A hypoxia-derived seven-gene signature reflects immune remodeling in diabetic kidney disease: integrated transcriptomic analysis and experimental validation.

AIMS: To develop and assess a hypoxia-associated gene signature for diabetic kidney disease (DKD), and to examine its relationship with immune remodeling and molecular heterogeneity. METHODS: GEO renal transcriptomic datasets were integrated to identify hypoxia-related differentially expressed genes. Weighted gene co-expression network analysis, support vector machine-recursive feature elimination, and LASSO regression were used to construct a diagnostic gene signature. Immune infiltration was evaluated using CIBERSORT and ssGSEA. The expression and biological relevance of the signature were further assessed in diabetic animal models, Nephroseq datasets, and plasma RT-qPCR samples. RESULTS: A seven-gene signature comprising EGF, CASP3, HDAC9, FN1, SERPINF1, VWF, and NT5E was identified. The signature showed good internal discrimination between DKD and control kidney samples (AUC = 0.879, 95 % CI: 0.820-0.937) and was associated with macrophage- and neutrophil-related immune remodeling. Score-defined subgroups and consensus clustering revealed distinct transcriptional and immune patterns. Animal models, external renal datasets, and plasma RT-qPCR analyses supported directionally consistent expression changes. CONCLUSIONS: This hypoxia-associated seven-gene signature showed diagnostic potential and reflected immune and molecular heterogeneity in DKD. Further prospective and mechanistic studies are required before clinical application.