Platelet-derived growth factor signaling modulates adult hair follicle dermal stem cell maintenance and self-renewal

Hair follicle regeneration is dependent on reciprocal signaling between epithelial cells and underlying mesenchymal cells within the dermal papilla. Hair follicle dermal stem cells reside within the hair follicle mesenchyme, self-renew in vivo, and function to repopulate the dermal papilla and regenerate the connective tissue sheath with each hair cycle. The identity and temporal pattern of signals that regulate hair follicle dermal stem cell function are not known. Here, we show that platelet-derived growth factor signaling is crucial for hair follicle dermal stem cell function and platelet-derived growth factor deficiency results in a progressive depletion of the hair follicle dermal stem cell pool and their progeny. Using αSMACreER T2 :Rosa YFP :Pdgfrα flox mice, we ablated Pdgfrα specifically within the adult hair follicle dermal stem cell lineage. This led to significant loss of hair follicle dermal stem cell progeny in connective tissue sheath and dermal papilla of individual follicles, and a progressive reduction in total number of anagen hair follicles containing YFP+ve cells. As well, over successive hair cycles, fewer hair follicle dermal stem cells were retained within each telogen hair follicle suggesting an impact on hair follicle dermal stem cell self-renewal. To further assess this, we grew prospectively isolated hair follicle dermal stem cells (Sox2GFP+ve αSMAdsRed+ve) in the presence or absence of platelet-derived growth factor ligands. Platelet-derived growth factor-BB enhanced proliferation, increased the frequency of Sox2+ve hair follicle dermal stem cell progeny and improved inductive capacity of hair follicle dermal stem cells in an ex vivo hair follicle formation assay. Similar effects on proliferation were observed in adult human SKPs. Our findings impart novel insights into the signals that comprise the adult hair follicle dermal stem cell niche and suggest that platelet-derived growth factor signaling promotes self renewal, is essential to maintain the hair follicle dermal stem cell pool and ultimately their regenerative capacity within the hair follicle.

of the hair follicle .The backs of the animals were depilated at P21 (telogen) and tamoxifen was administered by gavage at P21, P23 and P25 (1mg/day/animal). The subsequent anagen phase was documented and compared between tamoxifen-treated wild type (n = 3), heterozygous (n = 4) or Pdgfrα flox/flox (n = 3), carrying Cre, animals. For these, skin was harvested during midanagen (P33-P35). In another set of animals, the next two consecutive-induced hair cycles were evaluated and compared between tamoxifen-treated wild type (n = 7), heterozygous (n = 5) or Pdgfrα flox/flox (n = 5), carrying Cre, animals. A second depilation was performed at P50 (telogen) and dorsal skin was harvested at P60-62 (mid-anagen) for immunostaining. Based on the results observed from the mentioned collection time-points, we also collected skin from animals that entered 2 nd adult telogen (harvested around P45-P49) from tamoxifen-treated wild type (n = 2) or Pdgfrα flox/flox (n = 2), Cre +ve animals. (B) Example of genotyping before tamoxifen administration using multiplex PCR with primer 1, primer 2 and primer 3 (see map). Lanes 1 and 4 depict wild-type genotype; lanes 2-3 and 5-6 represent heterozygous animals (Pdgfrα +/flox ) and   To further confirm appropriate excision of the targeted sequence, DNA was purified from the electrophoresis gel using a gel PCR-product DNA purification kit. A fragment of approximately 580bp was obtained after sequencing, containing a sequence upstream of exon 1, followed by part of the original targeting vector (NCBI GQ424832.1) (ref. 3), which contained the loxP site, followed by a fragment downstream of exon 4, confirming successful recombination and deletion of exons 1 to 4 from the Pdgfrα gene (see figure S1).

Ex vivo hair follicle formation assay
We used the "patch" hair follicle formation assay 4,5 to assess the impact of PDGF on the ability of isolated hfDSCs to stimulate new hair follicle formation. Rat SKPs grown in the presence or absence of PDGF-AA and -BB were combined with epithelial aggregates extracted from backskin of newborn C57Bl/6 mice (P0; Charles River Laboratories). Dorsal skin was floated on 1mg/ml dispase for 30 min at 37°C in order to separate the dermis from the epidermis. The 6 dermis was discarded and the epidermis was transferred to a plate containing 0.25% trypsin (without EDTA) for 3-4 min at 37°C. To inactivate the enzyme, 10% FBS was added and then the epidermal sheet was gently scraped with a scalpel blade. Liberated epithelial cells were transferred to a 15 ml tube and centrifuged to 300g for 3 minutes and resuspended in DMEM.
For each graft 10,000 epithelial aggregates were combined with 500,000 SKPs and subcutaneously injected under the back skin of adult male nude mice (Charles River Laboratories). After 14 days, grafts were harvested and the number of HFs containing GFP +ve DPs within each graft were quantified (n = 3, with 4-6 grafts per treatment).

Immunofluorescence staining
To verify the location of PDGFRα positive cells within the adult dermal stem cell niche we examined anagen (P8, P28) or telogen (P56) backskin samples from PDGFRαH2BGFP mice (Jackson Laboratories). PDGFRαH2BGFP mice were created through knock-in of the histone H2Be-GFP fusion protein into the PDGFRα locus to allow for nuclear expression of GFP in cells that express PDGFRα 6 . Epithelial keratinocytes were labelled using keratin-5 (1:500; Covance).
Cell death and cell proliferation was done with anti-cleaved caspase-3 (1:500, Millipore) and anti-Ki67 (Millipore). Dorsal rat skin or human skin samples were cryosectioned into 20 µm 7 slices, and mouse skin into 55 µm for lineage tracing experiments. To verify that adult human SKPs preserved their phenotype after exposure to the PDGF-B ligand, single cells were cytospun (Shandon) on slides at 800 rpm for 3 min and subsequently fixed in cold 4% PFA for 5 min. Cells were incubated with 10% normal goat serum containing 0.05% TritonX-100 for 1 hr and then exposed to primary antibodies against versican, αSMA, fibronectin and FSP1 overnight at 4°C. Tissue sections and cells were visualized with Alexa-488, -555 or -647 conjugated secondary antibodies (Life Technologies). Nuclei were labeled with Hoechst (1 µg/ml, Sigma Aldrich). Images were obtained using an Axioplan Observer microscope (Zeiss) or TCS SP8 spectral confocal microscope (Leica).

Real-time quantitative RT-PCR
mRNA was isolated from FACS-sorted cell populations using αSMA dsRed :Sox2 GFP knock-in mice.
RNA (27.6µg) was reverse transcribed using a High Capacity cDNA synthesis kit (Applied Biosystems). cDNA samples were amplified for 14 cycles using TaqMan® PreAmp Master Mix Kit (Applied Biosystems) and Taqman probes for Pdgrfα, Pdgfrß and the endogenous control, Hprt (Life Technologies, in either FAM or VIC formats). Preamplification was tested by comparing the ∆∆CT for each TaqMan® probe from a non-limiting sample of unsorted dermal cells amplified to the same sample that did not undergo amplification. qPCR was performed using TaqMan® Fast Advanced Master Mix and samples were run using 7500 Fast Real Time PCR system (Applied Biosystems). Fold-change in gene expression was calculated using the ∆∆CT method with Hprt serving as housekeeping gene. Cells from each hair follicle mesenchymal compartment (dermal sheath, dermal papilla and dermal cup) were compared to fibroblasts from interfollicular dermis. Probe details are shown below: Hprt: ID: Mm03024075_m1 Chr.X: 52988078 -53021660 on Build GRCm38