Abstract

Identifying target genes that are causally linked to coronary artery disease (CAD) remains an elusive problem despite the plethora of CAD genome-wide association study (GWAS) risk loci. Here, we performed single nucleus multiomic (RNA + ATAC) sequencing in 44 human coronary arteries and bulk Hi-C in 16 coronaries from multi-ethnic participants. From sequencing 126,804 nuclei, we identified cell type specific transcriptional signatures, chromatin peaks, and gene regulatory networks through peak to gene linkage analysis. Single-cell quantitative trait loci (QTL) analysis identified cell-specific caQTL’s including cell-type specific caQTLs within CAD GWAS loci. The overlap of CAD GWAS loci with ATAC peaks demonstrated that vascular smooth muscle and endothelial cells harbor the greatest genetic risk. To identify proximal and distal putative target genes for CAD loci in a cell type specific manner, we built a enhancer-gene map by leveraging coronary artery ChiP-seq, cell specific RNA/ATAC-seq, and Hi-C based chromatin contact loops. Using this map, we identified distal target genes linked to CAD GWAS loci. Finally, given the effect of ancestry on CAD GWAS risk interpretation, we leveraged Hi-C data to compare the chromatin architecture in coronary tissue from 8 European Americans vs 8 African Americans. A meta-map, sample level, and local ancestry analysis identified ancestry-specific differences in chromatin looping and topologically associated domains. Collectively, our results provide a cell-type specific transcriptomic and epigenetic map which along with bulk chromatin contacts can be used to interpret and prioritize CAD GWAS candidates and nominate cell specific causal target genes.