Integration of single-cell and bulk transcriptomics reveals immune-related signatures in keloid.

Background: Keloid is a pathological dermatological condition that manifests as an overgrowth scar secondary to skin trauma. This study endeavored to excavate immune-related signatures of keloid based on single-cell RNA (scRNA) sequencing data and bulk RNA sequencing data.

Method: The keloid-relevant scRNA sequencing dataset GSE163973 and bulk RNA sequencing dataset GSE113619 were mined from the GEO database. The "Seurat" R package was utilized for data quality control, cell clustering, and investigation of marker genes of each cell cluster. The "SingleR" package helped match the marker genes of the corresponding cluster to specific cell types. Moreover, the R package "Monocle" was deployed for pseudotemporal ordering analysis, and the "clusterProfiler" was applied for functional and pathway enrichment analysis. The immune-related signatures were then identified, and potential targeted drugs were predicted via the DGIdb database. Verification of the immune-related signatures in clinical validation samples was implemented by RT-qPCR.

Results: Totally 23 cell clusters were screened and classified into 10 cell types based on the scRNA sequencing data. The keloid group had a significantly higher endothelial cell proportion than the control group. As enrichment analysis was applied in both differentially expressed genes (DEGs) of scRNA and bulk RNA sequencing data, we found they were enriched in multiple common immune-related pathways and biological processes. Meanwhile, we acquired three immune-related signatures (VCAM1, CALCRL, and HLA-DPB1) by intersecting the above DEGs with immune-related genes (IRGs). Then, we predicted 16 drugs potentially targeting the biomarkers through the DGIdb database. Finally, the outcome of RT-qPCR of clinical validation samples further verified the results.

Conclusion: In conclusion, we analyzed the cell types and functional differences in the keloid through scRNA and bulk RNA sequencing data. We identified three immune-related signatures (VCAM1, CALCRL, and HLA-DPB1) in keloid, providing a basis for further in-depth investigation of the molecular mechanisms of keloid and exploration of therapeutic targets.