Abstract
BACKGROUND: Diabetic retinopathy (DR) remains a leading cause of vision impairment worldwide, with limited therapeutic options and early detection methods. This study aims to identify novel biomarkers for DR to address the critical need for therapeutic targets. METHODS: We employed a multi-omics strategy combining scRNA-seq analysis of DR-related immune cell populations with bidirectional MR integrating quantitative trait loci (eQTL)/protein quantitative trait loci (pQTL) data to establish HLA-A's causal role in DR. Clinical validation utilized Chinese cohort peripheral blood samples (n = 81), complemented by molecular docking and cellular experiments testing resveratrol's modulation of HLA-A under metabolic stress conditions. RESULTS: Bidirectional Mendelian randomization identified HLA-A as a causal risk factor for DR, with consistent evidence from both expression quantitative trait loci (eqtl: OR = 1.121, 95% CI [1.057, 1.19], p = 1.584 × 10(-4). pqtl: OR = 1.136, 95% CI [1.095, 1.178], p = 1.05 × 10(-11)). Sensitivity analyses showed no evidence of horizontal pleiotropy (p = 0.234) or heterogeneity (p = 0.199). Shared causal variants (PP.H4 > 0.8) emerged from Bayesian colocalization. Clinically, peripheral blood HLA-A mRNA levels were significantly increased in DR patients (HbA1c ≥ 7.5%). Critically, in glucolipotoxic conditions, HLA-A knockdown suppressed pro-inflammatory cytokines in THP-1 cells, while glucolipotoxicity potently upregulated HLA-A expression, which was dose-dependently inhibited by resveratrol. CONCLUSION: This multi-omics study establishes HLA-A as a novel causal factor in DR pathogenesis and reveals resveratrol's potential therapeutic value through HLA-A modulation, providing new insights for targeted intervention strategies.