【作者】 郭洪；

【导师】 白云；

【作者基本信息】 第三军医大学 ， 遗传学， 2010， 博士

【摘要】 背景食管癌是人类最常见的恶性肿瘤之一,以其高发病率和高致死率成为威胁人类生命和健康的杀手。我国是食管癌大国,其中尤以鳞癌为主,每年有数十万的新发病例。因此,研究食管癌的发病机制,寻找其早期诊断、预防以及预后的标志就显得非常重要。食管癌是一个多基因参与、多因素作用的复杂疾病。近年来大量研究从基因多态性出发筛选出了很多与食管癌发病风险以及转移、预后相关的遗传标记。除了少部分位于蛋白编码序列或者启动子区域的位点功能得到阐明外,大部分的位点并没有发现其参与食管癌发病的机制,仅仅是作为一个分子标志物。microRNA(miRNA)是一种内源性的长22nt的单链小分子RNA,是一组不编码蛋白质的短序列RNA。通过与靶基因mRNA的3’端非翻译区(3’Untranslated Region, 3’UTR)的结合,在转录后水平降解mRNA或抑制翻译,从而调节基因的表达。miRNA与靶基因结合的关键部位称为“种子区域”(seed region, microRNA 5’末端开始2-7个核苷酸),与mRNA的3’UTR通过碱基配对结合。在人类基因组中目前已发现有超过一千万个SNP位点,在多态性数据库中可以发现其中有相当一部分落在了3’端非翻译区。最近的研究已经证实,这些位于3’端非翻译区的SNPs可能会通过影响miRNA的功能从而参与靶基因的表达调控。肿瘤细胞外基质与肿瘤基质细胞在肿瘤的发生发展过程中发挥着举足轻重的作用,基质金属蛋白酶通过对细胞外基质的降解来调节细胞的微环境。基质金属蛋白酶是一类依赖锌离子的蛋白水解酶,在人类己发现有20多种,几乎所有类型的人类肿瘤包括食管癌中都有基质金属蛋白酶,尤其是MMP1、2、3和9的表达异常。大多数基质金属蛋白酶由肿瘤细胞产生和分泌,但也可由肿瘤细胞周围的间质细胞分泌。研究表明,基质金属蛋白酶高表达与肿瘤进展,癌细胞侵袭和转移程度,病人的生存期等临床转归密切相关,是肿瘤病人预后的重要指标之一。前期大量研究关注基质金属蛋白酶基因的多态性位点与肿瘤的关联性,并发现部分位于启动子以及编码区的SNP位点与肿瘤的发生、转移以及临床分级分期相关。但迄今为止,还未有研究关注基质金属蛋白酶3’端非翻译区(3’UTR)的多态性位点与肿瘤易感性的关联研究。因此,本研究通过生物信息学预测,筛选基质金属蛋白酶基因3’端非翻译区中潜在的miRNA靶点,发掘这些靶点区域中的SNP位点,结合病例-对照研究,探讨基质金属蛋白酶基因3’端非翻译区中的基因多态性与食管癌发病的关系,并通过荧光素酶报告基因实验研究这些多态性位点的生物学功能。目的1、筛选与食管癌密切相关的基质金属蛋白酶基因3’UTR区域中的单核苷酸多态性位点,评价其对miRNA作用的影响。2、在重庆地区的食管鳞癌病例和正常对照中对以上位点进行关联研究。3、在细胞模型中,探讨阳性关联位点的生物学功能。方法与结果1、在NCBI数据库中下载与食管癌密切相关的基质金属蛋白酶基因(MMP1、2、3、7、8、9、10、11、12、13、14、15、21、26)的3’UTR序列,综合利用Diana、Targetscans、microCosm、miRanda、PicTar、microinspector以及PolymiRTS几个在线工具预测这些序列当中的miRNA结合位点。然后在dbSNP数据库中进行BLAST-SNP,筛选其中的SNPs位点,接下来使用RNAhybrid工具中的自由能总值(△△G)评价这些SNPs对“microRNA-靶基因”相互作用的影响。最后通过最小等位基因频率大于10%的标准选出一批位点进行下一步研究。最后共筛选到了位于基质金属蛋白酶基因3’UTR区域的13个SNP位点,在其中选取MMP8(rs12284255)、MMP2(rs7201)和MMP10 (rs470168)三个中国汉人中的高频位点进行下一步的研究。2、在444例食管鳞癌病人以及匹配的468例正常对照中,采用SNPshot的分型方法对以上筛选出来的SNP位点进行分型。结果发现对照人群中三个位点的基因型分布均符合哈代-温伯格平衡,将基因分型数据与HapMap数据库中的数据作比较,结果显示三个SNP位点在重庆地区正常汉族人群中的分布与中国北京汉人基本一致,与日本东京人无明显差异,但与欧裔美国人、非裔美国人以及非洲约鲁巴人相比有显著差异。采用病例-对照研究分析了三个SNP位点与食管鳞癌发病风险以及临床分期的关系。结果发现,MMP10(rs470168)位点和食管鳞癌发病风险间无相关性,而MMP8(rs12284255)和MMP2(rs7201)位点在共显性、显性、隐性和乘法模型中都显示出与食管鳞癌发病风险显著相关。MMP8(rs12284255)位点的AA基因型对食管鳞癌的发病风险有保护作用(OR为0.31, 95%CI为0.15-0.62),而MMP2(rs7201)位点的CC基因型是食管鳞癌发病的危险因素(OR为2.29,95%CI为1.44-3.64)。进一步在临床分期分析中发现,MMP8(rs12284255)位点在临床I/IIa分期组中A等位基因的频率为0.26,而在临床IIb/III/IV分期组中为0.17,CA基因型与较早的临床分期有关联(OR为0.48,95%为0.31-0.74)。MMP2(rs7201)位点在临床I/IIa分期组中C等位基因的频率为0.32,而在临床IIb/III/IV分期组中为0.45,CC基因型与较晚的临床分期有关联(OR为2.85,95%为1.65-4.23)。按照吸烟史对样本进行分层分析,结果发现,在不吸烟人群中,两个位点的基因型分布在食管癌患者组和对照组中没有显著差异(P值分别为0.29和0.39);而在吸烟人群中,两个位点的基因型分布与食管鳞癌发病风险显著相关(P值均小于0.001)。位点MMP8(rs12284255)和MMP10(rs470168)的单倍型分析分析显示,这两个位点不存在连锁不平衡关系(D’为0.423,r2为0.135),在单倍型的关联研究中也发现,四种单倍型在食管鳞癌病例组和对照组中均没有显著差异。3、第二部分研究分别探讨了MMP2(rs7201)和MMP8(rs12284255)两个位点的生物学功能。RT-PCR和Western-blot检测MMPs基因在食管癌组织中的mRNA和蛋白表达,统计相应SNP位点不同基因型与表达量的关系,探讨SNP位点影响基因表达调控的层面。然后分别构建SNP位点不同等位基因的报告基因载体pMIR-REPORT,与经生物信息学预测的相应miRNA的mimics/controls分别共转染到细胞株中,检测荧光素酶报告基因的表达,比较SNP位点不同基因型对“miRNA-靶基因”作用的影响,证实SNP位点的生物学功能。最后再选择基因分型为功能等位基因纯合子的细胞株,转染与生物信息学预测相应的miRNA的mimics和controls到细胞中,检测目的基因的蛋白表达,比较两组的差异,进一步验证SNP位点的功能意义。(1)MMP2(rs7201)位点的生物学功能研究检测并统计了rs7201位点不同基因型的食管鳞癌样本中MMP2基因的mRNA和蛋白表达水平。结果发现,三组的mRNA表达水平没有明显差异(P=0.85);而在蛋白表达水平,CC组明显高于CA组和AA组,CC组与AA组之间存在显著差异(P<0.05)。