整合免疫细胞转录组数据与孟德尔随机化揭示2型糖尿病及其并发症的新因果基因

原文

AIMS/HYPOTHESIS: The aim of this study was to facilitate the fine classification of diabetic complications and the prioritisation of drug targets via immune-cell-related gene expression data, distinguishing cell-type-related pleiotropy. METHODS: We analysed the combined clinical, genetic and immune-related features of diabetic complications, generating distinct clusters. Using 18,611 immune expression quantitative trait loci (eQTL) of 4487 genes, we investigated the causal effects of immune-cell-related gene expression on the risks of type 2 diabetes and its complications through Mendelian randomisation (MR) and colocalisation approaches. We then explored cell-type-related enrichment and specificity, and attenuated the cell-type-related pleiotropy for the top MR findings via multivariable MR methods. RESULTS: Clustering analyses identified diabetic neuropathy as a distinct cluster of diabetes complications, with distinct immune-related features. MR and colocalisation analyses revealed the expression of 425 and 123 unique genes associated with type 2 diabetes and its complications, respectively, with external validation performed using single-cell RNA-seq data. We further quantified the impacts of cell-type-related pleiotropy, demonstrating that the percentage of pleiotropic genes increased from 40.0% (classic pleiotropy) to 71.1% (classic and/or cell-type-related pleiotropy). Applying six multivariable MR (MVMR) methods substantially attenuated the cell-type-related pleiotropy for the top findings. Finally, we integrated clinical trial evidence with genetic evidence and prioritised ten immune-related drug targets for diabetic complications. CONCLUSIONS/INTERPRETATION: Our study supports a key role for immune mechanisms in diabetic complications and highlights promising therapeutic targets by distinguishing and minimising the influence of cell-type-related pleiotropy.