【作者】 苑晓燕；

【导师】 曹佳；

【作者基本信息】 第三军医大学 ， 卫生毒理学， 2007， 博士

【摘要】 目的肝癌是我国最常见的恶性肿瘤之一,其发病机制尚不十分清楚。现代遗传学观点认为,肝癌等复杂疾病属于多基因疾病,是由环境因素和机体自身的遗传因素协同作用,经过多个阶段而发生和发展的。鉴定肝癌的遗传易感基因将有助于揭示肝癌的发病机制,并为实现肝癌的风险预测及个体化医疗奠定理论基础。一方面,动物模型和人群流行病学资料皆显示,雌激素可作为肿瘤促进剂促进肝癌的发生。雌激素通过雌激素代谢酶完成体内转化,并通过其受体发挥其生物学作用,因而提示雌激素受体和雌激素代谢酶基因的多态性可能影响个体对肝癌的遗传易感性。另一方面,氧化应激在肝癌的发生中起着重要作用,而体内的抗氧化体系可通过多种途径来抑制和清除活性氧的产生,以减轻氧化损伤的程度。因此,体内抗氧化酶基因的多态性也可能影响个体对肝癌的易感性。为证实上述两假说,我们在第一部分及第二部分研究中采用基于候选基因的关联研究策略及病例-对照实验设计,选取参与雌激素生理作用的雌激素代谢酶、雌激素受体基因及参与自由基清除的抗氧化酶基因作为候选基因,系统的研究了这些基因的多态性与肝癌易感性的相关性。此外,在过去4年中,本课题组系统发掘了120余个人类重要功能基因的1400余个单核苷酸多态性(Single nucleotide polymorphism, SNP)位点(包括调控区、编码区和剪接区的SNPs),其中包括肿瘤转移抑制基因KAI1/CD82启动子区(-1572至+415)的SNPs。大量实验证实KAI1基因对大多数肿瘤的转移起抑制作用,与肿瘤的侵袭及预后密切相关。因此,揭示KAI1/CD82基因在肿瘤组织中的表达调控机制及生物学作用,可为肿瘤的治疗和判断预后提供新思路。为此,我们选取CD82启动子区的两个常见SNP位点(rs934178和T-262G)进行了系统的功能研究,以期揭示启动子区SNP位点对CD82表达调控的影响及机制。材料与方法1.关联研究中,研究对象包括434例肝癌患者和480例非癌症对照,病例采集于广西壮族自治区肿瘤医院,居住地局限于广西壮族自治区扶绥县及其周边地区,对照来自于同一地区的社区人群,对病例与对照组人群的性别、年龄、居住地等因素进行频数匹配。2.关联研究中,采用聚合酶链式反应-限制性酶切片段多态性(Polymerase chain reaction-restriction fragment length polymorphism, PCR-RFLP)方法对雌激素受体β(Estrogen receptorβ, ESR2)的4个SNP位点(G1082A、A1730G、rs1256054和T1057G)、雌激素代谢酶的8个SNP位点(CYP17 Msp AI、CYP19 Trp39Arg、CYP19 I/D、CYP1A1 MspI、CYP1A1 Ile462Val、CYP1B1 Val432Leu、COMT Ala72Ser和COMT Val158Met)和抗氧化酶的6个SNP位点(CuZnSOD A1934C、MnSOD Ala16Val、ECSOD Ala40Thr、ECSOD Arg213Gly、CAT C-262T和GPx-1 Pro198Leu)进行基因分型。采用片断分析技术对及CYP19的短串联重复标记(TTTA)n进行基因分型。3.在SNP的功能研究中,采用电泳迁移率实验(Electrophoretic mobility shift assay, EMSA)检测了CD82启动子区T-262G多态性位点对转录因子与DNA结合的影响。采用双荧光素酶报告基因瞬时转染HepG2及293细胞,检测调控区的两个SNP位点(rs934178和T-262G)对-1572至+493区段启动子活性的影响。4.在遗传关联分析中,采用PHASE软件构建单倍型;采用Arlequin软件包计算LD模式的两个参数(Lewontin’s D’及r2)。采用SPSS14.0统计软件的非条件Logistc回归模型计算各基因型与肝癌风险的相关性,分析时首先对HBV携带状态进行分层,然后对性别、年龄、吸烟、饮酒等混杂因素进行分层,结果经性别、年龄、吸烟、饮酒及吸烟量(年包数,pack-year)等因素校正,对所有统计结果进行Bonferroni多重检验校正。在功能研究中,采用SPSS14.0统计软件的单因素方差分析比较不同转染组荧光素酶相对活性的差异。结果1.在ESR2和雌激素代谢酶基因多态性与肝癌的遗传易感性研究中,基因分型结果显示,中国南方人群中不存在ESR2 T1057G多态性位点及CYP19 (TTTA)n多态性位点。其余位点的关联分析显示,经Bonferroni多重检验校正后,仅COMT Ala72Ser在总体中与肝癌易感性显著相关,但对HBV携带状态进行分层后,该多态性位点与HBV携带者或非HBV携带者的肝癌易感性均无显著关联;其余位点及ESR2、CYP1A1、材料与方法COMT的单倍型/双体型与肝癌的易感性也无显著关联。2.在抗氧化酶基因多态性与肝癌的遗传易感性研究中,基因分型结果显示,中国南方人群中不存在CuZnSOD A1934C多态性位点,ECSOD Arg213Gly多态性位点次要等位的频率<1%。关联分析结果显示,在总体样本中MnSOD Ala16Val、CAT C-262T和GPx-1 Pro198Leu多态性位点与肝癌的患病风险无显著关联;对受试者的HBV携带状态、性别、年龄、吸烟及饮酒状态进行分层后,上述位点与肝癌的患病风险仍无显著性关联。ECSOD Ala40Thr与总体及HBV携带者的肝癌患病风险无显著关联;而在非HBV携带者中,与Ala/Ala基因型相比,Ala/Thr基因型及Ala/Thr+Thr/Thr基因型均显著增加了肝癌的患病风险(OR = 2.13, 95% CI = 1.25-3.64, P = 0.0057;OR = 1.90, 95% CI = 1.15-3.15, P = 0.012)。对性别、年龄、民族、家族史、吸烟及饮酒状态进行分层后,上述显著性关联分别存在于非HBV携带者的男性组、>49岁组、非汉族组、吸烟组和饮酒组。基因-基因间交互作用分析显示:在总体及非HBV携带者中,与携带两个ECSOD 40Ala等位及一个MnSOD 16Ala等位的个体相比,携带至少一个ECSOD 40Thr等位及两个MnSOD 16Val等位的个体患肝癌的风险显著增加(OR = 2.32, 95% CI = 1.36-3.98, P = 0.0018;OR = 9.02, 95% CI =2.04-39.92, P = 0.0037)。基因-环境交互作用分析显示,MnSOD Ala16Val、ECSOD Ala40Thr、CAT C-262T和GPx-1 Pro198Leu与吸烟、饮酒、性别、年龄间不存在显著交互作用。3.在SNP的功能研究中,EMSA结果显示:含有G、T等位的寡核苷酸双链探针均可结合相同分子量的转录因子,但在结合能力上存在明显差异,G等位的结合能力高于T等位。HepG2及293细胞中的报告基因结果显示:CD82基因-1572至+493区域具有较强的启动子活性,四种单倍型的启动子活性具有明显差异,其中C-T单倍型的启动子活性显著高于T-G单倍型(P < 0.01)。