Abstract Body: Background: Genome-wide association studies (GWAS) have identified common variants at the Cerebral Cavernous Malformation 2 (CCM2) locus associated with coronary artery disease (CAD) risk. While rare CCM2 mutations cause cerebrovascular malformations, the role of CCM2 in atherosclerosis is unknown. Using Perturb-seq, we previously identified CCM2 as a master regulator of endothelial gene programs enriched in CAD risk loci. We hypothesized that CCM2 loss-of-function modulates atherogenesis through endothelial inflammatory pathways in a disease stage-dependent manner.
Methods: We integrated Bayesian fine-mapping of CAD GWAS data, single-cell RNA sequencing (scRNA-seq), functional assays in human aortic endothelial cells (ECs), and Ccm2 heterozygous knockout (Ccm2+/−) mice atherosclerosis models (PCSK9-D377Y + 12 weeks high-cholesterol diet). The ERK5 degrader INY-06-061 was used to assess pathway specificity.
Results: Ccm2+/− mice developed 14.5% less aortic atherosclerosis than wild-type controls after 12 weeks (p=0.002). Baseline aortic scRNA-seq revealed downregulated inflammatory programs in ECs. Concordantly, CCM2 knockout in human aortic ECs reduced monocyte adhesion and procoagulant activity through MEKK3-ERK5-KLF2/4 pathway activation. INY-06-061 reversed transcriptional and adhesion effects but not procoagulant activity. However, this protection did not persist in established disease. Ccm2+/− mice exhibited elevated inflammatory gene expression and increased plasma TNFα (9.0 vs. 46.5pg/ml, p=0.006). Human genetics supports this stage-dependent effect. Fine-mapping identified the coding variant V53I (rs2107732) as the likely causal variant (posterior inclusion probability=0.94). V53I was associated with decreased CAD risk in the general population (OR: 0.835, p=0.014) but increased risk in a meta-analysis of secondary prevention clinical trials (OR = 1.108, p < 0.001).
Conclusions: CCM2 loss-of-function showed opposing effects on atherosclerosis initiation versus progression, demonstrating that GWAS-identified variants can have disease stage-dependent effects with distinct implications for primary versus secondary prevention.