Type 2 diabetes (T2D) results from interaction between genetic variation and environmental exposures that influence insulin secretion in pancreatic islets and insulin resistance in skeletal muscle and other peripheral tissues. Genome-wide association studies (GWAS) have identified hundreds of loci that modulate risk for T2D and related insulin traits. These loci mostly occur in non protein-coding regions which makes identification of causal regulatory single nucleotide polymorphisms (SNPs) and their associated mechanisms difficult. Previous studies show that many active enhancers are transcribed into enhancer RNA (eRNA). To identify actively transcribed regulatory elements with high genomic resolution and prioritize SNPs, we generated eRNA profiles using cap analysis of gene expression (CAGE) across 57 islet and 146 skeletal muscle samples and 15 replicates of the human beta cell line EndoC-BH1. We computationally integrated these CAGE profiles with ChIP-seq profiles for five histone marks along with open chromatin ATAC-seq profiles to generate high-resolution cross-tissue chromatin states. We identify multiple CAGE-enriched chromatin states that correspond to actively transcribed promoter or enhancer regions, which are supported by diverse patterns of chromatin activity across an orthogonal set of 127 human epigenome maps. We find 11,856 CAGE+ promoter regions across islets and muscle and 54% are shared across the two tissues, whereas we observe 1,712 CAGE+ eRNA regions and only 7% are shared, indicating the cell-identity nature of these eRNA states. We find significant enrichment for GWAS and eQTL SNPs to overlap these CAGE+ chromatin states. Finally, at a subset of loci, we validate these findings using a modified high-throughput STARR-seq assay where we observe allelic activity. These results provide the highest-resolution epigenome maps for islets and muscle to date, and we are using them to fine-map GWAS SNPs.


S. Parker: None. A. Varshney: None.


American Diabetes Association/Pathway to Stop Diabetes (1-14-INI-07 to S.P.); National Institutes of Health (1R01DK117960)

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at http://www.diabetesjournals.org/content/license.