Abstract:

Objective To mine the public databases by using bioinformatics methods and construct the competitive endogenous RNA（ceRNA） immune regulatory network of atrial fibrillation （AF）， and to understand the mechanism of occurrence and development of AF. Methods The gene expression data of circular RNA （circRNA）（GSE129409）， micro RNA （miRNA） （GSE28594） and mRNA（GSE41177） in the AF patients and healthy controls were downloaded from Gene Expression Omnibus（GEO）.The differentially expressed circRNA， miRNA， and mRNA were discovered using the “limma” Data package in R software， and were then visualized by using the pertinent databases. The regulatory relationships between the differentially expressed circRNA， miRNA and mRNA were predicted by ENCORI，circBank，TargetScan and miRDB Databases， and the ceRNA regulatory networks were constructed based on the circRNA-miRNA pairs and the miRNA-mRNA pairs.The DAVID database was used for Gene Ontology （GO） and Kyoto Encyclopedia of Genes and Genomes （KEGG） enrichment analysis of differentially expressed mRNA（DEmRNA） to annotate its functions. The best gene characteristics were screened by receiver operating characteristic curve （ROC） curve and the area under curve（AUC） was calculated. CIBERSORT software was used to analyze the infiltration of the immune cells in AF. Results A total of 103 differential circRNAs， 37 differential miRNAs，and 296 mRNAs were identified （|log2（FC）|>1 and P<0.05）， among them 589 miRNAs were predicted to bind to the differentially expressed circRNAs， 9 miRNAs were obtained by intersecting the predicted miRNAs with differentially expressed miRNA（DEmiRNA）， 3 000 target genes of differentially expressed miRNAs were predicted， and 32 differential genes were obtained by intersecting the predicted target genes with differentially expressed genes （DEGs）. Finally，7 circRNAs，5 miRNAs，and 19 mRNAs were assembled into a circRNA-miRNA-mRNA network. The differential genes in the ceRNA network were mainly enriched in the biological processes such as protein degradation， cytosolic spitting action of cells， and protein tyrosine kinase activity.The KEGG enrichment analysis results showed that the DEGs were mainly enriched in the signaling pathways such as chemokine signaling pathway， hedgehog signaling pathway， T lymphocyte receptor signaling pathway and cell-cytokine interaction. The ROC curve results showed that MAL2， STT3B， SHISA3， ZBTB41， CPNE4， EPHA7， hsa_circ_0006562 hsa_circ_0024957， hsa-miR-199a-5p，and hsa-miR-142-3p were identified as the viable candidates for predicting AF （AUC>0.8）. Conclusion The construction of circRNA-miRNA-mRNA network provides a basis for the study of RNA interaction mechanism in AF，and circRNA may be a potential therapeutic target of AF.

Key words: Atrial fibrillation, Gene Expression Omnibus, Immune infiltration, Circular RNA, T lymphocytes