1137-P: Liver Regulatory Variants and Genes at Metabolic Trait Loci

Abstract

Genome-wide association studies (GWAS) have identified many loci that affect liver- and lipid-related metabolic traits, but the causal variants, target genes and mechanisms for most loci have not been characterized. GWAS loci are often enriched in tissue-specific transcriptional regulatory elements that drive physiological changes, and loci can be annotated with colocalized molecular quantitative traits for gene expression (eQTL) and accessible chromatin (caQTL) to identify candidate variants and genes. We previously profiled chromatin accessibility in 20 human liver samples, identified caQTL variants that colocalized with GWAS loci for metabolic traits, and predicted target genes based on promoter proximity, Hi-C chromatin contact with a promoter, and co-localization with liver eQTL. Here, we examined selected GWAS loci and tested variants for effects on gene regulation. We prioritized loci based on trait, gene function, and QTL characteristics and selected four variant-gene pairs to study. To evaluate effects of specific variants on gene expression, we performed transcriptional reporter assays in HepG2 cells. Three caQTL variants exhibited significant (P<0.05) allelic differences in transcriptional activity. rs13395911, near EFHD1 and associated with liver enzyme gamma glutamyl transferase, showed allelic differences in enhancer activity (P=0.008). In addition, rs9556404, near GPR180 and associated with triglycerides, and rs34003091, near ZNF329 and associated with LDL-cholesterol, also showed significant allelic effects on transcriptional activity (P<0.05). To study EFHD1 function in liver metabolism, we used gene-specific shRNA constructs in HepG2 cells and observed up to 90% reduction of EFHD1 mRNA expression compared to scramble shRNA. Subsequent metabolic assays will test role of EFHD1 on hepatic lipid and glucose metabolism. These data identified candidate regulatory variants that may be responsible for nearby GWAS loci and suggest genes as therapeutic targets for fatty liver disease. Disclosure G. Pandey: None. K. Currin: None. S. Vadlamudi: None. K. Broadaway: None. K. L. Mohlke: None.