sumo化修饰对胰岛β细胞凋亡的调控及机制研究

《sumo化修饰对胰岛β细胞凋亡的调控及机制研究》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

分类号学号D201078421学校代码10487密级博士学位论文SUMO化修饰对胰岛β细胞凋亡的调控及机制研究学位申请人：赖巧红学科专业：免疫学指导教师：王从义教授答辩日期：2013年5月 ADissertationSubmittedtoHuazhongUniversityofScienceandTechnologyfortheDegreeofDoctorSumoylationRegulatesPancreaticβCellApoptosisandtheUnderlyingMechanismsDoctorCandidate:QiaohongLaiMajor:ImmunologySupervisor:Prof.Cong-YiWangTheCenterforBiomedicalResearch,TongjiHospital,TongjiMedicalCollege,HuazhongUniversityofScienceandTechnology,1095JiefangAve,Wuhan430030,ChinaMay,2013 独创性声明本人郑重声明，本学位论文是本人在导师指导下进行的研究工作及取得的研究成果的总结。尽我所知，除文中已经标明引用的内容外，本论文不包含任何其他个人或集体已经发表或撰写过的研究成果。对本文的研究做出贡献的个人和集体，均已在文中以明确方式标明。本人完全意识到本人将承担本声明引起的一切法律后果。学位论文作者签名：日期：年月日学位论文版权使用授权书本学位论文作者完全了解学校有关保留、使用学位论文的规定，即：学校有权保留并向国家有关部门或机构送交论文的复印件和电子版，允许论文被查阅和借阅。本人授权华中科技大学可以将本学位论文的全部或部分内容编入有关数据库进行检索，可以采用影印、缩印或扫描等复制手段保存和汇编本学位论文。保密□，在_____年解密后适用本授权书。本论文属于不保密□√。（请在以上方框内打“√”）学位论文作者签名：指导教师签名：日期：年月日日期：年月日 目录主要缩略词表................................................................................................................1中文摘要........................................................................................................................3英文摘要........................................................................................................................9前言........................................................................................................................19第一部分UBC9信号缺失引发β细胞凋亡导致糖尿病的发生..........................21实验材料.........................................................................................................21实验方法.........................................................................................................25实验结果.........................................................................................................35小结.................................................................................................................51第二部分Ubc9信号缺失引发β细胞凋亡导致糖尿病的机制研究......................52实验材料.........................................................................................................52实验方法.........................................................................................................56实验结果.........................................................................................................59小结...............................................................................................................63参考文献................................................................................................................66综述............................................................................................................................71已发表文章..................................................................................................................87致谢..........................................................................................................................107 华中科技大学博士学位论文主要缩略词表英文缩写英文全称中文名称DNADeoxyribonucleicacid脱氧核糖核酸DMEMDulbecco'smodifiedEagle'smediumDulbecco改良的Eagle培养基DMSODimethylsulfoxide二甲基亚砜dldecilitre分升DAPIdiamidino-phenyl-indol二脒基苯基吲哚ELISAEnzyme-linkedimmunosorbentAssay酶联免疫吸附试验ECLEnhancedchemiluminescence增强的化学发光GTTGlucosetolerancetest葡萄糖耐受实验ggram克GSISglucose-stimulatedinsulinsecretion葡萄糖刺激的胰岛素分泌HEHematoxylin-eosin苏木素-伊红HEstaininghematoxylinandeosinstaining伊红－苏木素染色HRPHorseradishperoxidase辣根过氧化物酶HBSSHank'sBalancedSaltSolutionHank’s平衡盐溶液ITTinsulintolerancetest胰岛素耐受实验IFN-γInterferon-gamma干扰素-γIL-1βInterleukin-1beta白细胞介素-1βkgkilogram千克mgMilligram毫克minMinute分钟mlMillilitre毫升mRNAmessengerRNA信使RNAmmmillimeter毫米mMmilimol毫摩尔nicotinamideadeninedinucleotide还原型烟酰胺腺嘌呤NADPHphosphate二核苷酸磷酸1 华中科技大学博士学位论文nmnanometer纳米ngnanogram纳克ODOpticaldensity吸光度PBSPhosphatebufferedsaline磷酸盐缓冲液PCRPolymerasechainreaction聚合酶链反应PVDFPolyvinylidenefluoride聚偏二氟乙烯ROSReactiveoxygenspecies活性氧分子RPMIRoswellParkMemorialInstitutemedium罗斯维尔公园纪念研究所培养基RpmRevolution(s)PerMinute每分钟转数SDS-PAGESDS-polyacrylamidegelelectrophoresisSDS-聚丙烯酰胺凝胶电泳SUMOsmallubiquitin-likemodifier小泛素样调节物SPFSpecificPathogenFree无特定病原体STZStreptozotocin链尿佐霉素TNF-αTumornecrosisfactor-α肿瘤坏死因子αTerminaldeoxynucleotidyltransferase末端脱氧核苷酸TUNELdUTPnickendlabeling脱氧尿苷三磷酸切口末端标记TDT酶TerminalDeoxynucleotidy1Transferase末端转移酶μmmicrometre微米μlmillilitre微升2 华中科技大学博士学位论文中文摘要SUMO化修饰对β细胞凋亡的调控及机制研究博士研究生：赖巧红导师：王从义教授华中科技大学同济医学院附属同济医院生物医学研究中心一、研究背景近年来，流行病学调查显示，世界范围内的糖尿病发病率由2%上升到5%。特别是在一些发展中国家，例如中国，经济的迅速发展伴随着生活模式和生活环境的变化，使糖尿病的发病率每年增长3%。1型糖尿病（type1diabetes,T1D）又名胰岛素依赖型糖尿病(IDDM)或青少年糖尿病，是儿童与青少年常见的内分泌疾病。1型糖尿病是易感个体在环境因素的诱导下，由Th1细胞介导，以免疫性胰岛炎和胰岛β细胞受损为主要特征的自身免疫性疾病。2型糖尿病主要发生于成人，是由于体内高糖高脂毒性对β细胞的长期损害所致，表现为胰岛素抵抗和肥胖。外源性胰岛素治疗不能像自身分泌的内源性胰岛素那样精确地调节血糖水平，所以，糖尿病患者常伴有多种并发症：糖尿病肾病、酮症酸中毒、非酮症性高渗性昏迷、糖尿病性心脏病、糖尿病性脑血管病变、糖尿病性肢端坏疽、糖尿病性神经病变等。并且，糖尿病病情迁延，难以根除，需长期用药，导致家庭和社会负担沉重。我国用于治疗糖尿病的医疗支出约占医疗卫生总费用4%～5%。1型和2型糖尿病的发病机制至今尚未明确，但大量的研究证实氧化应激（Oxidativestress）和内质网应激（ERstress）在β细胞损坏导致糖尿病过程中发挥重要作用。氧化应激是导致β细胞损伤、凋亡的主要病理基础。在正常状态下机体可产生少量活性氧自由基（reactiveoxygenspecies,ROS），同时存在自由基清除体系，使自由基的产生和清除保持平衡。少量活性氧自由基参与代谢，如细胞内的信号转导,调控细胞的生长、分裂、分化、迁移、凋亡及衰老等生理活动，但过多活性氧自由基可导致组织、细胞发生病理改变，造成的机体伤害。胰腺β细胞抗氧化防御能力低下，细胞内自由基清除酶（Cu/Zn超氧化3 华中科技大学博士学位论文物歧化酶、Mn超氧化物歧化酶、过氧化氢酶和谷胱甘肽过氧化物酶）和ROS清除蛋白（硫氧化还原蛋白）水平较低，使较低的氧自由基浓度即可使胰腺β细胞受损。内质网应激是导致β细胞损伤、凋亡的另一个重要机制。β细胞作为内分泌细胞，拥有高度发达的内质网系统，同时也使β细胞对内质网应激高度敏感。在氧化应激等情况下，内质网内未折叠蛋白增多，发生未折叠蛋白反应（Unfoldproteinresponse,UPR），该反应首先激活膜上的3个信号蛋白：PERK（PKR-likeERkinase,双链RNA依赖的蛋白激酶样ER激酶），IRE-1（inositolrequiringenzyme1，跨膜蛋白激酶1）和ATF6（Activatingtranscriptionfactor6,活化转录因子6），通过其下游信号通路抑制蛋白质的合成，上调分子伴侣和折叠蛋白的表达，加速未折叠蛋白的降解等多种方式缓解内质网应激状态。当UPR超过细胞的承受能力，使细胞不能重新回到稳态时，则会启动细胞凋亡程序。通过CHOP通路、IRE1-2+TRAF2-ASK1-caspase12通路和Ca通路，使细胞发生凋亡。氧化应激和内质网应激所致β细胞损伤是引发及推动1型和2型糖尿病的重要环节，两者分别/共同作用，使β细胞失去分泌胰岛素的功能。本课题组的前期及其它课题组研究结果显示，在氧化应激和内质网应激状态下，SUMO高表达，多种抵抗应激的蛋白被SUMO化，显示蛋白SUMO化是一种重要的转录后抵抗应激的调节机制。SUMO化修饰过程与泛素化类似，包含E1活化酶,E2结合酶以及E3连接酶三个酶的级联反应。E1活化酶是一种异源二聚体,在哺乳动物中为SAE1/SAE2,E2结合酶为UBC9，UBC9是SUMO化修饰中唯一的E2结合酶，UBC9的表达变化直接影响到SUMO化功能，因此，UBC9是研究β细胞SUMO化功能的良好靶点。UBC9基因缺乏会导致小鼠胚胎期死亡。E3酶主要包括三类:PIAS家族、核孔蛋白RanBP2/Nup358和Pc2。通过三种酶的级联反应SUMO分子被转移至底物蛋白。本研究以UBC9为靶点，以诱导型条件性UBC9基因敲除小鼠，在成年小鼠β细胞内特异性地敲除UBC9基因作为模型，研究SUMO化修饰对β细胞功能的影响，进行了以下工作：4 华中科技大学博士学位论文二、诱导型UBC9基因敲除小鼠模型的建立fl/fl本研究用ubc9小鼠和他莫昔芬诱导型β细胞特异性RipCreER转基因小+fl/fl+fl/fl鼠进行杂交，建立RipCreERubc9小鼠。对8周龄RipCreERubc9小鼠腹腔注射他莫昔芬（50mg/kg/天），每天一次，连续5天，以注射玉米油溶剂（含10%乙醇）的同窝同性别小鼠作为对照，诱导成年小鼠的β细胞UBC9基因敲除。注射完成5天后处死小鼠，由胆总管向胰腺注入胶原酶P，消化挑拣胰岛。提取胰岛的DNA和蛋白，进行PCR和Westernblot分析胰岛的基因组型别及UBC9的表达，验证基因敲除效果。结果显示：对照组小鼠胰岛UBC9基因完整，表达UBC9蛋白高，而他莫昔芬组诱导组胰岛中染色体UBC9基因部分缺失，未检出UBC9蛋白表+fl/fl达，表明他莫昔芬诱导后，RipCreERubc9小鼠UBC9表达缺失；并且雌性小鼠的诱导完全缺失的比率为50%，雄性小鼠的诱导完全缺失的比率为75%。三、UBC9缺失导致小鼠早期发生糖耐量异常，晚期罹患1型糖尿病1、UBC9基因敲除小鼠早期发生糖耐量异常（1）UBC9基因敲除小鼠糖耐量降低：他莫昔芬或玉米油注射完毕后第9天（D9）行葡萄糖耐量实验，实验前小鼠空腹16小时，腹腔注射2g/kg葡萄糖溶液后于0min、30min、60min、90min、120min和180min采集小鼠尾部血样检测各时间点的血糖值。结果显示：①他莫昔芬组小鼠空腹血糖值较对照油组明显降低；②对照组小鼠血糖30min升到峰值，60min迅速下降，90min基本恢复到正常，而他莫昔芬组小鼠血糖30min峰值明显升高，60-120min血糖缓慢下降，180min时尚未恢复正常。（2）UBC9基因敲除小鼠胰岛素耐受正常：胰岛素的分泌减少与胰岛素抵抗均可造成糖耐量下降，为了进一步明确他莫昔芬组引起糖耐量下降的原因，我们比较了两组小鼠的胰岛素耐受情况。腹腔注射0.75U/kg胰岛素，于注射后0min、15min、30min、45min和60min检测血糖。结果显示：两组小鼠0min的随机血糖值无差异，30min时血糖均降到最低点，60min时血糖基本恢复正常，各时间点组间差异不显著。以上两个实验表明：胰岛β细胞特异性UBC9基因缺失导致β细胞分泌胰5 华中科技大学博士学位论文岛素的功能明显下降，造成糖耐量异常。2、UBC9基因敲除小鼠晚期血糖自发升高，血清胰岛素水平下降，出现自发糖尿病（1）UBC9基因敲除小鼠平均血糖值升高：监测他莫昔芬组和对照组小鼠的随机血糖值，一周2次。D7到D42天两组小鼠平均血糖持续在正常水平，他莫昔芬组血糖略高于对照组，但组间差异不显著；D42天之后，他莫昔芬组血糖值持续上升，D70天达478.6±12.4mg/dl，较对照组166.3±6.0mg/dl明显升高。（2）UBC9基因敲除小鼠血清胰岛素水平下降：采集小鼠心脏血液，ELISA检测血清胰岛水平，与对照组相比，他莫昔芬组小鼠血清胰岛素水平在D45天之后明显下降。（3）UBC9基因敲除小鼠糖尿病晚期发病率100%：他莫昔芬组小鼠在D43天开始出现血糖异常，D55天有66.7%小鼠血糖达到糖尿病标准，D65天所有小鼠均出现血糖异常和糖尿病相关症状，而对照组小鼠无1例发病。以上结果显示：胰岛β细胞特异性UBC9缺失可导致胰岛素分泌水平下降，造成血糖升高，长时间血糖升高在晚期造成小鼠罹患糖尿病。四、UBC9缺失导致胰岛形态和功能异常，β细胞发生凋亡1、胰腺组织内胰岛团变小，数目减少：小鼠胰腺石蜡切片HE染色示：对照组胰腺组织胰岛团大而圆，胞浆胞核着色均匀；他莫昔芬组胰岛团随着UBC9敲除时间的后移逐渐变小，并出现凋亡小体。胰岛素免疫组化染色和免疫荧光染色示：对照组胰岛素阳性染色的β细胞团大、圆，占整个切片胰腺组织面积的2%左右；他莫昔芬组胰岛素阳性染色细胞团逐渐减小，D45天仅占切片胰腺组织面积的0.9%；D75天减小到0.2%以下。结果显示：胰岛β细胞特异性UBC9缺失导致胰岛β细胞形态异常，数量减少。2、胰岛细胞功能异常：胶原酶P消化获得胰岛，体外培养过夜，进行葡萄糖刺激的胰岛素分泌（GSIS）实验，检测胰岛在低糖（3.3mM）和高糖（16.7mM）刺激下胰岛素的释放量。他莫昔芬组胰岛分泌能力明显下降。3、胰岛内β细胞发生凋亡：TUNEL和cleavedcaspase3染色检测显示他莫昔芬组D45天左右胰岛内2.3%的β细胞发生凋亡，对照组内仅见0.2%的凋亡细6 华中科技大学博士学位论文胞，组间差异显著。结果提示：胰岛β细胞特异性UBC9缺失使β细胞更容易发生凋亡。五、UBC9缺失小鼠对STZ诱导糖尿病的耐受力降低他莫昔芬和对照组小鼠腹腔内连续5天注射低剂量STZ（40mg/kg），两组小鼠随机血糖值均缓慢上升，他莫昔芬组血糖明显高于对照组，上升到近600mg/dl，而对照组血糖仅上升到近400mg/dl。结果提示：胰岛β细胞特异性UBC9缺失的β细胞对STZ耐受力明显降低。六、UBC9缺失导致β细胞凋亡的机制1、SUMO-NADPH氧化酶-ROS途径对β细胞凋亡的影响（1）UBC9基因缺失早期β细胞内ROS升高：UBC9基因敲除后D10天，胶原酶消化分离胰岛，培养过夜，细胞因子（IL-1β、TNF-α和IFN-γ）混合刺激6小时，25μM羧基-H2DCFDA中孵育30min，他莫昔芬组胰岛内ROS荧光强度明显高于对照组。结果说明：胰岛β细胞特异性UBC9缺陷在早期可导致β细胞内ROS表达增强。（2）UBC9基因缺失后期β细胞内ROS降低：UBC9基因敲除后D30-D45天，胰岛培养过夜，细胞因子刺激6小时，H2DCFDA中孵育30min，他莫昔芬组胰岛内ROS荧光强度明显低于对照组。结果说明：胰岛β细胞特异性UBC9缺陷导致β细胞大量损伤、凋亡，在诱导后期胰岛内ROS表达减弱。2、Ubc9基因缺陷导致β细胞内质网应激我们检测了β细胞内ERstress相关蛋白Grp78、Grp58、eIf2a、ATF6、IRE1a和CHOP等蛋白的表达，结果显示：ERstress可能介导了β细胞凋亡。七、结论1、UBC9基因缺失早期小鼠发生糖耐量异常，后期血糖升高，血清胰岛素水平下降，基因敲除小鼠全部自发糖尿病；2、UBC9基因缺陷导致小鼠胰岛内β细胞凋亡，引发胰岛形态和功能异7 华中科技大学博士学位论文常；3、UBC9基因缺陷小鼠对STZ诱导糖尿病的耐受力降低；4、UBC9基因缺陷小鼠β细胞内ROS和ERstress的异常表达是导致β细胞损伤、凋亡的重要原因。综上所述，SUMO化修饰对胰岛内β细胞的功能和活性起保护作用，该机制可能是通过SUMO1抑制NADPH氧化酶产生的ROS发挥的，同时，ERstress参与其中。去除SUMO化修饰，将导致β细胞凋亡，发生糖尿病。因此，SUMO化修饰的唯一E2结合酶UBC9可能成为临床治疗糖尿病的干预靶点。关键词：UBC9;ROS；ERstress;β细胞;糖尿病8 华中科技大学博士学位论文AbstractSumoylationRegulatesPancreaticbetaCellApoptosisandtheUnderlyingMechnismsPh.D.Candidate:QiaohongLaiAdvisor:ProfessorCong-YiWangTheCenterforBiomedicalResearch,TongjiHospital,TongjiMedicalCollege,HuazhongUniversityofScienceandTechnology,1095JiefangAve,Wuhan430030,China1ResearchbackgroundRecentepidemiologicstudiesrevealedthattheincidencefordiabetesinmostregionsofworldwidehasbeenincreasingby2%to5%.Particularly,insomedevelopingcountriessuchasinChina,therapideconomicdevelopmentalongwithchangesinlifestyleandpresumably,thelivingenvironmenthaverenderedChinawithanannualincreaseof3%fordiabetesprevalence.Fortype1diabetes(T1D,oncealsoknownasjuvenilediabetesorinsulin-dependentdiabetesmellitus),thediseaseismanisfestedbytheautoimmuneresponsesresultedfromthebreakdownofperipheraltolerance.Similarasotherautoimmunediseases,acharacteristicfeatureforT1Distheselectivetargetingofaspecifictypeofcells,theinsulinsecretingβcellsoftheisletsofLangerhansinthepancreas,byacertainpopulationofautoreactiveimmunecells.Incontrast,type2diabetesusuallyoccursinadultsandischacaterizedbyinsulinresistanceassociatedwithobesity,whicheventuallyleadscompletelossofβmassduetoglucotoxicityandlipotoxicity.GiventhatT1Distypicallydevelopedinchildrenandjuveniles,itsimpactonqualityoflifeisfarmoresignificantthanthatof9 华中科技大学博士学位论文type2diabetes.Althoughexogenousinsulintherapypartlycompensatesthefunctionofβcells,itcannotregulatebloodglucoseasaccuratelyastheactionofendogenousinsulin.Asaresult,long-termimproperlycontrolofbloodglucosehomeostasispredisposesdiabeticpatientstothedevelopmentofdiversecomplicationssuchasdiabeticretinopathy,nephropathy,neuropathy,footulcers,andcardiovasculardiseases.Moreover,duotoitslongtreatmentcycle,thepatientsarenecessaryforlifelongmedication,whichconstitutesasignificantburdenforbothfamiliesandsociety.Indeed,ithasbeenestimatdthattheannualmedicalexpensesfordiabeticpatientsinChinanowisaround4to5%ofthetotalhealthcosts.AlthoughtheunderlyingmechanismsleadingtobothT1DandT2Dhaveyettobefullyaddressed,extensivestudies,however,haveconsistentlydemonstratedthatbothoxidativestressandendoplasmicreticulum(ER)stressplayacriticalroleinβcelldestructionduringthecourseofdiabetest.Neverlethess,theexactregulatorymechanismsunderlyingβcelldestructionarelargelyremainedelusive.Intype1diabetes,autoimmuneresponsesagainstbetacellselfantigensleadtotheproductionofcopiousamountsofinflammatorycytokines,whichtheninduceoxidativestresstomediatebetacelldestruction.Indeed,animalswithbetacellspecificexpressionofantioxidantenzymesareresistanttodevelopingT1D,andbetacellsdeficientinNox2,themainintracellularROSproducer,areprotectedfromcytokine-oralloxan-inducedapoptosis.Assuch,oxidativestressresultingfromautoimmuneresponsesisbelievedtobetheleadingcauseforbetamasslossinT1Dpathogenesis.UnlikeT1D,betamasslossinT2Disslowermanifestedbyaperiodofbetacelldysfunction.SustainedexposureofisolatedisletstohighglucoseinducesincreasesintracellularROS,whichrendersbetacellsundergoingapoptosis.Ithasnowbecomeevidentthatoxidativestresscausedbychronicexposuretoelevatedglucoseorfattyacidcontributestobetacelldeathintype2diabetes.Collectively,betacelldeathcausedbyoxidativestressislikelythemajormechanismforbetamasslossforbothtype1andtype2diabetes.10 华中科技大学博士学位论文Underphysiologicalcondition,lowlevelsofreactiveoxygenspecies(ROS)arenecessaryforthemaintenanceofnormalmetabolismsuchasintracellularsignaltransduction,regulationofcellgrowth,division,differentiation,migration,apoptosisandsenescence.Ingeneral,thesuppressionsystemsareresponsiblefortheclearanceoffreeradicals,sotheproductionandclearanceoffreeradicalsmaintainabalance.However,undercertainpathologicalstates,excessiveproductionofsuperoxide(O2-),hydrogenperoxide(H2O2),hydroxylradical(·OH)andreactivenitrogenspecies(RNS）suchasNO,NO2,ONOO·willbreakthebalancebetweenoxidationandclearance,leadingtotissueinjury.Particularly,pancreaticbetacellspossesspoorantioxidantcapacityalongwithlowlevelofintracellularfreeradicalscavengingenzymessuchasCu/Znsuperoxidedismutase,Mnsuperoxidedismutase,andcatalaseandglutathioneperoxidase,andtherefore,theyareparticularlyvulnerabletooxidativestress.ERstressisalsoanotherimportantmechanismforbeta-celldamageandapoptosis.Asendocrinecells,betacellshaveahighlydevelopedendoplasmicreticulumsystemandbetacellsarehighlysensitivetoendoplasmicreticulumstress.Incircumstancessuchasoxidativestress,unfoldproteinincreasesintheendoplasmicreticulum,unfoldproteinresponse(UPR)occurswhichthenactivates3signalingproteins,thePKR-likeERkinase(PERK),theinositolrequiringenzyme1(IRE-1）andtheactivatingtranscriptionfactor6(ATF6),andbywhichtoreduceproteinsynthesis,increaseproductionofmolecularmateandexpressionoffoldprotein,acceleratethedegradationofunfoldproteinsandeasetheendoplasmicreticulumstress.ApoptosisofcellsbyCHOPpathway,IRE1-TRAF2-ASK1-caspase12pathwayandCa2+pathwaywouldinitiateoncetheUPRisnotstrongenoughtorebuildERhomeostasis.Thepancreaticβcellsareequippedwithhighlydevelopedendoplasmicreticulum(ER)tofulfilltherequirementforsecretinglargeamountofinsulin.ThisphysiologicalfeaturerendersβcellsparticularlyvulnerabletoERstress.Exhaustionofβcellsis11 