【作者】 刘文斌；

【导师】 曹佳；

【作者基本信息】 第三军医大学 ， 卫生毒理学， 2010， 博士

【摘要】 研究背景与目的:肺癌是全球性的恶性肿瘤,其发病率和死亡率在我国以及世界其它地区的肿瘤性疾病中均居第一位,严重危害人们的健康和生命。因此,深入研究肺癌发生的分子机制,尤其是癌前病变的分子机制,对于开展高危人群的筛选和利用有效的分子生物标记进行早期诊断,降低肺癌发生率和死亡率具有十分重要的意义。肺癌的发生是一个多阶段、多步骤的过程。例如:肺鳞癌由支气管粘膜上皮恶变所经历的病变包括:增生→鳞状上皮化生→轻度、中度、重度不典型增生→原位癌→浸润癌。既往的研究表明,在肺癌发生过程中需要一系列的遗传学改变的积累,包括DNA突变、缺失、扩增、重组、染色体畸变等引起的基因表达的改变。近年来,人们逐渐认识到基因表达调控的另一重要机制—DNA甲基化,在肺癌的发生发展过程中也起着重要的作用,但是相关的研究还不够深入。DNA甲基化是表遗传学修饰的主要形式,其主要特点是通过对CpG序列的胞嘧啶进行甲基化修饰来调控基因的表达,而DNA序列本身并不改变。已有的研究表明,DNA异常甲基化是人类肿瘤组织中常见的改变之一。一方面,基因组范围内散发的CpG二核苷酸序列普遍发生低甲基化,从而导致染色体的断裂、易位、丢失、以及部分原癌基因的激活;另一方面,抑癌基因的启动子区域CpG岛DNA发生高甲基化,造成相关基因的表达失活。目前,已有较多文献报道肺癌肿瘤组织和细胞株中许多肿瘤相关基因启动子区发生了高甲基化,这些基因涉及到细胞周期调控、细胞凋亡、DNA修复、细胞黏附等多条信号通道。可见,在肺癌的发生发展过程中,以DNA甲基化为主的表遗传学机制与遗传学机制占有同等重要的地位。但是,对肺癌癌前组织中DNA甲基化的研究相对较少,DNA异常甲基化在癌前病变中的作用机制也还不清楚。研究表明, 3-甲基胆蒽(3-methylcholanthrene, MCA)和二乙基亚硝胺(Diethylnitrosamine, DEN)诱导的大鼠肺鳞癌的发生与人类肺癌的发生经历了相似的形态学过程,在分子生物学水平上,也都有相似的变化,并导致相似的基因表达的改变。上述研究为利用此肺癌模型探讨人类肺癌发生的表遗传学机制提供了依据。基于以上考虑,我们以致癌物MCA和DEN诱导的大鼠肺鳞癌模型为基础,研究癌变过程中DNA甲基化的动态改变情况及作用机制,包括全基因组低甲基化动态变化趋势、各种肿瘤相关基因甲基化状态的改变及对蛋白表达的影响以及DNA甲基转移酶的改变情况等,初步揭示DNA甲基化在化学致癌过程中的作用,为阐明肺癌癌变的表遗传学机制奠定基础。材料和方法:1. Wistar大鼠90只,雌雄各半,6周龄,体重200g±20g,随机分为实验组80只,对照组10只。实验组每只大鼠左肺下叶一次性灌注0.1ml致癌物碘油混悬液(含10 mg MCA和10μl DEN),对照组每只大鼠灌注0.1ml碘油。于灌注后第15、35、55、65、75天随机抽取各实验组动物16只和对照组动物2只处死。取灌注部位病变组织,一分为二,一份抽提组织RNA和蛋白质,另一半置于4%多聚甲醛固定液中,常规石蜡包埋,制备成4μm和10μm厚的切片。利用激光捕获显微切割的方法获取正常和处于不同癌变阶段的组织细胞并提取DNA。2.采用抗5-甲基胞嘧啶(5-methycytosine, 5-mC)抗体免疫组化的方法,检测大鼠肺鳞癌癌变各阶段基因组甲基化水平,利用图像分析系统测量其平均光密度值和积分光密度值,从细胞形态学水平获取DNA低甲基化的发生情况;利用甲基化敏感性随机引物PCR(Methylation-sensitive arbitrarily primed PCR, MS-AP-PCR)方法分析了11例癌前病变组织和11例肿瘤组织及其配对的正常支气管上皮组织基因组DNA中异常甲基化的改变情况,观察癌变过程中DNA甲基化的差异;分离差异甲基化片段,克隆并进行测序;利用BLAST软件对序列的同源性进行分析,确定是否为已知基因。3.采用甲基化特异性PCR(Methylation-specific PCR, MSP)和测序的方法检测癌变过程中肿瘤相关基因p15、p16、p27、p57、RASSF1A、TSLC1、TIMP-3、E-cadherin、N-cadherin、DAPK1、FHIT、SOCS-3高甲基化的发生情况;采用免疫组化检测p16、p27、p57、RASSF1A、TSLC1、TIMP-3、N-cadherin、DAPK1、FHIT、SOCS-3蛋白在大鼠肺鳞癌癌变各阶段的表达水平,采用Western blot检测各蛋白在正常、癌变和鳞癌组织中的表达水平。4.采用免疫组化检测DNA甲基转移酶(DNA methyltransferase, DNMT)1、3a和3b在大鼠肺鳞癌癌变各阶段的表达水平,分析其与各肿瘤相关基因甲基化之间的关联;原代培养p16、p57、TSLC1基因甲基化大鼠肺鳞癌细胞,采用不同浓度(2μM、5μM、10μM)去甲基化试剂5-氮杂-2′-脱氧胞苷(5-aza-2′-deoxycytidine, 5-aza-dC)处理,MSP和逆转录PCR(Reverse transcription-PCR, RT-PCR)分别检测处理前后各基因甲基化和mRNA表达的情况。5.所有参数均采用SPSS13.0统计软件分析处理,各基因甲基化发生频率和各蛋白阳性率差异比较以及甲基化和表达的相关性分析采用χ2检验,方差分析用于分析组间均数差别,双侧概率检验,检验水平α为0.05及0.01两种。结果:1.灌注后15-75天分批处死大鼠,大体病理观察发现对照组左肺无明显改变,实验组大鼠左肺下叶有直径0.5cm-1cm不等的肿块。HE切片染色,光镜下观察发现,对照组无明显的增生性或肿瘤性病变;实验组出现从正常支气管上皮组织到肺鳞癌各阶段病变,包括支气管粘膜上皮过度增生、鳞状上皮细胞化生、不典型增生、原位癌以及广泛浸润癌,且多为一例病变组织中同时存在多个癌变阶段。癌变过程各阶段的计数结果:正常支气管上皮20例(包括对照组和实验组各10例)、上皮增生25例、鳞状化生27例、不典型增生37例、原位癌30例、浸润癌25例。2.抗5-mC抗体在支气管上皮细胞胞核表达,正常对照组胞核呈棕黄至黄褐色,随着癌变过程的延续,染色程度逐渐变淡,至浸润癌阶段胞核呈浅黄色。癌变各阶段细胞胞核染色基底层较腔细胞层深。支气管粘膜上皮增生、鳞状化生、不典型增生、原位癌、浸润癌5-mC免疫组化平均光密度值和积分光密度值呈总体下降趋势,与正常对照组比值均有显著差异(P < 0.01);正常与癌变各阶段支气管上皮基底层细胞平均光密度值和积分光密度值比腔细胞层细胞层高,差异有显著性意义(P < 0.01)。3.利用MS-AP-PCR方法分析了11例癌前病变组织和11例肿瘤组织,及其配对的正常支气管上皮组织DNA中异常甲基化的改变情况,结果发现多数标本中均能发现异常甲基化片段。在200bp-700bp范围内,我们一共回收了8个扩增片段,其中低甲基化片段7个,高甲基化片段1个。通过克隆、测序和BLAST分析显示,8个DNA片段中有7个与大鼠基因组序列有高度同源(99%-100%的同源性)。这些序列位于大鼠的染色体1、3、9、12、13、20和X上,与细胞周期调控、细胞生长分化相关。采用MSP和测序的方法对高甲基化片段Hyper-1进行了鉴定分析,结果发现Hyper-1在正常组织中未发现甲基化,而在肿瘤和癌变组织中均发现甲基化。