2116-P: A Pipeline to Explore Rare Variation Which Can Contribute to Extreme Obesity in American Indians

Identification of high-impact loss-of-function variants for common disease is challenging because these variants are typically rare and thus statistical power is low when sample sizes are limited. In this study, we utilized whole-exome variant detection, variant prediction programs, tissue expression data, and case-control analysis to identify potentially highly penetrant obesity-susceptibility variants for prioritization for functional testing. Whole-exome sequencing was performed on DNA samples collected from a longitudinal, population-based study of American Indians (N=5432; maximum BMI recorded at ≥18 years). Among the 1 million biallelic SNVs detected in this cohort, those that were exonic or splicing, have minor allele frequency (MAF) ≤0.05, predicted to be damaging (CADD score ≥15), and located in genes expressed in hypothalamus were selected for analysis. These selection criteria resulted in 94,444 SNVs, and odds ratios were calculated for them by comparing the group of individuals with BMI in the top 5th percentile for this population (N=271, BMI ≥53.13 kg/m2) vs. those below the median (N=2,717, BMI <35.92 kg/m2). The resulting 19,382 variants with positive OR values were then divided into 4 bins based on MAFs, and those with a OR in the top 1% of the distribution of each bin were prioritized. In the resulting list of 291variants in 266 different genes, a missense variant p.I330V (rs201597085; OR=2.66; CADD=20.3; MAF=0.02) in GNAS was the best candidate, since GNAS was the only gene linked to obesity in humans. GNAS is a highly complex imprinted locus, and the main transcript derived from this locus (Gs-alpha) encodes the alpha subunit of the stimulatory guanine nucleotide-binding protein. The XLas isoform, a large variant of Gs-alpha, is paternally expressed. An association analysis with accountance for parental origin of the alleles is ongoing. Future functional studies would entail examining the effects of this missense variant on G protein-coupled receptor signaling.