Abstract
Large case-control genome-wide association studies (GWASs) have detected loci associated with insomnia, but how these risk loci confer disease risk remains largely unknown. By integrating brain protein quantitative trait loci (pQTL) (N(pQTL1) = 376, N(pQTL2) = 152) and expression QTL (eQTL) (N = 452) datasets, with the latest insomnia GWAS summary statistics (N(case) = 109,548, N(Controls) = 277440), we conducted proteome/transcriptome-wide association study (PWAS/TWAS) and Mendelian randomization (MR) analysis, aiming to identify causal proteins involving in the pathogenesis of insomnia. We also explored the bi-directional causality between insomnia and several common diseases. As a result, the altered protein level of 28 genes in the brain was associated with the risk of insomnia in the discovery stage of PWAS, of which 18 genes' associations were replicated in the confirmatory stage of PWAS. Among them, four proteins (2-aminoethanethiol dioxygenase (ADO), calcium-modulating cyclophilin ligand (CAMLG), islet cell autoantigen 1 like (ICA1L) and latexin (LXN)) were found to be the most likely causal genes for insomnia with validations from TWAS, MR, and colocalization results. Specifically, the higher protein level of ADO, CALMG, and ICA1L was causally associated with a lower risk of insomnia. In comparison, the higher protein level of LXN was causally associated with an increased risk for insomnia. Moreover, genetically predicted insomnia was causally associated with an increased risk of developing cardiovascular diseases and depression. In conclusion, our study identified ADO, CAMLG, ICA1L, and LXN as potentially causal proteins in the pathogenesis of insomnia. This could provide insights into further mechanistic studies and therapeutic development for insomnia.