Cardiovascular CArGome Function and Variation

Serum response factor (SRF) binds 1,216 permutations of a DNA code known as a CArG box. Studies have demonstrated altered SRF or target gene expression in the heart and vessel wall, and conditional SRF knockout mice exhibit defective cardiovascular development. Thus SRF is a homeostatic switch for normal cardiovascular function through its activation of key target genes, the full complement of which is unknown. Accordingly, we have defined over 84,000 conserved CArG boxes across human and mouse genomes (the CArGome). We have defined >11,000 human CArG-SNPs, including some within haplotype blocks linked to type II diabetes and stroke. Most CArG-SNPs studied exhibit reduced SRF binding and transcriptional activity. Thus, we have begun to elucidate the CArG variome in humans and suggest such information will be of enormous value as the paradigm of genome wide association studies transitions to a 'clan genomics' approach. Interestingly, many CArG elements are found subjacent to long non-coding RNAs, including a novel gene we call MINCR. Here, we hypothesize that variations in CArG sequences (or lncRNAs such as MINCR) lead to reduced target gene expression/activity, thus modulating cardiovascular disease susceptibility. We will begin testing this novel hypothesis with three innovative aims utilizing our skills in molecular biology, human genomics, and mouse genetics. In the first aim, we will delineate the cardiovascular CArGome through RNA-seq/ChIP-seq studies in vascular smooth muscle cell-specific SRF gain/loss of function mice. In Aim 2, we will begin to validate the cardiovascular CArGome, define CArG haplotype blocks, and launch a CArGome database as a dynamic tool for researchers worldwide. In Aim 3, we will define CArG-dependent lncRNAs to augment the CArGome database and further analyze the function of the novel lncRNA, MINCR. Thus, the proposed studies will illuminate novel SRF target genes, including a growing number of long non-coding RNAs, and functional CArG variants therein that may contribute to cardiovascular disease. Such information fills an important gap in genetic association studies where non-coding SNP function is rarely addressed. Further, this experimental approach serves as a template for the development of similar datasets from other well-characterized cis-acting elements (e.g., AP1) known to have a role in cardiovascular pathobiology. (AHA Program: Grant-in-Aid)