxide to aid in product extraction. By-products of the reaction were filtered using a Buchner funnel and filtration flask, and a separation funnel and diethyl ether were used to extract and isolate the TK diol product. The solvent was removed by rotary evaporation and the product dried under vacuum overnight for a completely dry, solvent-free TK diol. Nuclear magnetic resonance spectroscopy (1H NMR, Bruker 400 MHz NMR) in dimethylsulfoxide (DMSO) and attenuated total reflectance Fourier transform infrared spectroscopy (ATR-FTIR) verified the chemical structure of the TK diol. Titration of a sample reacted with excess p-toluenesulfonyl isocyanate with tetrabutylammonium hydroxide was used to determine the hydroxyl (OH) number of the TK diol according to ASTM E1899-08.24 The molecular weight (Mn) was calculated from the OH number using Eq (1): (1) Mn=56,100OHNumber 2.3 Quasi-prepolymer synthesis and characterization A quasi-prepolymer was prepared according to methods previously described.25 Briefly, a 2.5:1 molar ratio of LTI:TK (3.75:1 NCO:OH equivalent ratio) was charged to a 100-mL boiling flask and purged with nitrogen while stirring in an oil bath at 45°C. TK diol was added to LTI drop-wise from a syringe through a 16G needle inserted through the rubber stopper. The reaction was allowed to proceed for 3 hours yielding an LTI-TK quasi-prepolymer. The NCO number was determined by titration according to ASTM D2572-97.26 2.4 Polyurethane/ceramic composite synthesis and characterization PTKUR/ceramic composites were fabricated by reactive liquid molding and catalyzed using a 5% FeAA solution in ε-caprolactone. The isocyanate index (NCO:OH equivalent ratio * 100) was 140 for all materials.8 TK diol, LTI-TK prepolymer, and 55 wt% MG or 60 wt% nHA particles were hand-mixed to yield a reactive paste. These concentrations of the ceramic particles were selected as the maximum values that could be added while maintaining a cohesive reactive paste. Once homogeneous, 0.06 wt% FeAA (in solution) was added to catalyze the reaction between the LTI-TK prepolymer and the TK diol. The morphology of the composite was verified by scanning electron microscopy (Hitachi S4200 SEM) following gold sputter coating of thin sections of sample (Cressington Q108) for 45 seconds at 30 mA. PTKUR films (without ceramic) synthesized using varying isocyanate indices were submerged in water for 2 weeks and water uptake measured periodically by weighing the samples. Swelling of films with indices of 110, 125, and 140 was calculated according to Eq (2), where Ms is the swollen mass and M0 is the initial mass. This information was used to determine effects of index on extent of crosslinking. (2) %Swelling=MS-M0M0×100% 2.4.1 Reaction kinetics and working time The reaction kinetics of the composite were assessed using methods described previously.8, 27 ATR-FTIR was used to evaluate the reaction rate of the isocyanate-terminated LTI-TK prepolymer with the other components of the composite individually by quantifying the disappearance of the isocyanate peak (around 2270 cm−1). The isocyanate peaks were calibrated to a standard curve of known NCO concentrations to find an initial rate constant for each reaction during the first 6 minutes. These rate constants along with the initial concentrations of each component were input into a Matlab program to calculate the number of isocyanate and hydroxyl equivalents versus time assuming second-order chemical kinetics. Isocyanate and hydroxyl conversion versus time were determined from the calculated numbers of equivalents. The working time for the MG composites was defined using a rheometer with 25-mm plates. A gap size of 1.5 mm and constant strain (1%) and frequency (1 Hz) were applied to the composite and the working time defined by the time of the G′ – G″ crossover point. This time was compared to the tack-free time which was defined as when the material no longer stuck to a metal spatula.7 2.4.2 Compressive mechanical properties Samples for compressive studies were prepared by injecting composites into 6 mm diameter tubes and compressing under a 0.96 kg weight to ensure cohesion throughout initial cure.28 Samples were cut to a height equal to 2 times their diameter (12 mm) using a Buehler IsoMet Low Speed Saw (Lake Bluff, Il). Modulus and strength were measured at various time points over a 2-week period to determine when the composites were completely crosslinked. Specimens were preloaded to 12 N and compressed at a rate of 25 mm min−1 using an MTS 858 Bionix Servohydraulic Test System (Eden Prairie, MN). The engineering stress was calculated by dividing the load by the platen-contacting surface area and the engineering strain determined by dividing the displacement by initial sample height. The slope of the linear-elastic portion of the resulting stress-strain curve was identified as the compressive modulus and the maximum stress as the compressive strength. When a maximum stress could not be identified, the stress at 10% strain was reported.29 2.4.3 Degradation The degradation characteristics of PTKUR were assessed in hydrolytic and oxidative conditions. An accelerated degradation medium comprising 20 wt% hydrogen peroxide in 0.1 M cobalt chloride in DI water simulated the environment produced by reactive oxygen species at the implant site.20, 30, 31 PTKUR films (17 mg) were immersed in 350 μL (1 mL/50 mg initial sample) degradation media and placed on a shaker table at 37°C. PTKUR degradation was compared to lysine-derived poly(caprolactone urethane) (PCLUR), which was expected to undergo minimal hydrolytic degradation. Oxidative media was changed every 72 hours when time points exceeded 3 days to ensure the presence of oxidizing radicals. Samples were washed 3X with 100 mL DI water, dried under vacuum for at least 48 hours, and weighed at various time points to determine the degradation rate. Samples were gold sputter-coated for 45 seconds and imaged using SEM to visualize the change in architecture with degradation. 2.5 Rheology Viscosity was characterized using a TA Instruments AR 2000ex rheometer fitted with 25-mm parallel plates at 25 °C. For the starting materials (TK diol and LTI-TK prepolymer), a small sample was injected between the plates which were subsequently depressed to a gap size of 500 μm. A frequency sweep was applied at a constant strain in the linear viscoelastic region (0.2 for the TK diol and 0.5 for the quasi prepolymer). A Cox-Merz transformation related the dynamic data to viscosity as a function of shear rate. The rheological properties of uncatalyzed (non-reactive) composites were found using a gap size of 1.5 mm. A constant strain of 1% was applied to the composite through a frequency sweep and a Cox-Merz transformation applied to characterize injectability. 2.6 In vitro characterization The surface chemistry of PTKUR polymer films was observed by water contact angle using a Ramé-Hart Goniometer (Mountain Lakes, NJ) to predict cellular behavior at the material interface. Cellular attachment was verified using SEM and proliferation was observed using a BCA Protein Assay kit. MC3T3 cells were seeded (2 × 104 cells/mL) onto thin sections of MG and nHA composites that were conditioned in complete αMEM medium with 10% FBS and 1% P/S overnight. Samples were submerged in 5% glutaraldehyde followed by 2% osmium tetroxide and an ethanol dehydration ladder to fix for SEM after 24 hours incubation. To measure proliferation, samples were taken from culture at 1, 4, and 7 days. Samples were transferred to a new well, washed with PBS, and the cells trypsinized. Cell pellets were lysed using RIPA buffer to extract the cellular protein. The BCA kit was used to quantify total protein at each time point.32 2.7 Implantation of PTKUR/ceramic composite cements in rabbits Animal experiments were performed in compliance with relevant laws and the guidelines of IBEX Preclinical Research, Inc. (Logan, UT). The animal protocol was approved by the Institutional Animal Care and Use Committee (IACUC) at IBEX Preclinical Research, and all surgical and care procedures were carried out under aseptic conditions per the approved IACUC protocol. PTKUR/ceramic composites were evaluated in cylindrical femoral condyle plug defects in eight New Zealand White rabbits weighing 4–5 kg. The reactive components (TK diol, FeAA catalyst, LTI-TK prepolymer, MasterGraft, and nHA) were gamma-irradiated using a dose of approximately 25 kGY prior to use. After administration of anesthesia, bilateral defects 6–8 mm deep × 5 mm diameter were drilled in the femoral condyle of the distal femurs of 8 rabbits. PTKUR/ceramic composites incorporating either MG or nHA (n=3) were mixed on site, injected into the defect, and allowed to cure for 10 minutes prior to closing the wound. Animals were euthanized and femurs harvested at 6 and 12 weeks to evaluate healing and polymer degradation. Micro-computed tomography (Scanco μCT 50) was performed with a voxel size of 17.2 μm and a threshold of 237 (386 mg HA/cm3) to match the intensity of the native trabecular bone surrounding the defect. Histology preparation was performed by Histion. Calcified samples were embedded in PMMA and sections taken from the center of the defect area; the sections were stained with Stevenel’s Blue or hematoxylin and eosin (H&E) to identify new bone formation and cellular activity at the defect site. 2.8 Statistical analysis Anova with post hoc comparisons using Tukey’s multiple comparisons test was applied to compression testing data to compare statistical differences with cure time. The Holm-Sidak multiple comparison test was used to evaluate significance in total protein over time for each composite individually, and the plot shows standard error of the mean (SEM). All other data was plotted with standard deviation, and p < 0.05 was considered statistically significant. 3. Results and Discussion 3.1 Thioketal diol and quasi-prepolymer characterization The TK diol was synthesized following the two-step reaction scheme in Figure 1A. The characteristic NMR peak for the methyl (1.59 ppm) and hydroxyl (4.8 ppm) groups of the TK diol indicate that the targeted product was achieved (Figure 1B), and an ATR-FTIR absorbance peak around 3400 cm−1 confirmed hydroxyl functionalization.20 The OH number was found to be 574 mg KOH/g, which corresponds to a molecular weight of 196 g mol−1 (Eq 1). These data confirm that the desired product with a theoretical molecular weight of 196.3 g mol−1 was achieved. This low-molecular weight TK diol had a viscosity of 0.11 Pa s at a shear rate of 5 s−1 and exhibited near Newtonian behavior at shear rates below 100 s−1 (Figure 1C). A quasi-prepolymer was synthesized to improve handling by increasing LTI viscosity, lowering the reaction exotherm, and minimizing phase separation during polymerization. TK diol was reacted with a 2.5 molar excess of LTI to form an LTI-TK prepolymer (Figure 2A). The excess of LTI greater than 2 renders this component a quasi-prepolymer, although it will be referred to as a prepolymer in this study.33 The LTI-TK prepolymer exhibited Newtonian behavior, but the viscosity of 61 Pa s (measured at 5 s−1, Figure 2B) was considerably greater than that measured for TK diol or LTI (0.036 – 0.061 Pa s).25 The %NCO number of the prepolymer determined by titration was 25.1%, which is slightly lower than the theoretical NCO number of 26.7% based on stoichiometry. 3.2 Composite characterization Crosslinked PTKUR composites (Figure 2C) incorporating either MG or nHA particles were fabricated according to the schematic in Figure 2D. Figure 2E shows the initial (e.g., uncatalyzed) dynamic viscosities of both MG and nHA composites up to shear rates of 100 s−1. Both materials exhibit shear thinning behavior that is more prominent at lower shear rates, which enhances injectability, and have viscosities of 20–25 Pa s at a shear rate of 100 s−1. SEM images of the composites showed minimal porosity was achieved using a low-toxicity, iron-based gelling catalyst (25:1 gel:blow, Figure 2F–G) compared to previously investigated amine-based catalysts with high blowing power (1:20 gel:blow).10, 13, 23, 34 PTKUR films were made by mixing TK and LTI-TK prepolymer with iron catalyst without incorporating ceramic particles. The polymer film exhibited a contact angle of 70.2° indicating a moderately hydrophobic surface. Films of indices 110, 125, and 140 all swelled less than 3.5% after soaking in water for 2 weeks and the differences between them were not significant. Since there was no difference in swelling and the swelling was less than 5% for all samples, all of the indices were considered suitable for use in vivo. An index of 140 was chosen for the studies in this work to ensure complete crosslinking and a more rigid composite as reported previously.8, 28, 35 The reactivity of the polymer was investigated using ATR-FTIR. The second-order rate constant (ki, Eq (3)) of each component was calculated based on the initial isocyanate concentration (C0) and the disappearance of the isocyanate peak (C).8, 36 (3) 1C=kit+1C0 The catalyst was reduced by half (compared to the in vivo studies) for the reactivity experiments to slow the reaction, which was necessary to investigate the reaction mechanisms. Figure 3A shows the calculation of the initial rate constant (ki) for each reaction from the slope of the 2nd order rate plot, in which the inverse concentration of NCO equivalents (g / equiv NCO) is plotted versus time. The plot is linear for the first 6 minutes of the reaction, which confirms that the reactions are second order as anticipated.8, 36 Further, the very small slope for MG, nHA, and water with LTI-TK indicates these components have very low reactivity, and thus they were not included in the conversion calculations. The relatively high rate constant for the LTI-TK/TK gelling reaction compared to the LTI-TK/water blowing reaction (25:1 gel:blow ratio) confirms the preferential gelling activity of the iron acetylacetonate (FeAA) catalyst compared to the triethylene diamine (TEDA) catalyst investigated previously (1:20 gel:blow).8, 35 The concentration of LTI-TK prepolymer (I) and TK diol (D) were calculated as: (4) dCDdt=dCIdt=-kDCDCIM where Cj is the concentration of each component (I or D, g equiv−1 min−1) and M is the mass of the composite (g). The conversions of LTI-TK prepolymer and TK diol were calculated from the second-order kinetic model as: (5) ξj=Cj0-CjCj0 Conversion of NCO and OH groups are shown in Figure 3B. The hydroxyl groups in the TK diol are completely converted and an excess of isocyanate functional groups remain, as anticipated from the high isocyanate index of 140. The excess isocyanate is anticipated to slowly react with the ceramic and environmental water, as reported previously for allograft bone composites9, due to the substantially lower reactivity of the LTI-TK prepolymer with these components. The tack-free time was determined by hand to be 6 minutes37 after mixing, which is comparable to the setting times for calcium phosphate cements. 2 MG composites achieved a maximum compressive yield strength of 40 ± 7 MPa and modulus of 936 ± 46 MPa after 1 week of curing in air at RT (Figure 4A–B). These composites had an initial strength of 7.7 MPa and modulus of 36 MPa after 16 hours curing at RT. nHA composites exhibited initial strength and moduli much greater than MG composites as expected due to the increased surface area-to-volume ratio of the mechanically robust nanoparticles.38, 39 These cements had an initial compressive yield strength and modulus of 31 ± 3 MPa and 452 ± 35 MPa, respectively. The composites reached a yield strength of 90 ± 6 MPa and modulus of 1267 ± 277 MPa after 1 week (Figure 4C–D). The mechanical properties of both composites increased over the first week, indicating that complete crosslinking was achieved 1 week after fabrication. The physical appearance of the composites post-compression supports this finding. MG composites up to 48 hours cure time experience some elastic recovery to their original shape around 30 minutes post-compression, where plastic deformation is more evident in the 1 and 2 week samples (Figure 4E). These changes in resilience are less apparent in the stronger nHA samples (Figure 4F). Trabecular bone is reported to have a compressive strength of 5–10 MPa and modulus of 50–400 MPa.40–42 Therefore, the initial compressive strength and modulus of MG composites are close to the properties of trabecular bone and nHA composites exceed these properties. Both composites are mechanically stronger than trabecular bone after 1 week. The degradation rate of PTKUR films under hydrolytic and oxidative conditions was measured in vitro. PTKUR was compared to PCLUR as this material has been shown to degrade slowly in vivo.5 PTKUR degraded completely after 4 days in vitro in oxidative media (Figure 5A inset) but experienced minimal hydrolytic degradation in PBS after 4 months (Figure 5A). SEM images of PTKUR after 24, 48, and 72 hours in oxidative media show morphological changes in the films in response to degradation, as evidenced by the formation of pores in the material (Figure 5B–D). PCLUR degraded minimally in PBS as expected and did not completely degrade in oxidative media until about 5 months. 3.3 In vitro characterization The osteoblast precursor MC3T3 cell line was used in all in vitro studies to assess cell attachment and proliferation. SEM images show that cells attached and spread on MG (Figure 6A) and nHA (Figure 6B) composites after 24 h culture. Cell proliferation on the films was assessed for up to 7 days post-seeding by measuring the change in total protein with time. Figure 6C shows that the cell population on MG composites increased with time, but the differences were not significant. Cells proliferated on nHA composites, as evidenced by the increase in total protein from day 1 to day 7. Hydroxyapatite is the primary mineral component in bone, and therefore MC3T3 cells were expected to adhere and proliferate on scaffolds comprising 60 wt% nHA.43 While MG contains only 15% HA, the beta-tricalcium phosphate (β-TCP) component is also an osteoconductive ceramic.44, 45 The slower proliferation rate of MC3T3 cells on MG composites could potentially be explained by the relatively large size (100 – 300 μm) of the MG microparticles, resulting in relatively large areas of polymer that is less osteoconductive than the ceramic. In contrast, phase-separation of the nHA and polymer components was not observed in the nHA composites, suggesting that the nHA is more uniformly distributed due to its smaller particle size. 3.4 Tissue and cellular response in the femoral condyle defect model The composites were injected into femoral condyle plug defects in rabbits to assess bone healing and cement resorption. In vivo x-ray imaging immediately following the surgery indicated good placement and complete fill of the defect with the materials. μCT images of MG and nHA cements at 6 and 12 weeks are shown in Figure 7. Trabecular densification was evident at the periphery of the defects, indicating that the material was integrated with the host bone and initiating a healing response. Low-magnification (2X) images of histological sections stained with Stevenel’s Blue stain show appositional growth of dense trabecular bone near the host bone-cement interface at 12 weeks (Figure 8). The materials were well-tolerated by the host tissue and no adverse reactions were evident. Higher magnification (20–40X) images show remodeling and integration of the cements with host bone near the surface of the cements at 6 and 12 weeks. Due to the relatively large size of MG particles (100 – 300 μm), the PKTUR (P) and MG particles (MG) could be distinguished in the histological sections. PTKUR resorption near the interface was observed, resulting in cellular infiltration and new bone (NB, red) formation. Osteoid (arrows) was observed near the surface of the residual PTKUR. While the nHA particles were too small to distinguish in the histological sections, similar phenomena were observed for nHA cements. Resorption of the cement (CM) near the host bone interface resulted in new bone formation and osteoid was evident near the surface of the cement. Resorption appeared to be cell-mediated, as indicated by the irregular morphology of the cement (black arrows, Figure 9) and the presence of osteoclast-like cells, identified as large (>50 μm) multi-nucleated (nuclei stained dark blue, Figure 9) cells, near the bone-cement interface. In contrast, negligible degradation was observed in the interior of the cement. These findings are consistent with the notion that resorption of the cements was surface-mediated by osteoclasts and/or macrophages through an ROS mechanism (Figure 5) as we have reported previously for PTKUR scaffolds implanted in cutaneous wounds.20 Due to their relatively large size (100–300 μm), MG particles can be observed in the SEM images as a distinct phase (“MG”, Figure 2F). Since osteoclasts are smaller than MG particles, resorption of the MG and PTKUR phases is anticipated to proceed at different rates. In contrast, the smaller nHA particles (100 nm) cannot be distinguished from the PTKUR component (Figure 2G). At the length scale of an osteoclast, the nHA composites comprise a single phase and are anticipated to resorb at a rate averaged over the resorption rates of the individual nHA and PTKUR components. Thus, differences in MG and nHA particle size may affect graft resorption. Due to the low (<10%) porosity of the cements, the rate of cellular infiltration and remodeling was slow. Increasing the porosity would be anticipated to accelerate infiltration of cells and consequent new bone formation.4 4. Conclusion In this study, a novel low-molecular weight thioketal diol crosslinker was synthesized to prepare cell-degradable bone cements with initial bone-like strength. The cements exhibited initial compressive strength exceeding that of trabecular bone, working times comparable to commercial bone cements (5 – 10 min), and degradation in response to reactive oxygen species secreted by cells. When implanted into femoral condyle plug defects in rabbits, the cements supported appositional new bone growth, osteoclast-mediated resorption, and integration with host bone. These findings highlight the potential of poly(thioketal urethane)/ceramic composite bone cements for repair of bone damaged by trauma or disease. Supplementary Material Fig Captions ToC The authors acknowledge Dr. Mike Larson and his team at IBEX Preclinical Research, Inc. for performing the rabbit surgeries and providing post-operative care. This research was supported in part by an appointment to the Student Research Participation Program at the U.S. Army Medical Research and Materiel Command administered by the Oak Ridge Institute for Science and Education through an interagency agreement between the U.S. Department of Energy and USAMRMC. This work was also supported by the National Institutes of Health under award numbers R01AR064304 and R01EB019409 and the National Science Foundation Graduate Research Fellowship Program under Grant No. 1445197. The rabbit study was funded by Medtronic Spinal and Biologics. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Institutes of Health or the National Science Foundation. Figure 1 Synthesis and characterization of low molecular weight thioketal diol (A) Synthesis scheme. (B) Characterization by NMR indicates that the targeted molecular structure was obtained. (C) Viscosity of the TK diol is independent of shear rate. Figure 2 Synthesis of poly(thioketal urethane) (PTKUR)/ceramic composites (A) Synthesis scheme for LTI-TK prepolymer. (B) Viscosity of the LTI-TK prepolymer is independent of shear rate. (C) Reaction of TK diol with LTI-TK prepolymer to form a crosslinked PTKUR network. (D) Fabrication of PTKUR/ceramic composites by mixing LTI-TK prepolymer, TK diol, and ceramic particles (MG or nHA). (E) The viscosity of uncatalyzed (non-reactive) LTI-TK/TK diol/ceramic mixtures decreases with increasing shear rate, providing evidence of shear-thinning behavior. (F–G) SEM images of (F) MG and (G) nHA composites show lack of porosity. Due to their relatively large size (100 – 300 μm), MG particles (light grey and labeled MG) can be distinguished from the PTKUR phase (dark grey). Figure 3 Kinetics of the setting reaction (A) Second-order reaction kinetics plot of inverse NCO concentration (1/[NCO], g / equivalents NCO) versus time. The specific reaction rate (k) was calculated from the slope of the 1/[NCO] versus time plot, which is anticipated to be linear for second-order chemical reactions. The rate constant of the LTI-TK prepolymer-TK diol reaction (□) is substantially greater than that measured for MG (△), nHA (○), or water (⋄). (B) Using the rate constant for the dominant reaction TK diol + LTI-TK prepolymer and the second-order kinetics model, the conversions of the NCO and OH functional groups were calculated versus time. Figure 4 Mechanical properties of PTKUR/ceramic composites under static compressive loading. (A) Yield strength and (B) modulus of PTKUR/MG composites measured versus time for up to two weeks. (C) Yield strength and (D) modulus of PTKUR/nHA composites measured versus time for up to two weeks. Maximum compressive properties were achieved after 1 week cure time. The physical appearance of (E) MG and (F) nHA composites after compressive testing supports this finding. Figure 5 Degradation of PCLUR and PTKUR films (A) PTKUR and PCLUR films were incubated in hydrolytic (open symbols) or oxidative (closed symbols). After 4 months, PCLUR and PTKUR films substantially degraded in oxidative medium, while no degradation was observed in hydrolytic medium. Furthermore, PKTUR films degraded after only 4 days incubation time in oxidative medium (inset). (B–D) SEM images show the effects of oxidative degradation on the architecture of the PTKUR films after (B) 24 h, (C) 48 h, and (D) 72 h. Figure 6 MC3T3 cells (arrows) attached and spread on (A) MG and (B) nHA composites after 24 h incubation. Scale bar = 50 μm. (C) Measurements of total protein versus time indicate that cells proliferated faster on nHA composites. Figure 7 Images of transverse μCT sections of PTKUR/MG and PTKUR/nHA composite cements explanted at 6 and 12 weeks Higher magnification images of the defect periphery show evidence of trabecular infiltration (single white arrows) and trabecular densification (double white arrows). Scale bar = 1 mm. Figure 8 Images of transverse histological sections of PTKUR/MG and PTKUR/nHA composite cements Low-magnification (2X) images of cements at 12 weeks show appositional growth of dense trabecular bone near the host bone-cement interface. Higher magnification (20–40X) images of PTKUR/MG cements at 6 and 12 weeks reveal evidence of residual MG (dark grey) particles, resorption of PTKUR (P, light grey), cellular infiltration (blue), osteoid (arrows), and new bone (NB, red) formation. Similar observations were made for PTKUR/nHA cements, but the nHA particles could not be distinguished due to their small size. Resorption of the cement (CM) was evident in the histological sections. Figure 9 Resorption of PTKUR/MG and PTKUR/nHA cements mediated by osteoclast-like cells at 6 and 12 weeks Osteoclasts are identified as large (>50 μm) multi-nucleated (nucleus stains dark blue) cells near the host bone-cement interface. 1 Kühn K-D Bone cements: up-to-date comparison of physical and chemical properties of commercial materials Springer 2000 2 Bohner M Eur Cell Mater 2010 20 3 10 3 Wagoner Johnson AJ Herschler BA Acta Biomaterialia 2011 7 16 30 20655397 4 Adhikari R Gunatillake PA Griffiths I Tatai L Wickramaratna M Houshyar S Moore T Mayadunne RT Field J McGee M Carbon T Biomaterials 2008 29 3762 3770 18632149 5 Dumas JE Davis T Holt GE Yoshii T Perrien DS Nyman JS Boyce T Guelcher SA Acta Biomaterialia 2010 6 2394 2406 20109586 6 Bonzani IC Adhikari R Houshyar S Mayadunne R Gunatillake P Stevens MM Biomaterials 2007 28 423 433 16979756 7 Dumas JE Zienkiewicz K Tanner SA Prieto EM Bhattacharyya S Guelcher SA Tissue Engineering Part A 2010 16 2505 2518 20218874 8 Page JM Prieto EM Dumas JE Zienkiewicz KJ Wenke JC Brown-Baer P Guelcher SA Acta biomaterialia 2012 8 4405 4416 22871639 9 Dumas JE Prieto EM Zienkiewicz KJ Guda T Wenke JC Bible J Holt GE Guelcher SA Tissue Engineering Part A 2014 20 115 129 23941405 10 Hafeman AE Zienkiewicz KJ Zachman AL Sung H-J Nanney LB Davidson JM Guelcher SA Biomaterials 2011 32 419 429 20864156 11 de Bruyn PP Kabisch WT American Journal of Anatomy 1955 96 375 417 13248819 12 Dumas JE BrownBaer PB Prieto EM Guda T Hale RG Wenke JC Guelcher SA Biomedical Materials 2012 7 024112 22456057 13 Prieto EM Talley AD Gould NR Zienkiewicz KJ Drapeau SJ Kalpakci KN Guelcher SA Journal of biomedical materials research Part B, Applied biomaterials 2015 10.1002/jbm.b.33349 14 Loi F Córdova LA Pajarinen J Lin T-h Yao Z Goodman SB Bone 2016 86 119 130 26946132 15 Schmidt-Bleek K Kwee BJ Mooney DJ Duda GN Tissue Engineering Part B: Reviews 2015 21 354 364 25742724 16 Nukavarapu S Freeman J Laurencin C Regenerative Engineering of Musculoskeletal Tissues and Interfaces Woodhead Publishing 2015 17 Sheweita SA Khoshhal KI Baghdadi HH Systems Biology of Free Radicals and Antioxidants Springer 2014 2973 2995 18 Garrett I Boyce B Oreffo R Bonewald L Poser J Mundy G Journal of Clinical Investigation 1990 85 632 2312718 19 Key LL Wolf WC Gundberg CM Ries WL Bone 1994 15 431 436 7917583 20 Martin JR Gupta MK Page JM Yu F Davidson JM Guelcher SA Duvall CL Biomaterials 2014 35 3766 3776 24491510 21 Wilson DS Dalmasso G Wang L Sitaraman SV Merlin D Murthy N Nat Mater 2010 9 923 928 20935658 22 LeGeros RZ Clinical orthopaedics and related research 2002 395 81 98 23 Talley AD Mcenery MAP Kalpakci KN Zienkiewicz KJ Shimko DA Guelcher SA Journal of Biomedical Materials Research Part B: Applied Biomaterials 2016 in press 24 A International, Journal 2016 E1899 16 25 Guelcher SA Srinivasan A Dumas JE Didier JE McBride S Hollinger JO Biomaterials 2008 29 1762 1775 18255140 26 A International, Journal 2010 27 Guo R Ward CL Davidson JM Duvall CL Wenke JC Guelcher SA Biomaterials 2015 54 21 33 25907036 28 Harmata AJ Uppuganti S Granke M Guelcher SA Nyman JS Journal of the mechanical behavior of biomedical materials 2015 51 345 355 26282077 29 A International, Journal 2016 30 Schubert MA Wiggins MJ Anderson JM Hiltner A Journal of Biomedical Materials Research 1997 34 519 530 9054535 31 Christenson EM Anderson JM Hiltner A Journal of Biomedical Materials Research Part A 2004 70A 245 255 32 Guo R Lu S Page JM Merkel AR Basu S Sterling JA Guelcher SA Advanced healthcare materials 2015 4 1826 1832 26121662 33 Dow polyurethanes - prepolymer definition http://dowac.custhelp.com/app/answers/detail/a_id/5082/~/dow-polyurethanes---prepolymer-definition accessed 05/23/2016, 2016 34 Guelcher SA Patel V Gallagher KM Connolly S Didier JE Doctor JS Hollinger JO Tissue Engineering 2006 12 1247 1259 16771638 35 Harmata AJ Ward CL Zienkiewicz KJ Wenke JC Guelcher SA Journal of Materials Research 2014 29 2398 2407 25798027 36 Parnell S Min K Cakmak M Polymer 2003 44 5137 5144 37 Ashida K Polyurethane and related foams: chemistry and technology CRC press 2006 38 Mi H-Y Jing X Salick MR Cordie TM Peng X-F Turng L-S Journal of Materials Science 2014 49 2324 2337 39 Rezwan K Chen Q Blaker J Boccaccini AR Biomaterials 2006 27 3413 3431 16504284 40 Fernando S McEnery M Guelcher S Advances in Polyurethane Biomaterials 2016 481 41 Karageorgiou V Kaplan D Biomaterials 2005 26 5474 5491 15860204 42 Muggli DS Burkoth AK Anseth KS Journal of biomedical materials research 1999 46 271 278 10380006 43 Buckwalter J Glimcher M Cooper R Recker R J Bone Joint Surg Am 1995 77 1256 1275 44 Gaasbeek RDA Toonen HG van Heerwaarden RJ Buma P Biomaterials 2005 26 6713 6719 15950278 45 Giannoudis PV Dinopoulos H Tsiridis E Injury 2005 36 S20 S27 16188545
PMC005xxxxxx/PMC5123598.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101316957 35616 Stem Cell Res Stem Cell Res Stem cell research 1873-5061 1876-7753 27879215 5123598 10.1016/j.scr.2016.09.004 NIHMS817581 Article A human VE-cadherin-tdTomato and CD43-green fluorescent protein dual reporter cell line for study endothelial to hematopoietic transition Jung Ho Sun 13 Uenishi Gene 123 Kumar Akhilesh 1 Park Mi Ae 1 Raymond Matt 1 Fink Dustin 1 McLeod Ethan 1 Slukvin Igor 124 1 Wisconsin National Primate Research Center, University of Wisconsin, Madison, WI 53715, USA 2 Department of Pathology and Laboratory Medicine, University of Wisconsin, Madison, WI 53792, USA 4 Correspondence: [email protected] 3 This authors contributed equally to this work 29 9 2016 13 9 2016 9 2016 13 9 2017 17 2 401405 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Human embryonic stem cell line WA01 was genetically modified using zinc-finger nucleases and the PiggyBac/transponson system to introduce a fluorescence reporter for VE-cadherin (VEC; tdTomato) and CD43 (eGFP). Phenotypic and functional assays for pluripotency revealed the modified hES cell reporter lines remained normal. When the cells were differentiated into hematoendothelial lineages, either by directed differentiation or direct reprogramming, flow cytometric and fluorescence microscopy showed that VEC+ endothelial cells express tdTomato and CD43+ hematopoietic progenitors express eGFP. Endothelial-to-hematopoietic transition (EHT) is a unique process during developmental process that gives rise to blood cells, including hematopoietic stem cells (HSCs). Modeling EHT in a culture in vitro is essential for identifying the molecular program involved in HSC specification. To track EHT from during human embryonic stem cells (hESCs) differentiation into hematoendothelial lineages, we made a VE-cadherin-tTdTomato/CD43-eGFP (VEC-tdTomato/CD43-eGFP) dual reporter fluorescent cell line based on the WA01 hESC line. VE-cadherin (CDH5, CD144) is the one of the most specific marker of endothelial cells [2–4], while CD43 (leukosialin/SPN) is a pan-hematopoietic marker [5] that is expressed by all hematopoietic progenitors during hPSC differentiation cultures [6, 7] and HSCs in the embryo and adult bone marrow [8–10]. A construct containing the CD43 promoter driving eGFP expression [11] was targeted to the AAVS1 locus by ZFN nuclease (Fig. 1A). After integrating the CD43-eGFP construct to the AAVS1 locus, we isolated and selected specific clones that had correct integration of the construct into the AAVS1 locus without random integration as determined by southern blot (Fig. 1B). We subcloned the −3394/+39 VE-cadherin promoter followed by the tdTomato into the PiggyBac vector and co-electroporated the construct with transposons into the selected CD43-eGFP reporter cell clones (Fig. 1A). After isolating of individual VEC-tdTomato/CD43-eGFP dual reporter clones, we selected and expanded clone designated H1.CD43/CD144DR. We confirmed the identity of the generated dual reporter (DR) clonal cell line by Short Tandem Repeat (STR) analysis (Table 1). DR cell line was assayed for pluripotent phenotype by measuring the expression of transcriptional factors, OCT4, SOX2, and NANOG. The DR cell line retained expression of pluripotency markers at the level similar to wild type H1 hESCs (Fig. 1C). As evaluated by embryoid body differentiation, the DR cell line showed capacity to produce all three germ layers following spontaneous differentiation (Fig. 1D). Also, we confirmed that generated DR cell line has normal karyotype (Fig. 1E) To determine the specificity and functionality of the DR cell line, we differentiated it for 8 days in chemically defined conditions [12] and analyzed for the expression of VEC and CD43 with antibodies and their respective fluorescent protein. Typically, VEC+ hemogenic endothelial progenitors are first detected on day 4 of differentiation, while CD43+ hematopoietic progenitors can be detected from day 5 of differentiation (Fig. 2A). Fluorescence microscopy showed tdTomato and eGFP were detected from day 4 and day 5, respectively, and continued through day 8 (Fig. 2A, B). Flow cytometry analysis revealed the tdTomato+ cells were exclusively detected in the VEC+ endothelial population (Fig. 2B) while GFP+ cells were almost exclusively detected in CD43+ hematopoietic cells (Fig. 2B). Consistent with the current understanding of hematopoietic development, CD43-GFP+ cells arose from VEC-tdTomato+ hemogenic endothelial cells from day 5. Recently we demonstrated that overexpression of ETV2 and GATA2 in undifferentiated hPSCs induces the formation of hemogenic endothelium which undergoes EHT [13]. Following infection of VEC-tdTomato/CD43-GFP dual reporter hESCs with ETV2 and GATA2-expressing lentiviruses, we observed expression of tdTomato+ within the endothelial population which subsequently underwent transition to eGFP+ blood cells between days 4 and 6 (Fig. 2C and 2D and supplementary movie 1). Overall, these results demonstrate that the dual VEC-tdTomato/CD43-eGFP reporter hESC line allows for easy visualization of EHT and provides a useful tool for assessing molecular factors involved in EHT regulation. Materials and Methods Maintenance and Hematopoietic differentiation of hESCs H1 (WA01, Wicell, Madison, WI) hESCs were cultured on vitronectin (Stem Cell Technologies), in E8 medium [14]. Cells were passaged every 5 days (80% confluency) using 0.5 mM EDTA in PBS. hESCs were differentiated in collagen IV-coated plate as previously described [12]. Vector construction ZFN vectors targeting the AAVS1 locus and AAVS1-SA-2A-PURO-CD43 promoter-eGFP vector are kindly provided by Paul Gadue [11]. The VE-Cadherin promoter-tdTomato construct was cloned to Piggybac transposon vector (Transposagen). The VE-Cadherin −3394/+39 promoter region was amplified by PCR from BAC clone. Gene targeting of the AAVS1 locus and PiggyBac system Cells were dissociated into single cells by treatment with TrypLE (Life technologies). One million cells were resuspended in 100 μl reagent (1 × 106 cells) of Amaxa human stem cell nucleofector kit 2 (Lonza) with 1 μg ZFN-left and right plasmid each and 10 μg AAVS targeting plasmid. For targeting PiggyBac, 0.5 μg transposase plasmid and 5 μg transposon plasmid were electroporated using program A-13 according to manufacture protocol (Amaxa). The electroporated cells were resuspended with E8 (Stem Cell Technologies) culture medium and rock inhibitor (10 μM, Tocris Y-27632) and then they were plated and cultured with E8 growth medium in 6-well plate. Puromycin and Zeocin selection (0.5 μg/ml, Life technologies) was started 3 days after electroporation. After 10 days, surviving colonies were singularized and sorted to 96 well plate as single cells, from which they were expanded individually. Southern blot Southern blot (SB) analysis was performed, as described in the protocol included in the DIG-labeling hybridization (Roche). Briefly, 10 μg genomic DNA was digested using restriction enzymes for O/N, separated on a 0.7% agarose gel, transferred to a nylon membrane (Amersham) with DIG-labeling probes. The external probe 1 was 600 base pair ApaLI fragment in 5′ external region, and internal probe1 was 600 base pair EcoRV fragment in internal region of AAVS-CD43 promoter-GFP. Embryonic body (EB) formation Cell harvested and cultured in ultra-low attachment well (Corning) to induce EB formation. EBs were cultured for 21 days with E6 media. Medium was changed every 2–3 days. EB were harvested at 14 days and 21 days and extracted RNA and analysis by RT-PCR with marker of three germ layers (Table 2). Flow cytometry Cells were dissociated into single cells by treatment with 1× Tryple. Cells analyzed using MACSQuant 10 (Miltenyi Biotech). Antibodies used included CD43-APC-Vio770 (Miltenyi Biotech) and VEC-VioBlue (Miltenyi Biotech). Intracellular staining to determine OCT4 (BD), NANOG (BD), and SOX2 (BD) expression was performed using FIX & PERM cell permeabilization reagents (BD). Time-Lapse Microscopy Endothelial to hematopoietic transition (EHT) was monitored by time-lapse microscopy using fluorescent optics. H1ESCs line expressing VE-Cadherin-Td tomato and CD43-GFP dual reporter were transduced with GATA2+ETV2 lentiviral constructs and cultured for 4 days until endothelial cluster were formed [13]. Time-lapse confocal imaging was performed over two days to capture blood formation. Time lapse movies were recorded using Nikon Eclipse Ti-E configured with an A1R confocal system, motorized stage (Nikon Instruments Inc. Melville, NY), and Tokai-Hit Stage Top Incubator (Tokai Hit CO., Ltd., Shizuoka-ken, Japan) at 37 °C and 5% CO2. Images were acquired using Nikon Elements (NIS – element C) imaging software for every 5 min with CFI Plan Fluor DLL 10× NA 0.5 WD 2.1MM objective (Nikon Instruments Inc. Melville, NY). The time-lapse serial images were converted to Quick-time movies (.mov) and analyzed using ImageJ software (NIMH, Bethesda, MD). Supplementary Material 1 2 We thank Dr. Gadue (The Children’s Hospital of Philadelphia) for providing AAVS1-SA-2A-PURO-CD43 promoter-eGFP vector. This work was supported by funds from the National Institute of Health (R01HL116221, U01HL099773, and P51 OD011106) and The Charlotte Geyer Foundation. Figure 1 Characterization of H1.CD43/CD144DR dual reporter H1 hESCs. (A) Schematic of the constructs used for targeting of CD43-eGFP reporter into AAVS locus and VEC-tdTomato reporter by PiggyBac system. (B) Southern blot analysis of ApaL1 or EcoRV digested genomic DNA of CD43-eGFP cell. Asterisk, wild type; filled triangle, targeted; arrow, off-targeting. (C) Flow analysis of intracellular staining. Expression of transcriptional factors, OCT4, SOX2, and NANOG, by dual reporter and wild type of hESCs. (D) Differentiation of dual reporter cells to three germ layers. Expression of three germ layers markers AFP (endoderm), PAX6 (ectoderm), ETV2 and T (mesoderm), and GAPDH (internal control) in day 0, day 14 and 21 EBs of dual reporter H1.CD43/CD144DR (DR) cell line and wild type of H1 hESCs analyzed by RT-PCR. (E) Normal Karyotype (46, XY) of generated DR hESC line. Figure 2 Differentiation of dual reporter cells to three germ layers and hematopoietic cells. Kinetic analysis of CD43-eGFP and VEC-tdTomato expressed during hematopoiesis by fluorescence microscopy (A) and flow cytometry (B). Cells were differentiated and checked for eGFP and tdTomato fluorescence everyday from day 3 to day 8. VEC-reporter cell expressed tdTomato and CD43-reporter cell expressed eGFP. Scale bar is 300 μM. Gray histogram is gated on the CD43−VEC− population, green histogram is gated on the CD43+ population and represents eGFP expression, and red histogram is gated on the VEC+ population and represents tdTomato expression. (C) GATA2+ETV2 induced cells with endothelial morphology can give rise to blood cells. Transition of VEC expressing cuboidal cells (endothelial cells) into round hematopoietic cells is associated with the loss of VEC expression (red) and acquisition of CD43 expression (green). (D) Cells were treated with GATA2+ETV2 virus and checked for CD43-eGFP and VEC-tdTomato at day 0, 4, 5, and 6 by flow cytometry. Table 1 Short Tandem Repeat (STR) profiling of dual reporter with the original cell line. STR locus H1 wild typea H1.CD43/CD144 Reporter (DR) FGA 20, 24 TPOX 8, 11 8, 11 D8S1179 12, 13 vWA 15, 17 15, 17 Amelogenin X, Y X, Y Penta_D 10, 13 CSF1PO 12, 13 12, 13 D16S539 9, 13 9, 13 D7S820 8, 12 8, 12 D13S317 8, 11 8, 11 D5S818 9, 11 9, 11 Penta_E 10, 12 D18S51 17, 18 D21S11 28, 32.2 TH01 9.3, 9.3 9.3, 9.3 D3S1358 15, 15 STR data obtained from Wicell (http://hpscreg.eu/docs/uploads_link/certificate_of_analysis/28f8d7dc69240e58dad1083d492983bf.pdf). Table 2 Primer sequences for three germ layers Target Forward primer Reverse primer PAX6 GAAGATGGTGATGGGATTTC CGTTGGACACGTTTTGATTG AFP GAATGCTGCAAACTGACCACGCTGGAAC TGGCATTCAAGAGGGTTTTCAGTCTGGA ETV2 TCTTTGAAGCGGTACCAGAG GGGACCTCGGTGGTTAGTT T GACAATTGGTCCAGCCTTG GGGTACTGACTGGAGCTGGT GAPDH GAAGGTGAAGGTCGGAGT GAAGATGGTGATGGGATTTC This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. 1 Thomson J Itskovitz-Eldor J Shapiro SS Waknitz MA Swiergiel J Marshall V Jones JM Embryonic Stem Cell Lines Derived from Human Blastocysts.pdf 1998 2 Breier G Breviario F Caveda L Berthier R Schnurch H Gotsch U Vestweber D Risau W Dejana E Molecular cloning and expression of murine vascular endothelial-cadherin in early stage development of cardiovascular system Blood 87 1996 630 641 8555485 3 Lampugnani MG Resnati M Raiteri M Pigott R Pisacane A Houen G Ruco LP Dejana E A novel endothelial-specific membrane protein is a marker of cell-cell contacts J Cell Biol 118 1992 1511 1522 1522121 4 Vittet D Buchou T Schweitzer A Dejana E Huber P Targeted null-mutation in the vascular endothelial-cadherin gene impairs the organization of vascular-like structures in embryoid bodies Proceedings of the National Academy of Sciences of the United States of America 94 1997 6273 6278 9177207 5 Remold-O’Donnell E Zimmerman C Kenney D Rosen FS Expression on blood cells of sialophorin, the surface glycoprotein that is defective in Wiskott-Aldrich syndrome Blood 70 1987 104 109 2439146 6 Choi KD Vodyanik M Slukvin II Hematopoietic differentiation and production of mature myeloid cells from human pluripotent stem cells Nat Protoc 6 2011 296 313 21372811 7 Vodyanik MA Thomson JA Slukvin II Leukosialin (CD43) defines hematopoietic progenitors in human embryonic stem cell differentiation cultures Blood 108 2006 2095 2105 16757688 8 Moore T Huang S Terstappen LW Bennett M Kumar V Expression of CD43 on murine and human pluripotent hematopoietic stem cells J Immunol 153 1994 4978 4987 7525720 9 Rybtsov S Batsivari A Bilotkach K Paruzina D Senserrich J Nerushev O Medvinsky A Tracing the origin of the HSC hierarchy reveals an SCF-dependent, IL-3-independent CD43(−) embryonic precursor Stem cell reports 3 2014 489 501 25241746 10 Inlay MA Serwold T Mosley A Fathman JW Dimov IK Seita J Weissman IL Identification of multipotent progenitors that emerge prior to hematopoietic stem cells in embryonic development Stem cell reports 2 2014 457 472 24749071 11 Tiyaboonchai A Mac H Shamsedeen R Mills JA Kishore S French DL Gadue P Utilization of the AAVS1 safe harbor locus for hematopoietic specific transgene expression and gene knockdown in human ES cells Stem Cell Res 12 2014 630 637 24631742 12 Uenishi G Theisen D Lee JH Kumar A Raymond M Vodyanik M Swanson S Stewart R Thomson J Slukvin I Tenascin C promotes hematoendothelial development and T lymphoid commitment from human pluripotent stem cells in chemically defined conditions Stem cell reports 3 2014 1073 1084 25448067 13 Elcheva I Brok-Volchanskaya V Kumar A Liu P Lee JH Tong L Vodyanik M Swanson S Stewart R Kyba M Yakubov E Cooke J Thomson JA Slukvin I Direct induction of haematoendothelial programs in human pluripotent stem cells by transcriptional regulators Nat Commun 5 2014 4372 25019369 14 Chen G Gulbranson DR Hou Z Bolin JM Ruotti V Probasco MD Smuga-Otto K Howden SE Diol NR Propson NE Wagner R Lee GO Antosiewicz-Bourget J Teng JM Thomson JA Chemically defined conditions for human iPSC derivation and culture Nat Methods 8 2011 424 429 21478862
PMC005xxxxxx/PMC5123666.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 0356504 3391 Dev Growth Differ Dev. Growth Differ. Development, growth & differentiation 0012-1592 1440-169X 21492154 5123666 10.1111/j.1440-169X.2011.01261.x NIHMS830815 Article Concordance and interaction of guanine nucleotide dissociation inhibitor (RhoGDI) with RhoA in oogenesis and early development of the sea urchin Zazueta-Novoa Vanesa 1 Martínez-Cadena Guadalupe 1 Wessel Gary M. 2 Zazueta-Sandoval Roberto 1 Castellano Laura 3 García-Soto Jesús 1† 1 Departamento de Biología, División de Ciencias Naturales y Exactas, Campus Guanajuato, Universidad de Guanajuato, Apdo. postal 187, Guanajuato, Gto. 36000, México 2 Department of Molecular and Cell Biology & Biochemistry, 185 Meeting Street, Box G, Brown University, Providence, RI 02912. USA 3 Departamento de Ciencias Aplicadas al Trabajo, División de Ciencias de la Salud, Campus León, Universidad de Guanajuato, Av. Eugenio Garza Sada No. 572, Col. Lomas del Campestre sección 2, León, Gto. CP 37150, México To whom correspondence should be sent: Guadalupe Martínez-Cadena, Departamento de Biología, División de Ciencias Naturales y Exactas, Campus Guanajuato, Universidad de Guanajuato, Apdo. postal 187, Guanajuato, Gto. 36000, México. Tel. 52 4737320006 ext. 8166, Fax 4737320006 ext. 8153, [email protected] † Dr. Jesús García-Soto died prematurely during the writing of this paper. He was an advisor to many students and a friend to all. 19 11 2016 4 2011 25 11 2016 53 3 427439 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Rho GTPases are Ras-related GTPases that regulate a variety of cellular processes. In the sea urchin Strongylocentrotus purpuratus, RhoA in the oocyte associates with the membrane of the cortical granules and directs their movement from the cytoplasm to the cell cortex during maturation to an egg. RhoA also plays an important role regulating the Na+-H+ exchanger activity which determines the internal pH of the cell during the first minutes of embryogenesis. We investigated how this activity may be regulated by a Guanine-nucleotide Dissociation Inhibitor (RhoGDI). The sequence of this RhoA regulatory protein was identified in the genome on the basis of its similarity to other RhoGDI species, especially for key segments in the formation of the isoprenyl-binding pocket and in interactions with the Rho GTPase. We examined the expression and the subcellular localization of RhoGDI during oogenesis and in different developmental stages. We found that RhoGDI mRNA levels were high in eggs and during cleavage divisions until blastula, when it disappeared, only to reappear in gastrula stage. RhoGDI localization overlaps the presence of RhoA during oogenesis and in embryonic development, reinforcing the regulatory premise of the interaction. By use of recombinant protein interactions in vitro, we also find that these two proteins selectively interact. These results support the hypothesis of a functional relationship in vivo and now enable mechanistic insight for the cellular and organellar rearrangements that occur during oogenesis and embryonic development. early development GTPases oogenesis RhoGDI sea urchin INTRODUCTION Rho GTPases are small G-proteins involved in the regulation of a variety of biological pathways (Bishop & Hall, 2000). They regulate as diverse pathways as the actin cytoskeleton (Ridley et al., 1992; Nobes & Hall, 1995), signaling pathways for transcriptional activation, including the JNK/stress-activated protein kinase (Coso et al., 1995; Minden et al., 1995) and p38 mitogen-activated protein kinase cascades (Minden et al., 1995) and the serum response factor pathway (Hill et al., 1995). Rho GTPases also regulate the phagocyte NADPH oxidase (Abo et al., 1991; Knaus et al., 1991), endocytosis (Lamaze et al., 1996; Leung et al., 1999; Jou et al., 2000), macrophage phagocytosis (Caron & Hall, 1998), morphogenesis (Barrett et al., 1997; Beane et al., 2006), and epithelial cell polarization (Kroschewski et al., 1999). Moreover, Ras-mediated cellular transformation is dependent on Rho GTPases (Qiu et al., 1995). These GTPases act as molecular switches in cell signaling, alternating between an inactive, cytosolic, GDP-bound state, and an active GTP-bound state usually associated with membranes, where effector targets reside (Wei et al., 1997; Ihara et al., 1998; DerMardirossian & Bokoch, 2005). More than fifteen mammalian Rho proteins have been described including RhoA–E and G, Rac1–3, two isoforms of Cdc42hs, and TC10 (Michaelson et al., 2001). Key to the complexity of these multiple small GTPases is the mechanisms of their regulation. The activity of the GTPases is controlled by guanine nucleotide exchange factors (GEFs), which catalyze the exchange of GTP for GDP; GTPase-activating proteins (GAPs), which accelerate GTP hydrolysis (Boguski & McCormick, 1993; Lamarche & Hall, 1994; Mackay & Hall, 1998), and guanine nucleotide dissociation inhibitors (GDIs). Three distinct biochemical functions have been described for RhoGDIs. First, GDIs modulate the cycling of Rho GTPases between cytosol and membranes. GDIs maintain GTPases as soluble cytosolic proteins forming complexes inserting the geranylgeranyl moiety present in the C-terminus of the Rho protein into the hydrophobic pocket formed by the immunoglobulin-like β sandwich of the GDI (Gosser et al., 1997; Keep et al., 1997; Longnecker et al., 1999; Hoffman et al., 2000; Scheffzek et al., 2000; Grizot et al., 2001). Second, they are capable of interacting with the GTP-bound form of the GTPase to inhibit the GTP hydrolysis, blocking the interaction between the Rho proteins and their effector targets. Third, GDIs inhibit the dissociation of GDP from Rho proteins, maintaining the inactive form of the GTPase and preventing its activation by GEFs (Olofsson, 1999; Zalcman et al., 1999). Three RhoGDI isoforms have been identified in humans. The expression of RhoGDI (or GDIα/GDI1) is ubiquitous (Ueda et al., 1990; Fukumoto et al., 1990), whereas Ly/D4GDI (or GDIβ/GDI2) is hematopoietic cell-selective (Scherle et al., 1993; Lelias, 2004), and RhoGDIγ (or GDI3) is specifically expressed in lung, brain and testis (Zalcman et al., 1996; Adra et al., 1997). Both RhoGDI and D4GDI are cytosolic and form complexes with several Rho GTPases (Ueda et al., 1990; Regazzi et al., 1992; Leonard et al., 1992; Ando et al., 1992; Hiraoka et al., 1992; Zalcman et al., 1996). By contrast, RhoGDIγ is associated with vesicular membranes and exhibits specificity for interactions with RhoB and RhoG (Zalcman et al., 1996; Adra et al., 1997; Michaelson et al., 2001; Faure & Dagher, 2001). In the sea urchin Strongylocentrotus purpuratus, RhoA is expressed in oocytes (Covián-Nares et al., 2004), eggs (Cuellar-Mata et al., 2000), and embryos in early stages of development (Nishimura et al., 1998; Manzo et al., 2003; Bean et al., 2006). During oogenesis, RhoA associates with the membrane of the cortical granule and directs the cortical granule movement from the cytoplasm to the cell cortex (Covián-Nares et al., 2004). Cortical granules are secretory vesicles that are synthesized in the Golgi and then accumulate throughout the cytoplasm (Anderson, 1968; Sathananthan et al., 1985). Once in the egg cortex, they play a fundamental role in preventing polyspermy at fertilization (Abbott & Ducibella, 2001). Upon fertilization, these vesicles undergo a Ca2+-dependent exocytosis releasing their content into the perivitelline space, creating a permanent block to subsequent sperm (Wessel et al., 2001; Wessel and Wong, 2009). In addition to cortical granule translocation, RhoA activity is required for the cell shape changes essential to initiate invagination (Bean et al., 2006) and it regulates the Na+-H+ exchanger activity which determines the internal pH of the cell during the first minutes of development (Rangel-Mata et al., 2007). The inhibition of RhoA blocks cytokinesis, arresting the embryos in the one-cell stage, and decreases the rate of protein synthesis after fertilization (Manzo et al., 2003). In the present work we investigated the expression and intracellular localization of RhoGDI during oogenesis and throughout embryogenesis in sea urchin to resolve its role in the regulation of the Rho GTPase activities. MATERIALS AND METHODS Animals Adult sea urchins (Strongylocentrotus purpuratus) were collected from Pamanes (Ensenada, Baja California, México) and were maintained in an aquarium at 12 °C. Adults were induced to shed gametes by intracelomic injection of 0.5 M KCl and eggs were collected by inverting the female over a beaker with artificial seawater (ASW: 450 mM NaCl, 26 mM MgCl2, 30 mM MgSO4, 10 mM CaCl2, 10 mM KCl, 0.1 mM EDTA, 10 mM Hepes, at pH 8.0). Oocytes were handled as described by Berg and Wessel (1997). Pre-meiotic oocytes and germinal vesicle (GV)-stage oocytes were isolated using a mouth pipette. To obtain embryos, fertilized eggs were cultured at 16°C in ASW supplemented with 10 mM p-aminobenzoic acid (PABA) to remove fertilization envelopes, and collected at necessary developmental stages. Ovaries, eggs, and embryos were fixed as described (Arenas-Mena et al., 2000) and maintained at −20°C. Cell fractionation Eggs were homogenized in MI buffer (220 mM gluconic acid, 500 mM glycine, 10 mM NaCl, 5 mM MgCl2, 5 mM EGTA, 2.5 mM ATP, 2 mM DTT, at pH 6.7) using a Potter homogenizer. Cortical granules bound to the plasma membrane and the vitelline layer (cortex) were separated from the soluble fraction and mixed membrane fraction by centrifugation at 700×g. The pelleted cortex was incubated with BA buffer (50 mM Tris-HCl, 2 mM EGTA, 450 mM KCl, at pH 9.1) and centrifuged at 700×g to separate cortical granules from the plasma membrane and vitelline layer. Embryos in different developmental stages were homogenized in MI buffer and centrifuged at 150,000×g to obtain soluble and insoluble fractions. Cloning of cDNA Complete cDNA sequence for RhoGDI from S. purpuratus was obtained by RT-PCR (GenBank Accession No. XM_776534) using an ovary cDNA library and specific primers (forward: 5′-ATG GCT GAA GAG GCA-3′ and reverse: 5′-TCA CTT CCA GTC-3′). RT-PCR conditions were as follows: after denaturation for 5 min at 95°C, PCR amplification was performed for 35 cycles of denaturation for 30 sec at 95°C, annealing for 30 sec at 55°C, and extension for 60 sec at 72°C. We obtained a 603 bp product that was isolated, cloned into pGEM-T EASY (Promega, WI, USA), and sequenced. RNA analysis Whole mount in situ RNA hybridization was performed using a digoxigenin-labeled RNA probe as previously described (Arenas-Mena et al., 2000). The plasmid named pRhoGDI that contains the insert of RhoGDI, was linearized with NcoI and transcribed with SP6 RNA polymerase to yield an antisense RNA probe with the DIG RNA Labeling Kit (SP6/T7) (Roche Applied Science, IN, USA). Ovaries, eggs, and embryos were fixed, hybridized, and the signals detected essentially as described (Ransick et al., 1993). Negative controls for these experiments included the use of non-relevant transcript probes provided by the manufacturer (neomycin) not present in this embryo and thus incapable of hybridizing even to genomic DNA sequences that might complicate sense and antisense probes. Northern blot analysis of RhoGDI mRNA expression during development was performed as described using 10 μg of total RNA from ovaries, eggs, and embryos at several developmental stages (LaFleur et al., 1998). A DIG-labeled RNA probe corresponding to the complete RhoGDI ORF was synthesized with DIG RNA Labeling Kit (Roche Applied Science, IN, USA) according to kit directions. The loading levels were ascertained by OD260 measurements of the samples, as well as by the intensity of ethidium bromide staining of rRNA bands. qPCR was performed on the 7300 Real-Time PCR system (Applied Biosystems, CA, USA) with the SYBER Green PCR Master Mix Kit (Applied Biosystems, CA, USA). Primer sets were designed using primer3 (http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi) to amplify products between 100 and 150 base pairs. cDNAs (kindly donated by Dr. Jia Song, Brown University, USA) were prepared with the TaqMan® Reverse Transcription Reagents kit (Applied Biosystems, CA, USA). All qPCR experiments were run in triplicate and repeated at least once. Data for each gene was normalized against ubiquitin RNA levels and represented as a fold change relative to the amount of gene-specific RNA present in the oocyte. Antibody Generation To generate antibodies against S. purpuratus RhoGDI we made a recombinant RhoGDI protein, consisting of the complete RhoGDI ORF, ligated to a 6- His tag using the pNo-TAT vector. The fusion construct was transformed into BL21 cells for overexpression, and identified by SDS–PAGE and immunoblot analysis using anti-His monoclonal antibodies (Amersham, PA, USA) diluted 1:3000. For large-scale purification, BL21 clones overexpressing the His-RhoGDI fusion protein were cultured at 37°C in LB added with 100 μg · ml−1 ampicillin. Cells were pelleted by centrifugation at 12,000×g, for 30 min, at 4°C, and washed once with phosphate buffered saline (PBS: 137 mM NaCl, 2.7 mM KCl, 8.1 mM Na2HPO4, 1.5 mM KH2PO4). Pellets were resuspended in buffer Z (8 M urea, 100 mM NaCl, and 20 mM HEPES, pH 8.0) containing 20 mM imidazole, and sonicated prior to purification on a Ni-agarose column (Life Technologies, CA, USA), according to the manufacturer’s directions. His-RhoGDI fusion protein was eluted with buffer Z containing 100 mM imidazole, and the purity of column elutant was verified by SDS–PAGE and immunoblot analysis using anti-His monoclonal antibodies. Fractions containing the fusion protein were pooled, extensively dialyzed against water and concentrated. Antigen was purified from acrylamide gel, homogenized with potter and resuspended in Freund’s adjuvant and injected subcutaneously into New Zealand White rabbits every 2 weeks for 3 months. One week following the last boost, plasma was collected from the central ear artery and SpRhoGDI antibodies were obtained as described previously (Harlow & Lane, 1988). The IgG fraction was purified using affinity columns (Hi Trap rProtein A FF; Amersham Biosciences, PA, USA). Immunoblot analysis Different cell fractions were separated by SDS-PAGE using a 12% acrylamide gel. Western blotting was performed using anti-SpRhoGDI as the first antibody and IgG anti-rabbit conjugated with peroxidase antibody as the secondary antibody (GE Healthcare, USA). Controls included using the preimmune serum as the primary antibody to assess the specificity of the signal. Western blot of total egg homogenate probed with anti-SpRhoGDI or with anti-SpRhoGDI that has been pre-incubated overnight at 4°C with 150 μg of recombinant RhoGDI protein was used to test the specificity of the antibody blocking its capacity to recognize the RhoGDI protein present in the sea urchin samples. 2-D PAGE experiments were analyzed using Ettan IPGphor3 System (GE Healthcare, USA). In all cases the RhoGDI antibody dilution used was 1:2000. To detect whether recombinant RhoGDI protein binds RhoA, experiments were performed using methods modified from Zong et al., (1999) and Fiedler et al., (2008). Briefly, pellets obtained from centrifugation of 1.5 ml overnight growth bacteria culture transformed with RhoGDI-His or His (pNo-TAT vector containing a 6- Histidines tag without insert) were lysed in RhoA lysis buffer, sonicated three times, and centrifuged at 16,000×g to remove insoluble material. The supernatants were rotated with 250 μl of Ni-agarose beads for 4 h at 4°C, and then washed three times with lysis buffer. Bead-bound RhoGDI-His and His were incubated with 1 ml of total egg homogenate overnight at 4°C. To elute proteins, beads were incubated with loading buffer after three last washes. Samples were used for Western blotting, subjected to SDS-PAGE, and transferred to PVDF membrane. Membrane was blocked with 3% BSA (for anti-RhoA, 1:500 dil., and anti-His, 1:3000 dil.) or 3% low fat milk (for anti-SpRhoGDI, 1:2000 dil.) in TBST buffer and incubated overnight at 4°C. After the membranes were blocked, they washed three times in TBST and then incubated for 1 h at room temperature with anti-SpRhoGDI or overnight at 4°C with a polyclonal antibody against RhoA (Santacruz, CA, USA) or an antibody against 6-Histidine tag (Amersham, PA, USA), rinsed three times in TBST, and then incubated 1 h at room temperature with HRP-conjugated anti-rabbit antibody (diluted 1:2000 in TBST). Immunofluorescence microscopy and immunogold analysis Whole-mount immunofluorescent labeling was performed in oocytes, eggs and embryos that were fixed as described (Arenas-Mena et al., 2000). For immunofluorescent labeling, fixed cells were permeabilized with 1% Triton X-100 and blocked for 15 minutes using 5% low-fat milk in TBST. Samples were incubated for 1 h with anti-SpRhoGDI (dilution 1:500) or overnight with anti-RhoA (1:25). They were washed with TBST to remove unbound antibody and then incubated with Alexa-488 (1:500) for 1 h. Fluorescence images were acquired with a Zeiss LSM 510 META confocal microscope attached to AIM software. Controls were made to examine the specificity of antibodies. The association of RhoGDI with RhoA was assessed by double immunolabeling with antibodies anti-SpRhoGDI and anti-RhoA as described (Berg & Wessel, 1997). Immunofluorescent labeling was performed on 8μm paraffin sections of S. purpuratus ovaries that were fixed and processed as previously described (Laidlaw & Wessel, 1994). Paraffin sections were rehydrated in phosphate-buffered saline containing 0.05% Tween-20 (PBST) and blocked for 15 minutes using 5% low-fat milk in PBST. Then, they were incubated for 1 h with anti-SpRhoGDI (dilution 1:500), washed with PBST to remove unbound antibody and then incubated with Alexa 488 (1:500) for 1 h. After washing, the section was processed for immunolabeling with anti-RhoA (1:25) and Alexa-549-conjugated anti-rabbit IgG (diluted 1:250) as a secondary antibody. Several controls were used to examine the specificity of both primary and secondary antibodies. These included using the secondary antibodies directly, and using anti-SpRhoGDI antibody with Alexa-488 as secondary antibody followed by Alexa-549, to identify non-specific signals. For immunogold electron microscopic analysis, eggs and 25-minute embryos were fixed with 2% paraformaldehyde and 0.5% glyceraldehyde, washed 3 times with ASW and postfixed with 0.01% OsO4 and 2% aqueous uranyl acetate. Cells were then processed for immunoelectron microscopy as indicated (Wessel et al., 2000). Incubation with both anti-SpRhoGDI (1:1) and anti-RhoA (1:1) was performed overnight at 4°C and with 20-nm colloidal gold-conjugated anti-rabbit secondary antibody (1:10) for 2 h. As controls, grids with eggs and embryos were processed as above except that the primary antibody was omitted. Samples were analyzed by electron microscopy using a Jeol electron microscopy (model 1011) at 80 kV and micrographs were taken on an AMT camera system. RESULTS A RhoGDI protein sequence is present in the S. purpuratus genome Blast analysis using RhoGDI protein sequences from different organisms in the NCBI database showed a possible RhoGDI protein sequence in S. purpuratus with a 48–50% identity compared to RhoGDI of various vertebrates and invertebrates. The candidate S. purpuratus RhoGDI is a 200 amino acid protein (with a predicted molecular weight of 23 kDa). The protein encoded by the cDNA was identified as a RhoGDI on the basis of similarity to the RhoGDI proteins distributed along the entire peptide sequence, especially in residues involved in the secondary structure of the protein, in the structure of the isoprenyl-binding pocket and in regions known to interact with the Rho GTPase (Figure S1). The β-strand-rich C-terminal domain of RhoGDIs has an immunoglobulin-like fold that contains a hydrophobic pocket for insertion of the isoprenyl moiety of the Rho GTPase (Hoffman et al., 2000). RhoGDI mRNA analysis in oocytes, eggs, and embryos To analyze mRNA accumulation we used both gel blot analysis and qPCR. Total RNA isolated from ovaries, eggs, and embryos (4 cells, 32 cells, early blastula, mesenchyme blastula, and gastrula) revealed two bands, a broad band approximately at 4.8 kb and a minor band at 4.2 kb upon prolonged exposure (Figure 1A). We believe this minor band to be an artifact of the rRNA density running at a similar size, and when polyA RNA was purified from unfertilized eggs and tested by RNA gel blots analysis, it revealed a single 4.8 kb band (Figure 1A). RhoGDI mRNA levels were high in eggs and during cleavage until early blastula when they disappeared until their re-accumulation in gastrulae. Although these RNA gel blots are not quantitative, they did show that the RhoGDI mRNA does not appear to get differentially spiced or modified over the course of development, a result that could have complicated interpretation of the final protein form and how it may interact with or regulate the Rho GTPase family. To determine the dynamic nature of RhoGDI mRNA accumulation quantitively we performed qPCR using ubiquitin as an internal standard for (Figure 1B). Ubiquitin has been used as a faithful standard for these types of studies in this embryo as well as in many other embryo types (Juliano et al., 2006). The flux of mRNA determined in this approach is generally in agreement with the RNA gel blot, the most significant changes being a decrease in RhoGDI during the beginning of morphogenetic transition in the embryo; embryos undergoing blastulation and mesenchyme cell ingression have the least amount of RhoGDI mRNA. qPCR showed high concentrations of RhoGDI mRNAs in oocyte, egg and cleavage stages but the highest level was in gastrulae. To determine which cells may accumulate the RhoGDI mRNA, we tested ovaries and embryos by in situ hybridization. Ovaries from various times during the breeding season were used as were embryos from different developmental stages (Figure 2). We observed a uniform level of RhoGDI mRNA throughout oocytes and eggs, while the somatic ovarian tissue was mostly devoid of signal. In embryos, the level of RhoGDI mRNA was uniform in all cells of the early embryo and the signal decreased throughout cleavage until blastula when we could not detect signal any further (Figure 2). The re-accumulation of RhoGDI mRNA in gastrulae detected by qPCR and RNA gel blots was also apparent by in situ hybridization. In addition to broad gastrulae labeling, a distinct enrichment was seen in cells at the tip of the archenteron, which may be small micromere derivatives, and/or a subpopulation of secondary mesenchyme (Figure 2). Immunodetection of RhoGDI To investigate the intracellular localization of RhoGDI we used an immunochemical approach. We raised and purified by affinity chromatography a polyclonal antibody against the sea urchin recombinant RhoGDI protein. This antibody recognizes a single protein of 24 kDa, close to the predicted size of 22.7 kDa based on sequence analysis, and this band is present in different cellular fractions from the egg (Figure 3A). As a control for this immunological approach, we found that specific immunoreactivity of the antibody was blocked with purified recombinant RhoGDI protein used as an antigen (Figure 3B). As a further test of the target protein specificity, we measured the isoelectric point of the protein using 2D-PAGE analysis. A plasma membrane fraction was used in this analysis (Figure 3A) and the isoelectric point for RhoGDI was determined to be 4.8 (Figure 3C), which coincides closely with the predicted value for this protein (pI of 4.86). This argues that Rho GDI is not modified significantly post-transitionally, at least either in shifting its mass, or charge. Intracellular localization of RhoGDI The demonstrated specificity of the affinity purified antibody enabled its use to investigate the intracellular localization of RhoGDI. Immunolocalization of whole cells was accomplished in oocytes, eggs and embryos to determine the intracellular distribution of the protein. Results show that before initiating maturation, the RhoGDI protein in young oocytes is present throughout the cytoplasm but concentrated in the cell periphery (Figure 4A). Full-sized oocytes displaying a prominent germinal vesicle in the cytoplasm (GV-stage oocytes) show RhoGDI in the cell periphery in a similar form to the small oocytes as well as a high concentration of the protein localized in the germinal vesicle. In eggs, the RhoGDI is mainly found in the cortical region (Figure 4A). A complementary control of the experiment was made using the anti-SpRhoGDI antibody pre-incubated with 150 μg of the recombinant protein and incubated overnight at 4°C, then this mixture was added to the fixed eggs as primary antibody and the immunodetection protocol was then continued. The result showed a significantly decreased immunofluorescence signal, implying that the major or only protein identified in these experiments was RhoGDI (Figure 4B). In embryos, RhoGDI overlaps the RhoA presence along in the plasma membrane, fertilization envelope and in regions of the cytoplasm (Figure 4C). Previous studies demonstrated that the fertilization envelopes often contained pieces of the plasma membrane from the egg that snapped off during its dramatic transition at fertilization (Wessel and Wong, 2009; and references therein). Blastulae show cells intensely labeled for RhoGDI (the optical section also shows a transverse section of the basal region of the epithelial cells intended to reveal the RhoGDI distribution in different cell profiles) and might be explained in terms of the initial steps of tissue differentiation and morphogenesis, particularly with regard to RhoA function (Beane et al., 2006). Overlapping presence of RhoGDI and RhoA during oocyte development As a complement of the intracellular localization of both RhoGDI and RhoA, we tested to what extent these proteins are present in the same cellular space. Antibodies against RhoA and SpRhoGDI were used to analyze whether these proteins overlapped in the cell during oocyte maturation. Sections of ovarian tissue carrying eggs and oocytes incubated with the two antibodies both showed a strong immunolabeling (Figure 4D). In young oocytes, the co-appearance is seen dispersed throughout the cytoplasm because in this development stage, the RhoGDI-RhoA complex is located in the cytoplasm; it can be seen as a yellow fluorescence signal in the cells (Fig. 4D, c). In GV-stage oocytes RhoGDI, but not RhoA, is present in the germinal vesicle region, suggesting a different set of Rho GTPase proteins regulated by this factor during late oocyte development and maturation (Fig. 4E, c). Immunoelectron microscopic observations with both anti-RhoA and anti-SpRhoGDI antibodies revealed that RhoGDI accumulates similarly to RhoA in cortical granules of mature eggs and in the fertilization envelopes of embryos (Fig. 5). Direct protein interaction between RhoGDI and RhoA in vitro To test the interaction between RhoGDI and RhoA more directly we performed co-immunoprecipitation experiments using total egg homogenates in a variety of different lysis buffers. In all cases we did not detect a co-immunoprecipitation possibly because of the transient nature of the predicted interaction in the presence of competitive interactions with Rho GEFs and GAPs and GTP. As an alternative approach, we performed in vitro binding experiments using methods modified from Zong et al., (1999) and Fiedler et al., (2008). Recombinant-produced proteins in vitro enables strict control over the binding conditions. When followed by Western blot analysis we find that recombinant RhoGDI binds native RhoA protein selectively (Figure 6, B and C). Immunodetection of recombinant RhoGDI with the α-His antibody (Panel A) displays a single band of 36 ±2 kDa corresponding to RhoGDI-His recombinant protein. Detection using anti-SpRhoGDI revealed a broad band with the same mobility as the RhoGDI-His recombinant protein (Panel B, lane: RhoGDI-His). In the same lane we also see more signal of higher relative mobility that could correspond to incomplete RhoGDI proteins produced by the heterologous expression system. Note that the lanes corresponding to His samples alone do not detect any signal due to the absence of a recombinant protein in these samples (Panels A and B, His lanes). Following immunoprecipitation with the anti-RhoGDI, the RhoGDI-His and His-alone samples were probed with anti-RhoA (Panel C), and a clear RhoA signal is visible in the RhoGDI-His lane but not in the lane corresponding to His alone. We do however see other RhoA immunoreactivity that corresponds to unknown activity. Positive controls for these experiments were made where the native RhoGDI and RhoA proteins were immunodetected in samples of total egg homogenate (B and C panels, respectively) and negative controls were performed by incubating the total egg homogenate with the Ni+ beads in absence of the recombinant RhoGDI protein and subsequent immunodetection with α-RhoGDI and α-RhoA antibodies (B and C panels, respectively). In this case, the negative control immunodetected with anti-RhoGDI showed a faint signal due to the presence of the native RhoGDI that remains in the sample even after exhaustive washes. In this experiment the protein loaded varied depending on the samples used. Overall, these results show that recombinant RhoGDI protein binds sea urchin RhoA protein in vitro suggesting that the Rho GDI identified in this study is indeed a regulatory factor for RhoA function. DISCUSSION Small GTPases have been known to activate many cellular regulatory processes. A major function of the small G-protein, Rho, revealed recently in mammalian and fungal cells is in coordinating microfilaments for vesicle trafficking (Ridley, 2001; Symons & Rusk, 2003). RhoA is involved in translocation of the cortical granules in sea urchin oocytes (Covián-Nares et al., 2004); cortical granules are secretory vesicles of the egg that play a fundamental role in preventing polyspermy at fertilization (Abbott & Ducibella, 2001). Cortical granule translocation and meiotic maturation are inseparable and once at the surface, the cortical granules remain attached to the plasma membrane for weeks without precocious secretion (Wessel et al., 2002). Upon fertilization, these vesicles undergo a Ca2+-dependent exocytosis releasing their content into the perivitalline space, creating a permanent block to subsequent sperm binding and fusion (Wessel et al., 2001). Rho proteins are under regulation by GDI proteins, which inhibit the dissociation of the nucleotide bound to these proteins. Covián-Nares et al. (2004) proposed that RhoA is synthesized early in oogenesis and is dispersed throughout the cytoplasm in an inactive state, likely through the sequestering action of RhoGDI. Our immunofluorescent labeling experiments supported this model since an overlapping signal was present throughout the cytoplasm in young oocytes and this process is possible only if there is a RhoGDI protein to bind RhoA, making it soluble. The RhoGDI distribution changes dramatically from young oocytes to GV-oocytes where it is possible to see its accumulation in the GV, independent of RhoA. This result suggests that RhoGDI interacts with other members of the small GTPases. Upon initiation of egg maturation, RhoA is activated by first dissociating from RhoGDI, and then exchanging GDP for GTP, stimulated by a GEF. RhoA is then recruited onto the cortical granules where it remains in an active, GTP-bound state. In oocytes, some cortical granules are likely free of RhoA suggesting that RhoA associates gradually with cortical granules during maturation and remains there until after maturation. Once in the cortical granules membrane, the GTP-bound form of RhoA interacts with downstream targets to stimulate formation of, and tethering to, microfilaments (Covián-Nares et al., 2004). In this way, RhoA directs the cortical granule movement from the cytoplasm to the cell cortex (Covián-Nares et al., 2004). In addition, RhoA plays an important role during early embryogenesis regulating the Na+-H+ exchanger activity which determines the internal pH of the cell during the first minutes of development (Rangel-Mata et al., 2007). The inhibition of RhoA blocks cytokinesis, arresting the embryos in the one-cell stage, and decreases the protein synthesis rate after fertilization (Manzo et al., 2003). Eukaryotic cells require filamentous actin to regulate their shape, growth, polarization, organelle movement, endocytosis/exocytosis, DNA replication and gene transcription, although we understand less about the roles of actin filaments in germ cells. Actin filaments that are associated with the plasma membrane are important for generating cell-surface specialization areas, and also provide the driving force for remodeling cell structure (Sun & Schatten, 2006; Mabuchi, 1994). For example, it has been shown that Cdc42 is involved in the actin assembly in sea urchin eggs (Nishimura & Mabuchi, 2003) and inhibition of RhoA blocks cortical granule translocation, and microfilaments undergo a significant disorganization in sea urchin oocytes (Wessel et al., 2002; Covián-Nares et al., 2004). RhoGDIs appear to inhibit the release of GDP from Rho proteins (Fukumoto et al., 1990) and perhaps more important than their effects on the GDP/GTP cycle is the ability of GDI’s to solublize the membrane-associated Rho proteins (Isomura et al., 1991). Our binding experiments showed that a recombinant RhoGDI protein consisting of the complete sea urchin RhoGDI ORF and a 6-His tag was able to bind sea urchin RhoA protein in vitro. These results suggest the possible interaction between RhoGDI and RhoA in the cell during oogenesis and embryogenesis to alter its activity and location within the cell. RhoGDI also can interact with other members of the Rho family GTPases such as Rac1/2 and Cdc42 and not only with Rho (DerMardirossian & Bokoch, 2005). In this context we studied the interaction between RhoGDI and RhoA by means of fluorescence and EM gold-labeling immunolocalization experiments. The results showed similar localization patterns for both RhoA and RhoGDI in oocytes, mature eggs and embryos in different developmental stages suggesting that RhoGDI is in the correct place for RhoA regulation. Analysis in different cell types has revealed that RhoGDI is in excess of any single Rho GTPase, but roughly equal to the total levels of RhoA, Rac1, and Cdc42 GTPases in these cells (Michaelson et al., 2001). For example, in human neutrophils, RhoA, Rac1/2, and Cdc42 are also equimolar with overall GDI (RhoGDI and D4GDI) levels, and exist largely as cytosolic GDI complexes, with no apparent pools of free GTPases (Chuang et al., 1993). RhoGDI interaction with members of the ERM family (Takahashi et al., 1997), the tyrosine kinase Etk (Kim et al., 2002) and the p75 neurotropin receptor (p75NTR) (Yamashita & Tohyama, 2003) has been described to induce the release of RhoA from RhoGDI. However, the ERM proteins might act to physically sequester RhoGDI (Allenspach et al., 2001). Immunodetection experiments using mature egg fractions showed that in our model, RhoGDI is interacting with plasma membrane and other proteins present in the mixed membrane fraction as well as a soluble fraction. These results are in agreement with the previous report about the ability of RhoGDIs to recruit and sequester Rho proteins (Isomura et al., 1991) and also that RhoGDI is capable to interact with membrane proteins (Allenspach et al., 2001). Future studies are important to establish which other proteins could be involved in RhoA signaling pathway activating or inactivating RhoGDI. Supplementary Material s1 Figure S1. Multiple aligments of S. purpuratus and representative RhoGDI proteins. Sequences were aligned using Clustal W program and default parameters, and the output file was subsequently edited manually (Rivero et al., 2001). Identical residues are in black boxes. Secondary structure elements are indicated on top the aligned sequences (Hoffman et al., 2000; Scheffzek et al., 2000), except for helix a1 (Golovanov et al., 2001). Open circles indicate residues important for the protein–protein interface, as determined from the human crystal structures of RhoGDIa, RhoGDIb, and RhoGDIg (Hoffman et al., 2000; Schefzek et al., 2000; Grizot et al., 2001). Black circles indicate residues lining the hydrophobic pocket as determined by Hoffman (Hoffman et al., 2000). Gray circles indicates residues of the “hydrophobic triad” critical for binding of the distal isoprene unit. Accession numbers are: Drosophila melanogaster NP_649162, Danio rerio NP_998626, Homo sapiens NP_004300, Mus musculus NP_598557, Strongylocentrotus purpuratus XP_001177559, Xenopus laevis NP_001079888. We are grateful to Dra. Rosana Sánchez (IBT-UNAM) and Dr. José Raúl Mena (CINVESTAV) for providing the gold conjugated antibodies; Dra. Carmen Beltrán, Dra. Claudia Treviño, Dra Georgina Reina López, and Dr. Mario Pedraza for valuable discussions and technical support. We thank to Dra. Guadalupe Zavala for electron microscopy facilities in IBT-UNAM, and Dr. Andres Saralegui for Confocal microscopy facilities in IBT-UNAM and for their technical support in the electron microscopy and confocal microscopy, respectively. Thanks to Jia Song, Julian Wong, Celina Juliano, Ekaterina Voronina, Mamiko Yajima, Eric Gustafson and Annie Gao by all their support and advise. This work was supported by the Consejo Nacional de Ciencia y Tecnología (CONACYT) of Mexico and by the Universidad de Guanajuato through the Programa de Fortalecimiento a la Investigación Institucional (J. G. S.). V.Z.N. is recipient of a doctoral scholarship from the CONACYT. We also gratefully acknowledge support from NIH (G. M. W.). Figure 1 Relative abundance of RhoGDI mRNA in different developmental stages of S. purpuratus. Total RNAs (10 μg) from indicated developmental stages purified mRNA (1μg) were submitted to electrophoresis, transferred to a Nylon membrane, and then hybridized with a RhoGDI-ORF probe. The RNA signals were revealed by peroxidase reporter activity. rRNAs were used as load control at the bottom (A). The relative amounts of RhoGDI-RNA were estimated by qPCR. All values were normalized against ubiquitin-RNA and represented as a fold change relative to the amount of RNA present in the eggs (B). Figure 2 Detection of RhoGDI-mRNA in whole cells from different developmental stages of S. purpuratus. RhoGDI-RNA was revealed by means of Sp-RhoGDI whole mount in situ RNA hybridizations in ovaries (A), mature eggs (B), and embryos at the indicated stages (C–F). All transcripts accumulate uniformly in the oocytes and eggs. Embryos accumulate transcripts levels through cleavage and then become enriched in the gastrula at the tip of the archenteron in the gastrula and remain enriched in these cells as they are subsequently incorporated into both coelomic pouches of the pluteus (data not shown). Negative controls for these experiments included the use of non-relevant transcript probes. All images were taken at the same magnification (400x). Scale bar, 50μm. Figure 3 Immunodetection of RhoGDI in different cellular fractions of S. purpuratus eggs. A, Western blot of cellular fractions of unfertilized eggs. 80 μg of protein were loaded in each lane. B, Specificity of the anti-SpRhoGDI antibody. Western blot of egg homogenate probed with anti-SpRhoGDI or with anti-SpRhoGDI that has been pre-incubated with 150 μg of recombinant RhoGDI protein. 80 μg of protein were loaded in each lane. C, Representative 2-D SDS-PAGE. The presence of RhoGDI was revealed with anti-SpRhoGDI antibody. The sample used was a plasma membrane fraction from unfertilized eggs. Molecular weight markers are indicated on the left and the pH range at the top of the gel. Arrows inside the figure show the putative subunit of RhoGDI, and numbers are the values of the respective molecular weight and isoelectric point. Figure 4 Immunodetection of RhoGDI protein in S. purpuratus oocytes, eggs and embryos and co-localization of RhoGDI and RhoA in oocytes and eggs using a polyclonal antibody against S. purpuratus RhoGDI. A, Left panels show the detection with SpRhoGDI antibody, and right panels, the detection controls using a preimmune serum. B, Detection with SpRhoGDI antibody pre-incubated with 150 μg of the recombinant protein as primary antibody. C, Left panels show the detection with SpRhoGDI antibody and right panels, the detection with RhoA antibody. D and E, An ovary section of S. purpuratus was analyzed by indirect immunofluorescence microscopy. Young oocyte (D) and GV-oocyte (E) cells were labeled with the SpRhoGDI antibody (b) and with the anti-RhoA antibody (a). Images of a and b were merged to show overlapping distributions of RhoGDI and RhoA (c). Phase contrast pictures show the cell morphology (d). RhoGDI is present also in the germinal vesicle whereas RhoA is not. Scale bar, 50 μm. Figure 5 Localization of RhoGDI and RhoA revealed by immunoelectron microscopy. The cortex of an unfertilized egg displays that RhoGDI is present in cortical granules and the plasma membrane (A), as revealed by immunogold staining. In embryos, gold labeling was found in the fertilization envelope (D). Cortex region of an unfertilized egg shows the typical appearance of cortical granules that contains RhoA (B). In embryos, RhoA is present in the fertilization envelope as is RhoGDI (E). No antibody signal is observed in unfertilized eggs (C) and embryos (F) incubated solely with gold-conjugated antibody. CG, cortical granule; FE, fertilization envelope; PM, plasma membrane. Scale bar, 500 nm. Figure 6 Interaction between recombinant RhoGDI and native RhoA in vitro. Western blot analysis using anti-His (α-His panel), anti-SpRhoGDI (α-RhoGDI panel) and anti-RhoA (α-RhoA panel) antibodies were used to test whether recombinant RhoGDI protein is able to bind a native RhoA protein from sea urchin egg homogenates. Control lanes were loaded with 80 μg of protein; other lanes were loaded with the amount of protein obtained in each case. Abbott AL Ducibella T 2001 Calcium and the control of mammalian cortical granule exocytosis Front Biosci 6 D792 D806 11438440 Abo A Pick E Hall A Totty N Teahan CG Segal AW 1991 Activation of the NADPH oxidase involves the small GTP-binding protein p21rac1 Nature 353 668 670 1922386 Adra CN Manor D Kho JL Zhu S Horiuchi T Van Aelst L Cerione RA Lim B 1997 RhoGDIγ: A GDP-dissociation inhibitor for Rho proteins with preferential expression in brain and pancreas Proc Natl Acad Sci U S A 94 4279 4284 9113980 Allenspach EJ Cullinan P Tong J Tang Q Tesciuba AG Cannon JL Takahashi SM Morgan R Burkhardt JK Sperling AI 2001 ERM-dependent movement of CD43 defines a novel protein complex distal to the immunological synapse Immunity 15 739 750 11728336 Anderson E 1968 Oocyte differentiation in the sea urchin Arbacia punctulata, with particular reference to the origin of cortical granules and their participation in the cortical reaction J Cell Biol 37 514 539 5690449 Ando S Kaibuchi K Sasaki T Hiraoka K Nishiyama T Mizuno T Asada M Nunoi H Matsuda I Matsuura Y Polakis P McCormick F Takai Y 1992 Post-translational processing of rac p21s is important both for their interaction with the GDP/GTP exchange proteins and for their activation of NADPH oxidase J Biol Chem 267 257098 25713 Arenas-Mena C Cameron AR Davidson EH 2000 Spatial expression of Hox cluster genes in the ontogeny of a sea urchin Development 127 4631 4643 11023866 Barrett K Leptin M Settleman J 1997 The Rho GTPase and a putative RhoGEF mediate a signaling pathway for the cell shape changes in Drosophila gastrulation Cell 91 905 915 9428514 Beane WS Gross JM McClay DR 2006 RhoA regulates initiation of invagination, but not convergent extension, during sea urchin gastrulation Dev Biol 292 213 25 16458878 Berg LK Wessel GM 1997 Cortical granules of the sea urchin translocate early in oocyte maturation Development 124 1845 1850 9165131 Bishop AL Hall A 2000 Rho GTPases and their effector proteins Biochem J 348 241 255 10816416 Boguski MS McCormick F 1993 Proteins regulating ras and its relatives Nature 366 643 654 8259209 Caron E Hall A 1998 Identification of two distinct mechanism of phagocytosis controlled by different Rho GTPases Science 282 1717 1721 9831565 Chuang TH Bohl BP Bokoch GM 1993 Biologically-active lipids are regulators of Rac-GDI complexation J Biol Chem 268 26206 26211 8253741 Coso OA Chiariello M Yu JC Teramoto H Crespo P Xu N Miki T Gutkin JS 1995 The small GTP-binding proteins Rac1 andCdc42 regulate the activity of the JNK/SAPK signaling pathway Cell 81 1137 1146 7600581 Covián-Nares F Martínez-Cadena G López-Godínez J Voronina E Wessel GM García-Soto J 2004 A Rho-signaling pathway mediates cortical granule translocation in the sea urchin oocyte Mech Dev 121 225 235 15003626 Cuéllar-Mata P Martínez-Cadena G López-Godínez J Obregón A García-Soto J 2000 The GTP-binding protein RhoA localizes to the cortical granules of Strongylocentrotus purpuratus sea urchin egg and is secreted during fertilization Eur J Cell Biol 79 81 91 10727016 DerMardirossian C Bokoch GM 2005 GDIs: central regulatory molecules in Rho GTPase activation Trends Cell Biol 15 356 363 15921909 Faure J Dagher MC 2001 Interactions between Rho GTPases and Rho GDP dissociation inhibitor (Rho-GDI) Biochimie 83 409 414 11368848 Fiedler SE Bajpai M Carr DW 2008 Identification and characterization of RhoA-interacting proteins in bovine spermatozoa Biology of Reproduction 78 184 192 17928627 Fukumoto Y Kaibuchi K Hori Y Fujioka H Araki S Ueda T Kikuchi A Takai Y 1990 Molecular cloning and characterization of a novel type of regulatory protein (GDI) for the rho proteins, ras p21-like small GTP-binding proteins Oncogene 5 1321 1328 2120668 Golovanov AP Chuang TH DerMardirossian C Barsukov I Hawkins D Badii R Bokoch GM Lian LY Roberts GCK 2001 Structure-activity relationships in flexible protein domains: Regulation of rho GTPases by RhoGDI and D4 GDI J Mol Biol 305 121 135 11114252 Gosser YQ Nomanbhoy TK Aghazadeh B Manor D Combs C Cerione RA Rosen MK 1997 C-terminal binding domain of Rho GDP-dissociation inhibitor directs N-terminal inhibitory peptide to GTPases Nature 387 814 819 9194563 Grizot S Fauré J Fieschi F Vignais PV Dagher MC Pebay-Peyroula E 2001 Crystal structure of the Rac1-RhoGDI complex involved in NADPH oxidase activation Biochemistry 40 10007 10013 11513578 Harlow E Lane D 1988 Antibodies: A Laboratory Manual Cold Spring Harbor Laboratory Press Cold Spring Harbor, NY Hill CS Wynne J Treisman R 1995 The Rho family GTPases RhoA, Rac1, and CDC42Hs regulate transcriptional activation by SRF Cell 81 1159 1170 7600583 Hiraoka K Kaibuchi K Ando S Musha T Takaishi K Mizuno T Asada M Menard L Tomhave E Didsbury J 1992 Both stimulatory and inhibitory GDP/GTP exchange proteins, smg GDS and Rho GDI, are active on multiple small GTP-binding proteins Biochem Biophys Res Commun 182 921 930 1734890 Hoffman GR Nassar N Cerione RA 2000 Structure of the Rho family GTP-binding protein Cdc42 in complex with the multifunctional regulator RhoGDI Cell 100 345 356 10676816 Ihara K Muraguchi S Kato M Shimizu T Shirakawa M Kuroda S Kaibuchi K Hakoshima T 1998 Crystal structure of human RhoA in a dominantly active form complexed with a GTP analogue J Biol Chem 273 9656 9666 9545299 Isomura M Kikuchi A Ohga N Takai Y 1991 Regulation of binding of rhoB p20 to membranes by its specific regulatory protein, GDP dissociation inhibitor Oncogene 6 119 124 1899476 Jou TS Leung SM Fung LM Ruiz WG Nelson WJ Apodaca G 2000 Selective alterations in biosynthetic and endocytic protein traffic in Madin-Darby canine kidney epithelial cells expressing mutants of the small GTPase Rac1 Mol Biol Cell 11 287 304 10637309 Juliano CE Voronina E Stack C Aldrich M Cameron AR Wessel GM 2006 Germ line determinants are not localized early in sea urchin development, but do accumulate in the small micromere lineage Dev Biol 300 406 15 16970939 Keep NH Barnes M Barsukov I Badii R Lian L Segal AW Moody PCE Roberts GCK 1997 A modulator of rho family G proteins, RhoGDI, binds these G proteins via an immunoglobulin-like domain and a flexible N-terminal arm Structure 5 623 633 9195882 Kim O Yang J Qiu Y 2002 Selective activation of small GTPase RhoA by tyrosine kinase Etk through its pleckstrin homology domain J Biol Chem 277 30066 30071 12023958 Knaus UG Heyworth PG Evans T Curnutte JT Bokoch GM 1991 Regulation of phagocyte oxygen radical production by the GTP-binding protein Rac2 Science 254 1512 1515 1660188 Kroschewski R Hall A Mellman I 1999 Cdc42 controls secretory and endocytic transport to the basolateral plasma membrane of MDCK cells Nat Cell Biol 1 8 13 10559857 LaFleur GJ Jr Horiuchi Y Wessel GM 1998 Sea urchin ovoperoxidase: oocyte-specific member of a heme-dependent peroxidase superfamily that functions in the block to polyspermy Mech Dev 70 77 89 9510026 Laidlaw M Wessel GM 1994 Cortical granule biogenesis is active throughout oogenesis in sea urchins Development 120 1325 1333 8026340 Lamarche N Hall A 1994 GAPs for rho-related GTPases Trends Genet 10 436 440 7871593 Lamaze C Chuang TH Terlecky LJ Bokoch GM Schmid SL 1996 Regulation of receptor-mediated endocitosis by Rho and Rac Nature 382 177 179 8700210 Lelias JM 2004 cDNA cloning of a human mRNA preferentially expressed in hematopoietic cells and with homology to a GDP-dissociation inhibitor for the Rho GTP-binding proteins Proc Natl, Acad Sci U S A 90 1479 1483 Leonard D Hart MJ Platko JV Eva A Henzel W Evans T Cerione RA 1992 The identification and characterization of a GDP-dissociation inhibitor (GDI) for the CDC42Hs proteins J Biol Chem 267 22860 22868 1429634 Leung SM Rojas R Maples C Flynn C Ruiz WG Jou TS Apodaca G 1999 Modulation of endocytic traffic in polarized Madin-Darby canine kidney cells by the small GTPase RhoA Mol Biol Cell 10 4369 4384 10588664 Longnecker K Read P Derewenda U Dauter Z Liu X Garrard S Walker L Somlyo AV Nakamoto RK Somlyo AP Derewenda ZS 1999 How RhoGDI binds Rho Acta Crystallogr D Biol Crystallogr 55 1503 1515 10489445 Mabuchi I 1994 Cleavage furrow: timing of emergence of contractile ring actin filaments and establishment of the contractile ring by filament bundling in sea urchin eggs Journal of Cell Science 107 1853 1862 7983152 Mackay DJK Hall A 1998 Rho GTPases J Biol Chem 273 20685 20688 9694808 Manzo S Martínez-Cadena G López-Godínez J Pedraza-Reyes M García-Soto J 2003 A Rho GTPase controls the rate of protein synthesis in the sea urchin egg Biochem Biophys Res Commun 310 685 690 14550256 Michaelson D Silletti J Murphy G D’Eustachio P Rush M Philips MR 2001 Differential localization of Rho GTPases in live cells: Regulation by hypervariable regions and RhoGDI binding J Cell Biol 152 111 126 11149925 Minden A Lin A Claret FX Abo A Karin M 1995 Selective activation of the JNK signaling cascade and c-Jun transcriptional activity by the small GTPases Rac and Cdc42Hs Cell 81 1147 1157 7600582 Nishimura Y Nakano K Mabuchi I 1998 Localization of Rho GTPase in sea urchin eggs FEBS Lett 441 121 126 9877178 Nishimura Y Mabuchi I 2003 An IQGAP-like protein is involved in actin assembly together with Cdc42 in the sea urchin egg Cell Motil Cytoskeleton 56 207 218 14584023 Nobes C Hall A 1995 Rho, Rac and Cdc42 GTPases regulate the assembly of multimolecular focal complexes associated with actin stress fibers, lamellipodia, and filopodia Cell 81 53 62 7536630 Olofsson B 1999 Rho guanine dissociation inhibitors: pivotal molecules in cellular signaling Cell signal 11 545 554 10433515 Qiu RG Chen J Kirn D McCormick F Symons M 1995 An essential role for Rac in Ras transformation Nature 374 457 459 7700355 Rangel-Mata F Méndez-Márquez R Martínez-Cadena G López-Godínez J Nishigaki T Darszon A García-Soto J 2007 Rho, Rho-kinase, and the actin cytoskeleton regulate the Na+-H+ exchanger in sea urchin eggs Biochem Biophys Res Commun 352 264 269 17113032 Ransick A Ernst S Britten RJ Davidson EH 1993 Whole mount in situ hybridization shows Endo 16 to be a marker for the vegetal plate territory in sea urchin embryos Mech Dev 42 117 124 8217840 Regazzi R Vallar L Ullrich S Ravazzola M Kikuchi A Takai Y Wollheim CB 1992 Characterization of small-molecular-mass guanine-nucleotide-binding regulatory proteins in insulin-secreting cells and PC12 cells Eur J Biochem 208 729 737 1327767 Ridley AJ Hall A 1992 The small GTP-binding protein rho regulates the assembly of focal adhesions and actin stress fibers in response to growth factors Cell 70 389 399 1643657 Ridley AJ Paterson HF Johnston CL Diekmann D Hall A 1992 The small GTP-binding protein rac regulates growth factor-induced membrane ruffling Cell 70 401 410 1643658 Ridley AJ 2001 Rho family proteins: coordinating cell responses Trends Cell Biol 11 471 477 11719051 Rivero F Dislich H Glöckner G Noegel AA 2001 The Dictyostelium family of Rho-related proteins Nucleic Acids Res 29 1068 1079 11222756 Sathananthan AH Ng SC Chia CM Law HY Edirisinghe WR Ratnam SS 1985 The origin and distribution of cortical granules in human oocytes with reference to Golgi, nucleolar and microfilament activity Annal N Y Acad Sci 442 251 264 Schatten H Walter M Biessman H Schatten G 1988 Microtubules are required for centrosome expansion and positioning while microfilaments are required for centrosome separation in sea urchin eggs during fertilization and mitosis Cell Motility and Cytoskeleton 11 248 259 Scheffzek K Stephan I Jensen ON Illenberger D Gierschik P 2000 The Rac-RhoGDI complex and the structural basis for the regulation of Rho proteins by RhoGDI Nat Struct Biol 7 122 126 10655614 Scherle P Behrens T Staudt LM 1993 Ly-GDI, a GDP-dissociation inhibitor of the RhoA GTP-binding protein, is expressed preferentially in lymphocytes Proc Natl Acad Sci U S A 90 7568 7572 8356058 Sun Q Schatten H 2006 Regulation of dynamic events by microfilaments during oocyte maturation and fertilization Reproduction 131 193 205 16452714 Symons M Rusk N 2003 Control of vesicular trafficking by Rho GTPases Curr Biol 13 R409 R418 12747855 Takahashi K Sasaki T Mammoto A Takaishi K Kameyama T Tsukita S Takai Y 1997 Direct interaction of the Rho GDP dissociation inhibitor with ezrin/radixin/moesin initiates the activation of the Rho small G protein J Biol Chem 272 23371 23375 9287351 Ueda T Kikuchi A Ohga N Yamamoto J Takai Y 1990 Purification and characterization from bovine brain cytosol of a novel regulatory protein inhibiting the dissociation of GDP from and the subsequent binding of GTP to rhoB p20, a ras p1-like GTP-binding protein J Biol Chem 265 9373 9380 2111820 Wei Y Zhang Y Derewenda U Liu X Minor W Nakamoto RK Somlyo AV Somlyo AP Derewenda ZS 1997 Crystal structure of RhoA-GDP and its functional implications Nat Struct Biol 4 699 703 9302995 Wessel GM Conner S Laidlaw M Harrison J LaFleur GJ Jr 2000 SFE1, a constituent of the fertilization envelope in sea urchins is made by oocytes and contains low-density lipoprotein-receptor-like repeats Biology of Reproduction 63 1706 1712 11090439 Wessel GM Brooks JM Green E Haley S Voronina E Wong J Zaydfudim V Conner S 2001 The biology of Cortical Granules Intern Rev Cytol 209 117 206 11580200 Wessel GM Conner S Berg L 2002 Cortical granule translocation is microfilamet and linked to meiotic maturation in the sea urchin oocyte Development 129 4315 4325 12183383 Wessel GM Wong JL 2009 Cell surface changes in the egg at fertilization Mol Reprod Dev 76 942 53 19658159 Yamashita T Tohyama M 2003 The p75 receptor acts as a displacement factor that releases Rho from Rho-GDI Nat Neurosci 6 461 467 12692556 Zalcman G Closson V Camonis J Honoré N Rousseau-Merck MF Tavitian A Olofsson B 1996 RhoGDI-3 is a new GDP dissociation inhibitor (GDI). Identification of a non-cytosolic GDI protein interacting with the small GTP-binding proteins RhoB and RhoG J Biol Chem 271 30366 30374 8939998 Zalcman G Dorseuil O Garcia-Ranea JA Gacon G Camonis J 1999 RhoGAPs and RhoGDIs, (His)stories of two families Prog Mol Subcell Biol 22 85 113 10081066 Zong H Raman N Mickelson-Young LA Atkinson SJ Quilliam LA 1999 Loop 6 of RhoA confers specificity for effector binding, stress fiber formation, and cellular transformation J Biol Chem 274 4551 4560 9988689
PMC005xxxxxx/PMC5123670.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 2985117R 4816 J Immunol J. Immunol. Journal of immunology (Baltimore, Md. : 1950) 0022-1767 1550-6606 27798167 5123670 10.4049/jimmunol.1601371 NIHMS820639 Article BATF modulates the Th2 locus control region and regulates CD4+ T cell fate during anti-helminth immunity Bao Katherine 1 Carr Tiffany 1 Wu Jianxuan 1 Barclay William 1 Jin Jingxiao 1 Ciofani Maria 1 Reinhardt R. Lee 2 1 Department of Immunology, Duke University Medical Center, Durham, NC 27710 2 Department of Biomedical Research, National Jewish Health, Denver, CO 80206 Correspondence to R.L.R. ([email protected]) 4 10 2016 26 10 2016 1 12 2016 01 12 2017 197 11 43714381 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. The AP-1 factor, basic leucine zipper transcription factor, ATF-like (BATF) is important for CD4+ T-helper (Th) 17, Th9 and follicular T helper (Tfh) cell development. However, its precise role in Th2 differentiation and function remains unclear, and the necessity of BATF in non-allergic settings of type-2 immunity has not been explored. Here, we show that in response to parasitic helminths, Batf−/− mice are unable to generate both Tfh and Th2 cells. As a consequence, Batf−/− mice fail to establish productive type-2 immunity during primary and secondary infection. Batf−/− CD4+ T cells do not achieve type-2 cytokine competency, which implies that BATF plays a key role in the regulation of interleukin(IL)-4 and IL-13. In contrast to Th17 and Th9 cell subsets where BATF binds directly to promoter and enhancer regions to regulate cytokine expression, our results show that BATF is significantly enriched at Rad50 hypersensitivity sites (RHS) 6 and 7 of the locus control region (LCR) relative to AP-1 sites surrounding type-2 cytokine loci in Th2 cells. Indeed, Batf−/− CD4+ T cells do not obtain permissive epigenetic modifications within the Th2 locus, which have been linked to RHS6 and RHS7 function. In sum, these findings reveal BATF as a central modulator of peripheral and humoral hallmarks of type-2 immunity, and begin to elucidate a novel mechanism by which BATF regulates type-2 cytokine production through its modification of the Th2 LCR. BATF helminth allergic immunity Tfh Th2 cells IL-4 AP-1 Th2 locus control region Introduction Over three billion people worldwide are afflicted with type-2 or allergic inflammation (1). In the context of parasitic helminth infections, type-2 inflammatory responses serve a protective role by reducing worm burden in exposed individuals. In contrast, allergic asthma and allergies often result from the improper activation of type-2 inflammation. Recent studies posit that type-2 immune responses are divided into two distinct arms (2–4). One arm is modulated by canonical T-helper (Th) 2 cells, which promote innate cell recruitment, mucus production, smooth muscle contractility, and tissue remodeling at mucosal barriers. The second arm is regulated by IL-4-producing follicular T-helper cells (Tfh) to promote humoral aspects of type-2 responses such as IgE and high-affinity IgG1. Type-2 cytokine production by both Th2 and Tfh cells underlies the unique hallmarks associated with type-2 immunity. However, type-2 cytokine regulation in Th2 cells is distinct from mechanisms used by Tfh cells. Th2 cells utilize canonical factors STAT6 and GATA3 for optimal IL-4 production (2, 5–10), while Tfh cells produce IL-4 independent of these classical Th2 lineage-determining factors (2, 5, 11, 12). One candidate factor that may regulate IL-4 expression in both Th2 and Tfh cells is the AP-1 factor basic leucine zipper transcription factor, ATF-like (BATF). BATF is essential for Tfh cell generation (13, 14), and modulates Tfh-mediated IL-4 expression by binding to the 3’ CNS2 enhancer of the Il4 locus (15–17). The role of BATF in Th2 cell generation and IL-4 production is less clear. Initial in vitro studies demonstrated that BATF was required for Th17 cell differentiation but not Th1 or Th2 cell development (13, 18, 19). Other studies also concluded that BATF had a minor role in Th2 differentiation in vitro (15), however, data from a different Batf−/− mouse line suggested a greater role for BATF in Th2 generation (14, 20). Although the reasons for such differences remain unclear, the genetic background of Batf-deficient strains has been proposed to contribute to the varied results (21). An in vivo role for BATF in allergic immunity is clear. Models of ovalbumin (OVA)-induced allergic airway inflammation show reduced type-2 cytokine expression and allergic hallmarks in Batf−/− lungs (15, 20, 22) regardless of the genetic background (BALB/c and C57BL/6) or different targeting strategies used for the generation of Batf−/− mice (14, 18). However, despite a clear decrease in allergic inflammation, a consensus as to the mechanisms involved has not been reached. Currently, increased Th1 cell accumulation and defects in Th9 development, Th17 cell generation, and Tfh cell function have all been used to explain the diminished type-2 inflammation found in Batf−/− mice (15, 20, 22). Although defects in Th2 cell function and generation have been implied as a result of data obtained from bulk cultures or total tissue extracts from Batf−/− lungs (20, 22), a recent study concluded that Th2 differentiation and IL-4 production remain largely intact after ovalbumin sensitization and challenge of Batf−/− mice (15). As such, further studies are needed to understand the full impact of BATF-mediated regulation in type-2 immunity and Th2 cell biology. To date, studies have focused on the role BATF plays in allergic airway models of type-2 inflammation. To extend our understanding of BATF in non-allergic settings of type-2 immunity, we utlilized the helminth Nippostrongylus brasiliensis, a mouse model of human hookworm infection. This model drives a robust type-2 immune response in both the lung and intestine of helminth colonized animals. Using cytokine reporter mice to assess Th2 and Tfh differentiation at the single cell level, we demonstrate that BATF is essential for both Tfh- and Th2-driven hallmarks of type-2 inflammation during helminth infection. BATF deficiency prevented Tfh cell formation, type-2 cytokine production from Th2 cells, and the recruitment of innate type-2 immune cells to mucosal sites of infection, all of which contribute to the defects observed in helminth clearance during primary and secondary infection. This study also reveals that BATF binds to the locus control region (LCR) within the Th2 cytokine locus and modulates early aspects of LCR activity during Th2 but not Th1 differentiation. Given that optimal type-2 cytokine expression is dependent on the Th2 LCR, this work has identified a novel mechanism of BATF-mediated regulation of type-2 cytokine expression in Th2 cells. Importantly, this mechanism is distinct from that described for BATF-mediated IL-4 production in Tfh cells. In sum, these findings demonstrate that BATF is a central regulator of both Tfh- and Th2-driven arms of type-2 immunity in response to helminth exposure. Materials and Methods Mice C57BL/6 Batf−/− mice used in this study were generated by Kenneth Murphy at Washington University (St. Louis, MO) and purchased from Jackson Laboratories (18). Batf−/− mice were breed at least 10 generations onto the C57BL/6 background and 8–10 generations onto the BALB/c background. IL-4 reporter mice: Il4KN2 and Il44get and IFN-gamma reporter mice: IfngGreat, were graciously provided by Richard Locksley (UCSF). All mice were maintained in pathogen-free animal facilities in accordance to guidelines established by the Division of Laboratory Animal Resources, Institutional Animal Care and Use Committee, and Duke University Medical Center. Infection and worm clearance N. brasiliensis was prepared as described (23). Mice were injected in the rear flank with 500 L3 larvae in saline solution. T cell isolation and culture Lymph nodes from naïve mice were obtained, and CD4+ T cells were negatively selected (Stemcell Technologies; 19860). CD4+ T cells of >90% purity, were cultured at a density of 3x106 cells/ml on plate-bound anti-CD3 and anti-CD-28 (5ug/ml; 145-2C11; 2ug/ml; 37.51 respectively). For Th1 cultures, cells were polarized with 10ug/ml anti-IL-4 (XMG1.2), 10ug/ml IL-12 (Biolegend: 577002), and 10ug/ml IL-2 (Biolegend: 575402); For Th2 cultures, cells were polarized with 10ug/ml anti-IFN-gamma (XMG1.2), 10ug/ml IL-4 (Biolegend: 574302), and 10ug/ml IL-2. All cultures used RPMI (Gibco) supplemented with 2% fetal calf serum, 55uM 2-mercaptoethanol, 100U Penicillin, and 100ug/mL Streptomycin. Cells were cultured for 3 (Th1) and 2 or 4 (Th2) days unless otherwise stated. Flow cytometry Mice were perfused with 15 mL of PBS. Mediastinal lymph nodes and lungs (left lobe only) were collected for analysis. Single-cell suspensions were prepared: lung was digested with: 250ug/ml Collagenase (Sigma: C7657), 50ug/ml Liberase (Roche: 145495), 1mg/ml Hyaluronidase (Sigma: h3506), 200ug/mL DnaseI (Sigma: DN25) in RPMI. Surface staining was performed with the following antibodies: Alexa Fluor 647 conjugated to anti-mouse/human GL7 (GL7); APC/Cy7 conjugated to anti-mouse CD90.2 (30-H12); PECy7 conjugated to anti-mouse PD-1 (RMP1-30); PE conjugated to Streptavidin, anti-mouse CD49b (DX5), anti-human/mouse/rat ICOS (C398.4A), to IL-4Rα (I015F8) and PerCP/Cy5.5 conjugated to anti-mouse CD8 (53-6.7), B220 (RA3-6B2) were from Biolegend; PE conjugated to anti-mouse SiglecF (E50-2440), anti-mouse CD131 (J0R050), anti-mouse CD95 (Jo2) were from BD Pharmingen. Brilliant violet 605 conjugated to anti-mouse CD4 (RM4-5); APC-eFluor 780 conjugated to anti-human/mouse CD45R/B220 (RA3-6B2) and biotinylated anti-mouse CD185/CXCR5 (SPRCL5) were from eBioscience; APC or PE conjugated to anti-human CD2 (S5.5) were from Invitrogen. Live lymphocytes were gated by DAPI exclusion, size and granularity based on forward and side scatter, and singlet gates. A lineage negative gate was used to identify ILC2 cells (Linneg: Ly6G−, B220−, CD8−, CD4−, SiglecF−, CD131−). Data was collected using a FACSCanto (BD Biosciences) and analyzed with FlowJo software (TreeStar). Intracellular staining Cells were stained for surface markers followed by Live/Dead exclusion (Invitrogen). For transcription factor staining, cells were fixed and permeablized using the FOXP3 staining kit (eBioscience) as recommended by the manufacturer. For cytokine staining, cells were incubated with PMA and ionomycin for 5 hours with monensin added during the final 2 hours of culture. Cells were harvested and fixed in 2% formaldehyde, and then permeabilized (0.5% Saponin, 2% FCS, 0.1% Sodium Azide in PBS) as previously described (24). Antibodies used to detect cytokines were FITC-conjugated anti-mouse-IFNγ (XMG1.2;) or PE-conjugated anti-mouse-IL-13 (eBio13A). Enzyme-linked immunosorbent assay 96-well plates were coated with Rat-anti-mouse-IgE (R35-72). Total IgE was detected by biotinylated anti-mouse-IgE (R35-118), followed by streptavidin-horseradish peroxidase and o-phenylenediamine. Concentrations of total IgE were determined by Emax-precision microplate reader (Molecular Devices) and compared to standard curves generated with purified IgE (MEB-38). Immunohistochemistry Mice were exposed to N. brasiliensis and sacrificed 10 days later. The draining mediastinal lymph nodes were isolated, fixed in 4% paraformaldehyde, placed in 30% sucrose, embedded and frozen in Optimum Cutting Temperature (O.C.T.) embedding compound (Sakura Finetek USA). Sections were cut at 5 microns and quenched of endogenous peroxidase activity with 1% H2O2 and 0.5% Sodium Azide in PBS. Slides were subsequently blocked for non-specific Fc Receptor binding (TruStain Fc block-clone 93, Biolegend) and endogenous biotin (Avidn/biotin block; Vector Labs). eGFP was detected by anti-mouse-GFP (Novus Biologicals: NB600-308), followed by biotinylated F(ab’)2 donkey anti-rabbit (1:500; Jackson Immuno Research Laboratories). Tyramide amplification was applied to maximize detection of biotinylated-anti-eGFP. Briefly, following addition of biotinylated F(ab’)2 antibodies, tissue sections were incubated with streptavidin-HRP followed by fluorescein isothiocyanate–tyramide from the TSA Fluorescein System according to the manufacturer's instructions (PerkinElmer: NEL 701001KT). Prior to CD4 detection, a second round of peroxidase quenching and blocking of biotin (Avidn/Biotin Block; Vector labs) was performed. CD4 was detected using biotinylated-anti-CD4 antibody (RM4-5; Biolegend), and signal was amplified with tyramide–Alexa Fluor 555 (Invitrogen). After a third round of blocking and quenching of biotin and peroxidase, B220 was detected by biotinylated-anti-B220 (RA3-6B2; Biolegend), followed by Dylight649 conjugated to anti-Streptavidin (Biolegend). Nuclei were counterstained with DAPI (4′, 6-diamidine-2′-phenylindole dihydrochloride; Roche) in PBS before mounting (Vectashield: Vector Laboratories) on coverslips. Digital images were collected with Zeiss Axio Imager. Images were converted to red-green-blue, colored and overlaid with Photoshop CS5 software (Adobe Systems). Concentrations of antibodies used are available on request. Quantification of IL-4 competent cells in the B cell follicles Individual images in the Fitc (anti-GFP), Tyramide-555 (anti-CD4), and Cy5 (B220) channels were captured using a 5x objective and processed in Fiji (ImageJ). First, B220 and CD4+ staining were overlaid and used to demarcate the B cell follicles. A region of interest (ROI) corresponding to the B cell follicle was outlined and saved. Total pixels were measured within this region of interest to obtain the total area of the follicular region. Next, thresholding was used to select GFP+ events that were above background within the GFP image. The saved ROI depicting the follicular region was applied to the adjusted GFP image, and total pixels within the threshold set for the GFP+ events was obtained. Total GFP+ pixels divided by the total pixels in the follicles gave the relative area of the follicles that was occupied by GFP+ (IL-4+) cells. H&E and PAS staining Lungs and small intestines were fixed in 4% formaldehyde overnight. Tissues were paraffin embedded, sectioned, and stained by the Duke University histology core. Digital images were acquired with Zeiss Axio Imager. DNA Methylation analysis and bisulfite sequencing Cells were lysed with proteinase K in lysis buffer (50mM Tris pH8, 100mM EDTA, 0.5% SDS), and genomic DNA was prepared by phenol-chloroform extraction and ethanol precipitation. Bisulfite treatment and purification was performed with the EpiTect (Qiagen) kit according to the manufacturer’s protocol. Primers used to amplify bisulfite-treated genomic DNA of RHS7 were previously described (25). Amplicons were cloned into the TOPO-TA pCR2.1 vector (Invitrogen) and individual clones were sequenced. Chromatin Immunoprecipitation (ChIP) 5x106 to 15x106 live cells were cross-linked with 1% paraformaldehyde for 10 minutes. Nuclei were generated with Farnham Lysis buffer. Nuclear lysates were isolated with RIPA buffers and fragmented with a Vibra-Cell VCX130PB (Sonics & Materials). For immunoprecipitation, anti-mouse IgG1 Dynabeads or Protein G were incubated overnight at 4°C with 5ug anti-BATF (WW8), anti-JunB (sc-46x), or anti-IRF4 (sc-6059), washed and incubated overnight with sonicated chromatin at 4°C. After immunoprecipitation, protein-DNA crosslinking was reversed by heating at 65°C, followed by proteinase K and RNase A treatment. DNA was purified using phenol-chloroform and ethanol precipitation and amplified by qPCR (LightCycler 480II, Roche) with indicated primers (see Supplementary Table 1) for conserved AP1 binding sites at the Th2 locus and a conserved non-AP-1 site within the locus (negative control). Conserved binding sites were identified with the Evolutionary Conserved Region (ECR) browser (http://ecrbrowser.dcode.org/), and promoter regions were defined to include approximately 500 basepairs upstream of the transcription start sites of Il4 and Il13. Native ChIP Chromatin was prepared from 5x106 cultured Th1 or Th2 cells and immunoprecipitated using antibodies specific for acetylated histone H3 (H3ac; EMD Millipore, 06-599), trimethylated histone H3 lysine 4 (H3K4me3; EMD Millipore; 07-473) following a protocol described previously (26) with modifications: MNase was purchased from NEB, and sonication step was omitted. Immunoprecipitated and input samples were quantified by SYBR green (Bio-Rad) qPCR (LightCycler 480II, Roche) with previously described primers against the Il4 promoter (27). Values were normalized to that obtained for β2-microglobulin (26) and enrichment was expressed as percent input. Results BATF is required for type-2 inflammation and immunity to Nippostrongylus brasiliensis Wild-type and Batf−/− mice were analyzed for hallmarks of type-2 inflammation 10 days after exposure to the helminth Nippostrongylus brasiliensis. Consistent with previous findings that showed BATF is required for antibody isotype class-switch recombination, we found that serum IgE was undetectable in Batf−/− mice (Fig. 1A) (13, 14). Furthermore, worms were cleared from the intestine in wild-type animals, but remained abundant in Batf−/− mice (Fig. 1B). Batf−/− mice also failed to induce mucus production in the lung and intestine during helminth infection (Fig. 1C). Together, these results indicated that Batf−/− mice exhibit significant impairments in both humoral and peripheral arms of type-2 inflammation. Lasting immunity to N. brasiliensis is dependent on CD4+ T cells and type-2 cytokine production (28–30). To determine the impact of BATF deficiency on anti-helminth immunity, we evaluated the generation and function of Tfh and Th2 memory cells with IL-4 reporter mice. Il44get mice report Il4 transcription by GFP expression and Il4KN2 mice report recent IL-4 protein production by human CD2 (huCD2) expression (5, 31). Dual IL-4 reporter mice (Il44get/KN2) were re-exposed to N. brasiliensis 21 days following primary infection (Fig. 2A). Wild-type mice elicited a rapid memory response indicated by increased CXCR5+, PD-1+ Tfh cell numbers (Fig. 2B, C) in the mediastinal lymph node and IL-4+ (CD4+, GFP+) Th2 cell numbers (Fig. 2D, E) in the lung five days after secondary exposure (2°). These responses were significantly above what is observed 5 days into the primary infection (1°). In contrast, Batf−/− mice showed no significant enhancement of Tfh cells in the draining lymph node or Th2 cell numbers in the lung upon primary or secondary exposure (Fig. 2C, E). Although B cells and antibodies do not appear to play a significant role in protection during the primary response, they are important for reducing worm burden upon re-infection (32, 33). In agreement, we found that previously exposed wild-type mice exhibited enhanced IgE production and expedited worm clearance during the secondary exposure to N. brasiliensis (Fig. 2F, G). In contrast, Batf−/− mice did not exhibit an increase in either humoral or cell-mediated immunity during the secondary response to helminths as serum IgE levels remained low and intestinal worm burden remained similar to that achieved during the primary infection (Fig.1B; 2F, G). These data demonstrate that BATF is required for the establishment of protective immunity against intestinal helminth infection. BATF is required for Tfh cell generation, follicular migration, and IL-4 production after infection with N. brasiliensis. To assess the humoral response in more detail, Tfh cell generation and germinal center formation were evaluated in mice exposed to N. brasiliensis. As expected, mediastinal lymph nodes from wild-type mice revealed robust Tfh cell (CXCR5+, PD1+) and germinal center B cell (CD95+, GL7+) generation eight days after helminth exposure (Fig. 3A, B). In contrast, Batf−/− mice were devoid of both Tfh cells and germinal center B cells. Consistent with previous studies (3, 34), GFP (IL-4) expression was enriched among CD4+CXCR5+ Tfh cells, while BATF deficiency resulted in the absence of CD4+, CXCR5+ T cells expressing GFP (Fig. 3C). As a whole, GFP+ CD4+ T cells in the lymph nodes were significantly reduced in Batf−/− mice. Immunohistochemical analysis of the mediastinal lymph nodes 10 days after helminth exposure confirmed the absence of CD4+ T cells in the follicles of Batf−/− mice (Fig. 3D). This is highlighted by the selective reduction of CD4+ staining (red) in the follicular regions of Batf−/− mice compared to wild-type animals (Fig. 3D, 3D inset). Importantly, GFP+, CD4+ T cells also were absent from the follicles of infected Batf−/− mice as shown by the significant reduction of GFP+ cells in B220+ regions of the lymph node (Fig. 3D, E). Failure of Tfh cells to migrate into the follicles would explain the defects we observed in IgE production and may help to explain the global reduction in class-switching observed with other isotypes in prior studies (13). In summary, our data demonstrate that impaired IL-4 production, Tfh cell generation, follicular migration and germinal center development collectively contribute to the humoral defects observed in Batf−/− mice after helminth exposure. BATF is required for type-2 inflammation The above studies clearly demonstrate that BATF is essential for the induction of type-2 humoral immunity during helminth exposure. To investigate whether BATF is also important for peripheral hallmarks of type-2 inflammation, mice were exposed to N. brasiliensis and inflammation in the lung was analyzed eight days later. Total numbers of eosinophils, basophils, group 2 innate lymphoid (ILC2) cells and CD4+ Th2 cells were reduced significantly in the lung of helminth exposed Batf−/− animals compared to wild-type mice (Fig. 4A). To determine if there was also impairment in type-2 cytokine production, IL-4 expression was assessed. As expected, eosinophils, basophils and a subset of CD4+ T cells in wild-type lungs were IL-4 competent (GFP+) (Fig. 4B). Additionally, a large fraction of wild-type CD4+ T cells expressed both IL-4 mRNA and protein (GFP+ and huCD2+ respectively) (Fig. 4C). Interestingly, although reduced in number, eosinophils and basophils from the lung of Batf−/− mice expressed GFP to a similar extent as wild-type cells (Fig. 4B). However, Batf−/− CD4+ T cells were significantly impaired in both IL-4 transcript and protein production (Fig. 4B, C). A similar defect in Th2 cell generation was observed in Batf−/− mice bred onto a BALB/c background (Supplemental Fig. 1A). Similar to C57BL/6 mice, the defect in Th2 function corresponded to an impairment in worm clearance (Supplemental Fig. 1B). A defect in Th2 cell commitment was further confirmed, as Batf−/− CD4+ T cells did not express high levels of GATA3 in the lungs or lymph nodes, a characteristic of canonical Th2 cells (Fig. 4D; Supplemental Fig. 2). In sum, BATF is required to establish IL-4 potential in Th2 cells, but this role appears specific to CD4+ T cells as type-2 innate immune cells maintain IL-4 competency in Batf−/− mice. BATF is required for Th2 cell differentiation IL-4 positively regulates its own expression in Th2 cells, and thereby stabilizes the Th2 cell fate. Since BATF deficiency resulted in impaired IL-4 expression by CD4+ T cells, low levels of IL-4 in the lymph nodes and lungs of Batf−/− mice may fail to reinforce cytokine expression and Th2 commitment. To determine if providing exogenous IL-4 could rescue IL-4 production in Batf−/− CD4+ T cells, CD4+ T cells were cultured under Th2-inducing conditions. In wild-type cultures, 35–38% of cells acquired IL-4 competency of which a subset secreted IL-4 protein (Fig. 5A). In comparison, Batf−/− CD4+ T cells exhibited negligible IL-4 competency and protein production (Fig. 5A). A similar defect was observed in BALB/c Batf−/− CD4+ T cells cultured under Th2 conditions (Supplemental Fig. 3). Therefore, even under conditions in which IL-4 is in excess, BATF was required for optimal IL-4 production. Similarly, IL-13 expression was also impaired among Batf−/− CD4+ T cells cultured under Th2 conditions (Fig. 5B). Because the threshold of GATA3 expression is highly associated with both immune cell-mediated IL-13 production (35, 36) and cell fate decisions (37), we assessed whether BATF deficiency impacted GATA3 expression in cells cultured under Th2 conditions. Similar to our results in vivo, Batf−/− CD4+ T cells demonstrated decreased GATA3 expression in vitro (Fig. 5C). We next sought to determine whether BATF was also required for Th1 differentiation and function. To do this we used an IFN-gamma cytokine reporter strain (IfnγGreat) to assess cytokine production. IfnγGreat reporter mice utilize a bicistronic reporter to generate both IFN-gamma and yellow fluorescent protein (YFP) expression from the same transcript (3, 38). In these mice, cells that produce IFN-gamma concomitantly express YFP. Both wild-type and Batf-deficient, ifnγGreat CD4+ T cells cultured under Th1 conditions readily expressed YFP and IFN-gamma protein (Fig. 6A). These results were consistent with previous findings showing that Th1 differentiation is uncompromised in Batf−/− CD4+ T cells cultured in vitro (13, 18, 19). To examine whether this trend held true in vivo, we exposed wild-type and Batf−/− IfngGreat mice to N. brasiliensis and quantified YFP expression in the lung 8 days later. Again, Batf-deficient mice showed no defect in IFN-gamma transcript expression among CD4+ T cells (Fig. 6B). In fact, a higher percentage of CD4+ T cells expressed IFN-gamma in Batf−/− mice relative to wild-type, a finding consistent with ovalbumin models of allergic lung inflammation (22). However, in the helminth model we did not observe a significant increase in the number of IFN-gamma-producing CD4+ T cells in the lung relative to wild-type animals. Together, these results demonstrate that BATF has a fundamental role in Th2 cell differentiation, but is not necessary for Th1 cell differentiation in vitro or in vivo. BATF deficiency limits permissive epigenetic modifications at the Th2 LCR and Th2 cytokine loci BATF binds directly to the Il17 locus to facilitate transcription factor recruitment and Il17 transcription early during Th17 differentiation (39). Similarly, BATF has been shown to bind directly to the Il9 promoter during Th9 differentiation (20). To test whether BATF regulated Il4 and Il13 in a similar fashion, we performed BATF Chromatin Immunoprecipitation (ChIP) in wild-type Th2 cells at putative conserved AP-1 binding sites across Il4 and Il13 loci and their promoters (Fig. 7A, B). We detected no enrichment of BATF binding among these sites (relative to the negative control region in wild-type Th2 cells) (Fig. 7C). Instead, BATF binding within the Th2 locus was enriched significantly upstream of Il4 and Il13 within the Th2 locus control region (LCR) (Fig. 7A, C). Within the LCR, there are several Rad50 hypersensitivity sites (RHS) that individually contribute to LCR hub activity and type-2 cytokine regulation (40, 41). We found that BATF binds specifically at RHS6 and RHS7 (Fig. 7C), two regions of the LCR critical for type-2 cytokine production (27, 40, 42). As BATF forms a heterodimer with JUN family members and these complexes cooperate with IRF4 to promote Th17 differentiation (43, 44), we assessed whether JunB and IRF4 binding was also enriched at RHS6 and RHS7. Indeed, both JunB and IRF4 bound at the same AP-1 sites within RHS6 and RHS7 that were enriched for BATF binding (Fig. 7D and E). Together, these findings indicate that BATF-JunB-IRF4 complexes are likely involved in the regulation of type-2 cytokine expression through the modulation of LCR activity. This mechanism is independent of BATF binding at proximal promoter and enhancer regions as described for other cytokine loci. To investigate a possible connection between BATF and the LCR, we focused on RHS6 and RHS7. RHS6 is required for permissive epigenetic modifications at the Th2 locus (27). As a result, RHS6-deficient mice are substantially impaired in the production of Th2 cytokines both from CD4+ T cells cultured in vitro and after OVA-induced allergic inflammation (27). Consistent with these results, we found decreased levels of permissive modifications, histone 3 pan-acetylation (H3Ac) (Fig. 7F, left) and H3 lysine 4 tri-methylation (H3K4me3) (Fig. 7F, right) at the IL-4 promoter in Batf−/− CD4+ T cells cultured under Th2 conditions. Demethylation of RHS7 occurs early during Th2 differentiation, but not during Th1 differentiation (40, 45). As expected, both wild-type and Batf−/− Th1 cells demonstrated a highly methylated RHS7 region (Fig. 7G, left). However, in contrast to the demethylated state of RHS7 in wild-type Th2 cells, RHS7 remained largely methylated in Batf−/− CD4+ T cells cultured under Th2 conditions (Fig. 7G, right). Given the importance of both RHS6 and RHS7 in LCR-mediated regulation of type-2 cytokine production, these data suggest that BATF works early to establish permissive epigenetic changes at the LCR and Th2 cytokine loci. Such epigenetic changes are critical to achieve type-2 cytokine competency in CD4+ T cells. Discussion Using a mouse model of intestinal helminth infection, we show that BATF is an essential regulator of Tfh cell generation and IgE production in non-allergic settings of type-2 immunity. In addition, BATF is required for canonical Th2 generation and Th2-mediated hallmarks of type-2 inflammation such as mucus production and innate cell recruitment to peripheral sites of inflammation. Thus, BATF serves as a central regulator of both Tfh (humoral) and Th2 (peripheral) cell-mediated hallmarks of type-2 immunity. This is the first evidence that BATF plays an important role in non-allergic settings of type-2 inflammation and highlights its necessity in establishing productive anti-helminth immunity and immunologic memory. Although findings in allergic models also support the importance of BATF in type-2 inflammation, no consensus has been reached regarding the mechanisms involved (15, 20, 22). One possibility may be through the expression of IL-9 by Th9 cells. Indeed, BATF has been described as a critical regulator of allergic inflammation through its collaboration with IRF4 in Th9 differentiation (20, 46). Although IL-9 is induced in response to N. brasiliensis, its functional contribution is overshadowed by IL-4 and IL-13 (2, 47, 48). Indeed, our data generated from Batf−/− mice more closely resemble that of helminth-infected Il4−/−Il13−/− animals than of helminth-infected Il9−/− mice (47). As such, BATF appears to be important for both Th9-dependent and -independent hallmarks of type-2 inflammation (49, 50). A second mechanism proposed to explain the reduced allergic inflammation observed in Batf−/− mice involved a specific defect in Tfh-driven IL-4 production, as Th2-derived IL-4 production was thought to be normal in this model (15). Our data suggests that this mechanism does not play a significant role during helminth-induced type-2 inflamamtion for three reasons. First, we observed a significant defect in Th2 cell generation and IL-4 production in helminth-exposed Batf−/− mice. This finding is consistent with the idea that Th2 cells are the predominant orchestrators of type-2 inflammation in mucosal tissues (51) and thus required for peripheral hallmarks of type-2 inflammation. Second, we found a complete absence of Tfh cells in helminth-exposed Batf−/− mice, consistent with the requirement for BATF in Tfh cell generation (13, 14). Thus, during helminth infection, defects attributed to Tfh cells likely result from their absence rather than their inability to express IL-4. Lastly, IL-13-producing cells are required to establish the hallmarks of type-2 inflammation observed during N. brasiliensis infection, and Tfh cells do not produce IL-13 in this model (2). Together, these findings make a specific-defect in Tfh-derived IL-4 an unlikely mechanism to explain the reduced type-2 inflammation observed after helminth infection in Batf−/− mice. Additional evidence that BATF deficiency likely impacts type-2 airway inflammation by affecting Th2 cells rather than Tfh cytokine production, comes from studies using mice that are deficient in the CNS2 enhancer of the Il4 locus. BATF binds to the CNS2 region and is critical for IL-4 production by Tfh cells (15). However, deletion of the CNS2 region ultimately has little impact on airway responsiveness to OVA (16, 52). Thus, BATF binding to the CNS2 region to promote IL-4 production in Tfh cells is neither sufficient nor necessary to mediate OVA-induced allergic airway inflammation. We believe the data is most consistent with a model where BATF deficiency results in a failure to generate both Tfh and Th2 cells during allergic and non-allergic settings of type-2 inflammation. Why different mechanisms have arisen to explain the diminished type-2 inflammation observed among Batf-deficient animals is interesting. Apart from the obvious reason that the helminth model is different from that of allergen sensitization, one possibility to explain the different results may lie in how T cell fate and function were determined. In this study, we assessed cytokine potential of Tfh and Th2 cells in vivo by histology or directly ex vivo by flow cytometry using sensitive reporter systems. Ex vivo manipulations were not required. In the ovalbumin model (15), cytokine potential was assessed after spleens were harvested from immunized mice and restimulated ex vivo with antigen for an additional three days. Thus, it is possible that extended culturing ex vivo does not reflect certain aspects of biology in vivo. On the other hand, extended culture conditions may reveal a unique pathway for Tfh generation that is absent during helminth exposure or one that occurs only in settings where prolonged and repeated T cell receptor engagement can occur. In fact, this ex vivo system resembles findings in vivo where Tfh-like cells are generated independently of B cells in mice repeatedly dosed with high concentrations of antigen (53). The use of extended cultures may also help to explain the different conclusions made surrounding the role of BATF in Th2 differentiation using similar OVA-induced airway inflammation models (15, 22). Genetic and histological comparisons between Tfh and Th2 cells isolated directly ex vivo and their cultured counterparts will shed further light on this topic. As stated earlier, prior literature reached contradictory conclusions about the importance of BATF in Th2 differentiation in vitro (21). The mixed results obtained by different groups investigating the role of BATF during Th2 differentiation do not appear to depend on the Batf−/− animal model as the Batf−/− mice and cells used in this study, which were originally generated at Washington University (18), generate a similar phenotype to that observed from mice generated using a completely different targeting strategy (14, 20). Similarly, the genetic background of the mice used can not fully explain the differences observed. CD4+ T cells from both BALB/c (19, 22) and C57BL/6 (14, 15, 18, 20) backgrounds have given similarly conflicting results. Indeed, no obvious role for the genetic background was observed in our studies as Batf−/− CD4+ T cells on the BALB/c background showed the same impairment in Th2 differentiation and IL-4 production as those on the C57BL/6 background. Thus, similar to the in vivo studies discussed above, antigen dose and cytokine availability may also explain the differences observed during in vitro Th2 differentiation. Similar to the methods used in this study, Th2 culture periods of 4–5 days have shown that BATF deficiency greatly impairs IL-4 production among T cells cultured under Th2 conditions (14, 20), while other publications report no effect (13, 15). However it should be noted that, when other type-2 cytokines were assessed during this 4–5 day period, a significant defect was still reported (15). Only when Th2 cultures were performed for a period of 7–15 days or when cells were subjected to prolonged restimulation prior to the assessment of cytokine potential did BATF deficiency consistently result in little or no defect in Th2 cell development (15, 18, 19). This suggests that extending the duration of T cell receptor engagement or increasing cytokine availability may be sufficient to bypass the early requirement for BATF in developing Th2 cells. Thus, a Batf-independent pathway for both Tfh and Th2 cell generation may occur in conditions when antigen and/or local type-2 cytokine availability is persistent. Whether such conditions occur naturally in vivo will be an interesting area for future investigation. The precise mechanism of BATF-mediated cytokine regulation in Th2 cells is still unclear. This study uncovers a unique BATF-dependent mechanism required for optimal Th2 differentiation that works through the Th2 LCR. We used ChIP to show that binding of BATF, JunB, and IRF4 is enriched at RHS6 and RHS7 of the Th2 LCR. This finding supports a model where BATF-JunB heterodimers pair with IRF4 to bind AP-1-IRF composite elements (AICE) at key locations within the LCR. Although BATF-IRF4 complexes are known to influence the expression of Il17 by binding directly at its promoter and enhancer regions (39, 43, 44), our results suggest that these complexes work at a distance to control IL-4 and IL-13 in developing Th2 cells. The results shown herein suggest that in Th2 cells, BATF is required for the binding of key lineage-determining factors to the LCR and Th2 locus. Specifically, BATF binds to essential DNase hypersensitivity sites, RHS6 and RHS7, both required for optimal type-2 cytokine expression in Th2 cells (27, 42, 54). To further support a role for BATF-mediated regulation of the LCR to coordinate type-2 cytokine production, we found that both IL-4 and IL-13 were impaired in Th2 polarized Batf−/− CD4+ T cells. Batf−/−cells cultured under Th2 conditions revealed an appreciably lower level of permissive histone modifications at the IL-4 promoter than wild-type cells. This defect was accompanied by impaired demethylation of the RHS7 region. In sum, the data presented demonstrate a requirement for BATF to facilitate chromatin remodeling at the Th2 locus early during Th2 differentiation. Dissecting precisely how BATF coordinates with other transcription factors at the LCR to modify the chromatin landscape in Th2 and Tfh cells will broaden our understanding of T cell lineage commitment and plasticity (55). It is also likely to strengthen our understanding of the relationship between Tfh cell and Th2 cell subsets in settings of type-2 inflammation (56, 57). Interestingly, basophils and eosinophils, which preferentially produce IL-4 and not IL-13 in vivo (2, 36), did not show a defect in IL-4 competency in Batf-deficient animals. This suggests that these innate cells do not require BATF binding at either the CNS2 region or the LCR to achieve IL-4 production. Indeed, the CNS2 region does not appear critical for IL-4 expression by these cells (16). This may indicate that BATF is required specifically in CD4+ T cells to achieve IL-4 competency, and the Il4 locus in these innate subsets becomes accessible independently from BATF. It is interesting to point out that the lack of IL-13 production in basophils and eosinophils despite IL-4 production further highlights the importance of BATF in the coordinate regulation of type-2 cytokines. Further work investigating the role of BATF and LCR activity in these various type-2 immune cells will provide insight into the specificity of this mechanism. Although much is currently understood with regard to the pathogenesis of allergic disorders and helminth infection, these diseases continue to impact human health on a global scale. As such, there is a need to identify molecular factors and mechanisms responsible for the initiation of both Tfh- and Th2-driven type-2 inflammation. Of interest, factors common to the development of both IL-4-producing Tfh cells and IL-4- and IL-13-producing Th2 cells are not well characterized, but represent critical targets for the development of therapeutics designed to ameliorate a spectrum of allergic pathologies. Furthermore, these findings have obvious implications among various infectious and allergic diseases that depend on the generation and maintenance of allergen-specific immunoglobulin and memory Tfh and Th2 cells. Taken together, BATF is an early regulator of type-2 inflammation, and works via several distinct mechanisms to regulate type-2 cytokine production. These results provide an intriguing platform to explore new therapies against type-2 cytokine-driven immunopathology. Supplementary Material 1 Funding: This work was supported by the National Institute of Health (AI119004) The authors thank Richard Locksley (UCSF) for providing various reporter strains; Sam Johnson and the Duke Light Microscopy Core Facility; The Duke Pathology core for processing of histological stains; Nancy Martin and Lynn Martinek for flow cytometry and sorting; and Ann Miller for expert technical and general support. Fig. 1 BATF deficiency impairs type-2 immunity against N. brasiliensis Wild-type and Batf−/− mice were exposed to N. brasiliensis. (A) Serum IgE and (B) worm burden ten days post exposure. (C) Periodic Acid-Schiff staining was performed on the lung and intestine of wild-type and Batf−/− mice eight days after helminth infection. Images are representative of multiple fields of view in each tissue and animal. P = 0.004 (two-tailed t-test), P = 0.0002 (two-tailed t-test). Data are representative of (A,B) 3 or (C) 2 independent experiments. (A,B) n= 4 mice per group, (C) n=2–3 mice per group. Data are represented as mean +/− SD. Fig. 2 BATF deficiency prevents the establishment of immunological memory (A) Model of experimental procedure: mice were exposed to N. brasiliensis either once (1°) or were rechallenged 21 days after initial infection (2°). Mice were then analyzed five or eight days after the last exposure. (B) Representative contour plots of total CD4+ T cells isolated from the lung-draining mediastinal lymph nodes five days post-secondary exposure. Gates indicate Tfh cells (CXCR5+, PD1+). (C) Total number of Tfh cells from the mediastinal lymph node five days after primary (1°) and secondary (2°) exposure. (D) Representative contour plots of total CD4+ T cells isolated from the lung five days post-secondary infection. Gates represent Th2 cells (CD4+, GFP+). (E) Total number of Th2 cells in the lung 5 days after primary (1°) and secondary (2°) infection. (F) Serum IgE post-secondary infection. (G) Worm burden post-secondary infection. P<0.05; P<0.005, P<0.0001 (two-tailed t-test). n= 3–6 per group. Data are represented as mean +/− SD. Fig. 3 BATF is required for Tfh cell generation, follicular migration, and germinal center formation Wild-type- and Batf−/−- Il44get/KN2 mice were exposed to N. brasiliensis and draining mediastinal lymph nodes were analyzed 8 days after infection. (A) Representative contour plots and quantification (bar graph) of CXCR5+PD1+ Tfh cells (gated through CD4+ T cells). (B) Representative contour plots and quantification (bar graph) of GC B cells identified as CD95+GL7+ (gated through B220+ cells). (C) Contour plots gated through CD4+ T cells depicting IL-4 competency (GFP-expression) in non-Tfh (CXCR5−; left gate) and Tfh (CXCR5+; right gate) cells. Graph shows the frequency of CXCR5 expression among CD4+GFP+ T cells. (D) Immunohistochemistry of mediastinal lymph nodes 10 days after helminth exposure. Sections were stained for IL-4 competency (anti-GFP; green) of CD4+ T cells (anti-CD4; red) and B cell follicles (anti-B220; blue). Top panel shows all three stains, and middle and bottom panels show only CD4 and GFP staining with dashed lines outlining the boarders of the B cell follicle. Bottom panels represent an enlargement of the boxed region in the corresponding middle panel. (E) Quantification of the area in the follicles in (D) comprised of GFP staining. Data points represent individual images of separate follicle regions from indicated mice. P<0.0001 (two-tailed t test). Data are cumulative of 3 (A, B, C) or representative of 2 (D, E) independent experiments. (A, B, D) n=3 per group, (C) n=4–5 per group, (E) n=2 per group with 1 wild-type (GFP-negative) mouse as a control. Data are represented as mean +/− SD. Fig. 4 Type-2 inflammation and Th2 cytokine production are impaired in Batf-deficient animals Wild-type- and Batf−/−- Il44get/KN2 mice were exposed to N. brasiliensis, and the lung (left lobe) was analyzed 8 days later. (A) Quantification of lung eosinophils, basophils, ILC2 cells and CD4+ T cells. (B) Histogram overlays show relative IL-4 competency (GFP+) of eosinophils, basophils, ILC2 cells, and CD4+ T cells from wild-type (reporter negative control, filled), wild-type Il44get/KN2 (dashed line) and Batf−/− Il44get/KN2 (solid line) mice. (C) Representative contour plots depicting IL-4 transcript (GFP+; left gate) or IL-4 transcript and protein (GFP+huCD2+; right gate) expression in CD4+ T cells. Bar graphs quantify total IL-4 transcript-expressing (GFP+; left) and total IL-4 transcript and protein-expressing (GFP+huCD2+; right) CD4+ T cells in the lung. (D) Representative contour plots of lung CD4+ T cells stained for GATA-3 and IL-4 protein (huCD2). Gates indicate: GATA3hihuCD2− (top left); GATA3hihuCD2+ (top right); and GATA3lohuCD2+ (bottom right). Bar graph depicts the proportion of CD4+ T cells that are GATA3hi. P=0.0132; P=0.0029; *P<0.0004 (two-tailed t-test). Data are representative of 4 independent experiments. (A,B,C,D) n=3–4 per group. Data are represented as mean +/− SD. Fig. 5 BATF is required for Th2 differentiation and function Naïve CD4+ T cells were isolated from indicated IL-4 reporter mice and cultured under Th2 conditions. (A) Representative contour plots of CD4+ T cells. Gates depict the proportion of IL-4 transcript-expressing (GFP; top left) and IL-4 mRNA- and protein-expressing (GFP, huCD2; top right) CD4+ T cells. Histogram overlays and bar graphs depict the relative levels and frequency of IL-4 transcript (GFP; left) and protein (huCD2; right) production in wild-type (reporter negative, filled), wild-type Il44get/KN2 (dashed) and Batf−/−Il44get/KN2 (solid) CD4+ T cells. (B) Th2 cultured CD4+ T cells, were stimulated with PMA/Ionomycin for 5 hours on the day of harvest. Representative contour plots of CD4+ T cells. Gates depict the proportion of IL-13 protein. Bar graphs show the relative frequencies of IL-13 protein production within CD4+ T cells. (C) Histogram overlay and bar graph depict the MFI of GATA-3 among indicated CD4+ T cell populations after Th2 culture. P<0.001 (two-tailed t-test). Data are representative of 3 (A,C) or 2 (B) independent experiments. (A,B,C) n= 2–4 replicate wells per group. Data are represented as mean +/− SD. Fig. 6 BATF is not required for Th1 differentiation or function (A) Th1 cultured CD4+ T cells, were stimulated with PMA/Ionomycin for 5 hours on the day of harvest. Representative contour plots of total CD4+ T cells from wild-type- and Batf−/−- IfngGreat mice with gates depicting the proportion of IFNg transcript (YFP). Histogram overlays and bar graphs depict the relative levels and frequency of IFNg transcript (YFP) and protein production. (B) Mice were exposed to N. brasiliensis and the lung (left lobe) was assessed 8 days later. Representative contour plots of isolated CD4+ T cells. Gate represents the percentage of IFNg transcript (YFP+) producing CD4+ T cells. Histogram overlay and bar graph depict the levels of IFNg transcript (YFP+) among indicated CD4+ T cell populations and the total number of IFNg transcript (YFP) producing cells, respectively. Ns: P>0.235 (two-tailed t-test). Data are representative of 3–4 (A) and cumulative of 2 (B) independent experiments. (A) n=2–3 replicate wells per group; (B) n=4–5 per group. Data are represented as mean +/− SD. Fig. 7 BATF binds to and modulates the activity of the Th2 LCR (A) Diagram of the Th2 locus with conserved AP-1 bindings sites underscored throughout. BATF binding is identified with arrows at RHS6 and RHS7. (B) List of conserved AP-1 motifs as labeled in (A), capitalized letters indicate consensus sequences. (C, D, E) Wild-type (filled bar) and Batf−/− (open bar) cells were cultured under Th2 conditions and prepared for ChIP analysis. BATF (C), JunB (D), and IRF4 (E) binding was evaluated at designated AP-1 sites along the Th2 locus relative to a non-AP1 site. (F) Bar graph represents ChIP analyses of permissive Histone 3 modifications at the IL-4 promoter. Bar graphs depict relative levels of H3Ac (left) and H3K4me3 (right) in wild-type (filled bar) and Batf−/− (open bar) cells. (G) Percent demethylation of wild-type (filled bar) and Batf−/− (open bar) Th1 (left) and Th2 (right) cells at indicated sites within the RHS7 amplicon. Data are representative of 2 (C, D, E) or 3 (F, G) independent experiments. Data are represented as mean +/− SD. 1 Pulendran B Artis D 2012 New paradigms in type 2 immunity Science 337 431 435 22837519 2 Liang HE Reinhardt RL Bando JK Sullivan BM Ho IC Locksley RM 2012 Divergent expression patterns of IL-4 and IL-13 define unique functions in allergic immunity Nature immunology 13 58 66 3 Reinhardt RL Liang HE Locksley RM 2009 Cytokine-secreting follicular T cells shape the antibody repertoire Nat Immunol 10 385 393 19252490 4 Bao K Reinhardt RL 2015 The differential expression of IL-4 and IL-13 and its impact on type-2 immunity Cytokine 5 Mohrs M Shinkai K Mohrs K Locksley RM 2001 Analysis of type 2 immunity in vivo with a bicistronic IL-4 reporter Immunity 15 303 311 11520464 6 Kaplan MH Schindler U Smiley ST Grusby MJ 1996 Stat6 is required for mediating responses to IL-4 and for development of Th2 cells Immunity 4 313 319 8624821 7 Takeda K Tanaka T Shi W Matsumoto M Minami M Kashiwamura S Nakanishi K Yoshida N Kishimoto T Akira S 1996 Essential role of Stat6 in IL-4 signalling Nature 380 627 630 8602263 8 Shimoda K van Deursen J Sangster MY Sarawar SR Carson RT Tripp RA Chu C Quelle FW Nosaka T Vignali DA Doherty PC Grosveld G Paul WE Ihle JN 1996 Lack of IL-4-induced Th2 response and IgE class switching in mice with disrupted Stat6 gene Nature 380 630 633 8602264 9 Pai SY Truitt ML Ho IC 2004 GATA-3 deficiency abrogates the development and maintenance of T helper type 2 cells Proceedings of the National Academy of Sciences of the United States of America 101 1993 1998 14769923 10 Zheng W Flavell RA 1997 The transcription factor GATA-3 is necessary and sufficient for Th2 cytokine gene expression in CD4 T cells Cell 89 587 596 9160750 11 Jankovic D Kullberg MC Noben-Trauth N Caspar P Paul WE Sher A 2000 Single cell analysis reveals that IL-4 receptor/Stat6 signaling is not required for the in vivo or in vitro development of CD4+ lymphocytes with a Th2 cytokine profile Journal of immunology 164 3047 3055 12 van Panhuys N Tang SC Prout M Camberis M Scarlett D Roberts J Hu-Li J Paul WE Le Gros G 2008 In vivo studies fail to reveal a role for IL-4 or STAT6 signaling in Th2 lymphocyte differentiation Proc Natl Acad Sci U S A 105 12423 12428 18719110 13 Ise W Kohyama M Schraml BU Zhang T Schwer B Basu U Alt FW Tang J Oltz EM Murphy TL Murphy KM 2011 The transcription factor BATF controls the global regulators of class-switch recombination in both B cells and T cells Nature immunology 12 536 543 21572431 14 Betz BC Jordan-Williams KL Wang C Kang SG Liao J Logan MR Kim CH Taparowsky EJ 2010 Batf coordinates multiple aspects of B and T cell function required for normal antibody responses The Journal of experimental medicine 207 933 942 20421391 15 Sahoo A Alekseev A Tanaka K Obertas L Lerman B Haymaker C Clise-Dwyer K McMurray JS Nurieva R 2015 Batf is important for IL-4 expression in T follicular helper cells Nature communications 6 7997 16 Vijayanand P Seumois G Simpson LJ Abdul-Wajid S Baumjohann D Panduro M Huang X Interlandi J Djuretic IM Brown DR Sharpe AH Rao A Ansel KM 2012 Interleukin-4 production by follicular helper T cells requires the conserved Il4 enhancer hypersensitivity site V Immunity 36 175 187 22326582 17 Tanaka S Tsukada J Suzuki W Hayashi K Tanigaki K Tsuji M Inoue H Honjo T Kubo M 2006 The interleukin-4 enhancer CNS-2 is regulated by Notch signals and controls initial expression in NKT cells and memory-type CD4 T cells Immunity 24 689 701 16782026 18 Schraml BU Hildner K Ise W Lee WL Smith WA Solomon B Sahota G Sim J Mukasa R Cemerski S Hatton RD Stormo GD Weaver CT Russell JH Murphy TL Murphy KM 2009 The AP-1 transcription factor Batf controls T(H)17 differentiation Nature 460 405 409 19578362 19 Tussiwand R Lee WL Murphy TL Mashayekhi M Kc W Albring JC Satpathy AT Rotondo JA Edelson BT Kretzer NM Wu X Weiss LA Glasmacher E Li P Liao W Behnke M Lam SS Aurthur CT Leonard WJ Singh H Stallings CL Sibley LD Schreiber RD Murphy KM 2012 Compensatory dendritic cell development mediated by BATF-IRF interactions Nature 490 502 507 22992524 20 Jabeen R Goswami R Awe O Kulkarni A Nguyen ET Attenasio A Walsh D Olson MR Kim MH Tepper RS Sun J Kim CH Taparowsky EJ Zhou B Kaplan MH 2013 Th9 cell development requires a BATF-regulated transcriptional network The Journal of clinical investigation 123 4641 4653 24216482 21 Sopel N Graser A Mousset S Finotto S 2016 The transcription factor BATF modulates cytokine-mediated responses in T cells Cytokine Growth Factor Rev 30 39 45 26970726 22 Ubel C Sopel N Graser A Hildner K Reinhardt C Zimmermann T Rieker RJ Maier A Neurath MF Murphy KM Finotto S 2014 The activating protein 1 transcription factor basic leucine zipper transcription factor, ATF-like (BATF), regulates lymphocyte- and mast cell-driven immune responses in the setting of allergic asthma The Journal of allergy and clinical immunology 133 198 206 e191 199 24290279 23 Voehringer D Shinkai K Locksley RM 2004 Type 2 immunity reflects orchestrated recruitment of cells committed to IL-4 production Immunity 20 267 277 15030771 24 Khoruts A Mondino A Pape KA Reiner SL Jenkins MK 1998 A natural immunological adjuvant enhances T cell clonal expansion through a CD28-dependent, interleukin (IL)-2-independent mechanism The Journal of experimental medicine 187 225 236 9432980 25 Kim ST Fields PE Flavell RA 2007 Demethylation of a specific hypersensitive site in the Th2 locus control region Proc Natl Acad Sci U S A 104 17052 17057 17940027 26 Hao B Krangel MS 2011 Long-distance regulation of fetal V(delta) gene segment TRDV4 by the Tcrd enhancer J Immunol 187 2484 2491 21784972 27 Williams A Lee GR Spilianakis CG Hwang SS Eisenbarth SC Flavell RA 2013 Hypersensitive site 6 of the Th2 locus control region is essential for Th2 cytokine expression Proceedings of the National Academy of Sciences of the United States of America 110 6955 6960 23569250 28 Urban JF Jr Maliszewski CR Madden KB Katona IM Finkelman FD 1995 IL-4 treatment can cure established gastrointestinal nematode infections in immunocompetent and immunodeficient mice J Immunol 154 4675 4684 7722320 29 Katona IM Urban JF Jr Finkelman FD 1988 The role of L3T4+ and Lyt-2+ T cells in the IgE response and immunity to Nippostrongylus brasiliensis J Immunol 140 3206 3211 2966208 30 Harvie M Camberis M Le Gros G 2013 Development of CD4 T Cell Dependent Immunity Against N. brasiliensis Infection Front Immunol 4 74 23518620 31 Mohrs K Wakil AE Killeen N Locksley RM Mohrs M 2005 A two-step process for cytokine production revealed by IL-4 dual-reporter mice Immunity 23 419 429 16226507 32 Obata-Ninomiya K Ishiwata K Tsutsui H Nei Y Yoshikawa S Kawano Y Minegishi Y Ohta N Watanabe N Kanuka H Karasuyama H 2013 The skin is an important bulwark of acquired immunity against intestinal helminths The Journal of experimental medicine 210 2583 2595 24166714 33 Liu Q Kreider T Bowdridge S Liu Z Song Y Gaydo AG Urban JF Jr Gause WC 2010 B cells have distinct roles in host protection against different nematode parasites J Immunol 184 5213 5223 20357259 34 Fairfax KC Everts B Amiel E Smith AM Schramm G Haas H Randolph GJ Taylor JJ Pearce EJ 2015 IL-4-Secreting Secondary T Follicular Helper (Tfh) Cells Arise from Memory T Cells, Not Persisting Tfh Cells, through a B Cell-Dependent Mechanism J Immunol 194 2999 3010 25712216 35 Klein Wolterink RG Serafini N van Nimwegen M Vosshenrich CA de Bruijn MJ Fonseca Pereira D Veiga Fernandes H Hendriks RW Di Santo JP 2013 Essential, dose-dependent role for the transcription factor Gata3 in the development of IL-5+ and IL-13+ type 2 innate lymphoid cells Proceedings of the National Academy of Sciences of the United States of America 110 10240 10245 23733962 36 O'Brien TF Bao K Dell'Aringa M Ang WX Abraham S Reinhardt RL 2016 Cytokine expression by invariant natural killer T cells is tightly regulated throughout development and settings of type-2 inflammation Mucosal Immunol 9 597 609 26349658 37 Scripture-Adams DD Damle SS Li L Elihu KJ Qin S Arias AM Butler RR 3rd Champhekar A Zhang JA Rothenberg EV 2014 GATA-3 dose-dependent checkpoints in early T cell commitment J Immunol 193 3470 3491 25172496 38 Reinhardt RL Liang HE Bao K Price AE Mohrs M Kelly BL Locksley RM 2015 A novel model for IFN-gamma-mediated autoinflammatory syndromes J Immunol 194 2358 2368 25637019 39 Ciofani M Madar A Galan C Sellars M Mace K Pauli F Agarwal A Huang W Parkurst CN Muratet M Newberry KM Meadows S Greenfield A Yang Y Jain P Kirigin FK Birchmeier C Wagner EF Murphy KM Myers RM Bonneau R Littman DR 2012 A validated regulatory network for Th17 cell specification Cell 151 289 303 23021777 40 Fields PE Lee GR Kim ST Bartsevich VV Flavell RA 2004 Th2-specific chromatin remodeling and enhancer activity in the Th2 cytokine locus control region Immunity 21 865 876 15589174 41 Lee GR Kim ST Spilianakis CG Fields PE Flavell RA 2006 T helper cell differentiation: regulation by cis elements and epigenetics Immunity 24 369 379 16618596 42 Lee GR Spilianakis CG Flavell RA 2005 Hypersensitive site 7 of the TH2 locus control region is essential for expressing TH2 cytokine genes and for long-range intrachromosomal interactions Nature immunology 6 42 48 15608641 43 Glasmacher E Agrawal S Chang AB Murphy TL Zeng W Vander Lugt B Khan AA Ciofani M Spooner CJ Rutz S Hackney J Nurieva R Escalante CR Ouyang W Littman DR Murphy KM Singh H 2012 A genomic regulatory element that directs assembly and function of immune-specific AP-1-IRF complexes Science 338 975 980 22983707 44 Li P Spolski R Liao W Wang L Murphy TL Murphy KM Leonard WJ 2012 BATF-JUN is critical for IRF4-mediated transcription in T cells Nature 490 543 546 22992523 45 Kim ST Fields PE Flavell RA 2007 Demethylation of a specific hypersensitive site in the Th2 locus control region Proceedings of the National Academy of Sciences of the United States of America 104 17052 17057 17940027 46 Staudt V Bothur E Klein M Lingnau K Reuter S Grebe N Gerlitzki B Hoffmann M Ulges A Taube C Dehzad N Becker M Stassen M Steinborn A Lohoff M Schild H Schmitt E Bopp T 2010 Interferon-regulatory factor 4 is essential for the developmental program of T helper 9 cells Immunity 33 192 202 20674401 47 Fallon PG Jolin HE Smith P Emson CL Townsend MJ Fallon R McKenzie AN 2002 IL-4 induces characteristic Th2 responses even in the combined absence of IL-5, IL-9, and IL-13 Immunity 17 7 17 12150887 48 Licona-Limon P Henao-Mejia J Temann AU Gagliani N Licona-Limon I Ishigame H Hao L Herbert DR Flavell RA 2013 Th9 Cells Drive Host Immunity against Gastrointestinal Worm Infection Immunity 39 744 757 24138883 49 McMillan SJ Bishop B Townsend MJ McKenzie AN Lloyd CM 2002 The absence of interleukin 9 does not affect the development of allergen-induced pulmonary inflammation nor airway hyperreactivity J Exp Med 195 51 57 11781365 50 Wilhelm C Hirota K Stieglitz B Van Snick J Tolaini M Lahl K Sparwasser T Helmby H Stockinger B 2011 An IL-9 fate reporter demonstrates the induction of an innate IL-9 response in lung inflammation Nat Immunol 12 1071 1077 21983833 51 Stetson DB Voehringer D Grogan JL Xu M Reinhardt RL Scheu S Kelly BL Locksley RM 2004 Th2 cells: orchestrating barrier immunity Adv Immunol 83 163 189 15135631 52 Harada Y Tanaka S Motomura Y Ohno S Yanagi Y Inoue H Kubo M 2012 The 3' Enhancer CNS2 Is a Critical Regulator of Interleukin-4-Mediated Humoral Immunity in Follicular Helper T Cells Immunity 36 188 200 22365664 53 Deenick EK Chan A Ma CS Gatto D Schwartzberg PL Brink R Tangye SG 2010 Follicular helper T cell differentiation requires continuous antigen presentation that is independent of unique B cell signaling Immunity 33 241 253 20691615 54 Koh BH Hwang SS Kim JY Lee W Kang MJ Lee CG Park JW Flavell RA Lee GR 2010 Th2 LCR is essential for regulation of Th2 cytokine genes and for pathogenesis of allergic asthma Proceedings of the National Academy of Sciences of the United States of America 107 10614 10619 20483988 55 Rothenberg EV 2014 The chromatin landscape and transcription factors in T cell programming Trends in immunology 35 195 204 24703587 56 Ballesteros-Tato A Randall TD Lund FE Spolski R Leonard WJ Leon B 2016 T Follicular Helper Cell Plasticity Shapes Pathogenic T Helper 2 Cell-Mediated Immunity to Inhaled House Dust Mite Immunity 44 259 273 26825674 57 Zaretsky AG Taylor JJ King IL Marshall FA Mohrs M Pearce EJ 2009 T follicular helper cells differentiate from Th2 cells in response to helminth antigens J Exp Med 206 991 999 19380637
PMC005xxxxxx/PMC5123694.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 9100013 2576 Ann Epidemiol Ann Epidemiol Annals of epidemiology 1047-2797 1873-2585 27793274 5123694 10.1016/j.annepidem.2016.08.009 NIHMS817863 Article Lipopolysaccharide-pathway proteins are associated with gallbladder cancer among adults in Shanghai, China with mediation by systemic inflammation Van Dyke Alison L. [email protected] Kemp Troy J. [email protected] Corbel Amanda F. [email protected] Zhu Bin [email protected] Gao Yu-Tang [email protected] Wang Bing-Sheng [email protected] Rashid Asif [email protected] Shen Ming-Chang [email protected] Hildesheim Allan a Hsing Ann W. [email protected] Pinto Ligia A. [email protected] Koshiol Jill [email protected] a Infections and Immunoepidemiology Branch, Division of Cancer Epidemiology and Genetics (DCEG), National Cancer Institute (NCI), 9609 Medical Center Drive, Bethesda, Maryland 20892, United States b HPV Immunology Laboratory, Leidos Biomedical Research, Inc., Frederick National Laboratory for Cancer Research, P.O. Box B, Frederick, Maryland 21702, United States c Biostatistics Branch, DCEG, NCI, 9609 Medical Center Drive, Bethesda, Maryland 20892, United States d Department of Epidemiology, Shanghai Cancer Institute, 2200 Xie-Tu Road, Shanghai, China 200032 e Department of General Surgery, Zhongshan Hospital, School of Medicine, Fudan University, 180 Fenglin Road, Shanghai, China f Department of Pathology, MD Anderson Cancer Center, 1515 Holcombe Blvd, Houston, Texas 77030, United States g Department of Pathology, Cancer Hospital, Fudan University, 399 Lingling Rd, Xuhui, Shanghai, China h Stanford Cancer Institute, Stanford University, 265 Campus Drive, Stanford, California 94305, United States Corresponding Author: Alison L. Van Dyke, MD, PhD, Postdoctoral Research Fellow, Infections & Immunoepidemiology Branch, National Cancer Institute-Division of Cancer Epidemiology & Genetics, 9609 Medical Center Drive, 6E210, Bethesda, MD 20892-9776, Phone: 240-276-6039, Fax: 240-276-7806, [email protected] 5 10 2016 31 8 2016 10 2016 01 10 2017 26 10 704709 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Purpose We examined inflammation as a mediator of associations between bacterial infection markers and gallbladder cancer (GBC). Methods Bacterial response proteins [lipopolysaccharide (LPS), soluble CD14 (sCD14), and LPS-binding protein (LBP)] were measured in 40 GBC cases and 126 gallstone controls with data on 63 serum inflammation markers. The relationships of LPS, LBP, and sCD14 with GBC were examined by logistic regression, which also was used to evaluate whether these associations are influenced by systemic inflammation as measured by a combinatorial inflammation score. Results The third versus the first tertiles of sCD14 and of LBP were associated with an increased GBC risk [Odds Ratio (95% Confidence Interval): 5.41 (2.00–16.75) for sCD14, and 6.49 (2.24–23.79) for LBP]. sCD14 and LBP were strongly associated with inflammation score (above versus below the median), which itself was associated with a >21-fold increased risk of GBC for the third vs. first tertiles. Associations between GBC and sCD14 and LBP were markedly attenuated when the inflammation score was included in the model. While LPS was not associated with GBC or inflammation, only 35% of cases and 22% of controls had detectable levels. Conclusions These findings suggest that these LPS-pathway proteins are associated with GBC via inflammation-related pathways. Gallbladder cancer Lipopolysaccharide-pathway Inflammation Mediation analysis INTRODUCTION In 2012, gallbladder cancer (GBC) was the fifth leading gastrointestinal cancer, resulting in over 178,000 new cases and over 142,000 deaths globally [1, 2]. GBC is often diagnosed after regional spread or distant metastases have already occurred. With a five-year survival rate of <5% for metastatic disease, it is one of the deadliest cancers as a result of late detection [3]. Little is known about GBC pathogenesis. The key risk factor associated with GBC development is the presence of gallstones [4]. Regular use of aspirin and other nonsteroidal anti-inflammatory drugs has been associated with decreased GBC risk, suggesting that inflammation may play a role in GBC development [5]. Relationships between inflammatory gene polymorphisms and GBC, in particular an IL8 SNP, have been reported [6]. Bacterial infections such as Salmonella enterica serovar Typhi (typhoid fever) and Helicobacter spp. also have been linked to GBC, although not all findings are consistent [7–14]. Moreover, Gram-negative bacteria have been detected in both gallstones and bile [15, 16]. We recently reported strong associations between GBC and local and systemic markers of inflammation [17], raising the possibility that chronic bacterial infection may play a role in GBC as these infections often induce chronic inflammation, systemically and locally. Bacteria that have been linked to GBC are Gram-negative bacteria that have an outer membrane containing lipopolysaccharide (LPS). Thus, LPS-pathway proteins may help elucidate the role of bacterial infections related to GBC. LPS assays are available, but conflicting results on detectability and reproducibility have been reported [18]. However, new pre-treatment kits have become available to improve LPS detection. In addition, two LPS-pathway proteins, including cluster of differentiation 14 (CD14) and LPS-binding protein (LBP), may serve as indirect measures of response to the LPS endotoxin [19, 20]. Thus, examining plasma levels of these LPS-pathway proteins may provide insight into the role of bacterial infection in GBC etiology. In this study, we measured plasma levels of LPS, sCD14, and LBP in GBC cases and GS controls in a case-control study in Shanghai, China to evaluate the role of these markers in GBC and to assess the degree to which these relationships are mediated by systemic inflammation. MATERIALS AND METHODS Study Population and Data Collection Details of the Shanghai Biliary Tract Cancer Study have been previously reported [5, 21]. Briefly, the study enrolled 368 GBC cases and 774 GS controls between June, 1997 and May, 2001. GBC cases with newly diagnosed GBC (ICD9 code: 156) between 18 and 75 years-old were ascertained from 42 hospitals in urban Shanghai, China, through a rapid reporting system to the Shanghai Cancer Registry. Case ascertainment rate (relative to all cases in Shanghai) and response rate of eligible cases were each over 90%. Control participants with documented gallstones who were undergoing either cholecystectomy or medical intervention in the same hospital as the index GBC case were enrolled and frequency-matched to GBC cases on 5-year age group and sex. Both cases and GS controls were permanent residents of urban Shanghai, China, without a prior history of cancer. The participation rate of eligible GS control patients was over 95%. This study was approved by the Institutional Review Boards at the National Cancer Institute and the Shanghai Cancer Institute, and informed consent was obtained prior to data collection. For the current analysis, we included 40 cases and 126 GS controls with previously-measured serum inflammation-related markers and available plasma for measurement of LPS, sCD14, and LBP levels. Multiplex Immune-Related Marker Panel (MIP) The multiplex Luminex-based system inflammation-related marker panel (MIP) has been previously described and validated [22]. The current study used data on 63 MIP markers measured in serum in a previous study utilizing the same population sample from the Shanghai Biliary Tract Cancer Study [23]. Detection of Plasma LPS, sCD14, and LBP Plasma LPS and LPS-pathway protein levels were quantified at the HPV Immunology Laboratory at Leidos Biomedical Research, Inc. The plasma LPS levels were quantified (EU/mL) using the EndoLISA (Hyglos GmbH, Cat#609033) kit, according to the protocol provided. Prior to quantification, the samples were pretreated as follows: diluted 1:10 with cation buffer (Charles Rivers Laboratories EndoSafe, Cat# BC1000) and water. Following heat treatment at 80°C for 10 minutes, the 1:10 diluted samples were further diluted in EndoSpecific buffer (Charles Rivers Laboratories EndoSafe, Cat# BG120), yielding a final dilution of 1:50. The plasma sCD14 levels were quantified (pg/mL) using the Quantikine ELISA kit (R&D Systems Cat# DC140) according to the provided protocol. The samples were diluted 1:800. The plasma LBP levels were quantified (pg/mL) using the Human LBP DuoSet ELISA kit (R&D Systems, Cat# DY870-05) according to the provided protocol. The samples were diluted 1:1500 with the exception of two samples that were diluted to 1:8000 because they exceeded the highest detectable level. For all three assays, the samples were assayed in duplicate; the concentration was calculated via either a four (LPS) or five (sCD14 and LBP) parameter logistic fit curve using the SoftMax Pro 6.3 (Molecular Devices, LLC) program. Blinded duplicates were included for 14 GBC cases and 20 GS controls. Statistical Analysis SAS V9.3 (Cary, North Carolina) was used for all analyses. Participant characteristics were compared between GBC cases and GS controls by univariate analysis using χ2 tests for categorical variables, t-tests for normally distributed variables, and nonparametric Wilcoxon rank-sum tests for continuous variables with a skewed distribution. Percent agreement in detection (detectable vs. undetectable) between repeated measures of LPS among GBC cases and GS controls combined and among GS controls only was calculated as the percentage of duplicate measures that were both detectable or both undetectable divided by the total number of duplicate measures. For percent agreement between duplicate measures of sCD14 and LBP, the first and second measures were classified as tertiles, respectively; percent agreement was reported as the percentage of measurements that were in the same tertile for the two measurements. In addition to percent agreement, performance parameters of the sCD14 and LBP assays among GBC cases and GS controls and also among GS controls included: coefficients of variation (CV) and intraclass correlation coefficients (ICC) for natural log-transformed duplicate measures. These measures could not be calculated for LPS due to low detection. χ2 tests were used to examine the relationships between LPS (detectable vs. undetectable) and sCD14 and LBP. Pearson rho (ρ) was used to estimate the correlation between the natural log transformed values of sCD14 and LBP. Unconditional logistic regression was used to estimate unadjusted odds ratios (OR) and 95% confidence intervals (95% CI) of associations between GBC cases relative to GS controls and sCD14 and LBP categorized as tertiles referent to the first tertile (as defined among GS controls). Given the low proportion of detection (38% among cases vs. 26% among GS controls), LPS was evaluated as detectable versus undetectable. Because of concerns of collinearity, LPS-pathway proteins were not included together in the models for GBC. Models were adjusted for sex, age group (≤54, 55–65, ≥65), and batch. To summarize systemic inflammation, a score was calculated. The MIP markers to be included in the score were chosen as the 20 MIP markers that were statistically significantly associated with GBC at P < 0.05 in an independent case-control study conducted in Chile (Chilean GBC Study) [23]. This group of markers included 13 proteins associated with increased GBC risk (CXCL13, CCL19, CCL20, CRP, CXCL11, ICAM-1, IL29-IFNG, IL-8, CXCL10, resistin, SAA, sTNFRI, and VCAM-1) and seven markers associated with a decreased GBC risk (CCL11, CCL22, sEGFR, sIL-RII, sTNFRII, CCL17, and TRAIL). The score was defined as the weighted sum of the quartile value of the MIP marker in the Shangi Biliary Tract Cancer Study (1, 2, 3, or 4). Weights assigned to the markers in the score calculation were −1 or +1 for odds ratios <1 and >1, respectively, in the Chilean GBC Study. Relationships between plasma LPS-pathway proteins and GBC and mediation by systemic inflammation were analyzed in a stepwise fashion via regression analysis (Figure 1)[24, 25]. First, unconditional logistic regression was used to investigate the relationships between plasma LPS-pathway proteins and GBC in separate models. Second, the relationship between each LPS-pathway protein (as tertiles) and systemic inflammation was analyzed by logistic regression (inflammation score as above vs. below median) adjusting for sex, age, and batch among GS controls only. Third, logistic regression was used to examine the relationship between inflammation score, as indicator variables (second and third tertiles vs. first tertile) or ordinal parameters (as tertiles), and GBC. Fourth, both the LPS-pathway marker and inflammation score were included in the model examining associations with GBC to evaluate mediation. The attenuation of associations between the LPS-pathway marker and GBC when inflammation score was entered into the model was considered a priori to be indicative of mediation of the associations between GBC and the LPS-markers by systemic inflammation. Significance of mediation was not assessed because of sample size limitations. RESULTS Participant Characteristics GBC cases and GS controls did not differ significantly in terms of demographic characteristics or medical history (Table 1). There was a statistically nonsignificant trend for higher average smoking packyears among GS controls in comparison with GBC cases [(Mean (SD): 23.2 (15.2) vs. 15.2 (14.7); P = 0.13). The majority of the GBC cases were diagnosed at stage III or IV (n = 24, 60%) with lymph node (n = 22, 55%) metastases, and ~17% (n = 7) of them had distant metastases. The most common type of stone for GS controls was mixed type (cholesterol + pigment stones) [n (%): 66 (52%)]; whereas, the stone type was missing or unclear for the 33% of GBC cases (n=13). Nine of these GBC cases without gallstone type information were stage III or above, and all of these cases either did not go to surgery or had no surgical information available. Plasma LPS, sCD14, and LBP Characteristics The percent agreement of LPS detection between duplicate measures was 53% assessed for GBC cases and GS controls combined (n = 36) or 59% when restricted to GS controls (n = 22). The percentages of agreement of replicate measures in tertiles among GBC cases and GS controls were 87% for sCD14 and 91% for LBP. Coefficients of variation were [mean (standard deviation)] 4.3% (5.3) for sCD14 and 4.3% (3.1) for LBP. When analyzed as natural log-transformed values between replicate measures among GS controls, sCD14 and LBP had ICCs of 0.96 and 0.99, respectively. While plasma natural log-transformed sCD14 and LBP levels were moderately correlated with each other (Pearson rho = 0.57; P <0.0001), neither marker (above vs. below the median) was associated with LPS detection on Chi-square analysis (P = 0.33 for sCD14 and 0.85 for LBP). Plasma LPS-Pathway Proteins, Systemic Inflammation, and GBC Plasma sCD14 and LBP levels (third versus the first tertiles) were strongly associated with GBC relative to GS controls after adjustment for sex, age, and batch with significant categorical trends in adjusted models; however, LPS detection was not associated with GBC (Table 2). While plasma sCD14 and LBP (second and third tertiles referent to the first tertile) were both associated with higher systemic inflammation score (above vs. below the median), LPS detection was not associated with systemic inflammation score (Table 3). The inflammation score was statistically significantly higher among GBC cases in comparison with GS controls [mean (standard deviation), range: 29 (9), 1 to 42 vs. 16 (8), 1 to 39; P <0.0001]. Inflammation score (third vs. the first tertile) was associated with an over 21-fold (95% CI: 5.91–143.59) increased risk of GBC with a statistically significant trend for inflammation score as a categorical variable. The relationships of plasma sCD14 and LBP with GBC were markedly attenuated and not significant after inflammation score (above vs. below the median) was included in the models (Table 2). The relationship between LPS detection and GBC remained null after inflammation score was included in Page 11 the model. In contrast, relationships between inflammation score (third vs. first tertile) and GBC were not altered substantively when plasma LPS detection was entered into the model or when sCD14 or LBP (above vs. below the median) was included in the model. DISCUSSION In summary, sCD14 and LBP were both associated with GBC and with a summary score of a panel of inflammation-related markers, associated with GBC in an independent study. This inflammation score strongly attenuated the associations between each protein and GBC, implying that inflammation mediates the relationships between these two LPS-pathway proteins and GBC. On the other hand, plasma LPS detection was not statistically significantly associated with GBC or systemic inflammation, and inclusion of inflammation score in the model did not markedly affect these null relationships. These findings argue that the relationships between sCD14 and LBP and GBC may be mediated by inflammation. The lack of a similar finding for LPS may suggest that sCD14 and LBP act through mechanisms independent of LPS, or may reflect misclassification in the LPS assay, which had a low level of detection. Bacteria have been hypothesized to contribute to GBC etiology. Bacterial DNA has been detected by RT-PCR in approximately 80% of cholesterol gallstones [26], and Gram-negative bacteria may be the primary colonizers [16]. While meta-analyses of two other Gram-negative bacteria, Salmonella enterica serovar Typhi (typhoid fever) and Helicobacter spp., showed overall associations with GBC and biliary tract cancer, the studies varied by method of carrier status detection [10, 14], and the association with Salmonella Typhi was extremely imprecise when the analysis was restricted to the cohort studies (summary estimate: 19.48, 95% CI: 0.27–1,418.18) [10]. Given these relationships between GBC and Gram-negative bacterial colonization and chronic bacterial infections, we studied LPS-pathway proteins as indirect measures of Gram-negative bacteria more broadly and found an increased risk of GBC with higher levels of sCD14 and LBP. Little research has focused on the relationship between the LPS-pathway and GBC. The gastrointestinal literature more broadly, specifically regarding colorectal and gastric cancers, also provides support for the roles of bacteria and LPS-pathway signaling in cancer development. CD14 and LBP functional genetic polymorphisms have been associated with colorectal cancer risk in a Chinese Han study population [27]. As described above, TLR4 is activated by LPS-pathway proteins, and TLR4 polymorphisms, some of which are in coding regions, have also been associated with gastric, prostate, colorectal, nasopharyngeal, and gallbladder cancers [28, 29]. The LPS-pathway gene polymorphisms in TLR2, TLR4, MD-2, LBP, and CD14 were associated with Helicobacter pylori infection and gastric cancer risk [30, 31]. Thus, there is evidence from other gastrointestinal cancers to suggest that the LPS-pathway may contribute to carcinogenesis through inflammatory processes. The mechanism underlying this inflammation-cancer link is complex, making it challenging to segregate acute and chronic inflammation stimulating cancer initiation and promotion from inflammation stimulated by the tumor itself. In addition, gallstones may induce gallbladder mucosal injury leading to inflammation and/or gallbladder mucosa bacterial colonization with subsequent innate immune system activation. Also, chronic infections may stimulate alteration of bile salts and proinflammatory and innate immune responses that together facilitate gallstone formation, as has been shown in animal and cell line studies [15, 32–34]. While it is possible that the presence of a tumor may stimulate bacterial growth, the results of the current study could not have been due to an effect of infections on gallstone development since we compared GBC cases to patients with gallstones. In addition, lymphoid infiltrate and precancerous histologic findings in the gallbladder, including mucosal hyperplasia, metaplasia and dysplasia have been more frequently observed in patients with Helicobacter pylori infection arguing for a role of bacterial involvement in GBC development [35]. When considered in the context of these findings, the relationship between the responses to Gram-negative bacteria and GBC warrants further exploration. This study has several notable strengths. It includes a unique analysis of circulating LPS and LPS-pathway proteins as indirect measures of chronic infection and gallbladder carcinogenesis utilizing well-characterized epidemiologic data. This study used GS controls as the reference group, allowing for the evaluation of GBC risk compared with patients with gallstones, the primary GBC risk factor. While systemic LPS showed marked inter-assay variability, both plasma sCD14 and LBP demonstrated strong reliability. Further the use of mediation analysis allowed for the examination of the impact of measures of systemic inflammation on these relationships to illuminate potential biological mechanisms in the GBC developmental process. Limitations of this study should be noted. The small sample size led to imprecise estimates and limited the type of mediation analysis that could be conducted; however, we were able to see significant risk estimates. This study population has been shown to have a low exposure to Salmonella Typhi, which has been hypothesized to increase GBC risk [12]; however, the measurement of LPS, sCD14, and LBP also served as broader markers of bacterial infection and of response to bacterial infection. Potentially the low level of LPS assay detection, which could contribute to misclassification bias, may have led to an underestimation of true associations with disease. Additionally, examination of circulating LPS-pathway proteins may not reflect factors acting in the gallbladder tissue itself to cause GBC. Similarly, our measurement of systemic inflammatory markers may reflect an etiologic association with GBC or stimulation of expression of systemic inflammatory markers by the tumor itself that is best followed up by analysis of measurements from a prospective cohort study. A number of future studies may be proposed from the findings in this project. In addition to optimization of LPS, LBP, and sCD14 measurements in bile, the plasma LPS endotoxin assay should be further developed for improved reproducibility and sensitivity. LPS, sCD14, LBP, and bacteria could be measured in gallbladder tissue via immunohistochemistry and/or PCR, and then be correlated with local inflammatory infiltrate, inflammatory markers in serum and bile, and corresponding pathology. Finally, a prospective approach to a cohort study is necessary to disentangle the temporality of these relationships between infection, inflammation, and gallbladder cancer. CONCLUSIONS These results suggest that systemic inflammation may mediate relationships between GBC and sCD14 and LBP; however, the impact of the tumor on systemic inflammation and LPS-pathway proteins cannot be ruled out. Collectively, these findings indicate that sCD14 and LBP may be contributing to GBC as sequela of bacterial infections or via independent mechanisms. FUNDING: We thank Ken Matsui, Ph.D. for discussions and review of the manuscript. Additionally, we honor the contribution of the late Christopher Owen Miller for his assistance with the LPS assay. This work was supported by general funds from the Intramural Research Program of the National Institutes of Health, National Cancer Institute, Division of Cancer Epidemiology and Genetics and the Office of Research on Women’s Health, National Institutes of Health. LIST OF ABBREVIATIONS CCL11 C-C motif ligand 11 CCL17 C-C motif ligand 17 CCL19 C-C motif ligand 19 CCL20 C-C motif ligand 20 CCL22 C-C motif ligand 22 CI Confidence interval COX-2 Cyclooxygenase-2 CRP C reactive protein CXCL10 C-X-C motif ligand 10 CXCL11 C-X-C motif ligand 11 CXCL13 C-X-C motif ligand 13 CV Coefficient of variation DNA Deoxyribonucleic acid ELISA Enzyme-linked immunosorbent assay EndoLISA Endotoxin-linked immunosorbent assay GBC Gallbladder cancer GS Gallstone ICAM-1 Intercellular adhesion molecule-1 ICC Intraclass correlation coefficient IL-8 Interleukin 8 IL-29 Interleukin 29 LBP LPS binding protein MD-2 Myeloid differentiation protein-2 MIP Multiplex Luminex-based system inflammation-related marker panel NFκB Nuclear factor kappa-light-chain-enhancer of activated B cells OR Odds ratio PGE2 Prostaglandin E2 RT-PCR Reverse transcriptase polymerase chain reaction SAA Serum amyloid A sCD14 Soluble cluster of differentiation 14 sEGFR Soluble epidermal growth factor receptor sIL-RII Soluble interleukin receptor 2 SNP Single nucleotide polymorphism sTNFRI Soluble tumor necrosis factor receptor 1 sTNFRII Soluble tumor necrosis factor receptor 2 TRAIL Tumor necrosis factor-related apoptosis-inducing ligand TLR Toll like receptor VCAM-1 Vascular cell adhesion molecule-1 FIGURE 1 Stepwise mediation testing with regression analysis The stepwise mediation testing via logistic regression analysis is outlined according to step and corresponding table where results are reported. Abbreviations: LPS=Lipopolysaccharide, GBC=gallbladder cancer, sCD14=soluble cluster of differentiation 14, LBP=lipopolysaccharide binding protein. Table 1 Shanghai Study Participant Characteristics by Gallbladder Case-Gallstone Control Status Characteristic Gallbladder Cancer Cases Gallstone Controls N (%) 40 (24) 126 (76) Sex Women 30 (75) 90 (71) Men 10 (25) 36 (29) Age at Interview in Years: mean (SD) a 65.2 (7.6) 63.0 (9.2) Age at Interview Group ≤54 4 (10) 18 (14) 55–65 13 (33) 37 (29) ≥66 23 (58) 71 (56) Highest Level of Education None 15 (38) 46 (37) Primary or Junior Middle School 11 (28) 28 (22) Senior Middle School 5 (13) 32 (25) Community College/University 9 (23) 20 (16) Gallstone Type Cholesterol 7 (18) 34 (27) Pigment 1 (3) 14 (11) Mixed 7 (18) 66 (52) Unclear 12 (30) 12 (10) Missing2 13 (33) 0 History of Inflammatory Bowel Disease (IBD) No 40 (100) 124 (98) History of Inflammatory Bowel Disease (IBD) Yes 0 2 (2) Crohn Disease 0 0 Ulcerative Colitis 0 1 (50) Other IBD 0 1 (50) Diabetes Mellitus No 36 (90) 107 (85) Yes 4 (10) 19 (15) Type I 0 0 Type II 4 (100) 16 (84) Unknown 0 2 (11) Not Ascertained 0 1 (5) Cigarette Smoking in Pack-years among Smokers: mean (SD) a 15.2 (14.7) 23.2 (15.2) Cigarette Smoking Status Never 31 (78) 98 (78) Former 6 (15) 10 (8) Current 3 (8) 17 (14) BMI 5 Years Ago in kg/m2: mean (SD) 24.5 (3.0) 24.0 (3.0) BMI 5 Years Ago Category (kg/m2) Underweight: 15.5 to 18.5 1 (3) 3 (2) Normal: >18.5 to <25 20 (51) 82 (67) Overweight: 25 to <30 17 (44) 33 (27) Obese: 30+ 1 (3) 4 (3) a Analyzed by nonparametric univariate Wilcoxon rank-sum test. 2 Nine of 13 (69%) of the cases where gallstone type was missing were stage III and above. Data on surgical information was either missing or indicated that surgery had not been conducted for all of the cases where gallstone type was missing. Table 2 LPS, sCD14, and LBP, Systemic Inflammation, and Gallbladder Cancer Relative to Gallstone Controls LPSa Cases N (%) Controls N (%) Sex, Age & Batch Adjusted OR (95% CI) Sex, Age, Batch, & Inflammation Scoreb Adjusted OR (95% CI) Undetectable 26 (65) 98 (78) 1.00 1.00 Detectable 14 (35) 28 (22) 1.94 (0.88–4.22) 1.74 (0.71–4.25) sCD14 Cases N (%) Controls N (%) Sex, Age & Batch Adjusted OR (95% CI) Sex, Age, Batch, & Inflammation Scoreb Adjusted OR (95% CI) Tertiles: 1st 6 (15) 40 (32) 1.00 1.00 2nd 7 (18) 43 (34) 1.19 (0.36–4.05) 0.57 (0.14–2.21) 3rd 27 (68) 43 (34) 5.41 (2.00–16.75) 1.53 (0.44–5.53) P-trend 0.0006 0.26 LBP Cases N (%) Controls N (%) Sex, Age & Batch Adjusted OR (95% CI) Sex, Age, Batch, & Inflammation Scoreb Adjusted OR (95% CI) Tertiles: 1st 4 (10) 43 (34) 1.00 1.00 2nd 10 (25) 41 (33) 2.57 (0.76–10.29) 0.99 (0.24–4.50) 3rd 26 (65) 42 (33) 6.49 (2.24–23.79) 1.97 (0.53–8.30) P-trend 0.0004 0.12 Inflammation Score Cases N (%) Controls N (%) Age & Sex Adjusted OR (95% CI) Sex, Age, LPS, & Batcha Adjusted OR (95% CI) Sex, Age, sCD14 c , & Batch Adjusted OR (95% CI) Sex, Age, LBP c , & Batch Adjusted OR (95% CI) Tertiles: 1st 2 (5) 40 (34) 1.00 1.00 1.00 1.00 2nd 3 (8) 42 (36) 1.82 (0.29–14.46) 2.00 (0.31–16.08) 1.69 (0.26–13.52) 1.71 (0.27–13.69) 3rd 35 (88) 34 (29) 21.86 (5.91–143.59) 22.57 (5.99–150.22) 21.77 (5.41–151.20) 19.57 (4.86–135.55) P-trend <0.0001 <0.0001 <0.0001 <0.0001 a LPS was analyzed as detectable versus undetectable levels. b Inflammation score analyzed as above versus below median. c sCD14 and LBP were analyzed as above versus below median. Table 3 LPS and LPS-Pathway Proteins and Inflammation Score among Gallstone Controls Marker Inflammation Score (Above vs. Below Median) Above Median N (%) Below Median N (%) Age, Sex & Batch Adjusted OR (95% CI) LPS: Undetectable Level 43 (78) 55 (77) 1.00 Detectable Level 12 (22) 16 (23) 0.96 (0.39–2.33) sCD14 a Above Median N (%) Below Median N (%) Age, Sex & Batch Adjusted OR (95% CI) Tertiles: 1st 7 (13) 33 (46) 1.00 2nd 18 (42) 25 (35) 3.93 (1.35–12.71) 3rd 30 (55) 13 (18) 19.04 (5.69–76.34) P-trend <0.0001 LBP b Above Median N (%) Below Median N (%) Age, Sex & Batch Adjusted OR (95% CI) Tertiles: 1st 6 (11) 37 (52) 1.00 2nd 23 (42) 18 (25) 6.72 (2.24–23.12) 3rd 26 (47) 16 (23) 11.48 (3.90–39.49) P-trend <0.0001 a sCD14 Cutoffs among GS controls: Tertiles <842760 (1st), 842760–1066040 (2nd), >1066040 (3rd) b LBP Cutoffs among GS controls: Tertiles: <9866475 (1st), 9866475–13396905 (2nd), >13396905 (3rd) STUDY HIGHLIGHTS 1) WHAT IS CURRENT KNOWLEDGE Gallbladder cancer has been associated with infections and inflammation. 2) NOVEL FINDINGS Lipopolysaccharide-pathway proteins are associated with gallbladder cancer. Inflammation may mediate these associations. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. 1 Ferlay J Soerjomataram I Dikshit R Eser S Mathers C Rebelo M Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012 Int J Cancer 2015 136 5 E359 86 25220842 2 Ferlay J Soerjomataram I Ervik M Dikshit R Eser S Mathers C GLOBOCAN 2012v1.0 Cancer incidence and mortality worldwide: IARC CancerBase No. 11 Lyon, France International Agency for Research on Cancer 2013 3 ACS Cancer facts and figures 2015 Atlanta, Georgia American Cancer Society 2015 4 Hsing AW Gao YT Han TQ Rashid A Sakoda LC Wang BS Gallstones and the risk of biliary tract cancer: a population-based study in China Br J Cancer 2007 97 11 1577 82 18000509 5 Liu E Sakoda LC Gao YT Rashid A Shen MC Wang BS Aspirin use and risk of biliary tract cancer: a population-based study in Shanghai, China Cancer Epidemiol Biomarkers Prev 2005 14 5 1315 8 15894693 6 Castro FA Koshiol J Hsing AW Gao YT Rashid A Chu LW Inflammatory gene variants and the risk of biliary tract cancers and stones: a population-based study in China BMC Cancer 2012 12 468 23057767 7 Dutta U Garg PK Kumar R Tandon RK Typhoid carriers among patients with gallstones are at increased risk for carcinoma of the gallbladder Am J Gastroenterol 2000 95 3 784 7 10710075 8 Glasinovic JC Marinovic I Vela P Mege RM Fernandez de la Reguera P Tobar E Cholecystectomy in young women from the south-oriental area of Santiago: comparison between 2 periods Rev Med Chil 1994 122 4 415 9 7809536 9 Mishra RR Tewari M Shukla HS Association of Helicobacter pylori infection with inflammatory cytokine expression in patients with gallbladder cancer Indian J Gastroenterol 2013 32 4 232 5 23446740 10 Nagaraja V Eslick GD Systematic review with meta-analysis: the relationship between chronic Salmonella typhi carrier status and gall-bladder cancer Aliment Pharmacol Ther 2014 39 8 745 50 24612190 11 Pandey M Shukla M Helicobacter species are associated with possible increase in risk of hepatobiliary tract cancers Surg Oncol 2009 18 1 51 6 18715780 12 Safaeian M Gao YT Sakoda LC Quraishi SM Rashid A Wang BS Chronic typhoid infection and the risk of biliary tract cancer and stones in Shanghai, China Infect Agent Cancer 2011 6 6 21535882 13 Walawalkar YD Gaind R Nayak V Study on Salmonella Typhi occurrence in gallbladder of patients suffering from chronic cholelithiasis-a predisposing factor for carcinoma of gallbladder Diagn Microbiol Infect Dis 2013 77 1 69 73 23790418 14 Zhou D Wang JD Weng MZ Zhang Y Wang XF Gong W Infections of Helicobacter spp. in the biliary system are associated with biliary tract cancer: a meta-analysis Eur J Gastroenterol Hepatol 2013 25 4 447 54 23470268 15 Peng Y Yang Y Liu Y Nie Y Xu P Xia B Cholesterol gallstones and bile host diverse bacterial communities with potential to promote the formation of gallstones Microb Pathog 2015 83–84 57 63 16 Stewart L Grifiss JM Jarvis GA Way LW Biliary bacterial factors determine the path of gallstone formation Am J Surg 2006 192 5 598 603 17071191 17 Koshiol J Castro F Kemp TJ Gao YT Roa JC Wang B Association of inflammatory and other immune markers with gallbladder cancer: Results from two independent case-control studies Cytokine 2016 83 217 25 27173614 18 Nalepka JL Greenfield EM Detection of bacterial endotoxin in human tissues Biotechniques 2004 37 3 413 7 15470896 19 Berner R Furll B Stelter F Drose J Muller HP Schutt C Elevated levels of lipopolysaccharide-binding protein and soluble CD14 in plasma in neonatal early-onset sepsis Clin Diagn Lab Immunol 2002 9 2 440 5 11874891 20 Pizzolato E Ulla M Galluzzo C Lucchiari M Manetta T Lupia E Role of presepsin for the evaluation of sepsis in the emergency department Clin Chem Lab Med 2014 52 10 1395 400 24897403 21 Hsing AW Gao YT McGlynn KA Niwa S Zhang M Han TQ Biliary tract cancer and stones in relation to chronic liver conditions: A population-based study in Shanghai, China Int J Cancer 2007 120 9 1981 5 17278101 22 Chaturvedi AK Kemp TJ Pfeiffer RM Biancotto A Williams M Munuo S Evaluation of multiplexed cytokine and inflammation marker measurements: a methodologic study Cancer Epidemiol Biomarkers Prev 2011 20 9 1902 11 21715603 23 Koshiol J Castro FA Kemp TJ Gao YT Roa JC Wang BS Association of inflammation in bile and serum with gallbladder cancer J Clin Oncol Submitted 24 Mackinnon DP Cox MC Commentary on “Mediation Analysis and Categorical Variables: The Final Frontier” by Dawn Iacobucci J Consum Psychol 2012 22 4 600 2 23180961 25 MacKinnon DP Fairchild AJ Fritz MS Mediation analysis Annu Rev Psychol 2007 58 593 614 16968208 26 Swidsinski A Khilkin M Pahlig H Swidsinski S Priem F Time dependent changes in the concentration and type of bacterial sequences found in cholesterol gallstones Hepatology 1998 27 3 662 5 9500691 27 Chen R Luo FK Wang YL Tang JL Liu YS LBP and CD14 polymorphisms correlate with increased colorectal carcinoma risk in Han Chinese World J Gastroenterol 2011 17 18 2326 31 21633598 28 Kutikhin AG Impact of Toll-like receptor 4 polymorphisms on risk of cancer Hum Immunol 2011 72 2 193 206 21081146 29 Srivastava K Srivastava A Kumar A Mittal B Significant association between toll-like receptor gene polymorphisms and gallbladder cancer Liver Int 2010 30 7 1067 72 20492496 30 Castano-Rodriguez N Kaakoush NO Pardo AL Goh KL Fock KM Mitchell HM Genetic polymorphisms in the Toll-like receptor signalling pathway in Helicobacter pylori infection and related gastric cancer Hum Immunol 2014 75 8 808 15 24929142 31 Zhao D Sun T Zhang X Guo Y Yu D Yang M Role of CD14 promoter polymorphisms in Helicobacter pylori infection--related gastric carcinoma Clin Cancer Res 2007 13 8 2362 8 17438094 32 Choi J Klinkspoor JH Yoshida T Lee SP Lipopolysaccharide from Escherichia coli stimulates mucin secretion by cultured dog gallbladder epithelial cells Hepatology 1999 29 5 1352 7 10216115 33 Hattori Y Tazuma S Yamashita G Ochi H Sunami Y Nishioka T Role of phospholipase A2 in cholesterol gallstone formation is associated with biliary phospholipid species selection at the site of hepatic excretion: indirect evidence Dig Dis Sci 2000 45 7 1413 21 10961723 34 Savard CE Blinman TA Choi HS Lee SK Pandol SJ Lee SP Expression of cytokine and chemokine mRNA and secretion of tumor necrosis factor-alpha by gallbladder epithelial cells: response to bacterial lipopolysaccharides BMC Gastroenterol 2002 2 23 12377103 35 Hassan EH Gerges SS El-Atrebi KA El-Bassyouni HT The role of H. pylori infection in gall bladder cancer: clinicopathological study Tumour Biol 2015 36 9 7093 8 25877756
PMC005xxxxxx/PMC5123703.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101631050 42499 EcoSal Plus EcoSal Plus EcoSal Plus 2324-6200 27735786 5123703 10.1128/ecosalplus.ESP-0006-2016 NIHMS780638 Article ANIMAL ENTEROTOXIGENIC ESCHERICHIA COLI Dubreuil J. Daniel Faculté de médecine vétérinaire, Université de Montréal, Québec, Canada Isaacson Richard E. Department of Veterinary and Biomedical Sciences, University of Minnesota, St Paul, MN Schifferli Dieter M. School of Veterinary Medicine, University of Pennsylvania, Philadelphia, PA 26 9 2016 10 2016 25 11 2016 7 1 10.1128/ecosalplus.ESP-0006-2016This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Enterotoxigenic Escherichia coli (ETEC) is the most common cause of E. coli diarrhea in farm animals. ETEC are characterized by the ability to produce two types of virulence factors; adhesins that promote binding to specific enterocyte receptors for intestinal colonization and enterotoxins responsible for fluid secretion. The best-characterized adhesins are expressed in the context of fimbriae, such as the F4 (also designated K88), F5 (K99), F6 (987P), F17 and F18 fimbriae. Once established in the animal small intestine, ETEC produces enterotoxin(s) that lead to diarrhea. The enterotoxins belong to two major classes; heat-labile toxin that consist of one active and five binding subunits (LT), and heat-stable toxins that are small polypeptides (STa, STb, and EAST1). This chapter describes the disease and pathogenesis of animal ETEC, the corresponding virulence genes and protein products of these bacteria, their regulation and targets in animal hosts, as well as mechanisms of action. Furthermore, vaccines, inhibitors, probiotics and the identification of potential new targets identified by genomics are presented in the context of animal ETEC. INTRODUCTION Enterotoxigenic Escherichia coli (ETEC) is the most common cause of E. coli diarrhea in farm animals (1). ETEC are characterized by the ability to produce two types of virulence factors; adhesins that promote binding and colonization of the intestinal epithelium and enterotoxins responsible for fluid secretion (2). The best-characterized adhesins are expressed in the context of fimbriae, such as the F4 (also designated K88), F5 (K99), F6 (987P), F17 and F18 fimbriae. Once established in the animal small intestine, ETEC produces enterotoxin(s) leading to diarrhea. Two major classes of enterotoxins are produced by ETEC: High-molecular weight heat-labile toxin (LT) inactivated by heating at 60°C for 15 min. and low-molecular weight heat-stable toxins (ST) which are stable to 100°C for 15 min (3). ETEC strains produce LT, STa, STb, and/or enteroaggregative heat-stable toxin 1 (EAST1). PATHOGENESIS ETEC in the environment enter animals via the oral route. The stomach, duodenum, and jejunum of animals generally do not contain coliform bacteria. The presence of ETEC in the environment of pigs, for example, is an important factor in the transmission of the pathogens as these bacteria are able to survive for at least 6 months when protected by manure (4, 5). In humans a dose of 108 to 1010 microorganisms are required for disease (6) but a lower infectious dose may be sufficient for transmission in animals. Bacteria transit and colonize the small intestine. Inside the animal it can attach to the intestinal epithelium through fimbrial or non-fimbrial adhesins (also called colonization factor (CF) antigens) by recognition of specific receptors on the small intestine and then colonize this habitat (2). There, it can multiply rapidly and reach up to 109 bacteria per gram of intestine. The degree of colonization determines whether or not disease will result from infection. Once established ETEC can synthesize and secrete one or more types of enterotoxins. Concomitant production of several enterotoxins was confirmed in piglets (7). ETEC are commonly associated with some animal species such as newborn (suckling) and weaned pigs and in newborn calves (8–10). ETEC infections are very rare or almost non-existent in other important farm animals like, rabbits, poultry or horses, but do occur in sheep. For now, no clear explanation of this species tropism can be given, as these animals do possess the receptors for the enterotoxins. We can only speculate that the gut environment of these animals is probably not appropriate for ETEC establishment and/or colonization. Close adhesion of ETEC to the intestinal epithelium permits efficient toxin delivery (11). Secretion of water and electrolytes in the intestinal lumen results from toxin activity. ETEC are known to cause rapid onset of secretory diarrhea leading to dehydration. Diarrhea is defined as soft to watery feces containing less than 10% dry matter. Lethal ETEC infections occur as a result of severe dehydration and electrolytes imbalance. ETEC affect various animal species causing profuse neonatal diarrhea in piglets, calves, sheep, and dogs and post weaning diarrhea (PWD) in piglets. In these animals, ETEC provoke diarrhea through a set of different enterotoxins. Colonization factors F5, F6, F17, and F41 are associated with ETEC strains causing neonatal diarrhea whereas F18 is generally associated with strains causing PWD, and F4 with both types of strains (12–19). ETEC-producing STa as the only toxin are associated with disease in neonatal pigs, calves, lambs, and dogs (5, 20). Pigs Diarrhea, one of the most common diseases in piglets worldwide, is transmitted from asymptomatic carrier piglets, or sows and piglets with diarrhea to naïve animals (20). Animals can be infected at early age (neonatal diarrhea) and after weaning (PWD). At birth, the pH of the piglet stomach and duodenum is relatively alkaline and production of digestive enzymes is poor favoring the establishment of infection. Diarrhea of newborn piglets is observed during the first 3–5 days of life. ETEC are frequently the primary and sole infectious agent (5). ETEC that have specific F antigens tend to be associated with a limited array of somatic O serogroups (Table 1). PWD on the other hand is common in piglets 3–10 days after weaning. It is an important cause of death in weaned pigs and occurs worldwide. The trend towards early weaning (at 3–4 weeks) may have been responsible for a concomitant increase in the occurrence of PWD. E. coli isolates from PWD are mostly ETEC and one or more strains can be found in the gastrointestinal tract of sick animals. Pigs with PWD typically have watery diarrhea that lasts from 1 to 5 days. This condition is a major cause of economic loss to the industry due to reduced growth rates and mortality. F18 and F4 are the fimbrial types commonly associated with PWD. F18 are typically associated with diarrhea of weaned pigs and F4 are associated with diarrhea in neonatal as well as weaned pigs. AIDA (adhesion involved in diffuse adherence), a non-fimbrial CF, has also been reported in E. coli from weaned pigs with PWD (21, 22). Some STb-positive strains or STb:EAST1 virulotypes of neonatal or weaned pigs may also be AIDA-positive and do induce diarrhea at least in experimental infections of neonatal pigs (22). Calves and Sheep Typically, ETEC in calves and lambs produce only STa and fimbrial adhesins F5 and F41 (1). Binding of F5 is age dependent and gradually decreases from 12 hours to two weeks of age. ETEC were implicated as the major cause of neonatal diarrhea in calves. These ETEC induce diarrhea in calves in the first four days of life, older calves or adult cattle being more resistant (10). Diarrhea accounts for more than half of all calf mortality on dairy farms. Dogs Most ETEC isolated from dogs with diarrhea are STa-positive but a small proportion of these are also STb-positive (23, 24). No LT-positive ETEC has been associated with diarrhea in this species. Many canine ETEC carry uncharacterized species-specific fimbriae. EPIDEMIOLOGY According to National Swine Surveys in the US during the nineties, diarrhea was a major cause of mortality and morbidity (25–27). In recent years, a general decrease in microbial porcine diarrhea in developed countries can likely be related to significant improvements in pig housing, management, sanitation, vaccination and biosecurity. Nevertheless, ETEC remain frequent agents of porcine diarrhea and continue to be diagnosed in neonatal and post weaning piglets that die from diarrhea in various countries, including the US (28–36), even though intestinal viruses became major topics of investigations. Notably, the presence of ETEC alone is not always sufficient for the disease to develop, indicating the influence of other factors such as feeding, weaning age, other infectious agents and season (37). The prevalence of the three major fimbriae (F4, F5 or F6) expressed by ETEC strains that colonize the intestines of neonatal piglets shows both temporal and geographic variations (Table 2). Although F4 became the major fimbriae of newborn piglet ETEC in the USA and Europe, F5 and F6 fimbriated ETEC remain a problem in some other countries. In the last 30 years, fimbriated isolates from sick pigs presented a less diverse F4- and O-serotype profile than 30–50 years ago. For example the F4ac variant became more predominant than the other F4 variants and O149 became a major O-serotype of ETEC in America, Europe and Australia (13, 28, 29, 38–42). It has been suggested that the observed variations of these serotypes over time are the result of vaccination pressures. Alternatively, successful selection of certain serotypes might relate to the changes the pig farming industry has undergone over the years. However, the adaptation of animal ETEC is not exclusively clonal (43). FIMBRIAE, A HISTORICAL PERSPECTIVE Fimbriae-mediated colonization of bacterial pathogens was first demonstrated in pigs, before similar studies were undertaken in other mammals or humans. The first described adhesive antigen of Escherichia coli strains isolated from animals was the F4 antigen (44), originally named K88 because it was thought to be a capsular antigen (K for “Kapsel” or capsule in german). Its proteinaceous nature and dependence on the presence of a plasmid was demonstrated later (45). This antigen was visible by electron microscopy as a surface–exposed filament that was thin and flexible and had hemagglutinating properties (46). Like other non-flagellar hair-like appendages on the bacterial surface, these fimbriae were also called pili. However, the designation of fimbriae is preferred by some because of historical precedence (47) or to distinguish adhesive (fimbriae) from conjugative (pili) organelles. That bacteria (48), and particularly E. coli (49), can agglutinate erythrocytes and that this property can be mediated by fimbriae (50), had been described previously. However, the fact that intestinal adhesion and colonization of E. coli in diarrheic animals (51) is mediated by fimbriae was first shown with F4 fimbriae (52). More importantly, the role of fimbriae as virulence factors was first demonstrated with these fimbriae (53). In a seminal paper, Smith and Linggood infected pigs with an enterotoxigenic E. coli (ETEC) that had either its enterotoxin- or its F4 antigen-determining plasmid missing to demonstrate that both the enterotoxin and F4 were needed to elicit severe diarrhea or death by dehydration. Moreover, in the absence of F4, significantly less E. coli could be isolated from the intestines, the difference being most impressive in the proximal portions of the small intestines. In this extensive investigation, the authors applied all the basic approaches for the study of bacterial virulence factors that were later reformulated as the molecular Koch’s postulates (54). For example, intestinal colonization and the induction of diarrhea in neonatal or weaned pigs were studied with a variety of strains, including plasmid-cured porcine ETEC strains into which the same or similar plasmids were reintroduced. Alternatively, nonpathogenic isolates were rendered pathogenic by introduction of the corresponding plasmids. To support their results, the authors even undertook competition experiments between the same strain carrying or missing either one or both plasmids. Additional studies by Jones and Rutter confirmed the role of F4 in ETEC adhesion and colonization of the small intestines of piglets (55). The latter study yielded data on the anti-adhesive properties of F4-specific antibodies. This result led the authors to the ingenious experiment demonstrating that neonatal piglets could be protected against ETEC by passive immunity acquired by colostral uptake from an immunized mother sow (56). Finally, it should be noted that the F4 gene cluster was the first fimbrial system shown to be expressed and functional when cloned into a K-12 strain of E. coli (57). This work and further studies on both the F4ac (58) and F4ab (59) genes and their products spurred the cloning (60) and genetic study of other fimbriae, such as the common type 1 and the P fimbriae (61, 62). THE ROLE OF FIMBRIAE IN ETEC PATHOGENESIS Even though the role of fimbriae in animal ETEC pathogenesis has been demonstrated with clinical strains cured of plasmids encoding the fimbrial genes, confirming experiments with defined isogenic mutants were never undertaken (54). Thus, it could be argued that the absence of other plasmid-encoded genes could have affected the results. Nevertheless, there is extensive additional evidence supporting the importance of the F4, F5 and F6 fimbriae as essential adhesive virulence factors of animal ETEC (32, 63–78). Evidence supporting the importance of the other fimbriae in Table 1 as adhesive virulence factors of ETEC is mainly epidemiological and indirect (e.g. passive protection with anti-F18 or F17 antibodies) (79–81). All the fimbriae in Table 1 have been found associated with enterotoxin (genes) on clinical isolates of animals with diarrhea. However, F17 fimbriae are also expressed by diarrheagenic non-ETEC strains or by extra-intestinal E. coli (82), (83–85). Clinical isolates of ETEC have clonal properties, since there is a preferential association of fimbriae with certain enterotoxins and O-serotypes (Table 1) that have lead to their classification into major pathotypes (86). Frequently, the same isolate carries the genes to express two or more fimbriae. Each fimbrial type carries at least one adhesive moiety that is specific for a certain host receptor, determining host species, age and tissue specificities (Table 1). Some fimbriae were not included in Table 1, because their involvement in the pathogenesis of ETEC remains unclear. The CS1541 fimbriae are expressed in vivo by porcine ETEC strains (87, 88), but they don’t bind to porcine enterocytes in vitro. Other fimbriae such as the F165 (1)(similar to Prs or F11), F165 (2)(same as F1C) and CS31A (protein capsule) fimbriae have been reported to be associated with animal ETEC strains (producing ST). However, these fimbriae are mainly expressed on non-ETEC strains of porcine or bovine origin (89–92). Most isolates expressing these fimbriae don’t co-express ETEC enterotoxins, but more typically express a cytotoxic necrotizing factor (CNF), as found in necrotoxic E. coli (NTEC) or the enteroaggregative E. coli (EAEC) heat-stable toxin (EAST-1, a diarrheagenic toxin), in the absence of an ETEC enterotoxin (93). The CS31A fimbriae don’t mediate bacterial adhesion to bovine or porcine enterocytes or their brush borders (94). Whether non-ETEC binding to enterocytes can be mediated by CS31A variants is not clear (95). No enteroadhesive property in relation to farm animals has been reported for both F165 fimbriae. Thus, both F165 and CS31A lack the typical properties of ETEC fimbriae. Some Dr-like fimbriae (Afa-7 and Afa-8) are expressed by diarrheagenic E. coli that are non-ETEC or extra-intestinal E. coli (83, 96). FIMBRIAL GENE CLUSTERS Like most studied bacterial fimbriae, the production of animal ETEC fimbriae requires sets of genes that are organized in clusters that include one and possibly more operons (Fig. 1). They all belong to various phylogenic clades of the chaperone-usher class of fimbriae (97). Most animal ETEC gene clusters for fimbriae are located on plasmids. These plasmids are quite large (40–100 kb) and usually also encode enterotoxin genes (98–104) (Table 1). In some cases, the fimbrial genes are adjacent to an enterotoxin gene, creating a pathogenicity islet that includes all the genetic determinants responsible for the symptoms of diarrhea (105, 106). However, ETEC strains typically carry several large plasmids encoding additional enterotoxins and fimbrial and non-fimbrial proteins, suggesting that the resulting redundancy of colonization and diarrheagenic factors must be beneficial to the survival or the transmission of ETEC. These same plasmids can also carry antibiotic resistance genes (107–109). The gene clusters for the fimbriae of animal ETEC (Fig. 1) encode proteins that have one of three essential functions for the production of fimbriae (110). First, two or more genes encode the structural components of the fimbriae. One protein, the major fimbrial subunit, forms most of the polymeric structure of the fimbriae, whereas the other components are incorporated as minor subunits. One (or sometimes more) subunit(s) of each fimbria carries at least one binding site for a specific mammalian host receptor. Second, the fimbrial biogenesis machineries all consist of two types of molecules, one or more periplasmic chaperones and one outer membrane protein, or usher (111). Finally, gene clusters for animal ETEC fimbriae encode typically regulatory proteins that are fimbriae-specific. Models depicting the subcellular locations of fimbrial subunits, chaperones and usher proteins involved in the biogenesis of the major fimbriae in animal ETEC are shown in Fig. 2. FIMBRIAL STRUCTURES The hair-like appearance of fimbriae is best observed by negative staining electron microscopy. Fimbriae, which can reach 2 μm in length, have been typically classified by their thickness (diameter), which varies from 2 to 7 nm (Table 3). Diameter values cited in the literature seem to vary according to the bacterial growth conditions and the staining techniques used by each investigator (112–115). One fimbrial thread or fimbria consists of the spiral arrangement of hundreds of protein subunits along a filamentous axis (116). The less the subunits are compacted along the axis, the less they share surfaces of interactions and the thinner and more flexible the fimbria appears. The structure of thick or thin fimbriae is best illustrated by a helix or spring that is compressed or stretched apart, every helical turn touching the next turn only in the former situation, and bending of the whole filamentous axis encountering the least amount of resistance in the latter situation. The model is also consistent with an axial hole (~ 2 nm), which is only visible on the electron micrograph of thicker fimbriae. Thick helical fimbriae can be stretched under certain in vitro conditions (117) or by using force to unwind the helix (118), accompanied by conformational changes in fimbrial subunits. It has been suggested that fimbrial stretching occurs in vivo to adjust and coordinate the lengths of the few hundred fimbrial threads anchoring the colonizing bacteria that are submitted to the intestinal peristalsis and its resulting shear force (119). Fimbriae-mediated bacterial adhesion to a target cell is enhanced by shear force, as described with F41 fimbriae (120). In addition, studies with adhesive subunits of a variety of fimbriae described a catch bond mechanism whereby tensile force induce stronger binding through the extension of the adhesin inter-domain linker chain and an allosteric conformational change of the binding pocket that closes around the receptor moiety like a Chinese finger trap (121). The components of animal ETEC fimbriae are listed in Table 3. Immune electron microscopy of animal ETEC fimbriae typically shows that minor subunits are located at the fimbrial tips as well as at discrete sites along the fimbrial threads (122, 123). Mechanical fragmentation of isolated type 1 fimbriae of E. coli resulted in increased adhesin-mediated binding, suggesting the uncovering of hidden or incorporated adhesive minor subunits (124). However, that minor subunits can effectively be incorporated into a fimbrial body remains controversial, since broken off fimbrial tips could increase binding by sticking along the sides of fimbrial threads. Moreover, as discussed later, the current biogenesis model of affinity-determined ordered delivery of fimbrial subunits to the periplasmic domains of the usher speaks against it (123, 125, 126). Bacterial mutants of minor subunits are typically poorly fimbriated or lack fimbriae (122, 123), in support of the involvement of minor subunits in the initiation steps of fimbrial elongation and tip localization. ADHESINS AND HOST RECEPTORS It is generally assumed that the fimbriae of ETEC strains act only as anchoring devices to serve bacterial colonization of the intestinal surface. Fimbriae-mediated ETEC colonization of the intestines in pigs induces also innate immune responses (127–129), but it is not clear what bacterial molecules are responsible, whether LPS, flagellin or other effectors. The pathogenesis of ETEC strains is specifically linked with the colonization of the small intestine and not of the large intestine. The distribution of the receptor(s) and the differential environmental signals regulating fimbrial expression in each intestinal segment (130, 131) determine the bacterial colonization sites. Many fimbriae mediate hemagglutination. Since fimbriae most frequently act like lectins by binding to carbohydrate moieties of glycoproteins or glycolipids, fimbrial receptors have frequently been studied with red blood cells of various animal species. Although hemagglutination remains a convenient way to study and classify fimbriae (132–134), some fimbriae of animal ETEC such as F6 don’t agglutinate red blood cells, or do so only after chemical-treatment (135). Another caveat is that the O- and N-glycosylation profiles and the glycosylated host molecules on red blood cells might be quite different from the ones found in the intestinal mucus and on enterocyte brush borders of the relevant animal species. The presence, modification or absence of some of these receptors in the mucus or on the brush borders varies with age, and these changes have been proposed to explain age-dependent intestinal colonization and ETEC-mediated diarrhea. For example, only newborn and weaned piglets had F4 receptors in their mucus, the latter more than the former (136), while these receptors were hardly detectable in the mucus of 6-month-old pigs (137). Similarly, intestinal cells from older pigs or calves were resistant to F5-mediated adhesion (138). This correlated to the age-dependent disappearance of the N-glycolyl group in intestinal glycolipids required for the F5 receptor activity (139, 140). Studies on the F6 receptors suggested that intestinal brush border receptors, particularly sulfatide, are released in the mucus of post-neonatal pigs inhibiting fimbriae-mediated adhesion and colonization (141–143). Lactotetraosylceramide, which was recently detected as a carbohydrate receptor moiety, was suggested to be inactivated by fucosylation in older pigs (144). In contrast, the adhesion and colonization by F18-fimbriated ETEC isolates was proposed to be dependent on receptors that develop progressively with age during the first 3 weeks after birth (145). F17-mediated bacterial binding to ileal mucus of older calves was decreased when compared with the binding to mucus of younger animals (146). The age-specific presentation and anatomical location in the intestines of the various receptor molecules for one fimbrial type don’t determine alone the susceptibility of neonatal or weaned animals to fimbriae-mediated colonization. Important are also the genetic makeup of breeds and individual animals determining whether a receptor is expressed or not, and the adaptive immune responses eliciting passive (colostral) or active protection against colonizing ETEC. The major intestinal receptors for animal ETEC fimbriae and their cognate fimbrial adhesins are listed in Table 4. F4 fimbriae and receptors Biochemical studies identified several different receptors for the F4 fimbriae and its serological variants (Table 4). That the antigenic classification of the F4 variants also determines their binding particularities (147, 148) is consistent with the identification of the major fimbrial subunit FaeG as the F4 adhesin (149, 150). By definition, the F4 fimbriae are polyadhesins (110), since FaeG accumulates two roles by constituting most of the fimbrial structure and by mediating bacterial adhesion, unlike the better studied model that have a separate major subunit and an adhesive minor subunit that locates only at the fimbrial tip. FaeG combines its structural requirements with its adhesive role by grafting an additional ligand domain on its Ig-like core (151). Recombinant FaeG inhibits bacterial F4-mediated binding to enterocytes (152) and substituting the phenylalanine at position 150 of FaeG for a serine drastically reduced the hemagglutinating property of the F4ab fimbriae, suggesting that this residue is important for intestinal binding (150). Binding studies with engineered chimeric F4ac/ad indicated that amino acids 125 to 163 of FaeG are essential for F4 variant-specific binding (153). The F4 receptor list in Table 4 is not exhaustive, other receptors of various molecular weights having been reported for the F4 fimbriae, as discussed in several reviews (154–156). Some of the additional mucin receptors might represent released degradation products of larger brush border receptors. Depending on the presence or absence of the different F4 receptors, up to 8 groups of receptor phenotypes were described (157), with six having been studied in more details (158)(Table 5). As shown with different glycoconjugate receptors, the three F4 variants demonstrate lectin activities specific for a minimal recognition sequence containing a ß-linked HexNAc, a terminal ß-linked galactose enhancing the binding (159). It is most likely that the context of this sequence on the different receptors is responsible for the binding specificities of the F4 variants. In contrast, the F4 aminopeptidase N receptor is glycosylated and biochemical evidence indicated that sialic acid was needed for binding (160), whereas transcription levels or single nucleotide polymorphisms (SNPs) could not explain the various binding profiles of three F4 variants (161). Recent studies characterized some of the porcine intestinal carbohydrate receptor moieties interacting with the F4 fimbriae (162). Crystal structure comparisons of the FaeG variants and of the FaeGad-lactose complex suggested different variant-specific binding pockets with a potential involvement of conformational changes for the adhesion process (163). Noticeably, all the intestinal ceramides that act as receptor for the F4, F5 and F6 fimbriae need to be hydroxylated (140, 162, 164, 165), indicating the importance of the lipid moiety in the binding properties of gangliosides with short carbohydrate chains. Moreover, the membrane-embedded lipid portion of a glycolipid receptor determines the orientation of the carbohydrate target on the surface of host cells, and thus, plays an essential role in the recognition by a fimbrial lectin (166). Finally, although binding studies have focused on FaeG whether some of the F4 minor subunits contribute to the adhesive properties of the fimbriae in animals remains an open question. Genetic studies located the receptors of the F4ab/ac fimbriae on porcine chromosome 13 (167, 168), in the MUC4 region (169–171). MUC4-mediated susceptibility was linked to the presence of high molecular weight glycoproteins (172). Based on linkage disequilibrium for MUC13 (173), and more specifically for six SNPs (two in MUC13) with the F4ab/ac receptor locus, this locus was located between the LMLN locus and microsatellite S0283 (174). Further studies confirmed a link between the F4ac receptor locus and MUC13, and pigs expressing at least one transcript predicted to encode a highly O-glycosylated MUC13 protein (MUC13B) were F4ac-susceptible, whereas pigs homozygous for the non-glycosylated allele (MUC13A) were F4ac resistant (175). In differing studies, transcription of either MUC13 or MUC20, another gene associate to F4ac binding (176), did not relate to the adhesive phenotype (177). Moreover, biochemical studies were unable to detect an interaction between MUC13 and F4ac and genotyping assays suggested that a yet uncharacterized M13-adjacent orphan gene participates in glycosylation of the F4ac receptor (178). In another investigation, the SNPs or transcription of 12 genes involved in the assembly of glycosphingolipid carbohydrates could not be associated to a F4 binding phenotype (179). In contradiction to early studies, recent data suggested that pigs carrying F4ab/ac receptors had greater average daily weight gains than pigs lacking these receptors or having the F4ad receptor (180), possibly contributing to the prevalence of F4 variants in western countries. F41 fimbriae and receptor DNA hybridization and gene expression studies indicated that the F41 fimbrial gene cluster is most similar to the F4 gene cluster, with the exception of the major subunit gene (181, 182). Thus, it is assumed that this fimbrial subunit acts as the F41 adhesin. An intestinal receptor for the F41 fimbriae remains to be identified, although it might include N-acetylglucosamine in a carbohydrate group that mimics one on glycophorin AM, as determined by hemagglutination assays (120, 183). A quantitative trait locus (QTL) with a suggested candidate gene (ST3GAL1) was found on Chromosome 4 (SSC4) (184). F5 fimbriae and receptors Similar to the F4 fimbriae, the major F5 subunit FanC was shown to be responsible for the hemagglutinating properties of the fimbriae (185, 186). Site-directed mutagenesis of two positively charged residues, lysine 132 and arginine 136, affected the interaction with erythrocytes known to share some of the sialylated glycolipid receptors with piglet and calf intestines (139, 140, 187–191). F6 fimbriae and receptors In contrast to F4 and F5, F6 fimbriae do not agglutinate mammalian red blood cells, but only bind to intestinal cells of neonatal piglets (141, 143, 192–194). The F6 minor subunit FasG binds specifically to porcine histone H1 proteins (143, 195, 196). FasG also mediates F6-binding to a glycolipid receptor, porcine intestinal sulfatide (142, 165). Out of twenty single arginine or lysine to alanine mutants, binding to sulfatide-containing liposomes was reduced in four cases (residues 17, 116, 118, 200) and abrogated for one mutant (lysine 117). All five mutants produced wild type levels of F6 fimbriae. It was proposed that one or more of these residues communicate with the sulfate group of sulfatide by hydrogen bonding and/or salt bridges (197). All the allelic FasG proteins with reduced binding to sulfatide still interacted like wild-type FasG with the protein receptors of porcine brush borders. At least two segments of FasG that did not include lysine117 were involved in this interaction, suggesting that different residues, and thus different domains of FasG, are required for binding to the protein and the sulfatide receptors (198). In addition to the two entero-adhesive properties of FasG, a third type of F6 binding occurs between the major subunit FasA and piglet brush border hydroxylated ceramide monohexoside (165). More recently, F6 fimbriae were shown to bind to lactotriaosylceramide and lactotetraosylceramide isolated from the intestines of 6 weeks old pigs (144). Whether these receptors are already expressed in younger neonatal piglets, which are the targets for F6-fimbriated ETEC, was not determined. F18 fimbriae and receptors For the F18 fimbriae, the minor subunit FedF of F18ab (199) was shown to act as an adhesin specific for porcine intestinal epithelial cells (200). The continuous FedF sequence from residue 60 to 109 was important for binding (201). Each of its charged residues was substituted for alanine, one at a time, and three mutations (replacing lysine 72, histidine 88 or histidine 89) significantly diminished bacterial binding to jejunal epithelial cells. Binding was abolished with a double mutant (lysine 72 and histidine 88) and the triple mutant. All these mutants produced wild type levels of fimbriae. The fedF sequence of 15 clinical isolates indicated 97% identity, with 8 sequences showing an asparagine substitution to glycine or aspartic acid at position 73 (201). Both substituted and “wild type” alleles were found in O139-serotyped strains, which typically express F18ab, and in O141-serotyped isolates, which usually express F18ac. Similar results were obtained with a study of 37 strains of F18-expressing E. coli of various countries (202). Thus, it is likely that F18ab and F18ac bind to the same receptor(s), as suggested in earlier studies showing that either F18ab or F18ac fimbriae inhibit F18-mediated bacterial attachment to enterocytes (203). Whether the adhesive function of F18 fimbriae is modulated by the reported allelic sequence profiles of fedF in clinical isolates remains an open question. Genetic studies have shown that susceptibility to F18ab-mediated enteroadhesion was inherited in pigs as a dominant trait and that it was linked to the α (1, 2) fucosyltransferase FUT1 gene on porcine chromosome 6 (204–206). These studies suggested that this glycosylase adds an essential fucose to one or more glycoconjugate receptors for the F18ab fimbriae. The role of FUT1 was consistent with the inhibition of F18-mediated receptor binding by a monoclonal antibody specific for the red blood group antigen H2, a trisaccharide that carries a terminal fucose (207), and confirmed biochemically by F18-binding to a variety of hexa- to nonaglycosylceramides from porcine intestinal epithelial cells, all carrying at least one terminal fucose (208). The crystal structure of FedF with its N-terminus bound to the blood group A and B type 1 hexaoses revealed a shallow glycan binding pocket with an adjacent polybasic loop proposed to stabilize F18 binding by interacting with intestinal cell membranes glycosphingolipids (209, 210). This polybasic loop was proposed to benefit particularly the intestinal binding of F18-ETEC to pig intestines, in the absence of inhibitory milk glycans that follows weaning (210). Several investigators in China detected that the homozygous resistance genotype occurred only in certain western breeds and associated with genotypes for other forms of resistance (211–214). F17 fimbriae and receptors The F17G minor fimbrial subunit is the adhesin of the F17 fimbriae (215). F17G mutants produce normal fimbriae that do not bind to calf intestinal epithelial cells. Binding inhibition assays suggested that the carbohydrate specificity of the F17a fimbriae is a terminal or an internal N-acetylglucosamine on O-linked oligosaccharides of bovine mucins or intestinal glycoproteins (146, 210, 216). The crystal structures of the immunoglobulin-like lectin domains of two F17 adhesins (F17G and the related adhesin GafD) that bind N-acetylglucosamine were resolved (217, 218). Although an f17G gene encoding a different N-terminal sequence was recently described, it was not carried by an ETEC strain (85). FIMBRIAL BIOGENESIS Animal ETEC use the export and assembly apparatus of protein subunits known as the chaperone-usher pathway of fimbrial biogenesis (110, 219). Although the earliest models of fimbrial biogenesis were developed by studies on K88, much of the current knowledge was obtained by more detailed investigations on the type 1 and P fimbriae. To produce fimbriae on the bacterial surface, the structural subunits have to be transported directionally through two membranes (Fig. 2). With the exception of the fimbriae-specific regulatory proteins, both the fimbrial subunit and biogenesis proteins have typical signal sequences allowing them to cross the cytoplasmic membrane by using the general secretory pathway, also designated sec-translocase (220). Having reached the periplasm, an exported fimbriae-specific chaperone associate with a fimbrial subunit that appears on the periplasmic side of the inner membrane and lends a beta-strand to complete the immunoglobulin fold of the subunit by a mechanism termed donor strand complementation (221) (222). The chaperones stabilize and protect the fimbrial subunits against proteolytic degradation and premature assembly in the periplasm. Most importantly chaperones keep the subunits in an assembly- and export-competent conformation, the energy for subunit export and assembly being maintained by the conformational state of the chaperone-associated subunits (223). After their export to the periplasm, usher proteins reach the outer membrane, where they assemble in a beta-barrel structure with the help of the BAM assembly machinery (224, 225). Fimbrial subunit release and delivery to the usher by the chaperone is coupled with the assembly of the subunit into a fimbrial fiber through a donor-strand complementation mechanism (110, 219, 226, 227). The usher molecule forms a gated channel that is required for the translocation of fimbrial subunits in a linear structure through the outer membrane (111, 221, 228). Although earlier findings suggested oligomeric and dimeric structures, recent work indicated that monomers are sufficient for fiber assembly and secretion, with a plug domain, a C-terminal domain required for filament assembly and an N-terminal domain responsible for recruiting the chaperone-subunit complexes (229). Incorporation of subunits into the linear fimbriae with the help of the usher protein results in the extension of fimbriae on the bacterial surface, where they take their final helical conformation (230). Thus, the usher also acts as an anchor for the elongating fimbrial fiber (58). Minor fimbrial subunits are frequently observed on the tip of fimbriae by immune electron microscopy. Since fimbriae grow from the base (231), tip-associated minor subunits have to be delivered to the usher before the major subunits. The role of these minor subunits can be essential in initiating fimbrial elongation, as described below for the individual fimbriae, when a mutation in a minor subunit gene results in the reduction or lack of fimbriation (122, 123). F4 biogenesis DNA and protein sequence differences of the serologically differentiated F4ab, F4ac and F4ad fimbriae are found between and within their major subunits FaeG (232). However, the sequences of the accessory genes of the F4ab and F4ac are identical (114, 233). The F4 chaperone FaeE (234–238), and the usher FaeD (239–241) are involved in the export of the adhesive major subunit FaeG and three to five minor subunits (FaeC, FaeF, FaeH, and possibly FaeI and FaeJ) (156, 221, 242–244). Genetic studies predict that the usher FaeD spans the outer membrane with 22 beta-strands, leaving relatively long N- and C-terminal ends in the periplasm (240, 241), which are likely used for subunit assembly and export as currently modeled (111, 229). FaeC, unlike FaeG, H and F, interacts only weakly or indirectly with the chaperone FaeE in the periplasm, but binds well to the usher FaeD (245). The model of F4 biogenesis suggests that FaeC locates at the fimbrial tip and plays an essential role in initiating fimbrial export and assembly, since there is no fimbriation in its absence (246–248). FaeF is thought to act as an adapter that links FaeC to the fimbriae. In addition, FaeF locates also at distinct distances along the fimbrial length, as does FaeH, and both are involved in fimbrial biogenesis since 40 to 100 times less fimbriae are expressed in their absence (122). The FaeI and FaeJ proteins share sequence similarities with the other subunits. However, their role is not clear and the corresponding mutants have no detectable phenotypes. The F41 biogenesis apparatus has not been studied in great details. However, it was shown that the F41 gene cluster contained a similar gene arrangement as the F4 gene cluster (181, 182). Moreover, F41 subunit-containing fimbriae were expressed by complementing its gene with the accessory genes from the F4 gene cluster, indicating that the F41 and F4 export and assembly system are closely related (249). F5 biogenesis The export of the F5 subunits is coordinated by the monomeric chaperone FanE (234, 250) and the usher FanD (251). The minor subunit FanF is positioned both at the tip and along the fimbrial shaft (252). This is consistent with the phenotype of a fanF mutant that produces only 0.1% fimbriae, all being short. FanG and FanH are additional minor components of the F5 fimbriae (253, 254). They associate with FanF and participate in the initiation and elongation of the fimbriae, since fanG and fanH mutants are nonfimbriated or produce shorter and less fimbriae (1–2% wild type levels). F6 biogenesis FasG, an adhesive minor subunit, is the first exported subunit of the F6 fimbriae, followed by FasF, the second minor subunit proposed to act as a linker molecule, and FasA, the major structural subunit (123, 195). FasG and FasF can be visualized as well at the tip as along the fimbrial shaft. In their absence, no fimbriae (or very rare short fibers for a fasF mutant) are expressed (106, 255). The outer membrane protein FasD is the F6 usher protein and mutagenesis studies suggested that its structure consists of a ß-barrel with 28 amphipathic ß–strands crossing the membrane (256). FasD was most accessible to proteases from the periplasmic side, implying the presence of a membrane-embedded usher with large periplasmic loops. In contrast to most other fimbrial systems, F6 fimbrial biogenesis involves three different chaperones (257). FasB is the chaperone associating with the major subunit FasA, whereas FasC acts independently of FasB as the FasG-specific chaperone. FasE, a chaperone like protein was also located in the periplasm. Although no FasE-associated Fas protein could be detected, FasE was shown to be required for optimal export of FasG. F18 biogenesis The serologically differentiated F18ac and F18ab fimbriae consist mainly of a major subunit FedA that has different protein sequences both between and within the serovars (258–261). Accessory genes and products have only been investigated for the F18ab fimbriae. However, by analogy to the F4 fimbriae, it is assumed that the F18ab and F18ac accessory genes have the same or very similar sequences. Based on their sequences and their requirement for fimbriation, the fedB and fedC genes were proposed to encode the usher and periplasmic chaperone proteins of the F18ab fimbriae (200). Two minor fimbrial subunits FedE and the adhesin FedF are not essential for fimbriation, but corresponding deletion mutants formed longer fimbriae, indicating that their products control fimbrial elongation (199). Interestingly, fedE mutants (like fedF mutants) don’t bind to porcine intestinal villi, suggesting that FedE is involved in the export and assembly of FedF. F17 biogenesis Analysis of the DNA sequence from the F17a gene cluster of a bovine ETEC strain revealed the presence of four genes (215). The chaperone F17a-D and the usher protein F17a-C are essential for the export and assembly of the major subunit F17a-A and the minor adhesive subunit F17a-G. An F17a-G mutant can make wild type levels of unaltered fimbriae that don’t bind, indicating the F17a-G adhesin is not required for fimbriation. A major subunit F17a-A mutant doesn’t bind either. This suggests that the major subunit is needed for the export, the final conformation or the presentation of F17a-G on the bacterial surface (215). REGULATION OF FIMBRIAL EXPRESSION Fimbrial gene clusters typically include one or two genes that specifically regulate the transcription of the genes in their cluster. It is generally assumed that fimbrial gene clusters have only one promoter per direction of transcription. However, when studied in more details, some gene clusters were found to carry multiple promoters and operons for the different accessory proteins (262). In addition to the fimbriae-specific regulators, each fimbrial gene cluster typically belongs to specific regulons that can be activated or repressed by global regulators such as H-NS (histone-like nucleoid structuring protein for temperature and osmolarity mediated signals) or CRP (cAMP receptor protein for catabolite repression). Some intriguing data highlighted the possibility that auto-inducer molecules expressed by the intestinal microbiota or host catecholamines have an effect on intestinal colonization by ETEC (263). Recent data indicated increased gene expression and production of F4 fimbriae by an ETEC strain grown in the presence of conditioned media and epinephrine (264), but more work is needed to dissect the corresponding mechanisms of regulation. Some fimbriae undergo phase variation. Transcriptional regulation ensures that all bacterial siblings synchronize their “on” or “off” switch for fimbrial expression. In contrast, a bacterial population regulated by phase variation always contains both “on” and “off” switched variants. The ratio of “on” and “off” variants depends on environmental growth conditions and one type of variant may be as scarce as mutants. It is thought that phase variation improves the survival rate of a bacterial population that is abruptly submitted to a new environment that selects for the scarcer phase variant. Fimbrial expression can be regulated post-transcriptionally. E. coli small RNA (sRNA) have been shown to regulate the transcription of some fimbriae (265, 266). However, it is not known whether sRNA also regulate ETEC fimbriae. F4 regulation The F4 fimbrial gene cluster carries two genes, faeA and faeB, encoding proteins that are similar to regulatory proteins of the P and S fimbriae (267). FaeA negatively controls F4 expression whereas FaeB doesn’t appear to regulate F4. Cis-active regulation of fimbrial expression is mediated by the level of methylation of three GATC sites upstream of faeB (268, 269). Dam (deoxyadenosine methylase) methylation of the first GATC site prevents the coordinate binding of the global regulator LRP and FaeA at this site; a regulatory mechanism thought to inhibit a lethal overproduction of fimbriae. Methylation of the two other GATC sites destabilizes Lrp/FaeA binding and methylation of the third site activates transcription of faeB (and downstream fae genes). The population of F4-encoded plasmids in a culture consists of a mixture of replicons that are either methylated (20%, responsible for high level expression) or nonmethylated (80%, responsible for low level expression) at the third GATC site (the first site remaining methylated and the second, nonmethylated). Thus, the methylation status of the third GATC site modulates the level of fimbrial expression. The presence of the two IS1 elements between faeA and faeB results in the lack of any detectable regulatory effects of FaeB on faeA or faeB transcription and in the constitutive expression and auto-activation of faeA. These effects were proposed to explain why the F4 fimbriae do not undergo phase variation as observed with the P fimbriae, despite the presence of similar regulatory proteins and GATC sites in cis (269). No regulatory genes have been identified for the F41 fimbriae. However, F41 expression is repressed at low temperature and in the presence of alanine, indicating that the F4 and similar F41 fimbrial systems are regulated in a different manner (270, 271). F5 regulation The production of F5 fimbriae or the transcription of the major subunit gene fanC is activated in the logarithmic growth phase, by oxygen, by low pH and by glycerol, whereas fewer fimbriae or fan transcripts are expressed in stationary phase, at high pH and in glucose-, pyruvate-, arabinose- or lactose-containing media (272, 273). Studies on several compounds as carbon source for bacterial growth indicated that acetate has a strong inhibitory effect on fimbriation and that together with glucose it essentially suppresses F5 expression (274). Fimbrial subunit production is drastically reduced in a cya mutant, confirming that F5 is regulated by catabolite repression (275). Curiously, unlike many other fimbrial systems, H-NS is not involved in the noticeable down-regulation of F5 expression at low temperature (270, 273). Moreover, two groups of F5-producing ETEC strains were distinguished based on the regulation of fimbriation by different growth conditions (276). At the 5′ end of the F5 gene cluster are two genes, fanA and fanB, that are transcribed in the same direction as the F5 genes and activate F5 expression (277). Studies on F5 gene transcription suggested that FanA and FanB act together as transcriptional antiterminators on two factor-dependent terminator sites after the fanA and fanB sequences (278). A third terminator after the fanC sequence includes a dyad symmetry for a potential stem-loop structure, suggesting that transcriptional termination at this site is factor-independent. Lrp (leucine-responsive regulatory protein) acts as a F5 transcriptional activator by binding to the fanA promoter (279). It is likely that the inhibitory effects of alanine and leucine on fan gene transcription and F5 fimbrial expression (279, 280) are mediated through Lrp. Most interestingly, the F5 gene cluster seems to consist of three operons, one for the fanA to fanD genes, a second one for the fanE and fanF genes and the third one for the fanG and fanH genes (262, 281). A stem-loop structure between fanE and fanF might act as an attenuator of fanF transcription. Putative promoters were mapped for the last two operons and the one upstream of fanE was adjacent to a CRP-binding consensus site, suggesting that the fanEF operon, like the fanABCD operon, is also regulated by catabolite repression. F6 regulation Early studies recognized that F6 fimbriae are best expressed in vivo in piglet intestines or in vitro when bacteria are grown to stationary phase, forming pellicles at the air-medium interface (64, 65). F6 fimbrial expression undergoes phase variation. Specific environmental signals or growth conditions regulate the rate of phase variation (282). The mechanisms and the potential cis- or trans-active elements regulating F6 phase variation are different from those of the F6-similar CS18 fimbriae of human ETEC (130, 131, 283, 284). Unlike CS18, no F6 DNA segment is directly regulated by DNA inversion (283, 285). Moreover, dam methylation is not involved in F6 expression or phase variation (131, 286). The apparent stability of the duplicated F6 fimbrial genes in the same clinical strain on a plasmid and on the chromosome (70, 104, 131) may suggest that the merodiploid fas genes confer some advantage to the host strain. Alternatively, the two locations of the F6 genes could be due to separate bacterial populations in the same culture, some having integrated the plasmid or a mobile DNA element carrying the F6 genes in their chromosome. Whether potentially duplicated or mobile DNA is required for phase variation in not known. However, phase variation being recA-independent, any mechanism of intrabacterial DNA exchange explaining phase variation would have to involve other recombinases (131). Expression of the major subunit FasA and of F6 is up-regulated by FasH (FapR) (104, 287). FasH shares sequence similarity with the DNA-binding domain of the AraC transcriptional activator, and more specifically with the Rns subfamily of positive regulators of fimbriae and other virulence factors of Enterobacteriaceae (288). A portion of the proximal IS1 sequence of the Tn1681 transposon located upstream of fasH is involved in activating fimbrial expression (105). The expression of fasH and fasA are both regulated in response to the carbon source and the nitrogen source (131, 284). Since these nutritional signals are differentially modulated in the intestinal environment, they may provide a mechanism to allow preferential colonization of different segments of the intestine by various enteropathogens (130, 284). F18 and F17 regulation No regulatory genes for F18 or F17 fimbriation have been described yet. Although special growth media seem to improve F18 fimbrial expression in vitro (203), not all the strains express F18 under these conditions (289). In contrast to F18ac, most F18ab fimbriae of clinical isolates are poorly expressed on commonly used media, suggesting a different mechanism of regulation for these two types of fimbriae. The F18 gene cluster is similar to the AF/R1 fimbrial gene cluster of rabbit attaching and effacing E. coli (290). Even though the F18 gene cluster lacks the upstream cis-active transcriptional regulators of AF/R1, it has an araC-like gene directly downstream fedF (291), suggesting regulation by a potential F18-specific protein. ENTEROTOXINS Two classes of ETEC enterotoxins have been described: heat-labile (LT) and heat-stable (ST). An E. coli strain may produce one or both of these types of toxins (292). Both types of toxins are plasmid-encoded. Nomenclature is based on toxin size, sequence and biological activity. LT is structurally arranged as an AB5 toxin where A is the enzymatically active subunit and the B subunits correspond to the receptor-binding moiety. LT is related to Vibrio cholerae toxin (cholera toxin, CT) a highly immunogenic molecule. In fact, antibodies against CT cross-react with LT. STs are poorly immunogenic and no immunological cross-reactivity has been observed between them. STs are single short polypeptides (less than 50 amino acids) and show no sequence similarity. Although the enteroaggregative E. coli (EAggEC or EAEC) enterotoxin (EAST1) was originally detected in human isolates of the corresponding E. coli pathotype (293), it was later shown to be also present in some porcine ETEC (294). Overall, all enterotoxins are associated with intestinal secretion of water and electrolytes in their normal hosts and/or in animal models. No significant pathological lesions or morphological changes in the intestinal mucosa result from the toxic activity of these enterotoxins (295). Many virulotypes of ETEC or classes of enterotoxin-producing E. coli that can be distinguished by their sets of virulence factors responsible for diarrheal diseases have been reported. Characteristics of ETEC responsible for enteric diseases in animals are listed in Table 6. HEAT-LABILE TOXIN Generalities In the 1950’s it was observed that some E. coli strains could cause diarrheal disease similar to cholera (296). Although most E. coli strains lack the genes for toxin production some strains secrete a heat-labile enterotoxin (LT) that is a homologue of CT produced by V. cholerae. These toxins share about 78% identity at the nucleotide level and their structures and function are very similar. The genes encoding LT are located on a large plasmid called pEnt (297). This plasmid can be transferred to non-pathogenic E. coli bacteria rendering them toxinogenic (298). LT enterotoxins are produced predominantly by human and porcine ETEC (1). In fact, the lion’s share of the knowledge on the structure and function of LT was obtained from human ETEC strains. In humans, LT causes a cholera-like disease with watery diarrhea and stomach cramps. In animals, LT-positive ETEC typically produces F4 fimbriae and STb suggesting a possible functional link between these virulence factors. However, STb and F4 are encoded on separate plasmids. Little heterogeneity among Ent genes found in porcine ETEC was observed (299). LT is part of an important group of toxins, the AB5 toxin family. Two subtypes of LT, LTI and LTII have been described. Differences between LTI and LTII are largely due to dissimilarity in their B subunit. LTI can be divided in LTIh and LTIp, produced respectively by human and porcine and human ETEC. These subtypes show slight differences in composition. LTII consists of three antigenic variants, LTIIa, LTIIb, and LTIIc (300–302) that are related to LTI in their A subunit but differ in their B subunits. In contrast to the more similar A subunit sequences, the amino acid sequences of the B subunit of LTI and LTII are highly divergent (303–305). LTII is antigenically distinct from LTI, with only 41% sequence identity with LTI but possess similar biological activities. LTII genes have been isolated from E. coli strains from humans, cows, buffalos, pigs, and ostriches. LTII toxins have mainly been observed causing disease in humans and calves (1, 306). Purified LTIIa and LTIIb caused severe diarrhea in neonatal pigs (302). LTII antigenic variants bind to various gangliosides whereas LTI binds preferentially ganglioside GM1 (Table 7). ETEC must be in close contact with the host cell to exert its effect as a semi-permeable filter could prevent toxicity expression (307). Efficient LT delivery to host cells most probably occurs via vesicles containing LT. Strains expressing LT have also been shown to have an advantage in colonization. In fact, Berberov et al., (2004) have shown that elimination of the genes for LT was associated with a concomitant reduction of toxicity and reduction in colonization of the intestine of gnotobiotic piglets demonstrating that this enterotoxin plays a role in adhesion (7). In the same way Johnson et al., (2009) have shown that LT promotes adherence of ETEC; the mechanism appearing to require the ADP-ribosyltransferase activity (308). Glucose, at an optimal concentration for LT expression, enhanced bacterial adherence through the promotion of LT production (309). Production and regulation ETEC strains do not produce similar amounts of LT (310). In fact, they can vary quite substantially in their production. On the other hand, we now know that ETEC strains that are kept frozen for long time periods just after their isolation show little change in toxin production (311). In general, conditions mimicking the human small intestine are optimal for production of LT. Growth condition influence the amount of LT produced as no detectable toxin was detected at temperatures lower than 26°C with the production increasing with temperature to reach a maximum at 37°C (312–314). Microaerophilic conditions as well as increased salt concentrations (optimum at 0.2M) promoted LT production (315). Alkaline pH is a signal for production and secretion of LT and growth medium with a pH of 8.6 resulted in optimal LT production (316). Glucose, which is found in the small intestine in appreciable concentration, increased the release of LT with maximum production at 2.5g/l (312, 314, 317). In contrast, short-chain fatty acids (in particular those with carbon chains between three and eight carbons) produced in large quantities in the colon impair production of LT (318). These conditions are believed to serve to indicate when LT production should be turn-off or turn-on. Structure LTI is a high-molecular-weight molecule (approx. 85 kDa) which activity is abolished after 15 min. at 60°C (Table 8). It consists of a bioactive A subunit and five B subunits assembled in a doughnut-shaped ring (319) that binds to GM1 ganglioside (Galβ1-3 GalNacβ1-4(NeuAc2-3) Galβ1-4Glcβ1-1 Ceramide) receptors found on the intestinal epithelial surface. Binding of LTI to GM1 located in lipid rafts on host cells is critical leading to expression of the toxic effect. Each B subunit binds cooperatively one GM1 ganglioside molecule. In addition, LT binds but more weakly to GD1b, asialo GM1, GM2, and a number of galactoproteins and galactose-containing glycolipids. In fact, this galactose-binding property was exploited for purification of LT (320). Recently, the B subunit has been shown to bind to blood group determinants (321–323). LT was observed to bind best to pig brush borders with type A blood (324) and to human erythrocytes with A and B glycolipids (325). In fact, a cohort study in Bangladesh found a high prevalence of ETEC-based diarrhea among children with blood type A or AB (326). However, it has to be proven if blood group antigens are functional receptors in vivo. In addition to its toxic activity, LT was shown to binds to the surface of E. coli cells (327, 328). LT localized to the cell surface by binding to lipopolysaccharides (LPS) found on the bacteria. LPS is present in the outer membrane of Gram-negative bacteria and consists of a characteristic lipid moiety called the lipid A, linked to a series of sugar residues (329). This LT-LPS association is independent of the A subunit (327). Even though it is bound to LPS, the B subunits remain able to bind to its mammalian cell surface receptor (330, 331). Free soluble LPS can significantly inhibit the binding of LT to the surface of ETEC strains (327). LPS lacking the O antigen was effective in blocking binding suggesting that the core sugars were responsible for binding. For full binding activity, the core sugars of LPS are required although some weaker binding activity can be observed for highly truncated LPS as 3-deoxy-D-manno-octulosonic acid (KDO) is the minimal requirement for binding (332). LT binds specifically to unphosphorylated E. coli KDO residues as was shown using V. cholerae phosphorylated KDO. LT may or may not associate with outer membrane vesicles (OMV) (333, 334) but the majority of LT activity in the extracellular environment is associated with OMVs (327, 332). OMVs are globular structures composed of lipid, of approximately 50–200 nm in diameter that are released from all Gram-negative bacteria studied yet (334, 335). Active LT molecules are found inside OMVs as well as associated with their surface (308, 328). LT within OMVs contributes to toxicity as LT can mediate internalization of entire vesicles (331). The LT-LPS association is robust and OMVs could play a role in protecting LT from proteolysis (328). Overall, more than 95% of LT is attached to the OMVs via LPS. Synthesis Individual LT toxin subunits are produced in the cytosol under the control of a joint promoter. The operon encoding the A- and B- LT subunits (eltAB) is flanked by highly conserved regions followed by variable sequences that mainly consist of partial insertion elements (336). Sequence analysis indicates that genes encoding LT were acquired by horizontal transfer from V. cholerae around 130 million years ago (337). The global regulator H-NS is involved in regulation by repressing LT expression at lower temperature (338). In high glucose condition, such as found in the duodenum, LT genes are transcribed and toxin is produced as a result of the inactivation of bacterial CRP. The CRP system ensures low level of cAMP but in presence of glucose this system is inactivated. In low glucose milieu, significant amounts of cAMP are produced and released by host cells in response to LT activity. This signal represses eltAB through CRP. Glucose limits LT production to where the toxin exerts its function (i.e. in the small intestine) corresponds to an effective targeting strategy. On the other hand, short-chain fatty acids found in the colon affect negatively LT production (339). The A (240 amino acids) and B (103 a.a.) subunits of LTI are synthesized with an N-terminal signal sequence in the cytoplasm. This sequence permits transport to the inner membrane (IM) and assembly as a holotoxin in the periplasm after the signal sequence is cleaved off. Disulfide- bond-A oxydoreductase (DsbA) aids in disulfide bond formation and peptidyl cis-trans isomerase ensures the formation of a cis-proline; together these steps facilitate folding (340). LT-A and -B subunits spontaneously assemble into holotoxins. Once formed, these complexes are remarkably stable remaining assembled from pH 2.0 to 11.0 (341, 342). Although strong acid conditions can dissociate the B pentamer, the free monomers re-oligomerize readily following neutralization. The A subunit (28 kDa) consists of an A1 fragment (22 kDa; a.a. 1–194) containing the active enzymatic site and an A2 fragment (5.5 kDa; a.a. 195–240) that links A1 to the B (11.6 kDa) subunits (Table 8). Before activation, the A1 and A2 fragments are connected through a short linker where nicking takes place. In contrast to CT this cleavage event is not required for LT toxicity to be expressed. Nevertheless, mutants unable to be nicked show a delayed toxic effect in cells in culture (343). These subunits upon nicking remain connected through a disulfide bond. In a human E. coli strain, LTI was shown to be transported across the OM by a type II secretion pathway (344). There has been no report of such a pathway in porcine ETEC. A type II secretion system consisting of a complex of 12–15 proteins spanning the IM and OM is found in numerous Gram-negative species (345). The genes encoding type II secretion apparatus are also regulated by H-NS (346). Thus, these genes are turned on by conditions that favor LT production. LT secretion rely on the B subunit but the A subunit is not involved in the process (347). Some studies have shown that secretion of LT was dependent of a protein called LeoA found to be a GTPase (348, 349). More recently, LeoA, B, and C were shown to be dynamin-like proteins responsible for potentiating ETEC virulence through membrane vesicle associated secretion (350). As few ETEC strains carry LeoA, a ubiquitous role in secretion of LT cannot be imparted to this protein. Internalization and mechanism of action Human LT enterotoxin was assimilated to a prototype and used to decipher the mechanism of action of this significant toxin. Thus, it was observed that LT must enter host cell’s cytosol to exert its toxic effect (Figure 3). LT-binding results in the clustering of GM1 gangliosides targeting more GM1. The pentamer is required for entry into cells of the intestinal epithelium and disruption of the holotoxin prevents intoxication of host cells (351). After binding of the B subunit to their specific cell receptor, internalization of LTI is mediated by receptor-mediated endocytosis. Upon endocytosis, the GM1-associated LT toxin relocates to early and recycling endosomes (352). It is then transported in a retrograde manner to the trans-Golgi network independent of the late endosome pathway (353) and the endoplasmic reticulum (ER). At the extremity of the A2 fragment an RDEL-sequence aids in the transport of the toxin to the ER. Inside this structure, the proteolytic cleavage and reduction of a disulfide bond within the A subunit provoke the release the A1 fragment in the cell cytosol (354). This fragment possesses an adenosine diphosphate (ADP)-ribosyltransferase activity that acts on the Gs heterotrimeric protein complexes found concentrated in the lipid rafts. In the cytosol, a host ADP-ribosylation factor (ARF), which is a 20-kDa regulatory GTPase, binds to the A1 subunit allowing it to bind to NAD in its active site (355). The Gs protein complex consists of a hormone stimulatory receptor, a regulatory Gs protein, and adenylate cyclase (AC) as the effector present in the basolateral membrane. Activation of Gs occurs when GTP is bound to Gsα. This component dissociates from Gsβ and Gsγ subunits and in turn activates its target, adenylate cyclase. Conversion of GTP to GDP by the intrinsic GTPase activity of Gsα acts as a turn-off switch inactivating the complex. The transfer of ADP-ribose from NAD by LT to the Gsα subunit results in inhibition of intrinsic GTPase activity. As stated previously, ADP-ribosylation is enhanced by an ADP-ribosylation factor (ARF6) that activates the A1 catalytic subunit. This modification results in activation of Gsα that turns on constitutively AC. In turn, AC produces the second messenger cyclic adenosine monophosphate (cAMP) from ATP. This second messenger targets protein kinase A (PKA) an activator of the membrane chloride channel cystic fibrosis transmembrane conductance regulator (CFTR) located in the apical membrane of epithelial cells. Activation of CFTR provokes the opening of this anion channel and results in the secretion of chloride (Cl−) and bicarbonate (HCO3−) ions from the cells into the intestinal lumen (356). The ions electrolytic balance across the epithelium is affected and the net effect is an increase in salt concentration in the intestine making it hypertonic. The osmotic pressure forces large amounts of water out of the intestinal cells. Activation of CFTR is the major way by which LT toxin provokes water efflux from the cells to induce diarrhea. Nevertheless, PKA also phosphorylates and opens a basolateral potassium channel (357). In the basolateral membrane, PKA also indirectly increases the activity of Na+/K+/2Cl− co-transporter (NKCC). This aids in the transcellular movement of chloride ions from the basolateral side of the intestinal lumen (356). Increased cAMP levels further contribute to enhance chloride secretion. In fact, cAMP inhibits the electroneutral absorption of Na+ from the intestinal lumen. The A1 subunit has also been implicated in stimulating arachidonic acid metabolism leading to the production of the secretagogue PGE2 in turn stimulating intestinal secretion (358). In contrast to CT, LTI does not stimulate the production of 5-hydroxytryptamine (5-HT or serotonin) (359). The osmotic gradient formed is responsible for the water flow outside of the cells and the observed diarrhea. Loosening of tight junctions (TJs) as a result of LT activity can also contribute to fluid loss into the intestinal lumen (360–362). In summary, LT takes advantage of the host cell’s machinery in order to exert its toxic effect and the effect of LT once initiated is irreversible. HEAT-STABLE ENTEROTOXINS Generalities E. coli heat-stable enterotoxins came to attention in the 1970s after it was observed that heat-inactivation of bacterial cultures from patients and animals suffering from diarrhea failed to eliminate enterotoxigenic activity (363, 364). Heat-stable toxins are peptidic molecules of less than 50 amino acids. Their small size and 3D-structure are responsible for resistance to boiling. Based on their sequences and their biochemical characteristics we have recognized in this group STa (also known as STI), STb (or STII) and EAST1, all associated with ETEC (295). STa STa peptide STa represents a family of toxins composed of a single peptide chain of approximately 2,000 Da. Toxins produced by human (STaH; 19 a.a.) and porcine (STaP; 18 a.a.) strains differ slightly in length and their amino acid sequence (Table 8). To date, STaP polypeptide has been observed in isolates from animal species including pigs, calves, lambs, chickens and horses and also from humans (365). In contrast, STaH is produced solely by human isolates. A 13 amino-acid peptide in the carboxy-terminus, which includes three disulfide bonds, corresponds to a highly conserved sequence and this sequence represents a common antigenic determinant. The amino-terminal sequence up to the first cysteine is not involved in toxicity (366). STa is particularly associated with ETEC that cause disease in neonatal animals. This toxin is also produced by ETEC implicated in PWD in pigs but rarely as the sole enterotoxin. STa toxins are synthesized as larger precursors (72 a.a.) that are later cleaved into the active mature toxin (367, 368). Six cysteine residues involved in disulfide bond formation are present at the same position in STaH and STaP (Table 8). The tertiary structure formed by the disulfide bonds is critical and required for full biological activity (369). Native STa toxins are poorly immunogenic. However, both homologous and heterologous antisera can neutralize toxicity (370). STa is an acidic peptide with a pI of 3.98 and a molecular weight of approximately 2000 Da (371). It is soluble in water and organic solvents including methanol and resists several proteases. The molecule is resistant to acidic but not to basic pH. Disruption of the disulfide bonds inactivates the toxin. Genetics Genes encoding STa (estA) are found on plasmids of varying molecular sizes (372). In animal ETEC isolates, it is common to find gene coding for STa, colonization factor, drug resistance and production of a colicin on the same plasmid. Genetic studies demonstrated the existence of two types of estA. This gene has an AT content of 70% and is associated with transposons that are carried on plasmids of a wide range of molecular weights. The gene cloned from a bovine isolate (STaP) was shown to be part of a transposon (Tn1681) which is flanked by inverted repeats of IS1 (373). STaP genes from ETEC isolated from other animal species (including humans) are part of the same transposon (374). Genes encoding STaH and STaP may be carried by a single human ETEC strain. Sequencing of the STa gene revealed it was identical in 52 ETEC strains of porcine origin (299). Synthesis of STa by E. coli is subject to catabolite repression and optimal yields of toxin are obtained in glucose-free media (375). Secretion and disulfide bonds formation STa is produced as a 72-amino acid precursor molecule referred to as pre-pro-STa (376, 377). This polypeptide consists of a 19 amino acid signal peptide (pre-STa), a 35 amino acid pro sequence and an 18 or 19 amino-acid mature STa. The pro-STa is translocated across the cytoplasmic membrane and requires secA-dependent transport. The signal sequence is cleaved by signal peptidase 1 (378). STa pro-region guides it into the periplasmic space (379), but this region does not seem to be involved in the extracellular transport of the peptide (380). Three intramolecular disulfide bonds are formed in the periplasm by DsbA protein prior to secretion (381–383). Proteolysis is then required to obtain biologically active 18 and 19 amino acids (378). STa toxin is secreted from the cell as it is synthesized. Mature STa molecules use TolC to cross the OM (384, 385) Formation of the three intramolecular disulfide bonds is not required for the mature toxin to pass through the OM (386). The three intramolecular disulfide bonds in STaH link cysteine 6 and 11, 7 and 15, and 10 and 18; in STaP disulfide bonds link cysteine 5 and 10, 6 and 14 and 9 and 17 (Table 8). These bonds stabilize the spatial structure (387). Using STa analogues it was demonstrated that the second disulfide bond is essential for toxicity whereas analogues without the first or the third disulfide bond showed only reduced toxicity (388). Structure and toxic domain STa tertiary structure consists of a folded peptide backbone assembled as a right-handed spiral from the first cysteine at the NH2-terminus to the last cysteine residue at the COOH-terminus (389) (Table 8). Three β-turns, located along the spiral, are stabilized by the three intramolecular disulfide bonds (390, 391). Overall, a 13 amino acid sequence from the amino-terminal cysteine to the carboxyl- terminal cysteine is essential for toxicity (392–394). This segment was defined as the toxic domain of STaH and STaP. Four amino acids (N-P-A-C) are conserved in STaP and STaH enterotoxins (Table 8). Receptor STa express its toxicity by elevation of cyclic GMP (cGMP) in intestinal epithelial cells (367, 395). Cloning the receptor from cDNA libraries of rat, pig, and human intestine led to the identification of the STa receptor (396). The deduced amino acid sequence and functional expression in mammalian cells indicated that the STa receptor is guanylate cyclase C (GC-C) belonging to the atrial natriuretic peptide receptor family (367). GC-C is a glycoprotein that is expressed primarily on intestinal epithelial cells. It is present on the brush border of villous and crypt intestinal cells. It consists of an extracellular receptor domain, a transmembrane domain and cytoplasmic domain including a kinase homology domain and a guanylate cyclase catalytic domain at the COOH-terminus (397). The endogenous agonist for GC-C was found to be a 15-amino acid hormone called guanylin (367). This hormone appears to play a role in fluid and electrolytes homeostasis in the gut. Guanylin is 50% homologous to STa containing 4 cysteine residues involved in disulfide bond formation essential for biological activity. This hormone is less potent than STa in activating GC-C and in stimulating chloride secretion (Cl−) (398, 399). STa is mimicking the hormone guanylin and the basal gut fluid homeostasis is altered through activation of GC-C. STa receptors are present throughout the human intestine and colon. Their number is decreasing along the longitudinal axis of the gut and binding of STa was noted in both crypts and villi of the small intestine and in crypts and surface epithelium of the colon (400). Binding is maximal on the villus and decreased along the villus-to-crypt axis (401). Based on the concentration of receptors, the posterior jejunum appears to be the major site responsible for STa hypersecretion of fluid. There is also good evidence that STa toxin binds to other receptors (295, 402). Mechanism of action STa is a potent toxin with rapid action but of short duration. For example, 6 ng of STa results in a positive fluid response in mouse intestine compared with 200 ng of STb or cholera toxin (CT) in the same model (403). STa exert its toxic activity through activation of an intracellular signaling cascade leading ultimately to watery diarrhea. STa receptor, is present on villus of the jejunum and ileum brush border of intestinal epithelial cells (400) (Figure 3). The binding of STa to the extracellular domain of GC-C and activation of the intracellular catalytic domain of GC-C results in hydrolysis of GTP and cellular accumulation of cGMP (404). Elevated cGMP level activates cGMP-dependent protein kinase II (cGMPKII) resulting in the phosphorylation of CFTR (395). Activation of CFTR induces secretion of Cl− and HCO3− and a net fluid secretion in the lumen of the intestine (356). Osmotically-driven water secretion results thereafter. Elevated cGMP also inhibits phosphodiesterase 3 (PDE3), resulting in an increased level of cAMP which activates protein kinase A (PKA). This enzyme phosphorylates CFTR as well as it inhibits Na+ re-absorption following phosphorylation of the Na+/H+-exchanger 3 (NHE3) (367). The effect of STa is reversible. STa and tight junctions TJs, a structure responsible for sealing the epithelium, could be held responsible, at least in part, for the electrolytes and fluid loss due to STa. TJs are highly organized structures where numerous proteins, including claudins and occludins, are involved in keeping closely associated areas of two neighboring cells whose membranes join together forming a virtually impermeable barrier to fluid and ions (405). Altering TJs could provoke loss of water and electrolytes within the intestinal lumen (406). Nakashima et al., (2013) observed that treating T84 polarized cell monolayers with STa elicited a reduction in transepithelial resistance (TER), indicating a loss of TJ integrity (407). However, no increase in paracellular permeability to a high molecular weight marker (FTIC-dextran) was noted. In contrast to STa, guanylin did not affect TER. Although both STa and guanylin induced cellular cGMP production, only STa reduced barrier integrity suggesting that STa causes not only an induction of water secretion, through channel activation, but also induces intestinal barrier dysfunction. The effect of STa on epithelial TJs contributes to the enterotoxic activity and most likely plays a role in the pathogenesis of STa-producing ETEC. EAST1 Generalities EAST1 is a peptidic toxin originally recognized in an enteroaggregative E. coli (EAEC) strain isolated from the stools of a Chilean child suffering from diarrhea (293, 408). The gene coding for this toxin was also observed in other diarrhea-causing E. coli, including ETEC, and in some other human enteric pathogens such as Salmonella (294, 409, 410). EAST1 is often compared to E. coli STa enterotoxin as it shares some physical and biological similarities. However, it does not hybridize with STa-specific DNA probes nor reacts with anti-STa antibodies (293). EAST1 is widespread among porcine ETEC in various countries (411–416). This toxin is not well characterized both in terms of function and contribution to ETEC-mediated disease. Diarrhea and sometimes death in a gnotobiotic piglet model resulted from infection with the prototype EAEC strain 17-2 (408, 417). EAST1 is associated with E. coli strains isolated from cattle and pigs with diarrhea (418–421). Alone EAST1 does not seem capable to produce disease (22) but together with LT it is efficient in producing diarrhea (7). The role of EAST1 in mediating diarrhea in animals remains controversial to this day (416). EAST1 peptide EAST1 is a heat-resistant 38 amino acid peptide with a molecular weight of 4100 Da and a calculated pI of 9.25 (Table 8) (422). Four cysteines at positions 17, 20, 24, and 27, are involved in the formation of two disulfide bridges required for toxicity expression. Unlike STa and STb, a classic signal peptide was not observed in the NH2-terminus of the predicted EAST1 sequence (423). EAST1 is immunologically distinct from STa and a polyclonal anti-STa antibody does not neutralize the biological activity of EAST1 (293). The toxic domain is comprised in a peptide spanning residues 8 to 29 (293, 423). Genetics astA gene was detected in EAEC strains 17-2 and in EAEC strain O-42 (424, 425). Variants differ by only one base at codon 21 (ACA→GCA), resulting in a change in the amino acid threonine to alanine (Table 8). Heterogeneity in virulence of EAEC strains 17-2 and O-42 (424) was reported. As strain O-42 was able to provoke diarrhea in volunteers, whereas strain 17-2 did not, it has been proposed that variant O-42 could contribute to the virulence of EAEC in a more significant way (425). Moreover, the toxin produced by strain O-42 can be observed more frequently in epidemiological studies (425, 426). Numerous other variants of EAST1 have been observed (294, 427). These molecules have usually been reported only once and probably represent less frequently distributed EAST1 variants. The toxicity of these variants has not been evaluated in animal models. The astA gene, a 117-bp-long DNA sequence encodes EAST1. There is no homology between astA and estA, the structural gene for STa. The G+C content of astA is 53%, which is similar to the mean value for E. coli (50.8%), while for estA (coding for STa) it is 30.6% (410). In EAEC strains astA is associated with a 60-MDa plasmid mediating aggregative adherence (410). astA was observed in one or more copy on plasmids of variable size but was also found on the chromosome of various bacteria, including ETEC (428). It has been detected in human, bovine, and porcine ETEC and it is commonly found on plasmids in F4-positive ETEC strains from pigs with diarrhea (421). The gene found in animals has more than 98% homology with the human isolates (7). Very little is known about the genetic regulation of EAST1 expression. Nucleotide sequences upstream of astA from porcine and bovine ETEC strains are identical, but are divergent from human ETEC. Even though results obtained by Yamamoto and Nakazawa point toward heterogeneity of DNA sequences between E. coli affecting humans and animals there seems to be a certain consensus among astA flanking sequences (429). This gene was also reported to be on transposon-like sequences, near insertion elements or inverted repeats, which could represent means by which the EAST1 gene is spread. Mechanism of action The small size and relative heat stability of EAST1 and the finding that cGMP is the molecule acting as a second messenger led to the comparison of EAST1 with STa toxin. Also, EAST1 protein shares 50% identity with the enterotoxic domain of STa (amino acid residues 6 to 18) (423). EAST1 is also structurally and functionally similar to guanylin, both having four cysteines and activating the production of cGMP (293). Interaction with GC-C could occur through the N-P-A-C motif common to STa and partially conserved for EAST1 (i.e. X-X-A-C) (Table 6). Hence, the mechanism of action of EAST1 is proposed to be identical to that of STa (Figure 3). Toxicity of EAST1 has been evaluated in Ussing chamber and in the suckling mouse assay (293, 428). No concordance between the presence of astA and toxicity of live bacterial strains or culture filtrates could be established for several strains tested (410). EAST1 did not stimulate an increase of intracellular cAMP or cGMP levels in human T84 or in porcine IPEC-J2 cell lines (430). In addition, 5-day-old gnotobiotic pigs challenged with E. coli strains, expressing EAST1 as the only toxin, did not developed diarrhea or clinical signs 72h post-inoculation. EAST1 alone seems not sufficient to cause diarrhea in 5-day old gnotobiotic pigs suggesting that EAST1 is likely not a virulence determinant in ETEC-associated diarrhea. An experimental infection with E. coli strains positive for EAST1 and AIDA and a F4/EAST1-positive strains, in gnotobiotic piglets, did not produce diarrhea either (431). Although the EAST1/AIDA strain used in the study was present in fecal shedding of challenged animals it was not markedly associated with intestinal epithelial surface. Overall, these studies showed that EAST1 toxin alone does not induce diarrhea in the animals tested. STb Generalities ETEC producing STb are associated almost exclusively with pigs and the majority of porcine ETEC produce STb enterotoxin. This toxin has also been detected in ETEC of human origin (432, 433). STb is recognized as a potent enterotoxin in weaned pigs (292, 434–436). STb peptide STb-positive E. coli strains have been isolated principally from pigs but also sporadically from cattle (including water buffaloes), chickens, dogs, cats, ferrets and humans (437). STb comprises 48 amino acids with four cysteine residues involved in disulfide bridges formation (Table 8). The enterotoxin has a Mr of 5,200 Da and bears no homology to STa or EAST1 enterotoxins. STb peptide is synthesized as a 71 amino-acid precursor comprising a 23 amino-acid signal sequence (438, 439). The first seven amino acids at the NH2-terminus of the mature toxin are not involved in either the structure or toxicity (440). The peptide spanning from Cys10 to Cys48 has full biological activity. STb isolated from various animal species have the same nucleotide and amino acid sequences. Nevertheless, a STb variant with a His to Asn substitution at position 12 was identified in E. coli isolates from pigs suffering from diarrhea (441). This variant shares structural and mechanistic properties with wild-type STb (442). No differences in biological activity of the variant have been reported. STb is a highly basic protein with an isolelectric point of 9.6 (443). It is soluble in water and some organic solvents but is insoluble in methanol and the toxin loses biological activity following ß-mercaptoethanol or trypsin treatment (363, 444, 445). It resists acid (pH 2), alkaline (pH 12) and 8 M urea treatments (444). STb is very susceptible to protease degradation (446, 447). STb is poorly immunogenic. A serological response can be obtained following immunization with either fusion proteins or proteins chemically coupled to STb (448–451). The anti-STb antibodies can neutralize STb toxicity but are unable to neutralize STa or CT toxins (452). Genetics The estB gene encodes STb. This gene is found on heterogeneous plasmids that often code for other properties including other enterotoxins, colonization factors, drug resistance, colicin production and transfer functions (372, 453). It is part of a transposon of approximately 9 kb designated Tn4521 (454–456). This transposon is flanked by defective IS2 elements but it is functional as estB can transpose from one plasmid to another. The structural gene for STb from different clinical isolates is uniform in size but the flanking sequences are heterogeneous suggesting that estB could be found on different transposons. Transposition of estB is probably one of the mechanisms by which this virulence factor is disseminated among ETEC. The promoter for estB expression is weak (457), capable of binding RNA polymerase, but seems to be a poor transcription initiator and hence very little STb is produced. Production is under the control of environmental conditions. STb synthesis by wild-type E. coli strains varies with the composition of the culture medium (444, 458) and a repressive effect of glucose on STb production was reported (459). Secretion and disulfide bond formation Intramolecular disulfide bonds must be correctly formed in order to produce an active STb toxin. The STb polypeptide is synthesized as a 71 amino acids precursor (438, 439). The NH2-terminus of pre-STb, residues 1–23, has characteristics of a signal sequence that is cleaved by a signal peptidase during export to the periplasm using the Sec export system (460). Thus, an 8.1-kDa precursor (pre-STb) is converted to a transiently cell-associated 5.2-kDa form consisting of 48 amino acids. Conversion of pre-STb to cellular STb depends on the secA gene product. Translocation of the precursor to the periplasm requires energy. The export of STb relies on the general export pathway of E. coli. STb is detected as a cell-associated molecule and an indistinguishable extracellular form becomes apparent, indicating that no proteolytic processing occurs during mobilization of STb from the periplasm to the culture supernatant. Mature STb is found preferentially in the culture supernatant (460). STb was absent from the culture supernatant of dsbA and tolC defective mutants, indicating that these genes are required for secretion (461). Two intramolecular disulfide bonds must be formed for the efficient secretion of STb (462). Elimination of either one of the bonds renders the toxin susceptible to periplasmic proteolysis. STb is exported across the OM through TolC involving accessory proteins. MacAB, an ABC transporter, interacts with TolC and participates in secretion (463, 464). MacAB probably binds the toxin in the periplasm and transports it through the pore formed by TolC (465, 466). MacA is a subunit of the MacAB transporter stimulating the MacB ATPase activity (467). The central region of STb from amino acid 19 to 36 is involved in the secretory process (468) and DsbA is necessary for STb to adopt a structure that can then cross the OM. Toxic domain and structure A nuclear magnetic resonance study established that STb is helical between residues 10 and 22 and residues 38 and 44 (Table 8) (440). The helical structure in the region 10–22 is amphipathic, exposing several polar residues to the solvent. The loop region between residues 21 and 36 contains a cluster of hydrophobic residues. The integrity of the disulfide bonds is crucial for the structure and function of the toxin as already discussed. Oligomerization was observed for STb resulting in the formation of hexamers and heptamers (469). The region responsible for this process comprises hydrophobic residues M, I and F, found close to or in the hydrophobic α-helix (a.a. 37 to 42). STb receptor A glycosphingolipid present in high number in the plasma membrane was shown to be the STb receptor (470, 471). Binding of STb to commercially available glycosphingolipids was evaluated (472). STb binding varied greatly depending on the molecule tested. Sulfatide (SO43-galactosyl-ceramide) was the molecule to which STb bound with greatest affinity. The reaction was dose-dependent and saturable. Total lipid extraction of pig jejunum epithelium and thin-layer chromatography indicated the presence of sulfatide on this tissue. A mass spectrometer analysis on the lipids isolated following high performance thin layer chromatography of pig jejunum epithelium confirmed the nature of the receptor (473). A dissociation constant of 2.4 ± 0.61 nM for the STb-sulfatide interaction was observed (474). The functionality of sulfatide, a widely distributed acidic glycosphingolipid on the intestinal epithelium, was finally proven in the rat ligated loop assay (472). The binding site of STb with sulfatide is comprised between a.a. 18 to a.a. 30. The region responsible is partly within the amphipathic α-helix and the flexible loop rich in glycine (475, 476). Mechanism of action STb is a toxin with rapid action but of moderate potency (477). In mouse intestinal loops, purified toxin elicits a response in 30 min and fluid accumulation reach a maximum after about 3 h (403). STb stimulated a cyclic nucleotide-independent secretion. This enterotoxin is thus a cytotonic toxin with properties and a mechanism of action different from STa. In vivo significant accumulation of Na+ and Cl− occur intraluminally following STb intoxication. As well, STb stimulates bicarbonate (HCO3−) secretion (478, 479). The level of prostaglandin E2 (PGE2) in the intestinal intraluminal fluid increases as a result of STb action (480). The quantity of PGE2 produced by intestinal cells is directly related to the dose of STb administered and the quantity of PGE2 correlated with the volumes of fluid released into the intestinal lumen. Levels of arachidonic acid is also elevated following STb intoxication, indicating that arachidonic acid metabolism is stimulated possibly through phospholipase A2 activity. The mode of action of STb may be somewhat similar to that of CT stimulating the release of both PGE2 and 5-HT and suggesting a potential effect on the enteric nervous system (3). STb could also act directly on the muscle cells of the ileal serosa increasing the spontaneous motility of the intestine resulting in contractions (403). Atropine could not abolish toxicity indicating it was not the result of the excitation of cholinergic nerves. Papaverine, which causes relaxation of smooth muscles, can inhibit STb implying that STb could act directly on muscle cells. Internalization of STb toxin within rat intestinal jejunum epithelium was observed (481) and the process was also confirmed using a confocal microscope and NIH-3T3 cells (482). When cells are intoxicated, STb binds through its galactose sulfate moiety to an acidic glycosphingolipid, sulfatide, a molecule widely distributed on intestinal epithelial cells (470)(Figure 3). STb stimulates a GTP-binding regulatory protein resulting in a Ca++ level increase inside the cell activating CAMKII (483). Activation of protein kinase C (PKC) is induced following the Ca++ increase and phosphorylation of the CFTR ensues (483, 484). PKC also inhibits Na+ uptake by acting on an unidentified Na+ channel. CAMKII opens a calcium-activated chloride channel (CaCC) and could as well be involved in phosphorylation of CFTR. The increased Ca++ levels also influence the activities of phospholipases A2 and C leading to the release of arachidonic acid from membrane phospholipids and formation of prostaglandin E2 (PGE2) and 5-HT (or serotonin), two secretagogues from enterochromaffin cells. Both compounds mediate the transport of H2O and electrolytes out of the intestinal cells by a yet unknown mechanism (452, 485–487). Using brush border membrane vesicles (BBMV) isolated from piglet jejunum and a membrane-potential-sensitive probe, STb was shown to permeabilize the intestinal tissue. It can do so by forming nonspecific pores confirming previous trypan blue and PLB studies (473, 488). An electrophysiological study using planar lipid bilayers (PLB) where the receptor for STb was reconstituted into large unilamellar vesicles made of PE osmotically fused to PE:PC:cholesterol (7:2:1) showed resolved channels currents (489). Thus, STb appears to be involved in the opening of a voltage-dependent channel. The previous observation that STb forms oligomers (469) can indicate that STb may allow formation of pores that could alter the cellular membrane. Permeabilization of intestinal cells in vitro was observed without cell death. The formation of pores/channels within the plasma membrane may constitute a signaling event triggering fluid secretion associated with diarrhea. In human (human colon tumor, HRT-18) and animal (rat ileum epithelium, IEC-18) cell lines, caspase-9, the initiator of mitochondrion-mediated apoptosis and caspase-3, an effector of caspase-9, were both activated following STb intoxication (490). DNA fragmentation was observed as well as condensation and fragmentation of nuclei. Overall the data indicated that STb toxin could induce, at least in these cell lines, a mitochondrion-mediated caspase-dependent apoptotic pathway. STb and tight junctions A significant reduction of TER parallel to an increase in paracellular permeability to BSA-FITC for STb-treated cells was noted in T84 human colon cells compared to untreated cells or cells treated with a non-toxic STb mutant (491). The increase in paracellular permeability was associated with a marked alteration of F-actin stress fibers. F-actin filaments dissolution and condensation observed in the presence of STb were accompanied by redistribution and/or fragmentation of ZO-1, claudin-1, and occludin. An 8 amino acid peptide (GFLGVRDG) present in STb sequence and corresponding to a consensus sequence of Vibrio cholerae Zona occludens toxin (Zot), affected T84 cells in the same way as STb (492, 493). A scrambled octapeptide (STb24-31) showed no effect compared to untreated cells. Further studies showed that STb provoked a redistribution of claudin-1, a protein playing a major role in TJ permeability. Claudin-1 was displaced from TJs and found in the cytoplasm. The loss of this protein from TJs was accompanied by its dephosphorylation (494). Thus, STb induces epithelial barrier dysfunction through changes in TJ proteins that could contribute to diarrhea. More studies are required to understand the pathways involved in STb-mediated alteration of TER and TJ proteins modulation. OTHER VIRULENCE FACTORS Although fimbriae with their adhesins and enterotoxins remain the characteristic virulence factors of enterotoxigenic Escherichia coli, additional encoded proteins can play various roles in the pathogenesis of these bacteria (495). For example and as mentioned earlier, the autotransporter protein AIDA is a non-fimbrial adhesin that is expressed by certain strains causing PWD in pigs (31, 496, 497). Iha is another adhesin usually associated with shigatoxin-producing E. coli that can be detected in some animal ETEC (495, 498). In addition to AIDA and Iha, it is likely that animal ETEC express other virulence factors, including some non-fimbrial adhesins and proteases described for certain human ETEC strains, such as the adhesins and invasins Tia and autotransporter protein TibA (499–502), the autotransporter adhesin TleA (503), the host-activated adhesin EaeH (504, 505), the two partner secreted EtpA adhesin that binds to the flagellar tip and acts as an adhesive bridge (506), the mucin-degrading proteases EatA, an autotransporter serine protease, and YghJ, a secreted metalloprotease (507, 508). Genomic studies and molecular epidemiology will help to evaluate the presence, distribution and frequencies of these or similar virulence factors in animal ETEC strains (509–511), whereas expression and functional investigations will be needed to identify new animal specific factors and potential host-adapted activation dependency (504). VACCINES AND COLONIZATION INHIBITORS Neonatal diarrhea vaccines Studies on ETEC fimbriae have helped to better understand the biology and role of these organelles in pathogenesis; they have also opened the door to new diagnostic, prophylactic, and therapeutic tools. Following on the seminal studies of Rutter and Jones (55, 56, 512, 513) demonstrating that colostral antibodies induced by maternal immunization protected neonatal piglets, many additional in vitro and in vivo studies confirmed that fimbriae are highly immunogenic proteins and that the induced antibodies protect by inhibiting adhesion to enterocytes and intestinal colonization (67–69, 73, 76, 81, 514–517). Studies with ETEC strains of veterinary relevance have led to the development of effective parenteral anti-adhesive vaccines based on four types of fimbriae, F4 (K88), F5 (K99), F6 (987P), and F41. In general, these vaccines have been quite successful in the prevention of neonatal diarrhea in piglets and calves. Vaccination of dams is a cost-effective health management strategy to prevent ETEC diarrhea in neonates (32, 518). However, immunity generated by the current generation of vaccines that are based on lacteal immunity is not effective in the prevention of PWD (discussed further below). Fimbriae are thought be very good immunogens because they are proteinaceous and contain a set of epitopes that are repeated 102 to 103 times on each fimbrial thread. Because bacteria have multiple copies of fimbriae on their surfaces, each bacterium can contain as many as 105 to 106 epitopes on each bacterial surface. On the other hand, recent data has shown that the use of a single fimbrial subunit from F4, the fimbrial adhesin subunit FaeG, does not provide good protection unless it is found in the polymerized fimbrial structure (232). Likewise, it has been shown that F5 needs to be assembled to induce immunity (519).Thus it is likely that vaccines for food animals produced using technologies that focus only on fimbrial subunits as immunogens will not result in good protection compared to polymerized, mature fimbrial strands. Notably, it has been shown that when foreign epitopes genetically engineered into fimbrial subunits and displayed in a polymeric form on attenuated live bacterial can be used to increase their immunogenicity (520–525). Because there are antigenic variants of some fimbriae, such as F4 (F4ab, F4ac, and F4ad), an important question is whether all three types are required in a vaccine for efficacy against all three types? Fortunately, there is strong immunologic cross reactivity between the three variants presumably based on the common “a” antigen. Thus cross-reactivities of the major and minor structural subunits of F4, results in good protection protection regardless of the origin of F4 in the vaccine if the vaccine contains polymerized F4. However, since there is no cross reactivity between F4, F5, F6, and F41 each of these types of fimbriae needs to be in the vaccine for the broadest level of protection. Among various advantages of using assembled ETEC fimbriae in vaccines are their relative resistance to enteric proteases and their ability to induce mucosal immunity; fimbriae are also fairly easy to extract intact from whole bacterial cells. Other approaches for vaccines for ETEC include the use of formalin-treated fimbriated ETEC, the use of live fimbriated but nontoxigenic ETEC, or the use of attenuated Salmonella enterica mutants expressing cloned fimbriae. These vaccines have been used to immunize animals by the oral route and to successfully induce anti-adhesive antibodies (526–530). Oral vaccinations, combined with parenteral applications, can increase and prolong the duration of lacteal immunity (529). A potential advantage of fimbriae of enteric pathogens is that they possess enteroadhesive properties, which they share with other mucosal immunogens such as the enterotoxins. The binding of fimbriae to their complementary intestinal receptors in the appropriate host species is important for the activation of mucosal immunity after oral immunization, as shown with the F4 fimbriae (531, 532). Other carrier bacteria have also shown some level of usefulness inducing anti-F5 antibodies including Lactobaccilus acidophilus (533). The use of plant-based vaccines also has the subject of investigations. The gene encoding the main subunit of F4 and its adhesin (faeG) has been cloned into tobacco, barley, and alfalfa genomes and these plants have been shown to express faeG (152, 534, 535). Using an in vitro binding assay that employed pig villi, the FaeG produced in tobacco and barley was shown to inhibit binding of F4 positive E. coli to the villi. Using the alfalfa-based vaccine, these investigators also showed that the vaccine could reduce shedding of ETEC in challenged pigs. The reduction in shedding was equal to vaccination with purified F4 fimbriae. Fimbrial specific monoclonal antibodies also have been used in protection against neonatal ETEC diarrhea. This approach has been most successfully applied for use in calves (536, 537) where monoclonal antibodies against F5 were administered to neonatal calves that also were challenged with an F5-expressing ETEC. F5-producing ETEC are the most common ETEC found in cattle. Calves receiving the F5 specific monoclonal antibody had reduced levels of diarrhea and mortality compared to non-vaccinated controls. However, passive vaccination with anti-fimbrial monoclonal antibodies is expensive and labor intensive (32). Post weaning diarrhea vaccines While the current generation of commercial vaccines confer excellent protection of neonatal piglets and calves against ETEC, these vaccines have not been shown to be effective against post weaning ETEC infections in pigs. PWD due to ETEC remains an important disease of young pigs and causes significant morbidity and mortality immediately after weaning. The reasons why the current vaccines are not effective for PWD are probably related to a loss of protective antibodies in piglets receiving colostrum and milk and the subsequent loss from circulation. Titers of anti-fimbrial antibodies in milk fade by the time pigs are weaned. Since it is believed that the mechanism of protection against ETEC is by blocking adhesion to enterocytes and by agglutination of the ETEC in the lumen of the intestines, the lack of antibodies in the intestinal lumen post weaning leads to a lack of protection against the intestinal colonization by the ETEC. Consistent with this is the observation that the feeding of anti-F18 antibodies to weaned pigs can be protective (80, 538). Spray dried serum-containing anti-F4 and anti-LT also has been shown to reduce post weaning diarrhea and shedding of ETEC (539). This could be a useful approach in the reduction of within herd spread of ETEC. The recent construction and use of nanobodies to inhibit the attachment of F18- and F4-fimbriated E. coli to pig enterocytes is an interesting approach that needs to be investigated in pigs (540, 541). Various active immunization studies have focused on the relevant F18 and F4 fimbriae to protect piglets against post weaning ETEC infections. Administration of non-attenuated live ETEC strains expressing F18ab or F18ac fimbriae to pigs shortly before or after weaning had some protective effects after challenge with ETEC, but most of the vaccinated pigs suffered mild to severe diarrhea (538, 542). Slower colonization of F18- fimbriated ETEC versus F4-fimbriated ETEC paralleled a slower induction of the humoral immune response (543). The oral administration of enteric-coated F4 or microencapsulated F18 fimbriae to newborn piglets at best marginally reduced intestinal colonization upon challenge after weaning (544, 545). All things considered, an efficient vaccine protecting against post weaning ETEC infections awaits further developments, including the design and evaluation of attenuated live bacteria and/or fimbrial protein vaccines that include adjuvants (546, 547). An impediment to creating better vaccines for weaning diarrhea is that unlike neonatal diarrhea, good and reproducible pig disease models are not available. A newer approach that has been developed is based on the creation of vaccines that incorporate the expected protective epitopes of fimbriae and enterotoxins in a single protein molecule (548–551). The incorporation of epitopes from F4, F18, and heat stable and heat labile enterotoxins in a single protein molecule has been shown to elicit antibodies that can neutralize attachment of ETEC to mucosal surfaces and neutralize the enterotoxins. Oral immunization of weaned pigs with F4 and F18 was shown to be better at priming a mucosal response than intramuscular administration. Induction of a primary immune response occurred only in pigs expressing the corresponding intestinal F4 receptor, suggesting that receptor binding may facilitate antigen uptake (532). Pigs with the F4 receptor were protected against a challenge with F4-fimbriated ETEC. However, parenteral priming with F4 induced suppression of a mucosal F4 recall response upon oral infection with F4+ fimbriated bacteria (552). In contrast, orally administered F4 was able to prime an immune response in both F4-susceptible and -resistant pigs, indicating that F4 given by the mucosal route does not induce oral tolerance (553). Since pigs are weaned when they are around 3–4 weeks of age, active protection against ETEC must be developed during the pig’s first month of life. However, one difficulty in eliciting protection during this timeframe is that most pigs also are receiving antibodies specific for the ETEC antigens in vaccines via colostrum and milk. Since protection is likely to depend upon local mucosal immunity and thus dependent upon oral immunization, any vaccine antigens orally administered to baby pigs must escape luminal antibodies (from the dam) specific to the vaccine antigens. Thus, strategies to elicit an active mucosal immune response in the presence of passive antibodies from the dam need to be developed. This could possibly include both oral and parenteral vaccinations. A recent review discusses immunization strategies and problems for ETEC-mediated PWD in pigs (554). An alternate approach that could be considered is the search for new, highly antigenic molecules that are not related to fimbriae but that are required for ETEC colonization and persistence in the small intestines of pigs. Because these antigens would not be part of the current generation of fimbrial specific vaccines for neonates, they could be administered directly to suckling pigs and would not be eliminated due to the maternal antibodies in milk and colostrum being received by piglets. The discovery of such antigens would need to be broadly expressed by ETEC causing PWD. While these antigens have not yet been identified, they could be discovered using extensive genomic data as a reverse genetics discovery tool. Phage display also has been used to screen for new potential antigens for vaccines (555). Inhibitors of colonization Identification and characterization of the binding moieties of ETEC fimbrial adhesins should be useful for the design of new prophylactic or therapeutic strategies. Studies describing potential receptor or adhesin analogues that interfere with fimbria-mediated colonization have been described (556–560). However, more studies including efficient inhibition of the relevant panoply of ETEC fimbriae are needed for this approach to be applied in agriculture. Although fat globule membranes of sow’s and cow’s milk were reported to contain receptors for ETEC fimbriae (556, 561–564), the postulated protective role of these receptors in the intestines of young animals remains unknown. Oral administrations of proteases that degrade intestinal receptors have been investigated with some success (565, 566). The use of probiotics is also being considered for use in the prevention of ETEC induced diarrhea in livestock and humans. Certain probiotics such as lactobacilli can bind to enterocytes without interfering with the attachment of F4-fimbriated ETEC (567). It was suggested that the co-aggregation of certain Lactobacillus isolates with the F4 ETEC decreases ETEC colonization. Probiotics were the most efficient in controlling diarrhea in calves when used in conjunction with fimbria-based vaccines (568). The in vivo relevance of F4-mediated adhesion inhibitors found in certain Lactobacillus culture supernatants remains to be determined (569, 570). Commensal bacteria such as Lactobacillus that express F5 have shown some degree of protectiveness against F5+ ETEC (533). Some commensal bacteria are being considered as directly blocking ETEC colonization of the intestines and might be useful in prevention or treatment of disease. Certain strains of Lactobacillus have shown the most promise to date (567, 569, 571, 572) along with strains of Bifidobacterium, E. coli, Bacillus, Enteroccoccus, Pediococcus and Saccharomyces (573–577). In recent years, the pathogenesis of ETEC has been linked to epithelial inflammatory responses in the form of cytokine expression (578). However, whether these responses are due to known or new ETEC virulence factors or to indirect effects on the microbiome and/or its metabolites remains to be determined. Probiotics, albeit frequently with dietary supplements, modulate detected host factors of inflammation (574, 579–581), possibly through their metabolites (582). More in vivo studies are needed to determine whether probiotics can be sufficiently protective and cost-effective with regard to ETEC diarrhea in farm animals. Genomics for future vaccines Several completed genomes of animal-source ETEC are publicly available. Shepard et al. first described the genomes of F4+ and F18+ porcine ETEC, and performed phylogenetic comparisons of a large collection of porcine ETEC (291). This study demonstrated complex genomes with remarkable plasmid complements encoding a variety of previously identified virulence factors including the F4 and F18 fimbrial operons. It was also evident that porcine ETEC are comprised of strains from multiple phylogenetic lineages that have acquired these plasmids, including strains within the E. coli phylogenetic groups A, B1, and D. However, it was also apparent from this study that the number of lineages containing porcine ETEC is somewhat limited compared to human ETEC, suggesting that a specific chromosomal background is required to harbor porcine ETEC plasmids enabling virulence in the host. Specific lineages may also have enhanced virulence potential. For example, the genome of an O157 porcine ETEC strain involved in an outbreak in pigs was analyzed, and it was LT+, STa+, and STb+, yet lacked common porcine ETEC fimbriae such as F4, F5, or F18 (510). The phylogenetic background of this strain was distinct from classical O157:H7 human clinical isolates, and was actually most similar to O78 strains of avian pathogenic E. coli (583). This again supports the concept that combinations of ETEC virulence factors in the appropriate phylogenetic background are required for enhanced virulence. Despite the apparent diversity of porcine ETEC, scanning of porcine ETEC genomes for antigenic candidates has revealed several candidates that are differentially present in porcine ETEC compared to porcine commensal E. coli found across porcine ETEC lineages and predicted to be surface exposed and accessible to the host (509). Therefore, reverse vaccinology exploiting available animal-source ETEC genomes could be an effective approach towards the development of subunit vaccines. The laboratories of the authors were supported by a Discovery Grant from the National Sciences and Engineering Research Council of Canada (139070) and Fonds de Recherche Nature et Technologies (Québec) to JDD, and grants from the USDA (2013–67015–21285) and NIH (AI098041) to DMS. The authors thank Mrs Jacinthe Lachance and Ms Deborah Argento for the artwork. Fig. 1 Genetic organization of animal ETEC fimbrial gene clusters Genes encoding similar products or products with similar functions were labeled with the same pattern; genes for the major fimbrial subunit (yellow), minor fimbrial subunits (blue), minor adhesive subunit (orange), chaperones (green), usher (red), regulators (purple) and mobile or conjugation elements (white). Fig. 2 Fimbrial biogenesis models The fimbriae consist all of the polymeric assembly of a major subunit (yellow) and of one or more minor subunits (blue), one of them being a tip adhesin (orange) for some fimbriae. For the K88 and K99 fimbriae, the major subunit is the adhesin. Usher proteins (red) locate in the outer membrane and channel the fimbrial subunits to the bacterial surface. All the fimbrial export systems use one periplasmic chaperone (green) for all the subunits, with the exception of the F6 fimbriae that have three chaperones, two being dedicated to two different fimbrial subunits. All the fimbrial proteins cross the inner membrane by using the general secretion (Sec) pathway (black), with the exception of fimbriae specific regulators that remain in the cytoplasm (not shown). Fig. 3 Mechanism of action of ETEC toxins on intestinal epithelial cells Signaling leading to water and electrolytes loss through activation of ion channels and loosening of tight junctions by the various toxins is described. CFTR: cystic fibrosis transmembrane regulator; AC; adenylate cyclase; ARF: ADP-ribosylation factor; PKA: protein kinase C; PKC: protein kinase C; GM1: ganglioside GM1; GC-C: guanylate cycles C; SFT: sulfatide; ER: endoplasmic reticulum; Gsα: α component of an heterotrimeric G protein; NHE3; Na+/H+-exchanger 3; PDE3: phosphodiesterase 3; cGMPKII: cGMP-dependent protein kinase II; cAMPKII: calmodulin-dependent protein kinase II; CaCC: calcium-activated chloride channel; P: phosphorylation. Table 1 Fimbriae of animal ETEC Fimbriae Enterotoxi ns1 Host Associated O-serotypes References F4ab, ac, ad (K88ab, ac, ad)2,3 LT, STa, STb neonatal and weaned piglets 8, 45, 138, 141, 147, 149, 157 (77, 584–586) F41 STa calves, lambs and goat kids 9, 20, 64, 101 (72, 586, 587) F5 (K99) STa, STb calves, lambs and goat kids, piglets 8, 9, 20, 64, 101 (77, 584, 586, 587) F6 (987P) STa, STb neonatal piglets 8, 9, 20, 46, 101, 138, 141, 147, 149 (77, 586–588) F18ac (2134P, 8813)2 LT, STa, STb weaned piglets 8, 25, 45, 108, 138, 141, 147, 149, 157 (203, 589) F17a (F[Y], Att25) STa, LT-IIa calves 19, 101 (590–592) 1 By expressing additional toxins (e.g. a Shiga-like toxin or a cytotoxic necrotizing factor) ETEC strains can share the pathogenic properties of non-ETEC strains 2 The F4 and F18 fimbriae each have their antigenic variants (the designation “a” describing the common antigenic determinants, and the second letter, the variant-specific determinants) (133, 203, 593). Shiga-like toxin-producing E. coli responsible for edema disease in weaned pigs express usually F18ab (F107) or F4 fimbriae 3 Fimbriae were classified and renamed according to their antigenic properties (594) Table 2 Numbers of F4-, F5- and F6-fimbriated ETEC isolated from piglets in various studies since 1979. Whenever possible (most studies), data including only neonatal piglets are shown. F4 F5 F6 Country Year1 References 52 7 6 Netherlands 1979 (595) 54 9 55 USA 1980 (77) 13 1 5 Hungary 1982 (596) 48 13 30 USA 1986 (586) 27 44 25 USA 1986 (597) 50 9 16 Norway 1986 (598) 20 4 13 USA 1987 (599) 71 19 9 Sweden 1988 (600) 02 32 23 Canada 1988 (587) 20 16 5 Canada 1989 (91) 31 6 25 Japan 1990 (601) 55 6 16 England 1990 (90) 29 96 484 Indonesia 1991 (602) 790 23 157 Poland 1992 (603) 766 54 47 England 1993 (604) 10 5 5 USA 1994 (605) 59 8 16 Denmark 1994 (606) 0 6 13 Spain 1997 (607) 13 8 13 Spain 1997 (608) 18 9 26 Korea 1999 (609) 280 2 6 USA 1999 (18) 64 3 11 Japan 2001 (610) 13 7 1 Canada 2003 (497) 14 1 1 Australia 2005 (611) 95 33 0 Vietnam 2006 (612) 40 6 1 Canada 2008 (441) 8 31 4 Korea 2008 (613) 19 15 1 Zimbabwe 2009 (30) 1 Year of publication 2 Only non-classical serotypes were studied Table 3 The structural proteins of animal ETEC Fimbriae Fimbrial diameter Major subunit Minor subunits References F4 2–4 nm FaeG FaeC,FaeF,FaeH,FaeI,(FaeJ) (46, 58, 59, 113, 122, 246) F41 3.2 nm Like F4 (no genes or protein names) (181, 182, 271) F5 ~ 3 nm* FanC FanF, FanG, FanH (112, 251–254, 614, 615) F6 7 nm FasA FasF,FasG (106, 123, 195, 616, 617) F18 6.7 nm FedA FedE,FedF (115, 199, 258) F17 3–4 nm F17-A F17-G (215, 590, 618) * Reported wider values were likely due to strand bundling Table 4 The fimbrial adhesins of ETEC and their receptors Fimbriae Adhesins Intestinal receptor molecules References F4ab FaeG(ab) b: Transferrin N-glycan (74 kDa)1 (159, 619) Galactosylceramide, sulfatide, sulf-lactosylceramide, globotriaosylceramide (162) bc: IMPTGP (210–240 kDa)1,3 (620, 621) bcd: Glycoproteins (45–70 kDa)2 (137) Aminopeptidase N (160, 161) F4ac FaeG(ac) bc: IMPTGP (210–240 kDa)1 (620, 621) Galactosylceramide (162) bcd: Glycoproteins (45–70 kDa)2 (137) Aminopeptidase N (160, 161) F4ad FaeG(ad) d: Neutral glycosphyngolipids1 (622) bcd: Glycoproteins (45–70 kDa)2 (137) Aminopeptidase N (160, 161) F41 Major subunit Unknown (erythrocytes: glycophorin AM) (120, 183) F5 FanC N-glycolylsialoparagloboside (139, 140, 189, 190) N-glycolyl-GM3 F6 FasG Sulfatide (142, 143, 165, 195) Proteins (32–35 kDa) FasA Ceramide monohexoside (hydroxylated galactosyl-cerebroside) F18ac FedF Unknown [F18ab: alpha(1,2)fucosyl-containing glycoconjugates] (205, 206) F17a F17-G Mucins, glycoproteins (170–200 kDa) (146) 1 As suggested by Billey et al. (155) 2 As suggested by van den Broeck et al. (156) 3 Intestinal mucin-type sialoglycoproteins Table 5 Groups of F4/F4 receptor phenotypes, as originally classified (I-IV or A-E), with an updated nomenclature that distinguishes receptors with fully (RFA) or partially (RPA) adhesive phenotypes (155, 158, 623–626). A (IV) A1: F4abRFA/F4acR+/F4adRFA A2: F4abRFA/F4acR+/F4adRPA B (III) B: F4abRFA/F4acR+/F4adR− C C1: F4abRPA/F4acR−/F4adRFA C2: F4abRPA/F4acR−/F4adRPA D (II) D1: F4abR−/F4acR−/F4adRFA D2: F4abR−/F4acR−/F4adRPA E (I) E: F4abR−/F4acR−/F4adR− Table 6 ETEC implicated in diarrheal diseases of animals Animal species Type of diarrhea Virulotypes Pig Neonatal STa:F41 STa:F6 STa:F5:F41 LT:STb:EAST1:F4 LT:STb:STa:EAST1:F4 STb:EAST1:AIDA Post weaning LT:STb:EAST1:F4 LT:STb:STa:EAST1:F4 STa:STb STa:STb:F18 STa:F18 Cattle Neonatal STa:F5:F41 STa:F41 Sheep Neonatal STa:F5:F41 STa:F41 Dog Neonatal STa:X* STb * X: Unknown fimbriae Adapted from Gyles and Fairbrother (20) Table 7 Receptors for ETEC toxins Toxin Subtypes Receptor(s) STa STaH Guanylate cyclase C STaP Guanylate cyclase C EAST1 Guanylate cyclase C STb Sulfatide LTI LTIh * GM1, GD1b, GM2 asialo GM1, galactoproteins, galactose-containing glycolipids LTIp LTII LTIIa * GD1b, GD1a, GT1b, GQ1b, GD2 LTIIb * GD1a, GT1b, GM3 LTIIc * GM1, GM2, GM3, GD1a * In order of decreasing binding strength Table 8 Structural characteristics of toxins produced by ETEC Toxin # a.a. M.W. (Da) Sequence/Arrangement Structure STa Subtype STaH 19 2,000 N S S N Y C C E L C C N P A C T G C Y Subtype STaP 18 2,000 N T F Y C C E L C C N P A C A G C Y EAST1 38 4,100 ? 17-2 strain MPSTQYIRRPASSYASCIWC TACASCHGRTTKPSLAT 0–42 strain MPSTQYIRRPASSYASCIWC TACASCHGRTTKPSLAT STb 48 5,200 STQSNKKDLCEHYRQIAKESCKKGFLGVRDGTAGACFG AQIMVAAKGC LTI 85,000 AB 5 B-subunit 103 A-subunit 240 Letters in bold and italics indicate the region involved in binding to the receptor. Letters in red indicate the change observed for EAST1 variants. 1 Nagy B Fekete PZ 2005 Enterotoxigenic Escherichia coli in veterinary medicine Int J Med Microbiol 295 443 454 16238018 2 Kopic S Geibel JP 2010 Toxin mediated diarrhea in the 21 century: the pathophysiology of intestinal ion transport in the course of ETEC, V. cholerae and rotavirus infection Toxins (Basel) 2 2132 2157 22069677 3 Dubreuil JD 2012 The whole Shebang: the gastrointestinal tract, Escherichia coli enterotoxins and secretion Curr Issues Mol Biol 14 71 82 22368232 4 van Beers-Schreurs HM Vellenga L Wensing T Breukink HJ 1992 The pathogenesis of the post-weaning syndrome in weaned piglets: a review Vet Q 14 29 34 1574833 5 Fairbrother JM Nadeau E Gyles CL 2005 Escherichia coli in postweaning diarrhea in pigs: an update on bacterial types, pathogenesis, and prevention strategies Anim Health Res Rev 6 17 39 16164007 6 Tacket CO Reid RH Boedeker EC Losonsky G Nataro JP Bhagat H Edelman R 1994 Enteral immunization and challenge of volunteers given enterotoxigenic E. coli CFA/II encapsulated in biodegradable microspheres Vaccine 12 1270 1274 7856290 7 Berberov EM Zhou Y Francis DH Scott MA Kachman SD Moxley RA 2004 Relative importance of heat-labile enterotoxin in the causation of severe diarrheal disease in the gnotobiotic piglet model by a strain of enterotoxigenic Escherichia coli that produces multiple enterotoxins Infect Immun 72 3914 3924 15213135 8 Liu W Yuan C Meng X Du Y Gao R Tang J Shi D 2014 Frequency of virulence factors in Escherichia coli isolated from suckling pigs with diarrhoea in China Vet J 199 286 289 24378293 9 Duan Q Yao F Zhu G 2012 Major virulence factors of enterotoxigenic Escherichia coli in pigs Ann Microbiol 62 7 14 10 Foster DM Smith GW 2009 Pathophysiology of diarrhea in calves Vet Clin North Am Food Anim Pract 25 13 36, xi 19174281 11 Fleckenstein JM Hardwidge PR Munson GP Rasko DA Sommerfelt H Steinsland H 2010 Molecular mechanisms of enterotoxigenic Escherichia coli infection Microbes Infect 12 89 98 19883790 12 Nagy B Casey TA Moon HW 1990 Phenotype and genotype of Escherichia coli isolated from pigs with postweaning diarrhea in Hungary J Clin Microbiol 28 651 653 1970575 13 Wittig W Fabricius C 1992 Escherichia coli types isolated from porcine E. coli infections in Saxony from 1963 to 1990 Zentralbl Bakteriol 277 389 402 1486239 14 Post KW Bosworth BT Knoth JL 2000 Frequency of virulence factors in Escherichia coli isolated from pigs with postweaning diarrhea and edema disease in North Carolina Swine Health Prod 8 119 120 15 Aarestrup FM Jorsal SE Ahrens P Jensen NE Meyling A 1997 Molecular characterization of Escherichia coli strains isolated from pigs with edema disease J Clin Microbiol 35 20 24 8968875 16 da Silva AS Valadares GF Penatti MP Brito BG da Silva Leite D 2001 Escherichia coli strains from edema disease: O serogroups, and genes for Shiga toxin, enterotoxins, and F18 fimbriae Vet Microbiol 80 227 233 11337138 17 Osek J 1999 Prevalence of virulence factors of Escherichia coli strains isolated from diarrheic and healthy piglets after weaning Vet Microbiol 68 209 217 10510040 18 Moon HW Hoffman LJ Cornick NA Booher SL Bosworth BT 1999 Prevalences of some virulence genes among Escherichia coli isolates from swine presented to a diagnostic laboratory in Iowa J Vet Diagn Invest 11 557 560 12968745 19 Dean-Nystrom EA Burkhardt D Bosworth BT Welter MW 1997 Presence of F18ac (2134P) fimbriae on 4P- Escherichia coli isolates from weaned pigs with diarrhea J Vet Diagn Invest 9 77 79 9087930 20 Gyles CL Fairbrother JM 2010 Escherichia coli 267 308 Gyles CL Prescott JF Songer JG Thoen CO Pathogenesis of Bacterial Infections in Animals Wiley-Blackwell New York, NY 21 Mainil JG Jacquemin E Pohl P Kaeckenbeeck A Benz I 2002 DNA sequences coding for the F18 fimbriae and AIDA adhesin are localised on the same plasmid in Escherichia coli isolates from piglets Vet Microbiol 86 303 311 11955780 22 Ngeleka M Pritchard J Appleyard G Middleton DM Fairbrother JM 2003 Isolation and association of Escherichia coli AIDA-I/STb, rather than EAST1 pathotype, with diarrhea in piglets and antibiotic sensitivity of isolates J Vet Diagn Invest 15 242 252 12735346 23 Drolet R Fairbrother JM Vaillancourt D 1994 Attaching and effacing Escherichia coli in a goat with diarrhea Can Vet J 35 122 123 24 Hammermueller J Kruth S Prescott J Gyles C 1995 Detection of toxin genes in Escherichia coli isolated from normal dogs and dogs with diarrhea Can J Vet Res 59 265 270 8548687 25 USDA.APHIS:VS 2001 Swine 2000. Part I: Reference of swine health and management in the United States, 2000, NAHMS 555 South Howes Fort Collins, Colorado 80521 970 490 8000 E-mail: [email protected] 26 USDA.APHIS:VS 1997 Swine ‘95 Study, Part III: 1990–1995 Changes in the U.S. Pork Industry 555 South Howes Fort Collins, Colorado 80521 970 490 8000 Internet: [email protected] 27 USDA.APHIS:VS 1992 National Swine Survey. Morbidity/Mortality and Health Management of Swine in the United States 555 South Howes Fort Collins, Colorado 80521 970 490 8000 Internet: [email protected] 28 Zhang W Zhao M Ruesch L Omot A Francis D 2007 Prevalence of virulence genes in Escherichia coli strains recently isolated from young pigs with diarrhea in the US Vet Microbiol 123 145 152 17368762 29 Smith MG Jordan D Chapman TA Chin JJ Barton MD Do TN Fahy VA Fairbrother JM Trott DJ 2010 Antimicrobial resistance and virulence gene profiles in multi-drug resistant enterotoxigenic Escherichia coli isolated from pigs with post-weaning diarrhoea Vet Microbiol 145 299 307 20688440 30 Madoroba E Van Driessche E De Greve H Mast J Ncube I Read J Beeckmans S 2009 Prevalence of enterotoxigenic Escherichia coli virulence genes from scouring piglets in Zimbabwe Trop Anim Health Prod 41 1539 1547 19347597 31 Moredo FA Piñeyro PE Márquez GC Sanz M Colello R Etcheverría A Padola NL Quiroga MA Perfumo CJ Galli L Leotta GA 2015 Enterotoxigenic Escherichia coli Subclinical Infection in Pigs: Bacteriological and Genotypic Characterization and Antimicrobial Resistance Profiles Foodborne Pathog Dis 12 704 711 26217917 32 Moon HW Bunn TO 1993 Vaccines for preventing enterotoxigenic Escherichia coli infections in farm animals Vaccine 11 213 200 8094931 33 Fairbrother JM 1999 Neonatal Escherichia coli diarrhea 433 441 Straw BE D’Allaire S Mengeling WL Taylor DJ Diseases of swine 8th Iowa State University Press Ames, Iowa 34 Vu Khac H Holoda E Pilipcinec E Blanco M Blanco JE Mora A Dahbi G López C González EA Blanco J 2006 Serotypes, virulence genes, and PFGE profiles of Escherichia coli isolated from pigs with postweaning diarrhoea in Slovakia BMC Vet Res 2 10 16549022 35 Chan G Farzan A DeLay J McEwen B Prescott JF Friendship RM 2013 A retrospective study on the etiological diagnoses of diarrhea in neonatal piglets in Ontario, Canada, between 2001 and 2010 Can J Vet Res 77 254 260 24124267 36 Okello E Moonens K Erume J De Greve H 2015 Enterotoxigenic Escherichia coli strains are highly prevalent in Ugandan piggeries but disease outbreaks are masked by antibiotic prophylaxis Trop Anim Health Prod 47 117 122 25311441 37 Goswami PS Friendship RM Gyles CL Poppe C Boerlin P 2011 Preliminary investigations of the distribution of Escherichia coli O149 in sows, piglets, and their environment Can J Vet Res 75 57 60 21461196 38 Sarrazin E Fritzsche C Bertschinger HU 2000 Hauptvirulenzfaktoren bei Escherichia coli-Isolaten von über zwei Wochen alten Schweinen mit Odemkrankheit und/oder Colidiarrhöe Schweiz Arch Tierheilkd 142 625 630 11103614 39 Westerman RB Mills KW Phillips RM Fortner GW Greenwood JM 1988 Predominance of the ac variant in K88-positive Escherichia coli isolates from swine J Clin Microbiol 26 149 150 3277990 40 González EA Vázquez F Ignacio Garabal J Blanco J 1995 Isolation of K88 antigen variants (ab, ac, ad) from porcine enterotoxigenic Escherichia coli belonging to different serotypes Microbiol Immunol 39 937 942 8789052 41 Frydendahl K 2002 Prevalence of serogroups and virulence genes in Escherichia coli associated with postweaning diarrhoea and edema disease in pigs and a comparison of diagnostic approaches Vet Microbiol 85 169 182 11844623 42 Abraham S Chin J Brouwers HJ Zhang R Chapman TA 2012 Molecular serogrouping of porcine enterotoxigenic Escherichia coli from Australia J Microbiol Methods 88 73 76 22093999 43 Osek J 2000 Clonal analysis of Escherichia coli strains isolated from pigs with post-weaning diarrhea by pulsed-field gel electrophoresis FEMS Microbiol Lett 186 327 331 10802192 44 Orskov I Orskov F Sojka WJ Leach JM 1961 Simultaneous occurrence of E. coli B and Lantigens in strains from diseased swine. Influence of cultivation temperature. Two new E. coli Kantigens: K 87 and K 88 Acta Pathol Microbiol Scand 53 404 422 14482368 45 Stirm S Orskov I Orskov F 1966 K88, an episome-determined protein antigen of Escherichia coli Nature 209 507 508 4958488 46 Stirm S Orskov F Orskov I Birch-Andersen A 1967 Episome-carried surface antigen K88 of Escherichia coli. 3. Morphology J Bacteriol 93 740 748 5335971 47 Duguid JP Anderson ES 1967 Terminology of bacterial fimbriae, or pili, and their types Nature 215 89 90 6053418 48 Kraus R Ludwig S 1902 Ueber Bakteriohaemagglutinine und Antihaemagglutinine Wien Klin Wochenschr 5 120 121 49 Guyot G 1908 Ueber die bakterielle Haemagglutination (Bakterio-Haemoagglutination) Centralbl f Bakt Abt I Orig XLVII 640 653 50 Duguid JP Smith IW Dempster G Edmunds PN 1955 Non-flagellar filamentous appendages (fimbriae) and haemagglutinating activity in Bacterium coli J Pathol Bacteriol 70 335 348 13295908 51 Smith T Orcutt ML 1925 The bacteriology of the intestinal tract of of young calves with special reference to the early diarrhea (“scours”) J Exp Med 41 89 106 19868978 52 Smith HW Halls S 1968 The production of oedema disease and diarrhoea in weaned pigs by the oral administration of Escherichia coli: factors that influence the course of the experimental disease J Med Microbiol 1 45 59 4918220 53 Smith HW Linggood MA 1971 Observations on the pathogenic properties of the K88, Hly and Ent plasmids of Escherichia coli with particular reference to porcine diarrhoea J Med Microbiol 4 467 485 4944321 54 Falkow S 1988 Molecular Koch’s postulates applied to microbial pathogenicity Rev Infect Dis 10 Suppl 2 S274 S276 3055197 55 Jones GW Rutter JM 1972 Role of the K88 antigen in the pathogenesis of neonatal diarrhea caused by Escherichia coli in piglets Infect Immun 6 918 927 4629393 56 Rutter JM Jones GW 1973 The K88 antigen of Escherichia coli–a model for vaccination with a virulence factor? J Med Microbiol 6 8 9 57 Mooi FR de Graaf FK van Embden JD 1979 Cloning, mapping and expression of the genetic determinant that encodes for the K88ab antigen Nucleic Acids Res 6 849 865 375197 58 Kehoe M Sellwood R Shipley P Dougan G 1981 Genetic analysis of K88-mediated adhesion of enterotoxigenic Escherichia coli Nature 291 122 126 7015144 59 Mooi FR Harms N Bakker D de Graaf FK 1981 Organization and expression of genes involved in the production of the K88ab antigen Infect Immun 32 1155 1163 7019089 60 Hull RA Gill RE Hsu P Minshew BH Falkow S 1981 Construction and expression of recombinant plasmids encoding type 1 or D-mannose-resistant pili from a urinary tract infection Escherichia coli isolate Infect Immun 33 933 938 6116675 61 Normark S Lark D Hull R Norgren M Båga M O’Hanley P Schoolnik G Falkow S 1983 Genetics of digalactoside-binding adhesin from a uropathogenic Escherichia coli strain Infect Immun 41 942 949 6136465 62 Orndorff PE Falkow S 1984 Organization and expression of genes responsible for type 1 piliation in Escherichia coli J Bacteriol 159 736 744 6146599 63 Moon HW Isaacson RE Pohlenz J 1979 Mechanisms of association of enteropathogenic Escherichia coli with intestinal epithelium Am J Clin Nutr 32 119 127 367139 64 Isaacson RE Nagy B Moon HW 1977 Colonization of porcine small intestine by Escherichia coli: colonization and adhesion factors of pig enteropathogens that lack K88 J Infect Dis 135 531 539 323375 65 Nagy B Moon HW Isaacson RE 1977 Colonization of porcine intestine by enterotoxigenic Escherichia coli: selection of piliated forms in vivo, adhesion of piliated forms to epithelial cells in vitro, and incidence of a pilus antigen among porcine enteropathogenic E. coli Infect Immun 16 344 352 326676 66 Isaacson RE Fusco PC Brinton CC Moon HW 1978 In vitro adhesion of Escherichia coli to porcine small intestinal epithelial cells: pili as adhesive factors Infect Immun 21 392 397 357285 67 Isaacson RE Dean EA Morgan RL Moon HW 1980 Immunization of suckling pigs against enterotoxigenic Escherichia coli-induced diarrheal disease by vaccinating dams with purified K99 or 987P pili: antibody production in response to vaccination Infect Immun 29 824 826 6111535 68 Morgan RL Isaacson RE Moon HW Brinton CC To CC 1978 Immunization of suckling pigs against enterotoxigenic Escherichia coli-induced diarrheal disease by vaccinating dams with purified 987 or K99 pili: protection correlates with pilus homology of vaccine and challenge Infect Immun 22 771 777 365769 69 Nagy B Moon HW Isaacson RE To C-C Brinton CC 1978 Immunization of suckling pigs against enteric enterotoxigenic Escherichia coli infection by vaccinating dams with purified pili Infect Immun 21 269 274 361566 70 Casey TA Schneider RA Dean-Nystrom EA 1993 Identification of plasmid and chromosomal copies of 987P pilus genes in enterotoxigenic Escherichia coli 987 Infect Immun 61 2249 2252 8097495 71 Morris JA Wells GAH Scott AC Sojka WJ 1983 Colonisation of the small intestine of lambs by an enterotoxigenic Escherichia coli producing F41 fimbriae Vet Rec 113 471 471 6359662 72 Morris JA Thorns CJ Wells GA Scott AC Sojka WJ 1983 The production of F41 fimbriae by piglet strains of enterotoxigenic Escherichia coli that lack K88, K99 and 987P fimbriae J Gen Microbiol 129 2753 2759 6138394 73 Runnels PL Moseley SL Moon HW 1987 F41 pili as protective antigens of enterotoxigenic Escherichia coli that produce F41, K99, or both pilus antigens Infect Immun 55 555 558 2880807 74 To SC 1984 Prevention of colibacillosis in neonatal swine with a 4-pilus E coli bacterin Mod Vet Pract 65 39 41 6374422 75 To SC 1984 F41 antigen among porcine enterotoxigenic Escherichia coli strains lacking K88, K99, and 987P pili Infect Immun 43 549 554 6420345 76 Nagy B 1980 Vaccination of cows with a K99 extract to protect newborn calves against experimental enterotoxic colibacillosis Infect Immun 27 21 24 6987168 77 Moon HW Kohler EM Schneider RA Whipp SC 1980 Prevalence of pilus antigens, enterotoxin types, and enteropathogenicity among K88-negative enterotoxigenic Escherichia coli from neonatal pigs Infect Immun 27 222 230 6102079 78 Moon HW 1990 Colonization factor antigens of enterotoxigenic Escherichia coli in animals Curr Top Microbiol Immunol 151 147 165 1973369 79 Imberechts H Bertschinger HU Nagy B Deprez P Pohl P 1997 Fimbrial colonisation factors F18ab and F18ac of Escherichia coli isolated from pigs with postweaning diarrhea and edema disease Adv Exp Med Biol 412 175 183 9192010 80 Zúñiga A Yokoyama H Albicker-Rippinger P Eggenberger E Bertschinger HU 1997 Reduced intestinal colonisation with F18-positive enterotoxigenic Escherichia coli in weaned pigs fed chicken egg antibody against the fimbriae FEMS Immunol Med Microbiol 18 153 161 9271166 81 Contrepois MG Girardeau JP 1985 Additive protective effects of colostral antipili antibodies in calves experimentally infected with enterotoxigenic Escherichia coli Infect Immun 50 947 949 2866162 82 Bertin Y Martin C Oswald E Girardeau JP 1996 Rapid and specific detection of F17-related pilin and adhesin genes in diarrheic and septicemic Escherichia coli strains by multiplex PCR J Clin Microbiol 34 2921 2928 8940423 83 Le Bouguénec C Bertin Y 1999 AFA and F17 adhesins produced by pathogenic Escherichia coli strains in domestic animals Vet Res 30 317 342 10367361 84 Dezfulian H Batisson I Fairbrother JM Lau PC Nassar A Szatmari G Harel J 2003 Presence and characterization of extraintestinal pathogenic Escherichia coli virulence genes in F165-positive E. coli strains isolated from diseased calves and pigs J Clin Microbiol 41 1375 1385 12682117 85 Bihannic M Ghanbarpour R Auvray F Cavalié L Châtre P Boury M Brugère H Madec JY Oswald E 2014 Identification and detection of three new F17 fimbrial variants in Escherichia coli strains isolated from cattle Vet Res (Faisalabad) 45 76 86 Harel J Lapointe H Fallara A Lortie LA Bigras-Poulin M Larivière S Fairbrother JM 1991 Detection of genes for fimbrial antigens and enterotoxins associated with Escherichia coli serogroups isolated from pigs with diarrhea J Clin Microbiol 29 745 752 1679765 87 Broes A Fairbrother JM Jacques M Lariviere S 1988 Isolation and Characterization of a New Fimbrial Antigen (Cs1541) from a Porcine Entero-Toxigenic Escherichia-Coli O8 - Kx105 Strain FEMS Microbiol Lett 55 341 347 88 Broes A Fairbrother JM Jacques M Larivière S 1989 Requirement for capsular antigen KX105 and fimbrial antigen CS1541 in the pathogenicity of porcine enterotoxigenic Escherichia coli O8:KX105 strains Can J Vet Res 53 43 47 2563336 89 Fairbrother JM Broes A Jacques M Larivière S 1989 Pathogenicity of Escherichia coli O115:K“V165” strains isolated from pigs with diarrhea Am J Vet Res 50 1029 1036 2672911 90 Woodward MJ Wray C 1990 Nine DNA probes for detection of toxin and adhesin genes in Escherichia coli isolated from diarrhoeal disease in animals Vet Microbiol 25 55 65 1978952 91 Contrepois M Fairbrother JM Kaura YK Girardeau JP 1989 Prevalence of CS31A and F165 surface antigens in Escherichia coli isolates from animals in France, Canada and India FEMS Microbiol Lett 50 319 323 2668111 92 Bertin Y Martin C Girardeau JP Pohl P Contrepois M 1998 Association of genes encoding P fimbriae, CS31A antigen and EAST 1 toxin among CNF1-producing Escherichia coli strains from cattle with septicemia and diarrhea FEMS Microbiol Lett 162 235 239 9627958 93 Valat C Forest K Auvray F Métayer V Méheut T Polizzi C Gay E Haenni M Oswald E Madec JY 2014 Assessment of Adhesins as an Indicator of Pathovar-Associated Virulence Factors in Bovine Escherichia coli Appl Environ Microbiol 80 7230 7234 25217019 94 Girardeau JP Der Vartanian M Ollier JL Contrepois M 1988 CS31A, a new K88-related fimbrial antigen on bovine enterotoxigenic and septicemic Escherichia coli strains Infect Immun 56 2180 2188 2899553 95 Mercado EC Rodríguez SM D’Antuono AL Cipolla AL Elizondo AM Rossetti CA Malena R Méndez MA 2003 Occurrence and characteristics of CS31A antigen-producing Escherichia coli in calves with diarrhoea and septicaemia in Argentina J Vet Med B Infect Dis Vet Public Health 50 8 13 12710494 96 Mainil JG Jacquemin E Pohl P Fairbrother JM Ansuini A Le Bouguénec C Ball HJ De Rycke J Oswald E 1999 Comparison of necrotoxigenic Escherichia coli isolates from farm animals and from humans Vet Microbiol 70 123 135 10591503 97 Nuccio SP Bäumler AJ 2007 Evolution of the chaperone/usher assembly pathway: fimbrial classification goes Greek Microbiol Mol Biol Rev 71 551 575 18063717 98 Franklin A Söderlind O Möllby R 1981 Plasmids coding for enterotoxins, K88 antigen and colicins in porcine Escherichia coli strains of O-group 149 Med Microbiol Immunol (Berl) 170 63 72 7033751 99 Isaacson RE Start GL 1992 Analysis of K99 plasmids from enterotoxigenic Escherichia coli FEMS Microbiol Lett 69 141 146 1537542 100 Bertin A 1992 Plasmid content and localisation of the STaI (STaP) gene in enterotoxigenic Escherichia coli with a non-radioactive polynucleotide gene probe J Med Microbiol 37 141 147 1629900 101 Mainil JG Bex F Dreze P Kaeckenbeeck A Couturier M 1992 Replicon typing of virulence plasmids of enterotoxigenic Escherichia coli isolates from cattle Infect Immun 60 3376 3380 1639505 102 Mainil JG Daube G Jacquemin E Pohl P Kaeckenbeeck A 1998 Virulence plasmids of enterotoxigenic Escherichia coli isolates from piglets Vet Microbiol 62 291 301 9791875 103 Fekete PZ Gerardin J Jacquemin E Mainil JG Nagy B 2002 Replicon typing of F18 fimbriae encoding plasmids of enterotoxigenic and verotoxigenic Escherichia coli strains from porcine postweaning diarrhoea and oedema disease Vet Microbiol 85 275 284 11852194 104 Schifferli DM Beachey EH Taylor RK 1990 The 987P fimbrial gene cluster of enterotoxigenic Escherichia coli is plasmid encoded Infect Immun 58 149 156 1967167 105 Klaasen P Woodward MJ van Zijderveld FG de Graaf FK 1990 The 987P gene cluster in enterotoxigenic Escherichia coli contains an STpa transposon that activates 987P expression Infect Immun 58 801 807 1968436 106 Schifferli DM Beachey EH Taylor RK 1991 987P fimbrial gene identification and protein characterization by T7 RNA polymerase-induced transcription and TnphoA mutagenesis Mol Microbiol 5 61 70 1673018 107 Boerlin P Travis R Gyles CL Reid-Smith R Janecko N Lim H Nicholson V McEwen SA Friendship R Archambault M 2005 Antimicrob ial resistance and virulence genes of Escherichia coli isolates from swine in Ontario Appl Environ Microbiol 71 6753 6761 16269706 108 Fekete PZ Brzuszkiewicz E Blum-Oehler G Olasz F Szabó M Gottschalk G Hacker J Nagy B 2012 DNA sequence analysis of the composite plasmid pTC conferring virulence and antimicrobial resistance for porcine enterotoxigenic Escherichia coli Int J Med Microbiol 302 4 9 22000740 109 Johnson TJ Shepard SM Rivet B Danzeisen JL Carattoli A 2011 Comparative genomics and phylogeny of the IncI1 plasmids: a common plasmid type among porcine enterotoxigenic Escherichia coli Plasmid 66 144 151 21843549 110 Zav’yalov V Zavialov A Zav’yalova G Korpela T 2010 Adhesive organelles of Gram-negative pathogens assembled with the classical chaperone/usher machinery: structure and function from a clinical standpoint FEMS Microbiol Rev 34 317 378 20070375 111 Werneburg GT Henderson NS Portnoy EB Sarowar S Hultgren SJ Li H Thanassi DG 2015 The pilus usher controls protein interactions via domain masking and is functional as an oligomer Nat Struct Mol Biol 22 540 546 26052892 112 Duchet-Suchaux M Bertin A Dubray G 1988 Morphological description of surface structures on strain B41 of bovine enterotoxigenic Escherichia coli bearing both K99 and F41 antigens J Gen Microbiol 134 983 995 2903218 113 Simons BL Mol O van Breemen JF Oudega B 1994 Morphological appearances of K88ab fimbriae and optical diffraction analysis of K88 paracrystalline structures FEMS Microbiol Lett 118 83 88 7912213 114 Bertin Y Girardeau JP Darfeuille-Michaud A Contrepois M 1996 Characterization of 20K fimbria, a new adhesin of septicemic and diarrhea-associated Escherichia coli strains, that belongs to a family of adhesins with N-acetyl-D-glucosamine recognition Infect Immun 64 332 342 8557360 115 Hahn E Wild P Schraner EM Bertschinger HU Häner M Müller SA Aebi U 2000 Structural analysis of F18 fimbriae expressed by porcine toxigenic Escherichia coli J Struct Biol 132 241 250 11243892 116 Brinton CC Jr 1965 The structure, function, synthesis and genetic control of bacterial pili and a molecular model for DNA and RNA transport in gram negative bacteria Trans N Y Acad Sci 27 8 Series II 1003 1054 5318403 117 Abraham SN Land M Ponniah S Endres R Hasty DL Babu JP 1992 Glycerol-induced unraveling of the tight helical conformation of Escherichia coli type 1 fimbriae J Bacteriol 174 5145 5148 1352770 118 Mortezaei N Epler CR Shao PP Shirdel M Singh B McVeigh A Uhlin BE Savarino SJ Andersson M Bullitt E 2015 Structure and function of enterotoxigenic Escherichia coli fimbriae from differing assembly pathways Mol Microbiol 95 116 126 25355550 119 Smyth CJ Marron MB Twohig JM Smith SG 1996 Fimbrial adhesins: similarities and variations in structure and biogenesis FEMS Immunol Med Microbiol 16 127 139 8988393 120 Brooks DE Cavanagh J Jayroe D Janzen J Snoek R Trust TJ 1989 Involvement of the MN blood group antigen in shear-enhanced hemagglutination induced by the Escherichia coli F41 adhesin Infect Immun 57 377 383 2563256 121 Le Trong I Aprikian P Kidd BA Forero-Shelton M Tchesnokova V Rajagopal P Rodriguez V Interlandi G Klevit R Vogel V Stenkamp RE Sokurenko EV Thomas WE 2010 Structural basis for mechanical force regulation of the adhesin FimH via finger trap-like beta sheet twisting Cell 141 645 655 20478255 122 Bakker D Willemsen PT Willems RH Huisman TT Mooi FR Oudega B Stegehuis F de Graaf FK 1992 Identification of minor fimbrial subunits involved in biosynthesis of K88 fimbriae J Bacteriol 174 6350 6358 1400188 123 Cao J Khan AS Bayer ME Schifferli DM 1995 Ordered translocation of 987P fimbrial subunits through the outer membrane of Escherichia coli J Bacteriol 177 3704 3713 7601834 124 Ponniah S Endres RO Hasty DL Abraham SN 1991 Fragmentation of Escherichia coli type 1 fimbriae exposes cryptic D-mannose-binding sites J Bacteriol 173 4195 4202 1676398 125 Li Q Ng TW Dodson KW So SS Bayle KM Pinkner JS Scarlata S Hultgren SJ Thanassi DG 2010 T he differential affinity of the usher for chaperone-subunit complexes is required for assembly of complete pili Mol Microbiol 76 159 172 20199591 126 Morrissey B Leney AC Toste Rego A Phan G Allen WJ Verger D Waksman G Ashcroft AE Radford SE 2012 The role of chaperone-subunit usher domain interactions in the mechanism of bacterial pilus biogenesis revealed by ESI-MS Mol Cell Proteomics 11 M111 015289 127 Luo Y Van Nguyen U de la Fe Rodriguez PY Devriendt B Cox E 2015 F4+ ETEC infection and oral immunization with F4 fimbriae elicits an IL-17-dominated immune response Vet Res (Faisalabad) 46 121 128 Gao Y Han F Huang X Rong Y Yi H Wang Y 2013 Changes in gut microbial populations, intestinal morphology, expression of tight junction proteins, and cytokine production between two pig breeds after challenge with Escherichia coli K88: a comparative study J Anim Sci 91 5614 5625 24126267 129 McLamb BL Gibson AJ Overman EL Stahl C Moeser AJ 2013 Early weaning stress in pigs impairs innate mucosal immune responses to enterotoxigenic E. coli challenge and exacerbates intestinal injury and clinical disease PLoS One 8 e59838 23637741 130 Edwards RA Puente JL 1998 Fimbrial expression in enteric bacteria: a critical step in intestinal pathogenesis Trends Microbiol 6 282 287 9717217 131 Edwards RA Schifferli DM 1997 Differential regulation of fasA and fasH expression of Escherichia coli 987P fimbriae by environmental cues Mol Microbiol 25 797 809 9379907 132 Burrows MR Sellwood R Gibbons RA 1976 Haemagglutinating and adhesive properties associated with the K99 antigen of bovine strains of Escherichia coli J Gen Microbiol 96 269 275 792386 133 Guinée PA Jansen WH 1979 Behavior of Escherichia coli K antigens K88ab, K88ac, and K88ad in immunoelectrophoresis, double diffusion, and hemagglutination Infect Immun 23 700 705 110679 134 Bijlsma IG Frik JF 1987 Haemagglutination patterns of the different variants of Escherichia coli K88 antigen with porcine, bovine, guinea pig, chicken, ovine and equine erythrocytes Res Vet Sci 43 122 123 3306851 135 Ike K Nakazawa M Tsuchimoto M Ide S Kashiwazaki M 1987 Hemagglutination by pilus antigen 987P of enterotoxigenic Escherichia coli Microbiol Immunol 31 1255 1258 2897616 136 Conway PL Welin A Cohen PS 1990 Presence of K88-specific receptors in porcine ileal mucus is age dependent Infect Immun 58 3178 3182 1976112 137 Willemsen PT de Graaf FK 1992 Age and serotype dependent binding of K88 fimbriae to porcine intestinal receptors Microb Pathog 12 367 375 1354324 138 Runnels PL Moon HW Schneider RA 1980 Development of resistance with host age to adhesion of K99+ Escherichia coli to isolated intestinal epithelial cells Infect Immun 28 298 300 6103878 139 Teneberg S Willemsen P de Graaf FK Karlsson KA 1990 Receptor-active glycolipids of epithelial cells of the small intestine of young and adult pigs in relation to susceptibility to infection with Escherichia coli K99 FEBS Lett 263 10 14 2185031 140 Yuyama Y Yoshimatsu K Ono E Saito M Naiki M 1993 Postnatal change of pig intestinal ganglioside bound by Escherichia coli with K99 fimbriae J Biochem 113 488 492 8514738 141 Dean EA 1990 Comparison of receptors for 987P pili of enterotoxigenic Escherichia coli in the small intestines of neonatal and older pig Infect Immun 58 4030 4035 1979318 142 Dean-Nystrom EA Samuel JE 1994 Age-related resistance to 987P fimbria-mediated colonization correlates with specific glycolipid receptors in intestinal mucus in swine Infect Immun 62 4789 4794 7927756 143 Dean EA Whipp SC Moon HW 1989 Age-specific colonization of porcine intestinal epithelium by 987P-piliated enterotoxigenic Escherichia coli Infect Immun 57 82 87 2562837 144 Madar Johansson M Coddens A Benktander J Cox E Teneberg S 2014 Porcine intestinal glycosphingolipids recognized by F6-fimbriated enterotoxigenic Escherichia coli Microb Pathog 76 51 60 25241919 145 Nagy B Casey TA Whipp SC Moon HW 1992 Susceptibility of porcine intestine to pilus-mediated adhesion by some isolates of piliated enterotoxigenic Escherichia coli increases with age Infect Immun 60 1285 1294 1347758 146 Mouricout M Milhavet M Durié C Grange P 1995 Characterization of glycoprotein glycan receptors for Escherichia coli F17 fimbrial lectin Microb Pathog 18 297 306 7476095 147 Bakker D Willemsen PT Simons LH van Zijderveld FG de Graaf FK 1992 Characterization of the antigenic and adhesive properties of FaeG, the major subunit of K88 fimbriae Mol Microbiol 6 247 255 1372075 148 Sun R Anderson TJ Erickson AK Nelson EA Francis DH 2000 Inhibition of adhesion of Escherichia coli k88ac fimbria to its receptor, intestinal mucin-type glycoproteins, by a monoclonal antibody directed against a variable domain of the fimbria Infect Immun 68 3509 3515 10816505 149 Jacobs AA Venema J Leeven R van Pelt-Heerschap H de Graaf FK 1987 Inhibition of adhesive activity of K88 fibrillae by peptides derived from the K88 adhesin J Bacteriol 169 735 741 2879831 150 Jacobs AA Roosendaal B van Breemen JF de Graaf FK 1987 Role of phenylalanine 150 in the receptor-binding domain of the K88 fibrillar subunit J Bacteriol 169 4907 4911 3312162 151 Van Molle I Joensuu JJ Buts L Panjikar S Kotiaho M Bouckaert J Wyns L Niklander-Teeri V De Greve H 2007 Chloroplasts assemble the major subunit FaeG of Escherichia coli F4 (K88) fimbriae to strand-swapped dimers J Mol Biol 368 791 799 17368480 152 Joensuu JJ Kotiaho M Riipi T Snoeck V Palva ET Teeri TH Lång H Cox E Goddeeris BM Niklander-Teeri V 2004 Fimbrial subunit protein FaeG expressed in transgenic tobacco inhibits the binding of F4ac enterotoxigenic Escherichia coli to porcine enterocytes Transgenic Res 13 295 298 15359606 153 Zhang W Fang Y Francis DH 2009 Characterization of the binding specificity of K88ac and K88ad fimbriae of enterotoxigenic Escherichia coli by constructing K88ac/K88ad chimeric FaeG major subunits Infect Immun 77 699 706 19015246 154 Jin LZ Zhao X 2000 Intestinal receptors for adhesive fimbriae of enterotoxigenic Escherichia coli (ETEC) K88 in swine–a review Appl Microbiol Biotechnol 54 311 318 11030565 155 Billey LO Erickson AK Francis DH 1998 Multiple receptors on porcine intestinal epithelial cells for the three variants of Escherichia coli K88 fimbrial adhesin Vet Microbiol 59 203 212 9549860 156 Van den Broeck W Cox E Oudega B Goddeeris BM 2000 The F4 fimbrial antigen of Escherichia coli and its receptors Vet Microbiol 71 223 244 10703706 157 Yan X Huang X Ren J Zou Z Yang S Ouyang J Zeng W Yang B Xiao S Huang L 2009 Distribution of Escherichia coli F4 adhesion phenotypes in pigs of 15 Chinese and Western breeds and a White DurocxErhualian intercross J Med Microbiol 58 1112 1117 19574416 158 Rampoldi A Bertschinger HU Bürgi E Dolf G Sidler X Bratus A Vögeli P Neuenschwander S 2014 I nheritance of porcine receptors for enterotoxigenic Escherichia coli with fimbriae F4ad and their relation to other F4 receptors Animal 8 859 866 24725922 159 Grange PA Mouricout MA Levery SB Francis DH Erickson AK 2002 Evaluation of receptor binding specificity of Escherichia coli K88 (F4) fimbrial adhesin variants using porcine serum transferrin and glycosphingolipids as model receptors Infect Immun 70 2336 2343 11953368 160 Melkebeek V Rasschaert K Bellot P Tilleman K Favoreel H Deforce D De Geest BG Goddeeris BM Cox E 2012 Targeting aminopeptidase N, a newly identified receptor for F4ac fimbriae, enhances the intestinal mucosal immune response Mucosal Immunol 5 635 645 22669578 161 Goetstouwers T Van Poucke M Nguyen VU Melkebeek V Coddens A Deforce D Cox E Peelman LJ 2014 F4-related mutation and expression analysis of the aminopeptidase N gene in pigs J Anim Sci 92 1866 1873 24663207 162 Coddens A Valis E Benktander J Ångström J Breimer ME Cox E Teneberg S 2011 Erythrocyte and porcine intestinal glycosphingolipids recognized by F4 fimbriae of enterotoxigenic Escherichia coli PLoS One 6 e23309 21949679 163 Moonens K Van den Broeck I De Kerpel M Deboeck F Raymaekers H Remaut H De Greve H 2015 Structural and functional insight into the carbohydrate receptor binding of F4 fimbriae-producing enterotoxigenic Escherichia coli J Biol Chem 290 8409 8419 25631050 164 Payne D O’Reilly M Williamson D 1993 The K88 fimbrial adhesin of enterotoxigenic Escherichia coli binds to beta 1-linked galactosyl residues in glycosphingolipids Infect Immun 61 3673 3677 8103039 165 Khan AS Johnston NC Goldfine H Schifferli DM 1996 Porcine 987P glycolipid receptors on intestinal brush borders and their cognate bacterial ligands Infect Immun 64 3688 3693 8751918 166 Karlsson K-A 1989 Animal glycosphingolipids as membrane attachment sites for bacteria Annu Rev Biochem 58 309 350 2673013 167 Python P Jörg H Neuenschwander S Hagger C Stricker C Bürgi E Bertschinger HU Stranzinger G Vögeli P 2002 Fine-mapping of the intestinal receptor locus for enterotoxigenic Escherichia coli F4ac on porcine chromosome 13 Anim Genet 33 441 447 12464019 168 Jørgensen CB Cirera S Anderson SI Archibald AL Raudsepp T Chowdhary B Edfors-Lilja I Andersson L Fredholm M 2003 Linkage and comparative mapping of the locus controlling susceptibility towards E. COLI F4ab/ac diarrhoea in pigs Cytogenet Genome Res 102 157 162 14970696 169 Peng QL Ren J Yan XM Huang X Tang H Wang YZ Zhang B Huang LS 2007 The g.243A>G mutation in intron 17 of MUC4 is significantly associated with susceptibility/resistance to ETEC F4ab/ac infection in pigs Anim Genet 38 397 400 17559554 170 Joller D Jørgensen CB Bertschinger HU Python P Edfors I Cirera S Archibald AL Bürgi E Karlskov-Mortensen P Andersson L Fredholm M Vögeli P 2009 Refined localization of the Escherichia coli F4ab/F4ac receptor locus on pig chromosome 13 Anim Genet 40 749 752 19392815 171 Jacobsen M Kracht SS Esteso G Cirera S Edfors I Archibald AL Bendixen C Andersson L Fredholm M Jørgensen CB 2010 Refined candidate region specified by haplotype sharing for Escherichia coli F4ab/F4ac susceptibility alleles in pigs Anim Genet 41 21 25 19799599 172 Nguyen VU Goetstouwers T Coddens A Van Poucke M Peelman L Deforce D Melkebeek V Cox E 2013 Differentiation of F4 receptor profiles in pigs based on their mucin 4 polymorphism, responsiveness to oral F4 immunization and in vitro binding of F4 to villi Vet Immunol Immunopathol 152 93 100 23084626 173 Zhang B Ren J Yan X Huang X Ji H Peng Q Zhang Z Huang L 2008 Investigation of the porcine MUC13 gene: isolation, expression, polymorphisms and strong association with susceptibility to enterotoxigenic Escherichia coli F4ab/ac Anim Genet 39 258 266 18454803 174 Rampoldi A Jacobsen MJ Bertschinger HU Joller D Bürgi E Vögeli P Andersson L Archibald AL Fredholm M Jørgensen CB Neuenschwander S 2011 The receptor locus for Escherichia coli F4ab/F4ac in the pig maps distal to the MUC4-LMLN region Mamm Genome 22 122 129 21136063 175 Ren J Yan X Ai H Zhang Z Huang X Ouyang J Yang M Yang H Han P Zeng W Chen Y Guo Y Xiao S Ding N Huang L 2012 Susceptibility towards enterotoxigenic Escherichia coli F4ac diarrhea is governed by the MUC13 gene in pigs PLoS One 7 e44573 22984528 176 Ji H Ren J Yan X Huang X Zhang B Zhang Z Huang L 2011 The porcine MUC20 gene: molecular characterization and its association with susceptibility to enterotoxigenic Escherichia coli F4ab/ac Mol Biol Rep 38 1593 1601 20890662 177 Schroyen M Stinckens A Verhelst R Geens M Cox E Niewold T Buys N 2012 Susceptibility of piglets to enterotoxigenic Escherichia coli is not related to the expression of MUC13 and MUC20 Anim Genet 43 324 327 22486505 178 Goetstouwers T Van Poucke M Coppieters W Nguyen VU Melkebeek V Coddens A Van Steendam K Deforce D Cox E Peelman LJ 2014 Refined candidate region for F4ab/ac enterotoxigenic Escherichia coli susceptibility situated proximal to MUC13 in pigs PLoS One 9 e105013 25137053 179 Goetstouwers T Van Poucke M Coddens A Nguyen VU Melkebeek V Deforce D Cox E Peelman LJ 2014 Variation in 12 porcine genes involved in the carbohydrate moiety assembly of glycosphingolipids does not account for differential binding of F4 Escherichia coli and their fimbriae BMC Genet 15 103 25277275 180 Yan XM Ren J Huang X Zhang ZY Ouyang J Zeng WH Zou ZZ Yang SJ Yang B Huang LS 2009 Comparison of production traits between pigs with and without the Escherichia coli F4 receptors in a White Duroc × Erhualian intercross F2 population J Anim Sci 87 334 339 18765849 181 Moseley SL Dougan G Schneider RA Moon HW 1986 Cloning of chromosomal DNA encoding the F41 adhesin of enterotoxigenic Escherichia coli and genetic homology between adhesins F41 and K88 J Bacteriol 167 799 804 2875061 182 Anderson DG Moseley SL 1988 Escherichia coli F41 adhesin: genetic organization, nucleotide sequence, and homology with the K88 determinant J Bacteriol 170 4890 4896 erratum appears in J Bacteriol 1989 Feb;171(2:1233.) 2902070 183 Lindahl M Wadstrom T 1986 Binding to erythrocyte membrane glycoproteins and carbohydrate specificity of F41 fimbriae of enterotoxigenic Escherichia coli FEMS Microbiol Lett 34 297 300 184 Yang B Huang X Yan X Ren J Yang S Zou Z Zeng W Ou Y Huang W Huang L 2009 Detection of quantitative trait loci for porcine susceptibility to enterotoxigenic Escherichia coli F41 in a White Duroc × Chinese Erhualian resource population Animal 3 946 950 22444814 185 Jacobs AA Simons BH de Graaf FK 1987 The role of lysine-132 and arginine-136 in the receptor-binding domain of the K99 fibrillar subunit EMBO J 6 1805 1808 2886335 186 Jacobs AA van den Berg PA Bak HJ de Graaf FK 1986 Localization of lysine residues in the binding domain of the K99 fibrillar subunit of enterotoxigenic Escherichia coli Biochim Biophys Acta 872 92 97 3089285 187 Ono E Abe K Nakazawa M Naiki M 1989 Ganglioside epitope recognized by K99 fimbriae from enterotoxigenic Escherichia coli Infect Immun 57 907 911 2465273 188 Smit H Gaastra W Kamerling JP Vliegenthart JF de Graaf FK 1984 Isolation and structural characterization of the equine erythrocyte receptor for enterotoxigenic Escherichia coli K99 fimbrial adhesin Infect Immun 46 578 584 6150011 189 Teneberg S Willemsen PT de Graaf FK Karlsson KA 1993 Calf small intestine receptors for K99 fimbriated enterotoxigenic Escherichia coli FEMS Microbiol Lett 109 107 112 8391499 190 Kyogashima M Ginsburg V Krivan HC 1989 Escherichia coli K99 binds to N-glycolylsialoparagloboside and N-glycolyl-GM3 found in piglet small intestine Arch Biochem Biophys 270 391 397 2648997 191 Teneberg S Willemsen PT de Graaf FK Stenhagen G Pimlott W Jovall PA Angström J Karlsson KA 1994 Characterization of gangliosides of epithelial cells of calf small intestine, with special reference to receptor-active sequences for enteropathogenic Escherichia coli K99 J Biochem 116 560 574 7531686 192 Dean EA Isaacson RE 1982 In vitro adhesion of piliated Escherichia coli to small intestinal villous epithelial cells from rabbits and the identification of a soluble 987P pilus receptor-containing fraction Infect Immun 36 1192 1198 6124505 193 Dean EA Isaacson RE 1985 Purification and characterization of a receptor for the 987P pilus of Escherichia coli Infect Immun 47 98 105 2856917 194 Dean EA Isaacson RE 1985 Location and distribution of a receptor for the 987P pilus of Escherichia coli in small intestines Infect Immun 47 345 348 2857152 195 Khan AS Schifferli DM 1994 A minor 987P protein different from the structural fimbrial subunit is the adhesin Infect Immun 62 4233 4243 7927679 196 Zhu G Chen H Choi BK Del Piero F Schifferli DM 2005 Histone H1 proteins act as receptors for the 987P fimbriae of enterotoxigenic Escherichia coli J Biol Chem 280 23057 23065 15840569 197 Choi BK Schifferli DM 1999 Lysine residue 117 of the FasG adhesin of enterotoxigenic Escherichia coli is essential for binding of 987P fimbriae to sulfatide Infect Immun 67 5755 5761 10531225 198 Choi BK Schifferli DM 2001 Characterization of FasG segments required for 987P fimbria-mediated binding to piglet glycoprotein receptors Infect Immun 69 6625 6632 11598031 199 Imberechts H Wild P Charlier G De Greve H Lintermans P Pohl P 1996 Characterization of F18 fimbrial genes fedE and fedF involved in adhesion and length of enterotoxemic Escherichia coli strain 107/86 Microb Pathog 21 183 192 8878015 200 Smeds A Hemmann K Jakava-Viljanen M Pelkonen S Imberechts H Palva A 2001 Characterization of the adhesin of Escherichia coli F18 fimbriae Infect Immun 69 7941 7945 11705982 201 Smeds A Pertovaara M Timonen T Pohjanvirta T Pelkonen S Palva A 2003 Mapping the binding domain of the F18 fimbrial adhesin Infect Immun 71 2163 2172 12654838 202 Tiels P Verdonck F Smet A Goddeeris B Cox E 2005 The F18 fimbrial adhesin FedF is highly conserved among F18(+)Escherichia coli isolates Vet Microbiol 203 Rippinger P Bertschinger HU Imberechts H Nagy B Sorg I Stamm M Wild P Wittig W 1995 Design ations F18ab and F18ac for the related fimbrial types F107, 2134P and 8813 of Escherichia coli isolated from porcine postweaning diarrhoea and from oedema disease Vet Microbiol 45 281 295 7483242 204 Vögeli P Bertschinger HU Stamm M Stricker C Hagger C Fries R Rapacz J Stranzinger G 1996 Ge nes specifying receptors for F18 fimbriated Escherichia coli, causing oedema disease and postweaning diarrhoea in pigs, map to chromosome 6 Anim Genet 27 321 328 8930072 205 Meijerink E Neuenschwander S Fries R Dinter A Bertschinger HU Stranzinger G Vögeli P 2000 A DNA polymorphism influencing alpha(1,2)fucosyltransferase activity of the pig FUT1 enzyme determines susceptibility of small intestinal epithelium to Escherichia coli F18 adhesion Immunogenetics 52 129 136 11132149 206 Meijerink E Fries R Vögeli P Masabanda J Wigger G Stricker C Neuenschwander S Bertschinger HU Stranzinger G 1997 Two alpha(1,2) fucosyltransferase genes on porcine chromosome 6q11 are closely linked to the blood group inhibitor (S) and Escherichia coli F18 receptor (ECF18R) loci Mamm Genome 8 736 741 9321466 207 Snoeck V Verdonck F Cox E Goddeeris BM 2004 Inhibition of adhesion of F18+ Escherichia coli to piglet intestinal villous enterocytes by monoclonal antibody against blood group H-2 antigen Vet Microbiol 100 241 246 15145502 208 Coddens A Diswall M Angström J Breimer ME Goddeeris B Cox E Teneberg S 2009 Recognition of blood group ABH type 1 determinants by the FedF adhesin of F18-fimbriated Escherichia coli J Biol Chem 284 9713 9726 19208633 209 Moonens K Bouckaert J Coddens A Tran T Panjikar S De Kerpel M Cox E Remaut H De Greve H 2012 Structural insight in histo-blood group binding by the F18 fimbrial adhesin FedF Mol Microbiol 86 82 95 22812428 210 Lonardi E Moonens K Buts L de Boer AR Olsson JD Weiss MS Fabre E Guérardel Y Deelder AM Oscarson S Wuhrer M Bouckaert J 2013 Structural Sampling of Glycan Interaction Profiles Reveals Mucosal Receptors for Fimbrial Adhesins of Enterotoxigenic Escherichia coli Biology (Basel) 2 894 917 24833052 211 Zhao Q Liu Y Dong W Zhu S Huo Y Wu S Bao W 2014 Genetic variations of TAP1 gene exon 3 affects gene expression and Escherichia coli F18 resistance in piglets Int J Mol Sci 15 11161 11171 24955792 212 Wang SJ Liu WJ Yang LG Sargent CA Liu HB Wang C Liu XD Zhao SH Affara NA Liang AX Zhang SJ 2012 Effects of FUT1 gene mutation on resistance to infectious disease Mol Biol Rep 39 2805 2810 21695432 213 Bao WB Ye L Zhu J Pan ZY Zhu GQ Huang XG Wu SL 2011 Polymorphism of M307 of the FUT1 gene and its relationship with some immune indexes in Sutai pigs (Duroc x Meishan) Biochem Genet 49 665 673 21626436 214 Bao WB Wu SL Musa HH Zhu GQ Chen GH 2008 Genetic variation at the alpha-1-fucosyltransferase (FUT1) gene in Asian wild boar and Chinese and Western commercial pig breeds J Anim Breed Genet 125 427 430 19134079 215 Lintermans PF Bertels A Schlicker C Deboeck F Charlier G Pohl P Norgren M Normark S van Montagu M De Greve H 1991 Identification, characterization, and nucleotide sequence of the F17-G gene, which determines receptor binding of Escherichia coli F17 fimbriae J Bacteriol 173 3366 3373 1675211 216 Sanchez R Kanarek L Koninkx J Hendriks H Lintermans P Bertels A Charlier G Van Driessche E 1993 Inhibition of adhesion of enterotoxigenic Escherichia coli cells expressing F17 fimbriae to small intestinal mucus and brush-border membranes of young calves Microb Pathog 15 207 219 217 Buts L Bouckaert J De Genst E Loris R Oscarson S Lahmann M Messens J Brosens E Wyns L De Greve H 2003 The fimbrial adhesin F17-G of enterotoxigenic Escherichia coli has an immunoglobulin-like lectin domain that binds N-acetylglucosamine Mol Microbiol 49 705 715 12864853 218 Merckel MC Tanskanen J Edelman S Westerlund-Wikström B Korhonen TK Goldman A 2003 The structural basis of receptor-binding by Escherichia coli associated with diarrhea and septicemia J Mol Biol 331 897 905 12909017 219 Geibel S Waksman G 2014 The molecular dissection of the chaperone-usher pathway Biochim Biophys Acta 1843 1559 1567 24140205 220 Lycklama A Nijeholt JA Driessen AJ 2012 The bacterial Sec-translocase: structure and mechanism Philos Trans R Soc Lond B Biol Sci 367 1016 1028 22411975 221 Mooi FR Wijfjes A de Graaf FK 1983 Identification and characterization of precursors in the biosynthesis of the K88ab fimbria of Escherichia coli J Bacteriol 154 41 49 6131880 222 Sauer FG Fütterer K Pinkner JS Dodson KW Hultgren SJ Waksman G 1999 Structural basis of chaperone function and pilus biogenesis Science 285 1058 1061 10446050 223 Jacob-Dubuisson F Striker R Hultgren SJ 1994 Chaperone-assisted self-assembly of pili independent of cellular energy J Biol Chem 269 12447 12455 7909802 224 Palomino C Marín E Fernández LA 2011 The fimbrial usher FimD follows the SurA-BamB pathway for its assembly in the outer membrane of Escherichia coli J Bacteriol 193 5222 5230 21784935 225 Hagan CL Silhavy TJ Kahne D 2011 β-Barrel membrane protein assembly by the Bam complex Annu Rev Biochem 80 189 210 21370981 226 Barnhart MM Sauer FG Pinkner JS Hultgren SJ 2003 Chaperone-subunit-usher interactions required for donor strand exchange during bacterial pilus assembly J Bacteriol 185 2723 2730 12700251 227 Van Molle I Moonens K Garcia-Pino A Buts L De Kerpel M Wyns L Bouckaert J De Greve H 2009 Structural and thermodynamic characterization of pre- and postpolymerization states in the F4 fimbrial subunit FaeG J Mol Biol 394 957 967 19799915 228 Thanassi DG Saulino ET Lombardo MJ Roth R Heuser J Hultgren SJ 1998 The PapC usher forms an oligomeric channel: implications for pilus biogenesis across the outer membrane Proc Natl Acad Sci USA 95 3146 3151 9501230 229 Huang Y Smith BS Chen LX Baxter RH Deisenhofer J 2009 Insights into pilus assembly and secretion from the structure and functional characterization of usher PapC Proc Natl Acad Sci USA 106 7403 7407 19380723 230 Saulino ET Bullitt E Hultgren SJ 2000 Snapshots of usher-mediated protein secretion and ordered pilus assembly Proc Natl Acad Sci USA 97 9240 9245 10908657 231 Lowe MA Holt SC Eisenstein BI 1987 Immunoelectron microscopic analysis of elongation of type 1 fimbriae in Escherichia coli J Bacteriol 169 157 163 2878917 232 Verdonck F Cox E Schepers E Imberechts H Joensuu J Goddeeris BM 2004 Conserved regions in the sequence of the F4 (K88) fimbrial adhesin FaeG suggest a donor strand mechanism in F4 assembly Vet Microbiol 102 215 225 15327796 233 Bakker D 1991 Studies on the K88 fimbriae of enteropathogenic Escherichia coli Vrije Universiteit Amsterdam, The Netherlands 234 Bakker D Vader CE Roosendaal B Mooi FR Oudega B de Graaf FK 1991 Structure and function of periplasmic chaperone-like proteins involved in the biosynthesis of K88 and K99 fimbriae in enterotoxigenic Escherichia coli Mol Microbiol 5 875 886 1713284 235 Mol O Visschers RW de Graff FK Oudega B 1994 Escherichia coli periplasmic chaperone FaeE is a homodimer and the chaperone-K88 subunit complex is a heterotrimer Mol Microbiol 11 391 402 8170401 236 Mol O Oudhuis WC Fokkema H Oudega B 1996 The N-terminal beta-barrel domain of the Escherichia coli K88 periplasmic chaperone FaeE determines fimbrial subunit recognition and dimerization Mol Microbiol 22 379 388 8930921 237 Mol O Oud RP de Graaf FK Oudega B 1995 The Escherichia coli K88 periplasmic chaperone FaeE forms a heterotrimeric complex with the minor fimbrial component FaeH and with the minor fimbrial component FaeI Microb Pathog 18 115 128 7643741 238 Van Molle I Moonens K Buts L Garcia-Pino A Panjikar S Wyns L De Greve H Bouckaert J 2009 The F4 fimbrial chaperone FaeE is stable as a monomer that does not require self-capping of its pilin-interactive surfaces Acta Crystallogr D Biol Crystallogr 65 411 420 19390146 239 Mooi FR Claassen I Bakker D Kuipers H de Graaf FK 1986 Regulation and structure of an Escherichia coli gene coding for an outer membrane protein involved in export of K88ab fimbrial subunits Nucleic Acids Res 14 2443 2457 2870470 240 Valent QA Zaal J de Graaf FK Oudega B 1995 Subcellular localization and topology of the K88 usher FaeD in Escherichia coli Mol Microbiol 16 1243 1257 8577257 241 Harms N Oudhuis WC Eppens EA Valent QA Koster M Luirink J Oudega B 1999 Epitope tagging analysis of the outer membrane folding of the molecular usher FaeD involved in K88 fimbriae biosynthesis in Escherichia coli J Mol Microbiol Biotechnol 1 319 325 10943563 242 Mooi FR Wouters C Wijfjes A de Graaf FK 1982 Construction and characterization of mutants impaired in the biosynthesis of the K88ab antigen J Bacteriol 150 512 521 6121788 243 van Doorn J Oudega B Mooi FR de Graaf FK 1982 Subcellular localization of polypeptides involved in the biosynthesis of K88ab fimbriae FEMS Microbiol Lett 13 99 104 244 Dougan G Dowd G Kehoe M 1983 Organization of K88ac-encoded polypeptides in the Escherichia coli cell envelope: use of minicells and outer membrane protein mutants for studying assembly of pili J Bacteriol 153 364 370 6129239 245 Mol O Oudhuis WC Oud RP Sijbrandi R Luirink J Harms N Oudega B 2001 Biosynthesis of K88 fimbriae in Escherichia coli: interaction of tip-subunit FaeC with the periplasmic chaperone FaeE and the outer membrane usher FaeD J Mol Microbiol Biotechnol 3 135 142 11200226 246 Oudega B de Graaf M de Boer L Bakker D Vader CE Mooi FR de Graaf FK 1989 Detection and identification of FaeC as a minor component of K88 fibrillae of Escherichia coli Mol Microbiol 3 645 652 2668694 247 Mooi FR van Buuren M Koopman G Roosendaal B de Graaf FK 1984 K88ab gene of Escherichia coli encodes a fimbria-like protein distinct from the K88ab fimbrial adhesin J Bacteriol 159 482 487 6086572 248 Mol O Oudega B 1996 Molecular and structural aspects of fimbriae biosynthesis and assembly in Escherichia coli FEMS Microbiol Rev 19 25 52 8916554 249 Korth MJ Apostol JM Jr Moseley SL 1992 Functional expression of heterologous fimbrial subunits mediated by the F41, K88, and CS31A determinants of Escherichia coli Infect Immun 60 2500 2505 1350275 250 Mol O Fokkema H Oudega B 1996 The Escherichia coli K99 periplasmic chaperone FanE is a monomeric protein FEMS Microbiol Lett 138 185 189 9026444 251 de Graaf FK Krenn BE Klaasen P 1984 Organization and expression of genes involved in the biosynthesis of K99 fimbriae Infect Immun 43 508 514 6141144 252 Simons BL Willemsen PT Bakker D Roosendaal B De Graaf FK Oudega B 1990 Structure, localization and function of FanF, a minor component of K99 fibrillae of enterotoxigenic Escherichia coli Mol Microbiol 4 2041 2050 1982454 253 Roosendaal E Jacobs AA Rathman P Sondermeyer C Stegehuis F Oudega B de Graaf FK 1987 Primary structure and subcellular localization of two fimbrial subunit-like proteins involved in the biosynthesis of K99 fibrillae Mol Microbiol 1 211 217 2897066 254 Simons LH Willemsen PT Bakker D de Graaf FK Oudega B 1991 Localization and function of FanH and FanG, minor components of K99 fimbriae of enterotoxigenic Escherichia coli Microb Pathog 11 325 336 1687753 255 Schifferli DM Beachey EH Taylor RK 1991 Genetic analysis of 987P adhesion and fimbriation of Escherichia coli: the fas genes link both phenotypes J Bacteriol 173 1230 1240 1671386 256 Schifferli DM Alrutz MA 1994 Permissive linker insertion sites in the outer membrane protein of 987P fimbriae of Escherichia coli J Bacteriol 176 1099 1110 7906265 257 Edwards RA Cao J Schifferli DM 1996 Identification of major and minor chaperone proteins involved in the export of 987P fimbriae J Bacteriol 178 3426 3433 8655537 258 Imberechts H Van Pelt N De Greve H Lintermans P 1994 Sequences related to the major subunit gene fedA of F107 fimbriae in porcine Escherichia coli strains that express adhesive fimbriae FEMS Microbiol Lett 119 309 314 7914179 259 Kennan RM Moncktor RP McDougall BM Conway PL 1995 Confirmation that DNA encoding the major fimbrial subunit of Av24 fimbriae is homologous to DNA encoding the major fimbrial subunit of F107 fimbriae Microb Pathog 18 67 72 7783599 260 Bosworth BT Dean-Nystrom EA Casey TA Neibergs HL 1998 Differentiation of F18ab+ from F18ac+ Escherichia coli by single-strand conformational polymorphism analysis of the major fimbrial subunit gene (fedA) Clin Diagn Lab Immunol 5 299 302 9605980 261 Byun JW Jung BY Kim HY Fairbrother JM Lee MH Lee WK 2013 Real-time PCR for differentiation of F18 variants among enterotoxigenic and Shiga toxin-producing Escherichia coli from piglets with diarrhoea and oedema disease Vet J 198 538 540 23992871 262 Lee JH Isaacson RE 1995 Expression of the gene cluster associated with the Escherichia coli pilus adhesin K99 Infect Immun 63 4143 4149 7558331 263 Yang Y Yao F Zhou M Zhu J Zhang X Bao W Wu S Hardwidge PR Zhu G 2013 F18ab Escherichia coli flagella expression is regulated by acyl-homoserine lactone and contributes to bacterial virulence Vet Microbiol 165 378 383 23693029 264 Sturbelle RT de Avila LF Roos TB Borchardt JL da Conceição RC Dellagostin OA Leite FP 2015 The role of quorum sensing in Escherichia coli (ETEC) virulence factors Vet Microbiol 180 245 252 26386492 265 Bak G Lee J Suk S Kim D Young Lee J Kim KS Choi BS Lee Y 2015 Identification of novel sRNAs involved in biofilm formation, motility, and fimbriae formation in Escherichia coli Sci Rep 5 15287 26469694 266 Khandige S Kronborg T Uhlin BE Møller-Jensen J 2015 sRNA-Mediated Regulation of P-Fimbriae Phase Variation in Uropathogenic Escherichia coli PLoS Pathog 11 e1005109 26291711 267 Huisman TT Bakker D Klaasen P de Graaf FK 1994 Leucine-responsive regulatory protein, IS1 insertions, and the negative regulator FaeA control the expression of the fae (K88) operon in Escherichia coli Mol Microbiol 11 525 536 8152376 268 van der Woude MW Braaten BA Low DA 1992 Evidence for global regulatory control of pilus expression in Escherichia coli by Lrp and DNA methylation: model building based on analysis of pap Mol Microbiol 6 2429 2435 1357527 269 Huisman TT de Graaf FK 1995 Negative control of fae (K88) expression by the ‘global’ regulator Lrp is modulated by the ‘local’ regulator FaeA and affected by DNA methylation Mol Microbiol 16 943 953 7476191 270 van der Woude MW Braster M van Verseveld HW de Graaf FK 1990 Control of temperature-dependent synthesis of K99 fimbriae FEMS Microbiol Lett 56 183 188 1970546 271 de Graaf FK Roorda I 1982 Production, purification, and characterization of the fimbrial adhesive antigen F41 isolated from calf enteropathogenic Escherichia coli strain B41M Infect Immun 36 751 758 6123486 272 Isaacson RE 1980 Factors affecting expression of the Escherichia coli pilus K99 Infect Immun 28 190 194 6103876 273 White-Ziegler CA Villapakkam A Ronaszeki K Young S 2000 H-NS controls pap and daa fimbrial transcription in Escherichia coli in response to multiple environmental cues J Bacteriol 182 6391 6400 11053383 274 Francis DH Ryan CJ Fritzemeier JD 1983 Effect of sodium acetate on expression of K99 pili by Escherichia coli Infect Immun 41 1368 1369 6136464 275 Isaacson RE 1983 Regulation of expression of Escherichia coli pilus K99 Infect Immun 40 633 639 6132879 276 Girardeau JP Dubourguier HC Gouet P 1982 Effect of glucose and amino acids on expression of K99 antigen in Escherichia coli J Gen Microbiol 128 2243 2249 6818326 277 Roosendaal E Boots M de Graaf FK 1987 Two novel genes, fanA and fanB, involved in the biogenesis of K99 fimbriae Nucleic Acids Res 15 5973 5984 2888075 278 Roosendaal B Damoiseaux J Jordi W de Graaf FK 1989 Transcriptional organization of the DNA region controlling expression of the K99 gene cluster Mol Gen Genet 215 250 256 2651880 279 Braaten BA Platko JV van der Woude MW Simons BH de Graaf FK Calvo JM Low DA 1992 Leucine-responsive regulatory protein controls the expression of both the pap and fan pili operons in Escherichia coli Proc Natl Acad Sci USA 89 4250 4254 1350087 280 de Graaf FK Klaasen-Boor P van Hees JE 1980 Biosynthesis of the K99 surface antigen is repressed by alanine Infect Immun 30 125 128 7002786 281 Inoue OJ Lee JH Isaacson RE 1993 Transcriptional organization of the Escherichia coli pilus adhesin K99 Mol Microbiol 10 607 613 7968538 282 van der Woude MW de Graaf FK van Verseveld HW 1989 Production of the fimbrial adhesin 987P by enterotoxigenic Escherichia coli during growth under controlled conditions in a chemostat J Gen Microbiol 135 3421 3429 2576874 283 Honarvar S Choi BK Schifferli DM 2003 Phase variation of the 987P-like CS18 fimbriae of human enterotoxigenic Escherichia coli is regulated by site-specific recombinases Mol Microbiol 48 157 171 12657052 284 Edwards RA Keller LH Schifferli DM 1998 Improved allelic exchange vectors and their use to analyze 987P fimbria gene expression Gene 207 149 157 9511756 285 Abraham JM Freitag CS Clements JR Eisenstein BI 1985 An invertible element of DNA controls phase variation of type 1 fimbriae of Escherichia coli Proc Natl Acad Sci USA 82 5724 5727 2863818 286 van der Woude M Braaten B Low D 1996 Epigenetic phase variation of the pap operon in Escherichia coli Trends Microbiol 4 5 9 8824788 287 Klaasen P de Graaf FK 1990 Characterization of FapR, a positive regulator of expression of the 987P operon in enterotoxigenic Escherichia coli Mol Microbiol 4 1779 1783 2077360 288 Munson GP Holcomb LG Scott JR 2001 Novel group of virulence activators within the AraC family that are not restricted to upstream binding sites Infect Immun 69 186 193 11119505 289 Nagy B Whipp SC Imberechts H Bertschinger HU Dean-Nystrom EA Casey TA Salajka E 1997 Biological relationship between F18ab and F18ac fimbriae of enterotoxigenic and verotoxigenic Escherichia coli from weaned pigs with oedema disease or diarrhoea Microb Pathog 22 1 11 9032757 290 Cantey JR Blake RK Williford JR Moseley SL 1999 Characterization of the Escherichia coli AF/R1 pilus operon: novel genes necessary for transcriptional regulation and for pilus-mediated adherence Infect Immun 67 2292 2298 10225886 291 Shepard SM Danzeisen JL Isaacson RE Seemann T Achtman M Johnson TJ 2012 Genome sequences and phylogenetic analysis of K88- and F18-positive porcine enterotoxigenic Escherichia coli J Bacteriol 194 395 405 22081385 292 Zhang W Berberov EM Freeling J He D Moxley RA Francis DH 2006 Significance of heat-stable and heat-labile enterotoxins in porcine colibacillosis in an additive model for pathogenicity studies Infect Immun 74 3107 3114 16714538 293 Savarino SJ Fasano A Robertson DC Levine MM 1991 Enteroaggregative Escherichia coli elaborate a heat-stable enterotoxin demonstrable in an in vitro rabbit intestinal model J Clin Invest 87 1450 1455 2010554 294 Ménard LP Dubreuil JD 2002 Enteroaggregative Escherichia coli heat-stable enterotoxin 1 (EAST1): a new toxin with an old twist Crit Rev Microbiol 28 43 60 12003040 295 Dubreuil JD 2015 Chapter 31, ••• 874 910 In Escherchia coli heat-stable enterotoxins. The comprehensive sourcebook of bacterial protein toxins Elsevier 296 De SN Bhattacharya K Sarkar JK 1956 A study of the pathogenicity of strains of Bacterium coli from acute and chronic enteritis J Pathol Bacteriol 71 201 209 13307349 297 Ochi S Shimizu T Ohtani K Ichinose Y Arimitsu H Tsukamoto K Kato M Tsuji T 2009 Nucleotide sequence analysis of the enterotoxigenic Escherichia coli Ent plasmid DNA Res 16 299 309 19767599 298 Scotland SM Day NP Rowe B 1983 Acquisition and maintenance of enterotoxin plasmids in wild-type strains of Escherichia coli J Gen Microbiol 129 3111 3120 6361215 299 Zhang C Rausch D Zhang W 2009 Little heterogeneity among genes encoding heat-labile and heat-stable toxins of enterotoxigenic Escherichia coli strains isolated from diarrheal pigs Appl Environ Microbiol 75 6402 6405 19684170 300 Guth BE Twiddy EM Trabulsi LR Holmes RK 1986 Variation in chemical properties and antigenic determinants among type II heat-labile enterotoxins of Escherichia coli Infect Immun 54 529 536 2429930 301 Nawar HF Berenson CS Hajishengallis G Takematsu H Mandell L Clare RL Connell TD 2010 Bindi ng to gangliosides containing N-acetylneuraminic acid is sufficient to mediate the immunomodulatory properties of the nontoxic mucosal adjuvant LT-IIb(T13I) Clin Vaccine Immunol 17 969 978 20392887 302 Casey TA Connell TD Holmes RK Whipp SC 2012 Evaluation of heat-labile enterotoxins type IIa and type IIb in the pathogenicity of enterotoxigenic Escherichia coli for neonatal pigs Vet Microbiol 159 83 89 22480773 303 Connell TD 2007 Cholera toxin, LT-I, LT-IIa and LT-IIb: the critical role of ganglioside binding in immunomodulation by type I and type II heat-labile enterotoxins Expert Rev Vaccines 6 821 834 17931161 304 Mekalanos JJ Sublett RD Romig WR 1979 Genetic mapping of toxin regulatory mutations in Vibrio cholerae J Bacteriol 139 859 865 479110 305 Pickett CL Twiddy EM Belisle BW Holmes RK 1986 Cloning of genes that encode a new heat-labile enterotoxin of Escherichia coli J Bacteriol 165 348 352 3511028 306 Connell TD Holmes RK 1992 Characterization of hybrid toxins produced in Escherichia coli by assembly of A and B polypeptides from type I and type II heat-labile enterotoxins Infect Immun 60 1653 1661 1548088 307 Dorsey FC Fischer JF Fleckenstein JM 2006 Directed delivery of heat-labile enterotoxin by enterotoxigenic Escherichia coli Cell Microbiol 8 1516 1527 16922869 308 Johnson AM Kaushik RS Francis DH Fleckenstein JM Hardwidge PR 2009 Heat-labile enterotoxin promotes Escherichia coli adherence to intestinal epithelial cells J Bacteriol 191 178 186 18978047 309 Wijemanne P Moxley RA 2014 Glucose significantly enhances enterotoxigenic Escherichia coli adherence to intestinal epithelial cells through its effects on heat-labile enterotoxin production PLoS One 9 e113230 25409235 310 Lasaro MA Rodrigues JF Mathias-Santos C Guth BE Régua-Mangia A Piantino Ferreira AJ Takagi M Cabrera-Crespo J Sbrogio-Almeida ME de Souza Ferreira LC 2006 Production and release of heat-labile toxin by wild-type human-derived enterotoxigenic Escherichia coli FEMS Immunol Med Microbiol 48 123 131 16965360 311 Rodighiero C Aman AT Kenny MJ Moss J Lencer WI Hirst TR 1999 Structural basis for the differential toxicity of cholera toxin and Escherichia coli heat-labile enterotoxin. Construction of hybrid toxins identifies the A2-domain as the determinant of differential toxicity J Biol Chem 274 3962 3969 9933586 312 Goins B Freire E 1988 Thermal stability and intersubunit interactions of cholera toxin in solution and in association with its cell-surface receptor ganglioside GM1 Biochemistry 27 2046 2052 3378043 313 Surewicz WK Leddy JJ Mantsch HH 1990 Structure, stability, and receptor interaction of cholera toxin as studied by Fourier-transform infrared spectroscopy Biochemistry 29 8106 8111 2261465 314 Hegde A Bhat GK Mallya S 2009 Effect of stress on production of heat labile enterotoxin by Escherichia coli Indian J Med Microbiol 27 325 328 19736401 315 Mekalanos JJ Collier RJ Romig WR 1979 Enzymic activity of cholera toxin. II. Relationships to proteolytic processing, disulfide bond reduction, and subunit composition J Biol Chem 254 5855 5861 221485 316 Gonzales L Ali ZB Nygren E Wang Z Karlsson S Zhu B Quiding-Järbrink M Sjöling Å 2013 Alkaline pH Is a signal for optimal production and secretion of the heat labile toxin, LT in enterotoxigenic Escherichia coli (ETEC) PLoS One 8 e74069 24058516 317 Pande AH Scaglione P Taylor M Nemec KN Tuthill S Moe D Holmes RK Tatulian SA Teter K 2007 Conformational instability of the cholera toxin A1 polypeptide J Mol Biol 374 1114 1128 17976649 318 Merritt EA Pronk SE Sixma TK Kalk KH van Zanten BA Hol WG 1994 Structure of partially-activated E. coli heat-labile enterotoxin (LT) at 2.6 A resolution FEBS Lett 337 88 92 8276119 319 Sixma TK Stein PE Hol WG Read RJ 1993 Comparison of the B-pentamers of heat-labile enterotoxin and verotoxin-1: two structures with remarkable similarity and dissimilarity Biochemistry 32 191 198 8418837 320 Clements JD Finkelstein RA 1979 Isolation and characterization of homogeneous heat-labile enterotoxins with high specific activity from Escherichia coli cultures Infect Immun 24 760 769 89088 321 Heggelund JE Haugen E Lygren B Mackenzie A Holmner Å Vasile F Reina JJ Bernardi A Krengel U 2012 Both El Tor and classical cholera toxin bind blood group determinants Biochem Biophys Res Commun 418 731 735 22305717 322 Mandal PK Branson TR Hayes ED Ross JF Gavín JA Daranas AH Turnbull WB 2012 Towards a structural basis for the relationship between blood group and the severity of El Tor cholera Angew Chem Int Ed Engl 51 5143 5146 22488789 323 Vasile F Reina JJ Potenza D Heggelund JE Mackenzie A Krengel U Bernardi A 2014 Comprehensive analysis of blood group antigen binding to classical and El Tor cholera toxin B-pentamers by NMR Glycobiology 24 766 778 24829308 324 Guidry JJ Cárdenas L Cheng E Clements JD 1997 Role of receptor binding in toxicity, immunogenicity, and adjuvanticity of Escherichia coli heat-labile enterotoxin Infect Immun 65 4943 4950 9393780 325 Sonnino S Mauri L Chigorno V Prinetti A 2007 Gangliosides as components of lipid membrane domains Glycobiology 17 1R 13R 326 Ahmed T Lundgren A Arifuzzaman M Qadri F Teneberg S Svennerholm AM 2009 Children with the Le(a+b−) blood group have increased susceptibility to diarrhea caused by enterotoxigenic Escherichia coli expressing colonization factor I group fimbriae Infect Immun 77 2059 2064 19273560 327 Horstman AL Kuehn MJ 2002 Bacterial surface association of heat-labile enterotoxin through lipopolysaccharide after secretion via the general secretory pathway J Biol Chem 277 32538 32545 12087095 328 Horstman AL Kuehn MJ 2000 Enterotoxigenic Escherichia coli secretes active heat-labile enterotoxin via outer membrane vesicles J Biol Chem 275 12489 12496 10777535 329 Raetz CR Whitfield C 2002 Lipopolysaccharide endotoxins Annu Rev Biochem 71 635 700 12045108 330 Kesty NC Kuehn MJ 2004 Incorporation of heterologous outer membrane and periplasmic proteins into Escherichia coli outer membrane vesicles J Biol Chem 279 2069 2076 14578354 331 Kesty NC Mason KM Reedy M Miller SE Kuehn MJ 2004 Enterotoxigenic Escherichia coli vesicles target toxin delivery into mammalian cells EMBO J 23 4538 4549 15549136 332 Horstman AL Bauman SJ Kuehn MJ 2004 Lipopolysaccharide 3-deoxy-D-manno-octulosonic acid (Kdo) core determines bacterial association of secreted toxins J Biol Chem 279 8070 8075 14660669 333 Bonnington KE Kuehn MJ 2014 Protein selection and export via outer membrane vesicles Biochim Biophys Acta 1843 1612 1619 24370777 334 Kuehn MJ Kesty NC 2005 Bacterial outer membrane vesicles and the host-pathogen interaction Genes Dev 19 2645 2655 16291643 335 Chatterjee D Chaudhuri K 2011 Association of cholera toxin with Vibrio cholerae outer membrane vesicles which are internalized by human intestinal epithelial cells FEBS Lett 585 1357 1362 21510946 336 Schlör S Riedl S Blass J Reidl J 2000 Genetic rearrangements of the regions adjacent to genes encoding heat-labile enterotoxins (eltAB) of enterotoxigenic Escherichia coli strains Appl Environ Microbiol 66 352 358 10618247 337 Yamamoto T Gojobori T Yokota T 1987 Evolutionary origin of pathogenic determinants in enterotoxigenic Escherichia coli and Vibrio cholerae O1 J Bacteriol 169 1352 1357 3546273 338 Yang J Tauschek M Strugnell R Robins-Browne RM 2005 The H-NS protein represses transcription of the eltAB operon, which encodes heat-labile enterotoxin in enterotoxigenic Escherichia coli, by binding to regions downstream of the promoter Microbiology 151 1199 1208 15817787 339 Takashi K Fujita I Kobari K 1989 Effects of short chain fatty acids on the production of heat-labile enterotoxin from enterotoxigenic Escherichia coli Jpn J Pharmacol 50 495 498 2674496 340 Wülfing C Rappuoli R 1997 Efficient production of heat-labile enterotoxin mutant proteins by overexpression of dsbA in a degP-deficient Escherichia coli strain Arch Microbiol 167 280 283 9094224 341 Streatfield SJ Sandkvist M Sixma TK Bagdasarian M Hol WG Hirst TR 1992 Intermolecular interactions between the A and B subunits of heat-labile enterotoxin from Escherichia coli promote holotoxin assembly and stability in vivo Proc Natl Acad Sci USA 89 12140 12144 1465452 342 Chung WY Carter R Hardy T Sack M Hirst TR James RF 2006 Inhibition of Escherichia coli heat-labile enterotoxin B subunit pentamer (EtxB5) assembly in vitro using monoclonal antibodies J Biol Chem 281 39465 39470 17038315 343 Grant CC Messer RJ Cieplak W Jr 1994 Role of trypsin-like cleavage at arginine 192 in the enzymatic and cytotonic activities of Escherichia coli heat-labile enterotoxin Infect Immun 62 4270 4278 7927684 344 Tauschek M Gorrell RJ Strugnell RA Robins-Browne RM 2002 Identification of a protein secretory pathway for the secretion of heat-labile enterotoxin by an enterotoxigenic strain of Escherichia coli Proc Natl Acad Sci USA 99 7066 7071 12011463 345 Johnson TL Abendroth J Hol WG Sandkvist M 2006 Type II secretion: from structure to function FEMS Microbiol Lett 255 175 186 16448494 346 Yang J Baldi DL Tauschek M Strugnell RA Robins-Browne RM 2007 Transcriptional regulation of the yghJ-pppA-yghG-gspCDEFGHIJKLM cluster, encoding the type II secretion pathway in enterotoxigenic Escherichia coli J Bacteriol 189 142 150 17085567 347 Hirst TR Sanchez J Kaper JB Hardy SJ Holmgren J 1984 Mechanism of toxin secretion by Vibrio cholerae investigated in strains harboring plasmids that encode heat-labile enterotoxins of Escherichia coli Proc Natl Acad Sci USA 81 7752 7756 6393126 348 Brown EA Hardwidge PR 2007 Biochemical characterization of the enterotoxigenic Escherichia coli LeoA protein Microbiology 153 3776 3784 17975086 349 Fleckenstein JM Lindler LE Elsinghorst EA Dale JB 2000 Identification of a gene within a pathogenicity island of enterotoxigenic Escherichia coli H10407 required for maximal secretion of the heat-labile enterotoxin Infect Immun 68 2766 2774 10768971 350 Michie KA Boysen A Low HH Møller-Jensen J Löwe J 2014 LeoA, B and C from enterotoxigenic Escherichia coli (ETEC) are bacterial dynamins PLoS One 9 e107211 25203511 351 Spangler BD 1992 Structure and function of cholera toxin and the related Escherichia coli heat-labile enterotoxin Microbiol Rev 56 622 647 1480112 352 Iglesias-Bartolomé R Trenchi A Comín R Moyano AL Nores GA Daniotti JL 2009 Differential endocytic trafficking of neuropathy-associated antibodies to GM1 ganglioside and cholera toxin in epithelial and neural cells Biochim Biophys Acta 1788 2526 2540 19800863 353 Wernick NL Chinnapen DJ Cho JA Lencer WI 2010 Cholera toxin: an intracellular journey into the cytosol by way of the endoplasmic reticulum Toxins (Basel) 2 310 325 22069586 354 Majoul I Ferrari D Söling HD 1997 Reduction of protein disulfide bonds in an oxidizing environment. The disulfide bridge of cholera toxin A-subunit is reduced in the endoplasmic reticulum FEBS Lett 401 104 108 9013867 355 O’Neal CJ Jobling MG Holmes RK Hol WG 2005 Structural basis for the activation of cholera toxin by human ARF6-GTP Science 309 1093 1096 16099990 356 Hug MJ Tamada T Bridges RJ 2003 CFTR and bicarbonate secretion by [correction of to] epithelial cells News Physiol Sci 18 38 42 12531931 357 Kunzelmann K Hübner M Schreiber R Levy-Holzman R Garty H Bleich M Warth R Slavik M von Hahn T Greger R 2001 Cloning and function of the rat colonic epithelial K+ channel KVLQT1 J Membr Biol 179 155 164 11220365 358 Nataro JP Kaper JB 1998 Diarrheagenic Escherichia coli Clin Microbiol Rev 11 142 201 9457432 359 Mourad FH O’Donnell LJ Dias JA Ogutu E Andre EA Turvill JL Farthing MJ 1995 Role of 5-hydroxytryptamine type 3 receptors in rat intestinal fluid and electrolyte secretion induced by cholera and Escherichia coli enterotoxins Gut 37 340 345 7590428 360 De Haan L Hirst TR 2004 Cholera toxin: a paradigm for multi-functional engagement of cellular mechanisms (Review) Mol Membr Biol 21 77 92 15204437 361 Berkes J Viswanathan VK Savkovic SD Hecht G 2003 Intestinal epithelial responses to enteric pathogens: effects on the tight junction barrier, ion transport, and inflammation Gut 52 439 451 12584232 362 Kreisberg RB Harper J Strauman MC Marohn M Clements JD Nataro JP 2011 Induction of increased permeability of polarized enterocyte monolayers by enterotoxigenic Escherichia coli heat-labile enterotoxin Am J Trop Med Hyg 84 451 455 21363985 363 Burgess MN Bywater RJ Cowley CM Mullan NA Newsome PM 1978 Biological evaluation of a methanol-soluble, heat-stable Escherichia coli enterotoxin in infant mice, pigs, rabbits, and calves Infect Immun 21 526 531 357288 364 Smith HW Gyles CL 1970 The effect of cell-free fluids prepared from cultures of human and animal enteropathogenic strains of Escherichia coli on ligated intestinal segments of rabbits and pigs J Med Microbiol 3 403 409 4919580 365 Nair GB Takeda Y 1998 The heat-stable enterotoxins Microb Pathog 24 123 131 9480795 366 Staples SJ Asher SE Giannella RA 1980 Purification and characterization of heat-stable enterotoxin produced by a strain of E. coli pathogenic for man J Biol Chem 255 4716 4721 6989817 367 Weiglmeier PR Rösch P Berkner H 2010 Cure and curse: E. coli heat-stable enterotoxin and its receptor guanylyl cyclase C Toxins (Basel) 2 2213 2229 22069681 368 Lima AA Fonteles MC 2014 From Escherichia coli heat-stable enterotoxin to mammalian endogenous guanylin hormones Braz J Med Biol Res 47 179 191 24652326 369 Okamoto K Okamoto K Yukitake J Kawamoto Y Miyama A 1987 Substitutions of cysteine residues of Escherichia coli heat-stable enterotoxin by oligonucleotide-directed mutagenesis Infect Immun 55 2121 2125 3305364 370 Takeda T Takeda Y Aimoto S Takao T Ikemura H Shimonishi Y Miwatani T 1983 Neutralization of activity of two different heat-stable enterotoxins (STh and STp) of enterotoxigenic Escherichia coli by homologous and heterologous antisera FEMS Microbiol Lett 20 357 359 371 Dreyfus LA Frantz JC Robertson DC 1983 Chemical properties of heat-stable enterotoxins produced by enterotoxigenic Escherichia coli of different host origins Infect Immun 42 539 548 6358024 372 Harnett NM Gyles CL 1985 Enterotoxin plasmids in bovine and porcine enterotoxigenic Escherichia coli of O groups 9, 20, 64 and 101 Can J Comp Med 49 79 87 3886109 373 So M McCarthy BJ 1980 Nucleotide sequence of the bacterial transposon Tn1681 encoding a heat-stable (ST) toxin and its identification in enterotoxigenic Escherichia coli strains Proc Natl Acad Sci USA 77 4011 4015 6254008 374 Sekizaki T Akashi H Terakado N 1985 Nucleotide sequences of the genes for Escherichia coli heat-stable enterotoxin I of bovine, avian, and porcine origins Am J Vet Res 46 909 912 2990268 375 Alderete JF Robertson DC 1977b Repression of heat-stable enterotoxin synthesis in enterotoxigenic Escherichia coli Infect Immun 17 629 633 20404 376 Stieglitz H Cervantes L Robledo R Fonseca R Covarrubias L Bolivar F Kupersztoch YM 1988 Cloni ng, sequencing, and expression in Ficoll-generated minicells of an Escherichia coli heat-stable enterotoxin gene Plasmid 20 42 53 3071819 377 Okamoto K Takahara M 1990 Synthesis of Escherichia coli heat-stable enterotoxin STp as a pre-pro form and role of the pro sequence in secretion J Bacteriol 172 5260 5265 2203746 378 Rasheed JK Guzmán-Verduzco LM Kupersztoch YM 1990 Two precursors of the heat-stable enterotoxin of Escherichia coli: evidence of extracellular processing Mol Microbiol 4 265 273 2187146 379 Yamanaka H Fuke Y Hitotsubashi S Fujii Y Okamoto K 1993 Functional properties of pro region of Escherichia coli heat-stable enterotoxin Microbiol Immunol 37 195 205 7686611 380 Sommerfelt H Haukanes BI Kalland KH Svennerholm AM Sanchéz J Bjorvatn B 1989 Mechanism of spontaneous loss of heat-stable toxin (STa) production in enterotoxigenic Escherichia coli APMIS 97 436 440 2567174 381 Okamoto K Baba T Yamanaka H Akashi N Fujii Y 1995 Disulfide bond formation and secretion of Escherichia coli heat-stable enterotoxin II J Bacteriol 177 4579 4586 7642482 382 Sanchez J Solorzano RM Holmgren J 1993 Extracellular secretion of STa heat-stable enterotoxin by Escherichia coli after fusion to a heterologous leader peptide FEBS Lett 330 265 269 8375497 383 Yang Y Gao Z Guzmán-Verduzco LM Tachias K Kupersztoch YM 1992 Secretion of the STA3 heat-stable enterotoxin of Escherichia coli: extracellular delivery of Pro-STA is accomplished by either Pro or STA Mol Microbiol 6 3521 3529 1474896 384 Yamanaka H Kameyama M Baba T Fujii Y Okamoto K 1994 Maturation pathway of Escherichia coli heat-stable enterotoxin I: requirement of DsbA for disulfide bond formation J Bacteriol 176 2906 2913 8188592 385 Yamanaka H Nomura T Fujii Y Okamoto K 1998a Need for TolC, an Escherichia coli outer membrane protein, in the secretion of heat-stable enterotoxin I across the outer membrane Microb Pathog 25 111 120 9790870 386 Yamanaka H Nomura T Fujii Y Okamoto K 1997 Extracellular secretion of Escherichia coli heat-stable enterotoxin I across the outer membrane J Bacteriol 179 3383 3390 9171378 387 Yamasaki S Hidaka Y Ito H Takeda Y Shimonishi Y 1988 Structure requirements for the spatial structure and toxicity of heat-stable enterotoxin (STh) of enterotoxigenic Escherichia coli Bull Chem Soc Jpn 61 1701 1706 388 Hidaka Y Ohmori K Wada A Ozaki H Ito H Hirayama T Takeda Y Shimonishi Y 1991 Synthesis and biological properties of carba-analogs of heat-stable enterotoxin (ST) produced by enterotoxigenic Escherichia coli Biochem Biophys Res Commun 176 958 965 1645548 389 Ozaki H Sato T Kubota H Hata Y Katsube Y Shimonishi Y 1991 Molecular structure of the toxin domain of heat-stable enterotoxin produced by a pathogenic strain of Escherichia coli. A putative binding site for a binding protein on rat intestinal epithelial cell membranes J Biol Chem 266 5934 5941 2005130 390 Gariépy J Judd AK Schoolnik GK 1987 Importance of disulfide bridges in the structure and activity of Escherichia coli enterotoxin ST1b Proc Natl Acad Sci USA 84 8907 8911 2827159 391 Gariépy J Lane A Frayman F Wilbur D Robien W Schoolnik GK Jardetzky O 1986 Structure of the toxic domain of the Escherichia coli heat-stable enterotoxin ST I Biochemistry 25 7854 7866 3801445 392 Carpick BW Gariépy J 1991 Structural characterization of functionally important regions of the Escherichia coli heat-stable enterotoxin STIb Biochemistry 30 4803 4809 2029521 393 Waldman SA O’Hanley P 1989 Influence of a glycine or proline substitution on the functional properties of a 14-amino-acid analog of Escherichia coli heat-stable enterotoxin Infect Immun 57 2420 2424 2568345 394 Yoshimura S Ikemura H Watanabe H Aimoto S Shimonishi Y Hara S Takeda T Miwatani T Takeda Y 1985 Essential structure for full enterotoxigenic activity of heat-stable enterotoxin produced by enterotoxigenic Escherichia coli FEBS Lett 181 138 142 3972100 395 Arshad N Visweswariah SS 2012 The multiple and enigmatic roles of guanylyl cyclase C in intestinal homeostasis FEBS Lett 586 2835 2840 22819815 396 Hasegawa M Shimonishi Y 2005 Recognition and signal transduction mechanism of Escherichia coli heat-stable enterotoxin and its receptor, guanylate cyclase C J Pept Res 65 261 271 15705168 397 Steinbrecher KA 2014 The multiple roles of guanylate cyclase C, a heat stable enterotoxin receptor Curr Opin Gastroenterol 30 1 6 24304979 398 Carpick BW Gariépy J 1993 The Escherichia coli heat-stable enterotoxin is a long-lived superagonist of guanylin Infect Immun 61 4710 4715 8104900 399 Forte LR Eber SL Turner JT Freeman RH Fok KF Currie MG 1993 Guanylin stimulation of Cl− secretion in human intestinal T84 cells via cyclic guanosine monophosphate J Clin Invest 91 2423 2428 8390480 400 Krause WJ Cullingford GL Freeman RH Eber SL Richardson KC Fok KF Currie MG Forte LR 1994 Distribution of heat-stable enterotoxin/guanylin receptors in the intestinal tract of man and other mammals J Anat 184 407 417 8014132 401 Cohen MB Mann EA Lau C Henning SJ Giannella RA 1992 A gradient in expression of the Escherichia coli heat-stable enterotoxin receptor exists along the villus-to-crypt axis of rat small intestine Biochem Biophys Res Commun 186 483 490 1378729 402 Sellers ZM Mann E Smith A Ko KH Giannella R Cohen MB Barrett KE Dong H 2008 Heat-stable enterotoxin of Escherichia coli (STa) can stimulate duodenal HCO3(−) secretion via a novel GC-C- and CFTR-independent pathway FASEB J 22 1306 1316 18096816 403 Hitotsubashi S Fujii Y Yamanaka H Okamoto K 1992b Some properties of purified Escherichia coli heat-stable enterotoxin II Infect Immun 60 4468 4474 1398961 404 Akabas MH 2000 Cystic fibrosis transmembrane conductance regulator. Structure and function of an epithelial chloride channel J Biol Chem 275 3729 3732 10660517 405 Guttman JA Finlay BB 2009 Tight junctions as targets of infectious agents Biochim Biophys Acta 1788 832 841 19059200 406 Suzuki T 2013 Regulation of intestinal epithelial permeability by tight junctions Cell Mol Life Sci 70 631 659 22782113 407 Nakashima R Kamata Y Nishikawa Y 2013 Effects of Escherichia coli heat-stable enterotoxin and guanylin on the barrier integrity of intestinal epithelial T84 cells Vet Immunol Immunopathol 152 78 81 23078906 408 Levine MM Prado V Robins-Browne R Lior H Kaper JB Moseley SL Gicquelais K Nataro JP Vial P Tall B 1988 Use of DNA probes and HEp-2 cell adherence assay to detect diarrheagenic Escherichia coli J Infect Dis 158 224 228 3292660 409 Paiva de Sousa C Dubreuil JD 2001 Distribution and expression of the astA gene (EAST1 toxin) in Escherichia coli and Salmonella Int J Med Microbiol 291 15 20 11403406 410 Savarino SJ McVeigh A Watson J Cravioto A Molina J Echeverria P Bhan MK Levine MM Fasano A 1996 Enteroaggregative Escherichia coli heat-stable enterotoxin is not restricted to enteroaggregative E. coli J Infect Dis 173 1019 1022 8603943 411 Osek J 2003 Detection of the enteroaggregative Escherichia coli heat-stable enterotoxin 1 (EAST1) gene and its relationship with fimbrial and enterotoxin markers in E. coli isolates from pigs with diarrhoea Vet Microbiol 91 65 72 12441232 412 Noamani BN Fairbrother JM Gyles CL 2003 Virulence genes of O149 enterotoxigenic Escherichia coli from outbreaks of postweaning diarrhea in pigs Vet Microbiol 97 87 101 14637041 413 Frydendahl K 2002 Prevalence of serogroups and virulence genes in Escherichia coli associated with postweaning diarrhoea and edema disease in pigs and a comparison of diagnostic approaches Vet Microbiol 85 169 182 11844623 414 Liu W Yuan C Meng X Du Y Gao R Tang J Shi D 2014 Frequency of virulence factors in Escherichia coli isolated from suckling pigs with diarrhoea in China Vet J 199 286 289 24378293 415 Toledo A Gómez D Cruz C Carreón R López J Giono S Castro AM 2012 Prevalence of virulence genes in Escherichia coli strains isolated from piglets in the suckling and weaning period in Mexico J Med Microbiol 61 148 156 21852524 416 Zajacova ZS Konstantinova L Alexa P 2012 Detection of virulence factors of Escherichia coli focused on prevalence of EAST1 toxin in stool of diarrheic and non-diarrheic piglets and presence of adhesion involving virulence factors in astA positive strains Vet Microbiol 154 369 375 21864997 417 Tzipori S Montanaro J Robins-Browne RM Vial P Gibson R Levine MM 1992 Studies with enteroaggregative Escherichia coli in the gnotobiotic piglet gastroenteritis model Infect Immun 60 5302 5306 1452364 418 Yamamoto T Echeverria P 1996 Detection of the enteroaggregative Escherichia coli heat-stable enterotoxin 1 gene sequences in enterotoxigenic E. coli strains pathogenic for humans Infect Immun 64 1441 1445 8606115 419 Bertin Y Martin C Girardeau JP Pohl P Contrepois M 1998 Association of genes encoding P fimbriae, CS31A antigen and EAST 1 toxin among CNF1-producing Escherichia coli strains from cattle with septicemia and diarrhea FEMS Microbiol Lett 162 235 239 9627958 420 Choi C Cho W Chung H Jung T Kim J Chae C 2001a Prevalence of the enteroaggregative Escherichia coli heat-stable enterotoxin 1 (EAST1) gene in isolates in weaned pigs with diarrhea and/or edema disease Vet Microbiol 81 65 71 11356319 421 Choi C Kwon D Chae C 2001b Prevalence of the enteroaggregative Escherichia coli heat-stable enterotoxin 1 gene and its relationship with fimbrial and enterotoxin genes in E. coli isolated from diarrheic piglets J Vet Diagn Invest 13 26 29 11243358 422 Ménard LP Lussier JG Lépine F Paiva de Sousa C Dubreuil JD 2004 Expression, purification, and biochemical characterization of enteroaggregative Escherichia coli heat-stable enterotoxin 1 Protein Expr Purif 33 223 231 14711510 423 Savarino SJ Fasano A Watson J Martin BM Levine MM Guandalini S Guerry P 1993 Enteroaggrega tive Escherichia coli heat-stable enterotoxin 1 represents another subfamily of E. coli heat-stable toxin Proc Natl Acad Sci USA 90 3093 3097 8385356 424 Nataro JP Deng Y Cookson S Cravioto A Savarino SJ Guers LD Levine MM Tacket CO 1995 Heter ogeneity of enteroaggregative Escherichia coli virulence demonstrated in volunteers J Infect Dis 171 465 468 7844392 425 Yamamoto T Wakisaka N Sato F Kato A 1997b Comparison of the nucleotide sequence of enteroaggregative Escherichia coli heat-stable enterotoxin 1 genes among diarrhea-associated Escherichia coli FEMS Microbiol Lett 147 89 95 9037769 426 Lai XH Wang SY Uhlin BE 1999 Expression of cytotoxicity by potential pathogens in the standard Escherichia coli collection of reference (ECOR) strains Microbiology 145 3295 3303 10589739 427 Veilleux S Dubreuil JD 2006 Presence of Escherichia coli carrying the EAST1 toxin gene in farm animals Vet Res 37 3 13 16336921 428 McVeigh A Fasano A Scott DA Jelacic S Moseley SL Robertson DC Savarino SJ 2000 IS1414, an Escherichia coli insertion sequence with a heat-stable enterotoxin gene embedded in a transposase-like gene Infect Immun 68 5710 5715 10992475 429 Yamamoto T Nakazawa M 1997 Detection and sequences of the enteroaggregative Escherichia coli heat-stable enterotoxin 1 gene in enterotoxigenic E. coli strains isolated from piglets and calves with diarrhea J Clin Microbiol 35 223 227 8968912 430 Ruan X Crupper SS Schultz BD Robertson DC Zhang W 2012 Escherichia coli expressing EAST1 toxin did not cause an increase of cAMP or cGMP levels in cells, and no diarrhea in 5-day old gnotobiotic pigs PLoS One 7 e43203 22905235 431 Zajacova ZS Faldyna M Kulich P Kummer V Maskova J Alexa P 2013 Experimental infection of gnotobiotic piglets with Escherichia coli strains positive for EAST1 and AIDA Vet Immunol Immunopathol 152 176 182 23068274 432 Lortie LA Dubreuil JD Harel J 1991 Characterization of Escherichia coli strains producing heat-stable enterotoxin b (STb) isolated from humans with diarrhea J Clin Microbiol 29 656 659 2037689 433 Handl CE Flock JI 1992 STb producing Escherichia coli are rarely associated with infantile diarrhoea J Diarrhoeal Dis Res 10 37 38 1619241 434 Whipp SC Moon HW Argenzio RA 1981 Comparison of enterotoxic activities of heat-stable enterotoxins from class 1 and class 2 Escherichia coli of swine origin Infect Immun 31 245 251 7011991 435 Loos M Geens M Schauvliege S Gasthuys F van der Meulen J Dubreuil JD Goddeeris BM Niewold T Cox E 2012 Role of heat-stable enterotoxins in the induction of early immune responses in piglets after infection with enterotoxigenic Escherichia coli PLoS One 7 e41041 22815904 436 Spitzer F Vahjen W Pieper R Martinez-Vallespin B Zentek J 2014 A standardised challenge model with an enterotoxigenic F4+ Escherichia coli strain in piglets assessing clinical traits and faecal shedding of fae and est-II toxin genes Arch Anim Nutr 68 448 459 25313936 437 Dubreuil JD 1997 Escherichia coli STb enterotoxin Microbiology 143 1783 1795 9202453 438 Lee CH Moseley SL Moon HW Whipp SC Gyles CL So M 1983 Characterization of the gene encoding heat-stable toxin II and preliminary molecular epidemiological studies of enterotoxigenic Escherichia coli heat-stable toxin II producers Infect Immun 42 264 268 6352495 439 Picken RN Mazaitis AJ Maas WK Rey M Heyneker H 1983 Nucleotide sequence of the gene for heat-stable enterotoxin II of Escherichia coli Infect Immun 42 269 275 6352496 440 Sukumar M Rizo J Wall M Dreyfus LA Kupersztoch YM Gierasch LM 1995 The structure of Escherichia coli heat-stable enterotoxin b by nuclear magnetic resonance and circular dichroism Protein Sci 4 1718 1729 8528070 441 Taillon C Nadeau E Mourez M Dubreuil JD 2008 Heterogeneity of Escherichia coli STb enterotoxin isolated from diseased pigs J Med Microbiol 57 887 890 18566148 442 Taillon C Hancock MA Mourez M Dubreuil JD 2012 Biochemical and biological characterization of Escherichia coli STb His12 to Asn variant Toxicon 59 300 305 22155058 443 Handl CE Harel J Flock JI Dubreuil JD 1993 High yield of active STb enterotoxin from a fusion protein (MBP-STb) expressed in Escherichia coli Protein Expr Purif 4 275 281 8374296 444 Dubreuil JD Fairbrother JM Lallier R Larivière S 1991 Production and purification of heat-stable enterotoxin b from a porcine Escherichia coli strain Infect Immun 59 198 203 1987033 445 Fujii Y Hayashi M Hitotsubashi S Fuke Y Yamanaka H Okamoto K 1991 Purification and characterization of Escherichia coli heat-stable enterotoxin II J Bacteriol 173 5516 5522 1885528 446 Whipp SC 1987 Protease degradation of Escherichia coli heat-stable, mouse-negative, pig-positive enterotoxin Infect Immun 55 2057 2060 3305361 447 Whipp SC 1991 Intestinal responses to enterotoxigenic Escherichia coli heat-stable toxin b in non-porcine species Am J Vet Res 52 734 737 1854098 448 Dubreuil JD Letellier A Harel J 1996 A recombinant Escherichia coli heat-stable enterotoxin b (STb) fusion protein eliciting neutralizing antibodies FEMS Immunol Med Microbiol 13 317 323 8739196 449 You J Xu Y He M McAllister TA Thacker PA Li X Wang T Jin L 2011 Protection of mice against enterotoxigenic E. coli by immunization with a polyvalent enterotoxin comprising a combination of LTB, STa, and STb Appl Microbiol Biotechnol 89 1885 1893 21085949 450 Zhang W Francis DH 2010 Genetic fusions of heat-labile toxoid (LT) and heat-stable toxin b (STb) of porcine enterotoxigenic Escherichia coli elicit protective anti-LT and anti-STb antibodies Clin Vaccine Immunol 17 1223 1231 20505006 451 Zhang W Zhang C Francis DH Fang Y Knudsen D Nataro JP Robertson DC 2010 Genetic fusions of heat-labile (LT) and heat-stable (ST) toxoids of porcine enterotoxigenic Escherichia coli elicit neutralizing anti-LT and anti-STa antibodies Infect Immun 78 316 325 19858307 452 Hitotsubashi S Akagi M Saitou A Yamanaka H Fujii Y Okamoto K 1992a Action of Escherichia coli heat-stable enterotoxin II on isolated sections of mouse ileum FEMS Microbiol Lett 69 249 252 1555760 453 Echeverria P Seriwatana J Taylor DN Tirapat C Rowe B 1985 Escherichia coli contains plasmids coding for heat-stable b, other enterotoxins, and antibiotic resistance Infect Immun 48 843 846 3888846 454 Hu ST Lee CH 1988 Characterization of the transposon carrying the STII gene of enterotoxigenic Escherichia coli Mol Gen Genet 214 490 495 2851097 455 Hu ST Yang MK Spandau DF Lee CH 1987 Characterization of the terminal sequences flanking the transposon that carries the Escherichia coli enterotoxin STII gene Gene 55 157 167 2822540 456 Lee CH Hu ST Swiatek PJ Moseley SL Allen SD So M 1985 Isolation of a novel transposon which carries the Escherichia coli enterotoxin STII gene J Bacteriol 162 615 620 2985539 457 Spandau DF Lee CH 1987 Determination of the promoter strength of the gene encoding Escherichia coli heat-stable enterotoxin II J Bacteriol 169 1740 1744 3549701 458 Erume J Berberov EM Moxley RA 2010 Comparison of the effects of different nutrient media on production of heat-stable enterotoxin-b by Escherichia coli Vet Microbiol 144 160 165 20097018 459 Busque P Letellier A Harel J Dubreuil JD 1995 Production of Escherichia coli STb enterotoxin is subject to catabolite repression Microbiology 141 1621 1627 7551030 460 Kupersztoch YM Tachias K Moomaw CR Dreyfus LA Urban R Slaughter C Whipp S 1990 Secretio n of methanol-insoluble heat-stable enterotoxin (STB): energy- and secA-dependent conversion of pre-STB to an intermediate indistinguishable from the extracellular toxin J Bacteriol 172 2427 2432 2158970 461 Foreman DT Martinez Y Coombs G Torres A Kupersztoch YM 1995 TolC and DsbA are needed for the secretion of STB, a heat-stable enterotoxin of Escherichia coli Mol Microbiol 18 237 245 8709843 462 Arriaga YL Harville BA Dreyfus LA 1995 Contribution of individual disulfide bonds to biological action of Escherichia coli heat-stable enterotoxin B Infect Immun 63 4715 4720 7591127 463 Kobayashi N Nishino K Yamaguchi A 2001 Novel macrolide-specific ABC-type efflux transporter in Escherichia coli J Bacteriol 183 5639 5644 11544226 464 Kobayashi N Nishino K Hirata T Yamaguchi A 2003 Membrane topology of ABC-type macrolide antibiotic exporter MacB in Escherichia coli FEBS Lett 546 241 246 12832048 465 Yamanaka H Tadokoro S Miyano M Takahashi E Kobayashi H Okamoto K 2007 Studies on the region involved in the transport activity of Escherichia coli TolC by chimeric protein analysis Microb Pathog 42 184 192 17350794 466 Yamanaka H Kobayashi H Takahashi E Okamoto K 2008 MacAB is involved in the secretion of Escherichia coli heat-stable enterotoxin II J Bacteriol 190 7693 7698 18805970 467 Tikhonova EB Devroy VK Lau SY Zgurskaya HI 2007 Reconstitution of the Escherichia coli macrolide transporter: the periplasmic membrane fusion protein MacA stimulates the ATPase activity of MacB Mol Microbiol 63 895 910 17214741 468 Okamoto K Yamanaka H Takeji M Fuji Y 2001 Region of heat-stable enterotoxin II of Escherichia coli involved in translocation across the outer membrane Microbiol Immunol 45 349 355 11471822 469 Labrie V Harel J Dubreuil JD 2001c Oligomerization of Escherichia coli enterotoxin b through its C-terminal hydrophobic alpha-helix Biochim Biophys Acta 1535 128 133 11342001 470 Rousset E Harel J Dubreuil JD 1998a Sulfatide from the pig jejunum brush border epithelial cell surface is involved in binding of Escherichia coli enterotoxin b Infect Immun 66 5650 5658 9826338 471 Chao KL Dreyfus LA 1997 Interaction of Escherichia coli heat-stable enterotoxin B with cultured human intestinal epithelial cells Infect Immun 65 3209 3217 9234777 472 Rousset E Harel J Dubreuil JD 1998b Binding characteristics of Escherichia coli enterotoxin b (STb) to the pig jejunum and partial characterization of the molecule involved Microb Pathog 24 277 288 9600860 473 Beausoleil HE Labrie V Dubreuil JD 2002a Trypan blue uptake by chinese hamster ovary cultured epithelial cells: a cellular model to study Escherichia coli STb enterotoxin Toxicon 40 185 191 11689240 474 Gonçalves C Berthiaume F Mourez M Dubreuil JD 2008 Escherichia coli STb toxin binding to sulfatide and its inhibition by carragenan FEMS Microbiol Lett 281 30 35 18279334 475 Labrie V Beausoleil HE Harel J Dubreuil JD 2001b Binding to sulfatide and enterotoxicity of various Escherichia coli STb mutants Microbiology 147 3141 3148 11700365 476 Fujii Y Okamuro Y Hitotsubashi S Saito A Akashi N Okamoto K 1994 Effect of alterations of basic amino acid residues of Escherichia coli heat-stable enterotoxin II on enterotoxicity Infect Immun 62 2295 2301 8188351 477 Dubreuil JD 2008 Escherichia coli STb toxin and colibacillosis: knowing is half the battle FEMS Microbiol Lett 278 137 145 17995951 478 Argenzio RA Liacos J Berschneider HM Whipp SC Robertson DC 1984 Effect of heat-stable enterotoxin of Escherichia coli and theophylline on ion transport in porcine small intestine Can J Comp Med 48 14 22 6538805 479 Weikel CS Nellans HN Guerrant RL 1986a In vivo and in vitro effects of a novel enterotoxin, STb, produced by Escherichia coli J Infect Dis 153 893 901 2422295 480 Fujii Y Kondo Y Okamoto K 1995 Involvement of prostaglandin E2 synthesis in the intestinal secretory action of Escherichia coli heat-stable enterotoxin II FEMS Microbiol Lett 130 259 265 7649449 481 Labrie V Harel J Dubreuil JD 2002 Escherichia coli heat-stable enterotoxin b (STb) in vivo internalization within rat intestinal epithelial cells Vet Res 33 223 228 11944810 482 Gonçalves C Dubreuil JD 2009 Effect of Escherichia coli STb toxin on NIH-3T3 cells FEMS Immunol Med Microbiol 55 432 441 19222570 483 Dreyfus LA Harville B Howard DE Shaban R Beatty DM Morris SJ 1993 Calcium influx mediated by the Escherichia coli heat-stable enterotoxin B (STB) Proc Natl Acad Sci USA 90 3202 3206 8475060 484 Fujii Y Nomura T Yamanaka H Okamoto K 1997 Involvement of Ca(2+)-calmodulin-dependent protein kinase II in the intestinal secretory action of Escherichia coli heat-stable enterotoxin II Microbiol Immunol 41 633 636 9310945 485 Harville BA Dreyfus LA 1995 Involvement of 5-hydroxytryptamine and prostaglandin E2 in the intestinal secretory action of Escherichia coli heat-stable enterotoxin B Infect Immun 63 745 750 7868242 486 Harville BA Dreyfus LA 1996 Release of serotonin from RBL-2H3 cells by the Escherichia coli peptide toxin STb Peptides 17 363 366 8735960 487 Peterson JW Whipp SC 1995 Comparison of the mechanisms of action of cholera toxin and the heat-stable enterotoxins of Escherichia coli Infect Immun 63 1452 1461 7890409 488 Gonçalves C Vachon V Schwartz JL Dubreuil JD 2007 The Escherichia coli enterotoxin STb permeabilizes piglet jejunal brush border membrane vesicles Infect Immun 75 2208 2213 17307947 489 Labrie V Potvin L Harel J Dubreuil JD Schwartz J-L Enterotoxin b of Escherichia coli (STb) forms ion channels in planar lipid bilayers A54 In (ed) 490 Syed HC Dubreuil JD 2012 Escherichia coli STb toxin induces apoptosis in intestinal epithelial cell lines Microb Pathog 53 147 153 22771838 491 Ngendahayo Mukiza C Dubreuil JD 2013 Escherichia coli heat-stable toxin b impairs intestinal epithelial barrier function by altering tight junction proteins Infect Immun 81 2819 2827 23716609 492 Popoff MR 2011 Multifaceted interactions of bacterial toxins with the gastrointestinal mucosa Future Microbiol 6 763 797 21797691 493 Schmidt E Kelly SM van der Walle CF 2007 Tight junction modulation and biochemical characterisation of the zonula occludens toxin C-and N-termini FEBS Lett 581 2974 2980 17553496 494 Nassour H Dubreuil JD 2014 Escherichia coli STb enterotoxin dislodges claudin-1 from epithelial tight junctions PLoS One 9 e113273 25409315 495 Wu XY Chapman T Trott DJ Bettelheim K Do TN Driesen S Walker MJ Chin J 2007 Comparative analysis of virulence genes, genetic diversity, and phylogeny of commensal and enterotoxigenic Escherichia coli isolates from weaned pigs Appl Environ Microbiol 73 83 91 17056683 496 Niewerth U Frey A Voss T Le Bouguénec C Baljer G Franke S Schmidt MA 2001 The AIDA autotransporter system is associated with F18 and stx2e in Escherichia coli isolates from pigs diagnosed with edema disease and postweaning diarrhea Clin Diagn Lab Immunol 8 143 149 11139209 497 Ngeleka M Pritchard J Appleyard G Middleton DM Fairbrother JM 2003 Isolation and association of Escherichia coli AIDA-I/STb, rather than EAST1 pathotype, with diarrhea in piglets and antibiotic sensitivity of isolates J Vet Diagn Invest 15 242 252 12735346 498 Tarr PI Bilge SS Vary JC Jr Jelacic S Habeeb RL Ward TR Baylor MR Besser TE 2000 Iha: a novel Escherichia coli O157:H7 adherence-conferring molecule encoded on a recently acquired chromosomal island of conserved structure Infect Immun 68 1400 1407 10678953 499 Fleckenstein JM Kopecko DJ Warren RL Elsinghorst EA 1996 Molecular characterization of the tia invasion locus from enterotoxigenic Escherichia coli Infect Immun 64 2256 2265 8675335 500 Elsinghorst EA Weitz JA 1994 Epithelial cell invasion and adherence directed by the enterotoxigenic Escherichia coli tib locus is associated with a 104-kilodalton outer membrane protein Infect Immun 62 3463 3471 8039917 501 Lindenthal C Elsinghorst EA 2001 Enterotoxigenic Escherichia coli TibA glycoprotein adheres to human intestine epithelial cells Infect Immun 69 52 57 11119488 502 Mammarappallil JG Elsinghorst EA 2000 Epithelial cell adherence mediated by the enterotoxigenic Escherichia coli tia protein Infect Immun 68 6595 6601 11083770 503 Gutiérrez D Pardo M Montero D Oñate A Farfán MJ Ruiz-Pérez F Del Canto F Vidal R 2015 TleA, a Tsh-like autotransporter identified in a human enterotoxigenic Escherichia coli strain Infect Immun 83 1893 1903 25712927 504 Kansal R Rasko DA Sahl JW Munson GP Roy K Luo Q Sheikh A Kuhne KJ Fleckenstein JM 2013 Transcriptional modulation of enterotoxigenic Escherichia coli virulence genes in response to epithelial cell interactions Infect Immun 81 259 270 23115039 505 Sheikh A Luo Q Roy K Shabaan S Kumar P Qadri F Fleckenstein JM 2014 Contribution of the highly conserved EaeH surface protein to enterotoxigenic Escherichia coli pathogenesis Infect Immun 82 3657 3666 24935979 506 Roy K Hilliard GM Hamilton DJ Luo J Ostmann MM Fleckenstein JM 2009 Enterotoxigenic Escherichia coli EtpA mediates adhesion between flagella and host cells Nature 457 594 598 19060885 507 Luo Q Kumar P Vickers TJ Sheikh A Lewis WG Rasko DA Sistrunk J Fleckenstein JM Payne SM 2014 Enterotoxigenic Escherichia coli secretes a highly conserved mucin-degrading metalloprotease to effectively engage intestinal epithelial cells Infect Immun 82 509 521 24478067 508 Kumar P Luo Q Vickers TJ Sheikh A Lewis WG Fleckenstein JM Payne SM 2014 EatA, an immunogenic protective antigen of enterotoxigenic Escherichia coli, degrades intestinal mucin Infect Immun 82 500 508 24478066 509 Shepard SM 2011 Phylogenetic and genomic characterization of porcine enterotoxigenic Escherichia coli M.S. University of Minnesota St. Paul, Minnesota 510 Wyrsch E Roy Chowdhury P Abraham S Santos J Darling AE Charles IG Chapman TA Djordjevic SP 2015 Comp arative genomic analysis of a multiple antimicrobial resistant enterotoxigenic E. coli O157 lineage from Australian pigs BMC Genomics 16 165 25888127 511 Luo Q Qadri F Kansal R Rasko DA Sheikh A Fleckenstein JM 2015 Conservation and immunogenicity of novel antigens in diverse isolates of enterotoxigenic Escherichia coli PLoS Negl Trop Dis 9 e0003446 25629897 512 Jones GW Rutter JM 1974 The association of K88 antigen with haemagglutinating activity in porcine strains of Escherichia coli J Gen Microbiol 84 135 144 4612097 513 Jones GW Rutter JM 1974 Contribution of the K88 antigen of Escherichia coli to enteropathogenicity; protection against disease by neutralizing the adhesive properties of K88 antigen Am J Clin Nutr 27 1441 1449 4611195 514 Isaacson RE 1978 K99 surface antigen of Escherichia coli: antigenic characterization Infect Immun 22 555 559 83300 515 Sherman DM Acres SD Sadowski PL Springer JA Bray B Raybould TJ Muscoplat CC 1983 Protecti on of calves against fatal enteric colibacillosis by orally administered Escherichia coli K99-specific monoclonal antibody Infect Immun 42 653 658 6358029 516 Snodgrass DR 1986 Evaluation of a combined rotavirus and enterotoxigenic Escherichia coli vaccine in cattle Vet Rec 119 39 42 3018989 517 Snodgrass DR Nagy LK Sherwood D Campbell I 1982 Passive immunity in calf diarrhea: vaccination with K99 antigen of enterotoxigenic Escherichia coli and rotavirus Infect Immun 37 586 591 6288567 518 Wittum TW Dewey CE 1996 Partial budget analysis of sow Escherichia coli vaccination Swine Health Prod 1 9 13 519 Ascón MA Ochoa-Repáraz J Walters N Pascual DW 2005 Partially assembled K99 fimbriae are required for protection Infect Immun 73 7274 7280 16239523 520 Bakker D van Zijderveld FG van der Veen S Oudega B de Graaf FK 1990 K88 fimbriae as carriers of heterologous antigenic determinants Microb Pathog 8 343 352 1699108 521 Chen H Schifferli DM 2003 Construction, characterization, and immunogenicity of an attenuated Salmonella enterica serovar typhimurium pgtE vaccine expressing fimbriae with integrated viral epitopes from the spiC promoter Infect Immun 71 4664 4673 12874347 522 Chen H Schifferli DM 2001 Enhanced immune responses to viral epitopes by combining macrophage-inducible expression with multimeric display on a Salmonella vector Vaccine 19 3009 3018 11282213 523 Chen H Schifferli DM 2000 Mucosal and systemic immune responses to chimeric fimbriae expressed by Salmonella enterica serovar typhimurium vaccine strains Infect Immun 68 3129 3139 10816454 524 Rani DB Bayer ME Schifferli DM 1999 Polymeric display of immunogenic epitopes from herpes simplex virus and transmissible gastroenteritis virus surface proteins on an enteroadherent fimbria Clin Diagn Lab Immunol 6 30 40 9874660 525 Thiry G Clippe A Scarcez T Petre J 1989 Cloning of DNA sequences encoding foreign peptides and their expression in the K88 pili Appl Environ Microbiol 55 984 993 2471451 526 Ascón MA Hone DM Walters N Pascual DW 1998 Oral immunization with a Salmonella typhimurium vaccine vector expressing recombinant enterotoxigenic Escherichia coli K99 fimbriae elicits elevated antibody titers for protective immunity Infect Immun 66 5470 5476 9784559 527 Attridge S Hackett J Morona R Whyte P 1988 Towards a live oral vaccine against enterotoxigenic Escherichia coli of swine Vaccine 6 387 389 3057757 528 Greenwood PE Clark SJ Cahill AD Trevallyn-Jones J Tzipori S 1988 Development and protective efficacy of a recombinant-DNA derived fimbrial vaccine against enterotoxic colibacillosis in neonatal piglets Vaccine 6 389 392 3057758 529 Moon HW Rogers DG Rose R 1988 Effects of an orally administered live Escherichia coli pilus vaccine on duration of lacteal immunity to enterotoxigenic Escherichia coli in swine Am J Vet Res 49 2068 2071 2907273 530 Morona R Morona JK Considine A Hackett JA van den Bosch L Beyer L Attridge SR 1994 Construction of K88- and K99-expressing clones of Salmonella typhimurium G30: immunogenicity following oral administration to pigs Vaccine 12 513 517 7913567 531 Bijlsma IG van Houten M Frik JF Ruitenberg EJ 1987 K88 variants K88ab, K88ac and K88ad in oral vaccination of different porcine adhesive phenotypes. Immunological aspects Vet Immunol Immunopathol 16 235 250 3324461 532 Van den Broeck W Cox E Goddeeris BM 1999 Induction of immune responses in pigs following oral administration of purified F4 fimbriae Vaccine 17 2020 2029 10217602 533 Chu H Kang S Ha S Cho K Park SM Han KH Kang SK Lee H Han SH Yun CH Choi Y 2005 Lact obacillus acidophilus expressing recombinant K99 adhesive fimbriae has an inhibitory effect on adhesion of enterotoxigenic Escherichia coli Microbiol Immunol 49 941 948 16301804 534 Joensuu JJ Kotiaho M Teeri TH Valmu L Nuutila AM Oksman-Caldentey KM Niklander-Teeri V 2006 Glycosylated F4 (K88) fimbrial adhesin FaeG expressed in barley endosperm induces ETEC-neutralizing antibodies in mice Transgenic Res 15 359 373 16779651 535 Joensuu JJ Verdonck F Ehrstrom A Peltola M Siljander-Rasi H Nuutila AM Oksman-Caldentey KM Teeri TH Cox E Goddeeris BM Niklander-Teeri V 2005 F4 (K88) fimbrial adhesin FaeG expressed in alfalfa reduces F4+ enterotoxigenic Escherichia coli excretion in weaned piglets Vaccine 536 Sherman DM Acres SD Sadowski PL Springer JA Bray B Raybould TJG Muscoplat CC 1983 Protec tion of calves against fatal enteric colibacillosis by orally administered Escherichia coli K99-specific monoclonal antibody Infect Immun 42 653 658 6358029 537 Acres SD Isaacson RE Khachatourians G Babiuk L Kapitany RA Vaccination of cows with purified K99 antigen, K99+ anucleated live E coli, and whole cell bacterins containing enterotoxigenic E coli for prevention of enterotoxigenic colibacillosis of calves 443 456 Acres SD Veterinary Infectious Disease Organization 538 Bertschinger HU Nief V Tschäpe H 2000 Active oral immunization of suckling piglets to prevent colonization after weaning by enterotoxigenic Escherichia coli with fimbriae F18 Vet Microbiol 71 255 267 10703708 539 Niewold TA van Dijk AJ Geenen PL Roodink H Margry R van der Meulen J 2007 Dietary specific antibodies in spray-dried immune plasma prevent enterotoxigenic Escherichia coli F4 (ETEC) post weaning diarrhoea in piglets Vet Microbiol 124 362 369 17524575 540 Moonens K De Kerpel M Coddens A Cox E Pardon E Remaut H De Greve H 2014 Nanobody mediated inhibition of attachment of F18 Fimbriae expressing Escherichia coli PLoS One 9 e114691 25502211 541 Moonens K Van den Broeck I Okello E Pardon E De Kerpel M Remaut H De Greve H 2015 Structural insight in the inhibition of adherence of F4 fimbriae producing enterotoxigenic Escherichia coli by llama single domain antibodies Vet Res (Faisalabad) 46 14 542 Sarrazin E Bertschinger HU 1997 Role of fimbriae F18 for actively acquired immunity against porcine enterotoxigenic Escherichia coli Vet Microbiol 54 133 144 9057257 543 Verdonck F Cox E van Gog K Van der Stede Y Duchateau L Deprez P Goddeeris BM 2002 Different kinetic of antibody responses following infection of newly weaned pigs with an F4 enterotoxigenic Escherichia coli strain or an F18 verotoxigenic Escherichia coli strain Vaccine 20 2995 3004 12126913 544 Felder CB Vorlaender N Gander B Merkle HP Bertschinger HU 2000 Microencapsulated enterotoxigenic Escherichia coli and detached fimbriae for peroral vaccination of pigs Vaccine 19 706 715 11115691 545 Snoeck V Huyghebaert N Cox E Vermeire A Vancaeneghem S Remon JP Goddeeris BM 2003 Enteri c-coated pellets of F4 fimbriae for oral vaccination of suckling piglets against enterotoxigenic Escherichia coli infections Vet Immunol Immunopathol 96 219 227 14592734 546 Alexa P Salajka E Salajková Z Máchová A 1995 Combined parenteral and oral immunization against enterotoxigenic Escherichia coli diarrhea in weaned piglets Vet Med (Praha) 40 365 370 8659089 547 Van der Stede Y Cox E Verdonck F Vancaeneghem S Goddeeris BM 2003 Reduced faecal excretion of F4+-E coli by the intramuscular immunisation of suckling piglets by the addition of 1alpha,25-dihydroxyvitamin D3 or CpG-oligodeoxynucleotides Vaccine 21 1023 1032 12547616 548 Hashish EA Zhang C Ruan X Knudsen DE Chase CC Isaacson RE Zhou G Zhang W 2013 A multiepitope fusion antigen elicits neutralizing antibodies against enterotoxigenic Escherichia coli and homologous bovine viral diarrhea virus in vitro Clin Vaccine Immunol 20 1076 1083 23697572 549 Ruan X Liu M Casey T Zhang W 2011 A Tripartite Fusion, FaeG-FedF-LT192A2:B, of Enterotoxigenic Escherichia coli (ETEC) Elicits Antibodies Neutralizing CT Toxin, Inhibiting Adherence of K88 (F4) and F18 Fimbriae, and Protecting Pigs against K88ac/LT ETEC Infection Clin Vaccine Immunol 550 Zhang C Zhang W 2010 Escherichia coli K88ac fimbriae expressing heat-labile (LT) and heat-stable (STa) toxin epitopes elicit antibodies that neutralize cholerae toxin and STa toxin and inhibit adherence of K88ac fimbrial E. coli Clin Vaccine Immunol 17 1859 1867 20980482 551 Zhang W Robertson DC Zhang C Bai W Zhao M Francis DH 2008 Escherichia coli constructs expressing human or porcine enterotoxins induce an identical diarrheal disease in a piglet infection model Appl Environ Microbiol 74 5832 5837 18658289 552 Bianchi AT Scholten JW van Zijderveld AM van Zijderveld FG Bokhout BA 1996 Parenteral vaccination of mice and piglets with F4+ Escherichia coli suppresses the enteric anti-F4 response upon oral infection Vaccine 14 199 206 8920700 553 Van den Broeck W Bouchaut H Cox E Goddeeris BM 2002 F4 receptor-independent priming of the systemic immune system of pigs by low oral doses of F4 fimbriae Vet Immunol Immunopathol 85 171 178 11943318 554 Melkebeek V Goddeeris BM Cox E 2013 ETEC vaccination in pigs Vet Immunol Immunopathol 152 37 42 23068270 555 Golchin M Aitken R 2008 Isolation by phage display of recombinant antibodies able to block adherence of Escherichia coli mediated by the K99 colonisation factor Vet Immunol Immunopathol 121 321 331 18036670 556 Martín MJ Martín-Sosa S Hueso P 2002 Binding of milk oligosaccharides by several enterotoxigenic Escherichia coli strains isolated from calves Glycoconj J 19 5 11 12652075 557 Mouricout M Petit JM Carias JR Julien R 1990 Glycoprotein glycans that inhibit adhesion of Escherichia coli mediated by K99 fimbriae: treatment of experimental colibacillosis Infect Immun 58 98 106 2403535 558 Nollet H Laevens H Deprez P Sanchez R Van Driessche E Muylle E 1999 The use of non-immune plasma powder in the prophylaxis of neonatal Escherichia coli diarrhoea in calves Zentralbl Veterinarmed A 46 185 196 10337234 559 Ofek I Hasty DL Doyle RJ 2003 Antiadhesion therapy 157 176 Bacterial adhesion to animal cells and tissues ASM Press Washington, D.C. 560 Ofek I Hasty DL Sharon N 2003 Anti-adhesion therapy of bacterial diseases: prospects and problems FEMS Immunol Med Microbiol 38 181 191 14522453 561 Atroshi F Alaviuhkola T Schildt R Sandholm M 1983 Fat globule membrane of sow milk as a target for adhesion of K88-positive Escherichia coli Comp Immunol Microbiol Infect Dis 6 235 245 6354573 562 Atroshi F Schildt R Sandholm M 1983 K 88-mediated adhesion of E. coli inhibited by fractions in sow milk Zentralbl Veterinarmed B 30 425 433 6353810 563 Lindahl M 1989 Binding of F41 and K99 fimbriae of enterotoxigenic Escherichia coli to glycoproteins from bovine and porcine colostrum Microbiol Immunol 33 373 379 2569154 564 Martín MJ Martín-Sosa S Alonso JM Hueso P 2003 Enterotoxigenic Escherichia coli strains bind bovine milk gangliosides in a ceramide-dependent process Lipids 38 761 768 14506839 565 Chandler DS Mynott TL 1998 Bromelain protects piglets from diarrhoea caused by oral challenge with K88 positive enterotoxigenic Escherichia coli Gut 43 196 202 10189844 566 Mynott TL Luke RK Chandler DS 1996 Oral administration of protease inhibits enterotoxigenic Escherichia coli receptor activity in piglet small intestine Gut 38 28 32 8566855 567 Spencer RJ Chesson A 1994 The effect of Lactobacillus spp. on the attachment of enterotoxigenic Escherichia coli to isolated porcine enterocytes J Appl Bacteriol 77 215 220 7961193 568 Avila FA Paulillo AC Schocken-Iturrino RP Lucas FA Orgaz A Quintana JL 1995 A comparative study of the efficiency of a pro-biotic and the anti-K99 and anti-A14 vaccines in the control of diarrhea in calves in Brazil Rev Elev Med Vet Pays Trop 48 239 243 8745746 569 Blomberg L Henriksson A Conway PL 1993 Inhibition of adhesion of Escherichia coli K88 to piglet ileal mucus by Lactobacillus spp Appl Environ Microbiol 59 34 39 8439162 570 Ouwehand AC Conway PL 1996 Purification and characterization of a component produced by Lactobacillus fermentum that inhibits the adhesion of K88 expressing Escherichia coli to porcine ileal mucus J Appl Bacteriol 80 311 318 8852678 571 Bomba A Nemcová R Gancarcíková S Herich R Kastel R 1999 Potentiation of the effectiveness of Lactobacillus casei in the prevention of E. coli induced diarrhea in conventional and gnotobiotic pigs Adv Exp Med Biol 473 185 190 10659357 572 Konstantinov SR Smidt H Akkermans AD Casini L Trevisi P Mazzoni M De Filippi S Bosi P de Vos WM 2008 Feeding of Lactobacillus sobrius reduces Escherichia coli F4 levels in the gut and promotes growth of infected piglets FEMS Microbiol Ecol 66 599 607 18537838 573 Gopal PK Prasad J Smart J Gill HS 2001 In vitro adherence properties of Lactobacillus rhamnosus DR20 and Bifidobacterium lactis DR10 strains and their antagonistic activity against an enterotoxigenic Escherichia coli Int J Food Microbiol 67 207 216 11518430 574 Daudelin JF Lessard M Beaudoin F Nadeau E Bissonnette N Boutin Y Brousseau JP Lauzon K Fairbrother JM 2011 Administration of probiotics influences F4 (K88)-positive enterotoxigenic Escherichia coli attachment and intestinal cytokine expression in weaned pigs Vet Res (Faisalabad) 42 69 575 Krause DO Bhandari SK House JD Nyachoti CM 2010 Response of nursery pigs to a synbiotic preparation of starch and an anti-Escherichia coli K88 probiotic Appl Environ Microbiol 76 8192 8200 20952649 576 Tsukahara T Tsuruta T Nakanishi N Hikita C Mochizuki M Nakayama K 2013 The preventive effect of Bacillus subtilus strain DB9011 against experimental infection with enterotoxcemic Escherichia coli in weaning piglets Anim Sci J 84 316 321 23590505 577 Jin LZ Marquardt RR Zhao X 2000 A strain of Enterococcus faecium (18C23) inhibits adhesion of enterotoxigenic Escherichia coli K88 to porcine small intestine mucus Appl Environ Microbiol 66 4200 4204 11010860 578 Devriendt B Stuyven E Verdonck F Goddeeris BM Cox E 2010 Enterotoxigenic Escherichia coli (K88) induce proinflammatory responses in porcine intestinal epithelial cells Dev Comp Immunol 34 1175 1182 20600278 579 Finamore A Roselli M Imbinto A Seeboth J Oswald IP Mengheri E 2014 Lactobacillus amylovorus inhibits the TLR4 inflammatory signaling triggered by enterotoxigenic Escherichia coli via modulation of the negative regulators and involvement of TLR2 in intestinal Caco-2 cells and pig explants PLoS One 9 e94891 24733511 580 Guerra-Ordaz AA González-Ortiz G La Ragione RM Woodward MJ Collins JW Pérez JF Martín-Orúe SM 2014 Lactulose and Lactobacillus plantarum, a potential complementary synbiotic to control postweaning colibacillosis in piglets Appl Environ Microbiol 80 4879 4886 24907322 581 Lee JS Awji EG Lee SJ Tassew DD Park YB Park KS Kim MK Kim B Park SC 2012 Effect of Lactobacillus plantarum CJLP243 on the growth performance and cytokine response of weaning pigs challenged with enterotoxigenic Escherichia coli J Anim Sci 90 3709 3717 22859771 582 Klose V Bayer K Bruckbeck R Schatzmayr G Loibner AP 2010 In vitro antagonistic activities of animal intestinal strains against swine-associated pathogens Vet Microbiol 144 515 521 20226602 583 Huja S Oren Y Trost E Brzuszkiewicz E Biran D Blom J Goesmann A Gottschalk G Hacker J Ron EZ Dobrindt U 2015 Genomic avenue to avian colisepticemia MBio 6 e01681 14 25587010 584 Guinée PA Agterberg CM Jansen WH Frik JF 1977 Serological identification of pig enterotoxigenic Escherichia coli strains not belonging to the classical serotypes Infect Immun 15 549 555 321354 585 Söderlind O Möllby R 1979 Enterotoxins, O-groups, and K88 antigen in Escherichia coli from neonatal piglets with and without diarrhea Infect Immun 24 611 616 381196 586 Wilson RA Francis DH 1986 Fimbriae and enterotoxins associated with Escherichia coli serogroups isolated from pigs with colibacillosis Am J Vet Res 47 213 217 2869725 587 Fairbrother JM Larivière S Johnson WM 1988 Prevalence of fimbrial antigens and enterotoxins in nonclassical serogroups of Escherichia coli isolated from newborn pigs with diarrhea Am J Vet Res 49 1325 1328 2902817 588 Suarez S Paniagua C Alvarez M Rubio P 1987 Features of enterotoxigenic Escherichia coli strains of porcine origin that express K88 and 987P fimbrial antigens Vet Microbiol 13 65 68 2880422 589 Dean-Nystrom EA Casey TA Schneider RA Nagy B 1993 A monoclonal antibody identifies 2134P fimbriae as adhesins on enterotoxigenic Escherichia coli isolated from postweaning pigs Vet Microbiol 37 101 114 7905220 590 Lintermans PF 1988 Characterization and purification of the F17 adhesin on the surface of bovine enteropathogenic and septicemic Escherichia coli Am J Vet Res 49 1794 1799 2907846 591 Contrepois M Bertin Y Pohl P Picard B Girardeau JP 1998 A study of relationships among F17 a producing enterotoxigenic and non-enterotoxigenic Escherichia coli strains isolated from diarrheic calves Vet Microbiol 64 75 81 9874105 592 Oswald E de Rycke J Lintermans P van Muylem K Mainil J Daube G Pohl P 1991 Virulence factors associated with cytotoxic necrotizing factor type two in bovine diarrheic and septicemic strains of Escherichia coli J Clin Microbiol 29 2522 2527 1774259 593 Orskov I Orskov F Sojka WJ Wittig W 1964 K antigens K88ab(L) AND K88ac(L) in E. coli - A new O antigen - 0147 and a new K antigen - K89(B) Acta Pathol Microbiol Scand 62 439 447 14227889 594 Ørskov I Ørskov F 1990 Serologic classification of fimbriae Curr Top Microbiol Immunol 151 71 90 1973373 595 Guinée PA Jansen WH 1979 Detection of enterotoxigenicity and attachment factors in Escherichia coli strains of human, porcine and bovine origin; a comparative study Zentralbl Bakteriol [Orig A] 243 245 257 596 Nagy B Orskov I Rátz F 1982 Pilus antigen 987P produced by strains of Escherichia coli serotypes O141:K–, H- and O8:K85:H– Acta Microbiol Acad Sci Hung 29 129 135 6127912 597 Evans MG Waxler GL Newman JP 1986 Prevalence of K88, K99, and 987P pili of Escherichia coli in neonatal pigs with enteric colibacillosis Am J Vet Res 47 2431 2434 2878634 598 Djønne BK Liven E 1986 Fimbriae in Escherichia coli isolated from the small intestine of piglets Acta Vet Scand 27 235 242 2879432 599 García D Cavazza ME Botero L Gorziglia M Urbina G Liprandi F Esparza J 1987 Preliminary characterization of Escherichia coli isolated from pigs with diarrhoea in Venezuela Vet Microbiol 13 47 56 2880421 600 Söderlind O Thafvelin B Möllby R 1988 Virulence factors in Escherichia coli strains isolated from Swedish piglets with diarrhea J Clin Microbiol 26 879 884 2454939 601 Takahashi K Takahashi N Yamamoto A Uemura T Imose J 1990 Isolation of Escherichia coli expressing 987P fimbrial antigen from suckling piglets Nippon Juigaku Zasshi 52 1113 1115 1980706 602 Supar Hirst RG Patten BE 1991 The importance of enterotoxigenic Escherichia coli containing the 987P antigen in causing neonatal colibacillosis in piglets in Indonesia Veterinary Microbiology 26 393 400 2031306 603 Osek J Truszczyński M 1992 Occurrence of fimbriae and enterotoxins in Escherichia coli strains isolated from piglets in Poland Comp Immunol Microbiol Infect Dis 15 285 292 1358502 604 Wray C McLaren IM Carroll PJ 1993 Escherichia coli isolated from farm animals in England and Wales between 1986 and 1991 Vet Rec 133 439 442 8291172 605 Shin SJ Chang YF Timour M Lauderdale TL Lein DH 1994 Hybridization of clinical Escherichia coli isolates from calves and piglets in New York State with gene probes for enterotoxins (STaP, STb, LT), Shiga-like toxins (SLT-1, SLT-II) and adhesion factors (K88, K99, F41, 987P) Vet Microbiol 38 217 225 7912467 606 Ojeniyi B Ahrens P Meyling A 1994 Detection of fimbrial and toxin genes in Escherichia coli and their prevalence in piglets with diarrhoea. The application of colony hybridization assay, polymerase chain reaction and phenotypic assays Zentralbl Veterinarmed B 41 49 59 7941847 607 Garabal JI Vázquez F Blanco J Blanco M González EA 1997 Colonization antigens of enterotoxigenic Escherichia coli strains isolated from piglets in Spain Vet Microbiol 54 321 328 9100332 608 Blanco M Blanco JE Gonzalez EA Mora A Jansen W Gomes TA Zerbini LF Yano T de Castro AF Blanco J 1997 Genes coding for enterotoxins and verotoxins in porcine Escherichia coli strains belonging to different O:K:H serotypes: relationship with toxic phenotypes J Clin Microbiol 35 2958 2963 9350767 609 Kwon D Kim O Chae C 1999 Prevalence of genotypes for fimbriae and enterotoxins and of O serogroups in Escherichia coli isolated from diarrheic piglets in Korea J Vet Diagn Invest 11 146 151 10098686 610 Matayoshi M Kaiga M Ooshiro S Nakazawa M 2001 Biological properties and virulence genes of enterotoxigenic Escherichia coli (ETEC) isolated from diarrheic piglets in Okinawa Prefecture J Jpn Vet Med Assoc 54 595 600 Japanese 611 Do T Stephens C Townsend K Wu X Chapman T Chin J McCormick B Bara M Trott DJ 2005 Rap id identification of virulence genes in enterotoxigenic Escherichia coli isolates associated with diarrhoea in Queensland piggeries Aust Vet J 83 293 299 15957392 612 Do TN Cu PH Nguyen HX Au TX Vu QN Driesen SJ Townsend KM Chin JJ Trott DJ 2006 Pathot ypes and serogroups of enterotoxigenic Escherichia coli isolated from pre-weaning pigs in north Vietnam J Med Microbiol 55 93 99 16388036 613 Lee SI Kang SG Kang ML Yoo HS 2008 Development of multiplex polymerase chain reaction assays for detecting enterotoxigenic Escherichia coli and their application to field isolates from piglets with diarrhea J Vet Diagn Invest 20 492 496 18599856 614 Isaacson RE 1977 K99 surface antigen of Escherichia coli: purification and partial characterization Infect Immun 15 272 279 401772 615 Altmann K Pyliotis NA Mukkur TK 1982 A new method for the extraction and purification of K99 pili from enterotoxigenic Escherichia coli and their characterization Biochem J 201 505 513 6124240 616 Isaacson RE Richter P 1981 Escherichia coli 987P pilus: purification and partial characterization J Bacteriol 146 784 789 6111559 617 de Graaf FK Klaasen P 1986 Organization and expression of genes involved in the biosynthesis of 987P fimbriae Mol Gen Genet 204 75 81 2875378 618 Lintermans P Pohl P Deboeck F Bertels A Schlicker C Vandekerckhove J Van Damme J Van Montagu M De Greve H 1988 Isolation and nucleotide sequence of the F17-A gene encoding the structural protein of the F17 fimbriae in bovine enterotoxigenic Escherichia coli Infect Immun 56 1475 1484 2897333 619 Grange PA Mouricout MA 1996 Transferrin associated with the porcine intestinal mucosa is a receptor specific for K88ab fimbriae of Escherichia coli Infect Immun 64 606 610 8550214 620 Erickson AK Baker DR Bosworth BT Casey TA Benfield DA Francis DH 1994 Characterization of porcine intestinal receptors for the K88ac fimbrial adhesin of Escherichia coli as mucin-type sialoglycoproteins Infect Immun 62 5404 5410 7960120 621 Erickson AK Willgohs JA McFarland SY Benfield DA Francis DH 1992 Identification of two porcine brush border glycoproteins that bind the K88ac adhesin of Escherichia coli and correlation of these glycoproteins with the adhesive phenotype Infect Immun 60 983 988 1347288 622 Grange PA Erickson AK Levery SB Francis DH 1999 Identification of an intestinal neutral glycosphingolipid as a phenotype-specific receptor for the K88ad fimbrial adhesin of Escherichia coli Infect Immun 67 165 172 9864211 623 Baker DR Billey LO Francis DH 1997 Distribution of K88 Escherichia coli-adhesive and nonadhesive phenotypes among pigs of four breeds Vet Microbiol 54 123 132 9057256 624 Bijlsma IG de Nijs A van der Meer C Frik JF 1982 Different pig phenotypes affect adherence of Escherichia coli to jejunal brush borders by K88ab, K88ac, or K88ad antigen Infect Immun 37 891 894 6752028 625 Hu ZL Hasler-Rapacz J Huang SC Rapacz J 1993 Studies in swine on inheritance and variation in expression of small intestinal receptors mediating adhesion of the K88 enteropathogenic Escherichia coli variants J Hered 84 157 165 8228168 626 Python P Jörg H Neuenschwander S Asai-Coakwell M Hagger C Bürgi E Bertschinger HU Stranzinger G Vögeli P 2005 Inheritance of the F4ab, F4ac and F4ad E. coli receptors in swine and examination of four candidate genes for F4acR J Anim Breed Genet 122 Suppl 1 5 14 16130451
PMC005xxxxxx/PMC5123741.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101573691 39703 Cell Rep Cell Rep Cell reports 2211-1247 27545877 5123741 10.1016/j.celrep.2016.07.024 NIHMS811464 Article Stimulation of Slack K+ channels alters mass at the plasma membrane by triggering dissociation of a phosphatase-regulatory complex Fleming Matthew R. 1 Brown Maile R. 1 Kronengold Jack 1 Zhang Yalan 1 Jenkins David P. 1 Barcia Gulia 4 Nabbout Rima 4 Bausch Anne E. 5 Ruth Peter 5 Lukowski Robert 5 Navaratnam Dhasakumar S. 2 Kaczmarek Leonard K. 136 1 Department of Pharmacology, Yale School of Medicine, New Haven, Connecticut, USA, 06520 2 Department of Neurology, Yale School of Medicine, New Haven, Connecticut, USA, 06520 3 Department of Cellular and Molecular Physiology, Yale School of Medicine, New Haven, Connecticut, USA, 06520 4 Department of Pediatric Neurology, Centre de Reference Epilepsies Rares, Hôpital Necker Enfants Malades, Assistance Publique-Hôpitaux de Paris (AP-HP), Paris, France 5 Department of Pharmacology, Toxicology and Clinical Pharmacy, Institute of Pharmacy University of Tübingen, 72076 Tübingen, Germany 6 Corresponding Author: Leonard K. Kaczmarek, Department of Pharmacology, Yale School of Medicine, 333 Cedar Street, New Haven, Connecticut 06520-8066, [email protected] 20 8 2016 18 8 2016 30 8 2016 25 11 2016 16 9 22812288 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Summary Human mutations in the cytoplasmic C-terminal domain of Slack sodium-activated potassium (KNa) channels result in childhood epilepsy with severe intellectual disability. Slack currents can be increased by pharmacological activators or by phosphorylation of a Slack C-terminal residue by protein kinase C. Using an optical biosensor assay, we find that Slack channel stimulation in neurons or transfected cells produces loss of mass near the plasma membrane. Slack mutants associated with intellectual disability fail to trigger any change in mass. The loss of mass results from the dissociation of the protein phosphatase 1 (PP1) targeting protein, Phactr-1, from the channel. Phactr1 dissociation is specific to wild-type Slack channels and is not observed when related potassium channels are stimulated. Our findings suggest that Slack channels are coupled to cytoplasmic signaling pathways, and that dysregulation of this coupling may trigger the aberrant intellectual development associated with specific childhood epilepsies. Introduction The Na+-activated K+ channel Slack (KNa1.1, KCNT1, SLO2.2) is widely distributed throughout the nervous system (Kaczmarek, 2013). KNa currents in neurons and in Slack-transfected cells are regulated by several pathways, including phosphorylation of a serine residue (S407) in cytoplasmic C-terminal domain, which results in a stimulation of current amplitude (Barcia et al., 2012; Santi et al., 2006). The large cytoplasmic C-terminal domain of Slack also interacts with the Fragile X mental retardation protein (FMRP), a RNA-binding protein that regulates activity-dependent protein translation (Brown et al., 2010; Zhang et al., 2012). The interaction of Slack with FMRP also stimulates channel activity. The amplitude of Slack currents can also be stimulated by a number of pharmacological agents, including bithionol and niclosamide (Biton et al., 2012; Yang et al., 2006). Mutations in ion channels can produce disorders of excitability, such as childhood epileptic seizures. Mutations in Slack have been described in malignant migrating partial seizures of infancy (MMPSI), a condition that produces infantile seizures coupled with very severe intellectual disability (Barcia et al., 2012; Kim et al., 2014). The majority of these mutations produce single amino acid substitutions within the cytoplasmic C-terminal domain of Slack. These gain-of-function mutations increase Slack current amplitude (Kim et al., 2014). It is unclear, however, why some epilepsy-associated mutations produce little intellectual deficit, while others, such as those in Slack, result in severe intellectual disability. Similar gain-of function mutations in another FMRP-interacting channel, the closely related Ca2+-activated K+ channel BK (KCNMA1, SLO1) (Deng et al., 2013), also produce seizures but do not result in intellectual disability (Du et al., 2005; N’Gouemo, 2014). This suggests that differences in the effects of these mutations on interactions with cytoplasmic signaling pathways, rather than current amplitude, may contribute to the differences in intellectual function. We have found that the large cytoplasmic domain of Slack interacts with a cytoplasmic signaling protein, Phactr1 (Phosphatase and Actin Regulator 1). Under normal conditions, stimulation of Slack channels causes Phactr1 to dissociate from the channel, resulting in a measurable loss of mass close to the plasma membrane of neurons. Mutant disease-causing Slack channels, however, fail to associate/dissociate with Phactr1. Our results suggest that failure to interact appropriately with its cytosolic signaling partners may underlie the severe intellectual disability associated with Slack mutations. Results Activation of Slack decreases mass at the plasma membrane To monitor the interactions of Slack channels with its potential cytoplasmic partners in real-time within living cells, we used resonance-wavelength grating (RWG) optical biosensors, a technique that has been used to monitor the activation of G-protein coupled receptors (Fang et al., 2007; Fleming and Kaczmarek, 2009; Lee, 2009). When cells adhere to these optical biosensors, changes in mass within ~150 nm of the biosensor alter the peak intensity of the reflected wavelengths of resonant light. Decreases or increases in protein density near the plasma membrane produce decreases or increases in the relative index of refraction, respectively (Fleming et al., 2014) (Figure 1A). RWG optical biosensors are sensitive enough to detect the binding of small molecules to proteins, providing a ready assay to detect the much larger changes resulting from the association/dissociation of channels with other proteins (Daghestani and Day, 2010; Lin et al., 2002). We first tested the actions of bithionol and niclosamide, two pharmacological activators that substantially enhance Slack currents (Biton et al., 2012; Yang et al., 2006). Treatment of Slack-expressing HEK293T cells with the Slo family channel activator bithionol (10 μM) (Yang et al., 2006) produced a progressive decrease in mass at the plasma membrane over several minutes following application (Figure 1B). The decrease in mass was sustained over the course of these experiments. Bithionol had no effect on untransfected cells, and DMSO vehicle had no significant effect on either untransfected or Slack-expressing cells (Figure 1B). Similar results were obtained using niclosamide (Biton et al., 2012) (Figure 1C). The observed decrease in mass upon channel stimulation is specific to Slack channel stimulation and not a general consequence of increased K+ conductance. Bithionol is a potent activator of other Slo family potassium channels, including BK channels and the very closely related Na+-activated K+ channel Slick (KNa1.2, KCNT2, Slo2.1) (Yang et al., 2006). Changes in mass were, however, evoked only in Slack-, but not in BK- or Slick- expressing cells (Figure 1D). As a further test to determine if ion flux plays a role in mass change, we pretreated Slack-expressing cells with the known Slack pore blocker Ba2+ (1 mM) (Bhattacharjee et al., 2003) for one hour prior to bithionol (10 μM) addition. Blocking K+ flux had no effect on the observed signal (Figure 1E). Changes in mass at the surface of the biosensor could reflect Slack-dependent changes in cell morphology or cellular movement. To exclude the possibility that the mass change upon Slack stimulation results from such changes we pretreated Slack-expressing cells with the actin polymerization inhibitor Latrunculin B (2 μM) (Wakatsuki et al., 2001) for one hour prior to channel stimulation by bithionol. Inhibition of actin polymerization did not attenuate the observed signal, excluding the possibility that the loss of mass results from actin-mediated changes in cell morphology (Figure 1F). To determine whether the Slack-stimulation induced loss of mass at the plasma membrane is an artifact of heterologous channel expression in HEK cells, we carried out the same experiment using primary cultures of cortical neurons from wild-type mice (Figure 1G–I). These neurons express Slack channels endogenously (Bhattacharjee et al., 2002; Brown et al., 2008). As in the cell line, a similar sustained decrease in mass was observed upon treatment with the Slack activator niclosamide (1 μM) (Biton et al., 2012) (Figure 1H–I). No significant change in mass was, however, produced on treatment of cortical neurons from animals in which the Slack gene had been deleted (Lu et al., 2015) (Figure 1G–I). Under physiological conditions, the stimulation of Slack occurs upon phosphorylation of residue S407 in the cytoplasmic C-terminus by protein kinase C (PKC), leading to a 2–3 fold increase in current (Barcia et al., 2012; Santi et al., 2006). Even in untransfected cells, however, treatment with PKC activators such as TPA (12-O-Tetradecanoylphorbol-13-acetate, 100 nM) leads to PKC translocation to the plasma membrane, increasing mass measured at the biosensor (Figure 2A). A change in mass produced by Slack stimulation in response to PKC activation was therefore calculated as the difference between the mass change produced by TPA in untransfected- and Slack-expressing cells (Figure 2B). We found that, like bithionol-treatment, stimulation of Slack channels by TPA led to a decrease in mass at the membrane (Figure 2B) and that the stimulatory effects of bithionol and phosphorylation are additive (Figure S1A and S1B). Mutation of serine 407 to an alanine abolished differences between Slack-expressing and control cells (Figure 2A and 2B), without altering bithionol sensitivity (Figure S1C and S1D). This indicates that, like bithionol, stimulation of Slack channels by phosphorylation at S407 decreases Slack-associated mass at the plasma membrane. Slack-induced changes in mass are absent in Slack MMPSI mutants To determine whether human Slack mutations that produce infantile seizures coupled to severe intellectual disability alter the normal pattern of mass distribution upon channel stimulation, we studied two previously described MMPSI mutations, Slack-R409Q and -A913T (Barcia et al., 2012). Expression of these mutant channels yield currents that resemble those of wild-type Slack, but their amplitude is increased by 2–3 fold and they cannot be further activated by PKC (Barcia et al., 2012). We first tested their sensitivity to pharmacological stimulation. Despite the fact that they are insensitive to TPA (Barcia et al., 2012), the currents of both mutant channels were activated by bithionol (10 μM) to the same extent as wild-type channels (Figure 2C). Surprisingly, however, we found that that both mutants failed to produce any change in mass in response to either bithionol (Figure 2D) or TPA (Figure 2E and 2F). The time course of changes in mass upon activation of PKC by TPA in cells expressing Slack-MMPSI mutant channels was identical to that in cells transfected with only control vector (Figure S2). The loss of Slack-associated changes in mass at the plasma membrane may therefore be linked to the pathogenesis of MMSPI. Slack interacts with FMRP, Cyfip1 and Phactr1 In addition to altering ion flux across the plasma membrane, the activation of ion channels can directly influence cytoplasmic events through protein-protein interactions between the channels and cellular signaling molecules (Calhoun and Isom, 2014; Deng et al., 2013; Ferron et al., 2014; Gross et al., 2011). Previous studies showed that the distal C-terminus of Slack interacts with FMRP, and that deletion of the Slack C-terminal domain distal to residue 804 results in functional channels that fail to interact with FMRP (Brown et al., 2010). We found that deletion of this same domain abolished the decrease in mass upon channel stimulation, indicating that this domain is required for the channel-protein interactions that lead to the loss of mass (Figure 3A). We therefore carried out experiments to test whether protein-protein interactions with FMRP or other cytoplasmic factors were required for the observed mass effects. To identify proteins other than FMRP that might interact with the C-terminus of Slack, we used the yeast 2-hybrid approach to identify two additional Slack-interacting proteins, Phactr1 (Phosphatase and Actin Regulator 1) and Cyfip1 (Cytoplasmic FMR1-Interacting Protein 1) (Figure S3). The interaction of Phactr1 with Slack was confirmed with co-immunopurification of Slack from transfected HEK cells (Figure 3C and 3D) and mouse brain (Figure S4), followed by Western blotting. Phactr1 did not co-immunoprecipitate with BK channels, indicating the specificity of this interaction with Slack (Figure 3E). Phactr1 is required for Slack-dependent changes in mass at the plasma membrane FMRP, Phactr1 and Cyfip1 are all expressed endogenously in HEK cells. To determine if one of these proteins is required for the decrease in mass at the plasma membrane upon Slack stimulation, we suppressed expression of each protein using RNAi. Conditions were optimized by detecting decreased protein expression of the target protein with individual or siRNAs in combination (Figure 4A). A decrease in levels of Phactr1, but not those of FMRP or Cyfip1, strongly suppressed the ability of Slack stimulation to produce a decrease in mass at the membrane (Figure 4B). RNAi treatments had no effect on cell viability (Figure 4C). To confirm that Phactr1 is required for changes in mass produced by Slack stimulation in real neurons, we repeated this experiment using mouse primary cortical neurons (Figure 4D). In these neurons, as in the HEK cells, down-regulation of Phactr1 strongly suppressed the effects of Slack stimulation. Phactr1 regulates the cellular localization of protein phosphatase 1 (PP1) (Allen et al., 2004; Wiezlak et al., 2012), and dissociation of Phactr1, presumably together with PP1, is likely to promote and maintain phosphorylation of the Slack channel. To determine whether the activity of PP1 is required for changes in mass at the plasma membrane, we tested the effects of the PP1 inhibitor tautomycetin (5 μM), which selectively inhibits PP1 over Protein Phosphatase 2A (Mitsuhashi et al., 2003). This agent did not, however, alter changes in mass at the plasma membrane on Slack stimulation, indicating that PP1 activity per se is not required (Figure 4E). The requirement for expression of Phactr1 for the observed change in mass on stimulation of Slack channels could represent the physical dissociation of Phactr1 from the channel, or could reflect Phactr1-dependent dissociation of some other factor. To measure changes in Phactr1/Slack association directly, we immunoprecipitated Slack channels before and after activation by TPA and bithionol. Levels of Phactr1 that were detected in the immunoprecipitate were very greatly reduced after channel stimulation (Figure 4F), confirming that loss of mass at the plasma membrane occurs when Phactr1 dissociates from Slack channels. Thus, PKC translocates to the membrane and phosphorylates Slack channels at S407 causing the release the Phactr1 and subsequent loss of mass at the membrane. Discussion Our findings indicate that the cytoplasmic C-terminal of Slack KNa channels interacts with several cytoplasmic signaling molecules including the phosphatase targeting protein Phactr1. On stimulation of the activity of Slack channels, either by the activation of PKC or by pharmacological agents, Phactr1 dissociates from the channel complex. It is possible that the change in mass at the plasma membrane that we measure both in transfected cells and cortical neurons in response to channel stimulation reflects the dissociation of other factors in addition to Phactr1. Nevertheless, suppression of the expression of either FMRP or Cyfip1 did not alter changes in mass at the membrane following Slack stimulation, indicating that these interacting proteins are not required for the observed effects. We have shown that the distal C-terminus of Slack is required for Phactr1/Slack interactions. This sequence of this region of Slack differs substantially from the corresponding regions in Slick and BK (Bhattacharjee et al., 2003; Salkoff et al., 2006), which apparently do not interact with Phactr1. Although the overall structure of Slack channels has recently been solved by cryo-electron microscopy (Hite et al., 2015), the distal cytoplasmic C-terminal region could not be resolved clearly, providing no insights into how this region may contribute to Phactr1 biding. The dissociation of Phactr1 is likely to regulate phosphorylation of the Slack channel. It may, however, potentially also regulate other channel-associated proteins such as FMRP. Alternatively, its dissociation from the channel may allow it to interact with other membrane or cytoplasmic targets. The disassociation of Phact1 on channel stimulation may therefore provide a signal that links neuronal excitability to downstream signaling pathways. The finding that the Phactr1/Slack interaction is entirely abolished in mutant Slack channels that produce childhood epilepsy with very severe developmental delay (Kim and Kaczmarek, 2014), suggests that this pathway may be crucial to normal neuronal development. Experimental Procedures Animals Rodents were handled in accordance with protocols approved by the Yale University Institutional Animal Care Committee. Expression Slack-B or Mutant Expression Vectors Slack-B S407A, R409Q, A913T, and Δ804 C-terminal truncation mutants were constructed as previously described (Barcia et al., 2012; Brown et al., 2010). Vector DNA concentrations were determined by a flourometric Qubit dsDNA Broad Range assay kit (Life Technologies). Transfections were performed with FugeneHD (Promega) transfection reagent. BIND Biosensor plates were seeded with 10,000 HEK293T cells/well. 0.2 μg DNA and 0.6 μL FugeneHD were diluted in 10 μL low serum OPTI-MEM medium (Life Technologies) for 15 minutes to allow DNA/FugeneHD complexes to form and added to HEK293T seeded plates and incubate under 95% Air/5% CO2 at 37° C for 48 hours. Two-electrode voltage clamp in Xenopus oocytes cRNA was created from Slack-B and mutant channel cRNA in pOX oocytes expression vector with a mMessage mMachine T3 kit (Ambion) and aliquoted in sterile water. Oocytes were isolated from Xenopus laevis frogs, injected with 50 – 100 ng of cRNA, and incubated in MND96 (Brown et al., 2010) at 18° C for up to 7 days before experiments were performed. Whole-oocyte currents were measured by a two-electrode voltage clamp amplifier (OC-725C, Warner Instruments.) Electrodes were filled with 3 M KCl and had resistance 0.1 – 1 MΩ. Standard bath solution was MND96. Ooctyes were depolarized by 400 ms pulses from a holding potential of −90 mV to test pulses of between −80 and +60 mV in 10 mV increments for 5 seconds. Data recording and analysis were performed using pClamp (Molecular Devices), Excel (Microsoft) and Origin 8.1 (OriginLab Corporation) software packages. BIND Scanner Assays Poly-D-lysine/Laminin coated TiO2 384-well BIND biosensor plates were utilized for all Scanner assays. For HEK293 cell based experiments, cells were counted with a hemocytometer and plated at a density of 10k cells/well in growth media and incubated for 24–48 hours. For primary cortical neuron based assays, neurons from postnatal day 2 or 3 male mice were prepared as described (Brewer and Torricelli, 2007) and incubated for 7 days. After 7 days incubation, growth media was removed and HBSS with 0.1% DMSO was added. Cells were allowed to adjust to room temperature for 1 hour before a baseline measurement was recorded. Measurements were recorded at 3.75 μm/pixel resolution for the central 1.0 mm2 area of each well. Compounds or DMSO controls were added, and 30 minutes later a second measurement was recorded. A density gradient map was generated for both records, and individual cells and the local background selected utilizing BIND Scanner analysis software. The change in the index of refraction for each cell and the local background as a function of the difference between the first and second images was calculated. The change in index of refraction for each cell was normalized to the local background, and a mean value calculated for each well. The change in peak wavelength of refraction was divided by the initial peak wavelength to give a physically relevant change in the index of refraction. For primary cortical neuron experiments, niclosamide addition was normalized to DMSO vehicle alone addition. Results were exported to Microsoft Excel and Origin Statistical Package for analysis. Statistical Analysis Data recording and analysis were performed using pClamp (Molecular Devices), Excel (Microsoft), Origin (Origin Lab Corporation), and BIND Scanner and BIND View (X-BODY Biosciences). The mean±SEM values are plotted. Where 2 groups were compared, 2-tailed student’s t-test was performed to compare statistical significance; where 3 or more groups were compared, 1-way ANOVA test was performed. See Supplemental Experimental Procedures for chemicals, cell culture, creation of mammalian expression vectors, animals, yeast two hybrid assay, RNAi against candidate Slack-B binding partners, cell viability assay, protein purification of RNAi samples, co-immunopurification experiments, western blotting, as well as detailed experimental procedures. Supplementary Material 1 Figure S1 (Related to Figure 2): Activation of Slack-B by the small molecule activator bithionol and by channel phosphorylation are independent, and the effect of bithionol activation is unchanged in Slack-S407A phosphorylation mutants. Slack transfected A) Xenopus oocytes or B) HEK cells were treated with either bithionol (10 μM), TPA (100 nM), or bithionol followed 20 minutes later by TPA. After 20 minutes incubation, either (A) two-electrode voltage clamp electrophysiology or (B) resonance wavelength grating optical biosensor assay was performed. For (A), the maximum percent increase in current from potentials of −80 mV to +60 mV was calculated (n= 3–4 oocytes/condition, p<0.01), and for (B) change in mass near the plasma was calculated (n=16 wells/condition, p<0.001). Slack or Slack-S407A transfected (C) Xenopus oocytes or (D) HEK cells were treated with bithionol (10 μM) and either (C) two-electrode voltage clamp electrophysiology or (D) resonance wavelength grating optical biosensor assay was performed 20 minutes after bithionol treatment. For (C), the maximum percent increase in current from potentials of −80 mV to +60 mV was calculated (n= 4 oocytes/condition), and for (D) change in mass near the plasma was calculated (n=16 wells/condition). Error bars ± SEM. Figure S2 (Related to Figure 2): TPA-induced mass redistribution in cells expressing MMPSI associated Slack mutant channels, is identical to that of cells transfected with only with vector. Slack, Slack-R409Q, or Slack-A913T transfected HEK cells and those transfected with vector alone were treated with the PKC activator TPA (100 nM) and the change in mass near the plasma membrane recorded (n=32 wells/condition). Error bars ± SEM. Figure S3 (Related to Figure 3): Protein-protein interactions between Slack-Phactr1 and Slack-Cyfip1 were identified by a yeast-two hybrid assay. Using a Gal-4 based yeast two-hybrid assay, the C-terminus of Slack was inserted into pGBKT7 and screened against a mouse brain cDNA library in pGADT7 vector. AH109 yeast cells were sequentially transformed with Slack-pGBKT7 and then the mouse cDNA library. Two of these clones from the initial screen containing (a) Phactr1 (aa1-435 NP_001289564) and (b) Cyfip1 (aa 1-326, NP_035500) demonstrated grown in restrictive conditions (-Leu, -Trp, -His). Figure S4 (Related to Figure 3): Phactr1 co-immunoprecipitates with Slack in mouse brain. Cell lysates were prepared from mouse brain in either wild-type or Slack knock-out mice. Co-immunoprecipitation was performed using anti-Slack antibodies or chicken IgY antibodies followed by western blotting with anti-Phactr1 antibodies. Left, middle, and right panels are from the same blot but were separated, reordered and adjusted in width for clarity of presentation. Also shown are GAPDH loading controls for cell lysates of wild-type or Slack knock-out mice (n=2/condition). 2 3 The authors are grateful to Yangyang Yan and Fred Sigworth at Yale University for donation of Slack-B, Slick and BK cell lines and technical assistance with protein purification, and to Steven Shamah and Brandt Binder at X-BODY Biosciences for the usage of the BIND Scanner system and technical assistance. This work is supported by grants from the NIH (HD067517) and from FRAXA to L.K.K. Figure 1 Stimulation of Slack channels alters mass distribution at the plasma membrane (A) Schematic diagram of Slack activation in a cell adherent to the biosensor. (B) Activation of Slack-expressing, but not untransfected, HEK cells with bithionol (10 μM) produced a sustained decreased in mass near the plasma membrane, n=32 wells/condition, p<0.001. (C) Changes in mass in Slack-expressing HEK after Slack activation with bithionol (10 μM) or niclosamide (500 nM) (n=16, p<0.001). (D) Stimulation of Slack, but not other Slo channel family members, with bithionol (10 μM) decreases mass near the plasma membrane, n=24, p<0.001. (E) Blocking ion flux through Slack channels with Ba2+ (1 mM) does not alter the signal upon channel activation, n=16. (F) Latrunculin B (2 μM) does not alter the signal upon Slack activation by bithionol (10 μM), n=12. (G) Slack protein expression is absent in brain synaptosomes isolated from Slack−/− mice (n=3/genotype). (H and I) Activation of Slack channels in WT, but not Slack−/−, murine cortical neurons decreases mass near the plasma membrane. (H) Representative images showing change in mass in cortical neurons with red representing decreases in mass and black representing no change in mass. (I) Quantification of (H), n=12, p<0.001. Error bars ±SEM. Figure 2 MMPSI and phosphorylation site Slack mutations prevent decreases in mass upon channel activation (A) Changes in mass measured 30 min after addition of the PKC activator TPA or its inactive analog 4α-phorbol, to untransfected cells or to cells expressing wild-type (WT) Slack or its phosphorylation site mutant S407A Slack. The increase in mass produced by TPA in untransfected cells or S407A cells is significantly reduced in cell expressing WT Slack. n=16, p<0.001. (B) The contribution of Slack to the mass change in a was calculated by the difference between the positive mass change after TPA addition in untransfected (black bar, A) and Slack-containing cells (red bar, B). (C) Percentage increases in Slack current amplitude produced by bithionol (10 μM, 20 min) in Xenopus oocytes expressing WT Slack, or the MMPSI mutants R409Q Slack or A913T Slack cRNA. n = 3 oocytes/condition. (D) Slack R409Q and A913T mutant channels produce no change in mass after activation by bithionol (10 μM), n = 16, p<0.001. (E) Changes in mass produced by activation of PKC by TPA in R409Q and A913T Slack expressing cells are not different from those in untransfected (Vector Control cells). (F) Differences in mass change between untransfected and Slack-expressing cells in response to TPA treatment. Mass change in R409Q and A913T mutant Slack-expressing cells was not different from untransfected cells but significantly different from WT Slack cells (n = 16, p<0.001). Error bars ± SEM. Figure 3 Slack interacts with Phacr1 (A) The changes in mass produced by PKC activation in cells expressing wild type Slack differs from that in cells expressing Δ804 Slack, which lacks the distal C-terminus, or untransfected cells (n = 16, p<0.001). Error bars ±SEM. (B) Data in (A) are plotted as the difference between untransfected cells and Slack-expressing cells. (C–F) Phactr1 binds to Slack channels but not BK channels. Cell lysates were prepared from Slack-Flag (C and D) or BK-Flag (E and F) expressing cells and coimmunoprecipitation was performed using anti-Flag antibodies followed by western blotting with anti-Phactr1 (C and E) antibodies or anti-Slack (D and F) antibodies (n=2/condition). GAPDH loading controls for cell lysates of Slack-Flag (C) and BK-Flag (E) inputs were performed. Figure 4 Phactr1 is required for Slack-induced changes in mass following phosphorylation or pharmacological stimulation (A) Blot showing decreased expression of Slack-associated proteins following RNAi against FMRP, Cyfip1, or Phactr1 in WT Slack-expressing cells. Lanes are s=scrambled siRNA negative control, 1 – 3 = three different individual siRNAs alone, 4 = equimolar combination of the 3 siRNAs, G=GAPDH positive control. Bottom blot shows that RNAi against Phactr1 did not decrease Slack expression. (B) RNAi against Phactr1, but not FMRP or Cyfip1, significantly reduced the mass change produced by bithionol(10 μM) in WT Slack cells (n=8, p<0.001, Error bars ± SEM). (C) RNAi against Slack-associated proteins, does not decrease cell viability in a Resazurin assay. A standard curve of cell survival was generated by utilizing 10 fold dilutions of Saponin from 0.1% to 0%, with 0.1% Saponin treated cells considered no cell survival, and 0% Saponin treated cells considered complete cell survival, n = 3 wells/condition, Z’ for assay = 0.60 (D) Changes in mass produced in mouse primary cortical neurons by the Slack activator niclosamide (1 μM) are suppressed in neurons treated with RNAi against Phactr1(n=12, p<0.01). (E) Pre-treatment of Slack-B expressing cells with the PP1 inhibitor tautomycetin (2 μM) does not alter bithionol-induced changes in mass in WT Slack cells (n=8). (F) Western blots demonstrating that treatment of WT Slack cells with either bithionol (10 μM) or TPA (100 nM) strongly reduces levels of Phactr1 that can be co-immunoprecipitated with Slack channels (n=2/condition). Author contributions: M.R.F. and L.K.K designed research. M.R.F., M.R.B, J.K., Y.Z. and D.P.G. performed research; G.H and R.N. generated MMPSI mutants; A.E.B, P.R and R.L generated Slack−/− animals; D.S.N carried out yeast-two hybrid experiments. M.R.F, M.R.B and L.K.K wrote the paper. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. Allen PB Greenfield AT Svenningsson P Haspeslagh DC Greengard P 2004 Phactrs 1–4: A family of protein phosphatase 1 and actin regulatory proteins Proc Natl Acad Sci U S A 101 7187 7192 15107502 Barcia G Fleming MR Deligniere A Gazula VR Brown MR Langouet M Chen H Kronengold J Abhyankar A Cilio R 2012 De novo gain-of-function KCNT1 channel mutations cause malignant migrating partial seizures of infancy Nat Genet 44 1255 1259 23086397 Bhattacharjee A Gan L Kaczmarek LK 2002 Localization of the Slack potassium channel in the rat central nervous system J Comp Neurol 454 241 254 12442315 Bhattacharjee A Joiner WJ Wu M Yang Y Sigworth FJ Kaczmarek LK 2003 Slick (Slo2.1), a rapidly-gating sodium-activated potassium channel inhibited by ATP J Neurosci 23 11681 11691 14684870 Biton B Sethuramanujam S Picchione KE Bhattacharjee A Khessibi N Chesney F Lanneau C Curet O Avenet P 2012 The antipsychotic drug loxapine is an opener of the sodium-activated potassium channel slack (Slo2.2) J Pharmacol Exp Ther 340 706 715 22171093 Brewer GJ Torricelli JR 2007 Isolation and culture of adult neurons and neurospheres Nat Protoc 2 1490 1498 17545985 Brown MR Kronengold J Gazula VR Chen Y Strumbos JG Sigworth FJ Navaratnam D Kaczmarek LK 2010 Fragile X mental retardation protein controls gating of the sodium-activated potassium channel Slack Nat Neurosci 13 819 821 20512134 Brown MR Kronengold J Gazula VR Spilianakis CG Flavell RA von Hehn CA Bhattacharjee A Kaczmarek LK 2008 Amino-termini isoforms of the Slack K+ channel, regulated by alternative promoters, differentially modulate rhythmic firing and adaptation J Physiol 586 5161 5179 18787033 Calhoun JD Isom LL 2014 The role of non-pore-forming beta subunits in physiology and pathophysiology of voltage-gated sodium channels Handb Exp Pharmacol 221 51 89 24737232 Daghestani HN Day BW 2010 Theory and applications of surface plasmon resonance, resonant mirror, resonant waveguide grating, and dual polarization interferometry biosensors Sensors (Basel) 10 9630 9646 22163431 Deng PY Rotman Z Blundon JA Cho Y Cui J Cavalli V Zakharenko SS Klyachko VA 2013 FMRP regulates neurotransmitter release and synaptic information transmission by modulating action potential duration via BK channels Neuron 77 696 711 23439122 Du W Bautista JF Yang H Diez-Sampedro A You SA Wang L Kotagal P Luders HO Shi J Cui J 2005 Calcium-sensitive potassium channelopathy in human epilepsy and paroxysmal movement disorder Nat Genet 37 733 738 15937479 Fang Y Li G Ferrie AM 2007 Non-invasive optical biosensor for assaying endogenous G protein-coupled receptors in adherent cells J Pharmacol Toxicol Methods 55 314 322 17207642 Ferron L Nieto-Rostro M Cassidy JS Dolphin AC 2014 Fragile X mental retardation protein controls synaptic vesicle exocytosis by modulating N-type calcium channel density Nat Commun 5 3628 24709664 Fleming MR Kaczmarek LK 2009 Use of optical biosensors to detect modulation of Slack potassium channels by G protein-coupled receptors J Recept Signal Transduct Res 29 173 181 19640220 Fleming MR Shamah SM Kaczmarek LK 2014 Use of label-free optical biosensors to detect modulation of potassium channels by G-protein coupled receptors J Vis Exp e51307 24562095 Gross C Yao X Pong DL Jeromin A Bassell GJ 2011 Fragile X mental retardation protein regulates protein expression and mRNA translation of the potassium channel Kv4.2 J Neurosci 31 5693 5698 21490210 Hite RK Yuan P Li Z Hsuing Y Walz T MacKinnon R 2015 Cryo-electron microscopy structure of the Slo2.2 Na(+)-activated K(+) channel Nature 527 198 203 26436452 Kaczmarek LK 2013 Slack, Slick and Sodium-Activated Potassium Channels ISRN Neurosci Kim GE Kaczmarek LK 2014 Emerging role of the KCNT1 Slack channel in intellectual disability Front Cell Neurosci 8 209 25120433 Kim GE Kronengold J Barcia G Quraishi IH Martin HC Blair E Taylor JC Dulac O Colleaux L Nabbout R 2014 Human slack potassium channel mutations increase positive cooperativity between individual channels Cell Rep 9 1661 1672 25482562 Lee PH 2009 Label-free optical biosensor: a tool for G protein-coupled receptors pharmacology profiling and inverse agonists identification J Recept Signal Transduct Res 29 146 153 19594401 Lin B Qiu J Gerstenmeier J Li P Pien H Pepper J Cunningham B 2002 A label-free optical technique for detecting small molecule interactions Biosens Bioelectron 17 827 834 12191932 Lu R Bausch AE Kallenborn-Gerhardt W Stoetzer C Debruin N Ruth P Geisslinger G Leffler A Lukowski R Schmidtko A 2015 Slack channels expressed in sensory neurons control neuropathic pain in mice J Neurosci 35 1125 1135 25609627 Mitsuhashi S Shima H Tanuma N Matsuura N Takekawa M Urano T Kataoka T Ubukata M Kikuchi K 2003 Usage of tautomycetin, a novel inhibitor of protein phosphatase 1 (PP1), reveals that PP1 is a positive regulator of Raf-1 in vivo J Biol Chem 278 82 88 12374792 N’Gouemo P 2014 BKCa channel dysfunction in neurological diseases Front Physiol 5 373 25324781 Salkoff L Butler A Ferreira G Santi C Wei A 2006 High-conductance potassium channels of the SLO family Nat Rev Neurosci 7 921 931 17115074 Santi CM Ferreira G Yang B Gazula VR Butler A Wei A Kaczmarek LK Salkoff L 2006 Opposite regulation of Slick and Slack K+ channels by neuromodulators J Neurosci 26 5059 5068 16687497 Wakatsuki T Schwab B Thompson NC Elson EL 2001 Effects of cytochalasin D and latrunculin B on mechanical properties of cells J Cell Sci 114 1025 1036 11181185 Wiezlak M Diring J Abella J Mouilleron S Way M McDonald NQ Treisman R 2012 G-actin regulates the shuttling and PP1 binding of the RPEL protein Phactr1 to control actomyosin assembly J Cell Sci 125 5860 5872 22976292 Yang B Gribkoff VK Pan J Damagnez V Dworetzky SI Boissard CG Bhattacharjee A Yan Y Sigworth FJ Kaczmarek LK 2006 Pharmacological activation and inhibition of Slack (Slo2.2) channels Neuropharmacology 51 896 906 16876206 Zhang Y Brown MR Hyland C Chen Y Kronengold J Fleming MR Kohn AB Moroz LL Kaczmarek LK 2012 Regulation of neuronal excitability by interaction of fragile X mental retardation protein with slack potassium channels J Neurosci 32 15318 15327 23115170
PMC005xxxxxx/PMC5123750.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101634614 42750 Microbiol Spectr Microbiol Spectr Microbiology spectrum 2165-0497 27837740 5123750 10.1128/microbiolspec.MCHD-0005-2015 NIHMS820940 Article Myeloid cell turnover and clearance Janssen William J. Bratton Donna L. Jakubzick Claudia V. Henson Peter M. * Departments of Pediatrics and Medicine, National Jewish Health, Denver; Departments Immunology, Medicine and Pediatrics, University of Colorado, Denver * Corresponding author: Peter M. Henson, [email protected] 5 10 2016 11 2016 25 11 2016 4 6 10.1128/microbiolspec.MCHD-0005-2015This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Given the dual and intrinsically contradictory roles for myeloid cells in both protective and yet also damaging effects of inflammatory and immunological processes we suggest that it is important to consider the mechanisms and circumstances by which these cells are removed, either in the normal unchallenged state or during inflammation or disease. In this essay we address these subjects from a conceptual perspective, focusing as examples on four main myeloid cell types (neutrophils, monocytes, macrophages and myeloid dendritic cells) and their clearance from the circulation or from naïve and inflamed tissues. While the primary clearance process appears to involve endocytic uptake into macrophages, various tissue cell types can also recognize and remove dying cells though their overall quantitative contribution is unclear. In fact, surprisingly, given the wealth of study in this area over the last 30 years, our conclusion is that we are still challenged with substantial lack of mechanistic and regulatory understanding of when, how and by what mechanisms migratory myeloid cells come to die, are recognized as needing to be removed and indeed the precise processes of uptake of either the intact or fragmented cells. This reflects the extreme complexity and inherent redundancy of the clearance processes and argues for substantial investigative effort in this arena. In addition, it leads us to a sense that approaches to significant therapeutic modulation of selective myeloid clearance is still a long way off. Introduction Few, if any, individual cells survive throughout the life of the animal, an observation that sets up the critical concepts of cell lifespan, turnover, removal and maintenance of homeostatic cell numbers. These issues are of special interest for understanding the properties of the myeloid cell lineage, which includes cells such as neutrophils, that may exhibit in the normal naive adult mammal the shortest lifespan of all, but yet are maintained in relatively constant numbers within the circulation. However our understanding of the underlying mechanisms for myeloid cell maintenance and removal is still substantially limited and also requires reexamination in light of new ideas about the ontogeny, characterization and distribution of the myeloid cells in general. Accordingly, this essay will focus on the concepts and questions that, we argue, are in need of exploration, rather than providing a detailed review of what is a huge literature. By focusing on four of the myeloid lineage cell types, (neutrophils, monocytes, macrophages and dendritic cells) we will also be able to bring to the fore many of the key issues that characterize this set of questions. Removal of cells implies cell death and destruction with uptake into phagocytes and subsequent digestion within the endosomes. An exception would be loss at extracorporeal sites such as lung, gut, skin etc where the cells may be removed physically. Various forms of programed cell death (PCD), often, but erroneously, subsumed under the term “apoptosis”, lead to uptake. Clearly, un-programed cell death (often called necrosis) can generate dead cells and cell debris that are also generally removed by being engulfed by phagocytic cell engulfment. Key to these processes is the necessary recognition of the dying cell or its constituents by the phagocyte – unique forms of “self-recognition” – that seem at first hand to defy the concepts of self/non-self that underpin how we usually think of “immunity”. In addition, and possibly of significant importance, stimulated cells that are still “living” may also exhibit such recognition signals that lead to their removal while still active (see the section on neutrophils) thus serving a potential regulatory role at the level of whole, living, cells. This removal by endocytic uptake inevitably puts the emphasis on myeloid cells themselves, especially macrophages, as key instruments of the cell and debris clearance (a legacy of Metchnikov’s phagocyte theories). However, it is increasingly clear that many non-myeloid tissue cell types can, either endogenously or after appropriate stimulation, exhibit these endocytic functions, including the engulfment of whole cells up to 15µm in diameter, i.e. clearcut phagocytosis. This point is also exemplified by the extensive literature on intact apoptotic cell clearance in C. elegans carried out by near-neighbor tissue cells in the absence of macrophages. A primitive clearance function would be an obvious requirement for tissue development in multicellular organisms, especially evident in substantial metamorphic alterations at different organizational phases seen in numerous animal groups. Implications from some of the observations noted below emphasize the possible unique elements of these endocytic clearance functions for tissue or inflammatory cells that would be in keeping with their early metazoan evolutionary development. In the context of understanding the life history and functions of mammalian myeloid cells, especially in the normal resolution of inflammatory processes to restore tissue homeostasis, the mechanisms underlying their recognition and removal become critical, and even therapeutically targetable. Accordingly in this discussion, we will first briefly address general issues of removal of cells and cell debris, which, as noted, also significantly involve a key function of phagocytic myeloid cells. Subsequent sections will focus on the four individual cell types, in each case with an emphasis on general points and concepts that, we suggest, are understudied and ripe for new experimental analysis. Recognition and removal of cells or of cell debris Before discussing turnover and removal of the different myeloid cell types, it is relevant to first consider the broader question of how dying, effete or fragmented/disrupted cells are recognized and then ingested. This is a form of self-recognition that, for the most part, does not in itself initiate inflammatory, immunological or defensive responses in the tissues and thus, must be distinguished from more classical innate or adaptive immune system recognition processes. Strikingly, this normal cell removal (including of myeloid cells) is occurring all the time in very large numbers (>1010 per day) in fashions that are essentially immunologically invisible. Recognition Extensive studies over the last 20 years have revealed numerous surface alterations that occur during programed cell death that contribute to recognition and removal of the dying cells. Many of these surface alterations are likely also to extend to cell debris if the cell undergoes disruption, either due to induction of externally induced “necrosis” or post-apoptotic cytolysis (often termed secondary necrosis). A general concept here is that as a consequence of such changes on apoptotic/PCD cells, their removal is usually so efficient that a) they are generally cleared before undergoing secondary necrosis, and b) it is very hard to detect apoptotic cells in tissues unless the clearance mechanisms are disrupted or overloaded. A key early change during programmed cell death is the cell’s inability to maintain the normal phospholipid asymmetry of the plasma membrane, resulting in the outward exposure of phosphatidylserines (PSs) from their normal location on the inner leaflet (1). Detection of this surface PS is an often-used assay for apoptosis and other forms of PCD and is also one of the most important recognition signals for normal removal of the cells. The PS species themselves may be in the form normally present on the inner leaflet or may become oxidized – leading to a various sets of recognition structures. These may be direct receptors on the engulfing phagocyte (for example BAI1, CD36, some scavenger receptors) or various sets of “bridge” molecules (opsonins) that bind the PS on the dying cell or cell debris (e.g. Protein S, MFGE8) but also link to the surface of the phagocyte to initiate ingestion via different sets of receptors (Mer or αv integrins as examples). Notably, such “bridge” molecules may be produced by the phagocytes themselves (e.g. MFGE8, thrombospondin-1) or may be present in the local environment (e.g. Protein S produced by endothelial cells). It is noteworthy that phospholipid asymmetry of the plasma membrane, and thus the normal inner leaflet distribution of PS is energy dependent (see for example, (2, 3)). PS may be exposed on the outer leaflet as a potential clearance recognition ligand under many different circumstances, including inevitably, when the plasma membrane is disrupted, i.e. during primary or secondary necrosis as well as in extrusion of numerous forms of exosomes, microparticles or even enveloped viruses. A number of specific molecules involved in altering the distribution of phospholipids in/on the cell membrane have recently been characterized (4) as well as mechanisms for maintaining the asymmetry under normal circumstances. In addition to PS exposure, many other molecular and structural changes on the surface of cells undergoing PCD have been described, each potentially also leading to interaction with the phagocyte via either direct receptor ligation or relevant bridge molecules. Examples include calreticulin from the dying cell endoplasmic reticulum, externally exposed chromatin and DNA, alterations in surface charge and very likely (but not studied) changes in glycosyl groups (glycocalyx). More complete consideration of these recognition processes is given elsewhere (5–7) but the overall picture indicates high levels of redundancy, further supported by the difficulty of experimentally blocking these systems in vivo. Also implicit in these comments is the expectation that different cell types and different circumstances that render cells effete and ready for removal and/or induce frank PCD or necrosis are likely to generate different combinations and permutations of these recognition structures and removal processes. Unfortunately, this point has not always been considered in the past, leading to studies of cell removal using standardized “apoptotic” or “necrotic” cells that may have only minimal bearing on their removal in vivo as well as ignoring the contribution to clearance of various forms of cell “debris”. These concepts apply also to normal myeloid cells that end their effective lifespans in the blood or tissues, and is especially true for cells exposed to inflammatory or toxic environmental processes. Removal As noted, the most effective route of cell removal in vertebrates appears to be phagocytosis by various categories of tissue macrophages. Importantly, however, the cells that remove effete cells in the bone marrow are poorly categorized or understood, though also important in clearing excess leukocytes that have developed at this site but which are not called into the circulation. This substantial clearance mechanism in the bone marrow extends beyond removal of myeloid cells to, for example, clearance of extruded nuclear material from developing erythrocytes (8). Ingestion of intact apoptotic cells (the normal process in homeostatic conditions) into macrophages appears in general to involve a somewhat unique phagocyte process that can be distinguished in a number of ways from “classical” phagocytosis mediated by normal innate or adaptive immunological recognition. The involved signal pathways for the apoptotic cell uptake are highly conserved between C. elegans, Drosophila and mammals (9–11) and show a significant requirement for the GTPase Rac with, in fact, a balancing suppressive action from the GTPase Rho. It has been suggested that the process represents a version of stimulated macropinocytosis (similarly Rac-requiring and similarly responsive to macropinocytosis inhibitors and generally resulting in early spacious phagosomes and concurrent uptake of surrounding fluid and its contents). In this case the usual 200nm size consideration for macropinocytosis is clearly overcome and there is also a requirement for initial tethering of the target cell to the phagocyte. To emphasize this distinction we have termed the process “efferocytosis”, (see (12, 13) and Figure 1). Not surprisingly, given the variability of the surface ligands and potential signaling responses, this concept of a primary uptake mechanism is likely oversimplified and we have argued that the specific uptake mechanisms, as well as the disposition of the phagosome and its contents are urgently in need of detailed investigation. Similarly, the increasing recognition that uptake of necrotic and disrupted cells (cell debris) is critical to understanding resolution of inflammation and normal tissue homeostasis requires investigation of the underlying mechanisms – to this point substantially understudied. Neutral, Inflammosuppressive and immunosuppressive responses to cell removal The massive ongoing uptake and removal of circulating cells (including myeloid cells) is normally immunologically and inflammationally silent (as is the similar daily removal of a massive load of retinal outer segments in the eye (14). Simplistically, this may be because the triggers for uptake (recognition signals) just do not stimulate proinflammatory responses in the phagocytosing cells. On the other hand in many experimental systems uptake of intact apoptotic cells has been shown to be actively anti-inflammatory and anti-immunogenic. This effect may in turn be mediated by intrinsic blockade of potential inflammatory responses (e.g. inhibition of proinflammatory transcription factors such as NFκB) or by initiation of anti-inflammatory mediators such as TGFβ or IL-10 that can act in autocrine or paracrine fashion to suppress responses of cells in the local environment. Examples of both possibilities abound (9, 11,15, 16). Notably, most of these studies have focused on the effects of apoptotic cells in suppressing stimulated inflammatory or immunogenic responses in the macrophages or dendritic cells rather than addressing the importance of mere lack of initiation of inflammatory responses when the phagocytes interact and ingest apoptotic cells. The potential difference here between efferocytosis and the phagocyte uptake of true foreign, or opsonin-coated particles is more evidence of the specialized receptors and signaling involved in efferocytosis as well as in the impact of the actual uptake process on the general response of the ingesting cell. These issues are of substantial importance to host response processes in general and in the questions underlying the initiation of autoimmunity. For example, there are numerous reports of associations between ineffective removal of apoptotic cells and induction of autoimmunity in both animals and man (see below). The implications are that if dying or effete cells are not removed (especially for example circulating granulocytes) they can undergo secondary necrosis, thereby releasing internal constituents such as chromatins or other nuclear contents that can initiate inflammation and immune responses via standard recognition and uptake as foreign stimuli, for example involving toll-like receptors. On the other hand, there are increasing reasons to believe that at least some forms of non- or post-apoptotic cell dissolution may in fact generate particles that can also be cleared efficiently without inflammatory or immunological sequelae. These concepts led to ideas of a contrasting dualism between reactions to apoptotic versus necrotic cells and even the teleology of “good” versus “bad” cell death. This is unfortunate, since the situation is clearly much more complex and the responses of phagocytes to cells and cell debris are much more nuanced and balanced. More realistically, the complexity and redundancy of ligands on dying cells and receptors on the phagocytes is likely to generate a wide spectrum of effects that vary from nothing more than uptake itself, to induction of anti-inflammatory processes, to frank initiation of inflammation and, in the absence of appropriate immunological tolerance, to autoimmunity. As an example, macrophage responses to ligation of receptors associated with PS recognition versus those associated with calreticulin recognition were shown to induce anti-inflammatory versus proinflammatory responses respectively (5). In real life however, both PS and calreticulin are likely to be present on a given apoptotic cell (as are many other potential ligands) and a given macrophage likely expresses various receptors either for these, or their bridge molecules, and the capacity for variable inflammatory or anti-inflammatory signaling. The net effects are going to be in balance, and certainly make it extremely difficult to mimic in vitro. Additionally, as mentioned earlier, it has to be noted that all apoptotic/PCD cells are not the same. They differ substantially by cell type, mode of PCD initiation, stage of cell death, and the nature of the local environment. This is likely even more true of “necrotic” cells, even when this arises from apoptosis when the cells are not cleared (secondary necrosis) and even more in the case of various forms of cell debris in general. In passing, it should be noted that most studies of necrotic cells have involved ex-vivo dissolution of cells either by sonication, heat or freeze/thaw cycles – processes not commonly seen in vivo. Intracellular fate of ingested cells Basically, ingested PCD cells and cell debris are effectively and rapidly digested within the endosome and as a consequence, are not commonly visible in tissue sections – i.e. the inability to detect apoptotic or dying cells does not disprove some ongoing level cell removal (note (17) as an example). Macrophages are suggested to mediate rapid phagosomal maturation and content digestion compared to dendritic cells (DC), which may reflect the increased ability of the latter to present antigens to the adaptive immune system. Since DC are also believed to ingest significantly through macropinocytic mechanisms and because this may itself result in slower endosomal maturation and digestive capability, in addition to immunological implications, a careful analysis of not only the uptake processes for dying and degraded cells, but also for their intracellular digestion, would seem to be needed. Circulating neutrophils, monocytes and dendritic cells Different cell types within the blood are normally maintained at a relatively constant numbers. Since each have finite lifespans in the circulation, this implies control of source input (the bone marrow in mature mammals) and removal as well as some mechanism for sensing the circulating concentrations. For erythrocytes, the sensor appears to be local tissue oxygenation with erythropoietin serving as a rheostat for bone marrow production and release. In the case of myeloid cells (for the purpose of this essay we will focus on neutrophils, monocytes and dendritic cells) the details are much less clear. The most well studied here are neutrophils, with less known about monocytes, and even less about circulating human dendritic cells. Neutrophil life span and cell turnover in the bloodstream Studies in the 1960s using tagged cells, suggested that mature neutrophils in the blood stream have a lifespan measured in hours. While this has recently been challenged and might be extended to days (18) the cells are undoubtedly short lived. A key question for all the circulating myeloid cells is whether their time in the circulation in normal circumstances, i.e. loss from the blood stream, is driven by their emigration into the tissues or clearance in the three main removal organs, liver, spleen and bone marrow. This latter fate also raises a significant question. Do the cells in the circulation undergo apoptosis and are then recognized and removed or are they recognized in some way as old or “effete” and then induced to undergo programed cell death by interaction with cells such as macrophages in the clearance sites, or even endothelial cells in the vasculature? Equally challenging is the question of how the circulating numbers of neutrophils are regulated. While certainly fluctuating, including in a circadian fashion, the numbers in normal, non-inflammatory conditions remain within certain limits, and estimates of 109 neutrophils/kg are released and removed daily (19). Given the life span, this inevitably means rates of input and output are relatively constant under homeostatic conditions, and implies a sensor of some sort to maintain the consistency. Input into the circulation involves release of mature neutrophils from stores in the bone marrow and upstream of this, production of the cells from hematopoetic precursors. Various candidates have been proposed for the mechanisms of sensing neutrophil numbers though hard and fast conclusions are lacking. One significant site of neutrophil removal from the blood is in the bone marrow itself, thereby setting up this as the possible site of sensing and control of the numbers in the blood in a so-called neutrophil “turnstile” driven by the IL-23/IL-17/G-CSF signaling axis (20). Investigation has shown that rhythmic release of neutrophils into the circulation (triggered by cycled exposure to light in mice) is followed by their return to the bone marrow and rhythmic reductions in the size and function of the hematopoietic niche (21). This suggests a rheostat for production that is turned down based on the numbers of returning neutrophils, and as return slows, production responds by ramping up again. The mechanism appears to depend on signals from engulfing bone marrow resident macrophages. Alternatively, other investigators describe the active production of G-CSF by bone marrow macrophages in response to engulfment of returning neutrophils, thereby spurring neutrophil production and release to replace the cells that have just been removed. (22). These data may certainly be reconciled, though as yet, have not been, as representing different phases and signals of the same finely tuned rhythmic process underlying neutrophil homeostasis. While there is some increasing evidence for constituent migration of at least a few neutrophils into tissues (as well as possibly most eosinophils), nonetheless, in the absence of an inflammatory response (considered below), most of the constant daily removal of neutrophils occurs within the spleen (red pulp), liver (Kupffer cells in the sinusoids) or bone marrow – notably all sites with relative low blood flow, perhaps a requirement for the attendant phagocytes to have enough time to interact with and ingest the target neutrophil. In keeping with this hypothesis, the a role for functional β2 integrins on the neutrophils themselves has been implicated as a requirement for their proper clearance (20). As far as the role of macrophages in carrying out the actual neutrophil removal, their programming state is critical in determining their capacity for cell clearance. For example, activation of peroxisome proliferator-activated receptors (PPARs) and liver X receptors (LXRs) in macrophages, have been shown in response to the engulfment of apoptotic cells, and in turn, their activation leads to expression of downstream targets promoting the uptake and disposal of apoptotic cells (23–25). Recent data show a special role for LXRα/LXRβ in neutrophil clearance, and in turn, in the maintenance of neutrophil homeostasis. The LXRs are expressed in bone marrow macrophages in an oscillatory fashion and their expression and activation was shown to follow the wave of returning neutrophils. Genetic mutations of these receptors resulted in loss of the rhythmic modulation of the hematopoietic niche suggesting a role within the neutrophil turnstile (21. Further, genetic loss of the LXRs was also shown to be key to neutrophil removal in the periphery: LXR deficient mice showed increases in circulating neutrophil numbers, slower turnover, and accumulation in the spleen and liver under homeostatic conditions {Hong, 2012 #12). Equally, perturbation of macrophage populations, either depleting their numbers from bone marrow and spleen (26), or genetically deleting macrophage receptors for the recognition of effete cells (27), similarly leads to neutrophilia though inflammation inevitably results from these manipulations making assessment of their role in homeostasis difficult. As noted above, the signals on the neutrophil for removal are still not precisely defined, but are hypothesized to depend on the fate of the neutrophil, as part of its aging, death or activation, rather than stochastic processes (15). Aging murine neutrophils are characterized as having relatively low CD62L and high CXCR4, and appear to interact with bone marrow macrophages leading to their preferential engulfment in comparison to younger neutrophils (21). Whether the aged neutrophils are alive or undergoing death at the time of engulfment, or whether death comes after ingestion has not been fully resolved, but experimental data exist to support both possibilities. As with most cell types, there are numerous possible pathways by which neutrophils undergo programed cell death (particularly relevant to inflammatory settings discussed below) including standard, caspase-induced apoptosis. Ex vivo, without additional stimuli, neutrophils undergo apoptosis relatively rapidly and, as a consequence, expose surface markers e.g. phosphatidylserines, that are recognized by phagocytes and initiate their engulfment into phagosomes and eventual digestion. Finally, the unique role of oxidants from the neutrophil NADPH oxidase acting either directly (e.g. generation of oxidant-modified phospholipids), or indirectly via influencing the form of programmed cell death (see below) deserves consideration in the generation of signals for engulfment. Monocytes in the circulation The current concept is that there are two major monocyte types within the circulation. In the mouse these can be distinguished by expression of Ly6C, and in humans most readily by CD14 and CD16. The Ly6Clow monocytes are considered to be largely restricted to the intravascular location and have been suggested to play roles of “patrolling” the vasculature (28). The luminal crawling of Ly6Clow monocytes on the endothelium requires firm adhesion with LFA-1 (CD11ca/CD18, αLβ2) integrin and CX3CR1 to ICAM1, ICAM2 and CX3CL1 on endothelial cells (28, 29). On the other hand, circulating Ly6Chigh monocytes have the ability to extravasate into the tissues, both constituitively and in response to inflammatory conditions, with subsequent ability to either maintain their monocyte characteristics or mature in a number of different ways, including development into monocyte-derived macrophages or monocytederived dendritic cells (30) even without inflammation, as has been noted in the gut, skin and lung (31). Tracer studies using BrdU have shown that monocytes in the circulation have a lifespan of approximately 4 days for Ly6Chigh monocytes and 11 days for Ly6Clow monocytes (31–33). Once again, the triggers for eventual removal of monocytes from the circulation are not clear, whether frank apoptosis, or surface changes leading to recognition and/or stimulation of apoptosis or uptake. Major sites for removal are likely macrophages in liver and spleen as well as phagocytic cells in the bone marrow, but this too requires detailed investigation. The proportion of the migratory Ly6Chigh monocytes that are lost from the circulation as a consequence of their extravasation versus clearance from the bloodstream directly is also not known. Notably, the Ly6Chigh monocytes which traffic into non-inflamed tissues and bypass macrophage differentiation can subsequently migrate, as monocytes, down lymphatics to lymph nodes, with presumed properties that include local antigen presentation (33). Once in the lymph nodes, the final disposition of the monocytes, as for other myeloid cells gaining access to lymph nodes via the afferent lymphatics may be of some importance and will be discussed in more detail below. Dendritic cells Unlike mouse, humans have fully differentiated DCs circulating in the blood, including those of the myeloid lineage. Although many transcriptional and functional parallels between murine tissue and human intravascular DC have been assessed the clearance and turnover rate of human DCs in the blood is currently unknown. Dendritic cells DC lifespan in lymphoid and non-lymphoid organs varies depending on the DC subtype and location examined. However, most emphasis to date has been on lymphoid DC in the lymph nodes (34) (34, 35). Nonetheless, BrdU studies have demonstrated that the lifespan of myeloid DC in the skin is approximately 10–14 days, but how generally this applies to other tissues needs to be explored. Tissue resident monocyte-derived DCs were previously hypothesized to require GM-CSF as bone marrow derived DCs in vitro require GM-CSF (36). However, GM-CSF deficient mice demonstrated that monocyte-derived TNFα and iNOS producing DCs did not require GM-CSF for development in vivo (37). Therefore, what mediators drive tissue trafficking monocytes to either remain as monocytes or differentiate into macrophage-like or dendritic cell like cells is currently unknown. The exact counterpart of this dichotomy is likely common to the human situation though may not be as clearly defined as in mice. When activated to migrate from their tissue locations into the lymph node it is generally recognized that the DCs are short-lived and undergo apoptosis and removal within 2–3 days. The stimuli/conditions driving this effect again need further investigation. Furthermore, inflammation in the tissue of origin did not substantially prolong the lifespan of migratory DCs within the draining-LN (34). On the other hand, some studies have suggested that upon TLR stimulation Bcl-2/Bcl-xL is up-regulated and DC survival is prolonged. In addition, in the absence of Bcl-2 DC turnover was increased (38). It has been hypothesized that the fast turnover rate of DCs functions to regulate antigen presentation to cognate T cells, ultimately limiting immune activation. Since DCs do not exit LNs via efferent lymphatics, removal occurs within the T cell zone of the LN node, but which cell-type performs this clearance function is unclear – possibly local DC themselves but perhaps more likely, medullary macrophages or stromal cells. The recent emphasis on studies of parabiotic mice to address issues of myeloid cell origin and turnover suggest that this approach could help answer many of these questions. For example, longer-lived cells (B cells, T cells and Ly6Clow monocytes reach equilibrium between both parabiotic partners. However, due to their shorter lifespan neutrophils and Ly6Chigh monocytes do not attain equilibrium in the blood even after 120 days of parabiosis (35). It was also shown that blood-derived DC precursors (preDCs) are rapidly cleared from circulation (residence time <2 hours) and do not fully equilibrate between the parabionts, i.e behave similarly to circulating ly6Chigh monocytes. Effects of neutrophil and monocyte cell emigration into tissues and clearance from inflammatory lesions Neutrophils in inflamed tissues It has generally been assumed that neutrophils remain in in the vasculature until called into inflammatory sites and that they are subsequently removed locally in the inflamed tissues, or if not so removed, remain as a major constituent of pus. Notably, relevant in vivo studies of their accumulation and fate are less clear, in part because of the difficulties in distinguishing neutrophil persistence from newly arriving cells in the face of normal efficient clearance of the dying cells. Further, there are increasing suggestions that neutrophils may under some inflammatory circumstances migrate back into the bloodstream (39) as well as down afferent lymphatics to gain access to, and accumulate in, lymph nodes (40). We will consider the issue of clearance of myeloid cells within the lymph nodes below in a separate section. Additional routes of clearance apply after migration onto epithelial surfaces. For example, within the inflamed alveolus, much of the post-inflammatory neutrophil clearance likely occurs through phagocytic uptake and digestion by alveolar macrophages just as in tissues (41) (Figure 2).. In other instances, neutrophils exit the body and escape phagocytic recycling. An interesting special case is the constant emigration of neutrophils to the mouth through the gingival crevice with removal via the saliva (42). Similarly, emigration into the bronchi results in removal by the mucociliary escalator, and emigration through the colonic epithelium results in elimination via the gastrointestinal tract. While critical for host defense at these sites of the body that interface with the external environment, the constant loss of such neutrophils, arguably, in response to inflammatory signals of relatively low intensity, have yet to be quantified or accounted for in overall recycling models. Importantly, whatever the actual lifespan of neutrophils in the circulation may turn out to be, various forms of inflammatory stimulation of neutrophils appears to delay the “spontaneous” development of their apoptotic pathways. This has mostly been studied ex vivo, where onset of “spontaneous” apoptosis without specific added stimuli usually occurs over 18 hours. Stimulation of the cells with pro-migratory or pro-inflammatory stimuli such as chemokines or TLR agonists, extend the time before apoptosis sets in (43). This has been suggested to have functional consequences in both optimizing neutrophil protective activities in the inflammatory process but with potentially negative consequences in disease states where neutrophil lifespan is prolonged and inflammation persists. Neutrophil apoptosis has been well studied in vitro caused by both intrinsic as well as extrinsic signals, and in the important context of phagocytizing microbes (43). Standard caspasemediated apoptotic signaling pathways, nuclear condensation, exposure of PS and other recognition ligands etc are all evident. Additionally, other signals, some unique to neutrophils and subverted from host defense activities, have also been described (15). Of some interest, neutrophils, though they do release microparticles, do not appear to generate the apoptotic blebs and larger apoptotic bodies that are seen in many parenchymal cell apoptotic processes. Equally, there is some indication that the cells remain intact for longer in the apoptotic state. Whether these are general features of myeloid cell PCD is not clear, as detailed study in monocytes and macrophages is less extensive. Teleologic suggestions have been raised about the need to maintain cell membrane integrity throughout the clearance process in the face of the potentially injurious contents of these cells and their potential to cause tissue injury. Clearly in circumstances in which removal of apoptotic neutrophils by macrophages is delayed, including a number of chronic disease states (see below), the cells undergo so-called post-apoptotic, or secondary necrosis in which the integrity of the plasma membrane is lost and intracellular contents including a variety of DAMPS and proteases are released. Of note, other forms of neutrophil programmed cell death are increasingly recognized, each likely resulting in differing signals to the inflammatory milieu and consequences. Neutrophils harboring large vacuoles, likely representing autophagosomes, evident during infections and autoimmunity, are thought to undergo autophagy-associated death (44). Characteristics include mitochondrial swelling, nuclear condensation and intact plasma membranes. Though precise mechanisms are yet to be elucidated, this form of death may represent a branch point from apoptosis in which caspase activation has been inhibited by survival signals, e.g. GM-CSF. In addition, neutrophils undergo a unique form of “death”, termed “netosis” in which the nuclear membrane is disrupted, with the extrusion of chromatin both into the cytoplasm and eventually into the extracellular environment (45). Of note, the original descriptions of this process relied on stimulation of the neutrophils with the phorbol diester, PMA, and criticisms of these observations have been levied accordingly. Recent studies using more physiologic stimuli, have shown that neutrophils can extrude nuclear DNA (mitochondrial DNA extrusion has also been described) while still clearly carrying on the business of being alive; they can still chemotax and actively phagocytose microbes (46). This cellular disruption is hypothesized to play an important role in host defense demonstrated both in vitro and in vivo (46). In the context of this essay, we clearly need also to think about mechanisms by which these disrupted cells and their contents are ultimately removed, and the inflammatory consequences of signals from their intracellular contents. Reactive oxygen species almost inevitably play important roles in cell death pathways, and special mention should be made of the role of the phagocyte NADPH oxidase in determining the fate of neutrophils. For instance, activation of the oxidase by TNFα and integrin engagement appears to enhance apoptosis by enhancing mitochondrial membrane permeability, whereas high levels of oxidase activation are implicated in autophagy-associated death as well as netosis (15, 43) Notably, mutations of the NADPH oxidase that render it nonfunctional are associated with prolonged survival of neutrophils (see below). Aside from playing a role in determining the programmed cell death pathway, activation of the NADPH oxidase leads to the production of a modified phosphatidylserine, lysophosphatidylserine. This lipid accumulates both in neutrophils that are activated and fully functional, as well as those aged in culture and apoptotic (47). It signals via a G-protein coupled receptor, G2A on macrophages for engulfment of neutrophils, alive or dead (48, 49). Monocytes and macrophages Origins As discussed elsewhere in this book, the current paradigm indicates two independent sources for macrophages – embryonic progenitors in the fetal liver and yolk sac and bone marrow-derived monocytes. Most so-called “tissue-resident” macrophages derive exclusively from the former, and colonize tissues during embryogenesis (or immediately after birth). Exceptions occur in barrier tissues, namely the lungs, skin and gut where mixed ontogenies may exist. For example, in the lungs alveolar macrophages arise exclusively from embryonic precursors, whereas extraluminal macrophages (often referred to as “interstitial” macrophages) may arise from circulating monocytes (50, 51) or embryonic precursors (33). In comparison, macrophages in the dermis and gut appear to primarily derive from circulating monocytes (52, 53), Whether commensal microbiota play a role in monocyte recruitment and subsequent development at these barrier sites during homeostasis remains unclear. Certainly, during inflammatory processes the recruitment, maturation, programing and fate of monocytes and macrophages changes considerably. Accordingly, in the following paragraphs homeostasis and inflammation will be discussed separately. Tissue resident macrophages Classic examples of embryonic-derived tissue macrophages include brain microglia, liver Kupffer cells and alveolar macrophages. These macrophages replicate locally throughout post-natal life, and unless they are depleted or destroyed (e.g. by gamma radiation, clodronate liposomes, etc) they are not replaced by monocytes from the circulation. While the mechanisms that implement and regulate the either the removal or local self-renewal of these macrophages are not known, it is noteworthy that their normal lifespans (i.e. in the absence of overt inflammation) are markedly different. For example, alveolar macrophages in the mouse have been shown to persist for at least 60 days and probably longer (52, 54) whereas resident peritoneal macrophages have considerably shorter life spans (approximately 15 days) (our unpublished observations). The programing states (sometimes termed “phenotype” - though we would argue that this term is misleading) of macrophages in these sites also varies considerably. Critically, both cell turnover rates and programing states are highly influenced by the local environment. To illustrate, the alveolar macrophage exists in an extraluminal compartment where it is bathed in a lipidic film at an air-liquid interface with ambient oxygen tension. In stark contrast, peritoneal macrophages exist in a near anoxic environment that is bounded by the mesothelium. Intriguingly, peritoneal macrophages transferred into the airspaces display cell surface molecules and functional properties characteristic of alveolar macrophages within a week (55). Thus the general concept of the macrophage as a cell that is highly adaptable to its environment (i.e. exhibiting enormous plasticity) is critical. Whatever the precise signals and processes involved, the environment is clearly paramount in determining macrophage life history. For the moment we must assume that the eventual death and removal of these resident cells likely will involve some form of programed cell death, but the triggers here for normal turnover are completely unknown, as to a significant degree are the removal processes. A little more is known for monocyte-derived macrophages that accumulate in tissues during the process of inflammation and thus removal of these cells is discussed in more detail below and contrasted where possible with the resident cell type. Inflammatory macrophages During inflammation, mononuclear phagocyte numbers increase dramatically. Two sources account for this – local replication of resident tissue macrophages and recruitment of monocytes from the circulation. As noted above, some of the monocytes that emigrate into an inflammatory site undergo maturational changes that result in their legitimate consideration as macrophages. This includes a huge increase in cell constituent synthesis, resulting a much larger cell, cytoplasm to nuclear ratio, increased numbers of mitochondria, lysosomes, membranes as well as a number of surface markers. As a consequence, in any inflammatory lesion there is a spectrum of cells in various stages of this maturation that are often lumped together under the general term “inflammatory monocytes”. Importantly, the relative contribution provided to the total mononuclear phagocyte pool by resident tissue macrophages versus recruited mononuclear phagocytes varies by anatomic site and is greatly influenced by the nature and magnitude of the stimulus. For example, administration of LPS into the lungs leads to massive recruitment of circulating monocytes yet induces only modest proliferation of resident alveolar macrophages (52). Conversely pleural infection with helminthic organisms drives marked proliferation of resident pleural macrophages but only minimal recruitment of monocytes from the circulation (56). Additionally, as discussed at length elsewhere in this book, the macrophages that exist in inflammatory conditions exhibit a variety of environment-induced programing states that change considerably over time and that are often associated in a contributory way with the progression of inflammation at its early stages and later with its resolution. Critically, as noted in Figure 3, these monocyte-derived macrophages as well as the increased numbers of monocytes accumulated in inflamed tissues, are eventually cleared as the inflammation resolves. Resolution of Inflammation As inflammation resolves, tissue macrophage and monocyte numbers return to homeostatic levels. The question of where the “inflammatory macrophages” go is long-standing and only partially answered. In basic terms, three potential fates exist; migration from the site, in situ death, and in the case of tissues that contact the external environment (such as the respiratory tract, gut lumen and skin) extracorporeal elimination (e.g. expectoration, sloughing, etc). As described above, monocytes clearly have the ability to migrate from tissues using the lymphatics as a conduit (33). However, whether monocytes can also exit tissues by direct egress back into the bloodstream remains unclear. That Macrophages emigrate from the peritoneum via the lymphatics has also been reported (57, 58). As mentioned above, monocyte maturation occurs across a continuum and it is likely that at least some of these emigrating macrophages represent inflammatory monocytes that have gained cell surface molecules typically associated with macrophages. In situ death of macrophages and subsequent efferocytic removal represents a clear mode of exit, and in some instances provides a critical means of host defense by eliminating ingested pathogens. Notably, in the peritoneum and lungs, macrophage death appears to be the dominant mechanism for macrophage elimination during resolving inflammation, with lymphatic egress playing only a minor role (52, 59). As mentioned above, numerous pathways for death exist, including apoptosis, autophagy, necroptosis and frank necrosis – all of which may be operative in driving elimination of macrophages from inflammatory sites. Critically, both the mode of death and the frequency with which cells die are likely to be influenced by the local environment tissue-specific factors, and perhaps even the origin of the cell itself. In this regard, studies of resolving inflammation in the peritoneum (59) and lungs (52) demonstrate that recruited monocyte-derived macrophages undergo programmed cell death, while at the same time resident macrophages (i.e. embryonic derived macrophages) persist – even though both populations of cells exist in the same environment. The developing theme of fundamental (transcription factor dependent) differences between embryonic derived self-renewing “resident” macrophages and the monocyte derived inflammatory macrophages under consideration here, may in fact be somewhat blurred and in need of more investigation. Thus, although study of resident alveolar macrophages shows their considerable ability to survive throughout the course of an acute inflammatory reaction, there is nonetheless some loss of these cells that are eventually replaced, in part apparently from a proportion of the monocyte-derived cells (52). Intriguingly, gene expression comparison of these, versus the persistent original “resident” macrophages shows that over time the monocyte derived cells became almost identical with the resident cells, differing in only a few genes (which did include some of the relevant transcription factors) (unpublished or some other attribution?). This observation, which is likely mirrored at other sites, such as the peritoneum, makes a strong case for the influence of the environment in molding the character of the local macrophage, whatever its proximal origin. The frequency of apoptotic macrophages that can be detected in resolving inflammatory lesions is generally low, suggesting their rapid removal and digestion. However, the question of “Who clears the dying macrophages?” remains largely unanswered. Both the remaining macrophages and non-professional phagocytes, such as epithelial cells have the capacity to engulf dying cells, though quantitative studies describing their relative contributions are lacking. The answer will almost certainly be organ/site specific and may also vary with the inflammatory stimulus. In any case, as described above for neutrophils, the process appears to involve exposure of “eat me” signals on the dying cell’s surface. Again, phosphatidylserine (PS) most likely plays a major role, although additional apoptotic-cell associated molecular patterns (ACAMPs) including changes in surface charge and glycosylation, and exposure of calreticulin on the cell surface may also be operative. Notably, activated macrophages can express PS on their surfaces, even when they are not apoptotic (60), thereby suggesting that PS exposure alone is insufficient to drive engulfment and that loss of “don’t eat me” signals (e.g. CD47, CD31) on the surface of the dying cell may also be required. The many molecules used by macrophages to recognize dying cells are described in detail elsewhere (5–7). However, it must be noted that the most of our knowledge derives either from studies performed in the steady state (as described above) or during the acute neutrophilic phase of inflammation. The local environment changes significantly as inflammation resolves, and it is therefore likely that phagocytic receptor expression and uptake mechanisms are also likely to change. Disposition of myeloid cells migrating to the lymph nodes One central theme in the removal of a number of myeloid cell types from tissues is migration/clearance down the lymphatics. Thus as noted, this appears to be one mode of removal of macrophages from the inflamed peritoneum and is clearly relevant for disposition of tissue migratory DC or constituitively migrating monocytes. However, this process only kicks the can down the street. Are the cells then removed in the local lymph node? Do some even pass on down the chain to further lymph nodes or even into the blood stream via the lymphatic ducts. We suggest that this last pathway for removal seems unlikely since mature macrophages are of substantial size and will not pass readily through capillary beds. Basically, programed cell death and local uptake by phagocytes in the lymph nodes seems most likely for this secondary process of clearance from the tissues. As noted above, the life span of migratory myeloid DC that have reached the lymph node is no more than a few days. However, reflecting a constant theme in this essay, the precise stimuli that initiate this, and the cells (macrophages, stromal cells, etc) and recognition processes involved in their subsequent phagocytic removal have not been clearly defined. Nonetheless, we must suppose this to be a quantitatively substantial clearance process with significant relevance for maintenance of tissue homeostasis. Defects and Disease To no great surprise, systems as essential to normal development and homeostasis are not immune from inherent or induced defects and these can include loss of function, as well as gain of function effects. In the case of myeloid cells, their additional role in the removal of damaged cells and debris leads to potential defects in both their own removal as well as their ability to mediate the removal process. A few examples will suffice to illustrate these points, but one can expect many others and a significant contribution to a wide variety of disease processes. Defects in removal of inflammatory myeloid cells. Chronic Obstructive Pulmonary Disease (COPD) In any inflammatory condition, inability to mediate the rapid and highly efficient removal of the accumulated neutrophils and monocytes would be expected to, and does, prolong the inflammation. Such effects can be caused by the effects of stimuli to the inflammatory cells that block their normal induction of programed cell death as well as in defects of the removal process itself. As noted above, neutrophil apoptosis can indeed be delayed by a host of potential inflammatory mediators as well as by stimulation of pattern recognition molecules from external or host-derived sources. We bring up COPD in this context because the long-term destruction of the lung tissue underlying this extremely common and important disease has been suggested to result from persistent low-grade inflammatory, especially neutrophil-driven, injury. Cigarette smoke, either directly or indirectly in a feed-forward inflammatory mediatorinduced fashion can prolong the life of neutrophils leading to accumulation in the lung tissue. However, it has additionally become apparent that COPD is associated with defects in the ability of macrophages in this tissue to recognize and remove apoptotic cells (i.e. the neutrophils), thereby adding an additional component to persistence of inflammation and its contribution to tissue destruction (61–63). Fundamentally, this example leads one to raise a challenge to the general possibility of similar feedback loops contributing to many forms of chronic inflammation, whether the targets are neutrophils selectively, or inflammatory myeloid (or lymphoid) cells more generally. In each case the mechanisms underlying the defects may be different, and may be either external or internal (or both), but, we would argue, are worth further exploration in the hope of ameliorating the ongoing process. Chronic Granulomatous Disease, CGD Another example, in this case with a genetic underpinning, is CGD. Various defects in subunits of the NADPH oxidase lead to loss of the oxidative burst (particularly relevant in neutrophils) with concomitant defects in bacterial and fungal host defense. However, in addition to problems with constant infections, the patients also exhibit persistent inflammatory and granulomatous problems (hence the name). Relevant to the current discussions, the defective neutrophil NADPH oxidase results in lack of normal changes on their surface during inflammation such as generation of lysophosphatidylserine that usually render the cells more palatable for uptake and removal by macrophages (48, 64). In addition, the genetic defects are also associated with altered programing of the macrophages towards decreased ability to ingest apoptotic cells. Consequently, these two effects act together to result in defective neutrophil clearance, and thus, persistent inflammation. Notably, the effect is seen most clearly in the context of inflammation, i.e. when the NADPH oxidase is normally activated, and to a much lower extent, at the level of normal neutrophil turnover in the blood. Autoimmune processes including Systemic Lupus Erythematosus (SLE) An intriguing observation that pervades the studies of mechanisms and extent of apoptotic cell clearance is that defects in these processes, for example loss of key receptors for apoptotic cell recognition and uptake, are frequently associated with generalize autoimmune disease. Examples include mice lacking the scavenger receptor CD36, bridge molecules MFG-E8 and C1q, Mer tyrosine kinase, and G-protein-coupled receptor kinase 6 (GRK6, (65)) which normally enhances apoptotic cell engulfment through Rac1 activation. When coupled with observations of defective apoptotic cell clearance in human systemic lupus erythematosus (SLE) the hypothesis that inefficient clearance, of for example dying neutrophils, results release of intracellular constituents such as nuclear contents that can then initiate autoimmune responses (66). Note also that CGD patients and mice, again showing inefficient clearance of neutrophils, are also prone over time to develop autoimmunity (67, 68). The extent to which defects in normal cell removal are generally associated with autoimmune responses and the additional controls that may serve to limit this extent are of considerable interest and potential importance. Hemophagocytic lymphohistiocytosis (HLH) In the other direction, a group of syndromes often termed HLH is associated with increased phagocytic uptake and removal of blood cells, including myeloid cells, throughout the body (69, 70). These can be either developmental or associated with viral infections and are accompanied and presumably induced by very high levels of systemic cytokines, particularly interferon gamma. Specific mechanisms by which the leukocytes (and erythrocytes) are altered or become recognizable by the phagocytes are not clear. The macrophages involved in the ingestion are likely substantially over stimulated by the cytokine storm, but whether the loss of normal discrimination against uptake of blood leukocytes is solely at the level of the targets or also in the functions of the phagocytes is also unclear. While the clinical conditions themselves are rare, and animal models are limited (69) full exploration of the underlying mechanisms would seem to have potential for general understanding of the controls normally involved in leukocyte maintenance and removal. Implications and conclusions – maintenance of homeostasis We are consistently impressed by the extraordinary variety of forms and mechanisms by which cells can be induced to undergo programed death, variety that is certainly manifest by the myeloid lineage as well as within other cell categories. This extensive repertoire extends also to the changes on the cell surface, or on cell debris that results, leading to recognition by phagocytes and mediating their removal. Increasingly it is also becoming apparent that these different ligands likely stimulate many different uptake mechanisms in the phagocytes. A sense of enormous redundancy results, suggesting the substantial importance to a multicellular organism of being efficient in its ability to remove cells no longer needed or undergoing aging, damage, abnormal function or death i.e. to maintain homeostasis. On the other hand, it also emphasizes the need consider each cell type, situation and environment separately, and explicitly, where possible, to study the removal processes in situ in the relevant environment in vivo. Not surprisingly, increasing understanding of the processes of myeloid cell removal has led to consideration of therapeutic approaches to its enhancement, for example in inflammation. Given the complexities noted above, these approaches are in their infancy and fraught with not only the redundancy problem, but also contrasting concerns relating to the balance between protective versus damaging effects of inflammatory leukocytes. We note only two potential examples here (of many) to illustrate the possibilities as well as issues involved. In our own studies, stimulation of PPARγ with pioglitazone is showing intriguing promise in overcoming some of the defects in neutrophil removal seen in chronic granulomatous disease by altering both the palatability of the neutrophils for uptake as well as the effectiveness of their uptake by the macrophages (unpublished observations). In another example, a flavone (wogonin) inducer of eosinophil apoptosis was recently reported to have attenuating effects on airways inflammation and allergic responses (71, 72). Other areas of investigation that would seem appropriate to this discussion and that have also perhaps not received the attention that would now be due, are the effects of aging of the animal and human on the clearance processes and at a cellular level, the effects of senescence of the cells (or precursors) themselves. In the case of the myeloid cells, this could apply both to the cells destined for removal as well as those doing the removal itself. Supported by: HL114381, HL109517, AI110408, HL34303, HL115334 Figure 1 Mechanisms for recognition and uptake of apoptotic cells Changes on the apoptotic cell surface, including exposure of phosphatidylserine and other normally internally located molecules are recognized by surface receptors on the phagocyte leading to tethering of apoptotic cell and transduction of uptake signals. Bridge molecules (opsonins) in the environment or produced by the phagocyte may also recognize the apoptotic surface changes and also a different set of receptors on the phagocyte to initiate tethering and/or signaling. These processes are significantly redundant and also highly regulated, by enhancing or inhibitory stimuli. Viable cells may also avoid removal by expressing “don’t eat me” stimuli that block the recognition and/or uptake processes. Figure 2 Uptake of apoptotic neutrophils by a macrophage in the alveolar air space during resolution of inflammation in the lung Figure 3 Time course of a standardized acute inflammatory response in the lung showing accumulation and removal of neutrophils and monocyte/macrophages The resident alveolar macrophages persist throughout the inflammation and do not substantially change in numbers. Some of the recruited Ly6chigh monocytes mature into macrophages (and DC) and some remain as monocytes. The macrophages undergo a variety of programing changes during the course of the inflammation participating in first its initiation and then its resolution. The monocytes and macrophages are cleared as the inflammation wanes, mostly by undergoing PCD and engulfment, though some of the monocytes migrate to the local lymph nodes and, at this site, some of the cells may also be cleared physically up the airways by the mucociliary escalator. 1 Fadok VA Voelker DR Campbell PA Cohen JJ Bratton DL Henson PM Exposure of phosphatidylserine on the surface of apoptotic lymphocytes triggers specific recognition and removal by macrophages J Immunol 1992 148 2207 2216 1545126 2 Takatsu H Tanaka G Segawa K Suzuki J Nagata S Nakayama K Shin HW Phospholipid flippase activities and substrate specificities of human type IV P-type ATPases localized to the plasma membrane J Biol Chem 2014 289 33543 33556 25315773 3 Suzuki J Nagata S Phospholipid scrambling on the plasma membrane Methods Enzymol 2014 544 381 393 24974298 4 Segawa K Nagata S An Apoptotic 'Eat Me' Signal: Phosphatidylserine Exposure Trends Cell Biol 2015 25 639 650 26437594 5 Gardai SJ Bratton DL Ogden CA Henson PM Recognition ligands on apoptotic cells: a perspective J Leukoc Biol 2006 79 896 903 16641135 6 Poon IK Lucas CD Rossi AG Ravichandran KS Apoptotic cell clearance: basic biology and therapeutic potential Nat Rev Immunol 2014 14 166 180 24481336 7 Ravichandran KS Beginnings of a good apoptotic meal: the find-me and eat-me signaling pathways Immunity 2011 35 445 455 22035837 8 Yoshida H Kawane K Koike M Mori Y Uchiyama Y Nagata S Phosphatidylserine-dependent engulfment by macrophages of nuclei from erythroid precursor cells Nature 2005 437 754 758 16193055 9 Birge RB Ucker DS Innate apoptotic immunity: the calming touch of death Cell Death Differ 2008 15 1096 1102 18451871 10 D'Mello V Birge RB Apoptosis: conserved roles for integrins in clearance Curr Biol 2010 20 R324 R327 20392425 11 Elliott MR Ravichandran KS Clearance of apoptotic cells: implications in health and disease J Cell Biol 2010 189 1059 1070 20584912 12 deCathelineau AM Henson PM The final step in programmed cell death: phagocytes carry apoptotic cells to the grave Essays Biochem 2003 39 105 117 14585077 13 Hoffmann PR deCathelineau AM Ogden CA Leverrier Y Bratton DL Daleke DL Ridley AJ Fadok VA Henson PM Phosphatidylserine (PS) induces PS receptor-mediated macropinocytosis and promotes clearance of apoptotic cells J Cell Biol 2001 155 649 659 11706053 14 Mazzoni F Safa H Finnemann SC Understanding photoreceptor outer segment phagocytosis: use and utility of RPE cells in culture Exp Eye Res 2014 126 51 60 24780752 15 Bratton DL Henson PM Neutrophil clearance: when the party is over, clean-up begins Trends in Immunology 2011 32 350 357 21782511 16 Erwig LP Henson PM Immunological consequences of apoptotic cell phagocytosis Am J Pathol 2007 171 2 8 17591947 17 Scott RS McMahon EJ Pop SM Reap EA Caricchio R Cohen PL Earp HS Matsushima GK Phagocytosis and clearance of apoptotic cells is mediated by MER Nature 2001 411 207 211 11346799 18 Pillay J den Braber I Vrisekoop N Kwast LM de Boer RJ Borghans JA Tesselaar K Koenderman L In vivo labeling with 2H2O reveals a human neutrophil lifespan of 5.4 days Blood 2010 116 625 627 20410504 19 Summers C Rankin SM Condliffe AM Singh N Peters AM Chilvers ER Neutrophil kinetics in health and disease Trends Immunol 2010 31 318 324 20620114 20 Stark MA Huo Y Burcin TL Morris MA Olson TS Ley K Phagocytosis of apoptotic neutrophils regulates granulopoiesis via IL-23 and IL-17 Immunity 2005 22 285 294 15780986 21 Casanova-Acebes M Pitaval C Weiss LA Nombela-Arrieta C Chevre R N AG Kunisaki Y Zhang D van Rooijen N Silberstein LE Weber C Nagasawa T Frenette PS Castrillo A Hidalgo A Rhythmic modulation of the hematopoietic niche through neutrophil clearance Cell 2013 153 1025 1035 23706740 22 Furze RC Rankin SM The role of the bone marrow in neutrophil clearance under homeostatic conditions in the mouse FASEB Journal 2008 22 3111 3119 18509199 23 A-Gonzalez N Bensinger SJ Hong C Beceiro S Bradley MN Zelcer N Deniz J Ramirez C Diaz M Gallardo G de Galarreta CR Salazar J Lopez F Edwards P Parks J Andujar M Tontonoz P Castrillo A Apoptotic cells promote their own clearance and immune tolerance through activation of the nuclear receptor LXR Immunity 2009 31 245 258 19646905 24 Majai G Sarang Z Csomos K Zahuczky G Fesus L PPARgamma-dependent regulation of human macrophages in phagocytosis of apoptotic cells Eur J Immunol 2007 37 1343 1354 17407194 25 Mukundan L Odegaard JI Morel CR Heredia JE Mwangi JW Ricardo-Gonzalez RR Goh YP Eagle AR Dunn SE Awakuni JU Nguyen KD Steinman L Michie SA Chawla A PPAR-delta senses and orchestrates clearance of apoptotic cells to promote tolerance Nat Med 2009 15 1266 1272 19838202 26 Gordy C Pua H Sempowski GD He YW Regulation of steady-state neutrophil homeostasis by macrophages Blood 2011 117 618 629 20980680 27 Hong C Kidani Y N AG Phung T Ito A Rong X Ericson K Mikkola H Beaven SW Miller LS Shao WH Cohen PL Castrillo A Tontonoz P Bensinger SJ Coordinate regulation of neutrophil homeostasis by liver X receptors in mice J Clin Invest 2012 122 337 347 22156197 28 Auffray C Fogg D Garfa M Elain G Join-Lambert O Kayal S Sarnacki S Cumano A Lauvau G Geissmann F Monitoring of blood vessels and tissues by a population of monocytes with patrolling behavior Science 2007 317 666 670 17673663 29 Carlin LM Stamatiades EG Auffray C Hanna RN Glover L Vizcay-Barrena G Hedrick CC Cook HT Diebold S Geissmann F Nr4a1-dependent Ly6C(low) monocytes monitor endothelial cells and orchestrate their disposal Cell 2013 153 362 375 23582326 30 Guilliams M Ginhoux F Jakubzick C Naik SH Onai N Schraml BU Segura E Tussiwand R Yona S Dendritic cells, monocytes and macrophages: a unified nomenclature based on ontogeny Nat Rev Immunol 2014 14 571 578 25033907 31 Yona S Kim KW Wolf Y Mildner A Varol D Breker M Strauss-Ayali D Viukov S Guilliams M Misharin A Hume DA Perlman H Malissen B Zelzer E Jung S Fate mapping reveals origins and dynamics of monocytes and tissue macrophages under homeostasis Immunity 2013 38 79 91 23273845 32 Hettinger J Richards DM Hansson J Barra MM Joschko AC Krijgsveld J Feuerer M Origin of monocytes and macrophages in a committed progenitor Nat Immunol 2013 14 821 830 23812096 33 Jakubzick C Gautier EL Gibbings SL Sojka DK Schlitzer A Johnson TE Ivanov S Duan Q Bala S Condon T van Rooijen N Grainger JR Belkaid Y Ma'ayan A Riches DW Yokoyama WM Ginhoux F Henson PM Randolph GJ Minimal differentiation of classical monocytes as they survey steady-state tissues and transport antigen to lymph nodes Immunity 2013 39 599 610 24012416 34 Kamath AT Henri S Battye F Tough DF Shortman K Developmental kinetics and lifespan of dendritic cells in mouse lymphoid organs Blood 2002 100 1734 1741 12176895 35 Liu K Waskow C Liu X Yao K Hoh J Nussenzweig M Origin of dendritic cells in peripheral lymphoid organs of mice Nat Immunol 2007 8 578 583 17450143 36 Serbina NV Salazar-Mather TP Biron CA Kuziel WA Pamer EG TNF/iNOS-producing dendritic cells mediate innate immune defense against bacterial infection Immunity 2003 19 59 70 12871639 37 Greter M Helft J Chow A Hashimoto D Mortha A Agudo-Cantero J Bogunovic M Gautier EL Miller J Leboeuf M Lu G Aloman C Brown BD Pollard JW Xiong H Randolph GJ Chipuk JE Frenette PS Merad M GM-CSF controls nonlymphoid tissue dendritic cell homeostasis but is dispensable for the differentiation of inflammatory dendritic cells Immunity 2012 36 1031 1046 22749353 38 Chen M Huang L Shabier Z Wang J Regulation of the lifespan in dendritic cell subsets Mol Immunol 2007 44 2558 2565 17267035 39 Woodfin A Voisin MB Beyrau M Colom B Caille D Diapouli FM Nash GB Chavakis T Albelda SM Rainger GE Meda P Imhof BA Nourshargh S The junctional adhesion molecule JAM-C regulates polarized transendothelial migration of neutrophils in vivo Nat Immunol 2011 12 761 769 21706006 40 Chtanova T Schaeffer M Han SJ van Dooren GG Nollmann M Herzmark P Chan SW Satija H Camfield K Aaron H Striepen B Robey EA Dynamics of neutrophil migration in lymph nodes during infection Immunity 2008 29 487 496 18718768 41 McCubbrey AL Curtis JL Efferocytosis and lung disease Chest 2013 143 1750 1757 23732585 42 Scott DA Krauss J Neutrophils in periodontal inflammation Front Oral Biol 2012 15 56 83 22142957 43 Geering B Stoeckle C Conus S Simon HU Living and dying for inflammation: neutrophils, eosinophils, basophils Trends Immunol 2013 34 398 409 23665135 44 Mihalache CC Simon HU Autophagy regulation in macrophages and neutrophils Exp Cell Res 2012 318 1187 1192 22245582 45 Brinkmann V Reichard U Goosmann C Fauler B Uhlemann Y Weiss DS Weinrauch Y Zychlinsky A Neutrophil extracellular traps kill bacteria Science 2004 303 1532 1535 15001782 46 Yipp BG Petri B Salina D Jenne CN Scott BN Zbytnuik LD Pittman K Asaduzzaman M Wu K Meijndert HC Malawista SE de Boisfleury Chevance A Zhang K Conly J Kubes P Infection-induced NETosis is a dynamic process involving neutrophil multitasking in vivo Nat Med 2012 18 1386 1393 22922410 47 Frasch SC Berry KZ Fernandez-Boyanapalli R Jin HS Leslie C Henson PM Murphy RC Bratton DL NADPH oxidase-dependent generation of lysophosphatidylserine enhances clearance of activated and dying neutrophils via G2A J Biol Chem 2008 283 33736 33749 18824544 48 Frasch SC Fernandez-Boyanapalli RF Berry KA Murphy RC Leslie CC Nick JA Henson PM Bratton DL Neutrophils regulate tissue neutrophilia in inflammation via the oxidant-modified lipid lysophosphatidylserine Journal of Biological Chemistry 2013 288 4583 4593 23293064 49 Frasch SC Fernandez-Boyanapalli RF Berry KZ Leslie CC Bonventre JV Murphy RC Henson PM Bratton DL Signaling via macrophage G2A enhances efferocytosis of dying neutrophils by augmentation of Rac activity J Biol Chem 2011 286 12108 12122 21297111 50 Schulz C Gomez Perdiguero E Chorro L Szabo-Rogers H Cagnard N Kierdorf K Prinz M Wu B Jacobsen SE Pollard JW Frampton J Liu KJ Geissmann F A lineage of myeloid cells independent of Myb and hematopoietic stem cells Science 2012 336 86 90 22442384 51 Scott CL Henri S Guilliams M Mononuclear phagocytes of the intestine, the skin, and the lung Immunol Rev 2014 262 9 24 25319324 52 Janssen WJ Barthel L Muldrow A Oberley-Deegan RE Kearns MT Jakubzick C Henson PM Fas determines differential fates of resident and recruited macrophages during resolution of acute lung injury Am J Respir Crit Care Med 2011 184 547 560 21471090 53 Bain CC Bravo-Blas A Scott CL Gomez Perdiguero E Geissmann F Henri S Malissen B Osborne LC Artis D Mowat AM Constant replenishment from circulating monocytes maintains the macrophage pool in the intestine of adult mice Nat Immunol 2014 15 929 937 25151491 54 Murphy J Summer R Wilson AA Kotton DN Fine A The prolonged lifespan of alveolar macrophages Am J Respir Cell Mol Biol 2008 38 380 385 18192503 55 Guth AM Janssen WJ Bosio CM Crouch EC Henson PM Dow SW Lung environment determines unique phenotype of alveolar macrophages Am J Physiol Lung Cell Mol Physiol 2009 296 L936 L946 19304907 56 Jenkins SJ Ruckerl D Cook PC Jones LH Finkelman FD van Rooijen N MacDonald AS Allen JE Local macrophage proliferation, rather than recruitment from the blood, is a signature of TH2 inflammation Science 2011 332 1284 1288 21566158 57 Bellingan GJ Caldwell H Howie SE Dransfield I Haslett C In vivo fate of the inflammatory macrophage during the resolution of inflammation: inflammatory macrophages do not die locally, but emigrate to the draining lymph nodes J Immunol 1996 157 2577 2585 8805660 58 Cao C Lawrence DA Strickland DK Zhang L A specific role of integrin Mac-1 in accelerated macrophage efflux to the lymphatics Blood 2005 106 3234 3241 16002427 59 Gautier EL Ivanov S Lesnik P Randolph GJ Local apoptosis mediates clearance of macrophages from resolving inflammation in mice Blood 2013 122 2714 2722 23974197 60 Marguet D Luciani MF Moynault A Williamson P Chimini G Engulfment of apoptotic cells involves the redistribution of membrane phosphatidylserine on phagocyte and prey Nat Cell Biol 1999 1 454 456 10559991 61 Petrusca DN Gu Y Adamowicz JJ Rush NI Hubbard WC Smith PA Berdyshev EV Birukov KG Lee CH Tuder RM Twigg HL 3rd Vandivier RW Petrache I Sphingolipid-mediated inhibition of apoptotic cell clearance by alveolar macrophages J Biol Chem 2010 285 40322 40332 20956540 62 Henson PM Vandivier RW Douglas IS Cell death, remodeling, and repair in chronic obstructive pulmonary disease? Proc Am Thorac Soc 2006 3 713 717 17065379 63 Hamon R Homan CC Tran HB Mukaro VR Lester SE Roscioli E Bosco MD Murgia CM Ackland ML Jersmann HP Lang C Zalewski PD Hodge SJ Zinc and zinc transporters in macrophages and their roles in efferocytosis in COPD PLoS One 2014 9 e110056 25350745 64 Fernandez-Boyanapalli R Frasch SC Riches DW Vandivier RW Henson PM Bratton DL PPARgamma activation normalizes resolution of acute sterile inflammation in murine chronic granulomatous disease Blood 2010 116 4512 4522 20693431 65 Nakaya M Tajima M Kosako H Nakaya T Hashimoto A Watari K Nishihara H Ohba M Komiya S Tani N Nishida M Taniguchi H Sato Y Matsumoto M Tsuda M Kuroda M Inoue K Kurose H GRK6 deficiency in mice causes autoimmune disease due to impaired apoptotic cell clearance Nat Commun 2013 4 1532 23443560 66 Munoz LE Janko C Schulze C Schorn C Sarter K Schett G Herrmann M Autoimmunity and chronic inflammation - two clearance-related steps in the etiopathogenesis of SLE Autoimmun Rev 2010 10 38 42 20817127 67 Rupec RA Petropoulou T Belohradsky BH Walchner M Liese JG Plewing G Messer G Lupus erythematosus tumidus and chronic discoid lupus erythematosus in carriers of X-linked chronic granulomatous disease Eur J Dermatol 2000 10 184 189 10725815 68 Sanford AN Suriano AR Herche D Dietzmann K Sullivan KE Abnormal apoptosis in chronic granulomatous disease and autoantibody production characteristic of lupus Rheumatology (Oxford) 2006 45 178 181 16249245 69 Brisse E Wouters CH Matthys P Hemophagocytic lymphohistiocytosis (HLH): A heterogeneous spectrum of cytokine-driven immune disorders Cytokine Growth Factor Rev 2014 70 Usmani GN Woda BA Newburger PE Advances in understanding the pathogenesis of HLH Br J Haematol 2013 161 609 622 23577835 71 Lucas CD Dorward DA Sharma S Rennie J Felton JM Alessandri AL Duffin R Schwarze J Haslett C Rossi AG Wogonin induces eosinophil apoptosis and attenuates allergic airway inflammation Am J Respir Crit Care Med 2015 191 626 636 25629436 72 Persson C Drug-induced Death of Eosinophils. Promises and Pitfalls Am J Respir Crit Care Med 2015 191 605 606 25767918
