Recent genome-wide association studies (GWAS) have identified multiple vascular cell-expressed genes linked to coronary artery disease (CAD), suggesting that smooth muscle cells (SMCs) and SMC-derived metaplastic cells are promising targets for novel antiatherosclerosis therapies. However, the disease-promoting pathways of most GWAS-identified genes are unknown, hindering their translation into therapeutic targets. This study integrated public GWAS data for CAD and single-cell RNA sequencing (scRNA-seq) analyses of human atherosclerotic plaques to identify 20 GWAS risk genes with a putative mechanism of action in SMCs or SMC-derived cells. Gene perturbation experiments in SMCs coaxed to plaque-relevant phenotypes revealed that the selected risk genes, despite encoding very different types of proteins, regulated shared sets of genes associated with contractile functions, cell cycle pathways, NFkappaB, and type I interferon signaling. By integrating information about GWAS gene effect direction and a deep analysis of cholesterol- and stretch-induced gene modules in SMCs, we find evidence that cholesterol-induced signaling is a pro-atherogenic disease mechanism in SMCs that is upregulated by detrimental and downregulated by protective GWAS genes. Overall, our study identifies a set of candidate disease mechanisms in SMCs that are regulated by multiple GWAS genes across several SMC assays. Furthermore, it provides proof-of-concept for using GWAS gene effect directionality to predict the pathogenic effect of candidate disease mechanisms that can be extended to other GWAS genes and cell types in the future.Competing Interest StatementA.K.U., D.D., C.P., J.M., G.B., J.B., K.H., M.T., and M.N. are employed at Novo Nordisk A/S.Footnotes* To predict which mesenchymal cell states observed in vivo in atherosclerotic lesions can be mirrored by in vitro samples in our experiments, we performed cell-type deconvolution analysis of bulk gene expression data from SMCs across our cellular assays. First, we generated pseudo-bulk samples based on scRNA-seq data from human atherosclerotic lesions and verified that the proportions of mesenchymal cell clusters in these pseudo-bulk samples, as estimated by the cell type deconvolution method Bisque, were similar to those observed in the original scRNA-seq dataset. Then, we estimated mesenchymal cell proportions for SMC samples transfected with control siRNA and compared them across our cellular assays.* https://doi.org/10.6084/m9.figshare.27650430.v1...

Mapping atherogenesis mechanisms in smooth muscle cells by targeting genes linked to coronary artery disease

10.1101/2024.11.11.623011