Single-cell reanalysis identifies macrophage-associated transcriptional and intercellular communication features in diabetic foot ulcers.

BACKGROUND: Diabetic foot ulcers (DFUs) are devastating diabetic complications characterized by an extremely high non-healing rate and elevated risk of life-threatening outcomes. To characterize the cellular landscape and macrophage-associated alterations between DFU tissue and diabetic non-ulcer foot (DNFU) skin, we performed a comprehensive reanalysis of publicly available single-cell RNA sequencing (scRNA-seq) datasets.

METHODS: A fully reproducible R-based analytical pipeline was used to process public scRNA-seq count matrices (GSE165816) from 5 DFU and 8 DNFU skin samples. A total of 27,844 cells (16,579 from DFU and 11,265 from DNFU skin) passed stringent quality control filters. Data normalization, integration, dimensionality reduction, unsupervised clustering, and marker-based cell type annotation were performed using the Seurat and Harmony packages. Sample-level comparisons included cell fraction analysis and UCell pathway scoring. Macrophages were further investigated through subclustering analysis, edgeR-based pseudobulk differential expression analysis, volcano plot visualization, GO over-representation analysis, enrichment analysis, and CellChat ligand-receptor interaction inference.

RESULTS: The integrated single-cell atlas resolved 10 major cell compartments including epithelial, stromal, endothelial, lymphatic endothelial, smooth muscle/pericyte, lymphoid, myeloid, mast cell, plasma cell, and melanocyte/Schwann cells. Sample-level cell fraction analysis revealed a nominal reduction in lymphatic endothelial cells in DFU; however, no cell type proportion differences remained statistically significant after false discovery rate (FDR) correction. Similarly, no significant differences were observed in macrophage UCell pathway scores for glycolysis, lactate metabolism, combined glycolysis-lactate metabolism, TNF-NF-κB signaling, or insulin/IGF signaling between DFU and DNFU skin following FDR correction. Macrophage subclustering identified DFU-enriched subpopulations, and macrophage pseudobulk analysis prioritized multiple DFU-associated genes. Notably, sex-linked genes including XIST, RPS4Y1, and DDX3Y reached statistical significance after FDR correction, indicating that a portion of the observed transcriptional signal may be attributed to demographic imbalance between groups. Exploratory enrichment analysis of nominal and ranked macrophage differentially expressed genes highlighted DFU-directional enrichment of translation, oxidative phosphorylation, electron transport, and antigen presentation terms, while DNFU-directional genes were enriched for transcriptional regulation and rhythmic process terms. CellChat analysis prioritized macrophage-associated CD99-CD99 and MIF-CD74/CXCR4 ligand-receptor pairs as key DFU-associated intercellular communication hypotheses.

CONCLUSION: This single-cell reanalysis comparing DFU and DNFU skin identifies putative ligand-receptor interaction hypotheses, transcriptomic alterations, and macrophage subpopulation shifts that require definitive validation in larger, clinically well-characterized, and spatially resolved patient cohorts.