Integrated Bioanalysis and In Vivo Validation of Mitochondrial Energy Metabolism–Related Genes in Diabetic Nephropathy

Abstract Background: Diabetic nephropathy (DN) is a leading cause of end-stage renal disease, and early diagnostic biomarkers remain limited. Given the critical role of mitochondrial energy metabolism in DN pathogenesis, this study aimed to identify mitochondrial energy metabolism–related diagnostic signatures and associated immune features. Methods: Three GEO microarray datasets (GSE30528, GSE96804, and GSE104948) were integrated after batch-effect correction. Differential expression analysis identified mitochondrial energy metabolism–related differentially expressed genes (MEMRDEGs). A diagnostic signature was developed using logistic regression for screening, SVM-RFE for feature refinement, and LASSO regression for final model construction, generating a seven-gene risk score model (PDK4, CKB, UCHL1, ADH1B, ALDH1L1, CR1, and PC). Diagnostic performance was evaluated using ROC analysis. Functional enrichment analyses (GO/KEGG) and immune infiltration estimation (ssGSEA) were performed to explore underlying biological pathways and immune characteristics. Regulatory networks (mRNA–TF and mRNA–miRNA) were further constructed for key genes, and immunohistochemistry was used to validate protein-level expression changes in DN tissues. Results: A total of 20 MEMRDEGs were identified. LASSO regression yielded a diagnostic model comprising seven genes, which demonstrated excellent discriminative ability (AUC > 0.9). Functional enrichment suggested that DN was characterized by activation of inflammatory, TGF-β, and PI3K–AKT–mTOR signaling pathways. Immune infiltration analyses revealed prominent immune microenvironment remodeling, with notable involvement of CD8 T cells and regulatory T cells (Tregs). Immunohistochemistry confirmed the protein-level expression alterations of the seven signature genes in DN kidney tissues. Conclusions: This study establishes a robust mitochondrial energy metabolism–based seven-gene diagnostic signature for DN and highlights metabolic–immune crosstalk associated with disease progression. The identified genes and immune-associated pathways may offer promising molecular targets for early diagnosis and therapeutic intervention in DN.