“Regulatory genetic variants and human diseases”
Genome-Wide Association Studies (GWAS) have systematically implicated noncoding DNA as risk factors for complex human diseases. These noncoding variants presumably disrupt the function of long-range regulatory elements, resulting in altered gene expression and disease etiology. Nevertheless, the functional follow-up of GWAS remain a significant challenge, partly due to the difficulty in i) identifying the causal variants underlying the association, ii) characterizing the spatio-temporal specificities of regulatory elements harboring causal variants, iii) ascribing the target genes of candidate enhancers, iv) identifying the molecular effectors that bind differentially to the alternative alleles of the causal variants, and v) characterizing the molecular, cellular, and organismal impact of the causal variants. We developed an integrated approach to mechanistically dissect GWAS, providing a paradigm framework for the functional follow-up of these genetic associations. Utilizing the association of SNPs within TCF7L2 and increased risk to type 2 diabetes (T2D) as a paradigm, we computationally predicted the causal variant of the association. Combing public datasets for epigenetic data we predicted that this variant lies within an active enhancer in pancreatic islets, enteroendocrine cells and adipose tissues, which we confirmed using reporter assays in appropriate cell lines. We computationally predict and experimentally show that Gata3 binds to the risk allele of this SNP and is required for the adipocyte enhancer activity of the risk allele. We show that the risk allele is associated with increased TCF7L2 expression and down-regulation of genes involved in adipogenesis. Utilizing genetically modified mice we show that increased adipocyte Tcf7l2 expression results in impaired adipogenesis, with insulin resistance and hepatic steatosis. Transcriptome analysis reveal that Tcf7l2 regulates major players in adipogenesis in mouse and human adipocytes. Using genomic editing we demonstrate that this SNP is sufficient to regulate TCF7L2 expression in adipocytes, directly linking this SNP to TCF7L2 expression levels. Finally, we uncovered a genetic by environment interaction. Our integrative approach posits that the mechanistic basis for the association of rs7903146 and T2D involves pleiotropic functions of an enhancer active in multiple tissues with important roles in glucose metabolism as well as a gene x environment interaction that predicts that risk allele carriers that become obese will have an additional phenotypic insult stemming from rs7903146.