Abstract 14675: High-Throughput CRISPR Screening Links CAD Loci to Endothelial Cell Programs and Causal Pathways

Abstract

Introduction: Genome-wide association studies (GWAS) have discovered >300 associations for CAD. Few loci are functionally characterized and may represent new mechanisms of disease. Hypothesis: Can a high-throughput, unbiased transcriptional screen for all candidate GWAS genes identify the causal genes and biological pathways at CAD GWAS loci in endothelial cells (ECs)? Methods: We applied CRISPRi-Perturb-seq to knock down the expression of all genes within 500 Kb of coronary artery disease GWAS loci (2,300 genes in total) and measure their effects on the transcriptome using single-cell RNA-seq. Results: We identified 60 programs of co-expressed genes, which represent core cellular pathways (i.e. ribosome biogenesis) and EC-specific pathways such as flow response and angiogenesis. The EC-specific programs show the greatest contribution to CAD heritability. 6 EC-specific programs have the greatest number of CAD GWAS candidate genes. One of the EC-specific programs had genes regulated by KLF -transcription factors and was enriched for shear-stress response genes. The novel CAD candidate genes in this program included multiple known mediators of the cerebral cavernous malformation (CCM)-signaling complex. 9 known members of the CCM-signaling cascade and several potentially novel mediators of CAD clustered together in the KLF Perturb-seq topic. These included genetic variants in the CCM2 , KLF4 , RAC1 , and HEG1 loci as well as genes will a similar transcriptional profile but no prior connection to the CCM-signaling complex. Conclusions: High-throughput functional analysis of 2,300 genes proximal to CAD GWAS loci prioritized pathways—such as angiogenesis and EC migration—that are regulated by multiple risk SNPs. Our study identifies new genes that likely influence risk for CAD, identifies convergence of CAD genes into certain pathways in endothelial cells, and demonstrates a generalizable strategy to connect disease variants to functions.