结论1.ESR2的G1082A、A1730G和rs1256054多态性位点不影响中国南方人群中个体对肝癌的遗传易感性;由上述3个SNP位点组成的ESR2单倍型及双体型也不影响个体对肝癌的易感性。2.CYP17 Msp AI、CYP19 I/D、CYP19 Trp39Arg、CYP1A1 Msp I、CYP1A1 Ile462Val、CYP1B1 Val432Leu、COMT Ala72Ser和COMT Val158Met等雌激素代谢酶的多态性位点不影响个体对肝癌的遗传易感性;由CYP1A1 Msp I、CYP1A1 Ile462Val组成的CYP1A1单倍型/双体型及由COMT Ala72Ser、COMT Val158Met组成的COMT单倍型/双体型也不影响个体对肝癌的易感性。3.MnSOD Ala16Val、CAT C-262T和GPx-1 Pro198Leu多态性位点不影响个体对肝癌的遗传易感性;ECSOD Ala40Thr多态性位点与非HBV携带者的肝癌易感性相关,携带至少一个40Thr等位个体患肝癌的风险显著增加;同时,ECSOD Ala40Thr与MnSOD Ala16Val多态性位点间存在显著的交互作用,协同增加肝癌的发生风险。4.CD82基因启动子区T-262G位点的G、T等位型均可结合相同的转录因子,但结合能力存在差异,rs934178和T-262G两个SNP组成的单倍型可影响CD82的启动子活性。更多还原

【Abstract】 Background & AimsHepatocellular carcinoma (HCC) is one of the most common malignancy in our country. The pathogenesis of HCC is still poorly understood. In fact, hepatocarcinogenesis is a long-term multistage process with the involvement of hereditary and environmental factors. The identification of susceptibility genes contributing to HCC would help to clarify pathophysiologic mechanisms relevant to hepatocarcinogenesis and would assist in predicting individual and population risk of HCC development.On the one hand, both animal models and human epidemiologic studies have documented that estrogen act as tumor promoters and might induce hepatocarcinogenesis. Estrogens realize their biogenesis, bioavailability and degradation via estrogen-metabolizing enzymes and exert their biological effects through the estrogen receptor (ER) of target cells. So we hypothesize that the estrogen-metabolizing enzymes and estrogen receptor ? may be the excellent candidate susceptibility genes for HCC and the genetic polymorphisms within these genes would result in differences in susceptibility to HCC. On the other hand, an increased oxidant burden may result in nuclear or mitochondrial DNA damage, DNA methylation, lipid peroxidation and regulation of signal transduction pathways and gene expression, which has been implicated in hepatocarcinogenesis. Several antioxidant enzymes assure protection against oxidative damage. So the genetic variation within antioxidant enzymes could also result in genotype-dependent differences in risk of HCC. To test these hypotheses, we investigated the association of the polymorphisms in estrogen-metabolizing enzymes genes, estrogen receptor ? and antioxidant enzymes genes with susceptibility to HCC in chapter I and chapter II.In the past 4 years, our group has screened SNPs systematically in 128 functionally important genes and discovered 1,400 SNPs in regulatory region, coding region and splicing region, including two common SNPs in the tumor metastasis suppressor gene, KAI1/CD82, regulatory region (-1,572bp to +415bp). Clarifying the influence of these two SNPs on gene expression may highlight a novel therapeutic target for cancer. So we analyzed the function of these two SNPs in the chapterⅢ.MethodsThe case-control study included 434 patients with HCC and 480 control subjects. All subjects were unrelated ethnic adult Chinese and residents in Fusui County and its surrounding regions at Guangxi province, a well-known high-risk region for HCC located in southern China. 