华中科技大学博士学位论文essentialfortheonsetofdiabetes,whichrequirestheresidualβcellsforcompensatedinsulinsecretion.Whilethiscompensatedactionisbeneficialforcontrolofbloodglucosehomeostasis,italsoincreasesERburdenassociatedwiththeinductionofunfoldedproteinresponse(UPR)andERstress,whichfurtherexacerbatesβcelldeath.AlthoughtheimplicationofERstressinβcelldeathhasbeenextensivelyemphasized,theunderlyingmechanisms,however,areyettobefullyelucidated.Assuch,understandingtheroleofERstressinthelossofβmassanddissectingthemechanismsunderlyingERstresswouldbeimportantfordevelopingtherapeuticapproachesaimedatpreventionandinterventionofdiabetes.Post-translationalattachmentofasmallubiquitin-likemodifier(SUMO)tothelysine(K)residue(s)ofatargetprotein(definedassumoylation)isanevolutionarilyconservedregulatorymechanism.FourSUMOproteinshavebeenidentifiedinhumans.Tobefunctionallyactive,SUMOneedstobehydrolyzedbyaSUMO-specificprotease(SENP)toexposeitsC-terminaldiglycinemotif,aprerequisiteforitscovalentconjugationtothesubstrateproteins.ThesumoylationprocessinvolvesaSUMO-activatingenzyme(E1,Uba2/Aos1),asingleSUMO-conjugatingenzyme(E2,Ubc9),andaSUMO-E3ligase(suchasthePIASfamilyorRanBP2).Sumoylationisareversibleprocessand,insomecases,dynamiccyclesofsumoylation/desumoylationoftargetproteinsarerequiredforcellularprocesses.Overthepastfewyears,sumoylationisnotonlyfoundtobeanimportantregulatorofthenormalfunctionofmanyvitalcellularproteins,butalsofoundtobeamajorplayerinthepathogenesisofhumandiseasessuchasHuntington’sdisease,Parkinson'sdisease,Alzheimer'sdisease,cancerandcardiovasculardiseases.Particularly,studiesincludingourshaveconsistentlydemonstratedthatsumoylationregulatesthecapacityofcellsagainstoxidativestress.OurrecentstudiesfurtherrevealedthatsumoylationregulatescellviabilitythroughrepressingintracellularROSgeneration.Whiletheseresultsareimportantandexciting,themechanismsbywhichsumoylationregulatesoxidaseactivity/ROSgeneration,andtherelatedimpacton12 华中科技大学博士学位论文betacells,aROSvulnerabletarget,areyettobeelucidated.2EstablishmentofinduciableUBC9geneknockoutmousemodelUbc9istheonlySUMOconjugatingenzymeessentialforsumoylation,andalterationsinUbc9expressiondirectlyreflectthecapacityofsumoylationfunction.Therefore,Ubc9isanidealtargetfordynamicmodulationofβ-cellsumoylationfunction.AslossofUbc9leadstoembryoniclethality,weusedthestrategytogenerateaconditionalUbc9knockoutmodel.Theknock-inallelecontainsaneo-FRTandtwoloxPsitesflankingUbc9exons2-4,whichcanbeidentifiedbyprobe5'andprobe3'.SevenchimericmicewereobtainedintheB6background.Theneo-FRTsiteinthechimericmicewasthendeletedbycrossingwithFlptransgenicmice,andwefloxhavenowobtainedtheUbc9micewithgermlinetransmissions.TheRip-CreERmicewereobtainedfromDr.DouglasMelton(HarvardUniversity),inwhichrecombinantCrecanbeinducedspecificallyinbetacellsbytamoxifen.Wefloxfl/flthencrossedUbc9-withRip-CreERtogenerateCre-Ubc9mice,andthemiceweretheninducedforbetacellspecificUbc9deficiencyat8wkoldwithtamoxifen(i.p.,50mg/kgbodyweight)for5consecutivedays.Littermatesinjectedwithequalvolumesofcornoil(controlvehicle)wereusedascontrols.Toisolatepancreaticislets,themiceweresacrificed5daysafterthelastinjection.Inbrief,thepancreaswasinflatedviathebileductwithadigestionbuffercontaining0collagenaseP.Thedistendedpancreaswasremovedanddigestedat37Cfor20to30-minwithshaking.Isletswerethenhand-pickedunderadissectingmicroscope.PCRandWesternblotanalysisofisletlysatesafterday2oflastinductionrevealedasignificantreductionforUbc9ascomparedwiththatofcornoilinducedmice,confirmingthatUbc9deficiencywasonlylimitedtobetacells.3Ubc9deficiencyleadstoabnormalglucosetoleranceanddiabetes13 华中科技大学博士学位论文onset3.1EarlyoccurrenceofimpairedglucosetoleranceinUbc9knockoutmice3.1.1ImpairedglucosetoleranceinUbc9knockoutmiceGlucosetoleranceteststartedday9(D9)afterTamoxifenorcornoilinjectionandmicewerefastedfor16hoursbeforetheexperiment.Bloodglucoselevelswererecordedafter0min,30min,60min,90min,120minand180minafterintraperitonealinjectionwithglucosesolution(2g/kg).ItwasfoundthatfastingbloodglucoselevelsofTamoxifenmiceweresignificantlylowerthanthatofthecontroloilgroup;②bloodglucoselevelreachesthehighestpeakat30minincontrolgroup,butrapidlydeclinedwithin60minandreturnedtonormallevelaround90min.Incontrat,bloodglucoselevelsignificantlyincreasedat30minintamoxifengroup,andmotimportantly,thebloodglucoselevelfailedtodeclineasthatofcontrolmice,after180minofchallenge,thebloodglucoselevelstillfailedtoreturntonormallevel.3.1.2Ubc9deficiencydoesnotresultininsulinresistanceInsulinresistanceoftwogroupsofmicewasthencomparedtofurtherclarifywhetherthefailureofglucosetoleranceintamoxifengroupwasduetolessinsulinsecretionorinsulinresistanceinvivo.Afterintraperitonealinjectionof0.75U/kginsulin0min,15min,30min,45minand60min,bloodglucoselevelswererecorded.Itwasnotedthattherewasnodifferenceinrandombloodglucoselevelsinbothgroupsofmice,bloodsugardroppedtothelowestpointat30minandreturnedtonormallevelat60min,andnosignificantdifferencesweredetectedateachtimepointexamined.Together,theseresultssuggestthatlossofUbc9impairsinsulinsecretioninbetacells.3.2LossofUbc9rendersmicetothedevelopmentofspontaneousdiabetes14 华中科技大学博士学位论文3.2.1LossofUbc9resultedintheincreaseofbloodglucoselevelsRandombloodglucosevaluesoftamoxifenandcontrolgroupofmiceweremonitoredeveryotherdayfor80days.Wefailedtoobserveasignificantdifferencefortheaveragebloodglucoselevelsbetweentwogroupsofmiceinthefirst30daysofUbc9deficiency.However,aprogressiveincreaseforbloodglucoseleveinTamoxifengroupwasnotedascomparedwiththatofcontrolgroup,andthemicestartedtoshowonsetofdiabetesaroundday43,allmicedevelopedspontaneousdiabetesonday65(478.6±12.4mg/dlinTamoxifengroupvs.166.3±6.0mg/dlincontrolgroup).3.2.2SeruminsulinlevelsdeclinedinUBC9geneknockoutmiceTofurtherdemonstratethattheincreaseofbloodglucoselevelsinUbc9deficiencymicewasduetothedecreaseofinsulinsecretion,weassayedseruminsulinlevelsusingtheheartbloodspecimensbyELISA.Inconsistentwithourexpectation,muchlowerlevelsofinsulinwerenotedinmiceinTamoxifengroupascomparedwiththatofmiceincontrolgroupafterday45ofTamoxifeninjection.3.2.3IncidenceofdiabetesinUbc9deficiencymiceWehavemonitoreddiabetesincidenceusing12Ubc9deficencymiceand12controlmice.OneUbc9deficientmicedevelopeddiabetesafterday43ofTamoxifeninduction;7miceshoweddiabetesonsetbetweendays50to55afterTamoxifeninduction,andallmicedevelopedspontaneousdiabetesonday65ofUbc9deficiency.Insharpcontrast,noenofthecontrolmicedevelopeddiabetes.4Ubc9playsanessentialroleforbetacellviability4.1Ubc9deficiencyaffectsthemorphology,numberandsizeoftheisletNext,histologicalanalysiswasconductedbyHEstainingofthepancreatic15 华中科技大学博士学位论文sections.Inlinewiththeaboveresults,micewithUbc9deficiencydiplayedshrinkedisletswithlossofbetamassascomparedwiththatofcontrolmice.InsulinstainingdemonstratedasignificantlylowernumberofinsulinproducingbetacellsinUbc9deficientsectionsascontrolsections.TUNElassayfurtherrevealedamassivebetacellapoptosisafterlossofUbc9.Tofurtherconfirmtheseresults,wedidcaspase3staininganddemonstratedthatthesecaspase3positivecellswerealsopositiveforinsulin,indicatingthatthesecaspase3positivecellswereinsulinproducingbetacells.4.2ImparedbetacellfunctioninUbc9deficientmicePancreaticisletswerehandpickedunderadissectionmicroscopeasdescribedearlieraftercollagenasePdigestion.Theisolatedisletswereculturedovernightinvitro,andthensubjectedtoglucose-stimulatedinsulinsecretion(GSIS)analysisbystimulationwithlowglucose(3.3mM)andhighglucose(16.7mM).Inconsistentwiththeaboveresults,thecapacityofpancreaticisletsforsecretionofinulinafterglucosechanllengewassignificantlyreducedinUbc9deficientmiceascomparedwiththatofcontrolmice.5LossofUbc9rendersmicewithhighersensitivitytostretozotocin(STZ)-inducedtoxicityToassesswhetherUbc9impactsSTZ-inducedcytoxicity,low-doseofSTZ(40mg/kg)wereadministeredviatailveinfor5consecutivedaysformiceinbothtamoxifengroupandcontrolgroup.Anincreaseforbloodglucosewasnotedformiceinbothgroups.However,theincreaseofbloodglucosewasmuchfasterandtheglucoselevelsweremuchhigherinmiceofTamoxifengroup(600mg/dl)ascomparedwiththatofcontrolmice,suggestingthatlossofUbc9enhancesthesensitivityofbetacellstoSTZtoxicity.ltshow:theSTZtoleranceofbetacellswithUBC9defectissignificantlyreduced.6Ubc9regulationofbetacellviabilityinvolvesROSaccumulationandER16 华中科技大学博士学位论文stress6.1LossofUbc9resultsinROSaccumulationinbetacells6.1.1EarlyUBC9genedeletionleadstoROSincreaseTodissectthemechanismsbywhichUbc9regulatesbetacellviability,wefirstexaminedintracellularROSaccumulation.Forthispurpose,pancreaticisletsweresiloatedfrommiceonday10afterTamoxifeninduction.Theisolatedisletswerefirstculturedovernight,thensubjectedtocytokinestimulation(IL-1β,TNF-αandIFN-gamma)for6hours.IntracellularROSaccumulationwasassessedbyincubationwith25mMcarboxy-H2DCFDAfor30min.Interestingly,ROSfluorescenceintensitywassignificantlyhigherinisletsfromtamoxifeninducedmicethanthatofcontrolmice,indicatingthatUbc9deficiencyleadsROSaccumulationinbetacells.6.1.2ROSreductioninlaterstageofUbc9deficiencymiceTofurtherdemonstratethatROSaccumulationbetacellspecific,isletsisolatedfrommiceafterdays30and45ofUbc9deficeincyweresubjectedtoROSaccumulationassay,inwhichtheamountoffunctionalbetacellsinUbc9deficientmiceshouldmuchlessascomparedwiththatofcontrolmiceduetoapoptosis.Indeed,ROSfluorescenceintensityinisletsfromtamoxifeninducedmicewassignificantlylowerthanthatofcontrolmice,demonstratingthatROSaccumulationisbetacellspecific.6.2Ubc9deficencyinducesERstressinbetacellsGiventheroleofERstressplayedinbetacellapoptosis,wenextexaminedtheERstressrelatedsignalingpathways.Forthispurpose,isletswereisolatedfrommiceafterday10ofTamoxifeninduction,andcelllysateswerepreparedandsubjectedtoWesternblotanalysisofGrp78、Grp58、eIf2a、ATF6、IRE1aandCHOPexpressions.ItwasnotedthtlossofUbc9significantlyenhancedIRE1aandEIf2aexpressionalongwithhigherlevelsofphosphorylatedIRE1aandEIf2a.Inconsistentwiththeseresults,muchhigherlevelsofGrp58andGrp78werenotedinUbc9deficientislets.17 华中科技大学博士学位论文Together,thesedatasupportthatlossofUbc9inducesERstress,whichthensynergizeswithROS-inducedoxidativestresstopromotebetacellsundergoingapoptosis.7Conclusions(1)LossofUbc9renderspancreaticbetacellsundergoingprogressiveapoptosisalongwiththedevelopmentofspontaneousdiabetes;(2)Ubc9deficiencyresultsinisletmorphologicalchangeandfunctionalabnormalityduetobetacelldeath;(3)MicedeficientinUbc9displayhighersensitivitytoSTZ-inducedcytoxicity;(4)Ubc9regulationofbetacellviabilityisassociatedwithmodulationofROSaccumulationandERstressinbetacells.Keywords:UBC9;ROS;ERstress;betacells;diabetes18 华中科技大学博士学位论文前言SUMO化修饰指由SUMO（smallubiquitin-likemodifier,小泛素样调节物）分子参与、对体内蛋白进行修饰并动态调节蛋白功能的生物学机制(1-2)。SUMO最早在酵母中被发现，其后证明普遍存在于各种真核生物（酵母、果蝇、线虫和脊椎动物），属高度保守的蛋白家族。其氨基酸序列与泛素同源性不高，但具有相同的蛋白折叠结构(3)(4)(5)(6)。人类基因组迄今已发现4个SUMO家族成员（SUMO1～4），分别位于2、17、21和6号染色体。其中，SUMO1、SUMO2、和SUMO3广泛存在于各种组织；SUMO4则局限存在于肾脏、胰腺和免疫组织，且其功能呈现组织或器官特异性。研究显示，SUMO4是灵长类动物进化过程中由SUMO1复制突变产生的(7-8)，在啮齿类动物（小鼠和大鼠）基因组中仅检出SUMO1、SUMO2、和SUMO3基因，未发现SUMO4基因，提示小鼠SUMO1是人类SUMO4的同源蛋白。SUMO化过程涉及SUMO活化酶（E1，Uba2-Aos1）、SUMO结合酶（E2，Ubc9）和SUMO连接酶（E3，PIAS家族蛋白和RanBP2）。SUMO与泛素分子相同，均以赖氨酸残基羧基端肽键与靶蛋白共价连接。然而，泛素化介导蛋白降解，而SUMO化参与靶蛋白转录后修饰。因此，SUMO化可能通过与泛素化竞争而稳定靶蛋白，避免后者被降解。另外，SUMO化具有可逆性，某些情况下，SUMO化与去SUMO化的动态变化可调节蛋白活性(9)。近来，大量文献显示SUMO化修饰在氧化应激（Oxidativestress）中有重要作用(10)。蛋白的SUMO化可阻断氧化应激所介导的细胞凋亡(8)。SUMO4可负调节细胞内氧化应激，清除细胞内ROS（Reactiveoxygenspecies）等(11-12)。同时，研究发现氧化应激是胰岛β细胞损伤、凋亡，糖尿病发生发展的主要病理基础（(13-15)。胰岛β细胞的抗氧化能力较差，对氧化应激敏感(16-19)，氧化应激主要通过以下机制损伤β细胞：（1）IL-1单独或联合IFNγ诱导β细胞内ROS和NO升高，通过死亡受体FAS和FASL上调中介β细胞凋亡(20)；（2）T细胞通过释放细胞毒性细胞因子如IL-1β、IFNγ和TNFα导致β细胞损伤(21-22)；（3）自身反应性T细胞通过激活FAS-caspase路径直接传递19 华中科技大学博士学位论文β细胞死亡信号，这也和ROS和NO相关(23-24)；（4）在动物体内，β细胞特异性表达抗氧化酶对1型糖尿病的发生有抵抗作用(25-27)；（5）缺乏NADPH氧化酶(ROS的主要来源)的β细胞明显减少对细胞毒性因子和alloxan诱导的凋亡(28)。本研究从唯一的SUMO化E2结合酶Ubc9入手，建立了胰腺β细胞特异性Ubc9基因敲除小鼠模型，期望能真实确切地揭示SUMO化修饰在糖尿病发生发展中的作用，并希望确立新的治疗糖尿病的有效靶标。20 华中科技大学博士学位论文第一部分UBC9信号缺失引发β细胞凋亡导致糖尿病的发生实验材料一、主要实验仪器5415R低温高速离心机德国Eppendorf公司Minispin台式高速离心机德国Eppendorf公司低温高速离心机，5415R德国Eppendorf公司－80℃超低温冰箱德国Nuaire公司－20℃低温冰箱德国Nuaire公司4℃低温冰箱中国Haier公司PCR仪德国Eppendorf公司倒置显微镜日本OlympusZEISS解剖显微镜德国ZEISS公司POLARstar全自动多功能酶标仪德国Omega公司CarlZEISS图像分析系统德国ZEISS公司血糖仪以色列SHIRAT公司血糖试纸以色列SHIRAT公司DK-8D型电热恒温水槽美国Labnet公司恒温水浴箱美国Labnet公司88－I型定时恒温磁力搅拌器美国Fisher公司（Thermix）CO2恒温培养箱德国Heraeus公司MS2型微型振荡器德国IKA公司WD-9405型水平摇床BioekelScientificUltrafree-MC10-kDa滤器Millipore公司DYY-8型稳压稳流电泳仪美国fisher公司，Bio-Rad公司水平电泳槽美国Bio－Rad公司21 华中科技大学博士学位论文小型垂直电泳槽美国Bio－Rad公司紫外分光光度仪德国Eppendorf公司酶标仪美国TECAN公司微量加样器德国Eppendorf公司电子分析天平瑞士Mettlertoledo公司超净工作台美国labconco公司转移电泳槽美国Bio－Rad公司二、主要实验材料动物：Rip-CreER转基因小鼠来自于Dr.DouglasMelton(HarvardUniversity)fl/f惠赠，Ubc9转基因小鼠（C57BL/6背景）自己构建并购买于JacksonLaboratory，均饲养于乔治亚瑞金斯大学(Georgiaregentsuniversity）SPF级实验动物中心。三、主要试剂DMEM培养基美国Gibco公司ECL化学发光试剂美国PIERCE公司HRP标记羊抗兔二抗美国Bio-Rad公司牛血清白蛋白美国fisher公司胎牛血清、小牛血清美国Gibco公司胰蛋白酶美国Sigma公司胶原酶P瑞士Roche公司D-葡萄糖美国Sigma公司PVDF膜美国Invitrogen生命技术公司RPMI1640美国Gibco公司二甲基亚砜（DMSO）德国Sigma公司低分子量标准蛋白分子量美国Bio-Rad公司台盼蓝粉（TypanBlue）德国Sigma公司无血清培养基（UltraCULTURETM）美国HyClone公司22 华中科技大学博士学位论文甘精胰岛素注射液（Lantus100U/ml）法国Sanofi-Aventis公司小鼠胰岛素ELISA检测试剂盒美国ALPCO公司链尿佐霉素（STZ）美国Sigma公司Hank’s平衡盐溶液（HBSS）美国Sigma公司山羊血清美国SantaCruz公司胰岛素一抗美国CellSignaling公司AlexaFluor488lgG荧光抗体美国JacksonImmunoResearch公司Cy2荧光抗体美国JacksonImmunoResearch公司RabbitABCstainingSystem(SC-2018)美国SantaCruz公司CleavedCaspase-3一抗美国CellSignaling公司ApoptagplusFluoresceininSite美国Chemiconinternational公司ApoptosisDetectionKit(S7111)蛋白酶K美国Sigma公司TMImage-iTLIVEGreenReactive美国Invitrogen生命技术公司OxygenSpeciesDetectionKit(I36007)含钙和镁的HBSS美国Gibco公司WizardSVGenomicDNAPurification美国Invitrogen生命技术公司System四、主要试剂配制1、SDS-PAGE和Westernblot试剂(1)配制30%丙烯酰胺：取29g丙烯酰胺、1gN，N’-亚甲双丙烯酰胺溶于60mlddH2O，37℃溶解，加ddH2O至100ml，过滤备用（pH≤7.0），室温避光保存。(2)1.5mol/LTris(pH8.8)：18.165gTris溶于80mlddH2O中，调pH至8.8，加双蒸水至100ml保存备用。(3)1mol/LTris(pH6.8)：12.11gTris溶于80mlddH2O中，调pH至6.8，加双蒸水至100ml保存备用。(4)10%过硫酸铵：0.1g过硫酸铵，加双蒸水至1ml（临用前新鲜配制）。(5)2×SDS凝胶加样缓冲液：6%β-ME、6%SDS、0.6%溴酚蓝、20%甘油、1223 华中科技大学博士学位论文ml10%SDS，加入4ml甘油，加入1.2mlβ-ME，加ddH2O至20ml，再加0.12g溴酚蓝备用。(6)5×Tris-甘氨酸电泳缓冲液：3.01gTris、18.8g电泳级甘氨酸（pH8.3）、10ml10%电泳级SDS溶于100ml双蒸水中，加双蒸水至200ml配成pH8.3，25mmol/LTris、250mmol/L甘氨酸、0.1%SDS。(7)考马斯亮蓝染液：甲醇250ml，去离子水200ml，冰乙酸50ml，溶于1.25g考马斯亮蓝R-250，滤纸过滤。(8)转移电泳缓冲液：加入Tris12.0g，甘氨酸57.65g，用去离子水定容至4000ml。(9)脱色液：甲醇250ml，去离子水200ml，冰乙酸50ml混合。(10)封闭液：5%脱脂奶粉，0.02%叠氮钠溶于PBS中备用。(11)漂洗液Ⅰ：配制0.01mol/LPBS（pH7.2）。(12)漂洗液Ⅱ:配制150mmol/LNaCl，50mmol/LTris-Cl（pH7.5）。(13)PBS-T:取2.5ml20%Tween-20加入1000mlPBS中。2、RIPA蛋白裂解液取1.5ml5MNaCl、0.5ml1MTris（pH7.2）、0.5ml10%SDS、0.5ml1%TritonX-100、0.5g1%Deoxycholate、0.5ml500mMEDTA，加水补至50ml3、4％多聚甲醛取多聚甲醛4g溶于100mlPBS，60℃水浴加速溶解。