Genetically Predicted Associations of CX3CL1 and UMOD with Acute Kidney Injury Risk: Evidence from Mendelian Randomization and Clinical Validation

Acute kidney injury (AKI) is a common clinical syndrome with poor prognosis and limited targeted therapies. This study aimed to systematically identify genetically supported candidate proteins associated with AKI and to explore potential biological pathways and phenotypic associations. Using data from the UK Biobank Pharma Proteomics Project (UKB-PPP), FinnGen, IEU Open GWAS, and other large-scale databases, we applied summary-data Mendelian randomization (SMR), Bayesian colocalization analysis, two-sample Mendelian randomization (MR), mediation analysis, and phenome-wide association study (PheWAS) to assess the causal relationships and mechanisms linking plasma proteins to AKI risk. We collected blood samples undergoing cardiac surgery with cardiopulmonary bypass to measure the levels of CX3CL1 and UMOD in the plasma and compare them with indicators such as serum creatinine and urinary [TIMP2]*[IGFBP7], to evaluate the efficacy of CX3CL1 and UMOD in diagnosing AKI. Two plasma proteins, CX3CL1 and uromodulin (UMOD), were identified as significantly and positively causally associated with AKI risk, supported by genetic colocalization evidence. Mediation analysis indicated that CX3CL1’s effect on AKI may be partially mediated by metabolites such as 1,6-anhydroglucose and Mannitol. PheWAS revealed associations of these proteins with hypertension, carbohydrate metabolism disorders, and infections, suggesting potential off-target effects that should be considered in drug development. The associations of these two proteins were further evaluated in human plasma samples. We found that patients with I/R-induced AKI exhibited significantly elevated plasma levels of CX3CL1 and UMOD compared to those who underwent ischemia-reperfusion without developing AKI (P< 0.05). Spearman's test demonstrated a significant correlation between serum creatinine and plasma CX3CL1 and UMOD in patients with I/R induced AKI (r = 0.848, P < 0.001; r = 0.794, P< 0.001, respectively). Plasma CX3CL1 and UMOD showed high specificity but low sensitivity in diagnosing AKI (sensitivity 68.0% and 60%;specificity 92.0% and 88%,respectively).This study systematically elucidates the genetic evidence and possible pathogenic mechanisms of CX3CL1 and UMOD as candidate proteins for AKI, CX3CL1 and UMOD demonstrated potential associations with I/R-induced AKI risk in the present cohort., enriching our understanding of AKI pathogenesis and providing theoretical and safety considerations for future targeted drug development.