Integrative multi-omics analysis identifies VCAN and ESM1 as candidate cell-type-associated biomarkers of NET-related glomerular injury in diabetic kidney disease

Abstract

Abstract Diabetic kidney disease (DKD) involves complex inflammatory and microvascular injury, but cell-specific molecular signatures linked to neutrophil extracellular traps (NETs) remain unclear. We aimed to identify NET-associated biomarkers of glomerular injury in DKD. Bulk glomerular transcriptomic datasets were analyzed using differential expression, weighted gene co-expression network analysis, and machine learning (LASSO and random forest). Single-cell RNA sequencing (scRNA-seq) defined cell-specific expression and intercellular communication. Candidates were evaluated in independent cohorts, in vitro high-glucose co-culture datasets, and db/db mice. Exploratory in silico drug-screening was also performed. VCAN and ESM1 emerged as core candidate genes. In bulk cohorts, VCAN was upregulated and ESM1 downregulated, yielding strong combined diagnostic performance. scRNA-seq revealed ESM1 decreases predominantly in glomerular endothelial cells, while VCAN increases in mesangial cells. NET-related activity enriched primarily in the endothelium, whereas VCAN-expressing cells associated with matrix remodeling. In vitro, ESM1 showed early endothelial-predominant decreases under injury. In db/db mice, renal VCAN and NET markers were elevated, while ESM1 levels were reduced and correlated with albuminuria severity. Molecular docking identified potential targeting compounds. In conclusion, VCAN and ESM1 are promising biomarkers reflecting distinct, cell-type-enriched patterns of NET-related glomerular injury in DKD, supporting a dual-track model of endothelial dysfunction and paracrine matrix remodeling.