Epigenetic Analyses of Human Left Atrial Tissue Identifies Gene Networks Underlying Atrial Fibrillation

Amelia W Hall, Nathan R Tucker, Mark Chaffin, Carolina Roselli, Honghuang Lin, Valerio Bianchi, Geert Geeven, Kenneth C Bedi, Kenneth Margulies, Wouter De Laat, Patrick T Ellinor.
Circ Genom Precis Med. 2019-11-11;13(6):e003085.
Abstract
Introduction: Atrial fibrillation (AF) often arises from dysfunction in the left atrium (LA), a tissue with limited annotation of the non-coding genome and its effects on gene expression. As many AF-associated genetic variants reside in these regions, this knowledge gap hinders efforts to understand AF biology. Methods: In order to generate a model of the LA non-coding genome, we profiled 7 histone post-translational modifications (active: H3K4me3, H3K4me2, H3K4me1, H3K27ac, H3K36me3; repressive: H3K27me3, H3K9me3), CTCF binding, and gene expression in 4 LA samples from individuals without structural heart disease or AF. We used MACS2 to identify peak regions (P < 0.01), applied a Markov model approach to classify regulatory elements, and then annotated with gene expression. In order to better relate this model with disease biology, we then intersected chromatin states with independently derived eQTL, DNA methylation, and HiC long-range chromatin interaction data from primary human LA and left ventricle. Results: Our model identified 21 states, which recapitulate regulatory motifs such as promoters, enhancers, and repressed regions. Genes were actively regulated by proximal chromatin states, and repressive states was associated with a significant reduction in gene expression (P < 2x10-16). Further, chromatin states were differentially methylated, with promoters being less methylated than repressed regions (P < 2x10-16). For enhancer-gene associations, we identified 142 significantly enriched pathways (P < 0.05; FDR). Intersecting AF, PR, and QT SNP loci with cell type specific promoters identified a gene interaction network dominated by CAV1, CDKN1A, CREBP, NKX2-5 and SYNPO2L. Discussion: Our profiling of the non-coding genome in human LA provides new insights into the gene expression and chromatin regulation within this AF-associated tissue. Further analyses of these data will be used to inform the relationship between GWAS variants and AF pathology.