【作者】 何苗；

【导师】 王子卫；

【作者基本信息】 重庆医科大学 ， 外科学， 2013， 博士

【摘要】 表观遗传研究不涉及DNA序列改变的，可随细胞分裂而遗传的基因组修饰作用，包括DNA甲基化（DNA methylation）、组蛋白修饰、染色质重塑、基因组印迹等。目前，DNA甲基化是表观遗传学在肿瘤领域的研究热点。肿瘤细胞基因组正常DNA甲基化模式异常改变，表现为全基因组广泛低甲基化及抑癌基因高甲基化。这些异常改变所导致的肿瘤相关基因功能异常可能参与肿瘤发生与发展。DNA甲基转移酶（DNAmethyltransferase, DNMT）是催化DNA甲基化修饰的生物酶，其异常表达可能是肿瘤基因组发生异常DNA甲基化的原因；特别是DNMT异常高表达所致的各种抑癌基因异常高甲基化并失活，可能直接参与肿瘤发生发展病理生理机制。5-氮杂-2’-脱氧胞苷是经典的DNMT抑制剂，其对DNMT功能的抑制可能通过逆转抑癌基因高甲基化而抑制肿瘤进展。本研究分析胃癌组织及细胞中各种DNMT表达特点及DNMT抑制剂对胃癌细胞的影响；并分析胃癌组织基因组异常DNA甲基化谱。期望本研究对胃癌表观遗传科研提供参考。第一章胃癌及胃粘膜组织的DNMT蛋白表达目的：分析胃癌与配对胃粘膜组织的DNA甲基转移酶（DNMT，包括DNMT1，DNMT2，DNMT3A，DNMT3B及DNMT3L）表达特点及差异。方法：采用免疫组织化学法对30例胃癌及配对胃粘膜组织行5种DNMT蛋白表达研究。结果：胃癌及配对胃粘膜组织均表达各种DNMT蛋白，其中胃粘膜组织增殖活跃区表达DNMTs较显著。胃癌中DNMT3A表达较配对胃粘膜组织减弱，余DNMT蛋白（DNMT1，DNMT2，DNMT3B及DNMT3L）在胃癌及胃粘膜的表达无显著差异。弥漫型胃癌的各种DNMT阳性表达率（除DNMT2外）较肠型胃癌增高。结论：胃癌DNMT蛋白表达并未较配对胃粘膜增强；胃癌DNMT表达与Lauren分型相关。第二章胃癌与胃上皮细胞DNMT蛋白表达及5-氮杂-2’-脱氧胞苷对胃癌细胞的影响目的：分析胃癌细胞及胃上皮细胞的DNMT蛋白表达及DNMT抑制剂5-氮杂-2’-脱氧胞苷对胃癌细胞的影响。方法：采用免疫荧光化学法及蛋白质免疫印迹法分析各种DNMT蛋白在胃癌细胞MKN-28，SGC-7901及胃上皮细胞GES-1的表达；采用噻唑兰比色法及流式细胞术分析DNMT抑制剂5-氮杂-2’-脱氧胞苷对胃癌与胃上皮细胞增殖率、细胞周期分布及凋亡率的影响。结果：胃癌细胞MKN-28，SGC-7901表达各种DNMT蛋白，但DNMT1，DNMT3A表达较弱。DNMT1，DNMT2，DNMT3L蛋白在胃癌细胞以胞核表达为主；而DNMT3A，DNMT3B蛋白为胞核胞浆共表达。各种DNMT蛋白在胃癌细胞MKN-28，SGC-7901与胃上皮细胞GES-1间表达无显著差异。单独应用常规浓度（0.5-5μmol/L）的5-氮杂-2’-脱氧胞苷在短期内（5天）对胃癌细胞MKN-28，SGC-7901增值、凋亡或细胞周期的影响甚小。结论：胃癌细胞表达DNMT蛋白未必较胃上皮细胞增强。单独使用DNMT抑制剂5-氮杂-2’-脱氧胞苷在短期内未必影响胃癌细胞增殖或凋亡。第三章胃癌与胃粘膜组织的基因组DNA甲基化谱目的：分析胃癌与配对胃粘膜组织的基因组DNA甲基化谱。方法：采用甲基化DNA免疫共沉淀芯片（MeDIP-chip）分析胃癌与配对胃粘膜组织的基因组DNA甲基化谱差异；采用甲基化DNA免疫共沉淀实时定量PCR（MeDIP-qPCR）验证部分MeDIP-chip结果；采用实时定量逆转录PCR（qRT-PCR）分析胃癌组织中部分异常低甲基化基因（ABL2, FGF18, TRAF2, EGFL7和RAB33A）的转录。结果：局部高甲基化及广泛低甲基化是胃癌与胃粘膜组织基因组DNA甲基化的共同特点。部分基因启动子只在胃癌组织呈高甲基化，但部分基因启动子却只在胃粘膜组织呈高甲基化。搜寻肿瘤相关基因后发现，胃癌组织基因组异常DNA甲基化特征既包括抑癌基因异常高甲基化，又包括癌基因或促癌基因异常低甲基化。MeDIP-qPCR证实促癌基因ABL2, FGF18, TRAF2, EGFL7和RAB33A确实在胃癌组织呈异常低甲基化（即在胃粘膜呈相对高甲基化）。qRT-PCR发现基因FGF18, TRAF2与EGFL7在胃癌组织转录较在配对胃粘膜组织增强，但ABL2与RAB33A在胃癌的转录却较胃粘膜减弱。结论：胃癌基因组异常DNA甲基化特征具有复杂性，并非仅限于抑癌基因高甲基化。更多还原

