【作者】 周燕虹；

【导师】 曹佳；

【作者基本信息】 第三军医大学 ， 卫生毒理学， 2009， 博士

【摘要】 研究背景与目的结直肠癌(CRC)是最常见的消化道肿瘤之一,近年来随着生活水平的不断提高,饮食习惯的改变以及工业化进程的加快,我国结直肠癌发病率呈逐年上升趋势,统计数据表明结直肠癌在中国的发病已经进入快速增长期。现代医学研究认为CRC是机体遗传因素与环境影响因素共同作用的一个多基因参与、多步骤、多阶段发展的结果。结直肠癌防治效果的关键在于早期预防及对结直肠癌患者的早期诊断。对结直肠癌遗传易感基因的筛查及鉴定不仅有利于丰富对结直肠癌发生机制的深入认识,更有利于对结直肠癌高危易感人群的大规模筛选及预测,帮助早期发现结直肠癌患者,指导高危人群进行预防,也有利于指导结直肠癌患者进行个性化治疗。流行病学和体内外功能研究证实雌激素、雌激素代谢酶和雌激素受体在结直肠癌的发生中发挥重要作用,提示雌激素受体及代谢酶基因的某些基因型有可能成为结直肠癌的侯选遗传易感基因。多项研究发现ERα及ERβ在CRC中表达降低,而这种低表达可能在APC依赖的微卫星不稳定性结直肠癌的发生中起重要作用。体外实验证明ERβ可通过增强p53信号通路促进结直肠癌细胞凋亡及抑制其增殖。最近多中心的队列结果研究显示雌激素替代治疗对结直肠癌的发生具有保护作用,雌激素也是随机对照研究证实的对结直肠癌及腺瘤发生具有明确预防作用的四类化学制剂之一。而体内的雌激素水平与雌激素合成及代谢酶密切相关。我们推测,雌激素受体及雌激素合成代谢酶基因多态性改变可能会影响到CRC的易感性。为了证实上述假说,我们采用基于侯选基因的关联研究策略及病例-对照设计,选取雌激素受体及代谢酶基因作为侯选基因,系统地研究了这些基因的多态性与CRC易感性的关联,并对有意义的阳性关联的多态位点进行初步的功能研究。材料与方法1.关联研究中,研究对象包括433例CRC患者及790例与之年龄、性别、居住地相匹配的非肿瘤对照。病例来源于第三军医大学三所附属医院普通外科住院治疗的结直肠癌患者,对照来源于与病例年龄、居住地、性别相匹配的上述三所医院同期因急性创伤住院的非肿瘤外伤病人。2.采用SNPlex方法对6个侯选基因(包括ERα、ERβ、CYP17、CYP19、HSD17B1、COMT)的14个SNP位点进行基因分型。基因分型数据采集和判读使用Data Collection软件GeneMapper软件完成。3.统计分析:采用非条件Logistic回归计算OR(Odds ratios)值和95% CI(Confidence interval),并对性别、年龄、是否吸烟、是否饮酒等混杂因素进行校正,以评估各基因型及多态性位点在结直肠癌发病中的相对风险度,并对统计结果采用FDR(False discovery rate)法进行多重检验校正。各SNP位点基因型风险主效应检测由SAS 9.0软件完成,等位风险检测(Alle test)和单倍型(haplotype)、双体型(diplotype)风险检测由UNPHASE软件完成。用Haploview4.1构建单倍型,并用UNPHASE软件计算单倍型和双体型疾病的关联;使用MDR及非条件logistic回归分析基因-基因交互作用。使用分层分析及非条件logistic回归分析基因-环境交互作用。以上所有检验均为双侧。4.在SNP功能研究中,首先采用生物信息学软件对阳性关联的基因功能进行预测,然后通过免疫组化检测不同的基因型个体阳性关联基因的蛋白表达情况,对阳性关联SNP位点进行初步研究。结果1.在雌激素受体α的基因多态性与CRC的遗传易感性研究显示各SNP位点的关联分析显示,经FDR多重检验校正后,位于ERα启动子区的rs2071454和第一外显子的rs2077647位点与结直肠癌的易感性显著相关。其中,与rs2071454位点的TT基因型相比,GT基因型及GG+GT基因型均显著增加了CRC的患病风险。rs2071454与年龄、性别、吸烟、饮酒等因素之间具有交互作用。与rs2077647位点的AA基因型相比,AG基因型及AG+GG基因型均显著增加了CRC的患病风险。rs2077647与年龄、性别、吸烟、饮酒等环境因素之间具有交互作用。rs2071454和rs2077647位点之间存在连锁不平衡(LD),位于同一个block内(D’=0.971, LOD=239.5, R2 =0.616)。这两个位点在重庆人群中分别构建出三种单倍型,分别是TA(60.2%)、GG(29.2%)和TG(10.0%)。以T-A单倍型为参考, G-G单倍型携带者患CRC的风险OR=1.231,95%CI=1.018~1.488。以TT-AA双体型为参考, GT-AG双体型携带者患CRC的风险OR=1.592,95%CI=1.198~2.116。2.在ERβ的各SNP基因多态性与结直肠癌的遗传易感性无关联。3.基因-基因交互logistic回归分析显示ERα的rs2077647与ERβ的rs3829768之间有相乘的交互作用。4.在雌激素代谢酶基因多态性与结直肠癌遗传易感性研究显示经FDR多重检验校正后,雌激素合成限速酶HSD17B1的rs676287和CYP19的rs10046位点与结直肠癌易感性显著关联。HSD17B1 rs676387位点: GT vs GG: OR=3.280,95%CI=2.326~4.625; TT vs GG : OR=3.018,95%CI=2.109~4.318; TT+GT vs GG: OR=2.002,95%CI=1.330~3.013。且rs676387与年龄、性别、吸烟及饮酒等因环境因素之间存在交互作用。CYP的rs10046位点: CT vs CC :OR=0.594 ,95%CI=0.442~0.797 ; TT+CT vs CC::OR=0.638,95%CI=0.480~0.849。且rs10046与年龄、性别、吸烟及饮酒存在交互作用。CYP17的rs248658位点和COMT的rs4680与CRC均无显著关联。基因-基因相乘的交互作用存在于下列基因位点之间:(1) CYP17的rs2486758与CYP19的rs10046;(2)HSD17B1的rs676387与CYP19的rs10046; (3)ERα的rs2071454与CYP19的rs10046;(4)ERα的rs2071454和HSD1B1的rs676387;(5)ERα的rs2077647与HSD1B1rs676387;(6) ERαrs2077647与CYP19 rs10046。5.应用生物信息学软件对阳性关联位点进行的功能预测提示:ERα的rs2071454位点T→G变异会引起两个转录因子(RAP1和CAP)结合位点缺失。病例标本免疫组化结果显示ERα在rs2071454三种基因型携带者中的表达呈现TT>GT及GG,结直肠癌组织>远端正常组织的趋势。结论1.本研究首次发现ERα5’端的rs2071454位点和exon1的rs2077647位点与结直肠癌遗传易感性关联;rs2071454与rs2077647位点构建的单倍型和双体型也同样与结直肠癌的易感性关联。rs2071454及rs2077647位点分别与年龄、性别、吸烟、饮酒存在交互作用。2.