Functional characterization of SIM1-associated enhancers
Haploinsufficiency of the single-minded homology 1 (SIM1) gene in humans and mice leads to severe obesity, suggesting that altered expression of SIM1, by way of regulatory elements such as enhancers, could predispose individuals to obesity. Here, we identified transcriptional enhancers that could regulate SIM1, using comparative genomics coupled with zebrafish and mouse transgenic enhancer assays. Owing to the dual role of Sim1 in hypothalamic development and in adult energy homeostasis, the enhancer activity of these sequences was annotated from embryonic to adult age. Of the seventeen tested sequences, two SIM1 candidate enhancers (SCE2 and SCE8) were found to have brain-enhancer activity in zebrafish. Both SCE2 and SCE8 also exhibited embryonic brain-enhancer expression in mice, and time course analysis of SCE2 activity showed overlapping expression with Sim1 from embryonic to adult age, notably in the hypothalamus in adult mice. Using a deletion series, we identified the critical region in SCE2 that is needed for enhancer activity in the developing brain. Sequencing this region in obese and lean cohorts revealed a higher prevalence of single nucleotide polymorphisms (SNPs) that were unique to obese individuals, with one variant reducing developmental-enhancer activity in zebrafish. In summary, we have characterized two brain enhancers in the SIM1 locus and identified a set of obesity-specific SNPs within one of them, which may predispose individuals to obesity.To identify potential SIM1 enhancers, we carried out a comparative genomic analysis on the SIM1 locus. We searched for conserved non-coding regions in this locus, defined as one gene upstream [activating signal cointegrator 1 complex subunit 3 (ASCC3)] and downstream [melanin-concentrating hormone receptor 2 (MCHR2)] of SIM1, for a total genomic distance of ∼1 mb (Fig. 1). Analysis of the SIM1 locus revealed a human–mouse synteny block that ends upstream of MCHR2 and separates/eliminates any non-coding conservation between human and mouse ∼93 kb upstream of MCHR2 (Fig. 1). This suggests that enhancers found within this synteny block likely regulate Sim1 in mice (42). Furthermore, while MCHR2 is expressed in the hypothalamus in humans and is modestly associated with polygenic obesity (43), MCHR2 and its ligand, MCH, are not present in mice.Using the ECR Browser (44), 488 evolutionary conserved sequences (ECRs) between human and mouse were found that were at least 70% conserved for at least 100 bp within the defined SIM1 locus (Supplementary Material, Table S1). ECRs were analyzed manually for repetitive sequences and any RNA coding evidence using the UCSC Genome Browser (45), removing any that contained either. The remaining 360 non-coding ECRs were then ranked by species conservation to prioritize for enhancer assays. Seventeen ECRs, conserved between human and frog, were chosen for subsequent enhancer assays (Supplementary Material, Table S2) and were termed SCEs.Using ECR Browser (44), we selected intronic and intergenic ECRs between MCHR2 and ASSC3 that were ≥100 bp long with at least 70% sequence identity between human and mouse. This analysis generated 488 unique ECRs (Supplementary Material, Table S1). We manually filtered out repetitive sequences, expressed sequence tags (ESTs) and any other coding sequences using the UCSC Genome Browser (45). To prioritize ECRs for functional assays, the remaining 360 ECRs were then ranked by species conservation. A total of seventeen human and frog ECRs (labeled as SCEs) were chosen for enhancer assays (Supplementary Material, Table S2).We would like to thank John L.R. Rubenstein for his gift of the mSim1 in situ construct and assisting in murine brain annotations, Len A. Pennacchio and Axel Visel (LBL) for providing E11.5 SCE2 mouse transgenic enhancer embryos and the Ahituv lab for helpful comments on the manuscript.Conflict of Interest statement. None declared.