PMC005xxxxxx/PMC5123759.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101651677 43556 Nat Plants Nat Plants Nature plants 2055-0278 27797352 5123759 10.1038/nplants.2016.169 NIHMS820857 Article The DNA demethylase ROS1 targets genomic regions with distinct chromatin modifications Tang Kai 1# Lang Zhaobo 12# Zhang Heng 2 Zhu Jian-Kang 12 1 Department of Horticulture and Landscape Architecture, Purdue University, West Lafayette, IN 47907, USA 2 Shanghai Center for Plant Stress Biology, and Center for Excellence in Molecular Plant Sciences, Chinese Academy of Sciences, Shanghai 201602, China Correspondence to: Zhaobo Lang ([email protected]) # Equal contribution Additional information Correspondence and requests for materials should be addressed to Z.L. 8 10 2016 31 10 2016 31 10 2016 30 4 2017 2 11 1616916169 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. The Arabidopsis ROS1/DEMETER family of 5mC DNA glycosylases are the first genetically characterized DNA demethylases in eukaryotes. However, the features of ROS1 targeted genomic loci are not well-understood. In this study, we characterized ROS1 target loci in Arabidopsis Col-0 and C24 ecotypes. We found that ROS1 preferentially targets transposable elements (TEs) and intergenic regions. Compared to most TEs, ROS1-targeted TEs are closer to protein coding genes, suggesting that ROS1 may prevent DNA methylation spreading from TEs to nearby genes. ROS1-targeted TEs are specifically enriched for H3K18Ac and H3K27me3, and depleted of H3K27me and H3K9me2. Importantly, we identified thousands of previously unknown RNA-directed DNA methylation (RdDM) targets upon depletion of ROS1, suggesting that ROS1 strongly antagonizes RdDM at these loci. In addition, we show that ROS1 also antagonizes RdDM-independent DNA methylation at some loci. Our results provide important insights into the genome-wide targets of ROS1 and the crosstalk between DNA methylation and ROS1-mediated active DNA demethylation. Introduction 5-methylcytosine (5mC) is an important epigenetic mark present in many eukaryotes, and is involved in many crucial biological processes, such as gene imprinting, regulation of gene expression, and genome stability 1–3. In plants, DNA methylation frequently occurs in three sequence contexts, including CG, CHG, and CHH (H represents either A, C or T). In Arabidopsis, DNA methylation is established and maintained by different pathways. DNA methylation (de novo methylation) is established by domains rearranged methyltransferase2 (DRM2), through the RNA-directed DNA methylation (RdDM) pathway, in which RNA polymerase IV (Pol IV)-dependent 24-nt small interfering RNAs (siRNAs) function to guide DRM2 to target loci. Recently, it was found that Pol IV-dependent 25–35nt precursors of the 24-nt siRNAs can trigger DNA methylation independently of the 24-nt siRNAs 4–7. Four different enzymes maintain DNA methylation after DNA replication, depending on the sequence context: mCG is maintained by DNA methyltransferase 1 (MET1), mCHG is maintained by chromomethylase3 (CMT3), and mCHH is maintained by CMT2 and DRM2. DNA methylation levels are dynamically regulated. DNA methylation can be passively lost due to lack of maintenance methylation, or can be actively removed by DNA demethylases 3. In plants, active DNA demethylation is initiated by the ROS1/Demeter family of proteins. ROS1 is the first genetically characterized DNA demethylase (the first enzyme in the active DNA demethylation pathway) in eukaryotes8. ROS1 can remove the 5mC base and nick the DNA backbone, leaving a single nucleotide gap that is filled with an unmethylated cytosine through a base excision repair pathway 9–13. An anti-silencing protein complex containing methyl-DNA binding protein 7 (MBD7), increased DNA methylation 1 (IDM1), IDM2, and IDM3, was recently discovered to regulate ROS1 targeting, and in turn DNA demethylation 14–16. In Arabidopsis, some TEs show lower expression levels in ros1 mutants due to increased DNA methylation 15,17. Some genes are silenced in ros1 mutants due to DNA hypermethylation of nearby TEs 14,17–19, suggesting that ROS1-mediated demethylation of TEs is important for regulation of gene expression by preventing nearby genes from being silenced. The methylomes of an Arabidopsis ros1/dml2/dml3 (rdd) triple mutant 25 and of a ros1 single mutant 14 have been analyzed, revealing thousands of genomic regions subjected to active DNA demethylation by ROS114. However, the features of ROS1 targets are not known. Previous studies suggested interactions between ROS1-mediated DNA demethylation and RdDM 20–22. Recently, studies revealed that the expression of ROS1 is finely tuned by RdDM and DNA demethylation pathways23,24, although the genome-wide interaction between ROS1-mediated DNA demethylation and RdDM has not been investigated. In this study, we generated and analyzed the DNA methylomes of ros1 mutants in both Col-0 and C24 ecotypes of Arabidopsis, and characterized ROS1 target loci in these two genetic backgrounds. Our analyses identified and characterized thousands of genomic loci that are regulated by both ROS1 and RdDM. Interestingly, we discovered thousands of previously unidentified RdDM targets by analyzing the DNA methylome of ros1/nrpd1 double mutant plants that are defective in both active DNA demethylation and RdDM. In addition, we show that besides antagonizing RdDM, ROS1 can also antagonize RdDM-independent DNA methylation at over a thousand genomic loci. These results provide important insights into the genome-wide effect of ROS1-mediated active DNA demethylation and the interaction between DNA demethylation and methylation in plants. Results Characterization of ros1 mutant methylomes in Col-0 and C24 ecotypes ros1-4 is an Arabidopsis mutant of Col-0 ecotype with T-DNA insertion in the ROS1 gene, causing complete loss of function of ROS1 26. ros1-1 is a loss-of-function mutant of ROS1 in C24 ecotype and has a single nucleotide substitution in ROS1 leading to a premature stop codon, and is likely a null allele 8. In this study, we generated single-base resolution maps of DNA methylomes of two-week-old seedlings of ros1-4 and ros1-1 mutants. Methylomes of Col-0 and C24 wild types at the same developmental stage were sequenced and served as controls. To identify potential genomic targets of ROS1 and compare the targets in different ecotypes, we identified differentially methylated regions (DMRs) in ros1-4 and ros1-1 mutants relative to their respective wild type plants. ros1-4 has 6902 hypermethylated DMRs (hyper DMRs) with an average length of 495 bp, and 1469 hypomethylated DMRs (hypo DMRs) with an average length of 193 bp, while 5011 hyper DMRs and 332 hypo DMRs were identified in ros1-1. The overwhelmingly higher numbers of hyper DMRs compared to hypo DMRs of ros1 mutants in both Col-0 and C24 are consistent with the ROS1 function in DNA demethylation. In ros1-4, 1887 (27%) hyper DMRs are in genic regions, 2878 (42%) in TE regions, 2010 (29%) in intergenic (IG) regions, and 127 (2%) in the category of others (Fig. 1a). Compared to the composition of randomly selected control regions that are composed of 27% TEs, 54% genes, and 18% IG regions (Fig. 1a), ros1-4 hyper DMRs have a decreased percentage in genes and increased percentages in TEs and IG regions, which is also observed in ros1-1 hyper DMRs (Fig. 1a). This indicates that ROS1 preferentially targets TEs and IG regions. ros1-1 and ros1-4 hyper DMRs are distributed throughout the five chromosomes of Arabidopsis (Supplementary Fig. 1a). For both ros1-1 and ros1-4 hyper DMRs, DNA hypermethylation was detected in all three contexts (CG, CHG and CHH) (Supplementary Fig. 1b). The length distribution of ROS1 targeted TEs is similar to that of all TEs (Supplementary Table 1), suggesting that ROS1 has no preference for short or long TEs. Interestingly, by analyzing the distance between TEs and its nearest genes, we found that ROS1-targeted TEs in both ros1-1 and ros1-4 tend to be located closer to genes relative to TEs that are not targeted by ROS1 (Fig. 1b). In both ros1-1 and ros1-4, DNA methylation is substantially increased around the boarders of TEs targeted by ROS1 (Fig. 1c). As expected, these TEs display decreased methylation in nrpd1 mutants, which are dysfunctional for RdDM due to disruption of NRPD1, the largest subunit of RNA polymerase IV (Fig. 1c). Interestingly, we noticed that, the hypermethylation in ros1 mutants extends from the TE borders to neighboring sequences before tempering off (Fig. 1c). These patterns support our previous hypothesis that ROS1 may counteract RdDM to prevent the spreading of methylation from highly methylated regions, such as TEs, to nearby genes 3. To investigate the influence of different genetic backgrounds on ROS1 targeting, we compared DMRs of ros1-4 and ros1-1, which are mutants in Col-0 and C24 ecotypes respectively. We found that only 27% hyper DMRs in ros1-4 are also hyper DMRs in ros1-1, suggesting that ROS1 targeting is greatly influenced by genetic backgrounds. Interestingly, 35% of TE-type and 28% of intergenic type hyper DMRs are shared between ros1-4 and ros1-1, but only 15% of genic type hyper DMRs are shared between the two mutants. Thus, ROS1 targeting seems relatively conserved in TE regions in Col-0 and C24 ecotypes. Since TEs and genes typically display similar levels of genetic variation, these findings suggest that chromatin features important for active DNA demethylation might be more conserved at TEs than genes between the two ecotypes. Several examples of shared hyper DMRs and non-shared hyper DMRs are displayed in Fig. 1d and Supplementary Fig. 1c. In summary, we identified, characterized, and compared targets of ROS1 in Col-0 and C24 genetic backgrounds. ROS1 targets in both Col-0 and C24 display a preference for TEs and intergenic regions, and the targeted TEs are located near genes. However, the specific genomic regions targeted by ROS1 are mostly distinct in Col-0 and C24 backgrounds. Chromatin features associated with ROS1 targets Histone modifications, such as histone methylation and acetylation, are known to interact with DNA methylation, therefore we determined which histone marks are associated with ROS1 targets. Compared to simulated regions, which are randomly selected genomic regions with the same length distribution as the DMRs, both total TEs and ROS1 hyper DMRs show a slight decrease in the level of H3 (Supplementary Fig. 2a–c), indicating a lower nucleosome density in TEs and ROS1 targets. ROS1 targets are negatively associated with most active histone marks compared to control regions, such as H3K36 di-/tri-methylation (H3K36me2/3), H3K4me2/3 and H3K9 acetylation (H3K9Ac) (Supplementary Fig. 2a–c), which was expected since a large proportion of ROS1 targets are within TEs (Fig. 1a). However, in contrast with most TEs, ROS1 targets are positively associated with the active histone mark H3K18Ac compared to control regions (Fig. 2a–c). Because only 42% of ros1-4 hyper DMRs are within TE regions (Fig. 1a), it is possible that the remaining 58% of ros1 targets that are not within TEs account for the enrichment of H3K18Ac. To investigate this possibility, we compared TE-, intergenic, and genic types of ros1 DMRs with simulated TEs, intergenic regions and genic regions, respectively. Consistently, we found that H3K18Ac is enriched in all types of ros1 DMRs (Supplementary Fig. 2d), suggesting that ros1 targets are indeed generally characterized by enrichment of H3K18Ac. The association with H3K18Ac is fully consistent with our previous finding that IDM1, an H3K18/23 acetyltransferase, is required for the demethylation of a subset of ROS1 targets 14. We identified additional histone marks that distinguish ROS1 target regions. As shown in Fig. 2, TEs in general are negatively associated with H3K27me3, and are positively associated with H3K27me and H3K9me2. In contrast, ros1 DMRs have the opposite features, in that they are associated with enrichment of H3K27me3 and depletion of H3K27me and H3K9me2 (Fig. 2a–c). Similarly, we compared these chromatin features of ros1 targets and corresponding simulation for each type of regions (TE, intergenic and genic regions). All types of ros1 targets are enriched of H3K27me3 compared to their respective simulated regions (Supplementary Fig. 2d). TE-type ros1 targets have decreased H3K27me and H3K9me2 ChIP signals compared to the corresponding simulated regions (Supplementary Fig. 2d). We did not observe decreased H3K27me and H3K9me2 signals for genic and intergenic ros1 targets, since the levels of these histone marks are already very low in simulated genic and intergenic regions (Supplementary Fig. 2d). These results support that ros1 targets are associated with enrichment of H3K27me3 and depletion of H3K27me and H3K9me2. A new class of RdDM targets De novo DNA methylation, especially in the CHH context, is established through the RdDM pathway, which requires DNA dependent RNA polymerase IV for small RNA production 27. In previous studies, RdDM targets have been identified through the identification of hypo DMRs in RdDM mutants compared to wild type plants. In this study, we identified 4580 hypo DMRs and 2348 hyper DMRs in nrpd1 (Pol IV largest subunit) mutant compared to wild type plants; the large number of hypo DMRs in the Pol IV mutant is consistent with the role of Pol IV in DNA methylation. The more than 2000 hyper DMRs in Pol IV may be related to reduced expression level of ROS1 in nrpd1 mutant 23,24 (see below). Homeostasis of DNA methylation is regulated by DNA methylation and active DNA demethylation processes 3,28. As diagrammed in Fig. 3a, regions identified as hypo DMRs (“type I”) in Pol IV mutants must be methylated in the wild type. The presence of methylation in the wild type implies that RdDM dominates over active DNA demethylation at these loci or that active DNA demethylation does not occur at these loci. We refer to the 4580 hypo DMRs in Pol IV mutant as type I RdDM targets. We hypothesized that DNA demethylation may be dominant over RdDM at some genomic loci. These RdDM targets (“type II”) would not be methylated in wild type plants due to the dominance of active DNA demethylation (Fig. 3a). To uncover type II RdDM targets, we introduced the nrpd1 mutation into the ros1-4 mutant, and compared the methylome of the ros1/nrpd1 double mutant with that of the ros1-4 mutant (Fig. 3a,b). The type II regions would be predicted to gain cytosine methylation in ros1-4 mutants due to loss of ROS1 function, however, this gained cytosine methylation would be lost in ros1/nrpd1 mutant due to the dysfunction in RdDM. In total, we identified 6069 hypo DMRs in ros1/nrpd1 compared to ros1. Out of the 6069 hypo DMRs, 3750 display DNA methylation in wild type (mC% >= 2%), and about 60% of these 3750 regions overlap with the 4580 type I RdDM targets. Importantly, and consistent with our hypothesis, there are 2319 hypo DMRs that do not display DNA methylation in the wild type (mC% < 2%). These represent the type II RdDM targets, which have not been identified previously. Similarly, by using published methylome data of ros1-1/nrpd1 in C24 background, we found 4966 hypo DMRs in ros1-1/nrpd1 double mutant compared to ros1-1, and 1656 of them are type II RdDM targets, demonstrating that type II RdDM targets exist in both Col-0 and C24 ecotypes. As shown in Fig. 3b, c and Supplementary Fig. 3a, b, methylation of type I loci is decreased in nrpd1 relative to wild type. In contrast, type II loci do not display a change in methylation level in nrpd1 (Fig. 3b, c and Supplementary Fig. 3c, d). However, introducing the nrpd1 mutation into the ros1-4 mutant revealed the role of RdDM in DNA methylation at the type II loci (Fig. 3a–c and Supplementary Fig. 3c, d). We evaluated 24-nt siRNA enrichment for type II loci and found that 39% of type II loci had 24-nt siRNAs (N>0 in either WT replicates), whereas 61% of type II loci did not display 24-nt siRNA reads (N=0 in both WT replicates). We cannot exclude the possibility that siRNA levels at these loci were too low to be detected by siRNA-seq. Similar with type I targets, the type II targets also have a decreased siRNA level in nrpd1 mutant relative to wild type (Supplementary Fig. 3e). This result further supports that type II loci are targets of RdDM. However, the siRNA level in type II loci is much lower than that in type I loci (Supplementary Fig. 3e), indicating weaker RdDM at type II loci. We examined Pol IV occupancy at type I and type II loci using previously published Pol IV ChIP-seq data29. Pol IV is enriched in type I loci (Fig. 3d). However, we did not observe a significant enrichment of Pol IV in type II loci (Fig. 3d). The low siRNA level and low Pol IV enrichment are consistent with weak RdDM effects at these type II loci. However, type II loci display Pol IV-dependent increased DNA methylation in ros1 mutant (Fig. 3c), suggesting enhanced RdDM at these loci in ros1 mutant. We performed small RNA-Seq in ros1-4 mutant and ros1-4/nrpd1 double mutant plants. We found that type II RdDM targets have a significantly elevated 24-nt siRNA level in ros1 relative to WT plants, and this increase in siRNAs can be suppressed by nrpd1 mutation (Fig. 3e). In contrast, type I RdDM targets do not display increased 24-nt siRNA levels in ros1 mutant compared to WT (Fig. 3f). The results suggest that RdDM becomes stronger at type II loci when ROS1 is removed. DRD1 is a component of the RdDM pathway, and a previous study showed that DRD1-mediated CHH methylation was positively correlated with the histone marks H3K27me3, H3K4me2, H3K4me3, and H3K36me3, and was negatively correlated with H3, H3K9me2 and H3K27me 30. Consistent with this previous study, type I and type II RdDM targets display a slight decrease in H3 enrichment (Supplementary Fig. 4a–c), suggesting a reduced nucleosome density in RdDM target loci than in control regions. Our results show that both type I and type II targets are associated with a depletion of euchromatic histone marks, including H3K4me2/3, H3K36me2/3 and H3K9meAc (Supplementary Fig. 4a–c). Type I and type II targets also display distinct chromatin features as shown in Fig. 4a–c. Chromatin features of type II RdDM targets are similar to ROS1 targets, including enrichment of H3K18Ac and H3K27me3 (Fig. 2 and 4), as expected. In contrast, type I targets display decreased H3K18Ac and slightly decreased H3K27me3 (Fig. 4). These distinct chromatin features are supported by examination of type I and type II targets for different categories of regions (TE, intergenic and genic regions) (Supplementary Fig. 4d). Consistently, we found enrichment of H3K18Ac and H3K27me3 in all categories of type II targets compared to the corresponding categories of type I targets. Also unlike type I targets, type II targets are depleted of H3K9me2 and H3K27me. The depletion of H3K9me2 and H3K27me was found only in TE and intergenic regions of type II targets compared to type I targets, but not in genic regions of type II targets (Supplementary Fig. 4d). In summary, type I RdDM targets show DNA methylation in the wild type, and they may or may not be regulated by ROS1. In contrast, the newly discovered type II RdDM targets are all regulated by ROS1 and are essentially depleted of DNA methylation in the wild type due to ROS1 activity. The two types of RdDM targets are also characterized by distinct small RNA profiles and histone modification marks. Relationship between ROS1-mediated DNA demethylation and RdDM pathway ROS1-mediated active DNA demethylation counteracts the RdDM pathway to prevent DNA hypermethylation at some specific loci 8,31,32. However, the genome-wide crosstalk between these two pathways has not been studied. To identify genomic regions targeted by both ROS1 and RdDM, we compared two groups of DMRs: hyper DMRs in ros1 mutant and hypo DMRs in nrpd1 mutants. We found that there are 1136 shared DMRs between ros1 hyper DMRs and nrpd1 hypo DMRs, suggesting that 16.5% (1136/6902) of ROS1 targets are antagonized by RdDM in wild type plants, however, this ratio increased to 60.1 % (4146/6902) by using hypo DMRs identified in ros1/nrpd1 mutant relative to ros1 (Supplementary Fig. 4e). These results suggest that the antagonistic effects between ROS1-mediated active DNA demethylation and RdDM have been underestimated, since type II RdDM targets were previously unappreciated. It has been reported recently that there is a regulatory link between RdDM and ROS1-mediated active DNA demethylation. It was found that ROS1 expression is dramatically reduced in RdDM mutants, including nrpd1, due to the change of DNA methylation at the promoter region of ROS1 gene 23,24. Thus, we speculated that the hyper DMRs in nrpd1 might be caused by reduced ROS1 expression. We found that 1026 of the 2348 hyper DMRs in nrpd1 overlapped with hyper DMRs in ros1, suggesting that nearly half of the hyper methylated loci in nrpd1 might be caused by reduction of ROS1 expression in nrpd1. ROS1 expression is reduced in not only nrpd1, but also other RdDM mutants, such as nrpe123. As shown in Fig. 5a,b and Supplementary Fig. 5a,b, hyper DMRs of different RdDM mutants, including nrpd1 and nrpe1, also have increased DNA methylation in ros1 mutants, suggesting that the decreased ROS1 expression level contributes to the hyper methylation in the two examined RdDM mutants. We then determined whether the methylome of the nrpd1 single mutant, which has a dramatically reduced ROS1 expression level, is similar to the methylome of ros1/nrpd1 double mutant (where ROS1 was knocked out). After comparing ros1/nrpd1 with nrpd1, we identified 3411 hyper DMRs in ros1/nrpd1 relative to nrpd1. Interestingly, only 35% of the 3411 hyper DMRs are included in the 6902 hyper DMRs in ros1-4, suggesting that ROS1 may have new target loci in nrpd1 mutant compared to those in wild type. This finding suggests that the remaining ROS1 expression in nrpd1 mutants still functions at thousands of loci, although other DMLs may also contribute to these hyper DMRs. ROS1 antagonizes RdDM-independent DNA methylation We identified 1026 shared hypermethylated loci in nrpd1 and ros1 (Fig. 6a), which are distributed across five chromosomes (Supplementary Fig. 6a). Although these loci have similar chromatin features to ROS1 targets, such as H3K18Ac, H3K4me2, and H3K4me3, they display slightly increased levels of H3 compared to general ROS1 targets (Supplementary Fig. 6b). At these loci, ROS1 prevents hypermethylation, and the methylation must be independent of RdDM since the methylation can occur in nrpd1 mutants. This indicates that there are RdDM-independent pathways responsible for the methylation and are antagonistic to ROS1 at these loci. Using previously published methylome data (Supplementary Table 2), we examined methylation levels of these 1026 loci in wild type, nrpd1, ros1, drm2, drm1drm2, cmt2, cmt3, cmt2cmt3, drm1/drm2/cmt2 (ddcmt2), drm1/drm2/cmt3 (ddcmt3), drm1/drm2/cmt2/cmt3 (ddcc), and met1 mutants. The nrpd1, ros1, and drm2 mutants display increased mCG, mCHG and mCHH levels at these loci (Fig. 6b–d), suggesting that MET1, DRM1, CMT2 and CMT3 may all contribute to the methylation at these loci. Indeed, we found that the mCG level of these loci is significantly reduced in met1 (Fig. 6b), whereas the mCHG levels are significantly reduced in cmt3 (Fig. 6c). Although CMT2 and CMT3 have been shown to function redundantly in mCHG methylation 33, it seems that the mCHG methylation at these loci mainly depends on CMT3 (Fig. 6c). For mCHH methylation level, there are no significant changes in cmt2 and cmt2cmt3 double mutants. However, mCHH is significantly reduced in ddcmt2 and ddcc mutants but is increased in drm2, drm1drm2 and ddcmt3 (Fig. 6d). This indicates that DRM1 and CMT2 may function redundantly at these regions. At these loci, mCHG and mCHH levels are increased in met1 mutant plants (Fig. 6c, d). This may be caused by the reduction in ROS1 expression in met1 mutants 34, such that the mCHG and mCHH methylation by CMT3, CMT2 and DRM1 could not be removed by ROS1. These results suggested that ROS1 antagonizes CMT3-, CMT2-, DRM1-, and MET1-mediated DNA methylation, which are independent of DRM2, the major DNA methyltransferase in the RdDM pathway (Supplementary Fig. 6c). We examined siRNA levels at these 1026 loci in wild type and nrpd1, and found that 24-nt siRNAs accumulate at the loci in the wild type, but are lost in nrpd1 mutant plants (Supplementary Fig. 6d). Since the siRNAs but not DNA methylation at these loci are dependent on Pol IV, the siRNAs at these loci would not be required for the methylation. This is consistent with a recent study showing that a reduction of siRNA levels in RdDM mutants does not substantially reduce CMT2-dependent CHH methylation 33. In summary, our study revealed that, besides RdDM, ROS1 can antagonize DNA methylation mediated by MET1, DRM1 and CMTs in an siRNA-independent manner. Discussion Among the four proteins in the ROS1/Demeter family in Arabidopsis, ROS1 is the major DNA demethylase in vegetative tissues. In this study, we showed that genome-wide, ROS1 preferentially targets TEs that are close to protein coding genes (Fig. 1b). We also showed that the sequences just outside the boarders of ROS1-targeted TEs have increased DNA methylation in ros1 mutants (Fig. 1c), suggesting that ROS1 prevents the spreading of DNA methylation from highly methylated TEs. Consistently, Yamamuro et al. reported that ROS1 is required for the expression of the EPF2 gene by preventing the spreading of methylation from a TE near the promoter of EPF2 18. In addition, ROS1 family demethylases can positively regulate fungal pathogen responsive genes via demethylating TEs located in or near their promoters 19. Together with these previous studies, our data support that ROS1 is involved in the regulation of gene expression by preventing DNA methylation spreading from nearby TEs. H3K18Ac is an active histone mark correlated with transcriptional activation 35. We found that ROS1 targets are positively associated with H3K18Ac (Fig. 2a–c), supporting our previous work showing that IDM1, an H3K18/23 acetyltransferase, can create a permissive chromatin environment important for ROS1 to access target loci 14. ROS1 targets were also found to be enriched with H3K27me3, but depleted of H3K27me and H3K9me2, in contrast to general TEs (Fig. 2a–c). This is consistent with a previous finding that there was a strong correlation between H3K18Ac and H3K27me3 in Arabidopsis 36, and is also consistent with findings in mammals that DNA demethylation process is coupled with decreased H3K9me2 and increased H3K27me3 37. Consistent with previously observed antagonism between ROS1 and RdDM, ROS1-targeted TEs display decreased DNA methylation in nrpd1 mutants (Fig. 1c). We hypothesized that ROS1-mediated DNA demethylation may be so strong at some loci that methylation does not accumulate in the wild type at these regions. These potential RdDM targets could not be identified by comparing RdDM mutants with wild type plants; however, in this study we discovered over two thousand of these “type II” RdDM targets by comparing ros1 and ros1/nrpd1 mutants. These RdDM targets have eluded previous attempts of RdDM target identification. Our discovery of type II loci suggests that the number of RdDM targets has been greatly underestimated. Overall siRNA enrichment and Pol IV occupancy were lower at type II targets compared to type I targets. More than half of the type II loci do not have any siRNA reads, and we did not observe any significant Pol IV enrichment at type II loci. The type II loci may be better targeted by RdDM in the ros1 mutant background. We observed increased siRNA levels at type II loci in ros1-4 mutant, indicating that RdDM becomes stronger in ros1 mutant plants. It is possible that demethylation or occupancy by ROS1 at these loci in wild type plants limits RdDM accessibility, thus leading to weak RdDM in the wild type, and that this inhibition of RdDM in the wild type is alleviated by ros1 mutation. ROS1 has been shown to have a similar binding affinity to both methylated and non-methylated DNA through a Lysine-rich Domain at the N terminus 38. Thus, it is possible that ROS1 antagonizes RdDM not only by removal of DNA methylation, but also by preventing the access of RdDM machinery to the target loci. In the future, it will be interesting to compare Pol IV occupancy at these loci in wild type and ros1 mutant plants to further investigate this possibility. It is well known that ROS1 expression is dramatically reduced in RdDM mutants 23. Our results suggested that the reduction in ROS1 expression in nrpd1 mutant plants induces DNA hypermethylation at over a thousand genomic regions. The DNA hypermethylation in RdDM mutants must also be caused by some RdDM-independent DNA methylation pathways. Our analysis suggested that four DNA methylases including DRM1, CMT2, CMT3, and MET1 contribute to the hypermethylation in RdDM mutants. This finding implies that ROS1 can also antagonize RdDM-independent DNA methylation. Interestingly, we noticed that the nrpd1 mutant has slightly increased DNA methylation at type II RdDM loci compared to the wild type (Fig. 3c): 198/2319 type II loci overlap with the 1026 hyper DMRs in nrpd1 mutant. Thus, RdDM-independent DNA methylation may compensate to methylate DNA at some type II loci when RdDM is lacking. Our findings suggest that the fine tuning of the plant methylome is complex and involves interactions between DNA methylation mediated by RdDM and RdDM-independent mechanisms, and DNA demethylation mediated by ROS1 family demethylases. Methods Plant materials Mutants including ros1-4, nrpd1-3 (SALK_128428), ros1/nrpd1 double mutant and nrpe1-11 (SALK_029919) are in the Col-0 background. ros1-4 and nrpd1-3 were crossed to generate ros1/nrpd1 double mutant. ros1-1, nrpd1(C24) and ros1-1/nrpd1 (C24) are mutants of C24 ecotype. Seeds were stratified for 2–3 d at 4 °C before being sown on 1/2 MS plates containing 2% (wt/vol) sucrose and 0.7% (wt/vol) agar. All of the plants were grown under long day conditions at 22 °C. Whole genome bisulfite sequencing and analysis DNA was extracted from 1 gram of 14-day-old seedlings using the Plant DNeasy Maxi Kit from Qiagen. And 5 μg of gDNA was used for library construction using Illumina’s standard DNA methylation analyses protocol and the TruSeq DNA sample preparation kit. The samples in Col-0 background were sequenced in the Genomics Core Facilities of the Shanghai Center for Plant Stress Biology, SIBS, CAS (Shanghai, China) with Illumina HiSeq2500. The samples in C24 background were sequenced in the Biosciences Core Laboratory of King Abdullah University of Science & Technology (KAUST) with Illumina HiSeq2000. For Col-0 background data analysis, low quality sequences (q < 20) were trimmed using trim in BRAT-BW 39, and clean reads were mapped to the TAIR10 genome using BRAT-BW and allowing two mismatches. To remove potential PCR duplicates, the remove-dupl command of BRAT-BW was used. DNA hypomethylated regions were identified according to Ausin et al. 40 with minor modification. In brief, only cytosines with 4X coverage in all libraries in the same background were considered. A sliding-window approach with a 200-bp window sliding at 50-bp intervals was used to identify DMRs. Fisher’s exact test was performed for methylated versus unmethylated cytosines for each context, within each window, with FDRs estimated using a Benjamini–Hochberg adjustment of Fisher’s p-values calculated in the R environment. Windows with an FDR ≤ 0.05 were considered for further analysis, and windows within 100 bp of each other were merged to larger regions. Regions were then adjusted to shrink to the first and last differentially methylated cytosines (DMCs). A cytosine was considered DMC if it showed at least a two-fold change in methylation percentage in the mutant. The regions were then filtered to include only those with at least 10 DMCs and with at least a twofold change in arithmetic mean of methylation percentage of all cytosine. For C24 data, clean reads were mapped to a pseudo-C24 genome using BRAT-BW allowing two mismatches. We used public data set of ros1-1/nrpd1(C24) double mutant in C24 background 41 to analyze type II RdDM targets. The pseudo-C24 genome was generated through the replacement of SNPs in the Col-0 genome with C24 variants (http://1001genomes.org/data/MPI/MPISchneeberger2011/releases/current//C24/Marker/C24.SNPs.TAIR9.txt). TE border analysis The analysis was according to previously described method 42: ros1-4 hyper DMR associated TEs were aligned at the 5′ end or the 3′ end. We discarded from the analysis 250 bp from the end opposite to the one used for alignment to avoid averaging the edges of shorter TEs with the middles of longer sequences. Histone feature analysis Histone features were analyzed according to a previously described method 4 with a minor modification: Briefly, the public data used for the analysis were downloaded from GEO (Accession No: GSE28398) 36. The reads were aligned to TAIR10 using Bowtie 43 allowing 3 mismatches. Only reads that were uniquely mapped to the genome were retained for the downstream analysis. To generate the relative histone signal distribution in the flanking 5-kb region of the mid-point of DMRs, the whole region (10050 bp long) was divided into 201 bins with a size of 50 bp and the 101th bin aligning at the middle point of each DMR. The number of depth in each of the 201 bins was summed. The relative histone modification signal (y axis) in each of the 201 bins was defined as: n( Histone_modification ) × N( Input )/[N( Histone_modification ) × n( Input )] where n is the sum of depth of the corresponding library in each bin and N is the number of mapped reads of the corresponding library. For box plots, DMRs were considered as the 1050 bp region from the DMR mid-point (+/−10 bins plus the mid-bin). In each region, the relative histone modification signal was calculated as above. The box plots were generated in R using function “boxplot” with parameter “range = 1.5, outline = F, notch = T”. The p-values were calculated in R using function “wilcox.test”. Small RNA analysis Small RNA samples were prepared from 14-day-old seedlings. The analysis pipeline was according to Zhang et al. 44. Pol IV ChIP-seq analysis The data sets we used are from published paper 29. According to this paper, WT is pure wild-type plants without any transgenes. nrpd1/NRPD1-3xFLAG is nrpd1 mutant with NRPD1-3xFLAG transgene. Data access The data discussed in this publication have been deposited in NCBI’s Gene Expression Omnibus and are accessible through GEO Series accession numbers GSE83802. Previously published data, including whole genome sequencing data and ChIP-seq data, used in this study were listed in Supplementary Table 2. Supplementary Material 1 Supplementary Figure 1. Characterization of ros1 hyper DMRs a. Chromosomal distribution of hyper DMRs in ros1-4 and ros1-1 mutants. b. Methylation changes in ros1 mutants relative to wild type at ros1 hyper DMRs are shown in the kernel density plots. Changes in mCG, mCHG and mCHH are shown separately. c. Methylation levels of ros1 hyper DMRs in ros1 mutants and their corresponding wild types. Left two panels show hyper DMRs in both ros1-4 and ros1-1. Middle two panels show hyper DMRs in only ros1-4 but not ros1-1. Right two panels show hyper DMRs in only ros1-1 but not ros1-4. 2 Supplementary Figure 2. Association of histone modifications with ROS1 targets a. Association of different histone modifications surrounding the mid-point of ros1-4 hyper DMRs, all TEs and simulated regions. Similar with general TEs, ros1-4 hyper DMRs are negatively associated with H3K9Ac, H3K4me2, H3K36me2, H3K36me3, and H3K4me3. b. Box plots display of the results in Supplementary Fig. 2a (p-value < 1e-15, one-tailed Wilcoxon rank sum test). c. Percentile plots of the same data as in Supplementary Fig. 2b. For each histone mark, simulated regions, TEs and ros1-4 DMRs were ranked based on their histone ChIP signals from low (left) to high (right) along X-axis. X-axis is ranking percentile, and Y-axis is ChIP signal. d. Box plot display of histone modifications of different types of ros1-4 hyper DMRs and the corresponding simulation regions (p-value < 1e-6, one-tailed Wilcoxon rank sum test). 3 Supplementary Figure 3. Methylation and siRNA levels of type I and type II RdDM targets a. Methylation levels of two type I RdDM targets in Col-0, nrpd1, ros1-4, and ros1/nrpd1. Whole genome bisulfite sequencing data are shown for two type I RdDM targets. The region on the left is the type of RdDM target shown in the upper panel of Fig. 3a, while the region on the right is the type of RdDM targets shown in the middle panel of Fig. 3b. b. Heat maps showing CG, CHG and CHH methylation levels of all type I RdDM targets in Col-0, nrpd1, ros1-4, and ros1/nrpd1. c. Methylation levels of type II RdDM targets. Methylation levels of two type II RdDM targets in Col-0, nrpd1, ros1-4, and ros1/nrpd1 were shown in left panels. Methylation levels of two type II RdDM targets in C24, nrpd1 (C24), ros1-1, and ros1-1/nrpd1 (C24 background) were shown in right panels. d. Heat maps showing CG, CHG and CHH methylation levels of all type II RdDM target in Col-0, nrpd1, ros1-4, ros1/nrpd1. e. Heat maps and box plots showing 24-nt siRNA abundance of type I (left panel) and type II (right panel) RdDM targets in Col-0 and nrpd1 (p-value < 1e-7, paired two-sample t-test). 4 Supplementary Figure 4. Association of different histone modifications with type I and type II RdDM targets a. Association of different histone modifications at regions surrounding the mid-points of type I and type II targets and simulation regions. b. Box plots showing the same results as in Supplementary Fig. 4a. (p-value < 0.005, significantly lower than that of simulation, one-tailed Wilcoxon rank sum test). c. Percentile plots of the same data as in Supplementary Fig. 4b. For each histone mark, simulated regions, type I and type II RdDM targets were ranked based on their histone ChIP signals from low (left) to high (right) along X-axis. X-axis is ranking percentile, and Y-axis is ChIP signal. d. Box plots showing histone features of different categories of type I and type II RdDM targets (p-value < 1e-6, one-tailed Wilcoxon rank sum test; NS, not significant). e. Venn diagrams showing relationships among different groups of hyper and hypo DMRs. 5 Supplementary Figure 5. Methylation levels of hyper DMRs of RdDM mutants a. Heat maps showing CG, CHG and CHH methylation levels of nrpd1 hyper DMRs in Col-0, nrpd1 and ros1-4. b. Heat maps showing CG, CHG and CHH methylation levels of nrpe1 hyper DMRs in Col-0, nrpe1 and ros1-4. 6 Supplementary Figure 6. Features of shared hyper DMRs between nrpd1 and ros1-4 a. Chromosomal distribution of shared hyper DMRs between nrpd1 and ros1-4. b. Box plots display of different histone modifications surrounding 1026 shared hyper DMRs, all ros1-4 DMRs and simulated regions (p-value < 0.0005, one-tailed Wilcoxon rank sum test). c. Diagram showing that the 1026 shared hyper DMRs are regulated by ROS1, RdDM, and RdDM-independent DNA methylation pathways in wild type, and might be also regulated by DMLs-mediated demethylation pathways. d. Heat map showing 24-nt siRNA abundance of the 1026 shared hyper DMRs in Col-0 and nrpd1. 7 Supplementary Table 1 Table of the percentiles of the length of ROS1-targeted TEs and all TEs. 8 Supplementary Table 2 Table of previously published data used in this study. This work was supported by National Institutes of Health Grant R01GM070795 and by the Chinese Academy of Sciences (to J.-K. Z.). Figure 1 Characterization of the DNA methylomes of ros1 mutants in Col-0 and C24 ecotypes a. Composition of the hyper DMRs in ros1-4, ros1-1 and of the corresponding simulated genomic regions. b. Box plot showing the distances between ROS1-targeted or non-targeted TEs and their nearest protein coding genes (p-value < 2.2e-16, one-tailed Wilcoxon rank sum test). c. DNA methylation levels of ros1 hyper DMR-associated TEs in wild type, ros1 and nrpd1 mutants. TEs were aligned at the 5′ end or the 3′ end, and average methylation for all cytosines within each 50 bp interval was plotted. d. Methylation levels at shared or non-shared hyper DMRs between ros1-1 and ros1-4. Integrated Genome Browser (IGB) display of whole-genome bisulfite sequencing data is shown in the screenshots. DNA methylation levels of cytosines were indicated with the heights of vertical bars on each track. Figure 2 Chromatin features associated with ROS1 targets a. Association of different histone modifications surrounding ros1-4 hyper DMRs. Association of histone modifications at total TEs and simulated regions served as controls. In contrast to total TEs, ros1-4 hyper DMRs are positively associated with H3K18Ac, H3K27me3, and negatively associated with H3K27me and H3K9me2. b. Box plots display of the results in Fig. 2a (p-value < 1e-15, one-tailed Wilcoxon rank sum test). c. Percentile plots of the same data as in Fig. 2b. For each histone mark, simulated regions, TEs and ros1-4 DMRs were ranked based on their histone ChIP signals from low (left) to high (right) along X-axis. X-axis is ranking percentile, and Y-axis is ChIP signal. Figure 3 Identification and characterization of type II RdDM targets a. Schematic hypothesis that different RdDM targets may be regulated differently by ROS1. Some type I RdDM targets are not regulated by ROS1 (Upper panel) whereas other type I RdDM targets are regulated by ROS1, although RdDM is more dominant at these loci (Middle panel). Type II RdDM targets are always regulated by ROS1, and ROS1 is more dominant at these loci (Lower panel). b. Methylation levels of type I and type II RdDM targets in Col-0, nrpd1, ros1-4, and ros1/nrpd1. The three panels are representative regions as diagramed in Fig. 3a respectively. c. Heat maps showing DNA methylation levels of all type I and type II RdDM target loci in different genotypes. d. Box plot showing Pol IV enrichment at type I and type II RdDM targets. Pol IV signal in wild type plants served as control. (p-value<2.2e-16, one-tailed Wilcoxon rank sum test; NS, not significant). e. Box plot and heat map showing 24-nt siRNA abundance of type II RdDM target loci in different genotypes (p-value<2.2e-16, paired two-sample t-test). f. Box plot and heat map showing 24-nt siRNA abundance of type I RdDM target loci in different genotypes. Figure 4 Chromatin features associated with type I and type II RdDM targets a. Association of different histone modifications at regions surrounding the mid-points of type I and type II targets. Simulation regions served as control regions. b. Box plots showing the same results as in Fig. 4a. (p-value<0.005, one-tailed Wilcoxon rank sum test; NS, not significant compared to simulation). c. Percentile plots of the same data as in Fig. 4b. For each histone mark, simulated regions, type I and type II RdDM targets were ranked based on their histone ChIP signals from low (left) to high (right) along X-axis. Figure 5 Reduced ROS1 expression contributes to DNA hypermethylation in RdDM mutants a. Heat map showing total C methylation levels of nrpd1 hyper DMRs in Col-0, nrpd1 and ros1-4. b. Heat map showing total C methylation levels of nrpe1 hyper DMRs in Col-0, nrpe1 and ros1-4. Figure 6 ROS1 antagonizes RdDM-independent DNA methylation a. IGB display of DNA methylation levels at shared hyper DMRs between nrpd1 and ros1-4. DNA methylation levels of cytosines were indicated with the heights of vertical bars on each track. b–d. The 1026 genomic regions with increased DNA methylation in both ros1 and nrpd1 mutants were used. Box plots of CG (b), CHG (c), and CHH (d) methylation level changes (mutant-WT) of these regions were shown in different mutants. ddcmt2 is drm1drm2cmt2 triple mutant, ddcmt3 is drm1drm2cmt3 triple mutant and ddcc is drm1drm2cmt2cmt3 quadruple mutant. (p-value<1e-10, one sample one-tailed Student’s t-test). Author contributions J.-K.Z, Z.L and K. T. designed the study, interpreted the data and wrote the manuscript. K.T. and Z.L. did the bioinformatics analysis. H.Z. performed sequencing experiments. Competing interests The authors declare no competing financial interests. 1 Collier J Epigenetic regulation of the bacterial cell cycle Curr Opin Microbiol 12 722 729 2009 19783470 2 He XJ Chen T Zhu JK Regulation and function of DNA methylation in plants and animals Cell Res 21 442 465 2011 21321601 3 Zhu JK Active DNA demethylation mediated by DNA glycosylases Annu Rev Genet 43 143 166 2009 19659441 4 Yang D-L Dicer-independent RNA-directed DNA methylation in Arabidopsis Cell Res 2015 10.1038/cr.2015.145 5 Zhai J A One Precursor One siRNA Model for Pol IV-Dependent siRNA Biogenesis Cell 163 445 455 2015 26451488 6 Ye R A Dicer-Independent Route for Biogenesis of siRNAs that Direct DNA Methylation in Arabidopsis Mol Cell 61 222 235 2016 26711010 7 Blevins T Identification of Pol IV and RDR2-dependent precursors of 24 nt siRNAs guiding de novo DNA methylation in Arabidopsis Elife 4 e09591 2015 26430765 8 Gong Z ROS1, a repressor of transcriptional gene silencing in Arabidopsis, encodes a DNA glycosylase/lyase Cell 111 803 814 2002 12526807 9 Agius F Kapoor A Zhu JK Role of the Arabidopsis DNA glycosylase/lyase ROS1 in active DNA demethylation Proc Natl Acad Sci USA 103 11796 11801 2006 16864782 10 Morales-Ruiz T DEMETER and REPRESSOR OF SILENCING 1 encode 5-methylcytosine DNA glycosylases Proc Natl Acad Sci USA 103 6853 6858 2006 16624880 11 Li Y An AP endonuclease functions in active DNA demethylation and gene imprinting in Arabidopsis [corrected] PLoS Genet 11 e1004905 2015 25569774 12 Martínez-Macías MI A DNA 3′ phosphatase functions in active DNA demethylation in Arabidopsis Mol Cell 45 357 370 2012 22325353 13 Li Y Duan CG Zhu X Qian W Zhu JK A DNA ligase required for active DNA demethylation and genomic imprinting in Arabidopsis Cell Res 25 757 760 2015 25906993 14 Qian W A histone acetyltransferase regulates active DNA demethylation in Arabidopsis Science 336 1445 1448 2012 22700931 15 lang Z The Methyl-CpG-Binding Protein MBD7 Facilitates Active DNA Demethylation to Limit DNA Hyper-Methylation and Transcriptional Gene Silencing Mol Cell 2015 10.1016/j.molcel.2015.01.009 16 Wang C Methyl-CpG-Binding Domain Protein MBD7 Is Required for Active DNA Demethylation in Arabidopsis Plant Physiol 167 905 914 2015 25593350 17 Zhu J Kapoor A Sridhar VV Agius F Zhu JK The DNA glycosylase/lyase ROS1 functions in pruning DNA methylation patterns in Arabidopsis Curr Biol 17 54 59 2007 17208187 18 Yamamuro C Overproduction of stomatal lineage cells in Arabidopsis mutants defective in active DNA demethylation Nat Commun 5 4062 2014 24898766 19 Le TN DNA demethylases target promoter transposable elements to positively regulate stress responsive genes in Arabidopsis Genome Biol 15 458 2014 25228471 20 Zheng X Zhu J Kapoor A Zhu JK Role of Arabidopsis AGO6 in siRNA accumulation, DNA methylation and transcriptional gene silencing EMBO J 26 1691 1701 2007 17332757 21 Penterman J Uzawa R Fischer RL Genetic interactions between DNA demethylation and methylation in Arabidopsis Plant Physiol 145 1549 1557 2007 17951456 22 Huettel B Endogenous targets of RNA-directed DNA methylation and Pol IV in Arabidopsis EMBO J 25 2828 2836 2006 16724114 23 Lei M Regulatory link between DNA methylation and active demethylation in Arabidopsis Proc Natl Acad Sci USA 112 3553 3557 2015 25733903 24 Williams BP Pignatta D Henikoff S Gehring M Methylation-sensitive expression of a DNA demethylase gene serves as an epigenetic rheostat PLoS Genet 11 e1005142 2015 25826366 25 Lister R Highly integrated single-base resolution maps of the epigenome in Arabidopsis Cell 133 523 536 2008 18423832 26 Penterman J DNA demethylation in the Arabidopsis genome Proc Natl Acad Sci USA 104 6752 6757 2007 17409185 27 Matzke MA Mosher RA RNA-directed DNA methylation: an epigenetic pathway of increasing complexity Nat Rev Genet 15 394 408 2014 24805120 28 Zhang H Zhu JK Active DNA demethylation in plants and animals Cold Spring Harb Symp Quant Biol 77 161 173 2012 23197304 29 Law JA Polymerase IV occupancy at RNA-directed DNA methylation sites requires SHH1 Nature 498 385 389 2013 23636332 30 Zemach A The Arabidopsis nucleosome remodeler DDM1 allows DNA methyltransferases to access H1-containing heterochromatin Cell 153 193 205 2013 23540698 31 He XJ A conserved transcriptional regulator is required for RNA-directed DNA methylation and plant development Genes Dev 23 2717 2722 2009 19903758 32 Gao Z An RNA polymerase II- and AGO4-associated protein acts in RNA-directed DNA methylation Nature 465 106 109 2010 20410883 33 Stroud H Non-CG methylation patterns shape the epigenetic landscape in Arabidopsis Nat Struct Mol Biol 21 64 72 2014 24336224 34 Mathieu O Reinders J Caikovski M Smathajitt C Paszkowski J Transgenerational stability of the Arabidopsis epigenome is coordinated by CG methylation Cell 130 851 862 2007 17803908 35 Eskandarian HA A role for SIRT2-dependent histone H3K18 deacetylation in bacterial infection Science 341 1238858 2013 23908241 36 Luo C Integrative analysis of chromatin states in Arabidopsis identified potential regulatory mechanisms for natural antisense transcript production Plant J 73 77 90 2013 22962860 37 Rose CM van den Driesche S Meehan RR Drake AJ Epigenetic reprogramming: preparing the epigenome for the next generation Biochem Soc Trans 41 809 814 2013 23697942 38 Ponferrada-Marín MI Martínez-Macías MI Morales-Ruiz T Roldán-Arjona T Ariza RR Methylation-independent DNA binding modulates specificity of Repressor of Silencing 1 (ROS1) and facilitates demethylation in long substrates J Biol Chem 285 23032 23039 2010 20489198 39 Harris EY Ponts N Le Roch KG Lonardi S BRAT-BW: efficient and accurate mapping of bisulfite-treated reads Bioinformatics 28 1795 1796 2012 22563065 40 Ausin I INVOLVED IN DE NOVO 2-containing complex involved in RNA-directed DNA methylation in Arabidopsis Proc Natl Acad Sci USA 109 8374 8381 2012 22592791 41 Huang CF A Pre-mRNA-splicing factor is required for RNA-directed DNA methylation in Arabidopsis PLoS Genet 9 e1003779 2013 24068953 42 Ibarra CA Active DNA demethylation in plant companion cells reinforces transposon methylation in gametes Science 337 1360 1364 2012 22984074 43 Langmead B Trapnell C Pop M Salzberg SL Ultrafast and memory-efficient alignment of short DNA sequences to the human genome Genome Biol 10 R25 2009 19261174 44 Zhang H An Rrp6-like protein positively regulates noncoding RNA levels and DNA methylation in Arabidopsis Mol Cell 54 418 430 2014 24726328
PMC005xxxxxx/PMC5123763.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101287858 34238 Curr Protoc Hum Genet Curr Protoc Hum Genet Current protocols in human genetics 1934-8266 1934-8258 27727435 5123763 10.1002/cphg.23 NIHMS799894 Article Pronuclear Injection-based Targeted Transgenesis Schilit Samantha L.P. 167 Ohtsuka Masato 236 Quadros Rolen 4 Gurumurthy Channabasavaiah B. 457 1 Department of Genetics, Harvard Medical School, 77 Avenue Louis Pasteur, Boston, MA, 02115, USA 2 Department of Molecular Life Science, Division of Basic Medical Science and Molecular Medicine, Tokai University School of Medicine, 143 Shimokasuya, Isehara, Kanagawa 259-1193, Japan 3 The Institute of Medical Sciences, Tokai University, 143 Shimokasuya, Isehara, Kanagawa 259-1193, Japan 4 Mouse Genome Engineering Core Facility, University of Nebraska Medical Center, Omaha, NE, 68198, USA 5 Developmental Neuroscience, Munroe Meyer Institute, University of Nebraska Medical Center, Omaha, NE, 68198, USA 7 Corresponding authors 6 Contributed equally Samantha L.P. Schilit: [email protected], tel +1 617-525-4567, fax +1 617-525-4533 Masato Ohtsuka: [email protected], tel +81 463-93-1121, fax +81 463-94-8884 Rolen Quadros: [email protected], tel +1 402-559-8187, fax +1 402-559-7328 Channabasavaiah B Gurumurthy: [email protected], tel +1 402-559-8187, fax +1 402-559-7328 28 7 2016 11 10 2016 11 10 2016 11 10 2017 91 15.10.115.10.28 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Microinjection of DNA expression cassettes into fertilized zygotes has been a standard method for generating transgenic animal models. While efficient, the injected DNA integrates randomly into the genome, leading to potential problems such as disruption of endogenous genes or regulatory elements, variation in copy number, and integration into heterochromatic regions that inhibit transgene expression. A recently developed method addresses such pitfalls of traditional transgenesis by targeting the transgene to predetermined sites in the genome that can safely harbor exogenous DNA. This method, called Pronuclear Injection-based Targeted Transgenesis (PITT), employs an enzymatic transfer of exogenous DNA from a donor vector to a previously created landing pad site in the animal genome. The DNA transfer is achieved through the use of molecular tools such as Cre-LoxP recombinase and PhiC31-attB/P integrase systems. Here, we provide protocols for performing PITT and an overview of the current PITT tools available to the research community. Pronuclear Injection Targeted Transgenesis PITT Cre-LoxP recombination PhiC31-attB/P integration INTRODUCTION Transgenic animals are invaluable for studying gene function and modeling human disease. The canonical method to create transgenic animals involves injecting exogenous DNA of interest (DOI) into fertilized zygotes (also called pronuclei; PN). These injected zygotes are then transferred to recipient females for gestation, leading to the birth of transgenic founder offspring. While efficient, this approach results in random integration of the DOI into the genome. The inability to control copy number and integration site can lead to multiple problems including disruption of endogenous genes, multiple integration sites, and repressed expression of the transgene by epigenetic silencing or proximity to heterochromatic regions and local regulatory elements. To overcome the challenges of random DOI integration, the transgene can be targeted to a specific locus by using gene targeting to embryonic stem (ES) cells. Transgenic ES cells are then injected into blastocysts, which are transferred to surrogate mothers to create chimeric founder animals. However, this method still has pitfalls, as it is time consuming, labor intensive, and more expensive than PN injection-based transgenesis. Using the best features from both of these standard techniques, we have developed Pronuclear Injection-based Targeted Transgenesis (PITT), a method in which the transgene can be inserted at a predetermined locus through PN injection (Ohtsuka et al., 2010). PITT includes two major steps. The first step involves generation of a “seed mouse” strain by inserting heterotypic recombination (LoxP) or integration (attP) sites at specific loci in the genome, which will ultimately serve as landing pads for DOI sequences. These seed mouse strains are currently available from us or RIKEN BRC mouse repository, as described in Table 3. In the second step, a donor DNA cassette that contains compatible LoxP or attB sites is injected into PN that have been isolated from the seed mouse strain. The donor cassette gets inserted at the landing pad sites through Cre-recombination or PhiC31-integration, respectively. An overall schematic of the PITT process is shown in Figure 1. Over the years, our group and other laboratories have improved this method and developed additional tools (Ohtsuka et al., 2012b, 2013, 2015; Tasic et al., 2011). This unit provides a detailed protocol for performing PITT. The PITT method involves four major steps: (1) designing and building of PITT donor DNA constructs (Basic Protocol 1); (2) synthesis and purification of DNA and RNA components for microinjection (Basic Protocol 2); (3) isolation of embryos from seed mice, microinjection of PITT components, and embryo transfer (Basic Protocol 3); and (4) genotyping of offspring to identify transgenic founders (Basic Protocol 4). A general overview of these steps is presented in Figure 2. BASIC PROTOCOL 1 Designing and building of PITT donor DNA constructs This protocol involves two major steps: (a) Designing the PITT donor construct and (b) Building the PITT donor vector. Designing the PITT donor construct The PITT donor DNA construct contains the DOI flanked by elements needed for integration into the endogenous landing pad. Table 1 describes these required sequence elements. There are currently two PITT platforms available, based on the enzyme used for the efficient targeted integration: Cre-PITT and PhiC31-PITT. There is also a third platform that uses Flp-FRT recombination, but due to poor efficiency of this system, it is not used as a major PITT platform. We instead currently use Flp-FRT recombination as an additional tool to remove excess sequences (that come from the vector backbone) after generation of founder transgenic mice. Platform selection depends on the DOI. For example, Cre-PITT would not be suitable if the DOI consists of LoxP sites, because it would result in aberrant recombination in the PITT donor DNA construct. Such a donor construct can be inserted using PhiC31-PITT. Similarly, presence of attB/P sites within the DOI would interfere with PhiC31-PITT, in which case Cre-PITT should be used. A PITT donor vector consists of two major classes of sequence elements: DOI elements and PITT elements. The DOI is the primary transgenic DNA cargo that needs to be inserted into the genome, whereas PITT elements help achieve the targeted insertion of the DOI into the genomic landing pad. The DOI and PITT elements are typically assembled in a standard bacterial plasmid backbone that contains essential plasmid features, such as an origin of replication and an antibiotic selection marker. A few previously developed PITT donor vectors are listed in Table 2. The composition of DOI elements depends on the transgenic project. A simple DOI may have an expression cassette with a promoter driving a cDNA and/or microRNA followed by polyA signal sequence. A more complex DOI may have an inducible reporter cassette followed by an internal ribosome entry site (or a viral 2A peptidase) with another expression cassette encoding a second reporter and a polyA terminator sequence. In a donor vector, PITT elements (such as LoxP or attB sites) flank the DOI. The choice and architecture of PITT elements in a donor vector depends on the chosen recombinase/integrase system and the seed mouse strain. For example, heterotypic LoxP sites or attB sites are included if Cre-PITT or PhiC31-PITT are used, respectively. Our recommendations for compatible combinations of PITT elements needed in a typical PITT donor vector and the corresponding seed mouse landing pads that are available to the scientific community are listed in Tables 2 and 3. Schematics of some donor vectors and landing pads in PITT seed mouse strains are shown in Figures 3 and 4, respectively. Once the DOI and PITT elements have been selected and the theoretical sequences of the donor vector have been designed, the vector can be built by custom synthesis from commercial vendors (such as Bio Basic, Integrated DNA Technologies, GENEWIZ, GeneArt, GenScript, or other companies). Alternatively, DOI elements can be cloned into preexisting plasmid donor vectors. We have made available donor vectors with certain PITT elements, such as recombination/integration sites, as well as some commonly used DOI elements, such as promoters and polyA signal sequences (Table 2). These plasmids contain multiple restriction enzyme sites that enable cloning of the desired cDNA or expression cassette. Building the PITT donor vector This step involves standard molecular biology and recombinant DNA techniques for donor vector cloning. The protocol steps described below provide a choice between two types of cloning methods: a conventional restriction endonuclease (RE)-based method and a more modern technique called Gibson assembly. Materials Reagents and solutions Donor vector backbone (selected from Table 2) Restriction enzymes for cloning and confirming positive clones (refer to Table 2 to identify specific enzymes needed for cloning into the selected donor vector) QIAquick PCR Purification Kit (Qiagen, catalog #28104) Standard gel electrophoresis reagents including low melting agarose (such as SeaPlaque GTG Agarose from Lonza [catalog #50111] or comparable products from other vendors), ethidium bromide (such as from Sigma Aldrich, catalog # E1510), 50× Tris-Acetate-EDTA (such as from Thermo Fisher Scientific, catalog #BP1332-20), DNA gel loading dye, and DNA ladder Modified TE: 10 mM Tris/0.1 mM EDTA (pH 8.0) (such as TE Buffer, 1× Solution pH 8.0, Low EDTA, from Affymetrix, catalog #75793) Alkaline Phosphatase, Calf Intestinal (CIP; NEB catalog #M0290) Phenol, TE-saturated (such as from Sigma-Aldrich [catalog #77607], Nacalai Tesque [catalog #26829-54], or comparable products from other vendors) 3M Sodium acetate (NaOAc) buffer solution (pH 5.2; such as from Sigma Aldrich [catalog #S7899], Nacalai Tesque [catalog #31150-64], or comparable products from other vendors) Ethanol (200 proof ethyl alcohol, such as from Decon Laboratories, catalog #07-678-005) High fidelity DNA polymerase (such as Phusion from NEB [catalog #M0530], KOD-plus from Toyobo [catalog #F0934K] or comparable products from other vendors) Primers with appropriate overhangs to amplify the desired cDNA or expression cassette while adding flanking sequences for cloning Quick Ligation Kit (NEB, catalog #M2200) Gibson Assembly Master Mix (NEB, catalog #E2611) Competent cells (such as 5-alpha Competent E. coli, High Efficiency, from NEB, catalog #C2987) LB Agar containing appropriate antibiotics (e.g. +100 μg/ml ampicillin or +25 μg/ml kanamycin) (such as from Sigma-Aldrich, catalog #L3147) Luria Broth (LB) medium (prepared as directed from MP Biomedicals, catalog #113002022) with appropriate drug selection Plasmid Mini Kit (Qiagen, catalog #12125) Primers for sequencing donor vector Equipment Standard equipment for agarose gel electrophoresis such as microwave for melting agarose, gel electrophoresis units, electrophoresis power supplies and gel imaging system (including LED gel illuminator) Microcentrifuge Vortex mixer Thermocycler (BioRad T100 or equivalent) Autoclave (if you are making your own LB media or plates) Standard equipment for growing and harvesting small volumes of bacteria including culture tubes/flasks and 37°C shaking incubator Decide on the optimal recombinase/integrase platform for your project. There are two major platforms for PITT: Cre and PhiC31. The Cre platform uses two heterotypic loxP sites (e.g. JTZ17 and Lox2272) flanking a DOI. The PhiC31 platform has an attB site on one end of the DOI. If your DOI contains LoxP or attB, do not choose the Cre or PhiC31 platforms, respectively. You may also choose to use both platforms in tandem (Ohtsuka et al., 2015). Select the appropriate plasmid backbone for your project based upon your platform of choice and the seed mouse you plan to use (refer to Table 2). Digest the selected plasmid using the specified restriction enzyme (Table 2) according to the manufacturer’s instructions. Purify the digested plasmid fragments using the QIAquick PCR Purification Kit or gel purification as follows: Electrophorese digested DNA in a 1% low-melting agarose gel. Excise the desired DNA fragment from the gel under LED light (~100 μl). Add 2 volumes (~200 μl) of modified TE to the sample and freeze at −80 °C for more than 20 minutes. Thaw the sample at room temperature, and then spin it down using a microcentrifuge at 4,000–6,000 rpm for 10 seconds. Transfer the supernatant to a new tube. Dephosphorylate the plasmid fragment with CIP at 37°C for 40 min. This step is important to prevent self-ligation of the vector during cloning. Clean the dephosphorylated DNA using the QIAquick PCR Purification Kit or phenol extraction and subsequent ethanol precipitation as follows: Add an equal amount of TE-saturated phenol, vortex, and centrifuge at 12,000 rpm for 5 minutes. Transfer the aqueous phase to a new tube and add 0.1 volumes 3M NaOAc (pH 5.2) and 2.5 volumes ethanol. Precipitate by centrifuging at 12,000 rpm for 10 minutes at room temperature. Remove the supernatant and wash the pellet with 70% ethanol. Centrifuge at 12,000 rpm for one minute at room temperature and remove the supernatant. Let the pellet air dry for 10 minutes. Resuspend the dried pellet in modified TE (e.g. 2 μl). Create your DOI. If you are amplifying cDNA or an expression cassette for your DOI, use a high fidelity polymerase and primers with suitable overhangs. For RE-based cloning, the overhangs should include RE sites that are compatible with the donor plasmid cloning site. For Gibson assembly, 15–80 bp overhangs that overlap the ends of the digested plasmid backbone should be used (refer to The NEBuilder Assembly Tool to design optimal primers (http://nebuilder.neb.com)). If you are custom synthesizing your DOI, include flanking sequences at both ends of your sequence that reflect the same overhangs in step 7a. Make sure to take the intended DOI orientation into account when designing primers. If you amplify your DOI by PCR: Electrophorese PCR products to confirm the correct size amplicon. RE-based cloning method only: Digest ~50–300 ng of the DOI fragment using the specified restriction enzyme(s) according to manufacturer’s instructions. For all DOIs amplified by PCR and/or digested by restriction enzymes: Clean the DNA sample using the QIAquick PCR Purification Kit or by gel purification, phenol extraction, and ethanol precipitation as follows: Electrophorese digested DNA in a 1% low-melting agarose gel. Excise the desired DNA fragment from the gel under LED light (~100 μl). Add 2 volumes (~200 μl) of modified TE to the sample and freeze at −80 °C for at least 20 minutes. Thaw the sample at room temperature, then spin it down using a microcentrifuge at 4,000–6,000 rpm for 10 seconds. Transfer the supernatant to a new tube. Add an equal amount of TE-saturated phenol, vortex, and centrifuge at 12,000 rpm for 5 minutes. Transfer the aqueous phase to a new tube and add 0.1 volumes 3M NaOAc (pH 5.2) and 2.5 volumes ethanol. Precipitate by centrifuging at 12,000 rpm for 10 minutes at room temperature. Remove the supernatant and wash the pellet with 70% ethanol. Centrifuge at 12,000 rpm for one minute at room temperature and remove the supernatant. Let the pellet air dry for 10 minutes. Resuspend the dried pellet in modified TE (e.g. 2 μl). Assemble your digested plasmid backbone (step 6) and DOI insert (steps 7b or 10). For RE-based cloning: Mix the vector and DOI fragments together and ligate using the Quick Ligation Kit (follow manufacturer’s instructions and incubate at room temperature for 5 minutes). For Gibson Assembly: Incubate samples with Gibson Assembly Master Mix according to manufacturer’s instructions. In brief, place mix in a thermocycler at 50°C for 60 minutes. Store samples between 4°C and -20°C until transformation. Transform sample into competent E. coli cells and spread mixture on an LB plate containing the appropriate antibiotic. Incubate at 37°C overnight. For projects where the DOI may render the plasmid unstable, transformation into a different bacterial strain, such as SURE (stop unwanted rearrangement events) E. coli, or growth at 30°C instead of 37°C, may be warranted. Pick colonies, grow in 5 ml LB media with the appropriate antibiotic while shaking at 37°C overnight, and extract plasmids using the Qiagen Plasmid Mini Kit. To avoid another transformation for positive clones, you can save the remaining bacterial cultures at 4°C until confirming transformants with positive clones. Confirm correct clones using restriction digestion analysis and sequencing. BASIC PROTOCOL 2 Synthesis and purification of DNA and RNA components for microinjection Microinjection components for PITT include the donor vector and an mRNA encoding Cre or PhiC31. While PITT donor vectors are unique for supporting a specific transgenic project, there are well-established plasmids available for generating mRNA encoding the enzymes (such as pBBI and pBBK, see Table 2 and Figure 5). You can also directly inject plasmids with DNA encoding the recombinase or integrase enzymes, but we have found empirically that mRNA is ~5-fold more efficient for PITT (Ohtsuka et al., 2013). We have observed that the plasmid is inferior to mRNA because the delay in translation of the recombinase/integrase results in a higher likelihood of mosaicism in founder mice and therefore less efficient germline transmission (Ohtsuka et al., 2013). Furthermore, injection of the Cre and PhiC31-encoding plasmids may even result in their own undesired insertion into the genome. For these reasons, the protocol below describes the mRNA technique. Materials Reagents and solutions Donor vector, transformed into E.coli (from Basic Protocol 1) Vector for mRNA synthesis of recombinase/integrase, transformed into E.coli Cre platform: pBBI, vector for iCre mRNA synthesis (addgene plasmid #65795; https://www.addgene.org/65795/; (Ohtsuka et al., 2013)) PhiC31 platform: pBBK, vector for PhiC31o mRNA synthesis (addgene plasmid #62670; https://www.addgene.org/62670/; (Ohtsuka et al., 2015)) Luria Broth (LB) medium (prepared as directed from MP Biomedicals, catalog #113002022) with appropriate drug selection HiSpeed Plasmid Midi Kit (Qiagen, catalog #12643) Phenol, TE-saturated (such as from Sigma-Aldrich [catalog #77607], Nacalai Tesque [catalog #26829-54], or comparable products from other vendors) 3M Sodium acetate (NaOAc) buffer solution (pH 5.2; such as from Sigma Aldrich [catalog #S7899], Nacalai Tesque [catalog #31150-64], or comparable products from other vendors) Ethanol (200 proof ethyl alcohol, such as from Decon Laboratories, catalog #07-678-005) Modified TE: 10 mM Tris/0.1 mM EDTA (pH 8.0) (such as TE Buffer, 1× Solution pH 8.0, Low EDTA, from Affymetrix, catalog #75793) Ultrafree-MC Filter (HV, 0.45 μM pore size; Millipore, catalog #UFC30HV00) Quant-iT PicoGreen dsDNA Assay Kit (Invitrogen, catalog #P11496) 1.5 ml RNase-free microfuge tubes XbaI to linearize mRNA synthesis vector (NEB, catalog #R0145) Nuclease-free water (not DEPC treated; Ambion, catalog #AM9937) Standard gel electrophoresis reagents including low melting agarose (such as SeaPlaque GTG Agarose from Lonza [catalog #50111] or comparable products from other vendors), ethidium bromide (such as from Sigma Aldrich, catalog # E1510), 50× Tris-Acetate-EDTA (such as from Thermo Fisher Scientific, catalog #BP1332-20), DNA gel loading dye, and DNA ladder Phenol/chloroform/isoamyl alcohol (25:24:1, v/v; pH 7.9, such as from Ambion [catalog #AM9730] or Nacalai Tesque [catalog #25970-14]) Chloroform (such as from Nacalai tesque [catalog #08402-55] or comparable products from other vendors) mMESSAGE mMACHINE T7 Ultra Kit (Ambion, catalog #AM1345) MEGAclear Transcription Clean-up Kit (Ambion, catalog #AM1908) Microinjection buffer (5 mM Tris pH 7.4, 0.1 mM EDTA) Dilute to final concentration from 1M Tris-HCl (Sigma Aldrich, catalog #T2663) and 0.5M EDTA (Sigma Aldrich, catalog #E7889) into sterile-filtered water (Sigma Aldrich, catalog #W1503). Sterilize using a 0.22 μM filter. Equipment Autoclave (if you are making your own LB media or plates) Standard equipment for growing and harvesting large volumes of bacteria including culture tubes/flasks, 37°C shaking incubator, and centrifuge with rotor and bottles for harvesting cells Vortex mixer Microcentrifuge NanoDrop spectrophotometer and/or fluorescence microplate reader Standard equipment for agarose gel electrophoresis such as microwave for melting agarose, gel electrophoresis units, electrophoresis power supplies and gel imaging system Preparation of plasmid DNA 1 Prepare donor and mRNA-encoding vectors using the Qiagen HiSpeed Plasmid Midi Kit according to the manufacturer’s protocol. We recommend growing these transformed bacterial strains in 200 ml LB medium with appropriate drug selection prior to harvesting (to achieve ~50–100 μg yield). 2 After concluding the Qiagen HiSpeed Plasmid Midi Kit protocol with TE elution, precipitate the DNA. Add an equal amount of TE-saturated phenol, vortex, and centrifuge at 12,000 rpm for 5 minutes. Transfer the aqueous phase to a new tube and add 0.1 volumes 3M NaOAc (pH 5.2) and 2.5 volumes ethanol. Precipitate by centrifuging at 12,000 rpm for 10 minutes at room temperature. Remove the supernatant and wash the pellet with 70% ethanol. Centrifuge at 12,000 rpm for one minute at room temperature and remove the supernatant. Let the pellet air dry for 10 minutes. Resuspend the dried pellet in 50–100 μl modified TE. Place the DNA sample at 4°C overnight. 3 Filter the donor vector using a pre-equilibrated Ultrafree-MC filter at 12,000 × g for 2 min at 4 °C. Filtering is critical to avoid clogging the injection needle during microinjection of one-cell stage embryos. 4 Determine plasmid concentration using the Quant-iT PicoGreen dsDNA Assay Kit. The final concentration may range from 200–2000 ng/μl. 5 Store plasmids at -20°C until ready to use. Preparation of mRNA 6 Transfer 30 μg of the mRNA synthesis vector (e.g. 30 µl × 1 µg/μl) to an RNase-free 1.5 ml microfuge tube. Add 120 units XbaI and incubate at 37°C for 3–3.5 hours in a total volume of 120 μl. Check 0.5 μl by agarose gel electrophoresis for complete digestion. 7 Wash and precipitate the digested sample. Add 280 μl nuclease-free water and 40 μl 3M NaOAc (pH 5.2) to the digested sample. Add 450 μl phenol/chloroform/isoamyl alcohol and vortex. Centrifuge at 12,000 rpm for 5 minutes at room temperature. Collect ~440 μl aqueous phase and add 450 μl chloroform. Vortex and centrifuge at 12,000 rpm for 5 minutes at room temperature. Collect ~435 μl aqueous phase and add approximately 2.0 volumes (~870 μl) ethanol. At this step, the DNA is visible in the solution. Centrifuge at 12,000 rpm for 10 minutes at room temperature. Remove the supernatant and wash the pellet with 70% ethanol. Centrifuge at 12,000 rpm for 1 minute at room temperature and remove the supernatant. Let the pellet air dry for 10 minutes. Resuspend the dried pellet in 23 μl nuclease-free water. 8 Use 1 μl DNA (diluted in 9 μl nuclease-free water) to determine the linearized plasmid concentration by NanoDrop or the Quant-iT PicoGreen dsDNA Assay Kit. 9 Dilute the sample to ~0.5 μg/μl with nuclease-free water. 10 Synthesize mRNA transcripts using the mMESSAGE mMACHINE T7 Ultra Kit according to the manufacturer’s protocol with the following adjustments to scale up the reaction for a large amount of stock solution: When combining in vitro transcription reagents included in the kit, make 100 μl total transcription reaction, using 5 μg of the linearized plasmid. Incubate the mixed sample at 37°C for 3 hours, add 5 μl TURBO DNase I, and incubate again for 15 minutes at 37°C. Divide the sample into two 52.5 μl aliquots. We have omitted the polyA tailing procedure following in vitro transcription because our plasmids already include a polyA sequence. 11 Recover mRNA using the MEGAclear Transcription Clean-up Kit according to the manufacturer’s protocol with the following adjustments: Treat each aliquot as its own sample throughout the purification. When eluting RNA, select RNA elution option 2, and perform the second optional elution procedure for a final volume of 100 μl for each eluate. Combine eluates from both aliquots (for a total of 200 μl) prior to doing the optional ethanol precipitation step with ammonium acetate (NH4Ac). We recommend performing a 15-minute centrifugation step at 4°C. Resuspend the pellet in 50–100 μl nuclease-free water (included in the kit). 12 Filter the mRNA using a pre-equilibrated Ultrafree-MC filter at 12,000 × g for 2 min at 4 °C. 13 Quantify mRNA concentration by NanoDrop and confirm good RNA quality by 1% agarose gel electrophoresis. A representative gel image of iCre and PhiC31 mRNA preps is shown in Figure 5. 14 Dilute the purified mRNA to approximately ~400–500 ng/μl and dispense 5 μl aliquots into RNase-free tubes. Store at -80°C until use. Preparation of microinjection solution 15 Thaw frozen stocks of the donor vector (200–2000 ng/μl) and mRNA (400–500 ng/μl) on ice. Gently mix the solutions by tapping the tube and centrifuge at 14,000 rpm for 3 minutes at 4°C. 16 Take 2 μl of donor vector from the top of the stock and dilute it with microinjection buffer to a concentration of about 20 ng/μl (at least twice as concentrated as desired for the final solution concentration of 10 ng/μl). 17 Take 2–4 μl of mRNA from the top of the stock and dilute it with microinjection buffer to a concentration of about 1–2 ng/μl for iCre and/or 15–30 ng/μl for PhiC31 (at least twice as concentrated as desired for the final solution concentration of 0.5–1 ng/μl for iCre and 7.5–15 ng/μl for PhiC31). Optional: At this stage, the mRNA can be filtered again using the Ultrafree-MC filter. 18 Centrifuge the diluted donor vector and mRNA solutions at 14,000 rpm for 3 minutes at 4°C. 19 Carefully transfer the appropriate volume (typically about 5–40 μl) of both the donor vector and mRNA solutions from the top of the tubes to a single RNase-free 1.5 ml tube. Mix gently by pipetting, and adjust final concentration with microinjection buffer if needed to make the final microinjection solution (typically ~100 μl). 20 Check the quality and concentration of nucleic acids using agarose gel electrophoresis on 3 μl of injection solution. 21 Either use the injection solution immediately or store at -80°C until use. BASIC PROTOCOL 3 Isolation of embryos from seed mice, microinjection of PITT components, and embryo transfer Targeted transgenesis follows similar technical steps to traditional transgenesis, but is unique in that it utilizes zygotes isolated from seed mice as opposed to from wild type animals. These seed mouse strains have been previously developed by us (Ohtsuka et al., 2010, 2012a, 2015) and other groups (Tasic et al., 2011), so the generation of PITT seed mice is not described in this unit. If a given project requires development of a new seed mouse, it can be made using standard gene targeting in mouse embryonic stem cells (Behringer et al., 2014; International Society for Transgenic Technologies, 2011) or more rapidly with CRISPR-based approaches (Quadros et al., 2015). The basic steps for this protocol include isolation of one-cell stage embryos from seed mice, microinjection of PITT components into mouse embryos, and transfer of injected embryos into pseudo-pregnant mice. These steps follow standard mouse transgenesis protocols that have been described previously (Pease and Saunders, 2011; Behringer et al., 2014). Notably, they are very similar to those described in a previous unit in Current Protocols in Human Genetics (see Basic Protocol 3 in Harms et al., 2014), and are given below with slight modifications relevant to PITT. We also extend techniques from the previous unit by providing an alternative protocol for producing fertilized zygotes in place of steps 1–11 (see Support Protocol 1). Materials Animals Egg donors (wild type C57BL/6 female mice; procured at 3 weeks old from Charles River Laboratories in Wilmington, MA). Homozygous seed female mice (bred in-house), instead of wild type females, can also be used. However, we have found that hemizygous embryos survive better than homozygous embryos after microinjection. Stud males (homozygous seed mice; 3–6 months old) A list of seed mouse strains can be found in Table 3. Pseudo-pregnant recipients (Crl:CD1(ICR) female mice; purchased at 5–6 weeks of age from Charles River Laboratories, Wilmington, MA; typically used for experiments when they are about 8 to 12 weeks of age) Vasectomized males (CD1 male mice; purchased at 5–6 weeks old from Charles River Laboratories, Wilmington, MA; vasectomized as previously described (Behringer et al., 2014)) Hormones Pregnant Mare’s Serum Gonadotropin (PMSG; National Hormone and Peptide Program, Harbor–UCLA Medical Center, Torrance, CA) Human Chorionic Gonadotropin (hCG; National Hormone and Peptide Program, Harbor–UCLA Medical Center, Torrance, CA) Hormones are supplied as lyophilized vials of 2000 IUs. Upon first use, reconstitute in 2 ml PBS (Millipore, catalog #BSS-1006-B). Aliquot this 20× stock solution (100 IU/100μl) into 100 µl vials and store at -80°C. On the day of the hormone injection, one vial is thawed and diluted to 2 ml with PBS to get a final concentration of 5 IU/0.1 ml. Each animal is administered 0.1 ml of this solution. The leftover solution is discarded. Reagents and solutions 35 × 10 mm Falcon Tissue Culture Dish (Corning, catalog #353001) EmbryoMax M2 Medium (1×), liquid, with phenol red (M2; for embryo handling and microinjection; Millipore, catalog #MR-015-D) EmbryoMax M2 Medium (1×), liquid, with phenol red and hyaluronidase (hyaluronidase; for dissociation of the cumulus oophorus complex; Millipore, catalog #MR-051-F) EmbryoMax KSOM Medium (1×) with ½ amino acids (KSOM; for embryo incubation; Millipore, catalog # MR-106-D) Falcon IVF/Organ Culture Dish (Corning, catalog #353653) Flexipet Pipette (130 μm, for collecting embryos; Cook Medical, catalog #K-FPIP-1130-10BS-5) Glass capillaries (4 in/1 mm) (World Precision Instruments, catalog #TW100F-4) MicroFil 28 gauge/97mm long (World Precision Instruments, catalog #MF28G) Microinjection buffer (5 mM Tris pH 7.4, 0.1 mM EDTA) Dilute to final concentration from 1M Tris-HCl (Sigma Aldrich, catalog #T2663) and 0.5M EDTA (Sigma Aldrich, catalog #E7889) into sterile-filtered water (Sigma Aldrich, catalog #W1503). Filter sterilize using a 0.22 μM filter. 1-cc tuberculin syringe Nunc Lab-Tek Chamber Slide System (Lab-Tek, catalog #177372) EmbryoMax Filtered Light Mineral Oil (Millipore, catalog #ES-005-C) Holding micropipets (Origio, catalog #MPH-SM-20) Equipment Individually Ventilated Cages (IVCs; Allentown, Lab Products, or Tecniplast) Heraeus HERAcell 150i Tri–gas incubator, equipped with Coda Inline filters Standard surgical equipment such as scissors, fine forceps, suturing material, anesthesia chambers, etc. (refer to Behringer et al., 2014) Large slide warmer (Spectrum Scientifics, catalog #3875) Dissecting scope (two examples are included below): Leica MZ 9.5 Condenser lens: PLAN 0.5×, model #10 446 157 Base: Model #10 445 367 Tilt head Heating glass (Live Cell Instrument, catalog #HG-T-Z002) with temperature controller (Live Cell Instrument, CU-301) Nikon SMZ1000 Condenser lens: PLAN APO 1× WD70 Base: C-DSDF, model #1002364 Mid-Piece: C-FMC, model #1009459 LV-TV Camera port Eyepiece: P-BERG w/c-w15/16 eyepiece, model #1007501 Heating glass (Live Cell Instrument, catalog #HG-T-Z002) with temperature controller (Live Cell Instrument, CU-301) Mouth pipetting apparatus (assembled as previously described: Gurumurthy et al., 2016). Glass pipette puller (Sutter Instrument Co. model #P97), outfitted with a 2.5mm × 2.5mm Box filament (catalog #FB255B) Microinjection scope (two examples are included below): Leica DM IRB, equipped with Narishige IM 300 microinjector, and Leica manual manipulators Eyepiece: HC PLAN 10×/22 w/tilt, model #11 507 804 Condenser lens: .30 S70 Objectives: C PLAN 4×/.10, model #11 506 074 N PLAN L20×/0.40 CORR, model #11 506 057 N PLAN L40×/0.55 CORR, model #11 506 059 Heating glass (Live Cell Instrument, catalog #HG-T-Z002) with temperature controller (Live Cell Instrument, CU-301) Nikon Eclipse TE 2000-E with DIC, equipped with Narishige IM 300 microinjector and NT-88-V3 micromanipulators Condenser lens: LWD 0.52 Objectives: PLAN 4×/0.10 WD30 PLAN APO 10×/0.45 WD4.0 PLAN FLUOR ELWD 20×/0.45 DIC L/NI PLAN FLUOR ELWD 40×/0.60 DIC M/NI Heating glass (Live Cell Instrument, catalog #HG-T-Z002) with temperature controller (Live Cell Instrument, CU-301) Isolation of one-cell stage embryos from seed mice 1 House mice in Individually Ventilated Cages (IVCs) on a 14-10 light cycle (on at 06:00, off at 20:00). 2 Intraperitoneally inject 10–20 donor female mice, each with 5.0 IU PMSG (in 0.1 ml volume) around 12:00 on Day 1. 3 On Day 3, approximately 48 hours after the PMSG injection, intraperitoneally inject each female mouse with 5.0 IU hCG (in 0.1 ml volume). Breed with stud males overnight. 4 On the morning of Day 4 (around 8:00), prepare the following dishes: Oviduct collection dish: 35 mm tissue culture dish with 2 ml M2 media (one dish per up to 10 females). Hyaluronidase dish: 35 mm tissue culture dish with 1.5 ml hyaluronidase media (one dish per up to 10 females). Wash dish: 35 mm tissue culture dish with 1.5 ml M2 media (at least two per injection batch). KSOM rinse dish: 35 mm tissue culture dish with 1.5 ml KSOM media (pre-equilibrated). Incubation dish: IVF/organ culture dish with 1 ml KSOM (two per injection batch, both pre-equilibrated). Embryo Transfer dish: 35 mm tissue culture dish with 1.5 ml M2 media. 5 Euthanize plugged donor females by your institutional animal care and use committee’s approved method. Euthanasia is performed approximately 20 hours after hCG injection (about 8:00 on Day 4). 6 Surgically dissect out the oviducts and place in the oviduct collection dish. Maintain tissue samples at 37°C on a heated slide warmer. Make sure all oviduct dissections are completed in less than 10 minutes after euthanasia. 7 Once all oviducts have been collected, clear the working space and begin dissociating the cumulus-oocyte complexes (COC), one at a time. The following steps are performed under a dissecting scope maintained at 37°C. Place oviducts in the hyaluronidase dish. Dissect the COCs by disrupting the ampulla with a pair of fine forceps. Continue to process the remaining oviducts while working efficiently. If all oviducts cannot be processed in under 10 minutes after they are dissected, it is better to plan a dissection of fewer animals in the future. 8 Once the last COC has been expelled from the ampulla, collect individual oocytes from the dish using the mouth pipetting apparatus. The first set of oocytes should begin to dissociate in the time it takes to harvest the rest of the COCs. 9 Using a 130 μm Flexipet pipette, transfer the oocytes to the wash dish (avoiding as much of the hyaluronidase media and as many of the cumulus cells as possible). This will inactivate the residual hyaluronidase. This step may be repeated a second time in order to remove residual cumulus cells and hyaluronidase. 10 Pool all harvested zygotes into a fresh wash dish. Collect the zygotes one by one using a flexipet pipette. Count the number of zygotes and unfertilized oocytes. Record this information to determine fertilization efficiency. Zygotes can be distinguished from unfertilized oocytes by the presence of two pronuclei. 11 Transfer only fertilized eggs (zygotes) to the KSOM rinse dish to wash the residual M2 media from the embryos. Next, transfer them to the incubation dish until needed (typically 30 minutes to 1 hour). Culture the incubation dish at 5% CO2 to maintain a pH range of 7.23–7.42. The time until microinjection should not exceed more than two hours, as zygotes that advance past the one-cell stage are no longer suitable for injection. Microinjection of PITT components into mouse embryos Microinjection needles Injection capillaries are made fresh on the morning of the microinjection using the pipette puller. It is essential to use sterile technique to keep the capillaries nuclease-free. The following program is used: Glass Heat Pull Velocity Pressure Time #TW100F-4 Ramp +5 70 120 200 100 delay Embryo microinjection PITT requires both a cytoplasmic and nuclear injection into embryos, because the mRNA must be translated in the cytoplasm and the donor DNA needs to be delivered to the nucleus for targeted integration. This injection approach is very similar to the protocols followed for mouse genome editing using CRISPR/Cas (Harms et al., 2014). 12 Prewash a 28 gauge MicroFil three times with sterile microinjection buffer. 13 Backfill 5–6 injection needles with 1–2 μl of microinjection solution each (from Basic Protocol 2, Preparation of Microinjection Solution) using the prewashed MicroFil connected to a 1-cc tuberculin syringe. 14 Affix the injection needle to the microinjector. The remaining prefilled needles should be stored on ice in a needle holder as an additional precaution to prevent RNA degradation during this step. The needle holder is made from a 150 mm tissue culture dish (Falcon, catalog #351058), outfitted with a 0.25 cm-diameter rod-shaped model of clay. The injection needles filled with the solution are pushed into the clay and the entire storage unit is placed directly in contact with the ice bath. 15 The following parameters are programmed into the Narishige IM 300 microinjector: Injection pressure Balance Hold Clear Clear Hold Inj. time 20 psi 2.2 psi 14 psi 0.20 sec 0.30 sec 0.08 sec 16 Prepare an injection slide using a Lab-Tek chamber and by making two side-by-side 150 μl drops of M2 media. Flatten these drops into discs with a pipette tip to minimize their height. Overlay the flattened drops with ~1 ml of mineral oil. Maintain the temperature at 37°C with the heating glass. 17 We usually inject about 25–30 zygotes per batch. Transfer zygotes to the injection slide. All zygotes must be injected within 10 minutes so the number of zygotes taken per batch depends on efficiency of the injector. A beginner may start with as few as 4 to 6 per batch and the most experienced technician can inject as many as 50 zygotes in 10 minutes. 18 Check the general morphology of the zygotes under the microscope for presence of zona pellucida, pronuclei, and both polar bodies. Discard embryos that contain more than two pronuclei. 19 Prior to injection, make sure that the needle is open by placing the injection needle next to an embryo. Press the “clear” button on the injector. If the embryo rotates freely, the needle is free of any obstruction. If the embryo does not move, gently remove the tip of the injection needle using a scraping motion against the holding pipette. Check the needle again for flow rate. The needle should be discarded if the embryo moves too much, but can be used if the embryo rotates freely. If the needle is still obstructed, try breaking off more of the tip. 20 Using the holding micropipet, place the first zygote in position and fix by applying negative pressure. 21 Align the embryo and the microinjection needle so that both the opening of the needle and the pronucleus of the embryo are in focus. 22 Perform the microinjection. Maintain positive pressure on the injection needle at all times. Penetrate the zona pellucida and oolemma with the injection needle. Move forward into the closest pronucleus. A slight swelling of the pronucleus may be seen once the plasma membrane is penetrated. Otherwise, press the injection foot pedal to observe a slight swelling of the pronucleus. Retract the tip of the needle to the cytoplasm and inject another volume of microinjection solution. Carefully withdraw the capillary from the zygote if it hasn’t already been removed from the force of the cytoplasm injection. 23 Proceed with all remaining zygotes. 24 After all zygotes have been injected, use the mouth pipetting apparatus to collect and transfer them to the Embryo Transfer dish. 25 Remove lysed zygotes. 26 Incubate surviving zygotes at 37°C in KSOM until embryo transfer (usually within the next 1–2 hours). Every new batch of microinjection solution should be assessed for toxicity by culturing about 30 injected zygotes overnight. In a successful injection session, 90–95% of zygotes should progress to the two-cell stage. If the solution batch is toxic, extensive lysis may be visible within an hour post injection. Transfer of injected embryos into pseudo-pregnant mice 27 Transfer injected embryos into pseudo-pregnant female mice. Obtain pseudo-pregnant mice by mating 8–12 week old CD1 females to vasectomized CD1 males on the day before microinjection between 12:00–16:00. On the morning of the injection day, use plug-positive females for oviduct transfers. Typically, 10–20 CD1 females are bred in each session to obtain an average of 4–8 plugged females. Transfer viable manipulated embryos into the oviducts of pseudo-pregnant foster mothers following established surgical procedures as previously described (Behringer et al., 2014). About 15–25 injected embryos are transferred per female. The optimal number of embryos transferred is 18 total per female (9 per side). SUPPORT PROTOCOL 1 Production of fertilized eggs through in vitro fertilization (IVF) As an alternative to steps 1–11 in Basic Protocol 3, zygotes may be produced by in vitro fertilization using sperm from homozygous PITT seed mice as described here. IVF is advantageous because it reduces the number of stud mice needed and provides more scheduling flexibility for the researcher. We have also found that IVF produces a large number of synchronized and high quality embryos for microinjection. Please note that the hormone treatment in this IVF protocol is slightly different from the natural mating protocol provided above (e.g., hormone concentrations, doses and timings of hormone administration). This is because these techniques are performed in different laboratories (natural mating: from CB Gurumurthy’s laboratory in the United States; IVF: from Masato Ohtsuka’s laboratory in Japan). Although slightly different, these protocols seem to work most optimally based on the hormones, animal sources, and housing conditions in these labs. Materials Animals Stud males (homozygous seed mice; 3–6 months old) A list of seed mouse strains can be found in Table 3. Egg donors (wild type C57BL/6 female mice; procured at 7 weeks old from CLEA Japan, Inc., Tokyo, Japan). Homozygous seed female mice (bred in-house), instead of wild type females, can also be used. However, we have found that hemizygous embryos survive better than homozygous embryos after microinjection. Hormones Pregnant Mare’s Serum Gonadotropin (PMSG; ASKA Animal Health Co., Ltd, Tokyo, Japan) Human Chorionic Gonadotropin (hCG; ASKA Animal Health Co., Ltd, Tokyo, Japan) Hormones are supplied as lyophilized vials. These are reconstituted in saline (Otsuka Normal Saline, Otsuka Pharmaceutical Factory, Inc.) to a final concentration of 7.5 IU/0.2 ml and stored at −30°C until use. Each animal is administered 0.2 ml of this solution. The leftover solution is discarded. Reagents and Solutions 35 × 10 mm Falcon Tissue Culture Dish (Corning, catalog #353001) HTF (Human Tubal Fluid) (ARK Resource, Kumamoto, Japan) EmbryoMax M2 Medium (1×), liquid, with phenol red (M2; for embryo handling and microinjection; Millipore, catalog #MR-015-D) 1 ml Tuberculin Syringe with needle (26G × ½″; TERUMO, catalog #SS-01T2613S) Liquid paraffin (Nacalai Tesque, catalog #26137-85) Equipment Standard surgical equipment such as scissors, fine forceps, suturing material, anesthesia chambers, etc. (refer to Behringer et al., 2014). Dissecting microscope (e.g., Olympus SZ11) with a hot plate (e.g., KM-1, Kitazato) Heraeus HERAcell 150i Tri–gas incubator, equipped with Coda Inline filters Intraperitoneally inject 7.5 IU of PMSG (in a 0.2 ml volume) into about 15 female mice at 18:00 on Day 1. The later injection time for IVF (compared to the earlier hormone treatment for natural mating) allows for the microinjection and surgery procedures to fall at a reasonable time (16:30 – 19:30) on the day of IVF. After 48 hours, on Day 3 at 18:00 pm, intraperitoneally inject 7.5 IU of hCG (in a 0.2 ml volume) into female mice. On the morning of Day 4, (8:00–8:30), about 30 min before egg collection, dissect the cauda epididymides from a stud homozygous male seed mouse as previously described (Takahashi and Liu, 2010) Instead of using the freshly-isolated epididymides as a source of sperm, cryopreserved sperm samples may also be used. However, we have observed that using cryopreserved sperm leads to poor fertilization rates and poor quality embryos. Cut the epididymides and use forceps to transfer the sperm into a 35 mm dish containing 37°C HTF medium (300 μl). Incubate the sperm in a 5% CO2 incubator for 1–1.5 hours to allow for capacitation. During sperm incubation, sacrifice the super-ovulated females following your institutional animal care and use committee’s approved method of euthanasia. Surgically remove the oviducts, and place them into 35 mm dishes containing M2 medium. Introduce the egg–cumulus cell complex into the medium by teasing the ampulla of the oviduct with a 26-gauge needle. Transfer the egg–cumulus cell complex to HTF drops (250 μl) and cover with liquid paraffin in a 35 mm dish. At 9:30, add 10 μl cultured sperm to the HTF drop containing oocytes. Using a dissecting microscope, confirm sperm motility. Incubate in a 5% CO2 incubator to allow IVF to occur. Determine success of IVF 5–6 hours later by observing fertilized pronuclei. With freshly isolated sperm, the fertilization rate should be greater than 80%. Proceed to microinjection starting at 16:30 (Basic Protocol 3, starting at step 12). BASIC PROTOCOL 4 Genotyping of offspring to identify transgenic founders Although performing a Southern blot is traditionally considered the gold standard for confirming successful targeted cassette insertions, PCR is reliable for identifying PITT founder mice. While the exact genotyping strategy depends on the PITT project, primer sets must be designed to amplify both landing pad-DOI junctions in order to identify the precise insertion site. Additional primer sets should be used to probe the DOI. A schematic of this standard PCR-based strategy is shown in Figure 6. With the exception of the chosen primer set, all genotyping PCRs follow the same standard protocol steps. There are two major steps in this protocol: (a) mouse tail DNA extraction and (b) PCR amplification followed by agarose gel electrophoresis. Materials Reagents and solutions Cell Lysis Soution (Qiagen, catalog #158908) Proteinase K (20 mg/ml, such as from 5 PRIME, catalog #2900150) Protein Precipitate solution (Qiagen, catalog #158912) Ethanol (200 proof ethyl alcohol, such as from Decon Laboratories, catalog #07-678-005) DNA Hydration Solution (Qiagen, catalog #158914) PCR 2× master mix (such as GoTaq Hot Start Green Master Mix from Promega [catalog #M5122] or other comparable vendors) Nuclease-free water (such as from Thermo Fisher Scientific, catalog #BP561-1) Primer mix Mix equal volumes of forward and reverse primers from 100 pmol/μl stocks, resulting in a final mix that is 50 pmol/μl with respect to each primer. (To be used at a ratio of 1 μl/100 μl of PCR master mix) Standard gel electrophoresis reagents including 1% TAE agarose gel with ethidium bromide, 50× Tris-Acetate-EDTA (such as from Thermo Fisher Scientific, catalog #BP1332-20), DNA gel loading dye, and DNA ladder. Equipment Heat block Vortex mixer Microcentrifuge Thermocycler (BioRad T100 or equivalent) Standard equipment for agarose gel electrophoresis such as microwave for melting agarose, gel electrophoresis units, electrophoresis power supplies and gel imaging systems Mouse tail DNA extraction 1 Collect ~2–3 mm tail pieces from potential founder mice and wild type controls in 1.5 ml microcentrifuge tubes. Add 300 μl Cell Lysis Solution containing 3 μl Proteinase K and incubate at 65°C overnight. To save time and reagent loss, make a master mix for the lysis solution and Proteinase K. This can then be distributed to 300 μl aliquots in separate microcentrifuge tubes. 2 Cool to room temperature. Add 100 μl of the Protein Precipitation Solution and vortex thoroughly for ~20 seconds. 3 Place the tubes on ice for 2–3 minutes, and then centrifuge at 13,000 rpm for 2–4 minutes. 4 Transfer supernatants to new tubes containing 800 μl of ethanol and mix by inverting the tubes 8–10 times. 5 Centrifuge at 13,000 rpm for 2 to 4 minutes. 6 Discard the supernatant, add 800 μl of 70 % ethanol, invert 8–10 times. 7 Centrifuge at 13,000 rpm for 2–4 minutes, and then discard supernatant. 8 Centrifuge at 13,000 rpm for 1 minute. 9 Manually aspirate the remaining 70% ethanol using a 200 μl pipette tip and air dry the DNA pellet for ~5 minutes (do not exceed 8 minutes). It is important to change tips between samples to avoid cross contamination. 10 Add 50 μl DNA Hydration Solution to the pellet and mix by flicking the side of the tube. Incubate the tubes at 65°C for 15–30 minutes to solubilize the DNA. PCR amplification and agarose gel electrophoresis 11 Perform PCR for all primer sets on potential transgenic founder mice and wild type control DNA samples. 12 Use the PCR 2× master mix manufacturer’s protocol to determine the exact parameters for your PCRs. The basic steps for GoTaq Hot Start Green Master Mix are outlined below: a Combine PCR master mix, nuclease-free water, mouse tail DNA (~1 μl, from step 10), and primer mix (both forward and reverse) for a total volume of 15 μl/sample. b Run the PCR reactions in a thermocycler. While the annealing temperature may slightly vary depending on the primer set, the following are standard PCR conditions: 95°C for 2 minutes 35 cycles: 95°C for 30 seconds, 55°C for 30 seconds, and 72°C for 1 minute Hold at 4°C. 13 Run all PCR products on a 1% TAE agarose gel. 14 Analyze gels to identify mice genotypes. c Ensure that no amplicons are visible for any reactions using wild type control DNA samples. d Identify transgenic founders as mice that demonstrate the correct size amplicons for all assayed primer sets. Example genotyping results are provided in Figure 6. COMMENTARY Background Information Historical perspectives on random and targeted transgenic technologies Transgenesis has revolutionized many fields of biology including genetics, gene regulation, medicine, and bioengineering since its inception in the early 1980s (Jones, 2011). The initial successes in developing this technology were reported by Drs. Jon Gordon and Frank Ruddle, who performed the first genetic transformation by injecting DNA into mouse pronuclei, resulting in the integration and stable germ line transmission of the exogenous DNA (Gordon et al., 1980; Gordon and Ruddle, 1981). However, this technology suffered from the random nature of exogenous DNA integration, including variation in copy number and chromosome position effects at the site of integration, which led to inconsistent gene expression (Hogan, 1983). To overcome this challenge, the next generation of transgenesis techniques involved insertion of transgenes into embryonic stem cells at a specific site in the genome and subsequent injection of targeted ES cells into blastocyst-stage embryos (Gossler et al., 1986). However, this technique required generation of transgenic embryonic stem cells including electroporation/transfection, selection, and screening, in addition to an additional generation of breeding to ensure germline transmission in the chimeric founders. These additional steps made the ES cell-based approach laborious and expensive. Taking the best features from these techniques, we decided to inject transgenic DNA into pronuclei, like traditional transgenesis, but direct integration to a specific site, like ES cell-based approaches. To facilitate targeted integration, our laboratory first developed a seed mouse containing a special landing pad in its genome that would enable efficient and targeted insertion of an injected DOI sequence in the presence of a recombination enzyme (Ohtsuka et al., 2010). We systematically tested and characterized a set of heterotypic LoxP sites to direct a DOI to a specific genomic locus using a non-reversible Cre recombination-mediated cassette exchange mechanism, and first described this technique, including the ideal landing pad and compatible DNA elements that would be suitable for such targeted transgenesis, in 2010 (Ohtsuka et al., 2010). Soon after, another group adapted pronuclear injection-based targeted transgenesis to be used with attB/P sites and PhiC31 integrase (Tasic et al., 2011). Our lab subsequently improved the PITT tools including developing a multiplexed PITT system that constitutes Cre-LoxP, PhiC31-attB/P, and Flp-FRT systems (Ohtsuka et al., 2015). Critical Parameters Here we address critical parameters of the microinjection solution and the seed mouse to maximize PITT efficiency. Microinjection solution PITT insertion efficiency and the viability of injected embryos are highly sensitive to the quality and concentration of nucleic acids in the microinjection solution. Nucleic acid quality We recommend filtering the nucleic acid solution with the Ultrafree-MC filter prior to preparing the injection mixture to prevent potential clogging of the injection needle. The concentration should be quantified after this step to account for decreased yield post filtration. It is also important to ensure high integrity of the synthesized mRNA. We recommend evaluating mRNA quality by gel electrophoresis both after mRNA synthesis and after injection mixture preparation. It is important to check mRNA quality after mRNA synthesis, because the synthesized mRNA solution contains variable amounts of incompletely synthesized RNAs. As a result, the concentration of the mRNA solution does not fully reflect the number of intact mRNA molecules within a given sample. Incomplete mRNA synthesis can be detected as a smear in the gel. Given that there may be trace RNase contamination in the donor vector solution, the microinjection solution should also be evaluated for mRNA integrity after it has been stored at room temperature or 4°C for 24 hours. Nucleic acid concentration Success of PITT is highly dependent on the concentration of nucleic acids. Although we have described the optimal concentrations of mRNA and donor vector for our own system, the ideal concentration should be optimized for each unique system. This is especially important because there is only a narrow range of recombinase/integrase concentrations that will maximize insertion efficiency while preserving embryo survival. Concentrations may vary because the methods of calculating nucleic acid concentration differ depending on the lab preference (NanoDrop, PicoGreen, UV absorbance, etc). In addition, regardless of the concentration of mRNA, the amount of recombinase/integrase translated in injected zygotes depends on the vector used for mRNA synthesis (e.g., Cre vs. iCre, PhiC31 vs. PhiC31o, or vector containing a polyA sequence vs. no polyA sequence and subsequent polyadenylation by the mMESSAGE mMACHINE T7 Ultra Kit). To determine the optimal concentration of mRNA in your system, inject pronuclei with microinjection solutions that test a range of mRNA concentrations. Develop these pronuclei in vitro and check the survival rate at the blastocyst stage (see Supplementary Tables S3 and S4 in Ohtsuka et al., 2010, Table S1 in Ohtsuka et al., 2015). Select the maximum concentration before blastocyst viability is compromised by more than 40–60%. This should be performed for each new batch of mRNA as well as for each unique microinjectionist to account for person-to-person technical variability. Given the investment in time and reagents to perform this optimization, we recommend preparing large batches of mRNA so one optimization can be performed for many experiments. While we have observed that increasing sizes of the DOI decreases insertion efficiency, we have not empirically defined the acceptable range for donor vector size. To date, we have performed PITT with donor vectors up to 14.4 kb in size (Ohtsuka et al., 2013). Seed mouse While we have generated many seed mice that have been used successfully for PITT, we have observed that some of these strains are less efficient at integrating exogenous DOI than others. Seed mouse parameters that may influence efficiency include 1) landing pad locus, 2) recombinase/integrase system, and 3) genetic background. Landing pad locus Rosa26 is the frequently used genomic locus for targeted donor vector insertion because it has been well established as an open chromatin region that is conducive to both efficient DOI integration and expression of inserted genes. However, several other loci can serve as landing pads including H11, H2-Tw3, Hprt, TIGRE, Actb and AAVS1. Recombinase/integrase system Although there are three available PITT platforms (Cre, PhiC31, and Flp), only Cre and PhiC31 systems have been proven to work in vivo. The Flp recombination system has been successfully used in vitro, but not yet in mouse embryos. In our experience, Cre and PhiC31 are equally effective at mediating targeted transgenesis. These two systems demonstrate comparable insertion efficiencies from embryonic stem cell studies (Ohtsuka et al., 2015). Furthermore, combining the Cre and PhiC31 platforms together, as has been done in our TOKMO-3 seed mouse, increases the targeted insertion efficiency (Ohtsuka et al., 2015). To use this improved PITT (i-PITT) strategy, mRNAs from both platforms should be added to the microinjection solution for simultaneous translation of Cre and PhiC31 in injected fertilized eggs. Seed mouse genetic background It has been reported that mouse genetic background influences traditional transgenic rates. For example, FVB/N mice are known to be highly susceptible to transgenesis whereas C57BL/6 strains have lower integration efficiencies (Auerbach et al., 2003). While we have not empirically tested the influence of genetic background on PITT, we predict that strain differences may influence insertion efficiency, rate of successful transplantation, and embryo survival. The genetic backgrounds of our seed mice are reported in Table 3. Troubleshooting If PITT mice are not obtained for a specific project, the following factors should be checked. Confirm that the donor vector, recombinase/integrase mRNA, and seed mouse are compatible. Ensure that the PITT elements in the DOI, recombinase/integrase of choice, and docking sites in the landing pad are all a part of the same Cre-PITT or PhiC31-PITT platform. To determine reagent compatibility for a given platform, please refer to our suggestions in Tables 2 and 3. Ensure that you have evaluated the quality and concentration of nucleic acids in your microinjection solution. Our suggestions for assessment and optimization may be found in the critical parameters section of this protocol. Consider the possibility that expression of the transgene of interest may affect embryo viability. If embryo health is influenced by expression of a specific transgene, engineer an inducible construct that will spatially and/or temporally control expression. If the above factors are satisfactory, then we recommend repeating the microinjection with more (150–200) zygotes. Alternatively, instead of transferring injected fertilized eggs to pseudo-pregnant surrogate females at the two-cell stage, eggs may be cultured in vitro until they develop into blastocysts, which can then be genotyped to screen for targeted insertion prior to transplantation. It is also possible to increase the concentration of nucleic acids in the injection solution or inject a greater volume of the solution if embryo viability has not been compromised at the current concentration (determined by if more than 60% of injected zygotes develop into blastocysts and/or more than 15% of transplanted eggs survive to birth). Anticipated Results PITT is used to generate targeted transgenic founder lines for a DOI of interest. While the outcome for every PITT project is unique depending on the selected DOI, seed mouse, and recombinase/integrase platform, here we provide the general range of anticipated results for targeted integration and gene expression of the donor cassette. Targeted integration of donor cassette Targeted integration efficiency depends on the PITT system used. We have observed efficiencies ranging from 1.9%–62.0% (Ohtsuka et al., 2010, 2012b, 2015). Efficiency will likely be greater than 10% in PITT projects that use mRNA injection instead of plasmid DNA encoding the recombinase/integrase and in projects that simultaneously employ both Cre and PhiC31 systems (Ohtsuka et al., 2013, 2015). While random insertions could presumably occur using this technology, this only happens rarely compared to traditional transgenesis that uses a linear DNA fragment (~2.4% vs. 10–20%) (Ohtsuka et al., 2010; Fielder et al., 2010). Excluding a minor fraction of cases, almost all PITT mice will have the integrated allele at the expected locus as a result of the proper recombination/integration events. The insertion is clean, without any unintended insertions or deletions of genomic DNA. However, it is important to note that several insertion alleles can be anticipated when simultaneously using several platforms, such as Cre and PhiC31 in i-PITT (Ohtsuka et al., 2015). Insertion of vector sequences at the landing pad In some currently available PITT designs, insertion of the DOI leaves trace vector backbone sequences near PITT landing pads. These excess sequences include prokaryote-derived elements from the vector that can inhibit stable transgene expression. To ameliorate this problem, many PITT platforms include FRT elements in the donor vectors that allow for genetic removal of these excess sequences by crossing PITT Tg founder mice with transgenic mice carrying a ubiquitously expressed Flp recombinase transgene (Ohtsuka et al., 2010, 2015). An example of a Flp transgenic mouse from JAX Mice that can be used for such purposes is B6.Cg-Tg(Pgk1-FLPo)10Sykr/J. Gene expression of donor cassette Unlike traditional transgenesis, each transgenic founder generated from a given PITT project will express the inserted transgene to similar levels. This is because the transgene resides in a predetermined locus with a single copy configuration, which allows for consistent expression between transgenic siblings and from generation to generation. Although a reliable transgene expression pattern can be obtained in PITT mice, the magnitude of cassette gene expression will vary between PITT projects based upon choice of locus, promoter, cDNA, and inclusion of additional sequence elements (such as multiple expression cassettes, microRNA sequences, internal ribosome entry sites, and polyadenylation signals). PITT can be used for ubiquitous expression, as well as tissue-specific expression when the DOI contains a tissue-specific promoter and is integrated into the Rosa26 locus (Tsuchida et al, in press). The one factor that may hinder reliable gene expression is the presence of proximal prokaryote-derived sequences from the donor vector backbone, which have been integrated as excess cargo in the landing pad. We have shown that the removal of this excess donor vector sequence ensures stable inheritance of cassette gene expression (Ohtsuka et al., 2010). In the unforeseen case that a transgene does not exhibit the expected expression pattern, it is possible that unannotated regulatory sequences reside in the DOI, such as enhancers or silencers. Such unknown DNA interactions can be overcome by incorporating insulator sequences at the both sides of DOI (Madisen et al., 2015). Time Considerations Targeted transgenesis experiments performed through ES cell-based approaches require at least one year to generate chimeric mice, which must be bred to ensure germline transmission of the transgene. On the other hand, PITT takes about 3–4 months to generate a founder mouse line, which invariably transmits the transgene to offspring. A typical time frame for the PITT experimental procedures is outlined below and depicted in Figure 2: Weeks 1 to 4: Designing and building of PITT donor DNA constructs Weeks 3 to 6: Synthesis and purification of DNA and RNA components for microinjection Weeks 4 to 6: Preparation of injection components and initiation of superovulation. Weeks 7 to 9: Isolation of embryos from seed mice, microinjection of PITT components, and embryo transfer Weeks 13 to 15: Genotyping of offspring to identify transgenic founders SLPS was supported by the NSF Graduate Research Fellowship DGE1144152. Any opinion, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation. MO acknowledges the staff from the Support Center for Medical Research and Education and Tokai University for microinjection and mRNA synthesis. We thank Hiromi Miura for technical assistance in preparation of mRNA, donor vector construction and seed mouse development. MO’s lab was supported by Grant-in-Aid for Young Scientists (B) (23700514), Grant-in-Aid for Scientific Research (25290035 and 16H04685) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT), and by MEXT-Supported Program for the Strategic Research Foundation at Private Universities, 2009–2013. This work was partially supported by an Institutional Development Award (IDeA) to CBG (PI: Shelley Smith) from the National Institute of General Medical Sciences of the National Institutes of Health under grant number P20GM103471. Figure 1 Schematic of the Cre-PITT system (A) Donor vectors containing a project-specific DNA of interest (DOI) flanked by mutant LoxP sites are injected along with Cre (plasmid or Cre-mRNA) into fertilized eggs collected from seed mice. (B) Blastocysts (top) and neonates (bottom) resulting from PITT of CAG-driven fluorescent reporters at the Rosa26 locus exhibit ubiquitous and high expression that is consistent between transgenic littermates and from generation to generation (adapted from Ohtsuka et al., 2010). The PhiC31-PITT platform follows a similar methodology except that it uses PhiC31 (instead of Cre) and attB/P (instead of LoxP elements). Figure 2 Overview of PITT steps. Figure 3 Examples of simple and complex PITT donor vectors Simple PITT donor vectors contain either (A) attB elements for the PhiC31 platform or (B) LoxP elements for the Cre-PITT platform. (C) This complex PITT donor vector contains PITT elements for both platforms (Cre-LoxP and PhiC31-attB). While not a part of the PITT platform, FRT elements in donor vectors serve as a tool for removing extra sequences in the founder mice using Flp recombinase. Figure 4 Examples of simple and complex landing pads in seed mouse strains (A–D) Simple landing pads include the critical elements to facilitate targeted insertion for a single PITT-platform. Examples of landing pads include LoxP variant recombination sites (Cre-PITT; A and B) or attP integration sites (PhiC31-PITT; C and D). (E) Complex landing pads can include a combination of both attP and LoxP elements for using PhiC31-PITT and Cre-PITT platforms either independently or together. While not a part of the PITT platform, FRT elements in landing pads serve as a tool for removing extra sequences in the founder mice using Flp recombinase. Figure 5 Generation of mRNA encoding Cre recombinase or PhiC31 integrase (A) Plasmids (pBBI = iCre; pBBK = PhiC31o) are linearized with XbaI prior to T7 RNA polymerase-initiated in vitro transcription. (B) A non-denaturing agarose gel image shows 1 μg of synthesized mRNAs that are run alongside a 100bp DNA ladder (M). The approximate sizes for iCre and PhiC31o mRNAs are 1.3 kb and 2.1 kb, respectively. Note that the mRNAs do not migrate perfectly with the DNA molecular weight marker. For accurate size analysis, RNA markers may be included in the gel. Figure 6 PCR-based genotyping of PITT transgenic founders (A) Schematic of PCR sets typically used for genotyping: 5′ junction PCR, 3′ junction PCR, and internal PCR. Primer sets are designed to amplify 200–800 bp. (B) A sample agarose gel run with a 100 bp ladder (M) that demonstrates a transgenic founder identified by a 5′ junction PCR (lane 1) compared to other tested mice (2–8) and a wild type control (wt). Table 1 Sequence elements used in the PITT donor vectors and seed mice Sequence element in the PITT donor vector Corresponding element in the landing pad of the seed mouse Notes PITT Platform Element Sequence Element Sequence Cre-PITT Lox2272 ATAACTTCGTATAGGATACTTTATACGAAGTTAT Lox2272 ATAACTTCGTATAGGATACTTTATACGAAGTTAT Donor DNA between the LoxP variants gets inserted through Cre-recombination-mediated cassette exchange. JTZ17 ATAACTTCGTATAGCATACATTATAGCAATTTAT JT15 AATTATTCGTATAGCATACATTATACGAAGTTAT PhiC31-PITT attB CCGCGGTGCGGGTGCCAGGGCGTGCCCTTGGGCTCCCCGGGCGCGTACTCCAC attP GTAGTGCCCCAACTGGGGTAACCTTTGAGTTCTCTCAGTTGGGGGCGTAG The entire donor plasmid gets inserted through PhiC31-integrase mediated integration. Table 2 List of representative* PITT donor vectors and plasmids for mRNA synthesis Vector name (Available from) DNA of interest (DOI) Landing pads (PITT Platform) Vector features Restriction site Compatible seed mouse strains References pAOM (Request from Authors) tdTomato expression cassette, “CAG-tdTomato-pA” JTZ17, Lox2272 (Cre-PITT) Contains “IRES-lacZ-pA” and “CAG-hyg-pA” cassettes in a pUC119-based vector. Has ampR for ampicillin resistance and FRT sites for Flp-mediated extra sequence excision. tdTomato can be replaced with a different DOI using AgeI and FseI. TOKMO-1 TOKMO-2 (Ohtsuka et al., 2010) pAOT (Request from Authors) eGFP expression cassette with a synthetic miRNA against the gene encoding tyrosinase, “CAG-eGFP-miR(Tyr-1/2)-pA” JTZ17, Lox2272 (Cre-PITT) Contains “IRES-lacZ-pA” and “CAG-hyg-pA” cassettes in a pUC119-based vector. Also has ampR for ampicillin resistance and FRT sites for Flp-mediated extra sequence excision. eGFP can be replaced with a different reporter using AgeI and BsrGI. The miRNA can be replaced with a different miRNA using BamHI and BglII TOKMO-1 TOKMO-2 (Ohtsuka et al., 2010) pA748 (Request from Authors) Sucrose counterselection cassettes with GFPuv, “sacB-GFPuv-sacB” cassette JTZ17, Lox2272 (Cre-PITT) Contains “IRES-eGFP-pA” and “CAG-hyg-pA” cassettes in a pUC119-based vector. Has ampR for ampicillin resistance and FRT sites for Flp-mediated extra sequence excision. The DOI cassette can be replaced with a different DOI using NotI. TOKMO-1 TOKMO-2 (Ohtsuka et al., 2010) pAWV (Addgene #62710) tdTomato expression cassette with a synthetic miRNA against the GFP gene, “CAG-tdTomato-miR(eGFP)-pA” JTZ17, Lox2272 (Cre-PITT) pBR322-based vector. Contains a “CAG-FLPe-pA” cassette to aid in self-removal of extra sequence by Flp-recombination. Has ampR for ampicillin resistance. The entire DOI cassette can be replaced with a different DOI using NotI. The miRNA region can be replaced with a different miRNA using SalI and BglII. TOKMO-1 TOKMO-2 (Ohtsuka et al., 2013; Miura et al., 2015) pAWK (Addgene #62713) Expression cassette for synthetic miRNA against the GFP gene, “CAG-miR(eGFP)-pA” JTZ17, Lox2272 (Cre-PITT) pBR322-based vector. Contains a “CAG-FLPe-pA” cassette to aid in self-removal of extra sequence by Flp-recombination. Has ampR for ampicillin resistance. The entire DOI cassette can be replaced with a different DOI using NotI. The miRNA region can be replaced with a different miRNA using SalI and BglII. TOKMO-1 TOKMO-2 (Ohtsuka et al., 2013; Miura et al., 2015) pBFD (Request from Authors) Dre expression cassette from the Thy1 promoter, “Thy1-Dre-pA” JTZ17, Lox2272 (Cre-PITT) pBR322-based vector. Has ampR for ampicillin resistance and FRT sites for Flp-mediated extra sequence excision. The entire DOI cassette can be replaced with a different DOI by EagI. TOKMO-1 TOKMO-2 (Ohtsuka et al., 2013) pBDR (Addgene #62663) Promoterless tdTomato cassette, “tdTomato-pA” JTZ17, Lox2272, attB (Cre-PITT and/or PhiC31-PITT) pIDTSMART-based vector. Has kanR for kanamycin resistance and mutant FRT sites (F14 and F15) for Flp-mediated extra sequence excision. The entire DOI cassette can be replaced with a different DOI using AgeI and EcoRI. The tdTomato can be replaced using AgeI and FseI. TOKMO-3 (Ohtsuka et al., 2015) pBHL (Request from Authors) Promoterless tdTomato cassette, “tdTomato-pA” JTZ17, Lox2272, attB (Cre-PITT and/or PhiC31-PITT) This plasmid is derived from pBDR but the order of attB and JTZ17 is reversed. The entire DOI cassette can be replaced with a different DOI using AgeI and EcoRI. The tdTomato can be replaced using AgeI and FseI. TOKMO-3 Unpublished pBBI (Addgene #65795) iCre expression cassette, “T7-iCre-AAAA…” – To be used for iCre mRNA synthesis. It is derived from pcDNA3.1 and has ampR for ampicillin resistance. Linearize with XbaI before mRNA synthesis. – (Ohtsuka et al., 2013) pBBK (Addgene #62670) PhiC31o expression cassette, “T7-PhiC31o- AAAA…” – To be used for PhiC31o mRNA synthesis. It is derived from pcDNA3.1 and has ampR for ampicillin resistance. Linearize with XbaI before mRNA synthesis. – (Ohtsuka et al., 2015) Key: AAAA… = polyA stretch; ampR = ampicillin resistance cassette; CAG = synthetic cytomegalovirus (CMV) early enhancer/ chicken ß-actin promoter/ rabbit ß-globin splice acceptor site for strong expression in mammalian cells; DOI = DNA of interest; Dre = encodes phage D6 recombinase; eGFP = encodes enhanced green fluorescent protein; FLPe = encodes enhanced FLP recombinase; GFPuv = encodes ultraviolet light-excitable green fluorescent protein; hyg = hygromycin-resistance cassette; iCre = encodes a codon-improved Cre recombinase; IRES = internal ribosome entry site; kanR = kanamycin resistance cassette; lacZ = encodes ß-galactosidase; pA = polyA sequence; PhiC31o = encodes a codon-optimized PhiC31 integrase; puro = puromycin resistance cassette; sacB = encodes levansucrase and confers sensitivity to sucrose as counterselection; T7 = T7 prokaryotic promoter; tdTomato = encodes a red fluorescent protein; Thy1 = promoter frequently used for expression in neurons. * Please note that this is not a comprehensive list of PITT donor vectors or plasmids for mRNA synthesis. Other versions of available plasmids may be found in our previous publications (Ohtsuka et al., 2010, 2012a, 2013, 2015). In addition, the construction of donor vectors for the R26 Cr4 mouse lines (described in Table 3) is in progress and these vectors can be requested from the authors. There are also other PhiC31-based vector systems reported in Tasic et al., 2011, which are available through Charles River Laboratories, USA, under the trade name TARGATT™ system. Table 3 List of representative* PITT seed mouse strains Common name Strain name (Genetic background) Landing pads used (PITT Platform) Locus Available from References TOKMO-1 Gt(ROSA)26Sor<tm1Maoh> (129/C57BL/6J mixed) JT15, Lox2272 (Cre-PITT) Rosa26 Authors (Ohtsuka et al., 2010) TOKMO-2 H2-T3<tm1Maoh> (129/C57BL/6J mixed) JT15, Lox2272 (Cre-PITT) H2-Tw3 Authors (Ohtsuka et al., 2010) R26 Cr4-Cre-PITT CRISPR(ROSA)(Cre-PITT)CBG> (129/C57BL/6N mixed) JT15, Lox2272 (Cre-PITT) Rosa26 Authors (Quadros et al., 2015) R26 Cr4-PhiC-PITT C57BL/6N- CRISPR(ROSA)(Cre-PITT)CBG> (C57BL/6N) attP (PhiC31-PITT) Rosa26 Authors Unpublished TOKMO-3 C57BL/6N-Gt(ROSA)26Sor<tm10(PITT)Maoh> (C57BL/6N) JT15, Lox2272, attP (Cre-PITT and/or PhiC31-PITT) Rosa26 RIKEN BioResource Center, RBRC06517 (Ohtsuka et al., 2015) R26 Cr4-CrePhiC-PITT CRISPR(ROSA)(Cre/PhiC31-PITT)CBG> (129/C57BL/6N mixed) JT15, Lox2272, attP (Cre-PITT and/or PhiC31-PITT) Rosa26 Authors Unpublished * Please note that this is not a comprehensive list of PITT seed mice. Specifically, there are other PhiC31-based seed mice reported in Tasic et al., 2011, which are available through Charles River Laboratories, USA, under the trade name TARGATT™ system. KEY REFERENCE Some sections of this protocol were adapted from Chapter 1 of Methods in Molecular Biology: “Development of Pronuclear Injection-Based Targeted Transgenesis in Mice Through Cre-LoxP Site-Specific Recombination” (Ohtsuka, 2014). LITERATURE CITED Auerbach AB Norinsky R Ho W Losos K Guo Q Chatterjee S Joyner AL 2003 Strain-dependent differences in the efficiency of transgenic mouse production Transgenic Research 12 59 69 12650525 Behringer R Gertsenstein M Nagy KV Nagy A 2014 Manipulating the mouse embryo: a laboratory manual Fielder TJ Barrios L Montoliu L 2010 A survey to establish performance standards for the production of transgenic mice Transgenic Research 19 675 681 19842055 Gordon JW Ruddle FH 1981 Integration and stable germ line transmission of genes injected into mouse pronuclei Science (New York, NY) 214 1244 1246 Gordon JW Scangos GA Plotkin DJ Barbosa JA Ruddle FH 1980 Genetic transformation of mouse embryos by microinjection of purified DNA Proceedings of the National Academy of Sciences of the United States of America 77 7380 7384 6261253 Gossler A Doetschman T Korn R Serfling E Kemler R 1986 Transgenesis by means of blastocyst-derived embryonic stem cell lines Proceedings of the National Academy of Sciences of the United States of America 83 9065 9069 3024164 Gurumurthy CB Takahashi G Wada K Miura H Sato M Ohtsuka M 2016 GONAD: A Novel CRISPR/Cas9 Genome Editing Method that Does Not Require Ex Vivo Handling of Embryos: GONAD: A Novel CRISPR/Cas9 Genome Editing Method Current Protocols in Human Genetics Haines JL Korf BR Morton CC Seidman CE Seidman JG Smith DR 15.8.1 15.8.12 John Wiley & Sons, Inc Hoboken, NJ, USA Available at: http://doi.wiley.com/10.1002/0471142905.hg1508s88 [Accessed May 17, 2016]. Harms DW Quadros RM Seruggia D Ohtsuka M Takahashi G Montoliu L Gurumurthy CB 2014 Mouse Genome Editing Using the CRISPR/Cas System Current Protocols in Human Genetics/Editorial Board, Jonathan L. Haines … [et Al.] 83 15.7.1 15.7.27 Hogan B 1983 Molecular biology. Enhancers, chromosome position effects, and transgenic mice Nature 306 313 314 6646213 International Society for Transgenic Technologies 2011 Advanced protocols for animal transgenesis: an ISTT manual Pease S Saunders TL Springer Heidelberg, New York Jones D 2011 Genetic engineering of a mouse: Dr. Frank Ruddle and somatic cell genetics The Yale Journal of Biology and Medicine 84 117 124 21698043 Madisen L Garner AR Shimaoka D Chuong AS Klapoetke NC Li L van der Bourg A Niino Y Egolf L Monetti C 2015 Transgenic Mice for Intersectional Targeting of Neural Sensors and Effectors with High Specificity and Performance Neuron 85 942 958 25741722 Miura H Inoko H Tanaka M Nakaoka H Kimura M Gurumurthy CB Sato M Ohtsuka M 2015 Assessment of Artificial MiRNA Architectures for Higher Knockdown Efficiencies without the Undesired Effects in Mice PloS One 10 e0135919 26285215 Ohtsuka M 2014 Development of pronuclear injection-based targeted transgenesis in mice through Cre-loxP site-specific recombination Methods in Molecular Biology (Clifton, NJ) 1194 3 19 Ohtsuka M Miura H Hayashi H Nakaoka H Kimura M Sato M Gurumurthy CB Inoko H 2013 Improvement of pronuclear injection-based targeted transgenesis (PITT) by iCre mRNA-mediated site-specific recombination Transgenic research 22 873 875 23529203 Ohtsuka M Miura H Mochida K Hirose M Hasegawa A Ogura A Mizutani R Kimura M Isotani A Ikawa M 2015 One-step generation of multiple transgenic mouse lines using an improved Pronuclear Injection-based Targeted Transgenesis (i-PITT) BMC genomics 16 274 25887549 Ohtsuka M Miura H Nakaoka H Kimura M Sato M Inoko H 2012a Targeted transgenesis through pronuclear injection of improved vectors into in vitro fertilized eggs Transgenic research 21 225 226 21437715 Ohtsuka M Miura H Sato M Kimura M Inoko H Gurumurthy CB 2012b PITT: pronuclear injection-based targeted transgenesis, a reliable transgene expression method in mice Experimental animals/Japanese Association for Laboratory Animal Science 61 489 502 Ohtsuka M Ogiwara S Miura H Mizutani A Warita T Sato M Imai K Hozumi K Sato T Tanaka M 2010 Pronuclear injection-based mouse targeted transgenesis for reproducible and highly efficient transgene expression Nucleic acids research 38 e198 20880997 Pease S Saunders TL 2011 Advanced Protocols for Animal Transgenesis Springer Berlin Heidelberg Berlin, Heidelberg Available at: http://link.springer.com/10.1007/978-3-642-20792-1 [Accessed June 22, 2014] Quadros RM Harms DW Ohtsuka M Gurumurthy CB 2015 Insertion of sequences at the original provirus integration site of mouse ROSA26 locus using the CRISPR/Cas9 system FEBS open bio 5 191 197 Takahashi H Liu C 2010 Archiving and Distributing Mouse Lines by Sperm Cryopreservation, IVF, and Embryo Transfer Methods in Enzymology 53 69 Elsevier Available at: http://linkinghub.elsevier.com/retrieve/pii/S0076687910760043 [Accessed May 17, 2016] Tasic B Hippenmeyer S Wang C Gamboa M Zong H Chen-Tsai Y Luo L 2011 Site-specific integrase-mediated transgenesis in mice via pronuclear injection Proceedings of the National Academy of Sciences of the United States of America 108 7902 7907 21464299 Tsuchida J Matsusaka T Ohtsuka M Miura H Okuno Y Asanuma K Nakagawa T Yanagita M Mori K Establishment of Nephrin Reporter Mice and its Use for Chemical Screening PLOS ONE
PMC005xxxxxx/PMC5123765.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 9432918 8591 Immunity Immunity Immunity 1074-7613 1097-4180 27742546 5123765 10.1016/j.immuni.2016.09.013 NIHMS820235 Article Deficient activity of the nuclease MRE11A induces T cell aging and promotes arthritogenic effector functions in patients with rheumatoid arthritis Li Yinyin 1 Shen Yi 1 Hohensinner Philipp 12 Ju Jihang 1 Wen Zhenke 1 Goodman Stuart B. 3 Zhang Hui 1 Goronzy Jörg J. 1 Weyand Cornelia M. 14 1 Division of Immunology and Rheumatology, Stanford University School of Medicine, Stanford, CA 94305 2 Department of Internal Medicine II/Cardiology, Medical University of Vienna, Vienna, Austria 3 Department of Orthopedic Surgery and Bioengineering, Stanford University School of Medicine, Stanford, CA 94305 Correspondence To: Cornelia M. Weyand, M.D., Ph.D., Division of Immunology and Rheumatology, Department of Medicine, Stanford University, CCSR Building Room 2225, MC-5166, 269 Campus Drive West, Stanford, CA 94305; Phone: (650) 723-9027, Fax: (650) 721-1251, [email protected] 4 Lead Contact 1 10 2016 11 10 2016 18 10 2016 18 10 2017 45 4 903916 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Immune aging manifests with a combination of failing adaptive immunity and insufficiently restrained inflammation. In patients with rheumatoid arthritis (RA), T cell aging occurs prematurely, but the mechanisms involved and their contribution to tissue-destructive inflammation remain unclear. We found that RA CD4+ T cells showed signs of aging during their primary immune responses and differentiated into tissue-invasive, pro-inflammatory effector cells. RA T cells had low expression of the double-strand-break repair nuclease MRE11A, leading to telomeric damage, juxtacentromeric heterochromatin unraveling, and senescence marker upregulation. Inhibition of MRE11A activity in healthy T cells induced the aging phenotype, whereas MRE11A overexpression in RA T cells reversed it. In human-synovium chimeric mice, MRE11Alow T cells were tissue-invasive and pro-arthritogenic, and MRE11A reconstitution mitigated synovitis. Our findings link premature T cell aging and tissue-invasiveness to telomere deprotection and heterochromatin unpacking, identifying MRE11A as a therapeutic target to combat immune aging and suppress dysregulated tissue inflammation. INTRODUCTION Preceded by a decade-long period of preclinical disease, rheumatoid arthritis (RA) manifests with a symmetrical polyarthritis causing irreversible cartilage and bone destruction and shortens life expectancy due to accelerated cardiovascular disease. Immune aging affects the general population after 50 years of age, but is accelerated in RA patients (Weyand et al., 2009), where it is already noticeable in antigen-unprimed naïve T cells (Koetz et al., 2000). Cells devote a significant proportion of their machinery to DNA surveillance and repair to prevent cellular aging or death associated with genome instability (Chow and Herrup, 2015). Predictable loss of telomeric sequences with each cell replication allows telomeres to serve as molecular clocks. By tallying the number of cell divisions, telomeres are believed to effectively force mutation-harboring cells into cell cycle arrest. Senescent T cells not only remain viable, but actively shape the tissue microenvironment by secreting cytokines and tissue remodeling factors (Weyand et al., 2014). However, despite several senescence features, aging human lymphocytes are not in replicative arrest (Yang et al., 2016) and continue to participate in clonal expansion, distinguishing lymphocyte aging from senescence (Akbar and Henson, 2011; Chou and Effros, 2013; Sharpless and Sherr, 2015). Reversibility of senescence in human end-differentiated effector T cells further supports the model that aging of lymphocytes reflects progressive differentiation more than true senescence (Di Mitri et al., 2011). Whether aging T cells acquire effector functions that mediate tissue inflammation is not understood. Abnormalities in the DNA damage sensing and repair machinery of RA T cells have raised the question of whether such defects are mechanistically linked to T cell aging and to arthritogenic effector functions (Shao et al., 2009; Shao et al., 2010). The MRN complex, composed of Meiotic Recombination 11 Homolog A (MRE11A), RAD50 and Nijmegen Breakage Syndrome 1 (NBS1), senses DNA double-strand breaks to amplify DNA repair (Lamarche et al., 2010). The core component of the complex, MRE11A, has double-stranded (ds)DNA exonuclease and single-stranded (ss)DNA endonuclease activity in both homologous recombination and nonhomologous end-joining (Xie et al., 2009). MRE11A is recruited to healthy telomeres, where its function is not understood. In S. cerevisiae, RAD50 or MRE11A loss shortens telomeres and abolishes S phase telomerase recruitment (Takata et al., 2005). In high eukaryotes, MRE11A specifically interacts with the shelterin protein TRF2 (Diotti and Loayza, 2011). Conversely, MRN complex hypomorphism in mouse embryonic fibroblasts does not affect telomere length (Attwooll et al., 2009) and reduces telomere dysfunction-induced foci (TIF), revealing the complexity of MRE11A function. Which role MRE11A plays in healthy and stressed human cells, particularly in long-lived lymphocytes, is unknown. Here, we report that in RA T cells, age-related telomeric defects manifested not only as shortening, but also as damage accumulation. Such T cells unfolded higher-order satellite heterochromatin, expressed increased levels of the cell cycle regulators p16 and p21 but lacked p53. Aged RA T cells were low-expressers for the DNA repair nuclease MRE11A and their chromosomal ends were MRE11Alow. In healthy T cells, decreasing MRE11A mRNA by treatment with short interfering RNAs (siRNAs) or pharmacologic inhibition of MRE11A’s nucleolytic activity promptly induced telomeric damage and upregulated the senescence markers p16, p21, and CD57, concomitant with unraveling of pericentromeric satellite DNA. Spontaneous or induced deficiency of MRE11A’s nucleolytic function had a profound impact on T cell behavior and rendered T cells tissue-invasive and pro-arthrogenic, whereas reconstitution of MRE11A protein in patient-derived T cells protected synovial tissue from inflammatory attack. These data provide mechanistic evidence for a role of the MRE11A nuclease in not only regulating aging but also differentiation of T cells into tissue-injurious effector cells. RESULTS Telomeres in RA T cells are not only shortened, but damaged Telomeric sequences in RA T cells are shortened relative to T cells from age-matched healthy individuals, and this has been attributed to increased proliferative pressure in an inflammatory environment (Koetz et al., 2000). However, T cell turnover measured by the proliferation marker Ki-67 correlates inversely with telomeric erosion (Schonland et al., 2003), suggesting alternative mechanisms underlying telomeric loss. Telo-FISH staining in metaphase nuclei confirmed that the vast majority of RA naïve CD4+ T cells had low-intensity telomeric signals (Figure 1A and 1B). Compared to healthy individuals, RA patients lacked high-brightness nuclei and almost all of their cells had a weak telomere signal (Figure 1C). The structural intactness of telomeric caps was examined through FISH-probe hybridization patterns in metaphase-arrested nuclei, revealing four structural patterns: telomeric apposition (long chromosomal arms opposed), telomeric fragmentation (signal doubling), telomeric loss (signal-free end) and telomeric fusion (physical linkage of two ends of the same or neighboring chromosomes) (Figure 1D). The majority of healthy metaphase T cell nuclei had structurally intact telomeres (56.5%). Thirty-one percent of nuclei displayed signal doublets and 12.5% displayed telomeric apposition. In contrast, RA T cell nuclei consistently contained damaged telomeres. End signal doubling, indicative of telomeric fragility, occurred in 41.9% of cells and 24.1% had opposed telomeres. Albeit encountered infrequently, RA samples showed evidence for severe structural damage, including fusions (2.8%) and total signal loss (1.6%) (Figure 1E). To understand the relationship between proliferation-induced telomeric shortening and TIF formation, we monitored TIF evolution in polyclonally expanding T cell populations by quantifying the recruitment of the DNA damage protein P53-Binding Protein 1 (53PB1) to telomeric ends marked through the shelterin protein TRF2. Naïve and memory CD4+ T cells from the same donor were placed under proliferative stress (Figure S1A) by repetitive polyclonal stimulation. Enforced proliferation markedly shortened telomeres in the naïve population, but lengths were relatively stable in memory counterparts, despite similar proliferation rates in both T cell subpopulations (Figure S1A, S1B, S1D). Naïve CD4+ T cells lost >1000 telomeric nucleotides over 3 weeks, but memory CD4+ T cell chromosomes were shortened by 300 base pairs (bp). Parallel monitoring of TIF by quantifying 53BP1 and TRF2 colocalization demonstrated rare foci in naïve CD4+ T cells, even after >1000 bp loss. In contrast, TIF appeared as early as day 7 in memory CD4+ T cells (Figure S1C). Overall, 53BP1 and TRF2 colocalization and erosion of telomeric sequences were unrelated (Figure 1F). Together, these data documented that telomeres in RA T cells are shortened, but more importantly, are sites of DNA damage. Premature aging of naïve CD4+ T cells in RA patients is related to telomeric damage The presence of circulating and synovial CD4+CD28null T cells in RA patients provides unequivocal evidence for in vivo T cell aging. Such end-differentiated T cells express CD57 and natural killer (NK) cell receptors (Warrington et al., 2001). To better define the aging process in naïve CD4+CD45RA+ T cells, we assessed the aging-associated cell cycle inhibitors p16, p21, and p53 (Sharpless and Sherr, 2015). Compared to healthy controls, RA T cells expressed higher levels of CDKN2A and CDKN1A mRNA and p16 protein (Figure 2A–B), while TP53 expression was reduced, suggesting a p16-dependent and p53-independent aging pathway (Figure 2A). The glucuronyltransferase gene family member CD57 is considered a senescence marker on T cells (Focosi et al., 2010). The frequency of CD4+CD45RA+CD57+ T cells was low in healthy individuals (7%), but more than doubled in RA patients (20%) (Figure 2C). DNA condensation into senescent-associated heterochromatin foci is a hallmark of cellular senescence, but early stages of cellular aging can be associated with the unravelling of constitutive heterochromatin into visibly extended structures (Swanson et. al., 2013). To identify heterochromatin changes, we analyzed CD4+CD45RA+ T cell populations for evidence of unraveling of pericentromeric and centromeric satellite DNA. Hybridization of a probe to satellite-II (sat-II) DNA in stimulated healthy CD4+CD45RA+ T cells yielded compact, round and centric satellite signals (Figure 2D). Conversely, RA T cells had distended pericentromeric and centromeric satellites, usually affecting multiple sites within a nucleus. Eighty percent of RA cells had clearly unraveled satellites. In contrast, 80% of control cells had tightly packaged condensed satellites (Figure 2E), demonstrating that pericentromeric and centromeric satellite heterochromatin in RA T cells undergoes decondensation, consistent with entry into early senescence. To investigate whether aging-associated T cell phenotypes are mechanistically linked to telomeric damage, we used siRNAs to decrease telomere-protecting shelterin TERF2 mRNA. Transfecting TERF2 siRNAs or a TERF2 mutant plasmid reduced TERF2 mRNA (Figure S3A); induced robust TIF formation (Figure S3B) and doubled CDKN2A and CDKN1A mRNA transcripts (Figure 2F, Figure S3C). TERF2 knockdown was sufficient to induce p16 and CD57 expression (Figure 2G–H), recapitulating conditions in RA patients. T cells from RA patients are MRE11A deficient The presence of telomeric damage foci and satellite DNA distention in RA T cells raised the question whether the DNA repair machinery was fully intact. We examined RA T cells for the expression of key molecules involved in the DNA double strand break repair, including Ataxia Telangiectasia and Rad3-Related Protein (ATR), Ataxia Telangiectasia Mutated (ATM), MRE11A, NBS1, RAD50 and DNA-dependent protein kinase catalytic subunit (DNA-PKcs) by flow cytometry. Intracellular staining of peripheral blood mononuclear cells (PBMCs) demonstrated reduced MRE11A protein expression in naïve and memory CD4+ T cells from RA patients (Figure 3A–B). In both T cell subpopulations, MRE11A protein expression declined in an age-dependent fashion in RA patients as well as healthy donors. However, MRE11A protein was consistently lower in RA patients’ cells; reduced by 40–50% in the naïve population and 35% in the memory population. CD4+ T cells of patients with the inflammatory polyarthritis, PsA, were indistinguishable from healthy T cells (Figure S4A). In RA patients, but not in PsA patients, CD8+ T cells, CD14+ monocytes and CD19+ B cells were also MRE11Alow expressors (Figure S4B, S4C, S4D). MRE11A low-expression was most pronounced in T cells from untreated RA patients, eliminating immunosuppressive therapy as the causative factor (Figure 3C–D). Total nuclear and telomere-bound MRE11A was analyzed by dual-color immunostaining with anti-MRE11A and anti-TRF2. Staining intensity for MRE11A protein was markedly reduced in RA T cell nuclei, in which the signal for MRE11A and TRF2 colocalization was especially low, suggesting that MRE11Alow RA T cells have telomeric tips almost depleted of the nuclease (Figure 3E–F). Thus, aging is associated with progressive decline in the expression of the double strand break repair nuclease MRE11A, a process which is notably accelerated in RA T cells. Diminished MRE11A activity causes telomeric damage The exonuclease and endonuclease MRE11A partners with the ATPase RAD50 and NBS1 to form the MRN complex, a DNA break sensor critically involved in multiple repair pathways (Lamarche et al., 2010). siRNA-mediated knockdown of each of the three MRN components, MRE11A, RAD50 and NBS1, produced partial depletion of both transcripts and protein (Figure S5A–S5B) and induced widespread DNA damage (Figure 4A). The amount of unrepaired DNA breaks was quantified by measuring the intensity of 53BP1-binding to damaged chromatin sites, where this checkpoint protein forms foci and regulates end-joining-mediated repair processes. MRE11A loss was most devastating for DNA integrity; 53BP1 was recruited to numerous DNA break sites in almost every single nucleus. In contrast, NBS1 knockdown left some nuclei unaffected. Partial loss of RAD50 was tolerated the best, producing the lowest 53BP1 signal. 53BP1 and TRF2 colocalization studies confirmed the functional relevance of MRE11A at the telomere (Figure S5C). Overall, transfection of MRE11A-specific siRNA produced robust DNA damage, with almost 50% of the 53BP1 signal localizing to telomeric ends. Only one-third of the 53BP1 signal copositioned with the shelterin protein TRF2 in NBS1 or RAD50 insufficient cells (Figure 4B). Alternatively, the function of MRE11A was impaired by treating with the small molecule inhibitor Mirin, which inhibits MRE11A’s exonuclease activity while sparing the endonuclease activity (Dupre et al., 2008; Shibata et al., 2014). Inhibition of exonuclease activity in healthy naïve CD4+ T cells resulted in massive DNA damage, of which a substantial proportion localized to the telomere (Figure 4C and Figure S5D). These data confirmed that the nucleolytic activity of MRE11A is particularly important in protecting telomeric integrity. Impairing MRE11A activity induces the aging profile of RA T cells To investigate whether MRE11Alow T cells are prone to accelerate the aging process, we targeted MRE11A, RAD50 and NBS1 with interfering RNA technology. MRE11A knockdown potently induced CDKN2A, NBS1 knockdown resulted in a significant, but small, increase in CDKN2A mRNA and RAD50 knockdown left CDKN2A expression unchanged. TP53 expression was unaffected by reducing MRE11A expression (Figure 4D). Treatment with the MRE11A inhibitor Mirin led to similar results (Figure 4E), indicating that MRE11A’s nucleolytic activity protected T cells from entering the aging program. Overall, the pattern of p16 induction paralleled the degree of telomeric damage induced by the genetic inhibition of the three molecules (Figure 2), with MRE11A trumping the two other components in TIF induction and p16 upregulation. Weaning of the inhibitory effects of the siRNAs and recovery of MRE11A expression were sufficient to reverse the upregulation of CDKN2A (encoding p16) and CDKN1A (encoding p21) (Figure S5E–S5F). CD57 expression in MRE11A siRNA-treated cells (Figure 4F) recapitulated the spontaneous aging profile in patient-derived T cells (Figure 2). Similarly, Mirin treatment more than doubled the frequency of CD57+CD4+ T cells (Figure 4G) and rapidly induced heterochromatin unpacking (Figure 4H). More than 80% of Mirin-treated cells had distended satellites, reproducing the constitutive satellite DNA unraveling in RA T cells (Figure 4I). These experiments mechanistically linked the nuclease MRE11A to the aging phenotype of T cells, including the upregulation of cell cycle regulators and the unraveling of satellite chromatin. Restoring MRE11A expression repairs telomeric damage and prevents T cell aging To examine whether MRE11A restoration is sufficient to revert the aging phenotype, an MRE11A construct was overexpressed in RA CD4+ naïve T cells (Figure 5A and 5B). Increasing the MRE11A protein concentration prevented telomeric 53BP1 recruitment; 53BP1 and TRF2 colocalization declined by 75% (Figure 5C and 5D), suggesting a direct role of MRE11A in telomere repair. Other components of the T cell aging program, such as p16 expression, were similarly affected (Figure 5E–5G). Sixty percent of control-transfected RA T cells were p16pos and frequencies were reduced to 20% after MRE11A overexpression (Figure 5H). Consistent with the data in Figure 4D, p53 expression appeared independent of MRE11A protein concentrations. These data confirmed that MRE11A has anti-aging function by securing telomeric intactness and controlling cell cycle regulators and that the aging signature of RA T cells is reversible. Inhibiting MRE11A’s nucleolytic activity renders T cells tissue-invasive and proinflammatory CD4+CD28null T cells participate in the inflammatory lesions in RA joints (Schmidt et al., 1996), but whether their accelerated aging is mechanistically connected to their proinflammatory functions is unknown. To establish tissue inflammation, T cells must migrate into the tissue site where they display proinflammatory effector functions, such as cytokine and chemokine release, activation of endothelial cells, stromal cells and innate immune cells. All phases of this process are captured in a humanized mouse model in which human synovial tissue is engrafted into immunodeficient mice and human CD45RO− PBMCs, either deriving from healthy individuals or from patients with RA, are adoptively transferred. Chimeric mice were reconstituted with either control or RA PBMCs and were treated with the MRE11A inhibitor Mirin or vehicle (Figure 6). Adoptively transferred PBMCs from healthy individuals formed sparse T cell infiltrates in the synovial grafts (Figure 6A). Inhibition of MRE11A’s nucleolytic activity enhanced T cell accumulation in the synovial tissue. MRE11Alow RA PBMCs spontaneously had a much higher degree of tissue invasiveness. Here, MRE11A inhibition marginally enhanced T cell recruitment into the synovium. Synovial T cell accumulation was similar in Mirin-treated control PBMCs and untreated RA PBMCs (Figure 6B–C). Inhibiting MRE11A activity resulted in upregulated expression of TNFSF11 (encoding RANKL), a molecule critically involved in rheumatoid bone destruction (Figure 6D). Also, the intensity of tissue inflammation was increased, as measured by the expression of the pro-inflammatory cytokines TNF, IL6 and IL1B (Figure 6E). Conversely, reducing MRE11A activity prompted increased tissue expression of the regulatory cytokines TGFB1 and IL10 (Figure 6F). Exposure of synovial tissue to the MRE11A inhibitor in the absence of transferred PBMCs had no impact on cytokine expression (Figure 6E–F). To examine a possible link between tissue invasiveness, inflammatory capability and cellular aging, we quantified tissue expression of CDKN2A, CDKN1A and TP53 transcripts and measured p16 protein expression by dual-color immunohistochemistry (Figure 6G–I). Mirin treatment left the cell cycle regulator expression in noninflamed synovial cells unaffected, but promptly induced the CDKN2Ahigh, CDKN1Ahigh, TP53low profile in tissue-infiltrating cells (Figure 6G). Loss of MRE11A function rendered T cells and non-T cells p16-positive (Figure 6H–I). RA T cells formed dense synovial infiltrates, with a visible enrichment of CD3+ p16+ cells. Also, MRE11Alow RA T cells had a hypermigratory phenotype in Transwell assays performed in the absence of chemokines (Figure 6J). Hypermotility was shared by T cells from old individuals and T cells treated with the MRE11A inhibitor Mirin (Figure 6J). Overall, healthy T cells with inhibited MRE11A nuclease activity resembled RA T cells. These experiments assigned a mechanistic role to the nuclease MRE11A in several disease-relevant capabilities of RA T cells; specifically, their tissue-invasive and aggressive inflammatory behavior. Restoring MRE11A in RA cells prevents pro-arthritogenic effector functions To test whether restoring MRE11A expression was sufficient to correct the pathogenic behavior of RA CD4+ T cells, we overexpressed MRE11A in naïve and memory CD4+ T cells prior to adoptive transfer. Forced MRE11A overexpression was strongly anti-inflammatory (Figure 7). Transfer of MRE11Ahigh RA T cells minimized tissue TRB transcript levels, diminished the density of the T cell infiltrate, reduced TNF, IL6, IL1B expression, increased TGFB1 and IL10 mRNA and prevented CDKN2A induction. Correcting MRE11A protein concentrations in the transferred T cells did not affect T cell lineage commitment, as indicated by the expression of the T cell transcription factors TBX21, GATA3, RORC and FOXP3 (Figure S6). These experiments confirmed that MRE11A is a critical regulator of pathogenic effector functions in tissue-residing RA T cells and that restoring MRE11A protein expression is sufficient to modify arthritogenic capabilities in naïve and memory CD4+ T cells. DISCUSSION RA is an HLA class II-associated autoimmune disease, in which arthritogenic T cells drive progressive inflammatory destruction of cartilage and bone. Patients with RA have a signature of premature immune aging, providing an excellent model system to study the interrelationship between failing T cell immunity and unremitting tissue inflammation. Here, we present data implicating defective DNA damage responses in accelerating T cell aging, which renders such T cells susceptible to differentiate into tissue-invasive, pro-athrogenic effector cells. Specifically, low expression of the double strand break repair nuclease MRE11A in RA T cells diminishes dsDNA exonucleolytic and ssDNA endonucleolytic activity, which appears critically important in maintaining intact telomeres, preventing entry into the aging program and averting tissue invasion of inflammation-biased T cells. These findings support the concept that telomeric erosion in MRE11Alow T cells is a consequence of insufficient repair and not proliferative stress and that telomeric damage is mechanistically linked to cellular behavior. In this study, we defined the T cell aging program in RA, which is distinct from cellular senescence (Chou and Effros, 2013). How T cells age and whether immune cells follow a “universal” aging program is insufficiently understood, but RA T cells and T cells from aged humans lack the irreversible cell cycle arrest that is considered a cardinal feature of senescence (Goronzy and Weyand, 2013; Sharpless and Sherr, 2015). In contrast, cell cycle passage of RA T cells is shortened and the entire population of patient-derived CD4 T cells is hyperproliferative (Fujii et al., 2009). Synovial T cells have all converted into effector memory T cells, but naïve CD4+ T cells in RA already display abnormalities, e.g., due to increased apoptotic susceptibility clonal expansion is relatively futile (Schonland et al., 2003). Long before differentiating into effector memory cells, RA T cells fail to repair DNA damage (Shao et al., 2009), are energy deprived (Yang et al., 2013) and chronically activate the DNA-PKcs-JNK pathway (Shao et al., 2010). The downregulation of MRE11A was a global phenomenon in RA T cells affecting the entire population and was not selective for a subpopulation. Flow cytometric quantification of MRE11A protein produced unimodal distribution patterns in both CD45RA+ and CD45RA− populations, indicating a broadly distributed defect amongst T cells. MRE11A overexpression rescued the pathogenic effector functions of both, naïve CD45RA+ and memory CD45RO+ CD4+ T cells, when adoptively transferred into chimeric mice. Experiments with isolated CD45RA+CD4+ T cells allowed focusing on early phases in the life cycle of T cells, as they begin to make the transition into effector and memory T cells, that eventually mediate the pathogenic functions in the tissue environment. The human CD8+, but not the CD4+ T cell compartment contains small populations of T cells that increase in frequency with age and that can reacquire a naïve-like phenotype after in vivo stimulation while maintaining the ability to rapidly respond to restimulation (Fuertes Marraco et al., 2015; Pulko et al., 2016). In healthy humans, such CD8 naïve-like T cells with effector function have relatively long telomeres, longer than memory T cells, indicating that they do not have a defect in telomeric maintenance. Cellular senescence, genomic instability, epigenetic alterations, stem cell exhaustion, loss of proteostasis, mitochondrial dysfunction and deregulated nutrient sensing are all critical components in organismal aging (Lepez-Otin et al., 2013). Current data defined the T cell aging signature as p16high, p21high and p53low. p16INK4A accumulation is characteristic of aging in other lymphoid and nonlymphoid cells (Chen et al., 2011; Cosgrove et al., 2014; Janzen et al., 2006; Liu et al., 2011; Molofsky et al., 2006; Signer et al., 2008; Sousa-Victor et al., 2014). More difficult to understand is the p21-gain and p53-loss combination, as DNA-damage-induced ATM-p53-p21 signaling typically drives growth arrest of senescent T cells (Bartkova et al., 2006; Fumagalli et al., 2014; Nakamura et al., 2008; Rodier et al., 2011). However, in cancer cells p21 expression can occur in the absence of p53 function (Jeong et al., 2010; Macleod et al., 1995). In RA T cells, p53 is consistently low (Maas et al., 2005; Shao et al., 2009), associating aging with a p53-independent, p16-dependent mechanism. Naïve RA CD4+ T cells include a considerable population of CD57+ cells. How CD57 alters the behavior of old versus young CD4+ T cells is unknown, but it could possibly affect tissue homing. Epigenetic changes to chromatin are essential to cellular aging. Senescence-associated distension of satellites (SADS) formation is a unifying process in most models of early senescence, believed to signify the loss of higher-order chromatin packaging (Swanson et al., 2013). The packaging of pericentromeric and centromeric satellite DNA, especially sat-II DNA, in RA T cells was visibly altered. Whether distensions in sat-II DNA are associated with transcription at these loci will need examination. So far, pericentromeres are considered transcriptionally inert chromatin conformations, but recent evidence emphasizes active and regulated transcription, especially in the context of development, tumorigenesis, senescence and cellular stress (Dejardin, 2015; Saksouk et al., 2015). MRE11A inhibition induced pericentromere unraveling, linking satellite packaging with DNA repair and emphasizing the nuclease’s role in protecting the epigenetic landscape. The MRN complex is a multifunctional enzyme conglomerate known to activate cell cycle checkpoints when sensing DNA breakage (Stracker and Petrini, 2011). In addition to a critical role in ATM function and double strand break repair, MRN is also implicated in suppression of lymphomagenesis (Balestrini et al., 2015) and regulating metastatic breast cancers (Gupta et al., 2013). In humans, MRE11A mutations cause an ataxia-telangiectasia-like disease, with considerable variability in clinical phenotypes (Taylor et al., 2015). In primary human T cells, MRE11A appears to have primarily a protective function in telomere maintenance. Reducing MRE11A abundance by 50% rapidly induced the formation of telomeric damage foci, long before such T cells shortened telomeric length. MRE11A deficiency was more effective in inducing telomeric damage foci than the knockdown of RAD50 or NBS1. Vice versa, replenishing MRE11A protein by forced overexpression promptly reversed telomeric damage. These data suggest MRE11A’s participation in telomere repair and protection in human T cells. MRE11A is an abundant protein in human T cells. A 50% reduction, however, had profound consequences, revealed by the reversal of several phenotypes after MRE11A overexpression. Although overexpression prevented TIF formation and suppressed p16, p21 and CD57 expression, the mechanistic connection between telomeric damage and T cell aging awaits further clarification. Pharmacological inhibition of MRE11A’s nucleolytic activity recapitulated all abnormalities of aged T cells, placing insufficient exonuclease activity at the pinnacle of the T cell aging program. Systemic inflammation was insufficient to explain MRE11A downregulation. In PsA, an aggressive multi-organ inflammatory syndrome, MRE11A expression was normal. Untreated RA patients had the lowest MRE11A concentrations; immunosuppressive therapy improved but did not normalize the nuclease’s expression, implicating mechanisms other than systemic inflammation in MRE11A suppression. In previous reports, the failure to degrade cytoplasmic DNA induced inflammatory responses in bone marrow macrophages to eventually induce polyarthritis (Kawane et al., 2006). Current data link DNA damage in T cells, not macrophages, to functional abnormalities driving synovitis. Mechanistically, MRE11A’s nuclease activity appears to be related to in vivo T cell trafficking and T cell control of the synovial microenvironment. Human synovium and human T cells in chimeric mice enabled us to investigate how insufficient MRE11A activity regulates nonlymphoid tissue infiltration and inflammation. RA T cells have been reported to be spontaneously hypermigratory (Hwang et al., 2015) and we found that nuclease inhibition rendered them hyper-motile and tissue-invasive. Once in the tissue, MRE11Alow T cells supported p16 and p21 expression, left p53 unaffected and greatly increased the abundance of classical pro-inflammatory cytokines. MRE11A restoration did not change T cell lineage commitment but appeared to regenerate an anti-inflammatory tissue environment, suggestive for a role of T cells in regulating joint-residing cells, such as synovial fibroblasts. The current study has implication for the understanding of the immune aging process in humans and could change the approach towards treating senescence-associated tissue inflammation. Experimental data focused attention on DNA damage sensing and repair as a key event in immune aging, particularly in RA. Impaired DNA damage responses affected T cells in their naïve state and fundamentally change their differentiation program, biasing them towards tissue-invasive and pro-inflammatory behavior. Inflammation per se, as in PsA patients, was insufficient to repress MRE11A expression and the associated T cell aging profile. MRE11A insufficiency appeared involved in multiple aspects of T cell aging, from unraveling of heterochromatin, to telomeric damage, changed expression of cell cycle regulators and altered tissue homing. Restoring MRE11A expression was sufficient to reverse aging features of RA T cells, identifying DNA damage repair, and particularly MRE11A, as a promising therapeutic target to treat immune-aging related disease. EXPERIMENTAL PROCEDURES Patients and control individuals RA patients recruited into the study fulfilled the 2010 diagnostic criteria for RA and were positive for anti-CCP antibodies and for rheumatoid factor (Table S1). Healthy controls were matched for age and gender. The following were exclusion criteria for their enrollment: personal or family history of autoimmune disease, malignancy, chronic infection or another inflammatory syndrome. The Institutional Review Board approved the study and written informed consent was obtained from all participants. T cell purification and culture Peripheral blood mononuclear cells (PBMCs) were separated from whole blood using Lymphocyte Separation Medium (Corning). Naïve CD4+ T cells were isolated from PBMCs using a human naïve CD4+ T cell isolation kit according to the manufacturer’s instructions (STEMCELL Technologies). This purification procedure minimizes the contamination of CD4+CD45RA+ cells with end-differentiated CD4+CD28− T cells, which accounted for 0.4% of cells in both patients and controls (Figure S2). Naïve CD4+ T cells were cultivated in RPMI 1640 medium (Fisher Scientific) supplemented with 10% FBS (Atlanta Biologicals) and Pen/Strep (Fisher Scientific). Cells were stimulated for 72 hrs using anti-CD3/CD28 beads (Life Technologies) at a ratio of 1 bead per 2 cells. Telomere FISH Telomere FISH was performed using a PNA probe (Panagene, Daejeon, Korea) as described (Batista et al., 2011; Gu et al., 2012). Proliferating T cells were treated overnight with 0.2 µg/ml colcemid (Life Technologies), swollen in KCl buffer (12.3 mM HEPES, 0.53 mM EGTA, 64.4 mM KCl), fixed in methanol/acetic acid (3:1) and dropped onto glass slides. Metaphase spreads were rehydrated in PBS, fixed in 4% formaldehyde and dehydrated in an alcohol series. Slides were incubated with hybridization mixture (70% formamide, 10 mM NaHPO4, pH 7.4, 10 mM NaCl, 20 mM Tris buffer, pH 7.5), placed on a heating block at 80°C for 5 min to denature chromosomal DNA and incubated for 30 min to 2 hrs at room temperature with the PNA probe (0.05 µg/ml). After washing, slides were mounted with ProLong Gold Antifade with DAPI (Life Technologies) and analyzed with a Leica microscope (Leica, Heidelberg, Germany). To visualize MRE11A binding, metaphase spreads were first incubated with anti-MRE11A (Cell Signaling Technology, 4847P; overnight, 4°C) and secondary Alexa Fluor® 488 goat anti-rabbit IgG (H+L) antibody (Life Technologies, A-11008) for 1 hr. After washing, slides were dehydrated in an alcohol series, hybridization mixture was applied and slides were placed on a heat block (80°C) for 5 min followed by 1 hr incubation as above. In situ DNA hybridization for SADS For DNA hybridizations, cells were denatured by incubating at 70°C for 2 min in 70% formamide. Preparations were immediately dehydrated through cold 70%, 95%, and 100% ETOH for 5 min each and then air-dried. Cells were hybridized with directly labeled oligonucleotides and heated at 70°C to 80°C for 10 min. Cells were incubated at 37°C in a humidified chamber overnight (Lawrence et al., 1988). Oligo sequences were sat-II: Cy3 direct label 5′-ATTCCATTCAGATTCCATTCGATC-3′. Confocal microscopy Cells were fixed in 4% paraformaldehyde, permeabilized with 0.05% Triton X-100 and incubated with primary antibodies to 53BP1 (Cell Signaling Technology, 4937S) and TRF2 (Abcam, ab13579) at 4°C for overnight. Incubation with secondary antibodies was performed at room temperature for 1 hr using Alexa Fluor 488 labeled goat anti-rabbit (Life Technologies, A-11008) and Alexa Fluor 546 labeled goat anti-mouse antibodies (Life Technologies, A-21123). The images were analyzed using the LSM710 microscope system with the ZEN 2010 software (Carl Zeiss) and a 63× oil immersion objective (Carl Zeiss). Telomere length measurement Genomic DNA was extracted directly from CD4+ T cells using a Mini Genomic DNA Kit (IBI Scientific) according to manufacturer’s instructions. Telomere length was determined as described (Cawthon, 2002). The primer sequences were as follows: tel 1, 5′-GGTTTTTGAGGGTGAGGGTGAGGGTGAGGGTGAGGGT-3′; tel 2, 5′-TCCCGACTATCCCTATCCCTATCCCTATCCCTATCCCTA-3′; 36B4u, 5′-CAGCAAGTGGGAAGGTGTAATCC-3′; 36B4d, 5′-CCCATTCTATCATCAACGGGTACAA-3′ Synovitis induction in chimeric mice NOD.Cg-PrkdcscidIl2rgtm1Wjl/SzJ (NSG) female mice (10 to 14 weeks old) (The Jackson Laboratory) were used as previously described (Seyler et al., 2005). Pieces of human synovial tissue were placed into a subcutaneous pocket on the upper dorsal midline. In this model, complete engraftment is reached within 1 week. Following engraftment, mice were injected intravenously with 10 million cells of CD45RO− PBMCs. Chimeric mice from the same litter and carrying the same synovial tissue were randomly assigned to one of two treatment arms: (A) vehicle (DMSO) control and (B) treatment with Mirin, 1 mg/kg/day. All treatments were delivered by daily intraperitoneal injection over a period of 9 days. For the MRE11A overexpression experiments, mice from the same litter were randomly assigned to two treatment arms and engrafted with aliquots of the same synovial tissue. CD45RO− PBMCs were prepared from RA patients and transfected with either a control plasmid (10 million cells transferred into group A chimeras) or a MRE11A overexpression plasmid (10 million cells transferred into group B chimeras). Transfected cells were rested for 24 hrs before the adoptive transfer. At the completion of the experiments, mice were sacrificed and synovial tissues were harvested and embedded in OCT (Tissue-Tek; Sakura Finetek) for histological studies or were shock frozen in liquid nitrogen for RNA extraction. Immunohistochemical staining Frozen tissue blocks were cut at 5 µm thickness and picked up on X-tra® Slides (Leica Biosystems). After air-drying for 30 min at room temperature, sections were immediately fixed with acetone at −20°C for 20 min. Slides were rehydrated in 1× PBS for 10 min. To quench endogenous peroxidase activity, sections were incubated with peroxidase reagent (3% H2O2 in 1× PBS) for 15 min and gently washed twice in 1× PBS for 5 min. Slides incubated with a primary antibody cocktail overnight at 2°C to 8°C. The cocktail contained both monoclonal mouse anti-human CD3, (Clone F7.2.38; 1:100, Dako) and anti-p16 ARC antibody (EP1551Y; 1:100, Abcam, ab51243). Upon finishing and rinsing, the secondary antibody cocktail [peroxidase anti-rabbit IgG (H+L) (Vector Laboratories, PI-1000) and alkaline phosphatase anti-mouse IgG (H+L) (Vector Laboratories, AP-2000)] was prepared and applied as recommended by manufacturer (Vector Laboratories) followed by the subsequent visualization of AP activity (Vector Red Alkaline Phosphatase Substrate Kit) (Vector Laboratories, SK-5100) and HRP activity (Vector DAB Peroxidase Substrate Kit) (Vector Laboratories, SK-4100). Stained tissues were mounted with hematoxylin (Sigma-Aldrich), dehydrated, covered and imaged via microscope. Statistical analysis Statistics were calculated using GraphPad Prism software (GraphPad Software). If not stated differently, a two-sided t-test was used to determine significance and p<0.05 was considered significant. p < 0.05, p < 0.01, *p < 0.001. Supplementary Material This work was supported by grants from the National Institutes of Health (R01 AR042547, R01 AI044142, R01 HL117913, R01 AI108906, P01 HL058000, R01 AI108891, R01AR055650, R01AR063717, R01 AG045779 and I01 BX001669) and from the Govenar Discovery Fund. Figure 1 Telomeres in RA T cells are shortened and damaged CD4+CD45RA+ T cells were stimulated for 72 hrs and metaphase nuclei were hybridized with a telomere specific probe (red). DNA is marked with DAPI. (A) Representative microscopy images for one patient with Rheumatoid arthritis (RA) and one healthy control (Con). (B) Fluorescence intensities for 5 patient-control pairs quantified in >30 nuclei for each donor. Fluorescence intensity in arbitrary units (a.u.). Results are mean ± SEM. (C) Distribution of fluorescence intensity (a.u.) strata from 5 RA (red) and 5 control samples (black) quantified in >30 nuclei for each donor. (D) Telomeric ends were analyzed for abnormal structures. Examples of double signal, apposition, fusion, and signal-free ends are shown. (E) Distributions of telomeric phenotypes in 10 patients and 10 controls. 150–200 nuclei were examined in each sample. Percentages of each damage pattern are presented. (F) Naïve and memory CD4+ T cells were separated and placed under proliferative stress by repetitive polyclonal stimulation. Loss of telomeric sequences was measured by PCR and, in parallel, telomeric damage foci were analyzed by dual-color immunostaining with antibodies to the DNA damage protein 53BP1 and the telomeric shelterin TRF2. 53BP1/TRF2 colocalization coefficients and telomeric length shortening in individual samples are correlated. P < 0.05, two-tailed Student’s t-test. See also Figure S1 and Figure S2. Figure 2 The aging profile of CD4+CD45RA+ T cells in RA CD4+CD45RA+ T cells from RA patients and age-matched controls were stimulated for 72 hrs. All data are mean ± SEM. (A, B) CDKN2A, CDKN1A, and TP53 transcript and p16 protein levels were measured in 7 RA-control pairs by RT-PCR and flow cytometry, respectively. (C) Expression of the aging marker CD57 assessed by flow cytometry. (D) SADS foci analyzed by confocal microscopy. Nuclei were hybridized with a sat-II–specific probe (red) and satellite DNA signals were examined in a minimum of 50 nuclei in each sample. Examples of condensed and threadlike distended satellites are shown in inserts. (E) Individual nuclei were scored as SADS-positive, if the pericentromeric or centromeric satellite heterochromatin had lost its tight packaging and was extended longitudinally. Data are from 3 RA-control pairs. (F) Telomeres were uncapped by transfecting healthy CD4+CD45RA+ T cells with TERF2 siRNA oligonucleotides. CDKN2A and CDKN1A transcripts were measured by RT-PCR. Results are from 3 independent experiments. (G) Flow cytometry for p16 expression. Representative histograms from control and TERF2 siRNA transfected cells are shown. Fluorescence Minus One control (FMO) is superimposed as grey area. Mean fluorescence intensities (MFI) of p16 are from 3 independent experiments. (H) CD57 expression was assessed by flow cytometry in 3 independent experiments. Representative data from control and TERF2 siRNA transfected cells are shown.P < 0.05, P < 0.01 and *P < 0.001, two-tailed Student’s t-test. See also Figure S3. Figure 3 MRE11Alow T cells in RA patients MRE11A protein expression was quantified by intracellular staining of PBMCs in a cohort of RA patients (RA) and age-matched controls (Con) with antibodies against MRE11A and lineage markers (CD4, CD45RA). (A, B) Flow cytometric measurement of MRE11A protein expression in relation to donor age for naïve CD4+CD45RA+ T cells and memory CD4+CD45RO+ T cells. (C) Representative histograms from naïve CD4+CD45RA+ T cells from a healthy individual, an untreated RA patient and a RA patient on therapy. (D) Mean ± SEM of MRE11A MFI are from 5 samples per group. (E, F) Localization of MRE11A to the telomere quantified by dual-color immunostaining with anti-MRE11A (green) and anti-TRF2 (red) in resting naïve CD4+CD45RA+ T cells of 5 RA patients and 5 age-matched controls. DNA is marked with DAPI (blue). (E) Representative image of immunostaining for MRE11A. (F) Staining intensities for total nuclear MRE11A and MRE11A-TRF2 colocalization measured in >50 nuclei from each of 5 different donors. P < 0.05, P < 0.01 and P < 0.001, two-tailed Student’s t-test. See also Figure S4. Figure 4 Genetic or pharmacologic inhibition of MRE11A induces T cell aging (A) The impact of MRN insufficiency on telomeric stability was tested by transfecting CD4+CD45RA+ T cells from healthy individuals with control, MRE11A, NBN, or RAD50 siRNA oligonucleotides. 48 hrs later, cells were stained with anti-53BP1 and anti-TRF2. Representative images of 53BP1 (green) and TRF2 (red) after knockdown of individual DNA repair proteins. DNA is marked with DAPI (blue). Merged images show colocalization of 53BP1 and TRF2. (B) Staining intensities (a.u.) for total 53BP1 and 53BP1/TRF2 colocalization were measured in individual nuclei. Mean ± SEM values are indicated. Results are from 5 independent knockdown experiments. (C) T cells were treated with the MRE11A inhibitor Mirin. Damage foci were analyzed by staining for 53BP1 and for 53BP1/TRF2 colocalization. Data are from 5 independent experiments. All data are mean ± SEM. (D, E) CDKN2A and TP53 transcripts were measured by RT-PCR after transfecting CD4+CD45RA+ T cells from healthy individuals with control, MRE11A, NBN, or RAD50 siRNA oligonucleotides for 48 hrs (D) or were treated with the MRE11A inhibitor Mirin (E). (F) Flow cytometry for CD57 expression. Representative data are from control and MRE11A siRNA transfected cells. Percentages of CD57 expressing cells from 3 independent experiments are presented as mean ± SEM. (G) T cells were treated with vehicle or the MRE11A inhibitor Mirin at the indicated doses and analyzed for CD57 expression by flow cytometry. Mean ± SEM from 3 independent experiments. (H, I) Sat-II DNA (red) hybridization in control and Mirin-treated T cells. Distention of satellite DNA in the nuclei of Mirin-treated T cells. Nuclei were scored for SADS as in Figure 2 in >40 nuclei per sample and quantified in 3 control and Mirin-treated samples. P < 0.05, P < 0.01 and P < 0.001, two-tailed Student’s t-test. See also Figure S5. Figure 5 Overexpression of MRE11A repairs telomeric damage and prevents T cell aging in RA Naïve CD4+CD45RA+ T cells from RA patients were transfected with control plasmids or a myc-MRE11A construct. After 48 hrs, transfection efficiency was monitored by qPCR (A) and Western blotting (B). The error bars in (A) represent the 95% confidence internal. (C) Representative images of 53BP1 (green) and TRF2 (red) after overexpression of control plasmids or myc-MRE11A. DNA is marked with DAPI (blue). Merged images show colocalization of 53BP1 and TRF2. (D) Staining intensities (a.u.) for 53BP1/TRF2 colocalization were measured in a minimum of 70 individual nuclei from 3 different patients. Mean ± SEM values are indicated. CDKN2A and TP53 transcript levels (E) and p16 protein levels (F) were measured in control (RA+Vec) and myc-MRE11A (RA+MRE11A) transfected cells from 3 different patients. Results are mean ± SEM. (G) Flow cytometry for p16 expression. Representative data are from cells with or without MRE11A overexpression. (H) Percentages of p16 expressing cells analyzed in 4 independent experiments are presented as mean ± SEM. P < 0.05, P < 0.01 and P < 0.001, two-tailed Student’s t-test. Figure 6 The nuclease MRE11A controls pro-arthritogenic effector functions Synovial inflammation was quantified by gene transcriptional analysis in tissue extracts or by immunohistochemical analysis of tissue sections. (A) TRB transcript levels measured by qPCR to assess T cell accumulation. (B) Representative sections of synovial tissues stained with anti-CD3 antibodies (brown). (C) Percentages of CD3+ T cells in randomly selected fields of synovial tissues sections presented as mean ± SEM. (D) TNFSF11, (E) TNF, IL6, IL1B and (F) TGFB1, IL10 mRNA expression measured by qPCR in tissue extracts. (G) Transcription analysis of the aging markers CDKN2A, CDKN1A, and TP53 by qPCR. (H) p16 (brown) and CD3 (red) were stained by dual-color immunohistochemistry in synovial tissue sections from vehicle and MRE11A inhibitor-treated chimeras. (I) Percentages of p16+CD3+ T cells in randomly selected fields of synovial tissue sections presented as mean ± SEM. (J) T cell migratory capacity was measured in Transwell migration assays in the absence of chemokine gradients. N=6 RA patient-control pairs; n=4 young (<30 yrs) and old (>65 years) healthy individuals; n=4 healthy control T cells with or without Mirin treatment. All data are mean ± SEM.P < 0.05, P < 0.01 and P < 0.001, two-tailed Student’s t-test. Figure 7 Restoring MRE11A expression in RA T cells prevents pro-arthritogenic effector function Pairs of NSG mice were engrafted with human synovial tissue and assigned to two treatment arms. CD45ROneg PBMC (A–D) and CD45RAnegPBMC (E–I) were prepared from RA patients and were transfected with either control plasmid or plasmid expressing MRE11A, and adoptively transferred into the chimeric mice. The intensity of synovial inflammation was compared by tissue gene expression analysis applying qPCR. (A, E and G) The density of the synovial T cell infiltrate was captured by TRB and TNFSF11 transcript levels. (B, C and H) Synovial cytokine production capability was assessed through TNF, IL6, IL1B, TGFB1 and IL10 transcript expression. (D, I) The impact of MRE11A overexpression on the tissue presence of aging markers was examined through CDKN2A and TP53 transcript levels. (F) Representative sections of synovial tissues stained with anti-CD3 antibodies (brown). Original magnification, X600 (insets in F). All data are mean ± SEM from at least 6 different synovial grafts.P < 0.05, P < 0.01 and *P < 0.001, two-tailed Student’s t-test. See also Figure S6. Highlights □ CD4+ T cells in rheumatoid arthritis have damaged telomeres and age prematurely. □ RA patient CD4+ T cells express low concentrations of the DNA repair nuclease MRE11A. □ MRE11A inhibition damages telomeres, unravels heterochromatin and induces T cell aging. □ MRE11Alow T cells are hypermotile, tissue-invasive and arthritogenic in vivo. In Brief Whether T cell aging contributes to tissue inflammation is unclear. In patients with rheumatoid arthritis, Li et al. identify prematurely aged T cells with damaged telomeres resulting from defective activity of the DNA break repair nuclease MRE11A. MRE11Alow T cells differentiate into hypermotile and tissue-invading effector cells that promote destructive synovitis. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. COMPETING FINANCIAL INTERESTS The authors declare that no conflict of interest exists. Additional experimental procedures are presented in Supplemental Materials. AUTHOR CONTRIBUTIONS YL, JJG and CMW designed the research and analyzed data. YL, YS, PH, JJ, ZW and HZ were responsible for the experimental work. SG recruited patients. CMW, YL and JJG wrote the manuscript. REFERENCES Akbar AN Henson SM Are senescence and exhaustion intertwined or unrelated processes that compromise immunity? Nat Rev Immunol 2011 11 289 295 21436838 Attwooll CL Akpinar M Petrini JHJ The Mre11 complex and the response to dysfunctional telomeres Mol Cell Biol 2009 29 5540 5551 19667076 Balestrini A Nicolas L Yang-Lott K Guryanova OA Levine RL Bassing CH Chaudhuri J Petrini JH Defining ATM-independent functions of the Mre11 complex with a novel mouse model Mol Cancer Res 2015 Epub ahead of print Bartkova J Rezaei N Liontos M Karakaidos P Kletsas D Issaeva N Vassiliou LVF Kolettas E Niforou K Zoumpourlis VC Oncogene-induced senescence is part of the tumorigenesis barrier imposed by DNA damage checkpoints Nature 2006 444 633 637 17136093 Batista LF Pech MF Zhong FL Nguyen HN Xie KT Zaug AJ Crary SM Choi J Sebastiano V Cherry A Telomere shortening and loss of self-renewal in dyskeratosis congenita induced pluripotent stem cells Nature 2011 474 399 402 21602826 Cawthon RM Telomere measurement by quantitative PCR Nucleic Acids Res 2002 30 e47 12000852 Chen HN Gu XY Liu YH Wang J Wirt SE Bottino R Schorle H Sage J Kim SK PDGF signalling controls age-dependent proliferation in pancreatic beta-cells Nature 2011 478 349 355 21993628 Chou JP Effros RB T cell replicative senescence in human aging Curr Pharm Design 2013 19 1680 1698 Chow HM Herrup K Genomic integrity and the ageing brain Nat Rev Neurosci 2015 16 672 684 26462757 Cosgrove BD Gilbert PM Porpiglia E Mourkioti F Lee SP Corbel SY Llewellyn ME Delp SL Blau HM Rejuvenation of the muscle stem cell population restores strength to injured aged muscles Nat Med 2014 20 255 264 24531378 Dejardin J Switching between epigenetic states at pericentromeric heterochromatin Trends Genet 2015 31 661 672 26431676 Di Mitri D Azevedo RI Henson SM Libri V Riddell NE Macaulay R Kipling D Soares MV Battistini L Akbar AN Reversible senescence in human CD4+CD45RA+CD27- memory T cells J Immunol 2011 187 2093 2100 21788446 Diotti R Loayza D Shelterin complex and associated factors at human telomeres Nucleus-Austin 2011 2 119 135 Dupre A Boyer-Chatenet L Sattler RM Modi AP Lee JH Nicolette ML Kopelovich L Jasin M Baer R Paull TT A forward chemical genetic screen reveals an inhibitor of the Mre11-Rad50-Nbs1 complex Nat Chem Biol 2008 4 119 125 18176557 Focosi D Bestagno M Burrone O Petrini M CD57(+) T lymphocytes and functional immune deficiency J Leukoc Biol 2010 87 1167 1167 Fuertes Marraco SA Soneson C Cagnon L Gannon PO Allard M Abed Maillard S Montandon N Rufer N Waldvogel S Delorenzi M Long-lasting stem cell-like memory CD8+ T cells with a naive-like profile upon yellow fever vaccination Sci Transl Med 2015 7 282ra248 Fujii H Shao L Colmegna I Goronzy JJ Weyand CM Telomerase insufficiency in rheumatoid arthritis Proc Natl Acad Sci U S A 2009 106 4360 4365 19255426 Fumagalli M Rossiello F Mondello C di Fagagna FD Stable cellular senescence is associated with persistent DDR activation Plos One 2014 9 e110969 25340529 Goronzy JJ Weyand CM Understanding immunosenescence to improve responses to vaccines Nat Immunol 2013 14 428 436 23598398 Gu P Min JN Wang Y Huang C Peng T Chai W Chang S CTC1 deletion results in defective telomere replication, leading to catastrophic telomere loss and stem cell exhaustion EMBO J 2012 31 2309 2321 22531781 Gupta GP Vanness K Barlas A Manova-Todorova KO Wen YH Petrini JH The Mre11 complex suppresses oncogene-driven breast tumorigenesis and metastasis Mol Cell 2013 52 353 365 24120666 Hwang SH Jung SH Lee S Choi S Yoo SA Park JH Hwang D Shim SC Sabbagh L Kim KJ Leukocyte-specific protein 1 regulates T-cell migration in rheumatoid arthritis Proc Natl Acad Sci U S A 2015 112 E6535 E6543 26554018 Janzen V Forkert R Fleming HE Saito Y Waring MT Dombkowski DM Cheng T DePinho RA Sharpless NE Scadden DT Stem-cell ageing modified by the cyclin-dependent kinase inhibitor p16(INK4a) Nature 2006 443 421 426 16957735 Jeong JH Kang SS Park KK Chang HW Magae J Chang YC p53-independent induction of G(1) arrest and p21(WAF1/CIP1) expression by ascofuranone, an isoprenoid antibiotic, through downregulation of c-Myc Mol Cancer Ther 2010 9 2102 2113 20587660 Kawane K Ohtani M Miwa K Kizawa T Kanbara Y Yoshioka Y Yoshikawa H Nagata S Chronic polyarthritis caused by mammalian DNA that escapes from degradation in macrophages Nature 2006 443 998 1002 17066036 Koetz K Bryl E Spickschen K O’Fallon WM Goronzy JJ Weyand CM T cell homeostasis in patients with rheumatoid arthritis Proc Natl Acad Sci U S A 2000 97 9203 9208 10922071 Lamarche BJ Orazio NI Weitzman MD The MRN complex in double-strand break repair and telomere maintenance FEBS Lett 2010 584 3682 3695 20655309 Lawrence JB Villnave CA Singer RH Sensitive, high-resolution chromatin and chromosome mapping in situ: presence and orientation of two closely integrated copies of EBV in a lymphoma line Cell 1988 52 51 61 2830981 Lepez-Otin C Blasco MA Partridge L Serrano M Kroemer G The hallmarks of aging Cell 2013 153 1194 1217 23746838 Liu Y Johnson SM Fedoriw Y Rogers AB Yuan H Krishnamurthy J Sharpless NE Expression of p16(INK4a) prevents cancer and promotes aging in lymphocytes Blood 2011 117 3257 3267 21245485 Maas K Westfall M Pietenpol J Olsen NJ Aune T Reduced p53 in peripheral blood mononuclear cells from patients with rheumatoid arthritis is associated with loss of radiation-induced apoptosis Arthritis Rheum 2005 52 1047 1057 15818671 Macleod KF Sherry N Hannon G Beach D Tokino T Kinzler K Vogelstein B Jacks T P53-dependent and independent expression of P21 during cell-growth, differentiation, and DNA-damage Genes Dev 1995 9 935 944 7774811 Molofsky AV Slutsky SG Joseph NM He SH Pardal R Krishnamurthy J Sharpless NE Morrison SJ Increasing p16(INK4a) expression decreases forebrain progenitors and neurogenesis during ageing Nature 2006 443 448 452 16957738 Nakamura AJ Chiang YJ Hathcock KS Horikawa I Sedelnikova OA Hodes RJ Bonner WM Both telomeric and non-telomeric DNA damage are determinants of mammalian cellular senescence Epigenet Chromatin 2008 1 6 Pulko V Davies JS Martinez C Lanteri MC Busch MP Diamond MS Knox K Bush EC Sims PA Sinari S Human memory T cells with a naive phenotype accumulate with aging and respond to persistent viruses Nat Immunol 2016 17 966 975 27270402 Rodier F Munoz DP Teachenor R Chu V Le O Bhaumik D Coppe JP Campeau E Beausejour CM Kim SH DNA-SCARS: distinct nuclear structures that sustain damage-induced senescence growth arrest and inflammatory cytokine secretion J Cell Sci 2011 124 68 81 21118958 Saksouk N Simboeck E Dejardin J Constitutive heterochromatin formation and transcription in mammals Epigenet Chromatin 2015 8 3 Schmidt D Goronzy JJ Weyand CM CD4(+) CD7(−) CD28(−) T cells are expanded in rheumatoid arthritis and are characterized by autoreactivity J Clin Invest 1996 97 2027 2037 8621791 Schonland SO Zimmer JK Lopez-Benitez CM Widmann T Ramin KD Goronzy JJ Weyand CM Homeostatic control of T-cell generation in neonates Blood 2003 102 1428 1434 12714521 Seyler TM Park YW Takemura S Bram RJ Kurtin PJ Goronzy JJ Weyand CM BLyS and APRIL in rheumatoid arthritis J Clin Invest 2005 115 3083 3092 16239971 Shao L Fujii H Colmegna I Oishi H Goronzy JJ Weyand CM Deficiency of the DNA repair enzyme ATM in rheumatoid arthritis J Exp Med 2009 206 1435 1449 19451263 Shao L Goronzy JJ Weyand CM DNA-dependent protein kinase catalytic subunit mediates T-cell loss in rheumatoid arthritis EMBO Mol Med 2010 2 415 427 20878914 Sharpless NE Sherr CJ Forging a signature of in vivo senescence Nat Rev Cancer 2015 15 397 408 26105537 Shibata A Moiani D Arvai AS Perry J Harding SM Genois MM Maity R van Rossum-Fikkert S Kertokalio A Romoli F DNA double-strand break repair pathway choice is directed by distinct MRE11 nuclease activities Mol Cell 2014 53 7 18 24316220 Signer RAJ Montecino-Rodriguez E Witte ON Dorshkind K Aging and cancer resistance in lymphoid progenitors are linked processes conferred by p16(Ink4a) and Arf Genes Dev 2008 22 3115 3120 19056891 Sousa-Victor P Gutarra S Garcia-Prat L Rodriguez-Ubreva J Ortet L Ruiz-Bonilla V Jardi M Ballestar E Gonzalez S Serrano AL Geriatric muscle stem cells switch reversible quiescence into senescence Nature 2014 506 316 321 24522534 Stracker TH Petrini JH The MRE11 complex: starting from the ends Nat Rev Mol Cell Biol 2011 12 90 103 21252998 Swanson EC Manning B Zhang H Lawrence JB Higher-order unfolding of satellite heterochromatin is a consistent and early event in cell senescence J Cell Biol 2013 203 929 942 24344186 Takata H Tanaka Y Matsuura A Late S phase-specific recruitment of Mrel11 complex triggers hierarchical assembly of telomere replication proteins in Saccharomyces cerevisiae Mol Cell 2005 17 573 583 15721260 Taylor AM Lam Z Last JI Byrd PJ Ataxia telangiectasia: more variation at clinical and cellular levels Clin Genet 2015 87 199 208 25040471 Warrington KJ Takemura S Goronzy JJ Weyand CM CD4+,CD28-T cells in rheumatoid arthritis patients combine features of the innate and adaptive immune systems Arthritis Rheum 2001 44 13 20 11212151 Weyand CM Fujii H Shao L Goronzy JJ Rejuvenating the immune system in rheumatoid arthritis Nat Rev Rheumatol 2009 5 583 588 19798035 Weyand CM Yang Z Goronzy JJ T-cell aging in rheumatoid arthritis Curr Opin Rheumatol 2014 26 93 100 24296720 Xie AY Kwok A Scully R Role of mammalian Mre11 in classical and alternative nonhomologous end joining Nat Struct Mol Biol 2009 16 U814 U831 Yang Z Fujii H Mohan SV Goronzy JJ Weyand CM Phosphofructokinase deficiency impairs ATP generation, autophagy, and redox balance in rheumatoid arthritis T cells J Exp Med 2013 210 2119 2134 24043759 Yang Z Shen Y Oishi H Matteson EL Tian L Goronzy JJ Weyand CM Restoring oxidant signaling suppresses proarthritogenic T cell effector functions in rheumatoid arthritis Sci Transl Med 2016 8 331ra338
PMC005xxxxxx/PMC5123791.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101257113 34616 Curr Protoc Microbiol Curr Protoc Microbiol Current protocols in microbiology 1934-8525 1934-8533 26528782 5123791 10.1002/9780471729259.mc14a04s39 NIHMS821995 Article UNIT 14A.4 Generation of Recombinant Vaccinia Viruses Earl Patricia L. Moss Bernard Wyatt Linda S. National Institute of Allergy and Infectious Diseases Bethesda, Maryland 13 10 2016 03 11 2015 03 11 2015 25 11 2016 39 14A.4.114A.418 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. This unit describes how to infect cells with vaccinia virus and then transfect them with a plasmid-transfer vector or PCR fragment to generate a recombinant virus. Selection and screening methods used to isolate recombinant viruses and a method for the amplification of recombinant viruses are described. Finally, a method for live immunostaining that has been used primarily for detection of recombinant modified vaccinia virus Ankara (MVA) is presented. This unit first describes how to infect cells with vaccinia virus and then transfect them with a plasmid-transfer vector or PCR fragment to generate a recombinant virus (see Basic Protocol 1). Also presented are selection and screening methods used to isolate recombinant viruses (see Basic Protocol 2) and a method for the amplification of recombinant viruses (see Basic Protocol 3). Finally, a method for live immunostaining that has been used primarily for detection of recombinant modified vaccinia virus Ankara (MVA) is presented (see Basic Protocol 4). HeLa S3 cells are recommended for large-scale growth of vaccinia virus. BS-C-1 cells may be used for xanthine-guanine phosphoribosyltransferase (XGPRT) and plaque size selection, fluorescent protein screening, transfection and determination of virus titer (UNIT 14A.3). For thymidine kinase (TK) selection, HuTK− 143B cells are used. With MVA, all steps are carried out in CEF or BHK-21 cells (UNIT 14A.3). CAUTION Proceed carefully and follow biosafety level 2 (BL-2) practices when working with standard vaccinia virus (see UNIT 14A.3 for safety precautions). [*Copy Editor: The original CPMB unit referenced CPMB Unit 16.15 for safety. The chapter editor asked that the authors include some of the safety information in the revised units – CPMB 16.16 and 16.17 – which are now CPMC Unit 14A.3 and 14A.4. As a result, the authors changed the safety citation here to “Unit 14A.3”, which doesn’t have nearly as much information as the original CPMB Unit 16.15. Perhaps the reason is that some of those early safety concerns are no longer relevant, but it would be good if you could double check with the authors and ask why they elected not to add more safety information. If guidelines have been relaxed in the pox virus field, that would be good to say explicitly.] NOTE: Carry out all procedures for preparation of virus in a biosafety cabinet. BASIC PROTOCOL 1: GENERATION OF A VACCINIA VIRUS VECTOR BY HOMOLOGOUS RECOMBINATION The foreign gene of interest is subcloned into a plasmid transfer vector (Fig. 14A.4.1) or assembled as a PCR fragment so that it is flanked by DNA from the vaccinia virus genome for insertion into a non-essential site. This recombinant plasmid is then transfected into BS-C-1 or other cells that have been infected with vaccinia virus. Homologous recombination occurs between the vaccinia virus genome and homologous sequences (Fig. 14A.4.1). The recombinant virus is obtained in a cell lysate, which is then subjected to several rounds of plaque purification using appropriate selection and/or screening protocols (see Basic Protocol 2). For MVA, the same transfection procedure is used and either CEF or BHK-21 (UNIT 14A.3) cells are substituted for BS-C-1 cells. For live immunostaining to detect recombinant MVA or standard virus, see Basic Protocol 4. Materials pRB21, pSC11, pSC65, pLW-44, pLAS-1, pLW-73, or other suitable vector (Table 16.17.1) BS-C-1, BHK-21, or CEF cells (UNIT 14A.3) Complete MEM-8 and -2.5 media (UNIT 14A.3) Vaccinia virus stock (UNIT 14A.3) Liposomal transfection reagent (such as Lipofectamine-2000, Life Technologies) Transfection buffer (see Reagents and Solutions) 2.5 M CaCl2 OptiMEM medium (Life Technologies) Dry ice/ethanol bath 12-well tissue culture plates 12 × 75–mm polystyrene tubes Disposable scraper, plunger from a 1 ml syringe, or rubber policeman, sterile 2-ml sterile microcentrifuge tubes (optional – use Sarstedt 2 ml screw-cap tubes) Additional reagents and equipment for subcloning, isolation of plasmid, PCR amplification, and tissue culture NOTE: All reagents and equipment coming into contact with live cells must be sterile, and aseptic technique should be used accordingly. Culture incubations should be performed in a humidified 37°C, 5% CO2 incubator unless otherwise specified. Prepare recombinant plasmid DNA 1a Subclone the gene of interest into the multiple cloning site (MCS) in pRB21, pSC11, pSC65, pLW-44, pLAS-1, pLW-73, or other suitable vector and isolate plasmid using standard procedures. Proceed to step 6. Prepare PCR fragment 1b Amplify the 2 flanking regions of the vaccinia virus genome between which the foreign gene will be inserted. 2b Amplify the foreign gene of interest and relevant vaccinia virus promoter. 3b Join the DNA containing the flanking regions and foreign gene by overlap PCR. 4b Verify the size integrity of the PCR fragment by gel electrophoresis. 5b Gel purify the DNA fragment. Proceed to step 6. Prepare vaccinia virus-infected cells 6 Seed wells of a 12-well tissue culture plate with 2.5 × 105 BS-C-1, BHK-21, or CEF cells in complete MEM-8 medium. Incubate to near confluency (usually overnight). UNIT 14A.3 details the culture of these cells. 7 Thaw an aliquot of vaccinia virus and sonicate in ice-water. Cool on ice and repeat sonication. (see UNIT 14A.3). Virus stocks are usually at a titer of ~1 × 109 pfu/ml, but may be significantly lower depending on the source. Vortexing usually breaks up any clumps of cells. However, if there are still visible clumps, chill to 0°C and sonicate 30 sec on ice. Sonication can be repeated several times but the sample should be allowed to cool on ice between sonications. 8 Dilute sonicated virus in complete MEM-2.5 to 0.5 × 105 pfu/ml. Aspirate medium from confluent monolayer of cells and infect with 0.5 ml diluted vaccinia virus (0.05 pfu/cell). Incubate 2 hr at 37°C. Approximately 30 min before the end of the infection period, prepare DNA according to the liposomal reagent or CaCl2 method as described below. Prepare DNA For liposomal reagent method (Lipofectamine) 9a Place 0.25 ml OptiMEM into each of 2 polystyrene tubes. To one tube add 5 μl of Lipofectamine-2000. To the other tube add 0.5 to 1 μg of the recombinant plasmid (from step 1a) or PCR fragment (from step 5b) (in <25 μl). Leave 5 min at room temperature. 10a Add the DNA solution to the lipofectamine solution and leave another 30 min at room temperature. For CaCl2 method 9b Place 0.5 ml transfection buffer into a 12 × 75–mm polystyrene tube and add 0.5 to 1 μg of the recombinant plasmid (from step 1a) or PCR fragment (from step 5b) (in <25 μl). A discussion of calcium phosphate transfection can be found in Critical Parameters. 10b Slowly add 13 μl of 2.5 M CaCl2 and mix gently. Leave 20 to 30 min at room temperature. Gentle mixing is essential; see Critical Parameters. A fine precipitate should appear. Transfect cells 11 Aspirate virus inoculum from monolayer of cells (step 8) and wash twice with PBS. 12a. For lipofectamine method: Add the DNA/lipofectamine solution (step 10a) to the cells and incubate 4 hr at 37°C. 12b For CaCl2 method: Add the precipitated DNA suspension (step 10b) to the cells and leave 30 min at room temperature, then add 1.5 ml complete MEM-8 and incubate 3 to 4 hr at 37°C. 13 Aspirate medium, replace with 1 ml MEM-2.5, and incubate 2 days at 37°C. 14 Dislodge cells with a disposable scraper, plunger from a 1ml syringe or sterile rubber policeman and transfer to a 2-ml sterile microcentrifuge tube. 15 Lyse the cell suspension by performing three freeze-thaw cycles, each time by freezing in a dry ice/ethanol bath, thawing in a 37°C water bath, and vortexing. 16 Store the cell lysate at −80°C until needed in the selection and screening procedure (see Basic Protocol 2). BASIC PROTOCOL 2: SELECTION AND SCREENING OF RECOMBINANT VIRUS PLAQUES For standard vaccinia strains, procedures are described involving xanthine-guanine phosphoribosyltransferase (XGPRT; Falkner and Moss, 1988) or thymidine kinase (TK; Mackett et al., 1984) for selecting virus plaques that contain recombinant DNA. A newer and simpler drug-free method (plaque selection) is based on repair of a mutation that caused the parental virus to form pin-point plaques (Blasco and Moss, 1995; see Background Information regarding choice of procedure). In addition, green fluorescent protein (GFP), β-galactosidase (Chakrabarti et al., 1985), or β-glucuronidase (GUS) screening (Carroll and Moss, 1995) can be used alone or in conjunction with TK selection to discriminate TK− recombinants from spontaneous TK− mutants. For each method, recombinant virus (obtained in the transfection; see Basic Protocol 1) is used to infect a monolayer culture of cells. For vaccinia virus, BS-C-1 cells are used because large plaques are obtained; for TK selection, it is necessary to use a cell line such as HuTK− 143B that is deficient in thymidine kinase. For MVA, CEF or BHK-21 cells are necessary. Medium containing 2.5% methylcellulose and the appropriate selective drugs, if applicable, is then pipetted onto the infected cell monolayer. Because of cell-to-cell spread of virus, each productively infected cell gives rise to an infected area on the monolayer. With vaccinia infection the cells are rounded and dead, and thus appear as colorless plaques. With MVA there is a buildup of cells resulting in appearance of a focus. If GFP (or other fluorescent protein) is used, fluorescent (recombinant) plaques are visualized with a fluorescence microscope. If β-galactosidase (or GUS) screening is used, the substrate Xgal (or Xgluc) is included in the overlay. Plaques containing infected cells that have expressed β-galactosidase or GUS turn blue; thus, blue (recombinant) plaques can be distinguished from clear (parental) plaques. Live immunostaining can also be employed (see Basic Protocol 4). A sterile toothpick or pipet tip is used to remove infected cells from the plaques/foci. The virus is placed in medium and is released by freeze-thaw cycling and sonication. Several rounds of plaque purification are used to ensure the absence of residual non-recombinant virus. TK selection has not been used with MVA. Materials BS-C-1, HuTK− 143B, BHK-21, or CEF confluent monolayer cells (UNIT 14A.3) and appropriate complete medium Complete MEM-2.5 medium (UNIT 14A.3) Selective agents (for XGPRT selection; filter sterilize, and store at –20°C): 10 mg/ml (400×) mycophenolic acid (MPA; Calbiochem) in 0.1 N NaOH 10 mg/ml (40×) xanthine in 0.1 N NaOH 10 mg/ml (670×) hypoxanthine in 0.1 N NaOH Transfected cell lysate (see Basic Protocol 1) 5 mg/ml 5-bromodeoxyuridine (BrdU) in H2O (for TK selection; filter sterilize and store at −20°C) 4% Xgal in dimethylformamide (optional, for β-galactosidase screening 2% Xgluc in dimethylformamide (optional, for GUS screening) Dry ice/ethanol bath 6-well, 35-mm tissue culture plates Cup sonicator (e.g., Ultrasonic Processor VCX-750 from Sonics and Materials) Fluorescence microscope (for GFP or other fluorescent protein screening) Pipet tips or toothpicks, sterile Sterile microcentrifuge tubes (optional – use Sarstedt 2 ml screw-cap tubes) Additional reagents and equipment for tissue culture and counting cells and serial dilution of virus. NOTE: All reagents and equipment coming into contact with live cells must be sterile, and aseptic technique should be used accordingly. Culture incubations should be performed in a humidified 37°C, 5% CO2 incubator unless otherwise specified. Prepare the cells 1 Trypsinize confluent monolayer culture and resuspend in appropriate complete medium as in UNIT 14A.3, steps 4 to 7 of Basic Protocol 1. For XGPRT selection, plaque selection, color or GFP screening, use BS-C-1 cells. For TK selection, use HuTK− 143B cells. For MVA, use BHK-21 or CEF cells. 2 Count cells using a cell counter or hemacytometer (APPENDIX 4A). The authors use a Nexcelom Cellometer cell counter 3 Plate 5 × 105 cells/well in a 6-well tissue culture plate (2 ml/well final). Incubate until confluent (this should take <24 hr). 4 Prepare cells as follows. For XGPRT selection, preincubate monolayer for 12 to 24 hr in filter-sterilized complete MEM-2.5 containing 1/400 vol 10 mg/ml MPA, 1/40 vol 10 mg/ml xanthine, and 1/670 vol 10 mg/ml hypoxanthine. For plaque or TK selection or color screening methods, do not preincubate. Prepare the lysate and infect cells 5 Thaw the transfected cell lysate and sonicate 20 to 30 sec in ice-water to break up clumps. Repeat, keeping lysate on ice between steps. 6 Add 100, 10, 1, or 0.1 μl of lysate to duplicate wells containing 1 ml complete MEM-2.5. Gently swirl to mix. Incubate 2 hr Perform methylcellulose overlay 7 Prepare selective medium containing 2.5% methylcellulose For XGPRT selection, include MPA, xanthine, and hypoxanthine For TK selection, include 1/100 vol of 5 mg/ml BrdU. For β-galactosidase or GUS screening, add 1/240 vol 4% Xgal or 1/200 vol 2% Xgluc, respectively. For plaque selection or fluorescent protein screening, make no additions. 8 Aspirate the virus inoculum from the infected cells (from step 6). Overlay each well with 2 ml appropriate selective medium containing 2.5% methylcellulose. Incubate 2 days. Obtain the plaques 9 Add 0.5 ml complete MEM-2.5 to sterile microcentrifuge tubes. When incubation period is complete (step 8), pick well-separated plaques by scraping and suction with a pipet tip or scraping with a toothpick. Transfer to a tube containing 0.5 ml complete MEM-2.5. Repeat for six to twelve plaques with separate pipet tips or toothpicks, placing each in a separate tube. 10 Vortex each virus-containing tube, then perform three freeze-thaw cycles, each time by freezing in a dry ice/ethanol bath, thawing in a 37°C water bath, and vortexing. 11 Place tubes containing virus into a cup sonicator containing ice-water and sonicate 20 to 30 sec at full power. If TK selection only has been used, plaque isolates should be tested by PCR, DNA dot-blot hybridization, or immunostaining because some plaques will contain spontaneous TK− mutations and not recombinant virus. Carry out several rounds of plaque purification 12 Prepare monolayers of the appropriate cell line as described in steps 1 to 4. One 6-well plate is needed for each plaque isolate. 13 Add 100, 10, 1, or 0.1 μl of lysate to duplicate wells containing 1 ml complete MEM-2.5. Gently swirl to mix. Incubate 2 hr. This should be performed with several plaque isolates. If XGPRT selection is used, cells must be preincubated with selective drugs and serial dilutions of the viral isolates must also contain selective drugs (step 4, substep a). Note that the parental non-recombinant virus will form tiny plaques in the presence of drug. 14 Aspirate medium from cell monolayers and overlay with selective medium (if appropriate) containing 2.5% methylcellulose. 15 Repeat steps 5 to 11 for three or more rounds of plaque purification to ensure a clonally pure recombinant virus. BASIC PROTOCOL 3: AMPLIFICATION OF VIRUS FROM A PLAQUE Recombinant virus from a plaque (obtained after the selection and screening protocols; see Basic Protocol 2) is amplified by infection of successively larger numbers of cells. Medium containing drugs for XGPRT or TK selection is usually used, up to and including the infection of cells in a 25-cm2 flask. Freeze-thaw cycling is carried out to release the recombinant virus from the cells. The titer of the vaccinia virus or MVA stock is determined as described in UNIT 14A.3. Materials Resuspended recombinant plaque (see Basic Protocol 2) Confluent monolayer cultures of appropriate cells in both a 6-well and a 25-cm2 tissue culture flask (UNIT 14A.3) Complete MEM-2.5 and -8 media (UNIT 14A.3) Selective agents (for XGPRT selection; filter sterilize, and store at –20°C): 10 mg/ml (400×) mycophenolic acid (MPA; Calbiochem) in 0.1 N NaOH 10 mg/ml (40×) xanthine in 0.1 N NaOH 10 mg/ml (670×) hypoxanthine in 0.1 N NaOH 5 mg/ml 5-bromodeoxyuridine (BrdU) in H2O (for TK selection; filter sterilize and store at −20°C) Dry ice/ethanol bath Spinner culture of HeLa S3 cells (UNIT 14A.3) Cup sonicator (e.g., Ultrasonic Processor VCX-750 from Sonics and Materials) 15-ml conical centrifuge tubes Sorvall centrifuge with H-6000A rotor (or equivalent) 162-cm2 tissue culture flask Additional reagents and equipment for tissue culture and counting cells (APPENDIX 4A) NOTE: All reagents and equipment coming into contact with live cells must be sterile, and aseptic technique should be used accordingly. Incubations should be performed in a humidified 37°C, 5% CO2 incubator unless otherwise specified. Infect a monolayer culture of cells with virus from a plaque 1 Place tube containing recombinant virus from a resuspended plaque into a cup sonicator containing ice-water and sonicate 20 to 30 sec at full power. Cool on ice and repeat. 2 Dilute 0.25 ml of lysate from step 1 with 0.75 ml complete MEM-2.5 containing selective agents, if appropriate (see step 3). Infect appropriate confluent monolayer culture in 6-well plate and incubate 30 min. If XGPRT selection is used, the monolayer culture is preincubated for 12 to 24 hr in complete MEM-2.5 containing MPA, xanthine, and hypoxanthine (see Basic Protocol 2, step 4, substep a). Infection should also be done in the presence of these drugs. 3 Overlay with 1 ml complete MEM-2.5 medium containing the appropriate selective agents. For XGPRT selection, include 1/400 vol 10 mg/ml MPA, 1/40 vol 10 mg/ml xanthine, and 1/670 vol 10 mg/ml hypoxanthine. For TK selection, include 1/200 vol of 5 mg/ml BrdU. Incubate 2 days or until cytopathic effect (cell rounding) is obvious. 4 Remove and discard 1 ml of the medium covering the cell monolayer. Scrape cells, transfer to microcentrifuge tube. 5 Perform three freeze-thaw cycles, each time by freezing in a dry ice/ethanol bath, thawing in a 37°C water bath, and vortexing. 6 Place tube containing cell suspension into a cup sonicator and sonicate 20 to 30 sec at full power. Scale up the culture 7 Dilute 0.5 ml of lysate from step 6 with 1.5 ml complete MEM-2.5 containing selective agents (see steps 2 and 3). Infect appropriate confluent monolayer culture in a 25-cm2 flask and incubate 30 min. 8 Overlay with 3 ml complete MEM-2.5 containing the appropriate selective agents (step 3). Incubate 2 days or until cytopathic effect is obvious. 9 Scrape cells, transfer to 15-ml conical centrifuge tube, and centrifuge 5 min at 1800 × g (2500 rpm in H-6000A rotor). Resuspend cells in 1 ml complete MEM-2.5 and repeat freeze-thaw cycling and sonication as described in steps 5 and 6. 10 Dilute 0.5 ml of lysate from step 9 with 4.5 ml complete MEM-2.5. Infect appropriate confluent monolayer culture in a 162-cm2 flask and incubate 30 min. 11 Overlay with 25 ml complete MEM-2.5 (selection is not required at this step) and incubate 2–3 days or until cytopathic effect is obvious. 12 Detach cells from the flask by shaking or scraping if necessary. Transfer to centrifuge tube by pipetting, then centrifuge 5 min at 1800 × g. Aspirate and discard supernatant. 13 Resuspend cells in 2 ml complete MEM-2.5 and carry out freeze-thaw cycling three times as described in step 5. 14 Determine titer of the viral stock as described in UNIT 14A.3. Freeze viral stock at −80°C. Test for sterility 15 Place 10 μl of lysate from step 13 into 2 ml bacterial growth medium (such as LB) in a 15 ml tube. Shake at 37°C for 1–2 days. Examine the turbidity of the liquid to ensure the absence of bacterial growth. BASIC PROTOCOL 4: LIVE IMMUNOSTAINING AND AMPLIFICATION OF MVA RECOMBINANTS Live immunostaining was developed for modified vaccinia virus Ankara (MVA) because this strain does not form discrete, easily recognizable plaques and because this technique helps to avoid the need for incorporation of selectable or screening marker genes. Immunostaining can be used for recombinant proteins that are expressed on the cell surface or in the cytoplasm. These protocols are also applicable to standard strains of vaccinia virus. The strong adherence of chicken embryo fibroblasts (CEF) to concavalin A–coated plastic dishes make them superior to BHK-21 or other cell lines that the authors have tried. Materials 162-cm2 flask of confluent CEF (UNIT 14A.3) Complete MEM-2.5, and -8 media (UNIT 14A.3) Transfected cell lysate (see Basic Protocol 1) Dry ice/ethanol bath Primary antibody to protein product of foreign gene Horseradish peroxidase–conjugated secondary antibody (to species of primary antibody) Concanavalin A–coated 6-well tissue culture plates (see Support Protocol 3) Cup sonicator (e.g., Ultrasonic Processor VCX-750 from Sonics and Materials) Inverted microscope Sterile toothpicks Cell scraper or plunger of 1-ml syringe 75- and 162-cm2 tissue culture flasks Additional reagents and equipment for culture, trypsinization, and immunostaining of CEF cells, titering of MVA, and preparation of MVA stocks (UNIT 14A.3) NOTE: All reagents and equipment coming into contact with live cells must be sterile, and aseptic technique should be used accordingly. Incubations should be performed in a humidified 37°C, 5% CO2 incubator unless otherwise specified. Prepare, infect, and immunostain CEF cells 1 Trypsinize confluent CEF monolayer cultures, seed cells in concanavalin A–coated 6-well tissue culture plates, and incubate until nearly confluent (UNIT 14A.3). 2 Thaw transfected cell lysate (Basic Protocol 1) and sonicate in a cup sonicator in ice-water for 30 sec at full power. Add 100, 10, 1, or 0.1 μl of lysate to duplicate wells of CEF containing 1 ml of complete MEM-2.5 medium. Gently swirl to mix. Incubate 2 hr. 3 Aspirate virus inoculum and overlay with complete MEM-2.5 containing 2.5% methylcellulose. Incubate 3 days. 4 Immunostain the unfixed cells using antibody to the protein product of the foreign gene (see UNIT 14A.3, Support Protocol 3, steps 7 to 13). If the protein of interest is expressed intracellularly, remove the fluid from the wells and place plate at −80°C for 1 hr. Allow to thaw and begin staining as described in UNIT 16.16. This procedure ruptures the cells in situ and allows the antibody to penetrate. Isolate recombinant virus 5 Examine the cell monolayer using an inverted microscope. Touch immunostaining foci with sterile toothpicks. Place toothpicks individually in small vials containing 0.5 ml of complete MEM-2.5, and break off each toothpick so that the sterile part is inside the tube. 6 Perform three freeze/thaw cycles on each tube, each time by freezing in a dry ice/ethanol bath, thawing in a 37°C water bath, and vortexing. Place tube into a cup sonicator containing ice-water and sonicate 30 sec at full power. 7 Replate onto new CEF monolayers and plaque purify the MVA recombinant a second time as in steps 2 to 6. Plaque purify an additional three times. In this protocol, plaque-purification is done under a liquid overlay containing methlycellulose. To check stability and purity, an extra plate can be held for 3 days and then fixed and immunostained as in UNIT 14A.3, Support Protocol 3. The presence of nonstaining foci, which may be detected at this time because of cytopathic effects, are due to wild-type virus or an unstable recombinant. Amplify plaque-purified virus 8 Infect 1 well of a 6-well plate of CEF or BHK-21 cells with 0.25 ml of plaque-purified MVA recombinant in a total volume of 2 ml MEM-2.5. Incubate 3 days. 9 Remove and discard 1 ml of medium covering the cell monolayer. Dislodge cells into the 1 ml of remaining medium with a cell scraper (or plunger of a 1 ml syringe) and transfer to a vial. Carry out three freeze/thaw cycles as in step 6 to produce a lysate. 10 Amplify the virus by inoculating 0.5 ml of lysate into 75-cm2 flask of CEF or BHK-21 cells containing 15 ml MEM-2.5. After 3 days harvest and lyse cells as in step 9. 11 Inoculate a 162-cm2 flask of CEF or BHK-21 cells containing 30 ml MEM-2.5 with 0.5 ml of lysate from step 10. 12 Determine titer and prepare large stock of recombinant MVA (UNIT 14A.3, Support Protocol 3 and Basic Protocol 5). Freeze viral stock in aliquots at −80°C. SUPPORT PROTOCOL 1: PCR AMPLIFICATION and DNA SEQUENCING OF INSERTED GENE This protocol describes PCR amplification of the DNA fragment inserted into the recombinant virus. The size of the DNA fragment is confirmed by gel electrophoresis and the integrity of the DNA sequence is then verified by sequencing. Materials Titered lysate of recombinant virus (see Basic Protocol 3) 6-well, 35-mm2 tissue culture plate with appropriate confluent monolayer cells (UNIT 14A.3) 2 ml screw-cap tubes Qiagen QIAamp DNA blood mini Kit (catalog # 51104) Appropriate oligonucleotides at the ends of the inserted DNA and as necessary within the inserted gene to allow sequencing of the entire insert Agarose gel electrophoresis apparatus and gel and means of visualizing DNA PCR thermal cycler Infect a monolayer of cells Infect a monolayer of cells in one well of a 6-well, 35-mm2 tissue culture plate with a multiplicity of infection of 5 plaque forming units of the recombinant virus. Incubate overnight at 37°C. Scrape cells with the plunger of a 1 ml syringe and transfer to a 2 ml screw-cap tube. Follow manufacturer’s instructions (Qiagen) for preparation of DNA. Amplify the DNA insert using appropriate oligonucleotides and run on agarose gel to verify the size of the insert Verify nucleotide sequence of the DNA fragment using standard methods. SUPPORT PROTOCOL 3: COATING PLATES WITH CONCANAVALIN A This protocol describes the preparation of concanavalin A–coated plates to be used growing CEF monolayers for live immunostaining as in Basic Protocol 4. Materials Concanavalin A (Sigma) Phosphate-buffered saline (PBS; APPENDIX 2A) 6-well, 35-mm tissue culture dishes Plastic bags for storage NOTE: All reagents and equipment coming into contact with live cells must be sterile, and aseptic technique should be used accordingly. Add 12 mg concanavalin A to 120 ml sterile PBS for a final concentration of 100 μg/ml. Add 1 ml of the PBS/concanavalin A solution to each well of twenty 6-well plates. Incubate 1 hr at room temperature. Aspirate the liquid from each well and rinse with 2 ml PBS. Remove the fluid from each well and let plates dry by storing open in a biological safety hood (to keep sterile). Store plates in a plastic bag (to preserve sterility) at room temperature (will keep for months). REAGENTS AND SOLUTIONS Use deionized, distilled water in all recipes and protocol steps. For common stock solutions, see APPENDIX 2A. Transfection buffer 0.14 M NaCl 5 mM KCl 1 mM Na2HPO4·2H2O 0.1% (w/v) dextrose 20 mM HEPES Adjust to pH 7.05 with 0.5 M NaOH and filter sterilize Store indefinitely at −20°C The pH of this buffer is critical and should be between 7.0 and 7.1. COMMENTARY Background Information Vaccinia virus has numerous advantages as an expression vector. The virus is easily propagated and has a wide host range in cultured cells and experimental animals; the large genome has the capacity to incorporate at least 25,000 bp (Smith and Moss, 1983); transcription occurs in the cytoplasm bypassing steps in mRNA processing and nuclear-cytoplasmic transport; high levels of expression can be achieved; and efficient recombination facilitates insertion of genes. The basic strategy for construction of recombinant vaccinia viruses using plasmid transfer vectors was described more than 30 years ago (Mackett et al., 1984) and is still the most popular method. An alternative strategy is to prepare the DNA by polymerase chain reaction without cloning in bacteria. Still another option involves recombineering artificial chromosomes (Domi and Moss, 2005). Although the Western Reserve (WR) strain of vaccinia virus has been used most extensively for construction of recombinant viruses, more attenuated strains such as Modified Vaccinia Virus Ankara may be preferred primarily for safety reasons and vaccine purposes. Plasmid-transfer vectors have three essential components: an expression cassette consisting of a natural or synthetic vaccinia virus promoter, restriction endonuclease sites for insertion of foreign genes, and flanking vaccinia virus DNA that determines the site of homologous recombination. An additional component may provide antibiotic selection or color screening. Various transfer vectors are listed in Table 14A.4.1 and one is presented in detail in Figure 14A.4.1. Homologous recombination (Fig. 14A.4.1) is the usual way of generating recombinant vaccinia viruses. Initially, recombination may result from a single cross-over event, resulting in integration of the circular plasmid and the creation of a tandem duplication; however, this intermediate is highly unstable. A second recombination event then occurs between the tandem repeats, resolving the structure into a small circular DNA molecule and either a wild-type or recombinant viral genome. An alternative method of direct ligation using an unique restriction endonuclese site in the vaccinia genome has also been described (Pfleiderer et al., 1995; Merchlinsky et al., 1997). This method avoids an E. coli cloning step and can potentially be used for direct cloning of cDNA libraries in vaccinia virus. Considerable attention has been devoted to promoters because they affect both the time and level of expression. One commonly used promoter, p7.5, contains tandem late and early promoters (Cochran et al., 1985) and provides a moderate level of expression throughout infection. The pmH5 promoter provides higher levels of both early and late expression than p7.5 (Wyatt et al., 1996). Strong natural late promoters include p11 (Bertholet et al., 1985) and pCAE (Patel et al., 1988). About a two-fold increase in expression may be achieved with the synthetic late promoter in pMJ601 and pMJ602 (Davison and Moss, 1990) or the synthetic early/late compound promoter in pSC59 and pSC65 (Chakrabarti et al., 1997; Table 14A.4.1). The ability to synthesize many different kinds of proteins, including those with transmembrane domains, is one advantage of the vaccinia virus expression system. Nevertheless, very high expression of certain genes might adversely affect virus replication. If difficulty is experienced in obtaining recombinant vaccinia with strong promoters, the weaker p7.5 or pmH5 promoter or the inducible vaccinia virus/bacteriophage T7 hybrid system should be tried (ELROY-STEIN AND MOSS, 2001). Many different methods for selecting recombinant viruses are available. The two simplest, and therefore preferable, methods are selection based on restoration of plaque size or co-expression of a fluorescent protein marker such as green fluorescent protein (GFP). Plaque size selection (Blasco and Moss, 1995) provides an extremely simple method of selecting recombinant vaccinia viruses that involves neither drugs nor reporter genes. The parental virus (vRB12) contains a deletion of most of the F13L gene, which prevents vaccinia virus from making normal-size plaques. The transfer vector pRB21 contains a segment of the F13L gene adjacent to the gene of interest, so that recombinant viruses will have a functional F13L gene and will make normal-size plaques that are easily distinguished from the pin-point plaques of vRB12. Therefore, one merely needs to pick large plaques to isolate recombinant viruses. However, it is important to check that the virus is a double cross-over recombinant. Co-insertion of GFP or other fluorescent protein marker under control of a vaccinia promoter adjacent to the gene of interest allows easy identification of recombinant plaques using a fluorescence microscope. An example is plasmid pLW-44 (Table 14A.4.1) in which GFP is stably inserted into the recombinant virus. A modification of this method, transient GFP selection, is performed using a transfer vector in which a portion of one of the two vaccinia virus flanking regions that surround the cassette containing GFP and the foreign gene is also inserted between the GFP and foreign gene (Fig. 14A.4.2). Thus, the GFP gene is flanked by direct repeats from one vaccinia virus flank. GFP-expressing plaques are isolated through several rounds of plaque purification to eliminate parental virus. However, the GFP gene is inherently unstable as recombination will readily occur between the two repeat sequences that surround the gene. Subsequent selection of non-GFP-expressing plaques allows isolation of recombinants that express only the foreign gene. Several rounds of plaque purification are necessary to ensure a pure clonal isolate. It is essential that these plaque isolates be characterized by immunostaining, PCR, or other methods to ensure that the desired gene has been incorporated. The TK selection method is based on the insertional inactivation of the thymidine kinase gene (Mortensen and Kingston, 2009). In the presence of active TK, added BrdU is phosphorylated and incorporated into viral DNA, where it causes lethal mutations. If TK− cells are used, then TK− virus will replicate normally in the presence of BrdU, whereas TK+ virus will not. Because the product of a single cross-over event is still TK+, only double cross-over events are selected. However, not all TK− viral plaques will be recombinants because spontaneous TK− mutants arise at a frequency of 1:10,000. Depending on the transfection efficiency, recombinants may comprise 10% to >90% of the TK− plaques. β-galactosidase or β-glucuronidase (GUS) screening is based on the coinsertion of the E. coli lacZ or GUS gene, under the control of a vaccinia virus promoter, into the vaccinia virus genome along with the gene of interest. If the TK gene is insertionally inactivated by such an event, recombinant viruses will be TK− and will make blue plaques on medium containing Xgal or Xgluc. This combination of color screening and TK selection will discriminate TK− recombinants from spontaneous TK− mutants. However, this method has been largely superceded by using fluorescent markers that do not require staining. XGPRT selection uses mycophenolic acid (MPA), an inhibitor of purine metabolism (see Mortensen and Kingston, 2009). Because MPA blocks the pathway for GMP synthesis, it interferes with the replication of vaccinia and severely reduces the size of vaccinia plaques. This effect can be overcome, however, by expression of the E. coli gpt gene in the presence of xanthine and hypoxanthine (i.e., XGPRT can use either of these as substrate for synthesis of GMP). Thus, coexpression of XGPRT provides a useful selection system. Usually the XGPRT gene is placed adjacent to the gene of interest and within the vaccinia virus DNA flanking sequences. Unlike the TK− situation, the viral products arising from both single and double cross-over events will be selected. Therefore, it is important to do successive plaque isolations so that the single cross-over events will be resolved. A reverse-selection procedure can be used to delete the XGPRT gene from a recombinant vaccinia virus or replace it with another gene (Isaacs et al., 1990). Another procedure, transient XGPRT selection, is performed using a transfer vector with the XGPRT gene outside of the vaccinia virus DNA flanking sequences (Falkner and Moss, 1990). Under these conditions, only the single cross-over recombinant virus will express XGPRT, providing substantial enrichment over the parental virus. However, when MPA selection is removed, the desired double cross-over recombinant virus without the XGPRT gene will have to be differentiated from the parental virus by PCR, immunostaining, or other methods. To avoid further impairment of MVA, foreign genes have been targeted to existing deletion sites. Deletion II encompasses parts of the HindIII N-K fragments; deletion III lies within the HindIII A fragment (Meyer et al., 1991). The TK gene has been successfully used as an insertion site for MVA (Carroll and Moss, 1997; T. Blanchard, pers. comm.), although there have been reports of some difficulties in isolating stable TK− MVA (Scheiflinger et al., 1996). Since there are no MVA-permissive TK− cell lines, the main attraction for using the TK site is the large number of available transfer vectors that target this site. When some foreign genes were placed under the control of the very strong synthetic early/late promoter, the resulting recombinant MVA were difficult to isolate or unstable (Wyatt et al., 1996). To date, the examples have been membrane proteins. In most cases, stable recombinants could be made by using a promoter of lower strength, e.g., pmH5 or p7.5. When a recombinant protein is somewhat toxic, non-expressing mutants that arise spontaneously can quickly overgrow the culture. To partially circumvent this problem, which has occurred with recombinant genes inserted in deletions II and III, a transfer vector, pLW-73, that directs recombination between two essential genes, I8R and G1L, was developed (Table 14A.4.1). Thus, instability resulting from any deletion of the inserted gene that includes part of either of the two essential flanking genes will result in a non-viable virus that cannot overgrow the culture. Critical Parameters Calcium phosphate transfections are described in detail in Kingston, Chen, and Rose, 2003. As discussed there, obtaining a very fine precipitate of calcium phosphate and DNA is critical for efficient transfection. Several factors can influence the quality of the precipitate. First, the pH of the transfection buffer should be between 7.0 and 7.1. Second, the CaCl2 should be added slowly and mixed gently (not vortexed), only until full mixing is achieved. The tube should then be left undisturbed until the solution is layered on the cells. Cationic lipids are simpler to use and provide similar or better transfection efficiencies. Although the Lipofectamine procedure was described in this unit, other liposomal or cationic lipids are also effective. During plaque purification and amplification of a new recombinant virus, it is important to maintain the appropriate selective pressure to prevent growth of any contaminating wild-type virus. After isolation, selection is not required. Check the purity of a recombinant virus by PCR amplification from the flanking DNA sequences. It is also advisable to perform DNA sequencing of the insert to ensure that no mutations have arisen, especially if a PCR product was used for transfection of the foreign gene. Immunostaining is another method that can be used to verify that all plaques/foci express the gene of interest. Non-expressing plaques/foci can be identified by microscopic visualization or double immunostaining. The latter technique requires an antibody to the expressed protein and one to vaccinia with different species specificities. This method has been used routinely with MVA recombinants. Freezing vaccinia virus stocks causes clumping of particles; thus, stocks should be sonicated after thawing. This is particularly important when plaque-purifying a virus, as each plaque should have arisen from a single virus. It is prudent to save half of each stock at each step in case of contamination or failure at a succeeding step. The authors strongly recommend preparation of a seed stock of the final recombinant virus that will be adequate for future needs. A common mistake is to continually passage the virus. If immunostaining is used instead of selection or color screening, the antibody must give a good signal to initially pick out the few small foci of cells infected with recombinant virus among a great excess of cells infected with the parental nonstaining virus. Before using an antibody for isolation of recombinant MVA, it can be tested by staining the MVA-infected cells after the first transfection step. Anticipated Results Approximately 1–5 × 1010 pfu of purified virus should be obtained per liter of 5 × 108 HeLa cells. Depending on the efficiency of the transfection, single, well-isolated plaques should be visible in cells infected with one of the recommended virus dilutions. With TK selection, from 10% to 90% of the plaques will contain recombinant virus. If GFP, β-galactosidase or GUS screening is also used, only fluorescent or blue recombinant virus plaques should be picked. With XGPRT selection, all plaques picked should contain recombinant virus. If the titer of recombinant virus is low, amplification can be achieved by a round of growth in the presence of MPA prior to plaquing (use the procedure described in amplification of a plaque). Cytopathic effects should be clearly visible at each step of amplification of vaccinia vius except with final infection (infected HeLa cells do not exhibit clear cytopathic effects). The titer of the final crude stock should be 1–2 × 109 pfu/ml. Time Considerations The infection/transfection procedure for generation of recombinant virus takes ~6 hr and cells are harvested after 2 days. Each round of selection and plaque purification takes 2–3 days, although little working time is required. Plaques can be picked and reinfections performed on the same day. Amplification of a single plaque isolate to a small high-titer crude stock will take ~7 days. Large stocks should then be prepared as described in UNIT 14A.3. Transfection of cells with the plasmid transfer vector requires ~2 hr to complete, and the cells are harvested after 2 days. Each round of isolation and plaque purification requires about 30 min for infection, 2 days for incubation, 3 hr for color or immunostaining, and 2 hr for isolating new recombinants and freeze-thaw cycling. Recombinant isolates can be picked and reinfections performed on the same day. From completion of plaque purification to the preparation and titering of a virus stock takes another 10 days. The work was supported by the Division of Intramural Research, NIAID, NIH. Figure 14A.4.1 Homologous recombination between a transfected plasmid and the vaccinia virus genome. The resultant recombinant virus stably incorporates a reporter gene as well as the foreign gene of interest. FL=left flank, FR=right flank, p=promoter Figure 14A.4.2 Homologous recombination between a transfected plasmid and the vaccinia virus genome demonstrating transient GFP selection. In the first recombination event, both GFP and the foreign gene are integrated into the vaccinia virus genome. Instability due to the presence of two direct repeats flanking the GFP gene result in a second recombination event upon passage that eliminates the GFP gene but maintains the foreign gene. FL=left flank, FR=right flank, p=promoter Table 16.17.1 Vaccinia Virus Transfer Vectors Vectora Promoterb Cloning Sitesc Insertion Sitesd Selection/screening Reference pGS20 p7.5 (E/L) BamHI; SmaI TK TK− Mackett et al., 1984 pSC11 p7.5 (E/L) SmaI; MCS TK TK−,β-gal Chakrabarti et al., 1985; Earl et al., 1990; Bacik et al., 1994 pMJ601, pMJ602 psyn (L) MCS TK TK−, β-gal Davison and Moss, 1990 pRB21 psyn (E/L) MCS F12L/F13L Plaque Blasco and Moss, 1995 pMC02 psyn (E/L) MCS TK TK−, GUS Carroll and Moss, 1995 pSC59 psyn (E/L) MCS TK TK− Chakrabarti et al., 1997 pSC65 psyn (E/L) MCS TK TK−, β-gal Chakrabarti et al., 1997 pLW-44 pmH5 (E/L) MSC Del II GFP Bisht et al., 2004 pLAS-1 pmH5 (E/L) MSC Del III Transient GFPe Wyatt et al., 2008 pLAS-2 pmH5 (E/L) MSC Del II Transient GFPe Wyatt et al., 2008 pLW-73 pmH5 (E/L) MSC I8R/G1L Transient GFPe Wyatt et al., 2009 pLW-76 pmH5 (E/L) MSC Del III rstr Transient GFPe L.Wyatt and B. Moss, unpubl. observ. pLW-7 psyn (E/L) MCS Del III Transient gpte Wyatt et al., 1996 pMC03 psyn (E/L) MCS Del III GUS Carroll and Moss, 1995 pLW-9 pmH5 (E/L) MCS Del III Transient gpte Wyatt et al., 1996 pLW-17 pmH5 (E/L) MCS Del II None L. Wyatt and B. Moss, unpub. observ. pLW-21 psyn (E/L) MCS Del II None L. Wyatt and B. Moss, unpub. observ. pLW-22 psyn (E/L) MCS Del II β-gal Ourmanov et al.,2000 pLW-24 p7.5 (E/L) MCS Del II None L. Wyatt and B. Moss, unpub. observ. a pRB21 was specifically designed for use with vaccinia virus vRB12, which has a deletion in the F13L gene. The plasmids pLW-44, pLAS-1, pLAS-2, pLW-73, and pLW-76 were designed for MVA. b Abbreviations: E, early; L, late; E/L, early and late. c SmaI digestion gives a blunt end for cloning any fragment that has been blunt-ended. MCS signifies multiple cloning sites. d Abbreviations: TK, thymidine kinase locus; F12L/F13L, between F12L and F13L open reading frames; Del II and III, sites of natural deletion in MVA, MCS, multiple cloning site. e Transient selection in which gpt or GFP gene is deleted from recombinant vaccinia virus during recombination; see Background Information. Key References Mackett et al., 1984. See above. Literature Cited Bacik I Cox JH Anderson R Yewdell JW Bennink JR 1994 TAP (transporter associated with antigen processing)–independent presentation of endogenously synthesized peptides is enhanced by endoplasmic reticulum insertion sequences located at the amino- but not carboxyl-terminus of the peptide J Immunol 152 381 387 8283027 Bertholet C Drillien R Wittek R 1985 One hundred base pairs of 5′ flanking sequence of a vaccinia virus late gene are sufficient to temporally regulate late transcription Proc Natl Acad Sci USA 82 2096 2100 3856886 Bisht H Roberts A Vogel L Bukreyev A Collins PL Murphy BR Subbarao K Moss B 2004 Severe acute respiratory syndrome coronavirus spike protein expressed by attenuated vaccinia virus protectively immunizes mice Proc Natl Acad Sci USA 101 6641 6646 15096611 Blasco R Moss B 1995 Selection of recombinant vaccinia viruses on the basis of plaque formation Gene 158 157 162 7607536 Carroll MW Moss B 1995 E. coli β-glucuronidase (GUS) as a marker for recombinant vaccinia viruses BioTechniques 19 352 355 7495543 Carroll MW Moss B 1997 Host range and cytopathogenicity of the highly attenuated MVA strain of vaccinia virus: Propagation and generation of recombinant viruses in a nonhuman mammalian cell line Virology 238 198 211 9400593 Chakrabarti S Brechling K Moss B 1985 Vaccinia virus expression vector: Coexpression of beta-galatosidase provides visual screening of recombinant virus plaques Mol Cell Biol 5 3403 3409 3939316 Chakrabarti S Sisler JR Moss B 1997 Compact, synthetic, vaccinia virus early/late promoter for protein expression BioTechniques 23 1094 1097 9421642 Cochran MA Puckett C Moss B 1985 In vitro mutagenesis of the promoter region for a vaccinia virus gene: Evidence for tandem early and late regulatory signals J Virol 54 30 37 3973982 Davison AJ Moss B 1990 New vaccinia virus recombination plasmids incorporating a synthetic late promoter for high level expression of foreign proteins Nucl Acids Res 18 4285 4286 2377486 Domi A Moss B 2005 Engineering of a vaccinia virus bacterial artificial chromosome in Escherichia coli by bacteriophage lambda-based recombination Nature Meth 2 95 97 Earl P Koenig S Moss B 1990 Biological and immunological properties of human immunodeficiency virus type 1 envelope glycoprotein: Analysis of proteins with truncations and deletions expressed by recombinant vaccinia viruses J Virol 65 31 41 Elroy-Stein O Moss B 2001 Gene Expression Using the Vaccinia Virus/T7 RNA Polymerase Hybrid System Curr Protoc Mol Biol 43 III 16.19 16.19.1 16.19.11 Falkner FG Moss B 1988 Escherichia coli gpt gene provides dominant selection for vaccinia virus open reading frame expression vectors J Virol 62 1849 1854 3130492 Falkner FG Moss B 1990 Transient dominant selection of recombinant vaccinia viruses J Virol 64 3108 3111 2159565 Isaacs SN Kotwal GJ Moss B 1990 Reverse guanine phosphoribosyltransferase selection of recombinant vaccinia viruses Virology 178 626 630 2219714 Kingston RE Chen CA Rose JK 2003 Calcium Phosphate Transfection Curr Protoc Mol Biol 63 I 9.1 9.1.1 9.1.11 Mackett M Smith GL Moss B 1984 General method for production and selection of infectious vaccinia virus recombinants expressing foreign genes J Virol 49 857 864 6321770 Meyer H Sutter G Mayr A 1991 Mapping of deletions in the genome of the highly attenuated vaccinia virus MVA and their influence on virulence J Gen Virol 72 1031 1038 2033387 Merchlinsky M Eckert D Smith E Zauderer M 1997 Construction and characterization of vaccinia direct ligation vectors Virology 238 444 451 9400616 Mortensen RM Kingston RE 2009 Selection of Transfected Mammalian Cells CurrProtoc Mol Biol 86 I 9.5 9.5.1 9.5.13 Ourmanov I Brown CR Moss B Carroll M Wyatt L Pleteva L Goldstein S Venson D Hirsch VM 2000 Comparative efficacy of recombinant Modified Vaccinia Virus Ankara expressing Simian Immunodeficiency Virus (SIV) gag-pol and/or env in macaque challenged with pathogenic SIV J Virol 74 2740 2751 10684290 Patel DD Ray CA Drucker RP Pickup DJ 1988 A poxvirus-derived vector that directs high levels of expression of cloned genes in mammalian cells Proc Natl Acad Sci USA 85 9431 9435 2849105 Pfleiderer M Falkner FG Dorner F 1995 A novel vaccinia virus expression system allowing construction of recombinants without the need for selection markers, plasmids and bacterial hosts J Gen Virol 76 2957 2962 8847500 Scheiflinger F Falkner FG Dorner F 1996 Evaluation of the thymidine kinase (tk) locus as an insertion site in the highly attenuated vaccinia MVA strain Arch Virol 141 663 669 8645102 Smith GL Moss B 1983 Infectious poxvirus vectors have capacity for at least 25,000 base pairs of foreign DNA Gene 25 21 28 6229451 Sutter G Wyatt LS Foley PL Bennink JR Moss B 1994 A recombinant vector derived from the host range-restricted and highly attenuated MVA strain of vaccinia virus stimulates protective immunity in mice to influenza virus Vaccine 12 1032 1040 7975844 Wyatt LS Shors ST Murphy BR Moss B 1996 Development of a replication-deficient recombinant vaccinia virus vaccine effective against parainfluenza virus 3 infection in an animal model Vaccine 14 1451 1458 8994321 Wyatt LS Earl PL Vogt J Eller LA Chandran D Liu J Robinson HL Moss B 2008 Correlation of immunogenicities and in vitro expression levels of recombinant modified vaccinia virus Ankara HIV vaccines Vaccine 26 486 493 18155813 Wyatt LS Earl PL Xiao W Americo JA Cotter CA Vogt J Moss B 2009 Elucidating and minimizing the loss by recombinant vaccinia virus of Human Immunodeficiency Virus gene expression resulting from spontaneous mutations and positive selection J Virol 83 7176 7184 19420086
PMC005xxxxxx/PMC5123796.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 2984771R 5824 Mil Med Mil Med Military medicine 0026-4075 1930-613X 27753561 5123796 10.7205/MILMED-D-15-00368 NIHMS748127 Article Impact of Operational Theater on Combat and Noncombat Trauma-Related Infections Tribble David R. MD * Li Ping MS * Warkentien LCDR Tyler E. MC, USN * Lloyd Col Bradley A. USAF, MC ** Schnaubelt Maj Elizabeth R. MC, USA * Ganesan Anuradha MD * Bradley William MS * Aggarwal Deepak MSE, MSPH * Carson M. Leigh MS * Weintrob Amy C. MD **** Murray COL Clinton K. MC, USA ** The Infectious Disease Clinical Research Program Trauma Infectious Disease Outcomes Study Investigative Team * Infectious Disease Clinical Research Program, Preventive Medicine & Biostatistics Department, Uniformed Services University of the Health Sciences, 4301 Jones Bridge Road, Bethesda, MD 20814 The Henry M. Jackson Foundation for the Advancement of Military Medicine, Inc., 6720A Rockledge Drive, Suite 100, Bethesda, MD 20817 * Walter Reed National Military Medical Center, 8901 Wisconsin Avenue, Bethesda, MD 20889 ** San Antonio Military Medical Center, 3551 Roger Brooke Drive #3600, Fort Sam Houston, TX 78234 *** Landstuhl Regional Medical Center, CMR 402, Box 1559, APO AE 09180, Landstuhl, Germany Corresponding Author: Dr. David R. Tribble, Uniformed Services University of the Health Sciences, 4301 Jones Bridge Road, Bethesda, MD 20814, [email protected], Phone: 301-816-8404, Fax: 301-816-8406 31 12 2015 10 2016 01 10 2017 181 10 12581268 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. The Trauma Infectious Disease Outcomes Study began in June 2009 as combat operations were decreasing in Iraq and increasing in Afghanistan. Our analysis examines the rate of infections of wounded U.S military personnel from operational theaters in Iraq and Afghanistan admitted to Landstuhl Regional Medical Center between June 2009 and December 2013 and transferred to a participating U.S. hospital. Infection risk factors were examined in a multivariate logistic regression analysis (expressed as odds ratios [OR]; 95% confidence intervals [CI]). The study population includes 524 wounded military personnel from Iraq and 4766 from Afghanistan. The proportion of patients with at least one infection was 28% and 34% from the Iraq and Afghanistan theaters, respectively. The incidence density rate was 2.0 (per 100 person-days) for Iraq and 2.7 infections for Afghanistan. Independent risk factors included large-volume blood product transfusions (OR: 10.68; CI: 6.73–16.95), high injury severity score (OR: 2.48; CI: 1.81–3.41), and improvised explosive device injury mechanism (OR: 1.84; CI: 1.35–2.49). Operational theater (OR: 1.32; CI: 0.87–1.99) was not a risk factor. The difference in infection rates between operational theaters is primarily due to increased injury severity in Afghanistan from a higher proportion of blast-related trauma during the study period. Combat-related infections injury severity trauma-related infections combat care military medicine INTRODUCTION It is widely recognized that individuals with combat-related injuries are at high risk for infectious complications.1–4 Combat-related injuries result in circumstances, such as breach of physical host defences, hypoxic tissue damage/necrosis, and implantation of foreign bodies, which greatly increase the risk of infection. The nature of trauma inflicted during combat (e.g., blast injuries) is generally more severe than injuries acquired in civilian settings.5–7 Since U.S. military personnel were first deployed to Iraq (Operations Iraqi Freedom [OIF] and New Dawn [OND] and Afghanistan (Operation Enduring Freedom [OEF]), over 52,100 service members have been wounded in action.8 These military operations saw changes in the patterns and severity of combat-related injuries, primarily due to the expanded utilization of improvised explosive devices (IEDs).2–4,9–11 Accordingly, combat care evolved through advancements in preventive measures and treatments, along with the implementation of the Joint Theater Trauma System (JTTS) in November 2004.12,13 In part due to the integrated approach to combat care executed by the JTTS, which emulates successful civilian trauma systems, the overall mortality rate of military personnel decreased during the conflicts in Iraq and Afghanistan (8.8%) when compared to World War II and Vietnam (22.8% and 16.5%, respectively).12 Nonetheless, the improved survival of wounded military personnel resulted in a rise of infectious complications and consequent effects on morbidity.2–4,14–17 In order to better characterize these infections, an observational cohort study of short- and long-term infectious consequences among U.S. service members injured during deployment (Department of Defense [DoD]-Department of Veterans Affairs, Trauma Infectious Disease Outcomes Study [TIDOS]) was initiated in 2009. An examination of data collected from the project reported an overall infection incidence density of 1.8 per 100 person-days among wounded U.S. military personnel medically evacuated to Landstuhl Regional Medical Center (LRMC) in Germany and transitioned to participating military treatment facilities (MTFs) in the U.S. between June and August 2009. In the analysis, infection rates were assessed with regards to level of care (LRMC and U.S. MTFs) and admitting units (intensive care versus non-critical ward); however, operational theater where the injury was sustained was not examined.3 For military personnel with injuries sustained during the recent conflicts, the risk of infection may differ between operational theaters due to diversities in injury mechanism (e.g., gunshot and blast wounds), environment (i.e., arid/urban landscape in Iraq versus the mountainous and agricultural settings in Afghanistan), or infectious exposures. Our objective was to assess infection rates among U.S. military personnel injured during deployment (combat and noncombat) with respect to operational theater (Iraq and Afghanistan). METHODS Study Design The TIDOS project is a multisite, observational cohort study initiated with the goal of describing the incidence, risk factors, and clinical outcomes of infectious complications associated with deployment-related injuries. The study commenced during a period when combat operations were decreasing in Iraq while simultaneously increasing in Afghanistan. Full details regarding the TIDOS project design have been previously published.3 Eligible subjects include U.S. service members injured during deployment (June 1, 2009 to January 31, 2015) and medically evacuated to LRMC followed by transition to a participating U.S. MTF. This study (IDCRP-024) is approved by the Infectious Disease Institutional Review Board of the Uniformed Services University of the Health Sciences in Bethesda, Maryland. Data Collection Trauma information (e.g., injury patterns and severity) were obtained through the DoD Trauma Registry (DoDTR).12 The TIDOS infectious disease module augmented the DoDTR data by providing detailed information on antimicrobial therapy, microbiology, and infectious outcomes from injury through initial hospitalization at a participating U.S. MTF: National Naval Medical Center and Walter Reed Army Medical Center in the National Capital Region (Walter Reed National Military Medical Center after September 2011), and Brooke Army Medical Center in San Antonio, Texas (San Antonio Military Medical Center after September 2011). Study Definitions and Endpoints Traumatic injuries sustained during deployment were categorized using injury codes defined by the Abbreviated Injury Scale (AIS), a consensus-derived anatomically-based injury scoring system.18 A composite injury severity score (ISS) was calculated for each patient based on the top three maximum AIS anatomical region values across all clinical facilities. Combat-related injuries were identified as traumatic injuries occurring within the operational theater that include the following injury mechanisms: blast, gunshot wound, motor vehicle/helicopter crash, fall/crush, and burns. Noncombat injuries, including sports and training injuries, were sustained while deployed and may include similar mechanisms (i.e., falls/crush, burns, and motor vehicle crashes), but were not directly related to combat operations. As described in Tribble et al.,3 infections were identified utilizing a combination of clinical findings and laboratory test results via review of medical records and were classified based upon the National Healthcare Safety Network (NHSN) standardized definitions for healthcare-associated infections.19 Furthermore, an infection was included if, in the absence of meeting a priori defined criteria, there was a clinical diagnosis associated with the initiation of directed antimicrobial therapy that was continued for more than five days. Infections were excluded from the analysis if medical records provided an alternative diagnosis combined with the termination of antimicrobial treatment. Multidrug-resistant (MDR) isolates were identified in accordance with definitions published by the NHSN.20 Isolates were classified as colonizing if they were collected via infection control-based surveillance. Isolates were considered infecting if they were collected as part of a clinical infection work-up and met infection clinical syndrome criteria. Statistical Analysis Tests of association for categorical variables were conducted using Chi-square and Fisher’s exact tests, while medians were compared by the Kruskal-Wallis test. Logistic regression models were used to assess the relationship between potential risk factors and outcomes (presence or absence of infection). The risk factor analysis was restricted to data collected from patients who transferred to participating U.S. MTFs. For each of the variables, the best-fitting parsimonious model was sought. A correlation analysis was also conducted to evaluate the relationship between potential risk factors. Models were compared on the basis of the Akaike Information Criterion and Hosmer and Lemeshow Goodness of Fit. Analysis was conducted with SAS® version 9.3 (SAS, Cary, NC). Data are expressed as odds ratios with 95% confidence intervals. Statistical significance was defined as p<0.05. RESULTS Study Population and Injury Patterns A total of 5290 wounded military personnel were admitted to LRMC between June 2009 and December 2013 (Figure 1), of which 4766 sustained injuries in Afghanistan (82% combat-related) and 524 in Iraq (54% combat-related). As previously mentioned, the start of the study period coincided with declining combat operations in Iraq (OIF ended on August 31, 2010 and peacekeeping support with OND began on September 1, 2010), while operations increased in Afghanistan. Specifically, 327 and 197 were wounded in support of OIF (62% combat-related) and OND (42% combat-related), respectively (Figure 2). For both theaters, the population was predominantly young enlisted men (>90%) serving in the U.S. Army (83% and 67%, respectively) or U.S. Marines (4% and 25%, respectively; Table 1). Furthermore, 52% and 72% of combat casualties sustained a blast injury in the Iraq and Afghanistan theaters, respectively, of which 67% and 78%, respectively, were the result of an IED. In addition, 14% of military personnel in Iraq and 33% in Afghanistan were injured while on foot patrol. The predominant injury mechanisms among personnel with noncombat trauma were falls (31% and 32% for Iraq and Afghanistan theaters, respectively) and sports-related injuries (21% and 15%, respectively). Combat-related injuries among personnel serving in the Afghanistan theater were more severe than noncombat trauma (Table 1), as indicated by the significantly higher ISS (median of 12 versus 4; p<0.0001) and greater proportion of admittance to the intensive care unit (ICU; 30% versus 5%; p<0.0001). The pattern of injury was also significantly different with a higher proportion of open injuries (skin/soft-tissue and fractures) sustained by combat casualties (84%; p<0.0001) compared to noncombat (27%). Furthermore, the proportion of wounded service members receiving massive transfusions of packed red blood cells plus whole blood (RBC; >10 units) within the first 24 hours was significantly higher among those with combat-related trauma compared to noncombat (15% versus 0.4%; p<0.0001). Lastly, significantly more patients with combat-related injuries were also prescribed prophylactic antibiotics for prevention of infections within 48 hours following injury (73% versus 23%; p<0.0001). A similar pattern was observed among military personnel injured in the Iraq theater (Table 1). Specifically, military personnel with combat-related trauma had significantly higher ISS (median of 10 versus 4; p<0.0001), occurrence of open injuries (76% versus 26%; p<0.0001), proportion of patients admitted to the ICU (25% versus 8%; p=0.015), any amount of RBCs transfused within 24 hours (24% versus 0.8%; p<0.0001), and receipt of prophylactic antibiotics within 48 hours (56% versus 18%; p<0.0001). Infection Characteristics From the population of 2513 patients transferred to participating U.S. MTFs (Table 2), 852 patients (34%) developed at least one infection (94% from Afghanistan and 6% from Iraq). Of the 54 patients from Iraq with at least one infection, 40 sustained injuries in OIF (78% combat-related) and 14 in OND (71% combat-related). A total of 2003 and 103 unique infections were diagnosed from military personnel wounded in Afghanistan and Iraq, respectively, of which 99% and 80% were combat-related, respectively (Table 2). Comparison of the incidence density rate ratio found that there was a significantly higher proportion of combat-related infections compared to noncombat in the Afghanistan theater (p<0.0001), but not in the Iraq theater of operation (p=1.0). Skin and soft-tissue infections (SSTIs) and pneumonia were common for patients injured in both theaters regardless of whether they had combat-related or noncombat trauma (Table 2). Specifically, 47% of unique infections were SSTIs among both combat and noncombat trauma patients from Afghanistan, whereas pneumonia contributed 14% of infections among combat casualties and 23% among noncombat trauma patients. Regarding patients from Iraq, 22% and 24% of unique infections in combat and noncombat trauma patients were SSTIs, respectively, while it was 21% and 29% for pneumonia, respectively. Among those with an infection, most were diagnosed after transfer to a participating U.S. MTF (70% and 67% of patients from Afghanistan and Iraq theaters, respectively). In addition, the overall rate of infections was higher among patients initially admitted to the LRMC ICU compared to the noncritical ward for both theaters of operation (p<0.0001). The rates were also statistically significant when combat casualties for both theaters and noncombat trauma from the Iraq theater were considered (p<0.001); however, the ratio was not significant among personnel with noncombat trauma from the Afghanistan theater (p=0.083). Between the two operational theaters regardless of combat status, the duration from injury to development of any type of infection was comparable (median: 6 days). When specific infection syndromes were considered, osteomyelitis had the longest duration from injury to diagnosis (median of 26 and 28 days for Afghanistan and Iraq, respectively), whereas pneumonia, bloodstream infections, and sepsis developed a median of 5 to 7 days after injury (data not shown). Among patients with diagnosed infections, the proportion who had only one infection while hospitalized was also similar between the Afghanistan and Iraq theaters (45% versus 43%, respectively); however, 21% of patients from Afghanistan were diagnosed with at least four infections compared to 9% of patients from Iraq (Table 2). When the data were restricted to noncombat trauma, 83% of personnel from the Afghanistan theater had only one infection compared to 46% from the Iraq theater. Overall, the incidence density rate (number of infections/100 person-days) was higher for military personnel serving in Afghanistan compared to patients with Iraq-related traumatic injuries (2.7 and 2.0, respectively). Clinical Microbiology and Post-Trauma Antibiotic Prophylaxis At admission to LRMC, 353 (7%) of wounded personnel were colonized with MDR gram-negative bacteria, as determined from surveillance groin swabs collected at hospital admission. Among the Afghanistan theater, 8% and 0.9% of personnel with combat-related and noncombat injuries were colonized at LRMC admission, respectively. For the Iraq theater, 6% of military personnel with combat-related injuries were colonized with MDR gram-negative bacteria compared to 0.8% of patients with noncombat injuries at LRMC admission. Among the 2513 patients that transferred to a participating U.S. MTF, 258 (10.2%) were colonized with MDR gram-negative bacteria at admission (10.7% and 4.6% of personnel injured in Afghanistan and Iraq, respectively). From patients injured in the Afghanistan theater, 2354 colonizing isolates were collected (27% MDR) across all levels of care, whereas 188 isolates were collected from Iraq (18% MDR). Overall, Escherichia coli and Klebsiella pneumoniae were the most common colonizing organisms for both the Afghanistan (43% of E. coli and 28% of K. pneumoniae were MDR, respectively) and Iraq theaters (30% and 12% MDR, respectively). In addition, Acinetobacter calcoaceticus baumannii (ACB) complex isolates were also frequently MDR from both Afghanistan (46%) and Iraq (22%). While not performed at LRMC, surveillance for methicillin-resistant Staphylococcus aureus (MRSA) using nares swabs was conducted at the U.S. MTFs and the rate of colonization was 4.3% overall (95% confidence interval: 3.5–5.1%) with no significant difference between theaters and combat versus noncombat. Specifically, 5.1% of patients from the Iraq theater who transferred to a participating U.S. MTF had MRSA colonization, while it was 4.2% from the Afghanistan theater. In addition, 4.1% and 6.2% of personnel with combat and non-combat injuries were colonized with MRSA. A total of 1542 unique infections (73.2% of 2106) had a corresponding infection work-up that yielded bacterial growth, of which 80.4% and 49.1% grew gram-negative and gram-positive organisms, respectively. Gram-negative organisms (susceptible and MDR) isolated from infection workups were predominantly collected from combat casualties (25% and 20% in Afghanistan and Iraq, respectively), compared to 5% and 6% of personnel with noncombat trauma, respectively. Overall, 68% of the patients with infections had gram-negative bacteria isolated in infection work-ups. Among personnel with combat-related injuries sustained in Afghanistan, the gram-negative organisms most commonly identified during infection workups were Pseudomonas aeruginosa, followed by E. coli and ACB complex, of which 10%, 73%, and 78% were MDR, respectively (Table 3). Combat casualties from Iraq had a similar microbiological profile with 14% of P. aeruginosa, 57% of A. baumannii, and 60% of E. coli determined to be MDR. S. aureus contributed 11.8% to the gram-positive organisms isolated from infection work-ups, of which 44% were MRSA. Overall, 1923 (77%) patients transferred to a participating U.S. MTF received prophylactic antibiotics, of which 1806 (94%) and 117 (6%) sustained injuries in Afghanistan and Iraq, respectively. Among the 852 patients who developed an infection, 761 (89%) received prophylactic antibiotics within 48 hours post-injury. In addition, 90% of patients who sustained injuries in Afghanistan and developed infections received prophylactic antibiotics, while it was 72% for Iraq. Patients who received prophylactic antibiotics more commonly had ISS >15 (65%) and sustained injuries via a blast mechanism (72%). Furthermore, 87% of patients with culture growth of MDR organisms (surveillance or infection work-up) received prophylactic antibiotics (data not shown). In a Chi-square analysis, the association of prophylactic antibiotics and occurrence of any infection was significant (odds ratio [OR]: 3.6; 95th confidence interval [CI]: 2.8–4.6; p<0.0001). The data remained significant when the Afghanistan and Iraq theaters were considered separately (p<0.0001 and p=0.03, respectively). Furthermore, there was also a significant association between prophylactic antibiotic use and isolation of MDR organisms via colonization surveillance or infection work-up among patients (OR: 2.6; CI: 2.0–3.4; p<0.0001). Risk Factor Analysis From the patients that transferred to U.S. MTFs, operational theater, circumstances of injury (i.e., combat-related and mechanism), composite ISS, RBC transfusion requirements within 24 hours of injury, open injury, branch of service, MDR gram-negative colonization at LRMC admission, post-trauma antibiotic prophylaxis, and admission to the ICU were examined in a logistic regression analysis (Table 4). The composite ISS and RBC requirements were evaluated as ranked variables. Injuries sustained via an IED blast mechanism and during combat, ISS >15, RBC requirements ≥1 unit, MDR gram-negative colonization, occurrence of open wounds, service in the U.S. Marines, use of prophylactic antibiotics, and admission to the ICU were significantly associated with development of infections in the univariate model (p<0.0001). In addition, injuries that occurred in the Iraq theater were significantly less likely to be associated with the development of infection compared to the Afghanistan theater (p<0.0001). In the multivariate model (Table 4), the risk for infection was highest among patients who received >20 units of RBCs within 24 hours of injury (p<0.0001). Injuries from IEDs (p<0.0001), post-trauma antibiotic prophylaxis (p=0.033), ISS >15 (p<0.002), and ICU admission (p<0.0001) were also significantly associated with the development of infections. Operational theater was not an independent risk factor for an infection following traumatic injury (p=0.193). Furthermore, when the analysis was repeated after separating Afghanistan into two time periods (June 2009–May 2012 and May 2012–December 2013), there was no significant association with operational theater (data not shown). DISCUSSION This analysis assessed characteristics and rates of infectious complications among 5290 U.S. service members with deployment-related injuries in association with two combat operational theaters (Iraq and Afghanistan). Overall, a higher rate of infection was observed with the Afghanistan theater compared to Iraq during a contemporaneous period (2.7 versus 2.0 infections per 100 person-days). After controlling for injury severity and other factors, there was no statistical association between operational theater and the risk of developing an infection. It is also notable that personnel with noncombat injuries also had high rates of infection (1.0 and 2.0 per 100 person-days for Afghanistan and Iraq, respectively). Our data corroborate prior analyses which reported associations of infectious consequences among wounded military personnel with the severity and mechanism of injury.14,15,21–23 Measures of injury severity (i.e., ICU admission, ISS, and hemorrhage as indicated by RBC transfusion requirements within 24 hours) primarily explain the statistical difference in infection rate independent of operational theater as data in our analysis suggest that injuries sustained in Afghanistan were generally more severe and likely due to the high proportion of blast injuries and dismounted status during the same time period. In addition, large-volume transfusions of blood products have been previously shown to be an independent risk factor for infection following deployment-related trauma, possibly due to inducing a transient state of immunosuppression.22,23 Furthermore, the association of prophylactic antibiotics within 48 hours is also consistent with use within a higher at-risk population.5 Patients who received prophylactic antibiotics were observed to have higher injury severity as indicated by increased proportion with ISS >15, ICU admissions, and predominance of blast injuries. Nonetheless, a more detailed exploration of antimicrobial regimens and related infection outcomes is warranted. The role of antecedent bacterial colonization and subsequent infection is unknown24 and the variable was not statistically significant in the risk factor analysis (p=0.08), our results suggest that colonization with MDR gram-negative organisms may be a risk factor and should be investigated further. The colonization data are consistent with our prior analysis that found similar annual rates of MDR gram-negative bacilli colonization over a three-year period (2009–2012).25 It is also noteworthy that the rate of MDR gram-negative colonization at admission to the U.S. MTFs was higher than the rate of MRSA colonization (10.2% versus 4.3%). Although a great deal of focus has been placed on injuries sustained during combat, a high rate of infection was found among personnel with noncombat injuries. Many noncombat injuries in an operational zone involve mechanisms such as motor vehicle collisions, falls, and burns, which often result in open wounds. In a prior retrospective analysis of 4566 military personnel with noncombat injuries not sustained in a war zone found that 8.2% had at least one related infection. Pneumonia was predominant (4%) with a lower proportion of cellulitis/wound infections (2.4%) and sepsis (0.9%).7 When considering all noncombat trauma patients from both theaters of operation, 15% had at least one infection, with pneumonia and SSTIs contributing the greatest proportion to unique infections. While the majority of previous analyses have not examined data on a per theater basis, infection rates from the recent conflicts have been published. Data from 16,742 deployment-injured patients were collected from a trauma registry and determined an infection rate of 5.5% (annual range: 0.6–10.9%).14 Moreover, an early evaluation from the TIDOS project reported 5% of LRMC admissions and approximately 27% of patients transferred to the U.S. developed infections.3 In our analysis, the proportion of infections among the total number of patients admitted to LRMC was consistent with the prior analyses (5%); however, the overall infection rate among wounded personnel who transferred to one of the participating U.S. MTFs was higher (34%). Analysis of data from the United Kingdom has also found a high rate of infection related to extremity injuries (24%), which corresponds to our finding of the predominance of SSTIs and minor contribution of osteomyelitis.26 In general, the rate of infections among wounded personnel from the Iraq theater was lower than Afghanistan (2.0 and 2.7 infections per 100 person-days, respectively). One reason for the differing infection rates may be the reduction of combat-related injuries as military operations ceased in Iraq during the study period (OIF concluded in August 2010 and was followed by OND peacekeeping efforts); however, it is important to note that despite the shift to peacekeeping efforts, a risk of combat-related injuries still occurs. Specifically, 42% of the injuries sustained in Iraq during OND were combat-related. Another possible explanation is that as combat operations transitioned, patterns of injury changed due to differences in military tactics (on both sides) affecting casualties. During 2010, as combat operations were concluding in Iraq and increasing in Afghanistan, the number of traumatic amputations substantially increased. Between 2010 and 2011, the amputation rate rose from 3.5 to 14 per 100 combat support facility trauma admissions, and was primarily the result of dismounted patrols encountering IEDs in Afghanistan.9,27 The consequent dismounted complex blast injuries were characterized by lower extremity amputations (unilateral or bilateral), upper extremity amputations, pelvic and urogenital injuries, and spinal injuries.27 United Kingdom military personnel were also greatly impacted by this injury pattern with 2.8% of combat casualties sustaining bilateral lower limb amputations over a six-year period.11 Due to the severity of these injuries, patients generally required large-volume blood transfusions (>10 units), debridements, and further surgical procedures in response to complications, such as infections.11,27,28 One example was the unexpected surge in invasive fungal wound infections among military personnel who sustained blast injuries in Afghanistan. Specifically, nearly 7% of the combat casualties admitted to LRMC between June 2009 and August 2011 were diagnosed with an invasive fungal wound infection.29–31 A similar emergence of invasive fungal wound infections was also reported among United Kingdom military personnel with blast injuries sustained in Afghanistan.32 While information is available on injury patterns and infection rates,2,3,14,15,28,33–35 further data defining the progression of infections and resultant short- and long-term outcomes are necessary. The findings in this military setting provide support for the identification of infection risk factors related to trauma sustained during deployment; however, the feasibility of using these factors in predictive modeling with clinical care application still needs to be assessed. Nonetheless, this information emphasizes the need for forward medical support in the deployed setting and a high index of suspicion for infectious complications following traumatic injuries, regardless of whether they are sustained during combat or noncombat. A limitation of this analysis that should be considered is that infection data were collected exclusively from patients who transferred to a TIDOS-participating U.S. MTFs (approximately 48% of subjects admitted to LRMC). In general, these patients experienced more severe injuries compared to those who transferred to U.S. MTFs other than the ones included in this analysis. Thus, the applicability of data reported herein to all U.S. injured service members is uncertain. Combat casualty care is continuously advancing as new technology and data become available; however, infectious complications remain a serious cause of morbidity. The implementation of epidemiologic and surveillance projects, such as TIDOS and the Multidrug-Resistant Organism Repository and Surveillance Network (MRSN),36 are integral in informing the military health system on these key issues. With the emergence of new challenges, such as MDR bacterial organisms, healthcare-associated transmission across evacuation and MTFs, and invasive fungal wound infections, further examination of infection predictive factors, microbiological findings, real-time surveillance and support for control of outbreaks of MDR bacterial organisms through the MRSN, and specific infectious disease syndromes among deployed service members is warranted to improve crucial elements of combat casualty care including trauma systems, infection control policies, early detection, and antimicrobial selection. Acknowledgments We are indebted to the Infectious Disease Clinical Research Program Trauma Infectious Disease Outcomes Study team of clinical coordinators, microbiology technicians, data managers, clinical site managers, and administrative support personnel for their tireless hours to ensure the success of this project. Funding sources: Supported by the Infectious Disease Clinical Research Program (IDCRP), a Department of Defense (DoD) program executed through the Uniformed Services University grant IDCRP-024, the National Institute of Allergy and Infectious Diseases, National Institutes of Health (NIH), under Inter-Agency Agreement Y1-AI-5072, and the Department of the Navy under the Wounded, Ill, and Injured Program. Figure 1 Flow chart for disposition of patients admitted to Landstuhl Regional Medical Center (LRMC) between June 2009 and December 2013 with deployment-related injuries. Figure 2 Flow chart for disposition of patients admitted to Landstuhl Regional Medical Center with deployment-related injuries sustained in the Iraq theater. Combat operations in Iraq (Operation Iraqi Freedom [OIF]) ended on August 31, 2010 and were followed by peacekeeping efforts (Operation New Dawn [OND]) which began on September 1, 2010 and ended on December 31, 2011. Table 1 Patient Demographics and Injury Characteristics, No. (%) by Theater of Operation (June 2009 – December 2013) Characteristic1 Total (N = 5290) Afghanistan (N = 4766) Iraq (N = 524) Noncombat Injury (N=852) Combat- related Injury (N=3914) p-value 2 Noncombat Injury (N=239) Combat-related Injury (N=285) p-value 3 Male 5160 (97.5) 813 (95.4) 3853 (98.4) <0.0001 223 (93.3) 271 (95.0) 0.382 Age, median (IQR) 24.9 (22.1, 29.6) 26.4 (22.6, 34.4) 24.4 (22.0, 28.3) <0.0001 28.1 (22.8, 35.8) 25.7 (22.6, 30.8) 0.012 Branch of Service <0.001 0.031 Army 3608 (68.2) 552 (64.7) 2619 (66.9) 187 (78.2) 250 (87.7) Marine 1206 (22.7) 151 (17.7) 1034 (26.4) 14 (5.8) 7 (2.4) Military Grade/Rank 0.003 0.462 Enlisted 4790 (90.5) 747 (87.6) 3573 (91.2) 213 (89.1) 257 (90.1) Officer / Warrant 379 (7.2) 80 (9.4) 256 (6.5) 22 (9.2) 21 (7.4) Mechanism of Injury <0.0001 <0.0001 Blast 2912 (55.0) 0 2764 (70.6) <0.0001 0 148 (51.9) <0.0001 IED 2262 (42.8) 0 2163 (55.3) 0 99 (34.7) Non-IED 650 (12.3) 0 601 (15.4) 0 49 (17.2) Gunshot wound only 965 (18.2) 0 918 (23.5) 0 47 (16.5) Motor vehicle collision only 195 (3.7) 10 (1.2) 110 (2.8) 6 (2.5) 69 (24.2) Helicopter crash 35 (0.7) 0 30 (0.8) 0 5 (1.8) Blunt object 118 (2.2) 79 (9.3) 3 (<0.1) 33 (13.8) 3 (1.1) Fall 384 (7.3) 273 (32.0) 31 (0.8) 74 (31.0) 6 (2.1) Sports 180 (3.4) 130 (15.3) 0 50 (20.9) 0 Other4 499 (9.4) 358 (42.0) 58 (1.5) 76 (31.8) 7 (2.5) Dismounted at time of injury 1708 (32.3) 283 (33.2) 1299 (33.2) <0.0001 85 (35.6) 41 (14.4) <0.0001 Injury Severity Score, median (IQR) 10 (5, 22) 4 (4, 8) 12 (6, 27) <0.0001 4 (4, 8) 10 (5, 21) <0.0001 RBC Transfusions: 1st 24 Hours <0.0001 <0.0001 0/missing units5 3927 (74.2) 838 (98.3) 2634 (67.2) 237 (99.1) 218 (76.4) 1–9 units 761 (14.3) 11 (1.2) 698 (17.8) 2 (0.8) 50 (17.5) 10 – 20 units 336 (6.3) 2 (0.2) 322 (8.2) 0 12 (4.2) >20 units 266 (5.0) 1 (0.1) 260 (6.6) 0 5 (1.7) Occurrence of open injury 3793 (71.7) 233 (27.3) 3281 (83.8) <0.0001 61 (25.5) 218 (76.4) <0.0001 ICU Admission 6 <0.0001 0.015 LRMC only 361 (6.8) 15 (1.8) 313 (8.0) 9 (3.8) 24 (8.4) U.S. MTFs ± LRMC 965 (18.2) 29 (3.4) 879 (22.5) 11 (4.6) 46 (16.1) Non-ICU 1180 (22.3) 135 (15.8) 940 (24.0) 43 (18.0) 62 (21.8) Prophylactic antimicrobial use within 48 hours of injury 3248 (61.3) 196 (23.0) 2849 (72.7) <0.0001 44 (18.4) 159 (55.7) <0.0001 ICU – intensive care unit; IED – improvised explosive device; IQR – Interquartile Range; LRMC – Landstuhl Regional Medical Center; MTFs – military treatment facilities; RBC – packed red blood cells plus whole blood 1 Data are missing for some of the variables (branch of service missing 47; military rank missing 75; mechanism of injury missing 2; ICU admittance missing 2784). The high amount of ICU admission missing is due to the transfer of approximately half of the patients to non-participating sites. 2 P-value compares the noncombat and combat-related injury data for Afghanistan. Missing values are not included in calculation. 3 P-value compares the noncombat and combat-related injury data for Iraq. Missing values are not included in calculation. 4 Other includes burns, machinery/equipment accidents, noncombat explosions, and crush injuries. 5 Missing RBC transfusion data are not randomly distributed. Patients with missing RBC data are characterized by lower injury severity scores and shock indices. In addition, the majority of patients with missing RBC data did not sustain a traumatic amputation and were not admitted to the LRMC ICU. 6 Admission to the ICU is recorded within the first week of care at each facility. Table 2 Characteristics of Inpatient Infections among Wounded Military Personnel after Transfer to a United States Military Treatment Facility (2009 – 2013)1 Afghanistan Theater Iraq Theater Noncombat Injury (N=180) Combat-related Injury (N=2138) Total (N=2318) Noncombat Injury (N=63) Combat-related Injury (N=132) Total (N=195) Total Days of Patient Observations, No. 3138 71,410 74,548 1043 4131 5174 Unique Infections, No. 30 1973 2003 21 82 103 Infections per 100 person-days, No. (95% CI) 1.0 (0.6–1.4) 2.8 (2.6–2.9) 2.7 (2.6–2.8) 2.0 (1.2–3.1) 2.0 (1.6–2.5) 2.0 (1.6–2.4) Patients admitted initially to LRMC ICU2 1.4 (0.8–2.3) 3.7 (3.5–3.9) 3.6 (3.5–3.8) 4.1 (2.4–6.7) 2.6 (2.1–3.3) 2.8 (2.3–3.5) Patients admitted initially to LRMC Ward2 0.7 (0.4–1.2) 0.7 (0.6–0.8) 0.7 (0.6–0.8) 0.8 (0.2–1.8) 0.8 (0.4–1.4) 0.8 (0.4–1.2) Incidence density rate ratio (95% CI): Combat versus Noncombat3 NA NA 2.9 (2.0–4.3) NA NA 0.99 (0.6–1.7) Incidence density rate ratio (95% CI): LRMC ICU versus Ward4 2.0 (0.9–4.6) 5.1 (4.3–6.0) 5.0 (4.3–5.9) 5.4 (1.9–18.9) 3.5 (1.8–7.3) 3.7 (2.2–6.8) Patients with ≥1 infection, No. (%) 24 (13) 774 (36) 798 (34) 13 (21) 41 (31) 54 (28) Infections per patient, No. (%) 5 1 event 20 (83.3) 339 (43.8) 359 (45.0) 6 (46.2) 17 (41.5) 23 (42.6) 2 events 3 (12.5) 174 (22.5) 177 (22.2) 6 (46.2) 14 (34.1) 20 (37.0) 3 events 0 97 (12.5) 97 (12.2) 1 (7.7) 5 (12.2) 6 (11.1) ≥ 4 events 1 (4.2) 164 (21.2) 165 (20.7) 0 5 (12.2) 5 (9.3) Level of care location for infection, No. (%) 5 LRMC only 5 (20.8) 84 (10.9) 89 (11.2) 3 (23.1) 6 (14.6) 9 (16.7) U.S. MTF only 19 (79.2) 536 (69.3) 555 (69.5) 7 (53.8) 29 (70.7) 36 (66.7) Both LRMC and U.S. MTF 0 154 (19.9) 154 (19.3) 3 (23.1) 6 (14.6) 9 (16.7) Type of Infection, No. (%) 6 Skin and soft-tissue infections (SSTI) 14 (46.7) 918 (46.5) 932 (46.5) 5 (23.8) 18 (22.0) 23 (22.3) Pneumonia 7 (23.3) 273 (13.8) 280 (14.0) 6 (28.6) 17 (20.7) 23 (22.3) Bloodstream infection 3 (10.0) 276 (14.0) 279 (13.9) 4 (19.1) 13 (15.9) 17 (16.5) Sepsis (excluding SIRS) 0 86 (4.4) 86 (4.3) 1 (4.8) 2 (2.4) 3 (2.9) Osteomyelitis 2 (6.7) 121 (6.1) 123 (6.1) 0 17 (10.7) 17 (16.5) CI – Confidence Interval; ICU – intensive care unit; LRMC – Landstuhl Regional Medical Center; MTF – military treatment facility; NA – Not applicable; SIRS – systemic inflammatory response system 1 Data are restricted to patients who transferred to a participating U.S. MTF following treatment at Landstuhl Regional Medical Center. 2 Total number of patients initially admitted to LRMC ICU was 1195 for Afghanistan (combat: 1152; noncombat: 43) and 88 for Iraq (combat: 69; noncombat: 19). Total number of patients initially admitted to LRMC ward was 1118 for Afghanistan (combat: 982; noncombat 136) and 107 for Iraq (combat: 63; noncombat: 44). 3 p-value for incidence density rate ratio of combat versus noncombat for Afghanistan and Iraq was <0.0001 and 1.0, respectively 4 p-value for incidence density rate ratio of LRMC ICU versus ward for Afghanistan was <0.0001 (combat: <0.0001; noncombat: 0.083) and for Iraq was <0.0001 (combat: <0.0001; noncombat: <0.001). 5 The total number of patients with ≥1 infection was used to calculate percent. P-values comparing the combat and noncombat data from Afghanistan and Iraq were 0.001 and 0.722 for infections per subject, respectively, and 0.010 and 0.448 for location of infection, respectively. 6 The total number of unique infections was used to calculate percent. Does not include miscellaneous infections, such as sinusitis, central nervous system infections, and urinary tract infections. Table 3 Most Common Gram-Negative Bacteria Isolated during Infection Workups among Wounded Military Personnel1 Number of Isolates Bacterial Organism Combat-related Injury % MDR Noncombat Injury % MDR Afghanistan Theater Pseudomonas aeruginosa 204 10 1 0 Escherichia coli 176 73 3 0 Acinetobacter calcoaceticus baumannii 142 78 2 50 Enterobacter cloacae 137 2 4 0 Stenotrophomonas maltophilia 68 51 0 0 Afghanistan Total 2 1070 33 16 6 Iraq Theater Pseudomonas aeruginosa 7 14 1 0 Acinetobacter calcoaceticus baumannii 7 57 1 0 Escherichia coli 5 60 0 0 Enterobacter cloacae 5 0 0 0 Haemophilus influenza 3 0 1 0 Iraq Total 2 40 23 4 0 MDR – multidrug-resistant 1 Data are collected from all infection work-ups (e.g., wound and blood cultures) among wounded personnel who were transferred from Landstuhl Regional Medical Center (LRMC) to a TIDOS-participating US military treatment facility (MTF) at all levels of care (LRMC and/or U.S. military treatment facilities). Patients often have serially positive cultures; however, an organism was counted only once per patient. An organism was counted at MDR if it was MDR at any time isolated during repeated isolation. 2 Only the top five organisms are reported. Total incorporates all organisms collected during infection workups, so it will be more than the sum of the listed organisms. Table 4 Results of Logistic Regression Models to Evaluate Risk Factors for Any Infectious Complications of Deployment-Related Traumatic Injury Parameter Univariate Odds Ratio (95% CI) P-value Multivariate Odds Ratio (95% CI) P-value Operational theater Afghanistan Reference Reference Iraq 0.56 (0.42–0.75) <0.0001 1.32 (0.87–1.99) 0.193 Combat-related injury 7.13 (5.10–9.99) <0.0001 0.60 (0.32–1.15) 0.124 Branch of Service 1 Army Reference – Marine 1.66 (1.41–1.95) <0.0001 – Other 0.83 (0.62–1.12) 0.220 – Mechanism of Injury Gunshot wound Reference Reference IED blast 3.26 (2.60–4.08) <0.0001 1.84 (1.35–2.49) <0.0001 Non-IED blast 0.99 (0.72–1.37) 0.937 0.91 (0.60–1.37) 0.639 Other 0.45 (0.33–0.62) <0.0001 1.49 (0.84–2.66) 0.174 Injury Severity Score ≤15 Reference Reference 16–25 7.7 (5.9–10.0) <0.0001 1.72 (1.23–2.42) 0.002 ≥ 26 28.5 (23.0–35.4) <0.0001 2.48 (1.81–3.41) <0.0001 RBC transfusion requirements 0/missing units2 Reference Reference 1–9 units 9.19 (7.45–11.33) <0.0001 2.50 (1.89–3.30) <0.0001 10 – 20 units 38.69 (29.57–50.64) <0.0001 5.66 (3.97–8.08) <0.0001 > 20 units 83.58 (59.50–117.41) <0.0001 10.68 (6.73–16.94) <0.0001 Injury Type Closed Reference Reference Open 6.84 (5.21–9.00) <0.0001 1.37 (0.93–2.03) 0.114 MDR Gram-negative Colonization at LRMC admission 2.72 (2.15–3.44) <0.0001 1.39 (0.97–1.99) 0.075 Use of prophylactic antibiotics within 1st 48 hours 6.43 (5.16–8.01) <0.0001 1.42 (1.03–1.97) 0.033 ICU Admission Non-ICU Reference Reference LRMC only 4.65 (3.49–6.19) <0.0001 1.98 (1.41–2.76) <0.0001 U.S. MTFs ± LRMC 13.59 (10.82–17.06) <0.0001 3.80 (2.85–5.05) <0.0001 CI – Confidence Interval; ICU – intensive care unit; IED – improvised explosive device; LRMC – Landstuhl Regional Medical Center; MDR – multidrug-resistant; MTFs – military treatment facilities; RBC – packed red blood cells plus whole blood 1 Due to stepwise selection, the branch of service parameters was not included in the multivariate model. 2 Missing RBC transfusion data are not randomly distributed. Patients with missing RBC data are characterized by lower injury severity scores and shock indices. In addition, the majority of patients with missing RBC data did not sustain a traumatic amputation and were not admitted to the LRMC ICU. A portion of this material was presented at the Advanced Technology Applications for Combat Casualty Care (ATACCC) 2011, August 15–18, 2011, Fort Lauderdale, FL; and the Military Health System Research Symposium (MHSRS), August 12–15 2013, Fort Lauderdale, FL. Disclaimer: The views expressed are those of the authors and do not reflect the official views or policies of the Uniformed Services University of the Health Sciences, Henry M. Jackson Foundation for the Advancement of Military Medicine, Inc., National Institutes of Health or the Department of Health and Human Services, Brooke Army Medical Center, Walter Reed National Military Medical Center, U.S. Army Medical Department, U.S. Army Office of the Surgeon General, the Department of Defense (DoD) or the Departments of the Army, Navy or Air Force. Mention of trade names, commercial products, or organizations does not imply endorsement by the U.S. Government. Author Contribution: I/We certify that all individuals who qualify as authors have been listed; each has participated in the conception and design of this work, the analysis of data (when applicable), the writing of the document, and the approval of the submission of this version; that the document represents valid work; that if we used information derived from another source, we obtained all necessary approvals to use it and made appropriate acknowledgements in the document; and that each takes public responsibility for it. Nothing in the presentation implies any Federal/DOD/DON endorsement. Authors acknowledge that research protocol (IDCRP-024) received applicable Uniformed Services University Institutional Review Board review and approval. REFERNCES 1 Murray CK Hinkle MK Yun HC History of infections associated with combat-related injuries J Trauma 2008 64 3 Suppl S221 S231 18316966 2 Murray CK Epidemiology of infections associated with combat-related injuries in Iraq and Afghanistan J Trauma 2008 64 3 Suppl S232 S238 18316967 3 Tribble DR Conger NG Fraser S Gleeson TD Wilkins K Antonille T Infection-associated clinical outcomes in hospitalized medical evacuees after traumatic injury: trauma infectious disease outcome study J Trauma 2011 71 1 Suppl S33 S42 21795875 4 Aronson NE Sanders JW Moran KA In harm's way: infections in deployed American military forces Clin Infect Dis 2006 43 8 1045 1051 16983619 5 Hospenthal DR Murray CK Andersen RC Bell RB Calhoun JH Cancio LC Guidelines for the prevention of infections associated with combat-related injuries: 2011 update: endorsed by the Infectious Diseases Society of America and the Surgical Infection Society J Trauma 2011 71 2 Suppl 2 S210 S234 21814089 6 Eardley WG Brown KV Bonner TJ Green AD Clasper JC Infection in conflict wounded Philos Trans R Soc Lond B Biol Sci 2011 366 1562 204 218 21149356 7 Yun HC Blackbourne LH Jones JA Holcomb JB Hospenthal DR Wolf SE Infectious complications of noncombat trauma patients provided care at a military trauma center Mil Med 2010 175 5 317 323 20486502 8 United States Department of Defense OIF/OEF Casualty Update accessed June 1, 2015 Available at http://www.defense.gov/news/casualty.pdf 9 Krueger CA Wenke JC Ficke JR Ten years at war: comprehensive analysis of amputation trends J Trauma Acute Care Surg 2012 73 6 Suppl 5 S438 S444 23192067 10 Hoencamp R Idenburg FJ Hamming JF Tan EC Incidence and epidemiology of casualties treated at the Dutch Role 2 enhanced medical treatment facility at multi national base Tarin Kowt, Afghanistan in the period 2006–2010 World J Surg 2014 38 7 1713 1718 24481991 11 Penn-Barwell JG Bennett PM Kay A Sargeant ID on behalf of the Severe Lower Extremity Combat Trauma Study Group Acute bilateral leg amputation following combat injury in UK servicemen Injury 2014 45 7 1105 1110 24598278 12 Eastridge BJ Jenkins D Flaherty S Schiller H Holcomb JB Trauma system development in a theater of war: experiences from Operation Iraqi Freedom and Operation Enduring Freedom J Trauma 2006 61 6 1366 1372 discussion 72–73. 17159678 13 Eastridge BJ Costanzo G Jenkins D Spott MA Wade C Greydanus D Impact of Joint Theater Trauma System initiatives on battlefield injury outcomes Am J Surg 2009 198 6 852 857 19969141 14 Murray CK Wilkins K Molter NC Li F Yu L Spott MA Infections complicating the care of combat casualties during Operations Iraqi Freedom and Enduring Freedom J Trauma 2011 71 1 Suppl S62 S73 21795880 15 Petersen K Riddle MS Danko JR Blazes DL Hayden R Tasker SA Trauma-related infections in battlefield casualties from Iraq Ann Surg 2007 245 5 803 811 17457175 16 Scott P Deye G Srinivasan A Murray C Moran K Hulten E An outbreak of multidrug-resistant Acinetobacter baumannii-calcoaceticus complex infection in the US military health care system associated with military operations in Iraq Clin Infect Dis 2007 44 12 1577 1584 17516401 17 Centers for Disease Control and Prevention Acinetobacter baumannii infections among patients at military medical facilities treating injured U.S. service members, 2002–2004 MMWR Morb Mortal Wkly Rep 2004 53 45 1063 1066 15549020 18 Champion HR Holcomb JB Lawnick MM Kelliher T Spott MA Galarneau MR Improved characterization of combat injury J Trauma 2010 68 5 1139 1150 20453770 19 Centers for Disease Control and Prevention CDC/NHSN Surveillance Definitions for Specific Types of Infections accessed June 1, 2015 Available at http://www.cdc.gov/nhsn/pdfs/pscmanual/17pscnosinfdef_current.pdf 20 Division of Healthcare Quality Promotion The National Healthcare Safety Network (NHSN) Manual. Patient safety component Protocol multidrug-resistant organism (MDRO) and Clostridium difficile-associated disease (CDAD) module Atlanta, GA Centers for Disease Control and Prevention accessed December 15, 2015 Available at http://www.cdc.gov/nhsn/PDFs/pscManual/12pscMDRO_CDADcurrent.pdf 21 Murray CK Wilkins K Molter NC Yun HC Dubick MA Spott MA Infections in combat casualties during Operations Iraqi and Enduring Freedom J Trauma 2009 66 4 Suppl S138 S144 19359957 22 Dunne JR Hawksworth JS Stojadinovic A Gage F Tadaki DK Perdue PW Perioperative blood transfusion in combat casualties: a pilot study J Trauma 2009 66 4 Suppl S150 S156 19359959 23 Dunne JR Riddle MS Danko J Hayden R Petersen K Blood transfusion is associated with infection and increased resource utilization in combat casualties Am Surg 2006 72 7 619 625 discussion 25–26. 16875084 24 Hospenthal DR Crouch HK English JF Leach F Pool J Conger NG Multidrug-resistant bacterial colonization of combat-injured personnel at admission to medical centers after evacuation from Afghanistan and Iraq J Trauma 2011 71 1 Suppl S52 S57 21795879 25 Weintrob AC Murray CK Lloyd B Li P Lu D Miao Z Active surveillance for asymptomatic colonization with multidrug-resistant gram negative bacilli among injured service members - a three year evaluation MSMR 2013 20 8 17 22 26 Brown KV Murray CK Claspar JC Infectious complications of combat-related mangled extremity injuries in the British military J Trauma 2010 69 suppl 1 S109 S115 20622604 27 Ficke JR Eastridge BJ Butler F Alvarez J Brown T Pasquina P Dismounted complex blast injury report of the Army Dismounted Complex Blast Injury Task Force J Trauma Acute Care Surg 2012 73 6 Suppl 5 S520 S534 28 Tintle SM Shawen SB Forsberg JA Gajewski DA Keeling JJ Andersen RC Reoperation after combat-related major lower extremity amputations J Orthop Trauma 2014 28 4 232 237 24658066 29 Rodriguez C Weintrob AC Shah J Malone D Dunne JR Weisbrod AB Risk factors associated with invasive fungal infections in combat trauma Surg Infect (Larchmt) 2014 15 5 521 526 24821267 30 Rodriguez CJ Weintrob AC Dunne JR Weisbrod AB Lloyd BA Warkentien T Clinical relevance of mold culture positivity with and without recurrent wound necrosis following combat-related injuries J Trauma Acute Care Surg 2014 77 5 769 773 25494431 31 Weintrob AC Weisbrod AB Dunne JR Rodriguez CJ Malone D Lloyd BA Combat trauma-associated invasive fungal wound infections: epidemiology and clinical classification Epidemiol Infect 2015 143 1 214 224 24642013 32 Evriviades D Jeffery S Cubison T Lawton G Gill M Mortiboy D Shaping the military wound: Issues surrounding the reconstruction of injured servicemen at the Royal Centre for Defence Medicine Philos Trans R Soc Lond B Biol Sci 2011 366 1562 219 230 21149357 33 Conger NG Landrum ML Jenkins DH Martin RR Dunne JR Hirsch EF Prevention and management of infections associated with combat-related thoracic and abdominal cavity injuries J Trauma 2008 64 3 Suppl S257 S264 18316970 34 Murray CK Obremskey WT Hsu JR Andersen RC Calhoun JH Clasper JC Prevention of infections associated with combat-related extremity injuries J Trauma 2011 71 2 Suppl 2 S235 S257 21814090 35 Segrt B Particularities of the therapeutic procedures and success in treatment of combat-related lower extremities injuries Vojnosanit Pregl 2014 71 3 239 244 24697009 36 Lesho EP Waterman PE Chukwuma U McAuliffe K Newmann C Julius MD The Antimicrobial Resistance Monitoring and Research (ARMoR): The US Department of Defense response to escalating antimicrobial resistance Clin Infect Dis 2014 59 3 390 397 24795331
PMC005xxxxxx/PMC5123799.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 9306780 2536 Health Econ Health Econ Health economics 1057-9230 1099-1050 27870306 5123799 10.1002/hec.3375 NIHMS821561 Article Health disparities by income in Spain before and after the economic crisis Coveney Max [email protected] García-Gómez Pilar [email protected] Van Doorslaer Eddy [email protected] Van Ourti Tom [email protected] a Erasmus School of Economics, Erasmus University Rotterdam, P.O. Box 1738, 3000 DR Rotterdam, the Netherlands; Tinbergen Institute, and NETSPAR b Institute for Health Policy and Management, Erasmus University Rotterdam, P.O. Box 1738, 3000 DR Rotterdam, the Netherlands 7 10 2016 11 2016 01 11 2017 25 Suppl 2 141158 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Little is known about what the economic crisis has done to health disparities by income. We apply a decomposition method to unravel the contributions of income growth, income inequality and differential income mobility across socio-demographic groups to changes in health disparities by income in Spain using longitudinal data from the Survey of Income and Living Conditions (SILC) for the period 2004–2012. We find a modest rise in health inequality by income in Spain in the five years of economic growth prior to the start of the crisis in 2008, but a sharp fall after 2008. The drop mainly derives from the fact that loss of employment and earnings has disproportionately affected the incomes of the younger and healthier groups rather than the (mainly stable pension) incomes of the over 65s. This suggests that unequal distribution of income protection by age may reduce health inequality in the short run after an economic recession. economic crisis health inequality Spain 1. Introduction The Great Recession that started in 2008 was for most OECD countries the worst economic contraction since the 1930s (Jenkins et al, 2013). While falling incomes and rising unemployment have been the most visible consequence of the crisis, an additional concern is whether any effects have been unequally spread across the income distribution. The importance of studying inequalities, both in income and other dimensions, is widely appreciated. While the European Union has targeted health inequality reduction as a key policy goal and warned of “the negative consequences for health, social cohesion and economic development if health inequalities are not effectively tackled” (European Commission, 2009, p. 5), the crisis has interfered with the execution of some of these policies (European Commission, 2013). The aim of this paper is to examine what has happened to the social gradient in health before and after the crisis. We focus on Spain, one of the EU countries confronted with a severe economic recession, and employ a decomposition method that has been used to examine the evolution of income, health and inequality during a period of rapid growth in China (Baeten et al, 2013). We use new EU Statistics on Income and Living Conditions (SILC) panel data spanning the period 2004–2012, including both a period of substantial economic growth (2004–2007) as well as of recession (2009–2012). We examine the extent to which the evolution of health disparities by income was associated with changes in income growth, in income inequality and in differential income mobility of various socio-demographic groups. The economic growth pattern in Spain since the mid–1990s can be summarized by a pre-crisis trend and a post-crisis trend. Figure 1 shows the evolution of unemployment (right axis), and GDP growth and real average annual wage growth (left axis) between 2002 and 2013. Prior to the crisis, the country experienced extended economic expansion with real GDP growing at approximately 3% per year and unemployment falling below 8% in 2007. Despite this extended period of labour demand, wage growth was minimal (Carrasco et al, 2011). The effects of the global financial crisis become obvious in Spain beginning in 2008: when GDP growth fell from 3% to below −3% between the first quarter of 2008 and the first quarter of 2009, while unemployment roughly doubled in the same period. Youth unemployment became particularly high, with unemployment in the 15–24 age group doubling between 2007 and 2009 to stand at more than 35% (OCED, 2009). Males, who were overrepresented in highly cyclical forms of employment, were hit disproportionally (Rica & Rebollo-Sanz, 2015). Jenkins et al (2013) reaches similar conclusions, noting that the largest fall in employment in Spain during the crisis period was concentrated among young people under the age of 25, especially young men.2 Particularly important to understanding both the boom and bust years in Spain is the expansion and collapse of the housing market. The nominal house price per square meter in Spain tripled between 1997 and 2007 (Bonhomme & Hospido, 2012). Parallel to this housing boom was an expansion in the construction sector. Between 1998 and 2008 the share of construction in Spain’s GDP increased by 4 percentage points to 10.7% (Gonzalez & Ortega, 2009). From 1997 to 2006 the share of construction in male employment rose from 14% to more the 20% (Bonhomme & Hospido, 2012). However, in 2008 as the effects of the subprime mortgage crisis in America spread and Europe began to enter recession, the Spanish housing sector crashed. Most of the sharp rise in unemployment in Spain was due to the collapse of this sector. From the first quarter of 2008 till the last quarter of 2009, construction experienced a 20% per annum drop in employment (Bentolila et al, 2012).3 Alongside the changes in average levels of income there have been changes in its distribution. In general, income inequality in Spain has followed a counter-cyclical pattern – during boom years income inequality decreased while during bust years it rose (Lacuesta and Izquierdo, 2012; Carrasco et al, 2011; Pijoan-Mas & Sanchez-Marcos, 2010). Bonhomme and Hospido (2012) show that the housing and construction sector is once again key to understanding these trends as the construction sector is one of the main employers of young, uneducated, relatively disadvantaged groups of (usually) men (Aparicio, 2010). A separate question is the extent to which the Great Recession has affected health. There is a large literature that documents that worsening economic conditions are associated with reduced mortality (e.g. Ruhm, 1996; Stuckler et al, 2009), but more recent evidence is mixed. Rhum (2015) suggests that the relationship may be disappearing over time, while others have found that mortality trends, except for suicides, continued to improve during the recent European-wide recession (Regidor et al, (2014); OECD, 2014). There is little evidence on the evolution of health inequalities during the recent crisis.4 Most observers appear to assume that it will widen existing gaps following reductions in welfare spending that increase the vulnerability of those with lower education levels, who are also more likely to be unemployed (European Commission, 2013). One strand of literature has focused on the comparison of income-related health inequalities (IRHI) across countries and over time. Van Doorslaer and Koolman (2004), for instance, documented the variation in degrees of IRHI for 13 EU countries in 1996 and showed that IRHI tends to be larger in countries with larger income inequality, but also that the relative income position of Europeans that are not working and not in good health, like the retired and the disabled, was critical. Van Ourti et al (2009) decomposed the evolution of IRHI between 1994 and 2001 for the same 13 EU countries. They found that the income elasticity of health was crucial for understanding the evolution of IRHI, although the period considered was one of economic growth for most European countries. An extended version of this decomposition was used by Baeten et al (2013) to decompose the evolution of IRHI in China into the contributions of various factors like income growth, income inequality, and income mobility, as well as a number of regional-demographic factors associated with health. They found that the substantial rise in IRHI over the period of double-digit income growth (1991–2006) was associated with rising income inequality, but especially with the adverse health and income experience of older women lacking pension or other social protection. It is this decomposition method that we use in this paper. Our findings indicate the following: inequality in health by income was slightly rising before the crisis, but started falling sharply after 2009 when the recession hit Spain. The main reason for this reversal is the differential effect of the crisis on the incomes of young and old Spaniards: while pensioner incomes were relatively shielded against the erosion in the post-crisis years, this does not hold for the incomes of younger groups. Loss of employment and of earnings in employment meant that these relatively healthier groups moved downwards in the income ranking, thereby lowering the association between health and income rank. As a result, IRHI in 2012 was lower than in the years prior to 2009, a somewhat surprising by-product of an otherwise discomforting period in recent Spanish history. 2. Decomposing the evolution of IRHI with a balanced cohort We use the decomposition method of Baeten et al. (2013) to estimate the evolution of IRHI and to shed light on the relative importance of (1) income growth, (2) the evolution of income inequality, (3) income mobility, and (4) the evolution in non-income factors (such as demographics) that are associated with health. This section describes the decomposition approach for a balanced cohort of individuals that we observe at the start (period 1) and end (period 2) of a time interval. 2.1 Choice of health inequality index We use the corrected concentration index (CC) (Erreygers, 2009) because it satisfies the mirror condition and it is insensitive to equal health additions (cf. absolute inequality) (Erreygers and Van Ourti 2011). When health is bounded between 0 and 1, it can be written as: (1) CC(ht∣yt)=8n2∑i=1nzithit where ht and yt are the health and income distribution in period t = 1,2, hit describes the health of individual i and zit is a weight that depends on the income rank of individual i. This weight takes the value 0 for the individual with median income, and increases (decreases) linearly for individuals with higher (lower) than median income levels.5 Descriptive model for health We use a simple descriptive6 model that links health linearly and additively to its associated factors: (2) hit=α+θ(yit)+x′itβ where α is an intercept parameter; θ(yit) is a non-linear function of income; xit represents a vector of K non-income variables (e.g. demographics), and β is its associated parameter vector.7 It is important to allow for a very general functional form for θ( ) since the actual functional form will largely determine the relative importance of the contribution of (a) income growth and (b) income inequality in our decomposition approach. 2.2 Evolution of IRHI over time Our interest lies in decomposing changes in IRHI. Combining equation (2) and (1) leads to8: (3a) CC(h2∣y2)-CC(h1∣y1)=8n2[∑i=1nzi2hi2-∑i=1nzi1hi1] (3b) =8n2∑i=1n{[zi2θ(yi2)-zi1θ(yi1)]+β[zi2x′i2-zi1x′i1]} Equation (3a–b) shows that we can disentangle the change in IRHI into a part due to changes in the association between the income rank and the income effect (zi2 θ(yi2) − zi1θ(yi1)) and a part due to changes in the association between the income rank and the non-income factors (zi2x′i2 − zi1x′i1). In order to isolate the contributions of (a) income growth, (b) the evolution of income inequality, (c) income mobility, and (d) the evolution in non-income factors, we construct two hypothetical health states in period 2 using equation (2) – health under average income growth ( hi2ag) and health under no income growth ( hi2ng). For the former, we calculate an individual’s health in period 2 in the scenario that everyone’s income changed proportionally to the average income gain (or loss) between period 1 and period 2. We denote this income as yi2ag. For the latter, we estimate an individual’s health in period 2 in the scenario that there was no income change between period 1 and period 2 ( yi2ng). In each scenario we allow all non-income variables to change as they actually did. 2.3 Decomposition method Given that CC(h2ag∣y2ag)=CC(h2ag∣y1) and yi2ng=yi1, the change in IRHI can be expressed as: (4) CC(h2∣y2)-CC(h1∣y1)=CC(h2∣y2)-CC(h2ag∣y1)︸incomeinequality&mobility+CC(h2ag∣y1)-CC(h2ng∣y1)︸averageincomegrowth+CC(h2ng∣y1)-CC(h1∣y1)︸non-incomefactors which can be further disentangled as the sum of 4 terms (note that zi2ag=zi2ng=zi1): (5) CC(h2∣y2)-CC(h1∣y1)=8n2∑i=1n{zi1[θ(yi2ag)-θ(yi1)]︸averageincomegrowth+[zi2θ(yi2)-zi1θ(yi2ag)]︸incomeinequality+(zi2-zi1)(∑k=1Kβkxi2k)︸incomemobility+zi1[∑k=1Kβk(xi2k-xi1k)]︸non-incomefactors} Equation (5) shows that the evolution of IRHI can be written as the sum of (a) average income growth, (b) the evolution of income inequality, (c) income mobility, and (d) the evolution in non-income factors. The first term, average income growth, captures the change in IRHI when everyone’s income changes in proportion to the average income change. As all incomes grow proportionally, there is no change in the rankings (zit’s). Therefore this term captures whether the health responsiveness to proportional income changes ( θ(yi2ag)-θ(yi1)) is, on average, larger or smaller for those with lower (negative zi1) versus higher incomes (positive zi1) in period 1. If the health responsiveness is larger for the initially richest part of the population, then this term will be positive. The sign (and magnitude) of this term depends on the functional form of θ( ), but also on whether incomes have increased or decreased on average. The second term captures the evolution of income inequality – that is, the health difference attributed to the difference between the true income in the second period and the income under the scenario of average income growth ( θ(yi2)-θ(yi2ag)). If the health returns from income growth are increasing with income (θ′ (.) >0), if there is no income re-ranking (zi2 = zi1) and if – relative to the average income growth scenario – the rich become richer while the poor loose, then the second term will be positive. In a scenario with income re-ranking, one cannot a priori assign a direction to term 2. Term 3 – ‘income mobility’ – captures the association between income re-ranking (zi2 − zi1) and the non-income factors in the second period, weighted by the βk coefficients. One can further decompose term 3 into separate contributions for each non-income variable since term 3 is additively separable. In our empirical application, the non-income variables are dummy variables. In this case, the contribution of each non-income dummy can be large (compared to the reference category) because (a) health is considerably higher or lower among the individuals belonging to the non-income dummy (βk), (b) income re-ranking is substantial for these individuals ((zi2 − zi1), and/or (c) a substantial share of the sample belongs to this non-income dummy ( ∑i=1nxi2k). Term 4 measures the association between changes in non-income factors and initial income ranks. If the non-income factors include age and location, then term 4 isolates the effect of ageing and migration on the change of IRHI. For example, if many people with high initial income ranks migrate to a location which is associated with better health then this term will be positive. In what follows we refer to the terms 1, 2, 3 and 4 as income growth, income inequality, income mobility, and non-income factors. 3. Data and empirical implementation EU-SILC data We use 6 rounds of the Spanish EU-SILC dataset spanning 2004–2012. It includes the period before and during the financial and economic crisis that affected Spain from 2009 onwards. As the EU-SILC dataset is set up as a rotating panel every year between 2004 and 2009, a new random sample is drawn and followed for 4 years, after which it is dropped. We use the term ‘rotation group’ for each of these random samples. For example, the first rotation group is drawn in 2004 and lasts till 2007; the second covers 2005–2008; and the sixth and last rotation group covers 2009–2012. Hence, for the full period of 2004–2012 we have 6 rotation groups in total, and these constitute different and independent samples.9 In total, we have 122,592 observations (see table I for more details). 3.1 Key variables The two main variables of interest are self-assessed health (SAH) and household income. The SAH responses derive from the question: “How is your health in general? Is it: (1) very good, (2) good, (3) fair, (4) bad, (5) very bad?” As our income variable we use total disposable household income during the previous 12 months. We adjust for household size and inflation by dividing by the square root of household size10 and by applying the Spanish CPI index with base year 2012. We remove observations with negative incomes.11 3.2 Estimating the health model We estimate the model for health in equation (2) using an interval regression (with threshold values imposed as in van Doorslaer and Jones (2003)), as the CC computation requires a health indicator measured on a cardinal scale.12 The predicted values are used as our main health indicator, and can be interpreted as predicted health utility index (HUI) scores (Van Doorslaer and Jones, 2003).13 We use a second degree polynomial for the income function φ(yit) as it is a parsimonious functional form that is sufficiently flexible to avoid predetermining the effect of proportional income changes on health.14 The remaining independent variables are dummy variables for each region in Spain and age category dummies for both males and females. The regions are listed in Table IV(b). Age is categorised into the following groups: 16 to 26 years, 26 to 36 years, 36 to 46 years, 46 to 56 years, 56 to 66 years, 66 to 76 years, 66 to 76 years, and more than 76 years of age. 3.3 Empirical implementation of decomposition method Because of the rotating design of EU-SILC we cannot directly compare IRHI measured for the same individuals in 2004 and 2012. This complicates both the implementation of the decomposition and the estimation of the empirical health model. The decomposition requires at least two observations of the same individual over time. We apply the decomposition to each of the 6 rotation groups separately and within each rotation group to a balanced panel only.15 While we calculate the decomposition for each of the 6 rotating panels, we only present three of these: a before crisis group: 2004–2007; a group covering both before and when the crisis occurs: 2007–2010; and a group that covers the crisis period: 2009–2012.16 We first pool the data from all 6 rotation groups and run the interval regression model described above.17 We remove the individuals belonging to the top 1% of incomes (calculated on the full pooled sample) as these observations have a disproportionate effect on the functional form of income.18 The estimated parameters of the pooled model are then used to decompose the 3 rotation groups which span the entire 2004–2012 period, leaving us with the observations per rotation group as shown in table II. Each of the rotating panels uses a different base year. In the 2004–2007 rotation group, we first compare the change in IRHI for 2004–2005, then 2004–2006, then 2004–2007. We next take the second rotation group (which spans 2007–2010) and compare the change in IRHI between 2007 and each following year. For the 2009–2012 rotation group, 2009 is the base year. In total there are then 9 changes of IRHI to be decomposed. We use the sample weights of the first year of each rotating panel provided with the EU-SILC data. In the interval regression model, we also allow for robust standard errors and cluster at the individual level. Statistical inference of the decomposition method is obtained after bootstrapping the entire procedure with 1,500 replications. The bootstrap sampling is bias-corrected, and clustered at the primary sampling unit of the EU-SILC. 4. Results 4.1 Summary statistics Table III displays variable means for each wave of rotation group 1, 4 and 6. Panel (a) includes variables most important to the analysis, whereas panel (b) provides additional background information on the labour market. The health variable refers to the predicted HUI score. Income is rising in each successive year for rotation group 1, as well as rotation group 4 until 2009. As income refers to the last 12 months, the drop observed in 2010 refers to an income fall in 2009, during which Spain was fully immersed in the economic crisis. In rotation group 6 income falls in each wave compared to the last. The effect of the crisis is also visible in the occupational category changes. The proportion unemployed in 2009 almost doubles from the previous year to approximately 11%. In subsequent years the proportion of employed individuals decreases every year. This does not appear to be due to ageing and retiring individuals; while the proportion of retirees does increase slightly, it is the unemployed category that shows the sharpest increase. Income inequality, as measured by the Gini coefficient was rather stable, although opposite trends can be observed before and after 2009. Income inequality appears to have been slightly falling during the “boom” years, and began to increase once the crisis started. This is in line with the findings of others, such as Jenkins et al (2013). Figure 2, with the CC per year for each rotation group, shows that IRHI has not been stationary over the sample period.19,20 Until 2009, there is a slightly significant upward trend.21 Since the beginning of the crisis, however, IRHI fell quite steeply. This is confirmed by comparatively large and significant falls in IRHI in the final two rotation groups. Column 1 of table IV a–b shows the coefficients from the interval regression for the age-sex and region dummies.22,23 Note that older age groups consistently report lower health than younger. Regional health differences, by contrast, are very small. Figure 3 shows decomposition results for rotation groups 2004–2007, 2007–2010 and 2009–2012 with 95% confidence intervals. 2004–2007 results Between 2004 and 2007, IRHI rose significantly. Panel (1) of figure 3 shows that income growth is important in understanding this rise. The income growth term, although small, indicates that health responsiveness to proportional income growth was larger for those with higher income in 2004. Despite being the largest term in all years, income mobility only becomes significant in the 2004–2007 comparison. This implies that income re-ranking occurring prior to 2007 was not systematically related to age, gender or location, while the elderly were on average (and just borderline significantly) more likely to experience negative income re-ranking between 2004 and 2007.24 As the elderly combine this move down the income ladder with the lowest predicted health, this led to a rise in IRHI. The evolution of income inequality and the non-income factors are unimportant for the IRHI change in this period. 2007–2010 Results IRHI grew significantly between 2007 and 2008, but returned in the subsequent two years to its 2007 level. The decreasing trend turns out to be almost entirely driven by the changing association between the age dummies and the income rank, while region is relatively unimportant (see income mobility term in panel (2) of figure 3 and panel (1) of figure 4). Closer inspection reveals that it is mainly influenced by the older, unhealthier, age groups. While initially, during the period of income growth prior to 2008, the elderly were falling in income rank, there is a reversal after 2008. Panel (2) of figure 4 shows that this was especially true for those over 75. The income rank of the older age groups, with poorer health, increased contributing to the fall in IRHI. Also significant between 2007 and 2010 is the contribution of income inequality. This suggests that the health effects of income gains – over and above proportional income growth – led to a rise in IRHI. Income growth is positive in each wave and remains small but significant. As average income falls in the final 2010 wave, so does the magnitude of income growth. 2009–2012 Results The final 4-year rotation group of the EU-SILC entirely reflects the crisis years. This is the period in which the largest drop in IRHI occurs and the trends observed in the 2007–2010 decomposition also emerge here. The significant fall in IRHI is primarily due to income mobility, which is largest in magnitude and significant in all years. Panels (1) and (2) of figure 5 demonstrate that it is the experience of certain older age groups – men and women aged 66 and above – which is the largest contributor to the decrease. By contrast, the younger and healthier individuals have a small but positive contribution. This leads to a fall in IRHI as those with poorer health became relatively richer. The 2009–2012 decomposition also reveals that both income growth and income inequality are significant drivers in the change of IRHI. The negative sign of income growth reflects the fact that had the average income fall between 2009 and 2012 been applied to everyone, those with high incomes would have had a larger fall in health than those with low incomes. Income inequality is positive however, indicating that the fall in income was disproportionately felt by the poor. Still, the overall effect of these terms relating to health responsiveness to income is small compared to income mobility. 5. Discussion The decomposition results reveal two very different trends in IRHI before and after the crisis. Prior to 2009 there was a trend of increasing inequality which was mostly driven by income growth but also by income mobility, with the elderly slightly moving down on the income ranks. After the start of the 2008 financial crisis we observe a sharp fall in IRHI. Income mobility is the main driver of this change: young and middle-aged healthier groups experienced a greater income drop, while on average the incomes of the elderly were less affected. This caused shifts in the income ranks in favour of the older, less healthy group, leading to a decrease in IRHI. Further decomposing the contribution of each regional-demographic group to income mobility reveals the relative importance of three distinct underlying mechanisms. Indeed, each sub-term depends on three elements – the partial association between the group and health (βk), the number of individuals in that particular group, and the changes in income ranks between the two periods for these individuals (zi2 − zi1). Any changes in income mobility result from some combination of these elements. Table IV(a) and (b) presents results for each of these three elements per demographic group and region. Column 1 reports the interval regression coefficients and Columns 2 to 7 the percentage shares of each regional-demographic group for the first and final years of each rotation group, while columns 8 to 10 report the income re-ranking for each regional-demographic group. The results for income re-ranking are obtained by running a simple no-constant OLS using the regional-demographic variables as explanatory variables for the change in individual z-scores between the two periods.25 A positive coefficient implies a rise in income rank between the two periods. Table IV(a) confirms that the income re-ranking of the elderly, in particular after the onset of the crisis, is most important for understanding changes in IRHI due to income mobility. Between 2004 and 2007, there was little re-ranking taking place, although the negative coefficients for the elderly indicate that, if anything, the elderly were slightly losing relative to young. In the final rotation group however the coefficients of the 65+ have become highly significant and switched sign. This, combined with the comparatively large negative coefficient of the 65+ in the health regression, and the sizable and increasing number of individuals in this category, leads to a large fall in income related health inequality.26 The primary reason that the elderly’s incomes were better protected during the crisis appears to be the old-age pension system. In Spain, the vast majority of pensioners receive their incomes from contributory pensions based on earnings prior to retirement (OECD, 2013). As a consequence, current economic conditions have little immediate effect on retiree incomes. Moreover, any potential changes to pension benefits are delayed by political processes and reforms are not applied retrospectively. Thus, in spite of a series of reforms that took place during the last decade in Spain, existing pensioners’ incomes have remained relatively untouched.27 5.1 Role of labour market status and occupation Our results thus far indicate that the pre-crisis rise and post-crisis fall in IRHI were largely related to differential income mobility. In this section we explore how income mobility is associated with labour market status and occupation.28 Again, we use OLS regression to analyse the correlation between labour market status and changes in the income ranks (see Table V). Prior to the crisis (column 1), we do not see that the changes in the z-scores are not significantly different across labour market states (except for the self-employed), but during the crisis years (column 2) every group, except the employed and unemployed, on average, moves up in the income ranking. Interestingly, the self-employed, the group with the greatest drop in the income ranks between 2004 and 2007, has gained the most after 2009. The retired and disabled groups also experienced gains, both of which receive “sticky” benefits that were not immediately affected by current economic conditions. Columns 3 and 4 repeat the exercise for employed individuals only to examine differences between occupations for those employed in the first wave of each rotation group, (in 2004 and 2009). We do not observe large differences in re-rankings across occupations in the pre-crisis years (column 3), but during the crisis years (column 4), all occupation groups fell relative to the Manager group. The largest significant drop occurs in the Elementary Occupation group, which contains manufacturing, mining and construction labourers. These findings are in line with previous evidence showing that it was those in the construction sector whose incomes fell the most after the onset of the crisis in Spain (Bentolila et al, 2012). 6. Conclusion We examine the evolution of IRHI in Spain both before and during the Great Recession, and decompose IRHI changes into four separately interpretable terms, reflecting the contribution of (i) income growth, (ii) income inequality changes, (iii) income mobility and (iv) changes in non-income terms. Our findings are as follows. First, while our approach only informs on health changes resulting from changes in the explanatory variables, our findings suggest that health inequality by income in Spain was rising in the four years of economic growth prior to the start of the crisis, but this rise was modest. By contrast, after 2008, it started falling at a faster pace. Second, there appear to be two reasons for this modest rise in IRHI prior to 2008 – income growth and to a lesser extent income mobility – suggesting that the health benefits associated with income growth were disproportionately concentrated amongst the already rich; and that the elderly, often in poorer health, fell slightly on the income ranks leading to increased disparities. Third, the falling health disparities by income mainly derived from the uneven distribution of the income consequences of the crisis. The incomes of younger, healthier groups were affected much more by rising unemployment than the incomes of the over 65s which mainly consist of pensions. Since contributory pensions are ‘sticky’ in Spain and therefore relatively unaffected in the first years of the crisis, pensioners improved their relative position in the income distribution substantially. Fourth, we study the role of labour market participation status and occupation and find that, in line with others studies, it was primarily the income deterioration of the unemployed and the employed, especially those in the construction sector, that was responsible for their fall in the income ranking. While the great recession caused a substantial deterioration in income, health policy makers can perhaps take solace in the fact that the Spanish pension system - at least in the short run - appears to have shielded some of the most vulnerable individuals. The EU has devoted special attention to reducing health inequalities and for decades countries have attempted to reduce pervasive and persistent health disparities in periods of economic growth. Ironically, our study reveals that the recent crisis has perhaps done more to cut back inequality than many years of pro-poor health policy making. This may be somewhat surprising, given the initial predictions of many observers and in light of media reports of crises hitting the most vulnerable population segments first. But in reality it can be understood as a logical outcome in the presence of sticky pensions and other welfare benefits in the immediate aftermath of a financial crisis. While employment rates and earnings levels are less protected in the short run, also pension and other benefits may be curtailed in the longer run as a consequence of fiscal constraints. It also remains to be seen whether the post-crisis evolution of income-related health inequality has been similar in other European countries with less sticky pension and other benefits. We would like to acknowledge the helpful comments and suggestions given by two anonymous referees. Pilar García-Gómez is a Postdoctoral Fellows of the Netherlands Organisation for Scientific Research—Innovational Research Incentives Scheme—Veni. Eddy Van Doorslaer and Tom Van Ourti acknowledge support from the National Institute on Ageing, under grant R01AG037398. We thank Eurostat for providing access to the EU-SILC data. We thank Dennis Petrie, Helena Hernández, seminar participants at Erasmus University Rotterdam, XXXIII Jornadas AES in Santander, 2015 iHEA Congress in Milan, Workshop on consequences of the economic crisis on health and health care systems in Madrid for further useful comments and suggestions. The usual caveats apply, and all remaining errors are our responsibility. 10. Appendices Table AI IRHI change within rotation groups Rotation Group IRHI Change 2004–2007 2004–2005 2004–2006 2004–2007 0.0019 0.0030 0.0047* 2005–2008 2005–2006 2005–2007 2005–2008 0.0031* 0.00071 0.0041* 2006–2009 2006–2007 2006–2008 2006–2009 0.0019* 0.0016 0.0027* 2007–2010 2007–2008 2007–2009 2007–2010 0.00447* 0.0037* −0.00001 2008–2011 2008–2009 2008–2010 2008–2011 0.00193 −0.00273 −0.00478* 2009–2012 2009–2010 2009–2011 2009–2012 −0.003* −0.0038* −0.0064* Source: EU-SILC, * p < 0.05. Table AII Decomposition results for rotation group 1 2004 –2005 2004–2006 2004–2007 IRHI change 0.00196 0.00339 0.00483 Income growth 0.00011 0.00065 0.00084 Income inequality −0.00016 0.00003 −0.00008 Income mobility 0.0014 0.00216 0.00354 Non-income factors 0.00061 0.00073 0.00094 Individual Contribution Income mobility Non-inc. factors Income mobility Non-inc. factors Income mobility Non-inc. factors M 16–26 −0.00001 0 −0.00002 0 0.00001 0 F 26–36 −0.00022 0.00018 −0.00012 0.00012 −0.00004 0.00025 M 26–36 0.00002 0.00016 −0.00026 0.00036 −0.00016 0.00047 F 36–46 −0.00006 0.00008 0.00011 −0.00014 −0.00051 −0.00007 M 36–46 −0.00008 −0.00028 −0.00025 −0.00063 0.00015 −0.00076 F 46–56 0.00076 −0.00065 0.00008 −0.00069 0.00013 −0.00102 M 46–56 −0.00032 0.00013 0.00015 0.00034 −0.00019 0.00026 F 56–66 0.00014 −0.00064 0.00027 -0.00097 0.00075 -0.00158 M 56–66 0.00062 −0.00068 −0.00033 −0.00145 0.00097 −0.00177 F 66–76 −0.00022 0.00021 0.00102 −0.00032 0.00038 −0.00113 M 66–76 0.00045 −0.00015 0.00094 −0.00043 0.00097 −0.00083 F 75+ 0.00012 0.00165 −0.00027 0.00313 0.00038 0.00495 M 75+ 0.00023 0.0006 0.00049 0.00143 0.00052 0.00215 Galicia −0.00012 0 0.00043 0 0.00028 0.00004 Asturias −0.00001 0 −0.00001 0 −0.00002 0 Cantabria 0.00002 0 0.00002 0 0.00002 0 País Vasco 0 0 0 0 0 0.00001 Navarra 0.00001 0 0.00001 0 0 0 La Rioja 0 0 0 0 0 0 Aragón 0 0 0 0 0 0 Castilla y León −0.00002 0 −0.00001 0 −0.00002 0 Castilla-La Mancha 0.00001 0 0 0 0.00001 0 Extremadura 0 0 −0.00002 0 −0.00002 0 Cataluña 0 0 0 0 −0.00001 0 Comunidad Valenciana 0.00001 −0.00001 0.00002 −0.00001 −0.00012 −0.00001 Baleares −0.00003 −0.00001 −0.00005 0 −0.00002 0.00001 Andalucía −0.00021 0.00001 −0.0002 −0.00001 −0.00021 0 Murcia 0.00015 0 −0.00002 −0.00001 0.0001 −0.00001 Ceuta y Melilla −0.00002 0 −0.00003 0 −0.00002 0 Canarias 0.00018 0 0.00023 −0.00002 0.00021 −0.00002 Madrid, F 16–26 and Employed used as control groups Table AIII Decomposition results for rotation group 4 2007–2008 2007–2009 2007–2010 IRHI change 0.00487 0.00413 0.00014 Income growth 0.00078 0.00135 0.00078 Income inequality 0.00017 0.00081 0.00205 Income mobility 0.00374 0.00215 −0.00217 Non-income factors 0.00018 0.00006 0.00007 Individual Contribution Income mobility Non-inc. factors Income mobility Non-inc. factors Income mobility Non-inc. factors M 16–26 0 −0.00002 0.00001 −0.00001 0.00001 −0.00002 F 26–36 −0.00015 −0.0001 0.00002 0 0.00012 0.00002 M 26–36 −0.00045 −0.00003 −0.00029 0.00002 0 0.00011 F 36–46 −0.0003 0.00005 0.00007 0 0.00018 −0.00015 M 36–46 0.00027 −0.00015 0.00031 −0.00028 0.00023 −0.00054 F 46–56 0.00086 −0.00018 −0.00013 −0.00019 0.00017 0.00008 M 46–56 0.00026 −0.0002 0.00032 0.00011 0.0004 −0.00006 F 56–66 0.00068 0.00027 0.00048 −0.00022 0.00091 −0.00094 M 56–66 0.00053 −0.00016 0.00037 −0.0003 0.00012 −0.00065 F 66–76 0.0008 −0.00097 0.00151 −0.00137 −0.00026 −0.00103 M 66–76 0.00058 −0.00032 0.00077 −0.00129 −0.00012 −0.00124 F 75+ 0.00006 0.00134 −0.00121 0.002 −0.00243 0.00257 M 75+ 0.0005 0.00066 −0.00029 0.0016 −0.00116 0.00196 Galicia 0.00013 −0.00003 0.00012 0.00001 −0.00044 −0.00002 Asturias 0.00001 0 0.00002 −0.00001 −0.00001 0 Cantabria 0 0 −0.00001 0 −0.00002 0 País Vasco 0.00003 0 0.00005 0 0.00004 0 Navarra 0 0 0 0 0 0 La Rioja 0 0 0 0 0 0 Aragón 0 0 −0.00001 0 −0.00001 0 Castilla y León 0.00001 0 −0.00001 0 −0.00002 0 Castilla−La Mancha 0 0 0.00001 0 0 0 Extremadura 0.00001 0 0.00002 0 0.00001 0 Cataluña 0.00001 0 0 0 −0.00001 0 Comunidad Valenciana 0.00009 0 0.00015 −0.00001 0.00019 −0.00001 Baleares 0.00002 0 0.00004 0 0.00002 0 Andalucía −0.00028 0.00001 −0.00033 0.00001 −0.0003 0.00001 Murcia 0.00002 0.00001 0.00008 0.00001 0.00008 0.00001 Ceuta y Melilla 0.00001 0 0.00001 0 0 0 Canarias 0.00004 0 0.00006 −0.00003 0.00012 −0.00004 Madrid, F 16–26 and Employed used as control groups Table AIV Decomposition results for rotation group 6 2009–2010 2009–2011 2009–2012 IRHI change −0.00244 −0.00367 −0.00643 Income growth −0.00043 −0.00119 −0.00182 Income inequality 0.00097 0.00103 0.00153 Income mobility −0.0033 −0.00385 −0.00611 Non-income factors 0.00032 0.00059 0.00029 Individual Contribution Income mobility Non-inc. factors Income mobility Non-inc. factors Income mobility Non-inc. factors M 16–26 0.00001 0 0 0 0 0 F 26–36 −0.00006 0.00015 0.00007 0.00032 0.00018 0.00051 M 26–36 0.00016 0.00007 0.00009 0.00013 0.00011 0.00021 F 36–46 0.00013 −0.00006 0.00007 −0.00011 0.00023 −0.00042 M 36–46 0.00038 −0.00003 −0.00002 0.00003 0.00042 0.00004 F 46–56 0.00033 −0.00029 0.00031 −0.00002 0.00031 −0.00015 M 46–56 0.00025 −0.00018 0.00029 −0.00032 0.0004 −0.00046 F 56–66 −0.00102 −0.00037 −0.00044 −0.00173 −0.00012 −0.00214 M 56–66 −0.00033 0.00017 0.00083 −0.00032 0.00022 −0.00073 F 66–76 −0.00015 −0.00026 −0.00142 −0.00051 −0.00191 −0.00076 M 66–76 −0.00013 −0.00095 −0.00043 −0.0014 −0.00118 −0.00166 F 75+ −0.00215 0.00121 −0.00223 0.00284 −0.00321 0.0037 M 75+ −0.00073 0.00088 −0.00099 0.00169 −0.00176 0.00217 Galicia −0.00042 0 −0.00042 −0.00001 −0.00009 −0.00001 Asturias 0 0 −0.00001 0 −0.00002 0 Cantabria 0 0 0.00001 0 0 0 País Vasco 0.00001 0 0.00003 0 0.00002 0 Navarra 0 0 0 0 0 0 La Rioja 0 0 0 0 0 0 Aragón −0.00001 0 −0.00001 0 −0.00001 0 Castilla y León 0.00002 0 0.00001 0 −0.00001 0 Castilla-La Mancha 0 0 0.00001 0 0 0 Extremadura 0 0 0.00001 0 0.00001 0 Cataluña 0 0 0.00001 0 0.00001 0 Comunidad Valenciana 0.00015 −0.00001 0.00012 −0.00002 0.00007 −0.00003 Baleares 0.00003 0 0.00002 −0.00001 0.00001 −0.00001 Andalucía 0.00011 0 0.00021 0.00001 0.00015 0.00002 Murcia 0.00012 0 0.00013 0 0.0001 0 Ceuta y Melilla 0.00003 0 0 0 0.00001 0 Canarias −0.00006 0 −0.00009 0.00002 −0.00007 0.00002 Madrid, F 16–26 and Employed used as control groups Figure A1 Decomposition of rotation groups 2, 3 & 5 Figure 1 Unemployment and GDP growth in Spain Source: GDP growth & unemployment data: Instituto National de Estadística (http://www.ine.es/prensa/pib_tabla_cntr.htm). Wage growth data: OECD data (https://stats.oecd.org/Index.aspx?DataSetCode=AV_AN_WAGE) Figure 2 IRHI per wave per rotation group Source: EU-SILC. Bars indicate 95% confidence intervals. Figure 3 Main decomposition results (bars indicate 95% confidence intervals) Figure 4 Decomposition of income mobility for rotation group 4 (bars indicate 95% confidence intervals) Figure 5 Decomposition of income mobility for rotation group 6 (bars indicate 95% confidence intervals) Table I Overview of the different rotation groups, Spanish EU-SILC, 2004–2012 Rotation group 1 2 3 4 5 6 2004 2005 2006 2007 2008 2009 2010 2011 2012 Table II Balanced panel observations per rotation group Rotation group Waves covered Individuals per wave Total observations per rotation group 1 2004–2007 4,193 16,772 2 2005–2008 4,996 19,984 3 2006–2009 5,099 20,396 4 2007–2010 5,575 22,300 5 2008–2011 5,617 22,468 6 2009–2012 5,168 20,672 122,592 Source: EU-SILC Table III Summary statistics (a) (b) Rotation Wave Household Incomea Ageb Female Health Labour Market Statusc Gini Employed Self- Employed Unemployed Other Disabled 1 2004 €23,962.08 47.2 53.10% 0.886 41.00% 7.40% 8.00% 25.80% 2.00% 0.31 1 2005 €25,156.27 48.3 53.10% 0.884 41.10% 8.30% 7.70% 24.40% 2.40% 0.31 1 2006 €25,602.62 49.3 53.10% 0.883 42.60% 8.10% 6.10% 24.20% 1.90% 0.30 1 2007 €26,056.85 50.2 53.10% 0.881 42.60% 8.00% 6.30% 22.80% 2.10% 0.30 4 2007 €27,554.90 46.7 51.30% 0.888 46.50% 6.50% 6.00% 24.20% 2.10% 0.28 4 2008 €28,775.31 47.6 51.30% 0.886 46.00% 6.90% 7.40% 21.70% 2.40% 0.29 4 2009 €29,900.84 48.5 51.30% 0.883 42.70% 6.50% 11.70% 19.80% 2.80% 0.29 4 2010 €28,475.24 49.5 51.30% 0.881 41.4 6.50% 12.30% 19.90% 3.20% 0.30 6 2009 €29,912.82 47.5 51.30% 0.885 40.10% 6.80% 11.60% 23.30% 2.80% 0.29 6 2010 €28,798.27 48.4 51.30% 0.882 38.80% 6.60% 13.50% 21.10% 3.40% 0.30 6 2011 €27,366.47 49.4 51.30% 0.88 39.00% 5.90% 12.00% 21.70% 3.20% 0.30 6 2012 €26,425.08 50.4 51.30% 0.878 37.10% 6.00% 14.90% 20.10% 3.20% 0.31 Source: EU-SILC. Weighted means calculated on balanced sample of rotation group 1, 4 & 6. a In 2010 Euros. b The yearly ageing of the balanced sample does not always equal one since survey months differed slightly from wave to wave. c Refers to current self-reported labour market status; ‘employed’ includes full- and part-time workers, while ‘other’ includes students, ‘domestic tasks’, and other ‘inactive’ persons. Table IV (a) Age/sex factors influencing term 3 Variable Coefficient Share of individuals in age-sex category (%) Re-ranking coefficienta,b Rotation 1 Rotation 4 Rotation 6 2004 2007 2007 2010 2009 2012 2004–2007 2007–2010 2009–2012 (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) Female 16–26 Reference 6.21 3.75 6.03 3.93 5.4 3.53 0.0152 0.0347 −0.0049 26–36 −0.0146* 8.81 8.19 9.27 8.41 8.98 8 0.000912 −0.0143 −0.023 36–46 −0.0286* 11.12 10.92 10.05 9.98 9.66 9.37 0.0106 −0.012 −0.00803 46–56 −0.0515* 8.65 9.55 8.4 9.27 8.65 9.28 −0.00071 −0.00066 −0.00675 56–66 −0.0816* 7.05 7.64 6.71 7.01 7.83 7.83 −0.00994 −0.0181 0.00274 66–76 −0.114*** 6.52 6.4 5.82 6.19 5.27 6.22 −0.00335 0.00199 0.0279* 75+ −0.156* 4.75 6.67 5.02 6.5 5.49 7.03 −0.00251 0.028 0.0354* Male 16–26 0.00117 5.97 4.11 5.98 3.94 6.12 4.45 0.0141 0.0272 −0.00259 26–36 −0.0125* 8.29 7.29 10.76 9.81 9.47 7.87 0.015 0.00154 −0.0222 36–46 −0.0272* 10.01 9.81 10.28 10.39 9.78 10.47 −0.00444 −0.0163 −0.0199 46–56 −0.0407* 8.22 8.9 7.79 8.69 8.45 8.78 0.00478 −0.00821 −0.0138 56–66 −0.0667* 6.53 7.12 5.97 6.38 7.19 7.61 −0.0136 −0.00323 −0.00266 66–76 −0.0853* 5.21 5.79 4.9 5.09 4.55 4.98 −0.0129 0.00211 0.0286* 75+ −0.118*** 2.65 3.87 3.03 4.4 3.17 4.55 −0.00728 0.0256* 0.0352** a Re-ranking coefficient refers to a no-constant regression where change in rank is regressed on the demographic variables in the final period b The coefficients are not jointly significant for the 2004–2007 model, while they are for the 2007–2010 and 2009–2012 model. * p<0.05, p<0.01, * p<0.001 Table IV (b) Region factors influencing term 3 Variable Coefficient Share of individuals in region category (%) Re-ranking coefficienta,b Rotation 1 Rotation 4 Rotation 6 2004 2007 2007 2010 2009 2012 2004–2007 2007–2010 2009–2012 (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) Region Madrid Reference 4.13 4.14 13.14 13.3 12.22 12.28 0.0455* −0.025 −0.00799 Galicia −0.0267*** 7.7 7.7 7.27 7.32 6.43 6.43 −0.00405 0.023 0.00503 Asturias −0.0048 3.36 3.35 3 2.98 2.71 2.7 0.00533 0.00578 0.00896 Cantabria 0.00449 0.9 0.9 1.28 1.29 1.44 1.42 0.0148* −0.0149 −0.00627 País Vasco 0.0023 3.69 3.63 3.66 3.67 3.94 3.95 −0.00055 0.0417*** 0.0148 Navarra 0.00135 2.01 2.01 1.65 1.65 1.3 1.3 0.00288 −0.00156 0.02* La Rioja −0.000513 0.83 0.83 0.69 0.69 0.88 0.96 0.00146 0.00245 −0.00226 Aragón −0.000675 3.44 3.41 3.36 3.35 3.23 3.23 −0.00034 0.0285** 0.0261* Castilla y León −0.0012 7.13 7.29 6.5 6.49 6.39 6.31 0.00815 0.0133 0.00948 Castilla-La Mancha −0.000617 4.74 4.67 4.89 4.82 5.17 5.2 −0.0287** −0.0122 −0.0146 Extremadura 0.00203 3.54 3.42 2.7 2.7 2.75 2.82 −0.0152* 0.0274* 0.0113 Cataluña 0.000396 16.9 16.9 15.09 14.93 15.12 15.19 −0.0153 −0.0109 0.0144 Valenciana −0.00439 12.6 12.69 10.11 10.13 12.44 12.22 0.0222 −0.0620* −0.0222 Baleares 0.00554* 2.28 2.25 2.19 2.19 1.92 1.87 −0.00624 0.0171 0.0133 Andalucía −0.00871*** 18.36 18.47 17.44 17.41 16.33 16.39 0.0128 0.0306* −0.015 Murcia −0.00974* 4.52 4.46 3.11 3.18 3.41 3.41 −0.0123 −0.022 −0.0267 Ceuta y Melilla −0.0148* 0.29 0.29 0.24 0.24 0.27 0.27 0.0052 −0.00152 −0.0028 Canarias −0.0119* 3.57 3.57 3.7 3.68 4.06 4.06 −0.0298* −0.0344* 0.012 a Re-ranking coefficient refers to a no-constant regression where change in rank is regressed on the regional variables in the final period b The coefficients are jointly significant for the 2004–2007 and 2007–2010 model, while they are only jointly significant at the 0.10 level for the 2009–2012 model. * p<0.05, p<0.01, * p<0.001 Table V Labour Market Status / Occupational re-ranking Variable Re-ranking coefficienta,b 2004–2007 2009–2012 2004–2007 2009–2012 (1) (2) (3) (4) Labour Market Status Employed 0.00472 −0.0142* - - Self Employed −0.0324* 0.0447* - - Unemployed 0.0109 −0.0675*** - - Other 0.00397 0.0152* - - Retired −0.00679 0.0287* - - Disabled 0.000519 0.0287 - - Occupation Managers - - −0.0222 0.0658* Military - - 0.0553 0.0136 Professionals - - 0.00953 0.0155 Technicians - - −0.0197 0.025 Clerks - - −0.00476 −0.0314 Service & Sales - - 0.0263 0.00399 Agricultural - - −0.0217 −0.0513 Trade - - −0.012 −0.0245 Machine Operators - - −0.00249 −0.0286 Elementary Occupation - - −0.00456 −0.0452 a Re-ranking coefficient refers to a no-constant regression where change in rank is regressed on economic status/occupation in the first period b The coefficients are jointly significant for the all models, except for the 2004–2007 occupation regression, where they are only jointly significant at the 0.10 level. * p<0.05, p<0.01, * p<0.001 2 In the years immediately following the crisis GDP growth rebounded slightly before falling back to negative growth, whereas unemployment steadily increased to above 25% in 2012. Real wages actually increased between 2009 and 2010. However, this was due to compositional effects, as less experienced workers with lower paying temporary contracts were the first to lose their jobs (Puente & Galan, 2014). As the crisis progressed however, real wages contracted. 3 The reasons for the extraordinary growth until 2006, and the collapse after 2008 in the construction sector are still up for debate. See Gonzalez and Ortega (2009) and Bentolila et al. (2012) for more details. 4 Bacigalupe and Escolar-Pujolar (2014) review four studies on Spain during the Great Recession, and conclude that the evidence points in the direction of increasing health inequalities. 5 zit takes (1 − n)/2 for the poorest individual and (n − 1)/2 for the richest individual. 6 It should be stressed that our goal is not to estimate a causal model of health; our sole aim is to decompose changes in the (partial) association between health and income rank. As we have neglected to include potentially endogenous variables such as education or lifestyle the non-linear income function features as the sole potentially endogenous variable. We have deliberately not addressed its potential endogeneity since we are interested in documenting the association between changes in the distribution of income and the evolution of IRHI in Spain. Turning to the underlying mechanisms is only sensible after the magnitude of this association has been established, and after the relative importance of “income growth”, “mean-preserving income changes”, and “income mobility” has been understood. 7 After results are presented we return to the assumption that equation (2) is deterministic. 8 The intercept parameters drop out in equation (3b) since ∑i=1nzit=0. 9 In 2004 and 2012, we observe only one rotation group (group 1 & 6); in 2005 and 2011 we simultaneously observe 2 rotation groups (group 1/2 & 5/6), in 2006 & 2010 3 rotation groups (group 1/2/3 & 4/5/6), while for the years 2007, 2008 and 2009, we simultaneously observe 4 separate rotation groups (group 1/2/3/4; 2/3/4/5 & 3/4/5/6). 10 Decompositions using alternative equivalence scales, such as the OECD-modified scale, did not significantly change the results. 11 Negative incomes can occur in the EU-SILC data due to debt, but make up less than 1% of the observations. They are problematic as in the hypothetical average income movement scenario these individuals will see their incomes drop when on average incomes rise. However, decompositions that included these observations did not change the qualitative features of our results. 12 This involves estimating an ordered probit model, with the thresholds imposed from the empirical distribution function of HUI in the Canadian National Population Health Survey 1994–1995 (HUI=1 equals maximum health and HUI=0 equals minimum health). Several studies using this approach (e.g. Van Doorslaer and Jones, 2003; Lauridsen et al., 2004; Lecluyse and Van Cleemput, 2006) have found the health inequality estimates to be rather insensitive to the threshold values imposed. 13 While the predicted HUI scores only reflect health changes resulting from changes in the explanatory variables, Van Doorslaer and Jones (2003) show that the interval regression approach is the preferred approach when calculating health inequality indices. One might also calculate the conditional predictions from the interval regression model given the observed SAH levels, but then the predicted HUI scores would no longer be a linear combination of the explanatory variables, and therefore not be amenable to our decomposition approach. 14 As explained before, the signs and size of term 1 and 2 (‘income growth’ and ‘evolution of income inequality’) largely depend on whether the health responsiveness to proportional income changes decreases or increases with rising incomes. This is left open with a second order income polynomial, but not with other popular choices in the empirical literature. For example, when one would favour the natural logarithm of income, one would impose that a proportional change in income has the same health effect for every individual (and hence one would force term 1 to be zero). 15 Summary statistics of the full unbalanced panel sample are similar to those of the balanced panel. Nor did the evolution of IRHI using the unbalanced panels for each rotating group differ markedly, suggesting that attrition is not an important driver of our main findings, although we cannot entirely rule out that explicitly accounting for mortality as in Petrie et al. (2013) would have disproportionally hit the older and poorer age groups. 16 Decomposition results for the 2005–2008, 2006–2009 and the 2008–2011 rotation groups are not presented for reasons of clarity and brevity. They are in line with the results presented and available upon request. 17 The assumption of constant coefficients may be questionable in the case of pre- and post-crisis Spain, since the relationship between income and health may have changed. To test the robustness of our main findings we also decomposed the periods 2004–2007 using coefficients estimated on pre-crisis observations (before 2009) only, and 2009–2012 using only post crisis observations (after 2008). This did not change our results. 18 While including the top 1% of incomes does change the function form of income due to some extreme outliers (in particular among very high incomes), it does not change the overall results of the decomposition. Nevertheless in order to achieve an income function that is not unduly influenced by outliers we remove the top 1% of incomes. 19 One should only use the 95% confidence intervals to compare IRHI between rotation groups, since different waves within each rotation group are dependent samples. 20 The fact that similar trends are observed using different rotation groups indicates that the trend is not simply driven by a particular rotation group. 21 Table A1 in the appendix shows the numerical changes of IRHI between waves for each rotation group, and indicates the significance of such changes. 22 The EU-SILC categorises Spain into 18 different regions: Galicia, Asturias, Cantabria, País Vasco, Navarra, La Rioja, Aragón, Madrid, Castilla y León, Castilla-La Mancha, Extremadura, Cataluña, Comunidad Valenciana, Baleares, Andalucía, Murcia, Ceuta y Melilla, and Canarias. 23 The coefficients of the income polynomial are suppressed. 24 The fact that term 3 is so large in magnitude but still insignificant indicates that there are a small amount of very large and influential income re-rankings occurring. An individual moving from the bottom of the income distribution to the top in turn affects the rankings of the rest of the sample as well. Term 3 therefore changes dramatically when this individual is left out of the sample in a bootstrap replication. 25 The change in z-scores is bounded between −2 and 2 since the z-scores have been normalized between −1 and 1. For example, the most extreme case of an individual going from the highest to the lowest rank would lead to zi2 − zi1 = (−1) −1= −2. 26 One may question whether IRHI due to natural ageing is interesting or important, since ageing is an unavoidable biological process. In this case, the decomposition method can be viewed in different ways. If we are interested in the evolution of total IRHI then the sum of all 4 terms should be considered. If we wish to exclude the effect of natural ageing then we should exclude the non-income factors term. If we wish to narrow our focus further, and ignore that part of the evolution of IRHI that is due to the mobility of different age groups then the income mobility term should also be excluded. 27 For a comprehensive overview of recent reforms of Spanish old-age and disability pensions see e.g. García-Gómez et al (2012). 28 Neither our model of health nor our decomposition accounts for individuals’ labour market status. We have repeated the decomposition with the inclusion of labour market status and the results are extremely similar to those presented here. This is because once age is controlled for labour market status has very little correlation with health, and consequently can explain only very little. JEL Classification Number: D30, D63, I14, I15 Aparicio Fenoll A 2010 High-school dropouts and transitory labor market shocks: The case of the Spanish housing boom 5139 Institute for the Study of Labor (IZA) Bacigalupe A Escolar A 2014 The impact of economic crises on social inequalities in health: what do we know so far? International journal of equity in health 13 52 10 1186 Baeten S Van Ourti T Van Doorslaer E 2013 Rising inequalities in income and health in China: Who is left behind? Journal of Health Economics 32 6 1214 1229 24189450 Bentolila S Cahuc P Dolado JJ Le Barbanchon T 2012 Two-Tier Labour Markets in the Great Recession: France versus Spain The Economic Journal 122 562 F155 F187 Bonhomme S Hospido L 2012 Earnings Inequality in Spain: Evidence from Social Security Data Trabajo presentado en el XV Encuentro de Economía Aplicada, A Coruña 7 8 Carrasco R Jimeno JF Carolina Ortega A 2011 Accounting for changes in the Spanish Wage Distribution: The Role of Employment Composition Effects Banco de España working paper 1120 Contoyannis P Forster M 1999 The distribution of health and income: a theoretical framework Journal of Health Economics 18 605 622 10621367 Denny K 2015 Has Subjective General Health Declined with the Economic Crisis? A Comparison across European Countries Working Papers 201506 School Of Economics, University College Dublin Erreygers G 2009 Correcting the Concentration Index Journal of Health Economics 28 504 515 18367273 Erreygers G Van Ourti T 2011 Measuring socioeconomic inequality in health, health care and health financing by means of rank-dependent indices: a recipe for good practice Journal of Health Economics 30 685 694 21683462 European Commission 2009 Solidarity in Health: Reducing Health Inequalities in the EU Brussels European Commission 2013 Health inequalities in the EU — Final report of a consortium 2013 Consortium lead: Sir Michael Marmot Brussels Garcia Gomez P Jimenez-Martin S Vall Castello J 2012 Health, disability and pathways into retirement in Spain Wise DA Social Security Programs and Retirement around the World: Historical Trends in Mortality and Health, Employment, and Disability Insurance Participation and Reforms University of Chicago Press Gonzalez L Ortega F 2013 Immigration and housing booms: Evidence from Spain* Journal of Regional Science 53 1 37 59 Jenkins SP Brandolini A Micklewright J Nolan B 2012 The great recession and the distribution of household income Oxford University Press Lacuesta A Izquierdo M 2012 The contribution of changes in employment composition and relative returns to the evolution of wage inequality: the case of Spain Journal of Population Economics 25 2 511 543 Lauridsen J Christiansen T Häkkinen U 2004 Measuring inequality in self-reported health—discussion of a recently suggested approach using Finnish data Health economics 13 7 725 732 15259049 Lecluyse A Cleemput I 2006 Making health continuous: implications of different methods on the measurement of inequality Health economics 15 1 99 104 16145723 Petrie D Allanson P Gerdtham UG 2011 Accounting for the dead in the longitudinal analysis of income-related health inequalities Journal of Health Economics 30 1113 1123 21820193 Pijoan-Mas J Sánchez-Marcos V 2010 Spain is different: Falling trends of inequality Review of Economic Dynamics 13 1 154 178 Puente S Galán S 2014 Analysis of Composition Effects on Wage Behaviour Banco de España, Economic Bulletin 2 Regidor E Barrio G Bravo MJ de la Fuente L 2014 Has health in Spain been declining since the economic crisis? Journal of epidemiology and community health 68 3 280 282 24153246 Rica S Rebollo-Sanz YF 2015 Gender Differentials in Unemployment Ins and Outs during the Great Recession in Spain 9135 Institute for the Study of Labor (IZA) Ruhm CJ 1996 Are recessions good for your health? w5570 National bureau of economic research Ruhm CJ 2015 Recessions, healthy no more? Journal of health economics 42 17 28 25839783 Stuckler D Basu S Suhrcke M Coutts A McKee M 2009 The public health effect of economic crises and alternative policy responses in Europe: an empirical analysis The Lancet 374 9686 315 323 Van Doorslaer E Koolman X 2004 Explaining the differences in income-related health inequalities across European countries Health economics 13 7 609 628 15259042 Van Doorslaer E Jones AM 2003 Inequalities in self-reported health: validation of a new approach to measurement Journal of health economics 22 1 61 87 12564718 Van Ourti T van Doorslaer E Koolman X 2009 The effect of income growth and inequality on health inequality: Theory and empirical evidence from the European Panel Journal of Health Economics 28 3 525 539 19185942 Vandoros S Hessel P Leone T Avendano M 2013 Have health trends worsened in Greece as a result of the financial crisis? A quasi-experimental approach The European Journal of Public Health 23 5 727 23417622 Mazzonna F Bucher-Koenen T 2013 The recent economic crisis and old-age health in Europe Active ageing and solidarity between generations in Europe Börsch-Supan A Brandt M Litwin H Weber G 233 242 OECD 2013 Pensions at a Glance 2013: OECD and G20 Indicators OECD Publishing http://dx.doi.org/10.1787/pension_glance-2013-en OECD 2014 Health at a Glance: Europe 2014 OECD Publishing http://dx.doi.org/10.1787/health_glance_eur-2014-en
PMC005xxxxxx/PMC5123804.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 2985117R 4816 J Immunol J. Immunol. Journal of immunology (Baltimore, Md. : 1950) 0022-1767 1550-6606 27798169 5123804 10.4049/jimmunol.1601401 NIHMS818996 Article Germinal center hypoxia potentiates immunoglobulin class switch recombination Abbott Robert K. * Thayer Molly * Labuda Jasmine * Silva Murillo * Philbrook Phaethon * Cain Derek W. † Kojima Hidefumi ‡ Hatfield Stephen * Sethumadhavan Shalini * Ohta Akio * Reinherz Ellis L. § Kelsoe Garnett † Sitkovsky Michail * * New England Inflammation and Tissue Protection Institute, Northeastern University, Boston, MA † Department of Immunology and Human Vaccine Institute, Duke University, Durham, NC ‡ Department of Immunology, Dokkyo Medical University School of Medicine, Tochigi, Japan § Laboratory of Immunobiology, Department of Medical Oncology, Dana Farber Cancer Institute and Harvard Medical School, Boston MA Address correspondence and reprint requests to Dr. Robert K Abbott, Department of Biology, Northeastern University, 360 Huntington Ave, Mugar Building Rm 031, Boston MA, 02115. [email protected], phone: 617-373-5737. Fax: 617-373-5834 29 9 2016 19 10 2016 15 11 2016 15 11 2017 197 10 40144020 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Germinal centers (GCs) are anatomic sites where B cells undergo secondary diversification to produce high affinity, class switched antibodies. We hypothesized that proliferating B cells in GCs create a hypoxic microenvironment that governs their further differentiation. Using molecular markers, we found GCs to be predominantly hypoxic. Compared to normoxia (21% O2), hypoxic culture conditions (1% O2) in vitro accelerated class switching and plasma cell formation and enhanced expression of GL-7 on B and CD4+ T cells. Reversal of GC hypoxia in vivo by breathing 60% O2 during immunization resulted in reduced frequencies of GC B cells, T follicular helper (TFH) cells and plasmacytes, as well as lower expression of ICOS on TFH. Importantly, this reversal of GC hypoxia decreased antigen-specific serum IgG1 and reduced the frequency of IgG1+ B cells within the antigen specific GC. Taken together, these observations reveal a critical role for hypoxia in GC B cell differentiation. Introduction Development of effective vaccination strategies requires a detailed understanding of the mechanisms that govern adaptive immunity. The germinal center (GC) is the anatomical site in which antigen-activated B- and T lymphocytes interact, initiating immunoglobulin class switch recombination (CSR), somatic hypermutation, and the affinity maturation associated with effective antibody responses (1). However, surprisingly little is known about the physiological mechanisms within the GC microenvironment that regulate these processes. It has become evident that local oxygen tension and the physiological response to hypoxia play roles in regulating inflammation (2) through synergistic and independent mechanisms including, but not limited to, hypoxia inducible factors (e.g. HIF-1α, HIF-2α and HIF-3α), NF-κB, mammalian target of rapamycin kinase (mTOR), and the unfolded protein response (UPR) (2-4). Moreover, the cellular response to tissue hypoxia often results in metabolites in the extracellular space, which have diverse signaling capacities that can facilitate both pro and anti-inflammatory responses (2, 5). Indications that hypoxia may be important in B cell physiology originally came from experiments that revealed HIF-1α is required for B cell development and prevention of autoimmunity (6). Furthermore, HIF-1α has also been detected in human tonsillar GCs (7). We hypothesized that rapidly proliferating B cells within GCs develop a hypoxic microenvironment that promotes CSR. We show that GCs contain hypoxic regions linked to accelerated class switching, plasma cell development and antibody secretion. Following vaccination, administration of clinically relevant (8) respiratory hyperoxia (60% O2) via supplemental oxygen dramatically suppressed the GC response and subsequent antibody production, revealing a previously unappreciated functional role of hypoxia within the GC microenvironment. Materials and Methods Animal studies 10-12wk female C57BL/6J mice were from Jackson Laboratories (Bar Harbor ME). Animal work was in accordance with IACUC at Northeastern University. Mice were immunized i.p. with Alum hydroxide (5mg/mouse) / NP(6)-OVA (1μg/mouse). 10% Alum sulfate (Sigma) in PBS was mixed 1:1 with diluted stock of NP(6)-OVA or NP(11)-CGG (Biosearch Technologies / produced in-house) and precipitated by 1M KOH, washed 3x, and injected in 200μl/PBS. For hyperoxia experiments, mice cages were placed in custom made chambers hooked up to an oxygen concentrator (AirSep). Statistics were calculated using two tailed students t-test in Microsoft excel. Flow cytometry Cells were FC blocked (10m/4°C) and stained (20m/4°C), washed 2x and fixed (FoxP3 kit, eBioscience). Hypoxyprobe compound was injected into mice i.v.(100mg/kg), circulated for 90 minutes, and detected with Hypoxyprobe mAB (4.3.11.3 (FITC)) was added for 1h/4°C after fixation. FACS buffer was 1xPBS/5%FCS. mABs (eBioscience, Becton Dickenson, Biolegend) were: IgG1 (A85-1) B220 (RA3-6B2), CD4 (RM4-5)(GK1.5), CD43 (S7), CD138 (281-2), CD16/32 (2.4G2), GL-7 (GL7), CD38(90), CD86 (GL-1), CD40 (1C10), CD73 (Ty/11.8), IgM(II/41), CXCR5 (SPRCL5), PD-1 (29F.1A12), ICOS (7E.17G9)(C398.4A), BCL-6(GI191E), FAS(Jo2), Zombie Green, FoxP3 (FJK16S), NP(36)-PE (Biosearch Technologies). Acquisition was on a Cytek DxP8 FACSCalibur, and analyzed on Flowjo X (Treestar). In Vitro Assays Napco 7000 incubators (37°C/5% CO2) were used. Freshly prepared media consisted of IMDM, 25mM HEPES, 10% FCS (defined high grade (GE Healthcare)), glutamine, 55mM 2-ME, and 100U/ml Pen/Strep. Media was equilibrated in incubators for at least 6h before experiments. Cells were cultured in a 24 well plate (Costar #3474) in 1mL. Splenocytes were Ficoll separated (GE Healthcare) and stained with CD43 FITC (CSR Assay), CD8/GL-7 FITC (CD4 Assay), or CD43/GL-7 FITC (B cell assay). Labeled cells were depleted using αFITC Microbeads (Miltenyi Biotec) and AutoMACS. For CSR assay αCD40 mAb (clone 1C10, 2.5μg/ml, Biolegend) and rmIL-4 (10ng/ml, R&D Systems) were used and seeded at 0.1× 106 cells/ml. For CD4 T cell stimulation cells were stimulated with αCD3 mAb (clone 2c11, 1μg/ml BD Biosciences) and seeded at 0.5× 106 cells/well. For B cell assay cells were stimulated with αCD40 mAb (1C10, 2.5μg/ml Biolegend) and αIgM ab (6μg/ml, Jackson Immunoresearch), and seeded at 0.5× 106 cells/ml. Visualization of Vessels and Histology Immunized mice were injected i.v. with 150μg Dylight 594 conjugated Tomato Lectin (Vector Labs). After 10-30m, the mice were euthanized and perfused with 30ml of 1% PFA followed by 30ml of PBS. Spleens frozen in OCT compound were cryo-sectioned (5μm), air dried and fixed in 1:1 mix of acetone/methanol at -20°C/10m. Slides were warmed to RT, Pap Pen applied and re-hydrated (0.1% Tween 20, 0.5% BSA in 1xPBS) for 20m, Fc blocked for 20m, primary stained for 3h, washed 3x and stained with αFITC Oregon Green (Invitrogen) for 1h, washed 3x and mounted. Images were acquired on a Nikon e80i microscope or a Zeiss LSM 710. ImageJ FIJI was used for analysis. Mask files were created for GCs (GL7+), follicles (B220+), and vasculature (tomato lectin+). Vasculature mask was expanded radially to 40μm and percent GC or follicle area covered was determined. ELISA and ELISPOTS For ELISA, 96 well plates (Costar 2595) were coated with 2μg/ml NIP 25- BSA in carbonate buffer for 1h/RT, washed 3x (0.5% BSA, 0.1%Tween 20 in PBS), blocked for 1h (0.5%BSA in PBS), washed 3x, and samples added for 1h. Plates were washed 3x and HRP-conjugated detection Abs for IgG1 and IgM (Bethyl Labs) were added for 1h, washed 3x and TMB substrate was added (BD Biosciences). Reaction was stopped using 2N H2SO4 and read on a Varioskan (Thermo Electron) at 450nm. Mid points of dilution curves were used and values calculated relative to standards (B-18 (IgM) and H33lγ1 (IgG1)). For ELISPOTS, Immobilon-P plates (Millipore) were coated overnight at 4°C with 2μg/ml NIP-25 BSA in carbonate buffer (pH 9.5). Plates were washed 5x (0.5%BSA/0.1% Tween 20/PBS), blocked in same buffer for 1h. After wash (5x) complete RPMI media was used to wash (2x) and cells were serially diluted, added and sat for 3h in incubator. Plate was washed with DI water and left in wash buffer overnight at 4°C. AP-conjugated antibodies to IgG1 or IgM (Southern Biotech) were added. Plates were washed (5x) with was buffer and with DI water (1x) and developed using BCIP/NBT reagent (Sigma Aldrich). RT-PCR for IgG1 circle transcripts RNA isolation was done using RNeasy mini kit (Qiagen) and Superscript III (Invitrogen) were used. For RT-PCR (done on Applied Biosciences 7300) initial step was 95C 9m followed by 40 cycles of PCR (94C 30s, 58C 1min) by using RT2 SYBR green PCR master mix (Life Technologies). Primers (Eurofins) were IgM reverse (CμR, 5′-AATGGTGCTGGGCAGGAAGT-3′) and IgG1 forward (Iγ1F, 5′-GGCCCTTCCAGATCTTTG AG-3′) (15). L32 (Qiagen) was used to normalize expression. Results The germinal center is hypoxic and poorly vascularized To ascertain whether the GC microenvironment is hypoxic, we immunized female C57BL/6J mice with the hapten-protein antigen, (nitrophenyl) acetyl-ovalbumin (NP-OVA) in aluminum hydroxide (alum) adjuvant. Identification of areas of tissue hypoxia or cells in hypoxic areas was determined by immunohistology using a well-documented molecular marker of tissue hypoxia in vivo, Hypoxyprobe (pimonidazole) (9). Injected intravenously, Hypoxyprobe creates thiol adducts with proteins under hypoxic conditions (<1.3% O2), which are then specifically recognized by a monoclonal antibody. To assess GC hypoxia in situ, we studied Hypoxyprobe labeling of splenic GCs (Fig.1A). These histologic studies indicated that GCs are enriched for hypoxic areas (Fig.1A). We subsequently confirmed Hypoxyprobe labeling of GC B cells by multi-color flow cytometric analysis of GC B cells from spleens (Fig. 1B, C) (Supplemental Fig. 1D), Peyer's Patches (Supplemental Figs. 1A-B) and mesenteric lymph nodes (Supplemental Fig. 1C) of immunized and naïve mice. To a lesser extent, GC associated TFH cells (CD4+CXCR5+Bcl-6+GL-7+) showed increased Hypoxyprobe staining compared to respective controls (Supplemental Figs. 1E-F), indicating hypoxia. In support of this demonstration of GC B- and T-cell hypoxia, we injected fluorescently-labeled tomato lectin that binds to blood vessel walls and then used fluorescence microscopy to measure the relative distances between GC B cells and the splenic blood vasculature (Figs. 1D-E). Due to the fact that earlier studies have shown that pO2 can drop to hypoxic levels 30-40 μm from blood vessels (10), this distance was used to evaluate the likelihood of a hypoxic environment at GC sites. We found that 8 days after NP-OVA immunization, the substantial majority of GC area lies ≥40 μm from the nearest blood vessel (Fig. 1F). Although the majority of GCs were located ≥40 μm from the nearest blood vessel, occasional GCs were adjacent to or even surrounded tomato lectin stained vessels (data not shown). However, in the three dimensional space of lymphoid tissue, GCs are preferentially sited in hypoxic regions (Fig. 1A-F). Hypoxic culture conditions promote Gl-7 expression To assess the functional role of GC hypoxia, we used in vitro experiments with defined gas mixtures. Briefly, resting B cells (CD43-GL7-) were isolated by magnetic depletion and stimulated in cultures containing anti-CD40 and anti-IgM antibodies in normoxic (21%O2) or hypoxic (1% O2) incubators. We found that the GL-7 activation marker was upregulated on a greater fraction of hypoxic B cells and to higher levels than normoxic controls (Fig. 2A-C). Although GL-7 expression can be upregulated on B- and T cells in vitro (11), it is widely used to identify GC B cells in vivo [1]. The GL-7 carbohydrate epitope may be influenced by oxygen levels directly, as it depends on the repression of CMP-Neu5Ac hydroxylase (12). Since GL-7 was recently shown to mark a subset of GC associated TFH (13), we assessed the effect of hypoxia on GL-7 up-regulation on CD4+ T cells in vitro. Upon stimulation with anti-CD3 antibody, we observed that hypoxic culture conditions increased the frequency of GL-7 expressing CD4+ T cells and also increased the expression level of GL-7 in CD4+ T cells when compared to normoxic controls (Supplemental Figs. 2A-C). Hypoxic culture conditions promote CSR and plasmacyte differentiation To identify any additional consequences of hypoxia and increased the expression of GL-7 by B cells activated in hypoxic cultures, we determined the frequency of cells exhibiting IgM+→IgG1+ class switch recombination after stimulating resting B cells with anti-CD40 antibody and IL-4 (14). We then placed matched cohorts under normoxic (21%O2) or hypoxic (1% O2) conditions for four days to determine the numbers of class switched IgG1+ B cells by flow cytometry. Interestingly, hypoxic culture conditions accelerated IgM+→IgG1+ class switch kinetics on day 3, while the frequency of IgG1+ B cells in hypoxic chambers failed to increase on day 4 (Figs. 3A-B). Consistent with the accelerated presence of IgG1+ B cells in hypoxic cultures on day 3 was a decrease in IgM+ B cells (Supplemental Fig. 3B). Furthermore, by CFSE labeling B cells at the start of culture, we assessed if IgG1+ B cells in hypoxic chambers proliferated comparably to normoxic controls. We found that the early increase in IgG1+ B cells in hypoxic cultures does not represent biased proliferation (Supplemental Figs. 3D), but rather increased rates of class switch recombination (IgM to IgG1) determined by the increase in IgM excision circles (Fig. 3C), which are hallmarks of CSR (15). Interestingly, the rate of B cell division in hypoxic chambers was accelerated on day 3 but slowed significantly and specifically in the IgG1+ compartment on day 4 (Supplemental Figs. 3C-D). Further analysis revealed that on day 4, the IgG1+ B cell compartment in hypoxia had significantly increased frequencies of apoptotic cells as determined by active caspase 3 staining, while apoptosis in the plasma cell compartment was unaffected (Supplemental Fig. 3E).This differential distribution of frequencies of apoptotic cells may account for the fact that we did not observe major differences in viable cells in hypoxic cultures (Supplemental Fig. 3A). Plasma cell differentiation, as determined by the appearance of CD138+B220- cells in hypoxic and normoxic cultures, was dramatically increased under conditions of low oxygen tension (Figs. 3D-E)(Supplemental Fig. 3F), suggesting that hypoxia accelerates both class switch recombination and differentiation to plasmacytes. This interpretation is supported by the observation that on day 4, hypoxic cultures had an increased frequency of IgG1+ antibody secreting cells (ASC) but not IgM+ ASCs as determined by ELISPOT (Figs. 3F-G). Taken together, the accelerated plasma cell differentiation and ∼50% increase in total IgG1+ ASC in hypoxic cultures may partially account for reduced expansion of IgG1+ B cells in these cultures on day 4 (Fig. 3B). Reversal of tissue hypoxia by treatment with respiratory hyperoxia during immunization suppresses CSR and the GC reaction If hypoxia promotes B cell differentiation during GC responses, we predicted that reversing tissue hypoxia during immunization would suppress the GC reaction. To test this prediction, we systemically reduced tissue hypoxia, as done clinically, by placing mice in chambers containing 60% O2 to induce respiratory hyperoxia (8). Female C57BL/6J mice immunized with NP-OVA or NP-chicken γ-globulin (NP-CGG) in alum and held under conditions of respiratory hyperoxia exhibited dramatic suppressions of total and NP-specific GC responses 8 days after immunization (Fig. 4A and Supplemental Fig. 4 A-C). Similarly, but to a lesser extent, TFH and GC associated TFH, defined as CD4+B220-CXCR5+PD1+GL7- and CD4+B220-CXCR5+PD1+GL7+, respectively, were reduced in frequency by hyperoxia (Fig. 4B and Supplemental Fig. 4E). Moreover, administration of respiratory hyperoxia in mice resulted in lower ICOS expression by both TFH and GC associated TFH (Fig. 4C and Supplemental Fig. 4D). Functional impairment of the GC response by respiratory hyperoxia was also manifested by fewer IgM+ and IgG1+ (nitroiodophenyl)acetyl-bovine serum albumin (NIP-BSA) specific ASCs (Figure 4D). To ascertain if plasma cells contributed to the reduced frequencies of ASCs in hyperoxic treated mice, we identified plasma cells using flow cytometry (CD138+B220-) and found that hyperoxia significantly suppressed plasma cell formation (Fig. 4E). This was in excellent correlation with our in vitro finding that hypoxic cultures accelerated plasmacytic differentiation (Figs. 3D-E). Functionally, serum concentrations of NIP-BSA-specific IgM and IgG1 antibody were significantly reduced in mice treated with respiratory hyperoxia (Figs. 4F-G). Moreover, among NP-specific GC B cells, hyperoxia significantly reduced the frequency of class switched IgG1+ GC B cells whereas the proportion of IgM+ GC B cells was unchanged (Figs. 4H-I). We subsequently confirmed the reduction in IgG1+ GC B cells by respiratory hyperoxia through histology (Fig. 4J). Taken together, the reduction in IgG1+ NP-binding GC B cells and diminished serum NIP-specific IgG1 (Figs. 4G-I) suggests that respiratory hyperoxia likely has a more robust effect on Ig class switching than on initial, extrafollicular interactions between activated T and B cells. Discussion In summary, we have described a previously unappreciated role for hypoxia in GCs that acts on both B- and T cells and promotes class switch recombination as well as plasmacyte differentiation. These data show that administration of clinically relevant (8) 60% O2 dramatically suppresses the GC reaction and production of IgG1+ antibody (Figs. 4A-J). Our observations of the GC microenvironment offer new avenues of investigation into the role of hypoxia and hypoxia-induced immune-regulatory pathways during vaccination. It would be interesting to determine if any HIF proteins are responsible for the effects of hypoxia in promoting CSR and the GC response. Future studies will be needed to asses if hypoxia plays a role in directly regulating activation induced cytidine deaminase (AID) at the transcriptional and/or post transcriptional level, as AID is known to be regulated through phosphorylation (16, 17) at serine 38 and tyrosine 140 which affects CSR (18, 19). Given our findings that class switched B cells cultured in hypoxic conditions appear to be more likely to undergo apoptosis, it would be interesting to determine if hypoxia within the GC promotes apoptosis and helps to “set the stage” within a GC for a competitive microenvironment to facilitate clonal competition and affinity maturation. This hypothesis is consistent with observations that both transgenic mice overexpressing the anti-apoptotic protein bcl-XL and mice that specifically lacking caspase 8 within GC B cells both exhibit delayed affinity maturation (20, 21). Our study also raises the question of whether hypoxia plays a role in regulation of somatic hypermutation (SHM) of variable region genes of GC B cells. We speculate that the hypoxic microenvironment within the GC may be an evolutionarily conserved regulatory mechanism to both promote as well as constrain potentially detrimental DNA-damaging genetic events such as CSR and SHM to the anatomical site of the GC. While there are documented examples of CSR and SHM outside of the GC (22-24), these examples come from either multi-valent T independent antigens such as NP-Ficoll which induce very strong BCR crosslinking or in disease states such as in lupus prone mice. While such examples exist, the GC is generally considered to be the “professional” site of CSR and SHM, and perhaps the hypoxic microenvironment within the GC helps promote this. In line with this thinking is the curious observation that while human GCs can be somewhat larger than murine GCs (25), many are the same size (26, 27)(and personal communication with G. Kelsoe), lie toward the center of the follicle (28), have been shown to express HIF-1α and are poorly vascularized (7). The development of hypoxia within the GC also raises the question if this hypoxic microenvironment is sufficient to drive neovascularization of secondary follicles over the course of a GC reaction. It is important to note that reversal of tissue hypoxia appears to affect multiple aspects of the GC reaction: GC B cell frequency, CSR, TFH and plasma cell induction. While insights from our in vitro CSR assay shows that hypoxia can have a direct effect on purified B cells in accelerating CSR and plasmacytic differentiation, it will be a goal of future studies to identify which cell types are most directly affected by oxygen tension in vivo (e.g. TFH or GC B cells). From our studies of reversing in vivo hypoxia utilizing supplemental oxygen, we observed greater reductions in GC B cell frequencies than in comparable reductions in TFH. These data coupled with our observation that hypoxia can accelerate overall proliferation of B cells in hypoxia in vitro, brings up the possibility that our observation of reduced TFH frequencies may be secondary to the reduction in GC B cells, as GC B cells have been shown to support the maintenance of TFH (29). However, it certainly is a possibility that GC TFH may be functionally affected in 60% O2 which contributes to reduction in CSR to IgG1 as IL-4 secretion has been shown to dramatically increase in T cells cultured in hypoxic conditions (31), and IL-4 is dramatically up-regulated in GC TFH when compared to non GC TFH (13). These interpretations raise a unique point that T and B cells appear to react quite differently in hypoxia, in that secondary diversification such as CSR and early proliferation and as well as apoptosis are promoted in B cells under hypoxic conditions, while proliferation of CD4 T cells is suppressed in hypoxic culture conditions (30) but cytokine production (such as IL-4 and IFN gamma) is increased (31). Extrapolating and coupling these observations to the function of TFH and B cells in the GC makes sense in the fact that GC B cells are intensely proliferating, undergoing secondary diversification, and highly apoptotic, while TFH are not heavily proliferative in the GC but able to provide B cell help through multiple mechanisms and thus limiting for clonal selection (32). It is certainly possible that hypoxia within the GC microenvironment may be a significant environmental requirement of the GC so that B cells and TFH develop their uniquely opposite but complimentary functions. Future studies will be needed to investigate these possibilities. Our study suggests the possibility of unintended, negative consequences of supplemental oxygenation following vaccination or during ongoing humoral responses to infection. On the other hand, our study raises the possibility that supplemental oxygen could be utilized to suppress GCs which produce pathogenic antibodies such as in systemic lupus erythematosus (33). We view this study as opening the door to further investigations of the functional role of hypoxia within the GC microenvironment. Supplementary Material 1 We thank Susan Ohman for careful reading of the manuscript. We would like to thank Dr. Erin Cram for use of her microscope. This work was supported by NIH Grants: U19 AI 091693 to E. Reinherz, M. Sitkovsky and G. Kelsoe. Abbreviations ASC antibody secreting cell CSR class switch recombination GC Germinal Center HIF hypoxia inducible factor mTOR mammalian target of rapamycin NIP-BSA (nitroiodophenyl) acetyl bovine serum albumin NP-OVA (nitrophenyl) acetyl ovalbumin NP-CGG (nitrophenyl) acetyl chicken gamma globulin TFH T follicular helper UPR unfolded protein response Figure 1 The germinal center (GC) microenvironment is hypoxic and poorly vascularized. A) Tissue histology of hypoxic marker Hypoxyprobe of splenic GCs. Magnification is 20x. B) Flow cytometric analysis of splenic GCs for Hypoxyprobe staining. Red histogram is GC gate, grey histogram is non GC B cell gate. Gated on scatter, B220+ CD4-. C) Quantification of GC B cells depicted in B. D) Tissue histology of splenic GCs of mice perfused with fluorescent tomato lectin to stain vasculature. E) Schematic of masking strategy for determining percent GC and follicle area within 40μm of lectin staining. F) Quantification of GC area that is within 40μm of perfused vessels. Each dot represents one GC or B cell follicle. Each bar represents an individual mouse. Bars placed at mean. Data is of splenic GCs 8 days following i.p. immunization with either NP-OVA/Alum or NP-CGG/Alum. Representative of at least two independent experiment. Four to ten mice per group. Figure 2 Hypoxic culture conditions increase GL-7 Expression on B cells. A) Representative flow cytometry plots of purified resting B cells stimulated with αIgM and αCD40 in hypoxic or normoxic culture conditions. Gated on lymphocyte, singlet, live, B220+, CD4-. B) Quantification of GL-7+ cells in A C) Relative expression of GL-7 on positive cells in A. Bars placed at geometric mean of quadruplicate wells from representative experiment. Representative of at least two independent experiments. =p<0.01 Figure 3 Hypoxia accelerates Ig class switch recombination and plasma cell formation in vitro. A-B) Flow cytometric analysis and kinetics of B cell stimulated in normoxic or hypoxic incubators to undergo class switch recombination with αCD40 and IL-4. Gated on lymphocyte, live, singlet, CD138-.C) RT-PCR analysis of looped-out circular IgG1 transcripts of B cells stimulated as in A. D, E) Representative flow cytometric plots and analysis and kinetics of plasma cell formation of cells stimulated as in A. Cells gated on lymphocyte, live, singlet. F,G) IgM and IgG1 antibody secreting cell (ASC) formation on day 4 of hypoxic or normoxic cultures stimulated as in A. =p<0.05, **p<0.01. Representative of two to ten independent experiments. Samples were run in quadruplicate. Figure 4 Respiratory hyperoxia suppresses the germinal center (GC) reaction during immunization. A) GC frequencies of mice 8 days following immunization breathing either normoxic (21%) or hyperoxic air (60%) from day 0 to 8 following immunization. Gated on lymphocyte, singlet, live, CD4-, B220+. B) Flow cytometric plots and frequencies of T follicular helper (TFH) cells of immunized mice 8 days following immunization and treated as in A. Gated on lymphocytes, singlet, live, B220-, FoxP3-, CD4+. C) Effect of breathing 60% O2 on relative ICOS expression of TFH. D) Representative IgM and IgG1 antibody secreting cell frequencies determined by ELISPOT. E) Effect of breathing 60% O2 on plasma cell frequencies, gated on lymphocytes, singlet, live, B220low. F) Effect of breathing 60% O2 on serum antigen specific IgM. G) Effect of breathing 60% O2 on serum IgG1. H-I) NP-binding GC IgM and IgG1 compartments and effect of breathing 60% O2. Gated on lymphocytes, singlet, live, CD4-, B220+, GL7+, CD38-. J) Histological images of IgG1 within GCs on day 8 of mice treated as in A. Bars placed at geometric mean. Representative plots of two independent experiments where mice were immunized with either NP-OVA/Alum or NP-CGG/Alum and analyzed 8 days following immunization. Each dot represents one mouse. Representative of at least two independent experiments. Five to ten mice per group. Disclosures: The authors R. Abbott, S. Hatfield, and M. Sitkovsky have a filed patent on this work. The online version of this article contains supplemental information. 1 Eisen HN 2014 Affinity enhancement of antibodies: how low-affinity antibodies produced early in immune responses are followed by high-affinity antibodies later and in memory B-cell responses Cancer immunology research 2 381 392 24795350 2 Eltzschig HK Carmeliet P 2011 Hypoxia and Inflammation New England Journal of Medicine 364 656 665 21323543 3 Semenza GL 2012 Hypoxia-inducible factors in physiology and medicine Cell 148 399 408 22304911 4 Wouters BG Koritzinsky M 2008 Hypoxia signalling through mTOR and the unfolded protein response in cancer Nat Rev Cancer 8 851 864 18846101 5 Eltzschig HK Sitkovsky MV Robson SC 2013 Purinergic signaling during inflammation The New England journal of medicine 368 1260 6 Kojima H Gu H Nomura S Caldwell CC Kobata T Carmeliet P Semenza GL Sitkovsky MV 2002 Abnormal B lymphocyte development and autoimmunity in hypoxia-inducible factor 1alpha -deficient chimeric mice Proceedings of the National Academy of Sciences of the United States of America 99 2170 2174 11854513 7 Piovan E Tosello V Indraccolo S Masiero M Persano L Esposito G Zamarchi R Ponzoni M Chieco-Bianchi L Dalla-Favera R Amadori A 2007 Differential regulation of hypoxia-induced CXCR4 triggering during B-cell development and lymphomagenesis Cancer research 67 8605 8614 17875700 8 Kallet RH Matthay MA 2013 Hyperoxic acute lung injury Respiratory care 58 123 141 23271823 9 Kizaka-Kondoh S Konse-Nagasawa H 2009 Significance of nitroimidazole compounds and hypoxia-inducible factor-1 for imaging tumor hypoxia Cancer science 100 1366 1373 19459851 10 Tsai AG Friesenecker B Mazzoni MC Kerger H Buerk DG Johnson PC Intaglietta M 1998 Microvascular and tissue oxygen gradients in the rat mesentery Proceedings of the National Academy of Sciences of the United States of America 95 6590 6595 9618456 11 Laszlo G Hathcock KS Dickler HB Hodes RJ 1993 Characterization of a novel cell-surface molecule expressed on subpopulations of activated T and B cells Journal of immunology (Baltimore, Md: 1950) 150 5252 5262 12 Naito Y Takematsu H Koyama S Miyake S Yamamoto H Fujinawa R Sugai M Okuno Y Tsujimoto G Yamaji T Hashimoto Y Itohara S Kawasaki T Suzuki A Kozutsumi Y 2007 Germinal center marker GL7 probes activation-dependent repression of N-glycolylneuraminic acid, a sialic acid species involved in the negative modulation of B-cell activation Molecular and cellular biology 27 3008 3022 17296732 13 Yusuf I Kageyama R Monticelli L Johnston RJ Ditoro D Hansen K Barnett B Crotty S 2010 Germinal center T follicular helper cell IL-4 production is dependent on signaling lymphocytic activation molecule receptor (CD150) Journal of immunology (Baltimore, Md: 1950) 185 190 202 14 Stavnezer J Guikema JE Schrader CE 2008 Mechanism and regulation of class switch recombination Annual review of immunology 26 261 292 15 Kinoshita K Harigai M Fagarasan S Muramatsu M Honjo T 2001 A hallmark of active class switch recombination: transcripts directed by I promoters on looped-out circular DNAs Proceedings of the National Academy of Sciences of the United States of America 98 12620 12623 11606740 16 Pasqualucci L Kitaura Y Gu H Dalla-Favera R 2006 PKA-mediated phosphorylation regulates the function of activation-induced deaminase (AID) in B cells Proceedings of the National Academy of Sciences of the United States of America 103 395 400 16387847 17 Basu U Chaudhuri J Alpert C Dutt S Ranganath S Li G Schrum JP Manis JP Alt FW 2005 The AID antibody diversification enzyme is regulated by protein kinase A phosphorylation Nature 438 508 511 16251902 18 McBride KM Gazumyan A Woo EM Schwickert TA Chait BT Nussenzweig MC 2008 Regulation of class switch recombination and somatic mutation by AID phosphorylation The Journal of experimental medicine 205 2585 2594 18838546 19 Cheng HL Vuong BQ Basu U Franklin A Schwer B Astarita J Phan RT Datta A Manis J Alt FW Chaudhuri J 2009 Integrity of the AID serine-38 phosphorylation site is critical for class switch recombination and somatic hypermutation in mice Proceedings of the National Academy of Sciences of the United States of America 106 2717 2722 19196992 20 Boulianne B Rojas OL Haddad D Zaheen A Kapelnikov A Nguyen T Li C Hakem R Gommerman JL Martin A 2013 AID and caspase 8 shape the germinal center response through apoptosis Journal of immunology (Baltimore, Md: 1950) 191 5840 5847 21 Takahashi Y Cerasoli DM Dal Porto JM Shimoda M Freund R Fang W Telander DG Malvey EN Mueller DL Behrens TW Kelsoe G 1999 Relaxed negative selection in germinal centers and impaired affinity maturation in bcl-xL transgenic mice The Journal of experimental medicine 190 399 410 10430628 22 Marshall JL Zhang Y Pallan L Hsu MC Khan M Cunningham AF MacLennan IC Toellner KM 2011 Early B blasts acquire a capacity for Ig class switch recombination that is lost as they become plasmablasts European journal of immunology 41 3506 3512 21932446 23 Cattoretti G Buttner M Shaknovich R Kremmer E Alobeid B Niedobitek G 2006 Nuclear and cytoplasmic AID in extrafollicular and germinal center B cells Blood 107 3967 3975 16439679 24 Odegard JM Marks BR DiPlacido LD Poholek AC Kono DH Dong C Flavell RA Craft J 2008 ICOS-dependent extrafollicular helper T cells elicit IgG production via IL-21 in systemic autoimmunity The Journal of experimental medicine 205 2873 2886 18981236 25 Gorfien JL Hard R Noble B Brodsky L 1999 Quantitative study of germinal center area in normal and diseased tonsils using image analysis The Annals of otology, rhinology, and laryngology 108 398 402 26 Wittenbrink N Klein A Weiser AA Schuchhardt J Or-Guil M 2011 Is there a typical germinal center? A large-scale immunohistological study on the cellular composition of germinal centers during the hapten-carrier-driven primary immune response in mice Journal of immunology (Baltimore, Md: 1950) 187 6185 6196 27 Steiniger B Trabandt M Barth PJ 2011 The follicular dendritic cell network in secondary follicles of human palatine tonsils and spleens Histochemistry and cell biology 135 327 336 21394653 28 Steiniger BS 2015 Human spleen microanatomy: why mice do not suffice Immunology 145 334 346 25827019 29 Baumjohann D Preite S Reboldi A Ronchi F Ansel KM Lanzavecchia A Sallusto F 2013 Persistent antigen and germinal center B cells sustain T follicular helper cell responses and phenotype Immunity 38 596 605 23499493 30 Ohta A Madasu M Subramanian M Kini R Jones G Chouker A Ohta A Sitkovsky M 2014 Hypoxia-induced and A2A adenosine receptor-independent T-cell suppression is short lived and easily reversible International immunology 26 83 91 24150242 31 Roman J Rangasamy T Guo J Sugunan S Meednu N Packirisamy G Shimoda LA Golding A Semenza G Georas SN 2010 T-cell activation under hypoxic conditions enhances IFN-gamma secretion American journal of respiratory cell and molecular biology 42 123 128 19372249 32 Victora GD Nussenzweig MC 2012 Germinal centers Annual review of immunology 30 429 457 33 Woods M Zou YR Davidson A 2015 Defects in Germinal Center Selection in SLE Front Immunol 6 425 26322049
PMC005xxxxxx/PMC5123813.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101302316 33345 Cell Host Microbe Cell Host Microbe Cell host & microbe 1931-3128 1934-6069 27773535 5123813 10.1016/j.chom.2016.09.015 NIHMS821053 Article N6-methyladenosine in Flaviviridae viral RNA genomes regulates infection Gokhale Nandan S. 1 McIntyre Alexa B.R. 312 McFadden Michael J. 1 Roder Allison E. 1 Kennedy Edward M. 1 Gandara Jorge A. 3 Hopcraft Sharon E. 4 Quicke Kendra M. 56 Vazquez Christine 1 Willer Jason 1 Ilkayeva Olga R. 7 Law Brittany A. 2 Holley Christopher L. 2 Garcia-Blanco Mariano A. 811 Evans Matthew J. 4 Suthar Mehul S. 56 Bradrick Shelton S. 8 Mason Christopher E. 391013 Horner Stacy M. 121314 1 Department of Molecular Genetics and Microbiology, Duke University Medical Center, Durham, NC 27710 2 Department of Medicine, Duke University Medical Center, Durham, NC 27710 3 Department of Physiology and Biophysics, Weill Cornell Medicine, New York, NY 10021 4 Department of Microbiology, Icahn School of Medicine at Mount Sinai, New York, NY 10029 5 Department of Pediatrics, Division of Infectious Diseases, Emory University School of Medicine, Atlanta, GA 30322 6 Emory Vaccine Center, Yerkes National Primate Research Center, Atlanta, GA 30329 7 Duke Molecular Physiology Institute, Duke University, Durham NC 27701 8 Department of Biochemistry and Molecular Biology, University of Texas Medical Branch, Galveston, TX 77555 9 The HRH Prince Alwaleed Bin Talal Bin Abdulaziz Alsaud Institute for Computational Biomedicine, Weill Cornell Medicine, New York, NY 10021 10 The Feil Family Brain and Mind Research Institute, Weill Cornell Medicine, New York, NY 10021 11 Programme in Emerging Infectious Disease, Duke-NUS Medical School, Singapore 12 Tri-Institutional Program in Computational Biology and Medicine, New York City, NY 10065 13 Co-corresponding 14 Lead Contact Correspondence: [email protected] (C.E.M; @mason_lab), [email protected] (S.M.H; @thehornerlab) 12 10 2016 20 10 2016 9 11 2016 09 11 2017 20 5 654665 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. SUMMARY The RNA modification N6-methyladenosine (m6A) post-transcriptionally regulates RNA function. The cellular machinery that controls m6A includes methyltransferases and demethylases that add or remove this modification as well as m6A-binding YTHDF proteins that promote the translation or degradation of m6A-modified mRNA. We demonstrate that m6A modulates infection by hepatitis C virus (HCV). Depletion of m6A-methyltransferases or an m6A-demethylase respectively increases and decreases infectious HCV particle production. During HCV infection, YTHDF proteins relocalize to lipid droplets, sites of viral assembly, and their depletion increases infectious viral particles. We further mapped m6A sites across the HCV genome and determine that inactivating m6A in one viral genomic region increases viral titer without affecting RNA replication. Additional mapping of m6A on the RNA genomes of other Flaviviridae, including dengue, Zika, yellow fever, and West Nile virus, identifies conserved regions modified by m6A. Together, this work identifies m6A as a conserved regulatory mark across Flaviviridae genomes. Graphical Abstract INTRODUCTION The chemical modification of RNA is an important post-transcriptional regulator of RNA. Of the many known RNA modifications, N6-methyladenosine (m6A) is the most abundant internal modification of eukaryotic messenger RNAs (mRNAs), contributing to RNA structure, localization, and function (Fu et al., 2014; Meyer and Jaffrey, 2014). m6A regulates many biological processes, including stress responses, fertility, stem cell differentiation, circadian rhythms, miRNA biogenesis, and cancer (Li and Mason, 2014; Saletore et al., 2012; Yue et al., 2015; Zhou et al., 2015). However, very little is known about its effects on microbial infection. m6A has long been known to be present in the RNA transcripts of viruses with nuclear replication such as influenza A virus, simian virus 40, Rous sarcoma virus, avian sarcoma virus, and adenovirus (Dimock and Stoltzfus, 1977; Kane and Beemon, 1985; Krug et al., 1976; Lavi and Shatkin, 1975; Sommer et al., 1976). More recently, we and others have shown that m6A serves as a positive regulator of human immunodeficiency virus-1 (HIV-1), a retrovirus with a nuclear replication step (Kennedy et al., 2016; Lichinchi et al., 2016; Tirumuru et al., 2016). However, a role for m6A in regulating the life cycle of viruses that replicate exclusively in the cytoplasm, such as viruses within the Flaviviridae family, has been unexplored. Flaviviridae, including Zika (ZIKV), dengue (DENV), West Nile (WNV), yellow fever (YFV), and hepatitis C virus (HCV), represent both established and emerging pathogens. They contain a positive-sense, single-stranded RNA genome that encodes a viral polyprotein, and utilize similar replication strategies. RNA-based regulation of these viral genomes plays a fundamental role in their infection, such as the liver-specific microRNA miR-122 for HCV replication, RNA structural elements for HCV and DENV replication, and 2’-O methylation of the 5’ cap of WNV RNA for immune evasion and WNV replication (Bidet and Garcia-Blanco, 2014; Hyde et al., 2014; Jopling et al., 2005; Mauger et al., 2015; Pirakitikulr et al., 2016). The cellular machinery that regulates m6A includes proteins that act as “writers,” “erasers,” and “readers” of m6A. The addition of m6A on mRNA, which occurs at a consensus motif DRAmCH (where D=G/A/U, R=G>A, and H=U/C/A), is mediated by a methyltransferase complex containing the methyltransferase-like (METTL) enzymes METTL3 and METTL14 and the co-factors Wilms tumor-1 associated protein (WTAP) and KIAA1429 (Fu et al., 2014; Liu et al., 2014; Meyer and Jaffrey, 2014; Schwartz et al., 2014; Yue et al., 2015). The removal of m6A from mRNA is catalyzed by the demethylases fat mass and obesity associated protein (FTO) or α-ketoglutarate-dependent dioxygenase AlkB homolog 5 (ALKBH5) (Jia et al., 2011; Zheng et al., 2013). The cytoplasmic YTH-domain family 1 (YTHDF1), YTHDF2, and YTHDF3 proteins bind to m6A through their C-terminal YTH domain. Functionally, YTHDF1 promotes the translation of m6A-modified mRNA, while YTHDF2 targets m6A-modified mRNAs for degradation (Wang et al., 2014; Wang et al., 2015). The function of YTHDF3 is still unknown. The discovery of these proteins and the development of high-throughput m6A-mapping techniques have led to many recent insights in the function of m6A (Dominissini et al., 2012; Fu et al., 2014; Linder et al., 2015; Meyer et al., 2012). Nonetheless, many aspects of the regulation of specific mRNAs by m6A remain unexplored. Here, we define a role for m6A in regulating the life cycle of HCV. We demonstrate that the m6A-methyltransferases negatively regulate the production of infectious HCV particles, and that the m6A-binding YTHDF proteins all relocalize to sites of HCV particle production and suppress this stage of viral infection. Importantly, we map m6A across the HCV RNA genome and show that preventing m6A at one of these regions enhances viral titer by increasing the interaction of the HCV RNA with the HCV Core protein. Finally, we describe viral RNA m6A-epitranscriptomic maps for several other Flaviviridae, including ZIKV, DENV, WNV, and YFV. Together, our data reveal that m6A regulates HCV infection and set the stage for the exploration of the function of m6A within the broader Flaviviridae family of viruses. RESULTS The m6A machinery regulates HCV particle production To determine if m6A regulates HCV infection, we depleted the m6A-methyltransferases METTL3 and METTL14 (METTL3+14) by siRNA in Huh7 liver hepatoma cells and infected these cells with HCV. Immunoblot analysis of cell extracts harvested at 72 hours post-infection (hpi) revealed that METTL3+14 depletion significantly increased the abundance of the HCV NS5A protein, a marker of viral replication, relative to its level in cells treated with non-targeting control siRNA (Fig. 1A). Conversely, depletion of the m6A-demethylase FTO decreased HCV NS5A levels relative to the control (Fig. 1A). Furthermore, we found that the percentage of HCV-positive cells increased after METTL3+14 depletion and decreased after FTO depletion (Fig. 1B-C and S1A). This change in HCV-positive cells occurred only after 24 hpi, suggesting that viral entry was unaffected by m6A machinery depletion. Depletion of the m6A machinery did not impair cell viability during infection (Fig. S1B). Additionally, HCV infection slightly reduced METTL3 protein levels in total cellular extracts, while METTL14 and FTO were unaffected (Fig. S1C-D). Thus, the m6A-methyltransferases negatively regulate HCV infection, while the m6A-demethylase positively regulates HCV infection. We next defined the stage of the HCV life cycle regulated by the m6A machinery. Depletion of METTL3+14 significantly increased the production of infectious virus and viral RNA in the supernatant as compared to control siRNA at 72 hpi (Fig. 1D-E). Conversely, depletion of FTO decreased infectious virus and HCV RNA in the supernatant (Fig. 1D-E), without altering the viral specific infectivity (Fig. S1E). Depletion of ALKBH5 did not affect viral titer or protein levels, indicating that this demethylase does not influence the HCV life cycle (Fig. S1F). We next tested if the altered HCV titer after m6A machinery depletion was due to altered viral RNA replication. In these experiments, we used Huh7.5 CD81 KO cells, in which the essential HCV entry factor CD81 (Zhang et al., 2004) was deleted by CRISPR/Cas9 resulting in cells permissive for HCV RNA replication and viral particle production following viral RNA transfection that are unable to support subsequent rounds of viral infection (Fig. S1G-I). In these cells, we depleted METTL3+14 or FTO by siRNA, transfected the cells with in vitro transcribed RNA of the HCV reporter virus JFH1-QL/GLuc2A, and measured HCV RNA replication by assaying for secreted Gaussia luciferase (Yamane et al., 2014). Depletion of METTL3+14 or FTO had no effect on Gaussia luciferase levels as compared to control over the time course, while our negative control RNA containing a point mutation in the viral RNA-dependent RNA polymerase (Pol−) did not replicate (Fig. 1F). These data indicate that m6A dynamics do not regulate HCV translation or RNA replication, but do regulate the production or release of infectious viral particles. Changes in expression of the m6A machinery have been shown to affect cellular gene expression (Dominissini et al., 2012; Meyer et al., 2012; Wang et al., 2014), which could indirectly regulate the HCV life cycle, for example by inducing antiviral interferon-stimulated genes (ISGs). While we did not find consistent changes in ISG mRNA levels following loss of the m6A machinery during HCV infection (48 hpi), FTO depletion did slightly increase the expression of IFITM1, which is known to restrict HCV entry (Fig. S1J) (Wilkins et al., 2013). This slight increase occurred at both 24 and 48 hpi, although the percentage of HCV-positive cells following FTO depletion is the same as control and METTL3+14 depletion at 24 hpi (Figs. 1B and S1J-K). Therefore, the observed changes in infectious virus following depletion of the m6A machinery are not solely a result of an altered antiviral response in these cells. Rather, these data suggest that m6A acts directly on the HCV RNA genome to regulate HCV particle production. The m6A-binding YTHDF proteins negatively regulate HCV particle production Given that the m6A machinery regulates infectious HCV particle production, we next tested if the known mediators of m6A function, the RNA-binding YTHDF proteins, similarly regulate the HCV life cycle. Depletion of any of the YTHDF proteins did not increase HCV NS5A protein levels at 48 hpi or HCV RNA replication of the HCV reporter (JFH1-QL/GLuc2A) over 72 hours in Huh7.5 CD81 KO cells. However, by 72 hpi, the levels of infectious HCV particles and HCV RNA in the supernatant were increased at least 2-fold as compared to control (Fig. 2A-D). Importantly, depletion of YTHDF proteins did not affect cell viability, nor did HCV infection alter their expression (Fig. S2A-B). Collectively, these data suggest that the YTHDF proteins negatively regulate infectious HCV production without affecting overall HCV RNA replication. We next tested whether YTHDF proteins bind to HCV RNA by RNA-immunoprecipitation (RIP). We found that FLAG-YTHDF ribonucleoprotein complexes (RNPs) enriched HCV RNA relative to the input, demonstrating that these proteins bind to viral RNA (Fig. 2E). Thus, YTHDF protein binding to HCV RNA may mediate their regulation of HCV particle production. This led us to examine the sub-cellular localization of the YTHDF proteins during HCV infection. YTHDF proteins relocalize to lipid droplets during HCV infection HCV particle assembly occurs around cytosolic lipid droplets in close association with ER membranes. Indeed, HCV RNA and proteins, including NS5A and Core (the capsid protein), as well as several host RNA-binding proteins that regulate HCV infection, accumulate around lipid droplets (Ariumi et al., 2011; Chatel-Chaix et al., 2013; Miyanari et al., 2007; Pager et al., 2013; Poenisch et al., 2015). Therefore, we analyzed the subcellular localization of YTHDF proteins after HCV infection in Huh7 cells by confocal microscopy. While YTHDF proteins were distributed in the cytoplasm in uninfected cells, in HCV-infected cells all three YTHDF proteins (both endogenous and over-expressed) were enriched around lipid droplets (Figs. 3A-B and S3A), where they colocalized with the HCV Core protein. We did not observe this relocalization in Huh7 cells stably expressing a sub-genomic HCV replicon that lacks the HCV structural genes and cannot produce viral particles (Fig. S3B) (Wang et al., 2003), suggesting that a fully productive HCV infection is required to trigger the relocalization of the YTHDF proteins around lipid droplets. HCV RNA is modified by m6A We and others recently mapped m6A on HIV-1 mRNA and showed that it regulates viral gene expression (Kennedy et al., 2016; Lichinchi et al., 2016; Tirumuru et al., 2016). Although m6A has not been found in RNAs from viruses that replicate in the cytoplasm, our findings above (Figs. 1-3), led us to hypothesize that the HCV RNA genome is modified by m6A during infection. To test this, we used an antibody that specifically recognizes m6A to perform methyl-RNA immunoprecipitation (MeRIP) on total RNA harvested from HCV-infected cells followed by RT-qPCR to detect enriched RNAs. HCV RNA in the eluate was specifically enriched by the anti-m6A antibody, but not IgG, as was a known m6A-modified mRNA SON, while an mRNA with little m6A-modification HPRT1, was not (Wang et al., 2014) (Fig. 4A). Ultra high pressure liquid chromatography coupled with tandem mass spectrometry (UPLC-MS/MS) analysis of viral RNA captured from HCV-infected Huh7 cells using specific antisense oligonucleotides proved that HCV RNA contains m6A, with a ratio of m6A/A of approximately 0.16% (Fig. S4A-B). Interestingly, the anti-m6A antibody did not enrich HCV RNA isolated from cell supernatants to the same degree as intracellular viral RNA (Fig. 4A). We next mapped the sites of the HCV RNA genome modified by m6A using MeRIP followed by sequencing (MeRIP-seq), as previously described (Dominissini et al., 2013; Meyer et al., 2012). We identified 19 peaks across the HCV RNA genome common to both experimental replicates (Fig. 4B, S6; Table S1). These data present evidence that HCV, which replicates exclusively in the cytoplasm, is marked by m6A during infection. As HCV replicates in the cytoplasm in association with intracellular membranes, for its RNA to undergo m6A-modification, the m6A-methyltransferases must also exist in the cytoplasm. Indeed, our immunoblot analysis of isolated nuclear and cytoplasmic fractions from mock or HCV-infected Huh7 cells reveals that METTL3, METTL14, and FTO are all present in both the nucleus and cytoplasm, where they could interact with viral RNA (Fig. S1C-D). This is in concordance with reports that have detected both METTL3 and m6A-methyltransferase activity in cytoplasmic extracts (Chen et al., 2015; Harper et al., 1990; Lin et al., 2016). Therefore, these data reveal that the m6A machinery are in the cytoplasm where they can modify cytoplasmic HCV RNA. As the cellular function of m6A is carried out by the YTHDF proteins, which bound to HCV RNA (see Fig. 2E), we hypothesized that one or more of the YTHDF proteins would bind to functionally relevant m6A sites on the HCV RNA genome. We directly mapped these YTHDF binding sites on the viral genome using photoactivatable ribonucleoside-enhanced crosslinking and immunoprecipitation (PAR-CLIP) in HCV-infected Huh7 single cell clones stably expressing these proteins or green fluorescent protein (GFP) (Fig. S4C) (Hafner et al., 2010; Kennedy et al., 2016). We identified 42 different sites on the HCV RNA genome that were bound by at least one YTHDF protein and not by GFP, and which contained the characteristic T-to-C transition that derives from reverse transcription of cross-linked 4SU residues (Table S2). Surprisingly, only two high-confidence YTHDF-binding sites (bound by >1 YTHDF protein) overlapped with the m6A peaks identified by all replicates of MeRIP-seq and only 55% of the YTHDF binding sites contained the DRAmCH motif required for m6A (Table S2). Taken together, these data build a map of m6A and YTHDF binding sites on the HCV RNA genome. m6A-abrogating mutations in the HCV E1 genomic region increase viral particle production To elucidate the functional relevance of a specific m6A site on the HCV genome, we made mutations in the genome to inactivate this modification. We identified only 1 region of the HCV genome, within the viral E1 gene, that both contains m6A and is also bound by YTHDF proteins at sites with consensus DRAmCH motifs (Table S1-2). This region of the genome has previously been shown to lack major RNA secondary structure (Pirakitikulr et al., 2016) and contains a cluster of 4 potential m6A sites (Fig. 5A). Importantly, comparative sequence analysis of the nucleotides in these sites revealed that the first m6A site is identical in 72 strains of genotype 2A, while the m6A motif in the latter three sites is conserved between 26 representative strains of HCV from different genotypes (Fig. S5A). We then mutated either the A or the C within the consensus site to a U in the four identified m6A sites in the E1 gene to construct HCV-E1mut. These mutations will abrogate the potential for m6A-modification (Kane and Beemon, 1987), without altering the encoded amino acid sequence (See Fig. 5A). To determine the role of these m6A sites in the HCV life cycle, we electroporated WT and E1mut HCV RNA into Huh7 cells and measured the production of infectious virus at 48 hpi. E1mut produced nearly 3-fold more viral titer in supernatant than WT, while the Pol− RNA did not produce any titer (Fig. 5B). E1mut also increased both intracellular and extracellular titer, suggesting that these mutations increased viral particle assembly (Fig. S5B). To determine if abrogation of the E1 m6A sites affected HCV RNA replication, we then measured replication of the WT or E1mut JFH1-QL/GLuc2A reporter after transfection into Huh7.5 CD81 KO cells. The E1 mutations did not alter HCV RNA replication over a time course, (Fig. 5C), nor did they alter the levels of viral Core protein (Fig. 5D). Together, these data suggest that m6A within the E1 gene negatively regulates infectious HCV particle production, similar to our findings with depletion of the m6A-methyltransferases and YTHDF proteins. While the YTHDF proteins bind to multiple sites on HCV RNA, comparison of the MeRIP-seq with the PAR-CLIP data suggest that their binding to the HCV RNA genome is not always m6A-dependent (Fig. 4B; Table S2). Therefore, to test whether the m6A-abrogating mutations in E1 affect binding by YTHDF proteins within this region, we measured FLAG-YTHDF2 binding to a reporter RNA containing 100 nucleotides of WT or E1mut, allowing us to isolate the interaction of a single m6A-region with a single YTHDF protein. Mutation of the m6A sites within the E1 region reduced binding of FLAG-YTHDF2 by 50% as compared to the WT by YTHDF2 RIP, while FLAG-YTHDF2 bound equally to a known m6A-modified mRNA SON in both conditions (Fig. 5E). Furthermore, depletion of YTHDF1 did not increase extracellular HCV RNA produced by cells infected with E1mut HCV over cells treated with control siRNA (Fig. S5C). The HCV Core protein binds to the HCV RNA genome during assembly of viral particles. Intriguingly, Core protein is known to bind to HCV RNA around the E1 region that contains our identified m6A sites (Shimoike et al., 1999). To test whether m6A in E1 influences Core binding to viral RNA, we immunoprecipitated Core RNP complexes from cells electroporated with WT or E1mut HCV RNA. We found that mutation of the m6A sites within the E1 region increases HCV RNA binding to the Core protein by nearly 2-fold as compared to WT (Fig. 5F). Together, these data suggest that YTHDF proteins bind to the m6A sites within the HCV E1 region to mediate the negative regulation of infectious HCV particle production, while the Core protein binds to viral RNA genomes lacking m6A within the E1 region for packaging into nascent viral particles. Mapping of m6A within the viral RNA genomes of the Flaviviridae family of viruses As we found that the HCV RNA genome contains m6A, we wanted to investigate the location of m6A on the RNA genomes of other members of the Flaviviridae family. We performed MeRIP-seq in duplicate on RNA isolated from Huh7 cells infected with DENV (DENV2-NGC), YFV (17D), WNV (TX), and ZIKV (PR2015 or DAK). Our data identified reproducible m6A-sites within all five viral genomes (Fig. 6A-E, S6; Table S3). Intriguingly, some of m6A sites on these viral genomes occurred within similar genetic regions among all the Flaviviridae (Fig. 6F). In particular, the NS3 and NS5 genes contained m6A peaks, reminiscent of the pattern on the HCV RNA genome, suggesting a conserved role for these sites in regulating these viral life cycles. Further, similar to HCV, DENV and ZIKV (PR2015) contained an m6A peak in the envelope gene. Therefore, these data suggest a potentially conserved set of m6A-epitranscriptome sites in the Flaviviridae family that could regulate viral RNA function, virulence, and transmission. DISCUSSION The function of m6A in regulating host and viral infection is only now emerging, even though nuclear-replicating viruses have been known to contain m6A since the 1970s (Dimock and Stoltzfus, 1977; Kane and Beemon, 1985; Krug et al., 1976; Lavi and Shatkin, 1975; Sommer et al., 1976). Recent studies have established a pro-viral role for m6A during HIV-1 infection (Kennedy et al., 2016; Lichinchi et al., 2016; Tirumuru et al., 2016). In our study, in which we define function for m6A and its cellular machinery in regulating the positive-strand RNA genome of the cytoplasmic virus HCV, we find that m6A negatively regulates HCV particle production. Further, we find that the positive-strand RNA genomes of other viruses within the Flaviviridae family, including two strains of ZIKV, are modified by m6A in conserved genomic regions. Taken together, this work reveals that Flaviviridae RNA genomes harbor RNA regulatory marks that could impact their life cycles and virulence. The known enzymes and RNA-binding proteins that regulate m6A also regulate the life cycle of HCV. Depletion of the m6A-methyltransferases METTL3 and METTL14 increases the rate of HCV infection by promoting infectious viral particle production without affecting viral RNA replication. Conversely, depletion of the m6A-demethylase FTO, but not ALKBH5, has the opposite effect (Fig. 1). These effects do not appear to be caused by dysregulated induction of host ISGs after depletion of the m6A machinery, as changes in ISG expression were minimal (Fig. S1). Instead, we hypothesize that the m6A machinery directly modulates the levels of m6A on the HCV genome to regulate its function, and this is supported by our finding that HCV RNA contains m6A. While it is known that m6A functions on host mRNAs to regulate their stability, translation, localization, and interactions with RNA-binding proteins (Fu et al., 2014), we hypothesize that the function of m6A in HCV RNA is not due to regulation of HCV RNA stability or translation because our studies of HCV RNA replication using a reporter virus found no change in reporter levels following depletion of the m6A machinery. Rather, our data suggest that m6A regulates infectious viral particle production through interactions of the viral RNA with host and viral proteins. Since the “writers” (METTL3+14) and an “eraser” (FTO) of m6A regulated HCV particle production, it was reasonable to hypothesize that the m6A-binding YTHDF “reader” proteins would have a similar effect. In fact, all three YTHDF proteins bound to HCV RNA at similar sites and their depletion increased HCV particle production suggesting that their effect on HCV particle production was due to binding HCV RNA (Fig. 2 and Table S2). While YTHDF1 and YTHDF2 have been found to have divergent functions on host mRNAs, in our study all three YTHDF proteins acted similarly to suppress HCV (Wang et al., 2014; Wang et al., 2015). Likewise, during HIV-1 infection, all three YTHDF proteins function similarly to each other, although they have been described to have both pro- and anti-HIV function (Kennedy et al., 2016; Tirumuru et al., 2016). During HCV infection, YTHDF regulatory function is likely related to their relocalization to lipid droplets, the sites of viral assembly (Fig. 3). Many RNA-binding proteins relocalize to lipid droplets in HCV-infected cells and regulate HCV particle production (Ariumi et al., 2011; Chatel-Chaix et al., 2013; Pager et al., 2013; Poenisch et al., 2015). In fact, many of these proteins are known to interact with YTHDF proteins, suggesting that these interactions could regulate HCV particle production (Schwartz et al., 2014; Wang et al., 2015). Consequently, it will be important in the future to identify any YTHDF protein-protein interactions enriched during HCV infection, which may point to a regulatory network of RNA-binding proteins that impact infectious HCV particle production. We found that about 50% of YTHDF protein binding sites identified on HCV RNA using PAR-CLIP overlapped with MeRIP-seq m6A peaks (Fig. 4). These results are similar to previous studies examining the overlap of YTHDF1 or YTHDF2 PAR-CLIP with MeRIP-seq data, which have found about a 60% overlap (Wang et al., 2014; Wang et al., 2015). We hypothesize that the non-overlapping YTHDF binding sites in HCV RNA represent m6A sites not detected by MeRIP-seq due to biological variation, technical noise, or potentially sites that might be bound by YTHDF proteins in an m6A-independent fashion. Indeed, a recent report found that YTHDF proteins bound to an in vitro transcribed, hence non-methylated, HCV RNA genome (Rios-Marco et al., 2016). Therefore, future studies could very well reveal functions of the YTHDF proteins that are independent of m6A during the HCV life cycle. To discern the function of an m6A site on HCV RNA during infection, we abrogated m6A-modification in the E1 region of HCV by mutation. This E1mut virus produced higher viral titers than the WT virus (Fig. 5), similar to what we found with METTL3+14 and YTHDF depletion, suggesting a conserved regulatory mechanism between both m6A and the YTHDF proteins at this site. Interestingly, the presence of these mutations in E1 increased HCV RNA binding to Core protein, while reducing binding to YTHDF2. This suggests that interactions of the HCV RNA with Core are regulated by m6A such that viral genomes lacking m6A in the E1 region are preferentially segregated for packaging into nascent virions. Therefore, we hypothesize that the presence or absence of m6A in E1 facilitates a competition between YTHDF protein and HCV Core binding to the viral genome, leading to the cellular retention or packaging of HCV RNA, respectively. Since RNA viruses can rapidly evolve under selection pressure, the maintenance of m6A sites on the HCV genome suggests that m6A must confer an evolutionary advantage to the virus. In HCV, whose pathology is characterized by chronic progression during infection in the liver, a slower replication rate has been linked to persistent infection through an evasion of immune surveillance (Bocharov et al., 2004). Therefore, m6A may boost viral fitness by allowing HCV to establish slow, persistent infections. Intriguingly, Pirakitikulr and colleagues recently identified a conserved stem loop in the E1 coding region, just downstream of our identified m6A sites, that suppresses viral particle production without affecting viral RNA replication (Pirakitikulr et al., 2016). This raises the possibility that within the E1 region, multiple RNA elements, including m6A, play a role in segregating the RNA genome between stages of the HCV life cycle. The function of the other m6A sites on the HCV RNA genome remains unknown. Since many of these sites do not overlap with YTHDF protein binding sites, they may directly modify HCV RNA structure, or recruit alternative m6A “readers” such as HNRNPA1/B2, eIF3, or even METTL3 itself (Alarcon et al., 2015; Lin et al., 2016; Meyer et al., 2015). They may also contribute to antiviral innate immune evasion, as the presence of m6A on RNA has been shown to reduce its activation of toll-like receptor 3 signaling (Kariko et al., 2005). While we did not identify m6A in the known poly-U/UC pathogen-associated molecular patterns in the 3’ untranslated region (UTR) of the HCV genome, we did find that YTHDF2 binds in close-proximity to this region (Table S2), and so future studies can now begin to discern if m6A plays a role in HCV innate immune evasion. We found that 4 other Flaviviridae (DENV, YFV, ZIKV, and WNV) also contained m6A within their viral genomes. As these viruses replicate in the cytoplasm, our data reveal that m6A-methyltransferases are functional in the cytoplasm. Indeed, similar to others, we detected the m6A machinery in cytoplasmic fractions (Fig. S1C) (Chen et al., 2015; Harper et al., 1990; Lin et al., 2016). Therefore, cellular mRNAs could also be dynamically regulated by m6A-modification following export into the cytoplasm. Interestingly, these viruses each had prominent m6A peaks in NS5, which encodes their viral RNA-dependent RNA polymerase, strongly suggesting the presence of a conserved RNA regulatory element here. Both DENV and ZIKV (PR2015) contained m6A peaks within their envelope genes, similar to HCV, and future studies to determine whether these m6A sites also affect production of infectious flaviviral particles will be of interest. While the genomic RNA structures for DENV, YFV, ZIKV, and WNV have not yet been determined, these viral genomes do contain specific RNA regulatory structures, especially within their UTRs. Notably, we found that two of the mosquito-transmitted viruses, DENV and YFV, both have m6A within their 3’UTRs (Fig. 6F). In DENV, the 3’UTR has two stem loops that regulate mosquito to human transmission (Villordo et al., 2015). Therefore, it is possible that m6A patterns and functionality in the mosquito-transmitted flaviviral genomes could contribute to vector-borne transmission. Finally, we observed clear differences in m6A patterns between the the Dakar and Puerto Rican isolates of ZIKV, which represent the African and Asian lineages, respectively (Haddow et al., 2012). As these lineages have differences in human disease, with increased pathogenicity ascribed to the Asian lineage of ZIKV (Weaver et al., 2016), the differences in regulation of these viruses by m6A could contribute to these varied infection outcomes. In summary, we present global m6A-profiling of RNA viruses within the Flaviviridae family. Additionally, we provide evidence that an exclusively cytoplasmic RNA is modified by m6A. Further, we present a role of this modification in regulating HCV RNA function at the level of infectious viral particle production. This work sets the stage to broadly study the role of m6A in Flaviviridae infection, transmission, and pathogenesis. This work also has the potential to uncover regulatory strategies to inhibit replication by these established and emerging viral pathogens. EXPERIMENTAL PROCEDURES Cell lines Human hepatoma Huh7, Huh7.5, and Huh7.5 CD81 KO cells were grown in Dulbecco's modification of Eagle's medium (DMEM; Mediatech) supplemented with 10% fetal bovine serum (HyClone), 25mM HEPES and 1X non-essential amino acids (cDMEM) (Thermo-Fisher). HCV-K2040 (1B) replicon cells (Wang et al., 2003) were cultured in cDMEM containing 0.2mg/ml geneticin (Thermo-Fisher). The identity of the Huh7 and Huh7.5 cell lines was verified using the Promega GenePrint STR kit (DNA Analysis Facility, Duke University), and cells were verified as mycoplasma free by the LookOut Mycoplasma PCR detection kit (Sigma). Huh7.5 CD81 KO cells were generated by CRISPR, as described before, with details given in the Supplemental Experimental Procedures (Hopcraft et al., 2016; Hopcraft and Evans, 2015). Viral infections and generation of viral stocks HCV. Infectious stocks of a cell culture-adapted strain of genotype 2A JFH1 HCV were generated and titrated by focus-forming assay (FFU), as described (Aligeti et al., 2015). HCV infections were performed at an MOI of 0.3 for 72 hours unless noted. WNV. Working stocks of WNV isolate TX 2002-HC (WNV-TX) were generated in BHK-21 cells and titered as described (Suthar et al., 2010). WNV infections (MOI 5) were performed in Huh7 cells for 48 hours. DENV and YFV. Preparation and titering of DENV2-NGC and YFV-17D stocks has been described (Le Sommer et al., 2012; Sessions et al., 2009). DENV and YFV infections (MOI 2) were performed for 24 hours in Huh7 cells. ZIKV. ZIKV_PR2015 (PRVABC59) stocks were prepared and titered as described (Quicke et al., 2016). ZIKV_DAK (Zika virus/A.africanus-tc/SEN/1984/41525-DAK) stocks were generated and titered by FFU assay in Vero cells (Le Sommer et al., 2012). ZIKV infections (MOI 2) were performed in Huh7 cells for 24 hours. Focus forming assay for HCV titer Supernatants were collected and filtered through a 0.45μM syringe filters. Serial dilutions of supernatants were used to infect naïve Huh7.5 cells in triplicate wells of a 48-well plate. At either 48 or 72 hpi, cells were fixed, permeabilized, and immunostained with HCV NS5A antibody (1:500; gift of Charles Rice, Rockefeller University). Following binding of HRP-conjugated secondary antibody (Jackson ImmunoResearch, 1:500), infected foci were visualized with the VIP Peroxidase Substrate Kit (Vector Laboratories), and counted at 40X magnification. Titer (FFU/ml) was calculated as described (Gastaminza et al., 2006). To measure intracellular HCV titer, cells pellets were washed in PBS, resuspended in serum free media, and subjected to five rounds of freeze-thaw in a dry ice/ethanol bath. Lysate was cleared by centrifugation, and focus forming assay was performed as described above. MeRIP-seq Poly(A)+ RNA purified from at least 75μg total RNA (Poly(A) Purist Mag kit; Thermo-Fisher) extracted from HCV-, DENV-, YFV-, WNV-, ZIKV (DAK)- and ZIKV (PR2015)-infected samples was fragmented using the Ambion RNA fragmentation reagent and purified by ethanol precipitation. Fragmented RNA was heated to 75°C for 5 minutes, placed on ice for 3 minutes, and then incubated with anti-m6A antibody (5μg, Synaptic Systems, #202111) conjugated to Protein G Dynabeads (50μl, Thermo-Fisher) in MeRIP buffer (50mM Tris-HCl pH 7.4, 150mM NaCl, 1mM EDTA, 0.1% NP-40) overnight at 4°C. Beads were then washed 5X with MeRIP buffer, and bound RNA was eluted in MeRIP buffer containing 6.7mM m6A sodium salt (Sigma). Eluted RNA was purified with the Quick-RNA miniprep kit (Zymo Research) and concentrated by ethanol precipitation. Sequencing libraries were prepared from this RNA, as well as input RNA, using the TruSeq RNA-sequencing kit (Illumina). Libraries were sequenced to 1×50 base-pair reads on the Illumina HiSeq2500 at the Weill Cornell Medicine Epigenomics Core Facility. Reads were aligned to combined human (hg19) and viral genomes using STAR with a mapping quality threshold of 20. Despite the poly(A)-enrichment, a significant number of reads mapped to the viral genomes. We identified peaks using MeRIPPeR (https://sourceforge.net/projects/meripper/), which defines peaks in m6A-IP over input control read counts using Fisher's Exact test, with a minimum peak size of 100 bases. The false discovery rate was set to <0.05 using a Benjamini-Hochberg correction. Intersections between the peaks called by two replicates provided the final set of peak calls. MeRIP-RT-qPCR followed this protocol, except that total RNA was not fragmented. Eluted RNA was reverse transcribed into cDNA and subjected to RT-qPCR. Statistical Analysis Student's unpaired t-test and two-way ANOVA (with Bonferroni's correction) were used for statistical analysis of the data using the Graphpad Prism software. Graphed values are presented as mean ± SD (n=3 or as indicated); P≤0.05, P≤0.01, and *P≤0.001. Supplementary Material ACKNOWLEDGEMENTS We thank Dr. Lemon and Dr. Weeks (University of North Carolina-Chapel Hill) and Dr. Rice (Rockefeller University) for reagents; the Duke University Light Microscopy Core Facility; the Epigenomics Core Facility at Weill Cornell; and members of the Horner and Mason labs for discussion and reading of this manuscript. This work was supported by funds from the NIH: R01AI125416 (SMH, CEM), 5P30AI064518 (SMH), T32-CA009111 (AR); R25EB020393, R01NS076465, R01AI125416, R01ES021006 (CEM); R01AI089526, R01AI101431 (MGB); R01DK0951250 (MJE); U19AI083019, R56AI110516 (M.S.S). Additional funding sources: SMH (Duke Whitehead Scholarship), CV (Ford Foundation), ABRM (Tri-Institutional Training Program in Computational Biology and Medicine), CEM ((STARR - I7-A765, I9-A9-071), the Irma T. Hirschl and Monique Weill-Caulier Charitable Trusts, Bert L and N Kuggie Vallee Foundation, WorldQuant, The Pershing Square Sohn Cancer Research Alliance, NASA (NNX14AH50G, 15-15Omni2-0063), the Bill and Melinda Gates Foundation (OPP1151054), and the Alfred P. Sloan Foundation (G-2015-13964)), MGB (U-TX STARs Award), MGB and SSB (funds from UTMB), MJE (Pew Charitable Trusts, USPHS-AI07647, ACS-RSG-12-176-01-MPC, and Burroughs Wellcome Fund. Figure 1 The m6A machinery regulates infectious HCV particle production (A) Immunoblot analysis of extracts of HCV-infected Huh7 cells (72 hpi) treated with siRNAs. NS5A levels were quantified relative to tubulin (n=3). P≤0.001 by unpaired Student's t-test (B) Percent of HCV+ cells by immunostaining of NS5A and nuclei (DAPI) after siRNA. n=3, with ≥5000 cells counted per condition. P≤0.05, P≤0.001 by two-way ANOVA with Bonferroni correction. (C) Representative fields of HCV-infected cells (NS5A+, green) and nuclei (DAPI, blue) at 72 hpi from (B). (D) Focus-forming assay of supernatants harvested from Huh7 cells at 72 hpi after siRNA treatment. (E) HCV RNA in supernatants harvested from Huh7 cells 72 hpi after siRNA treatment as quantified by RT-qPCR. Data in (D) and (E) are presented as percent of viral titer or RNA relative to control siRNA P≤0.001 by unpaired Student's t-test. (F) Gaussia luciferase assay to measure HCV luciferase reporter (JFH1-QL/GLuc2A) transfected in Huh7.5 CD81 KO cells after siRNA treatment. Pol−: lethal mutation in HCV NS5B polymerase. Values for (D) and (E) are the mean ± SEM of three experiments in triplicate. Values in (B) and (F) represent the mean ± SD (n=3) and are representative of three independent experiments. See also Figure S1. Figure 2 The m6A-binding YTHDF proteins negatively regulate infectious HCV particle production (A) Immunoblot analysis of extracts of HCV-infected Huh7 cells (48 hpi) treated with indicated siRNAs. (B) Focus-forming assay of supernatants harvested from Huh7 cells at 72 hpi after siRNA treatment. (C) HCV RNA in supernatants harvested from Huh7 cells 72 hpi after siRNA treatment was quantified by RT-qPCR. Data in (B) and (C) were analyzed as the percent of titer or HCV RNA relative to cells treated with control siRNA. (D) Gaussia luciferase assay to measure HCV luciferase reporter (JFH1-QL/GLuc2A) transfected in Huh7.5 CD81 KO cells after siRNA. (E) Enrichment of HCV RNA following immunoprecipitation of FLAG-tagged YTHDF from extracts of Huh7 cells after 48 hpi. (Left panel) Captured HCV RNA was quantified by RT-qPCR as the percent of input and graphed as fold enrichment relative to vector. (Right panel) Immunoblot analysis of immunoprecipitated extracts and input. Values represent the mean ± SEM of three (C) or four (B) experiments done in triplicate. For (D-E) data are representative of three experiments and show the mean ± SD (n=3). P≤0.05, P≤0.01, P≤0.001 by unpaired Student's t-test. See also Figure S2. Figure 3 YTHDF proteins relocalize to lipid droplets during HCV infection (A) Confocal micrographs of HCV-infected, or uninfected Huh7 cells (48 hpi) immunostained with antibodies to YTHDF (green) and HCV Core (red) proteins. Lipid droplets (grey) and nuclei (blue) were labeled with BODIPY and DAPI, respectively. Zoom panels are derived from the white box in the Merge. Scale bar, 10 μm. (B) Enrichment of YTHDF proteins around lipid droplets was quantified by using ImageJ from >10 cells analyzed and graphed in box and whisker plots, representing the minimum, 1st quartile, median, 3rd quartile, and the maximum. P≤0.01, P≤0.001 by unpaired Student's t-test. See also Figure S3. Figure 4 HCV RNA is modified by m6A (A) MeRIP-RT-qPCR analysis of intracellular or supernatant RNA harvested from HCV-infected Huh7.5 cells (72 hpi) and immunoprecipitated with anti-m6A or IgG. Eluted RNA is quantified as a percent of input. Values are the mean ± SD (n=3). P≤0.01, **P≤0.001 by unpaired Student's t-test. (B) Map of m6A binding sites in the HCV RNA genome by MeRIP-seq (representative of two independent samples) of RNA isolated from HCV-infected Huh7 cells. Read coverage, normalized to the total number of reads mapping to the viral genome for each experiment, is in red for MeRIP-seq and in blue for input RNA-seq. Red bars indicate m6A peaks identified in duplicate experiments by MeRIPPeR analysis (FDR corrected q-value <0.05). See also Figure S4, Figure S6, Table S1, and Table S2. Figure 5 m6A-abrogating mutations in E1 increase infectious HCV particle production (A) Schematic of the HCV genome with the mutation scheme for altering A or C residues (red arrows) to make the E1mut virus. Consensus m6A motifs (green) and inactivating mutations (red) are shown. Dashes represent nucleotides not shown. Genomic indices match the HCV JFH-1 genome (AB047639). (B) Focus-forming assay of supernatants harvested from Huh7 cells after electroporation of WT or E1mut HCV RNA (48 h) and analyzed as the percent of viral titer relative to WT (C) Gaussia luciferase assay to measure levels of the WT, E1mut or Pol− HCV luciferase reporter (JFH1-QL/GLuc2A) transfected in Huh7.5 CD81 KO cells. (D) Immunoblot analysis of extracts of WT, E1mut or Pol− JFH1-QL/GLuc2A transfected in Huh7.5 CD81 KO cells (E) Enrichment of WT or E1mut reporter RNA or SON mRNA by immunoprecipitation of FLAG-YTHDF2 or vector from extracts of Huh7 cells. Captured RNA was quantified by RT-qPCR and graphed as percent of input. (Right panel) Immunoblot analysis of anti-FLAG immunoprecipitated extracts and input. (F) Enrichment of WT or E1mut HCV RNA by immunoprecipitation of HCV Core from extracts of Huh7 cells electroporated with the indicated viral genomes (48 h). (Lower panel) Immunoblot analysis of anti-Core immunoprecipitated extracts and input. Data are representative of two (D, E) or three (B, C, F) experiments and presented as the mean ± SD (n=3). P≤0.05, P≤0.01, *P≤0.001 by unpaired Student's t-test. See also Figure S5. Figure 6 Mapping m6A in the RNA genomes of Flaviviridae Read coverage of Flaviviridae genomes of (A) DENV, (B) YFV, (C) ZIKV (DAK), (D) ZIKV (PR2015), and (E) WNV for one replicate of MeRIP-seq (red), and input RNA-seq (blue) from matched samples. Colored bars indicate m6A peaks identified by MeRIPPeR analysis. (n=2; FDR-corrected q-value <0.05). (F) Alignment of replicate m6A sites in the genomes of DENV (red), YFV (blue), ZIKV (DAK) (orange), ZIKV (PR2015) (green), and WNV (brown). See also Figure S6 and Table S3. Highlights The RNA genomes of HCV, ZIKV, DENV, YFV, and WNV contain m6A modification The cellular m6A machinery regulates HCV infectious particle production YTHDF proteins reduce HCV particle production and localize at viral assembly sites m6A-abrogating mutations in HCV E1 increase infectious particle production This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. AUTHOR CONTRIBUTIONS N.S.G., A.B.R.M., M.J.M., A.E.R, E.M.K., C.E.M., and S.M.H. designed experiments and analyzed the data. N.S.G., A.B.R.M., M.J.M., E.M.K., A.E.R., C.V., J.W., J.A.G., S.E.H., K.M.Q., B.A.L, O.R.I., S.B.B., S.M.H. performed the experiments, and C.L.H., M.J.E., M.S.S., and M.G.B. provided reagents. N.S.G., A.B.R.M, C.E.M., and S.M.H. wrote the manuscript. All authors contributed to editing. Accession Numbers. The raw sequencing data obtained from the MeRIP-seq and PAR-CLIP have been submitted to the NCBI Gene Expression Omnibus (GEO) and are available through accession number GSE83438. Additional experimental procedures can be found in the Supplemental Experimental Procedures. SUPPLEMENTAL INFORMATION Supplemental Information includes 6 figures, 3 tables, experimental procedures, and references, and can be found with this article online. REFERENCES Alarcon CR Goodarzi H Lee H Liu X Tavazoie S Tavazoie SF HNRNPA2B1 Is a Mediator of m(6)A-Dependent Nuclear RNA Processing Events. Cell 2015 162 1299 1308 26321680 Aligeti M Roder A Horner SM Cooperation between the Hepatitis C Virus p7 and NS5B Proteins Enhances Virion Infectivity. J Virol 2015 89 11523 11533 26355084 Ariumi Y Kuroki M Kushima Y Osugi K Hijikata M Maki M Ikeda M Kato N Hepatitis C virus hijacks P-body and stress granule components around lipid droplets. J Virol 2011 85 6882 6892 21543503 Bidet K Garcia-Blanco MA Flaviviral RNAs: weapons and targets in the war between virus and host. Biochem J 2014 462 215 230 25102029 Bocharov G Ludewig B Bertoletti A Klenerman P Junt T Krebs P Luzyanina T Fraser C Anderson RM Underwhelming the immune response: effect of slow virus growth on CD8+-T-lymphocyte responses. J Virol 2004 78 2247 2254 14963121 Chatel-Chaix L Germain MA Motorina A Bonneil E Thibault P Baril M Lamarre D A host YB-1 ribonucleoprotein complex is hijacked by hepatitis C virus for the control of NS3-dependent particle production. J Virol 2013 87 11704 11720 23986595 Chen T Hao YJ Zhang Y Li MM Wang M Han W Wu Y Lv Y Hao J Wang L m(6)A RNA methylation is regulated by microRNAs and promotes reprogramming to pluripotency. Cell Stem Cell 2015 16 289 301 25683224 Dimock K Stoltzfus CM Sequence specificity of internal methylation in B77 avian sarcoma virus RNA subunits. Biochemistry 1977 16 471 478 189800 Dominissini D Moshitch-Moshkovitz S Salmon-Divon M Amariglio N Rechavi G Transcriptome-wide mapping of N(6)-methyladenosine by m(6)A-seq based on immunocapturing and massively parallel sequencing. Nat Protoc 2013 8 176 189 23288318 Dominissini D Moshitch-Moshkovitz S Schwartz S Salmon-Divon M Ungar L Osenberg S Cesarkas K Jacob-Hirsch J Amariglio N Kupiec M Topology of the human and mouse m6A RNA methylomes revealed by m6A-seq. Nature 2012 485 201 206 22575960 Fu Y Dominissini D Rechavi G He C Gene expression regulation mediated through reversible m(6)A RNA methylation. Nat Rev Genet 2014 15 293 306 24662220 Gastaminza P Kapadia SB Chisari FV Differential biophysical properties of infectious intracellular and secreted hepatitis C virus particles. J Virol 2006 80 11074 11081 16956946 Haddow AD Schuh AJ Yasuda CY Kasper MR Heang V Huy R Guzman H Tesh RB Weaver SC Genetic characterization of Zika virus strains: geographic expansion of the Asian lineage. PLoS Negl Trop Dis 2012 6 e1477 22389730 Hafner M Landthaler M Burger L Khorshid M Hausser J Berninger P Rothballer A Ascano M Jr. Jungkamp AC Munschauer M Transcriptome-wide identification of RNA-binding protein and microRNA target sites by PAR-CLIP. Cell 2010 141 129 141 20371350 Harper JE Miceli SM Roberts RJ Manley JL Sequence specificity of the human mRNA N6-adenosine methylase in vitro. Nucleic Acids Res 1990 18 5735 5741 2216767 Hopcraft SE Azarm KD Israelow B Leveque N Schwarz MC Hsu TH Chambers MT Sourisseau M Semler BL Evans MJ Viral Determinants of miR-122-Independent Hepatitis C Virus Replication. mSphere 2016 1 Hopcraft SE Evans MJ Selection of a hepatitis C virus with altered entry factor requirements reveals a genetic interaction between the E1 glycoprotein and claudins. Hepatology 2015 62 1059 1069 25820616 Hyde JL Gardner CL Kimura T White JP Liu G Trobaugh DW Huang C Tonelli M Paessler S Takeda K A viral RNA structural element alters host recognition of nonself RNA. Science 2014 343 783 787 24482115 Jia G Fu Y Zhao X Dai Q Zheng G Yang Y Yi C Lindahl T Pan T Yang YG N6-methyladenosine in nuclear RNA is a major substrate of the obesity-associated FTO. Nat Chem Biol 2011 7 885 887 22002720 Jopling CL Yi M Lancaster AM Lemon SM Sarnow P Modulation of hepatitis C virus RNA abundance by a liver-specific MicroRNA. Science 2005 309 1577 1581 16141076 Kane SE Beemon K Precise localization of m6A in Rous sarcoma virus RNA reveals clustering of methylation sites: implications for RNA processing. Mol Cell Biol 1985 5 2298 2306 3016525 Kane SE Beemon K Inhibition of methylation at two internal N6- methyladenosine sites caused by GAC to GAU mutations. J Biol Chem 1987 262 3422 3427 3029112 Kariko K Buckstein M Ni H Weissman D Suppression of RNA recognition by Toll-like receptors: the impact of nucleoside modification and the evolutionary origin of RNA. Immunity 2005 23 165 175 16111635 Kennedy EM Bogerd HP Kornepati AV Kang D Ghoshal D Marshall JB Poling BC Tsai K Gokhale NS Horner SM Posttranscriptional m(6)A Editing of HIV-1 mRNAs Enhances Viral Gene Expression. Cell Host Microbe 2016 19 675 685 27117054 Krug RM Morgan MA Shatkin AJ Influenza viral mRNA contains internal N6- methyladenosine and 5'-terminal 7-methylguanosine in cap structures. J Virol 1976 20 45 53 1086370 Lavi S Shatkin AJ Methylated simian virus 40-specific RNA from nuclei and cytoplasm of infected BSC-1 cells. Proc Natl Acad Sci U S A 1975 72 2012 2016 166375 Le Sommer C Barrows NJ Bradrick SS Pearson JL Garcia-Blanco MA G protein-coupled receptor kinase 2 promotes flaviviridae entry and replication. PLoS Negl Trop Dis 2012 6 e1820 23029581 Li S Mason CE The pivotal regulatory landscape of RNA modifications. Annu Rev Genomics Hum Genet 2014 15 127 150 24898039 Lichinchi G Gao S Saletore Y Gonzalez GM Bansal V Wang Y Mason CE Rana TM Dynamics of the human and viral m6A RNA methylomes during HIV-1 infection of T cells. Nature Microbiology 2016 1 16011 Lin S Choe J Du P Triboulet R Gregory RI The m(6)A Methyltransferase METTL3 Promotes Translation in Human Cancer Cells. Mol Cell 2016 62 335 345 27117702 Linder B Grozhik AV Olarerin-George AO Meydan C Mason CE Jaffrey SR Single-nucleotide-resolution mapping of m6A and m6Am throughout the transcriptome. Nat Methods 2015 12 767 772 26121403 Liu J Yue Y Han D Wang X Fu Y Zhang L Jia G Yu M Lu Z Deng X A METTL3-METTL14 complex mediates mammalian nuclear RNA N6-adenosine methylation. Nat Chem Biol 2014 10 93 95 24316715 Mauger DM Golden M Yamane D Williford S Lemon SM Martin DP Weeks KM Functionally conserved architecture of hepatitis C virus RNA genomes. Proc Natl Acad Sci U S A 2015 112 3692 3697 25775547 Meyer KD Jaffrey SR The dynamic epitranscriptome: N6-methyladenosine and gene expression control. Nat Rev Mol Cell Biol 2014 15 313 326 24713629 Meyer KD Patil DP Zhou J Zinoviev A Skabkin MA Elemento O Pestova TV Qian SB Jaffrey SR 5' UTR m(6)A Promotes Cap-Independent Translation. Cell 2015 163 999 1010 26593424 Meyer KD Saletore Y Zumbo P Elemento O Mason CE Jaffrey SR Comprehensive analysis of mRNA methylation reveals enrichment in 3' UTRs and near stop codons. Cell 2012 149 1635 1646 22608085 Miyanari Y Atsuzawa K Usuda N Watashi K Hishiki T Zayas M Bartenschlager R Wakita T Hijikata M Shimotohno K The lipid droplet is an important organelle for hepatitis C virus production. Nat Cell Biol 2007 9 1089 1097 17721513 Pager CT Schutz S Abraham TM Luo G Sarnow P Modulation of hepatitis C virus RNA abundance and virus release by dispersion of processing bodies and enrichment of stress granules. Virology 2013 435 472 484 23141719 Pirakitikulr N Kohlway A Lindenbach BD Pyle AM The Coding Region of the HCV Genome Contains a Network of Regulatory RNA Structures. Mol Cell 2016 62 111 120 26924328 Poenisch M Metz P Blankenburg H Ruggieri A Lee JY Rupp D Rebhan I Diederich K Kaderali L Domingues FS Identification of HNRNPK as regulator of hepatitis C virus particle production. PLoS Pathog 2015 11 e1004573 25569684 Quicke KM Bowen JR Johnson EL McDonald CE Ma H O'Neal JT Rajakumar A Wrammert J Rimawi BH Pulendran B Zika Virus Infects Human Placental Macrophages. Cell Host Microbe 2016 20 83 90 27247001 Rios-Marco P Romero-Lopez C Berzal-Herranz A The cis-acting replication element of the Hepatitis C virus genome recruits host factors that influence viral replication and translation. Sci Rep 2016 6 25729 27165399 Saletore Y Meyer K Korlach J Vilfan ID Jaffrey S Mason CE The birth of the Epitranscriptome: deciphering the function of RNA modifications. Genome Biol 2012 13 175 23113984 Schwartz S Mumbach MR Jovanovic M Wang T Maciag K Bushkin GG Mertins P Ter-Ovanesyan D Habib N Cacchiarelli D Perturbation of m6A writers reveals two distinct classes of mRNA methylation at internal and 5' sites. Cell Rep 2014 8 284 296 24981863 Sessions OM Barrows NJ Souza-Neto JA Robinson TJ Hershey CL Rodgers MA Ramirez JL Dimopoulos G Yang PL Pearson JL Discovery of insect and human dengue virus host factors. Nature 2009 458 1047 1050 19396146 Shimoike T Mimori S Tani H Matsuura Y Miyamura T Interaction of hepatitis C virus core protein with viral sense RNA and suppression of its translation. J Virol 1999 73 9718 9725 10559281 Sommer S Salditt-Georgieff M Bachenheimer S Darnell JE Furuichi Y Morgan M Shatkin AJ The methylation of adenovirus-specific nuclear and cytoplasmic RNA. Nucleic Acids Res 1976 3 749 765 1272797 Suthar MS Ma DY Thomas S Lund JM Zhang N Daffis S Rudensky AY Bevan MJ Clark EA Kaja MK IPS-1 is essential for the control of West Nile virus infection and immunity. PLoS Pathog 2010 6 e1000757 20140199 Tirumuru N Zhao BS Lu W Lu Z He C Wu L N(6)-methyladenosine of HIV-1 RNA regulates viral infection and HIV-1 Gag protein expression. Elife 2016 5 Villordo SM Filomatori CV Sanchez-Vargas I Blair CD Gamarnik AV Dengue virus RNA structure specialization facilitates host adaptation. PLoS Pathog 2015 11 e1004604 25635835 Wang C Pflugheber J Sumpter R Jr. Sodora DL Hui D Sen GC Gale M Jr. Alpha interferon induces distinct translational control programs to suppress hepatitis C virus RNA replication. J Virol 2003 77 3898 3912 12634350 Wang X Lu Z Gomez A Hon GC Yue Y Han D Fu Y Parisien M Dai Q Jia G N6-methyladenosine-dependent regulation of messenger RNA stability. Nature 2014 505 117 120 24284625 Wang X Zhao BS Roundtree IA Lu Z Han D Ma H Weng X Chen K Shi H He C N(6)-methyladenosine Modulates Messenger RNA Translation Efficiency. Cell 2015 161 1388 1399 26046440 Weaver SC Costa F Garcia-Blanco MA Ko AI Ribeiro GS Saade G Shi PY Vasilakis N Zika virus: History, emergence, biology, and prospects for control. Antiviral Res 2016 130 69 80 26996139 Wilkins C Woodward J Lau DT Barnes A Joyce M McFarlane N McKeating JA Tyrrell DL Gale M Jr. IFITM1 is a tight junction protein that inhibits hepatitis C virus entry. Hepatology 2013 57 461 469 22996292 Yamane D McGivern DR Wauthier E Yi M Madden VJ Welsch C Antes I Wen Y Chugh PE McGee CE Regulation of the hepatitis C virus RNA replicase by endogenous lipid peroxidation. Nat Med 2014 20 927 935 25064127 Yue Y Liu J He C RNA N6-methyladenosine methylation in post-transcriptional gene expression regulation. Genes Dev 2015 29 1343 1355 26159994 Zhang J Randall G Higginbottom A Monk P Rice CM McKeating JA CD81 is required for hepatitis C virus glycoprotein-mediated viral infection. J Virol 2004 78 1448 1455 14722300 Zheng G Dahl JA Niu Y Fedorcsak P Huang CM Li CJ Vagbo CB Shi Y Wang WL Song SH ALKBH5 is a mammalian RNA demethylase that impacts RNA metabolism and mouse fertility. Mol Cell 2013 49 18 29 23177736 Zhou J Wan J Gao X Zhang X Jaffrey SR Qian SB Dynamic m(6)A mRNA methylation directs translational control of heat shock response. Nature 2015 526 591 594 26458103
PMC005xxxxxx/PMC5123825.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 2985117R 4816 J Immunol J. Immunol. Journal of immunology (Baltimore, Md. : 1950) 0022-1767 1550-6606 27794000 5123825 10.4049/jimmunol.1600926 NIHMS821999 Article AIRWAY EPITHELIAL KIF3A REGULATES Th2- RESPONSES TO AEROALLERGENS Giridhar Premkumar Vummidi † Bell Sheila M. † Sridharan Anusha † Rajavelu Priya † Kitzmiller Joseph A. † Na Cheng-Lun † Kofron Matthew § Brandt Eric B. ‡ Ericksen Mark ‡ Naren Anjaparavanda P. ¶ Moon Changsuk ¶ Khurana Hershey Gurjit K. ‡ Whitsett Jeffrey A. †1 † Perinatal Institute, Cincinnati Children’s Hospital Medical Center, Division of Neonatology, Perinatal and Pulmonary Biology, 3333 Burnet Avenue, Cincinnati, OH 45229-3039 ‡ Perinatal Institute, Cincinnati Children’s Hospital Medical Center, Division of Asthma Research, 3333 Burnet Avenue, Cincinnati, OH 45229-3039 § Perinatal Institute, Cincinnati Children’s Hospital Medical Center, Division of Developmental Biology, 3333 Burnet Avenue, Cincinnati, OH 45229-3039 ¶ Perinatal Institute, Cincinnati Children’s Hospital Medical Center, Division of Pulmonary Medicine, 3333 Burnet Avenue, Cincinnati, OH 45229-3039 Correspondence to: Jeffrey A. Whitsett, M.D., Division of Pulmonary Biology, MLC7029, Cincinnati Children's Hospital Medical Center, 3333 Burnet Avenue, Cincinnati, OH 45229-3039, Phone: 513-803-2790, FAX: 513-636-7868, [email protected] 12 10 2016 28 10 2016 1 12 2016 01 12 2017 197 11 42284239 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. KIF3A, the gene encoding Kinesin family member 3A, is a susceptibility gene locus associated with asthma; however, mechanisms by which KIF3A might influence the pathogenesis of the disorder are unknown. Herein, we deleted the mouse Kif3a gene in airway epithelial cells. Both homozygous and heterozygous Kif3a gene deleted mice were highly susceptible to aeroallergens from Aspergillus fumigatus and house dust mite, resulting in asthma-like pathology characterized by increased goblet cell metaplasia, airway hyper-responsiveness, and Th2-mediated inflammation. Deletion of the Kif3a gene increased the severity of pulmonary eosinophilic inflammation and expression of cytokines (Il-4, Il-13, Il-17A) and chemokine (Ccl11) RNAs following pulmonary exposure to Aspergillus extract. Inhibition of Kif3a disrupted the structure of motile cilia, impaired mucociliary clearance, barrier function, and epithelial repair, demonstrating additional mechanisms by which deficiency of KIF3A in respiratory epithelial cells contributes to pulmonary pathology. Airway epithelial KIF3A suppresses Th2 pulmonary inflammation and airway hyper-responsiveness following aeroallergen exposure, implicating epithelial microtubular functions in the pathogenesis of Th2-mediated lung pathology. Susceptibility Gene Asthma Cilia Kinesins Airway Hyper-responsiveness INTRODUCTION Asthma is an increasingly common, complex pulmonary disorder causing reversible airflow obstruction, mucus hyperproduction, and inflammation. Both environmental and genetic factors are associated with risk of asthma. While allergic asthma is often associated with atopy and Th2 immune activation, expression of IL-13, IL-4, and IL-5, and eosinophilic pulmonary inflammation (1), there is substantial heterogeneity in asthma phenotypes among patient populations. For example, Th1 activation is associated with obesity and asthma (2). Th2-related asthma patients can be further subdivided on the basis of increased Th2 cytokine and periostin expression, neutrophilic inflammation and responses to inhaled steroids, indicating that the heterogeneity of asthma phenotypes is highly relevant to therapeutic responses (3–5). Environmental exposures to pollutants, respiratory viruses, allergens, bacterial and fungal products are known to play important roles in the pathogenesis of asthma (6, 7, for review). Likewise, a number of genetic factors influence susceptibility and severity of asthma. Extensive genome-wide analyses from diverse populations identified numerous haplotypes and chromosomal loci associated with asthma and atopy, both allergic rhinitis and eczema being closely associated with increased risk of childhood asthma. Genome Wide Association Studies identified increased susceptibility to asthma associated with variations in genes controlling inflammation, including genes regulating Th2 lymphocyte recruitment and activation; for example, IL-33, ST2, TSLP, ORMDL3, GSDML, ADAM33, SPINK5, and others, that were associated with asthma susceptibility (8, 9). Polymorphisms in FCεRI-β, a high affinity receptor for IgE on mast cells, was linked to childhood asthma and allergic dermatitis (10). The human KIF3A gene locus has been repeatedly implicated in susceptibility to asthma and eczema (11–16). KIF3A is a component of a trimeric motor complex regulating microtubular function and transport and is required for formation and function of both motile cilia and non-motile primary and sensory cilia (17, 18). KIF3A plays pleotropic roles in the regulation of microtubular transport, influencing intracellular protein trafficking, as well as ciliary transport and function (17–19). Genomic deletion of Kif3a in the mouse is embryonic lethal (20–22). In the lung, motile cilia occur as clusters on apical surfaces of ciliated cells that coordinate mucociliary clearance in the airways. Whereas, primary cilia are singular, non-motor organelles present on many cell types, including pulmonary cells, that are known to mediate signal transduction through diverse signaling pathways including Shh, Wnt, Pdgf and others, influencing morphogenesis, homeostasis, and repair of many organs (23–26). In various experimental models, roles for KIF3A in the regulation of cell proliferation, apoptosis, differentiation, intracellular transport, cytoskeletal dynamics, and planar polarity have been demonstrated (27–31). While KIF3A gene polymorphisms have been correlated with asthma, allergic rhinitis, and eczema, cellular mechanisms underlying this association are unknown. In the present study, we selectively deleted the mouse Kif3a gene in airway epithelial cells. Loss of Kif3a enhanced pulmonary inflammation, airway hyper-responsiveness (AHR), and Th2-mediated inflammation following aeroallergen challenge with Aspergillus fumigatus and house dust mite extracts. KIF3A was required for mucociliary clearance, epithelial cell migration and repair, providing plausible mechanisms by which KIF3A influences susceptibility to asthma. MATERIALS AND METHODS Mice Animal protocols were approved by the Institutional Animal Care and Use Committee in accordance with NIH guidelines. Kif3afl/fl mice, generated by Marszaleck et al and Lin et al. (20, 21), and Ift88fl/fl mice generated by Haycraft et al. (32) were kindly provided by Samantha A. Brugmann (Department of Plastic Surgery, Cincinnati Children’s Hospital Medical Center). Scgb1a1-Cre mice were kindly provided by Dr. Steven Shapiro (33). Gt(ROSA)26Sortm4(ACTB-tdTomato,-EGFP)Luo/J and Shhtm1(EGFP/CreCjt/J (34) mice were purchased from Jackson Laboratories. Naphthalene Injury Mice were administered naphthalene (Sigma, 30 mg/ml in corn oil) via a single i.p. injection to deliver a dose of 275 μg/g body weight. Control animals were injected with corn oil. Animals were sacrificed at 2, 7, and 10 days post administration. Aspergillus and House Dust Mite (HDM) Extract Sensitization Anesthetized 6–8 week old mice were administered a dose of 10 μg Aspergillus fumigatus or house dust mite extract (25μg) (Greer Laboratories, Lenoir, NC) diluted in 50 μl of saline by intratracheal (i.t.) instillation 3 times weekly for 3 weeks. Control animals were dosed with saline. Mice were sacrificed 48 hs following the last exposure. AHR/flexiVent Airway responsiveness of anesthetized mice was assessed with a flexiVent apparatus (SCIREQ, Montreal, QC, Canada) using an invasive method. Mice were anesthetized and the tracheas cannulated with an 18-gauge blunt needle. Mice were ventilated at 150 breaths/min, 3.0 cm water positive end-expiratory pressure. Two total lung capacity perturbations were performed for airway recruitment before baseline measurement and subsequent methacholine (MCh) challenges were performed. Acetyl-α-methacholine chloride (Sigma, St. Louis, MO) was administered for 10s (60 breaths/min, 500 μl tidal volume) in increasing concentrations (12.5, 25 and 50 mg/ml) via nebulization via the tracheotomy. Dynamic resistance (R) was determined by fitting the data to a single compartment model of airway mechanics where Ptr = RV + EV+ PO (Ptr, tracheal pressure; V, volume; and PO, constant). Average of three highest R-values with a coefficient of 0.9 or greater was used to plot the dose-response curves. BALF Bronchoalveolar lavage fluid (BALF) was collected from the right lobes by lavage (3 times with 0.7 ml of normal saline). BALF was centrifuged and cell pellets were resuspended in HBSS and total cell numbers counted using a hemocytometer. Cytospins were prepared and stained with a Diff-Quik Staining Kit (Polysciences Inc.) to determine differential cell counts. Histology and immunohistochemistry Mouse lungs were inflation fixed in 4% paraformaldehyde (PFA) in PBS overnight. Fixed tissue was processed according to standard protocols for paraffin embedding and used for chemical or immunohistochemical staining. Depending on the primary antibody some sections were subjected to antigen retrieval using citrate buffer pH 6.0 or 10 mM Tris-EDTA buffer pH 9.0 with heating. For in vitro studies, cells grown on ibiTreat chambered coverslips (Ibidi Inc) were fixed with 4% PFA/PBS for 15 minutes and permeabilized with 0.2% Triton-X-100 in PBS for 7 minutes at room temperature. Primary antibodies were applied overnight at 4ºC. Primary antibodies used were acetylated tubulin (Sigma T7451, 1:3000), ARL13B (Proteintech 17711-1-AP, 1:200), E-cadherin (Cell Signaling 3195, 1:100), FOXJ1 (eBioscience 1409965-82, 1:200), FOXA3 (Santa Cruz sc-5361, 1:50), MUC5AC (Abcam Ab3649, 1:100), SCGB1A1 (CCHMC, 1:800), ACTA2 (Sigma A5228, 1:2000), TUBB4A (Biogenex MU178-UC, 1:200), Alpha tubulin (Sigma T6199, 1:200), and Phospho-Histone H3 (Santa Cruz, sc-8656-R, 1:100). FOXJ1 and FOXA3 were detected with species-specific biotinylated secondary antibodies and then visualized with a streptavidin-conjugated fluorophore. For all other antibodies, fluorophore-conjugated secondary antibodies were used including Alexa Fluor-488, Alexa Fluor-555, Alexa Fluor-568, Alexa Fluor-594, and Alexa Fluor-647 (Jackson ImmunoResearch and Life Technologies). For fluorescence stains, sections were stained with DAPI and mounted with ProLong Gold anti-fade reagent (Life Technologies). Bright-field images were obtained using a Zeiss Axio ImagerA2 microscope equipped with AxioVision Software. Confocal immunofluorescence images were obtained using a Nikon A1Rsi inverted laser confocal microscope and widefield images were obtained using a Nikon Ti-E inverted microscope equipped with an Andor Zyla 4.2 SCMOS camera. Images were analyzed using NIS Elements (Nikon) or Imaris (Bitplane) software. Scanning electron microscopy Three mouse lungs from each genotype were inflation fixed with ice cold 2% paraformaldehyde, 2% glutaraldehyde in 0.1M sodium cacodylate buffer (SCB), pH 7.3, for 30 min, followed by postfixation with fresh fixative at 4°C overnight. Fixed mouse lungs were sliced into 1–2 mm slabs, incubated with 1% osmium and 1.5% potassium ferrocyanide in 0.1M SCB, pH 7.3, for 2 hrs, dehydrated in a graded series of alcohol, washed with hexamethyldisilazane, and air dried in a chemical fume hood for up to 2 days. Mouse lung tissue slabs were mounted on specimen stubs and coated with palladium/gold film using a Denton Vacuum Desk IV sputter coater. Scanning electron micrographs of murine airways were acquired using a Hitachi field emission scanning electron microscope SU8010 at 5 kV. Mouse trachea and airway sample preparation and video microscopy for ciliary beat and mucociliary clearance Three mice of each genotype (control, heterozygote, or homozygote) from the Kif3aShh or Kif3aScg lines were analyzed. Tracheas and lungs from the Kif3aScgΔ/Δ mice were prepared as described (35). Tracheas from Kif3aShhΔ/Δ mice and controls were removed, washed, and prepared according to Francis and Lo (36). For clearance assays, samples were placed on a glass dish in medium containing 0.20-μm Fluoresbrite microspheres (Polysciences, Inc., Warrington, PA, USA). Ciliary dynamics were captured with a 40X objective using Differential Interference Contrast (DIC) on a Nikon Ti-E inverted microscope (Nikon Microscopy) equipped with a Andor Zyla 4.2 SCMOS camera or with the 60X objective using DIC on an Olympus 1X51 inverted microscope equipped with a Hamamatsu EM-CCD digital camera. Images up to 500 frames/s (fps) were recorded. To quantitate ciliary beat frequency (CBF) and cilia-generated flow, at least two videos were collected from each tracheal sample (n=3/genotype). Using ImageJ software (ImageJ, NIH), a line was marked perpendicular to the cilia captured in each video and a kymograph was created. The number of pixels between each wave peak was measured (one pixel = one movie frame) from which the number of beats per minute (i.e. Hz) was calculated. To measure cilia generated flow, MTrackJ ImageJ software was used to manually track fluorescence beads across the surface of the tracheal epithelia. RNA isolation and analysis Total RNA was isolated from snap-frozen left lung lobes using a tissue-homogenizer (OMNI-TH International, Kennesaw, GA) by pulsing the probe for about 20 seconds in 1 ml of TRIzol Reagent (Life Technologies) and extracting the RNA using DirectZOL RNA Miniprep R2072 (Zymo Research, CA, USA). RNA was isolated from purified epithelial cells and cells in culture using the Qiagen RNeasy Micro Kit, 74004 (Valencia, CA). RNA cleanup and on-column DNase digestion (Qiagen RNeasy Micro kit, Valencia, CA) was performed on samples before being reverse transcribed either with the First Strand Superscript Synthesis kit (Invitrogen) or iScript cDNA synthesis kit (Bio-Rad). Quantitative real-time PCR analysis was performed on cDNA samples using TaqMan probes (Applied Biosystems) (Table I) and EUK 18S rRNA (4352930), FoxJ1 (Mm00807215_m1), Arl13b (Mm01349328_m1), Kif3a (Mm01288585_m1), KIF3A (Hs00199901_m1), and CDH1 (Hs01023894_m1) on a ABI 2720 Thermal cycler (Applied Biosystems, NY, USA). Lung cell isolation and EpCAM sorting Lung tissues were enzymatically digested at 37°C for 1h using Dispase (Corning, Discovery Labware, Inc., Bedford, MA). After 1 hour, DNasel was added prior to passing the solution through a 19-gauge needle to remove tissue aggregates. Cells were re-suspended in MACS buffer and cell counts determined. For every 1X107 cells of lung suspension, 10 μl of EpCAM-biotin antibody (CD326 EpCAM-biotin, 130-101-859, MACS Miltenyl Biotec) was added. Cells were incubated with primary antibody for 15 min, 4°C, washed and re-suspended in MACS buffer, and incubated with the secondary antibody (streptavidin conjugated microbeads 130-048-102) for selection of epithelial cells. EpCAM positive cells were collected by magnetic separation using an AutoMACS (Miltenyl Biotec). In vitro assays HBEC3 cells (a kind gift of Dr. John D. Minna, UT Southwestern Medical Center) and BEAS2B cells were grown to confluence on 12-well tissue culture plates (TPP plasticware). At 48 hs post-plating, the cells were infected with lentiviral constructs expressing shRNA against KIF3A or scramble controls (MOI of 2) in the presence of polybrene (2 μg/ml, Sigma Aldrich) to determine cell migration and motility. The cells were stained at 72 hours post transduction with di-8-ANEPPS (Biotium) diluted 1:500 in culture-media. A scratch was made in each well of the plate using a p200 pipet tip. The cells were washed with 1X PBS and fed with culture media containing di-8-ANEPPS. For rescue experiments, HBEC3 transduced with lentiviral cells or scramble control shRNA were transfected with a construct expressing the full-length KIF3A cDNA (pCMV3-C-GFPSpark, Sino Biological Inc., Bejing, P.R. China) using Fugene HD (Promega) transfection reagent at 72 hours post transduction. Rescue of cell migration defects was assessed at 48 hours post-transfection. A Nikon A1Rsi inverted laser scanning confocal microscope, equipped with a motorized XY stage, Tokai Hit microplate incubator and Perfect Focus System was used to document cell migration. Cells were imaged at pre-selected XY coordinate points in each well, once every 10 minutes over a period of 16-22 hours. Cells were harvested and mRNA was isolated using Qiagen RNeasy Micro Kit (Qiagen) for qRT-PCR analysis. Cell migration was analyzed using a spot tracking algorithm on Imaris software (Bitplane). Flow cytometric analysis of lung cells For flow cytometric analysis, total lung cell suspensions were prepared as previously described (37) from Aspergillus fumigatus extract treated mice injected with Brefeldin A 16 hs prior to sacrifice. Lung tissues were enzymatically digested using Caseinolytic units of Dispase (Corning, Discovery Labware, Inc., Bedford, MA) to obtain single cell suspensions and stained using fluorochrome-labeled antibodies. The gating strategies used are described in Rajavelu et al (38). Eosinophils in lung cell suspensions were detected using SiglecF PE (BD Pharmingen E50-2440) and CCR3 FITC (BioLegend J073E5) positive cells in the CD45 leucocyte gate. For analysis of ILC2 cells the lineage cocktail CD3/Ly-6G(Ly-6C)/CD11b/CD45R/Ter-119 Alexa Fluor 700 (BioLegend 79923), IL-7Rα FITC (BioLegend A7R34), FLT3 APC (BioLegend A2F10), ST2 (BioLegend DIH9), IL-17RB (R&D 752101), and ICOS Pacific Blue (BioLegend C398.4A) were used. T cells were analyzed using CD45 Alexa Fluor 700 (BioLegend 30-F11) and CD3εFITC (BioLegend 145-2C11). For intracellular cytokine detection, mice were i.p. injected with Brefeldin A (Sigma) and lung lobes were processed for single cell suspension as mentioned above followed by surface staining, then permeabilized with Cytofix-Cytoperm solution (BD Pharmingen) and stained for IL-4 PE/Cy7 (BioLegend 11B11) and IL-17A PerCP/Cy5.5 (BioLegend TC11-18H10.1). The stained lung cell samples were acquired and analyzed on a Becton Dickinson FACSCanto III flow cytometer using FACS Diva software. In vivo capillary-epithelial permeability Experiments were performed as described by Davidovich, et al (39). In brief, 24 hours after i.t. exposure to Aspergillus extract (100 μg) or saline, mice were anesthetized by isoflurane and 0.3 ml of a 12 mg/ml solution of FITC-conjugated albumin (A9771; Sigma-Aldrich, St. Louis, MO) was injected via tail vein. At the end of 3 hours, blood was collected via direct cardiac puncture and BAL performed with 3 ml of normal saline. Albumin fluorescence in BALF and serum was determined using Biotek Synergy 2, Biotek Instruments Inc., Vermont USA) with absorption/emission wavelengths of 480/520 nm. Epithelial permeability was defined as the ratio of BALF to serum fluorescence. Statistical analysis Values are expressed as the mean ± SEM. Statistical analysis was performed using GraphPad Prism 6 software (GraphPad, La Jolla, CA). Data were analyzed using a 2-tailed, unpaired Student’s t test or 2-way ANOVA, and Bonferroni’s correction was used for multiple comparisons. P values of 0.05 or less were considered statistically significant. RESULTS Kif3a mRNA is widely expressed in many organs and tissue types and is widely expressed during fetal and postnatal lung development (40). In the developing mouse lung, Kif3a is expressed in multiple cell types, including cells of epithelial and mesenchymal origin. Since KIF3A is required for formation and function of both motile and primary cilia, we assessed the immunofluorescence staining of ARL13B and TUBB4A (β-tubulin-IV) during perinatal and postnatal lung development in the mouse. Primary cilia were detected in both mesenchymal and epithelial cells in the fetal lung at E13 and E16.5 as identified by expression of ARL13B and lack of TUBB4A (Figs. 1A, 2A); whereas, by E18.5 and postnatally, motile cilia co-expressing both proteins were readily detectable (Fig. 1B, 1C). The presence of abundant motile cilia in conducting airways made impossible the detection of potential primary cilia in multi-ciliated cells after E16.5 (Fig. 1C) (41). Primary cilia were not detected in club cells in the normal postnatal lung, consistent with previous findings (41). We utilized mice bearing Scgb1a1-Cre or Shh-Cre to selectively delete exon 2 of the Kif3a gene. Scgb1a1-Cre is expressed early during the differentiation of conducting airway epithelial cells (approximately E16–17). As indicated by expression of eGFP in Rosa-Tomato (Red/Green) reporter mice, Scgb1a1-Cre caused extensive recombination in both ciliated and non-ciliated airway epithelial cells at 8 weeks of age (Fig. 1D); recombination was not observed in alveolar regions. Using this driver, the homozygous gene deleted mice Kif3afl/fl;Scgb1a1-Cre+ (Kif3aScgΔ/Δ) and heterozygous Kif3afl/+;Scgb1a1-Cre+ (Kif3aScgΔ/+) mice were generated. Controls were either Kif3afl/fl or Kif3a+/+;Scgb1a1-Cre+ (Kif3aScg+/+). Shh-Cre is expressed throughout the developing lung epithelium beginning as early as E9.5 (42) resulting in targeting of the Kif3afl alleles in most airway epithelial progenitor cells in proximal and peripheral conducting airways and in the alveoli; henceforth termed Kif3aShhΔ/Δ and Kif3aShhΔ/+ and the corresponding controls Kif3afl/fl or Kif3aShh+/+. Both Kif3aScgΔ/Δ and Kif3aShhΔ/Δ mice were viable and present in normal Mendelian ratios after birth. The number and distribution of club, goblet, and cells staining for FOXJ1, a transcription factor selectively expressed in ciliated cells, were similar in Kif3afl/fl, Kif3aScgΔ/+ and Kif3aScgΔ/Δ, and Kif3aShhΔ/Δ mice (Figs. 1E, 1F, 2C, 2D). QRT-PCR demonstrated that Foxj1 mRNA levels were unchanged, while Arl13b was significantly decreased in adult mice (Fig. 1G, 1H). A marked decrease in staining of acetylated α-tubulin (TUBA1A) was observed in Kif3aScgΔ/Δ and Kif3aShhΔ/Δ mice while expression of SCGB1A1 persisted in the club cells (Figs. 1E, 1F, 2C, 2D, and data not shown). As an indication of the loss of KIF3A, the presence of primary cilia was evaluated in Kif3aShhΔ/Δ mice. ARL13B staining revealed a dramatic reduction in the number of primary cilia in the developing lung epithelial cells, whereas, no changes were detected in the adjacent mesenchyme (Fig. 2A, 2B). Cells disassociated from adult whole lungs isolated from Kif3aShhΔ/Δ and Kif3aShh+/+ were sorted using the epithelial marker EpCAM. Exon 2 specific primers were used to demonstrate a reduction in Kif3a mRNA in Kif3aShhΔ/Δ cells (Fig. 2E, 2F). Consistent with the reduction of acetylated tubulin staining, numbers, size, and shape of motile cilia were reduced in Kif3aShhΔ/+ and Kif3aShhΔ/Δ mice by scanning EM (Fig. 2G–I). Abnormalities in cilia were most prominent in the homozygous Kif3aShhΔ/Δ mice. To assess mucociliary clearance after deletion of Kif3a, video imaging of tracheal and large airway fluid dynamics and ciliary activity were measured in Kif3aShhΔ/Δ, Kif3aScgΔ/Δ and control mice. Consistent with the extensive loss of motile cilia from the airway epithelium, ciliary beat frequencies were markedly decreased in airways of both Kif3aShhΔ/+ and Kif3aShhΔ/Δ mice and fluorescent bead movement lost directionality (Fig. 2J, 2K, Video1.mov and Video2.mov). Taken together, these observations indicate that primary cilia are not required for cell fate specification of the airway epithelial cells and that deletion of Kif3a by either Shh-Cre or Scgb1a1-Cre (data not shown) inhibits ciliary function and inhibits mucociliary clearance. Enhanced Th2 Inflammation and AHR in Kif3a Gene Targeted Mice After Aeroallergen Exposure To assess the role of Kif3a during aeroallergen sensitization, adult Kif3aScg+/+, Kif3afl/fl, Kif3aScgΔ/+, and Kif3aScgΔ/Δ mice were repeatedly treated with relatively low concentrations of Aspergillus fumigatus or HDM extract. At baseline, under our pathogen-free vivarium conditions, no histological evidence of pulmonary inflammation was observed in Kif3aScgΔ/Δ and Kif3aScgΔ/+ mice, nor were there differences in AHR or inflammatory responses following exposure to saline (Fig. 3). Following intratracheal Aspergillus fumigatus extract sensitization, similar AHR and inflammatory responses were observed in control Kif3afl/fl and Kif3aScg+/+ mice. In contrast, AHR and the inflammatory responses were significantly increased in Kif3aScgΔ/Δ and Kif3aScgΔ/+ mice (Fig. 3). AHR was most increased in homozygous Kif3a deleted mice (Fig. 3A). Numbers of inflammatory cells, consisting primarily of eosinophils, were increased in the BALF from both Kif3aScgΔ/+ and Kif3aScgΔ/Δ mice after aeroallergen challenge (Fig. 3B, C). We observed no differences in plasma levels of Aspergillus specific IgG1 in the exposed mice of each genotype (Suppl. Fig. 1A). Goblet cell metaplasia and mucus hyper-production were observed in mice of all Kif3a genotypes after exposure to Aspergillus fumigatus extract and were more severe in the Kif3aScgΔ/Δ and Kif3aScgΔ/+ mice, (Fig. 3D–F, Suppl. Fig. 1B). Consistent with these findings, AHR was significantly increased in Kif3a gene deleted mice after house dust mite exposure (Suppl. Fig. 2). The increased inflammatory responses to Aspergillus fumigatus extract observed in the Kif3a deleted mice were supported by mRNA expression data, Table I. Histological and immunofluorescence studies demonstrating increased goblet cell metaplasia after exposure to Aspergillus extract were supported by increased expression of Spdef, Foxa3, Clca1, and Muc5b mRNAs in lungs from Kif3aScgΔ/Δ mice. Likewise, Il-13, Il-4, Il-17A, Ccl11, and Ear11 mRNAs were significantly increased in lungs of both Kif3aScgΔ/Δ and Kif3aScgΔ/+ mice, indicating Th2-mediated eosinophilic inflammation (Table I and Fig. 3D–F). We utilized flow analysis to identify the inflammatory cells present following Aspergillus extract administration to Kif3aScgΔ/Δ mice. Numbers of SiglecF+/CCR3+ cells in whole lung digests were increased in lungs of Kif3aScgΔ/Δ mice, consistent with the increased eosinophils seen in the BALF; likewise, IL-4+ and IL-17A+ T cells were increased, consistent with enhanced Th2 and Th17 responses in Kif3a gene deleted mice (Fig. 4). Like Kif3a, Ift88 (intraflagellar transporter protein 88) is a microtubule associated transport protein required for both primary and motile cilia formation in airway epithelial cells (43, 44). The Scgb1a1-Cre and Shh-Cre transgenes were bred into the Ift88fl/fl animals to create Ift88 gene deleted lung epithelium. At baseline, histological findings in the Ift88ScgΔ/Δ and Ift88ShhΔ/Δ mice were similar to those seen after epithelial deletion of Kif3a in which specification of lung epithelial cells and absence of motile cilia were observed (Fig. 5A–D, data not shown). Ift88fl/fl and Ift88ScgΔ/Δ were exposed to Aspergillus fumigatus or saline using the same sensitization protocol used with the Kif3a gene targeted mice. Pulmonary eosinophilic inflammation and Alcian blue staining of airway cells were similar in Ift88ScgΔ/Δ and controls after exposure to Aspergillus extract (Fig. 5E–H). While an increase in mRNAs associated with goblet cells was observed in Ift88ScgΔ/Δ mice after exposure, neither Ear11 or Th2 related cytokine mRNAs were increased to the levels seen in the Kif3a deleted mice (Table I). AHR was not evaluated in these animals since Ift88ShhΔ/Δ mice were found to have hyper-reactive airways at baseline (43). Taken together, both Kif3aScgΔ/+ and Kif3aScgΔ/Δ mice were more susceptible to Th2-mediated eosinophilic pulmonary inflammation than the Ift88ScgΔ/Δ mice indicating a distinct role for Kif3a in asthma-like pulmonary inflammation. KIF3A is Required for Epithelial Repair Since primary cilia play important roles in the regulation of proliferation, migration, and differentiation, we assessed airway epithelial repair by treating adult mice with naphthalene. Naphthalene is metabolized to a cellular toxicant by CYP2F2, a p450 enzyme that is selectively expressed in most non-ciliated conducting airway cells in the mouse lung (45, 46). CYP2F2 was normally expressed by club cells in Kif3a deficient airways. After naphthalene exposure, resistant progenitor cells rapidly migrate, proliferate, and differentiate to repair the airways. Seven days after naphthalene exposure, areas of damaged airway were present in both the control and Kif3a gene deleted airways (Fig. 6A). In control mice, the distribution of both ciliated and epithelial club cells was restored 10 days after naphthalene injection (Fig. 6B–D). Repair of the airway epithelium in Kif3aScgΔ/Δ mice was incomplete, as indicated by significantly decreased numbers of club cells and squamous cell metaplasia, demonstrating that KIF3A was required for normal repair of the conducting airway epithelium. KIF3A Mediates Epithelial Cell Migration To assess the role of KIF3A in cell proliferation and migration, KIF3A mRNA was inhibited in Human Bronchial Epithelial Cells (HBEC3) and BEAS-2B cells with lenti-viral shRNA and cell migration was assessed by video imaging. When KIF3A mRNA was selectively inhibited, cells adhered to the plates, proliferated and reached confluency at 72 hrs post-transduction. Primary cilia were identified by co-staining of ARL13B and acetylated α-tubulin (ac-TUBA4A) in BEAS-2B cells (Fig. 7C). Motile cilia were not detected in these cell lines. Inhibition of KIF3A was confirmed by qRT-PCR, whereas, expression of a non-targeted gene, CDH1 was unchanged (Fig. 7D). KIF3A knockdown suppressed formation of primary cilia and markedly inhibited cell movement and migration in “scratch” assays (Fig. 7A, 7B, see Video3.mov and Video4.mov). KIF3A shRNA did not alter proliferation of HBEC3 cells (Fig. 7E); however, phospho-Histone H3 (pHH3) staining was modestly decreased by KIF3A knockdown in BEAS2B cells at confluency (data not shown). Cytoplasmic staining of ac-TUBA4A was altered by inhibition of Kif3a, indicating a potential role of KIF3A in microtubule assembly (Suppl. Fig. 3A). Cell tracking analysis showed that inhibition of KIF3A did not cause cell death but inhibited cell movement and migration trajectory characteristics, e.g. displacement, directionality, speed, and persistence and expression of the human KIF3A cDNA substantially rescued cell motility and navigation (Suppl. Fig. 3B). Loss of KIF3A was associated with dramatic inhibition of cell movement and disruption of cytoplasmic microtubule organization indicating its important role in bronchial epithelial cell migration, findings consistent with the failure of airway epithelial repair seen after naphthalene induced injury in the Kif3aScgΔ/Δ mice in vivo. Increased Capillary Epithelial Permeability in Kif3aShhΔ/Δ mice Since loss of barrier function is associated with asthma-like pathology and Th2 inflammation (47), we tested whether capillary-epithelial permeability was altered in the Kif3aShhΔ/Δ mice in vivo. Twenty-four hours after exposure to one dose of Asperillus extract (i.t.), mice were injected with fluorescein-congugated albumin via tail vein followed by quantification of fluorescein in BALF and serum. Capillary-alveolar permeability was significantly increased in the Kif3a gene deleted mice, Fig. 6E. DISCUSSION Present findings demonstrate the important role of airway epithelial KIF3A in the pathogenesis of aeroallergen-induced inflammation and airway hyper-responsiveness, linking the activities of microtubules to innate immune responses to aeroallergens. Enhanced AHR, goblet cell associated gene expression and Th2-mediated eosinophilic inflammation observed after deletion of Kif3a, provide a plausible mechanistic link between the genetic association of KIF3A alleles, levels of KIF3A protein, microtubule function and asthma (15, 16). While complete deletion of Kif3a is lethal in embryonic development (20–22), present studies demonstrate that loss of Kif3a in airway epithelial cells impairs mucociliary clearance, epithelial repair following injury, and enhances Th2-inflammation that together may influence responses to aeroallergens. Aspergillus fumigatus infection causes Allergic Broncho-Pulmonary Aspergillosis (ABPA) associated with individuals with asthma, cystic fibrosis, and primary ciliary dyskinesia; patients with ABPA are also at risk for eczema, hives, hay fever, and sinusitis (48, 49). Increased AHR was seen in Kif3aScgΔ/+ and Kif3aScgΔ/Δ mice following repeated exposures to Aspergillus fumigatus or house dust mite extract. A major Aspergillus fumigataus allergen, ASPF13 promotes airway hyper-responsiveness, recruiting inflammatory cells to the bronchial submucosa and disrupting airway smooth muscle cell-extracellular matrix interactions (50). AHR was assessed by plethysmography, a measure of both smooth muscle hyperplasia and contractility, and mucus hyperproduction. In the present studies, we did not detect differences in alpha-smooth muscle actin (ACTA2) staining in bronchial smooth muscle and Acta2 mRNA was not significantly increased in the Kif3aScgΔ/Δ mice after Aspergillus sensitization. Thus, mechanisms involved in the observed AHR are presently unclear. However, expression of Il-13 and Il-4 RNAs were markedly increased in both heterozygous and homozygous Kif3a deleted mice after sensitization, both factors being known to directly activate IL-4Rα receptor signaling in bronchial smooth muscle cells causing AHR (51). Increased secretion of chemokines and cytokines recruiting inflammatory cells including eosinophils has been associated with increased smooth muscle cell activity in asthma (52). Accumulation of mucus in conducting airways seen in both Kif3aScgΔ/Δ and Kif3ScgΔ/+ mice may also contribute to AHR. Present studies demonstrate diverse functions of KIF3A in airway epithelial cells, including epithelial repair, innate immune responses, and mucociliary clearance, that may influence airway reactivity and Th2-inflammation. There is strong experimental evidence linking the loss of barrier function, epithelial injury, and mucociliary clearance in the pathogenesis of asthma (47, 53, 54). Patients with primary ciliary dyskinesia have recurrent airway infections related to poor mucociliary clearance, (55, 56) and expression of cilia-related genes, including KIF3A, were decreased during acute asthma (57). Reduced mucociliary clearance and increased susceptibility to infection related to motile ciliary dysfunction were proposed to contribute to the pathogenesis of asthma (58). KIF3A is known to play important roles in microtubule assembly and intracellular transport of multiple protein cargos, in addition to its known role in the formation of primary and motile cilia (17–19). It is therefore likely that changes in KIF3A levels or function play diverse roles in respiratory epithelial cell homeostasis. Aspergilla and HDM extracts contain proteases, antigens, and other inflammatory mediators that cause epithelial cell injury (59, 60). Decreased mucociliary clearance, increased uptake of antigen by dendritic cells, and disrupted microtubular transport within the epithelial cells may influence the increased Th2 responses seen following deletion of Kif3a. Changes in barrier function, modulation of cytokine and chemokine signaling from the epithelium, regulate recruitment and activation of dendritic and ILC2 cells in turn recruiting Th2-helper and Th17 cells mediating asthma-like pulmonary inflammation (37, for review, 61, 62). Increased Il-13, Il-4, Ccl11, Il-17A, and Ccl24 RNAs seen in lungs of the Kif3a deleted mice are consistent with the observed Th2-mediated inflammation. Ccl11 (eotaxin1), a potent eosinophil chemoattractant, was markedly induced in the Kif3a but not the Ifit88 deleted mice. Increased expression of Il-13 and Il-4 seen after Aspergilla extract in the Kif3aScgΔ/Δ mice is typical of canonical Th2 lymphocytic responses that influence goblet cell metaplasia and mucus hyperproduction (63). Likewise, increased Spdef, Foxa3, Muc5b, and Muc5ac in the Kif3aScgΔ/Δ mice is consistent with goblet cell metaplasia being related to the activation of Th2-induced IL-4R signaling and STAT6 activation that occurs following aeroallergen exposure (38, 64, 65). Repair of the respiratory epithelium was impaired and capillary-epithelial barrier function was decreased, factors that may contribute to the enhanced Th2-inflammation seen in the Kif3a gene deleted mice. Primary cilia influence cell migration, a process critical for repair (66, 67). Decreased expression of KIF3A inhibited migration of both BEAS-2B and HBEC3 cells in vitro, findings consistent with impaired epithelial repair seen in Kif3aScgΔ/Δ mice seen after exposure to naphthalene. These findings are supported by previous in vitro studies demonstrating the role of KIF3A in cell migration in kidney epithelial cells (27). Recent in vitro findings demonstrated that the disruption of microtubules seen after inhibition of KIF3A did not occur after inhibition of IFT88 (68). Inflammatory responses to Aspergillus were increased in both Ift88 and Kif3a deleted mice, although Th2-responses were more pronounced in the Kif3aScgΔ/Δ mice, supporting the concept that the microtubule associated proteins may have distinct as well as overlapping functions in innate immune regulation in the airway epithelial cells. Present findings demonstrate that KIF3A was required for suppression of Th2-mediated inflammatory responses, mucus hyperproduction, and AHR following aeroallergen exposure, findings that support the association of KIF3A gene polymorphisms with clinical susceptibility to allergic asthma and rhinitis (11–16). Previous clinical findings in nasal epithelial cells demonstrated decreased expression of KIF3A during acute asthma exacerbations (57). Since in the present study, increased lung inflammation and AHR were observed in haplo-insufficient mice, even a modest decrease in KIF3A expression may influence the susceptibility to Th2-mediated inflammation. Previous observations, that the human KIF3A gene locus is located contiguously with the IL-4/IL-13 genes (69, 70), genes known to regulate Th2 inflammation in asthma, complicate the interpretation of the importance of KIF3A alleles in the pathogenesis of atopy and asthma. While present studies do not exclude the possibility that asthma susceptibility related to this chromosomal region is associated with the other genes in that locus, present findings in our mouse models demonstrate a direct role for KIF3A in the regulation of aeroallergen induced Th2 inflammation, airway epithelial repair, and mucociliary clearance, processes that may influence susceptibility to asthma. Supplementary Material 1 2 The authors thank Shawn Grant, Jaymi Semona, Kalpana Srivastava, Brandy Ruff, Mehari E. Mengistu, Gail Macke, and Courtney Stockman for their scientific assistance, Ann Maher for typing and editing on this manuscript, and the CCHMC Viral Vector Core for making the lentiviruses. Figure 1 Conditional deletion of Kif3a in the airway epithelium At E16.5, primary cilia are detected by ARL13B staining in the fetal lung epithelium and mesenchyme (red arrows) and absence of TUBB4A (A). At E18.5 (B) and P07 (C), the presence of motile cilia (staining for ARL13B and TUBB4A) obscures detection of primary cilium. ARL13B staining was not detected in SCGB1A1 stained club cells (scale bars equal 50 μm). eGFP in Rosa-Tomato Red/Green reporter mice (D) were used to assess the extent of recombination by Scgb1a1-Cre at 8 weeks of age. Cilia were stained for TUBA1A (green) in Kif3afl/fl control mice (E). Cilia were lacking after deletion of KIF3A in Kif3aScgΔ/Δ mice (F). FOXJ1 staining was unchanged (red, E, F). Number and distribution of club cells staining with SCGB1A1 (white) were not altered in Kif3aScgΔ/Δ airways. qRT-PCR on whole lung cDNA from adult mice (G–H), n=6 per genotype, demonstrated decreased Arl13b mRNA (P<0.01 by t-test compared to controls), consistent with loss of Kif3a. Figure 2 Loss of primary cilia, ciliary beating, and mucociliary clearance in Kif3aShhΔ/Δ mice Kif3a was deleted under control of Shh-Cre. At E13, primary cilia, identified by ARL13B staining, were absent in Kif3aShhΔ/Δ airway epithelial cells (B, arrows), compared to Kif3aShh+/+ embryos (A, arrowheads). Figures are representative of n=3/genotype. Staining of ciliated (FOXJ1 positive nuclei, pink arrows) and club (SCGB1A1, white) cells is shown in the bronchiolar epithelial cells in adult airways (C, D), n=4/genotype. Kif3a mRNA levels were evaluated using exon-2 specific primers. A decrease in Kif3a mRNA was detected in EpCAM (+) but not EpCAM(-) sorted cells from adult Kif3aShhΔ/Δ mice, n=4/genotype, P <0.05 by t-test) (E–F). Motile cilia were identified in control, Kif3aShh+ and Kif3aShhΔ/Δ airways by scanning electron microscopy (G–I). Uniformly short cilia were observed in Kif3aShhΔ/Δ airways (n=3/genotype). Tracheas of adult mice were excised, perfused, and differential interference contrast images viewed longitudinally after introduction of fluorescent microspheres (0.20 μm). Movements of individual beads were followed by video-microscopy. Cilia beat frequency (CBF) was quantified (J). Cilia-generated flow was imaged by tracking the fluorescent microspheres (K). Data represent the mean ± SEM of n=3 mice per group, P < 0.001 by one-way ANOVA. Figure 3 Increased AHR and eosinophilic inflammation in Kif3a gene deleted mice Age-matched mice were sensitized with Aspergillus fumigatus antigen or saline. AHR is represented as resistance in response to methacholine. Aspergillus treated mice are represented as solid lines and saline treated mice as dashed lines (A). Total number of cells in BALF was determined (B). Eosinophil numbers were calculated (C). Data represent the mean ± SEM of 6–7 mice per group, P < 0.05, P < 0.01 and P < 0.001 by 2-way ANOVA. Increased inflammation and mucus were present in both Kif3a gene deletants shown by representative 40X tile scans after Alcian blue staining (D–F). Figure 4 Deletion of Kif3a in airway epithelial cells increased recruitment of IL4+ and IL17+ T cells Flow cytometric analysis of lung cells from Kif3afl/fl (■) and Kif3aScgΔ/Δ (●) mice after i.t. administration of Aspergillus fumigatus extract (10 μg, 9 times during a two-week period). Numbers of CD45+ leucocytes (A) and CD3+ T cells (B). Eosinophilic inflammation was assessed by SiglecF+/CCR3+ cells (C). Numbers of IL4+ and IL17+ CD3+ T cells (D and E) were significantly increased in Kif3aScgΔ/Δ mice. ILC2 (Lin−/IL7Rα+/ICOS+/ST2+/IL17RB+) cell numbers in Kif3afl/fl and Kif3aScgΔ/Δ were unaltered (F). Graphs represent mean ± S.E.M., p<0.05, p<0.005, ns (not significant), compared to controls using one way ANOVA, n= 3 animals per genotype. Figure 5 Response to Aspergillus sensitization in Ift88ScgΔ/Δ mice ac-TUBA1A (green) stained cilia in Ift88fl/fl controls (A) were nearly absent in Ift88ScgΔ/Δ mice (B). FOXJ1, a transcription factor expressed in ciliated cells, was unchanged (B). Number and distribution of club cells staining for SCGB1A1 (white) were not altered in Ift88ScgΔ/Δ airways (A–D). IFT88 staining (green) was nearly absent in Ift88ScgΔ/Δ mice (C vs D). Age-matched mice (6–8 weeks at initial sensitization) were sensitized 3 times per week for 3 weeks with 10 μg Aspergillus fumigatus antigen or saline by i.t., (n=6–7 mice per genotype). Alcian blue staining of whole left lobes was unchanged (E–F). The number of cells in BALF was not altered (G). Eosinophil numbers were calculated from total numbers of BALF cells and differential cell counts and were not statistically different (H) as determined by one-way ANOVA. Figure 6 KIF3A is required for repair of the airway epithelium Mice of each genotype were injected i.p. with one dose of naphthalene. Mice were sacrificed on day 7 (n=3) (A), and day 10 (n=4) (B, C). Lung sections were immunostained for SCGB1A1 (red) and ac-TUBA1A (green in A, B). In panel C, CDH1 (E-cadherin) is shown in (green) and TUBA1A in (white). Restoration of bronchial epithelial cells is shown in controls and persistence of squamous metaplasia and paucity of SCGB1A1 stained cells is shown in the Kif3aScgΔ/Δ mice. For quantitation of regenerated cells, 200X magnified images of comparable regions (5 proximal and 5 distal region) from control and mutant mice were evaluated using Image J Software (FIJI, plugin). Percentage of regenerated cells was calculated by counting SCGB1A1 stained cells and the total number DAPI stained epithelial nuclei. The number of club cells were significantly decreased in Kif3aScgΔ/Δ mice on day 10 (D). Decreased capillary-epithelial permeability in epithelial barrier function in Kif3aShhΔ/Δ mice (E). FITC labeled albumin was injected into the tail vein of 4–5 adult mice of each genotype 24 hours after i.t. treatment with one dose of Aspergillus extract (100 μg). The concentration of the FITC label in the serum and BALF was measured by 3 hours after injection and the ratio calculated, *P<0.001 by t-test. Figure 7 KIF3A is required for cell migration in vitro HBEC3 and BEAS2B cells were transduced with control and KIF3A shRNA lenti-viruses. Cells were stained with di-8-ANEPPS and the monolayer was wounded by mechanical scratch. Kif3a mRNA was decreased by the shRNAs, (D, n=4). Inhibition of Kif3a inhibited cell migration as observed by videography for 16 hours, data representative of n=3 independent experiments in BEAS2B (A) and HBEC3 (B). Co-staining for ARL13B and ac-TUBA4A showed loss of primary cilia in BEAS-2B cells (C). ARL13B staining was not detected in HBEC3 cells (not shown). Immunofluorescence data are representative of n=2 independent experiments. Average number of pHH3 nuclei detected per 1000 DAPI stained nuclei (E). Proliferation was unaltered by knockdown of Kif3a in HBEC3 cells. Data represent mean ± SEM of nuclei from 6-10 images per group. Table I Increased expression of RNAs related to Th2 inflammation and goblet cell differentiation qRT-PCR was performed on whole tissue RNA collected from the right lungs of mice sensitized with Aspergillus extract. Fold change in mRNA levels of Kif3aScgΔ/+ and Kif3aScgΔ/Δ compared to Kif3afl/fl and Ift88 gene deletants compared to Ift88fl/fl control mice is shown after normalization to 18S. Age-matched mice (n=6–7) per group are shown. Il17A RNA was barely detectable at baseline, and was readily detected in both Kif3a deletants. Gene Taqman Probe I.D. Fold Increase Relative to Controls (Mean ± SEM) Kif3aScgΔ/+ Kif3aScgΔ/Δ Ift88ScgΔ/Δ Acta2 Mm00725412_s1 1.7±0.6 3.3±1.2 5.5±3.4 Ccl11 Mm00441238_m1 36.4±21.1B 171.1±144.1 B 0.3±0.1 Ccl17 Mm00516136_m1 7.6±2.8 21.3±17.0 3.2±1.1 Ccl2 Mm00441242_m1 5.5±3.2 2.3±1.0 0.7±0.3 Ccl24 Mm00444701_m1 15.5±9.1 6.0±3.0 2.5±1.1 Cdh1 Mm00480906_m1 1.1±0.3 1.5±0.3 1.3±0.3 Clca1 Mm00777368_m1 2.4±0.6 14.4±11.2 A 2.5±0.9 Csf2 Mm01290062_m1 1.2±0.5 2.5±1.2 4.9±2.2 Cxcl1 Mm04207460_m1 1.6±0.7 4.1±1.9 5.0±2.4 Ear11 Mm00519056_s1 18.4±8.0 A 26.7±6.5 A 4.7±2.7 Foxa3 Mm00484714_m1 2.5±0.5 18.7±14.3 A 11.4±4A Ifnγ Mm01168134_m1 3.8±0.6 3.3±1.09 5.3±2.5 Il4 Mm00445259_m1 6.7±1.4 A 29.6±21.7 A 1.8±0.4 Il13 Mm00434204_m1 8.9±2.6A 25.4±11.4 A 1.9±0.5 Il17A Mm00439618_m1 78.7±52.6C 10.3±3.4D 3.5±1.2 Il33 Mm00434204_m1 2.3±0.6 3.3±0.9 2.1±1.1 Il5 Mm00439646_m1 3.4±0.6 9.5±6.6 2.2±0.7 Il6 Mm00446190_m1 4.4±2.7 2.7±0.7 3.8±1.2 Muc5ac Mm01276725_g1 4.1±2.1 8.5±5.6 3.1±1.8 Muc5b Mm00466391_m1 5.3±2.8 6.1±2.1 A 6.8±2.9 A Spdef Mm00600221_m1 4.7±1.1 A 10.8±5.9 A 7.2±3.7 A Tslp Mm01157588_m1 2.1±0.8 2.2±0.9 4.3±2.2 A P<0.05 and B P<0.01by one-way ANOVA followed by Bonferroni’s multiple comparison test. C P=0.024 by ANOVA on Ranks. D P=0.0388 by 2-tailed t-Test. 1 This work was supported by Grants HL095580 and HL110964 (JAW) and 2U19AI070235-11 (GKKH) 1 Lambrecht BN Hammad H 2014 Allergens and the airway epithelium response: gateway to allergic sensitization J Allergy Clin Immunol 134 499 507 25171864 2 Rastogi D Fraser S Oh J Huber AM Schulman Y Bhagtani RH Khan ZS Tesfa L Hall CB Macian F 2015 Inflammation, metabolic dysregulation, and pulmonary function among obese urban adolescents with asthma Am J Respir Crit Care Med 191 149 160 25457349 3 Corren J Lemanske RF Hanania NA Korenblat PE Parsey MV Arron JR Harris JM Scheerens H Wu LC Su Z Mosesova S Eisner MD Bohen SP Matthews JG 2011 Lebrikizumab treatment in adults with asthma N Engl J Med 365 1088 1098 21812663 4 Jia G Erickson RW Choy DF Mosesova S Wu LC Solberg OD Shikotra A Carter R Audusseau S Hamid Q Bradding P Fahy JV Woodruff PG Harris JM Arron JR G Bronchoscopic Exploratory Research Study of Biomarkers in Corticosteroid-refractory Asthma Study 2012 Periostin is a systemic biomarker of eosinophilic airway inflammation in asthmatic patients J Allergy Clin Immunol 130 647 654 e610 22857879 5 Woodruff PG Modrek B Choy DF Jia G Abbas AR Ellwanger A Koth LL Arron JR Fahy JV 2009 T-helper type 2-driven inflammation defines major subphenotypes of asthma Am J Respir Crit Care Med 180 388 395 19483109 6 Edwards MR Bartlett NW Hussell T Openshaw P Johnston SL 2012 The microbiology of asthma Nat Rev Microbiol 10 459 471 22669219 7 Parnia S Brown JL Frew AJ 2002 The role of pollutants in allergic sensitization and the development of asthma Allergy 57 1111 1117 12464038 8 Kontakioti E Domvri K Papakosta D Daniilidis M 2014 HLA and asthma phenotypes/endotypes: a review Hum Immunol 75 930 939 24994462 9 Martinez FD Vercelli D 2013 Asthma Lancet 382 1360 1372 24041942 10 Hill MR Cookson WO 1996 A new variant of the beta subunit of the high-affinity receptor for immunoglobulin E (Fc epsilon RI-beta E237G): associations with measures of atopy and bronchial hyper-responsiveness Hum Mol Genet 5 959 962 8817330 11 Hirota T Takahashi A Kubo M Tsunoda T Tomita K Sakashita M Yamada T Fujieda S Tanaka S Doi S Miyatake A Enomoto T Nishiyama C Nakano N Maeda K Okumura K Ogawa H Ikeda S Noguchi E Sakamoto T Hizawa N Ebe K Saeki H Sasaki T Ebihara T Amagai M Takeuchi S Furue M Nakamura Y Tamari M 2012 Genome-wide association study identifies eight new susceptibility loci for atopic dermatitis in the Japanese population Nat Genet 44 1222 1226 23042114 12 Kang Z Li Q Fu P Yan S Guan M Xu J Xu F 2015 Correlation of KIF3A and OVOL1, but not ACTL9, with atopic dermatitis in Chinese pediatric patients Gene 571 249 251 26127003 13 Lepre T Cascella R Ragazzo M Galli E Novelli G Giardina E 2013 Association of KIF3A, but not OVOL1 and ACTL9, with atopic eczema in Italian patients Br J Dermatol 168 1106 1108 23278845 14 Paternoster L Standl M Chen CM Ramasamy A Bonnelykke K Duijts L Ferreira MA Alves AC Thyssen JP Albrecht E Baurecht H Feenstra B Sleiman PM Hysi P Warrington NM Curjuric I Myhre R Curtin JA Groen-Blokhuis MM Kerkhof M Saaf A Franke A Ellinghaus D Folster-Holst R Dermitzakis E Montgomery SB Prokisch H Heim K Hartikainen AL Pouta A Pekkanen J Blakemore AI Buxton JL Kaakinen M Duffy DL Madden PA Heath AC Montgomery GW Thompson PJ Matheson MC Le Souef P St Pourcain B Smith GD Henderson J Kemp JP Timpson NJ Deloukas P Ring SM Wichmann HE Muller-Nurasyid M Novak N Klopp N Rodriguez E McArdle W Linneberg A Menne T Nohr EA Hofman A Uitterlinden AG van Duijn CM Rivadeneira F de Jongste JC van der Valk RJ Wjst M Jogi R Geller F Boyd HA Murray JC Kim C Mentch F March M Mangino M Spector TD Bataille V Pennell CE Holt PG Sly P Tiesler CM Thiering E Illig T Imboden M Nystad W Simpson A Hottenga JJ Postma D Koppelman GH Smit HA Soderhall C Chawes B Kreiner-Moller E Bisgaard H Melen E Boomsma DI Custovic A Jacobsson B Probst-Hensch NM Palmer LJ Glass D Hakonarson H Melbye M Jarvis DL Jaddoe VW Gieger C Strachan DP Martin NG Jarvelin MR Heinrich J Evans DM Weidinger S C Australian Asthma Genetics, C. Lifecourse Epidemiology, C. Genetics of Overweight Young Adults, EA Genetics 2012 Meta-analysis of genome-wide association studies identifies three new risk loci for atopic dermatitis Nat Genet 44 187 192 15 Kim JH Cha JY Cheong HS Park JS Jang AS Uh ST Kim MK Choi IS Cho SH Park BL Bae JS Park CS Shin HD 2011 KIF3A, a cilia structural gene on chromosome 5q31, and its polymorphisms show an association with aspirin hypersensitivity in asthma J Clin Immunol 31 112 121 20922562 16 Kovacic MB Myers JM Wang N Martin LJ Lindsey M Ericksen MB He H Patterson TL Baye TM Torgerson D Roth LA Gupta J Sivaprasad U Gibson AM Tsoras AM Hu D Eng C Chapela R Rodriguez-Santana JR Rodriguez-Cintron W Avila PC Beckman K Seibold MA Gignoux C Musaad SM Chen W Burchard EG Hershey GK 2011 Identification of KIF3A as a novel candidate gene for childhood asthma using RNA expression and population allelic frequencies differences PLoS One 6 e23714 21912604 17 Hirokawa N 2000 Stirring up development with the heterotrimeric kinesin KIF3 Traffic 1 29 34 11208056 18 Hirokawa N Noda Y Tanaka Y Niwa S 2009 Kinesin superfamily motor proteins and intracellular transport Nat Rev Mol Cell Biol 10 682 696 19773780 19 Kodani A Salome Sirerol-Piquer M Seol A Garcia-Verdugo JM Reiter JF 2013 Kif3a interacts with Dynactin subunit p150 Glued to organize centriole subdistal appendages EMBO J 32 597 607 23386061 20 Lin F Hiesberger T Cordes K Sinclair AM Goldstein LS Somlo S Igarashi P 2003 Kidney-specific inactivation of the KIF3A subunit of kinesin-II inhibits renal ciliogenesis and produces polycystic kidney disease Proc Natl Acad Sci U S A 100 5286 5291 12672950 21 Marszalek JR Ruiz-Lozano P Roberts E Chien KR Goldstein LS 1999 Situs inversus and embryonic ciliary morphogenesis defects in mouse mutants lacking the KIF3A subunit of kinesin-II Proc Natl Acad Sci U S A 96 5043 5048 10220415 22 Takeda S Yonekawa Y Tanaka Y Okada Y Nonaka S Hirokawa N 1999 Left-right asymmetry and kinesin superfamily protein KIF3A: new insights in determination of laterality and mesoderm induction by Kif3a −/− mice analysis J Cell Biol 145 825 836 10330409 23 Goetz SC Anderson KV 2010 The primary cilium: a signalling center during vertebrate development Nat Rev Genet 11 331 344 20395968 24 Lancaster MA Gleeson JG 2009 The primary cilium as a cellular signaling center: lessons from disease Curr Opin Genet Dev 19 220 229 19477114 25 Quinlan RJ Tobin JL Beales PL 2008 Modeling ciliopathies: Primary cilia in development and disease Curr Top Dev Biol 84 249 310 19186246 26 Veland IR Awan A Pedersen LB Yoder BK Christensen ST 2009 Primary cilia and signaling pathways in mammalian development, health and disease Nephron Physiol 111 p39 53 19276629 27 Boehlke C Kotsis F Buchholz B Powelske C Eckardt KU Walz G Nitschke R Kuehn EW 2013 Kif3a guides microtubular dynamics, migration and lumen formation of MDCK cells PLoS One 8 e62165 23658710 28 Boehlke C Kotsis F Patel V Braeg S Voelker H Bredt S Beyer T Janusch H Hamann C Godel M Muller K Herbst M Hornung M Doerken M Kottgen M Nitschke R Igarashi P Walz G Kuehn EW 2010 Primary cilia regulate mTORC1 activity and cell size through Lkb1 Nat Cell Biol 12 1115 1122 20972424 29 Carpenter BS Barry RL Verhey KJ Allen BL 2015 The heterotrimeric kinesin-2 complex interacts with and regulates GLI protein function J Cell Sci 128 1034 1050 25588831 30 Qiu N Xiao Z Cao L Buechel MM David V Roan E Quarles LD 2012 Disruption of Kif3a in osteoblasts results in defective bone formation and osteopenia J Cell Sci 125 1945 1957 22357948 31 Sipe CW Lu X 2011 Kif3a regulates planar polarization of auditory hair cells through both ciliary and non-ciliary mechanisms Development 138 3441 3449 21752934 32 Haycraft CJ Zhang Q Song B Jackson WS Detloff PJ Serra R Yoder BK 2007 Intraflagellar transport is essential for endochondral bone formation Development 134 307 316 17166921 33 Ji H Houghton AM Mariani TJ Perera S Kim CB Padera R Tonon G McNamara K Marconcini LA Hezel A El-Bardeesy N Bronson RT Sugarbaker D Maser RS Shapiro SD Wong KK 2006 K-ras activation generates an inflammatory response in lung tumors Oncogene 25 2105 2112 16288213 34 Harfe BD Scherz PJ Nissim S Tian H McMahon AP Tabin CJ 2004 Evidence for an expansion-based temporal Shh gradient in specifying vertebrate digit identities Cell 118 517 528 15315763 35 Delmotte P Sanderson MJ 2006 Ciliary beat frequency is maintained at a maximal rate in the small airways of mouse lung slices Am J Respir Cell Mol Biol 35 110 117 16484686 36 Francis R Lo C 2013 Ex vivo method for high resolution imaging of cilia motility in rodent airway epithelia J Vis Exp 37 Li BW Hendriks RW 2013 Group 2 innate lymphoid cells in lung inflammation Immunology 140 281 287 23866009 38 Rajavelu P Chen G Xu Y Kitzmiller JA Korfhagen TR Whitsett JA 2015 Airway epithelial SPDEF integrates goblet cell differentiation and pulmonary Th2 inflammation J Clin Invest 125 2021 2031 25866971 39 Davidovich N DiPaolo BC Lawrence GG Chhour P Yehya N Margulies SS 2013 Cyclic stretch-induced oxidative stress increases pulmonary alveolar epithelial permeability Am J Respir Cell Mol Biol 49 156 164 23526210 40 Du Y Guo M Whitsett JA Xu Y 2015 'LungGENS': a web-based tool for mapping single-cell gene expression in the developing lung Thorax 70 1092 1094 26130332 41 Jain R Pan J Driscoll JA Wisner JW Huang T Gunsten SP You Y Brody SL 2010 Temporal relationship between primary and motile ciliogenesis in airway epithelial cells Am J Respir Cell Mol Biol 43 731 739 20118219 42 Harris KS Zhang Z McManus MT Harfe BD Sun X 2006 Dicer function is essential for lung epithelium morphogenesis Proc Natl Acad Sci U S A 103 2208 2213 16452165 43 Gilley SK Stenbit AE Pasek RC Sas KM Steele SL Amria M Bunni MA Estell KP Schwiebert LM Flume P Gooz M Haycraft CJ Yoder BK Miller C Pavlik JA Turner GA Sisson JH Bell PD 2014 Deletion of airway cilia results in noninflammatory bronchiectasis and hyperreactive airways Am J Physiol Lung Cell Mol Physiol 306 L162 169 24213915 44 Pazour GJ Dickert BL Vucica Y Seeley ES Rosenbaum JL Witman GB Cole DG 2000 Chlamydomonas IFT88 and its mouse homologue, polycystic kidney disease gene tg737, are required for assembly of cilia and flagella J Cell Biol 151 709 718 11062270 45 Buckpitt A Chang AM Weir A Van Winkle L Duan X Philpot R Plopper C 1995 Relationship of cytochrome P450 activity to Clara cell cytotoxicity. IV. Metabolism of naphthalene and naphthalene oxide in microdissected airways from mice, rats, and hamsters Mol Pharmacol 47 74 81 7838135 46 Shultz MA Morin D Chang AM Buckpitt A 2001 Metabolic capabilities of CYP2F2 with various pulmonary toxicants and its relative abundance in mouse lung subcompartments J Pharmacol Exp Ther 296 510 519 11160638 47 Hammad H Lambrecht BN 2015 Barrier Epithelial Cells and the Control of Type 2 Immunity Immunity 43 29 40 26200011 48 Sehgal IS Dhooria S Bal A Agarwal R 2015 Allergic bronchopulmonary aspergillosis in an adult with Kartagener syndrome BMJ Case Rep 2015 49 Sharma B Sharma M Bondi E Sharma M 2005 Kartagener's syndrome associated with allergic bronchopulmonary aspergillosis MedGenMed 7 25 50 Balenga NA Klichinsky M Xie Z Chan EC Zhao M Jude J Laviolette M Panettieri RA Jr Druey KM 2015 A fungal protease allergen provokes airway hyper-responsiveness in asthma Nat Commun 6 6763 25865874 51 Perkins C Yanase N Smulian G Gildea L Orekov T Potter C Brombacher F Aronow B Wills-Karp M Finkelman FD 2011 Selective stimulation of IL-4 receptor on smooth muscle induces airway hyperresponsiveness in mice J Exp Med 208 853 867 21464224 52 Damera G Tliba O Panettieri RA Jr 2009 Airway smooth muscle as an immunomodulatory cell Pulm Pharmacol Ther 22 353 359 19121407 53 Georas SN Rezaee F 2014 Epithelial barrier function: at the front line of asthma immunology and allergic airway inflammation J Allergy Clin Immunol 134 509 520 25085341 54 Heijink IH Nawijn MC Hackett TL 2014 Airway epithelial barrier function regulates the pathogenesis of allergic asthma Clin Exp Allergy 44 620 630 24612268 55 Bush A Hogg C 2012 Primary ciliary dyskinesia: recent advances in epidemiology, diagnosis, management and relationship with the expanding spectrum of ciliopathy Expert Rev Respir Med 6 663 682 23234452 56 Leigh MW Pittman JE Carson JL Ferkol TW Dell SD Davis SD Knowles MR Zariwala MA 2009 Clinical and genetic aspects of primary ciliary dyskinesia/Kartagener syndrome Genet Med 11 473 487 19606528 57 Guajardo JR Schleifer KW Daines MO Ruddy RM Aronow BJ Wills-Karp M Hershey GK 2005 Altered gene expression profiles in nasal respiratory epithelium reflect stable versus acute childhood asthma J Allergy Clin Immunol 115 243 251 15696077 58 Thomas B Rutman A Hirst RA Haldar P Wardlaw AJ Bankart J Brightling CE O'Callaghan C 2010 Ciliary dysfunction and ultrastructural abnormalities are features of severe asthma J Allergy Clin Immunol 126 722 729 e722 20673980 59 Dagenais TR Keller NP 2009 Pathogenesis of Aspergillus fumigatus in Invasive Aspergillosis Clin Microbiol Rev 22 447 465 19597008 60 Gregory LG Lloyd CM 2011 Orchestrating house dust mite-associated allergy in the lung Trends Immunol 32 402 411 21783420 61 Fahy JV 2015 Type 2 inflammation in asthma--present in most, absent in many Nat Rev Immunol 15 57 65 25534623 62 Lambrecht BN Hammad H 2013 Asthma: the importance of dysregulated barrier immunity Eur J Immunol 43 3125 3137 24165907 63 Erle DJ Sheppard D 2014 The cell biology of asthma J Cell Biol 205 621 631 24914235 64 Chen G Korfhagen TR Karp CL Impey S Xu Y Randell SH Kitzmiller J Maeda Y Haitchi HM Sridharan A Senft AP Whitsett JA 2014 Foxa3 induces goblet cell metaplasia and inhibits innate antiviral immunity Am J Respir Crit Care Med 189 301 313 24392884 65 Chen G Korfhagen TR Xu Y Kitzmiller J Wert SE Maeda Y Gregorieff A Clevers H Whitsett JA 2009 SPDEF is required for mouse pulmonary goblet cell differentiation and regulates a network of genes associated with mucus production J Clin Invest 119 2914 2924 19759516 66 Veland IR Lindbaek L Christensen ST 2014 Linking the Primary Cilium to Cell Migration in Tissue Repair and Brain Development Bioscience 64 1115 1125 26955067 67 Blitzer AL Panagis L Gusella GL Danias J Mlodzik M Iomini C 2011 Primary cilia dynamics instruct tissue patterning and repair of corneal endothelium Proc Natl Acad Sci U S A 108 2819 2824 21285373 68 Boehlke C Janusch H Hamann C Powelske C Mergen M Herbst H Kotsis F Nitschke R Kuehn EW 2015 A Cilia Independent Role of Ift88/Polaris during Cell Migration PLoS One 10 e0140378 26465598 69 Cousins DJ Lee TH Staynov DZ 2002 Cytokine coexpression during human Th1/Th2 cell differentiation: direct evidence for coordinated expression of Th2 cytokines J Immunol 169 2498 2506 12193719 70 Frazer KA Ueda Y Zhu Y Gifford VR Garofalo MR Mohandas N Martin CH Palazzolo MJ Cheng JF Rubin EM 1997 Computational and biological analysis of 680 kb of DNA sequence from the human 5q31 cytokine gene cluster region Genome Res 7 495 512 9149945