4. p15和E-cadherin在正常和癌变各阶段均未发生甲基化。其余肿瘤相关基因在癌变过程中甲基化程度逐渐增强,至少一个基因发生甲基化的频率和平均甲基化基因个数不断增加。正常、增生、鳞状化生、不典型增生、原位癌和浸润癌各阶段组织中甲基化频率分别为:p16:0%、8.00%、22.22%、40.54%、50.00%和56.00%;p27:0%、0%、0%、0%、3.33%和8.00%;p57:0%、0%、11.11%、18.92%、26.67%和36.00%;RASSF1A:0%、0%、3.70%、8.11%、6.67%和16.00%;TSLC1:0%、0%、18.52%、24.32%、40.00%和48.00%;TIMP-3:10.00%、12.00%、18.52%、24.32%、30.00%和60.00%;N-cadherin:0%、0%、7.41%、18.92%、26.67%和36.00%;DAPK1:0%、0%、7.41%、10.81%、26.67%和36.00%;FHIT:0%、0%、3.70%、16.22%、43.33%和48.00%;SOCS-3:0%、0%、7.41%、16.22%、26.67%和48.00%。相反,除N-cadherin表达增强外,其余各肿瘤相关基因蛋白在正常、增生、鳞状化生、不典型增生、原位癌和浸润癌各阶段组织中表达不断下降,阳性表达率分别为:p16:90.00%、84.00%、70.37%、59.46%、40.00%和32.00%;p27:95.00%、88.00%、74.07%、67.57%、60.00%和44.00%;p57:95.00%、92.00%、81.48%、75.68%、63.33%和56.00%;RASSF1A:100.00%、92.00%、70.37%、64.86%、53.33%和48.00%;TSLC1:95.00%、88.00%、55.56%、48.65%、43.33%和40.00%;TIMP-3:100.00%、96.00%、77.78%、70.27%、53.33%和24.00%;N-cadherin:0%、0%、14.81%、21.62%、33.33%和44.00%;DAPK1:100.00%、100.00%、88.89%、81.08%、70.00%和56.00%;FHIT:100.00%、96.00%、81.48%、62.16%、43.33%和36.00%;SOCS-3:100.00%、100.00%、88.89%、86.49%、73.33%和56.00%。Western blot结果:未甲基化且免疫组化阳性的标本蛋白表达较强,甲基化且免疫组化阴性的标本蛋白表达较弱或无表达。p16、p57、TSLC1、TIMP-3、DAPK1、FHIT、SOCS-3基因甲基化与蛋白表达呈显著负相关。5.正常、增生、鳞状化生、不典型增生、原位癌和浸润癌各阶段DNA甲基转移酶的阳性表达率分别为:DNMT1:0%、16.00%、29.63%、40.54%、46.67%和64.00%; DNMT3a:0%、0%、18.52%、37.84%、40.00%和68.00%;DNMT3b:0%、0%、0%、5.41%、10.00%和12.00%。DNMT1与DNMT3a、DNMT3b表达均呈显著正相关(P < 0.05),DNMT3a与DNMT3b表达无相关(P > 0.05)。DNMT1、DNMT3a、DNMT3b的表达与肿瘤相关基因甲基化相关性分析结果显示:p16、p57、TSLC1基因甲基化分别与DNMT1和DNMT3a的表达呈显著正相关(P < 0.01),此外,RASSF1A、TIMP-3、N-cadherin基因甲基化与DNMT3a的表达呈显著正相关(P < 0.01)。DNMT1和DNMT3a表达阳性的标本中平均甲基化基因数目是2.43±1.85和2.65±1.82,显著高于表达阴性的标本中平均甲基化基因数目1.35±1.98和1.34±1.95。对p16、p57、TSLC1基因甲基化的大鼠肺鳞癌原代细胞进行去甲基化试剂处理后,其基因mRNA表达水平显著升高。结论:1.采用支气管灌注MCA和DEN两种致癌物成功建立Wistar大鼠肺鳞癌癌变模型。2.基因组甲基化水平的降低在肺癌癌变早期就已发生,高甲基化和低甲基化共同存在于癌变组织和肿瘤组织中,并且在MCA和DEN诱导的大鼠癌变过程中起到非常重要的作用。筛选出的差异性甲基化片段所在区域可能是化学致癌癌变过程中改变的敏感地区,这些序列可以探索作为接触致癌物质的潜在生物标志物。经鉴定的高甲基化片段,提示其可能来自新基因,并且高甲基化可能与相关基因转录抑制有关。3.细胞周期调控相关基因p16、p27、p57、RASSF1A,细胞黏附相关基因TSLC1、TIMP-3、N-cadherin,凋亡相关基因DAPK1、FHIT以及SOCS-3基因高甲基化导致的基因沉默是MCA和DEN诱导的大鼠肺鳞癌癌变过程中的早期和频繁的事件,并在此过程中起重要作用。4. DNMT1和DNMT3a表达的增加是癌变过程中一个具有特征的早期分子改变,共同调控基因甲基化的发生,参与了MCA和DEN诱导的大鼠肺鳞癌发生发展的全过程。5. DNA甲基化是肿瘤相关基因失活的重要机制,并且CpG岛高甲基化导致的基因沉默是一个可以逆转的过程。综上所述,MCA和DEN诱导的大鼠肺鳞癌模型可以从表遗传学机制来阐明致癌物在诱发癌变过程中的作用及其机制,丰富传统的“遗传学致癌理论”。建立的实验体系,为进一步全面评价环境因素与DNA甲基化相互作用参与肺癌癌变的发生奠定了基础,也将对传统的基于“突变检测”的致癌物的评价体系起到完善作用。同时,本研究为进一步研究相关基因如何通过表遗传学/遗传学的相互作用介导相关信号通路参与癌变提供资料,也为利用该模型筛选新的脱甲基化制剂用于肺癌治疗以及评价我国环境因素与DNA甲基化相互作用在肺癌发生中的机制提供方法学和实验模型。更多还原

【Abstract】 Backgroud and Objective:Lung cancer is the most common cause of cancer-related mortality in China, as well as in the United States and Europe. Effective early diagnosis and targeted therapies to reduce incidence and mortality would benefit from a better understanding of the key molecular changes from normal cells to precancerous lesions to malignant tumor cells. However, the molecular events responsible for lung cancer initiation and development are still largely unknown.The occurrence of lung cancer is a multiphase and multistep process. Take lung squamous cell carcinoma for example, it will experience