Integrative multi-omics analysis identifies microbial dysbiosis and functional metabolic reprogramming in acute kidney injury.

Background: Acute kidney injury (AKI) is a life-threatening syndrome with high morbidity and mortality, yet its early diagnosis and underlying mechanisms remain poorly defined. Emerging evidence implicates gut dysbiosis and microbial metabolic dysfunction in AKI pathogenesis via the gut-kidney axis, yet a comprehensive, multi-omics characterization of microbial functional alterations in general AKI populations remains lacking. Methods: We conducted a prospective multi-omics study including 16 patients with acute kidney injury (AKI) and 16 age- and sex-matched healthy controls (HCs). Plasma metabolomic profiling was performed using ultra-performance liquid chromatography coupled with quadrupole time-of-flight mass spectrometry (UPLC-QTOF/MS). Gut microbiome composition and function were characterized through whole-metagenome sequencing of stool samples. Differential taxonomic and metabolite features were identified using multivariate and univariate statistical analyses. Microbial functional potential was assessed across four hierarchical layers: Kyoto Encyclopedia of Genes and Genomes (KEGG) Orthologs (KOs) genes, pathways, gut-metabolite modules (GMMs), and gut-brain modules (GBMs), to achieve high-resolution mapping of metabolic pathways and taxon-specific functional contributions. Integrated microbe-metabolite-phenotype relationships were evaluated using Spearman correlation analysis. Results: Metabolomic profiling identified 65 differentially abundant metabolites between AKI patients and healthy controls (HCs), including 53 upregulated and 12 downregulated metabolites. These metabolites were mainly enriched in carbohydrate metabolism (e.g., starch and sucrose metabolism, fructose and mannose metabolism) and amino acid metabolism pathways. Among them, Maltol (C11918, AUC = 0.961), D-Quinovose (C02522, AUC = 0.926), and L-fucose (CO1019, AUC = 0.926) demonstrated the most robust diagnostic potential. Further feature selection using a random forest model identified an optimal panel of three metabolites, which achieved good discriminative performance (AUC = 0.859, 95% CI: 0.7073-1). Metagenomic analysis revealed significant gut microbiota dysbiosis in AKI, characterized by reduced α-diversity and distinct β-diversity compared to HCs. Taxonomic profiling showed depletion of key short-chain fatty acid-producing bacteria, including Conclusion: This multi-omics study reveals coordinated gut microbial dysbiosis and systemic metabolic reprogramming in AKI. The depletion of key commensals, rather than pathogen overgrowth, appears central to AKI-associated functional disruption. These findings highlight potential microbial and metabolic biomarkers and offer mechanistic insights into AKI pathogenesis.