PMC005xxxxxx/PMC5123830.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 7905840 6445 Pharmacol Ther Pharmacol. Ther. Pharmacology & therapeutics 0163-7258 1879-016X 27108948 5123830 10.1016/j.pharmthera.2016.03.015 NIHMS781451 Article Renoprotective Approaches and Strategies in Acute Kidney Injury Yang Yuan 1 Song Meifang 1 Liu Yu 1 Liu Hong 1 Sun Lin 1 Peng Youming 1 Liu Fuyou 1* Venkatachalam Manjeri A. 2 Dong Zheng 13* 1 Department of Nephrology, The Second Xiangya Hospital, Central South University, Changsha, Hunan, China 2 Department of Pathology, University of Texas Health Science Center at San Antonio, TX 2 Department of Cellular Biology & Anatomy, Medical college of Georgia at Augusta University and Charlie Norwood VA Medical Center, Augusta, GA * Corresponding Author: Zheng Dong, PhD, Department of Nephrology, The Second Xiangya Hospital, Central South University, Changsha, Hunan, China; Department of Cellular Biology and Anatomy, Medical College of Georgia and Charlie Norwood VA Medical Center, 1459 Laney Walker Blvd, Augusta, GA 30912. Phone: (706) 721-2825; Fax: (706) 721-6120; [email protected], Fu-you Liu, MD, Department of Nephrology, The Second Xiangya Hospital, Central South University, Changsha, Hunan, China. [email protected] 17 11 2016 22 4 2016 7 2016 01 7 2017 163 5873 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Acute kidney injury (AKI) is a major renal disease associated with a high mortality rate and increasing prevalence. Decades of research has suggested numerous chemical and biological agents with beneficial effects in AKI. In addition, cell therapy and molecular targeting have been explored for reducing kidney tissue damage and promoting kidney repair or recovery from AKI. Mechanistically, these approaches may mitigate oxidative stress, inflammation, cell death, and mitochondrial and other organellar damage, or activate cytoprotective mechanisms such as autophagy and pro-survival factors. However, none of these findings has been successfully translated into clinical treatment of AKI. In this review, we analyze these findings and propose experimental strategies for the identification of renoprotective agents or methods with clinical potential. Moreover, we propose the consideration of combination therapy by targeting multiple targets in AKI. Acute kidney injury Kidney protection Kidney repair Renoprotection Ischemia-reperfusion Nephrotoxicity Mitochondria Apoptosis Reactive oxygen species Vascular dysfunction Inflammation Introduction Acute kidney injury (AKI) is a syndrome characterized by the rapid loss of renal function resulting in the accumulation of end products of nitrogen metabolism (urea and creatinine) and/or decreased urine output (KDIGO, 2012). In clinic, AKI occurs mainly as the clinicopathological outcome of renal or extra-renal surgery, bacterial infection, and nephrotoxicity, Large epidemiological studies show a high incidence of AKI in hospitalized patients and in general population (Bellomo et al., 2012; Hsu et al., 2007; Lameire et al., 2013). AKI is considered to be an important independent risk factor for mortality (Uchino et al., 2006). Patients with uncomplicated AKI have a mortality rate of up to 10%. In contrast, patients presenting with AKI and multiorgan failure have been reported to have mortality rates of over 50%. If renal replacement therapy is required, the mortality rate rises further to as high as 80% (Shusterman et al., 1987; Liaño et al., 1998). In addition, AKI is an important factor in the development and progression of chronic kidney disease (CKD) (Chawla et al., 2014; Venkatachalam et al., 2015). Pathogenetically, AKI is generally described as the injury of renal tubular epithelial cell and vasculature, accompanied by the activation of a robust inflammatory response (Bonventre & Yang, 2011; Molitoris, 2014; Linkermann et al., 2014). In addition, depending on its severity and duration, the damage may spread to glomerulus and interstitium resulting in a full blown, lasting disease. Along with the mechanistic research, a number of agents have been shown for their renoprotective effects in AKI models (Table 1–5), which include some clinical drugs, herbs, active chemicals, hormones, cytokines and growth factors. Moreover, molecular and cell therapies have been attempted with some promising results. In experimental models, these agents and approaches protected kidneys by suppressing inflammation, preserving vasculature, and/or directly preventing tubular cell injury and death (Figure 1). However, up-to-date none of them has been successfully translated to the bedside or the use in patients (Jo et al., 2007). In this review, we have summarized the main renoprotective agents and analyzed their effects in AKI models and relevant mechanisms. We have also discussed the experimental strategies for the discovery of efficacious therapies for AKI, including the use of comorbid models and the test of combination therapies. I. Chemical Renoprotectants 1. Clinical drugs Some clinical drugs have been shown to be protective in experimental models of AKI. These include disease-modifying antirheumatic drugs (DMARD), cholesterol-cutting statins, neuroprotective agents for cerebral infarction, selective vitamin D receptor agonist (VDRA), tetracycline antibiotics, phosphodiesterase-5 (PDE5) inhibitors, angiotensin II receptor antagonist, mammalian target of rapamycin (mTOR) inhibitor, immunosuppressant drug, and steroid hormones (Table 1). A notable advantage of clinical drugs is that they have been thoroughly tested for safety in human use and, if effective, they can be relatively rapidly applied for AKI treatment. 1.1 Antirheumatic and statin drugs Leflunomide is known as an immunomodulating drug for the treatment of chronic inflammatory conditions, such as rheumatoid arthritis. In a rat model of renal ischemia-reperfusion injury (IRI), leflunomide markedly attenuated renal dysfunction and morphological alterations, and reduced oxidative stress (OS) (Karaman et al., 2006). Similarly, Etanercept (a soluble Tumor necrosis factor-alpha (TNF-α) receptor) showed anti-inflammatory and anti-apoptotic effects by lowering the expression of TNF-α and monocyte chemotactic protein-1 (MCP-1) in ischemic AKI rats (Choi et al., 2009). For statins, early postoperative statin use was associated with a lower incidence of AKI after cardiac surgery and decreased mortality risk as compared to preoperative statin use or acute statin withdrawal (Molnar et al., 2011; Billings et al., 2010). Several mechanisms have been suggested to contribute to the renoprotective effects of statins in AKI. Statins with their antioxidant, anti-inflammatory and anti-apoptotic effects may protect kidney against gentamicin-, cisplatin- and cyclosporine-induced nephrotoxicity, beyond their lipid-lowering capacity (Dashti-Khavidaki et al., 2013; Kostapanos et al., 2009). They may also block the activation of mitogen-activated protein kinase (MAPK) and the redox-sensitive NF-kB and activator protein-1 (AP-1) (Gueler et al., 2002). Also statins may ameliorate AKI by directly affecting renal vasculature, an observation that is particularly relevant to sepsis-associated AKI (Yasuda et al., 2006). 1.2 Neuroprotective drugs and Vitamin D receptor agonist Edaravone is a neuroprotective drug used for treating cerebral infarction through its antioxidant property. In ischemic AKI, edaravone showed renoprotective effects as indicated by decreased serum creatinine (SCr) and blood urea nitrogen (BUN), and increased Bcl-2 expression (Watanabe et al., 2004; Li et al., 2010). Paricalcitol, an agonist of the vitamin D receptor, protected against ischemic AKI by upregulating cyclooxygenase-2 (COX-2) and prostaglandin E2 (PGE2) to attenuate inflammation (Hwang et al., 2013). In line with this observation, Vitamin D deficiency aggravated AKI induced by Tenofovir, a widely used component of antiretroviral regimens for HIV treatment (Canale et al., 2013). 1.3 Inhibitor of phosphodiesterase type 5 Tadalafil and Sildenafil are inhibitors of PDE5, the enzyme responsible for cyclic GMP degradation. Clinically, they are common drugs prescribed for the treatment of erectile dysfunction (ED) and pulmonary hypertension. In ischemic AKI, Tadalafil significantly improved renal function and preserved renal histology, which was associated with the attenuation of AKI biomarkers including kidney injury molecule 1 (KIM-1) and neutrophil gelatinase-associated lipocalin (NGAL) (Sohotnik et al., 2013). Similarly, Sildenafil reduced contrast medium-induced AKI in rabbits (Lauver et al., 2014). 1.4 Angiotensin II receptor antagonist and anti-ulcer drug The angiotensin II receptor antagonist telmisartan was shown to attenuate the increases in BUN, SCr, malondialdehyde (MDA), TNF-α, NO and homocysteine levels in ischemic AKI (Fouad et al., 2010). Consistently, adrenomedullin (AM), a potent endogenous vasodilatory peptide hormone, also delayed the development of contrast-induced nephropathy (CIN) by negative regulation of the renin-angiotensin-aldosterone system (RAAS) (Charles et al., 2003; Inal et al., 2013). Interestingly, the combination of AM with AM binding protein-1 (AMBP-1) could markedly attenuate the inflammatory response in ischemic AKI, suggesting a mechanism of the renoprotective effect of AM (Shah et al., 2010). 1.5 Antibiotics and immunosuppressants Tetracyclines exhibit significant anti-inflammatory and antiapoptotic properties in AKI induced by hypoxia, azide, cisplatin, or ischemia. For example, minocycline induced accumulation of Bcl-2 in mitochondria and suppression of death-promoting molecules including Bax, Bak, and Bid (Wang et al., 2004), and reduced leukocytes infiltration, leukocyte chemotaxis, and the expression of intercellular adhesion molecule-1 (ICAM-1) (Kelly et al., 2004). Minocycline also reduced renal microvascular leakage which may be related to diminished activity of matrix metallopeptidase 2 (MMP-2) and MMP-9 on the perivascular matrix (Sutton et al., 2005). However, in a clinical study minocycline did not show significant protective effects against AKI that developed post-cardiac bypass surgery (Golestaneh et al., 2015). Doxycycline as another tetracycline antibiotics exhibited renoprotective effects by decreasing levels of IL-1β, TNF-α and MMP-2 in renal tissue against IRI induced by abdominal compartment syndrome (ACS) (Ihtiyar et al., 2011). Both minocycline and doxycycline were effective in mitigating liver and kidney injury to improve survival in the mouse model of hemorrhagic shock/resuscitation (Kholmukhamedov et al., 2014). Cyclosporin A, an immunosuppressant drug used in organ transplantation to prevent rejection, blocked the TNF-like weak inducer of apoptosis (TWEAK) expression and NF-κB activation in folic acid (FA)-induced AKI (Wen et al., 2012). Rituximab is a monoclonal antibody against the protein CD20 used in autoimmune diseases or anti-rejection treatment for organ transplants, which may suppress the inflammation in ischemic AKI (Hwang et al., 2013). In addition, treatment with mycophenolate mofetil together with polyphenolic bioflavonoids reduced tubular damage and attenuated the induction of inflammatory cytokines and renal inflammation (Jones et al., 2000). Thus, anti-inflammation appears to be a common mechanism underlying the renoprotective effects of antibiotics and immunosuppressants in AKI. 1.6 Inhibitors of mammalian target of rapamycin Mammalian target of rapamycin (mTOR) is a serine/ threonine protein kinase with multiple functions. On one hand, mTOR is a key to cell growth and proliferation by promoting protein synthesis. On the other hand, recent work has demonstrated that mTOR is a crucial, negative regulator of autophagy in response to nutritional status, growth factor and stress signals (Jung et al., 2010; Datta et al., 2014). In AKI, the role of mTOR varies according to experimental models. In ischemic AKI, the inhibition of mTOR by rapamycin impaired or at least delayed kidney repair and recovery by suppressing tubular cell growth and proliferation (Lieberthal et al., 2012). However, in cisplatin nephrotoxic AKI, rapamycin showed protective effects (Jiang et al., 2012). Rapamycin also protected renal tubular cells from apoptosis during ER-stress (Dong et al., 2015). Similarly, rapamycin ameliorated renal injury in diabetic mice and the underlying mechanism may be related to autophagy induction in podocytes (Xiao et al., 2014). In endotoxic AKI induced by lipopolysaccharide (LPS), another mTOR inhibitor Temsirolimus induced autophagy and protected against kidney injury, even after established endotoxemia (Howell et al., 2013). Therefore depending on the models, inhibition of mTOR may protect against or exacerbate AKI. The exact cause of the different effects is unclear, but apparently it results from the multiple functions of mTOR. While mTOR may protect by promoting cell growth and proliferation, it may also enhance injury by inactivating autophagy. Also it is important to recognize that inhibitors of mTOR (e.g. rapamycin) are immune suppressants that may diminish the inflammatory response in AKI contributing to the observed protective effects of rapamycin in vivo. 1.7 Other clinical drugs In addition to the drugs described above, several other clinically used drugs have shown renoprotective effects in AKI models. For example, suramin is an antiparasitic drug used for the treatment of trypanosomiasis. Zhuang and colleague demonstrated the beneficial effect of suramin in several kidney disease models, including ischemic AKI and renal fibrosis (Zhuang et al., 2009; Liu et al., 2011). Mechanistically, suramin may promote renal tubular cell proliferation and migration, processes important for kidney repair (Zhuang et al., 2005). Geranylgeranylacetone (GGA), a drug used in the treatment of gastric ulcers, ameliorated ischemic AKI via induction of heat shock protein 70 (Hsp70) (Suzuki et al., 2005). In addition, Fidarestat, an aldose reductase (AR) inhibitor used for treating diabetic complications, protected against LPS-induced endotoxic AKI probably by suppressing the inflammatory response (Kazunori et al., 2012). 2. Renoprotective Chemicals with Clinical Potential Heme oxygenase-1 (HO-1) and its activators: HO-1 is an inducible enzyme that converts heme into biliverdin and bilirubin, releasing iron and carbon monoxide. The potent cytoprotective role of HO-1 has been recognized for over 20 years (Nath et al., 1992). In kidneys, HO-1 is induced in various AKI models including ischemia-reperfusion, sepsis, and nephrotoxicity (Nath, 2014; Shimizu et al., 2000; Maines et al., 1993). Mechanistically, HO-1 is known to promote the anti-oxidative capacity of the cell. Moreover, it may also dilate blood vessels, increase perfusion, and suppress inflammation in AKI as a result of tissue protection or indirectly by modulating immune cell trafficking (Hull et al., 2015). Several studies have tested the effects of HO-1 induction in AKI. For example, tin chloride (SnCl2) ameliorated ischemic AKI as shown by the decrease in serum creatinine and BUN and in tubular damage (Toda et al., 2002), while the tin protoporphyrin/ Tin mesoporphyrin/ stannous mesoporphyrin (SnMP, a competitive inhibitor of HO) exacerbated AKI induced by cisplatin (Agarwal et al., 1995; Salom et al., 2007). Protein kinase C (PKC) inhibitors: PKC is a protein kinase family of multiple members, several of which are induced following renal IR injury in rats (Padanilam, 2001). In a rat model of kidney transplantation, the pan PKC inhibitor sotrastaurin attenuated tubular injury and accelerated renal recovery following transplantation (Fuller et al., 2012). In cisplatin nephrotoxicity, PKCδ was rapidly activated and the inhibition of PKCδ genetically or pharmacologically prevented kidney injury; notably PKCδ inhibitors also enhanced the chemotherapeutic effects of cisplatin in several tumor models, suggesting that blockade of PKCδ may be a “Kill two birds with one stone” strategy in cisplatin chemotherapy (Pabla et al., 2011). Other renoprotective chemicals: Renoprotective effects have also been shown for the Rho kinase inhibitor Y27632-lysozyme in ischemic AKI (Prakash et al., 2008). Moreover, zafirlukast, the antagonist of cysteinyl leukotriene-1 receptor (CysLT1R, a member of G protein-coupled receptors superfamily), was shown to alleviate ischemic AKI by reducing neutrophil infiltration as well as P-selectin overexpression in renal tissues (Hanan et al., 2012). Necrostatin-1, a specific inhibitor of the receptor-interacting protein 1 (RIP1) kinase, prevented necrotic cell death and partially preserved renal function during AKI induced by ischemia-reperfusion, contrast media, and cisplatin nephrotoxicity (Linkermann et al., 2013; Linkermann et al., 2012; Xu et al., 2015). In addition, the inhibitor of Na+/ Ca2+ exchange KB-R7943 may attenuate renal tubular cell death by suppressing the increases of renal endothelin-1 (ET-1) and catalase during ischemic AKI and contrast medium-induced nephrotoxicity (Yamashita et al., 2001; Yang et al., 2013). II. Herbs, Food and Dietary Nutrients A variety of herbs, food and dietary nutrients that showed renoprotective effects in AKI models (Table 2). 2.1 Herbs and derivatives Korean red ginseng is a traditional herbal medicine in China, Korea, and Japan, which was shown to attenuated renal dysfunction, cell apoptosis and tubular damage in cisplatin- and gentamicin-induced AKI mainly by reducing ROS and inflammation (Kim et al., 2014; Lee et al., 2013). Similarly, Radix Codonopsis and the extract saponins increased superoxide dismutase (SOD) level and decreased apoptosis index in a model of kidney transplantation (He et al., 2011), artemisia asiatica extract increased the level of HO-1 and Bcl-2 in the setting of acute renal IRI damage (Jang et al., 2015), and Ginkgo extract (ginaton) was shown to possess anti-oxidation and anti-inflammation activities through suppressing extrinsic apoptotic signal pathway induced by c-Jun N-terminal kinase (JNK) signal pathway (Wang et al., 2008). Interestingly, some bioactive extracts from herbs, such as flavonoids (naringin, quercetin, curcumin or hesperidin), flavanols (Catechin), Polyphenols (Resveratrol), and Saponin (Astragaloside IV), showed similar renoprotective effects with similar mechanisms. For example, quercetin, naringin, hesperidin, and catechin all reduced lipid perioxidation and restored the levels of antioxidant enzymes SOD and catalase in kidney tissues (Kahraman et al., 2003; Singh et al., 2004; Sahu et al., 2013; Singh et al., 2005). They also showed remarkable anti-inflammation effects (Shoskes, 1998). Resveratrol is known for its effects on life extension, cancer prevention, and antidiabetic effects (Howitz et al., 2003; Baur et al., 2006; Su et al., 2006). In the animal models of AKI induced by sepsis, IR, glycerol, or cisplatin, Resveratrol improved kidney microcirculation and protected tubular epithelium. Mechanistically, Resveratrol may work by scavenging reactive oxygen/nitrogen species (ROS/RNS), releasing nitric oxide (NO), activating sirtuin 1 (SIRT1) and inhibiting p53 to block apoptosis (Holthoff et al., 2012; Sener et al., 2006; Chander & Chopra, 2006; Kim et al., 2011; Chander & Chopra, 2006). Saponin prevented renal damage through inhibiting ROS and p38 kinase-associated apoptosis pathways in AKI induced by renal IRI or contrast medium (Gui et al., 2013). 2.2 Food and dietary nutrients Sulforaphane, an organosulfur compound enriched in cruciferous vegetables such as broccoli, protected against ischemic AKI probably by inducing the NF-E2-related factor-2 (Nrf2) antioxidative system (Yoon et al., 2008). Antioxidative activities were also shown for Sesame oil, which was renoprotective during aminoglycoside and iodinated contrast-induced AKI (Hsu et al., 2011; Hsu et al., 2010). At least two extracts from soybean have been shown to be renoprotective in AKI models. First, polyenylphosphatidycholine was shown to reduce serum levels of aspartate aminotransferase, BUN and NF-kB expression (Demirbilek et al., 2006). Second, isoflavone extracted from soybeans protected against ischemic AKI probably by inducing heme oxygenase (Watanabe et al., 2007). In addition, isoflavones, such as daidzein, formononetin, and genistein, may activate the expression of SIRT1 and PGC-1α to induce mitochondrial biogenesis, leading to accelerated recovery of mitochondrial and cellular functions for renoprotection (Rasbach & Schnellmann, 2008). III. Antioxidants and Mitochondrial Protectants Other chemicals with renoprotective effects in AKI include antioxidants and mitochondrial Protectants (Table 3). 3.1 Antioxidants Oxidant stress is a well-recognized pathogenic factor in AKI. ROS are produced excessively during AKI by several mechanisms. a. disruption of mitochondrial homeostasis results in electronic leak from the respiratory chain; b. macrophage phagocytosis of cellular debris leads to the release of a large amount of ROS; c. hypoxia-reoxygenation in kidney tissues decreases the cellular antioxidant activity (glutathione-GSH, antioxidant enzymes) resulting in redox imbalance (Funk et al., 2012; Samarasinghe et al., 2000; Martins et al., 2003). Consequently, excess ROS in cells induces oxidative damage of proteins, lipid membranes and biological macromolecules, and promotes inflammation and tissue damage. Glutathione (GSH) is a major cellular antioxidant that is synthesized by the precursors N-acetylcysteine (NAC), glutamine and glycine. NAC showed beneficial effects in various models of AKI and notably, in contrast-induced AKI patients (Kelly et al., 2008). In general, the renoprotective effect of NAC is attributed to improved levels of GSH and associated decrease of ROS in AKI (Duru et al., 2008; Briguori et al., 2011). However, in addition to antioxidation, glutamine may have other effects. For example, it may mitigate renal neutrophil infiltration and tubular cell apoptosis by inhibiting JNK and enhancing Hsp70 (Peng et al., 2013; Kim et al., 2009). Glycine is a classical cell plasma membrane protectant, which protects against kidney tubular cell death by a mechanism related to amino acid gated chloride channels rather than its anti-oxidant activity (Venkatachalam et al., 1996; Sogabe et al., 1996). In addition, cellular antioxidant enzymes, such as recombinant manganese superoxide dismutase (rMnSOD), reduced OS following contrast medium-induced AKI (Pisani et al., 2014). Consistently, deletion of extracellular SOD3 led to a more pronounced functional deterioration in AKI, supporting the beneficial effect of SOD (Schneider et al., 2010). 3.2 Mitochondrial protectants Pathologically, AKI is characterized by tubular cell injury and death. Under this condition, multiple forms of cell death are triggered and mediated by different pathways (Linkermann et al., 2014). Nonetheless, mitochondrial damage appears to be a common factor that induces tubular cell death in AKI. Mitochondrial permeability transition (MPT) at the inner membrane plays a critical role in tubular cell necrosis. As a result, the inhibition of MPT pharmacologically by cyclosporine A or genetically by cyclophilin D ablation led to an increased resistance of kidneys to ischemic AKI (Park et al., 2011; Feldkamp et al., 2009). At the outer membrane of mitochondria, Bax and Bak, two pro-apoptotic members of Bcl-2 family proteins, may co-operate to induce porous defects for the release of apoptotic factors, such as cytochrome c, leading to apoptosis. In ischemic AKI, GSK3β was suggested to activate Bax via phosphorylation and the pharmacological inhibitor of GSK3β, TDZD-8, could block Bax activation to afford significant renoprotective effects (Wang et al., 2010). Inerestingly, Nutlin-3, an murine double minute-2 (MDM2) inhibitor, was shown to directly antagonize Bax, resulting in the prevention of Bax/Bak oligomerization, inhibition of cytochrome c release, and suppression of apoptosis during cisplatin treatment of renal tubular cells (Jiang et al., 2007). Minocycline, a derivative of tetracycline, may up-regulate Bcl-2 in renal tubular cells to block Bax/Bak activation and apoptosis during hypoxia, ATP-depletion, and cisplatin injury (Wang et al., 2004). These studies support the therapeutic potential of the antagonists of Bax/Bak in AKI. Mitochondria are highly dynamic organelles that undergo fission and fusion (Brooks & Dong, 2007). In AKI, mitochondrial dynamics is disrupted, resulting in mitochondrial fragmentation, which can be partially prevented by mdivi-1, a mitochondrial fission inhibitor. Importantly, mdivi-1 provided significant protection against AKI (Brooks et al., 2009). This study not only supports a role of mitochondrial dynamics disruption in the pathogenesis of AKI but has also identified a new therapeutic strategy. Mechanistically, it was shown that the fragmented mitochondria are more sensitive to Bax insertion (Brooks et al., 2011). More recent work by Xiao et al has further shown the regulation of mitochondrial fragmentation by inner membrane protease OMA1 cleaving (Optic atrophy 1) OPA1 in ischemic AKI (Xiao et al., 2014). Several antioxidant agents have been reported to specifically target mitochondria and provide renoprotective effects in AKI. For example, Zorov and colleagues developed SKQR1, a positively charged mitochondrial-targeting compound carrying an antioxidative moiety, which showed renoprotective effects in rat models of ischemic and glycerol-induced AKI (Plotnikov et al., 2011). Mechanistically, SkQR1 may protect by inhibiting MPP and scavenging excessive ROS. Szeto and colleagues have synthesized SS-31, a mitochondria-targeted tetrapeptide with antioxidant property (Szeto et al., 2011). In ischemic AKI, SS-31 protected mitochondrial structure and function, reduced tubular cell death, and partially preserved renal function. Interestingly, the effects of SS-31 may be related to its interaction with cardiolipin (Birk et al., 2013), a specific type of liplid found in the inner membrane of mitochondria. In addition to limiting mitochondrial damage, another strategy is to promote mitochondrial biogenesis during and following AKI. In this regard, Schnellmann and colleagues reported that the SIRT1 activator SRT1720 could activate PGC-1α for mitochondrial biogenesis, leading to the accelerated recovery from ischemic AKI (Funk & Schnellmann, 2013). Their more recent work further demonstrated that formoterol, a potent β2-adrenergic agonist, induced renal mitochondrial biogenesis and enhanced renal recovery from ischemic injury. Remarkably, formoterol was effective even when given 24 hours after injury (Jesinkey et al., 2014), expanding the time window of treatment of clinical significance. IV. Hormones with Renoprotective Activities Several kinds of hormones are known for their protective effects in AKI (Table 4). 4.1 Sex hormones Several female sex hormones are known to be renoprotective in AKI. 17β-estradiol (E2), the primary female hormone, is a good example. E2 was shown to protect renal endothelial barrier function in AKI following cardiac arrest and cardiopulmonary resuscitation (Hutchens et al., 2010; Hutchens et al., 2012). Mechanistically, E2 may attenuate renal injury through the activation of phosphatidylinositol-3 kinase (PI3K)/Akt/endothelial nitric oxide synthase (eNOS) pathway (Satake et al., 2008) and by suppressing the renal sympathetic nervous system (SNS) (Tanaka et al., 2012). The pregnancy hormone Relaxin was also protective in AKI and the underlying mechanism may be related to the suppression of TNF-α-related inflammation and apoptosis (Yoshida et al., 2013; Yoshida et al., 2014). Similarly, Oxytocin attenuated ischemic AKI by decreasing TNF-α and oxidative damage (Tuğtepe et al., 2007). Renoprotective effect has also been demonstrated for AQGV, an oligopeptide related to the primary structure of human chorionic gonadotropin (beta-hCG), another pregnancy hormone (Khan et al., 2009). Currently, it is controversial whether the male hormone testosterone/ dihydrotestosterone is good or bad in AKI. Over a decade ago, Park and colleagues suggested a critical role for testosterone in the susceptibility of males to ischemic AKI (Park et al., 2004), and Attia, et al suggested that male gender increases sensitivity to renal injury due to lower renal NOS activity than female rats (Attia et al., 2003). Followup studies have further provided mechanistic insights into the effect of testosterone, referring to decreased expression of histone deacetylase HDAC11 that was accompanied by an increase in PAI-1 expression (Kim et al., 2013). However, a recent study showed a dramatic decrease of serum testosterone during ischemic AKI; further, infusion of testosterone during renal IR protected the kidneys (Soljancic et al., 2013). Interestingly, low dose of testosterone significantly decreased cisplatin-induced nephrotoxicity, while administration of high-dose testosterone enhanced it (Rostami et al., 2014), suggesting a dual role for testosterone at low- or high- doses, respectively. 4.2 Melanocortins Melanocortins are a group of hormones causing increased pigmentation, which includes alpha-melanocyte-stimulating hormone (α-MSH) and adrenocorticotropic hormone (ACTH). Star and Colleagues (Chiao et al., 1997) demonstrated the renoprotective effect of α-MSH in ischemic AKI in mice and rats. Mechanistically, although the initial work suggested suppression of neutrophil activation and infiltration in kidneys as a mechanism, a followup study indicated the involvement of neutrophil-independent mechanism (Chiao et al., 1998). AP214, an analogue of α-MSH, was protective in septic and ischemic AKI by reducing NF-kB activation and splenocyte apoptosis (Doi et al., 2008; Simmons et al., 2010). Somewhat paradoxically, renoprotective effects were also shown for melatonin, the physiological antagonist of α-MSH. It was suggested that melatonin protected kidneys by improving the migration and survival of "early outgrowth" endothelial progenitor cells (eEPCs) (Patschan et al., 2012), a function that is unrelated to that of melanogenesis (Valverde et al., 1995). Moreover the protective effect of α-MSH appears to be AKI model dependent, because it did not ameliorate mercuric chloride (HgCl2)-induced AKI (Miyaji et al., 2002). Similar to α-MSH, ACTH also demonstrated renoprotective effects in specific AKI models. For example, Gong and colleagues recently showed that ACTH alleviated TNF-induced AKI. Moreover, ACTH appeared to be more efficacious than α-MSH in renoprotection in the septic AKI model of cecal ligation puncture (Si et al., 2013). The beneficial effects of ACTH may derive from both steroid-dependent mechanisms and melanocortin 1 receptor (MC1R)-mediated anti-apoptotic effect. 4.3 Other hormones Beneficial effects of several other hormones have also been shown in some AKI models. For example, dexamethasone reduced mitochondrial damage, the release of proapoptotic proteins, and the production of pro-inflammatory cytokines in septic AKI following cecal ligation and puncture (Choi et al., 2013). In ischemic AKI, the stomach-derived peptide Ghrelin attenuated the vagus nerve-mediated systemic and kidney-specific inflammatory responses, resulting in significant preservation of renal histology and function (Rajan et al., 2012). Stanniocalcin-1 (STC1) is known as a regulator of calcium and phosphate transport and cellular calcium/phosphate homeostasis (Yeung & Wong, 2011). Sheikh-Hamad and colleagues demonstrated notable renoprotective effects of STC1 in ischemic AKI, which may be related to the induction of mitochondrial uncoupling protein 2 (UCP-2) and suppression of superoxide generation in ischemic AKI (Huang et al., 2012). Their latest work further suggested the activation of AMP-activated protein kinase (AMPK) as an upstream key to the effects of STC1 (Pan et al., 2015). Finally, the neuropeptide pituitary adenylate cyclase activating polypeptide (PACAP) prevented Bcl-2 decrease and apoptosis in ischemic AKI (Horvath et al., 2010), and consistently, PACAP deficiency was associated with an increased susceptibility to ischemic AKI (Szakaly et al., 2011). V. Cytokines and Growth Factors 5.1 Cytokines In general, increases of cytokines such as chemokines, TNF-α or ICAM-1 are implicated in the robust inflammatory response observed in AKI. Accordingly, blockade of these cytokines, their receptors or related signaling reduces inflammation and the associated kidney damage. In this regard, renoprotective effects in AKI have been demonstrated for ICAM-1 monoclonal antibodies, the CXCR4 (CXC chemokine receptor 4) inhibitor Plerixafor, and the TNF-α inhibitor pentoxifylline (Zuk et al., 2014; Ramesh et al., 2002; Kelly et al., 1996; Kelly et al., 1994). On the other hand, some other cytokines have notable renoprotective effects. For example, IL-10 is known to inhibit the increases of TNF-α, ICAM-1, and iNOS and protect against ischemic and cisplatin-induced AKI (Deng et al., 2001). In recent work, cardiotrophin-1 (CT-1), a member of the interleukin 6 (IL-6) family, showed significant protective effects in AKI induced by contrast medium (Quiros et al., 2013). In addition, some cytokines may directly protect renal tubules. For example, the lipocalin NGAL inhibited the activation of caspase-3 and reduced Bax expression and renal tubular cell apoptosis in ischemic AKI in rats (An et al., 2013). L-FABP (Liver-type fatty acid-binding proteins) attenuated aristolochic acid-induced nephrotoxicity likely through its antioxidant activity in renal tubules (Matsui et al., 2011). Furthermore, there are cytokines that are beneficial to hemodynamics or angiogenesis in AKI and kidney recovery following AKI. This is well-exemplified by soluble thrombomodulin (sTM) and Cartilage oligomeric matrix protein-angiopoietin-1 (COMP-Ang1), which improved microvascular erythrocyte flow rates and reduced microvascular endothelial leukocyte rolling and attachment during ischemic AKI (Sharfuddin et al., 2009). By improving peritubular capillary and enhancing renal tissue (re)perfusion, these cytokines were shown to alleviate ischemic kidney injury (Kim et al., 2006; Jung et al., 2009). 5.2 Growth factors Growth factors are signaling molecules between cells that promote cellular proliferation and differentiation by binding to specific receptors on the surface of their target cells. It is known that the activation of growth factor-mediated signaling pathways is important for the survival, migration and proliferation of renal tubular cell during AKI and subsequent renal recovery or repair (He et al., 2013; Tang et al., 2013; Zhou et al., 2013; Mason et al., 2014). In addition, various growth factors including insulin-like growth factor (IGF), epidermal growth factor (EGF), Milk fat globule-epidermal growth factor-factor VIII (MFG-E8), and hepatocyte growth factor (HGF), exerted beneficial effects in models of ischemic-, cisplatin-, HgCl2-, or glycerol- AKI (Miller et al., 1992; Yasuda et al., 2004; Friedlaender et al., 1995; Matsuda et al., 2011; Yen et al., 2015; Homsi et al., 2009; Chen et al., 2013). These growth factors, when added exogenously, protected against initial injury, enhanced kidney repair and accelerated recovery of renal function. In addition, HGF or IGF-1 expressing mesenchymal stem cells (MSCs) showed a high therapeutic efficacy in ischemic- or cisplatin- AKI models; notably, the efficacy appeared to rely on the growth factor expression on these cells, providing further support for the therapeutic potential of specific growth factors (Imberti et al., 2007; Chen et al., 2011). Renoprotective effects of haematopoietic growth factors in AKI have also been reported. Especially, Erythropoietin (EPO), named for its function of stimulating red blood cell generation, has been shown to protect against AKI in several models. The tissue protective effect of EPO appears to be largely independent on red blood cell production; instead, EPO may inhibit cell death and promote cellular repair and regeneration (Sharples & Yaqoob, 2006; Moore & Bellomo, 2011). In addition, granulocyte colony-stimulating factor (G-CSF) has been shown to ameliorate rhabdomyolysis-associated AKI, and interestingly the protective effect may be mediated by the induction of HO-1 (Wei et al., 2011). However, it is important to note that adverse effects of growth factors have also been reported. For example, IGF-1 enhanced the inflammatory response as indicated by increased neutrophil filtration in a rat model of ischemic AKI, which was associated with higher mortality rate (Fernández et al., 2001). More recently, it was shown that erlotinib (selective EGFR tyrosine kinase inhibitor) partially prevented cisplatin-induced AKI in rats, implying an injurious role for EGFR signaling (Wada et al., 2014). In addition, in post-AKI kidneys, growth factors may promote renal fibrosis. For example, EGFR mutant mice showed more severe AKI following renal ischemia (consistent with a protective role of EGFR signaling in acute injury), but these mice developed less interstitial fibrosis 28 days later, suggesting a role of EGFR signaling in renal fibrogenesis (Tang et al., 2013). Thus, in terms of AKI, the role played by a growth factor or its receptor-mediated signaling may depend on where and when the pathway is activated. This critical question requires detailed research using inducible, tissue-specific conditional gene knockout models (Chen et al., 2012). VI. Agents targeting gene expression 6.1 Transcription factors AKI is associated with a significant change in gene expression profile. Thus, it is not surprising that a number of transcription factors may participate in tissue injury as well as protection and repair. Here nuclear factor kappa B (NF-κB) and hypoxia inducible factors (HIF) are briefly discussed as examples. NF-κB is well-known as an inflammation promoting transcription factor that contributes to immune cell infiltration and cytokine production in AKI. In 2004, Cao and colleagues reported that transfection of NF-κB decoy oligodeoxynucleotides abolished NF-κB activation in ischemic AKI, resulting in decreases in MCP-1 and ICAM-1 expression, suppression of monocyte/ macrophage infiltration, and significant attenuation of tissue damage (Cao et al., 2004). Consistently, NF-κB activation was inhibited by pharmacologic agents such as milrinone and resveratrol or overexpression of SIRT1, resulting in a better preservation of renal histology and function in ischemic-AKI and cisplatin nephrotoxic (Jung et al., 2014; Jung et al., 2012). Blockade of NF-κB was also implicated in the protective effect of Nrf2 signaling (Jiang et al., 2014). In contrast to NF-κB, HIF are generally regarded as protective transcription factors in AKI. There are at least 3 members in the HIF family, i.e., HIF-1, -2, and -3. Functional HIF is a heterodimer protein consisting of α and β subunits. In response to hypoxia, HIF-α is stabilized and then associates with HIF-β to translocate into the nucleus to induce the transcription of target genes (Semenza, 2014). HIF-1 plays a pivotal role in the regulation of renal physiology and patho-physiology (Haase, 2013). Pharmacological as well as genetic up-regulation of HIF afforded renoprotective effects in ischemic and nephrotoxic AKI models (Matsumoto et al., 2003; Weidemann et al., 2008; Hill et al., 2008; Fähling et al., 2013; Conde et al., 2012), suggesting a therapeutic potential. The protective effect of HIF may involve the expression of genes for oxygen delivery, cell survival, and metabolic adaptation. It is noteworthy that HIF may function in different cell types in kidneys: while HIF-1 was generally believed to be the key HIF for renoprotection, recent work by Kapitsinou and colleague however suggests that HIF-2 of endothelial cells may be mainly responsible for the observed protective effects (Kapitsinou et al., 2014). From the point of therapeutics, it is important to note that HIF is also a critical factor for renal fibrosis following AKI (Kapitsinou et al., 2012), it is therefore critical to time the treatment to maximize the protective effect and minimize the fibrogenic effect. 6.2 microRNAs MicroRNAs are endogenously produced, small RNA molecules that negatively regulate target gene expression mainly by blocking their translation. Recent work has demonstrated the important roles played by microRNAs in renal development, physiology, and pathogenesis of various kidney diseases (Trionfini et al., 2015; Chung & Lan, 2015; Badal & Danesh, 2015; Marrone & Ho, 2014). The role of microRNAs in AKI was first demonstrated by using a conditional knockout model in which Dicer, a key enzyme for microRNA biogenesis, was ablated specifically from renal proximal tubules in mice. In this model, microRNAs were largely depleted from kidney tissues and remarkably, the animals were resistant to ischemic AKI (Wei et al., 2010). By microarray analysis, 13 microRNAs were shown to be significantly up- or down-regulated during ischemic AKI and the latest work has begun to delineate the regulations of these microRNAs and determine their pathological roles. For example, microRNA-687 was shown to be induced dramatically via HIF-1 in ischemic AKI and, upon induction, this microRNA targets phosphatase and tensin homolog (PTEN) to mediate tubular cell death and renal tissue damage (Bhatt et al., 2015). Interestingly, the microRNA expression profiles of bilateral ischemic AKI (Wei et al., 2010) was quite different from that of unilateral ischemia (Godwin et al., 2010), suggesting the sensitivity of microRNA expression. In cisplatin nephrotoxicity, microRNA-34a was shown to be induced via p53 and contributed to cell survival because antagonism of miR-34a with specific antisense oligonucleotides increased cell death during cisplatin treatment (Bhatt et al., 2010). In addition to these earlier studies, more recent studies have further identified miR-24, miR-127, miR-687, and miR-126 as critical regulators of ischemic AKI. For example, Lorenzen and colleagues demonstrated that the silencing of miR-24 ameliorated apoptotic responses and histologic tubular damage in ischemic AKI, resulting in a significant improvement in survival and kidney function (Lorenzen et al., 2014). Also as alluded above, blockade of miR-687 also protected against ischemic AKI (Bhatt et al., 2015). While the induction of some microRNAs have also been reported as beneficial in AKI. For example, miR-127 was shown to protect against ischemic AKI by targeting kinesin family member 3B (KIF3B), which is involved in the regulation of cell-matrix and cell-cell adhesion maintenance (Aguado-Fraile et al., 2012). In cisplatin nephrotoxicity, miR-34a appeared to promote renal tubular cell survival. Consistently, miR-155-deficient mice demonstrated heightened kidney toxicity following cisplatin treatment, supporting a protective role of this microRNA (Pellegrini et al., 2014). The recent work by Bijkerk and colleagues further suggested that overexpression of miR-126 in the hematopoietic compartment can facilitate vascular regeneration and renal recovery from AKI likely by mobilizing and homing hematopoietic stem and progenitor cells (Bijkerk et al., 2014). Thus, some microRNAs are protective whereas others being injurious in AKI, and targeting of specific microRNAs may offer an effective strategy for the treatment of AKI. 6.3 Epigenetic regulators A new development in AKI research is the recognition of the involvement of epigenetic regulation in kidney injury and subsequent recovery or repair (Tang, Dong, 2015; Tang, Zhuang, 2015). Epigenetics refers to heritable mechanisms that alter gene expression without changing DNA sequence. DNA methylation and post-translation histone modifications (e.g. acetylation) are major epigenetic mechanisms that may keep the chromatin in an ‘open’ or ‘closed’ configuration to facilitate or block gene expression. The earliest evidence for the contribution of epigenetic regulation in AKI came from the study of the effects of the inhibitors of histone deacetylase (HDAC). In 2008, we reported that two HDAC inhibtors, suberoylanilide hydroxamic acid and Trichostatin A, were toxic to renal tubular cells at relatively high concentrations (Dong et al., 2008), but at lower dosages they were protective against cisplatin-induced apoptosis in these cells (Dong et al., 2010). These studies suggested the involvement of epigenetic regulation in AKI and notably, the effect of HDAC inhibitors depended on their dosages. Consistently, MS-275 (another HDAC inhibitor) worsened AKI and prevented kidney repair in the mouse models of AKI induced by folic acid or rhabdomyolysis (Tang et al., 2014), whereas Trichostatin A and methyl-4-(phenylthio)butanoate were recently shown to be beneficial to ischemic AKI (Levine et al., 2015; Cianciolo et al., 2013). Thus, the effects of HDAC inhibitors depend on their specificity, dosages of use, and AKI models of test. Regardless, these studies support a role of epigenetic regulation in AKI and kidney repair following AKI. Recent studies have begun to delineate the specific epigenetic mechanisms in AKI. For example, Bomsztyk and colleagues have recently provided comprehensive information about the epigenetic modifications of histones in mouse models of AKI induced by renal ischemia/reperfusion and lipopolysaccharide (Mar et al., 2015). Further investigation in this area is expected to reveal specific epigenetic mechanisms that may provide effective therapeutic targets for AKI. A partial list of cytokines, growth factors and proteins with renoprotective effects in AKI is provided in Table 5. VII. Cell Therapy 7.1 Stem cells Depending on their differentiation potentials, bone marrow derived stem cells (BMSC) are classified into hematopoietic stem cells (HSCs) and MSCs. BMSC showed renoprotective effects in different AKI models in numerous studies. Earlier studies suggested that BMSC may differentiate into renal tubules for kidney repair after AKI (Kale et al., 2003). But later studies indicated that differentiation of BMSC into renal tubular cells for repair, if any, is a very rare event (Li et al., 2007; Duffield et al., 2005). In these studies, the protective effects were mainly attributable to Mesenchymal stem cells (MSCs/BM-MSCs) (Tögel & Westenfelder, 2010; Morigi & Benigni, 2013; Fleig & Humphreys, 2014). As alluded above, rather than differentiation into renal tubular cells, MSCs home to the injury sites and mainly function by producing paracrine factors that limit injury in renal tubules in AKI and/or facilitate the kidney repair. For example, knockdown of IGF-1 in MSCs led to a marked reduction of the cells’ protective ability in cisplatin-induced AKI (Imberti et al., 2007). Similarly, knockdown of VEGF in MSCs significantly reduced their efficacy in protection against ischemic AKI in rats (Tögel et al., 2009). Interestingly, Hu and colleagues further reported that MSCs mainly accumulated in lung and spleen, and their renoprotective effect in AKI may be related to the induction of T regulatory cells (Hu et al., 2013), suggesting a renoprotective mechanism for MSCs from distant organs, especially the spleen. In addition to bone marrow, MSCs derived from other tissues also showed the beneficial effects on AKI. For example, the Wharton's jelly-derived mesenchymal stromal cells (WJ-MSC) improved renal function following renal ischemia, which was associated with a stronger proliferative response, less apoptosis and less fibrotic lesions and HGF may be an important contributor to the effects of WJ-MSC (Du et al., 2012). Similarly, adipose tissue-derived MSCs ameliorated folic acid- and cisplatin-induced AKI by producing HGF, VEGF and other factors (Katsuno et al., 2013; Yasuda et al., 2012). In addition to MSCs, recent work has demonstrated the beneficial effect of the exosomes derived from MSCs in AKI induced by ischemia and cisplatin (Gatti et al., 2011; Bruno et al., 2012). Exosomes, containing specific proteins, mRNAs and microRNAs, are released from various cells and can fuse with neighboring cells to deliver their contents as a means of communication or supplementation. Thus, the exosomes from MSCs may offer a more efficient way to getting access to injured renal tubules for protection and kidney repair. Obviously, a focus of future investigation is to optimize the condition of MSCs or exosomes derived there from for therapeutic use. In this regard, several bioactive agents have been reported to enhance the renoprotective effects of MSCs. For example, Mias and colleagues reported that melatonin pretreatment could significantly increase the survival of MSCs, their paracrine activity of producing HGF and FGF, and the beneficial effect of MSCs in ischemic kidney (Mias et al., 2008). Genetic modification of MSCs is another option to improve the efficacy of renoprotection. For example, overexpression of CXCR4 (the alpha-chemokine receptor specific for SDF-1/CXCL12) improved the reparative ability of MSCs in AKI by enhancing their homing to injured kidneys and production of cytokines such as BMP-7, HGF, and IL-10 (Liu et al., 2013). 7.2 Endothelial progenitor cells EPCs are bone marrow–derived, circulating progenitor cells of the endothelial lineage (Asahara et al., 1997). Interestingly, patients suffering from sepsis-induced AKI showed a significantly higher level of circulating EPCs (Patschan et al., 2011). In AKI, microvascular endothelial cell dysfunction results in a decline of perfusion in peritubular capillaries, leading to the suppression of kidney repair or recovery. In 2006, Patschan and colleagues (Patschan et al., 2006) demonstrated the mobilization and homing of EPCs to injured kidneys in ischemic AKI. Importantly, transplantation or systemic administration of EPCs afforded renoprotective effect (Patschan et al., 2010). Interestingly, Li and colleagues showed that prior induction of hematopoietic stem and progenitor cells (HSPC) before application may provide a better protection by producing renotrophic factors including VEGF, IGF-1, and HGF that promote epithelial proliferation and tubular repair (Li et al., 2012). Similarly to that of MSCs, microvesicles or exosomes derived from EPCs were shown to attenuate ischemic AKI, notably, by harboring endothelial-protective miRNAs such as miR-126 and microRNA-dependent reprogramming of resident renal cells (Bitzer et al., 2012). 7.3 T lymphocytes In addition to their well-recognized injurious role, research in recent years has established a protective role for specific subsets of lymphocytes (Jang et al., 2015). Especially, the depletion of TXPβ(+)CD4(+)CD25(+)Foxp3(+) regulatory T cells (Tregs) after ischemic injury led to enhanced pro-inflammatory cytokines production, increased renal tubular damage, and reduced tubular proliferation, while infusion of Tregs enhanced kidney repair and recovery (Gandolfo et al., 2009; Kinsey et al., 2009). These and other follow-up studies indicate that the pathological role of T cells in AKI depends on the cell subtype and the stage of injury. How to specifically stimulate Tregs for renoprotection? Lai and colleagues identified the potential in N, N-dimethylsphingosine (DMS), a naturally occurring sphingosine derivative and sphingosine kinase inhibitor. DMS was shown to recruit Tregs and protect against ischemic AKI; notably, the protective effect of DMS was abolished when Tregs were depleted (Lai et al., 2012), suggesting that DMS protects kidneys by recruiting Tregs. Research in this direction may lead to the development of therapeutic agents for clinical application, whereas cell therapy using Tregs may be technically more challenging. VIII. Final thoughts on the strategies for identifying renoprotective agents As discussed, numerous agents and approaches have been reported to be effective in protecting against AKI in experimental models. However, most have yet to enter clinical trial (Faubel et al., 2012). For those tested in patients, none has been successfully translated into clinical use. The reason can be many, including the complexity of the pathogenesis of AKI, the heterogeneity of the patients, and the defects in the design of previous clinical trials, just to name a few. On the bench side, it is crucial to thoroughly verify the effects of potential protective agents before considering or proposing clinical tests. The verification needs to cross-checked against multiple AKI models and also considers comorbid factors. Currently, mouse and rat models are most commonly used for AKI research and for the test of potential renoprotective agents. Compared to mammals, the rodent models have notable merits, including the feasibility of transgenics. However, rodents are known to have major differences in the structural organization of kidneys. Especially, compared to mammals (e.g. dog), rodents have a relatively thicker renal medulla and a more complex vasculature that leads to the unique feature of “non-reflow” following ischemic injury. As such, many renoprotective agents shown in rodent ischemic models may fail in the models of higher animals since those agents mainly target the “non-reflow” phenomenon. Thus, it is important to verify the effect in rodent experiments by using higher animals, such as pig, dog, or sheep (Figure 2). Clinically, there are various causes of AKI, which may be broadly divided into sepsis, nephrotoxicity, and renal ischemia-reperfusion. It is noteworthy that these causes are not mutually exclusive and in many cases, they co-exist. For example, ischemic injury may be an important component in nephrotoxic AKI due to toxic damage of vasculature and ensuing ischemia in kidney tissues. Importantly, while the cause of AKI is known for some patients (e.g. renal ischemia following cardiac surgery or nephrotoxicity after cisplatin chemotherapy), the cause of AKI for the majority of patients is unclear at admission. Under these conditions, it would be ideal to have a treatment that has a broad therapeutic spectrum. To discover such therapies, it is necessary to examine the effect in AKI models of different pathogenic origins (Figure 3). If the renoprotective effect of an agent is verified in two or more models, the chance of success in clinical trials is higher. In addition, it is well recognized that AKI in young and otherwise healthy patients is mostly completely reversible. However, in clinic settings, a large portion of AKI patients also suffer from comorbid conditions, such as diabetes, hypertension, CKD, and/or aging. It is in this population of patients that AKI is severe, hard to recover, and likely to progress into end-stage renal disease or chronic kidney disease. Unfortunately, most previous studies investigated AKI in young and healthy adult animals without considering the comorbid factors that are known to have profound effects on the outcome in AKI patients. In this regard, AKI in aging has been studied for years (Rosner, 2013; Wang et al., 2014). Moreover, recent studies have begun to test comorbid models. For example, cisplatin nephrotoxicity has been investigated in tumor-bearing animal models (Pabla et al., 2011; Oh et al., 2014). and ischemic AKI examined in diabetic animals (Kelly et al., 2009; Peng et al., 2015; Gao et al., 2013). The comorbid models are obviously more complex; however, they are also more relevant to the patient condition and, as a result, renoprotective agents identified from these models are more likely to succeed at the bedside (Figure 4). Finally, depending on the etiology, AKI is mostly a combined result of the damage and dysfunction in kidney parenchymal and mesenchymal tissues, especially renal tubules, vasculatures, and immune response and inflammation. In view of such a complex pathogenesis, it is hard to envision a “silver bullet” for its optimal treatment. Rather, less specific, “dirty” drugs with multiple targets might be more effective. In this regard, cell therapy may be a good example. In addition, it is also important to consider the strategy of combination therapy, which takes advantage of the differential renoprotective effects of two or more agents. As presented in this review, various classes of renoprotective agents, including clinical drugs, herbs, natural or synthetic chemicals, bio-active proteins or peptides, and stem cells, have been described (Table 1–5). Notably, these agents have multiple and diverse mechanisms of protection, ranging from anti-oxidation, anti-inflammation, anti-apoptosis, and mitochondrial protection, to the activation of autophagy and other pro-survival pathways (Figure 1). Can the agents be used in combination to achieve better protective effects? Theoretically it is plausible. For example, it seems logical to combine a renal tubule protectant with an anti-inflammatory agent. However, the idea of combination therapy has rarely been tested, even in animal models (Liu et al., 2013). In summary, decades of research has gained significant insights into the pathogenesis of AKI. Along the research, various renoprotective agents have been identified. Further investigation may cross-check their efficacy in multiple AKI models and also in comorbid models containing comorbid factors. Moreover, therapeutic efficacy may be improved or optimized by combination therapies. This study was supported in part by grants from the National Natural Science Foundation of China (81430017), the Hunan Province Natural Science Foundation, China (No.2009TP-1066-2), the National Basic Research Program of China 973, program No. 2012CB517601, the scientific research project of Hunan Province education department (14C0911), and the National Institutes of Health and Department of Veterans Administration of USA. Abbreviations ACTH adrenocorticotropic hormone AIF apoptosis inducing factor AKI Acute kidney injury BMSC bone marrow derived stem cells CIN contrast-induced nephropathy COMP-Ang1 Cartilage oligomeric matrix protein-angiopoietin-1 CysLT1R cysteinyl leukotriene-1 receptor DMARD disease-modifying antirheumatic drugs eNOS endothelial nitric oxide synthase eEPCs endothelial progenitor cells HDAC histone deacetylase HSPC hematopoietic stem and progenitor cells IRI ischemia-reperfusion injury ICAM-1 intercellular adhesion molecule-1 JNK c-Jun N-terminal kinase KIF3B kinesin family member 3B KIM-1 kidney injury molecule 1 MAPK mitogen-activated protein kinase MCP-1 monocyte chemotactic protein-1 α-MSH alpha-melanocyte-stimulating hormone MFG-E8 Milk fat globule-epidermal growth factor-factor VIII MPT Mitochondrial permeability transition MSCs mesenchymal stem cells NGAL neutrophil gelatinase-associated lipocalin MDM2 murine double minute-2 MMP-2 matrix metallopeptidase 2 mTOR mammalian target of rapamycin PI3K phosphatidylinositol-3 kinase PACAP pituitary adenylate cyclase activating polypeptide RAAS renin-angiotensin-aldosterone system RANTES regulated upon activation normal T-cell expressed and secreted RIP1 receptor-interacting protein 1 TNF-α Tumor necrosis factor-alpha TWEAK TNF-like weak inducer of apoptosis VDRA vitamin D receptor agonist Fig. 1 Overview of renoprotective approaches in acute kidney injury. Insults, such as ischemia/reperfusion, sepsis, and various nephrotoxins, induces injury and death of renal tubular cells, vascular dysfunction, and inflammation, resulting in acute kidney injury and renal failure. Renoprotective agents may protect tubular cells, suppress inflammatory response, and/or maintain renal vasculture in AKI. Fig. 2 Experimental strategies for identifying renoprotective approaches for AKI: from rodent to mammalian models Fig. 3 Experimental strategies for identifying renoprotective approaches for AKI: from single to multiple models Fig. 4 Experimental strategies for identifying renoprotective approaches for AKI: from AKI-only to comorbid models Table 1 Clinical drugs with renoprotective effects in AKI No Name Characteristics Tested AKI model Mechanism 1 Leflunomide Pyrimidine synthesis inhibitor used in immunosuppressive diseases such as rheumatoid arthritis and psoriatic arthritis IRI in rat reduce oxidative stress 2 Etanercept TNF-α inhibitor used to treat autoimmune diseases IRI in rat lower expression of TNF-α and MCP-1 3 Statins drugs Inhibitors of HMG-CoA reductase used to lower cholesterol drug-, septic- and ischemic -induced AKI in rat or mice antioxidant, anti-inflammatory and anti-apoptotic 4 Edaravone Neuroprotective agent in acute brain ischemia and subsequent cerebral infarction IRI in rats increase Bcl-2 expression 5 Paricalcitol Analog of vitamin D2 active form, VDR agonist IRI in male C57BL/6 mice upregulate COX-2 and PGE2 6 Tadalafil, Sildenafil Phosphodiesterase type 5 inhibitor contrast-induced AKI in rabbits Inhibit protein kinase G 7 Milrinone Phosphodiesterase type 3 inhibitor IRI in mice Inhibit NF-κB activation 8 Fidarestat Aldose reductase inhibitor for diabetic complications LPS-induced endotoxic AKI suppress inflammation 9 Telmisartan Angiotensin II receptor antagonist used in hypertension IRI in rats decrease MDA, TNF-α, NO and homocysteine 10 Adrenomedullin A potent endogenous vasodilatory peptide hormone contrast induced AKI in rats negative regulation of the RAAS 11 Rituximab Monoclonal antibody against CD20 used in autoimmunity IRI in mice Suppression of inflammation 12 Cyclosporin A Immunosuppressant used in transplant medicine FA-induced AKI in mice block TWEAK expression and NF-κB activation 13 Mycophenolate mofetil Immunosuppressant used in transplant or autoimmune diseases IRI in rats attenuate the increase of cytokines RANTES and AIF 14 Temsirolimus Inhibitor of mammalian target of rapamycin (mTOR) septic-AKI in older adult mice induce autophagy 15 Doxycycline Tetracycline antibiotics for treating infections or inflammation IRI in a rat model of ACS decrease IL-1β, TNF-α and MMP-2 16 suramin An antiparasitic drug used in treatment of trypanosomiasis IRI in mice reduce tubular apoptosis and infiltrating leukocytes 17 Geranylgeranylac etone An antiulcer drug used in treatment of gastric ulcers IRI in rats induction of Hsp70 Table 2 Herbs, food and dietary nutrients with renoprotective effects in AKI No Name Characteristics Tested AKI model Mechanism 1 Korean Red Ginseng Perennial plants belonging to genus Panax of the Araliaceae AKI by cisplatin and gentamicin in rat reduce OS and inflammation 2 Radix Codonopsis (saponins) Perennial plants used frequently in traditional Chinese medicine IRI after kidney transplant in rat decrease lipid peroxidation and inhibit apoptosis 3 artemisia asiatica Wormwood, traditional uses include treating liver problems, joint pain, gastric reflux IRI in male C57BL/6 mice increase the level of HO-1 and Bcl-2 4 Ginkgo extract (ginaton) Herb extracts used for treating Alzheimer’s disease, memory loss, headache, et al. IRI in rats suppress extrinsic apoptotic pathway induced by JNK 5 naringin (flavonoids) A flavonoid in grapefruit metabolized to flavanone naringenin IRI in rats reduce TBARS, restore antioxidant enzymes 6 quercetin (flavonoids) A pigment with a molecular structure like or derived from flavone IRI in rats increase GSH levels and activities of SOD and CAT 7 hesperidin (flavonoids) A flavanone glycoside found abundantly in citrus fruits cisplatin-induced AKI in rats attenuate OS, inflammation, apoptosis/necrosis 8 curcumin (Flavonoids) A diarylheptanoid which is a member of the ginger family IRI in rats attenuate expression of RANTES, MCP-1 9 Catechin (Flavanols) Derivatives of flavans that are abundant in teas IRI in rats similar to naringin in rat kidney 10 Resveratrol (Polyphenols) A phenol found in red grapes, Japanese knotweed, etc septic-AKI in mice IRI in rats glycerol-ARF in rats cisplatin-AKI in mice Antooxidant, release NO, activate SIRT1 and inhibit p53 11 Astragaloside IV Marker compound in Astragali Radix IRI in rats inhibit OS and p38 MAPK phosphorylation 12 Sulforaphane A molecule within isothiocyanate from cruciferous vegetables H/R in HK2 RPTC IRI in mice induce Nrf2-dependent phase 2 enzymes 13 Sesame oil Extraction from sesame seeds containing Vit E, Vit B6, etc AAs and contrast-induced AKI in rats inhibit renal OS 14 Polyenylphosphatid ycholine A lecithin soybean extract IRI in rats reduce levels of AST, BUN and NF-kB 15 Isoflavones Phytoestrogens (plant estrogens) isolated from the soybean IRI in rats induce heme oxygenase Table 3 Other chemicals with renoprotective effects in AKI No Name Characteristics Tested AKI model Mechanism 1 N-acetylcysteine A precursor of the antioxidant glutathione AKI by contrast in human, various AKI models in mouse and rat reduce oxidative stress 2 Glutamine The abundant free amino acid in human blood while conditionally essential in states of illness or injury folic acid-induced AKI in CD-1 mice glycerol-induced AKI in rat inhibit JNK phosphorylation and enhancing Hsp70 3 Glycine The smallest amino acids found in proteins or natural products ATP-depleted MDCK cells, Menadione-induce d injury of RPTC target amino acid gated chloride channels 4 rMnSOD MnSOD recombinant generated by DNA technique contrast-induced AKI in rat reduce renal oxidative stress 5 TDZD-8 Pharmacological inhibitor of GSK3β ATP-depleted BUMPT cells, IRI in rats inhibit activation of GSK3β, Bax, and caspase 3 6 Nutlin-3 Small molecule antagonist of MDM2 Cisplatin-induced rat RPTC apoptosis suppress the activation of Bax/Bak 7 Minocycline Semisynthetic derivative of tetracycline hypoxia, et al-RPTC apoptosis, IRI in rats induction of Bcl-2 8 Mdivi-1 Selective cell-permeable inhibitor of mitochondrial fission protein DRP1 Azide, cisplatin- RPTC apoptosis, IRI in C57BL/6 mice attenuate mitochondrial fragmentation and apoptosis 9 OMA1 Mediator of mitochondrial inner membrane cleavage ATP-depleted RPTC, IRI in C57BL/6 mice mediate OPA1 proteolysis and mitochondrial fragmentation 10 SS-31 Synthetic cell-permeable tetrapeptide that targets and concentrates in mitochondrial inner membrane IRI in rats protect mitochondria by interacting with cardiolipin 11 SkQR1 Cationic rhodamine derivative linked to a plastoquinone molecule glycerol-, IR-induced AKI in rats inhibit MPP and scavenge ROS 12 SRT1720 Selective SIRT1 activator IRI in rats activate PGC-1α for mitochondrial biogenesis 13 Formoterol Specific β2-adrenergic agonist IRI in mice promote mitochondrial biogenesis and recovery 14 sotrastaurin Selective pan-PKC inhibitor kidney transplantation in rat inhibit the induced PKC in transplantation 15 Y27632 Coupled to lysozyme, selective Rho kinase inhibitor IRI in rats reduce KIM-1, vimentin, MCP-1 16 zafirlukast Antagonist of CysLT1R IRI in rats reduce neutrophil infiltration, P-selectin overexpression 17 Necrostatin-1 Specific inhibitor of RIP1 kinase contrast-induced AKI in mice prevent dilation of peritubular capillaries 18 KB-R7943 Inhibitor of Na+/ Ca2+ exchange IRI in mice contrast-induced AKI in rat suppress the increased ET-1 and catalase Table 4 Hormones with renoprotective effects in AKI No Name Characteristics Tested AKI model Mechanism 1 17β-estradiol The primary female hormone Ischemic AKI in mouse, rat activate PI3K/Akt/eNOS pathway, suppress renal SNS 2 Relaxin A hormone of insulin superfamily exists in ovary and breast of female or prostate and semen of male IRI in rats cisplatin-induced AKI in rat decrease plasma TNF-α levels and renal TNFR1 3 Oxytocin A neurohypophysial hormone stimulating uterine contraction during and after childbirth IRI in rats decrease TNF-α and oxidative damage 4 AQGV An oligopeptide related to the primary structure of beta-hCG IRI in mice decrease TNF-α, INF-γ, IL-6 and IL-10 5 Testosterone A androgen hormone secreted primarily by testicles IRI in rats attenuate the increase of urinary KIM-1and intrarenal TNF-α 6 α-MSH Hormones causing increased pigmentation, named as Melanocortins ischemic AKI in mice and rats suppressneutrophil activation and infiltration 7 ACTH septic AKI of cecal ligation puncture induce MC1R-mediated anti-apoptotic effect, 8 AP214 an α-MSH analogue septic AKI in mice, ischemic AKI in a porcine reduce NF-kB and splenocyte apoptosis 9 Melatonin the physiological antagonist of α-MSH ischemic AKI in C57Bl/6N mice improve the migration and survival of eEPCs 10 Ghrelin The hunger hormone produced in the gastrointestinal tract IRI in rats decrease kidney IL-6 and MPO activity, increase Bcl-2/Bax ratio 11 STC-1 A hormone regulating renal calcium/phosphate homeostasis IRI in mice activate AMPK induce UCP-2 of mitochondria 12 PACAP A hypophysiotropic hormone similar to vasoactive intestinal peptide IRI in rats prevent Bcl-2 decrease and apoptotic effects 13 Dexamethasone An artificial synthetic of Glucocorticoid hormone septic AKI in C57BL/6 mice reduce MD with preserved COI Table 5 Cytokines, growth factors and gene-interfered with renoprotective effects in AKI No Name Characteristics Tested AKI model Mechanism 1 IL-10 cytokine synthesis inhibitory factor Ischemic- and cisplatin- AKI in the mouse reduce levels of TNF-α, ICAM-1, and iNOS 2 CXCR4 antagonist Plerixafor, a small-molecule antagonist of CXCR4 IRI in rats reduce chemokines CXCL1, CXCL5 and IL-6 3 TNF-α inhibition Inhibitors of TNF-α production (GM6001, pentoxifylline), anti-TNF-α antibody, specific TNF-α knockout cisplatin- AKI in Swiss-Webster mice decrease levels of TNF-α, TGF-β, RANTES, MIP-2, MCP-1, and IL-1β 4 ICAM-1 inhibition Specific ICAM-1 knockout, Anti-1CAm-1 antibody IRI in mice IRI in rats attenuate neutrophil endothelial interactions 5 CT-1 A member of IL-6 family and a potent pleiotropic cytokine contrast-induced AKI in rats prevent tubular desepithelization and obstruction 6 NGAL A member of the lipocalin super family with diverse function IRI in rats inhibit activation of caspase-3 and expression of Bax 7 L-FABP A member of intracellular lipid-binding proteins involved in the transportation of fatty acids AA-induced AKI in mice suppress the production of HEL, HO-1, and receptor for AGEs 8 sTM A glycoprotein present on the membrane surface of endothelial cells in many organs, including lung, liver, and kidney. IRI in rats improve microvascular erythrocyte flow rates 9 COMP-Ang1 A soluble and potent Ang1 variant, act as the ligand for Tie2 tyrosine kinase receptor that is expressed on EC. unilateral ureteral obstruction-induce d renal fibrosis improve peritubular capillary and enhance renal tissue (re)perfusion 10 IGF A hormone similar in molecular structure to insulin IRI in rats cisplatin- RPTC cisplatin- or HgCl2- AKI in mice ameliorate acute tubular necrosis; produce pro-survival factor IGF-1 11 MFG-E8 A protein involved in marking apoptotic cells for phagocytosis IRI in C57BL/6 mice suppress renal inflammation 12 EGF A potent growth promoter to renal tubule cells, produced in large amounts in the kidney HgCl2- AKI in mice attenuate tubular necrosis 13 EGFR inhibitor erlotinib, a selective tyrosine kinase inhibitor that can block EGFR activity cisplatin- AKI in rats decrease apoptosis and proliferation of tubular cells 14 HGF A potent mitogen for parenchymal liver, epithelial and endothelial cells, as a ligand of MET oncoprotein glycerol-, gentamicin- AKI in rats attenuate tubulointerstitial injury, leukocyte infiltration and Th1 polarization 15 EPO Glycoprotein produced by the kidney that regulats red blood cell production in the bone marrow IR-, cisplatin- and contrast- AKI in mice or rats or pigs inhibit apoptosis and promote cellular regeneration 16 G-CSF Glycoprotein that stimulates bone marrow to produce granulocytes and stem cells glycerol- AKI In C57BL/6 mice induction of HO-1 17 NF-κB Blockade NF-κB decoy oligodeoxynucleotides; NF-κB inhibitor milrinone, resveratrol IRI in rats or mice decrease MCP-1 expression and monocyte infiltration 18 HIF-1 Ubiquitously expressed hypoxia-inducible transcription factor IR- or cisplatin- in rats ameliorate tubulointerstitial and vascular damage 19 HIF-2 Hypoxia-inducible transcription factor mainly expressed on endothelial cells IRI in mice protect from vascular damage and fibrosis 20 PKCδ knockdown A member of PKC subfamily involved in cell apoptosis cisplatin- AKI in mice, cisplatin-RPTCs Activated MAPKs for apoptosis and tissue damage 21 OMA1 knockdown a zinc metalloprotease located at mitochondrial inner membrane that is involved in mitochondrial inner membrane disruption in cell stress ATP-depleted RPTC, IRI in C57BL/6 mice mediate OPA1 proteolysis and mitochondrial fragmentation 22 Dicer deletion A key ribonuclease for microRNA production, Dicer deletion leads to a global downregulation of microRNAs IRI in mice depletion of the majority of microRNAs 23 miR-687, -24 blockade endogenous, noncoding, small RNAs that regulate expression and function of genes hypoxia-induced RPTC injury / apoptosis, IRI in mice attenuate cell cycle activation and apoptosis 24 miR-127, -34a, -155, -126 blockade ameliorate histologic tubular damage, apoptosis This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. Declarations All authors declare that they have no commercial or other conflicting interests. References Agarwal A Balla J Alam J Croatt AJ Nath KA Induction of heme oxygenase in toxic renal injury a protective role in cisplatin nephrotoxicity in the rat Kidney Int 1995 48 4 1298 1307 8569092 Aguado-Fraile E Ramos E Saenz-Morales D Conde E Blanco-Sanchez I Stamatakis K del Peso L Cuppen E Brune B Garcia Bermejo ML miR-127 Protects Proximal Tubule Cells against Ischemia/Reperfusion Identification of Kinesin Family Member 3B as miR-127 Target PLOS ONE 2012 7 e44305 22962609 An S Zang X Yuan W Zhuge Y Yu Q Neutrophil gelatinase-associated lipocalin (NGAL) may play a protective role against rats ischemia/reperfusion renal injury via inhibiting tubular epithelial cell apoptosis Ren Fail 2013 35 1 143 149 23151253 Asahara T Murohara T Sullivan A Silver M van der Zee R Li T Witzenbichler B Schatteman G Isner JM Isolation of putative progenitor endothelial cells for angiogenesis Science 1997 275 5302 964 967 9020076 Attia DM Goldschmeding R Attia MA Boer P Koomans HA Joles JA Male gender increases sensitivity to renal injury in response to cholesterol loading Am J Physiol Renal Physiol 2003 284 4 F718 F726 12488246 Badal SS Danesh FR MicroRNAs and their applications in kidney diseases Pediatr Nephrol 2015 30 5 727 740 24928414 Baur JA Sinclair DA Therapeutic potential of resveratrol the in vivo evidence Nat Rev Drug Discov 2006 5 6 493 506 16732220 Bellomo R Kellum JA Ronco C Acute kidney injury Lancet 2012 380 9843 756 766 22617274 Bhatt K Wei Q Pabla N Dong G Mi QS Liang M Mei C Dong Z MicroRNA-687 Induced by Hypoxia-Inducible Factor-1 Targets Phosphatase and Tensin Homolog in Renal Ischemia-Reperfusion Injury J Am Soc Nephrol 2015 26 7 1588 1596 25587068 Bhatt K Zhou L Mi QS Huang S She JX Dong Z MicroRNA-34a is induced via p53 during cisplatin nephrotoxicity and contributes to cell survival Mol Med 2010 16 9–10 409 416 20386864 Bijkerk R van Solingen C de Boer HC van der Pol P Khairoun M de Bruin RG van Oeveren-Rietdijk AM Lievers E Hematopoietic microRNA-126 protects against renal ischemia/reperfusion injury by promoting vascular integrity J Am Soc Nephrol 2014 25 8 1710 1722 24610930 Billings FT 4th Pretorius M Siew ED Yu C Brown NJ Early postoperative statin therapy is associated with a lower incidence of acute kidney injury after cardiac surgery J Cardiothorac Vasc Anesth 2010 24 6 913 920 20599398 Birk AV Liu S Soong Y Mills W Singh P Warren JD Seshan SV Pardee JD Szeto HH The mitochondrial-targeted compound SS-31 re-energizes ischemic mitochondria by interacting with cardiolipin J Am Soc Nephrol 2013 24 8 1250 1261 23813215 Bitzer M Ben-Dov IZ Thum T Microparticles and microRNAs of endothelial progenitor cells ameliorate acute kidney injury Kidney Int 2012 82 4 375 377 22846811 Bonventre JV Yang L Cellular pathophysiology of ischemic acute kidney injury J Clin Invest 2011 121 11 4210 4221 22045571 Briguori C Quintavalle C De Micco F Condorelli G Nephrotoxicity of contrast media and protective effects of acetylcysteine Arch Toxicol 2011 85 3 165 173 21140133 Brooks C Cho SG Wang CY Yang T Dong Z Fragmented mitochondria are sensitized to Bax insertion and activation during apoptosis Am J Physiol Cell Physiol 2011 300 3 C447 C455 21160028 Brooks C Dong Z Regulation of mitochondrial morphological dynamics during apoptosis by Bcl-2 family proteins a key in Bak? Cell Cycle 2007 6 24 3043 3047 18073534 Brooks C Wei Q Cho SG Dong Z Regulation of mitochondrial dynamics in acute kidney injury in cell culture and rodent models J Clin Invest 2009 119 5 1275 1285 19349686 Bruno S Grange C Collino F Deregibus MC Cantaluppi V Biancone L Tetta C Camussi G Microvesicles derived from mesenchymal stem cells enhance survival in a lethal model of acute kidney injury PLoS One 2012 7 3 e33115 22431999 Canale D de Braganca AC Goncalves J Shimizu MHM Volpini RA Andrade L Seguro AC Vitamin D deficiency aggravates tenofovir induced metabolic syndrome and renal failure Nephrol Dial Transplant 2013 28 115 Cao CC Ding XQ Ou ZL Liu CF Li P Wang L Zhu CF In vivo transfection of NF-kappaB decoy oligodeoxynucleotides attenuate renal ischemia/reperfusion injury in rats Kidney Int 2004 65 3 834 845 14871403 Chander V Chopra K Protective effect of nitric oxide pathway in resveratrol renal ischemia-reperfusion injury in rats Arch Med Res 2006 37 1 19 26 16314181 Chander V Chopra K Protective effect of resveratrol, a polyphenolic phytoalexin on glycerol-induced acute renal failure in rat kidney Ren Fail 2006 28 2 161 169 16538975 Charles CJ Lainchbury JG Nicholls MG Rademaker MT Richards AM Troughton RW Adrenomedullin and the renin-angiotensin-aldosterone system Regul Pept 2003 112 1–3 41 49 12667624 Chawla LS Eggers PW Star RA Kimmel PL Acute kidney injury and chronic kidney disease as interconnected syndromes N Engl J Med 2014 371 1 58 66 24988558 Chen J Chen JK Harris RC Deletion of the epidermal growth factor receptor in renal proximal tubule epithelial cells delays recovery from acute kidney injury Kidney Int 2012 82 1 45 52 22418982 Chen X Chen Z Wang H Xiong X Liu X Hu C Han Y Lu Y Wu Z Zhang Q Plasmid pUDK-HGF encoding human hepatocyte growth factor gene attenuates gentamicin-induced kidney injury in rats Exp Toxicol Pathol 2013 65 5 541 547 22551933 Chen Y Qian H Zhu W Zhang X Yan Y Ye S Peng X Li W Xu W Hepatocyte growth factor modification promotes the amelioration effects of human umbilical cord mesenchymal stem cells on rat acute kidney injury Stem Cells Dev 2011 20 1 103 113 20446811 Chiao H Kohda Y McLeroy P Craig L Housini I Star RA Alpha-melanocyte-stimulating hormone (α-MSH) protects against renal injury after ischemia in mice and rats J Clin Invest 1997 99 6 1165 1172 9077523 Chiao H Kohda Y McLeroy P Craig L Housini I Star RA Alpha-melanocyte-stimulating hormone inhibits renal injury in the absence of neutrophils Kidney Int 1998 54 3 765 774 9734601 Choi DE Jeong JY Lim BJ Na KR Shin YT Lee KW Pretreatment with the tumor nerosis factor-alpha blocker etanercept attenuated ischemia-reperfusion renal injury Transplant Proc 2009 41 9 3590 3596 19917350 Choi HM Jo SK Kim SH Lee JW Cho E Hyun YY Cha JJ Kang YS Cha DR Cho WY Kim HK Glucocorticoids attenuate septic acute kidney injury Biochem Biophys Res Commun 2013 435 4 678 684 23702481 Chung AC Lan HY MicroRNAs in renal fibrosis Front Physiol 2015 6 50 25750628 Cianciolo Cosentino C Skrypnyk NI Brilli LL Chiba T Novitskaya T Woods C West J Korotchenko VN McDermott L Day BW Davidson AJ Harris RC de Caestecker MP Hukriede NA Histone deacetylase inhibitor enhances recovery after AKI J Am Soc Nephrol 2013 24 6 943 953 23620402 Conde E Alegre L Blanco-Sánchez I Sáenz-Morales D Aguado-Fraile E Ponte B Hypoxia inducible factor 1-alpha (HIF-1 alpha) is induced during reperfusion after renal ischemia and is critical for proximal tubule cell survival PLoS One 2012 7 3 e33258 22432008 Dashti-Khavidaki S Moghaddas A Heydari B Khalili H Lessan-Pezeshki M Lessan-Pezeshki M Statins against drug-induced nephrotoxicity J Pharm Pharm Sci 2013 16 4 588 608 24210066 Datta K Suman S Fornace AJ Jr Radiation persistently promoted oxidative stress, activated mTOR via PI3K/Akt, and downregulated autophagy pathway in mouse intestine Int J Biochem Cell Biol 2014 57 167 176 25449263 Demirbilek S Karaman A Baykarabulut A Akin M Gürünlüoglu K Edali MN Polyenylphosphatidylcholine pretreatment ameliorates ischemic acute renal injury in rats Int J Urol 2006 13 6 747 753 16834655 Deng J Kohda Y Chiao H Wang Y Hu X Hewitt SM Miyaji T McLeroy P Nibhanupudy B Li S Star RA Interleukin-10 inhibits ischemic and cisplatin-induced acute renal injury Kidney Int 2001 60 6 2118 2128 11737586 Doi K Hu X Yuen PS Leelahavanichkul A Yasuda H Kim SM Schnermann J Jonassen TE Frøkiaer J Nielsen S Star RA AP214, an analogue of alpha-melanocyte-stimulating hormone, ameliorates sepsis-induced acute kidney injury and mortality Kidney Int 2008 73 11 1266 1274 18354376 Dong G Liu Y Zhang L Huang S Ding HF Dong Z mTOR contributes to ER stress and associated apoptosis in renal tubular cells Am J Physiol Renal Physiol 2015 308 3 F267 F274 25428129 Dong G Luo J Kumar V Dong Z Inhibitors of histone deacetylases suppress cisplatin-induced p53 activation and apoptosis in renal tubular cells Am J Physiol Renal Physiol 2010 298 2 F293 F300 19889954 Dong G Wang L Wang CY Yang T Kumar MV Dong Z Induction of apoptosis in renal tubular cells by histone deacetylase inhibitors, a family of anticancer agents J Pharmacol Exp Ther 2008 325 3 978 984 18310471 Du T Cheng J Zhong L Zhao XF Zhu J Zhu YJ Liu GH The alleviation of acute and chronic kidney injury by human Wharton's jelly-derived mesenchymal stromal cells triggered by ischemia-reperfusion injury via an endocrine mechanism Cytotherapy 2012 14 10 1215 1227 22920838 Duffield JS Park KM Hsiao LL Kelley VR Scadden DT Ichimura T Bonventre JV Restoration of tubular epithelial cells during repair of the postischemic kidney occurs independently of bone marrow-derived stem cells J Clin Invest 2005 115 7 1743 1755 16007251 Duru M Nacar A Yönden Z Kuvandik G Helvaci MR Söğüt S Protective effects of N-acetylcysteine on cyclosporine-A-induced nephrotoxicity Ren Fail 2008 30 4 453 459 18569921 Fähling M Mathia S Paliege A Koesters R Mrowka R Peters H Persson PB Neumayer HH Bachmann S Rosenberger C Tubular von Hippel-Lindau knockout protects against rhabdomyolysis-induced AKI J Am Soc Nephrol 2013 24 11 1806 1819 23970125 Faubel S Chawla LS Chertow GM Goldstein SL Jaber BL Liu KD Acute Kidney Injury Advisory Group of the American Society of Nephrology Ongoing clinical trials in AKI Clin J Am Soc Nephrol 2012 7 5 861 873 22442183 Feldkamp T Park JS Pasupulati R Amora D Roeser NF Venkatachalam MA Weinberg JM Regulation of the mitochondrial permeability transition in kidney proximal tubules and its alteration during hypoxia-reoxygenation Am J Physiol Renal Physiol 2009 297 F1632 F1646 19741014 Fernández M Medina A Santos F Carbajo E Rodríguez J Alvarez J Cobo A Exacerbated inflammatory response induced by insulin-like growth factor I treatment in rats with ischemic acute renal failure J Am Soc Nephrol 2001 12 9 1900 1907 11518783 Fleig SV Humphreys BD Rationale of mesenchymal stem cell therapy in kidney injury Nephron Clin Pract 2014 127 1–4 75 80 25343826 Fouad AA Qureshi HA Al-Sultan AI Yacoubi MT Al-Melhim WN Nephroprotective effect of telmisartan in rats with ischemia/reperfusion renal injury Pharmacology 2010 85 3 158 167 20150754 Friedlaender M Popovtzer MM Weiss O Nefesh I Kopolovic J Raz I Insulin-like growth factor-1 (IGF-1) enhances recovery from HgCl2-induced acute renal failure the effects on renal IGF-1, IGF-1 receptor, and IGF-binding protein-1 mRNA J Am Soc Nephrol 1995 5 10 1782 1791 7540432 Fuller TF Kusch A Chaykovska L Catar R Pützer J Hoff U Dragun D Protein kinase C inhibition ameliorates posttransplantation preservation injury in rat renal transplants Transplantation 2012 94 7 679 686 22932117 Funk JA Schnellmann RG Persistent disruption of mitochondrial homeostasis after acute kidney injury Am J Physiol Renal Physiol 2012 302 7 F853 F864 22160772 Funk JA Schnellmann RG Accelerated recovery of renal mitochondrial and tubule homeostasis with SIRT1/PGC-1α activation following ischemia-reperfusion injury Toxicol Appl Pharmacol 2013 273 2 345 354 24096033 Gandolfo MT Jang HR Bagnasco SM Ko GJ Agreda P Satpute SR Crow MT King LS Rabb H Foxp3+ regulatory T cells participate in repair of ischemic acute kidney injury Kidney Int 2009 76 7 717 729 19625990 Gao G Zhang B Ramesh G Betterly D Tadagavadi RK Wang W Reeves WB TNF-α mediates increased susceptibility to ischemic AKI in diabetes Am J Physiol Renal Physiol 2013 304 5 F515 F521 23283990 Gatti S Bruno S Deregibus MC Sordi A Cantaluppi V Tetta C Camussi G Microvesicles derived from human adult mesenchymal stem cells protect against ischaemia-reperfusion-induced acute and chronic kidney injury Nephrol Dial Transplant 2011 26 5 1474 1483 21324974 Godwin JG Ge X Stephan K Jurisch A Tullius SG Iacomini J Identification of a microRNA signature of renal ischemia reperfusion injury Proc Natl Acad Sci U S A 2010 107 32 14339 14344 20651252 Golestaneh L Lindsey K Malhotra P Kargoli F Friedman A El-Achkar TM Acute kidney injury after cardiac surgery is minocycline protective? J Nephrol 2015 28 2 193 199 25348221 Gueler F Rong S Park JK Fiebeler A Menne J Kunter U Postischemic acute renal failure is reduced by short-term statin treatment in a rat model J Am Soc Nephrol 2002 13 9 2288 2298 12191973 Gui D Huang J Liu W Guo Y Xiao W Wang N Astragaloside IV prevents acute kidney injury in two rodent models by inhibiting oxidative stress and apoptosis pathways Apoptosis 2013 18 4 409 422 23325448 Haase VH Mechanisms of hypoxia responses in renal tissue J Am Soc Nephrol 2013 24 4 537 541 23334390 Hanan H Hagar Raeesa Abd EI Tawab Cysteinyl leukotriene receptor antagonism alleviates renal injury induced by ischemia-reperfusion in rats J Surg Res 2012 178 1 e25 e34 22487384 He B Ying-Tian Z Xin-Gang Y Jing-Song S Guang-Hui W Tao L Protective effects of Radix Codonopsis on ischemia–reperfusion injury in rats after kidney transplantation Pediatr Surg Int 2011 27 11 1203 1212 21691763 He S Liu N Bayliss G Zhuang S EGFR activity is required for renal tubular cell dedifferentiation and proliferation in a murine model of folic acid-induced acute kidney injury Am J Physiol Renal Physiol 2013 304 4 F356 F366 23255615 Hill P Shukla D Tran MG Aragones J Cook HT Carmeliet P Maxwell PH Inhibition of hypoxia inducible factor hydroxylases protects against renal ischemia-reperfusion injury J Am Soc Nephrol 2008 19 1 39 46 18178798 Holthoff JH Wang Z Seely KA Gokden N Mayeux PR Resveratrol improves renal microcirculation, protects the tubular epithelium, and prolongs survival in a mouse model of sepsis-induced acute kidney injury Kidney Int 2012 81 4 370 378 21975863 Homsi E Janino P Amano M Saraiva Camara NO Endogenous hepatocyte growth factor attenuates inflammatory response in glycerol-induced acute kidney injury Am J Nephrol 2009 29 4 283 291 18824844 Horvath G Racz B Reglodi D Kovacs K Kiss P Gallyas F Jr Bognar Z Szabo A Effects of PACAP on mitochondrial apoptotic pathways and cytokine expression in rats subjected to renal ischemia/reperfusion J Mol Neurosci 2010 42 3 411 418 20229361 Howell GM Gomez H Collage RD Loughran P Zhang X Matthew R Rosengart Augmenting Autophagy to Treat Acute Kidney Injury during Endotoxemia in Mice PLoS ONE 2013 8 e69520 23936035 Howitz KT Bitterman KJ Cohen HY Lamming DW Lavu S Sinclair DA Small molecule activators of sirtuins extend Saccharomyces cerevisiae lifespan Nature 2003 425 6954 191 196 12939617 Hsu CY McCulloch CE Fan D Ordoñez JD Chertow GM Go AS Community-based incidence of acute renal failure Kidney Int 2007 72 2 208 212 17507907 Hsu DZ Li YH Chu PY Periasamy S Liu MY Sesame oil prevents acute kidney injury induced by the synergistic action of aminoglycoside and iodinated contrast in rats Antimicrob Agents Chemother 2011 55 6 2532 2536 21402854 Hsu DZ Liu CT Li YH Chu PY Liu MY Protective effect of daily sesame oil supplement on gentamicin-induced renal injury in rats Shock 2010 33 1 88 92 19487986 Hu J Zhang L Wang N Ding R Cui S Zhu F Xie Y Sun X Wu D Hong Q Li Q Shi S Liu X Chen X Mesenchymal stem cells attenuate ischemic acute kidney injury by inducing regulatory T cells through splenocyte interactions Kidney Int 2013 84 3 521 531 23615497 Huang L Belousova T Chen M DiMattia G Liu D Sheikh-Hamad D Overexpression of stanniocalcin-1 inhibits reactive oxygen species and renal ischemia/reperfusion injury in mice Kidney Int 2012 82 8 867 877 22695329 Hull TD Kamal AI Boddu R Bolisetty S Guo L Agarwal A Heme Oxygenase-1 Regulates Myeloid Cell Trafficking in AKI J Am Soc Nephrol 2015 26 9 2139 2151 25677389 Hutchens MP Fujiyoshi T Komers R Herson PS Anderson S Estrogen protects renal endothelial barrier function from ischemia-reperfusion in vitro and in vivo Am J Physiol Renal Physiol 2012 303 3 F377 F385 22622457 Hutchens MP Nakano T Kosaka Y Dunlap J Zhang W Herson PS Murphy SJ Anderson S Hurn PD Estrogen is renoprotective via a nonreceptor-dependent mechanism after cardiac arrest in vivo Anesthesiology 2010 112 2 395 405 20068453 Hwang HS Choi YA Park KC Yang KJ Choi HS Kim SH Pretreatment with rituximab prevents subsequent ischemia-reperfusion injury in mice kidney Nephrology Dialysis Transplantation 2013 28 96 Hwang HS Yang KJ Park KC Choi HS Kim SH Yang CW Pretreatment with paricalcitol attenuates inflammation in ischemia-reperfusion injury via the up-regulation of cyclooxygenase-2 and prostaglandin E2 Nephrol Dial Transplant 2013 28 5 1156 1166 23229926 Ihtiyar E Yaşar NF Erkasap N Köken T Tosun M Oner S Erkasap S Effects of doxycycline on renal ischemia reperfusion injury induced by abdominal compartment syndrome J Surg Res 2011 167 1 113 120 20080260 Imberti B Morigi M Tomasoni S Rota C Corna D Longaretti L Rottoli D Valsecchi F Benigni A Wang J Abbate M Zoja C Remuzzi G Insulin-like growth factor-1 sustains stem cell mediated renal repair J Am Soc Nephrol 2007 18 2921 2928 17942965 Inal S Koc E Okyay GU Pasaoglu O Gonul I Oyar E Pasaoglu H Guz G Protective effect of adrenomedullin on contrast induced nephropathy in rats Nephrology Dialysis Transplantation 2013 28 1 suppl 110 Jang HJ Jeong EK Kim SS Lee JH Oh MY Kang KS Protective effect of artemisia asiatica extract against renal ischemia-reperfusion injury in mice Exp Clin Transplant 2015 13 1 Suppl 377 382 Jang HR Rabb H Immune cells in experimental acute kidney injury Nat Rev Nephrol 2015 11 2 88 101 25331787 Jesinkey SR Funk JA Stallons LJ Wills LP Megyesi JK Beeson CC Schnellmann RG Formoterol restores mitochondrial and renal function after ischemia-reperfusion injury J Am Soc Nephrol 2014 25 6 1157 1162 24511124 Jiang M Pabla N Murphy RF Yang T Yin XM Degenhardt K White E Dong Z Nutlin-3 Protects Kidney Cells during Cisplatin Therapy by Suppressing Bax/Bak Activation J Biol Chem 282 4 2636 2645 Jiang M Wei Q Dong G Komatsu M Su Y Dong Z Autophagy in proximal tubules protects against acute kidney injury Kidney Int 2012 82 12 1271 1283 22854643 Jiang T Tian F Zheng H Whitman SA Lin Y Zhang Z Zhang N Zhang DD Nrf2 suppresses lupus nephritis through inhibition of oxidative injury and the NF-κB-mediated inflammatory response Kidney Int 2014 85 2 333 343 24025640 Jo SK Rosner MH Okusa MD Pharmacologic treatment of acute kidney injury why drugs haven't worked and what is on the horizon Clin J Am Soc Nephrol 2007 2 2 356 365 17699435 Jones EA Shoskes DA The effect of mycophenolate mofetil and polyphenolic bioflavonoids on renal ischemia reperfusion injury and repair J Urol 2000 163 3 999 1004 10688038 Jung CH Ro SH Cao J Otto NM Kim DH mTOR regulation of autophagy FEBS Lett 2010 584 7 1287 1295 20083114 Jung HS Joo JD Kim DW In JH Roh M Jeong JT Noh SJ Choi JW Effect of milrinone on the inflammatory response and NF-kB activeation in renal ischemia-reperfusion injury in mice Korean J Anesthesiol 2014 66 2 136 142 24624272 Jung YJ Kim DH Lee AS Lee S Kang KP Lee SY Jang KY Sung MJ Park SK Kim W Peritubular capillary preservation with COMP-angiopoietin-1 decreases ischemia-reperfusion-induced acute kidney injury Am J Physiol Renal Physiol 2009 297 4 F952 F960 19656917 Jung YJ Lee JE Lee AS Kang KP Lee S Park SK Lee SY Han MK Kim DH Kim W SIRT1 overexpression decreases cisplatin-induced acetylation of NF-κB p65 subunit and cytotoxicity in renal proximal tubule cells Biochem Biophys Res Commun 2012 419 2 206 210 22330808 Kahraman A Erkasap N Serteser M Köken T Protective effect of quercetin on renal ischemia/reperfusion injury in rats J Nephrol 2003 16 2 219 224 12768068 Kale S Karihaloo A Clark PR Kashgarian M Krause DS Cantley LG Bone marrow stem cells contribute to repair of the ischemically injured renal tubule J Clin Invest 2003 112 1 42 49 12824456 Kapitsinou PP Jaffe J Michael M Swan CE Duffy KJ Erickson-Miller CL Haase VH Preischemic targeting of HIF prolyl hydroxylation inhibits fibrosis associated with acute kidney injury Am J Physiol Renal Physiol 2012 302 9 F1172 F1179 22262480 Kapitsinou PP Sano H Michael M Kobayashi H Davidoff O Bian A Yao B Zhang MZ Harris RC Duffy KJ Erickson-Miller CL Sutton TA Haase VH Endothelial HIF-2 mediates protection and recovery from ischemic kidney injury J Clin Invest 2014 124 6 2396 2409 24789906 Karaman A Turkmen E Gursul C Tas E Fadillioglu E Prevention of renal ischemia/reperfusion-induced injury in rats by leflunomide Int J Urol 2006 13 11 1434 1441 17083399 Katsuno T Ozaki T Saka Y Furuhashi K Kim H Yasuda K Yamamoto T Sato W Tsuboi N Mizuno M Ito Y Imai E Matsuo S Maruyama S Low serum cultured adipose tissue-derived stromal cells ameliorate acute kidney injury in rats Cell Transplant 2013 22 2 287 297 22963874 Kazunori Takahashi Hiroki Mizukami Kosuke Kamata Wataru Inaba Noriaki Kato Chihiro Hibi Soroku Yagihashi Amelioration of Acute Kidney Injury in Lipopolysaccharide-Induced Systemic Inflammatory Response Syndrome by an Aldose Reductase Inhibitor, Fidarestat PLoS One 2012 7 e30134 22253906 Kelly AM Dwamena B Cronin P Bernstein SJ Carlos RC Meta-analysis effectiveness of drugs for preventing contrast-induced nephropathy Ann Intern Med 2008 148 4 284 294 18283206 Kelly KJ Burford JL Dominguez JH Postischemic inflammatory syndrome a critical mechanism of progression in diabetic nephropathy Am J Physiol Renal Physiol 2009 297 4 F923 F931 19656916 Kelly KJ Sutton TA Weathered N Ray N Caldwell EJ Plotkin Z Dagher PC Minocycline inhibits apoptosis and inflammation in a rat model of ischemic renal injury Am J Physiol Renal Physiol 2004 287 4 F760 F766 15172883 Kelly KJ Williams WW Jr Colvin RB Meehan SM Springer TA Gutierrez-Ramos JC Bonventre JV Intercellular adhesion molecule-1-deficient mice are protected against ischemic renal injury J Clin Invest 1996 97 4 1056 1063 8613529 Kelly KJ Williams WW Jr Colvin RB Bonventre JV Antibody to intercellular adhesion molecule 1 protects the kidney against ischemic injury Proc Natl Acad Sci U S A 1994 91 2 812 816 7904759 Khan NA Susa D van Den Berg JW Huisman M Ameling MH van den Engel S Roest HP Ijzermans JN Dik WA Benner R de Bruin RW Amelioration of renal ischaemia-reperfusion injury by synthetic oligopeptides related to human chorionic gonadotropin Nephrol Dial Transplant 2009 24 9 2701 2708 19633318 Kholmukhamedov A Czerny C Hu J Schwartz J Zhong Z Lemasters JJ Minocycline and doxycycline, but not tetracycline, mitigate liver and kidney injury after hemorrhagic shock/resuscitation Shock 2014 42 3 256 263 24978888 Kidney Disease, Improving Global Outcomes (KDIGO) Acute Kidney Injury Work Group KDIGO Clinical Practice Guidelines for Acute Kidney Injury Kidney Int 2012 2 Suppl 1 138 Kim DH Jung YJ Lee JE Lee AS Kang KP Kim W SIRT1 activation by resveratrol ameliorates cisplatin-induced renal injury through deacetylation of p53 Am J Physiol Renal Physiol 2011 301 2 F427 F435 21593185 Kim JI Jung KJ Jang HS Park KM Gender-specific role of HDAC11 in kidney ischemia- and reperfusion-induced PAI-1 expression and injury Am J Physiol Renal Physiol 2013 305 1 F61 F70 23657855 Kim W Moon SO Lee SY Jang KY Cho CH Koh GY Choi KS Yoon KH Sung MJ Kim DH Lee S Kang KP Park SK COMP-angiopoietin-1 ameliorates renal fibrosis in a unilateral ureteral obstruction model J Am Soc Nephrol 2006 17 9 2474 2483 16885409 Kim YJ Lee MY Son HY Park BK Ryu SY Jung JY Red ginseng ameliorates acute cisplatin-induced nephropathy Planta Med 2014 80 8–9 645 654 24963615 Kim YS Jung MH Choi MY Kim YH Sheverdin V Kim JH Ha HJ Park DJ Glutamine attenuates tubular cell apoptosis in acute kidney injury via inhibition of the c-Jun N-terminal kinase phosphorylation of 14-3-3 Crit Care Med 2009 37 6 2033 2044 19384198 Kinsey GR Sharma R Huang L Li L Vergis AL Ye H Ju ST Okusa MD Regulatory T cells suppress innate immunity in kidney ischemia-reperfusion injury J Am Soc Nephrol 2009 20 8 1744 1753 19497969 Kostapanos MS Liberopoulos EN Elisaf MS Statin pleiotropy against renal injury J Cardiometab Syndr 2009 4 1 E4 E9 19245508 Lai LW Yong KC Lien YH Pharmacologic recruitment of regulatory T cells as a therapy for ischemic acute kidney injury Kidney Int 2012 81 10 983 992 22189844 Lameire NH Bagga A Cruz D De Maeseneer J Endre Z Kellum JA Liu KD Mehta RL Pannu N Van Biesen W Vanholder R Acute kidney injury an increasing global concern Lancet 2013 382 9887 170 179 23727171 Lauver DA Carey EG Bergin IL Lucchesi BR Gurm HS Sildenafil citrate for prophylaxis of nephropathy in an animal model of contrast-induced acute kidney injury PLoS One 2014 9 11 e113598 25426714 Lee YK Chin YW Choi YH Effects of Korean red ginseng extract on acute renal failure induced by gentamicin and pharmacokinetic changes by metformin in rats Food Chem Toxicol 2013 59 153 159 23743120 Levine MH Wang Z Bhatti TR Wang Y Aufhauser DD McNeal S Liu Y Cheraghlou S Han R Wang L Hancock WW Class-specific histone/protein deacetylase inhibition protects against renal ischemia reperfusion injury and fibrosis formation Am J Transplant 2015 15 4 965 973 25708614 Li L Black R Ma Z Yang Q Wang A Lin F Use of mouse hematopoietic stem and progenitor cells to treat acute kidney injury Am J Physiol Renal Physiol 2012 302 1 F9 F19 21937606 Li L Truong P Igarashi P Lin F Renal and bone marrow cells fuse after renal ischemic injury J Am Soc Nephrol 2007 18 12 3067 3077 18003777 Li Y Xia AZ Xing SH Protective effect of edaravone against renal ischemia/reperfusion injury and compared with ischemic postconditioning in rats Yao Xue Xue Bao 2010 45 7 840 848 20931780 Liaño F Junco E Pascual J Madero R Verde E The spectrum of acute renal failure in the intensive care unit compared with that seen in other settings. The Madrid Acute Renal Failure Study Group Kidney Int 1998 66 Suppl S16 S24 Lieberthal W Levine JS Mammalian target of rapamycin and the kidney. I. The signaling pathway Am J Physiol Renal Physiol 2012 303 1 F1 F10 22419691 Linkermann A Bräsen JH Himmerkus N Liu S Huber TB Kunzendorf U Krautwald S Rip1 (receptor-interacting protein kinase 1) mediates necroptosis and contributes to renal ischemia/reperfusion injury Kidney Int 2012 81 8 751 761 22237751 Linkermann A Chen G Dong G Kunzendorf U Krautwald S Dong Z Regulated cell death in AKI J Am Soc Nephrol 2014 25 12 2689 2701 24925726 Linkermann A Heller JO Prókai A Weinberg JM De Zen F Krautwald S The RIP1-kinase inhibitor necrostatin-1 prevents osmotic nephrosis and contrast-induced AKI in mice J Am Soc Nephrol 2013 24 10 1545 1557 23833261 Liu N Patzak A Zhang J CXCR4-overexpressing bone marrow-derived mesenchymal stem cells improve repair of acute kidney injury Am J Physiol Renal Physiol 2013 305 7 F1064 F1073 23884141 Liu N Tolbert E Pang M Ponnusamy M Yan H Zhuang S Suramin inhibits renal fibrosis in chronic kidney disease J Am Soc Nephrol 2011 22 6 1064 1075 21617121 Liu P Feng Y Dong C Liu D Wu X Wu H Lv P Zhou Y Study on therapeutic action of bone marrow derived mesenchymal stem cell combined with vitamin E against acute kidney injury in rats Life Sci 2013 92 829 837 23499556 Lorenzen JM Kaucsar T Schauerte C Schmitt R Rong S Hübner A Scherf K Fiedler J Martino F Kumarswamy R Kölling M Sörensen I MicroRNA-24 Antagonism Prevents Renal Ischemia Reperfusion Injury J Am Soc Nephrol 2014 25 2717 2729 24854275 Maines MD Mayer RD Ewing JF McCoubrey WK Jr Induction of kidney heme oxygenase-1 (HSP32) mRNA and protein by ischemia/reperfusion possible role of heme as both promotor of tissue damage and regulator of HSP32 J Pharmacol Exp Ther 1993 264 1 457 462 8423544 Mar D Gharib SA Zager RA Johnson A Denisenko O Bomsztyk K Heterogeneity of epigenetic changes at ischemia/reperfusion- and endotoxin-induced acute kidney injury genes Kidney Int 2015 88 4 734 744 26061546 Marrone AK Ho J MicroRNAs potential regulators of renal development genes that contribute to CAKUT Pediatr Nephrol 2014 29 4 565 574 23996519 Martins PS Kallas EG Neto MC Dalboni MA Blecher S Salomão R Upregulation of reactive oxygen species generation and phagocytosis, and increased apoptosis in human neutrophils during severe sepsis and septic shock Shock 2003 20 3 208 212 12923490 Mason S Hader C Marlier A Moeckel G Cantley LG Met activation is required for early cytoprotection after ischemic kidney injury J Am Soc Nephrol 2014 25 2 329 337 24136921 Matsuda A Wu R Jacob A Komura H Zhou M Wang Z Aziz MM Wang P Protective effect of milk fat globule-epidermal growth factor-factor VIII after renal ischemia-reperfusion injury in mice Crit Care Med 2011 39 9 2039 2047 21666453 Matsui K Kamijo-Ikemorif A Sugaya T Yasuda T Kimura K Renal liver-type fatty acid binding protein (L-FABP) attenuates acute kidney injury in aristolochic acid nephrotoxicity Am J Pathol 2011 178 3 1021 1032 21356355 Matsumoto M Makino Y Tanaka T Tanaka H Ishizaka N Noiri E Fujita T Nangaku M Induction of renoprotective gene expression by cobalt ameliorates ischemic injury of the kidney in rats J Am Soc Nephrol 2003 14 7 1825 1832 12819242 Mias C Trouche E Seguelas MH Calcagno F Dignat-George F Sabatier F Piercecchi-Marti MD Daniel L Bianchi P Calise D Bourin P Parini A Cussac D Ex vivo pretreatment with melatonin improves survival, proangiogenic / mitogenic activity, and efficiency of mesenchymal stem cells injected into ischemic kidney Stem Cells 2008 26 7 1749 1757 18467662 Miller SB Martin DR Kissane J Hammerman MR Insulin-like growth factor I accelerates recovery from ischemic acute tubular necrosis in the rat Proc Natl Acad Sci U S A 1992 89 24 11876 11880 1465411 Miyaji T Hu X Star RA alpha-Melanocyte-simulating hormone and interleukin-10 do not protect the kidney against mercuric chloride-induced injury Am J Physiol Renal Physiol 2002 282 5 F795 F801 11934688 Molitoris BA Therapeutic translation in acute kidney injury the epithelial/endothelial axis J Clin Invest 2014 124 6 2355 2363 24892710 Molnar A Coca S Devereaux P Jain A Kitchlu A Luo J Statin use is associated with a lower incidence of acute kidney injury after major elective surgery J Am Soc Nephrol 2011 22 5 939 946 21493769 Moore E Bellomo R Erythropoietin (EPO) in acute kidney injury Ann Intensive Care 2011 1 1 3 21906325 Morigi M Benigni A Mesenchymal stem cells and kidney repair Nephrol Dial Transplant 2013 28 4 788 793 23258756 Nath KA Heme oxygenase-1 and acute kidney injury Curr Opin Nephrol Hypertens 2014 23 1 17 24 24275768 Nath KA Balla G Vercellotti GM Balla J Jacob HS Levitt MD Rosenberg ME Induction of heme oxygenase is a rapid, protective response in rhabdomyolysis in the rat J Clin Invest 1992 90 1 267 270 1634613 Oh GS Kim HJ Choi JH Shen A Choe SK Karna A Lee SH Jo HJ Yang SH Kwak TH Lee CH Park R So HS Pharmacological activation of NQO1 increases NAD+ levels and attenuates cisplatin-mediated acute kidney injury in mice Kidney Int 2014 85 3 547 560 24025646 Pabla N Dong G Jiang M Huang S Kumar MV Messing RO Dong Z Inhibition of PKCδ reduces cisplatin-induced nephrotoxicity without blocking chemotherapeutic efficacy in mouse models of cancer J Clin Invest 2011 121 7 2709 2722 21633170 Pabla N Dong G Jiang M Huang S Kumar MV Messing RO Dong Z Inhibition of PKCδ reduces cisplatin-induced nephrotoxicity without blocking chemotherapeutic efficacy in mouse models of cancer J Clin Invest 2011 121 7 2709 2722 21633170 Padanilam BJ Induction and subcellular localization of protein kinase C isozymes following renal ischemia Kidney Int 2001 59 5 1789 1797 11318949 Pan JS Huang L Belousova T Lu L Yang Y Reddel R Chang A Ju H DiMattia G Tong Q Sheikh-Hamad D Stanniocalcin-1 inhibits renal ischemia/reperfusion injury via an AMP-activated protein kinase-dependent pathway J Am Soc Nephrol 2015 26 2 364 378 25012175 Park JS Pasupulati R Feldkamp T Roeser NF Weinberg JM Cyclophilin D and the mitochondrial permeability transition in kidney proximal tubules after hypoxic and ischemic injury Am J Physiol Renal Physiol 2011 301 1 F134 F150 21490135 Park KM Kim JI Ahn Y Bonventre AJ Bonventre JV Testosterone is responsible for enhanced susceptibility of males to ischemic renal injury J Biol Chem 2004 279 50 52282 52292 15358759 Patschan D Hildebrandt A Rinneburger J Wessels JT Patschan S Becker JU Henze E Krüger A Müller GA The hormone melatonin stimulates renoprotective effects of "early outgrowth" endothelial progenitor cells in acute ischemic kidney injury Am J Physiol Renal Physiol 2012 302 10 F1305 F1312 22357919 Patschan D Krupincza K Patschan S Zhang Z Hamby C Goligorsky MS Dynamics of mobilization and homing of endothelial progenitor cells after acute renal ischemia modulation by ischemic preconditioning Am J Physiol Renal Physiol 2006 291 1 F176 F185 16478972 Patschan D Patschan S Wessels JT Becker JU David S Henze E Goligorsky MS Müller GA Epac-1 activator 8-O-cAMP augments renoprotective effects of syngeneic [corrected] murine EPCs in acute ischemic kidney injury Am J Physiol Renal Physiol 2010 298 1 F78 F85 19906949 Patschan SA Patschan D Temme J Korsten P Wessels JT Koziolek M Henze E Müller GA Endothelial progenitor cells (EPC) in sepsis with acute renal dysfunction (ARD) Crit Care 2011 15 2 R94 21396100 Pellegrini KL Han T Bijol V Saikumar J Craciun FL Chen WW Fuscoe JC Vaidya VS MicroRNA-155 deficient mice experience heightened kidney toxicity when dosed with cisplatin Toxicol Sci 2014 141 2 484 492 25015656 Peng J Li X Zhang D Chen JK Su Y Smith SB Dong Z Hyperglycemia, p53, and mitochondrial pathway of apoptosis are involved in the susceptibility of diabetic models to ischemic acute kidney injury Kidney Int 2015 87 1 137 150 24963915 Peng ZY Zhou F Wang HZ Wen XY Nolin TD Bishop JV Kellum JA The anti-oxidant effects are not the main mechanism for glutamine's protective effects on acute kidney injury in mice Eur J Pharmacol 2013 705 1–3 11 19 23454558 Pisani A Sabbatini M Riccio E Rossano R Andreucci M Capasso C De Luca V Carginale V Effect of a recombinant manganese superoxide dismutase on prevention of contrast-induced acute kidney injury Clin Exp Nephrol 18 3 424 431 Plotnikov EY Chupyrkina AA Jankauskas SS Pevzner IB Silachev DN Skulachev VP Zorov DB Mechanisms of nephroprotective effect of mitochondria-targeted antioxidants under rhabdomyolysis and ischemia/reperfusion Biochim Biophys Acta 2011 1812 1 77 86 20884348 Prakash J de Borst MH Lacombe M Opdam F Poelstra K Kok RJ Inhibition of renal rho kinase attenuates ischemia/reperfusion-induced injury J Am Soc Nephrol 2008 19 11 2086 2097 18650485 Quiros Y Sánchez-González PD López-Hernández FJ Morales AI López-Novoa JM Cardiotrophin-1 administration prevents the renal toxicity of iodinated contrast media in rats Toxicol Sci 2013 132 2 493 501 23335628 Rajan D Wu R Shah KG Jacob A Coppa GF Wang P Human ghrelin protects animals from renal ischemia-reperfusion injury through the vagus nerve Surgery 2012 151 1 37 47 21943641 Ramesh G Reeves WB TNF-alpha mediates chemokine and cytokine expression and renal injury in cisplatin nephrotoxicity J Clin Invest 2002 110 6 835 842 12235115 Rasbach KA Schnellmann RG Isoflavones promote mitochondrial biogenesis J Pharmacol Exp Ther 2008 325 2 536 543 18267976 Rosner MH Acute kidney injury in the elderly Clin Geriatr Med 2013 29 3 565 578 23849008 Rostami B Nematbakhsh M Pezeshki Z Talebi A Sharifi MR Moslemi F Eshraghi-Jazi F Ashrafi F Effect of testosterone on Cisplatin-induced nephrotoxicity in surgically castrated rats Nephrourol Mon 2014 6 5 e21546 25695037 Sahu BD Kuncha M Sindhura GJ Sistla R Hesperidin attenuates cisplatin-induced acute renal injury by decreasing oxidative stress, inflammation and DNA damage Phytomedicine 2013 20 5 453 460 23353054 Salom MG Cerón SN Rodriguez F Lopez B Hernández I Fenoy FJ Heme oxygenase-1 induction improves ischemic renal failure role of nitric oxide and peroxynitrite Am J Physiol Heart Circ Physiol 2007 293 6 H3542 H3549 (Salom et al., 2007) 17890422 Samarasinghe DA Tapner M Farrell GC Role of oxidative stress in hypoxia-reoxygenation injury to cultured rat hepatic sinusoidal endothelial cells Hepatology 2000 31 1 160 165 10613741 Satake A Takaoka M Nishikawa M Yuba M Shibata Y Okumura K Kitano K Tsutsui H Fujii K Kobuchi S Ohkita M Matsumura Y Protective effect of 17beta-estradiol on ischemic acute renal failure through the PI3K/Akt/eNOS pathway Kidney Int 2008 73 3 308 317 18004295 Schneider MP Sullivan JC Wach PF Boesen EI Yamamoto T Fukai T Harrison DG Pollock DM Pollock JS Protective role of extracellular superoxide dismutase in renal ischemia/reperfusion injury Kidney Int 2010 78 4 374 381 20505656 Semenza GL Oxygen sensing, hypoxia-inducible factors, and disease pathophysiology Annu Rev Pathol 2014 9 47 71 23937437 Sener G Tuğtepe H Yüksel M Cetinel S Gedik N Yeğen BC Resveratrol improves ischemia/reperfusion-induced oxidative renal injury in rats Arch Med Res 2006 37 7 822 829 16971220 Shah KG Rajan D Jacob A Wu R Krishnasastry K Wang P Attenuation of renal ischemia and reperfusion injury by human adrenomedullin and its binding protein J Surg Res 2010 163 1 110 117 20538296 Sharfuddin AA Sandoval RM Berg DT McDougal GE Campos SB Phillips CL Jones BE Gupta A Grinnell BW Molitoris BA Soluble thrombomodulin protects ischemic kidneys J Am Soc Nephrol 2009 20 3 524 534 19176699 Sharples EJ Yaqoob MM Erythropoietin and acute renal failure Semin Nephrol 2006 26 4 325 331 16949472 Shimizu H Takahashi T Suzuki T Yamasaki A Fujiwara T Akagi R Protective effect of heme oxygenase induction in ischemic acute renal failure Crit Care Med 2000 28 3 809 817 10752834 Shoskes DA Effect of bioflavonoids quercetin and curcumin on ischemic renal injury a new class of renoprotective agents [J] Transplantation 1998 66 2 147 152 9701255 Shusterman N Strom BL Murray TG Morrison G West SL Maislin G Risk factors and outcome of hospital-acquired acute renal failure. Clinical epidemiologic study Am J Med 1987 83 1 65 71 Si J Ge Y Zhuang S Wang LJ Chen S Gong R Adrenocorticotropic hormone ameliorates acute kidney injury by steroidogenic-dependent and -independent mechanisms Kidney Int 2013 83 4 635 646 23325074 Simmons MN Subramanian V Crouzet S Haber GP Colombo JR Jr Ukimura O Nielsen S Gill IS Alpha-melanocyte stimulating hormone analogue AP214 protects against ischemia induced acute kidney injury in a porcine surgical model J Urol 2010 183 4 1625 1629 20172543 Singh D Chander V Chopra K Protective effect of catechin on ischemia-reperfusion-induced renal injury in rats Pharmacol Rep 2005 57 1 70 76 15849379 Singh D Chopra K The effect of naringin, a bioflavonoid on ischemia-reperfusion induced renal injury in rats Pharmacol Res 2004 50 2 187 193 15177308 Sogabe K Roeser NF Venkatachalam MA Weinberg JM Differential cytoprotection by glycine against oxidant damage to proximal tubule cells Kidney Int 1996 50 3 845 854 8872959 Sohotnik R Nativ O Abbasi A Awad H Frajewicki V Abassi Z Phosphodiesterase-5 inhibition attenuates early renal ischemia-reperfusion-induced acute kidney injury assessment by quantitative measurement of urinary NGAL and KIM-1 Am J Physiol Renal Physiol 2013 304 8 F1099 F1104 23364806 Soljancic A Ruiz AL Chandrashekar K Maranon R Liu R Reckelhoff JF Juncos LA Protective role of testosterone in ischemia-reperfusion-induced acute kidney injury Am J Physiol Regul Integr Comp Physiol 2013 304 11 R951 R958 23552495 Su HC Hung LM Chen JK Resveratrol, a red wine antioxidant, possesses an insulin-like effect in streptozotocin-induced diabetic rats Am J Physiol Endocrinol Metab 2006 290 6 E1339 E1346 16434553 Sutton TA Kelly KJ Mang HE Plotkin Z Sandoval RM Dagher PC Minocycline reduces renal microvascular leakage in a rat model of ischemic renal injury Am J Physiol Renal Physiol 2005 288 1 F91 F97 15353401 Suzuki S Maruyama S Sato W Morita Y Sato F Miki Y Kato S Katsuno M Sobue G Yuzawa Y Matsuo S Geranylgeranylacetone ameliorates ischemic acute renal failure via induction of Hsp70 Kidney Int 2005 67 6 2210 2220 15882264 Szakaly P Laszlo E Kovacs K Racz B Horvath G Ferencz A Lubics A Kiss P Mice deficient in pituitary adenylate cyclase activating polypeptide (PACAP) show increased susceptibility to in vivo renal ischemia/reperfusion injury Neuropeptides 2011 45 2 113 121 21211837 Szeto HH Liu S Soong Y Wu D Darrah SF Cheng FY Zhao Z Ganger M Tow CY Seshan SV Mitochondria-targeted peptide accelerates ATP recovery and reduces ischemic kidney injury J Am Soc Nephrol 2011 22 6 1041 1052 21546574 Tanaka R Tsutsui H Kobuchi S Sugiura T Yamagata M Ohkita M Takaoka M Yukimura T Matsumura Y Protective effect of 17β-estradiol on ischemic acute kidney injury through the renal sympathetic nervous system Eur J Pharmacol 2012 683 1–3 270 275 22426161 Tang C Dong Z Epigenetic regulation in acute kidney injury: new light in a dark area Kidney Int 2015 88 4 665 668 26422622 Tang J Liu N Tolbert E Ponnusamy M Ma L Gong R Bayliss G Yan H Zhuang S Sustained activation of EGFR triggers renal fibrogenesis after acute kidney injury Am J Pathol 2013 183 1 160 172 23684791 Tang J Yan Y Zhao TC Gong R Bayliss G Yan H Zhuang S Class I HDAC activity is required for renal protection and regeneration after acute kidney injury Am J Physiol Renal Physiol 2014 307 3 F303 F316 24808536 Tang J Zhuang S Epigenetics in acute kidney injury Curr Opin Nephrol Hypertens 2015 24 4 351 358 26050122 Tang J Liu N Zhuang S Role of epidermal growth factor receptor in acute and chronic kidney injury Kidney Int 2013 83 5 804 810 23325080 Toda N Takahashi T Mizobuchi S Fujii H Nakahira K Akagi R Tin chloride pretreatment prevents renal injury in rats with ischemic acute renal failure Crit Care Med 2002 30 7 1512 1522 12130972 Tögel F Zhang P Hu Z Westenfelder C VEGF is a mediator of the renoprotective effects of multipotent marrow stromal cells in acute kidney injury J Cell Mol Med 2009 13 8B 2109 2114 19397783 Tögel FE Westenfelder C Mesenchymal stem cells a new therapeutic tool for AKI Nat Rev Nephrol 2010 6 3 179 183 20186233 Trionfini P Benigni A Remuzzi G MicroRNAs in kidney physiology and disease Nat Rev Nephrol 2015 11 1 23 33 25385286 Tuğtepe H Sener G Biyikli NK Yüksel M Cetinel S Gedik N Yeğen BC The protective effect of oxytocin on renal ischemia/reperfusion injury in rats Regul Pept 140 3 101 108 Uchino S Bellomo R Goldsmith D Bates S Ronco C An assessment of the RIFLE criteria for acute renal failure in hospitalized patients Crit Care Med 2006 34 7 1913 1917 16715038 Valverde P Benedito ESolano F Oaknin S Lozano JA Garcia Borron JC Melatonin Antagonizes α-Melanocyte-Stimulating Hormone Enhancement of Melanogenesis in Mouse Melanoma Cells by Blocking the Hormone-Induced Accumulation of the C Locus Tyrosinase Eur J Biochem 1995 232 1 257 263 7556159 Venkatachalam MA Weinberg JM Kriz W Bidani AK Failed Tubule Recovery, AKI-CKD Transition, and Kidney Disease Progression J Am Soc Nephrol 2015 26 8 1765 1776 25810494 Venkatachalam MA Weinberg JM Patel Y Saikumar P Dong Z Cytoprotection of kidney epithelial cells by compounds that target amino acid gated chloride channels Kidney Int 1996 49 2 449 460 8821829 Wada Y Iyoda M Matsumoto K Shindo-Hirai Y Kuno Y Yamamoto Y Suzuki T Saito T Iseri K Shibata T Epidermal growth factor receptor inhibition with erlotinib partially prevents Cisplatin-induced nephrotoxicity in rats PLoS One 2014 9 e111728 25390346 Wang J Wei Q Wang CY Hill WD Hess DC Dong Z Minocycline up-regulates Bcl-2 and protects against cell death in mitochondria J Biol Chem 2004 279 19 19948 19954 15004018 Wang J Wei Q Wang CY Hill WD Hess DC Dong Z Minocycline up-regulates Bcl-2 and protects against cell death in mitochondria J Biol Chem 2004 279 19 19948 19954 15004018 Wang X Bonventre JV Parrish AR The aging kidney increased susceptibility to nephrotoxicity Int J Mol Sci 2014 15 9 15358 15376 25257519 Wang Y Pei DS Ji HX Xing SH Protective effect of a standardized Ginkgo extract (ginaton) on renal ischemia/reperfusion injury via suppressing the activation of JNK signal pathway Phytomedicine 2008 15 11 923 931 18929474 Wang Z Havasi A Gall J Bonegio R Li Z Mao H Schwartz JH Borkan SC GSK3beta promotes apoptosis after renal ischemic injury J Am Soc Nephrol 2010 21 2 284 294 20093356 Watanabe M de Moura Neiva LB da Costa Santos CX Martins Laurindo FR de Fátima Fernandes Vattimo M Isoflavone and the heme oxygenase system in ischemic acute kidney injury in rats Food Chem Toxicol 45 12 2366 2371 Watanabe T Tanaka M Watanabe K Takamatsu Y Tobe A Research and development of the free radical scavenger edaravone as a neuroprotectant [Article in Japanese] Yakugaku Zasshi 2004 124 3 99 111 15049127 Wei Q Bhatt K He HZ Mi QS Haase VH Dong Z Targeted deletion of Dicer from proximal tubules protects against renal ischemia-reperfusion injury J Am Soc Nephrol 2010 21 5 756 761 20360310 Wei Q Hill WD Su Y Huang S Dong Z Heme oxygenase-1 induction contributes to renoprotection by G-CSF during rhabdomyolysis-associated acute kidney injury Am J Physiol Renal Physiol 2011 301 1 F162 F170 21511696 Weidemann A Bernhardt WM Klanke B Daniel C Buchholz B Câmpean V Amann K Warnecke C Wiesener MS Eckardt KU Willam C HIF activation protects from acute kidney injury J Am Soc Nephrol 2008 19 3 486 494 18256363 Wen X Peng Z Li Y Wang H Bishop JV Chedwick LR Singbartl K Kellum JA One dose of cyclosporine A is protective at initiation of folic acid-induced acute kidney injury in mice Nephrol Dial Transplant 2012 27 8 3100 3109 22294776 Xiao T Guan X Nie L Wang S Sun L Zhao J Rapamycin promotes podocyte autophagy and ameliorates renal injury in diabetic mice Mol Cell Biochem 2014 394 1–2 145 154 24850187 Xiao X Hu Y Quirós PM Wei Q López-Otín C Dong Z OMA1 mediates OPA1 proteolysis and mitochondrial fragmentation in experimental models of ischemic kidney injury Am J Physiol Renal Physiol 306 11 F1318 F1326 Xu Y Ma H Shao J Wu J Zhou L Han J A Role for Tubular Necroptosis in Cisplatin-Induced AKI J Am Soc Nephrol 2015 26 11 2647 2658 25788533 Yamashita J Ogata M Takaoka M Matsumura Y KB-R7943, a selective Na+/Ca2+ exchange inhibitor, protects against ischemic acute renal failure in mice by inhibiting renal endothelin-1 overproduction J Cardiovasc Pharmacol 2001 37 3 271 279 11243417 Yang D Yang D Jia R Tan J Na+/Ca2+ exchange inhibitor, KB-R7943, attenuates contrast-induced acute kidney injury J Nephrol 2013 26 5 877 885 23475466 Yasuda H Kato A Miyaji T Zhou H Togawa A Hishida A Insulin-like growth factor-I increases p21 expression and attenuates cisplatin-induced acute renal injury in rats Clin Exp Nephrol 2004 8 1 27 35 15067513 Yasuda H Yuen PS Hu X Zhou H Star RA Simvastatin improves sepsis-induced mortality and acute kidney injury via renal vascular effects Kidney Int 2006 69 9 1535 1542 16557230 Yasuda K Ozaki T Saka Y Yamamoto T Gotoh M Ito Y Yuzawa Y Matsuo S Maruyama S Autologous cell therapy for cisplatin-induced acute kidney injury by using non-expanded adipose tissue-derived cells Cytotherapy 2012 14 9 1089 1100 22731757 Yen TH Alison MR Goodlad RA Otto WR Jeffery R Cook HT Wright NA Poulsom R Epidermal growth factor attenuates tubular necrosis following mercuric chloride damage by regeneration of indigenous, not bone marrow-derived cells J Cell Mol Med 2015 19 2 463 473 25389045 Yeung BHY Wong CKC Stanniocalcin-1 Regulates Re-Epithelialization in Human Keratinocytes PLoS ONE 2011 6 e27094 22069492 Yoon HY Kang NI Lee HK Jang KY Park JW Park BH Sulforaphane protects kidneys against ischemia-reperfusion injury through induction of the Nrf2-dependent phase 2 enzyme Biochem Pharmacol 2008 75 11 2214 2223 18407246 Yoshida T Kumagai H Kohsaka T Ikegaya N Relaxin prorects against renal ischemia reperfusion injury Am J Physiol Renal Physiol 2013 305 8 F1169 F1176 23946288 Yoshida T Kumagai H Kohsaka T Ikegaya N Protective effects of relaxin against cisplatin-induced nephrotoxicity in rats Nephron Exp Nephrol 2014 128 1–2 9 20 25403022 Zhou D Tan RJ Lin L Zhou L Liu Y Activation of hepatocyte growth factor receptor, c-met, in renal tubules is required for renoprotection after acute kidney injury Kidney Int 2013 84 3 509 520 23715119 Zhuang S Lu B Daubert RA Chavin KD Wang L Schnellmann RG Suramin promotes recovery from renal ischemia/reperfusion injury in mice Kidney Int 2009 75 3 304 311 18843260 Zhuang S Schnellmann RG Suramin promotes proliferation and scattering of renal epithelial cells J Pharmacol Exp Ther 2005 314 1 383 390 15833899 Zuk A Gershenovich M Ivanova Y MacFarland RT Fricker SP Ledbetter S CXCR4 antagonism as a therapeutic approach to prevent acute kidney injury Am J Physiol Renal Physiol 2014 307 7 F783 F797 25080523
PMC005xxxxxx/PMC5123838.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101120028 22411 Dev Cell Dev. Cell Developmental cell 1534-5807 1878-1551 27825441 5123838 10.1016/j.devcel.2016.10.003 NIHMS822847 Article Neurotrophin signaling is required for glucose-induced insulin secretion Houtz Jessica 1 Borden Philip 13 Ceasrine Alexis 1 Minichiello Liliana 2 Kuruvilla Rejji 1 1 Department of Biology, Johns Hopkins University, 3400 N. Charles St, 224 Mudd Hall, Baltimore, Maryland 21218, USA 2 Department of Pharmacology, University of Oxford, OX1 3QT, Oxford, UK Corresponding author, Rejji Kuruvilla, [email protected] 3 Current address: Janelia Research Campus, Howard Hughes Medical Institute, Ashburn, Virginia 20147, USA 14 10 2016 7 11 2016 07 11 2017 39 3 329345 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Summary Insulin secretion by pancreatic islet β-cells is critical for glucose homeostasis, and a blunted β-cell secretory response is an early deficit in type 2 diabetes. Here, we uncover a regulatory mechanism by which glucose recruits vascular-derived neurotrophins to control insulin secretion. Nerve Growth Factor (NGF), a classical trophic factor for nerve cells, is expressed in pancreatic vasculature while its TrkA receptor is localized to islet β-cells. High glucose rapidly enhances NGF secretion, and increases TrkA phosphorylation in mouse and human islets. Tissue-specific deletion of NGF, TrkA, or acute disruption of TrkA signaling impairs glucose tolerance and insulin secretion in mice. We show that internalized TrkA receptors promote insulin granule exocytosis via F-actin reorganization. Furthermore, NGF treatment augments glucose-induced insulin secretion in human islets. These findings reveal a non-neuronal role for neurotrophins, and identify a new regulatory pathway in insulin secretion that can be targeted to ameliorate β-cell dysfunction. eTOC Blurb Blunted insulin secretion by pancreatic β-cell contributes to type 2 diabetes. Houtz et al. reveal that glucose stimulates paracrine neurotrophin signaling between pancreatic vasculature and islet beta-cells to control insulin secretion. These findings identify a non-canonical role for neurotrophins, classical neuronal growth and survival factors, in insulin secretion and glucose homeostasis. Introduction Insulin-producing β-cells are a small population of endocrine cells that comprise less than 2% of the pancreas, yet their function is imperative to maintenance of blood glucose homeostasis. In response to elevated blood glucose, β-cells secrete the hormone insulin which triggers glucose uptake by peripheral tissues. Blunted insulin secretion by β-cells is one of the earliest features of type 2 diabetes, a disease affecting almost 300 million people worldwide (Guariguata et al., 2014), observed even in pre-diabetic individuals and thought to precede the onset of overt hyperglycemia (Weir and Bonner-Weir, 2004). Strikingly, it has been postulated that by the time a diagnosis of diabetes is made, patients have lost almost 80% of β-cell function (Defronzo, 2009). Although loss of β-cell secretory responses precede a reduction in β-cell mass in type 2 diabetes (Defronzo, 2009; Weir and Bonner-Weir, 2004), research has predominantly focused on mechanisms governing β-cell proliferation (Vetere et al., 2014; Wang et al., 2015). While glucose is the primary stimulus for insulin secretion, β-cell secretion is potentiated by extrinsic signals that include fatty acids, amino acids, peptide hormones, and neurotransmitters that play essential roles (Henquin et al., 2003; Prentki et al., 2013; Rorsman and Braun, 2013). Glucose-stimulated insulin secretion (GSIS) involves the key steps of glucose uptake in β-cells, mitochondrial metabolism to alter the ATP/ADP ratio, closure of ATP-sensitive K+-channels and subsequent β-cell plasma-membrane depolarization, opening of voltage-gated Ca2+-channels, and Ca2+-dependent exocytosis of insulin granules (MacDonald et al., 2005). Extrinsic signals influence GSIS either by producing metabolic intermediates within β-cells, or by generating signaling second messengers that impinge on β-cell electrical activity and/or insulin exocytosis (Prentki et al., 2013; Rorsman and Braun, 2013). Because nutrient, hormonal, and neural inputs are precisely regulated by glucose itself, together they ensure insulin secretion remains glucose-dependent at much lower plasma glucose concentrations than are effective in vitro (Henquin et al., 2003; Prentki et al., 2013). Thus, a small meal-related increase in plasma glucose elicits a larger insulin response than predicted from in vitro dose-response curves, due to effects of glucose on non-glucose stimuli. Conversely, the exquisite glucose-dependence of these non-glucose insulin secretagogues also protects against hypoglycemia (Henquin et al., 2003). Neurotrophins are soluble peptide factors known predominantly for their functions in neuronal survival, axon growth, and synaptic communication (Huang and Reichardt, 2001). Although neurotrophins and their Trk receptor tyrosine kinases are expressed in non-neuronal tissues including the pancreas (Tessarollo, 1998), little is known about their in vivo functions outside of the nervous system. Autocrine signaling by the classical neurotrophin, Nerve Growth Factor (NGF), has been implicated in β-cell survival and secretion in cell cultures (Navarro-Tableros et al., 2004; Rosenbaum et al., 2001). However, whether neurotrophin signaling is essential for β-cell function in vivo where intercellular communications between β-cells and neighboring cell types are preserved, has not been addressed. The need to understand the physiological relevance of neurotrophins in islet function is underscored by evidence that altered neurotrophin secretion and/or signaling could contribute to the etiology of diabetes (Bullo et al., 2007; Kim et al., 2009; Schreiber et al., 2005). In humans, mutations in the TrkA gene encoding for the NGF receptor, cause a form of hereditary peripheral neuropathy called congenital insensitivity to pain and anhidrosis (CIPA) (Indo et al., 1996). Children with CIPA show decreased insulin secretion in response to a glucose challenge, suggesting that NGF signaling may play a role in insulin responses in humans (Schreiber et al., 2005). Furthermore, altered circulating NGF levels have been noted in type 2 diabetes (Bullo et al., 2007; Kim et al., 2009), although whether this reflects a cause or effect in disease pathogenesis remains undefined. Here, we uncover a fundamental role for neurotrophin signaling in controlling glucose-stimulated insulin secretion, and elucidate the cellular underpinnings. We found that NGF is robustly expressed in pancreatic vascular contractile cells, whereas its TrkA receptor is localized to islet β-cells. Elevated glucose rapidly increases NGF secretion, and stimulates TrkA phosphorylation in islets. Vascular-specific NGF deletion, pancreas-specific TrkA deletion, or acute inactivation of TrkA impairs glucose tolerance and attenuates GSIS in mice. TrkA activity promotes insulin granule localization to the β-cell plasma membrane via disassembly of a rigid F-actin barrier. Furthermore, Trk-mediated endosomal signaling, a critical determinant of neurotrophin actions in neurons, is conserved in β-cells and functionally important for insulin secretion. Finally, NGF potentiates GSIS in human islets. These findings elucidate a new pathway by which glucose promotes NGF/TrkA-mediated actin reorganization to trigger insulin secretion in β-cells. Results Pancreatic TrkA receptors are essential for glucose homeostasis and insulin secretion We recently reported that NGF signaling indirectly influences islet architecture and functional maturation by recruiting sympathetic nerves to developing islets (Borden et al., 2013). However, both NGF and its TrkA receptors are reportedly expressed in rat β-cells (Kanaka-Gantenbein et al., 1995; Rosenbaum et al., 1998), suggesting a cell-autonomous requirement for NGF signaling. To identify an intrinsic role for NGF signaling in pancreatic cells, we first defined TrkA localization in the mouse pancreas. Since available TrkA antibodies were inadequate for immunohistochemical analyses in the pancreas, we employed a ligand binding assay where mouse pancreatic tissue sections were incubated with biotinylated NGF followed by detection with Alexa-546-conjugated streptavidin. Highest levels of biotinylated NGF binding were observed in islets and co-localized with insulin immunostaining (Figure 1A). TrkA expression in β-cells was also confirmed by immunoblotting of FAC-sorted β-cells from MIP-GFP transgenic mice (Figure S1A), which express a GFP reporter under the insulin promoter (Hara et al., 2003). In addition to TrkA, NGF also binds the p75 receptor. Immunohistochemistry with a p75 antibody revealed p75 expression in nerve fibers surrounding islets, but not in endocrine cells (Figure 1B). Ligand binding was not detected in nerve fibers innervating the pancreas (Figure 1A), likely due to low levels of TrkA in mature axons (Miller et al., 1994), and the lower affinity of NGF for p75 (Rodriguez-Tebar et al., 1990). Together, these results suggest that NGF signaling in islets is primarily transduced by β-cell-localized TrkA receptors. To address the intrinsic role of TrkA in endocrine cells, we mated floxed TrkA (TrkAf/f) mice (Chen et al., 2005) with Pdx1-Cre transgenic mice (Hingorani and Tuveson, 2003), where Pdx1 is a transcription factor expressed in pancreatic progenitor cells. Pdx1-Cre;TrkAf/f mice would be expected to have an early and pancreas-wide deletion of TrkA. Pdx1-Cre;TrkAf/f mice survived to adulthood, had no gross morphological abnormalities, and had normal body weight (Figure S1B). Ligand binding and TrkA immunoblotting demonstrated reduced TrkA protein levels in mutant islets (Figures 1C–E), indicative of efficient TrkA deletion. Immunostaining with islet hormone markers, insulin and glucagon, revealed normal islet formation and cyto-architecture in neonatal Pdx1-Cre;TrkAf/f mice (Figure S1C). However, mutant islets showed a modest decrease in size by one month of age (Figures S1D,E), and this may be due, in part, to the trophic effects of NGF signaling in β-cells (Navarro-Tableros et al., 2004). When we assessed metabolic parameters at the whole animal level, 1.5–2 month-old Pdx1-Cre;TrkAf/f mice were slightly hyperglycemic when fed ad libitum, although fasted blood glucose levels were normal (Figures 1F,G). However, Pdx1-Cre;TrkAf/f mice showed significant glucose intolerance and decreased circulating insulin levels after glucose administration, compared to control TrkAf/f mice (Figures 1H,I). Although Cre recombinase activity has been reported in extra-pancreatic tissues, including the hypothalamus, in Pdx1-Cre transgenic mice (Song et al., 2010), there is little overlap with TrkA expression (Fagan et al., 1997). Additionally, the (Tg(Pdx1-CreTuv) transgenic mice that we employed have not been reported to carry a human growth hormone mini-gene, commonly found in several Cre lines, that elicits metabolic defects (Brouwers et al., 2014). Together, these results suggest that loss of TrkA in pancreatic tissues causes the defects in glucose homeostasis and insulin secretion in Pdx1-Cre;TrkAf/f mice. TrkA kinase activity is acutely required for GSIS In Pdx1-Cre;TrkAf/f mice, deletion of TrkA is initiated at early embryonic stages. Thus, the observations of impaired glucose tolerance and reduced insulin secretion in these mice could stem from developmental anomalies and/or acute deficits in insulin secretion. To distinguish between developmental versus acute effects of TrkA signaling, we employed a chemical-genetic approach to inducibly silence TrkA kinase activity in mature mice at 2 months of age. TrkAF592A knock-in mice express receptors with a mutated ATP binding pocket that can be selectively, rapidly and reversibly inhibited by a small molecule membrane-permeable inhibitor, 1NMPP1 (Chen et al., 2005) (Figure 2A). Although TrkAF592A mice have been reported to be hypomorphic (https://www.jax.org/strain/022362), these mice are viable and fertile, and TrkA signaling is effectively blocked by application of 1NMPP1 (Chen et al., 2005). In TrkAF592A mice, we always compared effects of 1NMPP1 administration to vehicle (DMSO) treatment in litter-mates. 1NMPP1 treatment attenuated TrkA phosphorylation in isolated TrkAF592A islets, but had no effect in wild-type islets (Figures 2B,C). Importantly, acutely disrupting TrkA activity in vivo by injecting 1NMPP1 (20 ng/g body weight, intra-peritoneally), 20 minutes prior to a glucose challenge, elicited glucose intolerance in adult TrkAF592A mice (Figure 2D), similar to Pdx1-Cre;TrkAf/f mice. The area under the curve (AUC) values of plasma glucose was significantly higher with 1NMPP1 treatment compared to vehicle (DMSO) injection (Figure 2E). Furthermore, TrkAF592A mice treated with 1NMPP1 showed reduced insulin secretion in the first phase of a glucose challenge, as well as dampened insulin levels in the sustained second phase (Figure 2F). Consistently, 1NMPP1 treatment also significantly decreased GSIS in TrkAF592A islets in static insulin secretion assays (Figure 2G). There were no differences in total islet insulin content and insulin sensitivity between 1NMPP1- and vehicle-injections in TrkAF592A mice (Figures 2H,I). Thus, glucose intolerance in 1NMPP1-treated mice likely does not arise from defects in insulin biosynthesis and insulin responsiveness. Notably, 1NMPP1 treatment did not alter glucose tolerance or insulin secretion in wild-type animals (Figures S2A–C), highlighting specific effects of 1NMPP1 in the context of the TrkAF592A mutation. Together, these results indicate that TrkA signaling acutely regulates glucose homeostasis and insulin secretion, in a manner independent of developmental effects. TrkA signaling promotes insulin granule exocytosis via F-actin remodeling Regulatory control of insulin secretion can occur at the level of glucose uptake and metabolism in β-cells, β-cell membrane depolarization, and insulin granule mobilization and exocytosis (MacDonald et al., 2005). A key question is, which step in this stimulus-secretion coupling pathway is affected by TrkA signaling? Exposure to high extracellular K+ is a well- established strategy to depolarize the β-cell plasma membrane, thus by-passing the need for glucose uptake and metabolism to trigger insulin secretion (Hatlapatka et al., 2009). Elevated potassium chloride (KCl) results in the opening of voltage-dependent Ca2+ channels to promote the exocytosis of secretion-ready insulin granules (Hatlapatka et al., 2009). To determine if the requirement for TrkA signaling in insulin secretion was downstream of glucose sensing and metabolism, we assessed the effects of 1NMPP1 on KCl-induced insulin secretion. 1NMPP1-mediated inhibition of TrkA activity abolished insulin secretion induced by high KCl (30mM) in TrkAF592A islets (Figure 3A). Thus, TrkA signaling is required in the step(s) of the insulin secretory pathway that is distal to membrane depolarization. To further probe the requirement for TrkA activity in insulin secretion, we employed ultra-structural analyses. Electron microscopy revealed an increase in insulin granule localization at the β-cell plasma membrane in response to an in vivo glucose challenge compared to the fasted state in TrkAF592A mice (Figures 3B,C, and F). However, 1NMPP1 injection, 20 minutes prior to glucose administration, suppressed glucose-induced recruitment of insulin granules to the cell periphery, but had no effect on insulin docking under basal conditions (Figures 3D,E, and F). 1NMPP1-treated β-cells had fewer docked insulin granules (quantified as granules within 50 nm of the plasma membrane) in the presence of high glucose. Since islet insulin content was unaffected by 1NMPP1 treatment (see Figure 2H), the reduction in surface-localized insulin granules is not due to defects in insulin biogenesis. Together, these results suggest that TrkA activity is necessary for glucose-dependent insulin granule positioning at the β-cell surface. Since a brief (20 minute) exposure to 1NMPP1 was sufficient to elicit glucose intolerance and attenuate GSIS in TrkAF592A mice, we reasoned that TrkA signaling likely influences insulin secretion by mechanisms independent of gene transcription. Actin rearrangement is a critical acute determinant of GSIS (Kalwat and Thurmond, 2013). Actin microfilaments are organized as a dense meshwork beneath the β-cell plasma membrane that restricts insulin granule access to the docking and fusion machinery (Orci et al., 1972). Glucose stimulation rapidly promotes filamentous actin (F-actin) remodeling to mobilize insulin granules to the cell periphery (Kalwat and Thurmond, 2013; Thurmond et al., 2003). However, how glucose stimulation regulates actin reorganization has remained unclear. Remodeling of the actin cytoskeleton is also a key cellular process by which neurotrophins control vesicular trafficking in neurons (Harrington and Ginty, 2013). Thus, we asked whether TrkA signaling promotes actin rearrangements in β-cells, a pre-requisite step in insulin granule positioning at the plasma membrane. Isolated TrkAF592A β-cells were pre-treated with 1NMPP1 for 20 minutes prior to stimulation with either low (2.8 mM) or high glucose (16.7 mM), and F- actin was visualized using Alexa-546-labeled phalloidin. Under basal conditions, both vehicle and 1NMPP1-treated β-cells showed a thick F-actin cortical ring (Figures 3G,I). High glucose treatment elicited a striking reduction in F-actin (Figures 3H,K), consistent with previous reports (Cai et al., 2012; Nevins and Thurmond, 2003). However, the glucose-induced dissolution of F- actin was prevented by 1NMPP1-mediated silencing of TrkA signaling (Figures 3J,K). Thus, TrkA activity is required for glucose-triggered actin reorganization in β-cells. Glucose-mediated actin remodeling in β-cells depends on the activities of several actin-modulatory proteins including the Rac1 GTPase (Kalwat and Thurmond, 2013). In clonal β-cell lines and in islets, elevated glucose stimulates Rac1 activity (Kowluru, 2011). β-cell-specific deletion of Rac1 inhibits F-actin disassembly and impairs glucose tolerance and GSIS in mice (Asahara et al., 2013). Rac1-deficient β-cells also show impaired glucose-dependent recruitment of insulin granules to the β-cell membrane (Asahara et al., 2013), similar to that seen with TrkA inhibition. However, how elevated glucose activates Rac1 remains unclear. Using an ELISA-based immunoassay, we observed that high glucose treatment for 15 minutes significantly increased GTP-bound Rac1 levels (1.8 ± 0.15-fold increase) in TrkAF592A islets, which was suppressed by TrkA inhibition with 1NMPP1 (Figure 3L). These results suggest that TrkA kinase activity is required for glucose-mediated activation of Rac1. We next reasoned that if indeed a perduring F-actin barrier was a key contributor to decreased GSIS upon TrkA inhibition, then forcing F-actin disassembly should alleviate impaired insulin secretion in 1NMPP1-treated islets. To test this prediction, we assessed insulin secretion in islets that were co-treated with 1NMPP1 and cytochalasin D, a cell-permeable fungal toxin that disrupts actin polymerization. As expected, 1NMPP1 treatment alone attenuated GSIS compared to vehicle (DMSO) in TrkAF592A islets (Figure 3M). Remarkably, normal insulin secretion in response to elevated glucose was observed in TrkAF592A islets treated with 1NMPP1 plus cytochalasin D (Figure 3M). Together, these results indicate that TrkA-mediated actin reorganization is a key mechanism contributing to glucose-induced insulin secretion. We also noted that cytochalasin D treatment alone had a striking potentiating effect on GSIS, in agreement with previous studies (Lacy et al., 1973; van Obberghen et al., 1973), which was significantly higher than that observed in cytochalasin D plus 1NMPP1-treated islets (Figure 3M). That 1NMPP1 treatment dampened GSIS in the presence of cytochalasin D suggests that TrkA inactivation might diminish the efficacy of cytochalasin D, and/or that TrkA signaling utilizes additional mechanisms that are independent of the actin cytoskeleton to influence insulin granule exocytosis. Endocytosed TrkA receptors mediate actin reorganization and insulin secretion In neurons, endosomal signaling from intracellular TrkA receptors is a key determinant of actin remodeling (Harrington et al., 2011). NGF promotes endocytosis of its TrkA receptors in nerve endings into signaling endosomes that are retrogradely transported to neuronal cell bodies to activate trophic signaling (Ascano et al., 2012). Endosomal TrkA signaling overcomes a dense peripheral actin network in axons to “carve a path” for retrogradely moving vesicles (Harrington et al., 2011). These findings raise the possibility of an analogous mechanism in β-cells where neurotrophin signaling endosomes might mediate changes in F-actin necessary for insulin granule mobilization. To address this possibility, we first asked whether TrkA receptors undergo ligand-dependent endocytosis in β-cells. We probed TrkA endocytosis using a cell surface biotinylation assay in MIN6 cells, a mouse insulinoma cell-line that exhibits many characteristics of primary β-cells, including GSIS, and have been widely used for biochemical analyses (Ishihara et al., 1993). NGF stimulation for 30 minutes markedly increased TrkA endocytosis compared to un-stimulated cells (Figures 4A,B). Furthermore, we visualized trafficking of surface TrkA receptors in primary β-cells using a well-established antibody-feeding assay (Ascano et al., 2009). Primary β-cells were infected with an adenoviral vector expressing FLAG-tagged chimeric receptors that have the extracellular domain of TrkB and the transmembrane and intracellular domains of TrkA (FLAG-TrkB:A). Chimeric Trk receptors respond to the TrkB ligand, Brain-Derived Neurotrophic Factor (BDNF), but retain the signaling properties of TrkA. Live-cell FLAG antibody feeding revealed predominantly surface localization of the receptors in the absence of ligand (Figure 4C). However, in BDNF-stimulated β-cells, Trk receptors accumulated in intracellular punctae that co-localized with Early Endosome Antigen 1 (EEA1), an early endosome marker (Figures 4D, E). Thus, ligand-mediated endocytosis of Trk receptors is a conserved mechanism that occurs in both β-cells and neurons. In neurons, TrkA signaling triggers its own internalization by recruiting and activating the downstream effector, Phospholipase C gamma (PLCγ) (Bodmer et al., 2011). Employing cell surface biotinylation to monitor internalization, we found that a selective PLCγ inhibitor, U73122, markedly decreased endocytosis of endogenous TrkA receptors in MIN6 cells (Figures 4F,G). Phosphorylation of TrkA receptors at a specific tyrosine residue, Y794, is necessary for PLCγ recruitment (Stephens et al., 1994). To assess the requirement of Y794 phosphorylation for endocytosis, we performed antibody-feeding assays in MIN6 cells transfected with mutant FLAG-TrkAY794F or control FLAG-TrkA receptors. Upon NGF treatment, FLAG-TrkA receptors were found in intracellular punctae, whereas mutant FLAG-TrkAY794F receptors remained largely at the cell surface (Figures 4H, 4I, 4J, and 4L). Furthermore, FLAG-TrkAY499F receptors, mutated at a site necessary for recruitment of the adaptor protein, Shc, and coupling to downstream MAP kinase and phosphatidylinositol 3-kinase (PI-3K) effector pathways, were internalized normally in response to ligand (Figures 4K and 4L). Thus, TrkA phosphorylation at Y794 and activation of PLCγ are required for endocytosis in β-cells, as in neurons. To investigate the effects of endocytosis-deficient Trk receptors on actin remodeling and insulin secretion in primary β-cells and islets, we generated adenoviral vectors expressing mutant FLAG-TrkAY794F or control FLAG-TrkA receptors, that were also doxycycline-inducible to precisely control expression. FLAG-tagged receptors were expressed in a doxycycline-dependent manner and also appeared normally on the cell surface in infected MIN6 cells (Figure S3A). To assess the effects of ectopic TrkA receptors on glucose-induced actin remodeling, isolated β-cells from TrkAF592A mice were infected with FLAG-TrkA adenoviruses, treated with doxycycline (100 ng/ml) for 24–30 hr, and then treated with 1NMPP1 for 20 minutes to silence endogenous TrkA activity. The FLAG-TrkA receptors are impervious to 1NMPP1 since they do not harbor the F592A mutation. β-cells were then exposed to low (2.8 mM) or high (16.7 mM) glucose, and F-actin visualized with phalloidin labeling. We found that ectopic expression of control FLAG-TrkA receptors promoted glucose-dependent actin reorganization in 1NMPP1-treated cells (Figures 4M, 4N and 4Q). In contrast, mutant FLAG-TrkAY794F receptors were unable to rescue the 1NMPP1-mediated impairment in actin reorganization (Figures 4O–Q). In insulin secretion assays, control FLAG-TrkA receptors elicited robust insulin secretion in response to glucose, despite the presence of 1NMPP1 (Figure 4R). However, GSIS was suppressed in islets expressing mutant FLAG-TrkAY794F receptors (Figure 4R). Furthermore, PLCγ inhibition also decreased GSIS in islets, similar to the effects of the non-internalizable TrkA receptors (Figure S3B). These results suggest that TrkA endocytosis, via PLCγ activity, is required for glucose-stimulated actin remodeling and insulin secretion. NGF expression in the vasculature, but not pancreatic anlage, is necessary for glucose homeostasis, GSIS, and actin remodeling Our findings that TrkA receptors in β-cells are required for GSIS prompted us to address the cellular source of NGF in the pancreas. Previous studies have reported NGF expression in β-cells (Rosenbaum et al., 1998), and antibody-mediated neutralization of endogenous NGF attenuated GSIS in dissociated β-cell cultures (Rosenbaum et al., 2001). To address whether β-cell-derived NGF is essential for glucose homeostasis and GSIS in vivo, we crossed mice carrying a floxed NGF allele (NGFf/f mice) (Muller et al., 2012) with Pdx1-Cre mice. Surprisingly, conditional loss of NGF from all pancreatic cell types including β-cells did not impair glucose tolerance (Figure 5A), and isolated Pdx1-Cre;NGFf/f islets showed normal insulin secretion in response to a glucose challenge (Figure 5B). These results suggest that autocrine NGF signaling in β-cells is not a major contributing factor to glucose homeostasis at the whole animal level and for GSIS in intact islets. To obtain a thorough profile of NGF expression in the pancreas, we next used NGFLacZ/+ knock-in mice where the LacZ transgene reports NGF expression (Liu et al., 2012). Using X-gal staining, we observed robust LacZ expression in large diameter blood vessels outside islets, and in intra-islet micro-vasculature (Figure 5C). To further define vascular cell types that express NGF, we performed double immunofluorescence for β-galactosidase (β-gal) and vascular markers. We observed β-gal immunofluorescence in vascular smooth muscle cells (VSMCs) labeled by smooth muscle actin (SMA) in large arteries outside islets (Figure 5D), and in smooth muscle cell-like pericytes labeled by PDGFRβ within the islet micro-vasculature (Figure 5E). Reporter expression was not detected in pancreatic endothelial cells, ducts, or exocrine tissue (Figures S4A,B). We did not detect LacZ expression in β-cells in tissue sections, in agreement with the lack of metabolic phenotypes in Pdx1-Cre;NGFf/f mice. However, low LacZ expression was seen in β-cells in dissociated islet cultures (Figures S4C,D), consistent with previous findings (Rosenbaum et al., 1998). In dissociated islet cultures, prominent NGF expression was observed in intra-islet pericytes that are also present (Figures S4E,E’). Together, these results suggest vascular contractile cells as a likely source of NGF in the pancreas. To address the role of vascular-derived NGF in glucose homeostasis and insulin secretion, we generated mutant mice to inducibly delete NGF from vascular contractile cells. We crossed NGFf/f mice to Myh11-CreERT2 transgenic mice, in which tamoxifen-induced Cre expression is specific to vascular smooth muscle cells including pericytes (Heinze et al., 2014; Wirth et al., 2008). To visualize tamoxifen-induced Cre activity in NGF-expressing cell types, we generated Myh11-CreERT2;R26-EYFP mice which were then mated to NGFLacZ/+ mice. Cre reporter fluorescence was observed in NGF-expressing large-diameter blood vessels outside islets and in intra-islet pericytes but not β-cells, upon tamoxifen injection in NGFLacZ/+;Myh11-CreERT2;R26-YFP mice (Figure S4F). qPCR analysis showed significant NGF depletion in islets isolated from tamoxifen-injected-Myh11-CreERT2;NGFf/f mice compared to vehicle treated controls (Figure S4G). When we evaluated metabolic parameters, tamoxifen-injected Myh11-CreERT2;NGFf/f mice were found to be glucose intolerant two weeks after induction of Cre activity (Figure 5F). Furthermore, isolated islets from tamoxifen-injected mice had attenuated GSIS ex vivo, compared to vehicle treatment (Figure 5G). These results indicate that glucose homeostasis and GSIS relies on NGF derived from pancreatic vascular contractile cells. We next asked whether vascular-specific inactivation of NGF influences glucose-induced cellular events in β-cells. Thus, we examined actin rearrangements in response to high glucose in dissociated islet cultures from vehicle- and tamoxifen-injected Myh11-CreERT2;NGFf/f mice. Tamoxifen injection would ablate NGF from intra-islet pericytes that are present in dissociated islet cultures, but not from β-cells. Visualization of F-actin with Alexa-546-labeled phalloidin revealed that while high glucose elicited cortical actin dissolution in β-cells from vehicle-injected mice (Figures 5H, 5I, and 5L), this effect of glucose was abrogated in β-cells from tamoxifen-injected mice (Figures 5J, 5K, and 5L), similar to our observations with TrkA inactivation. Since islet pericytes are the only cell types in the dissociated cultures that would be targeted by tamoxifen-induced Cre activity, these results indicate that NGF produced by intra-islet pericytes is necessary for glucose-induced actin remodeling in β-cells. Thus, although cultured β-cells express low levels of NGF, this expression is insufficient to mediate glucose-dependent cellular changes in the absence of vascular-derived NGF. Elevated glucose acutely enhances NGF secretion and TrkA phosphorylation Alterations in NGF levels have been noted in serum, nerves, and peripheral tissues in diabetic humans and animal models (Bullo et al., 2007; Kim et al., 2009; Meloni et al., 2012). Additionally, elevated glucose enhances NGF biosynthesis and secretion in cultured β-cells and islets (Pingitore et al., 2016; Rosenbaum et al., 1998). These results, together with the observed expression patterns of NGF and TrkA in the pancreas, prompted us to ask if the NGF signaling pathway might be regulated by glucose in vivo. To address this question, we measured serum NGF levels in mice that were fasted overnight and subjected to a glucose challenge administered intra-peritoneally. We observed a significant elevation (1.5-fold increase) in circulating NGF levels within 15 minutes of the glucose challenge (Figure 6A), suggesting that NGF secretion is acutely regulated by glucose in vivo. We were unable to directly profile local release of NGF from islets in response to glucose, because the levels were below the detection limit of our immunoassay, perhaps in part, due to the entrapment of secreted NGF in the extracellular matrix and/or because NGF-expressing pericytes comprise a small fraction of the islet cell population. However, in dissociated aortic tissue cultures, an abundant source of vascular smooth muscle cells, we found that NGF secretion was significantly enhanced (1.48-fold increase) by 15 minutes of exposure to elevated glucose (Figure 6B), similar to the increase in serum NGF observed with an in vivo glucose challenge. Together, these results highlight a direct and acute effect of glucose on vascular smooth muscle cells, the primary NGF-expressing cell types in the pancreas, to elicit NGF release. Given that elevated glucose acutely enhanced NGF secretion, we next asked if glucose influences TrkA activity in islets. Thus, we assessed TrkA phosphorylation in isolated islets treated with high glucose (16.7mM). 15 minutes of exposure to elevated glucose stimulated a robust increase (11.69 ± 1.9-fold) in islet TrkA phosphorylation levels, compared to basal conditions (2.8 mM glucose) (Figures 6C, D). NGF binds with high affinity to TrkA receptors with a reported Kd of 10−11M (i.e. 130 pg/ml) (Bibel and Barde, 2000; Sutter et al., 1979). Given that our measurements of circulating NGF was ~200 pg/ml, it is reasonable that modest increases in NGF levels with glucose stimulation would elicit large alterations in TrkA activity. Together, these findings indicate that NGF secretion and signaling in pancreatic islets are regulated by glucose, and suggest an instructive role for the NGF pathway in facilitating GSIS. We next addressed whether glucose-mediated regulation of the NGF pathway occurs in human islets. Similar to our findings in mouse islets, we found that elevated glucose significantly increased TrkA phosphorylation in human islets (Figure 6E, F), although the fold change (1.74 ± 0.27-fold) was modest compared to that in mouse islets. Although human and mouse NGF have similar binding affinities for TrkA (Altar et al., 1991), that human islets had higher basal TrkA phosphorylation compared to mouse islets could reflect differences in basal NGF levels, tyrosine phosphatase activity, age of islets, or manner of islet isolation and culture. Nevertheless, these findings support the notion that the ability of elevated glucose to activate NGF signaling in islets is conserved. NGF enhances glucose-dependent insulin secretion in cultured rodent β-cells and islets (Pingitore et al., 2016; Rosenbaum et al., 2001). Whether NGF has a similar effect on human islets has not been addressed. We found that NGF treatment augmented GSIS in human islets, but basal insulin secretion was unaffected (Figure 6G). These results support a physiological role for NGF in potentiating GSIS, although NGF is insufficient by itself to trigger insulin secretion. Discussion Here, we reveal a mechanism by which neurotrophin signaling influences endocrine functions. Our findings suggest a feed-forward model whereby NGF, secreted by the pancreatic vasculature in response to glucose, activates β-cell TrkA receptors to acutely promote glucose-stimulated insulin secretion in β-cells (Figure 7). TrkA signaling, specifically by internalized receptors, overcomes a peripheral F-actin barrier to boost insulin granule exocytosis in β-cells. Together, these findings identify a new regulatory pathway essential for insulin secretion and blood glucose homeostasis. The spatial pattern of NGF and TrkA expression in the pancreas, and effects of vascular- specific NGF deletion on GSIS support a paracrine effect of vascular-derived NGF on neighboring β-cells. The NGF expression pattern also implies that signals derived from large diameter blood vessels juxtaposed to islets and/or intra-islet pericytes should be sufficient to support adult islet function. Consistent with this notion, regression of intra-islet endothelial cells due to loss of VEGF signaling did not perturb adult insulin secretion, and elicited only modest elevations in blood glucose (D'Hoker et al., 2013; Reinert et al., 2013). In particular, impaired GSIS in isolated islets ex vivo from tamoxifen-injected Myh11-CreERT2;NGFf/f mice highlights an essential role for NGF produced by intra-islet pericytes in β-cell function. Recently, diphtheria toxin-mediated killing of islet pericytes was found to impair GSIS in mice (Sasson et al., 2016), supporting the significance of these cell types within the islet micro-environment. Although β-cells express NGF, we show that loss of β-cell-derived NGF does not compromise glucose homeostasis in vivo. Notably, NGF expressed in β-cells cannot compensate for loss of vascular-derived NGF in GSIS and actin remodeling. The reasons for differences between our findings and previous reports of autocrine NGF signaling in insulin secretion (Pingitore et al., 2016; Rosenbaum et al., 2001) remain unclear, although the possibility remains that dissociated islet cultures employed in previous studies may have included NGF-expressing pericytes, as we noted in this study. Together, our results support the notion that, while autocrine NGF signaling may be relevant for β-cell development or survival (Navarro-Tableros et al., 2004), paracrine NGF signaling is predominantly required for GSIS and glucose homeostasis. We identify TrkA signaling as a key mechanism by which glucose stimulation overcomes a dense F-actin meshwork to mobilize insulin granules to the β-cell membrane for efficient secretion. The disparate roles of actin in regulated exocytosis have been well-documented with evidence to support both negative and positive effects. On the one hand, F-actin tracks are needed for mobilizing vesicles from deeper reserve pools to the plasma membrane (Schuh, 2011). On the other hand, cortical F-actin acts as a physical barrier to limit exocytosis in diverse cell types (Porat-Shliom et al., 2013). In β-cells, the barrier function of the cytoskeleton has been proposed to be critical for maintaining low levels of insulin release under basal conditions (Kalwat and Thurmond, 2013). Importantly, a dense cytoskeletal meshwork at the cell periphery presents a layer of regulation for controlled insulin release under elevated glucose (Zhu et al., 2015). Each individual β-cell contains approximately 10,000 secretory granules of insulin but only a fraction (several hundreds) are released at a time in response to high glucose, emphasizing the precise regulation of release probability of insulin granules (Rorsman and Renstrom, 2003). The relevance of actin dynamics in glucose homeostasis is exemplified by genetic studies in mice where deletion of Rac1, or its effector, p21-activated kinase (PAK), elicits glucose intolerance and diminished insulin secretion (Asahara et al., 2013; Wang et al., 2011). Notably, a 10-fold increase in total cellular actin has been observed in islets from type 2 diabetes patients (Ostenson et al., 2006). Although a number of actin regulators of glucose-stimulated insulin secretion including Rac1, Cdc42, PAK, Focal Adhesion Kinase (FAK), cofilin, and gelsolin, have been identified in β-cells (Kalwat and Thurmond, 2013; Thurmond et al., 2003), the events upstream of these actin-modulatory proteins have remained unclear. Our results, together with a previous study in the MIN6 cell line implicating EphA receptors in actin dynamics and insulin secretion (Konstantinova et al., 2007), highlight receptor tyrosine kinase signaling as a critical link in relaying a glucose signal to the β-cell cytoskeleton. Although neurons and β-cells have distinct developmental origins (Pictet et al., 1976), they share remarkable similarities in terms of electrical properties, ion channel composition, and exocytosis machinery involved in regulated secretion (Arntfield and van der Kooy, 2011). In this study, we describe a non-canonical role for neurotrophins outside of the nervous system, and also define a conserved mechanism for neurotrophin actions via signaling endosomes in non-polarized β-cells. TrkA receptors, that were incapable of internalizing, failed to stimulate actin reorganization and insulin secretion in response to glucose. In neurons, TrkA internalization in axon terminals and subsequent endosomal signaling is obligatory for long-distance communication between distal axons, the site of action of target-derived NGF, and neuronal cell bodies, as recently reviewed in (Cosker and Segal, 2014). In β-cells, TrkA endocytosis may allow access to intracellular signaling effectors including actin modulatory proteins, prolong receptor activation, or allow localized responses at specific sub-cellular domains (Sorkin and von Zastrow, 2009). Interestingly, β-cell specific deletion of dynamin 2 resulted in a striking increase in F-actin density and impaired insulin secretion and glucose tolerance (Fan et al., 2015), similar to our observations with TrkA inactivation, raising the possibility that TrkA receptors may be a cargo for dynamin 2-mediated endocytosis in β-cells. Together, our findings suggest TrkA-harboring endosomes as a unique locus for regulatory control of insulin granule mobilization in β-cells. Intriguingly, we observed that although cytochalasin D normalized GSIS in 1NMPP1-treated islets, the secretory response in these islets was still blunted compared to the effects of cytochalasin D alone. These findings suggest that TrkA activity might be necessary to render the actin network more susceptible to cytochalasin D-mediated actin disassembly, perhaps by affecting actin-modulatory proteins that influence the ability of cytochalasin D to cap barbed ends or bind actin monomers (Cooper, 1987). Additionally, these observations could suggest contributions of actin-independent TrkA signaling mechanisms to GSIS. Two candidate TrkA pathways that might influence GSIS are the Ca2+ and cAMP second messenger systems. NGF- dependent insulin secretion in cultured rat β-cells was shown to be dependent on Ca2+ influx through voltage-gated Ca2+ channels (Rosenbaum et al., 2001). Furthermore, cAMP signaling via Epac2, a guanine nucleotide exchange factor and its target small GTPase, Rap1, is a key contributing mechanism to insulin exocytosis (Shibasaki et al., 2007). In neurons, Rap1 is localized to TrkA-containing endosomes (Wu et al., 2001), making it an attractive candidate endosomal effector of TrkA that might elicit insulin secretion. Monitoring TrkA-mediated Ca2+ and cAMP activity in β-cells and assessing the consequences of manipulating pathway effectors on insulin secretion and glucose homeostasis will be of interest in future studies. Our study revealed an unexpected physiological cross-talk between nutrient and neurotrophin signaling. Elevated glucose rapidly increased NGF secretion in vivo, and enhanced TrkA phosphorylation in both mouse and human islets. These results suggest an instructive mechanism by which glucose recruits the NGF signaling axis to augment insulin secretion. This is similar to the effects of glucose on intestinal hormones and neurotransmitters that subsequently act in concert to tightly regulate a post-prandial insulin secretory response. Future studies will be of interest in elucidating the glucose sensing and signaling mechanisms that underlie glucose-mediated neurotrophin secretion in pancreatic vascular cells. Previous studies on neurotrophin regulation of metabolism have primarily focused on central hypothalamic circuits that control appetite and energy balance (Xu and Xie, 2016). Despite expression of neurotrophins and their receptors in peripheral metabolic tissues, little is known about peripheral mechanisms by which neurotrophins influence metabolism. Recently, p75 receptors in adipocytes were found to regulate energy expenditure and obesity, although this effect was independent of the p75 extracellular domain and neurotrophin binding (Baeza-Raja et al., 2016). Here, we report a direct role for NGF/TrkA signaling within pancreatic endocrine cells in controlling insulin secretion and blood glucose homeostasis. These findings are of clinical relevance given that TrkA mutations have been linked to impaired glucose-stimulated insulin secretion in humans (Schreiber et al., 2005). Our results that NGF treatment potentiated GSIS in human islets suggest the potential utility of neurotrophins and small molecule receptor agonists in the treatment of type 2 diabetes. Experimental Procedures Details of immunohistochemistry, EM, receptor trafficking analyses, plasmids, adenoviral vectors, drug treatments, antibodies, insulin and NGF ELISAs, metabolic analyses and insulin secretion can be found in Supplemental Experimental Procedures. Mice and cell lines Procedures relating to animal care and treatment conformed to Johns Hopkins University Animal Care and Use Committee (ACUC) and NIH guidelines. Mice were housed in a standard 12:12 light-dark cycle. Mice were maintained on a C57BL/6 background, or mixed C57BL/6 and 129P, or C57BL/6 and FVB backgrounds. Both sexes were used for analyses at 1–2 months of age. TrkAF592A (TrkAf/f), Myh11-cre/ERT2, R26-EYFP and MIP-GFP mice were from Jackson Laboratory, and Pdx1-Cre mice (Tg(Pdx1-CreTuv) were obtained from NCI Frederick. NGFLacZ/+ mice were gifted by Dr. David Ginty (Harvard) and NGFf/f mice were previously generated in the Minichiello laboratory (Muller et al., 2012). MIN6 cells were obtained from Dr. Jun-ichi Miyazaki (Osaka University) and Dr. Donald Steiner (University of Chicago). In vivo metabolic analyses For glucose tolerance tests, 1–2-month-old mice were fasted overnight, with a blood glucose reading the evening before the assay serving as fed blood glucose measurement. Mice were injected with glucose (2g/kg, i.p). Blood glucose measurements were made from tail blood using OneTouch Ultra glucometer (Gu et al., 2010). For acute 1NMPP1 treatments, mice received i.p. injections with 20ng/g 1NMPP1 or DMSO, 20 min prior to glucose administration. For in vivo insulin and NGF secretion, mice were fasted overnight before being injected with glucose (3 g/kg, i.p.). Blood was collected from the tail, spun down, and the plasma fractions subjected to either insulin (Crystal Chem, 90080) or NGF ELISA (Millipore, CYT304). Reactions were assessed using a Tecan infinite 200 plate reader. For insulin sensitivity, mice were separated into individual cages with food the evening prior to the assay. The next morning, mice were treated with 0.75 U/kg of insulin (Novolin-R; Novo Nordisk), and blood glucose measurements were made from tail blood (Bruning et al., 1997). Islet insulin secretion Islets pooled from 4–5 mice were allowed to recover overnight in RPMI 1640 media containing 5% FBS and 5 U/l penicillin/streptomycin. Groups of 5–10 islets of similar size were handpicked into 24-well dishes, washed in KRHB containing 2.8 mM glucose and allowed to stabilize for 1 hr. Islets were pre-incubated with vehicle (DMSO) alone, 1NMPP1 (20μM), cytochalasin D (25μM) alone or 1NMPP1 (20 μM) plus cytochalasin D (25μM) for 20 minutes. Islets were then incubated in 2.8 mM or 16.7 mM glucose, or KCl (30 mM) in KRHB buffer for another 30 min. Supernatant fractions were removed, islets lysed in acid ethanol, and both cellular and supernatant fractions were subjected to insulin ELISA (Crystal Chem). For total insulin, islets were lysed with acid ethanol, and insulin content determined by ELISA and normalized to DNA, measured from the same lysates using a PicoGreen kit (Invitrogen). To assess effects of endocytosis-defective TrkA receptors on GSIS, TrkAF592A islets were incubated with high-titer adenoviruses expressing either FLAG-TrkAY794F or control FLAG-TrkA receptors, and treated with doxycycline (100 ng/ml) in RPMI 1640 media containing 5% FBS and 5 U/l penicillin/streptomycin for 30 hr. Islets were then incubated in 1NMPP1 (20μM) for 20 min to silence endogenous TrkA receptors, and insulin secretion was measured in low or high glucose conditions. Human islets were obtained from 10 cadaver donors through the Integrated Islet Distribution Program (IIDP) funded by NIDDK. Islets were allowed to recover overnight in CMRL 1066 medium containing 10% human serum albumin, 5 U/L penicillin/5μg/L streptomycin, 2mM GlutaMAX, and 1mM sodium pyruvate, washed in KRHB containing 2.8 mM glucose and allowed to stabilize for 1 hr. Insulin secretion was measured using ELISA (ALPCO) as described above. Statistical analyses All graphs and statistical analyses were done using GraphPad Prism software. Except where noted, Student’s t tests were performed assuming Gaussian distribution, two-tailed, unpaired, and a confidence interval of 95%. One-way or two-way ANOVA analyses were performed when more than two groups were compared. Supplementary Material supplement We thank Mehboob Hussain, Seth Blackshaw, Steve Leach, Samer Hattar, Haiqing Zhao, and Robert Johnston for helpful discussions. We thank David Ginty for NGFLacZ/+ mice, and Jun-ichi Miyazaki and Donald Steiner for MIN6 cells. This work was supported by NIH R01 (DK108267) to R.K., and in part by P30 DK079637 to JHU-UMD-Diabetes Research Center. J.H., A.C., and P.B. were supported by an NIH training grant (T32GM007231). Figure 1 Pdx1-Cre;TrkAf/f mice are glucose intolerant and have reduced insulin secretion (A) Biotinylated NGF (b-NGF) binding in insulin-positive β-cells. (B) p75 expression in nerve fibers at the islet perimeter. (C,D) b-NGF binding is reduced in Pdx1-Cre;TrkAf/f islets. Scale bar, 100 μm for (A–D). (E) Loss of TrkA in Pdx-Cre;TrkAf/f islets. Representative images in (A–E) are from at least 3 animals per genotype. (F,G) Pdx1-Cre;TrkAf/f mice have normal fasted blood glucose, but have high fed glucose. Means ± SEM for n=13 control, 14 mutant mice for (F), n=19 control, 15 mutant mice for (G). (H,I) Pdx1-Cre;TrkAf/f mice are glucose intolerant and have reduced insulin secretion. Means ± SEM for n=13 control, 14 mutant mice for (H), n=4 mice per genotype for (I). p<0.05, p<0.01, *p<0.001, t-test for (F,G, H, and I). Figure 2 Acute TrkA inactivation impairs glucose tolerance and insulin secretion (A) Chemical-genetic approach to silence TrkA kinase activity in TrkAF592A mice. (B,C) 1NMPP1 decreases TrkA phosphorylation in TrkAF592A islets, but has no effect in wild-type islets. Quantification of phospho-TrkA levels normalized to total TrkA. Values are relative to vehicle-treated islets. Means ± SEM from 3 experiments. (D) 1NMPP1 injection, prior to a glucose challenge, impairs glucose tolerance in TrkAF592A mice. n=6 vehicle- and n=8 1NMPP1-injected TrkAF592A mice. (E) Area under the curve (AUC) for glucose tolerance. (F) 1NMPP1 injection impairs glucose-induced insulin secretion in TrkAF592A mice. n=5 mice each for DMSO- and 1NMPP1-injections. (G) 1NMPP1-mediated attenuation of GSIS in isolated TrkAF592A islets. Means ± SEM from n=6 experiments. (H) Total islet insulin content is normal with 1NMPP1 treatment. Means ± SEM from n=5 experiments. (I) Normal insulin sensitivity in TrkAF592A mice treated with 1NMPP1. Means ± SEM from n=5 mice each for 1NMPP1 and vehicle injections. p<0.05, p<0.01,p<0.001, n.s. not significant, t-test for (C, D, E, F, H, and I) and two- way ANOVA with Bonferroni post-hoc test for (G) Figure 3 TrkA signaling regulates insulin exocytosis via actin remodeling (A) TrkA activity is necessary for KCl-induced insulin secretion. 1NMPP1 suppresses insulin secretion in response to high KCl (30 min) in TrkAF592A islets. Means ± SEM from n=5 experiments. (B–E) TrkA signaling is necessary for glucose-induced surface localization of insulin granules. Docked insulin granules in β-cells are indicated by red arrowheads. Scale bar, 1 μm. (F) Quantification of docked insulin granules per micron of plasma membrane. Means ± SEM from n=3 mice each for vehicle and 1NMPP1 injections. (G–J) 16.7 mM glucose (30 min) decreases F-actin in TrkAF592A β-cells. Glucose-dependent F-actin remodeling was prevented by 1NMPP1. F-actin is marked by Alexa-546-phalloidin and β-cells were identified by insulin immunostaining. Scale bar, 5μm. (K) Average fluorescence intensity for F-actin. Means ± SEM from n=7 experiments. (L) 1NMPP1 attenuated glucose-induced increase in Rac1 activity. Means ± SEM from n=5 experiments. (M) Cytochalasin D corrected the GSIS defect in 1NMPP1-treated TrkAF592A islets. Means ± SEM from n=7 experiments. p<0.05, p<0.01, p<0.001, n.s. not significant, two-way ANOVA with Bonferroni post-test for (A, F, K, L and M). Figure 4 TrkA endocytosis is necessary for glucose-induced actin changes and insulin secretion (A) NGF (100 ng/ml, 30 min) promotes TrkA internalization in MIN6 cells, assessed by a cell-surface biotinylation assay. Supernatants after neutravidin precipitation were probed for p85 for protein normalization. (B) Quantification of internalized TrkA. Means ± SEM from 4 experiments. (C,D) FLAG-TrkB:A chimeric receptors undergo BDNF-dependent internalization in β-cells, assessed by FLAG antibody feeding. Scale bar, 5 μm. (E) Trk internalization into early endosomes was determined by assessing EEA1 and FLAG co-localization. Means ± SEM from 3 experiments. (F, G) The PLCγ inhibitor, U73122, blocks NGF-dependent TrkA internalization in MIN6 cells. Total surface TrkA is shown in the “no stripping” lane. Means ± SEM from 7 experiments. (H–K) TrkA phosphorylation at Y794, the PLCγ docking site, is necessary for TrkA endocytosis. Scale bar, 5 μm. (L) Internal accumulation of FLAG-TrkA receptors was determined by assessing co-localization of FLAG with cytoplasmic GFP. Means ± SEM from 5 experiments. (M–Q) FLAG-TrkAWT, but not FLAG-TrkAY794F receptor, expression corrected 1NMPP1-mediated impairment of glucose-induced F-actin disassembly in isolated TrkAF592A β-cells. Scale bar, 5 μm. Means ± SEM from n=6 experiments. (R) Expression of wild-type FLAG-TrkA, but not FLAG-TrkAY794F receptors, rescued 1NMPP1-mediated impairment of GSIS in TrkAF592A islets. Means ± SEM from n=5 experiments. p<0.05, p<0.01, p<0.001, n.s. not significant, t-test for (B, E), one-way ANOVA and Tukey's post-hoc test for (G, L), and two-way ANOVA and Bonferroni post-hoc test for (Q, R). Figure 5 Vascular-specific NGF deletion impairs glucose homeostasis, GSIS, and actin remodeling (A) Normal glucose tolerance in Pdx1-Cre;NGFf/f mice. Values are means ± SEM from n=10 mice per genotype. (B) GSIS is unaffected in Pdx1-Cre;NGFf/f islets. Means ± SEM from n=11 NGFf/f and n=7 Pdx1-Cre;NGFf/f animals. (C) X-gal staining reports NGF expression in large diameter blood vessels outside islets (red arrows), and also within islets (red arrowheads) in NGFLacZ/+ mice. An islet is outlined in dashed lines, and shown below in higher magnification. Scale bar, 100 μm (D,E) Co-localization of β-gal with smooth muscle actin (SMA), and PDGFRβ, a pericyte marker. Insulin staining is in blue. Inset shows intra-islet pericytes. Scale bars, 100 μm for (D,E) and 50 μm for inset in E. Representative images in (C–E) are from analyzing five NGFLacZ/+ mice. (F) Vascular-specific NGF loss impairs glucose tolerance in tamoxifen-injected Myh11-CreERT2;NGFf/f mice. Means ± SEM for n=7 mice each for vehicle and tamoxifen injections. (G) GSIS is attenuated in tamoxifen-injected Myh11-CreERT2;NGFf/f islets. Means ± SEM from n=8 mice each for vehicle and tamoxifen injection. (H–L) Cultured β-cells from tamoxifen-treated Myh11-CreERT2;NGFf/f mice show persistent cortical F-actin in high glucose. Scale bar, 5 μm. Means ± SEM from n=7 experiments. p<0.05, *p<0.01, n.s. not significant, two-way ANOVA and Bonferroni post-hoc test for (B, G, and L) and t-test for (F). Figure 6 NGF signaling is acutely regulated by glucose, and exogenous NGF enhances insulin secretion in human islets (A) Circulating NGF levels are rapidly increased 15 min after a glucose injection. Values are the mean ± SEM from n=14 mice. (B) 16.7 mM glucose (15 min) enhances NGF secretion from cultured vascular smooth muscle cells. Means ± SEM from n=3 experiments. (C, D) TrkA phosphorylation is increased by high glucose (15 min) in mouse islets. Means ± SEM from n=3 experiments. (E,F) 16.7 mM glucose (15 min) enhances TrkA phosphorylation in human islets. p<0.05, t-test. (G) NGF treatment potentiates GSIS in human islets. Islets were treated with either 2.8 or 16.7 mM glucose in the presence or absence of NGF (100 ng/ml) for 30 min. p<0.05, *p<0.01, t-test for (A, B, D, F, G). Figure 7 Neurotrophin signaling acutely promotes glucose-induced insulin secretion via actin reorganization in β-cells (1) NGF is secreted by pancreatic vascular smooth muscle cells and intra-islet pericytes in response to elevated glucose. (2) Vascular-derived NGF activates TrkA receptors on islet β-cells. (3) TrkA phosphorylation on Y794 leads to association and activation of the downstream effector, PLCγ, which triggers receptor internalization. (4) Endosomal signaling from internalized TrkA receptors recruits and activates the actin modulatory protein Rac1, to (5) remodel a peripheral F-actin barrier, and (6) promote insulin granule exocytosis. Highlights Glucose acutely stimulates NGF secretion and signaling in the pancreas Vascular NGF and beta-cell TrkA receptors are essential for glucose homeostasis Internalized TrkA receptors promote insulin exocytosis via F-actin reorganization NGF augments glucose-stimulated insulin secretion in human islets Author Contributions J.H. and R.K. designed the study, analyzed data, and wrote the manuscript. J.H., P.B., and. A.C performed experiments. L.M provided mice, contributed to discussions, and manuscript edits. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. Altar CA Burton LE Bennett GL Dugich-Djordjevic M 1991 Recombinant human nerve growth factor is biologically active and labels novel high-affinity binding sites in rat brain Proceedings of the National Academy of Sciences of the United States of America 88 281 285 1846041 Arntfield ME van der Kooy D 2011 beta-Cell evolution: How the pancreas borrowed from the brain: The shared toolbox of genes expressed by neural and pancreatic endocrine cells may reflect their evolutionary relationship BioEssays : news and reviews in molecular, cellular and developmental biology 33 582 587 Asahara S Shibutani Y Teruyama K Inoue HY Kawada Y Etoh H Matsuda T Kimura-Koyanagi M Hashimoto N Sakahara M 2013 Ras-related C3 botulinum toxin substrate 1 (RAC1) regulates glucose-stimulated insulin secretion via modulation of F-actin Diabetologia 56 1088 1097 23412604 Ascano M Bodmer D Kuruvilla R 2012 Endocytic trafficking of neurotrophins in neural development Trends in cell biology 22 266 273 22444728 Ascano M Richmond A Borden P Kuruvilla R 2009 Axonal targeting of Trk receptors via transcytosis regulates sensitivity to neurotrophin responses The Journal of neuroscience : the official journal of the Society for Neuroscience 29 11674 11685 19759314 Baeza-Raja B Sachs BD Li P Christian F Vagena E Davalos D Le Moan N Ryu JK Sikorski SL Chan JP 2016 p75 Neurotrophin Receptor Regulates Energy Balance in Obesity Cell reports 14 255 268 26748707 Bibel M Barde YA 2000 Neurotrophins: key regulators of cell fate and cell shape in the vertebrate nervous system Genes & development 14 2919 2937 11114882 Bodmer D Ascano M Kuruvilla R 2011 Isoform-specific dephosphorylation of dynamin1 by calcineurin couples neurotrophin receptor endocytosis to axonal growth Neuron 70 1085 1099 21689596 Borden P Houtz J Leach SD Kuruvilla R 2013 Sympathetic Innervation during Development Is Necessary for Pancreatic Islet Architecture and Functional Maturation Cell reports 4 287 301 23850289 Brouwers B de Faudeur G Osipovich AB Goyvaerts L Lemaire K Boesmans L Cauwelier EJ Granvik M Pruniau VP Van Lommel L 2014 Impaired islet function in commonly used transgenic mouse lines due to human growth hormone minigene expression Cell metabolism 20 979 990 25470546 Bruning JC Winnay J Cheatham B Kahn CR 1997 Differential signaling by insulin receptor substrate 1 (IRS-1) and IRS-2 in IRS-1-deficient cells Molecular and cellular biology 17 1513 1521 9032279 Bullo M Peeraully MR Trayhurn P Folch J Salas-Salvado J 2007 Circulating nerve growth factor levels in relation to obesity and the metabolic syndrome in women European journal of endocrinology/European Federation of Endocrine Societies 157 303 310 Cai EP Casimir M Schroer SA Luk CT Shi SY Choi D Dai XQ Hajmrle C Spigelman AF Zhu D 2012 In vivo role of focal adhesion kinase in regulating pancreatic beta-cell mass and function through insulin signaling, actin dynamics, and granule trafficking Diabetes 61 1708 1718 22498697 Chen X Ye H Kuruvilla R Ramanan N Scangos KW Zhang C Johnson NM England PM Shokat KM Ginty DD 2005 A chemical-genetic approach to studying neurotrophin signaling Neuron 46 13 21 15820690 Cooper JA 1987 Effects of cytochalasin and phalloidin on actin The Journal of cell biology 105 1473 1478 3312229 Cosker KE Segal RA 2014 Neuronal signaling through endocytosis Cold Spring Harbor perspectives in biology 6 D'Hoker J De Leu N Heremans Y Baeyens L Minami K Ying C Lavens A Chintinne M Stange G Magenheim J 2013 Conditional hypovascularization and hypoxia in islets do not overtly influence adult beta-cell mass or function Diabetes 62 4165 4173 23974922 Defronzo RA 2009 Banting Lecture. From the triumvirate to the ominous octet: a new paradigm for the treatment of type 2 diabetes mellitus Diabetes 58 773 795 19336687 Fagan AM Garber M Barbacid M Silos-Santiago I Holtzman DM 1997 A role for TrkA during maturation of striatal and basal forebrain cholinergic neurons in vivo The Journal of neuroscience : the official journal of the Society for Neuroscience 17 7644 7654 9315886 Fan F Ji C Wu Y Ferguson SM Tamarina N Philipson LH Lou X 2015 Dynamin 2 regulates biphasic insulin secretion and plasma glucose homeostasis The Journal of clinical investigation 125 4026 4041 26413867 Gu C Stein GH Pan N Goebbels S Hornberg H Nave KA Herrera P White P Kaestner KH Sussel L 2010 Pancreatic beta cells require NeuroD to achieve and maintain functional maturity Cell metabolism 11 298 310 20374962 Guariguata L Whiting DR Hambleton I Beagley J Linnenkamp U Shaw JE 2014 Global estimates of diabetes prevalence for 2013 and projections for 2035 Diabetes research and clinical practice 103 137 149 24630390 Hara M Wang X Kawamura T Bindokas VP Dizon RF Alcoser SY Magnuson MA Bell GI 2003 Transgenic mice with green fluorescent protein-labeled pancreatic beta -cells American journal of physiology Endocrinology and metabolism 284 E177 183 12388130 Harrington AW Ginty DD 2013 Long-distance retrograde neurotrophic factor signalling in neurons Nature reviews Neuroscience 14 177 187 23422909 Harrington AW St Hillaire C Zweifel LS Glebova NO Philippidou P Halegoua S Ginty DD 2011 Recruitment of actin modifiers to TrkA endosomes governs retrograde NGF signaling and survival Cell 146 421 434 21816277 Hatlapatka K Willenborg M Rustenbeck I 2009 Plasma membrane depolarization as a determinant of the first phase of insulin secretion American journal of physiology Endocrinology and metabolism 297 E315 322 19470830 Heinze C Seniuk A Sokolov MV Huebner AK Klementowicz AE Szijarto IA Schleifenbaum J Vitzthum H Gollasch M Ehmke H 2014 Disruption of vascular Ca2+-activated chloride currents lowers blood pressure The Journal of clinical investigation 124 675 686 24401273 Henquin JC Ravier MA Nenquin M Jonas JC Gilon P 2003 Hierarchy of the beta-cell signals controlling insulin secretion European journal of clinical investigation 33 742 750 12925032 Hingorani SR Tuveson DA 2003 Targeting oncogene dependence and resistance Cancer cell 3 414 417 12781357 Huang E Reichardt L 2001 Neurotrophins: Roles in neuronal development and function Annu Rev Neurosci 24 667 736 Indo Y Tsuruta M Hayashida Y Karim MA Ohta K Kawano T Mitsubuchi H Tonoki H Awaya Y Matsuda I 1996 Mutations in the TRKA/NGF receptor gene in patients with congenital insensitivity to pain with anhidrosis Nature genetics 13 485 488 8696348 Ishihara H Asano T Tsukuda K Katagiri H Inukai K Anai M Kikuchi M Yazaki Y Miyazaki JI Oka Y 1993 Pancreatic beta cell line MIN6 exhibits characteristics of glucose metabolism and glucose-stimulated insulin secretion similar to those of normal islets Diabetologia 36 1139 1145 8270128 Kalwat MA Thurmond DC 2013 Signaling mechanisms of glucose-induced F-actin remodeling in pancreatic islet beta cells Experimental & molecular medicine 45 e37 23969997 Kanaka-Gantenbein C Tazi A Czernichow P Scharfmann R 1995 In vivo presence of the high affinity nerve growth factor receptor Trk-A in the rat pancreas: differential localization during pancreatic development Endocrinology 136 761 769 7835308 Kim HC Cho YJ Ahn CW Park KS Kim JC Nam JS Im YS Lee JE Lee SC Lee HK 2009 Nerve growth factor and expression of its receptors in patients with diabetic neuropathy Diabetic medicine : a journal of the British Diabetic Association 26 1228 1234 20002474 Konstantinova I Nikolova G Ohara-Imaizumi M Meda P Kucera T Zarbalis K Wurst W Nagamatsu S Lammert E 2007 EphA-Ephrin-A-mediated beta cell communication regulates insulin secretion from pancreatic islets Cell 129 359 370 17448994 Kowluru A 2011 Friendly, and not so friendly, roles of Rac1 in islet beta-cell function: lessons learnt from pharmacological and molecular biological approaches Biochemical pharmacology 81 965 975 21300027 Lacy PE Klein NJ Fink CJ 1973 Effect of cytochalasin B on the biphasic release of insulin in perifused rat islets Endocrinology 92 1458 1468 4571983 Liu Y Rutlin M Huang S Barrick CA Wang F Jones KR Tessarollo L Ginty DD 2012 Sexually dimorphic BDNF signaling directs sensory innervation of the mammary gland Science 338 1357 1360 23224557 MacDonald PE Joseph JW Rorsman P 2005 Glucose-sensing mechanisms in pancreatic beta-cells Philosophical transactions of the Royal Society of London Series B, Biological sciences 360 2211 2225 16321791 Meloni M Descamps B Caporali A Zentilin L Floris I Giacca M Emanueli C 2012 Nerve growth factor gene therapy using adeno-associated viral vectors prevents cardiomyopathy in type 1 diabetic mice Diabetes 61 229 240 22187379 Miller FD Speelman A Mathew TC Fabian J Chang E Pozniak C Toma JG 1994 Nerve growth factor derived from terminals selectively increases the ratio of p75 to trkA NGF receptors on mature sympathetic neurons Developmental biology 161 206 217 8293873 Muller M Triaca V Besusso D Costanzi M Horn JM Koudelka J Geibel M Cestari V Minichiello L 2012 Loss of NGF-TrkA signaling from the CNS is not sufficient to induce cognitive impairments in young adult or intermediate-aged mice The Journal of neuroscience : the official journal of the Society for Neuroscience 32 14885 14898 23100411 Navarro-Tableros V Sanchez-Soto MC Garcia S Hiriart M 2004 Autocrine regulation of single pancreatic beta-cell survival Diabetes 53 2018 2023 15277381 Nevins AK Thurmond DC 2003 Glucose regulates the cortical actin network through modulation of Cdc42 cycling to stimulate insulin secretion American journal of physiology Cell physiology 285 C698 710 12760905 Orci L Gabbay KH Malaisse WJ 1972 Pancreatic beta-cell web: its possible role in insulin secretion Science 175 1128 1130 4551150 Ostenson CG Gaisano H Sheu L Tibell A Bartfai T 2006 Impaired gene and protein expression of exocytotic soluble N-ethylmaleimide attachment protein receptor complex proteins in pancreatic islets of type 2 diabetic patients Diabetes 55 435 440 16443778 Pictet RL Rall LB Phelps P Rutter WJ 1976 The neural crest and the origin of the insulin-producing and other gastrointestinal hormone-producing cells Science 191 191 192 1108195 Pingitore A Caroleo MC Cione E Castanera Gonzalez R Huang GC Persaud SJ 2016 Fine tuning of insulin secretion by release of nerve growth factor from mouse and human islet beta-cells Molecular and cellular endocrinology 436 23 32 27424144 Porat-Shliom N Milberg O Masedunskas A Weigert R 2013 Multiple roles for the actin cytoskeleton during regulated exocytosis Cellular and molecular life sciences : CMLS 70 2099 2121 22986507 Prentki M Matschinsky FM Madiraju SR 2013 Metabolic signaling in fuel-induced insulin secretion Cell metabolism 18 162 185 23791483 Reinert RB Brissova M Shostak A Pan FC Poffenberger G Cai Q Hundemer GL Kantz J Thompson CS Dai C 2013 Vascular endothelial growth factor-a and islet vascularization are necessary in developing, but not adult, pancreatic islets Diabetes 62 4154 4164 23884891 Rodriguez-Tebar A Dechant G Barde YA 1990 Binding of brain-derived neurotrophic factor to the nerve growth factor receptor Neuron 4 487 492 2157470 Rorsman P Braun M 2013 Regulation of insulin secretion in human pancreatic islets Annual review of physiology 75 155 179 Rorsman P Renstrom E 2003 Insulin granule dynamics in pancreatic beta cells Diabetologia 46 1029 1045 12879249 Rosenbaum T Sanchez-Soto MC Hiriart M 2001 Nerve growth factor increases insulin secretion and barium current in pancreatic beta-cells Diabetes 50 1755 1762 11473035 Rosenbaum T Vidaltamayo R Sanchez-Soto MC Zentella A Hiriart M 1998 Pancreatic beta cells synthesize and secrete nerve growth factor Proceedings of the National Academy of Sciences of the United States of America 95 7784 7788 9636228 Sasson A Rachi E Sakhneny L Baer D Lisnyansky M Epshtein A Landsman L 2016 Islet pericytes are required for beta-cell maturity Diabetes Schreiber R Levy J Loewenthal N Pinsk V Hershkovitz E 2005 Decreased first phase insulin response in children with congenital insensitivity to pain with anhidrosis Journal of pediatric endocrinology & metabolism : JPEM 18 873 877 16279365 Schuh M 2011 An actin-dependent mechanism for long-range vesicle transport Nature cell biology 13 1431 1436 21983562 Shibasaki T Takahashi H Miki T Sunaga Y Matsumura K Yamanaka M Zhang C Tamamoto A Satoh T Miyazaki J 2007 Essential role of Epac2/Rap1 signaling in regulation of insulin granule dynamics by cAMP Proceedings of the National Academy of Sciences of the United States of America 104 19333 19338 18040047 Song J Xu Y Hu X Choi B Tong Q 2010 Brain expression of Cre recombinase driven by pancreas-specific promoters Genesis 48 628 634 20824628 Sorkin A von Zastrow M 2009 Endocytosis and signalling: intertwining molecular networks Nature reviews Molecular cell biology 10 609 622 19696798 Stephens RM Loeb DM Copeland TD Pawson T Greene LA Kaplan DR 1994 Trk receptors use redundant signal transduction pathways involving SHC and PLC-gamma 1 to mediate NGF responses Neuron 12 691 705 8155326 Sutter A Riopelle RJ Harris-Warrick RM Shooter EM 1979 Nerve growth factor receptors. Characterization of two distinct classes of binding sites on chick embryo sensory ganglia cells The Journal of biological chemistry 254 5972 5982 571871 Tessarollo L 1998 Pleiotropic functions of neurotrophins in development Cytokine & growth factor reviews 9 125 137 9754707 Thurmond DC Gonelle-Gispert C Furukawa M Halban PA Pessin JE 2003 Glucose-stimulated insulin secretion is coupled to the interaction of actin with the t-SNARE (target membrane soluble N-ethylmaleimide-sensitive factor attachment protein receptor protein) complex Mol Endocrinol 17 732 742 12554769 van Obberghen E Somers G Devis G Vaughan GD Malaisse-Lagae F Orci L Malaisse WJ 1973 Dynamics of insulin release and microtubular-microfilamentous system. I. Effect of cytochalasin B The Journal of clinical investigation 52 1041 1051 4573352 Vetere A Choudhary A Burns SM Wagner BK 2014 Targeting the pancreatic beta-cell to treat diabetes Nature reviews Drug discovery 13 278 289 24525781 Wang P Fiaschi-Taesch NM Vasavada RC Scott DK Garcia-Ocana A Stewart AF 2015 Diabetes mellitus--advances and challenges in human beta-cell proliferation Nature reviews Endocrinology 11 201 212 Wang Z Oh E Clapp DW Chernoff J Thurmond DC 2011 Inhibition or ablation of p21-activated kinase (PAK1) disrupts glucose homeostatic mechanisms in vivo The Journal of biological chemistry 286 41359 41367 21969371 Weir GC Bonner-Weir S 2004 Five stages of evolving beta-cell dysfunction during progression to diabetes Diabetes 53 Suppl 3 S16 21 15561905 Wirth A Benyo Z Lukasova M Leutgeb B Wettschureck N Gorbey S Orsy P Horvath B Maser-Gluth C Greiner E 2008 G12-G13-LARG-mediated signaling in vascular smooth muscle is required for salt-induced hypertension Nature medicine 14 64 68 Wu C Lai CF Mobley WC 2001 Nerve growth factor activates persistent Rap1 signaling in endosomes The Journal of neuroscience : the official journal of the Society for Neuroscience 21 5406 5416 11466412 Xu B Xie X 2016 Neurotrophic factor control of satiety and body weight Nature reviews Neuroscience 17 282 292 27052383 Zhu X Hu R Brissova M Stein RW Powers AC Gu G Kaverina I 2015 Microtubules Negatively Regulate Insulin Secretion in Pancreatic beta Cells Developmental cell 34 656 668 26418295
PMC005xxxxxx/PMC5123904.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 7500844 3382 Mol Cell Endocrinol Mol. Cell. Endocrinol. Molecular and cellular endocrinology 0303-7207 1872-8057 27663076 5123904 10.1016/j.mce.2016.09.019 NIHMS819488 Article Ovarian transcriptome associated with reproductive senescence in the long-living Ames dwarf mice Schneider Augusto 12* Matkovich Scot J. 3 Saccon Tatiana 1 Victoria Berta 2 Spinel Lina 2 Lavasani Mitra 45 Bartke Andrzej 6 Golusinski Pawel 278 Masternak Michal M. 28* 1 Faculdade de Nutrição, Universidade Federal de Pelotas, Pelotas, RS, Brazil 2 College of Medicine, Burnett School of Biomedical Sciences, University of Central Florida, Orlando, FL 3 Center for Pharmacogenomics, Department of Medicine, Washington University School of Medicine, St. Louis, MO, USA 4 Rehabilitation Institute of Chicago, Chicago, IL, USA 5 Department of Physical Medicine and Rehabilitation, Northwestern University Feinberg School of Medicine, Chicago, IL, USA 6 Departments of Internal Medicine and Physiology, Southern Illinois University School of Medicine, Springfield, IL, USA 7 Department of Biology and Environmental Studies, Poznan University of Medical Sciences, Poznan, Poland 8 Department of Head and Neck Surgery, The Greater Poland Cancer Centre, Poznan, Poland * Corresponding authors: Michal M. Masternak, Ph.D., University of Central Florida, Burnett School of Biomedical Sciences, College of Medicine, 6900 Lake Nona Blvd., Orlando, FL 32827, Phone: +1 (407) 266-7113, Fax: +1 (407) 266-7002, [email protected] and Augusto Schneider, Ph.D., Universidade Federal de Pelotas, Faculdade de Nutrição, Rua Gomes Carneiro, 1 Sala 239, CEP 96020-220, Pelotas, RS, Brazil, Phone: +55 53 39211270, [email protected] 7 10 2016 20 9 2016 5 1 2017 05 1 2018 439 328336 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. The aim of the current work was to evaluate the ovarian follicle reserve and the ovarian transcriptome in Ames dwarf (df/df) mice. The results suggest a delayed ovarian aging in df/df mice compared to normal (N) mice. Although a high number of genes were differentially expressed during aging of N mice, only a small fraction of these changed with aging in df/df mice. These alterations involved more than 500 categorized biological processes. The majority of these biological processes, including inflammatory/immune responses, were up-regulated with aging in N mice, while old df/df mice were characterized by down-regulation of these same processes in comparison to age matched N mice. However, biological processes related to DNA damage and repairing were commonly down-regulated with aging in both genotypes. In conclusion, delayed ovarian aging in long-living df/df mice was associated with reduced expression of genes related to the inflammatory and immune responses. ovarian aging GH IGF mRNA transcriptome Introduction The ovarian reserve is defined as the quiescent pool of primordial follicles, oocytes surrounded by a single layer of flattened granulosa cells (Peters, 1969), which is established at mid-gestation in humans (Baker, 1963) and just after birth in mice (Peters, 1969). Oocytes enclosed in primordial follicles will remain quiescent until recruited to grow at some point of the female reproductive life (Hirshfield, 1994). This process of continuous depletion of the ovarian follicular reserve and progressive reduction in fertility observed as female mammals as they age is known as ovarian aging (te Velde et al., 1998). The depletion of the ovarian reserve in women marks the beginning of the menopause (Faddy et al., 1992), although long before fertility rates start to decline due to a reduced ovarian reserve (van Noord-Zaadstra et al., 1991). The primordial follicle reserve decreases about 10 times from the ages of 0.5 to 1.5 years in female mice (Kevenaar et al., 2006) and is already completely depleted in 2.5 year-old females (Słuczanowska-Głąbowska et al., 2013). The activation of primordial follicles is a strictly regulated process coordinated by several paracrine and endocrine factors. One of the main pathways implicated is the phosphoinositide 3-kinase (PI3k)/protein kinase B (Akt1) signaling and its downstream effector Forkhead Box O3a (Foxo3a) (Castrillon et al., 2003; John et al., 2007; John et al., 2008). Foxo3a phosphorylation is associated with oocyte growth and the irreversible activation of primordial follicles. This event is coordinated by the surrounding granulosa cells producing kit ligand (KITL) in response to the activation of the mechanistic target of rapamycin complex 1 (mTORC1), which in turn activates the oocyte PI3k/Akt/Foxo3a pathway (Zhang et al., 2014). Despite these findings, it is still unknown which factors regulate the age at which somatic cells of the follicles begin their transformation and induce the oocyte to leave the resting pool (Zhang and Liu, 2015). Nevertheless, it seems clear that insulin-like growth factor I (IGF-I) signaling is involved in the initiation of follicle growth as demonstrated in growth hormone (GH) receptor disrupted mice treated with exogenous IGF-I (Slot et al., 2006) and caloric restricted rats with reduced circulating IGF-I levels (Li et al., 2011). Ames dwarf mice (df/df) are considered an excellent model for the study of aging as they can live 30-65% longer than their normal littermates (Brown-Borg et al., 1996). The df/df mice have a defective Prop1 (Prophet of Pit1) gene, impairing anterior pituitary gland development and resulting in deficient secretion of GH, thyroid-stimulating hormone (TSH) and prolactin (Sornson et al., 1996). These mice are characterized by severely low circulating IGF-I and reduced adult body size (Chandrashekar and Bartke, 1993). Interestingly, treatment of df/df mice with exogenous GH from 2 to 6 wk of age can abolish positive effects on longevity and insulin sensitivity (Panici et al., 2010). Female df/df mice have normal ovarian cyclicity and can reproduce under hormonal stimulation, although they have delayed puberty (Bartke et al., 2001). We had previously demonstrated that Foxo3a phosphorylation is reduced in oocytes enclosed in primordial follicles from df/df mice, and that they have an aberrant expression of several genes in the PI3k/Akt/Foxo3a pathway (Schneider et al., 2014). Since reduced Foxo3a activation is suggested as the main regulatory mechanism of ovarian aging (Zhang et al., 2014), we hypothesize that it also leads to decreased activation of primordial follicles in df/df mice, although there is no published histological evidence of delayed ovarian aging in df/df mice to date. Regulation of gene expression plays a key role in coordinating follicle development and aging within the ovary (Sharov et al., 2008). However, there is a very limited knowledge about the transcriptome of the aging ovary, since most knockout mouse models of important ovarian genes (e.g. PTEN and FoxO3) have complete ovarian failure early after puberty (Castrillon et al., 2003; John et al., 2008; Reddy et al., 2009) and do not allow the study of gene expression profiles in later ages. In addition, the use of newer techniques of next generation sequencing applied for RNA (RNA-Seq) has greatly improved the accuracy of transcriptomics studies in comparison to microarray (Matkovich et al., 2010) and very limited data regarding the aging ovary is available. Therefore the aim of the current work is 1) to evaluate the ovarian follicle reserve in aged df/df mice through histological sampling, and 2) to compare the transcriptomes of the aging ovaries in df/df and N mice. Materials and Methods Animals and tissue collection Normal (N; n=14) and Ames dwarf mice (df/df; n=14) (all females) were bred and maintained under temperature- and light-controlled conditions (22 ± 2°C, 12 hour light/12 hour dark cycle) (Masternak et al., 2004). The animals were anesthetized and euthanized after overnight fasting and the pair of ovaries was collected. The ovarian collection for histological evaluation was performed at 12-13 mo of age (4 N and 4 df/df mice) and ovaries were immediately stored in 10% formaldehyde. The ovarian collection for RNA extraction was performed at 5-6 mo of age (young; 5 N and 5 df/df mice) and at 21-22 mo of age (old; 5 N and 5 df/df mice) and ovaries were immediately stored at −80° C. All animal procedures employed in our presented work were approved by and performed in accordance to the guidelines from the Laboratory Animal Care and Use Committee (LACUC) at the Southern Illinois University School of Medicine (Springfield, IL). Quantification of ovarian follicles and histological analyses For the histological analysis, ovaries (n=8, 12 mo old mice) were removed from formaldehyde and embedded in paraffin (Paraplast, Leica Biosystems, Richmond, IL, USA). One ovary of the pair was serially sectioned at 5 μm using a semi-automated rotary microtome (RM2245, Leica Biosystems, San Diego, CA, USA). Sampling started at the beginning of the visible area of the ovarian surface until the end of the structure, and every 6th section was selected and placed on a standard histological slide for staining and counting (adapted from Myers et al. (2004)). The slides were dried at 56°C for 24 h, stained with hematoxylin-eosin, and mounted with coverslips and synthetic resin (Sigma Chemical Company®, St. Louis, MO, USA). Images of the ovarian sections were captured with a digital camera coupled to a microscope (Nikon Eclipse E200, Nikon Corporation, Japan) using the 10 and 40X objectives, assisted by the software Moticam 5.0 (Motic, Hong Kong, China). Only follicles containing an oocyte with clearly visible nucleus were counted in each slide. Follicle classification was based on Myers et al. (2004). Follicles were classified as primordial (oocyte surrounded by a single layer of flattened granulosa cells), in transition (oocyte surrounded by a layer of flattened granulosa cells and at least one cuboid granulosa cell), primary (oocyte surrounded by a single layer of cuboidal granulosa cells), secondary (oocyte surrounded by two or more layers of cuboid granulosa cells without a visible antrum) and tertiary (follicles with a clearly defined antral space and a with a layer of granulosa cells around the oocyte). To estimate actual follicle quantity, the number of follicles in each category was multiplied by six to account for the section sampling and by two to account for the fact that only one ovary of the pair was used. The diameter of the oocyte nucleus, oocyte and follicle for each category was measured for N and df/df mice (n=3/follicle type/mice). RNA Sequencing analyses The pair of ovaries from young (6 mo, n=5 per group) and old (22 mo, n=4 df/df and 5 N) was removed from the −80° C and homogenized with QIAzol (Qiagen, Valencia, CA, USA) using 0.5 mm zirconium oxide beads in the Bullet Blender 24 (Next Advance, Averill Park, NY, USA). Total RNA was extracted using a commercial column purification system (miRNeasy Mini Kit, Qiagen) and on-column DNase treatment (RNase-free DNase Set, Qiagen) following manufacturer's instructions. The quantity and quality of RNA samples was determined using BioAnalyzer and RNA Nano Lab Chip Kit (Agilent Technologies, Santa Clara, CA, USA). All samples had RNA Integrity Numbers higher than 7.0 and proceeded to further analysis (n=19). Transcriptomic profile of individual samples was performed using RNA-sequencing as previously described (Matkovich and Dorn, 2015) using 1 μg of total RNA per sample as an initial input. Briefly, Poly A RNA enrichment was performed using Dynabeads mRNA Purification Kit (Life Technologies, Carlsbad, CA, USA) and mRNA fragmentation using a fragmentation buffer (40 mM Tris acetate, 100 mM K acetate, 30 mM Mg acetate, pH 8.2; 2 min 30 s at 95 C). First strand cDNA was generated using Superscript III 1st Strand Synthesis System (Life Technologies) and the second strand was generated using DNA Polymerase I (New England Biolabs, Ipswich, MA, USA), E. Coli DNA Ligase (New England Biolabs) and RNAse H (Life Technologies). Finally, libraries were prepared using End-It DNA End Repair Kit (Epicentre, Madison, WI, USA) and Klenow Fragment (3′→5′ exo-) (New England Biolabs) for the ligation of Illumina sequencing adapters (Illumina-Index-AdapterA and 5’-phosphorylated Illumina-Index-AdapterB) using the LigaFast™ Rapid DNA Ligation System (Promega, Madison, WI, USA). Fragments between 150-300 bp were recovered after 2% agarose gel electrophoresis in order to remove adapter dimers by size selection (QIAquick Gel Extraction Kit, Qiagen). Twenty different indexes were added to individual libraries using Phusion DNA polymerase (New England Biolabs) during 16 cycles of amplification. After that, samples were quantified (Nanodrop, Thermo Scientific, Wilmington, DE, USA), combined in two mixtures, and submitted to sequencing on two flowcell lanes (10 samples/lane) on a HiSeq 2500 instrument (Illumina Inc., San Diego, CA, USA) at Washington University GTAC. The mapping of sequencing reads to the mouse transcriptome (Illumina iGenomes annotation for UCSC mm10, http://support.illumina.com/sequencing/sequencing_software/igenome.html) was performed using Tophat 2.1.0.0 and Bowtie 2.2.6 as its underlying alignment algorithm (Kim et al., 2013). mRNA abundance was generated using Cufflinks 2.1.1 (Trapnell et al., 2010) and is presented as FPKM (Fragments Per Kb of exon per Million reads mapped to mRNAs). The number of reads aligned to each known transcript/mRNA and its corresponding gene was generated using HTSeq 0.6.1 (Anders et al., 2015). Genes with an average less than 1/100,000th of the aligned reads in at least one group were eliminated from differential expression analyses (15,770 transcripts detected out of 23,457 transcripts in the mm10 database). Those mRNAs defined as significantly up and down-regulated (FDR < 0.02 and FC > 1.50, or FDR < 0.02 and FC < 0.75) were further analyzed for enrichment of gene ontology (GO) terms (biological processes, molecular function and cellular component) and pathway analysis using the Gage and Pathview packages in R (Luo et al., 2009; Luo and Brouwer, 2013), considering P values lower than 0.05 as significant. All RNA-Seq data are available at the Gene Expression Omnibus (GEO) at NCBI under accession number GSE84078. Statistical analyses The results are presented as mean ± standard error of the mean (SEM). All statistical analyses for ovarian morphology were performed using Graphpad Prism 6 (Graphpad Software Inc., La Jolla, CA, USA). A t-test was performed to observe differences in the number of follicles in each stage and oocyte, nucleus and follicle diameter between N and df/df mice. A P value lower than 0.05 was considered statistically significant. Statistical analyses of differentially expressed mRNAs was performed using the software R (3.2.2) and the Bioconductor package DESeq (1.2.0) (Anders and Huber, 2010) using the HTSeq output count. Read counts were normalized for library depth, and pairwise comparisons between genotypes and ages (age within each genotype and genotype within each age), measuring fold change, uncorrected P-values from the negative binomial distribution, and adjusted P values (false discovery rate; FDR) were obtained. Principal components analysis (PCA) was also performed using R to observe sample distribution in a two dimensional plot and eliminate outliers. Genes with a FDR < 0.02 and FC > 1.50 were considered up-regulated; and FDR < 0.02 and FC < 0.75 were considered down-regulated. Results Ovarian morphology Twelve mo df/df mice had an increased ovarian reserve as indicated by a three-fold higher number of follicles in the primordial stage compared to N mice (P<0.01; Figure 1). The increased number of follicles in the primordial stage indicates a deficiency in the transition to the later stages of follicle development, confirmed by the reduced number of follicles in the secondary stage in df/df mice compared to N mice (P<0.01; Figure 1). The number of follicles in the tertiary stage (90.0 ± 27.6) and total follicles (2388.0 ± 287.7) was not different between df/df and N and mice (P>0.05). Regarding follicle and oocyte diameters, we only observed a smaller diameter for oocytes included in primordial follicles in df/df (5.5 ± 0.3 μm) compared to N mice (6.8 ± 0.4 μm, P = 0.02) (Table 1). Other histological parameters were not different between genotypes (Table 1 and Figure 2). RNA sequencing In average 26,577,010 reads per sample were obtained and of these 15,182,570 (55.3%) were aligned to the mouse genome. Using PCA of the 500 most variable genes (Figure 3A) and unsupervised hierarchical clustering for the 200 most expressed genes (Figure 3B), we can observe that samples grouped in similar pattern by genotype and age, indicating a low level of sample variability. Only one sample from a young N female mouse (Sample NY4) was removed from the study due to an anomalous profile. For analysis of differential gene expression (DEG) we considered the genes that were regulated with aging (22 vs 6 mo) within each genotype and also genes that changed between genotypes (df/df vs N) at each age. The highest number of DEGs was observed between old and young N mice (1,120 exclusively in N mice and 352 genes in common with df/df mice), while df/df mice had 104 genes exclusively regulated with aging (Figure 4A). Considering the differences between genotypes, there were 288 DEGs at 6 mo between df/df and N mice, 157 DEGs at 22 mo and 72 DEGs between df/df and N mice at both 6 and 22 mo of age (Figure 4B). The top 20 most differentially regulated genes (highest fold change) in each category are presented in Table 2 and a list with all differentially regulated genes is provided in Supplemental Table S1. Processes related to inflammatory response were highly regulated with aging in N mice and between old df/df and old N mice and a heatmap containing the top 50 DEG in this category is showed in Suppl. Figure 2. The enriched Kegg pathways between df/df and N mice and with aging are represented in Supplemental Tables S2 and S3, respectively. There were 23 pathways enriched with aging exclusively in N mice, 8 exclusively in df/df mice and 26 pathways commonly enriched with aging in df/df and N mice. Regarding changes between genotypes, there were 16 pathways exclusively enriched in df/df and N mice at 6 mo old, 22 exclusively at 22 mo old, and only 3 in common at 6 and 22 mo old mice. We provided detailed diagrams for the transforming growth factor β (TGF-β) and oocyte meiosis pathways, which were downregulated with age in N mice (Suppl. Figures 2A and 3A), and up-regulated in old df/df compared to old N mice (Suppl. Figures 2B and 3B). On the other hand the toll-like receptor (TLR) pathway was up-regulated with age in N mice (Suppl. Figure 4A) and down-regulated in old df/df compared to old N mice (Suppl. Figure 4B). The main regulated Kegg pathways are summarized in Figure 5. The number of enriched GO Terms for biological processes between ages and genotypes is presented in Figures 4C and 4D, respectively. The majority of processes that changed with age (22 vs 6 mo) were shared by N and df/df mice (483 GO Terms in common). On the other hand, there were more enriched GO terms between df/df and N mice at old age (578 terms) than at young age (173 terms). There were a low number of processes commonly regulated in df/df and N mice at both 6 and 22 mo of age (only 35 terms). The enriched GO Terms for biological processes are presented in Supplemental Tables S4 and S5 for genotype and age differences, respectively. Discussion The present study evaluated for the first time the ovarian histological profile and transcriptome signature from df/df and N mice during aging. Although the expression of several genes changed with aging in N mice, only 31% of these genes also changed with aging in df/df mice, indicating that ovaries of df/df mice go through fewer gene expression changes as they age in comparison to N mice. As a result, we observed that several important processes related to inflammation/immune response up-regulated with age in N mice and were down-regulated in old df/df compared to N mice. This is compatible with our histological data, which indicate that the ovarian primordial follicle pool is almost three times larger in df/df than N mice, which strongly indicates preservation of a younger ovarian phenotype as these long-living df/df mice age. Indeed, the expression of the anti-Müllerian hormone (Amh) gene, an indicator of the ovarian reserve (Kevenaar et al., 2006), was about seven fold higher in the ovaries of old df/df compared to old N mice, although not different between young df/df and N mice (Suppl. Table 1). Similar results were observed for the expression of Zp1, Zp2 and Zp3 genes (Suppl. Table 1), which are exclusively expressed in oocytes and can also indicate the state of the follicle reserve (Epifano et al., 1995), further confirming a sharp decline in the oocyte reserve in N mice with aging, which was much less pronounced in df/df mice. We also observed a reduced primordial follicle oocyte diameter in df/df mice. The transition of follicles from the primordial to the primary stage is dependent on oocyte growth (Lintern-Moore and Moore, 1979). In this sense, it is possible that oocyte size can be an indicator of the rate of follicular depletion, since the FoxO3a and PTEN knockout mice have premature ovarian failure associated to an increased oocyte size in primordial follicles only (Castrillon et al., 2003; Reddy et al., 2008). Therefore, these histological and molecular findings further validate the df/df mice as a good model for ovarian aging, where the slower activation rate of primordial follicles resulted in fewer changes in gene expression as the mice age. One of the main characteristics of df/df mice is the severely reduced levels of serum IGF-I, therefore we also overlap the 72 DEGs between df/df and N mice at both ages with findings from in vitro IGF treated mice fibroblast cell lines (Dupont et al., 2001). However, no common target genes were identified, which can be due to the fact different cell types are being compared or also there are other characteristics unique to df/df mice, such as low insulin levels, decreased anti-inflammatory profile, adipose tissue cell composition and reduced oxidative stress (Bartke, 2005; Bartke, 2011; Masternak and Bartke, 2012), can also be responsible for its longevity phenotype and contribute to differential ovarian gene expression. The top 150 enriched gene ontology terms for biological processes up-regulated with aging in N mice were all related to the inflammatory/immune response. Interestingly, the top 150 down-regulated terms between old df/df and old N mice are also related to the inflammatory/immune response, with 138 overlapping terms. This indicates that many processes differentially regulated between old df/df mice and old N mice originate from genes that are up-regulated with age in N mice but remain unchanged in df/df mice, as we observed 70% less DEGs with age in df/df compared to N mice. For example, we observed that most genes in the TLR signaling pathway were up-regulated with aging in N mice, although were down-regulated in old df/df compared to old N mice. Therefore, our results point to regulation of the inflammatory response as a central player in ovarian aging. The df/df, GHRKO and calorie restricted mice have all been extensively characterized as having a reduced pro-inflammatory profile, which may represent one of the major mechanisms promoting increased insulin sensitivity and extended longevity in these mice (Masternak and Bartke, 2012), therefore this was also expected to be observed in the ovary. Previous studies suggested that the reduction of the ovarian reserve is associated with an increased pro-inflammatory status of the ovarian surface (Smith and Xu, 2008). This increased inflammatory activity in the aging ovary is claimed as one of the main reasons for the higher incidence of ovarian epithelial cancer in postmenopausal women (Smith and Xu, 2008). In fact there is evidence that during ovarian aging in mice, along with follicular depletion, there is intense stromal remodeling associated with the activation of pro-inflammatory and anti-proliferative pathways (Zimon et al., 2006). Increased expression of inflammatory/immune response associated genes was also previously observed for the ovaries of aging mice, although calorie restriction did not affect the same pathways (Sharov et al., 2008), as observed in our current study for df/df mice, which are GH-deficient. On the other hand, mice subjected to high-fat diet have an accelerated depletion of the ovarian reserve, associated with an increased macrophage infiltration in the ovarian stroma (Skaznik-Wikiel et al., 2016). We observed that macrophage chemotaxis, macrophage activation and macrophage differentiation were also among the down-regulated biological processes in old df/df compared to old N mice. Therefore, these findings point to an association between the depletion of the ovarian reserve and an increased ovarian pro-inflammatory status, which were both attenuated in the ovaries of aging df/df mice. However, controversy remains regarding whether the increased follicular atresia with aging is eliciting a pro-inflammatory response in the ovaries or if the increased overall pro-inflammatory status with aging is accelerating the depletion of the ovarian follicular reserve. Recently, it was demonstrated that IL-1 knockout mice have delayed ovarian aging (Uri-Belapolsky et al., 2014), suggesting an active role for the inflammatory response in the process of follicle depletion. We also observed that IL-1 production was one of the down-regulated biological processes in old df/df compared to old N mice. Most of the gene ontology terms down-regulated with aging in both N and df/df mice (237 matching terms) are related to biological processes such as DNA repair, DNA replication, response to DNA damage, chromatin organization, double-strand DNA break repair, spindle organization, and related processes. Maintenance of oocyte DNA integrity is essential for fertility preservation, and the extreme longevity of oocytes enclosed in primordial follicles renders them vulnerable to oxidative damage. It is well known that the age-related decline in reproductive performance in women is paralleled by a decrease in the ovarian reserve and an increase in chromosomally abnormal conceptions (Titus et al., 2013). Irradiation induced DNA damage in oocytes enclosed in primordial follicles leads to the activation of mitochondrial apoptotic pathways and the loss of the ovarian follicle reserve (Hanoux et al., 2007; Kerr et al., 2012). Therefore, it is proposed that the damaged oocyte DNA itself can be one of the causes for reduction in the size of the ovarian reserve with aging. In addition, impairment of DNA double strand break repair is associated with accelerated loss of ovarian follicular reserve (Titus et al., 2013). Expression of BRCA1, a key member of a family of DNA double strand break repair proteins, was reduced with aging in human oocytes (Titus et al., 2013). In our study we observed a three-fold reduction in BRCA1 gene expression with aging in N mice, although its expression did not change with aging in df/df mice (Suppl. Table 1). Therefore, it is suggested that maintenance of high levels of DNA protecting factors can be a key element for preserving the ovarian reserve for a longer period. We have previously demonstrated that the delayed ovarian aging in df/df mice is associated with an increased presence of the FoxO3a transcription factor in its non-phosphorylated form in the nucleus (Schneider et al., 2014), which is essential not only for maintaining the primordial follicles dormant (John et al., 2008) but also for increasing the protection of the cell against oxidative stress (Salih and Brunet, 2008). Additionally, mutations in the BRCA1 gene are also directly related to the development of ovarian carcinomas (Godlewski and Kapuscinska, 1996). There is no evidence that the GH/IGF-I axis can regulate BRCA1 gene expression, however BRCA1 can regulate IGF-I and IGF receptor gene expression and therefore, BRCA1 is an important regulator for tumor progression (Kang et al., 2012; Maor et al., 2000). In the current study we demonstrated that df/df mice have an increased ovarian primordial follicle reserve compared to N mice, which is suggested to occur due to a lower rate of primordial follicle activation, since there was a reduction in the number of secondary growing follicles. One of the main pathways involved activation of primordial follicles and maintenance of the quiescent stage is the TGF-β signaling pathway (Knight and Glister, 2006). We observed that the KEGG pathways representing TGF-β signaling and resumption of oocyte meiosis were up-regulated in old df/df in comparison to N mice, although the same pathways were down-regulated with age in N mice. Therefore, this indicates a potential role for this pathway in the reduced rate of activation of primordial follicles observed in df/df mice, which is compatible to the overall lower activation of growth signaling pathways in other tissues observed in dwarf mice models for aging studies (Swindell, 2007). Moreover, we could observe the terms sexual reproduction, fertilization, gamete generation, oogenesis and germ cell development among the gene ontology terms for up-regulated biological processes in old df/df compared to old N mice. This further indicates that growth signaling pathways play a central role in the reduced rate of activation of the primordial reserve and cell proliferation and are a possibly reason for the delayed ovarian aging in this model. In conclusion, the present study indicates delayed ovarian aging in long-living df/df mice in comparison to N mice, which was accompanied by a profound change in the ovarian transcriptome signature. Although there was a high number of DEGs with aging in N mice, only a small fraction of those changed with aging in df/df mice. We have shown that more than 500 different biological processes were down-regulated in the ovaries of old df/df in comparison to old N mice. Importantly, our study showed that the majority of the biological processes including inflammatory and immune responses were up-regulated with aging in N mice, and these were down-regulated in old df/df mice when compared to age matched littermates. This important observation confirms the delayed aging phenotype in these unique long living animals, showing beneficial regulation of longevity in the reproductive organs of female df/df mice. Supplementary Material 1 2 3 4 5 6 Acknowledgments This work was supported by Universidade Federal de Pelotas internal funds to AS; Washington University Department of Medicine internal funds to SJM; and National Institute on Aging of the National Institutes of Health (grants number R01AG032290, P01AG031736). This work was also supported by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq). Figure 1 Number of primordial, primary, secondary and tertiary follicles in the ovaries of 12 mo-old Ames dwarf (df/df) and Normal (N) mice. Differences at P<0.05 were considered as significant. Figure 2 Detailed histological images of each follicle stage (primordial, primary, secondary, and tertiary) and whole ovarian sections for Ames dwarf and Normal mice. The diameter of oocytes enclosed in primordial follicles was smaller in Ames dwarf than Normal mice. No further histological differences were observed between genotypes. O – oocyte, ON – oocyte nuclei, GC – granulosa cells, TC - theca cells, A – antrum, CL – corpus luteum, TF – tertiary follicle. Figure 3 Principal components analysis of the 500 most variable ovarian expressed mRNAs Ames dwarf (n=9; Young – DY dark blue; Old – DO light blue) and Normal mice (n=10; Young – NY dark green; Old – NO light green) (Panel A). Unsupervised hierarchical clustering of expression levels for the top 200 most expressed genes in ovaries of df/df (n=9; Young – DY and Old – DO) and N mice (n=10; Young – NY and Old – NO) (Panel B). Figure 4 Number of differentially expressed genes (DEGs) between Ames dwarf (df/df) mice with aging (22 vs 6 mo; A), and between aged (22 mo old) and young (6 mo old) df/df and Normal (N) mice (B). Number of regulated gene ontology (GO) biological process (BP) terms between df/df mice with aging (22 vs 6 mo; C), and between aged (22 mo old) and young (6 mo old) df/df and N mice (D). Figure 5 Schematic representation of the regulated set of genes between old Normal (N) and young N mice (22 vs 6 mo) and between Ames dwarf (df/df) and N mice at old age (22 mo. IGF-I – insulin-like growth factor type I, IGFR – insulin-like growth factor receptor, TLR – toll like receptor, IRAK - IL-1 Receptor-Associated Kinases, NFκB – nuclear factor kappa B, IκB – inhibitor of κB, TNF – tumor necrosis factor alpha, IL-1β – interleukin 1 beta, PI3K - phosphatidylinositide 3-kinase, AKT – protein kinase B, PKA – protein kinase B, CYCB2 – cyclin B2, CDK1 - cyclin-dependent kinase 1, TGFβR – transforming growth factor beta receptor. Orange arrows indicate genes changed with age in Normal (N) mice and blue arrows indicate genes changed in old (22 mo) df/df compared to old N mice. Table 1 Diameter of nuclei, oocyte and follicle of primordial (n=12/group), primary (n=12/group), secondary (n=12/group) and tertiary (n=12/group) follicles in 12 mo Ames Dwarf and Normal mice. Category Normal Ames Dwarf P value Primordial follicle Nucleus (μm) 2.6 (±0.3) 2.4 (±0.2) 0.46 Oocyte (μm) 6.8 (±0.4) 5.5 (±0.3) 0.02 Follicle (μm) 8.8 (±0.2) 7.9 (±0.5) 0.29 Primary follicle Nucleus (μm) 3.5 (±0.4) 3.0 (±0.2) 0.30 Oocyte (μm) 9.2 (±1.2) 7.5 (±0.5) 0.21 Follicle (μm) 14.0 (±1.5) 11.4 (±0.6) 0.11 Secondary follicle Nucleus (μm) 6.2 (±0.7) 5.8 (±0.5) 0.70 Oocyte (μm) 19.1 (±1.4) 22.7 (±1.2) 0.17 Follicle (μm) 44.0 (±2.5) 42.0 (±5.3) 0.74 Tertiary follicle Nucleus (μm) 9.6 (±0.8) 8.8 (±0.5) 0.07 Oocyte (μm) 107.4 (±20.0) 102.6 (±8.3) 0.82 Follicle (μm) 165.7 (±16.4) 147.3 (±6.6) 0.29 Data is presented as mean ± standard error of the mean Table 2 The top 20 most differentially regulated genes in each category. Category Direction of change Top 20 changed genes Change with age (22 vs. 6 mo) Normal mice Up-regulated Klk1b24, 1700012P22Rik, 1700018A04Rik, Zbbx, 4833428L15Rik, Klk1b21, Arhgap40, Cyp26a1, Uox, Cwh43, Pkp1, Agr2, A930001A20Rik, Rp1, Prss22, Gm609, 1700086L19Rik, Mogat1, Syt10, Slc34a2 Down-regulated C87977, Rfpl4, E330017A01Rik, Nlrp9b, Gm13084, Fbxw28, Wee2, Oog4, Mos, Nlrp4f, Tcl1b1, Amh, C87414, Pin1-ps1, Dppa3, Zp1, Klf17, Oas1e, Khdc1b, Obox5 Ames dwarf mice Up-regulated Prss29, Acsm3, Comp, Zpld1, Agr3, Msx3, Agr2, Mogat1, BC048546, Sectm1b, Ugt1a1, Cftr, Mmp8, Rasgrf1, Dhrs9, Pih1d2, Ovgp1, Cldn22, Cyp2b10, Stoml3 Down-regulated Pinlyp, Chga, Slc38a3, Otof, Slc30a3, Bfsp1, Khdc3, Nos2, Oas1d, AU015836, Zar1, Khdc1b, Dppa5a, Pou5f1, Gbx2, Padi6, E330034G19Rik, Fam78a, Amh, Zfp957 Change with genotype (Ames dwarf vs Normal mice) 6 mo (young) Up-regulated Klk1b24, Cyp26a1, Klk1b21, Arhgap40, 1700086L19Rik, Npbwr1, Gm266, 1500015O10Rik, Insl3, Cwh43, Rp1, Ptgds, Prom2, Caly, Slc34a2, Cyp21a1, Ect2l, Iqub, Hsd17b3, Mlc1 Down-regulated Wnt10b, Onecut2, Plekhg4, Sfrp4, Ptgfr, Cyp4f18, Wisp2, Adcyap1, Sstr1, Abcb1b, Elfn1, 2010003K11Rik, Tmem200a, Atp1a3, Lrrc30, Tnc, Adh7, Cyp11a1, Tmem178, Gm13003 22 mo (old) Up-regulated Hsd17b3, Ptgds, Elfn2, Cyp21a1, C7, C87977, Insl3, Fabp3, Serpina5, Svs5, Cntn6, Serpina3a, Serpina3c, Serpina3j, Itih2, Cdk15, Tcl1b1, Spink4, Wfikkn2, Nlgn1 Down-regulated Adcyap1, Wnt10b, Sfrp4, Onecut2, Wisp2, Abcb1b, Tnc, Atp1a3, Elfn1, Cyp11a1, Kcnab3, Tmem178, Ptgfr, Cma1, Ephx2, Sez6, Spp1, Stx11, Sgk1, Ctnna2 Highlights The ovarian reserve is three times larger in Ames dwarf (df/df) mice than normal (N) mice. More genes were differentially expressed during aging in N than df/df mice. Inflammatory response genes were up-regulated with aging in N mice Inflammatory response genes were down-regulated in old df mice compared to N mice. DNA damage and repairing genes were down-regulated with aging in both genotypes. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. Author contributions A.S., S.J.M., A.B. and M.M.M. contributed to study design, sample analysis and manuscript preparation. T.S., B.V., L.S., M.L. and P.G. contributed to sample analysis and manuscript revision. All authors reviewed and approved the manuscript. Competing interests The authors disclose no competing interests. References Anders S Huber W Differential expression analysis for sequence count data Genome Biol 2010 11 R106 20979621 Anders S Pyl PT Huber W HTSeq--a Python framework to work with high-throughput sequencing data Bioinformatics 2015 31 166 9 25260700 Baker TG A Quantitative and Cytological Study of Germ Cells in Human Ovaries Proc R Soc Lond B Biol Sci 1963 158 417 33 14070052 Bartke A Brown-Borg H Mattison J Kinney B Hauck S Wright C Prolonged longevity of hypopituitary dwarf mice Exp Gerontol 2001 36 21 8 11162909 Bartke A Minireview: role of the growth hormone/insulin-like growth factor system in mammalian aging Endocrinology 2005 146 3718 23 15919742 Bartke A Pleiotropic effects of growth hormone signaling in aging Trends Endocrinol Metab 2011 22 437 42 21852148 Brown-Borg HM Borg KE Meliska CJ Bartke A Dwarf mice and the ageing process Nature 1996 384 33 Castrillon DH Miao L Kollipara R Horner JW DePinho RA Suppression of ovarian follicle activation in mice by the transcription factor Foxo3a Science 2003 301 215 8 12855809 Chandrashekar V Bartke A Induction of endogenous insulin-like growth factor-I secretion alters the hypothalamic-pituitary-testicular function in growth hormone-deficient adult dwarf mice Biol Reprod 1993 48 544 51 8452930 Dupont J Khan J Qu BH Metzler P Helman L LeRoith D Insulin and IGF-1 induce different patterns of gene expression in mouse fibroblast NIH-3T3 cells: identification by cDNA microarray analysis Endocrinology 2001 142 4969 75 11606465 Epifano O Liang LF Familari M Moos MC Jr. Dean J Coordinate expression of the three zona pellucida genes during mouse oogenesis Development 1995 121 1947 56 7635043 Faddy MJ Gosden RG Gougeon A Richardson SJ Nelson JF Accelerated disappearance of ovarian follicles in mid-life: implications for forecasting menopause Hum Reprod 1992 7 1342 6 1291557 Godlewski D Kapuscinska M BRCA1 and BRCA2 genes – new risk factors in hereditary forms of breast cancer and ovarian carcinoma Reports of Practical Oncology and Radiotherapy 1996 1 53 55 Hanoux V Pairault C Bakalska M Habert R Livera G Caspase-2 involvement during ionizing radiation-induced oocyte death in the mouse ovary Cell Death Differ 2007 14 671 81 17082817 Hirshfield AN Relationship between the supply of primordial follicles and the onset of follicular growth in rats Biol Reprod 1994 50 421 8 8142559 John GB Shirley LJ Gallardo TD Castrillon DH Specificity of the requirement for Foxo3 in primordial follicle activation Reproduction 2007 133 855 63 17616716 John GB Gallardo TD Shirley LJ Castrillon DH Foxo3 is a PI3K-dependent molecular switch controlling the initiation of oocyte growth Dev Biol 2008 321 197 204 18601916 Kang HJ Yi YW Kim HJ Hong YB Seong YS Bae I BRCA1 negatively regulates IGF-1 expression through an estrogen-responsive element-like site Cell Death Dis 2012 3 e336 22739988 Kerr JB Hutt KJ Michalak EM Cook M Vandenberg CJ Liew SH Bouillet P Mills A Scott CL Findlay JK Strasser A DNA damage-induced primordial follicle oocyte apoptosis and loss of fertility require TAp63-mediated induction of Puma and Noxa Mol Cell 2012 48 343 52 23000175 Kevenaar ME Meerasahib MF Kramer P van de Lang-Born BM de Jong FH Groome NP Themmen AP Visser JA Serum anti-mullerian hormone levels reflect the size of the primordial follicle pool in mice Endocrinology 2006 147 3228 34 16556768 Kim D Pertea G Trapnell C Pimentel H Kelley R Salzberg SL TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions Genome Biol 2013 14 R36 23618408 Knight PG Glister C TGF-beta superfamily members and ovarian follicle development Reproduction 2006 132 191 206 16885529 Li L Fu YC Xu JJ Chen XC Lin XH Luo LL Caloric restriction promotes the reproductive capacity of female rats via modulating the level of insulin-like growth factor-1 (IGF-1) Gen Comp Endocrinol 2011 174 232 7 21945120 Lintern-Moore S Moore GP The initiation of follicle and oocyte growth in the mouse ovary Biol Reprod 1979 20 773 8 454765 Luo W Friedman MS Shedden K Hankenson KD Woolf PJ GAGE: generally applicable gene set enrichment for pathway analysis BMC Bioinformatics 2009 10 161 19473525 Luo W Brouwer C Pathview: an R/Bioconductor package for pathway-based data integration and visualization Bioinformatics 2013 29 1830 1 23740750 Maor SB Abramovitch S Erdos MR Brody LC Werner H BRCA1 suppresses insulin-like growth factor-I receptor promoter activity: potential interaction between BRCA1 and Sp1 Mol Genet Metab 2000 69 130 6 10720440 Masternak MM Al-Regaiey K Bonkowski MS Panici J Sun L Wang J Przybylski GK Bartke A Divergent effects of caloric restriction on gene expression in normal and long-lived mice J Gerontol A Biol Sci Med Sci 2004 59 784 8 15345726 Masternak MM Bartke A Growth hormone, inflammation and aging Pathobiol Aging Age Relat Dis 2012 2 Matkovich SJ Zhang Y Van Booven DJ Dorn GW 2nd Deep mRNA sequencing for in vivo functional analysis of cardiac transcriptional regulators: application to Galphaq Circ Res 2010 106 1459 67 20360248 Matkovich SJ Dorn GW 2nd Deep sequencing of cardiac microRNA-mRNA interactomes in clinical and experimental cardiomyopathy Methods Mol Biol 2015 1299 27 49 25836573 Myers M Britt KL Wreford NG Ebling FJ Kerr JB Methods for quantifying follicular numbers within the mouse ovary Reproduction 2004 127 569 80 15129012 Panici JA Harper JM Miller RA Bartke A Spong A Masternak MM Early life growth hormone treatment shortens longevity and decreases cellular stress resistance in long-lived mutant mice FASEB J 2010 24 5073 9 20720157 Peters H The development of the mouse ovary from birth to maturity Acta Endocrinol (Copenh) 1969 62 98 116 5394354 Reddy P Liu L Adhikari D Jagarlamudi K Rajareddy S Shen Y Du C Tang W Hamalainen T Peng SL Lan ZJ Cooney AJ Huhtaniemi I Liu K Oocyte-specific deletion of Pten causes premature activation of the primordial follicle pool Science 2008 319 611 3 18239123 Reddy P Adhikari D Zheng W Liang S Hamalainen T Tohonen V Ogawa W Noda T Volarevic S Huhtaniemi I Liu K PDK1 signaling in oocytes controls reproductive aging and lifespan by manipulating the survival of primordial follicles Hum Mol Genet 2009 18 2813 24 19423553 Salih DA Brunet A FoxO transcription factors in the maintenance of cellular homeostasis during aging Curr Opin Cell Biol 2008 20 126 36 18394876 Schneider A Zhi X Moreira F Lucia T Jr. Mondadori RG Masternak MM Primordial follicle activation in the ovary of Ames dwarf mice J Ovarian Res 2014 7 120 25543533 Sharov AA Falco G Piao Y Poosala S Becker KG Zonderman AB Longo DL Schlessinger D Ko M Effects of aging and calorie restriction on the global gene expression profiles of mouse testis and ovary BMC Biol 2008 6 24 18522719 Skaznik-Wikiel ME Swindle DC Allshouse AA Polotsky AJ McManaman JL High-Fat Diet Causes Subfertility and Compromised Ovarian Function Independent of Obesity in Mice Biol Reprod 2016 Slot KA Kastelijn J Bachelot A Kelly PA Binart N Teerds KJ Reduced recruitment and survival of primordial and growing follicles in GH receptor-deficient mice Reproduction 2006 131 525 32 16514195 Słuczanowska-Głąbowska S Laszczy Ska M Piotrowska K W GB Rumianowski B Masternak M Arum O Kucia M Kopchick JJ Bartke A Ratajczak MZ The effect of calorie restriction on the presence of apoptotic ovarian cells in normal wild type mice and low-plasma-IGF-1 Laron dwarf mice J Ovarian Res 2013 6 67 24063422 Smith ER Xu XX Ovarian ageing, follicle depletion, and cancer: a hypothesis for the aetiology of epithelial ovarian cancer involving follicle depletion Lancet Oncol 2008 9 1108 11 19012860 Sornson MW Wu W Dasen JS Flynn SE Norman DJ O'Connell SM Gukovsky I Carriere C Ryan AK Miller AP Zuo L Gleiberman AS Andersen B Beamer WG Rosenfeld MG Pituitary lineage determination by the Prophet of Pit-1 homeodomain factor defective in Ames dwarfism Nature 1996 384 327 33 8934515 Swindell WR Gene expression profiling of long-lived dwarf mice: longevity-associated genes and relationships with diet, gender and aging BMC Genomics 2007 8 353 17915019 te Velde ER Scheffer GJ Dorland M Broekmans FJ Fauser BC Developmental and endocrine aspects of normal ovarian aging Mol Cell Endocrinol 1998 145 67 73 9922101 Titus S Li F Stobezki R Akula K Unsal E Jeong K Dickler M Robson M Moy F Goswami S Oktay K Impairment of BRCA1-related DNA double-strand break repair leads to ovarian aging in mice and humans Sci Transl Med 2013 5 172ra21 Trapnell C Williams BA Pertea G Mortazavi A Kwan G van Baren MJ Salzberg SL Wold BJ Pachter L Transcript assembly and quantification by RNASeq reveals unannotated transcripts and isoform switching during cell differentiation Nat Biotechnol 2010 28 511 5 20436464 Uri-Belapolsky S Shaish A Eliyahu E Grossman H Levi M Chuderland D Ninio-Many L Hasky N Shashar D Almog T Kandel-Kfir M Harats D Shalgi R Kamari Y Interleukin-1 deficiency prolongs ovarian lifespan in mice Proc Natl Acad Sci U S A 2014 111 12492 7 25114230 van Noord-Zaadstra BM Looman CW Alsbach H Habbema JD te Velde ER Karbaat J Delaying childbearing: effect of age on fecundity and outcome of pregnancy BMJ 1991 302 1361 5 2059713 Zhang H Risal S Gorre N Busayavalasa K Li X Shen Y Bosbach B Brannstrom M Liu K Somatic cells initiate primordial follicle activation and govern the development of dormant oocytes in mice Curr Biol 2014 24 2501 8 25438940 Zhang H Liu K Cellular and molecular regulation of the activation of mammalian primordial follicles: somatic cells initiate follicle activation in adulthood Human Reproduction Update 2015 dmv037 Zimon A Erat A Von Wald T Bissell B Koulova A Choi CH Bachvarov D Reindollar RH Usheva A Genes invoked in the ovarian transition to menopause Nucleic Acids Res 2006 34 3279 87 16807318
PMC005xxxxxx/PMC5123907.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 7905589 6173 Neurotoxicology Neurotoxicology Neurotoxicology 0161-813X 1872-9711 27773601 5123907 10.1016/j.neuro.2016.10.011 NIHMS825866 Article MDMA Decreases Gluatamic Acid Decarboxylase (GAD) 67-Immunoreactive Neurons in the Hippocampus and Increases Seizure Susceptibility: Role for Glutamate Huff Courtney L. a Morano Rachel L. b Herman James P. b Yamamoto Bryan K. c Gudelsky Gary A. a a Division of Pharmaceutical Sciences, University of Cincinnati - James Winkle College of Pharmacy, Cincinnati, OH 45267 b Department of Psychiatry and Behavioral Neuroscience, University of Cincinnati - College of Medicine, Cincinnati OH, 45219 c Department of Pharmacology and Toxicology, Indiana University - School of Medicine, Indianapolis, IN 46202 Corresponding Author: Gary A. Gudelsky, Ph.D., James Winkle College of Pharmacy, University of Cincinnati, 231 Albert Sabin Way, Cincinnati, OH 45267-0514 29 10 2016 20 10 2016 12 2016 01 12 2017 57 282290 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 3,4-Methylenedioxy-methamphetamine (MDMA) is a unique psychostimulant that continues to be a popular drug of abuse. It has been well documented that MDMA reduces markers of 5-HT axon terminals in rodents, as well as humans. A loss of parvalbumin-immunoreactive (IR) interneurons in the hippocampus following MDMA treatment has only been documented recently. In the present study, we tested the hypothesis that MDMA reduces glutamic acid decarboxylase (GAD) 67-IR, another biochemical marker of GABA neurons, in the hippocampus and that this reduction in GAD67-IR neurons and an accompanying increase in seizure susceptibility involve glutamate receptor activation. Repeated exposure to MDMA (3×10mg/kg, ip) resulted in a reduction of 37–58% of GAD67-IR cells in the dentate gyrus (DG), CA1, and CA3 regions, as well as an increased susceptibility to kainic acid-induced seizures, both of which persisted for at least 30 days following MDMA treatment. Administration of the NMDA antagonist MK-801 or the glutamate transporter type 1 (GLT-1) inducer ceftriaxone prevented both the MDMA-induced loss of GAD67-IR neurons and the increased vulnerability to kainic acid-induced seizures. The MDMA-induced increase in the extracellular concentration of glutamate in the hippocampus was significantly diminished in rats treated with ceftriaxone, thereby implicating a glutamatergic mechanism in the neuroprotective effects of ceftriaxone. In summary, the present findings support a role for increased extracellular glutamate and NMDA receptor activation in the MDMA-induced loss of hippocampal GAD67-IR neurons and the subsequent increased susceptibility to evoked seizures. MDMA Glutamate GABA Excitotoxicity 1 Introduction MDMA, popularly known as ecstasy or, more recently, as Molly, is a synthetic, psychostimulant and a popular drug of abuse. It is well documented that the repeated administration of MDMA results in a persistent deficit in biochemical markers of 5-HT axon terminals in both laboratory animals (Gudelsky & Yamamoto, 2003) and humans (Kish et al., 2010) that has been viewed as evidence of 5-HT neurotoxicity (Capela et al., 2009). Results from several studies suggest that MDMA-induced neurotoxicity extends beyond 5-HT terminals to include cell bodies in brain regions such as the hippocampus. These studies adduce the MDMA-induced increases in activated calpain-1 and caspase-3, cytochrome C, BAX expression, TUNEL staining, and DNA fragmentation as evidence that MDMA promotes apoptotic cell death (Frenzilli et al., 2007; Soleimani Asl et al., 2012; Tamburini et al., 2006; Wang et al., 2009). Additional evidence to suggest that MDMA produces hippocampal cell loss includes the findings that the number of parvalbumin-IR interneurons in the DG is reduced following MDMA administration. The MDMA-induced reduction in parvalbumin-IR GABA neurons appears to involve 5-HT2A receptor-dependant and cyclooxygenase-dependent mechanisms (Anneken et al., 2013; Collins et al., 2015a). The involvement of glutamate in this apparent GABAergic neurotoxicity is supported by the finding that MDMA increases the extracellular concentration of glutamate in the hippocampus (Anneken & Gudelsky, 2012) and that MK-801 suppresses the MDMA-induced reduction in parvalbumin-IR GABA neurons (Collins et al., 2015a). Parvalbumin interneurons in the hippocampus function to provide strong inhibitory control of granule cell neuronal firing (Freund & Buzsáki, 1996). Loss of parvalbumin-IR cells has been reported in patients with epilepsy, underlining the importance of these neurons in maintaining balance between excitation and inhibition (Arellano et al., 2004; DeFelipe et al., 1993). In preclinical studies, Giorgi et al. (2005) and Abad et. al (2014) have reported that mice treated with MDMA exhibit an increased sensitivity to kainic acid induced seizures. However, the mechanism by which MDMA treatment results in an increase in seizure susceptibility has not been investigated. In the present study, we demonstrate that MDMA reduces the number of GAD67-IR neurons in the hippocampus and further investigatge the role of glutamate in the reduction of this biomarker of GABA neurons, as well as in the concomitant increase in seizure susceptibility. 2 Materials and Methods 2.1 Animals Adult, male Sprague-Dawley rats (275–325g) (Harlan Laboratories, Indianapolis, IN) were used in the studies. Animals were housed two per cage in a temperature and humidity controlled room with a 12-hr light/dark cycle and allowed food and water ad libitum. All procedures were in strict adherence to the National Institutes of Health guidelines and approved by the institutional animal care and use committee. Animals were acclimated for at least one week to the housing facilities and diet before being used in the study. 2.2 Drugs and Treatment MDMA was provided by the National Institute on Drug Abuse (Bethesda, MA), was dissolved in 0.15 M NaCl, and administered ip. Animals were treated with either a single injection of MDMA (1×10 mg/kg, ip), a binge regimen of MDMA, (10 mg/kg, ip, every 2 hr for a total of 3 injections), or vehicle and euthanized either 7 or 30 days after treatment. This binge regimen of MDMA has been well-documented to produce 5-HT neurotoxicity (Puerta et al., 2009; Shankaran et al., 2001), as well as reductions in parvalbumin-IR GABA neurons (Anneken et al., 2013). Ceftriaxone was purchased from Besse Medical (West Chester, OH), was dissolved in 0.15 M NaCl, and administered as a single daily injection of 200mg/kg, ip for 7 days, a dosage regimen similar to that used by others (Rasmussen et al., 2011; Verma et al., 2010). MDMA (3 × 10 mg/kg, ip) or vehicle was given 24 hr following the last injection of ceftriaxone. MK-801 was purchased from Sigma-Aldrich (St. Louis, MO), was dissolved in 0.15 M NaCl, and administered at 0.3mg/kg, sc 30 min prior to each injection of MDMA or vehicle. MK-801 + MDMA treated rats were maintained an elevated ambient temperature (approximately 27° C) in order to maintain MDMA-induced hyperthermia (see results). To determine the effect of MDMA on seizure susceptibility, animals were treated with a binge regimen of MDMA, (10 mg/kg, ip, every two hr for a total of 3 injections) or vehicle. Seizures were induced by kainic acid (Sigma-Aldrich, St. Louis, MO), 8mg/kg, sc, 7 or 30 days following MDMA treatment (Golden et al., 1995). 2.3 Tissue Preparation Rats were deeply anesthetized with Euthasol (100–150 mg/kg, ip) and transcardially perfused with 500 ml of physiological saline followed by 500 ml of 4% paraformaldehyde in 0.1M sodium phosphate-buffered saline (NaPBS) (pH=7.4). After perfusion, brains were removed and postfixed in the same fixative at 4° C overnight. Brains were then cryoprotected in 30% sucrose in 0.1 M NaPBS for at least 48 hr. Brains were frozen in the embedding medium and then transferred to the cryostat. Coronal sections (30 μm) of the hippocampus and the nucleus accumbens were cut and kept in the cryoprotective buffer at 4 °C. In the present study, analysis of GAD67-IR was restricted to the dorsal hippocampus (Bregma −3.0 to −3.7mm) and the core of the nucleus accumbens (Bregma +1.3 to +1.7mm). 2.4 Immunohistochemistry For GAD67-IR interneurons labeled with diaminobenzidine (DAB) chromogen, free-floating sections were stained with antibodies as follows: sections were rinsed in 50 mM potassium phosphate-buffered saline (KPBS), incubated for 10 min in 1% H2O2 in PBS, washed 5 × 5 min in KPBS, incubated for 60 min at room temperature in blocking solution (50mM KPBS with 0.2% Triton X-100, and 0.1% bovine serum albumin (BSA)), and incubated overnight at 4 °C in blocking solution containing monoclonal mouse anti-GAD67 antiserum (1:1000; MAB5406, Millipore, Temecula, CA). Sections were then rinsed 5 × 5 min in KPBS and incubated for 60 min with biotinylated secondary antibody (1:500, Vector) with 0.1% BSA in KPBS. The sections were again rinsed 5 × 5 min and incubated in ABC Elite kit (Vector) for 1 hr. After further washes, the sections were exposed to DAB (Sigma-Aldrich, St. Louis, MO; D5905) in 30% hydrogen peroxide for 3 min. The sections were transferred into KPBS before being mounted onto glass microscope slides in 0.5% gelatin. Slides were allowed to air dry overnight and then dehydrated in increasing concentrations of alcohol, cleared with xylene, and coverslipped with DPX (Sigma-Aldrich, St. Louis, MO). For GAD67-IR interneurons labeled with the Cy3 fluorophore, free-floating sections were stained with antibodies as follows: sections were rinsed in 50 mM KPBS, washed 5 × 5 min in KPBS, incubated for 60 min at room temperature in blocking solution (50mM KPBS with 0.2% Triton X-100, and 0.1% BSA), and incubated overnight at 4 °C in blocking solution containing monoclonal mouse anti-GAD67 antiserum (1:1000; MAB5406, Millipore, Temecula, CA). Sections were rinsed 5 × 5 min in KPBS and then incubated for 60 min in blocking solution containing Cy3-conjugated AffiniPure Donkey Anti-Mouse IgG (H+L) secondary antibody (1:500; The Jackson Laboratory). The sections were again washed 5 × 5 min with KPBS and 1 X 5 min in potassium phosphate buffer (KPB). The sections were transferred into KPB before being mounted onto glass microscope slides in 0.5% gelatin. Slides were allowed to air dry overnight and coverslipped with gelvatol (Sigma-Aldrich). 2.5 Image Analysis Quantitative analysis of the number of DAB-labeled GAD67-IR neurons within the brain regions specified (CA1, CA3, DG, and nucleus accumbens) was performed with Scion Image analysis software. Digital images of each side of the regions specified, as defined by the rat stereotaxic brain atlas of Paxinos and Watson, were captured at 5× magnification with a Carl Zeiss Imager Z.1 (Carl Zeiss Microimaging, Thornwood, New York). At least four images were obtained per animal (one left and one right side from each of two different sections). Using Scion Image analysis, the region of interest in each image was outlined, and the area of the region was measured. The software analyzed the number of particles detected within the outlined region. Particles were considered to be individual cells, and the cell count was divided by the area of the region for statistical analysis. Stereological analysis of the number of Cy3-labeled GAD67-IR neurons within the brain regions specified (CA1, CA3, and DG) was performed with NIS Elements imaging software. All images were collected with a Nikon A1R confocal on a Nikon Ti Eclipse inverted microscope controlled by NIS Elements interface. Images were captured with a 20x Plan Apo VC (NA 0.75) objective lens. During quantification, every sixth section for a total of six sections through the dorsal hippocampus were systematically sampled. Confocal images were acquired as Z-stacks (0.95 μm thickness) and the representative image is a maximum intensity projection image from the Z-stack. The numerical densities (ND) of GAD67-IR cells were determined using a modified optical fractionator technique (Gundersen et al., 1999; West et al., 1991) and calculated by the following equation: ND=∑Countsh(area)/SV, where ΣCounts is the number of GAD67-IR cells per counting frame, area is the area of the counting frame, h is the height of the optical dissector, and SV is the volumetric shrinkage factor (Jinno et al., 1998). Means were derived by averaging the ND from two sections from each animal in the CA1, CA3, and DG. In the present study, stereological analyses were restricted to the dorsal hippocampus (Bregma −3.0 to −3.7mm) and only GAD67-IR cells within the stratum pyramidale of CA1, the stratum radiatum of CA3, and the granular cell layer of the DG were quantified. Due to the lack of homogeneity in cell distribution within the hippocampus and the relative ease which with an absolute count of GAD67-IR neurons can be performed (Noori & Fornal, 2011), a single large counting frame was applied to each region of interest and an absolute cell count was performed within the frame. The optical dissector height (h) was 9.5μm (0.95μm interval x 10 slices) set ~2μm below the lookup section. 2.6 Analysis of Seizure Susceptibility Rats were transferred from the animal housing facility to the experimental procedure room on the evening prior to injection of kainic acid and allowed to acclimate overnight. Behaviors were videotaped for 3 hr beginning immediately after injection of kainic acid. Seizures were assessed from recorded behaviors using a modified Racine scale: no response (0), frozen posture, staring, and/or facial clonus, (1); myoclonic twitching and tremor (2), forelimb clonus with lordotic posture (3), forelimb clonus with rearing (4), forelimb clonus with rearing, jumping, and falling (5) (Hellier et al., 1998). Only animals which exhibited behaviors consistent with a stage three seizure or above were marked as having seized. Latency to seizure was recorded as the time (min) at which an animal first exhibited a stage 3–5 behavior. 2.7 Microdialysis and Glutamate Analysis Rats were implanted with a stainless steel guide cannula under anesthesia (ketamine/xylazine 70/6 mg/kg, ip) 72 hr prior to the insertion of a dialysis probe. On the evening prior to the experiment, a concentric style dialysis probe was inserted through the guide cannula into the dorsal hippocampus; the coordinates of the tip of the probe were: A/P, −3.6mm., L, 2.0mm, and D/V −4.0mm. The active portion of the membrane for the probes was 2.0mm. The probes were connected to an infusion pump set to deliver modified Dulbecco’s phosphate buffered saline containing 1.2mm CaCl2 and 5mM glucose at a flow rate of 1 μl/min overnight. On the morning of the experiment, the flow rate was increased to 2 μl/min, and the probes were allowed to equilibrate for 1.5hr. Three collections were then taken at 30 min intervals to establish baseline values; thereafter samples were collected every hr for the duration of the experiment. All experiments were performed at an ambient temperature of 24° C. Data were calculated as a percentage of the baseline value for glutamate which was obtained by averaging the three baseline samples. Glutamate was derivatized according to the method described by Donzanti and Yamamoto (1988). The HPLC consisted of an OPA-HS column (Part #28064. Grace Discovery Science) connected to an amperometric detector (Bioanalytical Systems, West Lafayette, IN) equipped with a glassy carbon target electrode set at +700 mV. The mobile phase consisted of 0.1M Na2HPO4, 50mg/L EDTA, 20% methanol, pH 6.4. Peak heights were recorded with an integrator, and the concentration of glutamate was calculated on the basis of known standards. 2.8 Core Body Temperature Measurement A BioMedic Data Systems DAS-7007s Reader System, including the Smart Probe Wand and Implantable Programmable Temperature Transponders 300 (IPTT-300), was utilized to monitor core body temperature in rats which were later assessed for GAD67-IR. Transponders were implanted sc in the shoulder region under isoflurane anesthesia 3–4 days prior to the experiment. Core body temperatures were recorded every 30 min beginning 1 hr prior to and ending 7 hr after the first injection of MDMA. All recordings were carried out while the animals were freely roaming around their cage with the lid removed. The transponder temperature was measured in triplicate (performed in rapid succession) in each rat at every time point, and the mean value recorded. The rats that received MK-801 and MDMA were maintained at an elevated ambient temperature of approximately 27° C in order to maintain MDMA-induced hyperthermia. 2.9 Statistical Analysis The effect of MDMA on GAD67-IR neurons was analyzed in each brain region of interest using either a one-way analysis of variance (ANOVA) or a two-way ANOVA. Multiple pairwise comparisons were performed using Student-Newman-Keul’s test. All seizure data were analyzed using chi-square analysis followed by the fisher’s exact test. Latency to seizure data was analyzed using a t-test. All microdialysis and body temperature data were analyzed using two-way repeated measures ANOVA, and multiple pairwise comparisons were performed using post hoc analysis with a Student-Newman-Keuls test. Treatment differences for all data were considered statistically significant at p < 0.05. 3 Results 3.1 MDMA reduces DAB-labeled GAD67-IR in the hippocampus The binge regimen of MDMA (3×10 mg/kg, ip) significantly (p<0.05) reduced GAD67-IR in subregions of the hippocampus at both 7 and 30 days after treatment (Figure 1A). GAD67-IR in the CA1, CA3, and DG was decreased by 43%, 58%, and 37%, respectively, 7 days after the binge regimen of MDMA when compared to values for vehicle-treated controls. Significant (p<0.05) reductions of ~40 to 50% in GAD67-IR were still evident in the CA1, CA3, and DG 30 days after the binge regimen of MDMA. A single injection of MDMA (10 mg/kg, ip) was not sufficient to produce a reduction in GAD67-IR at either 7 or 30 days in any region of the hippocampus examined. GAD67-IR was also assessed in the nucleus accumbens; there was no reduction in GAD67-IR in this brain region following the binge regimen of MDMA (data not shown). A one-way ANOVA indicated a significant main effect of treatment for each region of the hippocampus: CA1 (F(3,17)= 11.646, p<0.001), CA3 (F(3,17)=10.987, p<0.001), and DG (F(3,15)=23.635, p<0.001). There was no significant effect of treatment on number of GAD67 cells in the nucleus accumbens (F (1,9)=0.211, p=0.657). Representative images of the hippocampus are presented in Figure 1B. 3.2 Involvement of glutamate in the MDMA-induced loss of hippocampal GAD67-IR neurons The role of glutamate in the MDMA-induced loss of GAD67-IR in the hippocampus was evaluated by treatment of rats with the NMDA glutamate antagonist MK-801 or ceftriaxone, which has been shown to up-regulate the glutamate transporter GLT-1. In this experiment, the number of GABA neurons was quantified by stereological counting of GAD67-IR neurons in the CA1, CA3 and DG. MDMA treatment significantly (p<0.05) reduced the densities of GAD67-IR neurons in the pyramidal layer of CA1, the molecular layer of CA3, and the granular layer of the DG by 33%, 30% and 34%, respectively, when compared to the values for vehicle-treated rats (Figure 2). MDMA did not significantly reduce the number of GAD67-IR neurons in any hippocampal region of rats that had received MK-801 or ceftriaxone (Figure 2). MK-801 or ceftriaxone treatment alone did not affect the number of GAD67-IR neurons. A two-way ANOVA indicated a significant main effect of treatment on the densities of GAD67-IR neurons in each brain region: CA1 (F (1, 29)=7.17, P=0.012); CA3 (F (1, 28)=4.54, P=0.042); DG (F (1, 27)=7.16, P=0.012). 3.3 MDMA treatment increases the susceptibility to kainic acid-induced seizures Kainic acid-induced seizures were evaluated 7 and 30 days following treatment with the binge regimen of MDMA or vehicle. There was no statistical difference in seizure incidence of animals treated with vehicle 7 or 30 days prior to kainic acid administration; therefore, these data for control animals were pooled for the purposes of statistical comparison. Chi square analysis of the data indicated that there was a significant difference in the incidence of seizures across treatment groups (X2 = 19.48, p=0.001) (Figure 3A). Seizures were observed in 29% of rats treated with vehicle and in 89% and 100% of the rats treated with MDMA 7 days or 30 days earlier. Post hoc analysis indicated that seizure incidence was significantly increased in both groups of MDMA treated animals (7 days, p<0.001; 30 days, p<0.002). Animals treated with the binge regimen of MDMA, either 7 or 30 days prior to kainic acid, also exhibited a significant (p<0.05) reduction in seizure latency compared to animals treated with vehicle (Figure 3B). A one-way ANOVA indicated a significant main effect of treatment on seizure latency amongst groups (F (2, 27)=3.58, P=0.042). Post hoc analysis indicated that seizure latency was significantly reduced in both groups of MDMA treated animals (7 days, p<0.023; 30 days, p<0.048). 3.4 Involvement of glutamate in the MDMA-induced increase in seizure susceptibility Kainic acid-induced seizures were recorded 7 days following the administration of vehicle or MDMA in groups of rats that had also received prior treatment with vehicle or MK-801. Chi square analysis revealed significant differences in seizure incidence across treatment groups (X2 = 12.464, p=0.006). Vehicle+MDMA treated rats were significantly (p<0.05) more likely to exhibit kainic acid seizures than their appropriate control animals (vehicle+vehicle). However, treatment of rats with MK-801 abolished the increase in seizure susceptibility produced by MDMA. Seizure incidence was significantly (p<0.05) less in the MK-801+MDMA treated rats than in the vehicle+MDMA treated animals. Moreover, seizure incidence in MK-801+MDMA treated rats did not differ compared to that in the control group (MK-801+vehicle, p=1) (Figure 4A). In a separate experiment, the effect of ceftriaxone treatment also was determined on the MDMA-induced increase in seizure susceptibility. Seizure incidence was determined to be significant across treatment groups, as determined by Chi square analysis (X2 = 15.122, p=0.002). Again, seizure incidence was significantly (p<0.05) increased in the vehicle+MDMA group when compared to the vehicle+vehicle group. Importantly, treatment of rats with ceftriaxone prevented the increase in seizure susceptibility produced by MDMA. Seizure incidence in the ceftriaxone+MDMA treated animals was significantly less than that in the vehicle+MDMA treated rats (p<0.05). Furthermore, seizure incidence in the ceftriaxone+MDMA treated group was not significantly different than that in the ceftriaxone+vehicle treated control group (Figure 4B). 3.5 Neither MK-801 nor ceftriaxone alters MDMA-induced hyperthermia 2MK-801 has been shown to attenuate the hyperthermic response to MDMA, and this may confound the interpretation of its potential neuroprotective effects (Farfel & Seiden, 1995). For this reason, all rats treated with MK-801+MDMA were kept at an elevated ambient temperature of approximately 27° C during the MDMA treatment regimen in an effort to maintain the hyperthermic response to MDMA. Body temperatures in rats treated with the binge regimen of MDMA increased approximately 2–2.5° C during the course of treatment (Figure 5A, B). The magnitude of the MDMA-induced hyperthermia was not significantly (p>0.05) different in ceftriaxone-treated (Figure 5A) or MK-801-treated (Figure 5B) rats when compared to rats given MDMA alone. 3.6 Ceftriaxone suppresses the MDMA-induced increase in extracellular glutamate The ability of ceftriaxone to attenuate the MDMA-induced increase in extracellular glutamate in the hippocampus was assessed by in vivo microdialysis. Extracellular glutamate in the dorsal hippocampus was significantly (p<0.001) elevated during the treatment of rats with the binge regimen of MDMA. Extracellular glutamate was elevated compared to baseline values beginning 2 hr following the first injection of MDMA and remained elevated for the next 4 hr (Figure 6). In contrast, the administration of MDMA to rats treated for 7 days with ceftriaxone did not result in an increase in extracellular glutamate (p=0.136). A repeated measures ANOVA revealed a significant effect of treatment on extracellular glutamate (F(3,30)=15.80, p<0.001). 4 Discussion The key findings of the present study include the following: 1) exposure to MDMA results in a dose-dependent and persistent reduction in GAD67-IR neurons in multiple regions of the hippocampus, 2) treatment with MDMA results in an increase in susceptibility to kainic acid-induced seizures, 3) blockade of NMDA receptors or enhancement of glutamate re-uptake prevents the MDMA-induced loss of hippocampal GAD67-positive GABA neurons, as well as the increase in seizure susceptibility. Overall, the present findings substantiate a role for glutamate in the MDMA-induced loss of GAD67-IR neurons within the hippocampus and the subsequent increase in seizure susceptibility. It is well documented that the repeated administration of MDMA results in persistent deficits in markers of 5-HT axon terminals that has traditionally been viewed as 5-HT neurotoxicity (Gudelsky & Yamamoto, 2003; McCann & Ricaurte, 2004). The present findings provide evidence that MDMA neurotoxicity extends beyond 5-HT axon terminals. Herein we document that both two-dimensional and three-dimensional (stereological) analyses reveal a dose-dependent and persistent reduction in GAD67-IR in the CA1, CA3, and DG of the hippocampus following treatment with MDMA. To our knowledge, this is the first demonstration that MDMA reduces GAD67-IR in any brain region. These data are consistent with earlier reports that MDMA decreases the concentration of GABA in the hippocampus (Perrine et al., 2010) and reduces the number parvalbumin-IR GABAergic neurons in the hippocampus of rodents (Abad et al., 2014; Anneken et al., 2013; Collins et al., 2015a). The MDMA-induced reduction in parvalbumin-positive neurons in the rat was shown previously to be evident only within the DG of the hippocampus (Anneken et al., 2013; Collins et al., 2015a). In the present study, MDMA treatment resulted in a decrease in the number of GAD67-IR neurons throughout subregions of the hippocampus, including CA1, CA3, as well as DG. This finding suggests that the loss of biochemical markers of GABA neurons produced by MDMA extends not only beyond parvalbumin-IR neurons but also beyond the DG. The long-lasting reduction in GAD67-IR could be representative of either GABAergic neuron cell death or a persistent downregulation in GAD67 expression and change in cell phenotype. However, MDMA has been shown to increase biomarkers, e.g., TUNEL staining, caspase-3, and cytochrome C, of cell death in the hippocampus (Frenzilli et al., 2007; Soleimani Asl et al., 2012; Tamburini et al., 2006; Wang et al., 2009) and reduce GABA concentrations (Perrine et al., 2010). Therefore, it is tempting to speculate that the increase in markers of cell death following MDMA-treatment is due to the loss of GABAergic neurons. If, however, the MDMA-induced reduction of GAD67-IR is the result of down-regulation, the long-lasting deficits in GAD67 expression might still result in dysfunctional GABAergic systems in the hippocampus (i.e., functional neurotoxicity). Several studies have reported that hippocampal GABA neurons are vulnerable to glutamate-mediated excitotoxicity (Kerner et al., 1997; Moga et al., 2003; Nyiri et al., 2003), and MDMA has been shown to selectively increase extracellular glutamate in the hippocampus (Anneken & Gudelsky, 2012). Therefore, the MDMA-induced increase in glutamate has been suggested to contribute to the damage to hippocampal GABAergic neurons produced by MDMA. Supporting this view, Collins et al. (2015a) reported that the MDMA-induced reduction in parvalbumin-IR in the DG is dependent upon the activation of NMDA receptors. In the present study, treatment with either MK-801, an antagonist of NMDA receptors, or ceftriaxone, an inducer of GLT-1 expression (Rothstein et al., 2005), diminished the MDMA-induced reduction in GAD67-IR. Ceftriaxone is a beta-lactam antibiotic and has been shown to provide neuroprotection against glutamate-mediated excitotoxicity (Beller et al., 2011; Hota et al., 2008; Jagadapillai et al., 2014; Liu et al., 2013). Ceftriaxone presumably maintains glutamate homeostasis through an enhancement of glutamate re-uptake into astrocytes, thereby limiting excessive increases in extracellular glutamate. In the present study, ceftriaxone prevented the MDMA-induced increase in extracellular glutamate in the hippocampus. Thus, blockade of glutamate receptors or suppression of the MDMA-induced increase in extracellular glutamate prevented the MDMA-induced loss of GAD67-IR neurons. These findings further support a role for glutamate excitotoxicity in the MDMA-induced loss of GAD67-positive hippocampal neurons. Hyperthermia is a critical component of the neurotoxicity of amphetamine analogs, including MDMA (Broening et al., 1995; Malberg & Seiden, 1998). In the present study, rats were administered MK-801 concurrently with MDMA at an elevated ambient temperature to maintain MDMA-induced hyperthermia. The core body temperatures of rats treated with the combination of MK-801 and MDMA, as well as ceftriaxone and MDMA, were not significantly different from rats treated with MDMA alone at any time point. This finding excludes the possibility that the protection afforded by MK-801 and ceftriaxone against the MDMA-induced loss of GAD67-IR neurons is due to alterations in MDMA-induced hyperthermia and further supports glutamatergic mechanisms in the role of these drugs. The systemic administration of kainic acid induces status epilepticus in rats and generates a syndrome of seizures and brain damage that mimics human temporal lobe epilepsy (Goodman, 1998). Presently, we have demonstrated that susceptibility to kainic acid-induced seizure in rats is increased one week after MDMA treatment and persists up to 30 days. Additionally, rats exposed to repeated MDMA treatment exhibited reduced latency to seizure. These results are consistent with previous studies by Giorgi et al. (2005) and Abad et al. (2014) who demonstrated that MDMA increases seizure susceptibility in mice. However, these earlier studies did not address potential mechanisms underlying the MDMA-induced increase in seizure susceptibility. The increased sensitivity to kainic acid seizures reported here following exposure to MDMA is prevented by treatment with MK-801 or ceftriaxone. Thus, glutamate receptor mechanisms, presumably initiating excitotoxicity, appear to contribute to the increase in seizure susceptibility following MDMA administration, as well as the MDMA-induced loss of hippocampal GAD67-IR neurons. Cortical excitability is regulated by glutamate and GABA (Petroff, 2002); therefore, imbalances of these neurotransmitters can alter excitability and potentially increase seizure activity (Schousboe & White, 2009). Moreover, the hippocampus is sensitive to shifts in cortical excitability, and thus is especially prone to generating seizures (Stafstrom, 2010). Given that GABA interneurons provide strong inhibitory control of neuronal excitability, it seems reasonable to speculate that a causal relationship exists between the MDMA-induced loss of GAD67-IR neurons in the hippocampus and the associated pro-convulsant state. Indeed, loss of GABA interneuron activity in the hippocampus has been suggested to be associated with increased vulnerability to epileptogenic processes (Sloviter et al., 2001). The recent findings of Collins et al. (2015b) also support an association between a purported MDMA-induced loss of GABA interneurons in the DG and increased hippocampal neuronal excitability. These investigators reported that the threshold intensity to drive after-discharges in the DG is reduced in MDMA-treated rats in which there is a reduction in hippocampal parvalbumin-IR neurons. Collins and coworkers (2015b) also demonstrated that paired pulse inhibition (which has been attributed to GABAergic inhibition in the DG) is reduced following MDMA administration. An alternative possibility is that the MDMA-induced depletion of brain 5-HT underlies the increased seizure susceptibility, inasmuch as 5-HT modulates both glutamatergic and GABAergic neurotransmission (Ciranna, 2006). Indeed, it has been shown that a genetic manipulation of zebrafish can result in reduced 5-HT content and increased seizure activity (Sourbron et al., 2016). However, Giorgi et al. (2005) have reported that dosage regimens of MDMA that do not result in 5-HT depletion still result in increased seizure susceptibility. Moreover, the differential effects of MK-801 on the MDMA-induced depletion of brain 5-HT and increased seizure susceptibility also appear to preclude the involvement of 5-HT neurotoxicity in the alteration in seizure threshold. Although MK-801 was initially shown to prevent MDMA-induced 5-HT neurotoxicity (Farfel et al., 1992), it subsequently was shown not to afford neuroprotection when MDMA hyperthermia is maintained (Farfel & Seiden, 1995), as was done in the present study. Thus, MK-801 prevented the increase in seizure susceptibility, as well as the loss of GAD67-IR neurons, despite the likelihood that depletion of brain 5-HT was still evident. Although the data are only correlational in nature, they support the view that glutamate signaling, rather than 5-HT neurotoxicity, underlies the increase in seizure susceptibility that accompanies the MDMA-induced loss of GAD67-IR neurons. Nevertheless, other mechanisms also may contribute to the reduced seizure threshold produced by MDMA. For example, hippocampal gene expression may be altered (Weber et al., 2014) during or following the binge regimen of MDMA that would result in changes in cellular excitability. These changes may involve pro-inflammatory cytokines and the immune system (Scharfman, 2007). Several investigators have examined the relevance of the neurotoxic regimen of MDMA in rats to doses of the drug typically abused by humans (Baumann et al., 2007; Green et al., 2003; Green et al., 2009; McCann & Ricaurte, 2001). The doses of MDMA necessary to produce various physiological, neurochemical and/or behavioral effects in rats appears to be approximately 4 times the dose necessary in humans (Baumann et al., 2007; Green et al., 2009). On the basis of this “effect scaling”, the binge regimen of MDMA employed in the current study may be 4–5 times a single dose acutely administered recreationally by humans (100 mg or 1.5 mg/kg). However, MDMA pharmacokinetics in humans is non-linear (de la Torre et al., 2000). Consequently, humans acutely administering more than 1 Ecstasy tablet (more than 1.5 mg/kg) may exhibit plasma concentrations of MDMA that approach those of rats given 20–30 mg/kg (Green et al., 2009) which approximates the dosage regimen given rats in the present study. Human abusers of MDMA have been shown to exhibit long-lasting changes in EEG. Dafters et al. (1999) has reported that previous MDMA abuse correlates with a reduction in EEG coherence (synchrony) and an increase in EEG power in the alpha and beta frequency bands. These results are consistent with those of Gamma et al. (2000), who also reported global increases in alpha 1 and beta 2/3 power, as well as an increase in theta power in human users of MDMA. Such EEG changes have often been related to drug-induced cognitive deficits (Dafters et al., 1999; Gamma et al., 2000). However, EEG alterations also may be indicative of alterations in brain excitability (Dinner et al., 2002; Giorgi et al., 2005). Further clinical studies in human abusers of MDMA may be warranted to address whether seizure threshold is reduced in these individuals. This work was supported, in part, by USPHS DA07427. Figure 1 MDMA selectively decreases DAB-labeled GAD67-IR in the rat hippocampus. A) Rats were treated with MDMA (1x or 3×10 mg/kg, ip) or vehicle 7 or 30 days prior to sacrifice. GAD67-IR neurons were counted in 6 right and left images of each brain region from 6 rats/treatment group. Indicates p<0.05 compared to vehicle. B) Representative images of DAB-labeled GAD67 cells in the CA1, CA3, and DG in rats treated with vehicle or MDMA (3×10 mg/kg, ip). Figure 2 Glutamate mediates MDMA-induced reductions in GAD67-IR. Animals received ceftriaxone (CEF) (200 mg/kg, ip, daily) for 7 days prior to MDMA, MK-801 (0.3mg/kg, sc, 30 min prior to each injection of MDMA) or vehicle prior to either MDMA (3 × 10 mg/kg, ip) or vehicle for a total of six treatment groups. Animals were sacrificed 7 days after treatment with MDMA, and GAD67-IR was assessed using a stereological technique in the CA1 (A), CA3 (B), and DG (C). GAD67-IR neurons were counted in 2 sections in each brain region from 6–8 rats/treatment group. Indicates p<0.05 compared to the values for vehicle treated rats. Figure 3 MDMA increases susceptibility to kainic acid-induced seizures. Rats were given kainic acid (8mg/kg, sc) 7 or 30 days following the administration of either vehicle or MDMA (3×10mg/kg, ip). A) The percentage of rats exhibiting stage 3–5 seizures is depicted. N=8–19 rats/group. B) Of the animals that seized in panel A (n=5–17 rats/group), latency to seizure was recorded as the time (min) at which the animal first exhibited a stage 3–5 behavior. Indicates p<0.05 compared to the values for vehicle treated rats. Figure 4 MK-801 and ceftriaxone prevent the MDMA-induced increase in seizure susceptibility. Rats were treated with A) MK-801 (0.3mg/kg, sc) or vehicle 30 min prior to each injection of either MDMA (3×10 mg/kg, ip) or vehicle. B) Ceftriaxone (CEF) (200mg/kg, i.p.) or vehicle was given daily for 7 days prior to either MDMA (3 × 10 mg/kg, ip) or vehicle. Kainic acid (8mg/kg, sc) was administered 7 days after MDMA treatment. The percentage of rats exhibiting stage 3–5 seizures is depicted. N=5–10 rats/group. Indicates (p<0.05) compared to VEH/VEH. #Indicates (p<0.05) compared to VEH/MDMA. Figure 5 MDMA-induced hyperthermia was maintained in MK-801- and Ceftriaxone- treated animals. Core body temperatures were recorded every 30 min beginning 1 hr prior and ending 7 hr after the first injection of MDMA. A) Rats were treated with either a single daily injection of ceftriaxone (CEF) (200mg/kg, i.p.) or vehicle for 7 days prior to either MDMA (3 × 10 mg/kg, ip) or vehicle. B) Rats were treated with either MK-801 (0.3mg/kg, sc) or vehicle 30 min prior to each injection of either MDMA (3×10 mg/kg, ip) or vehicle. The values for animals that received MDMA in addition to either CEF or MK-801 were not significantly different from the values for MDMA-treated rats (p>0.05). N=6–8 rats/group. Figure 6 Ceftriaxone prevents the MDMA-induced increase in extracellular glutamate. Extracellular glutamate was determined in the hippocampus in rats treated daily with ceftriaxone (CEF), (200 mg/kg, ip,) or vehicle for 7 days prior to either MDMA (3 × 10 mg/kg, ip, as indicated by the arrows) or vehicle. N=7–11 rats/group. *Indicates values that differ significantly (p<0.05) from the corresponding average baseline values for vehicle treated animals. Highlights MDMA treatment results in a loss of GAD67-IR hippocampal neurons MDMA treatment reduces the threshold for kainic acid-induced seizures The MDMA-induced decrease in GAD67-IR hippocampal neurons and increase in seizure susceptibility appear to involve glutamate receptor mechanisms. Conflict of Interest: The authors declare no competing financial interests. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. Abad S Junyent F Auladell C Pubill D Pallas M Camarasa J Escubedo E Camins A 2014 3,4-methylenedioxymethamphetamine enhances kainic acid convulsive susceptibility Prog Neuropsychopharmacol Biol Psychiatry 54 231 242 24977329 Anneken JH Cunningham JI Collins SA Yamamoto BK Gudelsky GA 2013 MDMA increases glutamate release and reduces parvalbumin-positive GABAergic cells in the dorsal hippocampus of the rat: Role of cyclooxygenase J Neuroimmune Pharmacol 8 58 65 23179355 Anneken JH Gudelsky GA 2012 MDMA produces a delayed and sustained increase in the extracellular concentration of glutamate in the rat hippocampus Neuropharmacology 63 1022 1027 22842073 Arellano J Munoz A Ballesteros-Yanez I Sola R DeFelipe J 2004 Histopathology and reorganization of chandelier cells in the human epileptic sclerotic hippocampus Brain 127 45 64 14534159 Baumann MH Wang X Rothman RB 2007 3,4-methylenedioxymethamphetamine (MDMA) neurotoxicity in rats: A reappraisal of past and present findings Psychopharmacology (Berl) 189 407 424 16541247 Beller JA Gurkoff GG Berman RF Lyeth BG 2011 Pharmacological enhancement of glutamate transport reduces excitotoxicity in vitro Restor Neurol Neurosci 29 331 346 21846950 Broening HW Bowyer JF Slikker W Jr 1995 Age-dependent sensitivity of rats to the long-term effects of the serotonergic neurotoxicant (+/−)-3,4-methylenedioxymethamphetamine (MDMA) correlates with the magnitude of the MDMA-induced thermal response J Pharmacol Exp Ther 275 325 333 7562567 Ciranna L 2006 Serotonin as a modulator of glutamate- and GABA-mediated neurotransmission: Implications in physiological functions and in pathology Curr Neuropharmacol 4 101 114 18615128 Collins SA Gudelsky GA Yamamoto BK 2015a MDMA-induced loss of parvalbumin interneurons within the dentate gyrus is mediated by 5HT2A and NMDA receptors Eur J Pharmacol 761 95 100 25936514 Collins SA Huff C Chiaia N Gudelsky GA Yamamoto BK 2015b MDMA increases excitability in the dentate gyrus: Role of 5HT2A receptor induced PGE2 signaling J Neurochem Dafters RI Duffy F O’Donnell PJ Bouquet C 1999 Level of use of 3,4-methylenedioxymethamphetamine (MDMA or ecstasy) in humans correlates with EEG power and coherence Psychopharmacology (Berl) 145 82 90 10445376 de la Torre R Farre M Ortuno J Mas M Brenneisen R Roset PN Segura J Cami J 2000 Non-linear pharmacokinetics of MDMA (‘ecstasy’) in humans Br J Clin Pharmacol 49 104 109 10671903 DeFelipe J Garcia Sola R Marco P del Rio MR Pulido P Ramon y Cajal S 1993 Selective changes in the microorganization of the human epileptogenic neocortex revealed by parvalbumin immunoreactivity Cereb Cortex 3 39 48 7679938 Dinner DS Neme S Nair D Montgomery EB Jr Baker KB Rezai A Luders HO 2002 EEG and evoked potential recording from the subthalamic nucleus for deep brain stimulation of intractable epilepsy Clin Neurophysiol 113 1391 1402 12169320 Donzanti BA Yamamoto BK 1988 An improved and rapid HPLC-EC method for the isocratic separation of amino acid neurotransmitters from brain tissue and microdialysis perfusates Life Sci 43 913 922 2901021 Farfel GM Seiden LS 1995 Role of hypothermia in the mechanism of protection against serotonergic toxicity. I. experiments using 3,4-methylenedioxymethamphetamine, dizocilpine, CGS 19755 and NBQX J Pharmacol Exp Ther 272 860 867 7531765 Farfel GM Vosmer GL Seiden LS 1992 The N-methyl-D-aspartate antagonist MK-801 protects against serotonin depletions induced by methamphetamine, 3,4-methylenedioxymethamphetamine and p-chloroamphetamine Brain Res 595 121 127 1361410 Frenzilli G Ferrucci M Giorgi FS Blandini F Nigro M Ruggieri S Murri L Paparelli A Fornai F 2007 DNA fragmentation and oxidative stress in the hippocampal formation: A bridge between 3,4-methylenedioxymethamphetamine (ecstasy) intake and long-lasting behavioral alterations Behav Pharmacol 18 471 481 17762515 Freund TF Buzsáki G 1996 Interneurons of the hippocampus Hippocampus 6 347 470 8915675 Gamma A Frei E Lehmann D Pascual-Marqui RD Hell D Vollenweider FX 2000 Mood state and brain electric activity in ecstasy users Neuroreport 11 157 162 10683849 Giorgi FS Pizzanelli C Ferrucci M Lazzeri G Faetti M Giusiani M Pontarelli F Busceti CL Murri L Fornai F 2005 Previous exposure to (±) 3,4-methylenedioxymethamphetamine produces long-lasting alteration in limbic brain excitability measured by electroencephalogram spectrum analysis, brain metabolism and seizure susceptibility Neuroscience 136 43 53 16203101 Golden GT Smith GG Ferraro TN Reyes PF 1995 Rat strain and age differences in kainic acid induced seizures Epilepsy Res 20 151 159 7750511 Goodman PH 1998 Experimental Models of Status Epilepticus Peterson SL Albertson TE Neuropharmacology methods in epilepsy research CRC Press Green AR Gabrielsson J Marsden CA Fone KC 2009 MDMA: On the translation from rodent to human dosing Psychopharmacology (Berl) 204 375 378 19139850 Green AR Mechan AO Elliott JM O’Shea E Colado MI 2003 The pharmacology and clinical pharmacology of 3,4-methylenedioxymethamphetamine (MDMA, “ecstasy”) Pharmacol Rev 55 463 508 12869661 Gudelsky GA Yamamoto BK 2003 Neuropharmacology and neurotoxicity of 3,4-methylenedioxymethamphetamine Methods Mol Med 79 55 73 12506690 Gundersen HJ Jensen EB Kieu K Nielsen J 1999 The efficiency of systematic sampling in stereology--reconsidered J Microsc 193 199 211 10348656 Hellier JL Patrylo PR Buckmaster PS Dudek FE 1998 Recurrent spontaneous motor seizures after repeated low-dose systemic treatment with kainate: Assessment of a rat model of temporal lobe epilepsy Epilepsy Res 31 73 84 9696302 Hota SK Barhwal K Ray K Singh SB Ilavazhagan G 2008 Ceftriaxone rescues hippocampal neurons from excitotoxicity and enhances memory retrieval in chronic hypobaric hypoxia Neurobiol Learn Mem 89 522 532 18304843 Jagadapillai R Mellen NM Sachleben LR Jr Gozal E 2014 Ceftriaxone preserves glutamate transporters and prevents intermittent hypoxia-induced vulnerability to brain excitotoxic injury PLoS One 9 e100230 25014412 Jinno S Aika Y Fukuda T Kosaka T 1998 Quantitative analysis of GABAergic neurons in the mouse hippocampus, with optical disector using confocal laser scanning microscope Brain Res 814 55 70 9838044 Kerner JA Standaert DG Penney JB Jr Young AB Landwehrmeyer GB 1997 Expression of group one metabotropic glutamate receptor subunit mRNAs in neurochemically identified neurons in the rat neostriatum, neocortex, and hippocampus Brain Res Mol Brain Res 48 259 269 9332723 Kish SJ Lerch J Furukawa Y 2010 Decreased cerebral cortical serotonin transporter binding in ecstasy users: A positron emission tomography/[(11)C]DASB and structural brain imaging study Brain 133 1779 1797 20483717 Kosaka T Katsumaru H Hama K Wu JY Heizmann CW 1987 GABAergic neurons containing the Ca2+-binding protein parvalbumin in the rat hippocampus and dentate gyrus Brain Res 419 119 130 3315112 Liu CH Jiao H Guo ZH Peng Y Wang WZ 2013 Up-regulated GLT-1 resists glutamate toxicity and attenuates glutamate-induced calcium loading in cultured neurocytes Basic Clin Pharmacol Toxicol 112 19 24 22998524 Malberg JE Seiden LS 1998 Small changes in ambient temperature cause large changes in 3,4-methylenedioxymethamphetamine (MDMA)-induced serotonin neurotoxicity and core body temperature in the rat J Neurosci 18 5086 5094 9634574 McCann UD Ricaurte GA 2004 Amphetamine neurotoxicity: Accomplishments and remaining challenges Neurosci Biobehav Rev 27 821 826 15019431 McCann UD Ricaurte GA 2001 Caveat emptor: Editors beware Neuropsychopharmacology 24 333 336 11256359 Moga DE Janssen WG Vissavajjhala P Czelusniak SM Moran TM Hof PR Morrison JH 2003 Glutamate receptor subunit 3 (GluR3) immunoreactivity delineates a subpopulation of parvalbumin-containing interneurons in the rat hippocampus J Comp Neurol 462 15 28 12761821 Noori HR Fornal CA 2011 The appropriateness of unbiased optical fractionators to assess cell proliferation in the adult hippocampus Front Neurosci 5 140 22207833 Nyiri G Stephenson FA Freund TF Somogyi P 2003 Large variability in synaptic N-methyl-D-aspartate receptor density on interneurons and a comparison with pyramidal-cell spines in the rat hippocampus Neuroscience 119 347 363 12770551 Perrine SA Ghoddoussi F Michaels MS Hyde EM Kuhn DM Galloway MP 2010 MDMA administration decreases serotonin but not N-acetylaspartate in the rat brain Neurotoxicology 31 654 661 20800616 Petroff OA 2002 GABA and glutamate in the human brain Neuroscientist 8 562 573 12467378 Puerta E Hervias I Goni-Allo B Lasheras B Jordan J Aguirre N 2009 Phosphodiesterase 5 inhibitors prevent 3,4-methylenedioxymethamphetamine-induced 5-HT deficits in the rat J Neurochem 108 755 766 19187094 Rasmussen BA Baron DA Kim JK Unterwald EM Rawls SM 2011 Beta-lactam antibiotic produces a sustained reduction in extracellular glutamate in the nucleus accumbens of rats Amino Acids 40 761 764 20383795 Rothstein JD Patel S Regan MR 2005 Beta-lactam antibiotics offer neuroprotection by increasing glutamate transporter expression Nature 433 73 77 15635412 Scharfman HE 2007 The neurobiology of epilepsy Curr Neurol Neurosci Rep 7 348 354 17618543 Schousboe A White HS 2009 GLIA/ASTROCYTES \| Glial Modulation of Excitability via Glutamate and GABA Transporters Schwartzkroin PA Encyclopedia of basic epilepsy research Academic Press Oxford 397 401 Shankaran M Yamamoto BK Gudelsky GA 2001 Ascorbic acid prevents 3,4-methylenedioxymethamphetamine (MDMA)-induced hydroxyl radical formation and the behavioral and neurochemical consequences of the depletion of brain 5-HT Synapse 40 55 64 11170222 Sloviter RS Ali-Akbarian L Horvath KD Menkens KA 2001 Substance P receptor expression by inhibitory interneurons of the rat hippocampus: Enhanced detection using improved immunocytochemical methods for the preservation and colocalization of GABA and other neuronal markers J Comp Neurol 430 283 305 11169468 Soleimani Asl S Farhadi MH Moosavizadeh K Samadi Kuchak Saraei A Soleimani M Jamei SB Joghataei MT Samzadeh-Kermani A Hashemi-Nasl H Mehdizadeh M 2012 Evaluation of bcl-2 family gene expression in hippocampus of 3, 4-methylenedioxymethamphetamine treated rats Cell J 13 275 280 23508090 Sourbron J Schneider H Kecskes A Liu Y Buening EM Lagae L Smolders I de Witte P 2016 Serotonergic modulation as effective treatment for dravet syndrome in a zebrafish mutant model ACS Chem Neurosci 7 588 598 26822114 Stafstrom CE 2010 Pathophysiological Mechanisms of Seizures and Epilepsy: A Primer Rho JM Sankar R Stafstrom CE Epilepsy: Mechanisms, models, and translational perspectives CRC Press Boca Raton, FL 3 19 Tamburini I Blandini F Gesi M Frenzilli G Nigro M Giusiani M Paparelli A Fornai F 2006 MDMA induces caspase-3 activation in the limbic system but not in striatum Ann N Y Acad Sci 1074 377 381 17105935 Verma R Mishra V Sasmal D Raghubir R 2010 Pharmacological evaluation of glutamate transporter 1 (GLT-1) mediated neuroprotection following cerebral ischemia/reperfusion injury Eur J Pharmacol 638 65 71 20423712 Wang X Zhu SP Kuang WH Li J Sun X Huang MS Sun XL 2009 Neuron apoptosis induced by 3,4-methylenedioxy methamphetamine and expression of apoptosis-related factors in rat brain Sichuan Da Xue Xue Bao Yi Xue Ban 40 1000 2 1037 20067106 Weber GF Johnson BN Yamamoto BK Gudelsky GA 2014 Effects of stress and MDMA on hippocampal gene expression Biomed Res Int 2014 141396 24511526 West MJ Slomianka L Gundersen HJ 1991 Unbiased stereological estimation of the total number of neurons in thesubdivisions of the rat hippocampus using the optical fractionator Anat Rec 231 482 497 1793176
PMC005xxxxxx/PMC5123908.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101084138 22395 Infect Genet Evol Infect. Genet. Evol. Infection, genetics and evolution : journal of molecular epidemiology and evolutionary genetics in infectious diseases 1567-1348 1567-7257 27693401 5123908 10.1016/j.meegid.2016.09.021 NIHMS820600 Article Microgeographically diverse Plasmodium vivax populations at the Thai-Myanmar border Gupta Bhavna a Parker Daniel M. ab Fan Qi c Reddy B.P. Niranjan a Yan Guiyun d Sattabongkot Jetsumon e Cui Liwang a* a Department of Entomology, Pennsylvania State University, University Park, PA 16802, USA b Shoklo Malaria Research Unit, Mahidol-Oxford Tropic al Medicine Research Unit, Faculty of Tropical Medicine, Mahidol University, Mae Sot, Tak, Thailand c Dalian Institute of Biotechnology, Dalian, Liaoning Province, China d Program in Public Health, University of California Irvine, Irvine CA, USA e Mahidol Vivax Research Unit, Faculty of Tropical Medicine, Mahidol University, Bangkok, 10400 Thailand BG: [email protected] DP: [email protected] QF: [email protected] BR: [email protected] GY: [email protected] JS: [email protected] * Corresponding author, [email protected] 12 10 2016 28 9 2016 11 2016 01 11 2017 45 341346 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Malaria transmission along international borders of the Greater Mekong Subregion is a big challenge for regional malaria elimination. At the Thai-Myanmar border, Plasmodium falciparum cases have dropped dramatically; however, increasing P. vivax prevalence and the emerging reports on hidden malaria burden due to asymptomatic infections demand attention. We conducted cross-sectional surveys to detect asymptomatic malaria infections in a small village located at Thai-Myanmar border and genotyped P. vivax infections in order to understand the level of genetic diversity on such a microgeographic scale. PCR/RFLP and DNA sequencing identified high levels of genetic polymorphisms at both Pvmsp3α and Pvmsp3β loci among P. vivax infections. Combining the PCR/RFLP patterns of Pvmsp3α and Pvmsp3β, a total of 10 genotypes were observed among 17 samples, while concatenated DNA sequences of Pvmsp3α and 3β generated 14 haplotypes with haplotype diversity of 0.97. These markedly diverse parasites on a microgeographic scale suggest the circulation of a considerably large parasite population at the international border. Graphical Abstract Plasmodium vivax Thai-Myanmar Microgeographic Genetic diversity Pvmsp3α Pvmsp3β 1. Introduction Malaria remains a global public health problem, with an estimated 214 million cases and 0.4 million deaths in 2015 (WHO, 2015a). In recent years, malaria has been declining globally; 54% (57/106) malarious countries have reduced the incidence of malaria by >75% between 2000 and 2015 (WHO, 2015a). Countries within the Greater Mekong Subregion (Cambodia, China, Laos, Myanmar, Thailand and Vietnam) are aiming to achieve regional malaria elimination by 2030. Thailand has seen 50-75% reduction in malaria incidence in the last five years (WHO, 2015a). As in other countries of this region, the geographical distribution of malaria is highly heterogeneous (Cui et al., 2012). Central Thailand has been malaria-free for several decades; malaria predominantly occurs along the international borders shared with Myanmar, Cambodia and Malaysia. Provinces bordering Myanmar account for the highest burden of malaria in Thailand, and Tak Province among them, contributes ~70% of the total P. falciparum cases in the country (Wongsrichanalai et al., 2001). Among various possible reasons for this epidemiology in Thailand, importation of malaria cases from Myanmar poses a major challenge for controlling malaria in the border areas. Adding to the concerns of malaria control programs, huge burden of asymptomatic cases has been observed recently in Tak Province (Baum et al., 2015; Parker et al., 2015a; b). The situation is further worsened since the populations living there appear reluctant to seek proper health care (Sonkong et al., 2015). To obtain detailed knowledge of malaria epidemiology in the Thai-Myanmar border, we recently conducted cross-sectional surveys in 2011-2012 in the Tha Song Yang District of Tak Province (Parker et al., 2015b). This study revealed a high level of undiagnosed malaria cases at a microgeographic scale (in a village of ~500 residents and spanning ~321 m north to south and ~500 m east to west). Whereas an expert microscopist identified 34 positive cases (18 P. vivax and 16 P. falciparum), PCR revealed 81 parasite-infected blood samples (55 P. vivax and 20 P. falciparum, 5 mixed, and 1 P. malariae). The infections missed by microscopy have been considered as submicroscopic asymptomatic cases. Noticeably, this study revealed a threefold increase in P. vivax cases identified by PCR, while most of the P. falciparum cases were symptomatic. Similar trends were observed by Baum and coworkers (2015) from another sentinel village in Tak Province. Such microgeographic studies of malaria prevalence can potentially reveal the accurate burden of malaria, which can help present a big picture of the whole region where malaria cases are being under-diagnosed by routine microscopic examinations. Conversely, without an accurate estimation of malaria epidemiology, it would be difficult for the national malaria program to realize the goal of malaria elimination. For the last few years, Thailand's annual malaria incidence has continuously recorded P. vivax as the predominant species in the provinces along the Thai-Myanmar border (Parker et al., 2015b). Analyses in these regions have shown natural P. vivax populations to be highly genetically diverse with frequent multiple-strain infections. Yet, these studies have been conducted in different and distant geographical locations and on relatively large geographical scales (Cui et al., 2003; Rungsihirunrat et al., 2011; Putaporntip et al., 2014). The genetic diversity and transmission dynamics of malaria parasites at a microgeographic scale are much less understood. Thus, we determined the genetic diversity of P. vivax isolates from a small village in Tak Province using Pvmsp3α- and Pvmsp3β-based PCR/RFLP and sequencing techniques. Continuous surveillance of the genetic complexity in regions pursuing malaria elimination could reveal hidden reservoirs of malaria and also provide useful baseline molecular epidemiology data to monitor the effectiveness and progress in malaria elimination efforts. 2. Materials and methods 2.1. Study site and sample collection The study site is a small village (SO) in the Tha Song Yang District of Tak Province on the Thai side of Thai-Myanmar border (Fig. 1). The village spans over ~321 m from east to west and 500 m from north to south on a side of hill. This village has been inhabited by ~550 people with ~80% of residents being of the Karen ethnicity (Parker et al., 2015). Three mass blood surveys were conducted each year in 2012 and 2013. The study population included all the residents of the village present at the time of the surveys, the average population examined during six mass blood surveys was approximately 527. Some villagers could not be included in the surveys as they frequently moved out of the village and even crossed the international border in search of the work. The detailed demographic information of the study site has been published in Parker et al., 2015b. The identification codes for each family and the residents used by Parker and coworkers (Parker et al., 2015b) were maintained in the present study. Finger-pricked blood samples (100 μl each) on Whatman filter paper were obtained from all patients on enrolment and blood smears were prepared and stained with Giemsa. Written informed consent was obtained from the participants or guardians. This study was approved by the Institutional Review Boards of Pennsylvania State University and the Thai Ministry of Health. 2.2. Malaria diagnosis The presence of malaria parasites in all blood samples from participants were confirmed by microscopic examination of Giemsa-stained blood films and by PCR. Thick films were screened for 200 oil-immersion fields. For molecular identification, the DNA was extracted from dried blood spots using a QiaAmp DNA Mini Kit (Qiagen, Germany). Plasmodium species identification was carried out using genus- and species-specific nested PCR as described (Snounou et al., 1993) and samples with mixed infections were omitted from the analysis. The persons with P. vivax infections were visited by public health workers. The patients were advised and treated with antimalarials following the Thai national antimalarial drug policy. 2.3. PCR-RFLP genotyping of P. vivax isolates at the Pvmsp3α and Pvmsp3β loci Our earlier studies using two highly polymorphic genetic markers Pvmsp3α and Pvmsp3β have demonstrated high levels of genetic diversity of the P. vivax populations in western Thailand (Cui et al., 2003; Yang et al., 2006). Allelic diversity of the Pvmsp3α and Pvmsp3β genes was assessed using established PCR-RFLP techniques (Bruce et al., 1999; Yang et al., 2006). Pvmsp3α was amplified by nested PCR using primers P1 (5′-CAGCAGACACCATTTAAGG-3′) and P2 (5′-CCGTTTGTTGATTAGTTGC-3′) for the primary PCR and primers N1 (5′-GACCAGTGTGATACCATTAACC-3′) and N2 (5′- ATACTGGTTCTTCGTCTTCAGG-3′) for the nested PCR. Primary PCR of Pvmsp3β was performed with primers F1 (5′-GTATTCTTCGCAACACTC- 3′) and R1 (5′-CTTCTGATGTTATTTCCAG-3′), while nested reactions were done with primers F2 (5′-CGAGGGGCGAAATTGTAAACC-3′) and R2 (5′-GCTGCTTCTTTTGCAAAGG-3′). PCR was performed for 35 cycles using the following conditions: 94 °C for 20 s, 54 °C for 30 s, and 68 °C for 2.5 min. For RFLP analysis, the PCR products of Pvmsp3α and Pvmsp3β were digested with HhaI and PstI, respectively, and DNA was separated in a 1.5% agarose gel. Alleles were identified based on unique restriction banding patterns and were classified as described by earlier studies (Cui et al., 2003; Yang et al., 2006). If the molecular size of the bands was measured to be within 25 bp, they were considered the same. 2.4. Sequence analysis and phylogenetics Amplified fragments of Pvmsp3α and Pvmsp3β from the P. vivax samples were purified using the High Pure PCR cleanup microkit (Roche) and sequenced in both directions using BigDye Terminator v3.1. DNA sequences obtained were assembled using Lasergene software (DNASTAR) with manual editing, and Pvmsp3α and Pvmsp3β sequences were aligned with the Salvador I (Sal-I) reference gene sequences PVX_097720 and PVX_097680, respectively, using ClustalW. All sequences obtained in this study have been deposited in GenBank with accession numbers KX656749-KX656784. Phylogenetic analysis was conducted using MEGA v6.0 (Tamura et al., 2013).Unrooted trees were generated using the neighbor-joining method with the Jukes-Cantor distance option after applying partial deletion. Support for a node was assessed by 500 bootstrap resampling of the original dataset. The N- and C-termini of the Pvmsp3β sequence were determined as in Rayner et al. (2004) on the basis of the Sal-I strain sequence (PVX_097680). 3. Results 3.1. Detection of Plasmodium vivax using microscope and nested PCR method To determine the extent of P. vivax genetic diversity and transmission dynamics at a microgeographic scale along the Thai-Myanmar border, we analyzed parasite samples from mass blood surveys conducted during 2012 and 2013 in a small border village in western Thailand (Fig. 1). Our earlier molecular analysis identified a large number of asymptomatic infections of P. vivax infections in the study village residential population (Parker et al., 2015a). In the three cross sectional surveys of 2012, 10 single-species P. vivax infections were identified by PCR analysis using the 18S rRNA gene (4 in April, 2 in July-August, and 4 in October-December), while in the two surveys of 2013, 14 single-species P. vivax infections were identified (5 in March, and 9 in May - June) (Table 1). However, only six isolates were identified as P. vivax infections by microscopy. Of the total 24 P. vivax infections, two cases were probably recurrent as they occurred after three and five months of the first infections, respectively (Table 1). Twelve of the infected residents were males and ten were females, and they were grouped into three age groups: 0-5 years (children), 6-17 years (students) and 18 and above (adults) (Table1). Eight of the infections were detected each in the children and students groups, and six infections were observed among adults. All the adults were recorded as farmers who moved out of the village and crossed the international border frequently. Whereas age and sex were not correlated with the infections, 21 of the 22 infected persons were migrants or without Thai citizenship, indicating significant association of malaria infections cross-border migration. Infected individuals were from 16 families. Multiple members of the four families were found infected either at the same time or at different time points during the surveys (Table 1), leading to statistically significant clustering of the cases (Parker et al., 2015b). 3.2. Genetic diversity of the Pvmsp3α gene Pvmsp3α gene was successfully amplified in 21 of the 24 P. vivax isolates (Fig. 2A). Two PCR length variant types A (~1.9 Kb) and C (~1.1 Kb) were observed, whereas type B (1.5 Kb) was not observed in the tested samples (Cui et al., 2003) (Fig. 2A). Type A was observed in nine isolates, while type C was found in ten isolates. Two samples produced two bands, indicating mixed-strain infections (Fig. 2A). Pvmsp3α fragments from the 19 isolates were subjected to restriction digestion with HhaI. Totally three RFLP patterns (two in type A and one in C) were observed, and one ~950 bp fragment was present in all 20 isolates (Fig. 2A). In silico mapping of HhaI sites in Pvmsp3α of the Sal-I strain using RestrictionMapper (restrictionmapper.org) could not identify restriction site in the 1123 bp fragment from position 1912 to 3034, suggesting that this region may correspond to the ~950 bp fragment observed in all HhaI digested PCR products. Sequencing of single PCR bands from the 19 isolates generated fragments of 1072 - 1856 bp. The isolates with type C contained a ~750 bp deletion when aligned with type A isolates and the ~810 bp region from position 2009 to 2818 (corresponding to the Sal-I PVX_097720) was found conserved in all 19 isolates. A total of 35 single nucleotide polymorphisms (SNPs) were found among the 19 sequences, resulting in nine haplotypes with haplotype diversity of 0.77. 3.3. Genetic diversity of the Pvmsp3β gene PCR amplification of Pvmsp3β was successful in 22 samples, of which five samples showed mixed-strain infections (Fig. 2B). The remaining 17 samples generated three genotypic variants based on the size of amplified fragments (Fig. 2B). As previously described in the isolates from Tak Province by Yang et al. (2006), type A (1.7-2.2 Kb) and type B (1.4-1.5 Kb) were observed in the present study, however, type C (0.65 Kb) was not found (Fig. 2B). We observed some fragments in the range of 1.0 to 1.3 Kb that we have named as type C in the present study (Fig. 2B). Type A was observed in one sample, type B in nine samples and type C was observed in seven samples (Fig. 2B). Restriction digestion of PCR amplified products with PstI generated seven banding patterns including one in A, and four in B and two in C types (Fig. 2B) Seventeen isolates could be successfully sequenced ranging from 906 bp to 1866 bp. A large number of indels and polymorphic sites were observed in the Pvmsp3β sequences. However, the 752 bp in the N-terminal (143 to 894 bp) and 501 bp C-terminal portions (1890 to 2390 bp) were relatively less polymorphic. N-terminal region was obtained from 17 isolates, while C-terminal portion could be obtained from only 13 isolates. Thus, only N-terminal region was used for further analysis. Excluding gaps due to indels, a total of 175 SNPs were observed in the N-terminal that produced 13 haplotypes with haplotype diversity of 0.968. 3.4. Spread of P. vivax infections based on Pvmsp3α and Pvmsp3β genetic and phylogenetic relationships A high level of diversity was observed among P. vivax samples. Among two cases of possible recurrent infections, distinct genotypes were observed in one case, while in the other case no amplification was observed at both Pvmsp3α and Pvmsp3β loci. Combining both Pvmsp3α and Pvmsp3β RFLP patterns, a total of 10 genotypes were observed among 17 samples. However, when the Pvmsp3α and 3β sequences were concatenated in the 17 isolates (excluding four samples with mixed-strain infections), 14 haplotypes were obtained, giving haplotype diversity of 0.97. Phylogenetic analysis of the 19 Pvmsp3α sequences revealed three major clusters (Fig. 3A). Most of the infections detected in the same or the different members of the family were found in different clusters. Noticeably, the majority of the A and B RFLP-based subtypes formed separate groups (Fig. 2A). In contrast, the phylogenetic tree of Pvmsp3β showed even higher genetic diversity with multiple branches, and only three samples of the B subtypes were found clustered together in a group (Fig. 2B). We also constructed sequence relationship tree of 19 Pvmsp3α sequences with 158 published Pvmsp3α sequences from different geographical locations in Asia and America (18 from Thailand, 24 from Myanmar, 22 from South Korea, 17 from Sri Lanka, 4 from India, 33 from Peruvian Amazon, 21 from Venezuela, 6 from Brazil, 5 from Greece, Sal-I strain, Chesson strain, and one each from Pakistan, North Korea, Papua New Guinea, Bangladesh, Ong olé and Ecuador). All the sequences clustered into four distinct groups (Fig. S1). One group contained ten parasite samples from the present study along with the Chesson strain and one isolate from Thailand. Two other groups contained isolates from this study (SO) along with mixed strains from all the Asian and American countries, while the fourth group contained only South Korean isolates (Fig. S1). Nine SO isolates found in the first group were 100% identical to the Chesson strain. In the third group, one SO isolate showed 100% similarity with sequences from Myanmar and Thailand, while another SO isolate was identical to a Myanmar isolate. Similarly, a Pvmsp3β phylogenetic tree was constructed using 17 sequences from the present study and 58 published sequences including 32 from Thailand, 4 from China, 2 India, 2 Pakistan, 8 Brazil, 2 Ecuador, Sal-I, Chesson, one each from Papua New Guinea, Bangladesh, North Korea, South Korea, Ong olé and Sri Lanka (Fig. S2). No clear clustering pattern was observed as the tree showed many branches. The majority of the SO samples were found clustered with Thailand, Brazil, South Korea isolates and the Chesson strain (Fig. 2B), while only one SO isolate was identical to a previously published isolate from Thailand. 4. Discussion P. vivax is gaining increased attention due to its unique biology and changing epidemiology, which poses a big challenge to malaria control and elimination. In areas where both P. vivax and P. falciparum coexist, while P. falciparum has been found responsive to the control measures (especially artemisinin-based combination therapies), P. vivax incidences have significantly increased over the years (WHO, 2015b). A similar trend of increasing P. vivax/P. falciparum ratio has been observed in the Thai side of the Thai-Myanmar border in the past few years (Parker et al., 2015b). The high prevalence of P. vivax in this area might be the consequence of i) appearance of P. vivax resistance to chloroquine (Muhamad et al., 2011; Bhumiratana et al., 2013), ii) importation of infections from Myanmar, iii) the large number of asymptomatic cases that serve as the reservoir of transmission (Baum et al., 2015; Parker et al., 2015b), and iv) occurrence of relapsing cases. To develop effective P. vivax control measures that meet the goal of malaria elimination in this region, an in-depth understanding of genetic structure and transmission dynamics of these parasite species is highly needed. In the present study, we determined the genetic diversity of P. vivax isolates among asymptomatic infections in a malaria endemic village in Tak Province by genotyping two highly polymorphic merozoite surface protein genes Pvmsp3α and Pvmsp3β. While earlier studies of P. vivax clinical isolates from acute cases in this region have identified highly diverse P. vivax populations (Cui et al., 2003), the genetic diversity of asymptomatic infections, especially at a microgeographic scale, has not been characterized. In a small border village of western Thailand, we found that P. vivax isolates were highly diverse, showing 14 distinct genotypes from 17 isolates with haplotype diversity of 0.97 (combining both Pvmsp3α and Pvmsp3β). This is comparable to the high level of diversity (haplotype diversity of 1.0) recently reported from a neighboring district of the same province, where samples were collected from acute malaria cases from a larger geographical area (Putaporntip et al., 2014). As found earlier, there is significant clustering of malaria cases in the village (Parker et al., 2015b), and there were families with members being infected at the same time or in different times. Despite this, the infections recorded from the same or different members of the family were distinct genotypes, further adding to the genetic diversity pool of the parasites. Since malaria infections were significantly associated with people with no citizenship suggesting of migrant status, it is likely that this high level of genetic diversity in the P. vivax isolates might be the consequence of frequent migration of the villagers who acquired infections outside of the village (Parker et al., 2015b). Whereas different parasite genotypes occurring in the same individuals indicated the result of new infections, that also could be due to relapses from heterologous hypnozoites of different genotypes (Chen et al., 2007; Imwong et al., 2007). In this study, at least 20% of the asymptomatic infections showed multiple-strain infections, similar to the previous observations from the Tak Province (Cui et al., 2003; Putaporntip et al., 2014). If these were infectious reservoirs sustaining local malaria transmission, they could generate recombinant genotypes as earlier reported in P. falciparum populations from Tanzania (Tanabe et al., 2007), thus further increasing genetic diversity of the parasite populations. Altogether, the high prevalence of asymptomatic infections and highly diverse parasite populations suggest a relative large effective population size of the parasites, which needs to be considered for the design and deployment of more effective control measures. Both Pvmsp3α and Pvmsp3β used in this study are the most polymorphic genetic markers and PCR-RFLP technique based on these two genes have been widely used to identify multiple strains of natural P. vivax infections. However, according to a recent analysis, PCR-RFLP patterns of Pvmsp3α revealed less diversity as compared to that shown by sequencing (Rice et al., 2013). Similar observations were later observed in Pvmsp3β by different studies (Putaporntip et al., 2014; Li et al., 2015). This pattern has been further confirmed in this study, as we observed much higher diversity from the sequencing data as compared to RFLP patterns in both Pvmsp3α and Pvmsp3β genes. However, some of the identical haplotypes in both Pvmsp3α and Pvmsp3β had RFLP patterns belonging to the same genotypes (Figs. S1 & S2). Moreover, the majority of the isolates clustered together in the phylogenetic tree of Pvmsp3α shared common RFLP patterns, suggesting some levels of congruence between the two techniques. In corroboration with the earlier observations, Pvmsp3β was found more diverse with a large number of inserts and mutations (Rayner et al., 2004; Rungsihirunrat et al., 2011) as compared to Pvmsp3α. Moreover, a band size different from the ones previously reported for Pvmsp3b PCR amplification by Yang and coworkers (2006) was observed in this study, which may be related to the high transmission intensity in this region, or may have been generated from recombination events between asymptomatic, multiple clone infections. In addition to the higher level of genetic diversity in Pvmsp3β, this gene detected a higher number of mixed-strain infections than Pvmsp3α. The reasons for the higher diversity in Pvmsp3β are not known, which might be due to different selection pressures acting on these two genes (Gupta et al., 2015). However, these distinct observations between two genes suggested that using multiple loci for genetic diversity analysis may increase the resolution of genotypic characterization and eventually separate the common haplotype into two distinct haplotypes. Comparison of the Pvmsp3α and Pvmsp3β sequences with the published sequences from different countries, the SO isolates were clustered with the isolates from different countries. This might be the consequence of global distribution of the genotypes due to a high level of gene flow between different geographical regions. Interestingly, some of the SO isolates showed 100% similarity with the previously published sequences especially from Thailand and Myanmar. This implies the impact of frequent human migrations across the border between Thailand and Myanmar on malaria epidemiology. In conclusion, genotyping of asymptomatic P. vivax infections collected from a small village located at the Thai-Myanmar border showed markedly diverse parasite populations. This finding is highly encouraging, and such micro-geographically focused studies are needed to get an accurate estimate of the potential burden and the extent of genetic diversity of malaria parasites. The genetic diversity information obtained may also provide a valuable baseline to compare the extent and patterns of P. vivax diversity in later years, which could help assess the impact of ongoing malaria interventions on the P. vivax burden in that region. Therefore, continuous malaria genetic monitoring in the regions that are under malaria elimination phase should be an essential component of malaria control efforts. Supplementary Material 1 2 Acknowledgements This research was supported by National Institute of Allergy and Infectious Diseases, National Institutes of Health (U19AI089672). Fig. 1 Maps showing location of the study area. (A) Map of Thailand showing area of sample collection in Tak Province. (B) Satellite image of the SO village in Tak (wikimapia.org). (C) Village map created using spatial coordinates. Dots represent each house and the numbers written are the house numbers having P. vivax-positive patients and the figures in the parenthesis are the number of P. vivax infected samples detected from each house during these surveys. Fig. 2 PCR-RFLP typing of the P vivax isolates using (A) Pvmsp3α and (B) Pvmsp3β gene. Upper panel, undigested PCR products Lower panel, Pvmsp3α PCR products digested with Hha I and Pvmsp3β with PstI. Lane M is 2 log DNA marker. Sample names on each lane showing house number and the member of the family and the numbers in the parenthesis are showing the P. vivax infections that occurred twice in the same individual during the study period. Mixed infections are shown with asterisk on the sample names and samples with no amplification are circled. RFLP-based genotype of each sample is also shown in lower panels. Fig. 3 Phylogenetic relationships among the SO P. vivax isolates using (A) 19 sequences of Pvmsp3α and (B) 17 sequences of Pvmsp3β gene. Unrooted phylogenetic trees were constructed using neighbor-joining method in MEGA 6.0. Bootstrap support (>75%) is shown as the percentage from 500 pseudo replications. Corresponding PCR/RFLP genotype of each sample has been shown on both trees. The identical haplotypes sharing the same RFLP genotypes within a cluster are encircled. Table 1 Details of P. vivax infected persons from Suan Oi village in Tak province, their date of sampling, age, sex, occupation and nationality. S.No. Date of sampling (Month/Year) Identification code (Family-Member) Age (Years) Sex Occupation Ethnicity/Nationality 1 03/2013 64-5 4 F Child Foreigner 2 03/2013 64-6 1 F Child 3 08/2012 64-4 20 M Farmer 4 03/2013 15-3(1) 16 M Student Foreigner 5 04/2012 15-7 4 M Child 6 04/2012 15-4 14 M Student 7 06/2013 15-3(2) 16 M Student 8 03/2013 22-4 8 M Student Foreigner 9 05/2013 22-5 2 F Child 10 06/2013 86-5(1) 1 F Child Foreigner 11 11/2012 86-5(2) 1 F Child 12 04/2012 108-4 10 M Student Foreigner 13 04/2012 108-6 6 F Student 14 03/2013 74-4 30 F Farmer Foreigner 15 10/2012 76-6 7 M Student Foreigner 16 07/2012 83-7 3 F Child Foreigner 17 05/2013 89-2 36 F Farmer Foreigner 18 12/2012 36-6 4 M Child Foreigner 19 06/2013 45-5 10 M Student Foreigner 20 05/2013 41-2 49 F Farmer Foreigner 21 06/2013 58-3 23 M Farmer Thai Citizen 22 06/2013 68-4 5 M Child Foreigner 23 12/2012 9-4 7 F Student Foreigner 24 06/2013 14-2 19 M Farmer Foreigner • The cases of recurrent infection are shown in bold. • F- Female, M- Male Highlights Cross-sectional survey was conducted in a small village at Thai-Myanmar border Asymptomatic P. vivax infections were genotyped using PvMSP3α and 3β genes Data revealed highly diverse parasite populations at microgeographic scale Genetic diversity among P. vivax isolates was evaluated at a microgeographic scale Cross sectional surveys identified large number of asymptomatic P. vivax infections PCR/RFLP and DNA sequence polymorphisms in Pvmsp3α and Pvmsp3β gene were evaluated High level of genetic polymorphism was observed at both loci This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. References Baum E Sattabongkot J Sirichaisinthop J Kiattibutr K Davies D Jain A Lo E Lee M-C Randall AZ Molina DM Liang X Cui L Felgner PL Yan G Submicroscopic and asymptomatic Plasmodium falciparum and Plasmodium vivax infections are common in western Thailand - molecular and serological evidence. Malar. J 2015 14 95 doi:10.1186/s12936-015-0611-9 25849211 Bhumiratana A Intarapuk A Sorosjinda-nunthawarasilp P Maneekan P Koyadun S Border malaria associated with multidrug resistance on Thailand-Myanmar and Thailand-Cambodia borders : transmission dynamic , vulnerability , and surveillance. Biomed Res. Int 2013 2013 13 Bruce MC Galinski MR Barnwell JW Snounou G Day KP Polymorphism at the merozoite surface protein-3alpha locus of Plasmodium vivax: global and local diversity. Am. J. Trop. Med. Hyg 1999 61 518 525 10548283 Chen N Auliff A Rieckmann K Gatton M Cheng Q Relapses of Plasmodium vivax infection result from clonal hypnozoites activated at predetermined intervals. J. Infect. Dis 2007 195 934 41 doi:10.1086/512242 17330782 Cui L Mascorri CN Fan Q Rzomp KA Khuntirat B Zhou G Chen H Yan G Sattabongkot J Genetic diversity and multiple infections of Plasmodium vivax malaria in Western Thailand. Am. J. Trop. Med. Hyg 2003 68 613 619 12812356 Cui L Yan G Sattabongkot J Cao Y Chen B Malaria in the Greater Mekong Subregion: heterogeneity and complexity. Acta Trop 2012 121 227 239 doi:10.1016/j.actatropica.2011.02.016. Malaria 21382335 Gupta B Reddy BPN Fan Q Yan G Sirichaisinthop J Sattabongkot J Escalante A. a Cui L Molecular evolution of PvMSP3α Block II in Plasmodium vivax from diverse geographic origins. PLoS One 2015 10 e0135396 doi:10.1371/journal.pone.0135396 26266539 Imwong M Snounou G Pukrittayakamee S Tanomsing N Kim JR Nandy A Guthmann J Nosten F Carlton J Looareesuwan S Nair S Sudimack D Day NPJ Anderson TJC White NJ Relapses of Plasmodium vivax infection usually result from activation of heterologous hypnozoites. J. Infect. Dis 2007 195 927 933 doi:10.1086/512241 17330781 Li Y-C Wang G Meng F Zeng W He C Hu X Wang S Genetic diversity of Plasmodium vivax population before elimination of malaria in Hainan province. Malar. J 2015 14 78 doi:10.1186/s12936-015-0545-2 25888891 Muhamad P Ruengweerayut R Chacharoenkul W Rungsihirunrat K Na-Bangchang K Monitoring of clinical efficacy and in vitro sensitivity of Plasmodium vivax to chloroquine in area along Thai Myanmar border during 2009-2010. Malar. J 2011 10 44 doi:10.1186/1475-2875-10-44 21324161 Parker DM Carrara VI Pukrittayakamee S McGready R Nosten FH Malaria ecology along the Thailand–Myanmar border. Malar. J 2015a 14 388 doi:10.1186/s12936-015-0921-y 26437860 Parker DM Matthews SA Yan G Zhou G Lee M-C Sirichaisinthop J Kiattibutr K Fan Q Li P Sattabongkot J Cui L Microgeography and molecular epidemiology of malaria at the Thailand-Myanmar border in the malaria pre-elimination phase. Malar. J 2015b 14 198 doi:10.1186/s12936-015-0712-5 25962514 Putaporntip C Miao J Kuamsab N Sattabongkot J Sirichaisinthop J Jongwutiwes S Cui L The Plasmodium vivax merozoite surface protein 3β sequence reveals contrasting parasite populations in Southern and Northwestern Thailand. PLoS Negl. Trop. Dis 2014 8 e3336 doi:10.1371/journal.pntd.0003336 25412166 Rayner JC Huber CS Feldman D Ingravallo P Galinski MR Barnwell JW Plasmodium vivax merozoite surface protein PvMSP-3β is radically polymorphic through mutation and large insertions and deletions. Infect. Genet. Evol 2004 4 309 319 doi:10.1016/j.meegid.2004.03.003 15374528 Rice BL Acosta MM Pacheco MA Escalante A. a Merozoite surface protein-3 alpha as a genetic marker for epidemiologic studies in Plasmodium vivax: a cautionary note. Malar. J 2013 12 288 doi:10.1186/1475-2875-12-288 23964962 Rungsihirunrat K Chaijaroenkul W Siripoon N Seugorn A Na-Bangchang K Genotyping of polymorphic marker (MSP3α and MSP3β) genes of Plasmodium vivax field isolates from malaria endemic of Thailand. Trop. Med. Int. Heal 2011 16 794 801 doi:10.1111/j.1365-3156.2011.02771.x Snounou G Viriyakosol S Jarra W Pinheiro L do Rosario VE Thaithong S Brown KN High sensitivity of detection of human malaria parasites by the use of nested polymerase chain reaction. Mol. Biochem. Parasitol 1993 61 315 320 doi:10.1016/0166-6851(93)90077-B 8264734 Sonkong K Chaiklieng S Neave P Suggaravetsiri P Factors affecting delay in seeking treatment among malaria patients along Thailand-Myanmar border in Tak Province, Thailand. Malar. J 2015 14 1 8 doi:10.1186/1475-2875-14-3 25557741 Tamura K Stecher G Peterson D Filipski a. Kumar S MEGA6: Molecular Evolutionary Genetics Analysis Version 6.0. Mol. Biol. Evol 2013 30 2725 2729 doi:10.1093/molbev/mst197 24132122 Tanabe K Sakihama N Rooth I Björkman A Färnert A High frequency of recombination-driven allelic diversity and temporal variation of Plasmodium falciparum msp1 in Tanzania. Am. J. Trop. Med. Hyg 2007 76 1037 45 17556608 Wongsrichanalai C Sirichaisinthop J Karwacki JJ Congpuong K Miller RS Pang L Thimasarn K Drug resistant malaria on the Thai-Myanmar and Thai-Cambodian borders. Southeast Asian J. Trop. Med. Public Health 2001 32 41 49 11485094 World Health Organization World Malaria Report 2015a World Health Organization Control and elimination of Plasmodium vivax malaria: a technical brief 2015b World Health Organization Yang Z Miao J Huang Y Li X Putaporntip C Jongwutiwes S Gao Q Udomsangpetch R Sattabongkot J Cui L Genetic structures of geographically distinct Plasmodium vivax populations assessed by PCR/RFLP analysis of the merozoite surface protein 3α gene. Acta Trop 2006 100 205 212 17129568
PMC005xxxxxx/PMC5123919.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 0323470 5428 Kidney Int Kidney Int. Kidney international 0085-2538 1523-1755 27591083 5123919 10.1016/j.kint.2016.06.037 NIHMS805379 Article Loss of Zeb2 in mesenchyme-derived nephrons causes primary glomerulocystic disease Rasouly Hila Milo 12 Kumar Sudhir 1 Chan Stefanie 1 Pisarek-Horowitz Anna 1 Sharma Richa 1 Xi Qiongchao J. 3 Nishizaki Yuriko 4 Higashi Yujiro 4 Salant David J. 1 Maas Richard L. 3 Lu Weining 12* 1 Renal Section, Department of Medicine, Boston Medical Center, Boston University School of Medicine, Boston, MA 02118, USA 2 Graduate Program in Genomics and Genetics, Division of Graduate Medical Sciences, Boston University School of Medicine, Boston, MA 02118, USA 3 Genetics Division, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA 20115, USA 4 Institute for Developmental Research, Aichi Human Service Center, Kasugai, Aichi, Japan * Correspondence should be addressed to: Weining Lu, MD, Associate Professor of Medicine, Renal Section, EBRC 538, Boston Medical Center, Boston University School of Medicine, 650 Albany Street, Boston, MA 02118, USA, Tel: 617-414-1770, Fax: 617-638-7326, [email protected] 26 7 2016 31 8 2016 12 2016 01 12 2017 90 6 12621273 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Primary glomerulocystic kidney disease is a special form of renal cystic disorder characterized by Bowman’s space dilatation in the absence of tubular cysts. ZEB2 is a SMAD-interacting transcription factor involved in Mowat-Wilson syndrome, a congenital disorder with an increased risk for kidney anomalies. Here we show that deletion of Zeb2 in mesenchyme-derived nephrons with either Pax2-cre or Six2-cre causes primary glomerulocystic kidney disease without tubular cysts in mice. Glomerulotubular junction analysis revealed many atubular glomeruli in the kidneys of Zeb2 knockout mice, which explains the presence of glomerular cysts in the absence of tubular dilatation. Gene expression analysis showed decreased expression of early proximal tubular markers in the kidneys of Zeb2 knockout mice preceding glomerular cyst formation, suggesting that defects in proximal tubule development during early nephrogenesis contribute to the formation of congenital atubular glomeruli. At the molecular level, Zeb2 deletion caused aberrant expression of Pkd1, Hnf1β, and Glis3, three genes causing glomerular cysts. Thus, Zeb2 regulates the morphogenesis of mesenchyme-derived nephrons and is required for proximal tubule development and glomerulotubular junction formation. Our findings also suggest that ZEB2 might be a novel disease gene in patients with primary glomerular cystic disease. kidney development obstructive nephropathy pediatric nephrology proximal tubule ADPKD INTRODUCTION Cystic kidney disease is a group of genetically heterogeneous disorders that are characterized mainly by renal tubular dilatation.1,2 Glomerulocystic kidney disease (GCKD) is a special form of cystic kidney disease that confines cystic dilatation to the Bowman’s space.3–5 GCKD is defined as two- to three-fold dilatation of Bowman’s space in more than 5% of identifiable glomeruli in the plane of a kidney section, and the glomerular cysts in primary GCKD are mainly localized to the subcapsular region of the kidney.4–6 Glomerular cysts in association with tubular cysts can be found in several syndromes including tuberous sclerosis complex, orofaciodigital syndrome 1, and Meckel-Gruber syndrome.3,7–9 Glomerular cysts can also be found in patients with mutations in the UMOD and HNF1β genes,10,11 and in patients of autosomal dominant polycystic kidney disease with PKD1 mutations, other ciliopathies such as nephronophthisis, multicystic dysplastic kidney or urinary tract obstruction.3 However, the genetic basis of primary glomerulocystic kidney disease without tubular dilatation remains largely unknown. ZEB2 is a zinc finger E-box binding homeobox transcription factor that promotes epithelial-mesenchymal transition via a TGF-β/SMAD/ZEB/miR-200 signaling network.12–14 ZEB2 mutations in human cause Mowat-Wilson syndrome (MWS: OMIM 235730), a congenital disorder characterized by intellectual disability, craniofacial abnormalities, Hirschsprung’s disease, congenital heart defects, and an increased risk for congenital kidney anomalies.15 However, the pathological features of its kidney defects have not been examined and defined. 15,16 ZEB2 is also a known target for microRNA miR-200.17 Recently, it has been shown that upregulation of ZEB2 in Hnf1b knockout mice and downregulation of miR-200 in Dicer knockout mice are associated with glomerular and tubular cysts.18,19 However, no direct link between ZEB2 downregulation and renal cystic disease has been established and the pathological effect of ZEB2 downregulation in early nephrogenesis remains unknown. Here we report that deletion of Zeb2 in early mouse nephrogenesis with either Pax2-cre or Six2-cre resulted in primary glomerulocystic disease without tubular dilatation. We found that loss of Zeb2 caused renal proximal tubule hypotrophy and reduced glomerulotubular junction integrity and maturation, which contributed to the formation of congenital atubular glomeruli leading to glomerulocystic disease. By gene expression analysis, we found that Zeb2 nephron specific knockout kidney had aberrant expression of Pkd1, Hnf1β, and Glis3, three genes associated with glomerular cysts. These results suggest that Zeb2 regulates the morphogenesis of metanephric mesenchyme derived early nephrons and is required for proximal tubule development and normal glomerulotubular junction formation. RESULTS Because aberrant expressions of Zeb2 and miR-200 have been reported in young mice with a cystic kidney phenotype and ZEB2 loss-of-function mutations are associated with an increased risk for renal anomalies in MWS patients,15,18,19 we hypothesized that ZEB2 plays an important role in early kidney development. To test this hypothesis, we analyzed a Zeb2 floxed conditional knockout (cKO) mouse line20 as Zeb2 null mice die at E9.5 before kidney development begins.21 We first crossed the Zeb2 floxed homozygotes (Zeb2flox/flox) with a Pax2-cre+ deleter strain22 that expresses the Cre recombinase in both the metanephric mesenchyme and ureteric bud (Supplemental Figure 1a). After genotyping 96 three-week old weanlings of Zeb2flox/+;Pax2-cre+ heterozygous matings, we did not find any Zeb2flox/flox;Pax2-cre+ homozygotes (Supplemental Figure 1b). We then analyzed newborn mice and E18.5 embryos from timed-pregnant females. Five dead newborn Zeb2flox/flox;Pax2-cre+ homozygotes and 21/55 (38%) E18.5 homozygous embryos were found (Supplemental Figure 1b), suggesting that Zeb2flox/flox;Pax2-cre+ homozygotes die at birth. The E18.5 Zeb2flox/flox;Pax2-cre+ homozygous embryos did not display discernible gross structural defects of the kidney or ureter. However, histological examination of the kidneys from eight E18.5 Zeb2flox/flox;Pax2-cre+ homozygous embryos revealed that 100% of the homozygotes had renal cysts, while none of the seven littermate controls (0%) had a cystic phenotype (Table 1, Figure 1). The renal cysts first appeared at E16.5 as no cysts could be detected at E15.5 (Table 1, Figure 1). The renal cysts were apparently from a glomerular origin as 47% of the glomeruli were cystic in the E16.5 Zeb2 cKO embryos and 30% of the cysts had a visible glomerular tuft (Tables 1 and 2). This phenotype is sufficient for a diagnosis of glomerulocystic kidney disease (GCKD) according to the established criteria.4 To determine whether Zeb2 cKO embryonic kidneys also develop renal tubular cysts, we examined the kidney sections with both proximal tubule specific marker Lotus tetragonolobus lectin (LTL) and distal tubules and collecting duct specific marker Dolichos biflorus agglutinin (DBA). The cysts were negative for both LTL and DBA staining at E18.5 (Figure 1c and 1d), suggesting that the Zeb2flox/flox;Pax2-cre+ homozygous embryos develop primary glomerular cysts. To determine if glomerular cyst formation is due to a deletion of Zeb2 in the developing nephron, we examined ZEB2 expression during kidney development by analyzing a ZEB2-EGFP reporter mouse that has been reported previously.23 In addition to the stromal cells, ZEB2 was detected in a subset of cells in which it can be deleted by Pax2-cre in the developing nephron (Supplemental Figure 2), including the S-shaped bodies (Supplemental Figure 2e) and the glomeruli (Supplemental Figure 2f). By co-staining, we confirmed that ZEB2 is co-expressed in a subset of cells with PAX2 and JAG1, two markers of developing nephrons (Supplemental Figure 3). Zeb2flox/flox;Pax2-cre+ homozygous mice died at birth (probably due to high expression of Pax2-cre in the nervous system where ZEB2 also plays an important role),22,23 precluding a longitudinal analysis of glomerulocystic kidney disease progression after birth. A recent study shows that SIX2 regulates ZEB2 expression through miR-200.24 SIX2 is also a nephron progenitor marker and is expressed in the metanephric mesenchyme that gives rise to all segments of the mature nephron including the parietal and visceral epithelial cells in the glomeruli and the proximal and distal tubular epithelial cells.25 To further delineate the role of ZEB2 in nephron development and to study the pathological effect of Zeb2 deletion in mesenchyme-derived nephrons as well as glomerular cystic phenotype in postnatal mature kidney, we crossed Zeb2flox/flox mice with Six2-cre mice. The Zeb2flox/flox;Six2-cre+ homozygous mice survived after birth and were weaned at a Mendelian distribution. Histological examination of the kidneys from nine Zeb2flox/flox;Six2-cre+ homozygous embryos revealed that, like the Zeb2flox/flox;Pax2-cre+ embryos, all homozygous embryos had glomerular cysts starting at E16.5 (Figure 2a). At postnatal day 12 (P12), the glomerular cysts in the Zeb2flox/flox;Six2-cre+ mice were mainly located in the subcapsular region of the kidney that is consistent with the primary glomerulocystic kidney disease diagnosis (Figure 2b).4 The cysts were also negative for both LTL and DBA staining (Figure 2c and 2d), confirming their glomerular origin. In comparison, none of the five wild-type littermate controls had a cystic phenotype at P12 (Figure 2b). To determine the longitudinal effect of glomerulocystic kidney disease on mature kidney, we followed the Zeb2flox/flox;Six2-cre+ mice to adulthood. At 7 weeks of age, all Zeb2flox/flox;Six2-cre+ mice (n=5, 100%) developed macroscopic renal cysts that were visualized on the surface of the kidney (Figure 3a). Histological analysis revealed the presence of numerous large glomerular cysts in the area of the renal cortex and outer medulla (Figure 3b). Although no significant interstitial fibrosis was observed (Figure 4a), some remaining non-cystic glomeruli in adult Zeb2flox/flox;Six2-cre+ mice displayed glomerulosclerosis lesions (Figure 4b). Similar lesions were also observed in few non-cystic glomeruli in P8 Zeb2 cKO kidneys (Supplemental Figure 4). By 8 weeks of age, all Zeb2flox/flox;Six2-cre+ mice (n=5, 100%) developed albuminuria (Figure 4c and 4d) with significantly elevated serum blood urea nitrogen (BUN) levels (Figure 4e). Albuminuria was not directly caused by loss of ZEB2 in podocytes or reduced proximal tubule endocytosis as deletion of Zeb2 specifically in the podocyte using Nphs2-cre did not lead to proteinuria (data not shown), and the expressions of megalin and cubilin were upregulated in Zeb2 adult cKO kidneys (Supplemental Figure 5), which may increase endocytosis of albumin as previously reported.26 These data suggest that loss of Zeb2 in mesenchyme-derived nephrons alone is sufficient to cause glomerulocystic disease in mature kidneys, which leads to secondary glomerulosclerosis, albuminuria and renal failure in adult mice. One common cause of acquired glomerular cysts is the loss of glomerulotubular junction integrity leading to atubular glomeruli.27,28 Glomerulotubular integrity can be quantified by examining the connection of the proximal tubule to the Bowman’s capsule with the proximal tubule marker LTL.29 To determine whether the Zeb2 knockout mice have decreased glomerulotubular integrity, we examined glomeruli from five P12 Zeb2flox/flox;Six2-cre+ mice and five littermate controls by serial sections. We found that only 63/604 (10%) glomeruli had a visible LTL positive staining (i.e. glomerulotubular junction) in the Zeb2 cKO mice as compared to 115/324 (36%) in the wild-type littermate controls (Figure 5), a statistically significant decrease of glomerulotubular integrity, suggesting that Zeb2 deletion causes formation of atubular glomeruli. Acquired atubular glomeruli can be caused by atrophy of the proximal tubules in polycystic kidney disease.30 To determine if reduced glomerulotubular integrity in Zeb2 cKO is associated with renal proximal tubule atrophy (acquired defect) or hypotrophy (developmental defect), we analyzed the mRNA levels of markers for the proximal tubule (Hnf1a and Vil1), the podocyte (Nphs1 and Nphs2), and the collecting duct (Upk3a).31 We found that only the proximal tubule mRNA markers were significantly reduced in the Zeb2flox/flox;Six2-cre+ mice compared to the levels in the wild-type littermates at postnatal day 8 (P8) (Figure 6a). To determine if the reduction of proximal tubule mRNA at P8 is caused by an early developmental defect preceding the formation of glomerular cysts in Zeb2 cKO mice, we analyzed mRNA markers in E14.5 kidneys. Similar to P8, a significant decrease of the proximal tubular markers was detected in the E14.5 Zeb2flox/flox;Six2-cre+ kidneys as compared to their wild-type littermates (Figure 6b). Consistent with this finding, the expression of the proximal tubule brush border protein villin 1 (VIL1),31 was also downregulated in both E15.5 Zeb2flox/flox;Pax2-cre+ and E16.5 Zeb2flox/flox;Six2-cre+ mutant kidneys (Figure 6c and 6d). Finally, the mean kidney size of the E16.5 Zeb2 cKO embryos was smaller than that of their wild-type littermate controls (Figure 6e). Taken together, these data suggest that loss of Zeb2 in the mesenchyme-derived nephrons causes early proximal tubule developmental defects resulting in tubular hypotrophy, reduced glomerulotubular junction integrity, and congenital atubular glomeruli formation. Glomerular cysts are reported in several mouse models of renal cystic kidney disease, including Wwtr1, Glis3, Ofd1 and Pkhd1 knockout mice, the Hnf1β and Dicer cKO mice, and Pkd1 over expression transgenic mice.19,31–36 Interestingly, the Hnf1β cKO mice using Six2-cre develop glomerular cysts due to a drastic reduction in the levels of proximal tubular markers at E14.5, resembling the Zeb2 cKO phenotype.31 However, by immunostaining of JAG1, a marker for the renal vesicle and the S-shaped body,31 we did not observe differences between Zeb2 cKO and wild-type littermate controls (Supplemental Figure 6). To determine if loss of Zeb2 affects the expression of these six genes and the microRNA miR-200, we examined mRNA and microRNA levels in the kidney tissues of E14.5 and E18.5 Zeb2 cKO and wild-type controls. Although there was no significant difference in the expression levels of any of the six genes and miR-200 at E14.5 before glomerular cyst formation (Figure 7a), we detected a decreased expression of Glis3 and increased expression levels of Hnf1β and Pkd1 in E18.5 Zeb2 cKO compared to the wild-type littermate controls (Figure 7b). The expression level of Pkd1 mRNA was the most significantly upregulated in the Zeb2 cKO kidneys at postnatal day 8 (P8) (Figure 7c). Consistent with the mRNA levels, PKD1 coding protein polycystin-1 (PC1) was also found to be expressed at a higher level in the glomeruli of postnatal day 7 (P7) Zeb2flox/flox;Six2-cre+ mice but not at E16.5 and E17.5 when the glomerular cysts are initially observed (Figure 7d). Interestingly, the PC1 expression was upregulated in non-cystic glomeruli but not in the glomeruli with dilated Bowman’s space (Figure 7e). These data suggest that loss of Zeb2 in the kidney leads to upregulation of polycystin-1 expression in non-cystic glomeruli after the initial phase of glomerular cyst formation. Renal cystogenesis is often associated with increased cell proliferation.37 To determine if abnormal cell proliferation also plays a role in the formation of glomerular cysts in Zeb2 cKO mice, we quantified the proliferation of parietal epithelial cells in the Bowman’s capsule of Zeb2 cKO kidneys and wild-type littermate controls using the cell proliferation marker phospho-Histone H3 (pHH3). No significant difference of cell proliferation was observed in Zeb2 cKO kidneys compared to wild-type littermate controls (Supplemental Figure 7), suggesting that cell proliferation does not play an important role in Zeb2 cKO glomerulocystic phenotype. Apoptosis is part of normal kidney development during C-shaped body (CSB) and S-shaped body (SSB) formation.38,39 To determine if abnormal apoptosis in CSB and SSB may contribute to congenital atubular glomeruli formation in Zeb2 cKO mice, we performed TUNEL assays on E15.5 and E16.5 developing kidneys from five Zeb2 cKO and four littermate controls. Interestingly, 11/34 (32%) CSB/SSB were identified with at least one apoptotic cell in Zeb2 cKO kidneys, while 25/39 (64%) CSB/SSB were found in wild-type littermate controls (Supplemental Figure 8). These data suggest that loss of Zeb2 causes aberrant apoptosis in CSB and SSB during nephron development, which may contribute to proximal tubular hypotrophy and abnormal glomerulotubular junction maturation in Zeb2 cKO mice. DISCUSSION In this study, we found that ZEB2 is critical for the mesenchyme-derived nephron development. Conditional deletion of Zeb2 with Pax2-cre and Six2-cre, two Cre strains active in the metanephric mesenchyme derived nephron progenitor cells, resulted in the same glomerular cyst phenotype, indicating that ZEB2 regulates the morphogenesis of mesenchyme-derived nephrons and is required for normal nephron development. Decreased levels of proximal tubular markers at both mRNA and protein levels from E14.5 to E16.5 and reduced kidney size at E16.5 in the Zeb2 cKO mice suggest an early hypotrophy of the proximal tubule, leading to congenital atubular glomeruli formation, accumulation of glomerular filtrate and cystic expansion of the Bowman’s space. Although ZEB2 is also expressed in the stromal cells of the developing kidney, stromal cell-specific deletion of Zeb2 with Foxd1-cre did not result in a glomerular cystic phenotype (data not shown), which further supports the specific role of ZEB2 in early developing nephron. Consistent with our findings, haploinsufficiency of Zeb2 was also reported to be associated with delayed nephrogenesis in a transgenic rat model.40 Atubular glomeruli have been reported in polycystic kidney disease, obstructive uropathy, and many other glomerular or tubular diseases in humans and animal models.27,29,30,41 However, they are thought to be secondary to progressive injury to the proximal tubules inducing degenerative tubular cell changes and atrophy, eventually destroying proximal tubules at the glomerulotubular junctions.30,42 We found that loss of Zeb2 is associated with primary congenital atubular glomeruli due to developmental defects and hypotrophy of the proximal tubules and the glomerulotubular junction. Although the link between acquired atubular glomeruli and glomerular cyst formation was proposed two decades ago,28 our study now demonstrates that primary glomerulocystic disease can also be caused by congenital atubular glomeruli. ZEB2 is also known as SIP1 (SMAD-interacting protein 1), which interacts with activated SMAD proteins and is part of the TGF-β/SMAD signaling pathway.43,44 SMADs are highly expressed in developing kidney.45 Elevated TGF-β/SMAD signaling has also been detected in experimental Pkd1 mouse models and human patients with polycystic kidney disease.46 We found abnormal expression of Glis3 gene in Zeb2 cKO kidneys at E18.5 compared to their wild type littermates. Transcription factor GLIS3 interacts with TAZ protein (encoded by Wwtr1 gene), which is also part of the TGF-β/SMAD pathway regulating SMAD shuttling between the cytoplasm and the nucleus.34,47 Interestingly, both Glis3 and Wwtr1 knockout mice develop glomerular cysts as Zeb2 cKO mice.34,36 In our study, we also found abnormal expression of Hnf1β in the Zeb2 cKO kidneys at E18.5 compared to the wild-type littermate controls. HNF1B mutations cause glomerulocystic disease in human and Hnf1β mesenchyme specific cKO mice also develops glomerular cysts.11,31 A recent study shows that the expressions of both Zeb2 and Pkd1 are upregulated in another Hnf1β renal specific knockout mouse model with polycystic kidney disease,18 suggesting that an optimal and balanced ZEB2 expression in the kidney is required to prevent renal cysts formation. High levels of polycystin-1 are often detected in the kidney of ADPKD patients.48–52 Overexpression of polycystin-1 in mice also causes glomerular cysts.32,53 Hnf1β cKO and Dicer cKO mice develop renal cysts with increased levels of Pkd1 and decreased levels of miR-200.18,19 We found increased expression of Pkd1 mRNA and polycystin-1 in the glomeruli of Zeb2 cKO kidneys after the initial phase of glomerular cyst formation. Although miR-200 expression is repressed by ZEB254 and Zeb2 gene is a known target of miR-20055 in a double-negative feedback loop, we did not observe differential expression of miR-200 in Zeb2 cKO kidneys. Likewise, no differential expression of Zeb2 was observed in the Dicer cKO mice in which miR-200 is significantly downregulated.19 Therefore, Pkd1 upregulation in the non-cystic glomeruli of Zeb2 cKO mice is probably secondary to the glomerular stress (e.g. due to hyperfiltration of the non-cystic glomeruli) and not the result of a direct regulation of Pkd1 gene expression by the ZEB2/miR-200 signaling network. The effects of this increased expression of polycystin-1 in Zeb2 cKO mice are presently unknown and need further investigation. In conclusion, by studying animal models of a disease gene causing a human syndrome associated with an increased risk of congenital kidney anomalies, we identified Zeb2 as a novel gene important in proximal tubule development and glomerulotubular junction formation. Loss of Zeb2 in mesenchyme-derived nephrons in mice results in reduced glomerulotubular integrity, congenital atubular glomeruli and primary glomerular cystic disease. Future studies are needed to elucidate the molecular pathway of ZEB2 signaling during kidney development and the cell type and mechanism of Pkd1 overexpression in non-cystic glomeruli, and to determine whether MWS patients also develop glomerular cysts and whether a subset of patients with glomerulocystic kidney disease carry ZEB2 mutations. As a transcription factor, ZEB2 may also provide a starting point for further identification of new genes important in glomerulotubular junction development and that, when mutated, may cause primary glomerulocystic kidney disease in patients. MATERIALS AND METHODS Animals Zeb2 floxed conditional knockout mouse and ZEB2-EGFP reporter mouse were previously reported.20, 23 Pax2-cre+ mice were obtained from the MMRRC (#010569-UNC)22 and Six2-cre+ mice were purchased from the Jackson Lab (#009606).25 All animal studies were approved by the IACUC of Boston University. Histology The kidneys of mice at defined ages were dissected and fixed in 4% PFA and processed for paraffin embedding. Serial sections were cut and stained by H&E, Periodic acid–Schiff stain and Masson Trichrome stain using standard methods. Slides were viewed with an Olympus microscope and photographed using a DP72 digital camera. Quantification of glomerular cysts and kidney length The numbers of glomeruli with and without cysts were counted on an H&E-stained median sagittal kidney section from each animal. The glomerular cysts were quantified by counting the total number of glomeruli that are identified by glomerular tuft and by scoring them as cystic when a two- to threefold dilatation of the Bowman’s space was observed. Embryonic kidney length was measured using Olympus microscope and cellSens software. Immunostaining For immunohistochemistry (IHC), kidney sections were pretreated to quench endogenous peroxidase (3.0% hydrogen peroxide) and endogenous biotin (SP-2001, Vector Labs), and stained with biotinylated LTL (B-1325, Vector Labs) and DBA (B-1035, Vector Labs) following standard IHC protocol. For immunofluorescence staining, mouse kidneys were fixed in 4% PFA followed by incubation in 30% sucrose overnight at 4°C, embedded in OCT compound (Tissue-Tek), and cryosectioned at 10 μm. Frozen sections were permeabilized with 0.1% PBS- Triton X-100 for 10 min and blocked in 5% goat serum for 1 hour. Primary antibodies were incubated overnight at 4°C followed by secondary antibodies incubated at room temperature for 1 hour. Following primary antibodies were used: WT1 (sc-192, Santa Cruz), Polycystin-1 (sc-130554, Santa Cruz), Laminin (L9393, Sigma), GFP (GFP-1020, Aves), Villin-1 (2369, Cell Signaling), JAG1 (sc-8303, Santa Cruz), PAX2 (71-6000, Thermo Fisher), megalin (sc-16478, Santa Cruz), cubilin (sc-20609, Santa Cruz). Tissue sections were mounted in media containing DAPI and imaged by a Zeiss confocal microscope. Urine and Plasma Analysis Urine protein excretion was detected by SDS-PAGE followed by Coomassie blue staining and quantified using ImageJ. Urine creatinine was measured using a mouse creatinine assay kit (80350, Crystal Chem). Urine albumin/creatinine ratios were calculated. Serum BUN were measured using the Catalyst Dx Chemistry Analyzer by IDEXX. Assessment of glomerulotubular integrity The glomerulotubular integrity was assessed by analyzing the connection between the proximal tubules and the glomeruli on LTL-stained kidney sections as previously described,29, 30, 56 based on the knowledge that LTL-positive stained epithelial cells constitute part of the Bowman’s capsule with normal glomerulotubular junctions but are absent in atubular glomeruli. Briefly, consecutive serial sections of LTL-stained kidneys from Zeb2 cKO and wild-type controls (5 mice in each group) were used to make positive identification of atubular glomeruli. In order to quantify the glomerulotubular integrity, all glomeruli were counted on a single LTL-stained kidney section from each mouse and were divided into two categories on the basis of presence (positive) or absence (negative) of LTL staining in the Bowman’s capsule. Quantification of glomerulotubular integrity was presented as the percentage of LTL-positive glomeruli in total glomeruli counted in cKO mice and wild-type controls.29 Gene expression analysis Total RNA was extracted from kidneys using miRNeasy Micro kit (217084, Qiagen). cDNA was synthesized using Verso cDNA Synthesis Kit (AB-1453, Life Technologies) and TaqMan MicroRNA RT kit (4366596, Life Technologies). Gene expression was analyzed using 7500FAST real-time PCR machine with TaqMan probes (Life Technologies). Relative gene expression data were analyzed by the delta-delta-Ct method and were normalized to the either Gapdh or Ppp1r3c. Cell proliferation and apoptosis analyses Proliferative cells were identified using an anti-pHH3 antibody (9701, Cell Signaling). To quantify cell proliferation, all the glomeruli were counted on a single kidney section from each of the 8 mice and were divided into two categories on the basis of presence (positive) or absence (negative) of pHH3 staining in the Bowman’s capsule. Quantification is presented as the percentage of pHH3-positive glomeruli out of total glomeruli counted. Apoptosis was analyzed as previously described using the TUNEL assay.57 Apoptotic cells were detected using Apoptag Peroxidase In Situ Apoptosis Detection Kit (S7100, Millipore). To quantify apoptosis in the CSB and SSB, we analyzed all the CSB/SSB in one kidney section from 9 different embryos, and counted them as either “positive” when at least one apoptotic cell was present or “negative” when no apoptotic cell was visualized. The immunofluorescent TUNEL staining was performed using the ApopTag Red In Situ Apoptosis Detection Kit (S7165, Millipore). Statistical Analyses Data are given as mean and standard deviation (SD). A minimum of three mice were used for each analysis, unless stated otherwise. Statistical analysis was performed by using the Student t-test or the chi-square test, and significance was determined at p < 0.05. Supplementary Material supplement Supplemental Figure 1. Zeb2flox/flox;Pax2-cre+ conditional knockout mice die at birth. (a) Breeding scheme for generating Zeb2 conditional knockout mice using the Pax2-cre+ allele. (b) Distributions of genotypes in weaned mice at 3 weeks from 20 litters (i) and at E18.5 from 10 litters (ii). Observed: the number of mice with each genotype obtained after breeding experiments; Expected: the number of mice with each genotype expected based on the parents genotypes and normal Mendelian distribution. Supplemental Figure 2. ZEB2 expression in E16.5 developing mouse kidney. (a–f) ZEB2 expression (red) in the developing kidney was detected using the ZEB2-EGFP reporter mouse at E16.5 with an anti-GFP antibody (red). The structure of the nephrons and collecting ducts were delineated with an anti-Laminin antibody (green). Magnification of 100x (a), 200x (b), and 400x (c, d) are indicated. (e) Enlarged boxed area from c shows ZEB2 positive cells (arrows) in an S-shaped body (ssb). (f) Enlarged boxed area from d shows ZEB2 positive cells (arrows) in the Bowman’s capsule of a glomerulus (g). Supplemental Figure 3. ZEB2 and PAX2/JAG1 co-expression in E15.5 and E16.5 developing nephron from ZEB2-GFP transgenic mice. (a) Co-localization of ZEB2 (green) and PAX2 (red) on an E15.5 Zeb2-GFP reporter mouse kidney shows co-expression of ZEB2-GFP and PAX2 in the same cell nuclei (arrows) of developing nephron. (b) ZEB2 (green) is expressed in the same cells (arrow) labeled by membrane protein JAG1 (red) in the S-shaped body of an E16.5 Zeb2-GFP reporter mouse kidney. Supplemental Figure 4. Morphology of non-cystic glomeruli in Zeb2 cKO kidneys compared to wild-type littermate controls. PAS staining of P8 kidneys from Zeb2flox/flox;Six2-cre+ cKO and Zeb2+/+ wild-type littermate controls show difference of the histomorphometry of cKO mutant non-cystic glomeruli (red arrows) compared with those of wild-type mice, including increased number of sclerotic glomeruli (blue arrow) and glomerular capillary dilation (yellow arrow) in the cKO mutant kidneys. Magnification: 200x in (a) and 400x in (b). Red asterisk marks glomerular cyst in cKO mutant kidney. Supplemental Figure 5. Megalin and cubilin are upregulated in adult Zeb2 cKO renal proximal tubules. Immunohistochemistry (IHC) staining with anti-megalin and anti-cubilin antibodies show similar levels of megalin and cubilin expression (arrows) in the proximal tubules in embryonic E18.5 kidneys of Zeb2 cKO and wild-type littermate controls. The expressions of megalin and cubilin are significantly upregulated in the proximal tubules of 7-week old adult Zeb2 cKO in comparison to the age-matched wild-type littermate controls, suggesting increased proximal tubule endocytosis of albumin in Zeb2 cKO. The albuminuria in Zeb2 cKO is probably coming through those nephrons that are patent and connected to the hyperfiltrative glomeruli. Supplemental Figure 6. No major difference of JAG1 expression in E15.5 and E16.5 Zeb2 cKO kidney compared to littermate controls. (a) No major difference of JAG1 expression patterns in developing nephron (arrow) of Zeb2flox/flox;Pax2-cre+ cKO and wild-type littermates at E15.5. (b) No major difference of JAG1 expression patterns in developing nephron (arrows) of Zeb2flox/flox;Six2-cre+ cKO and wild-type littermates at E16.5. Supplemental Figure 7. No significant changes in cell proliferation in the Bowman’s capsule of Zeb2 knockout mice. (a) Staining with an antibody against phospho-Histone H3 (pHH3) as a marker of cell proliferation in E17.5 Zeb2flox/flox;Pax2-cre+ embryonic kidneys (two kidneys in each group, 400x magnification). (b) Staining with an antibody against pHH3 as a marker of proliferation in P12 Zeb2flox/flox;Six2-cre+ kidneys (two kidneys in each group, 400x magnification). (c) No significant difference in cell proliferation between the wild-type and Zeb2 conditional knockout mice in the Bowman’s capsule (cKO mutant n=235 glomeruli analyzed, wild-type n=220 glomeruli analyzed from four kidneys in each group, ns - non significant). Supplemental Figure 8. Decreased apoptosis in the C-shaped and S-shaped bodies in Zeb2 knockout mice. (a) TUNEL (rhodamine fluorochrome) staining shows reduced positive apoptotic cells (arrows) in Zeb2 conditional knockout kidneys using both Pax2-cre+ and Six2-cre+ (n=5 mutant kidneys and n=4 wild-type kidneys, 400x magnification). (b) Immunofluorescence staining with TUNEL (red) and an antibody against Laminin (green) to delineate the C-shaped bodies and the S-Shaped bodies. (c) Reduced apoptosis in the C-shaped bodies and S-shaped bodies of the Zeb2 conditional knockouts (n=34 mutant C-shaped and S-shaped bodies and n=39 wild-type C-shaped and S-shaped bodies analyzed, p-value < 10−2). We thank Drs. Xueping Fan, Kenn Albrecht, Marc Lenburg, and Matthew Layne for helpful discussion, Dr. Richard Lu for helping with Zeb2 floxed mouse transfer, Kathleen Dashner for cryostat and microtome technical support. This work is supported by NIH grants R01DK078226 (WL) and R01HD060050 (RLM), a Cooperative Research Grant from Massachusetts Life Sciences Center (WL), and is also supported in part by Research Grant #1-FY12-426 from the March of Dimes Foundation (WL). Figure 1 Deletion of Zeb2 with Pax2-cre leads to embryonic glomerulocystic kidney disease without tubular dilatation Congenital glomerular cysts were observed in the kidney of E16.5 (a) and E18.5 (b) Zeb2flox/flox;Pax2-cre+ homozygous embryos but not in the Zeb2+/+ littermate controls (H&E staining, 200x magnification). The cysts in the Zeb flox/flox;Pax2-cre+ kidneys at E18.5 are negative for the proximal tubules marker Lotus Tetragonolobus Lectin (LTL) (c), and the collecting duct and distal tubules marker Dolichos Biflorus Agglutinin (DBA) (d). Upper panels - 100x magnification; lower panels - enlarged images of boxed regions from upper panels (n≥3 for each group). Scale bars are shown. Abbreviations: g-glomerulus; cy-cyst. Figure 2 Mesenchyme-specific deletion of Zeb2 using Six2-cre also leads to glomerulocystic kidney disease without tubular dilatation (a) Dilated glomerular Bowman’s space (asterisk) was observed in the Zeb2flox/flox;Six2-cre+ embryonic kidneys at E16.5 but not in littermate controls (H&E staining, 200x magnification). (b) Glomerular cysts (asterisk) in the subcapsular region of the kidney cortex at P12 in the Zeb2flox/flox;Six2-cre+ kidneys but not in the littermate controls (H&E staining, 200x magnification). (c) IHC staining shows the cyst (cy) in a P12 Zeb2flox/flox;Six2-cre+ kidney is negative for the proximal tubules marker Lotus Tetragonolobus Lectin (LTL). Upper panels - 100x magnification; lower panels - high magnification of boxed regions in the upper panels (n≥3 for each group). (d) IHC staining shows the cysts (cy) in a P12 Zeb2flox/flox;Six2-cre+ kidney is negative for the collecting duct and distal tubules marker Dolichos Biflorus Agglutinin (DBA). Upper panels - 100x magnification; lower panels - high magnification of boxed regions in the upper panels (n≥3 for each group). Abbreviations: g-glomerulus; cy-cyst. Figure 3 Mesenchyme-specific deletion of Zeb2 with Six2-cre causes glomerulocystic kidney disease in adult mice (a) Macroscopic images of normal kidneys from a 7 weeks old control mouse (left panel) and pale and cystic kidneys (marked by asterisk) from a Zeb2flox/flox;Six2-cre+ homozygous mouse (right panel); n=5 in each group. (b) Histological images of kidneys from 7 weeks old control Zeb2+/+ mice and Zeb2flox/flox;Six2-cre+ cKO mice show large cysts in the renal cortex and outer medulla region in the cKO mice (n=3 in each group). Upper panels - 25x magnification; lower panels - 100x magnification with scale bars at 100 μm; g-glomerulus; cy-cyst. Figure 4 Mesenchyme-specific deletion of Zeb2 with Six2-cre causes glomerulosclerosis, albuminuria and renal failure in adult mice (a) Masson Trichrome staining (MTS) shows minimal fibrosis (blue) in the kidney of a 7 weeks old Zeb2flox/flox;Six2-cre+ mouse. Upper panels - 100x magnification; lower panels - high magnification of boxed regions in the upper panels (n=3 for each group). (b) Periodic acid Schiff (PAS) staining shows glomerulosclerosis (arrow) in non-cystic glomeruli of a 7 weeks old Zeb2flox/flox;Six2-cre+ mouse kidney. Upper panels - 100x magnification; lower panels - high magnification of boxed regions in the upper panels (n=3 for each group). (c) Representative SDS-page gel with Coomassie blue staining shows albuminuria (arrow) in two Zeb2flox/flox;Six2-cre+ 5 weeks old mice but not in two littermate controls (n=5 mice in each group); alb - albumin. (d) Increased albumin-to-creatinine ratio (ACR) in 5 weeks old Zeb2flox/flox;Six2-cre+ mice compared to Zeb2+/+ littermate controls (n=4 mice in each group, p < 0.05) (e) Significantly elevated BUN in 5 weeks old Zeb2flox/flox;Six2-cre+ mice compared to Zeb2+/+ littermate controls (n=3 mice in each group, *p-value< 0.001). Data are represented as means +/− standard deviation. Figure 5 Congenital atubular glomeruli in Zeb2 mesenchyme-specific knockout mice (a) LTL staining shows many glomerular Bowman’s capsule and proximal tubule connections (arrows) in P12 Zeb2+/+ wild-type control mice but very few connections in Zeb2flox/flox;Six2-cre+ mutant mice. Upper panels - 100x magnification; lower panels - 400x magnification; g -glomerulus. (n=5 in each group); Scale bars are shown. (b) Percentage of LTL-positive glomeruli out of total glomeruli counted in sagittal kidney sections from five Zeb2 cKO and five Zeb2+/+ littermate controls (n=604 glomeruli in Zeb2 cKO and n=324 glomeruli in wild-type were counted, p < 10−3). (c) Representative serial sections of P12 Zeb2+/+ wild-type kidneys (upper panels) and Zeb2flox/flox;Six2-cre+ cKO kidneys (lower panels) show the detection of glomerulotubular junctions (arrows) in the wild-type glomeruli but absence in many Zeb2 cKO glomeruli (* indicating the same glomerulus on serial sections). Figure 6 Decreased early proximal tubular marker expression in Zeb2 mesenchyme-specific knockout mice (a) At postnatal day 8 (P8), TaqMan assays show decreased mRNA expression levels of proximal tubular markers (Vil1 and Hnf1a) but not podocyte markers (Nphs1 and Nphs2) in Zeb2flox/flox;Six2-cre+ kidneys when normalized to a collecting duct marker Ppp1r3c. Collecting duct marker Upk3a was used as a control (n= 3 in each group, mean relative quantification adjusted to Ppp1r3c, p < 0.05). (b) At E14.5, TaqMan assays show decreased mRNA expression levels of proximal tubular mRNA markers (Vil1 and Hnf1a) in the cKO kidneys compared to wild-type kidneys. There are no differences for the metanephric mesenchyme markers (Pax2 and Wt1) and the collecting duct markers (Ppp1r3c and Upk3a) in cKO and control kidneys (mean relative quantification adjusted to Gapdh, n=3 in each group, p <0.05). (c) Immunofluorescent staining shows decreased expression of VIL1 (villin-1) protein in E15.5 Zeb2flox/flox;Pax2-cre+ cKO proximal tubules (arrowhead) as compared to the proximal tubules in wild-type littermates (arrows). (d) Decreased expression of VIL1 (villin-1) protein in E16.5 Zeb2flox/flox;Six2-cre+ cKO proximal tubules (arrowhead) as compared to the proximal tubules in wild-type littermates (arrows). (e) The Zeb2 cKO embryos have smaller kidney length at E16.5 as compared to wild-type controls (n=7 for cKO kidneys and n=6 for wild-type kidneys, p <10−3). Figure 7 Abnormal expression of known glomerulocystic disease genes in Zeb2 conditional knockout kidneys (a) TaqMan assays show no differences of the mRNA levels for the 6 genes and miRNA miR200 between E14.5 Zeb2 cKO kidneys and wild-type controls (mean relative quantification adjusted to Gapdh, n≥3). (b) TaqMan assays show significant upregulation of mRNA levels of Pkd1 and Hnf1β and downregulation of Glis3 mRNA in E18.5 Zeb2 cKO kidneys compared to wild-type controls (mean relative quantification adjusted to Gapdh, n=5, p <0.05) (c) Significant upregulation of Pkd1 mRNA at E18.5 and P8 but not E14.5 in Zeb2 cKO kidneys compared to wild-type controls (mean relative quantification adjusted to Gapdh, n=3, *p <0.05). (d) Immunofluorescence staining show increased levels of polycystin 1 (PC1) protein in the glomeruli (arrows) of Zeb2flox/flox;Six2-cre+ kidney at P7 (middle and lower panels), but not at E17.5 (upper panel). PC1 staining in the glomeruli was confirmed by co-expression of WT1, a glomerular podocyte marker (lower panel); Scale bars are shown. (e) Triple immunofluorescence staining with PC1, WT1 and DAPI in Zeb2 cKO kidney shows that increased level of polycystin 1 (PC1) protein was detected only in non-cystic glomeruli (glom 1) but not in an adjacent glomerulus (glom 2) with significantly enlarged Bowman’s space (cy). The podocytes (arrows) and parietal epithelial cells (arrowheads) are visible. Table 1 Cystic phenotype observed in H&E stained kidney samples of Zeb2flox/flox;Pax2-cre+ and wild-type littermates between E16.5 and E18.5 E16.5 E17.5 E18.5 Zeb2 +/+ Zeb2 flox/flox ; Pax2-cre + Zeb2 +/+ Zeb2 flox/flox ; Pax2-cre + Zeb2 +/+ Zeb2 flox/flox ; Pax2-cre + Embryos with kidney cysts (more than 1 cyst) 0/2 (0%) 3/3 (100%) 0/3 (0%) 3/3 (100%) 0/7 (0%) 8/8 (100%) Glomerular Cysts/Number of glomeruli 1/48 2%, (n=4)00 26/55 47%, (n=5) 0/40 0%, (n=3) 18/59 30.5%, (n=4) 1/363 0.2%, (n=13) 61/391 15.6%, (n=15) n= number of kidneys analyzed Table 2 Percentage of cysts with glomerular tufts on H&E stained kidney samples between E16.5 and E18.5. Cysts with Glomerular tufts/total cysts (n= number of kidneys analyzed) Zeb2flox/flox;Pax2-cre+ E16.5 26/85 (30%, n=5) E17.5 18/46 (39%, n=4) E18.5 61/226 (27%, n=15) n= number of kidneys analyzed DISCLOSURE All the authors declared no competing interests 1 Kurschat CE Muller RU Franke M An approach to cystic kidney diseases: the clinician’s view Nature reviews. Nephrology 2014 10 687 699 25266212 2 Bonsib SM The classification of renal cystic diseases and other congenital malformations of the kidney and urinary tract Archives of pathology & laboratory medicine 2010 134 554 568 20367308 3 Bissler JJ Siroky BJ Yin H Glomerulocystic kidney disease Pediatr Nephrol 2010 25 2049 2056 quiz 2056–2049 20091054 4 Lennerz JK Spence DC Iskandar SS Glomerulocystic kidney: one hundred-year perspective Archives of pathology & laboratory medicine 2010 134 583 605 20367310 5 Bernstein J Glomerulocystic kidney disease--nosological considerations Pediatr Nephrol 1993 7 464 470 8398663 6 Gusmano R Caridi G Marini M Glomerulocystic kidney disease in a family Nephrol Dial Transplant 2002 17 813 818 11981068 7 Feather SA Winyard PJ Dodd S Woolf AS Oral-facial-digital syndrome type 1 is another dominant polycystic kidney disease: clinical, radiological and histopathological features of a new kindred Nephrol Dial Transplant 1997 12 1354 1361 9249769 8 Galliani CA Gomez AM Panniello G Bisceglia M Selected case from the Arkadi M. Rywlin International Pathology Slide Series: Asymmetric, segmental glomerulocystic kidney in an infant with tuberous sclerosis complex Advances in anatomic pathology 2015 22 135 143 25664948 9 Bergmann C Fliegauf M Bruchle NO Loss of nephrocystin-3 function can cause embryonic lethality, Meckel-Gruber-like syndrome, situs inversus, and renal-hepatic-pancreatic dysplasia Am J Hum Genet 2008 82 959 970 18371931 10 Rampoldi L Caridi G Santon D Allelism of MCKD, FJHN and GCKD caused by impairment of uromodulin export dynamics Hum Mol Genet 2003 12 3369 3384 14570709 11 Bingham C Bulman MP Ellard S Mutations in the hepatocyte nuclear factor-1beta gene are associated with familial hypoplastic glomerulocystic kidney disease Am J Hum Genet 2001 68 219 224 11085914 12 Xiong M Jiang L Zhou Y The miR-200 family regulates TGF-beta1-induced renal tubular epithelial to mesenchymal transition through Smad pathway by targeting ZEB1 and ZEB2 expression Am J Physiol Renal Physiol 2012 302 F369 379 22012804 13 Tang O Chen XM Shen S MiRNA-200b represses transforming growth factor-beta1-induced EMT and fibronectin expression in kidney proximal tubular cells Am J Physiol Renal Physiol 2013 304 F1266 1273 23408168 14 Gregory PA Bracken CP Smith E An autocrine TGF-beta/ZEB/miR-200 signaling network regulates establishment and maintenance of epithelial-mesenchymal transition Molecular biology of the cell 2011 22 1686 1698 21411626 15 Garavelli L Mainardi PC Mowat-Wilson syndrome Orphanet journal of rare diseases 2007 2 42 17958891 16 Yamada Y Nomura N Yamada K The spectrum of ZEB2 mutations causing the Mowat-Wilson syndrome in Japanese populations Am J Med Genet A 2014 164A 1899 1908 24715670 17 Gregory PA Bert AG Paterson EL The miR-200 family and miR-205 regulate epithelial to mesenchymal transition by targeting ZEB1 and SIP1 Nature cell biology 2008 10 593 601 18376396 18 Hajarnis SS Patel V Aboudehen K Transcription Factor HNF-1beta Regulates MicroRNA-200 Expression Through a Long Noncoding RNA J Biol Chem 2015 19 Patel V Hajarnis S Williams D MicroRNAs regulate renal tubule maturation through modulation of Pkd1 J Am Soc Nephrol 2012 23 1941 1948 23138483 20 Higashi Y Maruhashi M Nelles L Generation of the floxed allele of the SIP1 (Smad-interacting protein 1) gene for Cre-mediated conditional knockout in the mouse Genesis 2002 32 82 84 11857784 21 Van de Putte T Maruhashi M Francis A Mice lacking ZFHX1B, the gene that codes for Smad-interacting protein-1, reveal a role for multiple neural crest cell defects in the etiology of Hirschsprung disease-mental retardation syndrome Am J Hum Genet 2003 72 465 470 12522767 22 Ohyama T Groves AK Generation of Pax2-Cre mice by modification of a Pax2 bacterial artificial chromosome Genesis 2004 38 195 199 15083520 23 Nishizaki Y Takagi T Matsui F Higashi Y SIP1 expression patterns in brain investigated by generating a SIP1-EGFP reporter knock-in mouse Genesis 2014 52 56 67 24243579 24 Wang CA Drasin D Pham C Homeoprotein Six2 promotes breast cancer metastasis via transcriptional and epigenetic control of E-cadherin expression Cancer Res 2014 74 7357 7370 25348955 25 Kobayashi A Valerius MT Mugford JW Six2 defines and regulates a multipotent self-renewing nephron progenitor population throughout mammalian kidney development Cell stem cell 2008 3 169 181 18682239 26 Terryn S Tanaka K Lengele JP Tubular proteinuria in patients with HNF1alpha mutations: HNF1alpha drives endocytosis in the proximal tubule Kidney Int 2016 89 1075 1089 27083284 27 Chevalier RL Forbes MS Generation and evolution of atubular glomeruli in the progression of renal disorders J Am Soc Nephrol 2008 19 197 206 18199796 28 Gibson IW Downie TT More IA Lindop GB Atubular glomeruli and glomerular cysts--a possible pathway for nephron loss in the human kidney? J Pathol 1996 179 421 426 8869291 29 Forbes MS Thornhill BA Chevalier RL Proximal tubular injury and rapid formation of atubular glomeruli in mice with unilateral ureteral obstruction: a new look at an old model Am J Physiol Renal Physiol 2011 301 F110 117 21429968 30 Galarreta CI Grantham JJ Forbes MS Tubular obstruction leads to progressive proximal tubular injury and atubular glomeruli in polycystic kidney disease Am J Pathol 2014 184 1957 1966 24815352 31 Massa F Garbay S Bouvier R Hepatocyte nuclear factor 1beta controls nephron tubular development Development 2013 140 886 896 23362349 32 Pritchard L Sloane-Stanley JA Sharpe JA A human PKD1 transgene generates functional polycystin-1 in mice and is associated with a cystic phenotype Hum Mol Genet 2000 9 2617 2627 11063721 33 Williams SS Cobo-Stark P James LR Kidney cysts, pancreatic cysts, and biliary disease in a mouse model of autosomal recessive polycystic kidney disease Pediatr Nephrol 2008 23 733 741 18286309 34 Kang HS Beak JY Kim YS Glis3 is associated with primary cilia and Wwtr1/TAZ and implicated in polycystic kidney disease Mol Cell Biol 2009 29 2556 2569 19273592 35 Ferrante MI Zullo A Barra A Oral-facial-digital type I protein is required for primary cilia formation and left-right axis specification Nat Genet 2006 38 112 117 16311594 36 Hossain Z Ali SM Ko HL Glomerulocystic kidney disease in mice with a targeted inactivation of Wwtr1 Proc Natl Acad Sci U S A 2007 104 1631 1636 17251353 37 Paul BM Vanden Heuvel GB Kidney: polycystic kidney disease. Wiley interdisciplinary reviews Developmental biology 2014 3 465 487 25186187 38 Ho J The regulation of apoptosis in kidney development: implications for nephron number and pattern? Frontiers in pediatrics 2014 2 128 25478553 39 Coles HS Burne JF Raff MC Large-scale normal cell death in the developing rat kidney and its reduction by epidermal growth factor Development 1993 118 777 784 8076517 40 El-Kasti MM Wells T Carter DA A novel long-range enhancer regulates postnatal expression of Zeb2: implications for Mowat-Wilson syndrome phenotypes Hum Mol Genet 2012 21 5429 5442 23001561 41 Forbes MS Thornhill BA Galarreta CI Chronic unilateral ureteral obstruction in the neonatal mouse delays maturation of both kidneys and leads to late formation of atubular glomeruli Am J Physiol Renal Physiol 2013 305 F1736 1746 24107422 42 Tanner GA Tielker MA Connors BA Atubular glomeruli in a rat model of polycystic kidney disease Kidney Int 2002 62 1947 1957 12427119 43 Conidi A Cazzola S Beets K Few Smad proteins and many Smad-interacting proteins yield multiple functions and action modes in TGFbeta/BMP signaling in vivo Cytokine & growth factor reviews 2011 22 287 300 22119658 44 Hegarty SV Sullivan AM O’Keeffe GW Zeb2: A multifunctional regulator of nervous system development Progress in neurobiology 2015 132 81 95 26193487 45 Vrljicak P Myburgh D Ryan AK Smad expression during kidney development Am J Physiol Renal Physiol 2004 286 F625 633 14656760 46 Hassane S Leonhard WN van der Wal A Elevated TGFbeta-Smad signalling in experimental Pkd1 models and human patients with polycystic kidney disease J Pathol 2010 222 21 31 20549648 47 Varelas X Sakuma R Samavarchi-Tehrani P TAZ controls Smad nucleocytoplasmic shuttling and regulates human embryonic stem-cell self-renewal Nature cell biology 2008 10 837 848 18568018 48 Peters DJ Spruit L Klingel R Adult, fetal, and polycystic kidney expression of polycystin, the polycystic kidney disease-1 gene product Lab Invest 1996 75 221 230 8765322 49 Geng L Segal Y Peissel B Identification and localization of polycystin, the PKD1 gene product J Clin Invest 1996 98 2674 2682 8981910 50 Palsson R Sharma CP Kim K Characterization and cell distribution of polycystin, the product of autosomal dominant polycystic kidney disease gene 1 Molecular medicine 1996 2 702 711 8972485 51 Ward CJ Turley H Ong AC Polycystin, the polycystic kidney disease 1 protein, is expressed by epithelial cells in fetal, adult, and polycystic kidney Proc Natl Acad Sci U S A 1996 93 1524 1528 8643665 52 Song X Di Giovanni V He N Systems biology of autosomal dominant polycystic kidney disease (ADPKD): computational identification of gene expression pathways and integrated regulatory networks Hum Mol Genet 2009 18 2328 2343 19346236 53 Thivierge C Kurbegovic A Couillard M Overexpression of PKD1 causes polycystic kidney disease Mol Cell Biol 2006 26 1538 1548 16449663 54 Bracken CP Gregory PA Kolesnikoff N A double-negative feedback loop between ZEB1-SIP1 and the microRNA-200 family regulates epithelial-mesenchymal transition Cancer Res 2008 68 7846 7854 18829540 55 Park SM Gaur AB Lengyel E Peter ME The miR-200 family determines the epithelial phenotype of cancer cells by targeting the E-cadherin repressors ZEB1 and ZEB2 Genes Dev 2008 22 894 907 18381893 56 Thornhill BA Forbes MS Marcinko ES Chevalier RL Glomerulotubular disconnection in neonatal mice after relief of partial ureteral obstruction Kidney Int 2007 72 1103 1112 17728704 57 Omori S Hida M Ishikura K Expression of mitogen-activated protein kinase family in rat renal development Kidney Int 2000 58 27 37 10886546
PMC005xxxxxx/PMC5123946.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 7500844 3382 Mol Cell Endocrinol Mol. Cell. Endocrinol. Molecular and cellular endocrinology 0303-7207 1872-8057 27793677 5123946 10.1016/j.mce.2016.10.025 NIHMS826478 Article DISORDERED ZONAL AND CELLULAR CYP11B2 ENZYME EXPRESSION IN FAMILIAL HYPERALDOSTERONISM TYPE 3 Gomez-Sanchez Celso E. 12 Qi Xin 2 Gomez-Sanchez Elise P. 3 Sasano Hironobu 4 Bohlen Martin O. 5 Wisgerhof Max 6 1 Endocrinology Division, G.V. (Sonny) Montgomery VA Medical Center 2 University of Mississippi Medical Center 3 Department of Pharmacology and toxicology and Medicine, University of Mississippi Medical Center, Jackson, MS 4 Department of Pathology, Tohoku University, Sendai, Japan 5 Department of Anatomical Sciences, University of Mississippi Medical Center, Jackson, MS 6 Division of Endocrinology, Henry Ford Health System, Detroit, MI Address Correspondence: Celso E. Gomez-Sanchez, Division of Endocrinology, University of Mississippi Medical Center, 2500 N. State St, Jackson, MS 39216, 601 368 3844, [email protected] 1 11 2016 25 10 2016 5 1 2017 05 1 2018 439 7480 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Three forms of familial primary aldosteronism have been recognized. Familial Hyperaldosteronism type 1 (FH1) or dexamethasone suppressible hyperaldosteronism, FH2, the most common form of as yet unknown cause(s), and FH3. FH3 is due to activating mutations of the potassium channel gene KCNJ5 that increase constitutive and angiotensin II-induced aldosterone synthesis. In this study we examined the cellular distribution of CYP11B2, CYP11B1, CYP17A1 and KCNJ5 in adrenals from two FH3 siblings using immunohistochemistry and immunofluorescence and obtained unexpected results. The adrenals were markedly enlarged with loss of zonation. CYP11B2 was expressed sporadically throughout the adrenal cortex. CYP11B2 was most often expressed by itself, relatively frequently with CYP17A1, and less frequently with CYP11B1. KCNJ5 was co-expressed with CYP11B2 and in some cells with CYP11B1. This aberrant co-expression of enzymes likely explains the abnormally high secretion rate of the hybrid steroid, 18-oxocortisol. 1.1 Introduction In 1953 the structure of a highly active mineralocorticoid isolated from the amorphous fraction of beef adrenal extracts (1) was elucidated and named electrocortin (2). It was soon renamed aldosterone by the group of Simpson and Tait in an academic-industry collaboration (3). Just 2 years later Jerome Conn from the University of Michigan described the first clinical case of mineralocorticoid excess due to an aldosterone-producing adenoma (4) and the first description of primary hyperaldosteronism as hypertension, hypokalemia due to potassium wasting in the urine, and hypomagnesemia caused by an aldosterone-producing adrenal adenoma. Primary hyperaldosteronism is now known to have several additional causes including bilateral adrenal hyperplasia, unilateral adrenal hyperplasia, adrenal carcinoma, rare extra-adrenal tumors producing aldosterone, and three familial forms of primary aldosteronism(5). Familial Hyperaldosteronism type 1 (FH1) or Glucocorticoid-Remediable Aldosteronism, the best characterized, is due to a gene duplication resulting from the crossover recombination of the promoter region and first 4 exons of the 11β-hydroxylase (Cyp11b1) and the last exons of the aldosterone synthase (Cyp11b2) genes, resulting in a chimeric gene expressed in the zona fasciculata and regulated by ACTH that produces an enzyme that synthesizes aldosterone (6). Familial hyperaldosteronism type 2 (FH2) is the most common FH. While the genetic basis remains unknown, many cases are in linkage with chromosome 7p22 (7, 8). FH3 has been ascribed to an inherited mutation of the KCNJ5 gene disrupting the selectivity filter of the G-protein activated inward rectifying potassium channel Kir3.4 (9). 1.2. Case reports In 1959, a 10 year old boy was described with severe hypertension, hypokalemia, polyuria and marked increase in the urinary excretion of aldosterone (9). His severe hypertension could not be controlled with the anti-hypertensive medications available in that era, so he was subjected to bilateral adrenalectomy, resulting in marked improvement of the hypertension, hypokalemia and associated symptoms. His adrenal glands were very large, the right and left adrenals weighing 8.0 and 9.0 g, respectively, with focal nodular hyperplasia and cells of a zona fasciculata phenotype laden with lipid vacuoles (9). Twenty-six years later two of his daughters, ages 7 and 4 years, were found to have severe hypertension, hypokalemic alkalosis, suppressed renin activity, and marked serum aldosterone elevation (10). Administration of dexamethasone resulted in an increase in aldosterone levels, ruling out glucocorticoid-remediable aldosteronism. The patients were lost to follow up for 8 years and presented again with the same clinical manifestations and elevated blood pressure in spite of intensive antihypertensive therapy including spironolactone and/or amiloride and potassium supplements (10). The striking biochemical findings in addition to severe hyperaldosteronism, were marked elevations in the urinary excretion of the hybrid steroids 18-hydroxycortisol and 18-oxo-tetrahydrocortisol, with normal levels of other conventional cortisol metabolites (10). Because the patients’ hypertension could not be controlled pharmacologically, including with spironolactone, they underwent bilateral adrenalectomy with normalization of the BP and serum potassium within 2 weeks. The adrenal glands were markedly enlarged with a combined left + right adrenal weight of 81 and 39 g (normal <12 g) and diffuse cortical hyperplasia (10). 1.3. Pathogenesis A significant advance in our understanding of the pathogenesis of aldosterone-producing adenomas came from the identification of somatic mutations in or near the selectivity filter of the G-protein coupled potassium channel, Kir3.4 coded by the KCNJ5 gene (11). Two mutations are most common, G151R and L168R. Sequencing of the KCNJ5 gene in the family described above (10) revealed a heterozygous T158A mutation in the three affected members (11). This mutation lies between the selectivity filter and the second transmembrane domain of the Kir3.4 potassium channel and results in the loss of potassium selectivity and depolarization of the cell membrane (11). The latter initiates the signals that increase aldosterone synthesis. We recently described the production of a highly selective mouse monoclonal antibody against the human CYP11B2 and a rat monoclonal antibody against the CYP11B1 enzyme. In this report we describe the adrenal distribution of staining of the CYP11B2 enzyme in the adrenals of the two patients described above and have done triple immunofluorescence of the CYP11B2, CYP11B1 and Kir3.4 potassium channel and triple immunofluorescence of the CYP11B1, CYP11B2 and 17α-hydroxylase enzyme. 2. Materials and Methods 2.1. Immunohistochemistry Slides were deparaffinized, subjected to antigen retrieval using a solution of EDTA 1mM and SDS 0.05% pH 9 for 45 min in a steamer, blocked with 5% goat serum in Tris 0.1M and 0.5% SDS at pH 7.4, and immunostained with the mouse monoclonal anti-human CYP11B2-41-17C antibody (1/1000), CYB5 monoclonal antibody (Acris, San Diego) (in tris 0.1 M, tween 20 0.05% with goat serum 5%, as previously reported (12). A slide was also stained with a specific monoclonal antibody against human HSD3B2 developed in our laboratory (unpublished). The slides were counterstained with hematoxylin. 2.2. Triple Immunofluorescence After deparaffinizing, antigen retrieval and blocking was done as described above, the slides were incubated overnight at 4C with a mixture of rat monoclonal anti human CYP11B1-80-7 (1/200), mouse monoclonal anti-human CYP11B2-41-17 (1/1,000) (as previously documented (12)), and rabbit anti-17α-hydroxylase (1/400) (13). The secondary antibodies were goat anti-mouse IgG H&L-Alexa 488, goat anti-rat IgG H&L-Alexa 594 and goat anti-rabbit-Alexa 647 (Jackson Immunoresearch Inc, Allentown, PA, USA) that have minimal crossreactivity to mouse, human and rat immunoglobulins. Coverslips were mounted using Vector Laboratories Vectashield mounting media with DAPI (Vector Labs, Burlingame, CA, USA). Similar triple immunofluorescence was done with a KCNJ5 antibody instead of the 17α-hydroxylase antibody using a sheep antibody from EMD-Millipore (AB9808)(1/10,000) (14) and processed as above. Immunofluorescence for the CYP17A1 (IgG2b), CyB5 (IgG1) and CYP11B2 (IgG1-biotin labeled) were done by incubating for the first two antibodies overnight, washing and incubating with specific goat anti-mouse IgG1-Alexa 488, goat anti-mouse IgG2b-Alexa 594 followed by washing and incubating for 30 min with mouse IgG, washing followed by incubation with mouse monoclonal CYP11B2-41-13B labeled with biotin for 1 hr, washed and incubated with avidin-Oyster 650 for 30 min. After washing they were mounted with Vectashield with DAPI (Vector Laboratories, Burlingame, CA). 3. Results 3.1 Similar to the description of their father’s adrenal (10), the adrenals of patients 1 (Fig 1A) and 2 (Fig 1B) stained with hematoxylin and eosin exhibited marked enlargement of the adrenal gland with complete loss of normal zonation composed primarily of lipid laden cells throughout the cortex and no clearly discernable zona glomerulosa type cells. The adrenals of patient 1, Fig 1C, E, G, and that of patient 2, Fig 1D, F, H have CYP11B2 immunoreactivity throughout the cortex in an irregular pattern. CYP11B2 immunoreactive cells are interspersed with cells not expressing the CYP11B2. Fig 2 shows immunofluorescence of the CYP11B1, CYP11B2 and KCNJ5: overlapped pictures show that cells immunoreactive for CYP11B1 and CYP11B2 are not only interspersed, but both enzymes are co-expressed in some cells. KCNJ5 was co-expressed with CYP11B2 and, in many cells, also with CYP11B1 in contrast to normal adrenals where KCNJ5 is expressed only in cells of the zona glomerulosa (11). Patient 2 showed similar co-expression and mingling of cells expressing CYP11B1 and CYP11B2, however in some areas there was greater co-expression of CYP11B1 with KCNJ5 than CYP11B2 with KCNJ5 (data not shown). 3.2 17α-Hydroxylase was expressed throughout the adrenal. As expected from normal adrenal zona fasciculata cells, it was expressed in all cells expressing CYP11B1, however, contrary to its normal expression pattern, 17α-hydroxylase was also found in a significant number of CYP11B2 expressing cells (Fig 3). In some areas 17α-hydroxylase seemed to be preferentially expressed with CYP11B2. In some cells there was clear expression of the three enzymes. 3.3 HSD3B2 was expressed strongly throughout the adrenal (Fig 4A and C). The cytochrome B5 showed just a few small areas of staining (Fig B and D). Immunofluorescence showed by in those areas CYB5 and CYP17A1 were co-expressed (Fig 4E) and in small areas CYB5 was also co-expressed with CYP11B2 (Fig 4F) 4. Discussion 4.1 Familial hyperaldosteronism type 3 is a rare form of hyperaldosteronism where various mutations of the potassium channel KCNJ5 cause the hypersecretion of aldosterone. The three members of the first family described had extremely severe hypertension and hypokalemia that was not controlled with high doses of spironolactone plus other antihypertensive agents and required a bilateral adrenalectomy for control of the hypertension and electrolyte abnormalities (9, 10). The three patients were heterozygous for the T158A mutation of the KCNJ5 gene. 4.2 The affected members of this family exhibited a marked increase in the secretion of aldosterone and 18-oxocortisol, but not of cortisol, and the secretion of aldosterone was not suppressed with the administration of dexamethasone (10). CYP17A1 is not normally expressed in the zona glomerulosa. It is co-expressed with CYP11B1 in normal adrenal zona fasciculata cell where their sequential action is required to synthesize cortisol. CYP11B2 is normally expressed only in the zona glomerulosa where it catalyzes the last steps in the synthesis of aldosterone (12). The synthesis of the hybrid steroids 18-hydroxycortisol and 18-oxocortisol requires the sequential action of CYP17A1, then CYP11B2 (15, 16). As the major blood flow in the adrenal gland is centripetal, from the zona glomerulosa, to the fasciculata, on through to the medulla, high concentrations of CYP17A1 products made downstream in the zona fasciculata do not normally become substrates for CYP11B2 directly. As in the normal adrenal there is no evidence of co-expression of these two enzymes, the synthesis of small amounts of 18-hydroxycortisol and 18-oxocortisol is thought to be by zona glomerulosa cells from circulating cortisol (17). In adrenal adenomas and in Familial Hyperaldosteronism type 1, there is a marked increase in the secretion of 18-oxocortisol (16, 18–20). In the case of Familial Hyperaldosteronism type 1, the formation of the chimeric enzyme from the unequal crossing over of the promoter region and first exons of the CYP11B1 and the last exons of the CYP11B2 results in the expression of CYP11B2 in the zona fasciculata where CYP17A1 is also expressed and allows the expression of the hybrid steroids 18-hydroxycortisol and 18-oxocortisol (6, 18). Co-expression of the CYP11B2, CYP11B1 and CYP17A1 enzymes, explains the increased production of the hybrid steroids in some aldosterone-producing adenomas (21). In this study, the CYP11B2 was widely but unevenly expressed throughout the grossly enlarged adrenal and some CYP11B2-positive cells co-expressed both the CYP11B1 and the CYP17A1 enzymes, explaining not only the marked increase in aldosterone, but the highest recorded plasma levels of 18-hydroxycortisol and 18-oxocortisol (10). In vivo KCNJ5 is normally expressed only in the zona glomerulosa of the adrenal (11) The human adrenal carcinoma HAC15 cell expresses all of the enzymes required to synthesize both aldosterone and cortisol. Transduction with a lentivirus carrying the KCNJ5-T158A mutant resulted in the decrease in the selectivity of the channel, an increase in intracellular sodium, resulting in the mobilization of calcium signaling for a marked increase in the expression of the CYP11B2 enzyme and synthesis of aldosterone (22). As HAC15 cells also express CYP11B1 and CYP17A1, it was not too surprising that the KCNJ5-T158A mutant HAC15 cells also produced significantly more of the hybrid steroid 18-oxocortisol. However the KCNJ5-T158A HAC15 cells also expressed more CYP11B1, in addition to CYP11B2, with no increase in StAR protein required to transport cholesterol in to the mitochondria, the initial step in steroidogenesis, or of other enzymes in the steroidogenic cascades to cortisol or aldosterone. CYP17A1 expression was decreased. While aldosterone production was greatly increased, cortisol was increased only slightly, albeit significantly, as was 18-oxocortisol, in the HAC15 expressing KCNJ5-T158A (22). KCNJ5 in the adrenals of these FH3 patients was co-expressed with CYP11B2, as expected, but was also abundantly expressed in cells expressing CYP11B1, most of which did not express detectable CYP11B2. The mechanism that normally limits KCNJ5 expression to zona glomerulosa cells is not known, but these findings suggest that it is not the same as that limiting CYP11B1 and CYP17A1 expression to the zona fasciculata. Nor does the expression of CYP11B2 appear to preclude the expression of CYP17A1. Unfortunately our studies in the HAC15 cells cannot answer these questions of zonal specificity of enzyme expression because they also co-express all of these enzymes. As these patients did not have hypercortisolism, the co-expression of the mutated KCNJ5 either had no effect on CYP11B1 expression, or, if CYP11B1 was increased in these patients’ adrenal cells, their cortisol production remained normal due to the appropriate function of the HPA axis and stimulation of StAR protein by ACTH. 4.3 The KCNJ5-T158A mutation resulting in a decreased selectivity of the channel for potassium over sodium results in a very severe syndrome with a grossly enlarged adrenal (9). In vitro this mutation causes a slightly less alteration in the selectivity compared to other mutations found in FH3 patients including KCNJ5-G151E or G151R (11),(23, 24). However patients with these more severe KCNJ5 mutations have a milder form of primary aldosteronism probably due to a toxic effect of increased calcium mobilization resulting in an increase in adrenal cell apoptosis and lower mass of the adrenal (24). Multiple families have now been described with germ line mutations of the KCNJ5 gene including G151E (24, 25), Y152C (26), I157S (27) and several other germ line mutations in patients with sporadic primary aldosteronism (28). A similar picture of heterogeneous expression of the CYP11B2 in a patient with the KCNJ5 Y152C mutation was reported in the supplemental material of the paper of Monticone et al (29), but with insufficient details to compare with these two patients. Patients with aldosterone-producing adenomas bearing a KCNJ5 mutation have been shown to have a high expression of CYP11B2, CYP11B1 and CYP17A1 enzymes (30) and these patients as a group excrete high amounts of the hybrid steroids (19). The co-expression of the CYP11B2 and the cytochrome B5 is very abnormal as the latter is only expressed in the zona reticularis (31). It would be interesting to speculate that the co-expression of the CYP11B2, CYP17A1 (which is extensive in these adrenals) and the cytochrome b5 might result in the production of newer steroids that remain to be identified. 4.4 In summary, immunohistochemical analysis of the cellular distribution of steroidogenic enzymes unique to normal zona glomerulosa or fasciculata cells in the grossly hyperplastic adrenal cortex of patients with hyperaldosteronism type 3 reveals the loss of zonation and mingling of cells that express CYP11B2 and/or CYP11B1, with and without CYP17A1, and some that express all three enzymes. KCNJ5 was abundantly expressed in almost all cells, including those that expressed CYP11B1 with no CYP11B2, suggesting that the mutation not only altered the potassium filter selectivity of the channel, but also altered a signal that designated specificity for CYP11B2-expressing cells, or that this family also has a second mutation that abolishes steroidogenic specialization by adrenal zonation. These studies were supported by NIH grant HL27255. Figure 1 Histology and immunohistochemistry of the adrenals of two patients with Familial Hyperaldosteronisms type 3 heterozygous for the KCNJ5-T158A mutation. Panels A, C, E, G: adrenal cortex from patient 1; Panels B, D, F and H adrenal cortex from patient 2. A & B: hematoxylin-eosin staining. C & D. CYP11B2 immunoreactivity throughout the cortex. E & G: sequentially higher magnifications of the boxes in panel C & E of patient 1. F & H: sequentially higher magnifications of the boxes in panel D & F of patient 2. Figure 2 Triple immunofluorescence of a section of the adrenal of patient 1 stained with the mouse monoclonal antibody against CYP11B2, rat monoclonal antibody against CYP11B1 and sheep polyclonal antibody against KCNJ5. The panels on the left show the individual staining and nuclear staining with DAPI. The right panels show simultaneous staining of nucleus with DAPI and the combination of antibodies as labeled. Co-expression of the steroidogenic enzymes is indicated by the orange color. Figure 3 Immunofluorescence staining of the adrenal of patient 1 with CYP11B1, CYP11B2 and CYP17A1. Panels on the left side are the individual immunoreactivity combined with nuclear staining with DAPI and right panels are overlap pictures of: Top right: CYP11B1 plus CYP11B2 which shows different cells staining for each enzyme and few cells showing co-expression of both enzymes. Middle right: CYP11B2 plus CYP17A1 showing separate cells staining of both enzymes in many cells and in a significant number of cells expressing both enzymes. Lower right panel: CYP11B1 and CYP17A1 showing immunoreactive for either enzyme or both in a significant number. Figure 4 Fig 4A and C showed the adrenal of patient 2 stained for HSD3B2 enzyme. Fig 4B and D showed the adrenal of patient 1 stained for the cytochrome B5. Fig 4E shows double immunofluorescence of CYP17A1 and cytochrome B5 in the adrenal of patient 2. Fig 4F shows double immunofluorescence of CYP11B2 and cytochrome B5 of the adrenal of patient 2. Highlights Immunohistochemistry of 2 cases of familial hyperaldosteronism 3 are described. CYP11B2 is expressed heterogeneously throughout the adrenal. CYP11B2 is co-expressed in cells with KCNJ5 and in many cells with CYP11B1. CYP11B2 is co-expressed in some cells with the 17α-hydroxylase. DISCLOSURE SUMMARY: No disclosures. Disclosures: The authors have nothing to disclose. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. 1 Grundy HM Simpson SA Tait JF 1952 Isolation of a highly active mineralocorticoid from beef adrenal extract Nature 169 795 796 14941045 2 Simpson SA Tait JF Wettstein A Neher R Van Euw J Reichstein T 1953 Isolierung eines neuen kristallisierten hormons aus nebennieren mit besonders hoher wirksamkeit auf den mineralostoffwechsel Experientia 9 333 333 3 Simpson SA Tait JF Wettstein A Neher R von Euw J Schindler O Reichstein T 1953 Konstitution des aldosterons, des neuen mineralocorticoids Experientia 10 132 133 4 Conn JW 1955 Primary aldosteronism, a new clinical syndrome J Lab Clin Med 45 3 7 13233623 5 Mulatero P Dluhy RG Giacchetti G Boscaro M Veglio F Stewart PM 2005 Diagnosis of primary aldosteronism: from screening to subtype differentiation Trends Endocrinol Metab 16 114 119 15808809 6 Lifton RP Dluhy RG Powers M Rich GM Cook S Ulick S Lalouel J-M 1992 A chimaeric 11β-hydroxylase/aldosterone synthase gene causes glucocorticoid-remediable aldosteronism and human hypertension Nature 355 262 265 1731223 7 Stowasser M Gordon RD Tunny TJ Klemm SA Finn WL Krek AL 1992 Familial hyperaldosteronism type II: five families with a new variety of primary aldosteronism Clin Expt Pharmacol Physiol 19 319 322 8 Lafferty AR Torpy DJ Stowasser M Taymans SE Lin JP Huggard PR Gordon RD Stratakis CD 2000 A novel genetic locus for low renin hypertension: familial hyperaldosteronism type II maps to chromosome 7 (7p22) J Med Genet 37 831 835 11073536 9 Therien B Mellinger RC Caldwell JR Howard PJ 1959 Primary aldosteronism due to adrenal hyperplasia; occurrence in a boy aged 10 years AMA J Dis Child 98 90 99 13660631 10 Geller DS Zhang J Wisgerhof MV Shackleton C Kashgarian M Lifton RP 2008 A novel form of human mendelian hypertension featuring nonglucocorticoid-remediable aldosteronism The Journal of clinical endocrinology and metabolism 93 3117 3123 18505761 11 Choi M Scholl UI Yue P Bjorklund P Zhao B Nelson-Williams C Ji W Cho Y Patel A Men CJ Lolis E Wisgerhof MV Geller DS Mane S Hellman P Westin G Akerstrom G Wang W Carling T Lifton RP 2011 K+ channel mutations in adrenal aldosterone-producing adenomas and hereditary hypertension Science 331 768 772 21311022 12 Gomez-Sanchez CE Qi X Velarde-Miranda C Plonczynski MW Parker CR Rainey W Satoh F Maekawa T Nakamura Y Sasano H Gomez-Sanchez EP 2014 Development of monoclonal antibodies against human CYP11B1 and CYP11B2 Molecular and cellular endocrinology 383 111 117 24325867 13 Gell JS Carr BR Sasano H Atkins B Margraf L Mason JI Rainey WE 1998 Adrenarche results from development of a 3beta-hydroxysteroid dehydrogenase-deficient adrenal reticularis The Journal of clinical endocrinology and metabolism 83 3695 3701 9768686 14 Oki K Plonczynski MW Lam ML Gomez-Sanchez EP Gomez-Sanchez CE 2012 The Potassium Channel, Kir3.4 Participates in Angiotensin II-Stimulated Aldosterone Production by a Human Adrenocortical Cell Line Endocrinology 153 4328 4335 22798349 15 Ulick S Chu MD 1982 Hypersecretion of a new corticosteroid, 18-hydroxycortisol in two types of adrenocortical hypertension Clin Exper Hyper A4 1771 1777 16 Ulick S Chu MD Land M 1983 Biosynthesis of 18-oxocortisol by aldosterone producing adrenal tissue The Journal of biological chemistry 258 5498 5502 6602132 17 Freel EM Shakerdi LA Friel EC Wallace AM Davies E Fraser R Connell JM 2004 Studies on the origin of circulating 18-hydroxycortisol and 18-oxocortisol in normal human subjects The Journal of clinical endocrinology and metabolism 89 4628 4633 15356073 18 Gomez-Sanchez CE Montgomery M Ganguly A Holland OB Gomez-Sanchez EP Grim CE Weinberger MH 1984 Elevated urinary excretion of 18-oxocortisol in glucocorticoid suppressible aldosteronism The Journal of clinical endocrinology and metabolism 59 1022 1024 6480807 19 Mulatero P di Cella SM Monticone S Schiavone D Manzo M Mengozzi G Rabbia F Terzolo M Gomez-Sanchez EP Gomez-Sanchez CE Veglio F 2012 18-hydroxycorticosterone, 18-hydroxycortisol, and 18-oxocortisol in the diagnosis of primary aldosteronism and its subtypes The Journal of clinical endocrinology and metabolism 97 881 889 22238407 20 Nakamura Y Satoh F Morimoto R Kudo M Takase K Gomez-Sanchez CE Honma S Okuyama M Yamashita K Rainey WE Sasano H Ito S 2011 18-oxocortisol measurement in adrenal vein sampling as a biomarker for subclassifying primary aldosteronism The Journal of clinical endocrinology and metabolism 96 E1272 1278 21593107 21 Nakamura Y Maekawa T Felizola SJ Satoh F Qi X Velarde-Miranda C Plonczynski MW Ise K Kikuchi K Rainey WE Gomez-Sanchez EP Gomez-Sanchez CE Sasano H 2014 Adrenal CYP11B1/2 expression in primary aldosteronism: Immunohistochemical analysis using novel monoclonal antibodies Molecular and cellular endocrinology 392 73 79 24837548 22 Oki K Plonczynski MW Luis Lam M Gomez-Sanchez EP Gomez-Sanchez CE 2012 Potassium Channel Mutant KCNJ5 T158A Expression in HAC-15 Cells Increases Aldosterone Synthesis Endocrinology 153 1774 1782 22315453 23 Scholl UI Lifton RP 2013 New insights into aldosterone-producing adenomas and hereditary aldosteronism: mutations in the K+ channel KCNJ5 Current opinion in nephrology and hypertension 22 141 147 23318698 24 Scholl UI Nelson-Williams C Yue P Grekin R Wyatt RJ Dillon MJ Couch R Hammer LK Harley FL Farhi A Wang WH Lifton RP 2012 Hypertension with or without adrenal hyperplasia due to different inherited mutations in the potassium channel KCNJ5 Proceedings of the National Academy of Sciences of the United States of America 109 2533 2538 22308486 25 Mulatero P Tauber P Zennaro MC Monticone S Lang K Beuschlein F Fischer E Tizzani D Pallauf A Viola A Amar L Williams TA Strom TM Graf E Bandulik S Penton D Plouin PF Warth R Allolio B Jeunemaitre X Veglio F Reincke M 2012 KCNJ5 Mutations in European Families With Nonglucocorticoid Remediable Familial Hyperaldosteronism Hypertension 59 235 240 22203740 26 Monticone S Hattangady NG Penton D Isales CM Edwards MA Williams TA Sterner C Warth R Mulatero P Rainey WE 2013 A Novel Y152C KCNJ5 mutation responsible for familial hyperaldosteronism type III J Clin Endocrinol Metab 98 E1861 1865 24037882 27 Charmandari E Sertedaki A Kino T Merakou C Hoffman DA Hatch MM Hurt DE Lin L Xekouki P Stratakis CA Chrousos GP 2012 A Novel Point Mutation in the KCNJ5 Gene Causing Primary Hyperaldosteronism and Early-Onset Autosomal Dominant Hypertension The Journal of clinical endocrinology and metabolism 97 E1532 1539 22628607 28 Murthy M Xu S Massimo G Wolley M Gordon RD Stowasser M O’Shaughnessy KM 2014 Role for germline mutations and a rare coding single nucleotide polymorphism within the KCNJ5 potassium channel in a large cohort of sporadic cases of primary aldosteronism Hypertension 63 783 789 24420545 29 Monticone S Hattangady NG Penton D Isales C Edwards MA Williams TA Sterner C Warth R Mulatero P Rainey WE 2013 A novel Y152C KCNJ5 mutation responsible for familial hyperaldosteronism type III The Journal of clinical endocrinology and metabolism 30 Azizan EA Lam BY Newhouse SJ Zhou J Kuc RE Clarke J Happerfield L Marker A Hoffman GJ Brown MJ 2012 Microarray, qPCR, and KCNJ5 sequencing of aldosterone-producing adenomas reveal differences in genotype and phenotype between zona glomerulosa- and zona fasciculata-like tumors The Journal of clinical endocrinology and metabolism 97 E819 829 22442279 31 Nakamura Y Xing Y Hui XG Kurotaki Y Ono K Cohen T Sasano H Rainey WE 2011 Human adrenal cells that express both 3beta-hydroxysteroid dehydrogenase type 2 (HSD3B2) and cytochrome b5 (CYB5A) contribute to adrenal androstenedione production The Journal of steroid biochemistry and molecular biology 123 122 126 21185375
PMC005xxxxxx/PMC5123955.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 0375410 5127 J Pediatr J. Pediatr. The Journal of pediatrics 0022-3476 1097-6833 27663215 5123955 10.1016/j.jpeds.2016.08.046 NIHMS818192 Article Complications of Endoscopic Retrograde Cholangiopancreatography in Pediatric Patients; A Systematic Literature Review and Meta-Analysis Usatin Danielle MD 1 Fernandes Melissa MD 1 Allen Isabel E. PhD 2 Perito Emily R. MD 12 Ostroff James MD 3 Heyman Melvin B. MD 1 1 Department of Pediatrics, University of California, San Francisco 2 Department of Epidemiology and Biostatistics, University of California, San Francisco 3 Department of Medicine, University of California, San Francisco Corresponding Author: Melvin B. Heyman, MD, 550 16th Street, 5th Floor, UCSF Box 0136, San Francisco, CA 94143, P:415-476-5892, F:415-353-2472, [email protected] 18 11 2016 20 9 2016 12 2016 01 12 2017 179 160165.e3 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Objectives To systematically review risks and summarize reported complication rates associated with performance of ERCP in children over the past two decades. Study design A systematic literature search of MEDLINE, Embase, and Web of Science from Jan 1995 to Jan 2016 was conducted for observational studies published in English. Studies reporting ERCP complications in patients <21 years without history of liver transplant or cholecystectomy were included. A summary estimate of the proportion of children who experienced complications following ERCP was derived using a random effects meta-analysis. Results Thirty-two studies involving 2612 children and 3566 procedures were included. Subjects’ ages ranged from 3 days to 21 years. Procedures were performed for biliary (54%), pancreatic (38%), and other (8%) indications. 56% of ERCPs were interventional. The pooled complication rate was 6% (95% CI: 4%– 8%). Procedural complications included post-ERCP pancreatitis (166, 4.7%), bleeding (22, 0.6%) and infections (27, 0.8%). The pooled estimate of post-ERCP pancreatitis was 3% (95%CI 0.02–0.05), and other complications were 1% (95%CI: 0.02–0.05). In neonatal cholestasis subgroup the pooled complication rate was 3% (95% CI: 0.01–0.07). Adult and pediatric gastroenterologists and surgeons performed the ERCPs. Available data limited the ability to report differences between pediatric-trained and other endoscopists. Conclusions Complications associated with pediatric ERCP range widely in severity and are reported inconsistently. Our review suggests 6% of pediatric ERCPs have complications. Further studies using systematic and standardized methodologies are needed to determine the frequency and risk factors for ERCP related complications. ERCP children pancreas hepatobiliary neonatal cholestasis endoscopist safety Endoscopic retrograde cholangiopancreatography (ERCP) is a specialized procedure that combines gastrointestinal endoscopy and fluoroscopy for diagnostic and therapeutic management of disorders of the pancreas and biliary tract. The procedure has been widely applied in adults for over 40 years. The first reported procedure in a child was in 1976 by Waye using an adult-sized duodenoscope (1). Smaller diameter duodenoscopes developed in the 1980s and 1990s led to expanding application of ERCP in children. ERCP allows less invasive access to the biliary tree and pancreatic duct than surgery or transhepatic procedures. Although the utility and feasibility of ERCP in pediatrics has been demonstrated in case reports and series, concerns about safety remain. In 2000, the North American Society for Pediatric Gastroenterology, Hepatology and Nutrition (NASPGHAN) subcommittee on Endoscopy Procedures published a narrative review of data on indications, technical considerations, risks and complications of ERCP in children (2). Since that time, the number of ERCPs performed on children has increased(3). In addition, many additional studies on ERCP in pediatric patients have made this procedure appear routine. However, complication rates appear to vary between case series, potentially dependent on multiple factors including patient selection, operator, and underlying disease factors (4–7). By conducting a systematic literature review, we examined complication rates for pediatric patients undergoing ERCP and compared complication rates by patient characteristics, endoscopist training, and center type. METHODS The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) I statement was used to identify and collate studies (8). We systematically searched MEDLINE/PubMed, Ovid Embase, and Web of Science for full text articles in which subjects <21 years of age underwent ERCP. Complications were reported as an outcome. We used the following search phrase ((technical AND (success OR successes OR outcome OR outcomes) OR quality assurance OR patient safety OR complications OR treatment outcomes OR intraoperative complications OR postoperative complications) AND ((pediatric OR child OR children OR childhood OR adolescent OR teen OR infant OR toddler)) AND (“Cholangiopancreatography, Endoscopic Retrograde“ OR Endoscopic Retrograde Cholangiopancreatography OR ERCP)) to identify articles. To find articles that may have been missed during the literature search, reference lists of candidate articles were also reviewed. The search was limited to English language texts from January 1995-January 2016. The final search was completed on February 10, 2016. The limitation to studies published since January 1995 was to avoid overlap with a previously published review(2). Study Selection Criteria Two independent reviewers screened all articles for methodological validity and relevance prior to inclusion in the review. Any disagreements between the two reviewers were resolved through discussion with a third reviewer. Our selection criteria were specified in advance and included the following: (1) published in English in a peer-reviewed journal; (2) available in full text; (3) included youth <21 years of age, excluding children who had undergone previous liver transplant or other hepatobiliary surgical procedure (eg, for choledochal cyst, cholecystectomy, for cholelithiasis); (4) observational study designs; (5) studies that examined the number and type of complications after an ERCP. If multiple articles were available from a single center, the most recently published article or the article containing the most comprehensive detail of study characteristics was selected for review. Article Review and Data Extraction Data were extracted from papers included in the review using a standardized data extraction tool created for this study in REDCap electronic data capture tools hosted at University of California, San Francisco(9). REDCap (Research Electronic Data Capture) is a secure, web-based application designed to support data capture for research studies, providing 1) an intuitive interface for validated data entry; 2) audit trails for tracking data manipulation and export procedures; 3) automated export procedures for seamless data downloads to common statistical packages; and 4) procedures for importing data from external sources. The data were extracted by two reviewers (D.U., M.F.). Once completed, any disagreements were arbitrated by a third reviewer (M.H.). The data extracted included details about the study population, study methods and outcomes of significance to the review question and specific objectives. The study center type, endoscopist type, anesthetic type, patient characteristics, indications and findings of ERCP, and percent of procedures which were interventional were collected. If not specifically stated in the manuscript, we attempted to determine the training background of the endoscopist by searching the internet to identify current position within hospital system in adult or pediatric gastroenterology program. This systematic review is registered on Prospero (http://www.crd.york.ac.uk/PROSPERO :CRD42016038065). Data synthesis Studies were categorized based on the author, year of publication, subjects’ age, procedure indication, and interventional or diagnostic procedure type. The complication prevalence for each study was summarized and compiled. Statistical analysis was performed using Microsoft Excel and STATA Version 13. Significant variations in study design and reporting amongst included publications precluded use of a standard definition for post-ERCP complications. We performed a random effects meta-analysis of the data using the Metaprop program (STATA 13) to provide a summary estimate of the proportion of children with complications following ERCP. We chose random effects to account for the variability among the studies, given that most were case reports and case series. Metaprop allowed for the inclusion of studies with complication proportions of 0 to 1(10). In addition to the overall complication rate, secondary analyses were performed to further understand factors impacting the summary estimate. A subgroup analysis was performed on cases from American centers. Additionally, a sensitivity analysis of all studies was performed following the exclusion of three papers that were felt to be outliers. Outliers were identified based on the findings from the inclusive summary estimate. Finally, a random effects meta-analysis of the complications that excluded post-ERCP-associated pancreatitis (PEP) was performed. Literature Search Results The PRISMA flow diagram was used to document the literature search process (Figure 1; available at www.jpeds.com). We identified 1932 articles and imported these into Endnote software. Duplicates were removed and any remaining duplicates were manually removed, leaving 1642 articles. A thorough review of all article titles and abstracts yielded 44 articles that were reviewed in full. Subsequently, 12 articles were excluded for the following reasons: not reporting complication rates, presentation of patients counted in other included study, and/or including patients who had undergone liver transplant or cholecystectomy. Of the studies that included patients with a mixture of patients who did and did not meet our exclusion criteria (n=3), it was not possible to distinguish the complication rates. As such, the entire study was excluded. Ultimately, 32 articles were identified (4–7, 11–38). RESULTS All 32 included studies were retrospective cohort studies or case series reporting on ERCP related complications in pediatric patients. From these articles, data was obtained on 2612 children and adolescents, who underwent a total of 3566 ERCPS. Some children accounted for multiple procedures within the same article. The median number of patients per study was 44, with a range of 3 to 276 patients. Studies differed in their definition of post-procedure observation periods, which may have impacted what was considered an ERCP related complication. Only 14 of the 32 studies specified their follow up period (7, 11–23). The range of follow up time was 2 days to 30 months. Fourteen of the 32 studies were conducted in the US (4–6, 13–16, 25–31). All 32 of the studies were from referral centers. Three of the studies were multicenter (14, 16, 18). We were unable to determine if the procedures were performed in stand-alone children’s hospitals. Four studies reported exclusively on 237 biliary ERCPs in infants (11, 19, 22, 35), and the rest included patients from birth to 21 years (Tables I and II; available at www.jpeds.com). Five studies did not report sex; of the remaining 27, 54% of included subjects were male (Table III; available at www.jpeds.com). Procedure characteristics ERCPs were done for biliary indications (54%), pancreatic indications (38%) or other indications including abdominal pain (8%). Overall, 56% of procedures were interventional, and the remainder were solely diagnostic (Table II). Adult gastroenterology-trained endoscopists performed a majority of ERCPs reported in these studies. Only 2 of the 32 reports specifically stated that a pediatric gastroenterologist performed the endoscopy (7, 31). One report cited a general surgeon who performed the ERCPs(5). Eleven studies did not specify who performed the procedure. ERCPs were performed under both general anesthesia and sedation in 19 studies. General anesthesia was exclusively used in 10 studies (6, 11, 16, 19, 21, 22, 30, 31, 34, 35, 38). Sedation was used exclusively in two studies(7, 12). ERCP complications Reported complications varied widely ranging from bleeding to post-ERCP pancreatitis (PEP) to fussiness. No deaths were reported in any of the studies as a consequence of ERCP. Five studies cited the ASGE lexicon for endoscopic adverse events by Cotton et al (39) to define complications and follow up period (6, 26, 27, 31, 36). The severity of complications was not documented in most studies. We aggregated complications into broad categories: PEP, infection, bleeding, and other (Table II). Out of all 3566 procedures performed, 291 (8.2%) involved complications. The pooled overall complication rate was 6% (95% CI: 4%–8%) (Figure 2). Analysis revealed significant heterogeneity among the studies (I2=80.54%, p<0.001), with three outlier studies (4, 20, 30). Kamelmaz et al was a small series and reported complications in two of the three performed procedures(4). Prasil et al reported one instance of bleeding and six episodes of PEP in 21 patients (20). Rescorla et al only included 6 subjects and was limited to patients with pancreatic trauma, of which 4 had complications (30). Removal of these three studies from the analysis resulted in the same pooled estimate of complications (6%; 95% CI: 4%–8%). PEP was reported in 166 of 3566 (4.7%) of procedures; bleeding was reported in 22 (0.6%); and infection was reported in 27 (0.8%). The pooled estimate of complications other than PEP was 1% (95% CI: 0%–3%). The pooled estimate of PEP as a complication was 3% (95% CI: 0.02–0.05). Further subgroup analyses were performed to identify important factors contributing to the heterogeneity of the studies including pediatric trained endoscopists, US centers, and neonatal cholestasis. In the two studies performed by pediatric trained endoscopists, there were 238 procedures with an overall complication rate of 4.6%. A pooled overall complication rate for the 14 US sites was 5% (95% CI: 2%–10%). Among the four studies reporting solely on ERCP in neonatal cholestasis, the complication rate was 4.2% out of 238 procedures, with a pooled complication rate of 3% (95% CI: 0.01–0.07) (Figure 3; available at www.jpeds.com). Excluding these four neonatal studies, the pooled estimate of the complication rate for the remaining studies was 6% (95% CI: 4% – 9%). DISCUSSION Our analysis documents that the mean prevalence of complications after ERCP performed in children aged 0–21 years is approximately 6%. In contrast, the estimated complication rates of upper endoscopy has been reported as 2.3% (40) and of colonoscopy, 1.1% (41), in the same patient population. Post-ERCP pancreatitis was the most commonly reported complication, and rates of other serious complications including bleeding and infection were less than 1% each. Since 2000, when the last comprehensive report on ERCP in pediatric patients was published (2), use of ERCP in pediatric patients with pancreatic and biliary disease have become increasingly prevalent(3). Numerous single center experiences with information on utility and feasibility of ERCP have been published, but to date no benchmarks for acceptable or expected complication rates exist. We report a complication rate that can be used in future ERCP research to explore factors such as the effects of subtype of procedure, patient characteristics or endoscopist training. Our review of the literature also demonstrates the heterogeneity of previous studies. This highlights the need for a more standardized approach to complication reporting, even in small studies. A few studies cite the ASGE recommended guidelines for determining complications (39). Although these guidelines provide definitions for AEs and levels of severity they have yet to be adapted to pediatric populations. We were unable to apply standardized definitions for what constitutes a complication of ERCP; the studies reviewed report a wide variety of complication types with no specific definition or severity for any of the complications reported. This suggests that certain issues still need to be defined, for example anatomic location of complication, timing in relation to the procedure, and severity. One unanswered question is whether the pediatric or adult endoscopist should be tasked with the procedure in pediatric patients, particularly infants and young children. Pediatric patients who require ERCP are often managed at pediatric referral centers where pediatricians, pediatric anesthesiologists, surgeons, radiologists, and pediatric intensivists coordinate care for these patients. Even though a pediatric-trained gastroenterologist might be more appropriate to perform the procedure in young patients, inadequate case volume both in training and in maintenance of skills is frequently cited as a reason for these procedures to be performed by adult-trained endoscopists (42–44). Our study is unable to answer this question, as only 2 reports of primarily pediatric trained gastroenterologists performing the ERCPs were available for inclusion in our review. Post-ERCP pancreatitis was the most common complication reported, but classification of this is challenging. Six of the 32 studies did not report on post-ERCP pancreatitis as a complication. We cannot discern whether they had no cases, or if they considered post-procedure pain and pancreatic enzyme elevation to be expected outcomes of pancreatic duct visualization. Institutions may not routinely measure amylase or lipase levels after ERCP, even if patients experience abdominal pain. Troendle et al recently investigated factors associated with post-ERCP pancreatitis, noticing this occurred in 10.9% of ERCPs. They found pancreatic duct injection, sphincterotomy, or a history of chronic pancreatitis placed subjects at higher odds of PEP(45). Our pooled complications rate was lower than this, but documentation of complications after ERCP in the pediatric population was inconsistent in the reviewed studies, indicating the need for future prospective studies in this area. Additional limitations of the existing literature are lack of consistency with reported follow-up times and minimal data available about the temporal relationship of complications to ERCP date. Studies varied, with some only reporting complications evident in the 1–2 days following the procedure, and others including the full follow up time reported, occasionally longer than a year. The intent of this review was to investigate what is known about the factors that impact the rate of complications in pediatric ERCP. However, we found that the majority of studies fail to relate complications to the other covariates of interest. For example, one might expect an increased rate of PEP with instrumentation of the pancreatic duct and/or pancreatic disease. However, few studies report complication rates by intervention type or by disease state. Furthermore, the effect of endoscopist training, sedation versus anesthesia, or indication on complications are not evaluable, as these factors are all reported separately. Despite these limitations, our report does provide a comprehensive picture of available literature on complications after pediatric ERCP. We hope our findings, and data missing from previously reported literature that would be clinically helpful, will provide groundwork for future studies to further our knowledge in this area. Prospective data collection across multiple centers is needed to determine which patient-, facility-, and physician-factors are important to optimize the safety and efficacy of pediatric ERCP. The Pediatric ERCP Database Initiative is a multicenter international database currently collecting data on all patients undergoing ERCP under the age of 18 years (http://www.utsouthwestern.edu/research/fact/detail.html?studyid=STU%20012014–086). This information will enhance our ability to provide a reliable consent process for these procedures. Efforts should also be made to standardize our definitions of post-ERCP complications, so that information will be generalizable to all centers performing this procedure in children. Supported by the Cystic Fibrosis Foundation (<grant number> [to D.U.]) and the National Institutes of Health (T32 DK007762 [to D.U. and M.F.] and K23 DK099253-01A1 [E.P.]). Figure 1 PRISMA flowchart of literature search Figure 2 Pooled estimate of ERCP complications in pediatric patients. Studies removed as outliers. Figure 3 (online only) Pooled Estimate of Proportion of Complications in ERCPs Performed Only in Neonatal Cholestasis. Pooled estimate takes into account the variability in study size as well as the heterogeneity of each of theses studies assigning a weight and then combines the proportions according to this weight. Table 1 Diagnostic ERCP in Neonatal Cholestasis Study Characteristics Author, year Patients (n) Procedures (n) Male (%) Age Means (years) Age Range (years) Endoscopist Training Complications (n) Aabakken, 2009 22 23 NR* 0.2 0.06–0.7 Adult 2 Petersen, 2009 140 140 NR* 0.16 0.04–0.48 Other/Unknown 5 Shanmugam, 2009 48 48 50 0.16 0.05–0.27 Adult 3 Shteyer, 2012 27 27 52 0.15 0.09–0.24 Adult 0 * NR= Not Reported Table 2 ERCP and Study Characteristics in Studies Including All Age Ranges Author, year Procedures (n) Patients (n) Age Mean (years) Age Range (years) Endoscopist Training‡ Diagnosis§ (n) Interventional (%) Complications (%) B P O Kamelmaz, 1999 3 3 12.8 6.5–16 O/U 0 3 0 33% 67% Rescorla, 1995 6 6 4.8 2–8 O/U 0 6 0 0% 67% Abukhalaf, 1995 16 16 10.5 0.17–18 O/U 9 4 0 25% 0% Zargar, 2003 16 16 12.6 7–16 Adult 7 0 9 100% 6% Prasil, 2001 21 20 11.3 4–17 Adult 15 5 0 48% 33% Green, 2007 26 19 13 7–16 O/U 11 14 1 88% 0% Tagge, 1997 26 26 10.1 0.5–19 O/U 21 5 0 50% 4% Hsu, 2000 34 22 10.7 1.5–17 O/U 6 18 0 68% 6% Paris, 2010 38 29 10.3 3–17 Adult 9 29 0 34% 11% Rocca, 2005 48 38 10 0.08–17 Adult 24 14 0 77% 6% Teng, 2000 50 42 NR* 0.16–15 O/U 31 11 0 12% 2% Taj, 2012 52 40 13.6 3–18 Adult 19 19 2 92% 2% Pfau, 2002 53 43 13.5 1–18 Adult 20 14 17 45% 6% Halvorson, 2013 70 45 12 6–17 Adult 32 22 0 93% 7% Keil, 2000 80 59 11.2 0.05–18 O/U 41 16 2 58% 9% Poddar, 2001 84 72 8.8 0.92–14 Pediatric 52 17 8 26% 7% Vegting, 2009 99 61 7 0.008–16.9 Adult 51 10 0 61% 8% Li, 2010 110 42 16 NR* Adult 0 42 0 100% 17% Issa, 2007 125 125 13.3 5–18 O/U 115 9 0 50% 4% Troendle, 2013 154 65 15.2 0.08–18.4 Pediatric 50 15 0 42% 3% Varadarajulu, 2004 163 116 9.3 0.08–17 Adult 60 49 7 47% 2% Agarwal, 2014 221 172 13.8 5–18 Adult 0 172 0 71% 4% Dua, 2008 224 185 NR* 0–18 Adult 71 43 71 33% 2% Otto, 2011 231 167 11.4 0.17–21 Adult 31 148 0 69% 5% Saito, 2014 235 220 4 0.02–20 O/U 181 5 32 3% 10% Limketkai, 2013 289 154 11.5 1–17 O/U 132 220 8 85% 6% Giefer, 2015 425 276 13.6 0.2–18 Both 194 210 11 81% 13% Enestvedt, 2013 429 296 14.9 0.25–21 Adult 268 51 92 64% 17% * NR= Not Reported ‡ O/U= Other/Unknown § B=Biliary, P=Pancreatic, O=Other Table 3 Study Characteristics of 32 Included Studies Author, year Center Study Period Patients (n) Procedures (n) Mean Age (years) Age Range (years) Aabakken, 2009 (14) University of Oslo 1999 – 2006 22 23 0.2 0.06 – 0.7 Abukhalaf, 1995 (27) Jordan University Hospital 1990 – 1993 16 16 10.5 0.17 – 18 Agarwal, 2014 (15) University of Hyderbad 2010 – 2011 172 221 13.8 5 – 18 Dua, 2008 (28) Children’s Hospital of Wisconsin 1994 – 2004 185 224 0 0 – 18 Enestvedt, 2013 (7) Children’s Hospital of Philadelphia 1993 – 2011 296 429 14.9 0.25 – 21 Giefer, 2015 (16) Seattle Children’s Hospital 1994 – 2011 276 425 13.6 0.2 – 18 Green, 2007 (6) De Vos Children’s Hospital, MI 2000 – 2005 19 26 13 7 – 16 Halvorson, 2013 (29) University of Maryland 2003 – 2011 45 70 12 6 – 17 Hsu, 2000 (17) UC Davis and U of South Carolina 1994 – 1996 22 34 10.7 1.5 – 17 Issa, 2007 (36) Qatif Central Hospital 1993 – 2005 125 125 13.3 5 – 18 Kamelmaz, 1999 (5) Marshall University, WV NR* 3 3 12.8 6.5 – 16 Keil, 2000 (37) University of Prague 1995 – 1999 59 80 11.2 0.05 – 18 Li, 2010 (20) Changhai Hospital, China 1997 – 2009 42 110 16 NR* Limketkai, 2013 (30) Johns Hopkins University 1998 – 2011 154 289 11.5 1 – 17 Otto, 2011 (31) University of Pittsburg 1992 – 2008 167 231 11.4 0.17 – 21 Paris, 2010 (21) St Justine University, Canada 1990 – 2007 29 38 10.3 3 – 17 Petersen, 2009 (22) Hanover Medical School, Germany 2001 – 2008 140 140 0.16 0.04 – 0.48 Pfau, 2002 (32) Rainbow Babies, OH NR* 43 53 13.5 1 – 18 Poddar, 2001 (8) Chandigarh, India 1995 – 2000 72 84 8.8 0.92 – 14 Prasil, 2001 (23) McGill University, Montreal children’s 1990 – 1999 20 21 11.3 4 – 17 Rescorla, 1995 (33) Riley Children’s Hospital, IN 1988 – 1993 6 6 4.8 2 – 8 Rocca, 2005 (24) Turin, IT 1996 – 2002 38 48 10 0.08 – 17 Saito, 2014 (37) Chiba University, Japan 1980 – 2011 220 235 4 0.02 – 20 Shanmugam, 2009 (25) London, Kings College Hospital 1997 – 2007 48 48 0.16 0.05 – 0.27 Shteyer, 2012 (38) Hadassah-Hebrew University, Israel 2000 – 2010 27 27 0.15 0.09 – 0.24 Tagge, 1997 (18) South Carolina 1990 – 1995 26 26 10.1 0.5 – 19 Taj, 2012 (39) Civil Hospital Karachi 2007 – 2010 40 52 13.6 3 – 18 Teng, 2000 (40) Kyushu University NR* 42 50 0 0.16 – 15 Troendle, 2013 (34) University of Texas Southwestern 2006 – 2012 65 154 15.2 0.08 – 18.4 Varadarajulu, 2004 (19) Birmingham and Charleston 1994 – 2002 116 163 9.3 0.08 – 17 Vegting, 2009 (41) Emma Children’s Hospital, Netherlands 1995 – 2005 61 99 7 0.008 – 16.9 Zargar, 2003 (26) Kashmir, India 1998 – 2001 16 16 12.6 7 – 16 * NR= Not Reported This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. The authors declare no conflicts of interest. 1 Waye JD Endoscopic retrograde cholangiopancreatography in the infant Am J Gastroenterol 1976 65 5 461 3 949055 2 Fox VL Werlin SL Heyman MB Endoscopic retrograde cholangiopancreatography in children. Subcommittee on Endoscopy and Procedures of the Patient Care Committee of the North American Society for Pediatric Gastroenterology and Nutrition J Pediatr Gastroenterol Nutr 2000 30 3 335 42 10749424 3 Pant C Sferra TJ Barth BA Deshpande A Minocha A Qureshi WA Trends in endoscopic retrograde cholangiopancreatography in children within the United States, 2000–2009 J Pediatr Gastroenterol Nutr 2014 59 1 57 60 24509307 4 Kamelmaz I Elitsur Y Pancreas divisum–the role of ERCP in children The West Virginia medical journal 1999 95 1 14 6 9949690 5 Green JA Scheeres DE Conrad HA Cloney DL Schlatter MG Pediatric ERCP in a multidisciplinary community setting – Experience with a fellowship-trained general surgeon Surgical Endoscopy and Other Interventional Techniques 2007 21 12 2187 92 17514391 6 Enestvedt BK Tofani C Lee DY Abraham M Shah P Chandrasekhara V Endoscopic retrograde cholangiopancreatography in the pediatric population is safe and efficacious J Pediatr Gastroenterol Nutr 2013 57 5 649 54 23760230 7 Poddar U Thapa BR Bhasin DK Prasad A Nagi B Singh K Endoscopic retrograde cholangiopancreatography in the management of pancreaticobiliary disorders in children J Gastroenterol Hepatol 2001 16 8 927 31 11555109 8 Moher D Liberati A Tetzlaff J Altman DG Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement PLoS medicine 2009 6 7 e1000097 19621072 9 Harris PA Taylor R Thielke R Payne J Gonzalez N Conde JG Research electronic data capture (REDCap)–a metadata-driven methodology and workflow process for providing translational research informatics support Journal of biomedical informatics 2009 42 2 377 81 18929686 10 Nyaga VN Arbyn M Aerts M Metaprop: a Stata command to perform meta-analysis of binomial data Archives of public health = Archives belges de sante publique 2014 72 1 39 25810908 11 Aabakken L Aagenaes I Sanengen T Aasen S Emblem R Bjornland K Utility of ERCP in neonatal and infant cholestasis Journal of laparoendoscopic & advanced surgical techniques Part A 2009 19 3 431 6 19216697 12 Agarwal J Nageshwar Reddy D Talukdar R Lakhtakia S Ramchandani M Tandan M ERCP in the management of pancreatic diseases in children Gastrointest Endosc 2014 79 2 271 8 24060520 13 Giefer MJ Kozarek RA Technical outcomes and complications of pediatric ERCP Surgical endoscopy 2015 29 12 3543 50 25673350 14 Hsu RK Draganov P Leung JW Tarnasky PR Yu AS Hawes RH Therapeutic ERCP in the management of pancreatitis in children Gastrointest Endosc 2000 51 4 Pt 1 396 400 10744808 15 Tagge EP Tarnasky PR Chandler J Tagge DU Smith C Hebra A Multidisciplinary approach to the treatment of pediatric pancreaticobiliary disorders J Pediatr Surg 1997 32 2 158 64 discussion 64–5 9044114 16 Varadarajulu S Wilcox CM Hawes RH Cotton PB Technical outcomes and complications of ERCP in children Gastrointest Endosc 2004 60 3 367 71 15332025 17 Li ZS Wang W Liao Z Zou DW Jin ZD Chen J A long-term follow-up study on endoscopic management of children and adolescents with chronic pancreatitis Am J Gastroenterol 2010 105 8 1884 92 20216535 18 Paris C Bejjani J Beaunoyer M Ouimet A Endoscopic retrograde cholangiopancreatography is useful and safe in children J Pediatr Surg 2010 45 5 938 42 20438931 19 Petersen C Meier PN Schneider A Turowski C Pfister ED Manns MP Endoscopic retrograde cholangiopancreaticography prior to explorative laparotomy avoids unnecessary surgery in patients suspected for biliary atresia J Hepatol 2009 51 6 1055 60 19726099 20 Prasil P Laberge JM Barkun A Flageole H Endoscopic retrograde cholangiopancreatography in children: A surgeon’s perspective J Pediatr Surg 2001 36 5 733 5 11329577 21 Rocca R Castellino F Daperno M Masoero G Sostegni R Ercole E Therapeutic ERCP in paediatric patients Dig Liver Dis 2005 37 5 357 62 15843086 22 Shanmugam NP Harrison PM Devlin J Peddu P Knisely AS Davenport M Selective use of endoscopic retrograde cholangiopancreatography in the diagnosis of biliary atresia in infants younger than 100 days J Pediatr Gastroenterol Nutr 2009 49 4 435 41 19680152 23 Zargar SA Javid G Khan BA Yattoo GN Shah AH Gulzar GM Endoscopic sphincterotomy in the management of bile duct stones in children American Journal of Gastroenterology 2003 98 3 586 9 12650791 24 Abukhalaf A The Role of Endoscopic Retrograde Cholangiopancreatography in Small Children and Adolescents Surgical laparoscopy & endoscopy 1995 5 4 296 300 7551282 25 Dua K Miranda A Santharam R Ramanujam S Werlin S ERCP in the evaluation of abdominal pain in children Gastrointest Endosc 2008 68 6 1081 5 18640674 26 Halvorson L Halsey K Darwin P Goldberg E The safety and efficacy of therapeutic ERCP in the pediatric population performed by adult gastroenterologists Dig Dis Sci 2013 58 12 3611 9 24026405 27 Limketkai BN Chandrasekhara V Kalloo AN Okolo PI 3rd Comparison of performance and safety of endoscopic retrograde cholangiopancreatography across pediatric age groups Dig Dis Sci 2013 58 9 2653 60 23709156 28 Otto AK Neal MD Slivka AN Kane TD An appraisal of endoscopic retrograde cholangiopancreatography (ERCP) for pancreaticobiliary disease in children: our institutional experience in 231 cases Surgical endoscopy 2011 25 8 2536 40 21359895 29 Pfau PR Chelimsky GG Kinnard MF Sivak MV Wong RCK Isenberg GA Endoscopic retrograde cholangiopancreatography in children and adolescents Journal of Pediatric Gastroenterology and Nutrition 2002 35 5 619 23 12454575 30 Rescorla FJ Plumley DA Sherman S Scherer LR 3rd West KW Grosfeld JL The efficacy of early ERCP in pediatric pancreatic trauma J Pediatr Surg 1995 30 2 336 40 7738761 31 Troendle DM Barth BA ERCP can be safely and effectively performed by a pediatric gastroenterologist for choledocholithiasis in a pediatric facility J Pediatr Gastroenterol Nutr 2013 57 5 655 8 24048163 32 Issa H Al-Haddad A Al-Salem AH Diagnostic and therapeutic ERCP in the pediatric age group Pediatr Surg Int 2007 23 2 111 6 17149628 33 Keil R Snajdauf J Stuj J Kalousova J Nevolova P Janik V Endoscopic retrograde cholangiopancreatography in infants and children Indian journal of gastroenterology : official journal of the Indian Society of Gastroenterology 2000 19 4 175 7 11059184 34 Saito T Terui K Mitsunaga T Nakata M Kuriyama Y Higashimoto Y Role of Pediatric Endoscopic Retrograde Cholangiopancreatography in an Era Stressing Less-Invasive Imaging Modalities Journal of Pediatric Gastroenterology and Nutrition 2014 59 2 204 9 24762457 35 Shteyer E Wengrower D Benuri-Silbiger I Gozal D Wilschanski M Goldin E Endoscopic retrograde cholangiopancreatography in neonatal cholestasis J Pediatr Gastroenterol Nutr 2012 55 2 142 5 22790901 36 Taj MA Leghari A Qureshi S Ghazanfar S Niaz SK Quraishy MS Endoscopic retrograde cholangiopancreatography: a therapeutic modality in children and adolescents JPMA The Journal of the Pakistan Medical Association 2012 62 2 98 101 22755366 37 Teng R Yokohata K Utsunomiya N Takahata S Nabae T Tanaka M Endoscopic retrograde cholangiopancreatography in infants and children Journal of gastroenterology 2000 35 1 39 42 10632539 38 Vegting IL Tabbers MM Taminiau J Aronson DC Benninga MA Rauws EAJ Is Endoscopic Retrograde Cholangiopancreatography Valuable and Safe in Children of All Ages? Journal of Pediatric Gastroenterology and Nutrition 2009 48 1 66 71 19172126 39 Cotton PB Eisen GM Aabakken L Baron TH Hutter MM Jacobson BC A lexicon for endoscopic adverse events: report of an ASGE workshop Gastrointest Endosc 2010 71 3 446 54 20189503 40 Thakkar K El-Serag HB Mattek N Gilger MA Complications of pediatric EGD: a 4-year experience in PEDS-CORI Gastrointest Endosc 2007 65 2 213 21 17258979 41 Thakkar K El-Serag HB Mattek N Gilger M Complications of pediatric colonoscopy: a five-year multicenter experience Clin Gastroenterol Hepatol 2008 6 5 515 20 18356115 42 Kapral C Duller C Wewalka F Kerstan E Vogel W Schreiber F Case volume and outcome of endoscopic retrograde cholangiopancreatography: Results of a nationwide Austrian benchmarking project Endoscopy 2008 40 8 625 30 18680074 43 Adler DG Lieb JG 2nd Cohen J Pike IM Park WG Rizk MK Quality indicators for ERCP Gastrointest Endosc 2015 81 1 54 66 25480099 44 Jorgensen J Kubiliun N Law JK Al-Haddad MA Bingener-Casey J Christie JA Endoscopic retrograde cholangiopancreatography (ERCP): core curriculum Gastrointest Endosc 2016 83 2 279 89 26708081 45 Troendle DM Abraham O Huang R Barth BA Factors associated with post-ERCP pancreatitis and the effect of pancreatic duct stenting in a pediatric population Gastrointestinal Endoscopy 2015 81 6 1408 16 25686874
PMC005xxxxxx/PMC5123977.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 7905589 6173 Neurotoxicology Neurotoxicology Neurotoxicology 0161-813X 1872-9711 27771255 5123977 10.1016/j.neuro.2016.10.010 NIHMS825354 Article Pb exposure prolongs the time period for postnatal transient uptake of 5-HT by murine LSO neurons Park Sunyoung 12 Nevin Andrew B.C. 1 Cardozo-Pelaez Fernando 1 Lurie Diana I. 1 1 Center for Structural and Functional Neuroscience, Center for Environmental Health Sciences, Department of Biomedical & Pharmaceutical Sciences, College of Health Professions and Biomedical Sciences, The University of Montana, Missoula, MT 59812 2 Business Planning Department, Kyowa Hakko Kirin Korea Co., Ltd. Seoul, Korea Corresponding Author: Diana I. Lurie, Ph.D., Dept. of Pharmaceutical Sciences, School of Pharmacy and Allied Health Sciences, Skaggs Building Room 383, The University of Montana, Missoula, MT 59812-1552, Phone: 406-243-2103, Fax: 406-243-5228, [email protected] 28 10 2016 19 10 2016 12 2016 01 12 2017 57 258269 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Pb exposure is associated with cognitive deficits including Attention Deficit Hyperactivity Disorder (ADHD) in children and alters auditory temporal processing in humans and animals. Serotonin has been implicated in auditory temporal processing and previous studies from our laboratory have demonstrated that developmental Pb decreases expression of serotonin (5-HT) in the adult murine lateral superior olive (LSO). During development, certain non-serotonergic sensory neurons, including auditory LSO neurons, transiently take up 5-HT through the serotonin reuptake transporter (SERT). The uptake of 5-HT is important for development of sensory systems. This study examines the effect of Pb on the serotonergic system in the LSO of the early postnatal mouse. Mice were exposed to moderate Pb (0.01mM) or high Pb (0.1mM) throughout gestation and postnatal day 4 (P4) and P8. We found that Pb exposure prolongs the normal developmental expression of 5-HT by LSO neurons and this is correlated with expression of SERT on LSO cell bodies. The prolonged expression of 5-HT by postnatal LSO neurons is correlated with decreased synaptic immunolabeling within the LSO. This Pb-associated decrease in synaptic density within the LSO could contribute to the auditory temporal processing deficits and cognitive deficits associated with developmental Pb exposure. Lead acetate serotonin reuptake transporter (SERT) serotonin superior olivary nuclei auditory system development 1.1 Introduction Lead (Pb) is a widespread environmental pollutant with neurotoxic effects causing neurobehavioral and cognitive deficits in humans (Finkelstein et al., 1998; Lanphear et al., 2000; Canfield et al., 2003). The developing central nervous system (CNS) is particularly susceptible to environmental Pb exposure (Landrigan and Todd, 1994; Moreira et al., 2001), suggesting that Pb may alter critical stages of development. Currently, the Centers for Disease Control (CDC) states that there is no safe blood lead level in children and that action be initiated for children with blood Pb levels above 5µg/dL (http://www.cdc.gov/nceh/lead/; 2/2/16). Blood Pb levels lower than 10 µg/dL have been shown to produce cognitive and neurobehavioral disorders such as attention deficit hyperactivity disorder (ADHD), and dyslexia (Glotzer et al., 1995; Lanphear et al., 2000; Bellinger and Bellinger, 2006; Braun et al., 2006; Bellinger, 2008). Even very low blood Pb levels (below 2 µg/dL) have been recently shown to be a risk factor for ADHD (Ha et al., 2009; Kim et al., 2013). Pb exposure is also associated with deficits in central auditory temporal processing (Finkelstein et al., 1998; Lurie et al., 2006; Jones et al., 2008). Of significance to the current study, children with either dyslexia or ADHD have also been shown to have deficits in auditory temporal processing (Breier et al., 2003; Facoetti et al., 2003; Putter-Katz et al., 2005; Wright and Conlon, 2009; Jafari et al., 2015). The cellular mechanism that underlies these Pb-induced cognitive and neurobehavioral dysfunctions are largely unknown, however serotonin (5-HT) may play a role. 5-HT has been implicated in the modulation of auditory temporal processing (Hurley et al., 2002; Hurley and Pollak, 2005; Papesh and Hurley, 2015) although the role of 5-HT during central auditory development has not been fully elucidated. Studies in our laboratory have found that low-level Pb exposure during development decreases 5-HT and vesicular monoamine transporter 2 (VMAT2) immunostaining in brainstem auditory nuclei, particularly in the lateral superior olive (LSO) (Fortune and Lurie, 2009). Thus the effect of Pb on cognitive function could be mediated through the serotonergic system. It is well established that during brain development, certain non-serotonergic neurons transiently express 5-HT. These neurons include the principal projection neurons of sensory systems such as the auditory, visual, and somatosensory systems (Gaspar et al., 2003). These neurons cannot synthesize 5-HT, instead, they accumulate 5-HT through uptake by SERT, a high affinity 5-HT transporter, which is also transiently expressed in these same neurons (Gaspar et al., 2003; Narboux-Neme et al., 2008; Thompson, 2008). In the auditory system, a subset of LSO neurons transiently express 5-HT from postnatal day 1 (P1) to P8 in wild-type mice, and from embryonic day 18 to P10 in Monoamine Oxidase A (MAOA) knockout mice (Cases et al., 1998; Thompson, 2006). Because Pb exposure decreases 5-HT within the developing LSO, we hypothesized that Pb disrupts the normal transient uptake of 5-HT by LSO neurons during early postnatal development. The current study was undertaken to determine whether developmental Pb exposure alters the normal transient uptake of 5-HT by developing LSO neurons. Brainstem sections from postnatal control and Pb-exposed mice were quantified for 5-HT and SERT and total brainstem levels of 5-HT were measured by HPLC analysis. Because our previous studies had found that Pb exposure reduced synaptic development in the adult LSO (Fortune and Lurie, 2009), we also quantified synaptophysin immunlabeling (SYP) in Pb exposed postnatal mice to determine if Pb-induced changes in the 5-HT system resulted in decreased numbers of synapses in the developing LSO neurons. We found that Pb prolongs the length of time that developing LSO neurons express 5-HT until at least P8. The expression of 5-HT in LSO neurons is correlated with the expression of SERT on LSO cell bodies, and total brainstem levels of 5-HT increased with the moderate dose of developmental Pb exposure. Pb also decreased immunoreactivity for SYP, demonstrating that the Pb-induced disruption of 5-HT accumulation in LSO neurons is correlated with impaired synaptic maturation within the LSO. Materials and Methods 2.1 Chronic lead exposure to CBA/CaJ mice Breeding pairs of CBA/CaJ mice were obtained from the Jackson Laboratory (Bar Harbor, MA). Mice were maintained in microisolator units and kept in the University of Montana specific pathogen free animal facility. Cages, bedding, and food were sterilized by autoclaving and mice were handled with aseptic gloves. Mice were allowed food and water ad libitum. All animal use was in accordance with NIH and University of Montana IACUC guidelines. Thirteen breeding pairs of CBA mice were randomly assigned to three groups having unlimited access to water (pH 3.0) containing 0 mM (control), 0.01 mM (moderate) or 0.1 mM (high) Pb acetate. Offspring were exposed to Pb throughout gestation and through the dam’s milk until sacrifice. The concentrations of Pb yielded blood Pb levels of 8.0 ± 0.4 µg/dL in moderate Pb, and 42.3 ± 1.97 µg/dL in high Pb P21 mice. Blood was collected from deeply anesthetized mice by retro-orbital puncture. Blood Pb levels were measured by the Montana Health Department in Helena, MT. The means for the No Pb group include values of <1.0 which were included in the data set as equal to 1.0 (data not shown) (Prins et al., 2010). 2.1.2 Tissue preparation for Immunohistochemistry Mice were deeply anesthetized using 2’,2’,2’-tribromoethanol (TBE) and perfused transcardially with 4% Na-periodate-lysine-paraformaldehyde fixative (PLP, final concentrations 0.01M sodium periodate, 0.075M lysine-HCl, 2.1% paraformaldehyde, 0.037M phosphate). Brains were removed and post-fixed in PLP overnight at 4°C, rinsed 3 times for 10 minutes each in phosphate buffered saline (PBS) and transferred to a 20% sucrose solution in PBS for 2 days at 4°C. Finally, brains were transferred to a 1:1 mixture of a 20 % sucrose solution in PBS and Optimal Cutting Temperature (O.C.T.) compound (Sakura Finetek, Torrance, CA) for 1–2 days at 4°C. Brains were embedded into 1.5 cm square embedding cups filled with O.C.T. compound, and then frozen in dry ice and 100% ethanol and stored at −20°C. Ten-micron tissue sections were cut on a Thermo Shandon Cryotome Cryostat (Thermo Shandon, Pittsburg, PA) and a one in three series was collected for each brain. Alternate sections were labeled with thionin and compared to the immunolabeled sections to confirm the identity and location of LSO neurons. 2.1.3 Immunohistochemistry Alternate sections from brains from no Pb, moderate Pb, and high Pb (n=5–8 per treatment group) were thawed to room temperature and rinsed in PBS three times for 10 minutes each. Sections were then permeabilized for 30 minutes in 0.5% Triton X-100 in PBS and blocked for 30 minutes with 4% of the appropriate normal serum (Vector Laboratories, Burlingame, CA) in PAB (1% sodium azide, 0.5% bovine serum albumin in PBS) and incubated with primary antibody for 1–4 days in a humid chamber at 4°C. The sections were rinsed in PBS three times for 10 minutes each and incubated with the appropriate secondary antibody, Alexa Fluor-488, 568, 594, 633; 1:400, or 488 Avidin-Biotin complex; 1:500 (Invitrogen, Grand Island, NY) in PAB for 1 hour at room temperature in the dark. Sections were then rinsed in PBS followed by distilled water and coverslipped with FluorSaverTM (Calbiochem®, San Diego, CA) and stored at 4°C. The primary antibodies used for immunohistochemitry were as follows: rabbit polyclonal anti-5-HT (1:10,000); goat polyclonal anti-5-HT (1:500), rabbit polyclonal anti-SERT (1:250), mouse monoclonal anti-Synaptophysin (1:12,000). 2.1.4 Antibodies The rabbit polyclonal against serotonin (5-HT) was raised in rabbit against serotonin coupled to bovine serum albumin with paraformaldehyde (Cat. No. 20080, ImmunoStar Inc., Hudson, WI). No cross-reactivity of serotonin antisera was seen with 5-hydroxytrytophan, 5-hydroxyindole-3-acetic acid, and dopamine (manufacturer’s specifications). The immunostar 5-HT antibody labels our mouse brainstem tissue in a staining pattern that is virtually identical to other studies that use the same antiserum to label the mouse superior olive and inferior colliculus (Hurley et al., 2002; Thompson, 2006) and has been previously well characterized in our system (Fortune and Lurie, 2009). In addition, preadsorption with the 5-HT/bovine serum albumin (BSA) conjugate protein (20 µg/ml, Cat. No. 20081, ImmunoStar Inc., Hudson, WI) abolishes all immunoreactivity in contrast to preadsorption with BSA, which does not affect immunostaining. The goat polyclonal against serotonin (5-HT) was used for the double-label experiments and was raised in goat against serotonin coupled to bovine serum albumin with paraformaldehyde (Cat. No. 20079, ImmunoStar Inc., Hudson, WI). No cross-reactivity of serotonin antisera was seen with 5-hydroxytrytophan, 5-hydroxyindole-3-acetic acid, and dopamine (manufacturer’s specifications). This goat polyclonal 5-HT antibody labels our mouse brainstem sections in a staining pattern similar to that of the rabbit polyclonal 5-HT antibody. In addition, preadsorption with the 5-HT/bovine serum albumin (BSA) conjugate protein (20 µg/ml, Cat. No. 20081, ImmunoStar Inc., Hudson, WI) eliminates all immunoreactivity, whereas preadsorption with BSA does not affect immunostaining (data not shown). The rabbit polyclonal antibody against the serotonin transporter (SERT) was raised in rabbit against synthetic peptide sequence corresponding to amino acids (602–622) of rat 5-HT transporter coupled to keyhole limper hemocyanin (Cat. No. 24330, ImmunoStar Inc., Hudson, WI). The SERT antibody labels our mouse brainstem sections in a staining pattern virtually identical to that observed in the mouse superior olive using the same antibody (Thompson and Thompson, 2009). The rabbit polyclonal against Tryptophan Hydroxylase (TPH) was raised in sheep against recombinant rabbit tryptophan hydroxylase, isolated as inclusion bodies from E. coli and purified by preparative SDS PAGE (Cat. No. AB1556, Chemicon, USA). The TPH antibody labels our mouse brainstem sections in a staining pattern that is virtually identical to that observed in the dorsal raphe using the same antibody (Maguire et al., 2014). The goat polyclonal against Monoamine Oxidase A (MAOA) was raised in goat against a peptide mapping near the C-terminus of MAO-A of human origin (Cat. No. SC18396; Santa Cruz Biotechnology, USA). The MAOA antibody co-localizes with the TPH antibody in dorsal raphe neurons in our mouse brain sections in a similar pattern to locus coerulius neurons that are double labeled with TPH antibodies and the same MAO-A antibody as used in the present study (Sader-Mazbar et al., 2013). The mouse monoclonal antibody against synaptophysin was raised against in the vesicular fraction of bovine brain against the SY38 epitope; a pentapeptide repeat structure in the C-terminal cytoplasmic tail of synaptophysin (Cat. No. MAB5258, Millipore, Temecula, CA). This synaptophysin antibody labels our mouse brainstem sections in a staining pattern that is virtually identical to the pattern seen in the ferret superior olive (Alvarado et al., 2004), and the rat superior olive (immunolabeled with a different mouse monoclonal antibody against synaptophysin (Sigma; (Caminos et al., 2007)). The Millipore synpatophysin antibody has also been characterized in our mouse system (Fortune and Lurie, 2009). 2.1.5 Tissue Imaging and analysis All fluorescent slides were viewed at either a 40x or a 60x objective using either a Bio-Rad Radiance 2000 Confocal microscope or an Olympus FV 1000 Fluoview Confocal microscope. Quantitative analysis for immunostaining was performed as previously described (Fortune and Lurie, 2009). Briefly, two to five sections per animal were analyzed from the middle of the nucleus of interest by an observer blinded as to treatment group. Images were collected and then converted from color tiff files to black and white 12-bit tiff files, and the integrated optical density (IOD) of the immunostaining was measured using MediaCybernetics Image-Pro software (Bethesda, MD). IOD measurements were used for quantification of immunostaining because it analyzes both the area of immunostained tissue that met threshold as well as the intensity of the immunostaining. A threshold of immunostaining in control (No Pb) animals was set for each antibody such that all immunoreactivity met threshold, and was used as a comparison with the Pb treatment groups. Immunostaining within the LSO in control and Pb-exposed mice was then quantified and averaged. To separately measure the IOD of 5-HT immunostaining in the LSO neuronal cell bodies and processes, a threshold for object size was also set such that either only neuronal cell bodies or processes were selected for IOD measurement. The immunostaining within three to five random areas (430 µm2) within the LSO in control and Pb-exposed mice was then quantified and averaged. Statistical differences in immunostaining between control and Pb-exposed mice were analyzed GraphPad Prism 6.0 software (GraphPad Software, La Jolla, CA). 2.1.6 Sample preparation for HPLC analysis At each time point of assessment, brains from mice (five per treatment group) were quickly dissected and the forebrain was removed with a coronal cut separating dorsally the superior and inferior colliculus. After dissection, the remaining basal cerebral cortex and cerebellum were gently removed from the separated brainstem under the dissecting microscope. The remaining brainstem includes the inferior colliculus, and the ventral and dorsal brainstem areas. The dissected tissue was immediately weighed, and quickly frozen in the liquid nitrogen. Tissue was collected from controls (n=5), moderate Pb (n=5) and high Pb (n=5) mice. The brainstems from each treatment group was combined and homogenized by sonication in 500 µl of perchloric acid (0.5M) containing 3,4-dihydroxybenzylamine (DBA, 31 ng/mL) (Branson Sonifier 150, Branson, Danbury, CT), and stored at −20 °C until ready to use. 2.1.7 RP-HPLC analysis of 5HT 5-HT levels were measured using reverse phase high performance liquid chromatography (RPHPLC) with electrochemical detection. The homogenates were centrifuged at 14,000 × g for 20 min at 4°C and the supernatants filtered through a Millex® hydrophilic LCR (PTFE) 0.45 um filter (Millipore, Bedford, MA). Sample filtrates were loaded into autosampler vials and were placed into an ESA Model 5 autosampler (ESA, Chelmsford, MA). An OmniSpher 5 C18 chromatographic column (Varian Inc., Lake forest, CA) with a length of 25 cm and an internal diameter of 4.6 mm was used to separate analytes. The mobile phase consisted of water:acetonitrile (9:1, vol/vol) containing 0.15 M monochloroacetic acid, 0.12 M sodium hydroxide, 0.6 mM EDTA and 1.30 mM sodium octyl sulfate, and the pH was adjusted to 3.2 with glacial acetic acid. A constant flow rate of 1ml/min was maintained and the column effluent was analyzed with a Model 5600A ESA CoulArray® electrochemical detector (ESA, Chelmsford, MA). Potentials of the three ESA Model 6210 four channel electrochemical cells, placed in series, were as follows: (channels 1 through 5)-50, 0, 25, 100, 200 mV, (6 through 12) 300 mV. 5-HIAA and 5-HT were monitored at 200 mV. The ratio of the peak heights produced by 5-HT were compared to the peak height produced by DBA (internal standard) in the samples and used to obtain the analyte levels from a calibration curve. Data was expressed as micrograms of analyte per gram of wet tissue weight (µg/g wt). 2.1.8 Statistical Analysis Data are expressed as mean ± SEM and were analyzed using a one-way analysis of variance (ANOVA) followed by Dunnett’s post hoc test; P < 0.05 was considered significant. Results 3.1 Mouse blood Pb levels The present study used three different doses of Pb in drinking water; the no Pb control, moderate Pb (0.01 mM), and high Pb (0.1 mM). The blood lead levels (mean ± SEM) of the mice were measured as: No Pb controls (≤1.38 ± 0.14 µg/dL), moderate Pb (8.0 ± 0.4 µg/dL), and high Pb (42.3 ± 1.97 µg/dL). Neither the high nor the moderate dose of Pb resulted in any lost litters or changes in postnatal body size and weight, and were considered sub-toxic doses. In addition, the blood lead levels of the high Pb animals have been used to demonstrate Pb’s neurotoxic effects in rodents (Gilbert et al., 1996; Lasley and Gilbert, 1996, 2000). Blood Pb levels in our moderate exposure are slightly higher than the newly established guidelines for blood lead levels in humans (5 µg/dL). 3.1.2 Pb exposure prolongs the time that developing LSO neurons transiently express 5-HT In order to determine whether developmental Pb exposure affects the normal transient expression of 5-HT by LSO neurons in early postnatal ages, control and Pb-exposed brainstem sections from postnatal day (P) 4 and P8 were immunostained for 5-HT and both LSO cell bodies and processes were quantified. Many 5-HT immunopositive cell bodies were observed in the LSO of control and high Pb mice at P4 (Figure 1A, C; 2A). Moderate Pb exposure resulted in LSO somata with significantly less 5-HT immunoreactivity (Figure 1B, 2A) compared with controls and the high Pb animals, suggesting that the moderate dose of Pb delayed the uptake of 5-HT by LSO neurons. In contrast to the 5-HT staining in LSO neuronal cell bodies, Pb exposure does not significantly change the 5-HT staining in LSO processes at P4 (Figure 2C). By P8, non-Pb exposed LSO neuronal cell bodies have lost their immunoreactivity for 5-HT, indicating that the transient expression of 5-HT by normal developing LSO neurons is over by P8 (Figure 1A1). However, both the moderate and the high dose of Pb result in continued expression of 5-HT in LSO neurons (Figure 1B1, C1, 2B). Similar to P4, 5-HT in LSO processes remains unaffected by Pb exposure at P8 (Figure 2D). Thus, Pb exposure prolongs the period of time that LSO neurons transiently express 5-HT through P8. However, LSO neurons do not continue to be immunopositive for 5-HT indefinitely. By P21, both control and Pb exposed LSO neurons show no expression of 5-HT (Fortune and Lurie, 2009). It is important to note that LSO neurons appear to be a specific target of Pb, because 5-HT immunostaining in processes within LSO remains unchanged with Pb exposure. In addition, LSO neurons were the only auditory neurons within the brainstem that immunolabeled for 5-HT during the early postnatal period. Neurons in the cochlear nucleus and the Medial Nucleus of the Trapezoid Body (MNTB) did not express 5-HT at any time during postnatal development (data not shown). 3.1.3 LSO neurons at P4 do not synthesize or degrade 5-HT To rule out any possibility that the Pb-induced alterations in 5-HT immunostaining in LSO neurons during early postnatal development is the result of changes in 5-HT synthesis and/or degradation, brainstem sections from control and Pb exposed mice were immunostained for the 5-HT synthesizing enzyme, tryptophan hydroxylase (TPH). LSO neurons have been found to be immunonegative for TPH in P6 MAOA k/o mice (Thompson and Thompson, 2009). In agreement with these studies, LSO somata were immunonegative for TPH in our P4 control mice (Figure 3A) as well as in Pb exposed mice (data not shown). As expected, neurons in the raphe nuclei located in the same brainstem section were immunopositive for TPH (Figure 3B). Thus, LSO neurons at P4 do not appear to be able to synthesize 5-HT. That being the case, then extracellular 5-HT must be taken up by transport in developing LSO neurons. In serotononergic neurons that synthesize 5-HT, the 5-HT that is taken up from the extracellular space by SERT is rapidly degraded by monoamine oxidase (MAOA) (Vitalis et al., 2002). It is possible that Pb is affecting MAOA in LSO neuronal cell bodies, resulting in less degradation of the 5-HT that is taken up by the cell bodies. MAO activity can be modulated by Pb (Devi et al., 2005) and therefore could be a potential mechanism by which Pb changes 5-HT levels in LSO neurons during postnatal development. We therefore examined whether the 5-HT-positive LSO neurons observed at P4 also contain MAOA. Double-labeling for 5-HT and MAOA in the LSO of P4 control mice demonstrated that 5-HT-positive LSO somata were immunonegative for MAOA (Figure 3C). LSO somata were immunonegative in Pb exposed mice as well (data not shown). In contrast, adjacent raphe neurons were immunopositive for both 5-HT and MAOA (Figure 3D). These results confirm that LSO neurons do not have the synthetic machinery necessary for either the synthesis or the degradation of 5-HT and suggest that the Pb-induced changes in 5-HT immunostaining within LSO neurons is due to the modulation of SERT on LSO neuronal cell bodies. 3.1.4 SERT expression in the developing LSO One possible explanation for the Pb-induced prolongation in 5-HT expression by LSO neurons may be that Pb affects the expression of SERT. It has been previously shown that LSO neurons take up 5-HT transiently during development through the transient expression of SERT on LSO neuronal cell bodies (Cases et al., 1998; Thompson, 2006). If Pb extends the period of time that LSO neurons express SERT, then we would expect to see an prolonged uptake of 5-HT into LSO neurons. We first immunostained brainstem sections from control and Pb exposed P4, P8, and P21 mice to determine if Pb affected the overall expression of SERT in LSO. Total levels of SERT immunoreactivity within the LSO (which can include SERT immunoreactivity within axons of passage from the Raphe and astrocytes) did not change with Pb exposure (Figures 4 & 5), although the amount of SERT immunostaining increases slightly from P4-P8 (Figure 5). However, because neuronal processes and astrocytes also express SERT (Hirst et al., 1998; Malynn et al., 2013) staining the entire LSO for SERT might not pick up subtle changes in SERT distribution that could occur on LSO neuronal cell bodies following Pb exposure. Therefore, brainstem sections from control and Pb exposed P4 and P8 mice were double immunostained for 5-HT and SERT. We found that LSO neuronal cell bodies and processes were double-labeled for both 5-HT and SERT in control P4 and P8 mice. Figure 6 illustrates that at P4, LSO neuronal cell bodies in control, moderate, and high Pb all double label for 5-HT and SERT, suggesting that 5-HT is taken up by LSO neurons by uptake through the SERT transporter. It is worth noting that SERT expression appears to be lower in the moderate dose of Pb at P4, which may explain the lower levels of 5-HT in LSO neurons compared to the controls and high dose of Pb. By P8, control LSO neuronal cell bodies are no longer double labeled for 5-HT and SERT, and these cell bodies no longer express 5-HT (Figure 6A1). Both the moderate and the high dose LSO neuronal cell bodies continue to be immunopositive both 5-HT and SERT, again suggesting that the prolonged expression of 5-HT in LSO neuronal cell bodies in Pb exposed mice is due to the prolonged expression of SERT (Figure 6B1 &B2). Taken together, these results indicate that Pb prolongs the period of normal developmental uptake of 5-HT through continued expression of SERT in LSO neurons. It is important to note that the effect of Pb on SERT appears to be specific for SERT expression on LSO somata. Total levels of SERT immunoreactivity within the LSO (which includes cell bodies and processes) did not change with Pb exposure (Figures 4 & 5). This suggests that Pb specifically targets SERT expression on LSO neuronal cell bodies. 3.1.5 Pb exposure decreases synaptophsyin (SYP) staining in LSO An important question is whether the Pb-induced extension of SERT expression on LSO neurons alters the structure and/or function of LSO neurons in the adult animal. Our previous studies found that both the moderate and high dose of Pb reduces SYP staining in LSO in P21 mice (Fortune and Lurie 2009) indicating that developmental Pb exposure reduces synaptic density in the adult LSO. However, it was not known when during development the decrease of SYP occurred. Therefore, brainstem sections from control and Pb treated mice were immunlabeled with SYP at P4 and P8 and synaptophysin staining quantified in LSO. At P4, no sSYP staining is observed in LSO (data not shown). By P8, the moderate dose of Pb induced a significant decrease in SYP labeling compared to controls at this time point (Figure 7). Interestingly, SYP labeling in the high Pb group was similar to controls at P8, but all Pb groups show decreased staining by P21. Thus, the moderate dose of Pb appears to induce an early decrease of SYP labeling LSO. 3.1.6 Moderate Pb exposure increases total 5-HT levels in the brainstem in adult animals LSO neurons appear to be a specific target for Pb. They express 5-HT through P8 presumably due to uptake through SERT, and show decreased SYP staining through adulthood. However, it is important to determine whether total brainstem levels of 5-HT are changed with Pb exposure, or whether Pb is only altering the expression of SERT. Therefore, 5-HT was measured in crude brainstem fractions (which contains all of the serotontergic raphe nuclei) of P 4 and P21 mice by HPLC analysis. Pb exposure did not result in significant changes in brainstem levels of 5-HT compared to controls at P4 (Figure 8). This suggests that the Pb-induced changes in the transient accumulation of 5-HT by LSO neurons at P4 are not the result of changes in total brainstem 5- HT levels. In contrast, at P21, the moderate, but not the high dose of Pb resulted in a significant increase in 5-HT (Figure 8). Discussion 4.1 Pb exposure prolongss the transient expression of 5-HT by LSO neurons The current study demonstrates that Pb exposure prolongs the transient expression of 5-HT by LSO neuronal cell bodies through postnatal day 8. In control mice, 5-HT expression in LSO cell bodies disappears by P8, however, the Pb-exposed mice continue to express 5-HT through P8. The expression of 5-HT by Pb-exposed LSO cell bodies is not permanent, by P21 both control and Pb-exposed LSO neurons are immune-negative for 5-HT. Thus Pb prolongs the normal developmental period during which LSO neuronal cell bodies are immunopositive for 5-HT. LSO neurons are not considered to be intrinsic serotonergic neurons. They project to the ipsilateral and contralateral inferior colliculus (IC) and use glycine and glutamate as a neurotransmitter (Thompson and Schofield, 2000). When brain levels of 5-HT are increased 6–9 fold compared to wild type (MAOA k/o mice), a subset of LSO neuronal cell bodies do transiently express 5-HT immunostaining starting from embryonic day 18 (E18). This immunoreactivity reaches maximal levels from birth (P0) to postnatal day 7 (P7) and then disappears by P10 (Cases et al., 1998). More recently, 5-HT immunoreactive LSO somata have been found in wild type mice at P1 which then disappears by P8 (Thompson, 2006). The 5-HT containing LSO neurons do not have the capacity for synthesizing 5-HT, as evidenced by the lack of tryptophan hydroxylase (TPH), a rate-limiting enzyme in 5-HT synthesis (Thompson and Thompson, 2009, and the current study). Instead, it is thought that LSO neurons take up extracellular 5-HT through SERT. In MAOA k/o mice, 5-HT and SERT staining is colocalized in LSO cell bodies at P0 and P5-6, and then all staining in the LSO somata for both proteins disappears by P15 (Thompson and Thompson, 2009). We see similar co-localization of 5-HT and SERT immunoreactivity postnatally, that then disappears. 4.1.1 Pb induced changes in SERT expression in LSO somata The current study found that the changes in 5-HT expression that were induced by Pb in LSO neurons were associated with similar changes in SERT expression. For example, at P4, control and Pb-exposed LSO neuronal somata are double-labeled with both 5-HT and the SERT. The moderate dose of Pb appears to result in a lower level of SERT expression on LSO neurons compared to the control and high dose of Pb, which could explain why LSO neurons exposed to the moderate dose of Pb have a delayed uptake of serotonin. Why the moderate dose of Pb differs from the high dose has yet to be determined, but previous studies in the developing hippocampus have documented that Pb has a dose-dependent bimodal effect on the developing hippocampus (Slomianka et al., 1989). Our results also show a dose-dependent bimodal effect in the developing LSO. By P8, control LSO cell bodies have lost their SERT immunostaining and are also immunenegative for 5-HT. A recent study of gene expression profiles in the developing Superior Olivary Complex (SOC) in the rat found that SERT was upregulated in the P4 SOC compared to the P25 SOC (Ehmann et al., 2013), lending additional support to the hypothesis that SERT expression is responsible for the transient uptake of 5-HT by developing LSO neurons. The current study found that developmental Pb exposure prolongs the expression of SERT on LSO cell bodies through P8, and these LSO neurons remain immunopositive for 5-HT at P8. In addition, Pb appears to specifically target LSO neurons. The serotonergic processes in LSO that are thought to originate from the raphe nuclei remain unaffected by Pb exposure. It will be very interesting to determine what happens in LSO if this transient uptake of 5-HT is eliminated using SSRI inhibitors. One might expect to see major alterations in synaptic structure in the LSO. 4.1.2 Pb and SERT expression The mechanism by which LSO somata transiently express SERT is not clear. The active (functional) SERT is located mainly in axon terminals and along the axons of 5-HT neurons (Zhou et al., 1998). In 5-HT neurons, the transcription factor Pet1 directly activates the transcription of genes that encode TPH, ADAC (L-amino acid decarboxylase), and SERT (Hendricks et al., 1999). However, this gene is not found in neurons transiently expressing SERT (Pfaar et al., 2002). In non-serotonergic thalamocortical neurons that transiently accumulate 5-HT through SERT, SERT proteins are transiently expressed in the axons and axon terminals, and are responsible for taking up 5-HT from the extracellular space and the 5-HT is then transported back to the cell bodies (Lebrand et al., 1996; Lebrand et al., 1998). However, the factors regulating SERT expression in non-serotonergic neurons are poorly understood. Nonetheless, there are several ways in which Pb might be affecting SERT expression. One mechanism by which Pb could be affecting SERT expression might be that Pb targets p38 MAP kinase (MAPK). Studies have found that inhibition of p38 MAP kinase decreases SERT proteins in the plasma membrane (Samuvel et al., 2005). In addition, when p38 MAPK expression is decreased by siRNAs, there is a concomitant decrease in cell surface expression of SERT. Taken together, these studies suggest that a decrease in p38 MAPK leads to a downregulation of SERT in the plasma membrane (Samuvel et al., 2005). Pb has been shown to activate the p38 MAPK pathway through phosphorylation of both ERK 1/2 and p38 MAPK (Cordova et al., 2004). Thus, it is possible that Pb activates p38 MAPK, leading to increased expression of SERT in LSO neuronal cell membranes. Although why developing LSO neuronal cell bodies appear to be a specific target for Pb remains to be determined. A second potential mechanism for how Pb might be affecting SERT expression involves thyroid hormones. Thyroid hormones have been shown to regulate the transient expression of SERT in thalamocortical neurons (Auso et al., 2001). In normal mice, the transient expression of the SERT gene disappears by P11, whereas its expression persists until P15 in hypothyroid rats (Auso et al., 2001). This effect of hypothyroidism is specific for the transient expression SERT, because there is no general delay in brain maturation and SERT is not changed in serotonergic raphe neurons. Prolonged expression of SERT in hypothyroid rats also leads to reduced axon terminal arborization and synaptogenesis within the cortical barrel fields, resulting in a smaller barrel area. Interestingly, Pb has been shown to reduce free thyroxine (FT4) levels in Pb exposed adolescents whose average blood Pb level is 7.3 ± 2.92 µg/dL (Dundar et al., 2006). In addition, Pb exposure reduces 131I uptake in animals and impairs the release of thyroidstimulating hormone (TSH) in children (Slingerland, 1955; Huseman et al., 1987). Therefore, a Pb-induced impairment in thyroid function is one potential mechanism by which Pb might prolong the transient expression of 5-HT and SERT in non-serotonergic LSO neurons. Future studies are needed to elucidate the precise mechanism by which Pb modulates the transient expression of SERT on LSO neuronal cell bodies. 4.1.3 Pb exposure decreases synaptophysin labeling in LSO The extension of the developmental window whereby Pb-exposed LSO neurons continue to express both SERT and 5-HT is correlated with decreased SYP labeling in both the P8 and P21 LSO. Thus Pb exposure results in decreased synaptic density within the adult LSO that can first be observed at P8. Recently, maternal Pb exposure in mice has been shown to decrease SYP expression in the hippocampus the Pb-exposed offspring at P21, indicating that Pb also affects synaptogenesis in non-sensory areas (Li et al., 2015). The transient expression of 5-HT by non-serotonergic neurons in sensory systems has been well studied. In other sensory systems such as rodent cortical barrel fields, precise regulation of the transient uptake of 5-HT by non-serotonergic sensory neurons has been shown to be critical for the proper formation of highly topographically organized sensory maps (Gaspar et al., 2003). Studies suggest that this transient uptake of 5-HT by non-serotonergic sensory neurons is necessary to remove 5-HT from the extracellular space, thereby maintaining proper extracellular concentrations of 5-HT during a critical period of development. It is well known that 5-HT functions as a neurotrophic factor during brain development prior to the time when it plays a role as a neurotransmitter (Luo et al., 2003). Depletion of 5-HT levels during development has a long lasting effect on synaptogenesis and brain maturation and therefore maintaining proper 5-HT levels during the critical period of brain development is crucial (Alvarez et al., 2002; Luo et al., 2003). For example, embryonic or neonatal depletion of 5-HT has been shown to decrease synaptic density, reduce spine density and complexity of cortical pyramidal neurons and hippocampal dentate granule cells (Mazer et al., 1997; Vitalis et al., 2007), and alter the neural morphology of somatosensory cortical barrel fields (Bennett-Clarke et al., 1994). In addition, knockout of SERT in thalamacortical neurons in mice results in long-term changes in spatial organization of cortical neurons, alterations in the pattern of thalamacortical neurons, and changes in the dendritic organization in sensory cortex (Chen et al., 2015). Thus disruption of SERT expression leads to excessive 5-HT and results in permanent impairments in the developing sensory cortex (Chen et al., 2015). The present study demonstrates that Pb exposure prolongs the normal uptake of 5-HT by LSO neurons. Thus, an intriguing hypothesis is that Pb shifts the extracellular concentrations of 5-HT during a critical window of neuronal development, thereby resulting in disrupted axonal arborization and synaptogenesis in LSO. In support of this, we found that Pb decreases synaptophysin immunostaining within the LSO of P8 and P21 mice. In addition, our moderate dose of Pb in the current study results in an increase in total 5-HT within the P21 brainstem, which might also contribute to the decreased SYP immunoreactivity observed in the P21 LSO (Fortune and Lurie, 2009). A limitation of the current study is that Pb was present in the drinking water throughout gestation and through P21. Thus we do not know the critical time during development when Pb has its effect on the LSO. Future studies will address this by exposing the developing mice to defined windows of Pb exposure. 4.1.4 5-HT and synaptogenesis Finally, in addition to its action as a trophic factor, 5-HT could modulate spontaneous activity in the LSO, thereby affecting LSO development. The application of 5-HT results in prolonged bursts of spontaneous inhibitory postsynaptic currents (IPSCs) in the developing gerbil LSO and intriguingly, this effect in not seen after P8 (Fitzgerald and Sanes, 1999). Thus, Pb could reduce the normal extracellular concentrations of 5-HT during the early postnatal period due to increased uptake into LSO neurons. This might lead to a decrease in IPSCs in the developing gerbil LSO, which could then alter proper axonal arborization and synapse formation. It is also worth noting that in the adult mouse brainstem, 5-HT does modulate the auditory brainstem response (ABR). Systemic serotonin depletion results in reduced ABR latencies, particularly for lower and mid-range frequencies (Papesh and Hurley, 2015). 4.1.5 Summary and Conclusion The current study demonstrates that Pb prolongs the normal developmental uptake of 5-HT by LSO neurons during a critical period of development, through the prolonged expression of SERT in LSO. SERT appears to be a target for Pb, and the extended uptake of 5-HT by LSO neurons is also correlated with the decreased expression of SYP and an increase in total 5-HT in the adult LSO. So how could the modulation of the serotonergic system by Pb be relevant to auditory processing in Pb-exposed children? Pb exposure in children has been linked to attention deficit disorder (Ha et al., 2009; Kim et al., 2013), and children with attention deficit disorder have deficits in the temporal processing of both speech and non-speech stimuli at the level of the brainstem (Jafari et al., 2015). Thus, developmental Pb exposure in humans could modulate 5- HT in the developing auditory system, leading to auditory temporal processing deficits. The effect of manipulating 5-HT levels on the developing human auditory system has not been examined, however, infants exposed to selective serotonin reuptake inhibitors (SSRIs) in utero show lower motor scores, more CNS stress signs, and higher arousal than unexposed infants (Salisbury et al., 2016). Of depressed pregnant women who seek treatment, one-third of them will take SSRIs. Increasing evidence suggests that exposure to SSRIs early in development alters the function of the serotonergic system and affects multiple systems including emotional, motor, and circadian (Maciag et al., 2006; McAllister et al., 2012). Thus, prenatal SSRI exposure has profound effects on infants, suggesting that prenatal regulation of 5-HT and SERT is essential for the normal development of the brain. Our study demonstrates that Pb exposure modulates the serotonergic system in the developing LSO. Additional studies are needed to fully elucidate the cellular and molecular mechanisms of the effect of Pb on the developing LSO, and the long-term consequences to the adult auditory system. We would like to thank Diane Brooks for her expert technical assistance. Supported by NIH NCRR P20 RR17670, NIH P20 RR015583 (D.I.L.; F.C.P.). NIA grant RO1AG031184-01 to F.C.P. Figure 1 5-HT immunofluorescence labeling in somata and processes within the LSO of postnatal day (P) 4 and P8 mice. A–C (P4): A) P4 LSO neurons are immunopositive for 5-HT (solid arrows) and some LSO processes are also immunopositive (broken arrows). B) P4 LSO neurons are immunopositive for 5-HT following moderate Pb treatment (arrows), but the immunoreactivity is decreased compared to the no Pb controls. 5-HT+ LSO processes are similar to the no Pb control (broken arrows). C) P4 high Pb LSO neurons have similar 5-HT immunoreactivity compared to the no Pb controls (arrows). Broken arrows=5-HT+ LSO processes. A1–C1 (P8): A1) By P8, control LSO neurons are no longer immunopositive for 5-HT (arrows) although there are many immunopositive LSO processes (broken arrows). Both the moderate (B1) and high (C1) Pb exposure results in 5-HT+ LSO neurons (arrows) and processes (broken arrows). Thus Pb extends the developmental window in which LSO neurons are immunopositive for 5-HT through P8. Bar=15 µ. Figure 2 Quantification of 5-HT immunostaining in LSO cell bodies and processes. At P4, moderate Pb exposure significantly decreases 5-HT expression in LSO neurons compared to controls while high Pb results in 5-HT expression that is similar to controls (A). At P8, the high dose of Pb results in significantly more 5-HT immunoreactivity in LSO cell bodies compared to controls, while the moderate dose of Pb shows a trend towards increased immunostaining (B). Neither the moderate nor the high dose of Pb affected the 5-HT immunostaining of LSO processes at either P4 (C) or P8 (D). Graphs represent mean IOD ± SEM. p<0.05; ANOVA with Dunnett’s post hoc test. Figure 3 LSO neurons of P4 mice do not express tryptophan hydroxylase (TPH) or monoamine oxidase-A (MAOA). A, B: The LSO (A) does not contain any TPH (green) immunopositive somata unlike neurons in the raphe pallidus (B) where there are many immunopositive neurons (arrows). C, D: Double immunolabeling of 5-HT (red) and MAOA (green) shows that the 5-HT positive neurons in the LSO (C) do not contain any MAOA positive somata (arrows), unlike neurons in the raphe pallidus where there is obvious double label (yellow) within many neurons (D; arrows). Scale bar = 30 µ. Figure 4 SERT immunolabeling in LSO. Pb does not result in significant changes in total SERT expression within the LSO at either P4 (A–C) or P8 (D–F) or P21 (G–I) (arrows). Controls (A, D, G), Moderate Pb (B, E, H), High Pb (C, F, I). Scale bar = 30µ (A–F), Scale bar = 50µ (GI). Figure 5 Quantification of SERT immunostaining confirms that Pb does not significantly affect total SERT immunoreactivity in the LSO. However, the amount of SERT immunoreactivity is higher at P8 (B) compared to P4 (A) and P21 (C). Graphs represent mean IOD ± SEM. p<0.05; ANOVA with Dunnett’s post hoc test. Figure 6 Micrographs of LSO double-labeled for 5-HT (red) and SERT (green) in the LSO in P4 and P8 mice. Neurons in LSO at P4 are double labeled for both 5-HT and SERT (arrows) in control, moderate Pb, and high Pb animals. By P8, control neurons are no longer double-labeled (A1), but LSO neuron in both the moderate (B1) and high Pb (C1) treatment groups continue to be double-labeled for both 5-HT and SERT. Scale bar = 15µ. Figure 7 Pb exposure decreases synaptophysin (SYP) immunolabeling in the LSO of P8 mice. A–C: Immunoreactivity for SYP (arrows) decreases with moderate Pb (B) compared to control (A). High Pb (C) does not result in changes in the SYP expression in the P8 LSO. D: Quantification of SYP immunostaining in the LSO confirms there is a statistically significant decrease in the moderate Pb treatment group. Graphs represent mean IOD ± SEM. p<0.05; ANOVA with Dunnett’s post hoc test. Scale bar = 30 µ. Figure 8 5-HT levels in brainstem, expressed as µg/g wet tissue. A) 5-HT levels in P4 mice. B) 5-HT levels in P21 mice. Graphs represent mean ± SEM. denotes significant change from the no Pb group at p<0.05; ANOVA with Dunnett’s post hoc test. Highlights Developmental Pb exposure prolongs transient expression of 5-HT in postnatal murine LSO neurons. Pb-induced 5-HT expression correlates with expression of SERT on LSO cell bodies. Prolonged expression of 5-HT after Pb exposure leads to decreased synapses in LSO. Decreased LSO synapses might contribute to cognitive deficits linked to Pb exposure. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. References Alvarado JC Fuentes-Santamaria V Henkel CK Brunso-Bechtold JK Alterations in calretinin immunostaining in the ferret superior olivary complex after cochlear ablation The Journal of comparative neurology 2004 470 63 79 14755526 Alvarez C Vitalis T Fon EA Hanoun N Hamon M Seif I Edwards R Gaspar P Cases O Effects of genetic depletion of monoamines on somatosensory cortical development Neuroscience 2002 115 753 764 12435414 Auso E Cases O Fouquet C Camacho M Garcia-Velasco JV Gaspar P Berbel P Protracted expression of serotonin transporter and altered thalamocortical projections in the barrelfield of hypothyroid rats Eur J Neurosci 2001 14 1968 1980 11860492 Bellinger DC Very low lead exposures and children's neurodevelopment Curr Opin Pediatr 2008 20 172 177 18332714 Bellinger DC Bellinger AM Childhood lead poisoning: the torturous path from science to policy J Clin Invest 2006 116 853 857 16585952 Bennett-Clarke CA Leslie MJ Lane RD Rhoades RW Effect of serotonin depletion on vibrissa-related patterns of thalamic afferents in the rat's somatosensory cortex The Journal of neuroscience : the official journal of the Society for Neuroscience 1994 14 7594 7607 7996198 Braun JM Kahn RS Froehlich T Auinger P Lanphear BP Exposures to environmental toxicants and attention deficit hyperactivity disorder in U.S. children Environ Health Perspect 2006 114 1904 1909 17185283 Breier JI Fletcher JM Foorman BR Klaas P Gray LC Auditory temporal processing in children with specific reading disability with and without attention deficit/hyperactivity disorder J Speech Lang Hear Res 2003 46 31 42 12647886 Caminos E Garcia-Pino E Martinez-Galan JR Juiz JM The potassium channel KCNQ5/Kv7.5 is localized in synaptic endings of auditory brainstem nuclei of the rat The Journal of comparative neurology 2007 505 363 378 17912742 Canfield RL Henderson CR Jr Cory-Slechta DA Cox C Jusko TA Lanphear BP Intellectual impairment in children with blood lead concentrations below 10 microg per deciliter N Engl J Med 2003 348 1517 1526 12700371 Cases O Lebrand C Giros B Vitalis T De Maeyer E Caron MG Price DJ Gaspar P Seif I Plasma membrane transporters of serotonin, dopamine, and norepinephrine mediate serotonin accumulation in atypical locations in the developing brain of monoamine oxidase A knock-outs The Journal of neuroscience : the official journal of the Society for Neuroscience 1998 18 6914 6927 9712661 Chen X Ye R Gargus JJ Blakely RD Dobrenis K Sze JY Disruption of Transient Serotonin Accumulation by Non-Serotonin-Producing Neurons Impairs Cortical Map Development Cell reports 2015 Cordova FM Rodrigues AL Giacomelli MB Oliveira CS Posser T Dunkley PR Leal RB Lead stimulates ERK1/2 and p38MAPK phosphorylation in the hippocampus of immature rats Brain research 2004 998 65 72 14725969 Devi CB Reddy GH Prasanthi RP Chetty CS Reddy GR Developmental lead exposure alters mitochondrial monoamine oxidase and synaptosomal catecholamine levels in rat brain Int J Dev Neurosci 2005 23 375 381 15927761 Dundar B Oktem F Arslan MK Delibas N Baykal B Arslan C Gultepe M Ilhan IE The effect of long-term low-dose lead exposure on thyroid function in adolescents Environmental research 2006 101 140 145 16360141 Ehmann H Hartwich H Salzig C Hartmann N Clement-Ziza M Ushakov K Avraham KB Bininda-Emonds OR Hartmann AK Lang P Friauf E Nothwang HG Time-dependent gene expression analysis of the developing superior olivary complex The Journal of biological chemistry 2013 288 25865 25879 23893414 Facoetti A Lorusso ML Paganoni P Cattaneo C Galli R Umilta C Mascetti GG Auditory and visual automatic attention deficits in developmental dyslexia Brain Res Cogn Brain Res 2003 16 185 191 12668226 Finkelstein Y Markowitz ME Rosen JF Low-level lead-induced neurotoxicity in children: an update on central nervous system effects Brain research Brain research reviews 1998 27 168 176 9622620 Fitzgerald KK Sanes DH Serotonergic modulation of synapses in the developing gerbil lateral superior olive J Neurophysiol 1999 81 2743 2752 10368394 Fortune T Lurie DI Chronic low-level lead exposure affects the monoaminergic system in the mouse superior olivary complex The Journal of comparative neurology 2009 513 542 558 19226511 Gaspar P Cases O Maroteaux L The developmental role of serotonin: news from mouse molecular genetics Nat Rev Neurosci 2003 4 1002 1012 14618156 Gilbert ME Mack CM Lasley SM Chronic developmental lead exposure increases the threshold for long-term potentiation in rat dentate gyrus in vivo Brain research 1996 736 118 124 8930316 Glotzer DE Freedberg KA Bauchner H Management of childhood lead poisoning: clinical impact and cost-effectiveness Med Decis Making 1995 15 13 24 guidelines C CDC blood lead guidelines. In 7898292 Ha M Kwon HJ Lim MH Jee YK Hong YC Leem JH Sakong J Bae JM Hong SJ Roh YM Jo SJ Low blood levels of lead and mercury and symptoms of attention deficit hyperactivity in children: a report of the children's health and environment research (CHEER) Neurotoxicology 2009 30 31 36 19100765 Hendricks T Francis N Fyodorov D Deneris ES The ETS domain factor Pet-1 is an early and precise marker of central serotonin neurons and interacts with a conserved element in serotonergic genes The Journal of neuroscience : the official journal of the Society for Neuroscience 1999 19 10348 10356 10575032 Hirst WD Price GW Rattray M Wilkin GP Serotonin transporters in adult rat brain astrocytes revealed by [3H]5-HT uptake into glial plasmalemmal vesicles Neurochemistry international 1998 33 11 22 9694037 Hurley LM Pollak GD Serotonin shifts first-spike latencies of inferior colliculus neurons The Journal of neuroscience : the official journal of the Society for Neuroscience 2005 25 7876 7886 16120790 Hurley LM Thompson AM Pollak GD Serotonin in the inferior colliculus Hearing research 2002 168 1 11 12117504 Huseman CA Moriarty CM Angle CR Childhood lead toxicity and impaired release of thyrotropin-stimulating hormone Environmental research 1987 42 524 533 3106028 Jafari Z Malayeri S Rostami R Subcortical encoding of speech cues in children with attention deficit hyperactivity disorder Clinical neurophysiology : official journal of the International Federation of Clinical Neurophysiology 2015 126 325 332 25066938 Jones LG Prins J Park S Walton JP Luebke AE Lurie DI Lead exposure during development results in increased neurofilament phosphorylation, neuritic beading, and temporal processing deficits within the murine auditory brainstem The Journal of comparative neurology 2008 506 1003 1017 18085597 Kim S Arora M Fernandez C Landero J Caruso J Chen A Lead, mercury, and cadmium exposure and attention deficit hyperactivity disorder in children Environmental research 2013 126 105 110 24034783 Landrigan PJ Todd AC Lead poisoning West J Med 1994 161 153 159 7941534 Lanphear BP Dietrich K Auinger P Cox C Cognitive deficits associated with blood lead concentrations < 10 microg/dL in US children and adolescents Public health reports 2000 115 521 529 11354334 Lasley SM Gilbert ME Presynaptic glutamatergic function in dentate gyrus in vivo is diminished by chronic exposure to inorganic lead Brain research 1996 736 125 134 8930317 Lasley SM Gilbert ME Glutamatergic components underlying lead-induced impairments in hippocampal synaptic plasticity Neurotoxicology 2000 21 1057 1068 11233752 Lebrand C Cases O Wehrle R Blakely RD Edwards RH Gaspar P Transient developmental expression of monoamine transporters in the rodent forebrain The Journal of comparative neurology 1998 401 506 524 9826275 Lebrand C Cases O Adelbrecht C Doye A Alvarez C El Mestikawy S Seif I Gaspar P Transient uptake and storage of serotonin in developing thalamic neurons Neuron 1996 17 823 835 8938116 Li N Zhang P Qiao M Shao J Li H Xie W The effects of early life lead exposure on the expression of P2X7 receptor and synaptophysin in the hippocampus of mouse pups Journal of trace elements in medicine and biology : organ of the Society for Minerals and Trace Elements 2015 30 124 128 Luo X Persico AM Lauder JM Serotonergic regulation of somatosensory cortical development: lessons from genetic mouse models Developmental neuroscience 2003 25 173 183 12966215 Lurie DI Brooks DM Gray LC The effect of lead on the avian auditory brainstem Neurotoxicology 2006 27 108 117 16162360 Maciag D Simpson KL Coppinger D Lu Y Wang Y Lin RC Paul IA Neonatal antidepressant exposure has lasting effects on behavior and serotonin circuitry Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology 2006 31 47 57 16012532 Maguire EP Mitchell EA Greig SJ Corteen N Balfour DJ Swinny JD Lambert JJ Belelli D Extrasynaptic glycine receptors of rodent dorsal raphe serotonergic neurons: a sensitive target for ethanol Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology 2014 39 1232 1244 24264816 Malynn S Campos-Torres A Moynagh P Haase J The pro-inflammatory cytokine TNF-alpha regulates the activity and expression of the serotonin transporter (SERT) in astrocytes Neurochemical research 2013 38 694 704 23338678 Mazer C Muneyyirci J Taheny K Raio N Borella A Whitaker-Azmitia P Serotonin depletion during synaptogenesis leads to decreased synaptic density and learning deficits in the adult rat: a possible model of neurodevelopmental disorders with cognitive deficits Brain research 1997 760 68 73 9237519 McAllister BB Kiryanova V Dyck RH Behavioural outcomes of perinatal maternal fluoxetine treatment Neuroscience 2012 226 356 366 23000627 Moreira EG Vassilieff I Vassilieff VS Developmental lead exposure: behavioral alterations in the short and long term Neurotoxicol Teratol 2001 23 489 495 11711252 Narboux-Neme N Pavone LM Avallone L Zhuang X Gaspar P Serotonin transporter transgenic (SERTcre) mouse line reveals developmental targets of serotonin specific reuptake inhibitors (SSRIs) Neuropharmacology 2008 55 994 1005 18789954 Papesh MA Hurley LM Modulation of auditory brainstem responses by serotonin and specific serotonin receptors Hearing research 2015 332 121 136 26688176 Pfaar H von Holst A Vogt Weisenhorn DM Brodski C Guimera J Wurst W mPet-1, a mouse ETS-domain transcription factor, is expressed in central serotonergic neurons Development genes and evolution 2002 212 43 46 11875656 Prins JM Brooks DM Thompson CM Lurie DI Chronic low-level Pb exposure during development decreases the expression of the voltage-dependent anion channel in auditory neurons of the brainstem Neurotoxicology 2010 31 662 673 20797405 Putter-Katz H Kishon-Rabin L Sachartov E Shabtai EL Sadeh M Weiz R Gadoth N Pratt H Cortical activity of children with dyslexia during natural speech processing: evidence of auditory processing deficiency J Basic Clin Physiol Pharmacol 2005 16 157 171 16285467 Sader-Mazbar O Loboda Y Rabey MJ Finberg JP Increased L-DOPA-derived dopamine following selective MAO-A or -B inhibition in rat striatum depleted of dopaminergic and serotonergic innervation British journal of pharmacology 2013 170 999 1013 23992249 Salisbury AL O'Grady KE Battle CL Wisner KL Anderson GM Stroud LR Miller-Loncar CL Young ME Lester BM The Roles of Maternal Depression, Serotonin Reuptake Inhibitor Treatment, and Concomitant Benzodiazepine Use on Infant Neurobehavioral Functioning Over the First Postnatal Month The American journal of psychiatry 2016 173 147 157 26514656 Samuvel DJ Jayanthi LD Bhat NR Ramamoorthy S A role for p38 mitogen-activated protein kinase in the regulation of the serotonin transporter: evidence for distinct cellular mechanisms involved in transporter surface expression The Journal of neuroscience : the official journal of the Society for Neuroscience 2005 25 29 41 15634764 Slingerland DW The influence of various factors on the uptake of iodine by the thyroid J Clin Endocrinol Metab 1955 15 131 141 13221634 Slomianka L Rungby J West MJ Danscher G Andersen AH Dose-dependent bimodal effect of low-level lead exposure on the developing hippocampal region of the rat: a volumetric study Neurotoxicology 1989 10 177 190 2616061 Thompson AM "Non-serotonergic" lateral superior olivary neurons of the neonatal mouse contain serotonin Brain research 2006 1122 122 125 17034765 Thompson AM Serotonin immunoreactivity in auditory brainstem neurons of the postnatal monoamine oxidase-A knockout mouse Brain research 2008 1228 58 67 18634763 Thompson AM Schofield BR Afferent projections of the superior olivary complex Microsc Res Tech 2000 51 330 354 11071718 Thompson AM Thompson GC Experimental evidence that the serotonin transporter mediates serotonin accumulation in LSO neurons of the postnatal mouse Brain research 2009 1253 60 68 19070605 Vitalis T Cases O Passemard S Callebert J Parnavelas JG Embryonic depletion of serotonin affects cortical development Eur J Neurosci 2007 26 331 344 17650110 Vitalis T Fouquet C Alvarez C Seif I Price D Gaspar P Cases O Developmental expression of monoamine oxidases A and B in the central and peripheral nervous systems of the mouse The Journal of comparative neurology 2002 442 331 347 11793338 Wright CM Conlon EG Auditory and visual processing in children with dyslexia Dev Neuropsychol 2009 34 330 355 19437207 Zhou FC Tao-Cheng JH Segu L Patel T Wang Y Serotonin transporters are located on the axons beyond the synaptic junctions: anatomical and functional evidence Brain research 1998 805 241 254 9733975
PMC005xxxxxx/PMC5124125.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 0372547 1869 Br J Nutr Br. J. Nutr. The British journal of nutrition 0007-1145 1475-2662 27609061 5124125 10.1017/S0007114516002944 NIHMS827488 Article The human milk oligosaccharide 2′-fucosyllactose attenuates the severity of experimental necrotising enterocolitis by enhancing mesenteric perfusion in the neonatal intestine Good Misty 12 Sodhi Chhinder P. 34 Yamaguchi Yukihiro 34 Jia Hongpeng 34 Lu Peng 34 Fulton William B. 34 Martin Laura Y. 34 Prindle Thomas Jr 34 Nino Diego F. 34 Zhou Qinjie 34 Ma Congrong 12 Ozolek John A. 56 Buck Rachael H. 7 Goehring Karen C. 7 Hackam David J. 34* 1 Divisions of Newborn Medicine, Children's Hospital of Pittsburgh, Pittsburgh, PA 15224, USA 2 Departments of Pediatrics, The University of Pittsburgh School of Medicine, Pittsburgh, PA 15224, USA 3 General Pediatric Surgery, Johns Hopkins University and Bloomberg Children's Center, Johns Hopkins Hospital, Baltimore, MD 21287, USA 4 Department of Surgery, Johns Hopkins University, Baltimore, MD 21287, USA 5 Pediatric Pathology, Children's Hospital of Pittsburgh, Pittsburgh, PA 15224, USA 6 Departments of Pathology, The University of Pittsburgh School of Medicine, Pittsburgh, PA 15224, USA 7 Abbott Nutrition, Columbus, OH 43215, USA * Corresponding author: D. J. Hackam, fax +1 410 502 5314, [email protected] 8 11 2016 9 9 2016 10 2016 26 11 2016 116 7 11751187 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Necrotising enterocolitis (NEC) is a common disease in premature infants characterised by intestinal ischaemia and necrosis. The only effective preventative strategy against NEC is the administration of breast milk, although the protective mechanisms remain unknown. We hypothesise that an abundant human milk oligosaccharide (HMO) in breast milk, 2′-fucosyllactose (2′FL), protects against NEC by enhancing intestinal mucosal blood flow, and we sought to determine the mechanisms underlying this protection. Administration of HMO-2′FL protected against NEC in neonatal wild-type mice, resulted in a decrease in pro-inflammatory markers and preserved the small intestinal mucosal architecture. These protective effects occurred via restoration of intestinal perfusion through up-regulation of the vasodilatory molecule endothelial nitric oxide synthase (eNOS), as administration of HMO-2′FL to eNOS-deficient mice or to mice that received eNOS inhibitors did not protect against NEC, and by 16S analysis HMO-2′FL affected the microbiota of the neonatal mouse gut, although these changes do not seem to be the primary mechanism of protection. Induction of eNOS by HMO-2′FL was also observed in cultured endothelial cells, providing a link between eNOS and HMO in the endothelium. These data demonstrate that HMO-2′FL protects against NEC in part through maintaining mesenteric perfusion via increased eNOS expression, and suggest that the 2′FL found in human milk may be mediating some of the protective benefits of breast milk in the clinical setting against NEC. Necrotising enterocolitis Human milk oligosaccharides 2′-Fucosyllactose Endothelial nitric oxide synthase Breast milk Necrotising enterocolitis (NEC) is a serious cause of morbidity and mortality in premature infants, and is a major public health concern due to the increasing number of premature infants(1). NEC is characterised by the sudden onset of ischaemia and necrosis in the small intestine typically, but can occur in the colon, affecting up to 12 % of premature infants, and has a mortality overall of up to 40 %(1–3), in part because current treatment options are limited to gut rest, broad-spectrum antibiotics and surgical resection of necrotic bowel(4). NEC is a gastrointestinal disease in its early stages, but quickly becomes a systemic disease, manifest by cardiorespiratory collapse and signs of systemic sepsis(1). Although the precise causes of NEC remain incompletely understood, our current understanding of the disease identifies a critical link between the administration of enteral formula and its development, as demonstrated by the finding that infants who are fed breast milk are largely protected from NEC development(5–8). In seeking to understand the mechanisms leading to NEC development, we(9–12) as well as others(13) have identified a critical role of the lipopolysaccharide (LPS) receptor toll-like receptor 4 (TLR4) in its pathogenesis. In a series of studies summarised in this recent review(14), we and others have shown that TLR4 activation in the intestinal epithelium leads to the loss of enterocytes through apoptosis(12,15,16), followed by delayed repair through inhibition of migration and TLR4-mediated loss of intestinal stem cells(17). These factors lead to the translocation of bacteria and LPS into the circulation where endothelial TLR4 activation occurs, resulting in a loss of endothelial nitric oxide synthase (eNOS), which leads to impaired perfusion and the development of experimental NEC(18). Several groups have sought to identify potential components of breast milk that could be responsible for exerting the protection against NEC in premature infants(12,18–22), and many of these studies have shed light on the beneficial activities of human milk oligosaccharides (HMO)(23–30). HMO are complex glycans that are abundant in human breast milk and not present in infant formula(23) and have beneficial effects on the intestine. There are various pathways by which HMO exert their protective effects on the gut, including selective consumption by protective gut microbes(28), as well as by binding to and thus preventing the adherence of bacteria to the intestinal epithelium(31). HMO have also been shown to be absorbed into the circulation and excreted into the urine, suggesting that these molecules could also exert systemic effects(32,33), and thus may play a protective role against a disease such as NEC, which has both intestinal and systemic components(1). Importantly, there has been no mechanism that links the potential benefits of HMO against NEC – a disease with largely systemic manifestations and not reliably linked to a particular microbe. In the current study, we seek to evaluate the role – if any – of an abundant HMO found in breast milk – namely, 2′-fucosyllactose (HMO-2′FL) – in the protection against experimental NEC. To do so, we now focus on our recent observation that TLR4 plays a critical role in NEC development through its ability to modulate intestinal perfusion via eNOS expression as described above(18). We specifically hypothesise that administration of HMO-2′FL will reduce the incidence and severity of NEC in newborn mice through effects on mesenteric perfusion via eNOS regulation. Methods Statement of ethics The described animal experiments in this study were carried out in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. Animal protocols were approved by the University of Pittsburgh's Animal Care and Use Committee (protocol 12040382) and Johns Hopkins University (protocol M014M362). Cells, materials, mice and reagents LPS (Escherichia coli 0111:B4 purified by gel filtration chromatography, >99 % pure) was obtained from Sigma-Aldrich. eNOS inhibitors – N5-(1-iminoethyl)-l-ornithine dihydrochloride (L-NIO) and diphenyleneiodonium chloride (DPI) – were obtained from Santa Cruz Biotechnology. The following antibodies were obtained: eNOS (Enzo Life Sciences) and platelet endothelial cell adhesion molecule 1 (PECAM-1; BD Biosciences). The fluorescein-labelled Lycopersicon esculentum (tomato) lectin (Vector Laboratories) was used as a tracer of intestinal perfusion(18). C57BL/6 and eNOS−/− (B6·129P2-Nos3tm1Unc/J) mice were obtained from the Jackson Laboratory. HMO-2′FL was produced through a proprietary fermentation process (Kyowa Hakko Bio Co. Ltd). Purity (95·3 %) was established by high-performance ion chromatography with pulsed amperometric detection using relative peak area comparisons. Endotoxin level (0·375 EU/mg) was estimated by limulus assay (limulus amebocyte lysate QCL-1000; Lonza). The cell line HUV-EC-C (HUVEC) (ATCC® CRL-1730™) was obtained from American Type Culture Collection (ATCC) and maintained in endothelial cell basal medium (Cell Applications, Inc.) with low serum growth supplement (Thermo Fisher) at 37°C with 5 % CO2. Where indicated, cells were treated with HMO-2′FL (dose 100 μg/ml) 1 h before LPS administration (dose 1 μg/ml), and the induction of TLR4-mediated pro-inflammatory cytokines and eNOS expression were measured by quantitative real-time PCR (qRT-PCR) 6 h after administration of LPS. Immunohistochemistry and immunofluorescence were performed as described by Afrazi et al.(34) and evaluated using a Nikon AZ-C2+ confocal microscope (Nikon Instruments). SDS-PAGE and qRT-PCR were performed as described by Good et al.(12). Induction of necrotising enterocolitis in neonatal mice All animal experiments were approved by the University of Pittsburgh Animal Care and Use Committee. Experimental NEC was induced in 7–10-d-old mice as previously described(10,34). In brief, we used formula gavage (Similac Advance infant formula (Abbott Nutrition):Esbilac (PetAg) canine milk replacer in a 2:1 ratio at a dose of 50 μl NEC formula/g body weight) five times/d, which was supplemented with enteric bacteria obtained from an infant with severe NEC as described in Good et al.(35) to simulate the dysbiosis seen in human NEC. As both the infant formula and the puppy milk replacer are commercially available, a brief description of each component of the NEC formula is as follows. Similac Advance contains the following ingredients: water, non-fat milk, lactose, high oleic safflower oil, soya oil, coconut oil, galactooligosaccharides and whey protein concentrate. Less than 0·5 % of C. cohnii oil, M. Alpina oil, β-carotene, lutein, lycopene, ascorbic acid, soya lecithin, monoglycerides, potassium citrate, calcium carbonate, potassium chloride, carrageenan, ferrous sulphate, magnesium chloride, choline chloride, choline bitartrate, taurine, m-inositol, calcium phosphate, zinc sulphate, potassium phosphate, d-α-tocopheryl acetate, niacinamide, calcium pantothenate, l-carnitine, vitamin A palmitate, cupric sulphate, thiamine chloride hydrochloride, riboflavin, pyridoxine hydrochloride, folic acid, manganese sulphate, phylloquinone, biotin, sodium selenate, vitamin D3, cyanocobalamin, salt, potassium hydroxide and nucleotides (AMP, cytidine 5′-monophosphate, disodium GMP, disodium uridine 5′-monophosphate). Esbilac canine milk replacer contains water, condensed, skimmed milk, soyabean oil, sodium caseinate, cream, calcium caseinate, l-methionine, l-arginine, calcium carbonate, choline chloride, lecithin, magnesium sulphate, potassium chloride, monopotassium phosphate, salt, tricalcium phosphate, carrageenan, dipotassium phosphate, dicalcium phosphate, ascorbic acid, ferrous sulphate, sodium hydroxide, zinc sulphate, vitamin A supplement, vitamin E supplement, niacin supplement, calcium pantothenate, copper sulphate, maltodextrins, thiamine hydrochloride, pyridoxine hydrochloride, riboflavin, manganese sulphate, vitamin D3 supplement, potassium citrate, potassium iodide, folic acid, vitamin B12 supplement and biotin. In addition to the formula used to induce NEC, mice were also subjected to 10 min of hypoxia (5 % O2, 95 % N2) via a chamber (Billups-Rothenberg) twice a day for 4 d. Where indicated, mice were supplemented with HMO-2′FL (5 mg/ml of formula, 0·25 mg/g body weight, once daily). The eNOS inhibitor DPI was dissolved in water–dimethyl sulfoxide (DMSO) (10:90), and a 25-mm stock was prepared of which 2 μl was added to each ml of NEC formula with each feeding as indicated. A 50-mm stock of L-NIO was made with water and 1 μl was added to each ml of NEC formula with each feeding. Necrotising enterocolitis severity assessment Mouse terminal ileal sections were assessed by histology for the degree of mucosal injury according to our previously published scoring system from 0 (normal) to 3 (severe injury)(36), gross morphology, weight loss and by the expressions of pro-inflammatory cytokines by qRT-PCR. Assessment of intestinal perfusion in neonatal mice In mice in which experimental NEC was induced, intestinal perfusion was assessed as previously described in the study by Yazji et al. (18). In brief, neonatal mice were intra-cardiacally injected with the fluorescently labelled tomato lectin (1 mg/ml, dose 5 μl/g body weight) for 5 min before euthanasia and intestinal harvest. Terminal ilea were co-stained with PECAM-1, an endothelial cell marker, and whole mounts were evaluated for PECAM-1 and tomato lectin fluorescent emission via confocal microscopy (Nikon AZ-C2+ confocal microscope). The quantification of the perfusion (villous perfusion index) was assessed as done previously(18). In brief, the sum total of tomato lectin fluorescence divided by the total volume of the mucosal vasculature is expressed as a percentage. Measurement of nitric oxide within the intestine Nitric oxide was measured within the intestine of mice in the indicated treatment groups using the OxiSelect™ In Vitro Nitric Oxide (Nitrite/Nitrate) Assay Kit (Cell Biolabs, Inc.) according to the manufacturer's instructions. 16S ribosomal RNA amplicon sequence analysis Raw paired-end reads output by the MiSeq platform was merged into consensus fragments by FLASH(37) and subsequently filtered for quality (maximum error rate 1 %) and length (minimum 200 bp) using Trimmomatic(38) and Quantitative Insights into Microbial Ecology (QIIME)(39,40). Spurious hits to the PhiX control genome were identified using Nucleotide Basic Local Alignment Search Tool (BLASTN) and removed. Passing sequences were trimmed of primers, evaluated for chimeras with UCLUST (de novo mode)(41) and screened for mouse-associated contaminants using Bowtie2(42) followed by a more sensitive BLASTN search against the GreenGenes 16S database(43). Chloroplast and mitochondrial contaminants were detected and filtered using the Ribosomal Database Project classifier(44) with a confidence threshold of 80 %. High-quality 16S sequences were assigned to operational taxonomic units (OTU) with a taxonomic lineage using Resphera Insight. To remove common bacterial contaminant species associated with DNA extraction kits, we first identified two well-known contaminant species present in all samples (Variovorax paradoxus and Janthinobacterium lividum) and computed Pearson's ρ coefficient between these two species and all other OTU. Those OTU with Pearson's ρ > 0·2 for both species were removed. We further removed low-abundance contaminants on the basis of taxa identified by Salter et al.(45). To normalise across samples, 16S profiles were subsampled to 1124 sequences per sample before downstream statistical comparisons. Differentially abundant taxa were detected using the negative binomial test with P-value correction using the false discovery rate(46,47). β-Diversity calculations and principal coordinate analysis (PCoA) was performed using QIIME(39,40). Differences in β-diversity were evaluated using the Mann–Whitney U test. Statistical analysis In our experimental NEC experiments, each mouse represents an individual symbol on a graph. Because of different numbers of pups randomised to each group, the numbers may vary between data sets, but each is shown in the relevant data set, and statistical analysis was performed using ANOVA for multiple groups or Student's t test for paired groups using PRISM version 6.0 (GraphPad). Statistical significance was accepted at P < 0·05. Quantitative real-time PCR Quantitative real-time PCR was performed as previously described using the Bio-Rad CFX96 Real-Time System (Bio-Rad)(36) using the primers listed in Table 1 relative to the housekeeping gene ribosomal protein large, P0 (RPL0). Results Necrotising enterocolitis severity in mice is reduced by the addition of the human milk oligosaccharide 2′-fucosyllactose to the formula As shown in Fig. 1, animals subjected to this experimental model developed gross evidence of small intestinal ischaemia, pneumatosis intestinalis and inflammation (Fig. 1(a)), as well as histological evidence of mucosal destruction and inflammatory influx (Fig. 1(b)), compared with breast-fed controls. Importantly, the administration of HMO-2′FL (0·25 mg/g body weight once daily during the model) significantly reduced the severity of NEC as compared with mice that were administered standard NEC formula, as was determined by assessment of the degree of gross and microscopic intestinal mucosal injury (Fig. 1(a) and (b)), preservation of daily weights (Fig. 1(c)), mucosal severity score (Fig. 1(d)) and by the expressions of the pro-inflammatory mediators that our laboratory(9–11,16) and others(48) have shown to be important in NEC, including inducible nitric oxide synthase (iNOS), IL-6, IL-1β and TLR4 (Fig. 1(e)). Taken together, the supplementation of HMO-2′FL in the formula protects against the development of experimental NEC. Given the possibility that there could be microbial effects, we next sought to describe the microbiota of the animals treated with HMO-2′FL. As shown in Fig. 2 and 3, taxonomic profiles at the family level reveal several well-represented taxa including Enterobacteriaceae, Lactobacillaceae and Clostridiaceae. We found Enterobacteriaceae to be generally more abundant in the experimental NEC group, whereas Lactobacillaceae appeared to be more abundant in the breast-fed group. Furthermore, differential abundance analysis between the mouse groups confirmed the associations with Enterobacteriaceae and Lactobacillaceae between the breast-fed controls as compared with animals that underwent experimental NEC; however, we did not see an association with HMO treatment (Fig. 2(b)). For Ruminococcaceae and Enterococcaceae, we observed some significant differences between HMO and their respective control groups (Fig. 2(b)). For example, Enterococcaceae was significantly more abundant in the breast-fed control group relative to the HMO-2′FL-treated group (Fig. 2(b)). We next wanted to analyse the microbial community diversity between the different treatment groups. β-Diversity analysis enabled us to compare total community composition among samples to look for broader associations beyond any single taxon. Using the Hellinger distance metric in QIIME, followed by PCoA, we observed a strong association with breast-fed compared with formula-fed (FF) animals (Fig. 3). Moreover, we can compare the β-diversity distances within and between the groups to see whether there are distinctive associations. In this case, we found that for both HMO-fed groups, the intestinal microbiomes of the mice were significantly more homogeneous than the breast-fed and FF groups (Fig. 3). The addition of the human milk oligosaccharide 2′-fucosyllactose prevented the inhibition on mesenteric perfusion observed in experimental necrotising enterocolitis in newborn mice In order to define the mechanisms by which HMO-2′FL could protect against the development of NEC, we next turned our attention to the effects of HMO-2′FL on the potential regulation of eNOS-mediated mesenteric blood flow by performing mesenteric micro-angiography in mice that underwent experimental NEC. To do so, we first assessed the degree of mesenteric perfusion in the mouse model of experimental NEC in the presence or absence of HMO-2′FL supplementation to the formula. As shown in Fig. 4, the intestinal perfusion in mice that were FF was significantly reduced compared with breast-fed controls, consistent with our previous findings(18). Importantly, the addition of HMO-2′FL restored perfusion to levels comparable with that of mice that were breast-fed on each day of the model, as observed in Fig. 4(a) and quantified in Fig. 4(b). The improvement in perfusion corresponded to a pink and healthy appearing intestine in breast-fed mice as compared with the dusky bowel seen in NEC, which was reversed in the presence of HMO-2′FL towards a healthy, pink and perfused appearance. Taken together, these findings illustrate that HMO-2′FL restores intestinal perfusion in NEC. We next sought to explore the potential mechanisms involved and focused on the effects of the vasodilatory molecule eNOS. Human milk oligosaccharide 2′-fucosyllactose maintains the expression of endothelial nitric oxide synthase in necrotising enterocolitis, and thus restores intestinal perfusion in neonatal mice Having now shown that HMO-2′FL restores mesenteric perfusion in NEC, we next sought to investigate the possibility that HMO-2′FL may affect the expression of eNOS. To do so, we assessed the expression of eNOS in the intestine of mice with and without NEC in the presence or absence of HMO-2′FL. As shown in Fig. 4, the expression of eNOS in the intestine was significantly reduced in FF mice compared with breast-fed control mice, and this was restored after HMO-2′FL induction (Fig. 4(c) and (d (i, ii))). The finding that HMO-2′FL could restore the expression of eNOS in the newborn gut in the setting of NEC raises the intriguing possibility that perhaps the protection against NEC by HMO-2′FL occurs via induction of eNOS expression. To assess this directly, we next utilised two approaches to inhibit eNOS – the use of pharmacological inhibitors and the use of eNOS knockout mice – and sought to determine whether HMO-2′FL could still protect against NEC development under conditions of eNOS inhibition. As shown in Fig. 5, we next treated mice with eNOS inhibitors L-NIO dihydrochloride and DPI. Importantly, under conditions of eNOS inhibition (verification in the online Supplementary Fig. S1(a)), HMO-2′FL failed to protect against NEC as demonstrated by histology, the pro-inflammatory cytokine IL-1β and NEC severity score (Fig. 5(a) and (c)), revealing that HMO-2′FL acts in part via eNOS activity. To assess whether eNOS expression was required for the protection exerted by HMO-2′FL against NEC, we next performed studies in eNOS−/− mice (Fig. 6, online Supplementary Fig. S1(b)) and as shown in the study by Shesely et al. (49). As shown in Fig. 6, eNOS−/− mice were significantly vulnerable to experimental NEC, consistent with our previous studies(18), and administration of HMO-2′FL did not confer protection as demonstrated by a disruption in the intestinal mucosal architecture, increase in pro-inflammatory cytokines and NEC severity, as well as significantly reduced intestinal perfusion (Fig. 6(a) and (d)). Taken together, these findings demonstrate that HMO-2′FL is protective against experimental NEC in a manner that is dependent on the expression and function of eNOS. Lipopolysaccharide decreases the expression of endothelial nitric oxide synthase in HUVEC endothelial cells and this decrease is attenuated by human milk oligosaccharide 2′-fucosyllactose In the final series of studies, we sought to explore in greater detail the potential mechanisms by which HMO-2′FL could regulate eNOS expression – and therefore mesenteric perfusion – in the newborn gut. To do so, we first examined whether HMO-2′FL could regulate the mRNA expression of eNOS in HUVEC endothelial cells in the presence of either LPS as a TLR4 ligand or LPS in the presence of HMO-2′FL to mimic the situation observed in the mesenteric endothelium. As shown in Fig. 7, TLR4 activation reduced the expression of eNOS mRNA in HUVEC (Fig. 7(a)), consistent with our previous findings in primary cultured endothelial cells(18). Importantly, the addition of HMO-2′FL restored eNOS expression as demonstrated by immunocytochemistry (Fig. 7(b)), consistent with the intestinal protein expression shown in Fig. 4(d), and demonstrated a clear role of HMO-2′FL in the regulation of eNOS gene expression and function. Discussion In the current study, we utilised an experimental mouse model of NEC, which recapitulates several of the features seen in premature infants with NEC, including gross evidence of small intestinal ischaemia, pneumatosis intestinalis and inflammation, as well as histological evidence of mucosal destruction(12,15,16,18), to demonstrate that the administration of formula containing an abundant HMO in human milk – namely, 2′FL – significantly reduces the severity of experimental NEC in newborn mice. In seeking to define the mechanisms involved, we and others have shown that TLR4 signalling in the gut leads to NEC via mucosal disruption, leading to LPS translocation and a loss of eNOS, which results in impaired perfusion and the development of intestinal ischaemia(18,50). We have now determined that HMO-2′FL-supplemented formula acts by regulating the degree of blood flow to the newborn intestine via eNOS, and that this novel pathway may explain some of the previously unrecognised benefits of this breast milk oligosaccharide. The current findings may shed light on the impact of other aspects of maternal breast milk production that could impact on the likelihood of NEC development to occur. Specifically, previous authors have noted that nearly 20 % of women in the population are homozygous for common mutations in the secretor gene fucosyltransferase 2 (FUT2) and are unable to produce HMO-2′FL(51), which has been shown to impact the intestinal microbiota of their infants(52). The studies on the relationship of FUT2 mutations and risk for NEC to date have focused predominantly on the phenotype or genotype of the baby rather than the mother(51,52). We now speculate that any increase in NEC in premature infants receiving milk from homozygous FUT2 mutant mothers may reflect a lack of endogenous 2′FL, suggesting that supplementing their milk with 2′FL could benefit their premature infants through NEC protection. Our findings extend our knowledge on the mechanisms by which HMO act on the neonatal intestine, and provide additional insights into the benefit of breast milk that is observed in premature infants at risk for developing NEC. Specifically, the molecular details underlying the capacity of two genera of gut bacteria to consume HMO have been demonstrated, including Bifidobacterium(53) and Bacteroidetes(54). Indeed, the presence of gram-negative bacteria including E. coli and Klebsiella has been associated with NEC outbreaks, and are commonly seen in the blood and stool of patients with this disease(55). HMO have also been shown to have structural homology to cell surface glycans, and can therefore act as molecular mimics, or decoys, by binding to luminal bacteria, and thereby prevent binding to the intestinal epithelium(56), thus potentially reducing the degree of bacterial inflammation and activation of toll-like receptors on the intestinal epithelium, which we have shown to be required for NEC development(10,15,17,34,36,57). Moreover, analysis of the microbiota revealed that supplementation of HMO-2′FL to the experimental NEC formula increased the faecal content of Ruminococcaceae, a family of bacteria of class Clostridia, which typically increase in premature infants as they mature and approach 36 weeks of postconceptional age(58). Although HMO-2′FL appears to affect the microbiota, these changes do not seem to be the primary mechanism of protection against NEC. The current findings extend the studies by Goehring et al.(32) who demonstrated that in addition to these local effects HMO are absorbed systemically in human infants, which in this case explains the effect on mesenteric perfusion through the activity of the vasodilator eNOS. Furthermore, 2′FL has been demonstrated in the plasma of breast-fed infants, although in lesser abundance than several other HMO(33), and in the urine of breast milk-fed premature infants(59). Although our studies did not directly test the systemic absorption of 2′FL, other authors have elegantly shown that HMO are absorbed in the intestine and can be measured in the serum and urine of rats, demonstrating the significance of the effects of HMO on other species(19,60). In determining which HMO are present in mouse breast milk, Prieto et al.(61) have shown that 2′FL is not present in mouse milk, while 3’FL indeed is. Importantly, the biological significance of the differential expression of HMO in breast milk across various species is unknown. It is noteworthy that the current study is supported by previous studies by Jantscher-Krenn et al.(19) who showed that the addition of the HMO disialyllacto-N-tetraose prevents NEC in neonatal rats. Although these studies did not provide a mechanism to explain these protective effects, and indeed the effects may be structure specific, we now speculate that the protection achieved was similar to that observed in the current study, and involved an enhanced mesenteric perfusion. Additional effects on the microbiota and on bacterial enterocyte signalling may also play a role in the protection in these two studies, and additional work will be required to determine these precise effects. It is possible that 2′FL gets absorbed into the circulation, attenuates endothelial TLR4 signalling and subsequently maintains eNOS levels. Moreover, it is likely that breast milk has various direct TLR4 inhibitors, as we have recently identified a class of small molecule inhibitors of TLR4(62), which are themselves oligosaccharides, suggesting that HMO-2′FL itself may inhibit TLR4 signalling. In support of this possibility, the expression of TLR4 was reduced in the mucosa of mice after administration of HMO-2′FL. Furthermore, it has recently been reported that HMO-2′FL modulates the expression of cluster of differentiation 14 (CD14), a molecule that is essential for LPS binding and optimal TLR4 signalling(63). The current study thus advances our understanding of the mechanisms by which breast milk may protect against the development of NEC. In summary, the current findings reveal an important role for HMO-2′FL in the protection against the development of NEC in newborn mice through the maintenance of mesenteric perfusion via a mechanism that requires the expression of eNOS. These findings expand our understanding about the protective benefits of breast milk for premature infants, in which HMO are present in high amounts, and also indicate that in addition to the established roles of HMO on the microbiota or on the epithelial surface, these molecules may have key roles when absorbed systemically. Although our preclinical studies were performed in neonatal mice rather than premature infants, the current studies also suggest that 2′FL is one of the components within human breast milk found to be protective against NEC, which remains an important goal of neonatal nutrition research. Furthermore, these studies reinforce the statements by the American Academy of Pediatrics Section on Breastfeeding and the European Society for Pediatric Gastroenterology, Hepatology and Nutrition that the preferred feeding for premature infants should be the mother's own milk with donor human milk from an established human milk bank as a second choice if the mother's milk supply is inadequate(64,65). Supplementary Material supplemental figure Acknowledgements M. G. is supported by K08DK101608 from the National Institutes of Health and the Children's Hospital of Pittsburgh of the UPMC Health System. D. J. H. is supported by a sponsored research grant from Abbott Nutrition, as well as R01DK083752 and R01GM078238 from the National Institutes of Health. M. G., C. P. S., R. H. B., K. C. G. and D. J. H. designed and/or performed experiments, analysed results, wrote and edited the manuscript; Y. Y., H. J., P. L., W. B. F., L. Y. M., T. P., D. F. N., Q. Z. and C. M. performed experiments and analysed data. Abbreviations 2′FL 2′-fucosyllactose eNOS endothelial nitric oxide synthase HMO human milk oligosaccharide LPS lipopolysaccharide NEC necrotising enterocolitis TLR4 toll-like receptor 4 Fig. 1 The addition of the human milk oligosaccharide (HMO) 2′-fucosyllactose to infant formula attenuates necrotising enterocolitis (NEC) severity in newborn mice. (a) Representative gross images of the intestine from wild-type neonatal mice that were either breast-fed (BF) or induced to develop NEC in the absence or presence of HMO in their feeds. (b) Representative haematoxylin–eosin micrographs of the terminal ileum of wild-type mice that were either BF, breast-fed with daily administration of HMO (BF + HMO) or received NEC formula with or without the addition of HMO to their feeds (FF or FF + HMO). (c) Daily body weights (g) of experimental animals for the duration of the NEC model. , BF; , FF; , BF + HMO; , FF + HMO. (d) NEC severity score (0–3) assigned by a pathologist blinded to the study conditions and treatment groups. (e) The ratio of the measured mRNA expression of the pro-inflammatory genes within the intestine including inducible nitric oxide synthase (iNOS), IL-6, IL-1β and toll-like receptor 4 (TLR4) of the indicated groups relative to the housekeeping gene RPLO, as in the study by Good et al.(12). * P < 0·05 v. BF control animals, ** P < 0·05 v. FF. Values are means with their standard errors. Scale bar is 50 μm. Representative of three separate experiments, where each mouse represents an individual symbol on a graph. qRT-PCR, quantitative real-time PCR. Fig. 2 Characterisation of the microbiota in mice treated with the human milk oligosaccharide (HMO) 2′-fucosyllactose (2′FL). (a) Taxonomic profiles at the family level for the microbiotas of pups treated with HMO-2′FL with necrotising enterocolitis (NEC) or breast-fed controls. , Other taxa; , Proteobacteria_Alphaproteobacteria_Sphingomonadales_Sphingomonadaceae; , Proteobacteria_Gammaproteobacteria_Pseudomonadales_Moraxellaceae; , Proteobacteria_Gammaproteobacteria_Pseudomonadales_Pseudomonadaceae; , Firmicutes_Bacilli_Bacillales_Staphylococcaceae; , Proteobacteria_Gammaproteobacteria_Pasteurellales_Pasteurellaceae; , Firmicutes_Clostridia_Clostridiales_Incertae sedis; , Firmicutes_Clostridia_Clostridiales_Ruminococcaceae; , Firmicutes_Bacilli_Lactobacillales_Enterococcaceae; , Firmicutes_Bacilli_Lactobacillales_Streptococcaceae; , Firmicutes_Clostridia_Clostridiales_Clostridiaceae 1; , Firmicutes_Bacilli_Lactobacillales_Lactobacillaceae; , Proteobacteria_Gammaproteobacteria_Enterobacteriales_Enterobacteriaceae. (b) Differential abundance analysis of the indicated treatment groups. , Enterobacteriaceae; , Lactobacillaceae; , Ruminococcaceae; , Enterococcaceae. BF, breast-fed; CNTRL, control; FF, formula fed. * Adjusted P < 0·05; adjusted P < 0·01; * adjusted P < 0·001. Fig. 3 β-Diversity analysis of the microbiota in mice treated with the human milk oligosaccharide (HMO) 2′-fucosyllactose. (a, b) β-Diversity analysis of the microbiota of the indicated treatment groups. Differences in β-diversity were evaluated using the Mann–Whitney U test. BF, breast-fed; CNTRL, control; FF, formula fed. * P < 0·05; P < 0·01; * P < 0·001. , BF + CNTRL; , BF + HMO; , FF + CNTRL; , FF + HMO. PC1, principal component 1; PC2, principal component 2. Fig. 4 Formula supplementation with the human milk oligosaccharide (HMO) 2′-fucosyllactose enhances mesenteric perfusion in experimental necrotising enterocolitis (NEC) via maintenance of intestinal endothelial nitric oxide synthase (eNOS) expression. (a) Representative confocal micrographs of terminal ileal whole mounts from wild-type neonatal mice from the indicated treatment groups: breast-fed controls (BF), breast-fed with HMO (BF + HMO, experimental NEC, formula fed (FF) and FF with HMO (FF + HMO), as well as mice that in addition to the experimental NEC model and HMO treatment also received the eNOS inhibitors N5-(1-iminoethyl)-l-ornithine dihydrochloride (L-NIO) (FF + HMO + L-NIO) or diphenyleneiodonium chloride (DPI) (FF + HMO + DPI) with their formula. After intra-cardiac injection with the fluorescent tracer tomato lectin, which correlates with intestinal blood flow (green), whole mounts were immunostained for platelet endothelial cell adhesion molecule 1 (PECAM-1) to assess the intestinal microvasculature (red). (b) Graph representing villus perfusion index as described in the Materials and methods section expressed as a percentage. (c) Intestinal eNOS mRNA expression by quantitative real-time PCR (qRT-PCR) in the indicated treatment groups relative to the housekeeping gene RPLO. (d) SDS-PAGE immunoprecipitation Western blot (IP-WB) performed on the terminal ileum of wild-type mice in the indicated treatment groups. IgG is shown as a loading control (i); densitometry quantification with Image J performed on three samples per group in (ii). * P < 0·05 v. breast-fed control animals, ** P < 0·05 v. FF. Values are means with their standard errors. Scale bar is 50 μm. Representative of at least three separate experiments, where each mouse represents an individual symbol on a graph. Fig. 5 The human milk oligosaccharide (HMO), 2′-fucosyllactose failed to protect against necrotising enterocolitis (NEC) in the presence of endothelial nitric oxide synthase inhibition. (a) Representative haematoxylin–eosin micrographs of the terminal ileum of wild-type mice in the indicated treatment groups: breast-fed controls (Ctrl), formula fed (FF), FF plus HMO (FF + HMO) and FF + HMO plus diphenyleneiodonium chloride (DPI) (FF + HMO + DPI) or FF + HMO and N5-(1-iminoethyl)-l-ornithine dihydrochloride (L-NIO) (FF + HMO + L-NIO). (b) IL-1β mRNA expression from the terminal ileum of the mice in the indicated treatment groups (, breast-fed mice; , FF mice with indicated treatment groups) relative to the housekeeping gene RPLO. * P < 0·05 v. breast-fed control (Ctrl) animals, P < 0·05 v. FF control (Ctrl), * P < 0·05 v. FF + HMO. (c) NEC severity score assessed by a pathologist blinded to the study conditions. * P < 0·05 v. breast-fed controls, P < 0·05 v. breast-fed + HMO + DPI, * P < 0·05 v. breast-fed + HMO + L-NIO. Values are means with their standard errors. Scale bar is 50 μm. Representative of at least three separate experiments, where each mouse represents an individual symbol on a graph. Fig. 6 The human milk oligosaccharide (HMO) 2′-fucosyllactose does not protect against necrotising enterocolitis (NEC) or enhance intestinal perfusion in endothelial nitric oxide synthase (eNOS)−/− mice. (a) Representative haematoxylin–eosin micrographs of the terminal ileum of eNOS−/− mice in the indicated treatment groups. (b) Representative confocal micrographs of terminal ileal whole mounts from eNOS−/− mice that received HMO and were either breast-fed (BF + HMO) or were formula-fed (FF + HMO). Whole mounts were stained for platelet endothelial cell adhesion molecule 1 (PECAM-1) to assess the microvasculature of the intestine and subjected to intra-cardiac injections with tomato lectin as a marker of intestinal perfusion (green). (c) Graph representing villus perfusion index as described in the Materials and methods section expressed as a percentage. (d) NEC severity score assessed by a pathologist blinded to the study conditions. (e) IL-1β mRNA expression from the terminal ileum of the mice in the indicated treatment groups. * P < 0·05 v. breast-fed control animals, ** P < 0·05 v. BF + HMO. Values are means with their standard errors. Scale bar is 50 μm. Representative of at least three separate experiments, where each mouse represents an individual symbol on a graph. qRT-PCR, quantitative real-time PCR. Fig. 7 Expression of endothelial nitric oxide synthase (eNOS) is enhanced by the human milk oligosaccharide (HMO) 2′-fucosyllactose in HUV-EC-C (HUVEC). (a) eNOS mRNA by quantitative real-time PCR (qRT-PCR) from HUVEC in the indicated treatment groups: vehicle control (Ctrl), lipopolysaccharide (LPS), HMO alone (HMO) and LPS with HMO (LPS + HMO) relative to the housekeeping gene RPLO. (b) Representative confocal micrographs of HUVEC with the indicated treatment groups stained for eNOS (red) and 4,6-diamidino-2-phenylindole, dihydrochloride (DAPI) (blue). * P < 0·05 v. vehicle control, ** P < 0·05 v. LPS. Values are means with their standard errors. Scale bar is 10 μm. Results representative of three separate experiments with over 50 high power fields per group imaged. Table 1 List of primers Gene Species Forward sequence Reverse sequence Amplicon size (bp) RPLO Mouse/human GGCGACCTGGAAGTCCAACT CCATCAGCACCACAGCCTTC 143 IL-6 Mouse GGCTAAGGACCAAGACCATCCAA TCTGACCACAGTGAGGAATGTCCA 138 TLR4 Mouse TTTATTCAGAGCCGTTGGTG CAGAGGATTGTCCTCCCATT 186 IL-1β Mouse AGTGTGGATCCCAAGCAATACCCA TGTCCTGACCACTGTTGTTTCCCA 175 iNOS Mouse CTGCTGGTGGTGACAAGCACATTT ATGTCATGAGCAAAGGCGCAGAAC 167 eNOS Mouse AGGACATATGTTTGTCTGCGGCGA AAATGTCCTCGTGGTAGCGTTGCT 155 eNOS Human ATGTTTGTCTGCGGCGATGTTACC TGTCTTCGTGGTAGCGTTGCTGAT 145 eNOSko Mouse TGGAAGGGAAGTGCAGCAAA GGCCAGTCTCAGAGCCATAC 151 RPL0, ribosomal protein large, P0; TLR4, toll-like receptor 4; iNOS, inducible nitric oxide synthase; eNOS,endothelial nitric oxide synthase. Supplementary material For supplementary material/s referred to in this article, please visit http://dx.doi.org/doi:10.1017/S0007114516002944 The authors declare that there are no conflicts of interest. References 1 Neu J Walker WA Necrotizing enterocolitis. N Engl J Med 2011 364 255 264 21247316 2 Fitzgibbons SC Ching Y Yu D Mortality of necrotizing enterocolitis expressed by birth weight categories. J Pediatr Surgery 2009 44 1072 1075 discussion 1075–1076 3 Lin PW Stoll BJ Necrotising enterocolitis. Lancet 2006 368 1271 1283 17027734 4 Obladen M Necrotizing enterocolitis – 150 years of fruitless search for the cause. Neonatology 2009 96 203 210 19407465 5 Sullivan S Schanler RJ Kim JH An exclusively human milk-based diet is associated with a lower rate of necrotizing enterocolitis than a diet of human milk and bovine milk-based products. J Pediatr 2010 156 562 567 e1 20036378 6 Lucas A Cole TJ Breast milk and neonatal necrotising enterocolitis. Lancet 1990 336 1519 1523 1979363 7 Newburg DS Walker WA Protection of the neonate by the innate immune system of developing gut and of human milk. Pediatr Res 2007 61 2 8 17211132 8 Meinzen-Derr J Poindexter B Wrage L Role of human milk in extremely low birth weight infants’ risk of necrotizing enterocolitis or death. J Perinatol 2009 29 57 62 18716628 9 Leaphart CL Cavallo J Gribar SC A critical role for TLR4 in the pathogenesis of necrotizing enterocolitis by modulating intestinal injury and repair. J Immunol 2007 179 4808 4820 17878380 10 Sodhi CP Neal MD Siggers R Intestinal epithelial toll-like receptor 4 regulates goblet cell development and is required for necrotizing enterocolitis in mice. Gastroenterology 2012 143 708 718 e1 e5 22796522 11 Good M Siggers RH Sodhi CP Amniotic fluid inhibits toll-like receptor 4 signaling in the fetal and neonatal intestinal epithelium. Proc Natl Acad Sci 2012 109 11330 11335 22733781 12 Good M Sodhi CP Egan CE Breast milk protects against the development of necrotizing enterocolitis through inhibition of toll-like receptor 4 in the intestinal epithelium via activation of the epidermal growth factor receptor. Mucosal Immunol 2015 8 1166 1179 25899687 13 Jilling T Simon D Lu J The roles of bacteria and TLR4 in rat and murine models of necrotizing enterocolitis. J Immunol 2006 177 3273 3282 16920968 14 Nino DF Sodhi CP Hackam DJ Necrotizing enter-ocolitis: mechanisms and management. Nat Rev Gastroenterol Hepatol 2016 In the press 15 Afrazi A Branca MF Sodhi CP Toll like receptor 4-mediated endoplasmic reticulum stress in intestinal crypts induces necrotizing enterocolitis. J Biol Chem 2014 289 9584 9599 24519940 16 Egan CE Sodhi CP Good M Toll-like receptor 4-mediated lymphocyte influx induces neonatal necrotizing enterocolitis. J Clin Invest 2015 126 495 508 17 Neal MD Sodhi CP Jia H Toll-like receptor 4 is expressed on intestinal stem cells and regulates their proliferation and apoptosis via the p53 up-regulated modulator of apoptosis. J Biol Chem 2012 287 37296 37308 22955282 18 Yazji I Sodhi CP Lee EK Endothelial TLR4 activation impairs intestinal microcirculatory perfusion in necrotizing enterocolitis via eNOS-NO-nitrite signaling. Proc Natl Acad Sci U S A 2013 110 9451 9456 23650378 19 Jantscher-Krenn E Zherebtsov M Nissan C The human milk oligosaccharide disialyllacto-N-tetraose prevents necrotising enterocolitis in neonatal rats. Gut 2012 61 1417 1425 22138535 20 Manzoni P Rinaldi M Cattani S Bovine lactoferrin supplementation for prevention of late-onset sepsis in very low-birth-weight neonates: a randomized trial. JAMA 2009 302 1421 1428 19809023 21 Dvorak B Halpern MD Holubec H Epidermal growth factor reduces the development of necrotizing enterocolitis in a neonatal rat model. Am J Physiol Gastrointest Liver Physiol 2002 282 G156 G164 11751169 22 Radulescu A Zorko NA Yu X Preclinical neonatal rat studies of heparin-binding EGF-like growth factor in protection of the intestines from necrotizing enterocolitis. Pediatr Res 2009 65 437 442 19127210 23 Newburg DS Neonatal protection by an innate immune system of human milk consisting of oligosaccharides and glycans. J Anim Sci 2009 87 26 34 19028867 24 Newburg DS Ruiz-Palacios GM Morrow AL Human milk glycans protect infants against enteric pathogens. Annu Rev Nutr 2005 25 37 58 16011458 25 Bode L Human milk oligosaccharides: prebiotics and beyond. Nutr Rev 2009 67 Suppl. 2 S183 S191 19906222 26 Bode L Human milk oligosaccharides: every baby needs a sugar mama. Glycobiology 2012 22 1147 1162 22513036 27 Bode L Kunz C Muhly-Reinholz M Inhibition of monocyte, lymphocyte, and neutrophil adhesion to endothelial cells by human milk oligosaccharides. Thromb Haemost 2004 92 1402 1410 15583750 28 Wang M Li M Wu S Fecal microbiota composition of breast-fed infants is correlated with human milk oligosaccharides consumed. J Pediatr Gastroenterol Nutr 2015 60 825 833 25651488 29 Zivkovic AM German JB Lebrilla CB Human milk glycobiome and its impact on the infant gastrointestinal microbiota. Proc Natl Acad Sci U S A 2011 108 Suppl. 1 4653 4658 20679197 30 Marcobal A Barboza M Froehlich JW Consumption of human milk oligosaccharides by gut-related microbes. J Agric Food Chem 2010 58 5334 5340 20394371 31 Ruiz-Palacios GM Cervantes LE Ramos P Campylobacter jejuni binds intestinal H(O) antigen (Fuc alpha 1, 2Gal beta 1, 4GlcNAc), and fucosyloligosaccharides of human milk inhibit its binding and infection. J Biol Chem 2003 278 14112 14120 12562767 32 Goehring KC Kennedy AD Prieto PA Direct evidence for the presence of human milk oligosaccharides in the circulation of breastfed infants. PLOS ONE 2014 9 e101692 24999728 33 Ruhaak LR Stroble C Underwood MA Detection of milk oligosaccharides in plasma of infants. Anal Bioanal Chem 2014 406 5775 5784 25059723 34 Afrazi A Sodhi CP Good M Intracellular heat shock protein-70 negatively regulates TLR4 signaling in the newborn intestinal epithelium. J Immunol 2012 188 4543 4557 22461698 35 Good M Sodhi CP Ozolek JA Lactobacillus rhamnosus HN001 decreases the severity of necrotizing enterocolitis in neonatal mice and preterm piglets: evidence in mice for a role of TLR9. Am J Physiol Gastrointest Liver Physiol 2014 306 G1021 G1032 24742987 36 Anand RJ Leaphart CL Mollen KP The role of the intestinal barrier in the pathogenesis of necrotizing enterocolitis. Shock 2007 27 124 133 17224785 37 Magoc T Salzberg SL FLASH: fast length adjustment of short reads to improve genome assemblies. Bioinformatics 2011 27 2957 2963 21903629 38 Bolger AM Lohse M Usadel B Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 2014 30 2114 2120 24695404 39 Kuczynski J Stombaugh J Walters WA Using QIIME to analyze 16S rRNA gene sequences from microbial communities. Curr Protoc Bioinformatics 2011 36 10.7.1 10.7.20 Chapter 10, Unit 10.17 40 Caporaso JG Bittinger K Bushman FD PyNAST: a flexible tool for aligning sequences to a template alignment. Bioinformatics 2010 26 266 267 19914921 41 Edgar RC Haas BJ Clemente JC UCHIME improves sensitivity and speed of chimera detection. Bioinformatics 2011 27 2194 2200 21700674 42 Langmead B Salzberg SL Fast gapped-read alignment with Bowtie 2. Nature Methods 2012 9 357 359 22388286 43 McDonald D Price MN Goodrich J An improved Greengenes taxonomy with explicit ranks for ecological and evolutionary analyses of bacteria and archaea. ISME J 2012 6 610 618 22134646 44 Wang Q Garrity GM Tiedje JM Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol 2007 73 5261 5267 17586664 45 Salter SJ Cox MJ Turek EM Reagent and laboratory contamination can critically impact sequence-based microbiome analyses. BMC Biol 2014 12 87 25387460 46 Benjamini Y Drai D Elmer G Controlling the false discovery rate in behavior genetics research. Behav Brain Res 2001 125 279 284 11682119 47 McMurdie PJ Holmes S phyloseq: An R package for reproducible interactive analysis and graphics of microbiome census data. PLOS ONE 2013 8 e61217 23630581 48 Jain SK Baggerman EW Mohankumar K Amniotic fluid-borne hepatocyte growth factor protects rat pups against experimental necrotizing enterocolitis. Am J Physiol Gastrointest Liver Physiol 2014 306 G361 G369 24407592 49 Shesely EG Maeda N Kim HS Elevated blood pressures in mice lacking endothelial nitric oxide synthase. Proc Natl Acad Sci U S A 1996 93 13176 13181 8917564 50 Yu X Radulescu A Zorko N Heparin-binding EGF-like growth factor increases intestinal microvascular blood flow in necrotizing enterocolitis. Gastroenterology 2009 137 221 230 19361505 51 Demmert M Schaper A Pagel J FUT 2 polymorphism and outcome in very-low-birth-weight infants. Pediatr Res 2015 77 586 590 25642664 52 Wacklin P Makivuokko H Alakulppi N Secretor genotype (FUT2 gene) is strongly associated with the composition of Bifidobacteria in the human intestine. PLoS ONE 2011 6 e20113 21625510 53 Garrido D Barile D Mills DA A molecular basis for bifidobacterial enrichment in the infant gastrointestinal tract. Adv Nutr 2012 3 415S 421 22585920 54 Marcobal A Sonnenburg JL Human milk oligosaccharide consumption by intestinal microbiota. Clin Microbiol Infect 2012 18 Suppl. 4 12 15 22647041 55 Kosloske AM Ulrich JA A bacteriologic basis for the clinical presentations of necrotizing enterocolitis. J Pediatr Surg 1980 15 558 564 7411368 56 Morrow AL Ruiz-Palacios GM Jiang X Human-milk glycans that inhibit pathogen binding protect breast-feeding infants against infectious diarrhea. J Nutr 2005 135 1304 1307 15867329 57 Gribar SC Sodhi CP Richardson WM Reciprocal expression and signaling of TLR4 and TLR9 in the pathogenesis and treatment of necrotizing enterocolitis. J Immunol 2009 182 636 646 19109197 58 La Rosa PS Warner BB Zhou Y Patterned progression of bacterial populations in the premature infant gut. Proc Natl Acad Sci U S A 2014 111 12522 12527 25114261 59 Underwood MA Gaerlan S De Leoz ML Human milk oligosaccharides in premature infants: absorption, excretion, and influence on the intestinal microbiota. Pediatr Res 2015 78 670 677 26322410 60 Santos-Fandila A Zafra-Gomez A Vazquez E Ultra high performance liquid chromatography-tandem mass spectrometry method for the determination of soluble milk glycans in rat serum. Talanta 2014 118 137 146 24274281 61 Prieto PA Mukerji P Kelder B Remodeling of mouse milk glycoconjugates by transgenic expression of a human glycosyltransferase. J Biol Chem 1995 270 29515 29519 7493992 62 Neal MD Jia H Eyer B Discovery and validation of a new class of small molecule Toll-like receptor 4 (TLR4) inhibitors. PLOS ONE 2013 8 e65779 23776545 63 He Y Liu S Kling DE The human milk oligosaccharide 2’-fucosyllactose modulates CD14 expression in human enterocytes, thereby attenuating LPS-induced inflammation. Gut 2016 65 33 46 25431457 64 Eidelman AI Schanler RJ Johnston M Breastfeeding and the use of human milk. Pediatrics 2012 129 e827 e841 22371471 65 Nutrition ECo Arslanoglu S Corpeleijn W Donor human milk for preterm infants: current evidence and research directions. J Pediatr Gastroenterol Nutr 2013 57 535 542 24084373