4、KRB储存液的配制：化学试剂F.W.[mM]g/LNaCL58.441378.0KCL74.554.70.35KH2PO4136.11.20.16MgSO4-7H2O246.481.20.3CaCL2-2H2O1472.50.37NaHCO384.01252.15、KRB工作液的配制：24 华中科技大学博士学位论文取100mlKRB储存液在37℃水浴中加热，加入0.05gBSA，10MmHEPES，调节PH值7.4。吸取25mlKRB储存液到另一50ml离心管中，加入250μl1.65M的葡萄糖液成为16.7mM葡萄糖工作液；余下的75mlKRB储存液加入150μl1.65M的葡萄糖液成为3.3mM葡萄糖工作液。6、STZ柠檬酸注射液的配制：柠檬酸2.1克加入双蒸水100ml中配成A液；柠檬酸钠2.94克加入双蒸水100ml中配成B液；用时将A液和B液按1:1.32混合（A液28ml，B液22ml），调节PH值为4.5，4℃保存；注射时配成1%STZ浓度。7、胶原酶P（Roche）的配制：D-Hank’s液中加入7.5mmol/LCaCL2，10mmol/LHEPES，NaOH调节PH值7.6，过滤后4℃保存备用。临用前按0.5mg/ml称取胶原酶P粉末，溶于该液，-20℃保存。8、10%Hank’s胰岛洗液的配制：将一瓶Hank’s(H1387，Sigma)盐溶于1000ml双蒸水中，加入0.35克NaHCO3，100ml胎牛血清（10%FBS），4℃预冷。9、石蜡切片柠檬酸修复储存液配制：0.1M柠檬酸溶液A:21.0g柠檬酸+1L蒸馏水；0.1M柠檬酸钠B:29.41g+1L蒸馏水。10、石蜡切片柠檬酸修复工作液配制：9mlA+41mlB+450ml双蒸水即为0.01MPH6.0柠檬酸盐缓冲液。11、他莫昔芬和玉米油注射剂：玉米油和乙醇按9:1混合，56℃水浴中加热溶解即为对照溶剂。称取适量他莫昔芬粉剂，56℃水浴中溶解于对照溶剂中，配成10mg/ml浓度，避光保存于4℃冰箱中（一周备用），注射前水浴加热至37℃。实验方法1、Ubc9和RipCre小鼠的基因鉴定25 华中科技大学博士学位论文剪取4-6周小鼠尾部组织0.5cm，按照WizardSVGenomicDNAPurificationSystem的步骤提取小鼠DNA，行PCR鉴定小鼠基因型：RipCre引物序列：R:5’-TTCCATGGAGCGAACGACGAGACC-3’，F:5’-AACCTGGATAGTGAAACAGGGGC-3’RipCre反应条件：95℃15min94℃1min58℃1.5min35cycles72℃2min72℃10min4℃PCR获得510bp条带表明为转基因小鼠。Ubc9引物序列：R:5’-TCTGAGGCTAAGATGTATGGCCAGCT-3’，F:5’-GCCTGACCTGGATCTTGCCCTC-3’Ubc9反应条件：94℃2min94℃30s60℃30s35cycles72℃30s72℃2min野生型条带为220bp；Loxp型为356bp。2、Ubc9基因敲除的诱导fl/fl1）Ubc9小鼠和他莫昔芬诱导性、胰岛β细胞特异性表达RipCre转基因小鼠+/-fl/-+/+fl/fl杂交，F1代RipCreUbc9小鼠相互杂交得到RipCreUbc9小鼠。+/+fl/fl2）8-12周龄的RipCreUbc9纯合子小鼠，体重约30克左右，每天腹腔注射1mg(100μl/d)的他莫昔芬，连续注射5天；同窝同性别小鼠作为对照连续5天注射100μl含10%乙醇的玉米油。3、Ubc9敲除的基因水平鉴定：取少量小鼠胰腺组织或者数个胰岛，按照WizardSVGenomicDNAPurificationSystem的步骤提取DNA，行PCR鉴定。Ubc9敲除的引物序列：26 华中科技大学博士学位论文R:5’-CTCAGGATTCCAGCACCACACG-3’，F:5’-TACAGTGCCCACCACCACCATT-3’反应条件：94℃3min94℃30s60℃30s35cycles72℃30s72℃10min25℃1min4、Ubc9敲除的蛋白水平Westernblot鉴定：1)胶原酶充盈消化胰腺，显微镜下挑拣小鼠胰岛，置于100-500μl加了蛋白酶抑制剂的Ripabuffer里，冰上静置30min，4℃超声碎裂细胞，4℃20000rpm离心10min，留取含蛋白的上清液检测；2）Bradford比色法测蛋白浓度，每孔加样20-40μl；3）100V恒压SDS-PAGE电泳，90-120min溴酚蓝达底端停止；4）剪6张大小与凝胶一致的滤纸和一张PVDF膜，转移缓冲液中浸泡3~5分钟（PVDF膜预先在100％甲醇中浸泡5min使之完全浸透，迅速转至转移缓冲液）；将转移电泳槽塑料支架平放，置放顺序为海绵、3块滤纸、PVDF膜、凝胶及另外3块滤纸和海绵，注意驱除夹层间气泡；用支架夹紧上述各层，置电转移槽中，PVDF膜一侧靠正极，凝胶一侧靠负极，并灌满转移缓冲液以淹没凝胶；5）接通电源，20mA转膜过夜；6）)转移结束，取出PVDF膜，做好标记，可切一小条进行蛋白质染色处理，以检查转移效果；7）取出PVDF膜，浸于5%脱脂奶粉封闭液中，封闭1小时，倒去封闭液；加入溶有一抗的封闭液（稀释倍数参见抗体说明书）4℃轻轻震荡过夜；8）回收一抗封闭液，4℃保存。加大量漂洗液（含0.05％Tween-20的1xPBS）洗3次，每次5min，轻轻震荡；加入溶有HRP标记的二抗的新鲜配制的封闭液，1：5000稀释；室温轻轻震荡1h；9）取出PVDF膜，于室温用漂洗液洗7次，每次5min；27 华中科技大学博士学位论文10）等量混合ECL试剂A液和B液。将膜片置混合液中于室温下振荡温育1min，每平方厘米膜片至少使用0.125ml。11）平头镊钳住膜片，垂直置于吸水纸，吸去过量试剂，置膜片于二层保鲜膜之间，小心赶尽气泡；12）将膜片吸附蛋白面朝上置于X光片盒中，暗室中压上X光片，曝光30秒显影，调节曝光时间，再次曝光显影。5、葡萄糖耐量实验（GTT）1）称取适量葡萄糖粉末溶于生理盐水中配成20%浓度，并用0.2μm滤器过滤备用，小鼠注射剂量为2g/kg。+/+fl/fl2）8-12周龄的RipCreUbc9纯合子小鼠，体重约30g左右，连续5天注射他莫昔芬（实验组）或玉米油（对照组）后一周开始检测。每组小鼠不少于6只。3）实验前一天下午5点将小鼠禁食16小时，保持正常饮水。4）实验当天上午9时，称取每只小鼠空腹体重并剪小鼠脚趾标记，根据小鼠体重计算葡萄糖的注射剂量。5）注射前采取小鼠尾部血样用血糖仪和血糖试纸检测小鼠的空腹血糖水平并记录（0min）。并于小鼠腹腔注射葡萄糖后30min，60min，90min，120min，180min等时间点同样方法检测小鼠血糖水平，并记录。6、胰岛素耐受实验（ITT）1）小鼠腹腔胰岛素注射剂量为0.75U/kg。胰岛素储存液0.1U/μl，用生理盐水1:1000稀释。+/+fl/fl2）8-12周龄的RipCreUbc9纯合子小鼠，体重约30克左右，连续5天注射他莫昔芬（实验组）或玉米油（对照组）后一周开始检测。每组小鼠不少于6只。3）小鼠实验当天随机饮食，下午2点实验。称取每只小鼠体重并剪小鼠脚趾标记，根据小鼠体重计算胰岛素的注射剂量。4）注射前采取小鼠尾部血样用血糖仪和血糖试纸检测小鼠基础血糖水平并记录（0min）。缓慢注射胰岛素后15min,30min,45min,60min分别检测小鼠血糖并记录。28 华中科技大学博士学位论文7、小鼠血清胰岛素的测定（ELISA）1）选取他莫昔芬或玉米油注射后不同时间段的小鼠作为检测对象，CO2处死小鼠后立即剪开胸廓用3ml注射器直接抽取心脏血液约500μl-1000μl，摘掉针头，将血液缓慢注入1.5mlEP管中，室温静置30min让其凝结。2）凝结后的血样室温4000g的速度离心10min，将上清液（血清）转移到新的EP管中，标记，-80℃保存。3）血清胰岛素水平的测定（ELISA）按ALPCO公司试剂盒步骤进行：将包被了胰岛素抗原的微孔条在室温下平衡，每孔分别加入10μl处理好的标准品、对照和样品，每孔设3个复孔。4）每孔加入WorkingStrengthConjugate液，封口膜封好，置于摇床上700-900rpm速度室温下振摇孵育2小时。5）倒掉反应液，用350ul稀释好的washbuffer反复清洗每孔6遍，每次清洗之后，倒掉washbuffer，然后在卫生纸上轻轻拍打，以去掉孔中残留的液体。6）每孔加入100ul底物，封口膜封好，置于摇床上700-900rpm速度室温下振摇孵育15min。反应孔中的液体会变成蓝色，蓝色的深浅程度与胰岛素的含量成正比。7）每孔中加入100ul反应终止液，用手轻轻拍打板架使液体充分混匀，并去除气泡。这时反应孔中的颜色会由蓝色变成黄色。在5min中内，用酶标仪读取各个孔在以450nm波长下的光吸收值（OD值）。8）用Excel软件分析实验结果，根据标准品的浓度值和光吸收数值，取对数值做标准曲线，根据标准曲线所得公式换算出检测样品的胰岛素浓度值(ng/ml)。29 华中科技大学博士学位论文Insulinstandardcurve8y=2.5873x+0.0972R=0.993465OD4321000.511.522.53Insulinconcentration(ng/ml)8、分离胰岛的胰岛素分泌实验（GSIS）1）加入500μl3.3mM葡萄糖工作液到5个1.5ml冻存管，标记为3.3-1，3.3-2，3.3-3，3.3-4，3.3-5；加入500μl16.7mM葡萄糖工作液到5个1.5ml冻存管，标记为，16.7-1，16.7-2，16.7-3，16.7-4，16.7-5。置于37℃孵箱中。2）准备2个100mm培养皿，分别加入3.3mM葡萄糖工作液。3）将RPMI1640中培养过夜的150个胰岛置于第一个培养皿中，37℃孵箱中平衡30min，移入第二个培养皿中。4）第二个培养皿中的胰岛在倒置显微镜下每管挑拣8-10个移入前面的10个冻存管中。37℃水浴中孵育60min，轻轻晃动。5）离心，收集每管上清300μl行ELISA检测，余下的胰岛检测蛋白质浓度并记录。9、低剂量STZ糖尿病模型建立+/+fl/fl1）8-12周RipCreUbc9纯合子小鼠，体重约30克，连续5天注射他莫昔芬或玉米油，停药5天，小鼠随机喂食，注射第一天称取小鼠体重，按40mg/kg计算每只小鼠每天注射1%STZ的剂量并记录；检测小鼠注射前血糖水平并记录；2）戴上口罩手套小心称取适量STZ粉剂装于干燥EP管中，置于冰上，与配制好的柠檬酸注射液一起带到动物房，新鲜配制1%STZ，按每只小鼠计算的所需剂量30min内注射完毕；3）连续注射5天，停药第3天采小鼠尾部血样检测小鼠血糖水平；血糖≥250mg/dl持续5天为造模成功。10、小鼠胰岛的分离纯化30 华中科技大学博士学位论文1）取胰岛前的准备工作：（1）水浴箱调到38℃，计时器放旁边，设好时间8min；（2）离心机1300rpm，1min调好；(3)10%Hank’s胰岛洗液装于50ml离心管中置于冰上；(4)配好的胶原酶P置于冰上解冻；(5)根据取材的小鼠只数准备数只空的50ml离心管，置于冰上，每管加入2ml配好的胶原酶P；（6）5ml注射器内抽3ml胶原酶P，与30G的注射针头连接，置于冰上。2）小鼠CO2处死后，快速打开胸腹腔，立即剪心放血；3）翻开肝脏，解剖显微镜下用显微镊子挑起胆总管，用显微镊夹夹紧胆总管胰头部，然后用连接注射器的30G针头插入胆总管管腔，缓慢推注3ml预冷的胶原酶P，看到胰腺前后叶均充分膨胀（图1）；图1小鼠胰腺的充盈消化：显微镜下找到胆总管，注射器针头穿入胆总管，注入配制的胶原酶P，见整个胰腺组织充盈膨胀。4）剥离完整的充盈胰腺组织，迅速放入加好2ml胶原酶P的50ml离心管内；并迅速置于38℃水浴箱内，轻轻摇晃，安静消化，计时8min。5）8min到时立即拿出50ml离心管，手腕用力振动管内胰腺组织，见其碎成细沙状，迅速倒入30ml4℃预冷的10%Hank’s洗液，振荡，置冰上；31 华中科技大学博士学位论文6）离心，速度达到1300rpm立即减速，停止；倒掉上清，留沉渣；加入30ml预冷的10%Hank’s洗液重悬沉渣，离心，再洗2次；7）沉渣中加入5ml10%Hank’s液，倒入硅化了的6mm玻璃培养皿中，1000μlEppedorf枪头剪去，反复吹吸沉渣使其分散均匀，显微镜下挑选胰岛（图2）；8）挑捡3-4次，胰岛分离干净，将胰岛移入10%胎牛血清的RPMI1640中，5%CO2，37℃培养箱中培养，以进行下一步实验。图2消化挑拣后培养的小鼠胰岛。11、石蜡切片HE染色1)剥离完整的小鼠胰腺组织,立即置入4%多聚甲醛内固定一周,常规石蜡包埋，6μm切片；2）二甲苯（Ⅰ）5min3）二甲苯（Ⅱ）5min4）95%乙醇（Ⅰ）3min5）95%乙醇（Ⅱ）3min6）80%乙醇1min7）蒸馏水1min8）苏木精染色30s9）流水缓冲至无色10）1%盐酸乙醇上下提插30次11）流水缓冲12）95%乙醇脱水30s13）0.5%曙红染色<1min14）蒸馏水稍洗15）95%乙醇上下提插30次32 华中科技大学博士学位论文16）95%乙醇上下提插30次17）100%乙醇上下提插30次18）100%乙醇上下提插30次19）二甲苯2min20）封片12、石蜡切片免疫荧光染色1）石蜡切片常规脱蜡至水（同HE染色）；2)3%H2O2室温孵育10min；3）蒸馏水冲洗，PBS浸泡5min；4）抗原修复：将盛有柠檬酸盐缓冲液的容器放入水浴箱中，水浴箱加热至96℃；将切片浸入预热至96℃的柠檬酸盐缓冲液中，计时20min；5）移出盛柠檬酸盐缓冲液的容器（切片浸于其中），冷却至室温；6）PBS浸泡5minx3次；7）5%的正常山羊血清（SantaCruz）室温封闭10min；8）倾去血清，勿洗，滴加1:100稀释的胰岛素一抗工作液，4℃孵育过夜；9）PBS浸泡5minx3次；10）滴加1:400-500稀释的AlexaFluor488lgG荧光抗体或Cy2荧光抗体，室温避光孵育2h；11）PBS浸泡5minx3次；12）DAPI（Sigma，1:2000稀释）染色1min；13）封片，荧光显微镜下观察、拍照。13、石蜡切片免疫组化染色：1）根据RabbitABCstainingSystem(SC-2018)（SantaCruz公司）步骤进行；2）石蜡切片常规脱蜡至水；3)3%H2O2室温孵育10min；4）蒸馏水冲洗，PBS浸泡5min；5）抗原修复同上6）PBS浸泡5minx3次；7）5%的正常山羊血清（SantaCruz）室温封闭10min；33 华中科技大学博士学位论文8）倾去血清，勿洗，滴加1:100稀释的Insulin一抗(CellSignaling)工作液，4℃孵育过夜；9）PBS浸泡5minx3次；10）滴加1:400-500稀释的生物素标记二抗，37℃孵育30min；11）PBS浸泡5minx3次；12）滴加AB酶试剂，37℃孵育30min；13）PBS浸泡5minx3次；14）显色剂显色（Insulin2min）；15）自来水充分冲洗；16）苏木精复染；17）梯度乙醇脱水，二甲苯透明；18）中性树胶封片；19）光学显微镜下观察，拍照。14、β细胞凋亡的TUNEL染色1）根据ApoptagplusFluoresceininSiteApoptosisDetectionKit(S7111)（Chemiconinternational公司）步骤进行；2）石蜡切片常规脱蜡至水；3）PBS浸泡5min；4）0.3%H2O2室温孵育10min；5）PBS浸泡3minx2次；6）蛋白酶K（Sigma）室温孵育15min；7)PBS浸泡2minx2次；28)轻轻甩掉玻片上水分，吸干组织周围液体，滴加75μl/5cm平衡液在组织上，室温孵育5min；29）轻轻甩掉玻片上水分，吸干组织周围液体，滴加55μl/5cmTdT酶工作液在组织上，湿盒内37℃孵育2小时；10）将玻片置入终止液里，37℃孵育30min，每10min将玻片轻提起和放下1次；11）PBS浸泡1minx3次；212）轻轻甩掉玻片上水分，吸干组织周围液体，滴加65μl/5cm地高辛连接工作34 华中科技大学博士学位论文液，湿盒内室温避光孵育30min；13）PBS浸泡4minx4次；14）滴加DAPI染30秒；15）封片，荧光显微镜下观察，拍照。15、细胞内ROS检测1）小鼠胰岛分离纯化，将胰岛移入10%RPMI1640中，5%CO2，37℃培养箱中培养过夜；2）第二天，将胰岛移入含钙和镁的HBSS中洗2遍；TM3）以下根据Image-iTLIVEGreenReactiveOxygenSpeciesDetectionKit(I36007)试剂盒步骤进行；4）在2ml含钙和镁的HBSS中加入5μl10mM的羧基-H2DCFDA中，混匀；5）将HBSS洗过的胰岛移入上液，37℃避光孵育30min；6）孵育的最后5min，加入Hoechst33342染核；7）HBSS洗3遍，将胰岛移入96孔板，每孔加入20个大小一致的胰岛，并补平各孔液平面，全自动多功能酶标仪读取各孔荧光强度值。16、统计分析所有数据用平均值±标准差表示，P<0.05则认为有明显统计学意义。小鼠Ubc9基因敲除后糖尿病的发病率用生存曲线表示。其余表达数据用SPSS11.5进行单向方差方法分析。实验结果一、Ubc9基因敲除小鼠的杂交和鉴定+/+fl/fl1、RipCreERUbc9纯合子小鼠的获得fl/fl+/-fl/-Ubc9和RipCreER转基因小鼠杂交得到F1代RipCreERUbc9小鼠。再+/-fl/-+/+fl/fl用F1代RipCreERUbc9小鼠相互杂交得到RipCreERUbc9小鼠。取+/+fl/fl+/+fl/flRipCreERUbc9小鼠作为亲本繁殖，所获后代均为RipCreERUbc9小鼠。35 华中科技大学博士学位论文+/+fl/-+/+fl/fl+/+fl/fl图3F1代小鼠基因型：RipCreERUbc9或RipCreERUbc9。挑选RipCreERUbc9雌雄小鼠合笼，繁殖后代即为纯合子小鼠。+/+fl/fl图4RipCreERUbc9纯合子小鼠繁殖3-4代后常规抽样复查小鼠基因型，明确有无Cre酶的缺失，确保实验的有效性。2、确定Ubc9基因敲除结果+/+fl/fl8周龄雄性或雌性RipCreERUbc9小鼠连续5天腹腔注射50mg/kg体重的他莫昔芬，每天一次；同窝同性别小鼠注射玉米油作为对照组。注射的最后一天定为D0天。于D5天处死小鼠，消化胰岛提取DNA行PCR检测，了解Ubc9基因水平的表达情况；并用RIPA裂解液裂解胰岛提取蛋白，跑PAGE胶作Westernblot检测Ubc9蛋白的表达。结果如下：PCR示他莫昔芬组胰岛表达Ubc9敲除条带，对照组不表达Ubc9敲除条带，提示他莫昔芬组Ubc9基因被敲除（图5）；-/-Westernblot中对照组高表达Ubc9蛋白，而他莫昔芬组（Ubc9）几乎未检测到Ubc9蛋白表达（图6）。以上结果显示：他莫昔芬诱导的Ubc9基因敲除是可靠有效的。图5PCR结果：他莫昔芬组高表达Ubc9敲除条带（320bpm），对照组不表达阳性条带。16-20：他莫昔芬组，21-22：对照组，7：阳性对照，N：阴性对照。36 华中科技大学博士学位论文-/-ControlUbc9Ubc9β-actin-/-图6Westernblot结果：对照组Ubc9蛋白高表达，他莫昔芬组（Ubc9）几乎未检测到Ubc9蛋白的表达。3、Ubc9基因敲除的性别差异性比较雌性和雄性小鼠同时注射他莫昔芬，发现雌性小鼠的基因敲除率为50%（12只小鼠中6只敲除），雄性小鼠的基因敲除率为75%（12只小鼠中9只敲除），雄性小鼠的基因敲除率高于雌性，虽然组间差异不显著（P>0.05），但考虑到他莫昔芬为雌激素受体调节剂，可能对雌鼠的生理功能影响更大，后续试验选雄性小鼠进行。二、Ubc9基因缺陷导致小鼠发生糖耐量异常1、空腹血糖和葡萄糖耐量实验异常他莫昔芬组或对照组小鼠在D9天进行葡萄糖耐量实验。实验前小鼠饥饿16小时，16小时后检测小鼠空腹血糖值（0min），他莫昔芬组为117.6±27.8mg/dl，对照组为91±12.7mg/dl，组间差异显著（P<0.05）；两组小鼠腹腔注射2g/kg葡萄糖溶液后血糖值迅速上升，30min时达到峰值，实验组小鼠为594.1±8.4mg/dl，对照组为453.4±85.1mg/dl，组间差异极显著（P<0.01）；随后对照组小鼠血糖迅速下降，60min时下降到237.4±58.9mg/dl，90min时下降到200mg/dl以下，180min时降到121.6±12.0mg/dl，基本恢复到正常随机血糖值；而他莫昔芬组血糖60min时仍高居558.4±33.7mg/dl，90min和120min时缓慢下降到400-500mg/dl间，180min时仍高达268.9±99.1mg/dl，各时间点与对照组相比间差异极显著（P<0.01）（图7）。37** 华中科技大学博士学位论文IPGTT700600500400300Ubc9-/-Glucose(mg/dl)Control200Blood10000306090120180Time(min)图7葡萄糖耐量实验：他莫昔芬组0min的空腹血糖值明显高于对照组；葡萄糖注射后30min，两组均达峰值，随后对照组血糖迅速下降而他莫昔芬组血糖居高，下降缓慢，180min时尚未恢复正常（N=12）。﹡：P<0.05。2、胰岛素耐受实验（ITT）正常为了明确他莫昔芬组小鼠葡萄糖耐量异常的产生是体内胰岛素的分泌减少还是体内产生了胰岛素抵抗所致，我们进一步进行了两组小鼠的胰岛素耐受实验。如图所示：他莫昔芬组小鼠实验前（0min）的随机血糖值为130.2±22.7mg/dl，和对照组小鼠142.2±32.2mg/dl比较，无差异显著性（P>0.05）；腹腔注射胰岛素后，两组小鼠血糖值均下降，30min达最低值，之后血糖值回升，60min基本恢复至实验前血糖水平，两组小鼠各实验点血糖值均无显著差异性（P>0.05）（图8）。以上两实验结果说明：Ubc9基因缺失所致的小鼠糖耐量下降是胰岛素的分泌不足产生，而不是外周器官发生胰岛素抵抗所致。38 华中科技大学博士学位论文ITT160140120100Ubc9-/-80control60（mg/dl）Bloodglucose40200015304560Time（min）图8胰岛素耐受实验：两组小鼠实验前的随机血糖值无差异，注射胰岛素后血糖下降，30min达最低值，之后血糖缓慢上升，60min恢复至实验前水平，实验组和对照组各时间点血糖值的差异不显著（P>0.05）。三、Ubc9基因缺陷小鼠发生糖尿病1、Ubc9基因敲除小鼠全部血糖升高，出现糖尿病症状从D3天开始监测他莫昔芬组和对照组小鼠的血糖变化，每周2次，血糖值≥250mg/dl为血糖升高，连续3天血糖值升高即认为小鼠血糖异常，发生糖尿病可能。12只他莫昔芬组小鼠在D43天出现1只随机血糖值≥250mg/dl，并持续3天以上；D48天出现4只血糖异常；D50天出现3只血糖异常；D63-65天余下的4只陆续发生血糖异常；并且，12只实验小鼠陆续出现多尿、多饮、竖毛等糖尿病症状。而12只对照组小鼠无一例发生血糖和症状的异常（图9）。39 华中科技大学博士学位论文100Ubc9-/-Control50DiabetesIncidence(%)05254565Days图9Ubc9基因敲除小鼠糖尿病的发生率：Ubc9-/-组D43天1只小鼠发病，D50-55天有7只小鼠发病，到D65天所有12只小鼠均出现糖尿病；而对照组无1例发生糖尿病。2、Ubc9基因敲除小鼠平均血糖变化曲线监测两组小鼠的随机血糖值，D7到D42天两组小鼠血糖持续在正常水平，他莫昔芬组血糖值略高于对照组，但组间差异不显著（P>0.05）。D42天之后，他莫昔芬组平均血糖值开始上升，D49天达335.8±25.4mg/dl，较对照组154.4±6.2mg/dl差异明显（P<0.01）；D63天他莫昔芬组血糖继续升高达423.5±26.2mg/dl，D70天达478.6±12.4mg/dl，分别较对照组169.9±6.2mg/dl和166.3±6.0mg/dl差异极显著（P<0.01）（图10）。并且随着血糖值升高，他莫昔芬组小鼠开始出现尿量增加，皮毛竖立等糖尿病症状。Ubc9-/-Control600500﹡﹡400﹡﹡300Glucose(mg/dl)200Blood10007142128354249566370Days40 华中科技大学博士学位论文图10平均血糖值变化曲线：D7-D42天他莫昔芬组血糖略高于对照组，组间差异不显著（P>0.05）；D42天之后他莫昔芬组血糖明显上升，显著高于对照组。﹡P<0.05。3、Ubc9基因敲除小鼠血清胰岛素水平变化比较他莫昔芬组各时间点D15、30、45、60和D75天血清内平均胰岛素水平。D45天他莫昔芬组胰岛素水平为0.30±0.10ng/ml，较对照组0.39±0.11ng/ml明显下降（P<0.05）；到D60天和D75天分别降为0.26±0.10ng/ml和0.24±0.09ng/ml，与对照组相比，差异极显著（P<0.01）（图11）。以上实验结果说明：Ubc9缺失导致小鼠胰岛素水平下降，血糖水平升高，最终出现自发糖尿病病症。0.50.450.4*0.35**0.30.250.20.150.10.05Seruminsulin(ng/ml)0ControlD15D30D45D60D75Daysaftertam/oilinjection图11他莫昔芬组各时间点的血清胰岛素水平，D45天较对照组明显下降，D60和D75天*下降更为明显。P<0.05。4、小鼠体重监测曲线监测两组小鼠体重变化，随着他莫昔芬组小鼠自发糖尿病，D60天后小鼠体重有下降趋势，但下降缓慢，在D75天之前未检测到与对照组的明显差异性（P>0.05）（图12）41 华中科技大学博士学位论文UBC9-/-Control30252015Weight（g）1050051015202530354045505560657075Daysaftertam/oilinjection图12小鼠体重变化曲线：两组小鼠体重变化无明显差异性（P>0.05）；D65天后实验组小鼠体重有下降趋势。四、Ubc9基因缺陷导致胰岛和β细胞形态、功能异常1、Ubc9基因缺陷后胰岛团变小，形态异常胰腺组织石蜡切片，HE染色结果显示：对照组胰岛团大而圆，边界清晰圆滑；内部的β细胞排列均匀整齐，细胞核圆，大小一致；胞浆着色均匀饱满。他莫昔芬组胰岛团随着Ubc9基因敲除的时间变化，逐渐缩小，边界断续呈锯齿状；岛内细胞数目减少，核大小不一，排列紊乱，可见染色质凝聚成点状或半月状，胞浆浓缩，嗜伊红染色，出现典型的凋亡小体（图13）。42 华中科技大学博士学位论文ControlHEx400Ubc9-/-D15HEx400Ubc9-/-D30HEx400Ubc9-/-D45HEx400Ubc9-/-D60HEx400Ubc9-/-D75HEx400图13HE染色：Ubc9-/-胰岛较对照胰岛逐渐变小，形态异常，岛内细胞减少，出现染色质聚集，胞浆浓缩等改变。箭头所示为胰岛内出现的凋亡小体。2、Ubc9基因缺陷后胰岛团变小，数目减少石蜡切片胰岛素免疫组化染色显示：对照组胰岛分布数目较多、胰岛团较大，β细胞团占切片组织面积的2%，胰腺内总重量约1.84±0.06mg；他莫昔芬组D1543 华中科技大学博士学位论文和D30天β细胞团总重量为1.71±0.05mg和1.63±0.04mg，与对照组相比差异不显著（P>0.05）；D45天β细胞团占切片组织面积的0.9%，总重量约0.76±0.02mg，较对照组明显降低（P<0.05）；D60和D75天β细胞团重量继续降为0.43±0.01mg和0.16±0.00mg，较对照组差异显著（P<0.05）（图14）。21.81.61.41.21*0.80.6*0.4*0.2Betacellmass(mg)0ControlD15D30D45D60D75Daysaftertam/oilinjection图14对照组胰岛分布数目较多，胰岛较大；他莫昔芬组胰岛数目逐渐减少，胰岛面积和重量均逐渐减小，与对照组差异显著。（P<0.05）3、Ubc9缺陷后β细胞数目减少，形态异常胰岛素免疫组化染色和荧光免疫染色显示：对照组胰岛大，圆，胰岛内β细胞分布均匀，占胰岛细胞85%以上，仅边缘可见少量非β细胞分布；他莫昔芬组β细胞逐渐减少，胰岛内未染色的非β细胞(α、δ、ε和PP细胞)占胰岛内细胞比例增加；并且，β细胞形态异常，核大小不一，胞浆内胰岛素分泌不均匀；由于大量β细胞的减少，胰岛萎缩塌陷变小，边缘呈不规则状或锯齿状（图15,16）。44 华中科技大学博士学位论文Controlx400Ubc9-/-D15x400Ubc9-/-D30x400Ubc9-/-D45x400Ubc9-/-D60x400Ubc9-/-D75x400图15胰岛素免疫组化染色：与对照组相比，他莫昔芬组胰岛内β细胞明显减少，胞核大小不一，胞浆胰岛素不均匀，非β细胞比例增加。45 华中科技大学博士学位论文ControlUbc9-/-D45Ubc9-/-D75InsulinDAPIMerged图16胰岛素和DAPI双染：与对照组相比，他莫昔芬组内分泌胰岛素的β细胞明显减少，非β细胞比例增加；整个胰岛团塌陷，不规则。4、Ubc9缺陷后β细胞分泌胰岛素的功能降低为进一步了解胰岛β细胞的分泌功能，我们消化获得胰岛，体外培养，进行了葡萄糖刺激的胰岛素分泌（GSIS）实验。在3.3mM的低葡萄糖工作液中，对照组分泌的胰岛素水平为2.37±0.38ng/ml，他莫昔芬组D45天的胰岛素分泌水平明显下降到1.84±0.18ng/ml（P<0.05）；在16.7mM的高葡萄糖刺激下，对照组胰岛迅速分泌胰岛素达13.49±1.30ng/ml，而D30天和D45天的他莫昔芬组胰岛分泌水平分别为10.69±1.44ng/ml和8.09±1.08ng/ml，与对照组相比显著降低（P<0.05）。可见，Ubc9基因缺陷明显影响了β细胞的分泌功能，特别是在高糖刺激下，β细胞的反应能力明显降低（图17）。46 华中科技大学博士学位论文GSIS161412*10*3.3mM816.7mM64*2Insulinsecretion(nginsulin/mgprotein)0controlD15D30D45Daysaftertam/oilinjection图17葡萄糖刺激的胰岛素分泌实验：3.3mM的低糖下他莫昔芬组D45天的胰岛素水平较对照组明显下降；16.7mM的高糖刺激下，他莫昔芬组D30天和D45天的胰岛分泌能力较对照组明显下降。P<0.05*。五、Ubc9基因缺陷致使β细胞凋亡1、TUNEL染色石蜡切片行TUNEL染色，在对照小鼠胰岛内未见TUNEL阳性细胞，而他莫昔芬组小鼠胰岛内可见多个β细胞凋亡，尤其在D45天明显，细胞核发生染色质聚集断裂，出现TUNEL阳性染色（图18）。47 华中科技大学博士学位论文InsulinDAPITUNELMergedControlUbc9-/-D453*2.521.510.50TUNEL+Insulin+cells/islet(%)ControlD45图18β细胞胰岛素和TUNEL染色显示：对照组胰岛内未见β细胞发生凋亡，而他莫昔芬组D45天胰岛内可见多个TUNEL阳性的β细胞，组间差异显著，P<0.05*。2、凋亡因子CleavedCasepase3染色为进一步证实胰岛内β细胞发生了凋亡，行凋亡因子CleavedCasepase3染色，在对照组胰岛内未见CleavedCasepase3阳性细胞，而他莫昔芬组胰岛内可见多个凋亡的β细胞胞浆内出现CleavedCasepase3红染（图19）。以上结果证明：Ubc9信号缺失导致胰岛内大量β细胞发生凋亡，这是胰岛变小、数目减少、形48 华中科技大学博士学位论文态异常，以及胰岛分泌功能下降的原因。InsulinDAPICasepase3MergedControlUbc9-/-D45图19β细胞Insulin和Cleavedcaspase-3染色：对照组胰岛内未见caspase-3阳性细胞；而他莫昔芬组D45天胰岛内可见多个β细胞胞浆内caspase-3阳性表达。六、Ubc9基因缺陷使β细胞对STZ的抵抗力降低1、Ubc9基因缺陷使注射STZ后小鼠血糖升高明显STZ（链脲菌素）对胰岛β细胞有毒性作用。我们从注射他莫昔芬或玉米油后D6天开始在两组小鼠腹腔内连续注射5天STZ（40mg/kg），并监测两组小鼠的血糖变化。最后一次STZ注射日设为D0天，血糖在D3天呈缓慢上升趋势，他莫昔芬组平均血糖于D6天升到278.11±21.56mg/dl，而对照组小鼠平均血糖为231.6±27.20mg/dl，尚在正常范围内；到D8天，他莫昔芬组小鼠血糖为362.56±29.40mg/dl，显著高于对照组255.4±33.18mg/dl（P<0.05）；D8-D20天，Ubc9-/-组小鼠血糖持续上升达600mg/dl，而对照组小鼠血糖上升较平缓，最终接近400mg/dl，组间差异显著（P<0.05）。可见，Ubc9缺失降低了小鼠β细胞对STZ毒性的抵抗能力，使其胰岛素分泌减少，血糖值升高较快、较高（图20）。