HLX and SLC25A20: Immunologic regulators bridging ankylosing spondylitis and uveitis via multi-omics integration and machine learning

Abstract

Background Ankylosing spondylitis (AS), a chronic inflammatory disorder affecting axial joints, is frequently complicated by uveitis. However, the molecular mechanisms linking AS to secondary uveitis remain poorly understood. Methods We integrated transcriptomic datasets from AS (GSE73754) and uveitis (GSE194060) cohorts to identify shared molecular pathways. Differential expression analysis, weighted gene co-expression network analysis (WGCNA), and machine learning (LASSO, SVM-RFE, random forest) were combined to prioritize biomarkers. Molecular docking was performed to evaluate drug-target interactions using Vina scores (≤−7 kcal/mol threshold). Results Cross-disease analysis revealed 697 overlapping dysregulated genes (481 upregulated, 216 downregulated), enriched in GTPase signaling and immune pathways. WGCNA identified disease-specific co-expression modules (AS: brown/tan modules, r=0.39/0.35; uveitis: brown module, r=0.49). Machine learning nominated HLX and SLC25A20 as core biomarkers, demonstrating robust diagnostic accuracy in discovery (AS AUC: 0.688/0.700; uveitis AUC: 0.867/0.838) and validation cohorts (AS AUC: 0.653/0.667; uveitis AUC: 0.662/0.736). Immune profiling linked HLX to neutrophil infiltration (r=0.55, p<0.01) and SLC25A20 to T helper cell regulation (r=0.36, p<0.01). Molecular docking identified high-affinity ligands for SLC25A20, including amiodarone (−8.0 kcal/mol) and estradiol (−7.7 kcal/mol), with folic acid showing dual binding potential (HLX: −7.5 kcal/mol; SLC25A20: −8.2 kcal/mol). Conclusion HLX and SLC25A20 emerge as immunologic regulators bridging AS and uveitis pathogenesis. These findings provide actionable targets for precision diagnostics and therapeutic development in AS-associated uveitis.