提示位点rs7201可能是在转录后水平发挥作用。在随后的荧光素酶报告基因实验中,选择miR-520g表达较低的KYSE150细胞株,将miR-520g mimics/controls与pMIR-REPORT-C/pMIR-REPORT-A共转染到细胞中,结果发现,与miR mimics + pMIR-REPORT-C组相比,miR mimics + pMIR-REPORT-A组中荧光素酶表达量明显下降(P<0.05)。证实了MMP2(rs7201)位点的不同等位基因能够影响miR-520g与靶位点的结合能力,A等位基因的亲和力明显高于C等位基因。MMP2(rs7201)位点在食管癌细胞株TE-1中的基因型为AA,分别转染miR-520g的mimics和controls到细胞中,检测MMP2的蛋白表达。结果发现,与转染之前相比,miR-520g的mimics转染组MMP2的蛋白表达明显降低,而controls组没有显著变化,验证了miR-520g对MMP2基因的调控作用。(2)MMP8(rs12284255)位点的生物学功能研究检测并统计了MMP8(rs12284255)位点不同基因型的食管癌样本中MMP8基因的mRNA和蛋白表达水平,结果发现,三组的mRNA表达水平没有明显差异(P=0.79);而在蛋白表达水平,CC组明显高于CA组和AA组,CC组与AA组之间存在显著差异(P<0.05)。提示位点rs12284255可能也在转录后水平发挥作用。在随后的荧光素酶报告基因实验中,选择miR-495表达较低的KYSE150细胞株,将miR-495 mimics/controls与pMIR-REPORT-G/pMIR-REPORT-T共转染到细胞中,结果发现,与miR mimics + pMIR-REPORT-G组相比,miR mimics + pMIR-REPORT-T组中荧光素酶表达量明显下降(P<0.05)。证实了MMP8(rs12284255)位点的不同等位基因能够影响miR-495与靶位点的结合能力,T等位基因(对应于正义链上rs12284255位点的A等位基因)的亲和力明显高于G(对应于正义链上rs12284255位点的C等位基因)等位基因。结论1、生物信息学预测并筛选到三个位于基质金属蛋白酶基因3’UTR区域的SNP位点,在重庆地区食管鳞癌病例和正常对照中的研究发现,MMP8(rs12284255)位点和MMP2(rs7201)位点与食管鳞癌的发病风险以及临床分期存在显著关联。2、位点MMP2(rs7201)的不同等位基因能够影响miR-520g与靶位点的结合能力,A等位基因的亲和力明显高于C等位基因。该多态性位点可能通过影响miR-520g与MMP2基因3’UTR靶位点的结合能力而参与食管癌的发病。3、位点MMP8(rs12284255)的不同等位基因能够影响miR-495与靶位点的结合能力,T等位基因(对应于正义链上rs12284255位点的A等位基因)的亲和力明显高于G等位基因(对应于正义链上rs12284255位点的C等位基因)。该多态性位点可能通过影响miR-495与MMP8基因3’UTR靶位点的结合能力而参与食管癌的发病。综上所述,本研究从与食管癌发生发展密切相关的基质金属蛋白酶基因家族出发,利用生物信息学在其3’UTR区域发掘位于miRNA结合位点的SNP位点,并通过病例-对照研究发现MMP8(rs12284255)位点和MMP2(rs7201)位点与食管鳞癌的发病风险以及临床分期存在显著关联,最后在细胞模型中证实了这两个SNP位点的生物学功能。研究为我们采用候选基因SNP关联分析策略开展散发性食管癌遗传易感性的研究提供了一条崭新的思路,关联研究结果可能可以用于食管癌的早期诊断,但这两个位点的关联性需要在更大的样本以及其它人群中得到验证。阳性关联SNP位点的功能研究结果不仅阐述了食管癌中基质金属蛋白酶异常表达的可能机制,还为miR-520g和miR-495参与食管癌发病的可能性提供了线索,值得进一步深入研究。更多还原

【Abstract】 BackgroundEsophageal cancer is regarded as one of the most common malignancy worldwide. Compared with other nations, there is higher incidence in China. Esophageal cancer is an extremely fatal disease. In spite of great advancement in cancer treatment, prognosis is still poor.Esophageal cancer is a complex disease involving many genes. Lots of genetic markers for tumorigenesis and progression of esophageal cancer have been identified in the case-control studies. However, the function of lots of the SNP markers hasn’t been illuminated, except the sites located in the promoter or coding regions.MicroRNAs (miRNAs) are endogenous 22nt small non-coding single stranded RNAs that bind with target mRNAs and function as post-transcriptional regulators of gene expression by either promoting mRNA degradation or translational silencing. The critical region for miRNA binding to mRNA is the ’seed region’ (2-7 nucleotides from the 5’ end of the miRNA), which most often binds to a target site in the 3’UTR of the genes by perfect Watsn-Crick complementarities. To date, approximately 10 million SNPs have been identified in the human genome, occurring on average every 100 to 300 base pairs. According to the SNP database, there are many SNP sites located in the 3’untranslated regions. Recent studies have discovered that these SNPs may influence the targeting of microRNA to protein-coding genes.The