Unveiling the role of oxidative stress in ANCA-associated glomerulonephritis through integrated machine learning and bioinformatics analyses.

Anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis (AAV) is a systemic autoimmune disease often leading to rapidly progressive glomerulonephritis. Oxidative stress plays a critical role in the development and progression of ANCA-associated glomerulonephritis (AAGN), but the underlying mechanisms remain poorly understood. Targeting genes related to oxidative stress may provide novel insights and supplementary therapeutic benefits for AAGN. In the current study, we obtained differentially expressed genes from AAGN-related microarray datasets in the Gene Expression Omnibus database, and oxidative stress-related genes (OSRGs) from the GeneCards and Gene Ontology databases to identify differentially expressed OSRGs. Then, by integrating weighted gene co-expression network analysis, and machine learning algorithms, we identified four upregulated hub OSRGs (all p < 0.01) with strong diagnostic potential (all AUC > 0.9)-CD44, ITGB2, MICB, and RAC2 - in the AAGN glomerular training dataset GSE104948 and validation dataset GSE108109, along with two hub OSRGs (all p < 0.05) with better diagnostic potential (all AUC > 0.7) - upregulated gene VCAM1 and downregulated gene VEGFA-in the AAGN tubulointerstitial training dataset GSE104954 and validation dataset GSE108112. The GSEA analysis suggested that these hub genes may play a role in inflammatory and immune response processes. Moreover, we constructed regulatory networks and identified drugs that potentially target these hub genes. It's to be noted that RAC2 and ITGB2 were associated with cyclophosphamide in the AAGN glomerular compartment, while VCAM1 and VEGFA were associated with dexamethasone in the tubulointerstitial compartment. This study offers novel insights into immune-associated OSRGs within the glomerular and tubulointerstitial compartments of AAGN which may serve as innovative targets for diagnosing and treating AAGN.