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 HCC were collected via structured questionnaire.Genotyping was performed by polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) assay for four SNPs in ESR2 gene (G1082A, A1730G, rs1256054 and T1057G ), eight in estrogen-metabolizing enzymes genes (CYP17 Msp AI, CYP19 Trp39Arg, CYP19 I/D, CYP1A1 MspI, CYP1A1 Ile462Val, CYP1B1 Val432Leu, COMT Ala72Ser and COMT Val158Met), and six in antioxidant enzymes genes (CuZnSOD A1934C, MnSOD Ala16Val, ECSOD Ala40Thr, ECSOD Arg213Gly, CAT C-262T and GPx-1 Pro198Leu). We also genotyped the (TTTA)n variable number of tandem repeats polymorphism in CYP19 gene through GeneScan technique.Electrophoretic mobility shift assay (EMSA) was conducted to identify potential transcription factors that bind to -262T or -262G probe. To determine whether the polymorphism in the human CD82 promoter will affect the transcription of this gene, PCR amplified promoter fragments of 2065bp with T/C at rs934178 site and T/G at T-262G site were obtained from four heterozygous individuals and subcloned into pGL3 basic vector. Transient transfection was performed in HepG2 and 293 cells using LipofectamineTM 2000 and luciferase activity was determined using a Dual-Luciferase Reporter Assay System.Associations between polymorphisms and risk of HCC were estimated by unconditional logistic regression analyses using SPSS software (version 14.0; SPSS Inc., Chicago, IL). The odds ratios (ORs) were calculated by logistic regression and adjusted for age, sex, status of smoking and drinking and pack-years of smoking where appropriate. Potential modification effects of the polymorphisms on HCC risk by the above risk factors were assessed by the addition of interaction terms in the logistic model and by separate analyses of subgroups of subjects determined by these factors. In view of the multiple testing, the correction factor n(m-1) (n loci with m alleles each) was applied to correct the significance level. Haplotype frequencies between cases and controls were compared by using theχ2 test. The comparisons between more than two groups were carried out using the analysis of variance (ANOVA) in chapter III.Results1. The ESR2 T1057G polymorphism and CYP19 (TTTA)n polymorphism did not exist in this Chinese population. For the other polymorphisms of ESR2 gene and estrogen-metabolizing enzymes genes, except for COMT Ala72Ser polymorphisms, there was no association between these polymorphisms and HCC susceptibility in overall sample. When the subjects were stratified by HBV carriers status, age, gender, family history, smoking status and drinking status, there were no associations observed between the above all polymorphisms and HCC risk. The haplotype and diplotype distribution of ESR2, CYP1A1 and COMT exhibited no significant difference between the cases and controls.2. The CuZnSOD A1934C polymorphism did not exist in this population and the minor allele frequency of ECSOD Arg213Gly polymorphism was too low. On the basis of logistic regression analysis with adjustment for age, sex, status of smoking and drinking, and pack-years of smoking, there were