Identification of KHSRP-Regulated RNAs in Esophageal Cancer by Integrated Bioinformatics Analysis.

The objective of this study was to identify key molecules that included long noncoding RNAs (lncRNAs), microRNAs (miRNAs), and mRNAs involved in esophageal cancer with KH-type splicing regulatory protein (KHSRP) knockdown. GSE99422 and GSE99423 from Gene Expression Omnibus database were extracted. After differentially expressed analysis of miRNAs, lncRNAs, and mRNAs, the lncRNAs-mRNAs interaction was obtained. Then the protein-protein interaction (PPI) network and module analyses were performed for differentially expressed mRNAs (DEmRNAs). Combined with miRWalk tool and DIANA-LncBase tool, regulating relationship between differentially expressed miRNAs (DEmiRNAs) and DEmRNAs/DElncRNAs were predicted. Finally, mRNA-miRNA-lncRNA regulatory network construction was established by Cytoscape software. Finally, the key genes were validated based on the Gene Expression Profiling Interactive Analysis GEPIA) database. Totally, 2,027 DEmiRNAs, 3,480 DElncRNAs, and 18,293 DEmRNAs were screened. The PPI network included 399 nodes and 1671 interaction pairs, and two function models (Cluster 1 and Cluster 2) were separately identified. The was a hub in the Cluster 1, which mainly enriched in the function of nuclear division. Then the competing endogenous RNA (ceRNA) network was constructed with 20 miRNAs, 66 lncRNA, and 202 mRNA. Here, lncRNA RP11-159D12.2 might function as a ceRNA in regulating expression of esophagus cancer through competitively binding to hsa-miR-4430. was mainly enriched in the function of mitotic nuclear division. Besides, the expression of and genes was upregulated in tumor tissues compared with normal controls. Also, the correlation analysis showed that the key genes ( and ) were validated to be positively correlated with KHSRP. , , hsa-miR-4430, and lncRNA RP11-159D12.2 were likely to be associated with esophageal cancer development.