Network-based prediction and functional validation of metformin for potential treatment of atrial Fibrillation using human inducible pluripotent stem cell-derived atrial-like cardiomyocytes

Abstract

Atrial fibrillation (AF) is a significant cause of morbidity and mortality, and effective therapeutic interventions are lacking. Here, we harness an integrative, network medicine approach to repurpose FDA-approved drugs for AF. We hypothesize that the use of an unbiased method for prioritizing AF drugs using patient transcriptomics data can help to identify alternative therapeutic strategies and mechanism-of-action for these drugs. To achieve this, we first characterized the molecular networks specific to AF by incorporating transcriptomic data of left atrial tissue. We quantified the network proximity of genes differentially expressed in AF to drug targets to identify putative drugs for repurposing. We identified nine high-confidence drug candidates that were validated using enrichment analysis of drug-gene signatures in human cell lines. We identified metformin for the potential treatment of AF and validated its use in human inducible pluripotent stem cell-derived atrial-like cardiomyocytes. We identified AF-specific dysregulated networks enriched in cardiac metabolism, ion transport, and immune pathways that were improved following metformin treatment. In summary, this study utilized network-based approaches for rapid identification of drugs that may be repurposed for AF treatment and validated metformin as a candidate drug using a robust human atrial cell model.