[Bioinformatics analysis of differently expressed genes in osteoblastic sarcoma and screening of key genes].

Objective: To screen the different expressed genes between osteosarcoma and normal osteoblasts, and find the key genes for the occurrence and development of osteosarcoma. Methods: The gene expression dataset GSE33382 of normal osteoblasts and osteosarcoma was obtained from Gene Expression Omnibus (GEO) database. The different expressed genes between normal osteoblasts and osteosarcoma were screened by limma package of R language, and the different expressed genes were analyzed by Kyoto encyclopedia of genes and genomes (KEGG) pathway enrichment analysis. The protein interaction network was constructed by the String database, and the network modules in the interaction network were screened by the molecular complex detection (MCODE) plug-in of Cytoscape software. The different expressed genes contained in the first three main modules screened by MCODE were analyzed by gene ontology (GO) using the BiNGO module of Cytoscape software. The MCC algorithm was used to screen the top 10 key genes in the protein interaction network. The gene expression and survival dataset GSE39055 of osteosarcoma was obtained from GEO database, and the survival analysis was performed by Kaplan-Meier method. The data of 48 patients with osteosarcoma treated in the First Affiliated Hospital of Fujian Medical University from January 2005 to December 2015 were selected for verification. The expression of STC2 protein in osteosarcoma was detected by immunohistochemical method, and the survival analysis was carried out combined with the clinical data of the patients. Results: A total of 874 different expressed genes were identified from GSE33382 dataset, including 402 down-regulated genes and 472 up-regulated genes. KEGG enrichment analysis showed that different expressed genes were mainly related to p53 signal pathway, glutathione metabolism, extracellular matrix receptor interaction, cell adhesion molecules, folate tolerance, and cell senescence. The top 10 key genes in the interaction network were GAS6, IL6, RCN1, MXRA8, STC2, EVA1A, PNPLA2, CYR61, SPARCL1 and FSTL3. STC2 was related to the survival rate of patients with osteosarcoma (P<0.05). The results showed that the expression of STC2 protein was related to tumor size and Enneking stage in 48 cases of osteosarcoma. The median survival time of 25 cases with STC2 high expression was 21.4 months, and that of 23 cases with STC2 low expression was 65.4 months. The survival rate of patients with high expression of STC2 was lower than that of patients with low expression of STC2 (P<0.05). Conclusions: Bioinformatics analysis can effectively screen the different expressed genes between osteosarcoma and normal osteoblasts. STC2 is one of the important predictors for the prognosis of osteosarcoma.目的： 筛选成骨型骨肉瘤与正常成骨细胞的差异表达基因，寻找成骨型骨肉瘤发生、发展的关键基因。 方法： 从Gene Expression Omnibus(GEO)数据库获得正常成骨细胞和骨肉瘤组织的基因表达数据集GSE33382，采用R语言的limma包筛选正常成骨细胞和成骨型骨肉瘤的差异表达基因，对差异表达基因进行京都基因与基因组百科全书(KEGG)通路富集分析。采用String数据库构建蛋白交互网络，应用Cytoscape软件的分子复合物检测(MCODE)插件筛选互作网络中的网络模块，应用Cytoscape软件中的BiNGO模块对MCODE筛选得到的前3个主要模块所包含的差异表达基因进行基因本体(GO)富集分析。采用MCC算法筛选蛋白互作网络中前10个关键基因。从GEO数据库获得骨肉瘤的基因表达和生存数据集GSE39055，采用Kaplan-Meier法进行生存分析。选取2005年1月至2015年12月福建医科大学附属第一医院收治的48例成骨型骨肉瘤患者的资料进行验证，采用免疫组化法测定骨肉瘤组织中STC2蛋白的表达，结合患者的临床资料进行生存分析。 结果： 从GSE33382数据集中共鉴定出差异表达基因874个，其中下调基因402个，上调基因472个。KEGG富集分析显示，差异表达基因主要与p53信号通路、谷胱甘肽代谢、细胞外基质受体相互作用、细胞黏附分子、抗叶酸盐耐受、细胞衰老等通路有关。互作网络中最重要的前10个关键基因为GAS6、IL6、RCN1、MXRA8、STC2、EVA1A、PNPLA2、CYR61、SPARCL1和FSTL3，其中只有STC2的表达与骨肉瘤患者的生存率有关(P<0.05)。验证结果显示，48例成骨型骨肉瘤组织中STC2蛋白表达与肿瘤大小、Enneking分期有关(均P<0.05)。25例STC2高表达，中位生存时间为21.4个月；23例STC2低表达，中位生存时间为65.4个月，STC2高表达组患者的生存率低于低表达组(P<0.05)。 结论： 生物信息学分析方法可有效筛查成骨型骨肉瘤和正常成骨细胞的差异表达基因，STC2是骨肉瘤预后判断的重要预测因子之一。.