several stages from bronchial epithelial to malignant lesions, including hyperplasia, squamous metaplasia, dysplasia (mild, moderate, and severe), carcinoma in situ (CIS) and infiltrating carcinoma. Previous studies have shown that there is an accumulation of many genetic changes that occur during the process of lung tumorigenesis, including DNA mutations, deletions, amplification, rearrangements, chromosome aberrations, and so on. In recent years, DNA methylation as another important mechanism for the regulation of gene expression has been shown to play an important role in the lung carcinogenesis, but the relevant research is still in the initial stage.Methylation of DNA, which occurs at the cytosine residues of cytosine- phospho-guanine (CpG) dinucleotides by an enzymatic reaction that produces 5-methycytosine (5-mC), is an extensively characterized mechanism for epigenetic gene regulation. Neoplastic cells may simultaneously harbor widespread genomic hypomethylation and regional areas of hypermethylation. The prevalence of global hypomethylation in many types of human cancers suggests that such hypomethylation plays an important role in tumorigenesis. Global DNA hypomethylation may induce neoplastic transformation, genomic instability, and abnormal chromosomal structures. Alterations in normal DNA methylation patterns are frequently observed in cancer cells, and hypermethylation in the promoter regions of tumor suppressor genes is associated with an epigenetically mediated gene silencing. This is a common feature of human carcinomas, and is a key to the tumorigenic process, contributing to all of the typical hallmarks of a cancer cell. With regard to abnormal DNA methylation in lung cancer, many literatures have found that tumor-related genes hypermethylation occurred in the tumor tissues. These genes are involved in a number of intracellular signal pathways: cell cycle regulation, cell apoptosis, DNA repair, cell adhesion and so on. Thus, epigenetic mechanisms representative of DNA methylation and genetic mechanisms are of equal importance during the process of lung cancer initiation, development, and progression. However, research about DNA methylation in precancerous tissues of lung cancer is relatively small and the underlying mechanisms are unclear.Previous studies have suggested that there are many similarities in not only the morphological process, but also the level of molecular biology, alterations in 3-methylcholanthrene (MCA) and diethylnitrosamine (DEN)-induced lung squamous cell carcinoma and human lung cancer. Therefore, the animal model may be particularly informative and helpful in gaining a better understanding of the epigenetic alterations that occur during human lung carcinogenesis.Based on these considerations, the aim of this study was to determine the role of DNA methylation alterations in the progression of MCA and DEN-induced rat lung carcinogenesis from normal bronchial epithelium to cancer. We focused on the changes in DNA methylation associated with the progression of lung carcinogenesis, including the global methylation status, the presence of