Single-cell eQTL-based Mendelian randomization identifies immune cell subtype-specific regulators of epigenetic aging and prioritizes candidate therapeutic targets.

Abstract

Epigenetic aging clocks offer precise measures of biological age, yet the causal contributions of immune gene expression within specific cell subtypes to epigenetic aging remain poorly understood. By integrating single-cell eQTL data from the OneK1K cohort with GWAS summary statistics for four epigenetic clocks (HannumAge Acceleration, IEAA, PhenoAge Acceleration, and GrimAge Acceleration), we performed two-sample Mendelian randomization across diverse immune cell subtypes, followed by colocalization analysis and gene-level phenome-wide association studies. We identified 11 eGene-cell type pairs surviving Bonferroni correction, including NUCKS1 in CD4 NC T cells and NCR3 in Classic Monocytes as risk-increasing eGenes for HannumAge Acceleration, and HSPA1B in Classic Monocytes as protective across multiple clocks. ANP32E in Classic Monocytes represented the strongest risk signal for GrimAge Acceleration (OR = 2.683), while BCAS4 in CD8 EM T cells was the strongest protective association (OR = 0.683). Colocalization confirmed NUCKS1 (PP.H4 = 87%) and NCR3 (PP.H4 = 69%) as high-confidence causal eGenes, and PheWAS revealed no genome-wide significant off-target associations for the prioritized targets, supporting their specificity. These findings establish cell subtype-specific causal roles for immune gene expression in epigenetic aging and prioritize NUCKS1, NCR3, and ANP32E as candidate targets for interventions aimed at promoting healthy aging.