49 华中科技大学博士学位论文低剂量STZ对小鼠血糖的影响700600**500***400*Ubc9-/-300Glucose(mg/dl)Control200Blood10002468101214161820DaysafterSTZinjection图20他莫昔芬和对照组对STZ的抵抗能力：对照组血糖升高较平稳，最高值接近400mg/dl；而他莫昔芬组血糖值升高较快，最高接近600mg/dl，组间差异明显（*P<0.05）。2、Ubc9基因缺陷的β细胞对STZ的抵抗力降低HE染色结果显示，STZ注射5天之后，他莫昔芬组β细胞较对照组排列紊乱，胰岛内出现凋亡小体的β细胞较多；胰岛素染色可见，胰岛内β细胞减少，β细胞和非β细胞比例降低。结果说明：β细胞在Ubc9基因缺陷后对STZ毒性更敏感，更易发生凋亡。ControlUbc9-/-HE染色Insulin50 华中科技大学博士学位论文图21他莫昔芬组β细胞较对照组β细胞对STZ的抵抗力降低，更易发生凋亡；他莫昔芬组胰岛内β细胞和非β细胞比例降低。小结1、UBC9缺陷导致小鼠空腹血糖升高，缺陷早期即出现糖耐量异常；2、UBC9信号缺失导致小鼠血糖自发升高，并发生发展为糖尿病；3、UBC9信号缺失致使β细胞凋亡，胰岛团变小，数目减少，形态异常；4、UBC9信号缺失影响β细胞的胰岛素分泌能力；5、UBC9缺陷的β细胞对STZ毒性的抵抗力降低。51 华中科技大学博士学位论文第二部分Ubc9信号缺失引发β细胞凋亡导致糖尿病的机制研究实验材料一、主要实验仪器倒置显微镜日本Olympus公司ZEISS解剖显微镜德国ZEISS公司Olympus荧光倒置显微镜日本Olympus公司POLARstar全自动多功能酶标仪德国Omega公司血糖仪以色列SHIRAT公司血糖试纸以色列SHIRAT公司DK-8D型电热恒温水槽美国Labnet公司恒温水浴箱美国Labnet公司88－I型定时恒温磁力搅拌器美国Fisher公司（Thermix）CO2恒温培养箱德国Heraeus公司MS2型微型振荡器德国IKA公司WD-9405型水平摇床德国BioekelScientific公司Ultrafree-MC10-kDa滤器美国Millipore公司DYY-8型稳压稳流电泳仪美国fisher公司水平电泳槽美国Bio－Rad公司小型垂直电泳槽美国Bio－Rad公司微量加样器德国Eppendorf公司电子分析天平瑞士Mettlertoledo公司超净工作台美国labconco公司5415R低温高速离心机德国Eppendorf公司－80℃超低温冰箱德国Nuaire公司52 华中科技大学博士学位论文Minispin台式高速离心机德国Eppendorf公司低温高速离心机，5415R德国Eppendorf公司微孔透析袋美国Bio－Rad公司50ml深低温离心管美国Bio－Rad公司二、主要实验材料动物：Rip-CreER转基因小鼠来自于Dr.DouglasMelton(HarvardUniversity)fl/f惠赠，Ubc9转基因小鼠（C57BL/6背景）自己构建并购买于JacksonLaboratory，均饲养于乔治亚瑞金斯大学(Georgiaregentsuniversity）SPF级实验动物中心。NIT-1胰岛细胞株：购买于美国ATCC公司，冻存于液氮中。293细胞株：购买于美国ATCC公司，冻存于液氮中。SUMO1腺病毒转染液，EMPTY腺病毒转染液：构建并购买于美国Q.BIOgene公司，-80℃保存。三、主要试剂HBSS(无NAHCO3)粉剂美国Sigma公司RPMI1640美国Gibco公司Hank’s平衡盐溶液（HBSS）美国Sigma公司1xDMEM(1g/lglucose)培养基美国Cellgro公司胎牛血清、小牛血清美国Gibco公司非必需氨基酸溶液美国Cellsignaling公司ABAM双抗美国Cellsignaling公司HEPES溶液美国Cellsignaling公司谷氨酸溶液美国Cellsignaling公司α-ME溶液美国Cellsignaling公司丙酮酸钠溶液美国Cellsignaling公司IL-1β美国Pepprotech公司TGF-α美国Pepprotech公司53 华中科技大学博士学位论文IFN-γ美国Pepprotech公司胶原酶P瑞士Roche公司PVDF膜美国Invitrogen生命技术公司ECL化学发光试剂美国PIERCE公司CsCl粉剂美国Sigma公司MgCl粉剂美国Sigma公司蔗糖美国Sigma公司二甲基亚砜（DMSO）德国Sigma公司台盼蓝粉（TypanBlue）德国Sigma公司TMImage-iTLIVEGreenReactive美国Invitrogen生命技术公司OxygenSpeciesDetectionKit(I36007)四、主要试剂的配制1、配制SDS-PAGE、Westernblot试剂和RIPA蛋白裂解液等与第一部分相同。2、10%DMEM（低糖）NIT-1细胞培养基配制1）从500ml1xDMEM（1g/l葡萄糖）培养基中吸出70ml，移入另一容器备用；2）在余下的430ml培养基中加入以下成分：50ml胎牛血清、5ml非必需氨基酸溶液、5mlABAM双抗、5mlHEPES溶液、丙酮酸钠溶液5ml、2.5ml谷氨酸溶液和500ulα-巯基溶液。3）混匀，4℃保存备用。3、10%DMEM293细胞培养基配制在500ml1xDMEM（含丙酮酸钠、谷氨酸和糖）内加入以上成分，方法同上。4、2%DMEM培养基配制从500ml1xDMEM（含丙酮酸钠、谷氨酸和糖）中移出15ml，加入10ml胎牛血清、5mlABAM双抗和500ulα-巯基溶液。混匀，4℃保存。5、腺病毒纯化试剂配制1）10mMTris-HCL(PH7.9):称取适量Trisbase粉剂溶于500ml双蒸水中，盐酸调节PH；2）比重1.4的CsCl溶液：53克CsCl粉剂溶于87ml10MmTris-HCL(PH7.9)；54 华中科技大学博士学位论文3）比重1.2的CsCl溶液：26.8克CsCl粉剂溶于92ml10MmTris-HCL(PH7.9)；4）配制10mMTris-HCL(PH7.9)，2mMMgCl2和4%蔗糖的腺病毒透析液：在4000ml双蒸水中加入160克蔗糖，8ml1mol的MgCl2和4ml10mMTris-HCL(PH7.9)，混匀，4℃保存透析。6、细胞因子配制：1）1mg/ml原液配制：装细胞因子粉剂的管子离心，用去离子水溶解为浓度1mg/ml原液，-80℃保存；2）储存液配制：IL-1β：1ul原液稀释500倍至2ug/ml；TGF-α：2.5ul原液稀释200倍至5ug/ml；IFN-γ：12.5ul原液稀释40倍至25ug/ml；配制好储存液后，10ul分装，-20℃保存备用，解冻一次；3）工作液配制：储存液1:1000稀释为工作液，加入细胞培养基中，终浓度IL-1β：2ng/ml；TGF-α：5ug/ml；IFN-γ：25ug/ml。7、免疫沉淀-Westernblot试剂配制1）PBS：NaCl20mM,KCL2.68mM,Na2HPO410mM,KH2PO41.76mM(pH7.4)，室温保存。2）1×lysisbuffer：Tris-HCl20mmol/L(pH7.5),NaCl150mM,EDTA1mM,EGTA1mM,TritonX-1001%,Sodiumpyrophosphate2.5mM,β-Glycerrophosphate1mM,NaVO41mM,Leupeptin1ug/ml,－20℃保存。（稀释成1×lysisbuffer后加入1mMPMSF）3）2×SDSsamplebuffer：Tris-Cl（pH6.8）125mM，SDS4%，甘油20%，DTT100mM，溴酚兰0.02%，室温保存。4）分离胶（下层胶）（10%SDS-PAGE）15ml：30%丙烯酰胺5ml，10×lowerbuffer(pH8.8)1.5ml，H2O8.5ml，10%AP15μl，TEMED15μl。5．10×lowerbuffer(pH8.8)100ml：42.25gTrisbase，10ml10％SDS，室温保存。6．压缩胶（上层胶）（4%SDS-PAGE）5ml：30%丙烯酰胺0.65ml，4×stackingbuffer(pH6.8)1.25ml，H2O3.05ml，10%AP5μl，TEMED5μl。7．4×stackingbuffer(pH6.8)100ml：6.06Trisbase，4ml10％SDS，室温保存。55 华中科技大学博士学位论文8．电泳缓冲液：Trisbase0.125M，甘氨酸0.96mM，SDS0.5%，室温保存。9．电转缓冲液：Tris25mM，甘氨酸0.2mM，甲醇10%，室温保存。10．10×TBS(pH7.6)：Trisbase2.42%，NaCl8%，室温保存。11．TBST：1×TBS，Tween-200.1%，室温保存。实验方法1、Ubc9基因敲除的诱导、Westernblot(免疫印迹)、小鼠糖耐量实验、胰岛的消化分离等方法同第一部分。2、NIT-1和293细胞的解冻复苏、冻存和培养按照一般的细胞培养操作（略）。3、腺病毒的扩增步骤1）293细胞解冻复苏，接种于10%DMEM，100mm直径的培养皿中；2）定期换液，细胞生长覆盖皿底达70%时，消化细胞，1:5传代培养；3）细胞生长覆盖皿底达95%时，吸走培养液，加入1ml的病毒转染液（1ml不加血清的DMEM中加入5ul的病毒保存液），十字形晃动培养皿，促进病毒扩散；4）在37℃,5%CO2孵箱中转染6小时，约20min拿出十字形晃动一次，保持转染液均匀分散在培养皿内；5）6小时后加入10%DMEM9ml，37℃,5%CO2孵箱中过夜；6）第二天，每盘细胞换10ml10%DMEM液体，37℃，5%CO2孵箱中培养；7）第三天，荧光显微镜下观察细胞的转染率，转染达90%以上时收集细胞，离心，丢弃上清，留细胞，-80℃冻存。4、腺病毒的纯化和过滤1）-80℃冻存的腺病毒置于37℃水浴箱中，等完全溶解后置于-20℃冷冻；反复解冻-冷冻3次，让病毒颗粒从293细胞中完全释放出来；2）4℃最大转数离心10min，收集病毒上清，丢弃沉渣，置冰上；3）在50ml透明低温离心管中缓缓加入8ml比重1.4的CsCl，在此上面再轻轻覆盖6ml比重1.2的CsCl，可见两层间的分界线；4）在冰上轻轻加入20ml收集的病毒上清液；动作轻，三层溶液不要混合了；5）4℃离心，SW28的转子23000rpm离心90min，减速率设为0；56 华中科技大学博士学位论文6）离心后可见薄的病毒层，吸出病毒层液，用28G的针头注入透析袋中，置于透析液中，4℃透析2小时，半小时更换一次透析液；7）28G针头吸出病毒液，分装在EP管中，-80℃保存。5、腺病毒滴度测定41）准备5ml293细胞悬液（2%DMEM），浓度4x10细胞/ml，种植在平底的96孔板中，每孔加50ul细胞液；2）37℃,5%CO2孵箱中培养12小时；3）准备10个1.5ml的EP管，按顺序作好标记，并在每管中加入0.9ml无血清的DMEM；-14）将0.1ml病毒存储液加入第一管中，混合均匀，标记为10；-25）从第一管中吸取0.1ml病毒稀释液加入第二管中，混合均匀，标记为10；-36）从第二管中吸取0.1ml病毒稀释液加入第三管中，混合均匀，标记为10；-1-107）依次进行，10个管中标记从10到10，注意每次更换枪头；8）倍比稀释完后，在孵育293细胞的96孔板中依次加入50ul病毒稀释液，每排加-3-10一个浓度，依次从10加到10，每排留2孔仅加DMEM作阴性对照；9）37℃，5%CO2孵箱中培养10天；10）荧光显微镜下观察并记录每稀释度的阳性孔数，根据公式算出病毒滴度PFU/ml。6、腺病毒转染NIT-1细胞1）复苏NIT-1细胞，离心洗2遍，10%DMEM（低糖）培养基接种于6孔板中，37℃，5%CO2孵箱中培养过夜；2）2-3天换培养液，细胞健康生长到70%融合度时，吸走培养液，温PBS洗2遍，估计孔中细胞数量，加入不同稀释倍数的病毒液转染过夜，转染方法同上293细胞；3）第二天荧光显微镜下观察细胞转染率，将细胞生长良好，转染率又较高的病毒稀释度作为该病毒的最佳转染滴度。7、免疫沉淀Western-blot步骤1）细胞转染后24－36小时后，吸净培养液，PBS漂洗一次；2）加入500μl预冷的1×lysisbuffer,于冰上裂解细胞5分钟；57 华中科技大学博士学位论文3）将细胞裂解液转移到1.5mleppondorf管内，4℃，13000g离心30min；4）将离心后的上清液分为两份：一份35μl，加入等体积的2×SDSsamplebuffer,混匀后于100℃煮10min，做为总细胞裂解液，取6－10μl进行电泳，Westernblot检测目的蛋白的表达；5）分A/Gbeads：管中加入5μlAgroseA/Gbeads及5μl（1μg）抗体，将抗体和beads混合，补充lysisbuffer，使每管能均匀分配到50μlbeads及抗体的混合物，将beads及抗体的混合物按每管50μ（l含5μlbeads及5μl抗体）分配到1.5mlEppendorf管中，再加入400μl1×lysisbuffer，备用；6）取蛋白上清液400μl加入已分好的beads中，使终体积达到850μl，将管子固定到混匀器上使混匀器匀速旋转（15rpm）免疫沉淀过夜；7）将免疫沉淀后的溶液于4℃3000rpm离心3分钟，去上清，加入500μl1×lysisbuffer洗涤beads，4℃3000rpm离心3分钟，弃上清，洗涤3次；8）最后一次洗涤完毕，弃上清，加入35μl1×lysisbuffer与等体积2×SDSsamplebuffer混合，于100℃煮沸10min，稍离心后取10μl左右上样到PAGE胶，进行电泳；9）电泳完毕，取下PAGE胶，与PVDF膜做成"三明治"形状，用湿转法20mA电转过夜；10）电转完毕，取下PVDF膜，加入5％脱脂奶粉，于脱色摇床摇荡（75rpm）封闭1小时以消除非特异背景；11）封闭完毕，用TBST洗掉牛奶，加入一抗，摇床摇荡孵育过夜，使一抗与特异蛋白结合；12）回收一抗，用TBST洗3次（75rpm），每次5－10分钟；13）加入二抗，摇床孵育1小时；14）用TBST洗3次，每次5－10分钟；15）将ECLA液和ECLB液各1ml，混匀，放入二抗孵育后的PVDF膜30秒，将PVDF膜平铺于曝光盒中，进暗房，用医学X光胶片曝光；16）经过显影、定影后的胶片，于室温下自然风干或烘干，描画蛋白marker便于分析。58 华中科技大学博士学位论文实验结果一、Ubc9基因缺陷使β细胞内氧化应激反应增强1、Ubc9缺失早期β细胞内ROS升高他莫昔芬或玉米油注射后D10天，处死小鼠，胶原酶消化、分离纯化胰岛，10%RPMI1640中，5%CO2，37℃培养过夜；第二天将胰岛移入10%DMEM中，并加入细胞因子IL-1β5ng/ml、TNF-α10ng/ml、IFN-γ100ng/ml刺激6小时，在25μM羧基-H2DCFDA中孵育30min，荧光显微镜下观察结果：他莫昔芬组（Ubc9-/-）胰岛细胞内ROS荧光较强，对照组ROS荧光较弱，组间差异显著（P<0.05）。结果表明：在Ubc9基因缺失早期，β细胞内ROS表达增强，提示氧化应激途径参与了β细胞凋亡，可能是导致β细胞损伤，诱发糖尿病的重要原因之一。胰岛胰岛内ROSUbc9-/-D10Control59 华中科技大学博士学位论文ROS6050*4030RFU(103)2010Superoxideproduction0controlUbc9-/-图22D10天β细胞内ROS荧光值比较：Ubc9-/-组ROS明显高于对照组。P<0.05*2、Ubc9基因缺失晚期β细胞内ROS降低他莫昔芬或玉米油注射后D50天，小鼠胰岛孵育过夜，加入细胞因子IL-1β5ng/ml、TNF-α10ng/ml、IFN-γ100ng/ml刺激6小时后，25μM羧基-H2DCFDA中孵育30min，荧光显微镜下观察结果：对照组ROS荧光较强，而Ubc9-/-胰岛细胞内ROS荧光较弱，组间差异显著（P<0.05）。结果显示：Ubc9缺失晚期，由于胰岛内β细胞大部分凋亡丢失，胰岛内ROS表达明显下降。胰岛胰岛内ROSUbc9-/-D50Control60 华中科技大学博士学位论文ROS4035302520RFU(103)15*105Superoxideproduction0controlUbc9-/-图23D50天β细胞内ROS荧光值比较：Ubc9-/-组ROS明显低于对照组。P<0.05*二、免疫共沉淀探索SUMO对ROS的调控途径1、Ad-Sumo1和Ad-Empty腺病毒转染NIT-1细胞为了明确SUMO化对ROS调控的具体途径，我们作了进一步的机制研究。体外构建的SUMO1和Empty腺病毒转染胰岛细胞NIT-1，转染率达90%。收集转染的NIT-1细胞进行免疫共沉淀研究，可见和对照相比，SUMO1抗体拖下了较多相结合的目的蛋白。我们正在对这些目的蛋白作进一步的分析之中。Ad-Sumo1Ad-Empty图24腺病毒载体Sumo1和Empty转染NIT-1细胞，转染率90%，腺病毒滴度500PFU/ml。61 华中科技大学博士学位论文—75KDa—50—37—25—15Ad-SUMO1Ad-EMPTY图25NIT-1细胞的免疫共沉淀结果：相对于Empty腺病毒，SUMO1拖下较多的目的蛋白。三、Ubc9基因缺陷导致β细胞的内质网应激我们检测了β细胞内与ERstress相关的蛋白表达，1和7为对照样本，2为D15天，3为D30天，4为D45天，5为D60天，6为D75天样本，可见，分子伴侣Grp78和Grp58在后期D60天和D75天的表达增强；p-eIf2a、ATF6和IRE1a也是在D45天以后的表达稍强（图26）。结果提示：在Ubc9缺失后期激活了ERstress的表达，ERstress的表达可能是导致β细胞凋亡的重要原因。1control2D15天3D30天4D45天5D60天6D75天7control图26Westernblot检测ERstress蛋白表达：Ubc9敲除后期，Grp78、Grp58、p-eIf2a、ATF6和IRE1a表达增强。62 华中科技大学博士学位论文小结1、氧化应激反应产生的活性氧自由基(ROS)可能是引起β细胞凋亡的重要原因；2、β细胞失去SUMO化保护，对ROS的抵抗力明显降低；3、SUMO化修饰是否通过NADPH氧化酶途径调控ROS的表达尚需免疫沉淀结果证实；4、内质网应激在Ubc9信号缺失晚期表达增强，是β细胞凋亡的原因还是结果尚需进一步研究。63 华中科技大学博士学位论文讨论一直认为细胞内SUMO化调节主要作用于细胞核内，介导基因的表达、蛋白的形成、蛋白和蛋白、蛋白和DNA之间相互作用等方面(3,29-30)。近来研究显示SUMO化调节在核外也扮演重要角色，包括调节线粒体形态和功能(31)、调控+红藻氨酸盐受体介导的突触信号传递(32)和调节K通道等(33-34)。SUMO化修饰在胰岛β细胞内也发挥重要的调节作用。小鼠和人类的胰岛以及胰岛瘤细胞株INS-1832/13中检测到SUMO1和SUMO3的表达(35)；并且E2连接酶Ubc9和SENP亚基在胰岛和INS-1832/13中也都有表达(35-36)。这是我们建立小鼠模型的理论基础。正如我们PCR和Westernblot所证实的一样，Ubc9诱导性敲除以后，小鼠胰岛内未检测到Ubc9的表达，而对照组Ubc9的表达正常。并且，SUMO化调节在胰岛素分泌方面发挥重要作用(37-38)。葡萄糖刺激下的胰岛素分泌是多因素多步骤调控下的复杂过程，许多基因及蛋白参与其中。线粒体在糖代谢物的刺激下生成较多的ATP，提高了β细胞内ATP：ADP比率，导致ATP敏感的KATP通道关闭，细胞膜去极化，电压依赖性钙通道（Voltage2dependentcalciumchannel,VDCC）开放，胞浆内游离钙离子浓度升高，刺激胰岛素胞外分泌。可见，K通道的阻滞提高了β细胞内胰岛素分泌，SUMO因子对K通道的阻滞调节促进了胰岛素的分泌；但也有相反的报道，SUMO1的过表达抑制了小鼠糖刺激的胰岛素分泌(37)。转录因子PDX-1在胰岛素基因的转录调节中发挥重要作用(39)，当β细胞受到葡萄糖刺激，胞浆内无活性的31kDPDX-1转入细胞核内生成46kD具有转录活性的核蛋白，这个过程有赖于PDX-1蛋白的SUMO化作用，并且，PDX-1调节胰岛素分泌需要募集去乙酰化酶p300(40-42)和组蛋白去乙酰化酶HDAC-1、2(40-41)，这两者都是SUMO化底物。并且，SUMO化调节介导了β细胞死亡和功能紊乱(43-45)。SUMO可与IkBα去磷酸化连接，避免了泛素对IkBα的降解，阻止了NFκB的转位和活化，对细胞内炎症和凋亡起保护作用(46)。并且，在β细胞内SUMO1的过表达保护了IL-1β诱导的凋亡；相反，SENP1的过表达促进了细胞死亡，可见，β细胞内SUMO化调节保护了糖尿病相关的代谢和炎症应激。我们的研究结果与以上结论一致，64 华中科技大学博士学位论文SUMO化作用保护了β细胞，去SUMO化作用导致了β细胞凋亡。在探寻β细胞凋亡的机制方面，我们认为氧化应激是重要原因。由于β细胞抗氧化防御能力低下，细胞内自由基清除酶和ROS清除蛋白水平较低，对氧化应激损伤特别敏感。并且，大量报道证实氧化应激是导致β细胞损伤的重要因素。我们在Ubc9缺陷早期检测到胰岛内强ROS表达，在β细胞大量凋亡后期发现ROS表达明显下降，证明ROS在β细胞凋亡中发挥了重要作用。并在更进一步的机制研究中，我们转染SUMO1腺病毒到胰岛细胞株NIT-1细胞中，希望通过SUMO1的过表达结合相关的底物蛋白，确定凋亡信号通路及靶蛋白。内质网应激也是导致β细胞凋亡的重要途径之一(47-48)。SUMO化修饰在内质网应激中的的作用仅见有限的报道。内质网应激中重要的因子XBP1的能被SUMO化(49)；AyaUemura等发现Ubc9与转录因子pXBP1(S)结合，提高了pXBP1(S)的表达稳定性；抑制Ubc9的表达，减少了pXBP1(S)和应激诱导的转录活性，提示Ubc9是内质网应激的一个新的调控子(50)。我们在研究中发现内质网应激可能参与了β细胞凋亡，但在Ubc9缺陷的早期和晚期到底是怎样发挥作用，我们还在继续研究中。65 华中科技大学博士学位论文参考文献1.GuoB,YangSH,WittyJ,SharrocksAD.SignallingpathwaysandtheregulationofSUMOmodification.BiochemSocTrans.2007Dec;35(Pt6):1414-8.2.TempeD,PiechaczykM,BossisG.SUMOunderstress.BiochemSocTrans.2008Oct;36(Pt5):874-8.3.VergerA,PerdomoJ,CrossleyM.ModificationwithSUMO.Aroleintranscriptionalregulation.EMBORep.2003Feb;4(2):137-42.4.JohnsonES.ProteinmodificationbySUMO.AnnuRevBiochem.2004;73:355-82.5.PraefckeGJ,HofmannK,DohmenRJ.SUMOplayingtagwithubiquitin.TrendsBiochemSci.2012Jan;37(1):23-31.6.DanciuTE,ChupretaS,CruzO,FoxJE,WhitmanM,Iniguez-LluhiJA.Smallubiquitin-likemodifier(SUMO)modificationmediatesfunctionoftheinhibitorydomainsofdevelopmentalregulatorsFOXC1andFOXC2.JBiolChem.2012May25;287(22):18318-29.7.GuoD,LiM,ZhangY,YangP,EckenrodeS,HopkinsD,etal.AfunctionalvariantofSUMO4,anewIkappaBalphamodifier,isassociatedwithtype1diabetes.NatGenet.2004Aug;36(8):837-41.8.LiM,GuoD,IsalesCM,EizirikDL,AtkinsonM,SheJX,etal.SUMOwrestlingwithtype1diabetes.JMolMed(Berl).2005Jul;83(7):504-13.9.MaisonC,RomeoK,BaillyD,DubarryM,QuivyJP,AlmouzniG.TheSUMOproteaseSENP7isacriticalcomponenttoensureHP1enrichmentatpericentricheterochromatin.NatStructMolBiol.2012Apr;19(4):458-60.10.KrumovaP,WeishauptJH.Sumoylationfights"aggregopathies".CellCycle.2012Feb15;11(4):641-2.11.WeiW,YangP,PangJ,ZhangS,WangY,WangMH,etal.Astress-dependentSUMO4sumoylationofitssubstrateproteins.BiochemBiophysResCommun.2008Oct24;375(3):454-9.66 华中科技大学博士学位论文12.WangCY,YangP,LiM,GongF.CharacterizationofanegativefeedbacknetworkbetweenSUMO4expressionandNFkappaBtranscriptionalactivity.BiochemBiophysResCommun.2009Apr17;381(4):477-81.13.WangX,HaiCX.ROSactsasadouble-edgedswordinthepathogenesisoftype2diabetesmellitus:isNrf2apotentialtargetforthetreatment?MiniRevMedChem.2011Oct;11(12):1082-92.14.HurJ,SullivanKA,SchuylerAD,HongY,PandeM,StatesDJ,etal.Literature-baseddiscoveryofdiabetes-andROS-relatedtargets.BMCMedGenomics.2010;3:49.15.BitarMS,Al-MullaF.ROSconstituteaconvergencenexusinthedevelopmentofIGF1resistanceandimpairedwoundhealinginaratmodeloftype2diabetes.DisModelMech.2012May;5(3):375-88.16.NerupJ,Mandrup-PoulsenT,MolvigJ,HelqvistS,WogensenL,EgebergJ.Mechanismsofpancreaticbeta-celldestructionintypeIdiabetes.DiabetesCare.1988Nov-Dec;11Suppl1:16-23.17.ParkEJ,ChoiKS,KwonTK.beta-Lapachone-inducedreactiveoxygenspecies(ROS)generationmediatesautophagiccelldeathingliomaU87MGcells.ChemBiolInteract.2011Jan15;189(1-2):37-44.18.SubasingheW,SyedI,KowluruA.Phagocyte-likeNADPHoxidasepromotescytokine-inducedmitochondrialdysfunctioninpancreaticbeta-cells:evidenceforregulationbyRac1.AmJPhysiolRegulIntegrCompPhysiol.2011Jan;300(1):R12-20.19.LimYA,GrimmA,GieseM,Mensah-NyaganAG,VillafrancaJE,IttnerLM,etal.InhibitionofthemitochondrialenzymeABADrestorestheamyloid-beta-mediatedderegulationofestradiol.PLoSOne.2011;6(12):e28887.20.KoekGH,LiedorpPR,BastA.Theroleofoxidativestressinnon-alcoholicsteatohepatitis.ClinChimActa.2011Jul15;412(15-16):1297-305.21.Suarez-PinzonWL,RabinovitchA.Approachestotype1diabetespreventionbyinterventionincytokineimmunoregulatorycircuits.IntJExpDiabetesRes.67 华中科技大学博士学位论文2001;2(1):3-17.22.SerrezeDV,ChapmanHD,PostCM,JohnsonEA,Suarez-PinzonWL,RabinovitchA.Th1toTh2cytokineshiftsinnonobesediabeticmice:sometimesanoutcome,ratherthanthecause,ofdiabetesresistanceelicitedbyimmunostimulation.JImmunol.2001Jan15;166(2):1352-9.23.KawasakiE,AbiruN,EguchiK.Preventionoftype1diabetes:fromtheviewpointofbetacelldamage.DiabetesResClinPract.2004Dec;66Suppl1:S27-32.24.KawasakiE,EguchiK.IsType1diabetesintheJapanesepopulationthesameasamongCaucasians?AnnNYAcadSci.2004Dec;1037:96-103.25.KubischHM,WangJ,LucheR,CarlsonE,BrayTM,EpsteinCJ,etal.Transgeniccopper/zincsuperoxidedismutasemodulatessusceptibilitytotypeIdiabetes.ProcNatlAcadSciUSA.1994Oct11;91(21):9956-9.26.ChenH,YuM,LiM,ZhaoR,ZhuQ,ZhouW,etal.Polymorphicvariationsinmanganesesuperoxidedismutase(MnSOD),glutathioneperoxidase-1(GPX1),andcatalase(CAT)contributetoelevatedplasmatriglyceridelevelsinChinesepatientswithtype2diabetesordiabeticcardiovasculardisease.MolCellBiochem.2012Apr;363(1-2):85-91.27.Ivanovic-MaticS,MihailovicM,DinicS,MartinovicV,BogojevicD,GrigorovI,etal.TheabsenceofcardiomyopathyisaccompaniedbyincreasedactivitiesofCAT,MnSODandGSTinlong-termdiabetesinrats.JPhysiolSci.2010Jul;60(4):259-66.28.XiangFL,LuX,StruttB,HillDJ,FengQ.NOX2deficiencyprotectsagainststreptozotocin-inducedbeta-celldestructionanddevelopmentofdiabetesinmice.Diabetes.2010Oct;59(10):2603-11.29.SeelerJS,DejeanA.NuclearandunclearfunctionsofSUMO.NatRevMolCellBiol.2003Sep;4(9):690-9.30.MullerS,LedlA,SchmidtD.SUMO:aregulatorofgeneexpressionandgenomeintegrity.Oncogene.2004Mar15;23(11):1998-2008.31.ZuninoR,SchaussA,RippsteinP,Andrade-NavarroM,McBrideHM.The68 