Abstract

Background Remote ischemic preconditioning (RIPC) is a promising, non-invasive strategy for reduction in ischemic injury after stroke, yet its molecular mechanisms and translatability from animal models to humans remain insufficiently defined. In particular, whether RIPC-induced gene programs causally contribute to reduced stroke risk in humans is unclear. Here, we aimed to establish a translational framework linking RIPC-induced gene regulation in mice to causal genetic and epigenetic determinants of stroke risk in humans. Methods Using a mouse transient middle cerebral artery occlusion (tMCAO) model, we performed multi-tissue transcriptomic profiling across central (cerebral cortex, hippocampus, striatum, hypothalamus, and insular cortex) and peripheral (spleen and peripheral blood mononuclear cells; n  = 5/group) tissues under five experimental conditions. Mouse differentially expressed genes were mapped to human orthologs and integrated with human expression quantitative trait loci data and large-scale, multi-ancestry stroke genome-wide association studies using Mendelian randomization to identify RIPC-regulated beneficial genes (RBGs). DNA methylation sites within RBGs were further evaluated, and machine-learning models were constructed to assess their predictive potential. Key genes were experimentally validated in the MCAO model. Results We identified 23 core RBGs showing significant causal associations with reduced stroke risk. Functional enrichment analyses linked these genes to pathways involved in neuroprotection, immune modulation, and tissue repair. Mendelian randomization analyses supported causal effects of specific RBG-associated DNA methylation sites on stroke risk. Predictive modeling based on combined RBG signatures demonstrated strong translational performance (AUC = 0.97). Experimental validation confirmed that RIPC upregulated key RBGs, including Abcc5, Fam83h, Pom121, Mst1r, and Metap1d ( n  = 12/group, p  < 0.05), accompanied by reduced in