结直肠癌和肺鳞癌预后相关基因的研究

发布时间：2018-07-10 06:14

  本文选题：结直肠发育 + 结直肠癌癌变　； 参考：《北京协和医学院》2016年博士论文

【摘要】：该博士学位论文由相互联系的两部分工作组成,其中第一部分分为三章。第一部分：结直肠癌预后相关基因的研究第一章：利用Spearman转移模型寻找结直肠癌预后相关的免疫相关基因免疫相关基因通过介导炎症和免疫逃逸在结直肠癌发生发展过程中起到非常重要的作用。虽然人类在探求癌症发病分子机制的过程中取得了巨大的进步,但是肿瘤的高度异质大大阻碍了人类对肿瘤复杂生物学进程的探索。胚胎发育和肿瘤进展在生物学行为及分子机制上有很多相似之处,这种相似性使胚胎发育成为一种没有显著异质性却可以研究肿瘤生物学行为的模型。本研究中我们提出造成免疫基因内协同性失调(定义为在癌进展中基因间共表达模式的破坏)的免疫相关基因,可能包含许多结直肠癌的预后相关信息。我们首先建立了137例人类结直肠的基因分子表达谱,样本涵盖了人类结直肠胚胎发育、癌前和癌的不同阶段。同时利用其中的60例样本建立miRNA基因分子表达谱。我们首次提出Spearman转移模型来量化人类结直肠从胚胎发育到癌前到癌变的转化过程中的协同性失调,并利用miRNA-mRNA分子调控网络和机器学习算法识别具有预后信息的重要基因分子。我们最终找到12个免疫基因集合,在5套独立的公共数据中验证了这些基因对预后的预测价值。利用log-rank检验,这12个基因集合与4套数据的总生存显著相关(GSE17536, n=177, p=0.0054:GSE17537, n=55, p=0.0039; GSE39582, n=562, p=0.13; GSE39084, n=70, p=0.11),并且与4套数据的无病生存显著相关(GSE17536,n=177,p=0.0018;GSE17537,n=55,p=0.016; GSE39582, n=557,p=4.4e-05; GSE14333, n=226,p=0.032)。Cox回归的结果证实这12个基因的表达水平是预后结直肠癌总生存(危害比：1.759；95%置信区间：1.126-2.746；p=0.013)和无病生存(危害比：2.116；95%置信区间：1.324-3.380；p=0.002)的独立因素。第二章：结直肠发育中上调的细胞周期相关基因可以预测晚期结直肠癌生存肿瘤可以被理解为在器官发育的过程中产生了分子生物学上的严重失调的一种特殊的器官。胚胎发育和癌变进程的细胞学行为和内在分子事件上存在着惊人的相似性。这种并非偶然的相似性提示我们,胚胎发育可以作为研究肿瘤分子生物学行为,同时规避肿瘤异质性所带来的干扰的一种理想模型。因此,基于全基因组分子表达谱,在胚胎发育中和癌阶段同时上调的或者同时下调的基因,可能蕴含重要的预后信息。本研究从基因表达数据库(Gene Expression Omnibus, GEO)中下载1,539例结直肠癌样本的表达谱数据。在660例结直肠癌和癌旁数据中利用meta分析的方法找到1,396个差异表达的探针。基于人类结直肠胚胎发育的数据,我们把这些差异表达探针分为27个模块中,然后利用基因集富集分析(gene set enrichment analysis, GSEA),我们在差异表达探针中找到393个在胚胎发育和肿瘤中同时高表达的V探针和207个在胚胎发育和肿瘤中同时低表达的A探针。值得注意的是,V探针中有28个细胞周期相关的探针和Ⅲ/Ⅳ期结直肠癌患者的总生存显著相关(GSE17536 cross validation, n=96, p=5.70e-03; GSE29621, n=36,p=1.70e-03; GSE39084, n=38,p=0.05; GSE39582, n=264,p=0.047; GSE17537, n=36, p=5.90e-03)。第三章：受异常启动子区甲基化和基因突变显著影响的发育基因可以预测晚期结直肠癌的总生存癌症的发生和发展是一个及其复杂的生物学过程,包括基因组(如基因突变,拷贝数变化)、DNA甲基化和转录组等很多层面的改变。因此,仅研究一个分子层面不足以解释肿瘤复杂的分子生物学机制。利用TCGA (the Cancer Genome Atlas)公共数据库的资源,我们对结直肠癌DNA拷贝数变化,启动子区域的甲基化,基因突变和基因表达等层面的数据进行了整合研究分析。在本研究中,我们通过各个基因层面上配对的样本找到在不同层面发生显著变化的差异表达基因。值得注意的是,GO (gene ontology)富集的结果显示具有差异启动子区甲基化的和具有基因突变的差异表达基因都可以显著的富集到发育进程的GO term中,暗示胚胎发育和肿瘤进展有着密不可分的联系。因此,利用重启动随机游走(random walk with restart),我们成功找到37个显著的发育相关基因,并在5套独立的表达谱芯片数据中进行验证。Kaplan-Meier和Cox回归分析的结果显示这些基因的表达水平与Ⅲ/Ⅳ期结直肠癌的总生存显著相关。第二部分：肺鳞癌预后相关基因研究肺鳞癌目前在临床中没有有效的靶向药物。因此对肺鳞癌,尤其是肺鳞癌癌前阶段的分子生物学改变的研究十分必要,这可能会为寻找重要的作用靶点或者靶向药物的开发起到巨大的推进作用。另外胚胎发育和癌变阶段在细胞生物学行为和分子特异性改变上有很多相似之处。肿瘤可以理解为一个在正常发育过程中出现严重异常的器官,暗示着胚胎发育可以作为一种研究肿瘤生物学行为的模型。在本研究中,我们收集了10个时间点人类肺组织,涵盖了从胚胎发育到癌前到癌形成的各个阶段,并建立了基因分子表达谱。我们找到癌前和癌阶段一致差异表达的基因,结合基于先验网络的贪婪搜寻算法,利用69例具有预后信息的肺鳞癌手术样本作为训练样本,成功的找出一个含有22个基因的最优子网。这22个基因的表达量于肺鳞癌患者的总生存显著相关。
[Abstract]:The doctoral thesis consists of two parts of interrelated work. The first part is divided into three chapters. Part one: Chapter 1: the first chapter of the study on the prognosis related genes of colorectal cancer: the immuno related genes associated with the prognosis of colorectal cancer by using the Spearman transfer model to mediate inflammation and immune escape in the colorectal cancer The carcinogenesis and development of cancer have played a very important role. Although great progress has been made in the process of exploring the molecular mechanisms of cancer, the high heterogeneity of the tumor greatly hinders human