Objective: To construct the competitive endogenous RNA (ceRNA) network related to gastric cancer and explore the molecular mechanism. Methods: The expression profiles of lncRNA, miRNA and mRNA in gastric cancer and paracancer tissues were analyzed by biochip technology, edgeR package in R software was used to filtrate differential expression genes (multiple change of >1.5 times, P<0.05) and volcano map was drawn. Based on the online miRNA-lncRNA prediction tool lncBase database and the miRNA Target gene prediction database (miRTarBase, target-scan, miRDB, starBase), the relationship between miRNA, lncRNA and mRNA was predicted. Cytoscape software was used to construct lncRNA-miRNA-mRNA ceRNA network and key genes (hub genes) were identified based on cytohubba calculation of degree score of each node. Then Hub genes related to the prognosis of gastric cancer were verified in the TCGA database. The GO and KEGG enrichment analysis of differentially expressed mRNA was performed using the online biological information annotation database DAVID, P<0.05 and false discovery rate (FDR)<0.05 were used as cut-off criteria. R software was used to download the RNA sequencing data and mirna-seq data of gastric cancer and adjacent tissues in TCGA database, edgeR package was used to screen out differentially expressed mRNA, miRNA and lncRNA, and some differentially expressed genes in our data were verified. In OncoLnc database, STAD project of TCGA data was selected and hub gene was input. Patients were divided into two groups based on the median value for hub genes and Kaplan-meier analysis was performed. Results: The differentially expressed 766 mRNA, 110 lncRNA and 10 miRNA were screened out, among them 90 mRNA, 4 lncRNA and 6 miRNA were used to construct the ceRNA network, and 2 of the 20 hub genes were related to the prognosis of patients. MLK7-AS1, SPP1, SULF1, hsa-miR-1307-3p were upregulated in gastric cancer tissues from our biochip, while MT2A, MT1X were downregulated, which were consistent with the results of TCGA gastric cancer database. The differentially expressed mRNAs were significantly enriched in the biological process (BP) and the mineral absorption pathway. CHST1 was negatively correlated while miR-183-5p was positively corelated with the survival of patients. Conclusion: The establishment of ceRNA network for gastric cancer is conducive to further understanding of the molecular biological mechanism. CHST1 and miR-183-5p can be used as prognostic factors of gastric cancer.
目的： 构建胃癌相关竞争内源性RNA(ceRNA)网络，探索胃癌发生、发展的分子机制。 方法： 应用生物芯片技术测定胃癌及对应癌旁组织的信使RNA(mRNA)、微小RNA(miRNA)和长链非编码RNA(lncRNA)表达谱，应用R软件中的edgeR包筛选差异表达基因并绘制火山图，基于miRNA－lncRNA在线预测工具lncBase数据库和miRNA靶基因预测数据库预测mRNA、miRNA和lncRNA的相互作用关系，采用Cytoscape软件构建胃癌lncRNA－miRNA－mRNA ceRNA网络，根据cytohubba计算各节点degree得分筛选关键基因(hub基因)。运用注释、可视化和集成发现数据库(DAVID)，对差异表达的mRNA进行基因本体论(GO)功能富集和京都基因与基因组百科全书(KEGG)通路富集分析。在OncoLnc数据库中，选择肿瘤基因组计划(TCGA)数据集的胃癌项目，输入hub基因，以标准化后基因表达量中位数为界值，将样本分为高表达组和低表达组，并进行生存分析。 结果： 共筛出差异表达的mRNA 766个，miRNA 10个，lncRNA 110个，其中90个mRNA、6个miRNA和4个lncRNA构成ceRNA网络，20个hub基因中有2个与患者预后有关。经TCGA胃癌数据库验证，LncRNA MAP3K20的反义RNA 1(MLK7－AS1)、分泌性磷酸化糖蛋白(SPP1)、硫酸酯酶1(SULF1)hsa－miR－1307－3p等基因在胃癌组织生物芯片中高表达，金属硫蛋白2A(MT2A)、金属硫蛋白1X(MT1X)等基因低表达，与TCGA胃癌数据库一致。生物学功能和通路分析显示，差异表达的mRNA主要在生物学过程(BP)和矿物吸收的通路中显著富集。在TCGA胃癌数据中，碳水化合物磺基转移酶1(CHST1)和miR－183－5p均与患者生存时间有关，CHST1表达量越高患者生存时间越短，而miR－183－5p表达量越高患者生存时间越长。 结论： 胃癌ceRNA网络的构建有助于深入了解胃癌的分子生物学机制，CHST1和mi－183－5p可作为胃癌预后的预测指标。.
Keywords: Carbohydrate sulfotransferase1; Competing endogenous RNA; Mi-183-5p; Stomach neoplasms.