在我们所研究人群的SNP位点中,没有发现ERβ多态性与结直肠癌的关联;基因-基因交互分析结果显示ERα的rs2077647与ERβ的rs3829768之间有交互作用。3.雌激素合成限速酶HSD17B1的rs676287和CYP的rs10046位点与结直肠癌的易感性显著关联。CYP17的rs248658位点和COMT的rs4680与结直肠癌均无关联。rs676387位点及rs10046位点分别与年龄、性别、吸烟和饮酒之间存在交互作用。4.基因-基因交互作用存在于下列基因位点之间:(1) CYP17的rs2486758与CYP19的rs10046;(2)HSD17B1的rs676387与CYP19的rs10046; (3)ERα的rs2071454与CYP19的rs10046;(4)ERα的rs2071454和HSD1B1的rs676387;(5)ERα的rs2077647与HSD1B1rs676387;(6) ERαrs2077647与CYP19 rs10046。5.对CRC阳性关联rs2071454位点功能的初步研究显示,该位点T→G变异会引起RAP1和CAP两个转录因子结合位点缺失,同时,可能也会影响ERα在大肠组织的表达下调,这个假说需要进一步进行基因功能研究验证。更多还原

【Abstract】 Background & aimsColorectal cancer(CRC) is one of the most frequent malignancys in our country, which is on the rise due to change in diet custom and the progress of industrialization. In shanghai, for example, the incidence of colorectal cancer has doubled since the last 30 years, making colorectal cancer the second most common malignancy. It is now generally accepted that the carcinogenesis and progression of colorectal cancer is the result of aggressive gene-environment interactions, a multi-step process involving effects of multi-genes. The key to improve the result of colorectal cancer management lies in early diagnosis and early treatment. The identification of susceptibility genes contributing to colorectal cancer would help to clarify pathophysiologic mechanisms relevant to carcinogenesis and would assist in predicting individual and population risk of colorectal cancer development.Both experimental and epidemiologic studies have clarified that estrogen and its receptors or metabolic enzymes play important roles in carcinogenesis and development of colorectal cancer, suggesting some genotypes of estrogen reptors or metabolic enzyemes as candidate cancer susceptibility genes. Several studies reports downregulation of ERαand ERβexpression in colorectal cancer, which is also found to have a role in the development of APC dependent MSI colorectal cancer. In vitro, ERβis found to induce the apoptosis of colorectal cancer cells and suppress its proliferation through p53 pathway. In recent years, several lines of epidemiologic, clinical and experimental evidences have been reported showing that estrogen hormones may have a role in preventing the carcinogenesis of colorectal cancer. The role of estrogen in a certain individual is in part determined by different expression pattern of ER subtypes and functional status of estrogen synthetic and metabolic enzymes. It’s our hypothesis that polymorphism of ER or estrogen synthetic and metabolic enzyme genes would have an association with colorectal cancer susceptibility.To validate these hypotheses, We investigated the association of the polymorphisms in estrogen receptor genes and estrogen-metabolizing enzymes genes with susceptibility to CRC in this case-control association study. The bio-function of the positive association SNP was also inspected preliminarily .MethodsThe case-control study included 433 patients with CRC and 790 control subjects. All CRC patients were unrelated ethnic adult Chinese and residents in Chongqing