Mammalian adult skin wounds typically heal by fibrotic repair without hair follicle (HF) regeneration. There is however, a great unmet medical need to regrow hair on healing skin in patients suffering from disfigurement caused by trauma, burns and other injuries. The hair follicle develops during embryogenesis from follicular epithelial cells directed by signals from the dermal papilla cells. Fibrosis and regeneration are currently considered the opposite end of wound healing process. It was assumed that, as part of the healing process, scarring and rapid collagen build up in damaged skin may be preferred to reactivation of embryonic development signaling.
However, in a recent study Ito et al., provided evidence that scar tissue can be stimulated to promote healing with HF regeneration via activation of a specific signaling pathway in the dermis, namely Sonic hedgehog (Shh) pathway.
In this study mice either had full thickness 1cm2 or 2.25cm2 wounds or small 4mm wounds. Gli1 expression was localized to the center of large wounds, in areas of hair placode/germ formation, but was completely absent from small wounds. Gli1 is indicative of Hedgehog (Hh) pathway activation.
The Gli1 signal was localized to both dermal papilla (DP) and epithelial hair germ cells, which concurs with its expression embryonic HF development. Shh was also upregulated at the site of HF neogenesis, but only in epidermal compartment of large wounds. It was not expressed in either epidermal or dermal compartments in small wounds.
To test the role of epithelial Shh in HF formation, authors genetically deleted Shh from epidermal cells employing Tamoxifen (TAM) inducible K14-CreER;Shh fl/fl mice. Healed wounds in those mice showed loss of DP and HF formation. Additionally deletion of Smoothened (Smo), an essential component of Shh pathway activation, in dermal cells of TAM-treated Pdgfra-CreER;Smo fl/fl mice also resulted in similar inhibition. These findings suggested activation of Shh signalling is vital for de novo hair follicle development.
Epidermal Shh overexpression in mouse skin resulted in extensive formation of DP and HF in small wounds with activation of Gli in the dermis and epidermis correlating with HF neogenesis. Large proportion of de novo HF grew downwards to form mature hair follicles with hair shafts, which was uncommon in small wounds from controls. New DP expressed Lef1, Noggin and AP, consistent with a DP identity.
In large wounds exogenous Shh expression also increased the area in which new hair follicles formed, whereas in WT untreated mice hair follicles only grew in the central region of large wounds.
Gene Ontology analysis of RNA-seq from the epidermal and dermal components of wounded skin in mice overexpressing Shh confirmed that genes involved in embryonic HF morphogenesis and DP formation both in the wound dermis and hair placode in the wound epidermis were enriched in comparison to controls. Upregulation of the Shh pathway was observed as well as upregulation of DP signature genes such as Bmp7, Enpp2, lamc3, and Trps1 in the wound dermis and Trp73, Vwa2, Samd5, Cxcl14, Nedd9, and Tnfaip3 in the epidermis.
A key characteristic of fetal scar-less wound healing is an increased ratio of type III versus type I collagen in skin. However, RNA-seq, collagen staining and transmission electron microscopy analysis of wounded skin overexpressing Shh showed that upregulation of the signaling pathway did not significantly change the overall extracellular matrix composition within the tissue. DP and HF were therefore forming within the scar tissue and the ratios of collagen types was not altered.
Expression of Smo from a promoter specifically active in dermal fibroblasts also resulted in DP formation in small wounds on mice. Epithelial cells proximal to the DP expressed genes consistent hair germ markers.
scRNA-Seq was used to functionally cluster the cell types in wound dermis. Expression of lineage markers demonstrated the presence of fibroblasts, schwann cells, immune cells, muscle cells and endothelial cells.
Expression of Hh pathway genes was observed in the fibroblast cluster of the Smo overexpressing but not WT. These fibroblast clusters also had upregulation of DP signature genes such as Hey1, Sema6a, Wif1, Cxcr4, Ggta1, Hck, Snrpn and Rasd1.
In contrast, activation of Hh pathway in epithelial cells only did not promote formation of DP in the wound area, there was also no signs of hair follicle differentiation. Epidermis from these cells was more reminiscent of basal cell carcinomas.
The scRNA-seq from myofibroblasts and dermis from small scarring wounds expressed Axin2, Wls, and canonical Wnt ligands such as Wnt2 and Wnt10α. Constitutive activation of Wnt signaling in small wounds was not sufficient to induce HF, however dermal depletion of Wls, which is essential for Wnt ligand secretion, inhibited HF formation in large wounds.
When Wnt-active fibroblasts were forced to overexpress SMO and therefore activate the Hh pathway this resulted in extensive DP formation in small wounds. In contrast control wounds without Hh activation in Wnt-active dermal cells underwent healing without HF. These results suggested that Hh activation in Wnt-active dermal cells promotes the regenerative dermal niche for HF formation.
In conclusion the studies demonstrate that the mechanisms of scarring and regeneration in skin are not as distant from one another as previously thought. Wound repair can be redirected to promote regeneration following injury by activating a key dermal Hh signaling pathway. This study is the first to show that de novo HFs can be regenerated in adult skin by modulating a Shh pathway in the dermis.