华中科技大学博士学位论文SUMOproteaseSENP5isrequiredtomaintainmitochondrialmorphologyandfunction.JCellSci.2007Apr1;120(Pt7):1178-88.32.MartinS,NishimuneA,MellorJR,HenleyJM.SUMOylationregulateskainate-receptor-mediatedsynaptictransmission.Nature.2007May17;447(7142):321-5.33.RajanS,PlantLD,RabinML,ButlerMH,GoldsteinSA.SumoylationsilencestheplasmamembraneleakK+channelK2P1.Cell.2005Apr8;121(1):37-47.34.BensonMD,LiQJ,KieckhaferK,DudekD,WhortonMR,SunaharaRK,etal.SUMOmodificationregulatesinactivationofthevoltage-gatedpotassiumchannelKv1.5.ProcNatlAcadSciUSA.2007Feb6;104(6):1805-10.35.DaiC,BrissovaM,HangY,ThompsonC,PoffenbergerG,ShostakA,etal.Islet-enrichedgeneexpressionandglucose-inducedinsulinsecretioninhumanandmouseislets.Diabetologia.2012Mar;55(3):707-18.36.LinX,LiangM,LiangYY,BrunicardiFC,FengXH.SUMO-1/Ubc9promotesnuclearaccumulationandmetabolicstabilityoftumorsuppressorSmad4.JBiolChem.2003Aug15;278(33):31043-8.37.RajanS,TorresJ,ThompsonMS,PhilipsonLH.SUMOdownregulatesGLP-1-stimulatedcAMPgenerationandinsulinsecretion.AmJPhysiolEndocrinolMetab.2012Mar15;302(6):E714-23.38.LiuLB,OmataW,KojimaI,ShibataH.TheSUMOconjugatingenzymeUbc9isaregulatorofGLUT4turnoverandtargetingtotheinsulin-responsivestoragecompartmentin3T3-L1adipocytes.Diabetes.2007Aug;56(8):1977-85.39.OhlssonH,KarlssonK,EdlundT.IPF1,ahomeodomain-containingtransactivatoroftheinsulingene.EMBOJ.1993Nov;12(11):4251-9.40.MosleyAL,OzcanS.Thepancreaticduodenalhomeobox-1protein(Pdx-1)interactswithhistonedeacetylasesHdac-1andHdac-2onlowlevelsofglucose.JBiolChem.2004Dec24;279(52):54241-7.41.MosleyAL,CorbettJA,OzcanS.Glucoseregulationofinsulingeneexpressionrequirestherecruitmentofp300bythebeta-cell-specifictranscriptionfactor69 华中科技大学博士学位论文Pdx-1.MolEndocrinol.2004Sep;18(9):2279-90.42.StanojevicV,HabenerJF,ThomasMK.Pancreasduodenumhomeobox-1transcriptionalactivationrequiresinteractionswithp300.Endocrinology.2004Jun;145(6):2918-28.43.LuX,YiJ.SUMO-1enhancingthep53-inducedHepG2cellapoptosis.JHuazhongUnivSciTechnologMedSci.2005;25(3):289-91.44.SudharsanR,AzumaY.TheSUMOligasePIAS1regulatesUV-inducedapoptosisbyrecruitingDaxxtoSUMOylatedfoci.JCellSci.2012Dec1;125(Pt23):5819-29.45.ShaoR,RungE,WeijdegardB,BilligH.InductionofapoptosisincreasesSUMO-1proteinexpressionandconjugationinmouseperiovulatorygranulosacellsinvitro.MolReprodDev.2006Jan;73(1):50-60.46.LeeYS,JangMS,LeeJS,ChoiEJ,KimE.SUMO-1repressesapoptosissignal-regulatingkinase1activationthroughphysicalinteractionandnotthroughcovalentmodification.EMBORep.2005Oct;6(10):949-55.47.O'Sullivan-MurphyB,UranoF.ERstressasatriggerforbeta-celldysfunctionandautoimmunityintype1diabetes.Diabetes.2012Apr;61(4):780-1.48.GurzovEN,OrtisF,CunhaDA,GossetG,LiM,CardozoAK,etal.SignalingbyIL-1beta+IFN-gammaandERstressconvergeonDP5/Hrkactivation:anovelmechanismforpancreaticbeta-cellapoptosis.CellDeathDiffer.2009Nov;16(11):1539-50.49.ChenH,QiL.SUMOmodificationregulatesthetranscriptionalactivityofXBP1.BiochemJ.2010Jul1;429(1):95-102.50.UemuraA,TaniguchiM,MatsuoY,OkuM,WakabayashiS,YoshidaH.UBC9RegulatestheStabilityofXBP1,aKeyTranscriptionFactorControllingtheERStressResponse.CellStructFunct.2013Apr18;38(1):67-79.70 华中科技大学博士学位论文综述Oxidativestressresultedfromexcessiveproductionofreactiveoxygenspecies(ROS)inthepathogenesisofpancreaticβcelldestruction1,21QiaohongLaiCong-YiWang*1TheCenterforBiomedicalResearch,TongjiHospital,TongjiMedicalCollege,HuazhongUniversityofScienceandTechnology,1095JiefangAve.Wuhan,430030,China2TheCenterforreproductionMedicine,TongjiHospital,TongjiMedicalCollege,HuazhongUniversityofScienceandTechnology,1095JiefangAve.Wuhan,430030,China*Correspondingauthor,email:cwang@.mcg.edu71 华中科技大学博士学位论文AbstractItiswildlyacceptedthatreactiveoxygenspecies(ROS)causedbyglucolipotoxicitycontributetoβcelldysfunctionanddamageindiabetes.Traditionally,thesourceofROSintheinsulinsecretingpancreaticβcellshasbeenconsideredtobethemitochondrialelectrontransportchain.RecentlysomeinformationandevidenceshowNADPHoxidase-dependentgenerationofROSwasanotherimportantresourceinphagocyteandpancreaticβ-cells.SeveralNOXisoformsareexpressedinthepancreaticisletsandinappropriateactivationofNOXenzymesmaydamagethepancreaticβ-cellsandleadstoprogressionfromthemetabolicsyndrometoType2diabetes.UnderstandingthefunctionalmechanismofNOXenzymesisimportanttoidentifyusefultargetsfornoveltherapeuticstrategies.72 华中科技大学博士学位论文IntroductionDiabetesisametabolicdiseasecharacterizedbyelevationofbloodglucoseconcentration,lipidabnormalitiesandvascularcomplications.Insulinresistanceandpancreaticβ-cellinsufficiencywithrespecttoinsulinproductionaremajorfeaturesintheprogressionofT2DM(51-52).Chronicexposuretoelevatedglucolipotoxicitycancausedamageindifferenttypesofcells,andoxidativestressmaybeacommonreasonincelldysfunction(52-54).Pancreaticβ-cellsdefectivefortheantioxidantdefenseabilitymaybeathighriskforoxidativedamagewithexcessivelevelsofmitochondrialandNADPHoxidaseROSgeneration.ThispointisimportanttotheonsetofT2DManditscomplications.Inthisreview,wewillemphasizedtheimportantroleforNADPHoxidasederivedROS,whichmayalterparametersofsignaltransduction,insulinsecretion,insulinactionandcellproliferationorcelldeath.WeintendtodiscussthedistributionofNADPHoxidaseisoformsinthepancreaticisletsandtheirfunctionsduringT2DMdevelopment,andthefeasibilityforNADPHoxidasetobeausefultargetfordiabeticinterventionstrategieswillbealsodiscussed.TheNADPHoxidasecomplexTheNOXfamilyisagroupofplasmaorsub-cellularmembrane-associatedenzymescomplexfoundinawidevarietyofeukaryoticcells.NowtheNOXfamilycomprisessevenmembers:NOX1,NOX2,NOX3,NOX4,NOX5,DUOX1andDUOX2(54).TheNOX2wasfirstfoundinneutrophils(55),whichisamajorsourcefortheintracellularROS.TheNOX2oxidasecomplexiscomposedoftwotransmembraneproteins(gp91phoxandp22phox)andfourcytosolicproteins(p47phox,p67phox,p40phox,andRac)(56).Themembranecomponentgp91phoxhassixtransmembranedomainsandisunstableintheabsenceofp22phox.Thetwocomponentsgp91phoxandp22phoxaretightlyconnectedwithaheterodimercomplexflavocytochromeb558.Inrestingcells,thisenzymecomplexsystemisinactive,anditscomponentsareseparatedbetweenthecytosolandthemembranes.Theflavocytochromeb558componentislocatedintheplasmamembraneandin73 华中科技大学博士学位论文specificgranules.Onactivation,somestimulisuchasgrowthfactors,cytokines,hormonesorbacterialcomponentsactivatedtheisoformsofproteinkinaseC,whichleadstothecytosoliccomponentsp47phox,p67phoxandp40phoxphosphorylatedandtranslocatedtotheplasmamembraneandinteractwithflavocytochromeb558.Simultaneously,Rac2dissociatesfromaRhoGDPinhibitorandinteractswithp67phox.Thecompleteassemblyofcomponentstransportselectronsthroughbiological•−membranesandcatalysethecytosolicmolecularoxygentoO2(57).NOX1wasthefirsthomologueofNOX2tobedescribed(58).ThesmallGTPaseRacmaybeinvolvedintheregulationofNOX1activity(59-60).NOX4isanenigmaticmemberoftheNOXfamilyofROS-generatingNADPHoxidases.NOX4hasawidetissuedistribution(kidney,endothelialcells,osteoclasts,smoothmusclecells,fibroblasts,mesangialcells,adipocytes,pancreaticisletsandembryonicstemcells)andexhibitsabout39%homologytoNOX2(61).Nox4canconstitutivelyproducehighlevelsofhydrogenperoxide.RecentstudieshavedemonstratedthatNOX4participatesintheregulationofcelldifferentiationinhumanandmousepre-adipocytes(62-64).WhileNOX4expressionhasbeenreportedinpancreaticislets(54),theroleofNOX4inbeta2+cellshasyettobedetermined.NOX5containsEF-handsandisactivatedbyCa(65),andprobablyalsobyPKC(proteinkinaseC)(66).DUOX1andDUOX2arecharacterisedbyEF-handcalcium-bindingdomainandNH2-terminalperoxidasedomain,respectively(65-66).Thesehomologuesareelectrontransportersthatreduce·−O2toO2.ExpressionofNADPHoxidaseisoformsandcomponentsinpancreaticisletsNADPHoxidasefamilycomponents,NOX1–3,areanimportantsourceofROSproductioninthepancreaticislets.AmorerecentstudydemonstratedthatNADPHoxidasecomponentsgp91phoxandp22phoxincreasedproductioninβcellsobtainedfromanimalmodelsoftype2diabetes(67).Oliveiraetal.foundthatNOX2subunitsgp91phox,p22phoxandp47phoxmRNAhasbeenshowntobeexpressedinthepancreaticisletsbyRT-PCR.P47phoxandp67phoxproteinsweredemonstratedin74 华中科技大学博士学位论文theratpancreaticislets(68)bywesternblotting,andp47phoxismainlyexpressesedinthecentraloftheislets,anareamainlypopulatedbybetacells(68).UchizonofoundthatisolatedratisletsconstitutivelyexpressmRNAsofNOX1,NOX2andNOXO1(homologueofp47phox),NOXA1(homologueofp67phox)andp40phoxcomponents(69).Recently,NOX1,NOX2andNOX4mRNAs,andNOXA1,p22phox,p47phoxandp67phoxproteinshavebeendetectedinhumanandratPSCs(pancreaticstellatecells)(70-71).Untilnow,theexactphysiologicalroleofNOXenzymesinthepancreaticisletsisyettobeaddressed.Mostofthestudiesusingthenon-selectiveflavoproteininhibitorDPIandtheantioxidant/NOXinhibitorapocynin(72)suggestedthatNOXenzymesplayaroleininsulinsecretion(51-52,73-74).OxidativestressandpancreaticβcellsdysfunctionGlucoseistransportedacrosstheplasmamembraneviatransportersGLUT1andGLUT2,andisrapidlyphosphorylatedbyaspecificglucokinasewithhighKmforglucose.Thetransportandphosphorylationofglucosedeterminemetabolicfluxthroughglycolysisintheβ-cell.Glucosewasfurthermetabolisedbypyruvatedehydrogenaseandpyruvatecarboxylase,resultinginenhancedtricarboxylicacidcycleactivityandsubsequentATPgeneration.ItleadstoanenhancedratioofATPto+ADPinthecytoplasm.ThiswillresultinclosureoftheATP-sensitiveKchannels+(KATP),decreasingthehyperpolarizingoutwardKflux.Thisresultsin2+depolarizationoftheplasmamembrane,influxofextracellularCa,arapidincrease2+inintracellularCa,andactivationofproteinkinases,whichthenmediateexocytosis2+ofinsulin(75-76).AnelevationofintracellularCacannotonlystimulatemitochondrialgenerationofROS,butactivateproteinkinaseC(PKC),whichmayactivateNADPHoxidasecomplexandenhanceoxidase-dependentgenerationofROSandthusinduceoxidativestressand/orapoptosis(76-77).Incontrastwithmostothermammaliancelltypes,βcellshaverelativelylowlevelsoffree-radical-detoxifyingandredox-regulatingenzymes(78-79)suchas75 华中科技大学博士学位论文catalase,andglutathioneperoxidase(80-81).Theconsequenceoflimitedscavenging2+systemsisthat,uponnormalorexcessiveCastimulation,ROSconcentrationinβ-cellsmayincreaserapidlyandbecomethetargetsofoxidativestress-mediatedtissuedamage(82-83).ExcessivelevelsofreactiveoxygenspeciesdirectlydamagecellsbyoxidizingDNA,proteinandlipids.SometimesexcessiveROSactivateavarietyofintracellularsignalingpathwayssuchasNF-kB,p38MAPK,JNK/SAPK,hexosamineandothers,whichmayleadtotheinsulinresistanceandimpairedinsulinsecretion.Accordingly,thereisalinkamongthehyperglycaemia-andFFA-inducedoxidativestressandthelatecomplicationsofdiabetesandβcelldysfunction.NOXfamilyandβ-celldysfunctioninType2diabetesUndernormalconditions,lowglucoseinducesareleaseofinsulinbyβ-cells.Underchronichigh-glucoseandhigh-FFAconditions,NOX/ROS-dependentβ-celldamagemayleadtoprogressionfromthemetabolicsyndrometoType2diabetes.ChronicoroverwhelmingERstressstimuliassociatedwiththemetabolicsyndromecandisruptproteinfoldingintheER,reduceinsulinsecretion,invokeoxidativestressandactivatecelldeathpathways(84).ImpairmentofnormalproteinfoldingintheERduringERstresscanredirectβ-cellstoapoptoticpathwaysviaNOX4.Furthermore,hyperglycaemiaisinvolvedincelldeaththroughtheROSinhibitionofphosphatasesoftheJNKpathway,NF-κBactivationandcaspaseactivation.ItalsoresultsinincreasedproductionofROSbyNOX1and/orNOX2andgenerationofAGEs,whichinturnareassociatedwithreducedtranscriptionofgenesinvolvedininsulinproduction.InType2diabetes,renin–angiotensinsystem(RAS)activityisup-regulatedinβ-cellsandinendothelialcellsofthepancreaticvasculature.ThephysiologicalroleofthepancreaticRASappearstoinvolveisletbloodflowregulationandaffectinsulinsecretionthatmaybetriggeredbyNOX1.Somestudiesshowchronichyperglycaemiainhibitsinsulinsecretion,activatesangiotensinIItype1·−receptor,andincreasesO2productionandp47phoxandp22phoxexpressioninarat76 华中科技大学博士学位论文insulin-producingcelllineandinisolatedhumanpancreaticislets(85-87).(4)NOXenzymeswerealsoimplicatedinthepancreatitisthatresultedintissueinflammationanddestruction(51,88).PancreaticfibrosisseemstoinvolvetheproductionofTGF-β1,aclassicprofibroticcytokine,andPSCactivation,whichmayinvolveNOX2and/orNOX4.PancreaticfibrosiswasalsoinhibitedbyDPIinvivo,suggestingaroleforNOXisoformsintheactivationofPSCs.(5)NOX2inmacrophagesandneutrophilscontributetoinflammationbyproducingpro-inflammatorycytokinesthatleadtodecreasedβ-cellmassandType2diabetes.ConclusionandoutlookTraditionally,mitochondriawerethoughttobethemostimportantsourceofROSinmetabolism(89).ItisnowbecomingclearNADPHoxidasesoftheNOXfamilywerealsoanimportantresourceofROSproduction.NADPHoxidaseisoformsdistributeinpancreaticislets,andmakearoleininsulinsecretion,β-celldysfunctionandapoptosis,pancreaticfibrosisandtype2diabetes.However,mostoftheexperimentalresultswerecomefromspecificNOXinhibitors.Therefore,studiesusingspecificNOXisoformknockoutmicewillsupplymorecompellingresultstous.Astheresearchevolves,NOXenzymeshavegraduallybeenrecognizedtobeamajorpotentialasdrugtargets(90)indiverseclinicalsettings,andasaresult,developingmorespecificNOXinhibitorswillbeanimportanttaskinthefuture.Reference1.GuoB,YangSH,WittyJ,SharrocksAD.SignallingpathwaysandtheregulationofSUMOmodification.BiochemSocTrans.2007Dec;35(Pt6):1414-8.2.TempeD,PiechaczykM,BossisG.SUMOunderstress.BiochemSocTrans.2008Oct;36(Pt5):874-8.3.VergerA,PerdomoJ,CrossleyM.ModificationwithSUMO.Aroleintranscriptionalregulation.EMBORep.2003Feb;4(2):137-42.4.JohnsonES.ProteinmodificationbySUMO.AnnuRevBiochem.77 华中科技大学博士学位论文2004;73:355-82.5.PraefckeGJ,HofmannK,DohmenRJ.SUMOplayingtagwithubiquitin.TrendsBiochemSci.2012Jan;37(1):23-31.6.DanciuTE,ChupretaS,CruzO,FoxJE,WhitmanM,Iniguez-LluhiJA.Smallubiquitin-likemodifier(SUMO)modificationmediatesfunctionoftheinhibitorydomainsofdevelopmentalregulatorsFOXC1andFOXC2.JBiolChem.2012May25;287(22):18318-29.7.GuoD,LiM,ZhangY,YangP,EckenrodeS,HopkinsD,etal.AfunctionalvariantofSUMO4,anewIkappaBalphamodifier,isassociatedwithtype1diabetes.NatGenet.2004Aug;36(8):837-41.8.LiM,GuoD,IsalesCM,EizirikDL,AtkinsonM,SheJX,etal.SUMOwrestlingwithtype1diabetes.JMolMed(Berl).2005Jul;83(7):504-13.9.MaisonC,RomeoK,BaillyD,DubarryM,QuivyJP,AlmouzniG.TheSUMOproteaseSENP7isacriticalcomponenttoensureHP1enrichmentatpericentricheterochromatin.NatStructMolBiol.2012Apr;19(4):458-60.10.KrumovaP,WeishauptJH.Sumoylationfights"aggregopathies".CellCycle.2012Feb15;11(4):641-2.11.WeiW,YangP,PangJ,ZhangS,WangY,WangMH,etal.Astress-dependentSUMO4sumoylationofitssubstrateproteins.BiochemBiophysResCommun.2008Oct24;375(3):454-9.12.WangCY,YangP,LiM,GongF.CharacterizationofanegativefeedbacknetworkbetweenSUMO4expressionandNFkappaBtranscriptionalactivity.BiochemBiophysResCommun.2009Apr17;381(4):477-81.13.WangX,HaiCX.ROSactsasadouble-edgedswordinthepathogenesisoftype2diabetesmellitus:isNrf2apotentialtargetforthetreatment?MiniRevMedChem.2011Oct;11(12):1082-92.14.HurJ,SullivanKA,SchuylerAD,HongY,PandeM,StatesDJ,etal.Literature-baseddiscoveryofdiabetes-andROS-relatedtargets.BMCMedGenomics.2010;3:49.78 