PMC005xxxxxx/PMC5124348.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 100887452 26737 Euro Surveill Euro Surveill. Euro surveillance : bulletin Europeen sur les maladies transmissibles = European communicable disease bulletin 1025-496X 1560-7917 27452806 5124348 10.2807/1560-7917.ES.2016.21.28.30283 NIHMS831071 Article The epidemiology and transmissibility of Zika virus in Girardot and San Andres island, Colombia, September 2015 to January 2016 Rojas DP 1 Dean NE 23 Yang Y 23 Kenah E 2 Quintero J 4 Tomasi S 4 Ramirez EL 5 Kelly Y 6 Castro C 7 Carrasquilla G 4 Halloran ME 89 Longini IM 2 1 Department of Epidemiology, University of Florida, Gainesville, FL, United States 2 Department of Biostatistics, University of Florida, Gainesville, FL, United States 4 Centro de Estudios e Investigacion en Salud, Fundacion Santa Fe de Bogota, Bogota, Colombia 5 Secretaria Municipal de Salud, Girardot, Colombia 6 IPS Universitaria, San Andres, Colombia 7 Secretaria Departamental de Salud, San Andres, Colombia 8 Department of Biostatistics, University of Washington, Seattle, WA, United States 9 Vaccine and Infectious Disease Division, Fred Hutchinson Cancer Research Center, Seattle, WA, United States Correspondence: Diana Patricia Rojas ([email protected]) 3 These authors contributed equally to this work 20 11 2016 14 7 2016 26 11 2016 21 28 10.2807/1560-7917.ES.2016.21.28.30283This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Transmission of Zika virus (ZIKV) was first detected in Colombia in September 2015. As of April 2016, Colombia had reported over 65,000 cases of Zika virus disease (ZVD). We analysed daily surveillance data of ZVD cases reported to the health authorities of San Andres and Girardot, Colombia, between September 2015 and January 2016. ZVD was laboratory-confirmed by reverse transcription-polymerase chain reaction (RT-PCR) in the serum of acute cases within five days of symptom onset. We use daily incidence data to estimate the basic reproductive number (R0) in each population. We identified 928 and 1,936 reported ZVD cases from San Andres and Girardot, respectively. The overall attack rate for reported ZVD was 12.13 cases per 1,000 residents of San Andres and 18.43 cases per 1,000 residents of Girardot. Attack rates were significantly higher in females in both municipalities (p < 0.001). Cases occurred in all age groups with highest rates in 20 to 49 year-olds. The estimated R0 for the Zika outbreak was 1.41 (95% confidence interval (CI): 1.15–1.74) in San Andres and 4.61 (95% CI: 4.11–5.16) in Girardot. Transmission of ZIKV is ongoing in the Americas. The estimated R0 from Colombia supports the observed rapid spread. Introduction First isolated in the Zika Forest of Uganda in 1947, Zika virus (ZIKV) is a flavivirus of the same genus as dengue virus and yellow fever virus. It is an arbovirus primarily transmitted by Aedes aegypti mosquitoes [1]. Although ZIKV has circulated in Africa and Asia since the 1950s, little is known about its transmission dynamics [2]. Recent outbreaks in Yap Island in Micronesia (2007), French Polynesia (2013), and other Pacific islands, including Cook Islands, Easter Island, and New Caledonia (2014), indicate that ZIKV has spread beyond its former geographical range [3–6]. In April 2015 ZIKV was isolated in the north-east of Brazil [7]. As of June 2016, around 500,000 Zika virus disease (ZVD) cases have been estimated in Brazil, and autochthonous circulation has been observed in 40 countries in the Americas. Further spread to countries within the geographical range of Ae. aegypti mosquitoes is considered likely [8]. Infection with ZIKV typically causes a self-limited dengue-like illness characterised by arthralgia, conjunctivitis, exanthema and low-grade fever [9]. While illness is believed to be mild or asymptomatic in ca 80% of the infections [10], an increase in rates of Guillain–Barre syndrome (GBS) has been observed during ZIKV outbreaks [8,11,12]. Furthermore, in October 2015, the Brazilian Ministry of Health reported a dramatic increase in cases of microcephaly in north-east Brazil where ZIKV had been circulating [13]. On the basis of the possible link between ZIKV, GBS and microcephaly, the World Health Organization (WHO) declared a public health emergency on 1 February 2016 [14,15]. In Colombia, the virus was first detected in mid-September 2015 in a municipality called Turbaco on the Caribbean coast. Turbaco is located 10.1 km from Cartagena (ca 20 min drive), a well-known commercial and tourism hub (Figure 1). In October 2015, ZIKV spread through the central region of the country, appearing in areas infested with Ae. aegypti and with endemic dengue transmission and ongoing circulation of chikungunya virus (CHIKV) since 2014. By April 2016, Colombia had reported over 65,000 cases of ZVD, making it the second country most affected by ZIKV after Brazil [16,17]. Up to April 2016, 280 cases of neurological complications including GBS as well as seven deaths possibly associated with ZVD had been reported in Colombia [18]. As of April 2016, there have been four confirmed cases of ZIKV congenital syndrome in the country [17]. In this paper we describe local ZIKV outbreaks between September 2015 and January 2016 in Girardot and San Andres island, two different geographical areas in Colombia for which detailed epidemiological data are available. We conduct an investigation to define the epidemiological features of these outbreaks and to estimate the corresponding transmission parameters. Methods Settings San Andres San Andres is the largest island in a Colombian archipelago in the Caribbean Sea located ca 750 km north of mainland Colombia and 230 km east of Nicaragua (Figure 1). The island has an area of 27 km2, a population of 54,513 inhabitants across 13,652 households, and a population density of 2,932 habitants per km2 in 2010 [19,20]. Tourism is the most important economic activity in San Andres, with two high touristic seasons: June to July and December to January. The average temperature is 27.3 °C, and 80% of the total annual rainfall of 1,700 mm occurs during the heavy rainy season between October and December. The weather is humid subtropical with occasional hurricanes. The population in San Andres has two main ethnic groups: Afro-Colombians (17.5%) and Raizal (an ethnic group of mixed Afro-Caribbean and British descent) (39.2%) [20]. The most productive breeding sites of Ae. aegypti in San Andres are unprotected water containers located in the households. San Andres has experienced low dengue transmission (annual incidence rates <1%) since 1983. Since 1995, the frequency of dengue outbreaks increased every two to five years with a mean annual incidence of 0.43 cases per 1,000 inhabitants between 1999 and 2010 [19]. In 2014, CHIKV began circulating in San Andres, and that year it reached an annual incidence of 3.65 cases per 1,000 inhabitants [21]. Girardot Girardot is a very central and well-connected municipality in continental Colombia. It is located 134 km (2 hours’ drive) from the capital city of Bogota, and it is a popular tourist destination for residents of Bogota (Figure 1). Girardot has 102,225 inhabitants across ca 23,000 households based on the most recent census from the National Statistics Department (NSD) [22], though the population can increase to 300,000 people during long weekends and high season holidays (June to July and December to January). Between 5 and 12 October 2015, a national beauty pageant in Girardot drew tourists from all regions in Colombia. Girardot is 289 m above sea level. The average temperature is 33.3 °C, and the relative humidity is 66%. The mean annual precipitation is 1,220 mm with a rainy season extending from May through October [23]. The most productive breeding sites of Ae. aegypti in Girardot are unprotected private water containers, such as water storage tanks used in the households during the dry and rainy seasons, while public spaces provide more breeding sites during the rainy season [24]. Girardot has experienced hyperendemic transmission of dengue since 1990 with simultaneous circulation of all four serotypes; the mean annual incidence was 5.72 per 1,000 inhabitants between 1999 and 2010 [19]. In late 2014, CHIKV started circulating in Girardot and that year it reached an annual incidence of 3.94 per 1,000 inhabitants, while in 2015 the annual incidence was 4.97 per 1,000 inhabitants [21,25]. Case definition and laboratory analysis We analysed surveillance data from nine local healthcare sites in San Andres and twenty-two local healthcare sites in Girardot, representing 100% of surveillance sites in both locations. Standardised case definitions used in both areas were defined by the Ministry of Health (MoH) and Colombian National Institute of Health (C-NIH) at the beginning of the ZIKV epidemic. According to these definitions, a suspected ZVD case is a person presenting with body temperature higher than 37.2 °C, maculopapular exanthema, and one or more of the following: arthralgia, headache, malaise, myalgia or non-purulent conjunctivitis and who lived or travelled to an area at risk for ZIKV transmission (usually below 2,000 m above sea level in Colombia) within 15 days of symptom onset. A laboratory-confirmed case is a suspected case with a ZIKV-positive reverse transcription-polymerase chain reaction (RT-PCR) result as determined by the C-NIH virology reference laboratory. ZIKV antibody testing was not done in Colombia due to high cross-reactivity with other endemic arboviruses. A clinically-confirmed case is defined by the Colombian authorities in the same way as a suspected case, except that the area of residence or travel within 15 days of symptom onset is an area with laboratory-confirmed ZIKV circulation [26]. Because the definition of a clinically-confirmed case in Colombia corresponded at the time of the study, to that of a probable case according to the WHO classification, we further refer to clinically-confirmed cases as probable cases in the context of this report [27]. At the start of the outbreaks in Girardot and San Andres, when local circulation of ZIKV had not yet been laboratory confirmed, only suspected cases were reported. Once the C-NIH confirmed the circulation of ZIKV in Girardot (on 27 January 2016, 3 months after the first local case report) and San Andres (on 22 October 2015, 45 days after the first local case report), the samples from suspected ZIKV cases were sent for laboratory confirmation if the respective cases fell into the risk groups defined by the C-NIH, including newborns and infants (age < 1 year), persons aged > 65 years, pregnant women, and individuals with comorbidities (e.g. diabetics, persons who were immunocompromised and/or with cardiovascular diseases) [23]. After ZIKV circulation was confirmed in the two areas, suspected cases whose acute samples tested positive were reclassified as laboratory-confirmed cases, while those with samples negative for ZIKV were reclassified as non-cases [26]. All reported suspected cases, who had not undergone laboratory testing were reclassified as probable cases [27]. Data collection The data in San Andres were collected initially using the C-NIH standard report form for dengue surveillance because from September up to October 2015 the outbreak in San Andres had an unknown aetiology. Once the C-NIH declared an alert on 14 October 2015 because ZIKV circulation had been observed in other areas of Colombia, reporting of ZVD became mandatory in the country, after which cases were reported by physicians at the healthcare sites using the standard report form for ZVD surveillance. The completeness of reporting is not known. We analysed a de-identified dataset based on place of residence with the following variables: age, sex, pregnancy status, date of symptom onset, date the case visited the healthcare facility, date the case was reported to the national surveillance system, and case type (suspected, laboratory confirmed, probable). Non-residents were excluded from the data [28]. Statistical analysis We calculated overall and age/sex-specific attack rates using population census data from NSD [22]. Surveillance data were analysed using R version 3.2.0 [29]. For descriptive results, categorical variables are presented as proportions and continuous variables by the median and interquartile range (IQR) or range. The relationship between attack rates and the variables age and sex was tested using log-linear models for case counts with age category (0–19 years-old, 20–49 years-old and >50 years-old), sex, and an interaction between age category and sex as independent variables, with population size as an offset. To estimate the basic reproductive number R0 in each population, we used maximum likelihood methods to fit a chain-binomial model to daily incidence data [30]. The model assumes a mean serial interval of 22 days (time between successive cases in a chain of transmission); the serial interval takes into account the infectious period in humans, the extrinsic latent period in mosquitoes, the mean infectious period in the mosquito, and the mean incubation period in humans [4,9,31,32]. Underreporting is assumed to be high (only 10% of cases reported) at the start of the outbreak and full reporting is assumed to be achieved in four weeks after the outbreak begins to grow. With this assumption, we aimed to take into account the respective delays in the two sites, between the ZVD outbreak start and the confirmation by the C-NIH of circulation of ZIKV. R0 is the median effective reproductive number during the growth phase of the epidemic, after accounting for early underreporting (see supplementary materials online for additional details on the model: https://github.com/dprojas/Zika). Results San Andres In San Andres, we identified 928 reported ZVD cases (Table 1). Of these cases, 52 (5.6%) were laboratory confirmed by RT-PCR on acute phase samples collected within five days of symptom onset, and 876 (94.4%) cases were probable. The dates of symptom onset among cases in San Andres ranged from 6 September 2015, to 30 January 2016 (Figure 2). Though the earliest case reported symptom onset on 6 September 2015, the local healthcare authorities did not receive laboratory confirmation of ZIKV until 22 October 2015. The distribution of this outbreak was bimodal. The first wave of the outbreak was before the C-NIH made an alert on 14 October 2015, about circulation of ZIKV in the country. The second wave started after the alert and the number of cases peaked in epidemiological week 45 (8 to 14 November), before the high tourist season started, and subsided in the last week of December. The second wave could be due to a reporting phenomenon. The median time between symptom onset and visiting a healthcare facility was 4 days (IQR: 1–16). Around 79% (733/928) of cases were reported to the national surveillance system on the same day that they visited the healthcare facility. The median age of reported ZVD cases in San Andres was 31 years-old (IQR: 15–47 years; range: 12 days–82 years). A total of 589 (63.5%) of the reported cases occurred in females. During the study period 238 dengue cases (incidence rate: 4.36 per 1,000 habitants) and 10 CHIKV cases (0.18 per 1,000 habitants) were reported in San Andres as expected in accordance with the trends and the historical data (data not shown). The overall attack rate for ZVD reported by local surveillance was 12.13 per 1,000 San Andres residents. The sex-specific attack rates were 15.34 per 1,000 females and 8.91 per 1,000 males; the difference was significant adjusting for age (p < 0.001). Cases occurred among all age groups, but the incidence of ZVD detected by local surveillance was highest among persons 20 to 49 years-old (Figure 3); there was significant heterogeneity across the age groups (p < 0.001). There was a significant interaction between age and sex (p < 0.001), consistent with the observation that attack rates were higher in females across all age groups 10 years-old and above, but lower for the younger age groups (Table 2). Thirty-three pregnant women with ZVD were reported in San Andres and are being followed according to national guidelines [33,34]. By June 2016, twenty-eight of them had given birth with two probable cases of congenital ZIKV syndrome reported. There were eight neurological syndromes reported in San Andres, including GBS and meningoencephalitis attributed to ZIKV and among them one death was reported. The incidence rate of neurological syndromes among ZVD cases in San Andres is 8.6 per 1,000 cases. Girardot In Girardot, we identified 1,936 reported ZVD cases (Table 1). Of these cases, 32 (1.7%) were laboratory confirmed by RT-PCR on acute phase samples collected within five days of symptom onset and 1,904 (98.3%) were probable. The date of symptom onset among cases in Girardot ranged from 19 October 2015 to 22 January 2016 (Figure 4). The first suspected case was reported on 23 October 2015, 19 days after the beauty pageant event started, with laboratory confirmation obtained on 27 January 2016. The number of cases peaked in epidemiological week 48 (29 November to 5 December) before the end-of-the-year tourist season, and subsided in early January. The median time between symptom onset and visiting a healthcare facility was 1 day (IQR: 1–2 days). Around 89% (755/1,936) of cases were reported to the national surveillance system on the same day they visited the healthcare facility. The median age of confirmed ZVD cases was 34 years-old (IQR: 24–46 years; range: 15 days–92 years). A total of 1,138 (58.8%) cases were female. During the study period 75 dengue cases (incidence rate: 0.73 per 1,000 habitants) and 200 CHIKV cases (1.95 per 1,000 habitants) were reported in Girardot as expected in accordance with the trends and the historical data (data not shown). The overall attack rate for confirmed ZVD detected by local surveillance was 18.43 per 1,000 Girardot residents. The sex-specific attack rates were 20.53 per 1,000 females and 16.07 per 1,000 males; the difference was significant adjusting for age (p < 0.001). Cases occurred among all age groups, but the incidence of ZVD detected by local surveillance was highest among persons 20 to 49 years-old (Figure 5); there was significant heterogeneity across the age groups (p < 0.001). Attack rates were higher in females in all age groups except in those 10 to 14 and 65 to 69 years-old; there was no significant interaction between age and sex (p = 0.20) (Table 2). Sixteen pregnant women with ZVD were reported in Girardot and are being followed according to national guidelines [33,34]. By June 2016, twelve of them had given birth with no complications or microcephaly reported. Nine cases with GBS have been reported after an initial suspected ZIKV infection; laboratory-confirmation of ZIKV is pending. There were no deaths attributed to ZIKV. The incidence rate of neurological syndromes among ZVD cases in Girardot is 4.6 per 1,000 cases. Basic reproductive number calculations Daily incidence data were used to estimate R0. The estimated R0 for the Zika outbreak in San Andres was 1.41 (95% confidence interval (CI): 1.15–1.74), and the R0 in Girardot was 4.61 (95% CI: 4.11–5.16) (Table 2 and Figure 6). Odds ratios for sex and age effects were obtained from the likelihood model, indicating increased odds of transmission among females and adults aged 20 to 49 years-old in both San Andres and Girardot (Table 2). The estimation procedure was also applied to daily incidence data from a published outbreak in Salvador, Brazil, that occurred between 15 February 2015, and 25 June 2015; 14,835 cases were reported with an overall attack rate of 5.5 cases per 1,000 Salvador residents [7]. The estimated R0 of the Zika outbreak in Salvador, Brazil was 1.42 (95% CI: 1.35–1.49). Sensitivity analyses are reported in the supplementary online materials (https://github.com/dprojas/Zika), including varying the incubation period in humans, the infectious period in humans, the infectious period in mosquitoes, the duration of underreporting, and the level of underreporting at the start of the outbreak. Discussion We report surveillance data on ZIKV outbreaks in two areas in Colombia between September 2015 and January 2016. The first area, San Andres, is a small, densely populated island that is relatively isolated from continental Colombia. The second area, Girardot, is a typical moderately sized Colombian municipality. Both regions have endemic transmission of dengue and experienced recent outbreaks of CHIKV. We describe key epidemiological features of the ZVD outbreaks and estimate R0 from daily incidence data. The overall attack rates for ZVD as detected by local surveillance were 12.13 cases per 1,000 residents of San Andres and 18.43 cases per 1,000 residents of Girardot. These attack rates are similar to those reported from Yap Island (14.3 per 1,000) [3] but higher than those reported in Salvador, Brazil (5.5 per 1,000) [7]. In both areas, significantly higher attack rates are observed among women, especially those of child-bearing age. The Colombian government issued an epidemiological alert in December 2015 to actively search for pregnant women with ZVD-like symptoms in areas with active transmission [33,34]. This effort may partially explain the findings, though differences in sex-specific attack rates persist when only cases occurring before December are considered. These results could be explained by male-to-female sexual transmission of ZVD, which is consistent with higher attack rates in females beyond child-bearing age. Given recent evidence from Brazil, in areas with ZIKV transmission, interventions aimed at preventing sexual-transmission of ZIKV to women are necessary because this mode of transmission could have a substantial influence on the overall dynamics of ZIKV epidemics [35–37]. Cases occurred in all age groups, but the most affected age group was 20 to 49 year of age, similar to previously published outbreaks in Yap Island, Micronesia, and in Salvador, Brazil [3,7]. As the population was fully susceptible to ZIKV transmission before the outbreaks, it is expected that all age groups would be affected. Forty-nine pregnant women with ZVD were reported from San Andres and Girardot. These women are being followed according to national guidelines [33,34] with two probable cases of congenital ZIKV syndrome reported from San Andres to the national authorities for analysis. Seventeen cases of neurological syndrome, including GBS and ZIKV-associated meningoencephalitis, were identified, similar to reports from French Polynesia and Brazil [12,38]. Laboratory-confirmation of these cases is challenging because neurological symptoms generally appear two weeks after acute symptoms [39] at which time ZIKV diagnosis by RT-PCR is not possible and serological tests are unreliable because of cross-reactivity with dengue [40,41]. As ZIKV can be detected in urine longer than in serum [42], using urine samples to confirm ZIKV in GBS cases may be an alternative [43]. These challenges underscore the need for reliable diagnostic tests that can detect ZIKV after the viraemic period. In each area of this study, daily incidence data were used to estimate R0. Our estimated R0 for the ZVD outbreak in San Andres was 1.41 (95% CI: 1.15–1.74), and the R0 for Girardot was 4.61 (95% CI: 4.11–5.16). Applying the same methods with previously published data, we estimated that the R0 for ZIKV in Salvador, Brazil, was 1.42 (95% CI: 1.35–1.49) [7]. We consider the estimate from San Andres to be the most reliable because it is a small, densely populated island and the outbreak occurred before the national epidemiological alert, while Girardot has a higher risk of importation because the population fluctuates during weekends and holidays. The relative magnitudes of R0 are consistent with the higher dengue transmission historically observed in Girardot vs San Andres [19]. Estimates of R0 in ZIKV are not widely available, though reports suggest an R0 of 4.3 to 5.8 in Yap Island and R0 of 1.8 to 2.0 in French Polynesia [44]. A recent manuscript considering the French Polynesian outbreak reported a range from 1.9 to 3.1 [45]. Relatively few cases were laboratory confirmed. One limitation of this study is that the majority of cases were probable, and the symptoms could be caused by other aetiologies such as dengue or CHIKV. Nonetheless, in the field we have observed that the diseases have different clinical manifestations. Dengue appears to coincide with high fever (> 38.5 °C), headaches, myalgia, and generalised pain. CHIKV is associated with joint pain and arthritis, and ZVD is associated with a very mild, low-grade fever (38 °C) or no fever, rash, and no generalised pain. This report only includes symptomatic cases who attended a healthcare facility and were captured by the surveillance systems. ZIKV usually causes a relatively mild illness lasting several days, and around 80% of infections are currently believed to be asymptomatic, so we are likely missing many mild or asymptomatic cases [10]. We also do not have a reliable estimate of underreporting at the study sites. Early underreporting seemed to be especially apparent in the Girardot outbreak compared with San Andres given that the circulation of ZIKV was not confirmed until January, 2016, and the sharp increase in cases in Girardot observed may be due to increased public awareness of the disease. This phenomenon can result in an overestimate of R0. Well-designed studies can provide valuable insight. Phylogenetic analyses of circulating ZIKV strains will be critical for understanding whether mutations in the viral genome are associated with an increased severity of disease, as manifested by microcephaly and GBS in this outbreak. Household studies can allow for more accurate estimation of transmission dynamics and enhance understanding of asymptomatic infection. Studies are required to understand the interactions between ZIKV, dengue, CHIKV, and other co-circulating arboviruses and their impact on disease. It is also necessary to increase surveillance of neurological syndromes associated with ZVD, such as GBS and encephalitis. The evidence for a causal relationship between ZIKV and microcephaly is strengthening [46–48]. Recent evidence from the French Polynesia outbreak suggests an estimated number of microcephaly cases possibly associated with ZIKV infection is around one per 100 women infected in the first trimester [49]. Currently the Colombian Government is following a cohort of pregnant women that reported ZVD-like symptoms anytime during their pregnancy. Those who are detected during the acute phase are being diagnosed with ZIKV RT-PCR. All women will be followed until the end of pregnancy, and the fetus will be evaluated during pregnancy, with a subsequent post-natal follow-up of twelve months [17]. The prospective collection of data through this and other similar national cohorts will be essential for assessing causality, determining risk factors, and estimating rates of birth defects. The results of this and other reports conclude that transmission of ZIKV may be widespread. Vector control has had limited success in controlling other arboviruses, such as dengue. A safe and efficacious vaccine, especially for women of child-bearing age, may be needed to reduce the disease burden. Supplementary Material Supplemental Materials This work was supported by National Institutes of Health (NIH) U54GM111274, NIH R37 AI032042 and the Colombian Department of Science and Technology (Fulbright-Colciencias scholarship to D.P.R). We want to thank the local health authorities from San Andres Providencia and Santa Catalina and Girardot; Especially Dr. Hayder Avendano Villa the Director of Health from San Andres, Providencia and Santa Catalina, Dr. Manuel Diaz Director of Health in Girardot and Dr. Ernesto Diaz Suarez from IPS Universitaria San Andres y Providencia. Figure 1 Location of the two Zika virus outbreak settings investigated, Colombia, September 2015–January 2016 Colombia figures in yellow on the map, with a dark square for the capital city Bogota. The two settings of Zika virus disease outbreaks investigated in this study are indicated by a star. On the map, the city of Cartagena is also shown, because in Colombia, Zika virus was first detected ca 10 km from this city, before spreading to other locations in the country. Figure 2 Daily Zika virus disease incidence in San Andres, Colombia, September 2015–January 2016 (n=928 cases) Cases include all reported cases, which were San Andres residents. Figure 3 Age- and sex-specific Zika virus disease attack rates for San Andres, Colombia, September 2015–January 2016 (n=928 cases) Figure 4 Daily ZVD incidence for Girardot, Colombia, October 2015–January 2016 (n=1,936 cases) Figure 5 Age- and sex-specific Zika virus disease attack rates for Girardot, Colombia October 2015–January 2016 (n=1,936 cases) Figure 6 Estimates of effective R (red) and model-fitted daily case numbers (green) for outbreaks of Zika virus disease in Colombia, September 2015–January 2016 (n=2,864 cases CI: confidence interval; R0: basic reproductive number. (A) Estimates of effective R (red) and model-fitted daily case numbers (green) for the outbreak of ZVD in San Andres, Colombia. The proportion of cases reported is assumed to increase linearly from 10% on and before 30 September 2015, to 100% in 4 weeks. Dashed curves (both red and green) are conservative 95% CIs. Histogram in grey shows the epidemic curve. The horizontal yellow line indicates the reference value of 1. The two vertical yellow lines indicate the time interval used for the estimation of R0. (B) As (A) for Girardot, Colombia. The proportion of cases reported increases on 19 October 2015. Table 1 Characteristics of reported cases of Zika virus disease in two areas of Colombia, September 2015–January 2016 Areas San Andres Girardot Total number of cases 928 1,936 Laboratory confirmed cases n (%) 52 (5.6%) 32 (1.7%) Probable cases n (%) 876 (94.4%) 1,904 (98.3%) Female n (%) 589 (63.5%) 1,138 (58.8%) Median age in years (IQR) 31 (15–47) 34 (24–46) Median time in days to visit healthcare facility from symptom onset (IQR) 4 (1–16) 1 (1–2) IQR: interquartile range. Table 2 Estimates of basic reproductive number (R0), sex-specific odds ratios (OR) and age-specific OR for transmission of Zika virus disease in San Andres and Girardot, Colombia, September 2015–January 2016 Parameter Estimate (95%CI) San Andres Girardot Estimate (95%CI) R0 1.41 (1.15–1.74) 4.61 (4.11–5.16) OR sex Male Reference Reference Female 1.71 (1.50–1.95) 1.28 (1.17–1.40) OR age in years 20–49 Reference Reference 0–19 0.86 (0.74–0.99) 0.37 (0.33–0.42) > 50 0.74 (0.63–0.88) 0.46 (0.41–0.52) CI: confidence interval. Conflict of interest None declared. Authors’ contributions DPR, NED, YY: Study design, data analysis, data interpretation, figures, writing and approval of this manuscript. EK: Study design, data analysis, data interpretation, approval of the manuscript. JQ, ST, ELR, YK, CC, GC: Data collection, data analysis, data interpretation, approval of this manuscript. MEH and IML: Study design, data analysis, data interpretation, writing and approving this manuscript. References 1 Staples JE Dziuban EJ Fischer M Cragan JD Rasmussen SA Cannon MJ Interim guidelines for the evaluation and testing of infants with possible congenital Zika virus infection — United States, 2016 MMWR Morb Mortal Wkly Rep 2016 65 3 63 67 PMID: 26820387 26820387 2 Haddow AD Schuh AJ Yasuda CY Kasper MR Heang V Huy R Genetic characterization of Zika virus strains: geographic expansion of the Asian lineage PLoS Negl Trop Dis 2012 6 2 e1477 PMID: 22389730 22389730 3 Duffy MR Chen T-H Hancock WT Powers AM Kool JL Lanciotti RS Zika virus outbreak on Yap Island, Federated States of Micronesia N Engl J Med 2009 360 24 2536 2543 PMID: 19516034 19516034 4 Hayes EB Zika virus outside Africa Emerg Infect Dis 2009 15 9 1347 1350 PMID: 19788800 19788800 5 Ioos S Mallet H-P Leparc Goffart I Gauthier V Cardoso T Herida M Current Zika virus epidemiology and recent epidemics Med Mal Infect 2014 44 7 302 307 PMID: 25001879 25001879 6 Tognarelli J Ulloa S Villagra E Lagos J Aguayo C Fasce R A report on the outbreak of Zika virus on Easter Island, South Pacific, 2014 Arch Virol 2016 161 3 665 668 PMID: 26611910 26611910 7 Cardoso CW Paploski IA Kikuti M Rodrigues MS Silva MM Campos GS Outbreak of exanthematous illness associated with Zika, Chikungunya, and dengue viruses, Salvador, Brazil Emerg Infect Dis 2015 21 12 2274 2276 PMID: 26584464 26584464 8 Pan-American Health Organization (PAHO). World Health Organization (WHO) Geographic distribution of confirmed cases of Zika virus (locally acquired) in countries and territories of the Americas, 2015–2016 PAHO/WHO Accessed 9 Jul 2016 Available from: http://www.paho.org/hq/index.php?option=com_content&view=article&id=11599:regional-zika-epidemiological-update-americas&Itemid=41691&lang=es 9 European Centre for Disease Prevention and Control (ECDC) Rapid risk assessment: Zika virus infection outbreak, French Polynesia, 14 February 2014 2014 Stockholm ECDC 10 Petersen EE Staples JE Meaney-Delman D Fischer M Ellington SR Callaghan WM Interim guidelines for pregnant women during a Zika virus outbreak — United States, 2016 MMWR Morb Mortal Wkly Rep 2016 65 2 30 33 PMID: 26796813 26796813 11 Musso D Zika virus transmission from French polynesia to Brazil Emerg Infect Dis 2015 21 10 1887 PMID: 26403318 12 Cao-Lormeau VM Blake A Mons S Lastère S Roche C Vanhomwegen J Guillain-Barré Syndrome outbreak associated with Zika virus infection in French Polynesia: a case-control study Lancet 2016 387 10027 1531 1539 PMID: 26948433 26948433 13 Schuler-Faccini L Ribeiro EM Feitosa IM Horovitz DD Cavalcanti DP Pessoa A Brazilian Medical Genetics Society–Zika Embryopathy Task Force Possible association between Zika virus infection and microcephaly - Brazil, 2015 MMWR Morb Mortal Wkly Rep 2016 65 3 59 62 PMID: 26820244 26820244 14 Lazear HM Stringer EM de Silva AM The emerging Zika virus epidemic in the Americas: Research priorities JAMA 2016 315 18 1945 1946 PMID: 26963564 26963564 15 World Health Organization (WHO) WHO director-general summarizes the outcome of the emergency committee regarding clusters of microcephaly and Guillain-Barre syndrome, 2016 2016 Geneva WHO Accessed 12 Mar 2016 Available from: http://www.who.int/mediacentre/news/statements/2016/emergency-committee-zika-microcephaly/en/ 16 Instituto Nacional de Salud (INS) Boletin epidemiologico semanal - semana epidemiologica 08 de 2016, 2016. INS:2016 Spanish Accessed 11 Mar 2016 Available from: http://www.ins.gov.co/boletin-epidemiologico/Boletn%20Epidemiolgico/2016%20Bolet%C3%ADn%20epidemiol%C3%B3gico%20semana%208.pdf 17 Pacheco O Beltrán M Nelson CA Valencia D Tolosa N Farr SL Zika virus disease in Colombia — preliminary report N Engl J Med 2016 PMID: 27305043 18 Instituto Nacional de Salud (INS) Vigilancia intensificada de sindromes neurologicos, (Guillain-Barre, polineuropatias ascendentes, entre otras afecciones neurologicas similares), en la fase epidemica de la infeccion por virus Zika en Colombia, 2015 – 2016, INS: 2016 Spanish Accessed 12 Mar 2016 Available from: http://www.ins.gov.co/Noticias/ZIKA/Bolet%C3%ADn%20semana%2006,%20s%C3%ADndromes%20neurol%C3%B3gicos%20-%20Zika.pdf 19 Padilla JC Rojas DP Sáenz-Gómez R Dengue en Colombia: Epidemiologia de la reemergencia a la hiperendemia 2012 First Bogota Guias de Impresion Spanish 20 Ministerio de Salud (MINSALUD) Analisis de situacion de salud (ASIS) Archipielago de San Andres, 2011 2011 Spanish MINSAD 21 Salas D Informe final del evento Chikungunya, Colombia 2014 2014 Spanish Accessed 12 Mar 2016 Available from: http://www.ins.gov.co/lineas-de-accion/Subdireccion-Vigilancia/Informe%20de%20Evento%20Epidemiolgico/Chikungu%C3%B1a%202014.pdf 22 Departamento Administrativo Nacional de Estadistica (DANE) Estimacion y proyeccion de poblacion Nacional, Departamental y Municipal total por area 1985–2020 2016 Spanish DANE 23 García-Betancourt T Higuera-Mendieta DR González-Uribe C Cortés S Quintero J Understanding water storage practices of urban residents of an endemic dengue area in Colombia: Perceptions, rationale and socio-demographic characteristics PLoS One 2015 10 6 e0129054 PMID: 26061628 26061628 24 Alcalá L Quintero J González-Uribe C Brochero H Productividad de Aedes aegypti (L.) (Diptera: Culicidae) en viviendas y espacios públicos en una ciudad end éndemica para dengue en Colombia Biomedica 2015 35 2 258 268 PMID: 26535548 26535548 25 Instituto Nacional de Salud (INS) Chikungunya week 52, 2015. INS: 2015 Accessed 12 Jul 2016 Available from: http://www.ins.gov.co/Noticias/Chikungunya/Resumen%20Chikungu%C3%B1a%20SEMANA%2052%202015.pdf 26 Ministerio de Salud (MINSAD), Instituto Nacional de Salud (INS) Circular conjunta externa no. 061 de 2015 2015 Spanish Accessed 12 Mar 2016 Available from: http://www.ins.gov.co/Noticias/SiteAssets/Paginas/Zika/Circular%20Conjunta%20061%20de%202015.pdf 27 World Health Organization (WHO) Case definitions- Zika Geneva WHO Accessed 12 Mar 2016 Available from: http://www.paho.org/hq/index.php?option=com_content&view=article&id=11117&Itemid=41532&lang=en 28 Instituto Nacional de Salud (INS) Report form Zika illness-Colombia, 2015. INS: 2015 Accessed 12 Mar 2016 Available from: http://www.ins.gov.co/lineas-de-accion/Subdireccion-Vigilancia/sivigila/Fichas%20de%20Notificacin%20SIVIGILA/Zika%20DEF%20895.pdf 29 R Core Team R: A Language and Environment for Statistical Computing 2016 Vienna, Austria R Foundation for Statistical Computing 30 Yang Y Sugimoto JD Halloran ME Basta NE Chao DL Matrajt L The transmissibility and control of pandemic influenza A (H1N1) virus Science 2009 326 5953 729 733 PMID: 19745114 19745114 31 Day JF Edman JD Scott TW Reproductive fitness and survivorship of Aedes aegypti (Diptera: Culicidae) maintained on blood, with field observations from Thailand J Med Entomol 1994 31 4 611 617 PMID: 7932609 7932609 32 Styer LM Carey JR Wang JL Scott TW Mosquitoes do senesce: departure from the paradigm of constant mortality Am J Trop Med Hyg 2007 76 1 111 117 PMID: 17255238 17255238 33 Ministerio de Salud (INS) Instituto Nacional de Salud. Circular externa 004 de 2016. INS: 2016 Spanish Accessed 12 Mar 2016 Available from: http://www.ins.gov.co/Noticias/ZIKA/Circular%20externa%200004_INS.PDF 34 Instituto Nacional de Salud (INS) Circular 007 de 2016, 2016. INS: 2016 Spanish Accessed 12 Mar 2016 Available from: http://www.ins.gov.co/Noticias/ZIKA/circular%200007-%202016_INS-DC%20Microcefalia.pdf 35 Kleber de Oliveira W Cortez-Escalante J De Oliveira WT do Carmo GM Henriques CM Coelho GE Increase in reported prevalence of microcephaly in infants born to women living in areas with confirmed Zika virus transmission during the first trimester of pregnancy—Brazil, 2015 MMWR Morb Mortal Wkly Rep 2016 65 9 242 247 PMID: 26963593 26963593 36 Bastos L Villela DA Carvalho LM Cruz OG Gomes MF Durovni B Zika in Rio de Janeiro: Assessment of basic reproductive number and its comparison with dengue bioRxiv 2016 055475 37 Chouin-Carneiro T Vega-Rua A Vazeille M Yebakima A Girod R Goindin D Differential susceptibilities of Aedes aegypti and Aedes albopictus from the Americas to Zika virus PLoS Negl Trop Dis 2016 10 3 e0004543 PMID: 26938868 26938868 38 World Health Organization Guillain-Barre syndrome - Brazil, 2016 2016 Geneva WHO Accessed 12 Mar 2016 Available from: http://www.who.int/csr/don/8-february-2016-gbs-brazil/en/ 39 Yuki N Hartung H-P Guillain-Barré syndrome N Engl J Med 2012 366 24 2294 2304 PMID: 22694000 22694000 40 Campos GS Bandeira AC Sardi SI Zika virus outbreak, Bahia, Brazil Emerg Infect Dis 2015 21 10 1885 1886 PMID: 26401719 26401719 41 Villamil-Gomez WE Gonzalez-Camargo O Rodriguez-Ayubi J Zapata-Serpa D Rodriguez-Morales AJ Dengue, Chikungunya and Zika co-infection in a patient from Colombia J Infect Public Health 2016 pii: S1876-0341(15)00221-X. [Epub ahead of print] PMID: 26754201 42 Gourinat A-C O’Connor O Calvez E Goarant C Dupont-Rouzeyrol M Detection of Zika virus in urine Emerg Infect Dis 2015 21 1 84 86 PMID: 25530324 25530324 43 Rozé B Najioullah F Fergé JL Apetse K Brouste Y Cesaire R GBS Zika Working Group Zika virus detection in urine from patients with Guillain-Barré syndrome on Martinique, January 2016 Euro Surveill 2016 21 9 pii=30154. PMID: 26967758 44 Nishiura H Kinoshita R Mizumoto K Yasuda Y Nah K Transmission potential of Zika virus infection in the South Pacific Int J Infect Dis 2016 45 95 97 PMID: 26923081 26923081 45 Kucharski AJ Funk S Eggo RMM Mallet H-P Edmunds WJ Nilles EJ Transmission dynamics of Zika virus in island populations: a modelling analysis of the 2013–14 French Polynesia outbreak PLoS Negl Trop Dis 2016 10 5 e0004726 PMID: 27186984 27186984 46 Mlakar J Korva M Tul N Popović M Poljšak-Prijatelj M Mraz J Zika virus associated with microcephaly N Engl J Med 2016 374 10 951 958 PMID: 26862926 26862926 47 World Health Organization (WHO) Zika situation report. 2016 2016 Geneva WHO Accessed 12 Mar 2016 Available from: http://www.who.int/emergencies/zika-virus/situation-report/10-march-2016/en/ 48 Brasil P Pereira JP Raja Gabaglia C Damasceno L Wakimoto M Ribeiro Nogueira RM Zika virus infection in pregnant women in Rio de Janeiro — preliminary report N Engl J Med 2016 3 4 [Epub ahead of print]. DOI: http://dx.doi.org/10.1056/NEJMoa1602412 PMID: 26943629 49 Cauchemez S Besnard M Bompard P Dub T Guillemette-Artur P Eyrolle-Guignot D Association between Zika virus and microcephaly in French Polynesia, 2013–15: a retrospective study Lancet 2016 387 10033 2125 2132 PMID: 26993883 26993883
PMC005xxxxxx/PMC5124365.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 0410462 6011 Nature Nature Nature 0028-0836 1476-4687 27556938 5124365 10.1038/nature19318 EMS69526 Article Serotonin engages an anxiety and fear-promoting circuit in the extended amygdala Marcinkiewcz Catherine A. 1 Mazzone Christopher M. 12 D’Agostino Giuseppe 3 Halladay Lindsay R. 4 Hardaway J. Andrew 1 DiBerto Jeffrey F. 1 Navarro Montserrat 5 Burnham Nathan 5 Cristiano Claudia 3 Dorrier Cayce E. 1 Tipton Gregory J. 1 Ramakrishnan Charu 6 Kozicz Tamas 78 Deisseroth Karl 6 Thiele Todd E. 15 McElligott Zoe A. 19 Holmes Andrew 4 Heisler Lora K. 3 Kash Thomas L. 12510 1 Bowles Center for Alcohol Studies, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, USA 2 Curriculum in Neurobiology, School of Medicine, University of North Carolina at Chapel Hill School of Medicine, Chapel Hill, NC 27599, USA 3 Rowett Institute of Nutrition and Health, University of Aberdeen, Aberdeen AB25 2ZD, UK 4 National Institute on Alcohol Abuse and Alcoholism, National Institutes of Health, Rockville, MD 20852-9411, USA 5 Department of Psychology & Neuroscience, College of Arts and Sciences, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, USA 6 Department of Bioengineering, Stanford University, Stanford CA 94305, USA 7 Hayward Genetics Center, Tulane University, New Orleans, LA 70112, USA 8 Department of Anatomy, Radboud University Nijmegen Medical Center, 6500HB Nijmegen, The Netherlands 9 Department of Psychiatry, School of Medicine, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, USA 10 Department of Pharmacology, School of Medicine, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, USA Please address all correspondence to: Thomas L. Kash, Bowles Center for Alcohol Studies, University of North Carolina at Chapel Hill, CB# 7178 Thurston Bowles Building, Chapel Hill, NC 27599-7178, [email protected] 11 8 2016 24 8 2016 01 9 2016 24 2 2017 537 7618 97101 This file is available to download for the purposes of text mining, consistent with the principles of UK copyright law. Summary paragraph Serotonin (5-hydroxytryptamine; 5-HT) is a neurotransmitter that has an essential role in the regulation of emotion. The precise circuits through which aversive states are orchestrated by 5-HT, however, have not yet been defined. Here we show that 5-HT from the dorsal raphe nucleus (5-HTDRN) enhances fear and anxiety and activates a subpopulation of corticotropin-releasing factor (CRF) neurons in the bed nucleus of the stria terminalis (CRFBNST). Specifically, 5-HTDRN projections to the BNST, via actions at 5-HT2C receptors (5-HT2CRs), engage a CRFBNST inhibitory microcircuit that silences anxiolytic BNST outputs to the ventral tegmental area (VTA) and lateral hypothalamus (LH). Further, we demonstrate that this CRFBNST inhibitory circuit underlies aversive behavior following acute exposure to selective serotonin reuptake inhibitors (SSRIs). This early aversive effect is mediated via the corticotrophin releasing factor type 1 receptor (CRF1R) given that CRF1R antagonism is sufficient to prevent acute SSRI-induced enhancements in aversive learning. These results reveal an essential 5-HTDRN→CRFBNST circuit governing fear and anxiety and provide a potential mechanistic explanation for the clinical observation of early adverse events to SSRI treatment in some patients with anxiety disorders1,2. Serotonin BNST anxiety fear 5-HT2C receptor CRF VTA In view of multiple converging lines of evidence pinpointing 5-HT as a critical neuromodulator of pathological fear learning3,4, we first interrogated the endogenous recruitment of the 5-HTDRN→BNST circuit by an aversive footshock stimulus. Using fluorogold to retrogradely label BNST-projecting 5-HT neurons in the DRN, we found that c-fos, an immediate early gene indicative of in vivo neuronal activation, was significantly elevated in 5-HTDRN→BNST neurons after footshock (Figure 1a-f). Using in vivo electrophysiology, we then probed the neuronal dynamics of the BNST during fear conditioning and recall and found evidence for engagement during both conditioning and recall (Extended Data Figure 1). To decipher the role of this 5-HTDRN→BNST circuit in aversive behavior, Channelrhodopsin2 (ChR2)-eYFP was selectively expressed in 5-HTDRN neurons through the delivery of a Cre-inducible viral vector in mice expressing Cre recombinase under the control of a serotonin transporter promoter (SertCre) for both in vivo and ex vivo analysis. We observed eYFP+ (5-HT) cell bodies in the DRN and eYFP+ fibers in both the dorsal and ventral aspects of the BNST (SertCre::ChR2DRN→BNST), confirming a direct projection of 5-HT neurons originating in the DRN to the BNST (Figure 1g-h)5. Optical stimulation of these fibers in BNST slices evoked 5-HT release, as measured by fast-scan cyclic voltammetry (FSCV) (Figure 1i-j). Furthermore, bath application of the SSRI fluoxetine reliably decreased the rate of 5-HT reuptake, confirming that photostimulation of SERT+ terminals in the BNST originating from the DRN induces 5-HT release (Figure 1k-l). We next examined whether this 5-HTDRN→BNST circuit is functionally relevant for fear and anxiety-like behavior. To investigate this, SertCre::ChR2DRN→BNST mice were implanted with bilateral optical fibers and photostimulated in the BNST (473 nm, 20 Hz) using a standard tone-shock fear conditioning paradigm. Optogenetic stimulation of this pathway was paired with a tone that co-terminated with a scrambled footshock. Cued fear was assessed 24 hours after, and contextual fear 48 hours after, the initial fear acquisition session (Figure 1m-n). While no changes were observed during fear acquisition, both cued and contextual fear recall were significantly heightened in photostimulated SertCre::ChR2DRN→BNST mice (Figure 1o-q). We next assessed anxiety-like behavior using well-characterized assays, the elevated plus maze (EPM) and novelty-suppressed feeding (NSF) tests. Upon stimulation with light, SertCre::ChR2DRN→BNST mice exhibited enhanced anxiety-like behavior in both the EPM and NSF (Figure 1r-s and Extended Data Figure 2a-b). Importantly, photostimulation did not induce hypolocomotion in the EPM or open field tests nor did it alter home-cage feeding, thus confirming that hypophagia in the NSF assay was due to anxiety and not a reduction in appetitive drive (Extended Data Figure 2c-e). One potential explanation of these results is that terminal stimulation in the BNST produces antidromic spikes in DRN cell bodies that release 5-HT in other brain regions, which could be also be driving these behaviors. In light of this, we probed the mechanism more deeply using converging approaches. To determine a receptor target through which 5-HT is signaling in the BNST, we then examined the impact of optogenetically evoked 5-HTDRN release on postsynaptic neuronal excitability and found a 3.05 ± 0.59 mV depolarization that was blocked by a 5-HT2CR antagonist (Figure 1t-u). In contrast to previous reports demonstrating co-release of 5-HT and glutamate from DRN projections to the nucleus accumbens6, we did not observe any time-locked light-evoked EPSCs in the BNST (data not shown). These results indicate that 5-HTDRN→BNST projections have a predominantly excitatory effect that is dependent on 5-HT2CR signaling. To examine the role of 5-HT2CR containing neurons in anxiety-like behavior, we next took advantage of a Htr2cCre mouse line (Extended Data Figure 3a-b)7. Using Designer Receptors Exclusively Activated by Designer Drugs (DREADDs)8, we found that activation of Gq signaling in 5-HT2CR-expressing neurons in the BNST significantly delayed the onset of feeding in the NSF assay without impacting home cage feeding behavior (Extended Data Figure 3c-g), thus phenocopying the effect observed with 5-HTDRN→BNST fiber stimulation during NSF. Taken together, these results provide converging evidence that activation of 5-HTDRN→BNST inputs elicits anxiety-like behavior via 5-HT2CR signaling. We then considered the neurochemical phenotype of these target 5-HTDRN→5-HT2CRBNST neurons and hypothesized that 5-HT via 5-HT2CR modulates the activity of neurons expressing the neuropeptide CRF. This hypothesis was based upon a previous analysis of 5-HT2CR knockout mice, which exhibit an anxiolytic phenotype associated with a reduction of c-fos in CRFBNST neurons9. Initially, using CRF reporter mice to a priori select CRF neurons for recordings, we found a heterogeneous 5HT-induced response of CRFBNST (Extended Data Figure 4a), with only a subset demonstrating a depolarization. Consistent with this, double fluorescence in situ hybridization revealed that only a subset of CRF neurons within the dorsal BNST (~70%) and ventral BNST (~43%%) express 5-HT2CRs (Extended Data Figure 4b-d). While CRF signaling within the BNST is classically associated with anxiety-like behavior10,11, more recent studies using circuit-based tools have found that optogenetic stimulation of GABAergic projections (which include CRFBNST neurons) to the VTA are anxiolytic12. This led us to hypothesize the existence of functionally distinct subsets of CRFBNST neurons that gate different behaviors and are differentially sensitive to 5HT. We used fluorescent retrograde tracer beads to label CRFBNST neurons as VTA-projecting or non-VTA-projecting (Figure 2a) and found that VTA-projecting CRF neurons (CRFBNST→VTA neurons) were hyperpolarized by an average of 5.73 ± 1.24 mV and non-VTA-projecting CRF neurons were depolarized by an average of 2.74 ± 0.39 mV during 5-HT bath application. Moreover, the excitatory response to 5-HT in non-VTA-projecting CRF neurons was reversed in the presence of a 5-HT2C receptor antagonist (Figure 2b). Furthermore, all CRFBNST→VTA neurons were non-responsive to the 5-HT2R agonist meta-Chlorophenylpiperazine (mCPP), while all non-VTA projecting CRF neurons were depolarized by mCPP by an average of 3.78 ± 1.17 mV (Extended Data Figure 4e-h). These findings suggest an anatomically distinct response to 5-HT by different subsets of CRFBNST neurons. The subset of CRFBNST neurons expressing 5-HT2CRs do not project to the VTA and are depolarized by 5-HT, whereas the CRFBNST→VTA neurons are hyperpolarized by 5-HT, via actions at another 5-HT receptor. To determine if this 5-HT-dependent mechanism extended to other anxiolytic efferents, we injected retrograde tracer beads into the lateral hypothalamus (LH) of CRF reporter mice and found 5-HT had similar bidirectional effects on non-LH and LH projecting CRFBNST neurons (Extended Data Figure 5a-c). Noting the functional similarities between these two populations, we used retrograde tracing to determine that roughly ~58% of CRFBNST neurons have projections to the LH or VTA (Extended Data Figure 5d-f). Notably, ~20-31% of these CRFBNST output neurons form parallel projections to these structures. In light of recent reports that CRFBNST neurons are exclusively GABAergic13, we hypothesized that non-VTA-projecting CRFBNST neurons may locally inhibit BNST→VTA neurons to promote fear and anxiety. To test this hypothesis, we injected CrfCre mice with a Cre-inducible ChR2 into the BNST and retrograde tracer beads into the VTA. We then recorded light-evoked IPSCs from non-ChR2 (ChR2-negative, retrograde tracer-positive) VTA-projecting BNST neurons (Figure 2c). Photostimulation produced action potentials in CRFBNST neurons and light-evoked IPSCs in non-ChR2 VTA-projecting neurons, indicating that CRFBNST neurons form local GABAergic synapses with BNST neurons that project to the VTA. Repeating these same experiments in CrfCRE::ChR2BNST mice with retrograde tracer beads in the LH, we found that we could light-evoke GABA currents in LH projecting neurons as well (Extended Data Figure 5g-i). Moreover, we observed that 5-HT increased GABAergic transmission on to BNST→VTA projecting neurons in a tetrodotoxin and 5-HT2CR antagonist dependent manner (Figure 2d-f and Extended Data Figure 5j-n). Similar effects of 5-HT on GABAergic transmission were found in BNST→LH projecting neurons (Extended Data Figure 5o-v). Furthermore, slice recordings in a CRF reporter line indicates that 5-HT does not increase GABAergic transmission on to the general population of CRFBNST neurons nor does it directly excite non-CRF VTA projecting neurons (Extended Data Figure 6). The 5-HT2R agonist mCPP also increased GABAergic but not glutamatergic transmission in the BNST (Extended Data Figure 7). Finally, to test if optically evoked 5-HT can inhibit BNST outputs to the VTA, we performed slice recordings in the BNST of SertCre::ChR2DRN→BNST mice and found that brief photostimulation of 5-HT terminals in the BNST increased sIPSCs on to VTA projecting BNST neurons in a manner similar to bath applied 5-HT (Extended Data Figure 8a-c). Together, these experiments indicate that CRFBNST neurons inhibit at least two major BNST outputs to the VTA and LH that are reported to be anxiolytic 12,14, providing mechanistic insight into the aversive actions of 5-HT signaling in the BNST. We next took advantage of an intersectional strategy for direct visualization of these non projecting, putatively local CRFBNST neurons15. By coupling retrograde Cre-dependence flpases (HSV-LSL1-mCherry-IRES-flpo) in the VTA and LH with INTRSECT(Creon/flpoff)-Chr2-eYFP in the BNST of Crfcre mice (CrfCre::Intrsect-ChR2BNST mice), we were able to genetically isolate non-VTA/LH projecting CRF neurons in the BNST. We also infused Cre-dependent HSV-mCherry vector in a subset of CrfCre::Intrsect-ChR2BNST mice as a control. In HSV-flp infused CrfCre::Intrsect-ChR2BNST mice, we observed a significant reduction in YFP+ cells in the ventral BNST (Extended Data Figure 8d-f), indicating that a large proportion of VTA and LH-projecting CRFBNST neurons are located in the ventral BNST. We also found that 5-HT robustly depolarized these CrfCre::Intrsect-ChR2BNST neurons compared to CRF neurons at large (Figure 2g-i). Furthermore, we observed light evoked IPSCs in the BNST of CrfCre::Intrsect-ChR2BNST mice, confirming local GABA release from these neurons (Extended Data Figure 8g). These results support the existence of a separate population of local CRFBNST neurons that is excited by 5-HT and increases local GABAergic transmission in the BNST, distinct from a population of CRFBNST neurons that project to and release GABA in the VTA or the LH (Extended Data Figure 8h-j). To probe the translational relevance of these BNST microcircuits, we adopted a pharmacological approach using SSRIs. SSRIs represent one of the most widely used classes of drugs for psychiatric disorders. One limitation of SSRIs is that acute administration can lead to negative behavioral states1,2, a finding that is recapitulated in rodent models3,16–20. Importantly, the BNST has been demonstrated to be an anatomical site of action for some of the aversive actions of SSRIs in rodents4. This provided the opportunity to test our model that 5-HT in the BNST drives aversive behavior through inhibition of BNST outputs to the VTA. We observed that an acute systemic injection of the SSRI fluoxetine increased GABAergic transmission on to VTA projecting neurons in the BNST (Figure 3a-d). We then interrogated the role of CRFBNST neurons in acute fluoxetine-enhanced anxiety using CrfCRE mice transduced in the BNST with the Cre-inducible Gi-coupled DREADD. We found that acute fluoxetine potentiated anxiety-like behavior, and this effect was blocked by chemogenetic inhibition of CRFBNST neurons (Figure 3e-h). To evaluate directly whether endogenous 5-HT acts on CRFBNST neurons to enhance cued fear memory, we used the same chemogenetic approach to silence CRFBNST neurons during fluoxetine treatment and subsequent fear conditioning (Figure 3i). Chemogenetic inhibition of CRFBNST neurons also significantly attenuated fluoxetine-induced enhancement of cued fear recall, providing proof of concept that augmentation of 5-HT via acute SSRI treatment recruits CRFBNST neurons to enhance fear-related behavior (Figure 3j-k). Next, using connectivity based chemogenetic approaches; we tested whether inhibition of BNST outputs to the VTA and LH is a critical component of 5-HT→BNST-induced aversive states. We observed that activation of Gq signaling in VTA- and LH-projecting BNST neurons, targeted by HSV-Cre-eYFP infused in the VTA and LH and Cre-dependent Gq-coupled DREADD infused in the BNST (HSVCre::hM3DqBNST), significantly attenuated fluoxetine enhancement of cued fear recall (Figure 3l-o). Together, these data provide compelling evidence that acute fluoxetine engenders aversive behavior by recruiting CRF neurons in the BNST that in turn inhibit putative GABAergic (anxiolytic and stress buffering) outputs from the BNST to the VTA and LH. Pharmacological interventions that target this circuit may improve adverse symptoms during the initial weeks of SSRI treatment. Based on the critical role for CRFBNST neurons in fluoxetine induced aversive behavior, we examined the impact of a systemic CRF1R antagonist on SSRI enhancement of cued fear recall. Notably, blocking the CRF system reduced this aversive state and abolished the increase in sIPSCs in LH-projecting neurons in the BNST during bath application of 5-HT (Extended Data Figure 9). This provides translational evidence that CRF1R antagonists given in concert with SSRIs could be a promising treatment for anxiety disorders. Taken together, these data reveal a discrete 5-HT responsive circuit in the BNST that underlies pathological anxiety and fear associated with a hyperserotonergic state (Extended Data Figure 10). SSRIs are currently a first-line treatment for anxiety and panic disorders but can acutely exacerbate symptoms, resulting in poor therapeutic compliance. Our results strongly implicate 5-HT engagement of a local BNST inhibitory microcircuit in acute SSRI induced aversive behaviors in rodents, and could potentially be involved in the early adverse events seen in clinical populations, emphasizing the need to identify compounds that selectively target both genetically-defined and pathway-specific cell populations. Methods Mice Mice were used in all experiments. For experiments involving Cre lines, mice were crossed for several generations to C57 mice before using. All wild-type mice were C57BL/6 mice obtained from The Jackson Laboratory (Bar Harbor, ME). For all behavior experiments except those involving Htr2cCre mice, male mice ranging in age from 8-16 weeks were used. Female Htr2cCre mice were used in chemogenetic manipulations. Both male and female mice aged 6-20 weeks were used for slice electrophysiology and anatomical tracing experiments. All behavioral studies or tissue collection for ex vivo slice electrophysiology were performed during the light cycle. All behavioral experiments in Htr2ccre mice were conducted at the University of Aberdeen and in accordance with the United Kingdom Animals (Scientific Procedures) Act of 1986. All in vivo electrophysiology experiments were conducted in accordance with all rules and regulations at the National Institute for Alcohol Abuse and Alcoholism at the National Institutes of Health. All other procedures were conducted in accordance with the National Institutes of Health guidelines for animal research and with the approval of the Institutional Animal Care and Use Committee at the University of North Carolina at Chapel Hill. All animals were group housed on a 12 hour light cycle (lights on at 7 a.m.) with ad libitum access to rodent chow and water, unless described otherwise. CRF-ires-Cre (Crfcre) were provided by Dr. Bradford Lowell (Harvard University) and were previously described21. C57BL/6J mice were obtained from the Jackson Laboratory (Bar Harbor, ME). To visualize CRF-expressing neurons, CrfCre mice were crossed with either an Ai9 or a cre-inducible L10-GFP reporter line (Jackson Laboratory)22 to produce CRF-Ai9 or CRF-L10GFP progeny, referred to throughout the manuscript as CRF-reporters. SertCre mice (from GENSAT) were a generous gift from Dr. Bryan Roth. Htr2cCre mice were supplied by Dr. Lora Heisler and are described in detail elsewhere7. Male mice were used for in vivo optogenetic behavioral experiments and for assessing the involvement of BNST CRF neurons on fluoxetine-induced enhancement of fear. Female 5-HT2C-Cre mice were used in chemogenetic manipulations. Both male and female mice were used for slice electrophysiology and anatomical tracing experiments. All behavioral studies or tissue collection for ex vivo slice electrophysiology were performed during the light cycle. Viruses and tracers All AAV viruses except INTRSECT constructs were produced by the Gene Therapy Center Vector Core at the University of North Carolina at Chapel Hill and had titers of >1012 genome copies/mL. For ex vivo and in vivo optical experiments, mice were injected with rAAV5-ef1α-DIO-hChR2(H134R)-eYFP or rAAV5-ef1α-DIO-eYFP as a control. Red IX retrobeads (Lumafluor) were used to fluorescently label LH - and VTA-projecting BNST neurons during ex vivo slice electrophysiology recordings. The retrograde tracer Fluoro-Gold (Fluorochrome) was used for anatomical mapping. Choleratoxin B (CTB) 555 and CTB 657 retrograde tracers (Invitrogen; C34776, and C34778, respectively) diluted to 0.5% (w/v) in sterile PBS were used per injection site for anatomical mapping of collateral projections from BNST to LH and VTA. For chemogenetic manipulations, mice were injected with 400 nl of rAAV8-hsyn-DIO-hM3D(Gq)-mCherry, rAAV8-hsyn-DIO-hM4D(Gi)-mCherry, or rAAV8-hsyn-DIO-mCherry bilaterally. HSV-hEF1α-mCherry, HSV-ef1α-LSL1-mCherry-IRES-flpo, and HSV-ef1α-IRES-Cre (supplied by Rachel Neve at the McGovern Institute for Brain Research at MIT) were injected bilaterally into the VTA and LH at a volume of 500500 nL per sitesite. The INTRSECT construct AAVdj-hSyn-Con/Foff-hChR2(H134R)-EYFP was infused at 500 nl per side into the BNST. All AAV constructs had viral titers >1012 genome particles/ml. Stereotaxic injections All surgeries were conducted using aseptic technique. Adult mice (2-5 months) were deeply anesthetized with 5% isoflurane (vol/vol) in oxygen and placed into a stereotactic frame (Kopf Instruments) while on a heated pad. Sedation was maintained at 1.5-2.5% isoflurane during surgery. An incision was made down the midline of the scalp and a craniotomy was performed above the target regions and viruses and fluorescent tracers were microinjected using a Neuros Hamilton syringe at a rate of 100 nl/min. After infusion, the needle was left in place for 10 minutes to allow for diffusion of the virus before the needle was slowly withdrawn. Injection coordinates (in mm, midline, Bregma, dorsal surface): BNST (±1.00, 0.30, -4.35), LH (±0.9 to 1.10, -1.7, -5.00 to -5.2), VTA (-0.3, -2.9, -4.6), DR (0.0, -4.65, -3.2 with a 23° angle of approach). When using retrobeads, injection volumes into the LH and VTA were 300 nl and 400 nl, respectively. Fluorogold injection volumes were 200 nl per target site. CTB volumes were 200200 nL per target site. An optical fiber was implanted in the BNST (±1.00, 0.20, -4.15) at a 10° angle for in vivo photostimulation studies. After fiber implantation, dental cement was used to adhere the ferrule to the skull. Following surgery, all mice returned to group housing. Mice were allowed to recover for at least 3 weeks before being used for chemogenetic behavioral studies, or 6 weeks for in vivo optogenetic studies. Drugs RS 102221, 5-HT and mCPP were from Tocris (Bristol, UK). For electrophysiology experiments, RS 102221 was made up to 100 mM in DMSO and then diluted to a final concentration of 5 µM in aCSF. 5-HT and mCPP were stocked at 10 and 20 mM, respectively, in ddH2O and diluted to their final concentations in aCSF. For electrophysiology experiments, clozapine-N-oxide (CNO; from Dr. Bryan Roth) was stocked at 100 mM in DMSO and diluted to 10 µM in aCSF. For behavior experiments, CNO was dissolved in 0.5% DMSO (in 0.9% saline) to a concentration of 0.1 mg/ml or 0.3 mg/ml and injected at 10 ml/kg for a final concentration of 1 or 3 mg/kg, i.p. Fluoxetine (Sigma) was made up in 0.9% NaCl to a concentration of 1 mg/ml and then injected at 10 ml/kg for a final concentration of 10 mg/kg, i.p. In vivo Electrophysiological Procedures Surgical Procedures Mice were anesthetized with 2% Isoflurane (Baxter Healthcare, Deerfield, IL) and implanted with 2x8 electrode (35um tungsten) micro-arrays (Innovative Neurophysiology Inc, Durham, NC) targeted at the BNST (ML: 0.8 mm, AP: ± 0.5 mm , and DV: -4.15 mm relative to Bregma). Following surgery, mice were singly housed and allowed at least one week to recover prior to behavioral testing. Fear Conditioning Fear conditioning took place in 27 × 27 × 11cm conditioning chambers (Med Associates, St. Albans, VT), with a metal-rod floor (Context A) and scented with 1% vanilla. Mice received 5 parings of a pure tone CS with a .6mA foot shock. 24 h following conditioning, mice underwent a CS recall test (10 presentations of the CS alone, 5 sec ITI), which was conducted in a Plexiglas cylinder (20cm diameter) and scented with 1% acetic acid (Context B). Stimulus presentations for both tests were controlled by MedPC (Med Associates Inc, St. Albans, VT). Cameras were mounted overhead for recording freezing behavior, which was scored automatically using CinePlex Behavioral Research System software (Plexon Inc, Dallas, TX). Electrophysiological recording and single unit analysis Electrophysiological recording took place during both fear conditioning and CS recall tests. Individual units were identified and recorded using Omniplex Neural Data Acquisition System (Plexon Inc, Dallas, TX). Neural data was sorted using Offline Sorter (Plexon Inc, Dallas, TX). Waveforms were isolated manually, using principal component analysis. To be included in the analyses, spikes had to exhibit a refractory period of at least 1 ms. Autocorrelograms from simultaneously recorded units were examined to ensure that no cell was counted twice. Single units were analyzed by generating perievent histograms (3 sec bins) of firing rates from 30 sec prior to CS onset until 30 sec after CS offset (NeuroExplorer 5.0, Nex Technologies, Madison, AL). Firing rates were normalized to baseline (30 sec prior to CS onset) using z-score transformation. Analysis included a total of 139 cells over three days of recording. Data reported for raw firing rates include only putative principal neurons (<10Hz). The formula for computing the suppression ratio was (average freezing rate) / (average freezing rate + average movement rate). Each cell was calculated individually. A value of .5 = no change in rate). Ex vivo Slice Electrophysiology Brains were sectioned at 0.07 (mm/s) on a Leica 1200S vibratome to obtain 300 µm coronal slices of the BNST, which were incubated in a heated holding chamber containing normal, oxygenated aCSF (in mM:124 NaCl, 4.4 KCl, 2 CaCl2, 1.2 MgSO4, 1 NaH2PO4, 10.0 glucose, and 26.0 NaHCO3) maintained at 30 ± 1°C for at least 1 hour before recording. Slices were transferred to a recording chamber (Warner Instruments) submerged in normal, oxygenated aCSF maintained at 28-30°C at a flow rate of 2 ml/min. Neurons of the BNST were visualized using infrared differential interference contrast (DIC) video-enhanced microscopy (Olympus). Borosilicate electrodes were pulled with a Flaming-Brown micropipette puller (Sutter Instruments) and had a pipette resistance between 3-6 MΩ. Signals were acquired via a Multiclamp 700B amplifier, digitized at 10 kHz and analyzed with Clampfit 10.3 software (Molecular Devices, Sunnyvale, CA, USA). Light-evoked action potentials In SertCre or CrfCre mice, fluorescently labeled neurons expressing ChR2 were visualized and stimulated with a blue (470 nm) LED using a 1 Hz, 2 Hz, 5 Hz, 10 Hz, and 20 Hz stimulation protocol with a pulse width of 0.5 ms. Evoked action potentials were recorded in current clamp mode using a potassium gluconate based internal solution (in mM: 135 K+ gluconate, 5 NaCl, 2 MgCl2, 10 HEPES, 0.6 EGTA, 4 Na2ATP, 0.4 Na2GTP, pH 7.3, 285–290 mOsmol). Light-evoked synaptic transmission In CrfCre mice with ChR2 in the BNST and retrograde tracer beads in the VTA or LH, we visualized non-ChR2-expressing, beaded neurons using green (532 nm) LED. Recordings were conducted in voltage clamp mode using a cesium-methansulfonate (Cs-Meth) based internal solution (in mM: 135 cesium methanesulfonate, 10 KCl, 1 MgCl2, 0.2 EGTA, 2 QX-314, 4 MgATP, 0.3 GTP, 20 phosphocreatine, pH 7.3, 285–290 mOsmol) so that we could detect EPSCs (-55 mV) and IPSCs (+10 mV) in the same neuron. After confirming the absence of a light-evoked EPSC signal, we measured light-evoked IPSCs during a single, 5 ms light pulse of 470 nm. In a subset of these experiments, SR95531 (GABAzine, 10 µM) was bath applied for 10 minutes to block IPSCs. Drug effects in CRFBNST neurons Crf-reporter mice were injected with retrograde tracer beads into the VTA (ML -0.5, AP -2.9, DV -4.6). We then recorded from beaded (VTA-projecting) and non-beaded (non-projecting) CRF neurons in the BNST. Acute drug effects were determined in current clamp mode in the presence of TTX using a potassium gluconate-based internal solution. After a 5-minute stable baseline was established, 5HT (10 µM) or mCPP (20 µM) was bath applied for 10 minutes while recording changes in membrane potential. The difference in membrane potential between baseline and drug application at peak effect (delta or Δ MP) was later determined. In a subset of mCPP experiments, slices were incubated with RS 102221 (5 µM) for at least 20 minutes before experiments began. Synaptic transmission Spontaneous inhibitory postsynaptic currents (sIPSCs) were assessed in voltage clamp using a potassium-chloride gluconate-based intracellular solution (in mM: 70 KCl, 65 K+-gluconate, 5 NaCl, 10 HEPES, 0.5 EGTA, 4 ATP, 0.4 GTP, pH 7.2, 285–290 mOsmol). IPSCs were pharmacologically isolated by adding kynurenic acid (3 mM) to the aCSF to block AMPA and NMDA receptor-dependent postsynaptic currents. The amplitude and frequency of sIPSCs were determined from 2 minute recording episodes at -70 mV. The baseline was averaged from the 4 minutes preceding the application of 5-HT (10 µM) or mCPP (10 µM) for 10 minutes. In a subset of these experiments, RS 102221 (5 µM) was added to the aCSF and slices were incubated in this drug solution for at least 20 minutes before experiments began. For miniature IPSCs (mIPSCs), TTX was included in the aCSF to block network activity. In SertCre::ChR2BNST mice with retrograde tracer beads in the VTA, sIPSCs were recorded as described above. After achieving a stable baseline, a 10 s, 20 Hz photostimulation was applied. For assessment of spontaneous excitatory postsynaptic currents (sEPSCs), a cesium gluconate-based intracellular solution was used (in mM: 135 Cs+-gluconate, 5 NaCl, 10 HEPES, 0.6 EGTA, 4 ATP, 0.4 GTP, pH 7.2, 290–295 mOsmol). AMPAR-mediated EPSCs were pharmacologically isolated by adding 25 μM picrotoxin to the aCSF. sEPSC recordings were acquired in 2 minute recording blocks at -70 mV. Fast-scan cyclic voltammetry (FSCV) Electrodes were fabricated as previously described and cut to 50-100 um in length23. Animal and slice preparation were as described above for electrophysiology and slices were perfused on the rig in ACSF. Using a custom built potentiostat (University of Washington Seattle), 5-HT recordings were made in the BNST using TarHeel CV written in lab view (National Instruments). Briefly a triangular waveform (-0.1 V to 1.3 V with a 10% phase shift at 1000 V/s, vs. Ag/AgCl) was applied to the carbon fiber electrode at a rate of 10 Hz. Slices were optically stimulated with 20 5-ms blue (490 nm) light pulses at a rate of 20 Hz down the submerged 40x objective. 10 cyclic voltammograms were averaged prior to optical stimulation for background subtraction. Voltammograms were digitally smoothed one time with a fast Fourier transform following data collection and analyzed with HDCV (UNC Chapel Hill). Fluoxetine (10 µM) was bath applied following a stable baseline (20 minutes). Behavioral Assays For chemogenetic manipulations, mice were transported to a holding cabinet adjacent to the behavioral testing room to habituate for at least 30 minutes before being pretreated with CNO (3 mg/kg, i.p. for CrfCre mice and 1 mg/kg, i.p. for Htr2CCre mice). All behavior testing began 45 minutes following CNO treatment, with the exception of fear conditioning training, which occurred 30 minutes after a CNO injection. When assessing the effect of fluoxetine on fear conditioning, fluoxetine (10 mg/kg, i.p.), or vehicle, was administered 1 hour before training (30 minutes before CNO treatment). For optogenetic manipulations, mice received bilateral stimulation (473 nm, ~10 mW, 5 ms pulses, 20 Hz) when specified. Unless specified, all equipment was cleaned with a damp cloth between mouse trials. All sessions were video recorded and analyzed using EthoVision software (Noldus Information Technologies) except where noted. Elevated Plus Maze Mice were placed in the center of an elevated plus maze and allowed to explore during a 5 minute session. Light levels in the open arms were ~14 lux. During optogenetic manipulations mice received bilateral stimulation during the entire 5 minute session. Mice that left the maze were excluded from analysis (n= 2 control, 1 ChR2 from optogenetic experiments). Open Field Mice were placed into the corner of a white Plexiglas open field arena (25 x 25 x 25 cm) and allowed to freely explore for 30 minutes. The center of the open field was defined as the central 25% of the arena. For optogenetic studies the 30 minute session was divided into three 10-minute epochs consisting of stimulation off, stimulation on, and stimulation off periods. Novelty-Induced Suppression of Feeding 48 hours prior to testing mice were provided with access to a single piece of Froot Loops cereal (Kellogg’s) in their home cage. 24 hours prior to testing, home cage chow was removed and mouse body weights were recorded. Water remained available ad lib. Beginning at least one hour before testing mice transferred to new clean cages so they were singly housed for the test session and body weights were recorded. During the test session mice were placed into an arena (25×25×25 cm) that contained a single Froot Loop on top of a piece of circular filter paper. Mice were monitored by a live observer and the latency for the mouse to begin eating the pellet was measured, allowing up to 10 minutes. All mice began eating within this time. Following the initiation of feeding, mice were removed from the arena and placed back into their home cages. Mice were then provided with 10 minutes of access to a pre-weighed amount of Froot Loops™ for a post-test feeding session. After this 10 minute post-test, the remaining Froot Loops were weighed and mice were returned to ad lib home cage chow. Mice were returned to group housing at the end of this session. For optogenetic experiments, mice received constant 20 Hz optical stimulation during both the latency to feed assay and the 10 minute post-test. During optogenetic experiments, one control mouse did not feed during the 10 minute NSF session and was excluded from the results. Home cage feeding SertCre mice were food deprived for 24 hours. On the day of the experiment, mice were acclimated to the behavior room for 1 hour. A single preweighed food pellet was placed in the home cage and the mice were allowed to eat for 10 minutes during optogenetic stimulation. At the end of the experimental session, the pellet was removed and weighed and mice were given ad lib access to food. Htr2CCre mice were acclimated in metabolic chambers (TSE Systems, Germany) for 2 days before the start of the recordings. After acclimation, mice were food deprived for 24 hours. Following fasting, mice received an IP injection of CNO 30 minutes before food presented again. Mice were recorded for 12 hours with the following measurements being taken every 30 minutes: water intake, food intake, ambulatory activity (in X and Z axes), and gas exchange (O2 and CO2) (using the TSE LabMaster system, Germany). Energy expenditure was calculated according to the manufacturer’s guidelines (PhenoMaster Software, TSE Systems). Fear Conditioning We used a three day protocol to assess both cued and contextual fear recall. On the first day, mice were placed into a fear conditioning chamber (Med Associates) that contained a grid floor and was cleaned with a scented paper towel (19.5% EtOH, 79.5% H2O, 1% vanilla). After a 3 minute baseline period, mice were exposed to a 30 second tone (3 KHz, 80 dB) that co-terminated with a 2 second scrambled foot shock (0.6 mA). A total of 5 tone-shock pairings were delivered with a random inter-tone interval (ITI) of 60-120 seconds. For optogenetic studies, light stimulation occurred only during the 30-second tones of this session. Following delivery of the last foot shock, mice remained in the conditioning chamber for a two minute consolidation period. 24 hours later, mice were placed into a separate conditioning box (Med Associates) that contained a white Plexiglas floor, a striped pattern on the walls, and was cleaned and scented with a 70% EtOH solution. After a 3 minute baseline period, mice were presented with 10 tones (30 seconds, 3 KHz, 80 dB) with a 5 second ITI. Mice remained in the chamber after the last tone for a two-minute consolidation period. 24 hours later (48 hours after training), mice were returned to the original training chamber for 5 minutes. For each session, freezing behavior was hand-scored every 5 seconds by a trained observer blinded to experimental treatment as described previously24. Freezing was defined as a lack of movement except as required for respiration. Immunohistochemistry and histology All mice used for behavioral and anatomical tracing experiments were anesthetized with Avertin and transcardially perfused with 30 ml of ice-cold 0.01M PBS followed by 30 ml of ice-cold 4% paraformaldehyde (PFA) in PBS. Brains were extracted and stored in 4% PFA for 24 hours at 4°C before being rinsed twice with PBS and stored in 30% sucrose/PBS until the brains sank. 45 µm slices were obtained on a Leica VT100S and stored in 50/50 PBS/Glycerol at -20°C. DREADD or ChR2-containing sections were mounted on slides, allowed to dry, coverslipped with VectaShield (Vector Labs, Burlingame, CA), and stored in the dark at 4°C. Tryptophan hydroxylase/Fluorogold/cfos triple labeling We stained free-floating dorsal raphe sections using indirect immunofluorescence sequentially for first tryptophan hydroxylase (TPH) and Fluoro-Gold(FG) and then c-fos. For TPH/FG, we washed sections 3X for 5 min with 0.01 M PBS, permeabilized them for 30 min in 0.5% Triton/0.01 M PBS, and washed the sections again 2X with 0.01 M PBS. We blocked the sections for 1 hour in 0.1% Triton/0.01 M PBS containing 10%(v/v) Normal Donkey Serum and 1%(w/v) Bovine Serum Albumin (BSA). We then added primary antibodies (1:500 Mouse anti-TPH [Sigma Aldrich T0678] and 1:3000 Guinea Pig anti-Fluoro Gold [Protos Biotech NM101]) to blocking buffer and incubated the sections overnight at 4 degrees C. The next day, we washed the sections 3X for 5 min with 0.01 M PBS, then incubated them with 1:500 with Alexa Fluor 647-conjugated Donkey anti-mouse and Alexa Fluor 488-conjugated Donkey anti-guinea pig secondary antibodies for 2 hr at RT, and washed the sections 4X for 5 min with 0.01 M PBS. We then proceeded directly to the c-fos tyramide signal amplification based immunofluorescent staining. We permeabilized the sections in 50% methanol for 30 min, then quenched endogenous peroxidase activity in 3% hydrogen peroxide for 5 min. Followed by two 10 min washes in 0.01 M PBS, we blocked the sections in PBS containing 0.3% Triton X-100 and 1.0 % BSA for 1 hour. c-fos primary antibody (Santa Cruz Biotechnology - sc-52) was added to sections at 1:3000 and sections were incubated for 48 hours at 4 degrees. On day 3, we washed the sections in TNT buffer (0.1 M Tris-HCl pH 7.5, 0.15 M NaCl, 0.05% Tween-20) for 10 min, blocked in TNB buffer (0.1 M Tris-HCl pH 7.5, 0.15 M NaCl, 0.5% Blocking reagent – PerkinElmer FP1020) buffer for 30 min. We then incubated the sections in secondary antibody (Goat anti-rabbit HRP-conjugated- PerkinElmer) 1:200 in TNB buffer for 30 min., washed the sections in TNT buffer 4X for 5 min, and then incubated the sections in Cy3 dye diluted in TSA amplification diluents for 10 min. We washed the sections 2X in TNT buffer, mounted them on microscope slides. We coverslipped the slides using Vectashield mounting medium. We acquired 4-5 2x4 tiled z-stack(5 optical slices comprising 7 µm total) images of the dorsal raphe from each naïve and shock mouse on a Zeiss 800 Upright confocal microscope. Scanning parameters and laser power were matched between groups. Images were preprocessed using stitching and maximum intensity projection and then analyzed using an advanced processing module in Zeiss Zen Blue that allows nested analysis of multiple segmented fluorescent channels within parent classes. Double and triple-labeled cells were validated in a semi-automated fashion. At least 4 sections per mouse were counted in this way. One mouse was identified as a significant outlier in the Shock group and was excluded from further analysis. SertCre::ChR2, and CrfIntrsect-ChR2 validation To verify expression of ChR2-expressing fibers in the BNST originating from DR serotonergic neurons, 300 µm slices used for ex vivo electrophysiological recordings containing the DR and BNST were stored in 4% paraformaldehyde at 4°C for 24 hours before being rinsed with PBS, mounted, and coverslipped with Vectashield mounting medium. Images showing eYFP fluorescence from the DR and BNST were obtained on a Zeiss 800 upright confocal microscope using a 10x objective and tiled z stacks. To validate the INTRSECT construct, mice received injections of HSV-hEF1α-mCherry or HSV-ef1α-LSL1-mCherry-IRES-flpo to both the LH and VTA bilaterally (N=4 and 5, respectively). Both groups received AAVDJ-hSyn-Cre-on/Flp-off-hChR2(H134R)-EYFP to the BNST bilaterally. Six weeks following injection, mice were perfused and tissue was collected as described above. To visualize YFP expression in the BNST of CrfCre::IntrsectBNST mice, free floating slices containing the BNST were rinsed three times with PBS for 5 minutes each. Slices were then incubated in 50% methanol for 30 minutes then incubated in 3% hydrogen peroxide for 5 minutes. Following three 10-minute washes in PBS, slices were incubated in 0.5% Triton X-100 for 30 minutes followed by a 10 minute PBS wash. Slices were blocked in 10% normal donkey serum/0.1% Triton X-100 for 1 hour, and then they were incubated overnight at 4°C with a primary chicken anti-GFP antibody (GFP-1020, Aves) at 1:500 in blocking solution. Following primary incubation, slices were rinsed three times with 0.01M PBS for 10 minutes each and incubated with a fluorescent secondary antibody (AlexaFluor 488 Donkey anti-chicken) at 1:200 in PBS for 2 hours at room temperature. Slices were then rinsed with four 10-minute PBS washes before being mounted onto glass slides and coverslipped with Vectashield with DAPI. A 3x4 tiled z stack (7 optical sections comprising 35 µm total) image from both the left and right hemispheres of the BNST was obtained at 20x magnification using a Zeiss 800 upright confocal microscope. Scanning parameters and laser power were matched between groups. Images were preprocessed using stitching and maximum intensity projection. The number of fluorescent cells in the dorsal and ventral aspects of the BNST were counted by a blinded scorer using the cell counter plug-in in FIJI (ImageJ). Each hemisphere was considered independently per mouse. One mouse in the flp-expressing group was a significant outlier for number of cells expressed in a ventral BNST hemisphere (ROUT, Q=0.1%) and all data from that mouse were excluded. Choleratoxin retrograde tracer studies in CRF reporter mice 3 male CRF-L10a reporter mice were injected with 200 nl of CTB 555 and CTB 647 bilaterally to the LH and VTA, respectively, as described above. 5 days following injection, mice were perfused as described above, the brains were extracted, and were stored in 4% paraformaldehyde for 24 hours at 4°C before being rinsed with PBS and transferred to 30% sucrose until the brains sank. 45 µm sections containing the BNST were collected as described above. Sections containing the BNST were mounted on glass slides and coverslipped using Vectashield. An image from the left and right hemispheres of a medial section of the BNST was obtained on a Zeiss 800 upright microscope using a 20x objective and 3x5 tiled z stacks (5 optical slices comprising 7 µm total). Images were preprocessed using stitching and maximum intensity projection, and were then analyzed using the cell counter function in FIJI (ImageJ). Only cells positive for GFP (putative CRF neurons) were considered. Cells were scored exclusively as either 555+ only (LH-projecting), 647+ only (VTA-projecting), 555+ and 647+ (projecting to both LH and VTA), or 555- and 647- (unlabeled; neither LH- nor VTA- projecting). The total number of CRF neurons scored was calculated as the sum of all four groups, and percentages of each type were calculated from this value. Each hemisphere was scored and plotted independently (N=6 images from 3 mice), and the dorsal and ventral BNST were considered separately. The average values were plotted as pie charts (ED 9). Double Fluorescence in situ hybridization (FISH) For validation of 2C-cre line and comparison of CRF/2C mRNA cellular colocalization, mice were anesthetized using isoflurane, rapidly decapitated, and brains rapidly extracted. Immediately after removal, the brains were placed on a square of aluminum foil on dry ice to freeze. Brains were then placed in a -80°C freezer for no more than 1 week before slicing. 12 µm slices were made of the BNST on a Leica CM3050S cryostat (Germany) and placed directly on coverslips. FISH was performed using the Affymetrix ViewRNA 2-Plex Tissue Assay Kit with custom probes for CRF, 5-HT2C, and Cre designed by Affymetrix (Santa Clara, CA). Slides were coverslipped with SouthernBiotech DAPI Fluoromount-G. (Birmingham, AL). 3x5 tiled z stack (15 optical sections comprising 14 µm total) images of the entire 12 µm slice were obtained on a Zeiss 780 confocal microscope for assessment of CRF/2C colocalization. A single-plane 40x tiled image of a CRF/2C slice was obtained on a Zeiss 800 upright confocal microscope for the magnified image shown in Extended Data 6b, right. 3x5 tiled z stack (7 optical sections comprising 18 µm) images of 2C/Cre slices were obtained on a Zeiss 800 upright confocal microscope for the 2C/Cre validation. All images were preprocessed with stitching and maximum intensity projection. An image of the BNST from 3 mice in each condition was hand counted for each study using the cell counter plugin in FIJI (ImageJ). Cells were classified into three groups: probe 1+, probe 2+, or probe 1 and 2 +. Only cells positive for a probe were considered. Results are plotted as average classified percentages across the three images. Group assignment No specific method of randomization was used to assign groups. Animals were assigned to experimental groups so as to minimize the influence of other variables such as age or sex on the outcome. Inclusion/exclusion criteria Pre-established criteria for excluding mice from behavioral analysis included 1) missed injections, 2) anomalies during behavioral testing, such as mice falling off the elevated plus maze, 3) damage to or loss of optical fibers, 4) statistical outliers, as determined by the Grubb’s test. Sample size A power analysis was used to determine the ideal sample size for behavior experiments. Assuming a normal distribution, a 20% change in mean and 15% variation, we determined that we would need 8 mice per group. In some cases, mice were excluded due to missed injections or lost optical fibers resulting in fewer than 8 mice per group. For electrophysiology experiments, we aimed for 5-7 cells from 3-4 mice. Statistics Data are presented as means ± SEM. For comparisons with only two groups, p values were calculated using paired or unpaired t-tests as described in the figure legends. Comparisons across more than two groups were made using a one-way ANOVA, and a two-way ANOVA was used when there was more than one independent variable. A Bonferonni posttest was used following significance with an ANOVA. In cases in which ANOVA was used, the data met the assumptions of equality of variance and independence of cases. If the condition of equal variances was not met, Welch’s correction was used. Some of the sample groups were too small to detect normality (<8 samples) but parametric tests were used because nonparametric tests lack sufficient power to detect differences in small samples (Graphpad Statistics Guide – www.graphpad.com). The standard error of the mean is indicated by error bars for each group of data. Differences were considered significant at p values below 0.05. All data were analyzed with GraphPad Prism software. Extended Data Extended Data Figure 1 In vivo recordings in BNST neurons during fear conditioning reveal opposite patterns of activation during acquisition and recall. (a) Representative neuron firing rate and (b) population Z score of the firing rate for BNST neurons (n=45 cells from 7 mice) 30 s before conditioned stimulus (tone), during the conditioned stimulus, and 30 seconds after the unconditioned stimulus. (c) Percentage time spent freezing during fear acquisition, cued fear recall and contextual fear recall. (d) Electrode placements for BNST recordings. (e) Raw firing rates during freezing (blue) versus movement (red) epochs were averaged across all putative principal neurons (firing rate <10Hz). Acquisition: Cells in BNST exhibited greater average firing rates during freezing epochs compared to movement epochs during CS3 (t44=2.88, p<0.01, Student’s unpaired two-tailed t-test), 4 (t44=3.14, p<0.01, Student’s unpaired two-tailed t-test), and 5 (t44=4.4, p<0.001, Student’s unpaired two-tailed t-test) (n=45 cells from 7 mice). CS Recall: Average firing rates during freezing epochs decreased over CS presentations such that firing during block 5 was significantly less than block 1 (t41=3.44, p=0.001, Student’s unpaired two-tailed t-test). Freezing firing rates during block 5 were also significantly less than movement epochs during block 5 (t41=4.03, p<0.001, Student’s unpaired two-tailed t-test) (n=42 cells from 7 mice). CX test: Average firing rate was significantly greater during movement versus freezing epochs during minutes 1 (t44=4.83, p<0.001, Student’s unpaired two-tailed t-test), 2 (t44=3.17, p<0.01, Student’s unpaired two-tailed t-test), and 5 (t44=4.36, p<0.001, Student’s unpaired two-tailed t-test) (n=45 cells from 7 mice). (f) Freezing-related changes in firing rates during the CS were determined by measuring the ratio of average firing rates during freezing versus movement epochs for each session. Acquisition: Activity during freezing epochs increased significantly relative to movement epochs during CS4 (t45=3.26, p<0.01, Student’s unpaired two-tailed t-test) and CS5 (t45=2.17, p<0.05, Student’s unpaired two-tailed t-test) (n=46 cells from 7 mice). CS Recall: Freezing significantly suppressed activity relative to movement epochs during the last two CS presentations (t47=5.29, p=<0.001, Student’s unpaired two-tailed t-test) (n=48 cells from 7 mice) CX test: Freezing significantly suppressed activity during minutes 1 (t44=6.06, p<0.001, Student’s unpaired two-tailed t-test), 2 (t44=2.92, p<0.01, Student’s unpaired two-tailed t-test), and 5 (t44=3.55, p=.001, Student’s unpaired two-tailed t-test) (n=45 cells from 7 mice). (g) Plots showing correlation between freezing behavior and firing rate of BNST neurons across sessions and for all sessions. Data are mean ± s.e.m. P<0.05 P<0.01; *P<0.001. Scale bar = 100 µm. Extended Data Figure 2 Effects of optogenetic stimulation of 5HT inputs to the BNST on feeding, anxiety and locomotion. (a-c) SertCre::ChR2DRN→BNST mice exhibited reduced probability (t15=2.67, p<0.05, Student’s unpaired two-tailed t-test, n=8 control, n=9 ChR2) and latency (t15=1.003, p>0.05, Student’s unpaired two-tailed t-test, n=8 control, n=9 ChR2) to enter the open arms of the EPM without exhibiting locomotor deficits. (d) Photostimulation of 5-HTDRN→BNST terminals had no effect on locomotor activity in the open field (n=9 control, n=11 ChR2) or (e) home cage feeding (n=4 control, n=6 ChR2). Data are mean ± s.e.m. P<0.05. Extended Data Figure 3 Chemogenetic activation of 5-HT2CR expressing neurons in the BNST increases anxiety-like behavior. (a) Confocal images of coronal BNST slices obtained from htr2cCre mice following double fluorescence in situ hybridization for 5-HT2CR and cre. Yellow arrows indicate cells in which there is colocalization, red arrows indicate cells in which only Cre is expressed and green arrows indicate cells in which only 5-HT2CR is expressed. (b) Pie chart representing the distribution of genetic markers in BNST neurons. (c) Experimental configuration in Htr2ccre::hM3DqBNST mice. (d) Coronal images showing cfos induction in 5-HT2CR expressing neurons in the BNST of Htr2cCre::hM3DqBNST or Htr2cCre::mCherryBNST mice following CNO injection. (e) Bath application of CNO depolarized 5HT2CR-expressing neurons expressing hM3Dq in slice (n=3 cells from 3 mice). (f) Chemogenetic stimulation of 5-HT2CR expressing neurons in BNST increased latency to feed in the NSF (t11=2.591, p<0.05, Student’s unpaired two-tailed t-test, n=6; mCherry, n=7 hM3Dq). (g) Chemogenetic activation of 5-HT2CR-expressing BNST neurons had no effect on home cage feeding (n=5 mCherry, n=6 hM3Dq). (h) Confocal images from Htr2cCre::mCherryBNST mice showing mCherry expression in 5-HT2CR-expressing soma in the BNST and fibers in the LH and VTA. Data are mean ± s.e.m. P<0.05. Scale bar = 100 µm. Extended Data Figure 4 Electrophysiological characterization of 5-HT responses and 5-HT receptor expression in CRFBNST neurons (a) A pie chart showing the distribution of CRFBNST neurons that were depolarized, hyperpolarized, or had no response to 5-HT (n=8 cells from 4 mice). (b) Coronal images of the BNST showing colocalization of 5-HT2CRs with CRF mRNA using double fluorescence in situ hybridization and (c-d) histograms showing the % of 5-HT2C neurons that express CRF and the % of CRF neurons that express 5-HT2CRs in the BNST (n=3 slices from 3 mice). (e) Recording configuration in CRFBNST neurons. (f) Slice electrophysiology in BNST of Crf reporter mice showing depolarization of all (VTA-projecting and non-projecting) CRF neurons following bath application of the 5-HT2 receptor agonist mCPP (n=12 cells from 6 mice) and blockade of this response by the 5-HT2C receptor antagonist RS 102221 (n=5 cells from 3 mice). (g) Change in membrane potential induced by mCPP (t12=2.18, p<0.05, One-sample t-test, n=12 cells from 6 mice) is blocked by a 5-HT2CR antagonist (n=5 cells from 3 mice). (h) mCPP selectively depolarizes non-VTA projecting CRFBNST neurons (n=3 cells from 2 mice non VTA-projecting CRF, n=5 cells from 4 mice VTA-projecting CRF). Data are mean ± s.e.m. P<0.05. Extended Data Figure 5 5-HT activates inhibitory microcircuits in the BNST that modulate outputs to the LH. (a) Recording configuration in CRF reporter mice infused with retrograde tracer beads in the LH. (b) Average traces of 5-HT induced depolarization in LH projecting vs non-projecting neurons (c) Histograms showing 5-HT induced depolarization in non-LH projecting BNST neurons (t4=4.425, p<0.05, One-sample t-test, n=5 cells from 3 mice) and hyperpolarization in LH-projecting neurons (t5=2.789, p<0.05, One-sample t-test, n=6 cells from 3 mice). (d) Confocal image of retrogradely CTB-labeled VTA (red) and LH (green) outputs in a CRF-L10a reporter (blue). (e-f) Pie charts depicting the percentage of LH-projecting only, VTA-projecting only, collateralizing, and CTB-negative (unlabeled) CRF in neurons in the dorsal and ventral aspects of the BNST (n=6 hemispheres from 3 mice). (g) Experimental schematic depicting viral infusions into the BNST and retrograde tracer bead infusions into the LH of CrfCre::ChR2BNST mice. (h) Recording configuration in CrfCre::ChR2BNST mice with LH tracer beads (i) Representative trace of light evoked IPSCs in LH projecting neurons (n=7 cells from 4 mice) and blockade of this light evoked response by GABAzine (n=2 cells from 2 mice). (j) Recording configuration in VTA projecting neurons in the BNST of C57BL/6 mice. (k-l) 5-HT has no effect on miniature IPSC frequency or amplitude in BNST→VTA projecting neurons (n=7 from 4 mice). (m-n) 5-HT has no effect on sIPSC frequency or amplitude in the presence of the 5-HT2CR antagonist RS102221 (n=5 cells from 4 mice). (o) Recording configuration in LH projecting neurons in the BNST of C57BL/6 mice (p) Representative traces showing an increase in sIPSC frequency in the presence of 5-HT for 6 cells from 3 mice (q-r) 5-HT increases sIPSC frequency but not amplitude in BNST→LH projecting neurons (F11,55=11.65, p<0.01, Repeated measures one-way ANOVA, n=6 cells from 3 mice). (s-t) 5-HT has no effect on miniature IPSC frequency or amplitude (n=5 cells from 3 mice). (u-v) 5-HT has no effect on sIPSC frequency or amplitude in the presence of RS102221 (n=6 cells from 4 mice). Data are mean ± s.e.m. P<0.05. Extended Data Figure 6 5-HT does not alter GABAergic transmission in CRF neurons nor does it directly excite non-CRF VTA projecting neurons in the BNST. (a) Recording configuration in CRFBNST neurons in a CRF reporter. (b-c) 5-HT has no effect on sIPSC frequency or amplitude in the total population of CRF neurons (n=5 cells from 3 mice). (d) Recording configuration in non-CRF, VTA projecting neurons in the BNST and average trace of 5-HT effect on membrane potential in non-CRF, VTA projecting neurons in the presence of TTX. (e) Histogram summarizing 5-HT effects on membrane potential in local and VTA projecting CRF neurons and local CRF neurons in the presence of the 5-HT2C receptor antagonist RS102221 (same data shown in Figure 2b) juxtaposed with the lack of effect of 5-HT on membrane potential in non-CRF, VTA projecting neurons (t4=0.9381, ns, One-sample t-test, n=5 cells from 3 mice). Data are mean ± s.e.m. P<0.01; *P<0.001. Extended Data Figure 7 The 5-HT2 agonist mCPP increases GABAergic but not glutamatergic transmission in the BNST. (a-b) mCPP increases sIPSC frequency (F15,30=1.863, p<0.001, Repeated measures one-way ANOVA, n=3 cells from 3 mice) but not amplitude in the BNST of C57BL/6 mice. (c-d) mCPP has no effect on sEPSC frequency or amplitude in the BNST of C57BL/6 mice (n=5 cells from 3 mice). Data are mean ± s.e.m. P<0.05. Extended Data Figure 8 Optogenetic and Intrsectional characterization of 5-HT-CRF circuits in the BNST and outputs to the midbrain (a) Experimental design and recording configuration from SertCre::ChR2DRN→BNST mouse with retrograde tracer beads in the VTA. (b) Representative traces for 5 cells from 3 mice depicting the increase in sIPSCs in VTA projecting neurons in the BNST following light-evoked 5-HT release (c) Histogram summarizing the effect of light evoked 5-HT release on sIPSC frequency in VTA projecting neurons (t4=4.890, p<0.01, One-sample t-test, n=5 cells from 3 mice). (d) Experimental configuration in CrfCre::Intrsect-ChR2BNST mice. (e) Representative images from 4 CrfCre::HSV-LSL1-mCherry-flpoVTA/LH mice and 4 CrfCre::HSV-LSL1-mCherryVTA/LH mice injected with Intrsect-ChR2-eYFP in the BNST. (f) Cell counts of eYFP+ neurons from HSV-LSL1-flpo and HSV-LSL1-mCherry injected CrfCre::Intrsect-ChR2BNST mice indicating the number of non-projecting CRF neurons compared to the total CRF population in the dorsal (top panel; t14=1.959, ns, Student’s unpaired two-tailed t-test, n=4 mice, 8 hemispheres per group) and ventral aspects of the BNST (bottom panel; t7=2.431, p<0.05, Student’s unpaired Welch’s corrected two-tailed t-test, n=4 mice, 8 hemispheres per group) (g) Recording configuration and light evoked IPSC showing local GABA release from non-projecting CRF neurons in the BNST. (h) Sterotaxic injection of ChR2 in Crfcre mouse (i-j) Light evoked IPSCs in the VTA and LH indicating that CRF projections to these regions are GABAergic. Data are mean ± s.e.m. P<0.05; P<0.01. Extended Data Figure 9 Pharmacological blockade of CRF1 receptors reduces fluoxetine induced aversive behavior and 5-HT enhancement of GABAergic transmission in the BNST. (a) Experimental schedule of injections and behavior. (b) CRF1R antagonist does not modify fear acquisition but reduces fluoxetine enhancement of cued fear recall (F1,20=13.70, p<0.01, Two-way ANOVA, n=6 per group). (c) Recording configuration in BNST neurons that project to the LH in C57BL/6 mice. (d) Bath application of a CRF1R antagonist blocks the 5-HT induced increase in sIPSC frequency in LH projecting neurons in the BNST (F10,30=0.2213, ns, Repeated measures one-way ANOVA, n=4 cells from 2 mice). (e) There was a reduction in sIPSC amplitude during 5-HT bath application and CRF1R blockade (F10,30=2.941, p<0.05, Repeated measures one-way ANOVA, n=4 cells from 2 mice). Data are mean ± s.e.m. P<0.01. Extended Data Figure 10 Model of a serotonin-sensitive inhibitory microcircuit in the BNST that modulates anxiety and aversive learning. Serotonin inputs to the BNST activate 5-HT2CRs expressed in non-projecting “local” CRF neurons. These “local” CRF neurons promote anxiety and fear by inhibiting anxiolytic outputs to the VTA and LH that are putatively GABAergic. Another discrete subset of CRF neurons, which are inhibited by 5-HT, send direct, inhibitory projections to the VTA and LH. These CRFBNST output neurons are GABAergic and putatively anxiolytic and stress buffering. Blue dashed lines indicate hypothesized additional synapses between CRFBNST neurons. Dashed red line indicates a putatively GABAergic synapse. Acknowledgements We acknowledge Bryan Roth for providing DREADD viral constructs and Sertcre mice, and Bradford Lowell for providing Crfcre mice. We also thank Alberto Lopez, Dan Perron, and Alexis Kendra for technical as sistance with stereotaxic surgeries on mice ,Bram Geenen for technical assistance with immunohistochemistry and Elyse Dankoski for technical assistance with the FSCV. This work was supported by NIH grants AA019454, AA011605 (T.L.K.), the Wellcome Trust (098012) and the Biotechnology and Biological Sciences Research Countil grant (BB/K001418/1) (L.K.H.) and by NIH grant K01AA023555 and the Alcohol Beverage Medical Research Fund (Z.A.M.). C.A.M. was supported by a postdoctoral NIAAA F32 fellowship (AA021319-02). C.M.M is supported by a predoctoral NIAAA F31 fellowship (F31AA023440). Figure 1 Optogenetic identification of a 5-HTDRN→BNST projection that elicits anxiety and fear-related behavior. (a) Experimental timeline for c-fos labeling of 5-HTDRN→BNST neurons following an aversive footshock stimulus. (b) Representative images of fluorogold (blue), tryptophan hydroxylase (violet), and c-fos (green) staining in the DRN for 13 mice. Scale bars: 100 µm. (c-f) Histograms depicting the number of double and triple labeled neurons in the DRN of naïve and shocked mice. (c) There were no significant differences in the number of BNST projecting 5-HTDRN neurons between groups. (d-f) Footshock lead to significant elevations in the number of c-fos+ (“activated”) 5-HT neurons (t11=2.975, p<0.05, Student’s unpaired two-tailed t-test, n=7 naïve and n=6 shock mice), c-fos+, fluorogold labeled neurons (t11=2.836, p<0.05, Student’s unpaired two-tailed t-test, n=7 naïve and n=6 shock mice), and triple labeled neurons (t11=2.374, p<0.05, Student’s unpaired two-tailed t-test, n=7 naïve and n=6 shock mice). (g) Experimental configuration for light-evoked FSCV experiments in SertCre::ChR2DRN→BNST mice (h) Coronal images showing ChR2-YFP expression in soma of the DRN and axons of the BNST. Scale bars: 500 µm. (i) Representative color plot of 5-HT release to optical stimulation (blue bar, 20 Hz 20 pulses) for 3 mice (j) Representative cyclic voltammogram at peak 5-HT (blue dashed line panel E) for 3 mice. (k) Representative Current vs. Time trace at baseline (black) and following 10 µM fluoxetine (red) for 3 mice. (l) Clearance half-life of 5-HT at baseline (white bar) and following 10 µM fluoxetine (red bar). (t2=8.43, p<0.05, Student’s paired two-tailed t-test, n = 3 slices from 3 mice) (m) SertCre mice were transduced in the DRN and implanted with bilateral optical fibers in the BNST. (n) Schematic of fear conditioning procedures in SertCre::ChR2DRN→BNST mice. (o-q) Photostimulation during fear acquisition had no effect on freezing behavior during fear learning but increased freezing during cued (t17=2.436, p<0.05, Student’s unpaired two-tailed t-test, n=10 control, n=9 ChR2) and contextual fear recall (t17=2.271, p<0.05, Student’s unpaired two-tailed t-test, n=10 control, n=9 ChR2). (r) Light delivery to the BNST reduced open arm time in the EPM (t15=2.79, p<0.05, Student’s unpaired two-tailed t-test, n=8 control, n=9 ChR2) and (s) increased latency to feed in the NSF (t17=2.19, p<0.05, Student’s unpaired two-tailed t-test, n=9 control, n=10 ChR2). (t) Action potentials generated by photostimulation in the DRN (5 Hz (top), 10 Hz (middle), 20 Hz (bottom), 473 nm). (u) Depolarization in cells (t8=5.20, p<0.01, One-sample t-test, n=9 cells from 4 mice) after photostimulation in the BNST (5 Hz, 10 s, 473 nm) and blockade of this response by 5 µM RS 102221 (t4=2.5, p>0.05, One-sample t-test, n=5 cells from 2 mice). Data are mean ± s.e.m. P<0.05; P<0.01; P<0.001. Figure 2 Serotonin activates a local population of CRFBNST neurons that inhibits outputs to the midbrain. (a) Recording scheme for CRF reporter mice injected with retrograde tracer beads in the VTA. (b) 5-HT depolarizes local CRF neurons (t5=7.06 , p<0.001, One-sample t-test, n=6 cells from 4 mice) in the BNST while hyperpolarizing CRFBNST→VTA neurons (t6=4.64, p<0.01, One-sample t-test, n=7 cells from 6 mice). Non VTA projecting CRF neurons are hyperpolarized by 5-HT in the presence of the 5-HT2CR antagonist RS102221 (t4=4.74, p<0.01, One-sample t-test, n=5 cells from 3 mice) (ci-ii) Schematic depicting infusions and recording configuration for CrfCre::ChR2BNST mice injected with retrograde tracer beads in the VTA. (ciii) Representative trace of light-evoked IPSC in beaded (i.e. VTA projecting), non-ChR2 expressing neurons in the BNST of CrfCre::ChR2 mice with retrograde tracer beads in the VTA (n=8 cells from 3 mice) and blockade of this response by GABAzine (F11,33=53.16, p<0.001, Repeated Measures One-way ANOVA, n=4 cells from 3 mice). (d) Recording scheme for C57BL/6 mice with retrograde tracer beads in the VTA or LH (e) Representative traces of sIPSCs in BNST neurons that project to the VTA before and after 5-HT application for 5 cells from 4 mice (f) Bar graphs showing magnitude of 5-HT effect on average sIPSC frequency in BNST neurons that project to the VTA (t4=3.257, p<0.05, One-sample t-test, n=5 cells from 4 mice) and in BNST neurons that project to the LH (t5=3.027, p<0.05, One-sample t-test, n=6 cells from 3 mice) and blockade of these responses by TTX and RS 102221. Effects on amplitude were non-significant. (g) Experimental scheme for experiments with CrfCre::Intrsect-ChR2BNSTmice. (h-i,) 5-HT significantly depolarizes non-projecting CRF (“Intrsect”) neurons in the BNST (t6=2.501, p < 0.05, One-sample t-test, n=7 cells from 5 mice) and produces a significant change in membrane potential in CRF Intrsect neurons compared to all CRF neurons (t26=2.08, p<0.05, Student’s unpaired two-tailed t-test, n=21 cells from 14 mice for experiments in all CRF neurons and n=7 cells from 5 mice for CrfCre::Intrsect-ChR2BNST experiments). Data are mean ± s.e.m. P<0.05; P<0.01; P<0.001. # donates P<0.05 for the Student’s unpaired two-tailed t-test between all CRF neurons and CRF Intrsect neurons in panel 2h. Figure 3 Acute fluoxetine elicits aversive behavior by engaging inhibitory CRF circuits in the BNST. (a) Schematic of recording for in vivo fluoxetine experiments in CRF reporter mice. (b) Representative traces of sIPSCs in VTA projecting neurons in the BNST for 5 experiments in 2 saline-treated mice and 7 experiments in 2 fluoxetine-treated mice. (c-d) Bar graphs showing that fluoxetine increases in sIPSC frequency (t10=2.55, p<0.05, Student’s unpaired two-tailed t-test, n=5 cells from 2 saline-treated mice, n=7 cells from 2 fluoxetine-treated mice), but not amplitude (t10=0.4752, p>0.05, Student’s unpaired two-tailed t-test, n=5 cells from 2 saline mice, n=7 cells from 2 fluoxetine mice) in VTA projecting neurons in the BNST. (e) Experimental configuration for assessment of anxiety in fluoxetine-treated CrfCre::hM4DiBNST mice and a coronal slice of the BNST expressing hM4Di-mCherry. Scale bar: 100 µm. (f) Confirmatory electrophysiology in the BNST showing hyperpolarization of hM4Di-mCherry-expressing cells following bath application of CNO (t5=4.32, p<0.01, One-sample t-test, n=6 cells from 4 mice) (g-h) Chemogenetic silencing of CRF neurons attenuates fluoxetine-induced anxiety like behavior on the elevated zero maze (F1,30=7.086, p<0.05, Two-way ANOVA, n=10 fluoxetine/hM4Di and n=8 for all other groups) without any concomitant locomotor effects. (i) Experimental configuration for fear conditioning experiments in CrfCre::hM4DiBNST mice. (j-k) Chemogenetic silencing of CRFBNST neurons had no effect on freezing behavior during fear learning but prevented fluoxetine enhancement of cued fear recall (F1,17=8.73, p<0.01, Two-way ANOVA, n=6 mCherry/vehicle and n=5 per group for all other groups). (l) Experimental configuration for assessment of the role of BNST outputs to the VTA and LH in fluoxetine-induced aversive behavior. (m) Confocal image of the BNST from HSVCre::hM3DqBNST mice. Scale bars: 500 µm.. (n-o) Chemogenetic activation of BNST neurons that project to the midbrain did not impact fear acquisition but attenuated fluoxetine induced enhancement of cued fear recall (F1,27=7.541, p<0.05, Two-way ANOVA, n=7 vehicle/hM3D and n=8 for all other groups). Data are mean ± s.e.m. P<0.05; P<0.01; *P<0.001. Author Contributions C.A.M., C.M.M., G.D., Z.A.M., L.K.H. and T.L.K. designed the experiments. A.H. and J.F.D performed triple label fos/tph/flg staining and image analysis. L.R.H. performed electrode placement surgeries and in vivo recordings during fear acquisition and recall. C.A.M. performed stereotaxic surgeries for evoked 5-HT electrophysiology and optogenetic behavior experiments. Z.A.M performed slice FSCV experiments and C.A.M performed evoked 5-HT electrophysiology experiments. C.A.M performed stereotaxic surgeries, behavior and data analysis for 5-HTDRN→BNST optogenetic experiments. C.A.M. performed all slice electrophysiology experiments and C.M.M and C.A.M. performed stereotaxic surgeries for these experiments (retrograde tracers, ChR2 infusions, and hM3D and hM4D infusions etc.) C.M.M, performed stereotaxic surgeries for chemogenetic manipulations in CRFBNST neurons that were used in fluoxetine fear conditioning experiments and C.A.M. performed behavior and data analysis. N.M. and J.F.D performed surgeries for chemogenetic manipulations in CRFBNST neurons that were used in fluoxetine anxiety (EZM) assays and N.B. and C.A.M. performed behavior and data analysis. C.M.M. and J.F.D. performed stereotaxic surgeries for HSVCre::hM3DBNST behavioral manipulations and C.A.M. performed behavior and data analysis. C.M.M. also performed imaging and analysis for optogenetic experiments, chemogenetic, and Intrsect experiments. C.R. and K.D. designed Intrsect viral contructs. G.D. performed surgeries, behavior and data analysis for htr2cCre::hM3DBNST experiments. C.A.M., C.M.M. and T.L.K wrote the manuscript with input from Z.A.M, L.H., J.F.D, J.A.H., G.D., T.T., A.H., L.K.H., T.K. The authors declare no financial conflict of interest. 1 Gorman J An open trial of fluoxetine in the treatment of panic attacks J Clin Psychopharmacol 1987 7 329 332 3500189 2 Westenberg HGM den Boer J Serotonin-influencing drugs in the treatment of panic disorder Psychopathology 1989 22 68 77 2498967 3 Burghardt NS Bush DEa McEwen BS LeDoux JE Acute selective serotonin reuptake inhibitors increase conditioned fear expression: blockade with a 5-HT(2C) receptor antagonist Biol Psychiatry 2007 62 1111 8 17524369 4 Ravinder S Burghardt NS Brodsky R Bauer EP Chattarji S A role for the extended amygdala in the fear-enhancing effects of acute selective serotonin reuptake inhibitor treatment Transl Psychiatry 2013 3 e209 23321806 5 Phelix C Liposits Z Paull W Serotonin-CRF interaction in the bed nucleus of the stria terminalis: a light microscopic double-label immunocytochemical analysis Brain Res Bull 1992 28 943 948 1638421 6 Liu Z Dorsal raphe neurons signal reward through 5-HT and glutamate Neuron 2014 81 1360 74 24656254 7 Burke LK Sex difference in physical activity, energy expenditure and obesity driven by a subpopulation of hypothalamic POMC neurons Mol Metab 2016 5 245 52 26977396 8 Armbruster BN Li X Pausch MH Herlitze S Roth BL Evolving the lock to fit the key to create a family of G protein-coupled receptors potently activated by an inert ligand Proc Natl Acad Sci U S A 2007 104 5163 8 17360345 9 Heisler LK Zhou L Bajwa P Hsu J Tecott LH Serotonin 5-HT(2C) receptors regulate anxiety-like behavior Genes Brain Behav 2007 6 491 6 17451451 10 Olive MF Koenig HN Nannini Ma Hodge CW Elevated extracellular CRF levels in the bed nucleus of the stria terminalis during ethanol withdrawal and reduction by subsequent ethanol intake Pharmacol Biochem Behav 2002 72 213 220 11900791 11 Huang MM Corticotropin-releasing factor (CRF) sensitization of ethanol withdrawal-induced anxiety-like behavior is brain site specific and mediated by CRF-1 receptors: relation to stress-induced sensitization J Pharmacol Exp Ther 2010 332 298 307 19843974 12 Jennings JH Distinct extended amygdala circuits for divergent motivational states Nature 2013 496 224 8 23515155 13 Dabrowska J Neuroanatomical evidence for reciprocal regulation of the corticotrophin-releasing factor and oxytocin systems in the hypothalamus and the bed nucleus of the stria terminalis of the rat: Implications for balancing stress and affect Psychoneuroendocrinology 2011 36 1312 26 21481539 14 Kim S-Y Diverging neural pathways assemble a behavioural state from separable features in anxiety Nature 2013 496 219 23 23515158 15 Fenno LE Targeting cells with single vectors using multiple-feature Boolean logic Nat Methods 2014 11 763 72 24908100 16 Dekeyne a Denorme B Monneyron S Millan M Citalopram reduces social interaction in rats by activation of serotonin (5-HT)2C receptors Neuropharmacology 2000 39 1114 1117 10727723 17 Belzung C Le Guisquet aM Barreau S Calatayud F An investigation of the mechanisms responsible for acute fluoxetine-induced anxiogenic-like effects in mice Behav Pharmacol 2001 12 151 162 11485052 18 Javelot H Efficacy of chronic antidepressant treatments in a new model of extreme anxiety in rats Depress Res Treat 2011 2011 531435 21808731 19 Liu J Acute administration of leptin produces anxiolytic-like effects: A comparison with fluoxetine Psychopharmacology (Berl) 2010 207 535 545 19823809 20 Mombereau C Gur TL Onksen J Blendy Ja Differential effects of acute and repeated citalopram in mouse models of anxiety and depression Int J Neuropsychopharmacol 2010 13 321 34 20003619 21 Krashes MJ An excitatory paraventricular nucleus to AgRP neuron circuit that drives hunger Nature 2014 507 238 42 24487620 22 Madisen L A robust and high-throughput Cre reporting and characterization system for the whole mouse brain Nat Neurosci 2010 13 133 140 20023653 23 Bath BD Subsecond adsorption and desorption of dopamine at carbon-fiber microelectrodes Anal Chem 2000 72 5994 6002 11140768 24 Hefner K Impaired fear extinction learning and cortico-amygdala circuit abnormalities in a common genetic mouse strain J Neurosci 2008 28 8074 8085 18685032
PMC005xxxxxx/PMC5124381.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 0355501 3206 Crit Care Med Crit. Care Med. Critical care medicine 0090-3493 1530-0293 19789445 5124381 10.1097/CCM.0b013e3181bc805d NIHMS153721 Article The Association Between Lymphotoxin Alpha (Tumor Necrosis Factor β) Intron Polymorphism and Predisposition to Severe Sepsis is Modified by Gender and Age Watanabe Eizo MD, PhD 1 Buchman Timothy G. MD, PhD, FCCM 12 Hirasawa Hiroyuki MD, PhD 3 Zehnbauer Barbara A. PhD, FACMG 4 1 Department of Surgery, Section of Acute Critical Care Surgery, Washington University School of Medicine 2 Center for Critical Care, Emory University 3 Department of Emergency and Critical Care Medicine, Graduate School of Medicine, Chiba University 4 Department of Pathology and Immunology, Washington University School of Medicine Corresponding Author: Timothy G. Buchman, Ph.D., M.D. Center for Critical Care, Emory University F524, 1364 Clifton Avenue, Atlanta GA, 30322 Telephone: 404-712-2609 Fax: (404) 712-2654 [email protected] 3 11 2009 1 2010 26 11 2016 38 1 181193 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Objective To investigate the significance of functional polymorphisms of inflammatory response genes by analysis of a large population of patients, both with and without severe sepsis, and representative of the diverse populations (geographic diversity, physician diversity, clinical treatment diversity) that would be encountered in critical care clinical practice. Design Collaborative case-control study conducted from July 2001 to December 2005. Setting A heterogeneous population of patients from 12 USA intensive care units (ICUs) represented by the Genetic Predisposition to Severe Sepsis (GenPSS) archive. Patients Eight hundred and fifty-four patients with severe sepsis and an equal number of mortality, age, gender, and race-matched patients also admitted to the ICU without evidence of any infection (matched nonseptic controls). Measurements and Main Results We developed assays for six functional single nucleotide polymorphisms (SNPs) present before the first codon of TNF at −308, IL1B at −511, IL6 at −174, IL10 at −819, and CD14 at −159, and in the first intron of LTA (also known as TNF-β) at +252 (LTA(+252)). The Project IMPACT™ critical care clinical database information management system developed by the Society of Critical Care Medicine and managed by Tri-Analytics, Inc. and Cerner Corporation was utilized. Template-directed dye-terminator incorporation assay with fluorescence polarization detection was used as a high-throughput genotyping strategy. Fifty-three percent of the patients were male with 87.3 % and 6.4 % of Caucasian and African American racial types, respectively. Overall mortality was 35.1 % in both severe sepsis (SS) and matched nonseptic control (MC) patients group. Average ages (SD) of the SS and MC patients were 63.0 (16.05) and 65.0 (15.58) years old, respectively. Among the 6 SNPs, LTA(+252) was most over-represented in the septic patient group (% severe sepsis; AA 45.6: AG 51.1: GG 56.7, P = .005). Moreover, the genetic risk effect was most pronounced in males, age > 60 yrs (P = .005). Conclusions LTA(+252) may influence predisposition to severe sepsis, a predisposition that is modulated by gender and age. Although the genetic influences can be overwhelmed by both comorbid factors and acute illness in individual cases, population studies suggest that this is an influential biological pathway modulating risk of critical illnesses. Genetic Predisposition to Disease Sepsis Lymphotoxin-alpha Gender Aging Sepsis remains a critical and costly public health problem. In the United States, more than 750,000 patients become septic each year, and more than 25% of these will die [1]. The annual total cost for care of sepsis in 1995 exceeded $16 billion [1] and patients suspected with severe sepsis are reported to account for 500,000 emergency department visits annually [2]. Recently, an evidence-based system for assessing quality of evidence of sepsis diagnosis and strength of recommendations for treatment [3, 4] has been used to develop an aggressive strategy for intervention for patients with this complication (Surviving Sepsis Campaign [www.survivingsepsis.com]). It is difficult, however, to identify prognostic factors that may predict response to the therapies implemented for severe sepsis. Advances in diagnosis, intervention, and prognostication of sepsis increasingly rely on tools that were recently confined to research laboratories much as molecular DNA typing, quantitative reverse transcriptase PCR and ELISA assays. Predisposition is complementary to diagnosis, intervention and prognostication. ‘Human Genetic Variation’ was selected as the ‘Breakthrough of the year 2007′ [5] and has been promoted as a strategy to identify patients predisposed to sepsis and other critical illnesses [6-11]. It is believed that the mortality of patients with severe sepsis correlates with markers of inflammatory responses [12-14] expressed as inflammatory cytokine levels in the blood. The inflammatory cytokine production of intensive care unit (ICU) patients with sepsis has been reported to be largely affected by interleukin-1 (IL-1) and tumor necrosis factor (TNF) related genetic polymorphisms in a single population study [10]. Assessing genetic predisposition is more challenging in population studies than the investigation of inheritance of Mendelian traits within specific kindreds. The challenges are related to the underlying composition of the populations in population association studies as well as the specific multigenic disease, trait or response under investigation. Sample sizes are often inadequate to achieve significant statistical power for predictive purposes and there is often reluctance to attribute seemingly significant results to chance [15, 16]. Further complicating the analyses is the fact that many genetic polymorphisms demonstrate ethnic differences in allele distributions [17]. This may be a factor in sepsis research because African American patients have been reported to have a higher population-based incidence of severe sepsis than either Caucasians or Hispanics, possibly attributable to differences in clinical, social, geographic, or access to healthcare services in the United States ICUs [18]. Gender of the patient with sepsis is also implicated as a prognostic factor for severe sepsis in both clinical [1, 19] and basic science studies [20]. Beneficial effects of estrogen through cytokine modulations of female patients with septic insults have been examined by experimental studies [20, 21]. Meanwhile, human [22, 23] and animal [24] studies have also demonstrated that the elderly experience more severe inflammatory responses and are more vulnerable to severe sepsis than the young due to dysregulation of cytokine storm. However, the biological basis of these differences is more complex than simply the presence or absence of sex hormones and may be modulated by genetic background and inflammatory gene expression [25, 26]. According to experts [15, 27, 28], issues in population genetics studies that complicate the correlation with clinical events in addition to inadequate statistical power include the heterogeneity of population subgroups, difficulty in acquiring matched control groups, genetic factors obscured by overwhelming disease pathophysiology, genetic variants that have undefined pathogenic consequences on gene function, and the low allele frequency of some polymorphic variants. Herein, we report on functional polymorphisms of six inflammatory response genes following analysis of a large and diverse population of patients both with and without severe sepsis from 12 medical centers in the USA. In this project, we used the Project IMPACT™ (PI) critical care clinical database information management system developed by the Society of Critical Care Medicine (SCCM) to facilitate the prospective collection of clinical samples for genetic analysis using a case-control design. Control patients were selected from those admitted to the same ICUs and having the same gender, age, race and outcome but absent severe sepsis. Simultaneously, we established a high-throughput, cost-effective genotyping of samples by introducing the template-directed dye-terminator incorporation assay with fluorescence polarization (TDI-FP) detection [29]. On the basis of the results of population association, we performed further assessment on the joint association between heritable and acquired factors with conventional tools and with recursive partitioning analysis. [Materials and Methods] Design and data source of ‘Genetic Predisposition to Severe Sepsis’ Project Critically ill patients from 12 intensive care units across the United States were studied as part of the Genetic Predisposition to Severe Sepsis (GenPSS) project. The study was designed to assess the significance of functional polymorphisms of inflammatory response genes by analysis of a large, diverse population of patients both with and without severe sepsis from geographically distinct medical centers with different clinicians and operating procedures. All participating clinical centers utilized the Project IMPACT (PI) database information management system developed by the Society for Critical Care Medicine (SCCM) and managed by Tri-Analytics, Inc. (http://www.trianalytics.com/, accessed 2Jan2009) and Cerner Corporation (Kansas City, MO). This database provided the separation of the patients’ identity and personal health information (PHI) from the central laboratory where genotyping was performed. PHI was scrubbed or masked from patient records by Tri-Analytics prior to delivery of clinical outcome data to the investigators. As the intended consequence of this study design (which also relied upon the use of waste blood, vide infra), the study was approved for conduct with waiver of written informed consent. The project design also meets the confidentiality and specimen access controls required of a typical NIGMS Human Genetic Cell Repository (http://ccr.coriell.org/Sections/Collections/NIGMS/?SsId=8 , accessed 2Jan2009). Patient Selection and Definitions The PI database was queried across the study period of July 2001 to December 2005 to identify specimens of blood submitted from 854 patients with unambiguous severe sepsis (SS) and an equal number of similar mortality, age (matched within a 10-yr margin), gender, and race matched controls (MC) without evidence of any infection during their ICU stay. The diagnosis of severe sepsis was made when a patient met the criteria proposed by the American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference [30]. We used a previously validated approach to the identification of severe sepsis involving the co-occurrence of International Classification of Diseases, Ninth Revision (ICD-9) codes for a bacterial or fungal infectious process and acute organ dysfunction using the following Boolean logic criteria (Fig. 1)[31]: Logic 1—Severe Sepsis at Admission/Physiology Within First 24 hrs: Criteria 1 AND 2 fulfilled Severe Sepsis at Admission 1a. Acute ICU admission diagnosis (ICD-9 codes); sepsis (septicemia-bacteremia) 038 OR Septic Shock 785.59 OR Shock-Septic 785.59 OR Bacteremia 790. (excludes anaphylaxis) OR 1b. APACHE 2 diagnostic criteria; Sepsis WITH Physiology, at least one of 2a. Systolic Blood Pressure; < 90mmHg as the lowest recorded systolic blood pressure in the first 24 hours OR use of vasopressor other than dopamine, such as phenylephrine; norepinephrine; and epinephrine. Dopamine is excluded because it continues to be used for purposes other than sepsis therapy. OR 2b. Acute renal failure as defined by APACHE 2 [32]; If a new rise in serum creatinine to 1.5 mg/dL or greater with oliguria (<135 mL urine any 8 hour period in the last 48 hours). The patient has to have had a recent period of oliguria. Discharge ICD coding was examined as a validity check to include ARF (584) with oliguria/anuria (788.5). OR 2c. Bilirubin > 2.0 mg/dL OR 2d. PaO2<90 torr for nonventilated patients receiving supplemental oxygen OR P/F ratio (PaO2/FIO2) < 225 for mechanical ventilated patients. Logic 2—Severe Sepsis developing after admission (beyond 24 hours) Project IMPACT™ coding as Severe Sepsis as a “complication not related to procedure”, with verified (a) Sepsis, severe (b) Sepsis-induced hypotension OR (c) Septic Shock. Selection of Matched Controls Controls were matched as follows: Participating centers submitted similar blood samples from critically ill patients who were not known to be septic, recognizing that some would become septic during the ICU stay. Most who were not septic on admission did not become septic during the ICU stay. All patients who did not meet severe sepsis criteria were placed into the matching pool, and those who had any evidence of infection were then excluded. This left a pool of potential non-sepsis match candidates who had no evidence of infection during their hospitalization as recorded in the PI database. Each septic patient was tentatively matched with all patients in the pool of the same gender, reported race/ethnicity and hospital discharge status (alive versus dead). Age matches were allowed +/− 10 years. Then, each severely septic patient was matched with one patient from its set of potential non-sepsis matches. If there was only one possible match for the septic patient, the match was assigned and then the match was removed from the pool. If more than a single matching patient was identified, then the match was made by selecting one of the potential matches at random, and then removing that match from the pool. The algorithm ran recursively until a match was made for each septic patient. In the event that a match corresponded to a sample that failed to yield DNA that could be analyzed, the algorithm was re-run against remaining matches in the control pool. Occasionally, patients would have more than one ICU admission during a single hospitalization. If the patient was severely septic on any ICU admission, the patient was judged severely septic for that hospitalization and therefore could not also be used as a matched control. Specimen Collection and Patient Confidentiality Fig. 2 illustrates our strategy for high-throughput, cost-effective genotyping of samples in this GenPSS project. Waste blood was collected from of ICU patients by participant ICU staff on blood spot cards (FTA®, Cat. No. WB120205, Whatman) as part of routine care. Typically, this waste blood was left in a syringe after sampling from an invasive arterial or central venous catheter. Cards were identified only by a bar code label. An identical bar code label was attached to the PI data sheet for each patient. The filter paper card was air-dried, inserted into a foil pouch (Cat. No. WB100037, Whatman), sealed in a cardboard mailing envelope (Cat. No. WB100016, Whatman), mailed to the genotyping laboratory via regular mail and received at a post office box for retrieval by lab personnel. Upon receipt at the GenPSS study center, the bar codes of the blood spot cards were scanned, entered into the local GenPSS database and then relabeled with a separate S/PAIR number. This S/PAIR re-coded blood spot card was forwarded to the genotyping lab. To promote confidentiality, genotyping lab personnel did not have information about the medical center, physician or patient from which the blood spot card originated. The PI personnel did not have access to genotyping determinations or the local GenPSS database at any time. The information management personnel queried the clinical PI database to identify specimens submitted from patients enrolled in GenPSS at participating ICUs without revealing any protected health information (PHI) to the genotyping lab personnel. Bar code identifiers were selected for PI records that corresponded to patients with severe sepsis and the matched control patients who did not have evidence of any infection at any time during their hospital course. DNA Isolation Using a 3 mm paper hole punch (McGill Inc., Marengo, IL), 2 FTA paper discs from the dried blood stain were punched out and placed in each tube of 8-strip thin wall tubes (MIDSCI, St. Louis, MO). DNA Elution Solution (PureGene, Qiagen (formerly Gentra Systems, Inc.) , MD) was added (200 μL per tube) and incubated overnight at 4°C. After incubation, the DNA elution solution was discarded. DNA purification solution (Gentra, Minneapolis, MN) 150 μL was added, incubated for 15 minutes at room temperature, mixed and discarded. This step was repeated 2 to 4 times until a colorless wash was obtained. DNA elution solution was added to the washed paper punches in each tube and incubated for 15 minutes. After removing as much solution as possible, DNA elution buffer was finally added to each tube. The sealed tubes were incubated for 15 minutes at 99°C within a MJ Research 225 Tetrad Thermal Cycler (GMI, Ramsey, MN) to facilitate release of the bound DNA from the paper discs. The eluted DNA was transferred to a fresh tube and stored at 4°C. Polymerase Chain Reaction We developed assays for 6 functional SNPs present in inflammation-related genes implicated as mediators of the sepsis cascade. These include SNPs present before the first codon of TNF at −308, IL1B at −511, IL6 at −174, IL10 at −819, and CD14 at −159 (TNF(−308)) [6-8], IL1B(−511) [33-35], IL6 (−174) [36, 37], IL10 (−819) [38, 39], and CD14 (−159) [40, 41]), and in the first intron of lymphotoxin-α (formerly known as TNF-β) at +252 (LTA(+252) [42, 43]. These 6 polymorphisms were selected from 22 candidate markers related to systemic inflammation based on our preliminary analysis of a separate pilot patient cohort. We focused on genes whose products had a known functional role in the inflammatory response [35, 39, 43-47]. Two microliters (μL) of genomic DNA (5 to 10 nanograms) was amplified in a mixture containing 2.5 μL of AmpliTaq Gold® PCR Master Mix (Applied Biosystems, Foster City, CA), 0.05 μL 50 mmol/L mixture of forward and reverse primer (Integrated DNA Technologies, Inc., Coralville, IA), 1.45 μL double-distilled water according to the conditions listed in Table 1a. To optimize the TDI-FP technique, PCR primers and conditions were designed to generate a homogeneous PCR product of 400 bp or smaller using Primer Premier ver. 5.0 software (Biosoft International, Palo Alto, CA). All reactions were performed in 0.2 mL black-skirted 96-well reaction plates (ABgene, Surrey, UK) with the MJ Research 225 Tetrad Thermal Cycler (BioRad, Hercules, CA). Amplicon Purification To eliminate unincorporated single-stranded PCR primers and dNTPs which may reduce the specificity of the TDI-FP reaction, E. coli exonuclease 1 (New England Biolabs, Inc., Ipswich, MA) (0.1 μL), shrimp alkaline phosphatase (SAP; Roche Molecular Biochemicals, Indianapolis, IN) (1.0 μL), 10× SAP reaction buffer (0.2 μL), pyrophosphatase (Perkin Elmer Life Sciences, Inc, Boston, MA) (0.15 μL), and double-distilled water (0.55 μL) were added to the total volume of PCR product (6.0 μL) and incubated at 37°C for 60 minutes, followed by 85°C for 25 minutes for enzyme inactivation. Pyrophosphatase was omitted from the purification of the PCR products of TNF(−308) and LTA(+252). Template-directed Dye-terminator Incorporation Reaction (TDI) For all assays, we used the single-base extension SNP genotyping kit AcycloPrime™ FP SNP Detection from Perkin Elmer Life Sciences, Inc, Boston, MA. The purified PCR product was combined with 0.05 μL AcycloPol™ thermostabile polymerase, 2.0 μL of 10 × AcycloPrime™ Reaction Buffer, 1.0 μL of specific AcycloTerminator™ mix consisting of R110/TAMRA-labeled dideoxynucleotide pairs for each SNP variant, 1.0 μL of the site-specific TDI oligonucleotide probe, 10 μmol/L (Table 1a; Integrated DNA Technologies, Inc., Coralville, IA), and 8.95 μL double-distilled water. The final reaction volume was 21 μL. The reactions were incubated according to the conditions specified in Table 1b. The TDI oligonucleotide probes were 20 to 30 nucleotides designed to complement the sequence immediately adjacent to the SNP position. Annealing temperatures were selected to be approximately 10°C below the melting temperature of the oligonucleotide probe DNA sequence. Fluorescence Polarization Determination Fluorescence polarization (FP) values were directly measured using an Analyst AD fluorescence reader (Molecular Devices, Sunnyvale, CA) with excitation and emission wavelengths of 580 nm and 605 nm for R110, and 552 nm and 575 nm for TAMRA, respectively. For LTA(+252), IL6(−174), and CD14(−159), 0.1 μL of single-stranded DNA binding protein (Epicentre Biotechnologies, Madison, WI), 0.2 μL of 10× SAP reaction buffer, and 1.7 μL water were added to each TDI reaction product and incubated at 37°C for 60 minutes before FP determination. For the remainder of the SNPs analyzed, FP was determined immediately following the TDI reaction. Between reaction steps (i.e., PCR, amplicon purification, TDI reaction, and incubation with single-stranded DNA binding protein) reaction plates were thermally sealed with adhesive sealing sheets (Brinkmann Instruments, Inc., Westbury, NY) and pulse centrifuged to minimize evaporative loss and cross contamination. Allele Assignment Allele assignment was objective and used SNPscorer software (Perkin Elmer Life Science, Inc at www.snpscoring.com, Samples that remained ambiguous following the first analysis were re-analyzed using identical methodology or direct DNA sequencing. Samples were considered of indeterminate genotype if allele assignment could not be made after two repetitions of TDI-FP methodology and sequencing. To confirm the accuracy of our approach, sequence-verified controls for each genotype were incubated in every TDI-FP assay performed. As a validity check, five percent of the TDI-FP calls were verified by conventional DNA sequencing. Gene and Clinical Data Analysis In addition to standard strategies, we employed recursive partitioning (RP) as a tool to identify relevant genetic and non-genetic predictors. We used HelixTree™ software (Golden Helix Inc, Bozeman, MT) for RP analysis. RP can be most easily viewed as conditional feature finding. Once a split is made based upon one gene/clinical feature, then the subsequent analysis is conditional on the presence or absence of that gene or clinical feature. Examination of a particular clinical outcome suggests that the effect of any feature will be, at least in part, dependent on the presence/absence of other features. Statistical Analysis Hardy-Weinberg equilibrium for the population distribution of the variant alleles was determined according to the approach described by Guo and Thompson [48]. Allelic chi-squares were also examined for each SNP. Strict Bonferroni correction is considered to be overly conservative, therefore a Bayesian formula was applied to obtain 0.95 posterior probability of a correct influence of association to a particular gene. In preliminary analysis, we considered differences significant at P < .05. In order to improve the likelihood that observed differences were not the result of a Type 1 error, we set final significance at P < .01. Results Characteristics of Severely Septic and Matched Control Patients The characteristics of 854 severely septic (SS) and matched non-septic control (MC) patients are listed in Table 2. In both groups, 53.6 % were male, 87.2 % and 6.4 % were Caucasian and African American patients, respectively, and overall survival was 64.9 % (Fig. 3). In spite of their similar mortality, the MC patients had no evidence of infection at any time during their hospitalization. Average ages (SD) of SS and MC groups were 63.0 (16.05) and 65.0 (15.58) years old, respectively (Fig. 3) and the distribution of the age in each cohort was similar (F = 1.06, P = 0.389). The non-septic patients were selected as directed matches in terms of gender, age race and survival from the same Project IMPACT ™ database (Fig 1). By design, the cohort of control ICU patients are well-matched on these selected variables to the patients with severe sepsis. Genotyped Specimens are Representative of the Total Severe Sepsis Population Distributions of gender, age, race and outcome between the genotyped SS and the total SS were compared (Fig. 3). The mean age (SD) was 64.4 (16.12) years old and 52.3 % were men in total SS of PI database (n = 21265; Fig. 3). In this total population, 79.0 % and 15.0 % were Caucasian and African-American, respectively. Caucasians were slightly overrepresented in the genotyped population, most likely the result of the patient populations in the institutions that chose to participate in the study. The distribution of age between the total SS and the genotyped SS was similar (F = 1.01, P = .885). The survival rate in total SS population (60 %) was slightly lower than the genotyped SS population (64%). Overall, as illustrated in Fig. 3, genotyped specimens are fairly representative of the total severe sepsis population. Genotypic Distributions of the Six Inflammation Gene Polymorphisms and Frequency of Severe Sepsis in the Genotype Categories Genotype call rate of the six SNPs was 91.4 - 99.0 % although genotypic distributions in TNF (−308), LTA(+252) and IL1B(−511) diverged from Hardy-Weinberg equilibrium (HWE) in the studied subjects (P = .0013, < .0001 and .0065, respectively; Table 3). As expected given their collocation on chromosome 6, TNF(−308) and LTA(+252) were in linkage disequilibrium (LD correlation R = .4523, D’ = .7819, P < .0001). LTA(+252) was the marker among the 6 SNPs which was the most over-represented in the septic patient group with codominant (additive) model analysis and therefore significantly associated with the sepsis susceptibility in terms of dose effect of alleles (P = .005; Fig 4). IL6(−174), CD14(−159) and IL1B(−511) also maintained the allele dose effect although their splits trend did not reach statistical significance. As for TNF(−308), the genotype distribution of variant homozygotes (AA) was too low (3.59 %; Table 3) to evaluate the frequency of severe sepsis occurrence. It appears that, in this large and geographically diverse population of the critically ill patients, LTA(+252) genotype influences the risk of developing severe sepsis. Lymphotoxin-α intron Polymorphism and Severe Sepsis after Gender, Age and Racial Stratification Table 5 shows the association of the LTA(+252) genotype and the other factors, i.e., age, gender, and race, on severe sepsis risk. In addition to the codominant model analysis, LTA(+252) variant was more frequent in the septic patient group with both recessive and dominant model analysis (P = .0044 and .0126, respectively). Severe sepsis risk did not segregate with the LTA(+252) among female ICU patients, but did it among male patients. Likewise, severe sepsis risk was associated with the LTA(+252) only among older (> 60 yrs) ICU patients. Furthermore, the statistical significance (P = .0016; Table 5) and the allele dose effect of the LTA(+252) on susceptibility to severe sepsis was maintained only among Caucasian patients. In each category (Male SS v Male MC, Female SS v Female MC, SS > 60 yrs v MC > 60 yrs, SS ≤ 60 yrs v MC ≤ 60 yrs, and Caucasian SS v Caucasian MC), the odds ratios of ‘AA v GG’ were 1.90, 1.22, 1.68, 1.36 and 1.55, respectively and were higher than the comparison of ‘AG v GG’ (1.40, 1.08, 1.18, 1.32, and 1.22, respectively), demonstrating reasonable trends in terms of LTA(+252) allele dose effect. Joint Association of the Lymphotoxin-α intron Polymorphism and Age and Gender on Susceptibility to Severe Sepsis Fig. 5 shows a tree diagram with the RP method indicating joint association of LTA(+252) and age and gender on susceptibility to severe sepsis. As shown in Fig. 5, in male population (P = .0026), LTA(+252) had significantly larger influence on severe sepsis risk than female population (P = .5870). In both gender groups, older (> 60 yrs) patients were more susceptible to severe sepsis, which was consistent with the results in Table 4. Although all populations exhibited the allele dose effect, there was statistical significance only in the older male population (P = .0051). Recursive partitioning indicated that the effect of LTA(+252) on severe sepsis risk was concentrated primarily in the older male members of the patient population. To verify the effects suggested by recursive partitioning, we formally tested the association between the sepsis case status and the marker a LTA(+252) adjusting for potential non-genetic covariates using a full versus reduced logistic regression approach based on an additive genotypic model. The full model consists of LTA(+252) allele counts along with the non-genetic covariates, and the reduced model consists only of the non-genetic covariates. This tests whether or not the allele count covariate adds significantly to the model. The non-genetic covariates considered for this test were age, pre-existing heart conditions, COPD, diabetes, malignant cancer, obesity, and whether the patient was admitted to the ICU due to trauma. The p-value of the full model was 4.53 × 10−22, and the p-value of the full versus reduced model test was 0.0016. Of the non-genetic covariates age, COPD, malignant cancer, and trauma were significant in the model with respective p-values of 0.00044, 0.00658, 0.00011, 2.42 × 10−11. When the same regression was run with only these significant covariates the p-value of the full model was 2.24× 10−22 and the p-value of the full versus reduced model was 0.001738. This indicates that there remains strong evidence that the LTA(+252) marker is associated with the sepsis response after accounting for these potential confounders. Discussion This study shows that in a large and heterogeneous population of critically ill patients who received medical care in diverse settings, a variant allele of the LTA gene is associated with increased risk of severe sepsis. This finding is important because prior studies have generally focused on care of a single ethnic group or care at a single medical center, and often report findings in relatively small study populations [49]. Moreover, this adverse effect of the variant allele of LTA(+252) on susceptibility to severe sepsis in this study is influenced by non-modifiable characteristics of gender and age (Fig. 5). The adverse effect of the LTA(+252)G on male patients is also reminiscent of the previous report by Schröder [19]. However, the surprising finding in the present study is that the influence of LTA(+252)G presence on the predisposition to severe sepsis is modified not only by gender (male) but also by age (elderly). Furthermore, the present study has more statistical power than the previous study plus the severe sepsis risk was evaluated with a completely matched control population among patients registered in the Project IMPACT™ (PI) database. This result suggests that inflammatory cytokine production is somewhat regulated by LTA(+252) in sepsis pathophysiology and the genetic risk association may be more pronounced among elderly male patients. Although the genotype distributions of LTA(+252) did not differ between young and elderly nor between male and female (Table 4), the largest genetic association appeared in older male patients. Interpretive caution is advised since a significant result in one subgroup and a lack of statistical significance in another does not necessarily signify a different effect according to subgroup. The interaction statistics between subgroups were not significant. Stewart, et al. and Marik, et al. have consistently shown that TNF production by the elderly was higher in severe biliary infections [23] and more generally in septic shock [22]. As for sexual dimorphism, Moxley, et al. demonstrated that LPS-stimulated TNF levels in peripheral blood of healthy female patients were lower than in blood of male patients [21]. Although the study design did not provide for measurements of cytokines, we speculate that their concentration might be significantly stratified by LTA(+252) especially in older male ICU patients. Furthermore, a recent report suggested that gene expression variance in critically ill patients due to age, gender, and ethnic background is greater than that due to infecting organism [50]. Our study failed to recruit a sufficient number of African-Americans with severe sepsis to reach reliable statistical inferences concerning any influence that race/ethnicity might have on sepsis predisposition phenotype among carriers of the various LTA(+252) alleles. Many of the contributing centers were community-based. Of note, there was no statistical difference in the frequency of African-Americans in those centers who did and who did not have genetic specimens contributed. Contrary to prior reports, many of the genetic polymorphisms besides LTA(+252) that had been already reported as predictors of severe sepsis did not demonstrate a statistically significant effect on sepsis risk. Thus variant allele carriers of CD14(−159) appeared somewhat less susceptible to severe sepsis in our result (Fig. 4), which is inconsistent with a published report [41] even if the influence of CD14(−159) in our result was not significant. This apparent discrepancy might be partially due to the larger size of our study or to differences in control selection. In our study, the US ICU patients from 12 the different centers who had no evidence of any infections were selected as control. In contrast, earlier studies that reported rare allele of CD14(−159) had adverse effect on sepsis risk chose healthy volunteers in a single center as their controls [40, 41, 51]. The early enthusiasm fueled by smaller studies that suggested the existence of a number of gene variants that might predispose to sepsis and/or an adverse outcome has been dampened by the realization that many of those studies were underpowered, their findings were likely falsely positive and their conclusions were unwarranted [49]. This concern applies to LTA (+252), which has previously been reported in smaller studies to both confer ( 5 studies) and not to infer (8 studies) an increased risk of sepsis [49]. One approach to minimizing the risk of falsely positive attributions is the calculation of a false positive report probability (FPRP) [52]. Even setting a conservative FPRP value of 0.2 still suggests that the observed influence of LTA(+252) on severe sepsis predisposition in our study to be noteworthy. A sufficiently powered, carefully described whole genome study in sepsis genomics appears important to understanding both sepsis biology as well as the interplay among genetic and non-genetic influences upon susceptibility and outcome. The importance of comprehensive clinical databases that collect not only demographic data and pre-existing conditions but also describe the septic event in detail—including all causative organisms, putative sources, treatments and apparent success or failure—cannot be overstated. Such process variables allow for rigorous disease-based grouping (e.g., analysis of patients with pneumococcal pneumonia) as opposed to the current stopgap strategy of syndrome-based grouping (e.g. severe vs. mild sepsis). Comprehensive clinical data contained in PI and similar databases enable not only close matching of cases and controls, but also some insight into selection bias that might affect the cases themselves. The specimens from patients with severe sepsis in the GenPSS archive appeared to be representative of the total severe sepsis population from the PI database (Fig. 3). Yet there was a difference in survival rate between ‘genotyped SS v total SS patients’ (64.87 v 60.72 %, P = .0102; Fig. 3) that might reflect differences in the approach of participating centers to the care of severely septic patients. Although the genotype failure rate approached 9% for some alleles, this appeared to be unrelated to the contributing center or to patient status (severe sepsis versus matched controls). Provided that the failure was unrelated to a specific allele, modeling suggests that the no additional SNPs would have become significant if genotype failures had not occurred Alternatively, there may be more complex genetic effects. In our ICU patients’ cohort, there are deviations from Hardy-Weinberg equilibrium (HWE) on LTA(+252). The Fisher’s Exact HWE test p-value is 1.21 × 10−5 and the HWE correlation R value is 0.10696. When cases and controls are considered separately, the Fisher’s exact HWE p-value is 0.00226 for cases and 0.00346 for controls for the LTA(+252) marker. Thus there is evidence to suggest that the association between the septic shock response and this marker is not driven by departures from HWE. It is important to note that the samples were not randomly drawn from a population, but instead were taken from patients admitted to the ICU. The LTA(+252) marker has been shown in previous studies to be associated with both sepsis and respiratory failure [53]. The likelihood of cases and controls suffering from either sepsis, respiratory failure or both is greater than would normally be found in the general “healthy” population, plausibly accounting for the departures from HWE. Plausible is different from certain, thus one can only conclude that for some genotypes evaluated, the alleles were not assorted randomly in the population under study. We believe that the intersection of clinical care and disease association genetics demands study of large (>100,000) populations whose clinical data are rigorously recorded and whose genomes are studied in their entirely. The costs of such an effort, previously prohibitive, are now mitigated by the US federal requirements for portable electronic records and advances in the efficiency of screening techniques. The risks of such an effort, primarily related to the potential for discrimination should genetic data leak to payors and insurers, have been mitigated by passage of the Genetic Insurability Nondiscrimination Act (GINA) signed into US law on April 24, 2008. Our mid-scale population association study supports the hypothesis that genetic predisposition to severe sepsis exists and is further modulated by both genetic (gender) and non-genetic (age) factors. Variation in the LTA(+252) gene appears to explain, in part, the susceptibility to sepsis and the effect may be greater in older Caucasian men. [Acknowledgments] Maureen Stark, formerly of Tri-Analytics and presently with Cerner Corporation provided tireless support. We thank the technicians and summer students who developed, qualified and performed the assays: Beverly Gibson, Brooke Stroup,Curtis Wilson, Xiaomin (May) Wu, Syamal (Dave) Bhattacharya, Michelle Mergler and Derek Bogdan. We are pleased to acknowledge the dedication and participation of the following individuals and institutions: Jackie O’Brien, Robert W. Taylor, St. John’s Mercy Medical Center, St. Louis MO; Brenda Snyder, Ellie Blasco, Michael Schwarz, North Colorado Medical Center, Greeley CO; Leslie DeSouza, Lori-Ann Kozikowski, Tom Higgins, Baystate Medical Center, Springfield MA, Dee Dee Boss, Gerald Plost, St. John Medical Center, Tulsa OK; Angela Dickson, St. John Medical Center, Longview WA; Mary Katherine Blackburn, Daniel Trahan, Learnard Chabert Medical Center, Houma LA; Howard Corwin, Dartmouth Hitchcock Medical Center, Lebanon NH; Terri Conner, Seton Medical Canter, Austin TX; Sandralee Blosser, Penn State Hershey Medical Center, Hershey PA; Christopher Dunatov, Milt L. McPherson, Northeast Medical Center, Concord NC; Sandy Hartenstein, Eastern Idaho Regional Medical Canter, Idaho Falls ID; Dianne Gergely, Devendra Amin, Morton Plant Hospital, Clearwater FL; Richard Riker, Maine Medical Center, Portland ME; Joanne Kuszat, Amy Fraccola, Rex Healthcare, Raleigh NC. An independent statistical opinion was provided by Christophe Lambert. Source of funding: This research was supported by NIH/GM (062809) and the Uehara Memorial Foundation. The funding sources had no role in study design, collection, analysis, or interpretation of data. Appendix: Report of power calculations Our study suggested that three SNPs maintained an allele dose effect although the splits trend did not reach statistical significance. In order to help the reader interpret the data and this statement, we performed and provide power calculations that vary with posited genetic effect and also with minor allele frequency assuming 854 case-control pairs, a severe sepsis rate of 5% in the ICU population, and a two-tailed α = 0.01. The calculations were performed with Quanto, a freely downloadable application offered by the University of Southern California and available at http://hydra.usc.edu/gxe/. Three models are presented. The first model, the dominant model, assumes that either one or two copies of the minor allele is sufficient to predispose to severe sepsis. The second model, the log-additive or coDominant model, assumes that two copies of the minor allele have double the probability of predisposing to severe sepsis. The third model, the recessive model, requires two copies of the minor allele to predispose to severe sepsis. Since there is no way of knowing how the minor allele might have its effect in advance, all three models are considered. For each of the models, a range of minor allele frequencies ranging from 0.18 to 0.38 is illustrated. A minor allele frequency of 0.18 corresponds approximately to the lowest frequency observed among the six minor alleles, that of TNF(−308). The highest frequency shown corresponds to the frequency of the minor allele of LTA(+252). The power calculation also depends on the degree of genetic predisposition. This is shown in the table as the RG, which can be thought of the increased odds of becoming septic if the model conditions are fulfilled. These are shown from 1.25 to 2.5 for each model. Each combination of model, minor allele frequency, and odds ratio produces a unique power value. A value of 0.8 corresponds to 80% , meaning that the study has an 80% probability of detecting the effect if the effect exists. Again, we set the p-value at a conservative 0.01. These data are calculated for single genes. They do not account either for gene-gene interactions or for multiple comparisons. Moreover they are power calculations only and have no influence on the analysis of the data. The table shows that if two copies of the minor allele are required to predispose to severe sepsis(recessive model), then either the minor allele must be relatively frequent or else the predisposition must be substantial (e.g. doubling the risk). Power Minor Allele Frequency Effect (Odds Ratio) Recessive coDominant Dominant 0.18 1.25 0.0429 0.5009 0.352 1.5 0.1671 0.9858 0.9258 1.75 0.3868 0.9999 0.9986 2 0.6284 0.9999 0.9999 2.25 0.8136 0.9999 0.9999 2.5 0.9209 0.9999 0.9999 0.23 1.25 0.0673 0.592 0.3837 1.5 0.2914 0.9952 0.942 1.75 0.6172 0.9999 0.9991 2 0.8563 0.9999 0.9999 2.25 0.961 0.9999 0.9999 2.5 0.9919 0.9999 0.9999 0.28 1.25 0.0995 0.6571 0.392 1.5 0.4383 0.998 0.9437 1.75 0.8019 0.9999 0.9991 2 0.96 0.9999 0.9999 2.25 0.9949 0.9999 0.9999 2.5 0.9995 0.9999 0.9999 0.33 1.25 0.1391 0.7015 0.382 1.5 0.5838 0.999 0.9354 1.75 0.9124 0.9999 0.9987 2 0.9913 0.9999 0.9999 2.25 0.9995 0.9999 0.9999 2.5 0.9999 0.9999 0.9999 0.38 1.25 0.1842 0.7299 0.3579 1.5 0.7076 0.9993 0.9168 1.75 0.9651 0.9999 0.9975 2 0.9984 0.9999 0.9999 2.25 0.9999 0.9999 0.9999 2.5 0.9999 0.9999 0.9999 Figure 1 Flow chart of patients’ selection in the GenPSS project GenPSS, Genetic Predisposition to Severe Sepsis Figure 2 Workflow of genotyping in the GenPSS project Figure 2 illustrates our strategy for high-throughput, cost-effective genotyping of samples in this GenPSS project. ICUs sent waste blood samples collected on a filter paper card (a). On this card, randomly- generated Bar-coded numbers (in the red circle*) were prepared to protect the confidentiality of individual patient. DNA for genotyping was isolated from these blood stains. First, we amplified the target region of DNA by PCR with primers specific for the sequences of each of our 6 selected markers. After a clean-up step of the PCR products to remove unused primers and nucleotides, we added a unique oligonucleotide probe molecule for the sequence-specific, single nucleotide extension step using dideoxyterminator nucleotides (ddNTPs). This is the TDI-FP as the high-throughput method for SNP analysis. PCR, Exo SAP, TDI reaction procedures were performed using a tetrad thermal cycler (b). FP values were directly measured using an Analyst AD fluorescence reader (c). In the result view which appears in the excel sheet (d), small dots indicate FP values obtained for individual patients’ samples and identification of specific genotype clusters from the TDI reaction. GenPSS, Genetic Predisposition to Severe Sepsis; PCR, polymerase chain reaction; Exo SAP, E. coli exonuclease 1 and shrimp alkaline phosphatase; TDI, template-directed dye-terminator incorporation; ddNTP, dideoxynucleotide triphosphate; FP, fluorescence polarization; DB, database; SNP, single nucleotide polymorphism. Figure 3 Representativeness of total patients with severe sepsis in Project IMPACT ™ critical care clinical database Top:Gender, age race and outcome of patients with severe sepsis (SS; solid black bars) who were genotyped in the present study (n = 865). Gender, age race and outcome of matched controls without infection (MC; gray shadow bars) who were genotyped in the present study (n = 865). Bottom: Gender, age race and outcome of patients with severe sepsis who appeared in the Project IMPACT™ (PI) critical care clinical database (n = 21265). Y-axes for all graphs show the relative frequencies of patients in each group. The boxplots shown above age distributions on each group indicate the median values and the inter-quartile range (sides of boxes). SS, patients with severe sepsis; MC, mortality, age, gender, and race-matched controls without evidence of any infection; PI, the Project IMPACT™ critical care clinical database; Race #1, American Indian/Alaska native/Australian Aborigine; Race #2, Asian/Pacific islander; Race #3, African American/African European/Haitian; Race #4, Latin/Hispanic; Race #5, Caucasian. Figure 4 Frequency of severe sepsis in genotype categories of the six single nucleotide polymorphisms SNP, single nucleotide polymorphism; TNF(−308), SNP at position −308 nucleotides 5’ of the first exon of tumor necrosis factor-α; LTA(+252), SNP at position 252 site of lymphotoxin-α; IL1B(−511), SNP at position −511 nucleotides 5’ of the first exon of interleukin-1β; IL6(−174), SNP at position −174 nucleotides before the first exon of IL-6; IL10(−819), SNP at position −819 nucleotides before the first exon of IL-10, CD14(−159), SNP at position −159 nucleotides before the first exon of CD14. Y-axes for all graphs show the frequencies of severe sepsis in each genotype category. P values in SNP were evaluated with chi-square test on codominant model analysis. Figure 5 Frequency of severe sepsis in the lymphotoxin alpha first intron single nucleotide polymorphism after gender and age stratification. LTA(+252), single nucleotide polymorphism at position 252 site of the lymphotoxin-α gene, Y-axes for all graphs shows the frequencies of severe sepsis. P values in SNP were evaluated with chi-square test on codominant model analysis. Table 1a PCR primers, Product Size, and Conditions Gene symbol SNP loci dbSNP rs# clustered ID Primersa PCR conditionsb TNF TNF(−308) rs1800629 TTTCTGAAGCCCCTCCCAGTT (F) CCCAAGGTGAGCAGAGGGAGA (R) 95°C, 10 min; 44 cycles of (95°C, 30 sec, 60°C, 30 sec, 72°C, 45 sec), 72°C, 5 min LTA LTA(+252) rs909253 CGTGCTTCGTGCTTTGGACTA (F) CCCAAGGTGAGCAGAGGGAGA (R) 95°C, 10 min; 40 cycles of (95°C, 30 sec, 60°C, 30 sec, 72°C, 45 sec), 72°C, 5 min IL1B IL1B(−511) rs16944 TGGCATTGATCTGGTTCATC (F) GTTTAGGAATCTTCCCACTT (R) 95°C, 10 min; 40 cycles of (95°C, 30 sec, 60°C, 30 sec, 72°C, 45 sec), 72°C, 5 min IL6 IL6(−174) rs1800795 GCGCTAGCCTCAATGACGACC (F) ATCTTTGTTGGAGGGTGAGGG (R) 95°C, 10 min; 40 cycles of (95°C, 30 sec, 64°C, 30 sec, 72°C, 45 sec), 72°C, 5 min IL10 IL10(−819) rs1800871 TACAGTAGGGTGAGGAAACC (F) GGTAGTGCTCACCATGACCC (R) 95°C, 10 min; 40 cycles of (95°C, 30 sec, 62°C, 30 sec, 72°C, 45 sec), 72°C, 5 min CD14 CD14(−159) rs2569190 GCTTAGGCTCCCGAGTCAACA (F) TGTCATTCAGTTCCCTCCTC (R) 95°C, 10 min; 40 cycles of (95°C, 30 sec, 63°C, 30 sec, 72°C, 45 sec), 72°C, 5 min a Primers are listed in a 5′-3′ orientation (F, forward primer; R, reverse primer). b Following amplification, PCR products were stored at 4°. dbSNP rs# clustered IDs are identified in NCBI, National Center for Biotechnology Information. SNP, single nucleotide polymorphism; TNF(−308), SNP at position −308 nucleotides 5′ of the first exon of the tumor necrosis factor-α gene; LTA(+252), SNP at position 252 site of the first intron of the lymphotoxin-α gene; IL1B(−511), SNP at position −511 nucleotides 5′ of the first exon of the interleukin-1β gene; IL6(−174), SNP at position −174 nucleotides before the first exon of the IL-6 gene; IL10(−819), SNP at position −819 nucleotides before the first exon of the IL-10 gene; CD14(−159), SNP at position −159 nucleotides before the first exon of the CD14 gene. Table 1b TDI-FP Oligonucleotide Probes, Dye-Terminator Combinations, and Reaction Conditions Gene symbol SNP loci Base change TDI-FP probesa Dye terminator TDI-FP reaction conditionsb TNF TNF(−308) G → A GAGGCAATAGGTTTTGA GGGGCATG (F) G/A 95°C, 2 min; 30 cycles of (95°C, 15 sec, 62°C, 30 sec) LTA LTA(+252) A → G TGTCACACATTCTCTGTT TCTGCCATG (F) G/A 95°C, 2 min; 25 cycles of (95°C, 15 sec, 61°C, 30 sec) IL1B IL1B(−511) C → T GTCTCTACCTTGGGTGCT GTTCTCTGCCTC (R) G/A 95°C, 2 min; 20 cycles of (95°C, 15 sec, 61°C, 30 sec) IL6 IL6(−174) G → C GTGCAATGTGACGTCCTT TAGCAT (R) G/C 95°C, 2 min; 30 cycles of (95°C, 15 sec, 62°C, 30 sec) IL10 IL10(−819) C → T TGTACCCTTGTACAGGTG ATGTAA (F) C/T 95°C, 2 min; 15 cycles of (95°C, 15 sec, 55°C, 30 sec) CD14 CD14(−159) C → T AATGAAGGATGTTTCAGG GAGGGGG G/A 95°C, 2 min; 25 cycles of (95°C, 15 sec, 55°C, 30 sec) a Probes are listed in a 5′-3′ orientation (F, forward primer; R, reverse primer). b All reaction products were stored at 4°. TDI-FP, template-directed dye-terminator incorporation with fluorescence polarization detection; SNP, single nucleotide polymorphism; TNF(−308), SNP at position −308 nucleotides 5′ of the first exon of tumor necrosis factor-α; LTA(+252), SNP at position 252 site of the first intron of lymphotoxin-α; IL1B(−511), SNP at position −511 nucleotides 5′ of the first exon of interleukin-1β; IL6(−174), SNP at position −174 nucleotides before the first exon of IL-6; IL10(−819), SNP at position −819 nucleotides before the first exon of IL-10, CD14(−159), SNP at position −159 nucleotides before the first exon of CD14. Table 2 Patient Characteristics (Severe Sepsis and Matched Controls) Genotyped patients with severe sepsis; SS (n = 854) Matched control patients; MC (n = 854) P values SAPS2 Prob. of Survival (mean±s.d.) 57.0 ± 28.8 70.5 ± 28.1 < .0001a Septic Shock (%) 7.5 NA NA Pre-existing conditions (%) Cardiovascular diseases 94.0 89.2 .0003b Diabetes mellitus 75.8 65.2 < .0001b COPD 70.5 58.2 < .0001b Malignant neoplasm 46.5 32.2 < .0001b Trauma 2.2 9.4 < .0001b Obesity 41.0 33.3 .0009 b Acute organ dysfunctions and sequelae (%) Respiratory (Ventilated) 63.1 48.6 < .0001b Cardiovascular 13.1 4.3 < .0001b Renal 25.6 6.9 < .0001b Hematologic 6.3 1.5 < .0001b Neurologic 3.2 3.0 .8890b Hepatic 2.5 .6 .0016b Number of acute organ dysfunctions (mean ±s.e.m) 1.14 ± .03 .65 ± .02 < .0001a DIC 4.0 2.3 .0529b DVT 6.6 4.7 .0928b GI bleeding 1.6 .7 .0720b Treatments (%) Renal replacement therapy 14.2 4.0 < .0001b Vasopressin 7.5 1.2 < .0001b Blood product administrations 18.5 11.2 < .0001b Enteral feeding 46.8 18.7 < .0001b a , P value of unpaired t-test (two-tailed); b , P value with Fisher’s exact test. Table 3 Genotypic distributions of the six SNPs SNP locus Genotype; n (%) Genoty pe call rate Hardy- Weinberg equilibrium test P value All/Septics/ Matched Controls TNF(−308) GG GA AA 0.001 Total genotyped patients n = 1561 1123 (71.94 %) 382 (24.27 %) 56 (3.59 % ) .9139 0.453 <0.0001 LTA(+252) AA AG GG < 0.0001 Total genotyped patients n = 1690 730 (43.20 %) 697 (41.24 %) 263 (15.56 %) .9895 0.002 0.003 IL1B(−511) CC CT TT 0.006 Total genotyped patients n = 1571 683 (43.48 %) 668 (42.52 %) 220 (14.00 %) .9198 0.041 0.071 IL6(−174) GG GC CC 0.169 Total genotyped patients n = 1684 655 (38.90 %) 769 (45.66 %) 260 (15.44 %) .9859 0.768 0.109 IL10(−819) CC CT TT 0.141 Total genotyped patients n = 158 897 (56.59 %) 576 (36.34 %) 112 (7.07 % ) .9280 0.749 0.071 CD14(−159) CC CT TT 0.703 Total genotyped patients n = 1570 443 (28.22 %) 789 (50.25 %) 338 (21.53 %) .9192 0.791 0.385 SNP, single nucleotide polymorphism; TNF(−308), SNP at position −308 nucleotides 5′ of the first exon of tumor necrosis factor-α; LTA(+252), SNP at position 252 site of lymphotoxin-α; IL1B(−511), SNP at position −511 nucleotides 5′ of the first exon of interleukin-1β; IL6(−174), SNP at position −174 nucleotides before the first exon of IL-6; IL10(−819), SNP at position −819 nucleotides before the first exon of IL-10; CD14(−159), SNP at position −159 nucleotides before the first exon of CD14. Table 4 Chi-square test on allelic frequencies and genotype distributions with respect to the predisposition to severe sepsis in the lymphotoxin alpha first intron polymorphism Patients’ subgroups (n; genotyped patients’ numbers) Chi-square test on LTA(+252) Odds ratio for incidence of severe sepsis Allelic frequency Genotypic distribution Chi-square (df=1) (P value) Chi-square of recessive model (AA v AG+GG) (df=1) (P value) Chi-square of dominant model (AA+AG v GG) (df=1) (P value) AG v GG [95% CI] AA v GG [95% CI] Total SS (n = 838) v Total MC (n = 852) 11.60 ( .0007) 8.10 ( .0044) 6.22 ( .0126) 1.25 [ .94, 1.67] 1.56 [1.17, 2.07] Male SS (n= 450) v Male MC (n = 458) 13.28 ( .0003) 8.95 ( .0028) 7.38 ( .0066) 1.40 [ .95, 2.06] 1.90 [1.30, 2.79] Female SS (n = 388) v Fe male MC (n = 394) 1.15 ( .2843) .93 ( .3349) .49 ( .4860) 1.08 [ .71, 1.66] 1.22 [ .80, 1.87] SS > 60 yrs (n = 477) v MC > 60 yrs (n = 545) 10.66 ( .0011) 9.52 ( .0020) 3.79 ( .0516) 1.18 [ .82, 1.73] 1.68 [1.16, 2.45] SS ≤ 60 yrs (n = 361) v MC ≤ 60 yrs (n = 307) 1.73 ( .1879) .55 ( .4596) 1.99 ( .1588) 1.32 [ .89, 1.96] 1.36 [ .88, 2.11] Caucasian SS (n = 730) v Caucasian MC (n = 744) 9.96 ( .0016) 7.69 ( .0056) 4.62 ( .0315) 1.22 [ .89, 1.67] 1.55 [1.13, 2.11] Non-Caucasian SS (n = 108) v Non-Caucasian MC (n = 108) 1.60 ( .2062) .49 ( .4822) 1.67 ( .1969) 1.48 [ .73, 3.00] 1.56 [ .77, 3.18] LTA(+252), single nucleotide polymorphism at position 252 site of lymphotoxin-α; df, degree of freedom; CI, confidence interval; SS, patients with severe sepsis; MC, mortality, age, gender, and race-matched controls without evidence of any infection. P values are examined with Chi-square test. Table 5 Genotype distributions of the lymphotoxin alpha first intron polymorphism stratified by gender, age and race SNP loci Genotype; n (%) Genotyp e call rate Hardy- Weinberg equilibriu m test P value LTA(+252) AA AG GG Total genotyed patients (n = 1690) 730 (43.20 %) 697 (41.24 %) 263 (15.56 %) .9895 < .0001 Male (n= 908) 388 (42.73 %) 371 (40.86 %) 149 (16.41 %) .9913 .0002 Female (n = 782) 342 (43.73 %) 326 (41.69 %) 114 (14.58 %) .9874 .0130 > 60 yrs (n = 1022) 440 (43.05 %) 434 (42.47 %) 148 (14.48 %) .9894 .0162 ≤ 60 yrs (n = 668) 290 (43.41 %) 263 (39.37 %) 115 (17.22 %) .9896 < .0001 Caucasian (n = 1474) 649 (44.03 %) 612 (41.52 %) 213 (14.45 %) .9893 .0006 Non-Caucasian (n = 216) 81 (37.50 %) 85 (39.35 %) 50 (23.15 %) .9908 .0039 LTA(+252), single nucleotide polymorphism at position 252 site of lymphotoxin-α. The authors have not disclosed any potential conflicts of interest. Special Instructions: none [References] 1 Angus DC Linde-Zwirble WT Lidicker J Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care Crit Care Med 2001 29 1303 1310 11445675 2 Wang HE Shapiro NI Angus DC National estimates of severe sepsis in United States emergency departments Crit Care Med 2007 35 1928 1936 17581480 3 Dellinger RP Levy MM Carlet JM Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock: 2008 Intensive Care Med 2008 34 17 60 18058085 4 Dellinger RP Levy MM Carlet JM Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock: 2008 Crit Care Med 2008 36 296 327 18158437 5 Pennisi E Breakthrough of the year. Human genetic variation Science 2007 318 1842 1843 18096770 6 McGuire W Hill AV Allsopp CE Variation in the TNF-alpha promoter region associated with susceptibility to cerebral malaria Nature 1994 371 508 510 7935762 7 Stuber F Udalova IA Book M −308 tumor necrosis factor (TNF) polymorphism is not associated with survival in severe sepsis and is unrelated to lipopolysaccharide inducibility of the human TNF promoter J Inflamm 1995 46 42 50 8832971 8 Mira JP Cariou A Grall F Association of TNF2, a TNF-alpha promoter polymorphism, with septic shock susceptibility and mortality: a multicenter study JAMA 1999 282 561 568 10450718 9 Tabrizi AR Zehnbauer BA Freeman BD Genetic markers in sepsis J Am Coll Surg 2001 192 106 117 quiz 145-106 11192910 10 Watanabe E Hirasawa H Oda S Extremely high interleukin-6 blood levels and outcome in the critically ill are associated with tumor necrosis factor- and interleukin-1-related gene polymorphisms Crit Care Med 2005 33 89 97 discussion 242-243 15644653 11 Villar J Perez-Mendez L Flores C A CXCL2 polymorphism is associated with better outcomes in patients with severe sepsis Crit Care Med 2007 35 2292 2297 17944017 12 Pinsky MR Vincent JL Deviere J Serum cytokine levels in human septic shock. Relation to multiple-system organ failure and mortality Chest 1993 103 565 575 8432155 13 Remick DG Bolgos GR Siddiqui J Six at six: interleukin-6 measured 6 h after the initiation of sepsis predicts mortality over 3 days Shock 2002 17 463 467 12069181 14 Oda S Hirasawa H Shiga H Sequential measurement of IL-6 blood levels in patients with systemic inflammatory response syndrome (SIRS)/sepsis Cytokine 2005 29 169 175 15652449 15 Cardon LR Bell JI Association study designs for complex diseases Nat Rev Genet 2001 2 91 99 11253062 16 Colhoun HM McKeigue PM Smith G Davey Problems of reporting genetic associations with complex outcomes Lancet 2003 361 865 872 12642066 17 Zehnbauer B Population genetics in critical illness Crit Care Med 2005 33 242 243 15644683 18 Barnato AE Alexander S Linde-Zwirble WT Racial Variation in the Incidence, Care, and Outcomes of Severe Sepsis Am J Respir Crit Care Med 2007 19 Schroder J Kahlke V Book M Gender differences in sepsis: genetically determined? Shock 2000 14 307 310 discussion 310-303 11028548 20 Choudhry MA Bland KI Chaudry IH Trauma and immune response--effect of gender differences Injury 2007 38 1382 1391 18048037 21 Moxley G Posthuma D Carlson P Sexual dimorphism in innate immunity Arthritis Rheum 2002 46 250 258 11817599 22 Marik PE Zaloga GP The effect of aging on circulating levels of proinflammatory cytokines during septic shock. Norasept II Study Investigators J Am Geriatr Soc 2001 49 5 9 11207836 23 Stewart L Grifiss JM Jarvis GA Elderly patients have more severe biliary infections: influence of complement-killing and induction of TNFalpha production Surgery 2008 143 103 112 18154938 24 Turnbull IR Wlzorek JJ Osborne D Effects of age on mortality and antibiotic efficacy in cecal ligation and puncture Shock 2003 19 310 313 12688540 25 Suffredini AF Systemic inflammation and sexual dimorphism: more than meets the eye Crit Care Med 2007 35 1610 1612 17522533 26 Trentzsch H Stewart D De Maio A Genetic background conditions the effect of sex steroids on the inflammatory response during endotoxic shock Crit Care Med 2003 31 232 236 12545021 27 Risch NJ Searching for genetic determinants in the new millennium Nature 2000 405 847 856 10866211 28 Terwilliger JD Haghighi F Hiekkalinna TS A bias-ed assessment of the use of SNPs in human complex traits Curr Opin Genet Dev 2002 12 726 734 12433588 29 Chen X Levine L Kwok PY Fluorescence polarization in homogeneous nucleic acid analysis Genome Res 1999 9 492 498 10330129 30 American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference: definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis Crit Care Med 1992 20 864 874 1597042 31 Weycker D Akhras KS Edelsberg J Angus DC Oster G Long-term mortality and medical care charges in patients with severe sepsis Crit Care Med 2003 31 2316 23 14501962 32 Knaus WA Draper EA Wagner DP APACHE II: a severity of disease classification system Crit Care Med 1985 13 818 829 3928249 33 Watanabe E Hirasawa H Oda S Cytokine-related genotypic differences in peak interleukin-6 blood levels of patients with SIRS and septic complications J Trauma 2005 59 1181 1189 discussion 1189-1190 16385298 34 Ma P Chen D Pan J Genomic polymorphism within interleukin-1 family cytokines influences the outcome of septic patients Crit Care Med 2002 30 1046 1050 12006801 35 Wen AQ Wang J Feng K Effects of haplotypes in the interleukin 1beta promoter on lipopolysaccharide-induced interleukin 1beta expression Shock 2006 26 25 30 16783194 36 Schluter B Raufhake C Erren M Effect of the interleukin-6 promoter polymorphism (−174 G/C) on the incidence and outcome of sepsis Crit Care Med 2002 30 32 37 11902285 37 Sutherland AM Walley KR Manocha S The association of interleukin 6 haplotype clades with mortality in critically ill adults Arch Intern Med 2005 165 75 82 15642878 38 Lowe PR Galley HF Abdel-Fattah A Influence of interleukin-10 polymorphisms on interleukin-10 expression and survival in critically ill patients Crit Care Med 2003 31 34 38 12544990 39 Allen ML Hoschtitzky JA Peters MJ Interleukin-10 and its role in clinical immunoparalysis following pediatric cardiac surgery Crit Care Med 2006 34 2658 2665 16932228 40 Hubacek JA Stuber F Frohlich D The common functional C(−159)T polymorphism within the promoter region of the lipopolysaccharide receptor CD14 is not associated with sepsis development or mortality Genes Immun 2000 1 405 407 11196689 41 Gibot S Cariou A Drouet L Association between a genomic polymorphism within the CD14 locus and septic shock susceptibility and mortality rate Crit Care Med 2002 30 969 973 12006789 42 Majetschak M Flohe S Obertacke U Relation of a TNF gene polymorphism to severe sepsis in trauma patients Ann Surg 1999 230 207 214 10450735 43 Stuber F Petersen M Bokelmann F A genomic polymorphism within the tumor necrosis factor locus influences plasma tumor necrosis factor-alpha concentrations and outcome of patients with severe sepsis Crit Care Med 1996 24 381 384 8625623 44 Wilson AG Symons JA McDowell TL Effects of a polymorphism in the human tumor necrosis factor alpha promoter on transcriptional activation Proc Natl Acad Sci U S A 1997 94 3195 3199 9096369 45 Fishman D Faulds G Jeffery R The effect of novel polymorphisms in the interleukin-6 (IL-6) gene on IL-6 transcription and plasma IL-6 levels, and an association with systemic-onset juvenile chronic arthritis J Clin Invest 1998 102 1369 1376 9769329 46 Terry CF Loukaci V Green FR Cooperative influence of genetic polymorphisms on interleukin 6 transcriptional regulation J Biol Chem 2000 275 18138 18144 10747905 47 Lin J Yao YM Yu Y Effects of CD14-159 C/T polymorphism on CD14 expression and the balance between proinflammatory and anti-inflammatory cytokines in whole blood culture Shock 2007 28 148 153 17515856 48 Guo SW Thompson EA Performing the exact test of Hardy-Weinberg proportion for multiple alleles Biometrics 1992 48 361 372 1637966 49 Clark MF Baudouin SV A systematic review of the quality of genetic association studies in human sepsis Intensive Care Med 2006 32 1706 1712 16957907 50 McDunn JE Husain KD Polpitiya AD Plasticity of the systemic inflammatory response to acute infection during critical illness: development of the riboleukogram PLoS ONE 2008 3 e1564 18270561 51 Agnese DM Calvano JE Hahm SJ Human toll-like receptor 4 mutations but not CD14 polymorphisms are associated with an increased risk of gram-negative infections J Infect Dis 2002 186 1522 1525 12404174 52 Wacholder S Chanock S Garcia-Closas M El Ghormli L Rothman N Assessing the probability that a positive report is false: an approach for molecular epidemiology studies J Natl Cancer Inst 2004 96 434 42 15026468 53 Waterer GW Quasney MW Cantor RM Septic shock and respiratory failure in community-acquired pneumonia have different TNF polymorphism associations Am J Respir Crit Care Med 2001 163 1599 1604 11401880
PMC005xxxxxx/PMC5124390.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 0372762 3389 Dev Biol Dev. Biol. Developmental biology 0012-1606 1095-564X 27729213 5124390 10.1016/j.ydbio.2016.10.006 NIHMS823477 Article AMP-activated protein kinase has diet-dependent and - independent roles in Drosophila oogenesis Laws Kaitlin M. a1 Drummond-Barbosa Daniela ab* a Department of Biochemistry and Molecular Biology, Johns Hopkins Bloomberg School of Public Health, 615 North Wolfe Street, Room W3118, Baltimore, MD 21205, USA b Department of Environmental Health Sciences, Johns Hopkins Bloomberg School of Public Health, 615 North Wolfe Street, Room W3118, Baltimore, MD 21205, USA * Corresponding author: Tel.: 410-614-5021; Fax: 410-955-2926, [email protected] (D.D.-B.) 1 Current address: Department of Neuroscience, 130 Clinical Research Building, 415 Curie Boulevard, University of Pennsylvania, Philadelphia, PA 19104, USA [email protected] (K.M.L.) 18 10 2016 10 10 2016 1 12 2016 01 12 2017 420 1 9099 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Multiple aspects of organismal physiology influence the number and activity of stem cells and their progeny, including nutritional status. Previous studies demonstrated that Drosophila germline stem cells (GSCs), follicle stem cells (FSCs), and their progeny sense and respond to diet via complex mechanisms involving many systemic and local signals. AMP-activated protein kinase, or AMPK, is a highly conserved regulator of energy homeostasis known to be activated under low cellular energy conditions; however, its role in the ovarian response to diet has not been investigated. Here, we describe nutrient-dependent and -independent requirements for AMPK in Drosophila oogenesis. We found that AMPK is cell autonomously required for the slow down in GSC and follicle cell proliferation that occurs on a poor diet. Similarly, AMPK activity is necessary in the germline for the degeneration of vitellogenic stages in response to nutrient deprivation. In contrast, AMPK activity is not required within the germline to modulate its growth. Instead, AMPK acts in follicle cells to negatively regulate their growth and proliferation, thereby indirectly limiting the size of the underlying germline cyst within developing follicles. Paradoxically, AMPK is required for GSC maintenance in well-fed flies (when AMPK activity is presumably at its lowest), suggesting potentially important roles for basal AMPK activity in specific cell types. Finally, we identified a nutrient-independent, developmental role for AMPK in cyst encapsulation by follicle cells. These results uncover specific AMPK requirements in multiple cell types in the ovary and suggest that AMPK can function outside of its canonical nutrient-sensing role in specific developmental contexts. Graphical abstract diet AMPK LKB1 germline stem cell follicle growth oogenesis Drosophila INTRODUCTION Adult stem cell lineages are sensitive to many physiological cues, including nutritional status. In species ranging from model organisms to humans, diet influences circulating factors, including nutrients, metabolites, and hormones, that can impinge on local signaling networks or act on stem cells directly to modulate their behavior (Ables et al., 2012). Although nutrient-sensing pathways are typically ubiquitously expressed, they can exert distinct effects on different cell types within stem cell lineages (Ables et al., 2012). A complete understanding of how nutrient sensors modulate stem cells and their descendants therefore requires dissecting how their roles in specific cell types contribute to the regulation of each lineage as a whole. In the Drosophila melanogaster ovary, stem cell populations continuously maintain oogenesis and have a well-described response to diet (Ables et al., 2012). Each ovary comprises a set of ovarioles, which are chronologically ordered arrays of developing follicles (or egg chambers) (Fig. 1A). Each follicle consists of a 16-cell germline cyst encapsulated by follicle cells that arise from germline stem cells (GSCs) and follicle stem cells (FSCs), respectively, within an anterior structure called the germarium (Fig. 1B). On a nutrient-rich diet, GSCs, FSCs, and their progeny divide and grow robustly, and oogenesis proceeds with minimal cell death. When females are shifted to a nutrient-poor diet, proliferation and growth of stem cells and their descendants slows uniformly (Drummond-Barbosa and Spradling, 2001), and GSC numbers decrease (Hsu and Drummond-Barbosa, 2009). In addition, early dividing cysts die frequently and most follicles degenerate instead of progressing through vitellogenesis (Drummond-Barbosa and Spradling, 2001). Several nutrient-dependent pathways play an active role within the GSC and/or FSC lineages to maintain robust rates of oogenesis on a rich diet. For example, germline insulin signaling is required for GSC division, cyst growth and follicle progression through vitellogenesis (Hsu et al., 2008; LaFever and Drummond-Barbosa, 2005). Target of rapamycin (TOR) signaling maintains GSC numbers and promotes GSC proliferation cell autonomously, and stimulates follicle growth by acting in both the germline cyst and in surrounding follicle cells (LaFever et al., 2010). The steroid hormone ecdysone acts directly on the germline to promote GSC maintenance and proliferation, survival of early dividing cysts (Ables and Drummond-Barbosa, 2010), and follicle growth and vitellogenesis (Ables et al., 2015; Buszczak et al., 1999; Carney and Bender, 2000). The role of pathways involved in actively sensing and responding to nutrient deprivation, however, remains less well understood. AMP-activated protein kinase (AMPK) is a heterotrimeric complex that is activated in response to low energy levels to control a number of cellular processes, including metabolism, protein homeostasis, and the cell cycle. The complex consists of a catalytic subunit (α) and two regulatory subunits (β and γ), and a single gene encodes each subunit in Drosophila. Studies in culture show that AMPK activity increases with high cellular levels of AMP or ADP (when ATP levels are low) and with activation by upstream kinases, including liver kinase B1 (LKB1). Activated AMPK stimulates catabolism and inhibits anabolic processes, thereby maintaining energy homeostasis (Hardie et al., 2016). Previous work in Drosophila showed that AMPK controls follicle cell growth (Haack et al., 2013); however, other potential roles of AMPK in oogenesis and its requirement in the germline have not been investigated. In this study, we uncovered both diet-dependent and -independent roles for AMPK in the Drosophila ovary using genetic mosaic analysis of available AMPKα alleles. We found that AMPK is required on a poor diet for slowing down GSC proliferation and follicle growth (through repression of follicle cell growth and proliferation). Surprisingly, AMPK is dispensable within the germline itself for follicle growth, in stark contrast to the germline requirement for insulin/TOR signaling in this process (LaFever et al., 2010). In addition, AMPK and LKB1 are intrinsically required for GSC maintenance in well-fed flies, suggesting that basal LKB1-dependent AMPK function can be important on a rich diet. Finally, we found that follicle cells intrinsically require AMPK to properly encapsulate germline cysts during follicle formation in a diet-independent fashion. This study underscores how widely acting nutrient-dependent pathways can have specific cellular requirements in multiple cell types within a tissue, some of which may be independent of their canonical role of tying cellular processes to the nutritional environment. MATERIALS AND METHODS Drosophila strains and culture conditions Fly stocks were maintained on standard cornmeal/molasses/yeast/agar medium at 22-25°C. For experiments, females (in the presence of wild-type males) were transferred daily onto either standard medium supplemented with wet yeast paste (“rich diet”) or molasses/agar (“poor diet”) (Armstrong et al., 2014). All AMPKα mutant stocks were maintained with a Y chromosome carrying a duplication including the AMPKα locus, Dp(1;Y)2E, y1 (DGRC #106089). The AMPKD2 FRT19A and FRT82B LKB1X5 null alleles were gifts from Jongkyeong Chung (Seoul National University) (Lee et al., 2007; Lee et al., 2006). The AMPKα1 null allele is described in Haack et al. (2013), and the AMPKα1 FRT19A recombinant chromosome was generated by standard crosses. AMPKaA is a lethal mutation described in Haelterman et al. (2014). The FRT82B LKB14A4-2 null allele was a gift from Daniel St. Johnston (University of Cambridge) (Martin and St Johnston, 2003). Other genetic elements are described in FlyBase (http://www.flybase.org). Genetic mosaic analysis Females of genotype y w His2Av::GFP hs-FLP FRT19A/AMPKα* FRT19A and hs-FLP; FRT82B Ubi-GFP/LKB1* were generated through standard crosses. (AMPKα* and LKB1* represent null or wild-type alleles of the AMPKα or LKB1 genes, respectively.) Zero- to 3-day-old females were maintained on dry yeast and heat shocked twice daily at 37°C for 3 days to induce mitotic recombination (Xu and Rubin, 1993). For GSC maintenance assays, flies were kept on a rich diet for 3 days after the final heat shock, then either maintained on a rich diet or shifted to a poor diet for an additional 4 days prior to dissection and processing. AMPKα* and LKB1* homozygous clones were recognized by the absence of green fluorescent protein (GFP), and GSCs were identified based on their anterior location and typical fusome morphology (de Cuevas and Spradling, 1998; Hsu et al., 2008). To quantify GSC loss, we analyzed all germaria containing GFP-negative cystoblasts and/or cysts, and calculated the percentage of germaria that no longer contained GFP-negative GSCs (i.e. “GSC loss events”), as described (Laws and Drummond-Barbosa, 2015). To measure GSC and follicle cell proliferation, flies were maintained on a rich diet for 4 days following the last heat shock, then either switched to a poor diet or maintained on a rich diet for an additional 3 days. The frequency of EdU-positive, GFP-negative GSCs or follicle cells was calculated as a percentage of the total number of GFP-negative GSCs or follicle cells, respectively, for multiple single plane images of follicle epithelia, as described (Laws and Drummond-Barbosa, 2015). To assess follicle growth, the size of follicles containing any number of GFP-negative follicle cells was compared to that of flanking follicles 7 days after the last heat-shock; a follicle was considered overgrown if larger than the follicle to its immediate posterior. Only follicles stage 10 or younger, and containing a single germline cyst were included in this analysis; the follicle cells analyzed are therefore part of follicle stem cell clones (Margolis and Spradling, 1995). This is a straightforward analysis because wild-type ovarioles always display progressively larger and more developed follicles from anterior to posterior, according to their chronological age; therefore, any significant deviation from this pattern is readily apparent in mosaic ovarioles. It is worth noting, however, that our analysis likely underestimates follicle overgrowth (see Fig. 3). Specifically, it is conceivable that in a significant number of ovarioles with mosaic AMPKα follicle cells, all follicles might have similar proportions of wild-type and mutant follicles cells (which are derived from wild-type and mutant follicle stem cells, respectively, in our experiments) relative to their neighbors and undergo similar overgrowth, retaining a normal pattern of relative follicle size. Follicle cell size analysis was performed in follicles in the mitotic program (stages 2-6) or later endoreplicative stages, and egg chambers were staged based on size, nuclear morphology, and yolk uptake (Spradling, 1993). The average size of follicle cells was determined by measuring the respective areas encompassing the same number of GFP-positive and GFP-negative cells in a single follicle cell monolayer, then dividing each area by the number of cells. Budding defects in region 3 were visually identified; those with excessive accumulation of follicle cells in region 3 were considered defective. Follicles were scored as misencapsulated if they contained two germline cysts within the same follicle cell monolayer. Immunofluorescence, EdU labeling, and microscopy Adult ovaries were dissected in Grace’s Insect Medium (Lonza), teased apart, and fixed for 13 minutes in 5.3% formaldehyde (Ted Pella) in Grace’s. Samples were rinsed and washed four times in 0.1% Triton X-100 (Sigma) in phosphate-buffered saline (PBS), or PBT, and blocked for at least 3 hours at room temperature or overnight at 4°C in 5% bovine serum albumin (BSA; Sigma) and 5% normal goat serum (NGS; Jackson ImmunoResearch) in PBT unless otherwise noted. Samples were incubated at 4°C overnight with primary antibodies in blocking solution at the following concentrations: mouse anti-Hts (1B1) (Developmental Studies Hybridoma Bank [DSHB], 1:10); mouse anti-Lamin C (LC28.26) (DSHB, 1:100); chicken anti-GFP (1:2000, Abcam); rabbit anti-pAMPK (1:200, Cell Signaling). After primary antibody incubation, samples were washed for 2 hours in PBT and incubated for 2 to 4 hours in Alexa Fluor 488-, 568-, or 633-conjugated goat species-specific secondary antibodies (1:200, Invitrogen) in blocking solution. Samples were mounted in Vectashield with DAPI (Vector Laboratories). Confocal images were acquired using a Zeiss LSM 700 microscope, and analyzed using either Zeiss ZEN 2009 or ImageJ software, and equally and minimally enhanced via histogram using Adobe Photoshop CS4. Although previous studies in Drosophila and human cell culture have used the aforementioned commercially available antibody against phosphorylated AMPKα (pAMPK) as a readout for AMPK activity (Castanieto et al., 2014; Lee et al., 2015; Vazquez-Martin et al., 2011), we detect a strong signal with this antibody in AMPKα mutant cells undergoing mitosis (Fig. S1), indicating that it is not a valid AMPK activity reporter in whole mount ovarian samples. EdU incorporation assays were performed as described (Ables and Drummond-Barbosa, 2013). Briefly, ovaries were dissected in Grace’s medium at room temperature, and incubated in 100 μM EdU (Invitrogen) in Grace’s medium for 1 hour prior to being teased apart, fixed, and stained as above. EdU was detected with AlexaFluor-594 via Click-It chemistry using the manufacturer’s instructions (Invitrogen) following secondary antibody incubation. RESULTS AMPK controls GSC proliferation in response to diet Downregulation of GSC division rates contributes to a reduction in egg production in response to a poor diet (Drummond-Barbosa and Spradling, 2001). Our previous work demonstrated an intrinsic requirement for insulin, TOR, and ecdysone signaling for robust GSC proliferation on a rich diet, consistent with their higher activity levels when nutrients are abundant (Ables and Drummond-Barbosa, 2010; Hsu et al., 2008; LaFever and Drummond-Barbosa, 2005; LaFever et al., 2010). It remains unknown, however, whether active repression of proliferation on a poor diet is also required. We therefore tested if AMPK, which becomes activated when resources are scarce (Towler and Hardie, 2007), is required for restricting GSC proliferation on a poor diet using genetic mosaic analysis of three independently generated AMPKα alleles: the lethal AMPKαA allele (Haelterman et al., 2014) and null AMPKαD2 and AMPKα1 alleles (Haack et al., 2013; Lee et al., 2007). We measured incorporation of the thymidine analog EdU, a marker for S phase, in control and AMPKα mutant mosaic ovarioles (Fig. 1C-E). The frequency of control mosaic GSCs in S phase decreased significantly when females were shifted from rich to poor diets (Fig. 1E). By contrast, the fraction of EdU-positive AMPKα mutant GSCs on either diet was statistically indistinguishable from that of control GSCs on a rich diet (Fig. 1E), showing that AMPKα mutant GSCs fail to downregulate proliferation on a poor diet. These results are consistent with a requirement for AMPK in actively repressing GSC proliferation in response to a poor diet. AMPK is required for GSC maintenance in well-fed flies on a rich diet In females on a poor diet, GSCs are lost more frequently from the niche, in part due to reduced TOR (LaFever et al., 2010; Sun et al., 2010) and ecdysone signaling in GSCs (Ables and Drummond-Barbosa, 2010), as well as to the non cell-autonomous effects of reduced insulin pathway activity (Hsu and Drummond-Barbosa, 2009, 2011; Yang et al., 2013) and amino acid levels (Armstrong et al., 2014). If AMPK activity were required to promote GSC loss on a poor diet, AMPKα mutant GSCs would be maintained better than controls on a poor diet, but similarly on a rich diet. To test these predictions, we compared the incidence of GSC loss events in control and AMPKα mutant mosaic germaria on both diets (Fig. 2A-C). In control mosaic females, where all cells are wild type, we detect GSC loss events in about 10% of germaria with a mosaic germline on a rich diet (Fig. 2C). Contrary to our prediction, GSCs homozygous mutant for AMPKα are lost approximately twice as frequently as controls on a rich diet, suggesting that the basal levels of AMPK activity present under those conditions are required for normal GSC maintenance. On a poor diet, AMPKα mutant GSCs also appear to be lost at higher frequencies relative to control GSCs, although these differences do not reach statistical significance (Fig. 2C). In addition, cleaved Dcp-1-positive GSCs are never detected in AMPKα or control mosaic germaria (see below), suggesting that mutant GSCs differentiate instead of undergoing apoptosis. In any case, the frequent loss of AMPKα mutant GSCs on a poor diet indicates that AMPK activity is not required for poor diet-induced GSC loss. Given that AMPK is required independently of low cellular energy conditions for GSC maintenance, it is possible that its activity on a rich diet is maintained by LKB1, a known upstream activator of AMPK (Hardie et al., 2016). Indeed, homozygous LKB1 mutant GSCs are lost from the niche significantly more frequently than control GSCs in well-fed germline mosaic females (Fig. S2). AMPK is not required in the germline for follicle growth Insulin/TOR signaling has a well-described role in the control of germline cyst growth both through an intrinsic germline requirement and non-autonomously through the regulation of follicle cells (LaFever et al., 2010). Because AMPK restricts cellular growth in times of nutrient deprivation (Yuan et al., 2013) and is a known negative regulator of TOR signaling (Hindupur et al., 2015), we asked if AMPK function is required to repress germline cyst growth on a poor diet. Surprisingly, we found that AMPK is dispensable in the germline for follicle growth. In both control and AMPK mutant mosaic ovarioles (Fig. 3A and B; n=>100 follicles with GFP-negative germline for each), all follicles carrying GFP-negative germline cysts develop at normal rates compared to flanking GFP-positive follicles. AMPK activity in follicle cells restricts germline cyst growth In addition to being intrinsically regulated, germline cyst growth is indirectly controlled by follicle cells. For example, both insulin/TOR signaling and the transcriptional factor Myc are required in follicle cells to regulate underlying germline cyst growth (LaFever et al., 2010; Maines et al., 2004). We therefore tested if AMPK activity in follicle cells might regulate follicle growth instead, by taking advantage of the linear and chronologically ordered arrangement of developing follicles of each ovariole that is characteristic of wild-type females. We analyzed ovarioles containing control versus AMPKα mutant mosaic follicle cell layers seven days after clone induction, and quantified the fraction of follicle cell mosaic ovarioles that contained one or more follicles with abnormal growth relative to its posterior (i.e. older) neighbor. In contrast to control mosaics, in which this is never observed, approximately 10% of ovarioles with AMPKα mutant mosaic follicle cells contain follicles that grow at a faster rate compared to the older, immediately posterior follicles on a rich diet (Fig. 3C and D). These overgrown follicles contain large patches of AMPKα mutant follicle cells, corresponding to ~50% or more of the follicle layer (>200 ovarioles analyzed for each genotype). These results suggest that basal AMPK levels are required in follicle cells for appropriate follicle growth in well-fed flies. On a poor diet, the frequency of follicle overgrowth triples (Fig. 3D), indicating that AMPK activity in follicle cells is particularly critical to further restrict follicle growth when nutrients are limiting. AMPK controls follicle cell proliferation and follicle cell size The requirement for AMPK in follicle cells to limit follicle growth could reflect a role of AMPK in restricting follicle cell proliferation, size, or both. In late stage follicles, where follicle cells undergo endoreplication (Spradling, 1993), we often observe markedly larger GFP-negative, AMPKα mutant follicle cells next to patches of normal sized GFP-positive, wild-type follicle cells in mosaic ovarioles (Fig. 4A-A’). We therefore measured the sizes of neighboring GFP-negative and -positive follicle cells in control and AMPKα mutant mosaic ovarioles on both rich and poor diets. As expected, in control mosaic ovarioles, GFP-negative and -positive follicle cells have similar sizes (Fig. 4B-C). By contrast, in AMPKα mutant mosaics GFP-negative follicle cells are significantly larger than neighboring wild-type, GFP-positive follicle cells regardless of diet or of whether they are proliferating (Fig. 4B) or endoreplicating (Fig. 4C). We conclude that AMPK restricts follicle cell growth on both rich and poor diets. Unlike the case for GSC maintenance, however, LKB1 function is not required for follicle cell growth regulation because LKB1 mutant follicle cells are comparable in size to neighboring wild-type follicle cells throughout mosaic ovarioles (Fig. S3). Since AMPK regulates follicle cell growth to similar extents on both rich and poor diets (Fig. 4B-C; compare differences between GFP-positive and GFP-negative follicle cells in AMPKα mosaics for each diet), this function of AMPK is not sufficient to explain the three-fold increase in oversized follicles with AMPKα mutant follicle cells on a poor diet (Fig. 3D). We therefore asked whether AMPK regulates follicle cell proliferation in a diet-dependent manner. In control mosaics, the frequency of EdU incorporation in GFP-negative follicle cells trends is lower on poor relative to rich diets (Fig. 5A-B). Unlike in controls, however, GFP-negative follicle cells in AMPKα mutant mosaics have similar frequencies of EdU incorporation on rich and poor diets (Fig. 5B), indicating that AMPK activity is required for follicle cells to downregulate proliferation on a poor diet. Taken together, these data support a model in which the regulation of follicle growth by AMPK is achieved through its intrinsic regulation of follicle cell growth and proliferation, which non-autonomously controls the growth of the underlying germline cyst. AMPK controls the vitellogenesis block under poor diet conditions Vitellogenesis is blocked in wild-type flies cultured on a poor diet due to the degeneration of early vitellogenic follicles. To determine if AMPK is required for this block in vitellogenesis, we compared the frequency of GFP-negative vitellogenic follicles in control versus AMPKα mosaic ovarioles. Approximately 40% of control mosaic ovarioles on a rich diet have vitellogenic follicles containing GFP-negative germline cysts, whereas this frequency drops to ~20% on a poor diet (Fig. 6A-B), consistent with reduced progression through vitellogenesis. By contrast, the fraction of AMPKα mutant mosaic ovarioles displaying vitellogenic follicles is not reduced on a poor diet compared to that on a rich diet (Fig. 6B). These results suggest that AMPK function contributes to the nutrient-sensitive vitellogenesis checkpoint. In addition to early vitellogenic follicle degeneration, switching females to a poor diet also induces the death of early germline cysts in the germarium (Drummond-Barbosa and Spradling, 2001). We therefore assessed whether AMPK was also required for early cyst death on a poor diet using cleaved Death caspase-1 (Dcp-1) as a marker (Florentin and Arama, 2012; Song et al., 1997) (Fig. 6C). We found, however, that Dcp-1-positive germline cysts were detected at similar frequencies in control and AMPKα mutant mosaic germaria on both rich and poor diets (Fig. 6D), indicating that the poor diet-induced increase in early cyst death is independent of AMPK activity. AMPK controls follicle cell development independently of diet Interestingly, we found that AMPK regulates follicle cell encapsulation of cysts in the germarium, a developmental process that is not diet-dependent (Drummond-Barbosa and Spradling, 2001). In control mosaics, germline cysts bud off of the germarium, and a monolayer of follicle cells encapsulates each 16-cell germline cyst (Fig. 7A). By contrast, AMPKα mutant follicle cells frequently fail to properly execute this budding event, and mutant follicle cells form sacs containing multiple germline cysts (Fig. 7B). At later stages, follicles containing multiple germline cysts are also observed (Fig. 7C). We can rule out that these phenotypes are an indirect consequence of the role of AMPK in limiting follicle cell size because mosaic Tsc1 mutant ovarioles, where follicle cell overgrowth also occurs, do not display these defects (LaFever et al., 2010). We quantified these phenotypes on both rich and poor diets, and found that although initial follicle budding defects in the germarium are observed more frequently on a poor diet (Fig. 7D), misencapsulated follicles occur at the same frequency on rich and poor diets (Fig. 7E). The similar number of mispackaged follicles on rich and poor diets indicates that these encapsulation events are not dependent on diet. It is possible that the global slowdown in oogenesis on a poor diet (Drummond-Barbosa and Spradling, 2001) leads to the observation of more follicles in the process of budding - and therefore to the increased visualization of any defects in this process. Therefore, we conclude that AMPK is developmentally required for follicle encapsulation independently of diet. DISCUSSION Based on its role as a nutrient sensor in the literature (Gowans and Hardie, 2014; Hardie, 2014; Hardie and Ashford, 2014; Hardie and Hawley, 2001; Hardie et al., 2012; Hardie et al., 2016), AMPK activity is expected to be low when flies are fed a rich diet and to increase in response to a poor diet. This study demonstrates that AMPK controls multiple diet-dependent steps during oogenesis, but also reveals unexpected roles in GSC maintenance and early follicle formation. AMPK is required in the germline to restrict GSC proliferation and vitellogenesis on a poor diet, and in follicle cells (but not in the germline) for follicle growth, consistent with the presumably higher levels of AMPK activity under low energy conditions. Surprisingly, AMPK is required for GSC maintenance on a rich diet, suggesting that basal levels of AMPK in the absence of apparent energy stress are required for optimal self-renewal. Finally, germline cyst encapsulation by follicle cells, which is a diet-independent process, also requires AMPK function, suggesting that developmental signals might regulate AMPK. A conserved role for AMPK in controlling germline and somatic cell proliferation in response to diet AMPK is required to downregulate GSC and follicle cell proliferation on a poor diet. Similarly, the Caenorhabditis elegans AMPKα homologs, aak-1 and aak-2, suppress germline proliferation during nutrient-dependent developmental arrests, and double mutants have hyperplastic germlines (Fukuyama et al., 2012; Narbonne and Roy, 2006). Failure to maintain germline quiescence during these nutrient-dependent arrests is catastrophic, leading to precocious entry into meiosis and sterility in surviving animals (Fukuyama et al., 2012; Narbonne and Roy, 2006). AMPK also appears to be important in the somatic gonad of several mammals to repress proliferation based on studies in culture. Both bovine and rat ovarian follicles cultured with AMPK activators have a reduction in granulosa cell proliferation (Kayampilly and Menon, 2009; Tosca et al., 2010), and inhibition of AMPK in cultured rat follicles leads to increased granulosa cell proliferation (Kayampilly and Menon, 2009). While further work is necessary to determine whether these observations hold true in vivo, these studies suggest that AMPK function in the somatic gonad is highly conserved. Unusual lack of germline requirement for AMPK in follicle growth regulation Insulin/TOR and Myc signaling are required in the germline itself for follicle growth in addition to also being required in follicle cells (LaFever et al., 2010; Maines et al., 2004). By contrast, AMPK activity restricts follicle cell growth and, non-autonomously, the growth of underlying germline cyst; however, it is surprisingly dispensable in the germline for cyst growth. This is the first example, to our knowledge, of a diet-dependent regulator that controls germline growth and development exclusively non-autonomously, via follicle cells. Interestingly, oocyte-specific deletion of AMPK in mice results only in mild fertility defects (Bertoldo et al., 2015), suggesting that the unusual lack of requirement for AMPK in the germline for follicle growth might be shared across species. TOR as a potential mediator of the diet-dependent role of AMPK in follicle cells Given the variety of phenotypes resulting from loss of AMPK in different types of germline and somatic cells, it is likely that distinct downstream targets are involved in each cell type. TOR is a cellular integrator of nutritional information that acts downstream of AMPK in many systems (Hardie et al., 2016); for instance, TSC2, an upstream inhibitor of TOR, contains a stimulatory AMPK phosphorylation site conserved from Drosophila to mammals (Kim and Lee, 2015). It is therefore conceivable that TOR signaling mediates a subset of AMPK effects in the Drosophila ovary. Previous studies demonstrated that either loss or overactivation of TOR signaling results in GSC loss (LaFever et al., 2010; Sun et al., 2010), raising the possibility that AMPKα mutant GSC loss is a consequence of hyperactive TOR signaling. Our data demonstrating that AMPK is not required in the germline for follicle growth, however, suggest that TOR regulation in later germline cysts is AMPK-independent. In genetic mosaic ovarioles, Tor mutant follicle cells are smaller than neighboring wild-type cells (LaFever et al., 2010), whereas AMPK mutant follicle cells are larger than control cells based on this study and on a previously published study using another AMPKα null allele, AMPKα3 (Haack et al., 2013). These data are consistent with a model in which AMPK acts as an upstream inhibitor of TOR signaling to control follicle cell size. It is unlikely, however, that disrupted TOR signaling is responsible for the follicle cell budding defect in AMPK mutant follicle cells, as these defects are not observed in TOR pathway mosaic ovaries (LaFever et al., 2010; Sun et al., 2010). Therefore, while TOR signaling may mediate the effects of AMPK in several ovarian processes, it is not the sole downstream effector of AMPKα in this system. Notch or hedgehog signaling may mediate the developmental functions of AMPK Our data reveal a diet-independent role for AMPK in the encapsulation of germline cysts by follicle cells during follicle formation, suggesting non-canonical upstream regulation of AMPK by developmental signals in this context. Other examples of energy-independent activation of AMPK have been previously described. For instance, energy-independent AMPK activation occurs in response to reactivate oxygen species (Mungai et al., 2011), and CAMKKβ can promote AMPK activation without elevated AMP (Hawley et al., 2005; Hurley et al., 2005; Woods et al., 2005). It will be interesting to investigate the specific signals regulating AMPK or acting downstream of it in the context of follicle formation. Hedgehog and Notch signaling have well-characterized roles in germline follicle budding (Forbes et al., 1996; Ruohola et al., 1991) and follicle cell specification (Chang et al., 2013; Nystul and Spradling, 2010), and are therefore logical candidates. A recent screen for Notch interactors in follicle cells uncovered interactions with multiple processes associated with metabolic state, including protein translation and degradation (Jia et al., 2015). Notch signaling is also regulated by autophagy, a process controlled by AMPK (Hardie et al., 2016). While it is clear that autophagy is upregulated in ovaries from nutrient-deprived flies, autophagy genes are also required for oogenesis under well-fed conditions (Barth et al., 2011). Follicle cell mutant clones of ATG1, a major autophagy-related gene, have fused egg chambers without stalk cells, reminiscent of both Notch and AMPK mutant phenotypes (Barth et al., 2012). Further, Notch controls the switch from mitosis to endoreplication in Drosophila (Deng et al., 2001; Lilly and Duronio, 2005; Lopez-Schier and St Johnston, 2001), and the large size of AMPK mutant follicle cells could reflect precocious entry into the endocycle. Future experiments should test a potential functional interaction between Notch signaling and AMPK activity in follicle cells. Supplementary Material ACKNOWLEDGEMENTS K.M.L. and D.D.-B. designed experiments, analyzed and interpreted data, and wrote the manuscript; K.M.L. performed all experiments. We thank the Developmental Studies Hybridoma Bank for antibodies and the Bloomington Stock Center (supported by National Institutes of Health P40 OD018537), D. St. Johnston, and J. Chung for Drosophila stocks. We are grateful to members of the Drummond-Barbosa lab for critical reading of the manuscript. This work was supported by National Institutes of Health R01 GM069875 (D.D.-B.). K.M.L. was supported by National Institutes of Health T32 CA009110 and the Elsa Orent Keiles Fellowship. Fig. 1 AMPK is required in GSCs for downregulation of proliferation on a poor diet (A) Diagram of Drosophila ovariole showing progressively more developed follicles, which bud off from an anterior germarium. Each follicle consists of a germline cyst (green) surrounded by somatic follicle cells (blue). Follicle cells undergo mitotic proliferation until stage 6 and transition to endoreplication at stage 7 of oogenesis. The oocyte begins yolk uptake, or vitellogenesis, during stage 8 (Spradling, 1993). (B) Diagram of the germarium, which houses germline stem cells (GSCs; dark green) juxtaposed to a somatic niche comprising cap cells (purple), a subset of escort cells (grey), and terminal filament cells (magenta). GSCs give rise to cystoblasts, which develop into 16-cell germline cysts that are encapsulated by follicle cells (blue) derived from a pair of follicle stem cells (dark blue) to form a new follicle. The fusome (red) is a special cellular structure present in early germ cells that becomes progressively more branched as cysts divide. (C and D) Maximum intensity projections of mosaic germaria showing GFP-negative GSCs (outlined) without (C) or with (D) EdU incorporation at 7 days after clone induction. GFP (green) labels wild-type cell nuclei; 1B1 (blue) labels fusomes and cell membranes; Lamin C (LamC; blue) labels cap cell nuclear envelopes; EdU (red) labels nuclei in S phase. Scale bar, 10 μm. (E) Average percentage of GFP-negative GSCs in control and AMPK mutant mosaic germaria that have incorporated EdU. These data combine three independent experiments, and sample sizes are indicated inside bars. Error bars represent S.E.M. *, p<0.01 by Chi-square test. Figure 2 AMPK is required intrinsically for GSC maintenance on a rich diet (A and B) Genetic mosaic germaria showing GFP-negative cystoblasts and cysts (dashed outline) derived from a GFP-negative GSC (solid outline) at 7 days after clone induction. The presence of GFP-negative germline cyst(s) in the absence of a GFP-negative GSC indicates a GSC loss event (B). GFP (green) labels wild-type cell nuclei; 1B1 (red) labels fusomes and cell membranes; Lamin C (LamC; red) labels cap cell nuclear envelopes. Scale bar, 10 μm. (C) Quantification of GSC loss events in control and AMPKα mutant mosaic germaria showing a significant increase in loss of AMPKα mutant GSCs at 7 days after clone induction on a rich diet. Differences on a poor diet do not reach statistical significance. Sample sizes from four independent experiments are indicated inside bars. Error bars represent S.E.M. p<0.05 by Student’s t test. Figure 3 AMPK function is required in follicle cells, but not in the germline, for follicle growth (A and B) Control (A) and AMPKα mutant mosaic (B) ovarioles with GFP-negative germline cysts (arrowheads) that grow normally relative to flanking GFP-positive cysts. (C) AMPKα mutant mosaic ovariole showing overgrowth of a follicle containing a wild-type germline cyst surrounded by AMPKα mutant follicle cells (open arrowhead) relative to the posterior, older follicle, which contains fewer GFP-negative follicle cells (arrow). GFP (green) labels wild-type cell nuclei; 1B1 (red) labels fusomes and cell membranes; Lamin C (LamC; red) labels cap cell nuclear envelopes. Scale bar, 10 μm. (D) Quantification of follicle overgrowth in follicle cell mosaic ovarioles at 7 days after clone induction. This phenotype is markedly enhanced on poor relative to rich diets. Sample sizes are shown inside bars and represent results from three independent experiments. Error bars represent S.E.M. p<0.05; p<0.01 by Student’s t test. Fig. 4 AMPK cell-autonomously controls follicle cell growth during mitotic and endoreplicative cycles (A) AMPKα mutant mosaic follicle cell layer showing larger mutant follicle cells surrounding wild type, GFP-positive follicle cells (outlined). GFP (green in A), labels wild-type cell nuclei; 1B1 (red) marks cell membranes; DAPI (blue) marks nuclei. GFP channel alone is show in grayscale in A’. Scale bar, 10 μm. (B and C) Quantification of GFP-positive and -negative follicle cell size at 7 days after clone induction in mitotic (B) or endoreplicative (C) follicle stages. Error bars represent S.E.M. p<0.05, p<0.01; p<0.001 by Student’s t test. Fig. 5 AMPK is required to restrict follicle cell proliferation on a poor diet (A) A genetic mosaic ovariole showing GFP-negative follicle cells with EdU incorporation. GFP (green) labels wild-type cell nuclei; 1B1 (blue) labels cell membranes; EdU (red) labels cells in S phase. GFP-negative follicle cell clones are outlined. Scale bar, 10 μm. (B) Percentage of control and AMPKα mosaic GFP-negative follicle cells in mitotic stages that incorporate EdU at 7 days after clone induction. Sample sizes from three independent combined experiments are indicated inside bars. Error bars represent S.E.M. , p<0.05 by Student’s t test. Fig. 6 AMPK is required for the inhibition of vitellogenesis, but not for early cyst death, in response to a poor diet (A) AMPKα mutant mosaic ovariole showing a vitellogenic GFP-negative germline cyst (arrowhead). GFP (green) labels wild-type cell nuclei; 1B1 (red) labels fusomes and cell membranes; DAPI (blue) labels nuclei. Scale bar, 50 μm. (B) Quantification of germline mosaic ovarioles containing vitellogenic GFP-negative germline cysts in control and AMPKα mosaic females. Error bars represent S.E.M. , p<0.05 by Student’s t test. (C) Maximum intensity projection of an AMPKα mosaic germarium showing a GFP-negative, cleaved Dcp-1-positive germline cyst (outlined). GFP (green) labels wild-type cell nuclei; 1B1 (blue) labels fusomes and cell membranes; Lamin C (LamC; blue) labels cap cell nuclear envelopes; cleaved Dcp-1 (red) is an apoptosis marker. Scale bar, 10 μm. (D) Percentage of germaria containing GFP-negative, cleaved Dcp-1-positive cystoblasts and/or cysts in control and AMPKα mosaics. We did not observe any cleaved Dcp-1 positive GSCs in our experiments. Sample sizes from three independent experiments are indicated inside bars. Error bars represent S.E.M. *p<0.001 by Chi-square test. Fig. 7 Follicle cell AMPK controls follicle encapsulation independently of diet (A) In control mosaic ovarioles, follicle cells envelop germline cysts to form follicles and also give rise to follicle cell stalks (arrow), which separate follicles for the remainder of oogenesis. (B) Example of AMPKα mutant follicle cells leading to abnormal follicle budding (arrowhead). (C) Example of AMPKα mutant mosaic follicle cells that have mispackaged multiple germline cysts with nurse cells of variable ploidy (asterisk). Given that the timing of our experiments is controlled such that we analyze exclusively stem cell clones, the presence of GFP-positive and GFP-negative nurse cells within the same follicle further indicates that distinct cysts (or parts thereof) were misencapsulated together. GFP (green) labels wild-type nuclei; 1B1 (red) labels fusomes and cell membranes; Lamin C (LamC; red) labels cap cell nuclear envelopes; DAPI (blue) labels nuclei. A single optical slice is shown in each panel; therefore, not all cells within the follicles are visible. Scale bar, 20 μm. (D and E) Graphs indicating the frequency of phenotypes shown in (B) and (C), respectively. Budding defects are more frequently observable on a poor diet (D); however, similar numbers of mispackaged follicles are generated regardless of diet (E) at 7 days after clone induction. Sample sizes represent data from four independent trials and are shown inside bars. Error bars, S.E.M. p<0.05, Student’s t test. HIGHLIGHTS AMPK restricts germline stem cell (GSC) proliferation on a poor diet AMPK inhibits vitellogenesis on a poor diet, but does not control early cyst death AMPK in follicle cells, but not in the germline, controls follicle growth Basal AMPK activity promotes GSC maintenance on a rich diet AMPK has a nutrient-independent, developmental role in follicle formation This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. REFERENCES Ables ET Bois KE Garcia CA Drummond-Barbosa D Ecdysone response gene E78 controls ovarian germline stem cell niche formation and follicle survival in Drosophila Dev Biol 2015 400 33 42 25624267 Ables ET Drummond-Barbosa D The steroid hormone ecdysone functions with intrinsic chromatin remodeling factors to control female germline stem cells in Drosophila Cell Stem Cell 2010 7 581 592 21040900 Ables ET Drummond-Barbosa D Cyclin E controls Drosophila female germline stem cell maintenance independently of its role in proliferation by modulating responsiveness to niche signals Development 2013 140 530 540 23293285 Ables ET Laws KM Drummond-Barbosa D Control of adult stem cells in vivo by a dynamic physiological environment: diet-dependent systemic factors in Drosophila and beyond Wiley Interdiscip Rev Dev Biol 2012 1 657 674 23799567 Armstrong AR Laws KM Drummond-Barbosa D Adipocyte amino acid sensing controls adult germline stem cell number via the amino acid response pathway and independently of Target of Rapamycin signaling in Drosophila Development 2014 141 4479 4488 25359724 Barth JM Hafen E Kohler K The lack of autophagy triggers precocious activation of Notch signaling during Drosophila oogenesis BMC Dev Biol 2012 12 35 23217079 Barth JM Szabad J Hafen E Kohler K Autophagy in Drosophila ovaries is induced by starvation and is required for oogenesis Cell Death Differ 2011 18 915 924 21151027 Bertoldo MJ Guibert E Faure M Rame C Foretz M Viollet B Dupont J Froment P Specific deletion of AMP-activated protein kinase (alpha1AMPK) in murine oocytes alters junctional protein expression and mitochondrial physiology PLoS One 2015 10 e0119680 25767884 Buszczak M Freeman MR Carlson JR Bender M Cooley L Segraves WA Ecdysone response genes govern egg chamber development during mid-oogenesis in Drosophila Development 1999 126 4581 4589 10498692 Carney GE Bender M The Drosophila ecdysone receptor (EcR) gene is required maternally for normal oogenesis Genetics 2000 154 1203 1211 10757764 Castanieto A Johnston MJ Nystul TG EGFR signaling promotes self-renewal through the establishment of cell polarity in Drosophila follicle stem cells Elife 2014 3 Chang YC Jang AC Lin CH Montell DJ Castor is required for Hedgehog-dependent cell-fate specification and follicle stem cell maintenance in Drosophila oogenesis Proc Natl Acad Sci U S A 2013 110 E1734 1742 23610413 de Cuevas M Spradling AC Morphogenesis of the Drosophila fusome and its implications for oocyte specification Development 1998 125 2781 2789 9655801 Deng WM Althauser C Ruohola-Baker H Notch-Delta signaling induces a transition from mitotic cell cycle to endocycle in Drosophila follicle cells Development 2001 128 4737 4746 11731454 Drummond-Barbosa D Spradling AC Stem cells and their progeny respond to nutritional changes during Drosophila oogenesis Dev Biol 2001 231 265 278 11180967 Florentin A Arama E Caspase levels and execution efficiencies determine the apoptotic potential of the cell J Cell Biol 2012 196 513 527 22351928 Forbes AJ Lin H Ingham PW Spradling AC hedgehog is required for the proliferation and specification of ovarian somatic cells prior to egg chamber formation in Drosophila Development 1996 122 1125 1135 8620839 Fukuyama M Sakuma K Park R Kasuga H Nagaya R Atsumi Y Shimomura Y Takahashi S Kajiho H Rougvie A Kontani K Katada T C. elegans AMPKs promote survival and arrest germline development during nutrient stress Biol Open 2012 1 929 936 23213370 Gowans GJ Hardie DG AMPK: a cellular energy sensor primarily regulated by AMP Biochem Soc Trans 2014 42 71 75 24450630 Haack T Bergstralh DT St Johnston D Damage to the Drosophila follicle cell epithelium produces “false clones” with apparent polarity phenotypes Biol Open 2013 2 1313 1320 24337115 Haelterman NA Jiang L Li Y Bayat V Sandoval H Ugur B Tan KL Zhang K Bei D Xiong B Charng WL Busby T Jawaid A David G Jaiswal M Venken KJ Yamamoto S Chen R Bellen HJ Large-scale identification of chemically induced mutations in Drosophila melanogaster Genome Res 2014 24 1707 1718 25258387 Hardie DG AMPK--sensing energy while talking to other signaling pathways Cell Metab 2014 20 939 952 25448702 Hardie DG Ashford ML AMPK: regulating energy balance at the cellular and whole body levels Physiology (Bethesda) 2014 29 99 107 24583766 Hardie DG Hawley SA AMP-activated protein kinase: the energy charge hypothesis revisited Bioessays 2001 23 1112 1119 11746230 Hardie DG Ross FA Hawley SA AMPK: a nutrient and energy sensor that maintains energy homeostasis Nat Rev Mol Cell Biol 2012 13 251 262 22436748 Hardie DG Schaffer BE Brunet A AMPK: An Energy-Sensing Pathway with Multiple Inputs and Outputs Trends Cell Biol 2016 26 190 201 26616193 Hawley SA Pan DA Mustard KJ Ross L Bain J Edelman AM Frenguelli BG Hardie DG Calmodulin-dependent protein kinase kinase-beta is an alternative upstream kinase for AMP-activated protein kinase Cell Metab 2005 2 9 19 16054095 Hindupur SK Gonzalez A Hall MN The Opposing Actions of Target of Rapamycin and AMP-Activated Protein Kinase in Cell Growth Control Cold Spring Harb Perspect Med 2015 5 a019141 Hsu HJ Drummond-Barbosa D Insulin levels control female germline stem cell maintenance via the niche in Drosophila Proc Natl Acad Sci U S A 2009 106 1117 1121 19136634 Hsu HJ Drummond-Barbosa D Insulin signals control the competence of the Drosophila female germline stem cell niche to respond to Notch ligands Dev Biol 2011 350 290 300 21145317 Hsu HJ LaFever L Drummond-Barbosa D Diet controls normal and tumorous germline stem cells via insulin-dependent and -independent mechanisms in Drosophila Dev Biol 2008 313 700 712 18068153 Hurley RL Anderson KA Franzone JM Kemp BE Means AR Witters LA The Ca2+/calmodulin-dependent protein kinase kinases are AMP-activated protein kinase kinases J Biol Chem 2005 280 29060 29066 15980064 Jia D Soylemez M Calvin G Bornmann R Bryant J Hanna C Huang YC Deng WM A large-scale in vivo RNAi screen to identify genes involved in Notch-mediated follicle cell differentiation and cell cycle switches Sci Rep 2015 5 12328 26205122 Kayampilly PP Menon KM Follicle-stimulating hormone inhibits adenosine 5′-monophosphate-activated protein kinase activation and promotes cell proliferation of primary granulosa cells in culture through an Akt-dependent pathway Endocrinology 2009 150 929 935 18927218 Kim M Lee JH Identification of an AMPK phosphorylation site in Drosophila TSC2 (gigas) that regulate cell growth Int J Mol Sci 2015 16 7015 7026 25826530 LaFever L Drummond-Barbosa D Direct control of germline stem cell division and cyst growth by neural insulin in Drosophila Science 2005 309 1071 1073 16099985 LaFever L Feoktistov A Hsu HJ Drummond-Barbosa D Specific roles of Target of rapamycin in the control of stem cells and their progeny in the Drosophila ovary Development 2010 137 2117 2126 20504961 Laws KM Drummond-Barbosa D Genetic Mosaic Analysis of Stem Cell Lineages in the Drosophila Ovary Methods Mol Biol 2015 1328 57 72 26324429 Lee IJ Lee CW Lee JH CaMKKbeta-AMPKalpha2 signaling contributes to mitotic Golgi fragmentation and the G2/M transition in mammalian cells Cell Cycle 2015 14 598 611 25590814 Lee JH Koh H Kim M Kim Y Lee SY Karess RE Lee SH Shong M Kim JM Kim J Chung J Energy-dependent regulation of cell structure by AMP-activated protein kinase Nature 2007 447 1017 1020 17486097 Lee JH Koh H Kim M Park J Lee SY Lee S Chung J JNK pathway mediates apoptotic cell death induced by tumor suppressor LKB1 in Drosophila Cell Death Differ 2006 13 1110 1122 16273080 Lilly MA Duronio RJ New insights into cell cycle control from the Drosophila endocycle Oncogene 2005 24 2765 2775 15838513 Lopez-Schier H St Johnston D Delta signaling from the germ line controls the proliferation and differentiation of the somatic follicle cells during Drosophila oogenesis Genes Dev 2001 15 1393 1405 11390359 Maines JZ Stevens LM Tong X Stein D Drosophila dMyc is required for ovary cell growth and endoreplication Development 2004 131 775 786 14724122 Margolis J Spradling A Identification and behavior of epithelial stem cells in the Drosophila ovary Development 1995 121 3797 3807 8582289 Martin SG St Johnston D A role for Drosophila LKB1 in anterior-posterior axis formation and epithelial polarity Nature 2003 421 379 384 12540903 Mungai PT Waypa GB Jairaman A Prakriya M Dokic D Ball MK Schumacker PT Hypoxia triggers AMPK activation through reactive oxygen species-mediated activation of calcium release-activated calcium channels Mol Cell Biol 2011 31 3531 3545 21670147 Narbonne P Roy R Inhibition of germline proliferation during C. elegans dauer development requires PTEN, LKB1 and AMPK signalling Development 2006 133 611 619 16407400 Nystul T Spradling A Regulation of epithelial stem cell replacement and follicle formation in the Drosophila ovary Genetics 2010 184 503 515 19948890 Ruohola H Bremer KA Baker D Swedlow JR Jan LY Jan YN Role of neurogenic genes in establishment of follicle cell fate and oocyte polarity during oogenesis in Drosophila Cell 1991 66 433 449 1907889 Song Z McCall K Steller H DCP-1, a Drosophila cell death protease essential for development Science 1997 275 536 540 8999799 Spradling AC Bate M Developmental genetics of oogenesis The Development of Drosophila melanogaster 1993 Cold Spring Harbor Laboratory Press Plainview, NY Sun P Quan Z Zhang B Wu T Xi R TSC1/2 tumour suppressor complex maintains Drosophila germline stem cells by preventing differentiation Development 2010 137 2461 2469 20573703 Tosca L Rame C Chabrolle C Tesseraud S Dupont J Metformin decreases IGF1-induced cell proliferation and protein synthesis through AMP-activated protein kinase in cultured bovine granulosa cells Reproduction 2010 139 409 418 19906888 Towler MC Hardie DG AMP-activated protein kinase in metabolic control and insulin signaling Circ Res 2007 100 328 341 17307971 Vazquez-Martin A Oliveras-Ferraros C Cufi S Menendez JA Polo-like kinase 1 regulates activation of AMP-activated protein kinase (AMPK) at the mitotic apparatus Cell Cycle 2011 10 1295 1302 21474997 Woods A Dickerson K Heath R Hong SP Momcilovic M Johnstone SR Carlson M Carling D Ca2+/calmodulin-dependent protein kinase kinase-beta acts upstream of AMP-activated protein kinase in mammalian cells Cell Metab 2005 2 21 33 16054096 Xu T Rubin GM Analysis of genetic mosaics in developing and adult Drosophila tissues Development 1993 117 1223 1237 8404527 Yang SA Wang WD Chen CT Tseng CY Chen YN Hsu HJ FOXO/Fringe is necessary for maintenance of the germline stem cell niche in response to insulin insufficiency Dev Biol 2013 382 124 135 23895933 Yuan HX Xiong Y Guan KL Nutrient sensing, metabolism, and cell growth control Mol Cell 2013 49 379 387 23395268
PMC005xxxxxx/PMC5124397.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 2984726R 1011 Biochem J Biochem. J. The Biochemical journal 0264-6021 1470-8728 27671892 5124397 10.1042/BCJ20160565 NIHMS821231 Article Sphingolipid biosynthesis upregulation by TOR Complex 2-Ypk1 signaling during yeast adaptive response to acetic acid stress Guerreiro Joana F. 1† Muir Alexander 23† Ramachandran Subramaniam 2 Thorner Jeremy 2* Sá-Correia Isabel 1* 1 iBB-Institute for Bioengineering and Biosciences, Department of Bioengineering, Instituto Superior Técnico, Universidade de Lisboa, 1049-001 Lisbon, Portugal 2 Division of Biochemistry, Biophysics and Structural Biology, Dept. of Molecular and Cell Biology 3 Chemical Biology Graduate Program, University of California, Berkeley CA 94720-3202 USA * To whom correspondence should be addressed: [email protected]; or, [email protected] † These authors contributed equally to this work. 8 10 2016 26 9 2016 1 12 2016 01 12 2016 473 23 43114325 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Acetic acid-induced inhibition of yeast growth and metabolism limits the productivity of industrial fermentation processes, especially when lignocellulosic hydrolysates are used as feedstock in industrial biotechnology. Tolerance to acetic acid of food spoilage yeasts is also a problem in the preservation of acidic foods and beverages. Thus, understanding the molecular mechanisms underlying adaptation and tolerance to acetic acid stress is increasingly important in industrial biotechnology and the food industry. Prior genetic screens for S. cerevisiae mutants with increased sensitivity to acetic acid identified loss-of-function mutations in the YPK1 gene, which encodes a protein kinase activated by the Target of Rapamycin (TOR) Complex 2 (TORC2). We show here by several independent criteria that TORC2-Ypk1 signaling is stimulated in response to acetic acid stress. Moreover, we demonstrate that TORC2-mediated Ypk1 phosphorylation and activation is necessary for acetic acid tolerance, and occurs independently of Hrk1, a protein kinase previously implicated in the cellular response to acetic acid. In addition, we show that TORC2-Ypk1-mediated activation of L-serine: palmitoyl-CoA acyltransferase, the enzyme complex that catalyzes the first committed step of sphingolipid biosynthesis, is required for acetic acid tolerance. Furthermore, analysis of the sphingolipid pathway using inhibitors and mutants indicates that it is production of certain complex sphingolipids that contributes to conferring acetic acid tolerance. Consistent with that conclusion, promoting sphingolipid synthesis by adding exogenous long-chain base precursor phytosphingosine to the growth medium enhanced acetic acid tolerance. Thus, appropriate modulation of the TORC2-Ypk1-sphingolipid axis in industrial yeast strains may have utility in improving fermentations of acetic acid-containing feedstocks. budding yeast weak acid adaptation lipid metabolism protein kinases protein phosphorylation INTRODUCTION Understanding the cellular response to acetic acid stress, and the ability to manipulate it, are increasingly important in two major industries. First, large-scale fermentations by the yeast Saccharomyces cerevisiae are contributing to an ever greater extent to the production of ethanol, other biofuels and bulk chemicals, and specialty chemicals and biologicals [1]. In this context, use of plant biomass as feedstock in place of simple sugars will have considerable economic impact [2, 3]. However, one impediment to using plant biomass-derived feedstocks is the presence of acetic acid in lignocellulosic hydrolysates, which inhibits yeast growth and hinders fermentation [4, 5]. Second, in the food industry, weak acids (in particular, acetic acid) are often used as preservatives. However, certain microbes are acetic acid-tolerant and, thus, able to survive and proliferate in foods and beverages preserved with this weak acid, causing spoilage that results in major economic loss [6]. At the low pH values reached in S. cerevisiae cultures, significantly below the pKa of acetic acid (~4.75), this weak acid is mainly in its undissociated state and, hence, uncharged, permitting its diffusion across the plasma membrane (PM), a process also facilitated through the aquaglyceroporin Fps1 [7, 8]. Upon entering the more basic environment of the cytosol, the acid dissociates into acetate anion and H+, which accumulate, thereby decreasing cytosolic pH, increasing membrane stretch (brought on by an increase in turgor pressure), and elevating reactive oxygen species [9, 10]. Ultimately, these stresses cause decreased metabolic activity [11], decreased cell growth and viability [12, 13], and decreased survival in stationary phase [14]. A number of adaptive responses of S. cerevisiae to acetic acid stress have been identified. Yeast cells challenged with acetic acid were found to increase vacuolar and plasma membrane (PM) H+-ATPase activity, presumably to allow recovery of the cytosolic pH to more physiological values [15, 16]. Exposure to high levels of acetic acid (≥100 mM) activates the Hog1 MAPK, which phosphorylates Fps1, thus promoting its endocytosis and thereby reducing acetic acid influx [7]. Exposure to acetic acid also induces the expression of the PM-localized members of the major facilitator superfamily Tpo2 and Tpo3 [17], via activation of the transcription factor Haa1 helping to reduce the concentration of the accumulated intracellular acetate. Genome-wide transcriptome analysis and functional genomics studies have revealed additional loci under Haa1 control, including the HRK1 gene, which encodes a H+-ATPase-activating protein kinase [17, 18], and other genes whose products contribute globally to yeast cell adaptation to acetic acid [19]. Also, phenotypic analysis of the S. cerevisiae genome-wide deletion collection identified a number of additional genes important for survival in response to acetic acid stress, including the protein kinase Ypk1 [20, 21]. The knowledge gathered so far on the understanding of molecular factors and cellular pathways involved in acetic acid tolerance has been used in the development of acetic acid-tolerant strains, either through the deletion or overexpression of single genes [22, 23], manipulation of genes that play a crucial role in the regulatory cascades that control acetic acid tolerance, such as manipulation of the Haa1-regulon [24–26], or the use of evolutionary engineering strategies to select strains that possess increased tolerance to the acid [27]. Manipulation of the composition of the growth medium in order to increase yeast tolerance to acetic acid has also been reported [20]. Ypk1 is a member of the AGC subfamily of protein kinases [28] and the ortholog of mammalian SGK1 [29]. Ypk1 activity requires phosphorylation of Thr504 in its activation loop by the protein kinases Pkh1 and Pkh2, which are associated with PM-localized structures called eisosomes [30, 31]. Ypk1 activity can be substantially increased by further phosphorylation at sites in its C-terminal domain (especially Ser644 and Thr662) by the PM-associated Target of Rapamycin (TOR) Complex 2 (TORC2) [31–33]. Certain membrane stresses, including sphingolipid depletion [32], heat shock [34], and hypotonic conditions [35], elevate TORC2 activity and consequently activate Ypk1, whereas other stresses, such as hypertonic conditions [36, 37], diminish TORC2 and Ypk1 activity. Identification of physiologically-relevant substrates of Ypk1 has shed light on the molecular mechanisms downstream of TORC2-Ypk1 activation that promote tolerance to these various membrane stresses [32, 38–40]. For example, upon sphingolipid depletion or in hypo-osmotic medium (which causes membrane stretch due to the increased turgor pressure), Ypk1 phosphorylates Orm1 and Orm2 [32], thus blocking the inhibition that these two proteins normally exert on L-serine:palmitoyl-CoA acyltransferase (SPT), the enzyme complex catalyzing the first step in sphingolipid biosynthesis [41], thus increasing flux into this pathway. Moreover, Ypk1 phosphorylates and stimulates Lac1 and Lag1, the catalytic subunits of the ceramide synthase complex [40], thereby ensuring that the SPT-generated long-chain (sphingoid) base (LCB), mainly phytosphingosine (PHS), and fatty acyl-CoA [typically very long chain fatty acids (VLCFA), mainly C26] are diverted more efficiently into ceramides for use in the production of complex sphingolipids [42]. Furthermore, TORC2-Ypk1-dependent phosphorylation of both Orm1 and Orm2 and Lac1 and Lag1 are required for yeast cell survival under the stresses of sphingolipid depletion and hypotonic shock [32, 40]. Given that TORC2-Ypk1 signaling is known to respond to membrane stretch, like that elicited by acetic acid treatment, and that functional genomics suggested a role for Ypk1 in the adaptive response to acetic acid, we investigated whether TORC2-Ypk1 activity responds to challenging yeast cells with an inhibitory sub-lethal exogenous acetic acid concentration and whether TORC2-Ypk1 function contribute to the ability of yeast cells to survive the stress imposed by exposure to this weak acid. If so, we sought to determine whether it is changes in the production of sphingolipids per se that is responsible for the role that TORC2-Ypk1 signaling plays in the maintenance of acetic acid tolerance. Our findings provide previously uncharacterized mechanistic insights about how yeast cells adapt to exposure to acetic acid. RESULTS TORC2-dependent phosphorylation of Ypk1 is activated upon challenge with acetic acid We first examined whether TORC2-Ypk1 signaling is modulated under acetic acid stress using a phosphosite-specific antibody that recognizes P-T662 in Ypk1 [33]. We have found (our unpublished results) that this antibody cross-reacts with an equivalent residue, P-T659, in the paralog Ypk2, which is also known to be phosphorylated in a TORC2-dependent manner [31]. Upon exposure of cells to 50 mM acetic acid, we observed a readily detectable and reproducible increase in the fraction of Ypk1 molecules phosphorylated at T622 (and a similar increase in P-T659 in Ypk2) (Fig. 1A). By contrast, under the exact same conditions, there was no observable increase in phosphorylation at T504 (Fig. 1B), the site regulated by Pkh1 and Pkh2 [30]. Thus, TORC2-dependent phosphorylation of Ypk1 and Ypk2 was specifically increased in response to acetic acid treatment, whereas Pkh1- and Pkh2-dependent phosphorylation was not. TORC2 and Ypk1 action are important for cells to cope with acetic acid stress We next investigated whether TORC2 activity and, more specifically, its activation of Ypk1 are important for cellular adaptation to acetic acid. First, growth of a tor2-as strain, which has chronically diminished TORC2 activity [43], was compared to growth of an otherwise isogenic TOR2+ strain on a defined medium (MM4) at pH 4 in the absence or presence of 60 mM acetic acid. Although both TOR2+ and tor2-as strains grew at very similar rates in the control medium lacking acetic acid, the tor2-as cells exhibited a much more protracted lag phase and a slower overall growth rate than the TOR2+ cells in the medium containing acetic acid (Fig. 1C). Second, and similarly, we used a Ypk1 mutant, Ypk1(T662A), that also displays diminished functionality [44]. Again, we observed that, although both YPK1+ and ypk1(T662A) strains grew at very similar rates in the control medium lacking acetic acid, the ypk1(T662A) cells exhibited a more protracted lag phase and a slower overall growth rate than the YPK1+ cells in the medium containing acetic acid (Fig. 1D). Thus, the stimulation of TORC2 in response to acetic acid, and its subsequent activation of Ypk1 by phosphorylation at T662, are physiologically important for the ability of yeast cells to adapt to sudden exposure to acetic acid. Thus, Ypk1 is the major downstream effector of TORC2 in this stress response. Phosphorylation of Ypk1 substrates is upregulated in response to acetic acid stress When Ypk1 becomes activated, among its well-documented in vivo substrates are Orm1 and Orm2, two protein inhibitors of SPT activity. It has been amply demonstrated that Ypk1-mediated phosphorylation of Orm1 occurs specifically on residues S51, S52 and S53 and, likewise, that Ypk1-mediated phosphorylation of Orm2 occurs at the equivalent residues (S45 S47 S48) [32]. Similarly, Ypk1 phosphorylates Lac1 and Lag1, the two highly related catalytic subunits of the ceramide synthase complex, at equivalent sites (S23 and S24 in both proteins) [40]. Therefore, to further verify that acetic acid stress activates Ypk1, we tested whether acetic treatment induced the Ypk1-mediated phosphorylation of these substrates. Modification of these proteins is most conveniently monitored by the fact that the mobility of the phosphorylated species is retarded, compared to the unphosphorylated state, when the proteins are resolved by SDS-PAGE or using the Phos-tag™ gel method. Indeed, we found that acetic acid treatment caused a prominent increase in the slower mobility Orm1 species (Fig. 2A, left panel), indicative of greatly enhanced phosphorylation. The appearance of these species was totally abrogated when cells expressed (as the sole source of Orm1) a mutant (orm1-P4) [41] in which the known Ypk1 phosphorylation sites are mutated to Ala (Fig. 2A, right panel), consistent with the conclusion that Ypk1-dependent phosphorylation is likely responsible for the observed mobility shift. However, in addition to S51A S52A S53A, the orm1-P4 allele also carries five other mutations (S29A S32A S34A S35A S36A) [41]. Hence, we could not rule out that some (or all) of the observed phosphorylated species might arise from modifications occurring at these other positions. For this reason, we also examined Lag1 and Lac1. Reassuringly, we found that acetic acid treatment also caused a marked increase in the slower mobility species for both Lag1 (Fig. 2B) and Lac1 (data not shown); moreover, these mobility shifts were completely abolished by mutation of just the two known Ypk1 sites in both proteins (S23A S24A), confirming that Ypk1 was responsible for the observed acetic acid-induced modifications. Ypk1 stimulation of the sphingolipid pathway is required for acetic acid tolerance Having found that TORC2-Ypk1 activation in response to acetic acid triggers the phosphorylation of targets that should result in an increase in both SPT and ceramide synthase activity raised the possibility that the role that TORC2 and Ypk1 play in promoting adaptation to acetic acid stress is upregulation of sphingolipid biosynthesis. This hypothesis is consistent with two additional observations. First, other enzymes involved in sphingolipid metabolism have been implicated genetically in the response to acetic acid stress [20, 45]. Second, lipidomic analysis has shown that readily detectable increases in sphingolipid species occur in response to acetic acid treatment [46]. If TORC2-Ypk1-dependent upregulation of sphingolipid synthesis is critical for adaptation to acetic acid stress, we reasoned that preventing TORC2-Ypk1-mediated phosphorylation of Orm1 and Orm2, or Lac1 and Lag1, or both, would compromise the ability of yeast cells to grow when challenged with acetic acid. In agreement with this conclusion, we found that an ORM1+ ORM2+ strain and otherwise isogenic cells carrying the orm1-P4 orm2-P2 alleles, in which all of the Ypk1 sites (in bold) in both proteins have been mutated to Ala [P4 = S29A S32A S34A S35A S36A S51A S52A S53A; P2 = S9A S15A T18A S46A S47A S48A] grew at very similar rates in the control medium lacking acetic acid; but, in marked contrast, and unlike the parental ORM1+ ORM2+ strain, the orm1-P4 orm2-P2 cells were unable to grow in the medium containing acetic acid (Fig. 2C). It has been shown in prior work that orm1-P4 orm2-P2 cells have decreased SPT activity and are hypersensitive to myriocin, an antibiotic that is a potent SPT inhibitor [41], and that TORC2-dependent and Ypk1-mediated phosphorylation of Orm1 and Orm2 is required to derepress sphingolipid biosynthetic activity [32]. Thus, the inability of orm1-P4 orm2-P2 cells to grow in medium containing high acetic acid is consistent with the need for Ypk1-evoked stimulation of LCB production to cope with this stress. It has been observed previously that basal ceramide synthase activity is quite high and that Ypk1-mediated phosphorylation stimulates this enzyme complex just 2-fold [40]. Moreover, cells carrying the lac1(S23A S24A) lag1(S23A S24A) alleles (lacking the Ypk1 sites in both proteins) are more sensitive to myriocin than otherwise isogenic LAC1+ LAG1+ cells, but only modestly so [40]. Nonetheless, although LAC1+ LAG1+ and lac1(S23A S24A) lag1(S23A S24A) cells grew quite comparably in the control medium lacking acetic acid, there was detectably poorer growth of the lac1(S23A S24A) lag1(S23A S24A) cells, compared to the LAC1+ LAG1+ control, in the medium containing acetic acid (Fig. 2D). Again, however, this effect was modest, compared to the severe growth inhibition observed upon challenge with an inhibitory acetic acid concentration when the blockade of SPT activity imposed by Orm1 and Orm2 cannot be alleviated by their Ypk1-mediated phosphorylation (Fig. 2E). Thus, elevating flux into the sphingolipid biosynthetic pathway via TORC2-dependent Ypk1-mediated activation of SPT is a prerequisite for achieving cell survival and adaptation in response to acetic acid stress. TORC2-Ypk1 activation is not dependent on protein kinase Hrk1 We next sought to determine whether any previously described acetic acid-evoked signaling system might be responsible for the activation of TORC2-Ypk1 by acetic acid stress that we observed. We focused our attention on the protein kinase Hrk1, an enzyme implicated in the regulation of the PM H+-ATPase (Pma1) [18] and expressed under the control of the transcription factor Haa1 in response to acetic acid stress [17], but otherwise a very poorly characterized member of the Npr1 sub-family of yeast protein kinases [47]. When cultivated in the presence of a mildly inhibitory concentration of acetic acid, hrk1Δ cells exhibit a protracted lag phase and an increase in the intracellular concentration of this weak acid, compared to HRK1+ cells [17]. Also, effects on the biosynthesis of complex sphingolipids are among the many cellular roles ascribed to the related protein kinase Npr1, a more well studied enzyme that is negatively regulated by TORC1 [48, 49]. Therefore, we examined whether Hrk1 is necessary for the TORC2-Ypk1 activation that occurs under acetic acid stress. However, upon challenge with 50 mM acetic, we found that TORC2-mediated phosphorylation of Ypk1 and Ypk2 was unaffected in hrk1Δ cells, compared to HRK1+ control cells (Fig, 3A) and, likewise, that Ypk1-mediated Orm1 phosphorylation was not affected (Fig. 3B). Thus, Hrk1 is not required for either TORC2 activation or Ypk1 signaling to the sphingolipid biosynthetic pathway upon acetic acid stress. Acetic acid tolerance requires production of complex sphingolipids Activation of TORC2-Ypk1 signaling increases metabolic flux into the sphingolipid pathway, which commences with the generation of LCBs, which are converted to ceramides and then to more complex sphingolipids. Therefore, we next investigated which of the intermediates or end-products of the sphingolipid biosynthetic pathway must be produced for cells to acquire acetic acid tolerance. Toward this end, we used either chemical and/or genetic means to block the sphingolipid pathway at discrete steps (Fig. 4A) and assessed the ability of the cells subjected to these blocks to cope with the challenge of acetic acid stress, as judged by their growth rate. We found, first, that a sublethal dose of myriocin, which causes only a minor effect on the specific growth rate and final biomass concentration of wild-type cells in medium lacking acetic acid, greatly exacerbated the growth inhibitory effect of exposure to 60 mM acetic acid at pH 4 (Fig. 4B), consistent with the need to upregulate flux into the sphingolipid pathway via generation of LCBs to adapt to acetic acid stress. Likewise, preventing utilization of the primary yeast LCB, PHS, by addition of fumonisin B1, a specific inhibitor of ceramide synthase [50], also greatly potentiated the growth inhibitory effect of acetic acid treatment (Fig. 4C). Similarly, using the Tet-Off system to block production of phosphatidylinositol:ceramide phosphoinositol transferase (Aur1), the enzyme responsible for inositolphosphorylceramide (IPC) synthesis [51], while causing no obvious growth defect in medium lacking acetic acid, caused a severe growth debility in the medium containing acetic acid (Fig. 4D). At last, we found in our work that ipt1Δ cells, which lack the enzyme (inositolphosphotransferase) necessary for mannose di(inositolphosphoryl)ceramide [M(IP)2C] formation, are no more acetic acid sensitive than IPT1+ control cells (Fig. 4E), suggesting that either IPC or mannosylinositol phosphorylceramide (MIPC), but not the M(IP)2C derived from it, are required for adaptation to acetic acid stress. Supplementation with phytosphingosine increases acetic acid tolerance If elevated intracellular LCB production in response to acetic acid challenge is required to drive higher levels of complex sphingolipids to permit cellular adaptation to this stress, then we reasoned that perhaps the simple expedient of supplementing the growth medium with the cell-permeable LCB PHS might be sufficient to increase flux into the sphingolipid biosynthetic pathway and thereby confer greater acetic acid tolerance. Supplementing the MM4 medium with 20–30 µM PHS did not impair the growth of wild-type (ORM1+ OMR2+) cells in medium lacking acetic, nor did it markedly improve the growth of the cells in the presence of 40 mM acetic acid (Fig. 5, top row). However, in otherwise isogenic cells carrying the orm1-P4 orm2-P2 alleles (where upregulation of flux into the sphingolipid pathway in response to acetic acid-evoked TORC2-Ypk1 signaling cannot occur), which had significantly impaired growth in the presence of 40 mM acetic acid, the presence of PHS in this concentration range (20–30 mM) permitted detectable improvement in the level of survival (Fig. 5, bottom rows). Thus, in the presence of an inhibitory concentration of acetic acid, LCBs (and the sphingolipids derived from them, based on the other results we have presented here) are indeed limiting factors for adaptation to this stressful condition. DISCUSSION In this work, we obtained several important new insights about the mechanisms by which S. cerevisiae cells combat exposure to an inhibitory but sub-lethal acetic acid concentration. We documented, for the first time, that under this stress TORC2-Ypk1 signaling is activated. Moreover, we demonstrated that TORC2-Ypk1-mediated stimulation of sphingolipid biosynthesis makes a major contribution to the ability of the cells to endure acetic acid stress because inhibition of either TORC2 or Ypk1, or blockade of sphingolipid production, substantially exacerbates the growth inhibitory effect of exposure to acetic acid. Conversely, we showed that even a simple manipulation to increase through-put through the sphingolipid pathway, such as supplying exogenously the specific pathway precursor PHS, improves the tolerance of a susceptible yeast strain to acetic acid. Our findings greatly extend previous genomewide studies that implicated genes involved in sphingolipid metabolism, among many other genes, in acquisition of tolerance to acetic acid [17, 20, 21]. Our results also explain the underlying biochemical basis for a prior observation, revealed by lipidomic profiling. It was found that exponentially-growing yeast cells adapted to a concentration of acetic acid that reduced their doubling time by 50% exhibited an increase in their total content of complex sphingolipids and extensive changes in the profile of the complex sphingolipids present, compared to cells cultivated in the same way, but not exposed to acetic acid [46]. The conclusions of our study are also in agreement with a recent report [52] providing evidence that the intrinsically high sphingolipid content in the PM of Zygosaccharomyces bailii, a hemiascomycete distantly related to S. cerevisiae, is an important contributor to the high acetic acid tolerance of this organism. In this regard, we demonstrated that, at least in S. cerevisiae, it is IPC and MIPC, and not M(IP)2C, that are required for acetic acid tolerance. Three observations suggest this. First, it has been observed in prior work that a mutation (csg2Δ) that impairs MIPC formation causes growth impairment at low pH values [21, 53], like the pH 4 MM4 medium that we used to test the effect of exogenous acetic acid concentration. Second, and more tellingly, it has been reported previously that sur1Δ cells, which also have reduced MIPC production, are more susceptible to the growth inhibitory effect of acetic acid [20]. Third, we observed in this work that a mutant with impaired M(IP)2C formation, exhibited no increase in acetic acid sensitivity when compared with the corresponding parental strain. Collectively, the results presented are consistent with the conclusion that it is the early complex sphingolipids IPC and MIPC that are the species primarily responsible for conferring acetic acid tolerance. In agreement with this scenario, loss-of-function mutations in SUR2 (an enzyme that generates PHS de novo) [53] and in YPC1 (an enzyme that can produce PHS from the breakdown of phytoceramides) [54] have lower levels of this LCB, and both genes are transcriptionally upregulated in a Haa1-dependent manner in response to acetic acid stress [17, 20]. This upregulation presumably leads to an increase in the rate of synthesis of those sphingolipid species most needed to best resist the effects of acetic acid-induced stress. In this same regard, a scs7Δ mutant, which lacks an enzyme necessary to alpha-hydroxylate the VLCFA in sphingolipids, contains an altered spectrum of IPC species and is more sensitive to acetic acid than the corresponding parental strain [20]. Given that TORC2, Ypk1 and the mechanism by which they regulate sphingolipid production appear to be largely conserved throughout the fungal clade [40], the response we have described here likely plays an important role in the tolerance to acetic acid of many other yeast species, such as Z. bailii [52]. Given what we have now established, several questions arise that are worthy of further investigation. First, it is unclear why elevation of complex sphingolipids permits growth on medium containing inhibitory concentrations of acetic acid. One suggestion is that an increase in the content of VLCFA-containing sphingolipids leads to a thicker and more dense PM, thereby increasing the free energy barrier for permeation of acetic acid [52], or enhancing cohesion among the lipids in the outer leaflet, which would be important to counteract the hypotonic shock reported to be induced by this weak acid [55]. Moreover, inadequate sphingolipid biosynthesis is known to deleteriously influence the trafficking of integral membrane proteins from the Golgi body to the PM [56, 57], and several proteins required for this vesicle-mediated protein sorting (Sur2, Sur4 and Ypc1) have known roles in sphingolipid metabolism [58]. Since several PM proteins (e.g., Fps1, Pma1, Tpo2, Tpo3), as well as Yro2 and its paralog Mrh1 [59], have been implicated in yeast adaptive response to acetic acid, it is possible that increasing sphingolipid production may increase the efficiency of the delivery to the PM of these proteins required for acetic acid tolerance. Another open question of interest is how acetic acid stress stimulates TORC2 activity. TORC2 could be coupled physically to a PM protein that serves as a direct sensor of acetic acid or its influx, in agreement with the recent observation that Tor2 (the catalytic component of TORC2) interacts physically with Sho1, an integral PM protein that is a putative osmosensor [60]. Additionally, TORC2 level or activity could be increased indirectly by changes elicited by acetic acid-sensing pathways or by changes in cellular properties brought on by the effects of this weak acid. Neither the transcription of TOR2, nor any of the genes encoding the other subunits of the TORC2 complex (AVO1, AVO2, AVO3/TSC11, BIT2, BIT61, SLM1 or SLM2), nor YPK1, is induced by acetic acid stress or dependent for their expression under acetic acid stress on the Haa1 transcription factor [17]. Thus, as we showed by immunoblotting for Ypk1 itself, the levels of the components of TORC2 are likely not affected when cells are challenged with acetic acid. Similarly, we found that the protein kinase Hkr1 that is upregulated by acetic acid treatment in a Haa1-dependent manner, is not required for the observed increase in TORC2-Ypk1 signaling that occurs in response to acetic acid stress. So, given that TORC2 is a PM-associated protein complex [61], TORC2 may instead sense the presence of acetic acid indirectly by responding to the ensuing membrane stretch, akin to the activation of TORC2 elicited by other hypo-osmotic stress conditions, which appears to depend on changes in the localization of the Slm1 and Slm2 subunits of TORC2 [35]. Alternatively, given that the MAPK Slt2/Mpk1 is also activated under hypo-osmotic conditions [62], it may be involved, directly or indirectly, in stimulating TORC2 activity. A third lingering question is whether TORC2-activated Ypk1 also contributes to acetic acid tolerance in other ways, aside from elevating the production of IPC and MIPC per se. For example, sphingolipids are largely found in the outer leaflet of the PM [41, 56, 57], and Ypk1 also affects the composition of the PM bilayer by negatively regulating the protein kinases Fpk1 and Fpk1, whose function is, in turn, to stimulate the flippases (mainly Dnf1 and Dnf2) that translocate aminoglycerophospholipids from the outer to the inner leaflet [38]. Thus, Ypk1-mediated down-modulation of flippase function may be physiologically important in altering the properties of the PM in response to acetic acid stress. One group has reported that Lag1-dependent de novo ceramide production (as well as the production of ceramides by the mitochondrially-localized complex sphingolipid hydrolase Isc1) is causal for the mitochondrial damage and increased cell inviability induced by acetic acid [45]. However, as we have shown before, activated Ypk1 phosphorylates Lag1 and Lac1 and stimulates ceramide production [40], which, as we have documented here, is protective for cells exposed to acetic acid. In this same vein, the apparent damage to mitochondria caused by acetic acid treatment reportedly increases cellular ROS levels to a harmful level [63, 64] and, consistent with the findings we present here, others have shown that TORC2-Ypk1 signaling and sphingolipid production are important for maintaining cellular ROS levels in a tolerable range [65], although the mechanism by which TORC2-Ypk1 signaling does so is unclear. Finally, from a practical standpoint, we have shown here that just promoting sphingolipid production by adding PHS to the growth medium provides a clear survival benefit to cells during adaptation to acetic acid stress. Thus, appropriate engineering to upregulate the sphingolipid pathway in situ may provide a readily accessible and rational means to enhance the acetic acid tolerance of industrially valuable S. cerevisiae strains, either alone or in combination with other genetic strategies that appear to improve the acetic acid tolerance of this organism [66–68]. MATERIALS AND METHODS Construction of yeast strains and growth conditions S. cerevisiae strains used in this study are listed in Table 1. Strains were constructed using standard yeast genetic manipulations [69]. When generating strains with chromosomal modification, integration of the DNA fragment into the correct genomic loci was confirmed by PCR using genomic DNA from isolated colonies as template and with oligonucleotides complementary to the integrated DNA fragment and complementary to genomic sequence at least 150 bases away from the integration site as primers for the reaction. Yeast cultures were grown in either Yeast Peptone Dextrose medium (YPD) (containing per liter: 10 g Bacto yeast extract, 20 g Bacto Peptone, and 20 g glucose) or MM4 medium adjusted to pH 4.0 (containing per liter: 1.7 g yeast nitrogen base without amino acids or ammonium sulfate, 20 g glucose, 2.65 g ammonium sulfate and 60 mg leucine, and further supplemented with 20 mg methionine, 20 mg histidine, 60 mg leucine, 20 mg uracil, or 30 mg lysine according to the strains' autotrophy). Plasmids and recombinant DNA methods Plasmids used in this study are listed in Table 2. All plasmids were constructed by standard laboratory methods [70]. Constructs generated in this study had all promoter and coding regions sequences confirmed by sequencing analysis. Preparation of cell extracts and immunoblotting of phosphorylated proteins Immunoblot analysis of phosphorylated Lac1, Lag1 and Ypk1 species was performed as previously described [37, 40]. Briefly, alkaline lysis and trichloroacetic acid precipitation [71] was used to make whole cell protein extracts. To detect Ypk1 phosphorylated species, 15 µL of the extract was resolved by standard SDS-PAGE using 8% acrylamide gels. To detect Lac1 and Lag1 phosphorylated species, 15 µL of TCA extract was resolved by Phostag SDS-PAGE [72] at 160 V using gels containing 8% acrylamide, 35 µM MnCl2, and 35 µM Phostag (Wako Chemicals USA, Inc.). After SDS-PAGE resolution, protein extracts were transferred to nitrocellulose and incubated with primary antibody in Odyssey buffer (Licor Biosciences). Membranes were washed and then incubated with appropriate secondary antibodies [anti-mouse, anti-rabbit or anti-goat IRDye680LT or IRDye800 conjugated IgG (Licor Biosciences)] diluted 1:10000 in Odyssey buffer with 0.1% Tween-20 and 0.02% SDS). Blots were imaged using an Odyssey infrared scanner (Licor Biosciences). Primary antibodies and dilutions used in this study were: 1:5000 mouse anti-FLAG (Sigma-Aldrich), 1:500 rabbit anti-pSGK (T256) (to recognize Ypk1 phosphorylated at T504) (Santa Cruz Biotechnology, Dallas, TX), 1:20000 rabbit anti-Ypk1(P-T662) (generous gift from Ted Powers, University of California, Davis) and 1:1000 goat anti-Ypk1 (Santa Cruz Biotechnology, Dallas, TX). Immunopurification and immunodetection of Orm1 Exponentially growing cells expressing 3xFLAG-Orm1 were pelleted and transferred into MM4 or MM4 with 50 mM acetic acid for 1 h. 40 mL of each culture was then harvested and washed with 1 mL of media. The cells were then frozen in liquid nitrogen and stored at −80 °C. The pellets were resuspended in 0.5 mL of ice cold IP buffer (50 mM Hepes-KOH pH 7.5, 150 mM KOAc, 2 mM MgOAc, 1 mM CaCl2, 15% glycerol, 4 mM NaVO4, 50 mM NaF, 20 mM Na-PPi, and 1× complete protease inhibitor [Roche, Basel, Switzerland]) with 0.1% digitonin. An equal volume of glass beads were added to each cell slurry and cells were lysed by shaking using a FastPrep 120 (Thermo Fisher Scientific). 0.5 mL of cold IP buffer with 1.9% digitonin was added and lysates were rotated for 45 min. at 4 °C. Lysates were clarified by centrifugation for 30 min at 13,200 rpm (16,100×g) in a microfuge (Eppendorf 5415D) at 4 °C. Lysate protein concentration was measured by Pierce BCA protein assay (Thermo Fisher Scientific) and protein concentration of each lysate was normalized. Immunoprecipitation was carried out by adding 25 µl of anti-FLAG agarose (Sigma) washed with 1 ml cold IP buffer with 0.1% digitonin and resuspended in 25 µL of the same buffer to each sample and rotating this slurry at 4 °C for 2.5 h. After rotation, the resin was washed four times with 1 ml cold IP buffer with 0.1% digitonin. Protein was eluted by adding 25 µl of PBS and 10 µl of 5× Laemlli buffer, and boiling for 8 min. Five µl of the IP extract were resolved by SDS-PAGE using 10% 75:1 acrylamide:bis-acrylamide gels resolved at 120 V. Proteins were subsequently transferred from SDS–PAGE gels to nitrocellulose membranes and blotted as described above. Yeast growth assays Growth assays were carried out either in solid agar media or in liquid broth batch cultures in flasks or microtiter plates. For growth assays in agar, cells were grown in MM4 media at pH 4.5 at 30°C, with orbital agitation (250 rpm) until culture OD600 = 0.7 ± 0.07 was reached. These cells were resuspended in sterile water to obtain suspensions with OD600 = 0.05 ± 0.005. These cell suspensions and dilutions of 1:5 and 1:10 were applied as 3 µl spots onto the surface of either MM4 medium (pH 4.5) or MM4 medium (pH 4.5), supplemented with 0.05% Tergitol Type NP-40, and with adequate concentrations of acetic acid and solvent (methanol) or PHS, and incubated at 30 °C for 2 to 3 days, according to the severity of growth inhibition. Plates were then scanned on a flat bed scanner and growth phenotypes examined. For liquid growth assays in flasks, cells cultivated until mid-exponential phase (OD600 = 0.8 ± 0.08) in MM4 growth medium (at pH 4.0) were used to inoculate cultures at an initial OD600 of 0.05 ± 0.005 in MM4 (at pH 4.0) or MM4 medium supplemented with 50 mM acetic acid and either solvent alone (methanol) or 0.4 µM myriocin. Cultures were then grown at 30 °C and the OD600 of the culture was determined at the indicated times. For liquid growth assays in microtiter plates, exponentially growing cells in MM4 medium were diluted to OD600 = 0.1 or OD600 = 0.05 in either MM4 medium or MM4 medium with 40 or 50 mM acetic acid and solvent [methanol or 50% ethanol (v/v)] or 500 µM fumonisin B1 or 0.2–1 µg/ml doxycycline. Then, 100 µl or 150 µl of each culture was placed in a well of a 96 flat bottom well plate. Cultures were grown with orbital shaking at 30°C in a Tecan Infinite M-1000 PRO plate reader (Tecan Systems Inc., San Jose, CA) or FilterMax F5 Multi-Mode Microplate Readers (Molecular Devices) for 36 or 72 hr, respectively. Absorbance measurements were taken every 15 min. Absorbance values were converted to OD600 values using a standard curve of absorbance values of cultures at known OD600 taken on the same plate reader. We thank Jonathan Weissman and Motohiro Tani for the generous gift of certain yeast strains used in this study, Garrett Timmons for technical assistance, Françoise Roelants for technical assistance and critical discussion, and other members of the Thorner Lab for various reagents, plasmids, and helpful advice. FUNDING This work was supported by Ph.D. grant SFRH/BD/80065/2011 (to J.F.G.) from the Fundação para a Ciência e a Tecnologia (FCT), by NIH Predoctoral Training Grant GM07232 and a Predoctoral Fellowship from the UC Systemwide Cancer Research Coordinating Committee (to A.M.), by NIH R01 Research Grant GM21841 and Senior Investigator Award 11-0118 from the American Asthma Foundation (to J.T.), by FCT contracts PTDC/BBB-BEP/0385/2014 and UID/BIO/04565/2013 (to I.S.C.) and by funds (Project N. 007317) from Programa Operacional Regional de Lisboa 2020 to the iBB. Abbreviations IPC inositolphosphorylceramide LCB long-chain base MIPC mannosyl-inositol-phosphorylceramide M(IP)2C mannosyl-diinositolphosphorylceramide PHS phytosphingosine PM plasma membrane TORC2 target of rapamycin complex 2 VLCFA very long chain fatty acids YPD yeast extract peptone dextrose Figure 1 Exposure to an inhibitory sub-lethal acetic acid concentration stimulates TORC2-dependent Ypk1 activation and both enzymes are important for cellular adaptation to this stress. (A) A Ypk1-3xFLAG expressing strain (YDB379) was grown to mid-exponential phase in MM4 medium and then cells were switched into control MM4 medium of MM4 medium supplemented with 50 mM acetic acid for 60 min. Cells were harvested and protein extracted. Extracts were resolved by SDS-PAGE, and blotted as in Materials and Methods. (B) Extracts from control or acetic acid treated wild-type (BY4741) cells were resolved by SDS-PAGE, and blotted as in Materials and Methods. (C) TOR2 (yKL4) or hypomorphic tor2-as (yKL5) strains were grown to exponential phase in MM4 medium and then diluted to OD600=0.05 in control MM4 medium or MM4 medium supplemented with 60 mM acetic acid and grown in batch culture for the indicated times. For liquid cultures, each was grown in triplicate and the error bars indicate the S.E.M. of replicates at each time point. (D) ypk1Δ (JTY6142) cells were transformed with plasmids expressing Ypk1 (pAM20) or Ypk1(T662A) (pFR221), which prevents TORC2 phosphorylation of Ypk1 at this residue and thus prevents full activation of Ypk1 by TORC2. Growth of cells expressing these constructs in MM4 medium containing acetic acid was determined as in (C). Figure 2 Exposure to an inhibitory sub-lethal acetic acid concentration triggers Ypk1-mediated phosphorylation of proteins whose modification stimulates sphingolipid production and is important for survival under this stress. (A) A strain (JTY5125) expressing either 3xFLAG-Orm1 (YDB146) or 3xFLAG-Orm1(P4), as indicated, was grown to mid-exponential phase in MM4 medium and then cells were switched into control MM4 medium or MM4 medium supplemented with 50 mM acetic acid for 60 min. Cells were harvested and protein extracted. Extracts were resolved by SDS-PAGE, and blotted as in Materials and Methods. (B) Wild-type (BY4741) cells expressing 3xFLAG-Lag1 (yAM159-A) or 3xFLAG-Lag1(S23AS24A) (yAM163-A) from their genomic loci were exposed to acetic acid, harvested and protein extracted as in (A). Extracts were resolved by SDS-PAGE, and blotted as in Materials and Methods. (C) Cells (JTY5215) expressing either 3xFLAG-Orm1 and 3xHA-Orm2 (JTY6140) or the 3xFLAG-Orm1(P4) and 3xHA-Orm2(P2) (JTY5215) mutants (which cannot be phosphorylated by Ypk1) were grown to mid-exponential phase in MM4 medium and then diluted to OD600=0.1 and grown in the wells of microtiter plates in either control MM4 medium or MM4 medium containing 50 mM acetic acid as described in Materials and Methods. For liquid cultures, each was grown in at least quadruplicate and the error bars indicate the SEM of replicates at each time point. (D) Liquid growth assays were performed as in (C) for LAC1 LAG1 (yAM205 - A) and Lac1(S23AS24A) Lag1(S23AS24A) (yAM207 - B) strains, as indicated. (E) Cells expressing either 3xFLAG-Orm1 and 3xHA-Orm2 (JTY6140) or 3xFLAG-Orm1(P4) and 3xHA-Orm2(P2) (JTY5215) were grown to mid-exponential phase and then serial dilutions were made and spotted onto MM4 agar plates containing the indicated acetic acid concentrations. Plates were imaged after 3 days of growth at 30 °C. Figure 3 The acetic acid-tolerance determinant protein kinase Hrk1 is not required for either TORC2 activation or Ypk1 action upon acetic acid stress. (A) Ypk1-3xFLAG (YDB379) or hrk1Δ Ypk1-3xFLAG (yAM336) strains were treated as in Fig. 1A. Ypk1 T662 (and Ypk2 T659) phosphorylation was determined by immunoblotting as in Fig. 1A. (B) 3xFLAG-Orm1 (YDB146) or hrk1Δ 3xFLAG-Orm1 (yAM336) strains were treated as above. Orm1 phosphorylation was determined by immunoblotting as in Fig. 2A. Figure 4 Perturbation of complex sphingolipid synthesis causes increased acetic acid sensitivity. (A) Schematic diagram of sphingolipid biosynthesis in S. cerevisiae. Adapted from (52). Pharmacological or genetic perturbations used in this study and their points of inhibition are shown. (B) Growth of wild-type (BY4741) yeasts in MM4 medium supplemented with or without acetic acid in the presence of vehicle (methanol) or 0.4 µM myriocin in the same solvent was determined by liquid growth assay as in Fig. 1C. (C) Wild-type (BY4741) yeasts were grown to exponential phase in MM4 medium and then diluted to OD600=0.05 and grown in MM4 medium supplemented with or without acetic acid in the presence of vehicle (methanol) or 500 µM fumonisin B1 in the same solvent. (D) Growth of a tet-AUR1 strain (MTY175) in MM4 medium supplemented with or without acetic acid in the presence of vehicle (50% v/v ethanol) or either 0.2 or 0.5 µM doxycycline in the same solvent, as indicated, was determined by liquid growth assay as in (C). (E) Growth of IPT1 (BY4741) or ipt1Δ (M24) yeast strains in MM4 medium supplemented with or without acetic acid was measured as in (B). Figure 5 Exogenously-supplied PHS enhances acetic acid tolerance of S. cerevisiae. Serial dilution of otherwise wild-type cells expressing either +93xFLAG-Orm1 and 3xHA-Orm2 (JTY6140) or the 3xFLAG-Orm1(P4) and 3xHA-Orm2(P2) mutants were grown to mid-exponential phase and then serial dilutions were made and spotted onto MM4 agar plates containing the indicated acetic acid and PHS concentrations. Plates were imaged after 3 days of growth at 30 °C. Table 1 Yeast strains used in this study. Strain Genotype Source/reference BY4741 MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 Research Genetics, Inc. BY4742 MATα his3Δ1 leu2Δ0 lys2Δ0 ura3Δ0 Research Genetics, Inc. YDB379 BY4741 Ypk1-3xFLAG::natNT2 J.S. Weissman, Univ. of California, San Francisco yKL4 BY4741 TOR2+::Hygr [40] yKL5z BY4741 Tor2(L2178A)::Hygr [40] JTY6142 BY4741 ypk1Δ::KanMX4 Research Genetics, Inc. YDB146 BY4741 3xFlag-Orm1 [41] yAM159-A BY4741 3xFLAG-Lag1::LEU2 [40] yAM163-A 3xFLAG-Lag1(S23A S24A)::LEU2 [40] JTY6140 BY4741 MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 can1Δ::pSte2-His5 lyp1Δ::pSte3- Leu5 3xFLAG-Orm1 3xHA-Orm2 [41] JTY5215 BY4741 MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 can1Δ::pSte2-His5 lyp1Δ::pSte3- Leu5 3xFLAG-Orm1-P4 3xHA-Orm2-P*2 [41] yAM205-A BY4742 Lac1::LEU2 Lag1::LEU2 [40] yAM207-B BY4742 Lac1(S23A S24A)::LEU2 Lag1(S23A S24A)::LEU2 [40] yAM327 BY4741 Ypk1-3xFLAG hrk1Δ::kanMX4 This study yAM336 BY4741 3xFlag-Orm1 hrk1Δ::kanMX4 This study M24 BY4741 ipt1Δ::kanMX4 Research Genetics, Inc. MTY175 BY4741, tetO7-AUR1::kanMX4 URA3 M. Tani, Kyushu University, Japan Table 2 Plasmids used in this study. Plasmid Description Source/reference pAM20 CEN, LEU2, PYPK1-Ypk1-myc [32] pFR221 CEN, LEU2, PYPK1-Ypk1(T662A)-myc [32] pAX126 CEN, LEU2, PLAC1-Lac1 This study pAX129 CEN, LEU2, PLAC1-Lac1(S23A S24A) This study Summary Statement We document that during yeast adaptation to acetic acid stress, TORC2-Ypk1 signaling is activated, contributing to yeast cell ability to endure this stress. Manipulation of the sphingolipid pathway is proposed as a way to increase yeast fitness in industrial biotechnology. AUTHOR CONTRIBUTIONS J.F.G. and A.M. performed the experimental work, S.R. provided invaluable guidance, A.M., J.F.G., I.S.C. and J.T. designed the experiments, analyzed the data, and wrote the paper. LITERATURE CITED 1 Erickson B Nelson Winters P Perspective on opportunities in industrial biotechnology in renewable chemicals Biotechnology journal 2012 7 176 185 21932250 2 Carroll A Somerville C Cellulosic biofuels Annual review of plant biology 2009 60 165 182 3 Chundawat SP Beckham GT Himmel ME Dale BE Deconstruction of lignocellulosic biomass to fuels and chemicals Annual review of chemical and biomolecular engineering 2011 2 121 145 4 Zha Y Muilwijk B Coulier L Punt PJ Inhibitory Compounds in Lignocellulosic Biomass Hydrolysates during Hydrolysate Fermentation Processes J Bioprocess Biotechniq 2012 2 1 11 5 Franden MA Pienkos PT Zhang M Development of a high-throughput method to evaluate the impact of inhibitory compounds from lignocellulosic hydrolysates on the growth of Zymomonas mobilis Journal of biotechnology 2009 144 259 267 19683550 6 Fleet GH Yeasts in foods and beverages: impact on product quality and safety Current opinion in biotechnology 2007 18 170 175 17275276 7 Mollapour M Piper PW Hog1 mitogen-activated protein kinase phosphorylation targets the yeast Fps1 aquaglyceroporin for endocytosis, thereby rendering cells resistant to acetic acid Molecular and cellular biology 2007 27 6446 6456 17620418 8 Casal M Cardoso H Leão C Mechanisms regulating the transport of acetic acid in Saccharomyces cerevisiae Microbiology 1996 142 1385 1390 8704978 9 Piper P Calderon CO Hatzixanthis K Mollapour M Weak acid adaptation: the stress response that confers yeasts with resistance to organic acid food preservatives Microbiology 2001 147 2635 2642 11577142 10 Mira NP Teixeira MC Microbial mechanisms of tolerance to weak acid stress Frontiers in microbiology 2013 4 416 24416030 11 Arneborg N Jespersen L Jakobsen M Individual cells of Saccharomyces cerevisiae and Zygosaccharomyces bailii exhibit different short-term intracellular pH responses to acetic acid Archives of microbiology 2000 174 125 128 10985752 12 Moon NJ Inhibition of the growth of acid tolerant yeasts by acetate, lactate and propionate and their synergistic mixtures Journal of Applied Microbiology 1983 55 453 460 13 Giannattasio S Guaragnella N Zdralevic M Marra E Molecular mechanisms of Saccharomyces cerevisiae stress adaptation and programmed cell death in response to acetic acid Frontiers in microbiology 2013 4 33 23430312 14 Burtner CR Murakami CJ Kennedy BK Kaeberlein M A molecular mechanism of chronological aging in yeast Cell Cycle 2009 8 1256 1270 19305133 15 Stratford M Nebe-von-Caron G Steels H Novodvorska M Ueckert J Archer DB Weak-acid preservatives: pH and proton movements in the yeast Saccharomyces cerevisiae International journal of food microbiology 2013 161 164 171 23334094 16 Carmelo V Santos H Sá-Correia I Effect of extracellular acidification on the activity of plasma membrane ATPase and on the cytosolic and vacuolar pH of Saccharomyces cerevisiae Biochimica et biophysica acta 1997 1325 63 70 9106483 17 Mira NP Becker JD Sá-Correia I Genomic expression program involving the Haa1p-regulon in Saccharomyces cerevisiae response to acetic acid Omics : a journal of integrative biology 2010 14 587 601 20955010 18 Goossens A de La Fuente N Forment J Serrano R Portillo F Regulation of yeast H(+)-ATPase by protein kinases belonging to a family dedicated to activation of plasma membrane transporters Molecular and cellular biology 2000 20 7654 7661 11003661 19 Mira NP Teixeira MC Sá-Correia I Adaptive response and tolerance to weak acids in Saccharomyces cerevisiae: a genome-wide view Omics : a journal of integrative biology 2010 14 525 540 20955006 20 Mira NP Palma M Guerreiro JF Sá-Correia I Genome-wide identification of Saccharomyces cerevisiae genes required for tolerance to acetic acid Microbial cell factories 2010 9 79 20973990 21 Kawahata M Masaki K Fujii T Iefuji H Yeast genes involved in response to lactic acid and acetic acid: acidic conditions caused by the organic acids in Saccharomyces cerevisiae cultures induce expression of intracellular metal metabolism genes regulated by Aft1p FEMS yeast research 2006 6 924 936 16911514 22 Lee Y Nasution O Choi E Choi IG Kim W Choi W Transcriptome analysis of acetic-acid-treated yeast cells identifies a large set of genes whose overexpression or deletion enhances acetic acid tolerance Applied microbiology and biotechnology 2015 99 6391 6403 26062532 23 Chen YY Stabryla L Wei N Improved Acetic Acid Resistance in Saccharomyces cerevisiae by Overexpression of the WHI2 Gene Identified through Inverse Metabolic Engineering Appl Environ Microb 2016 82 2156 2166 24 Chen YY Sheng JY Jiang T Stevens J Feng XY Wei N Transcriptional profiling reveals molecular basis and novel genetic targets for improved resistance to multiple fermentation inhibitors in Saccharomyces cerevisiae Biotechnol Biofuels 2016 9 25 Inaba T Watanabe D Yoshiyama Y Tanaka K Ogawa J Takagi H Shimoi H Shima J An organic acid-tolerant HAA1-overexpression mutant of an industrial bioethanol strain of Saccharomyces cerevisiae and its application to the production of bioethanol from sugarcane molasses Amb Express 2013 3 26 Sakihama Y Hasunuma T Kondo A Improved ethanol production from xylose in the presence of acetic acid by the overexpression of the HAA1 gene in Saccharomyces cerevisiae Journal of bioscience and bioengineering 2015 119 297 302 25282639 27 Meijnen JP Randazzo P Foulquie-Moreno MR van den Brink J Vandecruys P Stojiljkovic M Dumortier F Zalar P Boekhout T Gunde-Cimerman N Kokosar J Stajdohar M Curk T Petrovic U Thevelein JM Polygenic analysis and targeted improvement of the complex trait of high acetic acid tolerance in the yeast Saccharomyces cerevisiae Biotechnol Biofuels 2016 9 5 26740819 28 Pearce LR Komander D Alessi DR The nuts and bolts of AGC protein kinases Nature reviews. Molecular cell biology 2010 11 9 22 20027184 29 Casamayor A Torrance PD Kobayashi T Thorner J Alessi DR Functional counterparts of mammalian protein kinases PDK1 and SGK in budding yeast Current biology : CB 1999 9 186 197 10074427 30 Roelants FM Torrance PD Bezman N Thorner J Pkh1 and Pkh2 differentially phosphorylate and activate Ypk1 and Ykr2 and define protein kinase modules required for maintenance of cell wall integrity Molecular biology of the cell 2002 13 3005 3028 12221112 31 Kamada Y Fujioka Y Suzuki NN Inagaki F Wullschleger S Loewith R Hall MN Ohsumi Y Tor2 directly phosphorylates the AGC kinase Ypk2 to regulate actin polarization Molecular and cellular biology 2005 25 7239 7248 16055732 32 Roelants FM Breslow DK Muir A Weissman JS Thorner J Protein kinase Ypk1 phosphorylates regulatory proteins Orm1 and Orm2 to control sphingolipid homeostasis in Saccharomyces cerevisiae Proceedings of the National Academy of Sciences of the United States of America 2011 108 19222 19227 22080611 33 Niles BJ Mogri H Hill A Vlahakis A Powers T Plasma membrane recruitment and activation of the AGC kinase Ypk1 is mediated by target of rapamycin complex 2 (TORC2) and its effector proteins Slm1 and Slm2 Proceedings of the National Academy of Sciences of the United States of America 2012 109 1536 1541 22307609 34 Sun Y Miao Y Yamane Y Zhang C Shokat KM Takematsu H Kozutsumi Y Drubin DG Orm protein phosphoregulation mediates transient sphingolipid biosynthesis response to heat stress via the Pkh-Ypk and Cdc55-PP2A pathways Molecular biology of the cell 2012 23 2388 2398 22535525 35 Berchtold D Piccolis M Chiaruttini N Riezman I Riezman H Roux A Walther TC Loewith R Plasma membrane stress induces relocalization of Slm proteins and activation of TORC2 to promote sphingolipid synthesis Nature cell biology 2012 14 542 547 22504275 36 Lee YJ Jeschke GR Roelants FM Thorner J Turk BE Reciprocal phosphorylation of yeast glycerol-3-phosphate dehydrogenases in adaptation to distinct types of stress Molecular and cellular biology 2012 32 4705 4717 22988299 37 Muir A Roelants FM Timmons G Leskoske KL Thorner J Down-regulation of TORC2-Ypk1 signaling promotes MAPK-independent survival under hyperosmotic stress eLif 2015 4 e09336 38 Roelants FM Baltz AG Trott AE Fereres S Thorner J A protein kinase network regulates the function of aminophospholipid flippases Proceedings of the National Academy of Sciences of the United States of America 2010 107 34 39 19966303 39 Rispal D Eltschinger S Stahl M Vaga S Bodenmiller B Abraham Y Filipuzzi I Movva NR Aebersold R Helliwell SB Loewith R Target of rapamycin complex 2 regulates actin polarization and endocytosis via multiple pathways J Biol Chem 2015 40 Muir A Ramachandran S Roelants FM Timmons G Thorner J TORC2-dependent protein kinase Ypk1 phosphorylates ceramide synthase to stimulate synthesis of complex sphingolipids eLife 2014 3 e03779 41 Breslow DK Collins SR Bodenmiller B Aebersold R Simons K Shevchenko A Ejsing CS Weissman JS Orm family proteins mediate sphingolipid homeostasis Nature 2010 463 1048 1053 20182505 42 Schorling S Vallee B Barz WP Riezman H Oesterhelt D Lag1p and Lac1p are essential for the Acyl-CoA-dependent ceramide synthase reaction in Saccharomyces cerevisae Molecular biology of the cell 2001 12 3417 3427 11694577 43 Kliegman JI Fiedler D Ryan CJ Xu YF Su XY Thomas D Caccese MC Cheng A Shales M Rabinowitz JD Krogan NJ Shokat KM Chemical genetics of rapamycin-insensitive TORC2 in S. cerevisiae Cell reports 2013 5 1725 1736 24360963 44 Roelants FM Torrance PD Thorner J Differential roles of PDK1- and PDK2-phosphorylation sites in the yeast AGC kinases Ypk1, Pkc1 and Sch9 Microbiology 2004 150 3289 3304 15470109 45 Rego A Costa M Chaves SR Matmati N Pereira H Sousa MJ Moradas-Ferreira P Hannun YA Costa V Corte-Real M Modulation of mitochondrial outer membrane permeabilization and apoptosis by ceramide metabolism PloS one 2012 7 e48571 23226203 46 Lindberg L Santos AX Riezman H Olsson L Bettiga M Lipidomic profiling of Saccharomyces cerevisiae and Zygosaccharomyces bailii reveals critical changes in lipid composition in response to acetic acid stress PloS one 2013 8 e73936 24023914 47 Hunter T Plowman GD The protein kinases of budding yeast: six score and more Trends in biochemical sciences 1997 22 18 22 48 Shimobayashi M Oppliger W Moes S Jeno P Hall MN TORC1-regulated protein kinase Npr1 phosphorylates Orm to stimulate complex sphingolipid synthesis Molecular biology of the cell 2013 24 870 881 23363605 49 Jin N Mao K Jin Y Tevzadze G Kauffman EJ Park S Bridges D Loewith R Saltiel AR Klionsky DJ Weisman LS Roles for PI(3,5)P2 in nutrient sensing through TORC1 Molecular biology of the cell 2014 25 1171 1185 24478451 50 Wu WI McDonough VM Nickels JT Jr Ko J Fischl AS Vales TR Merrill AH Jr Carman GM Regulation of lipid biosynthesis in Saccharomyces cerevisiae by fumonisin B1 The Journal of biological chemistry 1995 270 13171 13178 7768913 51 Tani M Kuge O Hydroxylation state of fatty acid and long-chain base moieties of sphingolipid determine the sensitivity to growth inhibition due to AUR1 repression in Saccharomyces cerevisiae Biochemical and biophysical research communications 2012 417 673 678 22166213 52 Lindahl L Genheden S Eriksson LA Olsson L Bettiga M Sphingolipids contribute to acetic acid resistance in Zygosaccharomyces bailii Biotechnology and bioengineering 2016 113 744 753 26416641 53 Dudley AM Janse DM Tanay A Shamir R Church GM A global view of pleiotropy and phenotypically derived gene function in yeast Molecular systems biology 2005 1 2005 0001 54 Mao C Xu R Bielawska A Obeid LM Cloning of an alkaline ceramidase from Saccharomyces cerevisiae. An enzyme with reverse (CoA-independent) ceramide synthase activity The Journal of biological chemistry 2000 275 6876 6884 10702247 55 Mollapour M Piper PW Hog1p mitogen-activated protein kinase determines acetic acid resistance in Saccharomyces cerevisiae FEMS yeast research 2006 6 1274 1280 17156024 56 Dickson RC Thematic review series: sphingolipids. New insights into sphingolipid metabolism and function in budding yeast Journal of lipid research 2008 49 909 921 18296751 57 Hannich JT Umebayashi K Riezman H Distribution and functions of sterols and sphingolipids Cold Spring Harbor perspectives in biology 2011 3 58 Proszynski TJ Klemm RW Gravert M Hsu PP Gloor Y Wagner J Kozak K Grabner H Walzer K Bagnat M Simons K Walch-Solimena C A genome-wide visual screen reveals a role for sphingolipids and ergosterol in cell surface delivery in yeast Proceedings of the National Academy of Sciences of the United States of America 2005 102 17981 17986 16330752 59 Takabatake A Kawazoe N Izawa S Plasma membrane proteins Yro2 and Mrh1 are required for acetic acid tolerance in Saccharomyces cerevisiae Applied microbiology and biotechnology 2015 99 2805 2814 25503505 60 Lam MH Snider J Rehal M Wong V Aboualizadeh F Drecun L Wong O Jubran B Li M Ali M Jessulat M Deineko V Miller R Lee M Park HO Davidson A Babu M Stagljar I A Comprehensive Membrane Interactome Mapping of Sho1p Reveals Fps1p as a Novel Key Player in the Regulation of the HOG Pathway in S. cerevisiae Journal of molecular biology 2015 427 2088 2103 25644660 61 Berchtold D Walther TC TORC2 plasma membrane localization is essential for cell viability and restricted to a distinct domain Molecular biology of the cell 2009 20 1565 1575 19144819 62 Davenport KR Sohaskey M Kamada Y Levin DE Gustin MC A second osmosensing signal transduction pathway in yeast. Hypotonic shock activates the PKC1 protein kinase-regulated cell integrity pathway The Journal of biological chemistry 1995 270 30157 30161 8530423 63 Ludovico P Rodrigues F Almeida A Silva MT Barrientos A Côrte-Real M Cytochrome c release and mitochondria involvement in programmed cell death induced by acetic acid in Saccharomyces cerevisiae Molecular biology of the cell 2002 13 2598 2606 12181332 64 Semchyshyn HM Abrat OB Miedzobrodzki J Inoue Y Lushchak VI Acetate but not propionate induces oxidative stress in bakers' yeast Saccharomyces cerevisiae Redox report : communications in free radical research 2011 16 15 23 21605494 65 Niles BJ Joslin AC Fresques T Powers T TOR complex 2-Ypk1 signaling maintains sphingolipid homeostasis by sensing and regulating ROS accumulation Cell reports 2014 6 541 552 24462291 66 Zhang JG Liu XY He XP Guo XN Lu Y Zhang BR Improvement of acetic acid tolerance and fermentation performance of Saccharomyces cerevisiae by disruption of the FPS1 aquaglyceroporin gene Biotechnology letters 2011 33 277 284 20953665 67 Zhang MM Zhao XQ Cheng C Bai FW Improved growth and ethanol fermentation of Saccharomyces cerevisiae in the presence of acetic acid by overexpression of SET5 and PPR1 Biotechnology journal 2015 10 1903 1911 26479519 68 dos Santos SC Sa-Correia I Yeast toxicogenomics: lessons from a eukaryotic cell model and cell factory Current opinion in biotechnology 2015 33 183 191 25812478 69 Burke D Amberg D Strathern J Methods in Yeast Genetics: A Cold Spring Harbor Laboratory Course Manual Cold Spring Harbor Laboratory Press 2005 70 Green MR Sambrook J Molecular cloning : a laboratory manual 2012 Cold Spring Harbor, N.Y. Cold Spring Harbor Laboratory Press 71 Westfall PJ Patterson JC Chen RE Thorner J Stress resistance and signal fidelity independent of nuclear MAPK function Proceedings of the National Academy of Sciences of the United States of America 2008 105 12212 12217 18719124 72 Kinoshita E Kinoshita-Kikuta E Koike T Advances in Phos-tag-based methodologies for separation and detection of the phosphoproteome Biochimica et biophysica acta 2015 1854 601 608 25315852
PMC005xxxxxx/PMC5124426.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 9204265 8516 Lupus Lupus Lupus 0961-2033 1477-0962 27230555 5124426 10.1177/0961203316651738 NIHMS784984 Article Artesunate inhibits type I interferon induced production of macrophage migration inhibitory factor in patients with systemic lupus erythematosus Feng Xuebing 1+ Chen Weiwei 1+ Xiao Lihui 1 Gu Fei 1 Huang Jing 1 Tsao Betty P. 2 Sun Lingyun 1 1 Xuebing Feng, MD & PhD, Weiwei Chen, MS, Lihui Xiao, MD, Fei Gu, PhD, Jing Huang, MD, Lingyun Sun, MD & PhD: Department of Rheumatology and Immunology, The Affiliated Nanjing Drum Tower Hospital of Nanjing University Medical School, Nanjing, 200008, China 2 Betty P. Tsao, PhD: Division of Rheumatology and Immunology, Medical University of South Carolina, Charleston, SC 29425-8900, USA Corresponding authors: Xuebing Feng, MD & PhD, and Lingyun Sun, MD & PhD., Department of Rheumatology and Immunology, The Affiliated Drum Tower Hospital of Nanjing University Medical School, 321 Zhong Shan Road, Nanjing, Jiangsu 210008, China. Telephone: 0086-25-68182422. Fax number: 0086-25-68182428. [email protected] or [email protected] + These two authors contributed equally to this article. 11 5 2016 26 5 2016 1 2017 01 1 2018 26 1 6272 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Objective Macrophage migration inhibitory factor (MIF) is a key regulator of both atherosclerosis and systemic lupus erythematosus (SLE), yet factors leading to its overproduction remain unclear. To explore regulation of MIF in SLE, we studied effects and potential mechanisms of type I interferon (IFN) and artesunate (ART), an anti-malarial agent extracted from Chinese herbs, on levels of MIF. Methods Serum and peripheral blood cells from SLE patients and healthy controls were measured for MIF levels by ELISA and type I IFN inducible gene expressions by real-time PCR respectively, and assessed for associations by Spearman correlation. ART was added to human umbilical vein endothelial cells (HUVECs) cultures with or without prior IFNα-1b stimulation and to SLE peripheral blood mononuclear cells (PBMC) cultures. Protein levels of STATs and phosphorylated (p-) STATs in HUVECs were determined by Western blotting. Results Serum MIF levels were elevated in SLE patients and positively associated with disease activity (r = 0.86, p < 0.0001), accumulated damage (r = 0.34, p < 0.05), and IFN scores in SLE PBMCs (r = 0.74, p = 0.0002). The addition of IFNα-1b promoted MIF production in a time and dose-dependent manner in HUVEC cultures. ART could inhibit expressions of IFN inducible genes (LY6E and ISG15) in both HUVEC and SLE PBMC cultures, and suppress MIF production and over-expression of p-STAT1, but not p-STAT3 or STAT5, induced by IFNα-1b stimulation. IFNγ-induced expression of p-STAT1 in HUVECs was not inhibited by ART. Conclusion MIF could be regulated by type I IFN in SLE patients. ART counteracts the effect of IFNα to inhibit MIF production by blocking STAT1 phosphorylation and thus may 3 have therapeutic potential for SLE-associated atherosclerosis. Systemic lupus erythematosus artesunate macrophage migration inhibitory factor interferon signal transducer and activator of transcription 1 Introduction Systemic lupus erythematosus (SLE) is a chronic multisystem autoimmune disease characterized by the production of autoantibodies and end organ damage from immune complex deposition (1). In 1976, Urowitz et al. noted premature myocardial infarction among patients with SLE (2). Since this pioneering discovery, many follow-up studies have demonstrated that the incidence of coronary artery disease in women with SLE is much higher than age-matched general population, which could not be fully accounted for by traditional risk factors (3). Patients with SLE-associated atherosclerosis have an earlier occurrence, a longer duration of lupus and a higher damage-index score (4;5). However, the molecular mechanisms of accelerated atherosclerosis in SLE remain to be elucidated. Macrophage migration inhibitory factor (MIF) is released upon pro-inflammatory activation and exerts an upstream role in regulating the innate immune response. High-expression MIF alleles are associated with disease risk in SLE (6), while anti-MIF treatment reduces functional and histological indices of glomerulonephritis in different strains of lupus-prone mice (7), indicating that MIF is a key regulator for developing disease manifestations.. Of note, a growing body of literature has suggested that MIF also plays an important role in the pathogenesis of atherosclerosis. MIF expression within atheromatous plaques is closely associated with progression and instability in human disease and inhibition of MIF activity prevents atherosclerosis in animal models (8;9). Thus, MIF may be a common mediator for accelerated atherosclerosis in SLE. At present, it is unclear how MIF is modulated in SLE patients and there is a lack of preventive strategies targeting this process. Over the past decade, interferon-alpha (IFNα) or type I IFN has been identified to play a central role in the pathogenesis of SLE (10;11), and an association between type I IFN, vascular damage and progression of atherosclerosis has emerged (12). Artesunate (ART) is a semi-synthetic derivative of a Chinese herb named artemisinin that is commonly used as an anti-malarial agent in treating Chinese patients. Previously we have demonstrated that ART could inhibit the progression of lupus disease and reverse the pathologic lesion of nephritis in the MRL/lpr mouse model (13). Since ART could also exert an anti-angiogenic effect by decreasing the levels of vasoactive factors and inducing endothelial cell apoptosis, it is hypothesized that this drug may work on the regulation of atherosclerosis (14). In the current study, we showed for the first time that MIF level was associated with type I IFN gene expression signature in SLE patients. ART treatment down-regulated IFNα induced MIF production through the inhibition of phosphorylation of signal transducer and activator of transcription 1 (STAT1), and herein may have a potential role in ameliorating SLE-associated atherosclerosis. Materials and methods Patients and controls The demographics of SLE patients and normal controls were shown in Table 1. Thirty-five SLE in-patients (35.2±12.8 years old, including 34 female and 1 male) were enrolled in the study and all of them fulfilled the classification criteria of SLE proposed by the American College of Rheumatology (15). SLE disease activity was assessed by the Systemic Lupus Erythematosus Disease Activity Index (16), and damage was measured by Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index (SLICC) scores (17). The average dose of corticosteroids was 32.6 ± 24.0 mg/day (counted as the equivalent dose to methyl- prednisolone) at the time of blood draw. Twenty-one healthy donors (34.5±8.5 years old, 19 female and 2 male) were recruited as normal controls. The study was approved by the Ethics Committee at The Affiliated Drum Tower Hospital of Nanjing University Medical School. Informed consent was obtained from each patient and healthy donor. Cell culture Human umbilical vein endothelial cells (HUVECs) were purchased from Shanghai Fumeng Gene Biologicals (Shanghai, China). Cells were resuspended at 3 × 105 per well (1.5 × 105/ml) in DMEM (Gibco) supplemented with 10% fetal bovine serum (FBS) (Gibco) with or without the presence of different concentrations of IFNα-1b (200 U/ml, 1000 U/ml, 2000 U/ml) and incubated at 37 °C with 5 % CO2 for 6, 12 or 24 hours. Then the culture medium was replaced with a medium containing ART at the concentration of 5 or 20 µmol/L, with dexamethasone sodium phosphate (100 ng/ml, 0.19 µM) or fludarabine (50 µM) applied as a positive control. 12 and 24 hours later, cells and supernatants were harvested for further measurements. To confirm the results from HUVECs, peripheral blood mononuclear cells (PBMCs) from SLE patients were cultured with ART (5 or 20 µM) or dexamethasone sodium phosphate (100 ng/ml) at 3 × 105 per well for 24 hours. Gene expression measuring A 2 to 3 ml blood sample or 1 × 105 HUVECs was collected, and total RNA was extracted immediately using Trizol reagent (Invitrogen, USA). RNA was reverse-transcribed and quantified by real-time polymerase chain reaction (Q-PCR) using the PrimeScript RT-PCR and SYBR® Premix Ex Taq™ kit (TaKaRa Biotechnology) for the detection of type I IFN inducible gene (MX1, OASL, OAS1, ISG15 and LY6E) expression in triplicates in the ABI 7500 FAST real-time PCR detection system (Applied Biosystems, USA). All primers were purchased from TaKaRa (Dalian, China) and the sequences were the same as previously described (18). Human ribosomal protein, large, P0 (RPLP0) was used as the house-keeping gene to normalize cellular RNA amounts. After the real-time PCR procedure, mean CT value of a target gene was obtained for each sample. Relative gene expression values were presented as 2−ΔΔCT, and IFN scores were calculated using MX1, OASL, OAS1, ISG15 and LY6E mRNA expression as previously described (19). Because these 5 genes exhibited highly correlated expressions among each other (19), in in vitro studies two of them (LY6E and ISG15) were measured to represent the status of IFN inducible gene expressions. Protein analysis HUVECs were harvested and resuspended in lysis buffer for 30 minutes, and then centrifuged at 14,000 g for 10 minutes at 4°C. After centrifugation, cell lysates were subjected to 10% sodium dodecylsulfate -polyacrylamide gel electrophoresis and transferred onto polyvinylidene difluoride membranes (Millipore, USA). After blocking with 5% skim milk, the membranes were incubated with primary antibodies at 4°C overnight, including antibodies against phosphorylated (p-) STAT1, STAT1, p-STAT3, STAT3, p-STAT5, STAT5 and GAPDH (Cell Signaling Technology). After washing with Tris-buffered saline with Tween 20, the membranes were incubated with horseradish peroxidase conjugated secondary antibody (1:5000–1:10000) for 30 minutes and the bands were visualized in a luminol-based detection system with enhanced chemiluminescence (Millipore, USA). Anti-GAPDH antibody (Cell Signaling Technology) was used as a loading control. MIF levels in serum or cultured supernatants were determined by enzyme-linked immunosorbent assay (ELISA) (R&D, USA) with samples diluted in 1: 10 or 1: 20. Statistical analysis All data were analyzed using Prism version 6.0 software (GraphPad, San Diego, CA). Comparisons between two groups were made by unpaired t test or Mann-Whitney tests when the data were not normally distributed, and results were correspondingly shown as mean ± SD or median (interquartile range, IQRs). Spearman correlation was used to depict relationships between two factors. To determine whether the parameters were significantly altered after the treatment of ART in vitro, one-way analysis of variance (ANOVA) was applied, followed by Dunnett's multiple comparisons test. A p < 0.05 was considered to be statistically significant. Results Increased serum MIF levels in SLE patients Compared to healthy controls, serum MIF levels were significantly elevated in SLE patients [18.7 (IQRs 12.5, 29.9) ng/ml vs. 1.8 (1.0, 2.0) ng/ml, p < 0.0001] (Figure 1A). Positive correlations were detected between MIF levels and SLEDAI scores (r = 0.86, p < 0.0001), SLICC scores (r = 0.34, p = 0.048), as well as the dose of methylprednisolone at the time of blood draw (r = 0.45, p = 0.016) (Figure 1B, C, and D; respectively), implying that MIF is an important contributor to both disease activity and accumulated damage in SLE, which is reflected in the treatment dosage of steroids. The disturbance of MIF was related to aberrant IFN inducible gene expression in SLE patients To explore potential factors that participated in upregulation of MIF in SLE, expressions of 5 type I IFN inducible genes (MX1, OASL, OAS1, ISG15 and LY6E) in PBMCs from SLE patients and healthy controls were measured. Q-PCR analysis demonstrated that significantly higher expressions of all 5 IFN inducible genes were present in patients with SLE (p all < 0.001, except for OAS1 where p < 0.05) (Figure 2A–E). As shown in Figure 2F, IFN scores that calculated according to the expression levels of 5 IFN inducible genes were positively correlated to SLEDAI scores (r = 0.69, p = 0.0007). The increase of IFN scores was robustly associated with the elevation of MIF levels (r = 0.74, p = 0.0002) (Figure 2G), supporting a link between circulating MIF levels of SLE patients and elevated levels of type I IFN inducible gene expression in SLE PBMC. IFNα could directly modulate MIF levels in vitro To investigate whether type I IFN did have an effect on MIF regulation, a human endothelial cell line, HUVECs, was cultured with or without the addition of IFNα-1b. As expected, HUVECs expressed higher mRNA levels of type I IFN inducible genes (as represented by LY6E and ISG15) after IFNα stimulation (p < 0.05) (see the control group in Figure 3). MIF levels were significantly elevated at 24 hours (31.8 ± 4.5 ng/ml vs. 22.4 ± 1.8 ng/ml, p < 0.01) but not at 6 hours or 12 hours after IFNα treatment (Figure 4A). After stimulated by different concentrations of IFNα for 24 hours, MIF levels were increased in all three groups compared with that in control group and the effect was most significantly different at the highest dose in the 2000 U/ml group (60.6 ± 11.0 ng/ml vs. 22.4 ± 1.8 ng/ml, p < 0.001) (Figure 4B), indicating that type I IFN is capable of promoting MIF production in a time- and dose-dependent manner. ART down-regulated type I IFN inducible gene expression and MIF production To explore the role of ART on type I IFN and MIF, HUVECs were exposed to IFNα-1b and then treated by ART (5 and 20 µM) or dexamethasone (as a positive control). As shown in Figure 3, there was a dramatic decline of IFN inducible gene (LY6E and ISG15) expressions after ART and dexamethasone treatment, either with or without prior stimulation of IFNα (1000U/ml), indicating that ART could inhibit the activation of IFN pathway. After culturing with ART for 24 hours, a significant decline of MIF levels was observed (24.5 ± 0.5 ng/ml in ART 5 µM group, 24.4 ± 0.3 ng/ml in ART 20 µM group, all p < 0.001 compared to 29.4 ± 0.6 ng/ml in IFNα control group) (Figure 4D), suggesting that ART could counteract IFNα triggered MIF production. To further confirm the effect of ART, we assessed MIF production in SLE PBMC cultures in the presence or absence of ART at a dose of 20 µM. Similar to that in HUVECs, ART significantly decreased the mRNA expressions of LY6E and ISG15 (both p < 0.01) (Figure 5A–B) and down-regulated MIF levels (13.5 ± 3.5 ng/ml vs. 50.9 ± 6.5 ng/ml, p < 0.001) (Figure 5C). Thus, ART could not only act on endothelial cells, but also act on PBMCs to inhibit MIF production. ART inhibited MIF production through the blockade of STAT1 phosphorylation Given that the biological activities of IFNs are triggered by the STAT signaling cascade (20), we studied the activation of STATs after ART treatment. Western blotting demonstrated that IFNα-1b promoted tyrosine phosphorylation of SATA1 but not STAT3 or STAT5 in cultured HUVECs (Figure 6A). When treated by ART at a dose of 20 µM for 24 hours, IFNα-1b induced elevation of p-SATA1 was greatly down-regulated (Figure 6B). Consistently, MIF levels in cultured supernatants were also increased after IFNα-1b stimulation, while ART had a similar effect as that of fludarabine, a specific STAT1 inhibitor, in attenuating IFNα induced MIF production (Figure 6C). To rule out the effect of type II IFN, HUVECs were stimulated with 10ng/ml IFNγ for 24 hours and cultured with ART for another 24 hours. As shown in Figure 6D, ART treatment did not reverse IFNγ-induced STAT1 phosphorylation. Discussion Individuals with SLE have a striking increase in the incidence of premature atherosclerosis compared to their sex- and age-matched controls, which has been recognized as one of the leading cause of morbidity and mortality for this prototypical autoimmune disease (21). Recently, activation of type I IFN system has been shown to contribute, at least partially, to accelerated atherosclerosis in SLE patients (22), yet the mechanisms remain to be elucidated. In this study, we confirmed that MIF, a chemokine -like inflammatory regulator related to either atherosclerosis or lupus, was highly expressed in SLE patients, especially those with active diseases. Our data demonstrated for the first time that elevated type I IFN could lead to the abnormal production of MIF, while ART, an anti-malarial agent extracted from Chinese herbs, counteracted the effect of type I IFN by blocking STAT1 phosphorylation. Based on these data, we propose that ART may serve as a novel and feasible therapy for both SLE and SLE-associated arthrosclerosis. Our data also demonstrated that MIF was highly expressed in SLE patients and associated with both disease activity and accumulated organ damage. MIF, first described as a T-cell cytokine important for inhibiting macrophage migration, is now known to be secreted by both immune and nonimmune cell types through the autocrine or paracrine modes of action and to have diverse immunologic functions (23). There is an impressive body of evidence supporting a key role for MIF in the pathogenesis of SLE, as the circulating level and the tissue expression of MIF are elevated in SLE patients, and the MIF high-expression alleles are associated with increased prevalence of SLE (6;24). In addition, lupus-prone mice deficient in MIF have been reported to exhibit prolonged survival and marked reductions in renal and skin lesions despite no effects on T and B cells, while targeting the MIF–MIF receptor interaction by small-molecule antagonism down-regulates MIF-dependent pathways of tissue damage (7;25). MIF may directly accelerate atherosclerosis, since it is highly expressed in different stages of human atherosclerosis (26), and MIF inhibition induces the stabilization and even regression of atherosclerotic plaques in different experimental models (8;27;28). Thus, the elevation of MIF in SLE patients could not only exacerbate lupus diseases but also promote atherosclerosis. Our next question is why MIF levels fluctuated in SLE. In the past decade, IFNα has been proposed to play an important role in SLE etiology and pathogenesis. There are many studies including ours demonstrating that IFNα levels or type I IFN inducible gene expressions increase during SLE flares and multiple organ involvement (19;29). Recently, a link between type I IFN and premature atherosclerosis in SLE has been established by the data showing that IFNα priming promotes lipid uptake and macrophage-derived foam cell formation (12). In this study, we showed for the first time that MIF levels were related to increased IFN inducible gene expression in SLE patients. In vitro treatment of IFNα confirmed the relationship between type I IFN and MIF, implying that over-expressed type I IFN may promote atherosclerosis in SLE through the modulation of MIF. Currently, the available therapy for atherosclerosis primarily targets hyperlipidemia and prevention of thrombosis, even it has long been identified as having inflammatory components contributing to its pathogenesis. Unfortunately, routine statin use over 3 years has no significant effect on subclinical atherosclerosis progression in young SLE patients (30). With the exception of mycophenolate mofetil and hydroxychloroquine (31;32), the availability of beneficial treatments to treat SLE-associated atherosclerosis is greatly limited. Since MIF plays an important role in the progression of atherosclerosis, the design of new inhibitors targeting MIF and the study of their effects on the biologic functions of this factor will be extremely valuable and promising for treatment of lupus atherosclerosis (23). Artemisinins are a family of sesquiterpene trioxane lactone anti-malarial agents originally derived from Artemisia annua L., among which ART is the most studied analog. With its established safety record in millions of malarial patients, recently ART has also been investigated in other diseases such as virus or fungal infections, cancers and autoimmune diseases (33). Consistent with our previous findings (13), this kind of drug significantly ameliorates lupus-like disease in experimental murine models of SLE, with comparable or even superior efficacy to that of GC or immunosuppressant (34;35). The anti-inflammatory effects of ART could be attributed to the inhibition of various cytokines and diverse signaling pathways (36;37). Here we showed that ART directly inhibited the expression of IFN inducible genes and counteracted the effect of IFNα on MIF production, suggesting that ART act through the modulation of type I IFN pathway. Upon IFN-induced activation, the STAT proteins are phosphorylated and translocated into the nucleus to bind specific elements within the promoters of IFN-stimulated genes (20). Treatment of an artemisinin analog has been found to significantly suppress the phosphorylation of STAT-1, STAT-3, and STAT-5 but not STAT-4 and STAT-6 (35). However, only the phosphorylation of STAT-1 was found to be down-regulated by ART in this study, which was compatible to that of fludarabine, a selective STAT1 inhibitor. In summary, here we showed that the production of MIF was up-regulated by type I IFN in vitro. The IFNα-MIF axis could represent a new pathway for the development of arthrosclerosis in SLE. ART acted through the inhibition of STAT1 phosphorylation to counteract the effect of type I IFN, leading to the decline of MIF level, therefore may be helpful to relieve both lupus and atherosclerosis in SLE patients. We thank all the patients and healthy volunteers participating in this study. Funding This work was supported by National Natural Science Foundation of China (grant number 81373198), Jiangsu Provincial Special Program of Medical Science (BE2015602) and Jiangsu Province’s Key Provincial Talents Program. BPT was supported by RO1AR0 43814 and R21AR065626 from the NIAMS NIH. Figure 1 Serum MIF levels in SLE patients. A. Elevated MIF production in SLE patients compared with healthy individuals. Each symbol represents an individual sample. B. Associations of MIF levels with SLEDAI scores (r = 0.86, p < 0.001). C. Associations of MIF levels with SLICC scores (r = 0.34, p = 0.048). D. Associations of MIF levels with MP doses (r = 0.45, p = 0.016) Figure 2 Elevated IFN inducible genes expressions were related to both disease activities and MIF levels in SLE patients. A-E. Significantly higher expressions of 5 type I IFN inducible genes in SLE patients than in healthy donors. Each symbol represents an individual sample. p < 0.05, *p < 0.001. F. Association of SLEDAI scores with IFN scores (r = 0.69, p < 0.001). G. Associations of MIF levels with IFN scores (r = 0.74, p < 0.001). Figure 3 Down-regulated IFN inducible gene expression after ART treatment. HUVECs were cultured with or without the presence of IFN-1b for 24 hours, and then treated with ART or dexamethasone (DEX) for 24 hours. Similar to that of DEX, ART at both 5 µM and 20 µM significantly inhibited both natural and IFNα induced LY6E (A, C) and ISG15 (B, D) expressions in HUVEC. Data were present as mean ± SD. p < 0.05, **p < 0.001. Figure 4 ART counteracted IFNα triggered MIF production in HUVECs. HUVECs were stimulated by 1000 U/ml IFNα-1b for 6 hours, 12 hours and 24 hours (A), and then cultured with different concentrations of IFN-1b (200 U/ml, 1000 U/ml, 2000 U/ml) for 24 hours (B) to detect supernatant MIF levels. Next, HUVECs were cultured with the presence of 1000 U/ml IFN-1b for 24 hours, and treated with ART or dexamethasone (DEX) for another 12 and 24 hours (C, D). Data were present as mean ± SD. p < 0.05, p < 0.01, p < 0.001. Figure 5 ART inhibited MIF production in PBMC from SLE patients. A-B, compared with that in control group, ART at 20 µM significantly decreased the mRNA expressions of LY6E and ISG15. C, ART treatment significantly down-regulated the production of MIF levels in PBMC from SLE patients. p < 0.01, *p < 0.001 vs. the control group. Figure 6 ART down-regulated MIF production through the inhibition of STAT1 phosphorylation. A, Protein levels of STAT1, 3, 5 and p-STAT1, 3, 5 in HUVECs with or without IFNα stimulation. B, Effect of ART treatment on IFNα induced pSTAT1 production. HUVECs were cultured with or without the presence of 1000U/ml IFNα for 24 hours and then those with IFNα stimulation were treated by 20 µM ART for 24 hours. C, Similar to that of fludarabine, a selective STAT1 inhibitor, ART treatment significantly down-regulated IFNα induced MIF production. D. Effect of ART treatment on IFNγ induced pSTAT1 production. HUVECs were cultured with or without the presence of 10ng/ml IFNγ for 24 hours and then those with IFNγ stimulation were treated by 20 µM ART for 24 hours. Each lane represented an independent sample and data were also shown as mean ± SD. p < 0.05, p < 0.01, *p < 0.001. Table 1 Demographics of SLE patients and normal controls SLE patients (n = 35) Normal controls (n = 21) Age, years 35.2 ± 12.8 34.5 ± 8.5 Female (n) 34 19 Reasons for admission (n) Infection 4 Lupus nephritis 19 NPSLE+ 3 Hematologic involvement 5 Pulmonary involvement 3 Framingham risk score$≥ 0.01 (n) 16 SLEDAI score 8.91 ± 4.05 SLICC score 1.94 ± 1.63 Except where indicated otherwise, values were presented as mean ± SD. + NPSLE: Neuropsychiatric systemic lupus erythematosus. $ Framingham risk score was calculated according to the literature (38). Conflict of Interest Statement The authors declare that there are no competing interests. Reference List 1 Tsokos GC Systemic lupus erythematosus N Engl J Med 2011 365 22 2110 2121 22129255 2 Urowitz MB Bookman AA Koehler BE Gordon DA Smythe HA Ogryzlo MA The bimodal mortality pattern of systemic lupus erythematosus Am J Med 1976 60 2 221 225 1251849 3 Esdaile JM Abrahamowicz M Grodzicky T Li Y Panaritis C du BR Traditional Framingham risk factors fail to fully account for accelerated atherosclerosis in systemic lupus erythematosus Arthritis Rheum 2001 44 10 2331 2337 11665973 4 Roman MJ Shanker BA Davis A Lockshin MD Sammaritano L Simantov R Prevalence and correlates of accelerated atherosclerosis in systemic lupus erythematosus N Engl J Med 2003 349 25 2399 2406 14681505 5 Asanuma Y Oeser A Shintani AK Turner E Olsen N Fazio S Premature coronary-artery atherosclerosis in systemic lupus erythematosus N Engl J Med 2003 349 25 2407 2415 14681506 6 Sanchez E Gomez LM Lopez-Nevot MA Gonzalez-Gay MA Sabio JM Ortego-Centeno N Evidence of association of macrophage migration inhibitory factor gene polymorphisms with systemic lupus erythematosus Genes Immun 2006 7 5 433 436 16724072 7 Leng L Chen L Fan J Greven D Arjona A Du X A small-molecule macrophage migration inhibitory factor antagonist protects against glomerulonephritis in lupus-prone NZB/NZW F1 and MRL/lpr mice J Immunol 2011 186 1 527 538 21106847 8 Chen D Xia M Hayford C Tham e Semik V Hurst S Expression of human tissue factor pathway inhibitor on vascular smooth muscle cells inhibits secretion of macrophage migration inhibitory factor and attenuates atherosclerosis in ApoE−/− mice Circulation 2015 131 15 1350 1360 25677604 9 Sun H Zhang X Zhao L Zhen X Huang S Wang S Attenuation of atherosclerotic lesions in diabetic apolipoprotein E-deficient mice using gene silencing of macrophage migration inhibitory factor J Cell Mol Med 2015 19 4 836 849 25661015 10 Baechler EC Batliwalla FM Karypis G Gaffney PM Ortmann WA Espe KJ Interferon-inducible gene expression signature in peripheral blood cells of patients with severe lupus Proc Natl Acad Sci U S A 2003 100 5 2610 2615 12604793 11 Bennett L Palucka AK Arce E Cantrell V Borvak J Banchereau J Interferon and granulopoiesis signatures in systemic lupus erythematosus blood J Exp Med 2003 197 6 711 723 12642603 12 Li J Fu Q Cui H Qu B Pan W Shen N Interferon-alpha priming promotes lipid uptake and macrophage-derived foam cell formation: a novel link between interferon-alpha and atherosclerosis in lupus Arthritis Rheum 2011 63 2 492 502 21280004 13 Jin O Zhang H Gu Z Zhao S Xu T Zhou K A pilot study of the therapeutic efficacy and mechanism of artesunate in the MRL/lpr murine model of systemic lupus erythematosus Cell Mol Immunol 2009 6 6 461 467 20003822 14 Chen H Shi L Yang X Li S Guo X Pan L Artesunate inhibiting angiogenesis induced by human myeloma RPMI8226 cells Int J Hematol 2010 92 4 587 597 20945119 15 Hochberg MC Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus Arthritis Rheum 1997 40 9 1725 16 Bombardier C Gladman DD Urowitz MB Caron D Chang CH Derivation of the SLEDAI. A disease activity index for lupus patients. The Committee on Prognosis Studies in SLE Arthritis Rheum 1992 35 6 630 640 1599520 17 Gladman DD Goldsmith CH Urowitz MB Bacon P Fortin P Ginzler E The Systemic Lupus International Collaborating Clinics/American College of Rheumatology (SLICC/ACR) Damage Index for Systemic Lupus Erythematosus International Comparison J Rheumatol 2000 27 2 373 376 10685799 18 Feng X Huang J Liu Y Xiao L Wang D Hua B Identification of interferon-inducible genes as diagnostic biomarker for systemic lupus erythematosus Clin Rheumatol 2015 34 1 71 79 25344775 19 Feng X Wu H Grossman JM Hanvivadhanakul P FitzGerald JD Park GS Association of increased interferon-inducible gene expression with disease activity and lupus nephritis in patients with systemic lupus erythematosus Arthritis Rheum 2006 54 9 2951 2962 16947629 20 Brierley MM Fish EN Stats: multifaceted regulators of transcription J Interferon Cytokine Res 2005 25 12 733 744 16375601 21 Lateef A Petri M Unmet medical needs in systemic lupus erythematosus Arthritis Res Ther 2012 14 Suppl 4 S4 23281889 22 Karageorgas TP Tseronis DD Mavragani CP Activation of type I interferon pathway in systemic lupus erythematosus: association with distinct clinical phenotypes J Biomed Biotechnol 2011 2011 273907 22162633 23 Santos LL Morand EF Macrophage migration inhibitory factor: a key cytokine in RA, SLE and atherosclerosis Clin Chim Acta 2009 399 1–2 1 7 18838066 24 Foote A Briganti EM Kipen Y Santos L Leech M Morand EF Macrophage migration inhibitory factor in systemic lupus erythematosus J Rheumatol 2004 31 2 268 273 14760795 25 Hoi AY Hickey MJ Hall P Yamana J O'Sullivan KM Santos LL Macrophage migration inhibitory factor deficiency attenuates macrophage recruitment, glomerulonephritis, and lethality in MRL/lpr mice J Immunol 2006 177 8 5687 5696 17015758 26 Burger-Kentischer A Goebel H Seiler R Fraedrich G Schaefer HE Dimmeler S Expression of macrophage migration inhibitory factor in different stages of human atherosclerosis Circulation 2002 105 13 1561 1566 11927523 27 Noels H Bernhagen J Weber C Macrophage migration inhibitory factor: a noncanonical chemokine important in atherosclerosis Trends Cardiovasc Med 2009 19 3 76 86 19679264 28 Schober A Bernhagen J Thiele M Zeiffer U Knarren S Roller M Stabilization of atherosclerotic plaques by blockade of macrophage migration inhibitory factor after vascular injury in apolipoprotein E-deficient mice Circulation 2004 109 3 380 385 14691040 29 Niewold TB Clark DN Salloum R Poole BD Interferon alpha in systemic lupus erythematosus J Biomed Biotechnol 2010 2010 948364 20652065 30 Schanberg LE Sandborg C Barnhart HX Ardoin SP Yow E Evans GW Use of atorvastatin in systemic lupus erythematosus in children and adolescents Arthritis Rheum 2012 64 1 285 296 22031171 31 Richez C Richards RJ Duffau P Weitzner Z Andry CD Rifkin IR The effect of mycophenolate mofetil on disease development in the gld.apoE (−/−) mouse model of accelerated atherosclerosis and systemic lupus erythematosus PLoS One 2013 8 4 e61042 23577189 32 Jung H Bobba R Su J Shariati-Sarabi Z Gladman DD Urowitz M The protective effect of antimalarial drugs on thrombovascular events in systemic lupus erythematosus Arthritis Rheum 2010 62 3 863 868 20131232 33 Ho WE Peh HY Chan TK Wong WS Artemisinins: pharmacological actions beyond anti-malarial Pharmacol Ther 2014 142 1 126 139 24316259 34 Hou LF He SJ Li X Wan CP Yang Y Zhang XH SM934 treated lupus-prone NZB × NZW F1 mice by enhancing macrophage interleukin-10 production and suppressing pathogenic T cell development PLoS One 2012 7 2 e32424 22389703 35 Hou LF He SJ Li X Yang Y He PL Zhou Y Oral administration of artemisinin analog SM934 ameliorates lupus syndromes in MRL/lpr mice by inhibiting Th1 and Th17 cell responses Arthritis Rheum 2011 63 8 2445 2455 21484768 36 Li Y Wang S Wang Y Zhou C Chen G Shen W Inhibitory effect of the antimalarial agent artesunate on collagen-induced arthritis in rats through nuclear factor kappa B and mitogen-activated protein kinase signaling pathway Transl Res 2013 161 2 89 98 22749778 37 Cheng C Ho WE Goh FY Guan SP Kong LR Lai WQ Anti-malarial drug artesunate attenuates experimental allergic asthma via inhibition of the phosphoinositide 3-kinase/Akt pathway PLoS One 2011 6 6 e20932 21695271 38 Lloyd-Jones DM Wilson PW Larson MG Beiser A Leip EP D'Agostino RB Framingham risk score and prediction of lifetime risk for coronary heart disease Am J Cardiol 2004 94 1 20 24 15219502
PMC005xxxxxx/PMC5124491.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101465514 34788 Mol Plant Mol Plant Molecular plant 1674-2052 1752-9867 25983207 5124491 10.1016/j.molp.2015.05.003 NIHMS691471 Article SAUR Proteins as Effectors of Hormonal and Environmental Signals in Plant Growth Ren Hong Gray William M. 1 Department of Plant Biology, University of Minnesota, 250 Biological Sciences Center, 1445 Gortner Avenue, St. Paul, MN 55108, USA 1 To whom correspondence should be addressed. [email protected], fax 612-625-1738, tel. 612-624-3042 18 11 2016 15 5 2015 8 2015 27 11 2016 8 8 11531164 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. The plant hormone auxin regulates numerous aspects of plant growth and development. Early auxin response genes mediate its genomic effects on plant growth and development. Discovered in 1987, SMALL AUXIN UP RNAs (SAURs) are the largest family of early auxin response genes. SAUR functions have remained elusive, however, presumably due to extensive genetic redundancy. However, recent molecular, genetic, biochemical, and genomic studies have implicated SAURs in the regulation of a wide range of cellular, physiological, and developmental processes. Recently, crucial mechanistic insight into SAUR function was provided by the demonstration that SAURs inhibit PP2C.D phosphatases to activate plasma membrane (PM) H+-ATPases and promote cell expansion. In addition to auxin, several other hormones and environmental factors also regulate SAUR gene expression. We propose that SAURs are key effector outputs of hormonal and environmental signals that regulate plant growth and development. auxin SAURs plant growth and development cell expansion acid growth PP2C.D phosphatases PM H+-ATPases INTRODUCTION Auxin, derived from the Greek word αυξειν (auxein, meaning to grow or increase), was the first identified plant hormone. Since its discovery in the 1930s, auxin has fascinated plant biologists, who have sought to understand its biosynthesis, metabolism, transport, perception, signaling, and responses. Auxin influences nearly all aspects of plant growth and development through regulating cell division, expansion, differentiation, and patterning. In the nucleus, a co-receptor complex that is composed of TRANSPORT INHIBITOR RESPONSE1/AUXIN SIGNALING F-BOX PROTEINS (TIR1/AFBs) and AUXIN/INDOLE ACETIC ACID (AUX/IAA) transcriptional repressors perceives auxin to regulate gene transcription (reviewed in Wang and Estelle, 2014). These auxin-regulated genes then control downstream auxin responses. Early or primary auxin response genes are transcriptionally induced by auxin within minutes, and this induction does not require de novo protein synthesis (Abel et al., 1994). Most early auxin response genes are classified into three families: AUX/IAAs, GRETCHEN HAGEN3s (GH3s), and SAURs (reviewed in Hagen and Guilfoyle, 2002). SAURs were originally identified in a differential hybridization screen for genes that were rapidly induced by auxin in elongating soybean hypocotyl sections (McClure and Guilfoyle, 1987). Since then, SAURs have been identified in diverse plant species, including mung bean (Yamamoto et al., 1992), pea (Guilfoyle et al., 1993), tomato (Zurek et al., 1994), Arabidopsis (Gil et al., 1994), apple (Watillon et al., 1998), radish (Anai et al., 1998), maize (Yang and Poovaiah, 2000), rice (Jain et al., 2006), moss (Rensing et al., 2008), sorghum (Wang et al., 2010), potato (Wu et al., 2012), cotton (Yang et al., 2012), litchi (Kuang et al., 2012), tobacco (Wu et al., 2012), pepper (Wu et al., 2012), petunia (Wu et al., 2012), peach (Tatsuki et al., 2013), and poplar (Wang et al., 2014a). While substantial progress has been made toward understanding the functions of both Aux/IAA and GH3 proteins in auxin responses (Tiwari et al., 2001; Staswick et al., 2002; Tiwari et al., 2004; Calderón Villalobos et al., 2012), functional studies on SAURs have lagged behind. Nearly 30 years after their discovery, we have only just begun to unlock the secrets of the SAURs. In this review, we describe recent advances that implicate SAURs in regulating a wide range of cellular, physiological, and developmental processes. SAURs, the Largest Family of Early Auxin Response Genes Among early auxin response genes, the SAUR gene family is the most numerous. Genomic bioinformatic analyses have revealed that there are 81 SAURs (including two pseudogenes) in Arabidopsis (Hagen and Guilfoyle, 2002), 58 SAURs (including two pseudogenes) in rice (Jain et al., 2006), 18 SAURs in moss (Rensing et al., 2008), 71 SAURs in sorghum (Wang et al., 2010), 134 SAURs in potato (Wu et al., 2012), 99 SAURs in tomato (Wu et al., 2012), and 79 SAURs in maize (Chen et al., 2014). Typically, SAUR genes are not randomly distributed in the genome, as many of them are found in tandem arrays of extremely highly related genes in soybean (McClure et al., 1989), Arabidopsis (Hagen and Guilfoyle, 2002), rice (Jain et al., 2006), tomato (Wu et al., 2012), and maize (Chen et al., 2014). Tandem and segmental duplication events likely contributed to the expansion of the SAUR gene family (Wu et al., 2012; Chen et al., 2014). The genomic structures of SAUR genes show similar features. The vast majority of SAUR genes lack introns. Many also contain one or more auxin response elements (AuxREs) within their promoter region, and possess a downstream destabilizing (DST) element in the 3’ untranslated region (UTR) (Hagen and Guilfoyle, 2002; Jain et al., 2006; Wu et al., 2012; Chen et al., 2014). The DST consists of three conserved elements separated by non-conserved bases of variable length (GGA(N)xATAGAT(N)xGTA) (McClure et al., 1989; Newman et al., 1993; Sullivan and Green, 1996). This sequence is found in 30 of the 79 Arabidopsis SAUR genes (Supplementary Table 1). For at least some SAUR transcripts, the DST element confers instability (Sullivan and Green, 1996). However, the functional significance of these DST elements is uncertain, as Arabidopsis mutants defective in DST-mediated mRNA degradation exhibit no apparent phenotype (Johnson et al., 2000). SAUR genes encode small proteins that are unique to plants and contain no obvious characterized motifs indicative of a biochemical function. The predicted molecular weights of Arabidopsis SAUR proteins range from 9.3 to 21.4 kDa. SAUR proteins have been predicted to reside in the nucleus, cytosol, mitochondrion, chloroplast, and on the plasma membrane (Wu et al., 2012; Chen et al., 2014). Studies employing SAUR fusion proteins have provided evidence for SAUR localization to the nucleus (ZmSAUR2, Knauss et al., 2003; SAUR32, Park et al., 2007; SAUR36, Narsai et al., 2011), cytosol (OsSAUR39, Kant et al., 2009; SAUR55, Narsai et al., 2011; SAUR41, Kong et al., 2013; SAUR40 and SAUR71, Qiu et al., 2013), and plasma membrane (SAUR63, Chae et al., 2012; SAUR19, Spartz et al., 2012). While such findings should be interpreted with caution, as most of these studies were conducted with SAUR overexpression constructs and sometimes in heterologous systems, the findings suggest that different SAURs may localize to distinct cellular compartments. Protein multiple sequence alignment revealed that SAURs from different plant species contain a central domain specific to SAUR proteins (CDD superfamily cI03633, Marchler-Bauer et al., 2013). This ∼ 60 amino acid domain, referred to here as the SAUR domain, is highly conserved (Figure 1), suggesting that this domain is essential for SAUR protein function. The SAUR domain is dominated by hydrophobic amino acids and also contains short, highly conserved, charged patches and a nearly invariant cysteine residue. Several proline and aromatic amino acids are also exceptionally highly conserved. Outside of the SAUR domain, SAUR proteins have variable length N- and C-terminal extensions, which exhibit much lower degrees of conservation (Supplementary Figure 1). SAUR expression is regulated at multiple levels. Early studies revealed that several SAUR genes are transcriptionally induced by auxin in mung bean, pea, Arabidopsis, radish, and maize (reviewed in Hagen and Guilfoyle, 2002). More recent genomic methods of assessing gene expression have greatly expanded the number of known auxin-induced SAUR genes in Arabidopsis (Tian et al., 2002; Zhao et al., 2003; Goda et al., 2004; Nemhauser et al., 2004; Redman et al., 2004; Okushima et al., 2005; Chapman et al., 2012). However, this list likely remains incomplete as many SAUR genes are not represented on the Arabidopsis ATH1 gene chip and several of the SAUR probe sets that are on this chip are not specific to individual SAUR genes. With these caveats, it is estimated that about one-half of the Arabidopsis SAUR genes are upregulated to some extent by auxin treatment (Supplementary Figure 2). Interestingly, a small number of SAUR genes appear to be repressed by auxin (Supplementary Figure 2). In general, auxin-induced SAURs tend to be most highly expressed in shoots, while many auxin-repressed and non-responsive SAURs are preferentially expressed in roots (Paponov et al., 2008). A second layer of regulation of SAUR expression involves mRNA stability. At least some SAUR transcripts are highly unstable (McClure and Guilfoyle, 1989; Franco et al., 1990). The DST element in the 3’-UTR of SAUR genes functions as an mRNA instability determinant, responsible for the rapid turnover of SAUR mRNAs (Newman et al., 1993). In the cases examined, auxin does not appear to regulate SAUR transcript stability (Gil and Green, 1996). Interestingly, some SAUR transcripts have also been shown to be specifically polyadenylated by PAPS1, one of the four canonical nuclear poly(A) polymerases of Arabidopsis (Vi et al., 2013). This appears to be functionally significant, as overexpression of a GFP-SAUR19 transgene containing a heterologous 3’-UTR could suppress a subset of paps1 mutant phenotypes. In addition to regulation at the transcriptional and posttranscriptional levels, SAUR expression is also subject to posttranslational control. Several SAUR proteins have been found to be highly unstable. The first such report was with the maize SAUR2 (ZmSAUR2) protein (Knauss et al., 2003). 35S-methionine pulse-chase labeling experiments revealed that ZmSAUR2 is a short-lived protein with a half-life of about 7 minutes. More recently, the Arabidopsis SAUR19 and SAUR63 proteins were found to be highly unstable. Treatment with the 26S proteasome inhibitor MG132 causes accumulation of SAUR19 and SAUR63, suggesting the involvement of the ubiquitin-26S proteasome pathway in regulating SAUR degradation (Chae et al., 2012; Spartz et al., 2012). However, neither the mechanism nor regulation of SAUR ubiquitylation has been elucidated, although in the case of SAUR63, the degradation rate was reported to be faster under dim light than bright light (Chae et al., 2012). Interestingly, the SAUR19 and SAUR63 proteins are dramatically stabilized when expressed as fusions with GFP, GUS, or even some small epitope tags (StrepII) (Spartz et al, 2012; Chae et al., 2012). Roles of SAURs in Plant Growth and Development Since their discovery, SAURs have puzzled plant biologists. All plant genomes appear to contain large SAUR gene families. Because of likely genetic redundancy and the difficulties in generating higher order saur mutants due to the tight linkage of tandemly arrayed SAUR genes, it is challenging to employ loss-of-function genetic approaches to elucidate SAUR functions in plant growth and development. To circumvent this problem, gain-of-function genetic approaches that overexpress SAURs or stabilized SAUR fusion proteins have been utilized. Such studies have implicated SAURs in a wide range of cellular, physiological, and developmental processes involving hormonal and environmental control of plant growth and development. While the findings of overexpression studies should always be interpreted with caution, many of these findings have been supported by RNA silencing strategies that target multiple SAUR subfamily members, as well as additional biochemical and genetic findings. Below we review the evidence for SAUR-mediated regulation of several aspects of plant growth and development. Cell Expansion Cell expansion is a fundamental process essential for plant growth and development (reviewed in Braidwood et al., 2013). Hypocotyl growth is mainly caused by cell expansion, making it an excellent study system (Gendreau et al., 1997). Auxin plays a major role in regulating hypocotyl growth (Gray et al. 1998; Zhao et al. 2001). Many SAURs are highly expressed in elongating hypocotyls (McClure and Guilfoyle, 1987; Chae et al., 2012; Spartz et al., 2012; Stamm and Kumar, 2013), and it has long been hypothesized that SAURs may be involved in auxin-regulated cell expansion. Experimental results supporting a positive role for SAURs in cell expansion were obtained from several recent studies in Arabidopsis. Overexpression of SAUR36 (Stamm and Kumar, 2013), SAUR41 (Kong et al., 2013), or stabilized fusion proteins of SAUR19 (Spartz et al., 2012) or SAUR63 (Chae et al., 2012) promotes hypocotyl elongation as a result of increased cell expansion. By contrast, seedlings expressing artificial microRNAs (amiRNAs) targeting multiple members of the SAUR19 (Spartz et al., 2012) or SAUR63 (Chae et al., 2012) subfamilies have slightly shorter hypocotyls with smaller epidermal cells (Figure 2A). These findings suggest that these SAURs positively regulate cell expansion to promote hypocotyl growth. In addition to hypocotyl phenotypes, SAUR63 overexpression and amiRNA knockdown plants also exhibit longer and shorter stamen filaments, respectively. Likewise, SAUR19 and additional SAURs have also been implicated in leaf cell expansion, which we review below. What is the mechanism underlying SAUR-mediated cell expansion? Spartz et al. (2014) recently proposed that SAURs promote cell expansion via an acid growth mechanism. The acid growth theory was proposed in the 1970s to explain auxin-mediated cell expansion (Rayle and Cleland, 1970, 1980, 1992; Hager, 2003). According to this theory, auxin stimulates susceptible cells to excrete protons, lowering apoplastic pH and elevating membrane potential. Plasma membrane (PM) H+-ATPases are responsible for pumping protons into the apoplast. The resulting reduction in apoplastic pH activates cell wall-loosening enzymes such as expansins and hyperpolarizes the plasma membrane, thus increasing cell wall extensibility and promoting solute and water uptake to generate the turgor increase required for cell expansion. Until recently, however, the acid growth theory has lacked strong genetic support, and the underlying molecular mechanism(s) by which auxin activates PM H+-ATPase activity were uncertain. The Arabidopsis genome encodes 11 PM H+-ATPases called AHAs (ARABIDOPSIS H+-ATPases), with AHA1 and AHA2 being the most highly expressed isoforms (Baxter et al., 2003; Haruta et al., 2010). Takahashi et al. (2012) provided key insight into auxin-mediated PM H+-ATPase activation with the demonstration that auxin rapidly induces phosphorylation of the penultimate threonine residue of the C-terminal autoinhibitory domain of AHAs (Thr947 in AHA2). Phosphorylation of this residue coincides with 14-3-3 protein binding and PM H+-ATPase activation (Fuglsang et al., 1999; Kinoshita and Shimazaki, 1999). SAUR19 activates PM H+-ATPases by promoting AHA Thr947 phosphorylation (Spartz et al., 2014). This phosphorylation increase results in elevated PM H+-ATPase activity. Consistent with these findings, plants overexpressing stabilized SAUR19 fusion proteins exhibit remarkable phenotypic similarity to open stomata2 (ost2) mutants. ost2 is a gain-of-function allele of AHA1 that results in a constitutively active enzyme (Merlot et al., 2007). Like ost2 mutants, seedlings expressing SAUR19 fusion proteins exhibit increased elongation growth, reduced aploplastic pH, drought hypersensitivity due to improper guard cell regulation, and constitutive expression of pathogen defense genes (Spartz et al., 2014). These phenotypes are also elicited by fusicoccin, a fungal wilting toxin that specifically activates PM H+-ATPases (Marre, 1979; Singh and Roberts, 2004). SAUR19, as well as several additional SAURs, interact with several members of the 9-member PP2C.D family of protein phosphatases (Schweighofer et al., 2004; Spartz et al., 2014). Enzymatic assays revealed that SAUR binding inhibits phosphatase activity. Spartz et al. (2014) employed gain- and loss-of-function genetic studies to demonstrate that SAUR19 and PP2C.D phosphatases function antagonistically to regulate cell expansion by controlling AHA Thr947 phosphorylation and activity. For example, whereas seedlings overexpressing GFP-SAUR19 exhibit increased cell expansion, Thr947 phosphorylation, and PM H+-ATPase activity, PP2C.D1 overexpression confers reductions in all three processes. Importantly, both SAUR19-mediated growth promotion and PP2C.D1-mediated growth repression occur in a PM H+-ATPase-dependent manner. Consistent with these gain-of-function findings, seedlings expressing a PP2C.D2/5/7/8/9 amiRNA knockdown construct partially phenocopy GFP-SAUR19 overexpression seedlings. Furthermore, PP2C.D1 interacts with AHA2 on the plasma membrane and can catalyze AHA Thr947 dephosphorylation both in vitro and when heterologously expressed in yeast. Based on these results, the authors proposed that auxin induction of SAUR expression results in repression of PP2C.D-mediated dephosphorylation of PM H+-ATPase Thr947. Consequently, PM H+-ATPase equilibrium is shifted to the phosphorylated, active state, resulting in increased apoplastic acidification and cell expansion (Figure 2B). Curiously, in contrast to the above findings, some SAURs have been suggested to negatively regulate cell expansion. Transgenic seedlings overexpressing SAUR32 (also known as AAM1, ABOLISHED APICAL HOOK MAINTENANCE1) exhibit reduced hypocotyl elongation when grown either in the dark or under red light (Park et al., 2007). Likewise, saur36 T-DNA loss-of-function plants have larger leaves containing larger epidermal cells, suggesting that SAUR36 may negatively regulate cell expansion to inhibit leaf growth (Hou et al., 2013). However, Stamm and Kumar (2013) reported that SAUR36 overexpressing seedlings display a long hypocotyl phenotype. At this point, it is unclear whether these disparate findings with SAUR36 in leaves and hypocotyls are due to organ-specific regulation or some other factor. Future studies are required to elucidate the potential mechanisms of SAUR32 and SAUR36 inhibition of cell expansion. Nonetheless, it is tempting to speculate that different SAUR family members might act in an antagonistic fashion, a phenomenon frequently observed in other gene families, such as AUXIN RESPONSE FACTORS (ARFs) and ARABIDOPSIS RESPONSE REGULATORS (ARRs) (reviewed in Guilfoyle and Hagen, 2007; Argueso et al., 2010). Shade Avoidance Responses Under shade (a low red/far-red ratio) conditions, many plants exhibit increased elongation growth of stems and petioles, hyponastic leaves, reduced branching, increased leaf senescence, and early flowering. These growth responses are known as shade avoidance responses. In Arabidopsis seedlings, the most dramatic shade avoidance response is hypocotyl elongation (reviewed in Casal, 2013; de Wit et al., 2014). Mutant seedlings defective in auxin biosynthesis, transport, signaling, and response exhibit reduced hypocotyl growth in response to shade, indicating that auxin is an important regulator of shade avoidance responses. In response to shade, the PHYTOCHROME INTERACTING FACTOR (PIF4, PIF5, and PIF7) transcription factors activate expression of auxin biosynthetic genes to upregulate IAA production. A link between SAURs and shade avoidance response was suggested by Roig-Villanova et al. (2007). In Arabidopsis, PHYTOCHROME RAPIDLY REGULATED1 (PAR1) and PAR2 are atypical basic helix–loop–helix (bHLH) proteins, which lack the basic domain required for DNA-binding. Expression of PAR1 and PAR2 is upregulated by shade in seedlings, inhibiting shade avoidance responses. PAR1 and PAR2 repress the expression of SAUR15 (also known as SAUR-AC1), SAUR67 (misannotated as SAUR68), and several additional SAURs. PAR1 and PAR2 lack a DNA binding domain. Instead, PAR1 interacts with PIF4 to inhibit PIF4 DNA-binding activity (Hao et al., 2012). Many SAURs, including SAUR15, are potential direct targets of PIF4 (Oh et al., 2012). Therefore, PAR1-mediated transcriptional repression of SAUR15 likely involves PAR1 inhibition of PIF4-mediated transcriptional activation. Like PAR1 and PAR2, SAUR15 and SAUR67 are also induced by shade. However, this induction is transient, suggesting that PAR1 and PAR2 repression of SAUR expression may serve as a feedback mechanism for attenuating shade-induced SAUR expression. Additional evidence supporting a role of SAURs in shade-stimulated elongation growth was provided by Hornitschek et al. (2012). PIF4 and PIF5 function redundantly to promote shade-induced hypocotyl growth. The authors used chromatin immunoprecipitation sequencing (ChIP-seq) to identify the genomic binding sites for PIF5 in plants exposed to simulated shade for 2 hours. The potential direct PIF5 targets include the SAUR19 subfamily genes (SAUR19, 20, 21, 22, 23, and 24) as well as additional SAUR genes. In response to shade, expression of SAUR19, SAUR21, SAUR23, and SAUR24 is rapidly induced in elongating hypocotyls (Spartz et al., 2012). Upregulation of SAUR gene expression by shade could be achieved both directly by PIF4 and PIF5-mediated gene activation and indirectly by increased auxin levels controlled by PIF4, PIF5, and PIF7. Presumably, these increases in SAUR expression promote the cell expansion underlying the increased hypocotyl and petiole growth characteristic of shade avoidance growth. However, genetic studies are clearly needed to support this hypothesis. High Temperature-induced Growth Similar to shade, high temperature also induces rapid elongation growth. Arabidopsis seedlings exhibit elongated hypocotyls and petioles in response to high temperature. High temperature promotes auxin biosynthesis in Arabidopsis seedlings (Gray et al., 1998). This upregulation of auxin biosynthesis is controlled by PIF4, the primary PIF family member that controls high temperature-mediated growth (Koini et al., 2009). PIF4 directly regulates the expression of auxin biosynthesis genes TRYPTOPHAN AMINOTRANSFERASE OF ARABIDOPSIS1 (TAA1) and CYP79B2 (Franklin et al., 2011) and YUCCA8 (YUC8) (Sun et al., 2012) to increase auxin production in response to high temperature. Recent findings indicate that SAURs contribute to PIF4-mediated elongation growth at high temperature. SAURs in the SAUR19 subfamily positively regulate cell expansion to promote hypocotyl growth (Spartz et al., 2012; Spartz et al., 2014). High temperature rapidly induces the expression of these genes in hypocotyls (Franklin et al., 2011). However, pif4 null mutants exhibit dramatically reduced expression of these genes at high temperature, indicating that high temperature-induced SAUR gene expression is PIF4-dependent. Furthermore, SAUR19 overexpression rescues the hypocotyl elongation defects of pif4 mutants at high temperature. These results suggest that the SAUR19 subfamily genes function downstream of PIF4 to regulate hypocotyl growth in response to high temperature (Franklin et al., 2011). PIF4-dependent SAUR gene expression could be explained by two distinct mechanisms. As described above, PIF4 upregulates auxin biosynthesis in response to elevated temperatures. Since the SAUR19 subfamily genes are auxin inducible (Spartz et al., 2012), PIF4-mediated increases in auxin levels would induce the expression of these genes. Additionally, PIF4 may directly regulate the expression of the SAUR19 subfamily genes. Oh et al. (2012) provided experimental evidence to support this hypothesis, as SAUR19, 20, 21, 22, 23, and 24 were all identified as high-confidence PIF4 target genes by ChIP-seq analysis. Although targeted follow up experiments are needed to confirm that PIF4 directly binds to the promoters of these genes, it seems likely that both mechanisms contribute to PIF4-dependent SAUR gene expression to promote elongation growth at high temperature. Tropic growth SAUR-mediated changes in cell expansion have also been implicated in tropic growth responses – directional growth in response to environmental stimuli such as light and gravity. Light and gravity signals cause an asymmetric auxin distribution, resulting in differential growth (bending) of shoots and roots during phototropic and gravitropic responses. Shoots exhibit positive phototropism, growing toward the light source. Auxin accumulates on the shaded side of plant stems, promoting the differential cell expansion that results in stem bending toward the light. In response to gravity, shoots exhibit negative gravitropism (grow upward), while roots exhibit positive gravitropism (grow downward). A higher concentration of auxin on the lower side of shoots causes increased growth, while a higher concentration of auxin on the lower side of roots inhibits growth (reviewed in Goyal et al., 2013; Vandenbrink et al., 2014). Several of the first studies on SAUR genes reported increases in SAUR expression within the elongating cells of tropically stimulated stems. Using a tissue printing technique, McClure and Guilfoyle (1989) examined the distribution of SAUR transcripts in the elongating region of excised soybean hypocotyls during gravitropism. Upon gravity stimulation, SAUR transcripts changed from a symmetric distribution to an asymmetric distribution, accumulating on the lower side of hypocotyls. Likewise, increased GUS activity was observed on the lower side of gravistimulated tobacco stems expressing a pSAUR10A::GUS reporter (Li et al., 1991), and RNA in situ hybridization detected more abundant soybean SAUR transcripts in the epidermis and cortex on the lower side of hypocotyls during gravitropic response (Gee et al., 1991). In addition to gravitropism, asymmetric pSAUR10A::GUS activity was also observed in tobacco stems during phototropic growth, with the shaded side exhibiting increased GUS staining (Li et al., 1991). More recent transcriptomic studies have revealed that many SAUR genes are preferentially expressed on the lower, elongating sides of gravistimulated rice shoots and Arabidopsis stems (Hu et al., 2013; Taniguchi et al., 2014). In fact, of the 30 genes exhibiting differential expression between the upper and lower flanks of Arabidopsis stems following a 30-min gravistimulation, 14 were SAUR genes. Curiously, all 14 of these genes belong to the SAUR19 and SAUR63 clades of the phylogeny (Supplementary Figure 2). Overexpression studies have provided some functional support for the involvement of SAURs in tropic growth responses. Arabidopsis seedlings expressing a GFP-SAUR19 transgene driven by the cauliflower mosaic virus 35S promoter exhibit a dramatic delay in phototropic bending and a wavy root growth habit (Spartz et al., 2012). SAUR41 is specifically expressed in the endodermis of the root elongation zone on the upper side of gravistimulated roots, and ectopic expression of SAUR41 resulted in altered gravitropic root growth (Kong et al., 2013). Together with the above gene expression studies correlating SAUR expression with differential growth of the two sides of tropically stimulated roots and shoots, it seems likely that asymmetric SAUR expression may promote the cell expansion underlying tropic growth. Disrupting the normal expression pattern by expressing SAUR genes from heterologous promoters results in the loss of the ability to control differential cell expansion. Interestingly, a recent modeling study (Hohm et al., 2014) found that apoplastic acidification is an important component of phototropic bending, suggesting that SAUR-mediated activation of PM H+-ATPases may drive the differential cell expansion associated with tropic growth. Apical Hook Development Like tropic growth, the development, maintenance, and opening of apical hooks involves the regulated formation of auxin gradients to control differential cell expansion (reviewed in Abbas et al., 2013). Dicotyledonous seedlings grown in the dark develop an apical hook structure caused by differential cell expansion. The apical hook protects the shoot apical meristem from damage when seedlings grow through soil. Upon reaching the soil surface and perceiving light, hook opening is triggered. Gene expression and gain-of-function studies in Arabidopsis have suggested that SAURs are involved in apical hook development. The auxin regulated reporter DR5::GUS is preferentially expressed on the inner side of the apical hook (Abbas et al., 2013). Although SAUR32 expression does not appear to be auxin-inducible, SAUR32 expression in apical hooks closely resembles that of DR5::GUS (Park et al., 2007). While the apical hooks of a saur32 T-DNA null mutant appear wild-type, SAUR32 overexpression from the 35S promoter confers a hookless phenotype due to a defect in apical hook maintenance. Interestingly, SAUR32 overexpression seedlings also have short hypocotyls, suggesting that SAUR32 negatively regulates cell expansion. Together, these findings suggest that during apical hook development, SAUR32 might inhibit cell expansion of the inner side of the apical hook, which is important for hook maintenance. Paradoxically, however, overexpression of SAURs that promote hypocotyl cell expansion such as SAUR19 and SAUR36 also confers a hookless phenotype (Spartz et al., 2012; Stamm and Kumar, 2013). Although additional studies are clearly required, the above findings suggest that proper expression of both growth promoting and growth inhibiting SAUR genes is required to achieve the differential cell expansion necessary to achieve normal apical hook development. The challenge is to understand the molecular mechanisms behind SAUR-regulated differential growth in the apical hook. That said, initial genetic studies of the recently proposed SAUR-PP2C.D mediated regulation of PM H+-ATPase activity support a role for this regulatory module in apical hook development. When grown in the dark, pp2c.d1 and pp2c.d2 mutants, as well as the constitutively active ost2 PM H+-ATPase mutant, exhibit reduced apical hook angles and open cotyledon phenotypes (Spartz et al., 2014). Leaf Growth and Senescence Auxin plays a critical role in leaf growth and development, regulating leaf initiation, lamina formation, vein development, and leaf shape and size (reviewed in Scarpella et al., 2010; Byrne, 2012). Several lines of evidence have suggested that SAURs regulate leaf growth through controlling cell expansion or division, contributing to auxin-regulated leaf growth and development. Spartz et al. (2012) found that the SAUR19 subfamily genes positively regulate leaf growth. Plants expressing an artificial miRNA targeting SAUR19, 23, and 24 display reduced leaf area, while plants expressing stabilized SAUR19 fusion proteins exhibit larger leaves than wild-type. These changes in leaf size appear to be entirely due to altered cell size, suggesting that the SAUR19 subfamily genes positively regulate cell expansion to promote leaf growth. Strikingly, SAUR19 appears quite unique among leaf growth promoting genes. For example, a recent study examining 13 genes that confer increased leaf size when either overexpressed or mutated revealed that SAUR19 and EXPANSIN10 were the only ones that specifically affected cell expansion rather than cell division (Vanhaeren et al., 2014). Furthermore, when tested in pairwise combinations with the other growth promoting mutations/transgenes, SAUR19 overexpression resulted in synergistic increases in leaf size. These findings suggest that the SAUR19 subfamily genes are attractive targets for future genetic engineering efforts aimed at enhancing plant biomass. Unlike the SAUR19 subfamily genes, a few SAURs have been implicated as negative regulators of leaf growth. SAUR76 is highly expressed in roots, but only weakly expressed in leaves. Interestingly, while auxin treatment resulted in a dramatic upregulation in roots, no increase in SAUR76 expression was observed in leaves (Markakis et al., 2013). When overexpressed from the 35S promoter, SAUR76 conferred a reduction in leaf size. This effect appears to be the result of a reduction in cell number rather than cell size, suggesting that SAUR76 might negatively regulate cell division to inhibit leaf growth (Markakis et al., 2013). Similarly, SAUR36 also appears to negatively regulate leaf growth, as saur36 null mutants exhibit enlarged leaves (Hou et al., 2013). However, unlike SAUR76 overexpression, saur36 mutants exhibit an increase in leaf epidermal cell size, suggesting that SAUR36 negatively regulates leaf cell expansion rather than cell division. In addition to regulating leaf growth and development, auxin has also been implicated in leaf senescence, a process eventually leading to leaf death. However, the precise functions of auxin in leaf senescence are unclear, as contradictory results have been reported as to whether auxin negatively or positively regulates leaf senescence (reviewed in Jibran et al., 2013; Khan et al., 2014). Recent studies of auxin induced SAUR genes in Arabidopsis and rice suggest that auxin may promote leaf senescence through the expression of SAURs. Arabidopsis SAUR36, also known as SAG201 (SENESCENCE-ASSOCIATED GENE201), is highly upregulated during leaf senescence (Hou et al., 2013), and leaf senescence of two independent saur36 T-DNA mutants was significantly delayed. Furthermore, inducible overexpression of SAUR36 promoted premature leaf senescence. Interestingly, only SAUR36 expression constructs that lacked the DST element within the 3’-UTR conferred an early senescence phenotype, suggesting that post-transcriptional regulation of SAUR36 transcript levels may play an important functional role. Together, these findings convincingly demonstrate that SAUR36 positively regulates leaf senescence. Similar findings were obtained with OsSAUR39 in rice. Though loss-of-function support is lacking, OsSAUR39 expression is elevated in older leaves, and overexpression of OsSAUR39 promoted early senescence (Kant et al., 2009). Further studies are needed to provide mechanistic insight and to determine whether the upregulation of SAUR36 and OsSAUR39 expression during senescence is auxin-dependent. Root Growth and Development Auxin is essential for root growth and development, regulating stem cell specification and division, meristem size, cell elongation, and differentiation (reviewed in Sozzani and Iyer-Pascuzzi, 2014; Takatsuka and Umeda, 2014). In addition to the primary roots, auxin also regulates the initiation and growth and development of lateral roots. While the majority of SAUR studies discussed above have focused on expression patterns and functional studies in shoots, several of the aforementioned SAUR overexpression lines also exhibit root phenotypes. SAUR76 is expressed in the pericycle and endodermal layers of the root elongation zone and at lateral root initiation sites (Markakis et al., 2013). Auxin treatment strongly upregulates SAUR76 expression in roots. While saur76 null mutants exhibited no apparent phenotype, overexpression of SAUR76 conferred increases in root meristem size and primary root growth, suggesting that SAUR76 may positively regulate root growth. SAUR41, on the other hand, is specifically expressed in the quiescent center and cortical/endodermal initials of the stem cell niche of primary and lateral roots. Its expression pattern is expanded to the root endodermis in seedlings treated with auxin for 1 hour. Plants overexpressing SAUR41 exhibited increased primary root growth and lateral root numbers (Kong et al., 2013). Interestingly, expression of SAUR41 under the control of tissue/cell type-specific promoters (PIN1, WOX5, PLT2, and others) caused defects in root meristem patterning and auxin distribution within root tips (as assayed using the DR5rev::GFP reporter). As in shoots, there is some evidence for positive and negative acting SAURs in root growth and development. In contrast to the above findings with Arabidopsis SAUR41 and SAUR76, overexpression of OsSAUR39 in rice resulted in reductions in root elongation and lateral root development (Kant et al., 2009). To date, no molecular or biochemical mechanism has been elucidated to explain how SAUR proteins might regulate root growth and development. Calcium Signaling In response to cytosolic free Ca2+ increases, calcium-binding proteins, such as calmodulins (CaM) and CaM-like proteins (CML), sense Ca2+ signals and initiate signaling cascades to regulate many cellular, physiological, and developmental processes (reviewed in Dodd et al., 2010; Kudla et al., 2010). Numerous studies have suggested that Ca2+ is involved in auxin signaling or responses (reviewed in Vanneste and Friml, 2013). Auxin induces a rapid and transient increase of cytosolic Ca2+ concentration in various plant species. However, the functions of Ca2+ in auxin signaling or responses are poorly understood. SAUR proteins may provide a functional link between Ca2+ and auxin responses. Using recombinant 35S-labeled CaM to screen a maize root cDNA expression library, Yang and Poovaiah (2000) isolated ZmSAUR1 as a CaM-interacting protein that binds to CaM in a calcium-dependent manner. Deletion analyses revealed that the CaM-binding motif is located in the N-terminal region of ZmSAUR1. This study and subsequent studies have identified additional SAUR proteins (maize ZmSAUR2, Arabidopsis SAUR15 and SAUR70, and soybean SAUR10A5) capable of binding CaM (Knauss et al., 2003; Popescu et al., 2007). Our search for potential CaM-binding sites within the 79 Arabidopsis SAUR proteins using CALMODULIN TARGET DATABASE (http://calcium.uhnres.utoronto.ca/ctdb/ctdb/sequence.html) identified potential CaM-binding motifs in all family members except SAUR4, 6, 10, 18, 30, 33, 40, 41, 52, 57, 58, 67, and 74. To date, however, only SAUR70 has been shown to bind CaM or CaM-related proteins in planta (Popescu et al., 2007) and no functional significance has been ascribed to this potential activity of SAUR proteins. Auxin transport Several studies have indicated that SAURs can modulate polar auxin transport. In Arabidopsis, SAUR19 overexpression and SAUR19/23/24 amiRNA knockdown seedlings conferred increased and decreased basipetal indole-3-acetic acid (IAA) transport, respectively, in hypocotyls (Spartz et al., 2012). Similar findings have been obtained with SAUR41 and SAUR63 (Chae et al., 2012; Kong et al., 2013). In contrast, overexpression of OsSAUR39 in rice resulted in reduced basipetal IAA transport (Kant et al., 2009), again raising the possibility that different SAUR proteins may function antagonistically. The simplest explanation for increased auxin transport is the elevated PM H+-ATPase activity resulting from SAUR overexpression. As per the chemiosmotic model for IAA transport (Blakeslee et al., 2005), the increase in apoplastic acidification and the predicted increase in plasma membrane potential associated with SAUR overexpression would be expected to promote IAA transport. However, other mechanisms are also possible, including the aforementioned CaM binding activity of SAURs, as polar auxin transport is highly dependent on calcium (Vanneste and Friml, 2013). Regardless of the mechanism, the fact that SAUR proteins are capable of modulating IAA transport provides a potential explanation for the diversity of effects, including altered cell expansion, division, and patterning that result from ectopic SAUR expression. Hormonal Regulation of SAUR Expression - more than just Auxin In addition to auxin, brassinosteroids (BR), gibberellins (GA), jasmonate (JA), and abscisic acid (ABA) have been reported to regulate the expression of some SAUR genes, indicating that SAURs likely contribute to other hormone-regulated aspects of plant growth and development. Similar to auxin, BR is a growth-promoting hormone that regulates many physiological and developmental processes (reviewed in Fàbregas and Caño-Delgado, 2014; Wang et al., 2014b). A large body of evidence suggests that the BR and auxin pathways exhibit extensive crosstalk and converge at the transcriptional level to regulate many shared target genes (reviewed in Vert and Chory, 2011; Wang et al., 2012), including many SAUR genes (Goda et al., 2004; Nemhauser et al., 2004). In the BR signaling pathway, the BRI1-EMS-SUPPRESSOR1 (BES1)/BRASSINAZOLE RESISTANT1 (BZR1) family transcription factors regulate gene expression (He et al., 2005; Yin et al., 2005). Chromatin immunoprecipitation – DNA microarray (ChIP-chip) studies identified many SAUR genes as potential direct targets of BZR1 and BES1 (Supplementary Table 2) (Sun et al., 2010; Yu et al., 2011). Indeed, ChIP assays have confirmed that BZR1 and BES1 bind to the SAUR15 promoter (Yin et al., 2005; Sun et al., 2010; Walcher and Nemhauser, 2012), and BES1 binds to SAUR36 and SAUR59 promoters (Yu et al., 2011). While the precise roles of SAUR proteins in BR action remain unclear, given the well-established role of BR in promoting cell expansion, it seems likely that SAURs are downstream effectors that mediate at least some aspects of BR-mediated expansion growth. GA is also a growth-promoting hormone that regulates diverse physiological and developmental processes, such as seed germination and stem elongation. The DELLA repressors function as central regulators in the GA signaling pathway, and GA promotes DELLA protein degradation by the 26S proteasome to stimulate growth (reviewed in Claeys et al., 2014; Xu et al., 2014). Consistent with a possible role for SAURs in GA-dependent growth, Bai et al. (2012) recently found that 27 SAUR genes were upregulated following a 12h GA treatment (Supplementary Table 3). This increase in SAUR expression is likely the result of enhanced ARF, PIF, and BZR1/BES1 transcription factor binding to SAUR promoters following DELLA proteolysis, as DELLAs interact with, and inhibit the binding of each of these transcription factors (de Lucas et al., 2008; Feng et al., 2008; Bai et al., 2012; Oh et al., 2014). A second connection between SAUR proteins and GA-regulated growth is in seed germination. GA promotes germination, and RGA-LIKE2 (RGL2) is a major DELLA protein that represses seed germination in Arabidopsis. Stamm et al. (2012) reported that RGL2 upregulates SAUR36, which is highly expressed in dry and imbibed seeds, and ChIP-qPCR analysis detected RGL2 binding to the SAUR36 promoter. Consistent with the upregulation of SAUR36 by RGL2, GA represses SAUR36 expression (Stamm and Kumar, 2013). Genetic studies support a role for SAUR36 in regulating seed germination, although curiously, germination of both saur36 null mutants and SAUR36 overexpression lines was hypersensitive to paclobutrazol (PAC, a gibberellin biosynthesis inhibitor) and ABA treatments. Further research is required to understand the precise role of SAUR36 in seed germination. In contrast to the general trend of increased SAUR expression elicited by the above growth promoting hormones, many SAUR genes are repressed by the stress-related hormones jasmonate and ABA (Supplementary Table 4) (Nemhauser et al., 2006). ABA is a major regulator of the adaptive responses of plants to stresses, such as drought and high salt, which cause plant growth inhibition (reviewed in Yoshida et al., 2014). Many SAUR genes exhibit reduced expression in response to ABA treatment, as well as drought and osmotic stresses (Kodaira et al., 2011). To investigate the mechanisms underlying abiotic stress-regulated growth inhibition, Kodaira et al. (2011) studied two transcriptional repressors AZF1 (ARABIDOPSIS ZINC-FINGER PROTEIN1) and AZF2. Expression of AZF1 and AZF2 is induced by ABA, drought, and high salt, and AZF overexpression confers growth repression. Transcriptome analyses of DEX-inducible AZF1 and AZF2 overexpression plants identified 27 SAUR genes, whose expression is downregulated. These genes included nearly all members of the SAUR19 and SAUR63 subfamilies, and gel shift assays demonstrated that recombinant AZF1 and AZF2 proteins could bind to the promoters of SAUR20 and SAUR63, suggesting that at least these SAURs are direct targets of AZF-mediated repression. Consistent with these findings, expression of SAUR16 and SAUR63 is increased in the azf1 azf2 double mutant in response to high salt stress. As several studies have found that the SAUR19 and SAUR63 subfamily genes positively regulate cell expansion to promote plant growth (Chae et al., 2012; Spartz et al., 2012; Spartz et al., 2014), these findings suggest that AZF-mediated down-regulation of SAUR expression may contribute to the growth repression characteristic of ABA and abiotic stress treatments. Potentially, the molecular basis for this regulation could be the aforementioned SAUR regulation of PP2C.D phosphatase activity to modulate PM H+-ATPase activity. While both auxin treatment and SAUR19 overexpression promote increases in PM H+-ATPase C-terminal Thr947 phosphorylation and enzymatic activity, ABA treatment results in reduced Thr947 phosphorylation and ATPase activity (Hayashi et al., 2014). CONCLUSIONS AND FUTURE PERSPECTIVES Following a protracted lag phase, the functions of the large, highly conserved, and plant specific SAUR gene family are beginning to be elucidated. Data obtained from molecular, genetic, biochemical, and genomic studies have suggested that SAURs are involved in regulating a wide range of cellular, physiological, and developmental processes in response to hormonal and environmental signals. We suggest that SAURs play key roles in integrating hormonal and environmental signals into distinct growth and developmental responses (Figure 3). Based on our current knowledge, this regulation is accomplished largely through the induction or repression of SAUR gene transcription. In particular, the SAUR19 and SAUR63 subfamilies, both of which have been intimately linked with cell expansion, appear to be especially responsive to hormonal and environmental signals that regulate plant growth. That said, there is strong evidence that SAUR activities are also controlled post-transcriptionally, via the regulation of mRNA and protein stability. Determining the extent and mechanisms by which hormonal and environmental signals contribute to these modes of regulation is a crucial area for future research. Although considerable progress has been made in recent years, we still have only a rudimentary understanding of the functions and molecular mechanisms of SAUR-regulated plant growth and development. Much of our current knowledge is based on expression patterns and gain-of-function studies, which, at best, can only suggest potential functions. While SAUR loss-of-function studies are complicated by likely extensive genetic redundancy and tight linkage arrangements, modern genome editing techniques such as TALEN and CRISPR/Cas9 nucleases offer potential solutions. Molecular and biochemical studies are also desperately needed to provide additional mechanistic insight into SAUR protein function. The recent demonstration that SAUR19 inhibits PP2C.D phosphatase activity to control PM H+-ATPase activity is an exciting first step in this direction (Spartz et al., 2014). And indeed, the crucial role that PM H+-ATPases play in establishing electrochemical gradients across the plasma membrane could potentially control, either directly or indirectly, many of the growth responses that appear to be influenced by SAURs. However, the observations that different SAUR proteins exhibit distinct subcellular localizations strongly suggest that PM H+-ATPase regulation is only one of several SAUR functions. That said, the possibility that PP2C.D regulation is the primary function of SAUR proteins remains an intriguing prospect. There are 9 PP2C.D family members in Arabidopsis, and initial work suggests that these proteins are differentially localized (Tovar-Mendez et al., 2014). Thus, a SAUR-PP2C.D “code” may exist, whereby distinct SAUR-PP2C.D combinations in different tissues and subcellular compartments could modulate the phosphorylation status and activities of many downstream effectors to control diverse aspects of growth and development. Detailed PP2C.D genetic studies, characterization of PP2C.D expression patterns, and the identification of additional PP2C.D substrates will likely be informative in testing this possibility and potentially forging new links with SAUR proteins. In addition, the CaM binding activity exhibited by several SAUR proteins also warrants additional studies that may lead to the identification of distinct biochemical functions for SAUR proteins. It now seems apparent that SAUR proteins are crucial regulators of diverse aspects of plant growth and development. Unraveling the molecular networks that link specific hormonal and environmental signals to particular SAUR proteins and cellular/developmental outputs will undoubtedly improve our understanding of SAUR function, and may lead to novel strategies for manipulating plant growth. We expect many new and exciting findings on this long neglected family of plant proteins in the years to come. Supplementary Material We thank Angela K. Spartz and Mee Yeon Park for critical reading and comments on the manuscript, and thank the University of Minnesota - University Imaging Center for the Nikon A1 spectral confocal microscope. FUNDING This work was supported by grants to WMG from the National Institutes of Health (GM067203) and National Science Foundation (MCB-0817205). Figure 1 The SAUR domain Multiple sequence alignment of the SAUR domain of 247 SAUR proteins from Arabidopsis thaliana, Oryza sativa, Populus trichocarpa, and Physcomitrella patens was performed using Clustal Omega. The consensus residue(s) at each position are color coded. Dotted lines indicate the positions of short, nonconserved insertions present in a few SAUR family members that were removed from the alignment. Figure 2 SAURs activate PM H+-ATPases via inhibiting PP2C.D phosphatases to promote cell expansion (A) Seven-day-old light-grown Arabidopsis seedlings. Seedlings were stained in 10 µg /ml propidium iodide, and hypocotyl epidermal cells were observed under a Nikon A1 spectral confocal microscope. Asterisks (*) indicate the intercellular space of two adjacent epidermal cells. Scale bar = 100 µm. (B) A model for SAUR-regulated cell expansion. Adapted from Spartz et al. (2014). Figure 3 SAURs integrate hormonal and environmental signals to regulate plant growth and development Arrows and blunted lines indicate positive and negative regulation, respectively. Dashed lines indicate either positive or negative regulation. Hormones and environmental factors confer changes in SAUR expression. Particular SAURs then positively or negatively regulate specific cellular, physiological, and developmental processes. The indicated effects of the various hormones and environmental factors on SAUR expression reflect the overall trend. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. SUPPLEMENTARY DATA Supplementary data are available at Molecular Plant Online. No conflict of interest declared. REFERENCES Abbas M Alabadí D Blázquez MA Differential growth at the apical hook: all roads lead to auxin Front. Plant Sci 2013 4 441 24204373 Abel S Oeller PW Theologis A Early auxin-induced genes encode short-lived nuclear proteins Proc. Natl. Acad. Sci. U S A 1994 91 326 330 8278386 Anai T Kono N Kosemura S Yamamura S Hasegawa K Isolation and characterization of an auxin-inducible SAUR gene from radish seedlings DNA Seq 1998 9 329 333 10524760 Argueso CT Raines T Kieber JJ Cytokinin signaling and transcriptional networks Curr. Opin. Plant Biol 2010 13 533 539 20851038 Bai MY Shang JX Oh E Fan M Bai Y Zentella R Sun TP Wang ZY Brassinosteroid, gibberellin and phytochrome impinge on a common transcription module in Arabidopsis Nat. Cell Biol 2012 14 810 817 22820377 Baxter I Tchieu J Sussman MR Boutry M Palmgren MG Gribskov M Harper JF Axelsen KB Genomic comparison of P-type ATPase ion pumps in Arabidopsis and rice Plant Physiol 2003 132 618 628 12805592 Blakeslee JJ Peer WA Murphy AS Auxin transport Curr. Opin. Plant Biol 2005 8 494 500 16054428 Braidwood L Breuer C Sugimoto K My body is a cage: mechanisms and modulation of plant cell growth New Phytol 2013 201 388 402 24033322 Byrne ME Making leaves Curr. Opin. Plant Biol 2012 15 24 30 22079784 Calderón Villalobos LI Lee S De Oliveira C Ivetac A Brandt W Armitage L Sheard LB Tan X Parry G Mao H Zheng N Napier R Kepinski S Estelle M A combinatorial TIR1/AFB-Aux/IAA co-receptor system for differential sensing of auxin Nat. Chem. Biol 2012 8 477 485 22466420 Casal JJ Photoreceptor signaling networks in plant responses to shade Annu. Rev. Plant Biol 2013 64 403 427 23373700 Chae K Isaacs CG Reeves PH Maloney GS Muday GK Nagpal P Reed JW Arabidopsis SMALL AUXIN UP RNA63 promotes hypocotyl and stamen filament elongation Plant J 2012 71 684 697 22507274 Chapman EJ Greenham K Castillejo C Sartor R Bialy A Sun TP Estelle M Hypocotyl transcriptome reveals auxin regulation of growth-promoting genes through GA-dependent and -independent pathways PLoS One 2012 7 e36210 22590525 Chen Y Hao X Cao J Small auxin upregulated RNA (SAUR) gene family in maize: identification, evolution, and its phylogenetic comparison with Arabidopsis, rice, and sorghum J. Integr. Plant Biol 2014 56 133 150 24472286 Claeys H De Bodt S Inzé D Gibberellins and DELLAs: central nodes in growth regulatory networks Trends Plant Sci 2014 19 231 239 24182663 de Lucas M Daviere JM Rodriguez-Falcon M Pontin M Iglesias-Pedraz JM Lorrain S Fankhauser C Blazquez MA Titarenko E Prat S A molecular framework for light and gibberellin control of cell elongation Nature 2008 451 480 484 18216857 de Wit M Lorrain S Fankhauser C Auxin-mediated plant architectural changes in response to shade and high temperature Physiol. Plant 2014 151 13 24 24011166 Dodd AN Kudla J Sanders D The language of calcium signaling Annu. Rev. Plant Biol 2010 61 593 620 20192754 Feng S Martinez C Gusmaroli G Wang Y Zhou J Wang F Chen L Yu L Iglesias-Pedraz JM Kircher S Schafer E Fu X Fan LM Deng XW Coordinated regulation of Arabidopsis thaliana development by light and gibberellins Nature 2008 451 475 479 18216856 Franco AR Gee MA Guilfoyle TJ Induction and superinduction of auxin-responsive mRNAs with auxin and protein synthesis inhibitors J. Biol. Chem 1990 265 15845 15849 2394751 Franklin KA Lee SH Patel D Kumar SV Spartz AK Gu C Ye S Yu P Breen G Cohen JD Wigge PA Gray WM Phytochrome-interacting factor 4 (PIF4) regulates auxin biosynthesis at high temperature Proc. Natl. Acad. Sci. U S A 2011 108 20231 20235 22123947 Fuglsang AT Visconti S Drumm K Jahn T Stensballe A Mattei B Jensen ON Aducci P Palmgren MG Binding of 14-3-3 protein to the plasma membrane H+-ATPase AHA2 involves the three C-terminal residues Tyr946-Thr-Val and requires phosphorylation of Thr947 J. Biol. Chem 1999 274 36774 36780 10593986 Fàbregas N Caño-Delgado AI Turning on the microscope turret: a new view for the study of brassinosteroid signaling in plant development Physiol. Plant 2014 151 172 183 24547704 Gee MA Hagen G Guilfoyle TJ Tissue-specific and organ-specific expression of soybean auxin-responsive transcripts GH3 and SAURs Plant Cell 1991 3 419 430 1840920 Gendreau E Traas J Desnos T Grandjean O Caboche M Hofte H Cellular basis of hypocotyl growth in Arabidopsis thaliana Plant Physiol 1997 114 295 305 9159952 Gil P Green PJ Multiple regions of the Arabidopsis SAUR-AC1 gene control transcript abundance: the 3’ untranslated region functions as an mRNA instability determinant EMBO J 1996 15 1678 1686 8612592 Gil P Liu Y Orbović V Verkamp E Poff KL Green PJ Characterization of the auxin-inducible SAUR-AC1 gene for use as a molecular genetic tool in Arabidopsis Plant Physiol 1994 104 777 784 8159792 Goda H Sawa S Asami T Fujioka S Shimada Y Yoshida S Comprehensive comparison of auxin-regulated and brassinosteroid-regulated genes in Arabidopsis Plant Physiol 2004 134 1555 1573 15047898 Goyal A Szarzynska B Fankhauser C Phototropism: at the crossroads of light-signaling pathways Trends Plant Sci 2013 18 393 401 23562459 Gray WM Ostin A Sandberg G Romano CP Estelle M High temperature promotes auxin-mediated hypocotyl elongation in Arabidopsis Proc. Natl. Acad. Sci. U S A 1998 95 7197 7202 9618562 Guilfoyle TJ Hagen G Auxin response factors Curr. Opin. Plant Biol 2007 10 453 460 17900969 Guilfoyle TJ Hagen G Li Y Ulmasov TN Liu Z Strabala T Gee M Auxin-regulated Transcription Aust. J. Plant Physiol 1993 20 489 502 Hagen G Guilfoyle T Auxin-responsive gene expression: genes, promoters and regulatory factors Plant Mol. Biol 2002 49 373 385 12036261 Hager A Role of the plasma membrane H+-ATPase in auxin-induced elongation growth: historical and new aspects J. Plant Res 2003 116 483 505 12937999 Hao Y Oh E Choi G Liang Z Wang ZY Interactions between HLH and bHLH factors modulate light-regulated plant development Mol. Plant 2012 5 688 697 22331621 Haruta M Burch HL Nelson RB Barrett-Wilt G Kline KG Mohsin SB Young JC Otegui MS Sussman MR Molecular characterization of mutant Arabidopsis plants with reduced plasma membrane proton pump activity J. Biol. Chem 2010 285 17918 17929 20348108 Hayashi Y Takahashi K Inoue SI Kinoshita T Abscisic Acid Suppresses Hypocotyl Elongation by Dephosphorylating Plasma Membrane H+-ATPase in Arabidopsis thaliana Plant Cell Physiol 2014 55 845 853 24492258 He JX Gendron JM Sun Y Gampala SS Gendron N Sun CQ Wang ZY BZR1 is a transcriptional repressor with dual roles in brassinosteroid homeostasis and growth responses Science 2005 307 1634 1638 15681342 Hohm T Demarsy E Quan C Allenbach Petrolati L Preuten T Vernoux T Bergmann S Fankhauser C Plasma membrane H+-ATPase regulation is required for auxin gradient formation preceding phototropic growth Mol. Syst. Biol 2014 10 751 25261457 Hornitschek P Kohnen MV Lorrain S Rougemont J Ljung K Lopez-Vidriero I Franco-Zorrilla JM Solano R Trevisan M Pradervand S Xenarios I Fankhauser C Phytochrome interacting factors 4 and 5 control seedling growth in changing light conditions by directly controlling auxin signaling Plant J 2012 71 699 711 22536829 Hou K Wu W Gan SS SAUR36, a small auxin up RNA gene, is involved in the promotion of leaf senescence in Arabidopsis Plant Physiol 2013 161 1002 1009 23250625 Hu L Mei Z Zang A Chen H Dou X Jin J Cai W Microarray analyses and comparisons of upper or lower flanks of rice shoot base preceding gravitropic bending PLoS One 2013 8 e74646 24040303 Jain M Tyagi AK Khurana JP Genome-wide analysis, evolutionary expansion, and expression of early auxin-responsive SAUR gene family in rice (Oryza sativa) Genomics 2006 88 360 371 16707243 Jibran R Hunter DA Dijkwel PP Hormonal regulation of leaf senescence through integration of developmental and stress signals Plant Mol. Biol 2013 82 547 561 23504405 Johnson MA Perez-Amador MA Lidder P Green PJ Mutants of Arabidopsis defective in a sequence-specific mRNA degradation pathway Proc. Natl. Acad. Sci. U S A 2000 97 13991 13996 11087822 Kant S Bi YM Zhu T Rothstein SJ SAUR39, a small auxin-up RNA gene, acts as a negative regulator of auxin synthesis and transport in rice Plant Physiol 2009 151 691 701 19700562 Khan M Rozhon W Poppenberger B The role of hormones in the aging of plants - a mini-review Gerontology 2014 60 49 55 24135638 Kinoshita T Shimazaki K Blue light activates the plasma membrane H+-ATPase by phosphorylation of the C-terminus in stomatal guard cells EMBO J 1999 18 5548 5558 10523299 Knauss S Rohrmeier T Lehle L The auxin-induced maize gene ZmSAUR2 encodes a short-lived nuclear protein expressed in elongating tissues J. Biol. Chem 2003 278 23936 23943 12695517 Kodaira KS Qin F Tran LS Maruyama K Kidokoro S Fujita Y Shinozaki K Yamaguchi-Shinozaki K Arabidopsis Cys2/His2 zinc-finger proteins AZF1 and AZF2 negatively regulate abscisic acid-repressive and auxin-inducible genes under abiotic stress conditions Plant Physiol 2011 157 742 756 21852415 Koini MA Alvey L Allen T Tilley CA Harberd NP Whitelam GC Franklin KA High temperature-mediated adaptations in plant architecture require the bHLH transcription factor PIF4 Curr. Biol 2009 19 408 413 19249207 Kong Y Zhu Y Gao C She W Lin W Chen Y Han N Bian H Zhu M Wang J Tissue-specific expression of SMALL AUXIN UP RNA41 differentially regulates cell expansion and root meristem patterning in Arabidopsis Plant Cell Physiol 2013 54 609 621 23396598 Kuang JF Wu JY Zhong HY Li CQ Chen JY Lu WJ Li JG Carbohydrate stress affecting fruitlet abscission and expression of genes related to auxin signal transduction pathway in litchi Int. J. Mol. Sci 2012 13 16084 16103 23443112 Kudla J Batistic O Hashimoto K Calcium signals: the lead currency of plant information processing Plant Cell 2010 22 541 563 20354197 Li Y Hagen G Guilfoyle TJ An Auxin-Responsive Promoter Is Differentially Induced by Auxin Gradients during Tropisms Plant Cell 1991 3 1167 1175 12324587 Marchler-Bauer A Zheng C Chitsaz F Derbyshire MK Geer LY Geer RC Gonzales NR Gwadz M Hurwitz DI Lanczycki CJ Lu F Lu S Marchler GH Song JS Thanki N Yamashita RA Zhang D Bryant SH CDD: conserved domains and protein three-dimensional structure Nucleic Acids Res 2013 41 D348 D352 23197659 Markakis MN Boron AK Van Loock B Saini K Cirera S Verbelen JP Vissenberg K Characterization of a small auxin-up RNA (SAUR)-like gene involved in Arabidopsis thaliana development PLoS One 2013 8 e82596 24312429 Marre E Fusicoccin: A tool in plant physiology Ann. Rev. Plant Physiol 1979 30 273 288 McClure BA Guilfoyle T Characterization of a class of small auxin-inducible soybean polyadenylated RNAs Plant Mol. Biol 1987 9 611 623 24277197 McClure BA Guilfoyle T Rapid redistribution of auxin-regulated RNAs during gravitropism Science 1989 243 91 93 11540631 McClure BA Hagen G Brown CS Gee MA Guilfoyle TJ Transcription, organization, and sequence of an auxin-regulated gene cluster in soybean Plant Cell 1989 1 229 239 2485235 Merlot S Leonhardt N Fenzi F Valon C Costa M Piette L Vavasseur A Genty B Boivin K Müller A Giraudat J Leung J Constitutive activation of a plasma membrane H+-ATPase prevents abscisic acid-mediated stomatal closure EMBO J 2007 26 3216 3226 17557075 Narsai R Law SR Carrie C Xu L Whelan J In-depth temporal transcriptome profiling reveals a crucial developmental switch with roles for RNA processing and organelle metabolism that are essential for germination in Arabidopsis Plant Physiol 2011 157 1342 1362 21908688 Nemhauser JL Mockler TC Chory J Interdependency of brassinosteroid and auxin signaling in Arabidopsis PLoS Biol 2004 2 E258 15328536 Nemhauser JL Hong F Chory J Different plant hormones regulate similar processes through largely nonoverlapping transcriptional responses Cell 2006 126 467 475 16901781 Newman TC Ohme-Takagi M Taylor CB Green PJ DST sequences, highly conserved among plant SAUR genes, target reporter transcripts for rapid decay in tobacco Plant Cell 1993 5 701 714 8329900 Oh E Zhu JY Wang ZY Interaction between BZR1 and PIF4 integrates brassinosteroid and environmental responses Nat. Cell Biol 2012 14 802 809 22820378 Oh E Zhu JY Bai MY Arenhart RA Sun Y Wang ZY Cell elongation is regulated through a central circuit of interacting transcription factors in the Arabidopsis hypocotyl Elife 2014 3 e03031 Okushima Y Overvoorde PJ Arima K Alonso JM Chan A Chang C Ecker JR Hughes B Lui A Nguyen D Onodera C Quach H Smith A Yu G Theologis A Functional Genomic Analysis of the AUXIN RESPONSE FACTOR Gene Family Members in Arabidopsis thaliana: Unique and Overlapping Functions of ARF7 and ARF19 Plant Cell 2005 17 444 463 15659631 Paponov IA Paponov M Teale W Menges M Chakrabortee S Murray JA Palme K Comprehensive transcriptome analysis of auxin responses in Arabidopsis Mol. Plant 2008 1 321 337 19825543 Park JE Kim YS Yoon HK Park CM Functional characterization of a small auxin-up RNA gene in apical hook development in Arabidopsis Plant Sci 2007 172 150 157 Popescu SC Popescu GV Bachan S Zhang Z Seay M Gerstein M Snyder M Dinesh-Kumar SP Differential binding of calmodulin-related proteins to their targets revealed through high-density Arabidopsis protein microarrays Proc. Natl. Acad. Sci. U S A 2007 104 4730 4735 17360592 Qiu T Chen Y Li M Kong Y Zhu Y Han N Bian H Zhu M Wang J The tissue-specific and developmentally regulated expression patterns of the SAUR41 subfamily of small auxin up RNA genes: potential implications Plant Signal Behav 2013 8 e25283 23759547 Rayle DL Cleland R Enhancement of wall loosening and elongation by Acid solutions Plant Physiol 1970 46 250 253 16657445 Rayle DL Cleland RE Evidence that Auxin-induced Growth of Soybean Hypocotyls Involves Proton Excretion Plant Physiol 1980 66 433 437 16661450 Rayle DL Cleland RE The Acid Growth Theory of auxin-induced cell elongation is alive and well Plant Physiol 1992 99 1271 1274 11537886 Redman JC Haas BJ Tanimoto G Town CD Development and evaluation of an Arabidopsis whole genome Affymetrix probe array Plant J 2004 38 545 561 15086809 Rensing SA Lang D Zimmer AD Terry A Salamov A Shapiro H Nishiyama T Perroud PF Lindquist EA Kamisugi Y Tanahashi T Sakakibara K Fujita T Oishi K Shin IT Kuroki Y Toyoda A Suzuki Y Hashimoto S Yamaguchi K Sugano S Kohara Y Fujiyama A Anterola A Aoki S Ashton N Barbazuk WB Barker E Bennetzen JL Blankenship R Cho SH Dutcher SK Estelle M Fawcett JA Gundlach H Hanada K Heyl A Hicks KA Hughes J Lohr M Mayer K Melkozernov A Murata T Nelson DR Pils B Prigge M Reiss B Renner T Rombauts S Rushton PJ Sanderfoot A Schween G Shiu SH Stueber K Theodoulou FL Tu H Van de Peer Y Verrier PJ Waters E Wood A Yang L Cove D Cuming AC Hasebe M Lucas S Mishler BD Reski R Grigoriev IV Quatrano RS Boore JL The Physcomitrella genome reveals evolutionary insights into the conquest of land by plants Science 2008 319 64 69 18079367 Roig-Villanova I Bou-Torrent J Galstyan A Carretero-Paulet L Portoles S Rodriguez-Concepcion M Martinez-Garcia JF Interaction of shade avoidance and auxin responses: a role for two novel atypical bHLH proteins EMBO J 2007 26 4756 4767 17948056 Scarpella E Barkoulas M Tsiantis M Control of leaf and vein development by auxin Cold Spring Harb. Perspect. Biol 2010 2 a001511 20182604 Schweighofer A Hirt H Meskiene I Plant PP2C phosphatases: emerging functions in stress signaling Trends Plant Sci 2004 9 236 243 15130549 Singh J Roberts MR Fusicoccin activates pathogen-responsive gene expression independently of common resistance signalling pathways, but increases disease symptoms in Pseudomonas syringae-infected tomato plants Planta 2004 219 261 269 15014997 Sozzani R Iyer-Pascuzzi A Postembryonic control of root meristem growth and development Curr. Opin. Plant Biol 2014 17 7 12 24507488 Spartz AK Ren H Park MY Grandt KN Lee SH Murphy AS Sussman MR Overvoorde PJ Gray WM SAUR Inhibition of PP2C–D Phosphatases Activates Plasma Membrane H+-ATPases to Promote Cell Expansion in Arabidopsis Plant Cell 2014 26 2129 2142 24858935 Spartz AK Lee SH Wenger JP Gonzalez N Itoh H Inzé D Peer WA Murphy AS Overvoorde PJ Gray WM The SAUR19 subfamily of SMALL AUXIN UP RNA genes promote cell expansion Plant J 2012 70 978 990 22348445 Stamm P Kumar PP Auxin and gibberellin responsive Arabidopsis SMALL AUXIN UP RNA36 regulates hypocotyl elongation in the light Plant Cell Rep 2013 32 759 769 23503980 Stamm P Ravindran P Mohanty B Tan EL Yu H Kumar PP Insights into the molecular mechanism of RGL2-mediated inhibition of seed germination in Arabidopsis thaliana BMC Plant Biol 2012 12 179 23035751 Staswick PE Tiryaki I Rowe ML Jasmonate response locus JAR1 and several related Arabidopsis genes encode enzymes of the firefly luciferase superfamily that show activity on jasmonic, salicylic, and indole-3-acetic acids in an assay for adenylation Plant Cell 2002 14 1405 1415 12084835 Sullivan ML Green PJ Mutational analysis of the DST element in tobacco cells and transgenic plants: identification of residues critical for mRNA instability RNA 1996 2 308 315 8634911 Sun J Qi L Li Y Chu J Li C PIF4-mediated activation of YUCCA8 expression integrates temperature into the auxin pathway in regulating arabidopsis hypocotyl growth PLoS Genet 2012 8 e1002594 22479194 Sun Y Fan XY Cao DM Tang W He K Zhu JY He JX Bai MY Zhu S Oh E Patil S Kim TW Ji H Wong WH Rhee SY Wang ZY Integration of brassinosteroid signal transduction with the transcription network for plant growth regulation in Arabidopsis Dev. Cell 2010 19 765 777 21074725 Takahashi K Hayashi K Kinoshita T Auxin activates the plasma membrane H+-ATPase by phosphorylation during hypocotyl elongation in Arabidopsis Plant Physiol 2012 159 632 641 22492846 Takatsuka H Umeda M Hormonal control of cell division and elongation along differentiation trajectories in roots J. Exp. Bot 2014 65 2633 2643 24474807 Taniguchi M Nakamura M Tasaka M Morita M Identification of gravitropic response indicator genes in Arabidopsis inflorescence stems Plant Signal Behav 2014 9 e29570 Tatsuki M Nakajima N Fujii H Shimada T Nakano M Hayashi K Hayama H Yoshioka H Nakamura Y Increased levels of IAA are required for system 2 ethylene synthesis causing fruit softening in peach (Prunus persica L. Batsch) J. Exp. Bot 2013 64 1049 1059 23364941 Tian Q Uhlir NJ Reed JW Arabidopsis SHY2/IAA3 inhibits auxin-regulated gene expression Plant Cell 2002 14 301 319 11884676 Tiwari SB Hagen G Guilfoyle TJ Aux/IAA Proteins Contain a Potent Transcriptional Repression Domain Plant Cell 2004 16 533 543 14742873 Tiwari SB Wang XJ Hagen G Guilfoyle TJ AUX/IAA proteins are active repressors, and their stability and activity are modulated by auxin Plant Cell 2001 13 2809 2822 11752389 Tovar-Mendez A Miernyk JA Hoyos E Randall DD A functional genomic analysis of Arabidopsis thaliana PP2C clade D Protoplasma 2014 251 265 271 23832523 Vandenbrink JP Kiss JZ Herranz R Medina FJ Light and gravity signals synergize in modulating plant development Front. Plant Sci 2014 5 563 25389428 Vanhaeren H Gonzalez N Coppens F De Milde L Van Daele T Vermeersch M Eloy NB Storme V Inzé D Combining growth-promoting genes leads to positive epistasis in Arabidopsis thaliana Elife 2014 3 e02252 24843021 Vanneste S Friml J Calcium: The Missing Link in Auxin Action Plants 2013 2 650 675 27137397 Vert G Chory J Crosstalk in cellular signaling: background noise or the real thing? Dev. Cell 2011 21 985 991 22172668 Vi SL Trost G Lange P Czesnick H Rao N Lieber D Laux T Gray WM Manley JL Groth D Kappel C Lenhard M Target specificity among canonical nuclear poly(A) polymerases in plants modulates organ growth and pathogen response Proc. Natl. Acad. Sci. U S A 2013 110 13994 13999 23918356 Walcher CL Nemhauser JL Bipartite promoter element required for auxin response Plant Physiol 2012 158 273 282 22100645 Wang B Du Q Yang X Zhang D Identification and characterization of nuclear genes involved in photosynthesis in Populus BMC Plant Biol 2014a 14 81 24673936 Wang R Estelle M Diversity and specificity: auxin perception and signaling through the TIR1/AFB pathway Curr. Opin. Plant Biol 2014 21 51 58 25032902 Wang S Bai Y Shen C Wu Y Zhang S Jiang D Guilfoyle TJ Chen M Qi Y Auxin-related gene families in abiotic stress response in Sorghum bicolor Funct. Integr. Genomics 2010 10 533 546 20499123 Wang W Bai MY Wang ZY The brassinosteroid signaling network-a paradigm of signal integration Curr. Opin. Plant Biol 2014b 21C 147 153 Wang ZY Bai MY Oh E Zhu JY Brassinosteroid signaling network and regulation of photomorphogenesis Annu. Rev. Genet 2012 46 701 724 23020777 Watillon B Kettmann R Arredouani A Hecquet JF Boxus P Burny A Apple messenger RNAs related to bacterial lignostilbene dioxygenase and plant SAUR genes are preferentially expressed in flowers Plant Mol. Biol 1998 36 909 915 9520281 Wu J Liu S He Y Guan X Zhu X Cheng L Wang J Lu G Genome-wide analysis of SAUR gene family in Solanaceae species Gene 2012 509 38 50 22903030 Xu H Liu Q Yao T Fu X Shedding light on integrative GA signaling Curr. Opin. Plant Biol 2014 21 89 95 25061896 Yamamoto KT Mori H Imaseki H cDNA Cloning of Indole-3-Acetic Acid-Regulated Genes: Aux22 and SAUR from Mung Bean (Vigna radiata) Hypocotyl Tissue Plant Cell Physiol 1992 33 93 97 Yang T Poovaiah BW Molecular and biochemical evidence for the involvement of calcium/calmodulin in auxin action J. Biol. Chem 2000 275 3137 3143 10652297 Yang X Zhang X Yuan D Jin F Zhang Y Xu J Transcript profiling reveals complex auxin signalling pathway and transcription regulation involved in dedifferentiation and redifferentiation during somatic embryogenesis in cotton BMC Plant Biol 2012 12 110 22817809 Yin Y Vafeados D Tao Y Yoshida S Asami T Chory J A new class of transcription factors mediates brassinosteroid-regulated gene expression in Arabidopsis Cell 2005 120 249 259 15680330 Yoshida T Mogami J Yamaguchi-Shinozaki K ABA-dependent and ABA-independent signaling in response to osmotic stress in plants Curr. Opin. Plant Biol 2014 21 133 139 25104049 Yu X Li L Zola J Aluru M Ye H Foudree A Guo H Anderson S Aluru S Liu P Rodermel S Yin Y A brassinosteroid transcriptional network revealed by genome-wide identification of BESI target genes in Arabidopsis thaliana Plant J 2011 65 634 646 21214652 Zhao Y Christensen SK Fankhauser C Cashman JR Cohen JD Weigel D Chory J A role for flavin monooxygenase-like enzymes in auxin biosynthesis Science 2001 291 306 309 11209081 Zhao Y Dai X Blackwell HE Schreiber SL Chory J SIR1, an upstream component in auxin signaling identified by chemical genetics Science 2003 301 1107 1110 12893885 Zurek DM Rayle DL McMorris TC Clouse SD Investigation of Gene Expression, Growth Kinetics, and Wall Extensibility during Brassinosteroid-Regulated Stem Elongation Plant Physiol 1994 104 505 513 12232099
PMC005xxxxxx/PMC5124495.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101465514 34788 Mol Plant Mol Plant Molecular plant 1674-2052 1752-9867 25744359 5124495 10.1016/j.molp.2014.12.023 NIHMS655071 Article Focusing on the focus: what else beyond the master switches for polar cell growth? Qin Yuan 1 Dong Jua 23 1 Center for Genomics and Biotechnology, College of Crop Science, Fujian Agriculture and Forestry University, Fuzhou 350002, Fujian Province, China 2 Waksman Institute of Microbiology, Rutgers the State University of New Jersey, Pistataway, NJ, 08854, USA 3 The Department of Plant Biology and Pathology, Rutgers the State University of New Jersey, New Brunswick, NJ, 08901, USA Correspondence Author: Juan Dong ([email protected]) 18 11 2016 9 1 2015 4 2015 27 11 2016 8 4 582594 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Cell polarity, often associated with polarized cell expansion/growth in plants, describes the uneven distribution of cellular components, such as proteins, nucleic acids, signaling molecules, vesicles, cytoskeletal elements and organelles, which may ultimately modulate cell shape, structure and function. Pollen tubes and root hairs are model cell systems for studying the molecular mechanisms underlying sustained tip growth. The formation of intercalated epidermal pavement cells requires excitatory and inhibitory pathways to coordinate cell expansion within single cells and between cells in contact. Strictly controlled cell expansion is linked to asymmetric cell division in zygotes and stomatal lineages, which require integrated processes of pre-mitotic cellular polarization and division asymmetry. While small GTPases ROPs are recognized as fundamental signaling switches for cell polarity in various cellular and developmental processes in plants, the broader molecular machinery underpinning asymmetric division-required polarity establishment remain largely unknown. Here, we review the widely used ROP signaling pathways in cell polar growth and the recently discovered feedback loops with auxin signaling and PIN effluxers. We discuss the conserved phosphorylation and phospholipid signaling mechanisms for protein uneven distribution, as well as the potential roles of novel proteins and MAPKs in the polarity establishment related to asymmetric cell division in plants. Short Summary Establishing cell polarity is required for many developmental processes in plants, including sustained tip extension (pollen tubes and root hairs), diffuse cell growth (pavement cells) and regional cell expansion for asymmetric cell division (zygotes and stomatal lineages). We summarize the small GTPases-centered molecular mechanisms underpinning plant cell polarity and discuss the potential roles of other signaling pathways in asymmetric cell expansion. cell expansion cytoskeleton polarity polarity determination signal transduction Introduction Cell polar growth, including sustained unidirectional extension (polar growth) and diffused multidirectional expansion (diffuse growth), is the basis of growth and development of an organism. Polar growth, also termed as tip growth, is achieved by localized delivery of molecular and cellular materials to the growth site, and is often required for the generation of highly elongated tubular cells, such as fungal hyphae, animal neurons, plant root hairs and pollen tubes. Diffuse growth is a more universal form of cell expansion and found in almost all cell types in eukaryotic kingdoms. Although a large number of cell expansion undergo isotropic diffuse growth (the cell extension in all directions along the cell surface) (Kropf et al., 1998), here we consider and discuss anisotropic diffuse growth, the hallmark of plant cells, which describes the differential expansion of certain region or position of the cell. The molecular mechanism for the establishment and maintenance of cell polarity is one of the most fundamental and actively studied frontier areas in cell and developmental biology. Prior to visible cell polar growth, the establishment of cell polarity is necessary. Cell polarity is represented by the asymmetrical distribution of molecules, organelles or cytoskeletal elements along a particular axis of the cell (Grebe et al., 2001). Such organization is generally required for polar cell expansion during morphogenesis and/or specifying a distinct sub-region to fulfill a specific function in differentiated cells. While decades of work has led to extensive progress in deciphering the molecular components and genetic mechanisms governing cell polarity and regional cell growth in plants (Kania et al., 2014; Yang and Lavagi, 2012), recent research made an exciting wave of discoveries linking cell polarity to the establishment of physical and subsequent asymmetrical cell fate during cell division (Facette and Smith, 2012), which is of crucial importance for organ initiation, morphogenesis and development. Albeit differing from other organisms in lifestyle, body organization and cellular structure, plants seem to have evolved conserved core mechanisms (small GTPases) for cell polarity control (Thompson, 2013; Yang, 2008). This review summarizes recent progresses made in several plant model systems in studying the molecular mechanisms of polar cell expansion, and provides insights in to understanding their molecular linkages to specific developmental processes. Model Systems for Studying Polar Cell Expansion in Plants Benefiting from the amiability to experimental manipulations, as well as ample genetic and molecular markers, quite a few excellent model cell systems in plants, particular in Arabidopsis, have been established for studying cell polarity and polar growth. Root hair cells and pollen tubes are excellent model systems for tip growth (sustained, fast cell extension behavior) (Cardenas, 2009; Guan et al., 2013; Qin and Yang, 2011). Leaf epidermal pavement cells have been established as an ideal system for cell-cell coordination of interdigitated cell expansion (Chen and Yang, 2014; Lin et al., 2014). The stomata lineages and zygotes require cellular polarization for successful asymmetric cell division, therefore are emerging as new systems for studying cell polarity (Dong and Bergmann, 2010; Zhang and Laux, 2011). Root hairs and pollen tubes: model systems for sustained tip growth Root hairs (RHs), originated from a subset of root epidermal cells, extend fast by tip growth at the rates of 10–40 nm/s (Galway et al., 1997). Root hairs are important in sensing external biotic/abiotic conditions and nutritional status. Pollen tubes are developed from pollen grains after landing on the floral stigma. The tube tip (growth speed at ~300 nm/s) interacts with and penetrates through several types of pistil tissue to deliver sperm cargos for double fertilization. Both pollen tubes and root hairs are easy to grow and accessible to both genetic and cell biological analysis. These tip growing systems commonly require highly polarized intracellular organization and demand continuous exocytosis, which delivers cell membrane and wall materials to the apical dome for sustained growth (Hepler et al., 2001; Qin et al., 2007). The identified genes from these systems suggest that a Rho GTPase-based self-organizing signaling network interconnected with calcium homeostasis, F-actin cytoskeleton dynamics and polarized exocytosis plays a center role in the strictly controlled tip growth (Guan et al., 2013). Root hair growth also involves the production of reactive oxygen species (ROS) (Carol and Dolan, 2006; Molendijk et al., 2001). The clathrin-dependent endocytosis and cell wall modifications were also shown important for the pollen tube self-organizing growth (Bosch and Hepler, 2005; Rockel et al., 2008; Zhao et al., 2010). Pavement cells: interdigitating diffuse growth Unlike unicellular root hair and pollen tube, pavement cells in the leaf epidermis provide an excellent platform for investigating the mechanisms for cell shape determination, during which coordinated cell growth within a single cell and between adjacent cells is necessary and regulated by both developmental and intercellular signals (Smith, 2003; Wasteneys and Galway, 2003; Yang, 2008; Yang and Fu, 2007). Particularly, establishing cell polarity is critical for the morphogenesis of leaf epidermal pavement cells. Their jigsaw-puzzle appearance results from the intercalary growth of lobes and indentations, reminiscence of the convergent extension in animal cells (Price et al., 2006; Settleman, 2005). Accumulating evidence suggested that a Rho GTPase-orchestrated and cytoplasmic auxin signaling-mediated reorganization of cortical microtubules and fine actin microfilaments underpins the formation of cellular interdigitation in plants (Fu et al., 2005; Lin et al., 2012; Xu et al., 2014; Xu et al., 2010). Zygotes and stomatal lineage cells: cell expansion for asymmetric division Asymmetric cell division (ACD) is at the heart of plant development for continuously generating diverse cell types. One prerequisite for a successful ACD is cell polarization (De Smet and Beeckman, 2011), represented by unequal distribution of cellular components (no striking cell shape change, e.g. some ACDs in the root apical meristem (Petricka et al., 2009)) or by anisotropic cell expansion (elongation required, e.g. zygotes and stomatal lineages). Plants start their life from a single cell, zygote, formed by the fusion of an egg cell and a sperm cell. A zygote undergoes profound cellular reorganization to establish an apical-basal axis prior to division: from a symmetric state to an asymmetric stage with a large vacuole re-assembled at the basal end and the nucleus migrated to the apical end (Faure et al., 2002). The first asymmetric division then generates two daughter cells with distinct morphology and perspectives. The small apical daughter cell gives rise to the apical embryo lineage, and the large vacuolated basal daughter cell divides to produce an extra-embryonic structure, suspensor. Brown algae used to be a favorable system for studying zygote polarity because large populations of free-living and synchronously developing embryos can be obtained and the zygote polarity can be induced and altered by external signals (Brownlee and Bouget, 1998). The moss Physcomitrella patens has two types of bodies: a hypha-like body (protonema, undergoes tip growth) and a shoot-like body (gametophore, develops stems and leaves and requires ACD at the apical region), thus becomes a useful model system for studying polar cell growth and ACD (Vidali and Bezanilla, 2012). Facile tools are available for genic and genomic analysis in Physcomitrella (Strotbek et al., 2013). Arabidopsis zygotes for their nearly invariant division pattern have become an ideal system for studying the molecular basis underpinning cell polarity and asymmetric division at the initial and fundamental stage in plant development (Gallois, 2001). It was found that transcriptional programs in connection with the positional-cue regulated MAPK signaling regulate zygote polarity for the first asymmetric cell division (Lukowitz et al., 2004; Ueda et al., 2011). Stomata are gas exchange valves between plants and the atmosphere. Two guard cells surround to form a stomatal pore. The initiation and differentiation of guard cell population require a series of strictly controlled asymmetric cell division in both monocot maize and dicot Arabidopsis (Dong and Bergmann, 2010). In Arabidopsis, the ACD precursor cell, Meristemoid Mother Cell (MMC), expands to set up asymmetry; a procedure involves organelle shifting, nuclear migration, and asymmetric placement of the PPB. In maize, a stomatal complex is composed of a pair of guard cells (GCs) flanked by a pair of subsidiary cells (SCs). The formation of SCs requires the precursor cells, subsidiary mother cells (SMCs), to polarize and divide asymmetrically. It was proposed that external positional cues may guide the SMCs’polarization to divide asymmetrically (Facette and Smith, 2012). In both stomatal systems, cell polarity is obviously linked to patterned stomatal asymmetric divisions (Facette and Smith, 2012). In Arabidopsis, the bHLH transcription factor SPEECHLESS (SPCH) determines the initiation of the stomatal lineage asymmetries (MacAlister et al., 2007). The novel proteins BREAKING OF ASYMMETRY IN THE STOMATAL LINEAGE (BASL) (Dong et al., 2009) and POLAR LOCALIZATION DURING ASYMMETRIC DIVISION AND REDISTRIBUTION (POLAR) (Pillitteri et al., 2011)are employed to polarlyenrichat the cortical PM region during stomatal ACD. Another set of polarity proteins from maize were identified, PANGLOSS1 (PAN1) (Cartwright et al., 2009) and PAN2 (Zhang et al., 2012) receptor-like proteins. The PAN pathway has been linked to the functions of ROP signaling and actin organization (Humphries et al., 2011). Establishing Polarity: Cellular Machinery and Molecular Components in Symmetry Breaking In eukaryotes, including yeast, plants and animals, the establishment and maintenance of cell polarity are associated with common cellular components, such as the cytoskeletal elements, the endomembrane system and polarity molecules (Kania et al., 2014; Thompson, 2013). Due to the specialized wall structures, plants appear to have evolved their own polarization machineries and mechanisms, including cell wall modification, polar auxin transportation (Fowler and Quatrano, 1997; Grunewald and Friml, 2010)and plant-specific polarity factors, e.g. small GTPase ROPs (Rho of plants) (Yang and Fu, 2007) and BASL/POLAR ((Dong et al., 2009; Pillitteri et al., 2011). ROP GTPases are master molecular switches controlling cell polarization by orchestrating the behaviors of cytoskeleton, vesicle trafficking and calcium signaling (Craddock et al., 2012). In contrast to the much better understood cell polarization regulated by ROPs, how other polarity proteins coordinate asymmetric cell divisions in plants remain largely unknown. Other cellular components and signaling molecules, e.g. phospholipids and protein kinases, seem to be conserved and play important roles in symmetry breaking in both animal and plant cells. Small GTPases Small GTPases contain a large group of hydrolases implicated in a broad range of cellular events. GTPases function as biological switches, which constitutively cycle between the active GTP-bound form and the inactive GDP-bound form. Their activation is regulated by the guanine exchange factor (GEF) that promotes the substitution of GDP-to-GTP and the deactivation process is mediated by GTPase-activating proteins (GAPs) that stimulate the hydrolysis of GTP-to-GDP. Another type of negative regulators, guanine nucleotide dissociation inhibitors (GDIs) prevents the PM association and nucleotide exchange of Rho GTPases (Nagawa et al., 2010). GDIs also play a role in the polarized accumulation of Rho GTPases by mediating the recycling of Rho GTPase to specific membrane domains (Klahre et al., 2006; Lin et al., 2003). The Rho GTPase super family has evolved to several subfamilies, including Cdc42, Rac, Rho and plant specific ROP (Rho GTPase in plants). The Arabidopsis genome encodes 11 ROPs, and each seemed to function differently (Yang, 2002), with the best-described regulation in cell polarity and morphogenesis. ROP1, ROP3 and ROP5 redundantly control tip growth in pollen tubes (Gu et al., 2003; Li et al., 1999), while ROP2 overlaps with ROP4 to modulate tip growth in root hairs (Duan et al., 2010; Molendijk et al., 2001). A recent report showed that ROP3 also plays important roles in embryo development via maintaining PIN polarity at the PM (Huang et al., 2014). AlthoughROP6 and ROP2 are almost identical in sequence (94%), they function distinctly in the formation of interdigitated pavement cell shape. ROP2 promotes the formation of cortical diffuse F-actin and thus lobe outgrowth, whilst ROP6 rearranges MT organization to restrict cell growth and enhance indentation (Fu et al., 2005; Fu et al., 2009). These two antagonistic Rho GTPase pathways are controlled by feedback loops of ROPs with auxin signaling and regulated by Auxin-binding protein 1 (ABP1) and transmembrane kinase (TMK) receptors (Xu et al., 2014). Interestingly, ROP11 was recently found to regulate MT reorganization for secondary cell wall patterning in xylem cells (Oda and Fukuda, 2012). Cytoskeleton dynamics Microtubules (MTs) and actin filaments (F-actin) are linear proteinaceous polymers that comprise the complex cytoskeleton network in plant cells. MTs are frequently found aligned with wall microfibrils that are arranged in an orientation transverse to the direction of cell expansion (Baskin, 2001). Chemical drugs that disturbed MT structure/organization induced plant cells to lose their polarity and become isotropically swollen. These defects suggested that MT is essential for establishing and maintaining growth directionality (Mathur and Chua, 2000). Another important function that MTs involve in is the Preprophase band (PPB) formation. The PPB is composed by a highly ordered cortical array of MTs and marks the cortical division site (CDS) that corresponds to the future site of cell division plane (Pickett-Heaps and Northcote, 1966). During the PPB formation, cortical F-actin assembles alongside the PPB MTs and helps them to condense at CDS. Once assembled, the PPB MTs organize proteins and lipids at the CDS to guide the orientation of the phragmoplast extension for accurate cell-plate formation (Rasmussen et al., 2013). The PPB formation per se is not significantly affected by cell polarity, but the position of the PPB placement is likely guided by the polarity cue that induces and establishes cellular asymmetry (Facette and Smith, 2012). Functions of F-actin in cell polarity control are tightly associated with vesicle trafficking and deposition of materials to the plasma membrane (PM). In animals, F-actin, together with the associated proteins (such as spectrin, ankyrin and myosin) and the regulatory proteins (e.g. the small GTPase CDC42), helps to assemble vesicles at the Golgi and endosomes and to transport them across the cytoplasm (Musch et al., 2001). In the tip growing systems, the polarlylocalized fine F-actin at the apex is very dynamics and delivers signaling molecules, e.g. the Rho GTPases and their activators, to the apex, which in turn promotes the polymerization of F-actin (Figure 1a). Such a F-actin dynamics and vesicle trafficking coordinated feedback regulation is critical for robust cell polarity, not only conserved in yeast and animals, but also in plants (Charest and Firtel, 2006). Examples can be found in polarized growth of many plant cells. During trichome branching, the actin-dependent morphogenesis is regulated by SPIKE1 (SPK1), a ROP GEF, which activates the F-actin nucleating machinery comprised of theARP2/3 complex and its activator, the WAVE complex (Basu et al., 2008). In pollen tubes, the long actin cables, extending longitudinally through the shank region but not into the sub apical region of a tube, provide main tracks for intracellular delivery of organelles and vesicles (Cai and Cresti, 2009). F-actin appears to have more complicated roles in diffuse growth. The fine and bundled F-actins are localized differently to organize the MT orientation for restricted or promoted cell expansion, respectively (Figure 2a) (Mathur, 2006; Saedler et al., 2004). The indentation areas contain dense F-actin mesh and the vesicles with limited movement, whilst a fine F-actin mesh favors vesicle delivery, therefore promotes cell growth in the protruding regions. MTs co-localize with dense F-actin mesh and provide more physical support to enforce cell polarity (Mathur, 2006). Vesicular trafficking Many molecular components required for cell growth, such as proteins, enzymes, phospholipids, and polysaccharides, are produced, modified and transported through the endomembrane system, from which the vesicles containing cargo molecules are derived. The vesicles transit through the cytosol and arrive at the destination membrane, where the cargo molecules are released by membrane fusion. Therefore, polarized cell grow this highly dependent on vesicular trafficking that supplies the flow of macromolecules to the growth site (Kania et al., 2014). Vesicular trafficking in the secretory exocytosis and recycling endocytosis pathways are highly dynamic and key to the PM integrity. Exocytosis is mediated by an evolutionarily conserved complex, the exocyst, which is responsible for vesicle tethering to the PM (Hala et al., 2008). Mutations in the exocyst subunits resulted in defective tip growth (Cole et al., 2005; Synek et al., 2006). Other regulatory components involved in vesicle budding and docking, ARF (ADP-ribosylation factors) GTPases and Rab-GTPases, respectively, are also critical for tip growth (Preuss et al., 2006; Wen et al., 2005). In addition, the spatial and temporal changes of exocytosis are coordinated by the activities of cytoskeleton, ROP GTPases, calcium and phospholipids in root hairs and pollen tubes (Lee et al., 2008; Monteiro et al., 2005). Excessive signaling molecules and wall materials deposited into the plasma membrane must be retrieved by endocytosis (Goldstein et al., 1979). The clathrin-mediated endocytosis (CME) is the predominant endocyticroute in plants and has tremendous impacts on establishing and sustaining cell polarity (McMahon and Boucrot, 2011). This can be well exemplified by the process of PIN (PIN-FORMED) protein polarization. PIN proteins are auxin transporters that are polarly localized in plant cells to drive directional auxin flow. Blocking endocytosis by inhibitors resulted in lateral diffusion of PIN proteins from the polar domains. Recent studies illustrated that the auxin-dependent endocytosis of PIN proteins is under the orchestrated activities of the putative auxin receptor ABP1 and ROP GTPases (Chen et al., 2012; Lin et al., 2012; Murphy and Peer, 2012; Nagawa et al., 2012). Phosphorylation and phospholipid signaling Besides the small GTPase-centered cell polarization mechanisms, protein phosphorylation has been recognized as an important mechanism to regulate protein polarization in both animals and plants. In animals, the conserved PAR proteins are fundamental players to regulate cell polarity in different developmental and functional contexts (Goldstein and Macara, 2007). In C. elegans embryos, PAR polarity regulators are sorted into anterior and posterior cortical domains. The anterior PAR kinase (aPKC) phosphorylates the posterior PAR-2 to prevent its cortical association (Hao et al., 2006), and the posterior PAR-2 recruits PAR-1 kinases, which phosphorylates and locally dissembles anterior PAR-3 to prevent its accumulation at the posterior domain (Cuenca et al., 2003). Therefore, protein phosphorylation-mediated mutual exclusion plays a pivotal role in segregating the PAR polarity domains. Interestingly, in plant cells, polar localization of PIN proteins to the apical or basal domains of the PM also relies on their phosphorylation status (Friml et al., 2004; Huang et al., 2010; Zhang et al., 2010). The AGC-3 serine/threonine kinases, PINOID (PID), WAVY ROOT GROWTH1 (WAG1) and WAG2, redundantly phosphorylate PIN proteins to direct them to the apical membrane (Dhonukshe et al., 2010; Michniewicz et al., 2007; Sukumar et al., 2009). On the other hand, the trimeric protein phosphatase 2A (PP2A) and D6 protein kinase (D6PK) function antagonistically to dephosphorylate and phosphorylate PINs, respectively, and target them to the basal side (Dai et al., 2012; Michniewicz et al., 2007; Zourelidou et al., 2009). The loss-of-function mutations in these kinases and phosphatases commonly led to aberrant embryo patterning or agravitropic root growth, recapitulating the defects in the auxin transport mutants (Dhonukshe et al., 2010; Sukumar et al., 2009). Phospholipid signaling has also been recognized as a signature event in cell polarization. Different phosphoinositides are found to associate with different membrane compartments, e.g. PI(4,5)P2 and PI(3,4,5)P3 are generally found at the PM (Orlando and Guo, 2009). Multiple PAR proteins, including PAR-1, PAR-2 and PAR-3, contain motifs that can bind to phospholipids (Moravcevic et al., 2010; Wu et al., 2007a). Some lipid-binding domains are enriched with basic amino acids, raising the possibility of electrostatic interactions mediating the PAR proteins to associate with the PM (Hao et al., 2006; Krahn et al., 2010). In Drosophila epithelial cells, phosphoinositides, e.g. phosphatidylinositol- 4,5- diphosphate (PI(4,5)P2), are asymmetrically distributed. Inactivating the PI4P5 kinase Skittles (SKTL) resulted in strongly reduced PI(4,5)P2 levels in the epithelium, which disturbed the apical targeting of PAR-3 to the PM followed by cell polarity and shape defects (Claret et al., 2014). Interestingly, PI(4,5)P2 seemed to play a role in polarizing PIN proteins in plants as well. PINOID, the AGC kinase targeting PINs, is recruited to the PM through interacting the upstream kinase, 3-phosphoinositide- dependent protein kinase 1 (PDK1), which contains a pleckstrin homology (PH) domain that binds to PI(4,5)P2 phospholipids (Zegzouti et al., 2006). Thus, functions of membrane-localized phospholipids are connected to PID and PINs for auxin-mediated downstream signaling events (Anthony et al., 2004). In root hairs and pollen tubes, phospholipids are implicated in multiple events to regulate cell polarity, including actin cytoskeleton organization, clathrin-dependent endocytosis, vacuolar degradation and ion channel permeability (Ischebeck et al., 2011; Zhang et al., 2011; Zhao et al., 2010). Molecular Mechanisms Underpinning Polar Growth in Plants Differing from other mobile life styles, plants are sessile. Plant cells, once produced, are fixed in position and share cell walls with the neighboring cells, and the rigid cell walls impose spatial constrains on cell expansion and polarization. In response to various developmental and environmental cues, plant cells are capable of constantly adjusting their polar domains, which naturally requires intricate and flexible molecular mechanisms that modulate cell polarity. More than a decade of work disclosed that a few key principles in polarity signaling, e.g. positive feedback loops and mutual antagonism, are commonly applicable to animals, fungi and plants. Tip growth: ROP signaling integrates cytoskeleton dynamics, calcium gradient and ROS production The apical region of tip growth is defined by active ROP GTPases, which activate multiple downstream pathways to define the polar site and to precisely control exocytosis (Li et al., 1998; Molendijk et al., 2001). Disruption of active ROP GTPase inhibited tip growth, while overexpression of a constitutively active form, but not a native form, induced balloon-shaped (de-polarized) growth in root hairs and pollen tubes (Jones et al., 2002; Li et al., 1999). These data suggested that both positive and negative regulations are required for maintaining an optimal level of active ROPs to sustain polar growth. The rapid tip growth in root hairs and pollen tubes is restricted to the very apex and interestingly exhibits an oscillating manner (episodes of fast growth followed by slower growth) (Monshausen et al., 2008). These oscillatory phenomena were widely used in ordering of signaling events, but are not necessarily essential to growth. In pollen tubes, such growth oscillation is led by an oscillation of ROP1 activity (Figure 1b) (Hwang et al., 2005). ROP1 oscillation can be well explained by the activities of its downstream effectors, RIC3 and RIC4, through two counteracting pathways: 1) the RIC4 pathway promotes F-actin assembly and 2) the RIC3 pathway is related to calcium signaling and promotes F-actin disassembly (Gu, 2005). ROP1 also activates a downstream effector, RIP1/ICR1, which promotes the accumulation of exocytic vesicles to the growing tip (Lavy et al., 2007). These vesicles may contain ROP1 upstream components, such as Rop GEFs, PRK1 and PRK2, to further activate ROP1(Gu, 2006; Kaothien et al., 2005; Zhang and McCormick, 2007), forming a positive feedback loop. In the meanwhile, to achieve unidirectional growth, negative feedback mechanisms were found to limit the lateral propagation of ROP1 signaling. A global inhibition mechanism governed by ROP1 negative regulators, RopGAPs (RopGTPase activating proteins) and RhoGDIs (Rho guanine nucleotide dissociation inhibitors), was demonstrated (Hwang et al., 2008). Alateral inhibition mechanism mediated by ROP1 negative regulators was also proposed to act at the tube flank region to limit the lateral propagation of active ROP1 (Hwang et al., 2010). It has been well characterized that, in pollen tubes and root hairs, cytosolic calcium forms a tip-focused gradient and the elimination of it leads to growth arrest (Pierson et al., 1996). Interestingly, the concentration of apical calcium also fluctuates following the ROP1 activity oscillation (Hwang et al., 2005). It was suggested that ROP1 activates the downstream effector RIC3, which then mediates the influx of calcium across the PM. The tip-focused calcium in turn promotes F-actin dynamics to elevate exocytosis and ROP1 activities, forming a positive feedback loop. On the other hand, elevated calcium may suppress ROP1 activity via the negative feedback regulation, either through F-actin disassembly or through RhoGAPs (Gu, 2005; Yan et al., 2009). Then, how calcium gradient and oscillation are perceived? A couple of putative sensor proteins, such as calmodulin (CaM), CaM-like (CML), calcium-dependent protein kinase (CDPK), and calcineurin B-like protein (CBL), were found to mediate the downstream signaling responses and required for pollen tube growth (Myers et al., 2009; Rato et al., 2004; Yoon et al., 2006). However, the regulatory mechanisms of calcium sensors and their relationship with ROP1 signaling remain unclear. Another signaling pathway that integrates ROP and calcium in polar growth is by ROS (reactive oxygen species) production. Intracellular and extracellular ROS play important roles in multiple cellular responses, including the hypersensitive response, programmed cell death, hormonal signaling, stomata opening and ion channel activity (Mittler and Berkowitz, 2001). Plant NADPH oxidases, the key enzymes in the generation of superoxide radicals, are membrane bound proteins and polarized to the tips, the elimination of which causes the inhibition of root hair and pollen tube growth (Foreman et al., 2003; Jones et al., 2007; Kaya et al., 2014; Potocky et al., 2007). The roles of ROS were connected to calcium signaling, modulating cell wall properties and ROP feedback regulations (Kosami et al., 2014). To further elucidate how ROS determines cell polarity for growth, investigating the relationship between ROS production and ROP GTPase activation at specific site in polarized cells become necessary. Diffuse growth: auxin signaling linked to ROP-centered interdigitating cell growth In contrast to the tip-growing unicellular systems, interdigitation of pavement cells in the leaf epidermis involves coordinated intracellular and cell-cell signaling. Before pavement cell expands and matures, multiple polarity sites define the interdigitated regions for the formation of complementary lobes and indents (Figure 2a). The lobes of one cell expand into the indents of the neighboring cells, generating the jigsaw-puzzle appearance of pavement cells. How multiple polarity sites are initiated and coordinated within and between cells are of fundamental importance for unraveling the developmentally programmed cellular events. Intriguingly, recent progresses bridged the phytohormone auxin signaling with the ROP-centered polarity establishment and brought new insights into the important roles of self-organizing auxin signaling in promoting the intercalary growth of pavement cells (Figure 2b) (Lin et al., 2013; Nagawa et al., 2012; Xu et al., 2014; Xu et al., 2010). Experimental data showed that the formation of lobes and indentations in leaf pavement cells was promoted by auxin and compromised in the mutants defective in auxin biosynthesis or perceiving (Xu et al., 2010). It was found that the extracellular auxin is perceived by a cell surface receptor complex composed of the PM-associated auxin binding protein 1 (ABP1) and its partner, the transmembrane receptor–like kinases (TMKs) (Xu et al., 2014). Interestingly, the cytoplasmic auxin signaling activates two Rho GTPases, ROP2 and ROP6, which are polarized at lobes and indentations, respectively, to antagonistically promote or restrict local cell expansion during morphogenesis (Figure 2b) (Fu et al., 2005; Xu et al., 2010). Once activated by auxin signaling, ROP2 at the lobe protruding region interacts with RIC4 to promote the assembly of cortical F-actin for localized lobe outgrowth, in part resembling the behavior of ROP1 in pollen tube tips (Fu et al., 2005; Gu, 2005). On the other hand, the activated ROP6 in the indentation region directly activates RIC1 to promote well-ordered cortical MTs that locally restrict cell expansion to reinforce the indentation formation (Fu et al., 2005; Fu et al., 2009). Recently, a MT-severing protein, Katanin (KTN1), directly binds to RIC1 to detach branched MTs, further promoting MT ordering in the indentation procedure (Lin et al., 2013). Furthermore, as the antagonistic ROP6-RIC1 and ROP2-RIC4 pathways need to coordinate and reinforce the intercalated patterning, the lobe localized ROP2 inactivates RIC1 to suppress the well-ordered cortical MTs formation in the lobe regions (Fu et al., 2005). Meanwhile, the indentation localized ROP6-RIC1 triggered MT arrays to locally suppress ROP2 activation (Fu et al., 2005; Fu et al., 2009). Therefore, the jigsaw-puzzle shape of pavement cells is strictly regulated by a mutually exclusive system of antagonistic activity of ROP2-RIC4 and ROP6-RIC1 in the lobes and the indentations, respectively (Figure 2b). In principle, ROPs and PINs, as two classes of polarity proteins, control cell polarization in different manners and regulate different developmental aspects. Intriguingly, recent progress further emphasized that these two pathways do crosstalk (Hazak et al., 2010; Huang et al., 2014; Nagawa et al., 2012). In pavement cells, PIN1 is preferentially localized to the PM of the lobe regions, overlapping with ROP2, and mutations in PIN1 resulted in reduced ROP2 activity and defects in pavement cell interdigitation (Nagawa et al., 2012; Xu et al., 2010). The auxin-dependent local activation of ROP2 at the lobe tip, through RIC4-meditated F-actin accumulation, inhibits PIN1 endocytosis and promotes PIN1 enrichment at the lob tip, which further exports more auxin for ROP2 activation. Therefore, auxin-ABP1/TMKs-ROP2-PIN1 forms a positive feedback loop that reinforces the ROP2 self-organization and cell polarization (Nagawa et al., 2012; Xu et al., 2010). Another component ICR1, a Rho scaffold protein, likely functions downstream of ROP signaling and links the secretory system to the recruitment of PIN proteins to the PM polar domains (Hazak et al., 2010). Asymmetric cell division (ACD): MAPKs, transcription factors, auxin signaling polarity proteins, and ROPs in establishing cell polarity ACD provides the molecular basis for cell type diversification in plant development; therefore understanding the underpinning molecular mechanisms is critical. Among a few major cell systems for studying ACD (De Smet and Beeckman, 2011), the division of zygotes and stomatal lineage cells requires the precursor cell to elongate (symmetry breaking) and the division plane to be placed asymmetrically. However, in contrast to the above well studied polarity systems toward understanding the cellular machineries for morphogenesis, our knowledge of the cell polarization mechanisms for asymmetric cell division remains limited. Here, we will first discuss the possible functions of MAPK pathway and auxin signaling, at the transcriptional level and at the sub cellular level, for the establishment of cell polarity during asymmetric division of Arabidopsis zygotes. We will then discuss the functions of polarized proteins in stomatal lineage cells and their potential connections to the ROP signaling. While seeking genetic regulators in embryogenesis, a few mutants defective in the first zygote ACD were isolated in Arabidopsis, suggesting a pathway composed of the PM receptor kinases, downstream MAPK components, and nuclear transcription factors (Bayer et al., 2009; Jeong et al., 2011; Lukowitz et al., 2004). The common defects of these mutants were the failure in zygote elongation prior to the asymmetric cell division and the subsequent abnormal embryo patterning. The involved canonical mitogen-activated protein kinase (MAPK) pathway is composed of three tiers of kinases, the MAPKKK YODA (YDA), MAPK Kinases 4 and 5 (MKK4/5), and MAPK 3 and 6 (MPK3/6). This pathway is also critical for other developmental processes that require ACD, e.g. stomatal development (Bergmann et al., 2004)and root apical meristem organization (Smekalova et al., 2014). In the loss-of-function mutants of the YDA MAPK pathway, zygote elongation is compromised and two daughter cells are about equal size, leading to disturbed developmental fate adoption (Lukowitz et al., 2004). Upstream of YDA, the PM-located interleukin-1 receptor-associated kinase/Pelle-like kinase, SHORT SUSPENSOR (SSP) a paternal-derived signaling molecule, delivers signals to the cytoplasm (Smekalova et al., 2014). The extracellular peptides, Embryo Surrounding Factor 1 (ESF1), derived from the maternal central cell were placed upstream of YDA, working synergistically with SSP, to regulate early embryo patterning (Costa et al., 2014). Downstream of YDA, it was found that the putative RWP-RK transcription factor GROUNDED (GRD), also called RKD4 ((Waki et al., 2011)), functions in the nucleus to promote zygote elongation and the basal cell fate (Jeong et al., 2011). The molecular mechanisms of the YDA-centered pathway in promoting zygote cell expansion and ACD still remain elusive, but the downstream events may include the activation of the transcription factor WRKY2, which acts upstream of other transcription factors, such as the WOX family. The plant-specific WUSCHEL family transcription factors, WOX (WUSCHEL HOMEOBOX), were found to control the differential lineage fate after the zygote ACD (Figure 3a) (Haecker et al., 2004). WOX2 determines the apical lineage (embryo) patterning, whereas WOX8 (also named STIMPY-LIKE)together with its close homolog WOX9 (also called STIMPY) (Wu et al., 2007b) specify the basal lineage (the extra-embryonic suspensor) and also the apical lineage via non-cell autonomous activation of WOX2 (Breuninger et al., 2008). Prior to the zygote ACD and upstream of WOX8, a plant-specific zinc-finger transcription factor WRKY2 determines the asymmetric organelle distribution and subsequent division of the zygote, suggesting an intriguing connection between the WRKY2 and WOX8transcription factors and the downstream effectors in establishing cell polarity (Ueda et al., 2011). WRKY2 contains several consensus MAP kinase phosphorylation sites and its homologues, WRKY33 and WRKY34, are found targeted by MPK3 and MPK6(Guan et al., 2014; Mao et al., 2011). Although WRKY2 has been genetically placed downstream of MPK6 in pollen development (Guan et al., 2014), the direct connection between the YDA pathway and WRKY2 and the genetic network that WRKY2 controls in the process of zygote polarization a waits further investigation (Figure 3a). Is auxin involved in establishing zygote asymmetry? The early involvement of auxin is indicated by the lack of zygote elongation and occasional symmetric divisions caused by the mutations in Arabidopsis GNOM (GM), which encodes an ARF/GEF factor for vesicle trafficking of PIN1 (Friml et al., 2003; Geldner et al., 2004). In addition, the expressions of IAA, an auxin response factor, and ABP1 receptor were detected in the tobacco zygote, and the polar distribution and transportation of auxin begins at the zygote stage and affects the following asymmetric division (Chen et al., 2010). Although, after the first zygotic division, auxin was found to transport upwards to the apical cell, likely by apically localized PIN7 in the basal cell (Friml et al., 2003), it remains to be determined whether auxin regulates zygote cell polarity before the division. Besides auxin, otherpolarly distributed developmental determinants may also be important for zygote asymmetric division. For example, polarlyenriched actin patches, vesicles and cell wall modifiers involve in axis fixation in the zygote of Fucus (Kropf, 1997). In tobacco, specific arabinogalactan proteins (AGPs), potentially implicated in cell expansion, were preferentially localized at the apical pole of tobacco zygotes. Inactivating AGP activity by chemical inhibitors in in vitro cultured zygotes frequently resulted in more symmetrical divisions (Qin and Zhao, 2006). In stomatal ACD, the direct evidence supporting the critical roles of polarity elements came from the identification of the landmark proteins, BASL/POLAR in Arabidopsis (Dong et al., 2009; Pillitteri et al., 2011)and PAN1/PAN2 in maize (Cartwright et al., 2009; Humphries et al., 2011; Zhang et al., 2012). Both BASL and POLAR proteins are unknown of function and these two proteins largely overlap in accumulating into crescents at the cell cortex of stomatal ACD precursor cells (Figure 3b). Based on protein localization, loss-of-function mutant and overexpression phenotypes, BASL’s function was proposed to induce cortical cell expansion in the process of establishing physical asymmetry in the ACD precursor cell (Dong et al., 2009). As BASL-induced regional cell expansion was greatly compromised while ROP signaling was disturbed in Arabidopsis, it was also hypothesized that BASL might signal through or crosstalk with the ROP pathways to control cell polar expansion (Dong et al., 2009). But the key link between BASL and ROPs is still missing. Then, what activates the polarization of BASL and POLAR? Recently, Lau et al., showed that the SPCH transcription factor directly binds to the promoter regions of BASL and POLAR, as well as those of many other key stomatal regulators, emphasizing the direct roles of SPCH in controlling stomatal division and fate differentiation (Lau et al., 2014). This work did not address how BASL polarity is induced, but intriguingly one of the SPCH targets, ARK3/AtKINUa, encodes a plant-specific kinesin, the expression of which was found at the PPB of the asymmetrically dividing stomatal lineage cells. Interfering the ARK3 function by artificial microRNA resulted in disturbed stomatal ACD, phenocopying that of a basl mutant (Lau et al., 2014). It would be interesting to elucidate whether and how BASL and ARK3 may function coordinately to achieve division asymmetries in Arabidopsis. But, as BASL polarity appears prior to the asymmetric division and presumably before the appearance of ARK3 at the PPB, dissecting the BASL/POLAR polarity complex might more directly address how cellular asymmetry and polar cell expansion are achieved at the molecular level. In maize stomatal development, PAN1 and PAN2 are receptor-like proteins expressed in SMCs and polarized to the plasma membrane region adjacent to GMCs. In pan1 and pan2 mutants, the polarization of SMC divisions are affected, likely due to the defects of SMCs in perceiving the extrinsic signal from GMCs and/or the failure in transducing the polarization signal internally (Cartwright et al., 2009; Zhang et al., 2012). It is still unknown how PAN proteins are polarized, but new insights have been provided by their functional connection with ROP signaling and F-actin organization (Humphries et al., 2011). It was found that maize ROPs (2 and 9) physically bind to PAN1 and are asymmetrically distributed to the SMC-GMC contact site (Figure 3c). Enriched ROPs may stimulate local F-actin assembly to form patches and promote the localized accumulation and fusion of vesicles (Humphries et al., 2011). MAPKs as pivotal players in zygotic and stomatal ACD, besides modulating transcription factors, are they possibly involved in the regulation of cytoskeletal network and vesicle trafficking, the two fundamental processes in cell polarization? In fact, bi-directional signaling between MAPK cascades and the cytoskeleton was found important for cellular activities, such as cell division and polarized growth (Samaj et al., 2004). MAPKs physically associate with the MT cytoskeleton (Reszka et al., 1995) and, interestingly, the pool of MAPKs associated with MTs have higher activities (Morishima-Kawashima and Kosik, 1996). In plants, multiple lines of evidence supported that MAPKs are involved in cytoskeletal rearrangement during cytokinesis, e.g. MAPKs are co-localized with MTs and interact with the MT-associated protein, MAP65(Beck et al., 2011). The best-studied MAPK module in cytoskeletal regulation in plants is the tobacco NACK-PQR pathway during cytokinesis (Nishihama et al., 2002; Takahashi et al., 2004). The working model describes that the kinesin-related proteins interact with MAPKKK to activate the MAPK signaling cascade, which then phosphorylates MAPK65, a family of MT-crosslinking proteins, to control the rate of phragmoplast expansion (Nishihama et al., 2002; Takahashi et al., 2004). Arabidopsis MPK6, a kinase functionally redundant with MPK3 in zygote, root and stomatal development, was also shown to co-localize with MTs and disturbing MPK6 function caused disrupted division plane orientation in roots (Muller et al., 2010), resembling that of the loss-of-function yda (Smekalova et al., 2014). The p38 MAPK of animal cells and MPK1 of budding yeast are the prominent examples of MAPKs controlling the polarization of actin cytoskeleton (Mazzoni et al., 1993; Zarzov et al., 1996), but such a connection in plants has not been well established yet. One example is that the stress-induced MAPK (SIMK) in Madicago was translocated from the nucleus to the growth tip and co-localized with the enriched F-actin network during root hair formation. Altering F-actin dynamics affected SIMK activity and overexpression of active SIMK enhanced root hair growth (Samaj et al., 2002). The involvement of MAPK signaling in the secretory pathway was recently suggested by screening a RNA interference of kinases and phosphatases in human HeLacells (Farhan et al., 2010). The extracellular signal-regulated kinase, MAPK/ERK, may target Sec16, a key regulatory component for vesicle biogenesis (Farhan et al., 2010). In plants, there were signs of MAPKs functioning in vesicle delivery for root hair growth (Samaj et al., 2002), but clearly further advances require the identification of MAPK substrates that are possibly involved in cell morphogenesis. The cell biology of MAPKs has recently caught the attention of the field and is expected to expand in the near future (Samajova et al., 2013). Perspectives and Future Directions Rapid progress toward understanding the cellular signaling mechanisms for polarity formation using several model systems in plants has been achieved in recent years, there is yet much to learn about the mechanisms of how plants initiate, maintain, and alter cell polarity in response to developmental cues and prevailing environments. Important questions remain to be addressed, for instance, besides auxin, whether and how other plant hormones may influence ROP-centered polarity regulation. What environmental cues may inform and integrate into the intrinsic cellular reorganization machineries to influence cell morphogenesis in plant development? The coordination between positive feedback and mutual antagonism appears to be broadly used in organizing cell polarity in plants and other organisms. This is particularly well demonstrated by the small GTP-based signaling network in tip growth and diffuse growth of plant cells. Though, it is still not fully clear about how cytoskeleton dynamics and vesicular trafficking may feedback to the ROP signaling pathways, how ROP signaling are triggered by the intrinsic or extrinsic polarity cues, and how they are linked to the polarity signaling mechanisms. Mean while, we barely know anything about the feedback loops and antagonistic regulations in the polarity protein-centered asymmetry establishment for ACD. Considering the crucial roles of MAPKs in asymmetric cell division and possibly in polarity establishment in zygote, stomata and root apical meristem, it is anticipated that direct downstream substrates of MAPKs may function together or regulate cytoskeleton network and vesicle trafficking processes. It remains to be determined how the mechanisms discovered in the model systems, such as Arabidopsis, Drosophila and C. elegans, are modified in different developmental contexts of other organisms. Future studies would require the combination of genetic, molecular and cytological tools with computational modeling along experimental testing to provide new insights into these unresolved questions. We apologized for the work not being cited in this review due to the size limit. Y.Q. is supported by the Ministry of Science and Technology of China (2011CB944603; 2012CB944801), National Natural Science Foundation of China (31170290; 31470284) and Program for New Century Excellent Talents in Fujian Province University (JA14096). J.D. is supported by grants from the U.S. National Institute of General Medical Sciences (R01GM109080) and Rutgers University. Figure 1 An integrated model for tip growth of pollen tubes. (a) A diagram demonstrates the intracellular organization of a growing pollen tube. (b) Upper panel: a self-organizing ROP1 signaling network controls the oscillation of pollen tube tip growth. Bottom panel: the apical ROP1 activity oscillates ahead of tip growth, followed by the Ca2+ gradient oscillates. Figure 2 The ROP small GTPases-based molecular mechanism for coordinated cell polar growth to form puzzle-shaped pavement cells. (a) Schematic representation of the cytoskeletal architecture in intercalary pavement cells. (b) A simplified model for the auxin-controlled interdigitation through antagonistic ROP2 and ROP6 pathways that direct the formation of lobes and indentations, respectively. Figure 3 Zygote and stomatal lineage system: cell polarity formation and asymmetric cell division (ACD). (a) Schematic depiction of zygote ACD (left) and the hypothesized genetic mechanisms in polarity formation (right). (b–c) Schematic depiction of stomatal ACD in Arabidopsis (b) and maize (c). Thick dashed lines in (b–c) indicate where the polarity complexes, BASL/POLAR and PANs/ROPs, are localized. CV, central vacuole. N, nucleus. SMC, subsidiary mother cell. SC, subsidiary cell. GMC, guard mother cell. GCs, guard cells. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. References Anthony RG Henriques R Helfer A Meszaros T Rios G Testerink C Munnik T Deak M Koncz C Bogre L A protein kinase target of a PDK1 signalling pathway is involved in root hair growth in Arabidopsis The EMBO journal 2004 23 572 581 14749726 Baskin TI On the alignment of cellulose microfibrils by cortical microtubules: a review and a model Protoplasma 2001 215 150 171 11732054 Basu D Le J Zakharova T Mallery EL Szymanski DB A SPIKE1 signaling complex controls actin-dependent cell morphogenesis through the heteromeric WAVE and ARP2/3 complexes Proceedings of the National Academy of Sciences of the United States of America 2008 105 4044 4049 18308939 Bayer M Nawy T Giglione C Galli M Meinnel T Lukowitz W Paternal control of embryonic patterning in Arabidopsis thaliana Science 2009 323 1485 1488 19286558 Beck M Komis G Ziemann A Menzel D Samaj J Mitogen-activated protein kinase 4 is involved in the regulation of mitotic and cytokinetic microtubule transitions in Arabidopsis thaliana The New phytologist 2011 189 1069 1083 21155826 Bergmann DC Lukowitz W Somerville CR Stomatal development and pattern controlled by a MAPKK kinase Science 2004 304 1494 1497 15178800 Bosch M Hepler PK Pectin methylesterases and pectin dynamics in pollen tubes The Plant cell 2005 17 3219 3226 16322606 Breuninger H Rikirsch E Hermann M Ueda M Laux T Differential expression of WOX genes mediates apical-basal axis formation in the Arabidopsis embryo Developmental cell 2008 14 867 876 18539115 Brownlee C Bouget FY Polarity determination in Fucus: from zygote to multicellular embryo Semin Cell Dev Biol 1998 9 179 185 9599413 Cai G Cresti M Organelle motility in the pollen tube: a tale of 20 years Journal of experimental botany 2009 60 495 508 19112169 Cardenas L New findings in the mechanisms regulating polar growth in root hair cells Plant signaling & behavior 2009 4 4 8 19568333 Carol RJ Dolan L The role of reactive oxygen species in cell growth: lessons from root hairs Journal of experimental botany 2006 57 1829 1834 16720604 Cartwright HN Humphries JA Smith LG PAN1: a receptor-like protein that promotes polarization of an asymmetric cell division in maize Science 2009 323 649 651 19179535 Charest PG Firtel RA Feedback signaling controls leading-edge formation during chemotaxis Current opinion in genetics & development 2006 16 339 347 16806895 Chen D Ren Y Deng Y Zhao J Auxin polar transport is essential for the development of zygote and embryo in Nicotiana tabacum L. and correlated with ABP1 and PM H+-ATPase activities Journal of experimental botany 2010 61 1853 1867 20348352 Chen J Yang Z Novel ABP1-TMK auxin sensing system controls ROP GTPase-mediated interdigitated cell expansion in Arabidopsis Small GTPases 2014 5 Chen X Naramoto S Robert S Tejos R Lofke C Lin D Yang Z Friml J ABP1 and ROP6 GTPase signaling regulate clathrin-mediated endocytosis in Arabidopsis roots Current biology : CB 2012 22 1326 1332 22683261 Claret S Jouette J Benoit B Legent K Guichet A PI(4,5)P2 produced by the PI4P5K SKTL controls apical size by tethering PAR-3 in Drosophila epithelial cells Current biology : CB 2014 24 1071 1079 24768049 Cole RA Synek L Zarsky V Fowler JE SEC8, a subunit of the putative Arabidopsis exocyst complex, facilitates pollen germination and competitive pollen tube growth Plant physiology 2005 138 2005 2018 16040664 Costa LM Marshall E Tesfaye M Silverstein KA Mori M Umetsu Y Otterbach SL Papareddy R Dickinson HG Boutiller K Central cell-derived peptides regulate early embryo patterning in flowering plants Science 2014 344 168 172 24723605 Craddock C Lavagi I Yang Z New insights into Rho signaling from plant ROP/Rac GTPases Trends in cell biology 2012 22 492 501 22795444 Cuenca AA Schetter A Aceto D Kemphues K Seydoux G Polarization of the C. elegans zygote proceeds via distinct establishment and maintenance phases Development 2003 130 1255 1265 12588843 Dai M Zhang C Kania U Chen F Xue Q McCray T Li G Qin G Wakeley M Terzaghi W A PP6-type phosphatase holoenzyme directly regulates PIN phosphorylation and auxin efflux in Arabidopsis The Plant cell 2012 24 2497 2514 22715043 De Smet I Beeckman T Asymmetric cell division in land plants and algae: the driving force for differentiation Nature reviews. Molecular cell biology 2011 12 177 188 21346731 Dhonukshe P Huang F Galvan-Ampudia CS Mahonen AP Kleine-Vehn J Xu J Quint A Prasad K Friml J Scheres B Plasma membrane-bound AGC3 kinases phosphorylate PIN auxin carriers at TPRXS(N/S) motifs to direct apical PIN recycling Development 2010 137 3245 3255 20823065 Dong J Bergmann DC Stomatal patterning and development Current topics in developmental biology 2010 91 267 297 20705185 Dong J MacAlister CA Bergmann DC BASL controls asymmetric cell division in Arabidopsis Cell 2009 137 1320 1330 19523675 Duan Q Kita D Li C Cheung AY Wu HM FERONIA receptor-like kinase regulates RHO GTPase signaling of root hair development Proceedings of the National Academy of Sciences of the United States of America 2010 107 17821 17826 20876100 Facette MR Smith LG Division polarity in developing stomata Curr Opin Plant Biol 2012 15 585 592 23044038 Farhan H Wendeler MW Mitrovic S Fava E Silberberg Y Sharan R Zerial M Hauri HP MAPK signaling to the early secretory pathway revealed by kinase/phosphatase functional screening The Journal of cell biology 2010 189 997 1011 20548102 Faure JE Rotman N Fortune P Dumas C Fertilization in Arabidopsis thaliana wild type: developmental stages and time course The Plant journal : for cell and molecular biology 2002 30 481 488 12028577 Foreman J Demidchik V Bothwell JH Mylona P Miedema H Torres MA Linstead P Costa S Brownlee C Jones JD Reactive oxygen species produced by NADPH oxidase regulate plant cell growth Nature 2003 422 442 446 12660786 Fowler JE Quatrano RS Plant cell morphogenesis: plasma membrane interactions with the cytoskeleton and cell wall Annu Rev Cell Dev Biol 1997 13 697 743 9442885 Friml J Vieten A Sauer M Weijers D Schwarz H Hamann T Offringa R Jurgens G Efflux-dependent auxin gradients establish the apical-basal axis of Arabidopsis Nature 2003 426 147 153 14614497 Friml J Yang X Michniewicz M Weijers D Quint A Tietz O Benjamins R Ouwerkerk PB Ljung K Sandberg G A PINOID-dependent binary switch in apical-basal PIN polar targeting directs auxin efflux Science 2004 306 862 865 15514156 Fu Y Gu Y Zheng Z Wasteneys G Yang Z Arabidopsis Interdigitating Cell Growth Requires Two Antagonistic Pathways with Opposing Action on Cell Morphogenesis Cell 2005 120 687 700 15766531 Fu Y Xu T Zhu L Wen M Yang Z A ROP GTPase Signaling Pathway Controls Cortical Microtubule Ordering and Cell Expansion in Arabidopsis Current Biology 2009 19 1827 1832 19818614 Gallois P Future of early embryogenesis studies in Arabidopsis thaliana. Comptes rendus de l'Academie des sciences Serie III, Sciences de la vie 2001 324 569 573 Galway ME Heckman JW Jr Schiefelbein JW Growth and ultrastructure of Arabidopsis root hairs: the rhd3 mutation alters vacuole enlargement and tip growth Planta 1997 201 209 218 9084218 Geldner N Richter S Vieten A Marquardt S Torres-Ruiz RA Mayer U Jurgens G Partial loss-of-function alleles reveal a role for GNOM in auxin transport-related, post-embryonic development of Arabidopsis Development 2004 131 389 400 14681187 Goldstein B Macara IG The PAR proteins: fundamental players in animal cell polarization Developmental cell 2007 13 609 622 17981131 Goldstein JL Anderson RG Brown MS Coated pits, coated vesicles, and receptor-mediated endocytosis Nature 1979 279 679 685 221835 Grebe M Xu J Scheres B Cell axiality and polarity in plants--adding pieces to the puzzle Curr Opin Plant Biol 2001 4 520 526 11641068 Grunewald W Friml J The march of the PINs: developmental plasticity by dynamic polar targeting in plant cells The EMBO journal 2010 29 2700 2714 20717140 Gu Y A Rho family GTPase controls actin dynamics and tip growth via two counteracting downstream pathways in pollen tubes The Journal of cell biology 2005 169 127 138 15824136 Gu Y Members of a Novel Class of Arabidopsis Rho Guanine Nucleotide Exchange Factors Control Rho GTPase-Dependent Polar Growth The Plant cell 2006 18 366 381 16415208 Gu Y Vernoud V Fu Y Yang Z ROP GTPase regulation of pollen tube growth through the dynamics of tip-localized F-actin Journal of experimental botany 2003 54 93 101 12456759 Guan Y Guo J Li H Yang Z Signaling in pollen tube growth: crosstalk, feedback, and missing links Molecular plant 2013 6 1053 1064 23873928 Guan Y Meng X Khanna R LaMontagne E Liu Y Zhang S Phosphorylation of a WRKY transcription factor by MAPKs is required for pollen development and function in Arabidopsis PLoS genetics 2014 10 e1004384 24830428 Haecker A Gross-Hardt R Geiges B Sarkar A Breuninger H Herrmann M Laux T Expression dynamics of WOX genes mark cell fate decisions during early embryonic patterning in Arabidopsis thaliana Development 2004 131 657 668 14711878 Hala M Cole R Synek L Drdova E Pecenkova T Nordheim A Lamkemeyer T Madlung J Hochholdinger F Fowler JE An exocyst complex functions in plant cell growth in Arabidopsis and tobacco The Plant cell 2008 20 1330 1345 18492870 Hao Y Boyd L Seydoux G Stabilization of cell polarity by the C. elegans RING protein PAR-2 Developmental cell 2006 10 199 208 16459299 Hazak O Bloch D Poraty L Sternberg H Zhang J Friml J Yalovsky S A rho scaffold integrates the secretory system with feedback mechanisms in regulation of auxin distribution PLoS biology 2010 8 e1000282 20098722 Hepler PK Vidali L Cheung AY Polarized cell growth in higher plants Annu Rev Cell Dev Biol 2001 17 159 187 11687487 Huang F Zago MK Abas L van Marion A Galvan-Ampudia CS Offringa R Phosphorylation of conserved PIN motifs directs Arabidopsis PIN1 polarity and auxin transport The Plant cell 2010 22 1129 1142 20407025 Huang JB Liu H Chen M Li X Wang M Yang Y Wang C Huang J Liu G Liu Y ROP3 GTPase Contributes to Polar Auxin Transport and Auxin Responses and Is Important for Embryogenesis and Seedling Growth in Arabidopsis The Plant cell 2014 Humphries JA Vejlupkova Z Luo A Meeley RB Sylvester AW Fowler JE Smith LG ROP GTPases act with the receptor-like protein PAN1 to polarize asymmetric cell division in maize The Plant cell 2011 23 2273 2284 21653193 Hwang J-U Vernoud V Szumlanski A Nielsen E Yang Z A Tip-Localized RhoGAP Controls Cell Polarity by Globally Inhibiting Rho GTPase at the Cell Apex Current Biology 2008 18 1907 1916 19108776 Hwang JU Gu Y Lee YJ Yang Z Oscillatory ROP GTPase activation leads the oscillatory polarized growth of pollen tubes Molecular biology of the cell 2005 16 5385 5399 16148045 Hwang JU Wu G Yan A Lee YJ Grierson CS Yang Z Pollen-tube tip growth requires a balance of lateral propagation and global inhibition of Rho-family GTPase activity Journal of cell science 2010 123 340 350 20053639 Ischebeck T Stenzel I Hempel F Jin X Mosblech A Heilmann I Phosphatidylinositol-4,5-bisphosphate influences Nt-Rac5-mediated cell expansion in pollen tubes of Nicotiana tabacum The Plant journal : for cell and molecular biology 2011 65 453 468 21265898 Jeong S Palmer TM Lukowitz W The RWP-RK factor GROUNDED promotes embryonic polarity by facilitating YODA MAP kinase signaling Current biology : CB 2011 21 1268 1276 21802295 Jones MA Raymond MJ Yang Z Smirnoff N NADPH oxidase-dependent reactive oxygen species formation required for root hair growth depends on ROP GTPase Journal of experimental botany 2007 58 1261 1270 17301029 Jones MA Shen JJ Fu Y Li H Yang Z Grierson CS The Arabidopsis Rop2 GTPase is a positive regulator of both root hair initiation and tip growth The Plant cell 2002 14 763 776 11971133 Kania U Fendrych M Friml J Polar delivery in plants; commonalities and differences to animal epithelial cells Open biology 2014 4 140017 24740985 Kaothien P Ok SH Shuai B Wengier D Cotter R Kelley D Kiriakopolos S Muschietti J McCormick S Kinase partner protein interacts with the LePRK1 and LePRK2 receptor kinases and plays a role in polarized pollen tube growth Plant J 2005 42 492 503 15860008 Kaya H Nakajima R Iwano M Kanaoka MM Kimura S Takeda S Kawarazaki T Senzaki E Hamamura Y Higashiyama T Ca2+-activated reactive oxygen species production by Arabidopsis RbohH and RbohJ is essential for proper pollen tube tip growth The Plant cell 2014 26 1069 1080 24610725 Klahre U Becker C Schmitt AC Kost B Nt-RhoGDI2 regulates Rac/Rop signaling and polar cell growth in tobacco pollen tubes The Plant journal : for cell and molecular biology 2006 46 1018 1031 16805734 Kosami K Ohki I Nagano M Furuita K Sugiki T Kawano Y Kawasaki T Fujiwara T Nakagawa A Shimamoto K The Crystal Structure of the Plant Small GTPase OsRac1 Reveals Its Mode of Binding to NADPH Oxidase The Journal of biological chemistry 2014 289 28569 28578 25128531 Krahn MP Klopfenstein DR Fischer N Wodarz A Membrane targeting of Bazooka/PAR-3 is mediated by direct binding to phosphoinositide lipids Current biology : CB 2010 20 636 642 20303268 Kropf DL Induction of Polarity in Fucoid Zygotes The Plant cell 1997 9 1011 1020 12237373 Kropf DL Bisgrove SR Hable WE Cytoskeletal control of polar growth in plant cells Current opinion in cell biology 1998 10 117 122 9484603 Lau OS Davies KA Chang J Adrian J Rowe MH Ballenger CE Bergmann DC Direct roles of SPEECHLESS in the specification of stomatal self-renewing cells Science 2014 345 1605 1609 25190717 Lavy M Bloch D Hazak O Gutman I Poraty L Sorek N Sternberg H Yalovsky S A Novel ROP/RAC effector links cell polarity, root-meristem maintenance, and vesicle trafficking Current biology : CB 2007 17 947 952 17493810 Lee YJ Szumlanski A Nielsen E Yang Z Rho-GTPase-dependent filamentous actin dynamics coordinate vesicle targeting and exocytosis during tip growth The Journal of cell biology 2008 181 1155 1168 18591430 Li H Lin Y Heath RM Zhu MX Yang Z Control of pollen tube tip growth by a Rop GTPase-dependent pathway that leads to tip-localized calcium influx The Plant cell 1999 11 1731 1742 10488239 Li H Wu G Ware D Davis KR Yang Z Arabidopsis Rho-related GTPases: differential gene expression in pollen and polar localization in fission yeast Plant physiology 1998 118 407 417 9765526 Lin D Cao L Zhou Z Zhu L Ehrhardt D Yang Z Fu Y Rho GTPase signaling activates microtubule severing to promote microtubule ordering in Arabidopsis Current biology : CB 2013 23 290 297 23394835 Lin D Nagawa S Chen J Cao L Chen X Xu T Li H Dhonukshe P Yamamuro C Friml J A ROP GTPase-dependent auxin signaling pathway regulates the subcellular distribution of PIN2 in Arabidopsis roots Current biology : CB 2012 22 1319 1325 22683260 Lin D Ren H Fu Y ROP GTPase-mediated auxin signaling regulates pavement cell interdigitation in Arabidopsis thaliana Journal of integrative plant biology 2014 Lin Q Fuji RN Yang W Cerione RA RhoGDI is required for Cdc42-mediated cellular transformation Current biology : CB 2003 13 1469 1479 12956948 Lukowitz W Roeder A Parmenter D Somerville C A MAPKK kinase gene regulates extra-embryonic cell fate in Arabidopsis Cell 2004 116 109 119 14718171 MacAlister CA Ohashi-Ito K Bergmann DC Transcription factor control of asymmetric cell divisions that establish the stomatal lineage Nature 2007 445 537 540 17183265 Mao G Meng X Liu Y Zheng Z Chen Z Zhang S Phosphorylation of a WRKY transcription factor by two pathogen-responsive MAPKs drives phytoalexin biosynthesis in Arabidopsis The Plant cell 2011 23 1639 1653 21498677 Mathur J Local interactions shape plant cells Current opinion in cell biology 2006 18 40 46 16343888 Mathur J Chua NH Microtubule stabilization leads to growth reorientation in Arabidopsis trichomes The Plant cell 2000 12 465 477 10760237 Mazzoni C Zarov P Rambourg A Mann C The SLT2 (MPK1) MAP kinase homolog is involved in polarized cell growth in Saccharomyces cerevisiae The Journal of cell biology 1993 123 1821 1833 8276900 McMahon HT Boucrot E Molecular mechanism and physiological functions of clathrin-mediated endocytosis Nature reviews. Molecular cell biology 2011 12 517 533 21779028 Michniewicz M Zago MK Abas L Weijers D Schweighofer A Meskiene I Heisler MG Ohno C Zhang J Huang F Antagonistic regulation of PIN phosphorylation by PP2A and PINOID directs auxin flux Cell 2007 130 1044 1056 17889649 Mittler R Berkowitz G Hydrogen peroxide, a messenger with too many roles? Redox report : communications in free radical research 2001 6 69 72 11450985 Molendijk AJ Bischoff F Rajendrakumar CS Friml J Braun M Gilroy S Palme K Arabidopsis thaliana Rop GTPases are localized to tips of root hairs and control polar growth The EMBO journal 2001 20 2779 2788 11387211 Monshausen GB Messerli MA Gilroy S Imaging of the Yellow Cameleon 3.6 indicator reveals that elevations in cytosolic Ca2+ follow oscillating increases in growth in root hairs of Arabidopsis Plant physiology 2008 147 1690 1698 18583529 Monteiro D Castanho Coelho P Rodrigues C Camacho L Quader H Malho R Modulation of endocytosis in pollen tube growth by phosphoinositides and phospholipids Protoplasma 2005 226 31 38 16231099 Moravcevic K Mendrola JM Schmitz KR Wang YH Slochower D Janmey PA Lemmon MA Kinase associated-1 domains drive MARK/PAR1 kinases to membrane targets by binding acidic phospholipids Cell 2010 143 966 977 21145462 Morishima-Kawashima M Kosik KS The pool of map kinase associated with microtubules is small but constitutively active Molecular biology of the cell 1996 7 893 905 8816996 Muller J Beck M Mettbach U Komis G Hause G Menzel D Samaj J Arabidopsis MPK6 is involved in cell division plane control during early root development, and localizes to the pre-prophase band, phragmoplast, trans-Golgi network and plasma membrane The Plant journal : for cell and molecular biology 2010 61 234 248 19832943 Murphy AS Peer WA Vesicle trafficking: ROP-RIC roundabout Current biology : CB 2012 22 R576 R578 22835794 Musch A Cohen D Kreitzer G Rodriguez-Boulan E cdc42 regulates the exit of apical and basolateral proteins from the trans-Golgi network The EMBO journal 2001 20 2171 2179 11331583 Myers C Romanowsky SM Barron YD Garg S Azuse CL Curran A Davis RM Hatton J Harmon AC Harper JF Calcium-dependent protein kinases regulate polarized tip growth in pollen tubes The Plant journal : for cell and molecular biology 2009 59 528 539 19392698 Nagawa S Xu T Lin D Dhonukshe P Zhang X Friml J Scheres B Fu Y Yang Z ROP GTPase-dependent actin microfilaments promote PIN1 polarization by localized inhibition of clathrin-dependent endocytosis PLoS biology 2012 10 e1001299 22509133 Nagawa S Xu T Yang Z RHO GTPase in plants: Conservation and invention of regulators and effectors Small GTPases 2010 1 78 88 21686259 Nishihama R Soyano T Ishikawa M Araki S Tanaka H Asada T Irie K Ito M Terada M Banno H Expansion of the cell plate in plant cytokinesis requires a kinesin-like protein/MAPKKK complex Cell 2002 109 87 99 11955449 Oda Y Fukuda H Initiation of cell wall pattern by a Rho- and microtubule-driven symmetry breaking Science 2012 337 1333 1336 22984069 Orlando K Guo W Membrane organization and dynamics in cell polarity Cold Spring Harbor perspectives in biology 2009 1 a001321 20066116 Petricka JJ Van Norman JM Benfey PN Symmetry breaking in plants: molecular mechanisms regulating asymmetric cell divisions in Arabidopsis Cold Spring Harbor perspectives in biology 2009 1 a000497 20066115 Pickett-Heaps JD Northcote DH Organization of microtubules and endoplasmic reticulum during mitosis and cytokinesis in wheat meristems Journal of cell science 1966 1 109 120 5929804 Pierson ES Miller DD Callaham DA van Aken J Hackett G Hepler PK Tip-localized calcium entry fluctuates during pollen tube growth Developmental biology 1996 174 160 173 8626016 Pillitteri LJ Peterson KM Horst RJ Torii KU Molecular profiling of stomatal meristemoids reveals new component of asymmetric cell division and commonalities among stem cell populations in Arabidopsis The Plant cell 2011 23 3260 3275 21963668 Potocky M Jones MA Bezvoda R Smirnoff N Zarsky V Reactive oxygen species produced by NADPH oxidase are involved in pollen tube growth The New phytologist 2007 174 742 751 17504458 Preuss ML Schmitz AJ Thole JM Bonner HK Otegui MS Nielsen E A role for the RabA4b effector protein PI-4Kbeta1 in polarized expansion of root hair cells in Arabidopsis thaliana The Journal of cell biology 2006 172 991 998 16567499 Price MH Roberts DM McCartney BM Jezuit E Peifer M Cytoskeletal dynamics and cell signaling during planar polarity establishment in the Drosophila embryonic denticle Journal of cell science 2006 119 403 415 16418222 Qin Y Chen D Zhao J Localization of arabinogalactan proteins in anther, pollen, and pollen tube of Nicotiana tabacum L Protoplasma 2007 231 43 53 17602278 Qin Y Yang Z Rapid tip growth: Insights from pollen tubes Seminars in Cell & Developmental Biology 2011 22 816 824 21729760 Qin Y Zhao J Localization of arabinogalactan proteins in egg cells, zygotes, and two-celled proembryos and effects of beta-D-glucosyl Yariv reagent on egg cell fertilization and zygote division in Nicotiana tabacum L Journal of experimental botany 2006 57 2061 2074 16720612 Rasmussen CG Wright AJ Muller S The role of the cytoskeleton and associated proteins in determination of the plant cell division plane The Plant journal : for cell and molecular biology 2013 75 258 269 23496276 Rato C Monteiro D Hepler PK Malho R Calmodulin activity and cAMP signalling modulate growth and apical secretion in pollen tubes The Plant journal : for cell and molecular biology 2004 38 887 897 15165182 Reszka AA Seger R Diltz CD Krebs EG Fischer EH Association of mitogen-activated protein kinase with the microtubule cytoskeleton Proceedings of the National Academy of Sciences of the United States of America 1995 92 8881 8885 7568036 Rockel N Wolf S Kost B Rausch T Greiner S Elaborate spatial patterning of cell-wall PME and PMEI at the pollen tube tip involves PMEI endocytosis, and reflects the distribution of esterified and de-esterified pectins The Plant journal : for cell and molecular biology 2008 53 133 143 17971035 Saedler R Mathur N Srinivas BP Kernebeck B Hulskamp M Mathur J Actin control over microtubules suggested by DISTORTED2 encoding the Arabidopsis ARPC2 subunit homolog Plant & cell physiology 2004 45 813 822 15295064 Samaj J Baluska F Hirt H From signal to cell polarity: mitogen-activated protein kinases as sensors and effectors of cytoskeleton dynamicity Journal of experimental botany 2004 55 189 198 14673033 Samaj J Ovecka M Hlavacka A Lecourieux F Meskiene I Lichtscheidl I Lenart P Salaj J Volkmann D Bogre L Involvement of the mitogen-activated protein kinase SIMK in regulation of root hair tip growth The EMBO journal 2002 21 3296 3306 12093731 Samajova O Komis G Samaj J Emerging topics in the cell biology of mitogen-activated protein kinases Trends in plant science 2013 18 140 148 23291243 Settleman J Intercalating Arabidopsis leaf cells: a jigsaw puzzle of lobes, necks, ROPs, and RICs Cell 2005 120 570 572 15766519 Smekalova V Luptovciak I Komis G Samajova O Ovecka M Doskocilova A Takac T Vadovic P Novak O Pechan T Involvement of YODA and mitogen activated protein kinase 6 in Arabidopsis post-embryogenic root development through auxin up-regulation and cell division plane orientation The New phytologist 2014 203 1175 1193 24923680 Smith LG Cytoskeletal control of plant cell shape: getting the fine points Curr Opin Plant Biol 2003 6 63 73 12495753 Strotbek C Krinninger S Frank W The moss Physcomitrella patens: methods and tools from cultivation to targeted analysis of gene function The International journal of developmental biology 2013 57 553 564 24166438 Sukumar P Edwards KS Rahman A Delong A Muday GK PINOID kinase regulates root gravitropism through modulation of PIN2-dependent basipetal auxin transport in Arabidopsis Plant physiology 2009 150 722 735 19363095 Synek L Schlager N Elias M Quentin M Hauser MT Zarsky V AtEXO70A1, a member of a family of putative exocyst subunits specifically expanded in land plants, is important for polar growth and plant development The Plant journal : for cell and molecular biology 2006 48 54 72 16942608 Takahashi Y Soyano T Sasabe M Machida Y A MAP kinase cascade that controls plant cytokinesis Journal of biochemistry 2004 136 127 132 15496582 Thompson BJ Cell polarity: models and mechanisms from yeast, worms and flies Development 2013 140 13 21 23222437 Ueda M Zhang Z Laux T Transcriptional activation of Arabidopsis axis patterning genes WOX8/9 links zygote polarity to embryo development Developmental cell 2011 20 264 270 21316593 Vidali L Bezanilla M Physcomitrella patens: a model for tip cell growth and differentiation Curr Opin Plant Biol 2012 15 625 631 23022392 Waki T Hiki T Watanabe R Hashimoto T Nakajima K The Arabidopsis RWP-RK protein RKD4 triggers gene expression and pattern formation in early embryogenesis Current biology : CB 2011 21 1277 1281 21802301 Wasteneys GO Galway ME Remodeling the cytoskeleton for growth and form: an overview with some new views Annual review of plant biology 2003 54 691 722 Wen TJ Hochholdinger F Sauer M Bruce W Schnable PS The roothairless1 gene of maize encodes a homolog of sec3, which is involved in polar exocytosis Plant physiology 2005 138 1637 1643 15980192 Wu H Feng W Chen J Chan LN Huang S Zhang M PDZ domains of Par-3 as potential phosphoinositide signaling integrators Molecular cell 2007a 28 886 898 18082612 Wu X Chory J Weigel D Combinations of WOX activities regulate tissue proliferation during Arabidopsis embryonic development Developmental biology 2007b 309 306 316 17706632 Xu T Dai N Chen J Nagawa S Cao M Li H Zhou Z Chen X De Rycke R Rakusova H Cell surface ABP1-TMK auxin-sensing complex activates ROP GTPase signaling Science 2014 343 1025 1028 24578577 Xu T Wen M Nagawa S Fu Y Chen J-G Wu M-J Perrot-Rechenmann C Friml J Jones AM Yang Z Cell Surface- and Rho GTPase-Based Auxin Signaling Controls Cellular Interdigitation in Arabidopsis Cell 2010 143 99 110 20887895 Yan A Xu G Yang ZB Calcium participates in feedback regulation of the oscillating ROP1 Rho GTPase in pollen tubes Proceedings of the National Academy of Sciences 2009 106 22002 22007 Yang Z Small GTPases: versatile signaling switches in plants The Plant cell 2002 14 Suppl S375 S388 12045289 Yang Z Cell Polarity Signaling inArabidopsis Annual Review of Cell and Developmental Biology 2008 24 551 575 Yang Z Fu Y ROP/RAC GTPase signaling Curr Opin Plant Biol 2007 10 490 494 17709276 Yang Z Lavagi I Spatial control of plasma membrane domains: ROP GTPase-based symmetry breaking Curr Opin Plant Biol 2012 15 601 607 23177207 Yoon GM Dowd PE Gilroy S McCubbin AG Calcium-dependent protein kinase isoforms in Petunia have distinct functions in pollen tube growth, including regulating polarity The Plant cell 2006 18 867 878 16531501 Zarzov P Mazzoni C Mann C The SLT2(MPK1) MAP kinase is activated during periods of polarized cell growth in yeast The EMBO journal 1996 15 83 91 8598209 Zegzouti H Anthony RG Jahchan N Bogre L Christensen SK Phosphorylation and activation of PINOID by the phospholipid signaling kinase 3-phosphoinositide-dependent protein kinase 1 (PDK1) in Arabidopsis Proceedings of the National Academy of Sciences of the United States of America 2006 103 6404 6409 16601102 Zhang J Nodzynski T Pencik A Rolcik J Friml J PIN phosphorylation is sufficient to mediate PIN polarity and direct auxin transport Proceedings of the National Academy of Sciences of the United States of America 2010 107 918 922 20080776 Zhang J Vanneste S Brewer PB Michniewicz M Grones P Kleine-Vehn J Lofke C Teichmann T Bielach A Cannoot B Inositol trisphosphate-induced Ca2+ signaling modulates auxin transport and PIN polarity Developmental cell 2011 20 855 866 21664582 Zhang X Facette M Humphries JA Shen Z Park Y Sutimantanapi D Sylvester AW Briggs SP Smith LG Identification of PAN2 by quantitative proteomics as a leucine-rich repeat-receptor-like kinase acting upstream of PAN1 to polarize cell division in maize The Plant cell 2012 24 4577 4589 23175742 Zhang Y McCormick S A distinct mechanism regulating a pollen-specific guanine nucleotide exchange factor for the small GTPase Rop in Arabidopsis thaliana Proceedings of the National Academy of Sciences 2007 104 18830 18835 Zhang Z Laux T The asymmetric division of the Arabidopsis zygote: from cell polarity to an embryo axis Sexual plant reproduction 2011 24 161 169 21225434 Zhao Y Yan A Feijo JA Furutani M Takenawa T Hwang I Fu Y Yang Z Phosphoinositides regulate clathrin-dependent endocytosis at the tip of pollen tubes in Arabidopsis and tobacco The Plant cell 2010 22 4031 4044 21189293 Zourelidou M Muller I Willige BC Nill C Jikumaru Y Li H Schwechheimer C The polarly localized D6 PROTEIN KINASE is required for efficient auxin transport in Arabidopsis thaliana Development 2009 136 627 636 19168677
PMC005xxxxxx/PMC5124501.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101465400 34171 Sci Signal Sci Signal Science signaling 1945-0877 1937-9145 27803283 5124501 10.1126/scisignal.aag0240 NIHMS830528 Article Obligatory role for GPER in cardiovascular aging and disease^ Meyer Matthias R. 1‡* Fredette Natalie C. 1¶ Daniel Christoph 2 Sharma Geetanjali 1 Amann Kerstin 2 Arterburn Jeffrey B. 3 Barton Matthias 4* Prossnitz Eric R. 15* 1 University of New Mexico Health Sciences Center, Department of Internal Medicine, Albuquerque, NM 87131, USA. 2 Friedrich-Alexander-University of Erlangen-Nürnberg, Department of Nephropathology, 91054 Erlangen, Germany. 3 New Mexico State University, Department of Chemistry and Biochemistry, Las Cruces, NM 88003, USA. 4 University of Zürich, Molecular Internal Medicine, 8057 Zürich, Switzerland. 5 University of New Mexico Comprehensive Cancer Center, Albuquerque, NM 87131, USA. ‡ Current address: Division of Cardiology, Department of Internal Medicine, Triemli City Hospital, 8063 Zurich, Switzerland . ¶ Current address: Department of Physiology, University of Florida, Gainesville, FL 32611, USA * To whom correspondence should be addressed: [email protected] (M.R.M.), [email protected] (M.B.), [email protected] (E.R.P.). M.B. and E.R.P. are co-senior authors of this work. 18 11 2016 1 11 2016 1 11 2016 01 5 2017 9 452 ra105ra105 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Pharmacological activation of the heptahelical G protein-coupled receptor GPER by selective ligands counteracts multiple aspects of cardiovascular disease. We thus expected that genetic deletion or pharmacological inhibition of GPER would further aggravate such disease states, particularly with age. To the contrary, we found that genetic ablation of Gper in mice prevented cardiovascular pathologies associated with aging by reducing superoxide (.O2−) formation by NADPH oxidase (Nox) and reduced expression the Nox isoform Nox1. Blocking GPER activity pharmacologically with G36, a synthetic, small molecule, GPER-selective blocker (GRB), decreased Nox1 abundance and .O2− production to basal amounts in cells exposed to angiotensin II and in mice chronically infused with angiotensin II. Thus, this study revealed a role for GPER activity in regulating Nox1 abundance and associated .O2−-mediated structural and functional damage that contributes to disease pathology. Our results indicated that GRBs represent a new class of drugs that can indirectly reduce Nox activity and could be used for the treatment of chronic disease processes involving excessive .O2− formation, including arterial hypertension and diastolic heart failure. Introduction G protein-coupled receptors (GPCRs) exert both rapid and chronic effects (1). The G protein-coupled estrogen receptor GPER is a heptahelical receptor, originally designated GPR30, with amino acid homology to GPCRs for angiotensin II (Ang II) and chemokines, that is found in multiple cell types including vascular cells (2-4). GPER is localized predominantly to the endoplasmic reticulum and Golgi apparatus (5) and mediates cellular responses to estrogens, selective estrogen receptor modulators (SERMs), and xenoestrogens (6) through non-genomic as well as genomic mechanisms (5, 7-9). GPER activation results in the rapid mobilization of intracellular calcium (5, 10, 11) and activation of nitric oxide (NO) synthase (12), Akt (11, 13-16) and ERK (7, 11, 16), among other pathways (6). In addition, GPER, like many other GPCRs (1, 17), may also exhibit basal or intrinsic activities that contribute to the chronic regulation of genomic pathways. Such genomic effects have been suggested to be responsible for the increased vasoconstrictor tone observed in Gper-deficient mice (18). The beneficial cardiovascular effects of GPER-selective synthetic ligands in multiple disease models (4, 12, 19) have led to the currently prevailing concept that activation of GPER conveys organ protection (6, 11, 20-23). Reactive oxygen species (ROS) are short-lived intermediates of oxidative metabolism that are essential for cardiovascular homeostasis (24). Excessive ROS production, however, occurs in many chronic disease processes and aggravates vasoconstriction and cell growth, thereby contributing to increased vascular tone, myocardial hypertrophy, fibrosis, heart failure, and aging (25, 26). The NADPH oxidase (Nox) family represents the principal physiological source of ROS in the cardiovascular system and is composed of 7 catalytic subunits termed Nox1-5 and Duox1-2 (27-29). Of these, Nox1, Nox2 and Nox4 have been implicated in both experimental and human hypertension and heart failure, yet their role(s) in cardiovascular aging are less understood (27, 28). Given that studies utilizing the GPER-selective agonist G-1 have demonstrated that GPER activation conveys partial protection from vascular and myocardial disease (4, 13, 30) and given the central role of ROS in these chronic disease processes (24, 25, 27), we expected that pathologies characterized by increased bioavailability of ROS, particularly those associated with aging, would be exacerbated in the absence of functional GPER, resulting in even greater ROS production. We therefore set out to determine the functional and structural effects of genetic deletion (20) as well as pharmacologic inhibition (31, 32) of GPER on ROS-dependent pathologies affecting the cardiovascular system. Results GPER increases Nox-dependent vascular .O2− and vascular tone in aged arteries We first examined the effects of Gper deletion on vascular oxidative stress by measuring the production of the unstable free radical superoxide (.O2−) in the aorta of aged mice. To determine whether Nox enzymes are involved in the generation of .O2−, we used a peptide termed gp91dstat (33), which is derived from a gp91phox (now named Nox2) sequence in the region that interacts with the organizer protein p47phox, thus disrupting p47phox binding to and activation of associated catalytic Nox subunits, particularly Nox2 (34), but also Nox1 in vascular smooth muscle cells (VSMCs) (29, 35-37). We found that in aged wild-type mice, ~50% of .O2− formation was Nox-dependent as it was blocked by gp91ds-tat (Fig. 1A, left panel). In contrast to our expectation of exacerbated .O2− production, .O2− formation in aged Gper−/− mice was instead blunted by ~50-80% compared to wild-type mice (Fig. 1, A and B) and was unaffected by gp91ds-tat treatment (Fig. 1A), suggesting an inactive or absent Nox-mediated .O2−-producing pathway. We next determined the effects of aging on vascular tone, which is characterized by increased .O2− formation that inactivates endothelium-derived vasodilatory NO (38). Impaired endothelium-dependent vasodilation represents an important predictor of mortality in patients with heart failure and hypertension (38-41). As expected, NO-mediated endothelium-dependent relaxation induced by acetylcholine (42) was reduced in aged (Fig. 1C) compared to young wild-type mice (Fig. S1A). This impairment was completely reversed by incubating arteries with gp91ds-tat, restoring vasodilation to an extent similar to that observed in young mice (Fig. 1C and fig. S1A). Further supporting the notion that impaired NO bioactivity was a result of oxidative stress in aged wild-type mice, the smooth muscle sensitivity to NO alone, generated by an exogenous NO donor, was not affected by aging (Fig. S2). In contrast, aged Gper−/− mice were completely protected from the impairment in endothelium-dependent vasodilation observed in aged wild-type mice; in fact, the vasodilatory capacity was preserved and identical to that of young mice (Fig. 1C, fig. S1A and fig. S2). In agreement with these observations, we found that in aged wild-type mice, vascular contractions in response to Ang II (a vasoactive peptide that stimulates Nox (43, 44)) were partially (~50%) blocked by gp91ds-tat (Fig. 1D), whereas gp91ds-tat had no effect on Ang II-mediated contractions of arteries from aged Gper−/− mice. In line with the reduced .O2− formation in Gper−/− mice (Fig. 1, A and B, and fig. S1B), contractions in response to Ang II were attenuated in aged (Fig. 1D) as well as in young Gper−/− mice (fig. S1C). These findings, which contrast with the protective vascular role of Gper expression and/or GPER stimulation reported in previous studies (4, 12, 13, 30, 45), indicate instead that constitutive Gper expression is essential for increased vascular Nox bioactivity as well as Nox-mediated vasoconstriction and impaired endothelial cell function, particularly in the context of vascular aging. GPER deletion prevents structural and functional cardiac aging and myocardial dysfunction To determine whether Gper-dependent regulation of oxidative stress also plays a role in age-dependent structural and functional cardiac abnormalities, we next assessed .O2− production in the aging heart. Compared to wild-type mice, myocardial .O2− amounts were markedly lower in aged Gper−/− mice (Fig. 2A). Given that oxidative stress is centrally involved in the structural changes that occur with cardiac aging (25, 26), we next examined myocardial histopathology. Whereas aging increased the left ventricular (LV) wall-to-lumen ratio by ~60% in wild-type mice, Gper−/− mice were completely protected from age-dependent myocardial hypertrophy (Fig. 2B and fig. S3A). In addition, histological analyses of the myocardium of Gper−/− mice revealed an absence of cardiomyocyte hypertrophy (Fig. 2, C and D). Organ failure resulting from fibrosis accounts for at least one third of deaths worldwide (46), with myocardial fibrosis being a key feature of cardiac aging (25, 26). Aging in wild-type mice was associated with prominent and diffuse interstitial myocardial fibrosis and collagen IV accumulation, which again was generally absent in aged Gper−/− mice (Fig. 2, C, E and F). The cardioprotective effects of Gper deletion on myocardial fibrosis and hypertrophy were already detectable at 12 months of age (although the differences were less prominent due to the reduced disease pathology in the wild-type mice), resulting in a lower LV wall-to-lumen ratio (fig. S3A), reduced cardiomyocyte hypertrophy (fig. S3B) and reduced myocardial fibrosis, as assessed by Sirius Red (fig. S3C) and collagen IV (fig. S3D) staining, although the reduction in the former did not reach significance at this age. Given that Gper deletion prevented the structural cardiac abnormalities observed with aging, we next determined whether this translated into improved myocardial function in vivo. Echocardiography confirmed the marked increase in LV relative wall thickness and mass in wild-type mice compared to Gper−/− mice (Fig. 2G and 2H and table S1). Consistent with the reduced ventricular fibrosis and stiffness, analysis of LV filling and diastolic mitral valve annulus velocities (47) revealed improved diastolic function and lower LV filling pressures in aged Gper−/− mice (Fig. 2I and table S1). Together, the overall absence of myocardial fibrosis and hypertrophy in aged Gper−/− mice translated into increased ventricular elasticity, as indicated by improved LV diastolic filling. These differences were independent of changes in systolic LV function or systemic hemodynamics (Table S1). GPER is essential for .O2− production in murine and human VSMCs through Nox1 Cardiac fibrosis involves an age-dependent localized activation of the renin-angiotensin system (RAS) (41, 46, 48), with its primary vasoactive peptide Ang II also promoting premature senescence through the induction of Nox (49, 50). Moreover, Ang II-induced ROS promote redox-sensitive cell functions such as intracellular calcium mobilization and contraction (51). Having established that Gper deficiency abrogates Nox-generated .O2− production in aged mice, we next examined the underlying mechanisms of the molecular regulation in vascular smooth muscle cells (VSMCs) isolated from wild-type and Gper−/− mice (fig. S4, A and B). Consistent with the activation of Nox in intact arteries of wild-type mice (Fig. 1), Ang II-stimulated .O2− production in wild-type VSMCs was inhibited by gp91ds-tat (Fig. 3A, left panel). In cells lacking Gper, the Ang II-stimulating effect on .O2− generation was completely absent (Fig. 3, A and B), which was confirmed by electron paramagnetic resonance (EPR) spectroscopy using BMPO as a spin trap for .O2− (52-54) (fig. S5). Similarly, Ang II-induced, Nox-dependent mobilization of intracellular calcium (51) was absent in Gper-deficient VSMCs (Fig. 3, C and D). By contrast, intracellular calcium mobilization responses to the purinergic receptor agonist ATP (a Nox-independent stimulus (55)) were comparable in VSMCs from wild-type and Gper−/− mice, thus excluding inherent defects in calcium signaling in VSMCs lacking Gper (Fig. 3D). In addition, absence of Gper did not affect the expression of the genes encoding the Ang II AT1A and AT1B receptors (fig. S4C). We next sought to determine whether the effects of GPER on .O2− production observed in murine VSMCs extended to human VSMCs. Knockdown of GPER with siRNA abolished the ability of primary human VSMCs to generate .O2− in response to Ang II (Fig. 4A). To determine whether the effects of GPER were mediated through rapid non-genomic signaling alone or involved long-term genomic effects, we treated human VSMCs with the GPER-selective antagonist G36, a synthetic, small molecule GPER blocker (GRB) (32). Acute treatment (30 min) with gp91ds-tat, but not G36, abolished Ang II-stimulated .O2− production (Fig. 4B). In contrast, prolonged treatment with G36 (for 72 h) completely abrogated Ang II-induced .O2− formation (Fig. 4B), suggestive of mechanisms regulating gene transcription. Consistent with the lack of acute effects, G36 did not display direct antioxidant activity (fig. S6). Given that Nox inhibition by gp91ds-tat reduced vascular .O2− production in mice as well as in murine and human VSMCs only in the presence of GPER, we next determined whether the vascular abundance of Nox1, Nox2 or Nox4 catalytic subunits, which have been implicated in both experimental and human hypertension (27-29), was affected by intrinsic GPER activity. Although gp91ds-tat is traditionally thought only to disrupt Nox2 activity (33), studies have demonstrated that gp91ds-tat also blocks .O2−-mediated effects mediated by Nox1 in VSMCs, likely through its interaction with p47phox (56, 57). In fact, p47phox in VSMCs facilitates activation of Nox1 (58), the closest homologue of Nox2, but not that of Nox4 (59), and also mediates Ang II-induced, redox-dependent signaling (37, 60). Feed-forward mechanisms have also been observed in which ROS production by one Nox subtype or other sources results in the activation of additional Nox subtype(s), suggesting that inhibition of any intermediate could block downstream events (61, 62). We found that in human VSMCs treated with G36 for 72 h, the protein abundance of Nox1 was reduced by ~70% compared to solvent, whereas that of Nox2 and Nox4 was unaffected (Fig. 4C). Similarly, only the protein abundance of Nox1, but not that of Nox2 or Nox4, was substantially lower in murine VSMCs from Gper−/− mice as compared to wild-type mice (Fig. 4D). The reduced Nox1 protein abundance was commensurate with a similar reduction in the mRNA abundance in VSMCs isolated from Gper−/− mice, with gene expression of Nox2 and Nox4 again being unaffected (Fig. 4E). Gper deficiency also reduced Nox1 mRNA abundance in the aorta and myocardium of aged Gper−/− mice (Fig. 4F), both of which displayed markedly reduced .O2− bioactivity compared to wild-type mice (Fig. 1 and 2). To verify that the decreased Nox1 abundance indeed accounted for the inability of Gper-deficient VSMCs to generate .O2− in response to Ang II, we restored Nox1 abundance in these cells using a Nox1-expressing adenovirus. Reintroduction of Nox1 into Gper-deficient VSMC restored their capacity to generate .O2− in response to Ang II (Fig. 4G), further suggesting an obligatory role for GPER in the increase in Nox1 abundance and associated ROS-dependent cellular functions. Genetic ablation or pharmacologic inhibition of GPER prevents arterial hypertension To explore whether the protective effects of Gper deletion extended to cardiovascular disease conditions other than those associated with aging, we increased Nox1 abundance and activity in vivo by infusing mice with Ang II, a critical mediator of Nox1-depdendent .O2− production, vascular dysfunction and increased vascular tone (44). Animals lacking Gper were resistant to the Ang II-induced increase in blood pressure observed in wild-type mice (Fig. 5A). Furthermore, vascular .O2− generation and the increase in vascular Nox1 abundance in mice in response to Ang II infusion required the presence of Gper (Fig. 5, B to D). As previously reported in wild-type mice (44), .O2− generated in response to Ang II impaired endothelial cell function as evident from the blunted NO-dependent vasodilation in response to acetylcholine. By contrast, the attenuation of the vasodilator response was completely absent in Gper−/− mice infused with Ang II (Fig. 5E). In line with .O2−-mediated impairment of NO bioactivity, the inherent vascular smooth muscle sensitivity to NO, as determined with an exogenous NO donor, was not affected by either Ang II infusion or by GPER deficiency or inhibition (fig. S7). These data further confirm that Gper is required to increase Nox1 abundance and the resulting .O2− production, vascular dysfunction and increases in vascular tone. The Nox pathway has been recognized as a therapeutic target for ROS-dependent pathologies in humans (46, 63, 64). To determine in vivo whether decreasing Nox1 protein by pharmacological GPER inhibition can be achieved, we again utilized the GRB G36 (32). Not only did G36 treatment prevent the Ang II-mediated increase in Nox1 protein abundance (Fig. 5D), it also markedly reduced vascular .O2− production (Fig. 5, B and C) and restored the vasodilatory response (Fig. 5E). These effects of G36 resulted in a substantial inhibition of the Ang II-mediated increase in blood pressure (Fig. 5A). Given that GPER mediated increases in Nox1 protein abundance, these results identify GRBs as a member of a new class of drugs that act as Nox down-regulators. Discussion The results presented in the current study demonstrate that inhibiting GPER activity conveys protection from myocardial and vascular diseases associated with increased Nox1-derived oxidative stress, including cardiovascular aging and arterial hypertension. These data may seem counterintuitive at first when compared to the current body of evidence suggesting the protective role of GPER-selective agonists (GRAs) in the cardiovascular system (4). GRAs, such as G-1 (10), rapidly activate Nox-independent pathways thought to mediate salutary vascular effects, such as Akt and ERK (4). In particular, G-1, unlike the GRB G36, induces eNOS phosphorylation and the subsequent generation of NO, which mediates indirect antioxidant effects through inactivation of .O2− (12). Thus, both GRBs and GRAs improve disease outcome by reducing ROS activity through the same receptor, albeit through entirely distinct mechanisms. These findings place GPER at the center of the balance between the beneficial L-arginine-NO synthase pathway and harmful excessive ROS generation. Such dichotomous effects also exist for other hormones (such as insulin) with the ultimate (patho)physiogical effect dependent upon the stage of the disease process (65). Our results have identified new and important chronic functions of intrinsic GPER activity that determine Nox1 abundance. Indeed, regulation of gene expression through constitutive signaling is prevalent among G protein-coupled receptors (17). The GPER-dependent increase in Nox1 abundance and Nox1-dependent ROS formation was likely a key factor in the pathogenesis of increased vasoconstriction associated with hypertension or aging, as well as age-dependent cardiac remodeling. Indeed, arterial hypertension, left-ventricular hypertrophy and the associated diastolic dysfunction in humans are important predictors of the development of heart failure (26), the prevalence of which increases with age and has reached epidemic proportions (66). The present study has certain limitations. The results were obtained in experimental models of vascular and myocardial aging, heart failure and arterial hypertension. Investigating the association between oxidative excess and cardiovascular injury in humans is complicated because most of these patients are on ROS-inhibiting treatments (such as angiotensin converting enzyme inhibitors and statins), which makes it difficult to unmask mechanisms as described in the present study (67). Moreover, whether human Nox enzymes such as Nox5, which is not encoded in the rodent genome (27, 63), are involved in these disease processes requires further study. Given the results from the in vitro studies with primary vascular human cells, it remains to be shown whether G36 inhibits Nox1 activity in humans, which would also require monitoring the safety of GRB treatment. Because ROS are implicated in the progression of many chronic non-communicable diseases, GRBs may find application in a broader array of indications. In summary, this study has identified an obligatory role for the intrinsic activity of GPER as an activator of Nox1 expression that mediates structural and functional injury in vascular and myocardial diseases. Although the role of ROS in the aging process and chronic diseases is well recognized, simple scavenging of ROS with antioxidants has been largely unsuccessful therapeutically (68), likely due to the localized production and ensuing effects of ROS. Inhibition of Nox activity through the use of small molecules, some with limited selectivity (69), is currently being evaluated in clinical trials. The present study however introduces a new class of drugs, Nox down-regulators, which, as shown for G36, reduce Nox protein abundance, thus directly limiting .O2− production by one of its main sources. Therapeutically reducing the expression of Nox could provide an effective approach to target chronic disease conditions involving excessive Nox-mediated .O2− formation. The therapeutic application of GRBs may include prevention of organ injury due to Nox-dependent chronic diseases such as heart failure and arterial hypertension, while also treating or delaying pathologies associated with aging and rare diseases such as progeria syndromes (40). Materials and Methods Study Design The aim of this study was to explore whether and through which mechanisms GPER promotes cardiovascular oxidative stress induced by aging or Ang II. For this purpose, we used wild-type and Gper-deficient mice as a model of aging, as well as mice chronically infused with Ang II. In addition, the GRB G36 was employed to test whether pharmacological targeting of GPER reduced oxidative stress. Detailed mechanistic studies were carried out in primary VSMCs isolated from wild-type and Gper−/− mice, as well as in human VSMCs. Mice were randomly and equally assigned to different treatment groups. Animal studies were conducted in a controlled and non-blinded manner. Study designs included: (i) two-way factorial designs comparing Gper+/+ and Gper−/− mice, in the presence or absence of inhibitors; (ii) similar two-way factorial designs with repeated measures (concentrations, time); (iii) one-way factorial repeated measure designs within Gper+/+ cohort, controlling for baseline differences, over time and (iv) two-way factorial designs with repeated measures designs between the Gper+/+ and Gper−/− groups, controlling for baseline differences, over time. Transgenic mice and aging model Male Gper−/− mice (provided by Jan S. Rosenbaum, Proctor & Gamble Co.) were generated and backcrossed 10 generations onto the C57BL/6 background (Harlan Laboratories) as described (12). Wild-type C57BL/6 and Gper−/− littermates were housed at the Animal Resource Facility of the University of New Mexico Health Sciences Center under controlled temperature of 22–23 °C on a 12 h light-dark cycle with unrestricted access to standard chow and water. Mice aged 24 months, which show functional and structural changes resembling human cardiovascular aging (25), were used as a model of aging. Animals were euthanized at the age of 4, 12 or 24 months by intraperitoneal injection of sodium pentobarbital (2.2 mg g−1 body weight). All procedures were approved by and carried out in accordance with institutional policies and the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Chronic Ang II infusion model Micro-osmotic pumps (Alzet model 1002, Durect) were implanted subcutaneously in the midscapular region of wild-type and Gper−/− mice under isoflurane (3%) anesthesia. Pumps continuously delivered PBS or Ang II (MP Biomedicals) at a rate of 0.7 mg kg−1 per day for 14 days (43, 44). Three days prior to pump implantation, pellets continuously releasing the GRB G36 (32) (33 μg per day, Innovative Research of America) or placebo were implanted subcutaneously into the right hindlimb of a subset of wild-type mice. Vascular smooth muscle cells Primary aortic VSMCs from wild-type and Gper−/− mice (n = 13 per genotype) were isolated and cultured as described (12). Human aortic VSMCs (Lonza) were cultured according to the provider's recommendations. Experiments were performed with cells derived from passages 2 to 5 for murine and 2 to 8 for human VSMCs. For functional assays, cells at sub-confluence were rendered quiescent by overnight serum starvation. Measurement of superoxide (.O2−) by lucigenin-enhanced chemiluminescence Following euthanization, the aorta was immediately excised, carefully cleaned from perivascular adipose and connective tissue, opened longitudinally, and cut into segments of identical size (3 mm) in cold (4 °C) physiological saline solution (PSS, composition in mmol L−1: 129.8 NaCl, 5.4 KCl, 0.83 MgSO4, 0.43 NaH2PO4, 19 NaHCO3, 1.8 CaCl2, and 5.5 glucose; pH 7.4). Tissues were transferred and equilibrated in HEPES-buffered PSS (composition in mmol L−1: 134 NaCl, 6 KCl, 1 MgCl, 10 HEPES, 2 CaCl2, 0.026 EDTA, and 10 glucose; pH 7.4) in a humidified incubator at 37 °C for 60 min. In addition to intact isolated arteries, vascular smooth muscle cells (VSMCs) and a cell free .O2− generating system by adding the substrate xanthine (100 μmol L−1, Calbiochem) to xanthine oxidase (0.05 mU, Calbiochem) were employed. Chemiluminescence was measured in dark-adapted HEPES-PSS containing 5 μmol L−1 lucigenin (Enzo Life Sciences) at 37 °C (70). After equilibrating for 15 min, .O2− production was induced by Ang II (100 nmol L−1) (70). Where indicated, tissues or cells were pretreated with the Nox-selective inhibitor gp91ds-tat (Anaspec, 3 μmol L−1) (28, 33), the GRB G36 (32) (10 nmol L−1, 100 nmol L−1, or 1 μmol L−1), the .O2− dismutase mimetic tempol (100 μmol L−1, Tocris Bioscience) (71), or vehicle (DMSO 0.01%). Luminescence was measured 10-times in 20 sec intervals using a Synergy H1 multi-mode microplate reader (BioTek), and readings were averaged to reduce variability (70). A background reading was subtracted, and .O2− production normalized to surface area of vascular segments (72) or to VSMC number (70), respectively. In-situ detection of .O2− by dihydroethidium (DHE) The thoracic aorta was equilibrated in HEPES-PSS in a humidified incubator at 37 °C for 60 min, and treated with the Nox-selective inhibitor gp91ds-tat (3 μmol L−1) (28, 33) for 30 min when indicated. Tissues were frozen in optimum cutting temperature (O.C.T.) compound (Sakura Finetek), cut on a cryostat into 10 μm thick sections, and stored on glass slides at –80 °C. For staining, sections were incubated with DHE (5 μmol L−1, Invitrogen) in HEPES-PSS for 15 min at room temperature in the dark (73). In separate experiments, VSMCs were grown on poly-L lysine coated coverslips, which were incubated with DHE (5 μmol L−1) in HEPES-PSS for 30 min at 37 °C in the dark (73). Where indicated, VSMCs were pretreated with Nox-selective inhibitor gp91ds-tat (3 μmol L−1) (28, 33) for 30 min, and .O2− production was stimulated by Ang II (100 nmol L−1) for 20 min prior to imaging. Slides with VSMCs or aortic sections were carefully washed, mounted in HEPES-PSS with coverslips, and immediately imaged by epifluorescence microscopy (Axiovert 200M, Zeiss) using a rhodamine filter with exposure intensity adjusted to background fluorescence (73). Signal intensity was quantified using ImageJTM software (National Institutes of Health). .O2− detection by spin trapping combined with electron paramagnetic resonance (EPR) spectroscopy The nitrone 5-tert-butoxycarbonyl 5-methyl-1-pyrroline N-oxide (BMPO, Enzo Life Sciences) was used as the spin trap for .O2− generated from VSMC, which was monitored using EPR spectroscopy as described (52-54). Briefly, serum-starved VSMCs were suspended in serum-free medium supplemented with BMPO (50 mmol L−1) and diethylenetriaminepentaacetic acid (100 μmol L−1), and .O2− production was stimulated by Ang II (100 nmol L−1, 30 min at 37°C). Supernatant containing spin-trapped .O2− was snap-frozen in liquid nitrogen and stored at −80 °C for less than one week. After thawing, supernatant was immediately transferred to custom-made gas-permeable Teflon tubing (Zeus Industries), folded four times, and inserted into a quartz EPR tube open at each end. The quartz EPR tube was inserted into the cavity of an EPR spectrometer (EleXsys 540 X-band, Bruker) operating at 9.8 GHz and 100-kHz field modulation, and the spectra of BMPO-OOH, spin-trapped .O2−, was recorded after spectrometer tuning at room temperature. The EPR spectrum was acquired with a scan time of 40 s, and 20 scans were obtained and averaged to produce significant signal-to-noise ratio. Instrument settings were as follows: magnetic field, 3,509 G; scan range, 70 G; microwave power, 21 mW; modulation frequency, 100 kHz; modulation amplitude, 1.0 G; time constant, 20 ms. The EPR spectra were collected, stored, and manipulated using Xepr software (Bruker). Vascular reactivity studies The aorta was immediately excised after euthanization, carefully cleaned from perivascular adipose and connective tissue and cut into 2 mm long rings in cold (4 °C) PSS. Aortic rings were mounted in myograph chambers (multi-channel myograph system 620M, Danish Myo Technology) onto 200 μm pins (18, 74). A PowerLab 8/35 data acquisition system and LabChart Pro software (AD Instruments) were used for recording of isometric tension. Experiments to determine vascular reactivity of aortic rings were performed as described (18, 74). Briefly, rings were equilibrated in PSS (37 °C; pH 7.4; oxygenated with 21% O2, 5% CO2, and balanced N2) for 30 min and stretched step-wise to the optimal amount of passive tension for force generation. Functional integrity of vascular smooth muscle was confirmed by repeated exposure to KCl (PSS with substitution of 60 mmol/L potassium for sodium), with resulting contractions demonstrating no differences between groups. Selected arteries were pretreated with the Nox-selective inhibitor gp91ds-tat (3 μmol L−1) (28, 33) for 30 min. Contractions to Ang II (100 nmol L−1) were studied in the abdominal aorta in the presence of the NO synthase inhibitor L-NG-nitroarginine methyl ester (L-NAME, 300 μmol L−1, incubation for 30 min, Cayman Chemical) (75) to exclude Ang II-mediated release of NO (76). Ang II-induced contractions exhibit rapid desensitization in the mouse vasculature with a nearly complete loss of tension after about 2 min, thus preventing the recording of responses to increasing concentrations (75, 77). To study endothelium-dependent, NO-mediated relaxations, rings from the thoracic aorta were precontracted with phenylephrine (Sigma-Aldrich) to 80% of KCl-induced contractions, and responses to acetylcholine (0.1 nmol L−1 – 10 μmol L−1, Sigma-Aldrich) were recorded. Similarly, endothelium-independent, NO-mediated relaxations to sodium nitroprusside (SNP, 1 – 10 μmol L−1, MP Biomedicals) were determined. Precontraction did not differ between groups. To exclude any GPER-dependent effects on vasoconstrictor prostanoids (18), responses were obtained in the presence of the cyclooxygenase-inhibitor meclofenamate (1 μmol L−1, incubation for 30 min, Cayman Chemical). Contractions were calculated as the percentage of contraction to KCl, and relaxation was expressed as the percentage of phenylephrine-induced precontraction. Cardiac histopathology After sacrifice, hearts were excised, fixed in 4% paraformaldehyde and embedded into paraffin blocks. Histological sections (2 μm) were stained with hematoxylin-eosin, Sirius Red or by immunohistochemistry using a polyclonal goat antibody recognizing collagen IV (Southern Biotech) as described (78). Morphometric analysis of free left ventricle wall thickness and ventricular lumen area was performed using light microscopy at 400-fold magnification and cellSensTM software (Olympus), with left ventricular wall thickness based on analysis of 10 randomly selected measure points. Cardiomyocyte cross-sectional area was determined by analysis of 15 anterolaterally located cardiomyocytes using cellSens software. Myocardial fibrosis on Sirius Red or collagen type IV stained paraffin sections was graded using a semi-quantitative fibrosis score (0 = no staining; 1 = less than 25%; 2 = 26–50 %; 3 = 51–75%; 4 = more than 75% of cardiac tissue with positive staining). For each heart, the mean score evaluated on 10 power fields at 200-fold magnification was calculated. High-resolution, high-frequency echocardiography Mice were lightly sedated using inhaled isoflurane anesthesia, placed on a heat-pad to maintain body temperature, and echocardiography was performed using a Vevo® LAZR photoacoustic imaging system (VisualSonics) using high-resolution, high-frequency ultrasound at 40 mHz (47, 79). Conventional B-mode, M-mode, pulsed wave- and tissue-doppler images were acquired by an experienced, blinded operator to ensure a standardized, consistent technique, and LV dimensions were quantified as described (47, 79). LV ejection fraction was determined by speckle-tracking based wall motion analysis (47, 79, 80) using VevoStrain software (VisualSonics). Analysis of diastolic function included transmitral flow velocity waveforms obtained from pulsed-wave Doppler to calculate the ratio of early (E)-to-late (atrial, A) LV filling velocities (E/A ratio), and the mitral annulus diastolic velocity (e’ waves) obtained from pulsed-wave tissue Doppler imaging as well as the calculated E/e’ ratio as measures of diastolic function and LV filling pressures. Measurement of arterial blood pressure Systolic blood pressure was measured in conscious mice using a volume-pressure recording noninvasive monitoring system (CODA-6, Kent Scientific) as described (12), which produces blood pressure readings with similar sensitivity and specificity as invasive measurements (12). This blood pressure measurement technique has successfully been applied in the chronic Ang II infusion model (43, 44). Intracellular calcium mobilization VSMCs were loaded with 5 μmol L−1 indo1-AM (Invitrogen) and 0.05% pluronic F-127 (Invitrogen) in Hanks’ buffered salt solution (HBSS) supplemented with NaCl (150 mmol L−1), CaCl2 (2 mmol L−1), and HEPES (20 mmol L−1; pH 7.4) for 30 min at room temperature in the dark. Cells were washed and resuspended in HBSS (106 cells per mL), and calcium mobilization in response to Ang II (100 nmol L−1) and adenosine triphosphate (ATP, 1 μmol L−1, Sigma-Aldrich) was determined ratiometrically using λex 340 nm and λem of 405 and 490 nm at 37 °C in a QM-2000-2 spectrofluorometer (Photon Technology International). VSMC transduction with Nox1 adenovirus or transfection with GPER-targeted siRNA VSMCs from Gper−/− mice were plated at ~600,000 cells per T25 flask, washed in PBS, and infected with Nox1GFP and GFP control adenovirus constructs (59) at 400 MOI overnight in low serum (1% FBS) DMEM. Cells were allowed to recover for 48 h prior to experiments. Transduction efficiency was determined by GFP expression. For siRNA, human aortic VSMCs were transfected with siGPER (Dharmacon ON-Targetplus J-005563-08) or control (siCTL) siRNA (Dharmacon ON-Targetplus D-001810-02) using Lipofectamine 2000 (Invitrogen) for 6–8 h in serum-free medium, washed, and returned to normal medium as described by the manufacturer. Subsequent experiments were performed 72 h after transfection. Quantification of gene and protein expression RNA was extracted, reverse-transcribed and analyzed using SYBR Green-based detection of amplified gene-specific cDNA fragments by qPCR performed in triplicate as described (18). The following primer pairs have been used: 5′-CAT CCA GTC TCC AAA CAT GAC A-3′ (forward) and 5′-GCT ACA GTG GCA ATC ACT CCA G-3′ (reverse) for amplification of a specific cDNA fragment encoding mouse Nox1 (GenBank ID: NM_172203.1); 5′-ACT CCT TGG GTC AGC ACT GG-3′ (forward) and 5′-GTT CCT GTC CAG TTG TCT TCG-3′ (reverse) for amplification of a specific cDNA fragment encoding mouse Nox2 (GenBank ID: NM_007807.4); 5′-TGA ACT ACA GTG AAG ATT TCC TTG AAC-3′ (forward) and 5′-GAC ACC CGT CAG ACC AGG AA-3′ (reverse) for amplification of a specific cDNA fragment encoding mouse Nox4 (GenBank ID: NM_015760.4); 5′-GCG GTC TCC TTT TGA TTT CC-3′ (forward) and 5′-CAA AGG GCT CCT GAA ACT TG-3′ (reverse) for amplification of a specific cDNA fragment encoding mouse AT1A receptor (GenBank ID: NM_177322.3); 5′- TAT TTT CCC CAG AGC AAA GC-3′ (forward) and 5′-TGT TGC TTC CTT GTC CCT TG-3′ (reverse) for amplification of a specific cDNA fragment encoding mouse AT1B receptor (GenBank ID: NM_175086.3); and 5’-TTC ACC ACC ATG GAG AAG GC-3’ (forward) and 5’-GGC ATG GAC TGT GGT CAT GA-3’ (reverse) for amplification of a specific cDNA fragment encoding mouse GAPDH (GenBank ID: NM_008084.2), which served as the housekeeping control. For determination of protein abundance by Western blot, VSMCs were lysed in NP-40 buffer supplemented with protease inhibitor (1 μg/mL), 10% SDS, 0.5% sodium fluoride, and 0.5% sodium orthovanadate. 20 or 40 μg of lysate were loaded on 10% SDS-PAGE gel (Thermo Scientific), blotted onto polyvinylidene fluoride membrane (Millipore), and blocked with 3% newborn calf serum in Tris-buffered saline with Tween-20 (0.1%). Blots were incubated overnight at 4 °C with primary antibodies recognizing Nox1 (Sigma-Aldrich), Nox2 (Boster), or Nox4 (Boster), washed, incubated with secondary HRP-conjugated antibodies (1:5000) for 1 h at room temperature, and developed with Super Signal West Pico Chemiluminescent substrate (Thermo Scientific). Blots performed in duplicate were imaged and quantified using ImageJ densitometry analysis software. Immunofluorescence of Nox1 and GPER Aortic sections frozen in O.C.T. compound were fixed in 4% paraformaldehyde, blocked and permeabilized in PBS containing normal goat serum (3%) and TritonX-100 (0.01%, EM Science). Sections were incubated with rabbit antibody recognizing murine Nox1 (1:100, Sigma-Aldrich) or negative control IgG (1:100, Sigma-Aldrich) overnight at 4 °C, washed, incubated with goat antibody recognizing rabbit IgG conjugated to Alexa Fluor 488 (Invitrogen) for one hour, washed, mounted in Vectashield (Vector Laboratories), and imaged utilizing a Leica SP5 confocal microscope. Signal intensity was quantified using ImageJ software. VSMCs were stained with a rabbit antiserum recognizing murine GPER as described (12). Statistical analysis Statistical analysis for in vitro and in vivo experiments was performed using GraphPad Prism™ version 5.0 for Macintosh (GraphPad Software). When comparing two groups, the two-tailed, unpaired Student's t-test was performed. When comparing multiple groups, data were analyzed by two-way analysis of variance (ANOVA), with repeated measures as appropriate, followed by Bonferroni's post-hoc test to correct for multiple comparisons. Values are expressed as mean±sem; n equals the number of independent animals or cell preparations used. Statistical significance was accepted at a P value < 0.05. Supplementary Material Supplemental Information Acknowledgments We thank C. Hu, D. Cimino, M. Reutelshöfer, K. Schmitt and S. Söllner for expert technical assistance, J. Weaver for assistance with EPR spectroscopy, and B. Deeley (FUJIFILM VisualSonics Inc., Toronto, ON, Canada) for support with the echocardiography studies. We gratefully acknowledge K. K. Griendling and B. Lassègue (Emory University School of Medicine, Atlanta, GA, USA) for providing the Nox1/GFP adenovirus. We thank J. S. Rosenbaum (Proctor and Gamble Co.) for providing the male Gper−/− mice. Funding: This study was supported by the National Institutes of Health (NIH R01 CA127731 & CA163890 to E.R.P.), Dedicated Health Research Funds from the University of New Mexico School of Medicine allocated to the Signature Program in Cardiovascular and Metabolic Diseases (to E.R.P.), the Swiss National Science Foundation (grants 135874 & 141501 to M.R.M. and grants 108258 & 122504 to M.B.), and the Interdisciplinary Centre for Clinical Research (IZKF) Erlangen, project F1 (to K.A.). E.R.P. was also supported by the UNM Comprehensive Cancer Center (NIH grant P30 CA118100). N.C.F. was supported by NIH training grant HL07736. The EPR core facility of the University of New Mexico Biomedical Research and Integrative Neuroimaging Center is supported by NIH grant P30 GM103400, and the University of New Mexico & UNM Comprehensive Cancer Center Fluorescence Microscopy Shared Resource is supported by NIH grant P30 CA118100 as detailed: http://hsc.unm.edu/crtc/microscopy/acknowledgement.shtml. Biostatistics support was provided by the UNM Clinical and Translational Science Center supported by NIH grant UL1 TR001449. Fig. 1 Genetic deletion of Gper abrogates Nox activity and prevents enhanced vasoconstriction in vascular aging. Intact arteries of aged (24 month-old) wild-type (Gper+/+) and Gper−/− mice were analyzed. (A), (B) Nox activity was determined by measuring vascular superoxide (.O2−) production using chemiluminescence (A) or DHE fluorescence (B, scale bar, 200 μm). To quantify the amount of .O2− generated by Nox, subsets of arteries were treated with the Nox inhibitor gp91ds-tat (tat). (C), (D) Endothelium-dependent, NO-mediated vasodilation in response to acetylcholine (C) and contractions to angiotensin II (Ang II, D) in intact arteries in the presence or absence of gp91ds-tat (tat). Data are mean±sem; n = 3–4 mice per group in (A), n = 5–10 mice per group in (B), n = 4–5 mice per group in (C), (D). P < 0.05, P < 0.01 compared to control (CTL); †P < 0.05, ††P < 0.01, †††P < 0.001 compared to wild-type mice (ANOVA with Bonferroni post-hoc tests in (A), (D); repeated measures ANOVA with Bonferroni post-hoc tests in (C); Student's t-test in (B)). Fig. 2 Obligatory role for GPER in aging cardiomyopathy and diastolic dysfunction. Hearts of aged (24 month-old) wild-type and Gper−/− mice were analyzed. (A) Myocardial superoxide (.O2−) production determined by DHE fluorescence. Scale bar, 500 μm. (B) Myocardial hypertrophy as measured by left ventricular wall-to-lumen ratio in cardiac cross sections. Scale bar, 2 mm. (C) Representative histologic myocardial sections stained with hematoxylin-eosin (left, scale bar, 100 μm) and Sirius Red (right, scale bar, 200 μm). Cardiomyocyte hypertrophy analyzed by cross-sectional area (D), quantitation of interstitial fibrosis (E) and representative histologic sections stained for type IV collagen (F, scale bar, 100 μm). (G-I) Dimensions and function of the left ventricle (LV) determined by echocardiography. Representative parasternal M-mode images (G), calculated LV mass (H) and measures of LV filling pressures and diastolic dysfunction (E/e’ ratio, I) are shown. Data are mean±sem; n = 4–5 mice per group in (A), n = 6– 7 mice per group in (B), n = 6–8 mice per group in (C-F), n = 3 mice per group in (G-I). P < 0.05, P < 0.01, P < 0.001 compared to wild-type mice (Student's t-test). Fig. 3 GPER is required for Ang II-induced superoxide (.O2−) production and mobilization of intracellular calcium in VSMCs. Nox activity stimulated by Ang II was inhibited using gp91dstat (tat). (A), (B) Ang II-induced .O2− production detected by DHE fluorescence (A) and chemiluminescence (B). (C), (D) Mobilization of intracellular calcium ([Ca2+]i) determined utilizing the Ca2+ sensor dye indo1-AM in response to Ang II (representative tracing in C, cumulative data in D) and ATP (D). F405/F490, ratiometric detection of emitted light at 405 nm and 490 nm. P < 0.01, *P < 0.001 compared to control (CTL); †P < 0.05, ††P < 0.01, †††P < 0.001 compared to VSMCs isolated from wild-type (Gper+/+) mice. Data are mean±sem; n = 5–7 independent experiments per group in (A), n = 5 independent experiment per group in (B), 3–5 independent experiments per group in (D), all with VSMCs from 2 independent isolations. ANOVA with Bonferroni post-hoc tests in (A), (D); Student's t-test in (B). Fig. 4 GPER increases Nox1 abundance and activity. (A) Ang II-induced superoxide (.O2−) production detected by chemiluminescence in human VSMCs with GPER-targeted gene silencing (siGPER). P < 0.05 compared to control siRNA (siCTL). siGPER treatment decreased GPER protein abundance by 89+/-5% compared to siCTL treatment (representative data shown in inset). (B), (C) Ang II-induced .O2− production (detected by chemiluminescence, B) and protein abundance of Nox1, Nox2 and Nox4 (C) in human VSMCs treated with the GPER-selective antagonist G36 (1 μmol L−1) for 30 min (acute, B) or 72 h (chronic, B, C). Nox activity was inhibited using gp91ds-tat (tat). P < 0.05, P < 0.01 compared to control (CTL, DMSO 0.01%); †P < 0.05 compared to acute treatment. (D), (E) Protein (D) and mRNA (E) abundance of Nox1, Nox2 and Nox4 in VSMCs isolated from wild-type (Gper+/+) and Gper−/− mice. P < 0.01 compared to wild-type mice. (F) Nox1 mRNA abundance in aorta and myocardium of aged (24 month-old) Gper+/+ and Gper−/− mice. P < 0.05 compared to Gper+/+. (G) Ang II-induced .O2 − production detected by chemiluminescence in VSMCs isolated from Gper−/− mice and infected with Nox1-expressing adenovirus (AdNox1). P < 0.05 compared to vector control (AdGFP). Data are mean±sem; n = 3 independent transfections per group in (A), n = 4–8 independent experiments per group in (B), n = 3–4 independent experiments and Western blots per group in (C) and (D), n = 3-4 VSMC preparations in (E), n = 4–10 mice per group in (F), n = 3 independent infections per group in (G). ANOVA with Bonferroni post-hoc tests in (B); Student's t-test in (A), (C-G). Fig. 5 Genetic or pharmacologic ablation of GPER prevents Ang II-induced hypertension, oxidative stress, and increases in Nox1 abundance. Wild-type (Gper+/+) and Gper−/− mice were infused with Ang II (0.7 mg kg−1 per day) or vehicle (control, CTL) for 14 days. A subset of wild-type mice was also treated with the GRB G36. (A) Systolic arterial blood pressure in conscious animals. (B), (C) Vascular .O2− generation as measured by chemiluminescence (B) and DHE staining (C, scale bars, 300 μm). (D) Vascular Nox1 protein abundance detected by immunofluorescence. (E) Endothelium-dependent, NO-mediated vasodilation in response to acetylcholine. Data are mean±sem; n = 4–5 mice in (A), n = 3–5 mice in (B), n = 3–4 mice in (C), n = 3 mice in (D), n = 5–9 mice in (E). P < 0.05, P < 0.01, *P < 0.001 compared to genotype-matched CTL; †P < 0.05, ††P < 0.01, †††P < 0.001 compared to Ang II-treated wild-type mice (repeated measures ANOVA with Bonferroni post-hoc tests in (A), (E); ANOVA with Bonferroni post-hoc tests in (B), (D)). PE, phenylephrine. ^ This manuscript has been accepted for publication in Science Signaling. This version has not undergone final editing. Please refer to the complete version of record at http://www.sciencesignaling.org/. The manuscript may not be reproduced or used in any manner that does not fall within the fair use provisions of the Copyright Act without the prior, written permission of AAAS. Author contributions: M.R.M., N.C.F., C.D. and G.S. performed experiments; J.B.A. synthesized G36; M.R.M., N.C.F., C.D., G.S., M.B., and E.R.P. analyzed data; M.R.M., N.C.F., K.A., M.B., and E.R.P. interpreted results of experiments; M.R.M., M.B. and E.R.P. prepared figures and wrote the manuscript; all authors approved the final version of manuscript; M.R.M., M.B., and E.R.P. were involved in conception and/or design of research. Competing interests: M.R.M., G.S., M.B., and E.R.P. are inventors on a U.S. patent application for the therapeutic use of compounds targeting GPER. E.R.P. and J.B.A. are inventors on U.S. patent Nos. 7,875,721 and 8,487,100 for GPER-selective ligands and imaging agents. The other authors declare that they have no competing interests. References and Notes 1 Bond RA Ijzerman AP Recent developments in constitutive receptor activity and inverse agonism, and their potential for GPCR drug discovery. Trends Pharmacol Sci 2006 27 92 96 16406086 2 Owman C Blay P Nilsson C Lolait SJ Cloning of human cDNA encoding a novel heptahelix receptor expressed in Burkitt's lymphoma and widely distributed in brain and peripheral tissues. Biochem Biophys Res Commun 1996 228 285 292 8920907 3 Takada Y Kato C Kondo S Korenaga R Ando J Cloning of cDNAs encoding G protein-coupled receptor expressed in human endothelial cells exposed to fluid shear stress. Biochem Biophys Res Commun 1997 240 737 741 9398636 4 Prossnitz ER Barton M The G-protein-coupled estrogen receptor GPER in health and disease. Nat Rev Endocrinol 2011 7 715 726 21844907 5 Revankar CM Cimino DF Sklar LA Arterburn JB Prossnitz ER A transmembrane intracellular estrogen receptor mediates rapid cell signaling. Science 2005 307 1625 1630 15705806 6 Prossnitz ER Arterburn JB International Union of Basic and Clinical Pharmacology. XCVII. G Protein-Coupled Estrogen Receptor and Its Pharmacologic Modulators. Pharmacol Rev 2015 67 505 540 7 Filardo EJ Quinn JA Bland KI Frackelton AR Jr. Estrogen-induced activation of Erk-1 and Erk-2 requires the G protein-coupled receptor homolog, GPR30, and occurs via trans-activation of the epidermal growth factor receptor through release of HB-EGF. Mol Endocrinol 2000 14 1649 1660 11043579 8 Thomas P Pang Y Filardo EJ Dong J Identity of an estrogen membrane receptor coupled to a G protein in human breast cancer cells. Endocrinology 2005 146 624 632 15539556 9 Prossnitz ER Maggiolini M Mechanisms of estrogen signaling and gene expression via GPR30. Mol Cell Endocrinol 2009 308 32 38 19464786 10 Haas E Bhattacharya I Brailoiu E Damjanovic M Brailoiu GC Gao X Mueller-Guerre L Marjon NA Gut A Minotti R Meyer MR Amann K Ammann E Perez-Dominguez A Genoni M Clegg DJ Dun NJ Resta TC Prossnitz ER Barton M Regulatory role of G protein-coupled estrogen receptor for vascular function and obesity. Circ Res 2009 104 288 291 19179659 11 Sharma G Prossnitz ER Mechanisms of estradiol-induced insulin secretion by the G protein-coupled estrogen receptor GPR30/GPER in pancreatic beta-cells. Endocrinology 2011 152 3030 3039 21673097 12 Meyer MR Fredette NC Howard TA Hu C Ramesh C Daniel C Amann K Arterburn JB Barton M Prossnitz ER G protein-coupled estrogen receptor protects from atherosclerosis. Sci Reports 2014 4 7564 13 Deschamps AM Murphy E Activation of a novel estrogen receptor, GPER, is cardioprotective in male and female rats. Am J Physiol Heart Circ Physiol 2009 297 H1806 1813 19717735 14 Murphy E Estrogen signaling and cardiovascular disease. Circ Res 2011 109 687 696 21885836 15 Zekas E Prossnitz ER Estrogen-mediated inactivation of FOXO3a by the G protein- coupled estrogen receptor GPER. BMC Cancer 2015 15 702 26470790 16 Petrie WK Dennis MK Hu C Dai D Arterburn JB Smith HO Hathaway HJ Prossnitz ER G protein-coupled estrogen receptor-selective ligands modulate endometrial tumor growth. Obstet Gynecol Int 2013 2013 472720 24379833 17 Martin AL Steurer MA Aronstam RS Constitutive Activity among Orphan Class- A G Protein Coupled Receptors. PLoS One 2015 10 e0138463 26384023 18 Meyer MR Amann K Field AS Hu C Hathaway HJ Kanagy NL Walker MK Barton M Prossnitz ER Deletion of G protein-coupled estrogen receptor increases endothelial vasoconstriction. Hypertension 2012 59 507 512 22203741 19 Barton M Prossnitz ER Emerging roles of GPER in diabetes and atherosclerosis. Trends Endocrinol Metab 2015 26 185 192 25767029 20 Prossnitz ER Hathaway HJ What have we learned about GPER function in physiology and disease from knockout mice? J Steroid Biochem Mol Biol 2015 153 114 126 26189910 21 Sharma G Prossnitz ER GPER/GPR30 Knockout Mice: Effects of GPER on Metabolism. Methods Mol Biol 2016 1366 489 502 26585159 22 Blasko E Haskell CA Leung S Gualtieri G Halks-Miller M Mahmoudi M Dennis MK Prossnitz ER Karpus WJ Horuk R Beneficial role of the GPR30 agonist G-1 in an animal model of multiple sclerosis. J Neuroimmunol 2009 214 67 77 19664827 23 Sharma G Hu C Brigman JL Zhu G Hathaway HJ Prossnitz ER GPER deficiency in male mice results in insulin resistance, dyslipidemia, and a proinflammatory state. Endocrinology 2013 154 4136 4145 23970785 24 Cross CE Halliwell B Borish ET Pryor WA Ames BN Saul RL McCord JM Harman D Oxygen radicals and human disease. Ann Intern Med 1987 107 526 545 3307585 25 Dai DF Chen T Johnson SC Szeto H Rabinovitch PS Cardiac aging: from molecular mechanisms to significance in human health and disease. Antioxid Redox Signal 2012 16 1492 1526 22229339 26 Lakatta EG Arterial and cardiac aging: major shareholders in cardiovascular disease enterprises: Part III: cellular and molecular clues to heart and arterial aging. Circulation 2003 107 490 497 12551876 27 Brandes RP Weissmann N Schroder K NADPH oxidases in cardiovascular disease. Free Radic Biol Med 2010 49 687 706 20444433 28 Lassegue B San Martin A Griendling KK Biochemistry, physiology, and pathophysiology of NADPH oxidases in the cardiovascular system. Circ Res 2012 110 1364 1390 22581922 29 Gimenez M Schickling BM Lopes LR Miller FJ Jr. Nox1 in cardiovascular diseases: regulation and pathophysiology. Clin Sci (Lond) 2016 130 151 165 26678171 30 Lindsey SH Cohen JA Brosnihan KB Gallagher PE Chappell MC Chronic treatment with the G protein-coupled receptor 30 agonist G-1 decreases blood pressure in ovariectomized mRen2.Lewis rats. Endocrinology 2009 150 3753 3758 19372194 31 Dennis MK Burai R Ramesh C Petrie WK Alcon SN Nayak TK Bologa CG Leitao A Brailoiu E Deliu E Dun NJ Sklar LA Hathaway HJ Arterburn JB Oprea TI Prossnitz ER In vivo effects of a GPR30 antagonist. Nat Chem Biol 2009 5 421 427 19430488 32 Dennis MK Field AS Burai R Ramesh C Petrie WK Bologa CG Oprea TI Yamaguchi Y Hayashi SI Sklar LA Hathaway HJ Arterburn JB Prossnitz ER Identification of a GPER/GPR30 antagonist with improved estrogen receptor counterselectivity. J Steroid Biochem Mol Biol 2011 127 358 366 21782022 33 Rey FE Cifuentes ME Kiarash A Quinn MT Pagano PJ Novel competitive inhibitor of NAD(P)H oxidase assembly attenuates vascular O(2)(−) and systolic blood pressure in mice. Circ Res 2001 89 408 414 11532901 34 Csanyi G Cifuentes-Pagano E Al Ghouleh I Ranayhossaini DJ Egana L Lopes LR Jackson HM Kelley EE Pagano PJ Nox2 B-loop peptide, Nox2ds, specifically inhibits the NADPH oxidase Nox2. Free Radic Biol Med 2011 51 1116 1125 21586323 35 Ambasta RK Schreiber JG Janiszewski M Busse R Brandes RP Noxa1 is a central component of the smooth muscle NADPH oxidase in mice. Free Radic Biol Med 2006 41 193 201 16814099 36 Niu XL Madamanchi NR Vendrov AE Tchivilev I Rojas M Madamanchi C Brandes RP Krause KH Humphries J Smith A Burnand KG Runge MS Nox activator 1: a potential target for modulation of vascular reactive oxygen species in atherosclerotic arteries. Circulation 2010 121 549 559 20083677 37 Lavigne MC Malech HL Holland SM Leto TL Genetic demonstration of p47phox-dependent superoxide anion production in murine vascular smooth muscle cells. Circulation 2001 104 79 84 11435342 38 Munzel T Sinning C Post F Warnholtz A Schulz E Pathophysiology, diagnosis and prognostic implications of endothelial dysfunction. Ann Med 2008 40 180 196 18382884 39 Capell BC Collins FS Nabel EG Mechanisms of cardiovascular disease in accelerated aging syndromes. Circ Res 2007 101 13 26 17615378 40 Merideth MA Gordon LB Clauss S Sachdev V Smith AC Perry MB Brewer CC Zalewski C Kim HJ Solomon B Brooks BP Gerber LH Turner ML Domingo DL Hart TC Graf J Reynolds JC Gropman A Yanovski JA Gerhard-Herman M Collins FS Nabel EG Cannon RO 3rd Gahl WA Introne WJ Phenotype and course of Hutchinson-Gilford progeria syndrome. N Engl J Med 2008 358 592 604 18256394 41 Minamino T Komuro I Vascular aging: insights from studies on cellular senescence, stem cell aging, and progeroid syndromes. Nat Clin Pract Cardiovasc Med 2008 5 637 648 18762784 42 Furchgott RF Zawadzki JV The obligatory role of endothelial cells in the relaxation of arterial smooth muscle by acetylcholine. Nature 1980 288 373 376 6253831 43 Dikalova A Clempus R Lassegue B Cheng G McCoy J Dikalov S San Martin A Lyle A Weber DS Weiss D Taylor WR Schmidt HH Owens GK Lambeth JD Griendling KK Nox1 overexpression potentiates angiotensin II-induced hypertension and vascular smooth muscle hypertrophy in transgenic mice. Circulation 2005 112 2668 2676 16230485 44 Matsuno K Yamada H Iwata K Jin D Katsuyama M Matsuki M Takai S Yamanishi K Miyazaki M Matsubara H Yabe-Nishimura C Nox1 is involved in angiotensin II-mediated hypertension: a study in Nox1-deficient mice. Circulation 2005 112 2677 2685 16246966 45 Bopassa JC Eghbali M Toro L Stefani E A novel estrogen receptor GPER inhibits mitochondria permeability transition pore opening and protects the heart against ischemia-reperfusion injury. Am J Physiol Heart Circ Physiol 2010 298 H16 23 19880667 46 Rockey DC Bell PD Hill JA Fibrosis--a common pathway to organ injury and failure. N Engl J Med 2015 372 1138 1149 25785971 47 Ram R Mickelsen DM Theodoropoulos C Blaxall BC New approaches in small animal echocardiography: imaging the sounds of silence. Am J Physiol Heart Circ Physiol 2011 301 H1765 1780 21873501 48 Matsusaka T Katori H Homma T Ichikawa I Mechanism of cardiac fibrosis by angiotensin. New insight revealed by genetic engineering. Trends Cardiovasc Med 1999 9 180 184 10881748 49 Kunieda T Minamino T Nishi J Tateno K Oyama T Katsuno T Miyauchi H Orimo M Okada S Takamura M Nagai T Kaneko S Komuro I Angiotensin II induces premature senescence of vascular smooth muscle cells and accelerates the development of atherosclerosis via a p21-dependent pathway. Circulation 2006 114 953 960 16908765 50 Min LJ Mogi M Iwanami J Li JM Sakata A Fujita T Tsukuda K Iwai M Horiuchi M Cross-talk between aldosterone and angiotensin II in vascular smooth muscle cell senescence. Cardiovasc Res 2007 76 506 516 17706954 51 Nguyen Dinh Cat A Montezano AC Burger D Touyz RM Angiotensin II NADPH oxidase, and redox signaling in the vasculature. Antioxid Redox Signal 2013 19 1110 1120 22530599 52 Tsai P Ichikawa K Mailer C Pou S Halpern HJ Robinson BH Nielsen R Rosen GM Esters of 5-carboxyl-5-methyl-1-pyrroline N-oxide: a family of spin traps for superoxide. J Org Chem 2003 68 7811 7817 14510560 53 Shi H Timmins G Monske M Burdick A Kalyanaraman B Liu Y Clement JL Burchiel S Liu KJ Evaluation of spin trapping agents and trapping conditions for detection of cell-generated reactive oxygen species. Arch Biochem Biophys 2005 437 59 68 15820217 54 Ramiro-Diaz JM Giermakowska W Weaver JM Jernigan NL Gonzalez Bosc LV Mechanisms of NFATc3 activation by increased superoxide and reduced hydrogen peroxide in pulmonary arterial smooth muscle. Am J Physiol Cell Physiol 2014 307 C928 938 25163518 55 Guerra AN Gavala ML Chung HS Bertics PJ Nucleotide receptor signalling and the generation of reactive oxygen species. Purinergic Signal 2007 3 39 51 18404417 56 Schreiner CE Kumerz M Gesslbauer J Schachner D Joa H Erker T Atanasov AG Heiss EH Dirsch VM Resveratrol blocks Akt activation in angiotensin II-or EGF-stimulated vascular smooth muscle cells in a redox-independent manner. Cardiovasc Res 2011 90 140 147 21071431 57 Schroder K Helmcke I Palfi K Krause KH Busse R Brandes RP Nox1 mediates basic fibroblast growth factor-induced migration of vascular smooth muscle cells. Arterioscler Thromb Vasc Biol 2007 27 1736 1743 17541028 58 Banfi B Clark RA Steger K Krause KH Two novel proteins activate superoxide generation by the NADPH oxidase NOX1. J Biol Chem 2003 278 3510 3513 12473664 59 Lassegue B Sorescu D Szocs K Yin Q Akers M Zhang Y Grant SL Lambeth JD Griendling KK Novel gp91(phox) homologues in vascular smooth muscle cells : nox1 mediates angiotensin II-induced superoxide formation and redox-sensitive signaling pathways. Circ Res 2001 88 888 894 11348997 60 Brandes RP Miller FJ Beer S Haendeler J Hoffmann J Ha T Holland SM Gorlach A Busse R The vascular NADPH oxidase subunit p47phox is involved in redox- mediated gene expression. Free Radic Biol Med 2002 32 1116 1122 12031896 61 Al Ghouleh I Frazziano G Rodriguez AI Csanyi G Maniar S St Croix CM Kelley EE Egana LA Song GJ Bisello A Lee YJ Pagano PJ Aquaporin 1, Nox1, and Ask1 mediate oxidant-induced smooth muscle cell hypertrophy. Cardiovasc Res 2013 97 134 142 22997161 62 Li WG Miller FJ Jr. Zhang HJ Spitz DR Oberley LW Weintraub NL H(2)O(2)-induced O(2) production by a non-phagocytic NAD(P)H oxidase causes oxidant injury. J Biol Chem 2001 276 29251 29256 11358965 63 Drummond GR Selemidis S Griendling KK Sobey CG Combating oxidative stress in vascular disease: NADPH oxidases as therapeutic targets. Nat Rev Drug Discov 2011 10 453 471 21629295 64 Schramm A Matusik P Osmenda G Guzik TJ Targeting NADPH oxidases in vascular pharmacology. Vascul Pharmacol 2012 56 216 231 22405985 65 Nigro J Osman N Dart AM Little PJ Insulin resistance and atherosclerosis. Endocr Rev 2006 27 242 259 16492903 66 Kane GC Karon BL Mahoney DW Redfield MM Roger VL Burnett JC Jr. Jacobsen SJ Rodeheffer RJ Progression of left ventricular diastolic dysfunction and risk of heart failure. JAMA 2011 306 856 863 21862747 67 Montezano AC Touyz RM Molecular mechanisms of hypertension--reactive oxygen species and antioxidants: a basic science update for the clinician. Can J Cardiol 2012 28 288 295 22445098 68 Kris-Etherton PM Lichtenstein AH Howard BV Steinberg D Witztum JL P. A. Nutrition Committee of the American Heart Association Council on Nutrition, Metabolism, Antioxidant vitamin supplements and cardiovascular disease. Circulation 2004 110 637 641 15289389 69 Cifuentes-Pagano E Meijles DN Pagano PJ The quest for selective nox inhibitors and therapeutics: challenges, triumphs and pitfalls. Antioxid Redox Signal 2014 20 2741 2754 24070014 70 Guzik TJ Channon KM Measurement of vascular reactive oxygen species production by chemiluminescence. Methods Mol Med 2005 108 73 89 16028677 71 Muscoli C Cuzzocrea S Riley DP Zweier JL Thiemermann C Wang ZQ Salvemini D On the selectivity of superoxide dismutase mimetics and its importance in pharmacological studies. Br J Pharmacol 2003 140 445 460 14522841 72 Dikalov S Griendling KK Harrison DG Measurement of reactive oxygen species in cardiovascular studies. Hypertension 2007 49 717 727 17296874 73 Zanetti M d'Uscio LV Peterson TE Katusic ZS O'Brien T Analysis of superoxide anion production in tissue. Methods Mol Med 2005 108 65 72 16028676 74 Meyer MR Fredette NC Barton M Prossnitz ER Regulation of vascular smooth muscle tone by adipose-derived contracting factor. PLoS One 2013 8 e79245 24244459 75 Kretz M Mundy AL Widmer CC Barton M Early aging and anatomic heterogeneity determine cyclooxygenase-mediated vasoconstriction to angiotensin II in mice. J Cardiovasc Pharmacol 2006 48 30 33 16954818 76 Schena M Mulatero P Schiavone D Mengozzi G Tesio L Chiandussi L Veglio F Vasoactive hormones induce nitric oxide synthase mRNA expression and nitric oxide production in human endothelial cells and monocytes. Am J Hypertens 1999 12 388 397 10232499 77 Zhou Y Dirksen WP Babu GJ Periasamy M Differential vasoconstrictions induced by angiotensin II: role of AT1 and AT2 receptors in isolated C57BL/6J mouse blood vessels. Am J Physiol Heart Circ Physiol 2003 285 H2797 2803 12907424 78 Schlote J Schroder A Dahlmann A Karpe B Cordasic N Daniel C Hilgers KF Titze J Amann K Benz K Cardiovascular and renal effects of high salt diet in GDNF+/− mice with low nephron number. Kidney Blood Press Res 2013 37 379 391 24247178 79 Foster FS Hossack J Adamson SL Micro-ultrasound for preclinical imaging. Interface Focus 2011 1 576 601 22866232 80 Derumeaux G Ichinose F Raher MJ Morgan JG Coman T Lee C Cuesta JM Thibault H Bloch KD Picard MH Scherrer-Crosbie M Myocardial alterations in senescent mice and effect of exercise training: a strain rate imaging study. Circ Cardiovasc Imaging 2008 1 227 234 19808547