matrix outside the tumor cells and the sustentacular cells play important role in the tumorigenesis and progression. Matrix Metalloproteinases (MMPs) regulate the micro- environment through degrading the matrix. There are over 20 kinds of MMPs in human, which depend on Zn2+. Almost all of MMPs expressed abnormally in human cancers, especially MMP1, 2, 3, 9. Most of the MMPs are secreted by the tumor cells and the sustentacular cells. Lots of studies have uncovered that overexpression of MMPs correlated with progress, invasion and metastasis of cancer, so MMPs may also act as biomarkers for prognosis.Epidemiological studies have established a link between variations in MMPs genes and risk of cancer. Some SNP sites located in the promoter or coding regions were identified associated with tumorigenesis, metastasis and clinical stages. However, there is no association study focusing on the sequence variants in the 3’UTR of MMPs. In the study, we predicted the binding sits of miRNAs in the 3’UTR of MMPs by using bioinformatics approaches. Then after picking the SNP sites in these binding regions, we evaluated the relationship between the SNPs and risk of esophageal cancer in a case-control study. For the SNPs associated with ESCC, we verified the function of them in later luciferase reporter assay.Objectives1. Predicted the binding sits of miRNAs in the 3’UTR of MMPs by using bioinformatics approaches, picked the SNP sites in these binding regions and evaluated the effects of SNPs.2. Evaluated the associations between the SNPs and risk of esophageal cancer in a case-control study.3. For the SNPs associated with ESCC, we verified the biological functions of them in the cell model.Methods1. Downloaded the 3’ UTR sequences of MMP gene associated with esophageal cancer in NCBI database, used online prediction tools such as Diana, Targetscans, microCosm, miRanda, PicTar, microinspector and PolymiRTS to predict the miRNA combinding site of these sequences. Screened SNP sites in dbSNP database by BLAST-SNP, and used the total free energy (△△G) to evaluate the interaction between these SNPs and microRNA-target gene with RNAhybrid tool. If the minor allele frequency is over 10%, the SNPs could be used in the further study.2. Genotyped the SNPs by using SNPshot method in 444 ESCC patients and matched 468 healthy controls. Non-conditional logistic regression was used to revise the confounding factors such as age, sex and smoking status. We used odd ratio (OR) and 95% confidence interval (95% CI) to evaluate the associations between the SNPs and the risk of ESCC.3. RT-PCR and Western-blot were used to detect the mRNA and protein expression of MMPs in esophageal cancer tissues. The pMIR-REPORT vectors with different alleles of ESCC associated SNP sites were constructed. Then co-transfected the pMIR-REPORT vector and correlative miRNAs mimics/controls to a cell line and detected the expression of luciferase reporter gene. Compared the expression of luciferase reporter with different allele and verified the biological functions of the SNP sites. After selecting cell line with homozygous functional alleles of ESCC-associated SNP sites, transfected mimics/controls of corresponding miRNAs, detected protein expression and compared the two groups.Results1. 