Related Experiment Videos
Integrated single-cell transcriptomic and machine-learning analyses identify a CD36-associated macrophage state in
Xin Yang1, Jiangying Li2, Ying Zeng3
1Department of Nephrology, Ganzhou People's Hospital, Ganzhou, Jiangxi, China.
Background:
Diabetic nephropathy (DN) is the leading cause of end-stage renal disease worldwide and is characterized by progressive glomerulosclerosis, tubulointerstitial fibrosis, and a chronic inflammatory microenvironment. Although macrophage infiltration and immune dysregulation are recognized as central pathological features, the cell-type-specific transcriptional programs that drive inflammatory amplification in DN remain incompletely understood, particularly at single-cell resolution.
Objective:
To identify macrophage-associated transcriptional biomarkers and delineate intercellular communication networks in DN using an integrated workflow combining urine single-cell RNA sequencing, multi-algorithm machine learning, and tissue-level experimental validation.
Methods:
Urine scRNA-seq data from 17 healthy-control libraries (GSE157640 and GSE165396) and 8 diabetic nephropathy (DN) libraries (GSE266146) were processed using Seurat and integrated using Harmony. IL6-JAK-STAT3 pathway activity was quantified using AUCell. The GSE30122 cohort was stratified by disease status and randomly divided into training and held-out internal validation sets. Candidate-gene selection using LASSO, random forest, and Boruta was performed exclusively in the training set, and the discriminatory performance of the consensus genes was subsequently evaluated in the held-out internal validation set using receiver-operating-characteristic analysis. Cell-cell communication was computationally inferred using CellChat. Renal CD36 expression and its spatial association with F4/80-positive macrophages were assessed in an STZ-induced mouse model.
Results:
Single-cell analysis identified major epithelial and immune cell populations in the urine of patients with DN and controls. IL6-JAK-STAT3 pathway activity was significantly elevated in DN, with macrophages showing the highest enrichment. Machine learning identified CD36 as a consensus biomarker with an AUC of 0.894 in the external validation cohort. CD36 was predominantly expressed in macrophages, and CellChat analysis predicted inflammatory and profibrotic ligand-receptor communication patterns associated with CD36-detected macrophages. In DN mice, CD36 was significantly upregulated at both the mRNA and protein levels and co-localized with F4/80-positive macrophages in renal tissue.
Conclusions:
CD36 represents a macrophage-enriched candidate molecular marker associated with an IL6-JAK-STAT3-high transcriptional state and predicted intercellular communication patterns in DN. These findings support further investigation of CD36, although its clinical utility and therapeutic relevance require independent clinical and functional validation.
