Proteome-wide genetic analyses identify ASGR1 as a potential biomarker linking smoking with ischemic heart disease

Tobacco smoking increases the risk of multiple diseases, but the biological mechanisms involved remain incompletely understood. Here we characterize smoking-related protein profiles to identify potentially causal biomarkers mediating smoking's effects. In European-ancestry populations, genetically predicted smoking exposures were significantly associated with levels of 93 (total: 2,922) proteins, including higher levels of asialoglycoprotein receptor 1 (ASGR1). The positive association of ASGR1 levels with ischemic heart disease (IHD) was supported by colocalization (posterior probability of hypothesis 4 = 0.91), and was estimated to explain 6.35% (95% confidence interval 2.61-10.09%, P = 8.77 × 10^-4) of smoking's effects on IHD in two-step Mendelian randomization analyses. In network analyses, genetically predicted ASGR1 levels were inversely associated with plasma levels of more than 100 proteins, including growth arrest-specific 6, with all these proteins having rs55714927 (that is, a cis-protein quantitative trait locus for ASGR1) as a trans-protein quantitative trait locus. Sensitivity analyses in never-smokers identified a nonstatistically significant residual association between ASGR1 and IHD, which is consistent with possible horizontal pleiotropy, with body mass index implicated as a potential contributing pathway.