乳腺癌相关lncRNA-miRNA-mRNA共表达及关键基因网络构建预测  

Prediction of breast cancer related lncRNA-miRNA-mRNA co-expression and key genes based on network construction method

作　　者：侯敏[1] 蒋琳[1] 詹红梅[1] 皈燕[1] 赵妍丽[1] 马代远[1] 谭榜宪[1] HOUMin;JIANG Lin;ZHAN Hong-mei;BAN Yan;ZHAO Yan-li;MA Dai-yuan;TAN Bang-xian(Department of Oncology ,Affiliated Hospital of North Sichuan Medical College, Nanchong 637000, P. R. Chin)

机构地区：[1]川北医学院附属医院肿瘤科,四川南充637000

出　　处：《中华肿瘤防治杂志》2018年第1期26-33,共8页Chinese Journal of Cancer Prevention and Treatment

基　　金：四川省卫生计生委科研基金(16PJ135)

摘　　要：目的潜在治疗靶点的筛选鉴定是肿瘤研究的关键。本研究利用癌症和肿瘤基因图谱计划(The Cancer Genome Atlas,TCGA)中RNA-Seq数据构建乳腺癌相关长链非编码RNA(long non-coding RNA,lncRNA)-微小RNA(microRNA,miRNA)-mRNA共表达网络,并筛选验证其中的关键基因。方法对TCGA中1 213例乳腺样本RNA-Seq数据及临床资料进行整理。提取其中113例乳腺癌及癌旁配对样本分析其中lncRNA、miRNA及mRNA的表达差异。利用WGCNA计算上述差异基因在1 100例乳腺癌样本中表达的相关性并构建共表达网络。通过排列网络中基因与基因连接的数量关系分别筛选出关键的lncRNA、miRNA及mRNA,并利用本地标本库乳腺癌样本对关键基因在癌和癌旁的表达进行验证。最后,结合TCGA的临床资料采用Kaplan-Meier分析不同表达患者的生存率差异。结果在113例配对乳腺癌及癌旁样本中共发现4 582个差异基因,其中3 514个表达上调,1 068个表达下调。对上述差异基因的聚类分析可以有效地区分肿瘤与正常组织。构建的共表达网络包括739个节点及17 375个连接,其中lncRNA 65个,miRNA 3个,mRNA 671个,占整个差异表达基因的16.1%。网络中的关键基因CDK1(P<0.001)、RP11-214F16.8(P=0.048)和MIR27A(P=0.027)在肿瘤组织中的表达均高于正常癌旁组织。低表达RP11-214F16.8、MIR27A患者总体生存显著好于高表达患者,均P值<0.001;CDK1的表达与预后有一定程度的关联,但尚未达到有统计学意义的水平,P=0.061。结论乳腺癌相关lncRNA-miRNA-mRNA共表达网络为后续更深入研究乳腺癌关键基因及基因间的调控关系提供了参考和方向。通过共表达网络筛选关键基因是一种兼具高效性和生物学意义的方法。OBJECTIVE Screening and identification of potential therapeutic targets is a key task for cancer re- search. In the current study, RNA-seq data form TCGA database was used to construct lncRNA-miRNA-mRNA co-ex- pression network associated with breast cancer and identity the hub genes. METHODS The RNA-Seq data and clinical in- formation of 1 213 breast cancers patients was obtained and organized,among which 133 tumor and tumor-adjacent normal samples were extracted to analyze the differentially expressed lncRNA,miRNA and mRNA. WGCNA algorithm was con- ducted to calculate the correlation of the differentially expressed genes among 1 100 breast tumor samples and construct co-expression network. Hub lncRNA,miRNA and mRNA were screened over sorting the number of connection between each genes in the network. Additionally, the expression of each hub genes between tumor and adjacent normal tissues was further verified through local breast cancer tissue. Finally, the expression data and clinical information form TCGA were combined. Survival analysis was performed to investigate the prognostic value of the identified hub genes. RESULTS A total of 4 582 differentially expressed genes were screened,among which 3 514 genes ware up-regulated, 1 068 genes were down-regulated. The results of cluster analysis on above genes could differentiate tumor and normal tissue effectively. The co-expression network contained 793 nodes and 17 375 edges. There were 65 lncRNAs,671 mRNAs and 3 miRNA in the network which counted for 16.1% of all differentially expressed genes. Compared with the expression in the normal tis- sues,the expression of hub genes （CDK1, RP11-214F16. 8 and MIR27A） in the tumor tissues were significant higher （P〈0. 001,P= 0. 048,P= 0. 027）. The overall survival of patients with higher expressed RPll-214F16.8 and MIR27A was significant poor than the lower expression patients. To a certain extent, the expression of CDK1 was correlated with prognosis,but which was not statistically significant （P=0. 061）. CONC

关 键 词：乳腺癌 共表达网络 关键基因 lncRNA-miRNA MRNA 

分 类 号：R737.9[医药卫生—肿瘤]