前列腺癌诊断/预后和耐药分子标志物的筛选及其临床意义

目的：基于已发表的芯片数据通过生物信息学方法筛选差异表达基因，以发现前列腺癌诊断/预后和耐药相关分子标志物。方法：筛选GEO数据库中已发表的前列腺癌mRNA芯片数据GSE6956和前列腺癌细胞多烯紫杉醇耐药mRNA芯片数据GSE33455进行差异表达分析；通过生物学功能注释、基因通路富集分析、蛋白质相互作用网络（protein-protein interaction，PPI）分析等生物信息学方法发现和识别与差异表达基因相关的生物学功能和信号通路；比对TCGA数据库，验证差异表达基因在前列腺癌组织及癌旁组织中的表达，并通过Kaplan-Meier分析差异表达基因对前列腺癌患者生存率的影响；用qPCR方法验证差异表达基因在前列腺癌细胞株PC3及多烯紫杉醇耐药细胞PC3-DTX中的表达情况。结果：共筛选出227个在前列腺癌和前列腺癌多烯紫杉醇耐药细胞芯片数据中共同差异表达基因。差异表达基因主要富集到了癌症相关通路（Lysosome、Sphingolipid、FoxO、Acute myeloid leukemia），并主要参与细胞黏附、自噬和胞内蛋白转运等生物学过程。构建PPI网络选取18个连接度最高的基因作为Hub基因。Hub基因和共同差异表达基因中，上调基因CITED2、LRP12和RPL17-C18orf32与前列腺癌患者的不良预后显著相关。qPCR验证显示CITED2在多烯紫杉醇耐药细胞PC3-DTX中高表达。结论：通过生物信息学方法筛选出在前列腺癌组织和耐药细胞中共同差异表达，且与前列腺癌患者的不良预后密切相关的基因，为前列腺癌诊断/预后和耐药分子标志物的研究提供了新的思路。

Screening and validation of diagnosis/prognosis and drug resistance marker in prostate cancer

Objective: To screen the differentially expressed gene (DEGs) and to identify potential diagnostic/prognostic markers as well as drug resistance markers in prostate cancer (PC) by bioinformatics analysis of published microarray data sets. Methods:Differential expression analyses were performed in available mRNA microarray datasets including prostate cancer tissues dataset GSE6956 and prostate cancer cell taxotere resistance dataset GSE33455 from the GEO database. Gene Ontology enrichment analysis (GO), gene pathway enrichment analysis and protein-protein interaction (PPI) network analysis were performed to identify the biological function and signaling pathways related to DEGs. The expression level of DEGs in PC tissues and para-cancerous tissues was verified by comparing TCGA datasets. Kaplan-Meier method was adopted to detect the influence of DEGs on the survival of PC patients. The expression of DEGs in PC3 cells and taxotere resistant PC3-DTX cells was detected by qPCR. Results: There were a total of 227 genes that differentially co-expressed in taxotere resistant prostate cancer cells and prostate cancer tissues. The functional enrichment analysis showed that these differentially co-expressed genes were mainly enriched in cancer related pathways (Lysosome,Sphingolipid, FoxO, Acute myeloid leukemia) and involved in cell-cell adhesion, autophagy and intracellular protein transportation etc.PPI network screened 18 most connected genes as Hub genes. Among the Hub genes and differentially co-expressed genes, the upregulated CITED2, LRP12 and RPL17-C18orf32 were significantly associated with poor outcomes of PC patients. qPCR validated that CITED2 was upregulated in PC3 and PC3-DTX cells. Conclusion: The present study identified a number of DEGs that differentially co-expressed in prostate cancer tissues and drug resistance cells and significantly associated with the poor prognosis of PC patients by bioinformatical analysis. These results may provide a new idea for identifying potential diagnostic/prognostic and drug resistant markers for prostate cancer.