华中科技大学博士学位论文15.BitarMS,Al-MullaF.ROSconstituteaconvergencenexusinthedevelopmentofIGF1resistanceandimpairedwoundhealinginaratmodeloftype2diabetes.DisModelMech.2012May;5(3):375-88.16.NerupJ,Mandrup-PoulsenT,MolvigJ,HelqvistS,WogensenL,EgebergJ.Mechanismsofpancreaticbeta-celldestructionintypeIdiabetes.DiabetesCare.1988Nov-Dec;11Suppl1:16-23.17.ParkEJ,ChoiKS,KwonTK.beta-Lapachone-inducedreactiveoxygenspecies(ROS)generationmediatesautophagiccelldeathingliomaU87MGcells.ChemBiolInteract.2011Jan15;189(1-2):37-44.18.SubasingheW,SyedI,KowluruA.Phagocyte-likeNADPHoxidasepromotescytokine-inducedmitochondrialdysfunctioninpancreaticbeta-cells:evidenceforregulationbyRac1.AmJPhysiolRegulIntegrCompPhysiol.2011Jan;300(1):R12-20.19.LimYA,GrimmA,GieseM,Mensah-NyaganAG,VillafrancaJE,IttnerLM,etal.InhibitionofthemitochondrialenzymeABADrestorestheamyloid-beta-mediatedderegulationofestradiol.PLoSOne.2011;6(12):e28887.20.KoekGH,LiedorpPR,BastA.Theroleofoxidativestressinnon-alcoholicsteatohepatitis.ClinChimActa.2011Jul15;412(15-16):1297-305.21.Suarez-PinzonWL,RabinovitchA.Approachestotype1diabetespreventionbyinterventionincytokineimmunoregulatorycircuits.IntJExpDiabetesRes.2001;2(1):3-17.22.SerrezeDV,ChapmanHD,PostCM,JohnsonEA,Suarez-PinzonWL,RabinovitchA.Th1toTh2cytokineshiftsinnonobesediabeticmice:sometimesanoutcome,ratherthanthecause,ofdiabetesresistanceelicitedbyimmunostimulation.JImmunol.2001Jan15;166(2):1352-9.23.KawasakiE,AbiruN,EguchiK.Preventionoftype1diabetes:fromtheviewpointofbetacelldamage.DiabetesResClinPract.2004Dec;66Suppl1:S27-32.24.KawasakiE,EguchiK.IsType1diabetesintheJapanesepopulationthesameasamongCaucasians?AnnNYAcadSci.2004Dec;1037:96-103.79 华中科技大学博士学位论文25.KubischHM,WangJ,LucheR,CarlsonE,BrayTM,EpsteinCJ,etal.Transgeniccopper/zincsuperoxidedismutasemodulatessusceptibilitytotypeIdiabetes.ProcNatlAcadSciUSA.1994Oct11;91(21):9956-9.26.ChenH,YuM,LiM,ZhaoR,ZhuQ,ZhouW,etal.Polymorphicvariationsinmanganesesuperoxidedismutase(MnSOD),glutathioneperoxidase-1(GPX1),andcatalase(CAT)contributetoelevatedplasmatriglyceridelevelsinChinesepatientswithtype2diabetesordiabeticcardiovasculardisease.MolCellBiochem.2012Apr;363(1-2):85-91.27.Ivanovic-MaticS,MihailovicM,DinicS,MartinovicV,BogojevicD,GrigorovI,etal.TheabsenceofcardiomyopathyisaccompaniedbyincreasedactivitiesofCAT,MnSODandGSTinlong-termdiabetesinrats.JPhysiolSci.2010Jul;60(4):259-66.28.XiangFL,LuX,StruttB,HillDJ,FengQ.NOX2deficiencyprotectsagainststreptozotocin-inducedbeta-celldestructionanddevelopmentofdiabetesinmice.Diabetes.2010Oct;59(10):2603-11.29.SeelerJS,DejeanA.NuclearandunclearfunctionsofSUMO.NatRevMolCellBiol.2003Sep;4(9):690-9.30.MullerS,LedlA,SchmidtD.SUMO:aregulatorofgeneexpressionandgenomeintegrity.Oncogene.2004Mar15;23(11):1998-2008.31.ZuninoR,SchaussA,RippsteinP,Andrade-NavarroM,McBrideHM.TheSUMOproteaseSENP5isrequiredtomaintainmitochondrialmorphologyandfunction.JCellSci.2007Apr1;120(Pt7):1178-88.32.MartinS,NishimuneA,MellorJR,HenleyJM.SUMOylationregulateskainate-receptor-mediatedsynaptictransmission.Nature.2007May17;447(7142):321-5.33.RajanS,PlantLD,RabinML,ButlerMH,GoldsteinSA.SumoylationsilencestheplasmamembraneleakK+channelK2P1.Cell.2005Apr8;121(1):37-47.34.BensonMD,LiQJ,KieckhaferK,DudekD,WhortonMR,SunaharaRK,etal.SUMOmodificationregulatesinactivationofthevoltage-gatedpotassiumchannel80 华中科技大学博士学位论文Kv1.5.ProcNatlAcadSciUSA.2007Feb6;104(6):1805-10.35.DaiC,BrissovaM,HangY,ThompsonC,PoffenbergerG,ShostakA,etal.Islet-enrichedgeneexpressionandglucose-inducedinsulinsecretioninhumanandmouseislets.Diabetologia.2012Mar;55(3):707-18.36.LinX,LiangM,LiangYY,BrunicardiFC,FengXH.SUMO-1/Ubc9promotesnuclearaccumulationandmetabolicstabilityoftumorsuppressorSmad4.JBiolChem.2003Aug15;278(33):31043-8.37.RajanS,TorresJ,ThompsonMS,PhilipsonLH.SUMOdownregulatesGLP-1-stimulatedcAMPgenerationandinsulinsecretion.AmJPhysiolEndocrinolMetab.2012Mar15;302(6):E714-23.38.LiuLB,OmataW,KojimaI,ShibataH.TheSUMOconjugatingenzymeUbc9isaregulatorofGLUT4turnoverandtargetingtotheinsulin-responsivestoragecompartmentin3T3-L1adipocytes.Diabetes.2007Aug;56(8):1977-85.39.OhlssonH,KarlssonK,EdlundT.IPF1,ahomeodomain-containingtransactivatoroftheinsulingene.EMBOJ.1993Nov;12(11):4251-9.40.MosleyAL,OzcanS.Thepancreaticduodenalhomeobox-1protein(Pdx-1)interactswithhistonedeacetylasesHdac-1andHdac-2onlowlevelsofglucose.JBiolChem.2004Dec24;279(52):54241-7.41.MosleyAL,CorbettJA,OzcanS.Glucoseregulationofinsulingeneexpressionrequirestherecruitmentofp300bythebeta-cell-specifictranscriptionfactorPdx-1.MolEndocrinol.2004Sep;18(9):2279-90.42.StanojevicV,HabenerJF,ThomasMK.Pancreasduodenumhomeobox-1transcriptionalactivationrequiresinteractionswithp300.Endocrinology.2004Jun;145(6):2918-28.43.LuX,YiJ.SUMO-1enhancingthep53-inducedHepG2cellapoptosis.JHuazhongUnivSciTechnologMedSci.2005;25(3):289-91.44.SudharsanR,AzumaY.TheSUMOligasePIAS1regulatesUV-inducedapoptosisbyrecruitingDaxxtoSUMOylatedfoci.JCellSci.2012Dec1;125(Pt23):5819-29.81 华中科技大学博士学位论文45.ShaoR,RungE,WeijdegardB,BilligH.InductionofapoptosisincreasesSUMO-1proteinexpressionandconjugationinmouseperiovulatorygranulosacellsinvitro.MolReprodDev.2006Jan;73(1):50-60.46.LeeYS,JangMS,LeeJS,ChoiEJ,KimE.SUMO-1repressesapoptosissignal-regulatingkinase1activationthroughphysicalinteractionandnotthroughcovalentmodification.EMBORep.2005Oct;6(10):949-55.47.O'Sullivan-MurphyB,UranoF.ERstressasatriggerforbeta-celldysfunctionandautoimmunityintype1diabetes.Diabetes.2012Apr;61(4):780-1.48.GurzovEN,OrtisF,CunhaDA,GossetG,LiM,CardozoAK,etal.SignalingbyIL-1beta+IFN-gammaandERstressconvergeonDP5/Hrkactivation:anovelmechanismforpancreaticbeta-cellapoptosis.CellDeathDiffer.2009Nov;16(11):1539-50.49.ChenH,QiL.SUMOmodificationregulatesthetranscriptionalactivityofXBP1.BiochemJ.2010Jul1;429(1):95-102.50.UemuraA,TaniguchiM,MatsuoY,OkuM,WakabayashiS,YoshidaH.UBC9RegulatestheStabilityofXBP1,aKeyTranscriptionFactorControllingtheERStressResponse.CellStructFunct.2013Apr18;38(1):67-79.51.BellGI,PolonskyKS.Diabetesmellitusandgeneticallyprogrammeddefectsinbeta-cellfunction.Nature.2001Dec13;414(6865):788-91.52.KahnSE.Therelativecontributionsofinsulinresistanceandbeta-celldysfunctiontothepathophysiologyofType2diabetes.Diabetologia.2003Jan;46(1):3-19.53.KajimotoY,KanetoH.Roleofoxidativestressinpancreaticbeta-celldysfunction.AnnNYAcadSci.2004Apr;1011:168-76.54.BedardK,KrauseKH.TheNOXfamilyofROS-generatingNADPHoxidases:physiologyandpathophysiology.PhysiolRev.2007Jan;87(1):245-313.55.RaoMallaR,RaghuH,RaoJS.RegulationofNADPHoxidase(Nox2)bylipidraftsinbreastcarcinomacells.IntJOncol.2010Dec;37(6):1483-93.56.SongHY,JuSM,SeoWY,GohAR,LeeJK,BaeYS,etal.Nox2-basedNADPHoxidasemediatesHIV-1Tat-inducedup-regulationofVCAM-1/ICAM-1and82 华中科技大学博士学位论文subsequentmonocyteadhesioninhumanastrocytes.FreeRadicBiolMed.2011Mar1;50(5):576-84.57.KrauseKH.TissuedistributionandputativephysiologicalfunctionofNOXfamilyNADPHoxidases.JpnJInfectDis.2004Oct;57(5):S28-9.58.BanfiB,MaturanaA,JaconiS,ArnaudeauS,LaforgeT,SinhaB,etal.AmammalianH+channelgeneratedthroughalternativesplicingoftheNADPHoxidasehomologNOH-1.Science.2000Jan7;287(5450):138-42.59.SumimotoH,MiyanoK,TakeyaR.MolecularcompositionandregulationoftheNoxfamilyNAD(P)Hoxidases.BiochemBiophysResCommun.2005Dec9;338(1):677-86.60.ChengG,DieboldBA,HughesY,LambethJD.Nox1-dependentreactiveoxygengenerationisregulatedbyRac1.JBiolChem.2006Jun30;281(26):17718-26.61.ShioseA,KurodaJ,TsuruyaK,HiraiM,HirakataH,NaitoS,etal.Anovelsuperoxide-producingNAD(P)Hoxidaseinkidney.JBiolChem.2001Jan12;276(2):1417-23.62.GuichardC,MoreauR,PessayreD,EppersonTK,KrauseKH.NOXfamilyNADPHoxidasesinliverandinpancreaticislets:aroleinthemetabolicsyndromeanddiabetes?BiochemSocTrans.2008Oct;36(Pt5):920-9.63.SchroderK,WandziochK,HelmckeI,BrandesRP.Nox4actsasaswitchbetweendifferentiationandproliferationinpreadipocytes.ArteriosclerThrombVascBiol.2009Feb;29(2):239-45.64.MahadevK,MotoshimaH,WuX,RuddyJM,ArnoldRS,ChengG,etal.TheNAD(P)HoxidasehomologNox4modulatesinsulin-stimulatedgenerationofH2O2andplaysanintegralroleininsulinsignaltransduction.MolCellBiol.2004Mar;24(5):1844-54.65.BanfiB,MolnarG,MaturanaA,StegerK,HegedusB,DemaurexN,etal.ACa(2+)-activatedNADPHoxidaseintestis,spleen,andlymphnodes.JBiolChem.2001Oct5;276(40):37594-601.66.SumimotoH.Structure,regulationandevolutionofNox-familyNADPHoxidases83 华中科技大学博士学位论文thatproducereactiveoxygenspecies.FEBSJ.2008Jul;275(13):3249-77.67.NakayamaM,InoguchiT,SontaT,MaedaY,SasakiS,SawadaF,etal.IncreasedexpressionofNAD(P)HoxidaseinisletsofanimalmodelsofType2diabetesanditsimprovementbyanAT1receptorantagonist.BiochemBiophysResCommun.2005Jul15;332(4):927-33.68.OliveiraHR,VerlengiaR,CarvalhoCR,BrittoLR,CuriR,CarpinelliAR.Pancreaticbeta-cellsexpressphagocyte-likeNAD(P)Hoxidase.Diabetes.2003Jun;52(6):1457-63.69.UchizonoY,TakeyaR,IwaseM,SasakiN,OkuM,ImotoH,etal.ExpressionofisoformsofNADPHoxidasecomponentsinratpancreaticislets.LifeSci.2006Dec14;80(2):133-9.70.SakaiK,MatsumotoK,NishikawaT,SuefujiM,NakamaruK,HirashimaY,etal.Mitochondrialreactiveoxygenspeciesreduceinsulinsecretionbypancreaticbeta-cells.BiochemBiophysResCommun.2003Jan3;300(1):216-22.71.BrownleeM.Biochemistryandmolecularcellbiologyofdiabeticcomplications.Nature.2001Dec13;414(6865):813-20.72.TakeyaR,UenoN,KamiK,TauraM,KohjimaM,IzakiT,etal.Novelhumanhomologuesofp47phoxandp67phoxparticipateinactivationofsuperoxide-producingNADPHoxidases.JBiolChem.2003Jul4;278(27):25234-46.73.BulteauAL,Ikeda-SaitoM,SzwedaLI.Redox-dependentmodulationofaconitaseactivityinintactmitochondria.Biochemistry.2003Dec23;42(50):14846-55.74.MartensGA,CaiY,HinkeS,StangeG,VandeCasteeleM,PipeleersD.Glucosesuppressessuperoxidegenerationinmetabolicallyresponsivepancreaticbetacells.JBiolChem.2005May27;280(21):20389-96.75.NewsholmeEA,BlomstrandE.Branched-chainaminoacidsandcentralfatigue.JNutr.2006Jan;136(1Suppl):274S-6S.76.MorganD,Oliveira-EmilioHR,KeaneD,HirataAE,SantosdaRochaM,Bordin84 华中科技大学博士学位论文S,etal.Glucose,palmitateandpro-inflammatorycytokinesmodulateproductionandactivityofaphagocyte-likeNADPHoxidaseinratpancreaticisletsandaclonalbetacellline.Diabetologia.2007Feb;50(2):359-69.77.YuJH,KimKH,KimH.RoleofNADPHoxidaseandcalciumincerulein-inducedapoptosis:involvementofapoptosis-inducingfactor.AnnNYAcadSci.2006Dec;1090:292-7.78.CacicedoJM,BenjachareowongS,ChouE,RudermanNB,IdoY.Palmitate-inducedapoptosisinculturedbovineretinalpericytes:rolesofNAD(P)Hoxidase,oxidantstress,andceramide.Diabetes.2005Jun;54(6):1838-45.79.BeeharryN,ChambersJA,GreenIC.Fattyacidprotectionfrompalmiticacid-inducedapoptosisislostfollowingPI3-kinaseinhibition.Apoptosis.2004Sep;9(5):599-607.80.LenzenS,DrinkgernJ,TiedgeM.Lowantioxidantenzymegeneexpressioninpancreaticisletscomparedwithvariousothermousetissues.FreeRadicBiolMed.1996;20(3):463-6.81.TiedgeM,LortzS,DrinkgernJ,LenzenS.Relationbetweenantioxidantenzymegeneexpressionandantioxidativedefensestatusofinsulin-producingcells.Diabetes.1997Nov;46(11):1733-42.82.RobertsonRP,HarmonJ,TranPO,TanakaY,TakahashiH.Glucosetoxicityinbeta-cells:type2diabetes,goodradicalsgonebad,andtheglutathioneconnection.Diabetes.2003Mar;52(3):581-7.83.TanakaY,TranPO,HarmonJ,RobertsonRP.Aroleforglutathioneperoxidaseinprotectingpancreaticbetacellsagainstoxidativestressinamodelofglucosetoxicity.ProcNatlAcadSciUSA.2002Sep17;99(19):12363-8.84.MathewsCE,LeiterEH.ResistanceofALR/Ltisletstofreeradical-mediateddiabetogenicstressisinheritedasadominanttrait.Diabetes.1999Nov;48(11):2189-96.85.DuXL,EdelsteinD,RossettiL,FantusIG,GoldbergH,ZiyadehF,etal.85 华中科技大学博士学位论文Hyperglycemia-inducedmitochondrialsuperoxideoverproductionactivatesthehexosaminepathwayandinducesplasminogenactivatorinhibitor-1expressionbyincreasingSp1glycosylation.ProcNatlAcadSciUSA.2000Oct24;97(22):12222-6.86.LenzenS.Oxidativestress:thevulnerablebeta-cell.BiochemSocTrans.2008Jun;36(Pt3):343-7.87.GroverAK,KwanCY,SamsonSE.Effectsofperoxynitriteonsarco/endoplasmicreticulumCa2+pumpisoformsSERCA2bandSERCA3a.AmJPhysiolCellPhysiol.2003Dec;285(6):C1537-43.88.MathewsCE,LeiterEH.Constitutivedifferencesinantioxidantdefensestatusdistinguishalloxan-resistantandalloxan-susceptiblemice.FreeRadicBiolMed.1999Aug;27(3-4):449-55.89.MathewsCE,GraserRT,SavinovA,SerrezeDV,LeiterEH.UnusualresistanceofALR/Ltmousebetacellstoautoimmunedestruction:roleforbetacell-expressedresistancedeterminants.ProcNatlAcadSciUSA.2001Jan2;98(1):235-40.90.GeisztM,LetoTL.TheNoxfamilyofNAD(P)Hoxidases:hostdefenseandbeyond.JBiolChem.2004Dec10;279(50):51715-8.86 华中科技大学博士学位论文已发表文章ThesignificanceofantralfolliclesizepriortostimulationinpredictingovarianresponseinamultipledoseGnRHantagonistprotocolQiaohongLai,CaiChen,ZhijunZhang,ShuZhang,QilinYuandCong-YiWang*TheCenterforBiomedicalResearch,TongjiHospital,TongjiMedicalCollege,HuazhongUniversityofScienceandTechnology,1095JiefangAve.,Wuhan430030,China*CorrespondenceandreprintrequestsshouldbeaddressedtoDr.Cong-YiWang(Tel:86-27-8366-3485;email:cwang@georgiahealth.edu).87 华中科技大学博士学位论文AbstractThepurposeofthisstudywastoevaluatewhetherantralfolliclessizecanpredictovarianresponsesininfertilewomenundergoinginvitrofertilizationandembryotransfer（IVF-ET）inamultipledosegonadotrophinreleasinghormone(GnRH)antagonistprotocol.Thiswasaretrospectiveanalysisincluding214patientsand214cycles.Basalantralfollicularsizeandantralfollicularcount(AFC)weremeasuredbyvaginalultrasonographyonmenstrualday3beforeapplicationofgonadotropin.Usingamultipleregressionanalysis,AFCachievedthebestpredictivevalueinrelationtothenumberofoocyteobtained,followedbyantralfolliclesize,basalFSHandBMI(Log(totaleggno.)