Abstract 17184: Nearly One Third of GWAS CAD Are Putative SMC Klf4 or Oct4 Targets in Advanced Atherosclerotic Lesions

Abstract

Atherosclerosis is a disease of chronic inflammation and the leading cause of morbidity and mortality worldwide. Despite decades of research, our understanding of the mechanisms regulating plaque stability remains poor. Until recently, smooth muscle cells (SMCs) were thought to play an athero-protective role in lesion pathogenesis. However, rigorous lineage tracing has shown a significant portion (>80%) of SMC-derived cells within an advanced atherosclerotic lesion did not express detectable SMC markers. Furthermore, SMC-specific conditional KO of Klf4 resulted in reduction of several indices of plaque instability and SMC-derived macrophage-like cells. In contrast, conditional KO of Oct4 in SMCs presented a reduction of SMC and increased of Lgals3+ cells within lesions, as well as increase of several indices of plaque instability. Taken together, SMCs can have a beneficial or detrimental effect in lesion pathogenesis. Hypothesis: Combinatorial analysis of Oct4/Klf4 ChIP-seq and RNA-seq of advanced atherosclerotic lesions from our SMC-Klf4 and SMC-Oct4 mice provide a unique opportunity to identify genes that play a critical role in regulating beneficial or detrimental changes in SMC phenotype. Results: SMC-Oct4 target genes are enriched for migration pathways, while SMC-Klf4 target genes are enriched for inflammation pathways. Further, atherosclerosis pathway, among others, presented significant opposite patterns (SMC-Oct4-KO up-regulation and SMC-Klf4-KO down-regulation), and presented strong significant gene anti-correlation, corroborating our phenotypic observations. Remarkably, we cross-reference our ChIP-seq datasets with CAD GWAS and found that 54 of the 161 human loci were either Klf4 and/or Oct4 SMC targets, highly suggesting that CAD GWAS genes might be affecting SMC function in lesion development/pathogenesis. To validate our results, we used the STARNET dataset, and we observed that both Oct4 and Klf4 are significantly up-regulated in aorta of CAD patients compared to controls. In addition, co-expression analysis shows that one of the Klf4 network is highly related to human GWAS CAD, including genes such as LDLR, APOE, PCSK9, TRIB1, and others. Surprisingly, more than 35% of genes in this network also are present in our SMC Klf4 ChIP-seq dataset. Conclusions: our results suggests that: late stage atherosclerotic lesions of mouse and humans seems to be more similar than previously thought; and nearly 1/3 of CAD GWAS loci appear to be affecting SMC function and are potentially Klf4/Oct4-dependent.