PMC005xxxxxx/PMC5124510.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 0416575 7769 Toxicol Appl Pharmacol Toxicol. Appl. Pharmacol. Toxicology and applied pharmacology 0041-008X 1096-0333 27773686 5124510 10.1016/j.taap.2016.10.017 NIHMS828282 Article Differential modulation of FXR activity by chlorophacinone and ivermectin analogs Hsu Chia-Wen a Hsieh Jui-Hua b Huang Ruili a Pijnenburg Dirk c Khuc Thai a Hamm Jon d Zhao Jinghua a Lynch Caitlin a van Beuningen Rinie c Chang Xiaoqing d Houtman Rene c Xia Menghang a a NIH Chemical Genomics Center, National Center for Advancing Translational Sciences, National Institutes of Health, Bethesda, MD, USA b National Toxicology Program, National Institutes of Environmental Health Sciences, National Institutes of Health, Research Triangle Park, NC, USA c PamGene International B.V., Wolvenhoek 10, 5211 HH ‘s-Hertogenbosch, The Netherlands d Integrated Laboratory System, Inc., Morrisville, NC, USA * Address correspondence to: Menghang Xia, Ph.D., National Institutes of Health, National Center for Advancing Translational Sciences, 9800 Medical Center Drive, Bethesda, MD 20892, Phone: 301-217-5718, [email protected] 8 11 2016 20 10 2016 15 12 2016 15 12 2017 313 138148 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Chemicals that alter normal function of farnesoid X receptor (FXR) have been shown to affect the homeostasis of bile acids, glucose, and lipids. Several structural classes of environmental chemicals and drugs that modulated FXR transactivation were previously identified by quantitative high-throughput screening (qHTS) of the Tox21 10K chemical collection. In the present study, we validated the FXR antagonist activity of selected structural classes, including avermectin anthelmintics, dihydropyridine calcium channel blockers, 1,3-indandione rodenticides, and pyrethroid pesticides, using in vitro assay and quantitative structural-activity relationship (QSAR) analysis approaches. (Z)-Guggulsterone, chlorophacinone, ivermectin, and their analogs were profiled for their ability to alter CDCA-mediated FXR binding using a panel of 154 coregulator motifs and to induce or inhibit transactivation and coactivator recruitment activities of constitutive androstane receptor (CAR), liver X receptor alpha (LXRα), or pregnane X receptor (PXR). Our results showed that chlorophacinone and ivermectin had distinct modes of action (MOA) in modulating FXR-coregulator interactions and compound selectivity against the four aforementioned functionally-relevant nuclear receptors. These findings collectively provide mechanistic insights regarding compound activities against FXR and possible explanations for in vivo toxicological observations of chlorophacinone, ivermectin, and their analogs. Farnesoid X receptor nuclear receptor ivermetin moxidectin chlorophacinone diphacinone INTRODUCTION The farnesoid X receptor (FXR) (NR1H4) is a multi-functional nuclear receptor that governs the equilibrium of endogenous bile acids, glucose, and lipids (Kemper, 2011). In the cytoplasm, FXR activity can be stimulated by endogenous ligands such as chenodeoxycholic acid (CDCA) through its ligand-binding domain (LBD). Upon activation, agonist-bound FXR forms a heterodimer with retinoid X receptor (RXR) in the nucleus, where the DNA-binding domain (DBD) of FXR can bind to FXR response elements (FXREs) of target genes for transactivation. FXR transcription is tightly regulated by coactivator and corepressor proteins. These coactivators and corepressors alter gene expression by modulation of FXR-RXR interactions or post-translational modifications (PTMs) of the FXR target gene-regulating histones. Examples of FXR-interacting coregulators include cAMP-response-element-binding (CREB)-binding protein (CBP or CREBBP), E1A binding protein p300 (EP300 or p300), mediator of RNA polymerase II transcription subunit 1 (MED1), nuclear receptor co-repressor 2 (NCOR2/SMRT), peroxisome proliferator-activated receptor gamma (PPARγ) coactivator 1-alpha (PGC-1α), receptor-interacting protein 140 (RIP140), and steroid receptor coactivators (e.g., SRC1, SRC2, and SRC3) (Lien et al., 2014). Reducing FXR activity by small molecules, gene silencing, or gene knockout methods showed either beneficial or adverse effects in various in vivo models, depending on health states, age and diets (Lamers et al., 2014). Guggulsterone, an FXR antagonist, reduced the hepatic cholesterol levels in an FXR specific manner when mice were fed a high-cholesterol diet (Urizar et al., 2002). Ivermectin, a widely used antiparasitic drug, recently identified as an FXR modulator, was also shown to decrease cholesterol and serum glucose levels. As a result, insulin sensitivity improved in mice through interactions with FXR, suggesting that ivermectin or its analogues as possible treatments for diabetes (Jin et al., 2013; Hsu et al., 2014). Studies using FXR deficient mouse models showed both beneficial and adverse effects of FXR deficiency. Zhang et al. reported that compared to wild type controls, female Fxr−/− mice had increased energy expenditure and were resistant to diet-induced obesity (Zhang et al., 2012). On the other hand, several studies have demonstrated that Fxr−/− mice were more susceptible to liver injury induced by chemicals including bile acids, and developed severe non-alcoholic steatohepatitis (NASH)-like liver pathology, especially on a high-fat diet (Yang et al., 2007; Kong et al., 2009; Bjursell et al., 2013). Studies also indicated that FXR deficiency deregulated intestinal innate immunity, altered neurobehavior and promoted tumorigenesis in mice (Maran et al., 2009; Wang et al., 2010). In addition to FXR, constitutive androstane receptor (CAR), liver X receptor (LXR), and pregnane X receptor (PXR) also heterodimerize with the RXR and regulate transcription of genes involved in bile acid (Guo et al., 2003) and lipid (Handschin and Meyer, 2005) homeostasis, complicating interpretation of the results generated with FXR. Unlike effects mediated through other endocrine- or xenobiotic metabolism-regulating nuclear receptors such as estrogen receptors (ERs) and PPARs, effects of environmental chemicals on FXR have not been well characterized. Due to the lack of characterization and important role in homeostasis, FXR was selected as one of the targets for toxicological investigation by the Toxicology in the 21st Century (Tox21) program (Huang et al., 2011; Hsu et al., 2014). We previously profiled the chemicals present in the Tox21 10K chemical collection for their ability to modulate FXR agonism and antagonism on a quantitative high-throughput screening (qHTS) platform and identified over 3,000 environmental chemicals and drugs that potentially stimulate and/or inhibit CDCA-induced FXR transactivation (Hsu et al., 2014). The study identified several FXR-active structural classes, including anthracycline-based chemotherapeutic agents, avermectin and milbemycin-based anthelmintics, benzimidazole-based anti-fungal agents and fungicides, dihydropyridine-based calcium channel blockers, pyrethroid-based pesticides, and vinca alkaloid-based chemotherapeutic agents as FXR antagonists. In the present report, we confirm the FXR antagonist activity of the aforementioned compounds and focus on the characterization of chlorophacinone, ivermectin and their analogs in modulating FXR, CAR, LXRα, and PXR activities using in vitro and computational approaches. MATERIALS AND METHODS Chemicals and cell culture reagents FXR-bla HEK293 cells, components of individual thaw media, growth media, assay media, beta-lactamase assay reagents, and coactivator assays for CAR, FXR, and LXRα were purchased from Life Technologies (Carlsbad, CA, USA). CAR-luc cells were generated as described previously (Lynch et al., 2015). LXRα-luc and PXR-luc assays were acquired from Indigo Biosciences (State College, PA, USA) and Puracyp (Carlsbad, CA, USA), respectively. Compound solutions were prepared from NCATS’ in-house compound collections (e.g., Tox21 10K library). Powder compounds were purchased from SelleckChem (Houston, TX, USA) and Sigma-Aldrich (St. Louise, MO, USA) and sample quality control (QC) analysis was measured on Agilent LC/UV/MS system (Santa Clara, CA, USA) using a standard gradient of 4-100% acetonitrile in the presence of 0.05% trifluoroacetic acid over seven minutes on a 3 micron Luna C18 (3×75 mm) column. All 1536-well compound storage and assay plates were obtained from Greiner Bio One USA (Monroe, NC, USA). CellTiter-Fluor cell viability, CellTiter-Glo cell viability, and One-Glo luciferase assay reagents were acquired from Promega (Madison, WI, USA). QSAR analysis To identify important structural features of FXR antagonists that are active in both FXR-bla transactivation assay and FXR-SRC2 coactivator assay, QSAR analysis was conducted. To increase confidence of the actives, we used a >30% efficacy cutoff. The FXR antagonists identified in the FXR-bla transactivation assay with efficacy >30% were categorized into Group 1 and Group 2 depending on their activity in the FXR-SRC2 coactivator assay. Group 1 included compounds that had no activity in the FXR coactivator assay. Group 2 included compounds that had significant activity in the FXR coactivator assay with >30% efficacy and <3-fold EC50 change between the transactivation assay and coactivator assay. The chemical structures of compounds in both groups were curated to remove inorganic and organometallic salts. The curated structures were fed into CASE Ultra software (MultiCase Inc., Cleveland, OH) for QSAR analysis. CASE Ultra generates dataset-dependent chemical fragments and performs statistical analysis for identification of fragments that are differentiable between Group 1 and 2. The model was validated based on repeated 10-fold external cross validation, yielding area under the curve (AUC) values of receiver operating characteristic (ROC) curve, which estimates the prediction power of QSAR model. FXR-bla reporter gene assay FXR-bla HEK293 cells were seeded at 5000 cells per 5 μL per well in 1536-well black clear-bottom assay plates using a BioRAPTR flying reagent dispenser (FRD) (Beckman Coulter, Indianapolis, IN, USA), followed by incubation for 5 h in a humidified incubator at 37°C, 5% CO2. Added to the plates were 23 nL per well of compound solution prepared by cherry picking on a compound transfer workstation (Kalypsys, San Diego, CA, USA) and 1 μL of CDCA (50 μM final concentration)-containing medium for the FXR-bla antagonist assay using BioRAPTR FRD. After 16 h incubation in a humidified incubator set at 37°C, 5% CO2, 1 μL of CCF4-AM substrate was added to each well using a BioRAPTR and incubated at room temperature in the dark for additional 2 h. An excitation wavelength of 405 nm and two emission wavelengths of 530 nm (Ch1) and 460 nm (Ch2) were used to obtain FXR-bla reporter signals on an Envision plate reader (Perkin Elmer, Shelton, CT, USA), yielding beta-lactamase activity as fluorescence emission intensity ratios (Ch2/Ch1). Ratios were normalized to CDCA control wells as 100% FXR activity and DMSO-only wells (0.4% v/v DMSO) as 0% FXR activity. The concentration-response curves of the cherry-pick confirmation and powder confirmation plates were generated using an in-house data analysis protocol (Hsu et al., 2014) and GraphPad Prism software (GraphPad Software, La Jolla, CA, USA) to determine IC50 and efficacy values, respectively. CAR-luc, LXRα-luc, and PXR-luc reporter gene assays The CAR-luc and PXR-luc assays, generated by stably transfected parent cell lines with a nuclear receptor-driven luciferase reporter gene, were conducted as described previously (Shukla et al., 2011; Lynch et al., 2015). The LXRα-luc assay was optimized into a 1536-well format and conducted according to the manufacturer’s instructions. Briefly, CAR-luc, LXRα-luc, or PXR-luc cells were seeded into 1536-well white solid-bottom plates at 2000 to 2500 cells/well using BioRAPTR FRD, 23 nL of compound solution were added in each well, and the plates were then incubated for an indicated period in a humidified incubator at 37°C, 5% CO2. For antagonist mode, the CAR-luc, LXRα-luc, and PXR-luc cells had 1 μL of agonist-containing medium using BioRAPTR FRD added to every well, yielding final concentrations of 50 nM CAR agonist (CITCO), 0.4 μM LXRα agonist (T0901317), and 5 μM PXR agonist (rifampicin), respectively. The CAR-luc and PXR-luc cells were incubated with 4 μL per well of One-Glo luciferase reagent at room temperature for 20 min. The LXRα-luc cells were incubated with 5 μL per well of Luciferase Detection Reagent at room temperature for 10 min. The luciferase signals were acquired as luminescence intensity values on a ViewLux plate reader and normalized to DMSO-only wells (0.4% v/v DMSO) as 0% nuclear receptor activity and agonist control wells as 100% nuclear receptor activity. Concentration-response curves of each compound were generated using GraphPad Prism software to determine EC50, IC50, and efficacy values. Cell viability assays Assay plates of FXR-bla or LXRα-luc cells were added with 3 μL of CellTiter-Glo cell viability reagent per well using BioRAPTR FRD after fluorescence reading for FXR-bla cells and after compound treatment for LXRα-luc cells, followed by a 30-minute incubation in the dark at room temperature. Assay plates of CAR-luc or PXR-luc cells were added with 1 μL of CellTiter-Fluor cell viability reagent per well using BioRAPTR FRD before luminescence reading, followed by 30-minute incubation in a humidified incubator at 37°C, 5%CO2. The luminescence intensity values of CellTiter-Glo reagent and the fluorescence intensity values of CellTiter-Fluor reagent were acquired on a ViewLux plate reader and normalized to DMSO-only wells (0.4% v/v DMSO) as 100% cell viability and tetraoctylammonium bromide (TOAB) control wells as 0% cell viability. Concentration-response curves of each compound were generated using GraphPad Prism software to determine IC50 and efficacy values. Nuclear receptor coactivator assays for CAR, FXR, and LXRα The coactivator assays of CAR, FXR, and LXRα were miniaturized into 1536-well formats and conducted according to manufacturer’s instructions. Compound solutions of 23 nL prepared by cherry picking were added to a 6 μL mixture of the nuclear receptor of interest’s ligand binding domain (LBD) with detection reagents at indicated concentrations in 1536-well black solid bottom assay plates using a compound transfer workstation. The plates were centrifuged at 1000 rpm for 1 min, and incubated at room temperature in the dark for 30 min. The plates were then measured using an Envision plate reader with an excitation wavelength of 340 nm and two emission wavelengths of 495 nm and 520 nm. The CAR assay used 10 nM GST-tagged CAR-LBD protein, 5 nM Tb-labeled anti-GST antibody (goat), and 125 nM fluorescein-labeled PGC1α coactivator-derived peptide in the absence (agonist mode) and presence (antagonist mode) of 0.1 μM CITCO. The FXR assay used 5 nM GST-tagged FXR-LBD protein, 5 nM Tb-labeled anti-GST antibody (goat), and 500 nM fluorescein-labeled SRC2-2 coactivator-derived peptide in the absence (agonist mode) and presence (antagonist mode) of 50 μM CDCA. The LXRα assay used 5 nM GST-tagged LXRα-LBD protein, 10 nM Tb-labeled anti-GST antibody (goat), and 250 nM fluorescein-labeled TRAP220/DRIP-2 coactivator-derived peptide in the absence (agonist mode) and presence (antagonist mode) of 0.4 μM T0901317. Fluorescence ratios (520 nm/495 nm) were normalized to DMSO-only wells (0.4% v/v DMSO) as 0% nuclear receptor activity and agonist control wells as 100% nuclear receptor activity. Concentration-response curves of each compound were generated using GraphPad Prism software to determine EC50/IC50 and efficacy values. Microarray assay for real-time coregulator-nuclear receptor interaction (MARCoNI) Ligand-modulated interaction of FXR with coregulators was assessed using a PamChip® plate (PamGene International B.V.’s-Hertogenbosch, The Netherlands), which allows for analysis of 96 identical peptide microarrays in parallel (Koppen et al., 2009; Houtman et al., 2012; Aarts et al., 2013). Each array of 154 coregulator-derived motifs (PamChip #88101, PamGene, The Netherlands) was incubated with the assay mixture containing the receptor and detection antibody in the absence or presence of compound(s). In short, compounds were prediluted in DMSO. Next, assay mixtures (Coregulator Buffer G, PV4553, Invitrogen, USA) were prepared on ice in a master 96-well plate and contained 5 nM GST-tagged FXR ligand-binding domain (FXR LBD-GST, PV4834, Invitrogen, USA), 25 nM of Alexa488-conjugated GST-antibody (A11131, Invitrogen, USA), 60 μM CDCA (antagonist mode) and 10 or 20 μM test compound (final DMSO concentration was 2% v/v). Each condition was analyzed using four technical replicates (arrays). Incubation was performed in a fully automated microarray processing platform, PamStation®96, PamGene, The Netherlands) at 20 °C for 80 cycles (two cycles per min). After removal of the unbound receptor by washing each array with 25 μL of Tris-buffered saline, tiff images were obtained by the CCD camera which is part of the PamStation®96. Tiff images were analyzed for quantification of FXR binding (in arbitrary units of fluorescence) using Bionavigator software (PamGene, The Netherlands) and data analysis and visualization were performed by R software v2.15.3. The significance level of data (p<0.05; p<0.01; p<0.001) were determined by Student’s t-Test. RESULTS Confirmation of FXR Antagonist Activity in Receptor Transactivation and Coactivator Recruitment Assays A selection of FXR-active compounds identified from the primary screening of the Tox21 10K chemical collection (Hsu et al., 2014) were prepared by cherry picking and re-tested for their ability to inhibit chenodeoxycholic acid (CDCA)-induced FXR transactivation, cell viability of FXR-bla cells, and CDCA-induced recruitment of a SRC2 coactivator. In addition to the positive controls—(E)-guggulsterone and (Z)-guggulsterone, the selected compounds consists of structural classes including: anthracycline chemotherapeutics (e.g., daunorubicin and doxorubicin), avermectin and milbemycin anthelmintics (e.g., ivermectin and moxidectin), benzimidazole anti-fungal agent and fungicides (e.g., albendazole and nocodazole), conazole fungicides (e.g., econazole and sulconazole), dihydropyridine calcium channel blockers (e.g., nicardipine and nimodipine), 1,3-indandione rodenticides (e.g., chlorophacinone and diphacinone), organochloride pesticides (e.g., chlordane and dieldrin), pyrethroid pesticides (e.g., bifenthrin and cypermethrin), and vinca alkaloid chemotherapeutics (e.g., vinblastine and vincristine). Despite the fact that conazoles displayed apparent cytotoxicity, the majority of compounds in the aforementioned structural classes reduced CDCA-induced FXR transactivation at non-cytotoxic or slightly cytotoxic concentrations in the FXR-bla reporter gene assay measuring FXR-driven expression of the beta-lactamase reporter gene (Table S1). Seven out of ten of FXR-bla-active structural classes seemed to disrupt CDCA-induced recruitment of an SRC2 coactivator-derived peptide to the ligand binding domain (LBD) of FXR in the time resolved fluorescence resonance energy transfer (TR-FRET)-based biochemical assay (Table S2). All of the avermectin- and milbemectin-based anthelmintic drugs were identified by the coactivator assay as FXR antagonists with a wide range of IC50 values from 1.74 to 34.74 μM (Fig. 1A). Nicardipine, one of the dihydropyridine calcium channel blockers, was the most potent and efficacious with an IC50 of 1.59 ± 0.57 μM in the coactivator assay (Fig. 1B). 1,3-Indandione rodenticides (Fig. 1C) displayed weak FXR antagonist activity (efficacy <50%) in inhibiting CDCA-induced recruitment of SRC2-2 coactivator peptide to FXR-LBD. Most benzimide-based microtubule inhibitors and anthracycline-based chemotherapeutic agents such as albendazole and daunorubicin did not show CDCA-dependent reduction of FXR activity in the SRC2-2 coactivator assay (Table S2). Using QSAR Analysis to Identify Structural Features of FXR Antagonists Not all actives identified in the confirmatory FXR-bla transactivation assay were active in the FXR-SRC2 coactivation assay; therefore the validated results of the selected compounds in the two FXR antagonist assays were further analyzed to identify functional groups important for FXR-SRC2 binding interactions using quantitative structural activity relationship (QSAR) analysis. Compounds with >30% efficacy in the FXR assays were assigned as active compounds. We first categorized the identified FXR antagonists into Group 1 of compounds that were only active in the FXR-bla assay and Group 2 of compounds that were active in both the FXR-bla and the FXR-SRC2 coactivator assays with a <3-fold difference in EC50 values. In total, there were 53 and 62 compounds in Group 1 and Group 2, respectively. After chemical structure curation and data collapsing, in total, there were 43 and 60 compounds in Group 1 and Group 2 amenable for QSAR analysis, respectively. To identify the structural features crucial for modulating FXR-SRC2 interactions, SAR analysis were conducted based on the structural fragments of the test compounds and compound activities in the two FXR antagonist assays (Table 1). The area under the receiver operating characteristic (AUC-ROC) curve was over 0.78, indicating a high accuracy of the model (i.e., AUC ≤0.5 for random data). Fifteen avermectin anthelmintics (e.g., ivermectin) and cyclodiene pesticides (e.g., aldrin), with a common cyclodiene-containing fragment, were classified to Group 2. Several Group 2 compounds were environmental chemicals that contain a phenol group (e.g, 2,4-bis(2-methylbutan-2-yl)phenol and 2,2’-methylenebis(4-methyl-6-tert-butylphenol)) or a bisphenol group (e.g., bisphenol B and oxyphenisatin). Five phenylacetone-containing compounds such as dihydropyridine-based calcium channel blockers (e.g., cilnidipine and nimodipine) and 1,3-indandione-based rodenticides (e.g., chlorophacinone and diphacinone) were also Group 2 compounds. Among the Group 2 compounds, avermectin and milbemectin-based anthelmintics, 1,3-indandione-based rodenticides, and pyrethroid-based pesticides might cause FXR-dependent toxicities at their exposure levels and thus require more in-depth testing to confirm their FXR antagonist activity as shown by the primary screening and the cherry-pick confirmation data. New batches of representative compounds for each structural class were obtained in powder form from alternative sources and tested in the FXR-bla antagonist assay to further confirm their FXR antagonist activity. While avermectins/milbemectins and 1,3-indanediones had similar potencies in antagonizing CDCA-induced FXR transactivation in the cherry-pick confirmation and powder confirmation experiments (Table 2), the re-procured samples of all tested pyrethroids were inactive in the FXR-bla assay (data not shown). Next, new samples of (E)-guggulsterone, (Z)-guggulsterone, ivermectin, moxidectin (a milbemectin derivative), the other three FXR-active ivermectin analogs (abamectin, doramectin and selamectin), chlorophacinone, diphacinone, and additional four chlorophacinone analogs (2-acetyl-1,3-indanedione, 1,1-diphenylacetone, phenindione and pindone) (Fig. 2A) were first tested for their FXR antagonist activity in the FXR-bla, viability, and FXR coactivator assays. The FXR antagonist activities of (E)-guggulsterone and (Z)-guggulsterone were confirmed in the FXR-bla and the FXR coactivator assays (Table 2). Despite the fact that several avermectins and moxidectin were cytotoxic to FXR-bla cells, the IC50 values of the five compounds yielded a good correlation coefficient (R2 = 0.94) in inhibiting CDCA-induced FXR transactivation and recruitment of SRC2-2 coactivator-derived peptide. The three FXR-active 1,3-indandiones identified from the primary screening, chlorophacinone, diphacinone and pindone, reduced CDCA-induced FXR-bla expression and the first two compounds also antagonized CDCA-induced coactivator recruitment. Two structural analogs of chlorophacinone—2-acetyl-1,3-indanedione and phenindione— had no FXR antagonist activity in either FXR-bla or coactivator assays. 1,1-Diphenylacetone, a structural fragment present in chlorophacinone and diphacinone, showed weak FXR antagonist activity in the FXR-bla and FXR coactivator assays with an efficacy of <30%. Profiling of FXR-coregulator interactions of avermectins and 1,3-indandiones An allosteric conformational change of nuclear receptors upon ligand alters their affinity for coregulators, and the receptor-coregulator complex forms a physical and functional bridge with the gene transcription machinery. Depending on the nature of the ligand, coactivators and corepressors are recruited or displaced thus altering gene transcription levels. In this study we applied MARCoNI (Microarray Assay for Real-time Coregulator Nuclear receptor Interaction) in which a set of 154 immobilized peptides represents a variety of NR-interaction motifs from coregulators. Modulation of CDCA-activated FXR binding (antagonist mode) to these peptides by 12 compounds including (Z)-guggulsterone, ivermectin, chlorophacinone, and their analogs was analyzed. Coregulators display a variety of functions and are differentially expressed in a cell specific manner. Compounds dictate receptor conformation and subsequent affinity for coregulators. Compounds which display similar modulation are expected to act similarly while differential modulation is linked with an alternate mode of action. Fig. 2B shows the modulation of CDCA-activated FXR-coregulator interaction by each of the 12 compounds that were tested in powder confirmation studies. Modulation of each individual interaction is indicated by the modulation index (MI), or compound-induced log-fold change of binding vs. CDCA alone. To visualize similarity and dissimilarity in mode of action, the 154-point modulation signatures of the compounds, were subjected to R-based hierarchical clustering, which was used to classify coregulator interactions. (Z)-Guggulsterone, chlorophacinone, and ivermectin altered CDCA-induced binding between FXR and several known FXR-interacting coregulator motifs (Table 4). The three compounds reduced binding interactions between FXR and the coactivator binding motif of SRC1, SRC2, and SRC3 (NCOA1_677_700, NCOA2_677_700, and NCOA3_673_695). Additionally, ivermectin uniquely strengthened the binding interactions between FXR and one of the corepressor domains of N-CoR1 and SMRT (NCOR1_2251_2273 and NCOR2_2330_2352). A large cluster of interactions at the right side of the heat map (Fig. 2B) displays decreased FXR binding (indicated by blue) upon addition of compound. This cluster contains a number of motifs derived from coactivators. A particular example of this cluster, NCOA2_677_700 (SRC2-2), is shown in more detail in Fig. 3A. Binding of FXR, compared to the DMSO control, was significantly increased by addition of the natural FXR agonist, CDCA. This effect was significantly but not fully reversed by the FXR antagonist, (Z)-guggulsterone. All the FXR antagonists tested in the study displayed inhibitory effect on CDCA-induced FXR activity expressed as coactivator binding, to various extents. All members in the avermectin class displayed antagonism that was similar or superior to the positive control, (Z)-guggulsterone. Chloro- and di-phacinone also displayed strong antagonism. Chlorophacinone even showed the highest efficacy of up to full antagonism of CDCA-induced binding compared to all other tested compounds. Other 1,3-indanediones displayed much weaker antagonism than (Z)-guggulsterone. The selected motif is found in NCOA2, a histone acetyl transferase. This enzymatic activity enhances target gene availability and thus elevates gene transcription levels. A large number of motifs in this cluster are derived from coactivator proteins. Compound-induced antagonism of FXR function, i.e., expression of target genes, can be explained by decreased affinity and subsequent release of transcription facilitating coregulators. While all avermectins, chlorophacinone and diphacinone displayed strong passive antagonism, clustering revealed an additional class of interactions with different mechanisms of action at the far left of the heatmap (Fig. 2B), which is illustrated in detail in Fig. 3B. At basal level (DMSO), FXR interacted with this coregulator motif, but CDCA treatment released this binding. (Z)-Guggulsterone, and many FXR antagonists did not affect this binding; however, avermectins, except for moxidectin, reversed the CDCA effect. Selamectin, abamectin and ivermectin even induced an FXR binding level to this motif that exceeded that of the apo receptor. The selected motif and the other motif in this class of interactions in the heatmap are derived from NCORs. These corepressor proteins are well characterized for histone deacetylation and chromatin condensation, which disfavors gene transcription. Enhanced recruitment of these proteins to FXR by some of the compounds reveals a different mode of action which we refer to as active antagonism (i.e. an active way to reduce car speed is to press a foot to the brake). In addition to classification of the compound by weak and strong passive antagonism, the latter group can be further classified by absence or presence of displayed active antagonism. The FXR-coregulator binding interactions were tested with two concentrations of chlorophacinone (Fig. S1) or three concentrations of ivermectin (Fig. S2) in the presence of CDCA. Ivermectin showed concentration-dependent enhancement in CDCA-mediated FXR binding to several coregulators including CBP_2055_2007, EP300_2039_2061, NCOR1_2251_2273, NCOR2_2330_2352, and NRIP1_120_142 as well as reduction in CDCA-mediated FXR binding to several coregulators including NCOA6_875_897. Chlorophacinone enhanced FXR binding interactions to CNOT1_557_579 and RBL2_875_897_C879S/C894S at higher concentrations, while significantly attenuating FXR binding interactions to several coregulators such as EP300_69_91, LCOR_40_62, JHD2C_2054_2076, TIF1A_747_769, and UBE3A_649_671. Both compounds effectively reduced binding interactions between FXR and several coregulators including BRD8_254_276, CBP_57_80, NCOA1_677_700, NCOA2_677_700, NCOA2_733_755, NCOA3_673_695, NCOA3_725_747, MED1_591_614, MED1_632_655, NR0B2_106_128, PRGC1_130_155, PRGC_134_154, PRGC2_146_166, PRGC2_388_358, and WIPI1_119_141. Receptor Selectivity of avermectins and indandiones To better translate the compound activity of (Z)-guggulsterone, chlorophacinone, ivermectin, and their analogs from receptor assays to functional outcomes, these compounds were also tested for their ability to modulate other functionally relevant nuclear receptors, which coordinate with FXR to regulate downstream signaling, such as the constitutive androstane receptor (CAR), the liver X receptor alpha (LXRα), and pregnane X receptor (PXR) using cell-based luciferase reporter gene assays to measure luciferase expression under the control of the corresponding nuclear receptors and TR-FRET coactivator recruitment assays to measure the receptor-coactivator interactions. In the CAR-luc assay, guggulsterones and several indandiones were identified as potential CAR agonists and avermectins as potential CAR antagonists (Table 2). (E)-guggulsterone, (Z)-guggulsterone, chlorophacinone, diphacinone, and phenindione partially induced CAR transactivation but exerted no agonist effects on recruiting PGC-1α coactivator-derived peptide to CAR-LBD (Table 3). Avermectins were at least 5-fold more potent in inhibiting CITCO-induced coactivator recruitment than CITCO-induced CAR-luc expression with comparable cytotoxicity to CAR-luc cells. In the LXRα-luc assay, avermectins and several indandiones were identified as potential LXRα antagonists (Table 2). Avermectins completely suppressed T0901317-induced LXRα transactivation at slightly cytotoxic concentrations (IC50_Viability/IC50_LXRα-luc > 3), with doramectin as the most potent LXRα antagonist among the tested compounds. Chlorophacinone had similar IC50 values in reducing T0901317-induced LXRα activity in both cell-based and biochemical assays, while diphacione and phenindione were active only in the LXRα-luc assay (Table 2 and Table 3). Guggulsterones showed neither agonistic nor antagonistic effects on LXRα within the tested concentration range. In the PXR-luc assay, guggulsterones showed partial activities on both PXR agonist and antagonist modes. Moxidectin, diphacinone and phenidione had PXR agonist activity, while other avermectins, chlorophacinone and pindone had PXR antagonist activity (Table 2). Both isoforms of guggulsterones were more potent and less efficacious in inducing PXR-luc transactivation than inhibiting rifampicin-induced PXR activation. Doramectin and selamectin potently reduced rifampicin-induced PXR-luc expression with submicromolar IC50 values. Abamectin and pindone also inhibited rifampicin-induced PXR activity at low micromolar concentrations. Due to lack of available PXR coactivator recruitment assays, agonist and antagonist activities of the identified compounds at biochemical level were not explored in this study. DISCUSSION In this study, we combined in vitro and computational approaches to investigate the mechanisms of action of FXR-active chemicals and their analogs identified from a previous Tox21 screening (Hsu et al., 2014). A battery of in vitro cell-based reporter gene assays and biochemical coregulator assays in 1536-well plate or 154-plex microarray formats enabled high-throughput generation of data regarding compound activity and selectivity. Quantitative structure-activity relationship (QSAR) modeling method was employed to identify key structural features correlated to inhibited recruitment of an SRC2 coactivator-derived peptide to FXR. Reporter gene assays are useful experimental methods to assess compounds’ ability to potentially interfere with the physiological functions of targets of interest in a cellular environment, and many reporter gene assays can be adapted to high-throughput screening platforms to rapidly profile large numbers of chemicals. Our previous study (Hsu et al., 2014) used a cell-based FXR-responsive beta-lactamase reporter gene FXR-bla assay to profile the Tox21 10K chemical collection of environmental chemicals and drugs for FXR activity and also identified some potential assay artifacts including compound autofluorescence and compound cytotoxicity by comparing compound activities in available Tox21 10K qHTS data sets generated by a panel of beta-lactamase assays. In the follow-up studies, we utilized the same FXR-bla assay to test compounds prepared from same sources used for primary screening by cherry-picking and from re-procured alternatives. The weak concordance of pyrethroids in FXR activity tested in both cherry-picked and re-procured forms is likely due to the unknown breakdown products of pyrethroids in stocks used for primary screening. In addition, luciferase reporter gene assays were employed to explore compound selectivity against nuclear receptors (e.g., CAR, LXRα, and PXR) that also heterodimerize with the RXR and coordinate with FXR to regulate subsets of downstream effectors and signaling pathways involved in energy and xenobiotic metabolism (Cave et al., 2016). Avermectin- and milbemycin-based drugs (e.g., ivermectin and moxidectin) not only interacted with FXR but also CAR, LXRα, and PXR. The antiparasitic drug ivermectin was previously shown to stimulate FXR activity and to inhibit CDCA-induced FXR activity in cell-based and biochemical assays (Jin et al., 2013; Yu et al., 2013; Hsu et al., 2014). Ivermectin was also shown to decrease cholesterol and serum glucose levels, and improve insulin sensitivity in mice by stimulating FXR activity, suggesting the potential use of ivermectin or its analogs for diabetes treatment (Jin et al., 2013). However, the potential in vivo outcomes of ivermectin’s FXR antagonist activity have not been well characterized. All avermectins tested in our experiments showed antagonistic activity to FXR, CAR, LXRα, and PXR. Moxidectin, a milbemycin analog, antagonized transactivation of FXR, CAR and LXRα, but partially increased the expression of the PXR-luc reporter gene. Our results showed that the macrolactam moiety and the disaccharide moiety of avermectins and milbemycins are crucial for FXR antagonist activity and compound selectivity, respectively. It will be interesting to study how the antagonist activities of these compounds on FXR, CAR, LXRα, and PXR affect their in vivo outcomes. The 1,3-indandione-based rodenticides such as chlorophacinone and diphacinone are identified as FXR antagonists in our study. The two rodenticides are anti-coagulants that act through antagonizing the action of vitamin K. Exposures of chlorophacinone and diphacinone through oral, inhalation, or dermal routes cause toxic effects including hemorrhage, liver necrosis, developmental toxicities, and death. Our results showed that chlorophacinone and diphacinone had FXR antagonist, CAR agonist, and LXRα antagonist activities. The two compounds had distinct activity on PXR: Chlorophacinone partially decreased the SR 12813-induced PXR-luc expression and diphacinone partially stimulated PXR transactivation at high micromolar concentrations. 1,1-Diphenylacetone, which contains the common structural feature in both chlorophacinone and diphacinone, weakly inhibited CDCA-induced FXR transactivation and showed no activity to CAR, LXRα, and PXR. These results indicate the importance of the diphenyl indane moiety in antagonizing FXR activity. Since chlorophacinone and diphacinone are potent rodenticides that might cause severe lethality at concentrations lower than those exerting FXR-mediated toxicities, it will be valuable to use non-rodent species to study their in vivo FXR antagonist activities. The distinct functional profile of nuclear receptors in response to binding of ligands is largely determined by the ligand-specific recruitment of coregulators, which ultimately controls specific transcription of target genes and affects energy and xenobiotic metabolism. Upon ligand binding, FXR undergoes conformational changes to release corepressors such as the nuclear corepressor (NCoR) and recruits coactivators including steroid receptor coactivator-1 (SRC-1). Cholesterol 7α-hydroxylase (CYP7A1) is the rate-limiting enzyme in the conversion of cholesterol to bile acids. It is known that bile acids regulate the transcription of CYP7A1 by feedback repression through FXR binding. Studies have shown that the ability of various bile acids to recruit transcriptional coactivator protein (e,g, SRC-1) to the ligand binding domain of FXR is consistent with their ability to repress Cyp7A1 mRNA levels (Bramlett et al., 2000). On the other hand, in vitro pharmacological studies on liver cells demonstrated that theonellasterol antagonized the expression of FXR target genes including ABCC4 for the multidrug resistance-associated protein 4 (MRP-4) by abolishing the release of nuclear corepressor NCoR from the promoter of genes (Renga et al., 2012), and the in vivo studies using mice with obstructive cholestasis showed that theonellasterol increased MRP-4 expression in the liver and protected mice from liver injury. To determine whether a compound is a ligand of the nuclear receptor, and to assist the prediction of potential outcomes of modulating specific receptor-coregulator interactions, biochemical assays that measure binding interactions between a nuclear receptor and its corresponding coregulator(s) can be used. The two types of coregulator assays used in the present study generated rich information regarding FXR-coregulator interactions. The first coregulator assay is a homogeneous nuclear receptor binding assay that relies on time-resolved fluorescence resonance energy transfer (TR-FRET) between a donor dye for FXR and an acceptor dye for an SRC2 coactivator-derived peptide. Because the terbium-based donor dye used in the TR-FRET assay has a long fluorescence lifetime that can be detected at a later time point, the TR-FRET assay generates less false positives coming from compounds with a short fluorescence lifetime or precipitated compounds that scattered light. Unlike the TR-FRET assay, which measures a single FXR-coactivator interaction at a time, the second coregulator assay, MARCoNI, enables simultaneous measurement of 154 multiplex FXR-coregulator assays by fluorescence imaging of coregulator-bound FXR complexes on a microarray of immobilized peptides derived from various coregulator motifs. Our results showed that (Z)-guggulsterone, chlorophacinone, ivermectin, and some of their analogs antagonized CDCA-induced recruitment of a SRC2-derived peptide to FXR in the TR-FRET assay, but the FXR MARCoNI profiling study revealed that the three structural classes differentially modulated binding interactions between FXR and subsets of coregulator-derived peptides. Our results with ivermectin agree with previous findings, in which ivermectin was shown to partially promote the SRC-1,-2,-3 recruitment by FXR and also induced the recruitment of corepressor NCoR-2 and SMRT-2 by FXR (Jin et al., 2013). Computational approaches are powerful in identifying potential structural features important for modulating FXR activity. The HTS assays conducted in the current study, including the FXR-bla assay, the FXR TR-FRET coactivator assay, and the FXR MARCoNI profiling experiment generated rich information about the FXR activities of compounds and enabled us to develop and apply computational approaches such as QSAR and hierarchical clustering to rapidly analyze and interpret the HTS assay data. Despite the fact that some compounds identified from the FXR-bla antagonist screen were cytotoxic, many of the cytotoxic FXR antagonists also disrupted FXR-SRC2 interactions in the TR-FRET biochemical assay. By grouping compounds according to their FXR-bla and FXR TR-FRET assay results, our QSAR analysis show that not every compound modulates FXR signaling through SRC2; this identifies several structural features unique to SRC2-interacting compounds. Additionally, by hierarchical clustering of the FXR MARCoNI profiling data, we found that chlorophacinone, ivermectin, (Z)-guggulsterone, and their analogs differentially modulated the binding interactions between FXR and subsets of coactivator- or corepressor-derived peptides. In addition to using FXR antagonists, reducing FXR activity by gene knockout is another popular method to study FXR functions. FXR deficiency has been associated with a number of beneficial effects including decreased cholesterol levels. In general, Fxr−/− mice have normal phenotypes that are externally indistinguishable from their wild type littermates: FXR deficient mice are similar to wild type mice in terms of body weight, body composition, fertility, energy intake and expenditure as well as behaviors at a young age (Lamers et al., 2014). However, depending on diet, age and health status, distinguished phenotypes have been observed in Fxr−/− mice; some of which seem to have beneficial health effects. For example, female Fxr−/− mice had increased energy expenditure and were resistant to diet-induced obesity (Zhang et al., 2012). Aging Fxr−/− mice display late onset leanness associated with elevated energy expenditure and improved glucose control (Bjursell et al., 2013). Interestingly, depending on the cancer type, FXR deficiency resulted in either beneficial or adverse effects on tumor growth. Over-expression of FXR was previously found to be associated with larger tumor size and grade as well as lymph node metastasis in esophageal adenocarcinoma, while loss of FXR suppressed tumor growth (Guan et al., 2013). However, some studies showed that sustained high levels of bile acids in Fxr−/− mice may contribute to liver tumorigenesis (Yang et al., 2007) and FXR deficiency increased the size of small intestine adenocarcinomas in adenomatous polyposis coli mutant mice and colon carcinogen-treated C57BL/6 mice (Maran et al., 2009). Taken together, these in vivo findings further illustrate the complex nature of FXR biology. In summary, the current study provides mechanistic insights of in vitro FXR antagonist activity of guggulsterones, avermectin-based drugs, and 1,3-indandione-based rodenticides as well their ability to alter signaling activity of other functionally relevant nuclear receptors, including CAR, LXRα, and PXR. The in vitro results generated from the present study and the in vivo evidence of FXR deficiency from the literature warrant continuing exploration of the identified FXR antagonists in appropriate in vivo models. Supplementary Material 1 Fig. S1. Concentration response activity of chlorophacinone in FXR MARCoNI assays 2 Fig. S2. Concentration response activity of ivermectin in FXR MARCoNI assays 3 4 ACKNOWLEDGEMENT We thank Danielle vanLeer and Paul Shinn for compound management. We thank MultiCASE Inc. for providing their CASE Ultra software for the SAR analysis. FUNDING This work was supported by the U.S. Environmental Protection Agency (Interagency Agreement #Y3-HG-7026-03) and the interagency agreement IAG #NTR 12003 from the National Institute of Environmental Health Sciences/Division of the National Toxicology Program to the National Center for Advancing Translational Sciences, National Institutes of Health. Fig. 1 Selected FXR-active structural classes inhibited CDCA-induced association between FXR and a SRC2-derived coactivator peptide a. Concentration-response curves of avermectin and milbemycin-based anthelmintic drugs. b. Concentration-response curves of dihydropyridine-based calcium channel blockers. c. Concentration-response curves of 1,3-indandione-based rodenticides. All data points were collected in the presence of 50 μM CDCA and shown as mean ± SEM from three experiments. Fig. 2 FXR-coregulator interaction profiling of (Z)-guggulsterone, chlorophacinone, ivermectin, and their analogs a. Chemical structures of compounds tested in the FXR MARCoNI assays. b. Compounds and coregulators were clustered (hierarchical clustering, Euclidean distance, average linkage) based on modulation index (MI). Significance of the binding modulation is indicated (* p <0.05; p <0.01; * p <0.001). Fig. 3 Differential modulation of coactivators and corepressors by (Z)-guggulsterone, chlorophacinone, ivermectin, and their analogs a. Negative modulation of binding between FXR and NCOA2_677_700 (SRC2-2) by all test compounds. b. Positive modulation of binding between FXR and NCOR1_2251_2273 by ivermectin and its analogs. Each data point was presented as mean receptor binding ± SEM from three experiments. Bars are colored by MI standing for modulation index, i.e., compound-induced log-fold change of (CDCA-induced) FXR-coregulator interaction. Significance of the binding modulation (vs. CDCA) is indicated (* p <0.05; p <0.01; * p <0.001). Table 1 Structure-activity relationship (SAR) results of representative FXR-active Tox21 chemicals Structural fragment # in Group1/ # in Group2 Representative compound groups Example chemical structures 0/15 Avermectin-based anthelmintic and organochlorine pesticides 0/6 Environmental chemicals 0/4 Environmental chemicals and laxatives 0/5 Dihydropyridine-based calcium channel blockers and 1,3- indandione-based rodenticides 4/12 Imidazole antifungals 8/0 Anthracycline-based chemotherapeutic 8/2 Calcium sensitizers & dyes Table 2 Selectivity of guggulsterones, avermectins, and 1,3-indandiones against FXR, CAR, LXRα, and PXR in cell-based assays. Compound Name (CASRN) FXR-bla (antagonist) IC50, μM (Efficacy, %) Viability (FXR) IC50, μM (Efficacy, %) CAR-luc (agonist) EC50, μM (Efficacy, %) CAR-luc (antagonist) IC50, μM (Efficacy, %) Viability (CAR) IC50, μM (Efficacy, %) LXR-luc (antagonist) IC50, μM (Efficacy, %) Viability (LXR) IC50, μM (Efficacy, %) PXR-luc (agonist) EC50, μM (Efficacy, %) PXR (antagonist) IC50, μM (Efficacy, %) (E)- Guggulsterone (39025-24-6) 15.88 ± 3.79 (65 ± 10) Inactive (35 ± 4 at 38 μM) Inactive Inactive (48 ± 12 at 38 μM) Inactive (26 ± 2 at 38 μM) (35 ± 11 at 38 μM) (Z)- Guggulsterone (39025-23-5) 20.03 ± 2.76 (71 ± 5) Inactive (44 ± 1 at 38 μM) Inactive Inactive 11.62 ± 6.02 (58 ± 11) Inactive (24 ± 4 at 38 μM) (43 ± 15 at 38 μM) Abamectin (71751-41-2) 3.25 ± 0.88 (122 ± 8) 8.32 ± 0.25 (98 ± 0) Inactive 31.39 ± 19.00 (132 ± 9) Inactive 6.22 ± 7.17 (73 ± 40) (37 ± 24 at 38 μM) Inactive (41 ± 14 at 38 μM) Doramectin (117704-25-3) 3.07 ± 1.01 (125± 6) 9.96 ± 0.08 (98 ± 1) Inactive 43.23 ± 23.00 (134± 8) 12.47 ± 0.34 (103 ± 2) 2.18 ± 1.68 (95 ± 3) 29.87 ± 4.68 (61 ± 5) Inactive (45 ± 6 at 38 μM) Ivermectin (70288-86-7) 1.18 ± 0.98 (110 ± 31) 38.16 ± 3.90 (90 ± 5) Inactive 243.10 ± 143.39 (131 ± 9) 37.74 ± 17.86 (104 ± 1) 6.03 ± 4.39 (86 ± 4) (45 ± 10 at 38 μM) Inactive (45 ± 8 at 38 μM) Moxidectin (113507-06-5) 9.56 ± 3.08 (105 ± 6) 16.67 ± 7.70 (97 ± 1) Inactive 44.05 ± 0.91 (135 ± 8) 22.28 ± 2.05 (100 ± 2) 5.31 ± 3.45 (98 ± 1) 31.13 ± 2.60 (64 ± 2) 6.21 ± 2.59 (51 ± 10) Inactive Selamectin (220119-17-5) 3.33 ± 0.73 (116 ± 6) 17.90 ± 0.60 (75 ± 5) Inactive 29.04 ± 17.03 (58 ± 5) 21.63 ± 1.30 (54 ± 2) 10.84 ± 3.75 (69 ± 4) (49 ± 13 at 38 μM) Inactive 0.90 ± 0.31 (52 ± 11) 2-Acetyl-1,3- indanedione (1133-72-8) Inactive Inactive Inactive Inactive Inactive (49 ± 13 at 77 μM) Inactive Inactive Inactive Chlorophacinone (3691-35-8) 6.07 ± 0.89 (107 ± 2) 33.60 ± 0.46 (88 ± 4) 45.46 ± 0.58 (62 ± 2) Inactive Inactive 24.26 ± 12.46 (89 ± 11) Inactive Inactive (48 ± 18 at 77 μM) Diphacinone (82-66-6) 11.33 ± 2.61 (107± 8) Inactive (32 ± 8 at 77 μM) Inactive Inactive 35.73 ± 35.51 (67 ± 14) Inactive 60.67 ± 16.90 (29 ± 5) Inactive 1,1- Diphenylacetone (781-35-1) (28 ± 6 at 77 μM) Inactive Inactive Inactive Inactive Inactive Inactive Inactive Inactive Phenindione (83-12-5) Inactive Inactive 525.33 ± 148.42 (54 ± 5) Inactive Inactive 25.01 ± 7.08 (73 ± 11) Inactive (44 ± 7 at 77 μM) Inactive Pindone (83-26-1) (42 ± 8 at 77 μM) Inactive Inactive Inactive Inactive Inactive Inactive Inactive (39 ± 6 at 77 μM) Data are presented as mean ± SEM from three experiments. Compounds with efficacy values smaller than 20% were considered as inactive. Table 3 Selectivity of guggulsterones, avermectins, and 1,3-indandiones against FXR, CAR, LXRα, and PXR in the TR-FRET coactivator antagonist mode assays. Compound Name FXR Coactivator IC50, μM (Efficacy, %) CAR Coactivator IC50, μM (Efficacy, %) LXRα Coactivator IC50, μM (Efficacy, %) (E)-Guggulsterone (40 ± 4 at 38 μM) Inactive Inactive (Z)-Guggulsterone (34 ± 2 at 38 μM) Inactive Inactive Abamectin 2.14 ± 0.40 (116 ± 3) (49 ± 13 at 38 μM) 2.03 ± 1.62 (88 ± 47) Doramectin 1.15 ± 0.11 (98 ± 2) 1.84 ± 0.54 (55 ± 6) 2.47 ± 2.13 (98 ± 5) Ivermectin 0.76 ± 0.09 (124 ± 2) (40 ± 6 at 38 μM) 3.57 ± 2.67 (97 ± 2) Moxidectin 13.89 ± 0.92 (59 ± 1) 8.79 ± 1.35 (67 ± 8) 11.11 ± 5.30 (67 ± 3) Selamectin 0.94 ± 0.11 (117 ± 3) 2.76 ± 0.99 (61 ± 8) 29.13 ± 11.22 (57 ± 5) 2-Acetyl-1,3- indanedione Inactive Inactive Inactive Chlorophacinone 53.79 ± 11.66 (64 ± 12) Inactive 35.25 ± 10.58 (52 ± 15) Diphacinone (34 ± 5 at 77 μM) Inactive Inactive 1, 1 -Diphenylacetone (26 ± 5 at 77 μM) Inactive Inactive Phenindione Inactive Inactive Inactive Pindone Inactive Inactive Inactive Data are presented as mean ± SEM from three experiments. Compounds with efficacy values smaller than 20% were considered as inactive. Table 4 FXR MARCoNI data of known FXR coregulators. MARCoNI Assay ID Target CDCA + DMSO CDCA + Guggulsterone CDCA + Ivermectin CDCA + Chlorophacinone CBP_2055_2077 CBP 50 ± 0 1508 ± 254 184 ± 36 60 ± 13 EP300_2039_2061 p300 74 ± 22 82 ± 18 199 ± 99 88 ± 16 MED1_591_614 DRIP205 13670 ± 209 4221 ± 363 1100 ± 321 1319 ± 32 MED1_632_655 DRIP205 2056 ± 182 1078 ± 228 273 ± 273 50 ± 0 NCOA1_1421_1441 SRC1 9945 ± 208 3398 ± 334 1381 ± 216 680 ± 48 NCOA1_620_643 SRC1 50 ± 0 50 ± 0 50 ± 0 50 ± 0 NCOA1_677_700 SRC1 51347 ± 1459 18599 ± 2630 13835 ± 1149 5001 ± 276 NCOA2_628_651 SRC2 50 ± 0 50 ± 0 63 ± 11 50 ± 0 NCOA2_677_700 SRC2 39425 ± 928 13472 ± 1900 7761 ± 515 3781 ± 192 NCOA3_104_123_N-KKK SRC3 11268 ± 252 157 ± 63 141 ± 63 199 ± 233 NCOA3_673_695 SRC3 27836 ± 971 8847 ± 926 3919 ± 405 1809 ± 223 NCOA3_725_747 SRC3 6033 ± 282 2367 ± 288 626 ± 30 561 ± 30 NCOA6_1479_1501 ASC2 185 ± 44 279 ± 143 50 ± 0 50 ± 0 NCOR1_2251_2273 N-CoR1 50 ± 0 82 ± 39 2346 ± 288 61 ± 12 NCOR1_2376_2398 N-CoR1 50 ± 0 50 ± 0 50 ± 0 50 ± 0 NCOR2_2123_2145 SMRT 344 ± 26 238 ± 19 414 ± 45 261 ± 18 NCOR2_2330_2352 SMRT 75 ± 20 265 ± 65 4133 ± 419 167 ± 22 NR0B2_106_128 SHP 264 ± 40 337 ± 162 64 ± 17 50 ± 0 NR0B2_237_257 SHP 58 ± 6 51 ± 1 222 ± 175 69 ± 23 NRIP1_253_275_C263S RIP140 2060 ± 142 1045 ± 224 334 ± 108 337 ± 43 NRIP1_368_390 RIP140 6435 ± 412 3171 ± 217 1187 ± 139 1070 ± 76 NRIP1_488_510 RIP140 4184 ± 262 1614 ± 217 835 ± 105 524 ± 41 NRIP1_701_723 RIP140 51 ± 1 107 ± 58 194 ± 92 50 ± 0 NRIP1_805_831 RIP140 30916 ± 1213 10978 ± 1112 6980 ± 1034 2428 ± 172 NRIP1_924_946 RIP140 2257 ± 161 1595 ± 196 898 ± 161 657 ± 35 PCAF_178_200 PCAF 69 ± 16 127 ± 14 159 ± 74 116 ± 21 PRGC1_130_155 PGC1α 22951 ± 488 9079 ± 820 2971 ± 247 2025 ± 224 PRGC1_134_154 PGC1α 24330 ± 183 8668 ± 717 3649 ± 475 2425 ± 114 PRGC2_146_166 PGC1β 1106 ± 136 666 ± 111 483 ± 525 183 ± 39 PRGC2_338_358 PGC1β 482 ± 98 305 ± 60 109 ± 72 50 ± 0 Data are presented as mean ± SEM from three experiments. Highlights A subset of Tox21 chemicals was investigated for FXR antagonism In vitro and computational approaches were used to evaluate FXR antagonists Chlorophacinone and ivermectin had distinct patterns in modulating FXR activity This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. REFERENCES Aarts JM Wang S Houtman R van Beuningen RM Westerink WM Van De Waart BJ Rietjens IM Bovee TF Robust array-based coregulator binding assay predicting ERalpha-agonist potency and generating binding profiles reflecting ligand structure Chem Res Toxicol 2013 26 336 346 23383871 Bjursell M Wedin M Admyre T Hermansson M Bottcher G Goransson M Linden D Bamberg K Oscarsson J Bohlooly YM Ageing Fxr deficient mice develop increased energy expenditure, improved glucose control and liver damage resembling NASH PLoS One 2013 8 e64721 23700488 Bramlett KS Yao S Burris TP Correlation of farnesoid X receptor coactivator recruitment and cholesterol 7alpha-hydroxylase gene repression by bile acids Mol Genet Metab 2000 71 609 615 11136553 Cave MC Clair HB Hardesty JE Falkner KC Feng W Clark BJ Sidey J Shi H Aqel BA McClain CJ Prough RA Nuclear receptors and nonalcoholic fatty liver disease Biochim Biophys Acta 2016 1859 1083 1099 26962021 Guan B Li H Yang Z Hoque A Xu X Inhibition of farnesoid X receptor controls esophageal cancer cell growth in vitro and in nude mouse xenografts Cancer 2013 119 1321 1329 23280144 Guo GL Lambert G Negishi M Ward JM Brewer HB Jr. Kliewer SA Gonzalez FJ Sinal CJ Complementary roles of farnesoid X receptor, pregnane X receptor, and constitutive androstane receptor in protection against bile acid toxicity J Biol Chem 2003 278 45062 45071 12923173 Handschin C Meyer UA Regulatory network of lipid-sensing nuclear receptors: roles for CAR, PXR, LXR, and FXR Arch Biochem Biophys 2005 433 387 396 15581595 Houtman R de Leeuw R Rondaij M Melchers D Verwoerd D Ruijtenbeek R Martens JW Neefjes J Michalides R Serine-305 phosphorylation modulates estrogen receptor alpha binding to a coregulator peptide array, with potential application in predicting responses to tamoxifen Mol Cancer Ther 2012 11 805 816 22319200 Hsu CW Zhao J Huang R Hsieh JH Hamm J Chang X Houck K Xia M Quantitative high-throughput profiling of environmental chemicals and drugs that modulate farnesoid X receptor Sci Rep 2014 4 6437 25257666 Huang R Xia M Cho MH Sakamuru S Shinn P Houck KA Dix DJ Judson RS Witt KL Kavlock RJ Tice RR Austin CP Chemical genomics profiling of environmental chemical modulation of human nuclear receptors Environ Health Perspect 2011 119 1142 1148 21543282 Jin L Feng X Rong H Pan Z Inaba Y Qiu L Zheng W Lin S Wang R Wang Z Wang S Liu H Li S Xie W Li Y The antiparasitic drug ivermectin is a novel FXR ligand that regulates metabolism Nat Commun 2013 4 1937 23728580 Kemper JK Regulation of FXR transcriptional activity in health and disease: Emerging roles of FXR cofactors and post-translational modifications Biochim Biophys Acta 2011 1812 842 850 21130162 Kong B Luyendyk JP Tawfik O Guo GL Farnesoid X receptor deficiency induces nonalcoholic steatohepatitis in low-density lipoprotein receptor-knockout mice fed a high-fat diet J Pharmacol Exp Ther 2009 328 116 122 18948497 Koppen A Houtman R Pijnenburg D Jeninga EH Ruijtenbeek R Kalkhoven E Nuclear receptor-coregulator interaction profiling identifies TRIP3 as a novel peroxisome proliferator-activated receptor gamma cofactor Mol Cell Proteomics 2009 8 2212 2226 19596656 Lamers C Schubert-Zsilavecz M Merk D Medicinal chemistry and pharmacological effects of Farnesoid X Receptor (FXR) antagonists Curr Top Med Chem 2014 14 2188 2205 25388533 Lien F Berthier A Bouchaert E Gheeraert C Alexandre J Porez G Prawitt J Dehondt H Ploton M Colin S Lucas A Patrice A Pattou F Diemer H Van Dorsselaer A Rachez C Kamilic J Groen AK Staels B Lefebvre P Metformin interferes with bile acid homeostasis through AMPK-FXR crosstalk J Clin Invest 2014 124 1037 1051 24531544 Lynch C Zhao J Huang R Xiao J Li L Heyward S Xia M Wang H Quantitative high-throughput identification of drugs as modulators of human constitutive androstane receptor Sci Rep 2015 5 10405 25993555 Maran RR Thomas A Roth M Sheng Z Esterly N Pinson D Gao X Zhang Y Ganapathy V Gonzalez FJ Guo GL Farnesoid X receptor deficiency in mice leads to increased intestinal epithelial cell proliferation and tumor development J Pharmacol Exp Ther 2009 328 469 477 18981289 Renga B Mencarelli A D’Amore C Cipriani S D’Auria MV Sepe V Chini MG Monti MC Bifulco G Zampella A Fiorucci S Discovery that theonellasterol a marine sponge sterol is a highly selective FXR antagonist that protects against liver injury in cholestasis PLoS One 2012 7 e30443 22291955 Shukla SJ Sakamuru S Huang R Moeller TA Shinn P Vanleer D Auld DS Austin CP Xia M Identification of clinically used drugs that activate pregnane X receptors Drug Metab Dispos 2011 39 151 159 20966043 Urizar NL Liverman AB Dodds DT Silva FV Ordentlich P Yan Y Gonzalez FJ Heyman RA Mangelsdorf DJ Moore DD A natural product that lowers cholesterol as an antagonist ligand for FXR Science 2002 296 1703 1706 11988537 Wang XX Jiang T Shen Y Caldas Y Miyazaki-Anzai S Santamaria H Urbanek C Solis N Scherzer P Lewis L Gonzalez FJ Adorini L Pruzanski M Kopp JB Verlander JW Levi M Diabetic nephropathy is accelerated by farnesoid X receptor deficiency and inhibited by farnesoid X receptor activation in a type 1 diabetes model Diabetes 2010 59 2916 2927 20699418 Yang F Huang X Yi T Yen Y Moore DD Huang W Spontaneous development of liver tumors in the absence of the bile acid receptor farnesoid X receptor Cancer Res 2007 67 863 867 17283114 Yu DD Lin W Chen T Forman BM Development of time resolved fluorescence resonance energy transfer-based assay for FXR antagonist discovery Bioorg Med Chem 2013 21 4266 4278 23688559 Zhang Y Ge X Heemstra LA Chen WD Xu J Smith JL Ma H Kasim N Edwards PA Novak CM Loss of FXR protects against diet-induced obesity and accelerates liver carcinogenesis in ob/ob mice Mol Endocrinol 2012 26 272 280 22261820
PMC005xxxxxx/PMC5124517.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 9100095 8509 Mol Cell Neurosci Mol. Cell. Neurosci. Molecular and cellular neurosciences 1044-7431 1095-9327 27729244 5124517 10.1016/j.mcn.2016.10.001 NIHMS825871 Article NADPH OXIDASE ISOFORM EXPRESSION IS TEMPORALLY REGULATED AND MAY CONTRIBUTE TO MICROGLIAL/MACROPHAGE POLARIZATION AFTER SPINAL CORD INJURY Bermudez Sara Khayrullina Guzal Zhao Yujia Byrnes Kimberly R. Anatomy, Physiology and Genetics Department, Uniformed Services University, 4301 Jones Bridge Road, Bethesda, MD 20814, USA [email protected]; [email protected]; [email protected] * Correspondence: Kimberly R. Byrnes, Department of Anatomy, Physiology and Genetics Room C2115 4301 Jones Bridge Road, Bethesda, MD 20814, USA 1-301-295-3217 (telephone) 1-301-295-1786 (FAX) [email protected] 1 11 2016 10 10 2016 12 2016 01 12 2017 77 5364 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Spinal cord injury (SCI) results in both acute and chronic inflammation, as a result of activation of microglia, invasion of macrophages and activation of the NADPH oxidase (NOX) enzyme. The NOX enzyme is a primary source of reactive oxygen species (ROS) and is expressed by microglia and macrophages after SCI. These cells can assume either a pro- (M1) or anti-inflammatory (M2) polarization phenotype and contribute to tissue response to SCI. However, the contribution of NOX expression and ROS production to this polarization and vice versa is currently undefined. We therefore investigated the impact of SCI on NOX expression and microglial/macrophage polarization over time in a mouse model of contusion injury. Adult C57Bl/6 mice were exposed to a moderate T9 contusion SCI and tissue was assessed at acute, sub-acute and chronic time points for NOX isoform expression and co-expression with M1 and M2 microglia/macrophage polarization markers. Two NOX isoforms were increased after injury and were associated with both M1 and M2 markers, with an M1 preference for NOX2 acutely and NOX4 chronically. M2 cells were primarily found at acute time points only; the peak of NOX2 expression was associated with the decline in M2 polarization. In vitro, NOX2 inhibition shifted microglial polarization toward the M2 phenotype. These results now show that microglial/macrophage expression of NOX isoforms is independent of polarization state, but that NOX activity can influence subsequent polarization. These data can contribute to the therapeutic targeting of NOX as a therapy for SCI. Microglia inflammation spinal cord injury oxidative stress polarization Introduction Spinal cord injury (SCI) results in both acute (hours) and chronic (days to months) inflammation. This inflammation includes activation of microglia, invasion of macrophages and up-regulation of the NADPH oxidase (NOX) enzyme. This inflammatory process plays an important role in the secondary tissue damage and cell death that follows the initial mechanical insult. This contribution to the pathological outcome of the injury makes it an important therapeutic target. In a comprehensive study, a histopathological post-mortem assessment of injured human spinal cords showed an increased expression of NOX2 associated with a transient neutrophil population and activated microglia/macrophages with a peak expression at 3 days after injury (Fleming et al., 2006). SCI also results in acute and chronic up-regulation of NOX components in both rats and mice (Byrnes et al., 2006, Byrnes et al., 2011, Pajoohesh-Ganji et al., 2012). We have previously shown that NOX, particularly the NOX2 and 4 isoforms, are up-regulated after central nervous system (CNS) injury in a number of cells, including microglia and macrophages in a rats (Cooney et al., 2013, Cooney et al., 2014). The NOX family of proteins is a primary source of reactive oxygen species (ROS) and oxidative stress (Xiong et al., 2007). There are several isoforms of this enzyme, NOX1, NOX2, NOX3, NOX4, NOX5, DUOX1 and DUOX2, each with slightly different enzyme components and oxidative products. NOX2, for example, is composed of four cytosolic (gp40PHOX, p47PHOX, p67PHOX and GTP-binding protein p21-Rac1) and two membrane subunits (gp91PHOX and p22PHOX) (Cross and Segal, 2004). Activation of NOX typically involves two steps: up-regulation of expression of the individual protein components and PKC- or MAPK-mediated phosphorylation of cytosolic components (such as p47PHOX in the NOX2 isoform). Once phosphorylated, the cytosolic subunits translocate to the plasma membrane, where they assemble with the membrane subunits to form the active enzyme (Bey et al., 2004, Choi et al., 2005, El Benna et al., 1996, Zhao et al., 2005). Unlike NOX2, NOX4 is constitutively active, although it can be influenced by a number of factors including chemical changes. Active NOX2 produces O2-, which can be converted into other superoxide-derived oxidants such as hydrogen peroxide (H2O2), hydroxyl radicals and peroxynitrite (Patel et al., 2005), while NOX4 produces H2O2 directly (Takac et al., 2011). These ROS can have severe cytotoxic effects (Block et al., 2007, Qin et al., 2004) as well as facilitate pro-inflammatory pathways by activating MAPK and NFκB signaling, ultimately inducing transcription of pro-inflammatory cytokines and other inflammatory mediators such as tumor necrosis factor (TNF)α, inducible nitric oxide synthase (iNOS), and monocyte chemotractant protein (MCP)-1 (Aldskogius and Kozlova, 1998). ROS can also produce a positive feedback loop by inducing transcription and/or activation of NOX components, further increasing NOX activity (Brandes et al., 2001, El Benna et al., 1996, Pawate et al., 2004). In injured human spinal cords, activated microglia and macrophages were present at 1 day post injury, becoming the predominant inflammatory cells at and beyond 5 days post-injury; this response is sustained weeks to months (Fleming et al., 2006). Despite differences in lesion evolution and cavity formation in rats and mice, similarly to human injury, macrophages from the periphery and activated microglia appear in the spinal cord between 12 and 24 hours post-injury, peaking between 4 and 8 days post-injury (Carlson et al., 1998, Pajoohesh-Ganji et al., 2012, Popovich et al., 1997, Sroga et al., 2003). In a comprehensive study Kigerl et al. demonstrated that this microglial/macrophage response is primarily composed of 2 different cellular phenotypes: M1 and M2 (Kigerl et al., 2009). Classically activated, or M1, microglia/macrophages, express a number of identifiable cell surface markers, such as CD86 and CD16/32, and secrete pro-inflammatory molecules such as TNFα, interleukin (IL) 1β, ROS and nitrogen intermediates (Aloisi, 2001). Alternatively activated, or M2, microglia/macrophages have an anti-inflammatory phenotype and are associated with CD206/mannose receptor and arginase-1 expression and secretion of anti-inflammatory cytokines such as IL4 and transforming growth factor (TGF)-β (Kigerl et al., 2009). Acutely after injury, both M1 and M2 phenotypes are equally present in the damaged spinal cord, while, at later time points (i.e., beyond 2 weeks in a rodent SCI model), the M1 phenotype is dominant and M2 markers are often below levels of detection (Kigerl et al., 2009). It is currently unclear what factors in the injured spinal cord control the shifts in microglial and macrophage polarization. However, in 2011, Choi et al. (Choi et al., 2011) found that knockout of gp91PHOX or p47PHOX or administration of the non-specific NOX inhibitor apocynin could increase M2 marker expression in microglia in an in vivo inflammation model. Further, our previous work demonstrated that acute inhibition of the NOX2 enzyme with the specific inhibitor gp91ds-tat resulted in an elevation in the number of cells expressing the M2 marker CD206 (Khayrullina et al., 2015). The goal of this work was therefore to further explore this phenomenon and identify the NOX enzyme isoform expression profile after SCI and determine if there was a correlation between this expression profile and microglial/macrophage polarization. We now demonstrate that the expression of NOX2 and NOX4 enzymes is both temporally and polarization related. Further, we show that NOX enzyme inhibition in microglia can alter polarization status, although polarization does not alter NOX expression. Material and Methods Animal Handling and Surgical Methods Adult male C57Bl6 mice (20-25g, Taconic Farms, Derwood, MD) were used for all experiments. Mice were group housed and received food and water ad libitum with a 12:12 hour light cycle. A total of 72 male mice were used for this study; 3 injured mice and 1 sham injured mouse were removed from the study due to post-surgical complications or inadequate injury. All experiments complied fully with the principles set forth in the “Guide for the Care and Use of Laboratory Animals” and were approved by the Uniformed Services University IACUC. Moderate spinal cord contusion injury was performed using the Infinite Horizons Impactor (50 kdyne; Precision Systems and Instrumentation, Fairfax Station, VA) positioned over the exposed spinal cord at vertebral level T-9 while animals were anesthetized with isoflurane (4% induction, 2% maintenance). Sham injured mice underwent the same experimental procedures, but received a laminectomy only. Naïve mice underwent no surgery. Animals were allowed to recover on heating pads, after which they were returned to home cages (group housing, 4 mice/cage with cell-sorb bedding) and received acetaminophen (200mg/kg) in drinking water for 72 hours post-injury. Manual bladder expression was performed daily until normal bladder expression returned. At 1 (n = 5 injured, 5 sham), 4 (n = 4 injured, 4 sham), 7 (n = 5 injured, 5 sham), 28 (n = 3 injured, 4 sham), or 60 (n = 5 injured, 5 sham) days after surgery or no surgery (naïve; n = 4), mice were anesthetized (Euthasol, 0.22ml/kg, I.P.) and intracardially perfused with 50 ml of 0.9% saline followed by 100 ml of 10% buffered formalin. In injured or sham/naïve mice, a 5mm section of the spinal cord centered at the lesion epicenter was dissected, post-fixed in 10% buffered formalin overnight and cryoprotected in 30% sucrose for 48 hours. Tissue was cut into serial 20 μm thick transverse sections and collected at every 60μm (Supplementary Figure 1). At the above-mentioned time points, an additional 18 mice (naïve n = 4; 1 day n = 3; 4 days n = 4; 7 days n = 3; 28 days n = 4) were euthanized and tissue sample protein was obtained from a 5 mm section of the spinal cord centered at the lesion epicenter was dissected and immediately frozen on dry ice for western blotting. Immunochemistry Standard single or double fluorescent immunohistochemistry (IHC) and immunocytochemistry (ICC) was performed as described previously (Cooney et al., 2013). Primary antibodies were used as described in the Table 1. Appropriate secondary antibodies linked to AlexaFluor dyes (1:2000; Invitrogen, Carlsbad, CA) were incubated with tissue sections (serial 20μm thick transverse sections 60μm apart, supplementary Fig. 1) or fixed cells for 1 hour at room temperature. Slides were coverslipped using mounting media containing DAPI to counterstain for nuclei (Vector Labs, Burlingame, CA). Round coverslips were mounted onto slides using mounting media containing DAPI (Vector Labs). To ensure accurate and specific staining, negative controls were used in which the primary antibody was not applied, and only staining that labeled cells that were labeled with nuclei and had expected labeling patterns (i.e., classic microglia morphology) was confirmed as positive labeling. It should be noted that NOX specific antibodies from a number of sources were tested, as NOX antibodies have a history of lacking specificity and fidelity. Each antibody was tested in conjunction with cell specific antibodies and with western blotting to confirm that each resulted in appropriate, expected staining and/or expected band sizes in western blots (NOX2 - supplementary figure 2; NOX4 data not shown, but consistent with the testing in our previous work (Cooney et al., 2013)). Imaging and quantification ICC fluorescence was detected and photographed in at least 5 randomly chosen regions (20X) using an Olympus DP72 microscope with Olympus cellSens microscopy software (Olympus). IHC fluorescence was detected and photographed using an Olympus DP72 microscope with Olympus cellSens microscopy software (Olympus, Center Valley, PA) or NanoZoomer Digital Pathology system (Hamamatsu Photonics, K.K., Japan). For 8OHdG, fluorescence was quantified as previously described using pixel density measurement in Scion Image (Donnelly et al., 2009). For all other immunohistochemistry, semi-quantification was performed manually on 3 to 5 equally spaced sections per animal spanning a 1.7mm section encompassing the lesion center (Supplementary Figure 1) focused on the dorsal columns and surrounding gray matter at a 20X magnification, as this was the area most affected by the dorsal contusion. Cells were evaluated through the presence of a labeled nucleus and expected cellular morphology by two investigators blinded to group. The score system was used as follows: 0=no labeled cells observed, 1=<5, 2=5-10, 3=10-50, 4=>50 labeled cells observed (Supplementary Figure 3). Score for each section was then averaged with all other sections for each animal, and mean per group represents average for each animal in that group. This scoring system was chosen as standard cell counting was found to underestimate cell number due to overlapped cells (data not shown). In addition, given the injured nature of the tissue and common artifactual fluorescence, pixel density analysis was unable to specifically identify microglial/macrophage labeling or to detect double-labeling of cells. In all analyses, no significant difference was noted between sham and naïve data. Therefore, in all data presentation and analysis, sham and naïve data are combined and, in graphs, indicated as day 0. Western blot Analysis Protein from tissue samples was isolated in RIPA Buffer (Pierce, Rockford, IL) containing protease inhibitor (1%, Halt Protease Inhibitor Single-Use Cocktail, Thermo Scientific). The in-vitro sample protein was obtained 24 hours after treatment with LPS, gp91ds-tat or scrambled-tat, cells were scraped and lysed in Cell Lysis Buffer (Cell Signaling Technology, Danvers, MA) containing protease inhibitor (1%, Halt Protease Inhibitor Single-Use Cocktail, Thermo Scientific). Twenty-five μg of sample protein was run in a Mini-PROTEAN® TGX™ Precast Gel (Bio-Rad, Hercules, CA) and transferred to a Trans-Blot® Turbo™ (Bio-Rad) nitrocellulose membrane. Primary antibodies were probed overnight at 4°C as described in the table 1. Immune complexes were detected with appropriate secondary antibodies and chemiluminescence reagents (Pierce, Rockford, IL). GAPDH (0.5μg/ml; Millipore, Temecula, CA) or β-actin (0.5 ug/ml; Abcam) was used as a control for gel loading and protein transfer. NIH Image J was used to assess pixel density of resultant blots for quantitation. In Vitro Analysis The BV2 microglial cell line (a generous gift from Dr. Carol Colton) was cultured and replated to wells at passage 14 - 19. Cells were incubated at 37°C with 5% CO2 in Dulbecco's modified Eagle media (Gibco, Carlsbad, CA) with 10% fetal calf serum (Hyclone, Logan, UT), 1% L-glutamine (Gibco), 1% sodium pyruvate (Gibco), and 1% Pen/Strep (Fisher, Pittsburgh, PA). Cells were used 24 hours after plating for experimentation. BV2 cells were treated with the toll-like receptor 4 (TLR4) agonist lipopolysaccharide (LPS; 100ng/ml, Sigma, St. Louis, MO) or vehicle media for 24 hours. The NOX2 inhibitor gp91ds-tat peptide (50 μM, AnaSpec, Fremont, CA), scrambled ds-tat (50μM, AnaSpec) or recombinant rat IL-4 (7.5 μM, R&D systems) was applied one hour after LPS or control stimulation. All drugs were prepared and stored according to the manufacturer's guidelines. At 24 hours after treatment cells were fixed with 4% paraformaldehyde on round coverslips or protein was isolated. NOX isoforms, microglial polarization markers and oxidative stress markers were examined. Nitric Oxide Assay BV2 microglia were exposed to LPS for 1 hour prior to addition of gp91ds-tat (50μM) or scrambled ds-tat and incubated at 37°C and 5% CO2 for 24 hours. NO• production was assayed using the Griess Reagent Assay kit (Invitrogen) and absorption assessed at 540nm, according to the manufacturer's instructions. Statistics Power analysis was completed based on our previously published studies (Cooney et al., 2013, Cooney et al., 2014) and pilot work indicated that an n of 4/group for immunohistochemical analysis and an n of 3/group for western blotting would be sufficient to demonstrate a significant difference between groups with 80% power. Post-hoc analysis demonstrated that an n of 4 was sufficient to obtain this power in all studies; however, due to post-injury mortality in the 28 day injured group, the final n was 3. Analysis of this group did show that, while underpowered, no significant difference would have been observed without an extraordinarily large n (n = 10 – 15/group) for immunohistochemical analysis, and the result would likely be biologically insignificant. Therefore, 28 day data remains in the manuscript. All assays were carried out by investigators blinded to subject group. Quantitative data are presented as mean +/− standard error of the mean. Semi-quantitative score data were analyzed using the Kruskal-Wallis non-parametric test followed by Dunn's post-test. Quantitative in vivo and in vitro data were analyzed using one-way ANOVA with Dunnett's Multiple Comparison's post-test. All statistical tests were performed using the GraphPad Prism Program, Version 6.03 for Windows (GraphPad Software, San Diego, CA). A p value < 0.05 was considered statistically significant. Results SCI leads to a significant acute and chronic increase in oxidative stress The release of ROS can lead to oxidative damage to DNA, which can be detected using the 8OHdG antibody, and production of further oxidative compounds, such as peroxynitrate. In turn, peroxynitrate can lead to the nitrosylation of proteins, which can be detected using the 3NT antibody. We therefore investigated oxidative stress in the spinal cord at acute and chronic time points after SCI or in naïve tissue using immunohistochemistry and western blotting for 8OHdG and 3NT (Fig. 1A, B). A marked increase in nuclear staining of 8OHdG, particularly in large motor neurons within the gray matter, was observed in injured tissue in comparison to naïve tissue. This expression was most prevalent at 1 and 28 days post-injury. Immunolabeling was quantified as pixel density using stain intensity as a threshold; quantification confirmed a significant increase in immunofluorescence over naïve tissue at 24 hours and 28 days post-injury (p<0.05). Elevated 3NT was also observed in injured tissue from 24 hours to 60 days post-injury (Fig. 1E-G). Quantification of nitrosylated proteins in a western blot showed that this was significantly elevated above naïve levels by 1 day through 28 days peaking at 4 days post-injury (p<0.001). A number of enzymes are known to contribute to post-injury oxidative stress. In order to begin to understand whether NOX may play a role, we investigated if activation of NOX was observed at the same time point of elevated oxidative stress. Immunoblotting for the phosphorylated p47PHOX subunit showed a significant increase at 4 days post-injury in comparison to sham/naïve tissue (Fig. 1C-D). Moderate SCI results in changes in NOX isoform expression and microglia/macrophage polarization We next aimed to determine if NOX isotype expression was also altered following SCI. Immunolabeled NOX isotypes were semi-quantified using a 5 point scale based on the number of positive cells and show clear changes in NOX2 and NOX4 isotype expression after injury (Fig. 2). NOX2 expression showed a significant increase at 1 and 4 days, returning to levels that trended toward higher but were not significantly different than sham/naïve levels by 7 days after injury, with no further change. NOX4 also showed an initial increase, significantly elevating above sham by 4 days, but returned to sham levels by 28 days after SCI. A trend toward significance was observed at 1 and 7 days post-injury, but did not reach significance with the non-parametric Kruskal Wallis test. To further support the immunohistochemistry results, expression of NOX2 was confirmed at 4 days post-injury using immunoblotting techniques (supplementary Fig. 2). Using the same 5 point scale, microglia/macrophage polarization markers were evaluated and semi-quantified as well (Fig. 2). CD86 was used as a marker of M1 polarized macrophage/microglia. Its expression increased significantly after injury, with a slight reduction in expression at 7 days. Expression elevated again at 28 days and continued increasing during the rest of the study. CD206 was used as a marker of M2 polarized macrophage/microglia. This marker showed an increase in expression acutely but at 7 days post-injury expression levels returned to sham/naïve levels where it remained for the remainder of the study. Additional M1 and M2 markers were examined using immunohistochemistry and immunoblotting techniques to confirm these polarization trends (supplementary Fig. 2). Acute NOX2 expression is associated with M1 and M2 microglia/macrophages In order to assess any temporal or phenotypic dependent expression of NOX isoforms in macrophages/microglia, double immunostaning of polarization markers and NOX isoforms was performed. Single and double-labeled cells were manually semi-quantified spanning a representative area of the tissue always including the white and grey matter within the lesion epicenter and perilesional region (supplementary Fig. 1). In addition to double labeling, temporal changes in expression of each cell population were analyzed. Immunoreactive cells for each antibody were taken as a population and the proportion of this population expressing each marker at all different time points is reported. The overall pattern of staining suggests that at acute (24 hours to 4 days) time points, both M1 and M2 polarized cells were associated with NOX2, which was confirmed with confocal microscopy at higher magnifications (supplementary Fig. 4). NOX2 co-labeling with the M1 marker CD86 was slightly elevated over sham by 24 hours post-injury through 7 days post-injury, although this did not reach statistical significance (Fig. 3). At this 7 day time point, all of the cells expressing NOX2 were CD86+, but only about 75% of the CD86+ cell population expressed NOX2. Interestingly, at these acute time points, while 100% of the NOX2 positive cells were CD86+, 10 – 20% were also positive for the M2 marker CD206 (Fig. 4). Investigation of triple labeling for NOX2 and the M1 and M2 markers showed that at this time point, a subset of cells were positive for both M1 and M2 markers, suggesting an intermediate polarization state (Fig. 5). At 28 days post-injury, co-labeling decreased to resemble sham/naïve but returned to increased levels by 60 days post-injury. At chronic time points, about 30% of the M1 polarized cells expressed NOX2. CD206 was used as an M2 polarized macrophage/microglia marker. Standard double immunostaining was performed with CD206 and an antibody against NOX2 in tissue from sham-injured and all examined time points post-injury (Fig. 5). Although CD206 positive cells can be observed in tissue from sham/naïve and all the examined time points, co-labeling with NOX2 was most prominent at 4 and 7 days after injury with a peak at 4 days. Semi-quantification indicated that double-labeled population increased significantly from 1 day to 4 days post-injury; by 7 days the double labeled cell population diminished to sham injured levels. Up to 4 days post injury, the percent of CD206 positive cells expressing NOX2 had shifted from less than 10% to 25-35% while of all the NOX2+ cells, 40-50% were M2 polarized cells. This involved an increase from less than 10% at 1 day post injury to 40 – 50% by day 4. Both of these proportions returned to sham injured levels by 7 days, the time point at which M2 polarization decreased in the spinal cord. NOX4 isoform is associated with the M1 and M2 polarized phenotype Standard double immunostaining was also performed with an antibody against NOX4 and the M1 marker CD86 (Fig. 6). As with NOX2, single and double-labeled cells were manually semi-quantified spanning a representative area of the tissue always including the white and grey matter within the lesion epicenter and perilesional region. The overall pattern of staining suggests that NOX4 was observed in M2 polarized cells acutely after injury and in M1 polarized cells at more chronic time points. The number of M1 and NOX4 double labeled cells was increased slightly over sham, although without reaching statistical significance, during the study time frame from 1 to 28 days post-injury (Fig. 6), with a slight dip to sham levels observed at 7 days. However, by 60 days, the amount of M1 and NOX4 double labeled cells was significantly elevated over sham. The percent of M1 polarized cells expressing NOX4 changed from approximately 10% at 1 day post-injury to approximately 60% at 60 days post-injury. In a similar manner, of all the cells found to be NOX4+ in the tissue at the examined time points, approximately 10% at 1 day post-injury were M1 polarized cells. This proportion increased to 60% by the end of the study time frame. NOX4 and CD206 double staining was also evaluated (Fig. 7). At 1 and 4 days post injury, a noticeable increase in co-labeling was observed and found to be significant in the semi-quantification at both time points. During these acute time points, 40-50% of cells expressing NOX4 were M2 polarized cells. Of the CD206+ cells, approximately 50% expressed NOX4+ at the same time points. By 7 days, the double-labeled cell population declined to sham injured levels but climbed again at chronic time points and remained above sham/injured levels although not reaching significance as at acute time points. During these chronic time points the proportion of M2 cells expressing NOX4 was between 30 and 40% while of the NOX4+ cells present in the tissue at these time points, 20-25% were M2 polarized cells. In vitro, NOX2 activity is associated with changes in polarization and NOX2 inhibition alters polarization marker expression In order to understand if the observed alterations in NOX isoform expression and polarization changes are related, we moved to a cell culture model. First, the BV2 microglial cell line was cultured for 1 – 24 hours to the well characterized M1 polarizing agent LPS (Orihuela et al., 2016), and the levels of M1 and M2 markers were assessed. Phosphorylation of p47PHOX was found to be significantly elevated by 6 hours post-injury, indicating activation of the enzyme at that time point (Fig. 8A, B). At the same time, NO production, a marker for M1 polarization, was significantly elevated and continued to rise throughout the remaining time points (Fig. 8C). Conversely, the M2 marker CD206 was significantly decreased in LPS treated BV2 cultures by 16 hours, following the elevation of phosphorylated-p47PHOX (Fig. 8D). Next, to test the hypothesis that NOX2 contributes to polarization, BV2 cells were cultured following exposure to known polarizing agents (LPS, IL4) and to a NOX2 antagonist, gp91ds-tat and polarization markers were evaluated. NOX2 and NOX4 expression were not altered by LPS or IL4 administration nor by gp91ds-tat peptide administration (data not shown), suggesting that NOX isoform expression is not altered by microglial polarization. However, polarization markers were altered by administration of the NOX2 inhibitor. Resting microglia were found to be iNOS (M1) negative, but showed a moderate amount of baseline M2 marker (CD206) staining (Fig. 9). Treatment with LPS increased iNOS expression (Fig. 9A), although it did not lead to a significant reduction in CD206 protein expression as measured by western blotting. However, administration of the gp91ds-tat peptide 1 hour after LPS administration significantly elevated CD206 expression above LPS levels. iNOS expression remained elevated, as evidenced by both immunocytochemistry and western blotting. Discussion A number of studies have found evidence of oxidative damage in the spinal cord that can be detected within hours of injury, peaks at around 7 days, and remains present for weeks to months after (for review, see (Jia et al., 2012)). We have also previously shown that components of the NOX enzyme are elevated at both acute and chronic time points after SCI (Byrnes et al., 2006, Cooney et al., 2014). We now expand upon those findings and demonstrate that there is evidence of oxidative stress in the injured spinal cord from 1 to 28 days post-injury, and that two NOX isoforms, NOX2 and NOX4, are elevated in expression after SCI and are expressed in both M1 and M2 polarized microglia/macrophages. The overall pattern of staining demonstrates that at acute time points, M1polarized cells are associated with NOX2 whereas at later time points, M1 polarized cells seem most associated with NOX4. On the other hand, the transient M2 population is associated with both NOX isoforms. This association appears to change primarily due to the shift in the M1/M2 ratio over time, although our in vitro work suggests that activity of the NOX2 isoform may bear an influence on these observations. This sheds light into our previous in vivo work where acute NOX2 inhibition appears to limit the shift towards M1 polarization after SCI. NOX expression and activity has been noted to be elevated as early as 24 hours post-injury (Byrnes et al., 2006) and remains upregulated for months after injury (Byrnes et al., 2011). NOX has previously been shown to contribute to oxidative stress following injury, and we now show that 2 different markers of oxidative stress, DNA oxidation (8OHDG immunohistochemistry) and protein nitrosylation (3NT western blot and immunohistochemistry) are elevated following SCI. Both DNA oxidation and protein nitrosylation were found to be significantly elevated at 1 day post-injury. While DNA oxidation was found to be significantly elevated at 1 and 28 days, 3NT showed significant elevation at all time points after injury, peaking at 4 days and remaining significantly elevated through 28 days. Interestingly, 3NT peak elevation was found at the same point as peak in NOX2 expression and observation of activation of the NOX2 enzyme, as measured by phosphorylation of p47PHOX, suggesting that we are observing not just an increase in protein expression, but activity that is accompanied by an elevation in downstream oxidative events. The NOX2 isoform is expressed in microglia, macrophages, neurons and astrocytes (Cooney et al., 2013, Cooney et al., 2014), and this work now shows that the NOX2 expression is not restricted to a specific polarization phenotype, with up to 70% of M1 macrophages/microglia and up to 40% of M2 macrophages/microglia expressing NOX2. Following SCI in a rat model, we have shown that NOX2 peaks at 7 days post-injury, but is elevated above baseline at both 24 hours and 28 days (Cooney et al., 2014). Our current work is in agreement with this, although at a slightly earlier time frame (1 day and 4 day peaks) in the mouse model. It is interesting to note that by 7 days post-injury, 100% of the NOX2 positive cells were CD86 positive, despite the finding that NOX2 is expressed by a number of cells in the CNS. The NOX4 isoform, on the other hand, has been found to peak in microglia at approximately 24 hours in the rat model, decreasing in expression through 28 days (Cooney et al., 2014); in contrast, the mouse model demonstrates an elongated NOX4 expression, which remained elevated through 7 days before dropping at 28 days. NOX4 is also expressed in microglia, macrophages, neurons and astrocytes (Cooney et al., 2013, Cooney et al., 2014). This enzyme produces H2O2 and has been associated with pain after peripheral nerve injury (Kallenborn-Gerhardt et al., 2012) and glutamate release from microglia (Harrigan et al., 2008). It is considered to have the widest cellular distribution of the NOX's in the body (Altenhofer et al., 2012). Like NOX2, NOX4 expression is not restricted to a specific polarization phenotype, showing expression in up to 70% of M1 macrophages/microglia and up to 50% of M2 macrophages/microglia. Interestingly, this elevated expression in M1 macrophages/microglia was only observed at the final time point (60 days post-injury). Prior to that, NOX4 expression was markedly low in both macrophages and microglia, reaching at most 40-50%. This suggests that NOX4 is primarily expressed by other cells in the CNS, but may play a role in chronic inflammation. It has been shown that M2 polarized microglia and macrophages decrease in lesion areas in the CNS over time. In the spinal cord, Kigerl et al. performed a comprehensive investigation of M1/M2 polarization markers over time after injury, and showed that M2 markers are elevated during the first 2 weeks after injury and swiftly decline after that (Kigerl et al., 2009). Similar to our work, they found equivalent M1 and M2 staining within the first few days after SCI, often within the same cells. After a moderate traumatic brain injury in rats and mice, M2 markers also peaked at 5 days and were reduced dramatically thereafter (Turtzo et al., 2014, Wang et al., 2013), while M1 markers remained consistent throughout the post-injury time period (Bedi et al., 2013, Turtzo et al., 2014), which is similar to our findings in the injured mouse spinal cord. We now provide evidence that this M1/M2 staining may be present in the same cell (Fig. 4), demonstrating that microglia and macrophages exist in intermediate states that express both M1 and M2 markers along with NOX2 in the injured mouse spinal cord. In agreement with our findings, Kumar et al. (Kumar et al., 2015) also found a robust increase of NOX2 colocalization with Iba1-positive macrophage/microglia at 7 days post-injury in a moderate controlled cortical impact injury model in mice. Furthermore, this gp91PHOX expression was significantly co-localized with microglia expressing only M1 or both M1 and M2 markers but negligibly expressed in M2-like microglia. The significance of this intermediate polarization state is currently unclear. Our current in vitro data suggests that NOX activity influences polarization, as inhibition of NOX2 led to a significant increase in CD206 expression. In vivo we have previously observed that acute NOX2 inhibition limits the decline of CD206+ population at 7 days after SCI (Khayrullina et al., 2015). Our current in vivo data also seems to support this influence, demonstrating that the peak of NOX2 expression was followed by the reduction in CD206 expressing cells at 7 days post-injury. It is important to note that, despite the reduction in NOX2 expression at 7 days, CD206/M2 marker expression did not increase, as our in vitro data would suggest. In fact, this is the situation in the periphery, with tissue wounds or peripheral nerve injury, in which M2 markers are elevated in later stages of healing (Chen et al., 2015, Ferrante and Leibovich, 2012). While speculative at this point, we now propose that microglial polarization reflects the redox state of the lesion microenviroment. It is possible that despite low expression, NOX2 may still be active and contributing to oxidative stress and suppression of the M2 phenotype, or the NOX4 enzyme may contribute to this suppression as well. The observations of elevated oxidative stress markers at chronic time points support this proposal, but future research is needed to further explore this phenomenon. In addition, our in vitro p47PHOX phosphorylation data also provides a chronological insight into the interaction between NOX2 and M2 polarization, demonstrating that the peak of phosphorylation of p47PHOX was associated with the down-regulation of M2 expressing cells. Interestingly, our in vitro data did not show a sustained phosphorylation of p47PHOX, despite sustained changes in microglial polarization. This is likely due to a loss of stimulatory action of LPS due to degradation over time, unlike what will happen in the injured spinal cord with continual pro-inflammatory signals. However, despite the loss of obvious NOX2 activity, sustained changes in polarization were observed, suggesting that either NOX2 activity initiates a series of downstream events leading to sustained protein changes. Alternatively, additional signal transduction pathways may be at work to contribute to these sustained changes. For example, Kumar et al suggest that NOX2 activity acts like a switch driving M2 to an M1 polarization state after traumatic brain injury and proposes an upregulation of IL-4Rα expression as a mechanism for this phenomenon, although this upregulation was only found on infiltrating macrophages (Kumar et al.). Further, previous work has shown that NOX inhibition has a number of anti-inflammatory effects, which may be related to this shift in polarization. For example, in vitro, reduction of NOX activity can ameliorate microglial pro-inflammatory activity (Byrnes et al., 2006, Cheret et al., 2008, Peng et al., 2009) and reduce neuronal cell death (Gao et al., 2002). Reduction or knockout of the p47PHOX component also impairs cytokine and nitric oxide (NO) production in astrocytes and microglia stimulated with LPS and/or interferon (IFN) γ (Pawate et al., 2004). More specifically, diphenyleneiodonium (DPI), a non-specific NOX inhibitor, blocks NFκB activation in microglia, with subsequent reductions in iNOS and pro-inflammatory cytokine production (Min et al., 2004). Similarly, NFκB activation, pro-inflammatory cytokine production and inflammation were all reduced after SCI in mice following apocynin administration (Impellizzeri et al., 2011). Finally, in an in vivo model of inflammation utilizing LPS or Aβ injection, Choi et al. found that knockout of gp91PHOX or p47PHOX or administration of apocynin could increase M2 marker expression in microglia (Choi et al., 2011). In our hands the administration of the NOX2 inhibitor gp91ds-tat limits the decline of CD206+ after SCI (Khayrullina et al., 2015). These data are in agreement with our current in vitro and in vivo findings and suggest that, in the SCI model, the activity of NOX2 or 4 may be influencing the shift from an equal M1/M2 ratio towards a higher M1 expression rather than polarized microglia associated with any of the NOX isotypes. Conclusion In summary, our current data therefore provide a descriptive analysis of NOX isoform expression in combination with the post-injury M1/M2 ratio in the injured spinal cord. These data demonstrate that NOX expression, particularly NOX2 and NOX4, is increased after mouse SCI. Further, while both M1 and M2 microglia/macrophages express NOX isoforms, there is an influence of NOX on polarization. As polarization plays a significant role in outcome (Yao et al., 2014), these data support the idea that modulation of NOX activity after SCI may improve recovery. Supplementary Material 1 2 3 4 Acknowledgements The authors thank Fiona Brabazon for editorial contributions. This work was funded by the NINDS/NIH (Grant number 1R01NS073667-01A1). Abbreviations NOX NADPH oxidase LPS lipopolysaccharide DPI diphenyleneiodinium Figure 1 Oxidative stress markers significantly increase after injury 8OHdG immunohistochemistry was used as a marker of oxidative stress in naïve and injured tissue (A). Size bar = 2mm. Expression was measured by pixel density of immunolabeling using stain intensity as threshold. The staining showed a significant increase in expression over naïve tissue (n = 4) at 24 hours (n =5) and 28 days (n = 3) post-injury, with no significant difference at 4 days (n = 4), 7 days (n = 5) or 60 days (n = 5) post-injury (B). Phosphorylated p47PHOX was assessed by western blot in sham and injured tissue 4 days post-injury (n = 4/group). Quantification shows a significant increase in injured tissue over naïve (C). Representative blot is shown in D. Protein nitrosylation (3NT) was assessed by western blot in naïve (n = 3) and injured tissue at 24 hours (n = 3), 4 days (n = 3), 7 days (n = 3) and 28 days (n = 4) post-injury (E). Quantification shows a significant increase of 3NT in injured tissue over naïve at all time points, peaking at 4 days (F). 3NT was also assessed through immunohistochemistry and showed an elevated immunoreactivity in injured tissue from day 1 to day 60, particularly in grey matter (G). Size bar = 50μm. Bars represent mean +/- SEM. p<0.05, p<0.01, p<0.001, using One-Way ANOVA with Dunnett's Multiple Comparison post-test. Figure 2 Moderate SCI changes the expression of NOX isoforms and polarized microglia markers in the spinal cord NOX isoforms and M1 (CD86) and M2 (CD206) markers were assessed on a 5 point scale in naïve/sham tissue (n = 11) or at 1 (n = 5), 4 (n = 4), 7 (n = 5), 28 (n = 3) or 60 (n = 5) days after a moderate contusion injury. NOX2+ cells showed a significant increase at 1 and 4 days post-injury, after which expression declined, but remained elevated through 60 days post-injury over sham-injured tissue (A). The NOX4+ cell population showed a similar pattern, although significance was only observed at 4 days post-injury (B). CD86+ cells in the spinal cord increased significantly acutely and continued rising at chronic time points (C). CD206+ cells were increased at 4 days post-injury and decreased by 7 days but remained elevated over sham/injured levels at chronic time points (D). Bars represent mean score +/− SEM. p<0.05, p<0.01, p<0.001, using One-Way ANOVA with Dunnett's Multiple Comparison post-test. Figure 3 M1 polarized macrophages/microglia express NOX2 after SCI Macrophages/Microglia were labeled with CD86 (red) and immunostained with an antibody against NOX2 (green) in sham-injured tissue and at all examined time points post-injury (A). Co-labeled cells are indicated with arrows. CD86+/NOX2+ cells increased by 24 hours post-injury through 7 days post-injury. At 28 days post-injury staining resembled sham but returned to increased levels at 60 days post-injury (A). Bar = 50 μm. Double-immunolabeling was then semi-quantitated using a 5 point scale and is presented as total amount of double label (B), percent of CD86 positive cells that are also NOX2 positive (C) and percent of NOX2 positive cells that are also CD86 positive (D). NOX2 expression by M1 polarized cells increased acutely post-injury, and at 7 days after injury, 100% of the cells expressing NOX2 were CD86+. Of the CD86+ cell population, 75% expressed NOX2. Bars represent mean score or percent +/- SEM of naïve/sham tissue (n = 11) or 1 (n = 5), 4 (n = 4), 7 (n = 5), 28 (n = 3) or 60 (n = 5) days post-injury. p<0.05, *p<0.01, using One-Way ANOVA with Dunnett's Multiple Comparison post-test. Figure 4 NOX isoforms are associated with M2 polarized microglia/macrophage at acute time points M2 polarized macrophages/microglia were labeled with CD206 (red) and an antibody against NOX2 (green) in sham-injured tissue and at all examined time points post-injury. Four and 7 day post-injury tissue is shown. Co-labeled cells are indicated with arrows. CD206+ cells co-expression of NOX2 increases acutely after injury but returned to sham/naïve beyond 7 days post injury (A). Size Bar = 200 μm. Double-immunolabeling was then semi-quantitated using a 5 point scale and is presented as total amount of double label (B), percent of CD206 positive cells that are also NOX2 positive (C) and percent of NOX2 positive cells that are also CD206 positive (D). M2 polarized cells positive for NOX2 showed a peak at 4 days after SCI. At this point, the percent of CD206+ cells expressing NOX2 had shifted from less than 10% to 30-50%. Of the NOX2+ cells, 40-50% co-labeled with CD206 antibody. Bars represent mean score or percent +/− SEM of naïve/sham tissue (n = 11) or 1 (n = 5), 4 (n = 4), 7 (n = 5), 28 (n = 3) or 60 (n = 5) days post-injury. p<0.05, p<0.01, p<0.001, using One-Way ANOVA with Dunnett's Multiple Comparison post-test. Figure 5 Evidence of an intermediate polarization state Triple labeling for NOX2, CD32 and CD206 shows that NOX2 is present in cells in an intermediate state of polarization at 7 days post injury. Bar = 20 μm. Figure 6 M1 polarized macrophages/microglia express NOX4 post injury Macrophages/Microglia were labeled with CD86 (green) and immunostained with an antibody against NOX4 (red) in sham-injured tissue and at all examined time points post-injury. Co-labeled cells are indicated with arrows. CD86+/NOX4+ cells increased steadily from 1 to 60 days post-injury at all time points except at 7 days. At 7days post-injury staining resembled sham but returned to increased levels at 28 days post-injury (A). Bar = 200 μm. Double-immunolabeling was then semi-quantitated using a 5 point scale and is presented as total amount of double label (B), percent of CD86 positive cells that are also NOX4 positive (C) and percent of NOX4 positive cells that are also CD86 positive (D). NOX4/CD86 double labeled cell population increased gradually following the first day post-injury, reaching significance at 60 days. The percent of CD86+ cells that were also NOX4+ shifted from approximately 10% at 1 day post-injury to approximately 60% at 60 days post-injury. Of the NOX4+ cell population around 60% was also CD86+ at 60 days post-injury gradually increasing from approximately 10% at 1 day post-injury. Bars represent mean score or percent +/- SEM of naïve/sham tissue (n = 11) or 1 (n = 5), 4 (n = 4), 7 (n = 5), 28 (n = 3) or 60 (n = 5) days post-injury. . p<0.01, p<0.001, using One-Way ANOVA with Dunnett's Multiple Comparison post-test. Figure 7 M2 polarized macrophages/microglia express NOX4 Macrophages/microglia were labeled with CD206 (red) and immunostained with an antibody against NOX4 (green) in sham-injured tissue and at all examined time points post-injury. Co-labeled cells are indicated with arrows. CD206+ cells co-expression of NOX4 increases acutely after injury but returned to sham/naïve at 7 days post injury (A). Bar = 50 μm. Double-immunolabeling was then semi-quantitated using a 5 point scale and is presented as total amount of double label (B), percent of CD206 positive cells that are also NOX4 positive (C) and percent of NOX4 positive cells that are also CD206 positive (D). CD206/NOX4 double labeled population increased significantly at 1 and 4 days post-injury with 40-50% of NOX4+ cells co-labeling with CD206 antibody. Of the CD206+ cells approximately 50% were NOX4+ at the same time points . The M2/NOX4 double labeled population declined by 7 days but returned at chronic time points and remained above sham/injured levels. Bars represent mean score or percent +/− SEM of naïve/sham tissue (n = 11) or 1 (n = 5), 4 (n = 4), 7 (n = 5), 28 (n = 3) or 60 (n = 5) days post-injury. p<0.01, using One-Way ANOVA with Dunnett's Multiple Comparison post-test. Figure 8 Induction of NOX2 activity is associated with induction of the M1 marker NO production, and reduction of the M2 marker CD206 BV2 microglia were collected at 1, 6, 16 and 24 hours after stimulation with LPS or control media. CD206, p47PHOX phosphorylation and nitric oxide production were assessed using western blotting (A) or Griess Assay. Phosphorylation of p47PHOX peaked at 6 hours post-treatment (B). At the same time, NO production began to increase (C), followed by a peak at 16 hours and continued elevation by 24 hours. In contrast, CD206 was significantly decreased by 16 hours (D), following the peak in p47PHOX phosphorylation. Bars represent mean +/- SEM. All experiments completed in triplicate. Figure 9 NOX2 inhibition increases CD206 expression in LPS stimulated microglia BV2 microglial cells were stimulated with LPS and treated with the NOX2 inhibitor gp91ds-tat. BV2 microglia were labeled with antibodies against iNOS and CD206 (A). Bar = 100 μm. Microglia show initial CD206 but no iNOS expression. CD206 immunolabeling was diminished after LPS stimulation while iNOS staining increased. NOX2 inhibition with gp91ds-tat increased CD206 expression compared to untreated cells, with evidence of double labeling. Polarization markers expression was also assessed by western blotting (B). Quantification showed a basal M2 marker expression with no significant change in protein expression after LPS stimulation. CD206expression was significantly elevated above LPS levels by gp91ds-tat treatment. This treatment did not significantly alter iNOS expression. Bars represent mean +/− SEM. All experiments repeated in triplicate. p<0.01, *p<0.001 vs. control; $p<0.05 vs. LPS, using One-Way ANOVA with Dunnett's Multiple Comparison post-test. Table 1 Immunochemistry and western blot antibodies. Antibody Host Company Concentration Application 3-Nitrotyrosine Mouse Abcam 5 μg/ml, 2 μg/ml IHC, WB 8OHdG Rabbit Abcam 1 μg/ml IHC CD206 (Mannose receptor) Rat Serotec 20 μg/ml ICC CD206 (Mannose receptor) Rabbit Abcam 15 μg/ml, 1 μg/ml IHC, WB CD206 (Mannose receptor) Mouse Abcam 15 μg/ml IHC CD32 Goat R&D systems 5 μg/ml IHC CD86 Rabbit Abcam 5 μg/ml IHC CD86 Rat Abcam 5 μg/ml IHC iNOS Rabbit Abcam 20 μl/ml, 2 μg/ml ICC, WB Liver Arginase Goat Abcam 5 μg/ml IHC NOX2/gp91PHOX Mouse BD Transduction Laboratories 0.5 μg/ml, 0.5 μg/ml, 1 ug/ml ICH, ICC, WB NOX4 Rabbit Thermo Fisher Scientific 1 μg/ml, 1 μg/ml ICC, IHC Phospho-p47PHOX Rabbit Sigma-Aldrich 20 ug/ml WB Legend. IHC: Immunohistochemistry; ICC: Immunocytochemistry; WB: Western blot Highlights We examined NADPH oxidase expression and microglia polarization after injury Spinal cord injury leads to a significant increase in NOX2 and NOX4 expression M1 cells preferentially express NOX2 acutely and NOX4 chronically after injury The reduction in M2 polarization is associated with a peak in NOX2 expression Inhibition of NOX2 activity in vitro increases M2 polarization of microglia This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. References Aldskogius H Kozlova EN Central neuron-glial and glial-glial interactions following axon injury. Prog Neurobiol 1998 55 1 26 9602498 Aloisi F Immune function of microglia. Glia 2001 36 165 79 11596125 Altenhofer S Kleikers PW Radermacher KA Scheurer P Rob Hermans JJ Schiffers P The NOX toolbox: validating the role of NADPH oxidases in physiology and disease. Cell Mol Life Sci 2012 69 2327 43 22648375 Bedi SS Smith P Hetz RA Xue H Cox CS Immunomagnetic enrichment and flow cytometric characterization of mouse microglia. J Neurosci Methods 2013 219 176 82 23928152 Bey EA Xu B Bhattacharjee A Oldfield CM Zhao X Li Q Protein kinase C delta is required for p47phox phosphorylation and translocation in activated human monocytes. J Immunol 2004 173 5730 8 15494525 Block ML Zecca L Hong JS Microglia-mediated neurotoxicity: uncovering the molecular mechanisms. Nature reviews 2007 8 57 69 Brandes RP Viedt C Nguyen K Beer S Kreuzer J Busse R Thrombin-induced MCP-1 expression involves activation of the p22phox-containing NADPH oxidase in human vascular smooth muscle cells. Thromb Haemost 2001 85 1104 10 11434692 Byrnes KR Garay J Di Giovanni S De Biase A Knoblach SM Hoffman EP Expression of two temporally distinct microglia-related gene clusters after spinal cord injury. Glia 2006 53 420 33 16345062 Byrnes KR Washington PM Knoblach SM Hoffman E Faden AI Delayed inflammatory mRNA and protein expression after spinal cord injury. J Neuroinflammation 2011 8 130 21975064 Carlson SL Parrish ME Springer JE Doty K Dossett L Acute inflammatory response in spinal cord following impact injury. Exp Neurol 1998 151 77 88 9582256 Chen P Cescon M Zuccolotto G Nobbio L Colombelli C Filaferro M Collagen VI regulates peripheral nerve regeneration by modulating macrophage recruitment and polarization. Acta Neuropathol 2015 129 97 113 25421425 Cheret C Gervais A Lelli A Colin C Amar L Ravassard P Neurotoxic activation of microglia is promoted by a nox1-dependent NADPH oxidase. J Neurosci 2008 28 12039 51 19005069 Choi J Ifuku M Noda M Guilarte TR Translocator protein (18 kDa)/peripheral benzodiazepine receptor specific ligands induce microglia functions consistent with an activated state. Glia 2011 59 219 30 21125642 Choi SH Lee da Y Kim SU Jin BK Thrombin-induced oxidative stress contributes to the death of hippocampal neurons in vivo: role of microglial NADPH oxidase. J Neurosci 2005 25 4082 90 15843610 Cooney SJ Bermudez-Sabogal SL Byrnes KR Cellular and temporal expression of NADPH oxidase (NOX) isotypes after brain injury. J Neuroinflammation 2013 10 155 24344836 Cooney SJ Zhao Y Byrnes KR Characterization of the expression and inflammatory activity of NADPH oxidase after spinal cord injury. Free Radic Res 2014 48 929 39 24866054 Cross AR Segal AW The NADPH oxidase of professional phagocytes--prototype of the NOX electron transport chain systems. Biochimica et biophysica acta 2004 1657 1 22 Donnelly DJ Gensel JC Ankeny DP van Rooijen N Popovich PG An efficient and reproducible method for quantifying macrophages in different experimental models of central nervous system pathology. J Neurosci Methods 2009 181 36 44 19393692 El Benna J Han J Park JW Schmid E Ulevitch RJ Babior BM Activation of p38 in stimulated human neutrophils: phosphorylation of the oxidase component p47phox by p38 and ERK but not by JNK. Arch Biochem Biophys 1996 334 395 400 8900416 Ferrante CJ Leibovich SJ Regulation of Macrophage Polarization and Wound Healing. Advances in wound care 2012 1 10 6 24527272 Fleming JC Norenberg MD Ramsay DA Dekaban GA Marcillo AE Saenz AD The cellular inflammatory response in human spinal cords after injury. Brain 2006 129 3249 69 17071951 Gao HM Jiang J Wilson B Zhang W Hong JS Liu B Microglial activation-mediated delayed and progressive degeneration of rat nigral dopaminergic neurons: relevance to Parkinson's disease. J Neurochem 2002 81 1285 97 12068076 Harrigan TJ Abdullaev IF Jourd'heuil D Mongin AA Activation of microglia with zymosan promotes excitatory amino acid release via volume-regulated anion channels: the role of NADPH oxidases. J Neurochem 2008 106 2449 62 18624925 Impellizzeri D Mazzon E Esposito E Paterniti I Bramanti P Cuzzocrea S Effect of Apocynin, an inhibitor of NADPH oxidase, in the inflammatory process induced by an experimental model of spinal cord injury. Free Radic Res 2011 45 221 36 20954833 Jia Z Zhu H Li J Wang X Misra H Li Y Oxidative stress in spinal cord injury and antioxidant-based intervention. Spinal Cord 2012 50 264 74 21987065 Kallenborn-Gerhardt W Schroder K Del Turco D Lu R Kynast K Kosowski J NADPH oxidase-4 maintains neuropathic pain after peripheral nerve injury. J Neurosci 2012 32 10136 45 22836249 Khayrullina G Bermudez S Byrnes KR Inhibition of NOX2 reduces locomotor impairment, inflammation, and oxidative stress after spinal cord injury. Journal of Neuroinflammation 2015 12 172 26377802 Kigerl KA Gensel JC Ankeny DP Alexander JK Donnelly DJ Popovich PG Identification of two distinct macrophage subsets with divergent effects causing either neurotoxicity or regeneration in the injured mouse spinal cord. J Neurosci 2009 29 13435 44 19864556 Kumar A Alvarez-Croda DM Stoica BA Faden AI Loane DJ Microglial/Macrophage Polarization Dynamics following Traumatic Brain Injury. J Neurotrauma 2015 Kumar A Barrett JP Alvarez-Croda D-M Stoica BA Faden AI Loane DJ NOX2 drives M1-like microglial/macrophage activation and neurodegeneration following experimental traumatic brain injury. Brain, Behavior, and Immunity Min KJ Pyo HK Yang MS Ji KA Jou I Joe EH Gangliosides activate microglia via protein kinase C and NADPH oxidase. Glia 2004 48 197 206 15390122 Orihuela R McPherson CA Harry GJ Microglial M1/M2 polarization and metabolic states. Br J Pharmacol 2016 173 649 65 25800044 Pajoohesh-Ganji A Knoblach SM Faden AI Byrnes KR Characterization of inflammatory gene expression and galectin-3 function after spinal cord injury in mice. Brain Res 2012 1475 96 105 22884909 Patel M Li QY Chang LY Crapo J Liang LP Activation of NADPH oxidase and extracellular superoxide production in seizure-induced hippocampal damage. J Neurochem 2005 92 123 31 15606902 Pawate S Shen Q Fan F Bhat NR Redox regulation of glial inflammatory response to lipopolysaccharide and interferongamma. J Neurosci Res 2004 77 540 51 15264224 Peng J Stevenson FF Oo ML Andersen JK Iron-enhanced paraquat-mediated dopaminergic cell death due to increased oxidative stress as a consequence of microglial activation. Free Radic Biol Med 2009 46 312 20 19027846 Popovich PG Wei P Stokes BT Cellular inflammatory response after spinal cord injury in Sprague-Dawley and Lewis rats. The Journal of comparative neurology 1997 377 443 64 8989657 Qin L Liu Y Wang T Wei SJ Block ML Wilson B NADPH oxidase mediates lipopolysaccharide-induced neurotoxicity and proinflammatory gene expression in activated microglia. J Biol Chem 2004 279 1415 21 14578353 Sroga JM Jones TB Kigerl KA McGaughy VM Popovich PG Rats and mice exhibit distinct inflammatory reactions after spinal cord injury. The Journal of comparative neurology 2003 462 223 40 12794745 Takac I Schroder K Zhang L Lardy B Anilkumar N Lambeth JD The E-loop is involved in hydrogen peroxide formation by the NADPH oxidase Nox4. J Biol Chem 2011 286 13304 13 21343298 Turtzo LC Lescher J Janes L Dean DD Budde MD Frank JA Macrophagic and microglial responses after focal traumatic brain injury in the female rat. J Neuroinflammation 2014 11 82 24761998 Wang G Zhang J Hu X Zhang L Mao L Jiang X Microglia/macrophage polarization dynamics in white matter after traumatic brain injury. J Cereb Blood Flow Metab 2013 33 1864 74 23942366 Xiong Y Rabchevsky AG Hall ED Role of peroxynitrite in secondary oxidative damage after spinal cord injury. J Neurochem 2007 100 639 49 17181549 Yao A Liu F Chen K Tang L Liu L Zhang K Programmed death 1 deficiency induces the polarization of macrophages/microglia to the M1 phenotype after spinal cord injury in mice. Neurotherapeutics 2014 11 636 50 24853068 Zhao X Xu B Bhattacharjee A Oldfield CM Wientjes FB Feldman GM Protein kinase Cdelta regulates p67phox phosphorylation in human monocytes. Journal of leukocyte biology 2005 77 414 20 15591124
PMC005xxxxxx/PMC5124528.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 9010081 21515 J Nutr Biochem J. Nutr. Biochem. The Journal of nutritional biochemistry 0955-2863 1873-4847 27723473 5124528 10.1016/j.jnutbio.2016.05.005 NIHMS822017 Article Therapeutic properties of green tea against environmental insults Chen Lixia 1 Mo Huanbiao 2 Zhao Ling 3 Gao Weimin 4 Wang Shu 5 Cromie Meghan M 4 Lu Chuanwen 1 Wang Jia-Sheng 6 Shen Chwan-Li 7 1 1966 Services, Irving, TX, USA 2 Department of Nutrition, Byrdine F. Lewis School of Nursing and Health Professions, Georgia State University, Atlanta, GA, USA 3 Department of Nutrition, University of Tennessee, Knoxville, TN, USA 4 Department of Environmental Toxicology, The Institute of Environmental and Human Health, Texas Tech University, Lubbock, TX, USA 5 Department of Nutritional Sciences, Texas Tech University, Lubbock, TX, USA 6 Department of Environmental Health Science, University of Georgia, Athens, GA, USA 7 Department of Pathology, Texas Tech University Health Sciences Center, Lubbock, TX, USA Address reprint requests and correspondence to C.-L. Shen, 1A096B, 3601 4th street, Department of Pathology, Texas Tech University Health Sciences Center, Lubbock, TX 79430-8115, USA. [email protected] 2 11 2016 27 5 2016 2 2017 01 2 2018 40 113 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Pesticides, smoke, mycotoxins, polychlorinated biphenyls, and arsenic are the most common environmental toxins and toxicants to humans. These toxins and toxicants may impact on human health at the molecular (DNA, RNA, or protein), organelle (mitochondria, lysosome, or membranes), cellular (growth inhibition or cell death), tissue, organ, and systemic levels. Formation of reactive radicals, lipid peroxidation, inflammation, genotoxicity, hepatotoxicity, embryotoxicity, neurological alterations, apoptosis, and carcinogenic events are some of the mechanisms mediating the toxic effects of the environmental toxins and toxicants. Green tea, the non-oxidized and non-fermented form of tea that contains several polyphenols, including green tea catechins, exhibits protective effects against these environmental toxins and toxicants in preclinical studies and to a much-limited extent, in clinical trials. The protective effects are collectively mediated by antioxidant, anti-inflammatory, anti-mutagenic, hepato- and neuroprotective, and anti-carcinogenic activities. In addition, green tea modulates signaling pathway including NFκB and ERK pathways, preserves mitochondrial membrane potential, inhibits caspase-3 activity, down-regulates pro-apoptotic proteins, and induces the phase II detoxifying pathway. The bioavailability and metabolism of green tea and its protective effects against environmental insults induced by pesticides, smoke, mycotoxins, polychlorinated biphenyls, and arsenic are reviewed in this paper. Future studies with emphasis on clinical trials should identify biomarkers of green tea intake, examine the mechanisms of action of green tea polyphenols, and investigate potential interactions of green tea with other toxicant-modulating dietary factors. green tea pesticides cigarette smoke PCB mycotoxin heavy metal Introduction The biological, physical, and chemical environmental toxins/toxicants are introduced into the body via different routes. Food additives, contaminants, water pollutants, and drugs can be orally ingested. Airborne toxicants, particles, and tobacco smoke (active or passive) are inhaled, whereas cosmetic chemicals are absorbed through dermal contact [1]. Adverse health effects of these environmental toxicants on human bodies are determined by dose, route of exposure, toxicokinetic and toxicodynamic balance, and individual susceptibility. Acute toxic effects can be attributed to exposure to large quantities of a toxicant, whereas chronic adverse health effects can be caused by prolonged exposures to small quantities of a specific toxicant, which can ultimately result in bioaccumulation [2–7]. Exposure to toxicants can promote the formation of reactive radical (oxygen or nitrogen) species, which are inflammatory molecules inflicting oxidative stress upon cells. Depending on the molecular targets, toxicants may impact human health at the molecular (DNA, RNA, or protein), organelle (mitochondria, lysosome, or membranes), cellular (growth inhibition or cell death), tissue, organ, and overall systemic levels [2–7]. Pesticides, smoke, mycotoxins, endocrine-disrupting chemicals (e.g., polychlorinated biphenyl), and heavy metals (e.g. arsenic) have been listed as the most common toxins/toxicants to humans. Pesticides enter the body through inhalation of aerosols, dusts, and vapor, ingestion of food additives, and direct contact. Pesticides can damage vital organs, with the liver being the most susceptible due to its role in transforming, metabolizing and eliminating chemicals from the body [8]. Studies have found that many pesticides are potential hepatotoxicants. For example, chlorfenviphos, demeton-S-methyl (DSM), methiocarb, permethrin, chlorpyriphos, triazophos, and pirimicarb cause structural and functional changes in mammalian and avian hepatocytes [8]. Moreover, some neurotoxic pesticides have been associated with diseases characterized by neural damage or failure, such as Parkinson’s disease (PD) [9]. Smoke has long been recognized as a potent environmental toxicant and human health threat because it contains numerous chemical carcinogens and poisonous gases such as carbon monoxide, hydrogen cyanide, nitrogen and sulfur oxides, halogens, and organic acids [10]. The potential pathophysiological consequence associated with exposure to these poisonous gases may include the formation of carboxyhemoglobin, cyanide poisoning (cyanide blood), organic acid/ethanol intoxication [10], and enzymatic and morphologic alterations [11, 12]. Studies have found that inhalation of environmental smoke, including cigarette smoke, increases the risk of lung cancer, respiratory diseases, liver lesions, and liver cancer [11, 13] as well as the severity of liver damage in hepatitis patients [14]. Aflatoxins, mainly produced by Aspergillusflavus and A. parasiticus, are a subcategory of mycotoxins that, much like alcohol, often possess hepatotoxic properties [3]. Aflatoxin B1 (AFB1) exposure has been shown to cause acute, sub-acute, and chronic liver failure [14]. Furthermore, AFB1 is recognized as a potent carcinogen and mutagen [15]; the extent of aflatoxin contamination across regions of the United States has been correlated with incidences of hepatocellular carcinoma [15]. Many environmental contaminants act as endocrine disrupting chemicals (EDCs), capable of mimicking or blocking the action of hormones by binding to or interfering with their receptors. A subset of EDCs is known to affect metabolic processes if exposure occurs during early development, leading to obesity, type 2 diabetes mellitus and the metabolic syndrome. These chemicals are called “obesogens”. One class of the common obesogens is polychlorinated biphenyls (PCBs). PCBs are a major class of highly persistent organic pollutants (PCPs), widely used as synthetic chemical mixtures in industrial settings until it was banned in the United States and other developed countries beginning in 1970s. However, due to its resistance to degradation and bioaccumulation nature, the environmental and health impacts of PCBs are still of concerns. Epidemiological evidence now implicates exposure to PCBs in an increased risk of developing diabetes, hypertension, and obesity, all of which are clinically relevant to the onset and progression of cardiovascular disease. It is also suggested that PCBs exert their cardiovascular toxicity via additional mechanisms, including induction of chronic oxidative stress, inflammation, and endocrine disruption [16]. Exposure to inorganic and organic arsenic compounds in the environment remains a major public health problem, affecting hundreds of millions of people worldwide. Arsenic compounds affect almost every organ in the body, with health effects ranging from skin lesions and cancer to diabetes and lung disease [17, 18]. However, substantial knowledge gaps remain, particularly regarding the mechanisms by which arsenic induces such diverse health effects. Reactive oxygen species (ROS) generation is known to play a fundamental role in the arsenic-associated toxicity and carcinogenesis [19, 20]. Due to the inevitable human exposure to the aforementioned common environmental toxicants, there is a need for effective approaches to reduce or even eliminate their harmful impacts. Complementary and alternative approaches, such as dietary antioxidants or functional foods, could provide a safer way of protection or prevention than currently available options. Tea, the dried leaves of the Camellia sinensis species of theaceae family, is a popular beverage with an annual production of three billion kilograms worldwide [21]. Green tea is a non-oxidized and non-fermented product that is made by drying fresh leaves (roasting) at high temperatures to inactivate the oxidizing enzymes. Green tea contains several tea polyphenols – primarily green tea catechins (GTCs) – that accounts for 30–40% of the extractable solids of dried green tea leaves [21]. Tea catechins include (−) epigallocatechingallate (EGCG), (−) epicatechingallate (ECG), (−) epicatechin (EC), and (−) epigallocatechin (EGC) [21], among which EGCG is the most abundant and bioactive and the most studied. GTCs are known to increase the amount of anti-oxidative enzymes in the blood, and function as antioxidants to scavenge ROS such as superoxide, hydrogen peroxide (H2O2), and hydroxyl radicals [22, 23]. In the past decades, GTCs have demonstrated the ability to quench free radicals generated by oxidative environmental toxicants [24] and consequently, reduce toxicant-mediated cytological damage, mutation-mediated DNA damage, cancer, and apoptosis. This review will discuss the potential benefits of GTCs in the attenuation of the side effects and toxicity associated with common environmental toxicants including pesticides, smoke, mycotoxins, PCBs, and arsenic in in vitro and in vivo studies. Bioavailability and metabolism of green tea catechins The bioavailability of oral GTCs is generally less than 0.2% in humans and research animals [25–28]. Blood concentrations of GTCs peak at approximately 0.5 μM two to four hours after oral consumption of two cups of green tea [27]. The absolute oral bioavailability of EGCG is about 0.1% following the intake of 10 mg of green tea extract per kg body weight in humans and research animals [26, 28]. GTCs are metabolized in vivo through various metabolic transformations including methylation, glucuronidation, sulfation, oxidative degradation, and ring-fission metabolism [29–33]. The liver and intestine are generally considered to be the main organs to metabolize GTC. One third of GTCs in mesenteric plasma are in the form of glucuronide conjugates of catechin and 3′-O-methyl catechin (3′OMC), suggesting that glucuronidation and methylation occur in the intestinal tract [34]. The absorbed GTC and associated metabolites are first delivered to the liver where high levels of UDP-glucuronyltransferase [35, 36], sulfotransferase [37, 38], and catechol-O-methyltransferase (COMT) [39], among other enzymes, further metabolize GTC. After exiting the liver, GTCs and their metabolites are released into circulation system and distributed to different organs and tissues. Although GTCs have many metabolites in the human body, the biological activity of those metabolites remains unknown. Green tea modulates pesticide-related damage or disease Massive application of pesticides worldwide has conferred immense agricultural advances that in turn have led to improved nutrition and health. Most pesticides work via inhibition of pest growth and development or direct toxicity. Though researchers initially believed pesticides were harmless to living organisms, including humans, the advancement of technology has revealed many toxic effects, such as hematologic and immunological abnormalities, genotoxicity, embryo toxicity, neurological alterations, and hepatic dysfunction [40]. The hepatotoxicity of pesticides is related to metabolism through cytochrome P450 (CYP450) enzymes. The nephrotoxic effects include the formation of calculi, renal dysfunction, renal tubular acidosis, crystal uria, and hematuria. Additional adverse effects induced by pesticides, such ascyromazine, include high blood pressure, reduced body weight, and epithelial hyperplasia [40]. Most damage to membranes and tissues caused by pesticides is attributed to oxidative stress mediated by ROS such as hydroxyl radicals and H2O2 [40]. Table 1 lists the in vitro and in vivo studies showing the protective effects of green tea against different pesticides [9, 41–51]. Green tea extracts and polyphenols diminish pesticide-induced inhibition of cellular proliferation [9, 41, 42, 47, 49] and apoptosis [9, 41, 50], modulates intracellular signal transduction pathways [9], and elicits protective effect in a variety of neural cells [9, 45, 46]. For example, pretreatment with 500 μM L-theanine, a green tea ingredient, in SH-SY5Y neuroblastoma cells for 1 h significantly attenuated rotenone-induced loss of cell viability [9]. In PC12 cells, a commonly used cell line that retains the features of dopaminergic neurons, EGCG at low concentrations (1, 5, and 10 μM) significantly decreased paraquat-induced cell death, whereas higher concentrations of EGCG (50, 100, 200 μM) did not show any protective effect against paraquat damage [41]. In transformed RGC-5 retinal ganglion cells, EGCG (10–100 μM) and EC (10–75 μM) attenuated the rotenone-induced loss of cell viability [42]. In mouse hepatocytes and normal human epidermal keratinocytes (NHEK), 0.05–50 μg/mL green tea antioxidant (GTA) prevented the killing of hepatocytes by paraquat (1–10 mM) [47]. Tai et al. reported that EGCG protects SH-SY5Y from dichlorodiphenyl-trichloroethane (DDT)-induced cell death in a time-dependent manner [49]. Green tea bioactive compounds (i.e., L-theanine or EGCG) have also demonstrated the ability to inhibit pesticide-induced apoptosis [9, 41, 50]. For instance, L-theanine (500 μM) inhibited pesticide-induced apoptosis in SH-SY5Y cells as shown by the marked attenuation of caspase-3 activities [9, 41], chromatin condensation, and nuclear fragmentation [9, 41]. EGCG (1, 5, 10 μM) was found to markedly reduce pesticide-induced DNA fragmentation [41]. The potential mechanisms include the maintenance of the mitochondrial membrane potential, inhibition of caspase-3 activity, and the down-regulation of the expression of the pro-apoptotic protein, Smac, in the cytosol [41]. EGCG and catechins at ≥ 10.0 μM both reduced the effect of paraquat on the cell cycle rate [50]. Furthermore, both L-theanine and EGCG have been shown to protect against pesticide-mediated neuronal damage [9, 45, 46]. Glial cell line-derived neurotrophic factor (GDNF) and brain-derived neurotrophic factor (BDNF) act as potent and specific neurotrophic factors for midbrain dopaminergic neurons [52]. Cho et al. reported that L-theanine attenuated rotenone-mediated down-regulation of GDNF and BDNF in dopaminergic neurons [9]. Moldzio et al. also reported that EGCG significantly blocked rotenone-mediated increase in propidium iodide uptake, a measure of cell death, and 4-amino-5-methylamino-2′-7′-difluoroescein diacetate (DAF-FM) fluorescence intensity in organotypic striatal cultures from mice [45]. EGCG provides some protection for striatal slices by counteracting the rotenone-induced nitric oxide production. Pan et al. first reported that green tea polyphenols (GTP) exert a partial inhibitory effect on the uptake of 3H-dopamine (3H-DA) and 3H-1-methyl-4-phenylpyridinium (MPP+) by DA transporters (DAT) and provide partial neuro-protection from MPP+-induced injury in DAergic neurons, probably via their ability to block the DAT-dependent uptake of neurotoxin MPP+ [46]. Other studies examined the anti-oxidative activity of GTA, L-theanine, and EGCG in protection against pesticides. GTA prevents H2O2-induced cytotoxicity in cultured mouse hepatocytes (B6C3F1) and NHEK cells in a concentration-dependent manner [47]. Kamalden et al. found that green tea attenuates lipid peroxidation and rotenone toxicity in RGC-5 cells [42]. Heme oxygenase-1, a cellular stress protein expressed in brain and other tissues, responds to oxidative challenge [53]. L-theanine suppressed the production of ROS and the subsequent expression of heme oxygenase-1 induced by PD-related neurotoxins in SH-SY5Y cells [9]. L-theanine treatment also blocked the rotenone-induced oxidative stress and down-regulation of extra-cellular signal-regulated kinase 1/2 (ERK 1/2) phosphorylation [9]. Tanaka et al. reported that EGCG and green tea decreased the frequency of sister-chromatid exchanges (SCE) induced by paraquat, a generator of ROS, suggesting that green tea and polyphenol-containing foods may protect against ROS-induced genotoxicity [50]. Apart from in vitro testing, animal studies further confirmed the beneficial effects of green tea extracts (GTE, a mixtures of green tea polyphenols) against pesticide-induced toxicities including oxidative stress [40, 43, 46] in the liver [40], lung [43], and the nervous system (Table 1). Administration of the pesticides cyromazine and chlorpyrifos to rats led to significant elevations of transaminases and lactate dehydrogenase in the serum and a decrease in liver function. There were rises in catalase, hepatic SOD, and lipid peroxidation, and a decrease in glutathione (GSH), indicating increased oxidative stress [40]. Histopathological evaluation revealed that green tea supplementation mitigated mild hepatocyte necrosis and liver degeneration. Additionally, green tea normalized catalase, SOD, and liver function in the rats. Kim et al. showed that GTE significantly decreased pulmonary fibrosis induced by oxidative stress attributed toparaquat exposure in rats. GTE acted putatively by suppressing oxidative stress and endothelin-1 expression [43]. Lastly, green tea extract has been shown to suppress liver carcinogenesis induced by the pesticide pentachlorophenol (PCP) [48, 51]. In general, green tea extract contains, in descending order of abundance, EGCG, EGC, EC, ECG, and catechin. Umemura et al. found that a 2% green tea infusion prevented diethylnitrosamine (DEN)-induced and PCP-promoted hepatocellular tumors, and arrested the progression of cholangiocellular tumors, plausibly by counteracting carcinogen-mediated increases in serum alanine transaminase (ALT) activity and 8-oxodeoxyguanosine (8-oxodG) levels in the liver [51]. In a similar study, Sai et al. found that green tea acted as an anti-promoter against PCP-induced mouse hepatocarcinogenesis via its ability to prevent down-regulation of gap junction protein 1c (GJIC) [48]. Korany et al. [44] also showed that green tea pretreatment significantly reduced histopathological alterations induced by the pesticide fenitrothion. These studies, though limited in number, further elucidate the possible role of green tea in the inhibition of pesticide-induced carcinogenesis. In summary, preclinical studies showed that the protective effects of green tea against pesticide toxicity are, in part, due to its antioxidant and free radical scavenging activity. Most prominently, green tea extracts and polyphenols have been shown to decrease the toxicity of rotenone [42, 45], DDT [49], and paraquat [41, 47]. Future investigations should further examine these findings in preclinical and hopefully clinical studies since the latter is a noticeable research gap in the current literature. The role of green tea as a prophylactic agent prior to pesticide exposure may also warrant examination. Green tea modulates smoke related diseases or damage Smoking has been linked to malignant transformation or abnormal cell proliferation [54], and tobacco smoke is regarded as one of the leading causes of lung cancer [41, 47, 55]. Carcinogenic materials present in smoke produce DNA adducts and a mutagenic and carcinogenic response [56]. Toxins in cigarette smoke might also initiate and exacerbate tissue injury [23]. Another smoking related ailment is Chronic Obstructive Pulmonary Disease (COPD). Oxidative stress caused by a high concentration of free radicals and other oxidants in cigarette smoke can lead to inflammation, direct damage to epithelial cells, inactivation of anti-protease, and lipid peroxidation [57]. The protective activity of green tea and its constituents, including GTE, GTP, EGCG, ECG, EGC, EC, and catechins, on smoke-induced damage has been shown in cell culture, animal, and human studies (Table 2) [22, 51, 57–83]. Cellular studies have shown that EGCG and GTE protect against smoke-induced cellular proliferation [62, 71, 77], DNA damage [66, 80, 83], oxidative stress [22, 62, 71], and tumorigenesis [56, 71] through cell signaling pathways [71, 77]. More specifically, Syed et al. reported that EGCG pretreatment (20–80 μM) of normal human bronchial epithelial cells (NHBE) resulted in significant inhibition of cigarette smoke condensate (CSC)-induced cell proliferation [77]. Pre-, co-, and post-incubation of primary human osteoblasts with sub-toxic concentrations of GTE (0, 50, 100, and 200 μg/ml) or catechins (0, 50, 100, and 200 μM) dose-dependently reduced cigarette smoke medium (CSM)-mediated cellular damage and concomitantly increased the viability of human osteoblasts [62]. Aside from inhibiting smoke-induced cellular proliferation in normal cells, green tea polyphenols, such as GTP, EGCG, ECG, EGC, and EC, also reduced smoke-induced DNA damage [66, 71, 80]. Pretreatment with 20–100μg/mL GTP for 2 h resulted in a dose-dependent decrease in smoke-mediated DNA damage in A549 cells, and GTP concentration was inversely correlated with the extent of DNA damage observed [66]. In addition, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK)-induced single-strand DNA breaks were prevented by EGCG in A549 cells [80]. Rathore et al. reported that co-treatment with 10 μg/mL EC, ECG, EGC, and EGCG for 24 h reduced the NNK- and benzo[a]pyrene (B[a]P)-induced DNA damage in both human MCF10A breast epithelial cells and human MCF7breast cancer cells [71]. Zhou’s study demonstrated the strong anti-clastogenic effect of 0.025–0.075 mg/mL tea extract against extractable-respirable particulate in environmental tobacco smoke (ERP–ETS) in human peripheral blood lymphocytes [83]. The study also found that temperature and extraction soaking time impacted the potency of the tea extract. For example, increased soaking time (from 10, 30 to 60 min) at 100°C reduced the inhibitory activity of tea extract, whereas the same extended extraction period at lower temperatures increased the tea extract activity. Tea extract prepared by soaking at 80°C for 60 min produced the highest anti-clastogenic effect [83]. Further studies could help to confirm the precise mechanisms responsible for the inhibition of DNA damage. Studies have found that GTE has a protective effect against smoke-induced oxidation [22, 62, 71]. Protein oxidation induced by cigarette smoke in guinea pig lung microsomes was inhibited by 93% upon exposure to the green tea infusion [22]. Holzer et al. reported that osteoblasts pretreated or co-incubated with 0–200 μg/mL GTE or 0–200 μM catechins for 4 h prior to a 15-min exposure to smoke reduced ROS formation dose-dependently [62]. The GTE used for the Hozer’s study (Sunphenon® 90LB, Taiyo Kagaku, Japan) was obtained from the leaf of traceable green tea (Camelliasinensis) and consisted of >80% polyphenols, of which >80% are catechins, >40% EGCG, and <1% caffeine. Probable signal transduction pathways involved in EGCG’s protective effect against smoke induced injuries include nuclear factor-κB (NF-κB) [77], phosphoinositide 3-kinase (PI3K) signaling [77], extracellular signal-regulated kinase (ERK)1/2 pathway [71], and H2AX pathway [71]. EGCG treatment of NHBE cells suppressed smoke-induced phosphorylation of NF-κB/p65, p38, ERK1/2, and JNK1/2 MAPK in a dose-dependent manner [77]. EGCG not only inhibited DNA-binding of NF-κB, but also resulted in a significant decrease in smoke-induced NF-κB promoter activity [77]. The mechanism of NF-κB inactivation by EGCG involved IKKα phosphorylation and degradation of IkBα [77]. Rathore et al. reported that co-treatment with 10 μg/mL EC, ECG, EGC, and EGCG for 24 h reduced the NNK- and B[a]P-induced ERK1/2 and H2AX phosphorylation in both MCF10A cells and MCF7 cells [71]. In vitro studies have also shown that green tea possesses, in addition to altering signal transduction pathways, activity to block smoking-associated tumorigenesis [56, 71]. For instance, EC, EGC, EGCG and with the highest potency, ECG, have been reported to block ROS elevation and ERK pathway activation, and effectively suppress NNK- and B[a]P-induced cellular carcinogenesis [71]. Despite these promising findings, contradictory data exist on whether GTE inhibits mutagenicity [54]. As a potential mechanism for the chemopreventive activity of EGCG in renal epithelial cells, pretreatment with EGCG for 12 h preserved the level of gap junction intercellular communication (GJIC) protein, also known as connexin 43, in Madin-Darby canine kidney epithelial cells (MDCK) exposed to dimethylnitrosamine (DMN) [78]. Animal studies confirmed the anti-oxidative ability of EC, EGC, ECG, and EGCG in a variety of lung injury models. Chan et al. [57, 59] treated Sprague-Dawley rats with 2% Lung Chen Tea, a Chinese green tea, in the presence and absence of smoke for 56 days. Lung Chen Tea (2%) with the largest amount of EGCG was found to alter oxidative stress in the serum and lungs in the smoke-exposed group and protect against smoke-induced lung damage as determined by histological and morphometric analyses [57]. Concomitantly, 2% Lung Chen Tea may also help to prevent the smoke-induced up-regulation of matrix metalloproteinase-12 (MMP12) activity, leading to a reduction in ROS [59]. Yang et al. reported that catechins reduced oxidative stress by reducing the Bax/Bcl-2 ratio and preserving HSP70 chaperone protein expression in the lung of study animals [57]. Most animal studies – with a few exceptions [60, 81] – further confirmed the beneficial effects of EGCG or GTE on smoke related disease or damage. Administration of GTE, GTP, or EGCG during initiation, promotion, or progression inhibited NNK-induced lung tumorigenesis in rats, mice, and hamsters [84]. Lu’s study showed that the administration of tea polyphenols inhibited the progression of existing lung adenoma to adenocarcinoma [68]. In addition, GTP been shown to have protective effects against bladder tumors [69, 73, 75], hepatocellular tumors [51, 76], and esophageal tumors [70]. These studies provided additional evidence that GTP can possibly be used in conjunction with other therapeutics. Several clinical studies examined the effects of green tea on smoke-related DNA damage and gene mutations (Table 2) [61, 67, 79]. Lee et al. found that drinking green tea inhibited DNA mutations in blood samples from smokers and non-smokers [67]. Another study with 30 male and 90 female smokers found that consuming 4 cups or 960 mL of tea was associated with a significant decrease in urinary 8-hydroxydeoxyguanosine (8-OHdG), an indicator of DNA damage, as determined by ELISA [61]. In addition, green tea has also been shown to reduce the formation of the endogenous carcinogen, N-nitrosodimethylamine (NDMA) [79]. Despite the promising results of these clinical studies, additional data are necessary to not only understand the specific mechanisms involved, but also to fully explore the therapeutic potential of GTC. In summary, polyphenol-containing green tea may counteract environmental toxins by preventing extensive damage to DNA, proteins, and cells. GTP may reduce smoke-induced cancer risk. However, it is unclear if the action of GTP is specific for smoke-induced carcinogenesis. Besides cigarette smoke, other sources of inhaled pollution, for example, air pollutant, house dust, airborne fungi, allergic irritants and toxins, are also warranted more attention in future studies on how green tea may mitigate the potential risk of health. In addition, future studies may need to address the differential responses to green tea by individuals, and to compare the effects of various types of teas due to their diverse ingredients. Green tea and mycotoxin-related disease or damage Aflatoxin and fumonisins, low-molecular-weight secondary metabolites produced by filamentous fungi [85], are the most relevant mycotoxins associated with human diseases [86]. Aflatoxins, comprised of more than 20 fungal metabolites, were first isolated in dead turkey with turkey X disease, which was later reported to be associated with mold-contaminated peanuts [87]. The major naturally produced aflatoxins include B1, B2, G1 and G2 [88]. The letters B and G denote the blue and green fluorescence emitted upon exposure to UV light. The most toxic aflatoxin B1 (AFB1) is also a potent liver carcinogen widely used in carcinogenesis studies. Acute exposure to aflatoxins leads to acute aflatoxicosis characterized by acute hepatic injury, tissue edema, hemorrhage, and eventual death, whereas chronic exposure to aflatoxins gives rise to the development of malignancies [89]. A high level of aflatoxin exposure can promote hepatic cell necrosis. Fumonisins, first described in 1988, are produced by Fusarium verticillioides and F. proliferatum. Fumonisin B1 is the most abundant member in this family and has been associated with esophageal cancer [90], most prominently in areas with high incidence of esophageal cancer such as Transkei (South Africa), China, and northeast Italy [91]. The protective effects of green tea polyphenols, i.e. GTE and EGCG, against mycotoxin toxicity such as mutagenicity and DNA damage are summarized in Table 3 [92–107]. Salmonella typhimurium strains such as TA100, TA98, and TA97 have been used to test the effect of green tea on AFB1-induced mutagenicity [92, 103, 107]. Wang reported that GTE, GTP, and EGCG inhibited the mutagenicity of each promutagen in a dose-dependent manner. At the highest dose of GTP (500 μg/plate), the mutagenicity was inhibited by more than 95% [107]. Hour et al. determined that oolong tea, black tea, and green tea had anti-mutagenic properties and prevented cancer development, while GTE (0.1–100 μg/plate) showed the greatest inhibitory effect in strain TA100. In Snijman’s study, 0.8μM EGCG was also shown to prevent AFB-induced mutagenicity [103]. Additionally, the anti-mutagenic properties of flavonoids were related to their lipophilicity or hydrophilicity, which ultimately influences methylation, hydroxylation and glycosylation. Wang et al. reported that green tea reduced DNA damage caused by AFB1 [107] and in a dose-dependent manner, the frequency of SCE, chromosomal aberrations, and 6-thioguanine (6-TG)-resistance in Chinese hamster lung fibroblasts (V79); 100 μg/mL GTP induced 75% inhibition of SCE [107]. In addition, Mo’s group reported that the tea extracts inhibit aflatoxin synthesis through the down-regulation of the transcription of aflR and aflS, two genes involved in the regulation of aflatoxin biosynthesis [100]. Green tea also inhibits gene forward mutation and reduces the frequencies of SCE and chromosomal aberrations. Animal studies confirmed the effect of green tea on AFB1-induced carcinogenesis [98, 102, 108, 109]. Marnewick et al.[98] found that green tea enhanced hepatic microsomal fractions and protected from AFB1-induced mutogenesis in Male Fischer rats. Consumption of green tea during initiation or promotion phases of AFB1-induced carcinogenesis inhibited the glutathione S-transferase positive hepatic foci in male Fischer rats [102]. Li’s team [109] demonstrated that green tea was effective in inhibiting the hepatocarcinogenic effects of pre-inoculated diethylnitrosamine in rats. Boer et al. found that green tea markedly reduced the mutagenic potency of AFB1 and other carcinogens in a transgenic rodent model [108], while Tulayakul et al. showed that the GTE increased AFB1 detoxification via conjugation of AFB1 to GSH in the intestinal tissues, but not the liver. Dietary AFB1 has been identified as one of the major etiologic factors for primary liver cancer (mainly hepatocellular carcinoma) in the developing world, such as Southwest Asia and sub-Saharan Africa [106]. GTP has been shown as safe and effective chemopreventive agents in various AFB1-induced liver tumors. In a randomized, double-blinded, and placebo-controlled phase II chemoprevention trial in China, Wang’s team evaluated the efficacy of GTP intervention on urinary 8-OHdG in 124 high-risk individuals with aflatoxin exposure [106]. All participants tested positive for aflatoxin-albumin adducts and took either a placebo or a capsule containing 500 mg or 1,000 mg GTP daily for 3 months. Prior to the intervention and at the first and third months of the study, 24-hour urine samples were collected. The purity of GTP capsules was > 98.5% and the GTP consisted of a mixture of EGCG, EGC, ECG, EC, and catechin. Wang’s study demonstrated that (i) in GTP-treated groups, urinary EGC and EC levels displayed significant and dose-dependent increases, and (ii) at the end of 3-month intervention, urinary 8-OHdG concentrations decreased significantly in both GTP-treated groups [106], suggesting GTP is effective in diminishing oxidative DNA damage. A meta-analysis by Guyton et al. supported the preventive activity of green tea in liver cancer mediated by hepatotoxic contaminants including aflatoxins [110], but further studies are necessary to better understand the clinical applications of GTP upon mycotoxin exposure. AFB1 metabolism and specific AFB1 biomarkers were studied to evaluate the efficacy of chemopreventive agents such as GTP, in order to determine if they could provide mechanistic information for human intervention trials. From the same clinical study by Wang’s research team, Tang et al. further elucidated that daily administration of GTP (500 mg) significantly (15%) reduced levels of AFB1-albumin adducts (AFB–AA) after 3 months, compared with levels of the placebo and baseline groups. Similar results were also observed in the 1000 mg GTP group at 1 and 3 months after the intervention. In addition to AFB–AA, urinary aflatoxin M1 (AFM1) is another biomarker for AFB1 exposure that correlates well with dietary intake of AFB1 and the risk of human hepatocellular carcinoma (HCC). A reduction in urinary levels of AFM1 at 3 months of intervention was also observed in Tang’s study [104]. Intervention with 500 and 1000 mg GTP significantly elevated urinary levels of AFB1-mercapturic acid (AFB–NAC), the major detoxifying metabolic product of AFB1-8,9-epoxide. The increase in the AFB1–NAC:AFM1 ratio in GTP-treated groups further demonstrated the GTP-mediated induction of the phase II detoxifying pathway in AFB1 metabolism. Future studies should identify the optimal doses of green tea for the prevention of aflatoxin damage, taking into account variation in individual aflatoxin levels as a confounding factor that is largely attributed to differences in food products and most important in regions with high aflatoxin exposure. GTP has been considered as one of the major resources of chemopreventive agents against many types of cancers. Huang et al. reported that GTP have been shown to be relatively safe at target organs such as the liver [93]. In the same clinical trial, Wang and Luo reported that major GTP, especially EGCG, were detectable in plasma after 1- and 3-month GTP interventions; plasma ECG and EGCG and urinary EC and EGC may be used as reliable biomarkers to reflect the consumption of green tea or GTP supplementation at the population level [96, 106]. Green tea modulates toxicities associated with PCBs Some coplanar PCBs, e.g., PCB 126, exert their vascular toxicity primarily via stimulation of the aryl hydrocarbon receptor (AhR) and subsequent uncoupling of cytochrome P450 1A1 (CYP1A1), leading to oxidative stress and inflammation [111]. ROS induce the activity of antioxidant enzymes, such as superoxide dismutase (SOD), catalase, glutathione reductase (GSR), glutathione transferases (GST), thioredoxins (Trx) and thioredoxin reductases (TrxR) [111]. The antioxidant enzymes work together either to quench ROS or to reactivate enzymes through a crosstalk of multiple regulatory pathways including the aryl hydrocarbon receptor (AhR) and nuclear factor (erythroid-derived 2)-like 2 (Nrf2) transcription factors [111]. Petriello et al. recently have demonstrated both AhR and Nrf2 signaling pathways are activated by PCBs [111]. Table 4 summaries the protective effects of green tea against the toxicity of PCBs [112–116]. Anti-inflammatory food components, such as green tea EGCG, can work through both AhR- and Nrf2-mediated mechanisms to prevent PCB-induced inflammation. For example, EGCG can protect against the activation of vascular endothelial cells by coplanar PCBs [112, 115]. EGCG can suppress the expression of AhR-regulated CYP1A1 and induce Nrf2-regulated antioxidant enzymes, such as GST and NQO1, thus providing protection against PCB-induced inflammatory responses in cultured endothelial cells [112]. In addition, GTE has been shown to decrease oxidative stress in livers of mice exposed to PCB 126 via the induction of SOD1, GSR, NQO1 and GST [114]. In summary, epidemiological studies have linked exposure to POPs, such as PCBs, with an increased risk of developing diabetes, hypertension, and obesity, all of which lead to onset and progression of CVD [16]. Emerging evidence now suggests that green tea with its anti-inflammatory and anti-oxidative properties can decrease vascular toxicity and provide protection against PCBs-induced vascular toxicities. Green tea modulates arsenic-associated toxicities and cancer The beneficial effects of green tea on arsenic-associated toxicities were summarized in Table 5 [117–130]. Green tea and its principal polyphenol, (−)-EGCG, efficiently counteracted the cytotoxic effects of arsenic compounds in Chinese hamster lung fibroblast V69 cells [128]. This in vitro study is supported by subsequent in vivo studies [124]. Arsenic administration (3 mg/kg/day) for 14 days in rabbits resulted in significant oxidative stress, as revealed by reduction of whole blood GSH and elevation of thiobarbituric acid reactive substances (TBARS) and the index of nitrite/nitrate (NOx) levels. GTE administered to arsenic-treated rabbits for 14 days caused a significant elevation of the depleted GSH levels, which is likely attributed to the high polyphenol content of GTE [124]. Similarly, GT ameliorated arsenite (As III)-induced oxidative stress in Swiss albino mice by reducing the levels of lipid peroxides and protein carbonyl [130]. Co-administration of green tea extract also reduced arsenic (NaAsO2)-induced toxicity in liver, kidney and testicular and lipid peroxidation in experimental rats [122]. Interestingly, it has been reported that GTE significantly altered transport and uptake of As (III) across the Caco-2 cell model system, providing additional means of protection against arsenic exposure [119]. Arsenic exposure is associated with DNA damage, changes in ploidy of cells, and nonrandom chromosome aberrations, leading to development of cancer. Both green tea and its polyphenols ameliorated arsenic-induced genotoxicity in the Chinese hamster lung fibroblast cells (V79) as determined by micronucleus assay [128]. In addition, tea extracts are effective in counteracting the sodium arsenite-induced clastogenicity (chromatid breaks, in particular) and inducing phase II detoxification enzymes, such as SOD and catalase, in the same V79 cells, suggesting that the antioxidant function of tea in reducing clastogenicity may be partly due to the induction of these enzymes [129]. Moreover, tea extracts reduced As III-induced DNA damage in human lymphocytes, as determined by comet assays. Tea also quenched excessive ROS production, reduced the elevated levels of lipid peroxidation, and increased the activities of catalase, SOD, and glutathione peroxidase. Furthermore, tea enhanced recovery from DNA damage, as confirmed by unscheduled DNA synthesis and pronounced expression of DNA repair enzyme poly (ADP-ribose) polymerase [126]. Tea extracts have been shown to reduce oxidative stress and induce DNA repair, all of which led to protection of arsenic-induced cancer. GTE reduced arsenic-induced formation of 8OHdG and induced arsenic-suppressed DNA repair enzymes, such as PARP1, DNA β-polymerase, XRCC1, DNA ligase III, DNA protein kinase (catalytic subunit), XRCC 4, DNA ligase IV, and DNA topoisomerase Iiβ, in Swiss albino mice [127]. GTE (≥10 mg/ml aqueous) restored arsenic-induced mutagenic DNA breaks and liver damages in rats via upregulation of cytosolic SOD [117]. Similarly, supplementation of tea extracts (10 mg/mL water) with NaAsO2 (0.6 ppm)/100 g b.w. for 28 days in rats protected against arsenic-induced oxidative damages to DNA and small intestinal tissues by upregulation of antioxidant systems. In addition, in situ incubation of rat intestinal loop with NaAsO2 alone (250 μM) or with aqueous GTE (250 mg/mL) for 24 h showed that small intestinal epithelial cells were significantly protected by tea extract against arsenic-associated necrotic/mutagenic damages, suggesting that GTE have a direct role in free radical scavenging; the latter is associated with protection against mutagenic DNA breakages and tissue necrosis induced by arsenic [118]. Arsenic is metabolized to monomethylated arsenic (MMA), and subsequently becomes dimethylated arsenic (DMA). It was also reported that the incidence of lung tumors induced by lung tumor initiators (4NQO) and dimethylarsinic acid (DMA(V)), and 8-oxodG, were suppressed by co-treatment with EGCG [131]. Conclusions and summary Figure 1 summarizes the working mechanism of green tea in its protection against the five most common environmental toxins and toxicants. Green tea and its extracts reduce the environmental toxicant-induced oxidative stress and damages to DNA and cellular structures by down-regulation of HO-1, MMP12, NFκB, PI3K and ERK and up-regulation of antioxidant enzymes such as SOD and GSH, resulting in ROS quenching and lowering of oxidative products, including 8-OHdG. Green tea also suppresses toxicant-mediated carcinogenesis by preventing carcinogen-induced DNA damage, inhibiting cellular proliferation, inducing apoptosis and restoring connexin levels. On the other hand, green tea protects normal tissues and cells from toxicant-induced apoptosis by down-regulating caspases-3 and up-regulating SDO and GSH, thereby maintaining mitochondrial integrity. In addition, green tea attenuates signaling pathways such as NFκB and ERK pathways triggered by toxicants. The emerging evidence shows the promising protective/detoxifying impacts of green tea on environmental toxins. The lack of clinical evidence contributes to the need for such future studies. For instance, the development of predictive biomarkers for green tea consumption in the human population will offer a better understanding of the interaction between green tea and endogenous and exogenous factors that affect its bioavailability, and help to establish the safe doses of green tea consumption. Further development of molecular markers for the biological effects of green tea will also help to elucidate its underlying mechanisms of action. The interaction of green tea ingredients and other dietary factors with similar toxicant-modulating activities may also warrant further investigation. This study was supported by the National Center for Complementary and Integrative Health (NCCIH) of the National Institutes of Health, under grant U01AT006691 (C.L.S.). The contents of this manuscript are solely the responsibility of the authors and do not necessarily represent the official views of the NCCIH or the National Institutes of Health. Figure 1 Schematics representing the mechanisms of action for green tea in its protection against the five most common environmental toxins and toxicants. Green tea and its extracts suppress toxicant-induced cell proliferation and tumorigenesis by blocking carcinogen-mediated DNA damage. Green tea also quenches ROS and reduces oxidative stress by down-regulating MMP12, Bax/Bcl, HO-1, and NFκB, PI3K and ERK pathways and up-regulating SOD and GSH; the latter effects, coupled with down-regulation of caspases-3, offer additional protection of normal tissues from toxicant-induced apoptosis. Table 1 Green tea and pesticide-related injury First author, yr [ref] Study design Preparation of extract or GTP Used Treatments Effects of green tea Cellular studies Cho, 2008 [9] SH-SY5Y cells (neuroblastoma) Not available Rotenone Rotenone+ L-theanine ↑ Cell proliferation ↓ DNA fragmentation and apoptosis ↓ Heme oxygenase-1 up-regulation ↑ ERK1/2 phosphorylation ↑ GDNF and BDNF production Hou, 2008 [41] PC12 cells (pheochromocytoma cells) Pure EGCG Paraquat Paraquat+EGCG ↓ Apoptosis via maintaining mitochondrial membrane potential ↓ Caspase-3 activity ↓ expression of SMAC in cytosol Kamalden, 2012 [42] RGC-5 cells (retinal ganglion cells) Pure EGCG and EC Rotenone Rotenone+EGCG Rotenone+EC ↓ Rotenone-induced toxicity via JNK and p38 pathways ↓ Lipid peroxidation Ruch, 1989 [47] Mouse hepatocytes Human keratinocytes (NHEK cells) GTE extracted with methanol Paraquat Paraquat+GTE ↓ Apoptosis ↓ Inhibition of intercellular communication Tai, 2010 [49] SHSY-5Y cells Pure EGCG DDT DDT+EGCG ↓ DDT-induced toxicity Tanaka, 2000 [50] CNL cell (chronic neutrophilic leukemia) Purchased from Kurita industry Company, Tokyo, Japan Paraquat Paraquat+EGCG Paraquat+catechin Paraquat+EGCG+catechin ↓ Cell cycle rate ↓ ROS ↓ Paraquat-induced genotoxicity Moldzio, 2010 [45] Mesencephalic cultures Organotypic striatal cultures Pure EGCG Rotenone EGCG Rotenone+EGCG ↓ Nitric oxide production ↓ Rotenone-induced toxicity Pan, 2003 [46] Embryonic mesencephalic cells Green tea polyphenols Dopamine Dopamine+GTP ↓ Dopamine uptake ↓ MPP-induced dopamine neuron injury Animal studies Pan, 2003 [46] Male C57BL/6 mice Green tea polyphenols Control GTP ↓ Dopamine uptake in striatal synaptosomes ↓ MPP-induced dopamine neuron injury Kim, 2006 [43] Rats Green tea leaves extracted with ethanol, concentrated, diluted with water, and extracted with ethyl acetate Sham Paraquat Paraquat+GTE ↓ Paraquat-induced pulmonary fibrosis ↓ MDA, ET-1, and prepro-ET-1 mRNA expression ↓ Catalase activity Sai, 2000 [48] Male B6C3F1 mice 5-week-old 2% w/v green tea leaves were brewed with boiled water for 30 mins Control PCP GT infusion PCP+green tea infusion ↓ PCP-induced hepatocarcinogenesis ↑ Cell proliferation ↑ Production of GJIC and connexin32 Umemura, 2003 [51] Male B6C3F1 mice 5-week-old 2% w/v green tea leaves were brewed with boiled water for 30 mins Drinking water DEN DEN+GT infusion DEN+PCP DEN+PCP+GTE ↓ DEN-induced hepatocarcinogenesis Drinking water PCP GT PCP+GTE ↓ PCP-induced 8-oxodG levels in liver ↓ PCP-induced ALT activity in serum Korany, 2011 [44] Adult male Wistar rats Film coated tablets contains 30% polyphenol which produced by Arab Co. Control Fenitrothion Fenitrothion+GTE ↓ Fenitrothion-induced toxicity in rat parotid gland ↑ Caspase-3 cleaved activity Preservation of normal architecture Condensation of the chromatin ALT, alanine aminotransferase; BDNF, brain-derived neurotrophic factor; DDT, dichlorodiphenyl-trichloroethane; DEN, diethylnitrosamine; EC, epicatechin; EGCG, (−)-epigallocatechingallate; ERK, extracellular signal-regulated kinase; ET-1, endothelin-1; GDNF, glial cell line-derived eurotrophic factor; GJIC, gap junction intercellular communication; GTE, green tea extract; GTP, green tea polyphenols; JNK, c-Jun N-terminal kinases; MDA, malondialdehyde; MPP, 1-methyl-4-phenylpyridinium; 8-oxodG, 8-oxodeoxyguanosine; PCP, pentachlorophenol; ROS, reactive oxygen species; SMAC, second mitochondria-derived activator of caspases. ↑ increase; ↓ decrease. Table 2 Green tea and smoke-related injury First author, yr [ref] Study design Preparation of Extract or GTP Used Treatments Results Cellular studies Misra, 2003 [22] Guinea pig tissue microsomal suspension 1g green tea added to 10 mL of boiling water and brewed for 5 min. Control Smoke Smoke+tea infusion ↓ Smoke-induced oxidative damage of protein Holzer, 2012 [62] Primary human osteoblasts GTE (Sunphenon, Japan) Smoke Smoke+GTE Smoke+catechins ↑ Viability ↓ ROS formation Rathore, 2012 [71] MCF10A cells MCF7 cells (human breast epithelial cells) Pure EC, ECG, EGC, ECGC (Sigma-Aldrich Ltd., St. Louis, MO) NNK+B[a]P NNK+B[a]P+EC NNK+B[a]P+ECG NNK+B[a]P+EGC NNK+B[a]P+ECGC ↓ NNK+B[a]P-induced carcinogenesis ↓ NNK+B[a]P-induced ROS production ↓ Phosphorylation of ERK1/2 ↓ DNA damage Syed, 2007 [77] Normal human bronchial epithelial cells EGCG (>98% pure) (Mitsui Norin Co., Ltd Shizuoka, Japan) Smoke Smoke+EGCG ↓ Smoke-induced cell proliferation ↓ Smoke-induced activation of NF-κB, ↓ NF-κB-regulated proteins cyclin D1, MMP-9, IL-8, iNOS ↓ Smoke-induced phosphorylation of ERK1/2, JNK, p38 MAPK Leanderson, 1997 [66] A549 cells (human lung adinocarcinoma cells) Green tea extracted in 75°C water, 80% ethanol, chloroform, and ethyl acetate Control Smoke+H2O2+Iron Smoke+H2O2+Iron+EGCG ↓ Smoke- and H2O2-induced toxicity and DNA strand breaks ↓ Production of lipid peroxidation Weitberg, 1999 [80] A549 cells Pure EGCG (Sigma-Aldrich Ltd., St. Louis, MO) Control NNK NNK+EGCG ↓ NKK-induced tumor promotion ↓ NKK-induced single-strand DNA breaks Zhou, 2000 [83] Human peripheral blood lymphocytes 20 g of dry green tea leaves extracted in 400 mL of distilled water at 80°C, 90°C, and 100°C for 10, 30, and 60 min, respectively ERP/ETS ERP/ETS+GTE ↓ Smoke-induced mutations Takahashi, 2004 [78] MDCK cells (Mardin-Darby canine kidney cells) Pure EGCG (Sigma-Aldrich Ltd., St. Louis, MO) Control DMN EGCG DMN+EGCG ↑ Production of GJIC and connexin43 ↑ Phosphorylation of connexin43 Khoi, 2013 [64] Human endothelial ECV304 cells Pure EGCG (Sigma-Aldrich Ltd., St. Louis, MO) Nicotine Nicotine+EGCG ↓ Nicotine-induced ROS production ↓ Nicotine-induced cell invasion and MMP-9 activity ↑ Nicotine-induced NF-κB and AP-1 activation Animal studies Umemur, 2003 [51] Male B6C3F1 mice 5-week-old 2% w/v green tea leaves brewed with boiled water for 30 mins Drinking water DEN DEN+Green tea infusion DEN+PCP DEN+PCP+Green tea ↓ DEN-induced hepatocellular tumors Arrest the progression of cholangiocellular tumors Chan, 2009 [57] Sprague-Dawley rats 60 g dried Lung Chen tea leaves brewed in 600 mL hot water (not boiling) for 30 minutes Sham Smoke Sham+tea Smoke+tea ↓ Serum 8-isoprostane level ↓ Lung SOD activity ↓ Catalase activity Fiala, 2005 [60] Male Hartley guinea pigs 3-week-old EGCG (~94% pure) Sham Smoke Smoke+p-XSC Smoke+EGCG ↔ Smoke-induced lung cancer Witschi, 1998 [81] Male and female strain A/J mice Boiling deionized water poured over 62.5 g green tea leaves and left to stand for 15 min. Sham Smoke Smoke+GTE ↔ Smoke-induced lung tumor multiplicity Lu, 2006 [68] Female A/J mice 4–6 weeks old N=11–24 Polyphenon E containing 65% EGCG, 7% ECG, 3% EGC, 9% EC, 3% GCG Control NNK NNK+Polyphenon E ↓ Cell proliferation of adenomas ↑ Cell apoptosis of adenomas ↓ c-Jun and ERK1/2 phosphorylation Lubet, 2007[69] Female Fisher-344 rats and Sprague-Dawley rats 28-day-old Polyphenon E containing 64.3% EGCG, 3.1% EGC, 9.1% EC, 8.1% ECG, other polyphenols OH-BBN OH-BBN+Polyphenon E MNU MNU+Polyphenon E ↓ OH-BBN-induced urinary bladder tumors ↓ MMN-induced mammary cancer Sato, 2003 [75] Male Wistar rats 7-week-old Powdered green tea leaves BBN BBN+Green tea ↓ BBN-induced urinary bladder tumors Shimizu, 2011 [76] Male db/db mice Pure EGCG (Mitsui Norin Co. Ltd., Tokyo, Japan) Control EGCG DEN DEN+EGCG ↓ Tumor incidence and multiplicity ↓ Serum levels of insulin, IGF-1, IGF-2 ↓ p-IGF-1R protein ↓ Phospholylation of ERK and Akt ↓ Serum levels of free fatty acids Maliakal, 2011 [70] Female Wistar rats Boiled tap water added to tea powder (2% w/v) with intermittent stirring for 10 min and filtered NMBA NMBD+GTE ↓ Tumour multiplication, size, volume ↔ Tumour incidence Chan, 2012 [59] Sprague-Dawley rats 60 g dried Lung Chen tea leaves brewed in 600 mL hot water (not boiling) for 30 minutes Sham Smoke Sham+tea Smoke+tea ↓ Lung lipid peroxidation marker ↓ Smoke-induced serum MDA ↓ Neutrophil elastase concentration ↓ Smoke-induced MMP-12 Kaneko, 2003 [63] Female Syrian golden hamsters 6-week-old Not available Control BOP BOP+Catechin ↓ BOP-induced formation of 8-oxodG in the pancreas Abe, 2008[58] Male BrlHan:WIST@Jcl (GALAS) rats 5-week-old Green tea extract product Sunphenon BG (Taiyo Kagaku Company, Limited, Mie, Japan) DMBDD DMBDD+GTC ↓ DMBDD-induced liver and stomach cancers Kim, 2009[65] Male albino rats Green tea leaves extracted with 80% ethanol Control DMN Hepatic fibrosis control Hepatic fibrosis+GTE ↓ DMN-induced hydroxyproline ↓ DMN-induced MDA Yang, 2009[82] Female Wistar rats Decaffeinated catechins containing EGCG, ECG, GCG, EC, EGC, GC, catechin Non-cooking-oil-fumes Non-cooking-oil-fumes+catechins Cooking-oil-fumes (COF) Cooking-oil-fumes+catechins ↓ COF-enhanced ROS level ↓ COF-enhanced lung dityrosine ↓ COF-enhanced 4-HNE Level ↑ COF-decreased Bcl-2 and HSP70 Expression Roy, 2010[72] Male swiss albino mice GTP (Indfrag limited, Bangalore, India) Control GTP BTP DEN DEN+GTP ↓ DEN-induced COX-2 expression ↓ DEN-induced activation of NFκB and Akt Sagara, 2010[73] Female C3H/He mice GTP (Tokyo Kasei Industries, Tokyo, Japan) Control BBN GTP BBN+GTP ↓ Frequency of invasive tumors ↓ Tumor volume and microvessel density of intratumoral and stromal region Saiwichai, 2010[74] Rats Dried methanol extract dissolved in 500 mL of water at 50°C and washed with hexane, chloroform and ethyl acetate; residue dissolved in 50°C water and freeze dried to obtain mixture of EGCG, ECG, EGC, EC Sham Smoke Smoke+GTE ↓ Serum HMGB1 level Human studies Lee, 1997 [67] Questionnaire based study, Male Not available Non-smokers Smokers Smokers+Green tea ↑ SCE rates in smokers group ↔ Frequency of SCE in smokers Vermeer, 1999 [79] Healthy female volunteers Lyophilized green tea solids dissolved in boiling water (0.5 g in 100 mL); four cups (2 g) of tea per day in test week 4, and eight cups (4 g) of tea per day in test week 5. Fish meal rich in amines Fish meal+Green tea ↓ Nitrosation Hakim, 2008 [61] RCT Smokers (N=133) 4 cups (960 mL)/d of decaffeinated green tea product consisting of EGCG, EGC, EC, ECG, total catechins, gallic acid, total flavanol glycosides. Water Green tea ↓ Urinary 8-OHdG 4-HNE, 4-hydroxy-2-nonenal; 8-OHdG, 8-hydroxydeoxyguanosine; BBN, N-butyl-N-(4-hydroxybutyl)-nitrosamine; BOP, N-Nitrosobis(2-oxopropyl)amine; COX-2, cyclooxygenase-2; DEN, diethylnitrosamine; DMBDD, N-nitrosodiethylamine (DEN), N-methylnitrosourea (MNU), N-butyl-N-(4-hydroxybutyl)nitrosamine (BBN), 1,2-dimethylhydrazine (DMH) and 2,2′-dihydroxy-di-n-propylnitrosamine (DHPN); DMN, dimethylnitrosamine; EC, epicatechin; ECG, epicatechin-3-gallate; EGC, epigallocatechin; EGCG, (−)-epigallocatechin gallate; ERK, extracellul arsignal-regulated kinase; ERP/ETS, extractable respirable particulate/environmental tobacco smoke; GJIC, gap junction intercellul arcommunication; GTE, green teae xtract; GTP, green tea polyphenols; HMGBl,high-mobilitygroupproteinB1 ; IGF, insulin-like growth factor; IL-8, interleukin-8; iNOS, inducible nitric oxide synthase; JNK, c-Jun N-terminal kinases; MAPK, mitogen-activated protein kinases; MDA, malondialdehyde; MMP, matrixmetalloproteinases; MNU, methylnitrosourea; NF-κB, nuclearfactor-κB; NNK,4-(N-methyl-N-n-trosamino)-l-(3-pyridyl)-l-butanone; PCP, pentachlorophenolin; RCT, randomized controlled trial; ROS, reactive oxygen species; SCE, sister-chromatidex change; SOD, superoxide dismutase. ↑ increase; ↓ decrease; ↔ no difference Table 3 Green tea and mycotoxin-related injury First author, yr [ref] Study design Preparation of Extract or GTP Used Treatments Effects of green tea Cellular studies Hour, 1999[92] TA100, TA98, and TA97(Salmonella typhimurium strains) 12.5 g green tea leaves added to 500 mL boiling water and steeped for 15 min AFB1 AFB1+GTE AFB1+EGCG ↓ AFB1-induced mutagenesis Mo, 2013 [100] Aspergillusflavus 10 g green tea soaked in 100 mL of boiling demi-water for 0.5 h AFB1 AFB1+Green tea ↓ Aflatoxin production by down-regulating the transcription of afIR and afIS Snijman, 2007 [103] TA100 and TA98 Pure EGCG (Sigma-Aldrich, St. Louis, MO) AFB1 AFB1+EGCG ↓ AFB1-induced mutagenesis Wang, 1989 [107] TA100 and TA98 100 g of green tea powder suspended in water (0.75 L, 75°C), filtered, and extracted with 80% ethanol (0.75 L, 50°C); aqueous and ethanol extracts combined, concentrated and extracted with chloroform; remaining aqueous phase extracted with ethyl acetate; residue dissolved in water and freeze-dried to obtain GTP containing EC, EGC, ECG and EGCG.he oxidation of by atmospheric oxygen AFB1AFB1+GTP ↓ AFBl-induced mutagenesis ↓ AFBl-induced frequency of SCE ↓ AFBl-induced chromosomal aberrations Animal studies Ito, 1989 [94] Male Wistar rats 20 g green tea added to 1 L of boiling water and steeped for 10 min AFB1 AFB1+GTE ↓ AFB1-induced chromosome aberration Marnewick, 2004 [98] Male Fischer rats Boiled tap water added to tea leaves and stems (2 g/100 mL or 4 g/100 mL) AFB1 AFB1+Green tea ↓ Activation of AFB1-inducedmutagenesis Marnewick, 2009 [99] Male Fischer rats Boiled tap water added to tea leaves and stems (2 g/100 mL or 4 g/100 mL) DMSO DEN DEN+FB1 DEN+FB1+Green tea ↑ Oxygen radical absorbance capacity ↓ Serum cholesterol ↑ GSSG level and catalase activity ↓ GSH:GSSG ratio ↔ FB1-induced lipid peroxidation Qin, 1997 [101] Male Fischer rats Instant GT powder (Thomas J. Lipton, Englewood Cliffs, NJ) AFB1 AFB1+Green tea ↓ AFB1-induced hepatocarcinogenesis by modulation of AFB1 metabolism ↓ AFB1-DNA binding ↓ AFB1-induced GST-P-positive hepatocytes Qin, 2000 [102] Male Fischer rats Instant GT powder (Thomas J. Lipton, Englewood Cliffs, NJ) DMSO DMSO+Green tea DMSO+CCl4 AFB1 AFB1+CCl4 AFB1+CCl4 +Green tea ↓ Initiation and promotion of AFB1-induced hepatocarcinogenesis ↓ AFB1-induced GST-P-positive hepatic foci and area and volume ↑ GSH level Tulayakul, 2007 [105] Female piglets Green tea extracts (Taiyo Kagaku Co., Ltd., Mie, Japan) AFB1 AFB1+GTE ↑ AFB1 detoxification ↓ P450 enzyme activity in liver ↑ GST activity in intestine ↑ Conversion of AFB1 to aflatoxicol in the liver Human studies Huang, 2004 [93] Randomized, double blinded and placebo-controlled phase IIa chemoprevention Trial (design) GTP consisted of EGCG, ECG, EGC, and EC (Shili Natural Product Company, Guilin, Guangxi) Placebo Green tea (500 mg) Green tea (1000 mg) For 3 months Green intake at 1000 mg daily for 3 months was safe Luo, 2006 [95] Randomized, double blinded and placebo-controlled phase IIa chemoprevention GTP consisted of EGCG, ECG, EGC, and EC (Shili Natural Product Company, Guilin, Guangxi) Placebo Green tea (500 mg) Green tea (1000 mg) For 3 month ↓ 8-OHdG levels Luo, 2006 [96] Randomized, double blinded and placebo-controlled phase IIa chemoprevention trial GTP consisted of EGCG, ECG, EGC, and EC (Shili Natural Product Company, Guilin, Guangxi) Placebo Green tea (500 mg) Green tea group (1000 mg) For 3 month Metabolic profiling identified 106 metabolites, and 56 of them were chosen to construct discriminant functions. Tang, 2008[104] Randomized, double blinded and placebo-controlled phase IIa chemoprevention trial GTP consisted of EGCG, ECG, EGC, and EC (Shili Natural Product Company, Guilin, Guangxi) Placebo Green tea (500 mg) Green tea(1000 mg) For 3 month ↓ AFB-AA levels ↓ AFM1 ↑ AFB-NAC levels ↑ AFB-NAC:AFM1 8-OHdG, 8-hydroxydeoxyguanosine, AFB–AA, AFBl–albumin adducts; AFB–NAC, aflatoxin Bl–mercapturic acid; AFB1 aflatoxin B1; AFM1, aflatoxin M1; DEN, diethylnitrosamine; EGCG, (−)-epigallocatechin-3-gallate; FB, fumonisin B; GSH, glutathione; GSSG, oxidized glutathione; GST, glutathione S-transferase; GTE, green tea extract; GTP, green tea polyphenols; SCE, sister-chromatid exchange. ↑ increase; ↓ decrease; ↔ no difference Table 4 Green tea and PCB-related injury First author, yr [ref] Study design Preparation of Extract or GTP Used Treatments Effects of green tea Cellular studies Han, 2012[112] Primary vascular endothelial cells Pure EGCG (Cayman Chemical Co., Ann Arbor, MI) PCB PCB+EGCG ↓ CYP1A1 mRNA and protein expression ↓ Superoxide production ↓ MCP-1 and VCAM-1 ↑ Nrf2-controlled genes (GST and NQO1) Ramadass, 2003[115] Endothelial cells Pure EGCG (Sigma-Aldrich, Ltd., St. Louis, MO) PCB 77 (3,3′, 4,4′-tetrachlorobiphenyl) PCB 77+EGCG ↓ Oxidative stress↓ CYP1A1 activity Weisburger, 1994[116] in vitro systems of Jägerstad Not available PCB PCB+EGCG ↓ Formation of typical HCAs (MeIQx and PhIP) Animal studies Morita, 1997[113] Male rats Not available PCB PCB+GTE ↓ Liver distribution of PCB Newsome, 2014 [114] C57BL/6 mice GTE (Taiyo International Inc., Minneapolis, MN) PCB PCB+GTE ↑ SOD1, GSR, NQO1 and GST ↓ Oxidative stress EGCG, (−)-epigallocatechingallate; GTE, green tea extract; GSR, glutathione S-reductase; GST, glutathione S transferase; HCAs, Heterocyclic amines; MCP-1, monocyte chemotactic protein-1; MeIQx, 2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline; NQO1, NAD(P)H:quinoneoxidoreductase; Nrf2, nuclear factor erythroid 2 [NF-E2]-related factor 2; PCB, polychlorinated biphenyl; PhIP, 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine ; SOD1, Superoxide Dismutase 1; VCAM-1, vascular cell adhesion protein-1. ↑ increase; ↓ decrease. Table 5 Green tea and arsenic-related injury First author, yr [ref] Study design Preparation of Extract or GTP Used Treatments Effects of green tea Cellular studies Calatayud, 2011[119] Caco-2 cells GTE (Plantextrakt, Germany) As(III) As(III)+GTE ↑ TEER ↓ As(III) Papp Kim, 2015[120] BAE cells Pure EGCG As As+EGCG ↑ Cell cytotoxicity↑ ROS formation ↑ Lipid peroxidation ↓ Catalase activity Lee, 2011[121] HL60 cells Pure EGCG ATO ATO+EGCG ↑ Mitochondria-induced apoptosis ↑ ROS formation to damage cell Nakazato, 2005[123] Malignant B-cell lines Pure ECGC (WAKO Chemical Co. Tokyo, Japan) As2O3 As2O3+EGCG ↑ Apoptotic cell death ↑ ROS formation Sinha, 2005[125] V79 cells Green tea leaves extracted with hot water and extracted with ethyl acetate NaAsO2 NaAsO2+GTE ↓ Chromatid breaks ↑ SOD and CAT Sinha, 2007[126] Normal human lymphocytes Green tea leaves extracted with hot water and extracted with ethyl acetate As(III) As(III)+GTE ↓ DNA damage and ROS formation ↓ ROS formation and lipid peroxidation ↑ CAT, SOD, and GPx Sinha, 2003[128] V79 cells 2.5 g of dry tea brewed in 100 ml of boiled water for 5 min. GTE consisted of ECG, EGC, EGCG As As+GTE ↓ Apoptosis Sinha, 2005[129] V79 cells Green tea leaves extracted with hot water and extracted with ethyl acetate As As+GTE ↑ Repair activity Animal studies Acharyya, 2014[117] Rats Leaf dust of green tea dried, crushed, and extracted by distilled water As As+GTE ↓ Apoptotic hepatic degeneration ↓ DNA damage ↑ SOD1 protection Acharyya, 2015[118] Rats Leaf dust of green tea dried, crushed, and extracted by distilled water NaAsO2 NaAsO2+GTE ↓ Intestinal tissue degeneration ↓ DNA-breakages ↓ Necrotic damage Messarah, 2013[122] Rats 66 g dry leaves extracted per liter of tap water in drinking water NaAsO2 NaAsO2+GTE ↓ TBARS ↓ Oxidative stress ↑ Body weight Raihan, 2009[124] Rabbits 3.75 g green tea boiled with arsenic-free fresh drinking (100 mL) for 10 min. As As+GTE ↑ GSH ↓ TBARS and NOx Sinha, 2011[127] Swiss albino mice 0.5 g of dry tea leaves boiled with milli-Q water (5 mL) for 5 min. As(III) As(III)+GTE ↓ 8 OHdG and OGG1 ↑ DNA repair enzymes Sinha, 2010[130] Swiss albino mice 0.5 g of dry tea leaves boiled with milli-Q water (5 mL) for 5 min. As(III) As(III)+GTE ↓ Lipid peroxides ↓ Protein carbonyl ↑ CAT, SOD, GPx, GR, GST, and GSH As, arsenic; As(III), arsenite; As2O3, arsenic trioxide; ATO, arsenic trioxide; BAE, bovine aortic endothelial cells; Caco-2 cells, human epithelial colorectal adenocarcinoma cells; CAT, catalase; EGCG, (−)-epigallocatechingallate; GPx, glutathione peroxidase; GR, glutathione reductase; GSH, glutathione; GST, glutathione-S-transferase; GTE, green tea extract; HL60, human promyelocytic leukemia cells; NaAsO2, sodium arsenite; NOx, nitrogen oxides; OGG1, 8-oxoguanine DNA glycosylase; 8 OHdG, 8-hydroxy-2′-deoxyguanosine; Papp, apparent permeability; SOD, superoxide dismutase; TBARS, thiobarbituric acid reactive substances; TEER, transepithelial resistance; V79, male Chinese hamster lung fibroblasts; ↑ increase; ↓ decrease. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. 1 Toebak MJ Gibbs S Bruynzeel DP Scheper RJ Rustemeyer T Dendritic cells: biology of the skin Contact dermatitis 2009 60 2 20 19125717 2 Cohen M Environmental toxins and health–the health impact of pesticides Australian family physician 2007 36 1002 4 18075622 3 Hilgenkamp K Environmental health:ecological perspectives Jones & Bartlett Learning 2006 4 Lanphear BP Vorhees CV Bellinger DC Protecting children from environmental toxins PLoS medicine 2005 2 e61 15783252 5 Meadows-Oliver M Environmental toxins Journal of pediatric health care: official publication of National Association of Pediatric Nurse Associates & Practitioners 2006 20 350 2 6 Pan-Montojo F Schwarz M Winkler C Arnhold M O’Sullivan GA Pal A Environmental toxins trigger PD-like progression via increased alpha-synuclein release from enteric neurons in mice Scientific reports 2012 2 898 23205266 7 Stephenson J CDC report on environmental toxins: some progress, some concerns JAMA: the journal of the American Medical Association 2003 289 1230, 3 12633165 8 Zimmerman HJ Lewis JH Chemical- and toxin-induced hepatotoxicity Gastroenterology clinics of North America 1995 24 1027 45 8749910 9 Cho HS Kim S Lee SY Park JA Kim SJ Chun HS Protective effect of the green tea component, L-theanine on environmental toxins-induced neuronal cell death Neurotoxicology 2008 29 656 62 18452993 10 Alarie Y Toxicity of fire smoke Critical reviews in toxicology 2002 32 259 89 12184505 11 Moszczynski P Slowinski S Lisiewicz J [Enzyme deficiency in the lymphocytes of tobacco smokers] Przeglad lekarski 1989 46 719 22 2594989 12 Patel BP Rawal UM Shah PM Prajapati JA Rawal RM Dave TK Study of tobacco habits and alterations in enzymatic antioxidant system in oral cancer Oncology 2005 68 511 9 16037684 13 Bosetti C Turati F La Vecchia C Hepatocellular carcinoma epidemiology Best practice & research Clinical gastroenterology 2014 28 753 70 25260306 14 Bose T Bitter correlationship between autoimmune hepatitis and smoking Medical hypotheses 2015 84 118 21 25543266 15 Abbas HK Aflatoxin and Food Safety Abbas HK CRC Press 2005 16 Perkins JT Petriello MC Newsome BJ Hennig B Polychlorinated biphenyls and links to cardiovascular disease Environ Sci Pollut Res Int 2016 23 2160 72 25877901 17 Council CoIABoESaTDoEaLSNR Critical Aspects of EPA’s IRIS Assessment of Inorganic Arsenic 2014 18 Naujokas MF Anderson B Ahsan H Aposhian HV Graziano JH Thompson C The broad scope of health effects from chronic arsenic exposure: update on a worldwide public health problem Environ Health Perspect 2013 121 295 302 23458756 19 De Vizcaya-Ruiz A Barbier O Ruiz-Ramos R Cebrian ME Biomarkers of oxidative stress and damage in human populations exposed to arsenic Mutation research 2009 674 85 92 18984063 20 Shi H Shi X Liu KJ Oxidative mechanism of arsenic toxicity and carcinogenesis Mol Cell Biochem 2004 255 67 78 14971647 21 Yang CS Landau JM Effects of tea consumption on nutrition and health The Journal of nutrition 2000 130 2409 12 11015465 22 Misra A Chattopadhyay R Banerjee S Chattopadhyay DJ Chatterjee IB Black tea prevents cigarette smoke-induced oxidative damage of proteins in guinea pigs The Journal of nutrition 2003 133 2622 8 12888648 23 Sreekanth KS Sabu MC Varghese L Manesh C Kuttan G Kuttan R Antioxidant activity of Smoke Shield in-vitro and in-vivo The Journal of pharmacy and pharmacology 2003 55 847 53 12841947 24 Hara Y Green Tea: Health Benefits and Applications CRC Press 2001 25 Chen L Lee MJ Li H Yang CS Absorption, distribution, elimination of tea polyphenols in rats Drug Metab Dispos 1997 25 1045 50 9311619 26 Warden BA Smith LS Beecher GR Balentine DA Clevidence BA Catechins are bioavailable in men and women drinking black tea throughout the day The Journal of nutrition 2001 131 1731 7 11385060 27 Lee MJ Maliakal P Chen L Meng X Bondoc FY Prabhu S Pharmacokinetics of tea catechins after ingestion of green tea and (−)-epigallocatechin-3-gallate by humans: formation of different metabolites and individual variability Cancer Epidemiol Biomarkers Prev 2002 11 1025 32 12376503 28 Lambert JD Yang CS Mechanisms of cancer prevention by tea constituents The Journal of nutrition 2003 133 3262S 7S 14519824 29 Dou QP Molecular mechanisms of green tea polyphenols Nutr Cancer 2009 61 827 35 20155623 30 Lu H Meng X Li C Sang S Patten C Sheng S Glucuronides of tea catechins: enzymology of biosynthesis and biological activities Drug Metab Dispos 2003 31 452 61 12642472 31 Lu H Meng X Yang CS Enzymology of methylation of tea catechins and inhibition of catechol-O-methyltransferase by (−)-epigallocatechin gallate Drug Metab Dispos 2003 31 572 9 12695345 32 Vaidyanathan JB Walle T Glucuronidation and sulfation of the tea flavonoid (−)-epicatechin by the human and rat enzymes Drug Metab Dispos 2002 30 897 903 12124307 33 Sang S Lambert JD Ho CT Yang CS The chemistry and biotransformation of tea constituents Pharmacol Res 2011 64 87 99 21371557 34 Donovan JL Crespy V Manach C Morand C Besson C Scalbert A Catechin is metabolized by both the small intestine and liver of rats The Journal of nutrition 2001 131 1753 7 11385063 35 Boutin JA Thomassin J Siest G Cartier A Heterogeneity of hepatic microsomal UDP-glucuronosyltransferase activities. Conjugations of phenolic and monoterpenoid aglycones in control and induced rats and guinea pigs Biochem Pharmacol 1985 34 2235 49 3925953 36 Strassburg CP Nguyen N Manns MP Tukey RH UDP-glucuronosyltransferase activity in human liver and colon Gastroenterology 1999 116 149 60 9869613 37 Matsui M Homma H Biochemistry and molecular biology of drug-metabolizing sulfotransferase Int J Biochem 1994 26 1237 47 7851628 38 Shali NA Curtis CG Powell GM Roy AB Sulphation of the flavonoids quercetin and catechin by rat liver Xenobiotica 1991 21 881 93 1776264 39 Flohe L Catechol-O-methyltransferase Int Pharmacopsychiatry 1974 9 52 60 4599326 40 HEIKAL TMA-T HMOSSA ABDEL RASOUL MONAA MAREI GEHANIKH THE AMELIORATING EFFECTS OF GREEN TEA EXTRACT AGAINST CYROMAZINE AND CHLORPYRIFOS INDUCED LIVER TOXICITY IN MALE RATS Asian Journal of Pharmaceutical & Clinical Research 2013 6 48 41 Hou RR Chen JZ Chen H Kang XG Li MG Wang BR Neuroprotective effects of (−)-epigallocatechin-3-gallate (EGCG) on paraquat-induced apoptosis in PC12 cells Cell biology international 2008 32 22 30 17936647 42 Kamalden TA Ji D Osborne NN Rotenone-induced death of RGC-5 cells is caspase independent, involves the JNK and p38 pathways and is attenuated by specific green tea flavonoids Neurochemical research 2012 37 1091 101 22350517 43 Kim HR Park BK Oh YM Lee YS Lee DS Kim HK Green tea extract inhibits paraquat-induced pulmonary fibrosis by suppression of oxidative stress and endothelin-l expression Lung 2006 184 287 95 17235729 44 Korany NS Ezzat BA Prophylactic effect of green tea and Nigella sativa extracts against fenitrothion-induced toxicity in rat parotid gland Archives of oral biology 2011 56 1339 46 21696705 45 Moldzio R Radad K Krewenka C Kranner B Duvigneau JC Wang Y Effects of epigallocatechin gallate on rotenone-injured murine brain cultures Journal of neural transmission 2010 117 5 12 19707849 46 Pan T Fei J Zhou X Jankovic J Le W Effects of green tea polyphenols on dopamine uptake and on MPP+ -induced dopamine neuron injury Life sciences 2003 72 1073 83 12495785 47 Ruch RJ Cheng SJ Klaunig JE Prevention of cytotoxicity and inhibition of intercellular communication by antioxidant catechins isolated from Chinese green tea Carcinogenesis 1989 10 1003 8 2470525 48 Sai K Kanno J Hasegawa R Trosko JE Inoue T Prevention of the down-regulation of gap junctional intercellular communication by green tea in the liver of mice fed pentachlorophenol Carcinogenesis 2000 21 1671 6 10964098 49 Tai KK Truong DD (−)-Epigallocatechin-3-gallate (EGCG), a green tea polyphenol, reduces dichlorodiphenyl-trichloroethane (DDT)-induced cell death in dopaminergic SHSY-5Y cells Neuroscience letters 2010 482 183 7 20542083 50 Tanaka R Protective effects of (−)-epigallocatechin gallate and (+)-catechin on paraquat-induced genotoxicity in cultured cells The Journal of toxicological sciences 2000 25 199 204 10987127 51 Umemura T Kai S Hasegawa R Kanki K Kitamura Y Nishikawa A Prevention of dual promoting effects of pentachlorophenol, an environmental pollutant, on diethylnitrosamine-induced hepato- and cholangiocarcinogenesis in mice by green tea infusion Carcinogenesis 2003 24 1105 9 12807750 52 Siegel GJ Chauhan NB Neurotrophic factors in Alzheimer’s and Parkinson’s disease brain Brain research Brain research reviews 2000 33 199 227 11011066 53 Schipper HM Liberman A Stopa EG Neural heme oxygenase-1 expression in idiopathic Parkinson’s disease Experimental neurology 1998 150 60 8 9514830 54 Bothe H Gassmann K Gotz C Fritsche E Abel J Haarmann-Stemmann T Epigallocatechin-3-gallate does not affect the activity of enzymes involved in metabolic activation and cellular excretion of benzo[a]pyrene in human colon carcinoma cells Toxicology letters 2011 203 258 64 21458546 55 Roomi MW Roomi NW Kalinovsky T Rath M Niedzwiecki A Chemopreventive effect of a novel nutrient mixture on lung tumorigenesis induced by urethane in male A/J mice Tumori 2009 95 508 13 19856665 56 Kuttan R Kuttan G Joseph S Ajith TA Mohan M Srimal RC Antimutagenicity of herbal detoxification formula Smoke Shield against environmental mutagens Journal of experimental & clinical cancer research: CR 2004 23 61 8 15149152 57 Chan KH Ho SP Yeung SC So WH Cho CH Koo MW Chinese green tea ameliorates lung injury in cigarette smoke-exposed rats Respiratory medicine 2009 103 1746 54 19487113 58 Abe M Suzuki N Yoshida M Usuda K Furukawa S Juneja LR Possible carcinogenic risks of copper gluconate and their prevention by co-administered green tea catechins evaluated by a rat medium-term multi-organ carcinogenicity bioassay protocol Food and chemical toxicology 2008 46 1760 70 18313194 59 Chan KH Chan SC Yeung SC Man RY Ip MS Mak JC Inhibitory effect of Chinese green tea on cigarette smoke-induced up-regulation of airway neutrophil elastase and matrix metalloproteinase-12 via antioxidant activity Free radical research 2012 46 1123 9 22574903 60 Fiala ES Sohn OS Wang CX Seibert E Tsurutani J Dennis PA Induction of preneoplastic lung lesions in guinea pigs by cigarette smoke inhalation and their exacerbation by high dietary levels of vitamins C and E Carcinogenesis 2005 26 605 12 15579485 61 Hakim IA Chow HH Harris RB Green tea consumption is associated with decreased DNA damage among GSTM1-positive smokers regardless of their hOGG1 genotype The Journal of nutrition 2008 138 1567S 71S 18641208 62 Holzer N Braun KF Ehnert S Egana JT Schenck TL Buchholz A Green tea protects human osteoblasts from cigarette smoke-induced injury: possible clinical implication Langenbeck’s archives of surgery/Deutsche Gesellschaft fur Chirurgie 2012 397 467 74 63 Kaneko T Tahara S Takabayashi F Protective Effect of Fluvastatin, an HMG-CoA Reductase Inhibitor, on the Formation of 8-Oxo-2′-deoxyguanosine in the Nuclear DNA of Hamster Pancreas after a Single Administration of N-Nitrosobis (2-oxopropyl) amine Biological and Pharmaceutical Bulletin 2003 26 1245 8 12951466 64 Khoi PN Park JS Kim JH Xia Y Kim NH Kim KK (−)-Epigallocatechin-3-gallate blocks nicotine-induced matrix metalloproteinase-9 expression and invasiveness via suppression of NF-κB and AP-1 in endothelial cells International journal of oncology 2013 43 868 76 23835612 65 Kim HK Yang T-H Cho H-Y Antifibrotic effects of green tea on in vitro and in vivo models of liver fibrosis World journal of gastroenterology: WJG 2009 15 5200 19891020 66 Leanderson P Faresjo AO Tagesson C Green tea polyphenols inhibit oxidant-induced DNA strand breakage in cultured lung cells Free radical biology & medicine 1997 23 235 42 9199885 67 Lee IP Kim YH Kang MH Roberts C Shim JS Roh JK Chemopreventive effect of green tea (Camellia sinensis) against cigarette smoke-induced mutations (SCE) in humans Journal of cellular biochemistry Supplement 1997 27 68 75 9591195 68 Lu G Liao J Yang G Reuhl KR Hao X Yang CS Inhibition of adenoma progression to adenocarcinoma in a 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone-induced lung tumorigenesis model in A/J mice by tea polyphenols and caffeine Cancer research 2006 66 11494 501 17145898 69 Lubet RA Yang CS Lee MJ Hara Y Kapetanovic IM Crowell JA Preventive effects of polyphenon E on urinary bladder and mammary cancers in rats and correlations with serum and urine levels of tea polyphenols Molecular cancer therapeutics 2007 6 2022 8 17620432 70 Maliakal P Sankpal UT Basha R Maliakal C Ledford A Wanwimolruk S Relevance of drug metabolizing enzyme activity modulation by tea polyphenols in the inhibition of esophageal tumorigenesis Medicinal chemistry 2011 7 480 7 21801144 71 Rathore K Wang HC Green tea catechin extract in intervention of chronic breast cell carcinogenesis induced by environmental carcinogens Molecular carcinogenesis 2012 51 280 9 21882252 72 Roy P Nigam N Singh M George J Srivastava S Naqvi H Tea polyphenols inhibit cyclooxygenase-2 expression and block activation of nuclear factor-kappa B and Akt in diethylnitrosoamine induced lung tumors in Swiss mice Investigational new drugs 2010 28 466 71 19517063 73 Sagara Y Miyata Y Nomata K Hayashi T Kanetake H Green tea polyphenol suppresses tumor invasion and angiogenesis in N-butyl-(-4-hydroxybutyl) nitrosamine-induced bladder cancer Cancer epidemiology 2010 34 350 4 20362526 74 Saiwichai T Sangalangkarn V Kawahara K Oyama Y Chaichalotornkul S Narkpinit S Green tea extract supplement inhibition of HMGB1 release in rats exposed to cigarette smoke The Southeast Asian journal of tropical medicine and public health 2010 41 250 8 20578506 75 Sato D Matsushima M Preventive effects of urinary bladder tumors induced by N-butyl-N-(4-hydroxybutyl)-nitrosamine in rat by green tea leaves International journal of urology: official journal of the Japanese Urological Association 2003 10 160 6 12622713 76 Shimizu M Adachi S Masuda M Kozawa O Moriwaki H Cancer chemoprevention with green tea catechins by targeting receptor tyrosine kinases Molecular nutrition & food research 2011 55 832 43 21538846 77 Syed DN Afaq F Kweon MH Hadi N Bhatia N Spiegelman VS Green tea polyphenol EGCG suppresses cigarette smoke condensate-induced NF-kappaB activation in normal human bronchial epithelial cells Oncogene 2007 26 673 82 16862172 78 Takahashi H Nomata K Mori K Matsuo M Miyaguchi T Noguchi M The preventive effect of green tea on the gap junction intercellular communication in renal epithelial cells treated with a renal carcinogen Anticancer research 2004 24 3757 62 15736408 79 Vermeer IT Moonen EJ Dallinga JW Kleinjans JC van Maanen JM Effect of ascorbic acid and green tea on endogenous formation of N-nitrosodimethylamine and N-nitrosopiperidine in humans Mutation research 1999 428 353 61 10518007 80 Weitberg AB Corvese D The effect of epigallocatechin galleate and sarcophytol A on DNA strand breakage induced by tobacco-specific nitrosamines and stimulated human phagocytes Journal of experimental & clinical cancer research: CR 1999 18 433 7 10606191 81 Witschi H Espiritu I Yu M Willits NH The effects of phenethyl isothiocyanate, N-acetylcysteine and green tea on tobacco smoke-induced lung tumors in strain A/J mice Carcinogenesis 1998 19 1789 94 9806160 82 Yang CH Lin CY Yang JH Liou SY Li PC Chien CT Supplementary catechins attenuate cooking-oil-fumes-induced oxidative stress in rat lung The Chinese journal of physiology 2009 52 151 9 19777801 83 Zhou R Zhou Y Chen D Li S Haug A Effects of soaking temperature and soaking time during preparation of water extract of tea on anticlastogenicity against environmental tobacco smoke in the sister-chromatid exchange assay Toxicology letters 2000 115 23 32 10817628 84 Yang CS Wang X Green tea and cancer prevention Nutr Cancer 2010 62 931 7 20924968 85 Bennett JW Mycotoxins, mycotoxicoses, mycotoxicology and Mycopathologia Mycopathologia 1987 100 3 5 3683511 86 Bennett JW Klich M Mycotoxins Clinical microbiology reviews 2003 16 497 516 12857779 87 Squire RA Ranking animal carcinogens: a proposed regulatory approach Science 1981 214 877 80 7302565 88 Khlangwiset P Shephard GS Wu F Aflatoxins and growth impairment: a review Critical reviews in toxicology 2011 41 740 55 21711088 89 Groopman JD Kensler TW Wild CP Protective interventions to prevent aflatoxin-induced carcinogenesis in developing countries Annual review of public health 2008 29 187 203 90 Gonzalez HH Martinez EJ Pacin AM Resnik SL Sydenham EW Natural co-occurrence of fumonisins, deoxynivalenol, zearalenone and aflatoxins in field trial corn in Argentina Food additives and contaminants 1999 16 565 9 10789378 91 Peraica M Radic B Lucic A Pavlovic M Toxic effects of mycotoxins in humans Bulletin of the World Health Organization 1999 77 754 66 10534900 92 Hour TC Liang YC Chu IS Lin JK Inhibition of eleven mutagens by various tea extracts, (−)epigallocatechin-3-gallate, gallic acid and caffeine Food and chemical toxicology: an international journal published for the British Industrial Biological Research Association 1999 37 569 79 10478825 93 Huang TYJTLLHBMWZLYLY Wang K Shi J Cui Y Zhang ZQ Zhang L Sun S Martin C Wang JS Phase IIa Chemoprevention Trial of Green Tea Polyphenols in High-Risk Individuals of Liver Cancer: I. Design, Clinical Outcomes, and Baseline Biomarker Data Int J Cancer Prev 2004 1 269 80 94 Ito Y Ohnishi S Fujie K Chromosome aberrations induced by aflatoxin B1 in rat bone marrow cells in vivo and their suppression by green tea Mutation research 1989 222 253 61 2922009 95 Luo H Tang L Tang M Billam M Huang T Yu J Phase IIa chemoprevention trial of green tea polyphenols in high-risk individuals of liver cancer: modulation of urinary excretion of green tea polyphenols and 8-hydroxydeoxyguanosine Carcinogenesis 2006 27 262 8 15930028 96 Luo HC SB Gao W Yu J Tang L Wang JS Metabolic profiling in validation of plasma biomarkers for green tea polyphenols Metabolomics 2006 2 235 41 97 M IYI Suppressive Effect of (−)-Epigallocatechin Gallate on Aflatoxin B1-induced Chromosome Aberrations in Rat Bone Marrow Cells J Health Sci 2001 47 248 57 98 Marnewick JL Batenburg W Swart P Joubert E Swanevelder S Gelderblom WC Ex vivo modulation of chemical-induced mutagenesis by subcellular liver fractions of rats treated with rooibos (Aspalathus linearis) tea, honeybush (Cyclopia intermedia) tea, as well as green and black (Camellia sinensis) teas Mutation research 2004 558 145 54 15036128 99 Marnewick JL van der Westhuizen FH Joubert E Swanevelder S Swart P Gelderblom WC Chemoprotective properties of rooibos (Aspalathus linearis), honeybush (Cyclopia intermedia) herbal and green and black (Camellia sinensis) teas against cancer promotion induced by fumonisin B1 in rat liver Food and chemical toxicology: an international journal published for the British Industrial Biological Research Association 2009 47 220 9 19041360 100 Mo HZ Zhang H Wu QH Hu LB Inhibitory effects of tea extract on aflatoxin production by Aspergillus flavus Letters in applied microbiology 2013 56 462 6 23573983 101 Qin G Gopalan-Kriczky P Su J Ning Y Lotlikar PD Inhibition of aflatoxin B1-induced initiation of hepatocarcinogenesis in the rat by green tea Cancer letters 1997 112 149 54 9066721 102 Qin G Ning Y Lotlikar PD Chemoprevention of aflatoxin B1-initiated and carbon tetrachloride-promoted hepatocarcinogenesis in the rat by green tea Nutr Cancer 2000 38 215 22 11525600 103 Snijman PW Swanevelder S Joubert E Green IR Gelderblom WC The antimutagenic activity of the major flavonoids of rooibos (Aspalathus linearis): some dose-response effects on mutagen activation-flavonoid interactions Mutation research 2007 631 111 23 17537670 104 Tang L Tang M Xu L Luo H Huang T Yu J Modulation of aflatoxin biomarkers in human blood and urine by green tea polyphenols intervention Carcinogenesis 2008 29 411 7 18192689 105 Tulayakul P Dong KS Li JY Manabe N Kumagai S The effect of feeding piglets with the diet containing green tea extracts or coumarin on in vitro metabolism of aflatoxin B1 by their tissues Toxicon: official journal of the International Society on Toxinology 2007 50 339 48 17537474 106 Wang JS Luo H Wang P Tang L Yu J Huang T Validation of green tea polyphenol biomarkers in a phase II human intervention trial Food and chemical toxicology: an international journal published for the British Industrial Biological Research Association 2008 46 232 40 17888558 107 Wang ZY Cheng SJ Zhou ZC Athar M Khan WA Bickers DR Antimutagenic activity of green tea polyphenols Mutation research 1989 223 273 85 2500594 108 de Boer JG Protection by dietary compounds against mutation in a transgenic rodent The Journal of nutrition 2001 131 3082S 6S 11694652 109 Li Y Comparative study on the inhibitory effect of green tea, coffee and levamisole on the hepatocarcinogenic action of diethylnitrosamine Zhonghua zhong liu za zhi [Chinese journal of oncology] 1991 13 193 5 110 Guyton KZ Kensler TW Prevention of liver cancer Current oncology reports 2002 4 464 70 12354357 111 Petriello MC Newsome B Hennig B Influence of nutrition in PCB-induced vascular inflammation Environ Sci Pollut Res Int 2014 21 6410 8 23417440 112 Han SG Han SS Toborek M Hennig B EGCG protects endothelial cells against PCB 126-induced inflammation through inhibition of AhR and induction of Nrf2-regulated genes Toxicol Appl Pharmacol 2012 261 181 8 22521609 113 Morita K Matsueda T Iida T [Effect of green tea (matcha) on gastrointestinal tract absorption of polychlorinated biphenyls, polychlorinated dibenzofurans and polychlorinated dibenzo-p-dioxins in rats] Fukuoka Igaku Zasshi 1997 88 162 8 9194336 114 Newsome BJ Petriello MC Han SG Murphy MO Eske KE Sunkara M Green tea diet decreases PCB 126-induced oxidative stress in mice by up-regulating antioxidant enzymes J Nutr Biochem 2014 25 126 35 24378064 115 Ramadass P Meerarani P Toborek M Robertson LW Hennig B Dietary flavonoids modulate PCB-induced oxidative stress, CYP1A1 induction, and AhR-DNA binding activity in vascular endothelial cells Toxicol Sci 2003 76 212 9 12970578 116 Weisburger JH Nagao M Wakabayashi K Oguri A Prevention of heterocyclic amine formation by tea and tea polyphenols Cancer letters 1994 83 143 7 8062207 117 Acharyya N Chattopadhyay S Maiti S Chemoprevention against arsenic-induced mutagenic DNA breakage and apoptotic liver damage in rat via antioxidant and SOD1 upregulation by green tea (Camellia sinensis) which recovers broken DNA resulted from arsenic-H2O2 related in vitro oxidant stress J Environ Sci Health C Environ Carcinog Ecotoxicol Rev 2014 32 338 61 25436473 118 Acharyya N Sajed Ali S Deb B Chattopadhyay S Maiti S Green tea (Camellia sinensis) alleviates arsenic-induced damages to DNA and intestinal tissues in rat and in situ intestinal loop by reinforcing antioxidant system Environ Toxicol 2015 30 1033 44 24615952 119 Calatayud M Devesa V Montoro R Velez D In vitro study of intestinal transport of arsenite, monomethylarsonous acid, and dimethylarsinous acid by Caco-2 cell line Toxicology letters 2011 204 127 33 21554932 120 Kim JY Choi JY Lee HJ Byun CJ Park JH Park JH The Green Tea Component (−)-Epigallocatechin-3-Gallate Sensitizes Primary Endothelial Cells to Arsenite-Induced Apoptosis by Decreasing c-Jun N-Terminal Kinase-Mediated Catalase Activity PLoS One 2015 10 e0138590 26375285 121 Lee TC Cheng IC Shue JJ Wang TC Cytotoxicity of arsenic trioxide is enhanced by (−)-epigallocatechin-3-gallate via suppression of ferritin in cancer cells Toxicol Appl Pharmacol 2011 250 69 77 20950636 122 Messarah M Saoudi M Boumendjel A Kadeche L Boulakoud MS El Feki A Green tea extract alleviates arsenic-induced biochemical toxicity and lipid peroxidation in rats Toxicol Ind Health 2013 29 349 59 22301814 123 Nakazato T Ito K Ikeda Y Kizaki M Green tea component, catechin, induces apoptosis of human malignant B cells via production of reactive oxygen species Clin Cancer Res 2005 11 6040 9 16115949 124 Raihan SZ Chowdhury AK Rabbani GH Marni F Ali MS Nahar L Effect of aqueous extracts of black and green teas in arsenic-induced toxicity in rabbits Phytotherapy research: PTR 2009 23 1603 8 19370537 125 Sinha D Bhattacharya RK Siddiqi M Roy M Amelioration of sodium arsenite-induced clastogenicity by tea extracts in Chinese hamster v79 cells J Environ Pathol Toxicol Oncol 2005 24 129 40 15831085 126 Sinha D Dey S Bhattacharya RK Roy M In vitro mitigation of arsenic toxicity by tea polyphenols in human lymphocytes J Environ Pathol Toxicol Oncol 2007 26 207 20 18197836 127 Sinha D Roy M Antagonistic role of tea against sodium arsenite-induced oxidative DNA damage and inhibition of DNA repair in Swiss albino mice J Environ Pathol Toxicol Oncol 2011 30 311 22 22181980 128 Sinha D Roy M Dey S Siddiqi M Bhattacharya RK Modulation of arsenic induced cytotoxicity by tea Asian Pacific journal of cancer prevention: APJCP 2003 4 233 7 14507244 129 Sinha D Roy M Siddiqi M Bhattacharya RK Arsenic-induced micronuclei formation in mammalian cells and its counteraction by tea J Environ Pathol Toxicol Oncol 2005 24 45 56 15715508 130 Sinha D Roy S Roy M Antioxidant potential of tea reduces arsenite induced oxidative stress in Swiss albino mice Food and chemical toxicology: an international journal published for the British Industrial Biological Research Association 2010 48 1032 9 20096321 131 Mizoi M Takabayashi F Nakano M An Y Sagesaka Y Kato K The role of trivalent dimethylated arsenic in dimethylarsinic acid-promoted skin and lung tumorigenesis in mice: tumor-promoting action through the induction of oxidative stress Toxicology letters 2005 158 87 94 16039397
PMC005xxxxxx/PMC5124533.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101622459 42074 J Glob Antimicrob Resist J Glob Antimicrob Resist Journal of global antimicrobial resistance 2213-7165 2213-7173 27693863 5124533 10.1016/j.jgar.2016.08.002 NIHMS818059 Article Host response to Clostridium difficile infection: diagnostics and detection Usacheva Elena A. ab* Jin Jian-P. c Peterson Lance R. ab a Infectious Disease Research, NorthShore University HealthSystem, 2650 Ridge Ave., Evanston, IL 60201, USA b University of Chicago Pritzker School of Medicine, Chicago, IL, USA c Department of Physiology, Wayne State University School of Medicine, Detroit, MI 48201, USA * Corresponding author. Present address: NorthShore University HealthSystem, 2650 Ridge Ave., Walgreen SB #525, Evanston, IL 60201, USA. Tel.: +1 847 570 2063; fax: +1 847 733 5093, [email protected] (E.A. Usacheva) 23 9 2016 20 9 2016 12 2016 01 12 2017 7 93101 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Clostridium difficile infection (CDI) is a significant healthcare concern worldwide, and C. difficile is recognised as the most frequent aetiological agent of infectious healthcare-associated diarrhoea in hospitalised adult patients. The clinical manifestation of CDI varies from self-limited diarrhoea to life-threatening colitis. Such a broad disease spectrum can be explained by the impact of host factors. Currently, a complex CDI aetiology is widely accepted, acknowledging the interaction between bacteria and the host. Clostridium difficile strains producing clostridial toxins A and B are considered toxigenic and can cause disease; those not producing the toxins are non-pathogenic. A person colonised with a toxigenic strain will not necessarily develop CDI. It is imperative to recognise patients with active disease from those only colonised with this pathogen and to implement appropriate treatment. This can be achieved by diagnostics that rely on host factors specific to CDI. This review will focus on major aspects of CDI pathogenesis and molecular mechanisms, describing host factors in disease progression and assessment of the host response in order to facilitate the development of CDI-specific diagnostics. Graphical Abstract Clostridium difficile infection CDI Pathogenesis Host response Biomarker Diagnostics 1. Introduction Clostridium difficile infection (CDI) results from infection of the bowel by C. difficile, a Gram-positive, spore-forming, obligate anaerobic bacterium. Complications of CDI include severe infection, hypotension, shock and sepsis, ileus, megacolon and perforation, or death secondary to CDI [1,2]. Traditional risk factors for CDI include age >65 years, recent hospitalisation, increased length of hospital stay, long-term healthcare facility residence and antibiotic exposure [3–5]. However, only 25% of all cases of antibiotic-associated diarrhoea are associated with CDI [6]. This infectious disease is presently diagnosed at a rate of several hundred thousand cases per year in the USA [7,8]. It has been estimated that for every additional year of age after age 18 years, the risk of healthcare-associated CDI increases by ca. 2% [5]. For the last two decades, CDI has re-emerged in healthcare facilities with nearly a 10-fold increase in mortality [9,10]. It is now rivalling Staphylococcus aureus and vancomycin-resistant enterococci as a cause of nosocomial infection [11]. In the USA, C. difficile was associated with ca. 29 000 deaths in 2011 [4]. In addition to the morbidity and mortality associated with CDI, the US healthcare system expends considerable financial resources for care of this disease [12]. Traditionally, CDI has been considered as a hospital-acquired disease. Currently, however, only 20–25% of all CDI represents disease associated with healthcare exposure [12]. Recent epidemiology suggests the emergence of CDI into new populations having virtually no contact with healthcare settings, including healthy adults, children, pregnant women and patients who have not been subjected to antibiotic therapy [3,13,14]. An additional main challenge in CDI management is the rate of disease recurrence of 15–30% [15]. Critically, there is no specific diagnostic test for CDI. The current diagnostic strategy relies on combining clinical symptoms and signs (such as frequency of diarrhoea, antibiotic exposure and elevated white blood cell count) with a positive diagnostic test for toxigenic C. difficile. Remarkably, this organism can colonise 10 times as many patients asymptomatically than actually develop infection [16]. Combining a high colonisation rate with the fact that 20% of hospitalised patients can have diarrhoea from any cause [17–21] makes it imperative to develop CDI-specific diagnostics that goes beyond simply detecting C. difficile toxin(s)/gene(s). Here we briefly review the current state of understanding of the molecular bases of CDI and efforts to develop a CDI-specific test. 2. Clostridium difficile infection and the host microbiome CDI is characterised by severe alterations in the normal colonic bacterial flora [22]. The human colon is a complex and diverse ecosystem lined by a mucous membrane [23]. It has been postulated that the normal colonic microbiome provides some degree of protection against pathogenic organisms. The mechanism of this protection is incompletely understood but it has been described as ‘colonisation resistance’, with a healthy microbiome making it more difficult for C. difficile to colonise and infect the colon [24]. One of the mechanisms is competition for essential nutrients and attachment sites to the gut wall [25]. The other mechanism refers to bacterial populations within the gastrointestinal tract existing in two forms—free-floating cells and sessile bacteria within mucosa-associated biofilm communities. The sessile bacteria are often comprised of multispecies populations [23,26]. Mucosal communities interact closely with host epithelial cells and may have a greater influence on disease pathogenesis. Alteration of the normal gut microflora or colon microbial diversity (by antibiotic therapy) decreases the level of protection and represents the main risk factor for CDI [21,27]. However, not all classes of antimicrobials an carry equal risk for CDI predisposition [6]. Typically, disease onset occurs 4–9 days after the beginning of antimicrobial treatment [28]. Antibiotics having the highest CDI risk include clindamycin, cephalosporins, penicillins and, more recently, fluoroquinolones. Their frequent usage can increase cumulative antibiotic exposure over time leading to alteration of the indigenous gut microbiota, therefore providing a possibility of C. difficile colonisation and subsequent disease [6,18,29]. There is a belief that failure to restore the normally diverse microbial intestinal community may be related to disease recurrence in patients recovering from CDI [30,31]. Interestingly, a small study showed no difference between the faecal microbiota of asymptomatic C. difficile carriers and healthy subjects, but lower bacterial diversity in patients with C. difficile-related diarrhoea [32]. Animal studies demonstrated that certain bacterial species are present at low frequencies under gut homeostasis but they become more frequent and exhibit many opportunistic properties during dysbiosis [33]. 3. Colonisation Carriage of toxigenic C. difficile in asymptomatic patients is increasingly common. Numerous reports demonstrated a prevalence of asymptomatic C. difficile colonisation during hospital admission as high as 10–15% [19,34–37]. One explanation for such high rates compared with earlier reports is application of more sensitive methods for detection. Clearly, there are some patient populations (infants and the elderly) prone to high rates of C. difficile colonisation [10]. Thus, C. difficile is a frequent component of the faecal microbiota of newborn infants not causing disease [38]. Clostridium difficile colonisation in children is common but severe infection and death are much less frequent than in adults [39]. This might be explained by the fact that in infants the microbiota is insufficiently developed and colonisation resistance is not yet established [40,41]. In one study, the microbial flora composition of infants colonised with C. difficile had increased frequencies of Klebsiella pneumoniae, Ruminococcus gnavus and Clostridium nexile [40]. Their healthy non-colonised counterparts exhibited higher levels of Staphylococcus epidermidis, Escherichia coli and Bifidobacterium longum. It was suggested that specific microflora composition may promote C. difficile colonisation. Interestingly, the study reported that infants were colonised with a single clone of C. difficile for several months [40], but the clone could change as new infant food is introduced. High colonisation rates of 10–50% were also reported in elderly institutionalised adults [42]. Such observations may reflect the comprehensive influence of host factors such as age, co-morbidity, co-administered medications and functional status for disease severity [43]. 4. Pathogenesis of Clostridium difficile infection Clostridium difficile toxin(s) can cause disease within 1 day of the inciting event, usually initial antibiotic therapy, and for up to 2 months after discontinuation of treatment [18,29]. Clostridium difficile spores, the main mode of transmission due to strict anaerobic requirements of the organism, must interact with host epithelial tissue, germinate following interactions with small molecule germinants resulting in vegetative growth of the pathogen, and produce toxin(s) during the cycle of sporulation of a new bacterial generation [44] (Fig. 1). Toxin production and secretion increases after vegetative cells enter into the stationary growth phase [33,39]. A mouse model study demonstrated that during disease, CDI was localised in the large intestine and did not occur in the small intestine [45]. However, it was confirmed that germination of C. difficile spores occurred in the small intestine regardless of antibiotic pre-treatment. The cecum was determined as the site for optimal C. difficile growth, toxin production and disease after antibiotic treatment [45]. It appears that two interconnected events cause manifestation of clinical CDI [46]: (i) epithelial cell intoxication by C. difficile toxins (Fig. 1A); and (ii) inflammation-associated colon tissue damage (Fig. 1B). The latter, in turn, is simultaneously modulated by: (i) toxin–receptor interaction through activation of mucosal immune cells; and (ii) adhesion of C. difficile vegetative cells to the epithelium through non-toxin virulence factors. 4.1. Toxin aetiology of Clostridium difficile infection The main C. difficile toxins are monoglucosyltransferases. Toxin A (TcdA) was classically considered an important enterotoxin but not essential for virulence, whereas toxin B (TcdB) had more potent cytotoxic activity. Whilst some studies re-established the importance of the both toxins suggesting a synergised effect between them [47–49], recent studies clearly indicated that TcdB was the major virulence factor of C. difficile and did not require the presence of TcdA [50,51]. At a clinical level, variant strains of TcdA-negative, TcdB-positive C. difficile were indistinguishable from strains producing both toxins. Patients infected with such strains exhibited the full spectrum of CDI symptoms [50]. Alternatively, variant toxins have alterations in their substrate specificity that may have an impact on disease severity and outcome [50]. To date there have been no confirmed C. difficile strains that solely produce TcdA known to cause CDI [52]. However, there is evidence that the prevalence of clinically relevant CDI cases due to TcdA-negative, TcdB-positive strains has increased globally [53,54]. During cell infection, the toxins are subject to time-dependent degradation due to proteolysis and pH effects [50]. They modify the activity of members of the host Rho family of small GTPases [55], key cellular regulatory proteins, leading to actin filament depolymerisation, cytoskeletal disruption and subsequent intestinal epithelial cell death (Fig. 1A) [56–58]. Recent histopathological analysis of caecal and colonic tissues collected from infected mice showed that TcdB caused the majority of intestinal damage during infection, with TcdA causing more superficial and localised damage [51]. Besides the well characterised TcdA and TcdB toxins, a binary toxin CDT has been identified [59,60]. CDT is an ADP-ribosyltransferase that has been shown to disrupt the cytoskeleton of the cell, leading to cell rounding, loss of fluids and cell death [61,62]. It is a suspected C. difficile virulence factor shown to be involved in the formation of long microtubule protrusions in the host cell, facilitating bacterial attachment [63]. CDT is produced by some but not all toxigenic C. difficile strains and, in contrast to TcdA and TcdB, plays a minor role in CDI [51]. The CDT-positive but TcdA- and TcdB-negative strains were avirulent in the hamster infection model [64]. The contribution of the binary toxin to human disease is still being elucidated. 4.2. Inflammatory response An essential first step for CDI is colonisation of host mucosal surfaces [65]. The disease is characterised by tissue injury and an acute intestinal inflammatory response highlighted by neutrophil infiltration [66] (Fig. 1C). A study in mice demonstrated that the primary cellular immune response to toxin was oedema and polymorphonuclear cell infiltration [67]. The intense immune activation results in the endoscopic findings of the ‘volcano lesion’ and pseudomembranes. Inflammation-associated tissue damage is thought to be secondary to the intoxication of intestinal epithelial cells in CDI pathogenesis [52,68]. Following breakdown of the intestinal epithelial barrier, immune cells within the mucosa are activated by TcdA and TcdB, eventually leading to the release of inflammatory mediators [pro-inflammatory cytokines interleukin-1β (IL-1β), tumour necrosis factor-α (TNFα) and IL-8 from activated macrophages] [52,69] (Fig. 1B). The central role of inflammation in CDI pathogenesis is highlighted by the fact that the magnitude of the inflammatory response is the best predictor of CDI poor outcome, but not the overall bacteria/toxin burden within the intestine [70]. Of importance is the host immune system, as evidenced by the higher rates of infection and worsening disease severity among the elderly and other persons who lack the ability to mount an effective humoral immune response [5,66]. Therefore, all current data on CDI are moving towards emphasising a link between bacteria–host cellular immune interaction, homeostasis and the gut microbiome [69]. 5. Clostridium difficile infection diagnostics There is renewed interest in the development and validation of clinical prediction tools for CDI. Currently, laboratory testing available for CDI identification include sigmoidoscopy and colonoscopy, toxigenic culture (TC), cell cytotoxicity assay (CCTA), enzyme immunoassay (EIA), glutamate dehydrogenase (GDH) EIA, as well as real-time PCR (rtPCR) and loop-mediated isothermal amplification (LAMP) assay, which is less costly but uncommon in the clinical laboratory environment [71]. Together, clinical prediction algorithms and reliable laboratory diagnostics will greatly facilitate early diagnosis. Yet a number of CDI studies demonstrated that host response determines the character of the intestinal inflammation and clinical severity. 5.1. Bacteria detection Currently there is no ‘gold standard’ for the diagnosis of CDI, other than possible direct visualisation of characteristic lesions using colonoscopy [72]. The two laboratory reference methods most commonly cited are the stool CCTA, which detects the presence of ‘free’ C. difficile toxins, and TC, which detects C. difficile isolates producing toxins in vitro with the potential for producing toxins in vivo [20,73]. However, it must be noted that these methods only detect the presence of a disease-causing organism and when positive are not diagnostic of clinical infection. Of note, a recent large multicentre study of CDI reported CCTA as the best diagnostic indicator for CDI disease [20,22], however the sensitivity of this test can be as low as 50% when TC is used as the reference method [74]. CCTA appears to have a sensitivity of only 89% in cases where pseudomembranes are seen on colonoscopy [29]. A positive TC may still indicate a patient who is an infection risk to others [20]. Yet it should be clear that the current diagnosis requires clinical judgement along with the positive result of a diagnostic test. Also, TC and CCTA are laborious, time consuming, and suffer from suboptimal specificity and/or sensitivity [38,75]. Newer rapid nucleic acid amplification tests (NAATs) formatted in rtPCR and LAMP for detection of TcdB and TcdA genes offer improved sensitivity over immunoassays [76,77]. However, concerns relate to the biology of C. difficile and how detection of the genes correlates with expression of the toxins [38]. NAATs yield more positives than CCTA [20]. A number of studies confirmed that molecular detection of CDI is very sensitive although less specific, therefore leading to overdiagnosis of CDI [74,78]. Two meta-analyses of the performance of rtPCR reported its high sensitivity and specificity, however they were highly dependent on CDI prevalence [71,79]. According to Deshpande et al., the negative predictive value (NPV) of rtPCR was acceptable at a C. difficile prevalence of <10% [79]. This suggests the assay may serve as an effective screening test in endemic situations, with emphasis on a negative test result. However, Lloyd et al. reported the possibility that the LAMP assay may be more sensitive and specific than rtPCR [80]. The positive predictive value (PPV) and NPV for LAMP were better than rtPCR in settings where the CDI prevalence was <15%. It was suggested that rtPCR may be more suitable in epidemic conditions with a higher prevalence, and LAMP for settings with a lower C. difficile prevalence [80]. Nevertheless, the diagnosis of CDI either with rtPCR or LAMP should only be made in the presence of strong clinical symptoms consistent with CDI. This is highlighted by the fact that stool PCR tests remain positive for C. difficile for up to 30 days after successful treatment [81]. Whilst a negative result is adequate to rule out the presence of the disease [71], a positive stool test does not distinguish colonised patients from those with symptomatic disease [82]. Currently, 13 commercial NAATs have been approved by the US Food and Drug Administration (FDA) [83]. Some of these use PCR techniques, whilst the others utilise LAMP or helicase-dependent amplification method for detection of C. difficile toxins or regulatory genes [77]. The TcdB gene is usually chosen as a NAAT target since it is produced by almost all toxigenic C. difficile strains [60]. The TcdA gene is less frequently used because ca. 3% of European toxigenic strains have been reported to be TcdA-negative [84]. A higher prevalence of such strains has been reported throughout Asia [54,77]. However, a discussion of the NAATs for C. difficile target detection is beyond the scope of this review. The NAATs fail to discriminate between CDI and C. difficile asymptomatic colonisation. Therefore, colonised individuals may test falsely positive for CDI when evaluated for community-acquired diarrhoea caused by other enteric pathogens [16]. Also, the emergence of new C. difficile strains with altered toxins or genes can impact all currently existing CDI diagnostics [85]. This highlights the need for a test beyond simply detecting toxin(s) or gene(s). Of note, there is great variation between studies in the characteristics of C. difficile tests, suggesting possible, yet unidentified, human-related or microbe-related factors affecting test performance [20]. Therefore, it is invaluable to develop CDI diagnostics that include a biomarker correlated with active infection [38,82], and the most effective would be the one that relies on specific human response to CDI. 5.2. Host response There is growing evidence demonstrating the contribution of the host response (immune and inflammatory) to CDI outcome. Yet to date an assay for CDI host response has not been established, as well as an optimal assay to supplant or be used with other rapid testing. 5.2.1. Biological markers As noted earlier, the presence of C. difficile toxins stimulates a multifaceted immune response involving cytokines, chemokines and mucosal immune cells [65], eventually activating antibody production. The levels of such antibodies in serum or stool could be potential CDI biomarkers, however study results to date are complex. Approximately 60% of healthy older children and adults have detectable serum immunoglobulin G (IgG) and IgA antibodies to TcdA and TcdB even in the absence of C. difficile colonisation or active infection [66]. The level of anti-TcdA and anti-TcdB antibodies has been considered important in determining whether colonisation or clinical infection follows C. difficile spore acquisition [86]. Reduced anti-TcdA levels at the start of infection have been linked both to recurrence and increased 30-day mortality. However, Loo et al. did not find a significant association between levels of antibodies against TcdA and TcdB at the time of admission and subsequent healthcare-associated CDI [5,35]. The surface layer proteins (SLPs) are the outermost protein component of C. difficile responsible for adhesion to host tissue, known to be variable between strains, and may play an important role in intestinal colonisation and the persistence of CDI [65]. A study revealed that the levels of antibody to SLPs were similar in patients with CDI, asymptomatic carriers and controls [87]. In another study, IgG levels to SLPs were similar in cases and carriers but were higher compared with controls [88]. Of note, it has been shown that strains that adhere better to human intestinal cell lines proved to be more virulent in hamsters [89]. Concentrations of a systemic inflammation biomarker (C-reactive protein), white blood cells, serum creatinine, serum 25-hydroxyvitamin D and albumin have been also reported to be associated with CDI severity and possible independent mortality predictors [90–92]. However, these markers are non-specific to C. difficile disease [39,92]. It has been suggested that C. difficile is initially recognised in a Toll-like receptor 4- and MyD88-dependent manner, resulting in low serum expression of IL-23 [46]. Intoxication of primed host cells by TcdA and TcdB leads to robust IL-1β secretion that further enhances IL-23 production. However, increased serum IL-1β was also noted in patients with diarrhoea from other causes [46]. Therefore, these do not appear to be specific for CDI. Procalcitonin, a biomarker for bacterial infection, was evaluated in association with CDI severity [93]. In this small study, serum procalcitonin levels did show some promise as a biomarker for CDI severity. Further studies in a larger cohort need to be done. 5.2.2. Faecal biomarkers Faecal proteins are ideal biomarkers for gastrointestinal inflammation owing to direct contact with the intestinal mucosa (containing a large number of neutrophils) and the non-invasive sampling mode [94]. Faecal markers include a biologically heterogeneous group of substances that either leak from or are actively released by the inflamed mucosa [95]. Over the last several decades, a few faecal inflammation biomarkers have been investigated. Some of these markers have been shown to be produced in response to C. difficile toxins and therefore have been investigated as predictors of CDI disease severity (Table 1). A few publications have indicated that levels of faecal cytokines could be indicators of CDI severity. El Feghaly reported that the cytokines CXCL-5 and IL-8 as well as lactoferrin (LF) were more sensitive in identifying patients at risk than clinical parameter or score [70,96], but they were not specific for disease caused solely by C. difficile. The conclusion in children was that faecal inflammatory cytokines differentiate asymptomatic C. difficile colonisation from disease and are associated with CDI poor outcome [96]. To date, however, there are no commercial assays available for measurement of cytokines in stool. 5.2.2.1. Lactoferrin (LF) Multiple studies have validated faecal LF measurement as an accurate marker of intestinal inflammation and its usefulness for inflammatory diarrhoea [94,97–99]. This glycoprotein is resistant to proteolysis, is not degraded by C. difficile toxins, and is released following neutrophil activation. LF concentrations in stool and other fluids are proportional to the number of neutrophils recruited [98,100]. Van Langerberg et al. showed that a normal LF level effectively excludes inflammatory diarrhoea, therefore it was proposed as a screening test prior to microbiological assessment of faeces [98]. Faecal LF is not useful in discriminating between inflammatory diarrhoea caused by bacterial infection and non-inflammatory diarrhoea in children as well. A study of 1000 stools submitted for diagnostic testing found the sensitivity and specificity of LF detection to be only 75% and 60%, respectively [85], indicating that it is not a useful test for clinical evaluation of patients with potential CDI. 5.2.2.2. Calprotectin (CP) CP, a protein resistant to bacterial degradation, has been found within the cytosol of neutrophils, accounting for ca. 60% of their cytoplasmic proteins [101]. Under inflammatory conditions of the intestinal tract, CP is excreted in stool [100,101] and can be measured by commercial assays. To differentiate bowel diseases, some cut-off values of faecal CP have been established. However, other causes of inflammation, such as infection with enteric pathogens or disease from non-steroidal anti-inflammatory agents, have to be considered since they may raise the CP level [100]. Currently there is no reference method or standard for faecal CP [100] other than endoscopy, the ‘gold standard’ for assessing intestinal inflammation. Overall, CP or LF as faecal markers have emerged as new diagnostic tools to detect and monitor intestinal inflammation. Evaluation of the accuracy of faecal LF and CP as well as faecal occult blood testing showed statistically enhanced specificity and positive and negative likelihood ratios only for CP in predicting infectious diarrhoea [102]. Of note, LF and CP have been reported be able to differentiate inflammatory disease from functional bowel disorders [94]. Measuring both faecal LF and CP did not benefit detection [100]. Neither of these assays would differentiate CDI from other causes of diarrhoea. 5.2.2.3. pMK2 Clostridium difficile toxins activate the p38 pathway and its downstream kinase target MK2 [103]. Whilst p38 protein was suggested as a major regulator and essential to C. difficile toxin-induced inflammation, MK2 kinase (the p38 substrate) was proposed to be specifically involved in stress-induced inflammation. Phosphorylated MK2 (pMK2) phosphorylates specific molecules that regulate the actin cytoskeleton and stabilise cytokine mRNA transcripts. Tested stool specimens collected from CDI patients demonstrated that an elevated pMK2 level was significantly associated with the presence of toxigenic C. difficile [103]. However, like inflammatory cytokine levels, faecal pMK2 may be a general indicator of intestinal inflammation. It was also shown that MK2 can be activated by Shiga toxin and during influenza A virus infection [104]. 5.2.2.4. Phosphorylated p38 (pp38) pp38 has been tested as a biomarker for symptomatic CDI in the paediatric population [96]. Elevation of pp38 in the stools of children with CDI compared with their symptomatic controls was demonstrated. Consistent with previous data [103], the p38 pathway was suggested as specific for C. difficile-associated injury. Of note, in this study C. difficile bacterial burden was not associated with any clinical outcomes. Yet despite the observation that pp38 might be specific for C. difficile infection, it lacked sensitivity [96]. 5.2.2.5. Interleukin-23 Thirty-six major inflammatory markers present in the stools were examined in CDI and non-CDI patients [105]. CDI-positive stools exhibited significantly higher relative amounts of C5a, CD40L, granulocyte colony-stimulating factor (G-CSF), I-309, IL-13, IL-16, IL-27, monocyte chemoattractant protein 1 (MCP-1), TNFα and IL-8. However, the study suggested the importance of IL-8 and IL-23 in CDI immunopathogenesis. The relative amount of IL-23 was significantly higher in CDI-negative stools than in CDI-positive stools. On the other hand, the average concentration of IL-8 in CDI-positive stools was significantly higher than in CDI-negative stools. These two cytokines were detected in more CDI-positive stools than CDI-negative stools [105]. A novel proposed marker is justified by the fact that destruction of the actin cytoskeleton by C. difficile toxins results in accelerated dissociation of colonic epithelial cells leading to cell death (Fig. 1) [106]. The hypothesis is that the release of cytoskeleton contents of colon epithelial cells could serve as a specific indicator for the host response in CDI. To demonstrate this novel approach, we have obtained preliminary results showing the feasibility of detecting human non-muscle tropomyosin, a major cytoskeleton protein of colon epithelial cells, released into patients’ stool samples. The preliminary results presented a promising correlation with the presence of CDI [107]. The biomarker is currently under investigation, supported by the National Institutes of Health (NIH) (grant R21AI116659). 6. ‘Gold standard’ for Clostridium difficile infection host response assay Currently, there is no reliable single standing test for CDI diagnostics. The question raises what could be the choice of reference standard assay in development of a new CDI diagnostic test. The standard’s performance is crucial for defining true positives and true negatives. Use of suboptimal tests can blur CDI diagnosis owing to difficulty of distinguishing C. difficile and other infective or non-infective causes of diarrhoea [108]. CCTA has been traditionally chosen as the ‘gold standard’ assay for CDI confirmation, providing the final result in a day from sample submission [109,110], albeit with a sensitivity of <90% even in the presence of documented colitis [29]. Some investigators consider TC as more reliable for CDI although it takes 4–5 days. From a number of studies, the conclusion is drawn that TC detects more positive samples than CCTA [108], however the specificity is diminished. Planche and Wilcox discussed the importance of the fact that these two methods detect different targets [108]. CCTA detects the presence of C. difficile toxins, whilst TC detects C. difficile bacteria or spores. A faecal sample may be CCTA-negative but TC-positive. Conversely, a TC-positive result may occur in the absence of CDI. Interestingly, studies reported that only 90% [111] or 50% [112] of pseudomembranous colitis patients were CCTA-positive. Therefore, the use of a single laboratory test should not be considered sufficient as a reference standard for new testing to detect CDI. Ideally, the accuracy of laboratory diagnostic tests should be measured using reference assays utilising the same or equivalent targets [108]. On discrepant samples the use of assays with different targets likely will not improve assessment of the true accuracy of a diagnostic test. On note, a meta-analysis of the accuracy of rtPCR for CDI did not reveal differences in diagnostic sensitivity and specificity by the type of reference standard (CCTA/TC) [79]. In general, patients who are TC-positive but CCTA-negative appear different from those who are CCTA-positive [29]. Whilst a significant increase in the sensitivity of CDI detection is observed when TC is used, such advantage is diminished in regard to poor specificity of culture-based diagnosis secondary to C. difficile carriage [108]. A positive TC does not necessarily confirm CDI as the cause of diarrhoea. Therefore, none of these two methods are absolute for CDI diagnostics. If both reference tests were performed, a positive CCTA would confirm CDI and might be useful for the positive reference standard. Faecal samples negative by CCTA but positive by culture may be difficult to categorise, but it appears that many of these cases do not have CDI and it would be reasonable to exclude these from consideration of validation for a new diagnostic test. Therefore, clinical assessment of such cases would be important. Is rtPCR an alternative? The rtPCR-based tests potentially yield false-positive results, demonstrating moderate specificity and PPVs owing to their high sensitivity and the potential for detecting colonised, but not clinically infected, persons leading to overdiagnosis of CDI. The reported sensitivity and specificity has ranged from 83% to 94% and from 97% to 98%, respectively. Importantly, this approach is not designed to detect the disease but rather the causative agent, and performance depends on the disease prevalence. Therefore, it is not a reliable choice for ‘gold standard’ as a stand-alone test. For years, rapid and simple commercial EIAs detecting clostridial toxin(s) were the most frequently used CDI assays in clinical laboratories. However, their well confirmed characteristic is low sensitivity. The sensitivity and specificity of EIAs ranged from 32% to 83% and from 84% to 100%, respectively [20,74,113]. Thus, EIA is suboptimal for diagnosis of CDI. In addition, it has been demonstrated that EIAs are unable to detect some newer C. difficile strains, including epidemic clones [114]. The performance of toxin immunoassays varies markedly across manufacturers [110]. Therefore, the overall poor performance of toxin EIAs led to the recommendation to use them only as a part of a two- or three-stage algorithm [58,110]. Thus, these tests also are not a reliable choice for ‘gold standard’ as a stand-alone test. Another EIA, for GDH (a protein found on most C. difficile isolates), is more sensitive (76–100%) but less specific. For this reason, GDH is only recommended to be used in combination with other assays, such as a toxin immunoassay. Taken together, the conclusion is that EIAs cannot serve as ‘gold standard’ as well. Summarising from above, it is reasonable to hypothesise that validation of a new diagnostic test for true CDI, when linked to host response target, should rely on the combined outcome of two currently available laboratory diagnostic tests and relevant clinical signs. The CCTA should be the first choice as reference method for determining CDI cases, and a positive assay should be consistent with the diagnosis. Since the assay for detection of host response does not have an equivalent target as in CCTA, samples positive for TC and negative by CCTA should be excluded in the analysis. Those patients with both TC and CCTA tests negative should be considered as not having CDI. This will provide a reference standard where true positive and negative patients are recognised and those with an uncertain diagnosis are excluded. 7. Conclusion CDI is a globally important yet poorly diagnosed infectious disease. Current CDI diagnostics is limited to detection of the organism and/or its toxin product(s) in conjunction with clinical symptoms, not differentiating infected from colonised patients, thus leading to inaccurate diagnosis and antibiotic mis/overuse. This review outlines major knowledge of CDI molecular aetiology to form a holistic view of the disease and to advocate the development of specific and accurate diagnostics for CDI. It is now clear that there is a close link between pathogen and host response that controls disease progression and outcome. Therefore, two clear needs seem obvious for moving forward with this disease. One is the identification of a biomarker that is able to measure the effects of C. difficile toxin(s) on human colonic tissue. In addition to clinical signs and symptoms, the utility of a biomarker will significantly enhance the existing CDI diagnostic tools that entirely rely on organism detection. The second is to establish a reasonable set of criteria for the ‘gold standard’ of CDI diagnosis. In conclusion, in addition to pathogenic toxin production, the composition and function of the intestinal microbiome and host immune factors have direct impacts on C. difficile pathogenesis. To improve the care of patients with potential CDI, there remains a critical need for the optimal diagnostic test approach to this persistent infectious disease. Funding: This work was supported by an investigator-initiated grant from the NIH National Institute of Allergy and Infectious Diseases [NIH/NIAID grant R21AI116659]. Fig. 1 Complexity of the host response to Clostridium difficile infection (CDI). (A) Intoxication of host epithelial cells by C. difficile toxins produced by vegetative cells is primary to inflammatory response. Toxin production and secretion increases after vegetative cells enter into the stationary growth phase [22,39]. During cell infection, the toxins are subjects to time-dependent degradation due to proteolysis and pH effects. The toxin’s entry into the intestinal epithelial cells is one of the earliest pathogenic events. It leads to loss of structural integrity (actin skeleton disruption, disruption of tight junctions, reduced cell–cell contact), cell death and epithelium disruption. TLR, Toll-like receptor. (B) Inflammatory response by two mechanisms: (i) secondary to toxin intoxication (within a few hours after toxin exposure); and (ii) activation of intracellular cascades by non-toxin virulence factors such as surface layer proteins (SLPs), flagellar proteins (FliC and FliD), adhesins (cwp66, cwpV), fibronectin-binding proteins and cell surface polysaccharides [22]. (C) CDI clinical manifestation. The inflammatory response causes tissue damage: neutrophil accumulation (one of the major mechanisms) is responsible for pseudomembrane formation seen in severe colitis [38]; diarrhoea, toxic megacolon. Toxin B can also cause multiple organ dysfunction syndrome due to systemic toxin damage [50,68]. Table 1 Summary of faecal inflammatory biomarkers as possible predictors of Clostridium difficile infection (CDI) disease Biomarker Clinical indication/prediction Role in immunopathogenesis Specific/sensitive to CDI References Lactoferrin Colonic inflammation, CDI severity (when level is elevated) Innate inflammatory response; related to level of neutrophil translocation No/no [97,99–102,107] Calprotectin Intestinal inflammatory conditions (when level is elevated) Innate inflammatory response; correlates with level of released neutrophils No/no [103,105,106] IL-8 CDI severity (when elevated) Involved in the recruitment of neutrophils to sites of infection no/yes [73,99,109,107] IL-23 May relate to CDI recurrence (when level is decreased) Lack of a robust immunological response No/no [109] pMK2 Presence of toxigenic C. difficile (when level is elevated) Key mediator of p38-dependent inflammation No/– [107] pp38 Symptomatic CDI in paediatrics (when level is elevated) Activation of p38 protein pathway Yes/no [99] Highlights High rate of colonisation with Clostridium difficile. Toxin aetiology and inflammatory response are interconnected events in C. difficile infection (CDI). Bacteria detection versus biological marker for CDI diagnostics. Host response to CDI. Competing interests: None declared. Ethical approval: Not required. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. 1 Cohen SH Gerding DN Johnson S Kelly CP Loo VG McDonald LC Clinical practice guidelines for Clostridium difficile infection in adults: 2010 update by the Society for Healthcare Epidemiology of America (SHEA) and the Infectious Diseases Society of America (IDSA) Infect Control Hosp Epidemiol 2010 31 431 55 20307191 2 Burke KE Lamont JT Clostridium difficile infection: a worldwide disease Gut Liver 2014 8 1 6 24516694 3 Khanna S Pardi DS Clostridium difficile infection: management strategies for a difficult disease Therap Adv Gastroenterol 2014 7 72 86 4 Lessa FC Mu Y Bamberg WM Beldavs ZG Dumyati GK Dunn JR Burden of Clostridium difficile infection in the United States N Engl J Med 2015 372 825 34 25714160 5 Loo VG Bourgault AM Poirier L Lamothe F Michaud S Turgeon N Host and pathogen factors for Clostridium difficile infection and colonization N Engl J Med 2011 365 1693 703 22047560 6 Britton RA Young VB Interaction between the intestinal microbiota and host in Clostridium difficile colonization resistance Trends Microbiol 2012 20 313 9 22595318 7 Karas JA Enoch DA Aliyu SH A review of mortality due to Clostridium difficile infection J Infect 2010 61 1 8 20361997 8 Steiner C Barrett M Weiss A HCUP projections: Clostridium difficile hospitalizations 2001 to 2013. HCUP projections report # 2014-01 US Agency for Healthcare Research and Quality 2014 9 Kochanek KD Murphy SL Xu J Deaths: final data for 2011 National Vital Statistics Reports 2015 63 10 Lessa FC Gould CV McDonald LC Current status of Clostridium difficile infection epidemiology Clin Infect Dis 2012 55 Suppl 2 S65 70 22752867 11 Miller BA Chen LF Sexton DJ Anderson DJ Comparison of the burdens of hospital-onset, healthcare facility-associated Clostridium difficile infection and of healthcare-associated infection due to methicillin-resistant Staphylococcus aureus in community hospitals Infect Control Hosp Epidemiol 2011 32 387 90 21460491 12 Tabak YP Zilberberg MD Johannes RS Sun X McDonald LC Attributable burden of hospital-onset Clostridium difficile infection: a propensity score matching study Infect Control Hosp Epidemiol 2013 34 588 96 23651889 13 Jarvis WR Schlosser J Jarvis AA Chinn RY National point prevalence of Clostridium difficile in US health care facility inpatients, 2008 Am J Infect Control 2009 37 263 70 19278754 14 Zilberberg MD Tillotson GS McDonald C Clostridium difficile infections among hospitalized children, United States, 1997–2006 Emerg Infect Dis 2010 16 604 9 20350373 15 Drekonja D Reich J Gezahegn S Greer N Shaukat A MacDonald R Fecal microbiota transplantation for Clostridium difficile infection: a systematic review Ann Intern Med 2015 162 630 8 25938992 16 Galdys AL Nelson JS Shutt KA Schlackman JL Pakstis DL Pasculle AW Prevalence and duration of asymptomatic Clostridium difficile carriage among healthy subjects in Pittsburgh, Pennsylvania J Clin Microbiol 2014 52 2406 9 24759727 17 Teasley DG Gerding DN Olson MM Peterson LR Gebhard RL Schwartz MJ Prospective randomised trial of metronidazole versus vancomycin for Clostridium-difficile-associated diarrhoea and colitis Lancet 1983 2 1043 6 6138597 18 Johnson S Clabots CR Linn FV Olson MM Peterson LR Gerding DN Nosocomial Clostridium difficile colonisation and disease Lancet 1990 336 97 100 1975332 19 Alasmari F Seiler SM Hink T Burnham CA Dubberke ER Prevalence and risk factors for asymptomatic Clostridium difficile carriage Clin Infect Dis 2014 59 216 22 24755858 20 Planche T Wilcox MH Diagnostic pitfalls in Clostridium difficile infection Infect Dis Clin North Am 2015 29 63 82 25595842 21 Young VB Schmidt TM Antibiotic-associated diarrhea accompanied by large-scale alterations in the composition of the fecal microbiota J Clin Microbiol 2004 42 1203 6 15004076 22 Planche TD Davies KA Coen PG Finney JM Monahan IM Morris KA Differences in outcome according to Clostridium difficile testing method: a prospective multicentre diagnostic validation study of C difficile infection Lancet Infect Dis 2013 13 936 45 24007915 23 van der Waaij LA Harmsen HJ Madjipour M Kroese FG Zwiers M van Dullemen HM Bacterial population analysis of human colon and terminal ileum biopsies with 16S rRNA-based fluorescent probes: commensal bacteria live in suspension and have no direct contact with epithelial cells Inflamm Bowel Dis 2005 11 865 71 16189415 24 Bakken JS Fecal bacteriotherapy for recurrent Clostridium difficile infection Anaerobe 2009 15 285 9 19778623 25 Nord CE Kager L Heimdahl A Impact of antimicrobial agents on the gastrointestinal microflora and the risk of infections Am J Med 1984 76 99 106 26 Zoetendal EG von Wright A Vilpponen-Salmela T Ben-Amor K Akkermans AD de Vos WM Mucosa-associated bacteria in the human gastrointestinal tract are uniformly distributed along the colon and differ from the community recovered from feces Appl Environ Microbiol 2002 68 3401 7 12089021 27 Owens RC Jr Donskey CJ Gaynes RP Loo VG Muto CA Antimicrobial-associated risk factors for Clostridium difficile infection Clin Infect Dis 2008 46 Suppl 1 S19 31 18177218 28 Lyras D O’Connor JR Howarth PM Sambol SP Carter GP Phumoonna T Toxin B is essential for virulence of Clostridium difficile Nature 2009 458 1176 9 19252482 29 Gerding DN Olson MM Peterson LR Teasley DG Gebhard RL Schwartz ML Clostridium difficile-associated diarrhea and colitis in adults. A prospective case-controlled epidemiologic study Arch Intern Med 1986 146 95 100 3942469 30 Dethlefsen L Huse S Sogin ML Relman DA The pervasive effects of an antibiotic on the human gut microbiota, as revealed by deep 16S rRNA sequencing PLoS Biol 2008 6 e280 19018661 31 Dethlefsen L Relman DA Incomplete recovery and individualized responses of the human distal gut microbiota to repeated antibiotic perturbation Proc Natl Acad Sci U S A 2011 108 Suppl 1 4554 61 20847294 32 Rea MC O’Sullivan O Shanahan F O’Toole PW Stanton C Ross RP Clostridium difficile carriage in elderly subjects and associated changes in the intestinal microbiota J Clin Microbiol 2012 50 867 75 22162545 33 Peniche AG Savidge TC Dann SM Recent insights into Clostridium difficile pathogenesis Curr Opin Infect Dis 2013 26 447 53 23982235 34 Hutin Y Casin I Lesprit P Welker Y Decazes JM Lagrange P Prevalence of and risk factors for Clostridium difficile colonization at admission to an infectious diseases ward Clin Infect Dis 1997 24 920 4 9142793 35 Kyne L Warny M Qamar A Kelly CP Asymptomatic carriage of Clostridium difficile and serum levels of IgG antibody against toxin A N Engl J Med 2000 342 390 7 10666429 36 Leekha S Aronhalt KC Sloan LM Patel R Orenstein R Asymptomatic Clostridium difficile colonization in a tertiary care hospital: admission prevalence and risk factors Am J Infect Control 2013 41 390 3 23622704 37 Samore MH DeGirolami PC Tlucko A Lichtenberg DA Melvin ZA Karchmer AW Clostridium difficile colonization and diarrhea at a tertiary care hospital Clin Infect Dis 1994 18 181 7 8161624 38 Burnham CA Carroll KC Diagnosis of Clostridium difficile infection: an ongoing conundrum for clinicians and for clinical laboratories Clin Microbiol Rev 2013 26 604 30 23824374 39 Martin J The contribution of strains and hosts to outcomes in Clostridium difficile infection Infect Dis Clin North Am 2015 29 51 61 25582645 40 Rousseau C Poilane I De Pontual L Maherault AC Le Monnier A Collignon A Clostridium difficile carriage in healthy infants in the community: a potential reservoir for pathogenic strains Clin Infect Dis 2012 55 1209 15 22843784 41 Donskey CJ Kundrapu S Deshpande A Colonization versus carriage of Clostridium difficile Infect Dis Clin North Am 2015 29 13 28 25595843 42 Riggs MM Sethi AK Zabarsky TF Eckstein EC Jump RL Donskey CJ Asymptomatic carriers are a potential source for transmission of epidemic and nonepidemic Clostridium difficile strains among long-term care facility residents Clin Infect Dis 2007 45 992 8 17879913 43 Aronoff DM Editorial commentary: host–pathogen interactions in Clostridium difficile infection: it takes two to tango Clin Infect Dis 2014 58 1401 3 24604896 44 Paredes-Sabja D Shen A Sorg JA Clostridium difficile spore biology: sporulation, germination, and spore structural proteins Trends Microbiol 2014 22 406 16 24814671 45 Koenigsknecht MJ Theriot CM Bergin IL Schumacher CA Schloss PD Young VB Dynamics and establishment of Clostridium difficile infection in the murine gastrointestinal tract Infect Immun 2015 83 934 41 25534943 46 Cowardin CA Kuehne SA Buonomo EL Marie CS Minton NP Petri WA Jr Inflammasome activation contributes to interleukin-23 production in response to Clostridium difficile MBio 2015 6 pii: e02386-14 47 Kuehne SA Cartman ST Heap JT Kelly ML Cockayne A Minton NP The role of toxin A and toxin B in Clostridium difficile infection Nature 2010 467 711 3 20844489 48 Hirota SA Iablokov V Tulk SE Schenck LP Becker H Nguyen J Intrarectal instillation of Clostridium difficile toxin A triggers colonic inflammation and tissue damage: development of a novel and efficient mouse model of Clostridium difficile toxin exposure Infect Immun 2012 80 4474 84 23045481 49 Kuehne SA Collery MM Kelly ML Cartman ST Cockayne A Minton NP Importance of toxin A, toxin B, and CDT in virulence of an epidemic Clostridium difficile strain J Infect Dis 2014 209 83 6 23935202 50 Carter GP Rood JI Lyras D The role of toxin A and toxin B in the virulence of Clostridium difficile Trends Microbiol 2012 20 21 9 22154163 51 Carter GP Chakravorty A Pham Nguyen TA Mileto S Schreiber F Li L Defining the roles of TcdA and TcdB in localized gastrointestinal disease, systemic organ damage, and the host response during Clostridium difficile infections MBio 2015 6 e00551 26037121 52 Sun X Hirota SA The roles of host and pathogen factors and the innate immune response in the pathogenesis of Clostridium difficile infection Mol Immunol 2015 63 193 202 25242213 53 Cairns MD Preston MD Lawley TD Clark TG Stabler RA Wren BW Genomic epidemiology of a protracted hospital outbreak caused by a toxin A-negative Clostridium difficile sublineage PCR ribotype 017 strain in London, England J Clin Microbiol 2015 53 3141 7 26179308 54 Shin BM Kuak EY Yoo HM Kim EC Lee K Kang JO Multicentre study of the prevalence of toxigenic Clostridium difficile in Korea: results of a retrospective study 2000–2005 J Med Microbiol 2008 57 697 701 18480325 55 Just I Selzer J Wilm M von Eichel-Streiber C Mann M Aktories K Glucosylation of Rho proteins by Clostridium difficile toxin B Nature 1995 375 500 3 7777059 56 Voth DE Ballard JD Clostridium difficile toxins: mechanism of action and role in disease Clin Microbiol Rev 2005 18 247 63 15831824 57 Chumbler NM Farrow MA Lapierre LA Franklin JL Haslam DB Goldenring JR Clostridium difficile toxin B causes epithelial cell necrosis through an autoprocessing-independent mechanism PLoS Pathog 2012 8 e1003072 23236283 58 Rupnik M Heterogeneity of large clostridial toxins: importance of Clostridium difficile toxinotypes FEMS Microbiol Rev 2008 32 541 55 18397287 59 Stubbs S Rupnik M Gibert M Brazier J Duerden B Popoff M Production of actin-specific ADP-ribosyltransferase (binary toxin) by strains of Clostridium difficile FEMS Microbiol Lett 2000 186 307 12 10802189 60 Geric B Rupnik M Gerding DN Grabnar M Johnson S Distribution of Clostridium difficile variant toxinotypes and strains with binary toxin genes among clinical isolates in an American hospital J Med Microbiol 2004 53 887 94 15314196 61 Shen A Clostridium difficile toxins: mediators of inflammation J Innate Immun 2012 4 149 58 22237401 62 Sundriyal A Roberts AK Ling R McGlashan J Shone CC Acharya KR Expression, purification and cell cytotoxicity of actin-modifying binary toxin from Clostridium difficile Protein Expr Purif 2010 74 42 8 20433927 63 Schwan C Stecher B Tzivelekidis T van Ham M Rohde M Hardt WD Clostridium difficile toxin CDT induces formation of microtubule-based protrusions and increases adherence of bacteria PLoS Pathog 2009 5 e1000626 19834554 64 Geric B Carman RJ Rupnik M Genheimer CW Sambol SP Lyerly DM Binary toxin-producing, large clostridial toxin-negative Clostridium difficile strains are enterotoxic but do not cause disease in hamsters J Infect Dis 2006 193 1143 50 16544255 65 Vedantam G Clark A Chu M McQuade R Mallozzi M Viswanathan VK Clostridium difficile infection: toxins and non-toxin virulence factors, and their contributions to disease establishment and host response Gut Microbes 2012 3 121 34 22555464 66 Kelly CP Kyne L The host immune response to Clostridium difficile J Med Microbiol 2011 60 1070 9 21415200 67 Onderdonk AB Cisneros RL Bartlett JG Clostridium difficile in gnotobiotic mice Infect Immun 1980 28 277 82 7380566 68 Dobson G Hickey C Trinder J Clostridium difficile colitis causing toxic megacolon, severe sepsis and multiple organ dysfunction syndrome Intensive Care Med 2003 29 1030 12734650 69 Yacyshyn MB Yacyshyn B The role of gut inflammation in recurrent Clostridium difficile-associated disease Clin Infect Dis 2013 56 1722 3 23487370 70 El Feghaly RE Stauber JL Deych E Gonzalez C Tarr PI Haslam DB Markers of intestinal inflammation, not bacterial burden, correlate with clinical outcomes in Clostridium difficile infection Clin Infect Dis 2013 56 1713 21 23487367 71 O’Horo JC Jones A Sternke M Harper C Safdar N Molecular techniques for diagnosis of Clostridium difficile infection: systematic review and meta-analysis Mayo Clin Proc 2012 87 643 51 22766084 72 Gebhard RL Gerding DN Olson MM Peterson LR McClain CJ Ansel HJ Clinical and endoscopic findings in patients early in the course of Clostridium difficile-associated pseudomembranous colitis Am J Med 1985 78 45 8 73 Barbut F Surgers L Eckert C Visseaux B Cuingnet M Mesquita C Does a rapid diagnosis of Clostridium difficile infection impact on quality of patient management? Clin Microbiol Infect 2014 20 136 44 23565919 74 Peterson LR Mehta MS Patel PA Hacek DM Harazin M Nagwekar PP Laboratory testing for Clostridium difficile infection: light at the end of the tunnel Am J Clin Pathol 2011 136 372 80 21846912 75 Wilcox MH Overcoming barriers to effective recognition and diagnosis of Clostridium difficile infection Clin Microbiol Infect 2012 18 Suppl 6 13 20 23121550 76 Peterson LR Robicsek A Does my patient have Clostridium difficile infection? Ann Intern Med 2009 151 176 9 19652187 77 Shin BM Yoo SM Shin WC Evaluation of Xpert C. difficile, BD MAX Cdiff, IMDx C. difficile for Abbott m2000, and Illumigene C difficile assays for direct detection of toxigenic Clostridium difficile in stool specimens Ann Lab Med 2016 36 131 7 26709260 78 Smith B Peterson L Molecular diagnostic testing for Clostridium difficile infection: a double edged sword Medscape 11 5 2015 http://www.medscape.com/viewarticle/853529 79 Deshpande A Pasupuleti V Rolston DD Jain A Deshpande N Pant C Diagnostic accuracy of real-time polymerase chain reaction in detection of Clostridium difficile in the stool samples of patients with suspected Clostridium difficile infection: a meta-analysis Clin Infect Dis 2011 53 e81 90 21890762 80 Lloyd A Pasupuleti V Thota P Pant C Rolston DD Hernandez AV Accuracy of loop-mediated isothermal amplification for the diagnosis of Clostridium difficile infection: a systematic review Diagn Microbiol Infect Dis 2015 82 4 10 25752201 81 Surawicz CM McFarland LV Greenberg RN Rubin M Fekety R Mulligan ME The search for a better treatment for recurrent Clostridium difficile disease: use of high-dose vancomycin combined with Saccharomyces boulardii Clin Infect Dis 2000 31 1012 7 11049785 82 Koo HL Van JN Zhao M Ye X Revell PA Jiang ZD Real-time polymerase chain reaction detection of asymptomatic Clostridium difficile colonization and rising C. difficile -associated disease rates Infect Control Hosp Epidemiol 2014 35 667 73 24799643 83 US Food and Drug Administration Medical devices: products and medical procedures: in vitro diagnostics http://www.fda.gov/MedicalDevices/ProductsandMedicalProcedures/InVitroDiagnostics/ucm330711.htm accessed 12 September 2016 84 Bauer MP Notermans DW van Benthem BH Brazier JS Wilcox MH Rupnik M Clostridium difficile infection in Europe: a hospital-based survey Lancet 2011 377 63 73 21084111 85 Tenover FC Baron EJ Peterson LR Persing DH Laboratory diagnosis of Clostridium difficile infection. Can molecular amplification methods move us out of uncertainty? J Mol Diagn 2011 13 573 82 21854871 86 Solomon K Martin AJ O’Donoghue C Chen X Fenelon L Fanning S Mortality in patients with Clostridium difficile infection correlates with host pro-inflammatory and humoral immune responses J Med Microbiol 2013 62 1453 60 23722431 87 Drudy D Calabi E Kyne L Sougioultzis S Kelly E Fairweather N Human antibody response to surface layer proteins in Clostridium difficile infection FEMS Immunol Med Microbiol 2004 41 237 42 15196573 88 Sanchez-Hurtado K Corretge M Mutlu E McIlhagger R Starr JM Poxton IR Systemic antibody response to Clostridium difficile in colonized patients with and without symptoms and matched controls J Med Microbiol 2008 57 717 24 18480328 89 Dingle TC Mulvey GL Armstrong GD Mutagenic analysis of the Clostridium difficile flagellar proteins, FliC and FliD, and their contribution to virulence in hamsters Infect Immun 2011 79 4061 7 21788384 90 Abou Chakra CN Pepin J Valiquette L Prediction tools for unfavourable outcomes in Clostridium difficile infection: a systematic review PLoS One 2012 7 e30258 22291926 91 Walker AS Eyre DW Wyllie DH Dingle KE Griffiths D Shine B Relationship between bacterial strain type, host biomarkers, and mortality in Clostridium difficile infection Clin Infect Dis 2013 56 1589 600 23463640 92 Micic D Rao K Trindade BC Walk ST Chenoweth E Jain R Serum 25-hydroxyvitamin D levels are not associated with adverse outcomes in Clostridium difficile infection Infect Dis Rep 2015 7 5979 26500740 93 Rao K Walk ST Micic D Chenoweth E Deng L Galecki AT Procalcitonin levels associate with severity of Clostridium difficile infection PLoS One 2013 8 e58265 23505476 94 Abraham BP Kane S Fecal markers: calprotectin and lactoferrin Gastroenterol Clin North Am 2012 41 483 95 22500530 95 Dabritz J Musci J Foell D Diagnostic utility of faecal biomarkers in patients with irritable bowel syndrome World J Gastroenterol 2014 20 363 75 24574706 96 El Feghaly RE Stauber JL Tarr PI Haslam DB Intestinal inflammatory biomarkers and outcome in pediatric Clostridium difficile infections J Pediatr 2013 163 1697 704 e2 24011765 97 Kane SV Sandborn WJ Rufo PA Zholudev A Boone J Lyerly D Fecal lactoferrin is a sensitive and specific marker in identifying intestinal inflammation Am J Gastroenterol 2003 98 1309 14 12818275 98 van Langenberg DR Gearry RB Wong HL Ward M Gibson PR The potential value of faecal lactoferrin as a screening test in hospitalized patients with diarrhoea Intern Med J 2010 40 819 27 19849752 99 Vaishnavi C Bhasin D Kochhar R Singh K Clostridium difficile toxin and faecal lactoferrin assays in adult patients Microbes Infect 2000 2 1827 30 11165926 100 Sherwood RA Faecal markers of gastrointestinal inflammation J Clin Pathol 2012 65 981 5 22813730 101 Roseth AG Fagerhol MK Aadland E Schjonsby H Assessment of the neutrophil dominating protein calprotectin in feces. A methodologic study Scand J Gastroenterol 1992 27 793 8 1411288 102 Shastri YM Bergis D Povse N Schafer V Shastri S Weindel M Prospective multicenter study evaluating fecal calprotectin in adult acute bacterial diarrhea Am J Med 2008 121 1099 106 19028207 103 Bobo LD El Feghaly RE Chen YS Dubberke ER Han Z Baker AH MAPK-activated protein kinase 2 contributes to Clostridium difficile-associated inflammation Infect Immun 2013 81 713 22 23264053 104 Saenz JB Li J Haslam DB The MAP kinase-activated protein kinase 2 (MK2) contributes to the Shiga toxin-induced inflammatory response Cell Microbiol 2010 12 516 29 19951368 105 Darkoh C Turnwald BP Koo HL Garey KW Jiang ZD Aitken SL Colonic immunopathogenesis of Clostridium difficile infections Clin Vaccine Immunol 2014 21 509 17 24477852 106 Buda A Pignatelli M Cytoskeletal network in colon cancer: from genes to clinical application Int J Biochem Cell Biol 2004 36 759 65 15061130 107 Jin J-P Peterson L Host response in C. difficile infection: release of a cytoskeleton protein and its diagnostic application 2nd ASMET—The ASM Emerging Technologies Conference 8–11 March 2011 San Juan, Puerto Rico Washington, DC ASM Press 2011 108 Planche T Wilcox M Reference assays for Clostridium difficile infection: one or two gold standards? J Clin Pathol 2011 64 1 5 21118850 109 de Boer RF Wijma JJ Schuurman T Moedt J Dijk-Alberts BG Ott A Evaluation of a rapid molecular screening approach for the detection of toxigenic Clostridium difficile in general and subsequent identification of the tcdC 117 mutation in human stools J Microbiol Methods 2010 83 59 65 20674616 110 Eastwood K Else P Charlett A Wilcox M Comparison of nine commercially available Clostridium difficile toxin detection assays, a real-time PCR assay for C. difficile tcdB, and a glutamate dehydrogenase detection assay to cytotoxin testing and cytotoxigenic culture methods J Clin Microbiol 2009 47 3211 7 19710274 111 George WL Rolfe RD Harding GK Klein R Putnam CW Finegold SM Clostridium difficile and cytotoxin in feces of patients with antimicrobial agent-associated pseudomembranous colitis Infection 1982 10 205 8 7129642 112 Johal SS Hammond J Solomon K James PD Mahida YR Clostridium difficile associated diarrhoea in hospitalised patients: onset in the community and hospital and role of flexible sigmoidoscopy Gut 2004 53 673 7 15082585 113 Planche T Aghaizu A Holliman R Riley P Poloniecki J Breathnach A Diagnosis of Clostridium difficile infection by toxin detection kits: a systematic review Lancet Infect Dis 2008 8 777 84 18977696 114 Tenover FC Novak-Weekley S Woods CW Peterson LR Davis T Schreckenberger P Impact of strain type on detection of toxigenic Clostridium difficile: comparison of molecular diagnostic and enzyme immunoassay approaches J Clin Microbiol 2010 48 3719 24 20702676
PMC005xxxxxx/PMC5124544.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 9215515 20498 Neuroimage Neuroimage NeuroImage 1053-8119 1095-9572 27241482 5124544 10.1016/j.neuroimage.2016.05.064 NIHMS797359 Article Imaging Brain Source Extent from EEG/MEG by Means of an Iteratively Reweighted Edge Sparsity Minimization (IRES) Strategy Sohrabpour Abbas a Lu Yunfeng a Worrell Gregory b He Bin ac* a Department of Biomedical Engineering, University of Minnesota, Minneapolis, MN, USA b Department of Neurology, Mayo Clinic, Rochester, MN, USA c Institute for Engineering in Medicine, University of Minnesota, Minneapolis, MN, USA * Correspondence: Bin He, Ph.D., University of Minnesota, Department of Biomedical Engineering, 7-105 Hasselmo Hall, 312 Church Street, Minneapolis, MN 55455, [email protected] 24 6 2016 27 5 2016 15 11 2016 15 11 2017 142 2742 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Estimating extended brain sources using EEG/MEG source imaging techniques is challenging. EEG and MEG have excellent temporal resolution at millisecond scale but their spatial resolution is limited due to the volume conduction effect. We have exploited sparse signal processing techniques in this study to impose sparsity on the underlying source and its transformation in other domains (mathematical domains, like spatial gradient). Using an iterative reweighting strategy to penalize locations that are less likely to contain any source, it is shown that the proposed iteratively reweighted edge sparsity minimization (IRES) strategy can provide reasonable information regarding the location and extent of the underlying sources. This approach is unique in the sense that it estimates extended sources without the need of subjectively thresholding the solution. The performance of IRES was evaluated in a series of computer simulations. Different parameters such as source location and signal-to-noise ratio were varied and the estimated results were compared to the targets using metrics such as localization error (LE), area under curve (AUC) and overlap between the estimated and simulated sources. It is shown that IRES provides extended solutions which not only localize the source but also provide estimation for the source extent. The performance of IRES was further tested in epileptic patients undergoing intracranial EEG (iEEG) recording for pre-surgical evaluation. IRES was applied to scalp EEGs during interictal spikes, and results were compared with iEEG and surgical resection outcome in the patients. The pilot clinical study results are promising and demonstrate a good concordance between noninvasive IRES source estimation with iEEG and surgical resection outcomes in the same patients. The proposed algorithm, i.e. IRES, estimates extended source solutions from scalp electromagnetic signals which provide relatively accurate information about the location and extent of the underlying source. Extent Sparsity Iterative reweighting EEG MEG Inverse problem Source extent Convex optimization Introduction Electroencehalography (EEG)/magnetoencephalography (MEG) source imaging is to estimate the underlying brain activity from scalp recorded EEG/MEG signals. The locations within the brain involved in a cognitive or pathological process can be estimated using associated electromagnetic signals such as scalp EEG or MEG (He et al., 2011; Baillet et al., 2001). The process of estimating underlying sources from scalp measurements is a type of inverse problem referred to as electrophysiological source imaging (ESI) (Michel et al., 2004b; Michel & He, 2011; He & Ding, 2013). There are two main strategies to solve the EEG/MEG inverse problem (source imaging), namely the equivalent dipole models (Scherg & von Cramon 1985; He et al, 1987) and the distributed source models (Hämäläinen et al., 1984; Dale & Sereno, 1993; Pascual-Marqui et al., 1994). The dipole models assume that the electrical activity of the brain can be represented by a small number of equivalent current dipoles (ECD) thus resulting in an over-determined inverse problem. This, however leads to a nonlinear optimization problem, which ultimately estimates the location, orientation and amplitude of a limited number of equivalent dipoles, to fit the measured data. The number of dipoles has to be determined a priori (Bai & He, 2006). On the other hand, the distributed source models use a large number of dipoles (Hämäläinen et al., 1984; Dale & Sereno, 1993; Pascual-Marqui et al., 1994) or monopoles (He et al, 2002) distributed within the brain volume or the cortex. In such models, the problem becomes linear, since the dipoles (monopoles) are fixed in predefined grid locations, but the model is highly underdetermined as the number of unknowns is much more than the number of measurements. Given that functional areas within the brain are extended and not point-like, the distributed models are more realistic. Additionally, determining the number of dipoles to be used in an ECD model is not a straightforward process (Michel et al., 2004b). Solving under-determined inverse problems calls for regularization terms (in the optimization problem) or prior information regarding the underlying sources. Weighted minimum norm solutions were one of the first and most popular algorithms. In these models, the regularization term is the weighted norm of the solution (Lawson & Hanson, 1974). Such regularization terms will make the process of inversion (going from measurements to underlying sources) possible and will also impose additional qualities to the estimation such as smoothness or compactness. Depending on what kind of weighting is used within the regularization term, different solutions can be obtained. If a uniform weighting or identity matrix is used, the estimate is known as the minimum norm (MN) solution (Hämäläinen et al., 1984). The MN solution is the solution with least energy (L2 norm) within the possible solutions that fit the measurements. It is due to this preference for sources with a small norm that MN solutions are well-known to be biased towards superficial sources (Hämäläinen et al., 1994). One modification to this setback is to use the norm of the columns of the lead field matrix to weight the regularization term in such a manner to penalize superficial sources more than the deep sources, as the deep sources do not present well in the scalp potential (Lawson & Hanson, 1974). In this manner the tendency towards superficial sources is alleviated. This is usually referred to as the weighted minimum norm (WMN) solution. Another popular choice is the low resolution brain electromagnetic tomography (LORETA) (Pascual-Marqui et al., 1994). LORETA is basically a WMN solution where the weighting is a discrete Laplacian. The solution’s second spatial derivative is minimized so the estimation is smooth. Many inverse methods apply the L2 norm, i.e. Euclidean norm, in the regularization term. This causes the estimated solution to be smooth, resulting in solutions that are overly smoothed and extended all over the solution space. Determining the active cortical region by distinguishing desired source activity from background activity (to determine source extent) proves difficult in these algorithms, as the solution is overly smooth and poses no clear edges between background and active brain regions (pertinent or desired activity, epileptic sources for instance as compared to background activity or noise). This is one major drawback of most conventional algorithms including the ones discussed so far. In order to overcome the extremely smooth solutions, the L2 norm can be supplanted by the L1 norm. This idea is inspired from sparse signal processing literature where the L1 norm has been proposed to model sparse signals better and more efficiently, specifically after the introduction of the least absolute shrinkage selection operator (LASSO) (Tibshirani, 1996). While optimization problems involving L1 norms do not have closed form solutions, they are easy to solve as they fall within the category of convex optimization problems (Boyd & Vandenberghe, 2004). Selective minimum norm method (Matsura & Okabe, 1995), minimum current estimate (Uutela et al., 1999) and sparse source imaging (Ding & He, 2008) are examples of such methods. These methods seek to minimize the L1 norm of the solution. Another algorithm in this category, which uses a weighted minimum L1 norm approach to improve the stability and “spiky-looking” character of L1-norm approaches, is the vector-based spatio-temporal analysis using L1-minimum norm (VESTAL) (Huang et al., 2006). These algorithms encourage extremely focused solutions. Such unrealistic solutions root from the fact that by penalizing the L1 norm of the solution a sparse solution is being encouraged. As discussed by Donoho (Donoho, 2006) under proper conditions regularizing the L1 norm of a solution will result in a sparse solution; a solution which has only a few number of non-zero elements. Sparsity is definitely not a desired quality for underlying sources which produce EEG/MEG signals as EEG/MEG signals are the result of synchronous activity of neurons from a certain extended cortical region (Baillet et al., 2001; Nunez et al., 2000). To overcome the aforementioned shortcomings while still benefiting from the advantages of sparse signal processing techniques, new regularization terms which encourage sparsity in other domains have been proposed. The idea is to find a domain in which the solution has a sparse representation. This basically means that while the solution might not be sparse itself (as is usually the case for underlying sources generating the EEG/MEG) it still might be sparsely represented in another domain such as the spatial gradient or wavelet coefficient domain. This amounts to the fact that the signal still has redundancies that can be exploited in other domains. Haufe et al. (Haufe et al., 2008) penalized the Laplacian of the solution instead and showed focal results which are realistically extended. They have shown the effect of considering implicit domain sparsity in improving the results by comparing their estimation with that of conventional methods using L2 norm regularization terms or simple L1 norm terms. Ding (Ding, 2009) has tried penalizing the gradient instead of Laplacian. Due to the selected penalization term which penalizes the solution discontinuities or jumps, the solution is piecewise constant and needs thresholding to discard the low level semi-constant background activity. Liao et al. (Liao et al., 2012) have used the faced-based wavelet in the penalty term and have shown focal results. Chang et al. (Chang et al., 2010) and Zhu et al. (Zhu et al., 2014) proposed to impose sparsity on multiple domains to better capture the redundancies of the underlying sources and have shown positive results. Zhu et al. combined the gradient and wavelet transform in the regularization term and Chang et al. used two strategies to combine domain sparsity. The first strategy is to impose sparsity on the solution and the Laplacian of the solution, and the second strategy is to impose sparsity on the solution and its wavelet transform. Another piece of work worthy of mentioning is the Elastic Net (ENET) (Zou and Hastie, 2005) and ENET L (Vega-Hernández et al, 2008). In the ENET algorithm both the L1 norm and L2 norm of the solution are regularized to obtain estimations that are more robust than LASSO-type solutions. ENET L regularizes the L1 and L2 norm of the Laplacian to obtain smooth and focal solutions. While imposing sparsity on multiple domains improves the results and seems to be a good approach for estimating extended sources, the solutions presented in the discussed papers cannot yet determine the extent of the underlying source objectively, i.e. still a threshold needs to be applied to reject the background activity. Another successful class of inverse algorithms is Bayesian inverse techniques. In these methods, the problem is formulated within a Bayesian framework starting with prior distributions (of the dipole current density) to converge to a posterior distribution of the underlying source. Wipf and Nagarajan (Wipf and Nagarajan, 2009) have discussed this approach thoroughly. In this work many conventional algorithms such as MN, WMN and FOCUSS (Gorodnitsky et al., 1995) are re-introduced within this framework. Another well-known Bayesian algorithm is the maximum entropy on the mean (MEM) approach (Grova et al., 2006) where the cortical surface is clustered in a data driven manner to obtain active and inactive regions on the cortex by regularizing the mean entropy. The idea of MEM has been further pursued by Chowdhury et al. (Chowdhury et al., 2013) and Lina et al. (Lina et al., 2014) in a parcelization framework, where the cortical surface is divided into segments and then it is determined if each parcel is within the active source patch or not (through statistical analysis). The proposed method in the present work is not defined within the Bayesian framework, but as a series of convex optimization problems, as will be discussed. Model based algorithms inspired by the sparse signal processing literature have also been proposed in recent years. Spatial or temporal basis functions which are believed to model the underlying source activity of the brain are defined a priori in a huge data set called the dictionary (Bolstad et al., 2009; Limpiti et al., 2006). Later on a solution which best fits the measurements is sought within the dictionary. Haufe et al. proposed a Gaussian basis to spatially model extended brain sources (Haufe et al., 2011). These methods can be effective in solving the inverse problem as long as the underlying assumption about the basis functions holds true (since the solution is basically subsumed within the dictionary elements). For instance, the design of the basis function in (Haufe et al., 2011) might include very compact Gaussian kernels to provide a chance for the solver to select these kernels and give more compact estimates (although extent estimation is not pursued in that work). Furthermore, for a given spatial extent for the Gaussian kernel infinitely many different standard deviations can be assumed. If the kernel includes a good amount of such cases, in order to be unbiased and to avoid selecting parameters a priori (like subjective thresholding,) the dictionary can be huge and the problem might become intractable. However similar approaches undertaken by other groups have not been able to resolve this issue completely. Other studies (Chang et al., 2010; Liao et al., 2014; Zhu et al., 2014) adopted a similar approach using wavelets (wavelets that have many levels of precision and spatial extent) and their results still needed a minor thresholding to reject the weak sources. The proposed method in the present work does not assume any prior dictionaries or basis functions prior to solving the inverse problem. Mixed-norm estimates have also gained attention in recent years (Gramfort et al., 2012; Gramfort et al., 2013a). These algorithms have also been incorporated in an iterative reweighting scheme (Strohmeier et al., 2014, 2015). These algorithms define two-level (or multi-level) mixed norms (usually combining L1 and L2 norms) to obtain focal solutions. Since the regularization is enforced on the solution, very focal estimates are obtained. Basically not much information regarding the source extent can be extracted from these algorithms currently, although it is suggested that newer implementations or combination with other algorithms may provide such capabilities in the future (Gramfort et al., 2013a). Estimating the extent of the underlying source is a challenge but also highly desirable in many applications. Determining the epileptogenic brain tissue is one important application. EEG/MEG source imaging is a non-invasive technique making its way into the presurgical workup of epilepsy patients undergoing surgery. EEG/MEG source imaging helps the physician in localizing the location of activity and if it can more reliably and objectively estimate the extent of the underlying epileptogenic tissue, the potential merits to improve patient care and quality of life are obvious since EEG and MEG are noninvasive modalities. Another important application is mapping brain functions, elucidating roles of different regions of the brain responsible for specific functional tasks using EEG/MEG (He et al., 2013). In order to come up with an algorithm that is able to objectively determine the extent of the underlying sources, we move along the lines of multiple domain sparsity and also introduce the notion of iterative reweighting within the sparsity framework to achieve this goal. If the initial estimation of the underlying source is relaxed enough to provide an overestimation of the extent, it is possible to use this initial estimation and launch a series of subsequent optimization problems to gradually converge to a more accurate estimation of the underlying source. The sparsity is imposed on both the solution and the gradient of the solution. This is the basis of the proposed algorithm, which is called iteratively reweighted edge sparsity minimization (IRES). The notion of edge sparsity or imposing sparsity on the gradient of the solution is also referred to as the total-variation (TV) in image processing literature (Adde et al. 2005; Rudin et al., 1992). Recently some fMRI studies have shown the usefulness of working within the TV framework to obtain focal hot-spots within fMRI maps without noisy spiky-looking results (Dohmatob et al., 2014; Gramfort et al., 2013b). The results presented in the aforementioned works still need to set a threshold to reject background activity. These approaches are similar to the approach adopted in IRES with the difference that IRES initiates a sequence of reweighting iterations based on obtained solutions to suppress background activity and create clear edges between sources and background. One example has been presented in the supplementary materials to show the effect of thresholding on IRES estimates (Fig. S1). A series of computer simulations were performed to assess the performance of the IRES algorithm in estimating source extent from the scalp EEG. The estimated results were compared with the simulated target sources and quantified using different metrics. To show the usefulness of IRES in determining the location and extent of the epileptogenic zone in case of focal epilepsy, the algorithm has been applied to source estimation from scalp EEG recordings and compared to clinical findings such as resection and seizure onset zone (SOZ) determined from intracranial EEG by the physician. Materials and Methods Iteratively reweighted edge sparsity minimization (IRES) The brain electrical activity can be modeled by current dipole distributions. The relation between the current dipole distribution and the scalp EEG/MEG is constituted by Maxwell’s equations. After discretizing the solution space and numerically solving Maxwell’s equations, a linear relationship between the current dipole distribution and the scalp EEG/MEG can be derived: (1) φ=Kj+n0 where φ is the vector of scalp EEG (or MEG) measurements, K is the lead field matrix which can be numerically calculated using the boundary element method (BEM) modeling, j is the vector of current dipoles to be estimated and n0 models the noise. For EEG source imaging, φ is an M x 1 vector, where M refers to the number of sensors; K is an M x D matrix, where D refers to the number of current dipoles; j is a vector of D x 1, and n0 is an M x 1 vector. Following the multiple domain sparsity in the regularization terms, the optimization problem is formulated as a second order cone programming (SOCP) (refer to Boyd & Vandenberghe, 2004 for more details) in (2). While problems involving L1 norm minimization do not have closed-form solutions and may seem complicated, such problems are easy to solve as they fall within the convex optimization category (Boyd & Vandenberghe, 2004). There are many efficient methods for solving convex optimization problems. (2) jest=argminj\|\|Vj\|\|1+α\|\|j\|\|1subjectto(φ-Kj)T∑-1(φ-Kj)≤β Where V is the discrete gradient defined based on the source domain geometry (T x D, where T is the number of edges as defined by (4), later on), β is a parameter to determine noise level and Σ is the covariance matrix of residuals, i.e. measurement noise covariance. Under the assumption of additive white Gaussian noise (AWGN), Σ is simply a diagonal matrix with its diagonal entries corresponding to the variance of noise for each recording channel. In more general and realistic situations, Σ is not diagonal and has to be estimated from the data (refer to simulation protocols for more details on how this can be implemented). Under the uncorrelated Gaussian noise assumption it is easy to see that the distribution of the residual term will follow the χn2 distribution, where χn2 is the chi-squared distribution with n degrees of freedom (n is the number of recording channels, i.e. number of EEG/MEG sensors). In case of correlated noise, the noise whitening process in (2) will eliminate the correlations and hence is an important step. This de-correlation process is achieved by multiplying the inverse of the covariance matrix (Σ−1) by the residuals, as formulated in the constraint of the optimization problem in (2). In order to determine the value of β the discrepancy principle is applied (Morozov, 1966). This translates to finding a value for β for which it can be guaranteed that the probability (p) of having the residual energy within the [0 β] interval is high (p). Setting p=0.99 (Zhu et al., 2014; Malioutov et al., 2005), β was calculated using the inverse cumulative distribution function of the χn2 distribution. The optimization problem proposed in (2) is an SOCP-type problem that needs to be solved at every iteration of IRES. In each iteration, based on the estimated solution, a weighting coefficient is assigned to each dipole location. Intuitively, locations which have dipoles with small amplitude will be penalized more than locations which have dipoles with larger amplitude. In this manner, the optimization problem will gradually slim down to a better estimate of the underlying source. The details of how to update the weights at each iteration and why to follow such a procedure is given in Appendix A. Mathematically speaking, the following procedure is repeated until the solution does not change significantly in two consecutive iterations as outlined in (3): At iteration L: (3) jL=argminj\|\|WdL-1(Vj)\|\|1+α\|\|WL-1j\|\|1subjectto(φ-KjL)T∑-1(φ-KjL)≤β where WL and WdL are updated based on the estimation jL (refer to appendix A for details). The procedure is depicted in Fig. 1. The idea of data-driven weighting is schematically depicted. Although the number of iterations cannot be determined a priori, the solution converges pretty fast, usually within two to three iterations. One of the advantages of IRES is that it uses data-driven weights to converge to a spatially extended source. Following the idea proposed in the sparse signal processing literature (Candès et al., 2008; Wipf and Nagarajan, 2010), the heuristic that locations with smaller dipole amplitude need to be penalized more than other locations, will be formalized. In the sparse signal processing literature it is well known that under some general conditions (Donoho, 2006) the L1-norm can produce sparse solutions; in other words “L0-norm” can be replaced with L1-norm. In reality, “L0-norm” is not really a norm, mathematically speaking. It assigns 0 to the elements of the input vector when those elements are 0 and 1 otherwise. It is easy to imagine that when sparsity is considered, such a measure or pseudo-norm is intended (as this measure will impose the majority of the elements of the vector to be zero, when minimized). However this measure is a non-convex function and including it in an optimization problem makes it hard or impossible to solve, so it is replaced by the L1-norm which is a convex function and under general conditions the solutions of the two problems are close enough. When envisioning “L0-norm” and L1-norm, it is evident that while “L0-norm” takes a constant value as the norm of the input vector goes to infinity, L1-norm is unbounded and goes to infinity. In order to use a measure which better approximates the L0-norm and yet has some good qualities (for the optimization problem to be solvable), Fazel et al. (Fazel, 2002; Fazel et al., 2003) proposed that a logarithm function be used instead of the “L0-norm”. Logarithmic functions are concave but also quasi-convex (refer to (Boyd and Vandenberghe, 2004) for the definition and for more properties), thus the problem would be solvable. However, finding the global minimum (which is a promise in the convex optimization problems) is no more guaranteed. This means that the problem is replaced with a series of optimization problems, which could converge to a local minimum; thus the final outcome of the problem depends on the initial estimation. Our results in this paper indicate that initiating the problem formulated in (3) with identity matrices, provide good estimates in most of the cases, hopefully indicating that the algorithm might not be getting trapped in local minima. More detailed mathematical analysis is presented in Appendix A. Selecting the hyper-parameter α is not a trivial task. Selecting hyper-parameters can be a dilemma in any optimization problem and most optimization problems inevitably face such a selection. It is proposed to adopt the L-curve approach to objectively determine the suitable value for α (Hansen et al. 1990; He et al., 2011). Referring to Fig. 2 it can be seen how selecting different values for α can affect the problem. In this example a 20 mm extent source is simulated and a range of different α values ranging from 1 to 10−4 are used to solve the inverse problem. As it can be seen, selecting a large value for α will result in an overly focused solution (underestimation of the spatial extent). This is due to the fact that by selecting a large value for α the optimization problem focuses more on the L1-norm of the solution rather than the domain sparsity (TV term) so the solution will be sparse. In the extreme case when α is much larger than 1 the optimization problem turns into a L1 estimation problem which is known to be extremely sparse, i.e. focused. Conversely selecting very small values for α may result in spread solutions (overestimation of the spatial extent). Selecting an α value near the bend (knee) of the curve is a compromise to get a solution which minimizes both terms involved in the regularization. The L-curve basically looks at different terms within the regularization and tries to find an α for which all the terms are small and also changing α minimally along each axis will not change the other terms drastically. In other words the bend of the L-curve gives the optimal α as changing α will result in at least one of the terms in the regularization term to grow which is counterproductive in terms of minimizing (2). In this example an α value of 0.05 to 0.005 seems to give reasonable results) Another parameter to control for is the number of iterations. Although this cannot be theoretically dealt with now, it is suggested to continue iteration until the estimation of two consecutive steps do not vary much. The actual number of iterations needed is usually 2 to 4 iterations, as our experience with the data suggests. This is also reported by Candès et al (Candès et al., 2008). This means that within a few iterations an extended solution with distinctive edges from the background activity is reached. Fig. 3 shows one example. In this case a 10 mm extent source is estimated and the solution is depicted through 10 iterations. As it can be seen, the solution stabilizes after 3 iterations and it stays stable even after 10 iterations. It is also interesting that these iterations do not cause the solution to shrink and produce an overly concentrated solution like the well-known algorithm FOCUSS (Gorodnitsky et al., 1995). This is due to the fact that the regularization term in IRES contains TV and L1 terms which in turn balance between sparsity and edge-sparsity, avoiding overly spread or focused solutions. Neighborhood and edge definition In order to form matrix V which approximates some sort of total variation among the dipoles on the cortex, it is necessary to constitute the concept of neighborhood. Since the cortical surface is triangulated for the purpose of solving the forward problem (using the boundary element model) to form the lead field matrix K, there exists an objective and simple way to define neighboring relationship. The center of each triangle is taken as the location of the dipoles to be estimated (amplitude) and as each triangle is connected to only three other triangles via its edges, each dipole is neighbor to only three other dipoles. Based on this simple relation, neighboring dipoles can be detected and the edge would be defined as the difference between the amplitude of two neighboring dipoles. Based on this explanation it is easy to form matrix V (Ding, 2009) as presented in (4): (4) V=(v11⋯v1n⋮⋱⋮vT1⋯vTn){vij=1andvik=-1ifdipolejandkareneighborsoveredgeivij=0otherwise The number of edges is denoted by T. Basically each row of matrix V corresponds to an edge that is shared between two triangles and the +1 and −1 values within that row are located such as to differentiate the two dipoles that are neighbors over that particular edge. The operator V can be defined regardless of mesh size, as the neighboring elements can be always formed and determined in a triangular tessellation (always three neighbors). However reducing the size of the mesh to very small values (less than 1mm) is not reasonable as M/EEG recordings are well-known to be responses from ensembles of postsynaptic neuronal activity. Having small mesh grids will increase the size of V relentlessly without any meaningful improvement. On the other hand increasing the grid size to large values (>1 cm) will also give coarse grids that can potentially give coarse results. It is difficult to give a mathematical expression on this but a grid size of 3~4 mm was chosen to avoid too small a grid size and too coarse a tessellation. Computer simulation protocol In order to analyze IRES performance, a series of computer simulations were conducted in a realistic cortex model. The cortex model was derived from MR images of a human subject. The MR images were segmented into three layers, namely the brain tissue, the skull and the skin. Based on this segmentation a three layer BEM model was derived to solve the forward problem and obtain the lead field matrix, constituting the relation between current density dipoles and the scalp potential. The conductivity of the three layers, i.e. brain, skull and skin, were selected respectively as 0.33 S/m, 0.0165 S/m and 0.33 S/m (Oostendorp et al, 2000; Lai et al., 2005; Zhang et al., 2006). The number of recording electrodes used in the simulation is 128 channels. 100 random locations on the cortex were selected and at each location sources with different extent sizes were simulated, ranging from 10 mm to 30 mm in radius size. The amplitude of the dipoles were set to unity, so the cortical surface was partitioned into active (underlying source) and non-active (not included within the source) area. The orientation of the sources were taken to be normal to the cortical surface, as the pyramidal cells located in the gray matter responsible for generating the EEG signals are oriented orthogonal to the cortical surface (Baillet et al., 2001; Nunez et al., 2000). The orientation of the dipoles was accordingly fixed when solving the inverse problem. Different levels of white Gaussian noise were added to the generated scalp potential maps to make simulation more realistic. The noise power was controlled to obtain different levels of desired signal to noise ratio (SNR), namely 10 dB and 20 dB (another set of simulations with 6dB SNR was also conducted as explained in the next paragraph). These SNR values are realistic in many applications including epileptic source imaging. The inverse solutions were obtained using IRES and the estimated solutions were compared to the ground truth (simulated sources) for further assessment. The results of these simulations are presented in Figs. 4 to 6. In order to compare the effect of modeling parameters on the inverse algorithm and also to avoid the most obvious form of “inverse crime”, another series of simulations were also conducted for which the forward and inverse model was different (in addition to the previous results presented in Figs. 4 to 6). The mesh used for the forward problem was very fine with 1mm spacing consisting of 225,879 elements on the cortex. A BEM model consisting of three layers, i.e. brain, skull and skin with conductivities of 0.33 S/m, 0.015 S/m and 0.33 S/m was used for the forward model. For the inverse model a coarse mesh of 3mm spacing consisting of 28,042 elements was used. The inverse BEM model consists of three layers, i.e. brain, skull and skin with conductivities of 0.33 S/m, 0.0165 S/m and 0.33 S/m. Basically the conductivity ratio is changed by 10% in the inverse model compared to the forward model and also a different and much finer grid is used for the forward model in comparison to the inverse model (Auranen et al, 2005; Auranen et al., 2007). In this manner we reduced the dependency of IRES performance to modeling parameters such as grid size and conductivity values (by using different lead field matrices for forward and inverse). In addition to that, we used realistic noise recorded from a human subject as the additive noise so as to avoid using only white noise. Low SNR of 5~6 dB was also tested following (Gramford et al., 2013), to make sure IRES can be used in noisier conditions. The results of these simulations are presented in Fig. 7 and Fig. 8. To further assess the effect of slight differences in conductivities on inverse algorithms’ performance, another inverse model was formed and used as well. This inverse model is the same as the inverse model described in the previous paragraph, meaning that in these simulations the grids used for the forward and inverse model were different but the conductivity values were not. The results of the simulations are presented in Fig. S2 and Fig. S3 in the supplementary materials. Model violations such as non-constant sources (amplitude) and multiple simultaneous active sources were also tested. Additionally cLORETA (Wagner et al., 1996) and focal vector field reconstruction (FVR) (Haufe et al., 2008) were used to estimate solutions and the results from these inverse algorithms were compared with IRES (these results are presented in Figs. S4 and S5 of the supplementary materials). To further evaluate the performance of IRES non-constant sources (for which the dipoles within the source patch did not have constant amplitudes and varied in amplitude) and multiple simultaneously active sources were also simulated and tested. Results of IRES performance are presented in Fig. 9 for multiple cases in these scenarios. More detailed results and analyses are presented in Figs. S6 to S9 of the supplementary materials. Performance measures In order to evaluate IRES performance, multiple metrics were used. As extended sources are being considered in this study, appropriate metrics being able to compare extended sources need to be used. The first measure is the localization error (LE). The localization error calculates the Euclidean distance between the center of mass of the simulated and estimated sources. In order to compare the shape and relative position of the estimated and simulated sources the overlap metric is used. The amount of overlap between the estimated and simulated sources is calculated and divided by either the simulated source area or the estimated source area to derive a normalized overlap ratio (NOR). This normalized overlap shows how well the two distributions match each other. These measures should both be as close as possible to 1. If an overestimated or underestimated solution is obtained, one of the two measures can be close to 1 while the other decreases significantly. Another important measure is the area under curve (AUC) analysis (Grova et al., 2006). The curve mentioned is the receiver operating characteristics (ROC) curve (Kay, 2011). The AUC enables us to compare two source distributions (one is the estimated source distribution and the other one is the simulated source). The closer this AUC value is to 1, the better our estimation of the simulated source will be. Clinical data In order to determine if IRES could be used in practical settings and ultimately translated into clinical settings, the proposed algorithm was also tested in patients suffering from partial (focal) epilepsy. Three patients were included in this study. All clinical studies were conducted according to a protocol approved by the institutional review board (IRB) of Mayo Clinic, Rochester and the University of Minnesota. All patients had pre-surgical recordings with multiple inter-ictal spikes in their EEG recording. Two of the patients were suffering from temporal lobe epilepsy (TLE) and one was diagnosed with fronto-parietal lobe epilepsy. All three patients underwent surgery and were seizure free at one year post-operation follow-up. Two of the patients also had intracranial EEG recordings (before surgery) from which the seizure onset zone (SOZ) and spread activity electrodes were determined (Fig. 10). All patients had pre-surgical MRI as well as post-surgical MRI images (An example of pre/post-surgical MRI images is presented in Fig. S10 of the supplementary materials). The pre-surgical MRI was used to form individual realistic geometry head models, i.e. BEM models, for every patient. The BEM model composed of three layers. The conductivity of the three layers, i.e. brain, skull and skin, were selected respectively as 0.33 S/m, 0.0165 S/m and 0.33 S/m (Oostendorp et al, 2000; Lai et al., 2005; Zhang et al., 2006). The post-op MRI was used to locate and segment the surgical resection to later compare with IRES estimated solution. The EEG recordings were filtered with a band-pass filter with cut-off frequencies set at 1 Hz and 50 Hz. Inter-ictal spikes were searched for, through patient EEG recordings prior to operation, and were checked for scalp map consistency and temporal similarities. The spikes that were repeated more often were included in the analysis. The spikes were averaged around the peak of their mean global field power (MGFP) to produce a single averaged spike; on average 10 spikes were averaged for this study. These averaged spikes were then used for further source imaging analysis. The electrode montage used for these patients contained 76 electrodes (Yang et al., 2011) and the electrode locations were digitized and used in the inverse calculation. The estimated solution by IRES is compared to the surgical resection surface (as IRES is currently confined to cortical surface) and SOZ whenever available. Implementation The BEM based forward problem was solved using CURRY 7 (Compumedics, Chalotte, NC). The cortical surface was triangulated into 1 mm and 3 mm mesh for the computer simulations and clinical data analysis. In order to solve the SOCP problem which is the backbone of IRES a convex problem solver called CVX (Grant & Boyd, 2008; Grant & Boyd, 2013) was used. CVX contains many solvers including the self-dual-minimization (SeDuMi) (Strum, 1999) which is a MATLAB (Mathworks, Natick, MA) compatible package implementing an interior path method (IPM) for solving the SOCP problems. It takes about 2–4 minutes to solve (3) at each iteration on widely available desktop computers (3.4 GHz CPU and 4 Gbytes RAM). Although such a computation time is about 20 times that of MN-like algorithms, it is not too lengthy and IRES can be solved in reasonable time. While we have not developed a specific solver for IRES and used a general solver, i.e. CVX, it is reasonable to assume that the running time can be improved with tailored algorithms specifically designed for IRES. Results Computer Simulations Fig. 4 through Fig. 6 show simulation results using the same lead field matrix for forward and inverse problem. The results presented in figures 4 and 5 pertain to the case were the SNR of the simulated scalp potential is 20 dB and the same results are presented in figure 6 for the 10 dB case. As it can be seen in Fig. 4 IRES can distinguish between different source sizes and estimate the extent with good accuracy. In the left panel of Fig. 4, three different cases are presented with extents of 10, 20 and 30 mm, respectively. Comparing the simulated sources and estimation results shows that IRES can distinguish between different sources reasonably. The right panel in Fig. 4 shows the relation between the extent of the estimated and simulated source. The fitted line shows that IRES has small bias and minimal under/over-estimation on average. The variance of the estimated solutions is comparable to the estimated extent (about 50% of source extent). The overall trend is positive and indicates that IRES is relatively unbiased in estimating underlying source extent. Other measures such as LE, NOR and AUC are also important to assess the performance of IRES. In Fig. 5 the results of such different metrics can be seen. The localization error in sources with different extent is less than 5 mm over all. Note that the simulated sources were approximately categorized into three classes with average extent of 10, 20 and 30 mm, corresponding to the three colors seen in Fig. 4. This value is less for smaller sources and closer to 3 mm. Such a low localization error shows that IRES can localize the underlying extended sources with low bias. Actually there a few cases (where the LE is greater than 12 mm) for which the simulated sources were split between the two hemispheres with an unconventional geometry, thus the solution was a bit widespread and not good. However some of the cases simulated in deeper regions like the medial temporal lobe or the medial wall of the inter-hemisphere have errors in the range of 5–10 mm. Results of some difficult simulation cases are provided in Fig. 12 and discussed further in the Discussion section. Additionally, to better understand the combined effect of LE and extent estimation, the normalized overlaps should be studied. Looking at Fig. 5 one can see that the overlap between the estimated and simulated source is relatively high and over 70% for smaller sources (on average) and close to 85% for larger sources. The fact that both NOR values are high show that not only the estimated and simulated sources overlap extensively with each other, but the estimation is neither an overestimation nor an underestimation of the spatial extent (in either case only one of the NOR values would be high and the other would be small). The high AUC values for various source sizes also indicate the overall high sensitivity and specificity of IRES as an estimator. Fig. 6 shows the same results for the 10 dB case. Comparing Fig. 6 with Fig. 5 and Fig. 4 similar trends can be observed. Fig. 7 and Fig. 8 show simulation results when different forward and inverse models (in terms of grid size and conductivity) are used for two cases of 6 dB and 20 dB SNR. The results in Fig. 7 show an underestimation for the extent. The localization error is as low as 5 mm and the NOR is ~60%–70% on average. Comparing the same results when SNR is set to 20 dB, much better results can be obtained. Referring to Fig. 8 it can be seen that the extent estimation is with minimal bias with localization error of 3 mm (on average) and high NOR metrics (80% for both values on average). The slight decline in IRES performance in noisier conditions (Fig. 6 and Fig. 7) is expected due to increased levels of noise interference. A few examples when source amplitude is not constant and also when multiple sources are simultaneously active, are presented in Fig. 9. The variation of source amplitude was governed by a normal distribution, meaning that the amplitude of the dipoles decreased exponentially as the distance of the dipoles increased from the center of the source distribution (patch). More detailed analysis and explanations are presented in supplementary materials. These simulation results show that IRES is robust against noise and changes in model parameters such as grid size and conductivity. IRES can perform well when multiple sources are active or when source amplitude varies within the source extent. Clinical Data Analysis The patient data analysis is summarized in figures 10 and 11. Fig. 10 shows the results of a temporal lobe epilepsy patient who underwent invasive EEG recording and ultimately surgical resection. In this case different timing was tested to examine the effect of estimation results at different time instances around the peak. It can be seen in Fig. 10a, that five different timings (instances) were tested, two prior to peak time, the peak time and two after the peak corresponding to 50% and 70% rising phase, peak and 70% and 50% falling time. Looking at the estimated solution at different timings, it is clear that the solution is stable within tens of milliseconds around the peak and the epileptic activity does not propagate too much. This might be due to the fact that the averaged inter-ictal spikes arise from the irritative zone which is larger than the SOZ, so the propagation of activity might not even be observable using the average spike. Comparing the results of IRES with the resection in Fig. 10b it can be seen that the estimated solution coincides within the resection very well. Looking at the SOZ electrodes colored pink in Fig. 10b it is also observed that the SOZ electrodes are covered by the solution. SOZ region is very focal so it is challenging to obtain a solution which can be in full concordance with it and as it can be seen, sometimes the SOZ region is not continuous (electrodes are not always right next to each other); nonetheless the estimated solution by IRES is in concordance with the clinical findings. In order to further test the proposed IRES approach two more patients were studied; one temporal case with anterior tip of temporal lobe resection (and thus a smaller resection area) and an extra-temporal case (fronto-parietal). Referring to Fig. 11a it can be seen that the estimated solution includes most of the SOZ electrodes and coincides well with the resection. This implies that IRES does well for extra-temporal lobe cases as well as temporal cases. Given the fact that not all the SOZ electrodes are close to the resection, the estimated solution is a bit spread towards those SOZ electrodes and the neighboring regions thus extending beyond the resected region. In Fig. 11b the results of the second temporal case can be found. This patient did not have any intra-cranial EEG recordings. Fig. 11c reports the quantitative analyses for these three patients. In order to assess how well the estimated results match the clinical findings, the overlap between the solution and the resection and the SOZ was calculated. Then this overlap area was either divided by the solution area or the resection/SOZ area. The results are classified as resection and SOZ indicating the clinical finding, i.e. SOZ or resection, used to assess the estimated solutions. Looking at Fig. 11c it can be seen that the IRES solution generally covers the SOZ very well but also extends beyond the SOZ giving an overestimated solution. Looking at the right bar-plot it can be seen that this is not the case when comparing the IRES solution to resected area. Note that for temporal lobe epilepsy cases, due to more geometrical complexity of the cortex in the temporal region (compared to other locations within the brain) and the fact that mesio-temporal region is not directly recorded by the electrodes over the temporal region, Vector-based IRES (VIRES) was used instead of IRES (mathematical details of VIRES are given in Appendix B). VIRES basically relaxes the orientation constraint of the dipoles (being orthogonal to the cortical surface as implemented by IRES) and leaves that as a variable to be estimated. In this paper the clinical analysis presented was intended as a proof-of-concept study to show the feasibility of using IRES for epilepsy source imaging. Further investigation in a large number of patients and also with different number of electrodes, needs to be done in the future. Careful comparison of the IRES results to existing techniques is also necessary for future studies, although we have presented a comparison with cLORETA in one case (as an example) (Fig. S11 supplementary materials). Some recent literature in identifying extended sources of inter-ictal spikes is noteworthy (Chowdhury et al., 2013; Heers et al., 2015). In these works the MEM-type optimization alongside cortical parcelization is proposed to find extended sources. Testing IRES on Public Data Sets In order to further evaluate IRES, the algorithm was tested on the Brainstorm epilepsy data (Tadel et al., 2011) which is publicly available at (http://neuroimage.usc.edu/brainstorm). This tutorial includes the anonymous data of a patient suffering from focal fronto-parietal epilepsy, who underwent surgery and is seizure-free within a 5 year follow-up duration (http://neuroimage.usc.edu/brainstorm/Tutorials/Epilepsy). The data in this tutorial were originally analyzed and published in a paper by Dümpelmann et al. (Dümpelmann et al., 2012). The patient underwent iEEG recording prior to surgery. The iEEG study results as well as the post-operational MRI are not available in the data set but are presented in the published paper (Dümpelmann et al., 2012). The procedure outlined in the Brainstorm tutorial was followed to get the average spikes and the head model, with the exception of the head model conductivity (ratio) for which we used the conductivities we have used so far, throughout the paper. Fig. S12 (supplementary materials) shows the IRES solutions in this data set. Comparing these results with the clinical findings reported in (Dümpelmann et al., 2012), the obtained results are in well accordance with such findings. Additionally the source localization is performed using the cMEM algorithm (Grova et al., 2006) on another Brainstorm tutorial (http://neuroimage.usc.edu/brainstorm/Tutorials/EpilepsyBest?highlight=%28cMEM%29). Comparing IRES and cMEM results, it can be seen that the two solutions are concordant. Discussion In this paper a new inverse algorithm is proposed, namely the iterative reweighted edge sparsity (IRES). As the simulation results suggest, this algorithm is capable of distinguishing between sources with different sizes. The simulation results suggest that IRES not only localizes the underlying source but can also provide an estimate of the extent of the underlying source, as well. Moreover one of the main merits of IRES is that it produces extended sources without the need for any kind of thresholding. The initial clinical evaluation study in focal epilepsy patients also shows good concordance between clinical findings (such as resection and SOZ) with IRES solution. This suggests the practicality of IRES in clinical applications such as pre-surgical planning for pharmacoresistant focal epilepsy. Although we tested the IRES in imaging focal epilepsy sources, IRES is not limited to epilepsy source localization, but is applicable to imaging brain sources in other disorders or in healthy subjects from noninvasive EEG (or MEG). Merits and novelty of IRES and parameter selection Many algorithms have been proposed in the recent years that work within the sparse framework and are thus capable of producing relatively focal solutions. Some of these methods enforce sparsity on multiple domain like IRES (Chang et al., 2010; Haufe et al., 2008; Zhu et al., 2014), but neither of the aforementioned algorithms provide solutions with clear edges between background and desired activity and thus determining how to discard the tails of the solution distribution, i.e. thresholding, is difficult. IRES, on the other hand, achieves this by imposing sparsity on the spatial gradient which in turn creates visible edges. It is essential to note that some of the aforementioned algorithms do not intend to find the extent of the underlying sources like IRES and instead aim to model other physiologically plausible characteristics. Furthermore, it should not be assumed that all the existing methods in the literature resort to thresholding to separate brain activity from noisy background activity. Examples of these methods are the Gaussian dictionary based method by Haufe et al. (Haufe et al., 2011) and the Bayesian methods (Grova et al., 2006; Chowdhury et al., 2013; Lina et al., 2014). Algorithms formulated and operating within the Bayesian framework seem to be promising algorithms, such as MEM-type algorithms (Grova et al., 2006). Additionally the combination of cortex parcelization with these algorithms (Chowdhury et al., 2013; Lina et al., 2014) makes it even stronger. These algorithms use Otsu’s thresholding method (Otsu, 1979) to separate the background noise from active sources objectively. However, there is an implicit assumption in Otsu’s method that classes (say, background and desired signal) are distinguished enough, to be separated well with a threshold. In our experience this depends on the inverse method used. Most conventional methods do not provide strong discriminants. This means that the threshold might not be “unique” in practice. Furthermore, no parcelization is used in IRES. Additionally IRES does not work within the Bayesian framework and is formulated as a series of convex optimization problems. Bayesian methods usually have complex formulations and take long to run. Mixed norm methods have also proven to be effective in analyzing spatio-temporal activity of underlying brain sources (Gramfort et al., 2012; Gramfort et al., 2013a). However as mentioned before these methods enforce sparsity on solution and are thus highly focused (Gramfort et al., 2013a). IRES basically operates within a TV-L1 framework. This means that the sparsity is enforced on the edges as well as the solution itself. Within this framework data driven iterative reweighting are applied to IRES estimates at each step to get rid of small amplitude dipoles to obtain more accurate estimates. The formulation of IRES is simple yet effective and as tested by our extensive simulations, provides useful information about the source extent. Solutions derived based on sparse signal theory are proven to be mathematically optimal under certain mathematical conditions (Candès et al., 2006a) whether sparsity is applied on the solution or another appropriate domain such as wavelet coefficients (Candès et al., 2006b). This means that no other algorithm can provide solutions that are fundamentally better. Although these mathematical conditions do not hold for the electrophysiological source imaging (ESI) problem (due to ill-conditioned lead field matrices), still there is an increasing trend in applying sparse methods to ESI problems in recent years as indicated by the recent literature and the results presented in this work. Another feature of IRES is its iterative reweighting technique. This procedure has enabled IRES to improve the estimation by disregarding the locations that are more probable to lie outside the extent of the underlying source. Since the amplitude of the dipoles corresponding to locations outside the underlying source is smaller than dipoles closer to or within the underlying source (this is due to the formulation of the problem where spatially focused sources with zero background are preferred) it is reasonable to focus less on locations for which the associated dipoles have very low amplitudes. Additionally, after a few iterations, IRES converges to a solution which is zero in the background, i.e. a focused solution is obtained, so an extended solution is reached without applying any threshold. Due to the limited number of iterations and mostly due to the way existing convex optimization problems solvers work (Bolstad et al., 2009) a perfect zero background might not be obtained but the amplitude of dipoles located in the background is typically less than 3% of dipoles with maximum amplitude. The two features of IRES, namely the iteration and use of sparse signal processing techniques, makes IRES unique and can also explain the good performance of IRES in providing extended solutions which estimate the extent of the underlying source well. The parameters of the SOCP optimization problem in (3) need to be selected carefully, for IRES to work well. As discussed previously the L-Curve approach is adopted to determine α. The L-Curve in general is a tool to examine the dependency of level sets of the optimization problem to a certain parameter of the problem when the constraints are fixed. In our problem when β is determined (β intuitively determines noise level and is calculated based on the discrepancy principle) the constraint is fixed and then for different values of α the optimization problem is solved to find the optimum solution of the optimization problem for the given α. For the obtained solution j* the pair (\|\|j\|\|1* and \|\|Vj\|\|1) are graphed in a plot like Fig. 2 (\|\|j\|\|1 and \|\|Vj\|\|1are the two terms of the goal function in (2)). The curve obtained for different values of α (level set of the goal function) which are the best pair of (\|\|j\|\|1, \|\|Vj\|\|1) that can be achieved given a fixed constraint and a fixed α value, can show the dependency of the two terms on parameter α. If the curve has a clear pointed bending (knee), selecting the value of α around the knee is equal to selecting the Pareto-optimal pair (this means that for any other value of α that is selected, either \|\|j\|\|1 or \|\|Vj\|\|1 will be larger than the values of \|\|j\|\|1 or \|\|Vj\|\|1 of the knee). In this manner the (Pareto) optimal value of α is selected (Boyd and Vandenberghe, 2004). Another way to think about this is to imagine that for a fixed budget (β) the dependence curve of costs (\|\|j\|\|1 and \|\|Vj\|\|1) are obtained (also called “utility curve” in other fields such as microeconomics). The best way to minimize costs is to find the cost pair (\|\|j\|\|1 and \|\|Vj\|\|1) for which all other pairs are (Pareto) greater, meaning that either one will be larger (compared to \|\|j\|\|1 and \|\|Vj\|\|1). In the framework of Pareto-optimality the L-Curve formed from the level sets of the optimization problem, can determine the optimal value of α. The hyper-parameter β is determined using the discrepancy principle (Morozov, 1966). β ultimately determines the probability of capturing noise. Another way to look at this is to note that β determines how large the constraint space will be. In other words, selecting a larger β corresponds to searching for an optimal solution within a larger solution space; this translates into relaxing the parameters of the optimization problem. The manner in which the hyperparameters of IRES are defined is intuitive and bares physical meaning and thus tuning the hyperparameters can be done easily and objectively (as opposed to hyperparameters that are merely mathematical). Model Validity Considering the domain in which sparsity is imposed in IRES one might wonder if this model is completely right, that is to assume that underlying sources are spatially focused activities with constant value within their extent. It is not easy to answer this question. In practice EEG/MEG signals arise from the mass response of a large number of neurons which fire synchronously. Thus it is reasonable to assume that variations within the source extent, are hard to be detected from EEG/MEG recordings. Even in data recorded from intracranial EEG the region defined as SOZ by the physician seem to have a uniform activity, i.e., piecewise continuous activity over the recording grid (Lu et al., 2012b). From a simplistic modeling point of view (and as a first step) the assumption is defendable. Face-based wavelets (Zhu et al., 2014) and spherical wavelet transforms (Chang et al., 2010) have also been used to model sources but none of the clinical data analysis presented in these papers shows solutions with varying amplitudes. Ill-posedness of the problem and geometrical complexities of the cortex Note that while IRES provides information regarding the extent of the underlying source, it is still dealing with an underdetermined system and thus the extent information is not exact. Referring to Figs. 4 to 8 it can be seen that there is still a noticeable amount of variance in extent estimation (almost half of the true extent). Yet the fact that the general trend of IRES in estimating the extent is correct and distinguishes between sources with extents as small as 8 mm to as large as 50 mm is encouraging. Sensitivity to source location, depth and orientation The accuracy of ESI results can vary depending on the source location and orientation. It is well-known that deep sources are more difficult to resolve than superficial sources. Additionally tangential source orientations might be more difficult for EEG to detect (as normal orientations are more difficult to be detected by MEG). This adds further complexity to the already difficult inverse problem. IRES is not different from all other ESI algorithms in this aspect and will not function well under every circumstance. Some difficult cases where the simulated sources were deep or more tangentially located (in the inter-hemisphere wall or on the medial wall of the temporal lobe) are presented as examples in Fig. 12 (these cases are included in the statistical results presented so far). As it can be seen, specifically in the first row image, in highly noisy conditions (SNR of 6 dB, third column), IRES did not do very well in determining the extent or shape of the source. Still IRES does not totally fail in determining the location and extent of the source in these very difficult conditions. In less noisy conditions (second row), IRES does well. Future works and improvements on IRES In this work the focus was not to develop a specific solver for IRES. It is more a proof-of-concept project where the capabilities and usefulness of IRES are evaluated. As a result the CVX software which is a general solver for convex optimization problems was used. To name a few of the recent solvers and algorithms that have gained attention in the recent years one should mention the alternating direction method of multipliers (ADMM) (Boyd et al., 2011) and the fast iterative shrinkage thresholding algorithm (FISTA) (Beck and Teboulle, 2009). Implementing these algorithms for solving IRES can improve the speed and efficiency of the solver (compared to general solvers such as CVX). Referring to the clinical data analyzed in this study it is observed that the area of the estimated source is on average 2 times the SOZ area while comparable to resected area. Although we did not attempt to directly answer whether IRES can provide estimates that are comparable to SOZ or not, this is an important question that needs further investigation. This does not solely depend on IRES or the inverse algorithm per se, but also to the input fed into the inverse algorithm. In the clinical data analysis presented here, inter-ictal spikes have been analyzed. It is generally agreed upon that inter-ictal spikes arise from the irritative zone which is known to be larger than the SOZ (Rosenow & Luders, 2001). High frequency oscillations on the other hand are shown to be very focal and more concordant with SOZs compared to inter-ictal spikes (Worrell et al., 2008). Lu et al. performed a study to show that source localization based on HFO’s detected in scalp EEG are more accurate than inter-ictal spikes (Lu et al., 2014). Thus it is interesting to extract HFOs from scalp EEG recordings of focal epilepsy patients and feed them into IRES to see if better results can be achieved; better in the sense that the estimated solution area is comparable to SOZ. The idea of trying to determine the underlying epileptic source is important since one third of epilepsy patients do not respond to medication (Cascino, 1994). Surgical resection is a viable option for patients with focal epilepsy within this pharmacoresistant population. Currently the gold standard is to use intracranial EEG to determine SOZ (Engel, 1987). This is highly invasive with all the risks associated with such procedures. Being able to non-invasively determine the SOZ and assist the physician in determining the location and size of the epileptogenic tissue can improve the quality of life for many patients. In the current work, only a limited number of patients were studied as a proof-of-concept for potential clinical application of IRES to localize and image epileptogenic zone in patients. Further investigation in a large number of patients is needed to determine the usefulness of IRES in aiding pre-surgical planning in epilepsy patients. Additionally, it would be interesting to study the effect of electrode number on solution precision (within the new framework). While there are several studies suggesting that increasing the number of electrodes helps improve source localization results significantly (Brodbeck et al., 2011; Lantz et al., 2003; Michel et al., 2004a; Sohrabpour et al., 2015; Srinivasan et al., 1998), determining the relation between the number of electrodes and solution precision awaits future experimentation. Whether plateauing effects will be observed as suggested by Sohrabpour et al. (Sohrabpour et al., 2015) or not, can only be determined after further investigations (specifically in clinical data). In the current work, the solution space was limited to the cortical space. It is necessary to investigate if IRES can be generalized to include a solution space that encompasses the three dimensional brain volume. The presented form of IRES in this paper was intended for single time-points as opposed to spatio-temporal analysis. This is due to the fact that it was intended to show the feasibility of this algorithm and its applicability in real data recordings. Spatio-temporal algorithms (Gramfort et al., 2013a; Ou et al., 2009) are important as the dynamics of the underlying brain sources captured by EEG/MEG has to be studied properly to better understand brain networks. Following Ou et al. (Ou et al., 2009), a temporal basis can be extracted from the EEG recordings onto which the data is projected. This basis can be derived from principle component analysis (PCA), independent component analysis (ICA) and time-frequency analysis of the data (Gramfort et al., 2013a; Yang et al, 2011). In any case, IRES can be incorporated into the spatio-temporal basis and by no means is limited to single time-points, at all. We present here the IRES strategy and results for single time-points source imaging as this problem is fundamental to spatio-temporal source imaging. Spatio-temporal IRES imaging needs further investigation and will be pursued in the future. Conclusion We have proposed the iteratively reweighted edge sparsity minimization (IRES) strategy for estimating the source location and extent from EEG/MEG. We demonstrated, using sparse signal processing techniques, that it is possible to extract information about the extent of the underlying source objectively. The merits of IRES have been demonstrated in a series of computer simulations and tested in epilepsy patients undergoing intracranial EEG recordings and surgical resections. The present simulation and clinical results indicate that IRES provides source solutions that are spatially extended without the need to threshold the solution to separate background activity from active sources under study. This gives IRES a unique standing within the existing body of inverse algorithms. The present results suggest that IRES is a promising algorithm for source extent estimation from noninvasive EEG/MEG, which can be applied to epilepsy source imaging, determining the location and extent of the underlying epileptic source, as well as other brain source imaging applications. Supplementary Material 1 2 3 4 5 6 7 8 9 10 11 12 The authors would like to thank Dr. Benjamin Brinkmann and Cindy Nelson for assistance in clinical data collection, and Dr. Lin Yang for useful discussions. This work was supported in part by NIH EB006433, EY023101, HL117664, and NSF CBET-1450956, CBET-1264782. Appendix A Weighting Strategy for IRES Assuming that our problem is the following, where C is a convex set: (A1) xest=argminx\|\|x\|\|1subjecttox∈C It is reformulated as follows where ε is a positive number, (A2) argminln(∣x∣ε+1)x∈C In order to solve (A2) we linearize the logarithm function about the solution obtained in the previous step using the Taylor expansion’s series as follows, (A3) xk+1=argminxln(∣x∣ε+1)xεCln(∣x∣ε+1)=ln(∣xk∣ε+1)+1ε·(1∣xk∣ε+1)·(x-xk)+O(x2) Substituting the linearized logarithm into the optimization problem and noting that xk is treated as a constant (as the minimization is with respect to x) the following optimization problem is achieved, (A4) xk+1=argminx(x∣xk∣ε+1)=argminxWkxs.t.xεCWk=(x∣xk∣ε+1) Note that x was treated as a scalar here. In our case where x is a vector the weighting is derived for each element individually and finally placed into a diagonal matrix format. This is how (3) is derived. Treating (Vj)as a vector, i.e. y = Vj, the same procedure can be followed to update Wd. Appendix B Vector-based iteratively reweighted edge sparsity minimization (VIRES) Applying the same idea as in IRES to the case where the orientation is not fixed can be easily done. It is also possible to assume that each of the three dimensions of the dipole is an independent variable; in that case following the procedure described in Appendix A and (3) needs to be followed. Another way to approach the problem is to penalize the amplitude of the vector in the penalization terms as the variable under study is a vector now. It is necessary to note that matrix V which approximates the discrete gradient must be expanded three time using the Kronecker product, VNew = V ⊗ I3 where I3 is the 3×3 identity matrix. Following what was derived in (3) it is easy to get the following, (B1) jL=argminj∑i=1TWdL-1(i,i)\|\|VNew(i,:)j\|\|2+α∑p=1nWL-1(p,p)\|\|[jp(x),jp(y),jp(z)\|\|2subjectto(φ-Kj)T∑-1(φ-Kj)≤βWL,WdL→UpdatedbasedonjL Where T is the number of edges and n is the number of dipole locations. Also VNew(i,:) denotes the rows of VNew that correspond to the ith edge. These correspond to the 3i-2th to the 3*ith rows of VNew. Following the steps in Appendix A it is easy to obtain the update rule for the weights at iteration L as follows, (B2) WL(p,p)=(1\|\|[jpL-1(x),jpL-1(y),jpL-1(z)]\|\|2ε+1)forp=1,2,…,n (B3) WdL(i,i)=(1\|\|VNew(i,:)jL-1\|\|2ε+1)fori=1,2,…,T Fig. 1 Schematic diagram of the proposed method. Two novel strategies (edge sparse estimation and iterative reweighting) were proposed to accurately estimate the source extent. The edge sparse estimation is based on the prior information that source is densely distributed but the source edge is sparse. The source extent can thus be obtained by adding the edge-sparse term into the source optimization solution. The iterative reweighting is based on a multistep approach. Initially an estimate of the underlying source is obtained. Consequently the locations which have less activity (smaller dipole amplitude) are penalized based on the solutions obtained in previous iterations. This process is continued until a focal solution is obtained with clear edges. Fig. 2 Selecting the hyper-parameter α using the L-curve technique. In order to select α which is a hyper-parameter balancing between the sparsity of the solution and gradient domain, the L-curve technique is adopted. As it can be seen a large value of α will encourage a sparse solution while a small value of α encourages a piecewise constant solution which is over-extended. The selected α needs to be a compromise. Looking at the curve it seems that an α corresponding to the knee is optimum as perturbing α will make either of the terms in the goal function grow and thus would not be optimal. The L-curve in this figure is obtained when a source with an average radius of 20 mm was simulated. The SNR of the simulated scalp potential is 20 dB. Fig. 3 The effect of iteration. A 10 mm source is simulated and IRES estimation at each iteration, is depicted. As it can be seen the estimated solution converges to the final solution after a few iterations and more so the continuation of the iterations does not affect the solution, i.e. shrink it. The bottom right graph shows the norm of the solution (blue) and the gradient of the solution (green) and also the goal function (red) at each iteration. The goal function (penalizing terms) is the term minimized in (2). Fig. 4 Simulation results. In the left panel three different source sizes were simulated with extents of 10 mm, 20 mm and 30 mm (lower row). White Gaussian noise was added and the inverse was solved using the proposed method. The results are shown in the top row. The same procedure was repeated for random locations over the cortex. The extent of the estimated source is compared to that of the simulated source in the right panel. The SNR is 20 dB. Fig. 5 Simulation statistics. The performance of the simulation study is quantified using the following measures, localization error (upper left), AUC (upper right) and the ratio of the area of the overlap between the estimated and true source to either the area of the true source or the area of the estimated source (lower row). The SNR is 20 dB in this study. The simulated sources are roughly categorized as small, medium and large with average radius sizes of 10 mm, 20 mm and 30 mm, respectively. The LE, AUC and NOR are then calculated for the sources within each of these classes. The boxplots show the distribution of each metric to provide a brief statistical review of the distribution of these metrics for all of the data. For more explanation about the metrics and how to interpret them please refer to the methods section of the paper. Fig. 6 Simulation statistics and performance of the simulation study when the SNR is 10 dB. Results are quantified using the following measures, localization error, AUC and the ratio of the area of the overlap between the estimated and true source to either the area of the true source or the area of the estimated source. The statistics are shown in the left panel. In the right panel, the relation between the extent of the estimated and simulated source is delineated (top row). Two different source sizes namely, 10 mm and 15 mm, were simulated and the results are depicted in the right panel (bottom row). Fig. 7 Monte Carlos simulations for differing BEM models with 6 dB SNR (IRES). The estimated source extent is graphed against the true (simulated source) extent (A). Two examples of the target (true) sources (B) and their estimated sources (C) are provided. The localization Error (D), Normalized overlaps defined as overlap area over estimated source area (F) and overlap area over true source area (F), are presented to evaluate the performance of IRES (all data). The boxplots show the distribution of each metric to provide a brief statistical review of the distribution of these metrics for all of the data. For more explanation about the metrics and how to interpret them please refer to the methods section of the paper. Fig. 8 Monte Carlos simulations for differing BEM models with 20 dB SNR (IRES). The estimated source extent is graphed against the true (simulated source) extent (A). Two examples of the target (true) sources (B) and their estimated sources (C) are provided. The localization Error (D), Normalized overlaps defined as overlap area over estimated source area (F) and overlap area over true source area (F), are presented to evaluate the performance of IRES (all data). The boxplots show the distribution of each metric to provide a brief statistical review of the distribution of these metrics for all of the data. For more explanation about the metrics and how to interpret them please refer to the methods section of the paper. Fig. 9 Model violation scenarios. Examples of IRES performance when Gaussian sources (left panels) and multiple active sources (right panel) are simulated as underlying sources for a 6 dB SNR. Simulated (true) sources are depicted in the top row and estimated sources on the bottom row. More detailed analysis is provided in the supplementary materials. Fig. 10 Source extent estimation results in a patient with temporal epilepsy. (A) Scalp EEG waveforms of the inter-ictal spike in butterfly plot on top of the mean global field power (in red). (B) The estimated solution at Peak time (by VIRES) is shown on top of the ECoG electrodes and SOZ (left) and the surgical resection (right). (C) Scalp potential maps and estimation results of source extent at different latency of the interictal spike. Fig. 11 Source extent estimation results in all patients. (A) Estimated results by IRES in a parietal epilepsy patient compared with SOZ determined from the intracranial recordings (middle) and surgical resection (right). (B) Estimation results of source extent computed by VIRES in another temporal epilepsy patient compared with surgical resection. (C) Summary of quantitative results of the source extent estimation by calculating the area overlapping of the estimated source with SOZ and resection. The overlap area is normalized by either the solution area or resection/SOZ area. Fig. 12 IRES sensitivity to source location and depth. Simulation results for four difficult cases are presented in this figure for two SNRs, i.e. 20 and 6 dB. The sources were simulated in the medial wall located in the interhemispheric region, medial temporal wall and sulci wall. The orientation of some of these deep sources is close to tangential direction. Highlights A new inverse imaging strategy suitable for estimating extended sources from EEG/MEG is proposed. The sparsity of the source is exploited in multiple domains using an iterative method. No thresholding is required to obtain extended-source solutions. The proposed algorithm can estimate the source extent within reasonable error bounds. A potential application of the method is to estimate the source extent in epilepsy patients. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. Adde G Clerc M Keriven R 2005 Imaging methods for MEG/EEG inverse problem International Journal of Bioelectromagnetism 7 2 111 114 Auranen T Nummenmaa A Hämäläinen M Jääskeläinen I Lampinen J Vehtari A Sams M Bayesian inverse analysis of neuromagnetic data using cortically constrained multiple dipoles Human brain mapping 2007 28 10 979 994 17370346 Auranen T Nummenmaa A Hämäläinen M Jääskeläinen I Lampinen J Vehtari A Sams M Bayesian analysis of the neuromagnetic inverse problem with p-norm priors NeuroImage 2005 26 3 870 884 15955497 Bai X He B Estimation of independent brain electric sources from the scalp EEGs IEEE Trans Biomed Eng 2006 53 10 1883 1892 17019851 Baillet S Mosher J Leahy R Electromagnetic Brain Imaging IEEE Trans Signal Process 2001 18 14 30 Beck A Teboulle M A fast iterative shrinkage-thresholding algorithm for linear inverse problems SIAM J Imaging Sci 2009 2 1 183 202 Bolstad A Van Veen B Nowak R Space-Time event sparse penalizaion for magneto-/electroencephalography NeuroImage 2009 46 1066 1081 19457366 Boyd S Vandenberghe L Convex Optimization Cambridge Cambridge University Press 2004 Boyd S Parikh N Chu E Peleato B Eckstein J Distributed optimization and statistical learning via the alternating direction method of multipliers Found Trends Mach Learn 2011 3 1 122 Brodbeck V Spinelli L Lascano A Wissmeier M Vargas M Vulliemoz S Seeck M Electroencephalographic source imaging: a prospective study of 152 operated epileptic patients Brain 2011 134 10 2887 2897 21975586 Candès E Romberg J Tao T Stable signal recovery from incomplete and inaccurate measurements Commun Pure Appl Math 2006a 59 1207 23 Candès E Romberg J Tao T Robust Uncertainty Principle: Exact Signal Reconstruction From Highly Incomplete Frequency Information IEEE Trans Inf Theory 2006b 52 489 509 Candès E Wakin M Boyd S Enhancing Sparsity by Reweighting L1 Minimization J Fourier Anal Appl 2008 14 877 905 Cascino GD Commentary: how has neuroimaging improved patient care? Epilepsia 1994 35 S103 7 8206009 Chang W Nummenmaa A Hsieh J Lin F 2010 Spatially sparse source cluster modeling by compressive neuromagnetic tomography Neuroimage 53 146 160 20488248 Chowdhury R Lina J Kobayashi E Grova C MEG source localization of spatially extended generators for epileptic activity: comparing entropic and hierarchical Bayesian approaches PLoS ONE 2013 8 1 9 Dale A Sereno M Improved localization of cortical activity by combining EEG and MEG with MRI cortical surface reconstruction: a linear approach J Cog Neurosci 1993 5 162 176 Ding L He B Sparse Source Imaging in EEG with Accurate Field Modeling Human Brain Mapping 2008 29 1053 67 17894400 Ding L Reconstructing cortical current density by exploring sparseness in the transform domain Phys Med Biol 2009 54 2683 2697 19351982 Dohmatob E Gramfort A Thirion B Varoquaux G 2014 Benchmarking solvers for TV-l1 least-squares and logistic regression in brain imaging PRNI http://dx.doi.org/10.1109/PRNI.2014.6858516 Donoho DL Compressed Sensing IEEE Trans Inf Theory 2006 52 1289 1306 Dümpelmann M Ball T Schulze-Bonhage A 2012 sLORETA allows reliable distributed source reconstruction based on subdural strip and grid recordings Hum Brain Mapp 33 5 1172 1188 21618659 Engel J Jr Approaches to localization of the epileptogenic lesion Engel J Jr Surgical treatment of the epilepsies New York Raven Press 1987 75 9 Fazel M Matrix rank minimization with applications PhD thesis Electrical Engineering Department, Stanford University 2002 Fazel M Hindi H Boyd S Log-det heuristic for matrix rank minimization with applications to Hankel and Euclidean distance matrices Proceedings of Am. Control Conf June 2003 Gorodnitsky IF George JS Rao BD 1995 Neuromagnetic source imaging with FOCUSS: a recursive weighted minimum norm algorithm Electroencephalogr Clin Neurophysiol 95 21 231 251 8529554 Gramfort A Kowalski M Hämäläinen M Mixed-norm estimates for the M/EEG inverse problem using accelerated gradient methods Phys Med Biol 2012 57 7 1937 1961 3 22421459 Gramfort A Strohmeier D Haueisen J Hmlinen M Kowalski M Time– frequency mixed-norm estimates: sparse M/EEG imaging with non-stationary source activations NeuroImage 2013a 70 410 422 23291276 Gramfort A Thirion B Varoquaux G 2013b Identifying predictive regions from fMRI with TV-L1 prior Workshop on Pattern Recognition and NeuroImaging Grant M Boyd S Graph implementations for non-smooth convex programs, Recent Advances in Learning and Control (a tribute to M. Vidyasagar) Blondel V Boyd S Kimura H 95 110 Lecture Notes in Control and Information Sciences Springer 2008 Grant M Boyd S 2013 CVX: MATLAB Software for Disciplined Convex Programming http://cvxr.com/cvx/ Grova C Daunizeau J Lina J Bénar C Benali H Gotman J Evaluation of EEG localization methods using realistic simulations of interictal spikes Neuroimage 2006 29 734 53 16271483 Hämäläinen M Ilmoniemi R Interpreting measured magnetic fields of the brain: estimates of current distribution Helsinki University of Technology, Dept. of Technical Physics Report TKK-F-A559 1984 Hämäläinen M Ilmoniemi R Interpreting magnetic fields of the brain: Minimum norm estimates Medical and Biological Engineering and Computing 1994 32 35 42 8182960 Hansen PC 1990 Truncated singular value decomposition solutions to discrete ill-posed problems with ill-determined numerical rank SIAM J Sci Stat Comput 11 503 518 Haufe S Nikulin V Ziehe A Muller K Nolte G Combining sparsity and rotational invariance in EEG/MEG source reconstruction NeuroImage 2008 42 726 38 18583157 Haufe S Tomioka R Dickhaus T Sannelli C Blankertz B Nolte G Müller K Large-scale EEG/MEG source localization with spatial flexibility Neuroimage 2011 54 851 859 20832477 He B Musha T Okamoto Y Homma S Nakajima Y Sato T Electric dipole tracing in the brain by means of the boundary element method and its accuracy IEEE Trans Biomed Eng 1987 34 406 14 3610187 He B Yao D Lian J Wu D An Equivalent Current Source Model and Laplacian Weighted Minimum Norm Current Estimates of Brain Electrical Activity IEEE Trans on Biomed Eng 2002 49 277 288 He B Yang L Wilke C Yuan H Electrophysiological imaging of brain activity and connectivity – challenges and opportunities IEEE Trans Biomed Eng 2011 58 1918 31 21478071 He B Ding L Electrophysiological mapping and neuroimaging He B Neural Engineering New York Springer 2013 499 544 He B Coleman T Genin G Glover G Hu X Johnson N Grand challenges in mapping the human brain: NSF workshop report IEEE Trans Biomed Eng 2013 60 2983 92 24108705 Heers M Chowdhury R Hedrich T Dubeau F Hall J Lina J … Kobayashi E Localization accuracy of distributed inverse solutions for electric and magnetic source imaging of interictal epileptic discharges in patients with focal epilepsy Brain topography 2015 1 20 25526854 Huang MX Dale AM Song T Halgren E Harrington DL Podgorny I Canive JM Lewise S Lee RR 2006 Vector-based spatial-temporal minimum L1-norm solution for MEG NeuroImage 31 1025 1037 16542857 Kay S 2011 Fundamentals of Statistical Signal Processing: Detection Theory 15 Prentice-Hall New Jersey Lai D Drongelen W Ding L Hecox K Towle V Frim D He B Estimation of in vivo human brain-to-skull conductivity ratio from simultaneous extra- and intra-cranial electrical potential recordings J Clin Neurophysiol 2005 116 456 65 Lantz G Grave de Peralta R Spinelli L Seeck M Michel C Epileptic source localization with high density EEG: how many electrodes are needed Clin Neurophysiol 2003 114 63 9 12495765 Lawson CL Hanson RJ 1974 Solving Least Squares Problems Prentice Hall New York Liao K Zhu M Ding L Valette S Zhang W Dickens D 2012 Sparse representation of cortical current density maps using wavelets Phys Med Biol 57 6881 6901 23038163 Limpiti T Van Veen B Wakai R Cortical patch basis model for spatially extended neural activity IEEE Trans Biomed Eng 2006 53 9 1740 1754 16941830 Lina J Chowdhury R Lemay E Kobayashi E Grova C Wavelet-based localization of oscillatory sources from magnetoencephalography data IEEE Trans Biomed Eng 2014 61 8 2350 2364 22410322 Lu Y Yang L Worrell G He B Seizure source imaging by means of FINE spatio-temporal dipole localization and directed transfer function in partial epilepsy patients Clin Neurophysiol 2012a 123 1275 83 22172768 Lu Y Yang L Worrell G Brinkmann B Nelson C He B Dynamic imaging of seizure activity in pediatric epilepsy patients Clin Neurophysiol 2012b 123 2122 29 22608485 Lu Y Worrell G Zhang H Yang L Brinkmann B Nelson C He B Noninvasive imaging of the high frequency brain activity in focal epilepsy patients IEEE Trans Biomed Eng 2014 61 1660 67 24845275 Malioutov D Çetin M Willsky AS 2005 A sparse signal reconstruction perspective for source localization with sensor arrays IEEE Trans Signal Proces 55 3010 3022 Matsuura K Okabe Y 1995 Selective minimum-norm solution of the biomagnetic inverse problem IEEE Trans Biomed Eng 42 608 615 7790017 Michel C Lantz G Spinelli L Grave dePeralta R Landis T Seeck M 128-Channel EEG Source Imaging in Epilepsy: Clinical Yield and Localization Precision Clin Neurophysiol 2004a 21 71 83 Michel C Murray M Lantz G Gonzalez S Spinelli L Grave de Peralta R EEG Source Imaging Clin Neurophysiol 2004b 115 2195 2222 15351361 Michel C He B EEG Mapping and Source Imaging Schomer D Lopes da Silva F Niedermeyer’s Electroencephalography 6 Wolters Kluwer & Lippincott Williams & Wilkins Philadelphia Chapter 55 1179 1202 2011 Morozov VA 1966 On the solution of functional equations by the method of regularization Soviet Math Dokl 7 414 417 Nunez PL Silberstein RB 2000 On the relationship of synaptic activity to macroscopic measurements: Does co-registration of EEG with fMRI make sense? Brain Topogr 13 79 96 11154104 Oostendorp TF Delbeke J Stegeman DF 2000 The conductivity of the human skull: results of in vivo and in vitro measurements IEEE Trans Biomed Eng 47 1487 1492 11077742 Otsu N 1975 A threshold selection method from gray-level histograms Automatica 11 285–296 23 27 Ou W Hämäläinen M Golland P A distributed spatio-temporal EEG/MEG inverse solver Neuroimage 2008 44 932 946 18603008 Pascual-Marqui RD Michel C Lehmann D Low resolution electromagnetic tomography: a new method for localizing electrical activity in the brain Int J Psychophysiol 1994 18 49 65 7876038 Rosenow F Luders H Presurgical evaluation of epilepsy Brain 2001 124 1683 1700 11522572 Rudin L Osher S Fatemi E Nonlinear total variation based noise removal algorithms Physica D: Nonlinear Phenomena 1992 60 1 259 268 Scherg M von Cramon D A new interpretation of the generators of BAEP waves I-V: results of a spatio-temporal dipole model Electroencephalogr Clin Neurophysiol 1985 62 290 299 2408875 Sohrabpour A Lu Y Kankirawatana P Blount J Kim H He B Effect of EEG Electrode Number on Epileptic Source Localization in Pediatric Patients Clin Neurophysiol 2015 126 3 472 480 25088733 Srinivasan R Tucker D Murias M Estimating the spatial Nyquist of the human EEG Behavior Research Methods, Instruments, & Computers 1998 30 8 19 Stohmeier D Hauiesein J Gramfort A 2014 Improved MEG/EEG source localization with reweighted mixed-norms. Pattern Recognition in Neuroimaging (PRNI) 2014 International Workshop on. IEEE 2014 Stohmeier D Gramfort A Hauiesein J 2015 MEG/EEG source imaging with a non-convex penalty in the time-frequency domain Pattern Recognition in Neuroimaging (PRNI) 2015 International Workshop on. IEEE 2015 Sturm JF 1999 Using SEDUMI 1.02, a MATLAB toolbox for optimization over symmetric cones Optimization Methods and Software 11–12 625 653 Tadel F Baillet S Mosher JC Pantazis D Leahy RM 2011 Brainstorm: a user-friendly application for MEG/EEG analysis Comput Intell Neurosci 2011 879716 http://dx.doi.org/10.1155/2011/879716 (Epub 2011 April 13) 21584256 Tibshirani R Regression Shrinkage and Selection via Lasso J R Statist Soc 1996 58 267 88 Uutela K Hämäläinen M Somersalo E 1999 Visualization of magnetoencephalographic data using minimum current estimates NeuroImage 10 173 180 10417249 Vega-Hernández M Martínez-Montes E Sánchez-Bornot J Lage-Castellanos A Valdés-Sosa P Penalized least squares methods for solving the EEG inverse problem Stat Sin 2008 18 1535 1551 Wagner M Fuchs M Wischmann H Drenckhahn R Köhler T Smooth reconstruction of cortical sources from EEG or MEG recordings NeuroImage 1996 3 3 S168 Wipf D Nagarajan S A Unified Bayesian Framework for MEG/EEG Source Imaging NeuroImage 2009 44 3 947 966 18602278 Wipf D Nagarajan S Iterative reweighted l1 and l2 methods for finding sparse solutions IEEE J Sel Top Sign Proces 2010 4 2 317 329 Worrel G Gardner A Stead S Hu S Goerss S Cascino G Meyer F Marsh R Litt B High frequency oscillations in human temporal lobe: Simultaneous microwire and clinical macroelectrode recordings Brain 2008 131 928 37 18263625 Zhang Y van Drongelen W He B Estimation of in vivo human brain-to-skull conductivity ratio with the aid of intracranial electrical stimulation Appl Phys Lett 2006 89 1 3 Zhu M Zhang W Dickens DL Ding L Reconstructing spatially extended brain sources via enforcing multiple transform sparseness NeuroImage 2014 86 280 93 24103850 Zou H Hastie T Regularization and variable selection via the elastic net J R Stat Soc B Stat Methedol 2005 67 301 320
PMC005xxxxxx/PMC5124548.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101604520 41136 IEEE J Biomed Health Inform IEEE J Biomed Health Inform IEEE journal of biomedical and health informatics 2168-2194 2168-2208 27249841 5124548 10.1109/JBHI.2016.2574201 NIHMS794406 Article Integrative Analysis of Proteomic, Glycomic, and Metabolomic Data for Biomarker Discovery Wang Minkun Department of Electrical and Computer Engineering, Virginia Tech, Arlington, VA 22203, USA Yu Guoqiang Department of Electrical and Computer Engineering, Virginia Tech, Arlington, VA 22203, USA Ressom Habtom W. Senior Member, IEEE Department of Oncology, Georgetown University, Washington, DC 20057, USA Corresponding author: Habtom W. Ressom ([email protected]) 14 6 2016 27 5 2016 9 2016 01 9 2017 20 5 12251231 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Studies associating changes in the levels of multiple biomolecules including proteins, glycans, glycoproteins, and metabolites with the onset of cancer have been widely investigated to identify clinically relevant diagnostic biomarkers. Advances in liquid or gas chromatography mass spectrometry (LC-MS, GC-MS) have enabled high-throughput qualitative and quantitative analysis of these biomolecules. While results from separate analyses of different biomolecules have been reported widely, the mutual information obtained by partly or fully combining them has been relatively unexplored. In this study, we investigate integrative analysis of proteins, N-glycans, and metabolites to take advantage of complementary information to improve the ability to distinguish cancer cases from controls. Specifically, SVM-RFE algorithm is utilized to select a panel of proteins, N-glycans, and metabolites based on LC-MS and GC-MS data previously acquired by analysis of blood samples from two cohorts in a liver cancer study. Improved performances are observed by integrative analysis compared to separate proteomic, glycomic, and metabolomic studies in distinguishing liver cancer cases from patients with liver cirrhosis. Index Terms multi-omic data integration machine learning systems biology cancer biomarker discovery I. INTRODUCTION Characterizing the association of biomolecules such as proteins, glycans, glycoproteins, and metabolites with cancer has proven to be a promising strategy to discover candidate biomarkers. Glycosylation is one of the most common post-translational modifications of proteins. Altered patterns of glycosylation have been associated with various diseases and many currently used cancer biomarkers. In particular protein glycosylation is relevant to liver pathology because of the major influence of this organ on the homeostasis of blood glycoproteins. Characterizing glycan modifications of proteins in complex proteomes is challenging as glycosylation can occur on multiple sites of peptides involving the attachment of different glycans to each site. An alternative strategy to the analysis of glycoproteins is the study of proteins and protein-associated glycans [1, 2]. Metabolites are molecular fingerprints of what cells do at a particular point in time; they can reveal early signs of cancers when the chances for cure are highest. Because these biomolecules are members of strongly intertwined biological pathways and are highly interactive with each other, integrative analysis offers a great opportunity to help interpret such interactions and to identify reliable biomarkers. We previously performed separate analyses of proteins and N-linked glycans released from proteins in blood by using liquid chromatography coupled with mass spectrometry (LC-MS) [3]. Also, we used gas chromatography coupled with mass spectrometry (GC-MS) to analyze metabolites in blood [4]. We detected proteins, N-glycans, and metabolites significantly altered in hepatocellular carcinoma (HCC) cases compared to patients with liver cirrhosis using univariate statistical methods. However, multivariate statistical or machine learning methods are desirable to improve the ability to discriminate the cases from controls by taking advantage of the mutual information within the molecules detected by a single omic study as well as the combination of molecules from multiple omic studies. The integrative analysis will allow us to investigate if the synergy of the three omic studies leads to improved performance in distinguishing cases from controls compared to the a single omic study. We recently reported improvement achieved in discriminating HCC cases from cirrhotic controls using a panel of proteins and N-glycans selected by integrating proteomic and glycomic datasets [5]. In this paper, we consider three datasets we previously generated by proteomic, glycomic, and metabolomic analysis of blood samples from HCC cases and patients with liver cirrhosis to identify proteins, N-glycans, and metabolites that are significantly altered in HCC versus cirrhosis. The goal of this research is to evaluate the improvement in disease classification achieved by integrating the data from the three studies. To select multi-omic based features that lead to highly discriminant classification, we used a model, in which feature selection and classification methods are embedded. To accomplish this, we chose support vector machine-recursive feature elimination (SVM-RFE) [6] due to its wide application and flexibility to use as an embedded method that helps recognize relevant patterns in the feature space, while reducing dimensionality to overcome the risk of overfitting. Through a 10-fold cross-validation, we evaluated the classification performances of the features selected from each omic study as well as the combined features. In addition, we split the samples into training and test sets to evaluate the performance of the selected features on an independent set. We observed that improved performances can be achieved through the integrative analysis compared to a single omic study. The remaining part of this paper is organized as follows. Section II briefly summarizes the experimental design used for acquisition of proteomic, glycomic, and metabolomic datasets. Also, this section describes our feature selection and disease classification methods based on datasets acquired by the three omic studies. Section III presents the results we obtained in selecting optimal features from each omic study as well as the integrated multi-omic dataset. Section IV concludes the paper with summary and future goals. II. Materials and Methods A. Experimental Design The proposed integrative analysis is performed on LC-MS-based proteomic and glycomic datasets and GC-MS-based metabolomic dataset we acquired by analysis of blood samples from HCC cases and patients with liver cirrhosis recruited in Egypt and the U.S. [7, 8]. The participants in Egypt and the U.S. were recruited through protocols approved by the Ethics Committee at Tanta University Hospital and the Institutional Review Board at Georgetown University, respectively. Specifically, adult patients were recruited from the outpatient clinics and inpatient wards of the Tanta University Hospital (TU cohort) in Tanta, Egypt and from the hepatology clinics at MedStar Georgetown University Hospital (GU cohort) in Washington, DC, USA. The TU cohort consists of a total of 89 subjects (40 HCC cases and 49 patients with liver cirrhosis), and the GU cohort comprises of 116 subjects (57 HCC cases and 59 patients with liver cirrhosis). Fig. 1 depicts the overall workflow of our experimental design. Briefly, targeted quantitative analysis of selected proteins and N-glycans in blood samples was performed by multiple reaction monitoring (MRM) using a Dionex 3000 Ultimate nano-LC system (Dionex Sunnyvale, CA) interfaced to TSQ Vantage mass spectrometer (Thermo Scientific, San Jose CA). The targets were selected from our previous LC-MS-based untargeted proteomic and glycomic analyses and by text mining. Also, metabolites selected from a previous untargeted study were subjected for a targeted analysis in blood samples by selected ion monitoring (SIM) using an Agilent 7890A GC interfaced to a single quadrupole Agilent 5975C MSD (Agilent Technologies, Santa Clara, CA). The datasets from these omic studies were analyzed using Skyline [9], GPA [10], and SIMAT [11], respectively. Results from univariate statistical analysis have been previously reported in [4, 7, 8]. In the following, we introduce how we integrate the three datasets for feature selection that lead to improved performance on disease classification. B. Feature Selection and Classification Feature selection techniques can be generally organized into three categories: filter, wrapper, and embedded methods [12]. Filter methods are efficient and scalable to high-dimensional data analysis however they ignore feature dependencies and the interaction with the classifiers. Wrapper methods consider the model hypothesis search within the feature subset selection. A common drawback of these methods is that they have a higher risk of overfitting issue than filter methods and are very computationally intensive. Embedded methods have the advantage that they include the interaction with the classification model, while at the same time being far less computationally costly than wrapper methods. Because a thorough comparison among various feature selection methods and classifiers or determination of the most suitable ones is not the primary goal of this paper, we chose an embedded method implemented in SVM-RFE due to its wide application and flexibility for high dimensional data. Linear SVMs were trained to classify samples in case and control groups using features from each of the three omic studies (proteomics, glycomics, and metabolomics) separately and by combining features from the three. Equation (1) presents the decision function in SVM model for an input sample xt. D(xt)=w·xt+b, where (1) w=∑kαkykxk and b=〈yk−w·xk〉. The feature weight vector w determined by support vectors is used as feature ranking criterion by the recursive feature elimination (RFE) algorithm [6]. SVM-RFE eliminates redundant features iteratively and yields better and more compact feature subsets. The major steps include 1) training the SVM classifier; 2) ranking the features according to weight vector w of the learned SVM; 3) eliminating features with the smallest ranking criterion; 4) retraining SVM model with the remaining features; 5) estimating the performance of the model using cross-validation to check if the optimal subset is obtained. In this paper, we applied SVM-RFE to select highly discriminative sets of proteins, N-glycans, and metabolites as well as features selected from an integrated set consisting of proteins, N-glycans, and metabolites. At each iteration, we started from the entire feature list, trained an SVM classifier with linear kernel, and estimated the average classification accuracy based on a 10-fold cross-validation. The feature with minimum weight assigned by the classifier was removed at the end of each iteration until the feature subset was empty. Additionally, we split the samples into training and test sets. The performance of the features selected using the training set were evaluated on the test set. III. Results and Discussion A. Integrative Analysis of Proteins and N-Glycans We first perform the integrative analysis between proteomic and glycomic datasets with regards to their biological relations (i.e., glycosylation). Additional integration of metabolomic dataset is evaluated in the second part to further elucidate of the benefit of integrative analysis. Datasets from targeted analyses of 101 proteins (represented by Uniprot IDs) and 82 N-glycans (characterized by the number of five monosaccharides: GlcNAc, mannose, galactose, fucose, and NeuNAc) were considered here for integrative analysis. We used SVF-RFE to select the most relevant features based on analysis of the two separate datasets obtained from targeted analysis of proteins and N-glycans and a third dataset obtained by concatenating the two datasets. Fig. 2a depicts the distributions of the LC-MS datasets from the proteomic and glycomic studies. Fig. 2b presents the log-transformed datasets that resemble normal distributions. To make the two datasets compatible for integration, we performed Z-score normalization (Fig. 2c). This step ensures features from protein and glycan lists are treated equally in the feature selection procedure by SVM-RFE. Separate SVM-RFE models were trained for each of the three datasets. We started from the whole feature list in each dataset, and eliminated one feature in each iteration step till feature set was empty. At each step, we randomly partitioned the samples into 10 subsets. We tested the performance of classifying one of the 10 subsets using the SVM classifier trained based on the other nine subsets. The average classification performance (i.e., accuracy, sensitivity, and specificity) was evaluated at each iteration step. Figs. 3a and 3b depict the classification accuracy achieved at each iteration step for the top 50 features selected from the three datasets in the TU and GU cohorts, respectively. Also, the figures show the optimal number of features that leads to the best classification accuracy. We observed that, in most iteration steps, features selected from the integrated dataset yield higher accuracies compared to the same number of features selected from either the glycomic or proteomic dataset. Receiver operating characteristics (ROC) curves were estimated by varying the SVM threshold parameter (yr = ŵ · x − b̂). The 95% confidence intervals of area under the ROC (AUC) were calculated using bootstrap method with 1000 resampled replicates. Table I shows the disease classification performance with optimal subset of features in each dataset of the TU cohort. As shown in the table, SVM-RFE selected 29 out of 82 N-glycans and 15 out of 101 proteins as the optimal number of features. Among these, 13 glycans and 5 proteins were also selected as significantly altered in cases versus controls through univariate statistical test [7, 8]. Out of 183 integrated features, 7 proteins and 2 N-glycans in a panel were selected by SVM-RFE. The panel includes 2 that were also found significant in the univariate statistical analysis. The integrative analysis led to a significantly smaller number of features with a slight improvement on the disease classification accuracy compared to those selected by analysis of individual datasets. This phenomenon is observed consistently across the entire iteration steps, as illustrated in Fig. 3a. Similar results are obtained in the GU cohort (Table II), in which SVM-RFE selected 18 proteins and 5 N-glycans in a panel yielded better performance than 22 proteins or the 8 glycans selected by analysis of individual datasets. Among the 23 features selected by the integrative analysis, four N-glycans and 10 proteins were also reported as significant by univariate statistical analysis. As shown in Fig. 3b, the integrative analysis yielded improved performance compared to the analysis based on the individual datasets in the majority of the iteration steps. In both cohorts, we captured features with synergic contributions to the discrimination, which provide complementary information to univariate analysis. Although we did not observe overlapping features between the optimal sets of features in the two cohorts, we were able to achieve AUCs greater than 0.73 when we trained SVMs based on the data the integrated panel learned from TU cohort and tested it on the GU cohort, and vice versa. In addition, we investigated the performance for each dataset by setting the feature size to five. We compared the performances of the best five features selected by SVM-RFE from each of the three datasets. While the integrative analysis outperformed the analysis based on individual dataset in TU cohort (Table III), both the integrated features and the protein features led to similar performances in the GU cohort (Table IV). B. Integrative Analysis of Proteins, N-Glycans, and Metabolites We present here the improvement in disease classification by including a dataset from a targeted analysis of 50 metabolites in blood samples. Thus, a total of 233 features (101 proteins, 82 N-glycans, and 50 metabolites) were considered for integrative analysis. The same normalization method was applied when merging features from the new dataset. Table V presents the performance of features selected by SVM-RFE from the metabolites only and the improvement achieved by combining the metabolites with proteins and glycans in the TU cohort. From the 50 metabolites, SVM-RFE selected 14 that showed better performance than those selected from the protein and N-glycan list presented in Table I on the same TU cohort representing 89 participants. A panel consisting of 10 proteins, 5 glycans, and 6 metabolites selected from the integrated dataset outperformed all other panels selected by SVM-RFE from single omic dataset or by combining proteomic and glycomic datasets. Fig. 4a shows the classification accuracy at each iteration step for the top 50 features from three single datasets and two integrated datasets. We observe that the two integrated datasets (colored in red and magenta) have overall higher classification accuracies than any of the single omic datasets. Although the addition of metabolites to proteins and N-glycans did not improve the classification accuracy when relatively smaller number of features are selected, a more stable and discriminative performance is achieved as the feature size increases. We also evaluated the classification performance of the list concatenated from three feature subsets selected by SVM-RFE separately (i.e., 29 N-glycans, 15 proteins, and 14 metabolites). This approach resulted in a classification accuracy of 0.79 and an AUC of 0.94 with 95% CI at (0.86, 0.97), which is worse than the performance of 21 features selected by combining the three omic datasets prior to application of SVM-RFE as presented in TABLE V. We performed integrative analysis of proteomic, glycomic, and metabolomics datasets acquired by analysis of blood samples from 44 subjects in the GU cohort. Since the number of overlapping samples in the three omic datasets is different from the number of overlapping proteomic and glycomic datasets reported in Tables II and IV, we repeated all multivariate analyses for appropriate comparison. Table VI presents the performances of features selected from each of the three datasets as well as two integrated datasets. A panel of 10 features consisting of 4 proteins, 3 N-glycans, and 3 metabolites led to the best performance. Seven of these 10 features were also reported previously to have shown statistically significant changes in HCC vs. cirrhosis [4, 7, 8]. As illustrated in Fig. 4b, features selected from integrated datasets tend to have the best classification accuracy in most iterations. Integration of metabolites with proteins and N-glycans improves the classification accuracy as the number of features increases. Concatenating the three proteins and ten N-glycans, with the four metabolites selected independently from each omic dataset achieves a classification accuracy of 0.97 and an AUC of 0.99, which is about the same performance obtained by the ten features selected from the integrated omic dataset in the GU cohort, which resulted in accuracy of 0.98 and AUC of 0.99. We would like to emphasize that the performance evaluations presented in Tables V and VI represent the average 10-fold cross-validation results based on all samples in each cohort. Less sensitivity and specificity are expected when the selected features are tested on an independent set due to potential overfitting issue. To address this, we evaluated the model performance by using 70% of samples (balanced in case and control groups) as a training set and the remaining 30% as a testing set. When selecting features, we used the same 10-fold cross-validation on the 70% of samples (nine subsets for training, the remaining one for validation). Though the classification accuracies on the testing set using selected features, decease in both cohorts, improved classification performance is observed by using the integrative analysis compared to a single omic study (Table VII). IV. Conclusion In this study, we investigated the benefit of an integrative analysis of proteomic, glycomic, and metabolomic datasets in improving our ability to distinguish HCC cases from patients with liver cirrhosis. Through SVM-RFE, a panel of features was selected from 101 proteins, 82 N-glycans, and 50 metabolites acquired by targeted analysis of blood samples using LC-MS and GC-MS. Complementary to univariate statistical methods, the integrative analysis utilizes mutual information among features to select a panel of features with improved ability to discriminate biologically distinct groups. In this study, we observe that features selected by merging the proteomic, glycomic, and metabolomic datasets lead to better disease classification accuracy compared to those selected from one or two of the three datasets. We would like to emphasize that the improvement achieved by the integrative analysis was observed not only in using SVM-RFE, but also through other methods such as a sequential feature selection coupled with quadratic discriminant analysis. We believe that integration of multi-omic data by multivariate statistical or machine learning methods, combined with pathway-centric and network-based approaches, will help not only in identifying a panel of biomarkers that leads to improved diagnosis but also in gaining insight into the molecular mechanisms of cancer. Research supported by NIH Grants R01CA143420 and R01GM086746. Minkun Wang received the B.S. degree in electrical engineering from University of Science and Technology of China, Hefei, China, in 2012. He is currently working toward the PhD degree in the Department of Electrical and Computer Engineering at Virginia Tech. He is also a research assistant at the Lombardi Comprehensive Cancer Center, Georgetown University. His research focuses on applications of statistical and machine learning methods for omic data analysis including LC-MS data preprocessing, multi-omic data integration, and deconvolution of heterogeneous data. Guoqiang Yu received the B.S. degree in electronic engineering from Shandong University, Shandong, China in 2001, the M.S. degree in electrical engineering from Tsinghua University, Beijing, China in 2004 and the Ph.D. degree in electrical engineering from Virginia Tech in 2011. He is currently an assistant professor in Department of Electrical and Computer Engineering at Virginia Tech. His research interests include machine learning, signal and image processing, applied statistics, and their applications to developing bioinformatics and systems genetics tools for integrated modeling and analyses of various human diseases. Habtom W. Ressom received B.Sc. and M.Sc. degrees in Electrical Engineering from Addis Ababa University, Addis Ababa, Ethiopia in 1989 and 1992, respectively, and a Ph.D. degree in Electrical Engineering from University of Kaiserslautern, Kaiserslautern, Germany, in 1999. He is currently a Professor in the Department of Oncology and the Director of the Genomics and Epigenomics Shared Resource at Georgetown University Medical Center, Washington, DC, USA. His research interests focus on cancer biomarker discovery using multi-omic approaches. Figure 1 Workflow of integrative analysis of multi-omic data. Figure 2 The distributions of raw glycomic (orange) and proteomic (cyan) datasets (a); log-transformed data (b); data after log-transformation and Z-score normalization. Figure 3 Classification accuracy at each iteration step for the top 50 features from glycomic (green), proteomic (blue), and integrated datasets (red) in the TU and GU cohorts. The optimal numbers of features (indicated by triangles) correspond to the best classification accuracy (indicated by circles). Figure 4 Classification accuracy at each iteration step for the top 50 features from proteomic (blue), glycomic (green), metabolomic (yellow), integrated proteomic and glycomic (red), and integrated proteomic, glycomic, and metabolomic (matenga) datasets in the TU and GU cohorts. TABLE I Performance Comparison Based on the Optimal Number of Features Selected in the TU Cohort TU Cohort Glycomic Proteomic Integrated (P & G) Accuracy 0.77 0.84 0.87 Sensitivity 0.82 0.83 0.90 Specificity 0.75 0.84 0.85 AUC (95% CI) 0.87 (0.78, 0.93) 0.93 (0.83, 0.97) 0.92 (0.81, 0.97) Optimal Number of Features 29/82 15/101 9/183 Selected Featuresb [25000] [43000]a P01024a [53111]a [34100]a P02743 [63402]a [43202]a,c P02750 [53313] [53000]a P02753a P02743 [53323] [33101] P02763 P02763 [34110] [63403]a P03952 P05160 [53311]c [53311]c P04004a P06727 [43110]a [53010] P05160 P0C0L4 [63413]a [53302]a P06727 P22891 a [53411] [34101] P0C0L4 P35858 [63423] [63404]a P13598a [43000] a [53312] [29000] P13796 [26000] [53101]a [73514] P22891a [53201] [43202]a,c P27918 [2 10 000] P35858 a Significant (p value ≤ 0.05) in univariate statistical analysis b N-glycans are characterized by GlcNAc, mannose, galactose, fucose, and NeuNAc, and proteins are indicated by Uniprot IDs c Isomers with different retention times The results of best performing methods are marked in bold. TABLE II Performance Comparison Based on the Optimal Number of Features Selected in the GU Cohort GU Cohort Glycomic Proteomic Integrated (P & G) Accuracy 0.77 0.88 0.91 Sensitivity 0.79 0.86 0.89 Specificity 0.75 0.91 0.93 AUC (95% CI) 0.83 (0.71, 0.91) 0.95 (0.89, 0.98) 0.96 (0.89, 0.99) Optimal Number of Features 8/82 22/101 23/183 Selected Features b O75015 O75636a O75015 O75636 a P00748a P01023a P01023a P01034a P01877a P02741 P01877a P02771a [43100]a P02766 P02771a P04278 P05155 [53313] P02790 P04278 P05452a P08294 [53000]a P05155 P05452 P13796 P41222 a [43212] P06727 P13796 P61626a Q13201a [53411] P27169a P41222a Q15848a Q96KN2 [53312] P49747a P61626a [43100]a [53313] [53200] P61769a [53000]a [43200]a [63434] Q15848a [53411] [53200] Q96KN2 [53111] a Q9Y6R7a a Significant (p value ≤ 0.05) in univariate statistical analysis b N-glycans are characterized by GlcNAc, mannose, galactose, fucose, and NeuNAc, and proteins are indicated by Uniprot IDs The results of best performing methods are marked in bold. TABLE III Performance Comparison on the Top Rankning five Featured Selected in the TU Cohort. TU cohort Glycomic Proteomic Integrated (P & G)) Accuracy 0.68 0.79 0.83 Sensitivity 0.71 0.79 0.82 Specificity 0.67 0.79 0.85 AUC (95% CI) 0.77 (0.65, 0.59) 0.88 0.77, 0.94) 0.89 (0.80, 0.95) Number of Selected Features 5/82 5/101 5/183 a Significant (p value ≤ 0.05) proteins in univariate statistical analysis. N-glycans that found significant (p value ≤ 0.05) in univariate statistical analysis are shown in boxes. The results of best performing methods are marked in bold. TABLE IV Performance Comparison on the Top Rankning five Featured Selected in the GU Cohort. GU cohort Glycomic Proteomic Integrated (P & G) Accuracy 0.74 0.80 0.80 Sensitivity 0.74 0.82 0.82 Specificity 0.75 0.79 0.79 AUC (95% CI) 0.82 (0.70, 0.89) 0.85 (0.74, 0.92) 0.87 (0.77, 0.93) Number of Selected Features 5/82 5/101 5/183 a Significant (p value ≤ 0.05) proteins in univariate statistical analysis. N-glycans that found significant (p value ≤ 0.05) in univariate statistical analysis are shown in boxes. The results of best performing methods are marked in bold. TABLE V Performance Comparison Based on the Optimal Number of Features Selected in the TU Cohort TU Cohort Metabolomics Integrated (P + G + M) Accuracy 0.86 0.90 Sensitivity 0.91 0.91 Specificity 0.84 0.89 AUC (95% CI) 0.93 (0.84, 0.97) 0.99 (0.95, 0.99) Optimal # of Features 14/50 21/233 Selected Features b L-glutamic acida L-valinea L-(+) lactic acida N-acetyl-5-hydroxytryptamine L-threonine Diglycerol Urea Arachidic acid Trans-aconitic acid L-proline N, N-dimethyl-1 4-phenylenediamine D-glucose L-serine L-cystine P01024a P01591 P02743a P02763 P05160a P06727 P13591 P13598a P22891a P35858 [43000] [53000]a [63423] [28000] [66012]a L-glutamic acida L-valinea L-(+) lactic acida L-threonine Urea L-cystine a Significant (p value ≤ 0.05) in univariate statistical analysis b N-glycans are characterized by GlcNAc, mannose, galactose, fucose, and NeuNAc, and proteins are indicated by Uniprot IDs The results of best performing methods are marked in bold. TABLE 6 Performance Comparison Based on the Optimal Number of Features Selected in the GU Cohort (44 Samples) GU Cohort Proteomics Glycomics Metabolomics Integrated (P + G) Integrated (P+G+M) Accuracy 0.89 0.91 0.84 0.98 0.98 Sensitivity 0.94 0.87 0.85 0.95 0.96 Specificity 0.85 0.95 0.83 0.99 0.99 AUC (95% CI) 0.87 (0.72, 0.96) 0.97 (0.89, 0.99) 0.91 (0.77, 0.97) 0.99 (0.95, 0.99) 0.99 (0.95, 0.99) Optimal Number of Features 3/101 10/82 4/50 15/183 10/233 Selected Features b O75636a P00736 P00751a [43100]a Ethanolamine L-(+) lactic acida Oxalic acid Putrescine O75636a P01023a P02774a P04278 P16070 P41222a P80108a [53111]c [53313] [34110] [43110]a [43200]a [43201] [73514] [53111]a,c O75636a P01876a P14151 P41222a [43100]a [53101]a [53111]a Malonic acid Putrescine Sorbosea [53313] [53411]a [53000]a [43201] [34110] [63402] [53100] [53111]a [53302] a Significant (p value ≤ 0.05) in univariate statistical analysis b N-glycans are characterized by GlcNAc, mannose, galactose, fucose, and NeuNAc, and proteins are indicated by Uniprot IDs c Isomers with different retention times. The results of best performing methods are marked in bold. TABLE VII Classification Performance on Independent Samples Accuracy Prote- omics Glyc- omics Metabol- omics Integrated (P + G) Integrated (P+G+M) TU cohort 0.70 0.59 0.78 0.82 0.85 GU cohort 0.71 0.78 0.64 0.86 0.86 The results of best performing methods are marked in bold. References 1 Fuster MM Esko JD The sweet and sour of cancer: Glycans as novel therapeutic targets Nat. Rev. Cancer 2005 5 7 526 542 16069816 2 Blomme B Van Steenkiste C Callewaert N Van Vlierberghe H Alteration of protein glycosylation in liver diseases J. Hepatol 2009 50 3 592 603 19157620 3 Kulasingam V Diamandis EP Strategies for discovering novel cancer biomarkers through utilization of emerging technologies Nat. Clin. Pract. Oncol 2008 5 10 588 599 18695711 4 Nezami Ranjbar MR Luo Y Di Poto C Varghese RS Ferrarini A Zhang C Sarhan NI Soliman H Tadesse MG Ziada DH Roy R Ressom HW GC-MS based plasma metabolomics for identification of candidate biomarkers for hepatocellular carcinoma in Egyptian cohort PLoS One 10 6 e0127299 5 Wang M Yu G Ressom HW Integrative analysis of LC-MS based glycomic and proteomic data Engineering in Medicine and Biology Society (EMBC), 37th Annual International Conference of the IEEE 2015 IEEE 8185 8188 6 Guyon I Weston J Barnhill S Vapnik V Gene selection for cancer classification using support vector machines Mach. Learning 2002 46 1–3 389 422 7 Tsai T Wang M Di Poto C Hu Y Zhou S Zhao Y Varghese RS Luo Y Tadesse MG Ziada DH Ressom HW LC–MS profiling of N-glycans derived from human serum samples for biomarker discovery in hepatocellular carcinoma Journal of Proteome Research 2014 13 11 4859 4868 25077556 8 Tsai T Song E Zhu R Di Poto C Wang M Luo Y Varghese RS Tadesse MG Ziada DH Desai CS Shetty K Mechref Y Ressom HW LC–MS/MS based Serum Proteomics for Identification of Candidate Biomarkers for Hepatocellular Carcinoma Proteomics 2015 15 13 2369 2381 25778709 9 MacLean B Tomazela DM Shulman N Chambers M Finney GL Frewen B Kern R Tabb DL Liebler DC MacCoss MJ Skyline: An open source document editor for creating and analyzing targeted proteomics experiments Bioinformatics 2010 26 7 966 968 20147306 10 Wang M Yu G Mechref Y Ressom HW GPA: An algorithm for LC/MS based glycan profile annotation IEEE International Conference on Bioinformatics and Biomedicine Workshop (BIBM 2013) 2013 12 Shanghai, China 11 Nezami Ranjbar MR Di Poto C Wang Y Ressom HW SIMAT: GC-SIM-MS data analysis tool BMC Bioinformatics 2015 16 259 25591662 12 Saeys Y Inza I Larrañaga P A review of feature selection techniques in bioinformatics Bioinformatics 2007 23 19 2507 2517 17720704
PMC005xxxxxx/PMC5124554.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 9712133 21042 AIDS Behav AIDS Behav AIDS and behavior 1090-7165 1573-3254 27233248 5124554 10.1007/s10461-016-1437-3 NIHMS792655 Article Online Sex-Seeking among Men Who Have Sex with Men in Nigeria: Implications for Online Intervention Stahlman Shauna 1 Nowak Rebecca G. 2 Liu Hongjie 3 Crowell Trevor A. 45 Ketende Sosthenes 1 Blattner William A. 2* Charurat Manhattan E. 2 Baral Stefan D. 1 on behalf of the TRUST/RV368 Study Group 1 Johns Hopkins Bloomberg School of Public Health, Center for Public Health and Human Rights, Department of Epidemiology, Baltimore, USA 2 Division of Epidemiology and Prevention, Institute of Human Virology, University of Maryland, Baltimore, USA 3 Department of Epidemiology and Biostatistics, School of Public Health, University of Maryland, College Park, USA 4 U.S. Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, USA 5 Henry M. Jackson Foundation for the Advancement of Military Medicine, Inc., Bethesda, USA Corresponding author: Shauna Stahlman, MPH, PhD, E7133, 615 N. Wolfe St. Baltimore, MD 21205, Phone: 410-502-8975, Fax: 410-614-8371, [email protected] * Full Professor Co-senior authors The TRUST Study Group is constituted as follows: Principal investigators: William Blattner and Man Charurat (IHV, University of Maryland, Baltimore, MD, USA) Co-investigators: Alash’le Abimiku, Sylvia Adebajo, Julie Ake, Stefan Baral, Trevor Crowell, Charlotte Gaydos, Babajide Keshinro, Jerome Kim, Hongie Liu, Jennifer Malia, Nelson Michael, Ogbonnaya Njoku, Rebecca Nowak, Helen Omuh, Ifeanyi Orazulike, Sheila Peel, Merlin Robb, Cristina Rodriguez-Hart, Sheree Schwartz Institutions: Institute of Human Virology at the University of Maryland School of Medicine (IHV-UMB), Johns Hopkins Bloomberg School of Public Health (JHSPH), Walter Reed Army Institute of Research, U.S. Military HIV Research Program (MHRP), Department of Defense, Walter Reed Program, Nigeria (WRP), Institute of Human Virology Nigeria (IHVN), International Centre for Advocacy for the Right to Health (ICARH), The Initiative for Equal Rights (TIER), Population Council (Pop Council) 9 6 2016 11 2017 01 11 2018 21 11 30683077 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. The TRUST/RV368 project was undertaken to apply innovative strategies to engage Nigerian MSM into HIV care. In this analysis we evaluate characteristics of online sex-seekers from the TRUST/RV368 cohort of 1,370 MSM in Abuja and Lagos. Logistic regression and generalized estimating equation models were used to assess associations with online sex-seeking. Online sex-seeking (n=843, 61.5%) was associated with participation in MSM community activities, larger social and sexual networks, and higher levels of sexual behavior stigma. In addition, online sex-seeking was associated with testing positive for HIV at a follow-up visit (Adjusted Odds Ratio [aOR]=2.02, 95% Confidence Interval [CI]=1.37, 2.98) among those who were unaware of or not living with HIV at baseline. Across visits, online sex-seekers were marginally more likely to test positive for chlamydia/gonorrhea (aOR=1.28, 95% CI=0.99, 1.64). Online sex-seekers in Nigeria are at increased risk for HIV/STIs but may not be benefiting from Internet-based risk reduction opportunities. HIV STIs MSM Internet sex partners Africa Introduction Men who have sex with men (MSM) are among the populations at highest risk for HIV infection (1). In Nigeria, the prevalence of HIV among MSM is 11–48%, whereas the overall adult prevalence is 3% (2–4). MSM face high levels of stigma as a result of their sexual practices, including verbal, physical, and psychological abuse (5). High levels of sexual behavior stigma may act as a barrier to HIV testing and care for MSM, and this is particularly problematic in countries with concentrated key population epidemics where anti-gay legislation erects further barriers (6–10). In some parts of Nigeria, homosexual activity between men is punishable by death (11). As a result, there is a strong need for HIV prevention interventions to reach these highly stigmatized and high-risk men. In many countries, the prevalence of online sex-seeking is particularly high among MSM, with 34–50% of MSM who were sampled offline across North America, China, and Europe reporting use of the Internet to find sex partners (12–14). Data from Sub-Saharan Africa are more limited, but a previous study identified a prevalence of 39% for online sex-seeking in Lesotho and 44% in Swaziland (15). This high prevalence has partly been due to the increasing popularity of the Internet and smart phone technology, which has led to a widespread phenomenon of online sex-seeking (16, 17). Furthermore, in countries with high levels of sexual behavior stigma affecting MSM, highly stigmatized MSM may be more likely to seek sex online. A qualitative study conducted among MSM in Chengdu, China, noted that anonymity and safety were major contributing factors to using the Internet to find male sex partners (18). Another study of MSM in New York City indicated that MSM who wished to conceal their sexual orientation were more likely to seek male partners online than were MSM who were open about their orientation (19). In the Sub-Saharan African context there is limited information on the frequency and drivers of online sex-seeking among MSM. Recently, Internet-based interventions have been developed in the US, Europe, and other settings to engage online sex-seekers in preventative HIV interventions (20–22). Previous successes have included the ability for online interventions to increase HIV/STI knowledge, risk reduction, and HIV/STI testing, although the evaluation of most online interventions has depended on self-reported behavioral change (23–25). Overall, the availability of smartphone technology in Sub-Saharan Africa coupled with high levels of sexual behavior stigma (15, 17) presents opportunities to develop and optimize Internet-based technologies for HIV prevention programs. This study describes the patterns and drivers of online sex-seeking among a prospective cohort of MSM presenting for HIV testing and treatment in Abuja and Lagos, Nigeria. Our goals were to: 1) Improve the understanding of the characteristics of Nigerian MSM who use the Internet to find sex partners (i.e., socio-demographics, social networks, sexual networks, and sexual behavior stigma), and 2) Test the association of online sex-seeking with HIV/STI diagnoses and HIV treatment-related variables over time. Characteristics were chosen for analysis based on associations with HIV risk and online sex-seeking that were identified in recent studies (15, 26–29), and because these characteristics would inform the development or adaptation of online interventions to the Nigerian setting. Methods Study Population and Design Data were collected as part of the TRUST/RV368 study, which was implemented as a collaboration between the Institute of Human Virology (IHV) at the University of Maryland, Johns Hopkins University, the International Center for Advocacy on the Right to Health (a Nigerian CBO), IHV-Nigeria (PEPFAR implementing partner and Nigerian research center of excellence), and the U.S. Military HIV Research Program. Participants were recruited at two study sites in Abuja and Lagos from March 2013 to August 2015 using respondent-driven sampling (RDS) (4, 29). Briefly, RDS is a chain-referral process whereby initial recruits or “seeds” are identified in the target population during an initial recruitment (wave 0). These seeds recruit their peers, who then recruit additional peers, and so on until the desired sample size is reached, resulting in multiple waves of recruitment. RDS has been shown to be effective in populations of MSM who are less engaged in the MSM community as well as in HIV prevention activities (29, 30). MSM were eligible to participate if they presented to the study site with a valid RDS coupon, self-reported being assigned male sex at birth, were able to provide informed consent in English or Hausa, and reported a history of insertive or receptive anal intercourse in the previous 12 months. In addition, participants had to be aged 16 or older with men under the age of 18 considered emancipated minors who were exempt from parental consent for the purpose of this study. The men had to be willing to enroll and participate in follow up for 18 months, including completion of quarterly structured interviewer-administered questionnaires and HIV and STI testing and treatment monitoring. Ultimately, ten seeds recruited 1,371 baseline participants that resulted in up to 27 recruitment waves. Equilibrium was reached for several socio-demographic characteristics including age, sexual orientation, and education. Equilibrium was defined as the point at which the cumulative sample proportions came within 2% of the final sample proportions, and did not fluctuate more than 2% during the sampling of additional waves (31). For the current analysis, one participant was excluded due to missing data on the key variable of interest (report of using the Internet to find male sex partners). This study was approved by the Federal Capital Territory Health Research Ethics Committee, the University of Maryland Baltimore Institutional Review Board (IRB), and the Walter Reed Army Institute of Research IRB. Informed consent was obtained from all individual participants included in this study. Data Collection and Key Measures Participants were administered a structured survey instrument across seven different study visits. The survey instrument was validated in previous studies of MSM throughout Sub-Saharan Africa (30, 32–34) and was pre-tested in Lagos and Abuja before enrollment and initiation of the study. Survey measures of interest included socio-demographics, MSM social and sexual networks, experiences and perceptions of sexual behavior stigma in social and healthcare settings, and HIV treatment uptake. Online Sex-Seeking At visit 0, 2, 4, and 6, participants were asked, “Where (in what type of place) do you meet new male sexual partners?” and were asked to choose from a list of possible places. The possible response options were: private home, bar or club, private party, brothel, street or park, private vehicle, hotel or guest house, news advertisements or cards, online, school or work, or mosque or church. Participants could choose more than one location. At visit 1, 3, and 5 participants were asked how often they used the Internet to look for male sexual partners since their last visit. Those who indicated “yes” to meeting partners online or those who reported any recent Internet use for the purpose of looking for male sex partners were categorized as currently meeting male sex partners online. MSM Social and Sexual Networks Participants were asked to report the number of MSM that they knew and whether they participated in activities in their MSM community. They were also asked to report their number of male anal sex partners within the last 12 months and to give more detailed information about condom use, HIV status disclosure, and other information about their five most recent male sex partners. Using this information, we created a variable for “high risk sex” among participants not aware of living with HIV (i.e., self-reported HIV-negative or unknown status), which indicated whether a participant engaged in condomless anal sex with any of his five most recent male partners whose HIV status he did not know or who he knew to be HIV-positive. For participants who were aware of living with HIV (i.e., self-reported HIV positive), we created a variable for “serosorting”, which indicated whether the participant had engaged in condomless anal sex with any of his five most recent male partners whose HIV status he knew to be HIV-positive. Sexual Behavior Stigma Participants were asked whether they perceived or experienced stigma in personal, social, or healthcare settings because they have sex with men. This sexual behavior stigma was measured by a series of “yes” or “no” questions at visit 0, 2, 4, and 6. Sexual behavior stigma in personal settings was measured by asking participants whether they ever felt excluded by family members, felt like family members gossiped, or felt rejected by friends. Sexual behavior stigma in social settings was measured by asking whether participants knew of safe places in their community to socialize with other MSM, whether they did not feel protected by police, felt scared to walk around in public places, or were ever verbally harassed, blackmailed, or physically hurt because they have sex with men. Sexual behavior stigma in healthcare settings was measured by asking participants if they ever avoided or felt afraid to go to healthcare services because they were worried that someone may learn that they have sex with men, if they ever felt that they were not treated well in a health center because someone knew they had sex with men, or if they ever heard a healthcare worker gossiping about them because they have sex with men. These measures have been used in previous studies conducted in Sub-Saharan Africa and have been found to be both prevalent and associated with increased online sex-seeking (15, 34–36). HIV, Sexually-Transmitted Infections, and Treatment Uptake During visits 1–6, participants were screened regardless of symptoms for rectal and urogenital Neisseria gonorrhea (NG) and Chlamydia trachomatis (CT) using the Aptima Combo 2 Assay (Hologic, Bedford, MA). Urine and rectal swab samples were transported weekly at 2–8°C to the Defense Reference Laboratory in Abuja. STI test results were generated within 3 weeks of receipt of the samples and participants were called back to the clinic for an off study visit to receive treatment if the laboratory diagnosis was positive. Those who tested positive were treated with appropriate antibiotics and retested at follow-up visits. In addition, participants were tested for HIV using Abbott Determine HIV-1/2 test kits. The parallel testing algorithm was followed for high-risk individuals at baseline and among HIV-negative participants at each follow-up visit (37). Whole blood samples were collected for viral load testing at visits 1 – 6. HIV viral load was quantified using real-time polymerase chain reaction on the COBAS TaqMan platform (Roche Molecular Diagnostics, Pleasanton, CA). HIV suppression was defined as plasma HIV RNA <200 copies/mL. Current uptake of ARVs was defined by self-report of currently being treated for HIV or by having pharmacy records for ARVs at baseline or at subsequent visit. Those who tested positive for HIV received the standard of care clinical assessment. All HIV-positive participants were offered ART irrespective of their clinical status and CD4 count as part of a treatment as prevention strategy. Statistical Analysis Bivariate analysis was performed using logistic regression to assess associations between variables of interest with the outcome of meeting a male sex partner online at baseline. Pearson chi-square tests were used to compare prevalence of other venues for which online sex-seekers and non-online sex-seekers reported meeting male sex partners. Multivariate logistic regression models were used to assess the independent associations of MSM social and sexual networks as well as stigma with meeting a male sex partner online at baseline. In addition, multivariate models with generalized estimating equations (GEEs) and autoregressive correlation matrices were used to assess the association between meeting a male sex partner online and measurements of interest across study visits. GEEs were used to account for within-participant correlations and paired data of participants who had multiple study visits. Analyses were performed using SAS software version 9.4 (SAS Institute Inc., Cary, NC, USA). Results Venues for Meeting Male Sex Partners The prevalence of meeting male sex partners online at baseline was 61.5% (n=843), with a much higher prevalence in Lagos (86.3%; n=372) as compared with Abuja (50.2%; n=471). Overall 53% (n=725) reported using the Internet almost every day (Table 1). The most popular websites/web-applications reported by participants for meeting new male sex partners included 2GO (57%), Facebook (52%), and WhatsApp (31%) (Figure 1). Grindr, a popular sex-seeking web-application used by MSM in the US, was reported to be used by 4% of participants in this cohort. Online sex-seekers were more likely than those who did not meet male sex partners online to also meet male sex partners in a bar or club (54% vs. 37%, p<0.001), in a brothel (15% vs. 10%, p=0.03), in a private vehicle (30% vs. 22%, p<0.001), in a hotel or guesthouse (50% vs. 43%, p=0.01), through news advertisements or cards (10% vs. 3%, p<0.001), at school or work (62% vs. 53%, p<0.001), and at church or mosque (26% vs. 13%, p<0.001). However, they were less likely to meet male partners on the street or in a park (45% vs. 56%, p<0.001) and similarly likely to meet in a private home (62% vs. 65%, p=0.23) or party (57% vs. 52%, p=0.07). Socio-demographics Study participants who reported being Muslim (compared to Christian) (Odds Ratio [OR]=0.18, p<0.001), and ever being married or cohabiting with a male or female partner were less likely to report finding male sex partners online (OR=0.44, p<0.001)(Table 1). In contrast, those who were aged 20–26 years (compared to aged 16–19 years, OR=1.52, p=0.005), reported a secondary school (OR=21.26, p<0.001) or higher education (OR=48.7, p<0.001), owned a mobile phone (OR=6.28, p<0.001), and used the Internet almost every day (compared to sometimes, OR=2.14, p<0.001) were more likely to report using the Internet to find sexual partners. MSM Social/Sexual Networks In the bivariate analysis, those who reported a larger MSM network (OR=1.02, p=0.001), reported participation in MSM community activities (OR=1.81, p<0.001), reported a larger number of male anal sex partners (OR=1.05, p<0.05), and participated in an earlier RDS accrual wave (a marker of community “connectedness”(29, 30)) (OR=0.93, p<0.001) were more likely to seek-sex online (Table 2). After adjusting for age, education level, marital status, religion, and study site, the variables that remained significantly associated with online sex-seeking were participation in an earlier RDS accrual wave (Adjusted Odds Ratio [AOR]=0.97, p=0.03), participation in MSM community activities (AOR=1.45, p=0.02), and engagement in high risk sex (AOR=1.52, p=0.03). Sexual Behavior Stigma In the bivariate analysis, feeling excluded by family members (OR=1.73, p=0.005), feeling gossiped about by family members (OR=1.60, p=0.001), and feeling rejected by friends (OR=1.62, p=0.001) were associated with increased likelihood of online sex-seeking (Table 3). Online sex-seeking was also associated with believing that there was no safe place to socialize with other MSM (OR=2.07, p<0.001), not feeling protected by police (OR=2.42, p<0.001), feeling scared to walk around in public (OR=1.41, p=0.02), being verbally harassed (OR=1.71, p<0.001), being blackmailed (OR=2.39, p<0.001), and being physically hurt (OR=2.51, p<0.001). In addition, ever being afraid to seek healthcare services (OR=3.44, p<0.001), avoiding healthcare services (OR=4.35, p<0.001), being treated poorly by a healthcare worker (OR=4.80, p<0.001), and hearing a healthcare worker gossip about the participant’s sexual behavior (OR=3.12, p<0.001) were also associated with online sex-seeking. In the adjusted models, most of these associations remained significant except for feeling excluded by family members (AOR=1.56, p=0.06), feeling gossiped about by family members (AOR=1.41, p=0.06), being scared to walk around in public (AOR=1.26, p=0.19), and hearing healthcare workers gossip (AOR=1.47, p=0.29). HIV, Sexually-Transmitted Infections, and Treatment Uptake In the multivariate models, online sex-seeking was associated with testing positive for HIV among those presenting at the baseline study visit who were not already aware of living with HIV (AOR=2.02, p<0.001)(Table 4). Across study visits, online sex-seekers were also marginally more likely to test positive for chlamydia or gonorrhea (AOR=1.28, p=0.06). Among those living with HIV, no difference was observed in rates of ARV use (AOR=1.17, p=0.41) or viral suppression (AOR=1.36, p=0.08) between online sex-seekers and participants who did not seek sex online. Discussion The prevalence of 61.5% for online sex-seeking among TRUST/RV328 participants in Nigeria is even higher than the prevalence reported in previous years among MSM in other parts of the world such as Lesotho, Swaziland, North America, China, and Europe (12–15). This high prevalence suggests that online preventative interventions need to be developed, implemented, and evaluated in the African setting. We also found a link between sexual behavior stigma in social settings with online sex-seeking, which points to the impact of widespread societal prejudice, and is reflected in recent Nigerian legislation banning gay assembly and support for gay organizations. Our prior publication in this same population documented the negative impact of this law on service uptake and healthcare seeking (10). In addition, the current data support the concept that online sex-seeking provides a vehicle for non-public “hook ups” that protects MSM from potential prosecution. Notably, the participants who engaged in online sex-seeking were more likely to report sexual behavior stigma in social and healthcare settings, were younger, more educated, more engaged in the MSM community, and more likely to be at risk for HIV and STIs. This information may be useful for appropriately adapting online HIV prevention interventions to the Nigerian setting. In the current analysis we observed that participants who enrolled in later waves of RDS recruitment were less likely to have used the Internet for sex-seeking. We know from a separate analysis using these data that clients enrolled in later waves of RDS were less likely to have tested for HIV (29). An intervention to expand access to services such as HIV testing to this highly marginalized subset of the community with high rates of undetected HIV infection is a particular priority. Moreover, among self-reported HIV-negative participants, online sex-seeking was associated with high risk sex, which suggests that online interventions targeting behavioral change would have the potential to be highly impactful in this population. One potential intervention method could be to promote HIV status disclosure and negotiation of safer sex practices between partners before meeting (38). However, the feasibility of this approach would require pilot testing in a future study. In addition, the high overall prevalence of STIs including incident rectal STI infection observed in this cohort underscores the importance of presumptive STI testing and treatment (39), regardless of whether the interventions are implemented online or offline. Finally, we found that MSM who did not participate in the MSM community were less likely to meet sex partners online. MSM in this study reported meeting partners from a variety of other places aside from the Internet; although online sex-seekers tended to report a higher prevalence of meeting partners in several other venues suggesting an overall larger and more diverse sexual network. Alternatively, it is possible that participants discussed and selected physical meeting venues with partners who they first met online, which might also explain this finding if participants decided to report both the physical and online venues in their responses to these questions. Overall, these findings provide an important target for engagement in future online interventions, an approach that will require further understanding of the barriers to Internet access and intervention uptake. The finding of a lack of significant association between online sex-seeking and receipt of ARVs or viral suppression points to opportunities for intervention to support treatment as prevention goals, by using technology to improve adherence to medication and support retention in care (40, 41). Our finding that mobile phone ownership was associated with meeting sex partners online provides opportunity to engage mobile-app technology to achieve these goals. However, additional research should investigate the potential use of mobile-app social media for HIV-related interventions in West Africa given potential challenges including infrastructure, policy, and competing research and social networking business goals (40). There are potential limitations of this study. First, the associations reported here are exploratory and do not represent causality. Many associations were assessed using cross-sectional data at baseline and although some associations were assessed prospectively, in the absence of event-level data we cannot infer that participants’ sexual network data pertain to the partners that they met online. Recent research suggests that factors such as whether the person actually meets their Internet-found partner in person is differentially and more strongly associated with HIV risk, and we did not collect these data (42). Further, in RDS there can be bias introduced by the non-random selection of individuals out of a recruiter’s social network (43). However, because equilibrium was reached for several socio-demographic characteristics, this suggests a minimal overall bias due to non-random recruitment. Finally, the sub-cohort in Abuja versus Lagos suggests different patterns of socio-demographics and behavior that could have impacted our results. Lagos is the largest city in Nigeria and is a fast-growing urban center. Participants in Lagos were younger, more educated, more likely to be Christian compared to Muslim, and more likely to identify as gay compared to bisexual. Some of these socio-demographic characteristics may have contributed to the observation that MSM in Lagos were almost twice as likely to report online sex-seeking than participants in Abuja. However, in a sensitivity analysis, we found that the majority of relationships with online sex-seeking were consistent across both sites. The findings of the current analysis reinforce the universality of challenges in achieving WHO goals for MSM worldwide, but also offer the opportunity to engage best practices in providing services to key populations for high impact. In particular, online interventions that have been successfully developed and tested in North America and Europe may be transferrable to the African context given the similar HIV risk profile and high prevalence of online sex-seeking. Internet-based strategies in Nigeria could enrich the current findings with further qualitative and formative research studies to better understand societal drivers of online sex-seeking. Rigorously designed intervention trials that quantify impact of Internet-based strategies on measurable prevention and treatment uptake and outcomes will advance the HIV prevention agenda for high impact in the African setting. The study team would like to acknowledge the participants for taking part in this study given the significant stigma that exists affecting gay men and other men who have sex with men in Nigeria. We would also like to acknowledge Sara Kennedy for her leadership support in implementing the study. Marcy Gelman and Dr. Kevin Kapila from Fenway Health and Dr. Syliva Adebajo from the Population Council Nigeria completed training to increase the cultural and clinical competency of study and clinical staff for the TRUST Study. In addition, Ashley Grosso supported instrument development, and Erin Papworth provided training on respondent-driven method implementation. Source of Funding: This work was supported by a cooperative agreement (W81XWH-11-2-0174) between the Henry M. Jackson Foundation for the Advancement of Military Medicine, Inc., and the U.S. Department of Defense (DOD). This study is also supported by funds from the US National Institutes of Health under Award No. R01MH099001-01, the US Military HIV Research Program (Grant No. W81XWH-07-2-0067), Fogarty AITRP (D43TW01041), and the President’s Emergency Plan for AIDS Relief through cooperative agreement U2G IPS000651 from the HHS/Centers for Disease Control and Prevention (CDC), Global AIDS Program with IHVN. In addition, this work was supported by The American Foundation for AIDS Research (amfAR) and the Johns Hopkins University Center for AIDS Research (P30AI094189). Figure 1 Websites/Web-applications used to meet male sex partners, among 388 participants at visit 1 who reported using the Internet to meet male sex partners. Participants could indicate more than one response. Table 1 Prevalence of socio-demographic characteristics and bivariate associations with online sex-seeking among MSM in Nigeria at baseline (N=1,370) Characteristic n % OR 95% CI P-value Study Site Abuja 939 68.5 1.00 – – Lagos 431 31.5 6.26 4.63, 8.48 <0.001* Age, in years 16–19 244 17.8 1.00 – – 20–26 756 55.2 1.52 1.14, 2.04 0.005 27+ 370 27.0 1.04 0.75, 1.44 0.80 Sexual orientation Bisexual 860 62.9 1.00 – – Gay 507 37.1 0.89 0.71, 1.11 0.31 Education completed Primary school or less 127 9.3 1.00 – – Secondary school 839 61.6 21.26 10.65, 42.47 <0.001* More than secondary 396 29.1 48.70 23.72, 99.97 <0.001* Marital status Single/never married or cohabited 1198 87.6 1.00 – – Ever married/cohabited 170 12.4 0.44 0.32, 0.61 <0.001* Religion Christian 939 68.8 1.00 – – Muslim 425 31.2 0.18 0.14, 0.23 <0.001* Owns a mobile phone No 117 8.6 1.00 – – Yes 1244 91.4 6.28 4.03, 9.81 <0.001* Internet use Never 374 27.6 – – – Sometimes 258 19.0 1.00 – – Almost every day 725 53.4 2.14 1.49, 3.08 <0.001*** * p<0.05; p<0.01; * p<0.001 Note: Values may not sum to 100% due to rounding CI=Confidence Interval OR=Odds Ratio Table 2 Bivariate and multivariate associations of MSM social and sexual networks with online sex-seeking (N=1,370) Explanatory variable Median (IQR) N (%) OR 95% CI p-value AORa 95% CI p-value MSM social network MSM network size 10 (7 – 50) – 1.02b 1.01, 1.03 0.001 1.01b 1.00, 1.02 0.10 RDS accrual wave number 10 (6 – 15) – 0.93 0.92, 0.95 <0.001* 0.97 0.95, 1.00 0.03* Participation in MSM community activities – 364 (44.0) 1.81 1.36, 2.41 <0.001*** 1.45 1.05, 2.01 0.02* MSM sexual network Number of male anal sex partners, past 12 mo. 5 (3 – 10) – 1.05c 1.00, 1.11 0.0458* 1.02c 0.97, 1.07 0.55 High risk sex, among participants not aware of living with HIV – 233 (40.2) 1.16 0.83, 1.62 0.37 1.52 1.04, 2.22 0.03* Serosorting, among participants aware of living with HIV – 48 (19.8) 0.83 0.42, 1.64 0.59 0.66 0.29, 1.47 0.31 * p<0.05; p<0.01; * p<0.001 a Models adjust for age, education level, marital status, religion, and study site b Per 15 MSM c Per 5 male sex partners Note: Separate regression models were used for each explanatory variable AOR=Adjusted Odds Ratio CI=Confidence Interval IQR=Interquartile Range OR=Odds Ratio Table 3 Bivariate and multivariate associations between sexual behavior stigma and online sex-seeking (N=1,370) Stigma Meets Male Sex Partners Online (n/N) OR 95% CI p-value AORa 95% CI p-value Personal Family exclusion 105/145 1.73 1.18, 2.53 0.005 1.56 0.98, 2.48 0.06 Family gossiped 194/277 1.60 1.21, 2.13 0.001 1.41 0.99, 2.00 0.06 Friend rejection 187/266 1.62 1.21, 2.16 0.001** 1.47 1.03, 2.10 0.03* Social No safe place to socialize with other MSM 334/462 2.07 1.62, 2.64 <0.001*** 1.44 1.08, 1.93 0.01* Did not feel protected by police 182/236 2.42 1.75, 3.36 <0.001*** 1.57 1.07, 2.32 0.03* Felt scared to walk around in public 174/256 1.41 1.06, 1.88 0.02* 1.26 0.89, 1.79 0.19 Verbally harassed 311/445 1.71 1.35, 2.18 <0.001* 1.53 1.13, 2.05 0.005 Blackmailed 245/322 2.39 1.80, 3.18 <0.001* 2.19 1.55, 3.10 <0.001* Physically hurt 209/271 2.51 1.85, 3.42 <0.001*** 1.62 1.12, 2.35 0.01* Healthcare Afraid to seek services 321/401 3.44 2.61, 4.53 <0.001* 2.05 1.48, 2.82 <0.001* Avoided services 252/299 4.35 3.12, 6.08 <0.001* 2.15 1.46, 3.17 <0.001* Treated poorly 58/66 4.80 2.27, 10.12 <0.001*** 2.30 1.00, 5.27 0.0498* Healthcare worker gossiped 57/69 3.12 1.66, 5.86 <0.001*** 1.47 0.71, 3.04 0.29 * p<0.05; p<0.01; * p<0.001 a All stigma items entered into separate models. Models adjust for age, self-reported HIV status, sexual orientation, education level, marital status, religion, and study site AOR=Adjusted Odds Ratio CI=Confidence Interval OR=Odds Ratio Table 4 Multivariate adjusted associations between online sex-seeking and HIV/STI outcome variables among MSM in Nigeria study across visits (N=1,370) Outcome n/Na AOR 95% CI p-value HIV/STI Diagnoses Tested positive for HIV at any study visitb 256/734 2.02 1.37, 2.98 <0.001* Positive for CT/NGc 251/777 1.28 0.99, 1.64 0.06 Among those aware of living with HIV d Currently taking ARV 137/243 1.17 0.80, 1.72 0.41 Viral load suppressed 69/204 1.36 0.96, 1.91 0.08 * p<0.001 a Proportion of participants who indicated “yes” to the outcome variable among participants at baseline b Among those unaware of living with HIV at baseline. Model adjusts for age, sexual orientation, education level, marital status, religion, and study site. c Model adjusts for age, visit number, self-reported HIV status, sexual orientation, education level, marital status, religion, and study site. d Models adjust for age, visit number, sexual orientation, education level, marital status, religion, and study site. Note: Separate regression models were used for each outcome variable AOR=Adjusted Odds Ratio CI=Confidence Interval CT=Chlamydia trachomatis NG=Neisseria gonorrhea Disclaimer: The views expressed are those of the authors and should not be construed to represent the positions of the U.S. Army, the Department of Defense, or other funders. Conflicts of Interest: The authors do not have any conflicts of interest to declare. Ethical approval: All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. 1 Beyrer C Baral SD van Griensven F Global epidemiology of HIV infection in men who have sex with men Lancet 2012 380 9839 367 77 Epub 2012/07/24 22819660 2 UNICEF Country Statistics. 2013 9 1 2015 http://www.unicef.org/statistics/index_countrystats.html 3 Vu L Adebajo S Tun W High HIV prevalence among men who have sex with men in Nigeria: implications for combination prevention Journal of acquired immune deficiency syndromes 2013 63 2 221 7 Epub 2013/02/15 23406978 4 Charurat ME Emmanuel B Akolo C Uptake of Treatment as Prevention for HIV and Continuum of Care Among HIV-Positive Men Who Have Sex With Men in Nigeria Journal of acquired immune deficiency syndromes 2015 68 Suppl 2 S114 23 Epub 2015/02/28 25723975 5 Sekoni AO Ayoola OO Somefun EO Experiences of social oppression among men who have sex with men in a cosmopolitan city in Nigeria HIV/AIDS Auckland, NZ 2015 7 21 7 Epub 2015/01/08 25565899 6 Djomand G Quaye S Sullivan PS HIV epidemic among key populations in west Africa Current opinion in HIV and AIDS 2014 9 5 506 13 Epub 2014/07/11 25010898 7 Ekouevi DK Dagnra CY Goilibe KB [HIV seroprevalence and associated factors among men who have sex with men in Togo] Revue d’epidemiologie et de sante publique 2014 62 2 127 34 Epub 2014/03/13. Seroprevalence du VIH et facteurs associes chez les hommes ayant des rapports sexuels avec d’autres hommes au Togo 8 Knox J Sandfort T Yi H Reddy V Maimane S Social vulnerability and HIV testing among South African men who have sex with men International journal of STD & AIDS 2011 22 12 709 13 Epub 2011/12/17 22174050 9 Hladik W Barker J Ssenkusu JM HIV infection among men who have sex with men in Kampala, Uganda–a respondent driven sampling survey PloS one 2012 7 5 e38143 Epub 2012/06/14 22693590 10 Schwartz SR Nowak RG Orazulike I The immediate effect of the Same-Sex Marriage Prohibition Act on stigma, discrimination, and engagement on HIV prevention and treatment services in men who have sex with men in Nigeria: analysis of prospective data from the TRUST cohort The Lancet HIV 2015 2 7 e299 e306 Epub June 2, 2015 26125047 11 Itaborahy LP Zhu J State-sponsored homophobia. A world survey of laws: Criminalisation, protection and recognition of same-sex love International Lesbian Gay Bisexual Trans and Intersex Association, 2014 [ 14 5 2016 http://old.ilga.org/Statehomophobia/ILGA_SSHR_2014_Eng.pdf 12 Wu Z Xu J Liu E HIV and syphilis prevalence among men who have sex with men: a cross-sectional survey of 61 cities in China Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 2013 57 2 298 309 Epub 2013/04/13 23580732 13 Garofalo R Herrick A Mustanski BS Donenberg GR Tip of the Iceberg: young men who have sex with men, the Internet, and HIV risk American journal of public health 2007 97 6 1113 7 Epub 2007/04/28 17463378 14 Liau A Millett G Marks G Meta-analytic examination of online sex-seeking and sexual risk behavior among men who have sex with men Sexually transmitted diseases 2006 33 9 576 84 Epub 2006/03/17 16540884 15 Stahlman S Grosso A Ketende S Characteristics of men who have sex with men in southern Africa who seek sex online: a cross-sectional study Journal of medical Internet research 2015 17 5 e129 Epub 2015/05/27 26006788 16 Brown MJ Pugsley R Cohen SA Meeting Sex Partners Through the Internet, Risky Sexual Behavior, and HIV Testing Among Sexually Transmitted Infections Clinic Patients Archives of sexual behavior 2015 2 44 2 509 19 Epub 2015/01/09 25567074 17 Pew Research Center Emerging Nations Embrace Internet, Mobile Technology. 2014 14 5 2016 http://www.pewglobal.org/2014/02/13/emerging-nations-embrace-internet-mobile-technology/ 18 Feng Y Wu Z Detels R Evolution of men who have sex with men community and experienced stigma among men who have sex with men in Chengdu, China Journal of acquired immune deficiency syndromes 2010 53 Suppl 1 S98 103 Epub 2010/02/06 20104118 19 Schrimshaw EW Downing MJ Jr Siegel K Sexual venue selection and strategies for concealment of same-sex behavior among non-disclosing men who have sex with men and women Journal of homosexuality 2013 60 1 120 45 Epub 2012/12/18 23241205 20 Muessig KE Nekkanti M Bauermeister J Bull S Hightow-Weidman LB A Systematic Review of Recent Smartphone, Internet and Web 2.0 Interventions to Address the HIV Continuum of Care Current HIV/AIDS reports 2015 3 12 1 173 90 Epub 2015/01/30 25626718 21 Hirshfield S Chiasson MA Joseph H An online randomized controlled trial evaluating HIV prevention digital media interventions for men who have sex with men PloS one 2012 7 10 e46252 Epub 2012/10/17 23071551 22 Schnall R Travers J Rojas M Carballo-Diéguez A eHealth Interventions for HIV Prevention in High-Risk Men Who Have Sex With Men: A Systematic Review Journal of medical Internet research 2014 16 5 e134 24862459 23 Bauermeister JA Pingel ES Jadwin-Cakmak L Acceptability and preliminary efficacy of a tailored online HIV/STI testing intervention for young men who have sex with men: the Get Connected! program AIDS and behavior 2015 19 10 1860 74 Epub 2015/02/02 25638038 24 Lelutiu-Weinberger C Pachankis JE Gamarel KE Surace A Golub SA Parsons JT Feasibility, Acceptability, and Preliminary Efficacy of a Live-Chat Social Media Intervention to Reduce HIV Risk Among Young Men Who Have Sex With Men AIDS and behavior 2015 19 7 1214 27 Epub 2014/09/27 25256808 25 Ko NY Hsieh CH Wang MC Effects of Internet popular opinion leaders (iPOL) among Internet-using men who have sex with men Journal of medical Internet research 2013 15 2 e40 Epub 2013/02/27 23439583 26 Bien CH Best JM Muessig KE Wei C Han L Tucker JD Gay Apps for Seeking Sex Partners in China: Implications for MSM Sexual Health AIDS and behavior 2015 19 6 941 6 Epub 2015/01/13 25572834 27 Burnham KE Cruess DG Kalichman MO Grebler T Cherry C Kalichman SC Trauma symptoms, internalized stigma, social support, and sexual risk behavior among HIV-positive gay and bisexual MSM who have sought sex partners online AIDS care 2016 28 3 347 53 Epub 2015/10/16 26461452 28 Ko NY Tseng PC Huang YC Chen YC Hsu ST Seeking sex partners through the internet and mobile phone applications among men who have sex with men in Taiwan AIDS care 2016 1 5 Epub 2016/01/13 29 Baral SD Ketende S Schwartz S Evaluating Respondent-Driven Sampling as an Implementation Tool for Universal Coverage of Antiretroviral Studies Among Men Who Have Sex With Men Living With HIV Journal of acquired immune deficiency syndromes 2015 68 2 107 13 30 Stahlman S Johnston LG Yah C Respondent-driven sampling as a recruitment method for men who have sex with men in southern sub-Saharan Africa: a cross-sectional analysis by wave Sexually transmitted infections 2015 Epub 2015/10/02 31 Heckathorn DD Respondent-Driven Sampling: A New Approach to the Study of Hidden Populations Social Problems 1997 44 2 174 99 32 Wirtz AL Jumbe V Trapence G HIV among men who have sex with men in Malawi: elucidating HIV prevalence and correlates of infection to inform HIV prevention Journal of the International AIDS Society 2013 16 Suppl 3 18742 Epub 2014/01/01 24321110 33 Park JN Papworth E Billong SC Correlates of prior HIV testing among men who have sex with men in Cameroon: a cross-sectional analysis BMC public health 2014 14 1220 25424530 34 Fay H Baral SD Trapence G Stigma, health care access, and HIV knowledge among men who have sex with men in Malawi, Namibia, and Botswana AIDS and behavior 2011 15 6 1088 97 Epub 2010/12/15 21153432 35 Baral S Adams D Lebona J A cross-sectional assessment of population demographics, HIV risks and human rights contexts among men who have sex with men in Lesotho Journal of the International AIDS Society 2011 14 36 Epub 2011/07/06 21726457 36 Baral S Trapence G Motimedi F HIV prevalence, risks for HIV infection, and human rights among men who have sex with men (MSM) in Malawi, Namibia, and Botswana PloS one 2009 4 3 e4997 Epub 2009/03/28 19325707 37 Federal Ministry of Health Laboratory-based HIV Rapid Test Validation in Nigeria Abuja, Nigeria 2007 38 Grov C Hirshfield S Remien RH Humberstone M Chiasson MA Exploring the venue’s role in risky sexual behavior among gay and bisexual men: an event-level analysis from a national online survey in the U.S Archives of sexual behavior 2013 42 2 291 302 Epub 2011/10/21 22012413 39 Nowak RG Crowell TA Baral SD Incidence of STIs among MSM Engaged in Treatment as Prevention in Nigeria CROI Boston, MA 2016 40 Betjeman TJ Soghoian SE Foran MP mHealth in Sub-Saharan Africa International Journal of Telemedicine and Applications 2013 2013 482324 24369460 41 Okuboyejo S Eyesan O mHealth: Using Mobile Technology to Support Healthcare Online journal of public health informatics 2014 5 3 233 Epub 2014/03/29 24678384 42 Abara W Annang L Spencer SM Fairchild AJ Billings D Understanding internet sex-seeking behaviour and sexual risk among young men who have sex with men: evidences from a cross-sectional study Sexually transmitted infections 2014 90 8 596 601 Epub 2014/07/27 25063349 43 Wejnert C AN EMPIRICAL TEST OF RESPONDENT-DRIVEN SAMPLING: POINT ESTIMATES, VARIANCE, DEGREE MEASURES, AND OUT-OF-EQUILIBRIUM DATA Sociological Methodology 2009 39 1 73 116 20161130
PMC005xxxxxx/PMC5124559.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101128775 30137 J Proteome Res J. Proteome Res. Journal of proteome research 1535-3893 1535-3907 27058005 5124559 10.1021/acs.jproteome.6b00010 NIHMS829761 Article Profiling of Cross-Functional Peptidases Regulated Circulating Peptides in BRCA1 Mutant Breast Cancer Fan Jia † Tea Muy-Kheng M. ‡ Yang Chuan † Ma Li § Meng Qing H. ⊥ Hu Tony Y. †¶ Singer Christian F. ‡ Ferrari Mauro †∥ † Department of Nanomedicine, Houston Methodist Research Institute, Houston, Texas 77030, United States ‡ Department of Obstetrics and Gynecology and Comprehensive Cancer Center, Division of Senology, Medical University of Vienna, Vienna 1090, Austria § Department of Experimental Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, Texas 77030, United States ⊥ Department of Laboratory Medicine, The University of Texas MD Anderson Cancer Center, Houston, Texas 77030, United States ¶ Department of Cell and Developmental Biology, Weill Cornell Medical College of Cornell University, New York, New York 10021, United States ∥ Department of Internal Medicine, Weill Cornell Medical College of Cornell University, New York, New York 10021, United States * Corresponding Authors [email protected]. Phone: +1-713-441-5530. [email protected]. Phone: +1-713-441-8439. 18 11 2016 26 4 2016 6 5 2016 27 11 2016 15 5 15341545 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Women with inherited BRCA1 mutations are more likely to develop breast cancer (BC); however, not every carrier will progress to BC. The aim of this study was to identify and characterize circulating peptides that correlate with BC patients carrying BRCA1 mutations. Circulating peptides were enriched using our well-designed nanoporous silica thin films (NanoTraps) and profiled by mass spectrometry to identify among four clinical groups. To determine the corresponding proteolytic processes and their sites of activity, purified candidate peptidases and synthesized substrates were assayed to verify the processes predicted by the MERPOS database. Proteolytic processes were validated using patient serum samples. The peptides, KNG1K438-R457 and C 3fS1304-R1320, were identified as putative peptide candidates to differentiate BRCA1 mutant BC from sporadic BC and cancer-free BRCA1 mutant carriers. Kallikrein-2 (KLK2) is the major peptidase that cleaves KNG1K438-R457 from kininogen-1, and its expressions and activities were also found to be dependent on BRCA1 status. We further determined that KNG1K438-R457 is cleaved at its C-terminal arginine by carboxypeptidase N1 (CPN1). Increased KLK2 activity, with decreased CPN1 activity, results in the accumulation of KNG1K438-R457 in BRCA1-associated BC. Our work outlined a useful strategy for determining the peptide–petidase relationship and thus establishing a biological mechanism for changes in the peptidome in BRCA1-associated BC. breast cancer BRCA1 mutation circulating peptides biomarker kallikrein-2 Kininogen-1 INTRODUCTION Breast cancer (BC) ranks first among cancer-related deaths in women aged 20–60 years, and it is expected to account for about 29% of all new cancer cases among women in 2016.1 Hereditary factors dictate about 10% of BC cases, chief among them are mutations of the tumor suppressor gene BRCA1 identified in 1994.2 Genetic alterations in BRCA1 are responsible for approximately 50% of these hereditary malignancies.2,3 BRCA1 functions via a homologous recombination-mediated, double-stranded DNA-repair mechanism, which serves to maintain genome stability. DNA damage due to a malfunctioning repair system increases the risk and incidence of tumorigenesis.4,5 The probability of developing BC is significantly increased in women with highly penetrant germ-line mutation(s) in BRCA1. About 57% of female carriers of BRCA1 mutations develop BC by 70 years of age.6–11 These women also tend to develop BC at a younger age compared to their peers with sporadic BC. A notable portion (30–50%) of women carrying BRCA1 mutations never develop BC,6,12,13 but there are no clear features associated with BC development in this population, and only a few studies have attempted to determine the protein profile associated with BRCA1 mutant BC.14–16 Becker et al. detected 35 proteins that were overexpressed in BRCA1 cancer patients using surface-enhanced laser desorption/ionization time of flight (SELDI-TOF) mass spectrometry16 but reported only protein molecular weights without identifying specific proteins. Warmoes et al. identified several markers associated with BRCA1 deficiency in a proteomics study of mouse BRCA1-deficient mammary tumors, but these results have yet to be replicated with BRCA1 BC patients.15 Finally, a recent study profiled the plasma proteomes of four patients with BRCA1 mutant BC, four healthy carriers, and four healthy relatives and found an association with gelsolin although this was a small study.14 Better understanding of what drives BRCA1 mutant BC may lead to new biomarkers and new treatments. PARP inhibitors are currently under evaluation as targeted therapy for metastatic BRCA1 mutant BC in a phase III clinical trial;17 however, there are remaining questions in ongoing targeted therapy research, where better understanding may identify resistance mechanisms and potential therapy targets. Currently, genetic tests for BRCA1 and other BC-related gene mutations are used in the clinic to estimate risk and formulate prevention strategies. Although genetic testing identifies mutation status, it does not provide information about additional factors that influence disease development. Recent studies indicate the important role of proteases and peptidases during tumor angiogenesis, invasion, and metastasis.18 Biopsies are usually needed to evaluate tumor-resident proteases/peptidases for their disease biomarker potential; however, protease/peptidase cleavage products, due to size, likely enter the blood circulation where they may serve as more accessible information conduits than the enzymes themselves.19 Previous studies have illustrated the use of circulating peptides as potential biomarkers for cancer diagnostics and therapeutic evaluations;20–23 however, only a few studies have shown a direct correlation between those circulating peptides and their associated proteases/peptidases.24–26 Our group has developed nanoporous silica thin films (NanoTraps) for peptide enrichment prior to mass spectrometry (MS) analysis as a robust technology platform for accurate and reproducible biomarkers detection.27–30 We have used NanoTrap-MS to monitor peptides secretion at different stages of melanoma with lung metastases and to identify peptide markers for early detection in breast cancer.20,23 In this study, we applied NanoTraps to identify and profile circulating peptides that could distinguish BRCA1 carriers with breast cancer from the healthy carriers and sporadic BC. We further demonstrated a direct link between these peptides and their corresponding tumor-resident peptidases. MATERIALS AND METHODS Clinical Samples Collection The human specimens (132 serum samples) used in this study were collected at the Medical University of Vienna from patients who gave informed consent in a study approved by the Institutional Review Board. All specimens were collected as nonfasting samples in an outpatient setting. Specimens were allowed to clot at room temperature for 60 min before centrifugation. The serum was then collected and aliquoted immediately and stored at −80 °C. Serum samples from cancer patients were collected at the time of diagnosis or a few days after diagnosis. Retrospective samples were given coded labels, and ages were assigned at the time of collection for healthy controls, and at the time of cancer diagnosis for BC patients. Each sample was tested for the BRCA1 mutation(s). Patient characteristics are listed in Table 1, Supplementary Table S1, and the individual patient information, including age, histologic types, grading, TNM staging, and hormonal factors, is shown in Supplementary Tables S2–S4. Cell Lines MDA-MB-231, MCF-7, and MCF-10A cell lines were obtained from the ATCC. The HCC1937 cell line was kindly donated by Dr. Haifa Shen (Houston Methodist Research Institute). Two human BC cell lines, MDA-MB-231 and MCF-7, were maintained separately as adherent monolayers in Dulbecco's modified Eagle's medium (DMEM) medium with 10% fetal bovine serum (FBS) for the studies described below. The human BC cell line HCC1937, harboring BRCA1 mutations, was cultured in RPMI-1640 medium with 10% FBS. The human breast epithelial cell line MCF-10A was cultured in DMEM/F12 (1:1 vol/vol) medium supplemented with 5% horse serum, hydrocortisone (0.5 μg/mL), insulin (10 μg/mL), and epidermal growth factor (20 ng/mL). All cell cultures were also supplemented with penicillin (100 U) and streptomycin (100 μg/mL). Authentication of cell lines was conducted utilizing short tandem repeat (STR) DNA fingerprinting at MD Anderson Cancer Center's “Characterized Cell Line Core”. The BRCA1 mutation in HCC1937 (a C insertion at nucleotide 5382) was determined by DNA sequencing. Peptide Expression Profiling Sample buffer consisting of 50% acetonitrile (ACN; Sigma-Aldrich, St. Louis, MO, USA) and 0.1% trifluoroacetic acid (TFA; Sigma-Aldrich, St. Louis, MO, USA) was prepared in deionized water, and 2.2 μL was added to a 20 μL sample of serum that was thawed on ice. NanoTraps were fabricated as described previously.29 Five microliters of each serum was pipetted into the sample chamber. Samples were incubated at 25 °C in a humidified chamber for 30 min, after which the wells were washed four times with water. Five microliters of sample buffer were added to release peptides from the nanopores, and the peptides fractions were transferred into a tube for further analysis. MALDI-TOF MS Detection To prepare the samples for matrix-assisted laser desorption/ionization (MALDI)-TOF MS analysis, 0.5 μL of each sample was spotted on the MS target plate and allowed to dry completely. Once dry, 0.5 μL of matrix solution [5 g/L of acyano-4-hydroxycinnamic acid (CHCA) in 50% ACN and 0.1% TFA] was spotted on the target plate and left to air-dry. All of the samples were analyzed on an Applied Biosystems 4700 MALDI-TOF Analyzer (Applied Biosystems, Inc., Framingham, MA, USA), operated in positive ion mode with reflector (set laser intensity at 4300 and 5000 shots/sample and mass range 800–5000 Da, with target mass of 3000 Da). Peptide Identification by LC–MS/MS Reversed-phase chromatography was performed on an Agilent 1200 series HPLC autosampler. As gradient solvents for liquid chromatography (LC) analysis, 0.1% formic acid in water and 0.1% formic acid in acetonitrile were used. Samples were dried in a vacuum centrifuge and resuspended in solution (1% formic acid and 5 mM NH4OAc) prior to loading into the HPLC sample port. Analysis of the peptides was conducted on an Orbitrap-XL mass spectrometer (Thermo Scientific, Waltham, MA). The peptides were eluted using a linear gradient of 5–40% ACN over 75 min with flow rate at 0.3 μL/min. The electrospray source maintained at 2.1 kV. Acquisition parameters included: 1 FTMS scan at 60 000 resolution followed by three MS/MS product ion scans (in the ion trap) of two microscans each, 400–2000 Da mass range for MS1, 2000 ion counts as the threshold for triggering MS2, 0.5 Da for mass window of precursor ion selection, relative collision energy at 30%; +2, +3, +4, and +5 charge state for screening, 15 s as dynamic exclusion. The MS data obtained were processed using Proteome Discoverer (Version 1.4.1.14, Thermo Fisher Scientific, Germany) and screened against the SwissProt (SwissProt 010913 (538 849 sequences; 191 337 357 residues)) protein database using the Mascot search engine (Matrix Science, Boston, MA). The precursor and fragment mass tolerances were set to 15 ppm and 0.5 Da, respectively, with a 1.5 signal-to-noise ratio allowance. False discovery rates (FDRs) were determined by searching against a decoy database (0.01 FDR strict −0.05 FDR relaxed). Parameters for the searches were no enzyme, and allowance of nine missed cleavages, the oxidation of methionine and pyro-glutamate formation as the dynamic modification. BRCA1 shRNA Knockdown in MCF-7 and MDA-MB-231 Cells MCF-7 or MDA-MB-231 cells were transduced with lentiviral vectors carrying BRCA1_shRNA_1, BRCA1_shRNA_2, or a control shRNA (Ctrl_shRNA) (Dharmacon, Chicago, IL). The sequences of shRNA 1 and 2 were 5′-TAAGGGACCCTTGCATAGC-3′ and 5′-TTCAGTACAATTAGGTGGG-3′, respectively. The transduced cells were selected using 2 μg/mL puromycin (Invitrogen, Carlsbad, CA) for 48 h. Transduced cells were analyzed using immunoblotting to determine the level of BRCA1 expression. Preparing Conditioned Medium Once grown to approximately 80% confluence, the cells were washed three times with phosphate-buffered saline (PBS) and maintained in serum-free medium for an additional 24 h. Cells were removed by a two-step centrifugation process (300g, 5 min, 4 °C, and then 2000g, 10 min. Four °C) and lysed in Mammalian Protein Extraction Reagent (M-PER) (Pierce, Rockford, IL) containing protease inhibitors (Pierce). Clarified supernatant was collected and concentrated using 10K Millipore centrifugal devices (Amicon Ultra 10K, Millipore, Bedford, MA). Peptidase Expression Assays To examine KLK2 expression in serum samples, high-abundance proteins were first removed using Seppro IgY14 according to the manufacturer's instructions. The pass-through fraction containing KLK2 was measured using an in-house enzyme-linked immunosorbent assay (ELISA) according to the direct ELISA using primary antibody protocol provided by Abcam. The primary anti-KLK2 antibody was obtained from Abcam (Cambridge, MA). Expression levels of CFI were measured using ELISA assay according to the manufacturer's instructions (USCN Life Science Inc., Wuhan, PR China). Western blotting analysis was performed as follows: proteins were separated by gradient sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) (precast gels from Bio-Rad); separated proteins were transferred onto nitrocellulose membranes (Bio-Rad, Richmond, CA); the membranes were probed with peptidase-specific primary antibodies and peroxidase conjugated secondary antibodies. Signals were visualized by chemiluminescence. The ImageJ program was used analyze the density profiles for the protein bands and to normalize samples to the loading control. The primary antibodies used include anti-CPN1 (rabbit polyclonal antibody, Pierce) and anti-CFI (mouse monoclonal, Abcam). RNA Isolation and Quantitative PCR Total RNA was isolated from cells using TRIzol reagent (Life Technologies, Gaithersburg, MD) and reverse transcribed for quantitative real-time PCR. Expression of each peptidase was normalized to expression of GAPDH. The KLK2 primer sequences (forward) 5′-TCAGAGCCTGCCAAGATCAC-3′ and (reverse) 5′-CACAAGTGTCTTTACCACCTGT-3′ yielded a 250 bp PCR product. Depletion of KLK2 and CPN1 from Patient Serum and Conditioned Media Anti-KLK2 or anti-CPN1 antibodies were incubated for 2 h at 37 °C with either diluted patient serum or cell culture conditioned media (CM). Each mixture was incubated for 2 h at 37 °C with Protein A/G agarose beads (Pierce) and centrifuged to remove the antibody–peptidase complexes. Assay for Peptide Degradation in Sera and CM His-tagged KNG1 fragments (His3-KNG1E434-L461-His3 and His3-KNG1K438-R457-His3) were synthesized (98% purity) by GL Biochem (Shanghai, PR China). Prior to addition of the peptides, the serum samples were diluted 1/10 in Tris-Cl buffer, pH 7.5, and the CM were concentrated by buffer exchange into Tris-Cl buffer using 10 kDa centrifugation filters. The synthetic KNG1 fragments were then spiked into serum or concentrated CM at a final concentration of 100 μM, and incubated for 3 h at 37 °C. Cleaved peptide products were fractionated and analyzed via Nanotrap-MS. Statistical Analysis The MALDI-TOF MS data were processed using MarkerView software v. 1.2.1 (AB SCIEX, Concord, Canada), and normalized to the internal standard peptide ACTH 18–39 (Sigma-Aldrich, St Louis, MO). Comparisons of MS data sets (i.e., different patient cohorts) were performed using the unpaired t test with a p-value cutoff of 0.05. Principal components analysis-discriminant analysis (PCA-DA) was carried out with a Pareto Scaling for MS data analysis. Receiver operating characteristic (ROC) curves, used to assess the accuracy of biomarker analysis, were performed with a logistic regression model using SPSS version22 (Chicago, IL). Sensitivity against 1-specificity was plotted, and the area under the curve (AUC) values were computed. Mann–Whitney U analysis of the target peptides was performed. Youden Index (sensitivity-specificity +1) was calculated, and the optimal cutoff values were determined by the maximal Youden index. Kendall's tau-b analysis was used to test the correlations between patient characteristics and each peptide markers. One-way or two-way ANOVA was performed for each comparison of ELISA, Western blotting data, and age differences among the four clinical groups using GraphPad Prism v.6 (GraphPad Software, La Jolla, CA). Quantitative image analysis of immunoblots was conducted using the software ImageJ (Bethesda, MD). All numerical data are presented as mean ± standard deviation, mean ± standard error, or 95% confidence intervals. All statistical tests were considered statistically significant if P was less than 0.05. RESULTS Serum Peptide Profiling by Nanotrap-MS Clinical serum specimens (132 total) used in this study were collected from female patients who were tested for BRCA1 mutations (Table 1). Of these 132 patients, 55 were carriers of hereditary BRCA1 mutations, of whom 28 (median age = 48 years) were diagnosed with BC (BBC), and 27 (median age = 43 years) remained cancer-free (BH). Of the remaining patients, 39 (median age = 50 years) were diagnosed with sporadic breast cancer (SBC), and 38 (median age = 46) were healthy volunteers (WT). Samples in the four groups were age-matched, yielding a p-value of 0.124 when one-way ANOVA was performed to evaluate statistical differences. All patient serum samples were processed on NanoTraps as previously described,20,28 and the enriched peptide fractions were subsequently analyzed by MALDI-TOF MS in the mass range of 800–5000 m/z. Approximately 500 monoisotopic peaks were observed in each MS spectrum (Supplementary Figure S1). Of those, 62 peaks were confirmed by LC–MS/MS (Supplementary Table S5) and imported into MarkerView software for standard t test analysis (Supplementary Table S6). We performed pairwise comparisons of the MS spectra generated from the four sample cohorts. A comparison of the BBC and WT spectra showed a significant increase in the expression of seven peptides in the BBC samples (fold change >1.5, p < 0.01). The expression of two peptides was significantly increased in the BBC spectra when compared to that of SBC (fold change >1.5, p < 0.01). When we aligned these two pair-comparisons, the peak at 2365 m/z appeared as a common factor. Thus, we considered it a potential biomarker for BBC (Figure 1A). When we compared the BBC spectrum against that of BH, one peak at 2021 m/z appeared elevated in the BBC sample (Figure 1B). The median relative intensity and 95% confidence interval (CI) values of the two peptides’ relative intensity were shown in Table 1. There was no statistically significant correlation between the two circulating peptides and clinical variables (Supplementary Table S1). Taken together, these results implicate at least two putative circulating peptides were associated with BBC. As shown in Supplementary Figure S2, LC–MS/MS analysis identified the 2365 m/z peptide peak (KHNLGHGHKHERDQGHGHQR, KNG1K438-R457) as part of high-molecular weight KNG1 and the 2021 m/z peak (SSKITHRIHWESASLLR, C 3fS1304-R1320) as part of complement C3. To assess the diagnostic value of KNG1K438-R457 and C 3fS1304-R1320, we calculated the receiver operating characteristic (ROC) curves. AUC values were determined to be 0.794 (95% CI = 0.683–0.905) for KNG1K438-R457 for distinguishing BBC from SBC (Figure 1C and Supplementary Table S7), and 0.758 (95% CI = 0.638–0.877) for distinguishing BBC from WT (Figure 1D and Supplementary Table S7). The AUC values for KNG1K438-R457 and C 3fS1304-R1320 shown in Figure 1, panel E are 0.640 (95% CI = 0.489–0.791) and 0.685 (95% CI = 0.542–0.828), respectively, when comparing BBC to BH. The ROC curves were improved in a multivariate model using the two peptides, and AUC value was 0.739 (95% CI = 0.601–0.878) for distinguishing BBC from BH (Figure 1E). The optimal cutoff value (BBC vs WT) was obtained by the Youden index: the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy (ACC) were 0.750, 0.769, 0.700, 0.811, and 0.742, respectively, to differentiate patients with BBC from WT by KNG1K438-R457 (Figure 1F). By using the two peptides pattern, KNG1K438-R457 and C 3fS1304-R1320, to identify BBC from BH, the sensitivity, specificity, PPV, NPV, and ACC are 0.540, 1.000, 1.000, 0.675, and 0.764, respectively. The optimal cutoff values for different comparison between groups, BBC versus WT, BBC versus SBC, or BBC versus BH, are shown in Supplementary Table S8. Therefore, we proposed to investigate the clinical value of KNG1K438-R457 and C 3fS1304-R1320 for distinguishing BBC from WT, SBC, or BH. Peptidase Prediction and Validation We searched the peptidase database, MEROPS, for peptidases that have potential to cleave KNG1 to KNG1K438-R457 (Supplementary Table S9). Human kallikrein-2 (KLK2) was identified as being capable of cleaving KNG1 at two sites, R437-K438 and R457-G458. We observed increased amounts of the Des-Arg-KNG1 fragment in BBC versus BH (Supplementary Figure S3). Carboxypeptidase N (CPN) cleaves peptides at the carboxy-terminal arginine, and it was demonstrated to cleave arginine from C 3f to generate Des-Arg-C 3f.31 We believe CPN1, the catalytic subunit of CPN, to be the peptidase responsible for generating Des-Arg-KNG1-fragments. For the other peptide marker, C 3fS1304-R1320, a heptadecapetide, is known to be cleaved from C3b by complement factor I (CFI).31 To verify the suggested peptidase–substrate pairing, we synthesized two versions of the KNG1 fragments, His3-KNG1E434-L461-His3 and His6-KNG1K438-R457-His6, containing KLK2 and CPN1 cleavage sites, respectively. These peptides were added to a solution containing purified recombinant peptidase or clinical samples. The degraded peptide fragments were extracted and analyzed using Nanotrap-MS. Each prominent peak in the MS spectra, representing a cleaved peptide fragment, was matched to a cleavage prediction generated by FindPept (http://web.expasy.org/findpept/) (Supplementary Table S10). The 2365 m/z peak of KNG1K438-R457 was observed after incubating His3-KNG1E434-L461-His3 with purified KLK2 (Supplementary Figure S4 and Table S10). Given the complexity of serum, His3-KNG1E434-L461-His3 may also be cleaved by other peptidases. To demonstrate specificity to KLK2, we spiked the fragment into whole serum or KLK2-depleted serum, respectively, and then processed the samples through Nano-trap-MS. In the whole serum sample, a 3140 m/z peak (2365 m/z fragment with His-tags) appeared with increasing intensity and a corresponding decrease in intensity of the 4093 m/z peak (full-length peptide subtract). This observation suggests that 3140 m/z signifies a cleavage product of KLK2 (Figure 2A,B). After the depletion of KLK2 from whole serum, the signal corresponding to the cleavage product at 3140 m/z was dramatically lower; in contrast, the full-length substrate remained stable in KLK2-depleted serum. We also incubated whole serum with protein A/G agarose beads as a negative control, and the resulting cleavage product at peak 3140 m/z showed a signal comparable with that detected in whole serum. This result indicates the degraded product is KLK2 specific. The slightly lower intensity in the negative control may be due to nonspecific binding between KLK2 and protein A/G. Taken together, these results support the notion that KLK2 is the major peptidase largely responsible for cleaving KNG1 to generate KNG1K438-R457 under biological conditions. A similar experiment to examine CPN1 peptidase activity in serum revealed a 3031 m/z peak (the fragment at 2021 m/z, with His-tags) increased in intensity when the substrate was incubated with whole serum, whereas little to no fragment accumulated in CPN1-depleted sera (Figure 2C,D). The 3740 m/z peak lacks a histidine compared to fragment His6-KNG1K438-R457-His6, and was likely generated by a peptidase other than CPN1. As expected, the peak intensity at 3740 m/z remained unchanged irrespective of CPN1 depletion (Figure 2D). These observations suggested that KLK2 acts on KNG1 to generate KNG1K438-R457, which in turn serves as a substrate for CPN1, cleaving it to des-Arg-KNG1K438-R457. Specific Peptidase Activities Correlate with the Appearance of Peptide Fragments in Serum To demonstrate a direct correlation between the appearance of peaks at 3140 m/z and 3031 m/z to KLK2 and CPN1 cleavage, respectively, we added His-tagged peptides substrates into sera from each of the four sample cohorts and processed them through Nanotrap-MS. We observed a significant increase in the relative peak intensity of 3140 m/z in the BBC cohort, indicating an increase in KNG1 cleavage by KLK2 (Figure 3A), confirming the initial analysis with the endogenous KNG1 fragment. We conducted a similar experiment to evaluate the activity of CPN1, which, unlike KLK2, appeared to be lower in BBC compared to the WT group (Figure 3B). In fact, CPN1 activity appeared diminished in all three groups with BC, regardless of BRCA1-mutation status. This observation aligns with the appearance of CPN1-dependent fragments at 2209 and 2080 m/z, both of which were reduced in the BC groups (Supplementary Figure S3). These results indicate two possible regulatory factors that could give the accumulation of fragment KNG1K438-R457 in BBC: one is the increased peptidase activity of KLK2 and the other is reduced activity of CPN1, leading enhanced KNG1K438-R457 levels in BBC patients. Intracellular and Extracellular Peptidase Expression of BC Cells To determine the degree to which the BRCA1-associated peptidases are expressed within BC cells or released to the extracellular medium, we examined the expression of KLK2 in four different cell lines: MCF-10A (nontumorigenic), MCF-7 (tumorigenic, triple-positive, BRCA1 WT), MDB-MB-231 (tumorigenic, triple-negative, BRCA1 WT), and HCC1937 (tumorigenic, BRCA1-mutated). Compared to MCF-10A cells, we detected a 58-fold increase in KLK2 mRNA expression in HCC1937 cells, a 23-fold increase in MCF-7 cells, and a 16-fold increase in MDA-MB-231 cells (Figure 3C). To evaluate whether elevated KLK2 mRNA expression corresponded to increased secretion of the peptidase, we assessed the enzyme levels in CM of these cells. Consistent with KLK2 mRNA expression, secreted levels of KLK2 levels also appeared higher in HCC1937 CM, as determined by ELISA for detection of low-abundance KLK2 than immunoblotting (Figure 3D). CPN1 combined with KLK2 was demonstrated to regulate a peptide at 2365 m/z. CFI releases C 3f from C3b to form inactive iC3b (iC3b). We thus analyzed the expression and secretion of CPN1 and CFI, both BRCA1-associated peptidases. In all of the cell lines tested, CPN1 was almost entirely secreted (Figure 4A), as CPN1 signal in cell lysates was barely detectable (Figure 4B). Approximately 50% less CPN1 was detected in the CM of all three breast cancer cell lines (Figure 4C) compared to MCF-10A cells. CPN1 levels were higher in CM from BRCA1 mutant BC cells compared to BRCA1-wildtype BC cells. The level of secreted CFI decreased in CM from all of the cancer cell lines tested, and no significant difference was seen between BRCA1-wild-type and BRCA1 mutant cancer cells. These results indicate that tumor-resident factor(s) can regulate the expression of KLK2 and CPN1. Correlation between KLK2 and BRCA1 Expression To further determine whether KLK2 expression or activity was associated with BRCA1 status, we generated stable MCF-7 and MDA-MB-231 cell lines with shRNA-mediated knockdown of BRCA1 and measured KLK2 expression and activity in these cells. Puromycin was used to select cultures containing >90% stable viral transductants (Supplementary Figure S5). As shown in Figure 5, panels A and B, BRCA1 expression decreased by ~70% in both MCF-7 and MDA-MB-231 cells transduced with BRCA1_shRNA_2 compared to the controls. We therefore focused on MCF-7/MDA-MB-231 cells expressing BRCA1_shRNA_2 (MCF-7BRCA1- or MDA-MB-231BRCA1-) for further related experiments. KLK2 levels were elevated in MCF-7BRCA1- cells and tended to increase in MDA-MB-231shBRCA- cells (Figure 5C). KLK2-dependent peptide products of BRCA1-knockdown MCF-7 cells were analyzed by MS (Supplementary Figure S6), and after normalization with the internal standard peptide (ACTH), KLK2-dependent peptide products exhibited significantly higher levels in BRCA1 knockdown cells (Figure 5D). Thus, the result further demonstrated that the correlations between the activities and expressions of KLK2 and BRCA1 status. Peptidase Expression in Serum We examined the expression of KLK2, CFI, and CPN1 in sera to explore their implications for biological events (i.e., secretion into blood circulation). Because of the low abundance of KLK2 in circulation, high abundance proteins were depleted with an multifactor affinity column prior to ELISA to improve the sensitivity and specificity of KLK2 detection. The amount of KLK2 appeared significantly increased in BBC sera (Figure 6A). CFI levels were significantly reduced in BH sera (Figure 6B). Immunoblots revealed that CPN1 expression remained unchanged across the samples (Supplementary Figure S7). These results indicate a direct correlation of KLK2 and CFI expression with their reference peptides in serum. DISCUSSION Many of the efforts on BC prevention and early detection have focused on identifying BRCA1 mutation carriers,16,32 but we are only beginning to elucidate the mechanisms that increase BC risks for BRCA1 carriers. Only a limited number of studies have attempted to determine protein profiles associated with BRCA1 mutant BC,14–16 and to date no study has focused on peptide–peptidase interactions in BRCA1-related BC, largely due to the lack of tools for peptide profiling and limited access to populations with inherited BRCA1 mutant BC. In this study, we employed our NanoTraps technology coupling with MS to search circulating peptide candidates differentially presented in BRCA1 mutation carriers and also attempted to decipher the proteolytic mechanisms involved in producing these peptides. The peptide KNG1K438-R457 originates from domain 5 of HMW kininogen a 120-kDa glycoprotein, which is composed of heavy and light chains with domains 1–3 and domains 5 and 6, respectively.33,34 Amino acid residues 441–457 are part of a histidine-glycine-rich region of the protein, which was demonstrated to be responsible for binding to negatively charged surfaces.35 This may explain the preference of KNG1K438-R457 and its daughter fragments for the negatively charged NanoTraps used for peptide fractionation in this study. Small peptides cleaved from domain 5 have been indicated as biomarkers for bladder and gastric cancer.22,36 Although we believe proteases play an important role in generating biologically relevant peptides, little is known about the direct correlation between specific peptidases and their peptide products. This is the first report to our knowledge that shows a direct correlation for KLK2 and KNG1K438-R457. We report that the appearance of peptide fragments KNG1K438-R457 and C3fS1304-R1320 in serum depended on the presence of BRCA1 mutations, as KNG1K438-R457 was up-regulated in BRCA1 mutation carriers with BC. We also provide evidence that identifies KLK2 as the peptidase responsible for generating KNG1K438-R457. We also show that the peptidase CPN1 subsequently acts on KNG1K438-R457 at its C-terminal arginine residue (Figure 7A). Interestingly, KLK2 peptidase activity increased, while CPN1 activity decreased in sera from BRCA1 carriers, which resulted in the elevated level of KNG1K438-R457 in BRCA1-associated BC. Although the peptidase activity of CPN1 differed among the sample groups, its expression level remained steady. We speculate that changes in CPN1 peptidase activity may be influenced by other factors that affect its stability in serum (e.g., CPN2). Despite no obvious differences in CPN1 expression in sera among the four groups, we observed less CPN1 secretion by BC cell lines compared to nontumorigenic MCF-10A cells, although it is possible that such differences were obscured by signal saturation since CPN1 is highly abundant in serum. C 3fS1304-R1320 is cleaved from C3b as a result of C3 activation by CFI.31 The expression of CFI in BH samples was significantly lower than that in BBC samples, and its secretion was lower in BC cell lines compared to MCF-10A cells. It is straightforward to assume that peptidase abundance or activity changes in the tumor resulted in the detectable changes in cleaved peptides in the circulation, although tumor and blood changes were not completely identical. This may be partially due to the different methods used to measure peptidases in tumor cells and serum. In addition, BRCA1-associated BC biology remains only partially understood, we cannot definitively differentiate the peptidase activity of tumor-resident enzymes from that of their circulating counterparts. “Mapping” the peptide biomarker landscape of tumor formation and progression will require more information about other organs and tissue networks. We further demonstrated that KLK2 expression and activity are associated with BRCA1 status using shRNA to achieve stable knockdown of BRCA1 in wild type BC cells. We performed an ingenuity pathway analysis, which maps tentative network connections (Figure 7B) to identify possible mechanisms for how the peptidases are activated and how they relate to BRCA1 in BC development. The link between BRCA1 and KLK2 is better recognized than that between BRCA1 and CPN1 or CFI. Three proteins, E1A-binding protein p300 (EP300), androgen receptor (AR), and β-catenin (CTNNB1), were found to be directly connected to BRCA1 and KLK2. Overexpression of BRCA1 down-regulates cellular expression of the transcriptional coactivator EP300 in BC lines.37 Recent microarray analysis of prostate cancer cells identified KLK2 as an EP300-dependent gene.38 It was also recently reported that loss of BRCA1 leads to impaired expression of the nuclear protein CTNNB1 in BC, implicating it in connecting BRCA1 and KLK2.39 Another study revealed that CTNNB1 could enhance AR signaling, possibly affecting KLK2 expression.40 AR signaling is a third potential pathway connecting BRCA1 to KLK2. Studies with purified protein in vitro have shown that AR binds to a protein fragment of BRCA1, and that this interaction can allow activation of AR in prostate cancer cells.41 AR is also expressed in the ER and PR double negative cell line HCC1937.42 AR increases KLK2 mRNA expression in prostate cancer cells,43 and it differentially modulates KLK2 in different BC cells. KLK2 is strongly associated with BRCA1 through various pathways, and more studies are needed to gain a clearer understanding of their relationship and implications in breast cancer development and progression. We applied a simple, robust, and relatively noninvasive approach to identify BRCA1-associated BC peptide biomarkers KNG1K438-R457 and C 3fS1304-R1320. We also presented an analysis of their associated peptidases CFI and KLK2/CPN1. In both the tumor microenvironment and the circulation system, KLK2 cleaves KNG1 to produce KNG1K438-R457, and CPN1 removes the terminal residue to form KNG1K438–456. CFI cleaves C3 to produce C 3fS1304-R1320. Both peptides can be captured using NanoTraps, and their expression levels were associated with cancer status in BRCA1 carriers. We outline a new approach for profiling circulation peptide and determining their relationship with the activity of the corresponding peptidases. Most published cancer biomarkers fail to enter clinical practice. We believe that our strategy for discovering peptide–peptidase relationships in cancer may prove useful for biomarker discovery, but we acknowledge that our results are still in the early phase of biomarker discovery and that future prospective studies are required to validate our findings. We are currently conducting a prospective study to address this issue. Women carrying BRCA1 mutations typically present with BC at a younger age; therefore, the average age of the patients, whose samples are used in this study, is around 45 years. Including older women in the sample cohort would broaden the impact of these results. The long-term longitudinal information would also be greatly beneficial, particularly for cancer-free BRCA1 mutation carriers who maintain their high-risk status. We intend for this strategy to improve the early examination of cancer in the BRCA1 carriers based on the suggestions from the blood-based test. Supplementary Material 0003 0004 SI Part 1 SI Part 2 ACKNOWLEDGMENTS The work was primarily supported by research funding provided by DoD innovator award (DoD W81XWH-09–1–0212). T.Y.H. and M.F. also acknowledge the partial support from the following grants: U54CA143837, NIH U54CA151668, and DoD W81XWH-11–2–0168. We thank Sabitha Prabhakaran, Hanh H. Hoang, Christopher Bone, and Matthew Landry at the Office of Strategic Research Initiatives at Houston Methodist Research Institute for their suggestions, as well as Bo Ning at HMRI for the TOC graphic. Figure 1 Peptide biomarker levels in serum from clinical samples. Comparison of the relative intensities of MS peaks at (A) 2365 m/z (KNG1K438-R457) and (B) 2021 m/z (C 3fS1304-R1320), which represent peptide fragments cleaved from KNG1 and complement C3, respectively. Mean ± standard error is also shown. , P < 0.05; , P < 0.01; *, P < 0.001. (C) Peptide at 2365 m/z yielded AUC value of 0.794 (95% confidence interval [CI] is 0.683 to 0.905) in distinguishing BBC to SBC. (D) Peptide 2365 m/z yielded an AUC value of 0.758 (95% CI is 0.638 to 0.877) in distinguishing BBC to WT. (E) ROC curve generated by comparing the BBC versus BH sample cohorts. Blue dotted line, 2021 m/z; green dotted line, 2365 m/z; red line, multivariate model based on both peptide fragments (2365 m/z and 2021 m/z). The multivariate model AUC is 0.739 (95% CI is 0.601 to 0.878). (F) Sensitivity, specificity, PPV, NPV, ACC, and FDR of the two peptides. Figure 2 MS spectra showing the cleaved peptide fragments after incubation with whole serum or KLK2/CPN1 depleted serum. (A) Degraded products at 3140 m/z derived from His3-KNG1E434-L461-His3 in whole sera or KLK2-depleted sera. (B) The stability of full length of His3-KNG1E434-L461-His3 in whole sera and KLK2-depleted sera. (C) Degraded products with 3031 m/z generated from His6-KNG1K438-R457-His6 in whole sera and CPN1-depleted sera. (D) The stability of Des-His-KNG1K438-R457 in whole sera and KLK2-depleted sera. All the assays were performed by spiking synthesized His3-KNG1E434-L461-His3 or His6-KNG1K438-R457-His6 into biological fluids followed by NanoTraps enrichment. Figure 3 KLK2 and CPN1 activities in patient serum samples. (A) The relative intensity of the KLK2-dependent His-tagged cleavage product at 3140 m/z. (B) The relative intensity of the CPN1-dependent His-tagged cleavage product at 3031 m/z. (C) The fold changes of KLK2 mRNA expression in normal and breast cancer cells. (D) KLK2 expression in conditioned medium (CM) collected from normal and BC cells; equivalent amounts of proteins were coated on the ELISA plate. , P < 0.05; , P < 0.01; , P < 0.001. Figure 4 CPN1 and CFI expression in and secretion from MCF-10A, MDA-MB-231, MCF-7, and H1937 cell lines. (A) CPN1 and CFI expression in cell CM, and a silver stained gel to assess equal sample loading. (B) CPN1 and CFI expression in cell lysates (CL). (C) Histogram of quantitative image analysis of immunoblots of CPN1 and CFI and the silver stained gel performed by software ImageJ. , P < 0.01; *, P < 0.001. Figure 5 Expressions and activities of KLK2 in BRCA1 knocked down cells. (A) Immunoblotting images of BRCA1 expression and (B) relative BRCA1 levels in MDA-MB-231 and MCF-7 cells transduced with lentiviruses carrying BRCA1_shRNA_1, BRCA1_shRNA_2, or a Ctrl_shRNA. The density of each protein band was normalized to that of corresponding GAPDH. (C) Expressions and (D) activities of KLK2 in BRCA1 knockdown MDA-MB-231 and MCF-7 cells. , P < 0.05; **, P < 0.001. Figure 6 KLK2 and CFI expression in serum from clinical samples. (A) KLK2 expression levels in serum. Highly abundant serum proteins, such as albumin and IgG, were removed from serum before ELISA. (B) CFI expression levels in serum from four clinical groups. , P < 0.05; **, P < 0.01. Figure 7 (A) Schematic representation of the peptide generation pathway. KNG1 is first cleaved by KLK2 to form KNG1K438-R457, and subsequently the C-terminal arginine is removed by CPN1. (B) Ingenuity pathway maps between BRCA1 and KLK2. EP300, AR, and CTNNB1 are three key molecules that connect BRCA1 and KLK2. Table 1 Clinical Characteristics and Age Distribution of the Clinical Samples (n = 132) Used for Peptidomic Profiling age range peptide biomarker median (95% CI) groups n age median (95% CI) min max ER(+) PR (+) Her2 (+) KNG1K438-R457 C 3fS1304-R1320 wild type (WT) 38 46(44–50) 23 64 6.85(4.89–7.76) 38.16(32.23–47.95) sporadic breast cancer (SBC) 39 50(46–53) 31 63 26 20 2 5.76(5.27–7.20) 24.38(22.44–41.50) BRCA1 mut healthy (BH) 27 43(40–46) 31 57 8.36(7.63–9.79) 19.30(17.20–24.13) BRCA1 mut breast cancer (BBC) 28 48(42–50) 31 65 3 2 1 10.40(9.48–13.86) 27.24(26.58–49.63) Author Contributions J. F. and M.-K.M.T. contributed equally to this work. J.F., M.-K.M.T., T.Y.H., and M.F. designed the research plan. J.F. performed the experiments. M.-K.M.T. and C.F.S. collected the clinical samples. J.F. and M.-K.M.T. performed data analysis. J.F., M.-K.M.T., T.Y.H., and M.F. wrote the manuscript, and all authors contributed to the revision of the manuscript. All authors have given approval to the final version of the manuscript. The authors declare no competing financial interest. Supporting Information The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jproteome.6b00010. Peptide profiles obtained by MALDI-TOF MS after Nanotrap enrichment and fractionation from clinical serum samples; MS/MS spectra of the two peptides biomarkers; expression levels of daughter peptides from KNG1K438-R457 at 2209 m/z (KHNLGHGHKHERDQGHGHQ) and 2080 m/z (HNLGHGHKHERDQGHGHQ); mass spectra of peptide products of KNG1E434-L461 after cleavage by purified KLK2; phase contrast and GFP expression under a fluorescence microscope; mass spectra of KLK2 related peptide products in MCF-7 cell extraction before and after knockdown of MCF-7; immunoblotting image of CPN1 in serum samples (PDF) Demographic information for cancer patients; individual patient information in SBC; individual information on healthy women and healthy BRCA1 carriers; individual patient information on BBC; identification of peptides by LC−MS/MS; identified peptides t test analysis; area under the curve; actual numbers divided by optimal cutoff value; enzymes predicted by MEROPS to generate kininogen-1 fragments; matching peptides for unspecific cleavage (PDF) Supplementary Table S-5. Identification of peptides by LC-MS/MS (XLSX) Supplementary Table S-6. Identified peptides t-test analysis (XLSX) REFERENCES 1 Siegel RL Miller KD Jemal A Cancer statistics, 2016. Ca-Cancer J. Clin 2016 66 1 7 30 26742998 2 Miki Y Swensen J Shattuck-Eidens D Futreal PA Harshman K Tavtigian S Liu Q Cochran C Bennett LM Ding W Bell R Rosenthal J Hussey C Tran T McClure M Frye C Hattier T Phelps R Haugen-Strano A Katcher H Yakumo K Gholami Z Shaffer D Stone S Bayer S Wray C Bogden R Dayananth P Ward J Tonin P Narod S Bristow PK Norris FH Helvering L Morrison P Rosteck P Lai M Barrett JC Lewis C Neuhausen S Cannon-Albright L Goldgar D Wiseman R Kamb A Skolnick MH A strong candidate for the breast and ovarian cancer susceptibility gene BRCA1. Science 1994 266 5182 66 71 7545954 3 Wooster R Bignell G Lancaster J Swift S Seal S Mangion J Collins N Gregory S Gumbs C Micklem G Identification of the breast cancer susceptibility gene BRCA2. Nature 1995 378 6559 789 92 8524414 4 Gudmundsdottir K Ashworth A The roles of BRCA1 and BRCA2 and associated proteins in the maintenance of genomic stability. Oncogene 2006 25 43 5864 74 16998501 5 Futreal PA Liu Q Shattuck-Eidens D Cochran C Harshman K Tavtigian S Bennett LM Haugen-Strano A Swensen J Miki Y Eddington K Mcclure M Frye C Weaverfeldhaus J Ding W Gholami Z Soderkvist P Terry L Jhanwar S Berchuck A Iglehart JD J, M. G, B. D. Barrett JC Skolnick MH Kamb A Wiseman R BRCA1 mutations in primary breast and ovarian carcinomas. Science 1994 266 5182 120 2 7939630 6 Chen S Parmigiani G Meta-analysis of BRCA1 and BRCA2 penetrance. J. Clin. Oncol 2007 25 11 1329 33 17416853 7 Couch FJ Nathanson KL Offit K Two decades after BRCA: setting paradigms in personalized cancer care and prevention. Science 2014 343 6178 1466 70 24675953 8 Chen S Iversen ES Friebel T Finkelstein D Weber BL Eisen A Peterson LE Schildkraut JM Isaacs C Peshkin BN Corio C Leondaridis L Tomlinson G Dutson D Kerber R Amos CI Strong LC Berry DA Euhus DM Parmigiani G Characterization of BRCA1 and BRCA2 mutations in a large United States sample. J. Clin. Oncol 2006 24 6 863 71 16484695 9 Kroiss R Winkler V Bikas D Fleischmann E Mainau C Frommlet F Muhr D Fuerhauser C Tea M Bittner B Kubista E Oefner PJ Bauer P Wagner TM Younger birth cohort correlates with higher breast and ovarian cancer risk in European BRCA1 mutation carriers. Hum. Mutat 2005 26 6 583 9 16287141 10 Haffty BG Harrold E Khan AJ Pathare P Smith TE Turner BC Glazer PM Ward B Carter D Matloff E Bale AE Alvarez-Franco M Outcome of conservatively managed early-onset breast cancer by BRCA1/2 status. Lancet 2002 359 9316 1471 7 11988246 11 Castilla LH Couch FJ Erdos MR Hoskins KF Calzone K Garber JE Boyd J Lubin MB Deshano ML Brody LC Collins FS Weber BL Mutations in the BRCA1 gene in families with early-onset breast and ovarian cancer. Nat. Genet 1994 8 4 387 91 7894491 12 Struewing JP Hartge P Wacholder S Baker SM Berlin M McAdams M Timmerman MM Brody LC Tucker MA The risk of cancer associated with specific mutations of BRCA1 and BRCA2 among Ashkenazi Jews. N. Engl. J. Med 1997 336 20 1401 8 9145676 13 Antoniou AC Cunningham AP Peto J Evans DG Lalloo F Narod SA Risch HA Eyfjord JE Hopper JL Southey MC Olsson H Johannsson O Borg A Passini B Radice P Manoukian S Eccles DM Tang N Olah E Anton-Culver H Warner E Lubinski J Gronwald J Gorski B Tryggvadottir L Syrjakoski K Kallioniemi OP Eerola H Nevanlinna H Pharoah PD Easton DF The BOADICEA model of genetic susceptibility to breast and ovarian cancers: updates and extensions. Br. J. Cancer 2008 98 8 1457 66 18349832 14 Scumaci D Tamme L Fiumara CV Pappaianni G Concolino A Leone E Faniello MC Quaresima B Ricevuto E Costanzo FS Cuda G Plasma Proteomic Profiling in Hereditary Breast Cancer Reveals a BRCA1-Specific Signature: Diagnostic and Functional Implications. PLoS One 2015 10 6 e0129762 26061043 15 Warmoes M Jaspers JE Pham TV Piersma SR Oudgenoeg G Massink MP Waisfisz Q Rottenberg S Boven E Jonkers J Jimenez CR Proteomics of mouse BRCA1-deficient mammary tumors identifies DNA repair proteins with potential diagnostic and prognostic value in human breast cancer. Mol. Cell. Proteomics 2012 11 7 M111.013334 16 Becker S Cazares LH Watson P Lynch H Semmes OJ Drake RR Laronga C Surfaced-enhanced laser desorption/ionization time-of-flight (SELDI-TOF) differentiation of serum protein profiles of BRCA-1 and sporadic breast cancer. Annals of surgical oncology 2004 11 10 907 14 15383419 17 Litton JK Blum JL Im YH Martin M Mina LA Roche HH Rugo HS Visco F Zhang C Lokker NA Lounsbury DL Eiermann W A phase 3, open-label, randomized, parallel, 2-arm international study of the oral PARP inhibitor talazoparib (BMN 673) in BRCA mutation subjects with locally advanced and/or metastatic breast cancer (EMBRACA). Cancer Res 2015 33 15 OT1 1-12 18 Mason SD Joyce JA Proteolytic networks in cancer. Trends Cell Biol 2011 21 4 228 37 21232958 19 Liotta LA Ferrari M Petricoin E Clinical proteomics: written in blood. Nature 2003 425 6961 905 14586448 20 Fan J Huang Y Finoulst I Wu HJ Deng Z Xu R Xia X Ferrari M Shen H Hu Y Serum peptidomic biomarkers for pulmonary metastatic melanoma identified by means of a nanopore-based assay. Cancer Lett 2013 334 2 202 10 23200677 21 Profumo A Mangerini R Rubagotti A Romano P Damonte G Guglielmini P Facchiano A Ferri F Ricci F Rocco M Boccardo F Complement C3f serum levels may predict breast cancer risk in women with gross cystic disease of the breast. J. Proteomics 2013 85 44 52 23639844 22 Villanueva J Shaffer DR Philip J Chaparro CA Erdjument-Bromage H Olshen AB Fleisher M Lilja H Brogi E Boyd J Sanchez-Carbayo M Holland EC Cordon-Cardo C Scher HI Tempst P Differential exoprotease activities confer tumor-specific serum peptidome patterns. J. Clin. Invest 2005 116 1 271 84 23 Fan J Deng X Gallagher JW Huang H Huang Y Wen J Ferrari M Shen H Hu Y Monitoring the progression of metastatic breast cancer on nanoporous silica chips. Philos. Trans. R. Soc., A 1967 2012 370 2433 47 24 Diamandis EP Peptidomics for cancer diagnosis: present and future. J. Proteome Res 2006 5 9 2079 82 16944917 25 Diamandis EP Oncopeptidomics: a useful approach for cancer diagnosis? Clin. Chem 2007 53 6 1004 6 17517585 26 Petricoin EF Belluco C Araujo RP Liotta LA The blood peptidome: a higher dimension of information content for cancer biomarker discovery. Nat. Rev. Cancer 2006 6 12 961 7 17093504 27 Fan J Gallagher JW Wu HJ Landry MG Sakamoto J Ferrari M Hu Y Low molecular weight protein enrichment on mesoporous silica thin films for biomarker discovery. J. Visualized Exp No 2012 62 e3876 28 Hu Y Peng Y Lin K Shen H Brousseau LC 3rd Sakamoto J Sun T Ferrari M Surface engineering on mesoporous silica chips for enriching low molecular weight phosphorylated proteins. Nanoscale 2011 3 2 421 8 21135976 29 Hu Y Bouamrani A Tasciotti E Li L Liu X Ferrari M Tailoring of the nanotexture of mesoporous silica films and their functionalized derivatives for selectively harvesting low molecular weight protein. ACS Nano 2010 4 1 439 51 20014864 30 Fan J Niu S Dong A Shi J Wu HJ Fine DH Tian Y Zhou C Liu X Sun T Anderson GJ Ferrari M Nie G Hu Y Zhao Y Nanopore film based enrichment and quantification of low abundance hepcidin from human bodily fluids. Nanomedicine 2014 10 5 e879 e888 31 Ganu VS Muller-Eberhard HJ Hugli TE Factor C3f is a spasmogenic fragment released from C3b by factors I and H: the heptadeca-peptide C3f was synthesized and characterized. Mol. Immunol 1989 26 10 939 48 2531841 32 Weitzel JN Clague J Martir-Negron A Ogaz R Herzog J Ricker C Jungbluth C Cina C Duncan P Unzeitig G Saldivar JS Beattie M Feldman N Sand S Port D Barragan DI John EM Neuhausen SL Larson GP Prevalence and type of BRCA mutations in Hispanics undergoing genetic cancer risk assessment in the southwestern United States: a report from the Clinical Cancer Genetics Community Research Network. J. Clin. Oncol 2013 31 2 210 6 23233716 33 Zhang JC Claffey K Sakthivel R Darzynkiewicz Z Shaw DE Leal J Wang YC Lu FM McCrae KR Two-chain high molecular weight kininogen induces endothelial cell apoptosis and inhibits angiogenesis: partial activity within domain 5. FASEB J 2000 14 15 2589 600 11099478 34 Colman RW Jameson BA Lin Y Johnson D Mousa SA Domain 5 of high molecular weight kininogen (kininostatin) down-regulates endothelial cell proliferation and migration and inhibits angiogenesis. Blood 2000 95 2 543 50 10627460 35 Dela RAC Colman RW The sequence HGLGHGHEQQHGLGHGH in the light chain of high molecular weight kininogen serves as a primary structural feature for zinc-dependent binding to an anionic surface. Protein Sci 1992 1 1 151 60 1304876 36 Umemura H Togawa A Sogawa K Satoh M Mogushi K Nishimura M Matsushita K Tanaka H Takizawa H Kodera Y Nomura F Identification of a high molecular weight kininogen fragment as a marker for early gastric cancer by serum proteome analysis. J. Gastroenterol 2011 46 5 577 85 21298293 37 Fan S Ma YX Wang C Yuan RQ Meng Q Wang JA Erdos M Goldberg ID Webb P Kushner PJ Pestell RG Rosen EM p300 Modulates the BRCA1 inhibition of estrogen receptor activity. Cancer Res 2002 62 1 141 51 11782371 38 Ianculescu I Wu DY Siegmund KD Stallcup MR Selective roles for cAMP response element-binding protein binding protein and p300 protein as coregulators for androgen-regulated gene expression in advanced prostate cancer cells. J. Biol. Chem 2012 287 6 4000 13 22174411 39 Li H Sekine M Tung N Avraham HK Wild-type BRCA1, but not mutated BRCA1, regulates the expression of the nuclear form of beta-catenin. Mol. Cancer Res 2010 8 3 407 20 20215423 40 Chesire DR Ewing CM Gage WR Isaacs WB In vitro evidence for complex modes of nuclear beta-catenin signaling during prostate growth and tumorigenesis. Oncogene 2002 21 17 2679 94 11965541 41 Yeh S Hu YC Rahman M Lin HK Hsu CL Ting HJ Kang HY Chang C Increase of androgen-induced cell death and androgen receptor transactivation by BRCA1 in prostate cancer cells. Proc. Natl. Acad. Sci. U. S. A 2000 97 21 11256 61 11016951 42 Toth-Fejel S Cheek J Calhoun K Muller P Pommier RF Estrogen and androgen receptors as comediators of breast cancer cell proliferation: providing a new therapeutic tool. Arch. Surg 2004 139 1 50 4 14718276 43 Dong Y Zhang HT Gao AC Marshall JR Ip C Androgen receptor signaling intensity is a key factor in determining the sensitivity of prostate cancer cells to selenium inhibition of growth and cancer-specific biomarkers. Mol. Cancer Ther 2005 4 7 1047 1055 16020662
PMC005xxxxxx/PMC5124560.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 0322116 6595 Prev Med Prev Med Preventive medicine 0091-7435 1096-0260 27677441 5124560 10.1016/j.ypmed.2016.09.027 NIHMS821449 Article Trends in smoking prevalence and attributable mortality in China, 1991–2011 Li Shuangshuang a1 Meng Linghui b1 Chiolero Arnaud cd Ma Chuanwei a Xi Bo a* a Department of Epidemiology, School of Public Health, Shandong University, Jinan 250012, China b Department of Epidemiology, Capital Institute of Pediatrics, Beijing 10020, China c Division of Chronic Diseases, Institute of Social and Preventive Medicine (IUMSP), Lausanne University Hospital, Lausanne, Switzerland d Department of Epidemiology, Biostatistics, and Occupational Health, McGill University, Montreal, Canada * Corresponding author. [email protected] (B. Xi) 1 These authors contributed equally to this work. 18 11 2016 24 9 2016 12 2016 01 12 2017 93 8287 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Objective China is the largest producer of tobacco worldwide. We assessed secular trends in prevalence of smoking, average cigarettes per day, mean age of initiation, and mortality attributable to smoking among the Chinese population between 1991 and 2011. Design Data came from the China Health and Nutrition Survey, conducted eight times between 1991 and 2011. A total of 83,447 participants aged 15 years or older were included in this study. Trends in smoking were stratified by sex, age, and region (urban vs. rural). Results In 2011, 311 millions individuals were current smokers in China, with 295 million men and 16 million women, respectively. Between 1991 and 2011, the prevalence of current smoking decreased from 60.6% to 51.6% in men, and from 4.0% to 2.9% in women. However, during this period, the average number of cigarettes smoked per day per smoker increased from 15.0 to 16.5 in males, and from 8.5 to 12.4 in females. Further, age of smoking initiation decreased from 21.9 to 21.4 years in men and from 31.4 to 28.4 years in women. In 2011, 16.5% of all deaths in men and 1.7% in women were due to smoking. Between 1991 and 2011, the total number of deaths caused by smoking increased from 800,000 to 900,000. Conclusions During the past 20 years, a slight decrease in smoking prevalence was observed in the Chinese population. However, cigarette smoking remains a major cause of death in China, especially in men. Smoking Epidemiology Trends China 1. Introduction Cigarette smoking is the leading risk factor for many non-communicable diseases and premature mortality worldwide (Chen et al., 2015; WHO, 2015). Tobacco use was responsible for 4.8 million deaths in 2000 and 6.0 million in 2011 worldwide (Asma et al., 2014; Ezzati and Lopez, 2003). By 2030, tobacco use is expected to cause 8.3 million deaths, accounting for 10% of the all-cause mortality globally (Mathers and Loncar, 2006). China is the largest tobacco producer worldwide. The number of cigarettes produced in China increased from 0.5 trillion in 1980 to 2.6 trillion in 2013, corresponding to 43% of the current world’s tobacco production. 99% of cigarettes produced in China are consumed domestically and only 1% exported (Yang et al., 2008). It has been estimated that >300 million adults were smokers in China, accounting for a third of the world’s total number of smokers (No authors listed, 2011). As a result, China bears tremendous economic and disease burdens attributable to smoking. In 2008, smoking was estimated to have cost China about $5 billion for treatment of smoking-related diseases (direct costs) and $29 billion in total economic lost (direct and indirect costs) (No authors listed, 2014b; Yang et al., 2011). Tobacco has been estimated to account for 9.5% of disability-adjusted life-years and 16.4% of deaths among Chinese adults (Koplan et al., 2010; Yang et al., 2013). In consideration of the great number of smokers and the large health hazard attributable to smoking, the Chinese government has adopted a wide range of interventions to curb tobacco use, including tax increases, bans on advertising, and smoke-free laws. Furthermore, the Chinese government ratified the World Health Organization (WHO) Framework Convention on Tobacco Control (FCTC) in 2005. Despite these efforts, a number of serious challenges remain. The Global Adults Tobacco Survey (GATS) showed that 28.1% of Chinese adults (52.9% of men and 2.4% of women) were current smokers in 2010 (Li et al., 2011). In addition, among the 16 countries that had completed the survey, China ranked second for smoking among men, between Russia (60.2%) and Ukraine (50.0%) (Giovino et al., 2012). In China, patterns of tobacco use have evolved along social and economic development and progress of implementation of tobacco control policies. However, only few studies have documented trends in smoking and the impact on mortality in the Chinese population (Chen et al., 2015; Giovino et al., 2012; Gu et al., 2009; Lam et al., 1997; Qian et al., 2010). Monitoring trends in smoking and its impact is critical for policy makers in order to guide appropriate tobacco control interventions. Hence, based on data of successive national Chinese surveys (the China Health and Nutrition Surveys, CHNS), we assessed secular trends in smoking, average number of cigarettes smoked daily, mean age of smoking initiation, and mortality attributable to smoking in the Chinese population between 1991 and 2011. 2. Methods 2.1. Study design and subjects The CHNS is designed as a national, large scale survey to examine the health and nutritional status of the Chinese population over time. It is an international collaborative project between the Carolina Population Center of the University of North Carolina and the National Institute of Nutrition and Food Safety of the Chinese Center for Disease Control and Prevention since 1989. Surveys are carried out using a multistage random cluster sampling strategy to obtain population based data in nine Chinese provinces (Liaoning, Heilongjiang, Jiangsu, Shandong, Henan, Hubei, Hunan, Guizhou and Guangxi), which vary in geography, economic status, public resources, and health indicators. Detailed information of CHNS has been published elsewhere (Popkin et al., 2010). Between 1991 and 2011, a total of 83,447 participants aged 15 years or older in eight surveys (9394 in 1991, 8784 in 1993, 11,149 in 1997, 10,116 in 2000, 10,313 in 2004, 10,099 in 2006, 10,278 in 2009, and 13,314 in 2011) were included in data analyses. Characteristics of the survey participants are shown in the Supplemental Table 1. The participation rate was >98% in each survey. Written informed consents were obtained from all participants. This study was approved by the Institutional Review Board from both the University of North Carolina at Chapel Hill and the China Center for Disease Control and Prevention. 2.2. Definitions Between 1991 and 2011, smoking habits were assessed by 4 questions that were kept identical in each survey. The first question was “Have you ever smoked cigarettes (including hand-rolled, device-rolled or pipe)?” Responses included “never smoked” and “yes”. The second question was “Do you still smoke cigarettes now?” Participants who answered “yes” were defined as “current smoker”. The third question was “How many cigarettes do you smoke per day?”. The fourth question was “How old were you when you started to smoke?”. 2.3. Statistical analysis Categorical variables were expressed as percentage (SE), while continuous variables were presented as mean (SE). The estimates of current smoking or ever smoking, number of cigarettes used per day and the initiation age by sex, age, and region are described in this study. The sex-and age- specific China census distribution of the population in 2010 was used to standardize estimates in all surveys. Trends in prevalence of current smoking and ever smoking between 1991 and 2011 were examined using multiple logistic regression adjusted for sex, age and region; trends in number of cigarettes smoked per day and age of smoking initiation were assessed by similarly adjusted multiple linear regression. To calculate the population attributable risk (PAR) of smoking, we used the following formula: PAR = (P × [RR −1]) ÷ (P × [RR − 1] + 1), where P is the prevalence of smokers, and RR is the relative risk of disease or mortality (Gu et al., 2009). Estimates for RRs for all-cause mortality, cancer, respiratory disease and cardiovascular disease were extracted from a publication by Chen et al. (2015). All data were analyzed using the statistical package SPSS (version 16.0). Two-side p values of <0.05 indicated statistical significance. 3. Results Supplemental Table 1 presents characteristics of the study populations in each survey between 1991 and 2011. There were differences across the eight surveys, with increasingly old and urban populations. Table 1 shows trends in prevalence of current smoking by sex, age, and region between 1991 and 2011. In 2011, it was estimated that 311 millions individuals were current smokers in China, with 295 million men and 16 million women, respectively. The prevalence of smoking decreased from 1991 to 2011 in both sexes, i.e., from 60.6% to 51.6% in men and from 4.0% to 2.9% in women, respectively (for both sexes, p for trends <0.001) (Fig. 1A). Decreasing prevalence was found in all age and region (urban/rural) subgroups (p for trend <0.001) except for younger women aged 15–24 years for which the decrease did not reach statistical significance (p = 0.223). The prevalence of current smoking was highest for men aged 25–44 years old and 45–64 years old. In women, prevalence of smoking was much lower compared with men in each survey year; in 2011, the prevalence ratio of smokers was 1 woman for 18 men. In men, the prevalence of smoking was slightly higher in rural than urban areas in a consistent manner between 1991 and 2011. In women, while the prevalence was higher in urban than rural areas until 2004, the urban-rural difference vanished thereafter. For the prevalence of ever smoking, similar trends were found in the total population and in sex, age and region subgroups (Supplemental Table 2 and Fig. 1B). In addition, during the period of 1991 to 2011, the prevalence of former smoking increased from 3.7% to 8.4% in men, while it remained stable at 0.5% in women. Among current smokers, the number of cigarettes smoked per day increased from 15.0 to 16.5 in men and from 8.5 to 12.4 in women (Supplemental Table 3 and Fig. 1C). Similar upward trends were found in most age and region subgroups. In men, subjects aged 45–64 years experienced the highest increase in mean number of cigarettes smoked per day among all four age groups. In women, the greatest increase occurred in subjects aged 25–44 years. The mean age of smoking initiation declined in the whole population and in most age and region subgroups between 1991 and 2011(for most subgroups, p for trends <0.001). We restricted this analysis to subjects aged 25 years or older as the daily smoking rates were not stable for individuals aged 15–24 years. The decrease was more important in women (from 31.4 to 28.4 years old; difference of 3.0 years) compared to men (from 21.9 to 21.4 years old; difference of 0.5 years) (Supplemental Table 4 and Fig. 1D). In each survey, mean age of initiation was lower for men than women. Furthermore, mean age of initiation was lower for rural than urban female smokers while no obvious region difference was found in male smokers. In China, the total number of deaths from any cause increased from 5.9 million in 1990 to 7.0 million in 2010 (Yang et al., 2015). In 2011, the proportion of deaths due to smoking (i.e., the population attributable risk) was 16.5% in men and 1.7% in women. When applied to the whole population, this proportion corresponds to 900,000 deaths (800,000 in men and 100,000 in women). Between 1991 and 2011, the proportion of deaths due to smoking slightly decreased because of the small reduction in smoking prevalence on both sexes (Table 2). However, the number of deaths attributable to smoking increased markedly, from 800,000 in 1991 to 900,000 in 2011 (men: 700,000 to 800,000; women: 100,000 to 100,000). 4. Discussion In 2011, smoking remained common in the Chinese population (about 311 millions current smokers), especially in men. Between 1991 and 2011, the prevalence of smoking declined slightly in both sexes, and the number of cigarettes smoked daily increased and the age of smoking initiation decreased in all sex, age, and region subgroups. Importantly, the number of deaths attributable to smoking increased from 800,000 in 1991 to 900,000 in 2011 despite the slight decrease in the prevalence of smokers. Cigarette smoking remains a major public health threat in China. The prevalence of smoking was high in China, especially among men. In the early 1970s, a regional survey including 9351 middle-aged smokers in Shanghai found that 61% of men and 7% of women were smokers (Lam et al., 1997); in 1976, the corresponding figures from another regional survey in Xi’an were 56% and 12% (Chen et al., 1997). Based on national data from the National Health Service Survey (NHSS), it was reported that the prevalence of current smoking decreased from 59.6% in 1993 to 48.9% in 2003 in men, and from 5.1% in 1993 to 3.2% in 2003 in women (Qian et al., 2010). A recent nationally representative survey f(China Non-communicable Disease Surveillance) showed that 54.1% of men and 2.6% of women were current smokers in 2011 (Bi et al., 2015) and the Global Adult Tobacco Survey 2008–2010 showed that 52.9% of men and 2.4% of women were current smokers in China (Giovino et al., 2012). While these observations are consistent with our findings, the comparability between these surveys is limited by differences in study design, sample sources, geographic coverage, and definition of smoking. Major strengths of the CHNS data used in the current study are the same study design, same geographic regions, and same definition of smoking across all surveys between 1991 and 2011, which allows direct comparison of findings over time. The prevalence of smoking in China is particularly high compared to other countries around the world, at least for men (Giovino et al., 2012; Ng et al., 2014). In 2008–2010, data from 16 countries indicated that China had the second highest prevalence (52.9%) of smoking among men behind Russia (60.2%) (Giovino et al., 2012). In 2012, a large review including 187 countries suggested that 12 countries including China had a smoking prevalence of >40% in males and accounted for nearly 40% of all smokers globally (Ng et al., 2014). These findings underline the large scale of the smoking epidemic in China, and the urgent need to strengthen tobacco control interventions. The prevalence of smoking was much higher in men than women. In addition, men started to smoke cigarettes earlier than women. Although the Chinese society has been influenced by the western culture during the past decade, it remains quite conservative and rooted into the traditional Confucian idea. The society is quite tolerant to male smoking but not to female smoking (Kim, 2005). These socio-cultural characteristics might be a major factor in the largely higher smoking prevalence and earlier onset of smoking in men than women. In 2011, we estimate that cigarette smoking caused about 16.5% of all male deaths and 1.7% of all female deaths in China based on the prevalence of smoking found in our study and RRs of diseases due to smoking estimated by Chen et al. (2015) A previous study by Gu et al. reported that cigarette smoking accounted for 12.9% of deaths in men and 3.1% of deaths in women in China (Gu et al., 2009). Most recently, Chen et al. showed that 18% of deaths in men and 3% of deaths in women were attributable to tobacco use in China (Chen et al., 2015). The risk of mortality attributable to smoking was much smaller in women than in men mainly due to the much lower prevalence of smoking in men than women although the RRs of diseases due to smoking were similar for both sexes. In our study, the proportion of deaths due to smoking slightly de-creased over time because of the slight reduction in smoking prevalence on both sexes between 1991 and 2011. However, the absolute number of deaths attributable to smoking increased, consistent with the increasing size of the Chinese population during the past 20 years. It is likely that the absolute number of deaths in China attributable to tobacco will continue to rise because of the continued population growth, unless the prevalence of smoking decreases. This serious situation underlines the need to strengthen tobacco control interventions in China. However, the reported RR for all-cause mortality in the Chinese population was markedly lower (e.g., 1.21–1.33 for men) (Chen et al., 2012; Gu et al., 2009) than that reported in many western countries (e.g., 2.8 for men) (Thun et al., 2013). It seems that cigarettes smoking may pose less serious effect on health in Chinese population than in western countries. However, future studies are necessary to further determine the underlying biological mechanism on why smoking play different roles in human health across different populations. The slight decrease in the prevalence of current smoking suggested that China made some progress in tobacco control. However, major gaps still exist when compared to the WHO FCTC requirements. Several barriers impede tobacco control in China (Yang et al., 2015). First, there is a large conflict between the huge scale of the tobacco industry in China, largely controlled by the government, and tobacco control policies. Government is reluctant to scale down an industry that it controls and this industry is a large source of tax revenues. Second, effective interventions are not available to support smoking cessation. Although > 800 cessation clinics have been set up in China since 1996, few smokers are using these clinics. Third, warnings only cover 30% of the bottom of packages of cigarettes, compared to the size of at least 50% as recommended by the WHO FCTC (Kaul and Wolf, 2014; No authors listed, 2014a). Fourth, tax accounted for only 40–46% of the tobacco product retail price in 2011, which is much lower than the proportion required by the WHO FCTC (70%). (Yang et al., 2015) Thus, cigarettes in China are quite inexpensive, with an average price of ¥5.0 (around US $0.74) for one pack of locally manufactured cigarettes (Jena et al., 2012). Fifth, there was no national legislation in China to ban tobacco in public and work places until 2014, as well as no limitation on tobacco advertising. Finally, smoking is still socially accepted, and cigarettes are still regarded as a courtesy in most social events in China, such as weddings, funerals, and official activities. Expensive cigarettes are a usual gift for relatives, friends, guests or visitors (Rich and Xiao, 2012). Our study has several strengths. It includes a large number of participants (n = 83,447) with a large response rate in each survey (98% or more). Furthermore, the same standardized protocol was used to collect smoking information in each of the eight national surveys. In addition, strict quality control measures were applied in each survey, including trained examiners and calibrated instruments. Nevertheless, several limitations should also be noted. First, the CHNS only included nine provinces out of the 31 Chinese provinces, and our findings may not necessarily be generalized to the whole China. Second, information on cigarette smoking was based on self-report, and some recall or reporting bias is expected. As a result, the prevalence of smoking may be underestimated. Third, we only reported the trends in prevalence of cigarette smoking rather than the prevalence of any other forms of tobacco use. However, it has been previously shown that cigarettes account for >95% of all tobacco products used in China (Giovino et al., 2012). Fourth, due to the cross-sectional nature of our study design, the inferences should be drawn with caution. Fifth, our study provided secular trends in smoking estimates over time according to a limited number of characteristics (age, sex, and region). Future studies are needed to investigate other factors that might have influenced these changes. In conclusion, this study shows that although the prevalence of smoking slightly decreased since 1991, it remains very high in 2011, especially in men. Furthermore, the daily number of cigarette increased over time and the initiation age of smoking decreased during the past ~20 years. Importantly, the absolute number of deaths due to smoking increased between 1991 and 2011 because of the increasing size of the population. To control the epidemic of cigarette smoking, the Chinese government should strengthen a tobacco control measures and stop supporting the local tobacco industry. In addition, government should increase tobacco taxation, further implement bans on smoking in public place, ban advertisement promotion and sponsoring of tobacco product measures and further develop tobacco cessation programs. Fortunately, at the end of 2014, the China State Council released a long-anticipated guideline on national tobacco control, followed by tobacco control legislation. The potential for public health benefit of stricter tobacco control measures is huge: it was estimated that >12.8 million smoking-related deaths by 2050 in China could be prevented if the provisions recommended by the WHO FCTC were implemented (Levy et al., 2014). Supplementary Material sup Funding This study was supported by the National Institutes of Health (NIH) (grants R01-HD30880, DK056350, R24-HD050924, and R01-HD38700). The sponsors have no role in the study design, survey process, data analysis and manuscript preparation. We thank the National Institute of Nutrition and Food Safety of China Center for Disease Control and Prevention, and Carolina Population Center of the University of North Carolina at Chapel Hill for sharing their valuable data. We thank Pascal Bovet (Division of Chronic Diseases, Institute of Social and Preventive Medicine (IUMSP), Lausanne University Hospital, Lausanne, Switzerland) for help improve the manuscript. Abbreviations CHNS China Health and Nutrition Surveys FCTC Framework Convention on Tobacco Control GATS Global Adults Tobacco Survey RR Relative risk WHO World Health Organization Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.ypmed.2016.09.027. Fig. 1 Trends in (A) the prevalence of current smoking, (B) prevalence of ever smoking; (C) mean cigarettes smoked per day per smoker; and (D) average age of smoking initiation in the Chinese population, 1991–2011. Data are age- and sex- standardized to the 2010 China Census population. Table 1 Trends in the prevalence of current smoking in China, 1991–2011. 1991 1993 1997 2000 2004 2006 2009 2011 P for trendsa % SE % SE % SE % SE % SE % SE % SE % SE Total 31.0 0.5 30.6 0.5 28.9 0.4 28.8 0.5 27.8 0.4 26.5 0.4 27.5 0.4 25.8 0.4 <0.001 Age (years) 15–24 18.8 0.8 19.6 0.9 15.0 0.7 16.0 1.0 12.9 1.0 14.4 1.2 17.3 1.3 15.1 1.2 <0.001 25–44 34.7 0.7 32.9 0.8 33.5 0.7 31.2 0.7 30.4 0.8 27.6 0.8 28.6 0.8 25.9 0.7 <0.001 45–64 36.7 1.0 35.9 1.0 34.8 0.9 32.6 0.8 31.3 0.7 30.3 0.7 30.0 0.7 28.6 0.6 <0.001 ≥65 28.4 1.6 29.4 1.6 22.4 1.2 25.1 1.2 23.4 10.8 21.2 1.0 24.2 1.0 22.8 0.9 <0.001 Sex Men 60.6 0.7 59.4 0.8 53.9 0.7 55.2 0.7 53.2 0.7 51.8 0.7 53.4 0.7 51.6 0.6 <0.001 Women 4.0 0.3 4.1 0.3 3.9 0.3 4.3 0.3 3.8 0.3 3.4 0.3 3.5 0.3 2.9 0.2 <0.001 Region Urban 30.6 0.8 30.5 0.9 28.0 0.8 27.5 0.8 26.7 0.7 26.0 0.7 26.0 0.7 24.1 0.6 <0.001 Rural 31.2 0.6 30.7 0.6 29.3 0.5 29.5 0.6 28.3 0.6 26.8 0.5 28.2 0.6 27.0 0.5 <0.001 Men Age (years) 15–24 38.0 1.5 38.1 1.6 28.1 1.3 30.2 1.7 23.0 1.7 27.2 20.7 32.0 2.2 30.6 2.2 <0.001 25–44 71.8 1.0 68.5 1.1 64.6 1.0 63.7 1.1 61.6 1.2 57.9 1.2 58.8 1.3 55.2 1.2 <0.001 45–64 65.8 1.4 65.2 1.4 63.6 1.2 61.3 1.2 60.0 1.1 58.0 1.1 58.0 10.8 56.8 0.9 <0.001 ≥65 49.0 2.7 51.5 2.6 39.4 2.2 42.3 2.1 40.6 1.9 37.7 1.8 43.8 1.7 41.2 1.5 <0.001 Region Urban 59.5 1.3 58.7 1.4 52.5 1.2 52.7 1.2 50.4 1.2 50.2 1.2 51.0 1.2 48.3 1.0 <0.001 Rural 61.0 0.9 59.8 0.9 54.5 0.8 56.5 0.9 54.6 0.9 52.7 0.9 54.7 0.9 54.0 0.8 <0.001 Women Age (years) 15–24 0.4 0.2 0.8 0.3 0.5 0.2 0.3 0.2 0.2 0.2 – – 0.3 0.3 0.2 0.2 0.223 25–44 1.7 0.3 2.0 0.3 2.6 0.3 2.2 0.3 1.8 0.3 1.2 0.2 1.0 0.2 1.2 0.2 0.004 45–64 8.5 0.8 8.0 0.8 6.4 0.6 6.2 0.6 4.5 0.5 4.2 0.4 4.4 0.4 2.9 0.3 <0.001 ≥65 11.9 1.6 10.1 1.5 8.3 1.1 10.1 1.2 9.2 1.0 7.6 0.9 6.9 0.8 6.8 0.7 <0.001 Region Urban 4.7 0.5 4.7 0.6 4.5 0.5 4.5 0.5 4.6 0.5 3.6 0.4 3.2 0.4 2.6 0.3 <0.001 Rural 3.6 0.3 3.8 0.3 3.6 0.3 4.1 0.3 3.4 0.3 3.2 0.3 3.6 0.3 3.0 0.3 <0.001 SE: standard error of prevalence. a Adjusted for age, sex and region when appropriate. Table 2 Trends in mortality attributable to smoking in China, 1991–2011. Men Women Relative risk (95% CI)b Population attributable risk, % (95% CI) Relative risk (95% CI)b Population attributable risk, % (95% CI) 1991 2000 2011 1991 2000 2011 All causes 1.33 (1.28–1.39) 17.5 (15.3–20.0) 16.1 (14.0–18.5) 16.5 (14.4–19.0) 1.51 (1.40–1.63) 2.2 (1.8–2.8) 2.2 (1.8–2.8) 1.7 (3.5–17.8) Cancer All 1.51 (1.40–1.63) 24.7 (20.5–28.8) 22.9 (18.9–26.8) 23.4 (19.4–27.4) 1.58 (1.37–1.81) 2.5 (1.6–3.5) 2.5 (1.6–3.5) 1.9 (1.2–2.7) Lung 2.58 (2.17–3.05) 50.4 (42.9–56.9) 47.9 (40.5–54.4) 48.7 (41.2–55.2) 2.56 (2.02–3.26) 6.6 (4.4–9.2) 6.6 (4.4–9.2) 5.0 (3.4–7.1) Liver 1.26 (1.06–1.50) 14.3 (3.7–24.3) 13.1 (3.4–22.5) 13.5 (3.5–23.1) 1.40 (0.93–2.12) 1.8a 1.8a 1.3a Stomach 1.25 (1.05–1.49) 13.8 (3.1–24.0) 12.7 (2.8–22.2) 13.0 (2.9–22.7) 1.43 (0.92–2.23) 1.9a 1.9a 1.4a Oesophagus 1.58 (1.30–1.93) 27.2 (16.2–37.4) 25.2 (14.8–35.1) 25.8 (15.3–35.8) 1.05 (0.55–2.01) 0.2a 0.2a 0.2a Five minor sites 1.53 (1.15–2.04) 25.4 (8.8–40.1) 23.5 (8.0–37.7) 24.1 (8.3–38.4) 0.98 (0.55–1.75) a a a Cardiovascular disease All 1.24 (1.16–1.33) 13.4 (9.3–17.5) 12.2 (8.5–16.1) 12.6 (8.8–16.5) 1.44 (1.27–1.63) 1.9 (1.2–2.8) 1.9 (1.2–2.8) 1.5 (0.9–2.1) IHD 1.38 (1.23–1.54) 19.6 (12.9–25.8) 18.1 (11.8–23.9) 18.6 (12.1–24.5) 1.74 (1.42–2.12) 3.2 (1.9–4.8) 3.2 (1.9–4.8) 2.5 (1.4–3.7) Ischaemic stroke 1.41 (1.15–1.73) 20.9 (8.8–31.9) 19.2 (8.0–29.8) 19.7 (8.3–30.5) 1.15 (0.74–1.78) 0.7a 0.7a 0.5a Haemorrhagic stroke 1.03 (0.92–1.16) 1.9a 1.7a 1.8a 1.09 (0.86–1.39) 0.4a 0.4a 0.3a All respiratory 1.64 (1.42–1.90) 29.2 (21.3–36.7) 27.1 (19.6–34.3) 27.7 (20.1–35.1) 1.78 (1.46–2.17) 3.4 (2.0–5.0) 3.4 (2.0–5.0) 2.6 (1.5–3.8) All other medical diseases 1.20 (1.06–1.36) 11.4 (3.7–18.8) 10.4 (3.4–17.3) 10.7 (3.5–17.8) 1.41 (1.13–1.76) 1.8 (0.6–3.3) 1.8 (0.6–3.3) 1.4 (0.4–2.5) CI, confidence interval. a The PAR or 95% CI cannot be calculated. b Extracted from the publication by Chen et al. (2015). Contributors BX was involved in the study conception and design. SL, LM, AC, and CM provided statistical expertise. SL, LM and BX drafted the article; BX and AC revised it critically for important intellectual content. All authors approved the final version. BX is the guarantor. Declaration of interests None. Asma S Song Y Cohen J Eriksen M Pechacek T Cohen N Iskander J Centers for Disease, C., Prevention 2014 CDC grand rounds: global tobacco control MMWR Morb Mortal Wkly Rep 63 277 280 24699763 Bi Y Jiang Y He J Xu Y Wang L Xu M Zhang M Li Y Wang T 2015 Status of cardiovascular health in Chinese adults J Am Coll Cardiol 65 1013 1025 25766949 Chen ZM Xu Z Collins R Li WX Peto R 1997 Early health effects of the emerging tobacco epidemic in China. A 16-year prospective study JAMA 278 1500 1504 9363969 Chen Z Shin YS Beaglehole R 2012 Tobacco control in China: small steps towards a giant leap Lancet 379 779 780 22386013 Chen Z Peto R Zhou M Iona A Smith M Yang L Guo Y Chen Y Bian Z 2015 Contrasting male and female trends in tobacco-attributed mortality in China: evidence from successive nationwide prospective cohort studies Lancet 386 1447 1456 26466050 Ezzati M Lopez AD 2003 Estimates of global mortality attributable to smoking in 2000 Lancet 362 847 852 13678970 Giovino GA Mirza SA Samet JM Gupta PC Jarvis MJ Bhala N Peto R Zatonski W Hsia J 2012 Tobacco use in 3 billion individuals from 16 countries: an analysis of nationally representative cross-sectional household surveys Lancet 380 668 679 22901888 Gu D Kelly TN Wu X Chen J Samet JM Huang JF Zhu M Chen JC Chen CS 2009 Mortality attributable to smoking in China J Med]–>N Engl J Med 360 150 159 Jena PK Kishore J Bandyopadhyay C 2012 Prevalence and patterns of tobacco use in Asia Lancet 380 1906 (author reply 1906–1907) 23200498 Kaul A Wolf M 2014 Standardised packaging and tobacco-industry-funded research Lancet 384 233 234 Kim YH 2005 Korean adolescents’ smoking behavior and its correlation with psychological variables Addict Behav 30 343 350 15621405 Koplan JP An WK Lam RM 2010 Hong Kong: a model of successful tobacco control in China Lancet 375 1330 1331 20347129 Lam TH He Y Li LS Li LS He SF Liang BQ 1997 Mortality attributable to cigarette smoking in China JAMA 278 1505 1508 9363970 Levy D Rodriguez-Buno RL Hu TW Moran AE 2014 The potential effects of tobacco control in China: projections from the China SimSmoke simulation model BMJ 348 g1134 24550245 Li Q Hsia J Yang G 2011 Prevalence of smoking in China in 2010 J Med]–>N Engl J Med 364 2469 2470 Mathers CD Loncar D 2006 Projections of global mortality and burden of disease from 2002 to 2030 PLoS Med 3 e442 17132052 Ng M Freeman MK Fleming TD Robinson M Dwyer-Lindgren L Thomson B Wollum A Sanman E Wulf S 2014 Smoking prevalence and cigarette consumption in 187 countries, 1980–2012 JAMA 311 183 192 24399557 No authors listed 2011 China’s unhealthy relations with big tobacco Lancet 377 180 No authors listed 2014a Cigarette packaging in China–not going far enough Lancet 383 1360 No authors listed 2014b A step change for tobacco control in China? Lancet 384 2000 Popkin BM Du S Zhai F Zhang B 2010 Cohort profile: the China Health and Nutrition Survey—monitoring and understanding socio-economic and health change in China, 1989–2011 J Epidemiol]–>Int J Epidemiol 39 1435 1440 Qian J Cai M Gao J Tang S Xu L Critchley JA 2010 Trends in smoking and quitting in China from 1993 to 2003: National Health Service Survey data Bull World Health Organ 88 769 776 20931062 Rich ZC Xiao S 2012 Tobacco as a social currency: cigarette gifting and sharing in China Nicotine Tob Res 14 258 263 21849412 Thun MJ Carter BD Feskanich D Freedman ND Prentice R Lopez AD Hartge P Gapstur SM 2013 50-year trends in smoking-related mortality in the United States J Med]–>N Engl J Med 368 351 364 WHO 2015 Global health risks: mortality and burden of disease attributable to selected major risks http://www.who.int/healthinfo/global_burden_disease/GlobalHealthRisks_report_full.pdf Yang G Kong L Zhao W Wan X Zhai Y Chen LC Koplan JP 2008 Emergence of chronic non-communicable diseases in China Lancet 372 1697 1705 18930526 Yang L Sung HY Mao Z Hu TW Rao K 2011 Economic costs attributable to smoking in China: update and an 8-year comparison, 2000–2008 Tob Control 20 266 272 21339491 Yang G Wang Y Zeng Y Gao GF Liang X Zhou M Wan X Yu S Jiang Y 2013 Rapid health transition in China, 1990–2010: findings from the global burden of disease study 2010 Lancet 381 1987 2015 23746901 Yang G Wang Y Wu Y Yang J Wan X 2015 The road to effective tobacco control in China Lancet 385 1019 1028 25784349
