Bioinformatic analysis of differentially expressed profiles of lncRNAs and miRNAs with their related ceRNA network in endometrial cancer.

Abstract

Increasing evidence suggests that long non-coding riboneucleic acids (lncRNAs), as competing endogenous RNA (ceRNA), play a key role in the initiation, invasion, and metastasis of cancer. As a new hypothesis, the lncRNA-micro RNA (miRNA)-messenger RNA (mRNA), ceRNA regulatory network has been successfully constructed in a variety of cancers. However, lncRNA, which plays a ceRNA function in endometrial cancer (EC), is still poorly understood. In this study, we downloaded EC expression profiling from The Cancer Genome Atlas database and used the R software "edgeR" package to analyze the differentially expressed genes between EC and normal endometrium samples. Then, differentially expressed (DE) lncRNAs, miRNAs and mRNAs were selected to construct a lncRNA-miRNA-mRNA prognosis-related regulatory network based on interaction information. The Gene Ontology and Kyoto Encyclopedia of Genes and Genomes pathway enrichment analysis were performed on the genes in the network to predict the potential underlying mechanisms and functions of lncRNAs in EC. Kaplan-Meier method and the log-rank test were used for survival analysis. Based on the "ceRNA hypothesis," we constructed a co-expression network of mRNA and lncRNA genes mediated by miRNA in the process of tumor genesis. Furthermore, we successfully constructed a dysregulated lncRNA-associated ceRNA network containing 96 DElncRNAs, 27 DEmiRNAs, and 74 DEmRNAs. Through Kaplan-Meier curve analysis, we found that 9 lncRNAs, 3 miRNAs, and 12 mRNAs were significantly correlated with the overall survival rate of patients among all lncRNAs, miRNAs, and mRNAs involved in ceRNA (P < .05). Our research provides a new perspective for the interaction among lncRNAs, miRNAs, and mRNA and lays the foundation for further research on the mechanism of lncRNAs in the occurrence of EC.