differentially methylated DNA fragments, the alterations in methylation of tumor-related genes and their regulation of protein expression, DNA methyltransferases changes and their expressions association with tumor-related genes methylation, and demethylation experiments.Materials and Methods:1. Wistar rats of both sexes (6 weeks old, 200±20g) randomly assigned to two groups. The 80 animals assigned to Group 1 were treated with a single dose of 10mg MCA and 10μl DEN in iodized oil by left intra-bronchial instillation. The 10 animals assigned to Group 2 were instilled with 0.1 ml iodized oil without any carcinogen into the left lung as a control. Rats were killed by exsanguination under anesthesia on days 15, 35, 55, 65 and 75 after instillation. The left lungs were split into two halves. One was rapidly frozen in liquid nitrogen and kept at -80℃. The other was fixed by immersion in 4% formaldehyde in phosphate buffer for 24 hours, embedded in paraffin, sectioned to 4μm and 10μm thickness, and routinely processed for hematoxylin and eosin staining. Laser capture microdissection (LCM) was employed to obtain near-pure normal, precancerous and malignant cells for methylation analysis.2. The status of genomic methylation during rat lung carcinogenesis was detected by immunohistochemistry with anti-5-methycytosine (5-mC) antibody, and the mean optical density (OD) and integrated optical density (IOD) were measured by image analysis system. We also used methylation-sensitive arbitrarily-primed PCR (MS-AP-PCR) to screen differentially methylated DNA fragments in 11 tumor and 11 precancerous tissues, and their paired normal bronchial epithelial tissues isolated using LCM. Differentially methylated fragments products were separated cloned into vector and sequenced. Sequence data were used to determine genomic information, including homology to characterized or novel genes and cytogenetic map positions, employing the NCBI’s BLAST program.3. Methylation-specific PCR (MSP) and sequencing were used to detect the methylation status of tumor-related genes, including p15, p16, p27, p57, RASSF1A, TSLC1, TIMP-3, E-cadherin, N-cadherin, DAPK1, FHIT and SOCS-3 during MCA and DEN-induced rat lung carcinogenesis. Immunohistochemistry was used to examine the expressions of p16, p27, p57, RASSF1A, TSLC1, TIMP-3, N-cadherin, DAPK1, FHIT and SOCS-3 proteins during rat lung carcinogenesis. Normal tissues, precancerous and tumor tissues were examined by Western blot for the expressions of these proteins.4. Immunohistochemistry was used to examine the expressions of DNA methyltransferases (DNMTs) 1, 3a and 3b proteins during MCA and DEN-induced rat lung carcinogenesis. Correlation analyses were applied to study the relationship between DNMTs expressions and tumor-related genes methylation. After purification through a series of pathologic, morphological and immunologic identifications, primary rat lung squamous cell carcinoma cells with methylated p16, p57 and TSLC1 were cultured and exposed to different concentrations (2μM, 5μM, 10μM) of 5-aza-2′-deoxycytidine (5-aza-dC). DNA and RNA were harvested on day 3 after