Integrating islet methylation and patient electronic health record data to characterize enhancer-gene networks in Type-2 diabetes
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Genome-wide association studies (GWAS) have shown that Type-2 diabetes (T2D) risk variants are enriched within islet gene enhancers. These and other studies argue that T2D susceptibility largely arises from disrupted gene regulation in islet β cells, though the full extent of islet regulatory contributions remains unclear. In this dissertation, I present a framework combining islet DNA methylation with patient genotype-phenotype data to uncover enhancer-gene networks contributing to T2D risk. Using whole-genome bisulfite sequencing data from the islets of four non-T2D donors, I generated a reference islet methylome comprising ~35,000 hypomethylated regions (HMRs) shared across donors. Comparing HMR patterns across islets and diverse cell types identified those unique to (n = 4,858) or shared with (n = 30,863) islets. Most islet-specific HMRs exhibit enhancer-like chromatin signatures and are linked to genes and transcription factors critical for β cell function and identity. Moreover, islet-specific HMRs are enriched for genetic contribution to the heritability of T2D and blood glucose levels. Using genotype-phenotype data from Vanderbilt’s BioVU biobank, which links patient DNA to electronic health records, phenome- and lab-wide association scans replicated islet HMR associations with T2D and revealed new links to related metabolic clinical traits. I also developed a deviation of the GWAS approach, termed “HMR-WAS”, which restricts analyses to variants within HMRs to enhance detection of biologically meaningful signals with smaller effect sizes. This approach replicated a known T2D signal at the PAM locus undetected by GWAS. Luciferase reporter assays confirmed the enhancer function of the T2D-associated islet-specific HMR upstream of PAM and nearby non-coding regions at SLC2A2 and GCK, supporting a functional link between their regulatory activity and T2D risk. Overall, this work reveals enhancers of key metabolic genes regulating β cell function and provides a framework for identifying regulatory elements underlying complex diseases. In a separate chapter, I present findings from a collaborative study examining the methylomes of rodent β cell subtypes with distinct secretory and proliferative capacities, which implicate differential DNA methylation in islet progenitors as a determinant of β cell subtype diversity.