Abstract P484: Defining Disease Specific Cardiac Regulatory Networks

The transcriptional basis of homeostasis and disease is implied by the non-coding nature of genetic variation identified by GWAS. Functional genomic approaches to unveil the gene regulatory networks (GRNs) relevant to disease are therefore a high priority. Most models for transcriptional dysregulation presume that perturbation of a wild-type gene regulatory network causes disease risk. For example, the T-box transcription factor (TF) TBX5, is essential for atrial rhythm homeostasis and directly drives a physiologically relevant GRN composed of cardiac channel genes. Alternatively, the expression of many genes pertinent to cardiac pathology are upregulated after the removal of Tbx5, implicating a disease-specific GRN absent from the wild-type atrium. We applied TF-dependent noncoding RNA (ncRNA) profiling, using differential deep ncRNA sequencing from atria of wild-type and Tbx5 mutant mice, to identify TBX5-dependent enhancers and ncRNAs that were only activated following TBX5 removal. We hypothesized that these regulatory elements would reveal disease-response enhancers, essential for coping with atrial dysfunction. To identify the cell-specific regulatory elements, we generated cell-type specific Assay from Transposase-Accessible Chromatin (ATAC) datasets for left atrial tissue, cardiac fibroblasts, and cardiomyocytes. Overlap with the Tbx5 -repressed ncRNAs defined candidate cell-type specific regulatory elements. Candidate regulatory elements were identified upstream of Sox9, a known modulator of cardiac fibrosis, along with other cardiac stress-response pathways, including mediators of TGF-β signaling. Activation of the enhancer at Sox9 was confirmed in isolated cardiac fibroblasts treated with TGF-β. We hypothesized that the disease-acquired GRN in TBX5 mutant atria may be generalizable to other cardiac insults. We therefore examined the transcriptional and genomic changes in the left atria of the heart failure Transverse Aortic Constriction (TAC) mouse model. This analysis revealed remarkable correlation between differentially expressed genes and ncRNAs between TAC and TBX5 mutant disease models. The conservation of the coding and non-coding transcriptional response between arrhythmia and heart failure models supports a paradigm of a common disease-specific GRN that mediates the physiologic consequences of distinct cardiac diseases.