exploration of the complex biological processes of cancer. The development of embryos and the progression of tumors have many phases in biological behavior and molecular mechanisms. Similarly, this similarity makes the embryo develop into a model without significant heterogeneity but can study tumor biological behavior. In this study, we propose an immune related gene that causes co homosexual dysregulation within the immune gene, defined as the disruption of INTERGENE co expression patterns in cancer progression, which may contain a number of colorectal cancer prognosis. We first established 137 cases of human colorectal gene expression profiles, which covered the development of human colorectal embryos, the different stages of precancerous and cancer. At the same time, we used 60 of them to establish the molecular expression profiles of the miRNA gene. We first proposed the Spearman transfer model to quantify the human colorectal cancer from embryonic development to cancer. The synergetic disorder in the process of carcinogenesis, and the use of miRNA-mRNA molecular control networks and machine learning algorithms to identify important gene molecules with prognostic information. We finally found 12 sets of immune genes and verified the predictive value of these genes to the prognosis in 5 sets of independent public data. The use of log-rank test, 12 GSE17536, n=177, p=0.0054:GSE17537, n=55, p=0.0039; GSE39582, n=562, p=0.13; GSE39084, n=70, p=0.11) in the total survival of the 4 sets of data. The results of.Cox regression confirmed that the expression levels of these 12 genes were independent factors for the prognosis of total survival of colorectal cancer (1.759; 95% confidence interval: 1.126-2.746; p=0.013) and disease free survival (2.116; 95% confidence interval: 1.324-3.380; p=0.002). Second: the up-regulated cell cycle related groups in the development of colorectal cancer It is possible to predict that tumor survival in advanced colorectal cancer can be understood as a special organ in the process of organ development. The cytological behavior of embryonic development and the process of carcinogenesis and internal molecular events are astonishingly similar. This is not an accidental similarity that suggests us, Embryo development can be used as an ideal model to study the biological behavior of tumor molecules and to avoid the interference of tumor heterogeneity. Therefore, gene expression based on the whole genome molecular expression, up or down at the same time in the embryonic development and cancer stage may contain important prognostic information. In Gene Expression Omnibus (GEO), the expression profiles of 1539 colorectal cancer samples were downloaded. 