municipality and its surrounding regions, excluding FAP and HNPCC patients, treated at 3 affiliated hospital of the third military medical university. All control subjects were patients which were enrolled due to trauma with no signs of malignancies. At recruitment, informed consent was obtained from each subject, and personal information on demographic factors, medical history, tobacco and alcohol use, and family history of CRC were collected with structured questionnaire.1. Genotyping was performed by SNPlex assay for 5 SNPs in ERαgene, 5 SNPs in ERβ, and rs2486758 in CYP17, rs10046 in CYP19, rs676387 in HSD17B1, rs4680 in COMT. The data was collected with Data Collection and analyzed with Genemapper.2. statisticsThe Hardy-Weinberg equilibrium at each SNP site in control groups was analyzed with the goodness of fitχ2 test. unconditioned- Logistic regression analysis was used for the calculation of Odds ratio and 95% confidence interval, adjusted with sex, ages, and smoking and alcohol status . SAS astatistical package 9.0 and UNPHASE 3.07 was used to conducted the analyses. Analyes included evaluating the distribution of the alleles, genotypes , haplotypes and diplotypes in the population studied the independent associations of genetic polymorphisms with colorectal cancer risk , and the joint effect of genotypes on colorectal cancer risk.We carried out multiple hypothesis testing using the Benjamini-Hochberg method to control the false discovery rate (FDR) in the unconditional logistic regression analysis.Haplotypes were estimated using Haploview4.1. MDR and conditioned-Logestic regression analysis was used for gene-gene interaction analysis. Stratification and conditioned-logistic regression analysis were used for genetic-environmental interaction.All the analyses was two-side test.3. Bioinformatic software was used for prediction of function of positive genes, and IHC was used for protein expression analysis.Result:1. Both association analysis and FDR test confirmed that rs2071454 at the promoter region of ERαand rs2077647 at exon 1 were correlated with risk of CRC. As compared to carriers of TT genotype at rs2071454, carriers of GT or GG had a higher risk of CRC(GT vs TT:OR=1.472,95%CI=1.149~1.885; GG+GT vs TT: OR=1.408,95%CI=1.108~1.789)。There were gene-environment interaction between rs2071454 and age, sex, smoking and drinking.A higher CRC risk was also seen among carriers of AG or GG genotype at rs2077647 as compared to AA carriers,(AG vs AA:OR=1.460,95%CI=1.125~1.895; AG+GG vs AA: OR=1.453,95%CI=1.127~1.873)。There were gene-environment interaction between rs2077647 and age, sex, smoking and drinking. A linkage disequilibrium was found between rs2071454and rs2077647. There was three haplotypes in our subjects as these two SNPs were concerned, TA(60.2%)、GG(29.2%)and TG(10.0%). Carriers of GG haplotype is more likely to be susceptible to CRC(OR=1.231,95%CI=1.018~1.488).2. There was no significant association between 5 SNPs in ERβand CRC susceptibility.3. The gene-gene interaction analysis suggested a Multiplicative effect between rs2077647 in ERαand rs3829768 in ERβ, which was confirmed by logistic regression analysis. (GG+AG)/AA vs AA/AA: OR=1.837,95%CI=1.374~2.455; AA/(GG+AG)vs AA/AA: OR=OR=1.947, 95%CI=1.177~3.219。