Alcohol use disorder (AUD) is a common mental illness with high morbidity and disability. The discovery of laboratory biomarkers has progressed slowly, resulting in suboptimal diagnosis and treatment of AUD. This study aimed to identify promising biomarkers, as well as the potential miRNA-mRNA networks associated with AUD pathogenesis. RNA sequencing was performed on plasma-derived small extracellular vesicles (sEVs) from AUD patients and healthy controls (HCs) to harvest miRNAs expression profiles. Machine learning (ML) models were built to screen characteristic miRNAs, whose target mRNAs were analyzed using TargetScan, miRanda and miRDB databases. Gene Expression Omnibus (GEO) datasets (GSE181804 and GSE180722) providing postmortem hippocampal gene expression profiles of AUD subjects were mined. A total of 247 differentially expressed (DE) plasma-derived sEVs miRNAs and 122 DE hippocampal mRNAs were obtained. Then, 22 overlapping sEVs miRNAs with high importance scores were gained by intersecting 5 ML models. As a result, we established a putative sEVs miRNA-hippocampal mRNA network that can effectively distinguish AUD patients from HCs. In conclusion, we proposed 5 AUD-representative sEVs miRNAs (hsa-miR-144-5p, hsa-miR-182-5p, hsa-miR-142-5p, hsa-miR-7-5p, and hsa-miR-15b-5p) that may participate in the pathogenesis of AUD by modulating downstream target hippocampal genes. These findings may provide novel insights into the diagnosis and treatment of AUD.
Read full abstract