PMC005xxxxxx/PMC5125023.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101128834 31537 Integr Cancer Ther Integr Cancer Ther Integrative cancer therapies 1534-7354 1552-695X 27151592 5125023 10.1177/1534735416636222 NIHMS828457 Article A Systematic Review of Spiritually Based Interventions and Psychoneuroimmunological Outcomes in Breast Cancer Survivorship Hulett Jennifer M. PhD, APRN, FNP-BC, PPCNP-BC 1 Armer Jane M. PhD, RN, FAAN 2 1 University of Utah, Salt Lake City, UT, USA 2 University of Missouri, Columbia, MO, USA Corresponding Author: Jennifer M. Hulett, College of Nursing, University of Utah, 10 South 2000 East, Salt Lake City, UT 84112, USA. [email protected] 9 11 2016 4 5 2016 12 2016 01 12 2016 15 4 405423 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Objective This is a review of spiritually based interventions (eg, mindfulness-based stress reduction) that utilized psychoneuroimmunological (PNI) outcome measures in breast cancer survivors. Specifically, this review sought to examine the evidence regarding relationships between spiritually based interventions, psychosocial-spiritual outcomes, and biomarker outcomes in breast cancer survivors. Methods A systematic search of 9 online databases was conducted for articles of original research, peer-reviewed, randomized and nonrandomized control trials from 2005–2015. Data were extracted in order to answer selected questions regarding relationships between psychosocial-spiritual and physiological measures utilized in spiritually based interventions. Implications for future spiritually based interventions in breast cancer survivorship are discussed. Results Twenty-two articles were reviewed. Cortisol was the most common PNI biomarker outcome studied. Compared with control groups, intervention groups demonstrated positive mental health outcomes and improved or stable neuroendocrine-immune profiles, although limitations exist. Design methods have improved with regard to increased use of comparison groups compared with previous reviews. There are few spiritually based interventions that specifically measure religious or spiritual constructs. Similarly, there are few existing studies that utilize standardized religious or spiritual measures with PNI outcome measures. Findings suggest that a body of knowledge now exists in support of interventions with mindfulness-breathing-stretching components; furthermore, these interventions appear to offer potential improvement or stabilization of neuroendocrine-immune activity in breast cancer survivors compared to control groups. Conclusion From a PNI perspective, future spiritually based interventions should include standardized measures of religiousness and spirituality in order to understand relationships between and among religiousness, spirituality, and neuroendocrine-immune outcomes. Future research should now focus on determining the minimum dose and duration needed to improve or stabilize neuroendocrine-immune function, as well as diverse setting needs, including home-based practice for survivors who are too ill to travel to group sessions or lack economic resources. breast cancer interventions oncology psychoneuroimmunology religious review spirituality Introduction Despite the vast amount of research that has been conducted in breast cancer, the extended survivorship trajectory for breast cancer survivors remains unpredictable and complicated by breast cancer recurrence or treatment-related physical effects (ie, secondary lymphedema).1 The incidence of secondary lymphedema in breast cancer survivors ranges from 6% to 94% and is associated with a lifetime risk.2 Because of these posttreatment late effects, as many as 30% of breast cancer survivors have reported physical function decline, poorer mental health, social difficulties, and a poorer quality of life.3–5 When considering interventions appropriate for symptom management in breast cancer survivorship, it is recognized that breast cancer survivors are now an aging population with multifaceted problems beyond physiological symptoms, including economic, psychosocial, and spiritual stress.6–8 According to Pew Forum Research data, 92% of Americans believe in God or a higher power and 56% report religion is very important in their lives.9 Previous research has shown that older adults rely on religious practices (ie, prayer) while seeking treatment for chronic symptoms.10 To cope with chronic physical and psychological symptoms, studies have shown that 80% to 90% of breast cancer survivors report using spiritually based complementary and alternative medicine (CAM) therapies (eg, mindfulness, meditation, yoga, Tai Chi, Qigong, guided imagery, and affirmations) to manage long-term breast cancer treatment–related symptoms.11–15 Religious coping has shown an association with spiritual growth, better mental health, and positive outcomes following stressful life events.16,17 Studies of chronically ill populations, including breast cancer survivors, have reported that religious and spiritual factors (eg, relationship with God/higher power, prayer, and congregational/social support) are used for comfort and coping with cancer survivorship experiences.10,18,19 In the past 10 years, a number of CAM interventions have been conducted in breast cancer populations to examine changes in psychological and physiological outcomes in those receiving the interventions.20,21 Although many of these CAM interventions (eg, mindfulness-based interventions and Qigong) had a spiritual component, the relationships between the spiritual basis of these interventions and health outcomes have remained unclear. Moreover, previous data suggest that religious and spiritual variables influence mental health outcomes; however, a gap exists in the psychoneuroimmunological (PNI) research regarding the evidence linking religious and spiritual variables and physiological health outcomes. The purpose of this article is to review PNI-based interventions for health outcomes associated with religious and spiritually based interventions in breast cancer survivors. Despite numerous CAM interventions reported in the breast cancer literature, there remain few reported findings regarding the relationships between religious and spiritual variables and neuroimmune function changes associated with these religious and spiritual interventions. This article is divided into 2 sections: (1) An overview of the relationships between and among religious and spiritual variables related to health outcomes and (2) a systematic review of spiritually based interventions with PNI health outcomes in breast cancer survivors. Literature Review Religious and Spiritual Variables Related to Health Many terms have been used to describe spirituality for health research purposes including meaning, purpose in life, the mystical, the numinous, hope, value, optimism, emotional connectedness, transcendence, gratitude, and forgiveness.22–24 Religious and spirituality researchers have conceptualized spirituality based on their own cultural and philosophical traditions that may have either a religious or a secular basis.25,26 Consequently, the constructs of religiousness and spirituality have been difficult to clarify for health research purposes due to the lack of agreement among contemporary researchers who have viewed the 2 constructs as complex with overlapping aspects (i.e. searching for the sacred).27–29 In general, religion refers to denomination affiliation, religious identity, public religious practice (eg, attendance and group prayer) and specific beliefs in religious tenets (eg, afterlife).22,30 Religious practices are culturally based practices such as prayer, church attendance, meditation, or reading religious texts. Spirituality is viewed as a subjective experience of the sacred and refers to an emotional connectedness or relationship with God or the transcendent beyond the self.31–33 Additionally, recent data suggest that spirituality may be a dimension of personality.23,24 Self-transcendence refers to the ability to stand outside one’s immediate sense of time and place to view life from a larger perspective.24,34 Spiritual self-transcendence is further characterized by recognition that a synchronicity to life exists and fosters a sense of commitment to helping others.35 Three common themes have been represented in the literature regarding spirituality: a relationship with the transcendent and to others, the existence of a higher being, and an appreciation for the greater world.36 Although it has been generally agreed that religion and spirituality are separate constructs, the literature discusses these two constructs as interchangeable; furthermore, religious and spiritual variables have often been denoted as “RS.” Extending from this trend, religious and spirituality research now reflects measures with religious and spiritual construct overlap. It is recognized that there remains a lack of a “gold standard” definition providing separation of religious and spiritual constructs for health care research purposes.25,37 Therefore, for purposes of this discussion and where appropriate, religious and spiritual variables will hereafter be denoted as “RS.” Spiritual Experiences, Religious Practices, and Congregational Support Findings from a recent factor analysis of religious measures, spiritual measures, and health outcomes suggested that 3 dimensions represent the religious and spiritual variables related to health outcomes: spiritual experiences (ie, emotional connectedness to the transcendent), religious practices (ie, behaviors), and congregational support (ie, social support).27 Moreover, religious and spiritual dimensions are viewed as separate constructs that can be distinguished as positive and negative (eg, loving God/higher power vs a punishing God/higher power and positive/negative congregational support).27,38,39 Positive spiritual experiences are associated with better physical health in individuals with chronic disabilities, and negative spiritual experiences are associated with worse health. Among religious and spiritual variables, forgiveness appears to be a greater predictor of health outcomes.23 RS Variables and Health Outcomes in Breast Cancer Observation of breast cancer survivors (n = 763) over 10 years demonstrated that although a diagnosis of cancer was associated with a sense of vulnerability, there was a positive change in survivor participants’ worldview and perceptions of life’s meaning that persisted even 10 years postdiagnosis.40 A growing consensus in the literature has suggested that the presence of a spiritual dimension is an indicator of positive adaptation to cancer treatment and coping with cancer.41,42 Data have also shown that social isolation among women with newly diagnosed breast cancer was associated with higher mortality risk.43 Breast cancer survivors who reported positive perceptions of social support (eg, emotional reassurance, personal assistance, and advice) also demonstrated positive immune system benefits and decreased psychological distress.44,45 Data from breast cancer populations have suggested that increased stress correlates with decreased immune function (ie, natural killer [NK] cell activity and T cell response), resulting in a decreased ability to destroy cancerous tumors.46 Data have also shown that emotional distress experienced by breast cancer survivors correlated with impaired NK cell and cytokine function and further suggested that emotional distress negatively influenced the immune profile of breast cancer survivor participants.47 Pilot data from mindfulness-based interventions have shown that mindfulness exerted a positive influence on NK cell cytotoxicity, cytokine function,48–50 and cortisol patterns.49,51–54 These relationships between psychosocial and RS (eg, mindfulness) variables and health outcomes in populations with immune dysregulation (ie, breast cancer) are consistent with a psychoneuroimmunological (PNI) model of health. The PNI Model of Health The PNI model of health seeks to explain causal relationships between and among stress, brain function (ie, mind/thoughts), psycho-social-behavioral components (eg, spirituality), and physiological components (ie, neuroendocrine-immune system interactions).55–57 As stress influences perceptions, the resulting thought processes are then communicated from the brain to the immune system via neuroendocrine and hormonal pathways. The subsequent adaptive immune system response then manifests as a psychological or physiological symptom (ie, mental or physical illness) or health maintenance (ie, wellness).58,59 Psychoneuroimmunological research has established that neuroendocrine-immune function can be studied through measures of biomarker levels obtained in saliva and serum specimens. The 2 primary neuroendocrine-immune pathways associated with stress are the sympatho-adrenomedullary (SAM) axis, which includes the sympathetic nervous system (SNS); and the hypothalamic-pituitary-adrenocortical axis (HPAA).60 Chronic stress results in dysregulation of both axes.61 During acute stress the SAM axis activates the SNS to release norepinephrine which induces a “fight or flight” response.62 The HPAA responds by releasing endocrine-based glucocorticoids, primarily in the form of cortisol hormone.63 Repeated activation of the HPAA system results in increased allostatic load and has been shown to cause chronic immunosuppression associated with negative health outcomes.61,64,65 In chronically ill populations, suppressed immune function associated with long-term stress is further associated with increased susceptibility to illness, delayed wound healing, and prolonged recovery from illness.66 After prolonged HPAA vigilance, elevated cortisol activity becomes detrimental, as it becomes immunosuppressive and ultimately contributes to persistent immune dysregulation.58,67 Mindfulness-Based Interventions and Moving Meditations as Spiritually Based Interventions Among the most commonly reported psychosocial interventions in breast cancer research literature, mindfulness-based interventions (eg, mindfulness-based stress reduction [MBSR] and mindfulness-based cognitive therapies [MBCT]) have become prolific as adjunct treatments for post–cancer treatment–related symptoms.68,69 Traditional forms of MBSR stem from the contemplative, spiritually based Buddhist philosophy, which promotes the development of a nonjudgmental, accepting, and patient worldview and teaches relaxation through focused awareness on breathing.48 MBSR and MBCT interventions use meditation and gentle Yoga stretching to “maintain awareness moment by moment, disengaging oneself from strong attachment to beliefs, thoughts or emotions, and thereby developing a greater sense of emotional balance or well-being.”70(p1350) Similarly, there are other movement-meditative-breathing interventions (eg, Qigong and Tai Chi) that include encourage mindfulness and self-awareness. Since spiritual self-transcendence involves viewing life from a larger perspective beyond the self, mindfulness and meditation practices facilitate spiritual change.24 From this perspective, interventions that encourage mindfulness, meditation, and self-awareness through breath work and movement are viewed as facilitating spiritual transformation. Relaxation and Visualization Therapy Relaxation and visualization therapy (RVT) involves the induction of a relaxation using mental imagery of a desired object or outcome and includes progressive muscle relaxation, continued guided imagery, meditation, and deep breathing.71 RVT involves imagery of peaceful scenery and focused sensory awareness on individual muscle groups. RVT has been reported to improve quality of life, mood, reduced social conformity, and enhanced emotional expression.72,73 Guided Imagery Similar to RVT, guided imagery utilizes progressive relaxation of muscle groups combined with focused breathing techniques intended to calm the mind and prepare it for guided imagery.74 Guided imagery is a consciousness-focused practice intended to increase awareness with a relaxed, open mind for the purpose of confronting a specific concern or issue. Guided imagery is viewed as a way to allow patients to participate in their own healing and has been reportedly used for healing, symptom management, promotion of positive health behaviors, and making positive life changes. Guided imagery also includes aspects involving the “unconscious mind” including intuition, emotions, feelings, memories, values, beliefs, perceptions, and goals.75 Tai Chi In Western culture, Tai Chi is a moving meditation or a form of low-level exercise, which involves a sequence of fluid, graceful movements, focused breathing, posturing movement, and consciousness directed at relaxation.76,77 However, for some traditions, Tai Chi is a Chinese martial art, a Shamanic religious ritual, an exercise, and a relaxation technique that has been in existence for at least 5000 years.76 The practice of Tai Chi is meant to increase the participant’s mind-body connection through the awareness of the body’s energy and potential for self-healing, resulting in self-empowerment. The meanings of the Tai Chi movements and philosophy are conveyed using the 5 elements (ie, fire, earth, minerals, water, and wood) and seasons of the year to represent (spiritual) transformation.77 Yoga Yoga is an ancient Eastern traditional mind-body practice.78 Yoga is also considered a moving meditation and consists of breathing exercises, postures, relaxation, and meditation.79 Yoga encourages increased self-awareness and relaxation and has been observed to alter the stress response associated with thoughts and emotions, subsequently reducing psychological distress.79 Qigong Qigong, also a 5000 year-old Chinese mind-body tradition, conceptualizes health as the result of unimpeded flow of “qi” (ie, energy, spirit or life force) through “gong” (ie, achievement).80,81 The focus of Qigong is the prevention and healing of diseases by harmonizing the mind, body, and spirit.82 Specifically, 4 main components are involved in the stimulation and manipulation of the qi: consciousness, mindful focus on the body, breathing techniques, and specific movements.83 Unlike other moving meditations, Qigong can be practiced internally or externally. In external Qigong practice, a Qigong master may facilitate the clearing of qi blockages or balancing of qi for an individual.83 Given the preceding definitions, mindfulness-based interventions, RVT, guided imagery, Tai Chi, yoga, and Qigong all suggest common themes of mindfulness, self-awareness, and self-transcendence (ie, viewing life from a larger perspective beyond the self).24 Therefore, because these practices involve themes of spirituality and spiritual transformation, they were included in this review and are referred to as spiritually based interventions. Summary of Previous Spiritually Based Interventions and Health Outcomes MBSR Interventions and Health Outcomes Early pilot studies demonstrated that MBSR improves mindfulness,84–86 stress,50,87 mood, and psychological distress.49,84,88 MBSR has been shown to decrease depression and anxiety,89–91 and rumination among breast cancer survivors.92,93 Data have shown that MBSR improved sleep quality,54,85 reduced fatigue,21,54 improved quality of life,52,92 and improved perceptions of life being more meaningful.94 Although 2 studies have reported MBSR improves perceptions of spiritual growth in breast cancer survivors,94,95 few studies have examined relationships between mindfulness-based practices and RS constructs (ie, positive and negative spiritual experiences, congregational support, and religious practices). A previous meta-analysis and review revealed that the majority of early MBSR studies utilized exploratory, pilot designs, with small samples, varied intervention design, and lacked information regarding intervention therapy protocols.96,97 A more recent meta-analysis of MBSR studies (n = 9) with psychological measures in breast cancer survivors reported that MBSR interventions had a moderate to large effect (d = 0.76) on mental health outcomes.98 In a narrative review of MBSR interventions (n = 43) with PNI measures in heterogeneous cancer populations, Carlson69 noted that strong level 1 evidence offered support that MBSR improved anxiety, quality of life, depression, stress, spiritual growth, and well-being in cancer populations. Moreover, quantitative findings suggested MBSR improved PNI biomarker levels in cancer survivors (ie, decreased inflammatory cytokines and cortisol levels), but relationships remain unclear and the overall evidence was weak.69 Subnis et al99 conducted a systematic review of randomized control trials (RCTs), nonrandomized control trials (non-RCTs), and pretest/posttest designs (n = 24) among heterogeneous cancer populations who received cognitive-behavioral interventions or spiritually based interventions (eg, MBSR, yoga) and the associated PNI-based outcome measures. Analyses revealed that most of the psychosocial measures were measures of negative psychological states, with mood as the most common psychosocial measure of interest (11 of 24 studies) and cortisol as the most common biomarker of interest (11 of 24 studies). Moreover, effectiveness of interventions on PNI outcomes among cancer patients could not be determined due to lack of longitudinal designs, small sample sizes, wide variation in intervention protocols, and lack of adherence data. Finally, the review by Subnis et al99 was one of the first to address the lack of biomarker measures representing the SAM (ie, SNS) response to stress. The majority of PNI studies to date have investigated only the HPAA by using cortisol as a proxy measure of chronic stress. Qigong and Tai Chi Interventions In a systematic review of 23 studies (RCT and non-RCT), Chan et al100 reported that Qigong had a statistically significant effect on quality of life, symptom improvement, improved fatigue, improved well-being, and improved immune function. However, the overall quality of studies was ranked as poor in methodology, with 3 studies rated “A” and 9 studies rated as “B” (based on Oxford Centre for Evidence-Based Medicine’s levels of evidence).100,101 Moreover, half of the studies were not peer-reviewed and had numerous methodological flaws, including small sample sizes, lack of randomization designs, and disparities in intervention protocols (eg, duration, dose).100 In a meta-analysis by Zeng et al,102 Qigong (n = 5) and Tai Chi (n = 8) intervention data showed that Qigong did not have a statistically significant effect on depression or anxiety; however, it was associated with improved fatigue (P = .04) and improved quality of life (P = .008). Additionally, both Qigong and Tai Chi intervention data demonstrated reduction of cortisol levels (P < .05) and offered limited, but supportive, data suggesting positive associations with improved C-reactive protein and cytokine function. Zeng et al102 also found limitations in intervention designs and methodologies, as well as limited ability to pool data from multiple studies due to lack of heterogeneity between trials. Specific RS Interventions In a meta-analysis of RCTs and non-RCTs, Oh and Kim103 examined interventions (n = 14), including protocol descriptions of religious interventions, spiritual nursing care, spiritual counseling, spiritually focused meditation, and meaning-centered psychotherapy, in heterogeneous cancer patients (n = 889). Analyses of RS intervention data revealed significant effects on spiritual well-being (d = −0.48), meaning of life (d = −0.58), depression (d = −0.62), and anxiety (d = −0.82), as well as supported associations between RS interventions and improved mental health outcomes. However, findings and conclusions were limited by a lack of physiological measures, overlap of religious and spiritual constructs, heterogeneous samples, and varied intervention protocols.103 Gaps in the Literature Breast cancer survivors report perceived therapeutic benefits from using mindfulness-based (ie, spiritual) practices (ie, MBSR and yoga) to cope with posttreatment sequela, although there has been limited empirical evidence to support how spiritual practices influence physiological outcomes. A major gap in current PNI research is the limitations regarding the evidence linking objective measures of neuroendocrine-immune function (eg, biomarkers) to self-report measures of stress and RS variables (eg, religious and spiritual beliefs). The purpose of this review is to address the following questions: What psychosocial and RS measures have been used in spiritually based interventions in breast cancer survivors; and what were the associated mental health outcomes? What PNI outcome measures have been used in spiritually based interventions in breast cancer survivors, and what were the associated physiological biomarker outcomes? What is the evidence regarding the relationships between and among spiritually based interventions, psychosocial-spiritual outcomes, and PNI-based outcomes in breast cancer survivors? What are the implications for future research studies regarding spiritually based interventions, measures of RS, and PNI-based outcome measures? Methods Articles from 2005 through 2015 were retrieved from the following databases: PubMed, Medline, PsycInfo, PsycArticles, Cumulative Index to Nursing and Allied Health Literature (CINAHL), Academic Search Complete, American Theological Library Association (ATLA) Religion, Educational Resources Information Center (ERIC), and Google Scholar. Additionally, a hand search of peer-reviewed journals with a PNI focus (eg, Brain, Behavior, and Immunity [BBI], Psychoneuroendocrinology, and Psycho-Oncology), an ancestry search of references from selected articles, and a personal library search were done. Keywords The keywords search was developed from 6 main categories with related subcategories noted in parentheses: (1) spirituality (spiritual, spirituality, spiritual beliefs, spiritual practices), (2) religious (religious, religion, religiosity, religiousness, religious beliefs, religious practices), (3) cancer (neoplasm, oncology, breast, lymphedema), (4) stress, (5) intervention (spiritual, psychosocial, spiritual healing, spiritual therapies, faith healing, CAM, psychological therapy/therapies, spiritual coping, mind-body-therapies, Qigong, group therapy, mindfulness, and (6) PNI-based measures (biomarkers, immune function, immune system, leukocyte, lymphocyte, cytokines, natural killer cell, interferon-gamma, interleukin, tumor necrosis factor-alpha, cortisol, salivary alpha-amylase, neuroendocrine, neuroimmune, hormonal, and inflammatory/inflammation). Inclusion Criteria The inclusion criteria included the following: original research in peer-reviewed journals, full-text available online, clearly stated descriptions of samples and methodology, randomized control trials (RCTs), nonrandomized control trials (non-RCTs), human subjects, adults, and articles available in English. The initial search yielded 268 articles. Articles were further screened in several stages. First, titles of articles and abstracts were evaluated based on the inclusion criteria, which resulted in the selection of 158 articles. Next, selected articles were further evaluated by reading the full-text which resulted in 107 articles. Finally, articles were excluded for the following reasons: lack of PNI-based measures (n = 32), lack of RS-based PNI interventions (n = 20), poorly defined or highly customized psychosocial interventions (ie, difficult to replicate based on information provided (n = 10), lack of standardized psychosocial measures (ie, subjective measures) (n = 7), small sample size (n < 10) (n = 3), healthy participant sample (n = 2), and lack of a comparison group (n = 4). Article Selection Figure 1 provides a flowchart of the identification, screening, and article selection process, which resulted in 22 articles original articles with RS interventions in breast cancer survivors. Article data selected for extraction and summary included (1) first author and year of publication, (2) study design, (3) cancer stage, (4) sample size of intervention and control groups, (5) type and description of interventions, (6) psychosocial measure data, (7) RS measure data, (8) PNI measures, and (9) conclusion of findings. Results and Discussion Studies reviewed included RCTs (n = 19) and non-RCTs (n = 3). The majority of studies were from the United States (n = 12), followed by Canada (n = 3), India (n = 2), China (n = 1), Brazil (n = 1), Taiwan (n = 1), Sweden (n = 1), and the United Kingdom (n = 1). Participant sample sizes in the research studies ranged from 28 to 271 participants. RS interventions were a majority MBSR (n = 8), followed by yoga (n = 5), cognitive based stress management (CBSM) (n = 2), guided imagery (n = 2), relaxation visualization therapy (RVT) (n = 1), mindfulness-based cancer recovery (MBCR) (n = 1), Qigong (n = 1), Tai Chi/spiritual growth group (n = 1), and body-mind-spirit (n = 1). Interventions were described as lasting from 5 to 37 weeks in duration. The majority of research designs used a single intervention group with a wait list or usual care control group (n = 13). All studies utilized physiological measures, with cortisol being the most common biomarker measure (n = 12; 54.5%). Question 1: What psychosocial and RS measures have been used in spiritually based interventions in breast cancer survivors; and what were the associated health outcomes? The majority of studies (n = 20) included standardized, self-report psychosocial measures, with an average of 3 to 4 psychosocial instruments administered per study. The most common self-report psychosocial measures included depression (n = 8), stress (n = 7), quality of life (n = 7), anxiety (n = 6), fatigue (n = 5), mindfulness (n = 4), and mood (n = 2). Importantly, only 1 study examined specific RS measures: spiritual growth (ie, meaning of life).104 Because many of the studies reviewed have psychosocial (subjective) and PNI (objective) outcomes, the findings are separated and presented in 2 tables to illustrate the relationships between subjective and objective variables. Table 1 provides a brief summary of the statistically significant psychosocial-spiritual findings associated with each spiritually based intervention, while Table 2 focuses on the respective PNI outcomes associated with each spiritually based intervention. Both Tables 1 and 2 utilize upward- or downward-pointing arrows, respectively, to illustrate the reported direct or inverse relationships between the variables. Additionally, the number of studies in this review are noted in Tables 1 and 2 by their corresponding order of presentation in Table 3 where highlights of the reviewed studies are provided. The majority of spiritually based intervention studies demonstrated positive changes in one or more of the psychosocial-spiritual outcome measures examined. The most common positive psychosocial outcomes associated with spiritually based interventions were observed in measures of quality of life, depression, stress, anxiety, fatigue, and mood. The exception was a study by Robins et al77 in which measures of stress at study conclusion demonstrated an overall decrease in perceived stress among Tai Chi, spiritual intervention, and control groups compared with preintervention measures, however, the Tai Chi group continued to report more perceived stress compared to those in the spiritual intervention group and control group at postintervention. Of note, all participants, including the control group, were receiving adjuvant chemotherapy. Robins et al77 suggested that greater perceived stress among the Tai Chi participants may have indicated psychological distress that occurred due to increased self-awareness or “centering” secondary to the interventions. Moreover, the authors noted that although Tai Chi encourages mindfulness, an unintended initial effect of increased mindfulness may be temporary depression-like symptoms (ie, psychological distress) that occur during stressful life events (eg, post–cancer chemotherapy treatment sequela). Additionally, Robins et al77 noted that psychological distress may have been secondary to chemotherapy sequela (eg, fatigue) and/or secondary to learning Tai Chi movement; however, these potential confounders were not measured. Finally, 2 studies examined relationships between mindfulness-based interventions, psychosocial measures (eg, mood and stress), and telomere variables; however, the psychosocial relationships very small and were not statistically significant.117,118 Question 2: What PNI outcome measures have been used in spiritually based interventions in breast cancer survivors; and what were the associated physiological biomarker outcomes? Table 2 provides an overview of the spiritually based intervention studies with PNI-based biomarker outcomes. In this review, cortisol is the most commonly studied biomarker outcome studied in the spiritually based interventions. An interesting finding was that among the cortisol studies reviewed, 8 studies reported positive changes in cortisol activity (ie, decreased levels or healthier diurnal slope patterns compared to control groups), while 4 studies reported “no change” or “stable” cortisol levels compared with control groups. These can be somewhat mixed findings to interpret, as “no change” in cortisol levels is occasionally noted in studies in which elevated levels of cortisol are observed in the control group. These findings suggested that in some instances, “no change” or “stable” cortisol levels after a spiritually based intervention may reflect a positive buffering effect associated with the intervention. Additionally, some cortisol studies reported significant changes immediately postintervention, but not sustained at 1-month,109 3-month,110 and 6-month follow-up.111 Two biomarkers of emerging PNI interest, telomere length and telomere activity are noteworthy for this discussion. Recent data suggest that shorter telomere length is associated with poorer outcomes in chronic lymphocytic leukemia populations119 and may be predictive or poor prognosis in cancer patients.120 In this review, Carlson et al118 reported that MBSR appeared to preserve telomere length in breast cancer survivors in the intervention group, while telomere length in the control group was shortened. Interestingly, Lengacher et al117 did not report preserved telomere length among MBSR breast cancer survivor participants; instead, their study demonstrated that those receiving the MBSR intervention had increased telomerase activity. Of note, duration of these MBSR interventions ranged from 6 to 8 weeks. These studies offer significant direction to PNI research as both provide early data regarding the potential relationship between mindfulness-based interventions and telomere length/telomere activity in breast cancer populations. Although more empirical data are required, these studies are some of the first studies to report observations between spiritually based practices and physiological outcomes. Overall, a wide variety of neuroendocrine-immune biomarker measures represented in the studies were observed in this review. Positive and negative associations between the spiritually based interventions and PNI outcome variables for the intervention studies were examined and these relationships are depicted in Table 2 with upward- or downward-pointing arrows. Moreover, the studies reporting these associations in Table 2 are linked with their corresponding study number as presented in Table 3. Among studies reviewed, MBSR interventions have been the most utilized spiritually based interventions in breast cancer survivors. MBSR is associated with improved inflammatory cytokine activity, improved lymphocyte function, improved or stabilization of cortisol levels, and increased or preservation of telomere activity. Moreover, interventions similar to MBSR, including CBSM, MBCR, and yoga all demonstrated similar influences on cytokine function, lymphocyte production, and improved or stabilized cortisol activity. While the remaining interventions, including Qigong, body-mind-spirit, and relaxation visualization therapy showed preservation or improved cortisol function in breast cancer survivors, there are few of these studies to date and conclusions are limited. Findings suggest that in general, spiritually based interventions are associated with improved neuroendocrine-immune function, particularly cortisol and cytokine activity. Question 3: What is the evidence regarding the relationships between spiritually based interventions, psychosocial-spiritual outcomes, and PNI-based outcomes in breast cancer survivors? Details of each reviewed intervention study are provided in Table 3, including designs, sample sizes, intervention and control group descriptions, psychosocial-spiritual measures, PNI measures, and the statistical significance of the outcome variables. The comparison of findings is challenging due to the wide variation in intervention duration and the wide range of psychosocial measures, as well as variation in PNI measures outcomes. However, there is a growing body of evidence to suggest a positive pattern is emerging between spiritually based interventions and physiological health outcomes. The majority of studies in this review report positive psychosocial and mindfulness-spiritual outcomes, as well as positive biomarker outcomes across differing spiritually based interventions. While the immune system response to spiritually based interventions during chemotherapy and/or radiotherapy was not an initial question for the review, it was observed that five spiritually based intervention studies were conducted while participants were concurrently undergoing chemotherapy or radiotherapy for breast cancer treatment. These interventions included RVT, yoga, Qigong, Tai Chi/spiritual growth, and yoga/stretching. It was observed that yoga interventions report a limited, but positive, trend on immune function during chemotherapy and radiotherapy treatments.108,115 While comparison is limited due to differing designs and treatments, both yoga interventions utilized the same intervention dose and duration (6 weeks, 180 min/wk) and utilized cortisol measures in breast cancer participants with stage III or lower. Question 4: What are the implications for future research studies regarding spiritually based interventions, measures of RS, and PNI outcome measures? Although interventions with a spiritual basis were the primary focus of this review, only one breast cancer study reported utilization of actual RS measures. Hsiao et al104 reported an association between an 8-week body-mind-spirit intervention, spiritual growth, and healthier cortisol patterns among breast cancer participants, compared with the control group (P < .05). While there were no other breast cancer body-mind-spirit intervention studies for comparison in this review, a similar study in patients with chronic depression and anxiety reported no association between prayers and cortisol levels.121 These findings contrast with those reported by Bormann et al,122 who did find an association between prayer and decreased cortisol levels in HIV-positive individuals. There are few studies examining specific RS measures and associated PNI biomarker outcomes in breast cancer survivorship; therefore, the nature of these relationships remain unclear. Table 3 provides the differences in the spiritually based interventions, including duration and dosing protocols (ie, frequency of practice). Similar to findings by Subnis et al,99 little has been reported regarding intervention practice adherence. Additionally, it should be noted that spiritually based interventions with fewer weeks of duration have shown positive benefits similar to those with longer durations. This is important because intervention designs need to consider examination of the minimum dose and duration required to achieve a positive effect on immune function. Furthermore, these interventions add a degree of burden to chronically ill participants who may not feel well enough to leave home and travel to a group setting. Future intervention design should consider modifying spiritually based interventions for use in personal (ie, home-based) settings. Previously, spiritually based interventions with PNI measures were of exploratory design, cross-sectional, and lacking a comparison arm. This review found that studies reviewed were generally more rigorous in approach than previous studies as a growing number of interventions (n = 6) were greater than 10 weeks in duration and the majority performed repeat measures ranging from 3 to 24 months postintervention. Moreover, it was feasible to find a number of studies with comparison groups (22 RCTs/non-RCTs) that satisfactorily met review criteria, allowing a more comprehensive examination of findings than earlier reviews that lacked comparison groups. Findings from this review are consistent with other reviews that found spiritually based interventions exerted a positive influence on psychological health, particularly on symptoms of depression, anxiety, mood, stress, and perceptions of quality of life. However, previous reviews have been mixed on the extent to which spiritually based interventions influence PNI outcome measures. This review revealed a positive pattern between spiritually based interventions and a number of PNI biomarker outcomes, suggesting that spiritually based interventions offer positive mental and physiological health benefits. For breast cancer survivors, these studies suggested that engaging in spiritually based practices may improve or stabilize the immune profile dysregulation that occurs with breast cancer. Future questions to be answered based on these findings include the following: which components of the interventions are most influential on outcomes (eg, mindfulness, breathing, stretching); and, are all components necessary to achieve the immune profile benefits? Additionally, what could the minimum dose (eg, frequency and length of practice) and duration (eg, weeks, months, or lifetime) be for achieving and sustaining neuroendocrine-immune benefits? Moreover, which spiritually based interventions are most predictive of biomarker outcomes? Limitations This review found that spiritually based interventions utilized disease-specific variables (ie, treatment-type, stage) for study inclusion criteria. However, this design does not allow for individual differences in psychosocial-spiritual variables (eg, perceptions of distress, spiritual beliefs, coping skills, and lifestyle behavior patterns). For these reasons, and as Carlson69 indicated previously, recruitment of a study population based on disease characteristics rather than psychosocial-spiritual considerations may result in unintended participant stress and impede measurement of the intervention’s therapeutic value. Similar to previous reviews, there remains a continued need for larger sample sizes and the inclusion of study power calculations. The direction of future spiritually based intervention design needs to standardize methods for intervention dosing, frequency, and duration of treatment. This review observed an increase in the numbers of study designs with randomization and longitudinal measures. However, because of the relative “newness” of spiritually based interventions in health care research, a variety of methods and study protocols exist between these interventions. Subsequently, these study methods and protocols vary in the reporting of details, resulting in a lack of guidance for prospective researchers wishing to reproduce intervention findings. A recent National Institutes of Health123 mandate for grant submissions highlights the need for reporting detailed study protocols and statistical computation methods in order to facilitate scientific validation of study findings. Finally, this review did not utilize a standard measure of study quality as this was beyond the focus of the review. Additionally, rigor of study randomization and blinding practices were not examined. However, screening and selection of intervention studies reviewed were limited to studies with comparison groups; and the majority of studies reviewed were randomized control trials. Conclusion This review found a positive pattern of relationships between spiritually based interventions, mental health outcomes, and neuroendocrine-immune function in breast cancer survivors. However, there were limited and uncertain benefits regarding the impact of spiritually based interventions on neuroendocrine-immune function among individuals receiving chemotherapy and/or radiotherapy. There is a growing body of evidence supporting relationships between RS self-report measures and mental health outcomes; however, RS self-report measures were rarely utilized when psychosocial-spiritual and PNI outcome measures (ie, biomarkers) were examined. Specifically, the trend in PNI intervention studies has been to measure a wide variety of psychosocial variables, with minimal measures of specific RS variables. Therefore, future research must clarify how best to address the issue regarding overlap of RS constructs in order for intervention outcomes to be more meaningful. Additional biomarker studies are needed with specific and standardized measures of RS in order to understand the relationships between RS variables and PNI health outcomes. Similarly, future studies should examine which biomarkers offer the most utility in predicting breast cancer survivorship outcomes including biomarkers associated with the risk of post–breast cancer treatment late effects. Future psychosocial-spiritual healthcare research might consider intervention designs that personalize treatment based on psychosocial-spiritual needs, as well as diverse participant setting considerations. Additionally, research that can inform on the minimum necessary spiritual-intervention dose and duration will serve to reduce survivors’ burden in managing post-breast cancer treatment-related symptoms. Finally, intervention designs will need to consider home-based delivery options, which remain important for survivors who are too ill to travel, have a preference for private/individual practice settings, or lack economic resources for traveling and transportation for group interventions. The authors would like to acknowledge the Mizzou Advantage Fund, the University of Missouri Interdisciplinary Center on Aging: Research Enrichment and Dissemination (READ) program, and Sigma Theta Tau International–Alpha Iota Chapter for Jennifer Hulett’s doctoral research program support. Funding The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Research reported in this publication was supported (in part) by the National Institute of Nursing Research of the National Institutes of Health under Award Number T32NR013456. Figure 1 Flowchart of article identification, screening, and selection process. Table 1 Spiritually Based Interventions and Psychosocial-Spiritual Outcomes in Breast Cancer Survivors. Study Number (See Table 3) Spiritually Based Interventions Psychosocial-Spiritual Outcomes (↓ or ↑)a 1,4, 6, 8, 15, 18 CBSM, MBSR, Qigong, RVT, Yoga ↓ Depression 1, 4, 6, 8, 12 CBSM, MBSR, RVT, Yoga ↓ Anxiety 2, 4, 14, 15, 18, 19, 20 MBCR, MBSR, Qigong, Yoga ↑ Quality of life/vitality/vigor 4 Yoga ↓ Distress 4, 17 MBSR, Yoga ↓ Symptoms 2 MBSR ↑ Coping 5, 7 CBSM ↑ Relaxation 1, 6, 8, 12, 14, 18 CBSM, MBCR, MBSR, RVT, Yoga ↓ Stress 16 Tai Chi ↑ Stress 9, 12, 13 MBSR ↑ Mindfulness 11 BMS ↑ Spiritual growth/spiritual well-being 13 MBSR ↓ Rumination 14 MBCR ↑ Social support 15, 17, 18, 19, 20 MBSR, Qigong, Stretching, Yoga ↓ Fatigue 9, 14 MBCR, MBSR ↑ Mood 17 MBSR ↑ Cognitive function (postchemotherapy) Abbreviations: CBSM, cognitive-based stress management; MBCR, mindfulness-based cancer recovery; MBSR, mindfulness-based stress reduction; RVT, relaxation visualization therapy; BMS, body-mind-spirit. a ↓ indicates decreased and ↑ indicates increased. Table 2 RS Interventions and Biomarker Outcomes in Breast Cancer Survivors. Study Number (See Table 3) RS Intervention Biomarker Outcomes (↓, ↑, or Stable)a 18 Yoga ↓ Interleukin (IL)-1 7, 20 GI, Yoga ↓ IL-1β 3, 6 GI, CBSM ↑ (IL)-2 2, 17 MBSR ↓ IL-4 6 CBSM ↑ IL-4 2, 20 MBSR, Yoga ↓ IL-6 2 MBSR ↓ IL-10 2, 3, 7 GI, MBSR ↑Natural killer (NK) cell activity 4 Yoga ↓ Immunoglobulin A (IgA) 4 Yoga ↑ CD56 % 2, 5, 6, 8, 9b, 11, 14, 19 BMS, CBSM, MBCR, MBSR, Yoga ↓ Cortisol 1, 12, 15, 18 MBSR, Qigong, RVT, Yoga Stable or no change in cortisol 18, 20 Yoga Stable or ↓ tumor necrosis factor–α (TNF-α) 9, 13 MBSR ↓ Blood pressure 9 MBSR ↓ Pulse 9 MBSR ↓ Respirations 2, 6, 10, 17 CBSM, MBSR ↑ Lymphocyte subsets; T cells, Th1/Th2 7, 17 GI, MBSR ↑ CD4+/CD8+ 2, 6, 17 CBSM, MBSR ↑ Interferon-γ (IFN-γ) 21, 22 MBSR Stable (preserved) telomere length 17 MBSR Stable/no change in CD4+ T lymphocytes 17 MBSR Stable/no change CD3+ subsets (T1/T2) Abbreviations: RS, religious and spiritual; GI, guided imagery; CBSM, cognitive-based stress management; MBSR, mindfulness-based stress reduction; BMS, body-mind-spirit; RVT, relaxation/visualization therapy; MBCR, mindfulness-based cancer recovery. a ↓ indicates decreased and ↑ indicates increased. b Decreased cortisol not sustained at 1 month follow-up. Table 3 RS Interventions With PNI Measures in Breast Cancer Survivors. No. Study (First Author, Year) Design Cancer Stage Intervention Group (IG) (n) vs Control Group (CG) (n) Intervention Duration Psychosocial and RS Measures Objective (PNI) Measures Key Health Outcomes 1 Nunes (2007)71 RCT I–II Relaxation/visualization (IG) (RVT)* (n = 20) vs CG: Assessment only* (n = 14) All participants undergoing concurrent radiotherapy 24 consecutive days, 30-min sessions Stress, Anxiety, Depression (ISSL, STAI, BAI, BDI) Cortisol RVT improved depression and anxiety scores (P < .05). Change in anxiety negatively correlated with cortisol (r = −0.38). RVT had no effect on cortisol levels 2 Witek-Janusek (2008)105 Non-RCT 0–II MBSR (IG) (n = 38) vs CG: Assessment only (n = 28) 8 weeks; 150 min/wk Quality of Life, Coping, Mindfulness (QOLI-cv3, JCS, MAAS) Lymphocytes, NKCA, Cytokines (Interleukin [IL])-2, IL-4, IL-6, IL-10, and interferon-gamma (IFN-γ), cortisol MBSR group had lower levels of NKCA, IL-4, IL-6, IFN-γ (all P < .04) compared with CG. cortisol was lower in MBSR group (P = .002) than CG. No effect on mindfulness (P > .05). Treatment group reestablished positive immune function while control group had continued immune dysregulation 3 Lengacher (2008)74 RCT 0–II Guided imagery (IG) (n = 15) vs Usual care CG (n = 13) 6–7 weeks (2–3 weeks preop through 4 weeks postop) None Natural killer (NK) cells, cytokine (IL-2) Guided imagery positively influenced NK cell cytotoxicity after IL-2 activation at 4 weeks postoperative compared with CG (P < .05) 4 Rao (2008)79 RCT II–III Yoga (IG) (n = 33) vs Supportive therapy and exercise CG (n = 36) 4 weeks (Anxiety, Depression, Function) (STAI, BDI, FLIC) T-lymphocyte subsets, CD4%, CD8%, NK cells, and immunoglobulin (IgA, IgG, IgM) Yoga group demonstrated decreased anxiety (P = .04), depression (P = .01), decreased symptom severity (P = .01), decreased distress (P < .01) and improved QOL (P = .01) compared with CG. Less immune dysfunction was observed in yoga group (decreased CD56%, P = .02) and (decreased IgA, P = .001) compared with CG 5 Phillips (2008)106 RCT 0–III CBSM (IG) (n = 65) vs 1-day education CG (n = 63) 10 weeks; repeat measures over 12 months Current Status (MOCS-Relaxation) Cortisol Greater reductions in cortisol levels across 12 months in CBSM compared with CG, although effect was small (d = 0.20). 6 Antoni (2009)107 RCT 0–III CBSM (IG) (n = 65) vs Usual care CG: 1-day education (n = 63) 10 weeks Stress, Anxiety & Depression, & Negative Mood (IES, HADS, ABS) Cortisol, lymphocyte subsets, cytokines (IL-2, IFN-γ, IL-4) CBSM had improved cortisol patterns (P < .01), improved IL-2 (P < .05), and improved IFN-γ (P < .01) function compared with CG during the first 6 months. Results suggest CBSM group may have experienced a buffering effect of adjuvant therapy compared with the CG. Improved trends observed in psychosocial measures; however, no significant effects observed between changes in psychosocial measures and biomarker measures 7 Eremin (2009)72 RCT II–IV Relaxation and guided imagery (IG) (n = 40) vs Standard care CG (n = 40) 37 weeks None T-cell subsets and lymphokine activated killer cells, B lymphocytes and monocytes; cytokines IL-1beta (1β), IL-2, IL-4 and IL-6 and TNF-α At 8 weeks, significant correlations observed between imaging ratings and natural killer cell activity (r = 0.319, P = .02). Relaxation frequency (r = 0.308, P = .018) and imagery ratings (r = 0.308, P = .019) correlated significantly with blood IL-1β, CD4+, and CD8+ levels. 8 Vadiraja (2009)108 RCT I–III Yoga (IG) (n = 44) vs Supportive therapy CG* (n = 44) All participants undergoing concurrent radiotherapy 6 weeks (3) 1-hour sessions per week Anxiety & Depression, Stress (HADS, PSS) Cortisol Positive correlations between decreased AM cortisol levels and decreased anxiety (Cohen’s f = 0.31), depression (f = 0.31), and stress (f = 0.36) in yoga group compared with CG 9 Matchim (2010)109 Non-RCT 0–II MBSR (IG) (n = 15) vs Wait list CG (n = 17) 8 weeks 90 min/wk Repeat measures at 1 month Mood, Stress, Mindfulness (POMS, C-SOSI, FFMQ) Salivary cortisol, blood pressure (BP), pulse, and respirations Increased mindfulness decreased BP, pulse, and respirations observed in MBSR compared with CG (P = .05 to P = .001). Initial decrease in AM cortisol within MBSR group was statistically significant (P < .05), but was not sustained at 1-month follow-up 10 Lengacher (2013)50 RCT 0–III MBSR (IG) (n = 42) vs Usual care CG (n = 40) 6 weeks 2 h/wk Repeat measures at 12 weeks None Lymphocyte subsets, T helper 1 and 2 cells (Th1/Th2), NK cells, IFN-γ, IL-4 Positive associations between all immune subset recoveries in MBSR group compared with CG. Women who received MBSR had T cells more readily activated by the mitogen phytohemagglutinin and an increase in the Th1/Th2 ratio (P = .002). MBSR associated with a more rapid return to normal immune function compared with CG, particularly in early posttreatment recovery periods 11 Hsiao (2012)104 RCT 0–III Body-mind-spirit (BMS) (IG) (n = 26) vs Education CG (n = 22) 8 weeks 2 h/wk Repeat measures at 5 and 8 months Depression Meaning in Life (Purpose, Search) (BDI, MLQ-P, MLQ-S) Cortisol At 5 months, BMS was related to greater spiritual growth (search for meaning in life) (P < .01). At 8 months, the BMS group demonstrated healthier cortisol patterns compared with CG (P < .05). 12 Branstrom (2012; 2013)110,111 RCT Data not provided MBSR (IG) (n = 32) vs Wait list CG (n = 39) 8 weeks 2 h/wk Repeat measures at 3 and 6 months Stress, Anxiety & Depression, Mood, Coping; Mindfulness (PSS, HADS, IES-R, PSOM, CSES, FFMQ) Cortisol MBSR associated with lower stress (P = .06), lower anxiety (P = .09), and increased mindfulness (P < .01). Although not quite significant (r = −0.38, P = .06) a trend was observed that MBSR demonstrated a moderate effect on awakening cortisol levels and was sustained at 6-month follow-up. Nonsignificant effects between stress and cortisol (P = .06) were observed 13 Campbell (2012)92 NonRCT Data not provided MBSR (IG) (n = 19) vs Wait list CG (n = 16) 8 weeks 90 min/wk Mindfulness, Rumination (MAAS, RRQ-rs) Blood pressure (BP) MBSR may improve mindfulness, moderate effect between decreased rumination and decrease systolic BP (r = 0.35), no main effects observed 14 Carlson (2013)68 RCT 0–IV MBCR (IG) (n = 113) vs Supportive emotional therapy (SET) (IG) (n = 104) vs Usual care CG: 1-day stress management seminar (n = 54) 8 weeks 90 min/wk Mood, Stress, Quality of Life, Social Support (POMS-TMD, C-SOSI, FACT-B, FACT-G, MOS-SSS) Cortisol Cortisol patterns were stable over time in both SET (P = .003) and MBCR (P = .014) groups relative to the CG, who had more flattened cortisol slopes. Women in MBCR improved more over time on stress symptoms compared with both SET (P = .009) and control (P = .023) groups. Greater improvements in MBCR group in quality of life compared with SET (P = .006) and CG (P = .005); and in social support compared with the SET (P = .012) 15 Chen (2013)112 RCT 0–III Qigong (IG) (n = 49) vs Wait list CG (n = 47)* Participants currently undergoing radiotherapy 5 weeks 40 min/wk with 1- and 3-month repeat measures Depression, Fatigue, Quality of Life, Sleep Disturbance (CES-D, BFI, FACT-G, PSQI) Cortisol Qigong group reported less depression over time than women in CG (P = .05). Women who had elevated depressive symptoms at the start of radiotherapy reported less fatigue (P < .01) and better overall quality of life (P < .05) in the Qigong group compared with the CG. No significant changes observed in cortisol slopes 16 Robins (2013)77 RCT I–IIIa (n = 109) (no further data) Tai Chi (IG) vs Spiritual growth group (IG) vs Usual care CG Concurrent chemotherapy 10 weeks 90 min/wk Repeat measures at 1 week, 4.5 months, 6 months Stress, Quality of Life, Depression (IES, FACT-B, CES-D) Cytokine panel (IL-1β, IL-2, IL-4, IL-5, IL-6, IL-7, IL-8, IL-10, IL-12, IL-13, IL-17, TNF-α) Interesting patterns in biomarkers observed; however, no statistically significant effects observed between intervention and control groups while currently receiving chemotherapy 17 Reich (2014)113 RCT 0–III MBSR (IG) (n = 17) vs Usual care/wait list CG (n = 24) 6 weeks 2 hs/wk Measures at baseline and 6 weeks Symptoms (MDASI) Lymphocyte subsets, mitogen-stimulated subsets, cytokines After 6 weeks, multiple baseline biomarkers were significantly positively improvement in GI symptoms (P = .035) and fatigue (P = .035) in MBSR group. Regression modeling identified B-lymphocytes and IFN-γ as the strongest predictors of gastrointestinal symptom improvement (P < .01). CD4+, CD8+ were predictive of strongest predictor of cognitive/psychological improvement (P = .02). Lymphocytes and IL-4 were strongest predictors of fatigue improvement (P < .01) 18 Bower (2012; 2014)78,114 RCT 0–II Iyengar yoga (IG) (n = 14) vs Usual care CG (n = 15) 12 weeks 90 min, twice weekly Repeat measures at baseline, 12 weeks, and 24 weeks Fatigue, Depression, Sleep, Stress, Vigor (FSI, BDI-II, PSQI, PSS, MSFI) Cortisol, tumor necrosis factor-alpha (TNF-α), IL-1, IL-6, CRP Decreased fatigue in yoga group from baseline to posttreatment and sustained at 3-month follow-up compared to CG (P = .032). Yoga group had significant increases in vigor compared with CG (P = .011). Both groups had positive changes in depressive symptoms and perceived stress (P < .05). Yoga group showed improved immune biomarker functioning compared with controls (P < .05). Tumor necrosis factor patterns remained stable in yoga group, while CG levels increased (P = 0.28). Similar trend observed with IL-1, but nonsignificant (P = .16). No significant changes in CRP, IL-6 or diurnal cortisol patterns 19 Chandwani (2014)115 RCT 0–III Yoga (IG) (n = 53) vs Stretching (IG) (n = 56)* vs Wait list CG (n = 54)* *Concurrently undergoing radiotherapy 6 weeks Up to 180 min/wk Repeat measures at 1, 3, and 6 months postintervention Quality of Life, Fatigue, Depression, Sleep (MO-SF 36, BFI, CES-D, PSQI) Cortisol Yoga group demonstrated greater increases in physical component scale scores compared with CG at 1 and 3 months after radiotherapy (P = .01 and P = .01, respectively). At 1, 3, and 6 months, the yoga group had greater increases in physical functioning compared with both stretching and CG (P < .05), with stretching and CG differences at only 3 months (P < .02). Yoga and stretching groups had reduced fatigue (P < .05) at radiotherapy completion. No group differences for mental health and sleep quality. Cortisol slope was steepest for the yoga group compared with the stretching and CG at the end (P = .023 and P = .008, respectively) and 1 month after radiotherapy completed (P = .05 and P = .04, respectively) 20 Kiecolt-Glaser (2014)116 RCT 0–IIIa Yoga (IG) (n = 100) vs Wait list CG (n = 100) 12 weeks 90 min/twice weekly Repeat measures at 3 months postintervention Fatigue, Quality of Life, Depression, Sleep, Activity, Diet (MFSI-SF, MO SF-36, CES-D, PQSI, CHAMPS, FFQ) Cytokines (IL-1β, IL-6, TNF-α) At posttreatment, fatigue was not lower in yoga compared with CG (P > .05) but vitality was higher (P = .01). At 3 months posttreatment, fatigue was lower in the yoga group (P = .002), vitality was higher (P = .01), and IL-6 (P = .027), TNF-α (P = .027), and IL-1 (P = .037) were lower for yoga group compared with CG. No group differences in depression at either time points (P > .20). Frequency of yoga practice showed a stronger association with fatigue and vitality, but not depression; greater changes associated with more frequent practice. At 3 months posttreatment, increased yoga practice was associated with decreased IL-6 (P = .01) and IL-1 (P = .03), but not TNF-α production (P > .05) 21 Lengacher (2014)117 RCT 0–III MBSR (IG) (n = 74) vs Usual care (n = 68) 6 weeks intervention Repeat measures at 12 weeks Recurrence Concerns, Mindfulness, Stress, Anxiety, Depression (CARS, CAMS, PSS, STAI, CES-D) Telomere length, telomere activity Telomere activity (TA) increased steadily over 12 weeks in MBSR group (17%) compared with minimal increase in CG (approximately 3%, P < .01). No effects observed on Telomere length (P = .92). MBSR appears to increase TA in peripheral blood mononuclear cells. TA was not associated with change in mindfulness, stress, anxiety, or fear of recurrence 22 Carlson (2015)118 RCT I–III MBCR (IG) (n = 53) vs Supportive emotional therapy group (SET) (IG) (n = 49) vs Usual care (n = 26) (1-day seminar) 8 weeks 90 min/wk plus 6-hour, 1-day retreat 12 weeks 90 min/wk 1-day, 6-hour, stress management seminar Mood, Stress (POMS, C-SOSI) Telomere length No correlations between measures of mood (P = .80) and stress (P = .24) and changes in telomere length between MBCR and SET groups and the CG (P = .28), or across the 2 intervention conditions (P = .31) and (P = .55). Telomere length in the MBCR and SET groups remained preserved (positive outcome) while a decrease among the CG was significant (P = .04) Abbreviations: RS, religious and spiritual; PNI, psychoneuroimmunological; RCT, randomized control trial group; IG, intervention group; CRP, C-reactive protein; CG, control group; RVT, relaxation and visualization therapy; ISSL, Inventory of Stress Symptoms Lipp; STAI, State-Trait Anxiety Inventory; BAI, Beck Anxiety Inventory; BDI, Beck Depression Inventory; non-RCT, nonrandomized control trial group; MBSR, mindfulness-based stress reduction; QOLI-v3, Quality of Life Index Cancer Version 3, JCS, Jaloweic Coping Scale; MAAS, Mindfulness Attention Awareness Scale; FLIC, Functional Living Index of Cancer; MOCS-R, Measure of Current Status–Relaxation; IES, Impact of Events Scale; HADS, Hospital Anxiety and Depression Scale; ABS, Affects Balance Scale; PSS, Perceived Stress Scale; POMS, Profile of Mood States; C-SOSI, Calgary Symptoms of Stress Inventory; FFMQ, Five Facet Mindfulness Questionnaire; MDASI, MD Anderson Symptom Inventory; MLQ-P, Meaning in Life Questionnaire, Presence; MLQ-S, Meaning in Life Questionnaire; IES-R, Impact of Events Scale–Revised; PSOM, Positive State of Mind; CSES, Coping Self-Efficacy Scale; RRQ-rs, Rumination Reflection Questionnaire—revised; MBCR, mindfulness-based cancer recovery; POMS-TMD, Profile of Mood States–Total Mood Disturbance; FACT-B, Functional Assessment of Cancer Therapy–Breast; FACT-G, Functional Assessment of Cancer Therapy–General; MOS-SSS, Medical Outcomes Study–Social Support Survey; SOSI, Symptoms of Stress Inventory; NK, natural killer; IFN-γ, interferon-gamma; IL, interleukin; DHEA, dehydroepiandrosterone sulfate; FACIT-G, Functional Assessment of Cancer Illness Therapy–General; CES-D, Center for Epidemiological Studies–Depression Scale; BFI, Brief Fatigue Inventory; PSQI, Pittsburg Sleep Quality Index; FSI, Fatigue Symptom Inventory; BDI-II, Beck Depression Inventory–II; MFSI, Multidimensional Fatigue Symptom Inventory; MO SF-36, Medical Outcomes Short-Form 36; CHAMPS, Community Healthy Activities Model Program for Seniors; FFQ, Food Frequency Questionnaire; CARS, Concerns About Recurrence Scale; CAMS, Cognitive and Affective Mindfulness Scale–Revised. Authors’ Note The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health, The University of Utah, or the University of Missouri. Declaration of Conflicting Interests The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. 1 Battaglini C Dennehy C Groff D Kirk D Anton P Complementary therapies in the management of cancer treatment-related symptoms: the individualized prescriptive exercise intervention approach Med Sportiva 2006 10 2 49 57 2 Armer JM Stewart BR Post-breast cancer lymphedema: incidence increases from 12 to 30 to 60 months Lymphology 2010 43 118 127 21226414 3 Clough-Gorr KM Ganz PA Silliman RA Older breast cancer survivors: factors associated with self-reported symptoms of persistent lymphedema over 7 years of follow-up Breast J 2010 16 147 155 19968661 4 Foster C Wright D Hill H Hopkinson J Roffe L Psychosocial implications of living 5 years or more following a cancer diagnosis: a systematic review of the research evidence Eur J Cancer Care (Engl) 2009 18 3 223 247 19432917 5 Fu MR Rosedale M Breast cancer survivors’ experiences of lymphedema-related symptoms J Pain Symptom Manage 2009 38 849 859 19819668 6 Armer JM Stewart BR Wanchai A Lasinski BB Smith KM Cormier JN Rehabilitation concepts among aging survivors living with and at risk for lymphedema: a framework for assessment, enhancing strengths, and minimizing vulnerability Top Geriatr Rehabil 2012 28 260 268 7 Feinstein M Liu K Ning H Fitchett G Lloyd-Jones DM Burden of cardiovascular risk factors, subclinical atherosclerosis, and incident cardiovascular events across dimensions of religiosity: the multi-ethnic study of atherosclerosis Circulation 2010 121 659 666 20100975 8 Howlader N Noone A Krapcho M SEER Cancer Statistics Review, 1975–2011 Bethesda, MD National Cancer Institute 2014 9 Pew Center for Center Pew Research Religion and Public Life Project Religious beliefs and practices 2008 http://religions.pewforum.org/reports# Accessed February 16, 2015 10 Bearon LB Koenig HG Religious cognitions and use of prayer in health and illness Gerontologist 1990 30 249 253 2347508 11 Altschuler A Rosenbaum E Gordon P Canales S Avins A Audio recordings of mindfulness-based stress reduction training to improve cancer patients’ mood and quality of life—a pilot feasibility study Support Care Cancer 2012 20 1291 1297 21688161 12 Fouladbakhsh J Stommel M Gender, symptom experience, and use of complementary and alternative medicine practices among cancer survivors in the U.S. cancer population Oncol Nurs Forum 2010 37 E7 E15 20044334 13 Lengacher CA Bennett MP Kip KE Frequency of use of complementary and alternative medicine in women with breast cancer Oncol Nurs Forum 2002 29 1445 1452 12432415 14 Wyatt G Sikorskii A Wills CE An HS Complementary and alternative medicine: use, spending, and quality of life in early stage breast cancer Nurs Res 2010 59 58 66 20010046 15 Yates JS Mustian KM Morrow GR Prevalence of complementary and alternative medicine use in cancer patients during treatment Support Care Cancer 2005 13 806 811 15711946 16 Pargament KI The Psychology of Religion and Coping: Theory, Research, Practice New York, NY Guilford Press 1997 17 Ano GG Vasconcelles EB Religious coping and psychological adjustment to stress: a meta-analysis J Clin Psychol 2005 61 461 480 15503316 18 Hulett JM Armer JM Stewart BR Wanchai A Perspectives of the breast cancer survivorship continuum: diagnosis through 30 months post-treatment J Pers Med 2015 5 174 190 26030800 19 Gall TL Cornblat MW Breast cancer survivors give voice: a qualitative analysis of spiritual factors in long-term adjustment Psychooncology 2002 11 524 535 12476434 20 McCain NL Gray DP Elswick RK Jr A randomized clinical trial of alternative stress management interventions in persons with HIV infection J Consult Clin Psychol 2008 76 431 441 18540736 21 Lengacher CA Johnson-Mallard V Post-White J Randomized controlled trial of mindfulness-based stress reduction (MBSR) for survivors of breast cancer Psychooncology 2009 18 1261 1272 19235193 22 Aldwin CM Park CL Jeong Y-J Nath R Differing pathways between religiousness, spirituality, and health: a self-regulation perspective Psychol Relig Spiritual 2014 6 9 23 Johnstone B Yoon DP Cohen D Relationships among spirituality, religious practices, personality factors, and health for five different faith traditions J Relig Health 2012 51 1017 1041 22618413 24 Piedmont RL Ciarrochi JW Dy-Liacco GS Williams JEG The empirical and conceptual value of the spiritual transcendence and religious involvement scales for personality research Psychol Relig Spiritual 2009 1 162 179 25 Reinert KG Koenig HG Re-examining definitions of spirituality in nursing research J Adv Nurs 2013 69 2622 2634 23600849 26 Hsiao YC Chien LY Wu LY Chiang CM Huang SY Spiritual health, clinical practice stress, depressive tendency and health-promoting behaviours among nursing students J Adv Nurs 2010 66 1612 1622 20492024 27 Johnstone B Yoon DP Franklin KL Schopp L Hinkebein J Re-conceptualizing the factor structure of the Brief Multidimensional Measure of Religiousness/Spirituality J Relig Health 2009 48 146 163 19421866 28 Hill PC Pargament KI Advances in the conceptualization and measurement of religion and spirituality. Implications for physical and mental health research Am Psychol 2003 58 64 74 12674819 29 Miller WR Thoresen CE Spirituality, religion, and health. An emerging research field Am Psychol 2003 58 24 35 12674816 30 Fetzer Institute/National Institute on Aging Working Group Multidimensional Measurement of Religiousness/Spirituality for Use in Health Research: A Report of the Fetzer Institute/National Institute on Aging Working Group Kalamazoo, MI Fetzer Institute 1999 31 Emmons RA Paloutzian RF The psychology of religion Annu Rev Psychol 2003 54 377 402 12171998 32 Johnstone B Glass BA Support for a neuropsychological model of spirituality in persons with traumatic brain injury Zygon 2008 43 861 874 33 Rose S Is the term “spirituality” a word that everyone uses, but nobody knows what anyone means by it? J Contemp Relig 2001 16 193 207 34 Johnstone B Bodling A Cohen D Christ SE Wegrzyn A Right parietal lobe-related “selflessness” as the neuropsychological basis of spiritual transcendence Int J Psychol Relig 2012 22 267 284 35 Piedmont RL Does spirituality represent the sixth factor of personality? Spiritual transcendence and the five-factor model J Pers 1999 67 985 1013 36 White ML Peters R Schim SM Spirituality and spiritual self-care: expanding self-care deficit nursing theory Nurs Sci Q 2011 24 48 56 21220576 37 Koenig H King D Carson VB Handbook of Religion and Health Oxford, England Oxford University Press 2012 38 Bodling A Heneghan M Walsh JC Yoon DP Johnstone B The Brief Multidimensional Measure of Religiousness/Spirituality with an Irish sample: a factor analysis Int J Ther Rehabil 2013 20 72 78 39 Cohen D Yoon DP Johnstone B Differentiating the impact of spiritual experiences, religious practices, and congregational support on the mental health of individuals with heterogeneous medical disorders Int J Psychol Relig 2009 19 121 138 40 Bower J Meyerowitz B Bernaards C Rowland J Ganz P Desmond K Perceptions of positive meaning and vulnerability following breast cancer: predictors and outcomes among long-term breast cancer survivors Ann Behav Med 2005 29 236 245 15946118 41 Stefanek M McDonald PG Hess SA Religion, spirituality and cancer: current status and methodological challenges Psychooncology 2005 14 450 463 15376283 42 Vespa A Jacobsen PB Spazzafumo L Balducci L Evaluation of intrapsychic factors, coping styles, and spirituality of patients affected by tumors Psychooncology 2011 20 5 11 20336646 43 Kroenke CH Kubzansky LD Schernhammer ES Holmes MD Kawachi I Social networks, social support, and survival after breast cancer diagnosis J Clin Oncol 2006 24 1105 1111 16505430 44 Lutgendorf SK Sood AK Anderson B Social support, psychological distress, and natural killer cell activity in ovarian cancer J Clin Oncol 2005 23 7105 7113 16192594 45 Powell ND Tarr AJ Sheridan JF Psychosocial stress and inflammation in cancer Brain Behav Immun 2013 30 suppl S41 S47 22790082 46 Andersen BL Farrar WB Golden-Kreutz D Stress and immune responses after surgical treatment for regional breast cancer J Natl Cancer Inst 1998 90 30 36 9428780 47 Witek-Janusek L Gabram S Mathews HL Psychologic stress, reduced NK cell activity, and cytokine dysregulation in women experiencing diagnostic breast biopsy Psychoneuroendocrinology 2007 32 22 35 17092654 48 Carlson LE Speca M Patel KD Goodey E Mindfulness-based stress reduction in relation to quality of life, mood, symptoms of stress, and immune parameters in breast and prostate cancer outpatients Psychosom Med 2003 65 571 581 12883107 49 Carlson LE Speca M Faris P Patel KD One year pre–post intervention follow-up of psychological, immune, endocrine and blood pressure outcomes of mindfulness-based stress reduction (MBSR) in breast and prostate cancer outpatients Brain Behav Immun 2007 21 1038 1049 17521871 50 Lengacher CA Kip KE Post-White J Lymphocyte recovery after breast cancer treatment and mindfulness-based stress reduction (MBSR) therapy Biol Res Nurs 2013 15 37 47 22084404 51 Carlson LE Speca M Patel KD Goodey E Mindfulness-based stress reduction in relation to quality of life, mood, symptoms of stress and levels of cortisol, dehydroepiandrosterone sulfate (DHEAS) and melatonin in breast and prostate cancer outpatients Psychoneuroendocrinology 2004 29 448 474 14749092 52 Loizzo J Charlson M Peterson J A program in contemplative self-healing: stress, allostasis, and learning in the Indo-Tibetan tradition Ann N Y Acad Sci 2009 1172 123 147 19735246 53 Matousek RH Pruessner JC Dobkin PL Changes in the cortisol awakening response (CAR) following participation in mindfulness-based stress reduction in women who completed treatment for breast cancer Complement Ther Clin Pract 2011 17 65 70 21457893 54 Lengacher CA Kip KE Barta M A pilot study evaluating the effect of mindfulness-based stress reduction on psychological status, physical status, salivary cortisol, and interleukin-6 among advanced-stage cancer patients and their caregivers J Holist Nurs 2012 30 170 185 22442202 55 Ader R Cohen N Behaviorally conditioned immunosuppression Psychosom Med 1975 37 333 340 1162023 56 Zeller JM McCain NL Swanson B Psychoneuroimmunology: an emerging framework for nursing research J Adv Nurs 1996 23 657 664 8675880 57 Ray O How the mind hurts and heals the body Am Psychol 2004 59 29 40 14736318 58 Ader R Psychoneuroimmunology Curr Dir Psychol Sci 2001 10 94 98 59 Lutgendorf SK Costanzo ES Psychoneuroimmunology and health psychology: an integrative model Brain Behav Immun 2003 17 225 232 12831823 60 Besedovsky H DelRey A Sorkin E Immune-neuroendocrine interactions J Immunol 1985 135 2 suppl 750s 754s 4008933 61 Chrousos GP Gold PW The concepts of stress and stress system disorders. Overview of physical and behavioral homeostasis. JAMA. 1992;267:1244–1252 Erratum in: JAMA 1992 268 200 62 Sapolsky RM Romero LM Munck AU How do glucocorticoids influence stress responses? Integrating permissive, suppressive, stimulatory, and preparative actions Endocr Rev 2000 21 55 89 10696570 63 Hellhammer DH Wüst S Kudielka BM Salivary cortisol as a biomarker in stress research Psychoneuroendocrinology 2009 34 163 171 19095358 64 McEwen BS Mood disorders and allostatic load Biol Psychiatry 2003 54 200 207 12893096 65 Sephton SE Sapolsky RM Kraemer HC Spiegel D Diurnal cortisol rhythm as a predictor of breast cancer survival J Natl Cancer Inst 2000 92 994 1000 10861311 66 Glaser R Rabin B Chesney M Cohen S Natelson B Stress-induced immunomodulation: implications for infectious diseases? JAMA 1999 281 2268 2270 10386538 67 McEwen BS Protective and damaging effects of stress mediators N Engl J Med 1998 338 171 179 9428819 68 Carlson LE Doll R Stephen J Randomized controlled trial of mindfulness-based cancer recovery versus supportive expressive group therapy for distressed survivors of breast cancer (Mindset) J Clin Oncol 2013 31 3119 3126 Erratum in: J Clin Oncol, 2014, 32, 3686–3687 23918953 69 Carlson LE Mindfulness-based interventions for physical conditions: a narrative review evaluating levels of evidence ISRN Psychiatry 2012 2012 651583 23762768 70 Ludwig DS Kabat-Zinn J Mindfulness in medicine JAMA 2008 300 1350 1352 18799450 71 Nunes DFT Rodriguez AL da Silva Hoffmann F Relaxation and guided imagery program in patients with breast cancer undergoing radiotherapy is not associated with neuroimmunomodulatory effects J Psychosom Res 2007 63 647 655 18061756 72 Eremin O Walker MB Simpson E Immunomodulatory effects of relaxation training and guided imagery in women with locally advanced breast cancer undergoing multimodality therapy: a randomised controlled trial Breast 2009 18 17 25 19008099 73 Walker LG Walker MB Ogston K Psychological, clinical and pathological effects of relaxation training and guided imagery during primary chemotherapy Br J Cancer 1999 80 262 268 10390006 74 Lengacher CA Bennett MP Gonzalez L Immune responses to guided imagery during breast cancer treatment Biol Res Nurs 2008 9 205 214 18077773 75 Academy for Guided Imagery How guided imagery works Mechanisms of action according to basic principles of interactive guided imagerySM 2015 http://acadgi.com/abouttheacademy/FAQs/index.html Accessed February 10, 2016 76 Chen K-M Snyder M A research-based use of Tai Chi/movement therapy as a nursing intervention J Holist Nurs 1999 17 267 279 10690069 77 Robins JL McCain NL Elswick RK Jr Walter JM Gray DP Tuck I Psychoneuroimmunology-based stress management during adjuvant chemotherapy for early breast cancer Evid Based Complement Alternat Med 2013 2013 372908 23762127 78 Bower JE Greendale G Crosswell AD Yoga reduces inflammatory signaling in fatigued breast cancer survivors: a randomized controlled trial Psychoneuroendocrinology 2014 43 20 29 24703167 79 Rao RM Nagendra HR Raghuram N Influence of yoga on mood states, distress, quality of life and immune outcomes in early stage breast cancer patients undergoing surgery Int J Yoga 2008 1 11 20 21829279 80 Oh B Butow P Mullan B A critical review of the effects of medical Qigong on quality of life, immune function, and survival in cancer patients Integr Cancer Ther 2012 11 101 110 21715370 81 Chen K Yeung R Exploratory studies of Qigong therapy for cancer in China Integr Cancer Ther 2002 1 345 370 14664729 82 Oh B Butow P Mullan B Impact of medical Qigong on quality of life, fatigue, mood and inflammation in cancer patients: a randomized controlled trial Ann Oncol 2010 21 608 614 19880433 83 Cohen L Chen Z Arun B External QiGong therapy for women with breast cancer prior to surgery Integr Cancer Ther 2010 9 348 353 21106615 84 Brown KW Ryan RM The benefits of being present: mindfulness and its role in psychological well-being J Pers Soc Psychol 2003 84 822 848 12703651 85 Carlson LE Garland SN Impact of mindfulness-based stress reduction (MBSR) on sleep, mood, stress and fatigue symptoms in cancer outpatients Int J Behav Med 2005 12 278 285 16262547 86 Birnie K Garland SN Carlson LE Psychological benefits for cancer patients and their partners participating in mindfulness-based stress reduction (MBSR) Psychooncology 2010 19 1004 1009 19918956 87 Ott MJ Norris RL Bauer-Wu SM Mindfulness meditation for oncology patients: a discussion and critical review Integr Cancer Ther 2006 5 98 108 16685074 88 Altschuler A Rosenbaum E Gordon P Canales S Avins AL Audio recordings of mindfulness-based stress reduction training to improve cancer patients’ mood and quality of life—a pilot feasibility study Support Care Cancer 2012 20 1291 1297 21688161 89 Monti DA Peterson C Kunkel EJ A randomized, controlled trial of mindfulness-based art therapy (MBAT) for women with cancer Psychooncology 2006 15 363 373 16288447 90 Shapiro SL Carlson LE Astin JA Freedman B Mechanisms of mindfulness J Clin Psychol 2006 62 373 386 16385481 91 Speca M Carlson LE Goodey E Angen M A randomized, wait-list controlled clinical trial: the effect of a mindfulness meditation-based stress reduction program on mood and symptoms of stress in cancer outpatients Psychosom Med 2000 62 613 622 11020090 92 Campbell TS Labelle LE Bacon SL Faris P Carlson LE Impact of mindfulness-based stress reduction (MBSR) on attention, rumination and resting blood pressure in women with cancer: a waitlist-controlled study J Behav Med 2012 35 262 271 21667281 93 Shapiro SL Brown KW Biegel GM Teaching self-care to caregivers: effects of mindfulness-based stress reduction on the mental health of therapists in training Train Educ Prof Psychol 2007 1 105 115 94 Dobkin PL Mindfulness-based stress reduction: What processes are at work? Complement Ther Clin Pract 2008 14 8 16 18243937 95 Garland SN Carlson LE Cook S Lansdell L Speca M A non-randomized comparison of mindfulness-based stress reduction and healing arts programs for facilitating post-traumatic growth and spirituality in cancer outpatients Support Care Cancer 2007 15 949 961 17611782 96 Ledesma D Kumano H Mindfulness-based stress reduction and cancer: a meta-analysis Psychooncology 2009 18 571 579 19023879 97 Matchim Y Armer JM Stewart BR Mindfulness-based stress reduction among breast cancer survivors: a literature review and discussion Oncol Nurs Forum 2011 38 E61 E71 21356643 98 Zainal NZ Booth S Huppert FA The efficacy of mindfulness-based stress reduction on mental health of breast cancer patients: a meta-analysis Psychooncology 2013 22 1457 1465 22961994 99 Subnis UB Starkweather AR McCain NL Brown RF Psychosocial therapies for patients with cancer: a current review of interventions using psychoneuroimmunology-based outcome measures Integr Cancer Ther 2013 13 85 104 24105361 100 Chan CL Wang CW Ho RT A systematic review of the effectiveness of Qigong exercise in supportive cancer care Support Care Cancer 2012 20 1121 1133 22258414 101 Center for Evidence-based Medicine Center for evidenced-based medicine (CEBM) levels of evidence 2009 http://www.cebm.net/oxford-centre-evidence-based-medicine-levels-evidence-march-2009/ Accessed March 12, 2015 102 Zeng Y Luo T Xie H Huang M Cheng AS Health benefits of Qigong or Tai Chi for cancer patients: a systematic review and meta-analyses Complement Ther Med 2014 22 173 186 24559833 103 Oh PJ Kim SH The effects of spiritual interventions in patients with cancer: a meta-analysis Oncol Nurs Forum 2014 41 E290 E301 25158666 104 Hsiao F Jow G Kuo W The effects of psychotherapy on psychological well-being and diurnal cortisol patterns in breast cancer survivors Psychother Psychosom 2012 81 173 182 22399076 105 Witek-Janusek L Albuquerque K Chroniak KR Chroniak C Durazo-Arvizu R Mathews HL Effect of mindfulness based stress reduction on immune function, quality of life and coping in women newly diagnosed with early stage breast cancer Brain Behav Immun 2008 22 969 981 18359186 106 Phillips KM Antoni MH Lechner SC Stress management intervention reduces serum cortisol and increases relaxation during treatment for nonmetastatic breast cancer Psychosom Med 2008 70 1044 1049 18842742 107 Antoni MH Lechner S Diaz A Cognitive behavioral stress management effects on psychosocial and physiological adaptation in women undergoing treatment for breast cancer Brain Behav Immun 2009 23 580 591 18835434 108 Vadiraja H Raghavendra RM Nagarathna R Effects of a yoga program on cortisol rhythm and mood states in early breast cancer patients undergoing adjuvant radiotherapy: a randomized controlled trial Integr Cancer Ther 2009 8 37 46 19190034 109 Matchim Y Armer JM Stewart BR Effects of mindfulness-based stress reduction (MBSR) on health among breast cancer survivors West J Nurs Res 2010 33 996 1016 20956583 110 Bränström R Kvillemo P Moskowitz JT A randomized study of the effects of mindfulness training on psychological well-being and symptoms of stress in patients treated for cancer at 6-month follow-up Int J Behav Med 2012 19 535 542 21931958 111 Bränström R Kvillemo P Åkerstedt T Effects of mindfulness training on levels of cortisol in cancer patients Psychosomatics 2013 54 158 164 23218058 112 Chen Z Meng Z Milbury K Qigong improves quality of life in women undergoing radiotherapy for breast cancer Cancer 2013 119 1690 1698 23355182 113 Reich RR Lengacher CA Kip KE Baseline immune biomarkers as predictors of MBSR (BC) treatment success in off-treatment breast cancer patients Biol Res Nurs 2014 16 429 437 24477514 114 Bower JE Garet D Sternlieb B Yoga for persistent fatigue in breast cancer survivors Cancer 2012 118 3766 3775 22180393 115 Chandwani KD Perkins G Nagendra HR Randomized, controlled trial of yoga in women with breast cancer undergoing radiotherapy J Clin Oncol 2014 32 1058 1065 24590636 116 Kiecolt-Glaser JK Bennett JM Andridge R Yoga’s impact on inflammation, mood, and fatigue in breast cancer survivors: a randomized controlled trial J Clin Oncol 2014 32 1040 1049 24470004 117 Lengacher CA Reich RR Kip KE Influence of mindfulness-based stress reduction (MBSR) on telomerase activity in women with breast cancer (BC) Biol Res Nurs 2014 16 438 447 24486564 118 Carlson LE Beattie TL Giese-Davis J Mindfulness-based cancer recovery and supportive-expressive therapy maintain telomere length relative to controls in distressed breast cancer survivors Cancer 2015 121 476 484 25367403 119 Rossi D Lobetti Bodoni C Genuardi E Telomere length is an independent predictor of survival, treatment requirement and Richter’s syndrome transformation in chronic lymphocytic leukemia Leukemia 2009 23 1062 1072 19340005 120 Bisoffi M Heaphy CM Griffith JK Telomeres: Prognostic markers for solid tumors Int J Cancer 2006 119 2255 2260 16858685 121 Boelens PA Reeves RR Replogle WH Koenig HG A randomized trial of the effect of prayer on depression and anxiety Int J Psychiatry Med 2009 39 377 392 20391859 122 Bormann JE Aschbacher K Wetherell JL Roesch S Redwine L Effects of faith/assurance on cortisol levels are enhanced by a spiritual mantram intervention in adults with HIV: a randomized trial J Psychosom Res 2009 66 161 171 19154859 123 National Institute of Health Grants and funding: rigor and reproducibility 2015 http://grants.nih.gov/reproducibility/index.htm Accessed February 10, 2016
PMC005xxxxxx/PMC5125068.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 0372370 670 Annu Rev Microbiol Annu. Rev. Microbiol. Annual review of microbiology 0066-4227 1545-3251 19450140 5125068 10.1146/annurev.micro.091208.073431 NIHMS831066 Article Lipid Signaling in Pathogenic Fungi Rhome Ryan [email protected] 1 Del Poeta Maurizio [email protected] 123 1 Department of Biochemistry and Molecular Biology, Medical University of South Carolina, Charleston South Carolina 29425 2 Department of Microbiology and Immunology, Medical University of South Carolina, Charleston South Carolina 29425 3 Division of Infectious Diseases, Medical University of South Carolina, Charleston South Carolina 29425 21 11 2016 2009 28 11 2016 63 119131 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Lipid signaling in pathogenic fungi has been studied to determine the role of these pathways in fungal biology and human infections. Owing to their unique nature, they may represent targets for future antifungal treatments. Farnesol signaling was characterized as a quorum-sensing molecule, with exposure inhibiting filamentation. Research has shown involvement in both the Ras1-adenylate cyclase and MAP kinase pathways. In species of Aspergillus, farnesol exposure induces apoptosis-like changes and alterations in ergosterol synthesis. Eicosanoid production has been characterized in several pathogenic fungi, utilizing host lipids in some cases. The role in virulence is not known yet, but it may involve modulation of host lipids. Sphingolipid signaling pathways seem to center around the production of diacylglycerol in the formation of inositol phosphorylceramide. Diacylglycerol activates both melanin production through laccase and transcription of antiphagocytic protein, both of which are involved in virulence. farnesol prostaglandin Cryptococcus sphingolipid Aspergillus Candida INTRODUCTION Pathogenic fungi such as Candida, Aspergillus, and Cryptococcus continue to be a significant medical issue worldwide. The search for more effective and specific therapies for these infections continues to this day. Among the pathways and systems examined for potential new drug targets, lipid signaling has emerged as an attractive and interesting field. The lipid signaling pathways involving farnesol (FA), eicosanoids, and sphingolipids in these fungi have been the best characterized to date. FARNESOL SIGNALING One major player in lipid signaling pathways of pathogenic fungi is FA, a 15-carbon oxygenated lipid made up of isoprene moieties. FA has the distinction of being the first quorum-sensing molecule identified in eukaryotes. Hornby et al. (26) reported that a lipid-soluble molecule released into the media was responsible for inoculum size-dependent growth changes in Candida albicans, characterized by a reduction in yeast-to-mycelia transitions. Further evaluations of FA-related phenotypes revealed that both biofilm formation and hyphal development are suppressed in C. albicans. This phenomenon may not be limited to C. albicans, as there is evidence that FA reduces hyphal formation in C. dubliniensis (23) and alters gene expression and cell polarization in C. parapsilosis (55). FA secretion and biofilm formation have been documented for eight Candida species under a variety of conditions (66), although C. albicans seems to have the highest FA secretion. FA-mediated hyphal blockage was evaluated under a variety of different conditions, including several defined media and serum-containing conditions (43). C. albicans grown in all defined media tested showed similar FA-related reduction in germ tube formation, with about 1 μM of FA needed to show a 50% decrease. FA susceptibility was greatly reduced in the presence of serum (albumin), implying that the dynamics of FA signaling in relation to pathophysiology may be more complex than in vitro studies suggest. Mosel et al. (43) also showed that FA had little effect on existing germ tubes, suggesting the main role of FA is prevention of formation initiation. In addition to the effects on hyphal formation, FA is partially protective against oxidative stress (67), but not in a manner that induces classical antioxidant genes such as catalase or superoxide dismutases. A potential link between FA production and pathogenicity was alluded to by Navarathna et al. (45), who both increased FA production and enhanced virulence of C. albicans in mouse models of infection when the fungus was pretreated with low concentrations of azoles. Azoles are a class of antifungal drugs that inhibits the synthesis of ergosterol, a major component of fungal membranes. This azole-induced production of FA had been seen previously when it was treated with a variety of different azoles (27). Farnesyl pyrophosphate (FPP) is a common precursor to the synthesis of both FA and ergosterol, the major sterol component of fungal membranes. Because azole drugs disrupt ergosterol synthesis after the formation of FPP, treatment with azoles leads to an accumulation of FPP that then leads to an increased production of FA. Since this observation, several aspects of virulence have been examined with respect to FA. Navarathna et al. (46) created a mutant strain of C. albicans with reduced endogenous FA production and showed that this strain had reduced virulence in mouse models compared with wild-type and reconstituted strains. Additionally, exogenous FA administration increased murine mortality (46) and altered host cytokine production (46, 47, 58) during infection. Shchepin et al. (58) looked at synthetic compounds comparable to FA in quorum-sensing attributes in relation to mouse models of C. albicans infection. They showed that although the synthetic compounds varied in their function as virulence factors, neither of the compounds tested protected mice from infection. This could imply a strict structural requirement of these molecules for function. Taken together, these data further support the link between FA and Candida pathogenesis and highlight the need for future investigations into the exploitation of this molecular process for therapies. The various effects of FA on Candida are clear, but what signaling pathways underlie these phenotypes? The answer to this question may be the key to understanding the complex role of FA in the life cycle and pathogenesis of this fungus. Some clues to these processes come from microarray analyses. Cao et al. (6) looked at gene expression in C. albicans biofilms inhibited by FA exposure. They found that along with genes involved in processes such as drug resistance and heat shock, the expression of several genes associated with hyphal formation was altered when exposed to FA. Genes of interest in this group include TUP1 (dTMP uptake), PDE2 (phosphodiesterase), and CRK1 (Cdc2-related kinase). Tup1 is a transcriptional repressor that is upregulated in the presence of FA, whereas CRK1 and PDE2 are downregulated. TUP1 downregulation is associated with hyperfilamentation, while PDE2 and CRK1 have been hypothesized to positively effect filamentation. Taken together, FA affects the expression of these genes in a manner consistent with the effect of FA treatment on hyphal formation. Enjalbert et al. (14) studied effects of FA on gene expression profiles in strains of C. albicans that had resumed growth after stationary phase upon transfer to fresh media. In these cases, FA delayed this rebound of growth and changed the expression of genes that are normally affected (positively or negatively) during hyphal formation. Among the affected genes are cyclin genes HGC1, CLN3, and PCL2 and histone genes. For example, HGC1 is necessary for hyphal development (69) and the microarray data show that FA exposure downregulates HGC1. Another group examined the expression profile of C. albicans genes in the early yeast-to-hyphae transition period when grown in nonenriched media (8). They found that many genes affected by FA are also affected by high cell density, which would be expected from a quorum-sensing molecule. In general, they found that under these conditions expression profiles were similar to those seen in studies of gene expression C. albicans following phagocytosis (38). Although important to the molecular biology of signaling, gene expression alone cannot determine the specific roles of factors in the cellular processes underlying FA signaling. Further examination of some of the genes/proteins indicated in the microarray studies has yielded interesting results. Tup1 acts globally to repress hyphal formation (3). Tup1 interacts with other proteins such as Tcc1 (30) that function as corepressors after complexing with DNA binding proteins such as Nrg1 (5) and Rfg1 (29, 33), which are homologous to Saccharomyces cerevisiae proteins Nrg1p and Rox1p, respectively. When TUP1 is deleted, the strain shows constitutive filamentous growth with an inability to grow as yeast (3). Furthermore, activation of Tup1 and related factors results in a downregulation of genes that are involved in filament formation (4, 44). Nrg1 is a repressor of filament growth, and deletions of NRG1 have phenotypes similar to TUP1 deletions (5). nrg1−/− mutants show no virulence in mouse models of candidiasis. Another interesting aspect of the study examines hyphal genes that are affected by Nrg1, with Tup1-controlled genes ECE1 and HWP1 (4, 64) constitutively expressed in the nrg−/− mutant. RFG1 deletion showed a phenotype similar to the nrg−/−, and the double mutant showed colonies with a more pronounced wrinkled appearance than did the singular deletions. Examinations of the link between FA, Tup1, and these related DNA binding proteins have shown promising results. Kebaara et al. (31) recently established this link by examining the response of various knockout mutants to FA. The tup1−/− and nrg−/− mutants do not exhibit a morphological response to FA, whereas the rfg1−/− mutant does respond to FA. As seen before, treatment with FA increases TUP1 expression with a concomitant decrease in hyphae-specific genes such as HWP1 and RBT1. Interestingly, both tup1−/− and nrg1−/− mutants overproduced FA up to 19 times more than wild type did. The implications of this finding suggest a possible feedback mechanism between Tup1, Nrg1, and the production of FA. Other pathways seem to be involved in FA-related changes in the morphology of C. albicans. Ras1, a GTPase, activates the enzyme adenylate cyclase (AC), which produces the common second messenger cyclic-AMP (cAMP). cAMP activates protein kinase A (PKA), which phosphorylates many downstream factors. One of these is a transcription factor called Efg1, which induces the morphological transition to hyphae (18, 35). Davis-Hanna et al. (12) established that this Ras1-AC-PKA-Efg1 pathway is inhibited byFA and rescued from that inhibition in an Efg1-dependent manner upon addition of a cAMP analog, dibutyryl-cAMP. There is evidence that FA influences mitogen-activated protein (MAP) kinase cascades through reduced expression of genes such as HST7 (a MAP kinase kinase) and CPH1 (a transcription factor involved in hyphal formation) upon treatment, and the reduced cascade activity was measured indirectly by showing general amino acid permease 1 (GAP1) (56). Because Ras1 is a common factor of the MAP kinase pathway and the PKA-Efg1 pathway, it is possible that this is the molecule that FA acts upon. A summary of factors in farnesol signaling in C. albicans can be found in Figure 1. Another kinase gene involved in FA response and quorum sensing in C. albicans is CHK1, which encodes a histidine kinase. The deletion of CHK1 led to a strain (chk1−/−) that was unresponsive to FA treatment even at higher doses, whereas deletion of other histidine kinases such as Sln1p and Nik1p produced strains that responded to FA normally (34). This is possibly connected to previous findings in which CHK1 interacts with MAP kinase, one of its downstream regulators, or another yet unidentified mediator. Chk1 does not appear to be involved in quorum-related antifungal resistance (52). The focus of FA signaling thus far has been on C. albicans, but other pathogenic fungi have shown responses to FA as well. Recently, biofilm formation in Pneumocystis carinii has been described and is inhibited by exposure to FA in this pathogenic fungus as well (11). Of note, some Aspergillus species have shown various effects upon FA treatment. FA does not affect the hyphal morphology of A. nidulans, but it does inhibit conidiation of A. niger (37). Along with this conidiation inhibition, FA treatment decreases the intracellular levels of cAMP, which is produced by AC. Treatment with inhibitors of AC induces a phenotype similar to that induced by FA, which suggests that cAMP may be involved in the mechanism of reduced conidiation. A. nidulans does show an interesting phenotype upon FA treatment, which is consistent with morphological changes associated with apoptosis (57). This phenotype includes DNA fragmentation and nuclear condensation. Recent studies have further examined this phenotype and found a decrease in mRNA transcripts for proteins involved in transcription, translation, ergosterol biosynthesis, and ribosomal biogenesis along with an increase in transcripts for mitochondrial proteins. These studies found that treatment with FA induces autophagy and mitochondrial fragmentation. Ergosterol biosynthesis in A. nidulans seems to be affected, as evidenced by a punctate distribution of filipin staining (fluorescent compound that binds ergosterol) as opposed to the uniform distribution of untreated fungi. Because the studies of ergosterol polarization have involved sphingolipid-rich lipid rafts in C. albicans (41), it would be worthwhile to further explore the links between FA exposure and characteristics of lipid raft domains. Of interest to signaling studies is the observation that the protein kinase C (PKC)-deficient strain of A. nidulans (calC2) showed resistance to FA treatment. Some studies have shown links between PKC and autophagy (51), though not in Aspergillus. Could these FA-related observations be explained by a common cause involving PKC? One notable aspect of the research described above relating to Aspergillus species and FA is that Aspergillus species do not seem to produce detectable amounts of FA. This could mean that the phenotypes described are laboratory artifacts and that these phenomena do not occur in nature. More likely is the idea that FA, in addition to quorum sensing, is produced by Candida to compete with other organisms such as Aspergillus. This idea adds a different dimension to the FA studies, implying a complex system of regulating not only the growth of Candida itself but also its competition with other fungi and microbes for resources. The FA-mediated interaction between Candida and Aspergillus is not the only documented case of different microorganisms interacting through FA. Pseudomonas aeruginosa is a gram-negative bacteria that often coexists with Candida species in clinical situations such as cystic fibrosis. FA inhibits swarming motility in P. aeruginosa (42) and reduces the production of pyocyanin, a known virulence factor (10). In summary, many signaling genes and factors influenced by FA lead to the morphological changes observed. Although it can seem daunting, important common factors that appear in these studies suggest an overall theme of signaling and phenotype. The Ras1-cAMP connection appears to be common not only in Candida, but also in several pathways implicated in the FA response, including Aspergillus. In the past few years, significant advancements have furthered our understanding of this signaling molecule. Future research will revolve around the mechanism by which FA activates these pathways, alternative pathways involved in these morphological changes, the role of FA in complex microbial interactions, and whether FA can be exploited therapeutically. EICOSANOID SIGNALING Although a major player in fungal lipid signaling, FA is not the only lipid molecule that affects signaling pathways in fungi. Eicosanoids are one such family of lipids that plays roles in a number of pathogenic fungi. Eicosanoids are 20-carbon-long oxygenated lipids derived from fatty acids such as arachidonic acid. Included in the family of eicosanoids are groups of lipids, such as prostaglandins and leukotrienes, that have various effects related to immune function and inflammation in humans. Noverr et al. (49) first reported the occurrence of prostaglandins and leukotrienes in culture supernatants of Cryptococcus neoformans and C. albicans. Synthesis of the prostaglandins appeared to be from both de novo pathways and salvage of precursors (such as arachidonic acid) from the media. However, Wright et al. (68) confirmed that there is no arachidonic acid produced endogenously by Cryptococcus. C. albicans takes up arachidonic acid (13).Other pathogenic fungi synthesize these eicosanoids. Noverr et al. (50) showed that several different strains and species of pathogenic fungi produce both leukotrienes and prostaglandins in culture. This production was dramatically increased when arachidonic acid was added to culture media. Future studies confirmed that the prostaglandins purified from C. neoformans (15) and C. albicans (17) were of the prostaglandin E2 (PGE2) subtype. Once purified, these prostaglandins exhibited effects on mammalian cells (including cytokine modulation) similar to effects exhibited by mammalian PGE2. Inhibitors of cyclooxygenase (COX), a class of enzymes that synthesize prostaglandins from arachidonic acid, have varied effects on these fungi. Paracoccidioides brasiliensis shows reduced production of PGE2 when treated with COX inhibitors (2). In C. albicans, synthesis of PGE2 is stopped by COX and lipoxygenase inhibitors that are nonspecific with respect to the isoenzyme targeted (17). However, compounds that specifically inhibit the COX2 isoenzyme do no affect PGE2 production. Alem et al. (1) found a difference between PGE2 production by suspended Candida and Candida associated with a biofilm, with the latter producing more. Both of these conditions showed a block in PGE2 synthesis upon treatment with nonspecific COX inhibitors. Because C. albicans has no homolog to COX enzymes, alternative synthesis strategies were examined. Two such enzymes, the fatty acid desaturase Ole2 and the copper oxidase Fet3, were implicated when deletion of these enzymes produced strains with greatly reduced PGE2 synthesis. Given that deletion of either enzyme did not completely abolish PGE2 production, it seems that other synthesis pathways may exist. Taken together, the synthesis of PGE2 in C. albicans might occur through an enzyme with some general structural similarity but little homology to the known COX enzymes. As for the effect of COX inhibitors on C. neoformans, the answer is not as clear. Early reports (49) showed that COX inhibitor treatment decreased the amount of PGE2 in cultures; however, this seemed to result from a decrease in the number of viable cells. More recent studies (15) have shown that COX inhibitor treatment does not affect cryptococcal production of PGE2. Like C. albicans, C. neoformans has no homolog to known COX enzymes. These two observations suggest that an entirely different mechanism of PGE2 synthesis occurs in these fungi. Erb-Downward et al. (16) followed up on these ideas and found that several polyphenolic chemicals that typically inhibit lipoxygenase activity inhibit production of PGE2. Laccase, the enzyme in Cryptococcus responsible for melanin synthesis, binds phenolic substances and therefore was examined for a role in eicosanoids production. Deletion or inhibition of laccase resulted in a lack of PGE2 production, but laccase alone was unable to synthesize PGE2 from either arachidonic acid or prostaglandin H2. Interestingly, laccase converted prostaglandin G2 to PGE2. The multicopper oxidase Fet3 is homologous to laccase (17). These findings indicate that we are just beginning to understand prostaglandin production in Cryptococcus. Little is known about the direct effect of PGE2 on these fungi, though Levitin et al. (36) reported several alterations in the transcription profile of C. albicans upon exogenous exposure. Another class of lipids derived from omega-3-polyunsaturated fatty acids like arachidonic acid is the resolvins. In human cells, these compounds are anti-inflammatory and affect migration of neutrophils in later stages of inflammation. Recently, Haas-Stapleton et al. (22) described the synthesis of fungal resolvins in C. albicans. The synthesis of these resolvins was inhibited by lipoxygenase inhibitors, and production of these lipids by C. albicans affected neutrophil chemotaxis, phagocytosis, and intracellular killing at different concentrations. This finding is more evidence for the role of eicosanoid-derived lipids in fungal modulation of the host immune system. Phospholipases may also play a role in the eicosanoid signaling story. Phospholipases cleave fatty acid moieties from larger lipid molecules. Freeing arachidonic acid and other eicosanoids precursors from these molecules is the first step toward prostaglandin synthesis. Cryptococcus phospholipase (PLB1) is cell-wall-bound secreted, and deletion mutants deficient in this gene show cell wall defects (60) and reduced virulence in murine models of cryptococcosis (9). Noverr et al. (48) showed that this strain was deficient in production of certain prostaglandins and leukotrienes when given phospholipid precursors but not when given arachidonic acid directly. Owing to their observations that alveolar macrophages showed growth inhibition of the plb1 strain upon phagocytosis, they hypothesized that Plb1 is required for growth inside macrophages. Wright et al. (68) found that the prostaglandins synthesized during the macrophage-Cryptococcus interaction were derived from host (macrophage) arachidonic acid. Treatments that reduced production of phospholipases in Cryptococcus such as tipranavir have shown therapeutic effects on models of cryptococcosis (7). In summary, phospholipases may play an important role in prostaglandin metabolism in pathogenic fungi and should be further evaluated for their role in virulence and host-fungi interactions. As mentioned, Candida and Cryptococcus are not the only pathogenic fungi to produce eicosanoids. Genes identified in Aspergillus nidulans and A. fumigatus that are similar to COX genes also encode fatty acid oxygenases (63). The strains produced when these genes (ppoA, ppoB, and ppoC) were deleted showed deficits in PGE2 production. Further, a triple deletion mutant lacking all three of the PPO genes was resistant to oxidative stress and hypervirulent in a murine model of pulmonary aspergillosis. The authors hypothesize that the host immune system is further activated by the fungal prostaglandins in wild-type Aspergillus and that the triple deletion mutant is hypervirulent due to a lack of this activation. Eicosanoid production and signaling in pathogenic fungi has made interesting advances in recent years. Studies presented here have shown that these oxylipins are produced in many fungi that are pathogenic to humans, and that their production may be involved in the pathogenesis of these fungal infections. Future studies focused on signaling mechanisms in these fungi upon autocrine/paracrine exposure to self-produced eicosanoids would be interesting, especially studies examining more complex effects of these prostaglandins on other commensal microbes. Another aspect of the research that focuses on fungi-host interactions is the idea that the fungi are hijacking host lipids for their own processes. (This phenomenon is not unheard of; see for example research on Chlamydia trachomatis in Reference 65). The exact role that these prostaglandins play in pathogen virulence and host response would be crucial for the development of therapies that target these pathways. SPHINGOLIPID SIGNALING A class of lipids gaining new prominence in both mammalian and fungal research is the sphingolipid family. In mammals, these lipids, which contain a sphingoid backbone often bound to an acyl chain by an N-amide linkage, have shown numerous effects in processes such as apoptosis, stress responses, and cell proliferation. In recent years, the role of sphingolipids in pathogenic fungi, in terms of signaling, growth, and virulence, has become a rapidly growing field. Because many of these enzymes and products are structurally distinct from their mammalian counterparts, they have the potential for therapeutic targeting with minimal effects on the host. Advanced mathematical models have been created and experimentally tested using the sphingolipid biosynthetic pathway of C. neoformans and its relation to pathogenesis (19). One such enzyme that is unique to the fungal sphingolipid synthesis pathway is inositol phosphorylceramide (IPC) synthase. IPC synthase is responsible for the removal of an inositol-phosphate group from phosphatidylinositol and the transfer of that group to the terminal hydroxy group of phytoceramide. This reaction forms IPC and diacylglycerol (DAG) as a byproduct. Examinations of this enzyme in virulence of C. neoformans have shown that IPC synthase downregulation causes defects in melanin production (a known virulence factor in C. neoformans) and reduced growth inside alveolar macrophages. Both of these phenotypes may be responsible for the reduced virulence of strains with impaired IPC synthase (39, 40). The enzyme involved in the reverse process, inositol phosphosphingolipidphospholipase C (Isc1), is also involved in virulence. Isc1 removes the phosphorylinositol moiety from IPC, forming phytoceramide. When this enzyme is deleted, the resulting strain (Δisc1) is hypercapsulated and hypovirulent in immunocompromised mouse models of cryptococcosis and seems to be an obligate extracellular pathogen (59). The role of Isc1 likely revolves around macrophage interactions, as macrophage depletion results in the dissemination of the Δisc1 strain to the central nervous system. The production of DAG by IPC synthase is a major component of this enzyme’s role in virulence. DAG seems to regulate cryptococcal virulence factors in two ways. First, DAG is a component in the melanin synthesis pathway. DAG activates PKC in mammalian cells through binding to the C1 domain of PKC (28). Heung et al. (25) found that this activation occurs in C. neoformans as well, with DAG produced from IPC synthase reactions binding to the C1 domain of the cryptococcal PKC (Pkc1) (24) and causing an increase in kinase activity. Deletion of the specific C1 domain stops the DAG-dependent activation of Pkc1. This disruption of Pkc1 activity causes alterations in the cell wall, preventing the enzyme laccase (melanin-producing enzyme) from properly associating. It was shown here that this is the mechanism by which IPC synthase regulates melanin synthesis. Aside from the effects on laccase, the involvement of Pkc1 in cell wall integrity is another aspect of this pathway. Cell wall integrity is crucial for fungal growth and virulence, often becoming the target of antifungal therapies. Gerik et al. (21) deleted 10 genes in the Pkc1 signaling pathway and found that phosphatase Ppg1 and kinases Bck1 and Mkk2 are all required for proper cell wall integrity. Along with these defects were predictable impairments in melanin production and capsule synthesis, as well as protection from oxidative and nitrosative stresses (20). The second way that DAG production by IPC synthase regulates virulence factors is through the production of antiphagocytic protein (App1). App1 is not only involved in protection from phagocytosis, but also binds complement receptors CR2 and CR3 (61). The importance of App1 in virulence is clear from the phenotype of the deletion strain Δapp1, which is hypervirulent in complement-deficient mice but hypovirulent in mice with T-cell deficiencies (39). IPC synthase was indirectly involved in the transcription of APP1. DAG produced by IPC synthase activates the activating transcription factor 2 (Atf2), which binds to an ATF consensus sequence in the promoter region of APP1 (62). Upon binding, Atf2 upregulates the expression of App1, confirming another pathway by which DAG from IPC synthase regulates virulence of C. neoformans. Figure 2 summarizes the IPC regulation of virulence factors in C. neoformans. The future of sphingolipid pathway involvement in fungal disease is promising. Fluctuations in the pathway can be predicted with sophisticated mathematical models (19) and applied to pathogenesis. The IPC synthase-related pathways dominate the field of lipid signaling in pathogenic fungi to date, but other components of the fungal sphingolipid pathways are related to virulence. The deletion of the gene that encodes glucosylceramide synthase (GCS1) in C. neoformans yields a strain (Δgcs1) that is avirulent in inhalational mouse models of infection (54). This strain disseminates in a manner similar to Δisc1 when macrophages are depleted (32). Though the mechanism of this attenuated virulence is under examination (53), it is possible that lipid signaling plays a role in this phenotype as well. In C. albicans, enzymes known as Δ8-desaturases are responsible for the unique introduction of a double bond at the eighth carbon of the backbone of the fungal ceramide species. The deletion of this desaturase enzyme resulted in decreased hyphal growth and morphological alterations. Many enzymes in the sphingolipid pathway are still being characterized and deleted and may provide many more links between these lipids and virulence. CONCLUSIONS Lipid signaling in pathogenic fungi is a constantly growing field. New factors and fungal species are being studied constantly and the understanding of existing pathways is always improving. Aside from giving new insights into the biology of these important human pathogens, these pathways may represent a new class of antifungal treatments, if properly targeted. The discoveries presented here will likely be the foundation of many new advances in the years to come. This work was supported in part by the Burroughs Wellcome Fund; by grants AI56168, AI71142, and AI78493 (to M.D.P.) from the National Institute of Health; by RR17677 Project 2 (to M.D.P.) from the Centers of Biomedical Research Excellence Program of the National Center for Research Resources; by NIH grant C06 RR015455 from the Extramural Research Facilities Program of the National Center for Research Resources. R.R. is supported in part by a Medical Scientist Training Grant from the National Institutes of Health (GM08716) and by the Graduate Assistance in Areas of National Need (GAANN) training grant in Lipidology and New Technologies from the U.S. Department of Education. Dr. Maurizio Del Poeta is a Burroughs Wellcome New Investigator in Pathogenesis of Infectious Diseases. Farnesol (FA) isoprenoid lipid involved in quorum sensing in Candida and apoptosis in Aspergillus Eicosanoids a class of lipids derived from fatty acids like arachidonic acid that are involved in inflammation and immunomodulation in humans Sphingolipids a class of lipids including species with N-linked acylation of a sphingosine backbone and several variations of head groups Hyphae long branching filamentous cells in fungi and the major type of vegetative growth AC adenylate cyclase MAP mitogen-activated protein PGE2 prostaglandin E2 COX cyclooxygenase IPC inositol phosphorylceramide DAG diacylglycerol Pkc1 protein kinase C1 APP1 antiphagocytic protein 1 Figure 1 Known and hypothetical components of farnesol signaling in Candida albicans. Upon treatment with farnesol, Tup1 global transcriptional repressor associates with Tpp1, Nrg1, and Rfg1 to form a DNA binding complex. This complex inhibits gene expression of proteins involved in hyphal formation and may play a role in farnesol feedback pathways. Additionally, farnesol affects components of AC and MAP kinase signaling pathways. Farnesol may be through a common component, Ras, and inhibits these cascades that eventually lead to transcription of factors in hyphal formation. Abbreviations: AC, adenylate cyclase; cAMP, cyclic adenosine monophosphate; MAP, mitogen-activated protein; PKA, protein kinase A. Figure 2 Sphingolipid signaling in Cryptococcus neoformans. IPC synthase in C. neoformans produces DAG as a byproduct of IPC synthesis. DAG binds to the C1 domain of protein kinase A, which is important for cell wall integrity. This integrity is crucial for localization of laccase, the enzyme responsible for melanin synthesis. In addition, DAG also activates the transcription factor Atf2, which leads to transcription of App1. Both App1 and melanin are virulence factors in C. neoformans. Abbreviations: IPC, inositol phosphorylceramide; PI, phosphatidylinositol; App1, antiphagocytic protein 1; Pkc1, protein kinase C1 (with C1 domain); DAG, diacylglycerol. SUMMARY POINTS FA, a fungal quorum-sensing molecule, inhibits hyphal formation. FA exposure appears to alter gene expression in Candida and involve Ras, MAP kinase, and histidine kinase signaling pathways. In Aspergillus, FA treatment induces apoptosis and alters production of membrane lipids such as ergosterol. Eicosanoid production occurs in many pathogenic fungi. The prostaglandins are produced using host lipids and may be synthesized to manipulate the host immune system. Sphingolipid signaling affects transcription of antiphagocytic proteins and the production of melanin in Cryptococcus through the production of DAG by IPC synthase. DISCLOSURE STATEMENT The authors are not aware of any affiliations, memberships, funding, or financial holdings that might be perceived as affecting the objectivity of this review. LITERATURE CITED 1 Alem MA Douglas LJ Prostaglandin production during growth of Candida albicans biofilms J. Med. Microbiol 2005 54 1 5 15591247 2 Bordon AP Dias-Melicio LA Acorci MJ Biondo GA Fecchio D Prostaglandin E(2) production by high and low virulent strains of Paracoccidioides brasiliensis Mycopathologia 2007 163 129 135 17364218 3 Braun BR Johnson AD Control of filament formation in Candida albicans by the transcriptional repressor TUP1 Science 1997 277 105 109 9204892 4 Braun BR Johnson AD TUP1, CPH1 and EFG1 make independent contributions to filamentation in Candida albicans Genetics 2000 155 57 67 10790384 5 Braun BR Kadosh D Johnson AD NRG1, a repressor of filamentous growth in C. albicans, is down-regulated during filament induction EMBO J 2001 20 4753 4761 11532939 6 Cao YY Cao YB Xu Z Ying K Li Y cDNA microarray analysis of differential gene expression in Candida albicans biofilm exposed to farnesol Antimicrob. Agents Chemother 2005 49 584 589 15673737 7 Cenci E Francisci D Belfiori B Pierucci S Baldelli F Tipranavir exhibits different effects on opportunistic pathogenic fungi J. Infect 2008 56 58 64 17905438 8 Cho T Aoyama T Toyoda M Nakayama H Chibana H Kaminishi H Transcriptional changes in Candida albicans genes by both farnesol and high cell density at an early stage of morphogenesis in N-acetyl-D-glucosamine medium Nippon Ishinkin Gakkai. Zasshi 2007 48 159 167 9 Cox GM McDade HC Chen SC Tucker SC Gottfredsson M Extracellular phospholipase activity is a virulence factor for Cryptococcus neoformans Mol. Microbiol 2001 39 166 175 11123698 10 Cugini C Calfee MW Farrow JM 3rd Morales DK Pesci EC Hogan DA Farnesol, a common sesquiterpene, inhibits PQS production in Pseudomonas aeruginosa Mol. Microbiol 2007 65 896 906 17640272 11 Cushion MT Collins MS Linke MJ Biofilm formation by Pneumocystis Eukaryot. Cell 2008 8 197 206 18820078 12 Davis-Hanna A Piispanen AE Stateva LI Hogan DA Farnesol and dodecanol effects on the Candida albicans Ras1-cAMP signaling pathway and the regulation of morphogenesis Mol. Microbiol 2008 67 47 62 18078440 Showed farnesol role in Ras pathway. 13 Deva R Ciccoli R Schewe T Kock JL Nigam S Arachidonic acid stimulates cell growth and forms a novel oxygenated metabolite in Candida albicans Biochim. Biophys. Acta 2000 1486 299 311 10903481 14 Enjalbert B Whiteway M Release from quorum-sensing molecules triggers hyphal formation during Candida albicans resumption of growth Eukaryot. Cell 2005 4 1203 1210 16002646 15 Erb-Downward JR Huffnagle GB Cryptococcus neoformans produces authentic prostaglandin E2 without a cyclooxygenase Eukaryot. Cell 2007 6 346 350 17158733 16 Erb-Downward JR Noggle RM Williamson PR Huffnagle GB The role of laccase in prostaglandin production by Cryptococcus neoformans Mol. Microbiol 2008 68 1428 1437 18410494 17 Erb-Downward JR Noverr MC Characterization of prostaglandin E2 production by Candida albicans Infect. Immun 2007 75 3498 3505 17470538 18 Feng Q Summers E Guo B Fink G Ras signaling is required for serum-induced hyphal differentiation in Candida albicans J. Bacteriol 1999 181 6339 6346 10515923 19 Garcia J Shea J Alvarez-Vasquez F Qureshi A Luberto C Mathematical modeling of pathogenicity of Cryptococcus neoformans Mol. Syst. Biol 2008 4 183 18414484 20 Gerik KJ Bhimireddy SR Ryerse JS Specht CA Lodge JK PKC1 is essential for protection against both oxidative and nitrosative stresses, cell integrity, and normal manifestation of virulence factors in the pathogenic fungus Cryptococcus neoformans Eukaryot. Cell 2008 7 1685 1698 18689526 21 Gerik KJ Donlin MJ Soto CE Banks AM Banks IR Cell wall integrity is dependent on the PKC1 signal transduction pathway in Cryptococcus neoformans Mol. Microbiol 2005 58 393 408 16194228 Identifies PKC as crucial for cell wall integrity. 22 Haas-Stapleton EJ Lu Y Hong S Arita M Favoreto S Candida albicans modulates host defense by biosynthesizing the proresolving mediator resolvin E1 PLoS ONE 2007 2 e1316 18091990 23 Henriques M Martins M Azeredo J Oliveira R Effect of farnesol on Candida dubliniensis morphogenesis Lett. Appl. Microbiol 2007 44 199 205 17257261 24 Heung LJ Kaiser AE Luberto C Del Poeta M The role and mechanism of diacylglycerol-protein kinase C1 signaling in melanogenesis by Cryptococcus neoformans J. Biol. Chem 2005 280 28547 28555 15946943 Describes the mechanism of IPC-melanin-virulence connection through PKC and laccase. 25 Heung LJ Luberto C Plowden A Hannun YA Del Poeta M The sphingolipid pathway regulates Pkc1 through the formation of diacylglycerol in Cryptococcus neoformans J. Biol. Chem 2004 279 21144 21153 15014071 26 Hornby JM Jensen EC Lisec AD Tasto JJ Jahnke B Quorum sensing in the dimorphic fungus Candida albicans is mediated by farnesol Appl. Environ. Microbiol 2001 67 2982 2992 11425711 First identifies farnesol as a quorum sensing molecule. 27 Hornby JM Nickerson KW Enhanced production of farnesol by Candida albicans treated with four azoles Antimicrob. Agents Chemother 2004 48 2305 2307 15155241 28 Hurley JH Newton AC Parker PJ Blumberg PM Nishizuka Y Taxonomy and function of C1 protein kinase C homology domains Protein Sci 1997 6 477 480 9041654 29 Kadosh D Johnson AD Rfg1, a protein related to the Saccharomyces cerevisiae hypoxic regulator Rox1, controls filamentous growth and virulence in Candida albicans Mol. Cell Biol 2001 21 2496 2505 11259598 30 Kaneko A Umeyama T Utena-Abe Y Yamagoe S Niimi M Uehara Y Tcc1p, a novel protein containing the tetratricopeptide repeat motif, interacts with Tup1p to regulate morphological transition and virulence in Candida albicans Eukaryot. Cell 2006 5 1894 1905 16998076 31 Kebaara BW Langford ML Navarathna DH Dumitru R Nickerson KW Atkin AL Candida albicans Tup1 is involved in farnesol-mediated inhibition of filamentous-growth induction Eukaryot. Cell 2008 7 980 987 18424510 Proved the mechanistic connection with farnesol and Tup1. 32 Kechichian TB Shea J Del Poeta M Depletion of alveolar macrophages decreases the dissemination of a glucosylceramide-deficient mutant of Cryptococcus neoformans in immunodeficient mice Infect. Immun 2007 75 4792 4798 17664261 33 Khalaf RA Zitomer RS The DNA binding protein Rfg1 is a repressor of filamentation in Candida albicans Genetics 2001 157 1503 1512 11290707 34 Kruppa M Krom BP Chauhan N Bambach AV Cihlar RL Calderone RA The two-component signal transduction protein Chk1p regulates quorum sensing in Candida albicans Eukaryot. Cell 2004 3 1062 1065 15302838 35 Leberer E Harcus D Dignard D Johnson L Ushinsky S Ras links cellular morphogenesis to virulence by regulation of the MAP kinase and cAMP signaling pathways in the pathogenic fungus Candida albicans Mol. Microbiol 2001 42 673 687 11722734 36 Levitin A Whiteway M The effect of prostaglandin E2 on transcriptional responses of Candida albicans Microbiol. Res 2007 162 201 210 17433645 37 Lorek J Poggeler S Weide MR Breves R Bockmuhl DP Influence of farnesol on the morphogenesis of Aspergillus niger J. Basic Microbiol 2008 48 99 103 18383232 38 Lorenz MC Bender JA Fink GR Transcriptional response of Candida albicans upon internalization by macrophages Eukaryot. Cell 2004 3 1076 1087 15470236 39 Luberto C Martinez-Marino B Taraskiewicz D Bolanos B Chitano P Identification of App1 as a regulator of phagocytosis and virulence of Cryptococcus neoformans J. Clin. Investig 2003 112 1080 1094 14523045 40 Luberto C Toffaletti DL Wills EA Tucker SC Casadevall A Roles for inositol-phosphoryl ceramide synthase 1 (IPC1) in pathogenesis of C. neoformans Genes Dev 2001 15 201 212 11157776 41 Martin SW Konopka JB Lipid raft polarization contributes to hyphal growth in Candida albicans Eukaryot. Cell 2004 3 675 684 15189988 42 McAlester G O’Gara F Morrissey JP Signal-mediated interactions between Pseudomonas aeruginosa and Candida albicans J. Med. Microbiol 2008 57 563 569 18436588 43 Mosel DD Dumitru R Hornby JM Atkin AL Nickerson KW Farnesol concentrations required to block germ tube formation in Candida albicans in the presence and absence of serum Appl. Environ. Microbiol 2005 71 4938 4940 16085901 44 Murad AM Leng P Straffon M Wishart J Macaskill S NRG1 represses yeast-hypha morphogenesis and hypha-specific gene expression in Candida albicans EMBO J 2001 20 4742 4752 11532938 45 Navarathna DH Hornby JM Hoerrmann N Parkhurst AM Duhamel GE Nickerson KW Enhanced pathogenicity of Candida albicans pretreated with subinhibitory concentrations of fluconazole in a mouse model of disseminated candidiasis J. Antimicrob. Chemother 2005 56 1156 1159 16239285 46 Navarathna DH Hornby JM Krishnan N Parkhurst A Duhamel GE Nickerson KW Effect of farnesol on a mouse model of systemic candidiasis, determined by use of a DPP3 knockout mutant of Candida albicans Infect. Immun 2007 75 1609 1618 17283095 47 Navarathna DH Nickerson KW Duhamel GE Jerrels TR Petro TM Exogenous farnesol interferes with the normal progression of cytokine expression during candidiasis in a mouse model Infect. Immun 2007 75 4006 4011 17517874 48 Noverr MC Cox GM Perfect JR Huffnagle GB Role of PLB1 in pulmonary inflammation and cryptococcal eicosanoid production Infect. Immun 2003 71 1538 1547 12595473 49 Noverr MC Phare SM Toews GB Coffey MJ Huffnagle GB Pathogenic yeasts Cryptococcus neoformans and Candida albicans produce immunomodulatory prostaglandins Infect. Immun 2001 69 2957 2963 11292712 First reports prostaglandin synthesis in Cryptococcus and Candida. 50 Noverr MC Toews GB Huffnagle GB Production of prostaglandins and leukotrienes by pathogenic fungi Infect. Immun 2002 70 400 402 11748207 51 Ozpolat B Akar U Mehta K Lopez-Berestein G PKC delta and tissue transglutaminase are novel inhibitors of autophagy in pancreatic cancer cells Autophagy 2007 3 480 483 17507797 52 Perumal P Mekala S Chaffin WL Role for cell density in antifungal drug resistance in Candida albicans biofilms Antimicrob. Agents Chemother 2007 51 2454 2463 17502416 53 Rhome R McQuiston T Kechichian T Bielawska A Hennig M Biosynthesis and immunogenicity of glucosylceramide in Cryptococcus neoformans and other human pathogens Eukaryot. Cell 2007 6 1715 1726 17693597 54 Rittershaus PC Kechichian TB Allegood JC Merrill AH Jr Hennig M Glucosylceramide synthase is an essential regulator of pathogenicity of Cryptococcus neoformans J. Clin. Investig 2006 116 1651 1659 16741577 55 Rossignol T Logue ME Reynolds K Grenon M Lowndes NF Butler G Transcriptional response of Candida parapsilosis following exposure to farnesol Antimicrob. Agents Chemother 2007 51 2304 2312 17684006 56 Sato T Watanabe T Mikami T Matsumoto T Farnesol, a morphogenetic autoregulatory substance in the dimorphic fungus Candida albicans, inhibits hyphae growth through suppression of a mitogen-activated protein kinase cascade Biol. Pharm. Bull 2004 27 751 752 15133261 57 Semighini CP Hornby JM Dumitru R Nickerson KW Harris SD Farnesol-induced apoptosis in Aspergillus nidulans reveals a possible mechanism for antagonistic interactions between fungi Mol. Microbiol 2006 59 753 764 16420349 Showed farnesol induces apoptosis in Aspergillus. 58 Shchepin R Navarathna DH Dumitru R Lippold S Nickerson KW Dussault PH Influence of heterocyclic and oxime-containing farnesol analogs on quorum sensing and pathogenicity in Candida albicans Bioorg. Med. Chem 2008 16 1842 1848 18037299 59 Shea JM Kechichian TB Luberto C Del Poeta M The cryptococcal enzyme inositol phosphosphingolipid-phospholipase C confers resistance to the antifungal effects of macrophages and promotes fungal dissemination to the central nervous system Infect. Immun 2006 74 5977 5988 16988277 60 Siafakas AR Sorrell TC Wright LC Wilson C Larsen M Cell wall-linked cryptococcal phospholipase B1 is a source of secreted enzyme and a determinant of cell wall integrity J. Biol. Chem 2007 282 37508 37514 17947228 61 Stano P Williams V Villani M Cymbalyuk ES Qureshi A App1: an antiphagocytic protein that binds to complement receptors 3 and 2 J. Immunol 2009 182 84 91 19109138 62 Tommasino N Villani M Qureshi A Henry J Luberto C Del Poeta M Atf2 transcription factor binds to the APP1 promoter in Cryptococcus neoformans: stimulatory effect of diacylglycerol Eukaryot. Cell 2008 7 294 301 18083832 Showed link between DAG and App1 expression through Atf2. 63 Tsitsigiannis DI Bok JW Andes D Nielsen KF Frisvad JC Keller NP Aspergillus cyclooxygenase-like enzymes are associated with prostaglandin production and virulence Infect. Immun 2005 73 4548 4559 16040966 64 Tsuchimori N Sharkey LL Fonzi WA French SW Edwards JE Jr Filler SG Reduced virulence of HWP1-deficient mutants of Candida albicans and their interactions with host cells Infect. Immun 2000 68 1997 2002 10722594 65 van Ooij C Kalman L van Ijzendoorn Nishijima M Hanada K Host cell-derived sphingolipids are required for the intracellular growth of Chlamydia trachomatis Cell Microbiol 2000 2 627 637 11207614 66 Weber K Sohr R Schulz B Fleischhacker M Ruhnke M Secretion of E,E-farnesol and biofilm formation in eight different Candida species Antimicrob. Agents Chemother 2008 52 1859 1861 18332168 67 Westwater C Balish E Schofield DA Candida albicans-conditioned medium protects yeast cells from oxidative stress: a possible link between quorum sensing and oxidative stress resistance Eukaryot. Cell 2005 4 1654 1661 16215173 68 Wright LC Santangelo RM Ganendren R Payne J Djordjevic JT Sorrell TC Cryptococcal lipid metabolism: Phospholipase B1 is implicated in transcellular metabolism of macrophage-derived lipids Eukaryot. Cell 2007 6 37 47 17099081 69 Zheng X Wang Y Hgc1, a novel hypha-specific G1 cyclin-related protein regulates Candida albicans hyphal morphogenesis EMBO J 2004 23 1845 1856 15071502
PMC005xxxxxx/PMC5125069.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 9808648 21055 Prostaglandins Other Lipid Mediat Prostaglandins Other Lipid Mediat. Prostaglandins & other lipid mediators 1098-8823 22108026 5125069 10.1016/j.prostaglandins.2011.11.001 NIHMS831088 Article The presence of 3-hydroxy oxylipins in pathogenic microbes Sebolai Olihile M. a* Pohl Carolina H. a Kock Lodewyk J.F. a Chaturvedi Vishnu bc del Poeta Maurizio d a Department of Microbial, Biochemical and Food Biotechnology, University of the Free State, 205 Nelson Mandela Drive, Park West, Bloemfontein 9301, South Africa b Wadsworth Center, New York State Department of Health, 120 New Scotland Avenue, Albany, NY 12208, United States c Department of Biomedical Sciences, University at Albany School of Public Health, State University of New York, 1400 Washington Avenue, Albany, NY 12222, United States d Department of Biochemistry and Molecular Biology, Medical University of South Carolina, 173 Ashley Avenue, MSC 509, Charleston, SC 29425, United States * Corresponding author. Tel.: +27 51401 2004; fax: +27 86506 1588. [email protected] (O.M. Sebolai) 19 11 2016 11 11 2011 1 2012 28 11 2016 97 1-2 1721 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. There is a sufficient body of work documenting the distribution of 3-hydroxy oxylipins in microbes. However, there is limited information on the role of these compounds in microbial pathogenesis. When derived from mammalian cells, these compounds regulate patho-biological processes, thus an understanding of 3-hydroxy oxylipin function and metabolism could prove important in shedding light on how these compounds mediate cellular pathology and physiology. This could present 3-hydroxy oxylipin biosynthetic pathways as targets for drug development. In this minireview, we interrogate the relevant yeast and bacterial 3-hydroxy oxylipin literature in order to appreciate how these compounds may influence the inflammatory response leading to disease development. Bacteria 3-Hydroxy oxylipins Inflammation Pathogenesis Yeast 1. Biochemistry: definition, occurence and biosynthesis The word “oxylipin” describes a group of secondary metabolites that originate from the oxidation or further conversion of polyunsaturated fatty acids [1]. These lipid-based molecules are pivotal signal molecules documented to act in a hormone-like manner where they mediate a number of complex biological processes across a number of life domains. In terrestrial higher plants, oxylipins play a role in host defence mechanisms against pathogens and pests [2]. In mammalian cells, these molecules regulate cellular homeostasis and immune responses [3–5]. In marine algae, it is hypothesised that they may be involved in defence mechanisms [6], while in bacteria and fungi they may regulate virulence, biofilm formation via quorum-sensing mechanism [7,8], and play a role in sexual and asexual development [9]. Oxylipins constitute among others the eicosanoids and hydroxy oxylipins [9]. Given the broad scope of oxylipins, their distribution and function, this minireview is dedicated to hydroxy oxylipins and in particular, 3-hydroxy oxylipins in fungi and bacteria. The reader is referred to excellent reviews paying special attention to other oxylipins and lipid mediators that regulate important biological processes in cellular physiology and pathology [1,2,9–13]. 3-Hydroxy oxylipins (3-OH oxylipins) are fatty acid-based molecules characterised by a hydroxyl group on the beta-carbon atom, from the carboxylic group (Fig. 1). The carbon chain of 3-OH oxylipins may be branched and may vary considerably in length as well as in the degree of desaturation [9,14,15]. 3-Hydroxy oxylipins are also widely distributed in nature, occurring in mammals, bacteria and yeasts, including medically important pathogens [7,16–20]. In mammalian systems, production of 3-hydroxy oxylipins is mainly attributed to fatty acid oxidation disorders. Accumulation of these molecules in the blood is regarded as a major metabolic indicator of long chain hydroxyacyl coenzyme A dehydrogenase (LCHAD) deficiency in newborns and patients with liver failure [20]. The biosynthetic pathways for 3-OH oxylipins vary and although some remain poorly described, three generally accepted enzymatic routes have been reported (Fig. 2): fatty acid synthase (FAS) enzyme-system [21,22]. Here, the NADPH-dependent beta-ketoacyl-ACP reductase carries out the reduction of beta-ketoacyl-ACP to beta-hydroxyacyl-ACP, an enzymatic pattern similar to mitochondrial beta oxidation, however, incomplete [23]. The oxygen of the hydroxyl group inserted in the fatty acid chain originates from water. In this case, the produced 3-D hydroxyacyl-CoA enantiomer, cannot be, or is poorly, metabolised further by 3-hydroxyacyl-CoA dehydrogenase [24], and consequently accumulates inside the mitochondria. This compound is then excreted as a 3-D hydroxy oxylipin [25], direct hydroxylation of the fatty acid via a cytochrome P450 enzyme [26,27], with the oxygen molecule originating from the air. 2. Patho-biological functions of microbial 3-hydroxy oxylipins Microbial cell walls perform two critical roles in immunity, namely to provide protection from the extracellular environment, and interaction with the environment [28]. 3-Hydroxy oxylipins have been reported to be closely associated with cell walls of pathogens [7,16,18,19]. In bacteria, they are attached or bound to cell wall components, whereas in yeasts, they are mainly in a free form - coating or deposited on cell wall surfaces. Literature suggests that during infection, microbial cell wall components mediate key processes that could modulate the immune response leading to development of disease [29–31]. This minireview pays special attention to the role of 3-OH oxylipins in modulating the inflammatory response. Inflammation, usually a result of cytokine production, is a complex biological response that attempts to clear and heal vascular tissue of infection or other forms of damage [32,33]. Disease outcome may determine a shift in the balance maintained by both pro-inflammatory and anti-inflammatory cytokines. 2.1. Bacterial 3-hydroxy oxylipins 3-Hydroxy oxylipins occur as unique structural components of the sepsis-causing endotoxin (lipopolysaccharide layer; LPS), which is characteristic of Gram-negative bacteria [34]. The 3-OH oxylipin-containing Lipid A fraction is documented to be responsible for the toxic and immuno-modulating properties of LPS [30,35,36]. Upon shedding, the endotoxin triggers an innate immune response characterised by cytokine production. Here, the endotoxin is first recognised by receptor protein i.e. cluster of differentiation (CD)-14 and in turn, presented to toll-like receptor (TLR)-4 on surfaces of innate cells resulting in intracellular signalling [37,38]. This leads to the production of pro-inflammatory cytokines, i.e. interleukin (IL-) 1 and tumour necrosis factor alpha (TNF-alpha), and activation of mononuclear cells. These cytokines can then induce synthesis of mediator molecules viz. cyclo-oxygenase 2, phospholipase A2 and nitric oxide (NO) synthase - which up regulate inflammation [32,39]. Subsequently, these cytokines together with mediator molecules, acting through specific G-protein-coupled receptors, promote inflammation, causing widespread endothelial injury and platelet activation [40,41], and at high endotoxin levels, septic shock can be induced [41]. Interestingly, 3-hydroxy oxylipins are used as biomarkers for estimating the amount of endotoxins and Gram-negative bacteria in atmospheric bioaerosols [42]. Inhalation of bioaerosols-containing 3-hydroxy oxylipins i.e. entotoxin, can also initiate infectious processes that elicit allergenic and immunological responses [43,44]. Peden et al. [45] reported that a nasal challenge with LPS causes an eosinophil influx in nasal airways of atopic subjects, suggesting exposure may increase allergen-induced bronchial inflammation in asthmatics [43]. 3-OH oxylipins from Porphyromonas gingivalis constitute a major component of bioactive lipids reported to potentiate interleukin-1b-mediated secretory response in gingival fibroblasts. This organism is thought to be a major periodontal pathogen associated with inflammatory periodontal disease in adults [46]. 3-Hydroxy oxylipins also occur as complex molecules such as mycolic acid, which are 3-OH oxylipins with long alpha alkyl branched chains [22]. Here too, 3-OH oxylipins are associated with pathogenicity of Mycobacterium tuberculosis, the causative agent of tuberculosis. These compounds confer the pathogen with the ability to grow within macrophages and to avoid detection [47]. When this bacterium is lysed, mycolic acid is released from the cell wall. Regarded as pathogen-associated molecular patterns (PAMP), the released mycolic acid may then invoke an immune response [48–51]. Most of the damage in the lungs during tuberculosis is thought to be due to the up regulated inflammatory response. Here, it is hypothesised that IL-1, TNF-alpha and NO may induce oxidative damage to mitochondria by inhibiting the electron transport chain [41]. This inhibitory action results in less cellular energy and dysoxia. 2.2. Yeast 3-hydroxy oxylipins The lipopolysaccharide layer is not limited to bacteria. The presence of this cell wall component has been reported in a medically important higher basidiomycete, Antrodia camphorata [52]. Interestingly, this fungal LPS reverses immuno-regulating properties exerted by bacterial LPS. Nigam and co-workers were the first to provide evidence concerning the biological function of 3-OH oxylipins in mammalian cells [53]. In their study, 3-OH oxylipins were observed to act as a strong chemotactic agent - the potency of which is comparable with those of leukotriene B4 or fMet-Leu-Phe. In addition, 3-OH oxylipins affected signal transduction processes in human neutrophils and tumour cells in multiple ways, possibly via a G-protein receptor. Fluorescence studies conducted using a specific immunological probe against 3-OH oxylipins, revealed these compounds to be deposited on hyphal cell surfaces of the pathogen, Candida albicans, the causative agent of candidiasis [16,54,55]. In 2005, Ciccoli and co-workers elucidated a novel acetylsalicylic acid (ASA; aspirin) sensitive patho-biological process in C. albicans [17]. They found that this yeast converts arachidonic acid, released from infected host cells, to a 3-OH oxylipin i.e. 3-hydroxy eicosatetraenoic acid (3-HETE) via incomplete mitochondrial beta-oxidation. This 3-OH oxylipin, which is stereo-chemically similar to arachidonic acid, then acts as substrate for the host cyclooxygenase-2 (COX-2), leading to the production of potent pro-inflammatory 3-OH prostaglandin E2 (3-OH-PG E2) (Fig. 3). This novel compound could signal the expression of IL-6 gene, via the EP 3 receptor (PGE2 receptor 3) and raise cAMP levels via the EP 4 receptor. These results led this group of researchers to conclude that, these compounds have strong biological activities similar to and in some cases even more potent than those of the normally produced mammalian eicosanoids. This organism can also employ its own endogenously produced PG E2 to mediate pathogenesis [12,56]. Recently, the Nigam group also showed that 3-OH oxylipins can effect quorum sensing in C. albicans [8], a function used by microorganisms to measure population density and to regulate pathogenicity [8,57]. This group demonstrated that this yeast utilises 3-OH oxylipins, i.e. 3-OH-14:2 produced from 18:2, as a signal for expression of genes responsible for accelerating cell morphogenesis at a certain population density. Bio-prospecting studies into the presence of these compounds in other pathogenic yeasts led to the discovery of 3-OH oxylipins in capsules of Cryptococcus neoformans [18]. The cryptococcal capsule can inhibit phagocytosis and influence cytokine production, functions crucial for mounting an efficient immune response [58–60]. The study by Sebolai and co-workers [18,25] revealed a novel release mechanism of these compounds as “oily-droplets” into the surrounding environment. The release mechanism involved the participation of cell wall components namely, capsule and spiky capsular protuberances, as well mitochondria. This release mechanism was inhibited by ASA in a dose dependent manner. However, the function of these compounds upon release remains unknown. Could they also act as virulence factors alone or in association with the glucuronoxylomannans (GXM)? It has been established that GXM induces inflammation by activating TLRs [61,62]. 3. Concluding remarks and perspectives Over the years, microbial lipids have been shown to have bioactive functions mediating a number of cellular processes [10–13,63]. Most of our knowledge on 3-OH oxylipins stems from extensive studies conducted in non-pathogenic yeasts and studies focusing on bacterial endotoxins [9,35,64]. In yeast studies, the biological functions of these compounds were defined based on their role in facilitating cell aggregation, possibly for protection purposes [65], or for facilitating spore release from asci following sexual reproduction [9]. In addition, these molecules act as “toxins” secreted by lactic acid bacteria, where they are employed to appropriate environmental advantage against yeasts and molds in the bio-preserve of fermentation products [66]. As analysed in this minireview, we now can appreciate the role of 3-OH oxylipins, mainly associated with cell wall components or surfaces of medically important pathogens, as signal molecules, triggering inflammatory responses. The role of mitochondria in cancer development and programmed-cell death is well established [67–70]. As reported in literature, microbial mitochondria “house” enzymatic pathways that catalyse the biosynthesis of patho-biologically active 3-OH oxylipins [17,21,23,24]. This exposes mitochondria as targets for controlling biosynthesis and effects of 3-OH oxylipins, hence further highlighting the critical role of this organelle in cellular pathogenesis. Since the study by Ciccoli et al. [17] demonstrated that during infection, mammalian cyclooxygenases can serve as additional enzymes catalysing synthesis of 3-OH oxylipins, further contributing to inflammation, it will be interesting to determine if infected host cell’s mitochondria could serve as another 3-OH oxylipin production site particularly, in persons without fatty acid oxidation disorders [20]. Other questions that need to be answered are, could the actions of aspirin, a known anti-mitochondrial and anti-fungal [25,71–76], now be extended to control bacterial infections caused by the highly aerobic M. tuberculosis? Can aspirin inhibit the production of mycolic acids based on the structural similarities between aspirin and acyl-portions of the FAS biosynthetic pathway? In answering these questions, consideration should be taken in order to realise efficacy against pathogens without adversely affecting human mitochondria. In higher eukaryotic cell systems such as in humans, mitochondria are responsible for generation of cellular energy under strictly aerobic conditions [77,78], hence colonisation of lungs by highly aerobic pathogens such as M. tuberculosis and C. neoformans. During pulmonary cryptococcosis, cryptococcal phospholipase can degrade the phospholipid component of lung surfactants leading to increased inflammation via the production of eicosanoids [12]. And unlike in some lower eukaryotic cell systems, humans cannot switch to fermentation when oxygen is depleted thus mitochondrial damage can prove deadly. Could the change in form and complexity of 3-OH oxylipins from bacteria (bound or attached to cell wall components) to yeasts (in a “free” form and deposited onto cell walls) be indicative of an evolutionary development? According to the endosymbiotic theory, it is proposed that mitochondria are descendents of ancient bacteria [79]. Therefore, is it possible that the present day mitochondria, ancestral descendant of ancient sepsis-causing bacteria through this theory, adapted and found a novel way to shed virulence factors i.e. 3-OH oxylipins, from a safe or protected environment within eukaryotic cells? Though the theory is controversial, there is molecular evidence including phylogeny studies, in support of the theory [80]. Therefore, it would be interesting to determine if genes encoding enzymes involved in the biosynthesis of mitochondrially-produced 3-OH oxylipins are related or even conserved in both yeasts and bacteria. Our apologies to those authors whose work we may have overlooked in this minireview. The authors would also like to thank anonymous reviewers for their scholarly contributions. Our work on 3-OH oxylipins has been supported by the National Research Foundation of South Africa. Maurizio Del Poeta (MDP) is supported by National Institute of Health (NIH) awards AI056168, AI071142, AI078493, and AI087541. MDP is a Burroughs Welcome New Investigator in the Pathogenesis of Infectious Diseases. Fig. 1 The chemical structures of a typical 3-hydroxy oxylipin. (a) Depicts the R-enantiomer while; (b) depicts the S-enantiomer. Obtained with permission from Kock et al. [64]. Fig. 2 Biosynthetic pathways catalysing 3-hydroxy oxylipin production. (a) and (b) Depict enzymatic route similar to fatty acid synthase and beta oxidation, respectively while; (c) depicts direct hydroxylation of a fatty acid molecule. Fig. 3 A diagram showing the formation of potent inflammatory 3-hydroxy prostaglandins in host cells from 3-HETE produced via incomplete beta-oxidation from host-released arachidonic acid (AA) by the yeast Candida albicans. ASA, acetyl-salicylic acid; COX-2, cyclooxygenase-2; 3(R)-HETE, 3(R) hydroxyeicosatetraenoic acid; 3-OH PGs, 3-hydroxy prostaglandins. Obtained with permission from Kock et al. [64]. Conflict of interest There are no competing interests. 1 Brodhun F Feussner I Oxylipins in fungi FEBS J 2011 278 1047 63 21281447 2 Tsitsigiannis DI Keller NP Oxylipins as developmental and host-fungal communication signals Trends Microbiol 2007 15 109 18 17276068 3 Samuelsson B Leukotrienes: mediators of immediate hypersensitivity reactions and inflammation Science 1983 220 568 75 6301011 4 Needleman P Turk J Jakschik BA Morrison AR Lefkowith JB Arachidonic acid metabolism Annu Rev Biochem 1986 55 69 72 3017195 5 Curtis-Prior P The eicosanoids West Sussex John Wiley and Sons Ltd 2004 6 Bouarab K Adas F Gaquerel E Kloareg B Salaün J-P Potin P The innate immunity of a marine red algae involves oxylipins from both the eicosanoids and octadecanoid pathways Plant Physiol 2004 135 1838 48 15247395 7 Raetz CRH Whitfield C Lipopolysaccharide endotoxins Annu Rev Biochem 2002 71 635 40 12045108 8 Nigam S Ciccoli R Ivanov I Sczepanski M Deva R On mechanism of quorum sensing in Candida albicans by 3(R)-hydroxy-tetradecaenoic acid Curr Microbiol 2010 62 55 63 20509029 9 Kock JLF Strauss CJ Pohl CH Nigam S The distribution of 3-hydroxy oxylipins in fungi Prostaglandins Other Lipid Mediat 2003 71 85 96 14518554 10 Shea JM Del Poeta M Lipid signaling in pathogenic fungi Curr Opin Microbiol 2006 9 352 8 16798065 11 Singh A Del Poeta M Lipid signaling in pathogenic fungi Cell Microbiol 2011 13 177 85 21091925 12 Noverr MC Erb-Downward JR Huffnagle GB Production of eicosanoids and other oxylipins by pathogenic eukaryotic microbes Clin Microbiol Rev 2003 16 517 33 12857780 13 Erb-Downward JR Huffnagle GB Role of oxylipins and other lipid mediators in fungal pathogenesis Future Microbiol 2006 192 219 27 14 Bhatt RK Falck JR Nigam S Enantiospecific total synthesis of a novel arachidonic acid metabolite 3-hydroxyeicosatetraenoic acid Tetrahedron Lett 1998 39 249 52 15 Groza NV Ivanov IV Romanov SG Myagkova GI Nigam S A novel synthesis of 3(R)-HETE, 3(R)-HTDE and enzymatic synthesis of 3(R), 15(S)-DiHETE Tetrahedron 2002 58 9859 63 16 Deva R Ciccoli R Schewe T Kock JLF Nigam S Arachidonic acid stimulates cell growth and forms a novel oxygenated metabolite in Candida albicans Biochim Biophys Acta 2000 1486 299 311 10903481 17 Ciccoli R Sahi S Singh S Oxygenation by cyclooxygenase-2 (COX-2) of 3-hydroxyeicosatetraenoic acid (3-HETE), a fungal mimetic of arachidonic acid, produces a cascade of novel bioactive 3-hydroxy-eicosanoids Biochem J 2005 390 737 47 15869467 18 Sebolai OM Pohl CH Botes PJ 3-Hydroxy fatty acids found in capsules of Cryptococcus neoformans Can J Microbiol 2007 53 809 12 17668042 19 Marrakchi H Bardou F Lancèlle M-A Daffè M A comprehensive overview of mycolic acid structure and biosynthesis Daffè M Reyrat J-M The mycobacterial cell envelope Washington ASM Press 2008 41 62 20 Jones PM Bennett MJ Clinical applications of 3-hydroxy fatty acids analysis by gas chromatography–mass spectrometry Biochim Biophys Acta in press 21 Hiltunen JK Okubo F Kursu KJ Autio KJ Kastaniotis AJ Mitochondrial fatty acid synthesis and maintenance of respiratory competent mitochondria in yeast Biochem Soc Trans 2005 33 1162 5 16246072 22 Takayama K Wang C Besra G Pathway to synthesis and processing of mycolic acids in Mycobacterium tuberculosis Clin Microbiol Rev 2005 18 81 91 15653820 23 Venter P Kock JLF Kumar S Production of 3R-hydroxy-polyenoic fatty acids by the yeast Dipodascopsis uninucleata UOFS-Y128 Lipids 1997 32 1277 83 9438238 24 Finnerty WR Microbial lipid metabolism Ratledge C Wilkinson SG Microbial lipids 2 London Academic Press 1989 525 58 25 Sebolai OM Pohl CH Botes PJ van Wyk PWJ Kock JLF The influence of acetyl-salicylic acid on oxylipin migration in Cryptococcus neoformans var. neoformans UOFS Y-1378 Can J Microbiol 2008 54 91 6 18388977 26 Harayama S Kok M Neidle EL Functional and evolutionary relationships among diverse oxygenases Annu Rev Microbiol 1992 46 565 71 1444267 27 van Dyk MS Kock JLF Botha A Hydroxy long-chain fatty acids in fungi World J Microbiol 1994 10 495 504 28 Pommerville JC Cell structure and function in the bacteria and archaea Pommerville JC Alcamo’s fundamentals of microbiology: body systems edition Sudbury Jones and Bartlett Publishers 2010 98 128 29 Barry CE III Lee RE Mdluli K Mycolic acids: structure, biosynthesis, and physiological functions Prog Lipid Res 1998 37 143 79 9829124 30 Seydel U Oikawa M Fukase K Kusumoto S Brandenburg K Intrinsic conformation of lipid A is responsible for agonistics and antagonistic activity Eur J Biochem 2000 276 3032 9 31 Bose I Reese AJ Ory JJ Janbon G Doering TL A yeast under cover: the capsule of Cryptococcus neoformans Eukaryot Cell 2003 2 655 63 12912884 32 Dinarello CA Pro-inflammatory cytokines Chest 2000 118 503 8 10936147 33 Elenkov IJ Iezzoni DG Daly A Harris AG Chrousos GP Cytokine dysregulation inflammation and well-being Neuroimmunomodulation 2005 12 255 69 16166805 34 Alexander C Rietschel ET Bacterial lipopolysaccharides and innate immunity J Endotoxin Res 2001 7 167 72 11581570 35 Rietschel ET Kirikae T Schade FU Bacterial endotoxin: molecular relationships of structure to activity and function FASEB J 1994 8 217 25 8119492 36 Silipo A de Castro C Lanzetta R Parrilli M Molinaro A Lipopolysaccharides König H Claus H Varma A Prokaryotic cell wall compounds: structure and biochemistry Berlin Springer-Verlag 2010 133 53 37 Kitchens RL Role of CD14 in cellular recognition of bacterial lipopolysaccharides Chem Immunol 2000 74 61 82 10608082 38 Triantafilou M Triantafilou K The dynamics of LPS recognition: complex orchestration of multiple receptors J Endotoxin Res 2005 11 5 11 15826372 39 Hopkins SJ The pathophysiological role of cytokines Legal Med 2003 5 S45 57 12935551 40 Pober JS Endothelial activation: intracellular signaling pathways Arthritis Res 2002 4 S109 16 12110129 41 Annane D Bellissant E Cavaillon J Septic shock Lancet 2005 365 63 78 15639681 42 Lee AKY Chan CK Fang M Lau APS The 3-hydroxy fatty acids as biomarkers for quantification and characterization of endotoxins and Gram-negative bacteria in atmospheric aerosols in Hong Kong Atmos Environ 2004 6307 17 43 Sohy C Pons F Casset A Low-dose endotoxin in allergic asthmatics: effect on bronchial and inflammatory responses to cat allergen Clin Exp Allergy 2006 36 795 802 16776681 44 Poole JA Dooley GP Saito R Muramic acid, endotoxin, 3-hydroxy fatty acids, and ergosterol content explain monocyte and epithelial cell inflammatory responses to agricultural dusts J Toxicol Environ Health A 2010 73 684 90 20391112 45 Peden DB Tucker K Murphy P Eosinophil influx to the nasal airway after local, low-level LPS challenge in humans J Allergy Clin Immunol 1999 104 388 94 10452761 46 Mun J-Y Onorato A Nichols FC Structural confirmation of the dihydrosphinganine and fatty acid constituents of the dental pathogen Porphyromonas gingivalis Org Biomol Chem 2007 5 3826 33 18004463 47 Korf J Stoltz A Verschoor J de Baetselier P Grooten J The Mycobacterium tuberculosis cell wall component mycolic acid elicits pathogen-associated host innate immune responses Eur J Immunol 2005 35 890 900 15724242 48 Fenton MJ Macrophages and tuberculosis Curr Opin Hematol 1998 5 72 8 9515206 49 Means TK Wang S Lien E Yoshimura A Golenbock DT Fenton MJ Human toll-like receptors mediate cellular activation by Mycobacterium tuberculosis J Immunol 1999 163 3920 7 10490993 50 Gordon S Pattern recognition receptors: doubling up for the innate immune response Cell 2002 111 927 30 12507420 51 Modlin RL Hans D Brightbill HD Godowski PJ The toll of innate immunity on microbial pathogens N Engl J Med 1999 340 1834 5 10362831 52 Cheng J-JY Yang C-J Cheng C-H Wang YT Huang NK Lu M-K Characterization and functional study of Antrodia camphorata lipopolysaccharide J Agric Food Chem 2005 53 469 547 15656690 53 Nigam S Schewe T Kock JLF 3(R)-Hydroxy oxylipins - a novel family of oxygenated polyenoic fatty acids of fungal origin Adv Exp Med Biol 1999 469 663 8 10667396 54 Deva R Ciccoli R Kock JLF Nigam S Involvement of aspirin-sensitive oxylipins in vulvovaginal candidiasis FEMS Microbiol Lett 2001 198 37 43 11325551 55 Deva R Shankaranarayanan P Ciccoli R Nigam S Candida albicans induces selectively transcriptional activation of cyclooxygenase-2 in HeLa cells: pivotal roles of Toll-like receptors, p38 mitogen-activated protein kinase, and NF-kB J Immunol 2003 171 3047 55 12960330 56 Erb-Downward JR Noverr MC Characterization of prostaglandin E2 production by Candida albicans Infect Immun 2007 75 3498 505 17470538 57 Karatuna O Yagci A Analysis of quorum sensing-dependent virulence factor production and its relation with antimicrobial susceptibility in Pseudomonas aeruginosa respiratory isolates Eur Soc Clin Microbiol Infect Dis 2010 16 1770 5 58 Pathogenesis of Cryptococcus neoformans McClelland EE Casadevall A Eisenman CH Kavanagh K New insights in medical mycology New York Springer 2007 131 57 59 Monari C Bistoni F Casadevall A Glucuronoxylomannan, a microbial compound, regulates expression of costimulatory molecules and production of cytokines in macrophages J Infect Dis 2005 191 127 37 15593014 60 Monari C Bistoni F Vecchiarelli A Glucuronoxylomannan exhibits potent immunosuppressive properties FEMS Yeast Res 2006 6 537 42 16696649 61 Shoham S Huang C Chen JM Golenbock DT Levitz SM Toll-like receptor 4 mediates intracellular signaling without TNF-alpha release in response to Cryptococcus neoformans polysaccharide capsule J Immunol 2001 166 4620 6 11254720 62 Fonseca FL Nohara LL Radames JB Immunomodulatory effects of serotype B glucuronoxylomannan from Cryptococcus gattii correlate with polysaccharide diameter Infect Immun 2010 78 3861 70 20547742 63 Noverr MC Phare SM Toews GB Coffey MJ Huffnagle GB Pathogenic yeasts Cryptococcus neoformans and Candida albicans produce immunomodulatory prostaglandins Infect Immun 2001 69 2957 63 11292712 64 Kock JLF Sebolai OM Pohl CH van Wyk PWJ Lodolo EJ Oxylipin studies expose aspirin as antifungal FEMS Yeast Res 2007 7 1207 17 17623031 65 Sebolai O Kock JLF Pohl CH Bioprospecting for novel hydroxyoxylipins in fungi: presence of 3-hydroxy palmitic acid in Saccharomycopsis malanga Antonie Van Leeuwenhoek 2001 80 311 5 11827217 66 Sjogren J Magnusson J Broberg A Schnurer J Kenne L Antifungal 3-hydroxy fatty acids from Lactobacillus plantarum MiLAB 14 Appl Environ Microbiol 2003 69 7554 7 14660414 67 Crow MT Mani K Nam Y-J Kitsis RN The mitochondrial death pathway and cardiac myocyte apoptosis Circ Res 2004 95 957 70 15539639 68 Owens KM Modica-Napolitano JS Singh KK Mitochondria and cancer Singh KK Castello LC Mitochondria and cancer New York Springer 2009 1 22 69 Pastorino JG Hoek JB Integration of energy metabolism and control of apoptosis Singh KK Castello LC Mitochondria and cancer New York Springer 2009 103 30 70 Ralph SJ Neuzil J Mitochondria as targets for cancer therapy Singh KK Castello LC Mitochondria and cancer New York Springer 2009 211 50 71 Somasundaram S Rafi S Hayllar J Mitochondrial damage: a possible mechanism of the “topical” phase of NSAID induced injury to the rat intestine Gut 1997 41 344 53 9378390 72 Glasgow JFT Middleton B Moore R Gray A Hill J The mechanism of inhibition of β-oxidation by aspirin metabolites in skin fibroblasts from Reye’s syndrome patients and controls Biochim Biophys Acta 1999 1454 115 25 10354521 73 Norman C Howell KA Millar H Whelan JM Day DA Salicylic acid is an uncoupler and inhibitor of mitochondrial electron transport Plant Physiol 2004 134 492 501 14684840 74 Leeuw NJ Swart CW Ncango DM Acetylsalicylic acid as antifungal in Eremothecium and other yeasts Antonie Van Leeuwenhoek 2007 91 393 5 17094014 75 Sebolai OM Pohl CH Botes PJ Distribution of 3-hydroxy oxylipins and acetylsalicylic acid sensitivity in Cryptococcus species Can J Microbiol 2008 54 111 8 18388980 76 Kock JLF Swart CW Ncango DM Development of a yeast bio-assay to screen anti-mitochondrial drugs Curr Drug Discov Technol 2009 6 186 91 19496752 77 Corcoran CA Huang Y Sheikh MS Energy generating pathways and the tumor suppressor p53 Singh KK Castello LC Mitochondria and cancer New York Springer 2009 131 50 78 Scatena R Bittoni P Giardina B Mitochondrial respiration and differentiation Singh KK Castello LC Mitochondria and cancer New York Springer 2009 93 102 79 Gray MW Burger G Lang BF The origin and early evolution of mitochondria Genome Biol 2001 2 10181 5 80 Andersson GE Karlberg O Canback B Kurland CG On the origin of mitochondria: a genomics perspective Philos Trans R Soc Lond B 2003 358 165 79 12594925
PMC005xxxxxx/PMC5125070.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 9214969 2488 Methods Mol Biol Methods Mol. Biol. Methods in molecular biology (Clifton, N.J.) 1064-3745 1940-6029 21468997 5125070 10.1007/978-1-61779-086-7_16 NIHMS831078 Article Quantitation of Cellular Components in Cryptococcus neoformans for System Biology Analysis Singh Arpita Qureshi Asfia Del Poeta Maurizio 19 11 2016 2011 28 11 2016 734 317333 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Methods and procedures in molecular biology used to study fungal pathogenesis have significantly improved during the last decade. In this chapter, we provide step-by-step procedures for performing genetics and biochemical studies in the human pathogenic fungal microorganism Cryptococcus neoformans (Cn). These methods are employed for studying the pathobiology of Cn and for experimental validation of theoretical models of fungal pathogenicity. Cryptococcus neoformans Fungal infection Genetic Molecular biology Biochemistry Sphingolipid DNA RNA Protein Lipids 1. Introduction Cryptococcus neoformans (Cn) is the causative agent of cryptococcosis, a fungal disease acquired by inhalation of infectious particles from the environment. Cryptococcosis is a relatively frequent disease in immunocompromised subjects and in certain regions of the world such as sub-Saharan Africa in which the estimated number of deaths associated with cryptococcal disease, at half a million per year, is comparable with the number attributed to tuberculosis (1, 2). In the USA, the prevalence of cryptococcosis in HIV positive patients is 5–10%, which is approximately the same as that for meningococcal meningitis (3). Emerging groups at risk include patients suffering from chronic lymphatic leukemia, Hodgkin’s disease, chronic myelogenous leukemia, and multiple myeloma (4). The median overall survival of patients with lymphoproliferative disorders affected by cryptococcosis is 2 months, which is significantly shorter than the 9-month median survival of an AIDS patient with cryptococcosis (5). Cryptococcosis is also associated with organ transplantation (6, 7), and was documented in 2.8% of organ transplant recipients with an overall death rate of 42% (8). Some cases of cryptococcosis occur in patients with apparently normal immune function (9–12). One area of investigation that has significantly improved in the last 2 decades is the molecular biology of this microorganism. The development of molecular epidemiology and phylogeny and molecular technology for clinical diagnosis have significantly helped the clinicians to better manage this life-threatening disease. However, it was the advent of genetics and biochemistry of this microorganism that allowed basic and clinical investigators to address mechanistic questions and study the pathophysiology of cryptococcosis. This was (and still is) an essential step to define fungal features and characteristics necessary for the organism to cause disease (13). These fungal factors can then be exploited for the understanding of fungal pathogenicity and fungal interaction with the host cells and, ultimately, and for the development of new therapeutic strategies. With the rise of its importance as a human pathogen, there has been a concurrent rise in the ability to molecularly study its physiopathology. In Chapter 9, we provide a mathematical model of the regulation of melanin production by the sphingolipid pathway. In particular, we show that a specific enzyme of the sphingolipid pathway, inositol phosphoryl ceramide synthase 1 (Ipc1), regulates melanin formation in Cn through the production of diacylglycerol (DAG) and the consequent activation of protein kinase C 1 (Pkc1). Thus, the downregulation or/and deletion of IPC1 or/and PKC1 genes by homologous recombination should produce mutant strains that make less or no melanin. We would expect IPC and DAG lipid measurements to be decreased in the mutant in which Ipc1 is downregulated. Also in this mutant, Pkc1 enzymatic activity should be decreased. This experimental approach is necessary to validate the changes in the network behavior simulated by the mathematical model. Therefore, the deletion of the gene of interest by homologous recombination and confirmation by Southern or/and Northern blot of the isolated genomic DNA or total RNA, respectively, and the analysis of protein and lipid levels regulated by those genes are useful methods to refine and validate the mathematical model. In this chapter, we provide basic molecular methods for performing genetics and biochemistry studies in Cn. These methods can be employed to validate hypotheses and theoretical models of Cn pathogenicity or simply to study the pathobiology of this important human pathogen. 2. Materials 2.1. DNA Isolation from Cryptococcus neoformans 1 Yeast Peptone Dextrose (YPD) agar plates and YPD broth (see Note 1). 2 Sterile PBS 1×. 3 1 M Tris–HCl pH 7.5. 4 0.5 M, EDTA pH 8.0. 5 5 M NaCl. 6 100% Triton X-100. 7 20% SDS. 8 TENTS: 10 mM Tris–HCl, pH 7.5, 1 mM EDTA, pH 8.0, 100 mM NaCl, 2% Triton X-100, 1% SDS. 9 Acid washed 0.425–600 µm glass beads (SIGMA). 10 Phenol:chloroform:isoamyl alchohol =25:24:1 (SIGMA). 11 3 M Sodium acetate (NaOAC). 12 TE buffer, pH 8.0, sterile. 2.2. Biolistic Delivery in Cryptococcus neoformans 1 YPD agar + 1 M Sorbitol plates. 2 YPD agar + Nourseothricin/Hygromycin (100 µg/ml) plates. 3 0.6 µm Gold beads (BIORAD). 4 MacroCarriers (BIORAD). 5 Rupture Disks, 1,350 psi, (BIORAD). 6 Stopping Screens (BIORAD). 7 2.5 M CaCl2 sterile. 8 1.0 M spermidine (filter sterilize) (SIGMA), can be stored at −20°C. 9 100% Ethanol. 10 Isopropanol. 11 This instruction assumes the use of PDS-1000/He Biolistic Particle Delivery System from BIORAD. 2.3. Southern Hybridization of DNA Extracted from Cryptococcus neoformans 1 Denaturing solution: 1.5 M NaCl, 0.5 M NaOH. 2 Neutralizing Solution: 1 M Tris–HCl pH 8.0, 1.5 M NaCl. 3 Nytran SPC (0.45 µm Nylon Transfer Membrane) (Whatman). 4 Whatman 3MM Blotting Paper. 5 Paper towels, preferably single fold. 6 20 × SSC: 175.3 g of NaCl, 88.2 g of sodium citrate in 800 ml of double distilled water. Adjust pH with NaOH pellets and adjust the total volume to 1 L. Autoclave. Can be stored at room temperature. 7 20 × SSPE : 175.3 g of NaCl, 27.6 g of NaH2PO4·H2O, 7.4 g EDTA in 800 ml of double distilled water. Adjust the pH with NaOH pellets to 7.4. Final volume made up to 1 L. Autoclave. Can be stored at room temperature. 8 20% SDS. 9 Nonfat dry milk. 10 Prehybridizing solution: 10 ml 20× SSPE, 10 ml 20% SDS, 2 ml 10% nonfat dry milk in a total volume of 40 ml. Can be stored at 4°C with 0.02% sodium azide for 2–3 days. 11 Random Primers DNA labeling system kit (Invitrogen). 12 32P dCTP (Perkin Elmer). 13 Microspin G-25 column (Amersham Biosciences). 14 Sterile TE buffer, pH 8.0. 2.4. Highly Pure Total RNA Isolation from Cryptococcus neoformans (e.g., for Microarray Studies) 1 YPD agar plate and YPD broth. 2 Phosphate Buffered Saline (PBS) 1× sterile. 3 Tri Reagent (Molecular Research Centre). 4 BAN as a phase separation reagent, molecular biology grade (Molecular Research Centre). 5 RNeasy Mini Kit and RNeasy MinElute Cleanup Kit (Qiagen). 6 RNase Zap for removing RNase contamination from external surface (Ambion). 7 RNase/DNase free plastic wares. 2.5. Protein Extraction from Cryptococcus neoformans 1 YPD media and agar plates (made from YPD 50 g/L and agar 20 g/L). 2 Buffer for Cn cell lysis: 1 ml 1 M Tris–HCl pH 8, 9 ml H2O, 1.5 ml glycerol (13% v/v), 10 µl CLAP: chymostatin, leupeptin, antipain, and pepstatin A (each at 10 mg/ml in DMSO and stored at −20°C), and 20 µl 100 mM solution phenylmethylsulfonyl fluoride (PMSF) in isopropanol. 3 Glass beads, acid washed, 425 µm (30–40 US sieve) (Sigma). 2.6. Lipid Extraction 1 Mandala lipid extraction buffer: 150 ml ethanol, 150 ml distilled water, 50 ml diethyl ether, 10 ml pyridine, and 180 µl 14.2 N ammonium hydroxide. 2 Use glass tubes for all extraction steps (VWR) fit best in the ThermoSavant SPD2010 SpeedVac system we use). 3 Waters Sep-Pak Classic Silica cartridges (WAT 051900, 690 mg) for analytical scale or WAT036930 200 cc, 5 g cartridges for semipreparative scale lipid isolation and purification. 4 10″ × 10″ glass tank for thin layer chromatography (TLC). 5 3MM Whatman chromatography paper (Fisher). 6 TLC chromatography plates (Fisher M5628-5 or 05-713-329 depending on analytical or semipreparative purposes). 7 Soy glucosylceramide standard (Avanti Polar Lipids) made up to 3.5 mM (2.5 µg/µl) in chloroform/methanol (2:1). 8 Prepare 70% H2SO4 by adding 14 ml H2SO4 slowly to 6 ml water on ice, with mixing. Add 40 mg resorcinol to 20 ml 70% H2SO4. Stir well at room temperature with a magnetic stirrer bar. Pour solution into a glass TLC sprayer. 2.7. In Vitro Enzyme Activity Assay 1 NBD-C6-ceramide (Avanti Polar Lipids). 2 Lysis buffer: 25 mM Tris–HCl pH 7.4, 5 mM EDTA, 1 mM PMSF, and CLAP: chymostatin, leupeptin, antipain, and pepstatin A (each at 10 mg/ml in DMSO and stored at −20°C). 3 Silica gel 60 TLC plates (EM Sciences, Fisher). 2.8. Mass Spectrometry of Lipids 1 Commercially available synthetic lipid standards. 3. Methods 3.1. DNA Isolation (14, 15) 1 Inoculate a 10–15 ml YPD broth with a single colony from a fresh YPD agar plate and grow them for 20–24 h at 30°C with constant shaking. Pellet cells from this culture at 1,200 × g 4°C for 10 min. 2 Wash with sterile PBS 1× twice and resuspend in 1 ml of sterile double distilled water and transfer to a 2-ml screw cap tube (see Note 2). 3 Pellet cells in a Microcentrifuge for 30 s at 1,200 × g at room temperature. 4 Pour off the water; add 0.5 ml of TENTS and vortex at 7–8 speed for three times, 45 s each. This step assumes the use of Vortex Genie 2 from Scientific Industries. 5 Add two cups (1 cup =400–500 µl) of acid washed 0.45 µm glass beads (see Note 3) and 0.5 ml phenol–chloroform–Isoamyl alcohol (see Note 4). 6 This step assumes the uses of Bead Beater 16 from Scientific Industries. Tubes were vortexed/homogenized in a Bead Beater three times, 45 s each, with a gap of 45 s on ice, in between each cycle (see Note 5). 7 After homogenizing (or lysing) of the cells, centrifuge the cells for 10 min at 8,000 × g at room temperature to separate the cell debris and unbroken cells. 8 Remove the upper aqueous phase which now contains the DNA, to a fresh 1.5-ml Eppendorf and add 1 ml of ethanol 100% and keep at −20°C overnight (see Note 6). 9 Centrifuge the tube at 8,000 × g for 30 min at 4°C. Remove the supernatant, dissolve the pellet in 200 µl of TE containing RNase A at a concentration of 100 µg/ml and then incubate at 37°C for 20 min. 10 After incubation, add equal volume of phenol–chloroform–isoamyl alcohol and mix gently by inverting 4–6 times. Centrifuge at 8,000 × g, 10 min, 4°C. Remove the aqueous phase and repeat the step with the aqueous phase. 11 Add 20 µl of 3M NaOAC and 400 µl of ethanol (100%) to the final aqueous phase and incubate at −20°C for 30–60 min for complete precipitation. 12 After precipitation, centrifuge the tube at 8,000 × g at 4°C for 5 min. 13 Wash the DNA pellet twice with 200 µl of ice-cold 70% ethanol and air dry the pellet (see Note 7). 14 Dissolve the pellet in 30–50 µl of sterile TE gently and store at −20°C. 3.2. Biolistic Delivery in Cryptococcus neoformans (14, 16) 1 Spin down 19–20 h grown culture (15 ml) of the recipient strain and throw off 12 ml of the supernatant (see Note 8). 2 Plate 200–250 µl of this cell suspension on prewarmed YPD agar + 1 M sorbitol plates and let them dry for 4–5 h at 30°C (see Note 9). This should include a “non-shot” control plate. 3 During this time of incubation, prepare the shot. For preparing a stock of Gold Beads – 60 mg/ml, 30 mg was weighed out and dissolved in 100% ethanol, vortexed vigorously for 3 min, incubated at room temperature for 15 min and spun for 1 min. Discard the supernatant and suspend the gold beads in 1 ml of sterile water. Incubate or allow the particles to settle down, pellet and discard the supernatant. Add 500 µl of 50% Glycerol to make a final concentration of 60 mg/ml. This stock can be stored in 4°C. 4 Each Shot should be prepared as follows in the same sequence: 10 µl of 60 mg/ml of gold beads 1 µl of ≥1 µg/µl of DNA 10 µl of 2.5 M CaCl2 2 µl of 1.0 M Spermidine Vortex the mix for 3–5 min and let it settle for 5 min at room temperature. Spin for 20 s and take off the supernatant. Wash the Bead-DNA mix once with 500 µl of 100% ethanol by vortexing and spin down the Bead-DNA. Throw off the supernatant. Finally, resuspend the Bead-DNA in 25 µl of 100% ethanol (see Note 10). 5 The Macrocarriers should be prepared inside a Laminar Hood to prevent contamination. Dip the macrocarriers (one for each shot) in 100% ethanol. Blot off the excess liquid on a sterile wiper and keep in a sterile Petri dish until completely dry. 6 Vortex the Bead-DNA well so that the beads are uniformly coated with the DNA (see Note 11). Spread 10 µl of this mix, first onto the center of the macrocarrier then working outward, within 5mmto the edge in a slow circular motion. Let it dry. If there is any extra Bead-DNA, it can be added to each macrocarrier in the same fashion (see Note 12). 7 The machine (PDS-1000/He Biolistic Particle Delivery System) should be sterilized with 70% ethanol and dried before shooting. The chamber should be kept closed as much as possible. Open the Helium tank pressure valve and set the pressure regulator at 1,800–2,100 psi. 8 Soak the rupture disk, 1,350 psi in isopropanol, place in the retaining cap and screw the unit onto the gas acceleration tube of the machine with the retaining cap torque wrench (see Note 13). 9 Unscrew the macrocarrier cover lid and place a stopping screen on the stopping screen support. Place the macrocarrier on top (Bead-DNA side up) of the macrocarrier holder, invert and place on the fixed nest. The dried microcarriers should face toward the stopping screen. Screw the macrocarrier cover lid to the assembly until tightened and place this in the top slot inside the chamber. 10 Place the target shelf on the second to bottom shelf (see Note 14). Place the YPD agar + 1Msorbitol Petri dishes with cells, on this shelf without the lid on. 11 Close the chamber and set the vacuum switch at “VAC” position till the desired vacuum of 28.5–29″ is reached. Hold the vacuum chamber at this level of vacuum by quickly pressing the switch to “HOLD” position and press the “FIRE” switch to bombard the sample into the plate until the rupture disk pops. Vent the chamber and immediately cover the Petri dish with lid and remove it from the chamber. 12 Repeat the shooting until all the macrocarriers coated with Bead-DNA were utilized. All the parts should be cleaned and surface sterilized with 70% ethanol between two different DNA samples. 13 Incubate the “shot” along with a “non-shot control” plates for 2 h at 30°C (see Note 15). 14 Label Falcon 2054 tubes, one for each of the shot and non-shot plates. 15 Aliquot 1 ml of prewarmed YPD broth onto each plate. Rub the liquid broth across the whole surface of the plate with sterile hockey stick and scrape off the cells. Tilt the plate and pipette the liquid into the labeled Falcon tubes. 16 Plate 200–250 µl of the scraped liquid and spread uniformly onto prewarmed YPD Nourseothricin/Hygromycin plates. Incubate the plates at 30°C for several days. 3.3. Southern Hybridization (17) 1 After taking picture of the Gel (see Note 16), denature in the Denaturing Solution (use fresh) for 1 h at room temperature with constant shaking. 2 Neutralize the gel in Neutralizing Solution for 1–2 h. 3 Wash the gel with double distilled water. 4 Wet the membrane and the 3MM Whatman paper in 2× SSC until complete wet and assemble the transfer. Transfer overnight at room temperature or at least for 16–18 h. 5 Before removing the gel, mark with pencil the wells on the membrane (see Note 17). Keep the membrane on a filter paper presoaked with 6× SSC at room temperature and semidry. Auto cross-link for 1 min at 1,200 (µJ × 100; this instruction assumes the use of UV Stratalinker 1800 from Stratagene). The membrane, if not set for hybridization can be stored at 4°C for 2–3 days in a sealed bag. 6 Prehybridize the membrane in prehybridizing solution for 1–2 h at 65°C. 7 Labeling of probes: Spun down the contents of the Random Primer Labeling kit for 30 s in microcentrifuge after thawing. Boil 9 µl of DNA (for probe) for 5 min and cool it down on ice for 1 min. Add to the DNA 1 µl each of dATP, dGTP, dTTP, 2 µl of Random Primers, 5 µl of 32P dCTP, and lastly 1 µl of Klenow. Incubate the mix for 30 min at 37°C. After incubation, add 2 µl of Stop buffer and 20 µl of TE. Snap the tip of a microspin G 25 column and put in a 1.5-ml Eppendorf and spin for 30 s in a centrifuge inside the Laminar Hood and then run the probe through the column (see Note 18). Boil the probe for 5 min and cool it on ice for 1 min. Add 1 ml of 5× SSPE with a syringe into the probe and transfer it to the hybridizing chamber carefully. Hybridize overnight and wash sequentially with 50 ml of 0.1% SDS in 2× SSC for 20–30 min at 65°C, 50 ml of 0.5% SDS in 0.1× SSC thrice, each for 20–30 min at 65°C. 8 Dry the membrane over a filter paper and saran wrap and tape it on a cassette. Inside the darkroom put the film on top of the membrane and expose the film at −80°C overnight or at the least 4–5 h before developing. 3.4. Isolation of Total RNA (18) 1 Harvest cells (20–24 h) grown in the required media by pelleting down at 1,200 × g at 4°C for 10 min (see Note 19). 2 Wash the pelleted cells with sterile PBS twice and spin down at 1,200 × g, 4°C for 5 min. Drain out the PBS on a sterile wipe and flash freeze in a dry ice – ethanol bath and set for lyophilization (see Note 20). 3 Aliquot ~100 µl (about 50–75 mg) of lyophilized cells in a 2-ml screw cap tube and grind or smash the cells to powder form with the help of the spatula used to scoop out the lyophilized cells (see Note 21). Add 1–1.25 ml of Tri reagent. Cap the tubes properly and homogenize in Bead Beater 16 with pulses as follows 45 s thrice, 30 s once with a gap of 45 s between each cycle on ice. 4 Incubate the tubes for 10 min at room temperature. Centrifuge for 10 min at 4°C at 8,000 × g to pellet the cell debris and unbroken cells. 5 Transfer the supernatant to a fresh tube and add 50–60 µl of BAN (50 µl of BAN/ml of Tri reagent added) and shake vigorously for 20–30 s. Incubate for 5 min at room temperature and centrifuge at 8,000 × g at 4°C for 10 min. 6 Transfer the aqueous phase to a new tube and add equal volume of 70% Ethanol and mix gently and properly (see Note 22). 7 Load the aqueous phase (700 µl at a time) onto an RNeasy isolation column and spin for 30 s in a centrifuge at 8,000 × g at room temperature. If the volume exceeds 700 µl, the same column can be reloaded until the whole aqueous phase had passed through it. 8 Discard the flow-through and wash the column with RW1 buffer provided with the kit. Discard the flow-through and wash with 500 µl of RPE twice, spin for 30 s at 8,000 × g and discard the flow-through. Transfer the RNA isolation column to a new 2-ml collection tube and spin for 2 min at 8,000 × g at room temperature. 9 Elute the RNA in 50 µl of RNase free water in a fresh 1.5-ml Eppendorf. Re-elute the residual RNA in another aliquot of 50 µl of RNase free water in the same tube. 10 Concentrate the RNA with the column from RNeasy MinElute Cleanup kit following instructions of the manufacturer. Elute in a final volume of 20 µl of DNase–RNase free water. 3.5. Protein Extraction from Cn 1 These instructions assume the use of a Bead Beater 8. 2 Streak out Cn strains of interest (e.g., wt or mutant) onto YPD agar plate and incubate at 30°C for 48 h. 3 Pick a single colony into a 50-ml Corning Centrifuge tube containing 10 ml YPD media and allow to grow at 30°C with shaking for 24 h. 4 Centrifuge the culture 10 min at 1,200 × g at ambient temperature (20°C), wash once with doubly distilled water and then resuspend into 7 ml doubly distilled water. 5 Aliquot 1 ml each into 1.5-ml conical tubes with screw caps and centrifuge at 3,500 × g for 10 min at 25°C. 6 Meanwhile prepare the lysis buffer. 7 Following centrifugation, discard the supernatant and resuspend each pellet into 200 µl lysis buffer. 8 Add one “cupful” glass beads (see Note 3), then vortex and place on ice. 9 Place each tube into the beadbeater in a 4°C coldroom and beadbeat for 40 s, followed by 1 min on ice. Repeat this four times (see Note 23). 10 Centrifuge each tube at 3,500 × g for 12 min at 4°C. 11 Collect the supernatant and carry out Bio-Rad protein assay to determine the amount of protein. 3.6. Lipid Extraction 1 Under sterile conditions, fill 50-ml tube with 9 ml yeast-peptone (YP) and 1 ml 20% glucose. Add a single colony of the strain of interest (in this case Cn Gcs1REC) and incubate 48 h at 30°C, 250 rpm. 2 Centrifuge at 1,200 × g for 10 min at 4°C. Wash pellet twice with water then resuspend in 9 ml sterile water. 3 Count the cells after appropriate serial dilution and aliquot 5 × 108 cells per tube. Centrifuge 10 min 1,200 × g at 4°C. Suction out water carefully (see Note 24). 3.6.1. Mandala Extraction (for Extraction of Inositol-Containing Phospholipids and Phosphatidylcholine) (see Note 25) 4 Add 1.5 ml Mandala extraction buffer (19) to each tube. Vortex and sonicate 20 s each. 5 Incubate at 60°C in a water bath for 15 min, vortex and sonicate for 20 s each then reincubate at 60°C for 15 min. 6 Sonicate 20 s then centrifuge 10 min at 1,200 × g at 4°C. Using a glass Pasteur pipette, combine supernatant from two tubes together into a clean tube. 7 Evaporate the solvent in the Speedvac (see Note 25). 3.6.2. Bligh and Dyer Lipid Extraction (for Determination of Neutral Lipids) 8 Following evaporation, add 2 ml methanol and vortex. Sonicate if necessary. 9 Add 1 ml chloroform and vortex. Ensure there is one phase, even if turbid (see Note 26). 10 Incubate the samples at 37°C for 1 h. During this period, vortex each sample twice for 30 s. 11 Centrifuge at 1,200 × g for 5 min at room temperature, then transfer the lower phase to a clean tube with a glass Pasteur pipette. Add 1 ml Chloroform and 1 ml water and vortex twice for 30 s each. Recentrifuge samples at 1,200 × g for 5 min at room temperature. 12 Once again, using a glass Pasteur pipette transfer lower phase to a clean tube. Up to three tubes can be combined into one to lessen the amount of tubes being handled. 13 Evaporate the solvent in the Speedvac (see Note 25). 3.6.3. Additional Purification Steps (e.g., Isolation of Glucosylceramide Using a Silica Column) 3.6.3.1. Silica Column Purification 1 14 Resuspend the lipids in 1 ml chloroform/acetic acid (99:1). 15 Wash the SepPak cartridges with 15 ml chloroform (see Note 27). Apply sample (in 1 ml) and rinse with 1.5 ml chloroform/acetic acid (99:1). Collect flow-through after 0.5 ml has been allowed to collect into waste. 16 Add 15 ml chloroform/acetic acid (99:1) and collect 5 ml per tube. 17 Add 15 ml acetone and collect 5 ml per tube. Evaporate acetone from these tubes in the SpeedVac then resuspend in minimum amount acetone to combine into one tube. Reevaporate (see Note 28). 3.6.4. Base Hydrolysis 18 Add 0.5 ml chloroform, followed by 0.5 ml 0.6 M KOH in methanol to each sample. Vortex well and leave at room temperature for 1 h. 19 Add 0.325 ml 1 M HCl followed by 0.125 ml distilled water. Vortex well then centrifuge at 1,200 × g for 10 min at room temperature. Transfer lower organic phase to a clean tube. 20 Evaporate solvent in the SpeedVac. You should have a small dark brown pellet at this stage (see Note 25). 3.6.5. Silica Column Purification 2 21 Resuspend the pellet in 1 ml chloroform/acetic acid (99:1). Repeat steps 16 and 17. 22 Change eluting solvent to chloroform/methanol (95:5); add 10 ml and collect in two tubes. 23 Change eluting solvent to chloroform/methanol (90:10); add 15 ml and collect into 3 tubes. These are the tubes that will contain glucosylceramide, the lipid of interest for this example. Evaporate solvent using a SpeedVac. Do not combine the tubes (see Note 29). 24 Wash the column with 15 ml methanol and collect in case needed. 3.6.6. Thin Layer chromatography 25 Prepare a 10′ × 10′ glass TLC tank by adding chloroform/methanol/water (97.5:37.5:6) to a clean, dry tank lined with white chromatography paper. Apply a thin layer of vacuum grease around the top lip of the tank to ensure a good seal (see Note 30). Leave until paper is well saturated, usually at least 5 h to overnight. 26 Spot the soy standard onto a TLC plate 1.5 cm from the bottom using a 10 µl pipette, using 1, 2, and 3 µl in three separate lanes (equivalent to 2.5, 5, and 7.5 µg, respectively). 27 Resuspend the dried lipid from step 24 in 30 µl chloroform/methanol (2:1), and spot either 30 µl (analytical) or 5 µl (semipreparative scale) onto the TLC plate into a fourth lane. Allow solvent to evaporate in fume hood (~1–2 min) before placing the TLC plate in the tank. 28 Make sure the TLC tank is tightly closed. Allow the solvent front to migrate up to 1 cm from the top of the plate, before removing the plate from the tank. 29 Dry the TLC plate in the hood at room temperature prior to placing it in another tank containing only iodine crystals to allow visualization of the lipids. Alternatively, the plate can be sprayed with resorcinol in 70% H2SO4, and then placed in an oven for 10 min to allow a dark purple color to develop wherever sugar moieties are located on the lipids. 3.7. In Vitro Enzyme Activity Assay 1 This protocol describes the in vitro activity assay of Ipc1 (20) but could be adapted for assaying any enzyme from Cn. Ipc1 activity is measured by using the fluorescent ceramide analog NBD-C6-ceramide as substrate and monitoring the formation of NBD-C6-IPC, as described by Fischl et al. (21) with some modifications. 2 Grow wt and mutant Cn strains in YPD media in a shaker incubator for 24 h at 30°C. Harvest the cells by centrifugation and wash with sterile distilled water (see Note 24). 3 Resuspend the pellets in lysis buffer, add acid-washed glass beads for a volume equal to ¾ of the cell suspension and homogenize three times for 45 s, followed by 1 min on ice each time, using the Bead Beater 8. 4 Centrifuge at 2,500 × g for 10 min at 4°C, then transfer the supernatant (~100 µl) to a sterile 1.5-ml microcentrifuge tube for protein quantification. 5 Following protein determination, incubate 100 µg protein from the cell lysates for 30 min at 30°C in 50 mM bis-Tris–HCl buffer (pH 6.5) containing 1 mM phosphatidyl inositol, 5 mM Triton X-100, 1 mM MnCl2, 5 mM MgCl2, and 20 µM NBD-C6-ceramide in a final reaction volume of 100 µl. 6 Terminate the reaction by addition of 0.5 ml 0.1 N HCl in methanol. 7 Add 1 ml chloroform and 1.5 ml 1 M MgCl2, mix well and centrifuge at 1,000 × g for 10 min to separate the phases. 8 Analyze the chloroform-soluble product, NBD-IPC, by TLC on silica gel 60 plates (EM Science) as described above using chloroform/methanol/water (65:25:4). 9 Identify and quantify NBD-IPC by direct fluorescence using a Molecular Dynamics 840 Storm unit. 3.8. Mass Spectrometry of Lipids (22, 23) 1 This protocol describes MS and MS/MS of Cn glucosylcersamide but is applicable to any Cn lipid molecule. 2 Following Bligh and Dyer extraction described above under lipid extraction, MS and MS/MS scans of glucosylceramide were carried out on a Thermo Finnigan TSQ7000 triple quadrupole mass spectrometer equipped with electrospray ionization as described in ref. 6. 3 A 31 min method was used with A; water/0.2% formic acid/2 mM ammonium formate and B: methanol/0.2% formic acid/1 mM ammonium formate, on a 150 × 3 mm Spectra 3 µm C8SR column (Peeke Scientific) using gradient elution and addition of internal standards. 4 Include multiple reaction monitoring (MRM) for the characteristic production m/z 276.2. 5 Quantify Cn glucosylceramide using soy glycosylceramide (Avanti Polar lipids) for standard curve generation. 6 Normalize mass spectral data to inorganic phosphate determination. This work was supported by Grants AI56168 and AI72142 (to M.D.P) and was conducted in a facility constructed with support from the National Institutes of Health, Grant Number C06 RR015455 from the Extramural Research Facilities Program of the National Center for Research Resources. Dr. Maurizio Del Poeta is a Burroughs Wellcome New Investigator in Pathogenesis of Infectious Diseases. 1 1. All solutions should be prepared in water which has a resistivity of 18.3 MΩ-cm and total organic content of less than 5 ppb. This water is referred to double distilled water in this text. 2 Each tube should contain ~100 µl of cell pellet. 3 A cup was made by cutting out from the bottom till 0.5 ml marking of a 1.5-Eppendorf tube. Drive 23 gauge BD needle into the cup through the upper part to make a makeshift handle. 4 Tubes should be capped properly and the mouth should be wiped with kimwipes ensuring that tubes seal properly before vortexing. 5 The Vortex should have the single unit assembly during vortexing. 6 The volume of ethanol should be 2–2.5 times the volume of the aqueous phase. Incubating at −20°C at 2 h can also be done, however, the yield may be less. 7 The tube containing the DNA pellet can be covered by Para film, punctured and kept at 4°C to let the ethanol dry off. However, it should not be too much dried. 8 200–250 µl was to be used from this cell suspension, so this would suffice for 15–12 plates for biolistic delivery. If more number of shots is desired, the culture volume should increase proportionately as during shooting the recipient cell density should be high. 9 The cells should be spread in a monolayer over the plate. To do this, spread the cell suspension with a sterile glass hockey stick in a single direction. 10 10–15 µl extra ethanol was added to compensate for evaporation. It was always wise to include at least two extra shot when preparing for the Bead-DNA. 11 Spread Bead-DNA mix immediately as they have a tendency to settle down. Best is to spread from a continuously vortexed mix. 12 The macrocarrier should be used for shooting within 1–2 h of its preparation. 13 The rupture disk should not be kept for more than 30–60 s in the isopropanol and excess liquid should be blotted off as this may cause delamination. The rupture disk should also be wet while being loaded as the liquid reduces failure rate of the rupture disk. The retaining cap should be clean for any residual rupture disk part from previous shooting as this may cause rupture of the disk at a wrong pressure and thereby no delivery of the DNA into the Cryptococcal cells. 14 This distance is the best for delivering DNA into Cryptococcal cells. 15 If not transforming with any selectable marker like Nourseothricin/Hygromycin, these plates, after shooting can be incubated at 30°C directly, for several days. 16 The amount of DNA before restriction enzyme digestion is quantified by agarose gel electrophoresis and the DNA should be completely digested. 17 The total well should be marked with a pencil. 18 The amount of the DNA used as probe should be at least 100 ng. The angle of the Eppendorf with the G25 column after loading of the Probe should be the same as before in the microcentrifuge. 19 Be extremely cautious about RNase contamination. Wipe with RNase ZAP the whole external surface of the working area, pipettes, etc., before starting and change gloves frequently. If Minimal Media (YNB or DMEM) is to be used, it can be supplemented with 50 mMHepes, 1Msorbitol, and 10% FCS if required. 20 Lyophilization for a 75–100 ml culture should be at least for 24 h but not more than 48 h. 21 The lyophilized cells in powder form give better yield. 22 Do not let the tip touch into the interphase while transferring the aqueous phase. 23 It is important to go through four cycles on the beadbeater when lysing the Cn cells otherwise insufficient protein will be extracted. 24 At this stage, the cell pellet can be frozen at −80°C until ready for extraction. 25 To see the original references on how this protocol was established, see reference by Barbara Hanson (24). 26 The tubes can be left at 4°C overnight if there are time constraints. 27 For analytical scale, use WAT051900; 15 ml is equivalent to a 5 bed volume wash. 28 Dry down other tubes as well in case needed later, then store at −20°C. 29 Try to get as much compound down as possible by rinsing the walls of the glass tube with 9:1 chloroform:methanol. 30 You can add two weights on top to ensure the cover seals well. The weights can be 2 × 250 ml glass bottles filled with water. References 1 Harrison TS The burden of HIV-associated cryptococcal disease AIDS 2009 23 531 532 19240459 2 Park BJ Wannemuehler KA Marston BJ Govender N Pappas PG Chiller TM Estimation of the current global burden of cryptococcal meningitis among persons living with HIV/AIDS AIDS 2009 23 525 530 19182676 3 Hajjeh RA Conn LA Stephens DS Baughman W Hamill R Graviss E Pappas PG Thomas C Reingold A Rothrock G Hutwagner LC Schuchat A Brandt ME Pinner RW Cryptococcosis: population-based multistate active surveillance and risk factors in human immunodeficiency virus-infected persons. Cryptococcal Active Surveillance Group J Infect Dis 1999 179 449 454 9878030 4 Kaplan MH Rosen PP Armstrong D Cryptococcosis in a cancer hospital: clinical and pathological correlates in forty-six patients Cancer 1977 39 2265 2274 322854 5 White M Cirrincione C Blevins A Armstrong D Cryptococcal meningitis: outcome in patients with AIDS and patients with neoplastic disease J Infect Dis 1992 165 960 963 1569350 6 Kohno S Varma A Kwon-Chung KJ Hara K Epidemiology studies of clinical isolates of Cryptococcus neoformans of Japan by restriction fragment length polymorphism Kansenshogaku Zasshi 1994 68 1512 1517 7876673 7 Shaariah W Morad Z Suleiman AB Cryptococcosis in renal transplant recipients Transplant Proc 1992 24 1898 1899 1412904 8 Husain S Wagener MM Singh N Cryptococcus neoformans infection in organ transplant recipients: variables influencing clinical characteristics and outcome Emerg Infect Dis 2001 7 375 381 11384512 9 Fraser JA Giles SS Wenink EC Geunes-Boyer SG Wright JR Diezmann S Allen A Stajich JE Dietrich FS Perfect JR Heitman J Same-sex mating and the origin of the Vancouver Island Cryptococcus gattii outbreak Nature 2005 437 1360 1364 16222245 10 Seaton RA Verma N Naraqi S Wembri JP Warrell DA Visual loss in immunocompetent patients with Cryptococcus neoformans var. gattii meningitis Trans R Soc Trop Med Hyg 1997 91 44 49 9093627 11 Seaton RA Naraqi S Wembri JP Warrell DA Predictors of outcome in Cryptococcus neoformans var. gattii meningitis Qjm 1996 89 423 428 8758045 12 Findley K Rodriguez-Carres M Metin B Kroiss J Fonseca A Vilgalys R Heitman J Phylogeny and phenotypic characterization of pathogenic Cryptococcus species and closely related saprobic taxa in the Tremellales Eukaryot Cell 2009 8 353 361 19151324 13 Perfect JR Cryptococcus neoformans: a sugar-coated killer with designer genes FEMS Immunol Med Microbiol 2005 45 395 404 16055314 14 Casadevall A Perfect JR Cryptococcus neoformans 1998 Washington, DC ASM Press 381 405 15 Hull CM Heitman J Genetics of Cryptococcus neoformans Annu Rev Genet 2002 36 557 615 12429703 16 Toffaletti DL Rude TH Johnston SA Durack DT Perfect JR Gene transfer in Cryptococcus neoformans by use of biolistic delivery of DNA J. Bacteriol 1993 175 1405 1411 8444802 17 Sambrook J Fritsch EF Maniatis T Molecular cloning: a laboratory manual 1989 Cold Spring Harbor, New York Cold Spring Harbor Laboratory Press 18 Baker LG Specht CA Donlin MJ Lodge JK Chitosan, the deacetylated form of chitin, is necessary for cell wall integrity in Cryptococcus neoformans Eukaryot Cell 2007 6 855 867 17400891 19 Mandala SM Thornton RA Frommer BR The discovery of australifungin, a novel inhibitor of sphinganine N-acyltransferase from Sporormiella australis. Producing organism, fermentation, isolation, and biological activity J. Antibiot. (Tokyo) 1995 48 349 356 7797434 20 Luberto C Toffaletti DL Wills EA Tucker SC Casadevall A Perfect JR Hannun YA Del Poeta M Roles for inositol-phosphoryl ceramide synthase 1 (IPC1) in pathogenesis of C. neoformans Genes Dev 2001 15 201 212 11157776 21 Fischl AS Liu Y Browdy A Cremesti AE Inositolphosphoryl ceramide synthase from yeast Methods Enzymol 2000 311 123 130 10563317 22 Bielawski J Pierce JS Snider J Rembiesa B Szulc ZM Bielawska A Comprehensive quantitative analysis of bioactive sphingolipids by high-performance liquid chromatography-tandem mass spectrometry Methods Mol Biol 2009 579 443 467 19763489 23 Bielawski J Szulc ZM Hannun YA Bielawska A Simultaneous quantitative analysis of bioactive sphingolipids by high-performance liquid chromatography-tandem mass spectrometry Methods 2006 39 82 91 16828308 24 Hanson BA Lester RL The extraction of inositol-containing phospholipids and phosphatidylcholine from Saccharomyces cerevisiae and Neurospora crassa J Lipid Res 1980 21 309 315 6445928
PMC005xxxxxx/PMC5125071.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 7505689 5980 Mycopathologia Mycopathologia Mycopathologia 0301-486X 1573-0832 21971701 5125071 10.1007/s11046-011-9485-8 NIHMS831082 Article Detection of Antibody against Fungal Glucosylceramide in Immunocompromised Patients: A Potential New Diagnostic Approach for Cryptococcosis Qureshi Asfia Department of Biochemistry and Molecular Biology, Medical University of South Carolina, 173 Ashley Ave, BSB 512A, Charleston, SC 29425, USA Wray Dannah Division of Infectious Diseases, Medical University of South Carolina, Charleston, SC, USA Rhome Ryan Department of Biochemistry and Molecular Biology, Medical University of South Carolina, 173 Ashley Ave, BSB 512A, Charleston, SC 29425, USA Barry William Division of Infectious Diseases, Medical University of South Carolina, Charleston, SC, USA Del Poeta Maurizio [email protected] Department of Biochemistry and Molecular Biology, Medical University of South Carolina, 173 Ashley Ave, BSB 512A, Charleston, SC 29425, USA Division of Infectious Diseases, Medical University of South Carolina, Charleston, SC, USA Department of Microbiology and Immunology, Medical University of South Carolina, Charleston, SC, USA Department of Craniofacial Biology, Medical University of South Carolina, Charleston, SC, USA 19 11 2016 5 10 2011 6 2012 28 11 2016 173 5-6 419425 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. We have developed an ELISA to determine the value of anti-glucosylceramide antibody for the prediction of disseminated cryptococcosis in immunocompromised subjects and performed a clinical prospective study at the Medical University of South Carolina. The study enrolled a total of 53 patients who were free of active fungal diseases at the time of enrollment but at risk of developing one because they were all immunocompromised, e.g., (1) patients positive for HIV and (2) patients post- or awaiting solid organ transplantation. Among 53 patients enrolled, two patients developed invasive cryptococcosis, and in both patients, IgM anti-GlcCer was detected in sera using the ELISA at least 6 weeks prior to the clinical presentation of the brain disease. These results were corroborated by a cryptococcal antigen lateral flow assay, which was also positive in serum prior to the development of meningoencephalitis. However, a high number of positive results were also detected in patients with no evidence of cryptococcosis. This study highlights the potential utility of this new assay in early diagnostic testing algorithms for patients at risk for cryptococcosis, but further investigations are needed to validate the sensitivity and specificity of the glucosylceramide ELISA as a predictor of cryptococcosis. Glucosylceramide Antibody ELISA HIV Cryptococcosis Fungal infection Introduction Invasive fungal infections, such as cryptococcosis, candidiasis and aspergillosis, have dramatically increased in the last two decades and are associated with very poor prognosis [1]. Although significant progress has been made to improve molecular and genetic tools to study their development and progression, studies to provide better tools for an early diagnosis are urgently needed (reviewed in [2]). Our laboratory has focused on understanding the molecular mechanisms of fungal pathogenesis, particularly the role of bioactive lipids in this process, both at a mechanistic and translational level. We discovered that the sphingolipid pathway represents a rich reservoir for signaling molecules [3]. In particular, the reactions regulating complex sphingolipids such as inositol phosphoryl ceramide (IPC) impact virulence at several levels (reviewed in [4]). Additionally, the complex sphingolipid glucosylceramide (GlcCer) plays a crucial role in the survival of Cryptococcus neoformans (Cn) in the extracellular environment such as alveolar spaces. Specifically, GlcCer is required for Cn to cause disease once the fungus is introduced intranasally [5]. As a result, mice infected with mutant cells lacking the enzyme that forms GlcCer (Δgcs1) do not die of Cn infection, and this lack of pathogenicity has been traced to the formation of granulomas, which contain fungal cells in the lung. Thus, the production of GlcCer is required for the fungus to leave the lung and disseminate to the bloodstream and reach the brain. GlcCer is also required for the pathogenicity of other fungal pathogens [6–9]. The host (both mice and humans) responds to the Cn infection with the production of antibody against GlcCer [10], and administration of monoclonal antibody against GlcCer protects mice against lethal cryptococcal infection [11]. Interestingly, following intranasal infection of mice with 103 wild-type Cn cells, we detected IgM antibodies in the bloodstream prior to the dissemination of fungal cells to the brain [12]. Thus, we hypothesize that the detection of serum IgM against fungal GlcCer has future potential as an early diagnostic method of cryptococcosis. To investigate this, we analyzed serum samples of immunocompromised patients for the presence of anti-GlcCer antibody and evaluated the titer in the context of the clinical manifestations and other signs of Cn dissemination. Materials and Methods Patient Population Fifty-three (53) patients were enrolled in the clinical study which ran at the Medical University of South Carolina (MUSC) from January 2009 to January 2011. The sample population comprised the following groups: patients positive for HIV as well as patients who had undergone solid organ transplantation. At the time of enrollment, patients signed a consent document approved by the Institutional Review Board of MUSC to provide 5 ml of blood every 1–3 months. The collected blood was then sent to our laboratory for analysis. Sera was obtained by spinning blood at 3,000×g for 10 min at 4°C and stored at −80°C until used. When diagnosed with cryptococcal meningitis, patients were sampled weekly where possible. Each patient was compensated per sample. ELISA Assay The assay was performed in a 96-well microtiter plate (Maxisorp NUNC). First, the wells were coated with 50 µl/well of 160 µg/ml of soy glucosylceramide (Avanti Polar Lipids, Inc., Cat # 131304) in methanol and incubated overnight at 4°C. The plate was then blocked with 5% BSA in phosphate-buffered saline (PBS), incubated 1 h at 37°C and then washed three times with PBS/0.1% Tween 20 (PBST). Serum samples at a dilution of 1:32 in PBS were added and incubated 1 h at 37°. The positive control, α-GlcCer IgM [12] (0.8 mg/ml) was also diluted 1:32 in PBS. After three washes with PBST, the plate was incubated with 50 µl per well of either goat anti-human IgM-HRP (μ-chain specific) (Sigma Cat # A0420) diluted 1:50,000 with 1% BSA/PBS or goat anti-mouse secondary IgM (μ-chain specific) (Sigma Cat # A8786) diluted 1:30,000 with 1% BSA/PBS, for 1 h at 37°C. After three washes with PBST, the color was developed with 50 µl/well of 3, 3′, 5, 5′–tetramethylbenzidine (TMB) (Sigma Cat # T0440). The reaction was then stopped with 50 µl 2 M H2SO4 and the plate read at 450 nm with a VersaMax plate reader. Cryptococcal Antigen (CrAg) Lateral Flow Assay This assay is an immunochromatographic test system for the qualitative detection of capsular polysaccharide antigens of Cn in serum and cerebrospinal fluid (CSF). The strips for this test were obtained from Immuno Mycologics, Inc (IMMY; Norman, Oklahoma), and the test was performed according to the manufacturer’s directions at a dilution of 1:2, except for patient 23 whose samples were analyzed at a dilution of 1:2048. The presence of two lines, corresponding to Test and Control, indicates a positive result. A single control line indicates a negative result. Results and Discussion Over a period of 2 years, it was anticipated that patients would be enrolled at MUSC from five groups for the study: (1) patients positive for HIV (HIV+) (2) recipients of solid organ transplantation (SOT) (3) patients with lympho-proliferative disorders including leukemia and lymphoma (4) recipients of bone marrow transplantation and (5) patients immunosuppressed or receiving long-term immunosuppressive therapy for conditions such as sarcoidosis and connective tissue diseases. It was imperative that none of the patients enrolled in the study had any prior history of cryptococcosis or any other active fungal infections at the time of enrollment. While neutropenic patients are usually not at particular risk for developing cryptococcal infection, patients receiving immunosuppressive therapy and AIDS patients are more susceptible to cryptococcosis due to low (<50) CD4 T cell counts [13, 14]. A count less than 200 indicates the immune system is severely weakened, and the HIV+ person is at a much greater risk of opportunistic infections [15]. However, despite low levels of serum immunoglobulins in immunocompromised patients, normal or near normal antibody production has been observed [16, 17]. Actual enrollment numbers (Fig. 1) highlight the difficulty of patient enrollment in groups other than (1) and (2) due to logistical difficulties of gaining access to patients in other clinics beyond the infectious disease clinic. While clinicians were willing to allow recruitment from the other patient populations, conflicts with clinic space and scheduling impaired the ability to recruit from other groups. 79% of patients were HIV+, while the remaining 21% were patients who had received SOT, predominantly kidney or liver transplants (Fig. 1a). This latter group was also predominantly Caucasian. The breakdown of racial categories over both groups was as follows: male Caucasian 19%, male African–American 53%, male Asian 2%, female Caucasian 6% and female African–American 20% (Fig. 1b). These numbers can be considered to be somewhat representative of the HIV+ and transplantation communities in Charleston, South Carolina, but actual numbers, particularly among the African–American community, may be greater. All the patients in the HIV+ group had CD4 counts less than 200 except patients 16 and 18, whose CD4 counts were 220 and 292 counts, respectively, on enrollment. Those patients whose viral loads were <40 or not detected (nd) copies per milliliter had been compliant with HAART. For the ELISA, samples were diluted 1:32 in PBS. Absorbance cutoff signifying a positive IgM anti-GlcCer reading was set at 0.200 AU based on data from our previous studies in mice [12], as this mammalian model was considered a “true” negative control in being free of any previous (asymptomatic or symptomatic) fungal infections. In fact, since GlcCer is produced by a variety of fungi, such as Candida albicans, Cryptococcus neoformans, Aspergillus fumigatus and several dimorphic fungi (reviewed in [18], and [19]) it is expected that humans may develop IgM anti-GlcCer also when infected by these fungi. In our study, out of the fifty-three enrolled patients (Fig. 2), thirty were identified as positive for IgM anti-GlcCer from the ELISA, while sixteen were negative and seven were lost to follow-up (Fig. 2), often right after the enrollment visit. Of the thirty positive for IgM anti-GlcCer, twenty-one had a positive IgM anti-GlcCer reading for 2–6 months via the ELISA, whereas nine started showing a positive IgM anti-GlcCer value only in the last couple of months of the study. For these patients in the latter group, it remains to be seen whether they will develop a fungal disease. Of the twenty-one positive for IgM anti-GlcCer, two were diagnosed with cryptococcal meningitis and we will discuss these two patients in greater detail. Given the high number of positives, we examined further why this was the case and found three of these patients, predominantly male, had a history of oral candidiasis. For the other sixteen positive patients, we could not identify any previous fungal infections. However, all sixteen of these patients showed a relatively low positiveness (i.e., absorbance values between 0.200 and 0.400 AU), compared to the two patients who were diagnosed with cryptococcal meningitis (AU 0.300–0.700) suggesting that the test needs a better standardization with a much larger group. As mentioned, two patients (patient 23 and patient 11) were diagnosed with cryptococcal meningitis during the course of this study and their IgM anti-GlcCer results from the ELISA correlated well with the progression of disease, especially for the patient who was compliant with his highly active antiretroviral therapy (HAART) outside of the hospital setting. Patient 23 is a 62-year-old African–American male who joined the study in August 2009. He has a history of HIV, hepatitis, candidal esophagitis and chronic alcohol abuse. He presented with severe headache on visit 3 (December 2009), which had been ongoing for 2 weeks, with occasional nausea and vomiting. At that time, he tested positive for cryptococcal antigen in his cerebrospinal fluid and had an absolute CD4 count of only 2 (normal values are 500–1400) (Table 1). In fact, his CD4 counts were extremely low for the duration of this study, even though he was on HAART during this period. While his CD4 numbers did improve, they were still significantly below normal values. The patient’s white blood cell counts remained in the normal range (4.800–10.800 K), when he remained compliant with his HAART regimen, as well as amphotericin B and flucytocine. During the compliancy period, the IgM anti-GlcCer results remained below the positivity cutoff of 0.200 AU. This continued for almost a year (Table 1). However, it is possible that the patient had already developed an asymptomatic Cn lung infection at the time of enrollment, given the positive results of the IgM anti-GlcCer at enrollment and early sampling points. The CrAg assay results were positive for all the serum samples obtained for patient 23, suggesting again that either the patient had a low titer of capsular material in the bloodstream throughout the observation period and/or the CrAg assay may have issues of cross-reactivity. Studies are underway to study this potential complication against several Candida species, as well as Mycobacterium tuberculosis, Pneumocystis carinii, hepatitis A and C viruses, Staphyloccus aureus and Streptococcus pneumoniae. Thus, in this patient, the ELISA results overall trended well with severity of symptoms. Unfortunately, his viral load and IgM anti-GlcCer readings had gone back up at the end of the study when he was no longer complaint with HAART and/or with his antifungal maintenance therapy. Patient 11 is a 32-year-old African–American male who was in the study since its commencement in January 2009. He has a history of HIV/AIDS and no record of any previous fungal infection. He presented with positive serum cryptococcal antigen test at visit 7 (December 2009; Table 2) during the study after he had attended and been tested at an infectious disease clinic. At this time, he had endorsed a 2-week history of calvarial headaches, nausea and vomiting, as well as fevers and chills, and blurry vision. He was discharged following treatment with the diagnoses of cryptococcal meningitis, oral candidiasis, acute renal failure and AIDS. He had been on HAART regimen since July 2009, his CD4 counts were low (11), and his WBC count was at the lower end of the normal range, fluctuating between 4.5 and 5 K. We found that all samples but the first three collected from patient 11 were positive for CrAg (Table 2). Interestingly, also IgM anti-GlcCer antibody was negative on the first three samples and became negative again on visit 9 and 23. His viral load remained high and CD4 counts remained low throughout the study further suggesting the lack of compliance to HAART. Interestingly though, the IgM anti-GlcCer (Table 2) does not show a definitive trend as was observed for patient 23, as patient 11 did not remain compliant with his anti-retroviral regimen or his antifungal maintenance therapy over the course of the clinical study. This may suggest that a negativization of the IgM anti-GlcCer value may be a useful indicator that the patient is compliant and responding to the therapy. Unfortunately, several patients were positive for IgM anti-GlcCer antibody even if they were “clinically free” of cryptococcosis or any other fungal infection, suggesting that this test has a very low specificity and a low negative-predictive value. None of the patients enrolled in our study who remained free of cryptococcal meningitis showed positive CrAg assay results. Thus, the overall utility of the IgM anti-GlcCer ELISA as a potential diagnostic method, while it reliably predicted the onset of cryptococcal meningitis in two patients, is hampered by the fact that it was highly positive in patients with no signs or symptoms of active fungal infections. This may be due to cross-reactivity issues, and plans are underway to use the ELISA for the detection of IgM anti-GlcCer antibody against several other fungal and bacterial pathogens. Alternatively, the absorbance cutoff value of 0.200 AU could be increased once the test has been standardized with a much larger group of patients to get a true cutoff value in humans. In conclusion, the IgM anti-GlcCer assay reliably predicted the onset of cryptococcal meningitis in two patients, in which the IgM antibody anti-GlcCer appears to correlate well with the severity of the cryptococcal disease. Clearly, additional cases need to be evaluated in order to assess whether this method could be a predictor and/or a prognostic factor of cryptococcosis or other fungal diseases, given the high number of patients who tested positive in the ELISA despite having no evidence of cryptococcosis. This work was supported in part by the Burroughs Wellcome Fund and in part by the National Institutes of Health (grants AI56168, AI78493, AI71142 and AI87541 to M.D.P.). Dr Maurizio Del Poeta is a Burroughs Wellcome New Investigator in the Pathogenesis of Infectious Diseases. Fig. 1 Breakdown of patients enrolled in the clinical study at MUSC. a Patients were enrolled from 2 groups: HIV-positive (HIV+) and solid organ transplantation (SOT). b Racial breakdown of patients representing the immunocompromised population in Charleston SC Fig. 2 Flowchart showing that 4% of patients enrolled in the study developed cryptococcal meningitis; however, several other patients had positive results in the ELISA, indicating the need for further validation of the assay Table 1 IgM anti-GluCer results, CrAg assay data, CD4 counts and white blood cell (WBC) counts for patient 23 where CD4 and WBC counts are missing, and these were not available on the day of the blood draw Visit number Date IgM anti-GlcCer (abs 450 nm) CrAg result CD4/CUMM HIV 1RNA-viral load (copies/ml) WBC count (K) 1 8/19/2009 0.294 + 6 73003 3.87 2 11/23/2009 0.391 + 4 29379 3.42 3 12/11/2009 0.39 + 2 3437 6.31 4 12/22/2009 0.155 + 6.43 5 12/31/2009 0.18 + <40 6 1/15/2010 0.169 + 7 3/15/2010 0.147 + 3.05 8 4/28/2010 0.129 + 72 0 9 10/20/2010 0.137 + 55 274 3.91 10 1/19/2011 0.45 + 27 42460 3.12 Table 2 IgM anti-GluCer results, CrAg assay data, CD4 counts and white blood cell (WBC) counts for patient 11 where CD4 and WBC counts are missing, and these were not available on the day of the blood draw Visit number Date IgM anti-GlcCer (abs 450 nm) CrAg result CD4/CUMM HIV 1RNA-viral load (copies/ml) WBC count (K) 1 1/13/2009 0.12 − 2.53 2 4/27/2009 0.048 − 3 7/22/2009 0.057 − 4.92 4 9/25/2009 0.067 + 5 10/26/2009 0.464 + 12 80994 5 6 11/30/2009 0.372 + 7 12/14/2009 0.68 + 11 4.78 8 12/21/2009 0.344 + 4.49 9 12/30/2009 0.13 + 3.78 10 1/6/2010 0.392 + 11 1/20/2010 0.461 + 12 1/27/2010 0.295 + 13 2/1/2010 0.307 + 14 2/10/2010 0.652 + 6 56116 3.5 15 2/17/2010 0.516 + 16 2/26/2010 0.278 + 17 3/31/2010 0.491 + 18 4/30/2010 0.294 + 5 67269 2.9 19 5/26/2010 0.41 + 20 6/21/2010 0.291 + 21 7/22/2010 0.392 + 22 9/20/2010 0.469 + 5 40845 3.93 23 10/20/2010 0.187 + 24 11/22/2010 0.393 + 25 12/28/2010 0.524 + 5 118897 3.12 26 1/31/2011 0.384 + References 1 Perfect JR Casadevall A Heitman J Filler SG Edwards JE Jr Mitchell AP Fungal molecular pathogenesis: what can it do and why do we need it? Molecular principles of fungal pathogenesis 2006 Washington DC ASM Press 3 11 2 Diaz MR Nguyen MH Heitman J Kozel TR Kwon-Chung KJ Perfect JR Casadevall A Diagnostic approach based on capsular antigen, capsule detection, beta-glucan and DNA analysis Cryptococcus: from human pathogen to model yeast 2011 Washington DC ASM Press 3 Luberto C Toffaletti DL Wills EA Tucker SC Casadevall A Perfect JR Roles for inositol-phosphoryl ceramide synthase 1 (IPC1) in pathogenesis of C. neoformans Genes Dev 2001 15 2 201 212 11157776 4 Singh A Del Poeta M Lipid signalling in pathogenic fungi Cell Microbiol 2010 5 Rittershaus PC Kechichian TB Allegood JC Merrill AH Jr Hennig M Luberto C Glucosylceramide synthase is an essential regulator of pathogenicity of Cryptococcus neoformans J Clin Invest 2006 116 6 1651 1659 16741577 6 Pinto MR Rodrigues ML Travassos LR Haido RM Wait R Barreto-Bergter E Characterization of glucosylceramides in Pseudallescheria boydii and their involvement in fungal differentiation Glycobiology 2002 12 4 251 260 12042248 7 Oura T Kajiwara S Disruption of the sphingolipid Delta8-desaturase gene causes a delay in morphological changes in Candida albicans Microbiology 2008 154 Pt 12 3795 3803 19047747 8 Oura T Kajiwara S Candida albicans sphingolipid C9-methyltransferase is involved in hyphal elongation Microbiology 2010 156 Pt 4 1234 1243 20019081 9 Noble SM French S Kohn LA Chen V Johnson AD Systematic screens of a Candida albicans homozygous deletion library decouple morphogenetic switching and pathogenicity Nat Genet 2010 42 7 590 598 20543849 10 Rodrigues ML Travassos LR Miranda KR Franzen AJ Rozental S de Souza W Human antibodies against a purified glucosylceramide from Cryptococcus neoformans inhibit cell budding and fungal growth Infect Immunol 2000 68 12 7049 7060 11083830 11 Rodrigues ML Shi L Barreto-Bergter E Nimrichter L Farias SE Rodrigues EG Monoclonal antibody to fungal glucosylceramide protects mice against lethal Cryptococcus neoformans infection Clin Vaccine Immunol 2007 14 10 1372 1376 17715331 12 Rhome R Singh A Kechichian T Drago M Morace G Luberto C Surface localization of glucosylceramide during Cryptococcus neoformans infection allows targeting as a potential antifungal PLoS One 2011 6 1 e15572 21283686 13 Bender BS Davidson BL Kline R Brown C Quinn TC Role of the mononuclear phagocyte system in the immunopathogenesis of human immunodeficiency virus infection and the acquired immunodeficiency syndrome Rev Infect Dis 1988 10 6 1142 1154 3206059 14 Diamond RD Erickson NF 3rd Chemotaxis of human neutrophils and monocytes induced by Cryptococcus neoformans Infect Immunol 1982 38 1 380 382 6754617 15 Young B Dao CN CBuchacz K Baker R Brooks JT Increased rates of bone fracture among HIV-infected persons in the HIV outpatient study (HOPS) compared with the US general population, 2000–2006 Clin Infect Dis 2010 52 8 1061 1068 16 Irwin M Low CD4 counts: a variety of causes and their implications to a multi-factorial model of AIDS Br Med J online 2001 17 Dropulic LK Lederman HM Hayden RT Carroll KC Tang Y-W Wolk DM Overview of infections in the immunocompromised host Diagnostic microbiology of the immunocompromised host 2009 Washington DC ASM Press 18 Rhome R Del Poeta M Lipid signaling in pathogenic fungi Annu Rev Microbiol 2009 63 119 131 19450140 19 Warnecke D Heinz E Recently discovered functions of glucosylceramides in plants and fungi Cell Mol Life Sci 2003 60 5 919 941 12827281
PMC005xxxxxx/PMC5125077.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 9200019 21021 Cardiol Young Cardiol Young Cardiology in the young 1047-9511 1467-1107 26345374 5125077 10.1017/S1047951115001389 NIHMS830848 Article CHD associated with syndromic diagnoses: peri-operative risk factors and early outcomes Landis Benjamin J. Cooper David S. Hinton Robert B. Heart Institute, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio, United States of America Correspondence to: B. J. Landis, MD, Division of Pediatric Cardiology, Indiana University School of Medicine, Indianapolis, IN 46202, United States of America. Tel: 317 278 2807; Fax: 317 274 8679; [email protected] 20 11 2016 8 9 2015 1 2016 28 11 2016 26 1 3052 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. CHD is frequently associated with a genetic syndrome. These syndromes often present specific cardiovascular and non-cardiovascular co-morbidities that confer significant peri-operative risks affecting multiple organ systems. Although surgical outcomes have improved over time, these co-morbidities continue to contribute substantially to poor peri-operative mortality and morbidity outcomes. Peri-operative morbidity may have long-standing ramifications on neurodevelopment and overall health. Recognising the cardiovascular and non-cardiovascular risks associated with specific syndromic diagnoses will facilitate expectant management, early detection of clinical problems, and improved outcomes – for example, the development of syndrome-based protocols for peri-operative evaluation and prophylactic actions may improve outcomes for the more frequently encountered syndromes such as 22q11 deletion syndrome. CHD syndrome genetic CHD is present in 3–12 in 1000 births, but the incidence may be as high as 5% when strictly including all cardiovascular malformations such as bicuspid aortic valve.1–5 The genetic basis of CHD is well established4 – for instance, the Baltimore–Washington Infant Study in 1989 reported chromosomal abnormalities in nearly 13% of infants with CHD.6 More recent studies have observed that 20–30% of infants with CHD have a recognised genetic syndrome or significant non-cardiovascular anomaly.5,7,8 Even among patients with isolated CHD, there is evidence for heritability and increased familial recurrence risk that may be particularly important for certain classes of CHD such as heterotaxy, left ventricular outflow tract obstructive lesions, and atrioventricular septal defects.9,10 In a minority of cases, gene mutations in NKX2–5, GATA4, and NOTCH1 have been observed in families demonstrating Mendelian inheritance.11–13 With the advancement of genetic technologies including DNA microarray and high-throughput sequencing platforms detection of genetic causes of CHD continues to grow rapidly.14–16 It is critical that clinicians recognise the clinical relevance of a genetic diagnosis in order to improve outcomes, not only for syndromic patients but also for all CHD patients with informative genotypes. The peri-operative time period exposes patients to risk for significant complications that may have both immediate and long-term repercussions, including quality of life or neurocognitive outcomes.17,18 The aims of this review were to present the spectrum of peri-operative risks for patients with a genetic syndrome and CHD, comprehensively organise observations about the outcomes of patients with genetic syndromes, and synthesise our current understanding of the genetic basis of CHD as a tool for informing the peri-operative management of these patients. Advances in cardiac surgery, catheterisation, and intensive care have significantly reduced mortality associated with CHD,19 shifting the focus towards minimising short- and long-term morbidity. There are well-recognised peri-operative risks for all children undergoing cardiac surgery, including but not limited to myocardial dysfunction, arrhythmias, respiratory failure, infection, bleeding, thrombosis, kidney injury, and neurological injury.20 However, the CHD sub-population with syndromic disease often has important non-cardiovascular and functional – that is, non-structural – cardiovascular abnormalities that significantly modify these routine peri-operative risks or present additional risks that contribute to morbidity and mortality. It is certain that the cardiac surgeon, anaesthesiologist, intensivist, and cardiologist will frequently encounter children with a syndromic disorder. To our knowledge, the specific peri-operative risks that exist for patients with CHD and genetic syndromes have not previously been consolidated into a single source. Many large studies have enrolled syndromic patients to broadly evaluate the impact of a syndromic diagnosis on surgical outcomes Widely inclusive studies, which have analysed all types of paediatric cardiac surgical operations together, have observed that a syndromic diagnosis may not impact early operative mortality but does predispose to post-operative complications contributing to prolonged hospital length of stay.21–24 However, batching all types of CHD in this manner provides limited insight into risk factors, as both the genetic basis and the risk profiles of different cardiac lesions vary. Sub-classes of cardiac lesions that have been studied specifically include critical left ventricular outflow tract obstructive lesions and conotruncal defects. Detailed information about these studies, including study types, enrollment numbers, cardiac and genetic diagnoses, and early mortality and morbidity outcomes, is provided in Supplementary Table S1. Patel et al25 extensively reviewed early post-operative outcomes data for hypoplastic left heart syndrome/critical left ventricular outflow tract obstruction from both the Society of Thoracic Surgeons – ~1200 Norwood operations from 2002 to 2006 – and the Congenital Heart Surgeons’ Society ~700 stage 1 palliations from 1994 to 2001 – databases. In the Society of Thoracic Surgeons database, 15% of patients were documented to have a “genetic and/or significant non-cardiovascular abnormality”, which was associated with increased in-hospital mortality (26.7 versus 19.8%). Similarly, in the Congenital Heart Surgeons’ Society database, 8% had a “non-cardiac congenital abnormality or syndrome”, which was associated with increased early risk of mortality. These mortality data are consistent with two other single-centre reports (together 310 patients)26,27 and with data from the Pediatric Heart Network’s Single Ventricle Reconstruction trial including 549 patients undergoing Norwood operations.28 This evidence is countered only by a single series of 158 patients who underwent Norwood operation.29 The Society of Thoracic Surgeons data demonstrate that in-hospital mortality was not increased after stage 2 (~700 operations) or stage 3 palliations (~550 operations), recognising that stage 1 mortality may limit interpretation.25 Increased morbidity was observed after all stages of palliation.25,30 Michielon et al31 provided important perspective in a cohort of nearly 800 patients with conotruncal defects – tetralogy of Fallot with or without pulmonary atresia, double-outlet right ventricle, truncus arteriosus, or interrupted aortic arch – undergoing biventricular repair from 1992 to 2007. Uniquely, nearly every patient in the cohort (96%) underwent clinical evaluation by a geneticist and prospective molecular screening (93%) for 22q11 deletion or aneuploidy. A genetic diagnosis was established in ~27% of these patients and was associated with increased hospital mortality (17 versus 7%) and prolonged duration of intensive care. These findings were consistent with previous observations in 266 patients with tetralogy of Fallot with normal pulmonary artery anatomy.32 Similarly, a cohort of 350 patients with conotruncal defects undergoing primary or staged repair trended towards increased early mortality.33 Taken together, the presence of a genetic syndrome may negatively impact early post-operative survival, particularly in the context of more complex cardiac operations such as the Norwood operation. It is particularly clear that post-operative morbidity risk is consistently elevated across the spectrum of cardiac lesions. These are very important observations, but are based on data from heterogeneous groups of genetic syndromes, which limit generalisability to specific syndromes. Moreover, batching patients with non-cardiovascular malformations lacking a defined genetic syndrome together with those who have a defined genetic syndrome creates challenges. In order to understand the risk factors and clinically intervene to improve outcomes, more precise data are required. To this end, the remainder of this article focuses on outcomes and risk factors for specific syndromic CHD populations. The presence of a specific genetic syndrome impacts early peri-operative outcomes, and genetic syndromes often present with specific features posing significant peri-operative risks Down syndrome Down syndrome is present in at least one in 1000 live births and is caused by trisomy of chromosome 21 due to true aneuploidy, unbalanced translocation, or mosaicism.34,35 Approximately 40–50% of patients with Down syndrome present with CHD, most frequently atrioventricular septal defect, followed by ventricular septal defect, atrial septal defect, patent ductus arteriosus, and tetralogy of Fallot.34,36 Survival after cardiac surgery is generally favourable, as summarised in Table 1, with more detailed information in Supplementary Table S2; three large contemporary database reviews – encompassing a spectrum of cardiac operations and cumulatively including nearly 7000 patients with Down syndrome – demonstrated that in-hospital mortality risk decreased (Healthcare Cost and Utilization Project Kids’ Inpatient Database)37,38 or was not different (Society of Thoracic Surgeons database)39 when compared with children without Down syndrome. Cardiac lesions studied specifically in Down syndrome are atrioventricular septal defects, conotruncal defects (primarily tetralogy of Fallot), and single ventricle lesions. Poor outcomes after repair of atrioventricular septal defects were reported in early surgical eras,40,41 but recent evidence indicates that children with Down syndrome undergoing biventricular repair for complete atrioventricular septal defect have better37,42 or similar early mortality rates43–46 compared with patients without Down syndrome. Re-operation rates may be lower in Down syndrome, likely related to less complex atrioventricular valve and outflow tract anatomy.42–44,46 Increased risk for post-operative complete heart block is reported after ventricular septal defect repair39,47 but not after atrioventricular septal defect repair.42,48 Similar to complete atrioventricular septal defect repair, Down syndrome does not significantly impact early mortality after surgery for tetralogy of Fallot32,37,39,41,49 or conotruncal defects collectively (predominantly tetralogy of Fallot).31,33 In contrast, Down syndrome may significantly worsen outcomes for single ventricle lesions. Review of the Kids’ Inpatient Database found that early mortality was increased both after systemic-to-pulmonary shunt placement and after stage 2 palliation.37 Review of the Society of Thoracic Surgeons database also demonstrated increased hospital mortality for all stages of single ventricle palliation.39 Increased mortality (35%) after stage 3 palliation was observed in the Pediatric Cardiac Care Consortium database50 but was not corroborated by the Kids’ Inpatient Database or a smaller single-centre series.37,51 The reasons for poor outcomes after single ventricle palliations in these patients are undefined but likely related to predisposition for pulmonary hypertension, which may also contribute to prolonged hospitalisation after stage 2 and stage 3 palliations.39,51 Many features of Down syndrome impact peri-operative morbidity. Pulmonary and pulmonary vascular co-morbidities feature prominently (Table 2 and Supplementary Table S3). Congenital respiratory tract anomalies may be present at multiple levels and include macroglossia/glossoptosis, adenotonsillar hypertrophy, sub-glottic stenosis, laryngomalacia, tracheal stenosis, complete tracheal rings, and tracheobronchomalacia. Hypotonia can exacerbate anatomical narrowing. Patients are at risk for pulmonary hypertension due to chronic hypoventilation related to airway obstruction and sleep apnoea as well as intrinsic risk for pulmonary vascular disease.52–54 Craniofacial and upper airway anomalies can complicate peri-operative airway management and/or performance of trans-oesophageal echocardiography.55–57 Pulmonary abnormalities include pulmonary hypoplasia, interstitial lung disease secondary to chronic aspiration or infection, tracheal bronchus predisposing to recurrent right upper lobe collapse or pneumonia, sub-pleural cysts predisposing to pneumothorax, and lymphatic abnormalities including pulmonary lymphangiectasia.58–63 These airway co-morbidities manifest clinically as increased risk for post-operative respiratory complications,39,48,64 prolonged mechanical ventilation,51,64,65 pneumothorax,48 chylothorax,22,39 chylopericardium,66 and failed extubation.67 These observations mandate vigilant assessment and treatment of the pulmonary status in the post-operative period, which may be optimised by pre-operative consultation and testing, particularly in high-risk patients – for example, single ventricle lesions. Dysfunction of B- and T-lymphocytes and neutrophils may predispose to infections and exacerbate the inflammatory response to cardiopulmonary bypass.22,39,48,51,68–71 Congenital hypothyroidism occurs in ~1%, and thyroid screening at regular intervals, including at ages 6 and 12 months, is indicated because an additional 4–18% develop hypothyroidism.34,72,73 Pre-operative thyroid screening is indicated so that hypothyroidism can be treated pre-operatively. As thyroid levels decrease with cardiopulmonary bypass surgery and impact myocardial function and cardiovascular stability,74,75 intra-operative and post-operative parenteral therapy may be indicated. The risk for atlantoaxial instability calls for appropriate peri-operative precautionary measures to avoid neurological injury, especially in mid-to-late childhood.34,76 Increased risk for seizures – ~ 8% in the general Down syndrome population – should also be considered.77 Taken together, Down syndrome presents significant co-morbidities that can impact peri-operative outcomes. Fortunately, mortality outcomes have improved over time for the most-frequent lesions, but non-cardiovascular abnormalities continue to contribute to post-operative morbidity outcomes and require clinical vigilance and future research. 22q11 deletion syndrome Microdeletion of 22q11.2 causes several disorders with overlapping clinical phenotypes including DiGeorge syndrome, velocardiofacial syndrome, and conotruncal anomaly face syndrome, and is present in approximately one in 5000 live births.78,79 Suggestive features include long narrow face and small protuberant ears with thick and crumpled helices.80 CHD is present in at least 75%.81 The typical cardiac lesions are conotruncal defects and abnormalities of the aortic arch and brachiocephalic arteries, including type B interrupted aortic arch, truncus arteriosus, tetralogy of Fallot, pulmonary atresia with ventricular septal defect, isolated ventricular septal defect, and abnormal aortic arch sidedness and/or branching.82–84 Peri-operative outcomes are summarised in Table 1 with more detailed information in Supplementary Table S4. Early reports observed very high operative mortality in neonates with DiGeorge syndrome.85 Although increased hospital mortality was also observed in a more contemporary series of patients with conotruncal defects,33 there is strong evidence that 22q11 deletion no longer results in early mortality for the vast majority of cardiac lesions;31,32,86–88 however, substantial post-operative morbidity persists including slow recovery and increased frequency of cardiac events such as the need for re-operation.31,32,86,87 Notably, patients with pulmonary atresia with ventricular septal defect and major aortopulmonary collateral arteries have consistently demonstrated increased early mortality in the setting of 22q11 deletion.89–93 In addition, early operative mortality after Norwood stage 1 operation was observed in two of five patients in the Congenital Heart Surgeons’ Society database from 1994 to 2001, supporting the concept that genetic syndromes continue to impact high-risk operations.25 Congenital malformations including cleft palate, sub-mucous clefts, retrognathia, Pierre Robin sequence, congenital laryngeal web, and vascular ring may complicate airway management.81,94,95 Bronchomalacia and bronchospasm have been observed in patients with 22q11 deletion and pulmonary atresia with ventricular septal defect, which may be related to compression by aortopulmonary collateral vessels.96,97 Although prolonged mechanical ventilation was not observed after unifocalisation of the major aortopulmonary collateral arteries,98 increased post-operative respiratory complications including prolonged intubation and post-extubation stridor have been observed.99 Thymic aplasia occurs rarely (<1% of cases) and is associated with severe immune deficiency. More commonly, thymic hypoplasia causes mild-to-moderate immune deficiency. Complete preoperative immunological evaluation and blood product precautions – cytomegalovirus-negative and irradiated blood products – are indicated for all cases to prevent iatrogenic infection and graft versus host disease.80,85,100 Low T-lymphocyte counts are present in 75–80% of patients with 22q11 deletion.101 B-lymphocyte dysfunction with immunoglobulin deficiency also may have clinical significance.102,103 Frequent infectious complications including fungal infections have been reported23,85,88,91,99 but not uniformly.86,87,90 It has been suggested that prophylaxis with broad-spectrum antibiotics including antifungal agents may be indicated.104 Developmental hypoplasia of the parathyroid glands results in hypocalcaemia in 40–80% of patients and is often accompanied by hypomagnesaemia.105 Close peri-operative electrolyte monitoring is necessary to preserve cardiac function, avoid dysrhythmia, and prevent secondary seizures. Peri-operative seizures are linked to worse neurodevelopmental outcomes in 22q11 deletion.106 Annual assessment of thyroid function is recommended because hypothyroidism is present in 20–30% of patients; a routine preoperative thyroid screening approach similar to Down syndrome may be reasonable.80,105,107 Interestingly, the gene encoding glycoprotein Ib (GP1BB), which is responsible for autosomal-recessive Bernard–Soulier disease, is located within the 22q11 region. Patients with 22q11 deletion, and thus hemizygous deletion of GP1BB, may have abnormally large platelets and thrombocytopaenia (macrothrombocytopaenia).108,109 Platelet dysfunction has been described previously.110,111 Post-operative bleeding accounted for a significant proportion of post-operative deaths in patients with pulmonary atresia with ventricular septal defect.90,93 A complete haematological workup may be indicated before operations requiring small vessel anastomoses – for example, unifocalisation – and unexplained severe post-operative bleeding should trigger concern for Bernard–Soulier disease due to mutation of the nondeleted GP1BB allele.112 Renal and urinary tract abnormalities are present in 30–40% of patients, including renal agenesis, multi-cystic dysplastic kidneys, hydronephrosis, and vesicoureteral reflux.81,113 Increased need for post-operative dialysis has been observed.88 Autonomic dysfunction may in some cases explain post-operative hypotension refractory to usual therapy.114 Taken together, the developmental abnormalities associated with 22q11 deletion likely contribute to mortality after complex operations and morbidity across the spectrum of CHD surgery. Improvements in anticipatory management of common abnormalities – for example, immune dysfunction and hypocalcaemia – will continue to improve outcomes. Abnormalities that are less frequently recognised – for example,. haematological dysfunction – should be anticipated and acted upon when deviation from expected recovery is encountered. Heterotaxy syndrome Heterotaxy syndrome, a disorder of laterality characterised by abnormal thoracoabdominal situs, is frequently associated with CHD and is present in at least one in 10,000 live births.115 Mutations in genes such as DNAH5, ZIC3, CFC1, NODAL, ACVR2B, DNAI1, and LEFTY2, many of which are components of the Nodal signal transduction pathway, have been identified;116 familial recurrence is more frequently observed compared with other cardiac lesions.9 CHD is often complex, including complete atrioventricular canal defect, anomalous pulmonary and systemic venous return, and pulmonary outflow tract obstruction. Heterotaxy can be sub-classified as right atrial isomerism versus left atrial isomerism as determined by atrial appendage and bronchopulmonary anatomy.117 In general terms, right atrial isomerism typically has more severe CHD, often requires single ventricle palliation, and has worse survival in childhood.118–120 In right atrial isomerism, abnormal morphology and function of the sinoatrial node and the atrioventricular conduction system predispose to both tachyarrhythmia and bradyarrhythmia121–124 – for instance, supra-ventricular tachycardia has been observed in up to 25% of cases, including re-entrant mechanisms mediated by twin atrioventricular nodes.125–127 Atrioventricular block and sinus node dysfunction are more frequently observed in left atrial isomerism.125,127 In addition to arrhythmia concerns, non-compaction cardiomyopathy is described and may contribute to unexpected ventricular dysfunction.128 Peri-operative outcomes in heterotaxy are summarised in Table 1 and Supplementary Table S5. The complexities of both cardiovascular and noncardiovascular abnormalities likely contribute to poor outcomes.120 Increased mortality following any cardiac surgery has been observed in the Society of Thoracic Surgeons’ database.129,130 Mortality after initial single ventricle palliation is reported to range from 10 to 23%.121,129,131,132 In the setting of total anomalous pulmonary venous return, poor outcomes may be related to hypoplastic pulmonary veins and increased pulmonary vascular reactivity.131–133 Despite these challenges, there was similar survival between heterotaxy and non-heterotaxy patients undergoing primary repair for total anomalous pulmonary venous return but increased need for pulmonary vein re-operation.134 Mortality rates after stage 3 palliation ranged widely from as high as 19–43%123,135,136 to as low as 3–4% in recent studies.123,124,129 Complex anatomy can potentially complicate cardiac transplantation but did not impact early (or late) graft survival;137 however, early mortality was recently reported in two of five patients undergoing cardiac transplantation.138 Overall, there is strong evidence that heterotaxy confers significant peri-operative mortality risk. Post-operative respiratory morbidity was frequently observed.130 Up to 40% of patients with heterotaxy and CHD have dysfunctional airway cilia similar to primary ciliary dyskinaesia.139 Indeed, ciliopathy is a suspected developmental mechanism for cardiovascular and non-cardiovascular malformations.116 Respiratory ciliary dysfunction, diagnosed by nasal nitric oxide levels or nasal video microscopy, has been associated with post-operative respiratory complications, including failed extubation, respiratory failure, respiratory infection, stridor, pleural effusion, atelectasis, pneumothorax, or pulmonary oedema, as well as with the need for tracheostomy.140 It has been suggested that beta-agonist therapy may be effective by improving ciliary motility.140,141 Splenic abnormalities including asplenia (often left atrial isomerism) polysplenia (often right atrial isomerism) or the presence of accessory splenule are frequently observed.142 Asplenia clearly increases risk of bacterial infections in children.143 Splenic function in the setting of polysplenia may also be impaired and should be evaluated using scintigraphy.144 Sepsis was the cause of early post-operative mortality in 13% of deaths in a large heterotaxy population.145 Oropharyngeal malformations including micrognathia, choanal atresia, and cleft lip/palate can contribute to airway management difficulties.146,147 Renal anomalies including renal agenesis, cystic malformation, and horseshoe kidney are also frequently observed.132,147 The surgical outcomes in heterotaxy are improving, but persistent challenges include complex anatomy such as abnormal cardiac position, hypoplastic and anomalous pulmonary veins, and single ventricle morphology, predisposition for arrhythmia, and pulmonary and immunological dysfunction. Turner syndrome Turner syndrome occurs in approximately one in 2000 female live births and is caused by complete or partial absence of the X chromosome.148,149 Features include short stature, ovarian dysgenesis, webbed neck, low posterior hairline, and widely spaced nipples.150 There is a high rate of foetal mortality, often in the setting of foetal hydrops.151 Those surviving to birth often have cardiovascular malformations including bicuspid aortic valve, coarctation of the aorta, partial anomalous pulmonary venous return, persistent left superior caval vein, and hypoplastic left heart syndrome.152–155 Turner syndrome accounts for at least 5% of coarctation of the aorta among girls, which may indicate karyotype screening of all female neonates with coarctation.156 There is also significant long-term risk of aortic dilation and dissection that is likely under-recognised.157,158 Electrocardiographic abnormalities including prolonged QT interval are frequently encountered, but risk of life-threatening arrhythmia has not been established.159 Turner syndrome does not appear to increase mortality risk after repair of coarctation of the aorta but has been associated with longer hospitalisation (Table 1 and Supplementary Table S6).160 By comparison, mortality appears to be significantly increased in patients with hypoplastic left heart syndrome – for instance, 9 out of 11 infants with Turner syndrome undergoing Norwood stage 1 operation died by 4 months of age as per the Congenital Heart Surgeons’ Society database.25 In a retrospective single institution study, 8 out of 10 infants with Turner syndrome undergoing stage 1 palliation for hypoplastic left heart syndrome died before stage 2 operation, and both the survivors were mosaic XO.161 In a more recent series, all four patients with Turner syndrome undergoing stage 1 palliation survived to hospital discharge, but three were reported to have died before stage 3 palliation.160 A precise explanation for these outcomes has not been established thus far, but lymphatic abnormalities may contribute.161 Automatic karyotype screening in girls with hypoplastic left heart syndrome may be indicated because some features develop over time or may be subtle in mosaic cases. Predisposition to vascular complications were described in earlier case series that reported significant post-operative haemorrhage and risk for aortic rupture, possibly related to increased arterial tissue fragility and peri-operative systemic hypertension.162,163 Fortunately, improvements in surgical technique and intensive care have effectively reduced post-operative bleeding risk. Morphological abnormalities such as elongation of the transverse arch (present in 50% of cases) may impact surgical approach,152 which may lead to longer cross-clamp time during coarctation repair.160 Although unlikely to develop in the early post-operative period, there is established risk for dissection after surgical repair or transcatheter stenting of aortic coarctation.164–166 Small case series have provided evidence that balloon angioplasty or stent placement for coarctation is safe and effective in the short term,167,168 but covered stents may be the best approach in the context of intrinsically abnormal arterial tissue. The non-cardiovascular abnormalities that potentially impact peri-operative risk and outcomes include the lymphatic, renal, and endocrine systems. Lymphatic dysfunction can present as foetal lymphoedema or pulmonary lymphatic anomalies such as congenital pulmonary lymphangiectasia, which may predispose to post-operative chylothorax.169 Postnatal peripheral lymphoedema may be a clue to Turner syndrome diagnosis but has no clear clinical impact and usually resolves by 2 years of age without intervention.149 Abnormalities of the renal and urinary system are present in 30–40% of patients, including horseshoe kidney in 10%.149 Hypothyroidism develops in up to 25% of cases, most commonly autoimmune related, and annual thyroid screening is recommended starting at 4 years of age.149,170 In summary, Turner syndrome most clearly impacts outcomes for hypoplastic left heart syndrome. Further investigation is needed to explain these poor outcomes and develop novel approaches and interventions. Arteriopathy associated with Turner syndrome predisposes to hypertension and aortic complications, such as dissection, mandating acute peri-operative blood pressure management and longitudinal follow-up. Williams syndrome Williams syndrome occurs in approximately one in 10,000 live births171 and is associated with 7q11.23 microdeletion. Haploinsufficiency of the elastin gene (ELN) is responsible for the cardiovascular manifestations. Facial features during infancy include a short upturned nose with a flat nasal bridge, peri-orbital puffiness, and long philtrum and later develop into full lips, wide smile, and coarse appearance. Relative strengths in verbal skills and social personality may belie intellectual disability that is present in most cases.172 Familial supra-valvar aortic stenosis is associated with ELN mutations and presents with similar cardiovascular features but none of the non-cardiovascular features. The spectrum of vascular manifestations in Williams syndrome is consistent with generalised arteriopathy. The majority of patients with Williams syndrome have supra-valvar aortic stenosis (45–75%), which may be “hourglass” or “diffuse” type.173 Severe supra-valvar aortic stenosis is unlikely to regress and can be progressive,174–176 but mild stenosis is likely to remain stable.176–178 Additional vascular findings include branch pulmonary stenosis, peripheral pulmonary artery stenosis, supra-valvar pulmonary stenosis, and stenosis of the thoracic aorta, as well as bicuspid aortic valve and mitral valve prolapse.173 The pulmonary arterial lesions often spontaneously improve or resolve over time,174–176,178 but regression also is less likely when severe stenosis is present.179 Surgical repair of supra-valvar aortic stenosis in patients with Williams syndrome has good mortality outcomes with no significant difference in long-term survival compared with familial or sporadic supra-valvar aortic stenosis.180 On the other hand, early mortality can be as high as 20% for cases presenting with the combination of severe supra-valvar aortic stenosis and moderate-to-severe pulmonary stenosis.179,181 Balloon angioplasty of supra-valvar aortic stenosis has been dispelled due to lack of success.176 After transcutaneous stent placement for native or residual post-operative aortic coarctation, there is significant risk for developing re-stenosis, characterised by fibrosis and vascular smooth muscle cell proliferation.182,183 Indeed, patients with stenosis of the thoracic aorta have high re-intervention rates.184 The pulmonary arteries are also predisposed to re-stenosis, aneurysm formation, intimal flap formation, dissection, and rupture after catheter-based interventions.185,186 These outcomes indicate that arteriopathy may limit the effectiveness and increase risk factors when performing catheter-based interventions for arterial stenoses. It is critical to recognise the risk of sudden cardiac death in patients with Williams syndrome, particularly during procedural sedation or anaesthesia or coronary angiography.179,187–190 This risk is highest in those with coronary ostial stenosis or severe biventricular outflow tract obstruction. Among 242 patients with Williams syndrome undergoing 435 cardiac operations or catheter-based interventions, described in the Pediatric Cardiac Care Consortium database, 12 of 15 deaths occurred in the setting of biventricular outflow tract obstruction.185 Coronary ostial stenosis is present in at least 5% of cases and is more common in the “diffuse” type of supra-valvar aortic stenosis or when stenosis of the thoracic aorta is present.178,191 Potential mechanisms of coronary stenosis include adhesion of aortic valve leaflets, overhanging of the supra-valvar ring, or reactive changes to hypertension. Coronary artery stenosis can develop during childhood in the absence of supra-valvar aortic stenosis,192,193 and dilation and tortuosity of the coronary arteries are well recognised.194 These observations suggest primary arteriopathic mechanisms. QT interval prolongation is present in up to 15% of cases, which may predispose to ventricular dysrhythmia and also contribute to sudden death risk.195,196 As coronary stenosis can be sub-clinical, it is critical that patients undergo complete assessment of the coronary arteries when appropriate and that providers be cognizant of the risk factors for sudden death around the time of interventional procedures. Systemic hypertension develops in up to 50% of individuals, which is secondary to renal artery stenosis in some cases. In most cases, hypertension may rather be due to abnormal vascular function or morphology in the distal arteries, but the precise mechanisms are not well understood.197 Cerebral artery stenosis causing ischaemic stroke has been observed in children and should be suspected if neurological changes develop.198 Selecting target blood pressure ranges around the time of procedures can be complicated by the presence of pre-existing hypertension combined with coronary or cerebral artery stenosis, which requires highly attentive pre-operative and post-operative care. Owing to a 15–30% prevalence of sub-clinical hypothyroidism, often due to thyroid hypoplasia, thyroid function testing is recommended every 4 years, and pre-operative evaluation should include thyroid function tests and clinical evaluation for symptoms.199–202 Congenital hypothyroidism due to severe thyroid hypoplasia has also been reported.203 Airway management may be challenging due to facial dysmorphism.200 Based on a concern for mild myopathy in some patients, there have been recommendations to avoid the use of succinylcholine and closely monitor the effects of non-depolarising neuromuscular blockade.200 Anomalies of the kidneys and urinary tract, such as renal aplasia, kidney duplication, horseshoe kidney, and bladder diverticuli, are present in up to 40% of the cases.204,205 Proteinuria was observed in 25%of patients, suggesting that kidney function should be monitored closely.206 Although there is predisposition for episodic hypercalcaemia and hypercalciuria, particularly as neonates, nephrocalcinosis is uncommon.207 Taken together, severe vascular stenosis of the systemic and/or pulmonary arteries increase risk, and asymptomatic patients may be at risk for sudden cardiac death in the setting of occult coronary artery stenosis. These risks pertain to cardiac and non-cardiac procedures. Noonan syndrome and related disorders Noonan syndrome has a prevalence of one in 1000–2500 live births.208 Disease-causing mutations in genes associated with the RAS-MAPK signaling pathway, such as PTPN11 (most frequent), SOS1, RAF1, KRAS, NRAS, BRAF, SHOC2, and CBL, are identified in up to 60% of the cases.209 Cardiofaciocutaneous syndrome (BRAF, KRAS) and Costello syndrome (HRAS) are disorders related to Noonan syndrome with overlapping phenotypic features and genetic aetiologies.210 Neonatal features of Noonan syndrome include tall forehead, hypertelorism, arched eyebrows, low-set posteriorly rotated ears with thick helices, low posterior hairline, and excessive nuchal skin.209 Many of these features become more subtle over time, but short stature, pectus deformity, and neck webbing often remain prominent.208 Patients with Noonan syndrome often achieve normal intelligence,211 whereas cardiofaciocutaneous and Costello syndromes often have more significant developmental delay.210,212 At least 80% of patients with Noonan syndrome have cardiac lesions including pulmonary valve stenosis (50–60%) and secundum atrial septal defect (6–30%).208,213 Hypertrophic cardiomyopathy is present in ~ 20% of patients, especially RAF1 mutations, and portends worse survival than nonsyndromic hypertrophic cardiomyopathy;214,215 however, spontaneous regression occurred in nearly 20% of patients diagnosed in infancy.213 Fibromuscular dysplasia with clinically significant narrowing of the coronary arteries has been reported in the setting of Noonan syndrome and hypertrophic cardiomyopathy.216 Electrocardiographic abnormalities are frequently observed, including predominantly negative forces in the left pre-cordial leads, left axis deviation, and abnormal Q waves.217 Although there are no particular rhythm abnormalities associated with Noonan syndrome, individuals with Costello syndrome (HRAS mutation) develop atrial tachycardia (often multi-focal) in ~50% of cases.218 Early post-operative mortality outcomes have not been frequently reported in Noonan syndrome. Cardiac transplantation in the setting of Noonan syndrome is described, but outcome data are similarly scant.219 Longitudinal screening for occult hypertrophic cardiomyopathy may be indicated, particularly among those with PTPN11 or RAF1 mutations, in part to mitigate risk during cardiac and non-cardiac procedures. Systemic features most likely to impact perioperative outcomes are haematological and lymphatic abnormalities. Haematological abnormalities such as platelet dysfunction and coagulation factor deficiency are present in 30–65% of cases.209,220–223 Severe congenital thrombocytopaenia has been described.224 A recent study reported frequent easy bruising and post-surgical bleeding (15–25%), platelet dysfunction (80%), and factor VII deficiency (20%).225 Bleeding diathesis may predispose patients to spontaneous gastrointestinal or sub-arachnoid haemorrhage, which may respond to administration of recombinant factor VII.226,227 Owing to the risk of coagulopathy, complete blood count and basic coagulation testing is warranted before operations, haematology consultation should be considered, and aspirin may be avoided.208,209 Lymphatic abnormalities are observed in ~20% of cases.209 Peripheral lymphoedema often spontaneously resolves within the first several years but can have late onset.228 Similar to Turner syndrome, pulmonary lymphatic abnormalities including congenital pulmonary lymphangiectasia may predispose to chylothorax.169,229–231 Post-operative pericardial and pleural effusions were not significantly increased in a series of ~120 operations.213 Cutaneous leaking of lymphatic fluid from a femoral vascular access site due to lymphangiectasia has been reported during cardiac catheterisation.232 Taken together, Noonan syndrome and related disorders are notable for genotype–phenotype relationships such as the associations between RAF1 and hypertrophic cardiomyopathy and HRAS and atrial tachycardia. Bleeding and lymphatic abnormalities may complicate the peri-operative course. Additional peri-operative outcome studies are warranted. Marfan syndrome and related disorders Marfan syndrome is present in approximately one in 5000 live births and most commonly caused by mutations in the FBN1 gene, which encodes the extracellular matrix protein fibrillin-1.233 Skeletal abnormalities – for example, pectus deformity, long arms, short upper body segment, craniofacial dysmorphism, and arachnodactyly – and ocular abnormalities – such as ectopia lentis and myopia – are often present.234 Cardiovascular involvement consists of aortopathy, characterised by thoracic aortic aneurysm and risk for dissection, and mitral valve prolapse. Development and intellectual ability are typically normal. Although most patients with Marfan syndrome do not require cardiac surgery until adulthood,235 excellent operative survival has been demonstrated in children undergoing aortic root replacement.236–238 Peri-operative providers should recognise risk for pneumothorax and other pulmonary co-morbidities including pulmonary emphysema.239 Pectus deformity or severe scoliosis may also impact surgical approach and recovery. Some patients with particularly severe cardiovascular disease are referred to as having neonatal Marfan syndrome, which is associated with mutations in exons 24–32 of FBN1.240,241 Arachnodactyly, congenital contractures, and crumpled ears feature prominently in these neonates, who often present with severe mitral and tricuspid valve regurgitation, leading to cardiac failure and death within the first few months of life. Rare cases of surgical success including quadrivalvar replacement and cardiac transplantation have been reported.242,243 Loeys–Dietz syndrome, which is associated with mutations in the TGF-β receptor genes TGFBR1 and TGFBR2, has overlapping but distinct phenotypic features with Marfan syndrome.244 Systemic features include hypertelorism, bifid uvula, cleft palate, craniosynostosis, and velvety/thin skin. Talipes equinovarus and camptodactyly may also be diagnostic clues in a neonate.245 The major cardiovascular manifestations are generalised arterial tortuosity and risk for aneurysm and dissection. Additional cardiovascular lesions include bicuspid aortic valve, atrial septal defect, and mitral valve prolapse. Vascular disease in Loeys–Dietz syndrome is typically more severe than Marfan syndrome with risk of rapid progression and aortic dissection. Dissection is described as early as 6 months of age.246 There is also often more extensive arterial involvement, which may require complete aortic replacement. Tortuosity and aneurysm of the brachiocephalic and intra-cranial arteries may predispose to cerebrovascular events.247,248 Despite the aggressive vascular features of the disease, successful aortic root replacement in infancy has been reported.249 Furthermore, there were no operative deaths among two series of children with Loeys–Dietz syndrome undergoing aortic root replacement.246,250 Cardiovascular complications in the setting of complex CHD have included progressive pulmonary artery dilation and rupture and post-operative mitral leaflet rupture.245,251,252 Similar to Marfan syndrome, patients with Loeys– Dietz syndrome have increased risk of post-operative pneumothorax.247,250 Careful peri-operative positioning should be utilised due to risk of low bone mineral density and increased fracture risk as well as cervical spine anomalies.247,253–255 Tortuosity or aneurysm of the peripheral arteries also may impact vascular access.247 Taken together, early post-operative outcomes are generally favourable for these conditions, but the risk of recurrent aneurysm or dissection mandates lifelong surveillance. Loeys–Dietz syndrome has unusual characteristics that may not be well recognised due to the more recent discovery and characterisation of the disorder. Alagille syndrome Alagille syndrome has a prevalence of at least one in 70,000 live births and is associated with the Notch signaling pathway genes JAG1 (97% of cases) and NOTCH2 (1% of cases).256 The hallmark systemic manifestations include bile duct paucity, resulting in cholestasis, facial dysmorphism – deep-set eyes, prominent ears, triangular face with broad forehead, and pointed chin – vertebral anomalies, and ocular anomalies, often posterior embryotoxon. CHD is present in at least 90% of the cases. The most common cardiovascular findings are right-sided lesions including proximal branch pulmonary artery stenosis, peripheral pulmonary artery stenosis, tetralogy of Fallot, or pulmonary valve stenosis. Left-sided lesions and septal defects are also observed but are less frequent.257 In addition to the hallmark systemic features, renal anomalies are observed in ~40% of patients, which includes 20% with renal dysplasia and 5% risk of developing chronic renal failure.258 There is evidence that patients with Alagille syndrome have relatively poor longitudinal outcomes in the setting of tetralogy of Fallot or pulmonary atresia with ventricular septal defect;257,259 however, positive early outcomes were recently reported among 15 patients with pulmonary atresia and major aortopulmonary collateral arteries260 and six patients undergoing primary surgical reconstruction of peripheral pulmonary artery stenosis.181 Owing to congenital biliary anomalies, Alagille syndrome may present the unusual challenge of requiring paediatric cardiac surgery in patients with severe liver disease; two small case series have reported operative mortality in two out of four children with Alagille syndrome and end-stage liver disease undergoing cardiac surgery.261,262 It is increasingly clear that Alagille syndrome is a disorder characterised by diffuse arteriopathy and that arterial anomalies – aneurysm or stenosis – significantly contribute to poor outcomes. In a large cohort of 268 patients with Alagille syndrome, systemic arterial anomalies or intra-cranial vascular events were present in nearly 10% of patients, and vascular accidents were responsible for 34% of the observed mortality.263 Spontaneous haemorrhage in the gastrointestinal tract, nasal/oral mucosa, and uterine lining are also reported in the absence of liver failure. It is speculated that elevated levels of apolipoprotein E may impair normal haemostasis,264 but a primary arterial fragility may be likely. A unique case report of a child with recurrent aortopulmonary shunt dehiscence due to extensive atherosclerosis and plaque at the anastomosis site has prompted some to consider routine screening and treatment for dyslipidaemia to prevent exacerbation of arterial disease in these patients.265 Taken together, systemic arteriopathy presents significant challenges to both early and late survival outcomes. Trisomy 13 and 18 Patients with trisomy 13 – Patau syndrome – or trisomy 18 – Edwards syndrome – have severe co-morbidities and poor prognosis with >90% of the affected infants dying by age 1 year. Given the severe multi-systemic nature of these disorders, the presence of CHD may not impact overall survival.266 Cardiac lesions are most commonly septal defects, but left ventricular non-compaction associated with progressive heart failure has been described in trisomy 13.267,268 Despite poor overall survival, cardiac operations including palliative and complete repairs may be beneficial in select groups.269,270 The care for these patients and families requires a balanced multidisciplinary approach including palliative care specialists. CHARGE syndrome CHARGE syndrome is present in approximately one in 8500 live births.271 Most cases (~70%) are associated with mutation in the CHD7 gene, which encodes a chromodomain helicase DNA-binding protein, and are rarely associated with mutation in SEMA3E;272,273 22q11 deletion has also been described in patients clinically diagnosed with CHARGE syndrome.274 The major diagnostic criteria (“four C’s”) are coloboma, choanal atresia, cranial nerve dysfunction, and characteristic ear anomalies, external and middle ear anomalies.275 CHD is present in ~75% of patients and includes conotruncal and septal defects, including atrioventricular septal defects.272,276 Forebrain central nervous system malformations are frequently observed,277 yet significant intellectual disability is not guaranteed.275 Immunological dysfunction including severe T-cell deficiency has been reported.278 Renal anomalies are observed in ~30–40% cases and include solitary kidney, hydronephrosis, renal hypoplasia, duplex kidneys, and vesicoureteral reflux.275 Peri-operative outcomes have not been frequently reported, but sub-optimal longitudinal outcomes for patients with conotruncal defects have been suggested.31 A major peri-operative risk factor is the high frequency of anatomical and functional abnormalities of the respiratory tract. Upper airway anomalies – choanal atresia, cleft lip/palate, and micrognathia – and laryngotracheal anomalies – tracheoesophageal fistula, laryngomalacia, tracheomalacia, sub-glottic stenosis, laryngeal cleft, and anterior larynx – may complicate airway management.279,280 Cranial nerve dysfunction – for example, cranial nerves IX and X – leads to pharyngeal and laryngeal dysfunction and poor airway protection, a problem that may be exacerbated by high frequency of gastroesophageal reflux.281 Indeed, post-operative airway events are frequently encountered – 35% in a recent series – occurring most frequently after cardiac surgery.282 In an early case series, over half of deaths were attributed to pulmonary aspiration.283,284 Pituitary structural abnormalities may be associated with neonatal hypocortisolism and should be considered in cases of refractory hypotension.285,286 Rare genetic syndromes associated with CHD have features predisposing to poor perioperative outcomes that may be sub-optimally recognised by providers due to lack of familiarity Ellis–van Creveld syndrome Ellis–van Creveld syndrome is a rare autosomal-recessive disorder (EVC or EVC2 mutations) with increased occurrence among the Amish population inhabiting Pennsylvania, United States of America.287 Frequent characteristics include short stature, polydactyly, short ribs, and dysplastic nails, hair, and teeth. Notably, cognitive development is normal. CHD is present in ~60% and includes common atrium, atrioventricular septal defect, and systemic and pulmonary venous abnormalities.288,289 Overall, three noteworthy retrospective studies have analysed cardiac surgical outcomes. A case series of nine patients undergoing cardiac surgery at a single centre from 2004 to 2009 observed a preponderance of respiratory morbidity.288 Death occurred within 150 days after surgery in four out of nine patients, primarily due to respiratory failure. Respiratory complications, including three of five survivors requiring tracheostomy, were attributed to a thoracic dystrophy similar to Jeune syndrome. Increased procedure-related respiratory morbidity was also observed in the Pediatric Health Information System database from 2004 to 2011.290 In fairly stark contrast with these reports, a review of the Pediatric Cardiac Care Consortium database from 1982 to 2007 identified no operative mortality among 21 children undergoing cardiac surgery.289 The reason for the discrepancy between these reports is unclear. Notably, thoracic dystrophy may improve with somatic growth, suggesting benefit of deferring surgery for as long as possible.288 Together, these observations indicate the need for complete pulmonary evaluation and consideration of invasive haemodynamic assessment before cardiac operations. VACTERL VACTERL association likely represents a genetically heterogeneous population consisting of vertebral defects, anal atresia, cardiac defects, tracheoesophageal fistula with oesophageal atresia, renal anomalies, and limb defects.291 The renal anomalies include unilateral agenesis, horseshoe kidney, cystic disease, and dysplasia, and there is risk for chronic kidney disease with progression to end-stage renal disease.292 In a cohort of 46 patients, 31 had CHD, which was most frequently ventricular septal defect.293 Probably due to the currently imprecise nature of this diagnosis, there are little outcomes data available. PHACES PHACES association includes posterior fossa malformations, haemangioma – often large, segmental, and involving the head or neck – arterial anomalies, cardiac defects, eye abnormalities, and sternal defects.294 A genetic aetiology has not been established. Arterial manifestations include anomalous patterning, stenosis, occlusion, or aneurysm of the cervical and/or cerebral arteries, which are usually ipsilateral to the haemangioma.295,296 Aortic arch sidedness is also often ipsilateral to the haemangioma.297 Cardiovascular malformations are present in ~40% of patients, including aberrant subclavian artery, coarctation of the aorta (~20%), and ventricular septal defect (~15%).298 Coarctation morphology is often complex and is rarely associated with bicuspid aortic valve.294 Preparation for surgical repair of coarctation should include a complete evaluation of the aortic arch and head and neck arteries by cardiac catheterisation or other imaging modality to optimise surgical approach.294,299 Peri-operative providers should also recognise increased risk for subglottic haemangioma and risk for ischaemic stroke and seizures during infancy.300–302 Cri du chat syndrome Cri du chat syndrome (5p15 deletion) has a prevalence of approximately one in 15,000 live births.303 A distinguishing feature is the characteristic high-pitched cry. Neonatal craniofacial features include microcephaly and round face with large nasal bridge, hypertelorism, and micrognathia. Severe psychomotor and growth delay is observed in most cases. Tracheal intubation may be complicated by the presence of laryngeal abnormalities including small larynx, narrow diamond-shaped larynx, and laryngomalacia, and large, floppy epiglottis.304 CHD is present in ~ 20% of the patients, including patent ductus arteriosus, ventricular septal defect, atrial septal defect, and right ventricular outflow tract obstructive lesions including tetralogy of Fallot.305 Outcomes data are limited, but a review of the Pediatric Cardiac Care Consortium from 1982 to 2002 identified 18 children undergoing cardiac surgery, including five complete tetralogy of Fallot repairs, who had good overall survival with one operative death.305 Jacobsen syndrome Jacobsen syndrome has a prevalence of approximately one in 100,000 live births and is associated with a deletion on the long arm of chromosome 11 with break point at 11q23.306 The pathogenic gene for cardiovascular manifestations may be ETS1.16 Dysmorphic features include skull deformity – for example, trigonocephaly – hypertelorism, strabismus, low posteriorly rotated ears, and syndactyly. Intellectual disability and behavioural abnormalities are observed in the majority of cases. CHD occurs in ~50% of cases, primarily consisting of ventricular septal defect or left-sided obstructive lesion, including up to 5% with hypoplastic left heart syndrome.307,308 Importantly, there is often increased bleeding risk due to a platelet disorder – Paris–Trousseau syndrome – characterised by neonatal thrombocytopaenia, which can be severe but improves with age, and platelet dysfunction, which often persists.306,308 Pre-operative evaluation of platelet function using thromboelastography may be warranted. Airway management can be complicated by micrognathia and anterior laryngeal opening.309 Central hypothyroidism has been reported.310 Renal and urinary tract malformations, including dysplasia, hydronephrosis, and unilateral agenesis, occur rarely.306,307 Kabuki syndrome Kabuki syndrome has a prevalence of approximately one in 32,000 live births and in most cases is associated with mutations in the MLL2 gene, which encodes a histone methyltransferase.311,312 Its naming is derived from a characteristic appearance of long palpebral fissures with lower eyelid eversion and arched eyebrows, resembling masks worn in Kabuki theatre. Another characteristic finding is foetal finger pads. Intellectual disability is present in ~90% and seizures in 12–25%.313–315 Cardiac defects are present in ~50% of cases and include ventricular septal defect, atrial septal defect, left-sided obstructive lesions – most commonly coarctation of the aorta – and tetralogy of Fallot.313,314,316 Abnormalities in humoral immunity, including low levels of IgA, total IgG, or IgG sub-classes, were observed in ~50%, which may explain the predisposition to upper respiratory infections, and potentially impacts peri-operative risk.317 Cleft lip/palate including sub-mucous clefts occurs in ~50%.313 Renal abnormalities include renal dysplasia, agenesis, horseshoe kidney, ectopic kidney, and hydronephrosis.316 Smith–Magenis syndrome Smith–Magenis syndrome has a prevalence of approximately one in 25,000 live births318 and is associated with the deletion of 17p11.2.318,319 Craniofacial features include broad face with hypertelorism and upslanting eyes, prognathism, low-set ears, cleft lip/palate, and ocular abnormalities.320 Mild-to-moderate developmental delay is often observed along with characteristic neurobehavioural features such as sleep disturbance with inverted circadian rhythm and predilection for self-injury.320 CHD is present in ~30–40% and includes ventricular septal defect, atrial septal defect, right-sided lesions including tetralogy of Fallot, and total anomalous pulmonary venous return.320–322 The cardiovascular risk profile includes predisposition for dyslipidaemia, including hypercholesterolaemia.323 Post-operative ischaemic stroke in a young adult with premature cerebrovascular atherosclerosis has been reported.324 Immunoglobulin levels are low in ~20%.321 Hypothyroidism presents in ~30%.321 Epileptiform abnormalities are present in ~50%, and clinical seizures develop in ~20–30%.320,325 Renal and urinary tract anomalies are present in ~15% and include renal dysplasia, small kidney, vesicoureteral reflux, renal agenesis, and ureteral duplication.320,326 Wolf–Hirschhorn syndrome Wolf–Hirschhorn syndrome has a prevalence of approximately one in 20,000 live births and is associated with the deletion of 4p16.3.327,328 Patients characteristically have the appearance of a “Greek warrior helmet” with high forehead, prominent glabella, and protruding eyes with hypertelorism.328 Micrognathia, forehead haemangioma, and cleft lip/palate also occur with increased frequency. Severe developmental delay is uniformly observed, and seizures occur in ~90% of individuals starting at a young age.329 CHD is present in ~50%, most commonly atrial septal defect, pulmonary stenosis, ventricular septal defect, and patent ductus arteriosus, but more complex lesions have been reported.328,330,331 Defects in humoral immunity, including common variable immunodeficiency and isolated IgA deficiency, are frequently observed.332 Renal and urinary tract defects are observed in ~30% and include vesicoureteral reflux, renal agenesis, dysplasia, or hypoplasia, and horseshoe kidney.328 Cornelia de Lange syndrome Cornelia de Lange syndrome, also known as Brachmann–de Lange syndrome, has a prevalence of approximately one in 10,000 live births and is caused by mutations in the NIPBL, SMC1A, or SMC3 genes, which encode gene products involved in the function of cohesin, a protein complex involved in cell division.333 Patients have consistent craniofacial features including short neck, low posterior hairline, hirsute forehead, arched and confluent eyebrows, and thick and long eyelashes.334 Mild-to-moderate intellectual disability is frequent.335 CHD is present in ~30% and includes pulmonary valve stenosis, peripheral pulmonary artery stenosis, atrial septal defect, ventricular septal defect, left-sided obstructive lesions, and tetralogy of Fallot; there is also risk for progressive atrioventricular valve dysplasia.336,337 Airway management may be complicated by micrognathia, cleft palate, sub-mucous cleft, short, stiff neck, and restricted mouth opening.338 Recurrent infections including fungal infections are reported at increased frequency, and humoral deficiency and T-cell abnormalities have been observed.339 Thrombocytopaenia has been observed in ~20%.340 Renal and urinary tract anomalies are observed in ~40% of patients and most frequently renal dysplasia, pelvic dilation, and vesicoureteral reflux are observed.341 Seizures, often partial type, occur in ~25%.342 Holt–Oram syndrome Holt–Oram syndrome, which is characterised by the triad of atrial septal defect, conduction abnormality, and upper limb malformation – most commonly thumb – has a prevalence of approximately one in 100,000 live births and is caused by mutations in the cardiac transcription factor TBX5.343 Cardiac lesions include atrial septal defect, which is most common, ventricular septal defect, and more complex lesions such as conotruncal defects, atrioventricular canal defects, and left-sided obstructive lesions.343 The most frequent conduction abnormality is atrioventricular block, most commonly first degree, which may be present in the absence of structural CHD.344 Aside from the risk of atrioventricular block or other conduction disturbances, there are typically no other significant co-morbidities expected to complicate peri-operative care. Goldenhar syndrome Goldenhar syndrome, also known as oculo-auriculovertebral spectrum, occurs in up to one in 6000 live births.345 Although suspected to be due to abnormal development of the first and second branchial arches, the genetic cause is presently unknown; however, 22q11 deletion was recently reported in patients diagnosed with this disorder.346 The defining features include unilateral microtia, hemifacial microsomia with mandibular hypoplasia, ocular epibulbar dermoid, and cervical vertebral malformations.347 CHD is present in ~30% of cases and includes conotruncal defects, ventricular septal defect, and atrial septal defect.345 Significant craniofacial distortion and cervical vertebral anomalies may complicate airway management.348 Renal and urinary tract anomalies include ectopic or fused kidneys, renal agenesis, and vesicoureteral reflux.349 Conclusion The impact of a genetic syndrome and associated co-morbidities on the peri-operative course and outcomes cannot be understated (Table 3). Recognising the risk factors particular to specific genetic syndromes has the potential to prevent or ameliorate peri-operative complications and improve short-term and long-term outcomes (Table 2 and Supplementary Table S3). The development of peri-operative management protocols tailored to specific syndromes based on current knowledge may be an effective strategy to achieve these goals. Understanding the cause is essential to elucidate pathogenesis and develop new treatment strategies. As the capability to interrogate and comprehend the genetic basis of CHD improves and clinical availability of genetic testing proliferates, there are increasing opportunities for early diagnosis, risk stratification, genetic counselling, and anticipatory clinical care.350 We propose that these tasks may be most effectively achieved by the establishment of multi-disciplinary sub-specialty cardiovascular genetics services. In order to advance peri-operative management, there are present and future needs to integrate registries containing careful phenotyping and clinical outcomes data – for example, Society of Thoracic Surgeons database and Pediatric Heart Network – with registries containing comprehensive genetic data – for example, the Pediatric Cardiac Genomics Consortium.351,352 There are a limited number of exemplary studies that illustrate the value of performing comprehensive genetic evaluations and specifically reporting not only positive genetic testing results but also negative results to optimise interpretation and generalisability.31,33 This design may be more challenging to implement in large registries but should be considered for establishment and updating of registries as genetic testing advances. As clinical investigators continue to delineate the clinical significance of genetic diagnoses and apply the evidence to peri-operative care, there is promise for improvement in both short-term and long-term outcomes, such as neurodevelopment, quality of life, and general health into adulthood.17,18,353 Supplementary Material S1 S2 S3 S4 S5 S6 None. Financial Support This manuscript received no specific grant from any funding agency, commercial, or not-for-profit sectors. Table 1 Summary of post-operative mortality and hospital length of stay outcomes among four frequently encountered genetic syndromes. Down syndrome 22q11 deletion Heterotaxy syndrome Turner syndrome Lesion/operation Early mortality LOS Early mortality LOS Early mortality LOS Early mortality LOS All cardiac surgery Low37–39,71,* Low71 – – High129,130 High130 – – Septal defects AVSD Medium37,39–46,48,** Low42 – – – – – – VSD Low37,41,* High39,65 – – – – – – SV lesions Stage 1 palliation High37,39 – – – High129 – High25 – Stage 2 palliation High37,39 High39 – – – – – – Stage 3 palliation Medium37,39,50,51 High51 – – Medium124,129 – – – Conotruncal defects Collective Low31,33 – Medium31,33,88 Low88 – – – – TOF Low32,37,39,41,49 High39 Low32,86 – – – – – PA-VSD – – High91–93 – – – – – IAA or PTA – – Low87 High87 – – – – Other CoA Low37 – – – – Low160 High160 Cardiac transplantation – – – – Medium137,138 – – – TAPVR – – – – Low134 – – – AVSD = atrioventricular septal defect; CoA = coarctation of the aorta; IAA = interrupted aortic arch; LOS = length of stay (in-hospital); PA-VSD = pulmonary atresia with ventricular septal defect; PTA = persistent truncus arteriosus; SV = single ventricle; TAPVR = total anomalous pulmonary venous return; TOF = tetralogy of Fallot; and VSD = ventricular septal defect Classification of risk for poor early post-operative outcomes relative to patients without the syndromic diagnosis (high: studies reviewed only demonstrating increased mortality or LOS, medium: studies demonstrating increased or no difference in mortality or LOS, low: studies demonstrating no difference in mortality or LOS) * Some studies reported decreased mortality ** Studies reported increased mortality, decreased mortality, and no difference in mortality Table 2 Classes of risks and suggested peri-operative precautions/actions for specific syndromes. Class Syndromes Actions Cardiac rhythm HTX (SND, AV block, tachyarrhythmia), WS (LQT, ventricular ectopy), TS (LQT), Costello (atrial tachycardia), Holt–Oram (AV block) Maintenance of normal electrolyte levels, routine placement of temporary pacing wires Vascular (systemic) TS, WS, LDS, PHACES Pre-operative vascular imaging studies, documentation of pre-operative BP, patient-specific BP goals, ultrasound-guided arterial access Vascular (pulmonary) DS, HTX, EVC Pre-operative cardiac catheterisation, post-operative manoeuvers to minimise PVR Myocardial HTX (non-compaction cardiomyopathy), trisomy 13 (non-compaction cardiomyopathy) Intra-operative myocardial protection, anticipatory post- CPB management of ventricular dysfunction Respiratory Upper airway anomalies: DS, 22q11 deletion, CHARGE, PHACES, Cri du chat, Cornelia de Lange Lower airway disease: DS, EVC, MFS/LDS Pre-operative anatomic upper airway evaluation, extubation protocols, post-operative evaluation of airway protection mechanisms, otolaryngology/ pulmonary consultation Immunologic/infectious DS, 22q11 deletion, HTX, Kabuki, Smith–Magenis, Wolf–Hirschhorn, Cornelia de Lange Immunology consultation, broad-spectrum antimicrobial prophylaxis, minimise invasive monitoring Haematologic 22q11 deletion, NS, AGS, Jacobsen, Cornelia de Lange Haematology consultation, post-CPB antifibrinolytics, BP control, rapid access to blood products, liberal blood product administration Neurologic Seizure: DS, 22q11 deletion, Kabuki, Smith–Magenis, Wolf–Hirschhorn Seizure: neuroprotection, peri-operative EEG evaluation, normocalcaemia (22q11 deletion) Cerebrovascular: AGS, PHACES, LDS, WS, NS Cervical instability: DS, LDS Cerebrovascular: pre-operative cerebrovascular imaging, cerebral perfusion pressure monitoring, urgent imaging for neurological changes Cervical instability: appropriate positioning/support Endocrine Hypothyroidism: DS, TS, WS, PHACES, Jacobsen, Smith–Magenis Pituitary dysfunction: CHARGE Pre-operative thyroid function testing, endocrinology consultation as needed, steroid replacement Lymphatic DS, TS, NS Monitoring for chylothorax and sequelae if present, early transition to low and medium chain triglyceride diet/ formula, minimise central venous pressure AGS = Alagille syndrome; AV = atrioventricular; BP = blood pressure; CPB = cardiopulmonary bypass; DS = Down syndrome; EEG = electroencephalogram; EVC = Ellis–van Creveld; HTX = heterotaxy syndrome; LDS = Loeys–Dietz syndrome; LQT = prolonged QT interval; MFS = Marfan syndrome; NS = Noonan syndrome; PVR = pulmonary vascular resistance; SND = sinus node dysfunction; TS = Turner syndrome; and WS = Williams syndrome Table 3 Key points. Genetic syndromes often present specific cardiovascular and non-cardiovascular co-morbidities that negatively impact mortality and morbidity outcomes Diagnosis of a genetic syndrome allows for risk stratification, counseling on prognosis and recurrence risk, anticipatory peri-operative management, and therapy decisions Syndrome-specific protocols for peri-operative evaluation and prophylactic tactics may improve peri-operative outcomes. Particular attention should be given to immunological, haematological, vascular, and neurological risks. Cardiac anaesthesia during non-cardiac procedures should be considered in the context of certain genetic syndromes Improved peri-operative outcomes may translate to improved short-term and long-term outcomes and reduce long-term co-morbities and cost Design and reporting of surgical database registries and clinical trials should clearly define diagnostic criteria for genetic syndromes and specify positive and negative genetic testing results Integration of large clinical and genetic databases will advance clinical outcomes The development of cardiovascular genetics services will provide sub-specialty expertise on specific aspects of care of patients with genetic diagnoses, which over time will be increasingly encountered Conflicts of Interest None. Supplementary materials For supplementary materials referred to in this article, please visit http://dx.doi.org/10.1017/S1047951115001389 References 1 Hoffman JI Incidence of congenital heart disease: I. Postnatal incidence Pediatr Cardiol 1995 16 103 113 7617503 2 Reller MD Strickland MJ Riehle-Colarusso T Mahle WT Correa A Prevalence of congenital heart defects in metropolitan Atlanta, 1998–2005 J Pediatr 2008 153 807 813 18657826 3 Hoffman JI Kaplan S The incidence of congenital heart disease J Am Coll Cardiol 2002 39 1890 1900 12084585 4 Pierpont ME Basson CT Benson DW Jr Genetic basis for congenital heart defects: current knowledge: a scientific statement from the American Heart Association Congenital Cardiac Defects Committee, Council on Cardiovascular Disease in the Young: endorsed by the American Academy of Pediatrics Circulation 2007 115 3015 3038 17519398 5 Stoll C Dott B Alembik Y Roth M Associated noncardiac congenital anomalies among cases with congenital heart defects Eur J Med Genet 2014 58 75 85 25497206 6 Ferencz C Neill CA Boughman JA Congenital cardiovascular malformations associated with chromosome abnormalities: an epidemiologic study J Pediatr 1989 114 79 86 2521249 7 Eskedal L Hagemo P Eskild A A population-based study of extra-cardiac anomalies in children with congenital cardiac malformations Cardiol Young 2004 14 600 607 15679995 8 Fuller S Nord AS Gerdes M Predictors of impaired neurodevelopmental outcomes at one year of age after infant cardiac surgery Eur J Cardiothorac Surg 2009 36 40 47 19394849 9 Oyen N Poulsen G Boyd HA Recurrence of congenital heart defects in families Circulation 2009 120 295 301 19597048 10 Hinton RB Jr Martin LJ Tabangin ME Hypoplastic left heart syndrome is heritable J Am Coll Cardiol 2007 50 1590 1595 17936159 11 Schott JJ Benson DW Basson CT Congenital heart disease caused by mutations in the transcription factor NKX2–5 Science 1998 281 108 111 9651244 12 Chen Y Han ZQ Yan WD A novel mutation in GATA4 gene associated with dominant inherited familial atrial septal defect J Thorac Cardiovasc Surg 2010 140 684 687 20347099 13 Garg V Muth AN Ransom JF Mutations in NOTCH1 cause aortic valve disease Nature 2005 437 270 274 16025100 14 Fahed AC Gelb BD Seidman JG Seidman CE Genetics of congenital heart disease: the glass half empty Circ Res 2013 112 707 720 23410880 15 Zaidi S Choi M Wakimoto H De novo mutations in histone-modifying genes in congenital heart disease Nature 2013 498 220 223 23665959 16 Glessner J Bick AG Ito K Increased frequency of de novo copy number variations in congenital heart disease by integrative analysis of SNP array and exome sequence data Circ Res 2014 115 884 896 25205790 17 Marino BS Lipkin PH Newburger JW Neurodevelopmental outcomes in children with congenital heart disease: evaluation and management: a scientific statement from the American Heart Association Circulation 2012 126 1143 1172 22851541 18 Marino BS Tomlinson RS Wernovsky G Validation of the pediatric cardiac quality of life inventory Pediatrics 2010 126 498 508 20805147 19 Gilboa SM Salemi JL Nembhard WN Fixler DE Correa A Mortality resulting from congenital heart disease among children and adults in the United States, 1999 to 2006 Circulation 2010 122 2254 2263 21098447 20 Bronicki RA Chang AC Management of the postoperative pediatric cardiac surgical patient Crit Care Med 2011 39 1974 1984 21768801 21 Simsic JM Coleman K Maher KO Cuadrado A Kirshbom PM Do neonates with genetic abnormalities have an increased morbidity and mortality following cardiac surgery? Congenit Heart Dis 2009 4 160 165 19489943 22 Doell C Bernet V Molinari L Children with genetic disorders undergoing open-heart surgery: are they at increased risk for postoperative complications? Pediatr Crit Care Med 2011 12 539 544 21057364 23 Barker GM O’Brien SM Welke KF Major infection after pediatric cardiac surgery: a risk estimation model Ann Thorac Surg 2010 89 843 850 20172141 24 Kagen J Lautenbach E Bilker WB Risk factors for mediastinitis following median sternotomy in children Pediatr Infect Dis J 2007 26 613 618 17596804 25 Patel A Hickey E Mavroudis C Impact of noncardiac congenital and genetic abnormalities on outcomes in hypoplastic left heart syndrome Ann Thorac Surg 2010 89 1805 1813 (discussion 1813–1814) 20494032 26 Stasik CN Gelehrter S Goldberg CS Current outcomes and risk factors for the Norwood procedure J Thorac Cardiovasc Surg 2006 131 412 417 16434272 27 Jacobs JP O’Brien SM Chai PJ Management of 239 patients with hypoplastic left heart syndrome and related malformations from 1993 to 2007 Ann Thorac Surg 2008 85 1691 1696 (discussion 7) 18442568 28 Tabbutt S Ghanayem N Ravishankar C Risk factors for hospital morbidity and mortality after the Norwood procedure: a report from the Pediatric Heart Network Single Ventricle Reconstruction trial J Thorac Cardiovasc Surg 2012 144 882 895 22704284 29 Gaynor JW Mahle WT Cohen MI Risk factors for mortality after the Norwood procedure Eur J Cardiothorac Surg 2002 22 82 89 12103378 30 Hornik CP He X Jacobs JP Complications after the Norwood operation: an analysis of The Society of Thoracic Surgeons Congenital Heart Surgery Database Ann Thorac Surg 2011 92 1734 1740 21937021 31 Michielon G Marino B Oricchio G Impact of DEL22q11, trisomy 21, and other genetic syndromes on surgical outcome of conotruncal heart defects J Thorac Cardiovasc Surg 2009 138 565.e2 570.e2 19698836 32 Michielon G Marino B Formigari R Genetic syndromes and outcome after surgical correction of tetralogy of Fallot Ann Thorac Surg 2006 81 968 975 16488703 33 Anaclerio S Di Ciommo V Michielon G Conotruncal heart defects: impact of genetic syndromes on immediate operative mortality Ital Heart J 2004 5 624 628 15554034 34 Bull MJ Health supervision for children with Down syndrome Pediatrics 2011 128 393 406 21788214 35 Cocchi G Gualdi S Bower C International trends of Down syndrome 1993–2004: births in relation to maternal age and terminations of pregnancies Birth Defects Res A Clin Mol Teratol 2010 88 474 479 20589916 36 Freeman SB Taft LF Dooley KJ Population-based study of congenital heart defects in Down syndrome Am J Med Genet 1998 80 213 217 9843040 37 Evans JM Dharmar M Meierhenry E Marcin JP Raff GW Association between Down syndrome and in-hospital death among children undergoing surgery for congenital heart disease: a US population-based study Circ Cardiovasc Qual Outcomes 2014 7 445 452 24755908 38 Seifert HA Howard DL Silber JH Jobes DR Female gender increases the risk of death during hospitalization for pediatric cardiac surgery J Thorac Cardiovasc Surg 2007 133 668 675 17320563 39 Fudge JC Jr Li S Jaggers J Congenital heart surgery outcomes in Down syndrome: analysis of a national clinical database Pediatrics 2010 126 315 322 20624800 40 Morris CD Magilke D Reller M Down’s syndrome affects results of surgical correction of complete atrioventricular canal Pediatr Cardiol 1992 13 80 84 1535440 41 Reller MD Morris CD Is Down syndrome a risk factor for poor outcome after repair of congenital heart defects? J Pediatr 1998 132 738 741 9580782 42 St Louis JD Jodhka U Jacobs JP Contemporary outcomes of complete atrioventricular septal defect repair: analysis of the Society of Thoracic Surgeons Congenital Heart Surgery Database J Thorac Cardiovasc Surg 2014 148 2526 2531 25125206 43 Formigari R Di Donato RM Gargiulo G Better surgical prognosis for patients with complete atrioventricular septal defect and Down’s syndrome Ann Thorac Surg 2004 78 666 672 (discussion 72) 15276542 44 Al-Hay AA MacNeill SJ Yacoub M Shore DF Shinebourne EA Complete atrioventricular septal defect, Down syndrome, and surgical outcome: risk factors Ann Thorac Surg 2003 75 412 421 12607648 45 Alexi-Meskishvili V Ishino K Dahnert I Correction of complete atrioventricular septal defects with the double-patch technique and cleft closure Ann Thorac Surg 1996 62 519 524 (discussion 524–525) 8694616 46 Lange R Guenther T Busch R Hess J Schreiber C The presence of Down syndrome is not a risk factor in complete atrioventricular septal defect repair J Thorac Cardiovasc Surg 2007 134 304 310 17662766 47 Tucker EM Pyles LA Bass JL Moller JH Permanent pacemaker for atrioventricular conduction block after operative repair of perimembranous ventricular septal defect J Am Coll Cardiol 2007 50 1196 1200 17868813 48 Desai AR Branco RG Comitis GA Early postoperative outcomes following surgical repair of complete atrioventricular septal defects: is Down syndrome a risk factor? Pediatr Crit Care Med 2014 15 35 41 24201860 49 Mulder TJ Pyles LA Stolfi A Pickoff AS Moller JH A multicenter analysis of the choice of initial surgical procedure in tetralogy of Fallot Pediatr Cardiol 2002 23 580 586 12530488 50 Gupta-Malhotra M Larson VE Rosengart RM Guo H Moller JH Mortality after total cavopulmonary connection in children with the Down syndrome Am J Cardiol 2010 105 865 868 20211334 51 Furukawa T Park IS Yoshikawa T Outcome of univentricular repair in patients with Down syndrome J Thorac Cardiovasc Surg 2013 146 1349 1352 23522604 52 Levine OR Simpser M Alveolar hypoventilation and cor pulmonale associated with chronic airway obstruction in infants with Down syndrome Clin Pediatr 1982 21 25 29 53 Southall DP Stebbens VA Mirza R Upper airway obstruction with hypoxaemia and sleep disruption in Down syndrome Dev Med Child Neurol 1987 29 734 742 2961643 54 Stebbens VA Dennis J Samuels MP Croft CB Southall DP Sleep related upper airway obstruction in a cohort with Down’s syndrome Arch Dis Child 1991 66 1333 1338 1836718 55 Shott SR Down syndrome: common otolaryngologic manifestations Am J Med Genet C Semin Med Genet 2006 142C 131 140 16838306 56 Bezold LI Pignatelli R Altman CA Intraoperative transesophageal echocardiography in congenital heart surgery. The Texas Children’s Hospital experience Tex Heart Inst J 1996 23 108 115 8792541 57 Hilberath JN Oakes DA Shernan SK Safety of transesophageal echocardiography J Am Soc Echocardiogr 2010 23 1115 1127 (quiz 1220–1221) 20864313 58 Cooney TP Thurlbeck WM Pulmonary hypoplasia in Down’s syndrome N Engl J Med 1982 307 1170 1173 6214715 59 McDowell KM Craven DI Pulmonary complications of Down syndrome during childhood J Pediatr 2011 158 319 325 20846671 60 Rutigliani M Boccardo F Campisi C Immunohistochemical studies in a hydroptic fetus with pulmonary lymphangiectasia and trisomy 21 Lymphology 2007 40 114 121 18062612 61 Turan O Canter B Ergenekon E Koc E Atalay Y Chylothorax and respiratory distress in a newborn with trisomy 21 Eur J Pediatr 2001 160 744 745 11795686 62 Miera O Mildenberger E van Baalen A Fuhr N Neonatal chylothorax with trisomy 21 Z Geburtshilfe Neonatol 2004 208 29 31 15039889 63 Ochiai M Hikino S Nakayama H Nonimmune hydrops fetalis due to generalized lymphatic dysplasia in an infant with Robertsonian trisomy 21 Am J Perinatol 2006 23 63 66 16450276 64 Ip P Chiu CS Cheung YF Risk factors prolonging ventilation in young children after cardiac surgery: impact of noninfectious pulmonary complications Pediatr Crit Care Med 2002 3 269 274 12780968 65 Morray JP Mac Gillivray R Duker G Increased perioperative risk following repair of congenital heart disease in Down’s syndrome Anesthesiology 1986 65 221 224 2943193 66 Campbell RM Benson LN Williams WW Adatia I Chylopericardium after cardiac operations in children Ann Thorac Surg 2001 72 193 196 11465177 67 Harrison AM Cox AC Davis S Failed extubation after cardiac surgery in young children: prevalence, pathogenesis, and risk factors Pediatr Crit Care Med 2002 3 148 152 12780985 68 Kusters MA Verstegen RH Gemen EF de Vries E Intrinsic defect of the immune system in children with Down syndrome: a review Clin Exp Immunol 2009 156 189 193 19250275 69 Ram G Chinen J Infections and immunodeficiency in Down syndrome Clin Exp Immunol 2011 164 9 16 21352207 70 Malec E Mroczek T Pajak J Januszewska K Zdebska E Results of surgical treatment of congenital heart defects in children with Down’s syndrome Pediatr Cardiol 1999 20 351 354 10441690 71 Roussot MA Lawrenson JB Hewitson J Smart R De Decker HP Is cardiac surgery warranted in children with Down syndrome? A case-controlled review S Afr Med J 2006 96 Pt 2 924 930 17077919 72 Murphy J Philip M Macken S Thyroid dysfunction in Down’s syndrome and screening for hypothyroidism in children and adolescents using capillary TSH measurement J Pediatr Endocrinol Metab 2008 21 155 163 18422028 73 Gibson PA Newton RW Selby K Longitudinal study of thyroid function in Down’s syndrome in the first two decades Arch Dis Child 2005 90 574 578 15908619 74 Talwar S Khadgawat R Sandeep JA Cardiopulmonary bypass and serum thyroid hormone profile in pediatric patients with congenital heart disease Congenit Heart Dis 2012 7 433 440 22613232 75 Plumpton KR Anderson BJ Beca J Thyroid hormone and cortisol concentrations after congenital heart surgery in infants younger than 3 months of age Intensive Care Med 2010 36 321 328 19809804 76 Davidson RG Atlantoaxial instability in individuals with Down syndrome: a fresh look at the evidence Pediatrics 1988 81 857 865 2966926 77 Goldberg-Stern H Strawsburg RH Patterson B Seizure frequency and characteristics in children with Down syndrome Brain Dev 2001 23 375 378 11578846 78 Tezenas Du Montcel S Mendizabai H Ayme S Levy A Philip N Prevalence of 22q11 microdeletion J Med Genet 1996 33 719 79 Goodship J Cross I LiLing J Wren C A population study of chromosome 22q11 deletions in infancy Arch Dis Child 1998 79 348 351 9875047 80 Bassett AS McDonald-McGinn DM Devriendt K Practical guidelines for managing patients with 22q11.2 deletion syndrome J Pediatr 2011 159 332.e1 339.e1 21570089 81 Ryan AK Goodship JA Wilson DI Spectrum of clinical features associated with interstitial chromosome 22q11 deletions: a European collaborative study J Med Genet 1997 34 798 804 9350810 82 Goldmuntz E Clark BJ Mitchell LE Frequency of 22q11 deletions in patients with conotruncal defects J Am Coll Cardiol 1998 32 492 498 9708481 83 McElhinney DB Clark BJ 3rd Weinberg PM Association of chromosome 22q11 deletion with isolated anomalies of aortic arch laterality and branching J Am Coll Cardiol 2001 37 2114 2119 11419896 84 McElhinney DB Driscoll DA Levin ER Chromosome 22q11 deletion in patients with ventricular septal defect: frequency and associated cardiovascular anomalies Pediatrics 2003 112 Pt 1 e472 14654648 85 Marmon LM Balsara RK Chen R Dunn JM Congenital cardiac anomalies associated with the DiGeorge syndrome: a neonatal experience Ann Thorac Surg 1984 38 146 150 6465992 86 Mercer-Rosa L Pinto N Yang W Tanel R Goldmuntz E 22q11.2 deletion syndrome is associated with perioperative outcome in tetralogy of Fallot J Thorac Cardiovasc Surg 2013 146 868 873 23312975 87 O’Byrne ML Yang W Mercer-Rosa L 22q11.2 deletion syndrome is associated with increased perioperative events and more complicated postoperative course in infants undergoing infant operative correction of truncus arteriosus communis or interrupted aortic arch J Thorac Cardiovasc Surg 2014 148 1597 1605 24629220 88 McDonald R Dodgen A Goyal S Impact of 22q11.2 deletion on the postoperative course of children after cardiac surgery Pediatr Cardiol 2013 34 341 347 22864648 89 Reddy VM McElhinney DB Amin Z Early and intermediate outcomes after repair of pulmonary atresia with ventricular septal defect and major aortopulmonary collateral arteries: experience with 85 patients Circulation 2000 101 1826 1832 10769284 90 Mahle WT Crisalli J Coleman K Deletion of chromosome 22q11.2 and outcome in patients with pulmonary atresia and ventricular septal defect Ann Thorac Surg 2003 76 567 571 12902105 91 Carotti A Marino B Di Donato RM Influence of chromosome 22q11.2 microdeletion on surgical outcome after treatment of tetralogy of Fallot with pulmonary atresia J Thorac Cardiovasc Surg 2003 126 1666 1667 14666061 92 Carotti A Albanese SB Di Donato RM Unifocalization and repair of pulmonary atresia with ventricular septal defect and major aortopulmonary collateral arteries Acta Paediatr Suppl 2006 95 22 26 16801161 93 Carotti A Albanese SB Filippelli S Determinants of outcome after surgical treatment of pulmonary atresia with ventricular septal defect and major aortopulmonary collateral arteries J Thorac Cardiovasc Surg 2010 140 1092 1103 20850144 94 Yotsui-Tsuchimochi H Higa K Matsunaga M Nitahara K Shono S Anesthetic management of a child with chromosome 22q11 deletion syndrome Paediatr Anaesth 2006 16 454 457 16618302 95 McElhinney DB Jacobs I McDonald-McGinn DM Zackai EH Goldmuntz E Chromosomal and cardiovascular anomalies associated with congenital laryngeal web Int J Pediatr Otorhinolaryngol 2002 66 23 27 12363418 96 Yamagishi H Maeda J Higuchi M Bronchomalacia associated with pulmonary atresia, ventricular septal defect and major aortopulmonary collateral arteries, and chromosome 22q11.2 deletion Clin Genet 2002 62 214 219 12220436 97 Ackerman MJ Wylam ME Feldt RH Pulmonary atresia with ventricular septal defect and persistent airway hyperresponsiveness J Thorac Cardiovasc Surg 2001 122 169 177 11436051 98 Asija R Hanley FL Roth SJ Postoperative respiratory failure in children with tetralogy of Fallot, pulmonary atresia, and major aortopulmonary collaterals: a pilot study Pediatr Crit Care Med 2013 14 384 389 23439458 99 Ziolkowska L Kawalec W Turska-Kmiec A Chromosome 22q11.2 microdeletion in children with conotruncal heart defects: frequency, associated cardiovascular anomalies, and outcome following cardiac surgery Eur J Pediatr 2008 167 1135 1140 18172682 100 Jatana V Gillis J Webster BH Ades LC Deletion 22q11.2 syndrome – implications for the intensive care physician Pediatr Crit Care Med 2007 8 459 463 (quiz 64) 17873780 101 Kobrynski LJ Sullivan KE Velocardiofacial syndrome, DiGeorge syndrome: the chromosome 22q11.2 deletion syndromes Lancet 2007 370 1443 1452 17950858 102 Patel K Akhter J Kobrynski L Immunoglobulin deficiencies: the B-lymphocyte side of DiGeorge Syndrome J Pediatr 2012 161 950 953 22809661 103 Smith CA Driscoll DA Emanuel BS Increased prevalence of immunoglobulin A deficiency in patients with the chromosome 22q11.2 deletion syndrome (DiGeorge syndrome/velocardiofacial syndrome) Clin Diagn Lab Immunol 1998 5 415 417 9606003 104 Carotti A Digilio MC Piacentini G Cardiac defects and results of cardiac surgery in 22q11.2 deletion syndrome Dev Disabil Res Rev 2008 14 35 42 18636635 105 Cheung EN George SR Costain GA Prevalence of hypocalcaemia and its associated features in 22q11.2 deletion syndrome Clin Endocrinol 2014 81 190 196 106 Cheung EN George SR Andrade DM Neonatal hypocalcemia, neonatal seizures, and intellectual disability in 22q11.2 deletion syndrome Genet Med 2014 16 40 44 23765047 107 Stagi S Lapi E Gambineri E Thyroid function and morphology in subjects with microdeletion of chromosome 22q11 (del(22)(q11)) Clin Endocrinol 2010 72 839 844 108 Naqvi N Davidson SJ Wong D Predicting 22q11.2 deletion syndrome: a novel method using the routine full blood count Int J Cardiol 2011 150 50 53 20363518 109 Latger-Cannard V Bensoussan D Gregoire MJ Frequency of thrombocytopenia and large platelets correlates neither with conotruncal cardiac anomalies nor immunological features in the chromosome 22q11.2 deletion syndrome Eur J Pediatr 2004 163 327 328 15346916 110 Liang HP Morel-Kopp MC Curtin J Heterozygous loss of platelet glycoprotein (GP) Ib-V-IX variably affects platelet function in velocardiofacial syndrome (VCFS) patients Thromb Haemost 2007 98 1298 1308 18064328 111 Kato T Kosaka K Kimura M Thrombocytopenia in patients with 22q11.2 deletion syndrome and its association with glycoprotein Ib-beta Genet Med 2003 5 113 119 12644781 112 Nakagawa M Okuno M Okamoto N Fujino H Kato H Bernard-Soulier syndrome associated with 22q11.2 microdeletion Am J Med Genet 2001 99 286 288 11251994 113 Stewart TL Irons MB Cowan JM Bianchi DW Increased incidence of renal anomalies in patients with chromosome 22q11 microdeletion Teratology 1999 59 20 22 9988879 114 Shashi V Berry MN Hines MH Vasomotor instability in neonates with chromosome 22q11 deletion syndrome Am J Med Genet A 2003 121A 231 234 12923863 115 Lin AE Ticho BS Houde K Westgate MN Holmes LB Heterotaxy: associated conditions and hospital-based prevalence in newborns Genet Med 2000 2 157 172 11256661 116 Sutherland MJ Ware SM Disorders of left-right asymmetry: heterotaxy and situs inversus Am J Med Genet C Semin Med Genet 2009 151C 307 317 19876930 117 Jacobs JP Anderson RH Weinberg PM The nomenclature, definition and classification of cardiac structures in the setting of heterotaxy Cardiol Young 2007 17 Suppl 2 1 28 118 Lim JS McCrindle BW Smallhorn JF Clinical features, management, and outcome of children with fetal and postnatal diagnoses of isomerism syndromes Circulation 2005 112 2454 2461 16216960 119 Yildirim SV Tokel K Varan B Aslamaci S Ekici E Clinical investigations over 13 years to establish the nature of the cardiac defects in patients having abnormalities of lateralization Cardiol Young 2007 17 275 282 17615645 120 Freedom RM Jaeggi ET Lim JS Anderson RH Hearts with isomerism of the right atrial appendages – one of the worst forms of disease in 2005 Cardiol Young 2005 15 554 567 16297247 121 Gilljam T McCrindle BW Smallhorn JF Williams WG Freedom RM Outcomes of left atrial isomerism over a 28-year period at a single institution J Am Coll Cardiol 2000 36 908 916 10987619 122 Azakie A Merklinger SL Williams WG Improving outcomes of the Fontan operation in children with atrial isomerism and heterotaxy syndromes Ann Thorac Surg 2001 72 1636 1640 11722057 123 Stamm C Friehs I Duebener LF Improving results of the modified Fontan operation in patients with heterotaxy syndrome Ann Thorac Surg 2002 74 1967 1977 (discussion 78) 12643382 124 Kim SJ Kim WH Lim HG Lee JY Outcome of 200 patients after an extracardiac Fontan procedure J Thorac Cardiovasc Surg 2008 136 108 116 18603062 125 Wren C Macartney FJ Deanfield JE Cardiac rhythm in atrial isomerism Am J Cardiol 1987 59 1156 1158 3578058 126 Wu MH Wang JK Lin JL Supraventricular tachycardia in patients with right atrial isomerism J Am Coll Cardiol 1998 32 773 779 9741526 127 Anagnostopoulos PV Pearl JM Octave C Improved current era outcomes in patients with heterotaxy syndromes Eur J Cardiothorac Surg 2009 35 871 877 (discussion 877–888) 19233677 128 Wessels MW De Graaf BM Cohen-Overbeek TE A new syndrome with noncompaction cardiomyopathy, bradycardia, pulmonary stenosis, atrial septal defect and heterotaxy with suggestive linkage to chromosome 6p Hum Genet 2008 122 595 603 17938964 129 Jacobs JP Pasquali SK Morales DL Heterotaxy: lessons learned about patterns of practice and outcomes from the congenital heart surgery database of the society of thoracic surgeons World J Pediatr Congenit Heart Surg 2011 2 278 286 23804985 130 Swisher M Jonas R Tian X Increased postoperative and respiratory complications in patients with congenital heart disease associated with heterotaxy J Thorac Cardiovasc Surg 2011 141 637 644 20884020 131 Ota N Fujimoto Y Murata M Improving outcomes of the surgical management of right atrial isomerism Ann Thorac Surg 2012 93 832 838 (discussion 838–839) 22112795 132 Hashmi A Abu-Sulaiman R McCrindle BW Management and outcomes of right atrial isomerism: a 26-year experience J Am Coll Cardiol 1998 31 1120 1126 9562017 133 Jenkins KJ Sanders SP Orav EJ Individual pulmonary vein size and survival in infants with totally anomalous pulmonary venous connection J Am Coll Cardiol 1993 22 201 206 8509542 134 Morales DL Braud BE Booth JH Heterotaxy patients with total anomalous pulmonary venous return: improving surgical results Ann Thorac Surg 2006 82 1621 1627 (discussion 1627–1628) 17062215 135 Bartz PJ Driscoll DJ Dearani JA Early and late results of the modified Fontan operation for heterotaxy syndrome 30 years of experience in 142 patients J Am Coll Cardiol 2006 48 2301 2305 17161263 136 Humes RA Feldt RH Porter CJ The modified Fontan operation for asplenia and polysplenia syndromes J Thorac Cardiovasc Surg 1988 96 212 218 3398543 137 Larsen RL Eguchi JH Mulla NF Usefulness of cardiac transplantation in children with visceral heterotaxy (asplenic and polysplenic syndromes and single right-sided spleen with levocardia) and comparison of results with cardiac transplantation in children with dilated cardiomyopathy Am J Cardiol 2002 89 1275 1279 12031727 138 Jacobs JP Asante-Korang A O’Brien SM Lessons learned from 119 consecutive cardiac transplants for pediatric and congenital heart disease Ann Thorac Surg 2011 91 1248 1254 (discussion 1254–1255) 21440154 139 Nakhleh N Francis R Giese RA High prevalence of respiratory ciliary dysfunction in congenital heart disease patients with heterotaxy Circulation 2012 125 2232 2242 22499950 140 Harden B Tian X Giese R Increased postoperative respiratory complications in heterotaxy congenital heart disease patients with respiratory ciliary dysfunction J Thorac Cardiovasc Surg 2014 147 1291.e2 1298.e2 23886032 141 Shiima-Kinoshita C Min KY Hanafusa T Mori H Nakahari T Beta 2-adrenergic regulation of ciliary beat frequency in rat bronchiolar epithelium: potentiation by isosmotic cell shrinkage J Physiol 2004 554 Pt 2 403 416 14594991 142 Anderson C Devine WA Anderson RH Debich DE Zuberbuhler JR Abnormalities of the spleen in relation to congenital malformations of the heart: a survey of necropsy findings in children Br Heart J 1990 63 122 128 2317406 143 Waldman JD Rosenthal A Smith AL Shurin S Nadas AS Sepsis and congenital asplenia J Pediatr 1977 90 555 559 320299 144 de Porto AP Lammers AJ Bennink RJ Assessment of splenic function Eur J Clin Microbiol Infect Dis 2010 29 1465 1473 20853172 145 Serraf A Bensari N Houyel L Surgical management of congenital heart defects associated with heterotaxy syndrome Eur J Cardiothorac Surg 2010 38 721 727 20451403 146 Williams GD Feng A Heterotaxy syndrome: implications for anesthesia management J Cardiothorac Vasc Anesth 2010 24 834 844 20421166 147 Ticho BS Goldstein AM Van Praagh R Extracardiac anomalies in the heterotaxy syndromes with focus on anomalies of midline-associated structures Am J Cardiol 2000 85 729 734 12000048 148 Nielsen J Wohlert M Sex chromosome abnormalities found among 34,910 newborn children: results from a 13-year incidence study in Arhus, Denmark Birth Defects Orig Artic Ser 1990 26 209 223 2090319 149 Bondy CA Care of girls and women with Turner syndrome: a guideline of the Turner Syndrome Study Group J Clin Endocrinol Metab 2007 92 10 25 17047017 150 Frias JL Davenport ML Health supervision for children with Turner syndrome Pediatrics 2003 111 692 702 12612263 151 Surerus E Huggon IC Allan LD Turner’s syndrome in fetal life Ultrasound Obstet Gynecol 2003 22 264 267 12942498 152 Ho VB Bakalov VK Cooley M Major vascular anomalies in Turner syndrome: prevalence and magnetic resonance angiographic features Circulation 2004 110 1694 1700 15353492 153 Dawson-Falk KL Wright AM Bakker B Cardiovascular evaluation in Turner syndrome: utility of MR imaging Australas Radiol 1992 36 204 209 1445102 154 Bondy CA Congenital cardiovascular disease in Turner syndrome Congenit Heart Dis 2008 3 2 15 18373744 155 Gotzsche CO Krag-Olsen B Nielsen J Sorensen KE Kristensen BO Prevalence of cardiovascular malformations and association with karyotypes in Turner’s syndrome Arch Dis Child 1994 71 433 436 7826114 156 Wong SC Burgess T Cheung M Zacharin M The prevalence of turner syndrome in girls presenting with coarctation of the aorta J Pediatr 2014 164 259 263 24172638 157 Matura LA Ho VB Rosing DR Bondy CA Aortic dilatation and dissection in Turner syndrome Circulation 2007 116 1663 1670 17875973 158 Carlson M Airhart N Lopez L Silberbach M Moderate aortic enlargement and bicuspid aortic valve are associated with aortic dissection in Turner syndrome: report of the international Turner syndrome aortic dissection registry Circulation 2012 126 2220 2226 23032325 159 Bondy CA Ceniceros I Van PL Bakalov VK Rosing DR Prolonged rate-corrected QT interval and other electrocardiogram abnormalities in girls with Turner syndrome Pediatrics 2006 118 e1220 e1225 17015510 160 Cramer JW Bartz PJ Simpson PM Zangwill SD The spectrum of congenital heart disease and outcomes after surgical repair among children with Turner syndrome: a single-center review Pediatr Cardiol 2014 35 253 260 23933717 161 Reis PM Punch MR Bove EL van de Ven CJ Outcome of infants with hypoplastic left heart and Turner syndromes Obstet Gynecol 1999 93 532 535 10214828 162 Ravelo HR Stephenson LW Friedman S Coarctation resection in children with Turner’s syndrome: a note of caution J Thorac Cardiovasc Surg 1980 80 427 430 7412347 163 Brandt B 3rd Heintz SE Rose EF Ehrenhaft JL Clark EB Repair of coarctation of the aorta in children with Turner syndrome Pediatr Cardiol 1984 5 175 177 6531260 164 Oza NM Siegenthaler M Horvath K Serious aortic complications in a patient with Turner syndrome Eur J Pediatr 2013 172 703 705 22923005 165 Badmanaban B Mole D Sarsam MA Descending aortic dissection post coarctation repair in a patient with Turner’s syndrome J Card Surg 2003 18 153 154 12757343 166 Fejzic Z van Oort A Fatal dissection of the descending aorta after implantation of a stent in a 19-year-old female with Turner’s syndrome Cardiol Young 2005 15 529 531 16164796 167 Kataoka K Ozawa A Inage A Benson LN Transcatheter repair of native coarctation in children with Turner syndrome: three case reports and literature review Congenit Heart Dis 2006 1 315 320 18377500 168 Zanjani KS Thanopoulos BD Peirone A Alday L Giannakoulas G Usefulness of stenting in aortic coarctation in patients with the Turner syndrome Am J Cardiol 2010 106 1327 1331 21029833 169 Bellini C Boccardo F Campisi C Bonioli E Congenital pulmonary lymphangiectasia Orphanet J Rare Dis 2006 1 43 17074089 170 Gawlik A Gawlik T Januszek-Trzciakowska A Patel H Malecka-Tendera E Incidence and dynamics of thyroid dysfunction and thyroid autoimmunity in girls with Turner’s syndrome: a long-term follow-up study Horm Res Paediatr 2011 76 314 320 21997702 171 Stromme P Bjornstad PG Ramstad K Prevalence estimation of Williams syndrome J Child Neurol 2002 17 269 271 12088082 172 Pober BR Williams-Beuren syndrome N Engl J Med 2010 362 239 252 20089974 173 Collins RT 2nd Cardiovascular disease in Williams syndrome Circulation 2013 127 2125 2134 23716381 174 Wren C Oslizlok P Bull C Natural history of supravalvular aortic stenosis and pulmonary artery stenosis J Am Coll Cardiol 1990 15 1625 1630 2345244 175 Giddins NG Finley JP Nanton MA Roy DL The natural course of supravalvar aortic stenosis and peripheral pulmonary artery stenosis in William’s syndrome Br Heart J 1989 62 315 319 2803879 176 Del Pasqua A Rinelli G Toscano A New findings concerning cardiovascular manifestations emerging from long-term follow-up of 150 patients with the Williams-Beuren-Beuren syndrome Cardiol Young 2009 19 563 567 19941695 177 Wessel A Pankau R Kececioglu D Ruschewski W Bursch JH Three decades of follow-up of aortic and pulmonary vascular lesions in the Williams-Beuren syndrome Am J Med Genet 1994 52 297 301 7810560 178 Collins RT 2nd Kaplan P Somes GW Rome JJ Long-term outcomes of patients with cardiovascular abnormalities and Williams syndrome Am J Cardiol 2010 105 874 878 20211336 179 Stamm C Friehs I Moran AM Surgery for bilateral outflow tract obstruction in elastin arteriopathy J Thorac Cardiovasc Surg 2000 120 755 763 11003759 180 Deo SV Burkhart HM Schaff HV Late outcomes for surgical repair of supravalvar aortic stenosis Ann Thorac Surg 2012 94 854 859 22727249 181 Monge MC Mainwaring RD Sheikh AY Surgical reconstruction of peripheral pulmonary artery stenosis in Williams and Alagille syndromes J Thorac Cardiovasc Surg 2013 145 476 481 23228407 182 Apostolopoulou SC Kelekis NL Laskari C Kaklamanis L Rammos S Restenosis and pseudoaneurysm formation after stent placement for aortic coarctation in Williams syndrome J Vasc Interv Radiol 2002 13 547 548 11997368 183 Mookerjee J Roebuck D Derrick G Restenosis after aortic stenting Cardiol Young 2004 14 210 211 15691415 184 Collins RT 2nd Kaplan P Rome JJ Stenosis of the thoracic aorta in Williams syndrome Pediatr Cardiol 2010 31 829 833 20411252 185 Pham PP Moller JH Hills C Larson V Pyles L Cardiac catheterization and operative outcomes from a multicenter consortium for children with Williams syndrome Pediatr Cardiol 2009 30 9 14 19052807 186 Geggel RL Gauvreau K Lock JE Balloon dilation angioplasty of peripheral pulmonary stenosis associated with Williams syndrome Circulation 2001 103 2165 2170 11331257 187 Burch TM McGowan FX Jr Kussman BD Powell AJ DiNardo JA Congenital supravalvular aortic stenosis and sudden death associated with anesthesia: what’s the mystery? Anesth Analg 2008 107 1848 1854 19020129 188 Bird LM Billman GF Lacro RV Sudden death in Williams syndrome: report of ten cases J Pediatr 1996 129 926 931 8969740 189 Wessel A Gravenhorst V Buchhorn R Risk of sudden death in the Williams-Beuren syndrome Am J Med Genet A 2004 127A 234 237 15150772 190 Conway EE Jr Noonan J Marion RW Steeg CN Myocardial infarction leading to sudden death in the Williams syndrome: report of three cases J Pediatr 1990 117 593 595 2213386 191 Stamm C Li J Ho SY Redington AN Anderson RH The aortic root in supravalvular aortic stenosis: the potential surgical relevance of morphologic findings J Thorac Cardiovasc Surg 1997 114 16 24 9240289 192 Bonnet D Cormier V Villain E Bonhoeffer P Kachaner J Progressive left main coronary artery obstruction leading to myocardial infarction in a child with Williams syndrome Eur J Pediatr 1997 156 751 753 9365061 193 van Pelt NC Wilson NJ Lear G Severe coronary artery disease in the absence of supravalvular stenosis in a patient with Williams syndrome Pediatr Cardiol 2005 26 665 667 15549615 194 Martin EC Moseley IF Supravalvar aortic stenosis Br Heart J 1973 35 758 765 4718826 195 Collins RT 2nd Aziz PF Gleason MM Kaplan PB Shah MJ Abnormalities of cardiac repolarization in Williams syndrome Am J Cardiol 2010 106 1029 1033 20854969 196 Collins RT 2nd Aziz PF Swearingen CJ Kaplan PB Relation of ventricular ectopic complexes to QTc interval on ambulatory electrocardiograms in Williams syndrome Am J Cardiol 2012 109 1671 1676 22459308 197 Broder K Reinhardt E Ahern J Elevated ambulatory blood pressure in 20 subjects with Williams syndrome Am J Med Genet 1999 83 356 360 10232742 198 Kaplan P Levinson M Kaplan BS Cerebral artery stenoses in Williams syndrome cause strokes in childhood J Pediatr 1995 126 943 945 7776101 199 Stagi S Bindi G Neri AS Thyroid function and morphology in patients affected by Williams syndrome Clin Endocrinol 2005 63 456 460 200 Medley J Russo P Tobias JD Perioperative care of the patient with Williams syndrome Paediatr Anaesth 2005 15 243 247 15725324 201 Cambiaso P Orazi C Digilio MC Thyroid morphology and subclinical hypothyroidism in children and adolescents with Williams syndrome J Pediatr 2007 150 62 65 17188616 202 Committee on Genetics American Academy of Pediatrics: health care supervision for children with Williams syndrome Pediatrics 2001 107 1192 1204 11331709 203 Stagi S Manoni C Salti R Cecchi C Chiarelli F Thyroid hypoplasia as a cause of congenital hypothyroidism in Williams syndrome Horm Res 2008 70 316 318 18824871 204 Pankau R Partsch CJ Winter M Gosch A Wessel A Incidence and spectrum of renal abnormalities in Williams-Beuren syndrome Am J Med Genet 1996 63 301 304 8723124 205 Sforzini C Milani D Fossali E Renal tract ultrasonography and calcium homeostasis in Williams-Beuren syndrome Pediatr Nephrol 2002 17 899 902 12432430 206 Ingelfinger JR Newburger JW Spectrum of renal anomalies in patients with Williams syndrome J Pediatr 1991 119 771 773 1941384 207 Pober BR Lacro RV Rice C Mandell V Teele RL Renal findings in 40 individuals with Williams syndrome Am J Med Genet 1993 46 271 274 8488870 208 Roberts AE Allanson JE Tartaglia M Gelb BD Noonan syndrome Lancet 2013 381 333 342 23312968 209 Romano AA Allanson JE Dahlgren J Noonan syndrome: clinical features, diagnosis, and management guidelines Pediatrics 2010 126 746 759 20876176 210 Roberts A Allanson J Jadico SK The cardiofaciocutaneous syndrome J Med Genet 2006 43 833 842 16825433 211 Cesarini L Alfieri P Pantaleoni F Cognitive profile of disorders associated with dysregulation of the RAS/MAPK signaling cascade Am J Med Genet A 2009 149A 140 146 19133693 212 Axelrad ME Glidden R Nicholson L Gripp KW Adaptive skills, cognitive, and behavioral characteristics of Costello syndrome Am J Med Genet A 2004 128A 396 400 15264285 213 Prendiville TW Gauvreau K Tworog-Dube E Cardiovascular disease in Noonan syndrome Arch Dis Child 2014 99 629 634 24534818 214 Hickey EJ Mehta R Elmi M Survival implications: hypertrophic cardiomyopathy in Noonan syndrome Congenit Heart Dis 2011 6 41 47 21269411 215 Pandit B Sarkozy A Pennacchio LA Gain-of-function RAF1 mutations cause Noonan and LEOPARD syndromes with hypertrophic cardiomyopathy Nat Genet 2007 39 1007 1012 17603483 216 Ishikawa Y Sekiguchi K Akasaka Y Fibromuscular dysplasia of coronary arteries resulting in myocardial infarction associated with hypertrophic cardiomyopathy in Noonan’s syndrome Hum Pathol 2003 34 282 284 12673564 217 Croonen EA van der Burgt I Kapusta L Draaisma JM Electrocardiography in Noonan syndrome PTPN11 gene mutation – phenotype characterization Am J Med Genet A 2008 146 350 353 218 Lin AE Alexander ME Colan SD Clinical, pathological, and molecular analyses of cardiovascular abnormalities in Costello syndrome: a Ras/MAPK pathway syndrome Am J Med Genet A 2011 155A 486 507 21344638 219 Shaw AC Kalidas K Crosby AH Jeffery S Patton MA The natural history of Noonan syndrome: a long-term follow-up study Arch Dis Child 2007 92 128 132 16990350 220 Witt DR McGillivray BC Allanson JE Bleeding diathesis in Noonan syndrome: a common association Am J Med Genet 1988 31 305 317 3232698 221 Sharland M Patton MA Talbot S Chitolie A Bevan DH Coagulation-factor deficiencies and abnormal bleeding in Noonan’s syndrome Lancet 1992 339 19 21 1345952 222 Kitchens CS Alexander JA Partial deficiency of coagulation factor XI as a newly recognized feature of Noonan syndrome J Pediatr 1983 102 224 227 6822926 223 Wiegand G Hofbeck M Zenker M Budde U Rauch R Bleeding diathesis in Noonan syndrome: is acquired von Willebrand syndrome the clue? Thromb Res 2012 130 e251 e254 22985731 224 Nunes P Aguilar S Prado SN Severe congenital thrombocytopaenia – first clinical manifestation of Noonan syndrome BMJ Case Rep 2012 2012 pii: bcr1020114940 225 Artoni A Selicorni A Passamonti SM Hemostatic abnormalities in Noonan Syndrome Pediatrics 2014 133 e1299 e1304 24753526 226 Tofil NM Winkler MK Watts RG Noonan J The use of recombinant factor VIIa in a patient with Noonan syndrome and life-threatening bleeding Pediatr Crit Care Med 2005 6 352 354 15857538 227 Dineen RA Lenthall RK Aneurysmal sub-arachnoid haemorrhage in patients with Noonan syndrome: a report of two cases and review of neurovascular presentations in this syndrome Neuroradiology 2004 46 301 305 15034699 228 Ho WL Wang JK Li YW Radiological features of late-onset lymphoedema in Noonan’s syndrome Pediatr Radiol 2003 33 200 202 12612821 229 Fabretto A Kutsche K Harmsen MB Two cases of Noonan syndrome with severe respiratory and gastroenteral involvement and the SOS1 mutation F623I Eur J Med Genet 2010 53 322 324 20673819 230 Hernandez RJ Stern AM Rosenthal A Pulmonary lymphangiectasis in Noonan syndrome AJR Am J Roentgenol 1980 134 75 80 6766041 231 Goens MB Campbell D Wiggins JW Spontaneous chylothorax in Noonan syndrome. Treatment with prednisone Am J Dis Child 1992 146 1453 1456 1456257 232 Tsang HY Cheung YF Leung MP Chau KT Cutaneous oozing of lymphatic fluid after interventional cardiac catheterization in a patient with Noonan syndrome Catheter Cardiovasc Interv 2000 51 441 443 11108676 233 Dietz HC Loeys B Carta L Ramirez F Recent progress towards a molecular understanding of Marfan syndrome Am J Med Genet C Semin Med Genet 2005 139C 4 9 16273535 234 Loeys BL Dietz HC Braverman AC The revised Ghent nosology for the Marfan syndrome J Med Genet 2010 47 476 485 20591885 235 Gott VL Greene PS Alejo DE Replacement of the aortic root in patients with Marfan’s syndrome N Engl J Med 1999 340 1307 1313 10219065 236 Cattaneo SM Bethea BT Alejo DE Surgery for aortic root aneurysm in children: a 21-year experience in 50 patients Ann Thorac Surg 2004 77 168 176 14726057 237 Everitt MD Pinto N Hawkins JA Cardiovascular surgery in children with Marfan syndrome or Loeys-Dietz syndrome J Thorac Cardiovasc Surg 2009 137 1327 1332 (discussion 1332–1333) 19464442 238 Roubertie F Ben Ali W Raisky O Aortic root replacement in children: a word of caution about valve-sparing procedures Eur J Cardiothorac Surg 2009 35 136 140 19008114 239 Wood JR Bellamy D Child AH Citron KM Pulmonary disease in patients with Marfan syndrome Thorax 1984 39 780 784 6495247 240 Booms P Cisler J Mathews KR Novel exon skipping mutation in the fibrillin-1 gene: two “hot spots” for the neonatal Marfan syndrome Clin Genet 1999 55 110 117 10189088 241 Sutherell J Zarate Y Tinkle BT Novel fibrillin 1 mutation in a case of neonatal Marfan syndrome: the increasing importance of early recognition Congenit Heart Dis 2007 2 342 346 18377451 242 Strigl S Quagebeur JM Gersony WM Quadrivalvar replacement in infantile Marfan syndrome Pediatr Cardiol 2007 28 403 405 17687590 243 Krasemann T Kotthoff S Kehl HG Cardiac transplantation in neonatal Marfan syndrome – a life-saving approach Thorac Cardiovasc Surg 2005 53 Suppl 2 S146 S148 15704038 244 Drera B Ritelli M Zoppi N Loeys-Dietz syndrome type I and type II: clinical findings and novel mutations in two Italian patients Orphanet J Rare Dis 2009 4 24 19883511 245 Muramatsu Y Kosho T Magota M Progressive aortic root and pulmonary artery aneurysms in a neonate with Loeys-Dietz syndrome type 1B Am J Med Genet A 2010 152A 417 421 20101701 246 Williams JA Loeys BL Nwakanma LU Early surgical experience with Loeys-Dietz: a new syndrome of aggressive thoracic aortic aneurysm disease Ann Thorac Surg 2007 83 S757 S763 (discussion S785–S790) 17257922 247 Maccarrick G Black JH 3rd Bowdin S Loeys-Dietz syndrome: a primer for diagnosis and management Genet Med 2014 16 576 587 24577266 248 Malhotra A Westesson PL Loeys-Dietz syndrome Pediatr Radiol 2009 39 1015 19471917 249 Cleuziou J Eichinger WB Schreiber C Lange R Aortic root replacement with re-implantation technique in an infant with Loeys-Dietz syndrome and a bicuspid aortic valve Pediatr Cardiol 2010 31 117 119 19784694 250 Patel ND Arnaoutakis GJ George TJ Valve-sparing aortic root replacement in Loeys-Dietz syndrome Ann Thorac Surg 2011 92 556 560 (discussion 560–561) 21801912 251 Kawazu Y Inamura N Kayatani F Okamoto N Morisaki H Prenatal complex congenital heart disease with Loeys-Dietz syndrome Cardiol Young 2012 22 116 119 21774844 252 Nishida K Tamura S Yamazaki S Postoperative mitral leaflet rupture in an infant with Loeys-Dietz syndrome Pediatr Int 2014 56 e82 e85 25521989 253 Kirmani S Tebben PJ Lteif AN Germline TGF-beta receptor mutations and skeletal fragility: a report on two patients with Loeys-Dietz syndrome Am J Med Genet A 2010 152A 1016 1019 20358619 254 Tan EW Offoha RU Oswald GL Increased fracture risk and low bone mineral density in patients with Loeys-Dietz syndrome Am J Med Genet A 2013 161A 1910 1914 23825031 255 Fuhrhop SK McElroy MJ Dietz HC 3rd MacCarrick GL Sponseller PD High prevalence of cervical deformity and instability requires surveillance in Loeys-Dietz syndrome J Bone Joint Surg Am 2015 97 411 419 25740032 256 Turnpenny PD Ellard S Alagille syndrome: pathogenesis, diagnosis and management Eur J Hum Genet 2012 20 251 257 21934706 257 McElhinney DB Krantz ID Bason L Analysis of cardiovascular phenotype and genotype-phenotype correlation in individuals with a JAG1 mutation and/or Alagille syndrome Circulation 2002 106 2567 2574 12427653 258 Kamath BM Podkameni G Hutchinson AL Renal anomalies in Alagille syndrome: a disease-defining feature Am J Med Genet A 2012 158A 85 89 22105858 259 Blue GM Mah JM Cole AD The negative impact of Alagille syndrome on survival of infants with pulmonary atresia J Thorac Cardiovasc Surg 2007 133 1094 1096 17382662 260 Mainwaring RD Sheikh AY Punn R Reddy VM Hanley FL Surgical outcomes for patients with pulmonary atresia/major aortopulmonary collaterals and Alagille syndrome Eur J Cardiothorac Surg 2012 42 235 240 (discussion 240–241) 22402453 261 Bacha EA Hardin J Cronin DC Open-heart surgery in pediatric patients with end-stage liver disease Ann Thorac Surg 2004 78 e30 e33 15276586 262 Odim JN Wu J Laks H Banerji A Drant S Cardiac surgery in children with end-stage liver disease awaiting liver transplantation Ann Thorac Surg 2006 81 697 700 16427876 263 Kamath BM Spinner NB Emerick KM Vascular anomalies in Alagille syndrome: a significant cause of morbidity and mortality Circulation 2004 109 1354 1358 14993126 264 Lykavieris P Crosnier C Trichet C Meunier-Rotival M Hadchouel M Bleeding tendency in children with Alagille syndrome Pediatrics 2003 111 167 170 12509572 265 May L Hanley FL Connolly AJ Reddy S Atherosclerosis causing recurrent catastrophic aortopulmonary shunt dehiscence in a patient with Alagille syndrome Pediatr Cardiol 2013 34 1945 1948 22923029 266 Rasmussen SA Wong LY Yang Q May KM Friedman JM Population-based analyses of mortality in trisomy 13 and trisomy 18 Pediatrics 2003 111 Pt 1 777 784 12671111 267 McMahon CJ Chang AC Pignatelli RH Left ventricular noncompaction cardiomyopathy in association with trisomy 13 Pediatr Cardiol 2005 26 477 479 15549619 268 Yukifumi M Hirohiko S Fukiko I Mariko M Trisomy 13 in a 9-year-old girl with left ventricular noncompaction Pediatr Cardiol 2011 32 206 207 21082175 269 Maeda J Yamagishi H Furutani Y The impact of cardiac surgery in patients with trisomy 18 and trisomy 13 in Japan Am J Med Genet A 2011 155A 2641 2646 21990245 270 Graham EM Bradley SM Shirali GS Hills CB Atz AM Effectiveness of cardiac surgery in trisomies 13 and 18 (from the Pediatric Cardiac Care Consortium) Am J Cardiol 2004 93 801 803 15019900 271 Issekutz KA Graham JM Jr Prasad C Smith IM Blake KD An epidemiological analysis of CHARGE syndrome: preliminary results from a Canadian study Am J Med Genet A 2005 133A 309 317 15637722 272 Zentner GE Layman WS Martin DM Scacheri PC Molecular and phenotypic aspects of CHD7 mutation in CHARGE syndrome Am J Med Genet A 2010 152A 674 686 20186815 273 Lalani SR Safiullah AM Molinari LM SEMA3E mutation in a patient with CHARGE syndrome J Med Genet 2004 41 e94 15235037 274 Corsten-Janssen N Saitta SC Hoefsloot LH More clinical overlap between 22q11.2 deletion syndrome and CHARGE syndrome than often anticipated Mol Syndromol 2013 4 235 245 23885230 275 Blake KD Prasad C CHARGE syndrome Orphanet J Rare Dis 2006 1 34 16959034 276 Corsten-Janssen N Kerstjens-Frederikse WS du Marchie Sarvaas GJ The cardiac phenotype in patients with a CHD7 mutation Circ Cardiovasc Genet 2013 6 248 254 23677905 277 Lin AE Siebert JR Graham JM Jr Central nervous system malformations in the CHARGE Association Am J Med Genet 1990 37 304 310 2260555 278 Jyonouchi S McDonald-McGinn DM Bale S Zackai EH Sullivan KE CHARGE (coloboma, heart defect, atresia choanae, retarded growth and development, genital hypoplasia, ear anomalies/deafness) syndrome and chromosome 22q11.2 deletion syndrome: a comparison of immunologic and nonimmunologic phenotypic features Pediatrics 2009 123 e871 e877 19403480 279 Stack CG Wyse RK Incidence and management of airway problems in the CHARGE Association Anaesthesia 1991 46 582 585 1862904 280 Morgan D Bailey M Phelps P Ear-nose-throat abnormalities in the CHARGE Association Arch Otolaryngol Head Neck Surg 1993 119 49 54 8417743 281 Bergman JE Blake KD Bakker MK Death in CHARGE syndrome after the neonatal period Clin Genet 2010 77 232 240 20447140 282 Blake K MacCuspie J Hartshorne TS Postoperative airway events of individuals with CHARGE syndrome Int J Pediatr Otorhinolaryngol 2009 73 219 226 19058860 283 Blake KD Russell-Eggitt IM Morgan DW Ratcliffe JM Wyse RK Who’s in CHARGE? Multidisciplinary management of patients with CHARGE Association Arch Dis Child 1990 65 217 223 2317068 284 Wyse RK Al-Mahdawi S Burn J Blake K Congenital heart disease in CHARGE Association Pediatr Cardiol 1993 14 75 81 8469635 285 James PA Aftimos S Hofman P CHARGE Association and secondary hypoadrenalism Am J Med Genet A 2003 117A 177 180 12567418 286 Gregory LC Gevers EF Baker J Structural pituitary abnormalities associated with CHARGE syndrome J Clin Endocrinol Metab 2013 98 E737 E743 23526466 287 Baujat G Le Merrer M Ellis-van Creveld syndrome Orphanet J Rare Dis 2007 2 27 17547743 288 O’Connor MJ Rider NL Thomas Collins R Contemporary management of congenital malformations of the heart in infants with Ellis-van Creveld syndrome: a report of nine cases Cardiol Young 2011 21 145 152 21070693 289 Hills CB Kochilas L Schimmenti LA Moller JH Ellis-van Creveld syndrome and congenital heart defects: presentation of an additional 32 cases Pediatr Cardiol 2011 32 977 982 21533779 290 O’Connor MJ Tang X Collins RT Cardiac diagnoses, procedures, and healthcare utilisation in inpatients with Ellis-van Creveld syndrome Cardiol Young 2015 25 95 101 24168757 291 Solomon BD VACTERL/VATER association Orphanet J Rare Dis 2011 6 56 21846383 292 Ahn SY Mendoza S Kaplan G Reznik V Chronic kidney disease in the VACTERL association: clinical course and outcome Pediatr Nephrol 2009 24 1047 1053 19172300 293 Cunningham BK Hadley DW Hannoush H Analysis of cardiac anomalies in VACTERL association Birth Defects Res A Clin Mol Teratol 2013 97 792 797 24343877 294 Rao RP Drolet BA Holland KE Frommelt PC PHACES association: a vasculocutaneous syndrome Pediatr Cardiol 2008 29 793 799 18427881 295 Hess CP Fullerton HJ Metry DW Cervical and intracranial arterial anomalies in 70 patients with PHACE syndrome AJNR Am J Neuroradiol 2010 31 1980 1986 20705698 296 Heyer GL Dowling MM Licht DJ The cerebral vasculopathy of PHACES syndrome Stroke 2008 39 308 316 18174492 297 Bronzetti G Giardini A Patrizi A Ipsilateral hemangioma and aortic arch anomalies in posterior fossa malformations, hemangiomas, arterial anomalies, coarctation of the aorta, and cardiac defects and eye abnormalities (PHACE) anomaly: report and review Pediatrics 2004 113 412 415 14754961 298 Bayer ML Frommelt PC Blei F Congenital cardiac, aortic arch, and vascular bed anomalies in PHACE syndrome (from the International PHACE Syndrome Registry) Am J Cardiol 2013 112 1948 1952 24079520 299 Giardini A Gholam C Khambadkone S Kostolny M Need for comprehensive vascular assessment before surgical repair of aortic coarctation in PHACES syndrome Pediatr Cardiol 2010 31 291 293 19936584 300 Hartemink DA Chiu YE Drolet BA Kerschner JE PHACES syndrome: a review Int J Pediatr Otorhinolaryngol 2009 73 181 187 19101041 301 Metry DW Haggstrom AN Drolet BA A prospective study of PHACE syndrome in infantile hemangiomas: demographic features, clinical findings, and complications Am J Med Genet A 2006 140 975 986 16575892 302 Burrows PE Robertson RL Mulliken JB Cerebral vasculopathy and neurologic sequelae in infants with cervicofacial hemangioma: report of eight patients Radiology 1998 207 601 607 9609880 303 Cerruti Mainardi P Cri du chat syndrome Orphanet J Rare Dis 2006 1 33 16953888 304 Yamashita M Tanioka F Taniguchi K Matsuki A Oyama T Anesthetic considerations in Cri du chat syndrome: a report of three cases Anesthesiology 1985 63 201 202 4025870 305 Hills C Moller JH Finkelstein M Lohr J Schimmenti L Cri du chat syndrome and congenital heart disease: a review of previously reported cases and presentation of an additional 21 cases from the Pediatric Cardiac Care Consortium Pediatrics 2006 117 e924 e927 16585274 306 Grossfeld PD Mattina T Lai Z The 11q terminal deletion disorder: a prospective study of 110 cases Am J Med Genet A 2004 129A 51 61 15266616 307 Mattina T Perrotta CS Grossfeld P Jacobsen syndrome Orphanet J Rare Dis 2009 4 9 19267933 308 Favier R Jondeau K Boutard P Paris-Trousseau syndrome: clinical, hematological, molecular data of ten new cases Thromb Haemost 2003 90 893 897 14597985 309 Blaine Easley R Sanders D McElrath-Schwartz J Martin J Mark Redmond J Anesthetic implications of Jacobsen syndrome Paediatr Anaesth 2006 16 66 71 16409533 310 Pivnick EK Velagaleti GV Wilroy RS Jacobsen syndrome: report of a patient with severe eye anomalies, growth hormone deficiency, and hypothyroidism associated with deletion 11 (q23q25) and review of 52 cases J Med Genet 1996 33 772 778 8880580 311 Ng SB Bigham AW Buckingham KJ Exome sequencing identifies MLL2 mutations as a cause of Kabuki syndrome Nat Genet 2010 42 790 793 20711175 312 Niikawa N Kuroki Y Kajii T Kabuki make-up (Niikawa-Kuroki) syndrome: a study of 62 patients Am J Med Genet 1988 31 565 589 3067577 313 Schrander-Stumpel CT Spruyt L Curfs LM Defloor T Schrander JJ Kabuki syndrome: clinical data in 20 patients, literature review, and further guidelines for preventive management Am J Med Genet A 2005 132A 234 243 15690368 314 Digilio MC Marino B Toscano A Giannotti A Dallapiccola B Congenital heart defects in Kabuki syndrome Am J Med Genet 2001 100 269 274 11343317 315 Bogershausen N Wollnik B Unmasking Kabuki syndrome Clin Genet 2013 83 201 211 23131014 316 Armstrong L Abd El Moneim A Aleck K Further delineation of Kabuki syndrome in 48 well-defined new individuals Am J Med Genet A 2005 132A 265 272 15690370 317 Hoffman JD Ciprero KL Sullivan KE Immune abnormalities are a frequent manifestation of Kabuki syndrome Am J Med Genet A 2005 135 278 281 15887282 318 Greenberg F Guzzetta V Montes de Oca-Luna R Molecular analysis of the Smith-Magenis syndrome: a possible contiguousgene syndrome associated with del(17)(p11.2) Am J Hum Genet 1991 49 1207 1218 1746552 319 Elsea SH Girirajan S Smith-Magenis syndrome Eur J Hum Genet 2008 16 412 421 18231123 320 Edelman EA Girirajan S Finucane B Gender, genotype, and phenotype differences in Smith-Magenis syndrome: a meta-analysis of 105 cases Clin Genet 2007 71 540 550 17539903 321 Greenberg F Lewis RA Potocki L Multi-disciplinary clinical study of Smith-Magenis syndrome (deletion 17p11.2) Am J Med Genet 1996 62 247 254 8882782 322 Myers SM Challman TD Congenital heart defects associated with Smith-Magenis syndrome: two cases of total anomalous pulmonary venous return Am J Med Genet A 2004 131 99 100 15384100 323 Smith AC Gropman AL Bailey-Wilson JE Hypercholesterolemia in children with Smith-Magenis syndrome: del (17) (p11.2p11.2) Genet Med 2002 4 118 125 12180145 324 Chaudhry AP Schwartz C Singh AK Stroke after cardiac surgery in a patient with Smith-Magenis syndrome Tex Heart Inst J 2007 34 247 249 17622381 325 Goldman AM Potocki L Walz K Epilepsy and chromosomal rearrangements in Smith-Magenis Syndrome [del(17) (p11.2p11.2)] J Child Neurol 2006 21 93 98 16566870 326 Myers SM Challman TD Bock GH End-stage renal failure in Smith-Magenis syndrome Am J Med Genet A 2007 143A 1922 1924 17603799 327 Maas NM Van Buggenhout G Hannes F Genotype-phenotype correlation in 21 patients with Wolf-Hirschhorn syndrome using high resolution array comparative genome hybridisation (CGH) J Med Genet 2008 45 71 80 17873117 328 Battaglia A Filippi T Carey JC Update on the clinical features and natural history of Wolf-Hirschhorn (4p-) syndrome: experience with 87 patients and recommendations for routine health supervision Am J Med Genet C Semin Med Genet 2008 148C 246 251 18932224 329 Battaglia A Filippi T South ST Carey JC Spectrum of epilepsy and electroencephalogram patterns in Wolf-Hirschhorn syndrome: experience with 87 patients Dev Med Child Neurol 2009 51 373 380 19379291 330 von Elten K Sawyer T Lentz-Kapua S Kanis A Studer M A case of Wolf-Hirschhorn syndrome and hypoplastic left heart syndrome Pediatr Cardiol 2013 34 1244 1246 22639003 331 Tautz J Veenma D Eussen B Congenital diaphragmatic hernia and a complex heart defect in association with Wolf-Hirschhorn syndrome Am J Med Genet A 2010 152A 2891 2894 20830802 332 Hanley-Lopez J Estabrooks LL Stiehm R Antibody deficiency in Wolf-Hirschhorn syndrome J Pediatr 1998 133 141 143 9672528 333 Liu J Krantz ID Cornelia de Lange syndrome, cohesin, and beyond Clin Genet 2009 76 303 314 19793304 334 Jackson L Kline AD Barr MA Koch S De Lange syndrome: a clinical review of 310 individuals Am J Med Genet 1993 47 940 946 8291537 335 Kline AD Stanley C Belevich J Developmental data on individuals with the Brachmann-de Lange syndrome Am J Med Genet 1993 47 1053 1058 7507292 336 Selicorni A Colli AM Passarini A Analysis of congenital heart defects in 87 consecutive patients with Brachmann-de Lange syndrome Am J Med Genet A 2009 149A 1268 1272 19449412 337 Chatfield KC Schrier SA Li J Congenital heart disease in Cornelia de Lange syndrome: phenotype and genotype analysis Am J Med Genet A 2012 158A 2499 2505 22965847 338 August DA Sorhabi S Is a difficult airway predictable in Cornelia de Lange syndrome? Paediatr Anaesth 2009 19 707 709 19638123 339 Jyonouchi S Orange J Sullivan KE Krantz I Deardorff M Immunologic features of Cornelia de Lange syndrome Pediatrics 2013 132 e484 e489 23821697 340 Lambert MP Jackson LG Clark D The incidence of thrombocytopenia in children with Cornelia de Lange syndrome Am J Med Genet A 2011 155A 33 37 21204208 341 Selicorni A Sforzini C Milani D Anomalies of the kidney and urinary tract are common in de Lange syndrome Am J Med Genet A 2005 132 395 397 342 Verrotti A Agostinelli S Prezioso G Epilepsy in patients with Cornelia de Lange syndrome: a clinical series Seizure 2013 22 356 359 23473710 343 Bossert T Walther T Gummert J Cardiac malformations associated with the Holt-Oram syndrome – report on a family and review of the literature Thorac Cardiovasc Surg 2002 50 312 314 12375192 344 Newbury-Ecob RA Leanage R Raeburn JA Young ID Holt-Oram syndrome: a clinical genetic study J Med Genet 1996 33 300 307 8730285 345 Digilio MC Calzolari F Capolino R Congenital heart defects in patients with oculo-auriculo-vertebral spectrum (Goldenhar syndrome) Am J Med Genet A 2008 146A 1815 1819 18553555 346 Digilio MC McDonald-McGinn DM Heike C Three patients with oculo-auriculo-vertebral spectrum and microdeletion 22q11.2 Am J Med Genet A 2009 149A 2860 2864 19890921 347 Rollnick BR Kaye CI Nagatoshi K Hauck W Martin AO Oculoauriculovertebral dysplasia and variants: phenotypic characteristics of 294 patients Am J Med Genet 1987 26 361 375 3812588 348 Ozlu O Simsek S Alacakir H Yigitkanli K Goldenhar syndrome and intubation with the fiberoptic broncoscope Paediatr Anaesth 2008 18 793 794 18613937 349 Ritchey ML Norbeck J Huang C Keating MA Bloom DA Urologic manifestations of Goldenhar syndrome Urology 1994 43 88 91 8284889 350 Lin AE Salbert BA Belmont J Smoot L Total is more than the sum of the parts: phenotyping the heart in cardiovascular genetics clinics Am J Med Genet A 2004 131 111 114 15487010 351 Jacobs ML Jacobs JP Franklin RC Databases for assessing the outcomes of the treatment of patients with congenital and paediatric cardiac disease – the perspective of cardiac surgery Cardiol Young 2008 18 Suppl 2 101 115 19063780 352 Gelb B Brueckner M Chung W The Congenital Heart Disease Genetic Network Study: rationale, design, and early results Circ Res 2013 112 698 706 23410879 353 Lin AE Basson CT Goldmuntz E Adults with genetic syndromes and cardiovascular abnormalities: clinical history and management Genet Med 2008 10 469 494 18580689
PMC005xxxxxx/PMC5125078.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 7505689 5980 Mycopathologia Mycopathologia Mycopathologia 0301-486X 1573-0832 22076410 5125078 10.1007/s11046-011-9494-7 NIHMS504519 Article Pseudomonas aeruginosa Inhibits the Growth of Cryptococcus Species Rella Antonella Department of Biochemistry and Molecular Biology, Medical University of South Carolina, 173 Ashley Avenue, BSB 535E, Charleston, SC 29425, USA Department of Biomedical Science and Human Oncology, Hygiene Section, University of Bari, Bari, Italy Yang Mo Wei Department of Biochemistry and Molecular Biology, Medical University of South Carolina, 173 Ashley Avenue, BSB 535E, Charleston, SC 29425, USA Gruber Jordon Department of Biochemistry and Molecular Biology, Medical University of South Carolina, 173 Ashley Avenue, BSB 535E, Charleston, SC 29425, USA Montagna Maria Teresa Department of Biomedical Science and Human Oncology, Hygiene Section, University of Bari, Bari, Italy Luberto Chiara Department of Biochemistry and Molecular Biology, Medical University of South Carolina, 173 Ashley Avenue, BSB 535E, Charleston, SC 29425, USA Zhang Yong-Mei [email protected] Department of Biochemistry and Molecular Biology, Medical University of South Carolina, 173 Ashley Avenue, BSB 535E, Charleston, SC 29425, USA Department of Microbiology and Immunology, Medical University of South Carolina, Charleston, SC, USA Del Poeta Maurizio [email protected] Department of Biochemistry and Molecular Biology, Medical University of South Carolina, 173 Ashley Avenue, BSB 535E, Charleston, SC 29425, USA Department of Microbiology and Immunology, Medical University of South Carolina, Charleston, SC, USA Department of Craniofacial Biology, Medical University of South Carolina, Charleston, SC, USA Division of Infectious Diseases, Medical University of South Carolina, Charleston, SC, USA 20 11 2016 11 11 2011 6 2012 28 11 2016 173 5-6 451461 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Pseudomonas aeruginosa is a ubiquitous and opportunistic bacterium that inhibits the growth of different microorganisms, including Gram-positive bacteria and fungi such as Candida spp. and Aspergillus fumigatus. In this study, we investigated the interaction between P. aeruginosa and Cryptococcus spp. We found that P. aeruginosa PA14 and, to a lesser extent, PAO1 significantly inhibited the growth of Cryptococcus spp. The inhibition of growth was observed on solid medium by the visualization of a zone of inhibition of yeast growth and in liquid culture by viable cell counting. Interestingly, such inhibition was only observed when P. aeruginosa and Cryptococcus were co-cultured. Minimal inhibition was observed when cell–cell contact was prevented using a separation membrane, suggesting that cell contact is required for inhibition. Using mutant strains of Pseudomonas quinoline signaling, we showed that P. aeruginosa inhibited the growth of Cryptococcus spp. by producing antifungal molecules pyocyanin, a redox-active phenazine, and 2-heptyl-3,4-dihydroxyquinoline (PQS), an extracellular quorum-sensing signal. Because both P. aeruginosa and Cryptococcus neoformans are commonly found in lung infections of immunocompromised patients, this study may have important implication for the interaction of these microbes in both an ecological and a clinical point of view. Cryptococcus spp. Pseudomonas aeruginosa Quorum sensing 2-heptyl-3,4-dihydroxyquinoline Pyocyanin Introduction Cryptococcus neoformans is an important, widely distributed, fungal human pathogen that causes a life-threatening meningoencephalitis in immunocompromised hosts. In recent years, it has been shown to cause serious pulmonary infection also in individuals with competent immune system [1, 2]. The cryptococcal infection develops after exposure to fungal basidiospores or encapsulated yeast found ubiquitously in the environment, such as pigeon guano and eucalyptus trees. The port of entry of Cryptococcus species is the lung. When spores or yeasts are inhaled, the infection can be restricted to the lung or can disseminate to other tissues. Following inhalation, alveolar macrophages are the first line of defense against Cryptococcus, contributing to the clearance or containment of the fungal pathogen in the granuloma [3]. This occurs normally in immunocompetent subjects. In immunocompromised subjects, the yeast cells can disseminate from the lung to the brain within host cells (intracellularly) or in the bloodstream (extracellularly), leading to the development of life-threatening disease [4–6]. Thus, the lung is an important organ for the containment of cryptococcal infection. The respiratory tract is also a common organ infected by Pseudomonas aeruginosa (Pa), a ubiquitous Gram-negative bacterium commonly found in soil and water, following aspiration of the organism from the upper respiratory tract, especially in patients on mechanical ventilation. Alternatively, pneumonia may occur as a result of bacteremia spread to the lungs. This is observed commonly in patients following chemotherapy-induced neutropenia or in patients with AIDS [7, 8]. In Pa and in other bacteria, modulation and coordination of gene expression are influenced by population density via the production of small molecules that impact the production of virulence factors. This mechanism based on cell–cell communication is known as quorum sensing (QS). QS molecules are capable of not only controlling gene expression in Pa but also affecting host immune response and the growth of other microorganisms [9]. Several groups have demonstrated the capacity of Pa to inhibit yeast growth in vitro, namely Candida spp., Aspergillus fumigatus and Saccharomyces cerevisiae [10, 11]. Interest in the inhibitory effect of several strains of Pa against C. neoformans in vitro started in 1954, when Fisher et al. reported that strains of Pa produced a substance that inhibits the growth of C. neoformans in vitro, but the inhibitor has not been successfully separated [12]. Twenty years later, Teoh-Chan et al. reported an initial attempt to isolate and characterize the inhibitory substances involved in 1974 [13]. However, the literature about the inhibitory effect of Pa on C. neoformans is equivocal [12, 13]. In this study, we revealed the mechanism used by Pa to inhibit the growth of Cryptococcus spp. through the production of QS molecules. We found that the Pa PQS and, more profoundly, pyocyanin do significantly inhibit cryptococcal growth. Interestingly, this inhibition was mainly detected when C. neoformans was co-cultured with Pa, suggesting that physical contact triggers the production of antifungal molecules in Pa. Materials and Methods Experimental Strains Cryptococcus neoformans var. grubii serotype A strain H99, C. neoformans var. neoformans serotype D strain JEC21, C. gattii serotype B (strain MMRL 1336) and serotype C (strain MMRL 1343) [the latter two strains were a kind gift from Wiley Schell, Duke University Medical Center, Durham, NC, USA), and P. aeruginosa wild-type strains PAO1 and PA14 were used in this study. PAO1 and PA14 are the two commonly used laboratory strains of Pa. Numerous reports showed that the clinical isolate PA14 is significantly more virulent than PAO1 in a wide range of hosts [14, 15]. P. aeruginosa PAO1 mutant strains pqsAB, pqsB and pqsE and E. coli strain DH5a were used in this study. Cryptococcus species were grown at 30°C in YPD broth (2% peptone, 1% yeast extract, 2% dextrose, BD) with shaking at 250 rpm. PAO1 and PA14 were cultured in two different media LB and YPD. Pa showed similar profile of growth in both media. PAO1 mutant strains were grown in LB (Luria–Bertani) at 37°C with shaking at 250 rpm. YNB Plate Assay C. neoformans var. grubii serotype A, C. neoformans var. neoformans serotype D and C. gattii serotype B and serotype C were grown overnight, and the cells were washed twice with sterile water and resuspended in YNB broth (Yeast Nitrogen Broth, Sigma-Aldrich). 300 µl of 1 × 108 cells/ml were spread on YNB agar plates using a glass spreader. Then, sterile filter paper disks were placed on the plates. Both PAO1 or PA14 were grown overnight at 37°C in LB media, and 10-µl drops of an 8 × 109 CFU/ml Pa cultures were spotted on the filter disks, following by incubation at 30°C for up to 72 h. The same procedure was used for PAO1 mutant strains: pqsAB, pqsB and pqsE. Survival Assay of C. neoformans H99 Co-cultured with Pa Strains C. neoformans and Pa strains PAO1 or PA14 were grown at 30°C in YPD broth overnight. Different C. neoformans dilutions were performed (107, 106, 105, 104 and 103 cells/ml) and inoculated with 3 × 109 CFU/ml Pa strains onto 96-well plate. The plate was incubated at 30°C for 24 h. C. neoformans and Pa cells viability was determined by CFU counting on YNB and LB plates, respectively. YNB and LB plates were used to count C. neoformans and Pa cells in mixed culture conditions, respectively. On YNB plates, C. neoformans was growing faster than Pa, allowing us to count C. neoformans, whereas LB plates do not support the growth of C. neoformans and were chosen to count Pa. C. neoformans H99-PAO1 or PA14 Interaction Both PAO1 and PA14 were grown in 100 ml of YPD at 30°C for 24 h. C. neoformans was grown in 50 ml of YPD at 30°C for 24 h. Pa and C. neoformans cultures were washed twice with sterile water and harvested at 8,000×g for 20 min and at 3,000 rpm for 10 min, respectively. Pa cultures were split into two flasks: Pa was cultured alone in one flask, and Pa was co-cultured with C. neoformans in the other flask, for 48 h at 30°C. Pa alone and the Pa–C. neoformans co-culture were pelleted by centrifugation at 8,000×g for 20 min. The supernatants were filtered with vacuum-driven disposable filtration system with 0.22-µm pore size (Millipore), frozen and lyophilized using a bench top freeze dryer. 500 mg of dried co-culture supernatants (1.26 × 1015 CFU Pa strains and 1 × 109 CFU C. neoformans) and Pa supernatants (1 × 1015 CFU Pa strains) was resuspended in 1 ml of YPD broth. Cryptococcus neoformans was grown overnight and 300 µL of 1 × 108 cells/ml culture was spread on YPD agar plates using a glass spreader. Different amounts of dried co-culture supernatants (e.g., 12.5, 25, 50, 75 and 100 mg) and dried Pa supernatants (100 mg) resuspended in 200 µl of YPD were added into 8-mm wells made on the YPD plates. The plates were incubated at 30°C for up to 72 h. We used 100 mg of dried YPD broth resuspended in 200 µl of YPD as a control. U-Tube Assay A sterile membrane with 0.22-µm pore size was used to separate the two sides of the U-tube. In one side of the tube, 25 ml of YPD with 1.5 × 109 CFU/ml of either Pa strains was cultured, whereas in the other side, 25 ml of YPD with 1 × 105 cells/ml of C. neoformans was inoculated. Separately, Erlenmeyer flasks were used for controls, including co-cultures of C. neoformans with PAO1 or PA14, C. neoformans with E. coli and cultures of C. neoformans alone and PAO1 or PA14 alone, maintaining the same number of C. neoformans and Pa used in the U-tube. The U-tube and the flasks were incubated for 48 h at 30°C. C. neoformans and Pa cell viability was determined by CFU counting on YNB and LB plates, respectively. Fungal Viability Assay in Physiological Condition Overnight cultures of C. neoformans in YPD broth at 30°C and Pa strains in LB broth at 37°C were pelleted, washed with sterile water and resuspended in DMEM buffered with 50 mM HEPES (pH 7.2). 1.5 × 109 CFU/ml of PAO1 or PA14 and 2 × 104 cells/ml of C. neoformans were co-cultured at 37°C with 5% CO2 for 24 and 48 h. C. neoformans and Pa cells viability was determined by CFU counting. Mutant Generation The pqs mutant strains were created in strain PAO1 by gene replacement described previously [16]. Briefly, a DNA fragment, which contained a gentamicin resistance cassette flanked by the 5′ and 3′ ends of the target gene, was cloned into pEX18ApGW [16]. The resultant plasmid was conjugated from E. coli strain SM10 into the Pa wild-type strain PAO1 with the selection of gentamicin (30 µg/ml) on Pseudomonas isolation agar plate (BD, Becton, Dickinson and Company). Merodiploids formed via a single crossover event was resolved through 5% sucrose selection in the presence of gentamicin. The gentamicin resistance cassette was subsequently removed by Flp recombinase, resulting in an unmarked deletion mutant of the target gene. The presence of the deletion in the correct region was verified by PCR and sequencing. Pyocyanin Assay For pyocyanin quantification, cultures of PAO1 or pqs mutants were grown overnight in LB at 37°C with shaking (250 rpm) to reach OD600 of 2.0. Pyocyanin was extracted with chloroform from cell-free supernatants, acidified by 0.2 N HCl and assayed spectrophotometrically at 520 nm as previously described [17]. Concentrations, expressed in mg/l, were determined by multiplying the optical density at 520 nm by 17.072 [17]. Microtiter Assay C. neoformans was grown overnight, washed and resuspended in YNB buffered with 25 mM HEPES (pH 7.2). 2 × 104 cells/ml of C. neoformans were incubated in a 96-well plate containing 4-hydroxy-2-heptylquinoline (HHQ) [Qingdao Vochem], 3,4-dihydroxy-2-heptylquinoline (PQS) [Sigma-Aldrich], 2,4-dihydroxyquinoline (DHQ) [Fluka] and pyocyanin [Cayman Chemical] at different concentrations (1.5, 3.5, 7, 14, 28, 56, 112 and 225 µg/ml). The compounds were dissolved in DMSO. The final concentration of DMSO was less than 1–1.5%. The plate was incubated for 48 h at 30°C. The growth of yeast cells was assessed by monitoring optical density at 495 nm (OD495), and the surviving viable yeast cells were determined by CFU counting. Statistics Data from each experimental group were subjected to an analysis of normality and variance. Statistical significance between the means of different experimental data sets composed of normally distributed values was analyzed using two-tailed Student’s t test. For all statistical tests, standard deviation with P-values less than 0.05 was considered significant. All experiments were done twice, with similar results each time, or three times when statistics was applied. Results Pa Inhibited Growth of Cryptococcus spp Previous reports reveal that Pa inhibits the growth of different microbes [10–13]. However, the mechanism that underlines the inhibitory effects of Pa on Cryptococcus species has not been investigated. Our results showed that when PAO1 or PA14 was inoculated on paper disks on YNB plate, which contained a lawn of Cryptococcus species, a zone of inhibition developed around the disks within 24 h. The diameters of the zone of inhibition of PA14 against serotypes, A, B, C and D of Cryptococcus were 23, 25, 29 and 30 mm, respectively. In contrast, reduced diameters of the zone of inhibition of PAO1 against Cryptococcus species were observed (16, 12, 23 and 25 mm) (Fig. 1). Pa Fungicidal Effect Against C. neoformans in Co-culture To determine whether the inhibitory effects of Pa on C. neoformans growth were fungistatic or fungicidal, we performed a viability assay of C. neoformans in co-cultures with PAO1 or PA14. Different C. neoformans dilutions were mixed with Pa strains (Fig. 2). After co-incubation of PAO1 and C. neoformans, we observed, in the first 4 dilutions, 1 log of decrease in C. neoformans cell number compared with the start culture. The last dilution showed no remaining viable cells of C. neoformans (Fig. 2a). In contrast, in the first 3 dilutions of co-cultures of PA14 and C. neoformans, we observed 1 log of decrease in C. neoformans cells number, but there was no remaining viable cells of C. neoformans in the last two dilutions (Fig. 2c). These results showed that the inhibitory effect of Pa on C. neoformans is fungicidal and the killing effect is dependent on the relative cell density. To evaluate whether Pa growth is affected in co-culture with C. neoformans, the number of Pa cells was counted. The results showed that C. neoformans does not affect Pa growth (Fig. 2b, d). Both PAO1 and PA14 grew in the condition tested, and therefore, the different inhibitory effect is not the result of different growth rate of PAO1 and PA14. Enhanced Inhibition of C. neoformans by Co-culture Supernatants To understand the mechanism by which Pa inhibits fungal growth, we screened the effect of culture supernatants of PAO1 or PA14 and the effect of Pa–C. neoformans co-culture supernatants on C. neoformans growth. Supernatants were prepared as outlined in the Materials and Methods. We inoculated different amounts of dried co-cultured supernatants (e.g., 12.5, 25, 50, 75 and 100 mg) and dried Pa supernatants (e.g., 100 mg), on YPD plate, which contained a lawn of C. neoformans. As a control, we used dried YPD, to exclude any effects from the high concentration of salts. A diameter of zone of inhibition of 15, 16, 17 and 20 mm was observed with 25, 50, 75 and 100 mg of PAO1-C. neoformans supernatant, respectively (Fig. 3a), whereas we observed a diameter of zone of inhibition of 15, 18, 21 and 22 mm with PA14-C. neoformans supernatant (Fig. 3c). The dried PAO1 supernatant did not result in a clean zone of inhibition, whereas with dried PA14 supernatant, we estimated a zone of inhibition of 17 mm (Fig. 3b, d). These results suggest that the physical contact between Pa and C. neoformans triggers the production of antifungal molecules by Pa that inhibit C. neoformans growth. Pa supernatant alone was not able to inhibit C. neoformans growth as much as C. neoformans–Pa supernatant. The Cell–Cell Contact was Important for Pa to Inhibit C. neoformans To investigate whether Pa–C. neoformans cells contact was important for inhibition, we performed an assay in YPD media using a U-tube. We cultured Pa in one side and C. neoformans in the other side of the tube; the two sides were separated with a sterile membrane, which prevented mixing of Pa and C. neoformans but allowed free exchange of nutrients and extracellular molecules. The C. neoformans growth exhibited half a log reduction when cultured in the U-tube with PAO1 (Fig. 4a), whereas 1 log reduction was observed when C. neoformans was cultured with PA14 (Fig. 4c). A 4 log inhibition was observed when C. neoformans was co-cultured with Pa strains in Erlenmeyer flasks (Fig. 4a, c). These results suggest that the cell contact is important to elicit a strong inhibition of C. neoformans by Pa. When E. coli was co-cultured with C. neoformans, no inhibitory effects were observed (Fig. 4a, c). Under these conditions, Pa growth was not affected by C. neoformans like observed before (Fig. 4b, d). Pa Inhibited C. neoformans in Physiological Condition To determine whether Pa inhibits the growth of C. neoformans in physiological condition, we co-cultured PAO1 or PA14 with C. neoformans in DMEM (pH 7.2) at 37°C with 5% CO2 for 24 and 48 h. Both PAO1 and PA14 inhibited the growth of C. neoformans in DMEM; however, the degree of inhibition in DMEM (Fig. 5a, c) was significantly reduced compared to co-cultures in YPD (Fig. 4a, c). In DMEM, the growth of C. neoformans exhibited a 2 log reduction when co-cultured with PAO1 for 48 h (Fig. 5a), whereas a 4 log reduction was observed when C. neoformans was co-cultured with PAO1 in YPD (Fig. 4a). Similar results were observed with the PA14 strain (Figs. 4c, 5c). The reduced inhibition of C. neoformans growth by Pa could be the result of reduced replication of Pa in DMEM. Pa cell number counting showed that there was no significant increase in the number of Pa when co-cultured with C. neoformans in DMEM (Fig. 5b, d), suggesting that DMEM does not support the growth of Pa as well as YPD, in which Pa cell number increased by 1 log when co-cultured with C. neoformans (Fig. 4b, d). These results suggest that the inhibition of C. neoformans growth depends on the cell density of Pa and relative ratio of Pa and C. neoformans. Effects of PQS Mutants of PAO1 on the Growth of C. neoformans It has been shown that extracellular compounds secreted by Pa, including quorum-sensing molecules and phenazines (such as pyocyanin), affect the growth of different microbes [9, 11]. Pseudomonas quinolone signaling plays an essential role in activating the production of pyocyanin [18]. Proteins encoded by the pqsABCDE operon are required for both PQS synthesis and activation of pyocyanin production. Specifically, PqsA, PqsB, PqsC and PqsD are required for the formation of extracellular quinolones including 4-hydroxy-2-heptylquinoline (HHQ), 3,4-dihydroxy-2-heptylquinoline (PQS) and 2,4-dihydroxyquinoline (DHQ). PqsE is not required for the biosynthesis of quinolones but is essential for the activation of pyocyanin production. To determine whether the extracellular quinolones and pyocyanin are involved in the inhibition of C. neoformans, we tested the effects of pqs mutants of PAO1 on the growth of C. neoformans. Three mutant strains of PAO1 (pqsAB, pqsB and pqsE) were chosen to distinguish the effects of alkylquinolones (including HHQ and PQS), DHQ and pyocyanin. pqsAB mutant is defective in both PqsA and PqsB and does not produce any extracellular quinolones. pqsB mutant lacks alkylquinolones [19] but produces normal level of DHQ. pqsE mutant produces extracellular quinolones at the similar levels as the wild-type PAO1. In terms of pyocyanin production, pqsAB and pqsE mutants are defective, whereas pqsB secreted pyocyanin at a level that was about 50% of the wild-type PAO1 (Fig. 6a). No inhibition on the growth of C. neoformans by the pqsAB mutant was observed (Fig. 6b). pqsB exhibited inhibition on C. neoformans growth on YNB plates (14 mm, diameter of the zone of inhibition), although the zone of inhibition of pqsB mutant was smaller than that of the wild type (17 mm) (Fig. 6b). pqsE mutant marginally inhibited C. neoformans growth with a small zone of inhibition (9 mm) outside the paper disk (Fig. 6b). These results, combined with the different compositions of extracellular compounds secreted by the mutants (Fig. 6), suggest that the inhibition of C. neoformans growth by Pa is mainly dependent on pyocyanin and, to a lesser degree, alkylquinolones such as HHQ and PQS. The effects of quinolones and pyocyanin on C. neoformans were further investigated using a microtiter dilution method in YNB with different concentrations of HHQ, PQS, DHQ and pyocyanin (Fig. 7a). The viable C. neoformans cells were assessed by monitoring optical density at 495 nm and CFU. Both pyocyanin and PQS completely inhibited C. neoformans growth at 28 µg/ml. However, pyocyanin exhibited more pronounced inhibition than PQS at lower concentrations (less than 14 µg/ml), demonstrating that pyocyanin is a more potent antifungal agent than PQS (Fig. 7b). HHQ partially inhibited the growth of C. neoformans only at the highest concentration tested. No inhibition on C. neoformans by DHQ was observed. These results corroborated the data obtained with the pqs mutants in that pyocyanin is a major contributing factor in the inhibition of C. neoformans.We also determined the number of viable C. neoformans cells by withdrawing aliquots from the microtiter plates. Viable cells were present in wells with pyocyanin or PQS (data not shown), demonstrating that the inhibitory effects on C. neoformans growth are static but not fungicidal. Discussion In the present study, we investigated the ability of Pa strains PAO1 and PA14 to inhibit the growth of Cryptococcus species. Our results demonstrated that, in mixed culture conditions, Pa exhibited a fungicidal effect on C. neoformans, dependent on the relative cell density. Physical contact between Pa and C. neoformans was important for inhibition as demonstrated by experiments using U-tubes (Fig. 4), indicating that cell–cell contacts could activate the production of antifungal molecules by Pa that impeded C. neoformans growth. Our results also showed that the major antifungal molecule produced by Pa was the redoxactive metabolite, pyocyanin. In addition, alkylquinolones, such as HHQ and PQS, exhibited antifungal proprieties against C. neoformans. These findings are consistent with earlier reports that showed Pa inhibited the growth of Gram-positive bacteria and yeasts by heat-stable molecules involving quorum-sensing mechanism [20, 21]. Our results showed that PA14 exhibited more potent inhibition of Cryptococcus growth than PAO1. The p-value of paired t test is 0.02, demonstrating a significant difference in the diameters of the zone of inhibition of PA14 and PAO1 against Cryptococcus species (Fig. 1). PAO1 and PA14 are two commonly used laboratory strains of Pa. Strain PAO1 was first isolated in 1954 from a wound [22]. Numerous reports showed that the clinical isolate PA14 is significantly more virulent than PAO1 in multiple hosts, including mice, the nematode Caenorhabditis elegans, the insect Galleria mellonella and the plant Arabidopsis thaliana [14, 15, 23], suggesting that the greater antifungal activity of PA14 may be due to its increased virulence. Although the genomes of the two Pa strains are very similar (92% of the PA14 genome is present in PAO1, and 96% of PAO1 genome is present in PA14 [24]), unique genes of PA14 have been extensively studied for their roles in virulence. However, Lee et al. showed that the presence of these genes was not correlated with PA14 virulence, and the virulence of Pa is both combinatorial and multifactorial [24]. Therefore, future studies to identify PA14 genes that are essential for the increased inhibition of C. neoformans growth may help decipher the roles of different virulence factor on the ability of Pa to cause diseases. Previous reports have shown an interspecies competition between Pa and yeasts, and in this study, we have clearly demonstrated that Pa inhibits the growth of C. neoformans, whereas Pa growth is not affected by Cryptococcus strains. Hogan et al. [25] demonstrated that Pa formed a biofilmon Candida albicans filaments and killed the fungus; where as C. albicans produced farnesol, a major fungal QS molecule, which reduced the production of PQS and pyocyanin by Pa [18, 26]. Our data showed that the interaction of Pa and C. neoformans is not reciprocal; production of extracellular virulence molecules by Pa inhibited the growth of Cryptococcus strains, while Pa was not affected by the presence of C. neoformans. Our results also demonstrated that Pa exhibited a fungicidal effect when co-cultured with C. neoformans (Fig. 2), whereas the effects of pyocyanin and PQS were static, suggesting that other unidentified exoproducts, such as proteases, hemolysin and rhamnolipids, may be involved in fungicidal activity of Pa. Moreover, cell–cell contact with C. neoformans may be important in mounting a fungicidal response of Pa on C. neoformans. The lung of patients with cystic fibrosis is infected with a large spectrum of microbial pathogens. Over time, both the types of bacterial and their individual characteristics change [27, 28]. For instance, Staphylococcus aureus and Haemophilus influenzae are common bacteria found in CF patients at an early age. Eventually, more than 80% of CF patients harbor Pa. Filamentous fungal pathogens such as A. fumigatus and C. albicans are also frequently isolated from CF patients [29]. Fungal colonization could be the result of prolonged therapy with antibiotics and steroid, in addition to the defective mucus clearance. Hughes et al. [30] showed that in CF patients with Pa infection, only 10% had positive C. albicans skin tests, compared with 30% positivity in those free of Pa, suggesting the antifungal substance produced by Pa could prevent Candida infections. To date, there has been no report on the isolation of Cryptococcus species from the CF patients. Considering that both Cryptococcus and Pa are common lung pathogens, the lack of co-colonization could be the result of the antifungal effect of Pa on the growth of C. neoformans. Both pyocyanin and PQS accumulate intensively in the lung mucus of CF patients [31, 32], and these antifungal molecules may be important in the prevention of pulmonary cryptococcosis in CF patients. We thank all members of Del Poeta, Zhang and Luberto laboratories for discussion. This work was supported in part by Grants AI56168, AI71142, AI78493 and AI87541 (to M.D.P) from the National Institute of Health, in part by RR17677 (to M. D. P. and Y-M. Z) from the Centers of Biomedical Research Excellence Program of the National Center for Research Resources, and in part by NIH C06 RR015455 from the Extramural Research Facilities Program of the National Center for Research Resources. Dr. Maurizio Del Poeta is a Burroughs Wellcome New Investigator in the Pathogenesis of Infectious Diseases. Fig. 1 Pa inhibited the growth of Cryptococcus spp. PAO1 and PA14 were inoculated on paper disks, on YNB plates with C. neoformans var. grubii serotype A (Cn serotype A H99), C. gattii serotypes B (Cg serotype B MMRL 1336), C. gattii serotypes C (Cg serotype C MMRL1343) and C. neoformans var. neoformans serotype D (Cn serotype D JEC21). Control disks containing LB medium were placed on the plates (indicated with ctrl on the plates). The diameters of the zone of inhibition were determined after 72 h at 30°C Fig. 2 Pa has fungicidal effect against C. neoformans in co-culture. Different C. neoformans (Cn) dilutions were mixed with Pa strains and incubated in 96-well plates. The viability of Cn in co-culture with PAO1 (a) and PA14 (c) was determined by CFU counting on YNB plates. The viability of PAO1 (b) and PA14 (d) in co-culture with Cn was determined by CFU counting on LB plates. The inhibitory effect of Pa on Cn is fungicidal and is dependent on the relative cell density Fig. 3 Enhanced inhibition of C. neoformans by Pa in co-culture supernatants. Dried co-culture supernatants [Sup(Cn H99 + PAO1), Sup(Cn H99 + PA14)] (a, c) and dried Pa supernatants (SupPAO1, SupPA14) (b, d), resuspended in YPD broth, were inoculated into 8-mm wells on YPD plates, which contained a lawn of C. neoformans (Cn H99). The diameters of the zone of inhibition were determined Fig. 4 Cell–cell contact is important for Pa to inhibit C. neoformans. An assay in YPD media using a U-tube was performed. A sterile membrane was used to separate the two sides of the U-tube, in which Pa strains and C. neoformans (Cn) were cultured. Separately, Erlenmeyer flasks were used for controls, including co-cultures of Cn with PAO1 or PA14, Cn with E. coli and cultures of Cn alone and PAO1 or PA14 alone. The inhibitory effect of Pa on Cn, in the U-tube and in co-culture, was determined by CFU counting (a, c). When E. coli and Cn were co-cultured, no inhibitory effects of Cn were observed (a, c). The growth of PAO1 and PA14 was not affected by Cn (b, d). P values were calculated by Student’s t test, P < 0.05; P < 0.01; *P < 0.001 Fig. 5 Pa inhibited C. neoformans in physiological condition. PAO1 and PA14 were co-cultured with C. neoformans (Cn) in DMEM for 24 and 48 h. Viable Cn and Pa were measured by CFU on YNB and LB plates, respectively. Cn inhibition was observed with both Pa strains (a, c). Pa cells number counting showed no significant increase in the number of cells in DMEM (b, d). P-values were calculated by Student’s t test, P < 0.05 Fig. 6 Effects of pqs mutants of PAO1 on the growth of C. neoformans. The pyocyanin production was measured for all mutants (a). pqsB mutant secreted pyocyanin at a level that was about 50% of the wild-type PAO1 (a). PAO1, pqsAB, pqsB and pqsE were inoculated on paper disks, on YNB plates, which contained a lawn of C. neoformans (Cn). The diameters of the zone of inhibition were determined after 72 h at 30°C (b) Fig. 7 Antifungal activity of pyocyanin and PQS against C. neoformans. Chemical structures of 3,4-dihydroxy-2-heptylquinoline (PQS), 2,4-dihydroxyquinoline (DHQ), 4-hydroxy-2-heptylquinoline (HHQ) and pyocyanin (a). C. neoformans (Cn) was incubated in microtiter plates with different concentration of PQS, DHQ, HHQ and PYO. The growth of yeast cells was assessed by monitoring optical density at 495 nm (b) References 1 Chayakulkeeree M Perfect JR Cryptococcosis Infect Dis Clin North Am 2006 20 3 507 544 v–vi 16984867 2 Fraser JA Giles SS Wenink EC Geunes-Boyer SG Wright JR Diezmann S Same-sex mating and the origin of the Vancouver Island Cryptococcus gattii outbreak Nature 2005 437 7063 1360 1364 16222245 3 Mansour MK Levitz SM Interactions of fungi with phagocytes Curr Opin Microbiol 2002 5 4 359 365 12160853 4 Feldmesser M Tucker S Casadevall A Intracellular parasitism of macrophages by Cryptococcus neoformans Trends Microbiol 2001 9 6 273 278 11390242 5 Levitz SM Nong SH Seetoo KF Harrison TS Speizer RA Simons ER Cryptococcus neoformans resides in an acidic phagolysosome of human macrophages Infect Immun 1999 67 2 885 890 9916104 6 Chretien F Lortholary O Kansau I Neuville S Gray F Dromer F Pathogenesis of cerebral Cryptococcus neoformans infection after fungemia J Infect Dis 2002 186 4 522 530 12195380 7 Maschmeyer G Braveny I Review of the incidence and prognosis of Pseudomonas aeruginosa infections in cancer patients in the 1990s Eur J Clin Microbiol Infect Dis 2000 19 12 915 925 11205628 8 Boucher HW Talbot GH Bradley JS Edwards JE Gilbert D Rice LB Bad bugs, no drugs: no ESKAPE! An update from the Infectious Diseases Society of America Clin Infect Dis 2009 48 1 1 12 19035777 9 Gibson J Sood A Hogan DA Pseudomonas aeruginosa-Candida albicans interactions: localization and fungal toxicity of a phenazine derivative Appl Environ Microbiol 2009 75 2 504 513 19011064 10 Kerr J Inhibition of fungal growth by Pseudomonas aeruginosa and Pseudomonas cepacia isolated from patients with cystic fibrosis J Infect 1994 28 3 305 310 7522262 11 Kerr JR Taylor GW Rutman A Hoiby N Cole PJ Wilson R Pseudomonas aeruginosa pyocyanin and 1-hydroxyphenazine inhibit fungal growth J Clin Pathol 1999 52 5 385 387 10560362 12 Fisher AM Inhibition of growth of Cryptococcus neoformans by cultures of Pseudomonas aeruginosa Bull Johns Hopkins Hosp 1954 95 4 157 161 13199443 13 Teoh-Chan H Chau PY Ng MH Wong PC Inhibition of Cryptococcus neoformans by Pseudomonas aeruginosa J Med Microbiol 1975 8 1 77 81 805243 14 Choi JY Sifri CD Goumnerov BC Rahme LG Ausubel FM Calderwood SB Identification of virulence genes in a pathogenic strain of Pseudomonas aeruginosa by representational difference analysis J Bacteriol 2002 184 4 952 961 11807055 15 Tan MW Mahajan-Miklos S Ausubel FM Killing of Caenorhabditis elegans by Pseudomonas aeruginosa used to model mammalian bacterial pathogenesis Proc Natl Acad Sci USA 1999 96 2 715 720 9892699 16 Choi KH Schweizer HP An improved method for rapid generation of unmarked Pseudomonas aeruginosa deletion mutants BMC Microbiol 2005 5 30 15907219 17 Tully JG Jr Latteri A Paraplegia, syringomyelia tarda and neuropathic arthrosis of the shoulder: a triad Clin Orthop Relat Res 1978 134 244 248 18 Heeb S Fletcher MP Chhabra SR Diggle SP Williams P Camara M Quinolones: from antibiotics to autoinducers FEMS Microbiol Rev 2011 35 2 247 274 20738404 19 Zhang YM Frank MW Zhu K Mayasundari A Rock CO PqsD is responsible for the synthesis of 2, 4-dihydroxyquinoline, an extracellular metabolite produced by Pseudomonas aeruginosa J Biol Chem 2008 283 43 28788 28794 18728009 20 Machan ZA Taylor GW Pitt TL Cole PJ Wilson R 2-Heptyl-4-hydroxyquinoline N-oxide, an antistaphylococcal agent produced by Pseudomonas aeruginosa J Antimicrob Chemother 1992 30 5 615 623 1493979 21 Mowat E Rajendran R Williams C McCulloch E Jones B Lang S Pseudomonas aeruginosa and their small diffusible extracellular molecules inhibit Aspergillus fumigatus biofilm formation FEMS Microbiol Lett 2010 313 2 96 102 20964704 22 Holloway BW Genetic recombination in Pseudomonas aeruginosa J Gen Microbiol 1955 13 3 572 581 13278508 23 Rahme LG Stevens EJ Wolfort SF Shao J Tompkins RG Ausubel FM Common virulence factors for bacterial pathogenicity in plants and animals Science 1995 268 5219 1899 1902 7604262 24 Lee DG Urbach JM Wu G Liberati NT Feinbaum RL Miyata S Genomic analysis reveals that Pseudomonas aeruginosa virulence is combinatorial Genome Biol 2006 7 10 R90 17038190 25 Hogan DA Kolter R Pseudomonas-Candida interactions: an ecological role for virulence factors Science 2002 296 5576 2229 2232 12077418 26 Singh A Del Poeta M Lipid signalling in pathogenic fungi Cell Microbiol 2011 13 2 177 185 21091925 27 Oliver A Mutators in cystic fibrosis chronic lung infection: prevalence, mechanisms, and consequences for antimicrobial therapy Int J Med Microbiol 2010 300 8 563 572 20837399 28 Hoboth C Hoffmann R Eichner A Henke C Schmoldt S Imhof A Dynamics of adaptive microevolution of hypermutable Pseudomonas aeruginosa during chronic pulmonary infection in patients with cystic fibrosis J Infect Dis 2009 200 1 118 130 19459782 29 Bakare N Rickerts V Bargon J Just-Nubling G Prevalence of Aspergillus fumigatus and other fungal species in the sputum of adult patients with cystic fibrosis Mycoses 2003 46 1–2 19 23 30 Hughes WT Kim HK Mycoflora in cystic fibrosis: some ecologic aspects of Pseudomonas aeruginosa and Candida albicans Mycopathol Mycol Appl 1973 50 3 261 269 4199669 31 Taylor GW Machan ZA Mehmet S Cole PJ Wilson R Rapid identification of 4-hydroxy-2-alkylquinolines produced by Pseudomonas aeruginosa using gas chromatography-electron-capture mass spectrometry J Chromatogr B Biomed Appl 1995 664 2 458 462 7780603 32 Caldwell CC Chen Y Goetzmann HS Hao Y Borchers MT Hassett DJ Pseudomonas aeruginosa exotoxin pyocyanin causes cystic fibrosis airway pathogenesis Am J Pathol 2009 175 6 2473 2488 19893030
PMC005xxxxxx/PMC5125253.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 9010000 8514 Curr Opin Lipidol Curr. Opin. Lipidol. Current opinion in lipidology 0957-9672 1473-6535 27438680 5125253 10.1097/MOL.0000000000000334 NIHMS831029 Article Intestinal phospholipid and lysophospholipid metabolism in cardiometabolic disease Hui David Y. Department of Pathology, Metabolic Disease Research Center, University of Cincinnati College of Medicine, Cincinnati, Ohio 45237, USA Correspondence to: David Y. Hui, Department of Pathology and Laboratory Medicine, Metabolic Diseases Research Center, University of Cincinnati College of Medicine, 2120 E. Galbraith Road, Cincinnati, OH, USA. Tel: +1 513 558-9152; Fax: +1 513 558 1312; [email protected] 20 11 2016 10 2016 01 10 2017 27 5 507512 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Purpose of review Phospholipids are major constituents in the intestinal lumen after meal consumption. This article highlights current literature suggesting the contributory role of intestinal phospholipid metabolism toward cardiometabolic disease manifestation. Recent findings Group 1b phospholipase A2 (PLA2g1b) catalyzes phospholipid hydrolysis in the intestinal lumen. The digestive product lysophospholipid, particularly lysophosphatidylcholine (LPC), has a direct role in mediating chylomicron assembly and secretion. The LPC in the digestive tract is further catabolized into lysophosphatidic acid (LPA) and choline via autotaxin-mediated and autotaxin-independent mechanisms. The LPC and LPA absorbed through the digestive tract and transported to the plasma directly promote systemic inflammation and cell dysfunction, leading to increased risk of cardiovascular disease and obesity/diabetes. The choline moiety generated in the digestive tract can also be used by gut bacteria to generate trimethylamine, which is subsequently transported to the liver and oxidized into trimethylamine-N-oxide (TMAO) that also enhances atherosclerosis and cardiovascular abnormalities. Summary Products of phospholipid metabolism in the intestine through PLA2g1b- and autotaxin-mediated pathways directly contribute to cardiometabolic diseases through multiple mechanisms. The implication of these studies is that therapeutic inhibition of PLA2g1b and autotaxin in the digestive tract may be a viable approach for cardiovascular and metabolic disease intervention. Phospholipase A2 autotaxin trimethylamine-N-oxide lysophosphatidylcholine lysophosphatidic acid atherosclerosis diabetes INTRODUCTION The incidence of obesity continues to escalate in recent years and has skyrocketed to near pandemic levels over the past 2 decades [1,2], leading to significant increased risk of metabolic diseases including diabetes, cardiovascular and fatty liver diseases [3]. A major causative factor for the increasing prevalence of obesity and obesity-related metabolic disorder is the chronic consumption of fat-rich meals. Hence, tremendous resources have been expended to reduce obesity and lower the risk of cardiometabolic diseases via limiting the amount of dietary lipids absorbed through the gastrointestinal tract. Most of the efforts in reducing lipid absorption have focused on limiting triglyceride digestion in the intestinal lumen and the transport and absorption of fatty acids and cholesterol through the intestinal mucosa. It is important to note, however, that phospholipids are also abundantly present in the intestinal lumen during meal consumption. In fact, earlier studies have shown that phospholipid digestion is also required for intestinal lipid absorption [4]. The major enzyme responsible for phospholipid hydrolysis in the intestinal lumen is the group 1b phospholipase A2 (PLA2g1b) derived from the pancreas. However, compensatory enzyme(s) is/are also present to mediate intestinal phospholipid digestion and lipid absorption in the absence of PLA2g1b [5]. Interestingly, the Pla2g1b-null mice are resistant to diet-induced obesity, diabetes, hyperlipidemia, and atherosclerosis despite their near normal lipid absorption efficiency [6]. These observations suggested that the digested products generated from PLA2g1b hydrolysis of phospholipids may be different from those produced by the compensatory enzyme(s), and the PLA2g1b-derived metabolites may be bioactive mediators of cardiometabolic diseases. The physiological roles of PLA2g1b in metabolic disease promotion have been reviewed in this journal previously [6]. The current review is an update of the literature, focusing on PLA2g1b-derived metabolites in the digestive tract and their influence on cardiometabolic disease manifestation. INTESTINAL LYSOPHOSPHATIDYLCHOLINE AND CHYLOMICRON BIOSYNTHESIS Phospholipase A2 catalyzes the hydrolysis of the ester bond at the sn-2 position of phospholipids to yield free fatty acids and lysophospholipids. The absorption of lysophospholipids is dramatically reduced in the absence of PLA2g1b, thus indicating that the compensatory enzyme for phospholipid hydrolysis in the intestinal lumen is not a phospholipase A2, but most likely the phospholipase B that hydrolyzes phospholipids to phosphoglycerol in the distal intestine [7]. In the proximal intestine where normal lipid digestion and absorption occurs, the lysophospholipids produced by phospholipid hydrolysis, primarily lysophosphatidylcholine (LPC), are absorbed into enterocytes where they participate in intracellular lipid trafficking that is necessary for chylomicron assembly and secretion. In an elegant recent study, Siddiqi and Mansbach showed that the free fatty acids and LPC generated from phospholipid hydrolysis are transported intracellularly in association with caveolin-1 containing endocytic vesicles [8]. The LPC on the surface of these vesicles activates protein kinase C-ζ (PKCζ), liberating it from the vesicles to allow the targeting of the vesicles to the endoplasmic reticulum where the free fatty acids are used for triglyceride biosynthesis. At the same time, the activated PKCζ released from caveolin-1 containing endocytic vesicles can bind and phosphorylate the secretion-associated Ras-related GTPase 1B (Sar1b), resulting in the disruption of the multiprotein complex that includes fatty acid binding protein-1 (FABP1). The FABP1 released from the multiprotein complex can bind to the ER membrane and organize the budding of the pre-chylomicron transport vesicles to transport pre-chylomicrons to the Golgi. Despite the evidence indicating that LPC absorption is required for intracellular lipid trafficking and chylomicron biosynthesis in the intestine, two independent laboratories have provided indirect evidence implying that the re-esterification of LPC to PC is also necessary for normal lipid absorption and chylomicron assembly and transport. Both laboratories generated mouse models with defective expression of lysophosphatidylcholine acyltransferase-3 (LPCAT3), the enzyme responsible for resterification of LPC to phosphatidylcholine. The Jiang laboratory showed that the lack of LPCAT3 expression lower plasma levels of cholesterol, phospholipid, and triglyceride due to reduced lipid absorption [9]. Subsequent studies from the same laboratory documented that the dominant effect of LPCAT3 deficiency on plasma lipid levels is due to its defective expression in the intestine while liver LPCAT3 deficiency only impacts on plasma triglyceride levels [10]. Studies from the Tontonoz laboratory yielded similar results, with additional observations that identified a mechanism related to reduced production of arachidonoyl phospholipids and the remodeling of enterocyte plasma membrane phospholipid composition being the gatekeeper important for passive lipid absorption and chylomicron production [11,12]. Although results from these studies have led to the suggestion that inhibiting intestinal LPCAT3 activity may be a novel approach for treatment of dietary lipid-associated hyperlipidemia and metabolic disorders [10,12], the desirability of this strategy requires extensive and detailed exploration considering that intestinal LPCAT3 activity appears to be required for survival of mice on lipid-rich diets and its inactivation causes the undesirable effects of food intake cessation and starvation [12]. INTESTINAL LYSOPHOSPHOLIPID METABOLISM AND SYSTEMIC INFLAMMATION In addition to its re-esterification by LPCAT3 into phosphatidylcholine, LPC may also be hydrolyzed by lysophospholipase D enzymes in the intestine to generate lysophosphatidic acid (LPA) and choline. Chronic feeding of LDL receptor-null mice with a high fat-cholesterol Western type diet increases the levels of unsaturated LPA but not saturated LPA in the intestine [13]. Subsequent study revealed that unsaturated LPA is produced from unsaturated LPC hydrolysis by autotaxin, also called ectonucleotide pyrophosphatase/phosphodiesterase family member 2 (gene name ENPP2), but the enzyme responsible for converting saturated LPC to saturated LPA remains unknown [14]. The contribution of unsaturated LPA toward cardiometabolic diseases was illustrated by studies showing that adding LPA with unsaturated fatty acyl moiety to a normal chow diet directly leads to dyslipidemia and aortic atherosclerosis in Ldlr−/− mice whereas adding LPA with saturated fatty acyl moiety has no effect [14]. Intestinal-derived unsaturated LPA promotes dyslipidemia and atherosclerosis via several mechanisms including the induction of lipid and cholesterol biosynthesis genes, alteration of peroxisome proliferator-activated receptor signaling, mitochondria dysfunction, and oxidative stress [15]. Unsaturated LPA produced in the digestive tract and absorbed into plasma circulation also promotes inflammation as evident by increased serum amyloid A [15]. PLASMA LPC AND LPA IN INFLAMMATORY AND CARDIOMETABOLIC DISEASES Lysophosphatidylcholine is present in plasma circulation at relatively high levels and includes species with both saturated and unsaturated fatty acids. The source of plasma LPCs includes intestinal derived LPC generated through PLA2g1b digestion as well as those produced by other phospholipase A2 enzymes present in plasma and in other tissues. The LPC in circulation can also be hydrolyzed by autotaxin present in the plasma that is derived from lymphatic high endothelial venules, adipose tissues, ovary, lung, and the liver. Both saturated and unsaturated LPCs as well as saturated and unsaturated LPAs have direct roles in eliciting inflammatory response, capable of activating neutrophils, various classes of T lymphocytes, monocytes/macrophages, and vascular endothelial and smooth muscle cells. Most of the effects of LPC in inflammatory response is mediated through activation of NFκB signaling pathway via binding and activation of the G-protein coupled receptor G2A, whereas the LPA effects are mediated through binding and activation of six other G-protein coupled receptors LPAR1-LPAR6 as well as the orphan receptors GPR87 and P2Y10. The roles of LPC and LPA in inflammatory disorders and atherosclerosis have been reviewed recently [16,17]. Studies exploring the influence of autotaxin in diet-induced obesity and diabetes are discussed below. ROLE OF AUTOTAXIN IN OBESITY AND DIABETES In addition to its expression in the digestive tract where it catalyzes the conversion of unsaturated LPC to unsaturated LPA, autotaxin is also highly expressed and secreted by adipocytes to activate adipocyte differentiation [18]. Moreover, autotaxin expression in adipocytes is elevated in obese and diabetic mice and humans [18,19], thus leading to the suggestion that adipocyte-derived autotaxin and its reaction products may also play a role in mediating obesity and diabetes. This hypothesis gained further support by experiments showing that acute elevation of plasma LPA levels impairs glucose tolerance and glucose-induced insulin secretion in high fat diet fed mice [20]. Several laboratories have independently generated autotaxin gene modified mice to test this hypothesis. While homozygous autotaxin deficient (Enpp2−/−) mice die in utero with vascular defects, Nishimura et al. found that heterozygous Enpp2+/− mice are healthy and their plasma LPA levels are ~50% of that in wild type mice [21]. Importantly, these investigators found that, in comparison to control wild type mice, the heterozygous Enpp2+/− mice gained less weight and displayed better glucose tolerance and insulin sensitivity in response to high fat diet feeding. Adipocyte-specific autotaxin inactivation was also found to reduce adiposity and improve glucose tolerance and insulin sensitivity by these investigators [21]. Conversely, transgenic mice with increasing autotaxin levels in the circulation via liver-specific over-expression or locally in adipose tissues through transgenic overexpression displayed increased adiposity in response to high fat diet [21,22]. Curiously and in contrast to these studies, Dusaulcy et al. found that adipocyte-specific autotaxin knockout mice have higher fat mass and larger adipocyte size after high fat feeding despite the lower plasma LPA levels and improved glucose tolerance [23]. Thus, whereas the contribution of autotaxin and its metabolites toward glucose intolerance and insulin resistance is undisputed, inconsistencies regarding body weight gain have not been resolved. Potential differences in autotaxin expression levels in other tissues as well as differences in the composition of the diets or genetic background of the animals used in the different experiments may explain some of the discrepant results. For example, the levels of saturated and unsaturated LPA were not reported. Whether saturated and unsaturated LPA have similar effects on adiposity, glucose tolerance, and insulin resistance have not been assessed. It would also be interesting to determine autotaxin expression levels in the digestive tract of these animals, as well as the contribution of intestinal-derived autotaxin in modulating diet-induced body weight gain. INTESTINAL-DERIVED CHOLINE IN CARDIOMETABOLIC DISEASE A second product of autotaxin-mediated LPC to LPA conversion is choline. Choline can be metabolized into trimethylamine (TMA) by bacteria residing in the gastrointestinal tract. In particular, the anaerobic bacteria in the Firmicutes and Proteobacteria phyla are the most active in choline consumption and TMA generation while another prominent phyla in the digestive tract, the Bacteroidetes, do not produce TMA from choline [24]. The TMA produced by gut microbes is transported to the liver whereupon oxidization by flavin monooxygenase-3 (FMO3) converts the TMA to trimethylamine-N-oxide (TMAO). Interestingly, humans and mice consuming a low fat diet have low levels of Firmicutes and high levels of Bacteroidetes, while high fat diet reduces the number of Bacterioidetes and increases the number of Firmicutes and Proteobacteria [25–27]. These diet-induced changes in gut bacterial composition and the corresponding increase in TMAO may be one factor responsible for cardiometabolic disease risk associated with high fat feeding [28]. In fact, a large scale metabolomics screening study has identified TMAO as a biomarker for cardiovascular disease [28]. A causative relationship between TMAO and atherosclerosis was demonstrated in mice via fecal transplantation of gut microbes from atherosclerosis-prone and atherosclerosis-resistant strains of mice with high and low TMAO producing gut microbes, respectively. The results showed that recipients of high TMAO producing bacteria displayed increased atherosclerosis [*29]. The mechanism by which choline and TMAO promotes cardiovascular disease appears to be multifactorial. Firstly, oral supply of choline or TMAO have been shown to increase expression of CD36 and SR-A1 scavenger receptors in macrophages to promote lipid accumulation and foam cell formation [30]. TMAO also inhibiting reverse cholesterol transport to increase foam cell deposition in the vessel wall [31]. Secondly, TMAO also promotes vascular smooth muscle and endothelial cell inflammation by activating mitogen-activated protein kinase pathways and NFκB signaling cascade [32]; and thirdly, TMAO may also directly contribute to left ventricular diastolic dysfunction and portend adverse outcomes in patients with chronic systolic heart failure [33]. The diet-induced increase of plasma TMAO requires supply of choline, typically as phosphatidylcholine, to the digestive tract as intravenous injection of phosphatidylcholine or choline did not show similar increase of plasma TMAO [30]. These latter observations highlighted the contributory role of PLA2g1b, intestinal autotaxin, and metabolites derived from their enzymatic reactions, in diet-induced cardiovascular disease. Several recent studies suggested that TMAO may also contribute to diabetes. In one study, dietary TMAO supplementation was shown to exacerbate hepatic insulin signaling dysfunction and gluconeogenesis and adipose tissue inflammation, leading to increased insulin resistance, glucose intolerance, and hyperinsulinemia in high fat diet-fed wild type mice [34]. Another study showed that reducing TMAO levels via FMO3 inactivation improves plasma glucose and insulin levels while increasing TMAO via FMO3 overexpression enhances hepatic gluconeogenesis in LDL receptor-null mice [35]. However, interpretation of results from the latter study should take into consideration that modulating FMO3 expression may have effects independent of TMAO, including a direct role of FMO3 in modulating peroxisome proliferator-activated receptor-α and Kruppel-like factor 15 pathways [35] as well as FMO3 regulation of the forkhead box O1 transcription factor [36]. Therefore, pharmacologic inhibition of FMO3 to reduce TMAO levels for cardiometabolic disease intervention should be approached cautiously with considerations given to other potential side effects. The inhibition of PLA2g1b and autotaxin in the digestive tract may be more promising because inhibitors that act exclusively in the digestive tract are likely to have minimal adverse effects [37]. CONCLUSION Phospholipid and lysosphospholipid metabolism in the digestive tract contributes to high fat diet-induced cardiovascular and metabolic disease risk through multiple mechanisms (summarized in Figure 1). Firstly, phospholipid hydrolysis by PLA2g1b, while not required for lipid absorption, generates lysophospholipids such as LPC to promote hyperlipidemia, systemic inflammation and cell dysfunction. Secondly, autotaxin-mediated catabolism of unsaturated LPC to unsaturated LPA generates additional bioactive metabolites with pro-inflammatory properties to promote cell dysfunction. Thirdly, the choline moiety generated from the autotaxin enzymatic reaction provides the substrate for microbial metabolism to generate TMA, which is ultimately delivered to the liver to produce TMAO to further enhance atherosclerosis. These mechanisms indicate that inhibition of phospholipase-autotaxin pathway in the digestive tract may be a viable strategy for cardiometabolic disease management. Financial support and sponsorship No financial support was received for the preparation of this manuscript. Research conducted in the author’s laboratory is supported by grants DK069967, DK074932, HL118001, and HL131028 from the National Institutes of Health. FIGURE 1 Schematic diagram depicting the multiple roles of intestinal phospholipid digestion and metabolism in cardiometabolic diseases. PLA2g1b, group 1b phospholipase A2; LPC, lysophosphatidylcholine; LPA, lysophosphatidic acid; ATX/ENPP2, autotaxin/ectonucleotide pyrophosphatase/phosphodiesterase family member 2; TMA, trimethylamine; TMAO, trimethylamine-N-oxide; FMO3, flavin monooxygenase-3. KEY POINTS Lysophospholipids generated in the intestinal lumen from phospholipase A2 group 1b digestion of phospholipids modulate intestinal chylomicron assembly and systemic inflammation to promote hyperlipidemia, atherosclerosis, and metabolic diseases. Autotaxin-mediated conversion of unsaturated lysophospholipids to unsaturated lysophosphatidic acids in the intestinal lumen also contribute directly to systemic inflammation and cardiometabolic diseases. The choline moiety liberated from lysophosphatidylcholine-to-lysophosphatidic acid conversion is metabolized by gut bacteria to form trimethylamine that is ultimately oxidized to trimethylamine-N-oxide that directly enhances cardiovascular disease development. Phospholipid remodeling and metabolism in the intestinal lumen directly promotes cardiometabolic disease and inhibition of this pathway may be an option for intervention. Conflicts of Interest None REFERENCES AND RECOMMENDED READING 1 Ogden CI Carroll MD Kit BK Flegal KM Prevalence of childhood and adult obesity in the United States, 2011–2012 JAMA 2014 311 806 814 24570244 2 Kakkar AK Dahiya N Drug treatment of obesity: current status and future prospects Eur J Intern Med 2015 26 89 94 25634851 3 Guh DP Zhang W Bansback N The incidence of co-morbidities related to obesity and overweight: a systematic review and meta-analysis BMC Public Health 2009 9 88 19320986 4 Beil FU Grundy SM Studies on plasma lipoproteins during absorption of exogenous lecithin in man J Lipid Res 1980 21 525 536 7400685 5 Richmond BL Boileau AC Zheng S Compensatory phospholipid digestion is required for cholesterol absorption in pancreatic phospholipase A2 deficient mice Gastroenterology 2001 120 1193 1202 11266383 6 Hui DY Phospholipase A2 enzymes in metabolic and cardiovascular diseases Curr Opin Lipidol 2012 23 235 240 22327613 7 Takemori H Zolotaryov FN Ting L Identification of functional domains of rat intestinal phospholipase B/lipase. Its cDNA cloning, expression, and tissue distribution J Biol Chem 1998 273 2222 2231 9442065 8 Siddiqi S Mansbach CM Dietary and biliary phosphatidylcholine activates PKCz in rat intestine J Lipid Res 2015 56 859 870 This study shows that lysophospholipids generated in the intestinal lumen and absorbed into enterocytes are required for appropriate intracellular lipid trafficking that is necessary for chylomicron assembly and secretion 25713101 9 Li Z Jiang H Ding T Deficiency in lysophosphatidylcholine acyltransferase 3 reduces plasma levels of lipids by reducing lipid absorption in mice Gastroenterology 2015 149 1519 1529 26226572 10 Kabir I Li Z Bui HH Small intestine but not liver lysophosphatidylcholine acyltransferase 3 (Lpcat3) deficiency has a dominant effect on plasma lipid metabolism J Biol Chem 2016 291 7651 7660 This study reveals the importance of LPCAT3 expression in intestine in mediating dietary lipid absorption and plasma lipoprotein homeostasis 26828064 11 Rong X Wang B Dunham MM LPCAT3-dependent production of arachidonoyl phospholipids is a key determinant of triglyceride secretion eLIFE 2015 4 e06557 This paper shows that LPCAT3 mediated reesterification of LPC in the intestine remodels phospholipid composition and is required for appropriate chylomicron secretion 12 Wang B Rong X Duerr MA Intestinal phospholipid remodeling is required for dietary-lipid uptake and survival on a high fat diet Cell Metab 2016 23 492 504 This paper shows that intestinal LPCAT3 deficiency leads to poor survival of mice on high fat diet. The paper also reveals that LPCAT3 activity in the intestine is required for maintaining appropriate membrane phospholipid composition and chylomicron secretion 26833026 13 Chattopadhyay A Navab M Hough G A novel approach to oral apoA-I mimetic therapy J Lipid Res 2013 54 995 1010 23378594 14 Navab M Chattopadhyay A Hough G Source and role of intestinally derived lysophosphatidic acid in dyslipidemia and atherosclerosis J Lipid Res 2015 56 871 887 This paper shows that unsaturated LPA in the intestine is derived from autotaxin-mediated hydrolysis of unsaturated LPC whereas saturated LPC to saturated LPA conversion is autotaxin independent 25646365 15 Navab M Hough G Buga GM Transgenic 6F tomatoes act on the small intestine to prevent systemic inflammation and dyslipidemia caused by Western diet and intestinally derived lysophosphatidic acid J Lipid Res 2013 54 3403 3418 24085744 16 Sevastou I Kaffe E Mouratis M-A Aidinis V Lysoglycerophospholipids in chronic inflammatory disorders: The PLA2/LPC and ATX/LPA axes Biochim Biophys Acta 2013 1831 42 60 22867755 17 Smyth SS Mueller P Yang F Arguing the case for the autotaxin-lysophosphatidic acid-lipid phosphate phosphatase 3-signaling nexus in the development and complications of atherosclerosis Arterioscler Thromb Vasc Biol 2014 34 479 486 24482375 18 Ferry G Tellier E Try A Autotaxin is released from adipocytes, catalyzes lysophosphatidic acid synthesis, and activates preadipocyte proliferation. Up-regulated expression with adipocyte differentiation and obesity J Biol Chem 2003 278 18162 18169 12642576 19 Boucher J Quilliot D Praderes JP Potential involvement of adipocyte insulin resistance in obesity-associated up-regulation of adipocyte lysophospholipase D/autotaxin expression Diabetologia 2005 48 569 577 15700135 20 Rancoule C Attane C Gres S Lysophosphatidic acid impairs glucose homeostasis and inhibits insulin secretion in high fat diet obese mice Diabetologia 2013 56 1394 1402 23508306 21 Nishimura S Nagasaki M Okudaira S ENPP2 contributes to adipose tissue expansion and insulin resistance in diet-induced obesity Diabetes 2014 63 4154 4164 24969110 22 Federico L Ren H Mueller PA Autotaxin and its product lysophosphatidic acid suppress brown adipose differentiation and promote diet-induced obesity in mice Mol Endocrinol 2012 26 786 797 22474126 23 Dusaulcy R Rancoule C Gres S Adipose-specific disruption of autotaxin enhances nutritional fattening and reduces plasma lysophosphatidic acid J Lipid Res 2011 52 1247 1255 21421848 24* Romano KA Vivas EI Amador-Noguez D Rey FE Intestinal microbiota composition modulates choline bioavailability from diet and accumulation of the proatherogenic metabolite trimethylamine-N-oxide MBio 2015 6 e02481 This paper shows that choline metabolism to TMA in the gastrointestinal tract is dependent on the type and composition of the gut microbiota 25784704 25 Wu GD Chen J Hoffmann C Linking long-term dietary patterns with gut microbial enterotypes Science 2011 334 105 108 21885731 26 Hildebrandt MA Hoffmann C Sherrill-Mix SA High-fat diet determines the composition of the murine gut microbiome independently of obesity Gastroenterology 2009 137 1716 1724 e1711 1712 19706296 27 Turnbaugh PJ Backhed F Fulton L Gordon JI Diet-induced obesity is linked to marked but reversible alterations in the mouse distal gut microbiome Cell Host Microbe 2008 3 213 223 18407065 28 Tang WHW Wang Z Levison BS Intestinal microbial metabolism of phosphatidylcholine and cardiovascular risk New Engl J Med 2013 368 1575 1584 23614584 29** Gregory JC Buffa JA Org E Transmission of atherosclerosis susceptibility with gut microbial transplantation J Biol Chem 2015 290 5647 5660 This paper used fecal transplantation approach to document the role of gut microbiota composition in determining atherosclerosis susceptibility 25550161 30 Wang Z Klipfell E Bennett BJ Gut flora metabolism of phosphatidylcholine promotes cardiovascular disease Nature 2011 472 57 63 21475195 31 Koeth RA Wang Z Levison BS Intestinal microbiota metabolism of L-carnitine, a nutrient in red meat, promotes atherosclerosis Nat Med 2013 19 576 585 23563705 32* Seldin MM Meng Y Qi H Trimethylamine N-oxide promotes vascular inflammation through signaling of mitogen-activated protein kinase and nuclear factor-kB J Am Heart Assoc 2016 5 e002767 This paper identifies one mechanism by which TMAO promotes vascular inflammation and increased atherosclerosis 26903003 33 Tang WHW Wang Z Shrestha K Intestinal microbiota-dependent phosphatidylcholine metabolites, diastolic dysfunction, and adverse clinical outcomes in chronic systolic heart failure J Cardiac Fail 2015 21 91 96 34 Gao X Liu X Xu J Dietary trimethylamine N-oxide exacerbates impaired glucose tolerance in mice fed a high fat diet J Biosci Bioeng 2014 118 476 481 24721123 35* Shih DM Wang Z Lee R Flavin containing monooxygenase 3 exerts broad effects on glucose and lipid metabolism and atherosclerosis J Lipid Res 2015 56 22 37 This paper shows that FMO3 inhibition lowers plasma glucose and lipid levels to improve atherosclerosis, but the FMO3 effects are multi-factorial and not necessarily dependent on TMAO production 25378658 36* Miao J Ling AV Manthena PV Flavin-containing monooxygenase 3 as a potential player in diabetes-associated atherosclerosis Nat Commun 2015 6 6498 This paper shows that FMO3 regulates forkhead box O1 transcription factor in modulating atherosclerosis 25849138 37 Charmot D Non-systemic drugs: A critical review Curr Pharm Des 2012 18 1434 1445 22300258
PMC005xxxxxx/PMC5125295.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101505086 36963 Sci Transl Med Sci Transl Med Science translational medicine 1946-6234 1946-6242 27629486 5125295 10.1126/scitranslmed.aad9943 NIHMS830247 Article Enhanced T cell responses to IL-6 in type 1 diabetes are associated with early clinical disease and increased IL-6 receptor expression Hundhausen Christian 1 Roth Alena 12 Whalen Elizabeth 1 Chen Janice 1 Schneider Anya 13 Long S. Alice 1 Wei Shan 1 Rawlings Rebecca 1 Kinsman MacKenzie 1 Evanko Stephen P. 4 Wight Thomas N. 4 Greenbaum Carla J. 5 Cerosaletti Karen 1 Buckner Jane H. 1* 1 Translational Research Program, Benaroya Research Institute at Virginia Mason, 1201 Ninth Avenue, Seattle, WA 98101, USA 2 Hannover Medical School, Department of Pediatric Pneumology, Allergology and Neonatology, Carl-Neuberg-Strasse 1, 30625 Hannover, Germany 3 Central Clinic Augsburg, Neurological Clinic and Clinical Neurophysiology, Stenglinstrasse 2, 86156 Augsburg, Germany 4 Matrix Biology Program, Benaroya Research Institute at Virginia Mason, 1201 Ninth Avenue, Seattle, WA 98101, USA 5 Diabetes Research Program, Benaroya Research Institute at Virginia Mason, 1201 Ninth Avenue, Seattle, WA 98101, USA * To whom correspondence should be addressed: Benaroya Research Institute, 1201 Ninth Avenue, Seattle Washington 98101, Telephone: (206) 287-1033; [email protected] 17 11 2016 14 9 2016 14 9 2017 8 356 356ra119356ra119 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Interleukin-6 (IL-6) is a key pathogenic cytokine in multiple autoimmune diseases including rheumatoid arthritis and multiple sclerosis, suggesting that dysregulation of the IL-6 pathway may be a common feature of autoimmunity. The role of IL-6 in type 1 diabetes (T1D) is not well understood. We show that signal transducer and activator of transcription 3 (STAT3) and STAT1 responses to IL-6 are significantly enhanced in CD4 and CD8 T cells from individuals with T1D compared to healthy controls. The effect is IL-6-specific because it is not seen with IL-10 or IL-27 stimulation, two cytokines that signal via STAT3. An important determinant of enhanced IL-6 responsiveness in T1D is IL-6 receptor surface expression, which correlated with phospho-STAT3 levels. Further, reduced expression of the IL-6R sheddase ADAM17 in T cells from patients indicated a mechanistic link to enhanced IL-6 responses in T1D. IL-6-induced STAT3 phosphorylation was inversely correlated with time from diagnosis, suggesting that dysregulation of IL-6 signaling may be a marker of early disease. Finally, whole-transcriptome analysis of IL-6-stimulated CD4+ T cells from patients revealed previously unreported IL-6 targets involved in T cell migration and inflammation, including lymph node homing markers CCR7 and L-selectin. In summary, our study demonstrates enhanced T cell responses to IL-6 in T1D due, in part to, an increase in IL-6R surface expression. Dysregulated IL-6 responsiveness may contribute to diabetes through multiple mechanisms including altered T cell trafficking and indicates that individuals with T1D may benefit from IL-6-targeted therapeutic intervention such as the one that is being currently tested (NCT02293837). Introduction Type 1 diabetes (T1D) is a chronic, multifactorial autoimmune disease in which the pancreatic islet cells are destroyed, leading to lifelong dependence on exogenous insulin therapy. To date, there is no cure. However, understanding and treating factors of autoimmune inflammation may effectively treat or even prevent disease progression. One of the factors involved in autoimmune inflammation is interleukin-6 (IL-6), a multifunctional cytokine with a role in chronic inflammatory and autoimmune diseases. IL-6 can be produced by many cell types including stromal cells and cells of the immune system, with monocytes and neutrophils being major sources of IL-6 after bacterial or viral infection (1). Elevated IL-6 serum/tissue concentrations are a hallmark of rheumatoid arthritis, systemic lupus erythematosus and multiple sclerosis and often correlate with disease activity (2–4). Mice deficient for IL-6 are protected from experimental autoimmune encephalomyelitis (5) and blockade of the IL-6 receptor (IL-6R) suppresses collagen-induced arthritis (6), indicating that IL-6 can drive autoimmunity. Additionally, the successful treatment of rheumatoid arthritis, systemic juvenile idiopathic arthritis, or Castleman’s disease with the anti-IL-6R antibody tocilizumab demonstrates the benefit of targeting the IL-6/IL-6R axis in humans (7). IL-6 signals predominantly via the Janus kinase (JAK)/signal transducer and activator of transcription (STAT) pathway (8). Binding of IL-6 to the cell surface-expressed IL-6R leads to recruitment and dimerization of gp130, the common signal-transducing subunit for the IL-6 family of cytokines. gp130 dimerization activates JAK family kinases, which phosphorylate tyrosine residues on gp130. Subsequent docking of the phosphatase SHP2 (Src homology 2 domain-containing tyrosine phosphatase 2) and STAT transcription factors (STAT3 and STAT1) stimulates the MAPK (mitogen-activated protein kinase) and STAT pathways, respectively. The STAT proteins are phosphorylated, dimerize, and translocate to the nucleus, where they induce transcription of target genes (8). A unique feature of IL-6 biology is a signaling mechanism termed “trans-signaling”, which facilitates activation of cells lacking membrane-bound IL-6R (mbIL-6R) (9). This mode of IL-6 signaling occurs through engagement of gp130 with a complex of IL-6 and a soluble form of the IL-6R (sIL-6R). Whereas small quantities of sIL-6R protein are generated by translation of an IL-6R splice variant, most of the soluble receptor arises from proteolytic cleavage of the IL-6R ectodomain from the cell surface, a process referred to as “shedding”(9, 10). ADAM17, also known as TACE, has been identified as the major protease that mediates IL-6R shedding (11, 12). The pathological effects of IL-6 in autoimmunity are associated with the IL-6R-gp130-STAT3 axis; signaling via this pathway is essential for T helper 17 (TH17) cell differentiation and inhibition of regulatory T (Treg) cell development by suppression of FOXP3 expression (13, 14). Furthermore, IL-6 induced phosphorylation of STAT3 (pSTAT3) can mediate resistance of T effector (Teff) cells to suppression by Treg cells (15, 16). In T1D, the role of IL-6 is unclear. One early report demonstrated a significantly reduced incidence of diabetes with blockade of IL-6 in the nonobese diabetic (NOD)/Wehi mouse model of T1D (17). Data on serum IL-6 levels in T1D are inconsistent (18–20); however, other findings support the disease relevance of the IL-6 pathway, including increased IL-6 production by monocytes from type 1 diabetic subjects (21), increased numbers of TH17 cells in new-onset T1D (22), Teff resistance in T1D (23, 24), and association with a genetic variant in the IL6R gene (25). Here, we find that IL-6-induced pSTAT3 and pSTAT1 are significantly increased in peripheral blood CD4 and CD8 T cells from patients with T1D compared with healthy controls, due, in part to, increased surface expression of the IL-6R. We show that enhanced IL-6 signaling associates with early clinical disease and that reduced expression of ADAM17 is a likely cause for elevated surface IL-6R. We also provide evidence for IL-6-induced upregulation of lymph node and peripheral tissue homing receptors, indicating increased T cell migratory capacity. Together, our results suggest a role for the IL-6 pathway in the immune dysregulation of T1D and also potential benefit from IL-6 targeted therapies in this disease, one of which is currently in clinical trial (http://www.extend-study.org/). Results IL-6 signaling is enhanced in T cells from type 1 diabetic patients IL-6-mediated STAT3 phosphorylation was evaluated in peripheral blood mononuclear cells (PBMCs) stimulated with 2 ng of IL-6 for 10 min, a dose that consistently resulted in submaximal pSTAT3 signals. IL-6-responsive cells were detected by flow cytometry as a pSTAT3-positive population distinct from untreated or non-responsive cells (fig. S1). Although there was no difference in baseline pSTAT3 between controls and individuals with diabetes (fig. S2), we measured significantly increased IL-6-induced pSTAT3 signals in CD4 T cells from patients compared to controls (Fig. 1). This increase was observed in both the naïve (CD45RA+) and memory (CD45RA−) compartment and was evident by an increased frequency of pSTAT3-positive cells and a higher fold change in mean fluorescence intensity (MFI) (Fig. 1A). To exclude a dose-specific effect, we treated cells with varying IL-6 concentrations and found that pSTAT3 was consistently increased in CD4 T cells from T1D subjects (Fig. 1B). IL-6/pSTAT3 was also enhanced in CD8 T cells from patients, although the increase was restricted to the naïve CD8 subset (Fig. 1C). A strong positive correlation between IL-6 and pSTAT3 in CD4 and CD8 T cells from the same individuals demonstrated that enhanced IL-6 responsiveness was subject-specific and consistent across cell types (Fig. 1D). To validate the reproducibility of our phospho-flow assay, we measured IL-6/pSTAT3 in PBMCs from the same blood draw on two different occasions, which demonstrated excellent reproducibility of the assay. In addition, the IL-6/pSTAT3 signaling phenotype was stable over time (fig. S3). IL-6/pSTAT1 responses not only were relatively weak but also increased in T cells from T1D subjects and were highly correlated with IL-6/pSTAT3 (Fig. 1E and fig. S4). Furthermore, enhanced pSTAT3 appeared to be IL-6- specific because stimulation with IL-27 or IL-10, two cytokines that also signal via STAT3 and gp130 in the case of IL-27, showed similar pSTAT3 levels in controls and patients (Fig. 1, F and G). Together, these data suggest that T cells from patients with T1D are hyperresponsive to IL-6 stimulation. Enhanced T cell responses to IL-6 are associated with early disease We next sought to examine clinical features of the individuals with T1D. We found no correlations between the IL-6-induced pSTAT3 signaling level to age, body mass index (BMI), age at diagnosis, glycated hemoglobin (HbA1c), or blood glucose levels (Fig. 2A). In contrast, we found that time from diagnosis negatively correlated with the frequency of pSTAT3+ CD4 and CD8 cells (Fig. 2B). IL-6R surface expression is increased in T cells from patients and correlates with IL-6-induced pSTAT3 To determine the mechanisms that lead to enhanced T cell responses to IL-6 in T1D, we next assessed the expression of IL-6 signaling components by flow cytometry and real-time polymerase chain reaction (PCR). Analyzing samples from the same blood draws that had previously been assayed for pSTAT3 in response to IL-6, we found that baseline expression of mbIL-6R was significantly increased in individuals with diabetes (Fig. 3). This increase was seen in the naïve (CD45RA+) CD4 and CD8 T cell subsets, as well as in CD4 memory (CD45RA−) cells. CD8 memory cells displayed the lowest mbIL-6R expression, consistent with the low pSTAT3 response in this subset, and no difference was detected between controls and patients (Fig. 3A). Similar gp130 levels on T cells from control and diabetes subjects (Fig. 3B) implicated IL-6R expression in the enhanced IL-6 response. We found a significant positive correlation between mbIL-6R expression and IL-6/pSTAT3 in all T cell subsets, with the exception of the CD8 memory cells (Fig. 2C, and fig. S6). To gain insight into potential cell intrinsic factors that contribute to altered IL-6 response in T1D, we examined the IL-6R variant 358Ala (rs2228145 A>C) that has been associated with reduced IL-6R surface levels and protection in T1D (25). To determine whether this variant contributed to the altered response to IL-6 or IL-6R surface expression in T cells, we stratified our data based on rs2228145 genotype. However, in our cohort of established T1D, we did not observe an effect of the rs228145 C allele on mbIL-6R expression or IL-6 induced pSTAT3 (Fig. 3D). We also performed quantitative real-time PCR for expression of signaling components in unstimulated CD4+CD25− T cells. Transcript levels of the IL-6R itself, the kinases TYK2, JAK1, and JAK2, as well as the negative regulators of the IL-6 pathway, SOCS1 and SOCS3, did not differ between groups (Fig. 3E). IL-6, which induces transcription of IL-6R (26), was detected at similar levels in sera from controls and patients, and IL-1β and tumor necrosis factor (TNF), described to enhance IL-6R shedding (27) were not decreased in T1D (fig. S7). Together, the data show that enhanced T cell responses to IL-6 in T1D are largely determined by increased IL-6R surface levels, which appear to be caused by altered posttranslational regulation of the receptor, as opposed to soluble factors that induce gene expression. ADAM17 is down-regulated in T cells from individuals with T1D IL-6R surface expression is tightly regulated by proteolytic cleavage (shedding) through two proteases of the ADAM family, ADAM17 and ADAM10 (11). In T cells from healthy individuals, ADAM17 has been identified as the major protease that mediates IL-6R shedding after T cell receptor (TCR) activation (28). We performed quantitative real-time PCR and found that baseline messenger RNA (mRNA) expression of ADAM17, but not ADAM10, was significantly reduced in CD4+CD25− T cells from T1D patients (Fig. 4A). The T cells of T1D subjects also express less mature (active) ADAM17 by Western blot analysis of CD3 cell lysates (Fig. 4B and fig. S8), and protein expression was significantly reduced on the surface of resting CD8 T cells with a similar trend on CD4 T cells of T1D subjects (Fig. 4C). On the basis of these observations, we asked whether IL-6R processing by ADAM17 would be altered in T1D. To do this we designed an IL-6R shedding assay in which resting or anti-CD3/CD28-activated CD3 T cells from T1D subjects were cultured in the presence or absence of the ADAM17 inhibitor TAPI. After 4 hours, sIL-6R concentrations in the culture supernatant and surface expression of mbIL-6R and ADAM17 were measured by enzyme-linked immunosorbent assay (ELISA) and flow cytometry, respectively. The results of these experiments are shown in Fig. 4. Whereas sIL-6R could be detected in both the supernatant of resting and activated cells, TCR stimulation resulted in a 3.5-fold increase in the release of the soluble form of the receptor (88 pg/ml (resting) versus 308 pg/ml (activated), P = 0.007). However, in the presence of TAPI, anti-TCR-induced IL-6R shedding was efficiently blocked, with sIL-6R levels returning close to baseline concentrations (P = 0.007). Constitutive IL-6R shedding was also inhibited by TAPI, although to a lesser extent (Fig. 4D). Flow cytometric analysis showed that the accumulation of sIL-6R after TCR stimulation was matched by a significant decrease in mbIL-6R and up-regulated levels of surface ADAM17 (Fig. 4E and fig. S5). These data were comparable to results obtained from healthy controls (fig. S9, A and B), confirming a similar role of ADAM17 in IL-6R shedding from T cells of controls and patients. The relationship between sIL-6R, mbIL-6R and ADAM17 was further analyzed by linear regression demonstrating a significant inverse correlation between mbIL-6R and ADAM17, and between mbIL-6R and sIL-6R (Fig. 4F). A time course experiment suggested that the increased IL-6R surface expression in T cells from patients relative to controls is maintained in activated T cells over time (fig. S9C). Together, our data suggest that decreased ADAM17 expression, but not protease activity, in T cells from individuals with T1D contributes to higher IL-6R surface levels on T1D T cells. Transcriptome analysis of IL-6-treated CD4 cells reveals previously unreported target genes of IL-6 associated with T cell migration Having addressed mechanistic questions of enhanced IL-6 signaling in T1D, we sought to gain more insight into the functional consequences of our findings. Given the role of IL-6 in balancing Teff and Treg cells (13, 14), we measured frequency and cytokine production (IFN-γ, IL-17A) of both T cell subsets in PBMCs from controls and patients. For these studies we selected T1D individuals with enhanced IL-6 signaling and compared them to T1D and control subjects who do not have them this phenotype to specifically assess the impact of the IL-6/pSTAT3 response on T cell lineage. However, despite some variation and a trend toward reduced percentage of Treg cells in patients with higher IL-6/pSTAT3 signaling, no significant differences were detected (fig. S10). Because relatively little is known about IL-6-specific gene expression in human T cells, we next performed whole-transcriptome analysis of untreated and IL-6-stimulated CD4+CD25− T cells from a cohort of T1D patients studied above with available samples (n = 7). A total of 5836 Ensembl genes were significantly regulated by IL-6, with a fold change expression greater than 2 (in either direction). Of these, 3743 genes could be mapped by the Database for Annotation, Visualization and Integrated Discovery (DAVID) (29, 30) and were further analyzed: 56% of the 3743 genes were up-regulated, and 80% of all differentially expressed genes exhibited fold changes between 2 and 5 (Fig. S11). Performing Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway mapping (31), we identified multiple enriched pathways in both the up-regulated and down-regulated gene sets (table S1). Whereas most of the up-regulated pathways were related to metabolic processes such as purine or fatty acid metabolism, the pathway “cytokine-cytokine receptor interaction” stood out because of its specific role in immune cell biology: 40 genes formed a network with a cluster of chemokines and chemokine receptors involved in T cell migration and inflammatory responses. The highest up-regulated receptors were CCR5 and CXCR6 (fold changes of 6.3 and 5.3, respectively), whereas CCR1, CCR2 and CCR7 were up-regulated between 2.3 and 3.5 fold (Fig. 5, A and B). IL-6 also triggered expression of several chemokines including CCL24, CCL22, and CXCL10, albeit at very low copy numbers (Fig. 5A and tables S2 and S3). In addition to pathway mapping we performed Gene Ontology (GO) enrichment analysis (32, 33) and found that 36 biological process GO terms were enriched in the up-regulated gene set. Of those, six were directly related to cell motility or regulation of the immune system, with gene counts between 13 and 28, and enrichment scores between 2.7 and 7.0 (Fig 5C and table S4). Besides the previously identified chemokine receptors, other IL-6 induced genes implicated in T cell migration and activity included SELL (encoding L-selectin), MMP-7, MMP-9, syndecan-2 and syndecan-4 (table S5). We next performed hierarchical clustering for a set of IL-6 regulated genes, which we considered most relevant for immune cell function. We found that the effect of IL-6 was largely consistent across subjects and that functionally related genes, for example CCR7, SELL and CD27, clustered together (Fig 5D). RNA sequencing (RNA-seq) data validation by quantitative reverse transcription PCR (qRT-PCR) confirmed the role of IL-6 in gene expression of most of the tested genes, including CCR7, SELL, CCR5 and CXCR6 (fig. S12). Last, by combining phospho-flow cytometry with qRT-PCR data from the same subjects, we demonstrated that IL-6-induced mRNA levels of CCR7 and SELL in T cells correlate with IL-6/pSTAT3 signaling strength (Fig 5E). In summary, our data show that IL-6 significantly enhances expression of cell migration- and inflammation-associated genes in CD4 Teff cells from patients with T1D. Many of the IL-6-regulated genes that were identified through the RNA-seq experiment encode surface expressed proteins, which interact with extracellular matrix (CD62L) or respond to chemokine gradients (CCR7, CCR5, CCR2, and CXCR6), thereby facilitating cell adhesion and tissue-specific homing. To assess the effect of IL-6 on surface expression of these markers, we cultured purified T cells in the presence or absence of IL-6 for 48 hours and evaluated cell surface expression by flow cytometric analysis. Samples for these assays were selected from T1D and control subjects who we had previously characterized for their T cell responses to IL-6, and which reflected the range of responses seen. The results of this experiment are shown in Fig. 6 and fig. S13. Using this approach, we found that the chemokine receptors CXCR6 and CCR5 were significantly up-regulated by IL-6 on effector memory (TEM) but not central memory (TCM) cells, whereas we detected virtually no expression of CCR2 on any T cell subset (Fig. 6A and B, and fig. S14). In contrast, we found that IL-6 significantly up-regulated surface levels of CD62L and CCR7 as measured by MFI in naïve and memory CD4 and CD8 T cells, and in patients as well as in controls, with the exception of CCR7 levels on CD4 memory T cells from healthy subjects, which remained largely unchanged in response to IL-6 (Fig. 6C and D, and fig. S13, A and B). Notably, there was a strong positive correlation between IL-6/pSTAT3 and surface expression of CD62L, in CD4 naïve T cells (r = 0.7, P = 0.003) (Fig. 6E). This relationship was also seen in total CD8 T cells, although in subjects with T1D only (fig. S13C). To assess the functional consequences of IL-6 exposure on migration we performed Transwell migration assays with CD4 T cells treated with IL-6 to establish their response to chemokines CCL19 (ligand for CCR7), CCL5 (ligand for CCR5) and CXCL16 (ligand for CXCR6). As shown in fig. S15, T cells that had been treated with IL-6 for 48 hours exhibited significantly increased migration toward CCL5 (P = 0.0002) and CCL19 (P = 0.039) compared to unstimulated cells. Consistent with this, we found a positive correlation between the MFI of CCR7 in naïve CD4 cells and cell migration toward CCL19. This data provide a mechanistic link between IL-6 and T cell migration that strengthens the possibility that enhanced T cell responses to IL-6 in T1D may contribute to disease pathogenesis by altering homing of T cells to the sites of islet inflammation. Discussion Here, we demonstrate enhanced IL-6 responsiveness in both CD4 and CD8 T cells of individuals with established T1D. Examining this in the context of time from diagnosis, we found that increased IL-6/pSTAT3 responses were seen in patients with a disease duration of less than 10 years, but that IL-6 signaling declined in subjects with long-standing disease (>20 years). This suggests that dysregulated IL-6 signaling in T1D is not due to the acute metabolic changes that occur at the time of diagnosis nor is it a trait acquired over time, for example a consequence of persistent hyperglycemia, but instead a hallmark for early events during clinical disease manifestation. Mechanistic studies showed that IL-6R surface expression is the major determinant for IL-6 signaling strength and that reduced expression of the IL-6R sheddase ADAM17 contributes to elevated IL-6R levels on T cells from patients with T1D. We provide insight into potential functional consequences of our findings by identifying IL-6-mediated upregulation of genes involved in T cell trafficking such as CCR7, L-selectin, CCR5 and CXCR6, and by demonstrating that T cell migration toward chemokine ligands of these receptors is enhanced by IL-6 in vitro. IL-6 has been implicated in the development and progression of autoimmune diseases through both its role in the promotion of the innate immune response and its influence on T cell lineage, favoring the development of TH17 T cells and inhibiting Treg development (13, 14). Further, IL-6 and enhanced IL-6 signaling have been linked to Teff resistance to suppression by Treg (15, 16). In line with other reports (20, 34), we did not find convincing evidence of increased IL-6 production in T1D. Nonetheless, enhanced IL-6 responsiveness has the potential to promote inflammation and autoimmunity, even in the presence of homeostatic levels of IL-6. In relapsing-remitting multiple sclerosis patients, enhanced response to IL-6 as measured by pSTAT3 was found in CD4 T cells and also linked to increased IL-6R expression and a decrease in cleavage of mbIL-6R. In these subjects, Teff resistance to Treg-mediated suppression was observed and shown to be correlated to the enhanced IL-6R expression and IL-6-mediated STAT3 phosphorylation (16). Multiple factors can influence the response to IL-6. Here, we demonstrated that increased expression of mbIL-6R correlated with enhanced IL-6-induced pSTAT3 in diabetic subjects independent of the clinical characteristics of patients. The lack of difference in IL-6R mRNA levels between controls and patients suggests that post-translational regulation of IL-6R is altered in T1D through shedding of the receptor. The T1D-associated IL-6R single-nucleotide polymorphism rs2228145 has been shown to influence plasma IL-6R and mbIL-6R levels (35) presumably through the process of altered shedding (36) but did not affect surface IL-6R expression in our cohort. Our finding of reduced transcript and protein levels of ADAM17 in peripheral blood T cells from subjects with T1D indicates a potential mechanism for enhanced IL-6 signaling and raises the possibility that activated T cells in vivo may retain their capacity to respond to mbIL-6R because of diminished ADAM17 expression or activity. The increase in mbIL-6R is likely not the sole factor that contributes to the enhanced response to IL-6 in our T1D subjects. Cell-intrinsic factors may contribute to the altered response to IL-6 in some individuals, particularly those who do not demonstrate increased mbIL-6R but do show enhanced pSTAT3 in response to IL-6. To address this, we examined the expression of the major IL-6 signaling components in our subjects. We found similar expression of the JAKs and the negative regulators SOCS3 and SOCS1, but we cannot rule out that these may be important in individual cases, nor can we rule out the influence of genetic variants of signaling proteins including TYK2 (37) and STAT3 (38) or as-yet-unknown regulators of the IL-6 pathway. The pathogenic effects of IL-6 on T cells in autoimmunity have been ascribed to the cytokine’s role in promoting TH17 cell differentiation and suppressing Treg cell function. In T1D, several lines of evidence support the involvement of TH17 immunity, including reports on increased frequencies of IL-17-producing CD4 and CD8 T cells in the circulation of new-onset patients (22) or increased numbers of TH17 cells in the pancreatic lymph nodes of long-term patients (40). Yet, overall, these effects are subtle and may be limited to specific subsets of TH17 and Treg cells. In our cohort, the frequency and cytokine profiles (IL-17A, IFN-γ) of CD4+Foxp3−Helios− Teff and CD4+Foxp3+Helios+ natural Treg cells in the blood were similar between IL-6/pSTAT3 “high” and ”low” T1D subjects, and no significant difference to healthy controls was observed. The limited number of subjects available for this analysis and the subtlety of these changes in the periphery may have made it difficult to demonstrate the impact of T cell responsiveness to IL-6 on the balance of Teff and Treg cells in the periphery. It is likely that enhanced IL-6 responses promote pathogenic T cell function locally in the inflamed islet or pancreatic lymph node in concert with other proinflammatory cytokines such as TNF, IL-1β, IL-21 and IL-23. In this setting, enhanced IL-6 responses may alter the fate and function of islet specific T cells, resulting in increased pathogenicity through impaired Treg function, the resistance of Teff to suppression by Treg (15, 16), and enhanced cytotoxicity via the induction of granzyme B linked to IL-6 trans-signaling (41, 42). In addition, IL-6 has been linked to reduced apoptosis of antigen-specific CD4 T cells in mice (43); this may be a mechanism by which autoantigen-specific T cells from individuals with diabetes prolong their survival. In our study, STAT1 responses to IL-6 were also increased in T cells from individuals with diabetes. STAT1 and STAT3 are ascribed distinctive functions (44, 45), which may lead to multiple, subtle alterations in T cell differentiation and function, together significantly contributing to disease. Whole-genome transcriptome analysis of unstimulated and IL-6-stimulated CD4+CD25− T cells from peripheral blood of patients with T1D yielded additional insight into the functional consequence of enhanced T cell responses to IL-6 in T1D subjects. Notably, a cluster of 40 genes involved in T cell trafficking and inflammatory responses was increased by exposure to IL-6. Using flow cytometry in combination with Transwell migration assays we were able to assess the impact of IL-6 on surface expression of these proteins in T cell subsets and on CD4 T cell migration in response to their ligands. We found that CXCR6, CCR5 and CCR7 were significantly up-regulated by IL-6 in CD4 and CD8 T cells, with CXCR6 and CCR5 expression largely restricted to the effector memory compartment. In agreement with this finding, we detected increased CD4 T cell migration toward their ligands after IL-6 treatment of the cells. CXCR6 and CCR5 are both chemokine receptors associated with TH1 or cytotoxic T cell function (46, 47) and, together with their ligands, have also been implicated in T1D pathogenesis. For example, circulating levels of the CCR5 ligand CCL5 were inversely correlated with β-cell function in children with T1D (48) and CXCL16, the ligand for CXCR6, was identified as one of the candidate genes in the Idd4 susceptibility locus of the NOD mouse (49). CCR7 directs recruitment of T cells into inflamed pancreatic islets (50). Together, this suggests that enhanced T cell responses to IL-6 may contribute to islet inflammation and destruction by enhancing the ability of islet specific T cells to access the islet. In keeping with this idea, IL-6 deficient mice exhibited reduced T cell recruitment to the site of acute inflammation, caused in part by dysregulation of T cell chemokine receptor expression (51). We acknowledge that our study has limitations, one of which is that we are sampling peripheral blood, whereas the impact of the disease is at the islets and pancreatic lymph nodes. Further, this study focuses on the global T cell response and not on the response of islet-specific T cells. However, our observation likely applies to these cells in a manner similar to that seen globally, including having an impact on the fate, function, survival and localization of pathogenic islet-specific T cells in T1D subjects. In addition we have limited our studies to T cells but alterations in IL-6 signaling may extend beyond T cells. For example ADAM17 is expressed by myeloid cells; if ADAM17 expression is diminished on myeloid cells in T1D, this would have the potential to further enhance the response to IL-6 in T1D. Enhanced IL-6 signaling in some subjects may be promoted by the immunologic and/or metabolic milieu before or at the time of diagnosis, which may wane over time. In other individuals enhanced IL-6 responses may be mediated through genetic mechanisms. Further exploration of these questions will require longitudinal studies to assess whether enhanced IL-6 signaling precedes clinical disease onset or changes during the course of disease. Additional work that explores whether there is altered IL-6R expression in other cell types will be important in evaluating therapeutic interventions that target this pathway. Collectively, these questions are pertinent to understand what factors lead to disease progression, and whether IL-6 responsiveness can predict disease progression or response to therapy. Understanding the mechanism and role of altered IL-6 signaling in T1D may assist in targeting this pathway for therapeutic intervention either through blockade of IL-6, and the IL-6R or through the use of small molecule inhibitors of the signaling pathway. Material and Methods Study design Here, we posed the hypothesis that T cell responses to IL-6 are increased in T1D, thereby contributing to disease pathogenesis. The effects of IL-6 on CD4 and CD8 T cells were evaluated in PBMCs from healthy control and T1D subjects using a flow cytometry-based STAT phosphorylation assay. Our approach of using PBMCs as opposed to isolated T cells was validated by coculture experiments that show that the presence of antigen presenting cells did not modulate IL-6/pSTAT signals in the 10-min assays we performed. Mechanisms of dysregulated IL-6 signaling in T1D were interrogated by serum cytokine analyses and expression studies of IL-6 pathway components at the transcript and protein level. Functional consequences were assessed by immunophenotyping of peripheral blood lymphocytes and by transcriptome sequencing of IL-6-treated CD4 Teff cells from patients. Control and T1D samples were selected randomly but matched for age and gender. No selection was made on human leukocyte antigen genotype or time from diagnosis. Samples were blinded for analysis but were provided in a manner that guaranteed that samples from both groups would be tested on each assay day. This was done to avoid batch effects between controls and T1D subjects. Power calculations were performed to determine the sample size required to have 80% power to detect a significant difference. These calculations indicated that a sample size of 25 per group would provide 80% power for differences ≥20 % in IL-6-induced pSTAT3. Human samples Frozen PBMCs and matched serum samples were obtained from participants in the Benaroya Research Institute (BRI) Diabetes and Immune Mediated Disease Registry and Repository. Control subjects (n = 58) were selected on the basis of the absence of autoimmune disease or any family history of autoimmunity. Patients with T1D (n = 60) had a disease duration between 0.2 and 51 years. Subjects were matched for age (mean age: controls, 35.2 ±13.2 years; T1D patients, 32.7 ±14.8 years), and all experiments were performed in a blinded manner. High baseline pSTAT3 levels in T cells from some controls and patients obscured the detection of IL-6 responses; those subjects were excluded from the study (fig. S2). Characteristics of study participants are listed in table S6. The research protocols were approved by the Institutional Review Board at BRI (#07109-136). Flow cytometry For analysis of STAT phosphorylation, thawed PBMCs were rested in serum-free X-VIVO-15 medium for 1 hour, washed with phosphate buffered saline (PBS), and stimulated at 106 cells per100 μl X-VIVO-15 with recombinant human IL-6 (rhIL-6) (2 ng/ml) (BD, catalog no. 550071) for 10 min, rhIL-10 (5 ng/ml) (BD, catalog no. 554611)for 20 min) or rhIL-27 (10 ng/ml) (eBioscience, catalog no. 14-8279) for 20 min. Cells were fixed and permeabilized using Fix Buffer I and Perm Buffer III (BD), respectively. Subsequently, cells were stained simultaneously for CD4 (BD, clone SK3), CD45RA (BD, clone HI100), CD8 (Beckman Coulter, clone SFCl121Thy2D3), pSTAT3 pY705 (BD, clone 4/P-STAT3), and pSTAT1 pY701 (BD, clone 4a) and incubated at room temperature for 45 min. Cell surface staining for IL-6R (BD, clone M5), gp130 (BD, clone AM64), CCR7 (BioLegend, clone G043H7), CD62L (BioLegend, clone DREG-56), CXCR6 (BioLegend, clone K041E5), CCR5 (BD, clone 2D7/CCR5), CCR2 (BD, clone 48607), ADAM17 (R&D Systems, clone #111633) and ADAM10 (BioLegend, clone SHM14) was carried out without previous cell fixation. Cells were acquired on a BD FACSCanto II and data were analyzed using FlowJo version vX.06 (Tree Star). ELISA and serum cytokine analysis sIL-6R serum concentrations were determined using the Human IL-6R Platinum ELISA Kit (eBioscience). IL-6, IL-1β, IFN-γ, and TNF serum levels were determined using the High Sensitivity Human Cytokine Magnetic Bead Kit (Milliplex Multi-Analyte Profiling, Millipore) in combination with the Luminex platform. CD3 T cell isolation and culture Total CD3 T cells were purified by negative selection from thawed PBMCs of healthy control and diabetes patients using magnetic-activated cell sorting (MACS) technology (Miltenyi). To assess the effect of IL-6 on expression of surface molecules, cells were cultured in 96-well round-bottom plates (Nunc) at 2 × 106 cells per well/ 200 μl of RPMI medium [10% fetal calf serum (FCS)] in the absence or presence of IL-6 (10 ng/ml) for 48 hours. Cells were washed with PBS and stained for cell surface markers for 30 min at room temperature. IL-6R shedding assay Purified CD3 T cells were distributed to 96-well round bottom plates at 106 cells/ 200μl of medium. Cells were activated with anti-CD3/CD28 beads (Dynabeads, Life Technologies) at a bead-to-cell ratio of 1:1 for 4 hours. In some conditions, cells were pretreated with the ADAM17 inhibitor TAPI-1 (20μM; Selleck Chemicals) for 30 min. After 4 hours, supernatants were collected and sIL-6R concentrations were determined by ELISA. For flow cytometric analysis, cells were magnetically separated from beads, washed in PBS and stained for CD4, CD8, CD45RA, IL-6R, ADAM17 and ADAM10. Western blotting Per subject, 2 × 106 CD3 T cells were lysed in 100 μl EBC buffer (50 mM tris (pH 8.0), 120 mM NaCl, 0.5 % NP-40, 1 mM EDTA, 50 mM NaF, 1 mM Na3VO4, 1 mM 2-mercaptoethanol, and protease inhibitor (Roche complete Mini tablets)), and 10 μg of total protein was subjected to reducing SDS-polyacrylamide gel electrophoresis using NuPAGE Novex 4-12% Bis-Tris gels (Life Technologies). After protein transfer to polyvinylidene difluoride membrane, the membrane was blocked with 5% bovine serum albumin (BSA) in tris-buffered saline/0.05% Tween 20 (TBST) for 1 hour at room temperature and then probed with rabbit anti-ADAM17 polyclonal antibody (1 μg/ml) (Chemicon International) at 4°C overnight. For ADAM17 protein detection, the membrane was stained with a peroxidase-labeled anti-rabbit secondary antibody (Vector) at a 1:50,000 dilution in TBST for 1 hour at room temperature, followed by thorough washing with TBST and incubation with enhanced chemiluminescence substrate (ECL, Pierce). Signals were recorded by exposure to x-ray film (Research Products International). As endogenous control, protein levels of transcription factor IIB (TFIIB) were determined using rabbit anti-TFIIB polyclonal antibody C-18 (Santa Cruz Biotechnology). Transwell migration assay MACS purified CD4 T cells from healthy controls and subjects with T1D were seeded in a 96-well round-bottom plate at a density of 2 × 106 cells in RPMI medium supplemented with 10 % human serum. Cells were left untreated or were stimulated with IL-6 (10 ng/ml) for 48 hours. Cells were labeled with carboxyfluoroscein diacetate succinimidyl ester (CFSE), washed with PBS and added at 2 × 105 cells per 50 μl of serum-free RPMI (without phenol red, 1% BSA) to the inserts of a 96-well Fluoroblok Transwell plate (BD Falcon) fitted with a light-blocking 3 μm polyethylene terephthalate membrane. The bottom chamber of the Transwells was filled with medium containing 1 CCL5, CCL19 or CXCL16 (100 ng/ml). As standard, 1:2 serial dilutions of CFSE-labeled cells were added to some wells. Cell migration was quantified every half hour over a period of 3 hours by measuring fluorescence intensity from the bottom of the plate using a plate reader (EnSpire, Perkin Elmer). Real-time qRT-PCR CD4+CD25− T cells were isolated by negative selection from thawed PBMCs from healthy control and T1D subjects using MACS (Miltenyi). Cells were either left untreated or stimulated with IL-6 (10 ng/ml) for 24 hours in complete RPMI medium supplemented with 10% FCS. RNA was extracted from 1.5 × 106 to 2.0 × 106 cells using the RNeasy Mini Kit with on column DNA digestion (Qiagen). Superscript III (Life Technologies) was used to generate complementary DNA, and gene expression was measured by multiplex real-time PCR performed on an ABI 7500 Fast Real-Time PCR System. The following TaqMan expression assays were used: IL-6R (Hs01075666_m1), TYK2 (Hs00177464_m1), JAK1 (Hs01026983_m1), JAK2 (Hs00234567_m1), SOCS1 (Hs00705164_s1), SOCS3 (Hs02330328_s1), ADAM17 (Hs01041915_m1), ADAM10 (Hs00153853_m1), SELL (Hs00174151_m1), CCR7 (Hs1013469_m1), CCR5 (Hs00152917_m1), and CXCR6 (Hs00174843_m1). Expression of GTF2B (Hs00976258_m1) was measured for normalization. After log2 transformation of the ΔCt value, the resulting value was multiplied by an arbitrary number to obtain units of relative expression (52). RNA processing for sequencing CD4+ CD25− T cells from seven T1D patients were isolated and IL-6 stimulated as described above. RNA was extracted from 0.5 × 106 cells using the RNeasy Kit (Qiagen), and quality was assessed using the Bioanalyzer 2100 (Agilent). Sequencing libraries were constructed from total RNA using TruSeq RNA Library Prep Kit v2 (Illumina) and clustered onto a flow cell, using a cBOT amplification system with a HiSeq SR Cluster Kit v4 (Illumina). Single-read sequencing was carried out on a HiSeq 2500 sequencer (Illumina), using a HiSeq SBS Kit v4 to generate 58-base pair reads, with a target of about 10 million reads per sample. RNA seq data analysis FASTQ files were aligned to a human reference genome to generate gene counts. Analysis of the 64,253 Ensembl ID count data was performed using the edgeR package (53) in the R software environment (54). For each gene, a negative binomial general linear model (55) that is appropriate for count data was used for the two-group comparison (stimulated versus non-stimulated) while controlling for batch effects. The Ensembl IDs were filtered to those that had a trimmed mean of M values (TMM)- normalized count (53) of at least one in at least one library, which resulted in 21,129 Ensembl IDs used in the general linear model. A multi-dimensional scaling (56) plot was created to look at the major sources of variation in the data after filtering. The two-group comparison had 25.7% biological coefficient of variation and 5,836 Ensembl IDs had a false discovery rate (57) less than 0.05 and fold change greater than 2 (in either direction). Significantly regulated genes were subjected to enrichment analysis of KEGG pathways and GO terms using the bioinformatics resources DAVID (29, 30) and GOrilla (32), respectively. Redundancy of GO terms was removed with REViGO (33). Protein interaction networks were created using STRING (58) in combination with Cytoscape (59). Statistical analysis Statistical analysis was performed using GraphPad Prism 6. To assess statistical significance the Wilcoxon matched pairs test and the Mann-Whitney U test were used. Results were expressed as means +/− SD, and differences were considered statistically significant at P ≤ 0.05. Linear regression was performed by computing the Pearson correlation coefficient (r). Outliers were removed using the ROUT method (coefficient Q = 0.1%) (60). Supplementary Material Supplemental Fig. S1. Gating strategy for analysis of IL-6/pSTAT3 by flow cytometry Fig. S2. Baseline pSTAT3 Fig. S3. Reproducibility of phospho-flow assay and stability of IL-6/pSTAT3 phenotype Fig. S4. Correlation of IL-6/pSTAT3 and IL-6/pSTAT1 Fig. S5. IL-6R and ADAM17 cell surface staining Fig. S6. Correlation of mbIL-6R expression and IL-6/pSTAT3 in naïve and memory T cell subsets Fig. S7. Serum cytokine concentrations Fig. S8. Functional role of ADAM17 in basal and activation-induced IL-6R shedding Fig. S9. Western blot for ADAM17 protein in CD3 T cells Fig. S10. Frequency and cytokine profile of peripheral blood Teff and Treg cells Fig. S11. RNA-seq: Overview of IL-6 regulated genes Fig. S12. qRT-PCR validation of RNA-seq data Fig. S13. Effect of IL-6 on surface expression of CCR7 and CD62L in CD8 T cells Fig. S14. Gating strategy for the analysis of proinflammatory chemokine receptors on TCM versus TEM cells. Fig. S15. IL-6 promotes CD4 T cell migration Table S1. KEGG pathway mapping Table S2. KEGG pathway cytokine-cytokine receptor interaction: IL-6 induced genes Table S3. Normalized counts of IL-6 regulated genes Table S4. GO enrichment analysis Table S5. GO terms associated with T cell motility: IL-6 induced genes Table S6. Study participant characteristics Table S7. Source data (Excel) We thank the BRI Clinical Core Laboratory, in particular T.-S. Nguyen, for coordinating the sample supply. We also thank the BRI Genomics and Bioinformatics Cores, in particular V. Gersuk and S. Presnell, for performing RNA sequencing and data analysis. We appreciate the work of Diabetes Clinical Research Program clinical coordinators and thank R. McMurry, I. Frank, and Anne Hocking for technical assistance. In addition, we thank the participants in the BRI Diabetes and Immune Mediated Disease Registries. Funding: We acknowledge support by U01 AI101990, DP3 DK104466 and Juvenile Diabetes Research Foundation Collaborative Center for Cell Therapy grant 2-5RA-2014-150-Q-R to JHB, U19 AI050864-11 (to KC) and the Benjamin and Margaret Hall Foundation. Author contributions: C.H. designed and performed the experiments, analyzed the data and wrote manuscript; A.R. performed experiments and analyzed data; E.W. analyzed RNA sequencing data and was the consulting statistician; J.C. performed experiments; A.S. performed preliminary experiments; A.L. assisted with flow cytometry experiments and reviewed the manuscript; S.W. performed genotyping; R.R. and M.K. performed experiments; S.P.E. and T.N.W. assisted with Transwell migration assays; CG recruited diabetes patients and reviewed the data with respect to clinical data, and reviewed the manuscript; K.C. designed and coordinated experiments, and wrote manuscript; J.H.B conceived and oversaw project. J.H.B. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Competing interests: The authors declare that they have no competing interests. Data and materials availability: RNAseq data have been deposited to the Gene Expression Omnibus database under accession number GSE78922. Figure 1 IL-6-induced pSTAT3 and pSTAT1 are increased in T cells from patients with T1D Thawed and rested PBMCs from healthy controls and subjects with T1D were treated with recombinant cytokine (IL-6, IL-10, or IL-27) followed by staining for CD4, CD8, CD45RA, pSTAT3 (pY705), and pSTAT1 (pY701). (A) CD4 T cell response to IL-6 as determined by frequency of IL-6-induced pSTAT3-positive cells [pSTAT3+ (%)] and fold change in pSTAT3 MFI (FC MFI pSTAT3) compared to unstimulated cells (left and right panels, respectively); n = 24 (Ctrl) and n = 27 (T1D). (B) Dose-response curve showing IL-6-induced pSTAT3 (MFI) in total CD4 T cells from controls and patients; n = 5 (Ctrl) and n = 5 (T1D). (C) CD8 T cell response to IL-6 as determined by frequency of IL-6-induced pSTAT3-positive cells and fold change in pSTAT3 MFI (left and right panels, respectively); n = 24 (Ctrl) and n = 27 (T1D). (D) Linear regression showing positive correlation between IL-6-induced STAT3 activation in CD4 and CD8 T cells; n = 24 (Ctrl) and n = 27 (T1D). (E) pSTAT1 in response to IL-6 in total CD4 and CD8 T cells; n = 22 (Ctrl) and n = 23 (T1D). (F and G) pSTAT3 in total CD4 and CD8 T cells following stimulation with IL-27 (F) [n = 13 (Ctrl) and n = 13 (T1D)] or IL-10 (G) [n = 12 (Ctrl) and n = 14 (T1D)]. Statistical tests: Mann-Whitney U; r = Pearson correlation coefficient. Figure 2 T cell responses to IL-6 decrease with time from diagnosis (A) Linear regression showing that IL-6-induced pSTAT3 is independent from age at draw, BMI, age at diagnosis, blood glucose levels, and glycated hemoglobin (HbA1c). (B) Linear regression showing the inverse correlation between time from diagnosis (disease duration) and IL-6/pSTAT3 in CD4 and CD8 T cells from subjects with T1D (left and right panels, respectively); 21≤ n ≤29 (A) and n = 26 (B) ; r = Pearson correlation coefficient. Figure 3 Surface IL-6R levels are increased in T1D T cells and correlate with IL-6-induced pSTAT3 Unstimulated PBMC from controls and patients obtained from the same blood draw as shown in Fig. 1 were stained for CD4, CD8, CD45RA, gp130 and IL-6R. IL-6R surface expression (mbIL-6R) was calculated as the antigen-specific MFI minus the MFI of a matched isotype control (see also fig. S5). (A) mbIL-6R expression in CD4 and CD8 T cells (left and right panels, respectively); n = 24 (Ctrl) and n = 27 (T1D). (B) gp130 expression (MFI) in CD4 T cells; n = 24 (Ctrl) and n = 27 (T1D). (C) Linear regression showing the positive relationship between mbIL-6R expression and IL-6-induced pSTAT3 in total CD4 T cells from subjects with T1D and controls (left and right panel, respectively); n = 24 (Ctrl) and n = 27 (T1D). (D) Effect of IL-6R rs2228145 A/C polymorphism on IL-6R surface levels and IL-6-induced pSTAT3 in CD4 T cells from subjects with T1D; n = 24. (E) Real-time PCR analysis for baseline expression of IL-6 signaling components in CD4+CD25− T cells; n = 12 (Ctrl) and n = 12 (T1D). Statistical tests: Mann-Whitney U; r = Pearson correlation coefficient. Figure 4 Reduced expression of IL-6R sheddase ADAM17 in T cells from patients with T1D (A) Real-time PCR for ADAM17 and ADAM10 transcript in unstimulated CD4+CD25−_ cells form controls and patients with T1D; n = 12 (Ctrl) and n = 12 (T1D). (B) (Left) Western blot for ADAM17 in CD3 T cell lysates from controls and patients; TFIIB protein levels were determined as loading control. (Right) Densitometric analysis of the ratio between mature and pro-form of ADAM17; n = 3 (Ctrl) and n = 3 (T1D). (C) Flow cytometry showing ADAM17 surface expression on resting T cells from controls and patients; n = 13 (Ctrl) and n = 12 (T1D)] (D to F) IL-6R shedding assay demonstrating the role of ADAM17 in constitutive and TCR-activation induced shedding of mbIL-6R. CD3 T cells from patients with T1D were incubated for 4 hours in the presence or absence of anti-CD3/CD28 beads (anti-TCR) and the ADAM17 inhibitor TAPI. sIL-6R levels in the supernatant were determined by (D) ELISA and (E) surface expression of mbIL-6R and ADAM17 by flow cytometry. (F) Linear regression showing the inverse correlation between mbIL-6R and ADAM17 expression (left panel) and mbIL-6R and sIL-6R concentrations (right panel). Triangles, nonactivated cells at 0 hours; circles, nonactivated cells at 4 hours; squares, anti-CD3/CD28 bead-activated cells at 4 hours; n = 8 (Ctrl) and n = 8 (T1D). Statistical tests: Mann-Whitney U (A to C); Wilcoxon matched pairs (E and F); r = Pearson correlation coefficient. Figure 5 Transcriptome analysis of IL-6 treated CD4+CD25− T cells from patients with T1D (A) Network of enriched KEGG pathway cytokine- cytokine receptor interaction illustrates cluster of IL-6-induced genes involved in T cell trafficking. Chemokine receptors are shown in color. (B) Fold change up-regulation, expression level (logCPM) and P value of the chemokine receptors identified in (A). (C) GO analysis of IL-6-induced genes shows enrichment of terms associated with cell migration and regulation of inflammatory responses. (D) Heat map of immune-relevant genes demonstrating consistent up-regulation by IL-6 across subjects. (Left) Unstimulated cells. (Right) IL-6-stimulated cells from the same subjects. (E) Linear regression showing positive correlation between IL-6/pSTAT3 and expression level of CCR7 and SELL in CD3+ T cells; n = 6 (Ctrl) and n = 6 (T1D). r = Pearson correlation coefficient. Figure 6 IL-6 upregulates inflammatory homing makers CXCR6 and CCR5 on TEM cells Purified CD3 T cells were cultured in the absence or presence of IL-6 (10 ng/ml) for 48 hours, followed by surface staining and flow cytometric analysis. (A and B) Effect of IL-6 on frequency of CXCR6+ (A) and CCR5+ (B) CD4 and CD8 TEM cells. (C and D) Effect of IL-6 on CD62L (C) and CCR7 (D) expression in naïve (CD45RA+) and memory (CD45RA−) CD4 T cells. (E) Linear regression analysis demonstrating the relationship between IL-6/pSTAT3 and baseline CD62L (left panel) and IL-6-induced CD62L (right panel) in CD4 naïve T cells; n=7 (Ctrl) and n=11 (T1D) in all panels. Statistical tests: Wilcoxon matched pairs, Mann Whitney U; r = Pearson correlation coefficient. References and notes 1 Calabrese LH Rose-John S IL-6 biology: Implications for clinical targeting in rheumatic disease Nat Rev Rheumatol 10 720 727 2014 25136784 2 Hirano T Matsuda T Turner M Miyasaka N Buchan G Tang B Sato K Shimizu M Maini R Feldmann M Excessive production of interleukin 6/B cell stimulatory factor-2 in rheumatoid arthritis Eur J Immunol 18 1797 1801 1988 2462501 3 Linker-Israeli M Deans RJ Wallace DJ Prehn J Ozeri-Chen T Klinenberg JR Elevated levels of endogenous IL-6 in systemic lupus erythematosus. A putative role in pathogenesis J Immunol 147 117 123 1991 2051017 4 Maimone D Guazzi GC Annunziata P IL-6 detection in multiple sclerosis brain J Neurol Sci 146 59 65 1997 9077497 5 Samoilova EB Horton JL Hilliard B Liu TS Chen Y IL-6-deficient mice are resistant to experimental autoimmune encephalomyelitis: roles of IL-6 in the activation and differentiation of autoreactive T cells J Immunol 161 6480 6486 1998 9862671 6 Fujimoto M Serada S Mihara M Uchiyama Y Yoshida H Koike N Ohsugi Y Nishikawa T Ripley B Kimura A Kishimoto T Naka T Interleukin-6 blockade suppresses autoimmune arthritis in mice by the inhibition of inflammatory Th17 responses Arthritis Rheum 58 3710 3719 2008 19035481 7 Tanaka T Narazaki M Kishimoto T Therapeutic targeting of the interleukin-6 receptor Annu Rev Pharmacol Toxicol 52 199 219 2012 21910626 8 Heinrich PC Behrmann I Haan S Hermanns HM Muller-Newen G Schaper F Principles of interleukin (IL)-6-type cytokine signalling and its regulation Biochem J 374 1 20 2003 12773095 9 Rose-John S Heinrich PC Soluble receptors for cytokines and growth factors: generation and biological function Biochem J 300 Pt 2 281 290 1994 8002928 10 Scheller J Chalaris A Garbers C Rose-John S ADAM17: a molecular switch to control inflammation and tissue regeneration Trends Immunol 32 380 387 2011 21752713 11 Matthews V Schuster B Schutze S Bussmeyer I Ludwig A Hundhausen C Sadowski T Saftig P Hartmann D Kallen KJ Rose-John S Cellular cholesterol depletion triggers shedding of the human interleukin-6 receptor by ADAM10 and ADAM17 (TACE) J Biol Chem 278 38829 38839 2003 12832423 12 Althoff K Reddy P Voltz N Rose-John S Mullberg J Shedding of interleukin-6 receptor and tumor necrosis factor alpha. Contribution of the stalk sequence to the cleavage pattern of transmembrane proteins Eur J Biochem 267 2624 2631 2000 10785383 13 Bettelli E Carrier Y Gao W Korn T Strom TB Oukka M Weiner HL Kuchroo VK Reciprocal developmental pathways for the generation of pathogenic effector TH17 and regulatory T cells Nature 441 235 238 2006 16648838 14 Veldhoen M Hocking RJ Atkins CJ Locksley RM Stockinger B TGFbeta in the context of an inflammatory cytokine milieu supports de novo differentiation of IL-17-producing T cells Immunity 24 179 189 2006 16473830 15 Goodman WA Levine AD Massari JV Sugiyama H McCormick TS Cooper KD IL-6 signaling in psoriasis prevents immune suppression by regulatory T cells J Immunol 183 3170 3176 2009 19648274 16 Schneider A Long SA Cerosaletti K Ni CT Samuels P Kita M Buckner JH In active relapsing-remitting multiple sclerosis, effector T cell resistance to adaptive T(regs) involves IL-6-mediated signaling Sci Transl Med 5 170ra115 2013 17 Campbell IL Kay TW Oxbrow L Harrison LC Essential role for interferon-gamma and interleukin-6 in autoimmune insulin-dependent diabetes in NOD/Wehi mice J Clin Invest 87 739 742 1991 1899431 18 Rosa JS Flores RL Oliver SR Pontello AM Zaldivar FP Galassetti PR Sustained IL-1alpha, IL-4, and IL-6 elevations following correction of hyperglycemia in children with type 1 diabetes mellitus Pediatr Diabetes 9 9 16 2008 18211631 19 Dogan Y Akarsu S Ustundag B Yilmaz E Gurgoze MK Serum IL-1beta, IL-2, and IL-6 in insulin-dependent diabetic children Mediators Inflamm 2006 59206 2006 16864906 20 Wedrychowicz A Dziatkowiak H Sztefko K Wedrychowicz A Interleukin-6 (IL-6) and IGF-IGFBP system in children and adolescents with type 1 diabetes mellitus Exp Clin Endocrinol Diabetes 112 435 439 2004 15372363 21 Bradshaw EM Raddassi K Elyaman W Orban T Gottlieb PA Kent SC Hafler DA Monocytes from patients with type 1 diabetes spontaneously secrete proinflammatory cytokines inducing Th17 cells J Immunol 183 4432 4439 2009 19748982 22 Marwaha AK Crome SQ Panagiotopoulos C Berg KB Qin H Ouyang Q Xu L Priatel JJ Levings MK Tan R Cutting edge: Increased IL-17-secreting T cells in children with new-onset type 1 diabetes J Immunol 185 3814 3818 2010 20810982 23 Schneider A Rieck M Sanda S Pihoker C Greenbaum C Buckner JH The effector T cells of diabetic subjects are resistant to regulation via CD4+ FOXP3+ regulatory T cells J Immunol 181 7350 7355 2008 18981158 24 Lawson JM Tremble J Dayan C Beyan H Leslie RD Peakman M Tree TI Increased resistance to CD4+CD25hi regulatory T cell-mediated suppression in patients with type 1 diabetes Clin Exp Immunol 154 353 359 2008 19037920 25 Ferreira RC Freitag DF Cutler AJ Howson JM Rainbow DB Smyth DJ Kaptoge S Clarke P Boreham C Coulson RM Pekalski ML Chen WM Onengut-Gumuscu S Rich SS Butterworth AS Malarstig A Danesh J Todd JA Functional IL6R 358Ala Allele Impairs Classical IL-6 Receptor Signaling and Influences Risk of Diverse Inflammatory Diseases PLoS Genet 9 e1003444 2013 23593036 26 Takizawa H Ohtoshi T Yamashita N Oka T Ito K Interleukin 6-receptor expression on human bronchial epithelial cells: regulation by IL-1 and IL-6 Am J Physiol 270 L346 L352 1996 8638726 27 Franchimont N Lambert C Huynen P Ribbens C Relic B Chariot A Bours V Piette J Merville MP Malaise M Interleukin-6 receptor shedding is enhanced by interleukin-1beta and tumor necrosis factor alpha and is partially mediated by tumor necrosis factor alpha-converting enzyme in osteoblast-like cells Arthritis Rheum 52 84 93 2005 15641051 28 Briso EM Dienz O Rincon M Cutting edge: soluble IL-6R is produced by IL-6R ectodomain shedding in activated CD4 T cells J Immunol 180 7102 7106 2008 18490707 29 Huang da W Sherman BT Lempicki RA Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources Nat Protoc 4 44 57 2009 19131956 30 Huang da W Sherman BT Lempicki RA Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists Nucleic Acids Res 37 1 13 2009 19033363 31 Kanehisa M Goto S KEGG: kyoto encyclopedia of genes and genomes Nucleic Acids Res 28 27 30 2000 10592173 32 Eden E Navon R Steinfeld I Lipson D Yakhini Z GOrilla: a tool for discovery and visualization of enriched GO terms in ranked gene lists BMC Bioinformatics 10 48 2009 19192299 33 Supek F Bosnjak M Skunca N Smuc T REVIGO summarizes and visualizes long lists of gene ontology terms PLoS One 6 e21800 2011 21789182 34 Kulseng B Vatten L Espevik T Soluble tumor necrosis factor receptors in sera from patients with insulin-dependent diabetes mellitus: relations to duration and complications fo disease Acta Diabetol 36 99 105 1999 10436260 35 van Dongen J Jansen R Smit D Hottenga JJ Mbarek H Willemsen G Kluft C Penninx BW Ferreira MA Boomsma DI de Geus EJ AAGC Collaborators The contribution of the functional IL6R polymorphism rs2228145, eQTLs and other genome-wide SNPs to the heritability of plasma sIL-6R levels Behav Genet 44 368 382 2014 24791950 36 Garbers C Monhasery N Aparicio-Siegmund S Lokau J Baran P Nowell MA Jones SA Rose-John S Scheller J There interleukin-6 receptor Asp358Ala single nucleotide polymorphism rs2228145 confers increased proteolytic conversion rates by ADAM proteases Biochim Biophys Acta 1842 1485 1494 2014 24878322 37 Wallace C Smyth DJ Maisuria-Armer M Walker NM Todd JA Clayton DG The imprinted DLK1-MEG3 gene region on chromosome 14q32.2 alters susceptibility to type 1 diabetes Nat Genet 42 68 71 2010 19966805 38 Jakkula E Leppa V Sulonen AM Varilo T Kallio S Kemppinen A Purcell S Koivisto K Tienari P Sumelahti ML Elovaara I Pirttila T Reunanen M Aromaa A Oturai AB Sondergaard HB Harbo HF Mero IL Gabriel SB Mirel DB Hauser SL Kappos L Polman C De Jager PL Hafler DA Daly MJ Palotie A Saarela J Peltonen L Genome-wide association study in a high-risk isolate for multiple sclerosis reveals associated variants in STAT3 gene Am J Hum Genet 86 285 291 2010 20159113 39 Lin WW Yi Z Stunz LL Maine CJ Sherman LA Bishop GA The adaptor protein TRAF3 inhibits interleukin-6 receptor signaling in B cells to limit plasma cell development Sci Signal 8 ra88 2015 26329582 40 Ferraro A Socci C Stabilini A Valle A Monti P Piemonti L Nano R Olek S Maffi P Scavini M Secchi A Staudacher C Bonifacio E Battaglia M Expansion of Th17 cells and functional defects in T regulatory cells are key features of the pancreatic lymph nodes in patients with type 1 diabetes Diabetes 60 2903 2913 2011 21896932 41 Bottcher JP Schanz O Garbers C Zaremba A Hegenbarth S Kurts C Beyer M Schultze JL Kastenmuller W Rose-John S Knolle PA IL-6 trans-signaling-dependent rapid development of cytotoxic CD8+ T cell function Cell Rep 8 1318 1327 2014 25199826 42 Bhela S Kempsell C Manohar M Dominguez-Villar M Griffin R Bhatt P Kivisakk-Webb P Fuhlbrigge R Kupper T Weiner H Baecher-Allan C Nonapoptotic and extracellular activity of granzyme B mediates resistance to regulatory T cell (Treg) suppression by HLA-DR-CD25hiCD127lo Tregs in multiple sclerosis and in response to IL-6 J Immunol 194 2180 2189 2015 25637022 43 Rochman I Paul WE Ben-Sasson SZ IL-6 increases primed cell expansion and survival J Immunol 174 4761 4767 2005 15814701 44 Hirahara K Onodera A Villarino AV Bonelli M Sciume G Laurence A Sun HW Brooks SR Vahedi G Shih HY Gutierrez-Cruz G Iwata S Suzuki R Mikami Y Okamoto Y Nakayama T Holland SM Hunter CA Kanno Y O’Shea JJ Asymmetric Action of STAT Transcription Factors Drives Transcriptional Outputs and Cytokine Specificity Immunity 42 877 889 2015 25992861 45 Choi YS Eto D Yang JA Lao C Crotty S Cutting edge: STAT1 is required for IL-6-mediated Bcl6 induction for early follicular helper cell differentiation J Immunol 190 3049 3053 2013 23447690 46 Kim CH Kunkel EJ Boisvert J Johnston B Campbell JJ Genovese MC Greenberg HB Butcher EC Bonzo/CXCR6 expression defines type 1-polarized T-cell subsets with extralymphoid tissue homing potential J Clin Invest 107 595 601 2001 11238560 47 Loetscher P Uguccioni M Bordoli L Baggiolini M Moser B Chizzolini C Dayer JM CCR5 is characteristic of Th1 lymphocytes Nature 391 344 345 1998 9450746 48 Pfleger C Kaas A Hansen L Alizadeh B Hougaard P Holl R Kolb H Roep BO Mortensen HB Schloot NC D. Hvidore Study Group On Childhood Relation of circulating concentrations of chemokine receptor CCR5 ligands to C-peptide, proinsulin and HbA1c and disease progression in type 1 diabetes Clin Immunol 128 57 65 2008 18434252 49 Ivakine EA Gulban OM Mortin-Toth SM Wankiewicz E Scott C Spurrell D Canty A Danska JS Molecular genetic analysis of the Idd4 locus implicates the IFN response in type 1 diabetes susceptibility in nonobese diabetic mice J Immunol 176 2976 2990 2006 16493056 50 Shan Z Xu B Mikulowska-Mennis A Michie SA CCR7 directs the recruitment of T cells into inflamed pancreatic islets of nonobese diabetic (NOD) mice Immuno Res 176 2976 2990 2006 51 McLoughlin RM Jenkins BJ Grail D Williams AS Fielding CA Parker CR Ernst M Topley N Jones SA IL-6 trans-signaling via STAT3 directs T cell infiltration in acute inflammation Proc Natl Acad Sci U S A 102 9589 9594 2005 15976028 52 Schmittgen TD Livak KJ Analyzing real-time PCR data by the comparative C(T) method Nat Protoc 3 1101 1108 2008 18546601 53 Robinson MD Oshlack A A scaling normalization method for differential expression analysis of RNA-seq data Genome Biol 11 R25 2010 20196867 54 R. Development Core Team R: A language and environment for statistical computing R Foundation for Statistical Computing Vienna, Austria 2012 55 McCarthy DJ Chen Y Smyth GK Differential expression analysis of multifactor RNA-Seq experiments with respect to biological variation Nucleic Acids Res 40 4288 4297 2012 22287627 56 Ritchie ME Phipson B Wu D Hu Y Law CW Shi W Smyth GK limma powers differential expression analyses for RNA-sequencing and microarray studies Nucleic Acids Res 43 e47 2015 25605792 57 Benjamini Y Hochberg Y Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing Journal of the Royal Statistical Society Series B (Methodological) 57 289 300 1995 58 Snel B Lehmann G Bork P Huynen MA STRING: a web-server to retrieve and display the repeatedly occurring neighbourhood of a gene Nucleic Acids Res 28 3442 3444 2000 10982861 59 Shannon P Markiel A Ozier O Baliga NS Wang JT Ramage D Amin N Schwikowski B Ideker T Cytoscape: a software environment for integrated models of biomolecular interaction networks Genome Res 13 2498 2504 2003 14597658 60 Motulsky HJ Brown RE Detecting outliers when fitting data with nonlinear regression - a new method based on robust nonlinear regression and the false discovery rate BMC Bioinformatics 7 123 2006 16526949
PMC005xxxxxx/PMC5125396.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101676854 44921 Chin J Sociol Chin J Sociol Chinese journal of sociology 2057-150X 2057-1518 27909585 5125396 10.1177/2057150X16638602 NIHMS829281 Article Developmental Idealism, Body Weight and Shape, and Marriage Entry in Transitional China Xu Hongwei Institute for Social Research, University of Michigan, 426 Thompson St. ISR 2459, Ann Arbor, MI 48106-1248, [email protected] , Phone: (734) 615-3552, Fax: (734) 763-1428 15 11 2016 4 2016 01 4 2017 2 2 235258 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. New trends toward later and less marriage are emerging in post-reform China. Previous research has examined the changing individual-level socioeconomic and demographic characteristics shaping marriage entry in Chinese adults. Employing a cultural model known as developmental idealism (DI), this study argues that a new worldview specifying an ideal body type has become popular in the West and that this new worldview has been exported to China. This new part of the DI package is likely stratified by gender, has a stronger impact on women than on men, and has likely penetrated urban areas more than rural areas. Drawing on the 1991-2009 longitudinal data from the China Health and Nutrition Survey, this study employs discrete-time logit models to estimate the relationships between various body types and transition to first marriage in Chinese young adults 18-30 years old. Body weight status and body shape are measured by body mass index (BMI) and waist-to-hip ratio (WHR), respectively, and further divided into categories of underweight, normal, and obese. Regression results indicate that larger values of BMI and WHR were associated with delayed entry into first marriage in urban women, whereas being overweight or obese was associated with accelerated transition to first marriage in rural men. Not only were these associations statistically significant, but their strengths were substantively remarkable. Findings from this study suggest that both body weight and body shape have important implications for marital success, independent of individual-level socioeconomic and demographic characteristics, and contribute to evolving gender and rural-urban disparities, as China is undergoing a rapid nutrition transition. body weight body shape marriage entry obesity Introduction Despite its distinct traditions concerning family and marriage norms (e.g., patriarchal and patrilineal family organization, universal marriage, young age at marriage entry), new trends in marriage and family formation are occurring in China and resemble the second demographic transition in the West. One prominent example is the slow yet steady increase in age at first marriage during the past five decades. One set of estimates suggest a two-year increase in the median age for men (from 23 to 25 for rural men and from 25 to 27 for urban men) and a three-year increase for women (from 20 to 23 for rural women and from 22 to 25 for urban women) between 1970 and 2000 (Han, 2010). Other estimates show a 1.5-year increase in the singulate mean age at marriage for women (from 23.3 to 24.7) and a 1.4-year increase for men (from 25.1 to 26.5) between 2000 and 2010 (Jones and Yeung, 2014). Delayed entry into first marriage is particularly notable in well educated women (Ji, 2015; Qian and Qian, 2014) and economically disadvantaged men (Yu and Xie, 2015a). In light of these emerging trends, demographers have set out to investigate the new determinants of marriage entry in post-reform China. Most studies to date acknowledge, either explicitly or implicitly, the pivotal role of ideational changes – new attitudes toward family formation and gender ideology resulting from Westernization, modernization, and market transition – in affecting the timing of first marriage. Empirically, however, these studies tend to focus on the direct effects of individual-level demographic and socioeconomic characteristics on marriage entry, leaving ideational factors relatively underexamined. For example, Yu and Xie (2015a) posited that rising consumption aspirations, together with the shift of gender ideology toward gender equity within marriage, have contributed to the gendered role of economic prospects in marriage formation during the post-reform era. Drawing on data for urban respondents from the 2003 and 2008 Chinese General Social Surveys, they found a positive effect of employment status on transition to marriage in men. This effect is more pronounced in the late-reform cohort (born after 1974) than in the pre- (born before 1960) and early-reform (born between 1960 and 1974) cohorts. On the other hand, education exhibits a marriage-delaying effect for both men and women, and this effect is also stronger in younger cohorts. Several studies argued that the traditional norm of hypergamy persists despite the rapid increase in Chinese women's education and the resulting narrowed gender gap in educational attainment (Han, 2010; Mu and Xie, 2014; Qian and Qian, 2014). Therefore, the marriage-delaying effect of education is most evident in women with college or higher educational attainment because of the difficulty in finding suitable mates in the marriage market to maintain educational homogamy and female hypergamy. In this study, I investigate individual-level determinants of marriage entry beyond demographic and socioeconomic characteristics. I draw upon a theoretical framework, known as developmental idealism (Thornton, 2001; Thornton, 2005), for understanding body size and shape as new ideational forces behind the transition to first marriage in post-reform China. Taking advantage of the rural-urban gaps in the progress of the nutrition transition (Doak et al., 2002; Du et al., 2002) and the pace of modernization in China (Schafer and Kwon, 2012; Whyte, 2010), this study tests varying effects of body weight status on timing of first marriage by gender and rural-urban strata among the Chinese adults. Given the enduring norm of nearly universal marriage in a narrow band of appropriate ages, or at least the stated desires to do so, (Raymo et al., 2015), I carry out a prospective analysis of transition to first marriage in young Chinese adults 18-30 years old. Specifically, this study seeks to understand: (1) how body weight and body shape are related to marriage entry; and (2) how these relationships may vary by gender and rural-urban residence. Theoretical background Developmental idealism The centuries-old models of developmentalism and modernization theory, embraced by Western scholars, policy makers, and other elites since the Enlightenment, prescribe a linear, universal pathway of development for all societies with varying rates of progress. Furthermore, it is widely believed that the most developed countries are located in northwest Europe and its overseas diasporas (particularly the United States), while countries in the rest of the world are ranked at different lower levels of the developmental hierarchy. Thornton (2001; 2005) argued that the developmental paradigm, “reading history sideways,” and the social science research on family change together produced a package of ideas known as developmental idealism (DI). He (2001: 454-455) originally formulated four basic propositions of DI as follows: (1) modern society is good and attainable; (2) the modern family is good and attainable; (3) a modern family is both a cause and an effect of a modern society; and (4) in a modern society, individuals are free and equal and social relationships are based on consent. In other words, the DI package includes ideas that specify what the good life is and how to achieve it. In the family arena, DI suggests that the new patterns of marriage and family in Western countries – including such attributes as individualism, nuclear households, marriages at mature ages, courtship preceding marriage, gender equality, and planned and low fertility – are modern, good, and attainable and thus should be pursued (Thornton, 2001). As Thornton (2001; 2005; 2012) has repeatedly emphasized, the power of DI as an ideational force does not depend largely on whether its propositions (e.g., “modern society and modern family are good and attainable”) are true or false, or good or bad. In fact, the developmental paradigm itself has been challenged and abandoned by many scholars in recent decades (Davis and Harrell, 1993; Goldscheider, 1971; Greenhalgh, 1996; Tilly, 1984). However, what matters most is that once accepted by ordinary people and/or endorsed by governments, organizations, and other institutions, the interrelated values and beliefs of DI can cause changes in a wide array of human behaviors and relationships, including marriage and family. The ideas of DI have been spreading around the world since large-scale globalization began in the 19th century when new means of transportation and telecommunications were invented (Amin, 1989; Blaut, 1993). To the extent that modernization theory and developmentalism remain popular among non-academic elites and the general public worldwide to date, we should not ignore the ideational force of DI for changing marriage and family patterns (Thornton, 2001; Thornton, 2005). In China, various educators, revolutionaries, political parties, and governments have acted as powerful agents in disseminating the developmental paradigm throughout the 20th century (Davis and Harrell, 1993; Thornton, 2001). One of the most prominent examples is the introduction of the one-child policy in 1979 and its strict enforcement ever since in order to boost state development and modernization as China reentered the global capitalist system (Greenhalgh, 2008). As a result, the propositions of DI have reached the majority of ordinary people and remain persuasive in contemporary China. For example, consistent with the DI propositions, survey data have shown that Chinese people perceive a reverse causal relationship between development and low fertility (Thornton et al., 2012), and that endorsement of DI is positively associated with endorsement of modern family values, including gender egalitarianism and late marriage for both Han and Muslim Chinese women (Lai and Thornton, 2015). Body weight and shape Why and how do body weight and shape affect marital status in young Chinese adults? In this study, I argue that small body size (normal or even underweight as opposed to overweight and obese) and certain body shapes are currently seen as ideal in being both good in themselves – that is, aesthetically superior – and conducive to acquiring other good things such as health, high-paying jobs, and attractive spouses. Sobal and Stunkard (1989) suggested that this particular cultural schema of ideal body type originated in the West and can be traced back to more than a century ago, when Veblen (1889) recognized thinness as a status symbol of the leisure class. By the 1980s, many people in the West, especially high-status women, had accepted and internalized social norms equating thinness with attractiveness – that is, fatness is bad while thinness is healthy, beautiful, and good (Dornbusch et al., 1984; Sobal and Stunkard, 1989). As the prevalence of obesity continued to rise in the U.S. and other Western countries in the 1990s and eventually turned into a global epidemic (Finucane et al., 2011; Flegal et al., 2012), the social pressure for a thin body and the stigma attached to a fat body have become pervasive (Allon, 1982; Carr and Friedman, 2005; Puhl and Brownell, 2001). This particular cultural schema of ideal body weight and shape has likely become part of the DI package and acted as an ideational force to affect individuals’ social and economic outcomes. For example, overweight and obese persons are discriminated against at home by their parents (Crandall, 1995), at school by peers and teachers (Crosnoe and Muller, 2004; Neumark-Sztainer et al., 1998), in the job market and workplace by employers (Cawley, 2004; Morris, 2006), and at clinics and hospitals by health care professionals (Najman et al., 1982; Schwartz et al., 2003). The results of this broad array of discrimination experiences against the overweight and obese include low socioeconomic status (Sobal and Stunkard, 1989; McLaren, 2007), reduced earnings (Cawley, 2004), poor physical and mental health (Haslam and James, 2005; Pinhey et al., 1997), and even perceived unattractive personalities (Carr and Friedman, 2005). With respect to marriage and family, studies in Western countries often conjecture that being overweight or obese is widely stigmatized and hence discriminated against in the marriage market, whereas thinness is a socially or culturally desirable trait that increases attractiveness to potential mates. Earlier studies report mixed findings; some reveal a significant association but in different directions, depending on the research setting, whereas others report no association at all (Sobal et al., 1992). These studies rely largely on cross-sectional analysis that is unable to resolve the debate between “marital selection” (i.e., body weight affecting entry into or exit from marriage), and “marital causation” (i.e., marital status inducing weight changes) (Sobal et al., 1992). Assuming that body weight signals health status or carries certain cultural values during the process of assortative mating, more recent U.S. studies have adopted a longitudinal design in order to reduce the potential problem of reverse causality in understanding how body weight status affects the likelihood of being married (Averett and Korenman, 1999; Fu and Goldman, 1996; Gortmaker et al., 1993; Jæger, 2011; Jeffery and Rick, 2002). However, these longitudinal studies are limited by particular measurement and methodological challenges, and the findings on the effect of body weight on marital status remain inconclusive and sometimes contradict one another. For example, drawing upon data from the 1981-1988 National Longitudinal Study of Youth (NLSY), Gortmaker et al. (1993) found that men who were overweight at ages 16-24 were less likely to be married in the subsequent seven years; whereas, using the 1979-1991 NLSY data, Fu and Goldman (1996) found an increased likelihood of first marriage among overweight men. Other studies of smaller, non-representative longitudinal samples reported no significant association between body weight and marital status among young or middle-aged adults (Jæger, 2011; Jeffery and Rick, 2002), or differential effects – positive for elderly men but negative for elderly women (Jæger, 2011). Despite the inconclusive findings in the U.S., the new DI worldview of “what is thin is healthy, beautiful, and good and hence should be sought after” has likely been disseminated to China, where rapid economic growth and urbanization have fueled the nutrition transition; that is, a shift from significant malnutrition due to poverty to increases in obesity and degenerative diseases due to improved living standards (Popkin, 2002) is taking place. In fact, despite the persistence of malnutrition in some poor subpopulations (Doak et al., 2005), the rates of overweight and obesity are surging in the Chinese population. In less than two decades, the prevalence of overweight and obesity has doubled in both adults – from 13% in 1993 to 26% in 2009 (Xi et al., 2012) and children – from 5% in 1991 to 13% in 2006 (Cui et al., 2010). Given the potentially huge disease burden associated with obesity, the Chinese government has launched an ambitious program known as Healthy China 2020, which will not only provide universal healthcare access and treatment for the entire population by the year 2020 but also promote healthy diets and active lifestyles (Hu et al., 2011; Yang et al., 2013). Such government-sponsored health campaigns, together with other educational programs and the mass media, help to provide the intellectual justification and establish legitimacy for exporting the Western worldview on body image to China. Once accepted by Chinese young adults, this DI worldview may affect their marriage-related decision-making and behaviors, leading to my first research hypothesis as follows: Hypothesis 1: Excessive body weight and an unhealthy body shape reduce an individual's attractiveness and marriageability, thus delaying his/her marriage entry. The DI worldview on what body type is good in marriage is likely stratified by gender. Psychological studies of body image have shown that thinness is often viewed as a key characteristic of beauty for women but not for men in Western cultures (Haworth-Hoeppner, 2000). Men and women also differ in how body fat is distributed in that men tend to accumulate more abdominal fat while women generally store more lower-body fat. This sex difference in body fat distribution may contribute to sex differences in metabolic and cardiovascular risks (Cartier et al., 2009; Karastergiou et al., 2012). Lower-body fat has also been hypothesized to increase the supply of neurodevelopmental resources, and thus a lower waist-to-hip ratio is found to be predictive of higher cognitive test scores in women and their children (Lassek and Gaulin, 2008). Taken together, these findings suggest that body shape is indicative of health status, especially for women and their fertility, which in turn helps to explain the widespread men's preferences for women's low waist-to-hip ratios, in both Western and non-Western populations (Dixson et al., 2007; Swami et al., 2006). The gender aspect of the DI worldview on body image may have already been exported to China. A recent study of young adults aged 18-30 in urban China found evidence of discrimination in the job market against overweight and obese women but not men (Pan et al., 2011). Therefore, my second hypothesis states that: Hypothesis 2: The effects of body weight and body shape are stronger for women than for men. Lastly, any particular idea or proposition of DI from the West may not be unanimously accepted by non-Western populations. Indigenous populations can modify or even resist various aspects of DI based on their cultural heritage, religious beliefs, historical experiences, and socioeconomic conditions (Thornton, 2001). In contemporary China, the rural-urban divide is arguably the most fundamental socioeconomic and demographic marker (Whyte, 2010). During the post-reform era, inequalities between rural and urban populations have been rising in terms of their socioeconomic status (Zhao, 2006), access to quality health care (Xu and Short, 2011) and health outcomes (Hou, 2008; Zimmer et al., 2007), despite the overall improvement in living conditions and population health. The urban Chinese have been shifting towards a high-fat, high-energy-density and low-fiber diet and decreased physical activity in work and leisure, all at a faster pace than their rural peers (Du et al., 2002). Consequently, chronic diseases, many of which correlate with obesity, have become the major source of disease burden in urban China, whereas infectious diseases remain prevalent in rural areas (Zhao, 2006). Therefore, the particular DI worldview of “what is thin is healthy, beautiful, and good” has probably penetrated the urban areas more than the rural areas, leading to my third hypothesis: Hypothesis 3: The effects of body weight and body shape on marriage entry are stronger in urban areas than in rural areas. Data and measures Subjects for this study were adult participants ages 18-30 in the China Health and Nutrition Survey (CHNS), a panel survey that includes more than 4,000 households across 9 provinces in contemporary China. The CHNS data are not nationally representative, but the households were selected through a multistage, random cluster sampling process from a diverse set of nine provinces in northeast, central, and south China. All the individuals in the sampled households were interviewed. Together, these nine provinces are home to more than 40% of China's population, or 548.56 million people. The average response rate at the individual level is 88% across waves. Details on the design and sampling of CHNS are available elsewhere (Popkin et al., 2010). Due to both the high response rate and the diversity of population sampled, the CHNS data allows us to make inferences about a large proportion of the Chinese population. This study draws on data from the most recent seven waves of the survey: 1991, 1993, 1997, 2000, 2004, 2006, and 2009. The sample is restricted to young adults who were never married at the onset of each wave. The longitudinal data tracked the marital status of the same adult respondents ages 18 or older over time and thereby permitted constructing the dependent variable, a binary indicator of whether a respondent made the transition into marriage between two consecutive waves. The CHNS data did not fully capture the specific time points of entry into marriage, resulting in so-called “interval censoring.” Nevertheless, the temporal ordering of life events helps to alleviate the problem of reverse causality and facilitate the identification of the effects of body weight and shape on marital status. The key predictors in this study are body mass index (BMI) and waist-to-hip ratio (WHR) to capture overall body weight and shape, respectively. Both variables were derived from objective anthropometric measures taken by trained health workers using a portable stadiometer, providing accurate and reliable assessments of body weight and shape that were rarely available in previous research. BMI was calculated as the ratio of body weight to height squared (kg/m2). WHR was calculated as waist circumference (cm) divided by hip circumference (cm) and then rescaled to 0-100 to avoid too many decimal points in regression coefficients. In addition to these linear operationalizations, categorical variables were created to capture potential non-linear associations of body weight and shape with marital status. BMI was divided into three groups, including underweight (less than 18.5), normal (18.5 or greater but less than 23), and overweight or obese (23 or greater), according to the recommended cut-off points by the World Health Organization (WHO) for Asia-Pacific populations (WHO et al., 2000). WHR was dichotomized as central obese (0.9 or greater in men and 0.85 or greater in women) and not according to WHO's general recommendation (WHO, 2008) due to its lack of ethnicity-specific guidelines. Other variables were constructed in similar ways as has been done in previous research (Chen et al., 2010; Xu et al., 2013; Xu and Short, 2011). Rural-urban residence was measured dichotomously. An urban community is an administratively defined community known as a “street committee” (ju-wei-hui), with an average population of about 3,000, while a rural community refers to a natural village, with an average population of about 3,800. Age and education were measured continuously in years. Birth cohorts were divided into three groups: those born in 1970 or earlier, those born between 1971 and 1980, and those born in 1981 or afterwards, according to the major periods in China's recent history and the data distribution of this study. Household income per capita was measured in Chinese yuan (RMB), inflated to 2009 levels, and log transformed in regression analysis. Occupation was categorized into four groups, including farmers or other agricultural workers, unskilled workers, skilled workers or professionals, and unemployed or other miscellaneous. Self-rated health was included to control for subjectively assessed general health status and grouped into poor or fair, good, and excellent compared to others of same age. Regional variations were controlled by a set of dummy variables indicating residence in the northeastern, coastal, inland, and mountainous southern provinces. Method Success in transition into marriage is not well differentiated by comparing people's ultimate marital status by middle age, but is better reflected in the timing of marriage, since it remains a universal norm to get married in China. Thus, similar to Fu and Goldman (1996), this study employed a discrete-time model to analyze correlates of risk of first marriage. Specifically, body weight and shape along with other control variables measured at the ith wave among respondents who had never married yet (i.e. still at the risk of first marriage) are used to predict whether they had entered marriage by the (i+1)th wave through a logit link. Lagging covariates produced a clear ordering of life events for easier identification of the temporal process of mate selection. The discrete-time logit models permit the use of time-varying covariates and thus take into account the possibility that body weight and shape as well as other potentially important factors related to the marriage process may change over the life course. They also adjust for the fact that some adults remained single by age 30, known as the right-censoring. As in other longitudinal studies, sample attrition over time poses a potential source of bias in the CHNS. Less than 6% of observations had missing values on the dependent or independent variables in any given wave. Unfortunately, between about 30-50% had missing values for the same respondents in the following waves due to loss to follow-up, resulting in missing information on the change in marital status between two consecutive waves. However, exploratory analysis suggested that marital status in any given wave did not predict the probability of having missing values in the following wave after controlling for age, gender, education, and family income, indicating that missingness at random assumption and sequential ignorability are plausible (Gelman and Hill, 2007). Therefore, unlike previous studies that did not make any statistical adjustment, this study applied an inverse probability weighting technique to address the missing data problem (Fitzmaurice et al., 2004). Specifically, a dichotomous variable indicating missing values in a subsequent wave was regressed on a number of variables including age, birth cohort, education, family income, occupation, self-rated health, and region of residence in the current wave. Probabilities of dropping out of the study in the next wave were then predicted based on the regression estimates. The respondents who had a high probability of dropping out but remained in the survey were weighted upward, while those who had a low probability were weighted downward, resulting in more balanced data than without any statistical adjustment. These inverse probability weights were used in the discrete-time logit models of entry into first marriage to reduce bias and improve efficiency in the estimates. Preliminary analysis produced more significant coefficient estimates without inverse probability weighting, indicating more conservative and hence robust results after adjusting for missing data. Nevertheless, caution should be applied when interpreting the results. After dropping cases with missing information, the final sample consists of 1,749 and 919 person-year records for men and women, respectively. All the models were fit to men and women separately given notable gender differences in marriage selection and body weight profiles. The contrasts between rural and urban populations were achieved by interacting the dichotomous indicator of rural-urban residence with BMI and WHR instead of further stratifying the sample. This model specification preserves the statistical power by maintaining a sample size as large as possible and is more parsimonious by fixing the effects of other control variables for rural and urban respondents of the same gender. For each measure of body weight status, two models were fitted sequentially, the first one without and the second one with the interaction term. Robust standard errors were estimated using the Huber-White sandwich estimators to adjust for repeated measures of the same respondents over time. Preliminary analysis explored random effects models as an alternative analytical strategy. However, only a few respondents contributed to more than two observations since most of them entered marriage within three waves. Thus, random effects models were confronted by a convergence problem under certain specification and thus not pursued here. Results Descriptive statistics Descriptive statistics of the dependent variables are presented in Table 1. The number of records in a single wave decreased from 763 in 1991 to 184 in 2006 as the respondents either exited from the risk pool after entry into marriage or became right-censored after age 30. Men and women differed little in their average BMI (about 21) and distributions of overall body weight status (about 13-15% underweight, 70% normal, and 14-17% overweight or obese). Women had on average a lower WHR (79.4) than men (83.8), but they also suffered from a greater prevalence of central obesity (19.1%) than men (13.1%). These rates remain substantially lower than those in the U.S. or among Asian Americans (Schiller et al., 2012), but they are suggestive of an emerging obesity epidemic in China. Turning to other covariates, the average age was 21.1 in women and 22 in men, and more than half of the sample belonged to the 1971-1980 cohort. The average years of schooling were 9.1 for men and women. Men in the sample came from on average slightly wealthier households and were more likely to be farmers but less likely to be unskilled workers, unemployed, or engaged in other types of employment. Women had a slightly stronger tendency than men to rate their health as poor or fair. Only about one third or fewer respondents lived in urban areas, and women were more likely to come from the coastal region but less likely to live in the south compared to men. Using the midpoints between two consecutive waves to impute the age at first marriage, Figure 1 plots the Kaplan-Meier estimates of survival rates, that is, the cumulative proportion of the respondents who remained unmarried by a given age, stratified by gender and rural-urban residence. Rural women entered first marriage at a median age of 24, earlier than rural men (25) and urban women (25), followed by urban men (26). The gender gap in age at first marriage within rural or urban strata reflects the longstanding norm and expectation for men to build up and secure the economic foundation of marriage (Holmgren, 1985), despite China's development and modernization in the recent decades. On the other hand, later marriage for urban men and women than for their rural counterparts may partly result from the former's better educational and occupational attainments during the market transition (Wang and Yang, 1996). The estimated median age at first marriage in the CHNS sample was consistently higher for each group by about one year than the 2010 Census data (NBSC 2011), probably due to inaccurate imputations using the mid-point between two waves of the CHNS. However, the sample gender and rural-urban patterns still hold in the 2010 and 2000 census data (Han, 2010). Regression estimates Table 2 shows the coefficient estimates from the discrete-time logit models using BMI. The coefficient of BMI was not significant in Model 1 for either men or women, providing no support for Hypothesis 1. Urban residence was associated with a reduced likelihood of first marriage compared to rural residence (marginally significant in men), confirming the patterns revealed from the Kaplan-Meier estimates. After adding the interaction between BMI and urban residence (Model 2), the main effect of urban residence became marginally significant in women though not significant in men, partially confirming Hypothesis 2. The coefficient for the main effect of BMI remained insignificant, indicating no association in rural residents. The interaction term was marginally significant in both men and women, partially confirming Hypothesis 3 about a stronger effect of BMI in urban areas. However, the marginally significant interaction between BMI and rural-urban residence does not reveal conclusively whether or not the effect of BMI was itself significant in urban residents, especially given the insignificant main effect of BMI. To obtain the estimated net effect of BMI in urban residents, Model 2 was refitted by switching the reference group from rural to urban in both the main and the interaction terms with everything else unchanged. The mean coefficient estimates and the associated 95% confidence intervals for the effects of BMI in urban men and women were plotted in Figure 2 (top-left panel) together with those in rural residents. On average, every one unit increase in BMI was associated with about a 0.17 decrease in the log-odds (or 15.5% lower odds) of entry into first marriage in urban women, but did not affect any other group. Table 3 presents the coefficient estimates from the models using WHR. Again, the main effect of WHR was not significant either with or without the interaction term, providing no evidence for Hypothesis 1. The interaction between WHR and rural-urban residence was significant in women only, providing partial support to Hypotheses 2 and 3. After refitting Model 2 with urban residence as the reference group, the mean coefficient estimates and the associated 95% confidence intervals were again plotted in Figure 2 (top-right panel) for men and women, rural and urban separately. On average, every one unit increase in WHR (rescaled to the range of 0-100) was associated with about a 0.08 decrease in the log-odds (or 7% lower odds) of first marriage in urban women, but did not affect any other group. Table 4 reports the coefficient estimates from the models using categorical body weight status. The main effect of being overweight or obese on entry into first marriage was positively significant as in Model 1, but the interaction effect was not the same as in Model 2, making it unclear which group was the driving the results. Again, the uncertainty was lifted after refitting the same models using different reference categories. Compared to having normal weight, being overweight or obese had 1.54 smaller log-odds (or 78.6% lower odds) of first marriage in urban women (see the bottom-left panel in Figure 2), although this association was only marginally significant. By contrast, being overweight or obese was associated with a 0.46 increase in the log-odds (or 58.9% higher odds) in rural men. Table 5 presents the coefficient estimates from the models using the binary indicator of central obesity. Neither was the main nor the interaction effect significant for having a body shape classified as central obesity. Among other covariates, age was, not surprisingly, positively associated with the timing of first marriage, a pattern that is consistent with the prevailing norm of getting married as a marker of successful transition into adulthood in China. There is some evidence of an interesting cohort difference in that confronted by an increasingly heavier economic burden to start a new family, men of the post-80s generation entered first marriage much later than those born in the 1970s (Xue, 2013). Overall, socioeconomic variables had little impact, which may be attributed to the universal marriage norms predicting that people of low SES also manage to get married through assortative mating. Nevertheless, among men, those who were unemployed or engaged in miscellaneous occupations tended to delay their first marriage compared to agricultural workers, a finding that is consistent with prior research (Yu and Xie, 2015a). Health status was strongly valued by women in selecting their potential mates as better self-rated health increased the chance of entry into marriage in men but not in women. The different effects of cohort, occupation, and self-rated health between men and women are suggestive of the enduring traditional gendered norms and expectations about a husband's role as the breadwinner. In light of the stringent job and housing markets, especially in urban China, these findings imply likely increased economic hardship for young men in attracting potential mates in the near future. Discussion and conclusion Family demographers have observed emerging trends toward later and less marriage in China and other East Asian countries, which resembles the second demographic transition in the West (Raymo et al., 2015). Cultural models have been proposed and employed to explain such new trends of family formation in the West (Lesthaeghe, 2010; Thornton, 2001; Thornton, 2005; Thornton, 2010). Demographic research on marriage entry in post-reform China recognizes the shifting attitudes toward marriage and family but empirically tends to focus on individual-level socioeconomic and demographic determinants of marriage entry. Against this background, the current study makes two important contributions to the literature. First, I employ DI as a theoretical framework to examine the worldview of ideal body type as an ideational force to affect marriage entry in Chinese young adults. The findings of the significant impacts of body weight and body shape, independent of a battery of personal socioeconomic and demographic characteristics, help to produce a fuller picture of the determinants of transition to first marriage. Second, situated in transitional China, this study broadens the application scope of DI as a theoretical framework to explain changing marriage and family patterns in non-Western countries. Several studies have applied the DI framework for understanding cohabitation (Yu and Xie, 2015b), beliefs about fertility change (Thornton et al., 2012) and family values (Lai and Thornton, 2015), and perceptions about development and inequality (Xie et al., 2012). However, to the best of my knowledge, this is the first study that explicitly applies the DI framework to investigating marriage entry in China or East Asia. Overall, the empirical findings in this study suggest that, as part of the DI package, the worldview of thinness being healthy and beautiful and fatness being bad has been exported to China but penetrated into different subpopulations to varying degrees, thereby only affecting marriage entry in certain subgroups. Specifically, among the four sex- and residence-specific subgroups, urban women are most vulnerable to the negative impacts of the Western worldview regarding body weight and shape. Both a larger body weight and relatively more fat accumulated in the abdomen rather than the lower body significantly delayed the timing of first marriage in urban women. These findings are consistent with the positive views of good health and aesthetic superiority attached to women's low BMI and WHR by their potential mates (Buss, 2004; Jæger, 2011). The significant effect of WHR in this study also implies that research on the socioeconomic and demographic consequences of body type should not be restricted to examining overall body weight alone, but also incorporate body fat distribution as an independent factor. A recent study showed that the employment rate in urban China was about 10-17% lower in young overweight or obese women, classified based on BMI, compared to their normal or underweight counterparts (Schafer and Kwon, 2012). Thus, it will be interesting to further assess the relative effect of unhealthy body shape on labor market outcomes, independent of the effect of excessive body weight. Furthermore, the effect sizes of both body weight and shape were quite substantial. For example, assuming an urban woman with the average waist and hip circumstances in this sample (77 and 92 cm, respectively), her odds of first marriage would decline by about one fourth had her waist circumstance increased by 3 cm while her hip circumstance remained unchanged (equal to about a 3.3-unit increase in WHR rescaled to the range 0-100). Such a change can be easily achieved within a three-year period as the average annual growth in WHR was estimated to be 1.44 in Chinese women (Xu et al., 2012). Given that the prevalence of overweight and obesity continues to rise in China and Chinese women in particular tend to gain extra weight at a faster pace than men (Schafer and Kwon, 2012), these findings imply that urban women are facing growing discriminations based on their physiques in both the job (Pan et al., 2011) and marriage markets. It is worth noting that certain subgroups have likely not accepted the new worldview in favor of a thin rather than a fat body. The regression estimates indicate that neither rural women nor urban men were affected by body weight or shape with respect to their timing of first marriage. In contrast, there appeared to be a positive return to overweight and obesity in rural men since being overweight or obese was associated with an earlier entry into marriage than normal weight. These findings are consistent with the gendered discrimination against the overweight and obese in Western populations (Gortmaker et al., 1993; Fu and Goldman, 1996), and also underscore the uneven pace of the nutrition transition between rural and urban populations in China (Du et al., 2002). These between-group contrasts also highlight the persistent gender inequality and rural-urban disparity in a wide array of family behaviors despite China's success in economic growth and development. To the extent that family is an important social institution throughout a person's life course, any attempt to achieve gender or rural-urban equity is unlikely to succeed without addressing these gaps in the family domain. Several limitations remain in this study. First, the CHNS data are not nationally representative. Thus, the results from this study cannot be generalized to the entire Chinese population, although significant inferences can be made for a large proportion of the total population. Second, missing data due to sample attrition over time poses a potential threat to the accuracy of regression estimates in this study. The inverse probability weighting technique helps to adjust for the missing data, but it does not solve the problem once and for all. Third, the exact timing of entry into marriage was not captured but only measured to an interval of time. Thus, it is unclear whether there existed any systematical measurement error in this regard across different groups, which would lead to biased estimates in this study. These limitations can be addressed in future research that employs new large-scale longitudinal data of high quality. Fortunately, it will not be long before that data is available, as greater efforts have recently been devoted to such data collection (e.g., Gan, 2012; Xie, 2012). Despite these limitations, this study is among the first to reveal the heterogeneous returns to body weight status in contemporary China's marriage market. It also expands the scope of the existing Western literature on the relationship between body weight and marital status to developing countries. Capitalizing on the prospective design of the CHNS and its rich data, this study has improved upon previous research on body weight and marital status in several important aspects. First, measures of body weight status are derived from objective anthropometric data rather than self-reported data. Second, this study distinguishes the role of body shape from that of overall body weight, which has been extensively examined, in shaping marriage entry. Third, using temporally lagged covariates in discrete-time models of transition to first marriage allows me to better alleviate the potential problem of marital selection. Future research can benefit from including direct measures of personal attitudes toward body weight and body shape as additional covariates to better understand cultural influences on marriage behaviors. Figure 1 Kaplan-Meier survival estimates of entry into first marriage. Figure 2 Coefficient estimates (with 95% confidence intervals) for the effects of body weight and shape on risk of first marriage from discrete-time logit models Table 1 Descriptive statistics of the independent variables Women Men Body mass index (BMI) 20.7 20.9 Overall body weight (%) Underweight 15.0 13.0 Normal 71.0 70.3 Overweight/Obese 14.0 16.7 Body Shape Waist-to-hip ratio (WHR) 79.4 83.8 Central obese (%) 19.1 13.1 Age (years) 21.1 22.0 Cohort (%) <=1970 19.4 24.5 1971-1980 65.4 58.4 >=1981 15.2 17.2 Years of education 9.1 9.1 Household income per capita (RMB) 4311.7 4685.3 Occupation (% ) Unemployed/Other 28.4 24.7 Farmer 29.8 36.5 Unskilled worker 23.7 21.3 Skilled worker/Professional 18.1 17.5 Self-rated health (%) Poor/Fair 18.4 14.1 Good 60.3 62.4 Excellent 21.3 23.4 Urban (%) 35.4 29.0 Region (%) Northeast 11.5 9.4 Coastal 27.0 21.6 Inland 33.2 33.6 South 28.3 35.4 N 919 1749 Table 2 Estimated coefficients from discrete-time logit models of marriage entry on body weight (continuous measure) Women Men Model 1 Model 2 Model 1 Model 2 Urban (ref: Rural) −0.49 (0.25) * 3.00 (1.80) † −0.25 (0.14) † 1.32 (0.95) BMI −0.06 (0.04) 0.00 (0.05) 0.01 (0.02) 0.04 (0.03) Urban × BMI −0.17 (0.09) † −0.07 (0.04) † Age (years) 0.27 (0.04) * 0.27 (0.04) * 0.16 (0.02) * 0.16 (0.02) * Cohort ( ref: 1971-1980) <=1970 −0.72 (0.30) * −0.71 (0.29) * −0.46 (0.14) * −0.47 (0.14) * >=1981 −0.10 (0.32) −0.10 (0.33) −0.28 (0.17) † −0.30 (0.17) † Years of education −0.03 (0.04) −0.03 (0.04) −0.02 (0.02) −0.02 (0.02) Household income per capita (logged) 0.02 (0.16) 0.03 (0.16) −0.01 (0.07) 0.00 (0.07) Occupation (ref: Farmer ) Unemployed/Other −0.32 (0.36) −0.27 (0.36) −0.81 (0.17) * −0.83 (0.18) * Unskilled worker 0.48 (0.32) 0.52 (0.32) −0.05 (0.17) −0.05 (0.17) Skilled worker/Professional 0.34 (0.37) 0.38 (0.37) −0.21 (0.19) −0.20 (0.19) Self-rated health (ref: Poor/Fair ) Good 0.47 (0.30) 0.48 (0.30) 0.47 (0.17) 0.49 (0.17) Excellent 0.16 (0.34) 0.15 (0.34) 0.59 (0.19) 0.61 (0.19) * Region (ref: South ) Northeast 1.18 (0.39) 1.22 (0.39) 0.23 (0.21) 0.22 (0.21) Coastal 0.96 (0.34) 0.93 (0.34) 0.02 (0.15) 0.01 (0.15) Inland 1.07 (0.31) * 1.11 (0.32) * 0.22 (0.13) † 0.22 (0.13) † Constant −7.35 (1.52) * −8.67 (1.64) * −4.39 (0.73) * −4.97 (0.80) * N 919 919 1749 1749 Note: BMI = body mass index; robust standard errors are shown in parentheses. † p < .1 * p < .05 p < .01 * p < .001. Table 3 Estimated coefficients from discrete-time logit models of marriage entry on body shape (continuous measure) Women Men Model 1 Model 2 Model 1 Model 2 Urban (ref: Rural) −0.65 (0.28) * 5.51 (2.85) † −0.31 (0.15) * −0.90 (1.84) WHR −0.02 (0.02) 0.00 (0.02) −0.01 (0.01) −0.02 (0.01) Urban × WHR −0.08 (0.04) * 0.01 (0.02) Age (years) 0.21 (0.04) * 0.21 (0.04) * 0.13 (0.02) * 0.13 (0.02) * Cohort ( ref: 1971-1980) <=1970 −0.21 (0.44) −0.22 (0.44) −0.23 (0.21) −0.23 (0.21) >=1981 −0.44 (0.33) −0.44 (0.33) −0.43 (0.16) −0.43 (0.16) Years of education −0.03 (0.04) −0.03 (0.04) −0.03 (0.03) −0.03 (0.03) Household income per capita (logged) 0.07 (0.17) 0.07 (0.18) −0.11 (0.07) −0.11 (0.07) Occupation (ref: Farmer ) Unemployed/Other −0.63 (0.40) −0.57 (0.40) −0.80 (0.18) * −0.80 (0.18) * Unskilled worker 0.27 (0.37) 0.33 (0.37) 0.01 (0.17) 0.01 (0.17) Skilled worker/Professional 0.20 (0.41) 0.24 (0.41) 0.05 (0.20) 0.04 (0.20) Self-rated health (ref: Poor/Fair ) Good 0.32 (0.30) 0.34 (0.30) 0.53 (0.17) 0.53 (0.17) Excellent 0.23 (0.35) 0.22 (0.35) 0.70 (0.20) * 0.70 (0.20) * Region (ref: South ) Northeast 0.96 (0.44) * 0.90 (0.44) * 0.32 (0.23) 0.32 (0.23) Coastal 0.92 (0.36) ** 0.90 (0.36) * 0.13 (0.17) 0.13 (0.17) Inland 0.87 (0.33) * 0.87 (0.33) * 0.33 (0.14) * 0.33 (0.15) * Constant −5.23 (2.00) −7.26 (2.22) * −1.21 (1.02) −1.05 (1.15) N 609 609 1409 1409 Note: WHR = waist-to-hip ratio, rescaled from 0-1 to 0-100 by multiplying 100; robust standard errors are shown in parentheses. † p < .1 * p < .05 p < .01 * p < .001. Table 4 Estimated coefficients from discrete-time logit models of marriage entry on body weight (categorical measure) Women Men Model 1 Model 2 Model 1 Model 2 Urban (ref: Rural) −0.51 (0.25) * −0.40 (0.27) −0.28 (0.14) * −0.18 (0.16) Overall body weight (ref: Normal) Underweight −0.29 (0.32) −0.39 (0.43) 0.10 (0.16) 0.07 (0.21) Overweight/Obese −0.59 (0.35) † −0.23 (0.38) 0.31 (0.15) * 0.46 (0.18) * Urban × Underweight 0.22 (0.63) 0.05 (0.34) Urban × Overweight/Obese −1.31 (0.87) −0.45 (0.30) Age (years) 0.27 (0.04) * 0.28 (0.04) * 0.16 (0.02) * 0.16 (0.02) * Cohort ( ref: 1971-1980) <=1970 −0.76 (0.30) * −0.77 (0.29) −0.43 (0.14) −0.44 (0.14) >=1981 −0.11 (0.32) −0.12 (0.33) −0.31 (0.17) † −0.33 (0.17) † Years of education −0.03 (0.04) −0.03 (0.04) −0.02 (0.02) −0.02 (0.02) Household income per capita (logged) 0.04 (0.17) 0.05 (0.17) −0.01 (0.07) −0.01 (0.07) Occupation (ref: Farmer ) Unemployed/Other −0.27 (0.36) −0.25 (0.36) −0.80 (0.17) * −0.81 (0.17) * Unskilled worker 0.54 (0.32) † 0.58 (0.32) † −0.04 (0.17) −0.04 (0.17) Skilled worker/Professional 0.39 (0.38) 0.41 (0.37) −0.22 (0.19) −0.22 (0.19) Self-rated health (ref: Poor/Fair ) Good 0.45 (0.30) 0.45 (0.30) 0.48 (0.17) 0.48 (0.17) * Excellent 0.12 (0.34) 0.10 (0.34) 0.60 (0.19) * 0.61 (0.19) * Region (ref: South ) Northeast 1.20 (0.39) 1.25 (0.39) * 0.22 (0.21) 0.21 (0.21) Coastal 0.95 (0.34) 0.93 (0.34) −0.02 (0.15) −0.02 (0.15) Inland 1.07 (0.31) * 1.11 (0.32) * 0.20 (0.13) 0.20 (0.13) Constant −8.59 (1.32) * −8.78 (1.35) * −4.07 (0.66) * −4.08 (0.66) *** N 919 919 1749 1749 Note: Robust standard errors are shown in parentheses. † p < .1 * p < .05 p < .01 * p < .001. Table 5 Estimated coefficients from discrete-time logit models of marriage entry on body shape (categorical measure) Women Men Model 1 Model 2 Model 1 Model 2 Urban (ref: Rural) −0.65 (0.29) * −0.49 (0.30) −0.30 (0.15) * −0.29 (0.16) † Central obese (ref: No) −0.15 (0.31) 0.13 (0.36) −0.28 (0.18) −0.27 (0.21) Urban × Central obese −0.97 (0.70) −0.05 (0.38) Age (years) 0.20 (0.04) * 0.20 (0.04) * 0.13 (0.02) * 0.13 (0.02) * Cohort ( ref: 1971-1980) <=1970 −0.17 (0.44) −0.20 (0.44) −0.23 (0.21) −0.23 (0.21) >=1981 −0.44 (0.33) −0.45 (0.33) −0.43 (0.16) −0.43 (0.16) Years of education −0.03 (0.04) −0.04 (0.04) −0.03 (0.03) −0.03 (0.03) Household income per capita (logged) 0.09 (0.17) 0.09 (0.18) −0.11 (0.07) −0.11 (0.07) Occupation (ref: Farmer ) Unemployed/Other −0.62 (0.40) −0.57 (0.40) −0.81 (0.18) * −0.81 (0.18) * Unskilled worker 0.29 (0.36) 0.31 (0.37) 0.02 (0.17) 0.02 (0.17) Skilled worker/Professional 0.26 (0.41) 0.29 (0.41) 0.04 (0.20) 0.04 (0.20) Self-rated health (ref: Poor/Fair ) Good 0.32 (0.30) 0.35 (0.30) 0.52 (0.17) 0.52 (0.17) Excellent 0.25 (0.35) 0.25 (0.35) 0.69 (0.20) * 0.69 (0.20) * Region (ref: South ) Northeast 0.97 (0.44) * 0.96 (0.44) * 0.33 (0.24) 0.33 (0.24) Coastal 0.93 (0.36) ** 0.90 (0.36) * 0.13 (0.17) 0.13 (0.17) Inland 0.87 (0.33) ** 0.85 (0.33) * 0.33 (0.14) * 0.33 (0.15) * Constant −6.79 (1.40) * −6.85 (1.41) * −2.43 (0.69) * −2.43 (0.69) * N 609 609 1409 1409 Note: Robust standard errors are shown in parentheses. † p < .1 * p < .05 p < .01 * p < .001. References Allon N Wolman BB The stigma of overweight in everyday life. Psychological Aspects of Obesity: A Handbook 1982 130 174 Von Nostrand Reinhold New York Amin S Eurocentrism 1989 Monthly Review Press New York Averett S Korenman S Black white differences in social and economic consequences of obesity. International Journal of Obesity 1999 23 166 173 10078852 Blaut JM The Colonizer's Model of the World: Geographical Diffusionism and Eurocentric History 1993 Guilford New York Buss DM Evolutionary Psychology: The New Science of the Mind 2004 Allyn and Bacon Boston Carr D Friedman MA Is obesity stigmatizing? Body weight, perceived discrimination, and psychological well-being in the United States. Journal of Health and Social Behavior 2005 46 244 259 16259147 Cartier A Côté M Lemieux I Sex differences in inflammatory markers: What is the contribution of visceral adiposity? The American Journal of Clinical Nutrition 2009 89 1307 1314 19297456 Cawley J The impact of obesity on wages. The Journal of Human Resources 2004 39 451 474 Chen F Yang Y Liu G Social change and socioeconomic disparities in health over the life course in China: A cohort analysis. American Sociological Review 2010 75 126 150 20379373 Crandall CS Do parents discriminate against their heavyweight daughters? Personality and Social Psychology Bulletin 1995 21 724 735 Crosnoe R Muller C Body mass index, academic achievement, and school context: Examining the educational experiences of adolescents at risk of obesity. Journal of Health & Social Behavior 2004 45 393 407 15869112 Cui Z Huxley R Wu Y Temporal trends in overweight and obesity of children and adolescents from nine provinces in China from 1991–2006. International Journal of Pediatric Obesity 2010 5 365 374 20836722 Davis D Harrell S Chinese families in the Post-Mao era 1993 University of California Press Berkeley Dixson BJ Dixson AF Li B Studies of human physique and sexual attractiveness: Sexual preferences of men and women in China. American Journal of Human Biology 2007 19 88 95 17160976 Doak CM Adair LS Bentley M The underweight/overweight household: An exploration of household sociodemographic and dietary factors in China. Public Health Nutrition 2002 5 215 221 12027287 Doak CM Adair LS Bentley M The dual burden household and the nutrition transition paradox. International Journal of Obesity 2005 29 129 136 15505634 Dornbusch SM Carlsmith JM Duncan PD Sexual maturation, social class, and the desire to be thin among adolescent females. Journal of Developmental & Behavioral Pediatrics 1984 5 308 314 6529455 Du S Lu B Zhai F A new stage of the nutrition transition in China. Public Health Nutrition 2002 5 169 174 12027281 Finucane MM Stevens GA Cowan MJ National, regional, and global trends in body mass index since 1980: Systematic analysis of health examination surveys and epidemiological studies with 960 country-years and 9.1 million participants. The Lancet 2011 377 557 567 Fitzmaurice GM Laird NM Ware JH Applied Longitudinal Analysis 2004 John Wiley & Sons, Inc. New Jersey Flegal KM Carroll MD Kit BK Prevalence of obesity and trends in the distribution of body mass index among US adults, 1999-2010. Journal of the American Medical Association 2012 307 491 497 22253363 Fu H Goldman N Incorporating health into models of marriage choice: Demographic and sociological perspectives. Journal of Marriage and the Family 1996 58 740 758 Gan L Findings from the China Household Finance Survey 2012 Southwestern University of Finance and Economics Chengdu, China Gelman A Hill J Data Analysis Using Regression and Multilevel/Hierarchical Models 2007 Cambridge University Press New York Goldscheider C Population, Modernization, and Social Structure 1971 Little, Brown Boston Gortmaker SL Must A Perrin JM Social and economic consequences of overweight in adolescence and young adulthood. New England Journal of Medicine 1993 329 1008 1012 8366901 Greenhalgh S The social construction of population science: An intellectual, institutional, and political history of twentieth-century demography. Comparative Studies in Society and History 1996 38 26 66 Greenhalgh S Just One Child: Science and Policy in Deng's China 2008 University of California Press Berkeley, CA Han H Trends in educational assortative marriage in China from 1970 to 2000. Demographic Research 2010 22 733 770 Haslam DW James WPT Obesity. The Lancet 2005 366 1197 1209 Haworth-Hoeppner S The critical shapes of body image: The role of culture and family in the production of eating disorders. Journal of Marriage and the Family 2000 62 212 227 Holmgren J The economic foundations of virtue: Widow-remarriage in early and modern China. The Australian Journal of Chinese Affairs 1985 1 27 Hou X Urban-rural disparity of overweight, hypertension, undiagnosed hypertension, and untreated hypertension in China. Asia-Pacific Journal of Public Health 2008 20 159 169 19124310 Hu FB Liu Y Willett WC Preventing chronic diseases by promoting healthy diet and lifestyle: Public policy implications for China. Obesity Reviews 2011 12 552 559 21366840 Jæger MM “A thing of beauty is a joy forever”? Returns to physical atractiveness over the life course. Social Forces 2011 89 983 1004 Jeffery RW Rick AM Cross-sectional and longitudinal associations between body mass index and marriage-related factors. Obesity Research 2002 10 809 815 12181390 Ji Y Between tradition and modernity: “Leftover” women in Shanghai. Journal of Marriage and Family 2015 77 1057 1073 Jones GW Yeung W-JJ Marriage in Asia. Journal of Family Issues 2014 35 1567 1583 Karastergiou K Smith S Greenberg A Sex differences in human adipose tissues— the biology of pear shape. Biology of Sex Differences 2012 3 13 22651247 Lai Q Thornton A The making of family values: Developmental idealism in Gansu, China. Social Science Research 2015 51 174 188 25769860 Lassek WD Gaulin SJC Waist-hip ratio and cognitive ability: Is gluteofemoral fat a privileged store of neurodevelopmental resources? Evolution and Human Behavior 2008 29 26 34 Lesthaeghe R The unfolding story of the second demographic transition. Population and Development Review 2010 36 211 251 20734551 McLaren L Socioeconomic status and obesity. Epidemiologic Reviews 2007 29 29 48 17478442 Morris S Body mass index and occupational attainment. Journal of Health Economics 2006 25 347 364 16290234 Mu Z Xie Y Marital age homogamy in China: A reversal of trend in the reform era? Social Science Research 2014 44 141 157 24468440 Najman JM Klein D Munro C Patient characteristics negatively stereotyped by doctors. Social Science & Medicine 1982 16 1781 1789 7178923 National Bureau of Statistics of China China Statistical Yearbook 2011 2011 China Statistics Press Beijing Neumark-Sztainer D Story M Faibisch L Perceived stigmatization among overweight African-American and caucasian adolescent girls. Journal of Adolescent Health 1998 23 264 270 9814386 Pan J Qin X Liu G Does body size matter in urban employment? Evidence from China. Nankai Economics Studies 2011 164 68 85 Pinhey TK Rubinstein DH Colfax RS Overweight and happiness: The reflected self-appraisal hypothesis reconsidered. Social Science Quarterly 1997 78 747 755 Popkin BM The shift in stages of the nutrition transition in the developing world differs from past experiences! Public Health Nutrition 2002 5 205 214 12027286 Popkin BM Du S Zhai F Cohort profile: The China Health and Nutrition Survey - monitoring and understanding socio-economic and health change in China, 1989-2011. International Journal of Epidemiology 2010 39 1435 1440 19887509 Puhl R Brownell KD Bias, discrimination, and obesity. Obesity Research 2001 9 788 805 11743063 Qian Y Qian Z The gender divide in urban China: Singlehood and assortative mating by age and education Demographic Research 2014 31 1337 1364 Raymo JM Park H Xie Y Marriage and family in East Asia: Continuity and change. Annual Review of Sociology 2015 41 null Schafer MH Kwon S Cohorts and emerging health disparities: Biomorphic trajectories in China, 1989 to 2006. Journal of Health and Social Behavior 2012 53 515 532 23041792 Schiller JS Lucas JW Ward BW Summary Health Statistics for U.S. Adults: National Health Interview Survey, 2010 2012 National Center for Health Statistics Washington, DC Schwartz MB Chambliss HON Brownell KD Weight bias among health professionals specializing in obesity. Obesity Research 2003 11 1033 1039 12972672 Sobal J Rauschenbach BS Frongillo EA Marital status, fatness and obesity. Social Science & Medicine 1992 35 915 923 1411692 Sobal J Stunkard AJ Socioeconomic status and obesity: A review of the literature. Psychological Bulletin 1989 105 260 275 2648443 Swami V Caprario C Tovée MJ Female physical attractiveness in Britain and Japan: a cross-cultural study. European Journal of Personality 2006 20 69 81 Thornton A The developmental paradigm, reading history sideways, and family change. Demography 2001 38 449 465 11723944 Thornton A Reading History Sideways: The Fallacy and Enduring Impact of the Developmental Paradigm on Family Life 2005 University of Chicago Press Chicago Thornton A International family change and continuity: The past and future from the developmental idealism perspective. Demográfia 2010 53 21 50 Thornton A Ritzer G Developmental idealism. The Wiley-Blackwell Encyclopedia of Globalization 2012 John Wiley & Sons, Ltd. Hoboken, NJ Thornton A Binstock G Yount K International fertility change: New data and insights from the developmental idealism framework. Demography 2012 49 677 698 22403035 Tilly C Big Structures, Large Processes, Huge Comparisons 1984 Russell Sage Foundation New York Veblen T Theory of the Leisure Class 1889 Kelley Fairfield, NJ Wang F Yang Q Age at marriage and the first birth interval: The emerging change in sexual behavior among young couples in China. Population and Development Review 1996 22 299 320 Whyte MK One Country, Two Societies. Harvard Contemporary China Series 2010 Harvard University Press Cambridge, Massachusetts World Health Organizaion, International Association for the Study of Obesity, and International Obesity TaskForce The Asia-Pacific perspective: Redefining obesity and its treatment 2000 Sydney, Australia World Health Organization Waist Circumference and Waist-Hip Ratio: Report of a WHO Expert Consultation 2008 World Health Organization Geneva Xi B Liang Y He T Secular trends in the prevalence of general and abdominal obesity among Chinese adults, 1993–2009. Obesity Reviews 2012 13 287 296 22034908 Xie Y The User's Manual of the China Family Panel Studies (2010) 2012 Institute of Social Science Survey, Peking University Beijing Xie Y Thornton A Wang G Societal projection: Beliefs concerning the relationship between development and inequality in China. Social Science Research 2012 41 1069 1084 23017918 Xu H Short SE Health insurance coverage rates in 9 provinces in China doubled from 1997 to 2006, with a dramatic rural upswing. Health Affairs 2011 30 2419 2426 22147871 Xu H Short SE Liu T Dynamic relations between fast-food restaurant and body weight status: a longitudinal and multilevel analysis of Chinese adults. Journal of Epidemiology & Community Health 2013 67 271 279 22923769 Xue Y China's young adults still hold traditional values toward marriage. Chinese Social Sciences Today 2013 from http://www.csstoday.net/Item/41605.aspx Feburary 3, 2016 Yang G Wang Y Zeng Y Rapid health transition in China, 1990–2010: Findings from the Global Burden of Disease Study 2010. The Lancet 2013 381 1987 2015 Yu J Xie Y Changes in the determinants of marriage entry in post-reform urban China. Demography 2015a 52 1869 1892 26391287 Yu J Xie Y Cohabitation in China: Trends and determinants. Population and Development Review 2015b 41 607 628 Zhao Z Income inequality, unequal health care access, and mortality in China. Population and Development Review 2006 32 461 483 Zimmer Z Kaneda T Spess L An examination of urban versus rural mortality in China using community and individual data. The Journals of Gerontology: Psychological Sciences and Social Sciences 2007 62B S349 S357
PMC005xxxxxx/PMC5125438.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 100890393 21677 Org Lett Org. Lett. Organic letters 1523-7060 1523-7052 25723256 5125438 10.1021/acs.orglett.5b00068 NIHMS823356 Article Gargantulide A, a Complex 52-Membered Macrolactone Showing Antibacterial Activity from Streptomyces sp Rho Jung-Rae †⊥ Subramaniam Gurusamy ‡# Choi Hyukjae §▽ Kim Eun-Hee ‖ Ng Sok Peng ‡ Yoganathan K. ‡○ Ng Siewbee ‡○ Buss Antony D. ‡ Butler Mark S. ‡■ Gerwick William H. †§ † Center for Marine Biotechnology and Biomedicine, Scripps Institution of Oceanography, University of California, San Diego, La Jolla, California, United States ‡ MerLion Pharmaceuticals, 41 Science Park Road, #04-03B the Gemini, Singapore Science Park II, Singapore 117610, Singapore § Skaggs School of Pharmacy and Pharmaceutical Sciences, University of California, San Diego, La Jolla, California, United States ‖ Division of Magnetic Resonance, Korea Basic Science Institute, Ochang, Chungbuk 363-883, Korea * Corresponding Authors: [email protected].; [email protected] ⊥ Present Address: J.-R.R.: Department of Marine Biotechnology, Kunsan National University, Jeonbuk 573-701, Korea. # Present Address: G.S.: School of Chemical and Life Sciences, Nanyang Polytechnic, Singapore 569830, Singapore. ▽ Present Address: H.C.: Department of Pharmacy, Yeungnam University, Gyeongsangbuk-do 712-749, Korea. ○ Present Address: S.N. and K.Y.: Bioinformatics Institute, Agency for Science, Technology and Research (A*STAR), 138671, Singapore. ■ Present Address: M.S.B.: Institute for Molecular Bioscience, University of Queensland, St. Lucia 4072, Queensland, Australia. 23 11 2016 27 2 2015 20 3 2015 28 11 2016 17 6 13771380 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Gargantulide A (1), an extremely complex 52-membered macrolactone, was isolated from Streptomyces sp. A42983 and displayed moderate activity against MRSA. The planar structure of 1 was determined using 2D NMR, and its stereochemistry was partially established on the basis of NOESY correlations, J-based configuration analysis, and Kishi’s universal NMR database. Graphical abstract The bacterial membrane protein ATPase SecA facilitates the transfer of preproteins across the cell membrane and is essential for bacterial survival.1,2 There have been various SecA inhibitors identified, but none have advanced to clinical trials.2,3 A 384-well high throughput bioassay that used BIOMOL Green to measure free phosphate (Pi) release during the transport of SecA preprotein, proOmpA, through the SecYEG complex located on inverted membrane vesicles was developed to identify SecA inhibitors. An active MeOH extract derived from Streptomyces sp. A42983 was identified that displayed moderate activity in the SecA assay and whole cell activity against methicillin-resistant Staphylococcus aureus (MRSA) ATCC25923. Bioassay-guided fractionation led to the isolation of a novel and exceptionally complex macrolactone, gargantulide A (1), possessing a 52-membered ring. Gargantulide A (1) was isolated as a white powder (Figure 1) and was assigned a molecular formula of C105H200N2O38 based on the high resolution ESI-FT-MS ([M + 2H]2+ = 1049.6974, Δ = 1.2), which was consistent with seven degrees of unsaturation. The 1H NMR spectrum showed severely overlapped resonances from 0.5–5.5 ppm due to the presence of numerous oxymethines, methylenes associated with saturated hydrocarbon chains, and methyl groups. The 13C NMR spectrum similarly suffered from extensive signal degeneracy, but clearly contained one ketone group (δC 214.4), one ester or amide-type carbonyl (δC 175.1), two olefinic carbons (δC 123.3 and 146.5), and three anomeric carbons (δC 96.8, 102.3 and 104.3) as identified by their typical 13C NMR chemical shifts. Fortunately, a multiplicity-edited HSQC experiment allowed us to assign nearly all carbon signals to their corresponding proton(s) and resulted in the identification of 50 methine (including 39 oxymethine), 38 methylene, and 14 methyl carbons. Further HSQC analysis allowed identification of five partially overlapped carbon resonances near δC 34.0 [δC 34.22 (C-5, CH2) – δH 1.19/1.48; δC 34.01 (C-13, CH2) – δH 1.08/1.76; δC 34.04 (C-67, CH2) – δH 1.13/1.51; δC 34.07 (C-68, CH) – δH 1.43; δC 34.18 (C-1′, CH2) – δH 1.07/1.35] as well as three oxymethines in the cluster of δC 66.2 and δH 4.10 which were deduced from the molecular formula and proton integration data. Elucidation of the structure of gargantulide A (1) was initiated by considering the strong HMBC correlations between the protons of 14 methyl groups and their nearby carbons, thus allowing the identification of 12 subunits (A–L) as shown in Figure 2. Nine subunits were nicely revealed by well-separated signals in the expanded HMBC spectrum, and these assignments were confirmed by COSY and TOCSY data. Subunits B, I, and L which contained the nearly indistinguishable carbon resonances around δC 34.0 could not be unambiguously assigned by HMBC, but the separated resonances of their attached protons allowed these latter NMR resonances to be confidently assigned by the combination of COSY, TOCSY, and HSQC-TOCSY data (Figures S6–S14). Next, the structures of the two conventional pyranose residues were established as β-glucose (Glc) and β-mannose (Man) in a straightforward manner using 3JHH coupling constants from DQF-COSY, along with TOCSY and NOE correlations. Furthermore, these sugar components, cleaved by acid hydrolysis, were recognized as D-Glc and D-Man by comparison of retention times with the authentic sugars using GC/MS of their trimethylsilyl derivatives (Figure S1). One remaining sugar moiety was elucidated as 3,6-deoxy-3-methylamino pyranose by sequential COSY correlations from an anomeric proton to a methyl doublet, and HMBC correlations between subunit K of a methylamino group (δC 31.4; δH 2.78) and maG-3 carbon, and the anomeric proton and maG-5 carbon. The large coupling constants between protons in this moiety and NOEs of maG-1/maG-3, maG-1/maG-5, and mag-2/maG-4 led to 3,6-deoxy-3-methylamino glucose (maG). Following the assignment of the three sugar components, the linkage of the subunits defined above was accomplished by intensive inspection of 2D NMR spectra (COSY, TOCSY, HSQC, HMBC, and HSQC-TOCSY). As detailed below, HSQC-TOCSY correlations from oxymethine protons or other well-separated protons to neighboring carbons verified the connection of the subunits as shown in Figure 2. COSY and HSQC-TOCSY correlations from H2-2 allowed connection to H-5 in subunit A and allowed assignments of 13C resonances for C-3 to C-6. The subunits A and B were connected by two nearly overlapping resonances for C-7 (δC 32.7) and C-9 (δC 32.4), indicated by COSY correlation of H-7/H-8 and HSQC-TOCSY correlations from H-7 to C-10. Further connection of this spin system from H-13 in subunit B was prevented by TOCSY correlations overlapped with signals associated with other subunits. On the other hand, the linkages of subunits C, D, and E were readily determined by COSY and HMBC correlations due to distinct resonances as shown in Figure 2. Fortunately, the same resonances were located in the two TOCSY correlation sets propagating from each terminal proton of subunits B and C. HSQC-TOCSY correlations allowed assignment of three aliphatic methylene carbons between subunits B and C (C-14 to C-16). In a similar way, the methine H-47, which was recognized by the COSY and TOCSY correlations with H-46 in subunit F, played a crucial role in the connectivity of subunits F and G by the HSQC-TOCSY correlations. Successive connection from subunit G could be conducted by the extensive HMBC correlations of H-51/C-1, H-51/C-52, H-51/C-53, and H-51/C-1′. The assignment of the overlapped C-1′ resonance in subunit L was confirmed by NOE correlation between the corresponding proton and H-51 multiplet. Linkage of subunit G to H was evident from a combination of COSY correlations between two relatively distinct protons (Ha-53 and Hb-54), and weak HSQC-TOCSY correlations from the H-55 multiplet to the C-53 and C-54 resonances. Following subunit H, the consecutive methylene-oxymethine groups from C-58 to C-62 were apparent from a mutual sharing of TOCSY and HSQC-TOCSY correlations with three oxymethine 1H signals (H-57, H-59, and H-61). HSQC-TOCSY correlations associated with yet another oxymethine proton, H-65, enabled connection of subunits H and I. The terminal amino propyl moiety was attached to the distal side of subunit I by HSQC-TOCSY correlations starting with H-72. The assignment of the C-29 to C-33 segment adjacent to subunit E was enabled by three common HSQC-TOCSY correlations initiated from two different protons (H-29 and H-33). Their carbon chemical shifts were similar to those of C-62 to C-64 which were also involved in a 1,5-diol unit. However, extending COSY or TOCSY correlations from the H-33 signal was not possible because of both 1H and 13C spectral overlap. Interestingly, these overlapped signals corresponded to three unassigned oxymethines (around δC 66.2; δH 4.10) and three methylenes (around δC 46.5; δH 1.56), reminiscent of chemical shift and functionality to the alternating C-57 to C-62 methylene-oxymethine section. The insertion of these alternating groups between C-33 and C-40 was subsequently confirmed by COSY and TOCSY spectra and allowed connection between subunits E and F. Finally, HMBC correlation between the signal for H-2 and an ester carbonyl C-1 (δC 175.1), and placement of the three sugar components based on HMBC correlations as given in Figure 2, completed the planar structure of 1 as a triglycosylated 52-membered macrolactone, thereby accounting for the 7 degrees of unsaturation. Gargantulide A (1) contains 34 stereocenters on the macrolactone ring and the side chain, together with three different and fully substituted pyranose sugar residues, for a total of 49 chiral centers. A combination of NOE (NOESY) and coupling constant analysis (DQF-COSY and HECADE) was used to establish the configurations of the relatively rigid segments close to the three sugar residues (C-21–27 and C-47–57). Extending from the two stereochemically defined sugars (Man and Glc), the configurational assignment of the 1,3-diol (C-27–C-29), 1,3,5-triol (C-45–C-49 and C-57–C-61), 1,3,5,7-pentol (C-33–C-41) and a contiguous hydroxy/methyl/hydroxy/methyl set of substituents in the flexible chain (C-41–45) could be successfully accomplished by application of Kishi’s universal NMR database.4–7 Based on the stereochemically defined β-D-mannose unit, NOE correlations between Man-1/C-24 and Man-5/Me-22 allowed assignment of the absolute configuration of C-24 as S. Strong NOE correlations of H-23/H-25, H-23/H26, Me-22/H-24, Me-22/H-27, and H-24/H-27 allowed assignment of configurations for the C-23 to C26 segment (Figure 3A), and this was additionally supported by homo- and heteronuclear coupling constants between these proximal atoms.5 A small coupling constant between H-27 and Ha-28 (JHH ∼ 2.4 Hz) and an NOE correlation between H-26 and Ha-28 defined the configuration of C-27 as R. The configuration of C-29 could then be assigned as S by its carbon chemical shift value which matched typical values in Kishi’s NMR database for an oxymethine carbon in a 1,3-anti diol relationship [δC = 69.2, in CD3OD].6 Accordingly, the stereochemistry of the C-23 to C-29 fragment was completely assigned as 24S, 25S, 26S, 27R, and 29S. Additionally, a weak NOE signal between Man-1 and H-20 (appearing consistently in three repeated NOESY experiments) along with a strong NOE between H-21 and H-23 allowed deduction of C-21 as S. In a similar manner, the absolute configuration of C-49 was defined as R based on strong NOE correlations between Glc-1/H-49, Glc-1/H-50, and weak NOE correlations of Glc-2/H-47 and Glc-5/H-50. Despite the fact that the chemical shifts of Glc-2 (δH 3.18) and Glc-5 (δH 3.20) were nearly isochronous, their NOE correlations could be distinguished from one another by NOESY. Additional NOE correlations of H-49/H-50, H-49/H-51, H-50/Hb-53, H-51/Ha-48 and H-51/H-1′, H-52/Me-50 led to assignments of 47R, 49R, 50R, 51S, and 52S. Homo- and heteronuclear couplings also supported this determination as shown in Figure 3B. In particular, the S configuration of H-47 was supported by an NOE correlation with H-49 and a small coupling constant with Hb-48; this latter signal was present in a gauche orientation with the hydroxy group on C-47 as revealed by a large heteronuclear coupling with C-47. Furthermore, a large coupling constant between the well-resolved protons Hb-53 and Hb-54, and NOE interaction between Ha-53/H-1′, allowed deduction of the configuration at C-52. Small heteronuclear coupling constants between Ha-53/C-55 and Hb-53/C-55 indicated a gauche relationship with C-55 and these two protons, and also strong NOE correlations of H-55/Hb-54, Me-56/Ha-54, Me-56/Hb-54, maG-1/H-55, maG-1/H-56, and maG-1/H-57 defined C-55 as R (Figure 3B). Based on this configuration, the stereochemistry of the next carbon C-56 was apparent from J-based configuration analysis for the pair of interconverting rotamers (Figure 3C), revealing a 56R configuration.8 In turn, the configuration of C-57 could be established, as for C-27, by NOE correlation of H-56/Ha-58, the latter of which was gauche to H-57 (3JH58a-H57 = 2.2 Hz). Similarly, the configuration of the oxymethine carbon C-59 was assigned anti to C-57 by J-based analysis (Figure 3D). The configuration of C-61 as syn with C-59 was deduced by Kishi’s NMR database for a 1,3,5-triol moiety: the assignment of the carbon chemical shift of C-59 at δC 68.8 (CD3OD) indicated that it must possess either an anti/syn or syn/anti configuration.6 Additionally, the configuration of C-65 was suggested by the empirical rule for 1,5-diol systems in which the two central methylene protons are equivalent in 1,5-anti diols and nonequivalent in 1,5-syn diols.9 Because the H-63 signals were nonequivalent (δH 1.38 and 1.61), this implied that C-61 and C-65 are in a syn orientation, thus assigning C-65 as S. On the other hand, the maG pyranose moiety attached to C-55 could be established as L by weak NOE interactions of maG-1/H-57, maG-5/H-55, and maG-2/H-57 as well as a stronger NOE between maG-1 and H-55. Next, relative configurations for the C-33 to C-45 segment were assigned by Kishi’s NMR database values. On the basis of the 45, 47-syn diol moiety identified from the carbon chemical shift of C-47 (δc 71.42), the configuration of the C-45 to C-41 hydroxy/methyl/hydroxy/methyl/hydroxy sequence was determined following a previously reported procedure.10 In graphs comparing carbon chemical shifts of C-44 and Me-44 with the corresponding Kishi’s NMR database values (1a–h, s), the possible configuration of the first hydroxy/methyl/hydroxy/methyl sequence [C-45 to C-42 with Me-44] was indicated as 1a (ααββ), 1e (βαββ), 1f (βααα), 1g (βαβα), or 1h (βααβ). By applying the same method to the second hydroxy/methyl/hydroxy/methyl sequence [C-44 to C-41 with Me-42], stereoisomer 2d (αβββ) was selected as the best match. By combining these two results, the relative configuration of the C-45 to C-41 section was identified as either ααβββ or βαβββ. Of these two, the ααβββ possibility for C-45 to C-41 was preferred by measurement of small coupling constants for 3JH44–H45 and, 2JH44–H45, and a strong HMBC correlation of H-45/Me-44, thus indicating a threo relationship between the C-45 hydroxy group and Me-44. Furthermore, the relative tetraol configuration in the C-33 to C-41 section was assignable by the distinctive chemical shifts of the central carbons in the two overlapping 1,3,5-triol units. These depend on the relative orientation of the three hydroxy groups in each triol [δC 70.4 for 1,3-syn/3,5-syn; δC 66.3 for 1,3-anti/3,5-anti; δC 68.3 for 1,3-anti/3,5-syn or 1,3-syn/3,5-anti in CD3 OD].4 Comparison of these NMR database values with those observed (C-35 δC 66.35; C-39 δC 66.4) suggested a continuous anti/anti configuration in the C-33 to C-41 segment. Using the empirical rule described above for 1,5-diols, relay of this stereochemical information to C-29 was possible through observation of the equivalence of the methylene protons at H-31, thus suggesting a 29,33-anti-diol substructure. Finally, even though C-47 is the central carbon of a modified 1,3,5-triol moiety, its 13C NMR chemical shift is very close to that for the syn/syn configuration noted above, indicating a syn relationship to C-45, and thus completing the configurational assignments for the C-29 to C-51 segment. Compared with monazomycin,10 desertomycin,11 and mathemycin,12 glycosylated macrolactones with an amino group at their termini, gargantulide A is larger in ring size and features an unusual amino sugar residue. Gargantulide A (1) had an IC50 of 8 μg/mL in the HTS assay that measured the release of phosphate using BIOMOL Green from the reaction of inverted membrane vesicles, proOmpA, a precursor protein, and the SecA enzyme. In a secondary screening, 1 was not active at concentrations up to 160 μg/mL against the ATPase domain of SecA and the chaperone GroEL, which indicated that 1 might not be specifically inhibiting SecA. Gargantulide A (1) had an MIC of 2 μg/mL against S. aureus ATCC25923 that was not affected by the presence of 40% horse serum (MIC 1 μg/mL). Although 1 was not active against the E. coli DC2 (MIC >32 μg/mL), which is a mutant deficient in osmoregulation of periplasmic oligosaccharide synthesis, activity was rescued against an E. coli Imp (MIC 1 μg/mL), which is a mutant with increased membrane permeability. Gargantulide A (1) also had an MIC of 0.5 μg/mL against Clostridium difficile 6196 HMC and 4 μg/mL against C. difficile 6671 HMC. Gargantulide A (1) was profiled for antifungal activity and was found to be weakly active against Candida albicans ATCC90028 (IC50 64 μg/mL) and inactive against Saccharomyces cerevisiae SKY54 (MIC >128 μg/mL). No hemolysis was observed at concentrations up to 200 μg/mL. Overall this was a promising profile for a Gram-positive antibiotic, strengthened by the observation of excellent activity against multiple strains of methicillin-resistant S. aureus (MRSA), methicillin-resistant S. epidermidis (MRSE), vancomycin-resistant enterococci (VRE), and penicillin-resistant Streptococcus pneumonia (PRSP) (Table S2). Given this biological profile, the high molecular weight, and other physicochemical properties of gargantulide A (1), it was envisaged that 1 might become an intravenously dosed Gram-positive antibiotic or an orally administered treatment for the anaerobe C. difficile. Gargantulide A (1) was administered to female CD-1 mice using a lateral tail injection; however, the mice rapidly died in a state of rigid paralysis at a dose of 5 mg/kg. Mice treated with 5 mg/kg of 1 via subcutaneous injection showed no sign of immediate distress but died within 12 h. Unfortunately, this severe toxicity precluded any further development of 1 as an antibiotic. In summary, gargantulide A (1), discovered using a combination of a SecA HTS assay and whole cell screening, was deduced as a polyketide with a 52-membered macrolactone ring. The structure of this antibiotic was exceptionally complex in that it contained 105 carbon atoms, of which nearly half were chiral centers. Although the configuration of several chiral centers remains unassigned, most were determined using diverse methodologies. This compound displayed promising activity against pathogenic Gram-positive bacteria (MRSA, MRSE, VRE, PRSP, and C. difficile) but was found to be highly toxic to mice, precluding further development. Supplementary Material Supplemental We acknowledge NIH NS053398 and CA100851 (W.H.G.) and S. R. Whitton and other members of the MerLion microbiology team for fermentation and extract generation. Figure 1 Structure of Gargantulide A (1). Figure 2 Twelve subunits and their linkages by HMBC or HSQC-TOCSY correlations. Figure 3 Crucial NOE correlations observed near the three sugar residues and corresponding J-based configuration analyses. Supporting Information Experimental procedures and spectral data of 1. This material is available free of charge via the Internet at http://pubs.acs.org. Notes The authors declare no competing financial interest. 1 (a) Rao CVS De Waelheyns E Economou A Anné J Biochim Biophys Acta 2014 1843 1762 1783 24534745 (b) Gouridis G Karamanou S Sardis MF Schärer MA Capitani G Economou A Mol Cell 2013 52 655 666 24332176 (c) Lycklama A Nijeholt JA Driessen AJ Philos Trans R Soc London, Ser B: Biol Sci 2012 367 1016 1028 22411975 2 (a) Chatzi KE Sardis MF Economou A Karamanou S Biochim Biophys Acta 2014 1843 1466 1474 24583121 (b) Segers K Anné J Chem Biol 2011 18 685 698 21700205 3 (a) Chen W Chaudhary A Cui J Jin J Hsieh Y Yang H Huang Y Tai PC Wang B J Chin Pharm Sci 2012 21 526 530 (b) Cui J Jin J Hsieh YH Yang H Ke B Damera K Tai PC Wang B ChemMedChem 2013 8 1384 1393 23794293 (c) Akula N Trivedi P Han FQ Wang N Eur J Med Chem 2012 54 919 924 22703844 4 Kobayashi Y Tan CH Kishi Y Helv Chim Acta 2000 83 2562 2571 5 Hayashi N Kobayashi Y Kishi Y Org Lett 2001 3 2249 2252 11440591 6 Kobayashi Y Hayashi N Kishi Y Org Lett 2001 3 2253 2255 11440592 7 Bifulco G Dambruoso P Gomez-Paloma L Riccio R Chem Rev 2007 107 3744 3779 17649982 8 Matsumori N Kaneno D Murata M Nakamura H Tachibana K J Org Chem 1999 64 866 876 11674159 9 Miyata Y Matsunaga S Tetrahedron Lett 2008 49 6334 6336 10 Nakayama H Furihata K Seto H Otake N Tetrahedron Lett 1981 22 5217 5220 11 Bax A Aszalos A Dinya Z Sudo K J Am Chem Soc 1986 108 8056 8063 12 Mukhopadhyay T Vijayakumar EKS Nadkarni SR Fehlhaber HW Kogler H Petry S J Antibiot 1998 51 582 585 9711222
