Single-cell RNA sequencing reveals distinct immunology profiles in human keloid

Abstract

Keloids, characterized by skin fibrosis and excessive accumulation of extracellular matrix, remain a therapeutic challenge. In this study, we systematically capture the cellular composition of keloids by the single-cell RNA sequencing technique. Our results indicated that there are significant differences in most cell types present between 12 pairs of keloid and adjacent normal tissue. We found that fibroblasts, endothelial cells, mast cells, mural cells, and Schwann cells increased significantly in keloid. The proportion of mesenchymal fibroblast subpopulations in keloids was markedly higher than those in the surrounding normal skin tissue. Furthermore, we found that the immune profiles between two groups varied significantly. The proportion of macrophages in the keloid was significantly elevated compared to the surrounding normal tissue, while cDC2 cells significantly decreased. Hotspot and pseudotime trajectory analysis indicated two modules of macrophage cells (Module2: highly expresses RNASE1, C1QA, CD163, CD14, C1QC, FCGRT, MS4A7; Module10: highly expresses APOC1, CTSB, CTSL, TYROBP), which exhibited the characteristics of tumor-associated macrophages, were upregulated in more-advanced keloid cells. Subsequently, the analysis of cellular communication networks suggested that a macrophage-centered communication regulatory network may exist in keloids and that fibroblasts in keloids may facilitate the transition and proliferation of M2 macrophages, which contributes to further comprehension of the immunological features of keloids. Overall, we delineate the immunology landscape of keloids and present new insights into the mechanisms involved in its formation in this study.

Keywords: fibroblast; immunology profiles; keloid; macrophage; single-cell RNA sequencing.

Figure 1

Single-cell RNA-seq (scRNA-seq) reveals the cellular diversity and heterogeneity of keloid skin tissue. (A) Schematic representation of the experimental procedure. Keloids and adjacent normal skin tissues were harvested separately at the time of surgery (n=24). (B) Uniform Manifold Approximation and Projection (UMAP) of unbiased clustering of 427,171 cells reveals 11 cellular clusters. Clusters are distinguished by different colors. The number in the parenthesis indicates the cell count. (C) Hierarchical clustering of the clusters based on the average expression of the 2,000 most variable genes. (D) Comparison of the ratio of cells in keloids and adjacent normal skin tissues. (E) Representative molecular signatures for each cell cluster. SGCs, sweat gland cells; ECs, vascular endothelial cells; MPs, mononuclear phagocytes.

Figure 2

Fibroblasts of keloid and normal skin tissue subcluster into distinct cell populations. (A) Subclustering of keloid and normal tissue fibroblasts identified four distinct subtypes: mesenchymal (MF), pro-inflammatory (PF), secretory-papillary (SPF), and secretory-reticular (SRF). (B) Demonstration of the ratio of the four fibroblast subpopulations in keloids and adjacent normal tissue. (C) Color-coded UMAP plot is shown, and each fibroblast subcluster is defined. (D) Heatmap plots showing representative differentially expressed genes between the four fibroblasts in the keloids. (E) Gene ontology (GO) biological process enrichment analysis of differentially expressed genes in the four fibroblasts between keloid and normal scars. (F) GO enrichment analysis of differentially expressed genes in mesenchymal fibroblasts.

Figure 3

Characteristics of immunology profile in human keloids. (A) Color-coded UMAP plot is shown, and each immune cell subpopulation is defined as: macrophages, conventional type 1 dendritic cells (cDC1s), conventional type 2 dendritic cells (cDC2s), mature dendritic cells (MatureDCs), proliferating mononuclear phagocytes (ProliferatingMPs), plasmacytoid dendritic cells (pDCs), mast cells, neutrophils, and monocytes. (B, C) The proportions of immune cell subpopulations in keloid and adjacent normal tissue were shown. (D) Heatmap plots showing representative differentially expressed genes between the different types of immune cells in the keloids. (E, F) Demonstration of the ratio of M1 and M2 macrophage subpopulations in keloid and adjacent normal tissues. (G) Heatmap showing representative differentially expressed genes in M1 and M2 macrophage subpopulations. (H, I) Demonstration of representative modules of macrophages in keloid. (J) Pseudo-trajectory and cell source transition of macrophage cells between keloid and adjacent normal tissue.

Figure 4

T cells of keloid and normal skin tissues were subdivided into different cell subsets. (A) Six different subsets of T cells are identified in keloid and normal tissues: CD4+ naïve T cells (CD4NaiveT), CD4+ effector memory T cells (CD4Tem), CD8+ effector T cells (CD8Teff), CD8+ mucosa-associated invariant T cells (CD8MAIT), CD4+ regulatory T cells (Treg), and proliferating T cells. (B) The ratio of six T-cell subsets in keloid and adjacent normal tissues is shown. (C) Heatmap illustrating representative differentially expressed genes among the six T-cell subsets. (D) Augur analysis indicates that proliferative T cells are the cell type with the greatest variation among subsets. (E) Enrichment analysis of differentially expressed genes in the six T-cell subpopulations.

Figure 5

Typical images of immunohistochemical staining in the keloid (group K) and surrounding normal skin tissue (group N). Gross appearance: ×400.

Figure 6

Potential ligand–receptor interactions analysis in fibroblast subpopulations. (A) Heatmap showing the numbers of interpopulation communications with each other in keloid and adjacent normal tissue. (B) The ligand–receptor pairs exhibit significant changes in specificity between any one of the populations and any one type of fibroblast in adjacent normal tissue versus keloid. (C) Putative and relative signaling or transcriptional factors (TFs) within fibroblasts and other cell populations’ keloid and adjacent normal tissue.