A multi-layer functional genomic analysis to understand noncoding genetic variation in lipids

Abstract

A major challenge of genome-wide association studies (GWASs) is to translate phenotypic associations into biological insights. Here, we integrate a large GWAS on blood lipids involving 1.6 million individuals from five ancestries with a wide array of functional genomic datasets to discover regulatory mechanisms underlying lipid associations. We first prioritize lipid-associated genes with expression quantitative trait locus (eQTL) colocalizations and then add chromatin interaction data to narrow the search for functional genes. Polygenic enrichment analysis across 697 annotations from a host of tissues and cell types confirms the central role of the liver in lipid levels and highlights the selective enrichment of adipose-specific chromatin marks in high-density lipoprotein cholesterol and triglycerides. Overlapping transcription factor (TF) binding sites with lipid-associated loci identifies TFs relevant in lipid biology. In addition, we present an integrative framework to prioritize causal variants at GWAS loci, producing a comprehensive list of candidate causal genes and variants with multiple layers of functional evidence. We highlight two of the prioritized genes, CREBRF and RRBP1, which show convergent evidence across functional datasets supporting their roles in lipid biology.

Keywords: complex traits; fine-mapping; functional genomics; lipid biology; post-GWAS; regulatory mechanism; variant prioritization.

Figure 1

Schematic overview of the multi-layer functional genomic analysis We integrate GWAS summary statistics for five lipid phenotypes with eQTL and chromatin interaction data to identify potential genes mediating the GWAS loci, and use epigenomic annotations to identify regulatory mechanisms at these loci. For a GWAS locus indexed by a lead variant X, A, B, and C represent nearby eGenes across tissues, and SNPs around SNP X represent variants in the credible set for this locus.

Figure 2

Overlap between eQTL colocalized genes and Capture-C prioritized genes, and their enrichments in known lipid-associated genes (A) Numbers of genes identified by two approaches: eQTL colocalization (Coloc) and promoter Capture-C interaction (CapC). Capture-C interactions restricted to genes expressed in the tissue of interest (or in the union of adipose and liver for “all tissues”) are shaded. (B) Overlap between two list of prioritized genes (left: Capture-C prioritized genes; right: eQTL colocalized genes) with four external sets of genes previously associated with lipid biology (MGI knockout genes, ClinVar lipidemia-associated genes, genes implicated in rare burden of lipids, and genes from a lipid TWAS). Dashed lines represent enrichments using only genes expressed in the liver. (C) Enrichment in overlap between eQTL colocalized genes and Capture-C prioritized genes against what is expected by chance, assuming both gene sets are independent. Dashed lines represent genes expressed in the tissue of interest (or in the union of adipose or liver for “all”). Enrichment estimates and 95% confidence intervals shown in (B) and (C) are based on the Fisher exact test. (D) Fraction of colocalized loci that point to a single candidate gene when using eQTL data alone or using both eQTL and Capture-C data.

Figure 3

Tissue relevance of lipid-associated loci Partitioning heritability of lipid GWAS summary statistics on gene expression (A) and active chromatin marks (B) across tissues. Each plotted point represents a tested dataset for enrichment of heritability, with larger dots representing datasets with enrichment p value < 0.05. Each color represents a tissue group (Table S7), and the y-axis represents −log10 p value for enrichment of heritability.

Figure 4

TF enrichment identified by GREGOR and S-LDSC (A) Number of TFs enriched in the GREGOR analysis on GWAS loci for each of the five lipid traits. (B) Number of TFs enriched in S-LDSC analysis on each of the five lipid traits. (C) TF RXRA binds to the promoters of 26 colocalized genes (18 protein-coding). Colors represent the subsets of lipid phenotypes with colocalization. Larger node sizes represent smaller GWAS p values of colocalized loci.

Figure 5

Multi-layer functional integration to prioritize variants at GWAS loci (A) Variant annotation and prioritization scheme at each GWAS credible set. (B) Evidence for RRBP1 from functional genomics data. The LDL GWAS locus at this region (first row) is an eQTL for RRBP1 in the liver (second row). Variants in the credible set of this locus interact with the gene promoter in both adipose and HepG2 Capture-C data (third row). The interacting variant is also in an open chromatin peak in three liver-related cell types (fourth row). (C) Multiple sources of functional genomics data support CREBRF as a gene contributing to HDL levels. The HDL GWAS locus at this region (first row) is an eQTL for CREBRF in adipose (second row). Variants in the credible set at this locus interact with the CREBRF promoter in adipose (third row). The interacting variant is also in open chromatin in liver-related cell types (fourth row).