While genome-wide association studies (GWAS) have identified variants and genes associated with human disease, a comprehensive understanding of the genetic architecture of individual loci and the functional implications of these associations remains incomplete. In this work, we applied an integrated pipeline to chart the regulatory landscapes of obesity-associated loci within two cell types central to obesity etiology. In both adipocytes and hypothalamic neurons, we annotated gene expression, chromatin accessibility, and long-range chromatin interactions across multiple differentiation stages. Additionally, we generated a list of 2,396 variants in high LD with BMI lead SNPs and tested them in a massively parallel reporter assay to identify putatively causal variants modulating enhancer activity. We identified 94 variants within enhancers that displayed enhancer-modulating properties, many of which were active in both cell types. Our data show that individual GWAS loci harbor multiple candidate causal variants within distinct enhancers that display cross-tissue effects. Integrating the identified enhancer modulating variants (EMVars) with chromatin interactions and eQTL information generated a comprehensive list of genes predicted to underlie obesity GWAS associations. Aggregating our data across multiple time points allowed us to assign more candidate causal variants to genes compared to regulatory maps in a single cell type and to prioritize 232 genes with varying degrees of evidence for obesity risk importance. We used these insights during experimental dissection of a complex genomic interval on 16p11.2 where we observed EMVars at two independent GWAS loci exhibiting megabase-range, cross-locus Hi-C chromatin interactions and shared eQTL effects. We provide evidence that EMVars within these two loci converge to regulate a shared gene set. Together, our data chart the genetic architecture of obesity-associated loci and support a model in which many GWAS loci contain multiple variants that impair the activities of distinct enhancers across tissues, potentially with temporally restricted effects, to impact the expression of multiple genes. This complex network model has broad implications for ongoing variant to function efforts to mechanistically dissect GWAS.



Downloads Statistics