Multi-Omics Mendelian Randomization and Colocalization Reveal Key Glycolipid Metabolism-Related Genes in Gestational Diabetes Mellitus.

PURPOSE: Gestational diabetes mellitus (GDM) poses significant health risks, yet the causal genetic and epigenetic mechanisms linking glycolipid metabolism dysregulation to GDM remain elusive. This study aimed to identify key causal genes and regulatory pathways by integrating multi-omics data with large-scale genetic association studies.

PATIENTS AND

METHODS

We leveraged GDM genome-wide association study (GWAS) data from the FinnGen consortium (18,581 cases/263,483 controls) as the discovery dataset and the UK Biobank (193 cases/219,789 controls) for replication. These were integrated with blood-based quantitative trait loci (QTL) for DNA methylation (mQTL), gene expression (eQTL), and protein abundance (pQTL) of glycolipid metabolism-related genes. Findings were rigorously validated using colocalization analysis (PP.H4 > 0.5), cross-QTL mediation analysis, and tissue-specific eQTL data (adipose, liver, pancreas) from GTEx. Protein-protein interaction (PPI) networks and drug druggability analyses were conducted to assess translational potential.

RESULTS

The multi-omics SMR analysis identified 325 CpG sites, 58 gene expressions, and 6 protein abundances associated with GDM, which were refined by colocalization to 72 mQTLs, 6 eQTLs, and 3 pQTLs. Integrative analysis prioritized 6 core genes: HKDC1, CYP21A2, DCXR, GCDH, ACOT4, and GPX1. HKDC1 was identified as a core hub; its hypomethylation was associated with decreased expression and increased GDM risk, which was further validated in adipose, liver, and pancreatic tissues. CYP21A2 overexpression was confirmed as a risk factor, a finding substantiated by independent replication in the UK Biobank. Additionally, DCXR and GCDH were identified as novel susceptibility genes supported by significant gene-protein regulatory cascades. Finally, protein-protein interaction network and drug target analyses highlighted HKDC1 as a potential druggable target.

CONCLUSION

This study identifies glycolipid metabolism genes, particularly HKDC1, in GDM via a methylation-expression-phenotype axis, and nominates DCXR and GCDH as novel candidates. These findings reveal new molecular insights and therapeutic targets for GDM.