鉴定人类肝细胞癌潜在治疗靶基因：基于GEO和TCGA数据库的研究

Identification of Potential Therapeutic Target Gene for Human Hepatocellular Carcinoma: A GEO and TCGA Database Study

摘要: 目的：通过整合公共数据库分析筛选出与人肝细胞癌(HCC)相关的基因，为进一步研究HCC致病机制提供生物学依据并探讨基因在HCC中临床意义。方法：利用癌症基因组图谱数据库(TCGA)和GEO数据库获取肝细胞癌测序数据集及临床信息资料，采用生物信息学方法对数据进行提取、整理和分析，筛选出肝细胞癌差异表达基因(DEGs)，并利用DAVID数据库对DEGs进行基因本体论(GO)及通路富集分析(KEGG)，通过STRING数据库，Cytoscape及其插件cytoHubba，NetworkAnalyzer分析蛋白质互作网络的中心节点蛋白质，寻找关键(Hub)基因。结果：通过对TCGA数据、GSE29721、GSE84402数据集整合取交集获得了43个上调基因和137个下调基因，DEGs主要涉及免疫应答，趋化因子介导的信号通路、炎症反应及细胞表面受体信号通路等生物过程，介导趋化因子活性，免疫球蛋白受体结合等分子过程，主要富集于质膜外区域；KEGG分析结果显示DEGs主要富集于细胞因子–细胞因子受体相互作用等经典信号通路，筛选得到CDC20，RAD51AP1，RRM2，TTK，NUF2，NDC80，CCNB1，KIF11，UBEC2，BUB1B 10个Hub基因。其中HCC组织里NUF2的表达与正常组织中的表达有显著差异(P < 0.05)，且和年龄，肿瘤分级、临床分期显著相关。结论：通过生物信息学方法分析所得的Hub基因及信号通路可能是HCC潜在的治疗靶点，本文为HCC发病机制的进一步研究提供了理论依据。

Abstract: Objective: To screen out genes related to human hepatocellular carcinoma (HCC) by integrating public database analysis, so as to provide biological basis for further study of the pathogenesis of HCC and explore the clinical significance of genes in HCC. Methods: Hepatocellular carcinoma se-quencing data set and clinical information were obtained by using cancer Genome Atlas Database (TCGA) and GEO database. Bioinformatics methods were used to extract, collate and analyze the da-ta. Differentially expressed genes (DEGs) in hepatocellular carcinoma were screened, and gene on-tology (GO) and pathway enrichment analysis (Kyoto Encyclopedia of Genes and Genomes, KEGG) were performed on DEGs using DAVID database. Through the STRING database, Cytoscape and its plugin cytoHubba, NetworkAnalyzer analyzes proteins at the central nodes of the protein interac-tion network, looking for key (Hub) genes. Results: 43 up-regulated genes and 137 down-regulated genes were obtained by integrating the intersection of TCGA data, GSE29721 and GSE84402 data sets. DEGs mainly involves biological processes such as immune response, chemokine-mediated signaling pathway, inflammatory response and cell surface receptor signaling pathway, which me-diate chemokine activity. Molecular processes, such as immunoglobulin receptor binding, are mainly concentrated in the extramural region; KEGG analysis showed that DEGs were mainly en-riched in classic signaling pathways such as cytokine-cytokine receptor interaction. Ten Hub genes including CDC20, RAD51AP1, RRM2, TTK, NUF2, NDC80, CCNB1, KIF11, UBEC2 and BUB1B were screened. The expression of NUF2 in HCC tissues was significantly different from that in normal tis-sues (P < 0.05), and was significantly correlated with age, tumor grade and clinical stage. Conclusion: The Hub gene and signal pathway analyzed by bioinformatics methods may be potential therapeu-tic targets for HCC, which provides a theoretical basis for the progress of the pathogenesis of HCC.