PMC005xxxxxx/PMC5125441.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 9010000 8514 Curr Opin Lipidol Curr. Opin. Lipidol. Current opinion in lipidology 0957-9672 1473-6535 26855231 5125441 10.1097/MOL.0000000000000278 NIHMS831800 Article Expanding Role of Gut Microbiota in Lipid Metabolism Ghazalpour Anatole 1 Cespedes Ivana 23 Bennett Brian J. 45 Allayee Hooman 23 1 Department of Human Genetics, David Geffen School of Medicine of UCLA, Los Angeles, CA 90095 2 Department of Preventive Medicine, Keck School of Medicine, University of Southern California, Los Angeles, CA 90033 3 Institute for Genetic Medicine, Keck School of Medicine, University of Southern California, Los Angeles, CA 90033 4 Department of Genetics and Nutrition, University of North Carolina, Chapel Hill, NC 27599 5 Department of Nutrition Research Institute, University of North Carolina, Chapel Hill, NC 27599 Address correspondence and reprint requests to: Hooman Allayee, PhD, Institute for Genetic Medicine, Keck School of Medicine, University of Southern California, 2250 Alcazar Street, CSC202, Los Angeles, CA 90033, Phone: (323) 442-1736, Fax: (332) 442-2764, [email protected] 24 11 2016 4 2016 01 4 2017 27 2 141147 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Purpose of review This review highlights recent advances in the emerging role that gut microbiota play in modulating metabolic phenotypes, with a particular focus on lipid metabolism. Recent findings Accumulating data from both human and animal studies demonstrate that intestinal microbes can affect host lipid metabolism through multiple direct and indirect biological mechanisms. These include a variety of signaling molecules produced by gut bacteria that have potent effects on hepatic lipid and bile metabolism and on reverse cholesterol transport, energy expenditure, and insulin sensitivity in peripheral tissues. Additionally, host genetic factors can modulate the abundance of bacterial taxa, which can subsequently affect various metabolic phenotypes. Proof of causality for identified microbial associations with host lipid-related phenotypes has been demonstrated in several animal studies but remains a challenge in humans. Ultimately, selective manipulation of the gut microbial ecosystem for intervention will first require a better understanding of which specific bacteria, or alternatively, which bacterial metabolites, are appropriate targets. Summary Recent discoveries have broad implications for elucidating bacterially-mediated pathophysiological mechanisms that alter lipid metabolism and other related metabolic traits. From a clinical perspective, this newly recognized endocrine organ system can be targeted for therapeutic benefit of dyslipidemia and cardiometabolic diseases. gut microbiota lipid metabolism metabolic homeostasis Introduction It has become widely appreciated that our gut symbionts play integral roles in human health since perturbations of this bacterial community or the products they can produce have been associated with increased susceptibility to a variety of diseases (see Figure). The first indications of these associations were for colitis and inflammatory bowel disease, but altered gut microbial composition or function has now been established in the development of cardiometabolic phenotypes, including obesity and related abnormalities [1–6], and atherosclerosis [7]. There is also evidence that the microbiota can even be a potential contributor to risk of neurobehavioral conditions, such as autism [8]. In this review, we focus on recent studies that indicate an emerging role for gut microbiota in modulating lipid metabolism. Characterization of Gut Microbial Diversity The human intestinal tract is home to at least 1000 distinct species of bacteria, which collectively number over 100 trillion organisms. This diverse ecosystem is shaped by early life events but can evolve over time through interactions between its constituents as well as with exogenous factors or those that are endogenous to the host. Until recently, characterization of the gut microbiome relied mostly on conventional culture-based microbiological techniques, which was a major hindrance since the vast majority of bacteria in the gut are not readily amenable to cell culture. However, advances in next generation genomic technologies now allow us to identify and classify gut bacterial composition in an unprecedented manner. One widely used approach has typically involved in-depth sequencing of the variable regions of bacterial 16S rRNA genes to determine the diversity and proportion of bacterial taxa within the microbial community [9]. Based on the sequence data obtained, microbial richness and diversity are then organized into operational taxonomic units (OTUs). Although more challenging, recent studies have also begun to characterize microbial communities through unbiased metagenomics analyses, which involves untargeted shotgun sequencing of all genetic material recovered from the intestine or feces [10]. Both 16S and metagenomic analyses have revealed that the human gut is mostly comprised of a common core of bacteria from two major phyla, Firmicutes and Bacteriodetes, with the remainder of the gut microbiota being remarkably diverse. This diversity often includes less abundant representation from the phyla Proteobacteria, Verrumicrobia, Actinobacteria, Fusobacteria, and Cyanobacteria, as well as the domain Archaea [11]. It is also important to note that the human gut microbiome can also be dynamic and altered dramatically, for example, by antibiotic use, but less so by age, host genetics, chronic dietary patterns, and other environmental exposures [12–17]. Association of Gut Microbiota with Lipid Metabolism Early studies comparing germ free versus conventionally raised mice first supported a role for gut microbes in both affecting host energy metabolism and modulating lipid levels [18]; however, the design of these early studies did not permit identification of candidate microbes involved in promoting the observed phenotypic changes in conventionalized (microbe colonized) mice. Given the known clinical correlation between obesity, related metabolic disorders, and dyslipidemia, it is possible that the observed associations between gut bacterial taxa and lipid levels are mediated through effects on BMI or other metabolic disturbances. This notion is supported by a recent analysis in a population-based cohort that not only confirmed previously known associations between obesity and certain bacterial taxa, such as Akkermansia, Christensenellaceae, and Tenericutes [19, 20], but also demonstrated that some of the associations with microbial composition were shared between BMI and levels of triglycerides and high-density lipoproteins [••21]. Importantly, however, these analyses revealed microbial taxa whose proportions were associated with lipids independent of BMI as well, including novel associations with Eggerthella, Pasteurellaceae, and Butyricimonas. Surprisingly, only weak relationships were noted between microbial variation and total cholesterol or low-density lipoprotein cholesterol levels, suggesting that gut bacteria affect specific aspects of lipid metabolism and/or distinct classes of lipoproteins. Taken together, these observations provide new avenues for validation and follow up studies. Biological Mechanisms through which Gut Microbes May Affect Lipid Metabolism As with any gut microbiota study that is associative in nature, such as those in humans, a major challenge is elucidating the underlying biological mechanisms and proving whether the associations are due to a causal relationship. In this regard, evidence from animal studies supports the notion that the gut microbiome can mechanistically impact host lipid levels. For example, certain facultative and anaerobic bacteria in the large bowel produce secondary bile acids from the pool of bile salts secreted into the intestine (Figure). A small fraction of these bacterially derived bile acids is absorbed into the bloodstream and can modulate hepatic and/or systemic lipid and glucose metabolism through nuclear or G protein-coupled receptors (GPCRs), such as FXR or TGR5, respectively [22–24]. Another potential mechanism through which gut microbes could affect lipid metabolism may involve fermentation of nondigestable carbohydrates. Humans are not capable of breaking down many common forms of complex carbohydrates, whereas a subset of anaerobic bacteria found in the cecum and proximal colon can ferment several compounds, such as pectins, gums, hemicelluloses, and galactose-oligosaccharides [25]. One class of metabolites produced by these bacteria are short chain fatty acids (SCFAs), which can subsequently be metabolized by the host or alternatively act as hormones (Figure). SCFAs, such as acetate, propionate, and butyrate, are known to regulate intestinal immune homeostasis and serve as an energy source for colonic epithelial cells. However, SCFAs have been shown to have metabolic benefits as well, which are mediated, in part, through induction of intestinal gluconeogenesis [26]. SCFAs are also absorbed from the gut and can have potent effects on energy expenditure and insulin sensitivity in peripheral metabolic tissues through different GPCRs, such as GPR41 and GPR43 [27, 28]. It is also possible that gut bacteria generate intermediate precursors that are further metabolized by the host to products that exert direct effects on lipid levels. For example, recent studies have linked high levels of trimethylamine N-oxide (TMAO) to atherosclerosis in both mice and humans [7]. TMAO is derived secondarily through hepatic oxidation of trimethylamine (TMA), which is first produced through gut microbe-mediated metabolism of dietary choline and L-carnitine [29, 30] (Figure). Possible mechanisms for the pro-atherogenic effect of TMAO have been suggested to involve perturbations of reverse cholesterol transport, cholesterol and sterol metabolism, and/or the quantity and composition of bile acids [7, 29, •31, •32]. Interestingly, host DNA variation appears to only play a marginal role in the regulation of TMAO levels, particularly in humans, suggesting that dietary factors and/or gut bacterial composition are more important determinants [33]. The Role of Host Genetic Factors Another potentially important aspect to how gut microbiota can impact lipid metabolism may be related to genetic factors of the host. This concept is supported by evidence in humans demonstrating that gut bacterial composition has a significant heritable component and can vary across taxa or members of different phyla [34, ••35]. In the TwinsUK cohort, the abundance of the obesity-associated taxa Christensenellaceae was more highly correlated within monozygotic twins than dizygotic twins, and its heritability was shown to be independent of BMI [••35]. It is reasonable to assume that bacterial taxa that influence lipid metabolism could similarly have heritable components as well. Given the observed heritability of intestinal microbiota, attempts have been made to identify the genetic variants that are associated with gut bacterial composition. A targeted candidate approach with ~250 previously validated variants for lipid levels and BMI did not find any evidence for association of microbiome composition with these SNPs, either alone or as a risk score [••21]. This may be explained, in part, by the fact that the selection of variants to test was only based on their main effects on lipids and BMI. One might speculate that host genetic variants that are involved in regulating bacterial abundance within the host would be more likely to be associated with proportions of gut microbial taxa. However, given the enormous genetic diversity of intestinal bacteria and variability in dietary intake, it is likely that sample sizes in most human microbiome studies are still insufficient to permit identification of robust genetic associations. Moreover, linking these taxa proportions to changes in host lipid levels will pose additional challenges as well. By comparison, investigating the genetic determinants of bacterial composition has been more successful among inbred mouse strains, where environment (i.e. diet) and other confounding variables (i.e. age and sex) are tightly controlled. Notably, a recent genetic analysis with ~110 strains in the Hybrid Mouse Diversity Panel, all of which were maintained on equivalent dietary and housing conditions, identified seven host loci that were associated with common bacterial genera [••36]. Among the candidate genes at these loci were those implicated in processes related to innate immunity, glucose/insulin regulation, and the rapid acute-phase response to lipopolysaccharide. Additional noteworthy observations from this study were that one of the two host genetic loci that were associated with the proportion of Akkermansia muciniphila was also associated with gonadal fat mass and triglyceride levels [••36]. It is likely that similar genetic associations exist in humans as well, but their identification will require larger sample sizes, broader interrogation of genes, and/or new discovery approaches. The Challenges of Gut Microbiota Studies and Proving Causality For the most part, metagenomic analyses on the human gut microbiota have used fecal samples but it is clear from animal studies that certain anaerobic organisms, such as Akkermansia muciniphila, reside primarily in the mucosal layers of the gut and are not readily detected in analyses of only feces. Indeed, the microbial composition throughout the gut varies considerably, both with respect to the anatomic location along the intestinal tract and within a given site, according to the micro-environment. For example, at a given site of intestinal mucosa/lumen, distinct microbes can uniquely reside deep within the crypts, versus on the surface of the mucosal villi, versus within the fecal material. Thus, without invasive procedures to get samples from distinct anatomical regions through the intestines, obtaining a more complete picture of the full spectrum of gut microbiota, at least in humans, poses significant challenges. Next generation sequencing technology has been a major step forward by permitting more robust, time-efficient, and cost-effective characterization of intestinal microbiota. However, the evolutionary resolution provided by 16S sequencing is still limited with current platforms since bacterial composition with this approach is typically identified only down to the genus level. Untargeted metagenomics studies are beginning to overcome these challenges but these types analyses require much more sophisticated types of algorithms and are more expensive to carry out. Nonetheless, these efforts will improve our ability to identify and quantitate distinct species of bacteria and may have implications for understanding the pathological mechanisms through which specific bacteria affect both human health and disease processes. Fecal microbial composition studies are associative, and thus hypothesis generating. Ultimately, proof of causality for identified microbial associations with host phenotypes requires additional experimentation, such as manipulating gut bacterial composition and observing changes in physiological parameters that were identified in the initial associations. In this regard, intestinal microbial transplantation studies in both animal models and humans have provided evidence for a causal role of gut bacteria in treating various intestinal diseases, most notably Clostridium difficile infection. Bacterial transplantation experiments have also been shown to modulate metabolic and cardiovascular phenotypes. For example, studies in mice have elegantly demonstrated that transfer of gut microbes from either obese mice or humans can transmit obesity phenotypes to the recipients [34, •37]. By contrast, administration of Akkermansia muciniphila to an obesity-prone mouse strain significantly improved several metabolic parameters, including substantial decreases in total cholesterol, triglycerides, and, most strikingly, insulin resistance [••36]. A similar strategy has provided evidence that atherosclerosis susceptibility in mice could also be transmitted to a host by gut microbial transplantation [•38]. Although such approaches have yet to be implemented in humans for treating dyslipidemia or other cardiometabolic traits, fecal transplantation studies have shown that transfer of gut microbes from lean donors through a duodenal infusion into patients with metabolic syndrome can improve insulin sensitivity [39]. These observations underscore the therapeutic potential of interventions that alter gut microbial composition or function. Targeting the Gut Microbiota for Therapeutic Applications An important clinical implication from studies on gut microbiota is how to leverage findings for therapeutic purposes. Selective manipulation of the gut microbial ecosystem might provide new avenues to treat and/or prevent dyslipidemia and cardiometabolic diseases, but this will first require a better understanding of which specific bacteria, or alternatively, which bacterial metabolites, are the appropriate targets for intervention and manipulation. The simplest point of intervention may be to limit consumption of dietary constituents that either foster the growth of undesirable bacteria or serve as substrates for microbe-dependent generation of products that disrupt lipid homeostasis or other metabolic processes. Alternative viable therapeutic strategies may be the use of prebiotics or probiotics to produce a desired change in microbial composition and/or function that favorably impacts host lipid metabolism. Prebiotic therapy consists of ingestion of select nutrients or dietary constituents (nonmicrobial compositions) that provide a growth advantage of beneficial bacteria, whereas probiotic therapy involves the ingestion of one or more live bacterial strains, attempting to take advantage of the mutualism of microbes. Therapeutic intervention could also rely on the use of broad or class-specific antibiotics to eliminate bacterial species or their products associated with dyslipidemia and other metabolic disturbances. However, this approach is not a sustainable long-term option. Many gut microbial products are beneficial to the host and even infrequent antibiotic treatment, particularly in very young children whose gut microbiota has yet to be fully established, can adversely impact host global metabolism via changes in the gut microbial community [40] and facilitate the emergence of antibiotic-resistant bacterial strains. Another promising therapeutic approach may involve pharmaceutical targeting of gut microbe-specific biological processes. This concept was recently demonstrated in a series of elegant experiments with respect to the association of TMAO with atherosclerosis [••41]. Several important insights were revealed by this study. For example, Wang et al. designed a small molecule choline analog, 3,3-dimethyl-1-butanol (DMB), that competitively inhibited diverse and phylogenetically distant classes of microbial TMA lyases, which were previously identified as enzymes that catalyze the conversion of choline to TMA [42, 43]. Notably, these effects were observed in physiological polymicrobial cultures derived from both cecal contents of mice and fecal samples of healthy humans. Most importantly, chronic feeding of DMB to mice in the context of a high-choline diet led to shifts in the proportions of some bacterial taxa and substantial reductions in plasma TMAO levels, macrophage cholesterol accumulation, foam cell formation, and atherosclerotic lesions, without any evidence of toxicity or adverse cardiometabolic effects in the animals [••41]. Taken together, these results suggest that targeting gut microbial production of TMA through specific and non-lethal means may serve as a potential therapeutic approach for the treatment of cardiometabolic diseases and that microbial inhibitors in general may represent a novel therapeutic strategy for other disorders that involve intestinal dysbiosis. Conclusions For many years, the community of bacteria living in our gut was largely ignored. However, emerging evidence clearly demonstrates that our microbial symbionts play multiple fundamentally important roles in maintaining normal metabolic homeostasis. These discoveries have broad implications for elucidating bacterially-mediated pathophysiological mechanisms that alter lipid metabolism and other related metabolic traits. From a clinical perspective, this newly recognized endocrine organ system can be targeted for therapeutic benefit or prevention of cardiometabolic diseases and risk factors. The ability to manipulate the gut microbiome for improved health and prevention of diseases is still in the early phases of development, but recent rapid advances in gut microbiome studies highlight both the potential and promise of targeting intestinal microbes for therapeutic gain. None. Financial Support and Sponsorship Work in the author’s laboratories is supported, in part, by NIH grants R01ES021801, R01ES021801-S3, R01ES025786, P01ES022845, and R01HL128572, and U.S. EPA Grant RD83544101. The funders had no role in preparation, review, or approval of the manuscript. Figure Schematic illustration of organ systems and tissues that can be affected by the gut microbiota Multiple lines of evidence support a role for altered gut microbial composition or function as a contributor to the development of obesity and related metabolic abnormalities (i.e. type 2 diabetes), peripheral and coronary artery disease, and even neurobehavioral conditions such as autism. Recent observations of significant associations between proportions of specific intestinal bacteria taxa with lipid levels suggest a role for gut microbes in modifying host lipid metabolism. Gut microbe effects may be mediated through multiple mechanisms, including elaboration of lipopolysaccharide (LPS) or other bioactive metabolites that act fundamentally as hormones since they can circulate within the host and act at distant sites. Gut microbial production of short chain fatty acids (SCFAs) and secondary bile acids are two such examples that have been shown to affect lipid levels and other metabolic phenotypes. Evidence shows that gut bacteria can also generate intermediate precursors (e.g. trimethylamine) from certain dietary nutrients, that can then be further metabolized by the host to generate biologically active products (e.g. trimethylamine N-oxide), which then can exert direct effects on lipid metabolism and contribute to disease development or progression. Biological mechanisms impacted by gut microbial metabolites can involve reverse cholesterol transport, hepatic cholesterol and sterol metabolism, intestinal lipid transport, bile acid composition and pool size, glucose and insulin metabolism, energy harvest/expenditure, as well as others. Key Points The gut bacterial community is increasingly being recognized as an endocrine system that can modulate a variety of cardiometabolic processes, including host lipid metabolism. A variety of signaling molecules produced by gut bacteria have potent effects on hepatic lipid and bile metabolism and on reverse cholesterol transport, energy expenditure, and insulin sensitivity in peripheral tissues. Host genetic factors can modulate the abundance of bacterial taxa, which can subsequently affect metabolic phenotypes. Proof of causality for identified microbial associations with host lipid-related phenotypes has been demonstrated in animal studies but remains a challenge in humans. Manipulation of gut microbiota may serve as a novel therapeutic strategy for treatment and/or prevention of dyslipidemia and cardiometabolic diseases, but this will first require a better understanding of which specific bacteria are the appropriate targets for intervention. Conflicts of Interest None. References 1 Backhed F Ding H Wang T The gut microbiota as an environmental factor that regulates fat storage Proc. Natl. Acad. Sci. U. S. A 2004 101 15718 15723 15505215 2 Dumas ME Barton RH Toye A Metabolic profiling reveals a contribution of gut microbiota to fatty liver phenotype in insulin-resistant mice Proc. Natl. Acad. Sci. U. S. A 2006 103 12511 12516 16895997 3 Turnbaugh PJ Ley RE Mahowald MA An obesity-associated gut microbiome with increased capacity for energy harvest Nature 2006 444 1027 1031 17183312 4 Qin J Li Y Cai Z A metagenome-wide association study of gut microbiota in type 2 diabetes Nature 2012 490 55 60 23023125 5 Henao-Mejia J Elinav E Jin C Inflammasome-mediated dysbiosis regulates progression of NAFLD and obesity Nature 2012 482 179 185 22297845 6 Karlsson FH Tremaroli V Nookaew I Gut metagenome in European women with normal, impaired and diabetic glucose control Nature 2013 498 99 103 23719380 7 Wang Z Klipfell E Bennett BJ Gut flora metabolism of phosphatidylcholine promotes cardiovascular disease Nature 2011 472 57 63 21475195 8 Hsiao EY McBride SW Hsien S Microbiota modulate behavioral and physiological abnormalities associated with neurodevelopmental disorders Cell 2013 155 1451 1463 24315484 9 Weinstock GM Genomic approaches to studying the human microbiota Nature 2012 489 250 256 22972298 10 Human Microbiome Project Consortium A framework for human microbiome research Nature 2012 486 215 221 22699610 11 Human Microbiome Project Consortium Structure, function and diversity of the healthy human microbiome Nature 2012 486 207 214 22699609 12 Jernberg C Lofmark S Edlund C Jansson JK Long-term impacts of antibiotic exposure on the human intestinal microbiota Microbiology 2010 156 3216 3223 20705661 13 Muegge BD Kuczynski J Knights D Diet drives convergence in gut microbiome functions across mammalian phylogeny and within humans Science 2011 332 970 974 21596990 14 Wu GD Chen J Hoffmann C Linking long-term dietary patterns with gut microbial enterotypes Science 2011 334 105 108 21885731 15 Claesson MJ Jeffery IB Conde S Gut microbiota composition correlates with diet and health in the elderly Nature 2012 488 178 184 22797518 16 Yatsunenko T Rey FE Manary MJ Human gut microbiome viewed across age and geography Nature 2012 486 222 227 22699611 17 Schloissnig S Arumugam M Sunagawa S Genomic variation landscape of the human gut microbiome Nature 2013 493 45 50 23222524 18 Velagapudi VR Hezaveh R Reigstad CS The gut microbiota modulates host energy and lipid metabolism in mice J. Lipid Res 2010 51 1101 1112 20040631 19 Derrien M Van Baarlen P Hooiveld G Modulation of mucosal immune response, tolerance, and proliferation in mice colonized by the mucin-degrader Akkermansia muciniphila Front Microbiol 2011 2 166 21904534 20 Everard A Belzer C Geurts L Cross-talk between Akkermansia muciniphila and intestinal epithelium controls diet-induced obesity Proc. Natl. Acad. Sci. U. S. A 2013 110 9066 9071 23671105 21 Fu J Bonder MJ Cenit MC The gut microbiome contributes to a substantial proportion of the variation in blood lipids Circ. Res 2015 117 817 824 26358192 This study provides some of the first evidence in humans that variation in gut bacteria are associated with blood lipid levels, several of which were independent of obesity. Correlations were also observed with distinct classes of lipoproteins, suggesting that gut bacteria may affect specific aspects of lipid metabolism. 22 Watanabe M Houten SM Mataki C Bile acids induce energy expenditure by promoting intracellular thyroid hormone activation Nature 2006 439 484 489 16400329 23 Thomas C Gioiello A Noriega L TGR5-mediated bile acid sensing controls glucose homeostasis Cell Metab 2009 10 167 177 19723493 24 Ryan KK Tremaroli V Clemmensen C FXR is a molecular target for the effects of vertical sleeve gastrectomy Nature 2014 509 183 188 24670636 25 Macfarlane GT Macfarlane S Bacteria, colonic fermentation, and gastrointestinal health J. AOAC Int 2012 95 50 60 22468341 26 De Vadder F Kovatcheva-Datchary P Goncalves D Microbiota-generated metabolites promote metabolic benefits via gut-brain neural circuits Cell 2014 156 84 96 24412651 27 Brown AJ Goldsworthy SM Barnes AA The orphan G protein-coupled receptors GPR41 and GPR43 are activated by propionate and other short chain carboxylic acids J. Biol. Chem 2003 278 11312 11319 12496283 28 Kimura I Ozawa K Inoue D The gut microbiota suppresses insulin-mediated fat accumulation via the short-chain fatty acid receptor GPR43 Nat Commun 2013 4 1829 23652017 29 Koeth RA Wang Z Levison BS Intestinal microbiota metabolism of L-carnitine, a nutrient in red meat, promotes atherosclerosis Nat. Med 2013 19 576 585 23563705 30 Bennett BJ de Aguiar Vallim TQ Wang Z Trimethylamine-N-oxide, a metabolite associated with atherosclerosis, exhibits complex genetic and dietary regulation Cell Metab 2013 17 49 60 23312283 31 Shih DM Wang Z Lee R Flavin containing monooxygenase 3 exerts broad effects on glucose and lipid metabolism and atherosclerosis J. Lipid Res 2015 56 22 37 25378658 32 Warrier M Shih DM Burrows AC The TMAO-generating enzyme flavin monooxygenase 3 is a central regulator of cholesterol balance Cell Rep 2015 10 326 338 These two studies provided evidence that perturbing the expression of FMO3, the primary enzyme that produces the pro-atherogenic metabolite TMAO from a gut-microbiota generated intermediate, leads to alterations in plasma lipid levels and hepatic lipid and glucose metabolism. 33 Hartiala J Bennett BJ Tang WH Comparative genome-wide association studies in mice and humans for trimethylamine N-oxide, a proatherogenic metabolite of choline and L-carnitine Arterioscler. Thromb. Vasc. Biol 2014 34 1307 1313 24675659 34 Turnbaugh PJ Hamady M Yatsunenko T A core gut microbiome in obese and lean twins Nature 2009 457 480 484 19043404 35 Goodrich JK Waters JL Poole AC Human genetics shape the gut microbiome Cell 2014 159 789 799 25417156 This paper describes one of the first studies in humans to demonstrate the heritability of gut microbiota using a large number of monozygotic and dizygotic twin pairs. Follow-up experiments also provided evidence that certain bacterial taxa reduced weight gain when transplanted into germ-free mice. 36 Org E Parks BW Joo JW Genetic and environmental control of host-gut microbiota interactions Genome Res 2015 25 1558 1569 26260972 This study used a panel of inbred mouse strains to illustrate the heritability of gut microbial composition and to identify genetic loci that were associated with the abundance of various bacterial genera. Transplantation studies in an obese-prone strain also proved a causal role for Akkermansia muciniphila in metabolic homeostasis. 37 Ridaura VK Faith JJ Rey FE Gut microbiota from twins discordant for obesity modulate metabolism in mice Science 2013 341 1241214 24009397 38 Gregory JC Buffa JA Org E Transmission of atherosclerosis susceptibility with gut microbial transplantation J. Biol. Chem 2015 290 5647 5660 25550161 These two studies demonstrated that obesity or atherosclerosis phenotypes in mice can be transmitted through transfer of gut microbiota, thus demonstrating a direct causal role for intestinal bacteria in modulating these cardiometabolic traits and providing one potentially novel therapeutic approach. 39 Vrieze A Van Nood E Holleman F Transfer of intestinal microbiota from lean donors increases insulin sensitivity in individuals with metabolic syndrome Gastroenterology 2012 143 913 e917 916 e917 22728514 40 Cox LM Blaser MJ Antibiotics in early life and obesity Nat Rev Endocrinol 2015 11 182 190 25488483 41 Wang Z Roberts AB Buffa JA Non-lethal inhibition of gut microbial trimethylamine production for the treatment of atherosclerosis Cell 2015 163 1585 1595 26687352 This is one of the first studies to demonstrate that bacterial-specific processes can be pharmaceutical targeted for therapeutic benefit. An inhibitor of microbial TMA lyases was shown to reduce plasma levels of the pro-atherogenic choline-derived metabolite TMAO as well as aortic lesion formation in mice in the absence of any toxicity or adverse cardiometabolic effects. 42 Craciun S Balskus EP Microbial conversion of choline to trimethylamine requires a glycyl radical enzyme Proc. Natl. Acad. Sci. U. S. A 2012 109 21307 21312 23151509 43 Sandhu SS Chase T Jr Aerobic degradation of choline by Proteus mirabilis: enzymatic requirements and pathway Can. J. Microbiol 1986 32 743 750 3536045
PMC005xxxxxx/PMC5125445.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 9708816 20612 Depress Anxiety Depress Anxiety Depression and anxiety 1091-4269 1520-6394 24425049 5125445 10.1002/da.22233 NIHMS830531 Article Major depressive disorder subtypes to predict long-term course van Loo Hanna M. M.D. 1 Cai Tianxi Sc.D. 2 Gruber Michael J. M.S. 3 Li Junlong Ph.D. 2 de Jonge Peter Ph.D. 1 Petukhova Maria Ph.D. 3 Rose Sherri Ph.D. 3 Sampson Nancy A. B.A. 3 Schoevers Robert A. M.D., Ph.D. 1 Wardenaar Klaas J. Ph.D. 1 Wilcox Marsha A. Ed.D., Sc.D. 4 Al-Hamzawi Ali Obaid M.D., F.I.C.M.S. 5 Andrade Laura Helena M.D., Ph.D. 6 Bromet Evelyn J. Ph.D. 7 Bunting Brendan Ph.D. 8 Fayyad John M.D. 9 Florescu Silvia E. M.D., Ph.D. 10 Gureje Oye M.D., Ph.D., D.Sc., F.R.C.Psych. 11 Hu Chiyi M.D., Ph.D. 12 Huang Yueqin M.D., M.P.H., Ph.D. 13 Levinson Daphna Ph.D. 14 Medina-Mora Maria Elena Ph.D. 15 Nakane Yoshibumi M.D., Ph.D. 16 Posada-Villa Jose M.D. 17 Scott Kate M. Ph.D. 18 Xavier Miguel M.D., Ph.D. 19 Zarkov Zahari M.D. 20 Kessler Ronald C. Ph.D. 3* 1 Department of Psychiatry, University of Groningen, University Medical Center Groningen, Groningen The Netherlands 2 Department of Biostatistics, Harvard School of Public Health, Boston, MA, USA 3 Department of Health Care Policy, Harvard Medical School, Boston, MA, USA 4 Johnson and Johnson Pharmaceutical Research and Development, Titusville, NJ, USA 5 Al-Qadisia University College of Medicine, Diwania, Iraq 6 Section of Psychiatric Epidemiology-LIM 23 Department and Institute of Psychiatry, University of São Paulo Medical School, São Paulo, Brazil 7 State University of New York at Stony Brook, Stony Brook, New York, USA 8 Psychology Research Institute, University of Ulster, Londonderry, UK 9 Institute for Development Research, Advocacy, and Applied Care and St.George Hospital University Medical Center, Beirut, Lebanon 10 National School of Public Health, Management and Professional Development, Bucharest, Romania 11 University College Hospital, Ibadan, Nigeria 12 Shenzhen Institute of Mental Health and Shenzhen Kangning Hospital, Guangdong Province, People's Republic of China 13 Institute of Mental Health, Peking University, Beijing, People's Republic of China 14 Research and Planning, Mental Health Services, Ministry of Health, Jerusalem, Israel 15 Instituto Nacional de Psiquiatria Ramon de la Fuente, Mexico City, Mexico 16 Department of Social Work, The Faculty of Human Sociology, Nagasaki International University, Nagasaki, Japan 17 Universidad Colegio Mayor de Cundinamarca, Bogota, Colombia 18 Department of Psychological Medicine, Otago University, Dunedin, New Zealand 19 Department of Mental Health, Universidade Nova de Lisboa, Lisbon, Portugal 20 National Center of Public Health and Analyses Department Mental Health, Sofia, Bulgaria * Address correspondence to Ronald C. Kessler, Ph.D., Department of Health Care Policy, Harvard Medical School, 180 Longwood Avenue, Boston, MA 02115. Tel. (617) 432-3587, Fax (617) 432-3588, [email protected] 18 11 2016 14 1 2014 9 2014 28 11 2016 31 9 765777 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Background Variation in course of major depressive disorder (MDD) is not strongly predicted by existing subtype distinctions. A new subtyping approach is considered here. Methods Two data mining techniques, ensemble recursive partitioning and Lasso generalized linear models (GLMs) followed by k-means cluster analysis, are used to search for subtypes based on index episode symptoms predicting subsequent MDD course in the World Mental Health (WMH) Surveys. The WMH surveys are community surveys in 16 countries. Lifetime DSM-IV MDD was reported by 8,261 respondents. Retrospectively reported outcomes included measures of persistence (number of years with an episode; number of with an episode lasting most of the year) and severity (hospitalization for MDD; disability due to MDD). Results Recursive partitioning found significant clusters defined by the conjunctions of early onset, suicidality, and anxiety (irritability, panic, nervousness-worry-anxiety) during the index episode. GLMs found additional associations involving a number of individual symptoms. Predicted values of the four outcomes were strongly correlated. Cluster analysis of these predicted values found three clusters having consistently high, intermediate, or low predicted scores across all outcomes. The high-risk cluster (30.0% of respondents) accounted for 52.9-69.7% of high persistence and severity and was most strongly predicted by index episode severe dysphoria, suicidality, anxiety, and early onset. A total symptom count, in comparison, was not a significant predictor. Conclusions Despite being based on retrospective reports, results suggest that useful MDD subtyping distinctions can be made using data mining methods. Further studies are needed to test and expand these results with prospective data. Epidemiology Depression Anxiety/Anxiety Disorders Suicide/Self Harm Panic Attacks INTRODUCTION Patients with major depressive disorder (MDD) vary substantially in treatment response and illness course. Recognition of this variation has led researchers to search for depression subtypes defined either by presumed causes (e.g., postnatal depression),[1,2] clinical presentation (e.g., atypical or melancholic depression,[3,4]) or empirically-derived symptom profiles using cluster analysis,[5] factor analysis,[6] or latent class analysis,[7] in hopes that patients in subtypes would be sufficiently similar in psychopathological processes to help identify underlying molecular etiologies or predict treatment response.[7-9] However, subtyping distinctions up to now have not lived up to these expectations,[8,10] although some commentators suggest that subtyping using endophenotypes or intermediate phenotypes might hold more promise.[11,12] Another potentially useful approach to subtyping, given the goal of prediction, would be to define subtypes using recursive partitioning[13,14] and related data mining methods[15,16] that search for synergistic associations of predictors with illness course. Such methods have been used in other areas of medicine[17,18] and relatively simple applications have been used in psychiatry to predict depression treatment response[19-23] and suicidality.[24-26] The current report presents results of preliminary analyses designed to find symptom-based subtypes predicting course of major depressive disorder using more complex data mining methods than in previous studies. The analysis is preliminary because it uses retrospective data on depression course collected in cross-sectional population epidemiological surveys rather than longitudinal clinical studies. Results are nonetheless useful in providing a proof of concept of the approach in a large and diverse sample of subjects who were asked about potentially important subtyping variables in their index episodes and assessed for multiple indicators of subsequent depression persistence and severity. MATERIALS AND METHODS Sample Data come from the World Health Organization World Mental Health (WMH) surveys (www.hcp.med.harvard.edu/wmh), a series of well-characterized community epidemiological surveys[27-30] administered in six countries classified by the World Bank as high income (Israel, Japan, New Zealand, Northern Ireland, Portugal, United States,), five upper-middle income (Brazil, Bulgaria, Lebanon, Mexico, Romania), and five low/lower-middle income (Colombia, Iraq, Nigeria, Peoples Republic of China, Ukraine).[31] Most surveys feature nationally representative household samples, while two (Colombia, Mexico) represent all urban areas in the country, one selected states (Nigeria), and three selected Metropolitan Areas (Brazil, Japan, Peoples Republic of China). (Table 1) A total of 93,167 adults (age 18+) participated, 8,261 of whom met lifetime DSM-IV criteria for MDD. Sample sizes range from 2,357 (Romania) to 12,790 (New Zealand). The average weighted response rate was 73.7% (range: 55.1-95.2%). Weights adjusted for differential probabilities of selection and discrepancies with population socio-demographic/geographic distributions. Further details about WMH sampling and weighting are available elsewhere.[32] Measures Interview procedures Translation, back-translation, and harmonization of the interview schedule used standardized procedures.[33] Interviews were fully-structured and administered face-to-face in the homes of respondents by trained lay interviewers. Rigorous interviewer training and quality control procedures were employed.[34] The research presented here is in compliance with the Code of Ethics of the World Medical Association (Declaration of Helsinki). The institutional review board of the organization that coordinated the survey in each country approved and monitored compliance with procedures for obtaining informed consent and protecting human subjects. MDD DSM-IV MDD was assessed with the Composite International Diagnostic Interview (CIDI), Version 3.0,[35] a fully-structured diagnostic interview designed for administration by trained lay interviewers. The CIDI translation, back-translation, and harmonization protocol required culturally competent bilingual clinicians to review, modify, and approve key phrases describing symptoms. Clinical reappraisal studies conducted in several WMH countries found good concordance between lifetime DSM-IV/CIDI diagnoses of major depression and independent diagnoses based on blinded SCID clinical reappraisal interviews,[36] with area under the ROC curve (AUC) averaging .75 and LR+ averaging 8.8 (a level close to the threshold considered definitive for ruling in a clinical diagnosis from a screen).[37] Respondents with lifetime DSM-IV/CIDI MDD were asked retrospective questions about age-of-onset (AOO), whether their first lifetime depressive episode “was brought on by some stressful experience” or happened “out of the blue,” all DSM-IV Criterion A-D symptoms of MDE for the index episode (including separate questions about weight loss and weight gain, insomnia and hypersomnia, psychomotor agitation and retardation, and thoughts of death, suicide ideation, suicide plans, and suicide gestures-attempts), ICD-10 severity specifiers, questions to operationalize diagnostic hierarchy rule exclusions, and questions about symptoms during the index episode that might be markers of (i) dysthymia (inability to cope; social withdrawal), (ii) mixed episodes (sleep much less than usual and still not feel tired; racing thoughts), and (iii) anxious depression (feeling irritable; nervous-anxious-worried; having sudden attacks of intense fear or panic). Four retrospective questions were asked about subsequent lifetime MDD course: number of years since AOO when the respondent had an episode (i) lasting two weeks or longer or (ii) lasting most days throughout the year; (iii) a dichotomous measure of whether the depression was ever so severe that the respondent was hospitalized overnight (and, if so, age of first hospitalization); and (iv) a dichotomous measure of whether the respondent was currently disabled (at least 50% limitation in ability to perform paid work) because of depression. These are the four outcomes considered here. The two measures of years in episode were divided by number of years between age-at-interview (AAI) and AOO+1 to create continuous outcomes in the range 0-100%. Other predictors In addition to the information described above about the index episode, additional predictors included discretized information about the respondent's AOO in eight nested age categories selected for sensitivity in the age range with most onsets (less than or equal to ages 12, 15, 19, 24, 29, 34, 39, 59), similarly nested and discretized information about AAI-AOO, and a binary variable for respondent Family History Research Diagnostic Criteria Interview[38] reports for whether respondents’ parents had a history of major depression. Analysis methods Analysis of the de-identified WMH master dataset was approved by the Institutional Review Board of Harvard Medical School, the site of the WMH Data Coordination Center. An ensemble of 100 classification trees was used to find important interactions among predictors of the outcomes. The ensemble approach (i.e., combining results across a large number of replicates, each replicate estimated in a different simulated pseudo-sample) was used to reduce risk of over-fitting.[13-15] The recursive partitioning R package rpart[39] was used for this purpose. The minimum number of observations in a node for further splitting was set at 20 and the threshold complexity parameter (cp) at 0.01. The models to predict years in episode, which used a Poisson link function, were estimated among respondents where AAI-AOO was either 10+ years (years with episodes lasting most of the year) or 15+ years (years with any episode) based on preliminary inspection showing that outcome scores stabilized after these cut-points. Proportional hazards survival models were used to predict age at first hospitalization for depression among respondents who were not hospitalized for depression at AOO. Logistic regression models were used to predict current disability in the total sample. Each tree in the ensemble was built in a randomly selected bootstrap sample drawn without replacement from the sample and cross-validated among the remaining respondents to determine appropriate tree depth. Inspection of summary frequencies of unique terminal nodes (i.e., subgroups of respondents defined by the conjunction of the dichotomous predictors selected to optimize prediction of the outcome) across the 100 trees was used to select the interactions to retain in a second step of analysis. This second step fitted a separate generalized linear model (GLM) for the multivariate associations of all predictors with each outcome. Included here were additive associations of the individual predictors with the outcome, the interactions found to occur repeatedly in the tree models, and nested dichotomies to describe the total number of symptoms endorsed. The inclusion of the latter predictors was important to distinguish differential predictive effects of especially important symptoms from predictive effects of an overall symptom count. As some of the predictors in the GLM models were highly correlated, conventional regression methods yielded unstable results. Stepwise regression,[40] which is often used to address this problem, over-fits and performs poorly in new samples.[41] A number of data mining methods have been developed to improve on stepwise regression. We used one such method, the Lasso,[42] to address this problem. The Lasso is one of several penalized regression methods that trades off bias to increase the efficiency of estimation by constraining the sum of variance of nonzero values of standardized regression coefficients with coefficient shrinkage parameters We selected Lasso instead of alternatives, as this penalty handles high correlations among predictors by yielding a sparse model (i.e., forces coefficients of weak predictors to zero).[43] The R-package glmnet[44] was used to estimate the Lasso GLMs using the same link functions as in the regression tree models. Coefficients from the Lasso models were exponentiated to create incidence density ratios (IDRs) to predict proportion of years in episode, hazard ratios (HRs) to predict hospitalization, and odds-ratios (ORs) to predict disability. No confidence intervals were generated, as standard errors in such models are biased. The best-fitting Lasso coefficients were then used to generate predicted values of each outcome for all respondents. Based on evidence of strong correlations among these predicted values across outcomes, k-means cluster analysis was used to partition the sample into subtypes with similar multivariate profiles of predicted scores across the four outcomes using the R-package stats[45] and using 100 random starts for each number of clusters. Inspection of observed (as opposed to predicted) mean dichotomized outcome scores (percentages of respondents with high persistence and chronicity, hospitalization, and disability) and calculation of AUC (adjusted appropriately for the survival outcome,[46] were used to select an optimal number of clusters. Associations of cluster membership with dichotomized versions of outcomes were then examined by calculating relative-risk of the adverse outcomes in the high-risk versus other clusters, positive predictive value (PPV; the proportion of high-risk cluster respondents that experienced the adverse outcomes), and sensitivity (SN; the proportion of all adverse outcomes that occurred in the high-risk cluster). RESULTS Distributions of the outcomes The mean, median, and inter-quartile range (25th-75th percentiles) percentages of years after AOO when respondents in the analysis sample reported having a depressive episode lasting two weeks or longer were 25.8%, 13.0%, and 6.2-29.4%, respectively. The comparable percentages for years having a depressive episode lasting most days throughout the year were 9.5%, 0.0%, and 0.0-9.3%. Lifetime hospitalization for a depressive episode was reported by 4.3% of respondents and current disability due to depression was reported by 1.6% of respondents. Recursive partitioning The terminal nodes repeatedly predicting outcomes in recursive partitioning all involved two-way or three-way interactions between child-adolescent (before age 19) AOO, suicidality, and anxiety (nervous-anxious-worried, irritable, attacks of fear-panic) during index depressive episodes. The conjunction of later AOO (age 35+) with anxiety and suicidality also predicted chronicity. The cells defined by the conjunction of early onset, suicidality, and anxiety had either the highest or, in one case (disability), second highest scores on all outcomes across cells of the table defined by these predictors. (Detailed results are available on request.) Based on these results, all two-way and three-way interactions among AOO, anxiety, and suicidality were included in the Lasso GLMs. Lasso generalized linear models Four predictors of persistence, eight of chronicity, and 11 each of hospitalization and disability were retained in the GLMs with Lasso coefficients meaningfully different from zero. (Table 2) The vast majority (85%) of these coefficients were positive. The positive IDRs for years in episode were in the range 1.1-1.4. The positive HRs for hospitalization and ORs for disability were in the range 1.1-1.9. Only one predictor, severe dysphoria, was retained in all four models. Severe dysphoria was also the strongest predictor of chronicity (IDR=1.4) and one of the strongest predictors of hospitalization (OR=1.7). Four other predictors with consistently positive coefficients retained in three of the four models included suicidality (1.1-1.6), panic attacks (1.1-1.5), the multivariate profile of pediatric onset and anxiety (either nervousness-anxiety-worry or panic) (1.1-1.3), and parental history of major depression (1.2). One of these four, suicidality, was also among the strongest predictors of hospitalization (HR=1.6) and disability (OR=1.5), while panic was one of the strongest predictors of disability (OR=1.5). Other strong predictors of hospitalization included inability to cope (HR=1.9) and hypersomnia (HR=1.5), while inability to cope was also one of the strongest predictors of disability (OR=1.4). Early-AOO-suicidality also predicted disability, while later-AOO (older than age 34)-suicidality predicted chronicity. The latter represented a nonlinearity in the effect of the multivariate AOO-anxiety-suicidality profile. Cluster analysis Predicted values of each outcome were calculated for each respondent based on the GLM model coefficients. Spearman rank-order correlations among these predicted values were in the range .76-.89. Principal axis exploratory factor analysis showed that the correlations were consistent with the existence of a single underlying factor (factor loadings in the range .89-.94). Based on these results, k-means cluster analysis of transformed (to percentiles) predicted outcome scores searched for multivariate clusters defining differential risk of the outcomes. Inspection of mean percentile scores for solutions between three and eight clusters showed all solutions defined one class with the highest mean scores on all outcomes, a second class with lowest mean scores on all outcomes, and other classes with consistently intermediate mean scores on all outcomes. (Figure 1a-1f) Based on this observation, alternative three-cluster solutions were constructed from the original four- through eight-cluster solutions by collapsing the intermediate clusters. AUC was then compared across these solutions to predict dichotomous versions of the measures of years in episodes (distinguishing the 5-10 top percentiles of respondents with highest scores), hospitalization, and disability to see if classifications of high-risk or low-risk clusters were refined in solutions with more than three clusters. None of the collapsed solutions had higher AUCs than the original three-cluster solution (.64 for years in episode, .61 for years in episodes lasting more than half the year, .70 for hospitalization, and .72 for disability). The distribution of membership in the three-cluster solution was 30.7% high-risk, 35.6% intermediate-risk, and 33.7% low-risk. Respondents in the high-risk cluster were 2.1-5.1 times as likely as others and 2.5-11.3 times as likely as respondents in the low-risk cluster to have high levels of long-term MDD persistence and severity. (Table 3) Respondents in the high-risk cluster includes 52.9-69.7% of all those with high levels of long-term MDD persistence and severity and 68.4-71.1% of those with two or more such adverse outcomes. Cluster membership was strongly associated (Cramer's V greater than .50) with only one baseline predictor, suicidality (V=.54), and moderately associated (Cramer's V in the range .30-.50) with eight others, including one Criterion A depressive symptom (worthlessness/excessive guilt, V=.34), the ICD-10 severe dysphoria marker (V=.47), one symptom of dysthymia (inability to cope, V=.50), two of the three symptoms of anxiety (irritability, panic attacks, V=.30-.44), and the early-AOO multivariate symptom profiles retained in the Lasso GLMs (early AOO with either suicidality or anxiety, V=.35-.46). (Table 4) Scores on these variables were consistently higher in the high-risk than intermediate-risk cluster and in the intermediate-risk than the low-risk cluster. However, proportional high-risk versus intermediate-risk differences were relatively modest in most cases (1.1-1.4 risk-ratios) other than for panic (1.7) and the early-AOO multivariate symptoms profiles (2.0-3.2), while proportional intermediate-risk versus low-risk differences were consistently larger, with the highest risk-ratios for panic (2.8), inability to cope (2.5), suicidality (2.0), and the multivariate symptoms profiles (2.4-7.1). DISCUSSION The above results are limited by being based on retrospective data collected in fully-structured interviews excluding information on such potentially important predictors as temporally primary comorbid disorders and treatment status. Sample biases could also have been introduced by differential response related to predictors or predictor effects or differential mortality. The limitations involving use of a fully-structured interview and restricted predictors almost certainly led to downward bias in the estimated strength of associations, but the other limitations could have introduced either conservative or anti-conservative biases. Results should be considered only exploratory because of these limitations, although the results have value both as a proof of concept and as a source of ideas about prediction patterns that warrant analysis in future studies. Within the context of these limitations, three results emerged that could serve as a starting point for future prospective clinical studies. First, the recursive partitioning found an early-onset-anxious-suicidal subtype associated with all four outcomes (persistence, chronicity, hospitalization, disability) and a late-onset-anxious-suicidal subtype associated with chronicity. Second, the GLMs found that a number of index episode symptoms were significant predictors of all outcomes. The most consistent and powerful of these was severe dysphoria, while others included parental history of major depression, suicidality, panic attacks, and multivariate profiles of pediatric onset with anxiety and/or suicidality. Third, strong clustering was found in these predicted values across the outcomes, with the roughly 30% of respondents in the high-risk cluster accounting for more than two-thirds of cases with multiple indicators of high long-term persistence, chronicity, and severity. Several previous epidemiological studies examined baseline predictors of long-term course either in treatment[47,48] or community[49-51] samples, but did not attempt to search for depression subtypes. While these studies found several replicated predictors, including cooccurring anxiety, pain-physical comorbidity, and family history of depression,[50,52-54] no attempt was made in those studies to examine synergistic effects of predictor clusters other than for summary measures of overall depression symptom number. Importantly, we included a total count of depressive symptoms in our GLMs but this measure was not significant. As noted in the introduction, subtyping analyses more similar to those reported here have been done to predict treatment response[19,20] and naturalistic patterns of remission among patients[23] or in the placebo control group of a depression clinical trial.[21] A number of recent clinical studies have also used methods similar to ours either to predict suicidality during[22,25,26] or after termination of[24] treatment. However, none of those analyses used ensemble methods or combined recursive partitioning with GLM to assess both synergistic and additive predictor effects. In considering the possibility of future extensions to prospective studies, it is important to note that although we found an early-onset anxious-suicidal depression subtype that predicts all the outcomes (suicidality being the critical element in predicting disability and anxiety in predicting the other outcomes), we failed to find recursive partitioning profiles associated with a larger set of predictors despite the sample being much bigger than in existing prospective studies (i.e., affording good statistical power to detect synergistic symptom profiles if they existed) and the symptoms considered being quite broad. Taken together with the results of a recent secondary analysis that failed to find stable symptom-based MDD subtypes defined by internal consistency,[10] our failure to find more elaborate subtypes argues against the existence of complex MDD subtypes defined exclusively on the basis of synergistic associations among index depressive episode symptoms other than the early-onset anxious-suicidal subtype. It is important to note that broader MDD predictive subtypes not defined exclusively by index episode symptoms might be found in either of two other ways. One possibility would involve expanding the search for subtypes beyond symptoms of an index episode. Included here, for example, could be information about temporally primary comorbid mental disorders (e.g., early-onset distress, fear-circuitry, or impulse-control disorders), physical disorders (e.g., metabolic syndrome), socio-demographics, and (neuro)biological factors to define subtypes. We purposefully did not include such expansions here, as we wanted to focus on subtypes defined by index episode symptoms, but future analysis should do so to broaden the search for subtypes to include these other predictors. It would be interesting for future research to examine the possibility that the significant association found here between later-AOO-anxious-suicidal depression in the index episode with later chronicity but no other outcome might reflect the importance of a late-onset depression subtype that might occur in conjunction with a physical comorbidity, such as cardio-metabolic illness[55] associated with episodes of long duration but not high persistence or severity. Such a possibility can only be examined by broadening the search for subtypes to include comorbid physical disorders. The potential value of expanding the search for subtypes to include information about biomarkers is illustrated in recent studies showing that the course of atypical and melancholic depression is differentially predicted by HPA-axis, metabolic syndrome and inflammatory parameters[56] and that inflammatory dysregulation is associated with the onset of ‘mixed state depression’.[57] Such analyses have the potential to discover clinically meaningful and biologically valid disease clusters across a range of clinically relevant outcomes, an approach consistent with the recent call for what has been referred to as a stratified medicine approach[58] that bypasses the search for a gold standard and focuses instead on the discovery of subtypes associated with a range of clinically meaningful outcomes. A second possibility would be to look more closely within the high-risk cluster found in our analysis to search for embedded subtypes. To understand this suggestion it is important to recognize that the clusters we discovered cannot themselves be thought of as subtypes in the classical sense because they were discovered by clustering predicted outcome scores rather than the predictors themselves. A great many different combinations of predictors could yield the same predicted outcome scores. This means that further effort is needed to define subtypes within the high-risk cluster by considering multivariate profiles among the predictors that determine cluster membership so as to take into consideration the differential importance of these predictors within and across outcomes. No attempt was made to do this here, but it is clearly something that warrants future investigation in future studies based on the analysis of a more complete set of predictors. It would also be useful, finally, if future studies expanded the range of outcomes considered here. The four outcomes in our analysis were selected purely based on availability. Given the discovery that predicted values are strongly correlated across these outcomes, it would be useful to develop an understanding of the range of outcomes over which this consistency occurs. Such an investigation could be carried out informally using the simple correlational methods used here, or a more formal approach might be conceived along the lines of the canonical regression models used to study latent mediators in the development of comorbidity among mental disorders.[59-61] Or it might be possible to address this issue by adapting the data mining methods developed to discover what have been called master regulators[62] in molecular genetic studies of physical disorders.[63-65] Regardless of method, though, the discovery of common predictors of multiple indicators of persistence, chronicity, and severity call out for a more diverse and integrated analysis of clusters and within-cluster subtypes among the predictors of such outcomes. In thinking of these future developments, it is important to recognize that the recursive partitioning methods used here require a much larger sample size than is likely to exist in prospective clinical samples. This means that the most feasible way to extend the current results in prospective clinical studies would be to evaluate the significance of the synergistic symptom profiles found here rather than to attempt independent data mining exercises, although independent Lasso and cluster analyses using larger sets of predictors (possibly including measures of endophenotypes) and alternative indicators of outcomes would be quite feasible in such studies. Although it is unlikely that clinicians would be willing to collect such data for purposes of making the subtyping distinctions made here, it is conceivable that future studies will document powerful effects of other predictors that could be examined using similar methods and shown to have sufficiently important clinical implications that it would motivate clinicians to collect such information as a routine part of their initial evaluations to guide treatment planning. The technology described here holds great promise in facilitating analyses aimed at documenting such predictors. CONCLUSION Despite our analysis being based on retrospective reports, our results suggest that useful symptom-based MDD subtyping distinctions can be made with data mining methods that focus on prediction rather than internal consistency and that the resulting subtypes have meaningful relationships with course of illness. The practical value of this approach, though, can only be judged by replication with prospective data, ideally expanding the analysis to use a wider range of predictors and outcomes. ACKNOWLEDGEMENTS The World Health Organization World Mental Health (WMH) Survey Initiative is supported by the National Institute of Mental Health (NIMH; R01 MH070884), the John D. and Catherine T. MacArthur Foundation, the Pfizer Foundation, the US Public Health Service (R13-MH066849, R01-MH069864, and R01 DA016558), the Fogarty International Center (FIRCA R03-TW006481), the Pan American Health Organization, Eli Lilly & Company Foundation, Ortho-McNeil Pharmaceutical, Inc., GlaxoSmithKline, Sanofi Aventis and Bristol-Myers Squibb. Peter de Jonge is supported by a VICI grant (no: 91812607) from the Netherlands Research Foundation (NWO-ZonMW). We thank the WMH staff for assistance with instrumentation, fieldwork, and data analysis. A complete list of WMH publications can be found at http://www.hcp.med.harvard.edu/wmh/. Each WMH country obtained funding for its own survey. The São Paulo Megacity Mental Health Survey is supported by the State of São Paulo Research Foundation (FAPESP) Thematic Project Grant 03/00204-3. The Bulgarian Epidemiological Study of common mental disorders EPIBUL is supported by the Ministry of Health and the National Center for Public Health Protection. The Beijing, Peoples Republic of China World Mental Health Survey Initiative is supported by the Pfizer Foundation. The Shenzhen, People's Republic of China Mental Health Survey is supported by the Shenzhen Bureau of Health and the Shenzhen Bureau of Science, Technology, and Information. The Colombian National Study of Mental Health (NSMH) is supported by the Ministry of Social Protection. Implementation of the Iraq Mental Health Survey (IMHS) and data entry were carried out by the staff of the Iraqi MOH and MOP with direct support from the Iraqi IMHS team with funding from both the Japanese and European Funds through United Nations Development Group Iraq Trust Fund (UNDG ITF). The Israel National Health Survey is funded by the Ministry of Health with support from the Israel National Institute for Health Policy and Health Services Research and the National Insurance Institute of Israel. The World Mental Health Japan (WMHJ) Survey is supported by the Grant for Research on Psychiatric and Neurological Diseases and Mental Health (H13-SHOGAI-023, H14-TOKUBETSU-026, H16-KOKORO-013) from the Japan Ministry of Health, Labour and Welfare. The Lebanese National Mental Health Survey (L.E.B.A.N.O.N.) is supported by the Lebanese Ministry of Public Health, the WHO (Lebanon), National Institute of Health / Fogarty International Center (R03 TW006481-01), Sheikh Hamdan Bin Rashid Al Maktoum Award for Medical Sciences, anonymous private donations to IDRAAC, Lebanon, and unrestricted grants from AstraZeneca, Eli Lilly, GlaxoSmithKline, Hikma Pharm, Pfizer, Roche, Sanofi-Aventis, Servier and Novartis. The Mexican National Comorbidity Survey (MNCS) is supported by The National Institute of Psychiatry Ramon de la Fuente (INPRFMDIES 4280) and by the National Council on Science and Technology (CONACyT-G30544- H), with supplemental support from the PanAmerican Health Organization (PAHO). Te Rau Hinengaro: The New Zealand Mental Health Survey (NZMHS) is supported by the New Zealand Ministry of Health, Alcohol Advisory Council, and the Health Research Council. The Nigerian Survey of Mental Health and Wellbeing (NSMHW) is supported by the WHO (Geneva), the WHO (Nigeria), and the Federal Ministry of Health, Abuja, Nigeria. The Northern Ireland Study of Mental Health was funded by the Health & Social Care Research & Development Division of the Public Health Agency. The Portuguese Mental Health Study was carried out by the Department of Mental Health, Faculty of Medical Sciences, NOVA University of Lisbon, with collaboration of the Portuguese Catholic University, and was funded by Champalimaud Foundation, Gulbenkian Foundation, Foundation for Science and Technology (FCT) and Ministry of Health. The Romania WMH study projects “Policies in Mental Health Area” and “National Study regarding Mental Health and Services Use” were carried out by National School of Public Health & Health Services Management (former National Institute for Research & Development in Health, present National School of Public Health Management & Professional Development, Bucharest), with technical support of Metro Media Transilvania, the National Institute of Statistics – National Centre for Training in Statistics, SC. Cheyenne Services SRL, Statistics Netherlands and were funded by Ministry of Public Health (former Ministry of Health) with supplemental support of Eli Lilly Romania SRL. The Ukraine Comorbid Mental Disorders during Periods of Social Disruption (CMDPSD) study is funded by the US National Institute of Mental Health (RO1-MH61905). The US National Comorbidity Survey Replication (NCS-R) is supported by the National Institute of Mental Health (NIMH; U01-MH60220) with supplemental support from the National Institute of Drug Abuse (NIDA), the Substance Abuse and Mental Health Services Administration (SAMHSA), the Robert Wood Johnson Foundation (RWJF; Grant 044708), and the John W. Alden Trust. Additional support for preparation of this report was provided by Janssen Pharmaceuticals. Figure 1 Mean predicted outcome scores in the three-cluster through eight-cluster k-means1 Per = the percentile-transformed predicted score on the persistence outcome variable; Chr = the percentile-transformed predicted score on the chronicity outcome; Hos = the percentile-transformed cumulative predicted probability of hospitalization; Dis = the percentile-transformed predicted probability of disability. 1k-means cluster analysis of percentile-transformed predicted scores on the four outcomes for all respondents based on the Lasso GLM Table 1 WMH sample characteristics by World Bank income categoriesa Sample size Country by income category Surveyb Sample characteristicsc Field dates Age range Response rated Total With MDD I. High-income countries Israel NHS Nationally representative. 2002-4 21-98 72.6 4,859 284 Japan WMHJ2002-2006 Eleven metropolitan areas. 2002-6 20-98 55.1 4,129 219 New Zealandf NZMHS Nationally representative. 2003-4 18-98 73.3 12,790 1,908 N. Ireland NISHS Nationally representative. 2004-7 18-97 68.4 4,340 423 Portugal NMHS Nationally representative. 2008-9 18-81 57.3 3,849 379 United States NCS-R Nationally representative. 2002-3 18-99 70.9 9,282 1,562 Total 67.9 (39,249) (4,775) II. Upper-middle income countries Brazil – São Paulo São Paulo Megacity São Paulo metropolitan area. 2005-7 18-93 81.3 5,037 408 Bulgaria NSHS Nationally representative. 2003-7 18-98 72.0 5,318 283 Lebanon L.E.B.A.N.O.N Nationally representative. 2002-3 18-94 70.0 2,857 267 Mexico M-NCS All urban areas of the country (approximately 75% of the total national population). 2001-2 18-65 76.6 5,782 397 Romania RMHS Nationally representative. 2005-6 18-96 70.9 2,357 54 Total 74.8 (21,351) (1,409) III. Low and lower-middle income countries Colombia NSMH All urban areas of the country (approximately 73% of the total national population) 2003 18-65 87.7 4,426 476 Iraq IMHS Nationally representative. 2006-7 18-96 95.2 4,332 193 Nigeria NSMHW 21 of the 36 states in the country, representing 57% of the national population. The surveys were conducted in Yoruba, Igbo, Hausa and Efik languages. 2002-3 18-100 79.3 6,752 176 PRCe – Beijing/Shanghai B-WMH/S-WMH Beijing and Shanghai metropolitan areas. 2002-3 18-70 74.7 5,201 151 PRCe – Shenzhenf Shenzhen Shenzhen metropolitan area. Included temporary residents as well as household residents. 2006-7 18-88 80.0 7,132 452 Ukrainef CMDPSD Nationally representative. 2002 18-91 78.3 4,724 629 Total 81.4 (32,567) (2,077) a The World Bank. (2012). Data. Accessed June 5, 2012 at: http://data.worldbank.org/country. b NSMH (The Colombian National Study of Mental Health); IMHS (Iraq Mental Health Survey); NSMHW (The Nigerian Survey of Mental Health and Wellbeing); B-WMH (The Beijing World Mental Health Survey); S-WMH (The Shanghai World Mental Health Survey); CMDPSD (Comorbid Mental Disorders during Periods of Social Disruption); NSHS (Bulgaria National Survey of Health and Stress); LEBANON (Lebanese Evaluation of the Burden of Ailments and Needs of the Nation); M-NCS (The Mexico National Comorbidity Survey); RMHS (Romania Mental Health Survey); NHS (Israel National Health Survey); WMHJ2002-2006 (World Mental Health Japan Survey); NZMHS (New Zealand Mental Health Survey); NISHS (Northern Ireland Study of Health and Stress); NMHS (Portugal National Mental Health Survey); NCS-R (The US National Comorbidity Survey Replication). c Most WMH surveys are based on stratified multistage clustered area probability household samples in which samples of areas equivalent to counties or municipalities in the US were selected in the first stage followed by one or more subsequent stages of geographic sampling (e.g., towns within counties, blocks within towns, households within blocks) to arrive at a sample of households, in each of which a listing of household members was created and one or two people were selected from this listing to be interviewed. No substitution was allowed when the originally sampled household resident could not be interviewed. These household samples were selected from Census area data in all countries other than France (where telephone directories were used to select households) and the Netherlands (where postal registries were used to select households). Several WMH surveys (Belgium, Germany, Italy, Poland) used municipal resident registries to select respondents without listing households. The Japanese sample is the only totally un-clustered sample, with households randomly selected in each of the 11 metropolitan areas and one random respondent selected in each sample household. 19 of the 26 surveys are based on nationally representative household samples. d The response rate is calculated as the ratio of the number of households in which an interview was completed to the number of households originally sampled, excluding from the denominator households known not to be eligible either because of being vacant at the time of initial contact or because the residents were unable to speak the designated languages of the survey. The weighted average response rate is 73.7%. e People's Republic of China f For the purposes of cross-national comparisons, we limit the sample to those 18+. Table 2 Lasso GLM coefficients to predict subsequent course of DSM-IV major depressive disorder based on characteristics of the incident episodea Percent of years in episode Any episode IDRb Episode lasting most of year IDRb Hospitalized HRb Disabled ORb I. Criterion A symptoms of major depression Severe dysphoriac (ICD-10 severity specifier) 1.1 1.4 1.7 1.2 Anhedonia 1.1 Weight loss 0.9 Weight gain 1.1 0.8 Insomnia 1.3 Hypersomnia 1.5 Psychomotor agitation 1.2 Psychomotor retardation 1.2 Suicidality 1.1 1.6 1.5 II. Symptoms of dysthymia Inability to cope 1.9 1.4 III. Sym ptoms of anxiety Irritability 1.1 0.8 1.2 Panic 1.1 1.3 1.5 IV. Symptoms of mixed episode Racing thoughts 0.8 High energy 1.2 V. Multivariate symptom profiles AOO < 19 and suicidality 1.3 AOO < 19 and anxiety 1.1 1.3 1.2 AOO ≥ 35 and suicidality and anxiety 1.2 VI. Other predictorsd Endogenous 0.7 Parental history of depression 1.2 1.2 1.2 N (2,869) (3,958) (6,465) (8,261) a Based on Lasso GLM penalized regression models, with the size of penalty determined by 10-fold cross-validation to select the penalty yielding cross-validating results with minimum mean squared prediction error. No Confidence intervals are reported because standard errors of such simulated models are biased. See the text for a discussion of differences in link functions and sample sizes. b IDR = Incidence density ratio; HR=Hazard ratio; OR = Odds-ratio c This is not the DSM-IV Criterion A symptom of dysphoria but the ICD-10 symptom for somatic depression that the dysphoria is so severe that the patient has a lack of emotional reaction to events or activities that normally produce an emotional response. The DSM-IV symptom of dysphoria, in comparison, was not a significant predictor in any of the models. d An additional 12 predictors were included in the Lasso GLM models that had coefficients of either zero or near zero across all outcomes. These predictors are dysphoria, fatigue/loss of energy, worthlessness or excessive guilt, diminished ability to concentrate or indecisiveness, social withdrawal, nervousness-worry-anxiety, multivariate symptoms profiles of childhood (before age 13) onset with anxiety and/or suicidality, multivariate symptom profiles of AOO before 19 with anxiety and suicidality, other multivariate symptom profiles of AOO either before 13 or before 19 or after 34 with either anxiety and/or suicidality, little need for sleep, total number of symptoms, age of onset, and time between onset and age at interview. Table 3 Associations of cluster membership with positive screening characteristics Relative-riska in the high-risk cluster vs. All othersb Those in the low-risk cluster Positive predictive valuec Sensitivityc Est (95% CI) Est (95% CI) % (se) % (se) Percent of years in any episode Top 5 percentile 2.7 (1.7-3.7) 3.3 (1.7-4.9) 7.9 (0.8) 60.0 (4.3) Top 10 percentile 2.5 (1.9-3.1) 3.1 (1.9-4.2) 16.4 (1.1) 58.0 (3.0) Percent of years in episodes lasting most of the year Top 5 percentile 2.7 (1.8-3.5) 4.0 (1.9-6.1) 8.2 (0.8) 59.3 (3.5) Top 10 percentile 2.1 (1.6-2.5) 2.5 (1.7-3.3) 14.4 (0.9) 52.9 (2.5) Hospitalized 5.1 (3.4-6.7) 10.4 (4.3-16.6) 9.6 (0.8) 69.7 (3.5) Disabled 4.6 (2.5-6.7) 11.3 (2.9-19.8) 3.4 (0.4) 67.1 (4.9) Summary outcomes using top 5 percentile Anyd 3.2 (2.4-3.9) 4.5 (3.0-6.0) 25.4 (1.6) 64.2 (3.0) Multiplee 4.3 (1.8-6.9) 6.5 (0.9-12.1) 5.6 (0.8) 71.1 (5.6) Summary outcomes using top 10 percentile Anyd 2.6 (2.1-3.1) 3.1 (2.3-4.0) 33.0 (1.9) 59.3 (2.6) Multiplee 3.8 (2.3-5.3) 4.8 (1.9-7.7) 11.2 (1.1) 68.4 (4.3) a Relative-risk is the ratio of the percent of respondents in the high-risk cluster that experienced the adverse outcome compared to the percent in the other clusters or in the low -risk cluster. b Others = Respondents in either the intermediate-risk or low -risk clusters. c Positive Predictive Value is the percent of respondents in the high-risk cluster that experienced the adverse outcome; Sensitivity is the percent of observed adverse outcomes that occurred in the high-risk cluster. d These are dichotomous variables that differentiate respondents who had one or more of the following four adverse outcomes: in the top 5 percentile (or 10 percentile) of years with episodes, in the top 5 percentile (or 10 percentile) of years w ith episodes lasting most of the year, hospitalized, or disabled. e These are dichotomous variables that differentiate respondents who had two or more of the four adverse outcomes. Table 4 Symptoms associated with the high-risk, intermediate-risk, and low risk clusters High-risk Intermediate-risk Low-risk Total % (se) % (se) % (se) % (se) χ 2 2 Cramer's V I. Criterion A symptoms of major depression Dysphoria 99.6 (0.2) 98.9 (0.2) 96.1 (0.4) 98.2 (0.2) 57.0 0.11 Severe dysphoria 97.8 (0.4) 91.0 (0.7) 56.0 (1.1) 81.3 (0.5) 713.6* 0.47 Anhedonia 95.0 (0.5) 88.7 (0.7) 79.0 (0.9) 87.3 (0.5) 205.1* 0.20 Weight loss 74.9 (1.0) 74.8 (1.0) 76.4 (1.0) 75.4 (0.6) 1.8 0.02 Weight gain 18.0 (0.8) 14.8 (0.8) 12.0 (0.7) 14.8 (0.5) 28.5* 0.07 Insomnia 85.3 (0.9) 83.1 (0.8) 78.5 (0.9) 82.2 (0.5) 28.4* 0.07 Hypersomnia 10.5 (0.8) 9.0 (0.6) 9.7 (0.6) 9.7 (0.4) 2.4 0.02 Psychomotor agitation 16.9 (0.8) 17.7 (0.9) 14.8 (0.8) 16.5 (0.5) 6.4* 0.03 Psychomotor retardation 67.7 (1.0) 55.0 (1.0) 41.0 (1.1) 54.2 (0.6) 247.9* 0.22 Fatigue/loss of energy 89.4 (0.8) 86.6 (0.8) 81.2 (0.8) 85.7 (0.5) 54.4* 0.10 Worthlessness or excessive guilt 98.2 (0.3) 88.4 (0.7) 68.3 (1.0) 84.6 (0.4) 724.0* 0.34 Diminished ability to concentrate/indecisiveness 96.2 (0.4) 91.7 (0.6) 84.5 (0.7) 90.7 (0.4) 167.9* 0.16 Suicidality 98.8 (0.2) 76.9 (1.0) 38.2 (1.1) 70.6 (0.6) 1449.6* 0.54 II. Symptoms of dysthymia Inability to cope 86.0 (0.8) 59.5 (1.0) 24.0 (1.0) 55.7 (0.7) 1296.2* 0.50 Social withdrawal/tearfulness 99.2 (0.2) 94.4 (0.5) 88.0 (0.8) 93.7 (0.3) 261.3* 0.19 III. Symptoms of anxiety Irritability 77.7 (1.0) 62.9 (1.0) 41.1 (1.1) 60.1 (0.6) 541.3* 0.30 Nervousness-worry-anxiety 88.4 (0.9) 76.2 (1.0) 56.6 (1.1) 73.3 (0.6) 369.6* 0.29 Panic 66.4 (1.1) 38.1 (1.1) 13.5 (0.7) 38.5 (0.6) 842.1* 0.44 IV. Symptoms of mixed episode Little need for sleep 47.1 (1.2) 47.0 (1.2) 41.7 (1.1) 45.2 (0.7) 14.2* 0.05 Racing thoughts 10.6 (0.8) 12.4 (0.7) 15.7 (0.9) 12.9 (0.4) 16.6* 0.06 High energy 3.2 (0.4) 3.0 (0.4) 3.3 (0.4) 3.1 (0.2) 0.3 0.01 V. Multivariate symptom profiles AOO < 19 and suicidality 48.6 (1.2) 17.7 (0.9) 4.9 (0.5) 22.9 (0.6) 853.0* 0.43 AOO < 19 and anxiety 48.9 (1.2) 24.4 (1.1) 9.9 (0.7) 27.1 (0.6) 658.4* 0.35 AOO < 19 and suicidality and anxiety 47.8 (1.2) 14.9 (0.9) 2.1 (0.3) 20.7 (0.5) 919.8* 0.46 AOO ≥ 35 and suicidality and anxiety 16.4 (0.8) 21.9 (0.9) 9.1 (0.6) 15.9 (0.4) 179.2* 0.15 VI. Other predictors Endogenous 17.0 (0.9) 16.2 (0.9) 14.7 (0.8) 15.9 (0.6) 4.3 0.03 Parental history of depression 9.9 (0.7) 5.8 (0.6) 3.2 (0.4) 6.2 (0.4) 63.8* 0.11 N (2,520) (2,899) (2,842) (8,261) * Significant at the .05 level, two-sided test Disclosure Dr. Wilcox is an employee of Janssen Pharmaceuticals. Dr. Kessler has been a consultant for AstraZeneca, Analysis Group, Bristol-Myers Squibb, Cerner-Galt Associates, Eli Lilly & Company, GlaxoSmithKline Inc., HealthCore Inc., Health Dialog, Hoffman-LaRoche, Inc., Integrated Benefits Institute, J & J Wellness & Prevention, Inc., John Snow Inc., Kaiser Permanente, Lake Nona Institute, Matria Inc., Mensante, Merck & Co, Inc., Ortho-McNeil Janssen Scientific Affairs, Pfizer Inc., Primary Care Network, Research Triangle Institute, Sanofi-Aventis Groupe, Shire US Inc., SRA International, Inc., Takeda Global Research & Development, Transcept Pharmaceuticals Inc., and Wyeth-Ayerst. Kessler has served on advisory boards for Appliance Computing II, Eli Lilly & Company, Mindsite, Ortho-McNeil Janssen Scientific Affairs, Johnson & Johnson, Plus One Health Management and Wyeth-Ayerst. Kessler has had research support for his epidemiological studies from Analysis Group Inc., Bristol-Myers Squibb, Eli Lilly & Company, EPI-Q, GlaxoSmithKline, Johnson & Johnson Pharmaceuticals, Ortho-McNeil Janssen Scientific Affairs., Pfizer Inc., Sanofi-Aventis Groupe, Shire US, Inc., and Walgreens Co. Kessler owns 25% share in DataStat, Inc. The views and opinions expressed in this report are those of the authors and should not be construed to represent the views or policies of any of the sponsoring organizations, agencies, or the World Health Organization. REFERENCES 1 Cooper C Jones L Dunn E Clinical presentation of postnatal and non-postnatal depressive episodes. Psychol Med 2007 37 1273 1280 17349101 2 Cooper PJ Murray L Course and recurrence of postnatal depression. Evidence for the specificity of the diagnostic concept. Br J Psychiatry 1995 166 191 195 7728362 3 Fink M Rush AJ Knapp R DSM melancholic features are unreliable predictors of ECT response: a CORE publication. J ECT 2007 23 139 146 17804986 4 Uher R Dernovsek MZ Mors O Melancholic, atypical and anxious depression subtypes and outcome of treatment with escitalopram and nortriptyline. J Affect Disord 2011 132 112 120 21411156 5 Andreasen NC Grove WM The classification of depression: traditional versus mathematical approaches. Am J Psychiatry 1982 139 45 52 7055276 6 Romera I Delgado-Cohen H Perez T Factor analysis of the Zung self-rating depression scale in a large sample of patients with major depressive disorder in primary care. BMC Psychiatry 2008 8 4 18194524 7 Lamers F Burstein M He JP Structure of major depressive disorder in adolescents and adults in the US general population. Br J Psychiatry 2012 201 143 150 22700082 8 Baumeister H Parker G Meta-review of depressive subtyping models. J Affect Disord 2012 139 126 140 21885128 9 Carragher N Adamson G Bunting B Subtypes of depression in a nationally representative sample. J Affect Disord 2009 113 88 99 18644628 10 van Loo HM de Jonge P Romeijn JW Data-driven subtypes of major depressive disorder: a systematic review. BMC Med 2012 10 156 23210727 11 Glahn DC Curran JE Winkler AM High dimensional endophenotype ranking in the search for major depression risk genes. Biol Psychiatry 2012 71 6 14 21982424 12 Hasler G Northoff G Discovering imaging endophenotypes for major depression. Mol Psychiatry 2011 16 604 619 21602829 13 Strobl C Malley J Tutz G An introduction to recursive partitioning: rationale, application, and characteristics of classification and regression trees, bagging, and random forests. Psychol Methods 2009 14 323 348 19968396 14 Zhang H Singer BH Recursive Partitioning and Applications 2010 Springer New York, NY Second Edition. 15 Hastie T Tibshirani R Friedman J The Elements of Statistical Learning: Data Mining, Inference, and Prediction 2009 Springer New York, NY Second Edition 16 van der Laan MJ Rose S Targeted Learning: Causal Inference for Observational and Experimental Data 2011 Springer New York, NY 17 Chang YJ Chen LJ Chung KP Risk groups defined by Recursive Partitioning Analysis of patients with colorectal adenocarcinoma treated with colorectal resection. BMC Med Res Methodol 2012 12 2 22214198 18 Chao ST Koyfman SA Woody N Recursive partitioning analysis index is predictive for overall survival in patients undergoing spine stereotactic body radiation therapy for spinal metastases. Int J Radiation Oncology Biol Phys 2012 82 1738 1743 19 Andreescu C Mulsant BH Houck PR Empirically derived decision trees for the treatment of late-life depression. Am J Psychiatry 2008 165 855 862 18450930 20 Jain FA Hunter AM Brooks JO 3rd Predictive socioeconomic and clinical profiles of antidepressant response and remission. Depress Anxiety 2013 30 624 630 23288666 21 Nelson JC Zhang Q Deberdt W Predictors of remission with placebo using an integrated study database from patients with major depressive disorder. Curr Med Res Opin 2012 28 325 334 22292447 22 Rabinoff M Kitchen CM Cook IA Evaluation of quantitative EEG by classification and regression trees to characterize responders to antidepressant and placebo treatment. Open Med Inform J 2011 5 1 8 21603560 23 Riedel M Moller HJ Obermeier M Clinical predictors of response and remission in inpatients with depressive syndromes. J Affect Disord 2011 133 137 149 21555156 24 Ilgen MA Downing K Zivin K Exploratory data mining analysis identifying subgroups of patients with depression who are at high risk for suicide. J Clin Psychiatry 2009 70 1495 1500 20031094 25 Musil R Zill P Seemuller F Genetics of emergent suicidality during antidepressive treatment--data from a naturalistic study on a large sample of inpatients with a major depressive episode. Eur Neuropsychopharmacol 2013 23 663 674 23063133 26 Seemuller F Riedel M Obermeier M The controversial link between antidepressants and suicidality risks in adults: data from a naturalistic study on a large sample of in-patients with a major depressive episode. Int J Neuropsychopharmacol 2009 12 181 189 18662490 27 Kessler RC Üstün TB The WHO World Mental Health Surveys: Global Perspectives on the Epidemiology of Mental Disorders 2008 Cambridge University Press New York, NY 28 Alonso J Chatterji S He Y The Burden of Mental Disorders: Global Perspectives from the WHO World Mental Health Surveys 2013 Cambridge University Press New York, NY 29 Nock MK Borges G Ono Y Suicide: Global Perspectives from the WHO World Mental Health Surveys 2012 Cambridge University Press New York, NY 30 Von Korff MR Scott KM Gureje O Global Perspectives on Mental-Physical Comorbidity In the WHO World Mental Health Surveys 2009 Cambridge University Press New York, NY 31 World Bank Data, Country and Lending Groups 2009 Available from: http://go.worldbank.org/D7SN0B8YU0 2009 May 12 32 Heeringa SG Wells EJ Hubbard F Kessler RC Üstün TB Sample designs and sampling procedures. The WHO World Mental Health Surveys: Global Perspectives on the Epidemiology of Mental Disorders 2008 14 32 Cambridge University Press New York, NY 33 Harkness J Pennell BE Villar A Kessler RC Üstün TB Translation procedures and translation assessment in the world mental health survey initiative. The WHO World Mental Health Surveys: Global Perspectives on the Epidemiology of Mental Disorders 2008 91 113 Cambridge University Press New York, NY 34 Pennell B-E Mneimneh Z Bowers A Kessler RC Üstün TB Implementation of the world mental health surveys. The WHO World Mental Health Surveys: Global Perspectives on the Epidemiology of Mental Disorders 2008 33 57 Cambridge University Press New York, NY 35 Kessler RC Üstün TB The World Mental Health (WMH) Survey Initiative Version of the World Health Organization (WHO) Composite International Diagnostic Interview (CIDI). Int J Methods Psychiatr Res 2004 13 93 121 15297906 36 Haro JM Arbabzadeh-Bouchez S Brugha TS Concordance of the Composite International Diagnostic Interview Version 3.0 (CIDI 3.0) with standardized clinical assessments in the WHO World Mental Health surveys. Int J Methods Psychiatr Res 2006 15 167 180 17266013 37 Altman D Machin D Bryant T Statistics with Confidence: Confidence intervals and statistical guidelines 2000 BMJ Books London, UK Second Edition. 38 Endicott J Andreasen N Spitzer RL Family History Research Diagnostic Criteria 1978 Biometrics Research, New York State Psychiatric Institute New York, NY 39 Thernau T Atkinson B Ripley B Rpart: Recursive Partioning. R Package 4.1-0 2012 40 Draper N Smith H Applied Regression Analysis 1981 Wiley New York, NY Second Edition. 41 Berk RA Regression Analysis: A Constructive Critique 2003 Sage Newbury Park, CA 42 Tibshirani R Regression shrinkage and selection via the lasso. J Roy Stat Soc B 1996 58 267 288 43 Berk RA Statistical Learning from a Regression Perspective 2008 Springer New York, NY 44 Friedman J Hastie T Tibshirani R Regularization paths for generalized linear models via coordinate descent. J Stat Softw 2010 33 1 22 20808728 45 R Core Team R: A language and environment for statistical computing 2013 Available from: http://www.R-project.org/ 2013 November 8 46 Chambless LE Diao G Estimation of time-dependent area under the ROC curve for long-term risk prediction. Stat Med 2006 25 3474 3486 16220486 47 Judd LL Schettler PJ Coryell W Overt irritability/anger in unipolar major depressive episodes: past and current characteristics and implications for long-term course. JAMA Psychiatry 2013 70 1171 1180 24026579 48 Penninx BW Beekman AT Smit JH The Netherlands Study of Depression and Anxiety (NESDA): rationale, objectives and methods. Int J Methods Psychiatr Res 2008 17 121 140 18763692 49 Bottomley C Nazareth I Torres-Gonzalez F Comparison of risk factors for the onset and maintenance of depression. Br J Psychiatry 2010 196 13 17 20044653 50 Patten SB Wang JL Williams JV Predictors of the longitudinal course of major depression in a Canadian population sample. Can J Psychiatry 2010 55 669 676 20964946 51 Spijker J de Graaf R Bijl RV Determinants of persistence of major depressive episodes in the general population. Results from the Netherlands Mental Health Survey and Incidence Study (NEMESIS). J Affect Disord 2004 81 231 240 15337327 52 Coryell W Fiedorowicz JG Solomon D Effects of anxiety on the long-term course of depressive disorders. Br J Psychiatry 2012 200 210 215 21984801 53 Gerrits MM Vogelzangs N van Oppen P Impact of pain on the course of depressive and anxiety disorders. Pain 2012 153 429 436 22154919 54 Wardenaar KJ Giltay EJ van Veen T Symptom dimensions as predictors of the two-year course of depressive and anxiety disorders. J Affect Disord 2012 136 1198 1203 22177741 55 Vogelzangs N Beekman AT Boelhouwer IG Metabolic depression: a chronic depressive subtype? Findings from the InCHIANTI study of older persons. J Clin Psychiatry 2011 72 598 604 21535996 56 Lamers F Vogelzangs N Merikangas KR Evidence for a differential role of HPA-axis function, inflammation and metabolic syndrome in melancholic versus atypical depression. Mol Psychiatry 2013 18 692 699 23089630 57 Becking K Boschloo L Vogelzangs N The association between immune activation and manic symptoms in patients with a depressive disorder. Transl Psychiatry 2013 3 e314 24150223 58 Kapur S Phillips AG Insel TR Why has it taken so long for biological psychiatry to develop clinical tests and what to do about it? Mol Psychiatry 2012 17 1174 1179 22869033 59 Kessler RC Cox BJ Green JG The effects of latent variables in the development of comorbidity among common mental disorders. Depress Anxiety 2011 28 29 39 21225850 60 Kessler RC Ormel J Petukhova M Development of lifetime comorbidity in the World Health Organization World Mental Health Surveys. Arch Gen Psychiatry 2011 68 90 100 21199968 61 Kessler RC Petukhova M Zaslavsky AM The role of latent internalizing and externalizing predispositions in accounting for the development of comorbidity among common mental disorders. Curr Opin Psychiatry 2011 24 307 312 21602684 62 Chan SS Kyba M What is a master regulator? J Stem Cell Res Ther 2013 3 e114 63 Rabinowitz JD Silhavy TJ Systems biology: metabolite turns master regulator. Nature 2013 500 283 284 23925111 64 Rayner BS Figtree GA Sabaretnam T Selective inhibition of the master regulator transcription factor egr-1 with catalytic oligonucleotides reduces myocardial injury and improves left ventricular systolic function in a preclinical model of myocardial infarction. J Am Heart Assoc 2013 2 e000023 23902638 65 Ryu B Kim DS Deluca AM Comprehensive expression profiling of tumor cell lines identifies molecular signatures of melanoma progression. PloS One 2007 2 e594 17611626
PMC005xxxxxx/PMC5125510.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 0376340 5287 J Surg Res J. Surg. Res. The Journal of surgical research 0022-4804 1095-8673 27884342 5125510 10.1016/j.jss.2016.08.035 NIHMS810047 Article Optimizing neurogenic potential of enteric neurospheres for treatment of neurointestinal diseases Cheng Lily S. ab Graham Hannah K. a Pan Wei Hua ac Nagy Nandor ad Carreon-Rodriguez Alfonso ae Goldstein Allan M. a Hotta Ryo a# a Department of Pediatric Surgery, Massachusetts General Hospital, Harvard Medical School, 55 Fruit St., Boston, MA, 02114, USA b Department of Surgery, University of California San Francisco, 500 Parnassus Ave., San Francisco, CA, 94143, USA c Department of Pediatric Surgery, Xinhua Hospital, Shanghai Jiaotong University School of Medicine, 280 Chongqing S Rd, Huangpu, Shanghai, China d Department of Anatomy, Histology and Embryology, Faculty of Medicine, Semmelweis University, Tuzolto St. 58, Budapest 1094, Hungary e Laboratorio de Genética y Biomarcadores, Instituto Nacional de Salud Pública, Av Universidad 655, Santa María Ahuacatitlán, 62100 Cuernavaca, Mor., Mexico # Corresponding author Pediatric Surgical Research Laboratories, Massachusetts General Hospital, 185 Cambridge St., CPZN 6215, Boston, MA 02114, Phone: (617) 643-3040, Fax: (617) 726-5057, [email protected] 13 8 2016 12 8 2016 12 2016 01 12 2017 206 2 451459 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Background Enteric neurospheres derived from postnatal intestine represent a promising avenue for cell replacement therapy to treat Hirschsprung disease and other neurointestinal diseases. We describe a simple method to improve the neuronal yield of spontaneously-formed gut-derived neurospheres. Materials and Methods Enteric neurospheres were formed from the small and large intestines of mouse and human subjects. Neurosphere size, neural crest cell content, cell migration, neuronal differentiation, and neuronal proliferation in culture were analyzed. The effect of supplemental neurotrophic factors, including glial-derived neurotrophic factor (GDNF) and endothelin-3 (ET3), was also assessed. Results Mouse small intestine-derived neurospheres contained significantly more P75-expressing neural crest-derived cells (49.9 ± 15.3 vs. 21.6 ± 11.9%, p<0.05) and gave rise to significantly more Tuj1-expressing neurons than colon-derived neurospheres (69.9 ± 8.6 vs. 46.2 ± 15.6%, p<0.05). A similar pattern was seen in neurospheres isolated from human small and large intestine (32.6 ± 17.5 vs. 10.2 ± 8.2% neural crest cells, p<0.05; 29.7 ± 16.4 vs. 16.0 ± 13.5% enteric neurons, p<0.05). The addition of GDNF to the culture media further improved the neurogenic potential of small intestinal neurospheres (75.9 ± 4.0 vs. 67.8 ± 5.8%, p<0.05) whereas ET3 had no effect. Conclusions Enteric neurospheres formed from small intestine and supplemented with GDNF yield an enriched population of neural crest-derived progenitor cells and give rise to a high density of enteric neurons. Enteric nervous system Enteric neural stem cells Neurospheres Glial cell-derived neurotrophic factor Endothelin-3 INTRODUCTION The enteric nervous system (ENS) is a complex network of neurons and glia which controls many essential functions of the gastrointestinal (GI) tract (1). Diseases of the ENS encompass a broad spectrum of common GI disorders including esophageal achalasia, gastroparesis, slow-transit constipation, and Hirschsprung disease (2). Current treatment for these diseases is palliative rather than curative, although cell therapy holds promise as a novel potential therapy for this group of diseases (3). Recent evidence has shown that enteric neural stem/progenitor cells (ENSCs) can be isolated from the adult gut and propagated in culture as enteric neurospheres (4), and that these neurospheres can be transplanted to aneural gut, where they give rise to new neurons that may improve bowel function (5). However, in addition to neural and glial progenitors, enteric neurospheres also contain many other, non-neuroglial cell types (6). Many techniques have been employed to improve the neurogenic potential of enteric neurospheres. We have previously shown that co-transplantation of neurospheres with a serotonin receptor agonist can enhance neuronal differentiation and proliferation (7). Modified dissection techniques to isolate the enteric nerve plexuses from donor tissue have been described in order to generate a purer population (8). Cell sorting using neural crest cell markers have also been used to enrich the progenitor population within neurospheres (6), but this would be difficult to apply clinically and may eliminate bystander cells important to supporting neuronal growth. Previous reports have noted that the small intestine (SI) contains more than double the number of myenteric neurons than the large intestine (LI) (9), likely owing to the longer length of the SI. The length and redundancy of the SI make it an attractive tissue source, but neurospheres derived from SI and LI have not been compared. Other factors, such as glial cell-derived neurotrophic factor (GDNF) and endothelin-3 (ET3), are both known to play an important role in ENS development, but their effect on postnatal gut-derived neurospheres has not been explored. ET3, through its receptor endothelin receptor type B (EDNRB), promotes ENSC proliferation, while GDNF, through the receptor tyrosine kinase, rearranged during transfection (RET), promotes ENSC migration, proliferation, and neuronal differentiation (10). Mutations in either the RET/GDNF or EDNRB/ET3 pathway cause Hirschsprung disease in mice and humans (11). The goal of this study was to identify the optimal source for ENSCs and to explore methods to optimize their neurogenic potential. Our study identifies clinically useful observations for the use of neuronal cell therapy for the treatment of enteric neuropathies. MATERIALS AND METHODS Generation of Mouse Enteric Neurospheres With approval from the Institutional Animal Care and Use Committee, neurospheres were generated from male and female 3-week old C57BL/6 mice (Jackson Labs, Bar Harbor, ME) according to previously published protocols (7, 12). The muscularis propria, which contains the myenteric plexus, was isolated from SI (duodenum through ileum; Fig. 1 A-B) and LI (post-cecal colon through anus, Fig. 1 D-E). Primary neurospheres were dissociated with Accutase (StemCell Technologies, Vancouver, BC) at 37 °C for 30 minutes and re-plated at 50,000 cells/mL to form secondary neurospheres (Fig. 1 C, F). Generation of Human Enteric Neurospheres With approval from the Institutional Review Board, 1-3 cm2 pieces of SI or LI tissue was obtained from 5 patients (1 month-old to 21 years-old) undergoing bowel resection, including SI and LI tissue from a 17 year-old male undergoing ileocecal resection (Table 1). Neurospheres were generated based on previously published protocols (4, 13). In brief, the muscularis propria was isolated and digested for 90 minutes at 37 °C in dispase (250 μg/mL; StemCell Technologies) and collagenase XI (1 mg/mL; Sigma Aldrich, St. Louis, MO) and then filtered through a 70 μm filter. Cells were cultured in a 1:1 mix of mouse conditioned media (obtained from the supernatant of cultured mouse neurospheres) and human proliferation media, consisting of Neurocult Human Basal Medium (StemCell Technologies) supplemented with 10% Neurocult Human Proliferation Supplement (StemCell Technologies), 20 ng/mL epidermal growth factor, 20 ng/mL basic fibroblast growth factor, 0.0002% Heparin, 50 μg/mL metronidazole (Sigma Aldrich), 2 μL/mL Primocin (Invitrogen, Carlsbad, CA). After 7 days, primary neurospheres were dissociated with Accutase at 37 °C for 30 minutes and re-plated at 50,000 cells/mL in a 96-well round bottom plate (Corning, Kennebunk, ME), which was centrifuged at 480 g for 2 minutes to encourage cell aggregation. Secondary neurospheres formed after 7 days in culture. Tissue Preparation and Immunohistochemistry Tissue preparation and immunohistochemistry were performed as previously described (7). Cells and tissues were fixed in 4% paraformaldehyde. For cryosection, sections were cut at 12 μm thickness with a Leica CM3050 S cryostat (Leica, Buffalo Grove, IL). For immunohistochemistry, cells and tissues were permeabilized with 0.1% Triton X-100 and blocked with 10% donkey serum for 30 minutes. Primary antibodies included human anti-neuronal nuclear antibody-1 (Hu; 1:16,000; generous gift from Dr. Vanda Lennon), mouse anti-neuronal class III β-tubulin (Tuj1; 1:500; Covance, Dedham, MA), rabbit anti-p75 neurotrophin receptor (P75; 1:500; Promega, Madison, WI), rabbit anti-S100 calcium-binding protein B (S100; 1:100; NeoMarkers, Fremont, CA), and rabbit anti-α-smooth muscle actin (SMA; 1:100; Abcam, Cambridge, MA). Secondary antibodies included donkey anti-mouse Alexa Fluor 488, donkey anti-rabbit Alexa Fluor 546, and donkey anti-human Alexa Fluor 546 (Life Technologies, Carlsbad, CA). Cell nuclei were stained with DAPI (Vector Labs, Burlingame, CA). EdU incorporation was detected using the Click-iT EdU Imaging Kit (Invitrogen, Carlsbad, CA). Images were taken using a Nikon Eclipse TS100 or 80i microscope (Nikon Instruments, Melville, NY). Characterization of Neurospheres Cell migration was quantified by plating secondary neurospheres onto slides coated with 20 μg/mL fibronectin (Biomedical Technologies, Ward Hill, MA) and culturing for 7 days in either human or mouse differentiation media, consisting of Neurocult Human or Mouse Basal Medium supplemented with 10% Neurocult Human or Mouse Differentiation Supplement (StemCell Technologies), 10% fetal bovine serum, and 100 U/mL penicillin-streptomycin. Neurons were visualized using Tuj1 immunoreactivity and cell migration was measured as the distance from the edge of the neurosphere to the farthest neuronal nucleus using ImageJ software (National Institutes of Health, Bethesda, Maryland). The average distance to the 3 farthest neurons in each orthogonal direction was measured, accounting for 12 measurements per neurosphere with at least 3 neurospheres analyzed per condition. To quantify the number of neural crest-derived cells in the neurosphere, 7 day-old secondary neurospheres were dissociated with Accutase incubation for 30 minutes at 37 °C and centrifuged (800 g for 2 minutes; Shandon Cytospin 3) onto a poly-L-lysine-coated slide. Neural crest-derived cells were visualized using P75 immunoreactivity. To quantify neurogenesis and cell proliferation, dissociated neurospheres were re-plated at 5,000 cells/mL in differentiation media onto fibronectin-coated slides. Cells were cultured for 7 days and 10 μM EdU was added to the culture media 24 hours prior to fixation. Neurons were visualized with Tuj1 immunoreactivity and proliferation rate determined based on the proportion of cells incorporating EdU. To test the effect of supplemental factors, ET3 (100 ng/mL) and/or GDNF (50 ng/mL) were added to secondary neurospheres at the beginning of the culture period and again following neurosphere dissociation, at the time of plating onto fibronectin-coated slides. Four to 10 non-overlapping images were taken of each cell preparation and each preparation was repeated at least twice. Numbers of cells were counted using ImageJ software. Quantitative PCR Total mRNA was extracted from conditioned media of secondary neurospheres using the RNeasy Mini Kit (Qiagen, Santa Clarita, CA) and cDNA synthesized with the Superscript III Reverse Transcription Kit (Invitrogen). Gdnf, Et3, Ret, and Ednrb expression levels were measured using qPCR with Gapdh as the internal standard. Relative expression was calculated by ΔCt. The primers used where: Gdnf forward, GCCGGACGGGA; Gdnf reverse, CGTCATCAAAC; Et3 forward, GTGAGAGGATT; Et3 reverse, TGTCCTTGTAA; Ret forward, TCAAGGGATGCTTACTGGGAG; Ret reverse, GGTAGACGCCATAGAGATGCT; Ednrb forward, TCGGACTACAAAGGAAAGCC; Ednrb reverse, AGTAGAAACTGAACAGCCACC. Statistics Data are presented as mean ± standard deviation. Neuronal migration, neurosphere size, and relative expression were compared using Student's t-test. All other results were compared using repeated measures ANOVA. Intestinal tissue was pooled from 3-5 animals for each murine experiment. The number of replicates for each neurosphere culture condition is stated in the Results section. Statistical significance was considered at p<0.05. Statistical analysis was performed using JMP version 12 (SAS Institute, Cary, NC). RESULTS Neurospheres derived from mouse and human SI contain more ENSCs and generate more neurons than LI-derived neurospheres ENSCs were isolated from the muscular layers of the SI and LI of adult mice (Fig. 1A-B, D-E) and passaged to form secondary neurospheres (Fig. 1C, F). No difference was observed in the average diameter of neurospheres from SI and LI (153 ± 35 vs. 147 ± 36 μm, n=30; Fig. 1G). SI-derived neurospheres did contain significantly more P75+ neural crest-derived cells (49.9 ± 15.3%) than LI-derived neurospheres (21.6 ± 11.9%, p<0.05, n=5; Fig. 1H), representing a 2.3-fold greater density of P75+ cells from the SI. Neurons from both SI and LI migrated a similar distance in culture (1.55 ± 0.34 vs. 1.82 ± 0.23 mm, n=3; Fig. 1I). SI-derived neurospheres from mice gave rise to significantly more Tuj1+ neurons than LI (69.9 ± 8.6 vs. 46.2 ± 15.6%, p<0.05, n=7; Fig. 2A), but the rate of neuronal proliferation did not differ significantly (3.8 ± 2.6 vs. 6.4 ± 6.3%, n=7; Fig. 2B). In contrast, the percentage of glial cells did not differ between SI- and LI-derived neurospheres (20.2 ± 4.3 vs. 17.4 ± 5.8%, n=2; Fig. 2C). Interestingly, SI-derived neurospheres gave rise to significantly fewer SMA+ smooth muscle cells (15.6 ± 2.0 vs. 30.2 ± 1.5%, p<0.05, n=2; Fig. 2D). The overall gross morphology of cells did not differ between SI and LI sources (Fig. 2E-J). The same experiments were repeated using SI- and LI-derived neurospheres generated from human intestine obtained from 5 patients, including 3 SI and 3 LI samples (Table 1). Human SI-derived neurospheres generated 28.5 ± 18.0% Tuj1+ neurons with 6.4 ± 3.6% proliferating (Table S1; median age 17, range 16 to 21 years), while neurospheres from human LI generated 15.2 ± 2.6% Tuj1+ neurons with 12.5 ± 14.2% proliferating (Table S1; median age 6, range 1 month to 17 years). SI-derived neurospheres thus generated 88% more neurons than LI, although this difference did not reach statistical significance, possibly due to the large variability in results among human patients. From a single patient who underwent ileocecal resection, a direct comparison of SI and LI was performed. Ileum-derived neurospheres contained significantly more P75+ neural crest cells than cecum-derived neurospheres (32.6 ± 17.5 vs. 10.2 ± 8.2%, p<0.05; Fig. 3A) and gave rise to significantly more Tuj1+ neurons in culture (29.7 ± 16.4 vs. 16.0 ± 13.5%, p<0.05; Fig. 3B). As with mouse enteric neurospheres, neuronal proliferation from ileum and cecum-derived neurospheres did not differ significantly (8.1 ± 8.7% vs. 7.2 ± 9.7%; Fig. 3C), nor did neuronal migration (0.64 ± 0.11 vs. 0.77 ± 0.19 mm; Fig. 3D). Neurogenic potential is enhanced by GDNF The effect of factors known to promote ENSC proliferation during embryonic ENS development was tested by adding GDNF, ET3, or both to cultured mouse neurospheres. The proportion of P75+ neural crest-derived cells did not differ significantly between control, ET3, GDNF, or ET3+GDNF treated cultures for SI-derived neurospheres (56.3 ± 6.4 vs. 56.9 ± 6.3 vs. 55.5 ± 4.4 vs. 53.9 ± 5.7%, respectively, n=5; Fig. 4A) or LI-derived neurospheres (15.9 ± 8.4 vs. 17.6 ± 6.4 vs. 14.4 ± 5.7 vs. 19.7 ± 5.2%, respectively, n=5; Fig. 4C). However, SI-derived neurospheres cultured with GDNF gave rise to significantly more Tuj1+ neurons (75.9 ± 4.0%) compared to control neurospheres cultured with no additives (67.8 ± 5.8%, p<0.05, n=5; Fig. 4B), representing an increase of 12% in neuronal density. No significant difference in neuronal density was observed in the presence of ET3 (65.4 ± 7.9%) or ET3+GDNF (74.3 ± 5.7%). In contrast, LI-derived neurospheres cultured with either no additives, ET3, GDNF, or ET3+GDNF all gave rise to equivalent numbers of Tuj1+ neurons (36.4 ± 4.0 vs. 33.2 ± 4.0 vs. 31.6 ± 8.0 vs. 33.5 ± 6.9%, respectively, n=4; Fig. 4D). The rate of neuronal proliferation was not significantly different among the various treatment groups for SI-derived (4.1 ± 3.1 vs. 3.7 ± 2.5 vs. 3.6 ± 2.0 vs. 6.8 ± 4.5%, respectively, n=5; data not shown) or LI-derived neurospheres (8.8 ± 7.6 vs. 7.2 ± 7.2 vs. 4.8 ± 4.1 vs. 8.3 ± 7.4%, respectively, n=4; data not shown). In order to understand the differential effect of GDNF on SI and LI, we examined Gdnf and Et3 transcript levels in SI- and LI-derived neurospheres. The relative expression of Et3 transcript did not differ significantly between SI and LI (1.25 × 10−6 ± 4.15 × 10−7 vs. 4.88 × 10−6 ±1.46 × 10−6; Fig. 4E). Interestingly, however, the relative expression of Gdnf was significantly lower in SI neurospheres than in LI neurospheres (1.48 × 10−4 ± 1.41 × 10−5 vs. × 1.39 × 10−3 ± 2.26 × 10−4, p<0.05; Fig. 4F), representing a 10-fold difference in expression. The ratio of the relative expression of Gdnf compared to Et3 was also significantly lower in SI neurospheres than LI neurospheres (119 ± 11 vs. 285 ± 46; p<0.05). The relative expression of Ednrb did not differ significantly between SI and LI (7.98 × 10−4 ± 3.52 × 10−4 vs. 2.40 × 10−3 ± 2.65 × 10−5, data not shown), nor did the expression of Ret (2.88 × 10−5 ± 4.29 × 10−6 vs. 4.70 × 10−5 ± 3.52 × 10−6, data not shown). DISCUSSION Enteric neurospheres represent a potential treatment option for neurointestinal diseases by serving as an enriched source of ENSCs to replace missing or abnormal enteric neurons in patients with enteric neuropathies. Identifying the optimal source of these cells and developing efficient methodologies for their isolation and cultivation are important goals in order to achieve clinical application of this promising technology. Our results demonstrate that SI-derived neurospheres from mice and humans contain more than double the density of neural crest-derived progenitor cells as LI-derived neurospheres and subsequently give rise to significantly more neurons in culture. Furthermore, the neurogenic potential of those SI-derived neurospheres is substantially enhanced by the addition of the neurotrophic factor, GDNF. These findings will help to refine current strategies for generating ENSCs for use as cell therapy. SI represents a readily available source for ENSC harvest in the clinical setting. The SI is easily accessible for biopsy endoscopically or laparoscopically. When more tissue is needed, a large segment of SI can be surgically resected without adverse effects (14). Furthermore, SI is usually normally ganglionated in Hirschsprung disease and other enteric neuropathies, allowing harvest of autologous cells. Our findings suggest that SI may also possess other advantages over LI. The greater progenitor purity and neurogenic potential of SI-derived neurospheres compared to LI-derived neurospheres is an important observation. This may be due to the thinner muscular layer in the SI, as SI neurospheres produced significantly more neurons and less smooth muscle cells than their LI counterparts. The differences between SI and LI-derived neurospheres do not appear to be due to any innate differences in their neural crest-derived cells, as neuronal proliferation did not differ between SI and LI in our study. Similarly, the rate of neuronal proliferation has been shown to remain constant along different regions of the gut in embryonic mice (15). However, SI may inherently contain a greater density of ENSCs than LI, although this remains unclear. Though it accounts for a relatively small portion of the distal colon, the sacral neural crest contributes to the myenteric plexus of LI (16) and is limited in its ability to form enteric ganglia compared to the vagal neural crest (17). This difference in SI and LI innervation may explain some of the differences in enteric neurospheres formed from the two regions. Further study is needed to determine if the distal gut truly contains fewer ENSCs and whether this underlies the differences in SI- and LI-derived neurospheres. The balance of ET3 and GDNF plays a pivotal role in ENS development (10) and may continue to affect postnatal ENSCs in culture. The addition of GDNF to culture media had a modest but significant effect on the density of neurons generated from SI neurospheres. In contrast, GDNF had no observable neurogenic effect on LI neurospheres. Interestingly, both the relative expression of Gdnf and the ratio of Gdnf to Et3 expression were significantly lower in SI than LI neurospheres, which may account for the response to supplemental GDNF in the former. The relative expression of Et3 was not significantly different between the two, and supplemental ET3 had no observable neurogenic effect on SI- or LI-derived neurospheres. Since both GDNF and ET3 are expressed by the gut mesenchyme during embryonic development, an increased proportion of ENSCs and decreased proportion of bystander cells, such as mesenchymal (e.g. smooth muscle) cells, may diminish the relative expression of GDNF and ET3 in the neurosphere. As the population of neural progenitors is enriched in enteric neurospheres, our results suggest that it may be necessary to supplement factors normally produced by non-neural cells in order to maximize neurogenesis. While GDNF and ET3 have been shown to exert a synergistic effect on the growth of embryonic ENSCs (18), neither GDNF nor ET3 had an effect on postnatal ENSC density or neuronal proliferation. The lack of effect may be due to a number of factors. As discussed above, the effect of supplemental factors may not be seen until constitutive expression drops below a threshold level. In addition, changes in culture media and cell density are known to alter the effects of ET3 and GDNF (19), and the specific conditions used in this study may obscure observable changes. Finally, the effects of ET3 and GDNF may be time-sensitive. In normal development, ET3 stimulates ENSC proliferation only in early embryonic stages (20, 21), and GDNF stimulates ENSC proliferation early in ENS development and promotes neuronal differentiation at later embryonic stages (22). The effect of ET3 and GDNF on postnatal cells may be diminished as the pathways that promote postnatal enteric neurogenesis may differ from those involved in the embryonic mechanisms (23, 24). It is likely that many of the factors which promote postnatal ENSC growth remain undiscovered. More insight into the proliferation and differentiation of postnatal neurons will help develop techniques to maximize their growth in vitro. In summary, enteric neurospheres contain ENSCs with the potential to repopulate the aganglionic gut of patients with Hirschsprung disease. Compared to LI-derived neurospheres, SI-derived neurospheres contain a greater proportion of ENSCs and subsequently give rise to more neurons. The neurogenic potential of SI-derived neurospheres can be further augmented by the addition of GDNF. Optimization of enteric neurosphere culture improves the neurogenic, and subsequently, therapeutic potential of enteric neurospheres. Supplementary Material ACKNOWLEDGEMENTS We would like to thank Dr. Vanda Lennon (Mayo Clinic, Rochester, MN) for the kind gift of Hu antibody. FIGURE 1 Neurospheres can be generated from the small and large intestine of mice Enteric ganglia containing Tuj1+ and Hu+ neurons are found in both small intestine (SI, A) and large intestine (LI, D). Tuj1+ neurons in the myenteric plexus were isolated together with the surrounding SMA+ muscularis propria from SI (B) and LI (E) and neurospheres were formed (C, F; respectively). While neurospheres from both sources were similar in size (G), SI-derived neurospheres contained significantly more P75+ neural crest-derived cells (H). Migratory distance of neurons from SI- and LI-derived neurospheres was similar (I). Scale bar in A is 100 μm and applies to A-B and D-E. Scale bar in C is 100 μm and applies to C and F. p<0.05; SMA, smooth muscle actin; SI, small intestine; LI, large intestine FIGURE 2 Mouse SI- and LI-derived neurospheres differ in neurogenic potential SI-derived neurospheres give rise to a significantly greater percentage ofTuj1+ neurons than LI-derived neurospheres (A), but neuronal proliferation (B) and glial cell density (C) are not different. LI neurospheres give rise to a significantly greater percentage of smooth muscle cells (D). The morphology of Tuj1+ neurons, S100+ glia, and SMA+ smooth muscle cells from SI (E-G) are not grossly different from LI (H-J). Scale bar in E is 100 μm and applies to E-J. p<0.05; SMA, smooth muscle actin; SI, small intestine; LI, large intestine FIGURE 3 Human SI and LI-derived neurospheres differ in their neurogenic potential Neurospheres generated from human ileum contain significantly more P75+ neural crest-derived cells (A, E) and give rise to significantly more Tuj1+ neurons (B) than neurospheres from the cecum of the same patient. However, neuronal proliferation (C) and migration (D) are not statistically different. Proliferating neurons from SI (F) and LI (G) are immunoreactive to Tuj1 (green arrows) and incorporate EdU (red arrows). Scale bar in E is 100 μm. Scale bar in F is 50 μm and applies to F and G. p<0.05; SI, small intestine; LI, large intestine FIGURE 4 GDNF enhances neurogenic potential of SI-derived neurospheres Mouse SI- and LI-derived neurospheres were cultured with endothelin-3 (ET3, 100 ng/mL), glial cell-derived neurotrophic factor (GDNF, 50 ng/mL), both factors, or no additives (Control). Neither ET3 nor GDNF has an effect on the density of P75+ neural crest cells in SI-derived neurospheres (A). Addition of GDNF alone significantly increases the proportion of neurons generated from SI-derived neurospheres compared to control and ET3 (B). In LI-derived neurospheres, neither ET3 nor GDNF affect the density of P75+ neural crest cells or Tuj1+ neurons (C-D). The level of Et3 expression does not differ significantly between SI- and LI-derived neurospheres (E). However, the level of Gdnf expression in SI-derived neurospheres is significantly diminished relative to LI-derived neurospheres (F). p<0.05; SI, small intestine; LI, large intestine; ET3, endothelin-3; GDNF, glial cell-derived neurotrophic factor TABLE 1 Human tissue sources. Age Sex Operation Indication Tissue 1 month old M Colon resection necrotizing enterocolitis with colonic stricture Colon 6 years old F Colostomy closure cloaca Colon 16 years old M Ileostomy revision Crohn's disease Ileum 17 years old M Ileocecal resection Crohn's disease with ileocecal stricture Ileum, Colon 21 years old M Ileostomy closure ulcerative colitis Ileum This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. AUTHOR DISCLOSURE STATEMENTS The authors have no conflicting interests. AUTHOR CONTRIBUTIONS L.S.C., R.H., and A.M.G. designed the experiments. L.S.C. performed the experiments and wrote the manuscript. H.K.G. provided reagents and maintained mouse colonies. W.P. performed quantitative PCR. A.M.G. provided human tissue. All authors critically reviewed and approved of the final manuscript. DISCLOSURE L.S.C. is supported by the Society of University Surgeons Ethicon Surgical Research Fellowship Award. R.H. is supported by grants from the Tosteson Fund for Medical Discovery at Massachusetts General Hospital, the REACHirschsprung Foundation, and the American Neurogastroenterology and Motility Society. A.M.G. is supported by the National Institutes of Health (R01DK103785). REFERENCES 1 Furness JB The enteric nervous system and neurogastroenterology. Nat Rev Gastroenterol Hepatol 9 286 294 2 Burns AJ Thapar N Neural stem cell therapies for enteric nervous system disorders. Nat Rev Gastroenterol Hepatol 11 317 328 3 Burns AJ Goldstein AM Newgreen DF Stamp L Schafer KH Metzger M Hotta R Young HM Andrews PW Thapar N Belkind-Gerson J Bondurand N Bornstein JC Chan WY Cheah K Gershon MD Heuckeroth RO Hofstra RM Just L Kapur RP King SK McCann CJ Nagy N Ngan E Obermayr F Pachnis V Pasricha PJ Sham MH Tam P Berghe PV White paper on guidelines concerning enteric nervous system stem cell therapy for enteric neuropathies. Dev Biol 4 Metzger M Bareiss PM Danker T Wagner S Hennenlotter J Guenther E Obermayr F Stenzl A Koenigsrainer A Skutella T Just L Expansion and differentiation of neural progenitors derived from the human adult enteric nervous system. Gastroenterology 2009 137 2063 2073 e2064 19549531 5 Hetz S Acikgoez A Voss U Nieber K Holland H Hegewald C Till H Metzger R Metzger M In vivo transplantation of neurosphere-like bodies derived from the human postnatal and adult enteric nervous system: a pilot study. PLoS One 9 e93605 24699866 6 Binder E Natarajan D Cooper J Kronfli R Cananzi M Delalande JM McCann C Burns AJ Thapar N Enteric neurospheres are not specific to neural crest cultures: implications for neural stem cell therapies. PLoS One 10 e0119467 7 Hotta R Cheng LS Graham HK Nagy N Belkind-Gerson J Mattheolabakis G Amiji MM Goldstein AM Delivery of enteric neural progenitors with 5-HT4 agonist-loaded nanoparticles and thermosensitive hydrogel enhances cell proliferation and differentiation following transplantation in vivo. Biomaterials 88 1 11 26922325 8 Grundmann D Klotz M Rabe H Glanemann M Schafer KH Isolation of high-purity myenteric plexus from adult human and mouse gastrointestinal tract. Sci Rep 5 9226 25791532 9 Gianino S Grider JR Cresswell J Enomoto H Heuckeroth RO GDNF availability determines enteric neuron number by controlling precursor proliferation. Development 2003 130 2187 2198 12668632 10 Goldstein AM Hofstra RM Burns AJ Building a brain in the gut: development of the enteric nervous system. Clin Genet 83 307 316 23167617 11 Robertson K Mason I Hall S Hirschsprung's disease: genetic mutations in mice and men. Gut 1997 41 436 441 9391239 12 Hotta R Cheng LS Graham HK Pan W Nagy N Belkind-Gerson J Goldstein AM Isogenic enteric neural progenitor cells can replace missing neurons and glia in mice with Hirschsprung disease. Neurogastroenterol Motil 13 Almond S Lindley RM Kenny SE Connell MG Edgar DH Characterisation and transplantation of enteric nervous system progenitor cells. Gut 2007 56 489 496 16973717 14 Ji G Chu D Wang W Dong G The safety of donor in living donor small bowel transplantation--an analysis of four cases. Clin Transplant 2009 23 761 764 19689452 15 Young HM Turner KN Bergner AJ The location and phenotype of proliferating neural-crest-derived cells in the developing mouse gut. Cell Tissue Res 2005 320 1 9 15714282 16 Burns AJ Douarin NM The sacral neural crest contributes neurons and glia to the post-umbilical gut: spatiotemporal analysis of the development of the enteric nervous system. Development 1998 125 4335 4347 9753687 17 Burns AJ Champeval D Le Douarin NM Sacral neural crest cells colonise aganglionic hindgut in vivo but fail to compensate for lack of enteric ganglia. Dev Biol 2000 219 30 43 10677253 18 Barlow A de Graaff E Pachnis V Enteric nervous system progenitors are coordinately controlled by the G protein-coupled receptor EDNRB and the receptor tyrosine kinase RET. Neuron 2003 40 905 916 14659090 19 Hearn CJ Murphy M Newgreen D GDNF and ET-3 differentially modulate the numbers of avian enteric neural crest cells and enteric neurons in vitro. Dev Biol 1998 197 93 105 9578621 20 Woodward MN Kenny SE Vaillant C Lloyd DA Edgar DH Time-dependent effects of endothelin-3 on enteric nervous system development in an organ culture model of Hirschsprung's disease. J Pediatr Surg 2000 35 25 29 10646768 21 Shin MK Levorse JM Ingram RS Tilghman SM The temporal requirement for endothelin receptor-B signalling during neural crest development. Nature 1999 402 496 501 10591209 22 Chalazonitis A Rothman TP Chen J Gershon MD Age-dependent differences in the effects of GDNF and NT-3 on the development of neurons and glia from neural crest-derived precursors immunoselected from the fetal rat gut: expression of GFRalpha-1 in vitro and in vivo. Dev Biol 1998 204 385 406 9882478 23 Becker L Peterson J Kulkarni S Pasricha PJ Ex vivo neurogenesis within enteric ganglia occurs in a PTEN dependent manner. PLoS One 8 e59452 23527198 24 Uesaka T Nagashimada M Enomoto H Neuronal Differentiation in Schwann Cell Lineage Underlies Postnatal Neurogenesis in the Enteric Nervous System. J Neurosci 35 9879 9888 26156989
PMC005xxxxxx/PMC5125513.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101657097 43784 Crit Radic Soc Work Crit Radic Soc Work Critical and radical social work 2049-8608 2049-8675 27909583 5125513 10.1332/204986016X1473688814636 EMS70282 Article Re-Coopering anti-psychiatry: David Cooper, revolutionary critic of psychiatry Chapman Adrian Florida State University, London Centre, UK [email protected] 14 11 2016 11 2016 01 5 2017 4 3 421432 This file is available to download for the purposes of text mining, consistent with the principles of UK copyright law. This article offers an introduction to David Cooper (1931–86), who coined the term ‘anti-psychiatry’, and, it is argued here, has not so far received the scholarly attention that he deserves. The first section presents his life in context. The second section presents his work in detail. There follows a section on the critical reception of Cooper, and, finally, a conclusion that sets out ways in which he might be interesting and useful today. psychiatry anti-psychiatry mental health David Cooper R.D. Laing Life in context David Cooper was born in 1931 in Cape Town, South Africa. He graduated in medicine from the University of Cape Town in 1955. The son of a chemist, he never wanted to become a doctor and was probably responding to family pressure in becoming one. In South Africa, he was the lover of Ros de Lanerolle, who would later be a founder of the Anti-Apartheid Movement in Britain and would set up the Women’s Press publishing house in London. In his home country, Cooper was involved in underground resistance to the Apartheid regime. According to the radical Scottish psychiatrist R.D. Laing (whom Cooper met when he was working at the Seamen’s Hospital in South London, having come to England to train in psychiatry after university in Cape Town), Cooper had been a member of the then-outlawed Communist Party of South Africa (Mullan, 1995: 194), but he may not actually have held a party card. He did, however, spend time in China for political education (probably in the mid-1950s). Later, Cooper would travel to Cuba in the early days of the revolution. Debts to Marxism are evident throughout his work. For instance, in his first book, Psychiatry and anti-psychiatry (Cooper, 1967), he argues for a dialectical methodology in studying families, and in his last book, The language of madness (Cooper, 1980), he draws on the work of the Marxist Hungarian philosopher Ágnes Heller to think through the nature of radical needs. In his debt to Marx, we can draw a contrast with R.D. Laing, for whom Marxism was never a significant influence. In his more radical outlook, Cooper had affinities with the Psichiatria Democratica (Democratic Psychiatry) movement in Italy, of which Franco Basaglia was the key voice. Basaglia’s English biographer, John Foot (2014: 238), argues that the Italian’s political formulations could be rather crude, with him making, for instance, rather simplistic contrasts between the poor and the rich in thinking about mental health. Cooper’s linking of mental health with wider contexts could also lack nuance at times, and tend towards sloganeering (eg ‘The bourgeois state is a tranquillizer pill with lethal side-effects’ [Cooper, 1974: 35]). Nevertheless, Cooper always insisted on the need to understand the symptoms of the individual not simply in terms of individual psychology or family dynamics, but also in terms of broader social, institutional and economic forces, and, for him, Marx was a key starting point in doing so. His Marxism was far from that of the orthodox (‘Tankie’) or Eurocommunist member of the British Communist Party. Nor was he aligned with far left groups such as the International Socialists (later the Socialist Workers’ Party) or the International Marxist Group. He was very much a New Left countercultural revolutionary. This meant that the terrain of everyday life was as valid a region of political struggle as the factory or office, and sex and drugs went together with revolution. He was fond of the Belgian situationist Raoul Vaneigem’s line (quoted in Gale, 2001), popular during the May 1968 events in Paris: ‘He who speaks of revolution without living it in their daily life speaks with a corpse in his mouth.’ At one time, Cooper placed these words on a placard above the fireplace in his consulting room (Gale, 2001). Love, revolution, the need to overcome limits: Cooper’s politics were explicitly countercultural and he historically belongs to the 1960s (the 1960s, that is, as an historical period running at least up to the mid-1970s). His focus on love and on what goes on between people in households (especially in The death of the family [Cooper, 1974] and The grammar of living [Cooper, 1976]) involved him in both an appropriation of and a critique of psychoanalysis. However, unlike Laing, Cooper never trained in psychoanalysis. He trained in psychiatry in England in the 1950s and held several hospital appointments before leaving the National Health Service (in 1966) after his time at Shenley Hospital in Hertfordshire. He underwent two periods of psychoanalysis (with different therapists), both brief, and one of which was with the South African Kleinian analyst Leslie Sohn. In The death of the family (Cooper, 1974: 134), Cooper writes scathingly of one analyst (probably Sohn), claiming that he was jealous of his patient’s freedom in moving from commune to commune. Cooper was in analysis with Sohn while he (Cooper) was working at Shenley Hospital as Senior Registrar. There, he set up an experiment in ward democracy in a wing for male adolescents and young men. The Villa 21 experiment, however, which began in 1962, moved far beyond the then-popular idea of the therapeutic community (associated with Maxwell Jones) and proved unsettling for the hospital authorities, who closed the ward in 1966. Cooper soon came to reject both psychiatry and psychotherapy. In 1965, though, he set up the mental health charity the Philadelphia Association (PA) with R.D. Laing and others. The PA established a therapeutic community (outside state mental health provision) at Kingsley Hall, East London (1965–70), but Cooper showed little interest in the project. He left the organisation in 1971 in protest at the PA’s beginning of a formal training programme in psychotherapy. From his point of view, this smacked of failure to recognise the wider political nature of distress. As a member of the London-based Institute of Phenomenology, Cooper was instrumental in setting up the Dialectics of Liberation conference at the Roundhouse in North London. This event focused on the nature of violence and the possibility of liberation, and included presentations from Herbert Marcuse, Stokely Carmichael, Allen Ginsberg, Gregory Bateson and others, including Laing and Cooper. In his introduction to the book of the event, Cooper (1968a: 7) presents the therapists involved in the conference as ‘anti-psychiatrists’. Despite Laing’s objections, and despite very few other therapists ever identifying themselves as anti-psychiatrists, the term caught on in the media and among scholars of mental health. The conference is notable today not only as a high point of 1960s’ radicalism in the UK, but also as an example of the male focus of 1960s’ counterculture. One of the very few women to take part, the artist Carolee Schneerman, has spoken to me of the opposition to her involvement that she encountered from male participants (R.D. Laing and Paul Goodman in particular). Cooper, however, she remembers as supportive, someone with “no machismo about him”. Jean Paul Sartre was to have been the key speaker at the Dialectics conference but could not attend. Radical psychiatry in the long 1960s often drew on existentialism and phenomenology, and Cooper, like Laing, as well as Basaglia in Italy and Frantz Fanon in Algeria, was very influenced by Sartre. With Laing, Cooper (1964) wrote an introduction to Sartre’s later work, Reason and violence, and offered English-speaking readers ways into reading Critique of dialectical reason (1976), the first volume of which would not be translated until 1976. Cooper also wrote an introduction to Madness and civilisation (Foucault, 1967), the first English version of Michel Foucault’s Folie et déraison (1961). Madness, Cooper claimed, represents ‘a kind of lost truth’ (Foucault, 1967: vii). Through his work introducing English readers to the work of Foucault and Sartre, Cooper played an important part in mediating mid-20th-century French philosophy. Education was important to Cooper, who was involved in the Anti-university of London (1968–71), a radical educational initiative that, prior to the establishment of the low-cost, open access Open University and the expansion of adult education, offered very cheap courses to all-comers. Along with Cooper, teachers included Laing and the psychotherapist Joseph Berke, as well as Stuart Hall, the author Alexander Trocchi, the anthropologist Francis Huxley, the artist Jeff Nutall and the feminist Juliet Mitchell. More than just a low-cost form of education, however, the Anti-university aimed to question the division between teacher and student, and to orientate study explicitly to personal and political development. The putting into question of ‘the rules of the game’ was (as Cooper says on a news clip about the Anti-University) key to the meaning of the ‘anti-’ in education and other contexts – including, of course, psychiatry (Dorley-Brown, 2010). He was active not only in education, but also in what might be termed an early form of mental health users’ activism. Cooper met Basaglia at a conference in Portugal in 1974 and was impressed by his Psichiatria Democratica movement. The Portugal conference foreshadowed the setting up of the International Network of Alternatives to Psychiatry (INAP) in 1975, with which Cooper was involved. Following Cooper’s suggestion in a 1982 Brussels meeting of the INAP, Stephen Ticktin, who had acted as Cooper’s personal assistant and had shared a flat with him in Crouch End, North London, established the British Network for Alternatives to Psychiatry. Cooper was married to a French-Vietnamese psychiatric nurse, and the two had three children. He left his family, however, and by 1967, was in a relationship with Juliet Mitchell, who had been a patient of his and who would go on to write the influential Psychoanalysis and feminism (1974). The two lived for a while in the South of France. Cooper remarked that his The death of the family (Cooper, 1974) marked his departure from his own family (Ticktin, 1986: 16). In the book, he alludes to a period of madness (of de- and restructuring) in his own life. At this time, Cooper was practising private individual psychotherapy, and a former patient has written of his therapist’s evident distress (Gale, 2001). Cooper left England for Argentina in 1972 in order to promote opposition to psychiatry in the developing world. There he wrote The grammar of living (Cooper, 1976), but had little success in developing alternatives to psychiatry. In his later The language of madness (Cooper, 1980: 29), written in Paris, where he moved in 1975, he writes of undergoing a period of madness in Argentina. Although he lived in South America and grew up in South Africa, Cooper was very much a European intellectual and was at home in Paris. There he taught at the University of Vincennes (Paris VIII) and wrote the brief, untranslated, Qui sont les dissidents (Who are the dissidents) (Cooper, 1977), an essay composed with the assistance of his lover Marine Zecca. The text focuses on the nature of dissidence and how Western progressives (themselves living in a giant prison, ‘le Mega-Goulag de l’Ouest’ [‘the mega-gulag of the West’] [Cooper, 1977: 33]) ought to avoid criticism of the USSR that left the injustices of their own societies unaddressed. Paris was to be Cooper’s last city of residence. He died there in 1986. Like Laing, who also died young (in 1989), Cooper was a long-time heavy drinker – a significant factor in his early death. While he could be riotous when drinking, when he was sober, he was often perceived as the perfect gentleman (although he hated anyone referring to him as such). Stephen Ticktin used to tease Cooper by calling him ‘Dr Cooper and Mr Hyde’. Cooper was researching a book to be called The geometry of freedom (to be written with Marine Zecca) prior to his death. This was to be a project that moved beyond mental health to consider health needs more broadly in France, Italy and North Africa. I know of no manuscript copy of this last book. Nor do I know of any Cooper archival resources. Key concepts and interventions Probably the best reason for remembering Cooper is Villa 21 (1962–66), a residential unit consisting of young working-class men at Shenley Hospital in Hertfordshire, UK. There compulsory treatment was pared down to a daily meeting that included patients, doctors, nurses, social workers and volunteers interested in the project. Cooper’s emphasis was on fostering an atmosphere in which people – not reduced to their roles as ‘patient’, ‘doctor’, ‘social worker’ and so on – supported one another, and where the prescription of drugs was kept to the minimum. Oisin Wall (2013) provides a good basic history of Villa 21 that includes an interview with a former resident. One volunteer was the US author and left-wing activist Clancy Sigal, writer of Zone of the interior (Sigal, 2005), a satirical novel about anti-psychiatry based on his own experiences at both Villa 21 and his association with Laing at the much more famous therapeutic community at Kingsley Hall. For Sigal, Villa 21 was a place for those who, unlike the residents of Kingsley Hall, had no class privilege to cushion their voyages into madness and provide a route back into a basically secure world. In an article after Cooper’s death, Sigal (1986) argues that Cooper’s experiment in dissolving doctor–patient boundaries might not have succeeded, but was at the very least a noble failure. The hospital authorities, disturbed by the physical messiness of Villa 21 (not to say the challenge it presented to their view of mental health care), closed down the experiment after four years. Cooper came to believe that the future of radical psychiatry was beyond the hospital. We should remember, however, that the first experimental anti-psychiatric community was carried out not at Kingsley Hall, but in a state mental hospital, and it was Cooper who persuaded the hospital authorities to allow the project to go ahead. He recalls the experiment in his only book still in print, Psychiatry and anti-psychiatry (Cooper, 1967), which gives an account of his journey from being a psychiatrist to becoming an anti-psychiatrist. In this text, we find him, despite withering attacks on the hospitalisation of the mad and his questioning of the distinction between sanity and madness, following conventions of social-scientific writing and writing as a medically legitimated critic of psychiatry. There is, for instance, a chapter on methodology; another offers an extended case history. The final chapter gives us, in its entirety, an article (previously published in the British medical journal The Lancet) that presents evidence of Villa 21’s efficaciousness. In his next two books, The death of the family (Cooper, 1974) and The grammar of living (Cooper, 1976), Cooper moves away from the role of therapeutic professional. What people need most, he tells us, is a witness, someone who recognises the uniqueness of one’s experience and respects one’s individual autonomy. The witness need not be a professional; it could be – and is far more likely to be – someone with whom one lives in a commune, an anti-family in which the rigid roles of ‘mother’ and ‘father’ no longer exist (although mothering and fathering still do). This is a place in which children can bring up grown-ups, as well as vice versa, and in which people are not fixed in the contractual obligation of marriage or in one of the binaries of passive–active or heterosexual–homosexual. Therapy is understood as a matter of someone ridding him- or herself of haunting family ‘ghosts’, and this is best carried out in the commune. Some people, by dint of their hard-won, relative freedom from their internal families – the result of considerable work on themselves – will be in a better position than others to act as witnesses and to provide supportive guidance through experiences of de- and restructuring (see Cooper, 1976: 54–65; 1976b: 31–45). Therapy, for Cooper, becomes a matter not of specialised treatment, but rather of the practice of close attention to how people treat one another in everyday life. Consciousness-expanding drugs have a place, too, in opening up new therapeutically valuable experiences, as does meditation (Cooper, 1976: 30–38, 127–31; 1976b: 130). In The death of the family (Cooper, 1974), Cooper’s writing becomes more ludic, and he writes very much as a member of the counterculture. We can find the aphoristic ‘Guns have their place, of course, but the bed is perhaps the great unused secret weapon of the revolution’ (Cooper, 1974: 118), alongside diagrams representing stages of life and experience (Cooper, 1974: 12, 40, 121), Cooper’s own verse (Cooper, 1974: 125–6, 141–2), a cabalistic story (Cooper, 1974: 19–20), a story from Tibetan Buddhism (Cooper, 1974: 21–2), and a whole chapter of free-associative punning (Cooper, 1974: 84–90). As Cooper (1980: 19) puts it in his last book, The language of madness, ‘One would erect a mockery if one were to attempt to write systematically about a discourse that dismantles systematic thought’. There is a great energy about his writing, which is clearly different to that of R.D. Laing, with whom he is so often associated. Perhaps a good comparison would be between the jazz musicians John Coltrane and Miles Davis: there is something restless about Coltrane, who does not always know when to stop playing; Davis, by contrast, is more in control, more knowing – more like Laing to Cooper’s Coltrane. In Zone of the interior, Sigal (2005) satirises the Laing figure, Willie Last, who understands the madman as the new proletarian, the harbinger of revolution. This view, however, is much closer to Cooper’s than to Laing’s. In The language of madness, Cooper (1980: 23, emphasis in original) states: ‘all delusion is political statement … and all madmen are political dissidents’. Most notably, for Cooper, madness is both a resistance to and a sign of the repressive nature of the family. In his exploration of communal anti-families in The death of the family (Cooper, 1974) and The grammar of living (Cooper, 1976), we find him seeking to move beyond the bourgeois family. For him, the family is the origin of stunted experience and boxed-in social roles: ‘the ultimately perfected form of non-meeting’ (p 6). Growth is stunted among family members to the extent that, existentially, people tend to be more absent than present, and, therefore, the non-meeting of false selves characterises relationships (see Cooper, 1976: 54–65; 1976b: 31–45). In The language of madness (Cooper, 1980: 22, emphasis in original), madness is presented as ‘a movement out of familialism including family-modelled institutions) towards autonomy’. Madness presents a challenge to the family, which, for Cooper, stands at the basis of capitalism – and capitalism can only cope with psychotic dissent by invalidating the mad. The death of the family (Cooper, 1974) abounds with revolutionary optimism: it is possible to transcend the limitations of the bourgeois family; it is possible to make a revolution – and a new, revolutionary kind of self. In Cuba, he tells us, ‘they hope to abolish money in ten years. Everyone will be able to walk into shops and help themselves to whatever they need without paying’ (p 105). He writes of a visit in which ‘he found no evidence that people there would tolerate for long the imposition of sclerosed forms of non-life on them’ (p 148). What gives Cooper hope is the change in subjectivity brought about by the ‘The Guevarist doctrine of the New Man in Cuba’ (Cooper, 1974). To readers on the Left, of course, such optimism is likely to read rather plaintively today. It is likely, too, that readers today will recognise Cooper’s romanticisation of madness. This is the case despite his keen sympathy for those who suffer, and despite his writing about his own episodes of madness. In The language of madness (Cooper, 1980: 29), for instance, he writes of swimming naked ‘in the heart of a tempest that transformed miraculously the sand dunes into amiable and terrifying other humps [and] dinosauric monsters’. Cooper, at times, clearly associates madness with the purity of childhood – he can be very romantic and can romanticise, and he associates madness with truth-telling and full, transformatory experiences. (What about the madness of the lost, we might ask – those lost for weeks, months and years?) For Cooper, the environment in which an experience of madness takes place is crucial: ‘if one has to go mad, the tactic to learn in our society is one of discretion’, he says in Psychiatry and anti-psychiatry (Cooper, 1967: 33). The death of the family (Cooper, 1974) and The grammar of living (Cooper, 1976) are very much about constructing environments in which madness might be positively transformative. In The language of madness (Cooper, 1980), madness is presented as a resource, something almost everyone has that might be drawn upon for the purposes of individual and social transformation. Moreover, madness is distinguished from schizophrenia, a sort of failed madness: The madness about which I’m writing is the madness that is more or less present in each one of us and not only the madness that gets the psychiatric baptism by diagnosis of ‘schizophrenia’ or some other label invented by the specialised psycho-police agents of the final phase capitalist society. So when I use the word ‘madman’ here I’m not referring to a special race of people, but the madman in me is addressing the madman in you. (Cooper, 1980: 18) In The language of madness, we can discern the influence of Anti-Oedipus (1977) by Giles Deleuze and Félix Guattari, which, Cooper (1980: 138) says, is ‘a magnificent vision of madness as a revolutionary force, the decoding, deterritorializing refusal of fixity and outside definition by schizophrenia’. The influence of Guatari and Deleuze is present, too, in Cooper’s jibes at what he presents as the depoliticised therapeutic communities in England. He writes mockingly of ‘all the “inner voyages” going on’, and says that ‘one cannot fracture a macro-political reality of oppression and repression with introspective micro-groups of privileged children of the bourgeoisie’ (Cooper, 1980: 133–4). ‘What sense does it make’, asked Cooper at a 1981 conference in Leuven, Belgium, ‘to create ten happy islands in a world where everything keeps functioning just like before? In this way the institution is not being attacked. Madness is being recuperated, encapsulated by the system and loses its function to subversive activity’ (quoted in Laing, 1982 :3). Madness, then, needs to be integrated into every aspect of society and politicised. Inspired especially by the work of Franco Basaglia, in The language of madness, Cooper (1980: 116–52) styles himself as an advocate not of anti-psychiatry – a confusing term, he now argues, that has been co-opted by those really in support of psychiatry – but of non-psychiatry, a practice of politicised community activism, support and political education for those stigmatised by disabling labels. Cooper is still a revolutionary in his last book, but his stance also represents mounting pessimism in the counterculture and New Left. If revolution does not transpire by the end of the century, he remarks bitterly, then humanity should be ‘extinguished’ because ‘it will no longer be the human species’ (Cooper, 1980: 148) – presumably because it will have become so psychologically distorted. Critical reception It was never enough just to ‘talk the talk’ for Cooper, who, in an allusion to Gandhi’s dictum ‘Be the change you want to see in the world’, says in ‘Beyond words’, his contribution to the The dialectics of liberation, that ‘one must be the dialectic one wants to be’ (Cooper, 1968b: 193). It is easy, perhaps, to dismiss Cooper as hippy-dippy, as representing the excesses and lamentable enthusiasms of the New Left and counterculture. This, I think, is a significant reason why Cooper has been overlooked in the academy. R.D. Laing perhaps set the tone here with his remark that he found Cooper’s books ‘embarrassing’ (Mullan, 1995: 195). In considering the critical reception of Cooper, it must first be noted that he has received very little scholarly attention. He is mentioned in surveys or critiques of anti-psychiatry, but usually only in passing and en route to the analysis of the much better known critic of psychiatry R.D. Laing. Such is the case, for instance, in Peter Sedgwick’s (1982) Psychopolitics. Sedgwick (1982: 108) also tends to blur over the differences between Laing and Cooper, but does point out the latter’s greater radicalism in relation to the family. Liam Clarke (2004: 128–47) focuses at length on Villa 21, but is offhandedly dismissive of Cooper. Zbigniew Kotowicz (1997: 66) is concerned to distinguish Cooper’s supposed revolutionary extremity from the views and practice of Laing, whom he presents as less prey to the excesses of the 1960s. Laing’s son and biographer, Adrian, also distinguishes his father from Cooper. In his biography of Laing (1994: 187), he argues that there was only ever one anti-psychiatrist: Cooper. In a 2005 debate on the legacy of R.D. Laing organised by King’s College London’s Institute of Psychiatry, those arguing that Laing’s legacy has been pernicious presented him as an ‘anti-psychiatrist’, someone who stigmatised families. Adrian Laing and Anthony David, a consultant neuropsychiatrist and professor of cognitive neuropsychiatry, argued for a positive view of Laing’s legacy. They pointed out that Laing was not an anti-psychiatrist and ought not to be confused with David Cooper (A. Laing et al, 2005). Michael Staub (2011: 64), in Madness is civilisation, writes dismissively of Cooper ‘preaching to the choir in the counterculture and New Left’. Staub implies that Cooper is unwilling to enter into fruitful debate about the nature of psychiatry. However, Staub does not recognise the radical nature of Cooper’s project. Like Frantz Fanon (1967) in The wretched of the earth, Cooper addresses his ‘own’ people rather than those in power. He presents himself as a revolutionary addressing revolutionaries or would-be revolutionaries. In the Preface to Fanon’s (1967) The wretched of the earth, Jean-Paul Sartre (Sartre, 1967: 7–26) considers the scandalous nature of the text, which, he tells us, lies in it not being addressed to ‘us’, to white Western readers: Fanon speaks of the West but not to it – his is a Third World audience. Fanon’s message, then, is We do not aspire to be like you, and we do not need you. We shall construct our own lives, our own future. He cannot be assimilated into the values of liberal humanism. This is also so for Cooper in his writing. He also seeks to carve out a space beyond the coordinates of the dominant ideology. He is not interested, then, in a more humane form of psychiatry, or a less stuffy or more contemporary form of psychoanalysis. He does not want simply to ameliorate the bourgeois family. Rather, he writes in favour of a fundamentally different society and addresses himself to those who would construct it. He writes as a revolutionary, and perhaps it is this, above all, that makes Cooper embarrassing or irrelevant to his detractors. What can make Cooper uncomfortable reading at times, however, for even those most sympathetic to him, is his sexual politics. He has been attacked by Elaine Showalter (1985) in The female malady for abuse of the power invested in his status as a therapist. The attack focuses on a story in The grammar of living (Cooper, 1974). Cooper narrates how he met Marja, a young, disturbed Dutch woman – ‘quite tall and quite attractive with long blonde hair’ (p 98), he says at the opening of his brief account – who wanted to speak to him but struggled to do so. He took her home, made love with her and listened (later dividing witnessing responsibilities with someone from a nearby commune) as she spoke over several months about her life. Marja lived in another commune, and, Cooper tells us, eventually found a way to live independently and free of the psychiatric game of which she had previously been part. Cooper’s story ‘exemplified the combination of charisma in the male therapist and infantilism in the female’, writes Showalter (1985: 247); he ‘seemed blind to the ethical issues involved’. Clearly, in Cooper’s narrative, the young woman’s attractive appearance is bound up with her vulnerability and his desire to act as a witness. Cooper says very little about the episode, and we do not have Marja’s account. Showalter’s assumption is that the doctor–patient relationship was abused. By this time, however, Cooper makes it clear that he had become a commune-dweller, had given up psychiatry and had given up practice as a psychotherapist. While ethical issues are certainly raised by Cooper’s story, they ought not to be framed in terms of unprofessional behaviour on the part of a mental health professional. Nevertheless, it is reasonable to assume that much power and charisma would have accrued to someone occupying the position of witness/listener, and few today could read Cooper’s story without unease. Sigal’s (2005) Zone of the interior provides a sharp, satirical critique of anti-psychiatry’s failure to address matters of gender, and, specifically, how the appeal of madness – its ‘sexiness’ – gets mixed up with the physical appeal of distressed women to male anti-psychiatrists. At‘Meditation Manor’, which in Sigal’s novel bears strong resemblances to the actual ‘anti-hospital’ of Kingsley Hall, male guides to residents’ inner voyages become embroiled in what the narrator terms ‘Jurisdictional disputes’ over ‘curvy Jenny Potts’ and ‘luscious, mute Tanya’ (Sigal, 2005: 281). Madness, vulnerability and sexual appeal are sleazily commingled. Cooper’s 1960s’ optimism about the possibility of dissolving differences between the roles of carer and cared-for now seems unfortunate, and even, perhaps, an elegant form of camouflage under the cover of which a male carer might act unethically. Yet solely to identify Cooper’s sexual politics with the abuse of power would be unfair. Cooper, for instance, writes of melting boundaries between supposedly proper and improper sexuality (aligned with hetero- and homosexual activity), and he argues that women can be just as active, as penetrating, as men. Here, he seems remarkably contemporary. In his argument, too, that while mothering and fathering are necessary, these need not be identified with the biological father or mother (Cooper, 1976: 39–53), he again sounds like our contemporary. However, Cooper was a man of his time and it is not surprising that there is a sexist hue to his work. It should be noted, too, that he finds it difficult to move from analysis of the family to thinking about the relationship between the family and wider contexts. Also, like Laing, for Cooper, madness is essentially the same for a man and a woman: he does not attend carefully to the ways in which gender constructs experience. Nor does he pay close attention to race and ethnicity – despite being a socialist from Apartheid South Africa. What we can find in his work, though, is an attempt, albeit a flawed one, to connect everyday life with politics – and this attempt was influential. As the feminist Sheila Rowbotham (2001: 145) remarks in her memoir of the 1960s: ‘Social control was being presented by anti-psychiatry as being embedded in the texture of everyday life, an idea which the women’s movement was later to adapt’. Conclusion: re-Coopering anti-psychiatry In the age of ‘big pharma’, critical perspectives in psychiatry have assumed urgency (Double, 2006; Moncrief, 2009; Davies, 2013), and there has been a revival of interest in the work of R.D. Laing (eg Beveridge, 2011; Miller, 2012; McGeachan, 2014; Chapman, 2015). Laing’s (2010) Divided self is now a Penguin classic, and he is to be the subject of a movie, a biopic starring the former Dr Who star David Tenant. Cooper, however, has been overlooked. It is time, I suggest, for renewed interest in him. Certainly, no detailed account of the anti-psychiatric movement can be complete without sustained attention to Cooper. I would not suggest that we become Cooperians. Rather, I suggest that we return to Cooper’s texts and read them with a spirit of critical friendship. We have a body of work that justifies his place in the tradition of radical opposition to psychiatry and that can nourish thinking and action in the present. It might not be possible to overcome divisions between patients and doctors – or social workers and their clients – as Cooper believed. Nor is it necessarily desirable to do so. Writing in the opening issue of The Sixties journal, the editors remark that: In so many cases the aspiration for change [among 1960s’ radicals] was so much greater than the consequence – ‘the dream,’ however construed, was defeated or denied by entrenched powers. But the dream also often faced internal obstacles, or recklessly overreached. (Varon et al, 2008: 3) Cooper overreached, but his radical opposition to fixity can make us think again about how we might take up positions (eg ‘lecturer’, ‘social worker’, ‘researcher’, ‘therapist’, ‘user’ or ‘survivor’) that blinker and alienate, as well as illuminate, our vision and support us. His fellow-feeling with those considered crazy, together with his attempts to make madness intelligible, are laudable. While his inattention to gender and ethnicity is notable – easily notable for us now – his attention to the politics of everyday life and his belief that ‘treatment’ is best conceived in terms of how people treat one another remains relevant. Cooper’s revolutionary optimism might be thought risible or rather poignant, and there are dangers in overly optimistic political analysis (with the possibility of a reactionary position succeeding when unrealistic optimism fails). Yet hope can also sustain and inspire. Cooper is a figure from another age, certainly; however, what, we might speculate, might he have made of Cognitive Behavioural Therapy (CBT) or mindfulness or perhaps ‘recovery’ in mental health? It is likely that he would have seen them, at best, as false trails, therapies likely to take people away from social solidarity and ‘symptoms’ that might be painful, exciting and illuminating. At worst, he would have seen them, to use his words in The language of madness (Cooper, 1980: 8) (words that express a view akin to that of the current UK survivors’ group Recovery in The Bin [no date]) as strategies ‘invented by the specialised psycho-police agents of final phase capitalist society’. Sloganeering? Yes, perhaps. However, here, as elsewhere in his work, reading Cooper might spark us into thinking anew about the present-day terrain of mental health care. Acknowledgements I would like to warmly thank Dr StephenTicktin for supplying me with several important details about the life of David Cooper; and Dr Gavin Miller, director of the Medical Humanities Research Centre, University of Glasgow, for on-going dialogue concerning R. D. Laing and his associates. Staff at University of Glasgow’s Special Collections Department have been very helpful in aiding my consultation of archive material in the R. D. Laing Collection. The R. D. Laing estate has kindly given permission for me to refer to material in the Laing Collection. This article was written during a Wellcome Trust Research Bursary at the University of Glasgow (Grant reference: 108626/Z/15/Z). Beveridge A Portrait of the psychiatrist as a young man: The early writing and work of R.D. Laing, 1927–1960 2011 Oxford Oxford University Press Chapman A Dismemberment and the attempt at re-membering in R. D. Laing’s The Bird of Paradise Literature and Medicine 2015 33 2 393 418 Clarke L The time of the therapeutic communities: People, places and events 2004 London Jessica Kingsley Cooper D Psychiatry and anti-psychiatry 1967 London Tavistock Cooper D Introduction Cooper D The dialectics of liberation 1968a Harmondsworth Penguin 7 12 Cooper D Beyond words Cooper D The dialectics of liberation Harmondsworth Penguin 1968b 193 202 Cooper D The death of the family 1974 Harmondsworth Penguin Cooper D The grammar of living: An examination of political acts 1976 Harmondsworth Penguin Cooper D Qui sont les dissidents 1977 Paris Galilee Cooper D The language of madness 1980 Harmondsworth Penguin Davies J Cracked: Why psychiatry is doing more harm than good 2013 London Icon Books Deleuze G Guattari F Anti-oedipus: Capitalism and schizophrenia 1977 New York Viking Press Dorley-Brown C The anti-university London 1968 online video 2010 11 11 https://www.youtube.com/watch?v=Kbi_KgBA7-c Double D Critical psychiatry: The limits of madness 2006 Basingstoke Palgrave Macmillan Fanon F The wretched of the earth 1967 Harmondsworth Penguin Foot J Franco Basaglia and the democratic psychiatry movement in Italy, 1961–1968 Critical and Radical Social Work 2014 2 2 235 49 25984302 Foucault M Folie et déraison: Histoire de la folie à l’âge classique 1961 Paris Plon Foucault M Madness and civilisation: A history of insanity in the age of reason 1967 London Tavistock (intro D Cooper) Gale D Far out The Guardian 2001 8 9 http://www.theguardian.com/theguardian/2001/sep/08/weekend7.weekend Kotowicz Z R.D. Laing and the paths of anti-psychiatry 1997 London Routledge Laing RD Cooper D Reason and violence 1964 London Tavistock Laing RD Preface-Introduction to Asylum: To dwell in strangeness 1982 R.D. Laing Collection, University of Glasgow Special Collections MS Laing 221/15A, [unpublished] Laing A R.D. Laing: A biography 1994 London Peter Owen Laing Laing RD The divided self 2010 London Penguin Laing A David A Launer M Revely A Anti-psychiatry is dead, long live psychiatry, Maudsley Debates Kings College London Institute of Psychiatry 2005 29 6 http://www.kcl.ac.uk/ioppn/news/special-events/maudsley-debates/debate-archive-3-30.aspx McGeachan C ‘The world is full of big bad wolves’: investigating the experimental therapeutic spaces of RD Laing and Aaron Esterson History of Psychiatry 2014 25 3 283 98 25114145 Miller G RD Laing’s theological hinterland: the contrast between mysticism and communion History of Psychiatry 2012 23 2 139 55 23057224 Mitchell J Psychoanalysis and feminism 1974 London Allen Lane Moncrief J The myth of the chemical cure: A critique of psychiatric drug treatment 2009 Basingstoke Palgrave Macmillan Mullan B Mad to be normal: Conversations with R.D. Laing 1995 London Free Association Recovery in The Bin no date, https://recoveryinthebin.org/ Rowbotham S Promise of a dream: Remembering the sixties 2001 New York, NY Verso Sartre J-P Preface Fanon F The wretched of the earth 1967 Harmondsworth Penguin 7 26 Sartre J-P Critique of dialectical reason 1976 London NLB Sedgwick P Psychopolitics 1982 London Pluto Press Showalter E The female malady: Women, madness, and English culture 1830–1980 1985 New York, NY Pantheon Books Sigal C Society tomorrow: the doctor who opened the door to the people at villa 21, Shenley hospital’s schizophrenia wing The Guardian 1986 27 8 Sigal C Zone of the interior 2005 Hebdon Bridge Pomona Staub ME Madness is civilisation: When the diagnosis was social, 1948–1980 2011 Chicago, IL Chicago University Press Ticktin S Brother beast Asylum: A Magazine for Democratic Psychiatry 1986 1 3 14 16 Varon J Michael SF Mcmillian J Time is an ocean: the past and future of the sixties The Sixties: A Journal of History, Politics and Culture 2008 1 1 1 7 Wall O The birth and death of villa 21 History of Psychiatry 2013 24 3 326 40 24573448
PMC005xxxxxx/PMC5125516.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 0121103 280 Adv Exp Med Biol Adv. Exp. Med. Biol. Advances in experimental medicine and biology 0065-2598 2214-8019 20919658 5125516 NIHMS830889 Article Sphingolipid Signaling in Fungal Pathogens Rhome Ryan Del Poeta Maurizio * * Corresponding Author: Maurizio Del Poeta—Departments of Biochemistry and Molecular Biology, Microbiology and Immunology and Division of Infectious Diseases, Medical University of South Carolina, Charleston, South Carolina, 29425. [email protected] 20 11 2016 2010 28 11 2016 688 232237 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Sphingolipid involvement in infectious disease is a new and exciting branch of research. Various microbial pathogens have been shown to synthesize their own sphingolipids and some have evolved methods to “hijack” host sphingolipids for their own use. For instance, Sphingomonas species are bacterial pathogens that lack the lipopolysaccharide component typical but instead contain glycosphingolipids (Kawahara 1991, 2006). In terms of sphingolipid signaling and function, perhaps the best-studied group of microbes is the pathogenic fungi. Pathogenic fungi still represent significant problems in human disease, despite treatments that have been used for decades. Because fungi are eukaryotic, drug targets in fungi can have many similarities to mammalian processes. This often leads to significant side effects of antifungal drugs that can be dose limiting in many patient populations. The search for fungal-specific drugs and the need for better understanding of cellular processes of pathogenic fungi has led to a large body of research on fungal signaling. One particularly interesting and rapidly growing field in this research is the involvement of fungal sphingolipid pathways in signaling and virulence. In this chapter, the research relating to sphingolipid signaling pathogenic fungi will be reviewed and summarized, in addition to highlighting pathways that show promise for future research. Sphingolipid Synthesis Sphingolipid synthesis in pathogenic fungi is largely conserved among species. Early steps in the process, such as the condensation of palmitoyl-CoA with serine to form 3-ketodihydrosphingosine, are the same in Saccharomyces cerevisiae. S. cerevisiae is often the gold standard for model systems, but in this context, that system is limited by the fact that S. cerevisiae has very different sphingolipid metabolism from most of the significant fungal pathogens. Both S. cerevisiae and fungal pathogens like Candida albicans make phytoceramide for instance (a modified ceramide with a carbon 4 hydroxylation on its backbone instead of the 3,4 desaturation). Phytoceramide is often conjugated with very long chain fatty acids and can be used to make more complex sphingolipids such as inositol-phosphoryl ceramide (IPC). In addition to this pathway, most pathogenic fungi also synthesize a different ceramide that contains the 3,4 desaturation and cannot be referred to as phytoceramide. This ceramide is typically conjugated to a 16–18 carbon acyl chain and undergoes additional modifications. These modifications include an additional sphingoid backbone desaturation between the 8th and 9th carbons and a methylation of that backbone on the 9th carbon. Typically, the acyl chain is hydroxylated at the α-carbon position. This modified ceramide is the substrate for an enzyme called glucosylceramide synthase (Gcs1), which glycosylates this molecule on the hydroxyl group of the 1st carbon, creating glucosylceramide (GlcCer). Though there is no direct evidence of cross talk between the glucosylceramide synthase pathway and the pathway leading to the synthesis of inositol-phosphoryl ceramide-containing lipids, this possibility cannot be ruled out. The structures of major fungal sphingolipids are found in Figure 1. Cryptococcus Neofomans: Model of Sphingolipid Signaling in Fungi Cryptococcus neoformans is an encapsulated fungal pathogen that primarily affects immunocompromised patients (ex. HIV/AIDS patients, transplant candidates and patients on long-term steroid treatments). This environmental yeast is inhaled to the lungs, where it can live extracellularly or intracellularly (inside the phagolysosomes of alveolar macrophages). Infection with this organism is known as cryptococcosis. In some patients, the fungus disseminates to the bloodstream, seeding many organ systems, but eventually proliferating in the central nervous system. This scenario represents a significant medical emergency, as C. neoformans is the leading cause of fungal meningoencephalitis in the world and the disease is lethal if left untreated. Cryptococcus has several recognized virulence factors, such as the production of melanin and the polysaccharide capsule. Sphingolipid studies in C. neoformans have revealed some interesting implications for signaling pathways involving these molecules. Beyond understanding fungal biology on a cellular and biochemical level, the study of the fungal sphingolipid pathway is advantageous due to the fact that many of the enzymes and products are distinct in structure and function from their mammalian counterparts. This distinction makes them great candidates for drug targets. The best-studied example of this paradigm is inositol-phosphoryl ceramide synthase (Ipc1). This enzyme uses phytoceramide and phosphatidylinositol (PI) as substrates, transferring the phosphorylinositol moiety to phytoceramide. In addition to the generation of inositol-phosphoryl ceramide (IPC), diacylglycerol (DAG) is also released as a product of this reaction. Early studies on Ipc1 in C. neoformans implicated the enzyme in virulence pathways. In strains where Ipc1 is downregulated, melanin production is impaired and the strain has growth deficits when inside alveolar macrophages. When tested in mouse models of cryptococcosis, the strain lacking Ipc1 was less virulent in comparison to the wildtype C. neoformans. Studies on IPC metabolism have also given clues to the role of this reaction in virulence. Inositol phosphosphingolipid-phospholipase C (Isc1) is the enzyme that catalyzes the reverse reaction of Ipc1, which is to remove the phosphotidylinositol component from IPC. A strain of C. neoformans in which this enzyme is deleted (Δisc1) shows reduced virulence in immunocompromised mouse models. However, when macrophages are depleted in this model, Δisc1 will disseminate and cause meningoencephalitis. The Ipc1/Isc1 balance seems to play a role in the interaction between C. neoformans and the alveolar macrophages. Further studies into the mechanism underlying the connection between Ipc1 and virulence of C. neoformans have shown that the production of DAG is a common step of at least two separate determinants of virulence in this fungus. As mentioned, early studies downregulating Ipc1 showed impairments in melanin production. This interaction was found to be mediated by cryptococcal protein kinase C (Pkc1). DAG, the byproduct of Ipc1 activity, was found to bind to the C1 domain of Pkc1. This binding led to an increase in Pkc1 activity and that activation was abolished by the selective deletion of the C1 domain. It is known that Pkc1 and several pathway components are required for proper cell wall integrity, including the function of some cell wall-associated enzymes. One such enzyme is laccase, which is responsible for the synthesis of melanin. The defect in melanin synthesis observed in Ipc1-downregulated strains was caused by improper localization of laccase to the cell wall, due to reduction in DAG-dependent Pkc1 activity. Another way in which DAG production has been linked to virulence involves the fungal-macrophage interaction. C. neoformans has methods to avoid phagocytosis by alveolar macrophages when necessary, including the polysaccharide capsule. One such method is the production of anti-phagocytic protein 1 (App1). When App1 is deleted, the resulting strain (Δapp1) shows reduced virulence in immunocompromised mice. The production of App1 is driven, transcriptionally, by the presence of DAG. DAG binds and activates the transcription factor Atf2. Atf2 activation promotes the transcription of App1 and thus evasion of phagocytosis leading to increased virulence. App1, in addition to regulating phagocytosis, has been shown to bind host complement receptors CR2 and CR3, suggesting even more complex fungal-host interactions affected by production of DAG in C. neoformans. The downstream effects of Ipc1 activity are summarized in Figure 2. Sphingolipid Signaling in Other Pathogenic Fungi While many groups have discovered and characterized sphingolipid components of other pathogenic fungi, few have delved into the role of these lipids in signaling or cellular processes. One such fungus is Candida albicans. C. albicans is a dimorphic fungus that normally lives as a commensal organism in the human gut and urogenital tract. In immunocompromised patients, C. albicans can cause significant disease, including systemic dissemination. Recent evidence in Candida albicans has shown possible sphingolipid involvement in endocytosis and plasma membrane functions. Sur7 is a membrane bound enzyme known to be involved in sphingolipid membrane makeup in S. cerevisiae. When the homolog of Sur7 was deleted in C. albicans, the resulting strain showed defects in hyphal morphogenesis, endocytosis and cell wall formation. Confirming the role of sphingolipids in this process, blocking sphingolipid synthesis resulted in disruption of Sur7 patches in the plasma membrane. Though this has yet to show the definitive role of sphingolipids in this process, their involvement is clear. Also, the function and localization of some multidrug resistance proteins in C. albicans have been shown to be dependent on membrane sphingolipid composition. Conclusion The signaling pathway involving Ipc1 and the production of DAG is clearly related to virulence in more ways than one. What other sphingolipid metabolic pathways could be involved in virulence of C. neoformans and other pathogenic fungi? Recall that in addition to IPC-based sphingolipids, most pathogenic fungi synthesize glucosylceramide that is not based on a phytoceramide backbone. The synthesis of these GlcCers requires enzymes that introduce the desaturation of the sphingoid backbone between carbons 8 and 9 as well as the methylation of the ninth carbon. Examinations into the function of these enzymes as well as glucosylceramide synthase have suggested major roles in biology and virulence. In Candida, for instance, the sphingolipid Δ8-desaturase, responsible for the backbone desaturation at that position, is required for proper hyphal growth. The function of the methyltransferase responsible for the C9 methylation seen in most fungi was studied in the plant pathogen Fusarium graminerum. Disruption of the enzyme encoding this enzyme resulted in a strain that showed defects in virulence, growth and differentiation. When the gene for glucosylceramide synthase is deleted in C. neoformans, the resulting strain shows condition-dependent growth defects as well as a lack of virulence in inhalation mouse models. Taken together, these studies are beginning to uncover the roles of the “glucosylceramide branch” of sphingolipids in pathogenic fungi. Like enzymes involved in IPC production, many of these enzymes are unique to fungi and thus represent attractive possibilities as therapeutic targets. Though the signaling mechanisms involved are unclear, these observations may represent the beginning of new understandings into the role of sphingolipid signaling in pathogenic fungi. Figure 1 Chemical structures of C26 phytoceramide, C26 inositol phosphoryl ceramide, diacylglycerol, C18 α-hydroxy-Δ8, 9methyl-ceramide, C18 α-hydroxy-Δ8, 9methyl-glucosylceramide (fungal glucosylceramide) and C18 glucosylceramide (mammalial glucosylceramide). Figure 2 Sphingolipid signaling in Cryptococcus neoformans. Inositol phosphoryl ceramide synthase 1 (Ipc1) in C. neoformans produces diacylglycerol (DAG) in addition to inositol phosphoryl ceramide IPC). DAG binds to the C1 domain of protein kinase C1 (Pkc1), which is important for cell wall integrity. This integrity is crucial for localization of laccase, the enzyme responsible for melanin synthesis. In addition, DAG also activates the transcription factor Atf2, which leads to transcription of the antiphagocytic protein 1 (App1). Both App1 and melanin regulate pathogenicity of C. neoformans. Hanada K Sphingolipids in infectious diseases Jpn J Infect Dis 2005 58 3 131 148 15973004 Heung LJ Luberto C Del Poeta M Role of sphingolipids in microbial pathogenesis Infect Immun 2006 74 1 28 39 16368954 Kawahara K Sato N Tsuge K Seto Y Confirmation of the anomeric structure of galacturonic acid in the galacturonosyl-ceramide of Sphingomonas yanoikuyae Microbiol Immunol 2006 50 1 67 71 16428876 Kawahara K Moll H Knirel YA Structural analysis of two glycosphingoliids from the lipopolysaccharide-lacking bacterium Sphingomonas capsulata Eur J Biochem 2000 267 6 1837 1846 10712617 Heitman J GFS Edwards JEJ Mitchell AP Molecular Principles of Fungal Pathogenesis Washington American Society of Microbiology 2006 Marr KA New approaches to invasive fungal infections Curr Opin Hematol 2003 10 6 445 450 14564176 Rapp RP Changing strategies for the management of invasive fungal infections Pharmacotherapy 2004 24 2 Pt 2 4S 28S quiz 29S–32S 14992487 McQuiston TJ Haller C Del Poeta M Sphingolipids as targets for microbial infections Mini Rev Med Chem 2006 6 6 671 680 16787378 Rhome R McQuiston T Kechichian T Biosynthesis and immunogenicity of glucosylceramide in Cryptococcus neoformans and other human pathogens Eukaryot Cell 2007 6 10 1715 1726 17693597 Matmati N Hannun YA Thematic review series: sphingolipids. ISC1 (inositol phosphosphingolipid-phospholipase C), the yeast homologue of neutral sphingomyelinases J Lipid Res 2008 49 5 922 928 18305313 Obeid LM Okamoto Y Mao C Yeast sphingolipids: metabolism and biology Biochim Biophys Acta 2002 1585 2–3 163 171 12531550 Sims KJ Spassieva SD Voit EO Obeid LM Yeast sphingolipid metabolism: clues and connections Biochem Cell Biol 2004 82 1 45 61 15052327 Garcia J Shea J Alvarez-Vasquez F Mathematical modeling of pathogenicity of Cryptococcus neoformans Mol Sys Biol 2008 4 183 195 Rittershaus PC Kechichian TB Allegood J Glucosylceramide is an essential regulator of pathogenicity of Cryptococcus neoformans J Clin Invest 2006 116 6 1651 1659 16741577 Casadevall A Perfect JR Cryptococcus neoformans Washington, DC 381 405 ASM Press 1998 Luberto C Toffaletti DL Wills EA Roles for inositol-phosphoryl ceramide synthase 1 (IPC1) in pathogenesis of C. neoformans Genes Dev 2001 15 2 201 212 11157776 Shea J Kechichian TB Luberto C Del Poeta M The cryptococcal enzyme inositol phosphosphingolipid-phospholipase C (Isc1) confers resistance to the antifungal effects of macrophages and promotes fungal dissemination to the central nervous system Infect Immun 2006 74 10 5977 5988 16988277 Heung LJ Luberto C Plowden A The sphingolipid pathway regulates protein kinase C 1 (Pkc1) through the formation of diacylglycerol (DAG) in Cryptococcus neoformans J Biol Chem 2004 279 20 21144 21153 15014071 Heung LJ Kaiser AE Luberto C Del Poeta M The role and mechanism of diacylglycerol-protein kinase C1 signaling in melanogenesis by Cryptococcus neoformans J Biol Chem 2005 280 31 28547 28555 15946943 Gerik KJ Donlin MJ Soto CE Cell wall integrity is dependent on the PKC1 signal transduction pathway in Cryptococcus neoformans Mol Microbiol 2005 58 2 393 408 16194228 Luberto C Martinez-Marino B Taraskiewicz D Identification of App1 as a regulator of phagocytosis and virulence of Cryptococcus neoformans J Clin Invest 2003 112 7 1080 1094 14523045 Mare L Iatta R Montagna MT APP1 transcription is regulated by IPC1-DAG pathway and is controlled by ATF2 transcription factor in Cryptococcus neoformans J Biol Chem 2005 280 43 36055 36064 16129666 Tommasino N Villani M Qureshi A Atf2 transcription factor binds to the APP1 promoter in Cryptococcus neoformans: stimulatory effect of diacylglycerol Eukaryot Cell 2008 7 2 294 301 18083832 Stano P Williams V Villani M App1: an antiphagocytic protein that binds to complement receptors 3 and 2 J Immunol 2009 182 1 84 91 19109138 Alvarez FJ Douglas LM Rosebrock A The Sur7 protein regulates plasma membrane organization and prevents intracellular cell wall growth in Candida albicans Mol Biol Cell 2008 19 12 5214 5225 18799621 Pasrija R Panwar SL Prasad R Multidrug transporters CaCdr1p and CaMdr1p of Candida albicans display different lipid specificities: both ergosterol and sphingolipids are essential for targeting of CaCdr1p to membrane rafts Antimicrob Agents Chemother 2008 52 2 694 704 18056285 Oura T Kajiwara S Disruption of the sphingolipid Delta8-desaturase gene causes a delay in morphological changes in Candida albicans Microbiology 2008 154 Pt 12 3795 3803 19047747 Ramamoorthy V Cahoon EB Thokala M Sphingolipid C-9 methyltransferases are important for growth and virulence but not for sensitivity to antifungal plant defensins in Fusarium graminearum Eukaryot Cell 2009 8 2 217 229 19028992 Kechichian TB Shea J Del Poeta M Depletion of alveolar macrophages decreases the dissemination of a glucosylceramide-deficient mutant of Cryptococcus neoformans in immunodeficient mice Infect Immun 2007 75 10 4792 4798 17664261
PMC005xxxxxx/PMC5125517.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101517297 37639 Transl Stroke Res Transl Stroke Res Translational stroke research 1868-4483 1868-601X 27714669 5125517 10.1007/s12975-016-0502-6 NIHMS821788 Article A post-stroke therapeutic regimen with omega-3 polyunsaturated fatty acids that promotes white matter integrity and beneficial microglial responses after cerebral ischemia Jiang Xiaoyan 12* Pu Hongjian 2* Hu Xiaoming 123 Wei Zhishuo 2 Hong Dandan 2 Zhang Wenting 1 Gao Yanqin 12 Chen Jun 123 Shi Yejie 23 1 State Key Laboratory of Medical Neurobiology and Institutes of Brain Science, Fudan University, Shanghai 200032, China 2 Pittsburgh Institute of Brain Disorders & Recovery and Department of Neurology, University of Pittsburgh, Pittsburgh, PA 15213, USA 3 Geriatric Research, Educational and Clinical Center, Veterans Affairs Pittsburgh Health Care System, Pittsburgh, PA 15261, USA Address correspondence to: Dr. Jun Chen, Pittsburgh Institute of Brain Disorders & Recovery, University of Pittsburgh, S507 Biomedical Science Tower, 3500 Terrace Street Pittsburgh, PA 15213, USA, [email protected] or Dr. Yejie Shi, Pittsburgh Institute of Brain Disorders & Recovery, University of Pittsburgh, S510 Biomedical Science Tower, 3500 Terrace Street, Pittsburgh, PA 15213, USA, [email protected] 22 11 2016 7 10 2016 12 2016 01 12 2017 7 6 548561 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. White matter injury induced by ischemic stroke elicits sensorimotor impairments, which can be further deteriorated by persistent proinflammatory responses. We previously reported that delayed and repeated treatments with omega-3 polyunsaturated fatty acids (n-3 PUFAs) improve spatial cognitive functions and hippocampal integrity after ischemic stroke. In the present study, we report a post-stroke n-3 PUFA therapeutic regimen that not only confers protection against neuronal loss in the gray matter but also promotes white matter integrity. Beginning 2 hours after 60 minutes of middle cerebral artery occlusion (MCAO), mice were randomly assigned to receive intraperitoneal docosahexaenoic acid (DHA) injections (10 mg/kg, daily for 14 days), alone or in combination with dietary fish oil (FO) supplements starting 5 days after MCAO. Sensorimotor functions, gray and white matter injury, and microglial responses were examined up to 28 days after MCAO. Our results showed that DHA and FO combined treatment facilitated long-term sensorimotor recovery and demonstrated greater beneficial effect than DHA injections alone. Mechanistically, n-3 PUFAs not only offered direct protection on white matter components, such as oligodendrocytes, but also potentiated microglial M2 polarization, which may be important for white matter repair. Notably, the improved white matter integrity and increased M2 microglia were strongly linked to the mitigation of sensorimotor deficits after stroke upon n-3 PUFA treatments. Together, our results suggest that post-stroke DHA injections in combination with FO dietary supplement benefit white matter restoration and microglial responses, thereby dictating long-term functional improvements. myelin oligodendrogenesis corpus callosum microglial polarization Introduction White matter, consisting of axonal fiber bundles, myelin-ensheathed axons and myelin-producing oligodendrocytes, plays a fundamental role in transmitting nerve signals and coordinating communication between brain regions [1, 2]. In human stroke, white matter occupies about half of the lesion volume and is an important cause of long-term sensorimotor deficits and cognitive decline [3–5]. Many neuroprotective drugs that showed promise in preclinical testing failed in clinical stroke trials [6]. One major concern is that most, if not all, preclinical studies focus on the protection of gray matter. Therefore, strategies that battle both gray and white matter injury and/or boost white matter repair are urgently needed for the clinical translation of successful preclinical stroke therapies. White matter injury is characterized by demyelination and loss of axonal integrity [2, 7]. Demyelination, or destruction of the myelin sheath, if unrepaired, causes degradation of the naked axons, eventually leading to irreversible neurological disability [1]. White matter repair, including axonal regeneration, oligodendrogenesis, and the myelination of demyelinated or newly generated axons, rebuilds the neuronal circuits and restores axonal conduction [7–11]. Both white matter injury and repair are remarkably influenced by the functional status of the surrounding glial cells, such as astrocytes and microglia [12, 13]. For example, activated microglia can exert dualistic impacts on the white matter in a phenotype-dependent manner. Proinflammatory microglial responses are generally considered to exacerbate oligodendrocyte cell death and demyelination [14, 15], whereas the alternatively activated microglia (the so called “M2” microglia) can resolve local inflammation and promote remyelination, thereby facilitating white matter repair [16, 17]. To this end, therapeutic interventions that are capable of enhancing white matter restoration, either directly through actions on oligodendrocytes, or indirectly through modulation of microglial responses, are promising in improving the functional outcomes after stroke. Long-term prophylactic dietary supplementation of omega-3 polyunsaturated fatty acids (n-3 PUFAs) offers potent protection against ischemic brain injury [18–20]. Furthermore, acute treatment after the onset of stroke with n-3 PUFAs, e.g. docosahexaenoic acids (DHA), appears to be effective in ameliorating neurological deficits and reducing neuronal loss up to 7 days after cerebral ischemia [21–24]. We recently observed that repeated administration of n-3 PUFAs, beginning at 2 h after post-ischemic reperfusion, improved spatial learning and memory in a mouse stroke model, at least in part through enhancing the hippocampal integrity [25]. However, it remains unknown whether a post-stroke n-3 PUFA treatment regimen can promote white matter restoration and improve long-term sensorimotor function recovery. It has been reported that n-3 PUFAs directly protect oligodendrocytes against excitatory cell death [26]. Moreover, we found that n-3 PUFAs potently induce M2 polarization in cultured microglia [27]. Both of these actions by n-3 PUFAs could contribute to white matter protection and/or restoration after stroke. Therefore, the present study was designed to determine the efficacy of post-stroke administration of n-3 PUFAs on promoting white matter integrity and sensorimotor functions using a mouse model of transient focal cerebral ischemia (tFCI). We report here a post-stroke n-3 PUFA treatment regimen that improves white matter restoration by promoting both oligodendrocyte survival and beneficial microglial responses. Materials and methods Animals C57BL/6J mice were obtained from the Jackson Laboratory (Bar Harbor, Maine, USA). Animals were housed in a temperature- and humidity-controlled facility with a 12-h light-dark cycle. Food and water were available ad libitum. Efforts were made to minimize animal suffering and the number of animals used. Transient focal cerebral ischemia model tFCI was induced in adult male C57BL/6J mice (10–12 weeks old) by intraluminal occlusion of the left middle cerebral artery (MCA) for 1 h [28]. Experimental procedures were performed following the Stroke Therapy Academic Industry Roundtable (STAIR) guidelines. Briefly, mice were anesthetized with 3% isoflurane vaporized in 67%:30% N2O/O2 until they were unresponsive to the tail pinch test. Mice were then fitted with a nose cone blowing 1.5% isoflurane for anesthesia maintenance. A 7-0 suture with a silicon-coated tip was introduced into the common carotid artery, advanced to the origin of the MCA, and left undisturbed for 1 h. Rectal temperature was maintained at 37.0 ± 0.5°C during surgery with a temperature-controlled heating pad. To confirm the success of MCA occlusion and reperfusion, regional cerebral blood flow (rCBF) was measured using laser-Doppler flowmetry before, during, and after MCA occlusion (MCAO). Animals that did not show a CBF reduction of at least 75% of baseline levels or died immediately after ischemia induction or reperfusion (less than 10%) were excluded from further experimentation. Delayed n-3 PUFA treatment after stroke Immediately after the MCAO surgery, mice were randomly assigned to 3 groups with the use of a lottery-drawing box: 1) Vehicle control group. Mice were fed a regular laboratory rodent diet (Prolab Isopro RMH 3000 5P76; LabDiet, St. Louis, MO, USA) which has an inherently low n-3 PUFA concentration (0.36%), and received injections of 0.9% NaCl (300 μl per day, i.p. 2 h after MCAO, and then daily for 14 days). 2) DHA injection group. Mice were fed a regular diet, and received injections of DHA (10 mg/kg body weight, diluted with 300 μl of 0.9% NaCl, i.p. 2 h after MCAO, and then daily for 14 days). This dose of DHA injections was determined in a pilot study, which showed the therapeutic window of 2.5–10 mg/kg in the MCAO/reperfusion model (data not shown). 3) Combined DHA injection and fish oil dietary supplementation group. Mice were fed a diet supplemented with n-3 PUFAs (DHA and EPA, triple strength n-3 fish oil, Puritan’s Pride, Oakdale, NY, USA; final n-3 PUFA concentration 4%) [18] 5 days after MCAO for up to 28 days, and received injections of DHA (10 mg/kg body weight, diluted with 300 μl of 0.9% NaCl, i.p. 2 h after MCAO, and then daily for 14 days). We had quantified the food intakes by mice before and after MCAO (60 min)/reperfusion and found that their food intakes decrease after stroke, but fully recover at 4–5 days. Therefore, we started fish oil supplementations at 5 days after stroke to ensure all mice received approximately equal amount of n-3 PUFA supplementation every day. All outcome assessments were performed by investigators blinded to experimental group assignments. Neurobehavioral tests Before and after MCAO, sensorimotor functions of the mice were assessed by the cylinder and rotarod tests. The asymmetry of forelimb use was evaluated by the cylinder test as we described previously [28] before and at 3, 5, 7, 9, 11, 13, 15, 19, 23, 28 days after MCAO. Briefly, mice were placed in a transparent cylinder (9 cm in diameter and 15 cm in height) for 10 min. A camera was fixed above the cylinder to record all the forelimb movements of the mice. Videotapes were analyzed in slow motion, and forepaw (left/right/both) use during the first contact against the cylinder wall after rearing and during lateral exploration was recorded. Preference of the non-impaired forepaw (left) was calculated as a relative proportion of right forepaw contacts: (left-right)/(left+right+both)×100% (asymmetric rate). Uninjured mice typically show no preference for either forepaw, whereas injured mice have increased left forepaw preference depending on the severity of the injury. The rotarod test was performed before and at 3, 5, 7, 10 days after MCAO to assess motor functions [28]. Briefly, mice were forced to run on a machine with accelerated rotating drums (IITC Life Science Inc., Woodland Hills, CA, USA). The time at which mice fell off the rod was recorded (latency to fall). The rotating speed of the rod was set to start at 4 rpm and accelerate to 40 rpm in 300 seconds. Mice were trained for 3 trails per day from 3 days before the surgery. The average time of the 3 trails during the last day of training was recorded as pre-surgery baseline value. After surgery, 5 trials were performed on each testing day with intervals of 5 min between each trial, and the data for trials #3–5 were used to calculate the mean latency to fall on that day. Immunohistochemistry and image analysis At 14 or 28 days after MCAO, mice were deeply anesthetized and transcardially perfused with 0.9% NaCl followed by 4% paraformaldehyde in phosphate-buffered saline (PBS). Mouse brains were harvested and cryoprotected in 30% sucrose in PBS. Frozen serial coronal brain sections (25 μm thick) were prepared on a cryostat (Microm HM459, Thermo Scientific). Brain sections were blocked with 5% donkey serum in PBS for 1 h, followed by incubation with primary antibodies for 1 h at room temperature and overnight at 4°C. After a series of washing, sections were incubated with donkey secondary antibodies conjugated to Alexa Fluor 488 or Cy3 (1:1000, Jackson ImmunoResearch Laboratories, West Grove, PA, USA). Alternate sections from each experimental condition were incubated in all solutions except the primary antibodies to assess non-specific staining. Sections were mounted and coverslipped with Fluoromount-G containing 4′, 6-diamidino-2-phenylindole (DAPI; Southern Biotech, Birmingham, AL, USA). The following primary antibodies were used: rabbit anti-NeuN (1:500; EMD Millipore, Billerica, MA, USA), rabbit anti-myelin basic protein (MBP; 1:500; Abcam, Cambridge, MA, USA), mouse anti-nonphosphorylated neurofilaments (SMI-32; 1:1000; Abcam), mouse anti-adenomatous polyposis coli (APC; 1:400; EMD Millipore), rabbit anti-microtubule-associated protein 2 (MAP2; 1:200; Santa Cruz Biotechnology, Dallas, TX, USA), rat anti-CD16/32 (1:500; BD Biosciences, San Jose, CA, USA), goat anti-CD206 (1:500; R&D Systems, Minneapolis, MN, YSA), rabbit anti-Iba1 (1:1000; Wako, Richmond, VA, USA). Images were acquired using an inverted Nikon Diaphot-300 fluorescence microscope equipped with a SPOT RT slider camera and Meta Series Software 5.0 (Molecular Devices, Sunnyvale, CA, USA). Alternatively, images were captured with an Olympus Fluoview FV1000 confocal microscope using FV10-ASW 2.0 software (Olympus America, Center Valley, PA, USA). Analysis of the images was performed using the ImageJ software by an investigator blinded to experimental group assignments. Chronic brain atrophy was measured on NeuN-stained sections by subtracting the none-lesioned volume (NeuN-positive) of the ipsilateral cortex and striatum from that of the contralateral hemisphere in six brain slices (bregma 1.10 mm to −1.34 mm). The number of mature oligodendrocytes (APC+ cells), M1 microglia/macrophages (CD16/32+/Iba1+ cells), or M2 microglia/macrophages (CD206+/Iba1+ cells) was counted from 1–2 microscopic fields randomly selected from the peri-infarct area (within 300 μm to the infarct). The width of the corpus callosum (CC) was measured on MBP-stained brain sections (bregma 0.5 mm) as previously described [29]. The width of the MBP-immunopositive CC area was measured every 160 μm from the midline. MBP and SMI-32 fluorescence intensity were measured as described previously [26]. Briefly, 2 microscopic fields from the peri-infarct cortex and striatum, and 1 microscopic field from the peri-infarct CC were randomly selected from each brain and acquired using the same imaging settings. Images were then binarized and segmented under a consistent threshold (50%). The total black pixels in each image were then quantified. Examination of recently proliferated cells Recently proliferated cells were labeled with the S-phase marker 5-bromo-2′-deoxyuridine (BrdU; Sigma-Aldrich, St. Louis, Missouri, USA) [18]. Briefly, BrdU was i.p. injected twice a day at a dose of 50 mg/kg body weight at 3–6 days after MCAO. At 28 days after MCAO, mice were sacrificed and coronal brain sections were prepared as described above. Sections were pretreated with 2N HCl for 1 h at 37°C followed by 0.1 M boric acid (pH 8.5) for 10 min at room temperature. Sections were then blocked with M.O.M. kit (Vector, Burlingame, CA, USA) for 1 h, and incubated with purified mouse anti-BrdU antibody (1:1000; BD Biosciences) for 1 h at room temperature and then overnight at 4°C. After a serial of washing, sections were incubated with the 488-AffiniPure donkey anti-mouse IgG (1:1000; Jackson ImmunoResearch Laboratories) for 1 h at room temperature. Fluorescence images were captured as described above. BrdU immunopositive cells were counted using ImageJ and expressed as the number of cells in the designated fields divided by the area (mm2). Oligodendrogenesis was evaluated on BrdU/APC double-stained sections. At least 4 microscopic fields were randomly sampled in each section. Statistical analysis All data are presented as mean ± SEM. The statistical differences among means of multiple groups were assessed by one- or two-way ANOVA followed by the Bonferroni post hoc test. The Pearson product linear regression analysis was used to correlate the multiple histological parameters and sensorimotor behaviors. A p value of less than 0.05 was deemed statistically significant. Results Delayed treatment of DHA and FO after ischemic stroke improves long-term histological and functional outcomes The capability of improving long-term neurofunctional outcomes after stroke is imperative for a potential therapy to be translated from bench to bedside [30–33]. Our recent study revealed beneficial effects of post-stroke DHA injections combined with FO dietary supplementation against stroke-induced cognitive decline [25]. In the present study, we investigated whether the delayed DHA and FO treatments ameliorate long-term sensorimotor deficits after ischemic stroke (Fig. 1a). The cylinder test, which assesses spontaneous forelimb use of rodents, is sensitive in identifying stroke-induced sensorimotor asymmetry and suitable for evaluation of mild and chronic deficits [34]. MCAO induced prolonged asymmetric use of forelimbs in the cylinder test for up to 28 days in all groups of mice, for which an approximately 44% of spontaneous recovery was observed in vehicle-treated mice between day 3 and day 28 (Fig. 1b). Post-stroke DHA injections significantly reduced the asymmetric rate during the testing period (Fig. 1b; p≤0.001 vs. vehicle by two-way ANOVA). Notably, combined DHA and FO treatment demonstrated an even more potent beneficial effect than DHA injections alone (Fig. 1b; p≤0.01 DHA+FO vs. DHA by two-way ANOVA), which became prominent at day 9–28 after MCAO. In the rotarod test, MCAO caused motor deficits in vehicle-treated mice, which was prominent at 3 days after MCAO and largely recovered by 10 days after MCAO (Fig. 1c). Both DHA and DHA+FO treatments effectively reduced motor deficits at 3–10 days after MCAO (Fig. 1c; p≤0.001 DHA+FO or DHA vs. vehicle by two-way ANOVA). Compared to DHA injections alone, combined DHA+FO treatment showed significantly better efficacy (Fig. 1c; p≤0.05 DHA+FO vs. DHA by two-way ANOVA), perhaps as a result of the dramatically improved performance on day 10. The latency to fall of the mice receiving DHA+FO treatment gradually increased at 3–10 days after MCAO, reflecting motor learning during repeated tests. We further examined whether the delayed and repeated n-3 PUFA treatments after stroke could reduce brain tissue loss in the gray and white matter. At 28 days after MCAO, brain atrophy determined on NeuN-stained coronal sections was significantly reduced in mice receiving combined DHA and FO treatment, compared to vehicle-treated mice, whereas DHA treatment alone had no effects (Fig. 1c,d). The number of surviving neurons in the peri-infarct cortical regions was increased after both DHA and DHA+FO treatments compared to control stroke mice, where combined DHA and FO treatment conferred significantly greater effect than DHA injections alone (Fig. 2e). In addition to the protection against injury in the gray matter, delayed n-3 PUFA treatment after stroke partially preserved the integrity of white matter. The corpus callosum (CC), a structure rich in myelinated axons and vulnerable to ischemic injury [35, 36], displayed pathological changes in the gross morphology after MCAO, which is illustrated by the significantly reduced CC width (Fig. 2f). Importantly, post-stroke DHA treatment alone largely maintained the CC width at 28 days after MCAO; the combined treatment with FO showed no further effect (Fig. 2f–h). Post-stroke DHA and FO treatments ameliorate white matter injury We next investigated the microstructural changes of white matter components in the peri-infarct CC, cortex and striatum at 28 days after MCAO using histological indicators (Fig. 2a). Specifically, we evaluated the expression of MBP, a marker for myelin, together with immunohistochemistry using the SMI-32 antibody, which recognizes the nonphosphorylated epitope of neurofilament H, a marker of demyelination [37]. In sham-operated mice, MBP was abundantly expressed in the CC, cortex and striatum, whereas SMI-32 immunofluorescence was readily detectable in the cortex, but barely visible in the CC or striatum (Fig. 2b). The immunofluorescence of MBP and SMI-32 in the non-injured contralateral hemisphere showed similar patterns to that of the sham-operated non-ischemic mice, and n-3 PUFA treatments did not cause detectable alterations in the signal expression of MBP or SMI-32 (Supplementary Fig. 1). In the post-ischemic ipsilateral hemispheres, MCAO impaired the myelin sheath, which was visualized by a reduction of MBP and a concomitant increase of SMI-32 in all three regions examined in vehicle-treated mice (Fig. 2b). These pathological changes were quantified by measuring the ratio of SMI-32/MBP fluorescence intensity (Fig. 2c–e), an indicator of white matter injury and demyelination [14, 35]. Post-stroke DHA injections offered partial but significant protection against MCAO-induced myelin pathology in the CC, cortex and striatum (Fig. 2b–e). However, combined DHA and FO treatment almost abolished MCAO-induced elevations of SMI-32/MBP ratio, indicating a marked protective effect against white matter injury (Fig. 2b–e). Combined DHA and FO treatment enhances post-stroke oligodendrogenesis The improved white matter integrity observed at 28 days after MCAO in mice receiving n-3 PUFA treatments might result from reduction of white matter injury, and/or enhancement of white matter repair. After focal cerebral ischemia, regeneration of myelinating oligodendrocytes is crucial for remyelination, white matter restoration, and neurological recovery [2, 36]. To determine whether n-3 PUFA-associated protection of white matter involved an effect on oligodendrogenesis, we performed double-label immunostaining of BrdU and APC (marker for mature oligodendrocytes [38]) at 28 days after MCAO and quantified the numbers of APC+ and APC+/BrdU+ cells, respectively, in the peri-infarct CC, cortex, and striatum (Fig. 3a–c). In the non-injured contralateral hemisphere, oligodendrogenesis was barely detected at 28 days after MCAO (Supplementary Fig. 2). In the ipsilateral hemisphere, total oligodendrocytes (APC+ cells) were significantly increased in the peri-infarct cortex and striatum of mice receiving combined DHA and FO treatment, whereas DHA injections alone did not result in significant increases (Fig. 3e,f). In addition, DHA and FO combined treatment markedly enhanced post-MCAO oligodendrogenesis in the peri-infarct cortex and striatum, as evidenced by the increased number of mature oligodendrocytes that were expressing BrdU+/APC+ signals (Fig. 3h,i). Interestingly, although combined DHA and FO treatment did not enhance oligodendrogenesis significantly in the CC (Fig. 3g), the total number of oligodendrocytes in the CC was augmented (Fig. 3d; DHA+FO 792.04±42.38 cells/mm2 vs. vehicle 617.28±26.41 cells/mm2, p≤0.01), suggesting that the increase of cell numbers in this region might result from the increased cell survival of oligodendrocytes rather than elevated oligodendrogenesis. These results suggest that DHA injections alone preserved the myelin sheaths against ischemic injury, while combined DHA and FO treatment also promoted oligodendrogenesis. White matter integrity is linked to long-term sensorimotor recovery after stroke The observed improvement of white matter integrity in n-3 PUFA-treated mice likely contributes to post-stroke sensorimotor recovery. We performed Pearson product linear regression analysis to assess the correlation between the white matter histological parameters and neurofunctional performance. The asymmetric rate in the cylinder test showed a significant and strong positive correlation with SMI-32/MBP ratio (Fig. 4a–c; CC r=0.738, p<0.001; cortex r=0.799, p<0.001; striatum r=0.748, p<0.001), suggesting that the preservation of white matter integrity by n-3 PUFA treatments may contribute to the improved sensorimotor recovery. We further examined whether post-stroke oligodendrogenesis was linked to the neurofunctional improvement, and the results revealed interesting topographical differences. In the CC, the total number of oligodendrocytes showed a moderate but statistically significant negative correlation with asymmetric rate in the cylinder test (Fig. 4d; r=−0.481, p=0.032), whereas oligodendrogenesis was not associated with the sensorimotor deficits (Fig. 4g; r=0.124, p=0.614). In the cortex, the number of new born oligodendrocytes (Fig. 4h; r=−0.706, p=0.001), but not the total oligodendrocytes (Fig. 4e; r=−0.340, p=0.132), showed a strong negative correlation with the asymmetry rate. In the striatum, both total oligodendrocytes and newborn oligodendrocytes were moderately, but significantly linked to the levels of sensorimotor deficits (Fig. 4f,i; r=−0.589, p=0.008 and r=−0.575, p=0.006, respectively). In summary, these data indicate that the preservation of white matter integrity might causatively contribute to the improved sensorimotor recovery after stroke. However, the precise mechanisms, e.g. the relative contribution from the stimulated oligodendrogenesis versus oligodendrocyte protection, might differ in different brain regions. Delayed DHA and FO treatments after stroke regulate microglia/macrophage polarization To date, the effect of post-stroke n-3 PUFA treatment on microglial activation remains poorly understood. Ischemic stroke induces the polarization of microglia/macrophages, which may exert phenotypic-dependent impacts on tissue injury and repair [12, 39, 40]. After MCAO, an anti-inflammatory M2 phenotype is initially activated, which is transient and gradually overwhelmed by a persistent pro-inflammatory M1 phenotype [41]. We determined whether DHA injections or combined DHA and FO treatment influences microglial phenotypes by examining the expression of CD16/32 and CD206, markers for M1 and M2 microglia/macrophages, respectively [12]. Immunostaining was done at 14 days after MCAO, a stage when M1 microglia/macrophages peak and M2 microglia/macrophages subside [41]. In sham-operated mice, microglia displayed non-activated, ramified morphology in the CC, cortex and striatum, with low expression of Iba1 and extremely low to undetectable levels of CD16/32 or CD206 (Supplementary Fig. 3). In vehicle-treated mice at 14 days after MCAO (Fig. 5a), CD16/32 was abundantly expressed in Iba1+ microglia/macrophages in the peri-infarct CC, cortex and striatum, consistent with previous reports [41]. DHA injections significantly reduced the numbers of CD16/32+/Iba1+ cells, in all three regions examined; combined DHA and FO treatment had no further effects (Fig. 5b–d). In contrast, CD206 expression was barely detectable in the CC, cortex and striatum of vehicle-treated mice 14 days after MCAO (Fig. 6a). While DHA injections alone caused a moderate increase of CD206+/Iba1+ cell numbers in the cortex and striatum, combined DHA and FO treatment robustly increased M2 microglia/macrophages in the CC, cortex, and striatum (Fig. 6b–d). The phenotypic switch from M1 to M2 microglia elicited by n-3 PUFA treatments may contribute to the improved white matter integrity and sensorimotor recovery after stroke. In the CC, cortex, and striatum, the number of M1 microglia/macrophages at 14 days positively correlated with SMI-32/MBP ratio at 28 days after MCAO (Fig. 5e–g), whereas M2 microglia/macrophages negatively correlated with white matter injury (Fig. 6e–g). Interestingly, the number of M2 microglia/macrophages also showed significantly negative correlation with the asymmetric rate in the cylinder test (Fig. 6h–j); this correlation was absent for M1 microglia/macrophages (Fig. 5h–j). In summary, these data suggest that delayed and repeated treatment of DHA and FO after stroke was capable of modulating microglial polarization toward an anti-inflammatory M2 phenotype. Although both affected white matter integrity, M1 and M2 microglia/macrophages might play different roles in dictating the long-term sensorimotor functions. In particular, the M2 microglia/macrophages may be important for post-stroke neurological recovery. Discussion The present study is the first to investigate the therapeutic efficacy of delayed post-stroke n-3 PUFA treatment against long-term white matter injury and sensorimotor deficits. Our results demonstrated that significant and prolonged improvement of white matter integrity was achieved by combined post-stroke DHA and FO treatment, which let to improved sensorimotor recovery. Furthermore, the sustained white matter protection afforded by n-3 PUFAs is likely attributable to not only oligodendrocyte protection and enhanced oligodendrogenesis, but also a beneficial modulation of post-ischemia microglial responses. White matter is vulnerable to ischemic/reperfusion insult. Therefore, the damage or insufficient repair of white matter contributes to the development of long-term neurological deficits [42]. On the one hand, ischemia/reperfusion induces cell death of the myelin-producing oligodendrocytes and, consequently, white matter demyelination [7]. On the other hand, the continuous presence of oligodendrocyte progenitor cells (OPCs) in the brain provides an opportunity for oligodendrocyte regeneration and white matter repair [2, 43, 44]. Within one week after ischemic injury, endogenous OPCs actively proliferate in the peri-infarct areas [45, 46], suggesting that white matter integrity might be partially reestablished by oligodendrogenesis and replenishment of the lost oligodendrocytes. In the present study, post-stroke DHA injections alone or in combination with FO dietary supplements led to better preserved white matter integrity, where in several parameters, combined DHA and FO treatment demonstrated more robust protection than DHA treatment alone. The beneficial effects by n-3 PUFAs were achieved possibly through the direct preservation of oligodendrocytes or myelin sheaths, the enhancement of oligodendrogenesis and white matter repair, or both. DHA exerts potent protection against AMPA-induced cell death in cultured oligodendrocytes [26]. In addition, long-term elevation of brain n-3 PUFA levels by dietary supplementation or transgenic expression of a n-3 PUFA-generating enzyme potentiates post-ischemia oligodendrogenesis in rodent stroke models [18, 47]. An interesting finding of the present study is that n-3 PUFAs showed topographically different effects on oligodendrocytes in post-stroke brain. Specifically, in the cortex and striatum, DHA and FO combined treatment nearly doubled the number of new born oligodendrocytes compared to vehicle treatment, which likely accounted for the increased total oligodendrocytes (Fig. 3). In contrast, oligodendrogenesis in the CC was not altered following n-3 PUFA treatment despite the increased total number of oligodendrocytes (Fig. 3). These results suggested that in the CC, protection against oligodendrocyte cell death might be a major mechanism underlying the improved white matter integrity. Future studies are warranted to further investigate the contribution of various white matter injury and repair mechanisms to this observed topographical difference. Another interesting finding is that different correlation patterns between oligodendrocytes and post-stroke sensorimotor recovery were observed in different brain regions (Fig. 4). In the CC, the total surviving oligodendrocytes but not the levels of oligodendrogenesis displayed a significant but moderate correlation with sensorimotor performance. In the striatum, both oligodendrogenesis and total oligodendrocytes moderately correlated with sensorimotor performance. In the cortex where the majority of corticospinal tract fibers originate, only the newly generated oligodendrocytes were strongly linked to the sensorimotor recovery. This difference might result from the anatomical and cellular composition of the CC, cortex and striatum, as well as indicates their relative contributions to sensorimotor functions. The observation that a significant amount of BrdU+ cells expressed the mature oligodendrocyte marker APC was to our surprise, as the failure of the newly generated OPCs to fully differentiate into mature, myelinating oligodendrocytes was thought to be a major obstacle that limits remyelination of the post-ischemic white matter [45, 48, 49]. Future functional assessment on the white matter, e.g. action potential transmission in the corpus callosum [26, 35], are needed to explore whether these regenerated mature oligodendrocytes would indeed lead to improved myelination of axons and nerve fiber conduction. Inflammatory responses triggered by microglia/macrophages play an important role in the pathogenesis of stroke [50–52]. Ischemia/reperfusion rapidly activates microglia, which can exert different impacts on the injury and repair of gray and white matter, depending on their phenotypic polarizing state [12]. Under various pathological conditions such as ischemic stroke, traumatic brain injury or multiple sclerosis, M1 microglia/macrophages are generally considered to exacerbate oligodendrocyte cell death and destroy myelin through excessive proinflammatory responses [14, 27], whereas M2 microglia/macrophages facilitate remyelination and tissue repair [16, 36, 53]. A previous study using microglial cultures has shown that DHA and EPA both promote the M2 polarization of microglia by down-regulating M1 signature genes (e.g. TNF-α, IL-1α, CCL5) and up-regulating M2 signature genes (e.g. CD206, TGF-β) [27]. DHA and EPA also inhibit the production and release of proinflammatory mediators, such as TNF-α and NO, from activated microglia [27]. Moreover, DHA and EPA enhance microglial phagocytosis [27], which would facilitate post-injury clearance of tissue debris. These effects of n-3 PUFAs on cultured microglia may, at least in part, explain the beneficial effects of n-3 PUFAs on microglial responses and white matter protection observed in the present in vivo study. After ischemia/reperfusion, the initial M2 microglial polarization is gradually overwhelmed by the destructive M1 polarization, which may lead to progressive tissue injury [41]. In mice receiving post-stroke n-3 PUFA treatments, we observed enhanced polarization of microglia/macrophages towards M2 at 14 days after MCAO. How exactly n-3 PUFAs modulate microglial responses remains largely unknown, although recent studies suggest a role of DHA-containing lipid bodies and their functional interplay with mitochondria [54, 55]. We have previously found that n-3 PUFAs enhance Akt signaling after hypoxia/ischemic brain injury [56]. While Akt is known to exert anti-inflammatory effects on microglia and macrophages [14], it remains to be determined whether n-3 PUFAs induce M2 polarization after stroke. Interestingly, although both M1 and M2 microglia strongly correlated with white matter integrity (Fig. 5e–g, Fig. 6e–g), only M2 microglia significantly correlated with post-stroke sensorimotor recovery (Fig. 5h–j, Fig. 6h–j). A similar correlation between M2 microglia and post-stroke functional recovery was previously noted in aged mice after distal MCAO [53]. In contrast, the M1 microglia correlated poorly with post-stroke sensorimotor deficits (Fig. 5). These results strongly implicate the importance of M2 microglia/macrophages in long-term tissue remodeling, thus promoting the microglia M2 polarization may be a rational therapeutic strategy after brain injury. A potential limitation of the present study is the use of immunohistochemical markers for the identification of M1 and M2 microglia. Since many cells may co-express both M1 and M2 markers, there could be some discrepancies between immunohistochemical results and the actual functional states of cells [57]. Nevertheless, the present study provides an initial screening of n-3 PUFA actions on microglia in relation to long-term stroke outcomes. Future studies on the functional evaluation of microglia phenotypes are strongly warranted. To date, treatment to acute ischemic stroke remains largely limited to recombinant tissue-type plasminogen activator (tPA)-mediated endovascular thrombolysis. It is imperative for stroke research to develop therapies that have a wide treatment time window and ultimately lead to long-term neurological improvement [58–60]. While the protective effects of n-3 PUFAs against ischemic brain injury have been well established in the literature [18, 61], the majority of these previous studies were based on the preventative beneficial effect of n-3 PUFAs when delivered long before stroke onset. In studies that administered n-3 PUFAs within 1 h after post-ischemia reperfusion, neuroprotective effects were observed for up to 3 weeks after stroke [21, 23, 62]. In the current study, we extended the first injection of DHA to 2 h after reperfusion and further elevated brain n-3 PUFA levels in long term by FO dietary supplementation at 5 days after MCAO. Consistent with our recent finding that delayed DHA and FO treatments after MCAO promote cognitive recovery [25], the present study demonstrates long-term beneficial effect of n-3 PUFAs on white matter integrity and microglial responses. While DHA injections alone elicited some improvement in post-stroke histological and functional outcomes, such as improved sensorimotor recovery (Fig. 1), reduced neuronal death (Fig. 1), lessened white matter injury (Fig. 2), and promotion of microglia M2 polarization (Fig. 6), combination with FO dietary supplement demonstrated greater beneficial effects on these parameters. This additive protection might be attributed to two potential mechanisms. Firstly, chronic supplementation of FO could have increased brain n-3 PUFA contents more persistently, compared to DHA injections alone. Secondly, other non-DHA components in the FO, i.e. EPA, might play a major role. EPA normally exists in extremely low levels in the brain but can be dramatically elevated after long-term FO supplement [18]. To facilitate the translation of n-3 PUFA treatment for clinical use, future studies should take into consideration the confounding effects from aging and stroke comorbidities [63–65], and also test the therapeutic efficacy in models involving tPA thrombolysis [66–69]. In summary, our study demonstrates that delayed administration of n-3 PUFAs as late as 2 h after ischemia/reperfusion promotes white matter restoration. Combining DHA injections with FO dietary supplementation significantly elevates post-ischemia oligodendrogenesis and modulates microglial responses toward the beneficial M2 phenotype, both of which correlate with improved long-term sensorimotor functions. Delayed DHA injections combined with FO dietary supplementation thus may be a promising therapy to achieve white matter protection in stroke patients. Supplementary Material 12975_2016_502_MOESM1_ESM Supplementary Fig. 1 Delayed DHA and FO treatments do not affect white matter integrity in the non-injured contralateral hemisphere after MCAO a Representative images of double-label immunostaining of MBP (green) and SMI-32 (red) in the corpus callosum (CC), cortex (CTX) and striatum (STR) of the non-injured contralateral hemisphere at 28 days after MCAO. There was abundant MBP expression but barely any SMI-32 immunosignal in all three regions. Scale bar: 50 μm. b–d Summarized ratio of SMI-32 to MBP fluorescence intensity in the CC, CTX and STR of the contralateral hemisphere. There was no statistical difference among all groups. n=5 mice per group. 12975_2016_502_MOESM2_ESM Supplementary Fig. 2 Delayed DHA and FO treatments do not affect oligodendrogenesis in the non-injured contralateral hemisphere after MCAO a Representative images of double-label immunostaining of BrdU (green) and APC (red) in the corpus callosum (CC), cortex (CTX) and striatum (STR) of the non-injured contralateral hemisphere at 28 days after MCAO. Scale bar: 50 μm. b–d Quantification of the total viable oligodendrocytes in the CC, CTX and STR of the contralateral hemisphere. e–g Quantification of the newly generated oligodendrocytes in the CC, CTX and STR of the contralateral hemisphere. There was no statistical difference among all groups for either the total or newly generated APC+ cells in the three regions examined. n=4 mice per group. 12975_2016_502_MOESM3_ESM Supplementary Fig. 3 Microglia are not activated in mice receiving sham operation Shown are triple-label staining of Iba1 (green), CD16/32 or CD206 (red), and DAPI (blue) in the corpus callosum (CC), cortex (CTX) and striatum (STR) of the ipsilateral hemisphere at 14 days after sham operation. There was low Iba1 immunosignal and no detectable CD16/32 or CD206 immunosignal in all regions. Scale bar: 50 μm. Images represent data from n=5 mice per group. X.J. and H.P. contributed equally to this work. This project was supported by the US Department of Veterans Affairs (VA) RR&D Merit Review RX000420, the US National Institutes of Health grants NS045048, NS091175 and NS095671, the American Heart Association grant 13SDG14570025, and the Chinese Natural Science Foundation grants 81529002, 81171149, 81371306, 81571285 and 81100978. J.C. is a recipient of the VA Senior Research Career Scientist Award. The authors are indebted to Pat Strickler for excellent administrative support. Fig. 1 Delayed treatment of DHA and FO after ischemic stroke elicits long-term improvement in sensorimotor function and gray and white matter integrity Mice were subjected to 1 h of MCAO followed by DHA and FO treatments as described in Methods. a Illustration of the experimental timelines. Mice were pre-trained for the behavioral tests for 3 days before MCAO or sham operation. Two hours after MCAO, mice received DHA injections that lasted for 14 days. FO supplements were administrated 5–28 days after MCAO. Qd: once a day. Bid: twice a day. Sensorimotor functions were evaluated up to 28 days after MCAO by the cylinder and rotarod tests. Mice were sacrificed 14 or 28 days after MCAO for histological examinations. b The cylinder test was performed up to 28 days after MCAO and the asymmetric rate of the forelimb use was shown. c The rotarod test was performed up to 10 days after MCAO. p≤0.05, p≤0.01, *p≤0.001 by two-way ANOVA. n=6–9 mice per group. d Representative images of coronal brain sections showing NeuN immunosignal (red) at 28 days after MCAO. Dashes lines illustrate the chronic brain infarct (NeuN-negative area). Scale bar: 1 mm. e The volume of tissue atrophy in the ipsilateral hemisphere. n=7 mice per group. p≤0.05, p≤0.01, p≤0.001 by one-way ANOVA. f Representative images showing the morphology of the corpus callosum (CC) by MBP immunofluorescence at 28 days after MCAO. Red boxes indicate areas that were enlarged in the high-power images (the 2nd row). Dashed lines show the boundary of the CC. Scale bar: 1mm. g The width of the CC in the ipsilateral hemisphere measured every 160 μm from the midline. Shaded area shows the levels that were illustrated in the bar graph in h. h Summarized data of CC width at 800 μm and 960 μm from the midline. n=6–9 mice per group. p≤0.05, p≤0.01 by one-way ANOVA. Fig. 2 Post-stroke DHA and FO treatments mitigate white matter injury a A representative image of MBP immunofluorescence in a coronal brain section at the level of bregma 0.5 mm at 28 days after MCAO. Boxes show the peri-infarct areas of corpus callosum (CC), cortex (CTX) and striatum (STR) where images from (b) were taken. Scale bar: 1 mm. b Representative images of MBP (green) and SMI-32 (red) immunofluorescence in the peri-infarct corpus callosum, cortex and striatum of the ipsilateral hemisphere, or the corresponding areas in sham-operated mice at 28 days after MCAO. Scale bar: 50 μm. c-e. Quantification of the ratio of SMI-32 to MBP fluorescence intensity in the corpus callosum (c), cortex (d) and striatum (e) of the ipsilateral hemisphere. n=6–9 mice per group. p≤0.01, **p≤0.001 by one-way ANOVA. Fig. 3 Combined DHA and FO treatment after stroke enhances oligodendrogenesis a A representative image of MAP2 immunofluorescence (green) in a coronal brain section at 28 days after MCAO. Boxes illustrate the peri-infarct areas from the corpus callosum (CC), cortex (CTX) and striatum (STR) where images in c were taken. Scale bar: 1 mm. b A representative image showing a newly generated and matured oligodendrocyte identified by double-label immunostaining of BrdU (green) and the mature oligodendrocyte marker APC (red). c Representative images showing double-label immunostaining of BrdU (green) and APC (red) in the peri-infarct corpus callosum, cortex and striatum at 28 days after MCAO. Boxes illustrate areas that were enlarged in the 4th column. Arrow: BrdU+/APC+ cell (yellow). Arrowhead: BrdU+ newly generated cell that is negative for APC signal (green). Scale bar: 50 μm. d–f Quantification of total mature oligodendrocytes in the corpus callosum (d), cortex (e) and striatum (f). g–i Quantification of newly generated mature oligodendrocytes in the corpus callosum (g), cortex (h) and striatum (i). n=6–9 mice per group. p≤0.05, *p≤0.01 by one-way ANOVA. Fig. 4 White matter integrity correlates with sensorimotor recovery after ischemic stroke Pearson product linear regression analysis was performed to correlate post-stroke white matter histological parameters with the mice’s performance in the cylinder test. a–c Correlation of SMI-32/MBP ratio in the corpus callosum (a), cortex (b) and striatum (c) at 28 days after MCAO with the asymmetric rate of forelimb use in the cylinder test at 11–23 days after MCAO. d–i Correlation of total APC+ cell numbers (d–f) or BrdU+/APC+ cell numbers (g–i) in the corpus callosum, cortex and striatum at 28 days after MCAO with the asymmetric rate of forelimb use in the cylinder test at 11–23 days after MCAO. n=6–7 mice per group. Fig. 5 Delayed DHA and FO treatments after stroke reduce microglia/macrophage M1 polarization a Representative images showing double-label immunostaining of Iba1 (green) and CD16/32 (red) in the peri-infarct corpus callosum, cortex and striatum at 14 days after MCAO. Boxes indicate the regions that were enlarged in the 4th column. Arrow: Iba1+/CD16/32+ cell (yellow). Scale bar: 50 μm. b–d Quantification CD16/32+/Iba1+ cells in the corpus callosum (b), cortex (c) and striatum (d). n=3–5 mice per group. p≤0.05, p≤0.01, p≤0.001 vs. vehicle by one-way ANOVA. e-g Pearson correlation between CD16/32+/Iba1+ cell numbers at 14 days after MCAO and SMI-32/MBP ratios at 28 days after MCAO in the corpus callosum (e), cortex (f) and striatum (g). h–j Pearson correlation between CD16/32+/Iba1+ cell numbers in the corpus callosum (h), cortex(i) and striatum (j) at 14 days after MCAO and the asymmetric rate of forelimb use in the cylinder test at 11–23 days after MCAO. n=3–5 mice per group. Fig. 6 Delayed DHA and FO treatments after stroke promote microglia/macrophages M2 polarization a Representative images showing double-label immunostaining of Iba1 (green) and CD206 (red) in the peri-infarct corpus callosum, cortex and striatum at 14 days after MCAO. Boxes indicate the regions that were enlarged in the 4th column. Arrow: Iba1+/CD206+ cell (yellow). Scale bar: 50 μm. b–d Quantification CD206+/Iba1+ cells in the corpus callosum (b), cortex (c) and striatum (d). n=3–5 mice per group. p≤0.05, ***p≤0.001 by one-way ANOVA. e–g Pearson correlation between CD206+/Iba1+ cell numbers at 14 days after MCAO and SMI-32/MBP ratios at 28 days after MCAO in the corpus callosum (e), cortex (f) and striatum (g). h–j Pearson correlation between CD206+/Iba1+ cell numbers in the corpus callosum (h), cortex(i) and striatum (j) at 14 days after MCAO and the asymmetric rate of forelimb use in the cylinder test at 11–23 days after MCAO. n=3–5 mice per group. Author contributions Y.S., X.H., Y.G. and J.C. designed the research. X.J., H.P., Z.W. and W.Z. performed the research. X.J. and Y.S. analyzed the data. X.J., H.P., D.H., J.C. and Y.S. wrote the manuscript. All authors reviewed and edited the manuscript. Compliance with Ethical Standards All animal procedures were approved by the University of Pittsburgh Institutional Animal Care and Use Committee and performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Conflict of Interest The authors declare that they have no conflict of interest. 1 Fields RD A new mechanism of nervous system plasticity: activity-dependent myelination Nat Rev Neurosci 2015 16 12 756 67 26585800 2 Franklin RJ Ffrench-Constant C Remyelination in the CNS: from biology to therapy Nat Rev Neurosci 2008 9 11 839 55 18931697 3 Ho PW Reutens DC Phan TG Wright PM Markus R Indra I Is white matter involved in patients entered into typical trials of neuroprotection? Stroke 2005 36 12 2742 4 16269640 4 Kissela B Lindsell CJ Kleindorfer D Alwell K Moomaw CJ Woo D Clinical prediction of functional outcome after ischemic stroke: the surprising importance of periventricular white matter disease and race Stroke 2009 40 2 530 6 19109548 5 Khan MB Hoda MN Vaibhav K Giri S Wang P Waller JL Remote ischemic postconditioning: harnessing endogenous protection in a murine model of vascular cognitive impairment Transl Stroke Res 2015 6 1 69 77 25351177 6 Gladstone DJ Black SE Hakim AM Heart Stroke Foundation of Ontario Centre of Excellence in Stroke R Toward wisdom from failure: lessons from neuroprotective stroke trials and new therapeutic directions Stroke 2002 33 8 2123 36 12154275 7 Shi H Hu X Leak RK Shi Y An C Suenaga J Demyelination as a rational therapeutic target for ischemic or traumatic brain injury Exp Neurol 2015 272 17 25 25819104 8 Sakamoto M Miyazaki Y Kitajo K Yamaguchi A VGF, Which Is Induced Transcriptionally in Stroke Brain, Enhances Neurite Extension and Confers Protection Against Ischemia In Vitro Transl Stroke Res 2015 6 4 301 8 25921200 9 Plemel JR Keough MB Duncan GJ Sparling JS Yong VW Stys PK Remyelination after spinal cord injury: is it a target for repair? Prog Neurobiol 2014 117 54 72 24582777 10 Kuffler DP An assessment of current techniques for inducing axon regeneration and neurological recovery following peripheral nerve trauma Prog Neurobiol 2014 116 1 12 24380784 11 Van de Velde S De Groef L Stalmans I Moons L Van Hove I Towards axonal regeneration and neuroprotection in glaucoma: Rho kinase inhibitors as promising therapeutics Prog Neurobiol 2015 131 105 19 26093354 12 Hu X Leak RK Shi Y Suenaga J Gao Y Zheng P Microglial and macrophage polarization-new prospects for brain repair Nat Rev Neurol 2015 11 1 56 64 25385337 13 Rossi D Astrocyte physiopathology: At the crossroads of intercellular networking, inflammation and cell death Prog Neurobiol 2015 130 86 120 25930681 14 Wang G Shi Y Jiang X Leak RK Hu X Wu Y HDAC inhibition prevents white matter injury by modulating microglia/macrophage polarization through the GSK3beta/PTEN/Akt axis Proc Natl Acad Sci U S A 2015 112 9 2853 8 25691750 15 Wang G Zhang J Hu X Zhang L Mao L Jiang X Microglia/macrophage polarization dynamics in white matter after traumatic brain injury J Cereb Blood Flow Metab 2013 33 12 1864 74 23942366 16 Miron VE Boyd A Zhao JW Yuen TJ Ruckh JM Shadrach JL M2 microglia and macrophages drive oligodendrocyte differentiation during CNS remyelination Nat Neurosci 2013 16 9 1211 8 23872599 17 Mallucci G Peruzzotti-Jametti L Bernstock JD Pluchino S The role of immune cells, glia and neurons in white and gray matter pathology in multiple sclerosis Prog Neurobiol 2015 127–128 1 22 18 Zhang W Wang H Zhang H Leak RK Shi Y Hu X Dietary supplementation with omega-3 polyunsaturated fatty acids robustly promotes neurovascular restorative dynamics and improves neurological functions after stroke Exp Neurol 2015 272 170 80 25771800 19 Wang J Shi Y Zhang L Zhang F Hu X Zhang W Omega-3 polyunsaturated fatty acids enhance cerebral angiogenesis and provide long-term protection after stroke Neurobiol Dis 2014 68 91 103 24794156 20 Zhang M Wang S Mao L Leak RK Shi Y Zhang W Omega-3 fatty acids protect the brain against ischemic injury by activating Nrf2 and upregulating heme oxygenase 1 J Neurosci 2014 34 5 1903 15 24478369 21 Eady TN Khoutorova L Anzola DV Hong SH Obenaus A Mohd-Yusof A Acute treatment with docosahexaenoic acid complexed to albumin reduces injury after a permanent focal cerebral ischemia in rats PLoS One 2013 8 10 e77237 24194876 22 Eady TN Khoutorova L Obenaus A Mohd-Yusof A Bazan NG Belayev L Docosahexaenoic acid complexed to albumin provides neuroprotection after experimental stroke in aged rats Neurobiol Dis 2014 62 1 7 24063996 23 Eady TN Khoutorova L Atkins KD Bazan NG Belayev L Docosahexaenoic acid complexed to human albumin in experimental stroke: neuroprotective efficacy with a wide therapeutic window Exp Transl Stroke Med 2012 4 1 19 22980673 24 Williams JJ Mayurasakorn K Vannucci SJ Mastropietro C Bazan NG Ten VS N-3 fatty acid rich triglyceride emulsions are neuroprotective after cerebral hypoxic-ischemic injury in neonatal mice PLoS One 2013 8 2 e56233 23437099 25 Pu H Jiang X Hu X Xia J Hong D Zhang W Delayed Docosahexaenoic Acid Treatment Combined with Dietary Supplementation of Omega-3 Fatty Acids Promotes Long-Term Neurovascular Restoration After Ischemic Stroke Transl Stroke Res 2016 Epub ahead of print 26 Pu H Guo Y Zhang W Huang L Wang G Liou AK Omega-3 polyunsaturated fatty acid supplementation improves neurologic recovery and attenuates white matter injury after experimental traumatic brain injury J Cereb Blood Flow Metab 2013 33 9 1474 84 23801244 27 Chen S Zhang H Pu H Wang G Li W Leak RK n-3 PUFA supplementation benefits microglial responses to myelin pathology Sci Rep 2014 4 7458 25500548 28 Shi Y Zhang L Pu H Mao L Hu X Jiang X Rapid endothelial cytoskeletal reorganization enables early blood-brain barrier disruption and long-term ischaemic reperfusion brain injury Nat Commun 2016 7 10523 26813496 29 Rosenzweig S Carmichael ST Age-dependent exacerbation of white matter stroke outcomes: a role for oxidative damage and inflammatory mediators Stroke 2013 44 9 2579 86 23868277 30 Zuloaga KL Zhang W Yeiser LA Stewart B Kukino A Nie X Neurobehavioral and imaging correlates of hippocampal atrophy in a mouse model of vascular cognitive impairment Transl Stroke Res 2015 6 5 390 8 26040424 31 Soliman S Ishrat T Fouda AY Patel A Pillai B Fagan SC Sequential Therapy with Minocycline and Candesartan Improves Long-Term Recovery After Experimental Stroke Transl Stroke Res 2015 6 4 309 22 26004281 32 Brathwaite S Macdonald RL Current management of delayed cerebral ischemia: update from results of recent clinical trials Transl Stroke Res 2014 5 2 207 26 24338266 33 Zhu W Libal NL Casper A Bodhankar S Offner H Alkayed NJ Recombinant T cell receptor ligand treatment improves neurological outcome in the presence of tissue plasminogen activator in experimental ischemic stroke Transl Stroke Res 2014 5 5 612 7 24953050 34 Schaar KL Brenneman MM Savitz SI Functional assessments in the rodent stroke model Exp Transl Stroke Med 2010 2 1 13 20642841 35 Stetler RA Gao Y Leak RK Weng Z Shi Y Zhang L APE1/Ref-1 facilitates recovery of gray and white matter and neurological function after mild stroke injury Proc Natl Acad Sci U S A 2016 113 25 E3558 67 27274063 36 Han L Cai W Mao L Liu J Li P Leak RK Rosiglitazone Promotes White Matter Integrity and Long-Term Functional Recovery After Focal Cerebral Ischemia Stroke 2015 46 9 2628 36 26243225 37 Trapp BD Peterson J Ransohoff RM Rudick R Mork S Bo L Axonal transection in the lesions of multiple sclerosis N Engl J Med 1998 338 5 278 85 9445407 38 Fancy SP Harrington EP Baranzini SE Silbereis JC Shiow LR Yuen TJ Parallel states of pathological Wnt signaling in neonatal brain injury and colon cancer Nat Neurosci 2014 17 4 506 12 24609463 39 Hu X Liou AK Leak RK Xu M An C Suenaga J Neurobiology of microglial action in CNS injuries: receptor-mediated signaling mechanisms and functional roles Prog Neurobiol 2014 119–120 60 84 40 An C Shi Y Li P Hu X Gan Y Stetler RA Molecular dialogs between the ischemic brain and the peripheral immune system: dualistic roles in injury and repair Prog Neurobiol 2014 115 6 24 24374228 41 Hu X Li P Guo Y Wang H Leak RK Chen S Microglia/macrophage polarization dynamics reveal novel mechanism of injury expansion after focal cerebral ischemia Stroke 2012 43 11 3063 70 22933588 42 Pantoni L Garcia JH Gutierrez JA Cerebral white matter is highly vulnerable to ischemia Stroke 1996 27 9 1641 6 discussion 7 8784142 43 Liu X Ye R Yan T Yu SP Wei L Xu G Cell based therapies for ischemic stroke: from basic science to bedside Prog Neurobiol 2014 115 92 115 24333397 44 Duncan ID Brower A Kondo Y Curlee JF Jr Schultz RD Extensive remyelination of the CNS leads to functional recovery Proc Natl Acad Sci U S A 2009 106 16 6832 6 19342494 45 Chu M Hu X Lu S Gan Y Li P Guo Y Focal cerebral ischemia activates neurovascular restorative dynamics in mouse brain Front Biosci (Elite Ed) 2012 4 1926 36 22202008 46 Mandai K Matsumoto M Kitagawa K Matsushita K Ohtsuki T Mabuchi T Ischemic damage and subsequent proliferation of oligodendrocytes in focal cerebral ischemia Neuroscience 1997 77 3 849 61 9070757 47 Hu X Zhang F Leak RK Zhang W Iwai M Stetler RA Transgenic overproduction of omega-3 polyunsaturated fatty acids provides neuroprotection and enhances endogenous neurogenesis after stroke Curr Mol Med 2013 13 9 1465 73 23971733 48 Syed YA Baer AS Lubec G Hoeger H Widhalm G Kotter MR Inhibition of oligodendrocyte precursor cell differentiation by myelin-associated proteins Neurosurg Focus 2008 24 3–4 E5 49 Segovia KN McClure M Moravec M Luo NL Wan Y Gong X Arrested oligodendrocyte lineage maturation in chronic perinatal white matter injury Ann Neurol 2008 63 4 520 30 18393269 50 Seifert HA Pennypacker KR Molecular and cellular immune responses to ischemic brain injury Transl Stroke Res 2014 5 5 543 53 24895236 51 Li L Tao Y Tang J Chen Q Yang Y Feng Z A Cannabinoid Receptor 2 Agonist Prevents Thrombin-Induced Blood-Brain Barrier Damage via the Inhibition of Microglial Activation and Matrix Metalloproteinase Expression in Rats Transl Stroke Res 2015 6 6 467 77 26376816 52 Zhou Y Wang Y Wang J Anne Stetler R Yang QW Inflammation in intracerebral hemorrhage: from mechanisms to clinical translation Prog Neurobiol 2014 115 25 44 24291544 53 Suenaga J Hu X Pu H Shi Y Hassan SH Xu M White matter injury and microglia/macrophage polarization are strongly linked with age-related long-term deficits in neurological function after stroke Exp Neurol 2015 272 109 19 25836044 54 Chang PK Khatchadourian A McKinney RA Maysinger D Docosahexaenoic acid (DHA): a modulator of microglia activity and dendritic spine morphology J Neuroinflammation 2015 12 34 25889069 55 Tremblay ME Zhang I Bisht K Savage JC Lecours C Parent M Remodeling of lipid bodies by docosahexaenoic acid in activated microglial cells J Neuroinflammation 2016 13 1 116 27220286 56 Zhang W Liu J Hu X Li P Leak RK Gao Y n-3 Polyunsaturated Fatty Acids Reduce Neonatal Hypoxic/Ischemic Brain Injury by Promoting Phosphatidylserine Formation and Akt Signaling Stroke 2015 46 10 2943 50 26374481 57 Ransohoff RM A polarizing question: do M1 and M2 microglia exist? Nat Neurosci 2016 19 8 987 91 27459405 58 Pena I Borlongan CV Translating G-CSF as an Adjunct Therapy to Stem Cell Transplantation for Stroke Transl Stroke Res 2015 6 6 421 9 26482176 59 Liu J Wang Y Akamatsu Y Lee CC Stetler RA Lawton MT Vascular remodeling after ischemic stroke: mechanisms and therapeutic potentials Prog Neurobiol 2014 115 138 56 24291532 60 Ruan L Lau BW Wang J Huang L Zhuge Q Wang B Neurogenesis in neurological and psychiatric diseases and brain injury: from bench to bedside Prog Neurobiol 2014 115 116 37 24384539 61 Belayev L Khoutorova L Atkins KD Bazan NG Robust docosahexaenoic acid-mediated neuroprotection in a rat model of transient, focal cerebral ischemia Stroke 2009 40 9 3121 6 19542051 62 Hong SH Khoutorova L Bazan NG Belayev L Docosahexaenoic acid improves behavior and attenuates blood-brain barrier injury induced by focal cerebral ischemia in rats Exp Transl Stroke Med 2015 7 1 3 25642315 63 Fjell AM McEvoy L Holland D Dale AM Walhovd KB Alzheimer’s Disease Neuroimaging I What is normal in normal aging? Effects of aging, amyloid and Alzheimer’s disease on the cerebral cortex and the hippocampus Prog Neurobiol 2014 117 20 40 24548606 64 Mandava P Martini SR Munoz M Dalmeida W Sarma AK Anderson JA Hyperglycemia worsens outcome after rt-PA primarily in the large-vessel occlusive stroke subtype Transl Stroke Res 2014 5 4 519 25 24699843 65 Hafez S Coucha M Bruno A Fagan SC Ergul A Hyperglycemia, acute ischemic stroke, and thrombolytic therapy Transl Stroke Res 2014 5 4 442 53 24619488 66 Ansar S Chatzikonstantinou E Wistuba-Schier A Mirau-Weber S Fatar M Hennerici MG Characterization of a new model of thromboembolic stroke in C57 black/6J mice Transl Stroke Res 2014 5 4 526 33 24347404 67 Mandava P Shah SD Sarma AK Kent TA An Outcome Model for Intravenous rt-PA in Acute Ischemic Stroke Transl Stroke Res 2015 6 6 451 7 26385545 68 Lapchak PA Critical early thrombolytic and endovascular reperfusion therapy for acute ischemic stroke victims: a call for adjunct neuroprotection Transl Stroke Res 2015 6 5 345 54 26314402 69 Zhou Y Murugappan SK Sharma VK Effect of clot aging and cholesterol content on ultrasound-assisted thrombolysis Transl Stroke Res 2014 5 5 627 34 24488442
PMC005xxxxxx/PMC5125539.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101278296 36311 Expert Rev Clin Pharmacol Expert Rev Clin Pharmacol Expert review of clinical pharmacology 1751-2433 1751-2441 25916666 5125539 10.1586/17512433.2015.1034689 NIHMS830526 Article Pharmacologic and clinical evaluation of posaconazole Moore Jason N * Healy Jason R Kraft Walter K Department of Pharmacology and Experimental Therapeutics, Thomas Jefferson University, 132 South Tenth Street, Main Building, Room 1170, Philadelphia, PA 19107, USA * Author for correspondence: Tel.: +1 215 955 9081, Fax: +1 215 955 5681, [email protected] 19 11 2016 5 2015 28 11 2016 8 3 321334 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Posaconazole, a broad-spectrum triazole antifungal agent, is approved for the prevention of invasive aspergillosis and candidiasis in addition to the treatment of oropharyngeal candidiasis. There is evidence of efficacy in the treatment and prevention of rarer, more difficult-to-treat fungal infections. Posaconazole oral suspension solution has shown limitations with respect to fasting state absorption, elevated gastrointestinal pH and increased motility. The newly approved delayed-release oral tablet and intravenous solution formulations provide an attractive treatment option by reducing interpatient variability and providing flexibility in critically ill patients. On the basis of clinical experience and further clinical studies, posaconazole was found to be a valuable pharmaceutical agent for the treatment of life-threatening fungal infections. This review will examine the development history of posaconazole and highlight the most recent advances. antifungal immunosuppression invasive fungal infection pharmacokinetics pharmacology posaconazole triazole Invasive fungal infections (IFIs) are problematic for critically ill patients, with increased risks of morbidity and mortality. Prophylactic treatment is often advantageous because delayed antifungal treatment has been shown to increase mortality rates [1,2]. The Infectious Disease Society of America guidelines for the use of antimicrobial agents in neutropenic patients with cancer suggest that antifungal prophylaxis be used in high-risk patients, including those undergoing hematopoietic stem-cell transplantation (HSCT) or intensive chemotherapy for leukemia [3]. Unfortunately, safety and tolerability concerns often reduce the use of antifungals in the clinic. The currently available systemic triazoles are divided into two groups: the first generation (fluconazole and itraconazole) and the second generation (voriconazole, posaconazole and isavuconazonium; Figure 1). Although these agents all possess the same mechanism of action, each has differing antifungal activity, efficacy, pharmacokinetics and safety profiles, leading to unique therapeutic niches [4,5]. Posaconazole is a triazole antifungal that boasts an extended-spectrum of activity for prophylaxis and treatment of IFIs. Posaconazole has demonstrated efficacy as an antifungal prophylactic in HSCT recipients with graft versus host disease (GVHD) and in neutropenic patients with hematologic malignancy. In addition, posaconazole has been an effective salvage therapy option for patients who are nonresponsive to standard antifungal therapies [6], Overall, posaconazole covers a wide array of IFIs, including aspergillosis, candidiasis, fusariosis, mucormycosis, cryptococcosis, chromoblastomycosis, mycetoma and coccidioidomycosis [7]. Compared with the older azoles (fluconazole, itraconazole and voriconazole), posaconazole has a more favorable safety profile [8]. Furthermore, posaconazole's activity extends beyond that of other azoles, including voriconazole, for instance, that does not cover mucormycosis [9,10]. Chemical development Posaconazole is designated chemically as 4-[4-[4-[4-[[(3R,5R)-5-(2,4-difluorophenyl)tetrahydro-5-(1H-1,2,4-triazol-1-ylmethyl)-3-furanyl]methoxy]phenyl]-1-piperazinyl]phenyl]-2-[(1S,2S)-1-ethyl-2-hydroxypropyl]-2,4-dihydro-3H-1,2,4-triazol-3-one. Its empirical formula is C37H42F2N8O4 with a molecular weight of 700.8. It is synthesized solely as the (R,R,S,S) enantiomer via a three-step synthesis followed by a micronization step to enhance the dissolution rate. Posaconazole is structurally comparable with itraconazole (Figure 1). Modifications, including fluorine in place of chlorine and a furan ring in place of the dioxolane ring, result in an extended spectrum of antifungal activity [7]. Pharmacology Posaconazole, in addition to other triazoles, block ergosterol synthesis through 14 α-demethylase (CYP51) inhibition. Ergosterol depletion prevents proper fungal cell wall construction and causes the accumulation of methylated sterol precursors leading to cell death [11]. Posaconazole undergoes negligible oxidative Phase I metabolism (<2%); its metabolism is facilitated instead through a Phase II biotransformation via uridine disphosphate-glucuronosyltransferase enzyme pathway. No relevant active metabolites have been identified for posaconazole [12]. Posaconazole is both a substrate and an inhibitor of the p-glycoprotein efflux transporter [13]. The Biopharmaceutics Classification System of the US FDA Center for Drug Evaluation and Research lists posaconazole as a Class II compound, indicating that it is well absorbed but dissolves slowly (high permeability/low solubility) [14]. The apparent volume of distribution of posaconazole ranges from 5 to 25 l/kg, indicating extensive distribution and tissue penetration [7]. Posaconazole is highly protein bound (>98%), predominately to albumin in a concentration-dependent fashion. Posaconazole is predominately eliminated in the feces (77% of the radiolabeled dose), primarily eliminated as parent drug (66% of the radiolabeled dose) in healthy volunteers [15]. Fourteen percent of the radiolabeled dose is excreted in urine, primarily in the form of glucuronide conjugates. The mean half-life (t1/2) of posaconazole ranges from 25 to 35 h [16]. Posaconazole has shown in vitro activity against a wide variety of fungal pathogens, including Aspergillus spp., Candida spp., Coccidioides immitis and Fonsecaea pedrosoi. In addition, some species of Fusarium, Rhizopus and Mucor are sensitive to posaconazole [17]. Clinical Aspergillus fumigatus isolates have been identified that demonstrate resistance to posaconazole, specifically those that harbor CYP51 mutations [18]. There are no established pharmacokinetic guidelines with respect to plasma posaconazole for breakthrough IFIs [19]. A posaconazole concentration target greater than 0.50 mg/l is recommended for prophylactic treatment, with others suggesting a concentration target greater than 0.70 mg/l. Cardiothoracic transplant patients having posaconazole levels consistently exceeding 0.50 mg/l had therapeutically successful outcomes [20]. One report suggests that values exceeding 0.70 mg/l do not provide any further reduction in the clinical failure rate, as demonstrated in two randomized, active-controlled clinical studies using posaconazole oral suspension [21]. Oral suspension formulation The posaconazole oral suspension solution has been reviewed extensively [7,22]; its use has been eclipsed by the newer formulations. Briefly, there are several challenges with respect to the posaconazole oral suspension formulation. One, this formulation is limited by saturable absorption. In a clinical study involving healthy men, posaconazole oral suspension given 400 mg every 12 h or 200 mg every 6 h resulted in a 98 and 220% increase in bioavailability, respectively, compared with 800 mg given as a single dose [23]. Second, there is also significant pharmacokinetic variability in regard to nutrition as bioavailability increases at a low gastric pH along with high-fat meals. [24,25]. Acidic carbonated beverages have been shown to increase the bioavailability of posaconazole oral suspension [26]. Delayed-release oral tablet formulation The FDA approved a delayed-release oral posaconazole tablet in November 2013. This tablet was largely designed to overcome the absorption limitations associated with the oral suspension as seen in the clinic and in previous studies. The current posaconazole delayed-release tablet is designed to reduce active drug release at low gastric pH, while increasing release at the elevated pH of the intestine with an absolute bioavailability of 54% for the oral delayed-release tablet. It is suggested that the oral delayed-release tablet be taken with food, although it is not known if the oral bioavailability of the tablet improves under fed conditions [16]. A single- and multiple-dose study was performed in healthy subjects to ascertain optimal dosing and assess pharmacokinetics of this posaconazole tablet formulation [27,28]. During the single 100-mg dose study, the delayed-release tablet formulation had a higher maximum (peak) serum concentration (Cmax) compared with the oral suspension solution in the fasted state (0.39 vs 0.08 mg/l, respectively). In the fed state, the delayed-release tablet formulation still maintained a higher Cmax compared with the oral suspension formulation (0.33 vs 0.24 mg/l, respectively). For the multidose study, subjects were randomized to one of two cohorts. Cohort 1 consisted of either placebo or posaconazole 200 mg single dose on day 1, a 5-day washout period, and then 200 mg BID on day 6, 200 mg QD on days 7–14 and 200 mg BID on days 15–22. Cohort 2 received 400 mg, as opposed to 200 mg, and had the same schedule as cohort 1 until day 14. Median time to maximum concentration (Tmax) was 4 h for the 200-mg dose and 5 h for the 400-mg dose, while mean t1/2 following was similar for both doses (25 and 26 h for the 200- and 400-mg dose, respectively). With the 400-mg dose, the delayed-release tablet exhibited linear pharmacokinetics and steady-state concentrations were achieved after 7 days. Greater intersubject variability was noted in exposure values for the 400-mg dose compared with the 200-mg dose (54 vs 32%, respectively). A Phase Ib, multicenter dose-determining trial was conducted to examine the pharmacokinetics of posaconazole tablets in patients with acute myeloid leukemia (AML) or myelodysplasia [29]. The two cohorts were administered posaconazole tablets, 200 or 300 mg daily. The 300-mg cohort reached the predefined steady-state concentration goal of 0.50 and 2.50 mg/l in 97% of patients, whereas 79% of patients reached the goal in the 200-mg cohort. A Phase III multicenter trial used the 300-mg delayed-release posaconazole oral tablet to further examine the pharmacokinetics in AML or myelodysplasia patients along with recent HSCT recipients [30]. During the 28-day trial period, the predefined steady-state concentration goal, set at 0.50–3.75 mg/l, was achieved in 96% of patients, with 81% falling in the range between 0.50 and 2.50 mg/l. These concentration goals are in line with targeted recommendations for breakthrough IFIs. Intravenous formulation An intravenous posaconazole formulation developed as an aqueous solution containing the solubilizer sulfobutyl ether beta-cyclodextrin has also been approved for marketing in the USA [31]. Initially, a single-center, two-part rising single- and multiple-dose study in healthy adults was performed to evaluate the pharmacokinetics and safety of intravenous posaconazole [32]. For the first part of the study, six cohorts covered a single-dose posaconazole range from 50 to 300 mg by 30 min peripheral infusion. Intravenous posaconazole showed a greater-than-dose-proportional increase in exposure, whereby Cmax values ranged from 0.31 to 2.84 mg/l for the 50- and 300-mg single-dose administration, respectively. Part 2 was terminated early due to unacceptable rates of infusion site reactions. To expand the pharmacokinetic and safety profile of intravenous posaconazole, a two-part study was performed, one part classified as Phase Ib and the other as Phase III to bridge the new posaconazole intravenous solution to the previously approved posaconazole suspension [31]. The primary purpose of the Phase Ib trial was to identify the posaconazole dose that would attain an exposure target of 0.50–2.50 mg/l. A single-dose and two multiple-dose cohorts were established for the study, with the multiple dose cohorts used for evaluating the 200 or 300 mg once daily dose after a twice-daily loading dose on the first day. Subjects attaining steady-state exposure goals were 94 and 95% for the 200- and 300-mg dosing cohorts, respectively. Mean concentration average was 1.19 and 1.43 mg/l for the 200- and 300-mg dosing cohorts on day 14, respectively. Intravenous posaconazole demonstrated a similar safety profile to the oral suspension formulation. A 300-mg QD dose was recommended for the Phase III study. The pharmacokinetics and safety of the posaconazole intravenous formulations have also been investigated [33]. Patients with AML, myelodysplastic syndrome or HSCT were enrolled into a study that further examined the posaconazole intravenous formulation [34]. Posaconazole was intravenously administered 300 mg twice daily on day 1 and 300 mg once daily for 4–13 days afterward. Then, they were switched to posaconazole oral suspension 600 or 800 mg in divided doses for up to 23 days for a total treatment period of 28 days. The intravenous formulation resulted in higher trough concentrations of posaconazole than either of the dosages of oral suspension. Thus, this result shows that the posaconazole intravenous solution can be dosed at a level to reach satisfactory exposure for the treatment or prophylaxis of fungal infections. Safety & tolerability In general, posaconazole has a very good safety and tolerability profile [35,36]. Courtney et al. demonstrated the tolerability of oral suspension posaconazole in doses up to 400 mg twice daily in a Phase I study in healthy subjects [37]. Adverse effects were mild, such as fatigue and dry mouth. In the later phase clinical trials, posaconazole was also particularly well tolerated. The main side effects experienced by the participants were gastrointestinal distress (nausea, vomiting and diarrhea), neutropenia and elevated liver enzymes [34,38,39]. Patients suffering from mucositis, diarrhea or in the early post-transplant period in hematopoietic stem cell transplant therapy had reduced posaconazole levels when administered the oral suspension solution [19]. Overall, posaconazole has favorable safety profile compared with other currently approved systemic triazole antifungals (Table 1) [8]. Other than infusion site reactions, the intravenous formulation of posaconazole had few additional adverse effects. Infusion site reactions, specifically thrombophlebitis, were clinically acceptable at 30 min compared with 90 min for single-dose peripheral administration [32]. However, a decrease in infusion time from 90 to 30 min did not reduce infusion site reactions for multiple-dose administrations. It is recommended that infusion be performed by central line when multiple-dose posaconazole administration is necessary. Early pooled analysis of 18 controlled studies in healthy volunteers and patients receiving posaconazole oral suspension solution indicated low potential to prolong the corrected QT (QTc) interval, among other side effects [40]. An initial report identified minimal safety concerns regarding elevated hepatic function tests and QTc prolongation for the posaconazole delayed-release oral tablets [29]. However, an additional study has shown that the oral delayed-release tablet increases hepatic enzyme levels in addition to prolonging the QTc interval [41]. As such, patients need to be monitored of their hepatic enzyme levels and QTc intervals. Drug interactions The other triazole antifungals have numerous drug–drug interactions because they inhibit the p-glycoprotein transporter in addition to CYP P450 enzymes, thus resulting in increased concentration of other drugs [42]. However, posaconazole and fluconazole are less potent inhibitors than voriconazole and itraconazole. Posaconazole inhibits CYP3A4 and p-glycoprotein, whereas other triazoles may also affect CYP2C9 and CYP2C19 [22]. This has important implications for therapy selection. In the GVHD patient population, immunosuppressive drugs, cyclosporine and tacrolimus, are commonly used; as they are CYP 3A4 substrates, posaconazole will likely increase their plasma concentration, potentially resulting in toxicity (Figure 2). The coadministration of posaconazole with several CYP3A4 substrates is contraindicated, including substrates known to prolong the QTc interval (i.e., terfenadine, cisapride), in addition to HMG-CoA reductase inhibitors (i.e., statins) and ergot alkaloids [16]. Reduced posaconazole oral suspension exposure is noted with the concurrent use of metoclopramide, phenytoin or rifampin, and the H2 ranitidine [19]. Because the posaconazole oral suspension formulation is more readily absorbed at a lower pH, proton pump inhibitors and cimetidine have been shown to decrease the area under the curve of posaconazole [7,43]. Other histamine H2 receptor antagonists and antacids had no effect on the area under the curve. In contrast to the oral suspension formulation, the absorption profile of the delayed-release posaconazole tablet is not affected by gastric acidity or motility. To investigate, healthy volunteers were randomized to groups to receive a single 400-mg dose of the delayed-release posaconazole tablet alone; with metoclopramide to affect gastric motility; or with an antacid (aluminum and magnesium hydroxide), ranitidine or esomeprazole to affect gastric acidity in a crossover trial [44]. Exposure, Tmax and t1/2 were comparable whether posaconazole was administered alone or in combination with medications that affect gastric pH and motility. Additional studies demonstrated increased plasma concentrations associated with the delayed-release oral tablet as leukemia patients transitioned from posaconazole oral suspension to tablets had significantly higher posaconazole concentrations (median, suspension 0.75 mg/l, tablet 1.91 mg/l without clinically relevant hepatotoxicity) [45]. Use in special populations Posaconazole pharmacokinetics are comparable regardless of gender and are not significantly affected by ethnicity [16]. Furthermore, posaconazole pharmacokinetics do not differ significantly with age. Posaconazole is designated pregnancy category C, such that no adequate clinical studies have examined this population [16]. Posaconazole administration has led to skeletal malformations in rats at relative concentrations lower than human therapeutic dosing. Furthermore, posaconazole administration produced higher rates of bone resorption to occur in rabbits, with higher dosages causing a reduction in body weight gain and a reduction in liter size in females. Preclinical animal models also suggest that posaconazole may excrete into breast milk of lactating females. Prophylaxis studies indicate that mean steady-state posaconazole average concentration is consistent in the pediatric population with that of adults [16]. For prophylaxis of candidiasis, posaconazole has been indicated in several clinical contexts in pediatric patients [46]. For allogenic HSCT, posaconazole oral suspension (200 mg TID) is recommended for patients greater than grade II GVHD who are at least 13 years of age. Posaconazole oral suspension (200 mg TID) is recommended for AML and recurrent leukemia patients ages 13 and older after the last dose of chemotherapy until neutrophil recovery. Delayed-release posaconazole tablets are now indicated for patients 13 years or older. However, the use of posaconazole intravenous injection is not recommended for patients under the age of 18 due to preclinical safety concerns [16]. No posaconazole pharmacokinetic effects have been seen in patients administered 400 mg single-dose posaconazole oral suspension with mild-to-moderate renal impairment; thus, no further dose adjustment requirements are recommended for patients with mild-to-moderate renal impairment [16]. However, close monitoring for breakthrough IFI is recommended for patients with severe renal impairment (eGFR: <20 ml/min). Similar studies have not been conducted with the oral delayed-release posaconazole tablets; however, no dose adjustments are recommended for mild-to-moderate renal impairment. Posaconazole intravenous injection should be avoided in patients with moderate or severe renal impairment as excess accumulation of the intravenous vehicle, sulfobutylether-β-cyclodextrin, may be problematic. Accordingly, the serum creatinine levels of these patients should be closely monitored [16]. No posaconazole oral suspension dose adjustments are recommended for patients with mild-to-severe hepatic impairment [16]. Mean area under the curve values of a single dose of 400-mg posaconazole oral suspension range from 21 to 43% higher for patients with hepatic impairments. Respective Cmax and t1/2 values for patients with hepatic impairments vary compared with normal individuals. Similar studies have not been performed with the oral delayed-release tablets or intravenous injection; however, no dose adjustments are recommended for either of these formulations in patients with mild-to-severe hepatic impairment [16]. Immunosuppressant drugs, specifically cyclosporine and tacrolimus, need to have levels monitored when given in combination with posaconazole [47]. Although concurrent use of posaconazole and sirolimus is contraindicated by the manufacturer, a recent study has shown that posaconazole can be given in combination with sirolimus in a liver solid organ transplant patient. [48]. It is imperative that sirolimus concentration levels are monitored when given in conjunction with posaconazole. Clinical efficacy studies Phase II clinical studies Two main Phase II trials were completed in the clinical development program of posaconazole. For the indication of IFI prophylaxis, study P018893 was designed to establish the dose by comparing the pharmacokinetic and pharmacodynamic properties of different posaconazole dosing strategies [16,49,50]. This study focused on the treatment of patients with azole-refractory IFIs or patients experiencing febrile neutropenia and requiring empiric antifungal therapy. Ninety-eight patients were randomized to receive posaconazole as an oral suspension between 800 and 1600 mg every day divided into different dosing schedules. Patients continued to receive the drug for a maximum of 6 months or until febrile neutropenia resolved. Posaconazole was well tolerated and was comparably effective at all dosing schedules in the treatment of the fungal infections. These results agreed with previous pharmacokinetic data, suggesting that the absorption of posaconazole is saturable and therefore limited. Beyond a certain threshold, increasing the dose does not affect posaconazole exposure preventing additional therapeutic benefit and worsened adverse effects. For the indication of oropharyngeal candidiasis (OPC) treatment, study C/I96-209 was used to elucidate the dose while analyzing the pharmacologic profile [16,51]. This trial involved 463 HIV-infected patients with OPC. The goal of this trial was to compare treatment using various doses of posaconazole with an established dose of fluconazole, which is routinely used for this indication. Patients were randomized to receive either of the agents with posaconazole at the doses of interest and fluconazole at 200 mg once followed by 100 mg daily. Treatment was continued for a total of 14 days. Posaconazole was well tolerated and showed similar efficacy at the dose levels and compared with fluconazole. On the basis of these data, posaconazole 100 mg once daily dose was selected. Phase III clinical studies IFI prophylaxis The approval of posaconazole by the FDA and the EMA was based on several pivotal Phase III trials as shown in Table 2. Ullmann et al. compared the efficacy of posaconazole versus fluconazole in preventing IFIs in a multicenter, randomized, double-blind trial [39]. Six hundred patients with HSCT and GVHD or patients who were being treated with highly immunosuppressive agents were randomized to receive either agent for 16 weeks. One cohort received posaconazole as a 200-mg oral suspension three-times daily with placebo capsules once daily, whereas the second cohort received fluconazole as a 400-mg encapsulated tablet once daily with placebo oral suspension three-times daily. Posaconazole was shown to be noninferior to fluconazole in the prevention of IFIs (5.3 vs 9.0%) and was shown to be superior in the treatment of invasive aspergillosis specifically (2.3 vs 7.0%). Between both groups, adverse events were similar (36 vs 38%, respectively) as was the rate of discontinuation due to adverse events (34 vs 38%, respectively). Cornely conducted a study comparing the efficacy of posaconazole with fluconazole and itraconazole for IFI prophylaxis in patients treated for cancer who were projected to experience neutropenia [52]. Six hundred and two patients were randomized and received 200-mg posaconazole three-times daily as an oral suspension or one of the alternate azoles for up to 12 weeks during their rounds of chemotherapy; the choice of the alternate azole (400-mg fluconazole once daily or itraconazole 200 mg twice daily) was made by each investigator. Patients were monitored for fungal infections by a blinded independent data review committee. Posaconazole was shown to be superior in preventing IFIs compared with the alternate azoles (2 vs 8%, respectively). In addition, the mean time to IFI was longer with posaconazole (41 vs 25 days, respectively). Finally, the posaconazole group experienced lower mortality during the treatment period (16 vs 22%, respectively). Adverse effects were comparable for patients taking each agent. OPC treatment The efficacy of posaconazole versus fluconazole for patients with HIV/AIDS in the treatment of OPC has been evaluated [53]. Three hundred and fifty patients were randomized and received 200-mg posaconazole or fluconazole oral suspension on the first day as the loading dose, followed by 100 mg every day of the same drug for a total of 14 days. Blinded evaluators investigated each patient for clinical success, which was specified as cure or improvement of OPC. At the end of the trial, clinical success was seen in both groups at similar rates (91.7% for posaconazole and 92.5% for fluconazole). Of the patients considered clinically successful 42 days after trial completion, 31.5% of the posaconazole patients relapsed compared with 38.2% of the fluconazole patients. Also at the 42-day follow-up, mycological eradication was greater in patients of the posaconazole arm (35.6 vs 24.2%, respectively). Adverse events were comparable in each group. To investigate the role of posaconazole in azole-refractory OPC, Skriest analyzed patients infected with HIV with oropharyngeal or esophageal candidiasis resistant to standard azole treatments of fluconazole and itraconazole [38]. Patients received 400 mg twice daily for 3 days followed by 400 mg daily for 25 days or 400 mg twice daily for 28 days. Seventy five percent of the 176 patients experienced clinical success, defined as cure or improvement. Treatment success was relatively invariable between the groups taking different regimens. Four weeks after the last dose, 74% of 80 patients who had experienced a clinical response had relapsed (80% of patients on the daily dosing regimen, 68% of patients on the twice daily dosing regimen). Treatment of other approved fungal infections Additional studies were conducted to support the approval of posaconazole for the treatment of a broader range of fungal infections in Europe [17]. Three hundred and thirty patients with varying fungal infections, which were resistant to or intolerant of standard therapy, including other azoles, echinocandins and amphotericin, were administered posaconazole 200 mg four-times daily while hospitalized and 400 mg twice daily on an outpatient basis for up to 1 year. Overall 50% of patients responded to posaconazole. The investigators analyzed clinical success rates for each fungal infection individually, with the following infections showing the best efficacy with posaconazole: aspergillosis (42%), fusariosis (39%), chromoblastomycosis or mycetoma (82%), coccidioidomycosis (69%). Catanzaro et al. investigated the role of posaconazole in treating patients with coccidioidomycosis [54]. In 20 treatment-naïve patients with nonmeningeal disseminated or chronic pulmonary coccidioidomycosis, the study investigators administered 400 mg of posaconazole daily. Eighty five percent of the patients experienced a response, defined as a 50% or greater reduction in the mycoses study group score from baseline. Stevens et al. further analyzed the use of posaconazole in the treatment of treatment-resistant coccidioidomycosis [55]. Fifteen patients with refractory coccidioidomycosis were recruited for this trial. Patients were administered 800 mg of posaconazole in divided doses daily for up to a year. At the end of the treatment period, 73% of the patients responded to treatment: four showed a complete eradication of disease and seven showed a partial resolution. Posaconazole was well tolerated in both trials. Off-label indications Posaconazole has also been studied for infections that are rarer, often as salvage therapy in smaller, less controlled trials. While clinical experience with posaconazole is limited with these indications, there is evidence to support utilization. van Burik et al. retrospectively analyzed the use of posaconazole in mucormycosis [56]. Information from 91 patients was included in this study. Patients were treated with posaconazole, 80% for at least 30 days. Sixty percent of patients responded to treatment. Restrepo et al. investigated posaconazole's utility as salvage treatment of histoplasmosis [57]. Six patients who had failed other treatments were placed on posaconazole 800 mg per day in divided doses. All patients responded to treatment within a month as demonstrated by clinical improvement. Posaconazole has also demonstrated good activity in the treatment of fungal infection of the CNS. Pitisuttithum et al. studied 39 patients with CNS fungal infections [58]. Most of these patients had refractory disease (95%) or an HIV infection (74%). The infections were caused by a variety of fungi including Cryptococcus, Aspergillus and a variety of rarer fungi (Pseudallescheria boydii, C. immitis, Histoplasma capsulatum, Ramichloridium mackenziei, Apophysomyces elegans and Basidiomycetes sp.). Patients were administered the posaconazole oral suspension at 800 mg per day in divided doses for at least a month and up to 1 year. Clinical responses were seen in 48% of the patients infected with Cryptococcus and in 50% of the patients with the other fungal infections. This trial provides evidence that posaconazole can be useful as salvage therapy for a wider variety of fungal infections and in the CNS. Postmarketing surveillance The FDA required pediatric studies to be completed secondary to the approval of posaconazole. Completing one study that addressed the use of posaconazole in this population, Döring et al. compared itraconazole, voriconazole and posaconazole in pediatric patients with allogenic HSCT as an oral antifungal prophylactic agent [59]. One hundred and fifty patients between the age of 7 months and 18 years were divided evenly into three cohorts to receive 5 mg/kg twice daily itraconazole, 100 mg twice daily (or 200 mg twice daily if body weight >40 kg) voriconazole or 200 mg thrice daily posaconazole. Patients also received antiviral and antibacterial prophylaxis for other possible infections. At the end of the observation period (maximum of 220 days after HSCT), there were no deaths due to IFIs or even any cases of ‘proven’ or ‘probable’ IFIs in the patients. There were several ‘possible’ infections: two in the itraconazole group, three in the voriconazole group and none in the posaconazole group. The differences between ‘proven’, ‘probable’, and ‘possible’ were analyzed secondary to the National Institute of Allergy and Infectious Diseases Mycoses Study Group criteria. Ultimately, the three agents were considered noninferior to each other in terms of efficacy. Comparing the safety profiles, the adverse effect ratio was comparable between all three groups, with increased liver enzymes representing one of the most common events. Regulatory affairs Posaconazole has been approved to treat fungal infections in the USA and EU with the trade name Noxafil®. Posaconazole is marketed as Posanol in Canada [16]. Labeled indications in the USA and Europe are listed in Text Boxes 1 & 2. The FDA approved posaconazole for prophylaxis of the Aspergillus and Candida IFIs in immunocompromised patients of at least 13 years of age in September 2006, and it was approved for the treatment of OPC, including those cases refractory to other azole antifungals, in October 2006 [60,61]. For both indications, posaconazole was approved as a 40 mg/ml oral suspension. Posaconazole was subsequently approved as a delayed-release tablet in addition to an intravenous solution for the indication of invasive Aspergillus and Candida infection prophylaxis in November 2013 and March 2014, respectively [62,63]. In the USA, posaconazole's patent is scheduled to expire in 2019. The EMA has approved posaconazole as an oral suspension and as a gastro-resistant tablet for treatment of a wider variety of IFIs (aspergillosis, fusariosis, chromoblastomycosis and mycetoma, coccidioidomycosis, and OPC) including those cases refractory to standard antifungal therapy. In addition, posaconazole is approved for the prevention of IFIs in patients who receive immunosuppressive therapy after receiving HSCT or chemotherapy for AML or any other myelodysplastic syndrome [64]. For all of these indications, posaconazole was approved as an oral suspension and as a gastro-resistant tablet in October 2005. The FDA specified a number of post-marketing commitments from posaconazole. For the IFI indication, the FDA required a pediatric study in patients aged 0 months to 12 years based on the Pediatric Research Equity Act. For the OPC indication, a pediatric study is also required but in patients aged 0 months to 16 years. In addition, the FDA requested a study among patients receiving IFI prophylaxis at risk for low absorption to explore alternate dosing strategies and the utility of therapeutic drug monitoring (TDM) [60,61]. These studies were originally required by 2011. Since then, the FDA has granted a deferral extension, and the studies will need to be completed by 2019 with the exception of the study involving alternate dosing strategies and TDM. The FDA has released the company from the obligation to complete that study [65]. For the new formulations, pediatric studies are also required and will need to be completed within the next 10 years (Figure 3) [62,63]. Conclusion Posaconazole is a useful antifungal agent. This drug has shown efficacy in the role of IFI prophylaxis, the treatment of OPC and salvage treatment for refractory fungal infections. Overall, posaconazole has a favorable safety and tolerability profile. However, the initially approved oral suspension formulation has a wide interpatient variability with respect to absorption. The systemic availability of the oral suspension can be improved through administration with a high-fat meal, nutritional supplement or acidic beverage in addition to divided daily doses. The suspension has shown a reduction in systemic availability when administered with proton pump inhibitors in addition to pharmacotherapeutics that increase gastric motility. Both diarrhea and mucositis have also been shown to reduce the bioavailability of posaconazole. The administration of multiple daily doses with a full meal or nutritional supplement can be difficult for patients who have difficulties in swallowing. As such, the recent approval of a delayed-release tablet that requires administration only once daily and does not appear to be clinically affected by food, alterations in gastric pH or motility represents a new option for the prophylaxis of IFI. As it achieves consistently higher average serum concentrations, the delayed-release tablet is a better option than suspension for both prophylaxis and treatment of IFI. An intravenous formulation has been recently developed and approved, which provides even greater access to patient populations. Five-year view Unlike in the USA, posaconazole has not been approved in the EU as a delayed-release tablet or as a solution for intravenous administration. However, Merck has received ‘positive opinions’ from the Committee for Medicinal Products for Human Use inside the EMA [66,67]. This signifies that European Commission will review the new drug formulations further and likely approve it in the near future, following the precedent set by the FDA. Posaconazole has been increasingly studied for other fungal infections beyond those infections for which it was originally approved by the major regulatory bodies. Elewski et al. investigated posaconazole in the treatment of toenail onychomycosis and showed improved cure rates compared with the commonly used antifungal agent terbinafine [68]. Posaconazole has also shown some potential for use in Chagas disease and chronic granulomatous disease [69,70], which is a significant finding as both of these conditions currently have limited therapeutic options. With further investigation, posaconazole may be readily used for these indications in the future. Furthermore, the effect of posaconazole in pediatric patients will become clearer as the Pediatric Research Equity Act required studies are completed in the next few years. Favorable outcomes are expected given the positive safety data of the posaconazole oral suspension in pediatric patients. Thus, the use of the posaconazole formulations will likely increase over time in this subset of patients. TDM has been increasingly suggested for posaconazole oral suspension therapy, as interpatient concentration variability is high as is the risk for underexposure. Prophylactic target posaconazole concentrations range from at least 0.5 to 0.7 mg/l. As such, a general consensus has yet to be determined. A recent report from Hoenigl et al. suggests the need for TDM because personal on-site education of low posaconazole plasma concentrations can lead to >40% increase in sufficient plasma concentrations [71]. These findings are consistent with a previously reports that found that cases of breakthrough infections were all associated with low posaconazole plasma concentrations [19,21,72]. Bryant et al. (2011) was one of the first studies to report on TDM of prophylactic posaconazole in patients with AML or myelodysplastic syndrome [73]. In this study, greater than 7% of patients did not reach the lower target limit of 0.7 mg/l. Of the 21 patients, three developed ‘proven’ or ‘possible’ fungal infections; all three patients having posaconazole concentrations <0.5 mg/l. Because posaconazole is liable to interact with many other pharmaceutical agents, several sources recommend TDM for posaconazole during the duration of multimodal therapy. This may be particularly useful in cases where posaconazole is used with an agent with whom it has a drug–drug interaction, which is common, as many agents used for cancer, HIV and transplants are CYP3A4 substrates, for example, tacrolimus and cyclosporine. With time, a consensus will be reached concerning TDM, the appropriate concentration goals for effective posaconazole treatment. Also on the horizon, isavuconazonium is a novel triazole agent, which has recently been approved by the FDA for the treatment of aspergillosis and mucormycosis [74]. Preliminary data from Phase III studies suggest that it may have similar efficacy as voriconazole with fewer adverse effects in treating IFIs [75,76]. Isavuconazonium had a relatively quick regulatory timeline because the prognosis of mucormycosis is so poor at the moment that the FDA labeled isavuconazonium as an orphan drug and as a Qualified Infectious Disease Product for expedited review [77]. Following the approval of isavuconazonium, it will likely influence both the European and the US markets. It will present a good option for mucormycosis and may, in fact, be used instead of posaconazole for some patients suffering from invasive aspergillosis. Expert commentary Posaconazole has been recognized by the guidelines for quite a few therapeutic uses beyond its FDA approved indications. The Infectious Disease Society of America has included them in their guidelines for aspergillosis, cryptococcosis, candidiasis, skin and soft tissue infections, and antimicrobial prophylaxis in neutropenia [3–5,78,79]. The Infectious Disease Society of America guidelines on skin and soft tissue infections specifically recommend posaconazole for the treatment of aspergillus, mucormycosis and fusariosis. Posaconazole is also recommended for the general prophylaxis and treatment of IFIs in the respective guidelines for patients affected by HIV, hematopoietic cell transplantation and cancer (Figure 4) [80–82]. Numerous favorable pharmacoeconomic analyses have been published for posaconazole [83–86]. In these studies, the usage of posaconazole was primarily studied for the prevention of IFIs secondary to HSCT or secondary to febrile neutropenia in AML or myelodysplastic syndrome. Lyseng-Williamson (2011) specifically suggested that posaconazole was the dominant option compared with standard antifungal therapy in many markets if not simply a cost-effective choice [86]. Compared with the other azole antifungals, posaconazole occupies a unique place in therapy. It can be used in fungal infections resistant to fluconazole and itraconazole. Posaconazole can also be used for prophylaxis of a wider variety of fungal infections than voriconazole. As one of the newest agents in a long-standing class of pharmaceuticals, it will be some time before prescribing healthcare providers accumulate enough clinical experience with this agent and become accustomed to it. However, the addition of the delayed-release tablet and the intravenous formulation, as options will certainly accelerate this process as it allows for more convenient dosing. J Moore and J Healy were supported by National Institutes of Health Postdoctoral training grant no. T32GM008562. W Kraft's employer has received research grant support from Merck for clinical trials in which W Kraft was the Principal Investigator. Figure 1 Chemical structures of common triazole anti-fungal drugs and isavuconazonium. Figure 2 Key patient populations with posaconazole drug interactions. Adapted from references [7,16,19,22,42]. Figure 3 Regulatory history of posaconazole Adapted from references [49,51,62–65]. DRT: Delayed release tablet; EMA: European medicines agency; FDA: Food and drug administration; IFI: Invasive fungal infections; IV: Intravenous; OPC: Oropharyngeal candidiasis; PREA: Pediatric Research Equity Act. Figure 4 Posaconazole therapy algorithm Unless otherwise noted, listed dosages are for the posaconazole oral suspension. All other indications shown are off-label uses. Adapted from references [5,16,78,79]. †Denotes FDA approval for the posaconazole oral suspension. ‡Denotes FDA approval for all available posaconazole formulations. b.i.d.: Twice daily; DRT: Delayed-release tablet; iv.: Intravenous solution; q.d.: Once daily; q.i.d.: Four-times a day; SUS: Suspension; t.i.d.: Thrice daily. Table 1 Comparison of triazole adverse events. Fluconazole Itraconazole Voriconazole Posaconazole Gastrointestinal distress + ++ − + Liver dysfunction + ++ ++ − Dermatological toxicity + + + − Ocular toxicity − − ++ − − 0-5% +5-10% ++10-20% Adapted from Reference [8]. Table 2 Pivotal Phase III Trials. Indication Population Primary Endpoint Dose/duration Efficacy Adverse events Ref. IFI prophylaxis 600 patients with HSCT and GVHD Incidence of proven or probable IFI • POS 200 mg TID + placebo capsules OR • FLU 400 mg QD + placebo suspension (Incidence of IFI) POS: 5.3% FLU: 9.0% POS: 36% FLU: 38% [39] IFI prophylaxis 602 patients undergoing chemotherapy for AML or MDS Incidence of proven or probable IFI • POS 200 mg TID OR • FLU 400 mg QD or ITR 200 mg BID (Incidence of IFI) POS: 2% FLU/ITR: 8% POS: 52% FLU/ITR: 59% [52] IFI prophylaxis 213 patients with AML, MDS, or HSCT PK of POS IV POS 300 mg IV BID × 1, 300 mg QD × 4-13 days, then POS 600-800 mg in divided doses QD for up to 23 days (Clinical failure: death, IFI, Discontinuation) POS: 31% POS: 99%† [33,34] OPC treatment 350 patients with HIV infection and OPC Clinical success (cure or improvement after 28 days) • POS 200 mg × 1, 100 mg × 13 days, OR • FLU 200 mg ×1, 100 mg × 13 days (Clinical success) POS: 91.7% FLU: 92.5% POS: 64% FLU: 68% [53] OPC treatment 199 patients with HIV and azole-refractory OPC or EC Clinical success (cure or improvement after 28 days) • POS 400 mg BID × 3 days, POS 400 mg QD × 25 days OR • POS 400 mg BID × 28 days (Clinical success) Group 1: 75.3% Group 2: 74.7% Combined: 49% [38] IFI treatment 330 patients with IFIs resistant to standard therapy Global response status (clinical cure or improvement) POS 200 mg QID while hospitalized, then 400 mg BID up to 1 year (Global response status) 50% POS: 38%† [17] All studies involved the posaconazole suspension unless otherwise noted. † Adverse event calculation was pooled from this study and additional one(s). AML: Acute myeloid leukemia; BID: Twice daily; EC: Esophageal candidiasis; FLU: Fluconazole; GVHD: Graft vs host disease; HIV: Human immunodeficiency virus; HSCT: Hematopoietic stem cell graft; IFI: Invasive fungal infection; ITR: Itraconazole; IV: Intravenous; MDS: Myelodysplastic syndromes; OPC: Oropharyngeal candidiasis; POS: Posaconazole; QD: Once daily; QID: Four times a day; TID: Thrice daily. Box 1. Labeled FDA Indications for use Injection, delayed-release tablets and oral suspension Prophylaxis of invasive Aspergillus and Candida infections in patients who are at high risk of developing these infections due to being severely immunocompromised, such as HSCT recipients with graft versus host disease or those with hematologic malignancies with prolonged neutropenia from chemotherapy Oral suspension Treatment of oropharyngeal candidiasis, including oropharyngeal candidiasis refractory to itraconazole and/or fluconazole Box 2. Labeled EMA indications for use Fungal infections in adults Invasive aspergillosis in patients with disease that is refractory to amphotericin B or itraconazole or in patients who are intolerant of these medicinal products Fusariosis in patients with disease that is refractory to amphotericin B or in patients who are intolerant of amphotericin B Chromoblastomycosis and mycetoma in patients with disease that is refractory to itraconazole or in patients who are intolerant of itraconazole Coccidioidomycosis in patients with disease that is refractory to amphotericin B, itraconazole or fluconazole or in patients who are intolerant of these medicinal products OPC: as first-line therapy in patients who have severe disease or are immunocompromised, in whom response to topical therapy is expected to be poor Refractoriness is defined as progression of infection or failure to improve after a minimum of 7 days of prior therapeutic doses of effective antifungal therapy Prophylaxis of IFIs in the following patients: Patients receiving remission-induction chemotherapy for acute myelogenous leukemia or myelodysplastic syndromes expected to result in prolonged neutropenia and who are at high risk of developing IFIs Hematopoietic-stem-cell-transplant recipients who are undergoing high-dose immunosuppressive therapy for graft-versus-host disease and who are at high risk of developing IFIs Key issues Invasive fungal infections (IFIs) are a significant cause of morbidity and mortality for immunocompromised patients; unfortunately, the clinical usefulness of many antifungal agents is limited by safety and tolerability. Posaconazole is a relatively safe and well-tolerated broad-spectrum azole antifungal agent. Newer posaconazole formulations, oral tablet and intravenous, increase its availability to a larger patient population. Posaconazole is a cost–effective and dominant option compared with standard antifungal therapy in many markets. Posaconazole continues to be studied for additional fungal infections beyond original approval, including toenail onychomycosis, Chagas disease and chronic granulomatous disease. Financial & competing interests disclosure The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. References Papers of special note have been highlighted as: • of interest •• of considerable interest 1 Garey KW Rege M Pai MP Time to initiation of fluconazole therapy impacts mortality in patients with candidemia: a multi-institutional study. Clin Infect Dis 2006 43 1 25 31 16758414 2 Morrell M Fraser VJ Kollef MH Delaying the empiric treatment of candida bloodstream infection until positive blood culture results are obtained: a potential risk factor for hospital mortality. Antimicrob Agents Chemother 2005 49 9 3640 5 16127033 3 Freifeld AG Bow EJ Sepkowitz KA Clinical practice guideline for the use of antimicrobial agents in neutropenic patients with cancer: 2010 update by the infectious diseases society of america. Clin Infect Dis 2011 52 4 e56 93 21258094 4 Pappas PG Kauffman CA Andes D Clinical practice guidelines for the management of candidiasis: 2009 update by the infectious diseases society of America. Clin Infect Dis 2009 48 5 503 35 19191635 5 Walsh TJ Anaissie EJ Denning DW Treatment of aspergillosis: clinical practice guidelines of the infectious diseases society of America. Clin Infect Dis 2008 46 3 327 60 18177225 6 Kim JH Williams K Posaconazole salvage treatment for invasive fungal infection. Mycopathologia 2014 178 3-4 259 65 25103141 7 Li Y Theuretzbacher U Clancy CJ Pharmacokinetic/pharmacodynamic profile of posaconazole. Clin Pharmacokinet 2010 49 6 379 96 20481649 8•• Chau MM Kong DC van Hal SJ Consensus guidelines for optimising antifungal drug delivery and monitoring to avoid toxicity and improve outcomes in patients with haematological malignancy 2014. Int Med J 2014 44 12b 1364 88 Clinical guidelines for optimized antifungal regimens with safety and toxicity profiles. 9 Marty FM Cosimi LA Baden LR Breakthrough zygomycosis after voriconazole treatment in recipients of hematopoietic stem-cell transplants. N Eng J Med 2004 350 9 950 2 10 Rogers TR Treatment of zygomycosis: current and new options. J Antimicrob Chemother 2008 61 Suppl 1 i35 40 18063603 11 Torres HA Hachem RY Chemaly RF Posaconazole: a broad-spectrum triazole antifungal. Lancet Infect Dis 2005 5 12 775 85 16310149 12 Kim H Kumari P Laughlin M Use of high-performance liquid chromatographic and microbiological analyses for evaluating the presence or absence of active metabolites of the antifungal posaconazole in human plasma. J Chromat A 2003 987 1-2 243 8 13 Nagappan V Deresinski S Reviews of anti-infective agents: posaconazole: a broad-spectrum triazole antifungal agent. Clin Infect Dis 2007 45 12 1610 17 18190324 14 Gubbins PO Krishna G Sansone-Parsons A Pharmacokinetics and safety of oral posaconazole in neutropenic stem cell transplant recipients. Antimicrob Agents Chemother 2006 50 6 1993 9 16723557 15 Krieter P Flannery B Musick T Disposition of posaconazole following single-dose oral administration in healthy subjects. Antimicrob Agents Chemother 2004 48 9 3543 51 15328123 16 Product monograph – Noxafil (Posaconazole) 2014 Merck & Co. Inc. Whitehouse Station, NJ Available from: www.merck.com/product/usa/pi_circulars/n/noxafil/noxafil_pi.pdf 12 February 2015 17 Noxafil:EPAR-Scientific Discussion 2005 Agency EM. European Medicines Agency London, UK Available from: www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_Scientific_Discussion/human/000610/WC500037781.pdf 10 October 2014 18 Mavridou E Bruggemann RJ Melchers WJ Efficacy of posaconazole against three clinical Aspergillus fumigatus isolates with mutations in the cyp51A gene. Antimicrob Agents Chemother 2010 54 2 860 5 19917751 19 Dolton MJ Ray JE Chen SC Multicenter study of posaconazole therapeutic drug monitoring: exposure-response relationship and factors affecting concentration. Antimicrob Agents Chemother 2012 56 11 5503 10 22890761 20 Shields RK Clancy CJ Vadnerkar A Posaconazole serum concentrations among cardiothoracic transplant recipients: factors impacting trough levels and correlation with clinical response to therapy. Antimicrob Agents Chemother 2011 55 3 1308 11 21189337 21 Jang SH Colangelo PM Gobburu JV Exposure-response of posaconazole used for prophylaxis against invasive fungal infections: evaluating the need to adjust doses based on drug concentrations in plasma. Clin Pharmacol Ther 2010 88 1 115 19 20505665 22 Lipp HP Clinical pharmacodynamics and pharmacokinetics of the antifungal extended-spectrum triazole posaconazole: an overview. Br J Clin Pharma 2010 70 4 471 80 23 Ezzet F Wexler D Courtney R Oral bioavailability of posaconazole in fasted healthy subjects: comparison between three regimens and basis for clinical dosage recommendations. Clin Pharmacokinet 2005 44 2 211 20 15656699 24 Krishna G Moton A Ma L Pharmacokinetics and absorption of posaconazole oral suspension under various gastric conditions in healthy volunteers. Antimicrob Agents Chemother 2009 53 3 958 66 19075045 25 Krishna G Ma L Vickery D Effect of varying amounts of a liquid nutritional supplement on the pharmacokinetics of posaconazole in healthy volunteers. Antimicrob Agents Chemother 2009 53 11 4749 52 19738015 26 Walravens J Brouwers J Spriet I Effect of pH and comedication on gastrointestinal absorption of posaconazole: monitoring of intraluminal and plasma drug concentrations. Clin Pharmacokinet 2011 50 11 725 34 21973269 27•• Krishna G Ma L Martinho M A new solid oral tablet formulation of posaconazole: a randomized clinical trial to investigate rising single- and multiple-dose pharmacokinetics and safety in healthy volunteers. J Antimicrob Chemother 2012 67 11 2725 30 Clinical studies examining the pharmacokinetics and safety of the posaconazole oral tablet formulation. 22833639 28 Krishna G Ma L Martinho M Single-dose Phase I study to evaluate the pharmacokinetics of posaconazole in new tablet and capsule formulations relative to oral suspension. Antimicrob Agents Chemother 2012 56 8 4196 201 22615291 29 Duarte RF Lopez-Jimenez J Cornely OA Phase Ib study of new posaconazole tablet for prevention of invasive fungal infections in high-risk patients with neutropenia. Antimicrob Agents Chemother 2014 58 10 5758 65 25049247 30 Cornely OA Duarte RF Haider S Phase III pharmacokinetics (PK) and safety study of posaconazole (POS) tablet in patients at risk for invasive fungal infection. 23rd European congress of clinical microbiology and infectious diseases Berlin, Germany 2013 31•• Maertens J Cornely OA Ullmann AJ Phase IB study of the pharmacokinetics and safety of posaconazole intravenous solution in patients at risk for invasive fungal disease. Antimicrob Agents Chemother 2014 58 7 3610 17 Clinical studies examining the pharmacokinetics and safety of the posaconazole intravenous formulation. 24733463 32 Kersemaekers WM van Iersel T Nassander U Pharmacokinetics and safety study of posaconazole intravenous solution administered peripherally to healthy subjects. Antimicrob Agents Chemother 2015 59 2 1246 51 25512407 33 NDA 205-596 - Posaconazole Injection, Noxafil 2014 FDA Center for Drug Evaluation and Research. FDA/Center for Drug Evaluation and Research Washington, DC Available from: www.accessdata.fda.gov/drugsatfda_docs/nda/2014/205596Orig1s000MedR.pdf 15 December 2014 34 Cornely OA Haider S Grigg A Phase III Pharmacokinetics (PK) and Safety Study of Posaconazole (POS) IV in Patients (Pts) at Risk for Invasive Fungal Infection (IFI). Interscience conference of antimicrobial agents chemother Denver, CO 2013 35 Chandrasekar P Prophylaxis against Aspergillus is not perfect: problems and perils in stem cell transplantation. Med Mycol 2009 47 Suppl 1 S349 54 18663660 36 Jacinto PL Chandrasekar P Safety of posaconazole. Expert Opin Drug Saf 2013 12 2 265 74 23360443 37 Courtney R Pai S Laughlin M Pharmacokinetics, safety, and tolerability of oral posaconazole administered in single and multiple doses in healthy adults. Antimicrob Agents Chemother 2003 47 9 2788 95 12936975 38 Skiest DJ Vazquez JA Anstead GM Posaconazole for the treatment of azole-refractory oropharyngeal and esophageal candidiasis in subjects with HIV infection. Clin Infect Dis 2007 44 4 607 14 17243069 39 Ullmann AJ Lipton JH Vesole DH Posaconazole or fluconazole for prophylaxis in severe graft-versus-host disease. N Engl J Med 2007 356 4 335 47 17251530 40 Moton A Krishna G Wang Z Tolerability and safety profile of posaconazole: evaluation of 18 controlled studies in healthy volunteers. J Clin Pharma Ther 2009 34 3 301 11 41 Pettit NN Han Z delaCruz J Posaconazole (PCZ) Tablet Formulation Therapeutic Drug Monitoring (TDM) and Toxicity Analysis. Interscience conference on antimicrob agents chemother Washington, DC 2014 42 Neofytos D Avdic E Magiorakos AP Clinical safety and tolerability issues in use of triazole derivatives in management of fungal infections. Drug Healthc Patient Saf 2010 2 27 38 21701616 43 Cojutti P Candoni A Simeone E Antifungal prophylaxis with posaconazole in patients with acute myeloid leukemia: dose intensification coupled with avoidance of proton pump inhibitors is beneficial in shortening time to effective concentrations. Antimicrob Agents Chemother 2013 57 12 6081 4 24060872 44 Kraft WK Chang PS van Iersel ML Posaconazole tablet pharmacokinetics: lack of effect of concomitant medications altering gastric pH and gastric motility in healthy subjects. Antimicrob Agents Chemother 2014 58 7 4020 5 24798274 45 Jung DS Tverdek FP Kontoyiannis DP Switching from posaconazole suspension to tablets increases serum drug levels in leukemia patients without clinically relevant hepatotoxicity. Antimicrob Agents Chemother 2014 58 11 6993 5 25199774 46 Hope WW Castagnola E Groll AH ESCMID* guideline for the diagnosis and management of Candida diseases 2012: prevention and management of invasive infections in neonates and children caused by Candida spp. Clin Microbiol Infect 2012 18 Suppl 7 38 52 23137136 47 Sansone-Parsons A Krishna G Martinho M Effect of oral posaconazole on the pharmacokinetics of cyclosporine and tacrolimus. Pharmacotherapy 2007 27 6 825 34 17542765 48 Dahlan R Patel A Haider S Successful use of posaconazole to treat invasive cutaneous fungal infection in a liver transplant patient on sirolimus. Can J Infect Dis Med Microbiol 2012 23 2 e44 7 23730320 49 NDA 22-003 – Prophylaxis of invasive aspergillus and candida infections medical review 2006 FDA Center for Drug Evaluation and Research Washington, DC Available from: www.accessdata.fda.gov/drugsatfda_docs/nda/2006/022003s000_Noxafil_MedR.pdf 8 October 2014 50 PK/PD Study of posaconazole for empiric treatment of invasive fungal infections in neutropenic patients or treatment of refractory invasive fungal infections (Study P01893) (COMPLETED) 2013 ClinicalTrials.gov Washington, DC Available from: http://clinicaltrials.gov/show/NCT00034671 8 October 2014 51 NDA 22-027 – Invasive fungal infection and refractory invasive fungal infection medical review 2006 FDA center for drug evaluation and research Washington, DC Available from: www.accessdata.fda.gov/drugsatfda_docs/nda/2006/022027_noxafil_toc.cfm 8 October 2014 52 Cornely OA Maertens J Winston DJ Posaconazole vs. fluconazole or itraconazole prophylaxis in patients with neutropenia. N Eng J Med 2007 356 4 348 59 53 Vazquez JA Posaconazole for the management of mucosal candidiasis. Future Microbiol 2007 2 3 245 56 17661697 54 Catanzaro A Cloud GA Stevens DA Safety, tolerance, and efficacy of posaconazole therapy in patients with nonmeningeal disseminated or chronic pulmonary coccidioidomycosis. Clin Infect Dis 2007 45 5 562 8 17682989 55 Stevens DA Rendon A Gaona-Flores V Posaconazole therapy for chronic refractory coccidioidomycosis. Chest 2007 132 3 952 8 17573510 56 van Burik JA Hare RS Solomon HF Posaconazole is effective as salvage therapy in zygomycosis: a retrospective summary of 91 cases. Clin Infect Dis 2006 42 7 e61 5 16511748 57 Restrepo A Tobon A Clark B Salvage treatment of histoplasmosis with posaconazole. J Inf 2007 54 4 319 27 58 Pitisuttithum P Negroni R Graybill JR Activity of posaconazole in the treatment of central nervous system fungal infections. J Antimicrob Chemother 2005 56 4 745 55 16135526 59 Doring M Blume O Haufe S Comparison of itraconazole, voriconazole, and posaconazole as oral antifungal prophylaxis in pediatric patients following allogeneic hematopoietic stem cell transplantation. Eur J Clin Microbiol Infect Dis 2014 33 4 629 38 24173819 60 NDA 22-003 Approval 2006 FDA Center for Drug Evaluation and Research Washington, DC Available from: www.accessdata.fda.gov/drugsatfda_docs/appletter/2006/022003ltr.pdf 15 October 2014 61 NDA 22-027 Approval 2006 FDA Center for Drug Evaluation and Research Washington, DC Available from: www.accessdata.fda.gov/drugsatfda_docs/appletter/2006/022027s000ltr.pdf 15 October 2014 62 NDA 205053 Approval 2013 FDA center for drug evaluation and research Washington, DC Available from: www.accessdata.fda.gov/drugsatfda_docs/appletter/2013/205053Orig1s000ltr.pdf 15 October 2014 63 NDA 205596 Approval 2014 FDA Center for Drug Evaluation and Research Washington, DC Available from: www.accessdata.fda.gov/drugsatfda_docs/appletter/2014/205596Orig1s000ltr.pdf 15 October 2014 64 Summary of Product Characteristics 2010 European Medicines Agency London, UK Available from: www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_Product_Information/human/000610/WC500037784.pdf 15 October 2014 65 Postmarket Requirements and Commitments 2014 FDA Center for Drug Evaluation and Research Washington, DC Avaialbe from www.accessdata.fda.gov/scripts/cder/pmc/index.cfm 15 October 2014 66 CHMP issues positive opinion for tablet formulation of merck's NOXAFIL (posaconazole) 2014 Merck Newsroom Whitehouse Station Available from: www.mercknewsroom.com/news-release/corporate-news/chmp-issues-positive-opinion-tablet-formulation-mercks-noxafilposaconaz 23 October 2014 67 CHMP issues positive opinion for intravenous (IV) Formulation of Merck's NOXAFIL® (posaconazole) 2014 Merck Newsroom Whitehouse Station Available from: www.mercknewsroom.com/news-release/research-and-development-news/chmp-issues-positive-opinion-intravenous-iv-formulation-m 23 October 2014 68 Elewski B Pollak R Ashton S A randomized, placebo- and active-controlled, parallel-group, multicentre, investigator-blinded study of four treatment regimens of posaconazole in adults with toenail onychomycosis. Br J Dermatol 2012 166 2 389 98 21967490 69• Molina I Gomez i Prat J Salvador F Randomized trial of posaconazole and benznidazole for chronic Chagas’ disease. N Engl J Med 2014 370 20 1899 908 Examines potential benefit of posaconazole administration for Chagas disease. 24827034 70 Welzen ME Bruggemann RJ Van Den Berg JM A twice daily posaconazole dosing algorithm for children with chronic granulomatous disease. Pediatr Infect Dis J 2011 30 9 794 7 21772229 71 Hoenigl M Duettmann W Raggam RB Impact of structured personal on-site patient education on low posaconazole plasma concentrations in patients with haematological malignancies. Int J Antimicrob Agents 2014 44 2 140 4 25059446 72 Hoenigl M Raggam RB Salzer HJ Posaconazole plasma concentrations and invasive mould infections in patients with haematological malignancies. Int J Antimicrob Agents 2012 39 6 510 13 22481057 73 Bryant AM Slain D Cumpston A A post-marketing evaluation of posaconazole plasma concentrations in neutropenic patients with haematological malignancy receiving posaconazole prophylaxis. Int J Antimicrob Agents 2011 37 3 266 9 21236645 74 NDA 207501 Approval 2015 FDA Center for Drug Evaluation and Research Washington, DC Available from: www.accessdata.fda.gov/drugsatfda_docs/appletter/2015/207501Orig1s000ltr.pdf 17 March 2015 75 New drug against invasive mold disease as effective as existing drugs with fewer adverse effects. Interscience Conference on Antimicrobial Agents and Chemotherapy Washington, DC 2014 Available from: www.icaac.org/index.php/newsroom/92-icaac-2014/newsroom/310-new-drug-against-invasive-mold-disease-as-effective-as-existing-drugs-with-fewer-adverse-events 13 October 2014 76 Patterson T Selleslag D Mullane K A Phase III, randomized, double-blind, non-inferiority trial to evaluate efficacy and safety of isavuconazole versus voriconazole in patients with invasive mold disease (SECURE): Outcomes in Neutropenic Patients IDWeek 2014 Abstract 1210 77• Isavuconazole. Basilea Pharmaceutica 2014 Basilea Pharmaceutica Ltd Switzerland Available from: www.basilea.com/Portfolio/Isavuconazole/ 15 March 2015 Potential future direction of azole antifungal therapeutics. 78 Perfect JR Dismukes WE Dromer F Clinical practice guidelines for the management of cryptococcal disease: 2010 update by the Infectious Diseases Society of America. Clin Infect Dis 2010 50 3 291 322 20047480 79 Stevens DL Bisno AL Chambers HF Practice guidelines for the diagnosis and management of skin and soft tissue infections: 2014 update by the Infectious Diseases Society of America. Clin Infect Dis 2014 59 2 e10 52 24973422 80 Masur H Brooks JT Benson CA Prevention and treatment of opportunistic infections in HIV-infected adults and adolescents: Updated guidelines from the Centers for Disease Control and Prevention, National Institutes of Health, and HIV Medicine Association of the Infectious Diseases Society of America. Clin Infect Dis 2014 58 9 1308 11 24585567 81 Tomblyn M Chiller T Einsele H Guidelines for preventing infectious complications among hematopoietic cell transplantation recipients: a global perspective. Biol Blood Marrow Transplant 2009 15 10 1143 238 19747629 82 Clinical practice guidelines in oncology: Prevention and treatment of cancer-related infections. 2014 National comprehensive care network Abington PA Available from: www.nccn.org/professionals/physician_gls/PDF/infections.pdf 8 December 2014 83 Al-Badriyeh D Slavin M Liew D Pharmacoeconomic evaluation of voriconazole versus posaconazole for antifungal prophylaxis in acute myeloid leukaemia. J Antimicrob Chemother 2010 65 5 1052 61 20237074 84 de la Camara R Jarque I Sanz MA Economic evaluation of posaconazole vs fluconazole in the prevention of invasive fungal infections in patients with GVHD following haematopoietic SCT. Bone Marrow Transplant 2010 45 5 925 32 19802030 85 Heimann SM Cornely OA Vehreschild MJ Treatment cost development of patients undergoing remission induction chemotherapy: a pharmacoeconomic analysis before and after introduction of posaconazole prophylaxis. Mycoses 2014 57 2 90 7 23790060 86•• Lyseng-Williamson KA Posaconazole: a pharmacoeconomic review of its use in the prophylaxis of invasive fungal disease in immunocompromised hosts. Pharmacoeconomics 2011 29 3 251 68 Provides a cost–benefit analysis of posaconazole to other antifungal regimens. 21309616
PMC005xxxxxx/PMC5125546.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 0267200 431 Am J Med Am. J. Med. The American journal of medicine 0002-9343 1555-7162 26797082 5125546 10.1016/j.amjmed.2015.12.024 NIHMS759101 Article Rheumatoid arthritis presenting as acute myopericarditis Mitchell Adam J. MD [email protected] 1 Alexy Tamas MD [email protected] 1 Rubinsztain Leon MD [email protected] 3 Iqbal Ayesha MD [email protected] 3 Shah Amit MD [email protected] 1 Zafari Maziar MD, PhD [email protected] 1 Searles Charles D. MD 1 1 Division of Cardiology, Department of Medicine, Emory University 2 Department of Radiology, Atlanta VA Medical Center 3 Division of Rheumatology, Department of Medicine, Emory University Corresponding author: Charles D. Searles, Atlanta VA Medical Center, 1670 Clairmont Road, Decatur, GA 30033, Charles Searles: [email protected] 27 9 2016 18 1 2016 5 2016 01 5 2017 129 5 e17e18 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. myopericarditis myocarditis pericarditis rheumatoid arthritis pleuropericardial effusion To the Editor A 50-year-old man with hypertension, type-2 diabetes, gout, and treated latent tuberculosis infection presented to the Emergency Department with 12 hours of angina. Initial evaluation for acute coronary syndrome and pulmonary embolism were negative. However, 20 hours after hospital admission the patient developed substernal chest pain, 1mm ST-segment elevation in the inferior ECG leads, and repeat troponin was 4.67 ng/mL (normal: < 0.09 ng/mL). Transthoracic echocardiography demonstrated normal biventricular function with no regional wall motion abnormalities. Urgent coronary angiogram revealed angiographically normal coronary arteries and no evidence of plaque rupture, coronary embolization, or dissection. Erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) were 97 mm/Hr and 305.8 mg/L, respectively. The patient was treated for myopericarditis with indomethacin and colchicine, but continued to experience symptoms of pleuritic chest pain, dyspnea, malaise, poor appetite, night sweats, and weight loss over the next several weeks. Cardiac MRI eight weeks after hospital discharge confirmed myocardial inflammation with residual, hemodynamically insignificant pericardial effusion and the interval development of a large left sided pleural effusion (Figure). Laboratory investigation showed persistently elevated ESR (86 mm/Hr) and CRP (63.2 mg/L), positive rheumatoid factor (82 IU/mL), negative antinuclear antibody, and negative Quantiferon-GOLD. Diagnostic thoracentesis revealed bloody, lymphocyte predominant pleural fluid with negative bacterial, fungal, and viral cultures as well as negative cytology. The differential diagnosis for exudative pleural effusions includes malignancy, infection, and systemic inflammatory diseases. Bloody pleural effusions are due to underlying malignancy in approximately 50% of cases. Lymphocytic effusions (>80% lymphocytes) can be caused by malignancy (lymphoma), tuberculosis, sarcoidosis, chylothorax, rheumatoid arthritis (RA), and yellow-nail syndrome1,2. Our patient’s pleural fluid cultures and cytology were negative. Adenosine deaminase was low (14.5 U/L) and active tuberculosis was ruled out with three negative acid-fast bacilli sputum cultures. The serum anti-cyclic citrullinated peptide antibody level was found to be markedly elevated (>250 U), establishing the diagnosis of RA. The patient’s pleural effusion resolved spontaneously and has not recurred during six months of follow-up. Initial laboratory abnormalities, including anemia, thrombocytopenia, acute kidney injury, and elevated inflammatory markers, normalized without immunosuppressive therapy. However, over the following months the patient developed increasing arthralgia, particularly in his wrists and elbows, and was started on azathioprine. RA is a progressive, systemic inflammatory disease that usually develops insidiously. Cardiac manifestations of RA include pericarditis, myocarditis, and coronary vasculitis3. Postmortem studies document much higher rates of cardiac involvement than are observed clinically. For instance, while 30–50% of patients with RA have postmortem evidence of pericarditis4, pericarditis as a clinical manifestation of RA occurs in fewer than 10% of patients3. Cardiac MRI studies also suggest that subclinical myocardial abnormalities are not uncommon in RA patients5. Although it is not uncommon for patients with known RA to suffer cardiac or pulmonary complications, it is rare for the initial presentation of RA to be myopericarditis. Funding Sources: None Figure Cardiac MRI demonstrates thickening of the apical segments of the left ventricular myocardium with associated patchy delayed enhancement, as well as diffuse thickening and delayed enhancement of the pericardium. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. The authors have no conflicts of interest to disclose. All authors have access to the data and contributed to preparing the manuscript and it is not under consideration elsewhere. References 1 McGrath EE Anderson PB Diagnosis of pleural effusion: a systematic approach American journal of critical care : an official publication, American Association of Critical- Care Nurses 2011 20 2 119 127 quiz 128 2 Collins TR Sahn SA Thoracocentesis. Clinical value, complications, technical problems, and patient experience Chest 1987 91 6 817 822 3581930 3 Voskuyl AE The heart and cardiovascular manifestations in rheumatoid arthritis Rheumatology 2006 45 Suppl 4 iv4 iv7 16980723 4 Hurd ER Extraarticular manifestations of rheumatoid arthritis Seminars in arthritis and rheumatism 1979 8 3 151 176 370982 5 Kobayashi Y Giles JT Hirano M Assessment of myocardial abnormalities in rheumatoid arthritis using a comprehensive cardiac magnetic resonance approach: a pilot study Arthritis research & therapy 2010 12 5 R171 20836862
PMC005xxxxxx/PMC5125619.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 0410462 6011 Nature Nature Nature 0028-0836 1476-4687 26886794 5125619 10.1038/nature16952 NIHMS791522 Article Structural basis for activity regulation of MLL family methyltransferases Li Yanjing 12 Han Jianming 12 Zhang Yuebin 3 Cao Fang 4 Liu Zhijun 12 Li Shuai 5 Wu Jian 12 Hu Chunyi 12 Wang Yan 12 Shuai Jin 12 Chen Juan 12 Cao Liaoran 3 Li Dangsheng 6 Shi Pan 7 Tian Changlin 78 Zhang Jian 5 Dou Yali 4 Li Guohui 3 Chen Yong 12 Lei Ming 12 1 National Center for Protein Science Shanghai, State Key Laboratory of Molecular Biology, Institute of Biochemistry and Cell Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, 333 Haike Road, Shanghai 201210, China 2 Shanghai Science Research Center, Chinese Academy of Sciences, Shanghai 201204, China 3 Laboratory of Molecular Modeling and Design, State Key Laboratory of Molecular Reaction Dynamics, Dalian Institute of Chemical Physics, The Chinese Academy of Sciences, Dalian, Liaoning 116023, China 4 Department of Pathology, University of Michigan Medical School, 1301 Catherine, Ann Arbor, Michigan 48109, USA 5 Department of Pathophysiology, Key Laboratory of Cell Differentiation and Apoptosis of the Chinese Ministry of Education, Shanghai JiaoTong University School of Medicine, Shanghai 200025, China 6 Shanghai Information Center for Life Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai 200031, China 7 High Magnetic Field Laboratory, Chinese Academy of Sciences, Hefei 230031, China 8 National Laboratory for Physical Science at the Microscale and School of Life Sciences, University of Science and Technology of China, Hefei 230026, China Correspondence and requests for materials should be addressed to M.L. ([email protected]), Y.C. ([email protected]) or G.L. ([email protected]) 17 11 2016 17 2 2016 25 2 2016 28 11 2016 530 7591 447452 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. The mixed lineage leukaemia (MLL) family of proteins (including MLL1–MLL4, SET1A and SET1B) specifically methylate histone 3 Lys4, and have pivotal roles in the transcriptional regulation of genes involved in haematopoiesis and development. The methyltransferase activity of MLL1, by itself severely compromised, is stimulated by the three conserved factors WDR5, RBBP5 and ASH2L, which are shared by all MLL family complexes. However, the molecular mechanism of how these factors regulate the activity of MLL proteins still remains poorly understood. Here we show that a minimized human RBBP5–ASH2L heterodimer is the structural unit that interacts with and activates all MLL family histone methyltransferases. Our structural, biochemical and computational analyses reveal a two-step activation mechanism of MLL family proteins. These findings provide unprecedented insights into the common theme and functional plasticity in complex assembly and activity regulation of MLL family methyltransferases, and also suggest a universal regulation mechanism for most histone methyltransferases. Methylation of histone H3 Lys4 (H3K4), which is predominantly associated with actively transcribed genes1–3, is mainly mediated by MLL family histone lysine methyltransferases (HKMTs). Mammalian MLL family HKMTs contain six members (MLL1–MLL4, SET1A and SET1B)2–4, each of which has crucial yet non-redundant roles in cells4–6. MLL1 has been the most intensively studied because of its involvement by chromosomal translocations in a variety of acute lymphoid and myeloid leukaemias6,7. Recently, inactivating mutations in MLL3 (also known as KMT2C) and MLL4 (KMT2D) have been identified in several types of human tumours and in Kabuki syndrome8–12. In contrast to most SET [SU(VAR)3–9, E(Z) and TRX]-domain-containing methyltransferases, MLL1 protein alone exhibits poor HKMT activity13,14. The crystal structure of the MLL1 SET domain (MLL1SET) reveals an open conformation that is not efficient for the methyl transfer from the cofactor S-adenosyl-L-methionine (AdoMet) to the target lysine15. The optimal HKMT activity of MLL1 requires additional factors, WDR5, RBBP5 and ASH2L, which are shared core components of all MLL complexes and also evolutionarily conserved from yeast to humans13,16. Depletion of any of these components results in the global loss of H3K4 methylation to varying degrees16–18. Despite the importance of WDR5, RBBP5 and ASH2L, it is still unclear how these factors stimulate the HKMT activity of MLL proteins. In this work, our biochemical and structural analyses reveal how RBBP5–ASH2L binds and activates MLL family methyltransferases through a conserved mechanism. RBBP5–ASH2L binds and activates MLLs We first examined the effects of individual components (WDR5, RBBP5 and ASH2L) and their combinations on the HKMT activities of MLL family methyltransferases. We selected the carboxy-terminal conserved regions of MLL proteins containing both the WIN (WDR5-interaction) motif and the SET domain in activity assays14 (Fig. 1a). For simplicity, hereafter we use ‘MLL’ to represent the MLL WIN-SET fragment, and ‘MLLSET’ to represent the MLL SET domain unless stated otherwise. Consistent with previous observations14,15,19, activity assays showed that the RBBP5–ASH2L heterodimer substantially upregulated the HKMT activity of MLL1, and this activity was further stimulated by the addition of WDR5 (Fig. 1b and Extended Data Fig. 1a–c). By contrast, MLL2–MLL4, SET1A and SET1B can be fully activated by just RBBP5–ASH2L, and WDR5 was dispensable for activity regulation (Fig. 1b and Extended Data Fig. 1a–c). The stimulatory effect of RBBP5–ASH2L on MLL HKMT activities indicated a possible direct interaction between RBBP5–ASH2L and MLL proteins20–22. Indeed, a glutathione S-transferase (GST) pull-down assay clearly showed that all MLL proteins directly interact with RBBP5–ASH2L (Fig. 1c and Extended Data Fig. 2a). Among them, MLL2–MLL4 could be efficiently pulled down by GST–ASH2L–RBBP5, whereas SET1A and SET1B maintained a medium level of interaction with RBBP5–ASH2L (Fig. 1c and Extended Data Fig. 2a). By contrast, MLL1 only exhibited a very weak interaction with RBBP5–ASH2L under low-salt buffer conditions (Fig. 1c and Extended Data Fig. 2a). Fluorescence polarization analysis also revealed that MLL proteins interact with RBBP5–ASH2L with very different binding affinities ranging from ~100 nM (for MLL3) to more than 100 μM (for MLL1) (Extended Data Fig. 2b). Formation of the RBBP5–ASH2L heterodimer is a prerequisite for MLL binding, as neither RBBP5 nor ASH2L alone can stably associate with MLL proteins (Extended Data Fig. 2c, d). Notably, MLL proteins can also stabilize the RBBP5–ASH2L interaction when high-salt buffer was used in the pull-down assay (Extended Data Fig. 2a), consistent with the observation that the RBBP5–ASH2L interaction is highly sensitive to ionic strength (Extended Data Fig. 2e). Because MLL1 only maintained a weak direct interaction with RBBP5–ASH2L, we proposed that full activation of MLL1 by RBBP5–ASH2L requires the bridging molecule WDR5 that can interact with both MLL1 and RBBP5–ASH2L simultaneously. Consistent with this idea, the stimulatory effect of WDR5 on MLL1 HKMT activity is minimized when the protein concentration was increased in the assay (Extended Data Fig. 2f). Furthermore, the fusion of RBBP5 and MLL1 together achieved a robust HKMT activity that cannot be further stimulated by the addition of WDR5 (Extended Data Fig. 2g), suggesting that WDR5 per se is not directly involved in the MLL HKMT enzymatic reaction. Collectively, we conclude that RBBP5–ASH2L is the major functional unit that binds and activates MLL proteins. Conversely, WDR5 may have an indirect role in promoting HKMT activity by acting as a bridging molecule to facilitate the formation of MLL complexes under certain assay conditions, and this may explain the apparent discrepancy in reports about the role of WDR5 in the activity regulation of MLL complexes19–22. Complex structure of MLL3–RBBP5–ASH2L To determine the structural basis of how RBBP5–ASH2L activates MLL proteins, we first dissected the interactions among RBBP5, ASH2L and the SET domains of MLL proteins. Consistent with previous studies23,24, the ASH2L C-terminal SPRY (splA and ryanodine receptor) domain is sufficient to form a heterodimer with RBBP5 to stimulate the HKMT activity of MLL proteins (Fig. 1d, compare lanes 1 and 5). Three adjacent short motifs of RBBP5 were identified for the stimulation of MLL HKMT activity (residues 330–344, activation segment, AS), the interaction with ASH2L (residues 344–363, ASH2L-binding motif, ABM), and the association with WDR5 (residues 369–381, WDR5-binding motif, WBM)24,25 (Fig. 1a). A preformed RBBP5AS-ABM–ASH2LSPRY complex can stimulate MLL3SET HKMT activity to levels of ~70% of full-length RBBP5–ASH2L (Fig. 1d, compare lanes 2, 4 and 6), indicating that this minimized RBBP5AS-ABM–ASH2LSPRY heterodimer is essential for the stimulation of MLL3SET activity, and that other regions of RBBP5 might have a minor role in this process. We determined the crystal structure of this minimized ternary complex composed of MLL3SET, RBBP5AS-ABM and ASH2LSPRY (hereafter referred to as M3RA) in the presence of S-adenosyll-homocystein (AdoHcy) and a substrate peptide (H3 residues 1–9) (Fig. 2a, Extended Data Table 1 and Extended Data Fig. 3a). Notably, we crystallized the M3RA complex both with and without the H3 peptide in one asymmetric unit (Extended Data Fig. 3b). In the M3RA complex, RBBP5AS-ABM adopts an extended conformation that consists sequentially of two β-strands (activation segment) and a rigid coil (ABM), which respectively mediate the interactions with MLL3SET and ASH2LSPRY (Fig. 2a). The overall fold of MLL3SET is similar to other SET-domain proteins, and shares the conserved features of N- and C-terminal regions (SET-N and SET-C), an insertion region (SET-I) and post-SET motifs15,26,27 (Fig. 2b). The active site residues of MLL3SET, the conformation of the target lysine and an invariant water molecule, are almost identical to those of the active site of DIM-5 (ref. 28), suggesting a catalytically active configuration of MLL3SET (Fig. 2c). The ‘U’-shaped cofactor product AdoHcy binds into a well-defined surface pocket on MLL3SET through an extensive network of highly conserved interactions as observed in other SET-domain structures (Fig. 2d and Extended Data Fig. 3c, d). The H3 substrate peptide sits in an opposite groove on the surface of MLL3SET, and an intricate network of hydrogen bonds stabilizes the binding (Fig. 2e and Extended Data Fig. 3e). The unique geometry of the H3-binding groove specifically recognizes Thr3H3 and Arg2H3, defining the substrate specificity of MLL3SET (Extended Data Fig. 3f, g). Since all the H3-peptide-binding residues in MLL3SET are highly conserved in other MLL proteins (Extended Data Fig. 4), we conclude that all MLL proteins achieve the substrate specificity towards H3K4 through the same recognition mechanism as observed in the M3RA complex. Interfaces in the M3RA complex The structure of the M3RA complex reveals extensive interactions among ASH2LSPRY, RBBP5AS-ABM and MLL3SET. ASH2LSPRY recognizes RBBP5ABM through extensive salt-bridge and hydrogen-bonding interactions; the C-terminal portion of RBBP5ABM adopts a coiled conformation sitting on two arginine residues (Arg343 and Arg367) at the centre of the basic pocket of ASH2LSPRY (Fig. 3a). Mutations of ASH2L Arg343 and its interacting residues in RBBP5ABM (Glu349 and Asp353) completely abrogated the RBBP5ABM–ASH2LSPRY interaction (Extended Data Fig. 5a) and impaired the HKMT activity of the MLL3 complex (Extended Data Fig. 5b). The primary feature of the RBBP5AS–MLL3SET interaction is the inter-molecular β-sheet interactions involving two strands of the L-shaped RBBP5AS paring with β4 and an induced strand β7 immediately before helix αC of MLL3SET (Fig. 3b). Mutations of residues on this L-shaped motif partially decreased the HKMT activity of the MLL3 complex (Extended Data Fig. 5c). In addition to these binary contacts, the side chain of the conserved Arg4806 of MLL3SET sticks outside towards an acidic pocket formed by both RBBP5ABM and ASH2LSPRY, forming five salt-bridge and hydrogen-bonding interactions with Glu347RBBP5, Tyr313ASH2L and Gln354ASH2L (Fig. 3c). This extensive electrostatic network functions as an anchor point to fix the relative position of MLL3SET to ASH2LSPRY, and is crucial for assembly of the MLL3–RBBP5–ASH2L complex. Because the RBBP5AS-ABM–ASH2LSPRY-interacting residues are highly conserved in MLL-family proteins (Extended Data Fig. 4), we proposed that all MLLSET domains including MLL1SET should interact with RBBP5–ASH2L through the same molecular surface as observed in the M3RA complex. In support of this idea, alanine substitutions of the conserved arginine residues in all MLL proteins, which do not affect the overall fold of MLL proteins (Extended Data Fig. 5d), abolished the interaction between MLLSET and RBBP5–ASH2L (Fig. 3d), and substantially decreased the HKMT activities of all MLL complexes (Fig. 3e). Accordingly, mutations of the arginine-interacting residues on RBBP5 and ASH2L (RBBP5 Glu347 and ASH2L Gln354) also weakened the association of MLLSET with RBBP5–ASH2L, and reduced the HKMT activities of MLL complexes (Extended Data Fig. 5e, f). Together, our data confirmed that the electrostatic network observed at the MLL3SET–RBBP5–ASH2L interface is essential for the interaction between RBBP5–ASH2L and all MLL proteins. Interestingly, an inactivating mutation of the same key arginine residue in MLL4 (Arg5432Trp) that was identified in patients with non-Hodgkin lymphoma9 also disrupted the interaction between MLL4 and RBBP5–ASH2L and abolished the HKMT activity (Fig. 3d, e), indicating that loss of a stable MLL4–RBBP5–ASH2L association leads to lymphomagenesis. Difference between MLL1 and other MLL proteins The structure of the M3RA complex revealed that subtle sequence differences in the RBBP5–ASH2L-binding region (residues 4804–4814 in MLL3) are probably responsible for the ability of RBBP5–ASH2L to distinguish MLL1 from other MLL proteins (Fig. 4a). Most notably, the side chain of Val4809 in the SET-I motif of MLL3SET fits snugly in a shallow pocket formed by both RBBP5 and MLL3SET (Fig. 4b), which can also accommodate the corresponding residues of MLL2, MLL4, SET1A and SET1B at the equivalent positions, but not for the bulky residue Gln3867 of MLL1 (Fig. 4a, b). In addition, the side-chain methyl group of MLL3SET Thr4803 is surrounded by three hydrophobic resides of RBBP5AS (Leu339, Val343 and Tyr345) (Fig. 4c). By contrast, a large hydrophilic residue Asn3861 at this position in MLL1 is incompatible with RBBP5 binding (Fig. 4c). Thus, we proposed that two residues (Asn3861 and Gln3867) in MLL1 weaken the otherwise stable interaction between RBBP5–ASH2L and MLL1. Indeed, both MLL1-to-MLL2 (Asn3861Ile/Gln3867Leu) and MLL1-to-MLL3 (Asn3861Thr/Gln3867Val) double mutants of MLL1 re-gained stable interactions with RBBP5–ASH2L (Fig. 4d), and WDR5 had no further stimulatory effect on the HKMT activities of these mutants (Fig. 4e). Therefore, mutations of these two residues restore the strong RBBP5–ASH2L binding ability of MLL1 and thus bypass the requirement of WDR5 as the bridging molecule for the optimal HKMT activity of the MLL1 complex. This idea is further supported by the crystal structure of the MLL1SETN3861I/Q3867L-RBBP5AS-ABM-ASH2LSPRY complex (hereafter referred to as M1MRA, in which ‘M’ denotes mutant) (Fig. 4f and Extended Data Table 2). The structure of M1MRA highly resembles that of M3RA, with an identical interface as the one between MLL3SET and RBBP5–ASH2L (Fig. 4g), strongly indicating that the RBBP5–ASH2L heterodimer interacts with and activates all MLL proteins through a conserved mechanism. Notably, the equivalent residues of MLL1 Asn3861 in SET1A (Gln1600) and SET1B (Gln1816) also have large hydrophilic side chains and therefore are not optimal for RBBP5–ASH2L binding (Fig. 4a). This is consistent with the medium levels of interaction of SET1A and SET1B with RBBP5–ASH2L observed in the pull-down and fluorescence polarization assays (Extended Data Fig. 2a, b). Activation mechanism of MLL complexes Next we asked why MLL proteins by themselves are catalytically inactive, and how RBBP5–ASH2L stimulates their HKMT activities. One prevailing model suggests that the SET domain of MLL adopts an open conformation, and the interaction with regulatory factors induces MLL SET domain into a closed conformation15. To test this model, we crystallized apo MLL3SET and determined its structure in complex with AdoHcy (Extended Data Fig. 6a). Surprisingly, the apo structure of MLL3SET was almost indistinguishable from the active conformation of MLL3SET in the M3RA complex (Fig. 5a). In addition, we also determined the crystal structure of MLL1SETM(MLL1SETN3861I/Q3867L), the SET-I motif of which exhibits an even more closed conformation than that in the M1MRA complex (Fig. 5b and Extended Data Fig. 6b). Nevertheless, our data clearly showed that both MLL3 and MLL1M by themselves are catalytically inactive (Fig. 4e and Extended Data Fig. 1). This apparent discrepancy between the low enzymatic activity and the closed conformation of MLL1SETM or MLL3SET led us to propose that, in the absence of RBBP5–ASH2L, MLLSET might be highly dynamic, and the configuration of MLL SET-I motif captured in the crystal structure is a snapshot of a spectrum of conformations of a mobile motif. In support of this idea, normal mode analysis revealed a highly dynamic motion of the SET-I motif in apo MLL1SETM and MLL3SET, which is substantially suppressed upon the association with RBBP5–ASH2L (Supplementary Videos 1–5). To test this model experimentally, we use 19F-NMR (fluorine-19 nuclear magnetic resonance) to probe the structural dynamics of MLL3SET in solution. The 19F-NMR spectrum of Phe4827, a key residue at the substrate-binding site in the SET-I motif, displayed two peaks at different chemical shifts, defining at least two different conformations or states with dynamic exchanges (Fig. 5c). With titration of RBBP5–ASH2L, the 19F-NMR spectrum showed prominent changes with conformational equilibrium towards a single active state, indicating that RBBP5–ASH2L reduced the flexibility of SET-I to lock it in an active state (Fig. 5c). By contrast, Tyr4762 that is located in the SET-N motif exhibited no peak shift upon the addition of RBBP5–ASH2L (Fig. 5c). To provide further insight into this dynamic process, we carried out molecular dynamics simulation to investigate how RBBP5–ASH2L affects the structures of MLL3SET and MLL1SETM. Results showed that RBBP5–ASH2L reduces the root mean square fluctuation of helix αB and strand β7 of MLL3SET substantially (Fig. 5d). This coincides with our observation that the most variable element in apo MLLSET is the αB helix, illustrated by the superimposition of four apo MLLSET structures (Extended Data Fig. 6c). Furthermore, a flexible loop in apo MLL3SET (L6C) is induced to form strand β7 by pairing with strand β1 of RBBP5AS (Fig. 5e). Other than the structural variation of individual residues, molecular dynamics simulation also clearly showed that the cross-correlation within the SET-I motif was greatly enhanced upon RBBP5–ASH2L association (Fig. 5f and Extended Data Fig. 6d–f). The reduced flexibility of the SET-I motif may help cofactor binding and substrate recognition. Indeed, isothermal titration calorimetry (ITC) analysis showed that the binding affinities of cofactor to MLL3SET and MLL1SETM are markedly increased in the presence of RBBP5–ASH2L (Extended Data Fig. 7a, b). Furthermore, the association with RBBP5–ASH2L also facilitates MLL3 binding with the H3 substrate peptide (Extended Data Fig. 7c). Notably, Phe336 at the beginning of β1 in RBBP5AS stacks together with the side chains of MLL3SET/MLL1SETM Arg4845/Arg3903, Tyr4846/Phe3904 and Tyr4825/Tyr3883, and the latter makes a direct hydrogen-bonding interaction with AdoHcy (Fig. 5e). Molecular dynamics simulation revealed an obvious stabilizing effect of RBBP5–ASH2L on this network of interactions, which could explain the enhanced cofactor-binding ability for the M3RA complex (Extended Data Fig. 7d). Quantum mechanics/molecular mechanics (QM/MM) investigations further indicated that the presence of RBBP5–ASH2L facilitated the methyl transfer process from the cofactor AdoMet to the target lysine by lowering the energy barrier (Extended Data Fig. 7e). Taken together, we conclude that the RBBP5–ASH2L-induced conformational constraints on the SET-I motif help to stabilize MLLSET in a conformation competent for cofactor binding and substrate recognition. Structural comparison of the M3RA complex structures with and without the H3 peptide revealed a role of the substrate peptide in further stabilizing the active conformation of MLL3SET, which has been observed in other SET-domain methyltranferases28. After H3 binding, a marked local structural rearrangement occurs to loop LB5 between helix αB and strand β5 in the SET-I motif, leading to the completion of a narrow hydrophobic channel that orients the H3 Lys4 side chain for catalysis (Fig. 5g). Remarkably, the side chain of MLL3 Val4824 shifts ~4.1 Å and rotates ~50° relative to its position in the H3-peptide-free structure, enclosing the target lysine access channel (Extended Data Fig. 7f). Collectively, our studies suggest a novel two-step mechanism for MLLSET activation: interaction with the RBBP5–ASH2L heterodimer reduces the inherent flexibility of MLLSET and favours formation of a catalytically competent conformation; and then H3 substrate binding induces a local conformational change in the SET-I motif of MLLSET to achieve the fully active configuration that facilitates the methyl transfer process (Fig. 5h). Implications for other methyltransferases Structural comparison of the M1MRA and M3RA complexes with a large group of SET domain proteins reveals a striking similarity with other intrinsic active methyltransferases. In all SUV39- and SET2-family proteins, a short fragment amino-terminal to the pre-SET region (referred to as the activation segment) interacts with the SET-I motif in the same manner as RBBP5AS binding to MLL3SET (Extended Data Fig. 8a, b). Deletion of this activation segment from DIM-5 did not affect the overall fold of the protein but completely abrogated the HKMT activity of DIM-5, underscoring the importance of this segment in DIM-5 activity regulation (Extended Data Fig. 8c, d). Such an activation segment is also found in the EZH2 complex structure29, further supporting a conserved activation mechanism for a subset of SET-domain-containing methyltransferases. In summary, the present structural, biochemical and computational analyses provide new insights into the assembly and regulation mechanism of MLL family complexes. Our results suggest that a minimized RBBP5–ASH2L heterodimer is the structural unit to interact with and activate all MLL family histone methyltransferases. By contrast, WDR5 is not directly involved in the enzymatic stimulation of MLL complexes. WDR5 may serve as a recruitment module or crosstalk mediator to regulate H3K4 methylation in vivo30–34. METHODS No statistical methods were used to predetermine sample size. Protein expression and purification The SET domains of MLL family proteins (with or without the WIN motif), RBBP5, ASH2L, WDR5 and their truncations or mutants were purified as described before19. Escherichia coli Rosetta cells bearing expression plasmids were induced for 16 h with 0.1 mM IPTG at 18 °C, and the cells were collected by centrifugation. For MLL expression, 10 μM ZnSO4 was included in the media. The cell pellets were resuspended in lysis buffer (50 mM Tris-HCl, pH 8.0, 400 mM NaCl, 10% glycerol, 2 mM 2-mercaptoethanol, and home-made protease inhibitor cocktail). The cells were broken by sonication and cleared by ultracentrifugation at 100,000g for 30 min. The proteins were purified using Ni-NTA agarose beads (Qiagen) for His-tagged proteins or Glutathione Sepharose 4B beads (GE) for GST-tagged proteins, followed by enzyme digestion to remove the tags and gel-filtration chromatography. MLLSET, ASH2LSPRY, WDR5 and RBBP5 fragments were separated on Hiload Superdex 75, while full-length proteins of ASH2L and RBBP5 were separated on Hiload Superdex 200. The buffer for gel-filtration chromatography contains 25 mM Tris-HCl, pH 8.0, 150 mM NaCl except for MLLSET (which is in buffer 50 mM Tris-HCl, 300 mM NaCl and 10% glycerol, pH 8.0). The purified proteins were concentrated to 10–20 mg ml−1 and store at −80 °C. RBBP5 peptides were separated on Hiload Superdex 75 after tag digestion in buffer (100 mM NH4HCO3) and the peptide-containing fractions were lyophilized. The MLLSET–RBBP5AS-ABM–ASH2LSPRY complex was obtained by step-wise gel-filtration chromatography; binary complex of RBBP5AS-ABM–ASH2LSPRY was first purified, and then mixed with MLLSETM or MLL3SET, followed by separation on Hiload Superdex 75. Mutations were introduced by PCR-based site-directed mutagenesis, and mutated proteins were purified using the same protocol as described above. Crystallization, data collection and structural determination For structural studies, more than 50 different combinations of numerous RBBP5 fragments, ASH2LSPRY constructs, and SET domains from different MLL proteins were tested for crystallization. MLL3SET was crystallized in 100 mM Tris-HCl, pH 8.5, 3 M NaCl at 4 °C in the presence of 1 mM AdoHcy. Zinc single-wavelength anomalous dispersion (SAD) and native data sets of MLL3SET were collected at SSRF (Shanghai Synchrotron Radiation Facility in China) beamline BL17U at wavelengths of 1.2818 Å and 0.9793 Å, respectively. Data were indexed, integrated, and scaled using program HKL2000 (ref. 35). Crystals belong to space group P4132 and contain one MLL3SET per asymmetric unit. Zinc SAD phase determination, density modification and automatic model building were carried out using SHARP36. The initial model was further refined using the native data set diffracted at 2.8 Å. After several rounds of refinement in PHENIX package37 with manual rebuilding in COOT38, the final model has good stereochemistry with an R value of 18.0% and an Rfree of 22.9%. The MLL3SET–ASH2LSPRY–RBBP5AS-ABM complex was crystallized in 100 mM sodium cacodylate, pH 6.5, 10% PEG-3350, 0.1 M MgCl2 at 4 °C in the presence of 1 mM AdoHcy. The co-crystal with H3 peptide (ARTKQTARK) was obtained by soaking crystals in reservoir solution with 1 mM H3 peptide for 2 h before collection. A data set of 2.4 Å resolution was collected at Advanced Photon Source beamline 21ID-D at wavelength of 1.1272 Å. The crystal belongs to space group P21212 with cell dimension a = 80.342 Å, b = 236.076 Å, c = 44.416 Å. The complex structure was solved by molecular replacement using PHASER39 with ASH2LSPRY structure (PDB accession 3TOJ) and the MLL3 structure SET-N, SET-I, and SET-C motifs as search models. There are two MLL3SET–ASH2LSPRY–RBBP5AS-ABM complexes in one asymmetrical unit, and we can only observed H3 peptide in the density map of one complex. The model was further refined using PHENIX with manually rebuilding in COOT. MLL1SETN3861I/Q3867L crystals were grown by sitting drop vapour diffusion method at 4 °C in a solution containing 35% (v/v) tacsimate, pH 7.0, in the presence of 2 mM AdoHcy, and the crystals were cryo-protected in the same reservoir solution supplemented with 20% glycerol. Data sets were screened and collected at SSRF BL18U and BL19U. The structures were solved by molecular replacement (starting model PDB accession 2W5Y). The MLL1SETN3861I/Q3867L-ASH2LSPRY-RBBP5AS-ABM complex was crystallized at 200 mM NaCl, 20% PEG3350 in the presence of 2 mM AdoHcy. A data set of 1.9 Å resolution was collected at SSRF BL17U at wavelength of 0.9792 Å. The structures were solved by molecular replacement and further refined with PHENIX. All structure figures were generated using PyMOL (The PyMOL Molecular Graphics System, version 1.4.1 Schrödinger, LLC.). Histone methyltransferase assay In vitro methyltransferase assays were performed using an H3 peptide as the substrate. Two assay systems were used. The first one is the 3H-methyl-incorporation assay that measured the incorporation of 3H from [3H]AdoMet (S-adenosyl-L-[methyl-3H]-methionine) into the H3 peptide (9 mer: ARTKQTARK). Reactions were carried out at 22 °C for 1 h in the buffer containing 20 mM HEPES, pH 7.8, 5% glycerol, 5 mM dithiothreitol (DTT), 0.5 mM EDTA, 1 μCi [3H]AdoMet as previously described19. Unmodified H3 K4 peptides (0.25 mM) and 1 μM of WDR5, RBBP5, ASH2L and MLL proteins were used, except for SET1A (5 μM). For all activity assays, full-length WDR5, RBBP5 and ASH2L were used unless stated otherwise. MLL constructs containing both the WIN motif and SET domain were used. Each assay was performed in triplicate, and the mean ± s.d. was reported. The second assay system is to monitor the methylation kinetics of the H3 peptide substrate using MALDI–TOF (matrix-assisted laser desorption ionization–time-of-flight) mass spectrometry. Mass spectrometry analysis of the methylation process Methylation reactions were carried out in 20 mM HEPES, pH 7.8, 10 mM NaCl, 5 mM DTT, 250 μM AdoMet, 10 μM histone peptide (ARTKQTARKS) and 1 μM MLL complexes at 22 °C. The reaction was quenched at different time points by addition of trifluoroacetate (TFA) to 0.5%. Reaction mixture was diluted in 10 mg ml−1 CHCA (α-cyano-4-hydroxycinnamic acid) matrix and was spotted onto sample plate and air-dried. The molecular mass was measure by MALDI–TOF (AB SCIEX TOF/TOF 5800) operated in reflectron mode. Final spectra were the average of 200 shots per position at 200 different positions chosen randomly on each spot. To estimate the pseudo-first-order rate constants, we fit the decrease in the relative intensity of the unmodified peptide over time using a model for a single irreversible reaction [Lys4]t = [Lys4]0e−kt, in which [Lys4]0 is the initial concentration of the unmodified peptide, [Lys4]t represents the concentrations of the unmodified peptide at time t and k is the pseudo-first-rate constant. GST pull-down assays GST-fusion proteins and interacting partners were incubated with glutathione Sepharose 4B beads for 2 h at 4 °C in 100 μl buffer (50 mM Tris-HCl, 300 mM NaCl and 2 mM DTT, pH 8.0). After extensive wash with the same buffer, the bound proteins were eluted in elution buffer (50 mM Tris-HCl, pH 8.0, 300 mM NaCl and 15 mM reduced glutathione). The input samples and eluted samples were visualized on 12% SDS–PAGE by Coomassie blue staining. Initially, different pull-down buffers were tested and it was found that the interaction between ASH2L and RBBP5 could be disrupted by high ionic strength used in the pull-down assay, whereas the formation of the MLL–RBBP5–ASH2L trimeric complex is relatively insensitive to salt concentration. Thus, in most pull-down assays, buffer with 300 mM NaCl was chosen to assure undisrupted RBBP5–ASH2L interaction and also keep protein stable through pull-down experiments unless stated otherwise. Isothermal titration calorimetry The equilibrium dissociation constants of cofactor binding to MLLSET or MLLSET–RBBP5AS-ABM–ASH2LSPRY were determined by an ITC200 calorimeter (GE healthcare). The binding of proteins (20–200 μM) and cofactor AdoMet (0.5–2mM) were measure in the 25 mM Tris-HCl, pH 8.0, 300 mM NaCl at 20 °C. ITC data were analysed and fit using Origin 7 (OriginLab) using one-site model. Owing to instability of apo MLL protein during ITC experiments, curve fitting errors for apo MLL titration are relatively large, so the binding parameters of apo MLL proteins are rough estimations. Fluorescence polarization assay Different MLL proteins were diluted in 20 mM HEPES, pH 7.8, 150 mM NaCl, 10% glycerol, 0.5 mg ml−1 BSA to a serial of concentrations from 25 nM to 50 μM. The FAM-labelled RBBP5 peptide (residues 330–363) was mixed with ASH2LSPRY and used at a final concentration of 100 nM. The final volume was brought up to 100 μl with dilution buffer (20 mM HEPES, pH 7.8, 150 mM NaCl, 10% glycerol and 0.5 mg ml−1 BSA) and incubated in dark for 30 min. The fluorescence polarization values were measured using Synergy Neo Multi-Mode Reader (Bio-Tek) at 27 °C. Excitation wavelength was 485 nm and emission was detected at 528 nm. Fluorescence was quantitated with GEN 5 software and date was analysed with Prism 6. For MLL1, SET1A and SET1B, the binding is not saturated even at the highest protein concentration, so the calculated Kd should be an estimated lower limit of Kd value. 19F-NMR spectra measurements Expression of 19F-labelled proteins was achieved by an established protocol by incorporation of non-natural amino acid tfmF (l-4-trifluoromethylphenylalanine) into specific sites using genetic code TAG40. The 19F-labelled MLL3SET proteins were purified using the same protocol as for wild-type MLL3SET protein. The 19F-NMR spectra were obtained on a Bruker DMX Avance-500 MHz spectrometer equipped with a 5 mM PABBO room temperature probe. The spectra of 0.3 mM MLL3 F4827tfmF or 0.35 mM MLL3SET Y4762tfmF with or without RBBP5–ASH2L were collected at 293 K. The observation channel was tuned to 19F (470.54 MHz), with 512 free induction decay accumulations in every 3-s recycling delay. Each one-dimensional 19F-spectrum was acquired with a standard pulse program with a 90° pulse width of 16.75 μs and power at 35.9 W. 19F-chemical shifts were referenced to an external standard TFA. The free induction decay accumulations, which consisted of 20,480 complex points, was linear predicted to 40,560 points, backward linear predicated three points, and apodized with 20 Hz Lorentzian filter. All spectral processing was performed with Topspin 3.2 software. Normal mode analysis Normal modes were calculated using the NOMAD-Ref method41. For all MLL3SET, MLL1SETM , M3RA and M1MRA structures, default parameters in the method were used, including the analysis of ‘all atoms’, ‘default distance weight parameters for elastic constant’ of 5.0 Å, ‘ENM cutoff values’ of 1 Å, ‘average RMSD in output trajectories’ of 1.0 Å and ‘output’ of the lowest 16 modes. The first six trivial normal modes are discarded because they represent only translation and rotation. The motion patterns under certain mode and angle monitoring were achieved by using PyMOL. Molecular dynamics simulation To delineate how RBBP5–ASH2L modulates the dynamic behaviour of MLLSET domain, we performed molecular dynamic simulations (100 ns) of MLL3SET and MLL1SET in the presence or absence of RBBP5–ASH2L, respectively. The complex structure of MLLSET with RBBP5–ASH2L were centred into a 115 × 115 × 115 Å3 cubic box and dissolved with TIP3P waters. 0.1 M NaCl ions were used to neutralize the net charge of the system. While for the systems of MLLSET alone, we just removed the RBBP5–ASH2L from the complex to make sure the identical conformations of MLLSET before performing molecular dynamics simulations. The same procedures were used in setting up the MLLSET domains without RBBP5–ASH2L except for a smaller cubic box (83 × 83 × 83 Å3). All molecular dynamics simulations were performed using Gromacs 5.0.4 with Charmm36 force field. After the energy minimization of the whole system using the steepest descent algorithm, we first gradually heated the system to 300 K under NVT condition. Then we equilibrated the solvent and ions around the protein using NPT ensemble. In the equilibrations, the backbone of the protein was constrained with a harmonic potential of 1,000 kJ mol−1. The leap-frog integrator was used with an integration time-step of 2 fs. The Berendsen barostat was used to control the pressure at 1 bar with a coupling constant of 2 ps and the modified Berendsen (V-rescale) thermostat was employed to control the temperature of the systems at 300 K with a time constant of 0.1 ps. The Particle Mesh Ewald method was used to compute the electrostatic interactions with a real-space cut-off distance of 1 nm. The same cutoff value was chosen for treating the van der Waals interactions. After a 5 ns equilibration, we conducted the production molecular dynamics by changing the pressure and the thermostat coupling to Parrinello–Rahman and Nose–Hoover with coupling constants of 5 ps and 1 ps, respectively. The dynamical network analysis of MLL3SET and MLLSETM were performed using networkSetup in VMD. Cα atoms of MLLSET were defined as the node domains and the dynamical contact was drawn if two nodes were within a cutoff distance of 4.5 Å for at least 75% of the molecular dynamics trajectory. The cross correlation data were also calculated to weight edges in the dynamical network. The edge distances dij, which define the probability of information transfer across a given edge: dij = −log(\|Cij\|), were derived from pairwise correlations (Cij) using program Carma. To investigate how RBBP5–ASH2L affects methyl transfer process from the cofactor AdoMet to the target lysine of the H3 substrate, we performed QM/MM simulations to calculate the potentials of mean force for the methyl transfer reaction along the reaction coordinate (RC) of r(CM − Sδ) − r(CM − Nη1). Initial structures of the QM/MM simulations were derived from the snapshots of molecular dynamics trajectories in the presence of AdoMet and H3 substrates, simultaneously. Then each structure was solvated in an equilibrated 25 Å spherical water box represented by the TIP3P water model. The water box was centred at the centre of mass of the target lysine residue accepting the methyl group. In total, 20 atoms were selected as the QM zone, including the sulfur atom and the to-be-transferred methyl group on the peptide as well as the lysine residue. The simulation was performed in NVT ensemble at 300 K. The hybrid QM/MM method was used in the simulation. QM interactions are calculated using semi-empirical AM1 method and three GHO atoms (C4′ and CB, which connect the sulfur atom to the other two methyl group, and CD of the lysine residue) were selected as the boundary between QM and mM regions. The solvent boundary potential was treated by the generalized solvent boundary potential method and all atoms out of the water box were fixed. The umbrella sampling method was used to model the reaction process, with the reaction coordinate set as the difference between the sulfur atom on the peptide and the nitrogen atom on the QM lysine. The whole reaction process was distributed into 46 windows and the corresponding reaction coordinate ranged from −1.5 to 2.0 Å with an interval of 0.1 Å. Systems were restrained to each window with a force constant of 500 kcal mol−1 Å−2. Extended Data Extended Data Figure 1 Methyltransferase activity of MLL1–MLL4, SET1A and SET1B with the different combinations of WDR5, RBBP5 and ASH2L a, HKMT activities determined by the 3H-methylin-corporation assay. MLL constructs were chosen to contain both the WIN motif and the SET domain. Full-length WDR5, RBBP5 and ASH2L were used. The HKMT activities are normalized to the activity of the MLL–WDR5–RBBP5–ASH2L complexes setting at 100%. Mean ± s.d. (n = 3) are shown. b, Representative MALDI–TOF spectra at different time points for MLL complexes and apo MLL proteins clearly revealed that MLL complexes have much higher HKMT activities than apo MLL proteins. The peaks for unmodified (un) and mono-, di- and tri-methylated products are labelled. The minor peaks are sodium adducts of major peaks (+22 Da). Asterisks denote the adduct of un-peaks; filled circles denote the adduct of mono-peaks; and filled squares denote the adduct of di-peaks. c, Comparison of the overall rates of the methylation reactions catalysed by different MLL proteins in the presence of WDR5–ASH2L–RBBP5 or ASH2L–RBBP5. The overall rates were derived by fitting the decrease in the relative intensity of the unmodified H3 peptide peaks in MALDI–TOF mass spectra using one-phase exponential decay model [Lys4]t = [Lys4]0e−kt. Extended Data Figure 2 Interactions between MLL proteins and RBBP5–ASH2L a, GST pull-down assays showed direct interactions between MLL proteins and RBBP5–ASH2L. ASH2L C-terminal SPRY domain has been previously shown to interact with RBBP5. GST-fused ASH2LSPRY was incubated with full-length RBBP5 and different MLLSET proteins in the GST pull-down assay. Bound proteins were eluted and separated by SDS–PAGE. Three different salt concentration buffers were tested. b, Fluorescence polarization assay shows that MLL proteins can interact with RBBP5AS-ABM–ASH2LSPRY with different affinities. For MLL1, SET1A and SET1B, lower limits of the Kd values are reported because saturation of the binding could not be achieved in fluorescence polarization assays. c, GST–RBBP5 alone cannot pull down MLL proteins in the buffer with 300 mM NaCl. d, GST–ASH2LSPRY alone cannot pull down MLL proteins in the buffer with 300 mM NaCl. e, The RBBP5–ASH2L interaction is highly dependent on the salt concentration used in the assay. ITC measurements were carried out using ASH2LSPRY and RBBP5AS-ABM under buffer conditions with different salt concentrations. f, The requirement of WDR5 in methyltransferase activity of the MLL1 complex is sensitive to protein concentration. MLL1 (5 μM) could be markedly stimulated by equal amounts of ASH2L–RBBP5, and WDR5 had a minor stimulation effect. g, HKMT activities of RBBP5–MLL1 fusion proteins in the presence of ASH2L or ASH2L and WDR5. Full-length RBBP5 was fused to MLL1 (residues 3754–3969) with a GGSGGS linker. The addition of ASH2L substantially stimulated the HKMT activity of the RBBP5–MLL1 fusion protein, while further addition of WDR5 only had a marginal effect. Extended Data Figure 3 The overall structure of the MLL3SET–RBBP5AS-ABM–ASH2LSPRY–H3 complex a, The overall structure of the MLL3SET–RBBP5AS-ABM–ASH2LSPRY–H3 complex in cartoon diagram. ASH2L is in yellow-orange, RBBP5 in cyan, MLL3SET in salmon, the H3 peptide in yellow, and cofactor product (AdoHcy) in blue. The electron density (2Fo − Fc) map, contoured at 1σ, is shown for the RBBP5 fragment, the H3 peptide and AdoHcy. b, The electron density (2Fo − Fc) map, contoured at 1σ, is shown around the substrate-binding channel. There are two complexes in one asymmetric unit. One complex has clear electron density for H3 residues 2–7 (left), while the other exhibits no extra density in the substrate channel (right). c, Cofactor interaction network. Residues important for the AdoHcy–MLL3SET interaction are shown in stick models. Hydrogen bonds are indicated by dashed magenta lines. d, The space-filling model of MLL3SET shows that AdoHcy and H3 bind to the opposite surfaces on MLL3SET. The distance between the sulfur atom and ε-amine of Lys4 is shown. e, The binding interface between MLL3SET and H3. f, MLL3SET is in surface representation and coloured according to its electrostatic potential. Thr3 of H3 sits snugly on a shallow hydrophobic depression, which cannot accommodate residues with a large side chain. Arg2 is involved in electrostatic interactions with MLL3SET. g, Sequence alignment of histone methylation sites. Residues are numbered relative to the target lysine. Because only the Lys4 site of H3 contains a large basic residue and a small residue occupying the −2 and −1 positions respectively, Arg(−2) and Thr(−1) define the substrate specificity of MLL complexes. Extended Data Figure 4 Sequence alignment of MLL homologues from human, Drosophila and Saccharomyces cerevisiae The WDR5-interacting motif (WIN) and SET domain are aligned. Secondary structure assignments based on the MLL3 structure are shown as cylinders (α-helices) and arrows (β-strands) above the sequences. The WIN motif is coloured in blue, SET-N in green, SET-I in orange, SET-C in purple and post-SET in magenta. Conserved residues important for RBBP5–ASH2L interactions are highlighted in magenta. Four Zn-binding cysteine residues are highlighted in pale yellow. Residues important for cofactor binding are in brown; residues important for substrate H3 binding and maintenance of the active centre are in grey. Two glycine residues, which serve as the hinge for SET-I motif rotation, are indicated by blue dots. The residues with the corresponding MLL4 mutations found in Kabuki syndrome and non-Hodgkin lymphoma are indicated by stars. Extended Data Figure 5 The ternary interaction interface among MLL, RBBP5 and ASH2L a, Mutations of RBBP5 and ASH2L disrupted the interaction between ASH2LSPRY and RBBP5. Left, GST–RBBP5330–381 was used to pull down ASH2LSPRY and its mutants. Right, GST–ASH2LSPRY was used to pull down full-length RBBP5 and its mutants. Several control mutations (such as ASH2L(Q354A) and RBBP5(E347A)), which are not on the RBBP5–ASH2L interface, did not affect the interaction between ASH2L and RBBP5. b, ASH2L and RBBP5 mutants that disrupted the RBBP5–ASH2L interaction decreased the HKMT activities of the MLL3 complex. The activities of the mutant proteins are normalized to the wild-type MLL3–RBBP5–ASH2L complex. Mean ± s.d. (n = 3) are shown. c, Mutations of RBBP5AS residues decreased the HKMT activity of the MLL3 complex. d, Representative gel-filtration profiles for MLL and MLL mutant proteins indicate MLL mutant proteins have a similar fold to wild-type protein. e, GST–MLL3SET was used to pull down full-length RBBP5, ASH2LSPRY and their mutants. Mutations of RBBP5 Glu347 and ASH2L Gln354 in the ternary interface impaired the interaction with MLL3SET. Mutations of RBBP5AS residues (Phe336Ala, Glu338Ala/Leu339Ala) also decreased the interaction with MLL3SET to different degrees. f, RBBP5(Glu347Ala) and ASH2L(Gln354Ala) compromised the HKMT activities of all MLL complexes, indicating that RBBP5–ASH2L regulates MLL family proteins through the same mechanism. Activities of mutant complexes are normalized to the activity of wild-type MLLSET–RBBP5–ASH2L, setting at 100%. Mean ± s.d. (n = 3) are shown. Extended Data Figure 6 Activation mechanism of MLL proteins a, The structure of MLL3SET is shown in cartoon diagram. The electron density (2Fo − Fc) maps (contoured at 1σ) of AdoHcy are shown. b, The structure of MLL1SETM is shown in cartoon diagram. The electron density (2Fo − Fc) maps (contoured at 1σ) of AdoHcy are shown. c, Structural comparison of MLL1SET (PDB 2W5Y), MLL1SETM, MLL3SET and MLL4SET (PDB 4Z4P) suggests that the SET-I motif is intrinsically flexible, and can be captured in different configurations by crystallization. There are two highly conserved glycine residues serving as hinge points that connect the SET-I motif to the rest of MLLSET. The rotation of helix αB in the SET-I motif refers to an axis defined by the two hinge points of SET-I as indicated. d, Dynamic cross-correlation matrix for motions of all Cα atoms in apo MLL3SET and MLL3SET in the M3RA complex over the course of the simulation. The right panel shows enlarged cross-correlation maps of the SET-I motif. e, Dynamic cross-correlation matrix for motions of all Cα atoms in apo MLL1SETM and MLL1SETM in the M1MRA complex over the course of the simulation. The right panel shows enlarged cross-correlation maps of the SET-I motif. f, The most highly correlated residues of the SET-I motif by molecular dynamics simulation are indicated by red lines. Left panel is for apo MLL1SETM and right panel is for MLL1SETM in the MLL1MRA complex. Red lines are connected Cα atoms for pairs of residues with calculated correlation coefficients greater than 0.55. Extended Data Figure 7 Association of RBBP5–ASH2L increases the binding affinities of MLL to cofactor and substrate peptide a, ITC measurement of interactions of AdoMet with MLL3SET alone (blue) and the M3RA complex (red). The insets show the ITC titration data. b, Equilibrium dissociation constants between cofactor and MLL proteins obtained from ITC measurements. c, Fluorescence polarization assay shows that RBBP5–ASH2L increases the binding affinity between MLL3 and the H3 peptide substrate. d, Molecular dynamics simulation show dynamics of the cofactor binding pocket. Top, the distance between AdoHcy and Tyr4825; bottom, the distance between Arg4845 and Tyr4825. These distances are almost fixed in the M3RA complex, while the distances in apo MLL3 have large variations, explaining why the MLL3 complex has a higher binding affinity to cofactor than apo MLL3 does. e, The potentials of mean force for the methyl transfer reaction along the reaction coordinate range from −1.5 to 2.0 Å with an interval of 0.1 Å. It clearly shows that the MLLSET–RBBP5–ASH2L complex is more energetic favourable for the methyl transfer reaction than MLLSET alone. f, The space-filling surface model shows that the Ly4H3 binding channel exhibits open and closed conformations in the M3RA and M3RA–H3 structures. Extended Data Figure 8 A conserved activation mechanism for SET-domain-containing HKMTs a, Structural comparison of MLL3SET in the M3RA–H3–AdoHcy complex, and the SET domains of CLR4 (PDB 1MVH), DIM-5 (PDB 1PEG), EZH2 (PDB 5CH1), ASH1 (PDB 3OPE) and NSD1 (PDB 3OOI). Histone H3 peptide and AdoHcy in the CLR4 structure were modelled based on the M3RA–H3–AdoHcy complex structure. RBBP5AS and the corresponding activation segments in these proteins are almost identical in overall conformation (coloured in cyan). The recently reported EZH2 complex structure also revealed such an activation segment. Most notably, an aromatic residue (shown as stick model), equivalent to Phe336 in RBBP5, stacks with another two aromatic residues to form an aromatic cage to sandwich a conserved arginine. Another conserved hydrophobic residue (shown as stick model) is also important for the stable association between the activation segment and the SET-I motif. b, Sequence alignment of the activation segments of RBBP5 and several representative HKMTs, including members from the SUV39 and SET2 families. c, Gel-filtration profiles and SDS–PAGE for DIM-5 and DIM-5ΔAS show that activation segment does not affect protein folding. DIM-5ΔAS denotes DIM-5 (residues 51–319) that does not contain the activation segment. d, HKMT activities of DIM-5 and its mutants. Activities of mutant proteins are normalized to the activity of the wild-type protein setting at 100%. Mean ± s.d. (n = 3) are shown. Extended Data Table 1 Data collection and refinement statistics for MLL3SET and the MLL3SET–RBBP5AS-ABM–ASH2LSPRY complex MLL3SET Native† MLL3SET Peak (Zn-SAD)† MLL3SET-RBBP5AS-ABM-ASH2LSPRY† Data collection Space group P4132 P4132 P21212 Cell dimensions a, b, c (Å) 129.056 129.323 80.342,236.076,44.416 α, β, γ (°) 90 90 90, 90, 90 Wavelength (Å) 0.9793 1.2818 1.1272 Resolution (Å) 50-2.8 50-3.4 100-2.4 Rmerge 0.135(0.530)* 0.204(0.563) 0.110 (0.654) l/σl 31.6(7.0) 31.0(12.3) 32.6 (5.1) Completeness (%) 99.9(100) 99.9(100) 99.9(100) Redundancy 10.1(10.6) 13.3(13.9) 7.1 (7.3) Refinement Resolution (Å) 38.9-2.8 44.4-2.4 No. reflections 9521 33458 Rwork/Rfree (%) 18.0/22.9 18.0/22.8 No. atoms Protein 1198 5570 Ligand 27 54 Water 52 220 B-factors (Å2) Protein 36.22 46.65 Ligand 33.13 52.53 Water 32.30 41.28 R.m.s deviations Bond lengths (Å) 0.011 0.003 Bond angles (°) 1.039 0.654 * Values in parentheses are for the highest-resolution shell. † The data are collected from one crystal. Extended Data Table 2 Data collection and refinement statistics for MLL1SETN3861I/Q3867L and MLL1SETN3861I/Q3867L-RBBP5AS-ABM-ASH2LSPRY complex MLL1SETN3861I/Q3867L† MLL1SETN3861I/Q3867L-RBBP5AS-ABM-† Data collection Space group P3221 C2 Cell dimensions a, b, c (Å) 54.547,54.574,104.656 74.966,44,410, 117.792 α, β, γ (°) 90,90,122 90,106.157,90 Wavelength (Å) 0.9785 0.9792 Resolution (Å) 50-1.8 50-1.9 Rmerge 0.090(0.421)* 0.158 (0.548) l/σl 36.5(4.1) 11.3(2.4) Completeness (%) 100(100) 99.7(99.8) Redundancy 9.6(9.9) 3.6 (3.3) Refinement Resolution (Å) 28.1-1.8 37.5-1.9 No. reflections 17284 29678 Rwork/Rfree (%) 20.2/23.6 16.6/21.3 No. atoms Protein 1156 2804 Ligand 27 27 Water 142 361 B-factors (Å2) Protein 45.34 20.31 Ligand 33.19 27.14 Water 45.77 29.99 R.m.s deviations Bond lengths (Å) 0.011 0.008 Bond angles (°) 1.105 1.044 † The data are collected from one crystal. * Values in parentheses are for highest-resolution shell. We thank staffs of beamlines BL18U, BL19U1 and 17U at the National Center for Protein Sciences Shanghai and Shanghai Synchrotron Radiation Facility for their assistance in data collection. We are grateful to protein expression, protein purification and mass spectrometry facilities at the National Center for Protein Sciences Shanghai for their instrument support and technical assistance. This work was supported by grants from the Strategic Priority Research Program of the Chinese Academy of Sciences (XDB08010201 to M.L. and Y.C., XDB08030302 to C.T.), the Ministry of Science and Technology of China (2013CB910402 to M.L., 2013CB910401 to Y.C., 2012AA01A305 and 2012CB721002 to G.L., 2011CB910400 to C.T.), the National Science and Technology Major Project ‘Key New Drug Creation and Manufacturing Program’ of China (2014ZX09507002-005 to M.L.), the National Natural Science Foundation of China (31330040 to M.L., 31470737 to Y.C., and 91430110 to G.L.), the Basic Research Project of Shanghai Science and Technology Commission (14JC1407200 to Y.C.), the National Institutes of Health (R01 GM082856 to Y.D.), and Fundamental Research for the Central University (WK2340000064 to C.T.). Y.C. is a recipient of the Thousand Young Talents Program of the Chinese government. Figure 1 RBBP5–ASH2L interacts and activates MLL proteins a, Domain organization of human ASH2L, RBBP5, WDR5 and MLL proteins. Only the C-terminal domain of MLL is shown. DBM, DPY30 binding motif; PHD-WH, plant homeodomain-winged helix. Shaded areas denote the interacting domains among these proteins. b, The normalized HKMT activities determined by a 3H-methyl-incorporation assay. Mean ± s.d. (n = 3) are shown. c, GST pull-down assay shows MLL proteins directly interact with the RBBP5–ASH2LSPRY heterodimer. d, The normalized HKMT assays revealed that an activation segment of RBBP5 (residues 330–344) is crucial for the stimulation of MLL3 activity. FL, full-length. Mean ± s.d. (n = 3) are shown. Figure 2 Crystal structure of the M3RA complex a, The crystal structure of MLL3SET–RBBP5AS-ABM–ASH2LSPRY in complex with cofactor product AdoHcy and the H3 peptide. b, MLL3SET shares the conserved features of SET-N, SET-I, SET-C and post-SET motifs. c, Comparison of the active centre of MLL3 and DIM-5 (PDB accession 1PEG) complex structures. d, The AdoHcy binding pocket in MLL3SET. e, The substrate H3 binding channel. Hydrogen bonds are indicated by dashed magenta lines; purple sphere denotes a water molecule. Figure 3 Interfaces among MLL3SET, RBBP5AS-ABM and ASH2LSPRY a, Detailed view of the ASH2LSPRY–RBBP5ABM interface. b, The interface between MLL3SET and RBBP5AS. c, MLL3 Arg4806 forms an extensive salt-bridge and hydrogen-bonding network with ASH2L and RBBP5. d, Mutations of the conserved arginine in MLL proteins disrupt interactions with RBBP5–ASH2L, as shown by the GST pull-down assay in 300 mM NaCl buffer. e, Arginine mutations impair the HKMT activities of MLL family proteins. Activities of all complexes are normalized to the activity of wild-type MLL1–WDR5–RBBP5–ASH2L, and shown as mean ± s.d. (n = 3). Figure 4 Difference between MLL1 and other MLL proteins a, Sequence alignment of the RBBP5–ASH2L binding fragments from MLL family proteins. Two key residues that explain the RBBP5–ASH2L binding affinity difference between MLL1 and other MLL proteins are indicated by A and B sites. d, Drosophila; h, human. RA denotes RBBP5–ASH2L. b, The MLL3–RBBP5 interface around MLL3 Val4809. MLL1 Gln3867 (grey) cannot fit into this pocket. c, The MLL3–RBBP5 interface around MLL3 Thr4803, which is not compatible with MLL1 Asn3861 (grey). d, GST pull-down assay for the interactions of RBBP5–ASH2L with MLL1SET and its mutants. e, The normalized HKMT activities of MLL1WT and MLL1N3861I/Q3867L in the presence of full-length RBBP5–ASH2L and WDR5–RBBP5–ASH2L. Mean ± s.d. (n = 3) are shown. f, The overall structure of the M1MRA complex. g, Superposition of the structures of M1MRA and M3RA shows conserved interfaces between MLLSET and RBBP5–ASH2L. Figure 5 Activation mechanism of MLL proteins by RBBP5–ASH2L a, Structural comparison of the apo MLL3SET and MLL3SET in the M3RA complex. The structures are superimposed according to AdoHcy. b, Structural comparison of the apo MLL1SETM and MLL1SETM in the M1MRA complex. c, One-dimensional 19F-NMR measurements of MLL3SET with substitution of F4827tfmF (top) or Y4762tfmF (bottom) in the absence or presence of RBBP5–ASH2L. The locations of these two residues on MLLSET are shown. tfmF, l-4-trifluoromethylphenylalanine. d, Root mean square fluctuation (RMSF) of the SET domains in apo MLL3SET (black line) and in the M3RA complex (red line). e, RBBP5 Phe336 together with MLL3 Arg4845, Tyr4846 and Tyr4825 maintain a configuration that favours cofactor binding. f, The most highly correlated residues (correlation coefficients greater than 0.55) of SET-I in molecular dynamics simulation are indicated by red lines. g, Structural superimposition of the M3RA and M3RA–H3 complexes by the SET-I motifs highlights the local rearrangement of loop LB5. h, A working model for the activation of MLL family methyltransferases. Online Content Methods, along with any additional Extended Data display items and Source Data, are available in the online version of the paper; references unique to these sections appear only in the online paper. Supplementary Information is available in the online version of the paper. Author Contributions M.L. and Y.C. conceived and supervised the project. M.L. and Y.D. initiated the project. Y.L., J.H., C.H. and Y.C. purified the proteins, performed crystallization and determined the crystal structure. Y.L., J.H., F.C., C.H., J.W., Y.W. and Y.C. performed the biochemical assays. Z.L., P.S. and C.T. performed 19F-NMR experiments. S.L. and J.Z. performed normal mode analysis. Y.Z., L.C. and G.L. performed molecular dynamics and QM/MM simulation. D.L. and Y.D. contributed to manuscript preparation. G.L., Y.C. and M.L. analysed the data and wrote the manuscript. The atomic coordinates have been deposited in the Protein Data Bank (PDB) under the following accessions: 5F59 (MLL3SET), 5F6K (MLL3SET–RBBP5AS-ABM–ASH2LSPRY), 5F5E ( MLL1SETN3861I/Q3867L)) and 5F6L ( MLL1SETN3861I/Q3867L-RBBP5AS-ABM-ASH2LSPRY). The authors declare no competing financial interests. Readers are welcome to comment on the online version of the paper. 1 Barski A High-resolution profiling of histone methylations in the human genome Cell 129 823 837 2007 17512414 2 Shilatifard A Molecular implementation and physiological roles for histone H3 lysine 4 (H3K4) methylation Curr Opin Cell Biol 20 341 348 2008 18508253 3 Ruthenburg AJ Allis CD Wysocka J Methylation of lysine 4 on histone H3: intricacy of writing and reading a single epigenetic mark Mol Cell 25 15 30 2007 17218268 4 Ansari KI Mandal SS Mixed lineage leukemia: roles in gene expression, hormone signaling and mRNA processing FEBS J 277 1790 1804 2010 20236313 5 Wang P Global analysis of H3K4 methylation defines MLL family member targets and points to a role for MLL1-mediated H3K4 methylation in the regulation of transcriptional initiation by RNA polymerase II Mol Cell Biol 29 6074 6085 2009 19703992 6 Krivtsov AV Armstrong SA MLL translocations, histone modifications and leukaemia stem-cell development Nature Rev Cancer 7 823 833 2007 17957188 7 Ansari KI Mishra BP Mandal SS MLL histone methylases in gene expression, hormone signaling and cell cycle Front Biosci 14 3483 3495 2009 8 Parsons DW The genetic landscape of the childhood cancer medulloblastoma Science 331 435 439 2011 21163964 9 Morin RD Frequent mutation of histone-modifying genes in non-Hodgkin lymphoma Nature 476 298 303 2011 21796119 10 Li Y A mutation screen in patients with Kabuki syndrome Hum Genet 130 715 724 2011 21607748 11 Pleasance ED A small-cell lung cancer genome with complex signatures of tobacco exposure Nature 463 184 190 2010 20016488 12 Dalgliesh GL Systematic sequencing of renal carcinoma reveals inactivation of histone modifying genes Nature 463 360 363 2010 20054297 13 Dou Y Regulation of MLL1 H3K4 methyltransferase activity by its core components Nature Struct Mol Biol 13 713 719 2006 16878130 14 Patel A Dharmarajan V Vought VE Cosgrove MS On the mechanism of multiple lysine methylation by the human mixed lineage leukemia protein-1 (MLL1) core complex J Biol Chem 284 24242 24256 2009 19556245 15 Southall SM Wong PS Odho Z Roe SM Wilson JR Structural basis for the requirement of additional factors for MLL1 SET domain activity and recognition of epigenetic marks Mol Cell 33 181 191 2009 19187761 16 Steward MM Molecular regulation of H3K4 trimethylation by ASH2L, a shared subunit of MLL complexes Nature Struct Mol Biol 13 852 854 2006 16892064 17 Dou Y Hess JL Mechanisms of transcriptional regulation by MLL and its disruption in acute leukemia Int J Hematol 87 10 18 2008 18224408 18 Wysocka J WDR5 associates with histone H3 methylated at K4 and is essential for H3 K4 methylation and vertebrate development Cell 121 859 872 2005 15960974 19 Cao F An Ash2L/RbBP5 heterodimer stimulates the MLL1 methyltransferase activity through coordinated substrate interactions with the MLL1 SET domain PLoS ONE 5 e14102 2010 21124902 20 Shinsky SA Monteith KE Viggiano S Cosgrove MS Biochemical reconstitution and phylogenetic comparison of human SET1 family core complexes involved in histone methylation J Biol Chem 290 6361 6375 2015 25561738 21 Shinsky SA Cosgrove MS Unique role of the WD-40 repeat protein 5 (WDR5) subunit within the mixed lineage leukemia 3 (MLL3) histone methyltransferase complex J Biol Chem 290 25819 25833 2015 26324722 22 Zhang P Lee H Brunzelle JS Couture JF The plasticity of WDR5 peptide-binding cleft enables the binding of the SET1 family of histone methyltransferases Nucleic Acids Res 40 4237 4246 2012 22266653 23 Zhang P A phosphorylation switch on RBBP5 regulates histone H3 Lys4 methylation Genes Dev 29 123 128 2015 25593305 24 Chen Y Cao F Wan B Dou Y Lei M Structure of the SPRY domain of human Ash2L and its interactions with RbBP5 and DPY30 Cell Res 22 598 602 2012 22231628 25 Odho Z Southall SM Wilson JR Characterization of a novel WDR5-binding site that recruits RbBP5 through a conserved motif to enhance methylation of histone H3 lysine 4 by mixed lineage leukemia protein-1 J Biol Chem 285 32967 32976 2010 20716525 26 Zhang Y evolving catalytic properties of the MLL family set domain Structure 23 1921 1933 2015 26320581 27 Cheng X Collins RE Zhang X Structural and sequence motifs of protein (histone) methylation enzymes Annu Rev Biophys Biomol Struct 34 267 294 2005 15869391 28 Zhang X Structural basis for the product specificity of histone lysine methyltransferases Mol Cell 12 177 185 2003 12887903 29 Sirinupong N Brunzelle J Doko E Yang Z Structural insights into the autoinhibition and posttranslational activation of histone methyltransferase SmyD3 J Mol Biol 406 149 159 2011 21167177 30 Wang KC A long noncoding RNA maintains active chromatin to coordinate homeotic gene expression Nature 472 120 124 2011 21423168 31 Ang YS Wdr5 mediates self-renewal and reprogramming via the embryonic stem cell core transcriptional network Cell 145 183 197 2011 21477851 32 Gan Q WD repeat-containing protein 5, a ubiquitously expressed histone methyltransferase adaptor protein, regulates smooth muscle cell-selective gene activation through interaction with pituitary homeobox 2 J Biol Chem 286 21853 21864 2011 21531708 33 Dou Y Physical association and coordinate function of the H3 K4 methyltransferase MLL1 and the H4 K16 acetyltransferase MOF Cell 121 873 885 2005 15960975 34 Thompson BA Tremblay V Lin G Bochar DA CHD8 is an ATP-dependent chromatin remodeling factor that regulates β-catenin target genes Mol Cell Biol 28 3894 3904 2008 18378692 35 Otwinowski Z Minor W Processing of X-ray diffraction data collected in oscillation mode Methods Enzymol 276 307 326 1997 36 Vonrhein C Blanc E Roversi P Bricogne G Automated structure solution with autoSHARP Methods Mol Biol 364 215 230 2007 17172768 37 Adams PD PHENIX: building new software for automated crystallographic structure determination Acta Crystallogr D 58 1948 1954 2002 12393927 38 Emsley P Cowtan K Coot: model-building tools for molecular graphics Acta Crystallogr D 60 2126 2132 2004 15572765 39 Storoni LC McCoy AJ Read RJ Likelihood-enhanced fast rotation functions Acta Crystallogr D 60 432 438 2004 14993666 40 Shi P Site-specific 19F NMR chemical shift and side chain relaxation analysis of a membrane protein labeled with an unnatural amino acid Protein Sci 20 224 228 2011 21080424 41 Lindahl E Azuara C Koehl P Delarue M NOMAD-Ref: visualization, deformation and refinement of macromolecular structures based on all-atom normal mode analysis Nucleic Acids Res 34 W52 W56 2006 16845062
PMC005xxxxxx/PMC5125620.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 7600130 5844 Neurosci Lett Neurosci. Lett. Neuroscience letters 0304-3940 1872-7972 26971700 5125620 10.1016/j.neulet.2016.03.015 NIHMS831555 Article Distinct contributions of reactive oxygen species in amygdala to bee venom-induced spontaneous pain-related behaviors Lu Yun-Fei 12 Neugebauer Volker 4* Chen Jun 123 Li Zhen 12* 1 Institute for Biomedical Sciences of Pain and Institute for Functional Brain Disorders, Tangdu Hospital, The Fourth Military Medical University, Xi’an 710038, PR China 2 Key Laboratory of Brain Stress and Behavior, PLA, Xi’an 710038, PR China 3 Beijing Institute for Brain Disorders, Beijing 100069, P.R. China 4 Department of Pharmacology and Neuroscience, School of Medicine, Texas Tech University Health Sciences Center (TTUHSC), Lubbock, TX, USA * Corresponding author at: Department of Pharmacology and Neuroscience, School of Medicine, Texas Tech University Health Sciences Center (TTUHSC), Lubbock, TX, USA, Phone: (806) 743-3880, [email protected] (V. Neugebauer); Institute for Biomedical Sciences of Pain and Institute for Functional Brain Disorders, Tangdu Hospital, The Fourth Military Medical University, #569 Xinsi Road, Baqiao, Xi’an 710038, P.R. China. Tel: +86-29-84778642, Fax: +86-29-84777945; [email protected] (Z. Li) 23 11 2016 10 3 2016 21 4 2016 21 4 2017 619 6872 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Reactive oxygen species (ROS), such as superoxide and hydrogen peroxide, play essential roles in physiological plasticity and are also involved in the pathogenesis of persistent pain. Roles of peripheral and spinal ROS in pain have been well established, but much less is known about ROS in the amygdala, a brain region that plays an important role in pain modulation. The present study explored the contribution of ROS in the amygdala to bee venom (BV)-induced pain behaviors. Our data show that the amygdala is activated following subcutaneous BV injection into the left hindpaw, which is reflected in the increased number of c-Fos positive cells in the central and basolateral amygdala nuclei in the right hemisphere. Stereotaxic administration of a ROS scavenger (tempol, 10 mM), NADPH oxidase inhibitor (baicalein, 5 mM) or lipoxygenase inhibitor (apocynin, 10 mM) into the right amygdala attenuated the BV-induced spontaneous licking and lifting behaviors, but had no effect on BV-induced paw flinch reflexes. Our study provides further evidence for the involvement of the amygdala in nociceptive processing and pain behaviors, and that ROS in amygdala may be a potential target for treatment strategies to inhibit pain. Reactive oxygen species bee venom pain amygdala 1. Introduction Reactive oxygen species (ROS), such as superoxide and hydrogen peroxide, play essential roles in physiological plasticity [10,17], whereas under pathological conditions, the formation of ROS has been shown to be critical in apoptosis, stroke pathology, spinal cord injury and neurodegenerative disorders [6,22]. Moreover, ROS are also implicated in pathological pain states, including neuropathic pain, inflammatory pain and visceral pain [9,11,13,16,19,32,33]. Whereas previous studies mainly focused on the roles of peripheral and spinal ROS in the pathogenesis of inflammatory and neuropathic pain [7,16,33,37], the pain-related function of supraspinal ROS is still an understudied area [11,13,19,21]. As a part of the limbic system, the amygdala has traditionally been associated with negative emotions such as fear [14] and plays an important role in emotional responses to pain and in pain modulation [24–26]. The amygdala is composed of lateral (LA), basolateral (BLA) and central (CeA) nuclei [31], which serve distinct functions in the processing of pain-related information [24]. LA-BLA receive and integrate polymodal sensory, including nociceptive, information, which is transmitted to the amygdala output region (CeA) that also receives a nociceptive teaching signal from the brainstem parabrachial area [24]. These subnuclei of amygdala have been shown to be differentially activated in the formalin test and the acetic acid-induced visceral pain model based on different patterns of c-fos mRNA expression [23]. Our previous work showed that ROS in amygdala are involved in experimental models of visceral and arthritis pain [11,13,19]. However, the role of ROS in the amygdala in other experimental pain models remains to be determined. The bee venom (BV) test is a well-established experimental pain model. It is produced by subcutaneous injection of a BV solution, which can induce long-lasting spontaneous pain-related behaviors such as the spinally mediated paw flinch reflex and the supraspinally mediated paw licking and lifting behaviors [4,5,18,30,38]. The present study first mapped c-Fos changes in different subnuclei of amygdala following BV injection and then tested pharmacological agents to investigate the roles of ROS in the amygdala in BV-induced spontaneous pain-related behaviors. 2. Materials and Methods 2.1. Animals All experiments were conducted on male albino Sprague-Dawley rats (weighing 180–220g, 8–9 weeks old) purchased from the Laboratory Animal Center of Fourth Military Medical University (FMMU). The animals were housed in groups of 4–6 and maintained under standard conditions (12 h dark/light circle, temperature 22–26°C, air humidity 40–60%) with food and water available ad libitum. The experimental protocols were approved by the Institutional Animal Care and Use Committee of FMMU and animals were maintained and cared in line with EC Directive 86/609/EEC and the guidelines set forth by the International Association for the Study of Pain [39]. Every effort was made to minimize the number and suffering of the animals. 2.2. Immunohistochemistry Two hours after receiving subcutaneous BV (0.2 mg/50 μl, Sigma, USA) or physiological saline injection into the left hindpaw, rats were deeply anesthetized with sodium pentobarbital (50 mg/kg, intraperitoneal) and perfused transcardially with physiological saline followed by 4% paraformaldehyde. The whole brain was removed, post-fixed in the same fixate for 4 h and then transferred to a 30% sucrose solution in 0.01M phosphate buffer for cryoprotection. Coronal brain sections (40 μm thick) containing the amygdala were cut on a cryostat (CM1900, Leica, Germany). Every fifth slice was collected, yielding about 10–15 sections per rat. Brain slices were washed in 0.01 M phosphate-buffered saline (PBS) and incubated in 3% hydrogenperoxide for 10 min, and then incubated for 1 h in 1% bovine serum and 0.2% Triton X-100 in 0.01 M PBS. Sections were incubated overnight with rabbit anti-c-Fos polycolonal antibody (1:200, Cell Signaling). Sections were then washed and subjected to incubation with biotinylated goat anti-rabbit secondary antibody (1:200, ZSGB-Bio) for 3 h. Sections were washed again and incubated in avidin-biotin complex (1:200, Sigma) for 3 h. After several rinses in PBS, the immunostaining reactions were developed with an ABC kit (ZSGB-Bio, P.R. China). Reactions were stopped by repeated washes in PBS. Brain sections were mounted on slides and coverslipped. The number of c-Fos positive cells in CeA and BLA was counted with Image-Pro Plus digitizing software (Olympus, Japan). 2.3. Intra-amygdala drug application The rat was anesthetized with pentobarbital sodium (50 mg/kg, i.p.) and placed in a stereotaxic apparatus. A craniotomy over the right amygdala was performed and a guide cannula (outer diameter: 0.64 mm; inner diameter: 0.45 mm) was implanted on the dorsal margin of the CeA, using the following coordinates (in mm): 2.5 caudal to bregma; 4.0 lateral to midline; depth, 7.5. The cannula was fixed to the skull with dental acrylic. The behavioral tests were carried out 5 days after the surgery. For drug application, a microinjection tubing probe was inserted through the guide cannula so that the probe protruded by 1 mm. The probe was connected to a PE-10 polyethylene tube filled with dissolved drug solution. The drug solution was pushed into the brain tissue by injecting 1μl air into the polyethylene tube with a microsyringe. Then the polyethylene tube was heat sealed and the probe was left in place for 10 min before the behavioral tests. The following drugs were used: Tempol, baicalein, apocynin (Tocris). Drugs were dissolved in DMSO (30% in deionized water) and diluted in ACSF (containing [in mM] 117 NaCl, 4.7 KCl, 1.2 NaH2PO4, 2.5 CaCl2, 1.2 MgCl2, 25 NaHCO3 and 11 glucose) at 1:100 to their final concentration for injection into the amygdala. 2.4. Histology At the end of behavioral tests, all rats that received intracerebral cannulation were perfused as previously described above (see “Immunohistochemistry”). Brain sections containing the amygdala were sectioned at 20 μm and mounted on gel-coated slides. Brain sections were stained with 0.1% cresyl violet for identification of the tip of the cannula. 2.5. Quantitative measurement of pain-related behaviors A 30 × 30 × 30 cm transparent plastic box with transparent glass floor was placed on a supporting frame 30 cm high above the experimental table. The animals were habituated to the environment before the initiation of the experiment. Following s.c. BV injection into the left hindpaw, the number of spontaneous paw flinches, a spinally mediated reflex [4,18,38], and the lifting and licking time, which are supraspinally mediated [30], were counted for each 5-minute time block for 60 minutes. 2.6. Statistical analysis All data were expressed as mean ± SEM. One way ANOVA (post-hoc Fisher’s least significant difference) and two-way ANOVA with Bonferroni post-hoc tests were used where appropriate. P < 0.05 was chosen as the criterion for statistical significance. 3. Results 3.1. BV-induced c-Fos expression in the amygdala Two hours after s.c. saline injection into the left hindpaw, there were scatter c-Fos positive cells in the amygdala (CeA and BLA) (Fig. 1), with no difference of the number of c-Fos positive cells between the right and left amygdala. Importantly, two hours after s.c. BV injection into the left hindpaw, the number of c-Fos positive cells increased significantly in the right BLA (F1, 44 = 26.008, P < 0.001) and CeA (F1, 44 = 21.477, P < 0.001) compared to that received s.c. saline injection, and there were more c-Fos positive cells in the right amygdala (BLA, F1, 44 = 31.941, P < 0.001; CeA, F1, 44 = 27.685, P < 0.001) than that in the left. Increased c-Fos positive cells in CeA were mainly localized in the laterocapsular part (Fig. 1), which has been termed the “nociceptive amygdala” [26]. No difference of the number of c-Fos positive cells was detected between the right and left amygdala. 3.2. Effects of a ROS scavenger, NADPH oxidase inhibitor and lipoxygenase inhibitor on BV-induced pain-related behaviors Consistent with our previous findings [5], s.c. BV injection resulted in paw flinch reflexes, which lasted about 60 minutes, and paw licking or lifting behaviors, which lasted about 40 minutes. Stereotaxic administration of a ROS scavenger (tempol, 10 mM/1μl), which is a superoxide dismutase mimetic, or a 5- and 12-lipoxygenase inhibitor (baicalein, 5 mM/1μl), or a NADPH oxidase inhibitor (apocynin, 10 mM/1μl) into the right amygdala 10 min before s.c. BV injection had no effect on the BV-induced paw flinch reflex (F4, 34 = 0.177, P = 0.949; Fig. 3A and B). However, the paw lifting and licking behaviors were inhibited significantly (F4, 34 = 10.803, P < 0.001; Fig. 3C and D). 3.3. Histological analysis Histological analysis showed that the tips of the cannulae were mainly located within the region of the CeA, and with several tips located in BLA, which were on the lateral margin of CeA (Fig 2). 4. Discussion In the present study, we found that following subcutaneous injection of BV solution into the left hindpaw, the number of c-Fos positive cells increased significantly in the right CeA and BLA, indicating neuronal activity increases in the amygdala in this pain model. Stereotaxic administration of a ROS scavenger, NADPH oxidase inhibitor or lipoxygenase inhibitor into the right amygdala attenuated BV-induced supraspinally mediated paw lifting and licking behaviors, but had no effect on BV-induced spinal paw flinch reflexes. Our results show that ROS in amygdala play differential roles in BV-induced pain-related behaviors. The amygdala is composed by several subnuclei that serve distinct functions in the processing of pain-related information. The CeA receives nociceptive information directly from the spinal cord and the parabrachial area [8], whereas the LA receives polymodal sensory information from thalamus and cortex [28,29,31]. The LA-BLA network is believed to attach emotional significance to sensory information [28,29,31], and this highly processed information is then transmitted to the CeA, which modulates pain behaviors through projections to descending pain control center [26,27,35]. The involvement of amygdala in the processing of pain-related information has been previously demonstrated using electrophysiology [24–26], biochemistry [2], and neuronal activity markers such as c-Fos [36]. c-fos mRNA in the amygdala increased remarkably following noxious visceral or somatic stimulation [23] but suggested that the activation pattern in subnuclei of the amygdala can be different, depending on the type of the noxious stimulus. Intraplantar injection of formalin induced c-fos mRNA expression in the LA and BLA, but not CeA, although it should be noted that synaptic plasticity was measured in the CeA in the formalin pain model [1]. Intraperitoneal injection of acetic acid resulted in increased c-fos mRNA expression in the CeA rather than BLA [23]. In the present study, we found that BV injection increased the number of c-Fos positive cells in the BLA and CeA, particularly the laterocapsular part (CeLC), which further suggests that the activation of subnuclei of amygdala may be stimulus specific. Therefore, it is important to study roles of different subnuclei of amygdala in different experimental pain models. Interestingly, no differences of the expression of c-fos mRNA were detected between the left and right amygdala following intraplantar injection of formalin into the right hindpaw [23], whereas our study found more c-Fos positive cells in the CeA and BLA of the right than left amygdala. It is possible that the more pronounced increase in c-Fos expression in the right amygdala may be due to the side of BV injection, i.e., into the contralateral paw. Another possibility is that hemispheric lateralization of pain-related information processing in the amygdala, which has been suggested based on biochemical and behavioral [3] and electrophysiological [12] studies. Our pharmacological and behavioral experiments also confirm the right amygdala as an important site for antinociceptive drug effects. Although ROS play essential roles in physiological plasticity, ROS have also been implicated in pain mechanisms. Peripheral or spinal administration of exogenous ROS can exert pronociceptive effects, whereas the administration of ROS scavengers can have antinociceptive effects in pain states [15,34]. Moreover, intracerebral administration of ROS scavengers into the CeA inhibited visceral and somatosensory nociceptive responses of amygdala neurons induced by activation of metabotropic glutamate receptors [13,19] and attenuated visceral pain-related emotional responses [11]. Our present study further showed that intra-amygdala administration of a ROS scavenger, NADPH oxidase inhibitor or lipoxygenase inhibitor attenuated supraspinally mediated BV-induced lifting and licking behaviors. Consistent with our previous report that bilateral lesion of the amygdala exerts no effects on spinally mediated BV-induced paw flinch reflexes [20], the present study also found no evidence for the involvement of amygdala ROS in BV-induced spinal paw flinch reflexes. Taken together, ROS can contribute to the processing of pain-related information at different levels and inhibition of the synthesis of ROS or elimination of ROS with scavengers at different levels can inhibit pain-related behaviors that are mediated at different levels. It is possible that ROS needs to be targeted at different levels (systemically) to achieve maximum pain inhibition. In conclusion, our study provides further evidence for the involvement of the amygdala in pain processing and pain modulation and suggests that ROS in amygdala may be a useful target for developing treatments for pain. This work was supported by grants from the NSFC (31271184) to ZL and the National Basic Research Program of China (2013CB835103) to JC. Figure 1 c-Fos expression in amygdala following s.c. BV or saline injection Panels on the left show representative images of c-Fos immunoreactivity in the amygdala following s.c. BV or saline injection. Scale bars: 500 μm. Bar histograms on the right show quantitative analysis of the number of c-Fos positive cells in the CeA and BLA. *P < 0.001, BV compared to saline; ###P < 0.05, 0.001 left compared to right. Data are expressed as mean ± SEM. Figure 2 Drug application sites in the amygdala (A) Representative histological image of the drug application site. (B–E) Localizations of drug application sites for different compounds. Scale bars: 2 mm. Figure 3 Effects of ROS scavenger, NADPH oxidase inhibitor and lipoxygenase inhibitor on BV-induced pain-related behaviors (A, C) Time course of the effects of intra-amygdalar administration of a ROS scavenger (tempol), NADPH oxidase inhibitor (baicalein) or lipoxygenase inhibitor (apocynin) on BV-induced paw flinch reflex and licking and lifting behaviors. (B, D) Total number of paw flinches (B) and total duration of licking and lifting behaviors (D) based on data shown in A and C, respectively. *P < 0.001 compared to DMSO (vehicle control). Data are expressed as mean ± SEM. Bar histograms show averaged data and individual data points (open circles). Highlights The right amygdala can be activated by s.c. injection of BV into left hindpaw This activation is represented by increased c-Fos positive cells in right amygdala Intra-amygdala administration of ROS scavenger attenuated BV-induced flinch reflex NADPH oxidase and lipoxygenase inhibitor also attenuated BV-induced flinch reflex Conflict of interest: The authors declare they have no conflict of interest. Contributors: Y.F. and Z. carried out the experiments. Z., V. and J. drafted the manuscript. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. 1 Adedoyin MO Vicini S Neale JH Endogenous N-acetylaspartylglutamate (NAAG) inhibits synaptic plasticity/transmission in the amygdala in a mouse inflammatory pain model Mol Pain 6 2010 60 20860833 2 Carrasquillo Y Gereau RWt Activation of the extracellular signal-regulated kinase in the amygdala modulates pain perception J Neurosci 27 2007 1543 1551 17301163 3 Carrasquillo Y Gereau RWt Hemispheric lateralization of a molecular signal for pain modulation in the amygdala Mol Pain 4 2008 24 18573207 4 Chen J Lariviere WR The nociceptive and anti-nociceptive effects of bee venom injection and therapy: a double-edged sword Prog Neurobiol 92 2010 151 183 20558236 5 Chen J Luo C Li H Chen H Primary hyperalgesia to mechanical and heat stimuli following subcutaneous bee venom injection into the plantar surface of hindpaw in the conscious rat: a comparative study with the formalin test Pain 83 1999 67 76 10506673 6 Chinopoulos C Adam-Vizi V Calcium, mitochondria and oxidative stress in neuronal pathology. Novel aspects of an enduring theme FEBS J 273 2006 433 450 16420469 7 Gao X Kim HK Chung JM Chung K Reactive oxygen species (ROS) are involved in enhancement of NMDA-receptor phosphorylation in animal models of pain Pain 131 2007 262 271 17317010 8 Gauriau C Bernard JF A comparative reappraisal of projections from the superficial laminae of the dorsal horn in the rat: the forebrain J Comp Neurol 468 2004 24 56 14648689 9 Hassler SN Johnson KM Hulsebosch CE Reactive oxygen species and lipid peroxidation inhibitors reduce mechanical sensitivity in a chronic neuropathic pain model of spinal cord injury in rats J Neurochem 131 2014 413 417 25051888 10 Hu D Serrano F Oury TD Klann E Aging-dependent alterations in synaptic plasticity and memory in mice that overexpress extracellular superoxide dismutase J Neurosci 26 2006 3933 3941 16611809 11 Ji G Li Z Neugebauer V Reactive oxygen species mediate visceral pain-related amygdala plasticity and behaviors Pain 156 2015 825 836 25734993 12 Ji G Neugebauer V Hemispheric lateralization of pain processing by amygdala neurons J Neurophysiol 102 2009 2253 2264 19625541 13 Ji G Neugebauer V Reactive oxygen species are involved in group I mGluR-mediated facilitation of nociceptive processing in amygdala neurons J Neurophysiol 104 2010 218 229 20463194 14 Johansen JP Cain CK Ostroff LE LeDoux JE Molecular mechanisms of fear learning and memory Cell 147 2011 509 524 22036561 15 Keeble JE Bodkin JV Liang L Wodarski R Davies M Fernandes ES de Coelho CF Russell F Graepel R Muscara MN Malcangio M Brain SD Hydrogen peroxide is a novel mediator of inflammatory hyperalgesia, acting via transient receptor potential vanilloid 1-dependent and independent mechanisms Pain 141 2009 135 142 19059721 16 Kim HK Park SK Zhou JL Taglialatela G Chung K Coggeshall RE Chung JM Reactive oxygen species (ROS) play an important role in a rat model of neuropathic pain Pain 111 2004 116 124 15327815 17 Kishida KT Klann E Sources and targets of reactive oxygen species in synaptic plasticity and memory Antioxid Redox Signal 9 2007 233 244 17115936 18 Li KC Chen J Altered pain-related behaviors and spinal neuronal responses produced by s.c. injection of melittin in rats Neuroscience 126 2004 753 762 15183523 19 Li Z Ji G Neugebauer V Mitochondrial reactive oxygen species are activated by mGluR5 through IP3 and activate ERK and PKA to increase excitability of amygdala neurons and pain behavior J Neurosci 31 2011 1114 1127 21248136 20 Li Z Lu YF Li CL Wang Y Sun W He T Chen XF Wang XL Chen J Social interaction with a cagemate in pain facilitates subsequent spinal nociception via activation of the medial prefrontal cortex in rats Pain 155 2014 1253 1261 24699208 21 Little JW Chen Z Doyle T Porreca F Ghaffari M Bryant L Neumann WL Salvemini D Supraspinal peroxynitrite modulates pain signaling by suppressing the endogenous opioid pathway J Neurosci 32 2012 10797 10808 22875915 22 Maher P Schubert D Signaling by reactive oxygen species in the nervous system Cell Mol Life Sci 57 2000 1287 1305 11028919 23 Nakagawa T Katsuya A Tanimoto S Yamamoto J Yamauchi Y Minami M Satoh M Differential patterns of c-fos mRNA expression in the amygdaloid nuclei induced by chemical somatic and visceral noxious stimuli in rats Neurosci Lett 344 2003 197 200 12812839 24 Neugebauer V Amygdala pain mechanisms Handb Exp Pharmacol 227 2015 261 284 25846623 25 Neugebauer V Galhardo V Maione S Mackey SC Forebrain pain mechanisms Brain Res Rev 60 2009 226 242 19162070 26 Neugebauer V Li W Bird GC Han JS The amygdala and persistent pain Neuroscientist 10 2004 221 234 15155061 27 Ossipov MH Dussor GO Porreca F Central modulation of pain J Clin Invest 120 2010 3779 3787 21041960 28 Pape HC Pare D Plastic synaptic networks of the amygdala for the acquisition, expression, and extinction of conditioned fear Physiol Rev 90 2010 419 463 20393190 29 Phelps EA LeDoux JE Contributions of the amygdala to emotion processing: from animal models to human behavior Neuron 48 2005 175 187 16242399 30 Ren LY Lu ZM Liu MG Yu YQ Li Z Shang GW Chen J Distinct roles of the anterior cingulate cortex in spinal and supraspinal bee venom-induced pain behaviors Neuroscience 153 2008 268 278 18367341 31 Sah P Faber ES Lopez De Armentia M Power J The amygdaloid complex: anatomy and physiology Physiol Rev 83 2003 803 834 12843409 32 Salvemini D Little JW Doyle T Neumann WL Roles of reactive oxygen and nitrogen species in pain Free Radic Biol Med 51 2011 951 966 21277369 33 Schwartz ES Kim HY Wang J Lee I Klann E Chung JM Chung K Persistent pain is dependent on spinal mitochondrial antioxidant levels J Neurosci 29 2009 159 168 19129394 34 Schwartz ES Lee I Chung K Chung JM Oxidative stress in the spinal cord is an important contributor in capsaicin-induced mechanical secondary hyperalgesia in mice Pain 138 2008 514 524 18375065 35 Tracey I Mantyh PW The cerebral signature for pain perception and its modulation Neuron 55 2007 377 391 17678852 36 Traub RJ Silva E Gebhart GF Solodkin A Noxious colorectal distention induced-c-Fos protein in limbic brain structures in the rat Neurosci Lett 215 1996 165 168 8899739 37 Wang ZQ Porreca F Cuzzocrea S Galen K Lightfoot R Masini E Muscoli C Mollace V Ndengele M Ischiropoulos H Salvemini D A newly identified role for superoxide in inflammatory pain J Pharmacol Exp Ther 309 2004 869 878 14988418 38 You HJ Chen J Differential effects of subcutaneous injection of formalin and bee venom on responses of wide-dynamic-range neurons in spinal dorsal horn of the rat Eur J Pain 3 1999 177 180 39 Zimmermann M Ethical guidelines for investigations of experimental pain in conscious animals Pain 16 1983 109 110 6877845
PMC005xxxxxx/PMC5125629.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 0322116 6595 Prev Med Prev Med Preventive medicine 0091-7435 1096-0260 27717667 5125629 10.1016/j.ypmed.2016.10.004 NIHMS825277 Article Global tobacco prevention and control in relation to a cardiovascular health promotion and disease prevention framework: A narrative review Carroll Allison J. MS Labarthe Darwin R. MD, MPH, PhD Huffman Mark D. MD, MPH Hitsman Brian PhD * Department of Preventive Medicine, Northwestern University Feinberg School of Medicine, United States * Corresponding author at: Department of Preventive Medicine, Northwestern University Feinberg School of Medicine, 680 N. Lake Shore Drive, Suite 1400, Chicago, IL 60611, United States. [email protected] (B. Hitsman) 19 11 2016 4 10 2016 12 2016 01 12 2017 93 189197 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. The purpose of this review is to emphasize the role of tobacco prevention and control in cardiovascular health (CVH) promotion and cardiovascular disease (CVD) prevention, including the importance of these endpoints for measuring the full impact of tobacco-related policies, programs, and practices. In this review, we describe an overview of tobacco control interventions that have led to substantial declines in tobacco use and the relationship between these declines with CVH and CVD. We review interventions that have had success in high-income countries (HICs) as well as those that are gaining traction in low- and middle-income countries (LMICs). We emphasize the challenges to comprehensive tobacco prevention and control strategies faced by LMICs, and highlight the special role of cardiovascular health professionals in achieving CVH promotion and CVD prevention endpoints through tobacco control. Tobacco prevention and control strategies have a strong scientific basis, yet a distinct gap remains between this evidence and implementation of tobacco control policies, particularly in LMICs. Health professionals can contribute to tobacco control efforts, especially through patient-level clinical interventions, when supported by a health care system and government that recognize and support tobacco control as a critical strategy for CVH promotion and CVD prevention. Understanding, supporting, and applying current and evolving policies, programs, and practices in tobacco prevention and control is the province of all health professionals, especially those concerned with CVH promotion and CVD prevention. A new tobacco control roadmap from the World Heart Federation provides a strong impetus to the needed interdisciplinary collaboration. Tobacco control Smoking cessation Cardiovascular diseases Health personnel Global health 1. Introduction Cardiovascular disease (CVD) is the leading cause of death worldwide (WHO, 2012), and is especially prevalent in low- and middle-income countries (LMICs) where over 80% of CVD deaths occur (Hoyert and Zu, 2012). Comprehensive tobacco control is a critical global public health strategy for cardiovascular health (CVH) promotion and CVD prevention. Four fundamental strategies are distinguished in a CVH/CVD framework: 1) preserving CVH/low CVD risk; 2) controlling increased CVD risk; 3) detecting and treating acute CVD events; and 4) reducing disability and risk of recurrent CVD events. Tobacco control within each of these domains requires a different approach, with tobacco prevention being a key component for maintaining ideal CVH and tobacco cessation being a key component for restoring ideal CVH, controlling increased CVD risk, and treating individuals who have survived CVD events (Fig. 1). Here, we consider the role of comprehensive tobacco prevention and control within this CVH promotion and CVD prevention framework. We conducted an expert review using electronic databases (e.g., PubMed), manual searches of peer-reviewed journals most likely to publish global tobacco control research or commentaries (e.g., Tobacco Control, Global Heart), and reference lists of included articles, policy reports, and treatment practice guidelines. To better inform cardiovascular health professionals unfamiliar with tobacco prevention and control strategies, we review recent developments in this arena and the current state of global tobacco prevention and control. We include strategies that have led to the declines in tobacco use in high-income countries (HICs) and those that are gaining traction in LMICs, as well as topical challenges to global tobacco control in the context of CVH promotion and CVD prevention. We then describe the possible role of cardiovascular health professionals within the larger global tobacco control movement. We conclude with a focus on the tobacco endgame as a critical component of global CVH promotion and CVD prevention. 2. Viewing tobacco control from the perspective of CVH promotion and CVD prevention A priority for global CVH promotion and CVD prevention should be to end the tobacco epidemic through comprehensive, integrated, and coordinated tobacco prevention and control strategies (Reddy et al., 2012). The World Bank (1999) estimated that cumulative tobacco-attributable deaths through year 2050 would number 520 million with then-current prevention and intervention efforts, but could be reduced to 500 million if initiation rates were reduced by 50%, or to 340 million if adult consumption rates were reduced by 50%. Health professionals greatly contribute to CVD prevention by providing appropriate care for patients who use tobacco products. In the U.S., seven pharmacotherapies are approved for the treatment of tobacco use: nicotine replacement therapies (NRTs; patch, gum, lozenge, inhaler, and nasal spray) and non-NRT medications (bupropion and varenicline) (Cahill et al., 2013). Few randomized controlled trials of smoking cessation pharmacotherapies have been conducted in LMICs. Moreover, pharmacotherapy can be costly, less likely to be covered by insurance, and not as easily accessible in LMICs (Kishore et al., 2010). Cytisine, a natural compound with neurochemical properties similar to varenicline, shows promise as affordable and effective pharmaco-therapy options in LMICs (Hajek et al., 2013; West et al., 2015). In LMICs, health professionals should also note smokers’ use of alternative tobacco products, as use of these products may interfere with smoking cessation treatment adherence and abstinence. For example, waterpipe use predicted lower adherence to nicotine patch use in Syria (Ben Taleb et al., 2015). Health professionals’ advice to quit smoking can be a powerful motivator for smoking patients (Stead et al., 2013b), especially when delivered during hospitalization (Rigotti et al., 2012). Smokers who are advised to quit by a health care provider are 50% more likely to make a quit attempt (Davila et al., 2009). However, only about half of smoking patients in the U.S. report being advised to quit (CDC, 2007; Kruger et al., 2012). Encouraging health professionals to speak with patients about quitting, and facilitating access to evidence-based treatments, could increase the rate of quit attempts and successful cessation (McAfee, 2013). Globally, the cost of clinical interventions is significantly less than expenditures associated with continued tobacco use (Chen et al., 2012; John et al., 2009; Samet, 2010). Unfortunately, individual-level interventions alone do not address the greater tobacco epidemic, nor will they be sufficient to achieve the ultimate goal of a tobacco-free world. Comprehensive tobacco control comprises a multi-pronged approach that involves public health science, policy, and clinical practice devoted to cigarette smoking and other tobacco use prevention and cessation. Primordial strategies for tobacco control aim to prevent tobacco use initiation, particularly among children and adolescents, to promote CVH. Every day, nearly 100,000 children transition to regular smoking – 14,000–15,000 children per day in HICs and 68,000–84,000 in LMICs (The World Bank, 1999). Remedial strategies focus on smoking cessation at both the individual and population levels to control increased CVD risk and prevent recurrent CVD, as smoking cessation is beneficial at every age (Jha et al., 2013). Under these policies, primordial and remedial tobacco control strategies can both have major public health impacts: the first, to achieve ultimate success in eliminating the tobacco epidemic and promoting CVH; the second, to prevent tobacco-attributable deaths from CVD while reinforcing primordial strategies to end the tobacco epidemic (Fig. 1). 3. Developing and implementing the Framework Convention on Tobacco Control (FCTC) The Framework Convention for Tobacco Control (FCTC) represents a comprehensive approach to global tobacco control, and is the successful amalgamation of several international human rights proposals to address the global epidemic of tobacco use (WHO, 2003). With membership comprising 168 Signatories and 180 Parties,1 the FCTC is the most widely endorsed treaty in United Nations history. In 2008, the WHO summarized global tobacco control efforts under the FCTC in six strategies (MPOWER; WHO, 2008): Monitor implemented policies, Protect individuals from secondhand smoke, Offer cessation assistance, Warn about the health consequences of tobacco, Enforce bans on marketing, and Raise taxes and prices on tobacco products. To support implementation of MPOWER, Bloomberg Philanthropies provided a six-year $375 million dollar initiative for the WHO’s international efforts in 15 LMICs as part of the Tobacco Free Initiative (WHO, 2006, 2013b). By 2014, >2.8 billion people were newly protected by at least one, well-implemented MPOWER strategy (WHO, 2015). The tobacco control activities put forth by the FCTC are directly relevant to strategies for improving CVH and preventing CVD (Fig. 1). Primordial tobacco control, such as reducing sales to minors, is especially effective for preventing initiation of tobacco use and promoting CVH. Remedial tobacco control, including tobacco content regulation and clinical interventions for tobacco cessation, is designed to reduce CVD risk in adult smokers. Implementation of MPOWER policies reduced the number of smokers globally by 15 million in 2010, with a corresponding 7.5 million tobacco-attributable deaths averted (Levy et al., 2013). 4. Successes in global tobacco control: “best buys” To enact cost-effective interventions against noncommunicable diseases (NCDs) globally, the WHO Global Burden of Disease project proposed a list of “best buy” policies – strategies that require relatively low investment for their great projected impact and support both primordial and remedial tobacco control (WHO, 2011b; World Economic Forum and the Harvard School of Public Health, 2011). This list of best buys was adopted by the WHO as priority interventions for FCTC signatories. 4.1. Tobacco prices and taxes The most effective strategy for reducing tobacco uptake and promoting tobacco cessation is to increase the price of tobacco products (Savedoff and Alwaygn, 2015; WHO, 2015). Each 10% increase in the price of cigarettes is associated with 2–8% reductions in tobacco use in both HICs and LMICs (IARC, 2011). And yet, despite the efficacy of taxation as a tobacco control strategy in LMICs, the taxes placed on tobacco products in LMICs are less than half those in HICs (Jha, 2012; Jha and Peto, 2014). It is estimated that, if LMICs were to triple their excise taxes on tobacco products, they could achieve a 33% reduction in the prevalence of smoking (Jha and Peto, 2014), which would significantly decrease rates of tobacco-attributable CVD in these areas. In addition, raising taxes can increase revenues to fund other tobacco control activities, particularly in LMICs (Savedoff and Alwaygn, 2015). 4.2. Tobacco package labels Text and pictorial warnings on tobacco packages increase population awareness of the negative health consequences of tobacco use (Hammond, 2009). In a clinical trial, smokers randomized to have pictorial warnings on their cigarette packs had higher quit rates compared to those randomized to text-only warnings (Brewer et al., 2016). Plain packaging, which removes company branding and employs standard packaging and lettering, effectively strips tobacco companies of their proprietary marketing. From limited evidence, plain packaging appears to increase the salience of health warnings and decrease the appeal of cigarettes, particularly in LMICs (Hughes et al., 2016; Munafo et al., 2011; Stead et al., 2013a). Plain packaging was first implemented in Australia in 2012, is scheduled to be implemented in Ireland and England in 2016, and is under consideration in 10 other countries. Population studies in Australia show that introduction of plain packaging was associated with significantly reduced appeal of cigarette packs and increased motivation to quit smoking among both adolescents and adults (Wakefield et al., 2015; White et al., 2015). 4.3. Smoke-free policies and legislation Second-hand smoke is responsible for >600,000 deaths per year worldwide (Fichtenberg and Glantz, 2002; USDHHS, 2006). Secondhand smoke exposure affects many of the same pathophysiological pathways (e.g., inflammation, endothelial dysfunction) that cause CVD as mainstream tobacco smoke (Institute of Medicine Committee on Secondhand Smoke Exposure and Acute Coronary Events, 2010). Creating smoke-free workplaces is associated with approximately 4% reductions in tobacco use rates, and smokers who continued smoking reduced their consumption by 3 cigarettes per day on average – modest but meaningful effects (Fichtenberg and Glantz, 2002). More broadly, smoke-free legislation promotes denormalization of cigarettes by limiting their presence in public places. Further, it has been found to be associated with: reduced use of tobacco products among adolescents (Farrelly et al., 2013; Shang, 2015); 8–25% lower risk of CVD (Meyers et al., 2009); and an overall 13% reduction in rates of acute MI (OR: 0.87, 95% CI: 0.84–0.91) (Lin et al., 2013). A systematic review and meta-analysis found greater reductions in hospitalization rates for CVD with comprehensive versus partial smoke-free air laws around the world (for comprehensive laws, RR: 0.86, 95% CI: 0.83–0.89; for partial laws, RR: 0.92, 95% CI: 0.85–0.98) (Jones et al., 2014). 4.4. Mass media marketing Anti-smoking messaging via mass media has growing potential in this digital age of widespread media coverage. For example, researchers in the U.S. found that adolescents exposed to the anti-smoking messages of the National Truth Campaign had 20% lower risk of smoking initiation (HR: 0.80, 95% CI: 0.71–0.91) (Farrelly et al., 2009). For adults, campaigns that emphasize the health consequences of tobacco use and disparage the tobacco industry are associated with more negative attitudes about tobacco use (Durkin et al., 2012; Klesges et al., 2009). Mass media campaigns may have an even greater effect in LMICs, given their effectiveness among lower literacy populations (Durkin et al., 2012; Mullin et al., 2011). In sum, effective tobacco control strategies have been established on the basis of a large body of research, including economic analysis. However, their implementation is seriously lagging, especially in LMICs. 5. Recent and ongoing challenges and opportunities in global tobacco control 5.1. Barriers to tobacco control implementation in LMICs Although tobacco control policies have led to significant and often swift reductions in population smoking rates, and corresponding declines in CVD, there are important barriers to implementation of tobacco control policies in LMICs. Tobacco control research and policy are strategically interdependent (Mackay, 2013), however, there is often a lag in the process of policy research, implementation, evaluation, and optimization (FCTC Article 20; Gupta et al., 2012; Warner and Tam, 2012). To address this concern, the journal Tobacco Regulatory Science debuted in April 2015, to provide a place to publish research and evaluations concerning efficacy and effectiveness of tobacco control policies (Leischow et al., 2015). The science-policy gap in LMICs is only partly due to economic conditions (Oldenburg and Absetz, 2011). A systematic review identified five primary challenges to tobacco control implementation in LMICs: 1) information dissemination to the public regarding the health consequences of tobacco use; 2) tobacco companies taking advantage of trade disputes to establish and grow their market; 3) smuggling of tobacco products to increase sales and profits; 4) preventing tobacco taxes and smoke-free legislation through sponsored research and misinformation; and 5) competing interests and incentives within governments (Bump and Reich, 2013). Further research is needed to overcome these barriers. 5.2. The tobacco industry as a disease ‘vector’ The tobacco industry continues to fight public health efforts in the face of overwhelming evidence of the negative health consequences of tobacco use (Gilmore et al., 2015; Powers et al., 2004). The industry invests billions of dollars in marketing each year and has used deception in many forms to promote smoking initiation among youth and to maintain a high prevalence of tobacco use. For example, tobacco industry documents have revealed the industry’s efforts to pursue endorsement deals and contracts in Hollywood (Lum et al., 2008; Mekemson and Glantz, 2002) in order to capitalize on known associations between smoking in movies and uptake of cigarette smoking among youth (Charlesworth and Glantz, 2005). In recent years, these efforts have focused increasingly on LMICs (Henriksen, 2012; Wadland et al., 2011; WHO, 2011c). This record of behavior led to characterization of the tobacco industry as a disease “vector” – the mechanism by which the toxic exposure to tobacco is transmitted to its victim, the global population (Gilmore, 2012; Lee et al., 2012). Examples of toxic exposure include tobacco-oriented international trade and investment agreements, hostile litigation, and illicit trade of tobacco (e.g., smuggling), by which the tobacco industry attempts to prevent or dismantle public health policy efforts (e.g., graphic warning labels) and to leverage their resources to build up and exploit markets in vulnerable populations, especially in LMICs (Gilmore et al., 2015; Savell et al., 2014). An important step in combating the tobacco industry as a disease vector is to denormalize the industry by informing the public of its manipulative behavior and harmful influence (Gilmore et al., 2015). FCTC Article 5.3 details specific strategies to combat the industry, such as prohibiting industry contributions to events, activities, or individuals that would promote tobacco products (WHO, 2013c), banning point-of-sale advertising (i.e., promotional material placed near a check-out counter) directed to adolescents (Levy et al., 2015; Scheffels and Lavik, 2012), and implementing plain packaging and graphic labels on cigarette packages (Hammond and Parkinson, 2009) to limit the scope of tobacco industry marketing. Much work is needed in these areas to achieve a level of tobacco control to combat the industry’s growing influence in LMICs. 5.3. Alternative “safer” nicotine and tobacco products It is well established that there is no safe cigarette (NCI, 2001; USDHHS, 1964). Accordingly, tobacco control policies ban the use of misleading branding practices (e.g., “mild,” “light,” and “low-tar”) around the world (Elton-Marshall et al., 2010; Siahpush et al., 2011). In some countries, there is evidence of harm reduction, as opposed to harm elimination by total tobacco cessation, when smokeless tobacco products (e.g., snus, snuff, dissolvable tobacco) are used in place of cigarettes (Gartner et al., 2007). However, while switching to smokeless tobacco decreases risk of developing certain NCDs (e.g., lung cancer), it may increase risk for others (e.g., oral cancer) (Gray and Hecht, 2010); moreover, a national study of U.S. smokers found that using smokeless tobacco products was not associated with smoking cessation (Popova and Ling, 2013). Therefore, there may be no net harm reduction from smokeless tobacco products. Additional tobacco products that are used primarily in Eastern LMICs include bidis, supari, betel quid, gutka, and pan masala (Mahapatra et al., 2015). These products present unique challenges for tobacco control. Bidis, for example, tend to be smoked outdoors and therefore are less accountable under smoke-free legislation, and their packaging is not conducive to graphic warnings (Kumar et al., 2012b). Gutka and pan masala were legally banned in India in 2012, but tobacco vendor awareness of and compliance with the legislation is low (estimated 25–50%) (Pimple et al., 2014). Waterpipes, which originated in Indonesia but have recently gained popularity in HICs, are largely exempt from existing smoke-free legislation through legislative loopholes (e.g., hookah lounges are classified as retail shops rather than bars/restaurants) or vague phrasing (e.g., no clear definition of “smoking”) (Maziak, 2012; Noonan, 2010). Use of any of these alternative tobacco products conveys increased risk for NCDs, including CVD (Mahapatra et al., 2015; Maziak, 2012; Rahman and Fukui, 2000; Rastogi et al., 2005). Despite known health consequences, these products are not yet consistently regulated. They should be closely monitored and treated legally equivalent to cigarettes to prevent these products from fueling a new global epidemic of their own (O’Connor, 2012). Electronic nicotine delivery systems (ENDS, e-cigarettes) were patented in China in 2003. Since their introduction to the U.S. in 2007, ENDS use has climbed rapidly. Between 2013 and 2014, ever-ENDS use increased three-fold among U.S. middle school students (from 1.1% to 3.9%) and high school students (from 4.5% to 13.4%), surpassing cigarettes and becoming the most used tobacco product among adolescents (CDC, 2015). Cigarette smoking and ENDS use are highly related among adolescents and young adults (Agarwal and Loukas, 2015; Warner, 2016). Though a causal relationship has not been established, use of ENDS among adolescents is associated with use of combustible tobacco products within the next year (Leventhal et al., 2015). With growing evidence that nicotine use is a gateway to future marijuana and cocaine use, some are concerned that the uptake of these purely nicotine-delivery systems among youth may increase addiction rates to other substances in the younger generation (Kandel and Kandel, 2014). The long-term health consequences of ENDS use are unknown. Short-term studies show similar, if less severe, effects on the lungs compared to traditional cigarettes (Schweitzer et al., 2015; Vardavas et al., 2012). A meta-analysis found that adult smokers who use ENDS are 28% less likely to quit smoking than those not using ENDS (OR: 0.72, 95% CI: 0.57–0.91) (Kalkhoran and Glantz, 2016). This outcome may depend in part on the ENDS design, where smokers who use “open” devices (i.e., refillable “tank” systems) are more likely to quit smoking than users of “closed” devices (i.e., prefilled cartridges, “cigalikes”) (Chen et al., 2016). Furthermore, most ENDS users appear to be dual users (CDC, 2013; King et al., 2013), with use of ENDS serving to “bridge” between opportunities to smoke cigarettes, thereby sustaining tobacco dependence (Bell and Keane, 2012; O’Connor, 2012). Because continuing to smoke even 1–4 cigarettes per day is associated with nearly 300% greater risk of death due to ischemic heart disease versus non-smoking (Bjartveit and Tverdal, 2005), smokers who use ENDS to reduce but not abstain from cigarettes may remain at significant CVD risk. At the population and policy level, tobacco control advocates are concerned that ENDS threaten tobacco control efforts (Fairchild et al., 2014). ENDS can thwart tobacco control policies, including smoke-free legislation, advertising and promotion restrictions, and taxation. At the sixth session of the Conference of the Parties to the FCTC, the Parties “agreed to disagree” in the debate between the treatment benefit versus harm of ENDS by focusing on four regulatory objectives: 1) preventing uptake of ENDS among nonsmokers, 2) minimizing health risks of ENDS, 3) prohibiting false or deceptive promotion of ENDS, and 4) limiting the industry’s involvement in tobacco control efforts (Russell et al., 2016). 5.4. Harnessing mobile technology, Internet access, and social media for tobacco control New technologies offer opportunities to influence smoking behavior by facilitating widespread delivery of tobacco cessation treatment. Mobile-based health interventions (i.e., quitlines, text messaging services, and smartphone apps) hold great potential to provide low-cost tobacco treatment to the three-quarters of the global population carrying mobile phones (The World Bank, 2012), though research using technology for tobacco cessation remains in its infancy in HICs (Ubhi et al., 2016) and is essentially nonexistent in LMICs. The Internet offers many options for providing information and support for tobacco cessation. However, the tobacco industry has also taken advantage of this minimally regulated, inexpensive domain to exploit loopholes in tobacco control policies (Cohen et al., 2001; Hoek, 2004), including: cigarette sales to minors, as individuals online can more easily thwart efforts at age verification (Ribisl, 2012); discounts and coupons offered by online vendors to evade excise taxes (Kim et al., 2006); advertising and promotion via the Internet, where the tobacco industry has increased spending in recent years (Hrywna et al., 2007; Jenssen et al., 2009; U.S. Federal Trade Commission, 2009); and normalization of tobacco products using positive promotion via multiple, simultaneous channels (Nagler and Viswanath, 2013). Social media have also facilitated tobacco product promotion, particularly among youth. However, several potential approaches to tobacco control via social media could be considered. Proliferation of tobacco control messages on the Internet could promote tobacco industry denormalization and encourage tobacco cessation (Zhu et al., 2012). Social network websites can connect individuals on a potentially global scale – anti-tobacco messages could be posted as “pages” on Facebook, videos on YouTube, “tags” on Tumblr, or “tweets” on Twitter (Freeman and Chapman, 2007, 2010). Social media promote a sense of community (Cobb et al., 2010), and these channels can facilitate tobacco control campaigns and tobacco cessation (Hefler et al., 2012). 6. Ending the tobacco epidemic Discussion of the “tobacco problem,” as termed by the Institute of Medicine, has now advanced to “ending the epidemic” (USDHHS, 2010). Since the first U.S. Surgeon General’s Report in 1964, an estimated 8 million lives and 157 million life-years have been saved as a result of tobacco control efforts (Holford et al., 2014). Although the proportion of the population who are smoking has decreased significantly, with the growing global population the actual number of current daily smokers has increased (Ng et al., 2014), highlighting the need to persist in the development and optimization of interventions targeting tobacco cessation. Implementation of the above tobacco control strategies will necessarily create greater demand for cessation support (Raw et al., 2016). However, there appears to be little importance placed on providing cessation services, and implementation of cessation services is much lower in LMICs compared to HICs (Pine-Abata et al., 2013a). We support the position that many of the tenets of tobacco control could be promoted by health professionals in LMICs by providing widespread clinical cessation services and becoming advocates for tobacco control policies (Kumar et al., 2012a). To be successful, particularly in LMICs, this movement would require equipping health professionals with the necessary training to engage in comprehensive tobacco control efforts, addressing limitations to tobacco control within the health care system, and increasing government support for cost-effective distribution of tobacco cessation treatment. Unfortunately, many system-level barriers prevent successful adoption of tobacco control policies for offering cessation assistance (Samb et al., 2010). 6.1. Health care system barriers to tobacco control In LMICs, tobacco cessation services are not readily available in primary care or other general medical settings (Sivadasan Pillai and Ganapathi, 2012). Many tobacco cessation services are housed in tertiary care settings, which severely limits access to these tobacco services by the general public. In other cases, tobacco cessation assistance is offered through psychiatric or mental health services, which many tobacco users in LMICs may not wish to access due to stigma and taboo associated with receiving mental health treatment (Sivadasan Pillai and Ganapathi, 2012). These limitations could be addressed if expertise in tobacco cessation treatment were more widely available. Intervention at the level of primary care could reverse the current upward trend of tobacco use in LMICs (IOM, 2010). Reflecting these challenges, an analysis of Global Adult Tobacco Survey data collected from 16 LMICs found that only between 4% and 27% of smokers reported utilization of smoking cessation services (Wang et al., 2015). In the large majority of the LMICs surveyed, cessation counseling services were more likely to be used than pharmacotherapy, which was likely related to the higher costs and lower rates of insurance coverage of smoking cessation pharmacotherapy in those countries (Wang et al., 2015). These findings indicate that strengthening these services is a key component to smokers in LMICs receiving cessation assistance. There is a pressing need to address tobacco use and treatment in the medical and nursing school curricula in LMICs to equip all health professionals with the necessary skills to motivate and support tobacco quit attempts (Richmond et al., 1998). In 2013, approximately 20% of countries, the majority of which were LMICs, did not have a tobacco cessation training program for health professionals, and over 50% of those who did have at least one training program reported difficulties with funding stability for the program(s) (Kruse et al., 2015). Simply increasing health professionals’ knowledge and the rate of tobacco use screening is a critical first step for health professionals in LMICs (Shelley et al., 2014; Yan et al., 2008). More advanced strategies may involve active learning techniques, such as use of standardized patients, to increase health professionals’ confidence in performing behavioral cessation counseling (Wadland et al., 2011; Walsh et al., 2007). Another barrier to tobacco treatment delivery in LMICs may be the high rates of tobacco use among health professionals. As recently as 2010, smoking rates among students in the health professions exceeded 40% in 13 countries (Eriksen et al., 2012). Physicians who smoke, compared to physicians who have never smoked, are less likely to offer smoking cessation advice or assistance (Pipe et al., 2009). Strengthening tobacco control training and resources in the health care system could support health professionals to provide tobacco cessation interventions for patients. 6.2. Government support for tobacco cessation services Despite the widespread acknowledgement of the importance of tobacco cessation, implementation of strategies to “offer help to quit tobacco use” was lowest (<20% of FCTC Parties) among the six core measures of MPOWER (WHO, 2013a). An assessment of progress related to tobacco control policies concerning tobacco dependence and cessation found that only 44% of the 121 respondent countries had national treatment guidelines (Pine-Abata et al., 2013b). When present, guideline components included: behavioral cessation counseling for patients (56%), cessation support for healthcare workers (46%), a national treatment strategy (44%), a government official responsible for tobacco use treatment (41%), telephone quitlines (36%), and a national network of support for tobacco treatment (17%) (Pine-Abata et al., 2013a). National telephone quitlines are an effective and cost-effective modality for providing tobacco cessation clinical interventions (Stead et al., 2013c; Tomson et al., 2004). There are strong correlations between quitline spending, both for services and promotion, and the reach of effective treatment (Saul et al., 2014). As of 2009, however, few LMICs had implemented national quitlines, and those that had done so did not achieve sufficient population coverage and were underutilized (Wang et al., 2015; WHO, 2011a). 7. The Tobacco Endgame Ending the Tobacco Epidemic (USDHHS, 2010) cited “a historical opportunity to rekindle the momentum of previous decades and achieve the vision of a society free from tobacco-related death and disease.” Four strategic actions were proposed: 1) strengthening implementation of evidence-based interventions and policies in states and communities; 2) changing social norms around tobacco use; 3) leveraging health and human service systems and resources to create a society free of tobacco-related disease and death; and 4) accelerating research to expand the science base and monitor progress. These actions require the “ideal policy environment” for tobacco control within each FCTC Party, as described by Cairney and Mamudu (2014): 1) each country’s department of health is responsible for tobacco control policies and monitoring; 2) tobacco control is presented as a solvable public health problem; 3) public health groups are included, while tobacco interest groups are excluded from tobacco control activities; 4) the environment itself begets policy implementation; and 5) scientific evidence for tobacco control policies is widely accepted by government entities. Some researchers argue that bold new strategies are required to reduce and, ultimately, eliminate tobacco use and tobacco-related disease burden – for example, implementing policies that gradually reduce nicotine and addictive additives in tobacco products; raising the legal age of purchase and possession of tobacco products annually to create a smoke-free generation; and systematically limiting the tobacco industry’s production of tobacco products (McDaniel et al., 2015; Novotny, 2015). 8. Conclusion The 20th anniversary issue of Tobacco Control included a critical appraisal of the relation between the tobacco control movement and prevention of NCDs (among which CVD predominates) (Wipfli and Samet, 2012). Integration and synergy of these efforts would be supported by a global commitment to the United Nations Sustainable Development Goal 3.4 to reduce mortality due to NCDs (and CVD in particular) by one third by the year 2030 (http://www.un.org/sustainabledevelopment/health/) and the WHO goal of a 25% reduction in the risk of deaths due to NCDs by 2025 (Kontis et al., 2014). Comprehensive tobacco control will be essential to achieving these goals (Kontis et al., 2014; Raw et al., 2016). A significant development toward realizing a harmonization of tobacco control with NCD control was the adoption by the World Heart Federation (WHF) of its “Roadmap for Tobacco Control” in 2015 (Grainger Gasser et al., 2015). The Roadmap gives high prominence to tobacco control for achieving this goal and calls on many sectors of society to engage in its implementation. This global initiative brings tobacco control and CVH promotion/CVD prevention together under the aegis of the WHF and augurs a promising future for collaboration between the tobacco control and cardiovascular domains. (Further discussion on the integration of global tobacco control and NCD control is beyond the scope of this article; interested readers are referred to Additional Resources.2) Comprehensive tobacco control has improved and extended many lives over the past five decades, but many issues remain to be addressed to reach the ultimate goal: a tobacco-free world. Integration of primordial and remedial strategies for tobacco control promise to have a significant impact on CVH promotion and CVD prevention, particularly in LMICs. Policies such as tobacco control “best buys,” utilization of social media channels, and management of novel tobacco products are effective points of action. However, the importance of clinical interventions for tobacco control should not be overlooked. In the near future, greater reduction in CVD-related morbidity and mortality could be achieved through smoking cessation by providing health professionals with the appropriate skills and training in tobacco prevention and control, as well as supporting a health care system designed to promote the tobacco endgame as a critical strategy for global CVH promotion and CVD prevention. Funding Allison Carroll was supported by a Predoctoral Individual National Research Service Award (F31 HL129494). Fig. 1 Fundamental tobacco control strategies mapped onto cardiovascular health promotion and cardiovascular disease prevention. 1 These data are as of 4 March 2015, and last checked on 3 March 2016. For more information on FCTC Signatures and Ratification, please visit: https://treaties.un.org/pages/ViewDetails.aspx?src=TREATY&mtdsg_no=IX-4&chapter=9&lang=en. 2 Additional Resources: McKee, M., Haines, A., Ebrahim, S., Lamptey, P., Barreto, M.L., Matheson, D., Walls, H.L., Foliaki, S., Miranda, J.J., et al., 2014. Towards a comprehensive global approach to prevention and control of NCDs. Globalization and Health 10:74. doi: 10.1186/s12992-014-0074-8 Reddy, K.S., Yadav, A., Arora, M., Nazar, G.P., 2012. Integrating tobacco control into health and development agendas. Tob Control 21:281–6. doi: 10.1136/tobaccocontrol-2011-050419 Wipfli, H.L., Samet, J.M., 2012. Moving beyond global tobacco control to global disease control. Tob Control 21:269–72. doi: 10.1136/tobaccocontrol-2011-050322 World Health Organization, 2013. Global Action Plan for the Prevention and Control of Noncommunicable Diseases 2013–2020. World Health Organization, Geneva, Switzerland. Available from: http://apps.who.int/iris/bitstream/10665/94384/1/9789241506236_eng.pdf. Conflict of interest statement We do not have any financial interest/arrangement or affiliation with one or more organizations that could be perceived as a real or apparent conflict of interest in the context of the subject of this presentation. DRL and MDH receive grant support from the World Heart Federation to serve as a consultant and the senior program advisor, respectively, for the World Heart Federation’s Emerging Leaders program, which is supported by Boehringer Ingelheim and Bupa. BH has served on a scientific advisory board for Pfizer and receives study medication and placebo free of charge from Pfizer for use in ongoing National Institutes of Health funded clinical trials. Agarwal D Loukas A 2015 Examining e-cigarette use among college students Tob Regul Sci 1 166 174 http://dx.doi.org/10.18001/TRS.1.2.6 Bell K Keane H 2012 Nicotine control: e-cigarettes, smoking and addiction Int J Drug Policy 23 242 247 http://dx.doi.org/10.1016/j.drugpo.2012.01.006 22365155 Ben Taleb Z Ward KD Asfar T Bahelah R Maziak W 2015 Predictors of adherence to pharmacological and behavioral treatment in a cessation trial among smokers in Aleppo, Syria Drug Alcohol Depend 153 167 172 http://dx.doi.org/10.1016/j.drugalcdep.2015.05.038 26077603 Bjartveit K Tverdal A 2005 Health consequences of smoking 1–4 cigarettes per day Tob Control 14 315 320 16183982 Brewer NT Hall MG Noar SM Parada H Stein-Seroussi A Bach LE Hanley S Ribisl KM 2016 Effect of pictorial cigarette pack warnings on changes in smoking behavior: a randomized clinical trial JAMA Intern Med 176 905 912 http://dx.doi.org/10.1001/jamainternmed.2016.2621 27273839 Bump JB Reich MR 2013 Political economy analysis for tobacco control in low- and middle-income countries Health Policy Plan 28 123 133 http://dx.doi.org/10.1093/heapol/czs049 22585874 Cahill K Stevens S Perera R Lancaster T 2013 Pharmacological interventions for smoking cessation: an overview and network meta-analysis Cochrane Database Syst Rev 5 CD009329 http://dx.doi.org/10.1002/14651858.CD009329.pub2 Cairney P Mamudu H 2014 The global tobacco control ‘endgame’: change the policy environment to implement the FCTC J Public Health Policy 35 506 517 http://dx.doi.org/10.1057/jphp.2014.18 24831675 CDC 2007 Smoking-cessation advice from health-care providers - Canada, 2005 MMWR 56 708 712 17637597 CDC 2013 Tobacco product use among middle and high school students - United States, 2011–2012 MMWR 62 729 730 24005229 CDC 2015 Tobacco use among middle and high school students — United States, 2011–2014 MMWR 64 381 385 25879896 Charlesworth A Glantz SA 2005 Smoking in the movies increases adolescent smoking: a review Pediatrics 116 1516 1528 http://dx.doi.org/10.1542/peds.2005-0141 16322180 Chen PC Lee YC Tsai ST Lai CK 2012 A cost-benefit analysis of the outpatient smoking cessation services in Taiwan from a societal viewpoint Nicotine Tob Res 14 522 530 http://dx.doi.org/10.1093/ntr/ntr241 22180585 Chen C Zhuang YL Zhu SH 2016 E-cigarette design preference and smoking cessation: A U.S. population study Am J Prev Med http://dx.doi.org/10.1016/j.amepre.2016.02.002 Cobb NK Graham AL Abrams DB 2010 Social network structure of a large online community for smoking cessation Am J Public Health 100 1282 1289 http://dx.doi.org/10.2105/AJPH.2009.165449 20466971 Cohen JE Sarabia V Ashley MJ 2001 Tobacco commerce on the internet: a threat to comprehensive tobacco control Tob Control 10 364 367 http://dx.doi.org/10.1136/tc.10.4.364 11740029 Davila EP Zhao W Byrne M Webb M Huang Y Arheart K Dietz N Caban-Martinez A Parker D 2009 Correlates of smoking quit attempts: Florida Tobacco Callback Survey, 2007 Tob Induc Dis 5 10 http://dx.doi.org/10.1186/1617-9625-5-10 19563642 Durkin S Brennan E Wakefield M 2012 Mass media campaigns to promote smoking cessation among adults: an integrative review Tob Control 21 127 138 http://dx.doi.org/10.1136/tobaccocontrol-2011-050345 22345235 Elton-Marshall T Fong GT Zanna MP Jiang Y Hammond D O’Connor RJ Yong HH Li L King B 2010 Beliefs about the relative harm of “light” and “low tar” cigarettes: findings from the International Tobacco Control (ITC) China Survey Tob Control 19 Suppl 2 i54 i62 http://dx.doi.org/10.1136/tc.2008.029025 20935197 Eriksen MP Mackay J Ross H 2012 The Tobacco Atlas 4 American Cancer Society Atlanta, GA Fairchild AL Bayer R Colgrove J 2014 The renormalization of smoking? E-cigarettes and the tobacco “endgame” N Engl J Med 370 293 295 http://dx.doi.org/10.1056/NEJMp1313940 24350902 Farrelly MC Nonnemaker J Davis KC Hussin A 2009 The influence of the national truth campaign on smoking initiation Am J Prev Med 36 379 384 http://dx.doi.org/10.1016/j.amepre.2009.01.019 19211213 Farrelly MC Loomis BR Han B Gfroerer J Kuiper N Couzens GL Dube S Caraballo RS 2013 A comprehensive examination of the influence of state tobacco control programs and policies on youth smoking Am J Public Health 103 549 555 http://dx.doi.org/10.2105/AJPH.2012.300948 23327252 Fichtenberg CM Glantz SA 2002 Effect of smoke-free workplaces on smoking behaviour: systematic review BMJ 325 188 http://dx.doi.org/10.1136/bmj.325.7357.188 12142305 Freeman B Chapman S 2007 Is “YouTube” telling or selling you something? Tobacco content on the YouTube video-sharing website Tob Control 16 207 210 http://dx.doi.org/10.1136/tc.2007.020024 17565142 Freeman B Chapman S 2010 British American Tobacco on Facebook: undermining article 13 of the global World Health Organization framework convention on tobacco control Tob Control 19 e1 e9 http://dx.doi.org/10.1136/tc.2009.032847 Gartner CE Hall WD Vos T Bertram MY Wallace AL Lim SS 2007 Assessment of Swedish snus for tobacco harm reduction: an epidemiological modelling study Lancet 369 2010 2014 http://dx.doi.org/10.1016/s0140-6736(07)60677-1 17498798 Gilmore AB 2012 Understanding the vector in order to plan effective tobacco control policies: an analysis of contemporary tobacco industry materials Tob Control 21 119 126 http://dx.doi.org/10.1136/tobaccocontrol-2011-050397 22345234 Gilmore AB Fooks G Drope J Bialous SA Jackson RR 2015 Exposing and addressing tobacco industry conduct in low-income and middle-income countries Lancet 385 1029 1043 http://dx.doi.org/10.1016/S0140-6736(15)60312-9 25784350 Grainger Gasser A Welch C Arora M Greenland R Bhatti L Sanda L Moodie R Bianco E 2015 Reducing cardiovascular mortality through tobacco control Glob Health 10 123 133 http://dx.doi.org/10.1016/j.gheart.2015.04.007 Gray N Hecht SS 2010 Smokeless tobacco–proposals for regulation Lancet 375 1589 1591 http://dx.doi.org/10.1016/S0140-6736(10)60516-8 20452507 Gupta R Pandey MR Sharma D Malik A Santharaj W 2012 A checklist for CVD control in South Asia Glob Health 7 193 194 http://dx.doi.org/10.1016/j.gheart.2012.01.003 Hajek P McRobbie H Myers K 2013 Efficacy of cytisine in helping smokers quit: systematic review and meta-analysis Thorax 68 1037 1042 http://dx.doi.org/10.1136/thoraxjnl-2012-203035 23404838 Hammond D 2009 Tobacco Labelling and Packaging Toolkit: A Guide to FCTC Article 11 Waterloo, Canada Hammond D Parkinson C 2009 The impact of cigarette package design on perceptions of risk J Public Health (Oxf) 31 345 353 http://dx.doi.org/10.1093/pubmed/fdp066 19636066 Hefler M Freeman B Chapman S 2012 Tobacco control advocacy in the age of social media: using Facebook, Twitter and Change Tob Control http://dx.doi.org/10.1136/tobaccocontrol-2012-050721 Henriksen L 2012 Comprehensive tobacco marketing restrictions: promotion, packaging, price and place Tob Control 21 147 153 http://dx.doi.org/10.1136/tobaccocontrol-2011-050416 22345238 Hoek J 2004 Tobacco promotion restrictions: Ironies and unintended consequences J Bus Res 57 1250 1257 http://dx.doi.org/10.1016/S0148-2963(02)00449-6 Holford TR Meza R Warner KE Meernik C Jeon J Moolgavkar SH Levy DT 2014 Tobacco control and the reduction in smoking-related premature deaths in the United States, 1964–2012 JAMA 311 164 171 24399555 Hoyert DL Zu JQ 2012 Deaths: Preliminary Data for 2011 National vital statistics reports National Center for Health Statistics Hyattsville, MD Hrywna M Delnevo CD Lewis MJ 2007 Adult recall of tobacco advertising on the internet Nicotine Tob Res 9 1103 1107 http://dx.doi.org/10.1080/14622200701488442 17852764 Hughes N Arora M Grills N 2016 Perceptions and impact of plain packaging of tobacco products in low and middle income countries, middle to upper income countries and low-income settings in high-income countries: a systematic review of the literature BMJ Open 6 http://dx.doi.org/10.1136/bmjopen-2015-010391 IARC 2011 Effectiveness of Tax and Price Policies for Tobacco Control Lyon France Institute of Medicine Committee on Secondhand Smoke Exposure and Acute Coronary Events 2010r Secondhand Smoke Exposure and Cardiovascular Effects: Making Sense of the Evidence National Academies Press (US) Washington (DC) IOM 2010 Promoting Cardiovascular Health in the Developing World: A Critical Challenge to Achieve Global Health National Academies Press Washington D.C Jenssen BP Klein JD Salazar LF Daluga NA DiClemente RJ 2009 Exposure to tobacco on the internet: content analysis of adolescents’ internet use Pediatrics 124 e180 e186 http://dx.doi.org/10.1542/peds.2008-3838 19620193 Jha P 2012 Death and taxes: epidemiological and economic evidence on smoking Glob Health 7 139 142 http://dx.doi.org/10.1016/j.gheart.2012.05.004 Jha P Peto R 2014 Global effects of smoking, of quitting, and of taxing tobacco N Engl J Med 370 60 68 http://dx.doi.org/10.1056/NEJMra1308383 24382066 Jha P Ramasundarahettige C Landsman V Rostron B Thun M Anderson RN McAfee T Peto R 2013 21st-century hazards of smoking and benefits of cessation in the United States N Engl J Med 368 341 350 http://dx.doi.org/10.1056/NEJMsa1211128 23343063 John RM Sung HY Max W 2009 Economic cost of tobacco use in India, 2004 Tob Control 18 138 143 http://dx.doi.org/10.1136/tc.2008.027466 19131453 Jones MR Barnoya J Stranges S Losonczy L Navas-Acien A 2014 Cardiovascular events following smoke-free legislations: an updated systematic review and meta-analysis Curr Environ Health Rep 1 239 249 http://dx.doi.org/10.1007/s40572-014-0020-1 25328861 Kalkhoran S Glantz SA 2016 E-cigarettes and smoking cessation in real-world and clinical settings: a systematic review and meta-analysis Lancet Respir Med 4 116 128 http://dx.doi.org/10.1016/S2213-2600(15)00521-4 26776875 Kandel ER Kandel DB 2014 A molecular basis for nicotine as a gateway drug N Engl J Med 371 932 943 http://dx.doi.org/10.1056/NEJMsa1405092 25184865 Kim AE Ribisl KM Delnevo CD Hrywna M 2006 Smokers’ beliefs and attitudes about purchasing cigarettes on the Internet Public Health Rep 121 594 602 16972513 King BA Alam S Promoff G Arrazola R Dube SR 2013 Awareness and ever-use of electronic cigarettes among U.S. adults, 2010–2011 Nicotine Tob Res 15 1623 1627 23449421 Kishore SP Bitton A Cravioto A Yach D 2010 Enabling access to new WHO essential medicines: the case for nicotine replacement therapies Glob Health 6 22 http://dx.doi.org/10.1186/1744-8603-6-22 Klesges RC Sherrill-Mittleman DA DeBon M Talcott GW Vanecek RJ 2009 Do we believe the tobacco industry lied to us? Association with smoking behavior in a military population Health Educ Res 24 909 921 http://dx.doi.org/10.1093/Her/Cyp029 19528314 Kontis V Mathers CD Rehm J Stevens GA Shield KD Bonita R Riley LM Poznyak V Beaglehole R 2014 Contribution of six risk factors to achieving the 25 × 25 non-communicable disease mortality reduction target: a modelling study Lancet 384 427 437 http://dx.doi.org/10.1016/s0140-6736(14)60616-4 24797573 Kruger J Shaw L Kahende J Frank E 2012 Health care providers’ advice to quit smoking, National Health Interview Survey, 2000, 2005, and 2010 Prev Chronic Dis 9 E130 22814236 Kruse GR Rigotti NA Raw M McNeill A Murray R Piné-Abata H Bitton A McEwen A 2015 Tobacco dependence treatment training programs: an international survey Nicotine Tob Res http://dx.doi.org/10.1093/ntr/ntv142 Kumar G Mohan S Yadav L Arora M 2012a Global tobacco surveys: information for action by cardiologists Glob Health 7 99 105 http://dx.doi.org/10.1016/j.gheart.2012.06.005 Kumar G Sharma I Kabir Z 2012b Noncigarette smoking patterns, their health deffects and policy options Glob Health 7 175 178 http://dx.doi.org/10.1016/j.gheart.2012.04.002 Lee S Ling PM Glantz SA 2012 The vector of the tobacco epidemic: tobacco industry practices in low and middle-income countries Cancer Causes Control 23 Suppl 1 117 129 http://dx.doi.org/10.1007/s10552-012-9914-0 22370696 Leischow SJ Li DC Okamoto JM Liu H 2015 Tobacco regulatory science: implementing a new journal for tobacco control Tob Regul Sci 1 3 9 http://dx.doi.org/10.18001/TRS.1.1.1 Leventhal AM Strong DR Kirkpatrick MG Unger JB Sussman S Riggs NR Stone MD Khoddam R Samet JM 2015 Association of electronic cigarette use with Initiation of combustible tobacco product smoking in early adolescence JAMA 314 700 707 http://dx.doi.org/10.1001/jama.2015.8950 26284721 Levy DT Ellis JA Mays D Huang AT 2013 Smoking-related deaths averted due to three years of policy progress Bull World Health Organ 91 509 518 http://dx.doi.org/10.2471/BLT.12.113878 23825878 Levy DT Lindblom EN Fleischer NL Thrasher J Mohlman MK Zhang Y Monshouwer K Nagelhout GE 2015 Public health effects of restricting retail tobacco product displays and ads Tob Regul Sci 1 61 75 http://dx.doi.org/10.18001/TRS.1.1.7 26191538 Lin H Wang H Wu W Lang L Wang Q Tian L 2013 The effects of smoke-free legislation on acute myocardial infarction: a systematic review and meta-analysis BMC Public Health 13 529 http://dx.doi.org/10.1186/1471-2458-13-529 23721370 Lum KL Polansky JR Jackler RK Glantz SA 2008 Signed, sealed and delivered: “big tobacco” in Hollywood, 1927–1951 Tob Control 17 313 323 http://dx.doi.org/10.1136/tc.2008.025445 18818225 Mackay J 2013 The role of research on the development and implementation of policy Nicotine Tob Res 15 757 760 http://dx.doi.org/10.1093/ntr/nts197 23089484 Mahapatra S Kamath R Shetty BK Binu VS 2015 Risk of oral cancer associated with gutka and other tobacco products: a hospital-based case-control study J Cancer Res Ther 11 199 203 http://dx.doi.org/10.4103/0973-1482.143332 25879362 Maziak W 2012 The waterpipe: a new global threat to CV health? Glob Health 7 179 181 http://dx.doi.org/10.1016/j.gheart.2012.03.003 McAfee T 2013 Encouraging smokers to talk with their physicians about quitting JAMA 309 2329 2330 http://dx.doi.org/10.1001/jama.2013.5975 23700193 McDaniel PA Smith EA Malone RE 2015 The tobacco endgame: a qualitative review and synthesis Tob Control http://dx.doi.org/10.1136/tobaccocontrol-2015-052356 Mekemson C Glantz SA 2002 How the tobacco industry built its relationship with Hollywood Tob Control 11 Suppl 1 I81 I91 11893818 Meyers DG Neuberger JS He J 2009 Cardiovascular effect of bans on smoking in public places: a systematic review and meta-analysis J Am Coll Cardiol 54 1249 1255 http://dx.doi.org/10.1016/j.jacc.2009.07.022 19778665 Mullin S Prasad V Kaur J Turk T 2011 Increasing evidence for the efficacy of tobacco control mass media communication programming in low- and middle-income countries J Health Commun 16 Suppl 2 49 58 http://dx.doi.org/10.1080/10810730.2011.601395 21916713 Munafo MR Roberts N Bauld L Leonards U 2011 Plain packaging increases visual attention to health warnings on cigarette packs in non-smokers and weekly smokers but not daily smokers Addiction 106 1505 1510 http://dx.doi.org/10.1111/j.1360-0443.2011.03430.x 21401767 Nagler RH Viswanath K 2013 Implementation and research priorities for FCTC Articles 13 and 16: tobacco advertising, promotion, and sponsorship and sales to and by minors Nicotine Tob Res 15 832 846 http://dx.doi.org/10.1093/ntr/nts331 23291641 NCI 2001 Risks associated with smoking cigarettes with low machine-measured yields of tar and nicotine Institute, N.C U.S. Department of Health and Human Services, Public Health Service, National Institutes of Health Bethesda, MD Ng M Freeman MK Fleming TD Robinson M Dwyer-Lindgren L Thomson B Wollum A Sanman E Wulf S 2014 Smoking prevalence and cigarette consumption in 187 countries, 1980–2012 JAMA 311 183 192 24399557 Noonan D 2010 Exemptions for hookah bars in clean indoor air legislation: a public health concern Public Health Nurs 27 49 53 http://dx.doi.org/10.1111/j.1525-1446.2009.00826.x 20055968 Novotny TE 2015 The tobacco endgame: is it possible? PLoS Med 12 e1001832 http://dx.doi.org/10.1371/journal.pmed.1001832 26024483 O’Connor RJ 2012 Non-cigarette tobacco products: what have we learnt and where are we headed? Tob Control 21 181 190 http://dx.doi.org/10.1136/tobaccocontrol-2011-050281 22345243 Oldenburg B Absetz P 2011 Lost in translation: overcoming the barriers to global implementation and exchange of behavioral medicine evidence Transl Behav Med 1 252 255 http://dx.doi.org/10.1007/s13142-011-0051-1 21765879 Pimple S Gunjal S Mishra GA Pednekar MS Majmudar P Shastri SS 2014 Compliance to Gutka ban and other provisions of COTPA in Mumbai Indian J Cancer 51 Suppl 1 S60 S66 http://dx.doi.org/10.4103/0019-509x.147475 25526251 Pine-Abata H McNeill A Murray R Bitton A Rigotti N Raw M 2013a A survey of tobacco dependence treatment services in 121 countries Addiction 108 1476 1484 http://dx.doi.org/10.1111/add.12172 23451932 Pine-Abata H McNeill A Raw M Bitton A Rigotti N Murray R 2013b A survey of tobacco dependence treatment guidelines in 121 countries Addiction 108 1476 1484 http://dx.doi.org/10.1111/add.12158 23451932 Pipe A Sorensen M Reid R 2009 Physician smoking status, attitudes toward smoking, and cessation advice to patients: an international survey Patient Educ Couns 74 118 123 http://dx.doi.org/10.1016/j.pec.2008.07.042 18774670 Popova L Ling PM 2013 Alternative tobacco product use and smoking cessation: a national study Am J Public Health 103 923 930 http://dx.doi.org/10.2105/ajph.2012.301070 23488521 Powers CA Zapka JG Bognar B Dube C Hyder Ferry L Ferguson KJ O’Donnell JF Rigotti N Conley Thomson C 2004 Evaluation of current tobacco curriculum at 12 US medical schools J Cancer Educ 19 212 219 http://dx.doi.org/10.1207/s15430154jce1904_7 15725638 Rahman M Fukui T 2000 Bidi smoking and health Public Health 114 123 127 http://dx.doi.org/10.1038/sj.ph.1900625 10800151 Rastogi T Jha P Reddy KS Prabhakaran D Spiegelman D Stampfer MJ Willett WC Ascherio A 2005 Bidi and cigarette smoking and risk of acute myocardial infarction among males in urban India Tob Control 14 356 358 http://dx.doi.org/10.1136/tc.2005.011965 16183987 Raw M Mackay J Reddy S 2016 Time to take tobacco dependence treatment seriously Lancet 387 412 413 http://dx.doi.org/10.1016/s0140-6736(15)00950-2 26803442 Reddy KS Yadav A Arora M Nazar GP 2012 Integrating tobacco control into health and development agendas Tob Control 21 281 286 http://dx.doi.org/10.1136/tobaccocontrol-2011-050419 22345268 Ribisl KM 2012 Research gaps related to tobacco product marketing and sales in the family smoking prevention and tobacco control act Nicotine Tob Res 14 43 53 http://dx.doi.org/10.1093/ntr/ntr098 21690316 Richmond RL Debono DS Larcos D Kehoe L 1998 Worldwide survey of education on tobacco in medical schools Tob Control 7 247 252 9825419 Rigotti NA Clair C Munafo MR Stead LF 2012 Interventions for smoking cessation in hospitalised patients Cochrane Database Syst Rev 5 CD001837 http://dx.doi.org/10.1002/14651858.CD001837.pub3 Russell A Wainwright M Tilson M 2016 Means and ENDS – e-cigarettes, the framework convention on tobacco control, and global health diplomacy in action Glob Public Health 1 16 http://dx.doi.org/10.1080/17441692.2016.1152284 Samb B Desai N Nishtar S Mendis S Bekedam H Wright A Hsu J Martiniuk A Celletti F 2010 Prevention and management of chronic disease: a litmus test for health-systems strengthening in low-income and middle-income countries Lancet 376 1785 1797 http://dx.doi.org/10.1016/S0140-6736(10)61353-0 21074253 Samet JM 2010 Estimating the burden of smoking: premature mortality, morbidity, and costs Salud Publica Mex 52 Suppl 2 S98 107 21243216 Saul JE Bonito JA Provan K Ruppel E Leischow SJ 2014 Implementation of tobacco cessation quitline practices in the United States and Canada Am J Public Health 104 e98 e105 http://dx.doi.org/10.2105/AJPH.2014.302074 Savedoff W Alwaygn A 2015 The Single Best Health Policy in the World: Tobacco Taxes Center for Global Development Washington DC Savell E Gilmore AB Fooks G 2014 How does the tobacco industry attempt to influence marketing regulations? A systematic review PLoS One 9 e87389 http://dx.doi.org/10.1371/journal.pone.0087389 24505286 Scheffels J Lavik R 2012 Out of sight, out of mind? Removal of point-of-sale tobacco displays in Norway Tob Control http://dx.doi.org/10.1136/tobaccocontrol-2011-050341 Schweitzer KS Chen SX Law S Van Demark MJ Poirier C Justice MJ Hubbard WC Kim ES Lai X 2015 Endothelial disruptive pro-inflammatory effects of nicotine and e-cigarette vapor exposures Am J Phys Lung Cell Mol Phys 00411 2015 http://dx.doi.org/10.1152/ajplung.00411.2014 Shang C 2015 The effect of smoke-free air law in bars on smoking initiation and relapse among teenagers and young adults Int J Environ Res Public Health 12 504 520 http://dx.doi.org/10.3390/ijerph120100504 25584419 Shelley D Tseng TY Pham H Nguyen L Keithly S Stillman F Nguyen N 2014 Factors influencing tobacco use treatment patterns among Vietnamese health care providers working in community health centers BMC Public Health 14 68 http://dx.doi.org/10.1186/1471-2458-14-68 24450865 Siahpush M Borland R Fong GT Elton-Marshall T Yong HH Holumyong C 2011 Socioeconomic differences in the effectiveness of the removal of the “light” descriptor on cigarette packs: findings from the International Tobacco Control (ITC) Thailand Survey Int J Environ Res Public Health 8 2170 2180 http://dx.doi.org/10.3390/ijerph8062170 21776224 Sivadasan Pillai H Ganapathi S 2012 Tobacco cessation approaches and impact on CVD Glob Health 7 129 138 http://dx.doi.org/10.1016/j.gheart.2012.06.001 Stead M Moodie C Angus K Bauld L McNeill A Thomas J Hastings G Hinds K O’Mara-Eves A 2013a Is consumer response to plain/standardised tobacco packaging consistent with framework convention on tobacco control guidelines? A systematic review of quantitative studies PLoS One 8 e75919 http://dx.doi.org/10.1371/journal.pone.0075919 24146791 Stead LF Buitrago D Preciado N Sanchez G Hartmann-Boyce J Lancaster T 2013b Physician advice for smoking cessation Cochrane Database Syst Rev 5 CD000165 http://dx.doi.org/10.1002/14651858.CD000165.pub4 Stead LF Hartmann-Boyce J Perera R Lancaster T 2013c Telephone counselling for smoking cessation Cochrane Database Syst Rev 8 CD002850 http://dx.doi.org/10.1002/14651858.CD002850.pub3 The World Bank 1999 Curbing the epidemic: governments and the economics of tobacco control Tob Control 8 196 201 10478406 The World Bank 2012 Mobile Phone Access Reaches Three Quarters of Planet’s Population http://www.worldbank.org/en/news/press-release/2012/07/17/mobile-phone-access-reachesthree-quarters-planets-population 6 17 Tomson T Helgason AR Gilljam H 2004 Quitline in smoking cessation: a cost-effectiveness analysis Int J Technol Assess Health Care 20 469 474 15609797 U.S. Federal Trade Commission 2009 Cigarette Report for 2006 http://ftc.gov/os/2009/08/090812cigarettereport.pdf Ubhi HK Kotz D Michie S van Schayck OC Sheard D Selladurai A West R 2016 Comparative analysis of smoking cessation smartphone applications available in 2012 versus 2014 Addict Behav 58 175 181 http://dx.doi.org/10.1016/j.addbeh.2016.02.026 26950256 USDHHS 1964 Smoking and Health: A Report of the Surgeon General U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health Atlanta, GA.; Washington, D.C USDHHS 2006 The Health Consequences of Involuntary Exposure to Tobacco Smoke: A Report of the Surgeon General U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, Coordinating Center for Health Promotion, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health Atlanta, GA USDHHS 2010 Ending the Tobacco Epidemic: A Tobacco Control Strategic Action Plan for the U.S. Department of Health and Human Services Washington, D.C Vardavas CI Anagnostopoulos N Kougias M Evangelopoulou V Connolly GN Behrakis PK 2012 Short-term pulmonary effects of using an electronic cigarette: impact on respiratory flow resistance, impedance, and exhaled nitric oxide Chest 141 1400 1406 http://dx.doi.org/10.1378/chest.11-2443 22194587 Wadland WC WinklerPrins VJ Noel MM Thompson ME Rios-Bedoya CF 2011 Student performance on smoking cessation counseling with standardized patients Fam Med 43 422 428 21656398 Wakefield M Coomber K Zacher M Durkin S Brennan E Scollo M 2015 Australian adult smokers’ responses to plain packaging with larger graphic health warnings 1 year after implementation: results from a national cross-sectional tracking survey Tob Control 24 ii17 ii25 http://dx.doi.org/10.1136/tobaccocontrol-2014-052050 Walsh SE Singleton JA Worth CT Krugler J Moore R Wesley GC Mitchell CK 2007 Tobacco cessation counseling training with standardized patients J Dent Educ 71 1171 1178 17761623 Wang L Jin Y Lu B Ferketich AK 2015 A cross-country study of smoking cessation assistance utilization in 16 low and middle income countries: data from the global adult tobacco survey (2008–2012) Nicotine Tob Res http://dx.doi.org/10.1093/ntr/ntv139 Warner KE 2016 Frequency of E-cigarette use and cigarette smoking by American students in 2014 Am J Prev Med http://dx.doi.org/10.1016/j.amepre.2015.12.004 Warner KE Tam J 2012 The impact of tobacco control research on policy: 20 years of progress Tob Control 21 103 109 http://dx.doi.org/10.1136/tobaccocontrol-2011-050396 22345232 West R Raw M McNeill A Stead L Aveyard P Bitton J Stapleton J McRobbie H Pokhrel S 2015 Health-care interventions to promote and assist tobacco cessation: a review of efficacy, effectiveness and affordability for use in national guideline development Addiction 110 1388 1403 http://dx.doi.org/10.1111/add.12998 26031929 White V Williams T Wakefield M 2015 Has the introduction of plain packaging with larger graphic health warnings changed adolescents’ perceptions of cigarette packs and brands? Tob Control 24 ii42 ii49 http://dx.doi.org/10.1136/tobaccocontrol-2014-052084 WHO 2003 WHO Framework Convention on Tobacco Control World Health Organization Geneva, Switzerland WHO 2006 WHO Named as one of five Partners to Implement Michael Bloomberg’s $125 million Initiative to Promote Freedom from Smoking World Health Organization Geneva WHO 2008 WHO Report on the Global Tobacco Epidemic, 2008: The MPOWER Package, Research for International Tobacco Control World Health Organization Geneva, Switzerland WHO 2011a Developing and Improving National Toll-Free Tobacco Quit Line Services: A World Health Organization Manual World Health Organization Geneva, Switzerland 106 WHO 2011b Scaling up Action against Noncommunicable Diseases: How Much will it Cost? World Health Organization Geneva WHO 2011c WHO Framework Convention on Tobacco Control: Guidelines for Implementation of Article 5.3, Articles 8 to 14 World Health Organization Geneva, Switzerland WHO 2012 Cardiovascular Diseases World Health Organization Geneva, Switzerland WHO 2013a MPOWER in Action: Defeating the Global Tobacco Epidemic World Health Organization Geneva, SZ WHO 2013b Tobacco Free Initiative: Bloomberg Initiative to Reduce Tobacco Use WHO Geneva, Switzerland WHO 2013c WHO Report on the Global Tobacco Epidemic, 2013: Enforcing Bans on Tobacco Advertising, Promotion and Sponsorship World Health Organization Geneva, SZ WHO 2015 WHO Report on the Global Tobacco Epidemic, 2015: Raising Taxes on Tobacco World Health Organization Geneva, SZ Wipfli HL Samet JM 2012 Moving beyond global tobacco control to global disease control Tob Control 21 269 272 http://dx.doi.org/10.1136/tobaccocontrol-2011-050322 22170336 World Economic Forum and the Harvard School of Public Health 2011l The Global Economic Burden of Non-Communicable Diseases World Economic Forum Geneva, Switzerland Yan J Xiao S Ouyang D Jiang D He C Yi S 2008 Smoking behavior, knowledge, attitudes and practice among health care providers in Changsha city, China Nicotine Tob Res 10 737 744 http://dx.doi.org/10.1080/14622200801901930 18418795 Zhu SH Lee M Zhuang YL Gamst A Wolfson T 2012 Interventions to increase smoking cessation at the population level: how much progress has been made in the last two decades? Tob Control 21 110 118 http://dx.doi.org/10.1136/tobaccocontrol-2011-050371 22345233
PMC005xxxxxx/PMC5125778.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 9107782 8548 Curr Biol Curr. Biol. Current biology : CB 0960-9822 1879-0445 27780070 5125778 10.1016/j.cub.2016.08.047 NIHMS831430 Article Evolution of highly diverse forms of behavior in molluscs Hochner Binyamin [email protected] Glanzman David L. [email protected] 1 Department of Neurobiology, Silberman Institute of Life Sciences, Hebrew University, Jerusalem, Israel 2 Department of Integrative Biology and Physiology, University of California Los Angeles, USA 3 Department of Neurobiology, David Geffen School of Medicine, University of California, Los Angeles, USA 22 11 2016 24 10 2016 28 11 2016 26 20 R965R971 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Members of the phylum Mollusca demonstrate the animal kingdom’s tremendous diversity of body morphology, size and complexity of the nervous system, as well as diversity of behavioral repertoires, ranging from very simple to highly flexible. Molluscs include Solenogastres, with their worm-like bodies and behavior (see phylogenetic tree; Figure 1); Bivalvia (mussels and clams), protected by shells and practically immobile; and the cephalopods, such as the octopus, cuttlefish and squid. The latter are strange-looking animals with nervous systems comprising up to half a billion neurons, which mediate the complex behaviors that characterize these freely moving, highly visual predators. Molluscs are undoubtedly special — their extraordinary evolutionary advance somehow managed to sidestep the acquisition of the rigid skeleton that appears essential to the evolution of other ‘successful’ phyla: the exoskeleton in ecdysozoan invertebrates and the internal skeleton in Deuterostomia, including vertebrates. A skeletal body provides stability and enables, through lever action, efficient exploitation of muscle forces for the generation of rapid movements. By contrast, molluscs must use their muscles for both body support and movement. Having a skeleton also simplifies motor control, because motor commands are limited to a rather restricted number of control variables (degrees of freedom) dictated by the number of joints. Indeed, except for cephalopods, molluscs are not renowned for their motor capabilities. The flexibility and speed of cephalopod motor behaviors have been achieved through radical changes in morphology, and the expansion of neuronal number by 4–5 orders of magnitude relative to other molluscs. An intriguing hypothesis is that the vast diversity of molluscan morphology and behavior has resulted from the lack of skeletal constraints. This idea is supported by the discovery that gene families important in setting body morphology (like the Hox family) have ‘lost’ their collinear patterns of expression in the two behaviorally most advanced molluscan classes — gastropods and cephalopods. Furthermore, the recent sequencing of the genome and multiple transcriptomes of the California two-spot octopus by Albertin et. al. revealed that cephalopods may have achieved their supremacy among invertebrates in motor and cognitive abilities through the expansion of two developmentally important gene families (C2H2s and protocadherins), extensive transposable element activity, and genome rearrangements, rather than through gains in the platforms of core genes. These findings suggest that the diversity of molluscs arose from their more ‘modular’ developmental frameworks, which allow greater variability and independence in selecting evolutionary successful solutions. In this Primer we first discuss how the diversity among molluscs illustrates the co-evolution of body morphology and the nervous system in order to accommodate, in an embodied way, the different levels of behavioral complexity. We then use the examples of simple and complex forms of learning and memory in Aplysia californica and Octopus vulgaris to demonstrate the diversity of the learning-related molecular and cellular mechanisms in molluscs. Coevolution of nervous system, body morphology and behavior in molluscs A review of the main molluscan groups demonstrates how the anatomy of the nervous system has adapted to body morphology and lifestyle. The nervous system of the chitons, wormlike, shell-less molluscs (Figure 1), is organized in a ladder-like fashion of medullary cords, with neurons distributed along the cords. Such an arrangement permits better control of the locomotory behavior of these animals. This local organization is not true segmentation, suggesting that it is due to convergent evolution rather than to shared ancestry. This same principle seems to be general to all Mollusca groups, as demonstrated in Figure 1. In scaphopods, gastropods, and bivalves of the Conchifera (shell-bearing) class (Figure 1), the nervous system consists of distributed ganglia (clusters of nerve cells) that are interconnected by axonal fibers that run between the ganglia in connectives and commissures (Figure 1). In these animals, each ganglion is located close to the affector (sensory) or effector (motor) organs. In cephalopods, the central brain is organized into a set of closely interconnected ganglia (lobes), which retains the typical invertebrate organization of a circumesophageal ring (Figure 1), and comprises about 40 interconnected lobes, each with a more or less specific function. The lobes show the common structure of invertebrate ganglia: an outer layer of neuronal cell bodies, from each of which projects a single neurite that passes into a central neuropil, where it ramifies into dendritic and/or axonal terminal trees. However, a recent study of cephalopod brain neurogenesis reported that this brain is not embryologically derived from a ganglionic organization. Moreover, the peripheral nervous system of the Octopus arm, which contains two-thirds of the half billion neurons of the Octopus nervous system, is organized similarly to other molluscs. Specifically, the axial nerve cord of the arm, which controls the intrinsic muscular system, resembles the medullary cord organization in Chitons (Figure 1), probably to better regulate the wave-like arm movements characteristic of the Octopus. On the other hand, each of the 200–300 arm suckers has its own ganglion that mediates autonomous sensory-motor functions. Finally, a specific organization is not restricted by phylogenetic class. In the monoplacophorans, like Neopilina (Figure 1), which also belong to Chonchifera, the adult nervous system resembles the medullary cords typical of chitons in the Aculiferan (scale-bearing) group (Figure 1). Relationships among nervous systems, bodies and levels of behavior complexity The behavior of all molluscs, albeit to a lesser degree in cephalopods, involves simple reflexes that are mediated by Network A, schematized in Figure 2. Perhaps the best-known example is the defensive withdrawal response to a tactile stimulus. When touched, the animal retracts its body into its shell or, in the case of the gastropods, which have lost their shells, contracts its extended organs (tentacle, gill, siphon, etc.). This defensive withdrawal reflex (DWR) has been studied extensively in the marine snail Aplysia californica by Eric Kandel and others. (Kandel was awarded the Nobel Prize for Medicine in 2000 for his research in Aplysia.) This simple behavior, which is mediated by a single sensory modality, touch, shows non-associative and associative short- and long-term modulation, and is expressed in the same neurons that directly produce the reflex (Figure 2, Network C). Two molluscan classes, Bivalvia and Polyplacophora (chitons), are protected by shells and scales, respectively. These have evolved a unique defensive reflex that is mediated by simple, light-detecting eyes distributed along the mantle or within the shell. Instead of evolving complex computation, the optical properties of these eyes have been selected to detect specific visual parameters or features (light intensity or direction and velocity of passing shadows). The physical properties of the eye in these animals allow it to act as an optical alarm system that triggers, for example, a reflexive closure of the two valves of the ark clam (Figure 2, Network A). A somewhat more sophisticated level of defensive behavior is escape behavior, in which a tactile, light or odor stimulus evokes whole animal behavior (Figure 2, Network B). In this defensive behavior, particularly prominent in gastropods that have lost their protective shells (e.g. nudibranchs), the stimulus may activate a central pattern generator (CPG) that drives escape swimming. Two of the molluscan classes, the gastropods and cephalopods, possess true eyes and a well-developed head (cephalic) region. The development of the head is associated with the goal-directed locomotion shown in these groups. The directional information about the external world can be extracted through the visual and/or olfactory system (Figure 2, Network D). To direct the animal’s locomotion, the sensory receptors are concentrated in a pair of organs at the front of the animal’s head. The stereoscopic sensory information is processed in the cerebral ganglion or ‘brain’, whose size and associated structures correlate with the total potential amount of sensory input from the visual and the olfactory systems. A simple computational algorithm, which compares the chemical or light intensities between each of the two cephalic receptors and then computes the weighted motor outputs (Network D), may suffice for a simple goal-directed movement. The next level of behavioral complexity arises when behavior is based on the quality of the sensory signals rather than on their intensity. Here the ‘physical’ solution lies in having receptors for each modality (for example, an odor-specific sensory neuron) and a neural network that can process this multi-channel information. Certain gastropods, like the terrestrial Helix pomatia, possess a specific lobe-like structure, the protocerebrum, that can analyze the quality of incoming olfactory information. The protocerebrum has the glomerular organization that characterizes the neural structures dedicated to olfactory processing in nematodes, arthropods and vertebrates (Figure 2, Network E). This emphasizes the importance of specialized anatomical features within the nervous system for a specific computational function. Molluscan ‘image-forming’ eyes have evolved only in cephalopods. Seeing the external world with a camera-like eye requires projecting the image of the external world onto a two-dimensional array of photoreceptors (such as in the retina) whose density determines the level of optical resolution. Such camera-like eyes have been achieved by convergent and independent evolution in cephalopods and vertebrates. (In Nautilus the optical mechanism for projecting the image onto the retina circumvents the need for a lens by using pinhole optics). Analysis of the genetic basis of the independent evolution of these camera-like eyes indicates that about 70% of the octopus genes are conserved in the human eye. This suggests that the evolution of camera-like eyes used a limited number of protein-coding genes that were already present in our last common bilaterian ancestor, which lived 660–680 million years ago. The cephalopods’ camera-like eyes marvelously illustrate convergent evolution. Yet, the electrical responses of photoreceptors to light in the retinas of cephalopods and vertebrates differ, suggesting that there was no corresponding evolutionary pressure for invertebrates and vertebrates to converge on the same mechanism of visual transduction. Thus, neuronal network evolution and development are highly adaptive; the mechanisms necessary to achieve the same level of information processing can evolve in a variety of ways. As we show below, the synaptic mechanisms of learning and memory in two cephalopods, the Octopus and cuttlefish (Sepia officinalis), provide further support for this idea. Evolution of learning and memory in molluscs — extreme diversity in molecular solutions In this section, we address the variability in the neuronal mechanisms that mediate simple and complex behavioral flexibility in two well studied molluscs, Aplysia californica and Octopus vulgaris. In particular, we compare forms of synaptic plasticity underlying simple forms of learning and memory exhibited by the defensive withdrawal reflex (DWR) of the gastropod Aplysia with those expressed by the vertical lobe (VL) system of the Octopus. Network organization and learning-related plasticity of the DWR The core neuronal circuit regulating the DWR comprises only two types of neurons (Figure 2, Network A) — sensory and motor neurons. The network gains greater behavioral flexibility from a comparatively restricted population of interneurons and modulatory interneurons that are also activated by the sensory neurons (Figure 2, Network C; Figure 3A). Significantly, learning-related neurobiological changes in the DWR system are predominately restricted to the relatively small number of neurons that comprise the reflex circuit itself. The Aplysia DWR is evoked by activation of primary sensory neurons — the receptors of which reside in the animal’s skin — that respond to mechanical pressure. The most basic component of behavioral plasticity in the simple DWR network results from intrinsic synaptic alterations, including homosynaptic depression and homosynaptic facilitation, of the monosynaptic connection between the sensory neurons and the motor neurons. Homosynaptic plasticity in the DWR circuit Homosynaptic plasticity is induced at a synapse through its autonomous activity. In the DWR network this form of plasticity is best demonstrated by activity-dependent synaptic depression. Homosynaptic depression occurs upon repeated low frequency (at rates greater than ~1 per 3 min) activation of the sensory neuron, and is due to a progressive decrease in release of glutamate, the sensory neuron transmitter. This mechanism ensures that only the sensory neurons that are repeatedly activated by a weak, invariant tactile stimulus ‘learn’ to ignore the stimulus. Homosynaptic depression mediates short-term (< 1 h) habituation of the DWR. Another important form of activity-dependent homosynaptic plasticity of the sensory-motor synapse is long-term potentiation (LTP). This form of synaptic enhancement, induced by synchronous pre- and postsynaptic activity, was first proposed in 1949 as a mechanism for associative learning by the Canadian psychologist Donald Hebb. Hebbian LTP — initially discovered in the hippocampus, a brain structure known to be critically important for many forms of associative learning and memory — is mediated by N-methyl-D-aspartate- (NMDA)-type glutamate receptors (NMDARs). Hebbian/NMDAR-dependent LTP is not unique to the hippocampus; it has been described at synapses in other regions of the vertebrate nervous system, and is prominent at the Aplysia sensory–motor synapse. As its name suggests, LTP is a persistent form of synaptic plasticity and is thought to mediate types of learning and memory that last for hours. In Aplysia, NMDAR-dependent LTP mediates classical conditioning of the DWR. It is induced, for example, by siphon stimulation, the conditioned stimulus (CS), which activates the sensory neuron, paired with electrical shock of the tail, the unconditioned stimulus (US), which depolarizes the motor neuron through an excitatory interneuronal pathway. The simultaneous presynaptic activity and postsynaptic depolarization causes the opening of postsynaptic NMDAR channels, thereby producing LTP of the sensory–motor synapse. Heterosynaptic plasticity in the DWR circuit In the DWR network, the sensory neurons activate the motor neurons and, in parallel, also activate interneurons, some of which release various neuromodulatory transmitters. Some of these modulatory interneurons re-innervate the sensory neurons (Figure 2, Network C; Figure 3A). The best studied modulatory interneurons in Aplysia are the serotonergic facilitatory interneurons; these synapse onto the presynaptic terminals of the sensory neurons, as well as the cell bodies of both sensory and motor neurons. The neuromodulatory transmitters mediate various forms of short- and long-term heterosynaptic modulation of the sensory-motor synaptic connection and of the excitability of the sensory and motor neurons. Serotonin (5-HT) is released from the terminals of the facilitatory interneurons onto both the sensory and motor neurons in response to noxious or arousing stimulation. Because the sensory neurons mediate the perception of a tactile stimulus that triggers the DWR and of a noxious stimulus (in the laboratory this is commonly a shock applied to the animal’s tail), what differentiates a noxious stimulus from a harmless stimulus is the amount of released 5-HT (Figure 2, Network C; Figure 3A). This is a function of the number and intensity of the sensory cells activated. The forms of synaptic plasticity mediated by 5-HT are referred to as ‘heterosynaptic’ because they depend critically on mechanisms extrinsic to the sensory-motor synapse. Heterosynaptic, 5-HT-dependent facilitation of the sensory-motor synapses mediates behavioral enhancement of the DWR, most prominently, sensitization and dishabituation. Dishabituation, the enhancement of a habituated response, is distinct from sensitization because it requires a preceding form of learning (habituation); in contrast, sensitization is the enhancement of a nonhabituated response. Serotonin appears to mediate both forms of learning through the parallel activation of somewhat distinct biochemical and molecular pathways. Thus, sensitization involves 5-HT-dependent activation of cAMP-dependent kinase A (PKA), which phosphorylates and blocks specific types of K+ channels in the sensory neuron; the blockage of the K+ channels, in turn, causes prolongation of the presynaptic action potential. The broader action potential increases Ca2+ entry into the presynaptic terminal, thereby enhancing transmitter release. Dishabituation, by contrast, involves activation of presynaptic protein kinase C (PKC). Interestingly, the mechanism of presynaptic facilitation during dishabituation is independent of the broadening of the presynaptic action potential. Rather, 5-HT appears to facilitate synapses depressed during habituation through a different mechanism, possibly involving PKC-stimulated enhancement of presynaptic vesicle mobilization, which serves to counteract the vesicle depletion that occurs during habituation. In addition to the nonassociative forms of heterosynaptic facilitation that underlie behavioral sensitization and dishabituation, the DWR circuit exhibits an associative form of heterosynaptic facilitation, activity-dependent facilitation, which mediates classical conditioning together with NMDAR-dependent LTP. Activity-dependent facilitation (ADF) is produced by activation of the sensory-motor synapse in the presence of 5-HT. During classical conditioning, the CS-induced firing of the sensory neuron occurs in conjunction with release of 5-HT from facilitatory interneurons due to the US (tail shock), resulting in the enhancement of the biochemical cascade that mediates 5-HT’s effect on transmitter release. The Aplysia sensory–motor synapse exhibits a greater variety of plasticity than any other synapse known to neurobiology. Why is this so? A likely possibility is that the number of neurons in the DWR circuit is relatively restricted, and the same circuit that mediates the reflex also is used for storing short- and long-term memory traces. Consequently, each synaptic site must carry a greater ‘cognitive load’ if the animal is to express the full panoply of learning and memory forms that are expressed by Aplysia. Another important question is why certain forms of learning require multiple synaptic plasticity mechanisms. An illustrative example is classical conditioning of the DWR, which involves both NMDAR-dependent LTP and ADF. Possibly, NMDAR-dependent LTP is specialized for the mediation of postsynaptic modifications while ADF is better suited for modification of presynaptic properties. The VL association network: the role of homosynaptic plasticity The cephalopod VL is organized as a ‘fan-out fan-in’ matrix of connections among an extremely large number of neurons (~25,000,000). In contrast to the DWR circuit, the VL has evolved to deal efficiently with learned associations among the several sensory modalities that cephalopods use and is organized to work in parallel to the pathway that controls the motor behavior (Figure 2, Network F; Figure 3B). The sensory information of each sensory modality is first processed in feature-detecting networks (e.g. Figure 2, Network D) or in the optic lobes (see the Octopus brain anatomy in Figure 1), after which it undergoes further categorization in the superior frontal lobe (SFL) (see scheme in Figure 3B). The information is then transferred to the VL, where it is represented by sparse synaptic connections between the 1.8 million axons of SFL neurons entering the VL and the 25 million amacrine interneurons (AM) that comprise the input (‘fan-out’) synaptic layer of the VL. The AMs converge sharply onto the second synaptic layer (‘fan-in’) of large efferent neurons (LNs), the only output of the VL. The organization of the VL is similar to that of the insect mushroom body — which, like the VL, is involved in associating sensory information of different modalities — as well as the mammalian hippocampus. The cellular mechanism that evolved for forming associations among sensory stimuli in the Octopus VL is LTP. In the Octopus VL only those synapses in which the presynaptic terminals are intensely activated undergo LTP. The LTP in the Octopus VL occurs at the SFL-to-AM glutamatergic synapses (Figure 3B), whereas in Sepia it occurs at the cholinergic connections between the AMs and the large neurons that transmit the output of the VL to other brain regions. This difference represents a good example of versatility in the organization of different learning and memory networks in phylogenetically close animals, and suggests that the cellular and molecular mechanisms in molluscan nervous systems are disposed to significant variations. Mechanistically, LTP in the VL differs from that in the DWR circuit. Whereas LTP of the Aplysia sensory-motor synapse is mediated by postsynaptic NMDARs, LTP of the SFL–AM synapses in the Octopus VL is not blocked by standard NMDAR antagonists. On the other hand, LTP in about half of SFL–AM synapses in the VL is mediated by AMPA-type glutamatergic receptors, because it is blocked, by AMPA/kainate receptor antagonists. In fact, the LTP at the SFL–AM synapses of the Octopus appears to resemble the LTP of the glutamatergic synapses between the mossy fibers and CA3 neurons in the mammalian hippocampus, which is also NMDAR-independent and presynaptically expressed. Neuromodulation in the VL Similar to its effect on the sensory-motor synapses of Aplysia, 5-HT induces short-term presynaptic facilitation of the glutamatergic input to the AMs. Unlike in Aplysia, however, prolonged application of 5-HT does not lead to a long-term synaptic facilitation (LTF). But 5-HT indirectly reinforces LTP induction in the VL through its short-term facilitatory effects, suggesting a possible role in transmission of a reward signal into the VL. Mechanisms of long-term plasticity in Aplysia and the Octopus Perhaps the mechanistically best understood form of long-term synaptic memory, long-term facilitation (LTF) of the sensory-motor synapse can be induced by repeatedly exposing sensory-motor cocultures to spaced pulses of 5-HT. This in vitro ‘training’ is designed to approximate in vivo training, whereby repeated activation of the facilitatory interneurons by strong electrical shocks applied to the animal’s skin causes repeated, pulastile release of 5-HT, and thereby persistent sensitization of the DWR. The repeated exposure to 5-HT activates cAMP-dependent early genes within the sensory and motor neurons, which in turn triggers the protein synthesis-dependent long-term structural changes that underlie LTF. The studies of LTF in Aplysia were the first to show the involvement of cAMP-inducible genes in long-term synaptic enhancement and long-term memory (LTM). This mechanism is universally found to mediate LTM in invertebrates and mammals. It is therefore surprising that the persistent LTP in the association network of the Octopus VL appears independent of protein synthesis. Instead, it appears to employ a ‘molecular switch’, using a covalent state modification of existing molecules to maintain the long-term synaptic change. This molecular switch has evolved through adaptation of the nitric oxide (NO) system that, in invertebrates, mediates various forms of behavioral plasticity, such as feeding-related learning in the gastropod molluscs Lymnaea and Aplysia. In the Octopus VL LTP induction is not mediated by NO, because induction is not blocked by nitric oxide synthase (NOS) inhibitors. Nonetheless, such inhibitors transiently block the presynaptic expression of LTP. This suggests that the induction of LTP at the SFL–AM synapses involves activity-dependent NOS stimulation, likely in the AM neurons; the postsynaptically synthesized NO then diffuses retrogradely to produce the increase in probability of glutamate release that mediates the expression of LTP at these synapses. Recent results suggest a novel ‘self-activation’ mechanism, whereby NO can maintain LTP for more than 10 h by NO mediated reactivation of NOS. Why does LTF in Aplysia use a mechanism that depends on protein synthesis, while LTP in the Octopus VL does not? This distinction may be explained by differences in the organization of the two learning and memory systems (Figure 3). As suggested earlier, in the DWR network the memory is stored within the circuit that mediates the behavior; by contrast, the VL in the Octopus regulates the acquisition of memory, which is ultimately stored outside the VL. Therefore, in the Octopus the molecular memory switch in the VL must persist only long enough to maintain the VL memory until completion of the protein synthesis-dependent memory consolidation, possibly within the circuits that mediate Octopus behavior. In summary, although the simple DWR circuit of Aplysia and the dedicated learning and memory system in the VL of the Octopus and Sepia make use of functionally similar cellular forms of short- and long-term synaptic plasticity, there are significant differences in the specific molecular mechanisms used by these three systems to mediate long-term synaptic plasticity and LTM. This fact points to a high degree of variability in the evolutionary selection within and/or the ontological development of the learning systems of Aplysia and the Octopus. This variability suggests that the special expansion of developmental and regulatory gene families, together with the unique cephalopod expansion of posttranslational mechanisms, provide the molecular flexibility required to achieve the different forms of neuronal plasticity necessitated by the disparate behavioral repertoires exhibited by molluscs. We would like to acknowledge Yosef Yarom and Clifton Ragsdale for their valuable comments and suggestions. Figure 1 Diversity among molluscan nervous systems The extreme diversity of molluscan nervous systems shown in relation to their position in the phylogenetic tree (taken from Stögere et al., 2013). The examples illustrate similarities among different groups, as well as the differences within the same group (see text). The derivatives of the pedal ganglia are shown in red, the components of the pleural-parietal ganglia are shown in green, and the cerebral ganglia are uncolored (Figures adapted from Bullock & Horridge 1965, except Octopus brain, which is taken from Brusca & Brusca 1990) with permission of Sinauer Associates.. Figure 2 The neural bases of Molluscan behavioral diversity Network schemes showing the correlation between the level of complexity of molluscan behaviors and the number of sensory modalities involved in the behaviors (see text). The examples range from a simple defensive reflex (Network A) to the associative learning and memory network of the octopus (Network F). The neurons are schematized as a typical invertebrate monopolar neuron. Figure 3 Wiring diagrams for a simple and a complex learning and memory network The network organization of the defensive withdrawal reflex of Aplysia (A) and the associative learning and memory network of the Octopus vertical lobe (B). The vertical lobe comprises a ‘fan-out fan-in’ pattern of neural connectivity, which is also characteristic of the insect mushroom body and of an artificial machine learning classifier. VL, vertical lobe; SFL, superior frontal lobe; SFLn, superior frontal lobe neuron; AM, amacrine interneurons; LN, large efferent neuron; Ach, acetylcholine; Glut, glutamate. FURTHER READING Albertin CB Simakov O Mitros T Wang ZY Pungor JR Edsinger-Gonzales E Brenner S Ragsdale CW Rokhsar DS 2015 The octopus genome and the evolution of cephalopod neural and morphological novelties Nature 524 220 224 26268193 Brusca & Brusca 1990 Invertebrates 1 Sunderland Sinauer Associates Byrne JH Hawkins RD 2015 Nonassociative learning in invertebrates CSH Persp Biol 7 Eisthen HL 2002 Why are olfactory systems of different animals so similar? Brain Behav Evol 59 273 293 12207084 Glanzman DL 2010 Common mechanisms of synaptic plasticity in vertebrates and invertebrates Curr Biol 20 R31 R36 20152143 Haszprunar G Wanninger A 2012 Molluscs Curr Biol 22 R510 R514 22789994 Hawkins RD Byrne JH 2015 Associative learning in invertebrates CSH Persp Biol 7 Hochner B 2012 An Embodied View of Octopus Neurobiology Curr Biol 22 R887 R892 23098601 Kandel ER Dudai Y Mayford MR 2014 The molecular and systems biology of memory Cell 157 163 186 24679534 Kocot KM Cannon JT Todt C Citarella MR Kohn AB Meyer A Santos SR Schander C Moroz LL Lieb B Halanych KM 2011 Phylogenomics reveals deep molluscan relationships Nature 477 452 456 21892190 Ogura A Ikeo K Gojobori T 2004 Comparative analysis of gene expression for convergent evolution of camera eye between Octopus and human Genome Res 14 1555 1561 15289475 Shigeno S Parnaik R Albertin CB Ragsdale CW 2015 Evidence for a cordal, not ganglionic, pattern of cephalopod brain neurogenesis Zoo Lett 1 1 13 Shomrat T Graindorge N Bellanger C Fiorito G Loewenstein Y Hochner B 2011 Alternative sites of synaptic plasticity in two homologous fan-out fan-in learning and memory networks Curr Biol 21 1773 1782 22018541 Shomrat T Turchetti-Maia AL Stern-Mentch N Basil JA Hochner B 2015 The vertical lobe of cephalopods: an attractive brain structure for understanding the evolution of advanced learning and memory systems J Comp Physiol 201 947 956 26113381 Stöger I Sigwart JD Kano Y Knebelsberger T Marshall BA Schwabe E Schrdl M 2013 The continuing debate on deep molluscan phylogeny: evidence for Serialia (Mollusca, Monoplacophora + Polyplacophora) BioMed Res Internat 2013 18
PMC005xxxxxx/PMC5125820.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 9516420 21005 Clin Microbiol Infect Clin. Microbiol. Infect. Clinical microbiology and infection : the official publication of the European Society of Clinical Microbiology and Infectious Diseases 1198-743X 1469-0691 27619640 5125820 10.1016/j.cmi.2016.09.001 NIHMS824121 Article Therapeutic Manipulation of the Microbiota: Past, Present and Considerations for the Future Young Vincent B. MD/PhD [email protected] Department of Internal Medicine/Infectious Diseases Division, Department of Microbiology & Immunology, University of Michigan Medical School, Ann Arbor, MI 48109 29 10 2016 10 9 2016 11 2016 01 11 2017 22 11 905909 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. With the growing appreciation of the potential role of the indigenous microbiota in a variety of disease states there has been a concomitant interest in manipulating the microbiome for therapeutic effect. The most successful example of microbiota manipulation for the treatment of a disease is in the setting of recurrent infection with the bacterial pathogen Clostridium difficile. This review will provide historic perspectives on the development of microbiota transplantation and review the evidence for its use in recurrent C. difficile infection. Recent work has pointed to potential mechanisms by which microbiota restoration in the form of fecal transplantation has been efficacious. This includes studies of the microorganisms that are associated with successful fecal transplantation in human and animal studies and the focus on bacterial bile acid metabolism as a mechanism that mediates colonization resistance against the pathogen. The potential use of microbiota manipulation for other diseases such as the inflammatory bowel diseases and metabolic disorders will be discussed. The case will be made that the lessons learned from treatment of recurrent C. difficile infection may not necessarily translate to the use of fecal transplantation or other methods to alter the microbiome for the treatment of other diseases. A key conclusion that can be drawn from the review of the results of the use of fecal transplantation as a therapeutic modality is that understanding of the precise role of the microbiota in the pathogenesis of a specific disease is necessary prior to determining if microbiota manipulation represents a novel treatment therapy. Introduction Recent interest in microbiota manipulation There has been an explosion of interest in the role the intestinal microbiota, the complex community of microorganisms that inhabits the intestinal tract, in human health and disease (see Table). Corresponding to the increased attention of the role of the intestinal microbiota in influencing host health, there has been intense exploration of potential means to manipulate the intestinal microbiome to improve health. Perhaps the most visible form of this form of therapy is the use of fecal transplantation to treat recurrent infection with the bacterial pathogen Clostridium difficile [1]. The remarkable success of fecal transplantation to treat recurrent C. difficile infection (CDI) has prompted exploration of this therapy as potential treatment for a wide variety of illnesses including those outside the gastrointestinal tract [2, 3]. With regards to recurrent CDI itself, as of June 9, 2016 a PubMed search of the terms ([fecal transplant OR fecal transplantation] AND difficile), yielded 415 articles, 132 of which were reviews. It is interesting to note that 410 of these have been published in the past 6 years. With all of this recent interest, why publish another review of this topic? Given the significant interest, it is probably appropriate at this point to pause and reflect on the status of the field both in terms of the history of microbiota transplantation, the current practice especially with regards to recurrent CDI and perhaps most importantly look forward to upcoming treatments for recurrent CDI as well as other microbiota-related illnesses. While many reviews have been published on the technical aspects of fecal transplantation and the results of clinical trials, it is hoped that this review will provide a broader context with which to evaluate where we have been, where we are now and where we may head in the future with regards to manipulation of the indigenous microbiota. History Current discussions of microbiota manipulation through the use of fecal transplantation often invoke descriptions from the Roman philosopher/scientist Pliny the Elder and the traditional Chinese medicine practitioners Ge Hong and Li Shizhen. Although there may be some debate on the accuracy and provenance of these claims[4], it is clear that the idea of administration of feces, and whether intentional or not, the organisms contained within, as a treatment for gastrointestinal illness dates to antiquity. It is clear that the "modern era" of fecal transplantation as a therapy can be traced to the 1958 report by Eisman and colleagues who successfully treated four cases of pseudomembranous colitis associated with antibiotic treatment through the use of fecal transplant administered via enema [5]. It is interesting to note that this was two decades before Koch's postulates (see Table) were fulfilled to implicate C. difficile [6] as the cause of pseudomembranous colitis. At the time it was suggested that Staphylococcus aureus was a potential cause as all four of these patients had this organism isolated from their feces. These investigators attempted to develop a canine model of pseudomembranous colitis through the administration of Staphylococcus aureus to antibiotic treated animals but were unsuccessful [5]. The mention of Robert Koch reminds us that our understanding of infectious diseases has been influenced over the past 150 years by the rise of the "germ theory" as advanced by Koch based on the groundbreaking work of Louis Pasteur [7]. Although Koch’s Postulates were initially used to implicate bacteria as causes of infectious diseases, advances in laboratory science have allowed us to implicate viruses and fungi as etiologic agents. The development of molecular biological techniques even prompted the development of the so-called molecular Koch's postulates whereby the role of specific genetic loci in the pathogenesis of infectious diseases could be tested [8, 9]. All of these advances have prompted the scientific community to adopt the notion that the use of a reductionist approach to study pathogens would unravel all of the mysteries of infectious diseases. The past 20 years however, has seen another revolution in our understanding of the role of microbes in the causation of disease and just as importantly, the maintenance of health. Taking cues from microbiologists who had been studying the structure and function of complex microbial communities in natural environments such as soil and seawater, investigators started studying how host-associated microbial communities could affect their symbiotic partners. Through efforts such as the Human Microbiome Project and the MetaHIT initiative, application of modern technology has allowed detailed study of these microbial communities in many cases without the requirement of cultivation [10, 11]. Given these two contrasting yet complementary views of the role of microbes in helping maintain human health on the one hand and as causative agents of disease on the other, it is interesting to consider that C. difficile infection represents an ideal opportunity to apply both germ theory and concepts of the microbiome to a single clinical entity. Koch’s postulates were first fulfilled for C. difficile using a hamster model in 1977 [6]. At first glance this appears like a standard application of these postulates for a presumed infectious disease and indeed, this initial report was quickly followed by others that implicated the same organism in the etiology of pseudomembranous colitis [12, 13]. However, it is important to note that the "susceptible host" for testing Koch’s postulates for C. difficile was only susceptible after the administration of antibiotics. The obvious implication of this observation is that the antibiotic administration altered the indigenous intestinal microbiota in a manner that permitted colonization and expansion of C. difficile. In his 1958 case series Eisman proposed that "most of the recently reported cases [of pseudomembranous colitis] have followed the use of oral broad-spectrum antibiotics, suggesting that the intestinal flora was thus altered to permit the overgrowth of antibiotic-resistant Micrococcus progenies (staphylococcus) within the gut.”[5] Even if the implicated pathogen was not what we currently recognize as the most common etiologic agent for pseudomembranous colitis, the implication that alteration of the indigenous microbiota is important in the pathogenesis of this disease was clear to these earlier clinicians. Thus there are two roles for microbes in the pathogenesis of pseudomembranous colitis. On one hand you have the “classic” pathogen C. difficile, which has a number of pathogenic features that play a role in the etiology of disease such as the production of toxin [14]. On the other hand, the indigenous microbial community of the gut has an equally important role. It appears that the normal gut microbiota mediates a colonization resistance (see Table) against C. difficile, or at least limits the ability of C. difficile to result in disease in the event that is does colonize [15]. As a corollary, when the indigenous microbiota is disturbed, for example following the administration of antibiotics, this colonization resistance is destroyed and when exposed to the pathogen, colonization and disease ensues. Current studies of the microbiome in disease pathogenesis Having established C. difficile infection as a model system in which microbes play a dual role in the pathogenesis of disease (the pathogen on one hand, and the indigenous microbiota on the other) we will now turn our attention to current studies that are investigating the mechanisms by which these microbes interact. This naturally leads to a discussion of the use of microbiome transplantation as a therapeutic modality. However, as noted above there are a number of reviews as to the clinical utility and procedures used for microbiota manipulation in the form of fecal microbiota transplant (See annotated bibliography and [1, 2]) and thus this topic will not be discussed in detail here. Shortly after Koch’s postulates were fulfilled using a hamster model of experimental C. difficile infection, it was demonstrated that transfer of an intact microbiota could protect hamsters against experimental C. difficile challenge in antibiotic treated animals [16]. Administration of a cecal homogenate obtained from non-antibiotic treated hamsters would protect the majority of antibiotic-treated animals from developing a fatal colitis due to C. difficile. In the same study study, the requirement for the transfer of viable microbes was demonstrated as either heat-treated or filtered cecal homogenate was not effective. A number of approaches have been used to identify the specific organisms present in the intestinal microbiota that mediate colonization resistance. Since the wide application of culture-independent methods for the study of the microbiome [17], multiple studies have examined which microbes are associated with susceptibility or resistance to colonization with C. difficile. Studies in both humans and animal models of infection have noted that lowered overall diversity and decrease in microbial biomass of the intestinal microbiota is associated with loss of colonization resistance [18, 19]. While this observation is quite common, it should be noted that diversity in and of itself does not necessarily correspond with a specific functional state of the microbiome. In fact, many microbiome-associated diseases are characterized by the presence of a microbiota noted to have lower than normal diversity. It is likely that in many cases this is simply a reflection of a disturbed community structure and therefore disturbed function, generally loss of a beneficial function but potentially the abnormal presence of a deleterious function. Recurrent C. difficile infection is also associated with reduced microbiota diversity [18] and several investigators have noted that successful fecal transplantation is associated with a corresponding increase in intestinal microbiome diversity [20, 21]. As a corollary to the discussion above that decreased microbial diversity doesn’t necessarily result in abnormal microbiota function, increased diversity in and of itself does not necessarily result in with normal function. In our study of 14 patients with recurrent CDI who were treated with FMT, two patients who were successfully treated had continued reduced diversity while a patient who failed initial treatment had an increase in fecal microbiota diversity despite not having a return of normal function [20]. In a manner analogous to the use of Koch’s postulates for infectious disease causation, a number of investigators have searched for specific microbes that can mediate colonization resistance against C. difficile. Wilson and Freter conducted one of the initial studies examining the ability of specific members of the normal microbiota to interfere with C. difficile colonization [22]. In a series of studies using germ-free mice, they demonstrated that a wide variety of anaerobes appeared to be able to restore colonization resistance. In this study, two forms of hamster-derived microbes were tested. Complex communities derived from the cecae of normal hamsters were maintained in continuous flow cultures. Additionally, these investigators generated a mixture of anaerobes isolated by plating cecal contents on rich media and incubation under anaerobic conditions was generated. Both the naturally occurring complex community and the “synthetic” community (which consisted of up to 150 bacterial isolates) were able to significantly suppress (but not eliminate) colonization of germ free mice with C. difficile [22]. Our group reported that a single member of the family Lachnospiraceae (with a 16S gene most closely related to that of C. clostridioforme, a member of the normal intestinal microbiota of humans which can be a rare cause of invasive diseases such as bacteremia [23, 24]) could partially restore colonization resistance against C. difficile. Mono colonization of germ-free mice with this murine-derived strain significantly reduced the level of colonization with C. difficile, decrease the amount of toxin present in the gut and reduce overall mortality from experimental infection [25]. Lawley and colleagues reported on a mixture of six microorganisms that could fully restore colonization resistance in a murine model of C. difficile infection [26]. Other recent studies have started to define several of the molecular mechanisms by which specific members of the microbiota can interfere with colonization and growth of C. difficile within the intestine. One area of intense research is the role of microbe-mediated metabolism of bile acids (reviewed in [27]). Bile acids are secreted by the liver in a form where they are conjugated to either the amino acids glycine and taurine. These conjugated, primary bile acids are found in high concentration in the duodenum and small intestine. The importance of this is that the primary bile acid taurocholate is potent inducer of germination for C. difficile spores [28, 29]. Once conjugated, primary bile acids are present in the lumen of the small intestine, the can be acted on by members of the resident microbiota that posses hydrolases that can cleave the amino acid moiety, releasing the unconjugated form of the bile acid [27]. Still other microbes metabolize these unconjugated primary bile acids, for example through the activity of dehydroxylases, resulting in the formation of secondary bile acids such as deoxycholate and lithlocholate. Again, these secondary bile acids can influence the physiology of C. difficile within the intestine. In particular the secondary bile acid deoxycholate is quite inhibitory to the vegetative form of the pathogen [29, 30]. Therefore disruption of the intestinal environment by antibiotics can result in a significant alteration of bile acid metabolism, which in turn will affect the physiology of C. difficile in the gut. For example, loss of organisms with bile salt hydrolases and dehydroxylases can result in an accumulation of taurocholate, facilitating germination of C. difficile spores and also decrease levels deoxycholate with the concomitant loss of growth inhibition of vegetative C. difficile [31]. Sorg and Sonenshein proposed that restoration of bile acid metabolism could be a novel means to control this pathogen [32]. Indeed, Pamer and colleagues recently demonstrated that the targeted restoration of bile acid metabolism by the intestinal microbiota could restore colonization resistance [33]. It is likely that other mechanisms can contribute to the colonization resistance mediated by the normal intestinal microbiota. Future directions While the use of fecal microbiota transplantation is quite effective in the setting of recurrent C. difficile there is considerable effort being invested in trying to improve on this relatively crude method for microbiome manipulation. A number of investigators are trying to process fecal material to obtain a more refined therapeutic product [34]. One approach employed is to focus on the spore-forming organisms that are present in feces [35]. Perhaps the ultimate refinement of fecal transplantation would be defined mixtures of well-characterized bacterial isolates that carry out precise metabolic functions, for example bile acid metabolism, that are important for mediating colonization resistance against C. difficile. One could imagine a scenario by which the intestinal microbiota of a patient with recurrent C. difficile is analyzed for altered microbiome functions. Following this analysis, the microbiome could then be restored by the targeted administration of microbes that can restore these missing functions. This form of “personalized microbiome restoration” could help alleviate some of the concerns that the administration of bulk stool could inadvertently have long-term negative impacts on the overall function of the microbiota important for other homeostatic functions. The success of the use of microbiome manipulation to treat recurrent CDI has been accompanied by an increasing number of attempts to use fecal microbiota transplantation to ameliorate a wide variety of other diseases and conditions [2]. Inflammatory bowel disease IBD, obesity and diabetes are just a few of the conditions for which microbiome manipulation has been proposed and at least preliminary studies conducted. While it is beyond the scope of this work to comprehensively review these other applications of FMT, it is fair to suggest that at this point in time, that the use of FMT for these other applications such as treatment of IBD have not as successful as for recurrent CDI [36, 37]. In this respect, it is instructive to consider why this may be true as we continue to investigate the role of the microbiome in health and disease. In contrast to C. difficile infection, the role of microbiome alterations in the causation of other illnesses such as inflammatory bowel disease is often not as clear. Even where potentially causal associations have been discovered, the exact mechanisms that could be underlying this association are frequently not well defined. Even though there are likely to be more mechanisms by which FMT can succeed in CDI, at least the currently known mechanistic insights (for example, the role of altered bile acid metabolism) can guide future development of microbiome therapeutics for this condition and have met with preclinical success. Similar mechanistic insight and success is still being sought for most of the other diseases for which microbiome manipulation is being proposed. One aspect about using FMT for conditions other than CDI that has received minimal consideration is the fact that the standard methods for the preparation and use of fecal material for treatment of recurrentC. difficile infection may not be appropriate for other conditions. In most of the published reports of using FMT for CDI, the fecal material is not prepared under anaerobic conditions. Most often, the fecal material is homogenized in a liquid such as saline or water under ambient oxygen conditions. In many reports, homogenization is performed in a blender, which would highly aerate the fecal suspension. Since the vast majority of bacteria in feces are strict anaerobes, this type of preparation would likely kill the vegetative forms of any of these organisms. As such, it has been suggested that spore-forming anaerobes are the key component of feces responsible for restoring colonization resistance against C. difficile. However, there is no a priori reason to assume that spore-forming anaerobes play a critical role in the treatment of other diseases. If the key microbial functions that are helpful in other conditions lay within the oxygen sensitive fraction of the gut microbiome, then the methods used for treating recurrent CDI would not be effective. Without basic research bent on discovering the underlying mechanisms by which the indigenous microbiota contribute to a particular disease, it may be premature to test FMT as it is currently employed for recurrent CDI for the treatment of other microbiota-related conditions. Future of microbiota manipulation The study of the role of the indigenous microbial communities in influencing health and disease is an exciting scientific endeavor. We are positioned to radically alter our understanding of how microbial symbionts participate in the homeostasis of their mammalian host. The idea that intentionally manipulating the indigenous microbiota can be used to improve health is attractive and is generating considerable efforts in the academic and commercial spheres. I would like to suggest that a measured and comprehensive approach that employs both discovery and application science is the most reasonable way to advance this emerging therapeutic approach. While the early results in the setting of recurrent CDI are encouraging, the probable differences in the exact role that the microbiome plays in other diseases requires us to more completely understand these roles before widespread attempts at microbiome manipulation are started. We owe it to the advancement of scientific knowledge and the health of patients to carefully consider how to proceed in this exciting new area of medical therapeutics. Support for the work in Dr. Young’s lab comes from the following grants, U19 AI090871 and U01 AI124255. Additional support comes from the University of Michigan Host Microbiome Initiative. Table of Definitions Microbiota: The community of microorganisms that inhabit a specific environment. Microbiome: A characteristic microbial community and the reasonably well-defined habitat that they inhabit. Germ Theory: A theory states that some diseases are caused by microorganisms. Koch’s Postulates: A formal test of the germ theory for the causation of disease. In order to fulfill Koch’s postulates, the following four criteria must be met: The organism must always be present, in every case of the disease. The organism must be isolated from a host containing the disease and grown in pure culture. Samples of the organism taken from pure culture must cause the same disease when inoculated into a healthy, susceptible animal in the laboratory. The organism must be isolated from the inoculated animal and must be identified as the same original organism first isolated from the originally diseased host. Colonization resistance: The ability of a host-associated microbiota to prevent the establishment of an exogenous microbe within the community. In terms of infectious diseases, this exogenous microbe is often referred to as a potential pathogen, but colonization resistance need not be restricted to pathogens. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. This review is based on a talk entitled “Microbiota Transplantation” that was presented at ECCMID, Amsterdam, Netherlands 2016. Literature Cited 1 Rao K Young VB Fecal microbiota transplantation for the management of clostridium difficile infection Infectious disease clinics of North America 2015 29 109 122 25677705 2 Brandt LJ Fecal microbiota transplant: Respice, adspice, prospice J Clin Gastroenterol 2015 49 Suppl 1 S65 S68 26447968 3 Mangiola F Ianiro G Franceschi F Fagiuoli S Gasbarrini G Gasbarrini A Gut microbiota in autism and mood disorders World journal of gastroenterology : WJG 2016 22 361 368 26755882 4 Zhang F Luo W Shi Y Fan Z Ji G Should we standardize the 1,700-year-old fecal microbiota transplantation? The American journal of gastroenterology 2012 107 1755 author reply p 1755-1756 23160295 5 Eiseman B Silen W Bascom GS Kauvar AJ Fecal enema as an adjunct in the treatment of pseudomembranous enterocolitis Surgery 1958 44 854 859 13592638 6 Bartlett JG Onderdonk AB Cisneros RL Kasper DL Clindamycin-associated colitis due to a toxin-producing species of clostridium in hamsters J Infect Dis 1977 136 701 705 915343 7 Koch R An address on bacteriological research British medical journal 1890 2 380 383 8 Falkow S Molecular koch's postulates applied to bacterial pathogenicity--a personal recollection 15 years later Nat Rev Microbiol 2004 2 67 72 15035010 9 Falkow S Molecular koch's postulates applied to microbial pathogenicity Rev Infect Dis 1988 10 Suppl 2 S274 S276 3055197 10 Gevers D Knight R Petrosino JF The human microbiome project: A community resource for the healthy human microbiome PLoS biology 2012 10 e1001377 22904687 11 Arumugam M Raes J Pelletier E Enterotypes of the human gut microbiome Nature 2011 473 174 180 21508958 12 Rifkin GD Fekety FR Silva J Jr Antibiotic-induced colitis implication of a toxin neutralised by clostridium sordellii antitoxin Lancet 1977 2 1103 1106 73011 13 Lusk RH Fekety R Silva J Browne RA Ringler DH Abrams GD Clindamycin-induced enterocolitis in hamsters J Infect Dis 1978 137 464 475 649990 14 Pruitt RN Lacy DB Toward a structural understanding of clostridium difficile toxins a and b Frontiers in cellular and infection microbiology 2012 2 28 22919620 15 Theriot CM Young VB Interactions between the gastrointestinal microbiome and clostridium difficile Annu Rev Microbiol 2015 69 445 461 26488281 16 Wilson KH Silva J Fekety FR Suppression of clostridium difficile by normal hamster cecal flora and prevention of antibiotic-associated cecitis Infect Immun 1981 34 626 628 7309245 17 Bassis CM Young VB Schmidt TM Fredricks DN Methods for characterizing microbial communities associated with the human body The human microbiota: How microbial communities affect health and disease 2013 Hoboken, New Jersey John Wiley & Sons, Inc. 51 74 18 Chang JY Antonopoulos DA Kalra A Decreased diversity of the fecal microbiome in recurrent clostridium difficile-associated diarrhea J Infect Dis 2008 197 435 438 18199029 19 Buffie CG Jarchum I Equinda M Profound alterations of intestinal microbiota following a single dose of clindamycin results in sustained susceptibility to clostridium difficile-induced colitis Infect Immun 2012 80 62 73 22006564 20 Seekatz AM Aas J Gessert CE Recovery of the gut microbiome following fecal microbiota transplantation mBio 2014 5 21 Khoruts A Dicksved J Jansson JK Sadowsky MJ Changes in the composition of the human fecal microbiome after bacteriotherapy for recurrent clostridium difficile-associated diarrhea J Clin Gastroenterol 2010 44 354 360 20048681 22 Wilson KH Freter R Interaction of clostridium difficile and escherichia coli with microfloras in continuous-flow cultures and gnotobiotic mice Infect Immun 1986 54 354 358 3533778 23 Marvaud JC Mory F Lambert T Clostridium clostridioforme and atopobium minutum clinical isolates with vanb-type resistance in france J Clin Microbiol 2011 49 3436 3438 21775552 24 Finegold SM Song Y Liu C Clostridium clostridioforme: A mixture of three clinically important species Eur J Clin Microbiol Infect Dis 2005 24 319 324 15891914 25 Reeves AE Koenigsknecht MJ Bergin IL Young VB Suppression of clostridium difficile in the gastrointestinal tracts of germfree mice inoculated with a murine isolate from the family lachnospiraceae Infect Immun 2012 80 3786 3794 22890996 26 Lawley TD Clare S Walker AW Targeted restoration of the intestinal microbiota with a simple, defined bacteriotherapy resolves relapsing clostridium difficile disease in mice PLoS Pathog 2012 8 e1002995 23133377 27 Wahlstrom A Sayin SI Marschall HU Backhed F Intestinal crosstalk between bile acids and microbiota and its impact on host metabolism Cell Metab 2016 28 Sorg JA Sonenshein AL Bile salts and glycine as cogerminants for clostridium difficile spores J Bacteriol 2008 190 2505 2512 18245298 29 Wilson KH Efficiency of various bile salt preparations for stimulation of clostridium difficile spore germination J Clin Microbiol 1983 18 1017 1019 6630458 30 Theriot CM Koenigsknecht MJ Carlson PE Jr Antibiotic-induced shifts in the mouse gut microbiome and metabolome increase susceptibility to clostridium difficile infection Nature communications 2014 5 3114 31 Theriot CM Bowman AA Young VB Antibiotic-induced alterations of the gut microbiota alter secondary bile acid production and allow for clostridium difficile spore germination and outgrowth in the large intestine mSphere 2016 1 32 Sorg JA Sonenshein AL Inhibiting the initiation of clostridium difficile spore germination using analogs of chenodeoxycholic acid, a bile acid J Bacteriol 2010 192 4983 4990 20675492 33 Buffie CG Bucci V Stein RR Precision microbiome reconstitution restores bile acid mediated resistance to clostridium difficile Nature 2015 517 205 208 25337874 34 Orenstein R Dubberke E Hardi R Safety and durability of rbx2660 (microbiota suspension) for recurrent clostridium difficile infection: Results of the punch cd study Clin Infect Dis 2016 62 596 602 26565008 35 Khanna S Pardi DS Kelly CR A novel microbiome therapeutic increases gut microbial diversity and prevents recurrent clostridium difficile infection J Infect Dis 2016 214 173 181 26908752 36 Scaldaferri F Pecere S Petito V Efficacy and mechanisms of action of fecal microbiota transplantation in ulcerative colitis: Pitfalls and promises from a first meta-analysis Transplant Proc 2016 48 402 407 27109966 37 Kahn SA Rubin DT When subjects violate the research covenant: Lessons learned from a failed clinical trial of fecal microbiota transplantation The American journal of gastroenterology 2016 Selected annotated bibliography on the use of fecal microbiota transplantion for the treatment of recurrent C. difficile infection 38 Bojanova DP Bordenstein SR Fecal Transplants: What Is Being Transferred? PLoS Biol 2016 7 12 14 7 e1002503 eCollection 2016. 27404502 A thought provoking discussion that elements of feces beyond the bacterial component may play a role in the efficacy of the therapeutic efficacy of fecal transplants. 39 Tauxe WM Haydek JP Rebolledo PA Neish E Newman KL Ward A Dhere T Kraft CS Fecal microbiota transplant for Clostridium difficile infection in older adults Therap Adv Gastroenterol 2016 5 9 3 273 281 A case review that demonstrated that FMT for treatment of CDI in patients aged 65 years or older was generally safe and effective. 40 Ray A Jones C Does the donor matter? Donor vs patient effects in the outcome of a next-generation microbiota-based drug trial for recurrent Clostridium difficile infection Future Microbiol 2016 5 11 611 616 26986546 A study of a processed fecal microbiota-derived drug for the treatment of recurrent CDI. In this substudy, it was demonstrated that preparation of the therapeutic from four different donors did not have an effect on efficacy. 41 Furuya-Kanamori L Doi SA Paterson DL Helms SK Yakob L McKenzie SJ Garborg K Emanuelsson F Stollman N Kronman MP Clark J Huber CA Riley TV Clements AC Upper Versus Lower Gastrointestinal Delivery for Transplantation of Fecal Microbiota in Recurrent or Refractory Clostridium difficile Infection: A Collaborative Analysis of Individual Patient Data From 14 Studies J Clin Gastroenterol 2016 3 11 [Epub ahead of print] In this meta-analysis of 14 different studies, it was suggested that administration of fecal microbiota transplant via the low gastrointestinal tract was more effective than administration via the upper gastrointestinal route. 42 Lee CH Steiner T Petrof EO Smieja M Roscoe D Nematallah A Weese JS Collins S Moayyedi P Crowther M Ropeleski MJ Jayaratne P Higgins D Li Y Rau NV Kim PT Frozen vs Fresh Fecal Microbiota Transplantation and Clinical Resolution of Diarrhea in Patients With Recurrent Clostridium difficile Infection: A Randomized Clinical Trial JAMA 2016 1 12 315 2 142 149 26757463 In this noninferiority trial, 219 patients in a modified intention-to-treat analysis received either fresh or frozen-and-thawed fecal transplant via enema for treatment of recurrent or refractory CDI. There was no difference in efficacy or adverse events between patients who received fresh or prozen FMT. 43 Costello SP Tucker EC La Brooy J Schoeman MN Andrews JM Establishing a Fecal Microbiota Transplant Service for the Treatment of Clostridium difficile Infection Clin Infect Dis 2016 4 1 62 7 908 914 Epub 2015 Nov 30 26628567 This paper provides practical advice on setting up a frozen stool bank for use in treating recurrent or refractory CDI. Methodology for donor recruitment, donor screening, stool preparation and therapeutic delivery is described. 44 Youngster I Russell GH Pindar C Ziv-Baran T Sauk J Hohmann EL Oral, capsulized, frozen fecal microbiota transplantation for relapsing Clostridium difficile infection JAMA 2014 11 5 312 17 1772 1778 25322359 A clinical trial that demonstrated the efficacy of feces administered in encapsulated, frozen form for recurrent CDI. 45 Moore T Rodriguez A Bakken JS Fecal microbiota transplantation: a practical update for the infectious disease specialist Clin Infect Dis 2014 2 58 4 541 545 Epub 2013 Dec 23 24368622 Another “how-to” manual for providers utilizing FMT for treatment of recurrent CDI. There is a particular focus on the regulatory aspects of the use of FMT for this indication, including guidance on how to obtain an investigation new drug permit. 46 van Nood E Vrieze A Nieuwdorp M Fuentes S Zoetendal EG de Vos WM Visser CE Kuijper EJ Bartelsman JF Tijssen JG Speelman P Dijkgraaf MG Keller JJ Duodenal infusion of donor feces for recurrent Clostridium difficile N Engl J Med 2013 1 31 368 5 407 415 Epub 2013 Jan 16 23323867 The critical randomized trial that demonstrated the efficacy of FMT for the treatment compared to standard treatment with vancomycin.