=13.692+0.823×log(AFC)–0.661(antralfolliclesize)–0.326(D3FSH)–0.281(BMI),R=0.635,P<0.001).Alltreatmentcyclesweredividedintofourgroupsaccordingtothediameterofantralfollicles(i.e.,2-5mm,6-7mm,7-8mm,8-10mm)toevaluatetheinfluenceofvariousfactors.AntralfolliclesizewassignificantlycorrelatedwithrFSHdurationanddosage(P<0.01).Thewomenwith6-7mmantralfollicleshadalowerrateofcyclecancellationcomparedwiththeotherthreegroups(7.5%vs.16.7%,15.6%and16.0%,respectively).Basalantralfolliclesizewasagoodpredictortoevaluateovarianresponse.Themiddlesizeofantralfollicles(6-7mm)presentbeforeovarianstimulationachievedbestpredictivevalueinrelationtoIVFoutcomeswhereasthelargerantralfollicles(8-10mm)hadhighchanceoftransfercancellationandlowclinicalpregnancyrate.Keywords:antralfolliclesize/antralfolliclecount/cetrorelix/GnRHantagonist/IVF-ETIntroductionTherecruitmentanddevelopmentofmultiplefolliclesinresponsetogonadotrophinstimulationareessentialforthesuccessfultreatmentofinfertilitybytheassistedreproductivetechniques.Poorovarianresponsehasbeensuggestedtobeassociatedwithpoorpregnancyratesandhighcyclecancellationrates(Keayetal.,1997;88 华中科技大学博士学位论文Tarlatzisetal.,2003),whileexaggeratedovarianresponseisalsofoundtobeanriskfactorforbothovarianhyperstimulationsyndrome(OHSS)(AboulgharandMansour,2003)andmultiplepregnancy.Therefore,predictionofovarianresponsespriortostimulationisacriticalpointnecessarytobeaddressedduringthecounselingofpatientswithinfertility,whichcanalsoprovidehelptotailortheoptimaldosageofgonadotrophinadministrationforaparticularpatient.Traditionally,thepatientageandearlyphaseserumfolliclestimulatinghormone(FSH)levelsareconsideredtobethemostusefulparametersforthepredictionofovarianreserve(ScottandHofmannetal.,1995;Shararaetal.,1998;PadillaandGarcia,1989).Particularly,earlyFSHconcentrationisthoughttobethemostpowerfulpredictivemarkerforassessingovarianresponseascomparedwiththatofpatientage(Cahilletal.,1994;Sharifetal.,1998).However,followupstudieshavethenconsistentlydemonstratedthesignificantpredictivevalueforantralfolliclecount(AFC)inovarianresponseandpregnancyrateofpatientswithinvitrofertilization(IVF)treatment,inwhichAFCachievedthebestpredictivevalue,followedbybasalFSH,bodymassindex(BMI)andageofthepatient(Ngetal.,2000b).Otherthanthesemarkers,additionalhormonalandultrasoundmarkerssuchasseruminhibinB(Seiferetal.,1997;Tinkanenetal.,1999;Dziketal.,2000),serumanti-MÜllerianhormone(Seiferetal.,2002;vanRooijetal.,2002;Fanchinetal.,2003),ovarianvolume(Syropetal.,1995,1999;Lassetal.,1997)andovarianstromalbloodflow(Zaidietal.,1996;Engmannetal.,1999;KupesicandKurjak,2002;Kupesicetal.,2003;Popovic-Todorovicetal.,2003),havealsobeenassessed,buttheresultsarenotconclusive.Despitetheextensivestudiesindicatedabove,thereisnoreportintheliteraturetoassessthepredictivevalueoffolliclesizeinovarianresponsiveness.Haadsmaandcolleaguesreportedthatthenumberofsmallantralfollicles(2-6mm)isdeclinedwithage,whilethenumberoflargerfollicles(7-10mm)remainsconstant(HaadsmaM.Letal.,2007).Interestingly,after19daysofpituitaryconsumptionofendogenousFSHandluteinizinghormone(LH)bygonadotropin-releasinghormone(GnRH)89 华中科技大学博士学位论文agonist,higherpregnancyratewasonlynotedinthosepatientswithhighernumberoffolliclesbetween5and10mminsize,andpatientswithantralfolliclesdominantlybiggerthan11mminsizedisplayedhighercancellationrateduetoovarianlowresponse.Giventhatsmallantralfolliclesaresignificantlyrelatedtoage,AFCandvariousendocrineparameters,thenumberofsmallantralfolliclescouldrepresentthecapacityoffunctionalovarianreserve.Bykeepingthisinmind,wethusconductedthecurrentprospectivestudytoassesswhetherantralfolliclesizeisavaluablepredictivemarkerforovarianresponseandsuccessfulpregnancyrateinwomenundergoingIVFtreatment.MaterialsandMethodsSubjectsAtotalof214patientsand214cyclesbetweenJanuary2008andJuly2012wereincludedforthestudy.Allpatientshadregularmenstrualcycles(25-32days),andwereundergoingtreatmentforinfertilityduetotubal,endometriosis,male,unexplainedormixedfactors.Inaddition,allpatientsdidnotprocessovarianstimulation3monthspriortothiscycleanddidnotreceiveoralcontraceptivepill(OCP)pretreatmentbeforethiscycle.ConsentformswereobtainedfromallsubjectsandthestudieswereapprovedbytheTongjiHospitalHumanAssuranceCommittee.ProceduresforcontrolledovarianstimulationOnday3ofmenstrualperiod,abasicevaluationwasconductedbyultrasoundexamination.MedicationwastheninitiatedwithrecombinantFSH(rFSH)(Gonal-F,Serono)atthedayofultrasoundexamination,inwhichyoungerpatients(<35yearsold)tooktwoampoules(150IU)ofGonal-Fdaily,andolderpatients(≥35yearsold)tookthreeampoules(225IU)ofGonal-Fdaily.Thedosewasfixedforthefirst5daysofstimulation.After5consecutivedaysofmedication,transvaginalBultrasoundexaminationwasthenperformedtomonitorthedevelopmentoffollicles,andthedoseofrFSHwasoptimallyadjustedbasedonthenumberandsizeofdevelopingfollicles.90 华中科技大学博士学位论文TheGnRHantagonist,cetrorelix,wasnextadministereddailybys.c.injection(0.25mg/d)inthemorning(8:00-12:00AM)fromday6ofthestimulationcycletothedayofhumanchorionicgonadotropin(HCG)administration.AdditionaltransvaginalBultrasoundexaminationswerealsoperformedpostdays8,10and12ofmedication.OocyteretrievalproceduresGonal-Fandcetrorelixwereadministeredcontinuouslyuntilthreefolliclesreached≥17mm.HCG(10,000IU,Serono)wasthenadministrated,andserumconcentrationsforestradiol(E2),LH,andprogestone(P)weretestedonthedayofHCGadministration.ThehormonesweredeterminedbyanImmuliteAutomatedAnalyserSystem(ECL2012,Siemens,Germany)asinstructed.Oocyteswereretrieved34-38hafterinjectionofHCGandwerefertilizedinvitroaccordingtothestandardprocedures.Embryotransfer(ET)wascarriedout72hafteroocyteretrieval.Amaximumofthreeembryosweretransferredintoeachpatient.Progesterone(inoil)wasi.m.administereddaily(80mg/day,fromday1postoocyteretrieval)tomaintainlutealfunctionality.Clinicalpregnancywasdefinedaselevatedserumβ-hCG14daysafterETandthepresenceofgestationalsac(s)byultrasonography.StatisticalanalysisTheSPSS17.0forwindowswasusedforstatisticalanalysis.Basalantralfollicularsizeandcountweremeasuredbyvaginalultrasonographyonmenstrualday3priortogonadotropin(Gonal-F)administration.Antralfolliclewasdefinedasdiameter2-10mmfollicle.Thefollicularsizewasgivenasthemeanofthetwolargestdiametersinthesameplaneandperpendiculartoeachother(Wiklandetal.,1994).Normally,wemeasured5-8antralfolliclesforeachpatient.Dataonage,bodymassindex,basalFSHconcentration,thenumberandsizeofantralfollicles,theduration/dosageofrFSHused,serumconcentrationsforoestradiol,LHandprogesterone,thenumberoffollicleswithsize≥14mm,theendometrial91 华中科技大学博士学位论文thicknessonthedayofHCGadministration,thenumberofoocytesaspirated/fertilizedwererecorded.Thepatientswerealsofollowedforthenumberofgoodqualityembryosandtheratesoffertilization,implantationandthepresenceofclinicalpregnancy.Allvalueswereexpressedasmean±SDornumber(percentage).Continuousdatawerelogtransformedtocorrectforskewnesspriortofurtherstatisticalanalysis.Analysisofvariance(ANOVA)withmultiplecomparisons(Tukey2HSD)wasemployedforcontinuousdata,andorFisher’sexacttestswereusedforanalysisofcategoricaldata.CorrelationwasassessedbythePearsonmethodtoevaluatethepredictivevaluesofantralfolliclesizeontheIVFoutcome.Multiplelogisticregressionanalysiswasusedtoassesstheabilityofeachviableonthepredictionofovarianresponse.Inallcases,ap-value(two-tailed)<0.05wasconsideredwithstatisticalsignificance.ResultsDemographicinformationandclinicalcharacteristicsAlldemographicinformationandclinicalcharacteristicsforthestudyingsubjectsaresummarizedinTable1.Theageforthepatientsrangesfrom22to42yroldwithanaverageageof34.5±5.0yr.Amongall214subjectsrecruited,87ofwhich(40.7%)werepresentwithprimaryinfertility,andtherest127women(59.3%)wereassociatedwithsecondaryinfertility.Themeandurationofinfertilitywas7.74±5.3yr,rangingfrom3to11yr.Thelowestbodymassindex(BMI)wasnotedonly17.02,whilethehighestonereached27.02(average21.6±2.59).Inaverage,thenumberforantralfolliclewas8.3±4.5,andthesizeforantralfolliclewas6.9±1.5mm.TheaveragerFSHstimulationdayswere7.9±1.6day,andtheaveragerFSHampoulesusedforeachsubjectwere25.8±10.7.Ingeneral,oestradiollevelsonthedayofHCGadministrationwere1420.8±975.1pmol/l,and7.5±6.5oocyteswereretrievedfromeachsubject.92 华中科技大学博士学位论文TheinfertilecausativefactorsdonotaffectantralfolliclesizeAmongallsubjectsincludedinthecurrentstudy,108womenwerediagnosedwithtubalproblems,10withendometriosis,33withmalefactors,and44withmixedcauses.However,19ofwhichfailedtoreachaconfirmativediagnosis,andtherefore,theyweredefinedtothecategorywithunexplainedfactors.Surprisingly,wefailedtoobserveasignificantdifferencefortheaveragesizeofantralfolliclebetweenwomenwithdifferentcausesofinfertility(Table2),indicatingthatthecausativefactorsforinfertilitydonothaveaperceptibleimpactonantralfolliclesize.CorrelationanalysisoffactorsrelevanttoovarianresponseGiventhatantralfolliclecount(AFC)hasbeensuggestedtobeapivotalmarkerinpredictingovarianresponse,wethusfirstconductedanalysestoevaluateitscorrelationwithotherfactorsrelevanttoovarianresponse.Inconsistentwithpreviousreport,ourdatasetrevealedthatAFCwassignificantlycorrelatedwithpatientage,basalFSHconcentration,rFSHdosage,serumoestadiolandprogesteroneconcentration,follicleswithsize≥14mm,endometrialthicknessonthedayofHCGadministration,andthepotentialnumberofoocytestobeobtained.Together,theseresultsprovidedstrongevidencesupportingthatAFCisavaluablemarkerinpredictingovarianresponse(Table3).Wenextsoughttoaddressthecorrelationbetweenantralfolliclesizeandtheaboveindicatedfactorsrelatedtoovarianresponse.ItwasinterestinglynotedthatunlikeAFC,antralfolliclesizewasonlynegativelycorrelatedwiththedosage(R=-0.493)andduration(R=-0.465)ofrFSH(Table3).Thisisaveryexcitingdiscoverydemonstratingthatpatientswithsmallerbasalsizeforantralfolliclesonmenstrualday3wouldneedhigherrFSHdosageandlongerdurationduringIVFtreatment.Indeed,asimulationanalysisofourretrospectivedatafurtherconfirmedthistrend(datanotshown).Collectively,ourdatasuggestthatantralfolliclesizecouldbeanadditionalpredictivemarkerinovarianresponseotherthanAFC.93 华中科技大学博士学位论文MultipleregressionanalysisoffactorsinpredictingthenumberofoocytestobeobtainedTocharacterizethefactorswithpredictivevaluesforthenumberofoocytestobeobtained,AFC,antralfolliclesize,basalFSHconcentration,BMI,andtheageofpatientswereenteredinastepwisefashioninthemultipleregressionanalysis,inwhichthenumberofoocytesobtainedwasdefinedasadependentvariable.Itwas2notedthatAFCyieldedthehighestcoefficientdetermination(R=0.358)andpositiveregressioncoefficient(B=0.823)(Table4),indicatingthatAFCprovidesthebestpredictivevalueamongalloftheseparametersexamined.Indeed,inourclinicalpracticebasalAFCwasfoundwithhighpotencytopredicttheexactrangeforthenumberofoocytestobeobtainedforaparticularpatient.Importantly,ourdataalsoprovidedalogicalexplanationwhypoorrespondersalwayshavehighcyclecancellationrate,andwhypatientswithpolycysticovarysyndrome(PCOS)arealwaysmoresusceptibletothedevelopmentofovarianhyperstimulationsyndrome(OHSS).Excitingly,antralfolliclesizeonceagainwasfoundwithhighernegativeregressioncoefficient(B=-0.661)ascomparedwiththatofbasalFSHconcentration(B=-0.326)andBMI(B=-0.281)(Table4).TheseresultsprovidedadditionalevidencesupportingthatantralfolliclesizecouldbeabetterpredictivemarkerinthesettingofIVFtreatmentascomparedwiththatofbasalFSHconcentrationandBMI.Surprisingly,wefailedtodetectadiscernableimpactforthepatientageonthenumberofoocytestobeobtained.TheimpactofantralfolliclesizeonIVF-EToutcomesInviewoftheabovefindings,wenextperformedanalysistoassesstheimpactofantralfolliclesizeonIVF-EToutcomes.Forthispurpose,thesubjectswererecategorizedintofourgroupsaccordingtotheirantralfolliclesizeonmenstrualcycleday3,whichincludepatientswithantralfolliclesizes2-6mm,6-7mm,7-8mmand8-10mm,respectively.Wefailedtodetectasignificantdifferencebetweenpatientsin94 华中科技大学博士学位论文differentgroupsfortheage,BMI,basalFSHconcentration,HCGdayprogesteronelevelsandendometrialthickness(Table5).Remarkably,patientswithantralfolliclesize6-7mmshowedsignificantlyhigherAFC,oocytesretrieved,fertilizedoocytes,andgradeI/IIembryosascomparedwiththatofpatientsintheother3groups(Table5).Althoughwedidnotdetectanobviousdifferencefortheimplantationrateandclinicalpregnancyrateamonggroups,thetransfercyclecancellationrateinpatientswithantralfolliclesize6-7mm(7.5%)wassignificantlylowerthanthatintheotherthreegroups(16-17%)(Table5).Analysisoftheoutcomesinpatientswithantralfolliclesize8-10mmrevealedasignificantreductionfortherFSHdurationanddosage.However,numbersfortheoocytesobtained,fertilizedoocytes,andgradeI/IIembryosweresignificantlyreducedaswell.Furthermore,theimplantationrateandclinicalpregnancyratewerealsofoundlowerthanthepatientsinothergroups(Table5).Incontrast,wefailedtoobserveasignificantdifferencefortheoocytesobtained,fertilizedoocytes,gradeI/IIembryos,implantationrateandclinicalpregnancyratebetweenpatientswithantralfolliclesize2-6mmandpatientswithantralfolliclesize>7mm(the7-8mmand8-10mmgroup),butpatientswithantralfolliclesize2-6mmshowedsignificantlyhigherrFSHdosageandlongerrFSHduration(Table5).Takentogether,ourdataindicatethatantralfolliclesizeimpactstheoutcomesforIVF-ETtreatment,andpatientswithantralfolliclesize6-7mmarelikelypredisposedtothebestclinicaloutcomes.DiscussionPredictionofovarianresponsespriortostimulationisusefulinpatientcounseling,whichmayalsobehelpfulintailoringthedosageforgonadotrophinineachindividualpatient.Therefore,manyclinical,hormonalandultrasoundparametershavebeenextensivelyevaluatedfortheirfeasibilitytopredictovarianresponsesduringovarianstimulation.IthasnowgenerallyacceptedthatAFCisthusfarthe95 华中科技大学博士学位论文mostpowerfulmarkerforpredictingovarianresponse(Tomasetal.,1997;Changetal.,1998).Indeed,Changandcollegesfoundthatpatientswithantralfolliclenumber≤3hadasignificantlyhigherrateofcyclecancellationandhigherHMGdosageascomparedwiththosepatientswithantralfolliclenumber4-10or≥10.Similarly,Tomasandco-workersdemonstratedthatAFCisstronglycorrelatedwiththenumberofoocytestoberetrievedbeforeovarianstimulation.TheyconcludedthatAFCisabetterpredictorforovarianresponsethanthatofovarianvolumeoragealone.Ofnote,ourprospectivestudiesinthecurrentreportalsofoundthatAFConmenstrualcycleday3issignificantlycorrelatedwithpatientage,basalFSHconcentration,rFSHdosage,serumoestadiolandprogesteroneconcentration,antralfolliclesize≥14mm,endometrialthicknessonthedayofHCGadministrationandthenumberofoocytestobeobtained.ThesedataprovidedadditionalevidencesupportingthatAFCisavaluablemarkerinovarianresponseprediction.OtherthanAFC,antralfolliclesizemayalsoimpactthecapacityofovarianresponse.Indeed,Haadsmaandcollegesfoundthattheendocrinefunctionofanantralfollicleisrelatedtoitssize.Aprospectivestudyincluding474womenfurtherrevealedthatthenumberofsmallfollicles(2–6mm)inaparticularwomanmayrepresentherfunctionalovarianreserve.Pohlandco-workersstudied113patients,inwhichtheyhavecategorizedthepatientsbasedonantralfolliclesizeinto<5mm,5-10mm,11-20mmand>20mmfourgroupstoassesstheimpactofantralfolliclesizeonIVFoutcomes.Theyhaveinterestinglynotedthatpatientswithantralfolliclesize5-10mmdisplayedasignificantlyhigherpregnancyrate,whereasthosepatientswithantralfolliclessize>11mmpresentedahighercancellationrateduetolowovarianresponse.Together,thesestudiesprovidedsuggestiveevidencesupportingaroleforantralfolliclesizeinovarianresponseprediction,whiletheexactimpactofantralfolliclesizeonovarianresponseandIVFoutcomeisyettobefullyassessed.Inthecurrentreport,wehavesystematicallyassessedthefeasibilityandpotencyofantralfolliclesizeinovarianresponsepredictionandIVFoutcome.Byprospectivelystudyof214womenundergoingIVF-ETtreatment,wenotedthatantral96 