Anthracyclines induce global changes in cardiomyocyte chromatin accessibility that overlap with cardiovascular...

Breast cancer drugs including anthracyclines (ACs) and Trastuzumab increase the risk for cardiovascular diseases (CVDs) such as atrial fibrillation (AF) and heart failure (HF) that ultimately affect the heart muscle. These CVDs are associated with hundreds of genetic variants in non-coding regions of the genome. However, how these drugs affect the regulatory potential of the non-coding genome of the heart and CVD risk loci is unknown. We therefore measured global chromatin accessibility across iPSC-derived cardiomyocytes derived from four healthy individuals that we treated with topoisomerase II (TOP2) inhibiting ACs, Doxorubicin, Epirubicin, and Daunorubicin, and the anthracenedione, Mitoxantrone as well as the TOP2-independent monoclonal antibody Trastuzumab, for three and 24 hours. We identified tens of thousands of open chromatin regions that are differentially accessible in response to TOP2 inhibitor treatments over time, and no changes in response to Trastuzumab. Early AC-responsive regions are promoter-proximal and are enriched for regions bound by TOP2 beta. Late AC-responsive regions are enriched for AC response genes particularly those associated with mismatch repair. AC-response regions near AC response genes are enriched for FOS:JUNB transcription factor motifs. Three AC-induced cardiotoxicity-, 28 AF- and 14 HF-associated SNPs directly overlap late AC-responsive regions. Early AC-responsive regions are enriched for AF SNPs including rs3176326, which is also associated with HF, and is an eQTL for CDKN1A in heart tissue. This SNP associates with increased chromatin accessibility at a TOP2 beta-bound region, increased histone acetylation, and increased CDKN1A expression in response to all ACs. Our results demonstrate large-scale changes in chromatin accessibility in cardiomyocytes treated with ACs, which correspond to several regions harboring AF and HF risk loci. The identified drug-responsive chromatin regions can be used to annotate variants in cancer patient populations to contribute to risk estimation for CVD.