Quantitative trait meta-analysis identifies rare noncoding variants associated with altered hormone levels in PCOS

Abstract

Polycystic ovary syndrome (PCOS) is a complex genetic disorder characterized by hyperandrogenism, chronic anovulation, and polycystic ovarian morphology. PCOS affects up to 15% of premenopausal women worldwide, is the leading cause of anovulatory infertility, and is a major risk factor for type 2 diabetes. A number of susceptibility loci have been mapped for PCOS in genome-wide association studies (GWAS), but the risk alleles identified to date can account for only a small proportion of the estimated genetic heritability of PCOS. To test whether rare genetic variants can account for this missing heritability, we performed whole genome sequencing on DNA from 76 families with one or more daughters affected with PCOS. Variants were filtered for allele frequency (minor allele frequency ≤2%), call quality, consistency with Mendelian inheritance, and predicted deleteriousness. Associations between sets of rare variants and PCOS and its quantitative traits were assessed using sequence kernel association tests, grouping variants at the gene-level, accounting for relatedness, and adjusting for age and BMI. Quantitative trait associations were combined into a single test statistic using a modified Fisher’s method for correlated traits. After correcting for multiple testing (Pc), an association with altered reproductive and metabolic trait levels (testosterone, dehydroepiandrosterone, luteinizing hormone, follicle stimulating hormone, fasting insulin, and sex hormone binding globulin) was found for 30 rare, noncoding variants in DENND1A (P=2.66×10-5, Pc=0.017). Multiple GWAS and meta-analyses have previously found associations between common variants in DENND1A and PCOS, and overexpression of a DENND1A isoform produces a PCOS phenotype in theca cells. Our results indicate that rare noncoding variants in DENND1A contribute to elevated androgen levels in PCOS. These findings support the hypothesis that DENND1A plays a key role in the development of PCOS. This study also demonstrates how quantitative trait meta-analysis can be a powerful approach in rare variant association testing.