华中科技大学博士学位论文folliclesizeissignificantlycorrelatedwiththedosageanddurationofrFSHstimulation.Particularly,higherdosageandlongerdurationofrFSHstimulationwererequiredforthosewomenwithantralfolliclesize2-6mm.However,thedosageanddurationofrFSHstimulationweresignificantlyreducedforthosewomenwithantralfolliclesize8-10mm,buttheimplantationrateandclinicalpregnancyratewerereduced,andlowerestradiollevelswerenoticedinthesewomenaswell.Insharpcontrast,womenwithantralfolliclesize6-7mmshowedmuchmoreAFC,higherpotencyforoocytestoberetrievedalongwithhigherfertilizedoocytesandgradeI/IIembryos.Importantly,althoughwefailedtocharacterizeanobviousdifferencefortheimplantationrateandclinicalpregnancyratebetweenthesewomenandthosewithantralfolliclesize<6mmor>8mm,thetransfercyclecancellationratewassignificantlylowerinwomenwithantralfolliclesize6-7mm(7.5%vs.16-17%).Alltogether,ourdatasupportthatthebasalantralfolliclesize6-7mmcanproducethebestpredictivevalueonclinicalandlaboratoryresultsduringIVF-ETprocesses.Ithasbeensuggestedthatsubjectageisalsooneoftheimportantpredictiveparametersinovarianresponse.Indeed,womanfertilityisremarkablyreducedalongwiththeincreaseofageinbothspontaneousconceptions(Tietze,1957)andassistedreproductivemethods(Pearlstoneetal.,1992;Templetonetal.,1996).Thedecreaseoffertilitywithagingisapparentlyduetothedecreaseforthenumberofprimordialfollicles(Block,1952)andantralfollicleswithsize≥2mm(Reussetal.,1996).Inourdataset,subjectagewasfoundtobecorrelatedwithAFC.Unexpectedly,subjectagewasexcludedfromourmultipleregressionanalysis,suggestingthatagealonehaslimitedvalueinpredictingovarianresponseforwomenundergoingassistedreproductionmethods.BasalFSHconcentrationhasbeenshowntobeabetterpredictorofovarianresponsetostimulationthantheageofwomen(Cahilletal.,1994;Sharifetal.,1998).ItisgenerallysuggestedthathigherbasalFSHconcentrationwouldimpairovarianresponse.Inlinewiththisnotion,ourdataalsorevealedthatbasalFSHconcentration97 华中科技大学博士学位论文wasnegativelycorrelatedwithAFCandthenumberofoocytestobeobtained.Similarly,BMIisconsideredtobeanotherpredictivemarkerforovarianresponse.OurstudiesdemonstratethathigherBMIimpairsthenumberofoocytestobeobtained.Inconsistentwithourresults,obesepatientsaregenerallythoughttorequiresignificantlyhigherdoseofgonadotropininalongerdurationofstimulation,andtherefore,theyshouldbeadvisedtoreduceweightbeforeattemptingIVFtreatment(Dechaudetal.,1998).Ofimportantlynote,ourdatasuggestthatantralfolliclesizepossesseshigherpredictivepowerthanthatofbasalFSHconcentrationandBMI,asmanifestedbythatantralfolliclesizehassignificantlyhighernegativeregressioncoefficient.Giventhefactthatveryfewstudieshavebeenconductedrelevanttothistopic,thepredictiveimportanceofantralfolliclesizeinclinicalIVF-ETtreatmentisgenerallyneglected.Insummary,wehavedemonstratedevidencesupportingthatantralfolliclesizeisprobablyanewandvaluedpredictivemarkerforovarianresponseduringIVF-ETtreatment.Antralfolliclesize6-7mmcanachievethebestpredictivevalue.However,additionalperspectivestudieswithlargersamplesizewouldbenecessarytofurthervalidatetheconclusion.AcknowledgementsWewouldliketothankallstaffattheReproductiveMedicalCenterofTongjiHospitalfortheirhelptocollecttheinformationofourstudyingsubjects.ThisworkwassupportedbythePh.D.ResearchProjectInnovativeAwardfromHuazhongUniversityofScienceandTechnology.ReferencesAboulgharMA,MansourRT.Ovarianhyperstimulationsyndrome:classificationsandcriticalanalysisofpreventivemeasures.HumReprodUpdate.2003May-Jun;9(3):275-89.Review.98 华中科技大学博士学位论文BlockE.Quantitativemorphologicalinvestigationsofthefollicularsysteminwomen;variationsatdifferentages.ActaAnat(Basel).1952;14(1-2):108-23.ChangMY,ChiangCH,HsiehTT,SoongYK,HsuKH.Useoftheantralfolliclecounttopredicttheoutcomeofassistedreproductivetechnologies.FertilSteril.1998Mar;69(3):505-10CahillDJ,ProsserCJ,WardlePG,FordWC,HullMG.Relativeinfluenceofserumfolliclestimulatinghormone,ageandotherfactorsonovarianresponsetogonadotrophinstimulation.BrJObstetGynaecol.1994Nov;101(11):999-1002.DéchaudH,FerronG,AnahoryT,ArnalF,HumeauC,HédonB.Obesityandassistedreproductiontechniques.ContraceptFertilSex.1998Jul-Aug;26(7-8):564-7.DzikA,Lambert-MesserlianG,IzzoVM,SoaresJB,PinottiJA,SeiferDB.InhibinBresponsetoEFORTisassociatedwiththeoutcomeofoocyteretrievalinthesubsequentinvitrofertilizationcycle.FertilSteril.2000Dec;74(6):1114-7EngmannL,SladkeviciusP,AgrawalR,BekirJS,CampbellS,TanSL.Valueofovarianstromalbloodflowvelocitymeasurementafterpituitarysuppressioninthepredictionofovarianresponsivenessandoutcomeofinvitrofertilizationtreatment.FertilSteril.1999Jan;71(1):22-9FanchinR,SchonäuerLM,RighiniC,GuibourdencheJ,FrydmanR,TaiebJ.Serumanti-MüllerianhormoneismorestronglyrelatedtoovarianfollicularstatusthanseruminhibinB,estradiol,FSHandLHonday3.HumReprod.2003Feb;18(2):323-7.99 华中科技大学博士学位论文HaadsmaML,BukmanA,GroenH,RoeloffzenEM,GroenewoudER,HeinemanMJ,HoekA.Thenumberofsmallantralfollicles(2-6mm)determinestheoutcomeofendocrineovarianreservetestsinasubfertilepopulation.HumReprod.2007Jul;22(7):1925-31.Epub2007Apr16.HofmannGE,ScottRTJr,HorowitzGM,ThieJ,NavotD.Evaluationofthereproductiveperformanceofwomenwithelevatedday10progesteronelevelsduringovarianreservescreening.FertilSteril.1995May;63(5):979-83.Keay,S.D.,Liversedge,N.H.,Mathur,R.S.etal.(1997)Assistedconceptionfollowingpoorovarianresponsetogonadotrophinstimulation.BrJObstetGynaecol.,104,521-527.KupesicS,KurjakAPredictorsofIVFoutcomebythree-dimensionalultrasound.HumReprod.2002Apr;17(4):950-5.KupesicS,KurjakA,BjelosD,VujisicS.Three-dimensionalultrasonographicovarianmeasurementsandinvitrofertilizationoutcomearerelatedtoage.FertilSteril.2003Jan;79(1):190-7.LassA,SkullJ,McVeighE,MargaraR,WinstonRM.Measurementofovarianvolumebytransvaginalsonographybeforeovulationinductionwithhumanmenopausalgonadotrophinforin-vitrofertilizationcanpredictpoorresponse.HumReprod.1997Feb;12(2):294-7.MPohl,MHohlagschwandtner,AObruca,GPoschalko,MWeigert,andWFeichtinger.Numberandsizeofantralfolliclesaspredictivefactorsinvitrofertilizationandembryotransfer.JAssistReprodGenet,July1,2000;17(6):315-8.100 华中科技大学博士学位论文NgEHY,TangOSandHoPC(2000b)thesignificanceofthenumberofantralfolliclespriortostimulationinpredictingovarianresponsesinin-vitrofertilizationprogrammes.HumReprod15,1937–1942.PadillaSL,GarciaJE.Effectofmaternalageandnumberofinvitrofertilizationproceduresonpregnancyoutcome.FertilSteril.1989Aug;52(2):270-3PearlstoneAC,OeiML,WuTC.Thepredictivevalueofasingle,earlyhumanchorionicgonadotropinmeasurementandtheinfluenceofmaternalageonpregnancyoutcomeinaninfertilepopulation.FertilSteril.1992Feb;57(2):302-4Popovic-TodorovicB,LoftA,LindhardA,BangsbøllS,AnderssonAM,AndersenAN.Aprospectivestudyofpredictivefactorsofovarianresponsein'standard'IVF/ICSIpatientstreatedwithrecombinantFSH.AsuggestionforarecombinantFSHdosagenormogram.HumReprod.2003Apr;18(4):781-7.ReussML,KoltonS,TharakanT.Transvaginalultrasonographyingynecologicofficepractice:assessmentin663premenopausalwomen.AmJObstetGynecol.1996Nov;175(5):1189-94.SeiferDB,MacLaughlinDT,ChristianBP,FengB,SheldenRM.Earlyfollicularserummüllerian-inhibitingsubstancelevelsareassociatedwithovarianresponseduringassistedreproductivetechnologycycles.FertilSteril.2002Mar;77(3):468-71SeiferDB,Lambert-MesserlianG,HoganJW,GardinerAC,BlazarAS,BerkCA.Day3seruminhibin-Bispredictiveofassistedreproductivetechnologiesoutcome.FertilSteril.1997Jan;67(1):110-4.ShararaFI,ScottRTJr,SeiferDB.Thedetectionofdiminishedovarianreservein101 华中科技大学博士学位论文infertilewomen.AmJObstetGynecol.1998Sep;179(3Pt1):804-12.Review.SharifK,ElgendyM,LashenH,AfnanM.Ageandbasalfolliclestimulatinghormoneaspredictorsofinvitrofertilisationoutcome.BrJObstetGynaecol.1998Jan;105(1):107-12SyropCH,DawsonJD,HusmanKJ,SparksAE,VanVoorhisBJ.Ovarianvolumemaypredictassistedreproductiveoutcomesbetterthanfolliclestimulatinghormoneconcentrationonday3.HumReprod.1999Jul;14(7):1752-6SyropCH,WillhoiteA,VanVoorhisBJ.Ovarianvolume:anoveloutcomepredictorforassistedreproduction.FertilSteril.1995Dec;64(6):1167-71.TarlatzisBC,ZepiridisL,GrimbizisG,BontisJ.ClinicalmanagementoflowovarianresponsetostimulationforIVF:asystematicreview.HumReprodUpdate.2003Jan-Feb;9(1):61-76.Review.TempletonA,MorrisJK,ParslowW.Factorsthataffectoutcomeofin-vitrofertilisationtreatment.Lancet.1996Nov23;348(9039):1402-6.TinkanenH,BläuerM,LaippalaP,TuohimaaP,KujansuuEPrognosticfactorsincontrolledovarianhyperstimulation.FertilSteril.1999Nov;72(5):932-6TomasC,Nuojua-HuttunenS,MartikainenH.Pretreatmenttransvaginalultrasoundexaminationpredictsovarianresponsivenesstogonadotrophinsinin-vitrofertilization.HumReprod.1997Feb;12(2):220-3..TietzeC.ReproductivespanandrateofreproductionamongHutteritewomen.FertilSteril.1957Jan-Feb;8(1):89-97.Noabstractavailable.102 华中科技大学博士学位论文vanRooijIA,BroekmansFJ,teVeldeER,FauserBC,BancsiLF,deJongFH,ThemmenAP.Serumanti-Müllerianhormonelevels:anovelmeasureofovarianreserve.HumReprod.2002Dec;17(12):3065-71WiklandM,BorgJ,HambergerL,SvalanderP.SimplificationofIVF:minimalmonitoringandtheuseofsubcutaneoushighlypurifiedFSHadministrationforovulationinduction.HumReprod.1994Aug;9(8):1430-6.ZaidiJ,BarberJ,Kyei-MensahA,BekirJ,CampbellS,TanSL.Relationshipofovarianstromalbloodflowatthebaselineultrasoundscantosubsequentfollicularresponseinaninvitrofertilizationprogram.ObstetGynecol.1996Nov;88(5):779-84.103 华中科技大学博士学位论文Table1.BasicclinicalinformationofthesubjectsParametersaverageNo.ofpatients214Age(years)34.5±5.02BMI(kg/m)21.6±2.59BasalFSHconcentration(IU/l)7.2±2.9Durationofinfertility(years)7.74±5.30Infertility(%)Primary87(40.7)Secondary127(59.3)Sizeofantralfollicle(mm)6.9±1.5No.ofantralfollicle8.3±4.5DaysofrFSH7.9±1.6AmpoulesofrFSH25.8±10.7OestradiolonHCGday(pmol/l)1420.8±975.1No.ofoocytesretrieved7.5±6.5BMI=bodymassindexFSH=folliclestimulatinghormoneHCG=humanchorionicgonadotropinTable2.ComparetheantralfolliclesizeindifferentcausesofinfertilityCausesofinfertilityNo.ofwomenaverageTubalfactor1086.8±1.6Endometriosis106.9±0.8Malefactor337.2±1.5Unexplained197.3±1.2Mixed446.9±1.4Total2146.9±1.5104 华中科技大学博士学位论文Table3.Correlationcoefficientsofantralfolliclecountandsizewithsomeparametersofovarianresponses(afterlogtransformationofdata)antralfolliclesizeantralfolliclecountbAge(years)-0.013-0.304BMI-0.114-0.012bBasalFSHconcentration(IU/l)-0.119-0.228No.oftotalantralfollicle0.007-Antralfolliclesize(mm)-0.007brFSHduration(days)-0.4930.059bbrFSHdosage(ampoules)-0.465-0.280DayofHCGbFollicles≥14mm-0.0990.646bOestradiol(pmol/l)-0.1280.519bProgesterone(ng/ml)-0.0950.213aEndometrialthickness(mm)0.0490.152bNo.ofoocytesretrieved-0.1190.599aP<0.05.bP<0.01.BMI=bodymassindex.FSH=folliclestimulatinghormoneHCG=humanchorionicgonadotropinTable4.Multipleregressionanalysisevaluatingthevaluesofdifferentparametersinpredictingthenumberofoocytesobtained(afterlogtransformationofdata)2B(95%CI)βRchangP-valueConstant13.692---(6.159,21.226)Antralfolliclecount0.8230.5650.3580.000(0.668,0.978)Antralfolliclesize-0.661-0.1520.0150.005(-1.118,-0.205)BasalFSHconcentration-0.326-0.1450.0170.008(-0.567,-0.084)BMI-0.281-0.1120.0120.036(-0.544,-0.018)105 华中科技大学博士学位论文Ageexcludedintheequation:Log(totaleggno.)=13.692+0.823×log(antralfollicleno.)–0.661(antralfolliclesize)–0.326(D3FSH)–0.281(BMI).R=0.635.2AdjustR=0.392.CI=confidenceinterval;FSH=folliclestimulatinghormone;BMI=bodymassindex.Table5.ComparetheIVF-EToutcomesoffourgroupsofantralfolliclesize2-6mm6-7mm7-8mm8-10mmCycles52536148Age(years)34.6±5.534.7±4.734.4±4.634.5±5.2BMI21.6±2.422.1±2.721.7±2.621.2±2.5BasalFSH(IU/l)7.6±2.27.7±3.66.7±3.07.0±2.5bbbantralfolliclecount6.8±3.710.2±5.28.9±4.57.1±3.3rFSHduration(days)8.8±1.68.5±1.67.5±1.26.9±1.2bbbrFSHdosage(amp)31.7±12.127.8±11.123.7±8.719.9±6.3HCGdayOestradiol1513.4±1700.2±1323.0±1149.9±aa(pmol/l)1247.91154.6705.8590.2Progesterone(ng//ml)1.3±0.51.4±0.81.2±0.51.2±0.6Endometrial9.9±2.210.7±2.211.1±2.79.9±2.1thickness(mm)aa≥14mmfollicles5.7±3.86.6±3.85.3±2.84.4±3.1bboocytesretrieved7.2±5.810.3±9.47.1±4.75.3±4.1aba,bfertilizedoocytes4.5±4.16.0±5.73.9±2.82.6±2.4Fertilizationrate(%)245/391320/543249/455132/266(62.7)(58.9)(54.7)(49.6)bbEmbryostransferredno.1.9±1.02.3±0.91.8±1.01.7±1.0bbGradeI/IIembryos2.4±2.63.1±3.11.8±2.01.4±1.6Transfercycles9(17.3)4(7.5)10(16.4)8(16.7)cancelled(%)Implantationrate(%)18/101(17.8)24/121(19.8)27/117(23.1)11/83(13.3)Clinicalpregnancy(%)14/45(33.3)16/49(32.7)21/54(38.9)8/42(19.0)abP<0.05;P<0.01;BMI=bodymassindex.FSH=folliclestimulatinghormone.106 华中科技大学博士学位论文致谢我要以最诚挚的心意感谢我的导师王从义教授，感谢他对我这三年来的无私培养和付出。三年前，我仅仅是一名临床医生，对科研心存敬畏但知之甚少，是王老师将我领上科研之路并给我提供了到美国学习和培养的机会。在美国学习期间，王老师为我提供了优越的实验条件和自由宽松的学习氛围，使我在实验技术和科学理论方面都得到了很大提高。并且，每当我遇到困难，沮丧、迷茫的时候，王老师总是不断地帮我分析原因，给我鼓励，在王老师的支持和鼓励下，我才一步步走到今天。在这里我还要特别感谢我的师母杨萍老师，是她帮我打理好在美国生活学习所需的一切；是她手把手教会我各项实验技术；是她在我高兴和伤心时听我诉说；是她在异国他乡还惦记着我的生日；是她为我们准备满桌的饭菜还帮我照料孩子。回忆两年的点点滴滴真是太多太多，杨老师无微不至的关照，使我心中倍感温暖并永生难忘。还要特别感谢王老师团队的张述老师，当我实验遇到困难向他求助时，他总是在百忙之中及时、耐心、详细地给我分析和讲解，并将自己的经验和教训毫无保留地告诉我；并且在帮我精心修改论文的过程中，诚挚地提出了很多有益的建议和意见，可以说我的论文里也凝聚了张老师辛勤的汗水；还要感谢余其林老师、熊飞老师和向科老师，谢谢你们给我提供的各项帮助，使我在美国能够安心地工作并顺利完成博士学业。还要感谢在美国实验室同甘共苦的李金秀老师和杨飞博士，很高兴与你们成为朋友并互帮互助度过难忘的日子。还要感谢在美国结识的朋友们：毛锐、夏文芳、程波、张明生、周彬、庞俊峰、滕勇、陈峰、潘国栋、靳玉兰、白山、龚士强等，谢谢你们给我的工作和生活提供的各种无私的帮助并与我分享生活的喜怒哀乐。还要感谢我的爱人，谢谢他对我学业的帮助、体谅和支持，使我在异国他乡能够安心、顺利地工作学习；当我生活中遇到麻烦和困难时，他是我的坚强后盾，给我出谋划策，鼓励支持我；在我的论文写作期间，他承担起家里的所有事务，并且帮助我认真地修改论文。还有我的女儿胡诗尧小朋友，谢谢你在美国寂寞枯107 华中科技大学博士学位论文燥的日子里对妈妈的陪伴和支持。最后，还要感谢我的父母，当我远在异国他乡时，是他们帮助我照料留在国内的孩子；后来，又拖着年迈的身躯远涉重洋，辗转往返。为了论文和实验，我近几年都没能照料和问候年过半百的他们，还给他们增添了不少麻烦，心中甚感愧疚。在此，祝愿他们身体健康！108