13 SNPs were identified in 3’UTR of ESCC relevant MMPs. MMP8 (rs12284255), MMP2 (rs7201) and MMP10 (rs470168) were selected for further study.2. The three SNPs were genotyped by using SNPshot assay in 444 ESCC patients and 468 healthy controls. The genotype distributions of the three SNPs all conformed to H-W equilibrium in controls and were coincident with that of CHB (Chinese Han in Beijing) in HapMap database. However, they were different from that in CEU (Utah residents with Northern and Western European ancestry from the CEPH collection), ASW (African ancestry in Southwest USA) and YRI (Yoruban in Ibadan, Nigeria). Then the association between the three sites and risk/clinic stage of ESCC were evaluated. The results showed that site MMP8 (rs12284255) and MMP2 (rs7201) were both associated with risk of ESCC. The association was more significant under codominant genetic model for the two sites (P=0.001, P<0.001). Genotype AA of site MMP8 (rs12284255) was protective for ESCC (OR: 0.31, 95%CI: 0.15-0.62), however Genotype CC of site MMP2 (rs7201) increased the risk for ESCC (OR: 2.29, 95%CI: 1.44-3.64). When took account of smoking status, the association was lost for both sites in the non-smokers.The haplotype results showed that there was no linkage disequilibrium between site MMP10 (rs470168) and MMP8 (rs12284255) (D=0.423, r2=0.135). So the two sites might be not in a LD block. Further association study also showed there was no association between haplotypes and ESCC.Site MMP8 (rs12284255) and MMP2 (rs7201) were also associated with the clinic stage of ESCC. The frequency of allele A of site MMP8 (rs12284255) was 0.26 in stage I/IIa and 0.17 in stage IIb/III/IV. Genotype CA was associated with early stage of ESCC (OR: 0.48, 95% CI: 0.31-0.74). The frequency of allele C of site MMP2 (rs7201) was 0.32 in stage I/IIa and 0.45 in stage IIb/III/IV. Genotype CA was associated with a later stage of ESCC (OR: 2.85, 95% CI: 1.65-4.23).3. In the part two, the biological function of site MMP2 (rs7201) and MMP8 (rs12284255) were evaluated.(1) Biological function study of MMP2 (rs7201)The mRNA/protein expression of MMP2 in samples with different genotypes of site MMP2 (rs7201) were measured. The results showed that there was no significant difference between three groups on the mRNA expression level. However, the protein expression was higher in genotype CC than genotype CA or AA. The difference between genotype CC and AA was significant statistically (P<0.05). Then selected esophageal carcinoma cell line KYSE150 with low expression of miR-520g for further study. After cotransfecting with miR-520g mimics/controls and pMIR-REPORT-C/pMIR-REPORT-A, the expression of luciferase was obviously low in“miR mimics + pMIR-REPORT-A”group, compared with“miR mimics + pMIR-REPORT- C”group (P<0.05). The results revealed that the affinity of allele A to miR-520g was much higher than allele C.The genotype of site MMP2 (rs7201) for esophageal carcinoma cell line TE-1 was AA. We transfected Mimics/controls of miR-520g to KYSE450 cell respectively and detected the protein expression of MMP2. The results showed the expression of MMP2 decreased obviously after transfection with mimics of miR-520g, however, that remained invariable after transfection with controls.So different alleles of site MMP2 (rs7201) might influence the combination of miR-520g to 3’UTR of MMP2 gene.