initial treatment. The methylation status and mRNA expression of p16, p57 and TSLC1 were detected by MSP and reverse transcription-polymerase chain reaction (RT-PCR).5. Statistical analyses were performed with the SPSS 13.0 software. Chi-square analyses were applied to study the correlation between expression and CpG methylation, the differences in methylation, and differences in expression between normal, precancerous, and tumor tissues. Multiple comparisons in 5-mC immunostaining intensity were evaluated using one-way ANOVA and significant differences between two groups were analyzed by Tukey’s test. The level of significance was set at P < 0.05.Results:1. Rats treated with carcinogens had no obviously pathological lesions in any of the main organs, except the treated lung. No pathological changes were found in any of the control groups. We obtained different morphological tissues representative of hyperplasia, squamous metaplasia, dysplasia, CIS and infiltrating carcinoma in the multistep process of rat lung tumorigenesis by selecting different durations of time after instillation of the carcinogens to sacrifice the animals. All of the lung cancers induced by MCA and DEN were diagnosed as squamous cell carcinoma by a professional pathologist. In all, 20 samples of normal bronchial epithelium (including 10 samples from chemically-treated and 10 samples from untreated rats), 25 samples of hyperplasia, 27 samples of squamous metaplasia, 37 samples of dysplasia, 30 samples of CIS and 25 samples of infiltrating carcinoma were included.2. Anti-5-mC antibody was expressed in the nucleus of bronchial epithelial cells. Staining levels gradually decreased from normal with nucleus brown to tumor with nucleus yellow during the process of carcinogenesis. Nuclear staining in basal cells was deeper than that in luminal cells of normal and precancerous stages. The OD and IOD of combined 5-mC scores of different types of tissues decreased gradually during the progression of carcinogenesis (P < 0.01). The degree of global methylation was, in general, higher in basal cells compared to luminal cells of normal and precancerous tissues (P < 0.01).3. We chose 11 tumor and 11 precancerous tissues, and their paired normal bronchial epithelial tissues, for screening differentially methylated DNA fragments by MS-AP-PCR following LCM. Results showed that abnormal methylation fragments can be found in majority of samples. In the range of 200bp-700bp, we identified a total of eight differentially methylated DNA fragments, including seven hypomethylated and one hypermethylated DNA fragment in both precancerous tissues and tumor tissues. DNA sequence analysis of these fragments revealed that seven of the eight DNA fragments had significant homology matches (99-100% homology) to sequences in the GENBANK database after a BLAST search. These sequences were found to reside on rat chromosomes 1, 3, 9, 12, 13, 20 and X, indicating putative target genes in these regions. The methylation status of hypermethylated fragment CpG islands, which confirmed the MS-AP-PCR results.4. p15 and E-cadherin were unmethylated in