1396 differentially expressed probes were found by meta analysis in 660 colorectal and paracancerous data. Based on the data of human colorectal embryogenesis, we divided these differential expression probes into 27 modules. Using gene set enrichment analysis (GSEA), we found 393 V probes which were simultaneously highly expressed in embryonic development and tumors and 207 A probes with low expression in embryonic development and tumor. It is worth noting that there are 28 cell cycle related probes and stage III / IV junctions in V probes. GSE17536 cross validation, n=96, p=5.70e-03; GSE29621, n=36, p=1.70e-03; GSE39084, n=38, p=0.05; Chapter third: the developmental genes which are significantly affected by the methylation of the abnormal promoter region and the gene mutation can predict the late colorectal cancer. The occurrence and development of cancer's total survival is a complex biological process, including a variety of changes in the genome (such as gene mutation, copy number change), DNA methylation and transcriptional groups. Therefore, only one molecular level is not sufficient to explain the complex molecular biological mechanism of the tumor. The use of TCGA (the Cancer Genome Atlas) In this study, we found the differentially expressed genes that have changed significantly at different levels of DNA. It is noted that the results of GO (gene ontology) enrichment show that the differentially expressed genes with differential promoter methylation and gene mutation can be significantly enriched in the GO term of the development process, suggesting that there is an inseparable relationship between the development of the embryo and the progression of the tumor. For this reason, the restarted random walk (random walk with RES) is used. Tart), we successfully found 37 significant developmental related genes and validated.Kaplan-Meier and Cox regression analysis in 5 sets of independent expression chip data. The results showed that the expression levels of these genes were significantly related to the total survival of stage III / IV colorectal cancer. Second: lung squamous cell carcinoma prognosis related genes study lung squamous cell carcinoma at present. There is no effective targeting drug in the clinic. Therefore, it is necessary to study the molecular biological changes in the pre cancerous stage of lung squamous cell carcinoma, especially the squamous cell carcinoma of the lung, which may play an important role in the development of important targets or targeted drugs. In addition, the embryonic development and cancerous stage are in cell biological behavior and molecules. There are many similarities in specific changes. The tumor can be understood as a serious abnormal organ in normal development, suggesting that embryonic development can be used as a model to study the biological behavior of cancer. In this study, we collected 10 time points of human lung tissue, covering from embryo to cancer to cancer form. The gene molecular expression profiles were established at all stages of the formation. We found genes that were identical in precancerous and cancer stages, combined with a priori network based greedy search algorithm, and used 69 cases of lung squamous cell carcinoma with prognostic information as training samples to find an optimal subnet containing 22 genes. These 22 genes were found. The expression level was significantly correlated with the overall survival of patients with lung squamous cell carcinoma.
【学位授予单位】：北京协和医学院
【学位级别】：博士
【学位授予年份】：2016
【分类号】：R734.2;R735.34
，

本文编号：2112397