posted on 2021-04-13, 20:58authored byArushi Varshney, Yasuhiro Kyono, Venkateswaran Ramamoorthi Elangovan, Collin Wang, Michael R. Erdos, Narisu Narisu, Ricardo D'Oliveira Albanus, Peter Orchard, Michael L. Stitzel, Francis S. Collins, Jacob O. Kitzman, Stephen C. J. Parker
Identifying the
tissue-specific molecular signatures of active regulatory elements is critical
to understand gene regulatory mechanisms. Here, we identify transcription start
sites (TSS) using cap analysis of gene expression (CAGE) across 57 human
pancreatic islet samples. We identify 9,954 reproducible CAGE tag clusters
(TCs), ~20% of which are islet-specific and occur mostly distal to known gene
TSSs. We integrated islet CAGE data with histone modification and chromatin
accessibility profiles to identify epigenomic signatures of transcription
initiation. Using a massively parallel reporter assay, we validated the transcriptional
enhancer activity for 2,279 of 3,378 (~68%) tested islet CAGE elements (5%
FDR). TCs within accessible enhancers show higher enrichment to overlap type 2
diabetes genome-wide association study (GWAS) signals than existing islet
annotations, which emphasizes the utility of mapping CAGE profiles in
disease-relevant tissue. This work provides a high-resolution map of
transcriptional initiation in human pancreatic islets with utility for
dissecting active enhancers at GWAS loci.
Funding
We acknowledge support from the University of Michigan Rackham Predoctoral Fellowship (A.V.), T32 HG00040 from the National Human Genome Research Institute of the N.I.H. (P.O.), ADA Accelerator grant 1-18-ACE-15 (M.L.S.), NIH intramural support from project ZIA-HG000024 (F.S.C.), ADA Pathway to Stop Diabetes Grant 1-14-INI-07, NIH 1UM1DK126185-01 grant and R01 DK117960 (S.C.J.P.).