Transcriptome-wide association study-derived genes as potential visceral adipose tissue-specific targets for type 2 diabetes

Haibo Tang,Jie Wang,Peizhi Deng,Yalan Li,Yaoquan Cao,Bo Yi,Liyong Zhu,Shaihong Zhu,Yao Lu

DOI: https://doi.org/10.1007/s00125-023-05978-5

IF: 8.2

Diabetologia

Abstract:Aims/hypothesis: This study aimed to assess the causal relationship between visceral obesity and type 2 diabetes and subsequently to screen visceral adipose tissue (VAT)-specific targets for type 2 diabetes. Methods: We examined the causal relationship between VAT and type 2 diabetes using bidirectional Mendelian randomisation (MR) followed by multivariable MR. We conducted a transcriptome-wide association study (TWAS) leveraging prediction models and a large-scale type 2 diabetes genome-wide association study (74,124 cases and 824,006 controls) to identify candidate genes in VAT and used summary-data-based MR (SMR) and co-localisation analysis to map causal genes. We performed enrichment and single-cell RNA-seq analyses to determine the cell-specific localisation of the TWAS-identified genes. We also conducted knockdown experiments in 3T3-L1 pre-adipocytes. Results: MR analyses showed a causal relationship between genetically increased VAT mass and type 2 diabetes (inverse-variance weighted OR 2.48 [95% CI 2.21, 2.79]). Ten VAT-specific candidate genes were associated with type 2 diabetes after Bonferroni correction, including five causal genes supported by SMR and co-localisation: PABPC4 (1p34.3); CCNE2 (8q22.1); HAUS6 (9p22.1); CWF19L1 (10q24.31); and CCDC92 (12q24.31). Combined with enrichment analyses, clarifying cell-type specificity with single-cell RNA-seq data indicated that most TWAS-identified candidate genes appear more likely to be associated with adipocytes in VAT. Knockdown experiments suggested that Pabpc4 likely contributes to regulating differentiation and energy metabolism in 3T3-L1 adipocytes. Conclusions/interpretation: Our findings provide new insights into the genetic basis and biological processes of the association between VAT accumulation and type 2 diabetes and warrant investigation through further functional studies to validate these VAT-specific candidate genes.