normal, precancerous and tumor tissues. Other tumor-related genes methylation gradually increased during the process of carcinogenesis. The prevalence of DNA methylation of at least one gene and the average number of methylated genes were significantly increase from normal to precancerous and tumor tissues. The frequency of each tumor-related gene methylation in the stage of normal, hyperplasia, squamous metaplasia, dysplasia, CIS and infiltrating carcinoma was as follows: p16: 0%, 8.00%, 22.22%, 40.54%, 50.00% and 56.00%; p27: 0%, 0%, 0%, 0%, 3.33% and 8.00%; p57: 0%, 0%, 11.11%, 18.92%, 26.67% and 36.00%; RASSF1A: 0%, 0%, 3.70%, 8.11%, 6.67% and 16.00%; TSLC1: 0%, 0%, 18.52%, 24.32%, 40.00% and 48.00%; TIMP-3: 10.00%, 12.00%, 18.52%, 24.32%, 30.00% and 60.00%; N-cadherin: 0%, 0%, 7.41%, 18.92%, 26.67% and 36.00%; DAPK1: 0%, 0%, 7.41%, 10.81%, 26.67% and 36.00%; FHIT: 0%, 0%, 3.70%, 16.22%, 43.33% and 48.00%; SOCS-3: 0%, 0%, 7.41%, 16.22%, 26.67% and 48.00%. On the contrary, protein expression of tumor-related genes decreased during the process of carcinogenesis except the increase of N-cadherin expression. The rates of positive protein expression in the stage of normal, hyperplasia, squamous metaplasia, dysplasia, CIS and infiltrating carcinoma were as follows: p16: 90.00%, 84.00%, 70.37%, 59.46%, 40.00% and 32.00%; p27: 95.00%, 88.00%, 74.07%, 67.57%, 60.00% and 44.00%; p57: 95.00%, 92.00%, 81.48%, 75.68%, 63.33% and 56.00%; RASSF1A: 100.00%, 92.00%, 70.37%, 64.86%, 53.33% and 48.00%; TSLC1: 95.00%, 88.00%, 55.56%, 48.65%, 43.33% and 40.00%; TIMP-3: 100.00%, 96.00%, 77.78%, 70.27%, 53.33% and 24.00%; N-cadherin: 0%, 0%, 14.81%, 21.62%, 33.33% and 44.00%; DAPK1: 100.00%, 100.00%, 88.89%, 81.08%, 70.00% and 56.00%; FHIT: 100.00%, 96.00%, 81.48%, 62.16%, 43.33% and 36.00%; SOCS-3: 100.00%, 100.00%, 88.89%, 86.49%, 73.33% and 56.00%. Results of Western blot showed that samples with unmethylation and positive by immunohistochemistry showed strong protein expression, while samples with methylation and negative by immunohistochemistry showed weak or no expression. The correlation between methylation of p16, p57, TSLC1, TIMP-3, DAPK1, FHIT, SOCS-3 and their protein expressions was significant.5. The rates of positive DNMTs protein expression in the stage of normal, hyperplasia, squamous metaplasia, dysplasia, CIS and infiltrating carcinoma were as follows: DNMT1: 0%, 16.00%, 29.63%, 40.54%, 46.67% and 64.00%; DNMT3a: 0%, 0%, 18.52%, 37.84%, 40.00% and 68.00%; DNMT3b: 0%, 0%, 0%, 5.41%, 10.00% and 12.00%. The expression of DNMT1 was significant positively correlated with DNMT3a and DNMT3b expression (P < 0.05). The expression of DNMT3a was not correlated with DNMT3b expression (P > 0.05). Correlation analyses between DNMT1, DNMT3a, DNMT3b expression and 10 tumor-related genes with hypermethylation showed that the expression of DNMT1 and DNMT3a were significantly positively correlated with p16, p57 and TSLC1 gene methylation, respectively (P < 0.01). In addition, RASSF1A, TIMP-3, N-cadherin gene methylation was significantly positively correlated with the expression of DNMT3a (P < 0.01). The average