【Abstract】 Epigenetics is the study of inherited genetic changes that occurwithout changes to the DNA sequence. The mechanisms of epigeneticchange include DNA methylation—the most widely researched epigeneticalteration in human tumor cells—histone modification, and chromatinremodeling. Human tumor cells exhibit aberrant DNA methylation patternsincluding the hypermethylation of CpG islands in tumor-suppressor genes(TSGs) and a global loss of DNA methylation in the genome. Thesechanges, which are related to the inactivation of TSGs and the activation ofoncogenes or tumor-promoter genes (TPGs), may promote tumorprogression. Abnormal expressions of DNA methyltransferase (DNMT)may play important roles in the aberrant DNA methylation that occurs intumors. DNMT expression can be significantly higher in tumors than incontrol tissues, and the over-expression of DNMT may contribute to tumorprogression through the hypermethylation-mediated inactivation of TSGsin CpG islands. Because DNMT inhibitors can reverse themethylation-dependent TSG silencing,5-aza-2’-deoxycytidine (a DNMT inhibitor) should be useful for tumor treatment. In this study, we usedgastric cancer (GC) to analyze the expressions of DNMTs (DNMT1,DNMT2, DNMT3A, DNMT3B, and DNMT3L) and the effect of5-aza-2’-deoxycytidine on cell inhibition. And we studied the aberrantDNA methylation pattern of GC genome. We expect that this research willbe helpful for the future Epigenetics studies on GC.CHAPTER1The Expressions of DNMT Proteins in GastricCancer and Matched Mucosa TissuesObjective: To investigate the protein expressions of DNAmethyltransferases (DNMTs, including DNMT1, DNMT2, DNMT3A,DNMT3B and DNMT3L) in gastric cancer (GC) and matched gastricmucosa tissues.Methods: Immunohistochemistry was performed to detect the DNMTexpressions in30pairs of GC and matched mucosa tissues.Results: GC and matched mucosa tissues expressed all five kinds ofDNMT proteins. The proliferating zones of foveolar epithelia in gastricmucosa tissues expressed DNMT proteins significantly. The expression ofDNMT3A was weaker in GC tissues than in matched controls. Theexpressions of other DNMTs, however, did not differ significantly betweenthe GCs and controls. In GC tissues, the positive rates of DNMTexpressions (with the exception of DNMT2) were higher in diffuse type than in intestinal type.Conclusion: GC did not necessarily show stronger expressions ofDNMT proteins than the matched mucosa tissues. The DNMT expressionsin GC were associated with the Lauren’s type.CHAPTER2The Distributions of DNMT Proteins in Gastriccancer and the effect of5-aza-2’-deoxycytidine on Gastric cancercellsObjective: To investigate the protein expressions and distributions ofDNMTs in GC and gastric epithelial cells and to analyze the effect of5-aza-2’-deoxycytidine on GC’s growth rates, cell cycle distributions andapoptosis rates.Methods: Immunofluorescence and western blotting were performedto detect the DNMT distributions and expressions in GC cells (SGC-7901and MKN-28) and gastric epithelial cell (GES-1). MTT assay and flowcytometry were performed to analyze the reactions of GC cells to5-aza-2’-deoxycytidine treatment.Results: SGC-7901and MKN-28expressed5kinds of DNMTproteins. DNMT1, DNMT2and DNMT3L were distributed in the nuclei ofGC cells. DNMT3A and DNMT3B were distributed in both the nuclei andcytoplasm of GC cells. The DNMT expressions did not differ significantlybetween the GC cells (SGC-7901, MKN-28) and gastric epithelial cell (GES-1).5-Aza-2’-deoxycytidine with concentrations of0.5-5μmol/L(where it acts as a demethylation drug) had very little effect on the growthrates, cell cycle distributions, and apoptosis rates of SGC-7901or MKN-28cells when the drug was used alone in short term (5days).Conclusion: GC cells didn’t necessarily express DNMT proteinsstronger than gastric epithelial cells.5-aza-2’-deoxycytidine (0.5-5μmol/L)didn’t necessarily inhibit GC cell proliferation when it was used alone inshort term.CHAPTER3The aberrant DNA methylation pattern of Gastriccancer genomeObjective: To investigate the aberrant DNA methylation pattern of GCgenome.Methods: Methylated DNA immunoprecipitation (MeDIP) microarray(MeDIP-chip) was performed to analyze the genomic DNA methylation ofa pair of GC and matched mucosa tissue. MeDIP-qRCR assay wasperformed to validate parts of the MeDIP-chip findings in5pairs of gastrictissues. qRT-PCR was performed to analyze the mRNA expressions ofvalidated genes (such as ABL2, FGF18, TRAF2, EGFL7and RAB33A) inthe5pairs of gastric tissues above.Results: For both the GC and matched mucosa tissue, localhypermethylation and global hypomethylation of genomic DNA were the significant characteristics. And certain CpG islands and gene promoterswere hypermethylated only in GC tissue, whereas some other CpG islandsand gene promoters were hypermethylated only in matched mucosa tissue.After separating tumor related genes those showed differential DNAmethylation in promoters between the cancer and control from the genome,GC tissue was found to contain both aberrantly hypermethylated andhypomethylated TSGs, oncogenes, and tumor-promoter genes (TPGs).MeDIP-qPCR verified some tumor related genes (such as ABL2, FGF18,TRAF2, EGFL7and RAB33A) were aberrantly hypomethylated in GCtissues compared to in controls. And the qRT-PCR assay found that themRNA transcriptions of FGF18, TRAF2and EGFL7were significantlymore in GC tissues than in matched mucosa tissues, but the mRNAtranscriptions of ABL2and RAB33A were significantly less in GC than inmatched controls.Conclusion: Aberrant DNA methylation pattern of GC genome iscomplex, but not limited to hypermethylation of TSG only.更多还原