Distinct and shared genetic architectures of Gestational diabetes mellitus and Type 2 Diabetes Mellitus

Abstract

Gestational diabetes mellitus (GDM) affects more than 16 million pregnancies annually worldwide and is related to an increased lifetime risk of Type 2 diabetes (T2D). The diseases are hypothesized to share a genetic predisposition, but there are few GWAS studies of GDM and none of them is sufficiently powered to assess whether any variants or biological pathways are specific to GDM. We conducted the largest genome-wide association study of GDM to date in 12,332 cases and 131,109 parous female controls in the FinnGen Study and identified 13 GDM-associated loci including 8 novel loci. Genetic features distinct from T2D were identified both at the locus and genomic scale. Our results suggest that the genetics of GDM risk falls into two distinct categories - one part conventional T2D polygenic risk and one part predominantly influencing mechanisms disrupted in pregnancy. Loci with GDM-predominant effects map to genes related to islet cells, central glucose homeostasis, steroidogenesis, and placental expression. These results pave the way for an improved biological understanding of GDM pathophysiology and its role in the development and course of T2D.

Figure 1:. Genome-wide association results for GDM.

(A) Manhattan plot of GWAS of GDM in 12,332 cases and 131,109 parous female controls of Finnish ancestry. The x-axis reflects chromosomal positions and the y-axis reflects −log10(P) values for the two-tailed association test for each variant, presented on a log scale. Red dotted line indicates the significance threshold (P = 5 × 10−8). Colored SNPs represent the credible set members for the 13 genome-wide significant loci, with blue indicating loci previously associated with GDM and orange indicating novel associations. Labels indicate the gene nearest to the fine-mapped lead SNP. (B) Genetic correlations (SNP-rg) between GDM and other diseases, traits and biomarkers estimated using LD score regression. Depicted traits are significant after Bonferroni correction for 53 traits (p < 9.4e-4); results for all tested traits are reported in Supplementary Tables 15-16. Error bars show +/− 1 standard error. Colors indicate phenotype category.

Figure 2:. Classification of the genetic effects of SNPs in GDM and T2D.

Comparison of log odds ratios in GWAS of GDM (x-axis) and T2D in males (y-axis) for top-associated SNPs from GDM (13 SNPs) and (15 SNPs). Two distinct classes of SNP effects were identified by a Bayesian classifier in shared variants analysis: Class T (blue) containing SNPs with T2D-predominant genetic effects and Class G (red) with GDM-predominant effects (Supplementary Table 22). Grey SNPs were not confidently assigned to either class (posterior probability > 95%). Dotted ellipses indicate the 95% probability regions of the fitted bivariate effect size distributions with each class.

Figure 3:. Cell type specificity analysis of GDM and T2D highlights different cell associations.

Cell type specificity analysis was performed for GDM and for prior meta-analysis of T2D from Mahajan et al. using high quality murine single-cell RNA-seq datasets with FUMA (Supplementary Tables 26-29). Unadjusted P-values are reported for the association between relative gene expression in the given cell type and MAGMA gene-level associations in the GWAS. Results are shown for cell types that both (a) are significantly associated with at least one GWAS after correction for multiple testing of all cell types in all datasets, and (b) have putatively independent association conditional on other cell types in the same RNA-seq dataset. Colors indicate RNA-seq dataset source and significance.