numbers of methylated genes were significantly higher in samples with positive expression of DNMT1 and DNMT3a (2.43±1.85 and 2.65±1.82) than those with negative expression of DNMT1 and DNMT3a (1.35±1.98 and 1.34±1.95), respectively. After the primary tumor cell lines were treated with 5-aza-dC, the mRNA expression of p16, p57 and TSLC1 significantly increased compared to the untreated cells.Conclusions:1. Using bronchial instillation, two kinds of carcinogens MCA and DEN successfully established animal model for lung squamous cell carcinoma model in Wistar rats.2. Global hypomethylation not only occurred in tumors, but also in precancerous tissues. The decrease in the degree of genomic methylation and coexist of hypermethylation and hypomethylation may play an important role during carcinogenesis of lung squamous cell carcinoma in rats. Screened differential methylated fragments may locate in sensitive regions that are altered during MCA and DEN-induced lung carcinogenesis. In the future, these sequences could be explored as potential biomarkers of carcinogen exposure. Identification of Hyper-1 suggests that it might be new gene and the aberrant hypermethylation may be the main mechanism of gene inactivation.3. Dynamic changes in hyperthylation of cell cycle regulation-related genes p16, p27, p57, RASSF1A, cell adhesion-related gene TSLC1, TIMP-3, N-cadherin, cell apoptosis-related genes DAPK1, FHIT, and SOCS-3 gene, accounting for their defective expression, are an early and frequent event in tumorigenesis and play an important role during the progression of MCA/DEN-induced multistep rat lung carcinogenesis.4. DNMT1 and DNMT3a overexpression associated with accumulation of DNA methylation of multiple tumor-related genes are involved in multistage carcinogenesis from normal to early precancerous stages to malignant progression.5. DNA methylation is an important mechanism for tumor-related genes inactivation and CpG island hypermethylation leading to gene silencing is a reversible process.To sum up, the present investigation explored the underlying epigenetic mechanisms of chemically induced rat lung carcinogenesis and enriched the traditional“cancer genetics theory”. The experimental model used in present study may be particularly informative and helpful in gaining a better understanding of the epigenetic alterations that occur during genotoxic carcinogen-induced lung carcinogenesis. Established experimental system has laid a foundation for further comprehensive assessment of environmental factors and DNA methylation interactions associated with the occurrence of lung cancer. It also plays a key role in traditional evaluation system for carcinogenic compounds based on the“mutation detection”. At the same time, this study provide information for further study on how related genes on the signal pathway involved in cancer through epigenetic/genetic interactions. The experimental model can be used for screening new demethylation agents for lung cancer treatment and evaluating the interaction between DNA methylation and environmental factors in the mechanism of lung cancer in China.更多还原