(2) Biological function study of MMP8 (rs12284255)Detected the mRNA/protein expression of MMP8 in samples with different genotypes of site MMP8 (rs12284255). The results showed that there was no significant difference between three groups on the mRNA expression level. However, the protein expression was higher in genotype CC than genotype CA or AA. The difference between genotype CC and AA was significant statistically (P<0.05).Esophageal carcinoma cell line KYSE150 was picked for further study, because of low expression of miR-495. After cotransfecting with miR-495 mimics/controls and pMIR-REPORT-G/pMIR-REPORT-T, the expression of luciferase was obviously low in“miR mimics + pMIR-REPORT-T”group, compared with“miR mimics + pMIR- REPORT-G”group (P<0.05). The results revealed that tha affinity of allele T (corresponding to C allele of rs12284255 in the sense strand) to miR-495 was much higher than allele G (corresponding to C allele of rs12284255 in the sense strand). So different alleles of site MMP8 (rs12284255) might influence the combination of miR-495 to 3’UTR of MMP8 gene.Conclusions1. For the three SNPs located in 3’UTR of MMPs, we conducted a case-control design in Chinese Han population in Chongqing. These results demonstrated there was statistical significance of MMP8 (rs12284255) and MMP2 (rs7201) polymorphisms in the risk and clinical stage of esophageal squamous cell carcinoma.2. Different alleles of rs7201 could affect the ability of miR-520g combining to the target sites in MMP2. The affinity of A allele was higher than C allele. This polymorphism could participate the pathogenesis of ESCC by affecting this combination.3. Different alleles of MMP8 (rs12284255) could affect the ability of miR-495 combining to the target sites in MMP8. The affinity of T allele (corresponding to A allele of rs12284255 in the sense strand) was higher than G allele (corresponding to C allele of rs12284255 in the sense strand). This polymorphism could also participate the pathogenesis of ESCC.Above all, we discovered a set of SNP sites in the 3’UTR of MMPs genes. After a case-control study in a Chinese Han population in Chongqing, we found two sites MMP8 (rs12284255) and MMP2 (rs7201) were associated with risk and clinic stage of ESCC significantly. The biological functions of the two ESCC-associated SNP sites were studied in the later study. The results might be valuable for early detection of ESCC, however the association needs verification in larger populations or in other ethnic populations. The study might also elucidate possible mechanism of abnormal expression of MMPs in ESCC from a sight from miRNA.更多还原