Exploring the genomic and transcriptomic profiles of glycemic traits and drug repurposing

Abstract

Background: Type 2 diabetes is an increasingly prevalent metabolic disorder with moderate to high heritability. Glycemic indices are crucial for diagnosing and monitoring the disease. Previous genome-wide association study (GWAS) have identified several risk loci associated with type 2 diabetes, but data from the Taiwanese population remain relatively sparse and primarily focus on type 2 diabetes status rather than glycemic trait levels.

Methods: We conducted a comprehensive genome-wide meta-analysis to explore the genetics of glycemic traits. The study incorporated a community-based cohort of 145,468 individuals and a hospital-based cohort of 35,395 individuals. The study integrated genetics, transcriptomics, biological pathway analyses, polygenic risk score calculation, and drug repurposing for type 2 diabetes.

Results: This study assessed hemoglobin A1c and fasting glucose levels, validating known loci (FN3K, SPC25, MTNR1B, and FOXA2) and discovering new genes, including MAEA and PRC1. Additionally, we found that diabetes, blood lipids, and liver- and kidney-related traits share genetic foundations with glycemic traits. A higher PRS was associated with an increased risk of type 2 diabetes. Finally, eight repurposed drugs were identified with evidence to regulate blood glucose levels, offering new avenues for the management and treatment of type 2 diabetes.

Conclusions: This research illuminates the unique genetic landscape of glucose regulation in Taiwanese Han population, providing valuable insights to guide future treatment strategies for type 2 diabetes.

Keywords: Drug repurposing; Fasting glucose; Genome-wide association study; Glycemic traits; Hemoglobin A1c; Polygenic risk score; Transcriptome-wide association study; Type 2 diabetes.

Fig. 1

Study Workflow. The study conducted GWAS for fasting glucose and HbA1c using a community-based cohort from TWB and a hospital-based cohort from TCVGH, followed by a meta-analysis of the summary statistics from both cohorts. Fine mapping was performed to identify independent signals from the GWAS. The GWAS catalog was queried to distinguish known and novel variants. Pathway enrichment analysis was carried out to investigate the biological functions of the identified risk genes. Additionally, TWAS was conducted to explore risk genetic factors at the gene expression level. PRS were calculated for both glycemic traits to predict the risk of developing type 2 diabetes. Finally, drug repurposing strategies were applied to identify potential treatments for type 2 diabetes

Fig. 2

Miami plot of meta-GWAS and TWAS on HbA1c and fasting glucose. A Miami plot for HbA1c, with the upper panel showing meta-GWAS (TWB and TCVGH) and the lower panel showing TWAS results. B Miami plot for fasting glucose, with the upper panel showing meta-GWAS (TWB and TCVGH) and the lower panel showing TWAS results. Red horizontal dashed lines indicate significance thresholds (−log10(5 × 10^–8) for meta-GWAS, −log10(6.89 × 10^–6) for TWAS). Known loci are highlighted in blue, new associations from this study are in orange, and independent variants are in navy. We labeled only the top known and novel genes for each chromosome in the meta-GWAS

Fig. 3

Scatter plot of effect sizes and minor allele frequencies (MAF). A The effect sizes in TWB (x-axis) and TCVGH (y-axis) of the 3600 overlapping significant variants in the HbA1c meta-GWAS. B The MAF (x-axis) and effect size (y-axis) of the 5394 significant variants in the HbA1c meta-GWAS. C The effect sizes in TWB (x-axis) and TCVGH (y-axis) of the 3298 overlapping significant variants in the fasting glucose meta-GWAS. D The MAF (x-axis) and effect size (y-axis) of the 5332 significant variants in the fasting glucose meta-GWAS

Fig. 4

Comprehensive gene-set enrichment analysis of glycemic traits from meta-GWAS and TWAS. The top 10 enriched results from three databases were displayed, including GO: Biological Process, KEGG Pathway, and WikiPathways. A Enriched pathways for HbA1c-associated genes bases on meta-GWAS data; B Enriched pathways for fasting glucose-associated genes bases on meta-GWAS data; C Enriched pathways for HbA1c-associated genes bases on TWAS data; D Enriched pathways for fasting glucose-associated genes bases on TWAS data. Pathways with a false discovery rate (fdr) below 0.05 are highlighted in yellow, while others are shown in purple. The Rich factor is the ratio of the number of input genes annotated in a pathway to the total number of genes annotated in the same pathway

Fig. 5

Polygenic risk score analyses. A Association between glycemic traits PRS and clinical traits from TWB. Red horizontal dashed lines indicate significance thresholds (−log10(1.9 × 10^–4); B The risk of type 2 diabetes predicted by the PRS for HbA1c; C The risk of type 2 diabetes predicted by the PRS for fasting glucose