Integrated single-cell and bulk transcriptomic analysis identifies epithelial cell-related biomarkers and immune subtypes in focal segmental glomerulosclerosis.

backgroundFocal segmental glomerulosclerosis (FSGS) is a major cause of nephrotic syndrome and chronic kidney disease, characterized by podocyte injury and glomerular scarring. Its heterogeneity complicates early diagnosis and treatment.

methodsWe integrated bulk and single-cell RNA-seq datasets (GSE200818, GSE200828, and GSE176465) to investigate epithelial cell-related molecular alterations in FSGS. Differentially expressed genes (DEGs) were identified and analyzed for functional enrichment. Machine learning algorithms (random forest and support vector machine) were applied to construct a diagnostic gene signature. Immune cell infiltration was assessed using ssGSEA, and molecular subtypes were defined via consensus clustering. Independent datasets (GSE133288 and GSE108109) were used for validation.

resultsA total of 133 epithelial cell-associated DEGs were identified. Enrichment analysis indicated transcriptional regulation and viral infection pathways. Three diagnostic genes-B4GALT5, CSRNP1, and CYP3A5-were selected and shown to have excellent diagnostic accuracy (AUC > 0.97) in both discovery and validation cohorts, including podocyte-enriched samples. Immune profiling revealed increased infiltration of activated CD8⁺ T cells, NK cells, and dendritic cells in FSGS, with B4GALT5 positively correlated with multiple immune subsets. Consensus clustering stratified patients into two molecular subtypes with distinct immune features.

conclusionsThis integrative analysis identifies epithelial cell-related biomarkers and immune-associated subtypes in FSGS, providing promising tools for diagnosis and patient stratification. While robust across multiple datasets, further experimental validation is needed to confirm mechanistic relevance.