食管癌差异表达基因的生物信息学分析

Bioinformatics analysis of differentially expressed genes in esophageal cancer

目的:应用生物信息学技术挖掘食管癌差异表达基因,为食管癌治疗提供新靶点。方法:从GEO数据库中下载基因芯片数据集GSE20347、GSE92396和GSE1420,使用在线分析工具GEO2R筛选出食管癌组织与正常食管组织的差异表达基因。利用在线数据库DAVID和STRING分别进行功能、通路富集分析和蛋白互作分析,使用Cytoscape软件来筛选蛋白相互作用网络中的核心网络基因,并在TCGA数据库中验证其差异表达情况。结果:本实验共发现274个差异表达基因,269个基因有相同的表达趋势,其中96个上调基因,173个下调基因(调整后P<0.05,|log2FC|>1)。通过GO富集分析(P<0.05)发现它们主要参与了表皮发育、肽键交联结合、角质细胞分化、细胞外基质组织生成等生物学过程。分子功能包括细胞外基质结构组分、分子活性、钙离子及血小板源性生长因子结合等的功能;而细胞组成分析提示这些差异基因主要集中在细胞外基质。KEGG富集通路分析(P<0.05)显示主要的信号通路包括阿米巴病信号通路、细胞外基质作用信号通路、糖酵解和糖异生等信号通路过程。MCODE分析发现金属蛋白酶组织抑制因子-1 (TIMP metallopeptidase inhibitor 1,TIMP1)、基质金属蛋白酶-3 (matrix metallopeptidase 3,MMP3)、分泌性磷酸化蛋白-1(secreted phosphoprotein 1,SPP1)、丝氨酸蛋白酶抑制蛋白E1(serpin family E member 1,SERPINE1)、骨膜蛋白(periostin,POSTN)、类胰岛素生长因子结合蛋白-3(insulin like growth factor binding protein 3,IGFBP3)、Ⅲ型胶原α1(collagen typeⅢalpha 1 chain,COL3A1)、沉默桥粒芯糖蛋白-2(desmoglein 2,DSG2)这8个基因可能为诱导食管癌发生发展的关键基因(P<0.05)。通过TCGA数据库验证这8个基因,进一步筛选出COL3A1、IGFBP3、MMP3、SPP1、TIMP1关键基因,且这5个基因同时在食管鳞状细胞癌和腺癌中高表达(P<0.05)。结论:采用生物信息学方法能够有效分析食管癌与正常食管组织差异表达的基因,本次研究中共筛选出COL3A1、IGFBP3、MMP3、SPP1、TIMP1等5个关键基因,表明其可能是食管癌发病机制的新的生物标记物。

Objective:To identify differentially expressed genes in esophageal cancer as targets for further treatment of esophageal cancer. Methods:Gene data sets GSE20347,GSE92396 and GSE1420 were downloaded from the GEO database,and differentially expressed genes in esophageal cancer tissues and normal esophageal tissues were selected on the basis of GEO2 R analysis tools. GO analysis and KEGG pathway analysis were performed using the DAVID online database. the protein-protein interaction analysis was carried out via the STRING online database. Cytoscape software was used to screen the core network genes in the protein-protein interaction network,and the differential expression was verified in the TCGA database. Results:A total of 274 differentially expressed genes were revealed. 269 genes had the same expression trend,of which 96 were up-regulated and 173 were down-regulated(P<0.05 after adjustment,|log2 FC|>1). GO enrichment analysis(P<0.05) showed they were mainly involved in biological processes such as epidermal development,peptide bond cross-linking,keratinocyte differentiation,and extracellular matrix tissue formation. Molecular functions include extracellular matrix structural components and molecular activities,the function of calcium ion and platelet-derived growth factor binding,and cell composition analysis suggested that these differential genes are mainly concentrated in the extracellular matrix. The KEGG enrichment pathway analysis(P<0.05) showed that the main signaling pathways include those of amebiasis,the extracellular matrix,glycolysis and gluconeogenesis. MCODE analysis revealed that TIMP metallopeptidase inhibitor 1(TIMP1),matrix metallopeptidase 3(MMP3),secreted phos-phoprotein 1(SPP1),serpin family E member 1(SERPINE1),periostin(POSTN),insulin like growth factor binding protein 3(IGFBP3),collagen type Ⅲ alpha 1 chain(COL3 A1) and desmoglein 2(DSG2) may be the key genes for the development of esophageal carcinoma(P<0.05). These eight genes were verified by TCGA database,and COL3 A1,IGFBP3,MMP3,SPP1 and TIMP1 were further screened as key genes,all of which were highly expressed in esophageal squamous cell carcinoma and adenocarcinoma(P<0.05).Conclusion:Bioinformatics method can effectively screen the differentially expressed genes between esophageal cancer and normal esophageal tissue. In this study,five genes including COL3 A1,IGFBP3,MMP3,SPP1 and TIMP1 were screened,indicating that they may be new biomarkers for the pathogenesis of esophageal cancer.