Construction of a Competitive Endogenous RNA Network and Identification of Potential Regulatory Axis in Gastric Cancer

Background: Increasing studies has found that long non-coding RNAs (lncRNAs) play critical roles in carcinogenesis, but the underlying mechanisms remain unclear. The aim of this study is to construct a competitive endogenous RNA (ceRNA) network and to identify potential regulatory axis in gastric cancer (GC). Methods: Differentially expressed (DE) mRNAs, miRNAs, and lncRNAs were obtained by analyzing the RNA expression profiles of stomach adenocarcinoma (STAD) retrieved from The Cancer Genome Atlas (TCGA) database. The lncRNA-miRNA-mRNA regulatory networks of GC were constructed by comprehensive bioinformatics methods including functional annotation, RNA-RNA interactomes prediction, correlation analysis, and survival analysis. The interactions and correlations among ceRNAs were validated by experiments on cancer tissues and cell lines. Results: A total of 41 lncRNAs, 9 miRNAs, and 10 mRNAs were identified and selected to establish the ceRNA regulatory network of GC. Several ceRNA regulatory axes, which consist of 18 lncRNAs, 4 miRNAs, and 6 mRNAs, were obtained from the network. A potential ADAMTS9-AS2/miR-372/CADM2 axis which perfectly conformed to the ceRNA theory was further analyzed. qRT-PCR showed that ADAMTS9-AS2 knockdown remarkably increased miR-372 expression but reduced CADM2 expression, whereas ADAMTS9-AS2 overexpression had the opposite effects. Dual luciferase reporter assay indicated that miR-372 could bound to the ADAMTS9-AS2 and the 3'UTR of CADM2. Conclusion: The constructed novel ceRNA network and the potential regulatory axes might provide a novel approach of the exploring the potential mechanisms of development in GC. The ADAMTS9-AS2/miR-372/CADM2 could act as a promising target for GC treatment.