Endocrine Abstracts

JOINT3443

Background: Hypothyroidism is a common endocrine condition affecting up to 10% of the general population. Patients experience a high degree of heterogeneity in symptoms, co-morbidities, and treatment effect. We aimed to investigate the shared genetic architecture of hypothyroidism and its co-morbidities for a better understanding of the disease mechanisms.

Methods: We leveraged publicly available genome-wide association study (GWAS) summary statistics to perform seven multi-trait GWAS meta-analyses (MTAG) based on four co-morbidity categories (Category 1 – Cardiometabolic; Category 2 – Psychiatric; Category 3 – Reproductive; Category 4 – Immune-mediated polyautoimmunity syndromes). MTAG boosts statistical power by JOINT analysis of multiple traits, enabling detection of new genetic associations. Thus, we aimed (1) to identify new genetic associations for hypothyroidism, and (2) to explore the genetic overlap across 26 genetically correlated conditions. Furthermore, we carried out gene prioritization, based on the new associations identified, to detect and characterize the potentially functional genes using two parallel combined SNP-to-Gene methods; (1) Otargen, a GraphQL-based R-package for tidy data accessing and processing from Open Targets Genetics, including nearest gene, variant-to-gene and locus-to-gene strategies, and (2) combined S2G-Framework (cS2G).

Results: MTAG analyses identified 114 new variants in 92 loci for hypothyroidism (P <5x10^-8) compared to the input GWAS summary statistic data. Based on these new associations, a total of 1,134 genes were prioritized (1,114 genes by Otargen and 94 genes by cS2G). Out of 1,134 genes, 26 genes were prioritized as ‘top-scoring’ across the two methods, including 16 nearest protein-coding genes (GLIS3, NFE2L3, NRG1, AFF3, TAGAP, S1PR1, CHN2, FOXK2, SWAP70, FCRL3, UBASH3B, SLC25A37, ABO, CD2, KCTD5, IL12RB2). GLIS3, NRG1, KCTD5, and SLC25A37 had significant expression or regulation patterns in the thyroid. Among the genes of interest, for instance, GLIS3, expressed in early embryogenesis with a role in multiple organ development, was highlighted as its variants have previously been associated with multiple co-morbidities of hypothyroidism including rheumatoid arthritis, type 1 diabetes, depression, as well as thyroid-related traits blood cholesterol and sex hormone levels. GLIS3 loss-of-function mutations lead to hyperglycemia, hypoinsulinemia, and congenital hypothyroidism. Supporting our finding, it has also been briefly mentioned in a recent larger GWAS investigating the overlap between thyroid traits and the reproductive system.

Conclusion: Multi-trait analyses and gene prioritization approaches revealed new genetic associations and potential functional genes for hypothyroidism based on publicly available data, providing further fundamental insights into its genetic architecture and expanding the understanding of the common genetics and biological processes between hypothyroidism and its co-morbidities.