4. Polymorphisms at rs2486758 in CYP17, rs10046 in CYP19, rs676387 in HSD17B1, rs4680 in COMT was analyzed, rs676287 in HSD17B1 and rs10046 in CYP19 were found to be associated to CRC risk after FDR analysis. For rs676387 in HSD17B1: GT vs GG: OR=3.280,95%CI=2.326~4.625; TT vs GG : OR=3.018,95%CI=2.109~4.318; TT+GT vs GG: OR=2.002,95%CI=1.330~3.013。There were gene-environment interaction between rs676387 and age, sex, smoking and drinking.For rs10046 in CYP19: CT vs CC :OR=0.594 ,95%CI=0.442~0.797 ; TT+CT vs CC::OR=0.638,95%CI=0.480~0.849。There were gene-environment interaction between rs10046 and age, sex, smoking and drinking.There was no association between CRC risk and polymorphisms at rs2486758 in CYP17 or rs4680 in COMT .Gene-gene interactions exist between the following pairs of SNPs: (1) CYP17 rs2486758 and CYP19 rs10046;(2)HSD17B1 rs676387 and CYP19 rs10046; (3)ERαrs2071454 and CYP19 rs10046;(4)ERαrs2071454 and HSD1B1 rs676387;(5)ERαrs2077647 and HSD1B1 rs676387;(6) ERαrs2077647 and CYP19 rs10046。5. Bioinformation software was employed for alle function prediction, Mutation from T to G at rs2071454 in ERαwould would lead to the loss of two transcription factor binding site(RAP1 and CAP). IHC detection observed that the expression of ERαin cancer tissue was lower than in normal tissue. A lower expression of ERαwas observed in CRC cancer patients with GT or GG at rs2071454 in ERα. The expression of ERαwas downregulated both in cancer tissue.Conclusion1. There is a correlation between CRC susceptibility and polymorphisms at rs2071454 in 5’end of ERαand rs2077647 in exon1of ERα. The difference in CRC risk is also observed among different haplotypes or diplotypes at rs2071454 and rs2077647. A genetic environmental interaction exists between rs2071454 or rs2077647 polymorphisms and age, sex, smoking and drinking.2. There is no asscociation between ERβpolymorphisms and CRC. A Multiplicative interaction between rs2077647 in ERαand rs3829768 in ERβis also observed.3. Two SNPs rs676287 and rs10046 in two rate-limiting enzyme of estrogen synthesis HSD17B1 and CYP19 respectively have correlation with CRC susceptibility. No correlation between CRC risk and rs248658 in CYP17 or rs4680 in COMT is found. rs676387 or rs10046 polymorphisms had a gene-gene interaction with age, sex, smoking and drinking.4. Gene-gene interactions lie between the following pairs of SNPs: (1) CYP17 rs2486758 and CYP19 rs10046;(2)HSD17B1 rs676387 and CYP19 rs10046; (3)ERαrs2071454 and CYP19 rs10046;(4)ERαrs2071454 and HSD1B1 rs676387;(5)ERαrs2077647 and HSD1B1 rs676387;(6) ERαrs2077647 and CYP19 rs10046。5. Mutation in rs2071454 will cause the loss of two transcripting factor binding site, RAP1 and CAP, and thus likely leading to the downregulation ERαin colorectal mucosa. This hypothesis needs further genetic functional test.更多还原