no significant associations between three polymorphisms (MnSOD Ala16Val, CAT C-262T, GPx-1 Pro198Leu) and HCC susceptibility in overall sample, HBV carriers and non-HBV carriers. When the subjects were stratified by age, gender, family history, smoking status and drinking status, there were also no associations observed. For the ECSOD Ala40Thr polymorphism, no significant association was found between this polymorphism and HCC risk in overall subjects and HBV carriers. However, in non-HBV carriers, individuals with one 40Thr allele (Ala/Thr genotype) (OR = 2.13, 95% CI = 1.25-3.64, P = 0.0057) or at least one 40Thr allele (Ala/Thr and Thr/Thr genotype) (OR = 1.90, 95% CI = 1.15-3.15, P = 0.012) showed significantly higher risk to HCC, compared with Ala/Ala genotype. Although there was no significant interaction between ECSOD Ala40Thr and MnSOD Ala16Val in overall sample, HBV carriers or non-HBV carriers, the combined effects of ECSOD and MnSOD were greater than those for either polymorphism alone in overall sample and non-HBV carriers. Individuals with at least one ECSOD 40Thr allele and two MnSOD 16Val alleles showed increased risk compared with individuals with two ECSOD 40Ala alleles and at least one MnSOD 16Ala allele in overall sample(OR = 2.32, 95% CI = 0.13-0.80, P = 0.0018) and in non-HBV carriers (OR = 9.02, 95% CI =2.04-39.92, P = 0.0037). No interactions between genotypes of four polymorphisms and risk factors were obtained. No gene-gene interactions in these four polymorphisms were obtained.3. EMSA result showed that both -262T and -262G probe could form protein-DNA complex with HepG2 cell extracts. The G variant gives a more intense complex compared to the T probe, which was entirely competed by 200-fold molar excess of unlabeled G and T probes, respectively. The results indicated that the same protein bound with different affinity to G and T variants, respectively. Expression studies with reporter constructs showed significantly higher transcriptional activity of the C-T haplotype compared with T-G haplotype in HepG2 and 293 cells (P < 0.01).Conclusions1. The G1082A, A1730G and rs1256054 polymorphisms in ESR2 did not contribute to the differences in susceptibility to HCC in this Chinese population and the haplotypes and diplotypes based on these three polymorphisms also did not influence susceptibility to HCC.72. The eight polymorphisms in estrogen-metabolizing enzymes genes may not play a major role in mediating susceptibility to HCC. The haplotypes and diplotypes of CYP1A1 and COMT also did do not significantly confer an increased risk of HCC.3. The MnSOD Ala16Val, CAT C-262T and GPx-1 Pro198Leu polymorphisms did not contribute to the differences in susceptibility to HCC. ECSOD genotypes may modify the risk of HCC in non-HBV carriers. ECSOD and MnSOD had combined effects in mediating susceptibility to HCC.4. G and T variants at T-262G site in CD82 bound the same protein with different affinity. The G variant had higher affinity than the T variant. The haplotypes based on rs934178 and T-262G polymorphisms modified the transcriptional activity of CD82 promoter.更多还原