PMC005xxxxxx/PMC5125835.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 1246405 2835 Cell Immunol Cell. Immunol. Cellular immunology 0008-8749 1090-2163 27502364 5125835 10.1016/j.cellimm.2016.08.002 NIHMS808844 Article Phosphatidylcholine as a metabolic cue for determining B cell fate and function Brewer Joseph W. 2 Solodushko Viktoriya 1 Aragon Ileana 1 Barrington Robert A. 1 1 Department of Microbiology & Immunology, University of South Alabama Mobile, AL 36688 2 Department of Molecular & Cellular Sciences, Liberty University College of Osteopathic Medicine, Lynchburg, VA 24502 To whom correspondence should be addressed: Robert A. Barrington, Department of Microbiology & Immunology, University of South Alabama College of Medicine, MSB 2114, 5851 USA Drive North, Mobile, AL 36688, Telephone: (251) 461-1718; FAX: (251) 460-7931; [email protected] 7 8 2016 3 8 2016 12 2016 01 12 2017 310 7888 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. In activated B cells, increased production of phosphatidylcholine (PtdCho), the most abundant cellular phospholipid, is handled primarily by the CDP-choline pathway. B cell-specific deletion of CTP:phosphocholine cytidylyltransferase α (CCTα), the rate-limiting enzyme in the CDP-choline pathway, led to augmented IgM secretion and reduced IgG production, suggesting that PtdCho synthesis is required for germinal center reactions. To specifically assess whether PtdCho influences B cell fate during germinal center responses, we examined immune responses in mice whereby PtdCho synthesis is disrupted in B cells that have undergone class switch recombination to IgG1 (referred to as either Cγ1wt/wt, Cγ1Cre/wt or Cγ1Cre/Cre based on Cre copy number). Serum IgG1 was markedly reduced in naïve Cγ1Cre/wt and Cγ1Cre/Cre mice, while levels of IgM and other IgG subclasses were similar between Cγ1Cre/wt and Cγ1wt/wt control mice. Serum IgG2b titers were notably reduced and IgG3 titers were increased in Cγ1Cre/Cre mice compared with controls. Following immunization with T cell-dependent antigen NP-KLH, control mice generated high titer IgG anti-NP while IgG anti-NP titers were markedly reduced in both immunized Cγ1Cre/wt and Cγ1Cre/Cre mice. Correspondingly, the frequency of NP-specific IgG antibody-secreting cells was also reduced in spleens and bone marrow of Cγ1Cre/wt and Cγ1Cre/Cre mice compared to control mice. Interestingly, though antigen-specific IgM B cells were comparable between Cγ1Cre/wt, Cγ1Cre/Cre and control mice, the frequency and number of IgG1 NP-specific B cells was reduced only in Cγ1Cre/Cre mice. These data indicate that PtdCho is required for the generation of both germinal center-derived B cells and antibody-secreting cells. Further, the reduction in class-switched ASC but not B cells in Cγ1Cre/wt mice suggests that ASC have a greater demand for PtdCho compared to germinal center B cells. CCTα antibody-secreting cell antibody germinal centers T cell-dependent antigen unfolded protein response phospholipid biosynthesis INTRODUCTION B cell differentiation into antibody-secreting cells (ASC) invokes the unfolded protein response (UPR), a tightly organized, largely transcriptionally-controlled process resulting in enhanced secretory capacity. A distinctive feature of UPR transcriptional programming licenses B cells to enhance endoplasmic reticulum (ER) biogenesis [1], including components required for expanding the rough ER and Golgi to facilitate the marked demand for antibody secretion following activation and differentiation. Spliced X-box binding protein 1 (XBP1S) via IRE1 is essential for transcription of UPR genes[2], and consequently for the production of antibody by ASC. One key factor that can be regulated by XBP1S is choline cytidylyltransferase α (CCTα)[3], the predominant isoform of the rate-limiting enzyme in the cytidine phosphocholine (CDP-choline) pathway for phosphatidylcholine synthesis [4, 5]. Indeed, enforced expression of XBP1S in fibroblasts is sufficient to drive induction of phosphatidylcholine (PtdCho) by a mechanism involving regulation of CCTα [6]. The production of CCTα, in turn, increases the supply of CDP-choline for synthesis of phosphatidylcholine (PtdCho) that can be further processed into phosphatidylserine (PtdSer) and phosphatidylethanolamine (PtdEtn)[7, 8]. PtdCho, PtdSer and PtdEtn are all major lipid constituents for cellular membranes, with PtdCho the most abundant [9]. Thus, XBP1S can regulate the supply of membrane lipids, an ability that fits well with its essential function in ER biogenesis in numerous types of dedicated secretory cells such as ASC [10, 11], pancreatic acinar cells [12], Paneth cells [13] and salivary gland cells [14] that possess and utilize large quantities of rough ER for their functions. In vivo activation of B cells by either T cell-independent (TI) or –dependent (TD) antigens leads to differentiation of B cells into either short-lived plasmablasts [15] or to development of germinal centers that ultimately generate both long-lived ASC and memory B cells [16]. B cells stimulated with bacterial lipopolysaccharide (LPS), a TLR4-dependent model for T cell-independent responses, upregulate CCT activity approximately 2-fold while PtdCho production increases approximately 7-fold [9]. Similarly, LPS stimulation of CH12 lymphoma cells resulted in increased CCTα levels, though this was attributed to reduced protein turnover rather than transcriptional activation [5]. Importantly, CCTα-deficient B cells fail to upregulate PtdCho synthesis after LPS stimulation [17]. Thus, CCTα appears integral for B cell differentiation into ASC in response to T cell-independent stimuli. Interestingly, mice harboring B cells rendered CCTα-deficient following lineage commitment via CD19-Cre-induced gene deletion generated markedly reduced IgG and increased IgM in response to immunization with TD antigen [17]. IgM production was similarly increased in primary CCTα-deficient B cells in vitro upon stimulation with LPS, despite a corresponding reduction in B cell proliferation. However, reduced frequencies of splenic and peritoneal B cells were also noted in B cell-CCTα-deficient mice [17]. Both splenic marginal zones and the peritoneum contain B-1 cells [18], and B-1 cell-derived IgM is required for normal responses to TD-antigens [19]. This raises the possibility that a reduction of B-1 cells contributed to the impaired antibody responses observed in B cell-CCTα-deficient mice. Moreover, neither germinal center nor antigen-specific antibody levels were measured in those studies. Therefore, the significance of increased PtdCho production in antigen-specific B cell responses remains unknown. To resolve whether PtdCho production is required for B cell responses to TD antigens, humoral immunity was examined in conditional IgG1 B cell-CCTα-deficient (Cγ1-CCTα) mice in which CCTα is selectively eliminated in B cells that have undergone class switch recombination from IgM to IgG1. Importantly, B cell development appeared normal in all CCTαflox (Cγ1wt/wt, Cγ1Cre/wt, and Cγ1Cre/Cre) mice, and serum immunoglobulin (Ig) levels were similar between Cγ1Cre/wt and wild-type mice, with the exception of selective reduction in IgG1. Serum IgG1 levels in Cγ1Cre/Cre mice were also reduced, while these mice also unexpectedly exhibited decreased IgG2b and increased IgG3 titers as compared to control mice. In response to immunization with NP-KLH emulsified in alum, which generates an IgG1-dominant antibody response to NP, both antigen-specific IgM and IgG primary responses were impaired in Cγ1Cre-expressing mice as compared to CCTα-sufficient control mice. The reduced response was not due to failure of Cγ1-Cre-expressing mice to generate germinal centers since the frequency and number of GC was comparable between each of the three strains examined. Rather, the diminished antigen-specific IgG in Cγ1-Cre-expressing mice correlated with reductions in hapten-specific antibody-secreting cells (ASC). Examination of germinal center B cell populations revealed that, while the frequency and number of NP-specific IgM B cells in Cγ1-Cre-expressing mice was comparable to control mice, the frequency and number of NP-specific IgG1 germinal center B cells was significantly reduced in Cγ1Cre/Cre CCTα mice. Notably, though class-switched, hapten-specific ASC were reduced in Cg1Cre/wt mice, the frequency and number of class-switched hapten-specific germinal center B cells was not, suggesting a differential demand for PtdCho. No differences were observed in the affinity of NP-specific IgG after immunization, suggesting that increased PtdCho synthesis is not required for selection of antigen-specific B cells. In summary, these studies reveal that PtdCho is required for the generation of class switched B cells in germinal centers as well as the production of both IgG1 memory B cells and ASC. RESULTS Conditional Pycta1 mice were generated by crossing a mouse strain containing loxP-flanked Pycta1 alleles (Pycta1flox)[17] with the Cγ1-Cre strain [20] whereby Cre recombinase is expressed when B cells undergo class switch recombination from IgM to IgG1. Progeny from the mouse cross therefore generate B cells that selectively delete Pycta1 upon commitment to expressing IgG1. Pcyta1 encodes CTP:phosphocoline cytidylyltransferase α (CCTα), the rate-limiting in the CDP-choline pathway for synthesis of phosphatidylcholine (PtdCho) [4], the most abundant phospholipid component in cell membranes [9]. Because IgG1 B cells conventionally derive from immune responses, the requirement for phospholipid synthesis for the generation of antibody-secreting and memory B cells was measured following immunization with the well-characterized T cell-dependent (TD) hapten-carrier antigen NP-KLH. Conditional deletion of CCTα in IgG1 B cells does not alter B cell development To determine whether conditional deletion of Pycta1 alters B cell development prior to immunization, naïve littermate control mice (Cγ1-wt Pycta1flox/flox, referred to as wild-type) and Cγ1-Cre-expressing Pycta1flox/flox (referred to as Cγ1Cre/wt and Cγ1Cre/Cre) mice were compared. Developing bone marrow B cells were identified as B220+ CD23neg and were further distinguished phenotypically by CD24 and CD43 levels. As shown in Figure 1a, the frequencies of pre-pro (CD24neg, CD43+), pro (CD24int CD43+) and pre B (CD24+ CD43+/−) cells were comparable between control and Cγ1-Cre-expressing mice. Moreover, the frequencies of all developing B cells as well as mature recirculating (B220+ CD23+) B cells were also equivalent between control and experimental groups of mice. Peripheral B cell populations were similarly assessed, and the frequencies of total transitional (B220+ CD93+) and mature (B220+ CD93neg) B cells were again comparable between each of the strains (Figure 1b). Finally, the frequencies of follicular (CD21lo CD23+) and marginal zone (CD21+ CD23int) B cells were measured, and again both CCTα-sufficient and -deficient mice were indistinguishable. These data indicate that B cell development is normal in Cγ1-Cre-expressing mice. Impaired antibody-secreting cell (ASC) differentiation would most notably manifest as a reduction in serum immunoglobulin (Ig). Serum levels of IgM, IgG1, IgG2a, IgG2b, and IgG3 were therefore quantified in naïve 10 week-old mice. Serum titers of IgM, IgG2a, IgG2b and IgG3 were similar between wild-type and Cγ1Cre/wt mice (Figure 2). Interestingly, though Cγ1Cre/Cre mice also had comparable titers of IgM and IgG2a antibodies, IgG2b titers were significantly reduced and IgG3 titers were elevated compared to wild-type control mice. While serum IgG1 was detectable up to 1 to 10,000 dilution from control mice, IgG1 was detectable only at 1 to 100 serum dilution from Cγ1-Cre-expressing mice, indicating that CCTα, and PtdCho production, is necessary for the generation of normal serum IgG1. CCTα is required for humoral response to TD antigen To evaluate whether conditional deletion of Pycta1 in IgG1 B cells causes defects in Ig production upon antigen stimulation, mice were immunized intraperitoneally (i.p.) using the TD antigen NP-KLH emulsified in alum and examined three weeks later. As shown in Figure 3a, while wild-type mice generated mean NP-specific IgM and IgG titers of 1,673 and 56,000, Cγ1Cre/wt and Cγ1Cre/Cre mice had mean NP-specific IgM titers of 698 and 643, and mean NP-specific IgG titers of 5,929 and 5,000, respectively. Therefore, Cγ1-Cre-expressing mice exhibit an impaired IgM and IgG primary response to TD antigen. Serum IgG derives primarily from ASC in the bone marrow, and these bone marrow ASC derive from germinal center responses. To examine whether the production of germinal center-derived ASC was affected in Cγ1-Cre-expressing mice, NP-specific ASC were enumerated by ELIspot assay. Three weeks following immunization with NP-KLH in alum, both wild-type control and Cγ1-CCTα mice generated comparable frequencies of NP-specific IgM-secreting cells in the spleen (Figure 3b left panel, 26.7 ± 5.8, wt vs. 20.8 ± 4.1, Cγ1Cre/wt, and 16.0 ± 4.1, Cγ1Cre/Cre). Consistent with the reduced hapten-specific IgM antibody titers, both Cγ1Cre/wt and Cγ1Cre/Cre mice had reduced NP-specific IgM+ ASC in bone marrow compared to control mice (Figure 3c left panel, 1.6 ± 0.3, wt, 1.0 ± 0.5, Cγ1Cre/wt and 0.4 + 0.2, Cγ1Cre/Cre). By comparison, the production of NP-specific IgG secreting cells was reduced in Cγ1-Cre-expressing mice relative to wild-type control mice in both spleen (Figure 3b right panel, 39.2 ± 9.5, wt vs. 16.9 ± 5.7, Cγ1Cre/wt and 6.5 ± 3.0, Cγ1Cre/Cre) and bone marrow (Figure 3c right panel, 2.7 ± 1.4, wt vs. 1.0 ± 0.4, Cγ1Cre/wt and 0.3 ± 0.1, Cγ1Cre/Cre). Collectively, these results suggest that PtdCho production is required for the generation of class-switched ASC. Reduced NP-specific IgM during the primary response in Cγ1-CCTα mice was unexpected. To determine whether this could be due to impaired germinal centers, the frequency of splenic germinal centers was quantified. Using flow cytometry, the frequency of PNAhi CD95+ (germinal center) B cells was determined two weeks after immunization with TD antigen NP-KLH (Figure 4a, 4b). The frequency (5.1 ± 0.5, wt vs. 4.1 ± 0.7, Cγ1Cre/wt and 3.8 + 0.5, Cγ1Cre/Cre) and number (241,523 ± 36,718, wt vs. 193,479 ± 45,573, Cγ1Cre/wt and 200,235 ± 27,427, Cγ1Cre/Cre) of germinal center B cells was similar between immunized control and Cγ1-Cre-expressing mice. To complement these analyses, histological analysis was also performed. Approximately 36% of splenic follicles from immunized wild-type mice contained germinal centers, while approximately 39% and 44% of splenic follicles in immunized Cγ1Cre/wt and Cγ1Cre/Cre mice had germinal centers, respectively (Figure 4c). Thus, Cγ1-Cre-expressing mice were as capable as control mice of forming germinal centers in response to TD antigen challenge. Reduced antigen-specific B cell memory in Cγ1Cre/Cre mice Germinal center responses generate both antigen-specific ASC and memory B cells. To measure whether the production of memory B cells was also impaired in immunized Cγ1-Cre-expressing mice, the frequency and number of NP-specific germinal center B cells was measured (Figure 5). The frequency and number of IgM+ NP-specific germinal center B cells (Figure 5b and 5d) was similar between CCTα-sufficient and Cγ1-Cre-expressing mice (1.3 ± 0.4 and 2,659 ± 762, wt; 1.0 ± 1.0 and 1,754 ± 480, Cγ1Cre/wt; 1.7 ± 0.4 and 2,751 ± 788, Cγ1Cre/Cre). In contrast, while approximately 2% of germinal center B cells in wild-type mice were NP-specific IgG1+, the frequency was markedly reduced in Cγ1Cre/Cre mice (2.1 ± 0.5, wt vs. 0.3% ± 0.1, Cγ1Cre/Cre) (Figure 5c). The number of NP-specific IgG1 germinal center B cells was also reduced approximately 4-fold in Cγ1Cre/Cre compared to the number in wild-type mice (4,521 ± 1,313, wt vs. 796 ± 295, Cγ1Cre/Cre) (Figure 5e). These differences between wild-type and Cγ1Cre/Cre mice were evident in both CD38+ and CD38lo/neg germinal center cells (data not shown), suggesting that both light and dark zone responses were involved [21]. Surprisingly, the frequency and number of NP-specific IgG1 germinal center B cells did not differ statistically between Cγ1Cre/wt (1.4 ± 0.4 and 2,477 ± 740) and wild-type mice. From these data, we conclude that the generation of antigen-specific class-switched germinal B cells requires de novo PtdCho synthesis, but that this developmental process can proceed under conditions of reduced PtdCho production. To test whether the reduced antigen-specific IgG1 B cells and ASC had a functional consequence, mice were challenged and secondary responses were measured. As was observed in the primary response, the antigen-specific IgM response after challenge was comparable between wild-type and Cγ1-Cre-expressing mice (Figure 6). Thus, no differences in NP-specific IgM titers (1,300 ± 272, wt; 1,116 ± 210, Cγ1Cre/wt; 2,050 ± 665, Cγ1Cre/Cre) (Figure 6a left panel) nor frequency of NP-specific IgM ASC in bone marrow (5.5 ± 1.6, wt; 2.3 ± 1.3, Cγ1Cre/wt; 4.5 ± 3.7, Cγ1Cre/Cre) (Figure 6c left panel) were observed. There was a trend toward an increased frequency of NP-specific IgM ASC in spleen in Cγ1-Cre-expressing mice compared to wild-type mice, though these data were not statistically significant (8.7 ± 0.9, wt; 261.5 ± 121.3, Cγ1Cre/wt; 101.3 ± 67.2, Cγ1Cre/Cre) (Figure 6b left panel). In contrast to the IgM response, while wild-type mice produced 1,725 NP-specific IgG splenic ASC, the frequency was reduced at least 3-fold in Cγ1-Cre-expressing mice (Figure 6b right panel: 572 ± 83.7, Cγ1Cre/wt; 373.8 ± 83.2, Cγ1Cre/Cre). The frequency of NP-specific IgG bone marrow ASC was also reduced in Cγ1-Cre-expressing mice (12.2 ± 1.6, Cγ1Cre/wt; 9.2 + 2.3, Cγ1Cre/Cre) compared to wild-type mice (45.1 ± 5.8) (Figure 6c right panel). Therefore, PtdCho production is required for the generation of antigen-specific, class-switched ASC. To gauge whether the production of antigen-experienced B cells was also affected in Cγ1-Cre-expressing mice, germinal centers were interrogated (Figure 7). As expected, the frequency and number of NP-binding IgG1+ germinal center B cells increased relative to the frequency observed in primary germinal center responses (Figure 7a, c, e). Notably, the mean level of NP binding (MFI) appeared greater in IgG1 versus IgM B cells (Figure 7a). In contrast to the recall response of CCTα-sufficient control mice, the frequency and number of IgG1+ NP-specific germinal center B cells was reduced in Cγ1-Cre-expressing mice (IgG1, mean ± SEM: 7.7 ± 1.0 and 97,651 ± 26,303, wt; 3.7 ± 0.4 and 47,977 ± 6,426, Cγ1Cre/wt; 2.3 ± 0.7, 22,238 ± 4,404, Cγ1Cre/Cre). Thus, reduced primary responses of Cγ1-CCTα mice manifested as reduced recall germinal center responses. Antigen-specific IgG affinity is comparable in wild-type and Cγ1-Cre-expressing mice Reducing competition within germinal centers by loss of high-affinity B cells can allow lower-affinity B cells to be selected [22, 23]. To determine whether elimination of CCTα in B cells that class switch to IgG1 affects germinal center selection, relative changes in antigen-specific serum IgG antibody affinity were measured by determining the ratio of antibody titers binding low (NP2) to highly (NP20) haptenated BSA. The ratio of NP2/NP20 IgG titers from wild-type mice increased from approximately 0.5 to approximately 0.7 (Figure 8). Comparatively, the NP2/NP20 IgG titer ratio was lower after primary immunization in Cγ1Cre/wt mice, but increased to approximately 0.5 following secondary challenge. Cγ1Cre/Cre mice exhibited a trend toward more increased affinity following secondary challenge with antigen, however the increase was not statistically different from affinity changes in wild-type mice. Thus, no significant differences were observed in the relative serum affinity of IgG antibodies between wild-type and Cγ1-Cre-expressing mice. DISCUSSION Among the key events in B cells undergoing differentiation into ASC, upregulation of phospholipid biosynthesis facilitates the expansion of the ER and Golgi to accommodate an increased demand for Ig synthesis, assembly and secretion. Previous studies detailed that CCTα is the rate-limiting enzyme for PtdCho synthesis in B cells and that it was necessary to direct class switch recombination of B cells to the TI stimulus LPS in a proliferation-independent manner [17]. To address whether B cells responding to TD antigen in germinal centers differed in their dependence for CCTα, we utilized Cγ1-Cre CCTαflox/flox mice whereby CCTα would be deleted upon class switch recombination of B cells from IgM to IgG1. We observed that B cell development and germinal center development is normal in Cγ1-Cre-expressing mice; however, these mice had reduced serum IgG1 levels and reduced antigen-specific IgG1 antibody and ASC upon antigen challenge. Interestingly, class-switched hapten-specific germinal center B cells, though reduced in Cγ1Cre/Cre mice, were similar between wild-type and Cγ1Cre/wt mice, suggesting that proliferating germinal center B cells may demand less PtdCho than ASC. Surprisingly, the IgM antibody response in immunized Cγ1-Cre-expressing mice was also reduced compared to wild-type mice, though the frequency and number of hapten-specific IgM germinal center B cells was similar in all three strains of mice. These studies indicate that CCTα is required for the production of germinal center-derived class-switched ASC and memory cells. Further, these results are consistent with a model whereby PtdCho is rate-limiting for both cell proliferation and differentiation of germinal center B cells (Figure 9). Gene-targeted deletion of Pcyt1a is embryonically lethal (day 3.5)[24], whereas deletion of other CCT isoforms such as CCTβ have less profound effects [25], supporting that CCTα is the dominant isoform required for PtdCho synthesis in multiple tissues. The requirement for CCTα in B cells was assessed using CD19-mediated conditional deletion, and resulted in reduced numbers of peritoneal and splenic B cells, as well as significantly reduced serum IgG levels [17]. Upon challenge with TD antigen, CD19-CCTα mice notably failed to form germinal centers, likely due to the requirement for CCTα in the oligoclonal proliferative burst that initiates the germinal center response [26]. Both reduced B cell numbers and the failure to form germinal centers were likely responsible for impaired IgG production in CD19-CCTα mice following TD antigen challenge. Use of Cγ1-Cre-expressing mice in the current work offered an approach that does not affect B cell development or the formation of germinal centers, thereby allowing the requirement of CCTα and PtdCho synthesis in germinal center responses to be directly assessed. Therefore, antigen-specific B cells retain the ability to synthesize PtdCho to allow for expansion and generation of germinal centers. Differentiation of B cells into antibody-secreting cells requires the UPR transcription factor XBP1S. This requirement extends to both responses to TI antigens by marginal zone B cells and to TD antigens that occur through germinal centers. The link between XBP1S and PtdCho is 3-fold: XBP1 is sufficient to upregulate CCTα and PtdCho synthesis (3, 6), deletion of XBP1 impairs PtdCho synthesis in activated B cells [27] and blocking PtdCho synthesis in B cells through selective deletion of CCTα induces XBP1(S) (18). Interestingly, CCTα-deficient B cells stimulated in vitro with LPS fail to class switch Ig and also secrete more IgM compared to CCTα-sufficient B cells [17]. Therefore, disabling new PtdCho production in activated B cells drives them to upregulate XBP1 and secrete IgM. By comparison, conditional elimination of XBP1 in B cells leads to reduced basal Ig levels, as well as reduced specific Ig in response to immunization using TD antigen NP-KLH emulsified in alum. The reduced antigen-specific response was limited to the generation of ASC, as the development of antigen-specific B cells in germinal centers was not affected [28]. We demonstrate that conditional elimination of CCTα in germinal center B cells upon class switch recombination to IgG1 leads to a reduction in both antigen-specific ASC and B cells. Taken together, these separate studies would suggest that limiting PtdCho availability acts ahead of the need for XBP1S, likely due to the requirement for PtdCho in expansion of class-switched antigen-specific B cells prior to differentiation within the germinal center reaction. Interestingly, levels of serum IgM increase in CD19-CCTα mice following immunization, suggesting that reduced capacity to synthesize PtdCho in B cells may manifest as a type of hyper-IgM syndrome [17]. The expression of CCTα is cell-cycle regulated [4, 29], raising the possibility that inhibiting proliferation by impairing PtdCho synthesis can direct antigen-specific B cells to differentiate into ASC. Consistent with this idea, depletion of PtdCho levels leads to induction of XBP1S [17]. In Cγ1-CCTα mice, while the frequency of antigen-specific IgM germinal center B cells was comparable with those observed in primary-immunized control mice, levels of antigen-specific IgM were significantly reduced. In addition, antigen-specific IgM bone marrow ASC in Cγ1-CCTα mice were reduced compared with control mice. Thus, impairment of PtdCho synthesis in class-switched B cells affected the production of IgM-ASC but not IgM-B cells. The reduction in antigen-specific IgG observed in Cγ1-Cre expressing mice may contribute to a general reduction in germinal center efficiency by limiting IgG immune complexes, though this would not be expected to differentially affect the production of IgM-ASC. IgM B cells in germinal centers generally would not express Cre recombinase, since Cre is translated from Cγ1 transcripts via an internal ribosomal entry site. However, it is possible that a percentage of these germinal center B cells begin, but do not complete, class switch recombination. In this scenario, sterile transcripts could be generated, thereby allowing Cre recombinase to be expressed and CCTα to be subsequently deleted. The effect would be more profound for IgM-ASC than IgM B cells due to increased demands for PtdCho synthesis to expand the endoplasmic reticulum to accommodate enhanced secretory capacity. Depleting PtdCho in non-lymphoid cells initiates the UPR and leads to apoptosis [30, 31]. A similar phenotype could explain the reduction in antigen-specific IgG B cells and ASC in immunized Cγ1-Cre-expressing mice. Alternatively, the observation that both antigen-specific IgG B cells and ASC were similarly reduced in Cγ1Cre/Cre cohorts may suggest that PtdCho is required at the level of clonal expansion within the germinal center prior to commitment to differentiation into ASC. However, only antigen-specific IgG ASC and not B cells were reduced in Cγ1Cre/wt mice, raising the intriguing possibility that ASC are more sensitive to limited PtdCho supply than proliferating germinal center B cells. Notably, though titers of antigen-specific IgG were reduced in Cγ1-Cre-expressing mice, the affinity of the IgG response was comparable between immunized Cγ1-Cre-expressing mice and control animals. Therefore, impaired PtdCho synthesis had no effect on affinity of the IgG response. Previous studies examining selection of B cells within germinal centers found that reducing competition of high-affinity B cells allows lower-affinity B cells to be selected [22, 23]. Signaling via membrane IgG is more potent compared to IgM [32], possibly providing a competitive advantage for IgG B cells in the germinal center response. If true, it follows that reducing the number of IgG B cells through conditional deletion of CCTα could manifest as an increase in antigen-specific IgM B cells. This is not the case, as we observed approximately equivalent frequency and numbers of NP-specific IgM B cells within germinal centers of Cγ1-Cre-expressing mice and control mice. Because germinal centers also serve to potentiate affinity maturation, we speculate that in a setting whereby the frequency and number of IgG B cells is limiting, IgM B cells compete for antigen within the germinal centers thereby resulting in an increase in serum antibody affinity. Though affinity of antigen-specific IgM is very difficult to measure, the reduced NP-specific IgM bone marrow ASC observed in Cγ1-Cre-expressing mice would at least be consistent with this hypothesis, though we cannot exclude post-germinal center selection. In conclusion, these studies support a requirement for PtdCho synthesis in the generation of both memory and ASC during germinal center responses to TD antigens. This dependency likely reflects a dependence for PtdCho in both the proliferative burst of antigen-specific B cells as well as the need for the expansion of the ER, Golgi and plasma membranes required for ASC differentiation. Putting our studies into context with earlier work demonstrating that limiting PtdCho synthesis triggers induction of XBP1S in the B cell response to LPS, it is likely that the fate and/or function of responding B cells employing the UPR is dependent on the degree of proliferation involved in the differentiation process. MATERIALS AND METHODS Generation of mice with CCTα-deficient IgG1 B cells Pcyta1flox/flox mice [31] were crossed with Cγ1-Cre [20] mice to generate Cγ1Cre/Cre Pcta1flox/flox, Cγ1Cre/wt Pcta1flox/flox, and Cγ1wt/wt Pcta1flox/flox littermates. Notably, all examined aspects of B cell development and serum immunoglobulin levels were comparable between Cγ1wt/wt Pcta1flox/flox and non-floxed wild-type mice, and therefore they are collectively referred to as wild-type herein. Mice were housed at the University of South Alabama in an AAALAC-certified specific pathogen-free facility. Maintenance of breeding colonies and all procedures involving mice were performed according to protocols approved by the University of South Alabama Institutional Animal Care and Use Committee. Immunization and serum analysis Mice were immunized intraperitoneally (i.p.) with 0.05 mg Imject-precipitated (Thermo Scientific, Grand Island, NY) 4-hydroxy-3-nitrophenyl conjugated to keyhole limpet hemocyanin (NP5-KLH; Biosearch Technologies, Novato, CA). Booster immunizations were administered identically 3 weeks after primary immunization. Anti-NP Response and Affinity Measurements by ELISA Serum was collected from individual mice and NP-specific antibody titers were determined by sandwich ELISA. 96-well plates (Nunc; Thermo Scientific) were coated with 5 ug/well NP2-BSA or NP20-BSA. Plates were blocked with the addition of 5% dry milk (Carnation®) in PBS (Blotto). NP-specific IgG serum antibody in serially diluted samples was detected by horse radish peroxidase–conjugated goat anti–mouse IgG (Southern Biotechnology, Birmingham, AL). Between incubations, plates were washed with PBS containing 0.1% Tween. Color was developed using substrates 3, 3′, 5, 5′-tetramethylbenzemidine (TMB, Thermo Scientific) and hydrogen peroxide, and absorbance was measured at 405 nm using Softmax software package (Molecular Devices, Sunnyvale, CA). Antibody titer was determined as the reciprocal of the greatest dilution whose absorbance remained at least two-fold above background. For relative affinity measurements, the ratio of titers to NP2-BSA and NP20-BSA was calculated for individual mice using OD405 in the linear ranges of the assays as previously described [33]. For ELISA measuring IgM and IgG subclass titers in serum, a sandwich method was used by coating wells with goat anti-mouse Ig, and developing as above with subclass-specific IgM, IgG1, IgG2a, IgG2b or IgG3 horseradish peroxidase conjugated secondary antibodies (Southern Biotechnology). Enzyme-linked Immunospot Assay for NP-specific antibody-secreting cells (ASC) Frequencies of NP-specific ASCs were quantitated as previously described [33]. In brief, 24-well polystyrene plates (Corning Inc., Corning, NY) were coated with NP5-BSA. After extensive washing, plates were blocked using 1% BSA in PBS for 2 hours. Serial ten-fold diluted splenic mononuclear cells or BM cells (106 to 103 cells/well) were added in DMEM media with 2% fetal bovine serum and incubated overnight at 37 C. Each dilution of cells was assayed in duplicate. Plates were then washed with PBS containing 0.1% Tween and incubated with alkaline phosphatase–conjugated goat anti–mouse IgG antibody (Sigma-Aldrich). Plates were developed using 5-bromo-4-chloro-3-indolyl phosphate at 1 mg/ml in 0.6% agarose to produce blue-colored spots identifying NP-specific ASCs. Spots were counted to determine ASC frequency. As controls, each sample was also plated in wells coated with BSA or KLH. Few BSA-specific ASCs were detected in the BM or spleen and fewer than 100 KLH-specific ASCs were observed per recipient spleen (data not shown). Flow cytometry and cell sorting Single cell suspensions of bone marrow and splenic mononuclear cells (MNC) were isolated by density gradient centrifugation using Lympholyte M (Cedarlane Laboratories, Burlington, N.C.). To detect IgG1 B cells, MNCs were stained with biotinylated anti-IgG1 antibody (clone X-56, Miltenyi Biotec, Auburn, CA), followed by streptavidin-PerCP-Cy5.5 conjugate (BD Biosciences, San Jose, CA). Other antibodies used for flow cytometric analyses of B cell subsets in bone marrow, spleen and cervical lymph nodes included the following (BD Biosciences and eBioscience, San Diego, CA): CD19 (1D3), CD21/CD35 (7G6), CD23 (B3B4), CD24 (M1/69), CD38 (90), CD43 (S7), CD45R/B220 (RA3-6B2), CD93 (AA4.1), CD95 (Jo2), CD138 (281–2), IgM (II/41, R6-60.2), IgD (11–26), T and B cell activation antigen (GL-7), and PNA (Vector Laboratories, Burlingame, C.A.). To identify NP-specific B cells, cells were incubated with NP coupled to allophycocyanin (Thermo Scientific). Cells were analyzed by FACSCanto II and sorted using multi-laser FACSAria II-SORP (BD Biosciences) housed in the University of South Alabama College of Medicine Flow Cytometry Core Laboratory. Data were analyzed with FlowJo software (Tree Star Ashland, OR). Immunohistochemistry Isolated spleens were preserved in OCT compound on dry ice. 5-μm thick cryosections were examined following staining with biotinylated anti-B220 (BD Biosciences) and FITC-conjugated peanut agglutinin (Vector Laboratories). Streptavidin-alkaline phosphatase (Invitrogen) and HRP-conjugated anti-FITC (Thermo Scientific) were used as secondary staining reagents. Staining was visualized using a Nikon Eclipse microscope (Nikon Instruments Inc., Melville, NY) and images were analyzed using Nikon Elements software. Statistical analysis Data comparing three groups were analyzed by a 1-way ANOVA test with Tukey’s multiple comparisons test applied to compare individual group means. Statistical significance was determined by p value as indicated within the figure legends. This work was supported by NIH R01 GM61970. The authors thank Dr. Suzanne Jackowski (St. Jude Children’s Research Hospital, Memphis, TN) for providing CD19-Cre; Pcyta1flox/flox mice and Dr. Stefano Casola (IFOM, Milan, Italy) for providing Cγ1-Cre mice. Abreviations ASC antibody-secreting cell ER endoplasmic reticulum UPR unfolded protein response XBP1 X-box binding protein 1 XBP1S spliced X-box binding protein 1 IRE1 inositol requiring 1 PtdCho phosphatidylcholine CDP-choline cytidine diphosphocholine CCT choline cytidylyltransferase PtdEtn phosphatidylethanolamine PtdSer phosphatidylserine NP-KLH nitrophenyl-keyhole limpet hemocyanin TD T cell-dependent TI T cell-independent LPS lipopolysaccharide Figure 1 B cell development occurs normally in Cγ1-Cre-expressing mice B cell development in representative naïve 2 month-old wild-type (Cγ1wt/wt CCTαflox/flox, left panels) and Cγ1Cre/wt CCTαflox/flox (middle) and Cγ1Cre/Cre CCTαflox/flox (right) mice was assessed by flow cytometry. In (a), developing (B220+ CD23−) and mature (B220+ CD23+) bone marrow B cells were distinguished (top panels). Developing B cells were further assessed using CD43 and CD24 levels. Splenic B cell analyses are shown in (b): transitional B cells were defined as B220+ CD93+ (top panels); marginal zone (MZ) and follicular (FO) were distinguished by CD21 and CD23 (bottom panels). Figure 2 Antibody levels in naïve Cγ1-Cre-expressing mice Ig titers for IgM and IgG subclasses in serum from naïve 2 month-old wild-type (n=5, black bars), Cγ1Cre/wt (n=5, gray) and Cγ1Cre/Cre (n=5, white bars) mice were measured by ELISA. Data shown are means, with standard error indicated (p<0.003, *p<2 × 10−5, unpaired Student’s t test with Welch’s correction). Figure 3 Impaired antigen-specific IgG response and ASC formation in Cγ1-Cre-expressing mice Mice were immunized i.p. using 50 μg NP5-KLH emulsified in alum. Three weeks after immunization, ELISA (a) and ELIspot assays (b and c) were used to measure antigen-specific responses. In (a), IgM and IgG primary antibody responses (mean + standard deviation) in wild-type mice (black bars) and Cγ1-Cre-expressing mice (gray bars). NP-specific IgM (left panels) and IgG (total, right panels) antibody-secreting cells (ASC) in spleen (b) and bone marrow (c); filled circles represent data from individual wild-type mice, filled squares are data from Cγ1Cre/wt mice, and filled triangles are data from Cγ1Cre/Cre mice (all mice were CCTαflox/flox). Statistics were calculated using 1-way ANOVA with Tukey’s multiple comparison test (p<0.02; p<0.007; p<0.0002). Figure 4 Development of germinal centers occurs normally in Cγ1-Cre-expressing mice The frequency and number of germinal center B cells in spleen were determined by flow cytometry. The frequency of splenic germinal center B cells (a) identified as the PNAhi CD95+ fraction of the B220+ population. The absolute number of germinal center B cells (b) was calculated by multiplying the frequency by total splenocyte counts. Filled circles represent data from individual wild-type mice, filled squares are data from Cγ1Cre/wt mice, and filled triangles are data from Cγ1Cre/Cre mice (all mice were CCTαflox/flox). In (c), the frequency of germinal centers was determined by immunohistochemical analysis of splenic cryosections. Up to 4 sections from wild-type mice (n=6), Cγ1Cre/wt (n=6), and Cγ1Cre/Cre (n=6) mice were assessed. Figure 5 The frequency of germinal center-resident NP-specific IgG1+ B cells is reduced in Cγ1Cre/Cre mice NP-specific germinal center B cells were identified using flow cytometry. Briefly, in (a), single lymphocytes (upper left and middle-left plots) were gated on B cells (middle-right plot) in germinal centers (far right plot), then assessed for frequency of NP-binding IgG1 B cells (lower plots). Minus 1 staining negative controls (not shown) were used to establish gates for NP-binding. Representative data from multiple wild-type, Cγ1Cre/wt and Cγ1Cre/Cre mice are shown. The frequency and number of IgM+ (b and d) and IgG1+ (c and e) NP-binding splenic germinal center B cells. Statistical differences were determined with 1-way ANOVA using Tukey’s multiple comparison test (p<0.03). Figure 6 CCTα is required for IgG but not IgM antibody and ASC during secondary immune responses Mice were immunized i.p. using 50 μg NP5-KLH emulsified in alum. Five days after booster immunization, ELISA (a) and ELIspot assays (b) were used to measure antigen-specific responses. In (a), IgM (left panel) and IgG (right panel) primary antibody responses (mean + standard deviation) in wild-type (black bars) and Cγ1-Cre-expressing mice (gray bars). In (b and c), NP-specific IgM (left panels) and IgG (right panels) antibody-secreting cells (ASC) in spleen (b) and bone marrow (c); filled circles represent data from individual wild-type mice, filled squares are data from Cγ1Cre/wt mice, and filled triangles are data from Cγ1Cre/Cre mice (all mice were CCTαflox/flox). Statistical differences were determined with 1-way ANOVA using Tukey’s multiple comparison test (p<0.03; *p<0.0001). Figure 7 Germinal center recall responses are impaired in Cγ1-Cre-expressing mice NP-specific germinal center B cells were identified using flow cytometry as described earlier. Representative data from multiple wild-type, Cγ1Cre/wt and Cγ1Cre/Cre mice are shown in (a). The frequency (b, c) and number (d, e) of IgM+ (b, d) and IgG1+ (c, e) NP-binding splenic germinal center B cells for multiple mice are shown. Statistical differences are indicated (p<0.02, p≤0.002, *p=0.0006, 1-way ANOVA with Tukey’s multiple comparisons). Figure 8 Changes in relative antibody affinity in immunized Cγ1-CCTα mice Sera from individual mice were titrated in ELISA to determine the amount of high-affinity (NP2-BSA) and total (NP20-BSA) NP-specific IgG (a) and IgM (b) for wild-type (filled circles), Cγ1Cre/wt (filled squares) and Cγ1Cre/Cre (filled triangles) mice. The ratio of NP2:NP20 titers is expressed as mean ratios ± SEM and are compiled from three independent experiments. Figure 9 XBP1S and PtdCho in the humoral immune response When B cells are activated by antigen and T cell help (as well as by T cell-independent stimuli), synthesis of phosphatidylcholine (PtdCho) and other membrane lipids increases to meet the needs of membrane biogenesis in dividing cells. At this point, activated B cells can differentiate into short-lived antibody-secreting cells (ASC), a developmental process that is coordinated by the XBP1S transcription factor and that requires an increased supply of membrane lipids like PtdCho to fuel expansion of the secretory apparatus. Alternatively, in the case of B cells activated with T cell help, entry into the germinal center reaction ushers in additional rounds of cell division that require more membrane lipids, followed by the potential of either memory B cell development or XBP1S-dependent, lipid-demanding differentiation of long-lived ASC. Highlights CCTα is required for antigen-specific, germinal center-derived antibody-secreting cells CCTα is required for antigen-specific, germinal center-derived memory B cells Phosphatidylcholine is required for germinal center B cell responses This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. Conflict of interest: The authors declare that they have no conflicts of interest with the contents of the manuscript. Author contributions: JWB initially conceived the idea of the project, developed the necessary mouse colony and assisted with experimental design and manuscript preparation. VS conducted experiments and analyzed data. IA assisted with ELISA assays and with management of the mouse colony. RAB conceived and performed experiments, analyzed results, prepared data for publication and wrote the manuscript. 1 Brewer JW Hendershot LM Building an antibody factory: a job for the unfolded protein response Nat Immunol 2005 6 23 29 15611778 2 Walter P Ron D The unfolded protein response: from stress pathway to homeostatic regulation Science 2011 334 1081 1086 22116877 3 Sriburi R Jackowski S Mori K Brewer JW XBP1: a link between the unfolded protein response, lipid biosynthesis, and biogenesis of the endoplasmic reticulum J Cell Biol 2004 167 35 41 15466483 4 Jackowski S Fagone P CTP: Phosphocholine cytidylyltransferase: paving the way from gene to membrane J Biol Chem 2005 280 853 856 15536089 5 Fagone P Sriburi R Ward-Chapman C Frank M Wang J Gunter C Brewer JW Jackowski S Phospholipid biosynthesis program underlying membrane expansion during B-lymphocyte differentiation J Biol Chem 2007 282 7591 7605 17213195 6 Sriburi R Bommiasamy H Buldak GL Robbins GR Frank M Jackowski S Brewer JW Coordinate regulation of phospholipid biosynthesis and secretory pathway gene expression in XBP-1(S)-induced endoplasmic reticulum biogenesis J Biol Chem 2007 282 7024 7034 17213183 7 Brewer JW Phospholipids: “greasing the wheels” of humoral immunity Biochim Biophys Acta 2013 1831 642 651 23051607 8 Kent C Eukaryotic phospholipid biosynthesis Annu Rev Biochem 1995 64 315 343 7574485 9 Rush JS Sweitzer T Kent C Decker GL Waechter CJ Biogenesis of the endoplasmic reticulum in activated B lymphocytes: temporal relationships between the induction of protein N-glycosylation activity and the biosynthesis of membrane protein and phospholipid Arch Biochem Biophys 1991 284 63 70 1846517 10 Reimold AM Iwakoshi NN Manis J Vallabhajosyula P Szomolanyi-Tsuda E Gravallese EM Friend D Grusby MJ Alt F Glimcher LH Plasma cell differentiation requires the transcription factor XBP-1 Nature 2001 412 300 307 11460154 11 Wen XY Stewart AK Sooknanan RR Henderson G Hawley TS Reimold AM Glimcher LH Baumann H Malek LT Hawley RG Identification of c-myc promoter-binding protein and X-box binding protein 1 as interleukin-6 target genes in human multiple myeloma cells Int J Oncol 1999 15 173 178 10375612 12 Hess DA Humphrey SE Ishibashi J Damsz B Lee AH Glimcher LH Konieczny SF Extensive pancreas regeneration following acinar-specific disruption of Xbp1 in mice Gastroenterology 2011 141 1463 1472 21704586 13 Adolph TE Tomczak MF Niederreiter L Ko HJ Bock J Martinez-Naves E Glickman JN Tschurtschenthaler M Hartwig J Hosomi S Flak MB Cusick JL Kohno K Iwawaki T Billmann-Born S Raine T Bharti R Lucius R Kweon MN Marciniak SJ Choi A Hagen SJ Schreiber S Rosenstiel P Kaser A Blumberg RS Paneth cells as a site of origin for intestinal inflammation Nature 2013 503 272 276 24089213 14 Lee AH Chu GC Iwakoshi NN Glimcher LH XBP-1 is required for biogenesis of cellular secretory machinery of exocrine glands EMBO J 2005 24 4368 4380 16362047 15 Smith KG Hewitson TD Nossal GJ Tarlinton DM The phenotype and fate of the antibody-forming cells of the splenic foci Eur J Immunol 1996 26 444 448 8617316 16 McHeyzer-Williams LJ Driver DJ McHeyzer-Williams MG Germinal center reaction Curr Opin Hematol 2001 8 52 59 11138628 17 Fagone P Gunter C Sage CR Gunn KE Brewer JW Jackowski S CTP:phosphocholine cytidylyltransferase alpha is required for B-cell proliferation and class switch recombination J Biol Chem 2009 284 6847 6854 19139091 18 Hayakawa K Hardy RR Herzenberg LA Herzenberg LA Progenitors for Ly-1 B cells are distinct from progenitors for other B cells J Exp Med 1985 161 1554 1568 3874257 19 Boes M Esau C Fischer MB Schmidt T Carroll M Chen J Enhanced B-1 cell development, but impaired IgG antibody responses in mice deficient in secreted IgM J Immunol 1998 160 4776 4787 9590224 20 Casola S Cattoretti G Uyttersprot N Koralov SB Seagal J Hao Z Waisman A Egert A Ghitza D Rajewsky K Tracking germinal center B cells expressing germ-line immunoglobulin gamma1 transcripts by conditional gene targeting Proc Natl Acad Sci U S A 2006 103 7396 7401 16651521 21 Wang Y Carter RH CD19 regulates B cell maturation, proliferation, and positive selection in the FDC zone of murine splenic germinal centers Immunity 2005 22 749 761 15963789 22 Shih TA Meffre E Roederer M Nussenzweig MC Role of BCR affinity in T cell dependent antibody responses in vivo Nat Immunol 2002 3 570 575 12021782 23 Dal Porto JM Haberman AM Kelsoe G Shlomchik MJ Very low affinity B cells form germinal centers, become memory B cells, and participate in secondary immune responses when higher affinity competition is reduced J Exp Med 2002 195 1215 1221 11994427 24 Wang L Magdaleno S Tabas I Jackowski S Early embryonic lethality in mice with targeted deletion of the CTP:phosphocholine cytidylyltransferase alpha gene (Pcyt1a) Mol Cell Biol 2005 25 3357 3363 15798219 25 Jackowski S Rehg JE Zhang YM Wang J Miller K Jackson P Karim MA Disruption of CCTbeta2 expression leads to gonadal dysfunction Mol Cell Biol 2004 24 4720 4733 15143167 26 Kroese FG Wubbena AS Seijen HG Nieuwenhuis P Germinal centers develop oligoclonally Eur J Immunol 1987 17 1069 1072 3301368 27 McGehee AM Dougan SK Klemm EJ Shui G Park B Kim YM Watson N Wenk MR Ploegh HL Hu CC XBP-1-deficient plasmablasts show normal protein folding but altered glycosylation and lipid synthesis J Immunol 2009 183 3690 3699 19710472 28 Todd DJ McHeyzer-Williams LJ Kowal C Lee AH Volpe BT Diamond B McHeyzer-Williams MG Glimcher LH XBP1 governs late events in plasma cell differentiation and is not required for antigen-specific memory B cell development J Exp Med 2009 206 2151 2159 19752183 29 Jackowski S Cell cycle regulation of membrane phospholipid metabolism J Biol Chem 1996 271 20219 20222 8702749 30 van der Sanden MH Houweling M van Golde LM Vaandrager AB Inhibition of phosphatidylcholine synthesis induces expression of the endoplasmic reticulum stress and apoptosis-related protein CCAAT/enhancer-binding protein-homologous protein (CHOP/GADD153) Biochem J 2003 369 643 650 12370080 31 Zhang D Tang W Yao PM Yang C Xie B Jackowski S Tabas I Macrophages deficient in CTP:Phosphocholine cytidylyltransferase-alpha are viable under normal culture conditions but are highly susceptible to free cholesterol-induced death. Molecular genetic evidence that the induction of phosphatidylcholine biosynthesis in free cholesterol-loaded macrophages is an adaptive response J Biol Chem 2000 275 35368 35376 10944538 32 Horikawa K Martin SW Pogue SL Silver K Peng K Takatsu K Goodnow CC Enhancement and suppression of signaling by the conserved tail of IgG memory-type B cell antigen receptors J Exp Med 2007 204 759 769 17420266 33 Barrington RA Pozdnyakova O Zafari MR Benjamin CD Carroll MC B lymphocyte memory: role of stromal cell complement and FcgammaRIIB receptors J Exp Med 2002 196 1189 1199 12417629
PMC005xxxxxx/PMC5125837.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 7609829 6055 Neuropathol Appl Neurobiol Neuropathol. Appl. Neurobiol. Neuropathology and applied neurobiology 0305-1846 1365-2990 27424496 5125837 10.1111/nan.12337 NIHMS804377 Article Human adult neurogenesis across the ages: An immunohistochemical study Dennis CV B Med Sci (Hons) 1 Suh LS B Sci (Hons) 12 Rodriguez ML MBBS, FRCPA 1 Kril JJ PhD 1 Sutherland GT PhD 1 1 Discipline of Pathology, Sydney Medical School, University of Sydney, Camperdown, NSW 2006, Australia 2 Dementia Research Unit, School of Medical Sciences, University of New South Wales, Kensington, NSW 2052, Australia Corresponding author: Greg Sutherland, Discipline of Pathology, Rm 6211 Level 6W, Charles Perkins Centre D17, University of Sydney, NSW 2006, Australia; Tel +61 2 90367233; Fax +61 2 86271601; [email protected] 24 7 2016 28 8 2016 12 2016 01 12 2017 42 7 621638 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Aims Neurogenesis in the postnatal human brain occurs in two neurogenic niches; the subventricular zone (SVZ) in the wall of the lateral ventricles and the subgranular zone of the hippocampus (SGZ). The extent to which this physiological process continues into adulthood is an area of ongoing research. This study aimed to characterise markers of cell proliferation and assess the efficacy of antibodies used to identify neurogenesis in both neurogenic niches of the human brain. Methods Cell proliferation and neurogenesis were simultaneously examined in the SVZ and SGZ of 23 individuals aged 0.2–59 years using immunohistochemistry and immunofluorescence in combination with unbiased stereology. Results There was a marked decline in proliferating cells in both neurogenic niches in early infancy with levels reaching those seen in the adjacent parenchyma by four and one year of age, in the SVZ and SGZ, respectively. Furthermore, the phenotype of these proliferating cells in both niches changed with age. In infants, proliferating cells co-expressed neural progenitor (epidermal growth factor receptor), immature neuronal (doublecortin and beta III tubulin) and oligodendrocytic (Olig2) markers. However, after three years of age, microglia were the only proliferating cells found in either niche or in the adjacent parenchyma. Conclusions This study demonstrates a marked decline in neurogenesis in both neurogenic niches in early childhood, and that the sparse proliferating cells in the adult brain are largely microglia. adult neurogenesis human postmortem brain tissue immunohistochemistry microglial proliferation subventricular zone and subgranular zone Introduction It has been proposed that enhancement of endogenous neurogenesis using pharmaceutical or environmental measures could be an effective therapeutic strategy for neurodegenerative diseases [1, 2]. The ultimate success of these efforts would appear to be predicated on there being sufficient levels of neurogenesis in the aged human brain to respond to manipulation. The prolificity of adult neurogenesis in all mammals declines with age [3] although the timing of this and the extent to which it occurs in the human brain remains the subject of considerable debate. There are two major neurogenic niches in the adult human brain, the subventricular zone (SVZ) underlying the wall of the lateral ventricles and the subgranular zone (SGZ) in the dentate gyrus of the hippocampus. In both niches, pluripotent neural stem cells give rise to transit amplifier cells that differentiate into neuroblasts, astrocytes and oligodendrocyte precursor cells [3]. Studies in various lower mammals have shown that neuroblasts from the SVZ migrate, via the rostral migratory stream (RMS), and integrate as interneurons into the granule cell layer of the olfactory bulb where they are thought to modulate olfactory learning and performance [4, 5]. In contrast, neuroblasts from the SGZ migrate a short distance to the adjacent granule cell layer (GCL) [6] where survivors are thought to facilitate the temporal separation of spatial memories [7]. In pioneering work Curtis and colleagues used immunohistochemistry to identify the endogenous proliferation marker, proliferating cell nuclear antigen (PCNA), to demonstrate the presence of a functional RMS in humans [8]. In contrast, Sanai and colleagues used an alternative proliferation marker, Ki67 and the immature neuronal marker doublecortin (DCX) [9] to show that SVZ neurogenesis and migration along the RMS ceased by two years of age [10]. Knoth and colleagues used DCX and PCNA to demonstrate residual neurogenesis in the SGZ in adults up to 100 years of age [11]. To our knowledge no study has explored proliferation in the human SGZ using Ki67 across a wide age range. More recently an alternative methodology, radioactive carbon dating of neuronal DNA, has shown numerous adult-born neurons in the adult human hippocampus [12] and the SVZ [13], although the neuroblasts produced by the latter appear to migrate into the adjacent caudate nucleus [13] rather than entering the olfactory bulb via the RMS [14]. One of the major advantages of this technique is that unambiguous confirmation of proliferating cell fate can be confirmed using mature cell markers such as the neuronal marker, NeuN. However the technique is limited to one or two specialist laboratories, worldwide. Notwithstanding the contribution of radioactive carbon dating to this area, immunohistochemistry remains a much more accessible technique, although findings to date on the extent of adult neurogenesis have been marker-dependent. In an attempt to clarify the true extent of neurogenesis in both niches of the human brain we have compared and contrasted Ki67, PCNA and DCX immunostaining of postmortem tissue across a range of ages. Materials and methods Cases This project was conducted following approval from the Human Research Ethics Committee of the University of Sydney (HREC #13027). Formalin-fixed paraffin-embedded (FFPE) sections of the wall of the lateral ventricle and adjacent head of the caudate nucleus (CN) at the level of nucleus accumbens, and the hippocampus at the level of the lateral geniculate nucleus, were obtained from neurologically normal subjects (0.2 – 59 years; n = 23) (Table 1). Additional sections from the superior frontal gyrus (SFG) were also obtained from the three eldest adults (Table 1 #21 – 23). Tissue from eight individuals ≥ 24 years (adult) was obtained from the New South Wales Brain Tissue Resource Centre (NSW BTRC). Tissue from 15 individuals ≤16 years (juvenile), including 11 ≤4 years (infant), were obtained from the NSW Department of Forensic Medicine. Adult and juvenile tissues were fixed in 15 and 20% formalin, respectively, for 2–3 weeks, followed by paraffin embedding. NSW BTRC provided information on age, gender, postmortem interval (PMI), cause of death, brain pH, alcohol intake and liver pathology. Only age, PMI, gender and the cause of death were available for the remainder of cases (Table 1). Immunohistochemistry 7μm FFPE sections were dewaxed through graded alcohols prior to antigen retrieval and immunostaining. Antigen retrieval was performed in a decloaking chamber (BioCare medical DC2002, Concord, USA) for 30 minutes at 95°C using 10 mM Tris/1 mM EDTA buffer pH 9.0. A higher maximum temperature of 110°C was necessary for antigen retrieval for Ki67 due to its greater susceptibility to the effects of fixation [15]. Sections were washed in 0.05M Tris/0.015M NaCl/0.05% Tween-20 (TBST, pH 7.4) and incubated for 10 minutes in 3% H2O2 in methanol, followed by 15 minutes in 10% normal horse serum (NHS) before overnight incubation with the following primary antibodies at 4°C: Ki67 (Clone MIB-1; mouse monoclonal, 1:500; M7240, Dako, Denmark), doublecortin (DCX; goat polyclonal, 1:500, sc-8066, Santa Cruz Biotechnology, California) or PCNA (Clone PC10, mouse monoclonal, 1:500, sc-56, Santa Cruz Biotechnology). Primary and secondary antibodies were diluted in 1% NHS prepared in TBST. The slides were incubated in the secondary antibody (1:200; biotinylated anti-mouse BA-2000 and biotinylated anti-rabbit BA-1000, Vector Laboratories, Meadowbrook, Queensland) for 30 minutes, then with an avidin-biotin-peroxidase complex (1:100; PK-6100, Vector Laboratories) for 30 minutes. Negative controls, omitting the primary or secondary antibody, were run concurrently for all antibodies. Visualisation was with DAB in 5% H2O2 for 2–3 minutes followed by hematoxylin counterstaining. Immunofluorescence Immunofluorescence co-localisation studies with cell specific markers were performed using the same antigen retrieval protocol outlined above. Sections were then washed in 0.01M phosphate buffered saline (Bioline, Alexandria, Australia) with 0.1% Tween-20 (PBST), blocked for 20 minutes in 10% NHS and incubated overnight at 4°C in a primary antibody cocktail containing up to three antibodies raised in different species including: Ki67; DCX; PCNA; beta III tubulin (rabbit polyclonal, 1:200, ab18207, Abcam, Waterloo, NSW), ionized calcium binding adaptor molecule 1 (Iba1, rabbit polyclonal, 1:1000; 019-19741, Wako Pure Chemicals, Osaka, Japan), glial fibrillary acidic protein (GFAP, rabbit polyclonal, 1:200; Z0334, Dako, Denmark), epidermal growth factor receptor (EGFR, rabbit monoclonal, 1:200; 04-338, Merck Millipore, Billerica, USA) and oligodendrocyte lineage transcription factor 2 (Olig2, clone ERP2673 1:200; ab109186, Abcam). The sections were protected from UV light and incubated in a secondary antibody cocktail (1:200: AlexaFluor 488, donkey anti-goat IgG, A10055; AlexaFluor 568, donkey anti-mouse IgG, A10037; AlexaFluor 647, donkey anti-rabbit IgG, A31573) for 30 minutes. Tissue autofluorescence was quenched using 0.1% Sudan black B (B.D.H Laboratory Chemicals Group, United Kingdom) in 70% ethanol for 4 minutes. Sections were then cover slipped using ProLong® Gold Antifade Reagent with DAPI (P36935, Life Technologies, Mulgrave, Victoria) and left for 24 hours at room temperature, sealed with nail polish and stored at 4°C. Images were then obtained using confocal microscopy (Zeiss LSM 510 META Spectral microscope) at the Advanced Microscope Facility of the Bosch Institute (University of Sydney). Stereological quantification Ki67+, DCX+ and PCNA+ cells were counted within the entire SVZ and SGZ of each section using an Olympus BX53 microscope equipped with StereoInvestigator (v5.65). The SVZ was defined as the tissue extending from the ependymal layer (layer I) to the medial margin of the head of the CN which is delineated by a distinct myelin layer (layer IV) [16, 17]. Accordingly, all immunopositive cells in the SVZ were counted within the hypocellular layer (layer II) and the adjacent layer containing a ribbon of astrocyte cell bodies (layer III) [17] (Fig. S1A). As newborn neurons can migrate across the width of the GCL, the SGZ was arbitrarily defined as extending from the junction of the GCL and the molecular layer to the boundary between the polymorphic layer and the CA4 subregion of the hippocampus (the latter defined by the presence of pyramidal neurons) (Fig. S1B). The “Meander Scan” function was used at 40x magnification and immunopositive cells were recorded manually at each scan site. After the meander scan had sampled the entire area of interest, the area scanned and the total number of immunopositive cells counted were recorded and cell density calculated (cells/mm2). Ki67+ cells were also counted in the SFG of three neurologically normal adults using unbiased stereology. Cortical ribbons were outlined using a 4x objective and then randomly sampled using the optical dissector probe function of the software. Counts were then performed at 40x magnification with a meander sampling rate of 1% and a counting frame of 22,500 μm2. Images of the proximal RMS were obtained using a slide scanning microscope (Zeiss Axio Scan.Z1, Gena, Germany). Using Zen imaging software (v1.1.2.0, 2012 Blue Edition, Gena, Germany) the RMS was outlined prior to scanning and images, taken at 100x magnification, stitched together and exported. Confocal fluorescence microscopy was used to phenotype Ki67+ cells throughout the SVZ and SGZ of seven individuals (six juveniles including five infants). Neurogenic events were explored by triple staining with Ki67 and DCX and either EGFR (a putative transit amplifier cell marker) [18] or beta III tubulin (an immature neuronal marker) [19]. To determine the overall composition of proliferating cells, sections were triple-stained with Ki67, DCX and either GFAP (a mature astrocyte marker [20, 21], Iba1 (microglial marker) [22, 23] or Olig2 (an oligodendrocytic lineage marker) [24, 25]. Fluorescent images were imported using LSM Image Browser v3.5 (Zeiss) and Ki67+ cells were scored as positive or negative for DCX and the cell-specific markers by an investigator blinded to case status (CD). Statistical analysis Statistical analyses were performed using the statistical software package GraphPad Prism 6 (v6.0d, GraphPad Software, Inc. 2013). Juvenile versus adult group differences were assessed using a Student’s t-test. A ratio-paired t-test was used to compare the relative densities of markers across all individuals. The relationships between markers and age were evaluated by Pearson correlation. The level of significance was chosen at p = 0.05 for all tests. Results Baseline proliferation in the adult human brain The baseline level of cell proliferation in the adult brain outside the neurogenic niches was determined by counting Ki67+ cells in the SFG and the CN of three adults. Ki67+ cells were identified in all cases (mean density 2.9 ± 2.1 cells/mm2) and our previous work has shown these cells to be almost exclusively microglia [23]. No DCX+ cells were identified in the CN. Markers of proliferation and neurogenesis in the subventricular zone Ki67+ cells, exclusively with nuclear staining, were distributed uniformly along the SVZ at all ages (Fig. 1A, C, E, G). DCX+ cells were distributed unevenly along the SVZ, often present as aggregates (Fig. 1B and D). DCX staining varied with age, being somatic (perinuclear), dendritic (processes) or a combination of both in individuals ≤ 4 years of age, predominantly somatic in 7–16 year olds (Fig. 1B, D, F) and absent in individuals older than 16 (Fig. 1H). PCNA+ cells were distributed uniformly along the SVZ in all individuals. Staining was nuclear but varied considerably in intensity. PCNA staining of ependymal cell nuclei, more intense in adults, was seen in all cases. There was no staining of ependymal cells with either Ki67 or DCX (Fig. S2). Proliferating cells in the human subventricular zone decline with age There was a marked decrease in cell proliferation in the infant SVZ (Fig. 1). The density of Ki67+ cells decreased dramatically over the first year of life and by four years of age was similar to the density in adult SFG and CN (2.6 cells/mm2) (Fig. S3). Similarly, the density and total number of DCX+ cells in the SVZ declined with age. DCX+ cell density declined from 234.1 cells/mm2 in the youngest individuals (0.2 year-old) to 4.6 cells/mm2 in the 16 year old, although the highest density was seen in a three year-old (391.2 cells/mm2). DCX+ cells were not identified in individuals older than 16 years (Fig. S3). There was a positive correlation between Ki67+ and DCX+ cell densities in the juvenile SVZ (r2 = 0.39; p = 0.014) with DCX+ cells exceeding Ki67+ cells in 13/15 cases (mean = 2.6 fold; 95% CI = 1.7 – 4.2)(Fig. S3). The density of PCNA+ cells was greater than the density of Ki67+ cells in juvenile cases (mean = 3.9 fold; 95% CI = 1.8–8.5, p = 0.002). However, unlike Ki67+ and DCX+ cells, the PCNA+ cell density did not correlate with age (Fig. S3). The density of PCNA+ cells in adults (mean = 421.3 ± 121.1, range 0.76–799.3 cells/mm2, n = 8) was significantly higher than in juveniles (mean = 124.3 ± 26.0, range = 14.7–327.0 cells/mm2, n = 15, p = 0.005). The phenotype of proliferating cells in the SVZ changes with age Ki67+ cells in the juvenile SVZ (n = 6; Table 1) variably co-labelled with DCX (Fig. 2), beta III tubulin (Fig. 2A–D), EGFR (Fig. 2E–H), Olig2 (Fig. 2I–L), and Iba1 (microglia) (Fig. 2M–P) but not GFAP (astrocytes) (Fig. 2Q–T). The phenotype of Ki67+ cells in the SVZ differed with age (Fig. 3). In the youngest individual examined by confocal microscopy (1 year) Ki67 co-localised with DCX (65%). Triple labelling showed that Ki67+/DCX+ cells could be further subdivided into Ki67+/DCX+/EGFR+ (29%), Ki67+/DCX+/beta III tubulin+ (19%) and Ki67+/DCX+ only (17%) cells. Additionally, there were occasional Ki67+/DCX−/EGFR+ and Ki67+/DCX−/beta III tubulin+ cells. The remaining Ki67+ cells were either Olig2+ (20%) or Iba1+ (10%). In a 3.9 year-old Ki67+ cells co-labelled with DCX, Olig2 or Iba1 but no Ki67+/DCX+/EGFR+ or Ki67+/DCX+/beta III tubulin+ cells were identified. No Ki67+/DCX+ cells were seen at 14 years and older with all Ki67+ cells examined in the adult SVZ being Iba1+ (Fig. 3). Ki67+/GFAP+ cells were not seen at any age. Markers of proliferation and neurogenesis in the SGZ In juvenile brains, Ki67+ cells were common within, but not confined to, the SGZ. Ki67+ cells were also found in the molecular layer of the dentate gyrus and in the CA4 region of the hippocampus (Fig. 4A and C). Ki67+ cells were exceptionally rare in the adult SGZ and adjacent regions (Fig. 4E and G). The distribution and staining pattern of DCX+ cells differed with age. In juveniles, DCX+ cells were densely clustered within the PML near the GCL border with either somatic, dendritic or a combination of somatic and dendritic staining (Fig. 4B and D). Clustered DCX+ cells were not present in cases over three years of age and only sparse DCX+ cells were seen in the older juveniles and adults (2/8; Fig. 4F and H). Similar to the SVZ, nuclear PCNA staining varied in intensity and was distributed uniformly across all areas of the dentate gyrus (Fig. S2). Proliferating cells in the human SGZ decline with age As noted in the SVZ, there was a dramatic decline in the density of Ki67+ cells in the SGZ during the first year of life. In a 0.2 and 0.3 year-old there were 17.9 and 20.6 cells/mm2, respectively. At one year of age, the mean density was 3.77 ± 1.39 cells/mm2 (n = 3), similar to the Ki67+ cell density in the adult SFG and CN (baseline levels) (p = 0.66; Fig. S4). Ki67+ cells were only present in 2/8 adult SGZ and their mean density (0.27 ± 0.18 cells/mm2) was lower than in the adult SFG (2.6 cells/mm2; p = 0.006). As in the SVZ, the density of DCX+ cells exceeded that of Ki67+ cells in the SGZ in 14/15 juveniles (mean ratio = 18.0, 95% CI = 6.9 – 47.0). The density of both Ki67+ and DCX+ cells declined with age although the decline was more rapid for Ki67+. At one year of age, DCX+ cells were still numerous (mean = 255 ± 81 cells/mm2) but then either absent from, or very sparse in (n = 3, mean = 0.45 ± 0.30 cells/mm2), the eight adults examined (Fig. S4). The phenotype of proliferating cells in the SGZ changes with age In the six juveniles examined (n = 6, 1–14 years) Ki67+ cells in the SGZ variably co-expressed DCX, EGFR (Fig. 5A–D), Olig2 (Fig. 5E–H) and Iba1 but not beta III tubulin (Fig. 5M–P) or GFAP (Fig. 5Q–T). In younger individuals (1 and 1.5 years) the majority of Ki67+ cells co-expressed the transit-amplifier cell marker, EGFR. Ki67+/DCX+ and Ki67+/Olig2+ cells were also seen in infants up to 3.9 years. Some Ki67+/DCX+ cells co-expressed Olig2 but none co-expressed beta III tubulin or EGFR. In contrast to the SVZ, where 10% of Ki67+ cells in juveniles co-expressed Iba1, no Ki67+/Iba1+ cells were identified in the SGZ < 3.9 years of age. All Ki67+ cells in older individuals (14 and 54 year-olds) were Iba1+ (Fig. 6). Comparison of Ki67 and DCX in the subventricular and subgranular zones In all cases, Ki67+ cells were more numerous in the SVZ than in the SGZ, although the Ki67+ cell density varied considerably between cases. In the juvenile group, the average SVZ/SGZ density ratio was 9.7 (n = 13; 95% CI = 5.9–16.0). The two cases with no Ki67+ cells in the SGZ were excluded from this analysis. Since in adults, Ki67+ cells were only identified in the SGZ in 2/8 cases (37 and 54 year-olds) a meaningful density ratio between the two niches could not be determined. Rostral migratory stream In some individuals a RMS from the rostral extremity of the frontal horn of the lateral ventricle, could be identified by its distinctive SVZ-like cytoarchitecture (greater cellularity) and the presence of epdendymal islets [10, 26]. The RMS was bordered by the CN and genu of the corpus callosum (CC) (Fig. 7 and Fig. S5). Numerous Ki67+ and DCX+ cells were present throughout the RMS of the youngest individual (0.2 year-old; Fig. 7A, D) but were spatially distinct. DCX+ cells were seen more dorsomedially adjacent to the CC (Fig. 7A–C) while Ki67+ cells were largely localised to the ventrolateral aspect of the RMS adjacent to the head of the CN (Fig. 7D–E). In the 0.2 and one year-old approximately 15% of the DCX+ cells showed colocalisation with Ki67 (Fig. 7F, G). Similar to the SVZ, there was a dramatic decline in the number of DCX+ and Ki67+ cells in the RMS during the first year of life with only a single DCX+ cell and rare Ki67+ cells seen in older individuals. In all the adults examined (n = 8, 24–59 years) there was only one case (a 54 year-old) with a single DCX+ cell in the RMS (Fig. 7G) while the rare Ki67+ cells seen in the adult RMS co-localised with Iba1 (Fig. 7H). Ependymal islets were dispersed throughout the RMS at all ages. In individuals younger than 2.5 years there were DCX+ and Ki67+ cells (Fig. S5 D–F) both within (non-patent) and surrounding the islets. In contrast to Ki67 and DCX, the majority of RMS cells in all cases including ependymal cells of the ependymal islets were immunopositive for PCNA (Fig. S5A–C). Furthermore, in the infant brains PCNA staining was at two distinct intensities, a small subgroup of intensely stained cells with a similar density and distribution to the Ki67+ cells, and the majority of PCNA+ cells, including ependymal cells, that were moderately stained (Fig. S5A, B). In older brains intensely staining PCNA+ cells were rare, paralleling the density and distribution of Ki67+ cells, while the moderately staining population remained (a 54 year-old; Figure S5C). Discussion The potential benefits of therapeutically manipulating endogenous adult neurogenesis for patients with neurodegenerative diseases are immense. Both pharmaceutical and environmental manipulation has been shown to affect adult neurogenesis in many mammalian species but successful manipulation in humans is likely to depend on there being a sufficient level of neurogenesis at the ages that these diseases are prevalent. We had previously shown that there was no difference in cell proliferation in the SVZ of chronic alcoholics compared to neurologically normal controls [16]. In that study, and in previous studies examining the human SVZ and RMS [8, 10] the density of proliferative events depended on whether PCNA or Ki67 was used as a marker. Far fewer proliferating cells are identified in studies using Ki67 but epitope retrieval from fixed human brain tissue is known to be more difficult than for PCNA [15]. In order to clarify the true state of cell proliferation in adult neurogenic niches, fixed postmortem human brain tissue from neurologically normal infants to middle-aged subjects were examined. DCX was also included as an immature neuronal marker whose expression should partially overlap with both proliferative markers and confirm neurogenesis. Our results show that the density of Ki67+, DCX+ and Ki67+/DCX+ cells in the SVZ and SGZ is high prior to four and one year of age, respectively but then decreases markedly with age. In the SVZ this replicates the results of Sanai and colleagues in juveniles up to 18 months of age [10]. Furthermore the pattern of DCX+ immunostaining in the SGZ of juveniles and adults in this study is remarkably similar to the exponential decrease with ageing described by Knoth and colleagues [11]. Again, consistent with our results, Knoth et al. demonstrated that there was no relationship between PCNA+ and DCX+ cell density across a wide range of ages. PCNA+ cells were more frequent than Ki67+ cells both within and outside the neurogenic niches and PCNA also stained the nuclei of mature ependymal cells. This has been reported in other studies [16, 27–29] as has PCNA staining of mature leptomeningeal cells [30]. It is not clear whether staining in these non-proliferating cell populations is specific, specific but identifying a modified form of PCNA [31], or is non-specific, detecting a shared epitope on an unrelated protein. In the present study, non-specific ependymal staining due to cross reactivity with secondary anibodies was ruled out by the inclusion of a no primary antibody negative control. Furthermore the likelihood of non-specificity was somewhat mitigated by the PCNA immunostaining pattern being seen with multiple PCNA antibodies raised in different species and from different clones (data not shown). Unlike DCX and Ki67, the majority of cells within the adult RMS were PCNA immunopositive including the ependymal islet cells. However, in the infant RMS there was a combination of this adult pattern interspersed with a few intensely stained cells. These intensely stained cells resembled the Ki67 pattern of immunopositivity and may represent the true proliferative population. If staining in the majority of PCNA+ cells is specific, then PCNA must serve a distinct function in mature cells. Roles in DNA repair or transcription have been suggested [32] and a recent study showed that there are two pools of PCNA in mammalian cells, one involved in replication and a second soluble nucleoplasmic pool of unknown function [33]. Our findings are consistent with PCNA antibodies labelling two pools of PCNA in the infant human SVZ and RMS, with only the non-proliferative pool remaining in the adult. Notwithstanding the need for further work to delineate the functional significance of moderate PCNA expression in the dentate gyrus, ependymal layer and adjacent areas, our results strongly suggests that PCNA-based IHC studies will overestimate proliferative events in human postmortem brain tissue. Notwithstanding DCX immunostaining of human astrocytes [34] and mature murine neurons [35, 36], our findings in the SVZ and SGZ combined with those of Sanai et al. in the SVZ [10] suggest that within the confines of the neurogenic niches, the combination of DCX and Ki67 immunostaining gives an accurate depiction of adult neurogenesis. This conclusion is supported by studies that employed either Ki67 [37] or DCX [11] in the SGZ. These studies also reported changes in the subcellular location of DCX [11, 37]. Boekhoorn et al. suggested that somatic DCX staining was related to PMI [37], however Knoth and colleagues, having matched for PMI, considered that somatic staining identified the most immature progenitor cells, notwithstanding these cells being found throughout the GCL [11]. Similar to Knoth et al, we found that somatic DCX staining was identified throughout the GCL independent of the PMI. Co-localisation studies enabled us to explore the changing cell phenotype in the SVZ and SGZ with age. In infants there was evidence of neurogenesis with transit-amplifier cells (Ki67+/DCX+/EGFR+ and Ki67+/DCX−/EGFR+) and neuroblasts (Ki67+/DCX+/beta III tubulin+) in both niches consistent with previous studies [3, 38, 39] as well as oligodendrogenesis with Ki67+/DCX+/−/Olig2+ cells [40]. GFAP was also expected to co-localise with Ki67 in the neurogenic niches due to the likely presence of both the GFAP-δ+ type B (neural stem) cells [41] and newborn astrocytes [3]. The lack of co-localisation may be partially explained by a largely quiescent neural stem cell population at the ages we examined. In this respect, our findings agree with a previous study that demonstrated no Ki67+/GFAP-δ+ cells in the aged human (82 year-old) SVZ [42]. Whereas, the extent of ongoing astrogliogenesis in the adult human SVZ or SGZ, unlike in the rodent SVZ [43], is unknown. In terms of the RMS our results are consistent with previous studies that suggested that chains of migrating neuroblasts are not seen in the human RMS after 2 years of age [10]. They are also in agreement with studies using atmospheric radioactive carbon from nuclear weapon testing that showed no evidence of adult-born neurons in the olfactory bulb [13, 14]. Ernst et al. suggested that SVZ-derived neurons deviated into the adjacent CN rather than continuing along the RMS to the olfactory bulb in the adult human brain [13], although our current and previous work in alcoholic and normal brains [16, 23] found no evidence of migration of proliferative cells into the CN. One caveat with our characterisation of the RMS is that it relied on a single coronal section at the level of the CN. This contrasts with more extensive analyses undertaken previously [8, 10, 26] and our findings must be considered in this context. Nevertheless, the distinct spatial separation of the majority of Ki67+ and DCX+ cells in the infant RMS suggests that proliferation, including neurogenesis, may continue along the length of the RMS rather than being confined to the SVZ [26]. A proliferative RMS has been shown in other species [44], [45], [46] and provides an alternative explanation to the traditional view that a proportion of neuroblasts remain mitotically active for their entire journey along the RMS[8]. In contrast to our findings in the SGZ, Spalding and colleagues, using atmospheric radioactive carbon techniques, suggested that human SGZ neurogenesis continues throughout life [12]. These apparently discordant results may be explained, at least in part, by differences in sampling techniques. Firstly Spalding et al. homogenised the entire hippocampus, approximately 7cm long, from one side of each brain they studied, while most IHC studies typically examine 5–48 μm sections. Usng an independent stereological measurement of the neurogenic volume of the human hippocampus [47], Spalding et al.’s estimation of 700 new neurons per day would equate to 5.2 neurons/mm3/day. Now given that the half-life of DCX expression in the human SGZ is unknown, an extrapolation from rat data [48], suggests that 5.2 neurons/mm3/day would equate to approximately 150 DCX+ cells/mm3. This figure is comparable to the 90–150 DCX+ cells/mm3 (0.45 ± 0.30 cells/mm2 or one DCX+ cell every 2 to 3 sections) found in the SGZ of 3/8 adults in the current study. Therefore the results of SGZ studies utilising either exogenous markers such as 14C [12] and BrdU [11] or endogenous IHC markers, may well be in agreement. At issue is whether the generation of 700 new neurons a day or 0.004% of the total granule cell population [12] is functionally significant. This estimate by Spalding and colleagues is approximately 10-fold lower than for rats (0.03%) [49] and mice (0.06%)[50] but is comparable to macaques (0.004 – 0.02%) [51, 52]. The marked differences between infants and adults in the current study suggests that in humans, the decline in neurogenesis with age is more rapid than in other mammals and that in adults, neurogenesis is functionally insignificant. Our conclusion is supported by a recent meta-analysis of hippocampal neurogenesis studies encompassing seven mammalian species, including humans, that suggested that age-related decline in humans is underestimated and that from middle age “neurogenesis occurs at relatively low, and perhaps negligible levels” [53]. Several hypotheses have been proposed to explain the differences in adult neurogenesis between humans and lower mammals. Ongoing neurogenesis in the adult SVZ may reflect the dominant role of olfaction in macro-osmic rodents [54] and sheep [55] compared with micro-osmic primates [14, 56]. In the adult rodent SGZ neurogenesis may be necessary to erase ‘superfluous’ memories and ensure that the hippocampus remains receptive to new memories [7]. In infant humans, a similar process may be the basis of infantile amnesia [7]. Since, unlike lower mammals, primates including humans, do not rely on the hippocampus for remote memory recall or reconsolidation [57], functionally significant neurogenesis in the adult SGZ may be unnecessary. Away from the SVZ and the SGZ, glial proliferation occurs at low levels even in healthy, adult human brains [58]. This has been largely attributed to oligodendrocyte precursor cell [59] and more recently, to microglia [23]. The rate of turnover of oligodendrocyte precursor cells in the adult human is unclear, in part due to the lack of a definitive marker. While NG2 (Chondroitin Sulphate Proteoglycan 4) has been used [59], this antigen is extremely sensitive to fixation, precluding its use in FFPE tissue. Our results demonstrate that within the adult human SVZ and SGZ the vast majority of proliferating cells are microglia and the degree of proliferation is similar to that noted in the surrounding brain parenchyma [23]. Similarly, Doorn et al. demonstrated that microglia are the predominant proliferating cells in the adult hippocampus [60] Conclusions Although commonly used,, PCNA is not aspecific marker of proliferation in postmortem human brain tissue. A combination of Ki67 and DCX immunoreactivity demonstrates that in humans, the density of proliferating cells, and neuroblasts in particular, declines markedly with age in the SVZ and SGZ and by four years of age is no greater than in the adjacent parenchyma. Although this low level of neurogenesis in neurologically normal adult humans suggests that any further reduction is unlikely to significantly contribute to the pathogenesis of neurodegenerative diseases the reported increases in neurogenesis in patients with conditions such as of Huntington’s disease [27] and epilepsy [61] suggests that therapeutic intervention to augment neurogenesis may be possible. Supplementary Material Supp Fig S1 Fig. S1. Counting frame for unbiased stereology of the SVZ and SGZ Micrographs of the SVZ and SGZ show the regions (red) where Ki67, DCX and PCNA were quantified. (A) The SVZ was defined as the area immediately beneath the ependymal layer of the lateral ventricle and the myelin layer that separates the SVZ from the underlying parenchyma. (B) The SGZ was defined as the area between the granule cell layer of the hippocampus and the boundary between the polymorphic layer and the CA4 subregion of the hippocampus which was determined by the presence of pyramidal neurons. Supp Fig S2 Fig. S2. PCNA immunostaining in the human SVZ and SGZ PCNA immunostaining of the SVZ shows a similar staining pattern in four cases of different ages (A, C, E and G) with nuclear staining of variable intensity as well as staining of the post-mitotic ependymal layer. In the SGZ, PCNA immunostaining shows PCNA+ cells throughout all regions of the hippocampus in juveniles (B, D and F) and weak staining in a 54 year-old individual (H). 200x magnification. Scale bar = 100μm. Supp Fig S3 Fig. S3. Quantification of endogenous markers of proliferation and neurogenesis in the human SVZ Scatter plots show the density of (A) Ki67+ and (B) DCX+ in the SVZ of 23 neurologically normal donors aged 0.2–59 years. A solid line in (A) represents the mean density of Ki67+ cells in adult cortex and dotted lines indicate the standard deviation (2.90 ± 2.1 cells/mm2). (C) A histogram shows the density of immunopositive cells for Ki67, DCX and PCNA in each case. Supp Fig S4 Fig. S4. Quantification of endogenous markers of proliferation and neurogenesis in the human SGZ Scatter plots show the density of (A) Ki67+ and (B) DCX+ cells in the SGZ of 18 neurologically normal donors aged 0.2–59 years. A solid line in (A) represents the mean density of Ki67+ cells in adult cortex and dotted lines indicate the standard deviation. (C) A histogram shows the density of immunopositive cells for Ki67, DCX and PCNA in each case. Supp Fig S5 Fig. S5. Endogenous markers of proliferation in the RMS Photomicrographs show traces of the RMS extending rostrally between the corpus callosum (CC) and caudate nucleus (CN). Higher magnifications of regions corresponding to black rectangles are shown in insets. Collages of overlapping fields show PCNA+ cells within the RMS of (A) a 0.2 year-old. The upper inset shows occasional intensely stained cells among many moderately stained PCNA+ cells. The lower inset shows a non-patent ependymal islet with an invagination of putative proliferative (SVZ-like) tissue. Intensely stained cells are present within the islet, the ependyma and adjacent tissue. (B) a one year-old with a single intensely stained cell adjacent to a moderately stained ependymal islet and (C) a 54 year-old with moderately stained PCNA+ cells throughout the RMS. The inset shows the typical moderately intense immunostaining of the ependymal islets and adjacent SVZ-like tissue. (D–F) Collages of overlapping fields show Ki67 immunostaining in the human RMS of (D) a 0.2 year-old with upper inset showing Ki67+ cells clustered towards the CN and the lower inset showing an ependymal islet with Ki67+ cells within the islet as well as the adjacent SVZ-like tissue; (E) one year-old with rare Ki67+ cells (black arrows) in the RMS adjacent to the CN and (F) a 54 year-old with a single Ki67+ cell (black arrow). CVD, GTS and JJK designed the study. CVD and LSS contributed to data acquisition. DS, MLK and JJK contributed to provision of clinical and pathological data. CVD and GTS carried out data analysis and wrote the first draft. All authors contributed to manuscript review. The authors would like to thank the donors and their families for their kind gift that has allowed this research to be undertaken and the New South Wales Brain Tissue Resource Centre (NSW BTRC) and the NSW Department of Forensic Medicine for providing tissue samples. We would like to acknowledge Dr Louise Cole (Core Facilities Manager, Bosch Institute Advanced Microscopy Facility, The University of Sydney) for her support and assistance with the confocal microscopy. The NSW BTRC is part of the NSW Brain Bank Network and is supported by the University of Sydney, National Health and Medical Research Council (NHMRC), Schizophrenia Research Institute and the National Institutes of Alcoholism and Alcohol Abuse (NIAAA; R28 AA012725). This work was supported by the NIAAA (R28 AA012725) and the NHMRC (grant #605210). Abbreviations CC Corpus callosum CN Caudate nucleus DCX Doublecortin EGFR Epidermal growth factor receptor FFPE Formalin-fixed paraffin-embedded GCL Granule cell layer GFAP Glial fibrillary acidic protein Iba1 ionized calcium binding adaptor molecule 1 IHC Immunohistochemistry PCNA Proliferating cell nuclear antigen PML Polymorphic layer RMS Rostral migratory stream SFG Superior frontal gyrus SGZ Subgranular zone SVZ Subventricular zone Fig. 1 Immunostaining of endogenous markers of proliferation and neurogenesis in the human SVZ Representative photomicrographs showing differences in immunostaining for Ki67 (A, C, E and G) and DCX (B, D, F and H) with age. Immunostaining of the SVZ from a 0.3 year-old individual shows (A) numerous Ki67+ cells with a nuclear staining pattern and (B) clusters of DCX+ cells with a combination of somatic and dendritic staining. A similar pattern is shown in a 1.3 year-old individual for both Ki67 (C) and DCX (D). A 16 year-old individual shows (E) a single Ki67+ cell (black arrow) in the SVZ and (F) shows a single DCX+ cell (black arrow) with weak somatic staining. A 54 year-old individual shows (G) a single Ki67+ cell (black arrow) and (H) no DCX+ cells. 400x magnification. Scale bar = 50μm. Insets show digital enlargement of immunopositive cells. Fig. 2 Phenotype of proliferating cells in the juvenile SVZ Confocal micrographs of the SVZ from a 1 year-old individual shows Ki67+ cells co-localising with the immature neuronal marker beta III tubulin and the neuroblast marker DCX (A–D), the transit-amplifier marker EGFR (E–H), the oligodendrocytic marker Olig2 (I–L), the microglial marker Iba1 (M–P) but not the astrocytic marker GFAP (Q–T). 400x Magnification. Scale bar = 50μm. Insets show higher magnification of Ki67+ cells. Fig. 3 Phenotype of proliferating cells in the adult SVZ Confocal micrographs of the SVZ from a 54 year-old individual shows Ki67+ cells co-localising with the microglial marker Iba1 (A–C) but not the immature neuronal marker beta III tubulin (D–F) or the astrocytic marker GFAP (G–I). (J) A graph displaying the number of Ki67+ cells co-localising with different cell specific markers in the SVZ across a range of ages, note as quantification was performed on serial tissue sections totals could exceed or be less than 100%. (A–I) Magnification = 200x, Insets - 600x. Scale bar = 50 μm. Fig. 4 Immunostaining of endogenous markers of proliferation and neurogenesis in the human SGZ Representative photomicrographs showing differences in immunostaining for Ki67 and DCX with age with the polymorphic layer to the right side in all images. Immunostaining shows the SGZ of a 0.3 year-old individual with (A) numerous Ki67+ cells with a nuclear staining pattern and (B) abundant DCX+ cells with predominantly somatic staining. A 1.3 year-old shows (C) a single Ki67+ cell and (D) numerous DCX+ cells with predominantly cytoplasmic staining. A16 year-old individual shows no Ki67+ cells (E) and a single DCX+ cell with dendritic and somatic staining (arrow). A 54 year-old individual shows the same pattern for both Ki67 (G) and DCX (H) as the 16 year-old individual. 400x magnification. Scale bar = 50μm. Fig. 5 Phenotype of proliferating cells in the juvenile SGZ Confocal micrographs of the SGZ from a 1 year-old individual shows Ki67+ cells co-localising with the transit-amplifier marker EGFR (A–D), the oligodendrocytic marker Olig2 (E–H), but not the microglial marker Iba1 (I–L), the immature neuronal marker beta III tubulin (M–P) or the astrocytic marker GFAP (Q–T). 400x Magnification. Scale bar = 50 μm. Insets show digital enlargement of Ki67+ cells. Fig. 6 Phenotype of proliferating cells in the adult SGZ Confocal micrographs of the SGZ from a 54 year-old individual shows Ki67+ cells co-localising with the microglial marker Iba1 (A–C) but not the immature neuronal marker beta III tubulin (D–F) or the astrocytic marker GFAP (G–I). (J) A graph displaying the number of Ki67+ cells co-localising with different cell specific markers in the SGZ across a range of ages. N.B. as quantification was performed on serial tissue sections totals could exceed or be less than 100%. (A–I) Magnification = 200x. Insets at 600x magnification. Scale bar = 50 μm. Fig. 7 Characterisation of the RMS Photomicrographs show traces of the rostral migratory stream with distinctive ependymal islets extending rostrally from and confluent with the frontal horn of the lateral ventricle between the corpus callosum (CC) and head of the caudate nucleus (CN). Higher magnifications of regions corresponding to black rectangles are shown in insets. Collages of overlapping fields show DCX+ cells throughout the dorsomedial portion of the RMS in a (A) 0.2 year-old and (B) one year-old brain but only a single DCX+ cell (nuclear) in the (C) 54 year-old. (D – E) Ki67+ cells are seen in ventrolateral portion of the RMS in (D) 0.2 year-old but markedly fewer Ki67+ cells within the RMS of (E) a one year-old. (F) a confocal micrograph showing the disparity of Ki67 (red) and DCX (green) within the RMS of a 0.2 year-old individual. Orthogonal projections of z-stacks show (G) a rare Ki67+/DCX+ cells in the infant RMS whilst and (H) a Ki67+ cellsin the adult RMS co-positive for the microglia marker, Iba1. Table 1 Details of cases included in this study Case ID Classification Sex Age (Years) PMI (Hours) Cause of Death Brain pH Immunofluorescence analysis performed 1 Juvenile Male 0.2 44 Undetermined n/a No 2 Juvenile Female 0.3 52 Undetermined n/a No 3 Juvenile Male 1.0 18 Respiratory n/a Yes 4 Juvenile Female 1.0 47 Undetermined n/a No 5 Juvenile Male 1.0 46 Drowning n/a No 6 Juvenile Female 1.3 32 Infection n/a Yes 7 Juvenile Male 1.3 20 Respiratory n/a No 8 Juvenile Male 2.0 60 Drowning n/a Yes 9 Juvenile Male 2.5 90 Drowning n/a No 10 Juvenile Female 3.0 43 Drowning n/a Yes 11 Juvenile Female 3.9 35 Respiratory n/a Yes 12 Juvenile Female 7.0 41 Respiratory n/a No 13 Juvenile Male 7.0 15 Respiratory n/a No 14 Juvenile Male 14 18 Trauma n/a Yes 15 Juvenile Female 16 18 Drowning n/a No 16 Adult Male 24 43 Cardiac 6.57 No 17 Adult Female 29 40 Cardiac 6.83 No 18 Adult Male 37 15 Cardiac 6.46 No 19 Adult Male 43 66 Respiratory 6.57 No 20 Adult Male 47 38 Cardiac 6.2 No 21 Adult Female 49 15 Cardiac 6.97 No 22 Adult Male 54 28 Cardiac 6.38 Yes 23 Adult Male 59 29 Respiratory 6.61 No Statement of Conflict of Interest None of the authors have any conflict of interest in respect of the manuscript contents. 1 Falk A Frisen J New neurons in old brains Ann Med 2005 37 480 6 16278161 2 Taylor CJ Jhaveri DJ Bartlett PF The therapeutic potential of endogenous hippocampal stem cells for the treatment of neurological disorders Frontiers in cellular neuroscience 2013 7 5 23372544 3 Capilla-Gonzalez V Herranz-Perez V Garcia-Verdugo JM The aged brain: genesis and fate of residual progenitor cells in the subventricular zone Frontiers in cellular neuroscience 2015 9 365 26441536 4 Mouret A Murray K Lledo PM Centrifugal drive onto local inhibitory interneurons of the olfactory bulb Ann N Y Acad Sci 2009 1170 239 54 19686142 5 Nissant A Bardy C Katagiri H Murray K Lledo PM Adult neurogenesis promotes synaptic plasticity in the olfactory bulb Nat Neurosci 2009 12 728 30 19412168 6 van Praag H Schinder AF Christie BR Toni N Palmer TD Gage FH Functional neurogenesis in the adult hippocampus Nature 2002 415 1030 4 11875571 7 Akers KG Martinez-Canabal A Restivo L Yiu AP De Cristofaro A Hsiang HL Wheeler AL Guskjolen A Niibori Y Shoji H Ohira K Richards BA Miyakawa T Josselyn SA Frankland PW Hippocampal neurogenesis regulates forgetting during adulthood and infancy Science 2014 344 598 602 24812394 8 Curtis MA Kam M Nannmark U Anderson MF Axell MZ Wikkelso C Holtas S van Roon-Mom WM Bjork-Eriksson T Nordborg C Frisen J Dragunow M Faull RL Eriksson PS Human neuroblasts migrate to the olfactory bulb via a lateral ventricular extension Science 2007 315 1243 9 17303719 9 Couillard-Despres S Winner B Schaubeck S Aigner R Vroemen M Weidner N Bogdahn U Winkler J Kuhn HG Aigner L Doublecortin expression levels in adult brain reflect neurogenesis Eur J Neurosci 2005 21 1 14 15654838 10 Sanai N Nguyen T Ihrie RA Mirzadeh Z Tsai HH Wong M Gupta N Berger MS Huang E Garcia-Verdugo JM Rowitch DH Alvarez-Buylla A Corridors of migrating neurons in the human brain and their decline during infancy Nature 2011 9 28 478 7369 382 6 21964341 11 Knoth R Singec I Ditter M Pantazis G Capetian P Meyer RP Horvat V Volk B Kempermann G Murine features of neurogenesis in the human hippocampus across the lifespan from 0 to 100 years PLoS One 2010 5 e8809 20126454 12 Spalding KL Bergmann O Alkass K Bernard S Salehpour M Huttner HB Bostrom E Westerlund I Vial C Buchholz BA Possnert G Mash DC Druid H Frisen J Dynamics of hippocampal neurogenesis in adult humans Cell 2013 153 1219 27 23746839 13 Ernst A Alkass K Bernard S Salehpour M Perl S Tisdale J Possnert G Druid H Frisen J Neurogenesis in the striatum of the adult human brain Cell 2014 156 1072 83 24561062 14 Bergmann O Liebl J Bernard S Alkass K Yeung MS Steier P Kutschera W Johnson L Landen M Druid H Spalding KL Frisen J The age of olfactory bulb neurons in humans Neuron 2012 74 634 9 22632721 15 Lyck L Dalmau I Chemnitz J Finsen B Schroder HD Immunohistochemical markers for quantitative studies of neurons and glia in human neocortex J Histochem Cytochem 2008 56 201 21 17998570 16 Sutherland GT Sheahan PJ Matthews J Dennis CV Sheedy DS McCrossin T Curtis MA Kril JJ The effects of chronic alcoholism on cell proliferation in the human brain Exp Neurol 2013 247 9 18 23541433 17 Quinones-Hinojosa A Sanai N Soriano-Navarro M Gonzalez-Perez O Mirzadeh Z Gil-Perotin S Romero-Rodriguez R Berger MS Garcia-Verdugo JM Alvarez-Buylla A Cellular composition and cytoarchitecture of the adult human subventricular zone: a niche of neural stem cells J Comp Neurol 2006 494 415 34 16320258 18 Doetsch F Petreanu L Caille I Garcia-Verdugo JM Alvarez-Buylla A EGF converts transit-amplifying neurogenic precursors in the adult brain into multipotent stem cells Neuron 2002 36 1021 34 12495619 19 Menezes JR Luskin MB Expression of neuron-specific tubulin defines a novel population in the proliferative layers of the developing telencephalon J Neurosci 1994 14 5399 416 8083744 20 Pixley SK de Vellis J Transition between immature radial glia and mature astrocytes studied with a monoclonal antibody to vimentin Brain Res 1984 317 201 9 6383523 21 Middeldorp J Hol EM GFAP in health and disease Prog Neurobiol 2011 93 421 43 21219963 22 Ito D Imai Y Ohsawa K Nakajima K Fukuuchi Y Kohsaka S Microglia-specific localisation of a novel calcium binding protein, Iba1 Brain Res Mol Brain Res 1998 57 1 9 9630473 23 Dennis CV Sheahan PJ Graeber MB Sheedy DL Kril JJ Sutherland GT Microglial proliferation in the brain of chronic alcoholics with hepatic encephalopathy Metab Brain Dis 2014 29 1027 39 24346482 24 Marshall CA Novitch BG Goldman JE Olig2 directs astrocyte and oligodendrocyte formation in postnatal subventricular zone cells J Neurosci 2005 25 7289 98 16093378 25 Menn B Garcia-Verdugo JM Yaschine C Gonzalez-Perez O Rowitch D Alvarez-Buylla A Origin of oligodendrocytes in the subventricular zone of the adult brain J Neurosci 2006 26 7907 18 16870736 26 Wang C Liu F Liu YY Zhao CH You Y Wang L Zhang J Wei B Ma T Zhang Q Zhang Y Chen R Song H Yang Z Identification and characterization of neuroblasts in the subventricular zone and rostral migratory stream of the adult human brain Cell Res 2011 11 21 11 1534 50 21577236 27 Curtis MA Penney EB Pearson AG van Roon-Mom WM Butterworth NJ Dragunow M Connor B Faull RL Increased cell proliferation and neurogenesis in the adult human Huntington’s disease brain Proc Natl Acad Sci U S A 2003 100 9023 7 12853570 28 Curtis MA Penney EB Pearson J Dragunow M Connor B Faull RL The distribution of progenitor cells in the subependymal layer of the lateral ventricle in the normal and Huntington’s disease human brain Neuroscience 2005 132 777 88 15837138 29 van den Berge SA van Strien ME Korecka JA Dijkstra AA Sluijs JA Kooijman L Eggers R De Filippis L Vescovi AL Verhaagen J van de Berg WD Hol EM The proliferative capacity of the subventricular zone is maintained in the parkinsonian brain Brain 2011 134 3249 63 22075520 30 Funato H Yoshimura M Ito Y Okeda R Ihara Y Proliferating cell nuclear antigen (PCNA) expressed in human leptomeninges J Histochem Cytochem 1996 44 1261 5 8918901 31 Ino H Chiba T Expression of proliferating cell nuclear antigen (PCNA) in the adult and developing mouse nervous system Brain Res Mol Brain Res 2000 78 163 74 10891596 32 Stoimenov I Helleday T PCNA on the crossroad of cancer Biochem Soc Trans 2009 37 605 13 19442257 33 Zessin PJ Sporbert A Heilemann M PCNA appears in two populations of slow and fast diffusion with a constant ratio throughout S-phase in replicating mammalian cells Sci Rep 2016 6 18779 26758689 34 Verwer RW Sluiter AA Balesar RA Baayen JC Noske DP Dirven CM Wouda J van Dam AM Lucassen PJ Swaab DF Mature astrocytes in the adult human neocortex express the early neuronal marker doublecortin Brain 2007 130 3321 35 18055496 35 Klempin F Kronenberg G Cheung G Kettenmann H Kempermann G Properties of doublecortin-(DCX)-expressing cells in the piriform cortex compared to the neurogenic dentate gyrus of adult mice PLoS One 2011 6 e25760 22022443 36 Worlitzer MM Viel T Jacobs AH Schwamborn JC The majority of newly generated cells in the adult mouse substantia nigra express low levels of Doublecortin, but their proliferation is unaffected by 6-OHDA-induced nigral lesion or Minocycline-mediated inhibition of neuroinflammation Eur J Neurosci 2013 38 2684 92 23734736 37 Boekhoorn K Joels M Lucassen PJ Increased proliferation reflects glial and vascular-associated changes, but not neurogenesis in the presenile Alzheimer hippocampus Neurobiol Dis 2006 24 1 14 16814555 38 Ihrie RA Alvarez-Buylla A Lake-front property: a unique germinal niche by the lateral ventricles of the adult brain Neuron 2011 70 674 86 21609824 39 Kohler SJ Williams NI Stanton GB Cameron JL Greenough WT Maturation time of new granule cells in the dentate gyrus of adult macaque monkeys exceeds six months Proc Natl Acad Sci U S A 2011 108 10326 31 21646517 40 Galvez-Contreras AY Quinones-Hinojosa A Gonzalez-Perez O The role of EGFR and ErbB family related proteins in the oligodendrocyte specification in germinal niches of the adult mammalian brain Frontiers in cellular neuroscience 2013 7 258 24381541 41 Roelofs RF Fischer DF Houtman SH Sluijs JA Van Haren W Van Leeuwen FW Hol EM Adult human subventricular, subgranular, and subpial zones contain astrocytes with a specialized intermediate filament cytoskeleton Glia 2005 52 289 300 16001427 42 van den Berge SA Middeldorp J Zhang CE Curtis MA Leonard BW Mastroeni D Voorn P van de Berg WD Huitinga I Hol EM Longterm quiescent cells in the aged human subventricular neurogenic system specifically express GFAP-delta Aging Cell 2010 9 313 26 20121722 43 Sohn J Orosco L Guo F Chung SH Bannerman P Mills Ko E Zarbalis K Deng W Pleasure D The subventricular zone continues to generate corpus callosum and rostral migratory stream astroglia in normal adult mice J Neurosci 2015 35 3756 63 25740506 44 Malik SZ Lewis M Isaacs A Haskins M Van Winkle T Vite CH Watson DJ Identification of the rostral migratory stream in the canine and feline brain PLoS One 2012 7 e36016 22606243 45 Alonso M Ortega-Perez I Grubb MS Bourgeois JP Charneau P Lledo PM Turning astrocytes from the rostral migratory stream into neurons: a role for the olfactory sensory organ J Neurosci 2008 28 11089 102 18945916 46 Gritti A Bonfanti L Doetsch F Caille I Alvarez-Buylla A Lim DA Galli R Verdugo JM Herrera DG Vescovi AL Multipotent neural stem cells reside into the rostral extension and olfactory bulb of adult rodents J Neurosci 2002 22 437 45 11784788 47 Boldrini M Santiago AN Hen R Dwork AJ Rosoklija GB Tamir H Arango V John Mann J Hippocampal granule neuron number and dentate gyrus volume in antidepressant-treated and untreated major depression Neuropsychopharmacology 2013 38 1068 77 23303074 48 Brown J Cooper-Kuhn CM Kempermann G Van Praag H Winkler J Gage FH Kuhn HG Enriched environment and physical activity stimulate hippocampal but not olfactory bulb neurogenesis Eur J Neurosci 2003 17 2042 6 12786970 49 Cameron HA McKay RD Adult neurogenesis produces a large pool of new granule cells in the dentate gyrus J Comp Neurol 2001 435 406 17 11406822 50 Kempermann G Kuhn HG Gage FH Genetic influence on neurogenesis in the dentate gyrus of adult mice Proc Natl Acad Sci U S A 1997 94 10409 14 9294224 51 Jabes A Lavenex PB Amaral DG Lavenex P Quantitative analysis of postnatal neurogenesis and neuron number in the macaque monkey dentate gyrus Eur J Neurosci 2010 31 273 85 20074220 52 Kornack DR Rakic P Continuation of neurogenesis in the hippocampus of the adult macaque monkey Proc Natl Acad Sci U S A 1999 96 5768 73 10318959 53 Lazic SE Modeling hippocampal neurogenesis across the lifespan in seven species Neurobiol Aging 2012 33 1664 71 21621300 54 Mak GK Enwere EK Gregg C Pakarainen T Poutanen M Huhtaniemi I Weiss S Male pheromone-stimulated neurogenesis in the adult female brain: possible role in mating behavior Nat Neurosci 2007 10 1003 11 17603480 55 Low VF Faull RL Bennet L Gunn AJ Curtis MA Neurogenesis and progenitor cell distribution in the subgranular zone and subventricular zone of the adult sheep brain Neuroscience 2013 244 173 87 23587842 56 Kornack DR Rakic P Cell proliferation without neurogenesis in adult primate neocortex Science 2001 294 2127 30 11739948 57 Hattori S Chen L Weiss C Disterhoft JF Robust hippocampal responsivity during retrieval of consolidated associative memory Hippocampus 2015 25 655 69 25515308 58 Bhardwaj RD Curtis MA Spalding KL Buchholz BA Fink D Bjork-Eriksson T Nordborg C Gage FH Druid H Eriksson PS Frisen J Neocortical neurogenesis in humans is restricted to development Proc Natl Acad Sci U S A 2006 103 12564 8 16901981 59 Geha S Pallud J Junier MP Devaux B Leonard N Chassoux F Chneiweiss H Daumas-Duport C Varlet P NG2+/Olig2+ cells are the major cycle-related cell population of the adult human normal brain Brain Pathol 2010 20 399 411 19486010 60 Doorn KJ Drukarch B van Dam AM Lucassen PJ Hippocampal proliferation is increased in presymptomatic Parkinson’s disease and due to microglia Neural plasticity 2014 2014 959154 25197578 61 Liu YW Curtis MA Gibbons HM Mee EW Bergin PS Teoh HH Connor B Dragunow M Faull RL Doublecortin expression in the normal and epileptic adult human brain Eur J Neurosci 2008 28 2254 65 19046368
PMC005xxxxxx/PMC5125851.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 9708549 20727 Parasitol Int Parasitol. Int. Parasitology international 1383-5769 1873-0329 27510768 5125851 10.1016/j.parint.2016.08.002 NIHMS817649 Article Real-time PCR detection and phylogenetic relationships of Neorickettsia spp. in digeneans from Egypt, Philippines, Thailand, Vietnam and the United States Greiman Stephen E. 12 Vaughan Jefferson A. 2 Elmahy Rasha 3 Adisakwattana Poom 4 Van Ha Nguyen 5 Fayton Thomas J. 6 Khalil Amal I. 3 Tkach Vasyl V. 2* 1 Department of Biology, Georgia Southern University, Statesboro, Georgia 30460, U.S.A 2 Department of Biology, University of North Dakota, Grand Forks, North Dakota 58202, U.S.A 3 Department of Zoology, Tanta University, 31527 Tanta, Egypt 4 Department of Helminthology, Mahidol University, Bangkok 10400, Thailand 5 Institute of Ecology and Biological Resources, VAST, Nghiado, Caugiay, Hanoi, Vietnam 6 Department of Coastal Sciences, University of Southern Mississippi, Ocean Springs, Mississippi, U.S.A * Address correspondence to Vasyl V. Tkach, [email protected] 28 9 2016 7 8 2016 2 2017 01 2 2018 66 1 10031007 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Neorickettsia (Rickettsiales, Anaplasmataceae) is a genus of obligate intracellular bacterial endosymbionts of digeneans (Platyhelminthes, Digenea). Some Neorickettsia are able to invade cells of the digenean's vertebrate host and are known to cause diseases of domestic animals, wildlife, and humans. In this study we report the results of screening digenean samples for Neorickettsia collected from bats in Egypt and Mindoro Island, Philippines, snails and fishes from Thailand, and fishes from Vietnam and the USA. Neorickettsiae were detected using a real-time PCR protocol targeting a 152 bp fragment of the heat shock protein coding gene, GroEL, and verified with nested PCR and sequencing of a 1853 bp long region of the GroESL operon and a 1371 bp long region of 16S rRNA. Eight unique genotypes of Neorickettsia were obtained from digenean samples. Neorickettsia sp. 8 obtained from Lecithodendrium sp. from Egypt; Neorickettsia sp. 9 and 10 obtained from two species of Paralecithodendrium from Mindoro, Philippines; Neorickettsia sp. 11 from Lecithodendrium sp. and Neorickettsia sp. 4 (previously identified from Saccocoelioides lizae, from China) from Thailand; Neorickettsia sp. 12 from Dicrogaster sp. Florida, USA; Neorickettsia sp. 13 and SF agent from Vietnam. Sequence comparison and phylogenetic analysis demonstrated that the forms, provisionally named Neorickettsia sp. 8-13, represent new genotypes. We have for the first time detected Neorickettsia in a digenean from Egypt (and the African continent as a whole), the Philippines, Thailand and Vietnam based on PCR and sequencing evidence. Our findings suggest that further surveys from the African continent, SE Asia, and the Island countries are likely to reveal new Neorickettsia lineages as well as new digenean host associations. Graphical Abstract Neorickettsia Southeast Asia Africa North America real-time PCR molecular phylogeny Digenea 1. Introduction Obligate intracellular bacteria in the genus Neorickettsia (Rickettsiales, Anaplasmataceae) are endosymbionts of digeneans. Neorickettsia are vertically transmitted through all stages of complex digenean life cycles. Additionally, they are capable of being horizontally transmitted to the vertebrate hosts of the digenean, both human and wildlife, where they can cause disease [1]. These diseases are potentially debilitating, e.g., Sennetsu fever in humans (Neorickettsia sennetsu), or even fatal, e.g., salmon dog poisoning (Neorickettsia helminthoeca) and Potomac horse fever (Neorickettsia risticii)[1]. Greiman et al. [2] conducted the first large scale molecular screening of digeneans for Neorickettsia from multiple countries and continents. Their screening revealed 7 new genotypes of Neorickettsia, bringing the known number of species/genotypes of the bacteria to 20. Additionally, they detected Neorickettsia from digeneans for the first time from the Australian continent and China. Records resulting from PCR detection of neorickettsial DNA are known only from North and South America, eastern Asia, Antarctica, and Australia [2], [3] and [4]. Africa and Europe are the last continents where Neorickettsia has yet to be found. In this study, we screened DNA extracts of lecithodendriid digeneans from bats collected in Egypt and the Philippines, and other digenean samples collected from fishes and snails in the United States, Thailand, and Vietnam. As a result we found 6 new genotypes of Neorickettsia and two previously identified genotypes of Neorickettsia. We have for the first time detected Neorickettsia in digeneans from the African continent, the Philippines, Vietnam, and Thailand using PCR-based detection and DNA sequencing. Also, we have conducted molecular phylogenetic analyses in order to reveal the interrelationships among the newly discovered genotypes with previously known species/genotypes of Neorickettsia. 2. Materials and methods 2.1 Sample collections Digenean samples were collected from bats, fishes and snails from the United States (Salt Springs, Florida, 29°21′01″N, 81°43′57″ W), Vietnam (Cat-Ba Island) 20°43′27.86″N, 107°2′58.61″E), Egypt (30°1′39″N, 31°12′37″E), Philippines (Mindoro Island, 12°47′14.99″N, 120°54′57.96″E), and Thailand (Aranyaprathet, Sa Kae, 13°36′37″N, 102°34′14″E and Talay Thai Seafood market, Sumut Sakhon (13°32′56.29″N, 100°15′26.11″E). Fishes from Vietnam and Thailand were purchased from local and commercial fish markets. 2.2 Sample processing Live adult worms were rinsed in saline, examined briefly, killed with hot water, and fixed in 80% ethanol that allowed for both morphological examination and molecular study. For morphological identifications, fixed worms were stained in aqueous alum carmine or Mayer's hematoxylin; dehydrated in a graded ethanol series; cleared in clove oil (carmine) or methyl salicylate (Mayer's); and mounted permanently in Damar Gum. Genomic DNA was extracted one of two ways. Samples collected from Florida were extracted using the Qiagen DNAeasy tissue kit (Qiagen, Inc., Valencia, California) following the manufacturer's instructions. Samples collected from Egypt, Thailand, Vietnam, and Philippines were first homogenized by direct sonication using a UP100H compact ultrasonic processor (Hielscher USA, Inc., Ringwood, New Jersey) at an amplitude of 90% for 20 seconds and further DNA extracted using the guanidine thiocyanate method according to Tkach and Pawlowski [5]. 2.3 Molecular screening DNA extracts were first tested for the presence of Neorickettsia using a real-time PCR protocol as described by Greiman et al. [2]. Five microliters of each DNA extract were used. The real-time PCR amplified a 152-bp portion of the 3' end of the heat shock protein coding gene, GroEL. Samples that tested positive with real-time PCR were verified using a substantially modified nested PCR protocol as described by Greiman et al. [6]. Five microliters of each DNA extract were used for the first PCR reaction and 1 μl of the first PCR product was used for the nested PCR. A 1470-bp long fragment of the 16S rRNA gene was first amplified, followed by the nested PCR step which amplified a 1371-bp fragment using internal primers. The same nested PCR primers were used in sequencing reactions along with internal forward and internal reverse primers [6]. DNA of Neorickettsia sp. from the digenean Plagiorchis elegans, obtained from a laboratory life cycle, was used as a positive control [7]. Pure water was used for negative controls in both real-time and nested PCRs. A nested PCR amplifying a 1940-bp fragment of the GroESL operon was carried out for members of the “Neorickettsia risticii” clade (Fig. 1, subclade B). Five microliters of each DNA extract were used for the first PCR reaction and 2 μl of the first PCR product was used for the nested PCR. The primers used for the initial PCR were; hs10F 5′-CTCAAATGAAACAAT-CCGTTTGTTTGTAGC-3′ and hs2090R 5′-CATTCCACCCATGCCA-CCACCAGGCAT-3′. The primers used for the nested PCR were hs90F 5′-GTAGGTCTTGAAAAATATCACAGCG-3′ and hs2030R 5′-GTAGTCACTAGAACACTAGCAACAGA-3′. The same nested PCR primers were used for sequencing along with internal forward primers; hs120F 5′- TACGATATTTGATTCTGTAGGTCATTAG -3′and hs910F 5′- TGGTTCAATTTCTGCTAACGGCAAT-3′ and internal reverse primers; hs700R 5′ GCTTTTTCATTCGCCTGTGAGGTAGCCT-3′ and hs1620R 5′- CTTTAACCTCAACTTCTGTAGCACCAC-3′ designed by SEG. 2.4 Digenean host identification The digenean hosts were identified based on morphology using stained whole mounts and using partial sequences of the nuclear large ribosomal subunit gene (28S). Digenean DNA was amplified from the extracted DNA by the PCR protocol described by Greiman et al. [2]. 2.5 DNA sequencing PCR amplicons of both Neorickettsia and digeneans were purified using the Zymo DNA Clean & Concentrator™-5 (Zymo Research, Irvine, CA) or ExoSap PCR clean-up enzymatic kit from Affimetrix (Santa Clara, CA) according to the manufacturer's instructions. The PCR products were cycle-sequenced using ABI BigDye™ chemistry, ethanol precipitated, and run on an ABI Prism 3100™ automated capillary sequencer. Contiguous sequences of Neorickettsia were assembled using Sequencher™ ver. 4.2 (GeneCodes Corp., Ann Arbor, MI) and submitted to GenBank under accession numbers XXXX. 2.6 Phylogenetic analysis Newly obtained sequences of neorickettsiae from GenBank were used in the phylogenetic analyses (Figs. 1, 2). The sequences were initially aligned with the aid of ClustalW as implemented in the BioEdit program, version 7.0.1 [8]. Two distinct alignments were prepared and two analyses were done. The alignments were manually refined in MacClade, version 4 [9]. The first phylogenetic analysis was run using a larger dataset including all new and previously available 16S sequences of Neorickettsia (a total of 30 sequences), one sequence of closely related Anaplasmataceae sp., and Wolbachia pipientis as the outgroup (Fig. 1). The analysis was carried out using Bayesian inference as implemented in the MrBayes program, version 2.01 [10] with the following nucleotide substitution parameters: lset nst=6, rates=invgamma, prset (prior assumptions) pinvarpr=fixed (0.6170) shapepr=fixed (0.4360), that correspond to a “three-parameter model (TPM3) [11] model including estimates of the proportion of invariant sites (I) and rate variation among sites with a number of rate categories (G). Posterior probabilities were approximated over 1,500,000 generations, log-likelihood scores plotted, and only the final 75% of trees were used to produce the consensus trees by setting the “burnin” parameters at 375,000 generations. The TPM3+I+G model was used for the analysis based off of the results obtained from jModelTest, version 0.1.1 [12], [13]. The second phylogenetic analysis was run using a much smaller dataset based on the partial sequences of the GroESL operon, including only species/genotypes of Neorickettsia closely related to Neorickettsia risticii (Fig. 2), a total of 13 sequences. Neorickettsia helminthoeca was used as an outgroup based on the outcome of the 16S analysis (Fig. 1). The analysis was carried out using Bayesian inference (BI) as implemented in the MrBayes program, version 2.01 [10] with the following nucleotide substitution parameters: lset nst=6, rates=propinv, prset (prior assumptions) pinvarpr=fixed (0.6380), that correspond to a Tamura-Nei (TrN) model including estimates of the proportion of invariant sites (I). Posterior probabilities were approximated over 1,000,000 generations, log-likelihood scores plotted, and only the final 75% of trees were used to produce the consensus trees by setting the “burnin” parameters at 250,000 generations. The TrN+I model was used for the analysis based off of the results obtained from jModelTest, version 0.1.1 [12], [13]. 3. Results In northeastern Egypt, during the fall of 2012, authors RE and AIK collected a total of 50 bats (Pipistrellus kuhli marginatus) infected with adult digeneans. In northern Vietnam, during the fall of 2013, authors SEG and NVH dissected a total of 75 fishes, representing at least 6 fish species. From those dissections, a total of 30 fishes were infected with either larval or adult digeneans. In southeastern Thailand, during the winter of 2014, authors SEG, JAV, PA, and VVT dissected a total of 114 fishes. From those dissections, 50 fishes were infected with either larval or adult digeneans. In spring of 2013 (March), TJF obtained 8 Mugil cephalus via Hawaiian sling from Salt Springs, Florida. Two-thirds of these hosts harbored coinfections of F. apiensis and Saccocoelioides sp. and one individual harbored a single mature specimen of Intromugil alachuaensis. Digenean samples were screened for Neorickettsia. Screening revealed eight unique genotypes of Neorickettsia from digenean samples: Neorickettsia sp. 8 obtained from Lecithodendrium sp. from Egypt; Neorickettsia sp. 9 and 10 obtained from two species of Paralecithodendrium from Mindoro, Philippines; Neorickettsia sp. 11 from Lecithodendrium sp. and Neorickettsia sp. 4 (previously identified from Saccocoelioides lizae, from China) from Thailand; Neorickettsia sp. 12 from Florida, USA; Neorickettsia sp. 13 and SF agent from Vietnam (Table 1). Genotype numbering is the continuation of that proposed by Greiman et al. [2]. The 16S alignment included a total of 1,258 sites, of which 1,249 could be aligned unambiguously. Positions of ambiguous homology were excluded from the analysis. Bayesian analysis produced a tree where all Neorickettsia sequences clearly fall into a well-defined clade, with 100% support (Fig. 1), composed of three subclades – subclade I (1 genotype), subclade II (7 genotypes), and subclade III (22 genotypes). Subclade IIIA included N. helminthoeca, agent of salmon dog poisoning and subclade IIIB included N. risticii, agent of Potomac horse fever and N. sennetsu, agent of Sennetsu fever in humans The topology of the consensus tree generated by this analysis is essentially identical to the phylogeny published by Greiman et al. [2], with the only exception being the inclusion of the new genotypes of Neorickettsia discovered in this study. The GroESL alignment included a total of 1,822 sites, of which all could be aligned unambiguously. Bayesian analysis produced a well-structured consensus tree with strong support of nearly all topologies (Fig. 2). Based on this analysis, Neorickettsia sp. 8 from Egypt and Neorickettsia spp. 9 and 10 from the Philippines appeared in a cluster that formed a sister clade to all other genotypes of Neorickettsia (N. sennetsu, N. risticii, Neorickettsia sp. from Plagiorchis elegans, Neorickettsia sp. from Fasciola hepatica, and SF agent). 4. Discussion Greiman et al. [2] emphasized that the known distribution of Neorickettsia is very geographically uneven, with no well-documented records supported by PCR/sequence data from Africa, Europe, most of South America, most of Asia, and nearly all island countries. In this study we have further expanded taxonomic and geographic coverage of screening of digeneans for Neorickettsia. As suggested by Greiman et al. [2] it resulted in discoveries of novel forms of Neorickettsia from additional digenean taxa and geographic regions including the first record from Africa. Digeneans in the family Lecithodendriidae parasitic in bats appear to be common hosts of Neorickettsia, with a total of 7 species of lecithodendriids known to harbor the bacteria including 4 added by the present study [1], [2]. This is a relatively large digenean family comprising at least 12 genera parasitizing mammals (mostly bats) and occasionally birds, on all continents [14]. Currently, lecithodendriids have been found to harbor the bacterial endosymbiont in Egypt, Thailand, North America (USA), South America (Argentina), and the Philippines, with those in the Americas known to harbor N. risticii, the etiological agent of Potomac horse fever. Mugil cephalus (grey mullet) is the 2nd intermediate or definitive host of four digenean species harboring Neorickettsia in this study (Saccocoelioides lizae, Saccocoelioides sp., Forticulcita apiensis, and Stellantchasmis falcatus (host of SF agent)) and two digenean species (Saccocoelioides beauforti and S. lizae) harboring Neorickettsia in the study by Greiman et al. [2]. Although this is a globally distributed marine fish, its digenean species appear to be much more restricted in their distribution. Saccocoelioides lizae has only been recorded off the cost of southeastern China [15]; however, we have now found it off the coast of Thailand. The distribution of Saccocoelioides sp. from northern Vietnam is unknown, however, the sequence of Neorickettsia sp. 13 differs from Neorickettsia sp. 4 identified from S. lizae from Thailand (this study) and China [2] by only 3 bp, likely indicating that its distribution is limited to the marine habitats of the region. Based on the high sequence similarity, Neorickettsia sp. 13 from Vietnam is likely the same species as Neorickettsia identified from China and Thailand. The last Neorickettsia hosting digenean species from mullet, F. apiensis was recently described from Salt Springs, Florida [16]. Salt Springs is an artesian spring that discharges slightly saline water in Lake George. This haploporid is likely endemic to the region, and therefore the genotype/species of Neorickettsia (sp. 12) is likely also unique to this area. The least resolved part of the 16S tree is the polytomy in the sub-clade B of clade III that includes several genotypes occupying a derived position in relation to N. sennetsu (Fig. 1; also in [2]). The GroESL phylogenetic analysis helped resolve the relationships among a majority of these taxa and has produced a tree with strong branch support for all topologies (Fig. 2). The GroESL operon was used for the second phylogenetic analysis following its use by Dumler et al. [17] for the members of the family Anaplasmataceae. The new genotypes (Neorickettsia spp. 8, 9, and 10) (from Egypt and the Philippines, respectively) identified in our study, fall into a clade that is sister to the well-supported clade including N. risticii, N. sennetsu, Neorickettsia (P. elegans agent), Neorickettsia sp. (from F. hepatica) and SF agent. This brings up the potential question regarding the pathogenicity of the new genotypes of Neorickettsia from Egypt and the Philippines. Neorickettsia risticii, N. sennetsu, and SF agent are all known to cause potentially debilitating or even fatal diseases in the vertebrate hosts of digeneans [1]. Neorickettsia spp. 8, 9, and 10 from lecithodendriids from Egypt and the Philippines, are closely related to these three species of Neorickettsia and therefore have the potential of causing yet unknown disease in wildlife or domestic animals. Further studies of these three genotypes may reveal their potential as pathogens. The GroESL phylogeny also indicates that the genotype of Neorickettsia from Deropegus aspina, originally called Neorickettsia sp. by Greiman et al. [2], is in fact N. risticii. Additionally, Greiman et al. [6] identified the species of Neorickettsia from P. elegans as N. risticii. However, our GroESL phylogeny (Fig. 2) corroborates the suggestion by Greiman et al. [2] regarding the status of this genotype as a separate species related to N. risticii. Lastly, Mitreva et al. (unpublished; GenBank LNGI01000001) sequenced the complete genome of a species of Neorickettsia from Fasciola hepatica. Based on our GroESL phylogeny the species falls into the same clade as our species from P. elegans. However, due to the high 16S sequence similarity (99.6%) between these two forms, it remains unknown whether they represent two distinct species or different strains of one species. The GroESL phylogenetic tree shows close associations among Neorickettsia and their digenean hosts' 2nd intermediate or definitive hosts (Fig. 2). In the sub-clade including N. sennetsu and N. risticii, the 2nd intermediate host of all digenean hosts (at least where the digenean host is known) is an arthropod, and the definitive hosts for all but one are mammals (bats) and potentially birds. The digenean host of N. sennetsu, the causative agent of the human disease Sennetsu fever in Southeast Asia, is currently unknown. Greiman et al. [2] hypothesized a mammal or fish as a definitive host. The topology of the GroESL phylogenetic tree obtained in this study strongly suggests that it also should be a digenean that uses an arthropod as a second intermediate host. This is in contrast with the previous notion that the digenan host of N. sennetsu utilizes a fish second intermediate host. This previous hypothesis was based on the discovery that people became ill with N. sennetsu after consuming raw mullet infected with digenean metacercariae in Japan [18]. However, based on our GroESL phylogeny, and the transmission biology of the closely related species N. risticii, we hypothesize that the tissue of the fish may become infected with the bacteria in a similar manner to how horses become infected with N. risticii. Horses are dead-end hosts to the digeneans infected with N. risticii, and therefore, the mullet may also be dead end hosts to the digenean harboring N. sennetsu, and become infected when the fish ingests an arthropod infected with the metacercariae of the digenean harboring the bacteria. Further studies of trematodes from different host taxa are needed in the regions endemic for Sennetsu in order to answer this intriguing question of public health importance. Acknowledgements This project was funded by the grant R15AI092622 from the National Institutes of Health, USA to Vasyl V. Tkach and Jefferson A. Vaughan. We are grateful for the Florida Fish and Wildlife Conservation Commission for issuing the FNW-13-05 (renewal) that allowed for the collection of freshwater fish in Florida. The authors are also grateful to Dr. Richard Heard for the provision of microscopy and sampling equipment. Lastly, SEG was partially funded through a National Science foundation Postdoctoral Research Fellowship in Biology (1523410). Fig 1 Phylogenetic interrelationships among 30 genetic lineages of Neorickettsia based on Bayesian analysis of partial 16S rRNA sequences. Numbers above internodes indicate posterior probabilities. An asterisk () at the end of a taxon name indicates a new sequence obtained in this study. GenBank numbers are given here for all taxa. Fig. 2 Cladogram based on Bayesian analysis of partial GroESL sequences depicting phylogenetic interrelationships among 14 genetic lineages of N. risticii and N. risticii-like neorickettsiae. Posterior probabilities greater than 80% are shown. Digenean second intermediate host groups and vertebrate definitive host groups are indicated by symbols. An asterisk () at the end of a taxon name indicates a new sequence obtained in this study. GenBank numbers are given here for all taxa. Table 1 Neorickettsia genotypes found in this study, their digenean host associations and corresponding digenean life cycle information. Neorickettsia species Digenean family Digenean genus and species Digenean life cycle stage Vertebrate host (this study) Definitive host 1st/2nd intermediate hosts Origin Neorickettsia sp. 8 (XXXX) Lecithodendriidae Lecithodendrium sp. Adult ________ Bats Aquatic snail / aquatic arthropod Egypt Neorickettsia sp. 9 (XXXX) Lecithodendriidae Paralecithodendrium sp. Adult Horsfield's bat (Myotis horsfieldii) Bats Aquatic snail / aquatic arthropod Philippines Neorickettsia sp. 10 (XXXX) Lecithodendriidae Paralecithodendrium sp. Adult Large-eared horseshoe bat (Rhinolophus philippinensis) Bats Aquatic snail / aquatic arthropod Philippines Neorickettsia sp. 11 (XXXX) Lecithodendriidae Lecithodendrium sp. Cercariae ________ Bats Aquatic snail / aquatic arthropod Thailand Neorickettsia sp. 4 (XXXX) Haploporidae Saccocoelioides lizae Adult Striped mullet (Mugil cephalus) Fishes Unknown Thailand Neorickettsia sp. 12 (XXXX) Haploporidae Forticulcita apiensis Adult Striped mullet (Mugil cephalus) Fishes Unknown USA Neorickettsia sp. 13 (XXXX) Haploporidae Saccocoelioides sp. Adult Striped mullet (Mugil cephalus) Fishes Unknown Vietnam Neorickettsia sp. SF agent (XXXX) Heterophyidae Stellantchasmus falcatus metacercariae Striped mullet (Mugil cephalus) Mammals Aquatic snail/fish Vietnam Highlights Eight unique isolates of Neorickettsia discovered in SE Asia, Africa and North America. Seven isolates represent new species level genetic lineages. This is the first report of Neorickettsia from digeneans in Africa and three SE Asian countries. Phylogenetic relationships Neorickettsia has been studied 16S and GroESL genes. Host associations of digeneans harboring Neorickettsia are discussed. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. References 1 Vaughan JA Tkach VV Greiman SE Neorickettsial endosymbionts of the Digenea: diversity, transmission and distribution Adv. Parasitol 2012 79 253 297 22726644 2 Greiman SE Tkach VV Pulis E Fayton TJ Curran SS Large scale screening of Digeneans for Neorickettsia endosymbionts using real-time PCR reveals new Neorickettsia genotypes, host associations and geographic records PLoS ONE 2014 9 e98453 24911315 3 Ward NL Steven B Penn K Methe BA Detrich WH Characterization of the intestinal microbiota of two Antarctic notothenioid fish species Extremophiles 2009 13 679 685 19472032 4 Tkach VV Schroeder JA Greiman SE Vaughan JA New genetic lineages, host associations and circulation pathways of Neorickettsia endosymbionts of digeneans. Acta Parasitol 2012 57 285 292 22875676 5 Tkach VV Pawlowski J A new method of DNA extraction from the ethanol-fixed parasitic worms Acta Parasitol 1999 44 147 148 6 Greiman SE Tkach VV Vaughan JA Transmission rates of the bacterial endosymbiont, Neorickettsia risticii, during the asexual phase of its digenean host, Plagiorchis elegans, within naturally infected lymnaeid snails Parasit. Vectors 2013 6 303 309 24383453 7 Greiman SE Tkach M Vaughan JA Tkach VV Laboratory maintenance of the bacterial endosymbiont, Neorickettsia sp., through the life cycle of a digenean, Plagiorchis elegans Exp. Parasitol 2015 157 78 83 26160679 8 Hall TA BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT Nucleic Acids Symp. Ser 1999 41 95 98 9 Maddison DR Maddison WP MacClade 4: Analysis of phylogeny and character evolution. Version 4.08a 2005 Available: http://macclade.org 10 Huelsenbeck JP Ronquist F MRBAYES: Bayesian inference of phylogeny Bioinformatics 2001 17 754 755 11524383 11 Kimura M Estimation of evolutionary distances between homologous nucleotide sequences. Proc. Natl. Acad. Sci. U.S.A 1981 78 454 458 6165991 12 Guindon S Gascuel O A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood Syst. Biol 2003 52 696 704 14530136 13 Posada D jModelTest: Phylogenetic Model Averaging Mol. Biol. Evol 2008 25 1253 1256 18397919 14 Lotz JM Font WF Bray RA Gibson DI Jones A Family Lecithodendriidae Luhe, 1901. Keys to the Trematoda 2008 3 527 536 CABI Publishing Wallingford U.K. 15 Liu SF One new species of Saccocoelioides (Digenea: Haploporidae) from Liza carinatus in Taiwan Strait J. Oceanogr 2002 21 37 44 16 Andres MJ Curran SS Fayton TJ Pulis EE Overtsreet RM An additional genus and two additional species of Forticulcitinae (Digenea: Haploporidae) Folia Parasitol 2015 62 17 Dumler JS Barbet AF Bekker CPJ Dasch GA Palmer GH Ray SC Rikihisa Y Rurangirwa FR Reorganization of genera in the families Rickettsiaceae and Anaplasmataceae in the order Rickettsiales: unification of some species of Ehrlichia with Anaplasma, Cowdria with Ehrlichia and Ehrlichia with Neorickettsia, descriptions of six new species combinations and designation of Ehrlichia equi and ‘HE agent’ as subjective synonyms of Ehrlichia phagocytophila Int. J. Syst. Evol. Microbiol 2001 51 2145 2165 11760958 18 Fukuda T Rickettsial mononucleosis (synonyms: Hyuganetsu, kagaminetsu) Kitasato Arch. Exp. Med 1958 31 51 56
PMC005xxxxxx/PMC5125855.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 0372762 3389 Dev Biol Dev. Biol. Developmental biology 0012-1606 1095-564X 27634568 5125855 10.1016/j.ydbio.2016.09.008 NIHMS817847 Article The Hyaloid Vasculature Facilitates Basement Membrane Breakdown During Choroid Fissure Closure in the Zebrafish Eye James Andrea ab Lee Chanjae a Williams Andre M. a Angileri Krista ab Lathrop Kira L. bd Gross Jeffrey M. abc* a Department of Molecular Biosciences, Institute for Cellular and Molecular Biology, The University of Texas at Austin, Austin TX, 78712 b Department of Ophthalmology, Louis J. Fox Center for Vision Restoration, The University of Pittsburgh School of Medicine, Pittsburgh, PA 15213 c Department of Developmental Biology, The University of Pittsburgh School of Medicine, Pittsburgh, PA 15213 d Department of Bioengineering, University of Pittsburgh Swanson School of Engineering, Pittsburgh, PA 15213 * Correspondence: Jeffrey Gross, 3501 Fifth Ave, BST3 Rm 2051, The University of Pittsburgh School of Medicine, Pittsburgh, PA 15261, 412-383-7325. [email protected] 23 9 2016 12 9 2016 15 11 2016 15 11 2017 419 2 262272 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. A critical aspect of vertebrate eye development is closure of the choroid fissure (CF). Defects in CF closure result in colobomas, which are a significant cause of childhood blindness worldwide. Despite the growing number of mutated loci associated with colobomas, we have a limited understanding of the cell biological underpinnings of CF closure. Here, we utilize the zebrafish embryo to identify key phases of CF closure and regulators of the process. Utilizing Laminin-111 as a marker for the basement membrane (BM) lining the CF, we determine the spatial and temporal patterns of BM breakdown in the CF, a prerequisite for CF closure. Similarly, utilizing a combination of in vivo time-lapse imaging, β-catenin immunohistochemistry and F-actin staining, we determine that tissue fusion, which serves to close the fissure, follows BM breakdown closely. Periocular mesenchyme (POM)-derived endothelial cells, which migrate through the CF to give rise to the hyaloid vasculature, possess distinct actin foci that correlate with regions of BM breakdown. Disruption of talin1, which encodes a regulator of the actin cytoskeleton, results in colobomas and these correlate with structural defects in the hyaloid vasculature and defects in BM breakdown. cloche mutants, which entirely lack a hyaloid vasculature, also possess defects in BM breakdown in the CF. Taken together, these data support a model in which the hyaloid vasculature and/or the POM-derived endothelial cells that give rise to the hyaloid vasculature contribute to BM breakdown during CF closure. Introduction In all vertebrates, eye development begins with the evagination of optic primordia from the diencephalon. The optic primordia subsequently adopt vesicle-like structures, and as the lateral edges of the vesicles begin to fuse, each invaginates to form a bilayered cup. The outer layer of the cup gives rise to the retinal pigment epithelium (RPE) and the inner layer gives rise to the retina. Fusion between the prospective RPE and retina occurs at the choroid fissure (CF), a distinct region of the ventral optic cup. Prior to fusion, the hyaloid vasculature enters the eye through the CF and the retinal ganglion cell axons exit through it. Once these developmental events have occurred, CF closure (CFC) is critical for containment of the retina and RPE within the optic cup. Defects in ventral optic cup formation or CFC result in colobomas (Graw, 2003; Gregory-Evans et al., 2004; Fitzpatrick and van Heyningen, 2005; Chang et al., 2006). The incidence of colobomas ranges from 2.6 in 10,000 births in the U.S. to 7.5 in 10,000 births in China, and colobomas are estimated to be present in 3-11% of all blind children worldwide (Onwochei et al., 2000). Colobomas are also present in over 50 human genetic disorders (OMIM), often associated with other ocular abnormalities like microphthalmia (Bermejo and Martinez-Frias, 1998). Despite a significant amount of genetic research to identify coloboma loci, causative mutations have been identified in less than 20% of coloboma patients (Gregory-Evans et al., 2004; Fitzpatrick and van Heyningen, 2005; Chang et al., 2006). Moreover, we lack a comprehensive mechanistic understanding of the cellular and molecular regulation of CF closure in the human eye or in any of the animal model systems utilized for modeling human eye development and disease. Studies from a number of laboratories, and from both human genetics and experimental analyses in a variety of animal systems, have identified a suite of gene products required for CFC (reviewed in (Bibliowicz et al., 2011; Morris, 2011; Gestri et al., 2012). These include: Pax2 (Sanyanusin et al., 1995; Torres et al., 1996; Macdonald et al., 1997; Bower et al., 2012), GDF3 and GDF6 (Asai-Coakwell et al., 2007; Ye et al., 2010), CHD7 (Bosman et al., 2005; Lalani et al., 2006; Bajpai et al., 2010), SALL2 (Kelberman et al., 2014), YAP1 (Williamson et al., 2014), VSX2 (Bar-Yosef et al., 2004), SMOC1 (Rainger et al., 2011), Sox11 (Pillai-Kastoori et al., 2014), Jnk1/2 (Weston et al., 2003), MAB21L2 (Rainger et al., 2014; Deml et al., 2015), and Vax1 and Vax2 (Barbieri et al., 2002; Take-uchi et al., 2003), as well as components of the Hedgehog (Schimmenti et al., 2003; Koudijs et al., 2008; Lee et al., 2008), Fibroblast growth factor (FGF) (Cai et al., 2013; Chen et al., 2013a; Atkinson-Leadbeater et al., 2014), Wnt (Liu et al., 2016), and Retinoic acid (RA) (Matt et al., 2008; See and Clagett-Dame, 2009; Lupo et al., 2011) signaling pathways. While the mechanisms underlying their requirements during CFC have not been fully elucidated in each case, many of these gene products act early in eye development to modulate key events in retinal growth and cell survival, optic cup morphogenesis and patterning that are ultimately required for CFC. Defects in any of these processes result in colobomas; however, these are likely an indirect consequence of defects in ventral optic cup formation or patterning, and not a direct function of the mutated gene product during CFC. Indeed, we have a far more limited understanding of the later events in eye development that function to close the CF. Results from several studies link an inability to degrade the basement membrane (BM)/basal lamina that lines the CF to defects in CFC and colobomas. For example, deficiencies in RA signaling during the later stages of eye development in rats result in a retention of the BM lining the CF and colobomas (See and Clagett-Dame, 2009). Knockout or mutations in Pax2/pax2 (Torres et al., 1996; Macdonald et al., 1997) or Vax2 (Barbieri et al., 2002) also result in retention of the CF BM and colobomas. Much like BM breakdown, the mechanism of tissue fusion that ultimately seals the tightly apposed sides of the CF during CFC is also poorly understood. Roles for adhesion regulators like N-cadherin (Erdmann et al., 2003; Masai et al., 2003) and α-catenin (Chen et al., 2012) have been identified in mediating CFC, but the morphogenetic events leading to tissue fusion in the CF and their cellular underpinnings have not been elucidated in any system. To begin to identify the cellular and molecular underpinnings of these later aspects of CFC, we utilized the zebrafish embryo and a combination of fixed-sample and in vivo imaging in wild-type (WT) and mutant lines. We determined the spatial and temporal aspects of BM breakdown and tissue fusion during CFC, identifying unique characteristics of CF cells throughout the closure process. We identify BM membrane breakdown defects in the CF of talin1 and cloche mutants, and these defects correlate with malformations in, or absence of, the periocular mesenchyme (POM)-derived hyaloid vasculature. Taken together, these data support a model in which the hyaloid vasculature itself, or the POM-derived endothelial cells that generate the hyaloid vasculature, facilitate BM breakdown during CFC. MATERIALS AND METHODS Animals Zebrafish were maintained as described (Westerfield, 1995). Embryos were obtained and staged as described (Kimmel et al., 1995). We utilized the transgenic lines Tg(fli1a:eGFP) (Lawson and Weinstein, 2002), Tg(sox10:eGFP) (Wada et al., 2005), Tg(sox10:mRFP) (Kirby et al., 2006), Tg(kdrl:mCherry) and Tg(kdrl:moesin-GFP) (Wang et al., 2010). tln1hi3093Tg/+ embryos (Amsterdam et al., 2004) were obtained from ZIRC (Eugene, OR, USA) and crossed with Tg(kdrl:mCherry). tln1hi3093tg mutants were genotyped using these primers: 5’-ccaaacctacaggtggggtc-3’ and 5’-taccagcatttactcaacaggaac-3’. clochem378 embryos (Stainier et al., 1995) were a kind gift of Dr. Beth Roman (University of Pittsburgh) and mutants were identified based on lack of blood flow at 48hpf. All protocols used within this study were approved by the Institutional Animal Care and Use Committee of The University of Texas at Austin and The University of Pittsburgh School of Medicine, and conform to the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Microinjection of mRNAs Capped mRNAs were synthesized with a mMESSAGE mMACHINE Sp6 kit (Life Technologies) and were injected into one-cell-stage embryos with the following amounts: 150 pg of membrane-GFP or 100pg of utrophin-GFP (Burkel et al., 2007). In situ hybridization In situ hybridizations were performed as described (Jowett and Lettice, 1994). A talin1 in situ probe was synthesized from a partial cDNA cloned by RT-PCR with the following primers: 5’-tagcagtggcacagtcccgtattg-3’ and 5’-ttttttcttcctccaggcttg-3’. Embryos were subsequently cryosectioned and imaged, or immunostained in the case of fli1a:GFP and imaged. Immunohistochemistry Immunohistochemistry was performed as described (Uribe and Gross, 2007). Briefly, zebrafish embryos were fixed overnight in 4% paraformaldehyde, or 30 minutes at room temperature (for β-catenin), and then cryosectioned. Primary antibodies were used at the following dilutions: rabbit polyclonal anti-laminin (Sigma, #L9393), 1:200; mouse monoclonal anti-β-catenin (BD Biosciences, 610153), 1:250; and rabbit polyclonal anti-GFP (Life Technologies, A11122), 1:50. Phalloidin (Alexa Fluor 488 or 633, Invitrogen), DAPI or Sytox-Green were applied along with, or just prior to, the application of Cy2, Cy3 or Cy5 conjugated Goat anti-mouse or anti-rabbit IgG secondary antibodies (Jackson Labs). Imaging For sectioned embryos, imaging was performed with a Zeiss LSM5 Pascal or Olympus FV1200 confocal microscope. In situ cryosections were imaged utilizing a Leica DM2500 with a 100X oil immersion objective (Numerical aperture: 1.25). Brightfield images were captured on a Leica MZ16F stereomicroscope mounted with a DFC480 digital camera. In vivo time-lapse imaging was performed on a Leica TCS SP5 II confocal microscope equipped with a 25X (Numerical aperture: 0.95) water immersion objective as previously described (Hartsock et al., 2014), or an Olympus FV1200 with a 20X (Numerical aperture: 1.0) water immersion objective. TUNEL Staining Embryos were fixed in 4% paraformaldehyde and cryosectioned. The sectioned tissues were permeabilized with 0.1% TritonX-100 in 0.1% sodium citrate and then treated with the TUNEL reaction mixture (Roche). Images were captured with a Zeiss LSM5 Pascal. RESULTS Temporal and spatial dynamics of basement membrane breakdown during CF closure To identify the cellular and molecular underpinnings of CFC in zebrafish, we first performed a detailed analysis of the temporal and spatial dynamics of CFC. In mouse and hamster, BM breakdown correlates with CFC (Geeraets, 1976; Hero, 1989; Hero, 1990). Laminin-111 (Lam-111) immunostaining was used as a marker for the BM lining the CF to determine whether this is also the case in zebrafish. Embryos were sectioned sagittally, at varying depths along the proximal-distal axis of the eye (Fig. 1A,N), and the presence or absence of Lam-111 assessed. We designated the vitreous cavity as “central” along the proximal-distal axis of the CF, as this was a fiducial marker easily identifiable in the continually growing eye over time. For Lam-111 immunohistochemistry, we then examined sections taken at 12um intervals proximal and distal of this central point. Previous histological data suggested that CFC starts at 36 hour post fertilization (hpf) in zebrafish (Schmitt and Dowling, 1994), so we examined 36hpf embryos in our initial assays (Fig. 1J-M). When Lam-111 distribution was examined in these eyes however, the BM was already absent/degraded in the central/proximal region of the CF (Fig. 1L), and was present or partially degraded in more distal regions (Fig. 1J,K). Examination of eyes from younger embryos (31hpf) revealed that, despite the tight apposition of the sides of the CF, the BM remained intact at all proximal-distal depths (Fig. 1B-E; inset in Fig. 1D). By 34hpf, BM breakdown had initiated and the BM separating the closely apposed sides of the CF was discontinuous (Fig. 1F-I, inset in Fig. 1H). Spatially, BM breakdown began in the central/proximal CF (Fig. 1F-I), and proceeded bi-directionally from this point (Fig. 1 J-M). Laminin-111 staining was mostly absent in the CF by 48hpf (Fig. 1N-P) except in the most distal region of the eye in some embryos, where it was no longer detected at 60hpf (data not shown). TEM studies in mouse (Hero, 1989; Hero, 1990; Hero et al., 1991) and hamster (Geeraets, 1976) have also reported a correlation between cell death in the CF and CFC, although the functional relevance of this is unknown. TUNEL staining during zebrafish CFC did not identify any apoptotic cells in the CF (data not shown). Temporal and spatial dynamics of tissue fusion during CF closure Next, we wanted to determine when tissue fusion occurs during CFC and we turned to in vivo imaging to address this question. Embryos were injected with membrane-GFP (Fig. 2) orutrophin-GFP (data not shown) mRNA at the one-cell stage and embryos were imaged at various proximal-distal positions throughout the CF over time. Our initial imaging strategy focused on 24-40hpf embryos, hypothesizing that fusion would likely occur between ~32-34hpf, correlating closely with basement membrane breakdown. However, during this time frame, fusion appeared to begin only at the very end of the imaging window (data not shown). Thus, we shifted our imaging window slightly later, to 32-52hpf, to better observe the initiation and completion of CFC (Fig. 2A; n=4 embryos; Supplemental Movies 1-3). As above, for axial orientation along the proximal-distal axis we again designated the central CF as the vitreous cavity; however, because there is substantial growth and morphogenesis occurring in the eye over these time points, and because we did not utilize a lineage marker to track individual cells, it is not possible to directly compare the locations of CF cells from early time points (i.e. BM breakdown assays) to later time points analyzed in these fusion assays. From the time-lapse movies, fusion of the CF in the most proximal regions imaged appeared to initiate at ~37.5hpf and to be complete by 44hpf, based on the tight apposition between the two sides of the CF and the appearance of a seemingly continuous RPE (Fig. 2D). In the distal/central CF, the two sides of the CF became tightly apposed at ~46 to 47hpf (Fig. 2C), while in the most distal CF, fusion had yet to begin even at 49hpf (Fig. 2B). Absence of Lam-111 staining in the central CF and the tight apposition of the sides of the CF at 36hpf (Fig. 1L) initially suggested that tissue fusion might occur immediately after BM breakdown; however, in vivo imaging data indicated a significantly later timing for the fusion process. To address this apparent contradiction, we performed immunohistochemistry on sagittal sections taken at 16um intervals along the proximal-distal axis of the CF, identifying the margins of CF cells with phalloidin, which marks F-actin, and adherens junctions with β-catenin immunohistochemistry (Fig. 3). Co-localization of F-actin and β-catenin provides a reliable indication that a stable adherens junction had formed, and thus, that fusion had occurred (Gumbiner, 2000; Halbleib and Nelson, 2006; Hartsock and Nelson, 2008; Oda and Takeichi, 2011). Examining single 1um optical sections at 44hpf revealed that F-actin and β-catenin were co-localized in the central/proximal region of the CF, indicating that the CF was indeed fused at this time (Fig. 3D). Proximal to this however, the CF was still open at 44hpf (Fig. 3E), despite in vivo imaging data that indicated the two sides of the CF might already have undergone fusion (Fig. 2). Tissue fusion proceeded bi-directionally from the central/proximal region, with F-actin and β-catenin co-localization in the central (Fig. 3M) and proximal (Fig. 3O) aspects of the CF at ~47hpf (Fig. 3M), and the distal/central CF at 49hpf (Fig. 3Q). Fusion is complete by ~54hpf in most embryos, with the exception of the most distal CF, which is fully closed by 72hpf (data not shown). At all regions of fusion, F-actin and β-catenin formed a “seam” at the site of fusion (e.g. Fig. 3M), which then dissipated within 1-2 hours thereafter (e.g. Fig. 3R). Previously, in cell culture it was observed that there are areas of organized actin outside of the apical zonula adherens that are required for proper contractile activity (Wu et al., 2014). Examination of the CF in each single 1um optical slice through an individual 16um section plane revealed a progressive co-localization between F-actin and β-catenin within the CF where there are points in which β-catenin is present without F-actin and vice versa, indicating that both the traditional apical cortical ring (seam) and lateral adhesion clusters (puncta) are present in CF cells during CF fusion (Fig. 3U). Periocular mesenchyme-derived endothelial cells contribute to CFC TEM studies in mouse have shown that BM breakdown correlates with a population of “amoeboid phagocytic cells” present in the CF, and these cells possess pseudopodia which adhere to the BM at regions where it is being actively degraded (Hero, 1990; Hero et al., 1991). In Mitf−/− mutants, there is a decrease or complete absence of these phagocytic cells in CF regions where BM breakdown did not occur. Based on these observations, the authors hypothesized that these phagocytic cells were of mesenchymal origin and that they contribute to BM breakdown. Periocular mesenchyme (POM) cells migrate into the developing eye to form a variety of ocular and extraocular structures (Gage et al., 2005). Moreover, POM cells have been reported to function during CFC in a variety of contexts, and in a variety of model systems, with defects in POM function or survival resulting in colobomas (Evans and Gage, 2005; Kim et al., 2007; McMahon et al., 2009; See and Clagett-Dame, 2009; Bajpai et al., 2010; Lupo et al., 2011). Despite these studies, the cellular mechanisms through which POM cells contribute to CFC remain largely unknown. We hypothesized that POM might play a direct role in mediating CFC. To test this hypothesis, we utilized transgenic zebrafish lines in which GFP is expressed in distinct populations of POM cells. Tg(sox10:eGFP) expresses GFP in neural crest cells (Wada et al., 2005); during CFC, sox10:GFP+ cells were detected transiently in the CF, but not in regions where BM breakdown was occurring and were no longer present within the CF at 37hpf (Fig. 4A-A’’; Fig. S1A-A”). In the eye, subsets of POM cells migrate through the CF to generate the hyaloid vasculature, and these are labeled in Tg(fli1a:eGFP) embryos (Lawson and Weinstein, 2002; Alvarez et al., 2007; Hartsock et al., 2014). In comparison to sox10:GFP+ cells, which rapidly transit the CF to enter the anterior chamber, fli1a:eGFP+ cells are retained in the CF to generate the hyaloid. Moreover, observing kdrl:mCherry+ hyaloid cells in membrane-GFP injected embryos highlights that hyaloid cells are retained in the CF as late as 40hpf (Supplemental Movie 4). Interestingly, fli1a:eGFP+ cells localized to regions of the CF where the BM was absent or was punctate (Fig. 4B,C; Fig. S1B-E), and they possessed pronounced F-actin accumulations that localize to regions of BM breakdown (Fig. 4C-C’’’). POM is required for basement membrane breakdown during choroid fissure closure The data presented thus far support a model in which POM cells could play a role during CFC, possibly utilizing an F-actin dependent process to facilitate BM breakdown. To test this model, we analyzed Talin function during CFC. Talin is a scaffold protein that links integrins to the actin cytoskeleton, playing a critical role in integrin activation (Legate and Fassler, 2009; Moser et al., 2009; Desiniotis and Kyprianou, 2011). In zebrafish, talin1 is enriched in the CF, in both POM cells lining the CF and retinal/RPE cells that comprise the fissure (Fig. 5A, D-F). A talin1 loss of function mutant was previously identified, but its role during eye development has not yet been examined (Amsterdam et al., 2004). tln1hi3093 mutants possessed colobomas at 3dpf (Fig. 5B,C), with the distal aspect of the CF more affected, reminiscent of an iris coloboma in humans (Chang et al., 2006). BM degradation in the CF of tln1 mutants was deficient at 48hpf, a time point at which BM degradation was complete in most CF regions of phenotypically wild-type siblings (Fig. 5G-J). Distribution of talin1 in the POM and retinal/RPE cells within the CF, and correlation between fli1a:eGFP expressing cells and regions of BM breakdown indicated that the POM-derived hyaloid vasculature might be affected in tln1 mutants. To assess hyaloid formation we crossed the kdrl:mCherry transgene into the tln1 mutant line and imaged the hyaloid as previously described (Hartsock et al., 2014). Indeed, the structure of the hyaloid was disrupted in the distal/central region of the CF with the hyaloid severely hypotrophic relative to that in wild-type siblings (Fig. 5K-N). These findings support a model in which BM breakdown during CFC requires an actin-mediated process and indicate that POM-derived endothelial cells and/or the hyaloid itself may play a direct role in facilitating BM breakdown. To test this model we utilized the cloche mutant which lacks all early ocular vasculature (Stainier et al., 1995; Liao et al., 1997; Dhakal et al., 2015), thus eliminating any potential contribution of POM-derived endothelial cells to CFC. As in tln1 mutants, BM degradation was deficient in cloche mutants and Lam-111 staining was detected in both proximal and distal regions of the CF at 51hpf (Fig 6C,D), a time at which it is absent from wild-type siblings (Fig 6A,B). Lam-111 staining was absent in cloche mutants by 3dpf (data not shown) however, indicating that the absence of the hyaloid vasculature delayed, but did not completely prevent, BM breakdown. DISCUSSION The molecular and cellular underpinnings of CFC are largely unknown, despite the relatively high incidences of colobomas in the human population and the increasing number of identified loci associated with isolated or syndromic colobomas (e.g. (Graw, 2003; Gregory-Evans et al., 2004; Fitzpatrick and van Heyningen, 2005; Chang et al., 2006). CFC can be segregated into three principal stages (Fig. 7A). During Stage I, optic cup morphogenesis proceeds such that the correct number of cells is generated, and the retina/RPE aspects of the CF properly develop and become positioned in the ventral aspect of the optic cup. Stage II of CFC involves breakdown of the basement membrane/basal lamina lining the neuroepithelial and RPE components of the CF. Stage III is comprised of tissue fusion events that serve to close the CF. Many genes mutated in human coloboma patients have been shown to mechanistically function during Stage I when examined in animal model systems; these include SALL2 (Kelberman et al., 2014), PAX2 (Sanyanusin et al., 1995; Torres et al., 1996; Macdonald et al., 1997; Bower et al., 2012), CHD7 (Bosman et al., 2005; Lalani et al., 2006; Bajpai et al., 2010), and GDF3 and GDF6 (Asai-Coakwell et al., 2007; Ye et al., 2010). Moreover, in vivo imaging studies have also begun to identify cell movements and cell populations that generate distinct components of the CF (Picker et al., 2009). By comparison, virtually nothing is known about the mechanisms underlying the Stage II BM breakdown or Stage III tissue fusion events of CFC. Here, we have utilized zebrafish to determine the spatial and temporal characteristics of BM breakdown (Fig. 7B) and tissue fusion (Fig. 7C) and thereby provide a model in which the cellular and molecular underpinnings of these processes can now be elucidated. Previous studies in a variety of model systems have led to the hypothesis that defects in Stage II, an inability to degrade the CF BM, result in colobomas. Indeed, knockout or loss of function mutations in Pax2/pax2 (Torres et al., 1996; Macdonald et al., 1997) and Vax2 (Barbieri et al., 2002) result in retention of the CF BM and colobomas. Similarly, in Fatty Liver Shionogi mice, the CF BM is retained and colobomas result (Tsuji et al., 2012). These data lead to a model in which BM breakdown is required for normal CFC; however there is very little known about how this process occurs in vivo. As discussed above, TEM studies in mouse have shown that BM breakdown correlates with a population of cells that are possibly mesenchymal in origin (Hero, 1990; Hero et al., 1991). POM cells, comprised of lateral plate mesenchyme and cranial neural crest derivatives, migrate into the developing eye to form a variety of ocular and extraocular structures (Gage et al., 2005; Williams and Bohnsack, 2015). POM cells are known to be required for CFC, with defects in lmx1b, Tfap2a/tfap2a, foxc1, nlz1 and Pitx2/pitx2 resulting in colobomas (Gage et al., 1999; Brown et al., 2009; Gestri et al., 2009; McMahon et al., 2009; Bassett et al., 2010; Lupo et al., 2011). Retinoic acid (RA) signaling also contributes to CFC, by acting both directly on the ventral optic cup, as well as regulating gene expression within the POM (Lupo et al., 2011). Indeed, defects in RA signaling have been shown to cause a reduction in Pitx2 expression in the POM, resulting in retention of the BM lining the CF and colobomas (See and Clagett-Dame, 2009). Of particular interest are the POM-derived endothelial cells that migrate through the CF and into the vitreous to form the hyaloid (Saint-Geniez and D'Amore, 2004; Alvarez et al., 2007; Hartsock et al., 2014). Our data demonstrate that regions of BM breakdown correlate with actin enrichment in POM cells within the CF and that these cells are likely POM-derived endothelial cells that express fli1a-driven transgenes and give rise to the hyaloid vasculature (Fig. 4). Roles for endothelial cells during tissue morphogenesis have been identified in the developing liver; for example, endothelial cells are required for hepatic outgrowth, acting prior to the formation of functional vessels (Matsumoto et al., 2001), and they influence the apical-basal polarity of developing hepatocytes (Sakaguchi et al., 2008). We identified actin enrichment in POM cells coincident with areas of BM degradation (Fig 4), and, when combined with the classic TEM studies of Isabelle Hero (Hero, 1990; Hero et al., 1991), we hypothesized that POM cells actively degrade the CF BM as they transit through the CF. We examined BM breakdown in talin1 and cloche mutants (Figs. 5,6). talin1 encodes a scaffold protein linking integrins to the actin cytoskeleton, but little is known about function during eye development. cloche mutants lack all POM-derived ocular vasculature but surprisingly, despite being microphthalmic, the early phases of eye development are fairly normal in cloche, with pronounced retinal defects not detectable until after 48hpf, well beyond the window for BM breakdown in most of the CF (Dhakal et al., 2015). Both mutants possessed defects in BM breakdown (Fig. 5,6), thus supporting a model in which POM-derived endothelial cells and/or the hyaloid vasculature itself, facilitate BM breakdown during CFC. Notably, BM breakdown was delayed, but did eventually complete in cloche mutants. These data suggest that the ocular vasculature is not the only cell or tissue type required for BM breakdown and that there is likely a hyaloid-independent component to BM breakdown. When compared to cloche, the more severe BM breakdown and coloboma phenotypes in tln1 mutants also suggest that whatever cell or tissue type is required for efficient BM breakdown, it is likely to do so in an actin-dependent fashion. The TEM studies of Hero noted cellular protrusions from retinal cells in the CF that correlated with regions of BM breakdown (Hero, 1990) and thus, it is possible that retinal and/or RPE components of the CF also participate in BM breakdown and fill such a role in cloche mutants. Mechanistically, podosomes and invadosomes are known to mediate focal BM breakdown in a variety of developmental and cell biological contexts, including endothelial cells (Murphy and Courtneidge, 2011; van den Dries et al., 2014) making them interesting candidates for further analysis. Future studies will examine the BM breakdown and CFC phenotypes in loss-of function mutants for podosome/invadosome components like tks4/sh3pxd2, mmp2 and mmp14. With respect to tissue fusion during CFC (Stage III), several gene products have been identified that are required for fusion in the CF, but the mechanisms underlying the fusion process are not yet clear (Erdmann et al., 2003; Masai et al., 2003; Chen et al., 2012). In addition to defects arising from loss of function of retinal genes, disruptions to RPE specification and differentiation have also been shown to result in colobomas. For example, disruption of Mitf or an RPE-specific loss of β-catenin produces colobomas during mouse eye development (Scholtz and Chan, 1987; Westenskow et al., 2009), supporting the idea that the RPE also contributes to CFC. CFC defects arising in these models are thought to result from early Stage I defects in which overall optic cup development is perturbed, but it is also possible that CFC defects reflect a role for the RPE in tissue fusion itself. As in mouse and hamster (Geeraets, 1976; Hero, 1990; Hero et al., 1991), the RPE is inverted into the CF during the early stages of CFC in zebrafish (Fig 3 and data now shown) supporting a potential role in mediating fusion. Through in vivo imaging, F-actin staining and β-catenin immunohistochemistry, we determined the spatial and temporal characteristics of tissue fusion in zebrafish. While in vivo time lapse imaging provided a general indication to the timing of CFC, β-catenin immunohistochemistry provided a more accurate molecular readout of fusion between the tightly apposed sides of the CF. Given the transient enrichment of β-catenin at regions of nascent adhesion, cadherin-mediated mechanisms are likely to be involved in CF fusion. Indeed, multiple N-cadherin mutants exist in zebrafish that present with colobomas (Liu et al., 2001; Erdmann et al., 2003; Masai et al., 2003) although the cell biological mechanisms leading to these defects have not yet been resolved. Other cadherins expressed in the zebrafish eye include R-cadherin (Liu et al., 1999a; Liu et al., 1999b; Babb et al., 2005), cadherin 6 (Liu et al., 2008), and protocadherin 9 and 17 (Liu et al., 2009; Chen et al., 2013b), and these could also play a role in tissue fusion during CFC. Finally, the rapid turnover of β-catenin within the CF presents an excellent model to study how formation and maintenance of adherens junctions are regulated in vivo. Supplementary Material 1 2 3 4 5 06 Acknowledgements We are grateful to Jeff Essner for kdrl:mCherry and kdrl:moesin-GFP transgenics, Beth Roman for cloche, Adam Kwiatkowski for providing antibodies, members of the Gross lab for helpful suggestions and comments on this work, and Roky Erickson for technical support. This work was supported by a CAREER Award from the National Science Foundation (IOS-0745782), grants from the National Eye Institute (RO1-EY18005 to JMG, and F32-EY23910 to AH), and NIH CORE Grant P30 EY08098 to the Department of Ophthalmology. In vivo imaging was performed on a confocal microscope funded by NIH S10-RR028951. Zebrafish were obtained from ZIRC, which is supported by NIH-NCRR Grant P40 RR012546. We acknowledge additional support from the Eye and Ear Foundation of Pittsburgh and from an unrestricted grant from Research to Prevent Blindness, New York, NY. Figure 1 Temporal and spatial dynamics of basement membrane breakdown during choroid fissure closure in zebrafish (A) Schematic depicting the approximate level of sections in B-M along the proximal-distal axis of the CF. The vitreous cavity was defined as central, and sections were taken at 12um intervals proximally and distally from this point. (B-M) Sagittal sections along the proximal-distal axis of the retina, immunostained for Lam-111 expression. (B-E) 31 hpf, (F-I) 34 hpf, (J-M) 36 hpf. Insets in D, H show high magnification views of the regions in the dashed boxes. (N) Schematic depicting the plane of section for 48hpf embryos in O,P. (O,P) Representative sagittal sections along the proximal-distal axis of the retina immunostained for Lam-111. Scale bars = 20 μm. Figure 2 in vivo imaging of choroid fissure closure in zebrafish (A) Schematic depicting the approximate level of sections in B-D along the proximal-distal axis of the CF. The vitreous cavity was defined as central, and optical sections were taken at 16um intervals proximally and distally from this point. (B-D) membrane-GFP injected embryos were imaged throughout the CF. Single micron optical slices are shown from 44-49hpf and at three distinct proximal-distal regions of the CF. (B) Distally, the CF remains open until at least 49hpf. (C) Distal/centrally, the CF appears to close between 46-47hpf. (D) Proximally, the CF already appears to be closed at 44hpf. Orange arrowheads in B,C mark open CF. White arrow in C marks what appears to be a closed CF. Dashed line outlines the RPE. Scale bar = 50 μm. Figure 3 Temporal and spatial dynamics of tissue fusion during choroid fissure closure in zebrafish Single micron optical sections from sagittal cryosections stained with phalloidin (green) and anti-β-catenin (red) at distinct proximal-distal regions of the CF over time. As in Figure 2, the vitreous cavity was defined as central, and sections were taken at 16um intervals proximally and distally from this point. (A-E) At 44hpf, the CF is fused in central/proximal sections (white arrow) based on co-localization between F-actin and β-catenin in a fusion ‘seam’. (F-J) At 45hpf, the two sides of the CF are tightly apposed but no fusion outside of the central-proximal region is detected. (K-O) At 47hpf, a fusion seam is present within central and proximal sections, and there are punctate regions of co-localization in distal/central sections. (P-T) At 49hpf, the fusion seam has disappeared in the central and proximal CF regions while it is appearing in the distal/central region. (U) Single 1um optical sections from one section plane aligned distal (left) to proximal (right) demonstrate a progressive co-localization of F-actin and β-catenin in the CF and formation of the fusion seam. Scale bar = 50μm (A-T) and 5μm (U). Figure 4 Periocular mesenchymal cells contribute to CFC (A-C) Sagittal views of the CF stained with anti-GFP (green), Lam-111 (red) and/or phalloidin (blue). (A) Few sox10:eGFP+ cells are detected in the CF, 37hpf section pictured. (B) fli1a:eGFP+ cells are retained in the CF. 36hpf section pictured. (C) fli1a:eGFP+ cells possess F-actin accumulations that localize to regions of BM breakdown. Arrows denote puncta of F-actin where Lam-111 is low or absent. 34hpf section pictured. Scale bar = 20μm (A,B) and 10μm (C). Figure 5 talin1 is required for CFC in zebrafish (A) talin1 is expressed within the POM and retinal/RPE cells lining the CF at 33hpf (arrow). (B,C) Lateral views of the eye of tln1hi3093Tg mutant (C) and wild-type sibling (B) at 3dpf. tln1 mutants possess colobomas. (D-F) Distal section through the eye of a 32hpf embryo demonstrating talin1 expression within the retina/RPE cells lining the CF (arrows) and the hyaloid vasculature (marked by GFP expression from fli1a:eGFP; arrowhead). (G-J) Sagittal sections through the eyes of 48hpf tln1 mutants and siblings stained with Lam-111 (red) and Sytox-green (DNA; green). BM degradation is disrupted in tln1 mutants at 48hpf. (K-N) Maximum projection images of the distal hyaloid in tln1 mutants and siblings demonstrating severe hypotrophy of the hyaloid in the tln1 mutant at 44.5hpf. Scale bar = 20μm (G-J, K,M) and 50μm (L,N). Figure 6 POM-derived endothelial cells facilitate BM breakdown during CFC (A-D) Sagittal sections through the eyes of 51hpf clochem378 mutants and siblings stained with Lam-111 (red) and Sytox-green (DNA; green). BM degradation is disrupted in clochem378 mutants in both the (C) proximal and (D) distal regions of the CF when compared to (A,B) siblings. Scale bar = 25um. Figure 7 Schematics depicting key stages and events of CFC (A) During Stage I, tissue growth and optic cup morphogenesis generates an appropriate number of cells, which are correctly patterned and positioned in the optic cup. The opposing sides of the CF become closely apposed. During Stage II, the basement membrane lining the CF is degraded through a process involving periocular mesenchyme cells. During Stage III, tissue fusion between opposing sides of the CF closes the fissure. (B) Schematic of basement membrane breakdown (green = BM) within the CF from 31-48hpf. BM breakdown initiates in the central/proximal CF and proceeds bi-directionally, being complete by 48hpf except in the most distal regions of the CF. (C) Schematic of tissue fusion (blue = fusion) in the CF from 41-54hpf. The hyaloid vasculature is depicted in red in each image. Fusion initiates in the central/proximal CF and proceeds bi-directionally, being complete by ~54hpf except in the most distal regions of the CF. Highlights Utilizing zebrafish as a model to identify the cellular and molecular underpinnings of choroid fissure closure The hyaloid vasculature and/or the POM-derived endothelial cells that give rise the hyaloid vasculature contribute to basement membrane breakdown during CF closure This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. References Alvarez Y Cederlund ML Cottell DC Bill BR Ekker SC Torres-Vazquez J Weinstein BM Hyde DR Vihtelic TS Kennedy BN Genetic determinants of hyaloid and retinal vasculature in zebrafish BMC developmental biology 2007 7 114 17937808 Amsterdam A Nissen RM Sun Z Swindell EC Farrington S Hopkins N Identification of 315 genes essential for early zebrafish development Proceedings of the National Academy of Sciences of the United States of America 2004 101 35 12792 7 15256591 Asai-Coakwell M French CR Berry KM Ye M Koss R Somerville M Mueller R van Heyningen V Waskiewicz AJ Lehmann OJ GDF6, a novel locus for a spectrum of ocular developmental anomalies American journal of human genetics 2007 80 2 306 15 17236135 Atkinson-Leadbeater K Hehr CL McFarlane S Fgfr signaling is required as the early eye field forms to promote later patterning and morphogenesis of the eye Developmental dynamics : an official publication of the American Association of Anatomists 2014 243 5 663 75 24478172 Babb SG Kotradi SM Shah B Chiappini-Williamson C Bell LN Schmeiser G Chen E Liu Q Marrs JA Zebrafish R-cadherin (Cdh4) controls visual system development and differentiation Developmental dynamics : an official publication of the American Association of Anatomists 2005 233 3 930 45 15918170 Bajpai R Chen DA Rada-Iglesias A Zhang J Xiong Y Helms J Chang CP Zhao Y Swigut T Wysocka J CHD7 cooperates with PBAF to control multipotent neural crest formation Nature 2010 463 7283 958 62 20130577 Bar-Yosef U Abuelaish I Harel T Hendler N Ofir R Birk OS CHX10 mutations cause non-syndromic microphthalmia/ anophthalmia in Arab and Jewish kindreds Human genetics 2004 115 4 302 9 15257456 Barbieri AM Broccoli V Bovolenta P Alfano G Marchitiello A Mocchetti C Crippa L Bulfone A Marigo V Ballabio A Vax2 inactivation in mouse determines alteration of the eye dorsal-ventral axis, misrouting of the optic fibres and eye coloboma Development 2002 129 3 805 13 11830579 Bassett EA Williams T Zacharias AL Gage PJ Fuhrmann S West-Mays JA AP-2alpha knockout mice exhibit optic cup patterning defects and failure of optic stalk morphogenesis Human molecular genetics 2010 19 9 1791 804 20150232 Bermejo E Martinez-Frias ML Congenital eye malformations: clinical-epidemiological analysis of 1,124,654 consecutive births in Spain American journal of medical genetics 1998 75 5 497 504 9489793 Bibliowicz J Tittle RK Gross JM Toward a better understanding of human eye disease insights from the zebrafish, Danio rerio Progress in molecular biology and translational science 2011 100 287 330 21377629 Bosman EA Penn AC Ambrose JC Kettleborough R Stemple DL Steel KP Multiple mutations in mouse Chd7 provide models for CHARGE syndrome Human molecular genetics 2005 14 22 3463 76 16207732 Bower M Salomon R Allanson J Antignac C Benedicenti F Benetti E Binenbaum G Jensen UB Cochat P DeCramer S Update of PAX2 mutations in renal coloboma syndrome and establishment of a locus-specific database Human mutation 2012 33 3 457 66 22213154 Brown JD Dutta S Bharti K Bonner RF Munson PJ Dawid IB Akhtar AL Onojafe IF Alur RP Gross JM Expression profiling during ocular development identifies 2 Nlz genes with a critical role in optic fissure closure Proceedings of the National Academy of Sciences of the United States of America 2009 106 5 1462 7 19171890 Burkel BM von Dassow G Bement WM Versatile fluorescent probes for actin filaments based on the actin-binding domain of utrophin Cell motility and the cytoskeleton 2007 64 11 822 32 17685442 Cai Z Tao C Li H Ladher R Gotoh N Feng GS Wang F Zhang X Deficient FGF signaling causes optic nerve dysgenesis and ocular coloboma Development 2013 140 13 2711 23 23720040 Chang L Blain D Bertuzzi S Brooks BP Uveal coloboma: clinical and basic science update Current opinion in ophthalmology 2006 17 5 447 70 16932062 Chen S Lewis B Moran A Xie T Cadherin-mediated cell adhesion is critical for the closing of the mouse optic fissure PLoS One 2012 7 12 e51705 23240058 Chen S Li H Gaudenz K Paulson A Guo F Trimble R Peak A Seidel C Deng C Furuta Y Defective FGF signaling causes coloboma formation and disrupts retinal neurogenesis Cell research 2013a 23 2 254 73 23147794 Chen Y Londraville R Brickner S El-Shaar L Fankhauser K Dearth C Fulton L Sochacka A Bhattarai S Marrs JA Protocadherin-17 function in Zebrafish retinal development Developmental neurobiology 2013b 73 4 259 73 22927092 Deml B Kariminejad A Borujerdi RH Muheisen S Reis LM Semina EV Mutations in MAB21L2 result in ocular Coloboma, microcornea and cataracts PLoS genetics 2015 11 2 e1005002 25719200 Desiniotis A Kyprianou N Significance of talin in cancer progression and metastasis International review of cell and molecular biology 2011 289 117 47 21749900 Dhakal S Stevens CB Sebbagh M Weiss O Frey RA Adamson S Shelden EA Inbal A Stenkamp DL Abnormal retinal development in Cloche mutant zebrafish Developmental dynamics : an official publication of the American Association of Anatomists 2015 244 11 1439 55 26283463 Erdmann B Kirsch FP Rathjen FG More MI N-cadherin is essential for retinal lamination in the zebrafish Developmental dynamics : an official publication of the American Association of Anatomists 2003 226 3 570 7 12619142 Evans AL Gage PJ Expression of the homeobox gene Pitx2 in neural crest is required for optic stalk and ocular anterior segment development Human molecular genetics 2005 14 22 3347 59 16203745 Fitzpatrick DR van Heyningen V Developmental eye disorders Current opinion in genetics & development 2005 15 3 348 53 15917212 Gage PJ Rhoades W Prucka SK Hjalt T Fate maps of neural crest and mesoderm in the mammalian eye Investigative ophthalmology & visual science 2005 46 11 4200 8 16249499 Gage PJ Suh H Camper SA Dosage requirement of Pitx2 for development of multiple organs Development 1999 126 20 4643 51 10498698 Geeraets R An electron microscopic study of the closure of the optic fissure in the golden hamster The American journal of anatomy 1976 145 4 411 31 1266776 Gestri G Link BA Neuhauss SC The visual system of zebrafish and its use to model human ocular diseases Developmental neurobiology 2012 72 3 302 27 21595048 Gestri G Osborne RJ Wyatt AW Gerrelli D Gribble S Stewart H Fryer A Bunyan DJ Prescott K Collin JR Reduced TFAP2A function causes variable optic fissure closure and retinal defects and sensitizes eye development to mutations in other morphogenetic regulators Human genetics 2009 126 6 791 803 19685247 Graw J The genetic and molecular basis of congenital eye defects Nature reviews. Genetics 2003 4 11 876 88 Gregory-Evans CY Williams MJ Halford S Gregory-Evans K Ocular coloboma: a reassessment in the age of molecular neuroscience Journal of medical genetics 2004 41 12 881 91 15591273 Gumbiner BM Regulation of cadherin adhesive activity The Journal of cell biology 2000 148 3 399 404 10662767 Halbleib JM Nelson WJ Cadherins in development: cell adhesion, sorting, and tissue morphogenesis Genes & development 2006 20 23 3199 214 17158740 Hartsock A Lee C Arnold V Gross JM In vivo analysis of hyaloid vasculature morphogenesis in zebrafish: A role for the lens in maturation and maintenance of the hyaloid Developmental biology 2014 394 2 327 39 25127995 Hartsock A Nelson WJ Adherens and tight junctions: structure, function and connections to the actin cytoskeleton Biochimica et biophysica acta 2008 1778 3 660 9 17854762 Hero I The optic fissure in the normal and microphthalmic mouse Experimental eye research 1989 49 2 229 39 2767170 Hero I Optic fissure closure in the normal cinnamon mouse. An ultrastructural study Investigative ophthalmology & visual science 1990 31 1 197 216 2298539 Hero I Farjah M Scholtz CL The prenatal development of the optic fissure in colobomatous microphthalmia Investigative ophthalmology & visual science 1991 32 9 2622 35 1869414 Jowett T Lettice L Whole-mount in situ hybridizations on zebrafish embryos using a mixture of digoxigenin- and fluorescein-labelled probes Trends in genetics : TIG 1994 10 3 73 4 8178366 Kelberman D Islam L Lakowski J Bacchelli C Chanudet E Lescai F Patel A Stupka E Buck A Wolf S Mutation of SALL2 causes recessive ocular coloboma in humans and mice Human molecular genetics 2014 23 10 2511 26 24412933 Kim TH Goodman J Anderson KV Niswander L Phactr4 regulates neural tube and optic fissure closure by controlling PP1-, Rb-, and E2F1-regulated cell-cycle progression Developmental cell 2007 13 1 87 102 17609112 Kimmel CB Ballard WW Kimmel SR Ullmann B Schilling TF Stages of embryonic development of the zebrafish Developmental dynamics : an official publication of the American Association of Anatomists 1995 203 3 253 310 8589427 Kirby BB Takada N Latimer AJ Shin J Carney TJ Kelsh RN Appel B In vivo time-lapse imaging shows dynamic oligodendrocyte progenitor behavior during zebrafish development Nature neuroscience 2006 9 12 1506 11 17099706 Koudijs MJ den Broeder MJ Groot E van Eeden FJ Genetic analysis of the two zebrafish patched homologues identifies novel roles for the hedgehog signaling pathway BMC developmental biology 2008 8 15 18284698 Lalani SR Safiullah AM Fernbach SD Harutyunyan KG Thaller C Peterson LE McPherson JD Gibbs RA White LD Hefner M Spectrum of CHD7 mutations in 110 individuals with CHARGE syndrome and genotype-phenotype correlation American journal of human genetics 2006 78 2 303 14 16400610 Lawson ND Weinstein BM In vivo imaging of embryonic vascular development using transgenic zebrafish Developmental biology 2002 248 2 307 18 12167406 Lee J Willer JR Willer GB Smith K Gregg RG Gross JM Zebrafish blowout provides genetic evidence for Patched1-mediated negative regulation of Hedgehog signaling within the proximal optic vesicle of the vertebrate eye Developmental biology 2008 319 1 10 22 18479681 Legate KR Fassler R Mechanisms that regulate adaptor binding to beta-integrin cytoplasmic tails Journal of cell science 2009 122 Pt 2 187 98 19118211 Liao W Bisgrove BW Sawyer H Hug B Bell B Peters K Grunwald DJ Stainier DY The zebrafish gene cloche acts upstream of a flk-1 homologue to regulate endothelial cell differentiation Development 1997 124 2 381 9 9053314 Liu C Widen SA Williamson KA Ratnapriya R Gerth-Kahlert C Rainger J Alur RP Strachan E Manjunath SH Balakrishnan A A secreted WNT-ligand-binding domain of FZD5 generated by a frameshift mutation causes autosomal dominant coloboma Human molecular genetics 2016 25 7 1382 91 26908622 Liu Q Babb SG Novince ZM Doedens AL Marrs J Raymond PA Differential expression of cadherin-2 and cadherin-4 in the developing and adult zebrafish visual system Visual neuroscience 2001 18 6 923 33 12020083 Liu Q Chen Y Pan JJ Murakami T Expression of protocadherin-9 and protocadherin-17 in the nervous system of the embryonic zebrafish Gene expression patterns : GEP 2009 9 7 490 6 19616130 Liu Q Londraville R Marrs JA Wilson AL Mbimba T Murakami T Kubota F Zheng W Fatkins DG Cadherin-6 function in zebrafish retinal development Developmental neurobiology 2008 68 8 1107 22 18506771 Liu Q Marrs JA Raymond PA Spatial correspondence between R-cadherin expression domains and retinal ganglion cell axons in developing zebrafish The Journal of comparative neurology 1999a 410 2 290 302 10414534 Liu Q Sanborn KL Cobb N Raymond PA Marrs JA R-cadherin expression in the developing and adult zebrafish visual system The Journal of comparative neurology 1999b 410 2 303 19 10414535 Lupo G Gestri G O'Brien M Denton RM Chandraratna RA Ley SV Harris WA Wilson SW Retinoic acid receptor signaling regulates choroid fissure closure through independent mechanisms in the ventral optic cup and periocular mesenchyme Proceedings of the National Academy of Sciences of the United States of America 2011 108 21 8698 703 21555593 Macdonald R Scholes J Strahle U Brennan C Holder N Brand M Wilson SW The Pax protein Noi is required for commissural axon pathway formation in the rostral forebrain Development 1997 124 12 2397 408 9199366 Masai I Lele Z Yamaguchi M Komori A Nakata A Nishiwaki Y Wada H Tanaka H Nojima Y Hammerschmidt M N-cadherin mediates retinal lamination, maintenance of forebrain compartments and patterning of retinal neurites Development 2003 130 11 2479 94 12702661 Matsumoto K Yoshitomi H Rossant J Zaret KS Liver organogenesis promoted by endothelial cells prior to vascular function Science 2001 294 5542 559 63 11577199 Matt N Ghyselinck NB Pellerin I Dupe V Impairing retinoic acid signalling in the neural crest cells is sufficient to alter entire eye morphogenesis Developmental biology 2008 320 1 140 8 18539269 McMahon C Gestri G Wilson SW Link BA Lmx1b is essential for survival of periocular mesenchymal cells and influences Fgf-mediated retinal patterning in zebrafish Developmental biology 2009 332 2 287 98 19500562 Morris AC The genetics of ocular disorders: insights from the zebrafish’, Birth defects research. Part C, Embryo today : reviews 2011 93 3 215 28 Moser M Legate KR Zent R Fassler R The tail of integrins, talin, and kindlins Science 2009 324 5929 895 9 19443776 Murphy DA Courtneidge SA The ‘ins’ and ‘outs’ of podosomes and invadopodia: characteristics, formation and function Nature reviews. Molecular cell biology 2011 12 7 413 26 21697900 Oda H Takeichi M Evolution: structural and functional diversity of cadherin at the adherens junction The Journal of cell biology 2011 193 7 1137 46 21708975 Onwochei BC Simon JW Bateman JB Couture KC Mir E Ocular colobomata Survey of ophthalmology 2000 45 3 175 94 11094243 Picker A Cavodeassi F Machate A Bernauer S Hans S Abe G Kawakami K Wilson SW Brand M Dynamic coupling of pattern formation and morphogenesis in the developing vertebrate retina PLoS biology 2009 7 10 e1000214 19823566 Pillai-Kastoori L Wen W Wilson SG Strachan E Lo-Castro A Fichera M Musumeci SA Lehmann OJ Morris AC Sox11 is required to maintain proper levels of Hedgehog signaling during vertebrate ocular morphogenesis PLoS genetics 2014 10 7 e1004491 25010521 Rainger J Pehlivan D Johansson S Bengani H Sanchez-Pulido L Williamson KA Ture M Barker H Rosendahl K Spranger J Monoallelic and biallelic mutations in MAB21L2 cause a spectrum of major eye malformations American journal of human genetics 2014 94 6 915 23 24906020 Rainger J van Beusekom E Ramsay JK McKie L Al-Gazali L Pallotta R Saponari A Branney P Fisher M Morrison H Loss of the BMP antagonist, SMOC-1, causes Ophthalmo-acromelic (Waardenburg Anophthalmia) syndrome in humans and mice PLoS genetics 2011 7 7 e1002114 21750680 Saint-Geniez M D'Amore PA Development and pathology of the hyaloid, choroidal and retinal vasculature The International journal of developmental biology 2004 48 8-9 1045 58 15558494 Sakaguchi TF Sadler KC Crosnier C Stainier DY Endothelial signals modulate hepatocyte apicobasal polarization in zebrafish Current biology : CB 2008 18 20 1565 71 18951027 Sanyanusin P Schimmenti LA McNoe LA Ward TA Pierpont ME Sullivan MJ Dobyns WB Eccles MR Mutation of the PAX2 gene in a family with optic nerve colobomas, renal anomalies and vesicoureteral reflux Nature genetics 1995 9 4 358 64 7795640 Schimmenti LA de la Cruz J Lewis RA Karkera JD Manligas GS Roessler E Muenke M Novel mutation in sonic hedgehog in non-syndromic colobomatous microphthalmia American journal of medical genetics. Part A 2003 116A 3 215 21 12503095 Schmitt EA Dowling JE Early eye morphogenesis in the zebrafish, Brachydanio rerio The Journal of comparative neurology 1994 344 4 532 42 7929890 Scholtz CL Chan KK Complicated colobomatous microphthalmia in the microphthalmic (mi/mi) mouse Development 1987 99 4 501 8 3665767 See AW Clagett-Dame M The temporal requirement for vitamin A in the developing eye: mechanism of action in optic fissure closure and new roles for the vitamin in regulating cell proliferation and adhesion in the embryonic retina Developmental biology 2009 325 1 94 105 18955041 Stainier DY Weinstein BM Detrich HW 3rd Zon LI Fishman MC Cloche, an early acting zebrafish gene, is required by both the endothelial and hematopoietic lineages Development 1995 121 10 3141 50 7588049 Take-uchi M Clarke JD Wilson SW Hedgehog signalling maintains the optic stalk-retinal interface through the regulation of Vax gene activity Development 2003 130 5 955 68 12538521 Torres M Gomez-Pardo E Gruss P Pax2 contributes to inner ear patterning and optic nerve trajectory Development 1996 122 11 3381 91 8951055 Tsuji N Kita K Ozaki K Narama I Matsuura T Organogenesis of mild ocular coloboma in FLS mice: failure of basement membrane disintegration at optic fissure margins Experimental eye research 2012 94 1 174 8 22182670 Uribe RA Gross JM Immunohistochemistry on cryosections from embryonic and adult zebrafish eyes CSH protocols 2007 2007: pdb prot4779 van den Dries K Bolomini-Vittori M Cambi A Spatiotemporal organization and mechanosensory function of podosomes Cell adhesion & migration 2014 8 3 268 72 24658050 Wada N Javidan Y Nelson S Carney TJ Kelsh RN Schilling TF Hedgehog signaling is required for cranial neural crest morphogenesis and chondrogenesis at the midline in the zebrafish skull Development 2005 132 17 3977 88 16049113 Wang Y Kaiser MS Larson JD Nasevicius A Clark KJ Wadman SA Roberg-Perez SE Ekker SC Hackett PB McGrail M Moesin1 and Ve-cadherin are required in endothelial cells during in vivo tubulogenesis Development 2010 137 18 3119 28 20736288 Westenskow P Piccolo S Fuhrmann S Beta-catenin controls differentiation of the retinal pigment epithelium in the mouse optic cup by regulating Mitf and Otx2 expression Development 2009 136 15 2505 10 19553286 Westerfield M The Zebrafish Book: A Guide for the Laboratory use of Zebrafish (Danio rerio) 1995 University of Oregon Eugene, OR Weston CR Wong A Hall JP Goad ME Flavell RA Davis RJ JNK initiates a cytokine cascade that causes Pax2 expression and closure of the optic fissure Genes & development 2003 17 10 1271 80 12756228 Williams AL Bohnsack BL Neural crest derivatives in ocular development: discerning the eye of the storm Birth defects research. Part C, Embryo today : reviews 2015 105 2 87 95 Williamson KA Rainger J Floyd JA Ansari M Meynert A Aldridge KV Rainger JK Anderson CA Moore AT Hurles ME Heterozygous loss-of-function mutations in YAP1 cause both isolated and syndromic optic fissure closure defects American journal of human genetics 2014 94 2 295 302 24462371 Wu SK Gomez GA Michael M Verma S Cox HL Lefevre JG Parton RG Hamilton NA Neufeld Z Yap AS Cortical F-actin stabilization generates apical-lateral patterns of junctional contractility that integrate cells into epithelia Nature cell biology 2014 16 2 167 78 24413434 Ye M Berry-Wynne KM Asai-Coakwell M Sundaresan P Footz T French CR Abitbol M Fleisch VC Corbett N Allison WT Mutation of the bone morphogenetic protein GDF3 causes ocular and skeletal anomalies Human molecular genetics 2010 19 2 287 98 19864492
PMC005xxxxxx/PMC5125861.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101623795 42112 Arthritis Rheumatol Arthritis & rheumatology (Hoboken, N.J.) 2326-5191 2326-5205 27390295 5125861 10.1002/art.39797 NIHMS800623 Article An Autotaxin-LPA-IL-6 Amplification Loop Drives Scleroderma Fibrosis Castelino Flavia V. M.D. 12 Bain Gretchen Ph.D. 4 Pace Veronica A. B.S. 12 Black Katharine E. M.D. 13 George Leaya B.A. 13 Probst Clemens K. B.S. 13 Goulet Lance B.S. 4 Lafyatis Robert M.D. 5 Tager Andrew M. M.D. 13 1 Center for Immunology and Inflammatory Diseases, Massachusetts General Hospital and Harvard Medical School, Boston, MA 02114 2 Division of Rheumatology, Allergy and Immunology, Massachusetts General Hospital and Harvard Medical School, Boston, MA 02114 3 Division of Pulmonary and Critical Care Medicine, Massachusetts General Hospital and Harvard Medical School, Boston, MA 02114 4 PharmAkea Inc, San Diego, CA 92130 5 Division of Rheumatology, Boston University, Boston, MA 02118 Correspondence: Flavia V. Castelino, M.D. or Andrew M. Tager, M.D., Center for Immunology and Inflammatory Diseases, Massachusetts General Hospital, 149 13th Street, Room 8301, Charlestown, MA 02129, [email protected], [email protected], Phone : (617) 643-6385 (Castelino), (617) 724-7368 (Tager) 9 7 2016 12 2016 01 12 2017 68 12 29642974 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Objective We previously implicated the lipid mediator lysophosphatidic acid (LPA) in dermal fibrosis in systemic sclerosis (SSc). Here we identify the role of the LPA-producing enzyme autotaxin (ATX), and connect the ATX-LPA and IL-6 pathways in SSc. Methods We evaluated a novel ATX inhibitor, PAT-048, on fibrosis and IL-6 expression in the bleomycin (BLM) mouse model of dermal fibrosis. We utilized SSc patient and control dermal fibroblasts to evaluate LPA induction of IL-6, and IL-6 induction of ATX. We next evaluated whether LPA-induced ATX expression is dependent on IL-6, and whether baseline IL-6 expression in SSc fibroblasts is dependent on ATX. Finally, we compared ATX and IL-6 expression in SSc and healthy subject skin. Results PAT-048 markedly attenuated BLM-induced dermal fibrosis when initiated before or after fibrosis development. LPA stimulated human dermal fibroblast IL-6 expression, and IL-6 stimulated fibroblast ATX expression, connecting the ATX-LPA and IL-6 pathways in an amplification loop. IL-6 knockdown abrogated LPA-induced ATX expression in fibroblasts, and ATX inhibition attenuated IL-6 expression in fibroblasts and the skin of BLM-challenged mice. Both ATX and IL-6 expression were increased in SSc skin, and LPA-induced IL-6 levels and IL-6-induced ATX levels were increased in SSc fibroblasts compared to controls. Conclusion ATX is required for the development and maintenance of dermal fibrosis in the BLM model and enables two major mediators of SSc fibrogenesis, LPA and IL-6, to amplify each other’s production. Our results suggest that concurrent inhibition of these two pathways may be an effective therapeutic strategy for SSc fibrosis. systemic sclerosis fibrosing disorders dermal fibroblasts Systemic sclerosis (SSc) is a potentially fatal autoimmune disease of unknown etiology, characterized by multi-organ fibrosis that is refractory to current therapies. We previously showed that signaling of the lipid mediator lysophosphatidic acid (LPA) specifically through one of its receptors, LPA1 is required for dermal fibrosis in the bleomycin (BLM) SSc model. Both genetic deletion and pharmacologic antagonism of LPA1 protected against BLM-induced dermal fibrosis [1]. In human skin, tissue injury increases LPA [2], and serum levels of arachidonoyl (20:4)-LPA are elevated in SSc patients [3]. Although safety was the primary endpoint of a recently reported Phase 2a study of SAR100842, an LPA1 receptor antagonist in SSc, and subjects were only randomized to drug or placebo for 8 weeks, clinically meaningful improvements were noted in skin fibrosis assessed by the modified Rodnan Skin Score (mRSS) and in symptoms assessed by the scleroderma-health assessment questionnaire (S-HAQ) [4]. Here we sought to identify the enzyme(s) required for LPA production in dermal fibrosis. Since LPA signals through multiple receptors, of which six are currently recognized [5], inhibition of LPA production by targeting these enzyme(s) would have the potential advantage of inhibiting pro-fibrotic effects of LPA mediated by receptors other than LPA1. Two major enzymatic pathways are described for LPA production [6]. LPA can be produced from phosphatidic acid (PA) by phospholipase A1 (PLA1) and phospholipase A2 (PLA2) family members, including Lipase H (LIPH). Alternatively, PLA1 and PLA2 can first convert phospholipids to lysophospholipids, which can then be converted to LPA by autotaxin (ATX). These two pathways appear non-redundant. LIPH produces the LPA that is required for hair growth [7], while ATX is required for embryonic vascular and nervous system development [8, 9]. Several prior investigations have suggested a role for ATX in fibrotic diseases [10, 11]. Here we further support a role for ATX in pathological fibrosis by demonstrating that ATX expression is elevated in the fibrotic skin of patients with SSc, and that ATX activity is required for BLM-induced dermal fibrosis. We also investigated the regulation of ATX expression in dermal fibrosis. ATX expression in breast cancer cells is induced by Stat3, a transcription factor downstream of IL-6 [12]. Several studies have implicated IL-6 in SSc dermal fibrosis [13, 14], and a phase II/III study of tocilizumab, a monoclonal antibody to the IL-6 receptor, was recently completed with encouraging results [15]. LPA in turn is a potent inducer of IL-6 [16]. We therefore hypothesized that ATX, LPA, and IL-6 participate in an amplification loop in SSc dermal fibrosis, in which LPA induces IL-6 expression, IL-6 in turn induces ATX expression, and ATX catalyzes further LPA generation. MATERIALS AND METHODS Study Approval Animal studies were performed according to protocols approved by the Massachusetts General Hospital (MGH) or PharmAkea Inc. IACUC. All procedures and protocols for human studies were approved by the MGH or Boston University (BU) IRB, and written informed consent was obtained from all participants prior to inclusion in the study. Animals For all mouse experiments, C57Bl/6 mice were purchased from the NCI-Frederick Mouse Repository. All experiments used age-matched, female mice at 6–8 weeks of age maintained in a specific pathogen–free environment. All experiments were performed in accordance with NIH guidelines, and protocols approved by the MGH or PharmAkea Inc. IACUC. Patient Samples Diffuse SSc skin biopsy samples were collected at MGH and BU Rheumatology clinics per approved IRBs, respectively. The patients fulfilled the 2013 American College of Rheumatology (ACR)/European League Against Rheumatism (EULAR) Systemic Sclerosis criteria. The mRSS was calculated for each subject, by grading skin thickness as 0 (none), 1 (mild), 2 (moderate) or 3 (severe) at 17 body areas for a maximum total of 51. BLM injections and skin harvests BLM (Fresenius Kabi) was dissolved in PBS at 10 μg/ml and sterilized by filtration. BLM or PBS (100 μl) was injected subcutaneously into two locations on the shaved backs of C57Bl/6 mice, once per day for 3, 7, 14 or 28 doses. Mice were then sacrificed and full thickness 6 mm dermal punch biopsies were obtained from each injection site. One skin sample was fixed in 10% formalin and embedded in paraffin for histology and IHC studies; the other was frozen immediately at −80°C for hydroxyproline and quantitative PCR (qPCR) analyses. Determination of pharmacokinetics and pharmacodynamics of PAT-048 in mouse plasma The selective ATX inhibitor, PAT-048, ((3-[6-Chloro-7-fluoro-2-methyl-1-(1-propyl-1H-pyrazol-4-yl-1H-indol-3-ylsulfanyl)]-2-fluoro-benzoic sodium salt)) (US Patent number 2013/0150326 A1) was administered to n=3 mice/time point by oral gavage (10 mg/kg in 0.5% methylcellulose) and blood was collected by cardiac puncture under anesthesia in EDTA vacutainer tubes at 0.5, 1, 2, 4, 8, 16 and 24h post-dosing. Plasma samples were prepared and stored at −80°C prior to analysis of PAT-048 concentrations by LC-MS/MS and autotaxin activity by the TOOS choline release assay. Known amounts of PAT-048 were added to mouse plasma to yield a concentration range from 0.8 to 4000 ng/ml. Plasma samples were precipitated using 70:30 acetonitrile: methanol containing an internal standard. The supernatant (150 μl) was mixed with 100 μl water then centrifuged at 4000 × g for 10 min at 4°C. The analyte mixture (10 μl) was injected using a Leap PAL autosampler. Calibration curves were constructed by plotting the peak-area ratio of analyzed peaks against known concentrations. The lower limit of quantitation was 20 ng/ml. Linear regression analysis with a 1/x2 weighting was performed. Autotaxin activity in each plasma sample was analyzed using the TOOS choline release method by adding 20 μl plasma to 75 μl LysoPLD buffer and 5 μl 4 mM 14:0 LPA (final concentration 200 μM) and incubating at 37°C for 1.5h. ATX activity was detected by measuring the liberated choline using an enzymatic photometric method described previously [17]. Plasma from non-dosed animals was used to determine maximum ATX activity. PAT-048 administration in the BLM model PAT-048 was dissolved in 0.5% methylcellulose, and a dose of 20 mg/kg per mouse, or methylcellulose alone (vehicle), was administered by oral gavage to C57Bl/6 mice, once daily for up to 28 days. PAT-048 was administered from the onset of BLM challenge in a ‘preventive’ regimen, or beginning either 7 or 14 days after the onset of BLM challenge in two ‘delayed’ regimens. For all PAT-048 regimens, BLM or PBS was injected subcutaneously for 28 consecutive days, and skin samples were obtained at the completion of the experiment as described above. Histology and dermal thickness measurement Multiple 5 μm sections of paraffin-embedded skin samples were de-paraffinized, rehydrated and stained with H&E or Masson’s trichrome stains according to the standard protocols of our laboratory [18]. Dermal thickness was determined using 20× magnification photomicrographs of H&E-stained sections by measuring the distance between the epidermal-dermal junction and the dermal-fat junction in five randomly selected sites/high power field (HPF), for ten HPF per section. Hydroxyproline Assay Hydroxyproline content was determined as a measure of skin collagen using the standard protocol of our laboratory [19]. Briefly, skin samples were homogenized in PBS and hydrolyzed overnight in 6N HCl at 120°C. A 25 μl aliquot was desiccated, re-suspended in 25 μl H2O and added to 0.5 ml of 1.4% chloramine T (Sigma), 10% n-propanol, and 0.5 M sodium acetate, pH 6.0. After 20-minute incubation at room temperature, 0.5 ml of Erlich’s solution (1M p-dimethylaminobenzaldehyde (Sigma) in 70% n-propanol, 20% perchloric acid) was added. After 15 minute incubation at 65°C, absorbance was measured at 550 nm and hydroxyproline concentration determined against a standard curve. Assay results were expressed as μg hydroxyproline/6 mm punch biopsy of skin. Immunohistochemistry (IHC) Multiple 5 μm sections of paraffin-embedded skin samples (mouse or human skin) were cut onto slides, and then de-paraffinized and rehydrated. Immunolabeling of α-smooth muscle actin (α-SMA) and phospho-Smad2 was performed with primary rabbit anti-mouse α-SMA antibody (Abcam) and primary rabbit anti-mouse phospho-Smad2 antibody (Cell Signaling). Immunolabeling of IL-6 was performed with primary rabbit anti-mouse IL-6 antibody or primary rabbit anti-goat IL-6 antibody (Cell Signaling), respectively, using the DakoCytomation system per manufacturer’s instructions. Immunolabeling of CD3 or F4/80 (Abcam) was performed with primary rabbit anti-mouse CD3 or primary rat anti-mouse F4/80, respectively. Appropriate biotinylated secondary antibodies were used, followed by detection with Vectastain avidin/biotin complex conjugated with horseradish peroxidase (Vector Laboratories) and color development with AEC (Dako). Cells positive for α-SMA, pSmad2, CD3 and F4/80 were counted in 10 randomly selected, non-overlapping high-power fields. Blinded sections were photographed with imaging software and IL-6 positive staining was quantified using ImageJ analysis software to calculate the percentage of skin surface area that stained positively for IL-6. Human dermal fibroblast isolation and culture Fibroblasts were explanted and cultured from skin biopsies taken from diffuse SSc or healthy control subjects. Dermal fibroblasts were used between passages 2 and 6. Quantitative PCR RNA was extracted from human dermal fibroblasts or mouse skin using RLT buffer or Trizol, respectively and the RNeasy Mini Kits (Qiagen, Valencia, CA, USA) according to the manufacturer’s instructions. Quantitative real-time PCR analyses of RNA were performed using Mastercycler Realplex2 (Eppendorf, Hauppauge, NY, USA). ATX activity assay ATX activity was measured by colorimetric assessment of choline production from ATX-mediated LPC cleavage, as described previously [17]. Briefly, samples were incubated with 2 mM 1-myristoyl (14:0)–LPC (Avanti Polar Lipids Inc., Alabaster, AL), in buffer containing 100 mM Tris-HCl, pH 9.0, 500 mM NaCl, 5 mM MgCl2, 30 μM CoCl2, 0.05% Triton X-100 [17, 20]. Samples were then analyzed for liberated choline colorimetrically following the addition of choline oxidase, horseradish peroxidase, and TOOS reagent (N-ethyl-N-(2-hydoroxy-3-sulfoproryl)-3-methylaniline, Sigma). Absorbance was read at 555 nm and converted to nanomoles of choline by comparison to a choline standard curve. Activity is reported as nmol/mL/minute choline. siRNA IL-6 transfection siRNA targeting human IL-6 (OnTargetPlus IL-6 SMARTPool reagent) and control siRNA (OnTargetPlus non-targeting pool) were obtained from Dharmacon and used according to the manufacturers’ instructions. Transfection into explanted human dermal fibroblasts was achieved using oligofectamine (Invitrogen). The siRNA was transfected at 10nM and 25nM. After transfection, dermal fibroblasts were stimulated with LPA 10 μM and ATX expression was evaluated by qPCR. IL-6 and ATX Quantification by ELISA ATX protein levels in mouse skin were determined using a commercially available ELISA kit (Echelon Biosciences, Salt Lake City, UT, USA) according to the manufacturer’s instructions. Human dermal fibroblasts were stimulated with LPA 10 μM or IL-6 20 ng/ml or both IL-6 and recombinant soluble IL-6 receptor, respectively over a time course of 24 hours and supernatants were collected and stored at −20°C. Protein levels of IL-6 or ATX in human samples were measured by ELISA (R&D Systems, USA) according to the manufacturer’s protocol. PAT-048 in vitro stimulation Primary dermal fibroblasts from three SSc patients were isolated and rested in serum-free media overnight. Fibroblasts were treated with PAT-048 (1 μM). Fibroblasts were harvested at 3 and 6 hours post stimulation for qPCR analysis. Statistics Statistical significance was determined by ANOVA or by two-tailed Student’s t test using GraphPad Prism software. A p value ≤ 0.05 was considered significant. For all figures with error bars, data represent the mean ± SEM. RESULTS We found that ATX mRNA (Figure 1A) and protein (Figure 1B) expression were both elevated in the dermis of BLM-injected mice compared to PBS-injected controls, with expression peaking early at Day 3 following the onset of daily BLM challenges. We next investigated the efficacy of PAT-048 (Figure 2A), a potent small molecule ATX inhibitor [21], in the BLM model. PAT-048 has an IC50 and IC90 of 20 nM and 200 nM in mouse plasma, respectively (Figure 2B). Following a single 10 mg/kg oral dose, mouse plasma concentration peaked with a Cmax of 16 μM at 30 min, and decreased to a trough of ~100 nM (Figure 2C). At this dose, PAT-048 showed 75% inhibition of ATX activity at trough (24 hours) (Figure 2D). When PAT-048 was dosed at 20 mg/kg once daily for 5 days in order to reach steady state, we observed >90% inhibition of plasma ATX activity 24 hours post-dosing (data not shown). To exceed the IC90 at trough, we administered 20 mg/kg PAT-048 once daily. We evaluated a ‘preventive’ regimen, in which PAT-048 administration began at the onset of BLM challenges, and two ‘delayed’ regimens, in which PAT-048 treatment was delayed until 7 or 14 days after the onset of BLM challenges (Figure 3A), at which time dermal hydroxyproline levels are already increased (data not shown). Preventive PAT-048 attenuated BLM-induced dermal fibrosis and collagen accumulation as demonstrated by H&E-stained (Figure 3B, left panels) and Masson’s trichrome-stained (Figure 3B, right panels) skin sections, respectively. Dermal thickness and hydroxyproline measurements demonstrated treatment efficacy in all three regimens studied (Figures 3C, 3D). Preventive PAT-048 attenuated the BLM-induced increase in dermal thickness by 55%, while delayed PAT-048 begun on day 7 or 14 attenuated the BLM-induced increase in dermal thickness by 49% or 48%, respectively (Figure 3C). Similarly, preventive PAT-048 attenuated the BLM-induced increase in hydroxyproline by 62%, while delayed PAT-048 begun on day 7 or 14 attenuated the BLM-induced increase in hydroxyproline by 45% and 43%, respectively (Figure 3D). The efficacy of the delayed regimens suggest an ongoing requirement for the ATX-LPA pathway in the progression of pathologic dermal fibrosis, and suggest that targeting this pathway may be an effective therapeutic strategy in fibrotic diseases of the skin even when treatment is initiated after the onset of established fibrosis. Dermal fibrosis in SSc patients is characterized by the accumulation of myofibroblasts [22, 23], which is recapitulated in the BLM model. Myofibroblast differentiation is characterized by the acquisition of features of smooth muscle cells, such as the expression of α-smooth muscle actin (α-SMA). We therefore investigated whether ATX inhibition with PAT-048 also reduced the dermal accumulation of α-SMA+ myofibroblasts in the BLM model. Preventive PAT-048 treatment significantly attenuated the BLM-induced increase in α-SMA+ cells, by 45% (Supplemental Figures 1A, B). TGF-β is considered to be the major cytokine directly promoting myofibroblast development [24]. We therefore investigated whether the PAT-048-induced reduction in skin myofibroblasts was associated with reduced dermal activation of the TGF-β pathway, by enumerating dermal cells that demonstrated nuclear Smad2 phosphorylation. Preventive PAT-048 treatment of BLM-challenged mice produced a trend toward reduced numbers of dermal pSmad2+ cells (of 30%) compared to vehicle-treated mice, but this reduction was not statistically significant (Supplemental Figures 1C, D). We next determined whether ATX expression is driven by an ATX-LPA-IL-6 amplification loop, as we hypothesized, by investigating human dermal fibroblasts as a potential source of ATX in SSc dermal fibrosis. Stimulation of dermal fibroblasts with LPA significantly induced IL-6 mRNA expression, which peaked at 6 hours post-stimulation (Figure 4A). IL-6 significantly increased the ATX activity that was present in dermal fibroblast culture supernatants, consistent with fibroblast secretion of ATX, which also peaked at 6 hours post-stimulation (Figure 4B). At this 6-hour time point, IL-6 induced fibroblast ATX secretion, as reflected by ATX activity levels in the culture supernatants, in a dose-dependent manner (Figure 4C). We next investigated whether LPA could induce ATX expression, and if so, whether LPA-induced ATX expression was dependent on its ability to induce IL-6, as we hypothesized. LPA significantly induced dermal fibroblast ATX mRNA expression, and this ATX expression was indeed abrogated by siRNA knockdown of IL-6 (Figure 4D). LPA also increased the expression of ATX in the supernatants of dermal fibroblasts in a dose-dependent manner (Figure 4E). Lastly, IL-6 and LPA concurrently had synergistic effects on ATX expression in dermal fibroblasts (Figure 4F). Taken together, our in vitro studies demonstrate a cyclic relationship between LPA, IL-6 and ATX production by dermal fibroblasts, in which LPA induces IL-6 expression, IL-6 in turn induces ATX expression, and ATX catalyzes further LPA generation. If an ATX-LPA-IL-6 amplification loop is similarly present in dermal fibrosis in vivo, we hypothesized that ATX inhibition, by inhibiting LPA production, would significantly inhibit induction of IL-6 in the BLM mouse model. Immunostaining of skin sections with anti-IL-6 antibody demonstrated increased IL-6 expression in BLM-induced dermal fibrosis that was attenuated by preventive PAT-048 treatment (Figure 5A). Multiple cell types appeared to stain positively for IL-6, consistent with previous descriptions of skin IL-6 immunostaining in SSc patients and in mouse models of skin injury [25–27]. These prior investigations reported positive IL-6 staining for numerous cell types, of both infiltrating leukocyte and resident cell lineages, including neutrophils, macrophages, keratinocytes, endothelial cells, and fibroblasts. Quantitative analyses of dermal IL-6 immunostaining in our model demonstrated that the percentage of dermal area with positive staining increased from 12 to 60% in PBS- vs. BLM-challenged mice that were treated with vehicle (Figure 5A). In contrast, percentages of dermal area staining positively for IL-6 increased only from 14% to 20% in PBS- vs. BLM-challenged mice that were treated with PAT-048. As previously described [28], we also observed that BLM challenge induced dermal IL-6 mRNA expression. As determined by QPCR, BLM-induced increases in skin IL-6 mRNA expression were significantly decreased, by 65%, in mice treated with preventive PAT-048 (Figure 5B). Experiments using IL-6-deficient mice have demonstrated that IL-6 is critically required for chemokine production and leukocyte recruitment into inflamed or injured skin, through IL-6-induced endothelial cell chemokine expression [29]. We therefore investigated whether PAT-048-induced reductions in skin IL-6 expression in the BLM model were associated with reduced leukocyte accumulation, by performing immunostaining with anti-CD3 and F4/80 antibodies to identify T cells and macrophages, respectively. Preventive PAT-048 attenuated BLM-induced CD3+ T cell accumulation in the skin (Figure 5C). Quantification of CD3+ cells in the skin of PBS- or BLM-challenged mice treated with PAT-048 or vehicle indicated that ATX inhibition reduced BLM-induced T cell accumulation by 44% (Figure 5D). In contrast, there were no significant differences in skin F4/80+ macrophages between PAT-048- and vehicle-treated groups (data not shown). The attenuation of BLM-induced IL-6 expression and T cell accumulation we observed with ATX inhibition suggests that the cyclic relationship we observed between ATX, LPA and IL-6 production in vitro also contributes to dermal fibrosis in the BLM model in vivo. To begin to investigate whether an ATX-LPA-IL-6 amplification loop could also contribute to human SSc fibrosis, we first compared ATX and IL-6 expression in the skin of patients with diffuse SSc and control subjects. We then investigated whether components of the amplification loop we observed in human dermal fibroblasts are increased in dermal fibroblasts from patients with diffuse SSc compared with fibroblasts from controls. We assessed ATX and IL-6 mRNA expression in skin biopsies from 7 age-matched diffuse SSc patients and 5 healthy control subjects. The SSc subjects had a mean age of 52.4 years, a disease duration of less than 3 years, and moderate skin involvement with a mean mRSS of 27 out of 51 (the highest possible value). The healthy controls had a mean age of 49.2 years. ATX mRNA expression was significantly increased 2.6-fold in SSc skin compared to controls (Figure 6A). IL-6 protein expression, assessed as the percentage of dermal area staining positively for IL-6 in IHC analyses, was also elevated in the dermis of SSc patients (27% IL-6 positive staining) compared to controls (2% IL-6 positive staining) (Figure 6B), consistent with prior studies [25]. We then compared the ability of LPA to induce IL-6 expression, and IL-6 to induce ATX expression in SSc and control dermal fibroblasts. SSc dermal fibroblasts produced significantly more IL-6 in response to LPA than did control fibroblasts at 6 hours post-stimulation (Figure 6C, left panel). Similarly, SSc dermal fibroblasts produced significantly more ATX in response to IL-6 than control fibroblasts at 12 and 24 hours post-stimulation (Figure 6C, right panel). The augmentation of ATX expression in response to IL-6 in SSc dermal fibroblasts was more prolonged than the augmentation of IL-6 expression in response to LPA in these cells, suggesting that differences in IL-6 signaling pathways may be more pronounced in SSc dermal fibroblasts than differences in LPA signaling pathways. To determine the contribution of ATX-produced LPA to IL-6 production by SSc dermal fibroblasts at baseline, we treated SSc fibroblasts with PAT-048 in vitro and assessed IL-6 mRNA expression. As shown in Figure 6D, ATX inhibition with PAT-048 reduced SSc fibroblast IL-6 mRNA expression significantly at 3 hours, and further at 6 hours. Taken together, these data suggest that an ATX-LPA-IL-6 amplification loop is operative in dermal fibroblasts and augmented in fibroblasts from SSc patients compared with control subjects. As noted above, multiple cell types in addition to fibroblasts stain positively for IL-6 expression in our and prior IHC analyses of scleroderma dermal fibrosis [25, 26]. Whether IL-6 production in other cell types is driven by the same ATX-LPA-IL-6 amplification loop that we have described in dermal fibroblasts will be the subject of future studies. Previous studies have shown that IL-6 signaling in fibroblasts either requires, or is augmented by the soluble form of the IL-6 receptor (sIL-6R) [25, 30]. IL-6 signaling initiated by complexes of this cytokine and its soluble receptor is known as trans-signaling [31, 32] and mediates or contributes to IL-6 signaling in cells with no or low expression of membrane-bound IL-6R. To evaluate whether IL-6-induced ATX expression is mediated by IL-6 trans-signaling in dermal fibroblasts, we compared ATX protein levels induced in SSc and control human dermal fibroblasts by IL-6 alone with that induced by IL-6 plus sIL-6R. The addition of sIL-6R produced a significant increase in ATX expression induced by IL-6 in control fibroblasts and a trend toward an increase in SSc fibroblasts (Supplemental Figure 2), suggesting IL-6 trans-signaling may contribute to the ATX-LPA-IL-6 amplification loop that we have observed in dermal fibroblasts. DISCUSSION Our data reveal an essential role for ATX in mediating dermal fibrosis. ATX expression is increased in the fibrotic skin of diffuse SSc patients and in the BLM SSc dermal fibrosis model. Pharmacological inhibition of ATX reduced dermal fibrosis in the animal model when initiated before or after established fibrosis. We additionally identify a novel amplification loop connecting ATX and the LPA it produces to the pro-inflammatory cytokine IL-6. In this amplification loop, LPA drives IL-6 expression and IL-6 in turn induces ATX expression, leading to increased LPA production. Together with prior evidence implicating LPA and IL-6 in SSc [1, 4, 13–15], our results identify the ATX-LPA-IL-6 axis as a pathway fundamental to the development and progression of SSc fibrosis. As noted, two major enzymatic pathways produce LPA, one involving ATX and the other involving LIPH. In skin, ATX is expressed in hair follicles and is one of the most highly expressed genes in dermal papilla [33]. Despite high ATX expression, hair follicle morphogenesis is unaffected by genetic deletion of ATX in dermal papilla, demonstrating that the ATX pathway of LPA production is dispensable in this dermal appendage. LIPH upregulation is noted in ATX-deficient dermal papilla precursors, suggesting the LIPH pathway of LPA production can compensate for the loss of ATX and support normal hair growth. In contrast, an absolute requirement for the LIPH pathway for hair growth is suggested by the recent demonstration that LIPH mutations in humans cause an inherited form of hypotrichosis simplex, characterized by diffuse hair loss [34]. As opposed to ATX expression being redundant for physiological hair growth in the dermis, our results demonstrate a requirement for ATX activity for the development of pathologic dermal fibrosis in the BLM SSc model. Consistent with ATX contributing to dermal fibrosis in human SSc, ATX expression was elevated in the skin of SSc patients. The cellular source(s) of dermal ATX during fibrogenesis remain to be identified. We demonstrate evidence of ATX expression by human dermal fibroblasts and increased ATX expression in SSc dermal fibroblasts, suggesting that these cells may be an important source of ATX in SSc dermal fibrosis. ATX has also been noted to be secreted by adipocytes [35], and mice deficient for ATX expression specifically in adipocytes have demonstrated that ATX secreted by these cells is responsible for 38% of plasma LPA levels [36]. Loss of dermal adipose tissue is a characteristic feature of the fibrotic skin of SSc patients, and this loss is recapitulated in the BLM model of dermal fibrosis [37]. After the initiation of BLM injections, dermal adipocyte numbers are preserved at day 3, when we observed maximal dermal ATX levels, but are significantly reduced at day 5, and remain depressed through day 21 [37], the same period in which we observed a fall in dermal ATX levels. These observations suggest that adipocytes could be an important source of ATX at early time points during the development of dermal fibrosis as well. Together with our prior demonstration of a requirement for LPA signaling through LPA1 in the BLM model [1], our demonstration that BLM-induced dermal fibrosis also requires the LPA-producing activity of ATX firmly establishes an important pro-fibrotic role for the LPA pathway in this model. Patients with SSc have been reproducibly classified into molecular subsets based on dermal gene expression patterns, with different signaling pathways underlying each subset [38]. Of these subsets, the BLM model of dermal fibrosis has recently been shown to share gene expression patterns with the inflammatory subset of SSc patients [39]; our findings in this mouse model may consequently be most relevant to this subset of patients. Despite this molecular heterogeneity of SSc, and our human sample sizes being relatively small, our finding of increased ATX expression in the skin of SSc patients, together with recent evidence that an LPA1 receptor antagonist can reduce dermal fibrosis in SSc patients [4], suggests the LPA pathway may importantly contribute to human SSc dermal fibrosis as well. The ability of an ATX inhibitor such as PAT-048 to attenuate the progression of dermal fibrosis when treatment is initiated after fibrosis is already established suggests that ATX inhibition may also be effective in the treatment of SSc patients with already existing fibrosis. Multiple other molecular pathways are implicated in SSc fibrogenesis, however, raising the question of whether therapies targeting a single pathway will ultimately be able to reverse fibrosis in this disease. Such therapies may be effective in patients whose fibrosis is driven predominantly by the targeted pathway, if such patients could be identified. Otherwise, combination therapy targeting different pathways implicated in SSc may be required to effectively treat fibrosis in this disease, as suggested for idiopathic pulmonary fibrosis [40]. We have connected two major pro-fibrotic pathways, those of LPA and IL-6, in dermal fibrosis through IL-6-induced ATX expression. In so doing, we have identified a potential amplification loop in which LPA induces IL-6 expression, and IL-6, in turn, induces ATX expression, which catalyzes additional LPA production. Our demonstration of this amplification loop, and our findings that indicate this loop is augmented in SSc compared to control dermal fibroblasts, suggest that combined therapies targeting the ATX/LPA and IL-6 pathways may be a particularly effective future therapeutic strategy for SSc. Supplementary Material Supp Fig S1 Supplemental Figure 1 Skin accumulation of myofibroblasts and activation of TGF-β signaling are attenuated by PAT-048 in BLM-challenged mice. (A) Representative images of α-SMA immunostaining of skin of PBS- or BLM-challenged mice treated with vehicle or preventive PAT-048 for 28 days. Magnification 200x. (B) Quantification of α-SMA+ cells, indicative of myofibroblasts, in the dermis. p ≤ 0.05 by ANOVA. (C) Representative images of pSmad2 immunostaining of skin of PBS- or BLM-challenged mice treated with vehicle or preventive PAT-048 regimen for 28 days. Magnification 200x. (D) Quantification of pSmad2+ cells, indicative of cells responding to active TGF-β, in the dermis. Data not statistically significant by ANOVA. Supp Fig S2 Supplemental Figure 2 Contribution of trans-signaling to IL-6-induced ATX expression in dermal fibroblasts. ATX expression in SSc and control human dermal fibroblasts induced by IL-6 alone and IL-6 plus soluble IL-6 receptor at 6 hours. Data represent mean ± SEM of triplicate measurements in each group. p ≤ 0.05 by ANOVA. This work was supported by NIH K08-AR062592 and Scleroderma Foundation Grant to F.V.C, and a Scleroderma Research Foundation Grant and NIH R01-HL095732 to A.M.T. Figure 1 ATX is elevated early in the BLM mouse model (A) QPCR analysis of ATX mRNA and (B) ELISA measurements of ATX protein expression in the dermis of C57Bl/6 mice after 3 to 28 daily BLM or PBS challenges. Data are representative of three independent experiments with n = 5 mice per group. p ≤ 0.01, p ≤ 0.001 by two-tailed Student’s t test. Figure 2 Structure and pharmacokinetics of PAT-048 (A) Chemical structure of PAT-048 (3-[6-Chloro-7-fluoro-2-methyl-1-(1-propyl-1H-pyrazol-4-yl-1H-indol-3-ylsulfanyl]-2-fluoro-benzoic sodium salt). (B) Concentration-response curve of PAT-048 inhibition of ATX activity in mouse plasma. (C) PAT-048 plasma concentrations in mice over time after a single oral dose of 10 mg/kg. (D) Inhibition of ATX activity in mice after a single PAT-048 oral dose of 10 mg/kg. Figure 3 ATX is required for dermal fibrosis in the SSc BLM mouse model (A) Mice treated with vehicle or PAT-048 in a ‘preventive’ regimen from the onset of 28 daily PBS or BLM challenges, or two ‘delayed’ regimens begun at 7 (Delayed PAT-048 #1) or 14 days (Delayed PAT-048 #2) after PBS or BLM challenge (B) H&E (left panels) and Masson’s trichrome (right panels) staining of the skin of mice treated with vehicle or preventive PAT-048. Arrows depict dermal thickness. Magnification 100x; scale bar: 100 μm. (C) Dermal thickness measured in 10 high power fields (HPF)/skin sample of mice treated with vehicle or PAT-048 in a preventive regimen, or two ‘delayed’ regimens. n = 5 mice per group. Preventive group data are combined from three independent experiments; ‘delayed’ group data are from two experiments. (D) Skin hydroxyproline content from the same experiments in (C). p ≤ 0.05, p ≤ 0.01, p ≤ 0.001 by ANOVA. Figure 4 Amplification loop of LPA, IL-6 and ATX production by human dermal fibroblasts (A) Human dermal fibroblasts (n = 3 subjects) stimulated with LPA (10 μM) demonstrated increased IL-6 mRNA expression peaking at 6 hours. (B) Human dermal fibroblasts (n = 3 subjects) stimulated with IL-6 (20 ng/ml) demonstrated increased ATX activity, peaking at 6 hours. (C) IL-6 induced ATX activity in a dose-dependent manner. (D) Transfection of human dermal fibroblasts with siRNA targeting IL-6 abrogated LPA (10 μM)-induced ATX mRNA expression. Data represent mean±SEM of triplicate measurements. (E) LPA induced ATX activity in a dose-dependent manner. (F) IL-6 (20 ng/ml) and LPA (10 μM) had synergistic effects on ATX mRNA expression. p≤0.05, p≤0.01, p≤0.001 by ANOVA. Figure 5 Skin IL-6 expression and T cell accumulation are attenuated by PAT-048 in BLM-challenged mice (A) Left panel: representative images of IL-6 immunostaining of skin of PBS- or BLM-challenged mice treated with vehicle or preventive PAT-048 for 28 days. Magnification 50× (top), 200× (bottom). Right panel: Quantification of IL-6 staining using ImageJ analysis software. (B) IL-6 mRNA expression in the skin of PBS- or BLM-challenged mice (n=5 mice per group). (C) Representative images of CD3 immunostaining of skin of PBS- or BLM-challenged mice treated with vehicle or preventive PAT-048 for 28 days. Magnification 50× (top), 200× (bottom). (D) Quantification of number of CD3+ cells per HPF in 10 HPF. p ≤ 0.05, p ≤ 0.01 by ANOVA. Figure 6 The ATX-LPA-IL-6 amplification loop is augmented in SSc (A) ATX mRNA expression in skin of 7 SSc patients and 5 healthy controls. (B) Left panel: representative images of IL-6 protein expression in skin of SSc patients and healthy controls (n = 3 subjects per group). Magnification 50× (top), 200× (bottom). Right panel: quantification of IL-6 immunostaining using ImageJ analysis software. Data represent mean ± SEM of triplicate measurements in each group. p ≤ 0.01 by ANOVA. (C) Left and right panels: LPA (10 μM)-induced IL-6 protein expression, and IL-6 (20 ng/ml)-induced ATX protein expression, respectively, as measured by ELISA in the supernatants of SSc and control human dermal fibroblast cultures (n = 3 subjects per group). (D) IL-6 mRNA expression in SSc dermal fibroblasts at baseline or after 3 or 6 hours of PAT-048 (1 uM) treatment. Data represent mean ± SEM of triplicate measurements in each group. p ≤ 0.05, *p ≤ 0.01 by ANOVA. Conflict of interest: Drs. Bain and Goulet receive income from and own equity in PharmAkea Inc. Dr. Tager is a member of the scientific advisory board of PharmAkea Inc. 1 Castelino FV Seiders J Bain G Brooks SF King CD Swaney JS Lorrain DS Chun J Luster AD Tager AM Amelioration of dermal fibrosis by genetic deletion or pharmacologic antagonism of lysophosphatidic acid receptor 1 in a mouse model of scleroderma Arthritis Rheum 2011 63 5 1405 1415 21305523 2 Mazereeuw-Hautier J Gres S Fanguin M Cariven C Fauvel J Perret B Chap H Salles JP Saulnier-Blache JS Production of lysophosphatidic acid in blister fluid: involvement of a lysophospholipase D activity J Invest Dermatol 2005 125 3 421 427 16117781 3 Tokumura A Carbone LD Yoshioka Y Morishige J Kikuchi M Postlethwaite A Watsky MA Elevated serum levels of arachidonoyl-lysophosphatidic acid and sphingosine 1-phosphate in systemic sclerosis Int J Med Sci 2009 6 4 168 176 19521548 4 Khanna D Denton C Jagerschmidt A Jasson M Distler O Allanore Y SAR100842, an Antagonist of Lysophaphatidic Acid Receptor 1, As a Potential Treatment for Patients with Systemic Sclerosis: Results from a Phase 2a Study Arthritis and Rheumatology Abstract Supplement 2014 66 10 S386 5 Choi JW Herr DR Noguchi K Yung YC Lee CW Mutoh T Lin ME Teo ST Park KE Mosley AN LPA receptors: subtypes and biological actions Annu Rev Pharmacol Toxicol 2010 50 157 186 20055701 6 Aoki J Inoue A Okudaira S Two pathways for lysophosphatidic acid production Biochim Biophys Acta 2008 1781 9 513 518 18621144 7 Shimomura Y Congenital hair loss disorders: rare, but not too rare J Dermatol 2012 39 1 3 10 22044263 8 Tanaka M Okudaira S Kishi Y Ohkawa R Iseki S Ota M Noji S Yatomi Y Aoki J Arai H Autotaxin Stabilizes Blood Vessels and Is Required for Embryonic Vasculature by Producing Lysophosphatidic Acid J Biol Chem 2006 281 35 25822 25830 16829511 9 van Meeteren LA Ruurs P Stortelers C Bouwman P van Rooijen MA Pradere JP Pettit TR Wakelam MJO Saulnier-Blache JS Mummery CL Autotaxin, a Secreted Lysophospholipase D, Is Essential for Blood Vessel Formation during Development Mol Cell Biol 2006 26 13 5015 5022 16782887 10 Watanabe N Ikeda H Nakamura K Ohkawa R Kume Y Aoki J Hama K Okudaira S Tanaka M Tomiya T Both plasma lysophosphatidic acid and serum autotaxin levels are increased in chronic hepatitis C Journal of clinical gastroenterology 2007 41 6 616 623 17577119 11 Oikonomou N Mouratis MA Tzouvelekis A Kaffe E Valavanis C Vilaras G Karameris A Prestwich GD Bouros D Aidinis V Pulmonary autotaxin expression contributes to the pathogenesis of pulmonary fibrosis American journal of respiratory cell and molecular biology 2012 47 5 566 574 22744859 12 Azare J Doane A Leslie K Chang Q Berishaj M Nnoli J Mark K Al-Ahmadie H Gerald W Hassimi M Stat3 mediates expression of autotaxin in breast cancer PLoS One 2011 6 11 e27851 22140473 13 Feghali CA Bost KL Boulware DW Levy LS Mechanisms of pathogenesis in scleroderma. I. Overproduction of interleukin 6 by fibroblasts cultured from affected skin sites of patients with scleroderma J Rheumatol 1992 19 8 1207 1211 1404155 14 Hasegawa M Fujimoto M Matsushita T Hamaguchi Y Takehara K Sato S Serum chemokine and cytokine levels as indicators of disease activity in patients with systemic sclerosis Clinical rheumatology 2011 30 2 231 237 21049277 15 Khanna D Denton C van Laar JM Jahreis A Cheng S Spotswood H Siegel J Furst DE on behalf of FaSScinate Trial Safety and Efficacy of Subcutaneous Tocilizumab in Adults with Systemic Sclerosis: Week 24 data from a Phase 2/3 Trial Arthritis and Rheumatology Abstract Supplement 2014 66 10 S386 16 Boucharaba A Serre C-M Gres S Saulnier-Blache JS Bordet J-C Guglielmi J Clezardin P Peyruchaud O Platelet-derived lysophosphatidic acid supports the progression of osteolytic bone metastases in breast cancer J Clin Invest 2004 114 12 1714 1725 15599396 17 Nakamura K Ohkawa R Okubo S Tozuka M Okada M Aoki S Aoki J Arai H Ikeda H Yatomi Y Measurement of lysophospholipase D/autotaxin activity in human serum samples Clinical biochemistry 2007 40 3–4 274 277 17222397 18 Tager AM Lacamera P Shea BS Campanella GS Selman M Zhao Z Polosukhin V Wain J Karimi-Shah BA Kim ND The lysophosphatidic acid receptor LPA(1) links pulmonary fibrosis to lung injury by mediating fibroblast recruitment and vascular leak Nat Med 2008 14 1 45 54 18066075 19 Tager AM Kradin RL LaCamera P Bercury SD Campanella GS Leary CP Polosukhin V Zhao LH Sakamoto H Blackwell TS Inhibition of pulmonary fibrosis by the chemokine IP-10/CXCL10 American journal of respiratory cell and molecular biology 2004 31 4 395 404 15205180 20 Umezu-Goto M Kishi Y Taira A Hama K Dohmae N Takio K Yamori T Mills GB Inoue K Aoki J Autotaxin has lysophospholipase D activity leading to tumor cell growth and motility by lysophosphatidic acid production J Cell Biol 2002 158 2 227 233 12119361 21 Amira Pharmaceuticals Patent Autotaxin Inhibitors and Uses Thereof vol. N/A San Diego, USA World Intellectual Property Organization 2012 22 Sappino AP Masouye I Saurat JH Gabbiani G Smooth muscle differentiation in scleroderma fibroblastic cells Am J Pathol 1990 137 3 585 591 1698026 23 Kissin EY Merkel PA Lafyatis R Myofibroblasts and hyalinized collagen as markers of skin disease in systemic sclerosis Arthritis Rheum 2006 54 11 3655 3660 17075814 24 Hinz B Phan SH Thannickal VJ Galli A Bochaton-Piallat ML Gabbiani G The myofibroblast: one function, multiple origins Am J Pathol 2007 170 6 1807 1816 17525249 25 Khan K Xu S Nihtyanova S Derrett-Smith E Abraham D Denton CP Ong VH Clinical and pathological significance of interleukin 6 overexpression in systemic sclerosis Ann Rheum Dis 2012 71 7 1235 1242 22586157 26 Desallais L Avouac J Frechet M Elhai M Ratsimandresy R Montes M Mouhsine H Do H Zagury JF Allanore Y Targeting IL-6 by both passive or active immunization strategies prevents bleomycin-induced skin fibrosis Arthritis research & therapy 2014 16 4 R157 25059342 27 Lin ZQ Kondo T Ishida Y Takayasu T Mukaida N Essential involvement of IL-6 in the skin wound-healing process as evidenced by delayed wound healing in IL-6-deficient mice J Leukoc Biol 2003 73 6 713 721 12773503 28 Kitaba S Murota H Terao M Azukizawa H Terabe F Shima Y Fujimoto M Tanaka T Naka T Kishimoto T Blockade of interleukin-6 receptor alleviates disease in mouse model of scleroderma Am J Pathol 2012 180 1 165 176 22062222 29 Romano M Sironi M Toniatti C Polentarutti N Fruscella P Ghezzi P Faggioni R Luini W van Hinsbergh V Sozzani S Role of IL-6 and its soluble receptor in induction of chemokines and leukocyte recruitment Immunity 1997 6 3 315 325 9075932 30 Le TT Karmouty-Quintana H Melicoff E Weng T Chen NY Pedroza M Zhou Y Davies J Philip K Molina J Blockade of IL-6 Trans signaling attenuates pulmonary fibrosis J Immunol 2014 193 7 3755 3768 25172494 31 Taga T Hibi M Hirata Y Yamasaki K Yasukawa K Matsuda T Hirano T Kishimoto T Interleukin-6 triggers the association of its receptor with a possible signal transducer, gp130 Cell 1989 58 3 573 581 2788034 32 Chalaris A Garbers C Rabe B Rose-John S Scheller J The soluble Interleukin 6 receptor: generation and role in inflammation and cancer Eur J Cell Biol 2011 90 6–7 484 494 21145125 33 Grisanti L Rezza A Clavel C Sennett R Rendl M Enpp2/Autotaxin in dermal papilla precursors is dispensable for hair follicle morphogenesis J Invest Dermatol 2013 133 10 2332 2339 23677168 34 Pasternack SM von Kugelgen I Muller M Oji V Traupe H Sprecher E Nothen MM Janecke AR Betz RC In vitro analysis of LIPH mutations causing hypotrichosis simplex: evidence confirming the role of lipase H and lysophosphatidic acid in hair growth J Invest Dermatol 2009 129 12 2772 2776 19536142 35 Rancoule C Dusaulcy R Treguer K Gres S Attane C Saulnier-Blache JS Involvement of autotaxin/lysophosphatidic acid signaling in obesity and impaired glucose homeostasis Biochimie 2014 96 140 143 23639740 36 Dusaulcy R Rancoule C Gres S Wanecq E Colom A Guigne C van Meeteren LA Moolenaar WH Valet P Saulnier-Blache JS Adipose-specific disruption of autotaxin enhances nutritional fattening and reduces plasma lysophosphatidic acid J Lipid Res 2011 52 6 1247 1255 21421848 37 Marangoni RG Korman B Wei J Wood TA Graham L Whitfield ML Scherer PE Tourtellotte WG Varga J Myofibroblasts in cutaneous fibrosis originate from adiponectin-positive intradermal progenitors Arthritis Rheumatol 2014 38 Milano A Pendergrass SA Sargent JL George LK McCalmont TH Connolly MK Whitfield ML Molecular subsets in the gene expression signatures of scleroderma skin PLoS One 2008 3 7 e2696 18648520 39 Sargent JL Li Z Aliprantis AO Greenblatt M Lemaire R Wu MH Wei J Taroni J Harris A Long KB Interspecies comparative genomics identifies optimal mouse models of systemic sclerosis Arthritis Rheumatol 2016 40 Wuyts WA Antoniou KM Borensztajn K Costabel U Cottin V Crestani B Grutters JC Maher TM Poletti V Richeldi L Combination therapy: the future of management for idiopathic pulmonary fibrosis? Lancet Respir Med 2014 2 11 933 942 25439569
PMC005xxxxxx/PMC5125862.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 8006324 5984 Mol Biochem Parasitol Mol. Biochem. Parasitol. Molecular and biochemical parasitology 0166-6851 1872-9428 27496178 5125862 10.1016/j.molbiopara.2016.08.001 NIHMS813436 Article A unified approach towards Trypanosoma brucei functional genomics using Gibson assembly McAllaster Michael R. 1¶ Sinclair-Davis Amy N. 1 Hilton Nicholas A. 1 de Graffenried Christopher L. 1* 1 Department of Molecular Microbiology and Immunology, Brown University, Providence, RI, 02912 * To whom correspondence should be addressed: Phone: +1 (401) 863-9775, [email protected] ¶ Current address: Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, MO 63110. 30 8 2016 3 8 2016 Nov-Dec 2016 01 11 2017 210 1-2 1321 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Trypanosoma brucei is the causative agent of human African trypanosomiasis and nagana in cattle. Recent advances in high throughput phenotypic and interaction screens have identified a wealth of novel candidate proteins for diverse functions such as drug resistance, life cycle progression, and cytoskeletal biogenesis. Characterization of these proteins will allow a more mechanistic understanding of the biology of this important pathogen and could identify novel drug targets. However, methods for rapidly validating and prioritizing these potential targets are still being developed. While gene tagging via homologous recombination and RNA interference are available in T. brucei, a general strategy for creating the most effective constructs for these approaches is lacking. Here, we adapt Gibson assembly, a one-step isothermal process that rapidly assembles multiple DNA segments in a single reaction, to create endogenous tagging, overexpression, and long hairpin RNAi constructs that are compatible with well-established T. brucei vectors. The generality of the Gibson approach has several advantages over current methodologies and substantially increases the speed and ease with which these constructs can be assembled. Graphical Abstract Gibson assembly is a single-step process for rapidly assembling constructs from multiple DNA segments. We adapt the approach to generate the essential plasmids necessary for T. brucei functional genomics. Gibson assembly RNAi T. brucei 1. Introduction Trypanosoma brucei is a protist parasite that causes enormous harm to both humans and livestock in Sub-Saharan Africa [1]. The parasite has been the focus of intense research efforts, originally focusing on morphological analysis and observational studies to understand the parasite’s life cycle and how it evades the host immune response [2,3]. The advent of molecular biology has opened up a host of powerful tools for studying trypanosomes, including gene tagging and inducible expression using the tetracycline suppressor system [4,5]. Unlike several closely-related kinetoplastid parasites, T. brucei has retained the machinery necessary for RNA interference, which allows straightforward access to loss of function experiments [6,7]. These tools have been used to develop high-throughput approaches for gene analysis using inducible whole-genome RNAi and next generation sequencing (RIT-seq) [8]. This method has been used in a host of screens to identify genes involved in the parasite’s transition from the mammalian-infectious bloodstream form to the insect-resident (procyclic) form and genes that are essential for resistance to several trypanocidal agents [9,10]. Protein-protein interaction approaches such as in vivo biotinylation using a mutant of the bacterial biotin ligase BirA (BioID) and high-throughput GFP tagging have begun to uncover components of many enigmatic cytoskeletal structures that are essential for cell polarity and motility [11–14]. Advances in genomic and proteomic methods have also identified key components of different cellular pathways that merit further study [15–19]. The initial characterization of T. brucei proteins relies on gene tagging for localization and tetracycline-inducible RNAi to establish function. There are numerous approaches for gene tagging, including constitutive overexpression, tetracycline-inducible overexpression, and tagging of the endogenous locus (Figure 1A, 1B) [5,11,20–24]. Epitope tags such as the HA and Ty1 tags are commonly used, along with GFP for imaging of live cells [11,25]. Tetracycline-inducible RNAi plasmids originally employed flanking T7 promoters and tetracycline suppressors to produce double-stranded RNA suitable for triggering mRNA degradation, although single-stranded long hairpin RNAs (lhRNAs) are now favored due to lower levels of background expression and improved basepairing [26–30]. While these approaches are effective, there are significant shortcomings that can decrease throughput when a large number of genes need to be assessed. Most available tagging and overexpression plasmids lack multiple restriction sites for cloning, which can require blunting or other workarounds. Endogenous tagging methods either require multiple assembly steps to clone portions of the gene of interest to direct the tag or use smaller targeting segments such as overhangs in primers, which can make PCR difficult and decrease homologous recombination efficiency, especially if the second allele is being targeted [21,31]. For the production of RNAi hairpins, intermediate steps are frequently necessary to assemble the hairpin prior to insertion into the final tetracycline-inducible plasmid [28,29]. A rapid, general strategy for making these constructs would allow more genes to be studied in a more cost-effective manner. Gibson assembly is a single-step isothermal reaction that rapidly assembles segments of DNA with overlapping termini (Figure 1C) [32]. The method employs a mixture of three enzymes: a 5′ exonuclease that exposes overhanging sequence for specific annealing of the complimentary DNA segments, a DNA polymerase that fills in the overhangs, and a DNA ligase that links the segments. The method requires 15–20 base pairs of homologous sequence, so specificity can easily be encoded within non-annealing overhangs in PCR primers. Gibson-compatible segments can also be generated by restriction digest, which is especially useful for including plasmid backbones in the assembly reaction. Gibson assembly is remarkably robust and has been used to assemble whole genomes from small DNA segments, showing the generality of the method [33]. In this work, we show how the Gibson approach can be used to create many of the constructs used for gene analysis in T. brucei, including tetracycline-inducible expression, endogenous replacement, and RNAi (Figure 1D). The Gibson approach removes many obstacles such as restriction enzyme incompatibilities, the need for intermediate ligation steps, and the limited size of targeting segments, allowing the rapid assembly of the best-suited constructs for probing gene function. 2. Materials and Methods 2.1 Molecular biology Enzymes used in this study were from New England Biolabs (Ipswich, MA) and chemicals from Thermo Fisher Scientific. PCR was performed with Q5 High Fidelity Polymerase in Q5 buffer (NEB). Plasmids were prepared for transfection with GeneJET Plasmid Midiprep Kit (LifeTechnologies). PCR primers used in this work are included in Supplemental Figure 1, while the sequences of all the constructs are in Supplemental Figure 2. 2.2 Gibson Assembly Gibson assembly, also known as isothermal chew-back-anneal assembly, was conducted as described [32]. Briefly, typically 10 μg of vector backbone was digested to generate linearized vector for assembly reactions, followed by treatment with 10 units of calf intestinal alkaline phosphatase (NEB) for 1 h at 37 °C. PCR was conducted using Q5 polymerase (NEB) with the manufacturer’s standard conditions to generate fragments bearing 20 bp overhangs of desired homology to the flanking region. Both vector fragments and PCR products were purified by gel purification (Zymoclean Gel DNA Recovery Kit, Zymogen) prior to assembly. Fragments were combined in either commercially available (New England Biolabs Gibson Assembly Master Mix) or homemade Gibson Assembly Master Mix. The homemade Master Mix was prepared by combining 699 μL water, 320 μL 5x isothermal reaction buffer (500 mM Tris-Cl, pH 7.5, 250 mg/mL PEG-8000, 50 mM MgCl2, 50 mM DTT, 1 mM each of four dNTPs, 5 mM beta-NAD), 0.64 μL T5 Exonuclease (Epicentre, 10 U/μL), 20 μL Phusion DNA polymerase (NEB, 2 U/μL) and 160 μL Taq DNA ligase (NEB, 40 U/μL). This solution was divided into 15 μL aliquots and stored at −20 °C. Vector to fragment ratios were variable; typically,100 ng of linearized vector was added to the mixture with a 2-fold excess of each PCR fragment. PCR fragments less than 100 bp were added at 4–6-fold excess over the vector. The mixture was incubated at 50 °C for 1 h and 10 μL was transformed into chemically competent E. coli. Plasmid DNA was isolated from colonies and assayed for the correct vector assembly by colony PCR and DNA sequencing. For sequencing pTrypSon RNAi constructs, 10 μg of DNA was linearized by overnight incubation with either PacI or AscI, followed by purification using DNA Clean and Concentrator Kit (Zymo Research). The linearized DNA was then submitted for conventional Sanger sequencing. 2.3 Cell culture Experiments were performed in wild type procyclic T. brucei brucei 427 strain and 427 cells carrying the machinery necessary for tetracycline inducibility (29–13). 427 cells were cultured at 28 °C in Cunningham’s medium supplemented with 10% fetal calf serum (Sigma Aldrich). The 29–13 cells were cultured at 28 °C in Cunningham’s medium supplemented, 15% tetracycline free-fetal calf serum (Clontech), 50 μg/mL hygromycin and 15 μg/mL neomycin. Cell growth was monitored using a particle counter (Z2 Coulter Counter, Beckmann Coulter). 2.4 Antibodies Antibodies were obtained from the following sources: AB1 from Keith Gull (Oxford University, UK), anti-Leishmania donovani Centrin4 from Hira L. Nakhasi (Food and Drug Administration, USA), anti-Ty1 from Cynthia He (NUS, Singapore), 1B41 (Linda Kohl, CNRS, France). The monoclonal antibody against TbCentrin2 and antibodies against TbPLK have been described previously [38,45]. Mouse anti-tubulin (clone B-5-1-2) was purchased from Sigma. 2.5 Cloning and cell line assembly All DNA constructs were validated by sequencing prior to transfection. Verified constructs were introduced into cells using electroporation with a GenePulser Xcell (BioRad) and clonal cell lines were generated by selection and limiting dilution. For plasmids based on pLEW100, 30 μg was used in each transfection, while 20 μg was used for endogenous tagging vectors. 2.6 Western blot Cells were harvested, washed once in PBS, then lysed in SDS-PAGE loading buffer. 3×106 cell equivalents of lysate per lane were fractionated using SDS-PAGE, transferred to nitrocellulose, and blocked for 1 h at RT. For antibody detection, blocking and antibody dilution were done in TBS supplemented with 5% non-fat milk and 0.1% Tween-20. Primary antibodies were incubated overnight at 4 °C, followed by washing in TBS containing 0.1% Tween-20, and incubation with secondary antibodies conjugated to HRP (Jackson Immunoresearch). Clarity (BioRad) ECL substrate and a BioRad Gel Doc XR+ documentation system were used for detection. 2. 7 Fluorescence microscopy Cells were harvested, washed once in PBS, then adhered to coverslips. For immunofluorescence, 6×105 cells per coverslip were washed once in PBS, then adhered to coverslips by centrifugation, followed by immersion in −20 °C methanol for 20 min. The cells were then air dried and rehydrated in PBS. The cells were blocked overnight at 4 °C in blocking buffer (PBS containing 3% BSA). Primary antibodies were diluted in blocking buffer and incubated for 1 h at RT, then washed 4 times in PBS and placed in blocking buffer for 20 min. Alexa 488- or 568-conjugated secondary antibodies (Life Technologies) were diluted in blocking buffer and incubated for 1 h at RT. Cells were washed and mounted in Fluoromount G with DAPI (Southern Biotech). Coverslips were imaged using a Zeiss Observer Z1 equipped with a CoolSNAP HQ2 camera (Photometrics) and a Plan-Apochromat 63x/1.4 oil immersion lens (Zeiss). AxioVision Rel. 4.8 was used to control the microscope for acquisition. All images were quantified in ImageJ and assembled for publication using Photoshop CC2015 and Illustrator CC2015. 2.8 RNAi Cultures of TbPLK and TbCentrin2 RNAi cells were seeded at 1 × 106 cells/ml and induced by adding 1 μg/ml of doxycycline, while 70% ethanol was added to control cells. Cells were maintained at 28 °C and reseeded every 48 h with fresh media and doxycycline if necessary. Cells were counted every 24 h and samples taken for Western blot analysis. All cell counts are the average of three biological replicates and the error bars are the standard error. 3. Results Tetracycline-inducible systems are commonly employed as a means to avoid toxicity and to control the degree of overexpression. The plasmid pLEW100 and its derivatives are frequently used for this purpose due to their extremely low background and highly tunable levels of expression [5]. The most common form of this plasmid contains the luciferase gene flanked by the restriction sites for HindIII and BamHI. Expression constructs are generated by excision of the luciferase by double digestion followed by insertion of a gene prepared by PCR amplification with compatible restriction sites in the primers or by digestion of an existing gene product to liberate compatible ends. Introduction of epitope tags such as HA and Ty1 into ectopically expressed constructs allows for detection without the need for specific antibodies. The limited choice of restriction enzymes for inserting genes into pLEW100 can cause problems if a site is present within the insert, especially in the case of HindIII, which lacks isoschizomers. To overcome this issue, a set of primers with 5′ overhangs that correspond to 20 base pairs at the termini of pLEW100 after HindIII/BamHI digestion can be used to insert genes via Gibson assembly. Since the PCR product does not have to be treated with restriction enzymes to generate cohesive ends for ligation, any gene can be introduced using this method. As proof of principle, we generated a construct in pLEW100 that contains the gene TOEFAZ1 (Tb927.11.15800) with three copies of the Ty1 epitope tag on its N-terminus (Figure 2A). TOEFAZ1 is found on the tip of the extending new flagellum attachment zone (FAZ) during cell division and appears to be essential for cytokinesis [34]. The triple-Ty1 tag and the TOEFAZ1 gene were PCR amplified with primers containing overhangs for Gibson assembly (Figure 2B). The pLEW100 plasmid was digested with HindIII and BamHI to produce termini that were compatible with the overhangs present in the PCR products. The two amplicons and the digested vector were assembled with the Gibson reagent and then transformed into competent bacteria. Individual transformants carrying the completed construct were identified initially by PCR screen and then by restriction digest. After validation by DNA sequencing, the construct was linearized with NotI and integrated into the rDNA spacer of the 29.13 cell line, which contains the tetracycline suppressor and T7 polymerase, allowing inducible expression [5]. In the absence of tetracycline, no Ty1-positive signal was visible (Figure 2C). In cultures treated with 1 μg/mL doxycycline (a tetracycline analog), cells that were undergoing cell division showed Ty1-positive labeling on the anterior end of the new FAZ, which is the expected localization for TOEFAZ1. Inducible overexpression constructs are useful for initial characterization of proteins at high expression levels and for testing dominant negative phenotypes, but tagging the endogenous locus is the best approach for validating protein localization. Employing the native regulatory elements present in the untranslated regions minimizes alterations in expression level and cell cycle regulation that may be present. This is especially true for trypanosomes, where most regulation of protein expression occurs at the post-transcriptional level via mRNA-binding proteins that function via the 3′ UTRs [35–37]. There are two main approaches to endogenous tagging: PCR with primers containing long (~80 bp) non-annealing overhangs that amplify a selection marker, intergenic region, and a tag such as GFP or an epitope tag [11,20,22]. The primer overhangs then target the tagging construct to the correct locus which inserts using homologous recombination after direct transfection of the PCR product. Alternatively, the tagging construct can be cloned into a plasmid that contains a larger (~500 bp) targeting sequence, which is then digested to release the completed construct prior to transfection [13,38,39]. The primer approach benefits from speed of assembly; only successful PCR is necessary to generate a construct that can then be directly introduced into cells. However, the shorter segments used for targeting can decrease the efficiency of integration. This is especially problematic if both copies of a gene need to be modified because targeting the second allele is at best half as efficient as the first. Since the targeting segments are encoded in PCR primers, it is difficult to verify their sequence prior to transfection. In our experience, UTRs isolated from the Lister 427 strains of T. brucei, which are frequently used for in vitro experiments, can vary from the published sequence, including insertions and deletions. The plasmid approach addresses most of these issues- the longer targeting segments improve the rate of homologous recombination and the DNA sequences can be isolated directly from the targeted strain and confirmed prior to transfection. The only drawback is the need for iterative cloning of the targeting sequences, which can dramatically increase the time it takes to assemble a construct. Gibson assembly allows the rapid assembly of plasmid-based targeting constructs from multiple DNA segments, which removes the biggest obstacle to their use. We have previously described a plasmid that introduces a triple-Ty1 tag at the N-terminus of targeted genes by using a 500 bp portion of the 5′ UTR and the first 500 bp of the gene without the start codon for targeting (Figure 3A) [13]. Between the two targeting segments, the vector contains the blasticidin resistance gene followed by the alpha/beta tubulin intergenic region and the Ty1 tag, flanked by PacI/HindIII restriction sites on the 5′ side and BamHI/NsiI on the 3′ side. Iteratively cloning the 5′ UTR between the PacI/HindIII sites and the beginning of the gene between the BamHI/NsiI sites generated a completed construct, which could be excised for transfection using PacI/NsiI. Using Gibson assembly, the two targeting segments can be amplified by PCR with primers containing overhangs for the plasmid backbone and either terminus of the drug-intergenic-Ty1 segment. To test this method, we attempted to create an N-terminal triple-Ty1 tagging construct for FC1 (Tb927.11.1340), a component of the flagella connector in T. brucei [34]. A Gibson reaction containing two targeting PCR products, the drug-intergenic-Ty1 segment liberated from a previously assembled vector using HindIII/BamHI digestion, and backbone plasmid digested with PacI/NsiI produced the completed vector, yielding the desired product in a single step (Figure 3B). After DNA sequencing to confirm the construct, we liberated the targeting segment by restriction with PacI and NsiI and transfected it into procyclic trypanosomes. After selection with blasticidin and clonal dilution, we were able to isolate resistant cells. Immunofluorescence showed Ty1 labeling at the tip of the new flagellum in dividing cells that colocalized with the flagellar connector marker AB1, confirming the correct tagging of FC1 (Figure 3C) [40]. Gibson assembly not only speeds up the construction of our plasmid tagging constructs but also allows flexibility in terms of the location and type of tag appended to the gene. To demonstrate rapid de novo assembly, we created a novel C-terminal triple-Ty1 tag for the gene Tb927.3.4400, which does not tolerate N-terminal tagging because of a putative internal start site (Figure 4A) [34,41]. While alterations at the C-terminus are likely to disrupt regulation mediated by the 3′ UTR, there are instances such as internal starts and N-terminal signal sequences that require alternate approaches. To create the C-terminal tagging construct, we assembled 6 DNA segments using Gibson assembly. For targeting, we used the same digested backbone plasmid we used to make N-terminal constructs, the PCR-amplified terminal 500 bp of the gene lacking the stop codon, and the first 500 bp of the 3′ UTR for targeting using PCR (Figure 4B). For C-terminal tagging, the positions of the selection marker and epitope tag must be inverted around the intergenic sequence. We generated this segment from separate PCR products containing sequence for puromycin, the alpha/beta tubulin intergenic region, and for triple-Ty1 (Figure 4B). Combining these 6 DNA segments using Gibson assembly yielded the completed tagging construct in a single step, demonstrating the flexibility of the reaction. After DNA sequencing, the validated targeting construct was liberated from the vector by PacI/NsiI digestion and transfected into parasites. Puromycin selection and clonal dilution yielded viable parasites, which were fixed for immunofluorescence and stained with anti-Ty1 antibody. The Ty1 labeling colocalized with a marker of the basal body, suggesting that the protein is an element of the T. brucei cytoskeleton (Figure 4C). Considering the value of Gibson assembly for speeding up the production of tagging constructs, we next sought to improve the speed and ease of constructing plasmids for RNAi. Currently, there are two main strategies for production of long hairpin RNAi (lhRNAi) constructs in T. brucei: a primer containing a large non-annealing 5′ overhang with a restriction site that allows the PCR product to be digested and ligated together in a head-to-tail orientation, and a Gateway-mediated recombination strategy that inserts inverted repeats of the targeted sequence on either side of a “stuffer” sequence to provide a hairpin [28,29,42]. Both methods require intermediate steps; either a restriction/ligation step in the first case or TOPO coning and Gateway recombination steps in the second. The Gateway approach, which employs the pLEW100 plasmid backbone, is especially attractive due to the tight tetracycline inducibility that can be achieved. We developed a strategy using Gibson assembly that allows the single-step assembly of lhRNAis in pLEW100. The Gateway-mediated lhRNAi approach employs a pLEW100-derived vector known as pTrypRNAiGate that contains a stuffer sequence flanked by two AttB1/AttB2 site-specific recombination sites. Incubation with a shuttle vector containing a targeting DNA segment and a recombinase leads to incorporation of the targeting segment onto either side of a stuffer in a head-to-tail fashion. The Gibson-mediated process mimics the Gateway RNAi system by using the AttB1/AttB2 recombination sites to guide the assembly reaction. Using the pTrypRNAiGate plasmid as a guide, we made three modifications to the base pLEW100 vector to generate a vector compatible with Gibson assembly reactions [28]. First, the 3′ BamHI site in pLEW100 was replaced with the HindIII restriction site to create identical AttB1 recombination sites in pTrypRNAiGate and make the vector compatible for downstream Gibson assembly reactions (Figure 5A). Next, we altered the stuffer sequence that separates the two inverted repeats by inserting a small segment that contains the restriction sites for PacI and AscI (Figure 5A). These restriction sites allow the vector to be linearized to prevent secondary structure formation during DNA sequencing. Finally, we inserted XhoI sites flanking the AttB2 recombination sites to allow reuse of the stuffer region with a single restriction digest reaction (Figure 5A). We called the modified pLEW100 vector pTrypSon. As proof of concept we created a pTrypSon plasmid that targets TbPLK (Tb927.7.6310), an essential kinase involved in cytoskeletal biogenesis [43–45]. We selected a unique 500 bp gene segment of TbPLK suitable for RNAi using the program RNAit [46]. The segment was amplified from genomic DNA by PCR with primer overhangs homologous to the AttB1 and AttB2 sequences within the plasmid and stuffer, respectively. A Gibson reaction containing HindIII-digested pTrypGate, the two sections of the stuffer sequence, and the TbPLK-specific RNAi segment generated the completed lhRNAi hairpin plasmid in a single step (Figure 5B). After DNA sequencing to confirm the correct assembly of the hairpin, we digested the vector with NotI and transfected it into 29.13 cells to test its efficacy in depleting TbPLK. After selection of resistant clones, we induced RNAi by the addition of 1 μg/mL doxycycline and monitored TbPLK expression and cell growth. Western blotting of induced and uninduced lysates showed a dramatic decrease of TbPLK levels within 8 h of lhRNAi expression, which is expected because TbPLK is degraded at the end of each cell cycle (Figure 5C). Depletion of TbPLK lead to total block in cell division within 16 h of RNAi induction, which is similar to previously-published reports (Figure 5D) [43–45]. To demonstrate that we could recycle components of pTrypSon and further simplify the process of generating lhRNAi plasmids we generated a second pTrypSon construct, in this case directed against TbCentrin2 (Tb927.8.1080) (Figure 5E) [47]. We excised the completed stuffer segment from the TbPLK lhRNAi pTrypSon plasmid by restriction digest with XhoI, so that only a single PCR is necessary to generate the gene-specific targeting fragment. Gibson assembly with the TbCentrin2 targeting segment, stuffer, and HindIII-digested pTrypSon yielded the completed construct (Figure 5F). Prior to sequencing, the construct was linearized with PacI to avoid interference from secondary structure imparted by the hairpin structure of the DNA. Once the construct was confirmed, the plasmid was linearized with NotI and transfected into 29.13 cells as was previously done with the TbPLK construct. Resistant clones were isolated and treated with doxycycline to assess the effect of TbCentrin2 depletion. TbCentrin2 was depleted within 24 h of RNAi induction and was completely absent by 48 h (Figure 5G). Cell growth was inhibited by 72 h and remained suppressed for the remainder of the experiment (Figure 5H). 4. Discussion The recent advances in high-throughput screens have emphasized the need for more streamlined approaches for generating DNA constructs for probing the function of candidate genes in T. brucei. Gibson assembly provides a single, unified approach for generating the ideal constructs necessary for rapidly characterizing genes with minimal modifications to currently employed plasmids. In this work, we have shown that overexpression, endogenous tagging, and RNAi plasmids can be readily assembled using Gibson assembly in a single step. Using the Gibson reagent diminishes the need for restriction enzymes, which are only necessary for generating segments as components for assembly reactions. Other commonly-used enzymes such as T4 DNA ligase are also unnecessary, further simplifying the process of generating constructs. Considering that this method does not require unique recombination sites, such as in the Gateway approach, essentially any plasmid can be adapted for Gibson assembly. The restriction sites used for conventional restriction/ligation strategies can be used to create compatible termini for assembly with PCR products or other DNA segments generated from digestion of other plasmids. The generality of Gibson assembly allows it to be used to simplify any potential cloning strategy. For example, our previous approach for endogenous tagging using a plasmid relied on iterative cloning of two targeting segments, which took approximately 7–10 days including sequencing[13,38]. With the Gibson assembly approach, the whole construct can be assembled in a single step and directly transformed, providing viable colonies to sequence the next day. The Gibson reagent can be purchased from a variety of different vendors, although it is more cost-effective to create the enzyme mixture from its individual components, which can be done easily to generate large quantities of reagent [32]. The only increased cost for the Gibson approach is the slightly longer primers containing the overlapping sequences that are necessary for the assembly process. The 15–20 base pair overhangs are larger than those used for conventional cloning, although they are substantially shorter than those used for direct endogenous tagging by PCR (~80 base pairs) [11,20]. The increased cost of primers is offset by the decreasing cost of DNA synthesis and the diminished reliance on restriction enzymes and ligase. Designing primers for Gibson assembly is straightforward, although for constructs generated from a large number of DNA segments on-line programs are available to assist with designing overhangs and calculating melting temperatures (http://nebuilder.neb.com/). Gibson assembly improves the speed and ease of generating each type of construct in this work. For the assembly of overexpression constructs from one or two segments of DNA, Gibson assembly is most useful when restriction enzyme incompatibilities are an issue or when a tag needs to be appended in situ, although we have observed anecdotally that larger inserts also clone more easily using the approach when compared to traditional restriction/ligation strategies. When generating endogenous tagging cassettes, Gibson assembly allows the use of larger, fully sequenced targeting segments and provides plasmids that can be archived, unlike current PCR-based strategies. The larger targeting segments substantially enhance the homologous recombination efficiency of the tagging construct. The PCR approaches for endogenous tagging are more rapid than our plasmid/Gibson approach, but the more general utility of our method makes it an important alternative, especially if the second allele of a gene is being targeted. It is also possible to rapidly create new constructs containing different tags and selection markers based on the needs of the specific experiment. For RNAi, the Gibson method allows the assembly of ideal lhRNAi constructs in a tightly inducible plasmid, all in a single step, making it the best strategy currently available. The ability to use a single reagent to assemble all of these constructs makes Gibson assembly an essential tool to expand the scale of trypanosome functional genomics. Supplementary Material 1 2 We would like to thank Richard Bennett for the use of his microscope, and all the labs that provided essential reagents. This work was supported by startup funds from Brown University and the National Institutes of Health (NIGMS P20 GM104317, NIAID R01AI112953, NIAID R21AI115089-01). Abbreviations lhRNAi long hairpin RNAi BioID proximity-dependent biotin identification Figure 1 Overview of the Gibson assembly method [A] Three commonly used plasmids for functional genomics in T. brucei. Expression- A conventional expression vector for inducible control of a gene of interest. Endogenous tagging- A vector that introduces a tag and a selection marker to an endogenous locus. Long hairpin RNAi (lhRNAi)- A plasmid that provides inducible expression of a long hairpin RNAi for depletion of a protein of interest. [B] The specific components comprising each of the constructs described in A. [C] A schematic of the Gibson assembly process. [D] An overview of how multiple DNA segments can be assembled into a completed construct. Figure 2 Gibson assembly of N-terminally tagged TOEFAZ1 in an inducible expression plasmid [A] Schematic of the plasmid, showing the triple-Ty1 tag (3X Ty1; 1), the TOEFAZ1 gene (TOEFAZ1 ORF; 2), and the pLEW100 vector backbone for inducible expression (pLEW100; 3). Sizes for each DNA segment are shown in parentheses. [B] Agarose gel showing the DNA segments used for Gibson assembly (1–3, as labeled in A, and the product (P) of the assembly digested with HindIII and BamHI to show the tagged insert. The asterisk in lane 3 denotes the plasmid backbone of pLEW100, which was used for the Gibson reaction. [C] Trypanosomes containing the completed plasmid where treated with vehicle control (70% EtOH) or 1 μg/mL doxycycline overnight and then fixed and stained with antibody against the FAZ (FAZ; red), anti-Ty1 (Ty1-TOEFAZ1; green), and DAPI to label DNA (DNA; blue). Scale bar is 5 μm. Figure 3 Gibson assembly of an endogenous tagging construct for FC1 [A] Schematic of the plasmid, showing the 5′ UTR targeting segment (1; 5′ UTR), a segment (3) containing the blasticidin selection marker (BSD), alpha/beta tubulin intergenic rection (INTER), triple-Ty1 tag (3X Ty1), the first 500 bp of FC1 (2; FC1 Cod). and the endogenous tagging vector backbone for inducible expression (ET vector, 4). Sizes for each DNA segment are shown in parentheses. [B] Agarose gel showing the DNA segments used for Gibson assembly (1–4, as labeled in A), and the product (P) of the assembly digested with PacI and NsiI to show the completed endogenous tagging insert. The asterisk in lane 3 denotes the BLA-INTER-3X Ty1 DNA segment, which was excised from a previously-assembled construct by BamHI and HindIII restriction digest. The asterisk in lane 4 denotes the ET plasmid backbone, which was isolated by PacI and NsiI digest. [C] Cells containing the FC1 endogenous tagging construct were fixed and stained with an antibody against the flagella connector (FC; red), anti-Ty1 (Ty1-FC1; green), and DAPI to label the DNA (DNA; blue). Cells were imaged by brightfield and fluorescence microscopy. Scale bar is 5 μm. Figure 4 De novo assembly of a tagging construct for C-terminal triple-Ty1 tagging using Gibson assembly [A] Schematic of the plasmid, showing the last 500 bp of the 4400 gene, which functions as a targeting segment (1; 4400 Cod), the triple-Ty1 tag (2; 3X Ty1), the alpha/beta tubulin intergenic region (3; INTER), the puromycin resistance gene (4; PAC), a 500 bp segment of the 3′ UTR of 4400 used for targeting (5; 3′ UTR), and the endogenous tagging vector backbone for inducible expression (ET vector, 6). Sizes for each DNA segment are shown in parentheses. [B] Agarose gel showing the DNA segments used for Gibson assembly (1–6, as labeled in A), and the product (P) of the assembly digested with PacI and NsiI to show the tagged insert. The asterisk in lane 6 denotes the plasmid backbone of the ET vector, which was used for the Gibson reaction. [C] Cells containing the 4400 endogenous tagging construct were fixed and stained with an antibody that detects the basal body and bilobe structure (BB + Bilobe; red), anti-Ty1 (Ty1-FC1; green), and DAPI to label the DNA (DNA; blue). Cells were imaged by brightfield and fluorescence microscopy. Scale bar is 5 μm. Figure 5 One-step assembly of lhRNAi constructs using Gibson assembly [A] Schematic of the TbPLK pTrypSon. It contains the gene sequence for triggering RNAi (1; TbPLK RNAi), one copy of AttB2 and the first half of the stuffer region (2), the second half of the stuffer region and a second copy of AttB2 in an inverted position compared to the sequence in 2 (3), an inverted second copy of the gene sequence (1), and the tetracycline-inducible plasmid (4). [B] Agarose gel showing the DNA segments used for Gibson assembly (1–4), as labeled in A, and the product (P1–5) of the assembly. P1 shows the digestion with HindIII, which liberates the full lhRNAi hairpin, P2 shows digestion with XhoI, which excises the AttB2-stuffer region, P3 and P4 show digestion with AscI and PacI, respectively, while P5 is the uncut plasmid. [C] Cells containing TbPLK pTrypSon were treated with doxycycline to initiate RNAi or vehicle control. Cell lysates were harvested from both conditions at various time points and then probed by anti-TbPLK Western blotting, (TbPLK) with anti-tubulin (Tubulin) used as a loading control. [D] TbPLK lhRNAi cells were treated with doxycycline to induce RNAi (TbPLK RNAi) or vehicle control (Control) and their growth was monitored by cell counting over 24 h. [E] The schematic of the TbCentrin2 pTrypSon. It contains the gene sequence for triggering RNAi (1; TbC2 RNAi), two copies of AttB2 flanking the stuffer region (2), an inverted second copy of the gene sequence (1), and the tetracycline-inducible plasmid (3). [F] Agarose gel showing the DNA segments used for Gibson assembly (1–3), as labeled in A, and the product (P) of the assembly, which was digested with HindIII to liberate the completed lhRNAi. The asterisk in 2 shows the stuffer sequence, which was isolated from the TbPLK pTrypSon by XhoI digest. The asterisk in 3 shows the tetracycline-inducible plasmid backbone, which was used for the assembly. [G] Cells containing the completed lhRNAi construct for TbCentrin2 depletion were treated with doxycycline to initiate RNAi or vehicle control. Cell lysates were harvested from both conditions at various time points and then probed by anti-TbCentrin2 Western blotting, (TbCentrin2) with anti-tubulin (Tubulin) used as a loading control. [D] TbCentrin2 pTrypSon cells were treated with doxycycline to induce RNAi (TbCentrin2 RNAi) or vehicle control (Control) and their growth was monitored by cell counting over 96 h. Highlights Gibson Assembly allows the rapid assembly of multicomponent plasmids. Currently used plasmids can be employed without modifications. Tagging constructs and RNAi hairpins can be generated in a single 1-hour reaction. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. 1 Fèvre EM Wissmann BV Welburn SC Lutumba P The burden of human African trypanosomiasis PLoS Negl Trop Dis 2 2008 e333 10.1371/journal.pntd.0000333 19104653 2 Matthews KR 25 years of African trypanosome research: From description to molecular dissection and new drug discovery Mol Biochem Parasitol 200 2015 30 40 10.1016/j.molbiopara.2015.01.006 25736427 3 Vickerman K Developmental cycles and biology of pathogenic trypanosomes Br Med Bull 41 1985 105 114 3928017 4 Wirtz E Clayton C Inducible gene expression in trypanosomes mediated by a prokaryotic repressor Science 268 1995 1179 1183 7761835 5 Wirtz E Leal S Ochatt C Cross GA A tightly regulated inducible expression system for conditional gene knock-outs and dominant-negative genetics in Trypanosoma brucei Mol Biochem Parasitol 99 1999 89 101 10215027 6 Ngô H Tschudi C Gull K Ullu E Double-stranded RNA induces mRNA degradation in Trypanosoma brucei Proc Natl Acad Sci USa 95 1998 14687 14692 9843950 7 Lye L-F Owens K Shi H Murta SMF Vieira AC Turco SJ Retention and Loss of RNA Interference Pathways in Trypanosomatid Protozoans PLoS Pathog 6 2010 e1001161 13 10.1371/journal.ppat.1001161 21060810 8 Alsford S Turner DJ Obado SO Sanchez-Flores A Glover L Berriman M High-throughput phenotyping using parallel sequencing of RNA interference targets in the African trypanosome Genome Res 21 2011 915 924 10.1101/gr.115089.110 21363968 9 Mony BM Macgregor P Ivens A Rojas F Cowton A Young J Genome-wide dissection of the quorum sensing signalling pathway in Trypanosoma brucei Nature 2013 10.1038/nature12864 10 Alsford S Eckert S Baker N Glover L Sanchez-Flores A Leung KF High-throughput decoding of antitrypanosomal drug efficacy and resistance Nature 2012 1 6 10.1038/nature10771 11 Dean S Sunter J Wheeler RJ Hodkinson I Gluenz E Gull K A toolkit enabling efficient, scalable and reproducible gene tagging in trypanosomatids Open Biology 5 2015 140197 140197 10.1098/rsob.140197 25567099 12 Sunter JD Varga V Dean S Gull K A dynamic coordination of flagellum and cytoplasmic cytoskeleton assembly specifies cell morphogenesis in trypanosomes J Cell Sci 128 2015 1580 1594 10.1242/jcs.166447 25736289 13 Morriswood B Havlicek K Demmel L Yavuz S Sealey-Cardona M Vidilaseris K Novel Bilobe Components in Trypanosoma brucei Identified Using Proximity-Dependent Biotinylation Eukaryotic Cell 12 2013 356 367 10.1128/EC.00326-12 23264645 14 Obado SO Brillantes M Uryu K Zhang W Ketaren NE Chait BT Interactome Mapping Reveals the Evolutionary History of the Nuclear Pore Complex PLoS Biol 14 2016 e1002365 30 10.1371/journal.pbio.1002365 26891179 15 Gadelha C Zhang W Chamberlain JW Chait BT Wickstead B Field MC Architecture of a Host-Parasite Interface: Complex Targeting Mechanisms Revealed Through Proteomics Mol Cell Proteomics 14 2015 1911 1926 10.1074/mcp.M114.047647 25931509 16 Shimogawa MM Saada EA Vashisht AA Barshop WD Wohlschlegel JA Hill KL Cell Surface Proteomics Provides Insight into Stage-Specific Remodeling of the Host-Parasite Interface in Trypanosoma brucei Mol Cell Proteomics 14 2015 1977 1988 10.1074/mcp.M114.045146 25963835 17 Urbaniak MD Martin DMA Ferguson MAJ Global quantitative SILAC phosphoproteomics reveals differential phosphorylation is widespread between the procyclic and bloodstream form lifecycle stages of Trypanosoma brucei J Proteome Res 12 2013 2233 2244 10.1021/pr400086y 23485197 18 Ericson M Janes MA Butter F Mann M Ullu E Tschudi C On the extent and role of the small proteome in the parasitic eukaryote Trypanosoma brucei BMC Biol 12 2014 1 17 10.1186/1741-7007-12-14 24417977 19 Erben ED Fadda A Lueong S Hoheisel JD Clayton C A Genome-Wide Tethering Screen Reveals Novel Potential Post-Transcriptional Regulators in Trypanosoma brucei PLoS Pathog 10 2014 e1004178 10.1371/journal.ppat.1004178 24945722 20 Shen S Arhin GK Ullu E Tschudi C In vivo epitope tagging of Trypanosoma brucei genes using a one step PCR-based strategy Mol Biochem Parasitol 113 2001 171 173 11254965 21 Oberholzer M Morand S Kunz S Seebeck T A vector series for rapid PCR-mediated C-terminal in situ tagging of Trypanosoma brucei genes Mol Biochem Parasitol 145 2006 117 120 10.1016/j.molbiopara.2005.09.002 16269191 22 Kelly S Reed J Kramer S Ellis L Webb H Sunter J Functional genomics in Trypanosoma brucei: A collection of vectors for the expression of tagged proteins from endogenous and ectopic gene loci Mol Biochem Parasitol 154 2007 103 109 10.1016/j.molbiopara.2007.03.012 17512617 23 Bangs JD Brouch EM Ransom DM Roggy JL A soluble secretory reporter system in Trypanosoma brucei. Studies on endoplasmic reticulum targeting J Biol Chem 271 1996 18387 18393 8702482 24 Wirtz E Hartmann C Clayton C Gene expression mediated by bacteriophage T3 and T7 RNA polymerases in transgenic trypanosomes Nucleic Acids Res 22 1994 3887 3894 7937108 25 Bastin P Bagherzadeh Z Matthews KR Gull K A novel epitope tag system to study protein targeting and organelle biogenesis in Trypanosoma brucei Mol Biochem Parasitol 77 1996 235 239 8813669 26 LaCount DJ Barrett B Donelson JE Trypanosoma brucei FLA1 is required for flagellum attachment and cytokinesis J Biol Chem 277 2002 17580 17588 10.1074/jbc.M200873200 11877446 27 Wang Z Morris JC Drew ME Englund PT Inhibition of Trypanosoma brucei Gene Expression by RNA Interference Using an Integratable Vector with Opposing T7 Promoters J Biol Chem 275 2000 40174 40179 10.1074/jbc.M008405200 11013266 28 Kalidas S Li Q Phillips MA A Gateway® compatible vector for gene silencing in bloodstream form Trypanosoma brucei Mol Biochem Parasitol 178 2011 51 55 10.1016/j.molbiopara.2011.03.002 21420443 29 Atayde VD Ullu E Kolev NG A single-cloning-step procedure for the generation of RNAi plasmids producing long stem–loop RNA Mol Biochem Parasitol 184 2012 55 58 10.1016/j.molbiopara.2012.04.003 22542486 30 Inoue M Nakamura Y Yasuda K Yasaka N Hara T Schnaufer A The 14-3-3 proteins of Trypanosoma brucei function in motility, cytokinesis, and cell cycle J Biol Chem 280 2005 14085 14096 10.1074/jbc.M412336200 15653691 31 Barnes RL McCulloch R Trypanosoma brucei homologous recombination is dependent on substrate length and homology, though displays a differential dependence on mismatch repair as substrate length decreases Nucleic Acids Res 35 2007 3478 3493 10.1093/nar/gkm249 17478508 32 Gibson DG Young L Chuang R-Y Venter JC Hutchison CA Smith HO Enzymatic assembly of DNA molecules up to several hundred kilobases Nat Methods 6 2009 343 345 10.1038/nmeth.1318 19363495 33 Gibson DG Smith HO Hutchison CA Venter JC Merryman C Chemical synthesis of the mouse mitochondrial genome Nat Methods 7 2010 901 903 10.1038/nmeth.1515 20935651 34 McAllaster MR Ikeda KN Lozano-Núñez A Anrather D Unterwurzacher V Gossenreiter T Proteomic identification of novel cytoskeletal proteins associated with TbPLK, an essential regulator of cell morphogenesis in Trypanosoma brucei Mol Biol Cell 26 2015 3013 3029 10.1091/mbc.E15-04-0219 26133384 35 Clayton CE Life without transcriptional control? From fly to man and back again Embo J 21 2002 1881 1888 10.1093/emboj/21.8.1881 11953307 36 Erben ED Fadda A Lueong S Hoheisel JD Clayton C A Genome-Wide Tethering Screen Reveals Novel Potential Post-Transcriptional Regulators in Trypanosoma brucei PLoS Pathog 10 2014 e1004178 15 10.1371/journal.ppat.1004178 24945722 37 Clayton CE ScienceDirect Gene expression in Kinetoplastids Curr Opin Microbiol 32 2016 46 51 10.1016/j.mib.2016.04.018 27177350 38 de Graffenried CL Anrather D Von Raussendorf F Warren G Polo-like kinase phosphorylation of bilobe-resident TbCentrin2 facilitates flagellar inheritance in Trypanosoma brucei Mol Biol Cell 24 2013 1947 1963 10.1091/mbc.E12-12-0911 23615446 39 Turnock DC Izquierdo L Ferguson MAJ The de novo synthesis of GDP-fucose is essential for flagellar adhesion and cell growth in Trypanosoma brucei J Biol Chem 282 2007 28853 28863 10.1074/jbc.M704742200 17640865 40 Briggs LJ McKean PG Baines A Moreira-Leite F Davidge J Vaughan S The flagella connector of Trypanosoma brucei: an unusual mobile transmembrane junction J Cell Sci 117 2004 1641 1651 10.1242/jcs.00995 15075226 41 Nilsson D Gunasekera K Mani J Osteras M Farinelli L Baerlocher L Spliced Leader Trapping Reveals Widespread Alternative Splicing Patterns in the Highly Dynamic Transcriptome of Trypanosoma brucei PLoS Pathog 6 2010 e1001037 13 10.1371/journal.ppat.1001037 20700444 42 Sunter J Wickstead B Gull K Carrington M A New Generation of T7 RNA Polymerase-Independent Inducible Expression Plasmids for Trypanosoma brucei PLoS ONE 7 2012 e35167 9 10.1371/journal.pone.0035167 22511983 43 Kumar P Wang CC Dissociation of cytokinesis initiation from mitotic control in a eukaryote Eukaryotic Cell 5 2006 92 102 10.1128/EC.5.1.92-102.2006 16400171 44 Hammarton TC Kramer S Tetley L Boshart M Mottram JC Trypanosoma brucei Polo-like kinase is essential for basal body duplication, kDNA segregation and cytokinesis Mol Microbiol 65 2007 1229 1248 10.1111/j.1365-2958.2007.05866.x 17662039 45 de Graffenried CL Ho HH Warren G Polo-like kinase is required for Golgi and bilobe biogenesis in Trypanosoma brucei J Cell Biol 181 2008 431 438 10.1083/jcb.200708082 18443217 46 Redmond S Vadivelu J Field MC RNAit: an automated web-based tool for the selection of RNAi targets in Trypanosoma brucei Mol Biochem Parasitol 128 2003 115 118 10.1016/S0166-6851(03)00045-8 12706807 47 He CY Pypaert M Warren G Golgi duplication in Trypanosoma brucei requires Centrin2 Science 310 2005 1196 1198 10.1126/science.1119969 16254149
PMC005xxxxxx/PMC5125876.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101623795 42112 Arthritis Rheumatol Arthritis & rheumatology (Hoboken, N.J.) 2326-5191 2326-5205 27563728 5125876 10.1002/art.39837 NIHMS809280 Article Receptor activator of nuclear factor kappa-B (RANK) independent osteoclast formation and bone erosion in inflammatory arthritis O’Brien William M.D. 1* Fissel Brian M. M.S. 1* Maeda Yukiko PhD 2 Yan Jing Ph.D. 1 Ge Xianpeng D.D.S./Ph.D. 1 Gravallese Ellen M. M.D. 2 Aliprantis Antonios O. M.D./Ph.D. 1% Charles Julia F. M.D./Ph.D. 1# 1 Department of Medicine, Division of Rheumatology, Immunology and Allergy, Brigham and Women’s Hospital and Harvard Medical School, One Jimmy Fund Way, Rm650A, Boston, MA 02115 2 Department of Medicine, Division of Rheumatology, University of Massachusetts Memorial Medical Center and University of Massachusetts School of Medicine # Correspondence and reprint requests should be sent to either: Julia F. Charles, Department of Medicine, Division of Rheumatology, Immunology and Allergy, Brigham and Women’s Hospital, One Jimmy Fund Way, Rm614A, Boston, MA 02115, [email protected], Ph: (617)-525-1224, Fax: (617) 525-1013 * equal contribution % Current contact information: Merck Research Laboratories, 33 Ave Louis Pasteur, Boston, MA 02115, [email protected], 617-992-3040 20 8 2016 12 2016 01 12 2017 68 12 28892900 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Objective Pro-inflammatory molecules promote osteoclast-mediated bone erosion by upregulating local RANKL production. However, recent evidence suggests that combinations of cytokines, such as TNFα plus IL-6, induce RANKL-independent osteoclastogenesis. This study sought to better understand TNFα/IL-6 induced osteoclast formation, and to determine whether RANK is absolutely required for osteoclastogenesis and bone erosion in murine inflammatory arthritis. Methods Myeloid precursors from wild-type (WT) mice, or mice with either germline or conditional deletion of Rank, Nfatc1, Dap12 or Fcrg, were treated with either RANKL, or TNFα plus IL-6. Osteoprotegerin, anti-IL-6 receptor (IL6R) and hydroxyurea were used to block RANKL, the IL6R and cell proliferation, respectively. Clinical scoring, histology, micro-computed tomography and qPCR were employed to evaluate K/BxN serum transfer arthritis in WT and RANK-deleted mice. Loss of Rank was verified by qPCR and by performing osteoclast cultures. Results TNFα/IL-6 generated osteoclasts in vitro that resorbed mineralized tissue through a pathway dependent on IL6R, NFATc1, DAP12 and cell proliferation, but independent of RANKL or RANK. Bone erosion and osteoclast formation were reduced, but not absent, in arthritic mice with inducible deficiency of RANK. TNFα/IL-6, but not RANKL, induced osteoclast formation in bone marrow and synovial cultures from RANK-deficient animals. Multiple IL-6 family members (IL-6, LIF, OSM) were upregulated in the synovium of arthritic mice. Conclusion The persistence of bone erosion and synovial osteoclasts in RANK-deficient mice, and the ability of TNFα/IL-6 to induce osteoclastogenesis, suggest more than one cytokine pathway exists to generate these bone resorbing cells in inflamed joints. Introduction Periarticular bone erosion is a hallmark manifestation of rheumatoid arthritis (RA) and other inflammatory arthritidies, which can occur soon after disease onset (1, 2). Unhindered joint inflammation leads to proliferation of the synovium and destruction of bone, resulting in deformity and functional deterioration (3). While controlling synovitis reduces inflammation and can arrest the progression of erosions in some cases, patients in remission or with low disease activity may continue to accrue erosions (4, 5). Osteoclasts are the only cell type well accepted to resorb bone. In RA patients, these cells are found at sites where the synovium engages the periosteal surface at the edge of the articular cartilage, as well as in subchondral and trabecular bone (1, 6, 7). These large multi-nucleated cells differentiate from myeloid precursors upon stimulation with macrophage colony-stimulating factor (MCSF) and receptor activator of NF-κB ligand (RANKL) (8). MCSF promotes the survival and expansion of precursors, while stimulation of RANK by RANKL initiates canonical and non-canonical NF-κB pathways, as well as the mitogen-activated kinase (MAPK) pathway (9). RANKL-driven osteoclast differentiation requires co-stimulatory signals initiated by two adaptor molecules, DNAX-activated protein 12 (DAP12) and Fc receptor common γ subunit (FcRγ) (10, 11), which contain immunoreceptor tyrosine-based activation motifs (ITAM). Tyrosine phosphorylation of the ITAM motif enables activation of phospholipase-Cγ2, which increases intracellular calcium. This calcium signal promotes activation of nuclear factor of activated T cells cytoplasmic 1 (NFATc1), the master regulatory transcription factor of osteoclast differentiation (9, 12). Pro-inflammatory cytokines such as IL-1, IL-6, and TNFα drive joint inflammation and damage (13, 14). These cytokines promote osteoclast differentiation through induction of RANKL on synovial fibroblasts (1, 15) and may directly activate osteoclasts. For example, TNFα sensitizes precursors to RANKL, and IL-1 promotes terminal osteoclast differentiation (16–18). Whether inflammatory cytokines promote osteoclastogenesis independent of RANKL has been controversial. Some reports suggest that TNFα and stromal cell-derived factor 1 (SDF1), can promote osteoclast formation (19–23) and a recent study showed that TNFα and IL-6 drive osteoclast formation (24). However, seminal studies demonstrated that arthritic mice lacking RANKL or the RANK receptor do not form osteoclasts (7, 25), and humans treated with denosumab, a RANKL-specific blocking antibody, are protected from inflammatory bone erosion (26). However, it remains possible that bone erosion in the absence of RANKL signaling may occur under certain inflammatory conditions. Here, we show that TNFα/IL-6 can drive osteoclastogenesis in RANK-deficient cells, excluding participation of this receptor. The TNFα/IL-6 pathway acts on an identical myeloid precursor population, and has similar requirements for ITAM co-stimulation and NFATc1, as classic RANKL-RANK driven osteoclastogenesis. Bone erosion and osteoclast formation after induction of KBxN serum transfer arthritis were reduced, but not absent, in mice with inducible deficiency of RANK. Moreover, other IL-6 family cytokines, Oncostatin M (OncM) and Leukemia Inhibitory Factor (LIF) were upregulated in the synovium of mice with KBxN serum transfer arthritis. Taken together, our results provide evidence that osteoclasts can develop in vitro in response to TNFα/IL-6 in the absence of RANK. Moreover, we show that osteoclast formation and bone erosion occur without RANK in an inflammatory arthritis model where IL-6 family cytokines are elevated. These data may aid in the design of rationale therapies to reduce bone destruction in inflammatory arthritis patients. Materials and Methods Mice Tnfrsf11afl/fl (Rankfl/fl) mice were a gift of Dr. Josef Penninger (27). Nfatc1fl/fl mice were described in (28). Rankfl/fl and Nfatc1fl/fl mice were mated to Mx1-Cre mice (29) (Jackson Laboratories, Bar Harbor, ME). Dap12−/− and Fcrg−/− mice were provided by Dr. Mary Nakamura and Dr. Jessica Hammerman, respectively (30). Reporter mT/mG mice purchased from Jackson Laboratories were mated to Rankfl/fl;Mx1-Cre+mice to generate Rankfl/fl;Mx1-Cre+; mTmG mice. All mouse strains were on a C57BL/6 background and maintained under specific pathogen-free conditions. The number of animals used for each experiment is noted in the figure legends. Animal research was conducted with the approval of the Institutional Animal Care and Use Committee of Harvard Medical School and conformed to relevant guidelines and laws. Deletion of Rank and Nfatc1 Rankfl/fl;Mx1-Cre+ and Nfatc1fl/fl;Mx1-Cre+ mice were treated at 8 weeks of age with 250 μg of polyinosinic-polycytidylic acid (poly I:C; Sigma-Aldrich) every other day for 3 doses to induce deletion of floxed alleles. Mice were utilized 4 weeks after polyI:C treatment. Mice with Mx1-Cre driven deletion of Rank and Nfatc1 are referred to as RankΔ/Δ and Nfatc1Δ/Δ mice, respectively. Similarly treated sex-matched, littermate Rankfl/fl;Mx1-Cre− mice and Nfatc1 fl/fl;Mx1-Cre− mice were used as controls and are referred to as RankWT and Nfatc1WT, respectively. Rankfl/fl;Mx1-Cre+; mTmG and control Rankfl/fl; Mx1-Cre−; mTmG mice were given poly I:C to generate RankΔ/Δ mTmG and RankWT mTmG mice, respectively. In vitro Assays for Osteoclast Differentiation Osteoclast assays were performed as described (31, 32). Bone marrow cells (BMCs) were plated in triplicate in 96-well flat-bottom plates at a density of 3 × 104 cells/well with 10 ng/ml MCSF (R&D Systems, Minneapolis, MN). After 2 days, adherent bone marrow-derived macrophages (BMMs) were cultured in OC media supplemented with 10 ng/ml MCSF and either 5 ng/ml murine RANKL (PeproTech, Rocky Hill, NJ), 50ng/ml murine TNFα (PeproTech), 50ng/ml murine IL-6 (R&D Systems) or combined TNFα/IL-6. After 3 days, fresh cytokines were added. In some experiments, osteoprotegerin (OPG) (R&D Systems), anti-mouse IL-6 Receptor antibody (cMR16-1; Genentech, San Francisco, CA) (33) or control mouse IgG2a antibody (BioXCell, West Lebanon, NH) were added. Where indicated, cell proliferation was monitored using the alamarBlue assay according to the manufacturer’s instructions (Thermo Fisher Scientific). After 5 days, cells were fixed and stained for tartrate-resistant acid phosphatase (TRAP). Osteoclasts were quantified by counting TRAP-positive cells with 3 or more nuclei. All images were taken using a Nikon TMS-F inverted microscope at a final magnification of 100X. In vitro Assays for Osteoclast Resorption To measure resorptive function, osteoclasts were grown on either commercially available human bone particles (OsteoAssay Wells, Lonza, Basel, Switzerland) or hydroxyapatite coated wells (OsteoLogic slides, BD Biosciences, Franklin Lakes, NJ). Osteoclasts were cultured as above for 14 days in the presence of cytokines, which were replenished every 2–3 days. For the OsteoAssay, resorption activity was measured using an enzyme-lined immunoassay (ELISA) to detect carboxy-terminal collagen crosslinks (CTX) (Immunodiagostic Systems, Boldin, UK) in culture supernatants. For OsteoLogic plates, cells were removed and wells were stained with 5% (w/v) silver nitrate (Von Kossa reagent) and 1% pyrogallol. Images were taken using a Nikon TMS-F inverted microscope at a final magnification of 100X. Osteoclast differentiation of FACS sorted osteoclast precursors Single-cell suspensions of mouse BMCs were treated with Fc-block, and incubated with FITC-conjugated antibodies against CD3, B220, and Ter119, plus CD11b-APC and Ly6C-PE-Cy7 (all antibodies from BioLegend, San Diego, CA). After exclusion of CD3/B220/Ter119+ cells, osteoclast precursors were sorted based on the intensity of CD11b and Ly6C as in (32). To compare response of osteoclast precursors to different cytokine combinations, FACS-sorted precursors were plated in triplicate at either 2.5 or 5 × 103/well in 96-well plates and cultured as above. Real-time Quantitative PCR (qPCR) Analysis Osteoclasts for gene expression analysis were generated from BMMs as above, mRNA was isolated with Trizol reagent, and cDNA was generated with the Affinity script qPCR cDNA synthesis kit (Agilent Technologies, Santa Clara, CA). Synovial samples were prepared as described (34). qPCR reactions were performed using Fast SYBR Green Master Mix (Life Technologies) on an Mx3005P qPCR system (Agilent Technologies). cDNA levels for each target gene were normalized to the house keeping genes hypoxanthine phosphoribosyltransferase (Hprt) or glyceraldehyde 3-phosphate dehydrogenase (Gapdh). Samples with no target amplification were assigned a value of zero. Primer sequences are listed in Table 1. K/BxN Serum-Transfer Arthritis model Arthritis was induced in RankWT and RankΔ/Δ mice 4 weeks after polyI:C treatment by injection with 150 μl of pooled K/BxN arthritogenic serum (35) by intraperitoneal injection on days 0 and 2. Disease severity was assessed every 2 days by a single blinded investigator using a clinical scoring system and calipers to measure ankle thickness as described in (36) and mice were euthanized on day 14. Micro-computed tomography (μCT) analysis for bone erosion Fore paws were scanned on a Scanco μCT-35 with an isotropic voxel size of 7 μm and 3-dimensional images were generated with software supplied by the manufacturer using a global threshold that set the bone/marrow cut-off at 352.3 mg HA/cm3. The severity of periarticular erosions was determined blindly on the 3D images using a semi-quantitative method. Four sites in the wrist joint were scored: the distal ulnar epiphysis, and the bases of the third, fourth, and fifth metacarpals (Figure S1A). Each site received a score of 0–3 as described in Figure S1B. The maximal score per paw was 12, and the scores of the left and right fore paw of each mouse were determined by two blinded observers and averaged for a final erosion score. See Figures S1C and S1D for intra- and inter-observer reproducibility. Histology Hindpaws were fixed, decalcified and embedded in paraffin. Inflammation and bone erosions were assessed on H&E stained sagittal sections of the midfoot. Arthritic changes were assessed and scored as in (7) with minor modifications. TRAP staining was used to identify osteoclasts (32). TRAP staining was scored on a scale of 0 to 5 as follows: 0 - no TRAP-positive cells, 1 - rare TRAP-positive cells in BM or soft tissues, 2 - few TRAP-positive cells seen in areas of resorption, 3 - moderate TRAP-positive cells at at least one site of resorption or attached to periarticular sites without bone erosion, 4 - moderate TRAP-positive staining at multiple areas of resorption or within resorption pits, 5- marked TRAP-positive cells at most or all areas of bone resorption. The average score of the two hindpaws was reported as the individual score for the mouse. For fluorescence images, cryosections of OCT embedded hindpaws were mounted in VECTASHIELD Antifade Mounting Medium with DAPI (Vector Laboratories). Fluorescence images and photomicrographs were obtained using a KEYENCE BZ-X710 and Leica DM 2000 respectively. Brightness and contrast were digitally enhanced for images in Figure 4C. Synovial Cultures Pooled synovial tissue from the wrist, ankle, and midfeet was collected 14 days after the initiation of K/BxN serum transfer arthritis, minced and incubated with 1 mg/ml collagenase type 4 (Sigma-Aldrich) and 2 mg/ml dispase 2 (Roche, Basel, Switzerland). Digests were filtered through a 100 micron Falcon cell strainer (Corning, Corning, NY) followed by RBC lysis. Cells were cultured in OC media with the indicated cytokines at a density of 5 × 104 cells/well in a 96-well plate for 10 days with cytokines replenished every 2–3 days. Statistical Analysis Values are presented as the mean ± SEM. An unpaired t test or Mann-Whitney test, as appropriate, was used for comparisons between 2 groups. A one-way or two-way ANOVA with Sidak’s or Tukey’s post-test was performed for comparisons between multiple groups for single or multiple variables, respectively, as specified in the figure legends. P values less than 0.05 were considered significant. Results TNFα and IL-6 promote osteoclast differentiation Recent data suggests that other inflammatory cytokines may substitute for RANKL in osteoclast differentiation assays. In particular, the combination of TNFα and IL-6 promotes osteoclast differentiation in vitro from mouse myeloid precursors (24). This pathway was not inhibited by OPG, suggesting that RANKL is not required. However, the signaling requirements leading to osteoclast formation in response to TNFα/IL-6 have not been resolved. To investigate this pathway further, we performed in vitro osteoclast differentiation assays from mouse myeloid precursors (Figures 1A, B). MCSF alone, or MCSF with IL-6, did not induce osteoclast formation, while MCSF and TNFα led to the formation of TRAP-positive mononuclear cells and few multinuclear cells. In contrast, both RANKL and TNFα/IL-6 in the presence of MCSF induced large TRAP-positive multinucleated cells. The efficiency of osteoclast formation was greater for RANKL than TNFα/IL-6 at the concentration of cytokines employed. While OPG completely blocked osteoclast formation in response to RANKL, it had no effect on TNFα/IL-6 induced osteoclastogenesis. Similar to RANKL-induced osteoclasts, those generated in response to TNFα/IL-6 expressed markers of terminal differentiation, including Nfatc1, Itgb3, Atp6v0d2, Ctsk, Calcr, Acp5, Dcstamp, and Slc4a2 (Figure 1C). Interestingly, few differences in gene expression were observed between BMMs cultured with TNFα alone vs. TNFα/IL-6 at either day 4 of differentiation (Figure 1C) or day 2 of differentiation (data not shown). In RANKL osteoclast formation assays, the majority of precursor activity in murine bone marrow resides within a population of CD11blow/−Ly6CHi myeloid cells (32). We verified that this precursor population also formed osteoclasts in response to TNFα/IL-6 (Figure 1D). Taken together, these data indicate that in vitro, TNFα/IL-6 induces the formation of osteoclast-like cells from the identical population of precursors, and with similar morphologic and gene expression characteristics, as RANKL. RANK is not required for TNFα/IL-6 induced osteoclastogenesis TNFα/IL-6 induced osteoclastogenesis is not inhibited by OPG, a decoy receptor for RANKL (Figures 1A,B and (24)). However, this does not exclude a requirement for the RANKL receptor, RANK. Moreover, the additional molecular requirements leading to osteoclast formation downstream of TNFα/IL-6 have not been defined. As expected, RANKL was unable to induce osteoclastogenesis from RankΔ/Δ BMMs (Figure 2A, B). In contrast, both RankWT and RankΔ/Δ BMMs formed osteoclasts in response to TNFα/IL-6. Since RANK is also required for osteoclast resorbing activity, RankΔ/Δ BMMs stimulated with TNFα/IL-6 were tested for their ability to resorb calcified matrices. Osteoclasts derived from TNFα/IL-6 treated RankWT and RankΔ/Δ BMMs formed resorption pits when cultured on hydroxyapatite coated plates (Figure 2C) and generated type I collagen fragments, an indicator of bone resorption, when grown on human bone chips (Figure 2D). These data indicate that RANK is dispensable for both the formation and resorbing activity of TNFα/IL-6 induced osteoclasts in vitro. TNFα/IL-6 induced osteoclastogenesis requires IL6R, NFATc1, costimulatory pathways and cell proliferation We confirmed the requirement for the IL6R for TNFα/IL-6 induced osteoclastogenesis. The IL6R blocking antibody, cMR16-1, but not an isotype control, dose dependently inhibited osteoclast formation in response to TNFα/IL-6 (Figure 3A), but had no effect on RANKL driven osteoclastogenesis (data not shown). Next, we investigated whether the transcription factor NFATc1 and co-stimulatory pathways are required for TNFα/IL-6 induced osteoclast formation as they are for RANKL induced osteoclasts (11, 28, 30, 37). Neither RANKL nor TNFα/IL-6 was capable of inducing osteoclast formation in NFATc1-deficient BMMs (Figure 3B, C). Likewise, both RANKL and TNFα/IL-6 required the ITAM adapter molecule DAP12, to induce osteoclast formation (Figure 3C). Similar to what has been previously shown for RANKL (11), FcRγ was not required for TNFα/IL-6 to promote osteoclastogenesis (Figure S2). Both TNFα and TNFα/IL-6 treated BMMs expressed markers of osteoclast differentiation (Figure 1C), but only the latter formed multinuclear osteoclasts (Figures 1A,B and Figures 2C,D). Removing IL-6 on day 2 of culture reduced, but did not eliminate, RANKL-independent osteoclastogenesis (Figure S3A), suggesting that IL-6 may have pro-proliferative effects on precursors (38, 39). Indeed, RANKL and TNFα inhibit proliferation compared to MCSF alone, while IL-6 promotes proliferation in the presence of TNFα to levels similar to MCSF alone (Figure S3B). Accordingly, the addition of hydroxyurea, a small molecule that inhibits cell proliferation without affecting viability, significantly reduced TNFα/IL-6 induced osteoclast formation (Figure 3D). In contrast, hydroxyurea had no effect on RANKL mediated osteoclastogenesis. In summary, the data in Figure 3 indicate that TNFα/IL-6 promotes osteoclast formation through a pathway dependent on IL6R, DAP12, NFATc1 and cell proliferation. RankΔ/Δ mice are not fully protected from inflammatory bone erosion The ability of TNFα/IL-6 to support osteoclastogenesis suggests that RANK-independent osteoclast formation and bone erosion should be evident in a sufficiently inflamed environment. To test this hypothesis we induced maximal KBxN serum transfer arthritis in RankWT and RankΔ/Δ mice. In this series of experiments, the Rank gene was not deleted until the mice were 8 weeks, ensuring they were skeletally mature at the time of arthritis induction. Arthritis was induced 4 weeks after polyI:C treatment, at which time we see evidence of excellent Rank deletion (Figure S4). The K/BxN serum transfer model was chosen because both TNFα and IL-6 are robustly induced within 3–6 days after serum transfer (40). Indeed, we confirmed that IL-6, as well as the other IL-6 family member cytokines Oncostatin M (OSM) and Leukemia inhibitory factor (LIF), are upregulated in the synovium of WT mice with K/BxN serum transfer arthritis (Figure 4A). Inflammation, as measured by clinical score, paw thickness measurements and hindpaw histopathology, did not differ between RankWT and RankΔ/Δ mice given K/BxN arthritogenic serum (Figure 4B–C). Histopathologic bone erosions scores were reduced, but not absent, in RankΔ/Δ mice (Figures 4C, D) while cartilage erosions were indistinguishable (Figure S5A). Osteoclasts, as identified by TRAP staining, were reduced in the inflamed synovium of RankΔ/Δ mice, but similar to bone erosions were not absent as expected (Figures 4C). In contrast, TRAP-positive osteoclasts were not observed in the bone marrow compartment of RankΔ/Δ mice (Figures 4C, S5B). Micro-CT was used to confirm the presence of cortical erosions in the forepaws of RankΔ/Δ mice. Similar to histopathology, micro-CT revealed a partial reduction in erosion scores in RankΔ/Δ mice (Figure 4D) when scored according to the method described in Figure S1. These data suggest that a significant fraction of the inflammatory bone erosion observed in the K/BxN serum transfer arthritis model proceeds independent of the RANK pathway. A possible explanation for the observation that RankΔ/Δ mice were only partially protected from inflammatory bone erosion was that our methodology resulted in incomplete deletion of the conditional Rank allele. This could result in a population of RANK-positive osteoclast precursors within the inflamed synovium. Thus, the osteoclasts and erosion we observed may indeed have been generated via a RANK pathway. The following experiments however indicate that incomplete Rank gene deletion was not a likely reason for our result. First, a greater than 99% reduction in Rank mRNA was observed in BM cells isolated from arthritic RankΔ/Δ compared to RankWT mice (Figure 5A). Second, RANKL was unable to generate osteoclasts from cells isolated from either the BM or inflamed synovium of arthritic RankΔ/Δ mice (Figures 5B–D). In contrast, cells isolated from the inflamed synovium formed osteoclasts readily in response to TNFα/IL-6 (Figures 5C). Lastly, serum transfer arthritis was given to RankΔ/Δ mTmG and RankWT mTmG mice. The mT/mG reporter facilitates discrimination of cells that have expressed the Cre recombinase through Cre-mediated exchange of a red TdTomato (mT) for GFP (mG). Serial frozen-sections were stained for TRAP and visualized with direct fluorescence microscopy. As expected, no GFP-positive cells were detected in samples from RankWT mTmG mice (data not shown). In contrast, GFP-positive cells were detected throughout the synovium of RankΔ/Δ mTmG mice, indicative of Cre-expression and thus Rank deletion (Figure 5D). Importantly, TRAP-positive multinucleated osteoclasts colocalized with GFP-expression on serial sections within erosions of RankΔ/Δ mTmG mice. The digitally enlarged images of boxed areas on the right most image of Figure 5D highlight two examples of GFP+ multinucleated osteoclasts. This indicates that the osteoclasts observed in the inflamed joints of RankΔ/Δ mTmG mice previously expressed Cre and therefore likely had recombined the floxed Rank allele. Taken together, the data in Figure 5 suggest that our strategy resulted in robust Rank gene deletion and that the osteoclasts and erosions observed in the arthritic joints of RankΔ/Δ mice arose from a RANK-independent pathway. Discussion In the present study, we demonstrate that TNFα/IL-6 can generate osteoclast-like cells capable of bone resorption in vitro, confirming a recent report (24). TNFα/IL-6 induced osteoclast formation, and the resorbing capability of these cells, is entirely independent of RANK signaling. However, the pathways driving osteoclast formation by RANKL and TNFα/IL-6 share several features. Both pathways require co-stimulation through ITAM adapter proteins, and the transcription factor NFATc1, to drive osteoclastogenesis from CD11blow/− Ly6Chi myeloid precursors. The requirement for DAP12 suggests that an ITAM co-signal is required to induce osteoclastogenesis regardless of whether the first signal for osteoclast differentiation is RANKL or TNFα/IL6. We hypothesize that, analogous to RANKL stimulation, ITAM-mediated calcineurin activation is required to drive NFATc1 nuclear translocation and autoamplification in TNFα/IL6 driven osteoclastogenesis. Interestingly, BMMs cultured with TNFα/IL-6 did not display the expected increased levels of osteoclast marker genes compared to those treated with TNFα alone. However, only BMMs treated with TNFα/IL-6 formed multinucleated cells capable of resorption. This suggests IL-6 may facilitate multinucleation by promoting the expansion of pre-osteoclasts, increasing the population available for fusion. Consistent with this, IL-6 increased the proliferation of TNFα treated BMM, and the cell proliferation inhibitor hydroxyurea attenuated TNFα/IL-6 induced osteoclastogenesis. While pro-proliferative effects of IL-6 appear to be important, we cannot rule out additional mechanisms by which IL-6 promotes osteoclastogenesis. The in vivo relevance of the observation that TNFα/IL-6 drives osteoclast formation is supported by the outcome of KBxN serum transfer arthritis in RankΔ/Δ mice. In this robust model of inflammatory arthritis, TNFα, IL-6 and other IL-6 family members, are highly upregulated ((40) and data presented here). In the absence of RANK, we found that mice still developed bone erosion and TRAP-positive osteoclasts in inflamed joints, although these parameters were reduced. TNFα/IL-6 induced osteoclast formation from RankΔ/Δ synovial cultures in vitro, and our in vivo model strongly suggests that RANK independent osteoclast formation can occur in vivo in inflammatory conditions. Erosions with TRAP-positive osteoclasts are observed in arthritic RankΔ/Δ mice, and we further demonstrate, using a marker of Cre recombinase expression, that multinucleated TRAP-positive cells located in erosions expressed Cre and thus are RANK deficient. Importantly, our data strongly suggest that the osteoclasts and erosions seen in RANK deficient mice are not solely due to incomplete deletion of the Rank gene by Mx1-Cre. This is the first study to look at inflammatory bone loss in adult mice with inducible RANK deficiency. This strength of our study represents an advance over previous models where inflammatory arthritis was induced in animals with germline deficiency of Rankl, Rank, or c-Fos. These strains show severe congenital osteopetrosis, with an appendicular skeleton composed of calcified cartilage (7, 25, 41). In our report Rank was not deleted until the mice reached skeletal maturity and had developed normal bone with well formed cortices. Possible explanations for the difference between our findings and prior studies that showed complete RANK-RANKL dependence for osteoclastogenesis include differences in the mineralized bone matrix and maturation of hematopoetic precursors to osteoclasts. Recent work by Danks et al. supports the relative importance of RANKL-RANK signaling as a critical inducer of periarticular osteoclastic bone erosion (15). In this study, inflammatory arthritis was induced in mice in which the gene encoding RANKL (Tnfsf11) was deleted from synovial fibroblasts using Col6a1-Cre (Tnfsf11Col6a1). Similar to our model, Tnfsf11Col6a1 mice have a normal skeleton and in two models of inflammatory arthritis, erosions were significantly reduced, but not absent. In contrast to our report, TRAP-positive osteoclasts could not be identified in the inflamed joints of Tnfsf11Col6a mice. However, it is challenging to exclude that osteoclasts might have been present within the joint at some point prior to necropsy. Taken together, our results, along with previous studies, confirm a very important role for the RANKL:RANK axis in generating the osteoclasts that lead to inflammatory bone erosion. However, in a sufficiently inflamed environment, other cytokines may compensate to form osteoclast-like cells independent of RANK. Interestingly, denosumab, a RANKL blocking antibody, reduced erosions and bone turnover markers at 12 months in a phase II clinical trial (26). These data would seem to question the relevance of our data to human RA. However, it should be noted that patients in this study all received background methotrexate therapy, which may reduce the levels of TNFα and IL-6 required to induce osteoclasts in the absence of RANKL signaling. Other in vitro models of RANK-independent osteoclastogenesis have been reported. Experimental conditions where investigators have observed RANKL or RANK independent osteoclast formation in vitro include treatment of precursors with 1) TNFα in combination with IL-1 or TGFβ (19, 22), 2) other TNF-family members, such as April or Light, or insulin-like growth factors (20) or 3) secreted osteoclastogenic factor of activated T cells (SOFA). In addition, TNFα treatment of RBPJ deficient myeloid precursors is sufficient to induce osteoclastogenesis (21), as is overexpression of constitutively active IKKβ or NFATc1 (42, 43). Recently, Yokoto et al. demonstrated that TNFα/IL-6 could support osteoclast differentiation in vitro, and that infusion of these cytokines into mice led to local inflammatory bone loss (24). However, their report could not exclude the possibility that endogenous RANKL is generated in these microenvironments by TNFα/IL-6, supporting osteoclast differentiation and bone erosions in vivo in a RANK-dependent manner. Therefore, generation of functional RANK independent osteoclast-like cells in vitro with TNFα/IL-6, and in an animal model of arthritis, as we have shown here, supports the hypothesis that RANK signaling is not absolutely required for the osteoclastogenesis and bone loss observed during inflammation. While IL-6 is sufficient to promote osteoclastogenesis in the presence of TNFα, other activators of the IL-6R signaling subunit gp130, such as OSM and LIF, might also play an important role in the context of K/BxN serum transfer arthritis. Inflammation in this model does not depend on IL-6, as mice deficient in this cytokine develop inflammation similar to wild-type (40). In contrast, LIF-deficient mice are partially protected from inflammation and bone erosion in this model (44). Whether IL-6 contributes to the development of bone erosion in this model has not been established in our study. One might expect to observe IL-6 dependent erosions in the absence of RANK, a hypothesis we are currently addressing. However, since LIF and OSM, in addition to IL-6, are upregulated in the synovium of mice with K/BxN serum transfer arthritis, redundancy among IL-6 family members may make this experiment challenging. Our work and that of others demonstrates that a variety of inflammatory cytokines contribute directly to the generation of bone eroding osteoclasts. Our findings suggest that TNFα/IL-6, cytokines known to be elevated in RA, can drive RANK-independent osteoclast formation in vivo and in vitro. The demonstration of RANK-independent osteoclastogenesis has implications for prevention of erosions in inflammatory joint diseases and suggests that cytokine blockade, either with TNFα blockers or IL-6R blockers, could have the additional benefit of inhibiting RANK independent osteoclast formation. Supplementary Material Supp Fig S1 Figure S1. A semi-quantitative method for scoring forepaw erosions on microCT (A) Three dimensional reconstructions of fore paws imaged at 7 micron resolution were scored at 4 anatomic locations as noted by circles in the panels. (B) Erosions were scored on a 0–3 scale, as follows: 0 – normal cortical bone; 1 –a large cortical erosion without perforation into the marrow cavity (black arrow); 2 – a large cortical erosion with a small perforation (white arrow); and 3 – a large cortical erosion with a large perforation or multiple small perforations (red arrow). (C–D) Both intra- and inter-observer reproducibility was excellent with r=0.8944 and 0.9343, respectively. Supp Fig S2 Figure S2. TNFα/IL-6 osteoclastogenesis does not require Fcrγ in vitro (A) Quantitation of TRAP-positive MNCs generated from Fcrγ or WT BMMs in the presence of MCSF, or MCSF plus RANKL or TNFα/IL-6. The average of triplicate cultures of 3 biologic replicates is shown. (B) Representative TRAP stained cultures of WT and Fcrγ−/− BMMs cultured with MCSF plus RANKL or TNFα/IL-6. NS by two-way ANOVA with Sidak’s post-test for multiple comparisons. Supp Fig S3 Figure S3. IL-6 promotes BMM proliferation in the presence of TNFα (A) Quantitation of TRAP-positive MNCs generated from WT BMMs cultured with TNFα, plus IL-6 from day 0–4 or day 0–2 of culture. Data at each time point represent mean ± SEM from 3 biological replicates. *, p<0.01 by Student’s t test. (B) WT BMMs were cultured without cytokines for 1 day (No Stim), or in the indicated cytokines for 1–3 days. At the indicated time points, alamarBlue was added and fluorescence measured after 3 hours. Each data point represents the mean ± SEM of 3 biological replicates. No difference was observed comparing MCSF+TNFα+IL-6 to MCSF+TNFα, but for all other comparisons, p <0.001 by two way ANOVA with Tukey’s post-test. Supp Fig S4 Figure S4. Loss of RANK function within 4 weeks after polyI:C treatment of Rankfl/fl;Mx1-Cre+ mice Eight week old Rankfl/fl;Mx1-Cre− and Rankfl/fl;Mx1-Cre+ mice were treated with polyI:C and analyzed 4 weeks later. (A) 3D reconstruction of micro-CT scans demonstrates massive accumulation of trabecular bone in the secondary spongiosa of Rankfl/fl;Mx1-Cre+ mice. (B) Representative TRAP stain of BMMs cultured with MCSF and RANKL, demonstrating that loss of RANKL responsiveness in Rankfl/fl;Mx1-Cre+ mice occurs within 4 weeks of polyI:C treatment, 100X magnification. Supp Fig S5 Figure S5. Histology of serum transfer arthritis in RANK deficient mice (A) Cartilage damage is not altered by RANK deficiency. Scoring of cartilage damage on histologic sections from RankWT and RankΔ/Δ mice with K/BxN serum transfer arthritis. NS, not significant by Student’s t-test. (B) Representative images of TRAP stained distal tibia from RankWT (top panel) and RankΔ/Δ (bottom panel) mice with K/BxN serum transfer arthritis, 50X magnification. Scale bar represents 200μm. RankWT bones have osteoclasts both in erosions (white arrowheads) and in the bone marrow cavity (black arrowheads), whereas osteoclasts are seen only in erosions in RankΔ/Δ mice. We thank Genentech for the gift of cMR16-1 antibody. We thank Dr. Mary Nakamura (Dap12−/−) and Dr. Jessica Hammerman (FcRg−/−) for providing bones and Dr. Josef Penninger for providing the Rankfl/fl mice. Dr. Peter Nigrovic provided K/BxN serum. Imaging courtesy of the Confocal Microscopy Core Facility at Brigham and Women’s Hospital (Boston, MA). This work was supported by the following grants: This work was supported by NIH grants K08 AR062590 (JFC), R01 AR060363 (AOA, JFC) and R01 AG046257 (AOA, JFC), R01 AR055952 (EMG) and grants from the Burroughs Wellcome Fund (AOA), the Rheumatology Research Foundation (WOB, AOA, JFC) and the Bettina Looram Fund (JFC). Figure 1 BMMs differentiate into osteoclasts upon stimulation with TNFα/IL-6 (A) Representative TRAP stain of BMMs cultured with the indicated cytokines in the presence or absence of OPG, 100X magnification. (B) Quantitation of TRAP-positive multinucleated (≥3) cells (MNCs) generated as in panel A (Each data point represents average of 3 wells, n=3–6 mice per condition). Results in (A) and (B) are representative of 3 independent experiments. (C) qPCR for the indicated markers of osteoclast differentiation on BMMs cultured for 4 days with TNFα alone or with TNFα/IL-6, n=3 mice. (D) TRAP-stain of FACS sorted CD11b−/lowLy6chigh osteoclast precursors differentiated with TNFα/IL-6 or RANKL, n=3 mice. Results in (C) and (D) are representative of 2 independent experiments. Multiple comparison adjusted p values are reported as follows: , p<0.05; p≤0.0003; **, p≤0.0001 by one-way ANOVA with Sidak’s post-test for multiple comparisons. Figure 2 TNFα/IL-6-induced OC differentiation and resorption is RANK independent (A) TRAP stain of RankWT and RankΔ/Δ BMMs cultured with RANKL or TNFα/IL-6. (B) Quantitation of TRAP-positive MNCs generated as in (A) from triplicate cultures of each biologic replicate, n=3–6 mice per condition. Results in (A) and (B) are representative of 3 independent experiments. (C) Von Kossa stain of calcium phosphate coated plates incubated with RankWT and RankΔ/Δ BMMs cultured in the presence of the indicated cytokines, n=3 mice per condition. Results are representative of 2 independent experiments (D) CTX ELISA on culture supernatants of RankWT and RankΔ/Δ BMMs cultured in the presence of the indicated cytokines. Data are pooled from 2 independent experiments, n= 2–6 mice per condition. Multiple comparison adjusted p values are reported as follows: , p<0.03; ***, p≤0.0001 by Sidak’s multiple comparisons test of a two-way ANOVA (B) and one-way ANOVA (D). Figure 3 TNFα/IL-6-induced OC differentiation requires IL-6R signaling, NFATc1, DAP12, and cell proliferation (A) TRAP stain of WT BMMs cultured with TNFα, TNFα/IL-6, or TNFα/IL-6 plus either IgG2a isotype control or anti-IL6R antibody, 100X magnification. Quantitation of TRAP-positive MNCs cultured with TNFα/IL-6 alone, or TNFα/IL-6 with 0.2, 2 and 20μg/mL of IgG2a isotype control or anti-IL6R antibody (right panel). Triplicate cultures of each biologic replicate, n=3 mice, were quantified for each condition. (B) TRAP stain of Nfatc1WT or Nfatc1Δ/Δ BMMs cultured with RANKL or TNFα/IL-6 (C) Quantitation of TRAP-positive MNCs generated from Nfatc1WT or Nfatc1Δ/Δ BMMs (left panel), or Dap12+/+ or Dap12−/− BMMs (right panel), cultured with the indicated cytokines. Triplicate cultures from n=3 mice for Nfatc1Δ/Δ and n=2 for Dap12−/− were quantified. (D) Quantitation of TRAP-positive MNCs from BMMs cultured with the indicated cytokines in the presence and absence of HU. Triplicate cultures from n=3 mice were quantified for each condition. Results are representative of 3 independent experiments except for (E). Multiple comparison adjusted p values are reported as follows: , p<0.02; , p< 0.005; , p <0.0002, **, p≤0.0001 by Tukey’s (A) and Sidak’s (C) multiple comparisons test of a two-way ANOVA, and Sidak’s multiple comparisons test of a one-way ANOVA (D). Figure 4 RankΔ/Δ mice are not fully protected from inflammatory bone erosion (A) qPCR for IL6, Osm, and Lif on mRNA generated from the synovial tissue of WT mice 0 and 10 days after the initiation of K/BxN serum transfer arthritis (n=4–5 per time point). (B) Clinical scores (left panel) and change in paw thickness (right panel) from baseline of RankWT vs RankΔ/Δ mice after induction of K/BxN arthritis (n=10 and 12, respectively). (C) Upper panel shows H&E (left column) and TRAP stains (right 2 columns) of hindpaws from RankWT vs RankΔ/Δ 14 days after the initiation of K/BxN serum transfer arthritis. Arrows indicate erosion sites; arrowheads indicate osteoclasts in the marrow space of RankWT mice. Lower panel shows qualitative scoring of the histology shown for inflammation (left), bone erosion (middle), and TRAP staining (right) (D) Representative microCT images of forepaws from RankWT vs RankΔ/Δ 14 days after the initiation of K/BxN serum transfer arthritis with graph of the mean micro-CT erosion score for each mouse (lower panel). , p<0.05, , p<0.005 by Student’s t test. Figure 5 Loss of RANK expression in the BM and synovium of RankΔ/Δ mice with K/BxN serum transfer arthritis (A) qPCR for Rank expression in BM mRNA from RankWT vs RankΔ/Δ mice with arthritis (n=8–10); , p<0.0001 by Mann-Whitney test. (B) TRAP-stain of RANKL treated BMMs from RankWT vs RankΔ/Δ mice with arthritis. (C) Representative images of synovial cultures from RankWT vs RankΔ/Δ mice with arthritis cultured with RANKL or TNFα/IL-6 stained with TRAP (left panel) with quantitation of TRAP-positive multinucleated (≥3) cells (MNCs) (right panel). Results are representative of three independent experiments with the exception of A, where results are pooled from two independent experiments. Multiple comparison adjusted p values are reported as follows , p≤0.0002 by two-way ANOVA with Sidak’s multiple comparison test. E. Serial frozen sections of the hindpaw of a RankΔ/Δ mTmG mouse with arthritis stained for TRAP (left panel) or visualized by direct fluorescence (middle panel), 200X. S, synovium; B, bone; C, cartilage; JS, joint space; , erosion. Digital overlay of TRAP and fluorescence images (right panel), with far left panels showing 3x digitally magnified image of corresponding boxed area (red channel removed for clarity). Table 1 List of Primers used for qRT-PCR in this study All primers were designed using Primer-Blast(45). Acp5_290L 5′-CCAAAGAGATCGCCAGAACC-3′ Acp5_468R 5′-GTTTCCAGCCAGCACATACC-3′ Atp6vod2-307− 5′-TGCAGAGCTTCTTCCTCATCT-3′ Atp6vod2-307+ 5′-CAAAGCCAGCCTCCTAACTC-3′ Calcr F 5′-GCCTCCCCATTTACATCTGC-3′ Calcr R 5′-CTCCTCGCCTTCGTTGTTG-3′ Ctsk F 5′-GGGCTCAAGGTTCTGCTGC-3′ Ctsk R 5′-TGGGTGTCCAGCATTTCCTC-3′ Dcstamp_1255 5′-TTCTCGTGTCAGTCTCCTTCTAC-3′ Dcstamp_1440 5′-GGATTGTCTGCTTCTTCCTAATC-3′ Gapdh F 5′-TTCACCACCATGGAGAAGGC-3′ Gapdh R 5′-GGCATGGACTGTGGTCATGA-3′ Hprt-2 F 5′-GTTAAGCAGTACAGCCCCAAA-3′ Hprt-2 R 5′-AGGGCATATCCAACAACAAACTT-3′ mIL6_670 5′-GCACTAGGTTTGCCGAGT-3′ mIL6_446 5′-GCCAGAGTCCTTCAGAGA-3′ Itgb3 F 5′-TCCAGACCCTGGGTACCAAG-3′ Itgb3 R 5′-GCCAATCCGAAGGTTGCTAG-3′ mLIF F 5′-CAACTGGCACAGCTCAAT-3′ mLIF R 5′-TCTGTCATGTTAGGCGCA-3′ Nfatc1ex3-F 5′-TGCCTTTTGCGAGCAGTATCT-3′ Nfatc1ex3-R 5′-CAGGCAAGGATGGGCTCATAT-3′ mOSM F 5′-CAGTATGCAGACACGGCT-3′ mOSM R 5′-TTTTGGAGGCGGATATAG-3′ Slc4a2 exon 8 F 5′-CCCATGAGGTGTTTGTGG-3′, Slc4a2 exon 8 R 5′-TCCACATCCTCCTCGAATTT-3′. Tnfrsf11a F 5′-GCGCAACAGTGTTTCCACAG-3′ Tnfrsf11a R 5′-CGCTTGGATCACAGTAAGGCT-3′ Disclosures Anti-mouse IL-6 Receptor antibody (cMR16-1) was a gift of Genentech, San Francisco, CA. 1 Schett G Gravallese E Bone erosion in rheumatoid arthritis: mechanisms, diagnosis and treatment Nat Rev Rheumatol 2012 8 11 656 64 23007741 2 Ollier WE Harrison B Symmons D What is the natural history of rheumatoid arthritis? Best Pract Res Clin Rheumatol 2001 15 1 27 48 11358413 3 Navarro-Compan V Landewe R Provan SA Odegard S Uhlig T Kvien TK Relationship between types of radiographic damage and disability in patients with rheumatoid arthritis in the EURIDISS cohort: a longitudinal study Rheumatology (Oxford) 2015 54 1 83 90 25065011 4 Cohen G Gossec L Dougados M Cantagrel A Goupille P Daures JP Radiological damage in patients with rheumatoid arthritis on sustained remission Ann Rheum Dis 2007 66 3 358 63 16935911 5 Molenaar ET Voskuyl AE Dinant HJ Bezemer PD Boers M Dijkmans BA Progression of radiologic damage in patients with rheumatoid arthritis in clinical remission Arthritis Rheum 2004 50 1 36 42 14730597 6 Gravallese EM Manning C Tsay A Naito A Pan C Amento E Synovial tissue in rheumatoid arthritis is a source of osteoclast differentiation factor Arthritis Rheum 2000 43 2 250 8 10693863 7 Pettit AR Ji H von Stechow D Muller R Goldring SR Choi Y TRANCE/RANKL knockout mice are protected from bone erosion in a serum transfer model of arthritis Am J Pathol 2001 159 5 1689 99 11696430 8 Jones D Glimcher LH Aliprantis AO Osteoimmunology at the nexus of arthritis, osteoporosis, cancer, and infection J Clin Invest 2011 121 7 2534 42 21737885 9 Nakashima T Hayashi M Takayanagi H New insights into osteoclastogenic signaling mechanisms Trends in endocrinology and metabolism: TEM 2012 23 11 582 90 22705116 10 Humphrey MB Lanier LL Nakamura MC Role of ITAM-containing adapter proteins and their receptors in the immune system and bone Immunol Rev 2005 208 50 65 16313340 11 Koga T Inui M Inoue K Kim S Suematsu A Kobayashi E Costimulatory signals mediated by the ITAM motif cooperate with RANKL for bone homeostasis Nature 2004 428 6984 758 63 15085135 12 Sitara D Aliprantis AO Transcriptional regulation of bone and joint remodeling by NFAT Immunol Rev 2010 233 1 286 300 20193006 13 Shen J Shang Q Tam LS Targeting inflammation in the prevention of cardiovascular disease in patients with inflammatory arthritis Translational research : the journal of laboratory and clinical medicine 2015 14 Siebert S Tsoukas A Robertson J McInnes I Cytokines as therapeutic targets in rheumatoid arthritis and other inflammatory diseases Pharmacological reviews 2015 67 2 280 309 25697599 15 Danks L Komatsu N Guerrini MM Sawa S Armaka M Kollias G RANKL expressed on synovial fibroblasts is primarily responsible for bone erosions during joint inflammation Ann Rheum Dis 2015 16 Kim JH Jin HM Kim K Song I Youn BU Matsuo K The mechanism of osteoclast differentiation induced by IL-1 J Immunol 2009 183 3 1862 70 19587010 17 Lam J Takeshita S Barker JE Kanagawa O Ross FP Teitelbaum SL TNF-alpha induces osteoclastogenesis by direct stimulation of macrophages exposed to permissive levels of RANK ligand J Clin Invest 2000 106 12 1481 8 11120755 18 Yamashita Y Ukai T Nakamura H Yoshinaga Y Kobayashi H Takamori Y RANKL pretreatment plays an important role in the differentiation of pit-forming osteoclasts induced by TNF-alpha on murine bone marrow macrophages Archives of oral biology 2015 60 9 1273 82 26099662 19 Azuma Y Kaji K Katogi R Takeshita S Kudo A Tumor necrosis factor-alpha induces differentiation of and bone resorption by osteoclasts J Biol Chem 2000 275 7 4858 64 10671521 20 Hemingway F Taylor R Knowles HJ Athanasou NA RANKL-independent human osteoclast formation with APRIL, BAFF, NGF, IGF I and IGF II Bone 2011 48 4 938 44 21193069 21 Li S Miller CH Giannopoulou E Hu X Ivashkiv LB Zhao B RBP-J imposes a requirement for ITAM-mediated costimulation of osteoclastogenesis J Clin Invest 2014 124 11 5057 73 25329696 22 Kim N Kadono Y Takami M Lee J Lee SH Okada F Osteoclast differentiation independent of the TRANCE-RANK-TRAF6 axis J Exp Med 2005 202 5 589 95 16147974 23 Kim HR Kim KW Kim BM Jung HG Cho ML Lee SH Reciprocal activation of CD4+ T cells and synovial fibroblasts by stromal cell-derived factor 1 promotes RANKL expression and osteoclastogenesis in rheumatoid arthritis Arthritis & rheumatology 2014 66 3 538 48 24574213 24 Yokota K Sato K Miyazaki T Kitaura H Kayama H Miyoshi F Combination of tumor necrosis factor alpha and interleukin-6 induces mouse osteoclast-like cells with bone resorption activity both in vitro and in vivo Arthritis & rheumatology 2014 66 1 121 9 24431283 25 Li P Schwarz EM O’Keefe RJ Ma L Boyce BF Xing L RANK signaling is not required for TNFalpha-mediated increase in CD11(hi) osteoclast precursors but is essential for mature osteoclast formation in TNFalpha-mediated inflammatory arthritis J Bone Miner Res 2004 19 2 207 13 14969390 26 Cohen SB Dore RK Lane NE Ory PA Peterfy CG Sharp JT Denosumab treatment effects on structural damage, bone mineral density, and bone turnover in rheumatoid arthritis: a twelve-month, multicenter, randomized, double-blind, placebo-controlled, phase II clinical trial Arthritis Rheum 2008 58 5 1299 309 18438830 27 Schramek D Leibbrandt A Sigl V Kenner L Pospisilik JA Lee HJ Osteoclast differentiation factor RANKL controls development of progestin-driven mammary cancer Nature 2010 28 Aliprantis AO Ueki Y Sulyanto R Park A Sigrist KS Sharma SM NFATc1 in mice represses osteoprotegerin during osteoclastogenesis and dissociates systemic osteopenia from inflammation in cherubism J Clin Invest 2008 118 11 3775 89 18846253 29 Kuhn R Schwenk F Aguet M Rajewsky K Inducible gene targeting in mice Science 1995 269 5229 1427 9 7660125 30 Mocsai A Humphrey MB Van Ziffle JA Hu Y Burghardt A Spusta SC The immunomodulatory adapter proteins DAP12 and Fc receptor gamma-chain (FcRgamma) regulate development of functional osteoclasts through the Syk tyrosine kinase Proc Natl Acad Sci U S A 2004 101 16 6158 63 15073337 31 Coury F Zenger S Stewart AK Stephens S Neff L Tsang K SLC4A2-mediated Cl-/HCO3- exchange activity is essential for calpain-dependent regulation of the actin cytoskeleton in osteoclasts Proc Natl Acad Sci U S A 2013 110 6 2163 8 23341620 32 Charles JF Hsu LY Niemi EC Weiss A Aliprantis AO Nakamura MC Inflammatory arthritis increases mouse osteoclast precursors with myeloid suppressor function J Clin Invest 2012 122 12 4592 605 23114597 33 Okazaki M Yamada Y Nishimoto N Yoshizaki K Mihara M Characterization of anti-mouse interleukin-6 receptor antibody Immunology letters 2002 84 3 231 40 12413742 34 Walsh NC Reinwald S Manning CA Condon KW Iwata K Burr DB Osteoblast function is compromised at sites of focal bone erosion in inflammatory arthritis J Bone Miner Res 2009 24 9 1572 85 19338457 35 Monach PA Mathis D Benoist C The K/BxN arthritis model Current protocols in immunology Coligan John E 2008 Chapter 15 Unit 15 22 36 Nigrovic PA Malbec O Lu B Markiewski MM Kepley C Gerard N C5a receptor enables participation of mast cells in immune complex arthritis independently of Fcgamma receptor modulation Arthritis Rheum 2010 62 11 3322 33 20662064 37 Asagiri M Sato K Usami T Ochi S Nishina H Yoshida H Autoamplification of NFATc1 expression determines its essential role in bone homeostasis J Exp Med 2005 202 9 1261 9 16275763 38 Scherzad A Steber M Gehrke T Rak K Froelich K Schendzielorz P Human mesenchymal stem cells enhance cancer cell proliferation via IL-6 secretion and activation of ERK1/2 International journal of oncology 2015 47 1 391 7 25997536 39 Yu RY Wang X Pixley FJ Yu JJ Dent AL Broxmeyer HE BCL-6 negatively regulates macrophage proliferation by suppressing autocrine IL-6 production Blood 2005 105 4 1777 84 15507530 40 Ji H Pettit A Ohmura K Ortiz-Lopez A Duchatelle V Degott C Critical roles for interleukin 1 and tumor necrosis factor alpha in antibody-induced arthritis J Exp Med 2002 196 1 77 85 12093872 41 Redlich K Hayer S Ricci R David JP Tohidast-Akrad M Kollias G Osteoclasts are essential for TNF-alpha-mediated joint destruction J Clin Invest 2002 110 10 1419 27 12438440 42 Otero JE Dai S Alhawagri MA Darwech I Abu-Amer Y IKKbeta activation is sufficient for RANK-independent osteoclast differentiation and osteolysis J Bone Miner Res 2010 25 6 1282 94 20200955 43 Takayanagi H Kim S Koga T Nishina H Isshiki M Yoshida H Induction and activation of the transcription factor NFATc1 (NFAT2) integrate RANKL signaling in terminal differentiation of osteoclasts Dev Cell 2002 3 6 889 901 12479813 44 Upadhyay A Senyschyn D Santos L Gu R Carroll GJ Jazayeri JA K/BxN serum transfer arthritis is delayed and less severe in leukaemia inhibitory factor (LIF)-deficient mice Clin Exp Immunol 2012 169 2 71 8 22774981 45 Ye J Coulouris G Zaretskaya I Cutcutache I Rozen S Madden TL Primer-BLAST: a tool to design target-specific primers for polymerase chain reaction BMC Bioinformatics 2012 13 134 22708584
PMC005xxxxxx/PMC5125894.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 101623795 42112 Arthritis Rheumatol Arthritis & rheumatology (Hoboken, N.J.) 2326-5191 2326-5205 27337150 5125894 10.1002/art.39785 NIHMS794992 Article Human Gut-Derived Prevotella histicola Suppresses Inflammatory Arthritis in Humanized Mice Marietta Eric V PhD 1 Murray Joseph A MD 1 Luckey David H BS 2 Jeraldo Patricio R. PhD 4 Lamba Abhinav 1 Patel Robin MD 3 Luthra Harvinder S MD 5 Mangalam Ashutosh PhD 2* Taneja Veena PhD 25 1 Department of Medicine, Division of Gastroenterology and Hepatology, Mayo Clinic College of Medicine 2 Department of Immunology, Mayo Clinic College of Medicine 3 Department of Medicine, Division of Clinical Microbiology, Mayo Clinic College of Medicine 4 Department of Surgery, Mayo Clinic College of Medicine 5 Department of Medicine, Division of Rheumatology, Mayo Clinic College of Medicine Address correspondence and reprint requests to: Veena Taneja, Department of Immunology, Mayo Clinic, 200 First St. SW, Rochester, MN55905. Fax 507-266-0981, Tel 507 284 4541, [email protected] * Present Address: Department of pathology, University of Iowa 15 6 2016 12 2016 01 12 2017 68 12 28782888 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Objective The gut microbiome regulates host immune homeostasis. Rheumatoid arthritis (RA) is associated with intestinal dysbiosis. In this study we used a human gut-derived commensal to modulate immune response and treat arthritis in a humanized mouse model. Methods We have isolated a commensal bacterium, Prevotella histicola, native to the human gut that has systemic immune effects when administered enterally. Arthritis-susceptible HLA-DQ8 mice were immunized with type II collagen and treated with P. histicola; disease incidence, onset and severity were monitored. Changes in the gut epithelial proteins and immune response as well as systemic cellular and humoral immune responses were studied in treated mice. Results DQ8 mice when treated with P. histicola in prophylactic or therapeutic protocols exhibited significantly decreased incidence and severity of arthritis as compared to controls. The microbial mucosal modulation of arthritis was dependent on the regulation by CD103+ dendritic cells and myeloid suppressors, CD11b+Gr-1, and by generation of T regulatory cells, CD4+CD25+FoxP3+, in the gut, resulting in suppression of antigen-specific Th17 response and increased transcription of IL-10. Treatment with P. histicola led to reduced intestinal permeability by increasing expression of enzymes that produce antimicrobial peptides as well as tight junction proteins, Zo-1 and Occludin. However, the innate immune response via TLR4 and TLR9 were not affected in treated mice. Discussion Our results demonstrate that enteral exposure to P. histicola suppresses arthritis via mucosal regulation. P. histicola is a unique commensal that can be explored as a novel therapy for RA and may have low/no side effects. Rheumatoid arthritis (RA) is a chronic inflammatory joint disease that requires both genetic and environmental factors (1). Among the known genetic factors, the strongest association is with the presence of certain alleles of HLA class II molecules (2). Using transgenic mice expressing RA-associated HLA-DR4/DQ8 genes, we have developed a humanized model of inflammatory arthritis that shares similarities with human disease in sex-bias, autoantibody profile and phenotype (3). Our recent data suggests that gut microbial composition of naïve 0401 and 0402 mice shares similarities with the human mucosal microbiome (4) and the 0401 genotype may be associated with a dysbiosis of the gut microbiome. MHC polymorphism has been shown to impact gut flora in humans and mice (5–7). Studies in patients with RA have shown dysbiosis with one study showing decreased Bacteroides-Porphyromonas-Prevotella species compared to healthy controls (8, 9). How certain commensals suppress T-cell proliferation is not well understood; further studies are needed to more precisely determine their effects on the immune response in inflammatory diseases. We have isolated a gram negative anaerobe commensal bacterium, Prevotella histicola, native to oral, nasopharyngeal, gastrointestinal, and genito-urinary mucosal surfaces. P. histicola is a recently discovered species with taxonomic similarity to Prevotella melanogenica and Prevotella veroalis. While commensals like Bifidobacterium species and some species of Prevotella have been studied for their impact on the immune system, there are virtually no reports on the biological effects of P. histicola. Our studies suggest that P. histicola has immune modulating properties and suppresses inflammatory cytokines. We tested if orally administered P. histicola can modulate the immune response in the gut and if that can be translated systemically to control the arthritis in DQ8 mice. Our data suggests that oral feeding of P. histicola in a therapeutic protocol (after induction of arthritis) to DQ8 transgenic mice leads to resistance to develop disease and limits the disease severity. P. histicola alone did not lead to any enteric or other pathology in transgenic mice. These studies provide experimental support for the exploration of commensals as treatment options for systemic diseases, including RA. Material and Methods Isolation and Identification of Prevotella Species Biopsies from individuals were taken from the proximal small bowel and bacterial cultures grown on KV agar plates for isolation of individual colonies. Isolates were cultured on sheep blood agar plates and incubated at 35°C anaerobically for 2 days. Bacterial genomic DNA was extracted using the QIAamp DNA Mini kit (Qiagen). Real time, rapid cycle Light Cycler PCR with SYBR Green I detection (Roche Applied Science, Indianapolis, IN) was used to amplify 527-bp of the 16S rRNA gene. Universal bacterial 16S rRNA gene primers were used (Microseq® 500 16S rRNA gene PCR kit). PCR cycling was followed by a post-amplification melting curve analysis to verify the amplicon before sequencing. Sequencing was performed with the BigDye terminator version 1.1 Taq kit and an ABI 3730XL DNA sequencer (Applied Biosystems). Bidirectional sequence data were aligned using Sequencher (Gene Codes Corp). The generated consensus sequences were compared to those of the National Center for Biotechnology Information’s (NCBI) GenBank database. Identity ≥ 99% between the query sequence and the GenBank database with a difference > 0.4% between species was used for identification to the species level. Transgenic Mice Transgenic mice were generated as described previously (10). All mice used lacked endogenous class II molecules (AEo) and expressed DQB10302/DQA10301 (DQ8.AEo) on B6/129 background. Mice of both sexes (8–12 weeks of age) were used in this study and were bred and maintained in the pathogen-free Immunogenetics Mouse Colony at the Mayo Clinic, Rochester, MN in accordance with the Institutional Animal Use and Care Committee. All the experiments included transgene negative littermates as controls and were carried out with the approval of the Animal Use and Care Committee. Induction and Evaluation of Collagen Induced Arthritis (CIA) CIA was induced by immunization with type II collagen (CII) (100μg of CII emulsified 1:1 with complete Freund’s adjuvant) as previously described (10). Mice were monitored for the onset and progression of CIA using a grading system (range 0–3) wherein each paw was scored; 0= no swelling, 1=1or 2 digit swollen, 2=2 or more digits swollen and 3= swollen paw. The mean arthritic score was determined using arthritic animals only. Mice were divided in 2 groups for reproducibility. Histopathology of representative paws from each group was done to determine arthritis induction. DBA/1 mice were immunized with CII and the paw thickness was measured with calipers before and after induction of arthritis to determine the arthritis severity. Treatment with Commensal Bacterium Organisms (P. histicola, and P. melanogenica) were stored at −70C in skim milk, inoculated onto an Anaerobe Laked Sheep Blood Agar with Kanamycin and Vancomycin (KV) (Becton, Dickson) and incubated anaerobically in an anaerobic jar with AnaeroPack® system (Mitsubishi Gas Chemical America) and incubated at 37°C for 2–3 days. The bacterium was then swabbed into 10 mL of tryptic soy broth (TSB) and anaerobically incubated for 2 days prior to inoculation. The identity of the organisms was verified by PCR. Transgenic mice were then orally gavaged on alternate days with 1 × 109 live bacteria suspended in 100 microliters of TSB (anaerobic) bacterial culture. The dose was chosen based on the fact that a higher dose did not provide any additional benefit. For the preventive protocol, bacteria were administered 10 days prior to immunization with CII and continued for 6 weeks post-immunization. For the therapeutic protocol, mice were treated 2 weeks post CIA induction and continued for 6 weeks. Control sham gavage consisted of administering 100ul of bacterial media alone. DBA/1 mice were treated 7 days post immunization. Prevotella did not colonize the gut (not shown). Isolation of Lamina Propria Cells Intestinal tissue was cut longitudinally using a scalpel blade, and washed six times using CMF solution (88% 1X Hanks balanced salt solution, 10% HEPES-bicarbonate buffer, 2% FBS. A 1 hour collagenase digestion using Complete RPMI-10/collagenase (1.33mg/ml) solution released lymphocytes from the intestinal tissue. This mixture was passed through a nylon filter, centrifuged, and the pellet containing the lamina propria cells was suspended in Complete RPMI-10 with gentamycin. Intestinal Permeability All transgenic mice were kept on standard diet. Changes in intestinal permeability were determined using 4-KDa FITC-labeled dextran. Mice were deprived of food for 3 hours, then gavaged with FITC–labeled dextran (0.6 mg/g body weight). Three hours later, mice were bled and serum collected. FITC-dextran was determined at 490 nm. Gut permeability was tested in age and sex matched treated and control mice 8 weeks after induction of arthritis. Staining for Tight Junction Proteins Various parts of the gut, duodenum, jejunum, ileum and colon, were frozen at the termination of the experiments and tested for the expression of tight junction proteins Zonulin-1 (ZO-1) and Occludin by immunofluorescence using purified anti-Zonulin-1 (Life Technologies), Alexa Flour 594 conjugated anti-occludin (Life Technologies), and FITC conjugated anti rabbit IgG (Jackson ImmunoResearch Laboratories). Caco-2 cells (3×105) were cultured on 22×22 mm coverslips in 6 well tissue culture plates at 37°C in humidified 5% CO2 incubator till the cultures were confluent monolayers. They were then incubated with or without P. histicola (100 μl of 1×108 CFU/ml) for 4 hours, fixed with 10% formaldehyde and evaluated for Zo-1 and Occludin expression. rtPCR for Cytokine and Chemokine expression RNA was extracted from cells using RNAeasy columns (Qiagen) and cDNA prepared using RNase H-reverse transcriptase (Invitrogen) and cDNA generated by standard methods. The expression level of each gene was quantified using the threshold cycle (Ct) method normalized for Actin, a housekeeping gene. Affymetrix mouse PAMM073 microarrays were as per manufacturer’s instructions. The data was analyzed as per the online resources of the manufacturer. Collagen Specific ELISA Mice were bled after CII-immunization before and after treatment with P. histicola. Titers of sera IgG antibodies against CII were measured by standard ELISA and are shown as optical density. T Cell Proliferation Assay For the T cell proliferation assay, mice were immunized with 200ug of CII emulsified 1:1 in CFA (Difco) intradermally at the base of the tail and proliferation was done as described (10). For some experiments, CD4+ cells (5×106) sorted from lymph nodes of CII-primed mice that were treated with P. histicola or media only were cultured in vitro in the presence or absence of the antigen and CD11c+ dendritic cells (5×105) harvested from spleens. Stimulation index of 2 or more was taken as a positive response. Response to CPG (cytosine connected to guanine through phosphodiester bonds) (1 μg/ml) and LPS (lipopolysaccharide) (5μg/ml) was tested in mice receiving P. histicola treatment for 10 days prior to CII-immunization or 10 days post immunization and followed for 2 weeks of treatment. Also, mice in in vivo protocol were tested for response to LPS and CPG. Flow Cytometry The expression of DQ in transgenic mice was analyzed by flow cytometry using mAb IVD12 (anti-DQ). Conjugated antibodies for CD3, CD4, CD11c, CD19, CD25, GITR, Gr-1 and B220 (BD Biosciences, CA) were also used. All experiments were done with cells pooled from 2 mice/strain and repeated 2–3 times. Intracellular staining for FoxP3 and IL-10 was performed using specific antibodies obtained (eBioscience, San Diego, CA) as per the manufacturer’s instructions. Phycoerythrin-conjugated (PE) rat IgG2a (eBioscience) was used as the isotype control for FoxP3 staining. Analysis was done using the Cell Quest program (Beckton Dickinson). Cytokines Cytokines were measured using the Bio-Plex protein array system with the mouse cytokine 23-plex panel as per manufacturer’s instructions, and were analyzed with Bio-Plex manager 2.0 software (Bio-Rad laboratories, Hercules, CA). Some cytokines were also tested by Capture ELISA using commercial kits (BD biosciences). Statistical Analysis The difference in the incidence of arthritis between groups was analyzed using Chi square test. Antibody levels, onset of arthritis, mean scores for arthritic mice and various cells were compared using non-parametric Student’s T test. RESULTS Gut-Derived P. histicola Modulates Immune Response P. histicola and P. melanogenica isolated from the duodenum of an individual were cultured and tested for pathogenic properties. None of the mice developed weight loss greater than 5% of original weight. Also, none of the mice developed any gut pathology, villous atrophy, or crypt hyperplasia. Mice gavaged with P. histicola showed significantly reduced IL-2, IL-17, TNFα, and increased IL-4 and IL-10 compared to sham controls (Figure 1). Also proinflammatory chemokines involved in various autoimmune diseases, GM-CSF and MCP-1, were suppressed demonstrating a pro-biotic effect of P. histicola isolate in the DQ8 mice. On the other hand P. melanogenica gavaged mice did not show any significant changes in cytokines but did show reduced MIP-1b and MCP-1 levels compared to sham controls. P. histicola Suppresses Arthritis in Susceptible DQ8 Mice by Modulating Cellular and Humoral Response P. histicola was tested for treating CIA in DQ8 mice. HLA-DQ8 mice were immunized with CII and some mice were gavaged with P. histicola either before immunization (prophylactic) or after the development of arthritis (therapeutic) or media. Mice gavaged with P. histicola without CII immunization were used as controls (Figure 2). Both the prophylactic (P. histicola and CII) as well as therapeutic (CII and P. histicola) protocols resulted in statistically significant lower incidences of arthritis (P<0.05) as compared to sham controls (P<0.05). Mice treated with P. histicola developed milder arthritis with delayed onset compared to sham controls. However, P. melanogenica did not provide any protection from developing CIA. No inflammation in the small intestine or colon of P. histicola treated mice was observed while only mild shortening of villi with mild infiltration occurred with P. melanogenica treatment (Figure 2B). We next determined if treatment with P. histicola modulates the antigen-specific systemic immune response thereby resulting in protection from CIA. Sera from mice before and after treatment were tested for the presence of anti-CII antibodies in mice used in the therapeutic protocol. Mice treated with P. histicola showed a significant reduction in anti-CII antibodies. In addition, the antigen-specific T cell response was lower in treated mice with and without arthritis compared to arthritic controls (Figure 2C, D). Treatment with P. histicola Changes the Systemic But Not the Innate Immune Response Treating mice with immunomodulatory agents are known to cause suppression of innate responses leading to infections (11). We assessed whether treatment with P. histicola modulates the innate immune responses in in vivo model by culturing splenocytes with LPS or CpG. Although immunized and treated mice generated a higher response than naïve mice, administration of P. histicola did not significantly change the response to CpG or LPS (Figure 3A). A near significant decrease in the antigen specific T cell response to CII was observed in treated mice, suggesting that P. histicola suppresses CIA by changing the systemic immune response rather than causing immune suppression (Figure 3B). Treated mice showed significantly lower levels of serum IL-17 as well as the regulating cytokines IL-9, IL-13 and IL-12(p40) as compared to sham treated mice (Figure 3C).. P. histicola Treatment Modulates Immune Response in the Gut We next evaluated whether P. histicola treatment affects the mucosal immune system locally in the gut in vivo by determining mRNA transcripts for various cytokines from all parts of the gut (Figure 4). Treated mice as well as sham mice did not show significant changes in cytokine expression in the ileum The jejunum and colon of treated mice showed most cytokines being suppressed, P. histicola treated mice without arthritis showed much higher IL-10 and lower TGFβ levels compared to controls. The duodenum had high expression of most cytokines in both groups. Heat maps of intestinally derived mRNA transcripts of Th17 regulatory network revealed that P. histicola treatment led to changes in cytokine expression; arthritic control and treated mice showed similarities compared to non-arthritic treated mice (not shown). We further compared the effect of treatment on the duodenum, jejunum, ileum, and colon in the treated, arthritic or non-arthritic, and control groups (Figure 4A–D shows data with 5 fold or more difference). P. histicola treated mice showed suppression of all cytokines in the jejunum (except IL-10), and colon compared to the duodenum (Fig 4 B, D). Cytokines in P. histicola treated non-arthritic mice compared to controls showed more than 5-fold reductions in IL-17 and other pro-inflammatory cytokines with an increase in IL-10 (Fig 4 B–D). Interestingly, the major difference between P. histicola treated arthritic and non-arthritic mice was the increase in the levels of cytokines in the duodenum and decrease in the ileum along with an increase in the regulatory cytokines, IL-10 and FoxP3, in the jejunum and colon of the latter (not shown). Treatment with P. histicola Generates Treg Cells via Dendritic Cells and Modulates Antigen Presentation Next we determined if the P. histicola treatment modulates arthritis via IL-10 producing T regulatory cells in the gut and also systemically. Treated mice showed a consistent, although non-significant, increase in the total number of splenic CD4+ Tregs and IL-10 producing Tregs as compared to control mice even though the CD4+ cell numbers were similar (Figure 4E). CD103+ intestinal dendritic cells (DCs) maintain a tolerant state in the intestine by inducing regulatory T cell differentiation (12). Mice treated with P. histicola had increased numbers of CD103 expressing DCs in the lamina propria (P<0.05) (Figure 4F). Increased CD11c+CD103+ cells were also observed in splenocytes of treated mice, P<0.05 (Figure 4G) which might suggest that these DCs could have migrated from the intestine. Splenic DCs from treated mice could modulate the in vitro T cell response by culturing sorted CD4+ cells from the treated and control mice in a crisscross manner in the presence or absence of CII (Figure 5A). Antigen presentation by DCs from P histicola treated mice showed a significantly lower response and CD4+ cells produced undetectable levels of IL-17 (Figures 5A, B) suggesting both DCs and T reg cells may be involved in modulation of antigen-specific response. This data corroborates the increased CD103+ DCs and Tregs, and altered cytokine profile in the gut and periphery (Figure 4). P. histicola treated mice showed an increase in absolute numbers of both CD103+ DCs as well as myeloid CD11b+Gr-1+ cells, suggested to be suppressors (13, 14), (Figure 5C), although the differences were significant only in percentages and not absolute numbers. P. histicola Treatment Lowers the Gut Permeability Arthritis-susceptible humanized mice have enhanced gut permeability (4). Comparison of gut permeability in naïve, treated and control mice showed that. P. histicola treated mice had a significantly lower gut permeability compared to controls, P=0.02 (Figure 5D). Also, treated mice showed an increase in the expression of tight junction protein Zo-1 in the colon, ileum and jejunum as compared to sham control mice (Figure 5E). P. histicola treated arthritic and non-arthritic mice had much higher expression of the tight junction protein Occludin compared to sham treated mice (Figure 5F) suggesting that P. histicola regulates tight junction proteins. Suppression of Arthritis in P. histicola Treated DBA/1 mice To confirm our findings, we also tested this therapy in a commonly used mouse model of arthritis, DBA/1 mice. All CII-immunized mice developed arthritis. However, mice treated with P. histicola 7 days post-immunization developed a significantly milder disease compared to sham controls (Figure 6A), suggesting that amelioration of arthritis by this commensal is not genotype specific. P. histicola Treatment Leads to Increased Expression of Zo-1 by Epithelial Cells We next determined the effect of P. histicola on human derived epithelial Caco-2 cells by measuring expression of various chemokines and cytokines transcripts (Figure 6). The treatment with P. histicola increased Zo-1 and Occludin expression in Caco-2 cells as compared to media control, though it was not significant (Figure 6B). There was an increase in mRNA transcripts in treated epithelial cells for IL17RB, a receptor not associated with production of pro-inflammatory IL-17A, required for production of Th1 cytokine (Figure 6C). However, the most significant increase was observed in the expression of glyceraldehyde 3-phosphate dehydrogenase (GAPDH) while ACTB did not show any difference (not shown). P. histicola is Novel and Differs from Prevotella copri A recent study has shown an expansion of P. copri in new onset RA patients (9). We used whole-genome average Nucleotide Identity (ANI) to compare differences between the sequences from various known P. histicola strains and P. copri with the P. histicola isolate used in this study. As shown (Figure 6D), the P. histicola isolate used in this study is very different from P. copri with only 69% ANI. The previously available P. histicola strains, JCM_15637_JCVI, JCM_15637_UT and F0411, and the isolate used here have 97% ANI. We assembled and generated a Maximum Likelihood phylogenetic tree of the 16S ribosomal RNA gene of all the different genomes tested which suggested that the P. histicola isolate used in this study is novel. Comparison of the functional ORFs between the 2 species showed very low identity in genes when tested by BLASTp. Discussion Recent studies have highlighted the impact of dysbiosis in the gut microbiome on systemic inflammatory diseases including RA in patients and arthritis model in transgenic mice (4, 9, 15, 16). Interestingly, while one study showed an association with Prevotella species, another showed a lack of Prevotella, Bacteroidetes phylum, in fecal samples suggesting various species of Prevotella might have different effects on arthritis (8, 9). A comparison with P. copri, a disease associated species, suggested that our P. histicola isolate as well as other P. histicola strains differ from P. copri. Moreover, we isolated P. histicola from duodenum while other studies used stool samples; commensals isolated from different sites may have different functions. In transgenic mice, dysbiosis of the gut microbiome is associated with pro-inflammatory conditions locally implicating a bottom-up approach (driven by the gut microbiome) such that the adaptive immune system may be modulated by the gut immune system (4). Studies with germ-free and specific pathogen-free mice suggesting disruptions to gut microbiota can modulate systemic phenotype further support this contention (17). A role of the gut residing bacteria in causation of arthritis was shown in germ-free mice where a single species could promote expansion of intestinal Th17 cells, resulting in development of arthritis (17, 18). Further studies showed that inflammatory microbiota driven signals favor maintenance and proliferation of autoimmune CD4+ T cells (19). These studies strongly highlight the concept that the gut microbiota plays a role in causation of arthritis, thereby suggesting that commensals can be used for modulating immune responses locally and systemically. We took advantage of the HLA transgenic mouse model to test if a gut-derived commensal, that is observed with lower numbers in arthritis-susceptible transgenic mice and human RA (4, 8), can be used for treating CIA as a preclinical model for RA. Our hypothesis was that it would modulate dysbiosis and result in immune homeostasis in the gut that can be translated systemically for modulating disease outcome. Our additional data with P. melanogenica, which belongs to the Bacteroidetes phyla and was also isolated from duodenal biopsy samples of patients with celiac disease, suggest that not all Prevotella species may be suppressive. Arthritis-susceptible HLA-DQ8 mice treated with the two species of Prevotella demonstrated that treating mice with P. histicola leads to resistance to disease development and limits severity of disease without causing any pathology while P. melanogenica did not have this effect. We believe that these studies are relevant to humans as the effect of the treatment can be studied in vivo under normal physiological conditions and secondly, the DQ8-restricted effector arm of the immune response has similarities with human disease in phenotype and autoantibody profile. (3, 10, 20, 21). The major drawback of the available biologic drugs used for treating RA is that they suppress the immune response such that an individual’s capability of fighting infection is undermined. Our data indicates that treatment with P. histicola does not lead to the suppression of innate responses via CPG, sequences found in bacterial and viral genomes that bind to TLR9, and via LPS, a ligand for TLR4. However, studies to show that mice treated with P. histicola are not at risk for developing infectious diseases need to be done. Commensal bacteria and probiotics have been shown to exert their anti-inflammatory effect through production of IL-10, and the Th2 cytokines IL-25, IL-33 or thymic stromal lymphopoietin (TSLP) as well as via induction of regulatory cells (22–29). There are several putative mechanisms by which luminal applied microbiome therapy could affect inflammation distal from the gut, 1) regulatory cytokines produced by Tregs or suppressive DCs in the gut travel to the target organ, 2) expansion of regulatory cells that traffic to the site of inflammation and 3) change in gut permeability. Our data suggested that all of these 3 mechanisms could be occurring in mice treated with P. histicola. A comparison of various parts of the guts demonstrated that treatment with P. histicola led to suppression of cytokines in the jejunum, colon and ileum although the duodenum showed an increased expression in comparison to controls suggesting that P. histicola may control immune response differently in various parts of the gut. This change in expression of cytokines was associated with an increase in lamina propria Treg cells, CD103+ DCs and CD11c+F4/80+ cells, the latter having been shown to produce IL-10 and induce differentiation of T cells into Treg cells in lamina propria (30). Suppressive DCs can stimulate CD4+ T cells and reestablish the Th1/Th2 ratio to a “normal” level (31). Our data demonstrated that P. histicola has a potent modulatory effect upon the systemic production of inflammatory cytokines. IL-13 and IL-17 are involved in pathogenesis of RA and CIA in humanized mice (10, 32, 33) it is likely that P. histicola suppresses CIA by modulating the immune response to inflammatory cytokines as both these cytokines were produced lower than controls in in vivo model. An increase in IL-10 levels and T regulatory cells in LP and spleen do support this contention and further provide one of the mechanisms by which P histicola modulates arthritis phenotype. Arthritis as well as the medications used for treating RA, NSAIDS, are associated with increased gut permeability (34, 35). The observations of lowered gut permeability and increased expression of ZO-1 in treated compared to control mice as well as in epithelial cells suggest that P. histicola protects by preserving gut epithelium integrity in the context of inflammation. Further, P. histicola-mediated increase in expression of GAPDH and IL17RB in epithelial cells may be involved in protection from inflammation. Recently, GAPDH-derived anti-microbial peptides (AMPs) have been identified (36, 37). IL-17RB is a cytokine receptor that does not bind IL-17A, known to be involved in RA, and is not associated with RA pathogenesis rather leads to Th2 response (38, 39). Recent work has suggested that transient increase in colonic permeability in the presence of normal commensal organisms may provide protection against subsequent colitis, again suggesting the importance of commensal organisms in immune homeostasis and a beneficial or anti-inflammatory response in the context of an inflammatory stimulus. The hygiene hypothesis suggests a reduced bacterial burden has led to an increase in autoimmunity. However, the colon, which is replete with large quantities of bacteria, is less likely to be affected by this hygiene than the upper GI tract, which is the first portal of entry of foreign bacteria and bacterial products. Administering commensals to exact an effect on systemic immune responses through their interaction with the small intestine may be more germane for modifying systemic autoimmune responses and could provide an experimental framework to explain how the increase in environmental hygiene could result in an increase of autoimmune diseases. Our data suggests that P. histicola has potent probiotic properties, at least in this mouse model, and should be explored further for its beneficial effects for treating inflammation. Funding Source: The work was supported by funds from the Department of Defense grant, W81XWH-10-1-0257, and NIH grant AR30752 to VT. The technology used in this manuscript was invented by EVM, JM, AM, SB, and VT. We thank Dr. Chella David for providing transgenic mice and helpful suggestions for this study, Dr. Susan Barton helped in isolating and Melissa J Karau helped in culturing Prevotella histicola. The work presented here was supported by the funds from the Department of Defense grant W81XWH-10-1-0257 and NIH grant AR30752 to VT Figure 1 Oral treatment with P. histicola modulates systemic immune response in DQ8.Aβo mice. Serum levels of cytokines and chemokines were tested in naïve DQ8 transgenic mice gavaged with bacterial media alone, Sham (■) P. histicola ( ) or P. melanogenica ( ) every other day for two weeks. denotes p<0.05 comparison between P. histicola treated and sham mice (N=5–7 each experimental group). Figure 2 Oral treatment with P. histicola protects DQ8 mice from Collagen Induced Arthritis. P. histicola was administered to DQ8 mice in therapeutic protocol (CII+ P. histicola) (N=21) or as a prophylactic measure (P. histicola + CII), (N=12). P. melanogenica was administered as a species control for the therapeutic protocol (CII + P. melanogenica) (N=10). Other controls consisted of oral gavage with P. histicola without CII immunization (P. histicola) (N=12) and CII immunization with gavage of bacterial culture media alone (CII+ Media) (N=18) in DQ8 mice. (A) Incidence of arthritis in various experimental groups described above. Difference between CII-immunized and media versus P. histicola treated mice, incidence and severity, P<0.05 (B) Hematoxylin and eosin staining of the small intestine is shown from a representative mouse for each of the three treatment groups (P. histicola treated, sham (CII+ media) treated, and P. melanogenica treated). N=3 each group (C) Levels of serum anti-CII (IgG) antibodies before and after the administration of P. histicola using the therapeutic protocol. (D) T cell proliferative response to CII in vitro at the termination of the experiment at 12 weeks using splenocytes in mice with collagen-induced arthritis (CIA+) and without (CIA-). Figure 3 Treatment with P. histicola modulates immune response in an antigen-specific manner in CIA. Splenocytes harvested from DQ8 mice in different conditions as indicated were tested for in vitro proliferative response to A) CPG and LPS, and B) CII in mice treated with P. histicola, prophylactic ( P. histicola +CII) and therapeutic (CII+ P. histicola). C) Serum cytokines produced by DQ8 transgenic mice induced for arthritis and treated with culture media (sham) or P. histicola in therapeutic protocol (P. hist.) (N=4 each group). Figure 4 Anti-inflammatory effects of P. histicola treatment upon the Intestinal immune system. Fold change in cytokine transcript levels in arthritic (CIA+) and non-arthritic (CIA-) DQ8 mice treated with P. histicola treatment (therapeutic) as compared to the control group in the A) duodenum, B) jejunum, C) ileum, and D) colon. Expression levels of cytokine transcripts in duodenum, jejunum, ileum and colon of arthritic mice immunized with CII and gavaged media (Sham), and mice immunized with CII and treated with P. histicola in therapeutic protocol, arthritic (group 1) and non-arthritic (group 2) (N=3 each group). Control group was used as the reference. E) the absolute numbers of regulatory T cells (CD4+CD25+FoxP3+) producing IL-10 in the spleens of mice immunized with CII and treated with P. histicola or bacterial media alone (sham) Sham. FACS histogram of regulatory DCs, CD11c+CD103, in F) lamina propria cells and G) splenocytes. Also, shown is FACS histogram of CD4+GITR+ regulatory T cells in spleens of CII-immunized and treated or sham mice. (N=3–4 mice per group, experiment representative). Figure 5 CD4+ cells from P. histicola treated mice generate lower antigen-specific T cell response compared to controls. A) In vitro T cell response to CII in splenic CD4+ cells of DQ8 mice immunized with CII and gavaged with P. histicola (P. hist.) with CD11c+ cells (DCs) from same mouse or control mouse, CII immunized mice gavaged with bacterial media alone (sham) or vice versa, B) IL-17 production in supernatant from the culture in 5A, and C) regulatory dendritic cells, CD11c+CD103+ cells and myeloid suppressors, CD11b+Gr-1+. (N=4 each group) were enumerated from splenocytes from sham and P. histicola treated mice. D) Gut permeability, done by FITC-Dextran, in naïve (N=12) and CII-immunized mice treated with P. histicola (N=12) or not (Sham) (N=8. E) Expression of ZO-1 in intestinal sections of control and P. histicola treated DQ8 mice induced for arthritis. F) Expression of Occludin in arthritic DQ8 mice (a), arthritic mice administered P. histicola (b) and naïve mice treated with P. histicola (c). Sections are shown at 60× magnification. Mean florescence intensity (MFI) of expression is depicted below. Figure 6 P. histicola treatment suppresses arthritis by increasing expression of tight junction protein. A) DBA/1 mice treated with P. histicola develop milder arthritis compared to controls. DBA/1 mice were induced for arthritis and treated with P. histicola or media 7 days post immunization (N=10) and monitored for arthritis. From day 30 onwards, a significant difference in paw severity was observed between treated and control mice. P. histicola treated Caco-2 cells show increase expression of B) Zo-1 (left panel) and Occludin (right panel) and merged (middle panel) and C) fold expression of mRNA transcripts of certain cytokines in treated as compared to Caco-2 cells in media only. D) The ANI plot comparing whole-genome average nucleotide identity as calculated using the windowed blast method and maximum. E) Likelihood phylogenetic tree of the 16S rRNA gene of all the available different genomes and the P. histicola isolate from the present study. No authors declare any conflicts 1 Klareskog L Padyukov L Lorentzen J Alfredsson L Mechanisms of disease: Genetic susceptibility and environmental triggers in the development of rheumatoid arthritis Nature clinical practice 2006 2 425 33 2 Gregersen PK Silver J Winchester RJ The shared epitope hypothesis. An approach to understanding the molecular genetics of susceptibility to rheumatoid arthritis Arthritis and rheumatism 1987 30 1205 13 2446635 3 Behrens M Trejo T Luthra H Griffiths M David CS Taneja V Mechanism by which HLA-DR4 regulates sex-bias of arthritis in humanized mice J Autoimmun 2010 35 1 9 20061120 4 Gomez A Luckey D Yeoman CJ Marietta EV Berg Miller ME Murray JA Loss of sex and age driven differences in the gut microbiome characterize arthritis-susceptible 0401 mice but not arthritis-resistant 0402 mice PLoS One 2012 7 e36095 22553482 5 Toivanen P Vaahtovuo J Eerola E Influence of major histocompatibility complex on bacterial composition of fecal flora Infection and immunity 2001 69 2372 7 11254595 6 Vaahtovuo J Toivanen P Eerola E Bacterial composition of murine fecal microflora is indigenous and genetically guided FEMS microbiology ecology 2003 44 131 6 19719658 7 De Palma G Capilla A Nadal I Nova E Pozo T Varea V Interplay Between Human Leukocyte Antigen Genes and the Microbial Colonization Process of the Newborn Intestine Current issues in molecular biology 2009 12 1 10 19478349 8 Vaahtovuo J Munukka E Korkeamaki M Luukkainen R Toivanen P Fecal microbiota in early rheumatoid arthritis The Journal of rheumatology 2008 35 1500 5 18528968 9 Scher JU Sczesnak A Longman RS Segata N Ubeda C Bielski C Expansion of intestinal Prevotella copri correlates with enhanced susceptibility to arthritis Elife 2013 2 e01202 24192039 10 Taneja V Behrens M Mangalam A Griffiths MM Luthra HS David CS New humanized HLA-DR4-transgenic mice that mimic the sex bias of rheumatoid arthritis Arthritis Rheum 2007 56 69 78 17195209 11 Winthrop KL Risk and prevention of tuberculosis and other serious opportunistic infections associated with the inhibition of tumor necrosis factor Nature clinical practice Rheumatology 2006 2 602 10 12 Jaensson E Uronen-Hansson H Pabst O Eksteen B Tian J Coombes JL Small intestinal CD103+ dendritic cells display unique functional properties that are conserved between mice and humans J Exp Med 2008 205 2139 49 18710932 13 Marigo I Dolcetti L Serafini P Zanovello P Bronte V Tumor-induced tolerance and immune suppression by myeloid derived suppressor cells Immunol Rev 2008 222 162 79 18364001 14 del Rio ML Bernhardt G Rodriguez-Barbosa JI Forster R Development and functional specialization of CD103+ dendritic cells Immunol Rev 2010 234 268 81 20193025 15 Hakansson A Molin G Gut microbiota and inflammation Nutrients 2011 3 637 82 22254115 16 Vaahtovuo J Munukka E Korkeamaki M Luukkainen R Toivanen P Fecal microbiota in early rheumatoid arthritis The Journal of rheumatology 2008 35 1500 5 18528968 17 Wu HJ Ivanov II Darce J Hattori K Shima T Umesaki Y Gut-residing segmented filamentous bacteria drive autoimmune arthritis via T helper 17 cells Immunity 2010 32 815 27 20620945 18 Ivanov II Atarashi K Manel N Brodie EL Shima T Karaoz U Induction of intestinal Th17 cells by segmented filamentous bacteria Cell 2009 139 485 98 19836068 19 Chappert P Bouladoux N Naik S Schwartz RH Specific gut commensal flora locally alters T cell tuning to endogenous ligands Immunity 2013 38 1198 210 23809163 20 Taneja V Taneja N Behrens M Griffiths MM Luthra HS David CS Requirement for CD28 may not be absolute for collagen-induced arthritis: Study with HLA-DQ8 transgenic mice Journal of Immunology 2005 174 1118 25 21 Taneja V Taneja N Paisansinsup T Behrens M Griffiths M Luthra H CD4 and CD8 T cells in susceptibility/protection to collagen-induced arthritis in HLA-DQ8-transgenic mice: Implications for rheumatoid arthritis Journal of Immunology 2002 168 5867 75 22 Liu YJ Soumelis V Watanabe N Ito T Wang YH de Malefyt RW TSLP: an epithelial cell cytokine that regulates T cell differentiation by conditioning dendritic cell maturation Annual review of immunology 2007 25 193 219 23 Moore KW de Waal Malefyt R Coffman RL O’Garra A Interleukin-10 and the interleukin-10 receptor Annual review of immunology 2001 19 683 765 24 Saenz SA Taylor BC Artis D Welcome to the neighborhood: epithelial cell-derived cytokines license innate and adaptive immune responses at mucosal sites Immunol Rev 2008 226 172 90 19161424 25 Braat H van den Brande J van Tol E Hommes D Peppelenbosch M van Deventer S Lactobacillus rhamnosus induces peripheral hyporesponsiveness in stimulated CD4+ T cells via modulation of dendritic cell function The American journal of clinical nutrition 2004 80 1618 25 15585777 26 Pessi T Isolauri E Sutas Y Kankaanranta H Moilanen E Hurme M Suppression of T-cell activation by Lactobacillus rhamnosus GG-degraded bovine casein International immunopharmacology 2001 1 211 8 11360922 27 Pessi T Sutas Y Saxelin M Kallioinen H Isolauri E Antiproliferative effects of homogenates derived from five strains of candidate probiotic bacteria Applied and environmental microbiology 1999 65 4725 8 10543777 28 Schultz M Linde HJ Lehn N Zimmermann K Grossmann J Falk W Immunomodulatory consequences of oral administration of Lactobacillus rhamnosus strain GG in healthy volunteers The Journal of dairy research 2003 70 165 73 12800870 29 Konieczna P Groeger D Ziegler M Frei R Ferstl R Shanahan F Bifidobacterium infantis 35624 administration induces Foxp3 T regulatory cells in human peripheral blood: potential role for myeloid and plasmacytoid dendritic cells Gut 2012 61 354 66 22052061 30 Denning TL Wang YC Patel SR Williams IR Pulendran B Lamina propria macrophages and dendritic cells differentially induce regulatory and interleukin 17-producing T cell responses Nat Immunol 2007 8 1086 94 17873879 31 Surana NK Kasper DL The yin yang of bacterial polysaccharides: lessons learned from B. fragilis PSA Immunol Rev 2012 245 13 26 22168411 32 Hitchon CA Alex P Erdile LB Frank MB Dozmorov I Tang Y A distinct multicytokine profile is associated with anti-cyclical citrullinated peptide antibodies in patients with early untreated inflammatory arthritis The Journal of rheumatology 2004 31 2336 46 15570632 33 Chabaud M Lubberts E Joosten L van Den Berg W Miossec P IL-17 derived from juxta-articular bone and synovium contributes to joint degradation in rheumatoid arthritis Arthritis Res 2001 3 168 77 11299057 34 Bjarnason I Takeuchi K Intestinal permeability in the pathogenesis of NSAID-induced enteropathy J Gastroenterol 2009 44 Suppl 19 23 9 19148789 35 Edwards CJ Commensal gut bacteria and the etiopathogenesis of rheumatoid arthritis The Journal of rheumatology 2008 35 1477 14797 18671318 36 Branco P Francisco D Chambon C Hebraud M Arneborg N Almeida MG Identification of novel GAPDH-derived antimicrobial peptides secreted by Saccharomyces cerevisiae and involved in wine microbial interactions Appl Microbiol Biotechnol 2014 98 843 53 24292082 37 Seo JK Lee MJ Go HJ Kim YJ Park NG Antimicrobial function of the GAPDH-related antimicrobial peptide in the skin of skipjack tuna, Katsuwonus pelamis Fish Shellfish Immunol 2014 36 571 81 24412436 38 Hwang SY Kim HY Expression of IL-17 homologs and their receptors in the synovial cells of rheumatoid arthritis patients Mol Cells 2005 19 180 4 15879699 39 Petersen BC Budelsky AL Baptist AP Schaller MA Lukacs NW Interleukin-25 induces type 2 cytokine production in a steroid-resistant interleukin-17RB+ myeloid population that exacerbates asthmatic pathology Nat Med 2012 18 751 8 22543263
PMC005xxxxxx/PMC5125895.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 9214969 2488 Methods Mol Biol Methods Mol. Biol. Methods in molecular biology (Clifton, N.J.) 1064-3745 1940-6029 27812879 5125895 10.1007/978-1-4939-6670-7_17 NIHMS831172 Article Senescence-like Phenotypes in Human Nevi Joselow Andrew 12 Lynn Darren 1 Terzian Tamara PhD 1 Box Neil F. PhD 1 1 University of Colorado, Anschutz Medical Campus, Charles C. Gates Center for Regenerative Medicine and Department of Dermatology, Anschutz Medical Campus, Aurora, CO 2 Tulane University, School of Medicine Corresponding Author: Neil Box, [email protected] 22 11 2016 2017 01 1 2017 1534 175184 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Summary Cellular senescence is an irreversible arrest of cell proliferation at the G1 stage of the cell cycle in which cells become refractory to growth stimuli. Senescence is a critical and potent defense mechanism that mammalian cells have to suppress tumors. While there are many ways to induce a senescence response, oncogene-induced senescence (OIS) remains key to inhibiting progression of cells that have acquired oncogenic mutations. In primary cells in culture, OIS induces a set of measurable phenotypic and behavioral changes, in addition to cell cycle exit. Senescence-associated β-Galactosidase (SA-β-Gal) activity is a main hallmark of senescent cells, along with morphological changes that may depend on the oncogene that is activated, or on the primary cell type. Characteristic cellular changes of senescence include increased size, flattening, multi-nucleation, and extensive vacuolation. At the molecular level, tumor suppressor genes such as p53 and p16INK4a may play a role in initiation or maintenance of OIS. Activation of a DNA damage response and a senescence-associated secretory phenotype could delineate the onset of senescence. Despite advances in our understanding of how OIS suppresses some tumor types, the in vivo role of OIS in melanocytic nevi and melanoma remains poorly understood and not validated. In an effort to stimulate research in this field, we review in this chapter the known markers of senescence and provide experimental protocols for their identification by immunofluorescent staining in melanocytic nevi and malignant melanoma. nevus oncogene senescence DNA damage melanoma immunofluorescence cancer protocol Galactosidase p16INK4A p53 BRAF Introduction Cellular proliferation is highly regulated and key failsafe mechanisms prevent the unwanted expansion of cell number that can lead to cancer. Tumorigenesis may be constrained through two major mechanisms: apoptosis or cell cycle exit, including transient cell cycle arrest, quiescence and senescence (1). Apoptosis (programmed cell death) is a well-described pathway in humans involving energy-dependent biochemical processes that ultimately lead to the cell’s own “suicide”. This process results in membrane blebbing, pyknosis, and eventual removal through the immune activity. Quiescence is defined by reversible cell cycle exit and an extended period of metabolic inactivity (2). Senescence, on the other hand, is an irreversible arrest that restricts cellular proliferation in response to physiological stress signals, mitogen levels, or nutrients which would otherwise trigger cell duplication (3–5). Senescence was first identified in cells with shortened telomeres and telomeric dysfunction resulting in premature cell cycle arrest (6). Subsequently, it was shown that overexpression of an oncogene, RAS, could induce senescence in cultured human cells through a telomerase-independent mechanism (7). Numerous studies have since corroborated these findings, now commonly referred to as oncogene-induced senescence, or OIS. OIS is thought to occur when the aberrant activation of oncogenes (e.g. RAS, MYC, etc.) leads to depletion of cellular energy and replication reserves. The excess oncogenic load causes stalled replication forks and reactive oxygen species-induced damage, activating the cellular senescence program (3–5). Melanocytic nevi were initially considered to be an excellent example of senescence in vivo. These pigmented lesions are viewed as precancerous, and result from a focal and limited proliferation of melanocytes that is driven by the mutational activation of an oncogene such as BRAF. BRAFV600E is the most common oncogenic mutation found in human melanomas (8, 9). After forming a nevus, nevomelanocytes undergo a permanent exit from the cell cycle that prevents them from progressing to melanoma (10, 11). In support of this model, primary cultured melanocytes transduced with a BRAFV600E expressing lentivirus experience an initial burst of proliferation followed by morphologic changes and increased levels of SA-β-Gal (10). These changes are one of the most obvious hallmarks of senescent melanocytes in culture and include loss of the elongated fusiform profile characteristic of highly proliferative melanocytes, gain of a flattened egg-shape, and an appearance of multinucleation and vacuolation. On the other hand, nevomelanocytes present in common acquired melanocytic nevi have not been reported as morphologically distinct from adjacent normal skin melanocytes when examined histologically. Rather, it is the presence of nesting, or clustered growth of melanocytes, that distinguishes junctional nevi from solar lentigo. Analysis of the morphology of melanocytes in normal skin and in nevi using confocal microscopy on thick sections (50 µm), and some of the antibodies and markers described here, will allow characterization of any true senescent features of nevomelanocytes. Beyond morphology, numerous molecular markers of senescence have been utilized in various cell types that may be also useful for the study of senescence in nevus and melanoma formation. Senescent cells, typically those with a flattened egg-shape and a vacuole-rich cytoplasm, exhibit unusually high lysosomal β-Gal activity (2). Some have speculated that oncogene activation in the presence of intact tumor suppressors may cause cell hypertrophy and subsequent compensatory activation of lysosomal enzymes, including β-Gal (11–12). β-Gal gene silencing experiments demonstrated that its expression functions as a marker or an indicator of senescence rather than as a contributor to the process (13). Additionally, DNA damage (p53, γ-H2AX), proliferation (Ki67 and C-MYC), cell cycle regulation (p16INK4A, p14ARF, p15INK4B, FBX031 and p53) and chromatin structural changes (Senescence-Associated Heterochromatin Foci or SAHF) indicators were used in melanocytes and other cell types as putative senescence markers (2, 14, 15). A tissue specific hyper-secretory phenotype was also identified in senescent cells (16–18). A list of markers used in the literature is presented in Table 1. To validate the utility of some of them in distinguishing nevi (senescence) from melanoma (senescence bypass or escape), we performed immunofluorescent staining on paraffin sections from 10 nevus and 10 melanoma samples. We selected Ki67 (proliferation) and γ-H2AX (DNA repair foci) as representative biomarkers for senescence (Figure 1), and we observed successful staining using the protocol provided below. Positive staining for these proteins was detected in both nevus and melanoma samples, although it was clear that melanomas had considerably higher proliferation and DNA repair foci. We also observed successful staining for β-Galactosidase using the β-Gal antibody (ab9361; Abcam; data not shown), which may provide an alternative to the SA-β-Gal biochemical assay that requires the less commonly available frozen sample compared to paraffin-embedded tissues generated in most pathology practices. MART-1 is used in clinical diagnosis and identifies melanocytes and melanoma cells. Since known molecular indicators of senescence are not exclusive to this process, the use of multiple markers may be necessary for identification of senescence in vivo. There is a great need for further studies in this area, since conflicting reports have questioned the sensitivity and specificity of these biomarkers for recognizing senescent nevus cells. Several publications could not distinguish between nevi and melanoma using some of these reported markers (15, 19, 20). Only Ki67 appeared to consistently distinguish human nevi from a panel of primary and metastatic melanomas. Likewise, p16INK4A is reliably used in clinical diagnosis to distinguish between nevi and melanoma. It is possible that there are differences between in vivo with in vitro biomarkers of senescence. Alternatively, nevi and melanoma are not necessarily mutually exclusive, since approximately 25% of melanomas appear in pre-existing nevi (20–22). Heterogeneity of cell properties and mutation content is a well-established idea for tumors that has not been considered in reference to nevus senescence and evasion or bypass of senescence. The concept that nevi are irreversibly senescent has not been well documented, particularly since low mitotic activity has been found in congenital, common acquired, and dysplastic nevi (18, 22). It is anticipated that quantitative differences may be observed between nevi and melanoma, necessitating studies that consider staining patterns in greater detail. If senescence-like phenotypes can be reliably identified in nevi, it will lead to improvements in histological detection of melanoma and to novel avenues for anti-melanoma therapy. Materials Primary Antibodies: Ki67 (VP-K452; Vector), γ-H2AX (20E3; Cell Signaling), p16INK4a (N20 and C20; Santa Cruz), MART-1 (18-7263; Invitrogen), β-Gal (ab9361; Abcam) Secondary Antibodies: Secondary anti-IgG antibodies conjugated with Alexa Fluor 594 and Alexa Fluor 488 (Invitrogen) 100%, 95% and 70% Ethanol Xylene (247642-4L-CB; Sigma-Aldrich) Sodium Citrate Buffer 10X, pH 6.0 (C9999;Sigma) Pressure Cooker (CPC-600; Cuisinart) PBST: 0.1% surfactant Tween-20 (BP337-500; Fisher), and 1XPBS (D8537; Sigma-Aldrich) Blocking Solution: 10% γ-globulin free BSA/PBST/2 % normal immune serum: Add 100 µl of normal goat serum (S-1000; Vector) to 4.9 mL BSA/PBST Hydrophobic pen (Immedge; Vector) Methods 1. Dewaxing sections Place slides in a dipping rack and transfer successively to 3 plastic jars containing xylene while incubating in each for 5 minutes (See Notes 1, 2). 2. Rehydrate slides Bring to absolute alcohol by placing the slide rack in 100% EtOH baths twice for 5 minutes each. Start the rehydration by transfering successively the slide rack to two jars containing 95% EtOH 5 minutes each time then into two successive 70% EtOH consecutive baths twice for 3 minutes each. Place slide rack in water then under slowly running water for 2 minutes. 3. Pressure Cooker Antigen Retrieval Pour 10mM Sodium Citrate, pH 6.0 into heat-resistant coplin jar (See Notes 3, 4). Load slides into the citrate filled slide containers, close lid and put in the pressure cooker. Prepare the pressure cooker by adding water and place the plastic coplin jar with slide containers into the cooker (See Note 5). Start pressure cooker on HIGH setting: P=10psi, T=201-220°C for 15 min (See Note 6). Remove slide containers and cool down for 20-30 min at room temperature. While cooling, prepare blocking solution (described below). Transfer slides to a rack in a plastic jar containing water and wash under running water for 3 minutes. Transfer slide rack to two PBS wash tubs, for 5 minutes each. 4. Blocking Background Prepare a humid chamber to prevent evaporation of blocking solution and later on the primary antibody. A large slide box with a layer of moist tissue paper or a pool of water in the bottom could be used to create the moist chamber. Take slide out of 1XPBS, dry back of slide with a Kimwipe, gently dab front of slide around tissues, and aspirate PBS from around sample carefully (Do NOT touch the tissue!) Delineate each tissue sample with a hydrophobic pen to help retain the reagents on the section. Add 1XPBS onto slide immediately so they don’t dry out. Aspirate carefully the PBS and immediately replace with the blocking buffer to cover each section completely. This step will block sites of non-specific affinity for immunoglobulins (See Note 7). Close the humid chamber and incubate for one hour at room temperature. 5. Primary Specific Antibody Quick vortex the antibody stock tube. Dilute the specific antibody in the blocking solution. Dilution ratios change depending on which antibody is used and are normally provided by the manufacturer. If double staining is desired, add both primaries at the same time. Label slides at this point with (−) no primary antibody (control) and (+) primary antibody to be added (See Note 8). Aspirate the blocking solution from the (+) –labeled sections and leave the (−)–labeled sections intact (See Note 9). Add just enough diluted antibody to cover the (+)-labeled tissue sample. Incubate in closed humid chamber overnight at 4°C 6. Washes Tap off the primary antibody from each slide onto a Kimwipe and place in a 1XPBS tub. Wash in 1XPBS three times, 5 min each to remove the unbound antibody. In the meantime, prepare the secondary antibody as described below. Aspirate the liquid from both (+) and (−)-labeled sections, and immediately add the diluted secondary antibody (See Note 9). 7. Secondary Antibody Quick vortex the Alexa Fluor-conjugated antibody stock tubes (See Note 10). Dilute the secondary antibody to 1:1000 in blocking solution. If double staining is desired, make sure that each primary antibody is generated in a different species of animal so that the secondary antibodies cannot cross hybridize and that there is no spectral overlap. We chose Alexa Fluor 488 (green) and Alexa Fluor 594 (red) as fluorescent conjugates for our secondary antibodies. Incubate for one hour at room temperature in a humid chamber in the dark. 8. Washes/Coverslip Tap off the primary antibody, stick slides in racks and place these in a plastic jar containing 1XPBS wash. Wash slides 3 × 5 min in 1XPBS to remove the unbound antibody. Remove liquid from all sections and add one drop of permanent mounting medium containing the DNA dye DAPI (intercalates into the DNA) for nuclear visualization. Coverslip and allow an even spread of the mounting medium over the section by applying pressure gently to the slide cover with forceps. Make sure that there are no bubbles formed. If a hardening mounting medium is used, it is best to allow the sample to cure overnight, so next day microscopic examination is performed. Place in slide tray and keep in a slide folder at 4°C until ready to view but keep in mind that the fluorochromes’ fluorescence intensity diminishes over time. NFB was funded by NIH/NCI 5R01CA74592, NIH/NCI 1R01CA190533, NIH/NIAMS 1R03AR066880, and NIH/NCATS UL1 RR025780. TT was funded by NIH/NIAMS 1K01AR063203-01, NIH/NC 1R03CA191937, CCTSI Research grant from NIH/NCATS UL1 RR025780, ACS IRG 57-001-53 from the American Cancer Society and a Dermatology Foundation research grant. Figure 1 Hematoxylin and Eosine (H&E) and immunofluorescent staining of biomarkers in human nevus and melanoma sections, 20X. (A) H&E staining. In these examples, the nevus shows extensive nesting while the melanoma has a high density of melanocytes at the dermal-epidermal junction. (B) MART-1 and γ-H2AX immunostaining labelling melanocytes (cytoplasmic) and DNA damage (nuclear), respectively. (C) MART-1 and Ki-67 double immunostains label melanocytes and detect proliferation, respectively. Double positivity marks proliferative melanocytes. Table 1 Summary of literature findings for cellular senescence associated markers. Marker Association Cell/Tissue Studied Reference SA-β-Gal Senescence induction HUVEC* (24) p16 (INK4A) Cell Cycle Regulation Melanocytes (25, 26) p15(INK4B)` Cell Cycle Regulation Prostate Tissue` (20) ARF (p14)` Cell Cycle Regulation Melanocytes (25) Ki67 Proliferation Melanocytes (15) C-MYC Proliferation Cancer Cells (27) SAHF ** Chromatin Remodeling Melanoma Cells (28, 29) γ-H2AX DNA Damage Melanocytes (15, 19) p53 DNA Damage Melanocytes (19) FBXO31 Senescence induction Melanoma Cells (26) IGFBP7 Senescence induction Nevi, Melanoma Cells (26) PML-IV Bodies Senescence induction Melanoma Cells (28) miR-203 Senescence induction Melanoma Cells (30) Dec1 Senescence induction Cancer Cells (31) DcR2 Senescence induction Cancer Cells (32) RASG12V Senescence induction Retroviral Cells (33) * Human umbilical vein endothelial cells (HUVEC) ** Senescence Associated Heterochromatic Foci (SAHF): H3K9Me, HP1, H2A 1 Allow racks to drip when transferring between the plastic jars to reduce carry over to keep concentrations in the contents of the jars unchanged and minimalize 2 Keep xylene and EtOH tubs clean by replacing every four to five runs. 3 Allow an even distribution of slides in the citrate container for best epitope retrieval. 4 Tap each slide once or twice before placing in the Sodium Citrate solution. Excess water carry over from the slides can alter the pH of the buffer. 5 To save time, fill the pressure cooker with a few cups of water and set to “Keep Warm” while rehydrating the slides. 6 Ensure the pressure cooker is in fact holding pressure. If so, there should be no steam leaking from the cooker. 7 More solution may be required to fully cover large tissue sections; however, overloading sections is unnecessary and can lead to seepage. Avoid creating bubbles if possible. 8 By convention the negative control is closest to the label. 9 Don’t allow sections to dry. This is best achieved by working with one slide at a time rather than aspirating, or removing all the slides from PBS, at once. 10 After vortexing, collect the secondary antibody from the surface. Denser aggregates of antibody settle at the bottom of the tube and can contribute to background artifact in the staining. References 1 Lowe SW Cepero E Evan G Intrinsic tumour suppression Nature 2004 432 7015 307 315 15549092 2 Larsson LG Cellular senescence--a barrier against tumor development? Semin Cancer Biol 2011 21 6 347 348 22137673 3 Collado M Serrano M Senescence in tumours: evidence from mice and humans Nat Rev Cancer 2010 10 1 51 57 20029423 4 Campisi J d'Adda di Fagagna F Cellular senescence: when bad things happen to good cells Nat Rev Mol Cell Biol 2007 8 9 729 740 17667954 5 Kuilman T Michaloglou C Mooi WJ Peeper DS The essence of senescence Genes Dev 2010 24 22 2463 2479 21078816 6 Harrington L Robinson MO Telomere dysfunction: multiple paths to the same end Oncogene 2002 21 4 592 597 11850784 7 Serrano M Lin AW McCurrach ME Beach D Lowe SW Oncogenic ras provokes premature cell senescence associated with accumulation of p53 and p16INK4a Cell 1997 88 5 593 602 9054499 8 Davies H Bignell GR Cox C Stephens P Mutations of the BRAF gene in human cancer Nature 2002 417 6892 949 954 12068308 9 Pollock PM Harper UL Hansen KS Yudt LM Stark M Robbins CM High frequency of BRAF mutations in nevi Nat Genet 2003 33 1 19 20 12447372 10 Michaloglou C Vredeveld LC Soengas MS Denoyelle C Kuilman T van der Horst CM BRAFE600-associated senescence-like cell cycle arrest of human naevi Nature 2005 436 7051 720 724 16079850 11 Collado M Serrano M Senescence in tumours: evidence from mice and humans Nat Rev Cancer 2010 10 1 51 57 20029423 12 Blagosklonny MV Cell senescence and hypermitogenic arrest EMBO Reports 2003 4 4 358 362 12671679 13 Blagosklonny MV Cell cycle arrest is not senescence Aging 2011 3 2 94 101 21297220 14 Lee BY Han JA Im JS Morrone A Johung K Goodwin EC Senescence-associated beta-galactosidase is lysosomal beta-galactosidase Aging Cell 2006 5 2 187 195 16626397 15 Tran S Rizos H Human nevi lack distinguishing senescence traits Aging 2013 2 98 99 23518674 16 Parrinello S Coppe JP Krtolica A Campisi J Stromal-epithelial interactions in aging and cancer: senescent fibroblasts alter epithelial cell differentiation J Cell Sci 2005 118 3 485 496 15657080 17 Freund A Orjalo AV Desprez PY Campisi J Inflammatory networks during cellular senescence: causes and consequences Trends Mol Med 2010 16 5 238 246 20444648 18 Patil CK Mian IS Campisi J The thorny path linking cellular senescence to organismal aging Mech Ageing Dev 2005 126 10 1040 1045 16153470 19 Tran SL Haferkamp S Scurr LL Gowrishankar K Becker TM Desilva C Absence of distinguishing senescence traits in human melanocytic nevi J Invest Dermatol 2012 132 9 2226 2234 22513787 20 Zhang G Herlyn M Human nevi: no longer precursors of melanomas? The Journal of investigative dermatology 2012 132 9 2133 2134 22895446 21 Bevona C Goggins W Quinn T Fullerton J Tsao H Cutaneous melanomas associated with nevi Arch Dermatol 2003 139 12 1620 1624 14676081 22 Tsao H Bevona C Goggins W Quinn T The transformation rate of moles (melanocytic nevi) into cutaneous melanoma: a population-based estimate Arch Dermatol 2003 139 3 282 288 12622618 23 Marks R Dorevitch AP Mason G Do all melanomas come from "moles"? A study of the histological association between melanocytic naevi and melanoma Australas J Dermatol 1990 31 2 77 80 2095738 24 Kurz DJ Decary S Hong Y Erusalimsky JD Senescence-associated (beta)-galactosidase reflects an increase in lysosomal mass during replicative ageing of human endothelial cells J Cell Sci 2000 113 Pt 20 3613 3622 11017877 25 Bennett DC Genetics of melanoma progression: the rise and fall of cell senescence Pigment Cell Melanoma Res 2015 29 2 122 140 26386262 26 Ross AL Sanchez MI Grichnik JM Nevus senescence ISRN Dermatology 2011 2011 642157 22363855 27 Wang H Mannava S Grachtchouk V Zhuang D Soengas MS Gudkov AV c-Myc depletion inhibits proliferation of human tumor cells at various stages of the cell cycle Oncogene 2008 27 13 1905 1915 17906696 28 Goel VK Ibrahim N Jiang G Singhal M Fee S Flotte T Melanocytic nevus-like hyperplasia and melanoma in transgenic BRAFV600E mice Oncogene 2009 28 23 2289 2298 19398955 29 Webster MR Xu M Kinzler KA Kaur A Appleton J O'Connell MP Wnt5A promotes an adaptive, senescent-like stress response, while continuing to drive invasion in melanoma cells Pigment Cell Melanoma Res 2015 28 2 184 195 25407936 30 Noguchi S Mori T Otsuka Y Yamada N Yasui Y Iwasaki J Anti-oncogenic microRNA-203 induces senescence by targeting E2F3 protein in human melanoma cells J Biol Chem 2012 287 15 11769 11777 22354972 31 Haferkamp S Becker T Scurr L p16INK4a-induced senescence is disabled by melanoma-associated mutations Aging Cell 2008 7 5 733 745 18843795 32 Zhang Z Rosen DG Yao JL Huang J Liu J Expression of p14ARF, p15INK4b, p16INK4a, and DCR2 increases during prostate cancer progression Mod Pathol 2006 19 10 1339 1343 16799475 33 Aird K Zhang R Detection of senescence-associated heterochromatin foci (SAHF) Methods Mol Biol 2013 965 185 196 23296659
PMC005xxxxxx/PMC5125913.txt	LICENSE: This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. 0117605 668 Annu Rev Genet Annu. Rev. Genet. Annual review of genetics 0066-4197 1545-2948 27617971 5125913 10.1146/annurev-genet-120215-035146 NIHMS829246 Article Transition Metals and Virulence in Bacteria Palmer Lauren D. 1 Skaar Eric P. 12 1 Department of Pathology, Microbiology, and Immunology, Vanderbilt University Medical Center, Nashville, Tennessee 37212 2 Tennessee Valley Healthcare System, US Department of Veterans Affairs, Nashville, Tennessee 37212; [email protected] 15 11 2016 7 9 2016 23 11 2016 28 11 2016 50 6791 This file is available for text mining. It may also be used consistent with the principles of fair use under the copyright law. Transition metals are required trace elements for all forms of life. Due to their unique inorganic and redox properties, transition metals serve as cofactors for enzymes and other proteins. In bacterial pathogenesis, the vertebrate host represents a rich source of nutrient metals, and bacteria have evolved diverse metal acquisition strategies. Host metal homeostasis changes dramatically in response to bacterial infections, including production of metal sequestering proteins and the bombardment of bacteria with toxic levels of metals. Presumably, in response, bacteria have evolved systems to subvert metal sequestration and toxicity. The coevolution of hosts and their bacterial pathogens in the battle for metals has uncovered emerging paradigms in social microbiology, rapid evolution, host specificity, and metal homeostasis across domains. This review focuses on recent advances and open questions in our understanding of the complex role of transition metals at the host-pathogen interface. host-pathogen interface metal acquisition metal homeostasis nutritional immunity transition metal INTRODUCTION Metals are essential for all forms of life. Metal cofactors serve structural and catalytic roles in biological processes, including precursor biosynthesis, DNA replication, transcription, respiration, and responses to oxidative stress. Most organisms require the first-row transition metals manganese, iron, cobalt, nickel, copper, and zinc. Excepting zinc, these metals have unfilled d-orbitals and thus are redox active. Their ability to easily cycle between oxidation states contributes to both their catalytic properties and their toxicity. Their divalent cations bind ligands similarly, but each metal has different properties such that they are only occasionally functional substitutes. Thus, protein mismetallation often ablates function, necessitating exquisitely sensitive metal homeostasis systems. In bacterial pathogens, homeostasis strategies include complex metal-responsive transcriptional regulatory networks, as reviewed elsewhere (20, 62, 78, 130). Understanding how metal homeostasis proteins recognize and respond to the correct metals is an active area of research in bioinorganic chemistry with implications for bacterial and host metal metabolism during infection. In the context of infection, bacterial and host metal metabolism determine disease pathogenesis with three emerging principles. First, although the host represents a rich nutrient metal source for bacteria, the host immune system can block bacterial metal acquisition in a process termed nutritional immunity. Second, hosts barrage bacteria with toxic concentrations of metals. Finally, dysregulation of host metal homeostasis—through genetic mutation or nutrition—alters susceptibility to infection. This review summarizes findings, highlights recent work, and discusses questions to be pursued in the study of metals at the interface of bacterial pathogens and their hosts. BACTERIAL METAL ACQUISITION AND HOST NUTRITIONAL IMMUNITY In order to obtain their required nutrient metals, bacteria have evolved three main classes of metal acquisition systems: elemental metal import, extracellular metal capture by siderophores, and metal acquisition from host proteins (including metal piracy from host nutritional immunity proteins). In response, mammalian hosts have evolved strategies to restrict bacterial metal acquisition. The protein transferrin was first reported to inhibit microbial growth by iron chelation in egg whites and human plasma in the 1940s, and lactoferrin was later shown to have similar properties at mucosal surfaces (105, 120, 121) (Figure 1) (see sidebar, Nutritional Immunity Protein Moonlights as an Antimicrobial). Thirty years later, Eugene Weinberg popularized the term nutritional immunity to describe the phenomenon, which has since been expanded to include host restriction of zinc, manganese, and nonmetal nutrients such as amino acids (65, 140, 149). Growing evidence suggests that this battle for metals represents an interface for coevolution of hosts and bacteria. As posited by the Red Queen hypothesis in evolutionary biology, with each development in nutritional immunity defense, pathogens evolve a new offense (11, 133). The following sections discuss our understanding of nutrient transition metals at the host-pathogen interface, focusing on iron, zinc, and manganese. Iron and the Host-Pathogen Battle: The Vanguard in Understanding Iron is the most widely used transition metal in biological systems, presumably because it is the most abundant element on Earth, making up 80% of the Earth's core. Bacterial pathogens require iron for essential processes, including the pentose phosphate pathway, DNA replication, transcription, and cofactor biosynthesis. For example, biosynthesis of the cofactor heme (Fe2+-protoporphyrin IX) is important in Brucella abortus systemic infection and Staphylococcus aureus systemic and bone infection (5, 57, 58). Additionally, the iron-sulfur cluster chaperone Nfu was found to be important in mouse and moth models of infection by S. aureus and Acinetobacter baumannii, respectively (86, 150). Iron is incorporated into cofactors in its soluble ferrous form (Fe2+). Many bacteria import Fe2+ with the ferrous transporter Feo, which is important for colonization and/or virulence in a number of bacteria, including the gastrointestinal pathogens Helicobacter pylori, Salmonella enterica, and Campylobacter jejuni (77) (Figure 2a). However, Fe2+ is largely unavailable for bacterial pathogens due to host restriction, and the physiochemical properties of iron make Fe3+ the predominant form in the host environment, as describe below. Two-thirds of iron in humans is sequestered in erythrocytes as heme bound to the oxygen-carrying protein hemoglobin (24). Under oxic conditions and at physiological pH in serum, free iron is insoluble as Fe3+. In vertebrates, nearly all Fe3+ is bound by transferrin in serum, by ferritin in cells, or by lactoferrin in milk, saliva, tears, mucus, and neutrophil secondary granules. These factors combined result in a low free-iron concentration in human serum, which is further restricted in response to bacterial infection (24). Bacteria have evolved mechanisms to circumvent host iron restriction. For the simplest solution, the Lyme disease pathogen Borrelia burgdorferi evolved a complete lack of iron requirement by exploiting the interchangeability of transition metals and replacing iron with manganese in many metalloenzymes (110). Pathogens that have not weaned themselves off iron often rely on cheating, thievery, and piracy to acquire iron in spite of nutritional immunity. Siderophores, Evasion of Host Interference, and Cheating Many bacterial pathogens use siderophores to overcome iron limitation in the host environment. Siderophores are low molecular weight, high-affinity Fe3+-binding compounds secreted and imported by bacteria for iron acquisition. Siderophores have exceptionally high-iron binding affinity, which allows some siderophores to steal iron from host proteins. For example, enterobactin produced by Enterobacteriaceae binds Fe3+ (Kd 10−51–10−49 M) more tightly than transferrin (Kd 10−24–10−20 M), thus outcompeting transferrin chelation (23, 81). In Gram-negative bacteria, siderophores are transported across the outer membrane by a TonB-dependent system, then trafficked through the periplasm and imported into the cytoplasm by ATP-binding cassette (ABC) transporters (93) (Figure 2a). Gram-positive bacteria import Fe3+-siderophores with lipoprotein substrate-binding proteins and ABC transporters (93) (Figure 2b). Iron is released from the siderophores by one of two main mechanisms. Siderophores with high Fe3+ binding affinity, such as enterobactin or bacillibactin produced by Bacillus anthracis, are hydrolyzed by specific esterases before Fe3+ is reduced to Fe2+. The fate of the hydrolyzed siderophores is not well understood; they may be further degraded or recycled (93). For other siderophores, ferric reductases reduce Fe3+ to Fe2+ that is then released due to its lower binding affinity or by competitive sequestration (93). Deferrated siderophores can then be recycled. Pseudomonas aeruginosa uses a dedicated periplasmic export system to recycle siderophores (66). Recent studies in Mycobacterium tuberculosis determined that one system exports both de novo synthesized and recycled siderophores from the cytoplasm and is required to prevent cytoplasmic siderophore accumulation and poisoning (69) (Figure 2c). Siderophore iron acquisition and host interference illustrate an evolutionary arms race in bacterial pathogenesis. The vertebrate host produces the siderophore-binding protein lipocalin 2, also known as siderocalin or neutrophil gelatinase–associated lipocalin, during the innate immune response to bacterial infection to sequester siderophores such as enterobactin or bacillibactin (1, 44). In response, bacteria evolved to elaborate stealth siderophores that evade lipocalin 2 binding, including the highly glycosylated salmochelin produced by pathogenic enterobacteria and petrobactin produced by B. anthracis (1, 43). Remarkably, expression of the salmochelin biosynthetic locus confers virulence to a nonpathogenic Escherichia coli, implying stealth siderophore production is a key virulence determinant (43). Many pathogens encode multiple siderophore biosynthetic systems, suggesting that diverse siderophores allow context-specific replicative advantages. Studies on siderophore production in enterobacterial pathogens have provided insight into these fitness benefits. During S. enterica gastrointestinal infection, S. enterica uses salmochelin to gain a competitive advantage over commensals but is outcompeted for siderophores by the probiotic E. coli strain Nissle, which reduces S. enterica colonization (33, 113). Although enterobactin can be sequestered by lipocalin 2, it has a higher binding affinity for Fe3+ than stealth siderophores salmochelin and yersiniabactin. This high Fe3+ binding efficiency renders enterobactin essential in certain host niches. In the lung, enterobactin is not required for Klebsiella pneumoniae pneumonia but is required for invasion into surrounding tissue; its higher Fe3+ binding affinity enables K. pneumoniae to acquire iron from transferrin in the perivascular space (9). Counterintuitively, uropathogenic E. coli use enterobactin to resist lipocalin 2 in human urine, apparently by extracting Fe3+ from Fe3+-catechol-lipocalin 2 complexes (122). Human urine varies in characteristics such as pH and aryl metabolite (catechol) content. In donors with high urine pH, lipocalin 2 complexes with catechols and Fe3+ to sequester iron from invading pathogens. Uropathogenic E. coli can overcome this sequestration owing to the exceptionally high Fe3+ binding affinity of enterobactin (9). Given the large number of siderophores identified thus far, understanding mechanisms driving siderophore diversity warrants further investigation. In addition to selective pressures associated with maintaining multiple siderophore biosynthetic systems, siderophore production in P. aeruginosa serves as a model to understand how cheating drives evolution. Because siderophores are secreted, bacteria can import siderophores for iron acquisition without bearing the metabolic cost of production. In this way, siderophores are public goods. Experimental evolution in P. aeruginosa showed that relatedness and low competition selects for cooperation (i.e., siderophore production), whereas local competition selects for cheaters (53). Additional research found that the nutritional environment and regulatory cross talk between social behaviors (i.e., quorum sensing and production of multiple siderophores) further modulate the selection for cooperation versus cheating in P. aeruginosa (49, 117). Importantly, long-term P. aeruginosa infection in the cystic fibrosis lung was recently shown to select for siderophore cheating, eventually driving some populations to lose siderophore production and thus reduce fitness in the iron-limited host (7). The competitive dynamics leading to complete loss of siderophore production demonstrate that bacterial communities are susceptible to the tragedy of the commons, an economics term first applied to evolutionary biology by Garrett Hardin in 1968 to describe each individual seeking maximum benefit from common goods at the expense of population-wide fitness (60). The P. aeruginosa findings suggest that the intensity of competition in the host environment can drive bacterial populations to extinction. Iron Piracy Subverts Nutritional Immunity In addition to acquiring iron from heme, some pathogens have developed specialized systems to obtain iron from nutritional immunity proteins through a process termed iron piracy. The transferrin-binding proteins TbpA/TbpB that are present in species, including Neisseria spp., Haemophilus influenzae, and Moraxella catarrhalis, are TonB-dependent receptors that bind transferrin and extract Fe3+ for transport into the cell (96). Additionally, some pathogens express lactoferrin receptors, including pathogenic Neisseria spp., Treponema pallidum, H. pylori, and Streptococcus pneumoniae (34, 59, 96). With these systems, enterprising bacteria convert host proteins that evolved to restrict bacterial growth into nutrient iron sources. Selective pressure at the receptor-ligand binding sites between host transferrin and bacterial transferrin-binding proteins has driven evolution at the host-pathogen interface. Like the hemoglobin-IsdB interaction, transferrin-binding proteins restrict host range for numerous pathogens. The obligate human pathogens of Neisseria spp. are host restricted based on their transferrin-binding proteins, a fact that has been exploited to develop murine models for meningitis by introducing or expressing human transferrin (102, 148). Structural analysis of TbpB from the porcine pathogen Actinobacillus pleuropneumoniae revealed a single conserved residue required for transferrin binding, whereas surrounding residues varied extensively, presumably due to host selective pressure (95). Elegant evolutionary work showed that transferrin genetic variation among apes is the result of rapid evolution to counteract bacterial iron piracy (10). This work emphasizes the utility of metal acquisition by bacterial pathogens as a model to uncover rapid evolution at protein interfaces between host and pathogen. Iron Thievery: Bacterial Heme Acquisition Because the majority of iron in the human body is sequestered in erythrocytes, many bacterial pathogens evolved mechanisms to access heme iron. These pathogens first lyse erythrocytes, bind hemoglobin or other host heme proteins, and then extract heme for direct use or degradation for liberation of free iron (27). Following erythrocyte lysis, mammalian hosts employ hemopexin and haptoglobin to bind and sequester free heme and hemoglobin, respectively, to protect against their oxidative activities. Bacterial heme acquisition systems have surface-associated receptors that bind heme or hemoproteins and pass heme through the periplasm and/or cell wall to a membrane heme transporter. Heme acquisition systems are well described for many Gram-negative pathogens and rely on an outer membrane receptor such as HmuR that binds heme or hemoproteins, transporting heme through the periplasm by a TonB-ExbB-ExbD system (19) (Figure 2a). Heme is then transported into the cytoplasm by an ABC transporter. Some bacteria additionally release hemophores, such as heme assimilation system HasA in Serratia marescens and P. aeruginosa, which extracts heme from hemoglobin with 10−11 M binding affinity and relays it to the outer membrane hemophore receptor HasR (76, 144). M. tuberculosis uses a secreted hemophore with a unique fold, demonstrating the diversity of these systems (131). Many Gram-positive pathogens use the iron-regulated surface determinant Isd heme uptake system with dedicated cell wall sortase and heme passage proteins to acquire iron from free heme or heme complexed to hemoglobin or hemoglobin-haptoglobin (87) (Figure 2b). B. anthracis additionally encodes secreted hemophores IsdX1 and IsdX2 that bind heme extracellularly for import but are not required for virulence, implying that there are functional redundancies in B. anthracis heme acquisition (84). The recent discovery of an iron-regulated autolysin in the opportunistic pathogen Staphylococcus lugdunensis suggests that Gram-positive bacteria remodel their cell walls to accommodate heme acquisition proteins (42). Heme acquisition is required for virulence in Staphylococcus aureus, B. anthracis, and Streptococcus pyogenes (27). In Corynebacterium diphtheriae, one of at least two redundant heme transport–associated Hta uptake systems is tethered to the cytoplasmic membrane rather than relying on cell wall sortases. The roles that the two Hta systems play in virulence remain unknown (4). The importance of heme as an iron source during infection has driven evolution at the host-pathogen interface. S. aureus evolved to use hemoglobin as its preferred iron source in infection, and its IsdB hemoglobin-binding protein evolved to preferentially bind human hemoglobin over hemoglobin from other animals (109, 123). Additionally, P. aeruginosa lineages colonized in cystic fibrosis patients evolved to prefer heme over the siderophore pyoverdine as an iron source (85). These examples suggest that the interaction between bacterial heme acquisition proteins and host hemoglobin is also susceptible to selective pressure in the host-pathogen arms race. Therefore, bacterial heme acquisition may offer opportunities for further understanding host-pathogen evolution. Zinc Turf Wars: Sequestration and Potential for Piracy Following iron, zinc is the second most abundant transition metal cofactor. Zinc binding by proteins can be structural, regulatory, or catalytic, and zinc is required for many metalloproteases. Bacterial zinc uptake (Znu) systems are ABC transporters required for virulence in a variety of pathogens, including S. pneumoniae, Acinetobacter baumannii, Listeria monocytogenes, and P. aeruginosa (29, 36, 64) (Figure 3a). Recent work uncovered two novel zinc acquisition systems in Yersinia spp. In Yersinia pseudotuberculosis, a type VI secretion system (T6SS) imports zinc by transporting a zinc-binding protein effector bound to zinc, and maximal virulence requires both the Znu and T6SS systems, thus expanding known roles for T6SS (139). In Y. pestis, Bobrov et al. (16) demonstrated that the siderophore yersiniabactin binds zinc and is transported into the cell using a dedicated Zn2+-yersiniabactin importer, YbtX. Importantly, in a model of septicemic plague, Y. pestis displays a virulence defect only in strains lacking both the znu transporter and yersiniabactin biosynthesis genes. This pivotal finding established the importance of siderophore noniron metal acquisition in bacterial pathogenesis. Bacterial zinc acquisition systems are required during infection to counteract limitation by nutritional immunity. During the acute phase response to inflammation, serum zinc drops due to enhanced zinc uptake and sequestration by metallothionein in liver cells (55). Additionally, members of the helix-loop-helix (EF-hand type) Ca2+-binding S100 protein family of the vertebrate innate immune system exert nutrient limitation through extracellular zinc chelation (Figure 3b). The importance of metal chelation by S100 proteins for innate immunity was first established with the demonstration that S100A7 (psoriasin) protects human skin from infection by zinc chelation (51). Psoriasin also contributes to mucosal immunity on the tongue and in the vagina (92, 94). Similarly, S100A12 (calgranulin C), which binds both Zn2+ and Cu2+, is induced in H. pylori infection and limits bacterial growth through zinc chelation (56). Molecular imaging by mass spectrometry demonstrated that Staphylococcus aureus liver abscesses are severely depleted of zinc and manganese and depletion depends on heterodimers of S100A8/S100A9 (calprotectin) (30). Calprotectin is an abundant neutrophil protein that can chelate Zn2+ and Mn2+ and efficiently competes away these nutrient metals from numerous bacterial pathogens in vitro (32). In addition to S. aureus, calprotectin zinc limitation is important for controlling infection by bacterial pathogens, including L. monocytogenes, A. baumannii, K. pneumoniae, and H. pylori (3, 47, 64, 147). Recent work showed that Fe2+ binding by calprotectin contributes to inhibition of bacterial growth in vitro, an activity that remains to be investigated during infection (98). Counterintuitively, calprotectin enhances pathogenesis of some bacteria by diverse mechanisms. For example, in pneumococcal pneumonia, calprotectin appears to protect S. pneumoniae from host-induced zinc toxicity in the lung (2). Calprotectin enhances S. enterica growth in the inflamed gut, presumably because S. enterica can outcompete the gut microbiota for limited zinc (79). Subinhibitory concentrations of calprotectin also increase H. pylori biofilm formation (46). In addition, calprotectin promotes pneumonia co-infection of P. aeruginosa and S. aureus by abrogating production of anti-staphylococcal factors by P. aeruginosa (137). This finding offers a potential explanation for co-infection by P. aeruginosa and S. aureus in cystic fibrosis patients, where calprotectin is abundant. The differential effects of calprotectin zinc sequestration likely reflect differences in host environments, bacterial zinc requirements, and acquisition systems. Recent work identified the potential for bacterial zinc piracy from S100 proteins during infection. The Neisseria meningitidis outer membrane receptor CbpA was shown to bind calprotectin and extract zinc in a TonB-dependent manner (127). This finding suggests that bacteria can exploit host zinc nutritional immunity proteins analogously to iron acquisition from transferrin and lactoferrin. In this case, interactions between CpbA and calprotectin may also be susceptible to rapid evolution, establishing a new front in the study of the evolutionary battle for nutrient metals. Manganese Restriction: A War in Two Theaters Manganese is an essential trace element for all forms of life and serves as a cofactor for enzymes involved in oxidative stress, DNA replication, and central metabolism, including manganese superoxide dismutase and pyruvate kinase. Bacterial manganese import systems generally belong to the MntH/Nramp or ABC transporter families and are important during infection in a number of pathogenic bacteria, including S. enterica, S. aureus, Neisseria gonorrhoeae, B. abortus, Streptococcus spp., and Yersinia spp. (71) (Figure 4a,b). In the Lyme disease pathogen B. burgdorferi, the BmtA manganese transporter is required for virulence. Curiously, BmtA is not related to bacterial manganese importers but, rather, to the ZIP (Zrt/Irt-like protein) family transporters in eukaryotes (106). The divergence of B. burgdorferi manganese importers from other pathogens may reflect its unique evolutionary strategy to substitute its iron requirement with manganese. Interestingly, the requirement for manganese acquisition systems can depend on the mode of infection or the infected organ. For S. aureus, mutants lacking MntH and ABC transport systems have a lower burden in the liver during systemic infection but maintain their burden in the kidney (73). In Y. pestis, mutants lacking both transport systems display no defect in pneumonic plague but are attenuated in bubonic plague (108). These findings contribute to the understanding that the requirement for metal import depends on the specific host environment. The host innate immune system has evolved both extra- and intracellular manganese restriction mechanisms (71) (Figure 4c). The Mn2+-binding activity of calprotectin limits available manganese for many bacterial pathogens (32). In S. aureus, calprotectin-mediated manganese restriction inhibits the bacterium's oxidative stress defense program by limiting manganese superoxide dismutase activity (72). In turn, the S. aureus high-affinity manganese import systems help combat calprotectin limitation during infection (73). In contrast to what occurs in the liver, S. aureus kidney abscesses remain manganese depleted in calprotectin-deficient animals, implicating additional host measures of manganese sequestration (73). During S. enterica infection of the gut, manganese acquisition is required for infection, where manganese restriction is IL-22 dependent but calprotectin-independent (35). Therefore, open areas for investigation include the role of calprotectin-dependent manganese sequestration in other infections and the remaining calprotectin-independent mechanisms of host manganese chelation. Intracellular pathogens also face host manganese limitation in the phagosome. In response to infection, neutrophils and macrophages express the natural resistance–associated macrophage protein 1 (Nramp1), which depletes the phagosome of manganese and, to a lesser extent, iron (67, 142). Bacterial manganese acquisition systems are important for resisting Nramp1-mediated manganese restriction and may contribute differentially, such as in S. enterica infection of stimulated macrophages. The fact that polymorphisms in nramp1 affect host susceptibility to infection emphasizes the importance of manganese limitation. The commonly used laboratory mouse strains C57BL/6 and BALB/c are Nramp1-deficient and rapidly succumb to S. enterica infection, in contrast to Nramp1-sufficient mouse strains (135). Human nramp1 polymorphisms affect susceptibility to tuberculosis, meningococcal diseases, and leprosy (14). Additionally, nramp1 expression in macrophages induces the host siderophore-sequestering protein lipocalin 2, uncovering a regulatory link between multiple mechanisms of nutritional immunity (45). Therefore, continued study of nutrient-limiting proteins may expose additional multimetal regulatory networks involved in nutritional immunity. Cobalt, Nickel, and Copper: Neglected Transition Metals in Nutritional Immunity Less is known about the roles of cobalt, nickel, and copper acquisition in bacterial pathogenesis. Nickel is well established as an essential cofactor for virulence factors, including nickel-iron hydrogenase and urease in H. pylori, where the NixA nickel transporter is required for virulence (41, 100). Urease is also expressed by Proteus mirabilis and Staphylococcus saprophyticus, other causative agents of urinary tract infections (21). Cobalt is an enzymatic cofactor for virulence determinants, including S. pneumoniae GlcNAc deacetylase and Helicobacter pylori arginase (15, 89). Dedicated cobalt transporters have not been identified, but many bacterial pathogens encode nickel-cobalt transporters (115). In S. aureus, the nickel transporter Nik is required for urease activity and efficient kidney colonization, and the cobalt and nickel transporter Cnt is additionally required for infection in systemic and urinary tract infections in mice (63, 114). Recent work demonstrated that Cnt transports metals in complex with staphylopine, a nicotianamine-like metallophore capable of binding nickel, cobalt, zinc, copper, and iron (50). Characterization of staphylopine marks the first description of a nicotianamine-like molecule in bacteria and the first broad-spectrum metallophore. The role of copper in bacterial pathogenesis is an exciting area of current investigation. Copper is important during aerobic growth as a cofactor of cytochrome c oxidase and copper-zinc superoxide dismutase. It is unclear whether bacterial pathogens actively transport copper and whether copper is restricted by host nutritional immunity. In mycobacteria, porins in the outer membrane are known to be important for copper transport (124). Mounting evidence demonstrates that the siderophore yersiniabactin binds Cu2+ in physiologically relevant environments and that Cu2+-yersiniabactin can be imported in uropathogenic E. coli using the TonB-dependent siderophore importer FyuA (26, 75). Copper acquisition by the fungal pathogen Cryptococcus neoformans is critical for meningoencephalitis virulence but not pneumonia, demonstrating that some host niches are copper limited (129). These findings raise questions as to the mechanism of copper limitation in these niches (e.g., cerebrospinal fluid). Although the host immune protein S100A12 binds copper in addition to zinc, its antimicrobial activity against H. pylori appears to be restricted to zinc chelation (56). Improved understanding of the importance of cobalt, nickel, and copper for bacterial pathogenesis will determine whether the host actively limits these metals through nutritional immunity and whether bacterial pathogens are affected by their limitation or can counteract host-mediated limitation. HOST-IMPOSED METAL TOXICITY Although the first-row transition metals are required trace elements for biological systems, they are also inherently toxic. One potential mechanism of toxicity is Fenton chemistry by the redox-active transition metals iron, copper, and manganese. In the Haber-Weiss reaction of Fenton chemistry, Fe2+ reacts with hydrogen peroxide to form Fe3+, a hydroxyl radical, and hydroxide; Fe3+ can then react with hydrogen peroxide to regenerate Fe2+ in addition to a hydroperoxyl radical and a proton. These oxidative species inhibit cell growth by damaging proteins, DNA, and lipids. Another mechanism of toxicity is protein mismetallation. The prevalence of metalloproteins in the bacterial cell and their relatively compatible binding ligands necessitate tight control of metal homeostasis for continued metabolism. Complementary to nutrient metal limitation, hosts have evolved mechanisms to deploy toxic levels of metal ions for bacterial control, and bacteria use multiple metal exporters to resist host intoxication (20, 39, 54). Copper Ions as Ammunition Copper has long been recognized as antimicrobial and is often embedded in textiles and medical devices as a microbicide. Although copper can carry out Fenton chemistry, its toxicity is caused by mismetallation in bacterial species investigated thus far. The primary targets of its toxicity in E. coli and N. gonorrhoeae are iron-sulfur cluster proteins in branched chain amino acid and heme biosynthesis, respectively; in S. pneumoniae, the primary target is the manganese-containing aerobic ribonucleotide reductase (38, 68, 83) (Figure 5a). In a possible twist on Fenton chemistry, copper was also found to potentiate nitrosative stress in N. gonorrhoeae (37). Host bactericidal mechanisms take advantage of copper toxicity to create poisonous microenvironments for intracellular bacteria. During bacterial infection, interferon-γ induces expression of copper transport protein 1 (CTR1) to import copper from the environment to ATOX1, which then shuttles copper to the ATP7A transporter at the membrane of the phagolysosome for copper accumulation (74, 143, 145) (Figure 5b). Accordingly, copper resistance is required for virulence of many intracellular pathogens. N. gonorrhoeae, S. enterica, and mycobacteria employ copper exporters and extracellular copper oxidation (39, 119) (Figure 5a). Investigation of copper resistance has uncovered surprising mechanisms. For example, in response to copper toxicity, mycobacteria produce the bacterial metallothionein MymT to sequester copper in the cytoplasm (52). Uropathogenic E. coli use Cu2+-yersiniabactin to protect against copper toxicity and oxidative stress through at least two mechanisms: (a) binding to yersiniabactin prevents catechol-mediated Cu2+ reduction to Cu+, which could participate in Fenton chemistry, and (b) Cu2+-yersiniabactin exhibits extracellular and catalytic superoxide dismutase–like activity (25, 26). Our understanding of metal resistance beyond export mechanisms is in its infancy, but findings thus far suggest it may be fertile ground for discovery. Zinc Toxicity Expands the Host Armory Zinc cannot participate in Fenton chemistry and its mechanism of toxicity likely involves mismetallation. Zinc intoxication competitively inhibits manganese binding to an importer in S. pneumoniae and the glycolytic enzymes phosphofructokinase and glyceraldehyde-3-phosphate dehydrogenase in S. pyogenes (88, 104) (Figure 6). Like copper toxicity, zinc intoxication as a host defense mechanism was first described during mycobacterial infection. Transcriptional profiling of M. tuberculosis in human macrophages identified signs of metal intoxication in bacteria and in the host, revealing a burst of free zinc in macrophages and intraphagosomal zinc accumulation (18). Zinc efflux by the cadmium-zinc-cobalt (Czc) family exporter was also found to be important for S. pyogenes in a mouse model of infection (103). Human neutrophils respond to S. pyogenes internalization by mobilizing zinc, apparently increasing free cytosolic zinc (103). Interestingly, an increase in cytosolic zinc due to the transporter ZIP8 is an innate immune signaling mechanism in macrophages and lung epithelial cells. Specifically, zinc accumulation modulates nuclear factor (NF)-κB activation and inflammation (80). It is therefore unclear whether S. pyogenes zinc poisoning in the cytosol of neutrophils is an evolved mechanism of zinc intoxication or a side effect of the host using zinc as an immune signal. S. pneumoniae also likely experiences zinc intoxication during infection, based on the protective effect of zinc-chelating calprotectin and the increase in zinc/manganese ratios in the lung, brain, blood, and nasopharynx that occur during infection (2, 88). Further research may determine whether the rise in cytosolic zinc in neutrophils and macrophages is the main mechanism for zinc intoxication. Potential Intoxicants: Manganese, Iron, Nickel, and Cobalt There are currently no described mechanisms of host intoxication using the remaining first-row transition metals, but all are toxic at high levels and therefore could be harnessed by the host for intoxication. In a Mongolian gerbil model of stomach colonization, H. pylori requires resistance to the cadmium, zinc, and nickel (Czn) exporter, suggesting that H. pylori encounters metal stress in the host (125). In addition to elemental iron, Fe2+ in heme is toxic to bacterial cells (27). One could imagine host-imposed heme toxicity as a response to infection; consistent with this, B. anthracis induces heme efflux pumps during infection (126). Manganese is often considered to be protective against toxicity. However, manganese efflux is important in multiple pathogens, revealing distinct metal niches in the host. For example, resistance to manganese toxicity by the manganese exporter MntX is conserved in N. meningitidis but not N. gonorrhoeae, suggesting that manganese homeostasis differs in their infectious niches (134). MntX appears to be important for maintaining the appropriate manganese/iron ratio, which is particularly important in low-iron environments (134). Similarly, the manganese efflux protein MntE is required for S. pneumoniae pathogenesis in murine nasal passages and dissemination to blood (116). S. pyogenes also requires MntE for virulence and resistance to manganese toxicity and hydrogen peroxide stress, apparently due to dysregulation of metal homeostasis (132). The maintenance of manganese efflux systems suggests that although the host may not impose manganese intoxication per se, it may manipulate metal ratios to disrupt bacterial metal homeostasis. HOST METAL HOMEOSTASIS AND BACTERIAL INFECTIONS Host metal status can dramatically affect susceptibility to bacterial infections. Changes in host metal homeostasis can cause increased metal availability for bacteria, which is thought to be particularly important for opportunistic pathogens that have not evolved the host-specific metal acquisition systems described above (141). Data establishing the link between host metal homeostasis and infection are often correlative and do not address bacterial mechanisms of increased pathogenicity; therefore, animal models of genetic and environmental variation in metal homeostasis offer potential avenues for molecular characterization. This section briefly reviews how genetic and dietary differences in host metal homeostasis can affect bacterial infection. Iron Overload Lowers Defenses to Infection Human genetic polymorphisms in iron homeostasis can confer increased susceptibility or resistance to bacterial infection. The hereditary blood disorders β-thalassemia and hemochromatosis can cause iron overload and are associated with increased rates of infection (141). In patients with β-thalassemia, which is caused by mutations in the hemoglobin β-chain, infectious disease is the second most common cause of death (>50% caused by Enterobacteriaceae) (138). Although iron overload is a consistent feature in β-thalassemia, its relationship to infection is muddied due to the prevalence of splenectomy and blood transfusion. However, in a murine model of β-thalassemia, the iron overload feature is isolated and confers increased susceptibility to infection by S. enterica and L. monocytogenes (6) (Table 1). Hemochromatosis is also associated with increased infection by multiple, normally noninvasive pathogens. The importance of iron has been corroborated in mouse models of infection by Vibrio vulnificus and nonpigmented Y. pestis (plague vaccine strains) (112, 146). Until recently, it was unclear which aspect of hemochromatosis caused this increase in infection: liver injury, high basal iron overload, or the inability to reduce serum iron in response to infection. Using a mouse model of hemochromatosis by hepcidin deficiency, acute hypoferremia in response to infection was isolated as the critical mechanism of increased infection by V. vulnificus (8). Increased susceptibility could be partially rescued by dietary iron depletion or administration of hepcidin agonists, suggesting therapeutic potential to treat patients with hereditary iron overload and infection (8). Excess iron can also develop from environmental exposure through occupational hazards and diet. For example, workers exposed to high levels of iron-containing dust have increased rates of respiratory tract infection (107). A cluster of pneumonia cases caused by the opportunistic pathogen Acinetobacter calcoaceticus was linked to iron inhalation by foundry workers (31). Excess iron can be induced by dietary iron overload, generally due to inappropriate iron supplementation. Dietary iron overload has caused shifts from commensalism to pathogenesis by Yersinia enterocolitica, and from latent to acute forms of tuberculosis (48, 91). The fact that zinc and manganese are similarly restricted by nutritional immunity suggests that dietary overload may also increase susceptibility to infection. Dietary Metal Deficiencies and Infection There are well-established links between infection and dietary deficiencies of iron, copper, and zinc. Iron deficiency is specifically associated with H. pylori infection, which is a leading cause of gastric cancer. Work in a gerbil model showed that iron restriction by host dietary deficiency or limitation in vitro enhanced H. pylori pathogenesis via increased expression and elaboration of virulence factors (101). H. pylori infection can also induce iron deficiency, highlighting the complex nature of trace metals at the host-pathogen interface. In contrast, sufficient dietary copper and zinc are required to resist infection generally, which has been the subject of previous reviews (39, 55). Copper and zinc are important for immune development and can serve as bacterial intoxicants, and intracellular zinc levels serve as an immune signal. Copper deficiency is associated with neutropenia and increases susceptibility to infection by Mannheimia haemolytica and S. enterica in animal models (70, 99). In addition to neutrophil development, dietary copper is important for phagosomal copper intoxication: Copper supplementation of guinea pig food led to increased Mycobacterium tuberculosis killing (145). Approximately one-third of the world's population suffers from zinc deficiency, which is associated with increased rates of diarrhea and pneumonia (111). Prophylactic zinc supplementation was found to reduce incidence of persistent diarrhea by 33% and pneumonia by 41% in a pooled analysis; zinc as a therapeutic reduced prevalence of diarrhea by 34% but had no effect on pneumonia (13). Because zinc is critical for immune function, the role of enhanced bacterial virulence in a zinc-deficient host is unclear. Although rodent studies addressing the effect of zinc deficiency on pneumonia and diarrheal infection are limited, zinc deficiency was shown to increase susceptibility to L. monocytogenes, pneumococcal pneumonia, and diarrheal infection by enteroaggregative E. coli (EAEC), and to alter the time course of Mycobacterium bovis infection (17, 22, 90, 128). Notably, EAEC had increased expression of putative virulence factors in the cecal contents of zinc-deficient mice, suggesting that host zinc deficiency can alter bacterial pathogenesis (17). CONCLUDING REMARKS Transition metal biology at the host-pathogen interface is an active area of cross-disciplinary research spanning chemistry, biochemistry, microbiology, immunology, human genetics, and environmental and nutrition sciences. The field presents a number of opportunities for therapeutic potential. As secreted public goods, bacterial siderophores present a promising therapeutic target with reduced selective pressure for resistance, because the bacterium that evolves resistance does not necessarily benefit. For example, gallium (Ga2+) quenching of P. aeruginosa siderophores reduced pathogenesis in a Galleria mellonella model and retained efficacy longer than the conventional antibiotics gentamicin and ciprofloxacin (118). If a bacterium were to evolve production of gallium-resistant siderophore, it would be secreted and available to the entire bacterial population. Therefore, the producer of the gallium-resistant siderophore would not gain a selective advantage. In this way, targeting bacterial public goods decouples resistance and selective pressure. The fact that gallium siderophore quenching takes advantage of its interchangeability with biologically relevant transition metals suggests that gallium may also be toxic to the host due to host protein mismetallation. Therefore, other methods to target siderophores present attractive therapeutic opportunities. For example, vaccines against bacterial siderophores could similarly decouple resistance and selective pressure without causing metal intoxication of the host. Additional areas of interest include bacterial metal homeostasis, evolution of bacterial pathogenesis, the role of nutritional immunity in modulating bacterial social behavior, and the effect of host metal nutrition on bacterial virulence. Recent work defining a low molecular weight thiol as the labile zinc buffer in Bacillus subtilis has expanded our understanding of metal homeostasis in bacteria, but the role of metal buffering and metallochaperones in pathogenesis remains largely unexplored (82). Recent work on the Acinetobacter baumanii response to zinc starvation identified a putative Zn metallochaperone and that intracellular histidine contributes to the labile zinc pool, expanding our understanding of zinc homeostasis in this opportunistic pathogen (97). We expect that additional targets of metal piracy will open new avenues for the study of rapid evolution at the host-pathogen interface. Additionally, polymicrobial infections can modulate the metal environment in the host. Infection by respiratory syncytial virus causes transferrin release, increasing available iron and promoting P. aeruginosa infection and biofilm formation (61). Future work will likely continue to uncover the role of metal homeostasis in bacterial pathogenesis and additional mechanisms by which metals influence coinfections. The effect of host dietary metals on bacterial virulence also remains open for exploration. For example, future work could expand the effect of nutrient metals on gut microbial ecology and colonization resistance to pathogens. Transposon mutagenesis followed by selection and high-throughput sequencing of insertion sites could help determine whether bacterial pathogens can take advantage of altered nutrient metal landscapes in copper- and zinc-deficient hosts. Although the requirement of transition metals for enzyme catalysis likely predates life itself, the battle for metals at the host-pathogen interface remains a hotbed of evolution and continues to present opportunities for bettering human health. ACKNOWLEDGMENTS We thank members of the Skaar laboratory for critical reading of the manuscript and Sarah A. Marcus for helpful discussion. The writing of this manuscript was supported through fellowships to L.D.P. from the National Institutes of Health (NIH) (5T32HL094296-08/1F32AI122516-01) and research grants to E.P.S. from the NIH (R01 AI069233, R01 AI107233, R01 AI101171), US Department of Veterans Affairs (I01 BX002482/BX/BLRD), Defense Advanced Research Projects Agency, and American Asthma Foundation. Figure 1 Iron at the host-pathogen interface. In the host, Fe3+ is limited because of chelation by lactoferrin (LF) at the mucosal surface, transferrin (TF) in blood and tissue, and ferritin (F) in the cellular cytoplasm and at low levels in serum. Some bacteria can acquire iron from these host ferroproteins. In the bloodstream, bacteria can liberate iron from erythrocytes by hemolysis followed by extraction of heme from hemoprotein complexes, including hemoglobin (Hb), hemoglobin-haptoglobin (HP), and hemopexin (HPX). Intracellular pathogens are iron starved by Nramp1-mediated iron efflux from the phagosome. Nramp1 expression further induces lipocalin 2 production to bind and sequester some bacterial siderophores, whereas stealth siderophores evade binding. Figure 2 Iron acquisition in bacterial pathogens. (a) Gram-negative pathogens use porins or TonB-dependent outer membrane (OM) transporters to mediate the passage of iron complexes to the periplasm and, subsequently, the Fe2+ transporter Feo or ABC transporters to transport iron complexes into the cytoplasm. Pseudomonas aeruginosa reduces Fe3+-siderophores in the periplasm and exports deferrated siderophores from the periplasm via an ABC transporter (process not shown). P. aeruginosa and Serratia marescens use the secreted hemophore HasA and its receptor HasR for heme acquisition. Some bacteria also encode transferrin or lactoferrin OM receptors to actively pirate Fe3+ from host nutritional immunity proteins. (b) Gram-positive pathogens acquire iron using siderophores or dedicated heme/hemoprotein uptake systems. Pathogenic Staphylococcus spp. and Bacillus spp. use the Isd system to mediate the passage of heme through the cell wall to a dedicated transporter. Additional species-specific features include the Staphylococcus lugdunensis autolysin IsdP and Bacillus anthracis–secreted hemophores IsdX1 and IsdX2. Corynebacterium diptheriae uses the lipoprotein heme transporters HtaA/HtaB. (c) Portions of the mycobacterial iron acquisition process have been elucidated. Mycobacteria couple Fe3+-siderophore import with reduction and Fe2+ release; the deferrated siderophores can then be exported by the MmpL4/5-MmpS4/5 system for recycling. Heme import requires the secreted hemophore Rv0203 and the CM importers MmpL3/11. Abbreviations: ABC, ATP-binding cassette; CM, cytoplasmic membrane; Hb, hemoglobin; HP, hemoglobin-haptoglobin; LF, lactoferrin; TF, transferrin. Figure 3 Zinc sequestration and bacterial zinc import. (a) Pathogenic bacteria acquire zinc using ABC transporters ZnuABC (Gram-negative bacteria) and AdcABC (Gram-positive bacteria; not shown). Zinc is transported through the outer membrane (OM) using the TonB-dependent transporter ZnuD. Additional zinc acquisition systems include a T6SS-secreted zincophore protein YezP in Yersinia pseudotuberculosis, transport of yersiniabactin-Zn2+ in Y. pestis, and an OM receptor CbpA that binds calprotectin (CP) in Neisseria meningitidis. (b) During infection, zinc is limited by secretion of zinc-chelating S100 proteins: S100A7 is secreted at epithelial surfaces, whereas S100A12 and CP are secreted by innate immune proteins such as neutrophils. Some bacteria can acquire zinc from CP. Abbreviations: ABC, ATP-binding cassette; CM, cytoplasmic membrane; T6SS, type VI secretion system. Figure 4 Intra- and extracellular manganese restriction. (a) Gram-negative pathogens import manganese with the SitABCD or MntH transporters; the mechanism of transport across the outer membrane (OM) is unknown. (b) Gram-positive pathogens import manganese with the MntABC or MntH transporter. (c) The phagosome membrane protein Nramp1 effluxes manganese from the phagosome, limiting availability for intracellular pathogens. Extracellularly, manganese is sequestered by calprotectin (CP), which is secreted by innate immune cells, including neutrophils, in response to infection. Abbreviations: ABC transporter, ATP-binding cassette transporter; CM, cytoplasmic membrane. Figure 5 Copper intoxication and bacterial resistance. (a) Evidence thus far suggests that copper is toxic because it displaces metal cofactors in certain proteins, including iron-sulfur clusters and manganese in ribonucleotide reductase (RNR). Bacteria have evolved diverse mechanisms to withstand copper toxicity, including copper efflux. In uropathogenic Escherichia coli, extracellular yersiniabactin-Cu2+ prevents copper reduction to the more reactive Cu+ and can convert superoxide to hydrogen peroxide. E. coli also utilizes the periplasmic copper oxidase CueO to detoxify Cu+. Mycobacteria also produce a metallothionein, MymT, to sequester cytoplasmic copper. (b) In response to infection, phagocytic cells, including macrophages, import copper into the cytosol with the transporter CTR1. Copper is then shuttled by ATOX1 to the phagolysosomal membrane, where it is transported into the phagolysosome by ATP7A to intoxicate bacterial cells. Figure 6 Zinc intoxication in innate immune cells. In response to infection, innate immune cells accumulate zinc in the cytoplasm through ZIP8-mediated import and in the phagosome by an unknown mechanism. Zinc accumulation in the cytoplasm modulates nuclear factor (NF)-κB activation, dampening proinflammatory responses. Bacterial zinc poisoning inhibits manganese import and the manganese-utilizing glycolytic enzymes (Gly) phosphofructokinase and glyceraldehyde-3-phosphate dehydrogenase. Bacteria express zinc efflux proteins in response. Table 1 Selected bacterial species with improved virulence in host environments with metal dysregulation Bacterial species Host metal dysregulation Host Effect on infection Reference Acinetobacter calcoaceticus Occupational inhaled iron exposure Human Cluster of A. calcoaceticus pneumonia cases with two deaths 31 Enteroaggregative Escherichia coli Dietary zinc deficiency Mouse Enhanced disease 17 Helicobacter pylori Dietary iron deficiency Gerbil Enhanced bacterial burdens and carcinogenesis 101 Listeria monocytogenes Iron overload from β-thalassemia Mouse Increased mortality and bacterial burdens 6 Dietary zinc deficiency Rat Reduced immune response 22 Mannheimia hemolytica Dietary copper deficiency Mouse Reduced lethal dose 70 Mycobacterium bovis Dietary zinc deficiency Guinea pig Enhanced bacterial burdens early in infection 90 Mycobacterium tuberculosis Dietary iron overload Human Shift from latent to active infection 48 Salmonella enterica Iron overload from β-thalassemia Mouse Increased mortality 6 Dietary zinc deficiency Rat Increased mortality 99 Streptococcus pneumoniae Dietary zinc deficiency Mouse Increased mortality and bacterial burdens 128 Vibrio vulnificus Iron overload by injection Mouse Reduced lethal dose 146 Hepcidin deficiency Mouse Increased bacteremia and mortality, partially rescued by dietary iron depletion 8 Yersinia enterocolitica Dietary iron overload Human Infection by commensal species 91 Yersinia pestis Iron overload from hemochromatosis Human, mouse Restored virulence of attenuated vaccine strain 112 NUTRITIONAL IMMUNITY PROTEIN MOONLIGHTS AS AN ANTIMICROBIAL Beyond its role in nutritional immunity, lactoferrin can inhibit bacterial growth by a variety of mechanisms including generation of antimicrobial peptides and binding of bacterial cell wall components (136). Bacterial pathogens have evolved systems to protect against these activities. Lactoferrin generates the antimicrobial peptides lactoferricin and lactoferrampin during bacterial infection. Recent work revealed rapid evolution in primates at the binding interface of lactoferricin and bacterial proteins that prevent lactoferrin-mediated killing (12). Additionally, the Staphylococcus aureus heme acquisition protein IsdA protects against lactoferrin-mediated killing by inhibiting lactoferrin's serine protease activity (28). Thus, resistance to a host nutritional immunity protein moonlighting as an antimicrobial agent relies on a bacterial metal acquisition protein moonlighting as a protease inhibitor. These evolutions/counter-revolutions exemplify the Red Queen hypothesis in bacterial pathogenesis. SUMMARY POINTS Although bacterial pathogens use vertebrate hosts as nutrient metal sources, hosts possess multiple mechanisms to restrict bacterial access to transition metals through nutritional immunity. Bacteria evade host metal sequestration through metal acquisition strategies, including stealth siderophores and metal piracy from host proteins. The coevolution of pathogens and vertebrate hosts at the nutrient metal interface illustrates the Red Queen hypothesis in evolutionary biology. Vertebrate hosts barrage bacteria with metals during intracellular infection, exploiting the toxicity of high concentrations of copper and zinc. The host intoxicates intraphagosomal bacteria by transporting high levels of copper with ATP7A. The mechanism of zinc intoxication and its relationship to zinc innate immune signaling is still being elucidated. Resistance to copper and zinc toxicity is important for virulence in a number of pathogens. Host metal dysregulation through genetic disease or dietary metal intake affects susceptibility to infection by many bacterial pathogens. In general, iron overload, zinc deficiency, and copper deficiency increase susceptibility to infection. During H. pylori infection, iron deficiency accelerates production of bacterial virulence factors and carcinogenesis, demonstrating that changes in dietary metals can affect bacterial virulence. The interchangeability of transition metal binding requires all organisms, including bacterial pathogens and their vertebrate hosts, to maintain appropriate metal homeostasis to prevent toxicity and mismetallation. The role of metal homeostasis at the host-pathogen interface remains an important area of investigation. Transition metals: d-block elements in rows 3–12 of the periodic table that generally have partially filled d-orbitals and produce many oxidation states Siderophores: small molecules that bind Fe3+ with high affinity secreted by microbes to acquire iron from the environment Transferrin: a protein that sequesters iron in blood, transporting iron to the rest of the body following absorption in the duodenum Red Queen hypothesis: a concept referencing Lewis Carroll's Through the Looking Glass to describe evolutionary pressures on organisms in close association Heme: Fe2+-protoporphyrin IX, a cofactor for cytochrome oxidases used for respiration in bacteria and eukaryotes ATP-binding cassette (ABC) transporters: a family of proteins that transport molecules across the cytoplasmic membrane via ATP hydrolysis Public goods: a term originating from economics to describe resources that may be costly to produce but are freely accessible to all Tragedy of the commons: refers to loss of a public resource from overuse, from economist William Forster Lloyd's phrase describing overgrazing of common lands Hemopexin: a protein that binds free heme to protect against its oxidative effects and recycle the body's iron Haptoglobin: a protein that binds free hemoglobin to protect against its oxidative effects; most hemoglobin-haptoglobin is then removed by the spleen TonB-ExbB-ExbD system: a protein complex that harnesses the proton motive force for active transport across the outer membrane in Gram-negative bacteria Type VI secretion system (T6SS): a system used by Gram-negative bacteria to transport effector proteins into host or bacterial target cells Metallothioneins: a family of low molecular weight cytosolic proteins rich in cysteines that bind heavy metals on their thiol groups S100 protein family: Ca2+-binding proteins with a helix-loop-helix (EF-hand type) conformation that often bind additional metals and serve as nutritional immunity proteins Natural resistance–associated macrophage protein 1 (Nramp1): a protein that exports iron and manganese from the phagosome and enhances resistance to intracellular pathogens β-thalassemia: a hereditary blood disorder caused by defects in synthesis of the hemoglobin β-chain that can cause iron overload Hemochromatosis: hereditary blood disorder caused by mutations in transferrin receptor–interacting genes and characterized by iron overload Iron overload: an accumulation of iron in the body due to genetic diseases, repeated blood transfusion, or excess iron consumption DISCLOSURE STATEMENT The authors are not aware of any affiliations, memberships, funding, or financial holdings that might be perceived as affecting the objectivity of this review. LITERATURE CITED/RELATED RESOURCES 1 Abergel RJ Wilson MK Arceneaux JE Hoette TM Strong RK Anthrax pathogen evades the mammalian immune system through stealth siderophore production. PNAS 2006 103 18499 503 17132740 2 Achouiti A Vogl T Endeman H Mortensen BL Laterre PF Myeloid-related protein-8/14 facilitates bacterial growth during pneumococcal pneumonia. Thorax 2014 69 1034 42 25179663 3 Achouiti A Vogl T Urban CF Rohm M Hommes TJ Myeloid-related protein-14 contributes to protective immunity in gram-negative pneumonia derived sepsis. PLOS Pathog 2012 8 e1002987 23133376 4 Allen CE Schmitt MP HtaA is an iron-regulated hemin binding protein involved in the utilization of heme iron in Corynebacterium diphtheriae. J. Bacteriol 2009 191 2638 48 19201805 5 Almiron M Martinez M Sanjuan N Ugalde RA Ferrochelatase is present in Brucella abortus and is critical for its intracellular survival and virulence. Infect. Immun 2001 69 6225 30 11553564 6 Ampel NM van Wyck DB Aguirre ML Willis DG Popp RA Resistance to infection in murine β-thalassemia. Infect. Immun 1989 57 1011 17 2925237 7 Andersen SB Marvig RL Molin S Johansen HK Griffin AS Long-term social dynamics drive loss of function in pathogenic bacteria. PNAS 2015 112 10756 61 Evolutionary analysis of long-term infection showed that local competition drives siderophore cheating and community loss-of-function. 26240352 8 Arezes J Jung G Gabayan V Valore E Ruchala P Hepcidin-induced hypoferremia is a critical host defense mechanism against the siderophilic bacterium Vibrio vulnificus. Cell Host Microbe 17 47 57 9 Bachman MA Lenio S Schmidt L Oyler JE Weiser JN Interaction of lipocalin 2, transferrin, and siderophores determines the replicative niche of Klebsiella pneumoniae during pneumonia. mBio 2012 3 e00224 11 Evolutionary and biochemical analyses showed that rapid evolution drives host-pathogen interactions in metal acquisition. 23169997 10 Barber MF Elde NC Escape from bacterial iron piracy through rapid evolution of transferrin. Science 2014 346 1362 66 25504720 11 Barber MF Elde NC Buried treasure: evolutionary perspectives on microbial iron piracy. Trends Genet 2015 31 627 36 26431675 12 Barber MF Kronenberg Z Yandell M Elde NC Antimicrobial functions of lactoferrin promote genetic conflicts in ancient primates and modern humans. PLOS Genet 2016 12 5 e1006063 27203426 13 Bhutta ZA Black RE Brown KH Gardner JM Gore S Prevention of diarrhea and pneumonia by zinc supplementation in children in developing countries: pooled analysis of randomized controlled trials. J. Pediatr 1999 135 689 97 10586170 14 Blackwell JM Searle S Mohamed H White JK Divalent cation transport and susceptibility to infectious and autoimmune disease: continuation of the Ity/Lsh/Bcg/Nramp1/Slc11a1 gene story. Immunol. Lett 2003 85 197 203 12527228 15 Blair DE Schuttelkopf AW MacRae JI van Aalten DM Structure and metal-dependent mechanism of peptidoglycan deacetylase, a streptococcal virulence factor. PNAS 2005 102 15429 34 16221761 16 Bobrov AG Kirillina O Fetherston JD Miller MC Burlison JA Perry RD The Yersinia pestis siderophore, yersiniabactin, and the ZnuABC system both contribute to zinc acquisition and the development of lethal septicaemic plague in mice. Mol. Microbiol 2014 93 759 75 This study discovered a dedicated Zn2+-siderophore importer and established its importance during infection. 24979062 17 Bolick DT Kolling GL Moore JH Jr de Oliveira LA Tung K Zinc deficiency alters host response and pathogen virulence in a mouse model of enteroaggregative Escherichia coli–induced diarrhea. Gut Microbes 2014 5 618 27 25483331 18 Botella H Peyron P Levillain F Poincloux R Poquet Y Mycobacterial P1-type ATPases mediate resistance to zinc poisoning in human macrophages. Cell Host Microbe 2011 10 248 59 21925112 19 Braun V Hantke K Recent insights into iron import by bacteria. Curr. Opin. Chem. Biol 2011 15 328 34 21277822 20 Braymer JJ Giedroc DP Recent developments in copper and zinc homeostasis in bacterial pathogens. Curr. Opin. Chem. Biol 2014 19 59 66 24463765 21 Burne RA Chen YY Bacterial ureases in infectious diseases. Microbes Infect 2000 2 533 42 10865198 22 Carlomagno MA Coghlan LG McMurray DN Chronic zinc deficiency and listeriosis in rats: acquired cellular resistance and response to vaccination. Med. Microbiol. Immunol 1986 175 271 80 3093828 23 Carrano CJ Raymond KN Ferric ion sequestering agents. 2. Kinetics and mechanism of iron removal from transferrin by enterobactin and synthetic tricatechols. J. Am. Chem. Soc 1979 101 5401 4 24 Cassat JE Skaar EP Iron in infection and immunity. Cell Host Microbe 2013 13 509 19 23684303 25 Chaturvedi KS Henderson JP Pathogenic adaptations to host-derived antibacterial copper. Front. Cell Infect. Microbiol 2014 4 3 24551598 26 Chaturvedi KS Hung CS Crowley JR Stapleton AE Henderson JP The siderophore yersiniabactin binds copper to protect pathogens during infection. Nat. Chem. Biol 2012 8 731 36 Demonstrated that yersiniabactin secreted by uropathogenic Escherichia coli binds copper to resist toxicity during infection. 22772152 27 Choby JE Skaar EP Heme acquisition and synthesis in bacterial pathogens. J. Mol. Biol 2016 In press. doi: 10.1016/j.jmb.2016.03.018 28 Clarke SR Foster SJ IsdA protects Staphylococcus aureus against the bactericidal protease activity of apolactoferrin. Infect. Immun 2008 76 1518 26 18227165 29 Corbett D Wang J Schuler S Lopez-Castejon G Glenn S Two zinc uptake systems contribute to the full virulence of Listeria monocytogenes during growth in vitro and in vivo. Infect. Immun 2012 80 14 21 22025520 30 Corbin BD Seeley EH Raab A Feldmann J Miller MR Metal chelation and inhibition of bacterial growth in tissue abscesses. Science 2008 319 962 65 18276893 31 Cordes LG Brink EW Checko PJ Lentnek A Lyons RW A cluster of Acinetobacter pneumonia in foundry workers. Ann. Intern. Med 1981 95 688 93 7305146 32 Damo SM Kehl-Fie TE Sugitani N Holt ME Rathi S Molecular basis for manganese sequestration by calprotectin and roles in the innate immune response to invading bacterial pathogens. PNAS 2013 110 3841 46 This structural analysis of calprotectin established the molecular basis for high-affinity manganese binding. 23431180 33 Deriu E Liu JZ Pezeshki M Edwards RA Ochoa RJ Probiotic bacteria reduce Salmonella Typhimurium intestinal colonization by competing for iron. Cell Host Microbe 2013 14 26 37 23870311 34 Dhaenens L Szczebara F Husson MO Identification, characterization, and immunogenicity of the lactoferrin-binding protein from Helicobacter pylori. Infect. Immun 1997 65 514 18 9009306 35 Diaz-Ochoa VE Lam D Lee CS Klaus S Behnsen J Liu JZ Salmonella mediates oxidative stress and thrives in the inflamed gut by evading calprotectin-mediated manganese sequestration. Cell Host Microbe 2016 19 814 25 27281571 36 Dintilhac A Alloing G Granadel C Claverys JP Competence and virulence of Streptococcus pneumoniae: Adc and PsaA mutants exhibit a requirement for Zn and Mn resulting from inactivation of putative ABC metal permeases. Mol. Microbiol 1997 25 727 39 9379902 37 Djoko KY Franiek JA Edwards JL Falsetta ML Kidd SP Phenotypic characterization of a copA mutant of Neisseria gonorrhoeae identifies a link between copper and nitrosative stress. Infect. Immun 2012 80 1065 71 22184419 38 Djoko KY McEwan AG Antimicrobial action of copper is amplified via inhibition of heme biosynthesis. ACS Chem. Biol 2013 8 2217 23 23895035 39 Djoko KY Ong CL Walker MJ McEwan AG The role of copper and zinc toxicity in innate immune defense against bacterial pathogens. J. Biol. Chem 2015 290 18954 61 26055706 40 D'Orazio M Mastropasqua MC Cerasi M Pacello F Consalvo A The capability of Pseudomonas aeruginosa to recruit zinc under conditions of limited metal availability is affected by inactivation of the ZnuABC transporter. Metallomics 2015 7 1023 35 25751674 41 Eaton KA Krakowka S Effect of gastric pH on urease-dependent colonization of gnotobiotic piglets by Helicobacter pylori. Infect. Immun 1994 62 3604 7 8063376 42 Farrand AJ Haley KP Lareau NM Heilbronner S McLean JA An iron-regulated autolysin remodels the cell wall to facilitate heme acquisition in Staphylococcus lugdunensis. Infect. Immun 2015 83 3578 89 26123800 43 Fischbach MA Lin H Zhou L Yu Y Abergel RJ The pathogen-associated iroA gene cluster mediates bacterial evasion of lipocalin 2. PNAS 2006 103 16502 7 17060628 44 Flo TH Smith KD Sato S Rodriguez DJ Holmes MA Lipocalin 2 mediates an innate immune response to bacterial infection by sequestrating iron. Nature 2004 432 917 21 15531878 45 Fritsche G Nairz M Libby SJ Fang FC Weiss G Slc11a1 (Nramp1) impairs growth of Salmonella enterica serovar typhimurium in macrophages via stimulation of lipocalin-2 expression. J. Leukoc. Biol 2012 92 353 59 22706314 46 Gaddy JA Radin JN Cullen TW Chazin WJ Skaar EP Helicobacter pylori resists the antimicrobial activity of calprotectin via lipid A modification and associated biofilm formation. mBio 2015 6 e01349 15 26646009 47 Gaddy JA Radin JN Loh JT Piazuelo MB Kehl-Fie TE The host protein calprotectin modulates the Helicobacter pylori cag type IV secretion system via zinc sequestration. PLOS Pathog 2014 10 e1004450 25330071 48 Gangaidzo IT Moyo VM Mvundura E Aggrey G Murphree NL Association of pulmonary tuberculosis with increased dietary iron. J. Infect. Dis 2001 184 936 39 11528590 49 Ghoul M West SA Diggle SP Griffin AS An experimental test of whether cheating is context dependent. J. Evol. Biol 2014 27 551 56 24480013 50 Ghssein G Brutesco C Ouerdane L Fojcik C Izaute A Wang S Biosynthesis of a broad-spectrum nicotianamine-like metallophore in Staphylococcus aureus. Science 2016 352 6289 1105 9 27230378 51 Glaser R Harder J Lange H Bartels J Christophers E Schroder JM Antimicrobial psoriasin (S100A7) protects human skin from Escherichia coli infection. Nat. Immunol 2005 6 57 64 15568027 52 Gold B Deng H Bryk R Vargas D Eliezer D Identification of a copper-binding metal-lothionein in pathogenic mycobacteria. Nat. Chem. Biol 2008 4 609 16 18724363 53 Griffin AS West SA Buckling A Cooperation and competition in pathogenic bacteria. Nature 2004 430 1024 27 15329720 54 Guilhen C Taha MK Veyrier FJ Role of transition metal exporters in virulence: the example of Neisseria meningitidis. Front. Cell Infect. Microbiol 2013 3 102 24392357 55 Haase H Rink L Multiple impacts of zinc on immune function. Metallomics 2014 6 1175 80 24531756 56 Haley KP Delgado AG Piazuelo MB Mortensen BL Correa P The human antimicrobial protein calgranulin C participates in control of Helicobacter pylori growth and regulation of virulence. Infect. Immun 2015 83 2944 56 25964473 57 Hammer ND Cassat JE Noto MJ Lojek LJ Chadha AD Inter- and intraspecies metabolite exchange promotes virulence of antibiotic-resistant Staphylococcus aureus. Cell Host Microbe 2014 16 531 37 25299336 58 Hammer ND Reniere ML Cassat JE Zhang Y Hirsch AO Two heme-dependent terminal oxidases power Staphylococcus aureus organ-specific colonization of the vertebrate host. mBio 2013 4 e00241 13 23900169 59 Hammerschmidt S Bethe G Remane PH Chhatwal GS Identification of pneumococcal surface protein A as a lactoferrin-binding protein of Streptococcus pneumoniae. Infect. Immun 1999 67 1683 87 10085004 60 Hardin G The tragedy of the commons. Science 1968 162 1243 48 61 Hendricks MR Lashua LP Fischer DK Flitter BA Eichinger KM Respiratory syncytial virus infection enhances Pseudomonas aeruginosa biofilm growth through dysregulation of nutritional immunity. PNAS 2016 113 1642 47 26729873 62 Higgins KA Giedroc D Insights into protein allostery in the CsoR/RcnR family of transcriptional repressors. Chem. Lett 2014 43 20 25 24695963 63 Hiron A Posteraro B Carriere M Remy L Delporte C A nickel ABC-transporter of Staphylococcus aureus is involved in urinary tract infection. Mol. Microbiol 2010 77 1246 60 20662775 64 Hood MI Mortensen BL Moore JL Zhang Y Kehl-Fie TE Identification of an Acinetobacter baumannii zinc acquisition system that facilitates resistance to calprotectin-mediated zinc sequestration. PLOS Pathog 2012 8 e1003068 23236280 65 Hood MI Skaar EP Nutritional immunity: transition metals at the pathogen-host interface. Nat. Rev. Microbiol 2012 10 525 37 22796883 66 Imperi F Tiburzi F Visca P Molecular basis of pyoverdine siderophore recycling in Pseudomonas aeruginosa. PNAS 2009 106 20440 45 19906986 67 Jabado N Jankowski A Dougaparsad S Picard V Grinstein S Gros P Natural resistance to intracellular infections: natural resistance–associated macrophage protein 1 (Nramp1) functions as a pH-dependent manganese transporter at the phagosomal membrane. J. Exp. Med 2000 192 1237 48 11067873 68 Johnson MD Kehl-Fie TE Rosch JW Copper intoxication inhibits aerobic nucleotide synthesis in Streptococcus pneumoniae. Metallomics 2015 7 786 94 25730343 69 Jones CM Wells RM Madduri AV Renfrow MB Ratledge C Self-poisoning of Mycobacterium tuberculosis by interrupting siderophore recycling. PNAS 2014 111 1945 50 24497493 70 Jones DG Suttle NF The effect of copper deficiency on the resistance of mice to infection with Pasteurella haemolytica. J. Comp. Pathol 1983 93 143 49 6841690 71 Juttukonda LJ Skaar EP Manganese homeostasis and utilization in pathogenic bacteria. Mol. Microbiol 2015 97 216 28 25898914 72 Kehl-Fie TE Chitayat S Hood MI Damo S Restrepo N Nutrient metal sequestration by calprotectin inhibits bacterial superoxide defense, enhancing neutrophil killing of Staphylococcus aureus. Cell Host Microbe 2011 10 158 64 21843872 73 Kehl-Fie TE Zhang Y Moore JL Farrand AJ Hood MI MntABC and MntH contribute to systemic Staphylococcus aureus infection by competing with calprotectin for nutrient manganese. Infect. Immun 2013 81 3395 405 23817615 74 Kim HW Chan Q Afton SE Caruso JA Lai B Human macrophage ATP7A is localized in the trans-Golgi apparatus, controls intracellular copper levels, and mediates macrophage responses to dermal wounds. Inflammation 2012 35 167 75 21336677 75 Koh EI Hung CS Parker KS Crowley JR Giblin DE Henderson JP Metal selectivity by the virulence-associated yersiniabactin metallophore system. Metallomics 2015 7 1011 22 25824627 76 Krieg S Huche F Diederichs K Izadi-Pruneyre N Lecroisey A Heme uptake across the outer membrane as revealed by crystal structures of the receptor-hemophore complex. PNAS 2009 106 1045 50 19144921 77 Lau CK Krewulak KD Vogel HJ Bacterial ferrous iron transport: the Feo system. FEMS Microbiol. Rev 2015 40 273 98 26684538 78 Lisher JP Giedroc DP Manganese acquisition and homeostasis at the host-pathogen interface. Front. Cell Infect. Microbiol 2013 3 91 24367765 79 Liu JZ Jellbauer S Poe AJ Ton V Pesciaroli M Zinc sequestration by the neutrophil protein calprotectin enhances Salmonella growth in the inflamed gut. Cell Host Microbe 2012 11 227 39 22423963 80 Liu MJ Bao S Galvez-Peralta M Pyle CJ Rudawsky AC ZIP8 regulates host defense through zinc-mediated inhibition of NF-κB. Cell Rep 2013 3 386 400 23403290 81 Loomis LD Raymond KN Solution equilibria of enterobactin and metal-enterobactin complexes. Inorg. Chem 1991 30 906 11 82 Ma Z Chandrangsu P Helmann TC Romsang A Gaballa A Helmann JD Bacillithiol is a major buffer of the labile zinc pool in Bacillus subtilis. Mol. Microbiol 2014 94 756 70 25213752 83 Macomber L Imlay JA The iron-sulfur clusters of dehydratases are primary intracellular targets of copper toxicity. PNAS 2009 106 8344 49 19416816 84 Maresso AW Garufi G Schneewind O Bacillus anthracis secretes proteins that mediate heme acquisition from hemoglobin. PLOS Pathog 2008 4 e1000132 18725935 85 Marvig RL Damkiaer S Khademi SM Markussen TM Molin S Jelsbak L Within-host evolution of Pseudomonas aeruginosa reveals adaptation toward iron acquisition from hemoglobin. mBio 2014 5 e00966 14 Independent Pseudomonas aeruginosa isolates from cystic fibrosis patients lost siderophore acquisition, evolving a heme iron preference. 24803516 86 Mashruwala AA Pang YY Rosario-Cruz Z Chahal HK Benson MA Nfu facilitates the maturation of iron-sulfur proteins and participates in virulence in Staphylococcus aureus. Mol. Microbiol 2015 95 383 409 25388433 87 Mazmanian SK Skaar EP Gaspar AH Humayun M Gornicki P Passage of heme-iron across the envelope of Staphylococcus aureus. Science 2003 299 906 9 12574635 88 McDevitt CA Ogunniyi AD Valkov E Lawrence MC Kobe B A molecular mechanism for bacterial susceptibility to zinc. PLOS Pathog 2011 7 e1002357 22072971 89 McGee DJ Zabaleta J Viator RJ Testerman TL Ochoa AC Mendz GL Purification and characterization of Helicobacter pylori arginase, RocF: unique features among the arginase superfamily. Eur. J. Biochem 2004 271 1952 62 15128304 90 McMurray DN Yetley EA Response to Mycobacterium bovis BCG vaccination in protein- and zinc-deficient guinea pigs. Infect. Immun 1983 39 755 61 6339390 91 Melby K Slordahl S Gutteberg TJ Nordbo SA Septicaemia due to Yersinia enterocolitica after oral overdoses of iron. BMJ (Clin. Res. Ed.) 1982 285 467 68 92 Meyer JE Harder J Sipos B Maune S Kloppel G Psoriasin (S100A7) is a principal antimicrobial peptide of the human tongue. Mucosal Immunol 2008 1 239 43 19079183 93 Miethke M Marahiel MA Siderophore-based iron acquisition and pathogen control. Microbiol. Mol. Biol. Rev 2007 71 413 51 17804665 94 Mildner M Stichenwirth M Abtin A Eckhart L Sam C Psoriasin (S100A7) is a major Escherichia coli-cidal factor of the female genital tract. Mucosal Immunol 2010 3 602 9 20571488 95 Moraes TF Yu RH Strynadka NC Schryvers AB Insights into the bacterial transferrin receptor: the structure of transferrin-binding protein B from Actinobacillus pleuropneumoniae. Mol. Cell 2009 35 523 33 19716795 96 Morgenthau A Pogoutse A Adamiak P Moraes TF Schryvers AB Bacterial receptors for host transferrin and lactoferrin: molecular mechanisms and role in host-microbe interactions. Future Microbiol 2013 8 1575 85 24266357 97 Nairn BL Lonergan ZR Wang J Braymer JJ Zhang Y Calcutt MW The response of Acinetobacter baumannii to zinc starvation. Cell Host Microbe 2016 19 826 36 Identified a putative zinc metallochaperone and the contribution of histidine to the labile zinc pool. 27281572 98 Nakashige TG Zhang B Krebs C Nolan EM Human calprotectin is an iron-sequestering host-defense protein. Nat. Chem. Biol 2015 11 765 71 26302479 99 Newberne PM Hunt CE Young VR The role of diet and the reticuloendothelial system in the response of rats to Salmonella typhilmurium infection. Br. J. Exp. Pathol 1968 49 448 57 4882436 100 Nolan KJ McGee DJ Mitchell HM Kolesnikow T Harro JM In vivo behavior of a Helicobacter pylori SS1 nixA mutant with reduced urease activity. Infect. Immun 2002 70 685 91 11796600 101 Noto JM Gaddy JA Lee JY Piazuelo MB Friedman DB Iron deficiency accelerates Helicobacter pylori–induced carcinogenesis in rodents and humans. J. Clin. Invest 2013 123 479 92 23257361 102 Oftung F Lovik M Andersen SR Froholm LO Bjune G A mouse model utilising human transferrin to study protection against Neisseria meningitidis serogroup B induced by outer membrane vesicle vaccination. FEMS Immunol. Med. Microbiol 1999 26 75 82 10518045 103 Ong CL Gillen CM Barnett TC Walker MJ McEwan AG An antimicrobial role for zinc in innate immune defense against group A Streptococcus. J. Infect. Dis 2014 209 1500 8 Showed zinc mobilization in the neutrophil cytoplasm as one mechanism for host zinc poisoning of pathogens. 24449444 104 Ong CL Walker MJ McEwan AG Zinc disrupts central carbon metabolism and capsule biosyn-thesis in Streptococcus pyogenes. Sci. Rep 2015 5 10799 26028191 105 Oram JD Reiter B Inhibition of bacteria by lactoferrin and other iron-chelating agents. Biochim. Biophys. Acta 1968 170 351 65 4974829 106 Ouyang Z He M Oman T Yang XF Norgard MV A manganese transporter, BB0219 (BmtA), is required for virulence by the Lyme disease spirochete, Borrelia burgdorferi. PNAS 2009 106 3449 54 19218460 107 Palmer K Coggon D Does occupational exposure to iron promote infection? Occup. Environ. Med 1997 54 529 34 9326155 108 Perry RD Craig SK Abney J Bobrov AG Kirillina O Manganese transporters Yfe and MntH are Fur-regulated and important for the virulence of Yersinia pestis. Microbiology 2012 158 804 15 22222497 109 Pishchany G McCoy AL Torres VJ Krause JC Crowe JE Jr Specificity for human hemoglobin enhances Staphylococcus aureus infection. Cell Host Microbe 2010 8 544 50 21147468 110 Posey JE Gherardini FC Lack of a role for iron in the Lyme disease pathogen. Science 2000 288 1651 53 10834845 111 Prasad AS Discovery of human zinc deficiency: its impact on human health and disease. Adv. Nutr 2013 4 176 90 23493534 112 Quenee LE Hermanas TM Ciletti N Louvel H Miller NC Hereditary hemochromatosis restores the virulence of plague vaccine strains. J. Infect. Dis 2012 206 1050 58 22896664 113 Raffatellu M George MD Akiyama Y Hornsby MJ Nuccio SP Lipocalin-2 resistance confers an advantage to Salmonella enterica serotype Typhimurium for growth and survival in the inflamed intestine. Cell Host Microbe 2009 5 476 86 19454351 114 Remy L Carriere M Derre-Bobillot A Martini C Sanguinetti M Borezee-Durant E The Staphylococcus aureus Opp1 ABC transporter imports nickel and cobalt in zinc-depleted conditions and contributes to virulence. Mol. Microbiol 2013 87 730 43 23279021 115 Rodionov DA Hebbeln P Gelfand MS Eitinger T Comparative and functional genomic analysis of prokaryotic nickel and cobalt uptake transporters: evidence for a novel group of ATP-binding cassette transporters. J. Bacteriol 2006 188 317 27 16352848 116 Rosch JW Gao G Ridout G Wang YD Tuomanen EI Role of the manganese efflux system mntE for signaling and pathogenesis in Streptococcus pneumoniae. Mol. Microbiol 2009 72 12 25 19226324 117 Ross-Gillespie A Dumas Z Kummerli R Evolutionary dynamics of interlinked public goods traits: an experimental study of siderophore production in Pseudomonas aeruginosa. J. Evol. Biol 2015 28 29 39 25421271 118 Ross-Gillespie A Weigert M Brown SP Kummerli R Gallium-mediated siderophore quenching as an evolutionarily robust antibacterial treatment. Evol. Med. Public Health 2014 2014 18 29 24480613 119 Rowland JL Niederweis M A multicopper oxidase is required for copper resistance in Mycobacterium tuberculosis. J. Bacteriol 2013 195 3724 33 23772064 120 Schade AL Caroline L Raw hen egg white and the role of iron in growth inhibition of Shigella dysenteriae, Staphylococcus aureus, Escherichia coli and Saccharomyces cerevisiae. Science 1944 100 14 15 17783793 121 Schade AL Caroline L An iron-binding component in human blood plasma. Science 1946 104 340 122 Shields-Cutler RR Crowley JR Hung CS Stapleton AE Aldrich CC Human urinary composition controls antibacterial activity of siderocalin. J. Biol. Chem 2015 290 15949 60 25861985 123 Skaar EP Humayun M Bae T DeBord KL Schneewind O Iron-source preference of Staphylococcus aureus infections. Science 2004 305 1626 28 15361626 124 Speer A Rowland JL Haeili M Niederweis M Wolschendorf F Porins increase copper susceptibility of Mycobacterium tuberculosis. J. Bacteriol 2013 195 5133 40 24013632 125 Stahler FN Odenbreit S Haas R Wilrich J van Vliet AH The novel Helicobacter pylori CznABC metal efflux pump is required for cadmium, zinc, and nickel resistance, urease modulation, and gastric colonization. Infect. Immun 2006 74 3845 52 16790756 126 Stauff DL Skaar EP Bacillus anthracis HssRS signalling to HrtAB regulates haem resistance during infection. Mol. Microbiol 2009 72 763 78 19400785 127 Stork M Grijpstra J Bos MP Manas Torres C Devos N Zinc piracy as a mechanism of Neisseria meningitidis for evasion of nutritional immunity. PLOS Pathog 2013 9 e1003733 Neisseria meningitidis CbpA binds calprotectin and supports zinc-dependent growth, suggesting a potential for zinc piracy. 24204275 128 Strand TA Hollingshead SK Julshamn K Briles DE Blomberg B Sommerfelt H Effects of zinc deficiency and pneumococcal surface protein A immunization on zinc status and the risk of severe infection in mice. Infect. Immun 2003 71 2009 13 12654820 129 Sun TS Ju X Gao HL Wang T Thiele DJ Reciprocal functions of Cryptococcus neoformans copper homeostasis machinery during pulmonary infection and meningoencephalitis. Nat. Commun 2014 5 5550 25417972 130 Troxell B Hassan HM Transcriptional regulation by Ferric Uptake Regulator (Fur) in pathogenic bacteria. Front. Cell Infect. Microbiol 2013 3 59 24106689 131 Tullius MV Harmston CA Owens CP Chim N Morse RP Discovery and characterization of a unique mycobacterial heme acquisition system. PNAS 2011 108 5051 56 21383189 132 Turner AG Ong CL Gillen CM Davies MR West NP Manganese homeostasis in group A Streptococcus is critical for resistance to oxidative stress and virulence. mBio 2015 6 e00278 15 25805729 133 van Valen L A new evolutionary law. Evol. Theory 1973 1 1 30 134 Veyrier FJ Boneca IG Cellier MF Taha MK A novel metal transporter mediating manganese export (MntX) regulates the Mn to Fe intracellular ratio and Neisseria meningitidis virulence. PLOS Pathog 2011 7 e1002261 21980287 135 Vidal SM Pinner E Lepage P Gauthier S Gros P Natural resistance to intracellular infections: Nramp1 encodes a membrane phosphoglycoprotein absent in macrophages from susceptible (Nramp1 D169) mouse strains. J. Immunol 1996 157 3559 68 8871656 136 Vogel HJ Lactoferrin, a bird's eye view. Biochem. Cell Biol 2012 90 233 44 22540735 137 Wakeman CA Moore JL Noto MJ Zhang Y Singleton MD Prentice BM The innate immune protein calprotectin promotes Pseudomonas aeruginosa and Staphylococcus aureus interaction. Nat. Commun 2016 7 11951 Zinc sequestration by calprotectin promotes co-infection by Pseudomonas aeruginosa and Staphylococcus aureus.. 27301800 138 Wanachiwanawin W Infections in E-β thalassemia. J. Pediatr. Hematol. Oncol 2000 22 581 87 11132234 139 Wang T Si M Song Y Zhu W Gao F Type VI secretion system transports Zn2+ to combat multiple stresses and host immunity. PLOS Pathog 2015 11 e1005020 26134274 140 Weinberg ED Nutritional immunity: host's attempt to withhold iron from microbial invaders. JAMA 1975 231 39 41 1243565 141 Weinberg ED Iron availability and infection. Biochim. Biophys. Acta 2009 1790 600 5 18675317 142 Wessling-Resnick M Nramp1 and other transporters involved in metal withholding during infection. J. Biol. Chem 2015 290 18984 90 26055722 143 White C Lee J Kambe T Fritsche K Petris MJ A role for the ATP7A copper-transporting ATPase in macrophage bactericidal activity. J. Biol. Chem 2009 284 33949 56 19808669 144 Wolff N Izadi-Pruneyre N Couprie J Habeck M Linge J Comparative analysis of structural and dynamic properties of the loaded and unloaded hemophore HasA: functional implications. J. Mol. Biol 2008 376 517 25 18164722 145 Wolschendorf F Ackart D Shrestha TB Hascall-Dove L Nolan S Copper resistance is essential for virulence of Mycobacterium tuberculosis. PNAS 2011 108 1621 26 21205886 146 Wright AC Simpson LM Oliver JD Role of iron in the pathogenesis of Vibrio vulnificus infections. Infect. Immun 1981 34 503 7 7309236 147 Zaia AA Sappington KJ Nisapakultorn K Chazin WJ Dietrich EA Subversion of antimicrobial calprotectin (S100A8/S100A9 complex) in the cytoplasm of TR146 epithelial cells after invasion by Listeria monocytogenes. Mucosal Immunol 2009 2 43 53 19079333 148 Zarantonelli ML Szatanik M Giorgini D Hong E Huerre M Transgenic mice expressing human transferrin as a model for meningococcal infection. Infect. Immun 2007 75 5609 14 17893132 149 Zhang YJ Rubin EJ Feast or famine: the host-pathogen battle over amino acids. Cell Microbiol 2013 15 1079 87 23521858 150 Zimbler DL Park TM Arivett BA Penwell WF Greer SM Stress response and virulence functions of the Acinetobacter baumannii NfuA Fe-S scaffold protein. J. Bacteriol 2012 194 2884 93 22467784 RELATED RESOURCES 1 Nriagu JO Skaar EP Trace Metals and Infectious Diseases 2015 MIT Press Cambridge, MA 2 Skaar EP Raffatellu M Metals in infectious diseases and nutritional immunity. Metallomics 2015 7 926 28 26017093

Subsets and Splits

No community queries yet

The top public SQL queries from the community will appear here once available.