Genetically determined platelet traits impact stroke risk through multiple mechanisms and cell types

Abstract

Stroke remains a leading cause of death and disability worldwide. Current antiplatelet and anticoagulant treatments are prone to failure. Heritable blood and platelet traits contribute to stroke risk, but related mechanisms are not fully understood. Platelets bind to damaged endothelial walls to initiate thrombosis, but erythrocyte and leukocyte recruitment are involved in stroke pathogenesis. We aimed to identify causal blood-related cells and mechanisms that modulate stroke risk. By two sample Mendelian Randomization (MR), increased platelet count heightened stroke risk (Odds ratio [OR] 1.03 per 1 SD unit increase in platelet count, P=1x10-2). However, these effects were relatively weak and complicated by similar effects from erythrocyte and leukocyte traits. To ascertain key blood traits that influence stroke, we applied Bayesian Model Averaging (MR-BMA) and identified platelet count and mean platelet volume as key positive regulators for stroke risk. We validated an epidemiologic association between increased platelet count and higher stroke risk among a large patient cohort. Taken together, these findings indicate that platelet traits are the most critical risk factors for stroke, among analyzed blood cell traits. To deconvolute multiple underlying genetic mechanisms by which platelet traits impact stroke risk, we clustered platelet count variants using noise-augmented directional clustering (NAvMix). We identified 13 clusters, two of which were highly predictive for increased stroke risk (OR 1.31 per SD unit increase in platelet count, P<1x10-6). Pathway analyses on eQTLs linked to variants in these subclusters indicated enrichment for endothelial cell adhesion or platelet reactivity. Colocalization analysis of stroke and platelet count loci identified genes implicated in platelet reactivity (RIN3) and peroxisome biogenesis (PEX6/PEX29). These findings reflect complex mechanisms underlying platelet trait variation and reveal key pathways that influence stroke risk through multiple cell types and biological mechanisms, including platelet biology and endothelial cell adhesion. An approach combining novel MR methods with subclustering may be a viable method to ascertain causal mechanisms related to other closely related exposure traits.