Integrative transcriptomic and genomic insights into diabetic kidney disease: evidence from multi-omics analysis and experimental validation

Diabetic kidney disease (DKD) remains a critical challenge in diabetes management, necessitating a deep understanding of its molecular underpinnings for better diagnosis and treatment strategies. This study was conducted to identify and validate novel biomarkers for DKD by integrating multi-omics analysis and experimental validation. Through weighted gene co-expression network analysis and Mendelian randomization analysis, 11 genes were identified as being causally associated with DKD. ADARB2, GOLPH3L, LRG1, and PEX6 were identified as characteristic genes through machine learning methods, including least absolute shrinkage and selection operator (LASSO) regression and SVM algorithms. Receiver operating characteristic curve analysis demonstrated that the characteristic genes had high predictive accuracy for DKD. Functional enrichment analyses indicated that dysregulation of key genes was associated with inflammatory and immune responses in both peripheral blood mononuclear cells and kidney single-cell populations. Peripheral blood samples from DKD patients and healthy controls were collected to assess the reliability of identified key genes in humans. Kidneys from wild-type and db/db mice were harvested to further validate the reliability of key genes at the tissue level in animal models using reverse transcription-quantitative polymerase chain reaction (RT-qPCR). After RT-qPCR validation, the robustness of LRG1 and PEX6 was confirmed in both human peripheral blood and mouse kidney tissues. The significance of ADARB2 was confirmed in the kidney tissue of DKD mouse models. This study highlights the power of multi-omics analyses in elucidating complex disease pathogenesis and identifying biomarkers, thereby laying a foundation for the development of DKD-targeted therapeutics.