Identification of atrial fibrillation-associated lncRNAs and exploration of their functions based on WGCNA and ceRNA network analyses.

ABSTRACT

Atrial fibrillation (AF) is a common cardiac arrhythmia that induces serious complications. However, pharmacological treatments of AF remain challenging. This study aimed to screen crucial long non-coding RNAs (lncRNAs), microRNAs (miRNAs) and messenger RNA (mRNAs) for AF using the microarray datasets (lncRNAs and mRNAs GSE79768, GSE115574; miRNAs GSE68475) collected from the Gene Expression Omnibus database. Weighted correlation network analysis of GSE79768 and GSE115574 datasets identified five modules were highly related to AF status. Among 118 module-related differentially expressed mRNAs, FBXW7, EGFR, CXCR2, ROCK1 and UBE2D1 were considered as hub genes according to the gene significance, module membership and the topological characteristics for the nodes in the protein-protein interaction network. lncRNA MIR100HG and LINC01105 may function by co-expressing with (MIR100HG-ROCK1/FBXW7/UBE2D1, LINC01105-EGFR) mRNAs or sponging miRNAs to regulate mRNAs (LINC01105-miR-125a-3p-EGFR, MIR100HG-miR-200b-3p- FBXW7, MIR100HG-miR-561-3p-CXCR2, MIR100HG-miR-548z-UBE2D1). Connectivity Map and Comparative Toxicogenomics Database searches predicted dexamethasone may treat AF by reversing the expression of MIR100HG; artemisinin may reverse the expression of hub DEGs. In conclusion, our results may provide novel molecular mechanisms and potential therapeutic targets and drugs for AF.