Understanding the role of mesenchymal cell populations in human skin in health and disease

Abstract

Keloids are benign overgrowths of fibrotic scar tissue that can occur after an injury to the skin. They are characterised by excessive fibroblast proliferation, extracellular matrix (ECM) formation and invasion into the adjacent normal skin. Mesenchymal cells, in particular fibroblasts, are thought to be the primary cellular mediators in keloid scar formation. However, mesenchymal cells are highly heterogeneous, and to date, the distinct mesenchymal cell subsets involved in keloid scarring remain to be determined. Therefore, we sought to identify and characterise different mesenchymal cell populations both in healthy human dermis and in keloid pathogenesis. A 15-colour flow cytometry panel was developed to analyse the mesenchymal cell populations directly from ex vivo human skin and keloid tissue. We identified a novel population of CD26+ FAP+ fibroblasts present in both normal skin and in keloid scars, as well as a small population of CD271+ mesenchymal cells, which is previously undescribed in keloid lesions. Multicolour immunofluorescence microscopy revealed that CD26+ FAP+ fibroblasts are the main cell population within keloid scars, while the CD271+ mesenchymal cell population is located in the perivascular space surrounding CD146+ pericytes. RNA sequencing on uncultured FACS sorted CD26+ FAP+ fibroblasts from healthy human skin and keloid scar has revealed substantial differences in the gene expression profiles between the healthy and pathological state of CD26+ FAP+ fibroblasts including striking upregulation of ECM genes, matrix metalloproteinases, and altered expression of Wnt signalling pathway components in keloid scarring. RNA sequencing of isolated CD271+ cells from keloid scar exposed expression of genes typical of perivascular cells, such as the genes for the vascular basement membrane, as well as chemokines and chemokine receptors associated with angiogenesis and cell recruitment. While this study revealed significant differences between uncultured fibroblasts, we also provide evidence for the striking similarity of cultured keloid and healthy dermal fibroblast lines. These fibroblasts lines have almost identical cell surface marker expression and have highly similar gene expression profiles. Furthermore, functional in vitro assays revealed that most keloid characteristics were lost in culture and only the increased invasiveness of keloid fibroblasts remained in vitro. Finally, we identified the transcription factor TWIST as a potential new mediator in keloid pathogenesis by multicolour immunofluorescence microscopy. Furthermore, we have shown elevated expression of the TWIST downstream target BMI1 as well as the potential TWIST upstream inducer β-catenin, which indicates a potential role of a TWIST pathway in keloid scarring. In conclusion, this is the first study to isolate and investigate mesenchymal cell subsets from fresh human skin and keloid tissue in contrast to previous work using whole tissue or cultured fibroblasts. This enabled the identification of a novel CD26+ FAP+ fibroblast population as a major cellular contributor to the excessive ECM deposition in keloid scarring and a CD271+ perivascular cell population in keloid scarring. Furthermore, this work also highlights the limitations of previous in vitro studies of keloid scarring by revealing the striking similarities of cultured cell lines. Overall, insight into the phenotypes, localisation and functions of mesenchymal cell populations in healthy human dermis and in keloid scarring may ultimately lead to more targeted therapeutic strategies for keloid scarring.