Identifying a novel biological mechanism for alcohol addiction associated with circRNA networks acting as potential miRNA sponges

Abstract

Our lab and others have shown that chronic alcohol use leads to gene and miRNA expression changes across the mesocorticolimbic (MCL) system. Circular RNAs (circRNAs) are noncoding RNAs that form closed‐loop structures and are reported to alter gene expression through miRNA sequestration, thus providing a potentially novel neurobiological mechanism for the development of alcohol dependence (AD). Genome‐wide expression of circRNA was assessed in the nucleus accumbens (NAc) from 32 AD‐matched cases/controls. Significant circRNAs (unadj. p ≤ 0.05) were identified via regression and clustered in circRNA networks via weighted gene co‐expression network analysis (WGCNA). CircRNA interactions with previously generated mRNA and miRNA were detected via correlation and bioinformatic analyses. Significant circRNAs ( N = 542) clustered in nine significant AD modules (FWER p ≤ 0.05), within which we identified 137 circRNA hubs. We detected 23 significant circRNA–miRNA–mRNA interactions (FDR ≤ 0.10). Among these, circRNA‐406742 and miR‐1200 significantly interact with the highest number of mRNA, including genes associated with neuronal functioning and alcohol addiction ( HRAS, PRKCB, HOMER1, and PCLO). Finally, we integrate genotypic information that revealed 96 significant circRNA expression quantitative trait loci (eQTLs) (unadj. p ≤ 0.002) that showed significant enrichment within recent alcohol use disorder (AUD) and smoking genome‐wide association study (GWAS). To our knowledge, this is the first study to examine the role of circRNA in the neuropathology of AD. We show that circRNAs impact mRNA expression by interacting with miRNA in the NAc of AD subjects. More importantly, we provide indirect evidence for the clinical importance of circRNA in the development of AUD by detecting a significant enrichment of our circRNA eQTLs among GWAS of substance abuse.

Abstract

We explore circRNA expression changes within the postmortem brains of chronic alcohol abusers and their role in regulating mRNA expression through acting as miRNA sponges. Our methods included integrating WGCNA and genotypic information with a multistep bioinformatics and statistical analysis to identify significant circRNA–miRNA–mRNA interactions associated with alcohol dependence (AD). Overall, we implicate circRNA as potential regulators of known AD risk genes by intersecting correlated expression with in silico binding predictions and GWAS enriched eQTL identification.