Abstract

Objective: Neutrophil extracellular trap formation (NETosis) is a pivotal pathogenic process in lupus nephritis (LN), but the precise mechanistic role of neutrophil extracellular traps (NETs) in LN remains unclear. This study aims to identify potential diagnostic biomarkers for LN through bioinformatics analysis of genes associated with NETs and explore their correlations with clinical parameters, providing new insights for LN diagnosis and treatment.

Methods: 1. Differentially expressed genes (DEGs) were screened from the GSE32591 dataset, followed by functional enrichment (GO/KEGG) and machine learning (LASSO, RF, SVM-RFE) to prioritize hub genes. The diagnostic value of LN-related biomarkers was further assessed by the receiver operating characteristic (ROC) curves in the validation dataset. LN samples were subjected to unsupervised clustering (WGCNA) and single-sample gene set enrichment analysis (ssGSEA). 2. Immunohistochemical (IHC) staining of ITGB2 and CYBB proteins was performed on renal tissue specimens from 18 LN patients and 6 controls, with Image-Pro Plus 6.0 quantifying the average optical density (AOD) to validate the differential expression of key genes. Correlation analysis between protein expression levels and clinical parameters was conducted using GraphPad Prism 10.

Results: Bioinformatics analysis revealed DEGs enriched in immune response, viral infection, and NF-κB pathways. Machine learning prioritized ITGB2, CYBB, and G0S2, with ITGB2/CYBB showing high diagnostic accuracy. Unsupervised clustering based on CYBB/ITGB2 expression profiles stratified LN patients into two molecular subtypes. ssGSEA revealed immune cell subtypes: Cluster B (high CYBB/ITGB2 expression) exhibited Treg/Th17 imbalance and hyperactivation of myeloid cells, suggesting a NETs-driven immune microenvironment heterogeneity. IHC confirmed elevated expression of ITGB2(LN: 0.179±0.018 vs. controls: 0.159±0.028, P < 0.05) and CYBB (LN: 0.152 ± 0.011 vs. controls: 0.144 ± 0.010, P < 0.05) in renal tissues, with CYBB showing glomerular upregulation (LN: 0.180±0.012 vs. controls: 0.156±0.017, P <0.001). Correlation analysis demonstrated that ITGB2 levels were negatively correlated with the chronicity index (CI) (r = -0.610, P = 0.007), while CYBB expression was positively correlated with serum creatinine (r = 0.606, P = 0.008) and inversely correlated with the estimated glomerular filtration rate (eGFR) (r = -0.571, P = 0.013).

Conclusion: Through integrated bioinformatics and machine learning algorithms, CYBB and ITGB2 were identified as key regulatory genes of NETs in LN. Further, IHC CYBB exhibited significantly elevated expression in glomeruli, with significant correlations to SCr and eGFR. ITGB2 expression showed a negative correlation with CI, highlighting its distinct clinical relevance as a therapeutic target.