An important role of podoplanin in hair follicle growth

Podoplanin (PDPN) is a glycoprotein that is expressed by various cell types, including keratinocytes, fibroblasts, and lymphatic endothelial cells. We found that PDPN is expressed in the hair follicle (HF) keratinocyte region and HF stem cell area during the late anagen phase but not during the telogen phase in mice. Importantly, keratinocyte-specific PDPN deletion in mice (K5-Cre;PDPNflox/flox) promoted anagen HF growth after depilation-induced HF regeneration as compared to control mice. RNA sequencing, followed by gene ontology analysis, showed down-regulation of focal adhesion and extracellular matrix interaction pathways in HF stem cells isolated from K5-Cre;PDPNflox/flox mice as compared to control mice. Furthermore, HF keratinocytes isolated from K5-Cre;PDPNflox/flox mice exhibited a decreased ability to interact with collagen type I in cell adhesion assays. Taken together, these results show that PDPN deletion promotes HF cycling, possibly via reduced focal adhesion and concomitantly enhanced migration of HF stem cells towards the bulb region. They also indicate potential new therapeutic strategies for the treatment of conditions associated with hair loss.


Introduction
Podoplanin (PDPN) is a 38-kDa mucin-type transmembrane glycoprotein consisting of a heavily glycosylated extracellular domain, a single transmembrane domain, and a short, nine amino acid cytoplasmic tail [1,2]. Expression of PDPN is upregulated in various tumor types, including squamous cell carcinoma, angiosarcoma, hemangioblastoma, malignant mesothelioma, and brain tumors [3][4][5], and correlates with increased tumor cell motility and metastasis [6]. While PDPN expression on cancer-associated fibroblasts (CAFs) of human lung adenocarcinomas is associated with poor prognosis [7][8][9], increased PDPN expression in human colorectal CAFs was a significant indicator of good prognosis [10]. Although not detected in the normal interfollicular epidermis, PDPN is expressed in epidermal keratinocytes during wound healing, in psoriasis and during mouse skin carcinogenesis [11][12][13]. PDPN is also expressed in the basal cell layer of the sebaceous gland and the outer root sheath cells of hair follicle (HF) a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 keratinocytes [14]. The HF is a mini-organ of the skin that continuously cycles through rapid growth (anagen phase), apoptosis-driven regression (catagen phase) and relative quiescence (telogen phase) [15]. HF stem cells, which reside within the bulge region of the HF, promote the repetitive regeneration of the follicle during the hair cycle [16]. The role of PDPN in the regulation of HF cycling has remained unknown. In this study, we first characterized PDPN expression throughout the depilation-induced hair cycle in mice. We next investigated whether keratinocyte-specific PDPN deletion in mice (K5-Cre;PDPN flox/flox mice) might have an effect on HF growth. To identify the molecular and cellular mechanisms, we also performed RNA sequencing analysis of HF stem cells isolated from K5-Cre;PDPN flox/flox and control mice. Our results reveal that PDPN regulates HF cyling.

Ethics statement
This study was approved by the Institutional Review Board at the Seoul National University Hospital (approval number C-1203-031-400), and all subjects provided written informed consents. All experimental procedures using human materials were conducted according to the principles described in the Declaration of Helsinki.

Mouse models
To generate K5-Cre;PDPN flox/flox knockout mice on the C57BL/6 background, we crossed keratin 5 (K5)-Cre-ERT2 mice obtained from the MMRRC repository (University of Missouri, USA) [17] with PDPN floxed (PDPN flox/flox ) mice (Ozgene Pty Ltd, Perth Australia). PDPN flox/flox mice were designed with the loxP sites flanking exon 2 and were used as control mice. To investigate the effects of PDPN on HF keratinocytes during the hair cycle after depilation-induced hair regeneration, the back skin of 8-week-old female K5-Cre;PDPN flox/flox mice in the telogen phase was depilated using wax as described [15,18], resulting in the synchronized induction of new anagen follicle growth. The number of mice used for the experiment is indicated in the figure legends. Mice were sacrificed with an overdose of anaesthesia (160 mg kg −1 ketamine; 0.4 mg kg −1 medetomidine) at days 2, 5, 8, 19, and 22 after depilation, and back skin was taken for histological analysis. To measure the bulb diameter, 3 images/mouse were acquired of hematoxylin and eosin (H&E)-stained paraffin sections and the bulb diameter was measured at the level of the largest diameter ("Auber's line") of the hair bulbs with a clearly visible dermal papilla (DP) [19]. For quantitative analysis, the Image J software (National Institutes of Health, Bethesda, Maryland, USA) was used. All experimental procedures were conducted according to animal protocols approved by the Kantonales Veterinaeramt Zuerich.

Isolation of HF stem cells from mouse back skin
At day 12 (late-anagen phase) after depilation, the back skin of control and K5-Cre; PDPN flox/flox mice (n = 5 each) was dissected, minced using scissors, and digested in Dulbecco's Modified Eagle Medium supplemented with 2% fetal bovine serum (FBS), 1.

RNA sequencing
Total RNA was isolated from the sorted cells using the RNeasy Plus Micro Kit (Qiagen, Hilden, Germany) and RNA quality was assessed using bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). For the preparation of sequencing libraries, RNA was reverse transcribed to double-stranded cDNA and then amplified using the NuGEN Ovation RNA-Seq System according to the manufacturer's instructions. The quality of libraries was assessed with the High Sensitivity D1000 ScreenTape system (Agilent). For sequencing, Illumina HiSeq 2500 v4 was used to generate paired-end reads of 126 nt length. For data processing, the raw reads were first cleaned by removing adapter sequences, trimming low quality ends, and filtering reads with low quality (phred quality <20) using Trimmomatic [20]. Sequence alignment of the resulting high-quality reads to the Mus musculus reference genome (build GRCm38) was carried out using STAR (Version 2.5.1b) [21]. Gene expression values were computed with the function featureCounts from the Bioconductor package Rsubread [22]. Differential expression analysis was performed using the generalized linear model implemented in the Bioconductor package EdgeR [23]. Differential expression was assessed using an exact test adapted for overdispersed data. Genes showing altered expression with adjusted (Benjamini and Hochberg method) p-value < 0.05 were considered differentially expressed.

Cell adhesion assays
HF keratinocytes isolated from K5-Cre;PDPN flox/flox and control mice were labelled with 6 μM calcein for 10 min at 37˚C and then washed twice in PBS. Calcein-labelled keratinocytes (10 4 cells/well) were seeded on 25 μg/ml type I collagen (PureCol, Advanced BioMatrix, San Diego, CA, USA) or 0.1% bovine serum albumin (negative control)-coated 96-well plates (black, clear bottom, Corning, NY, USA) and incubated for 1 h at 37˚C. After washing twice with PBS, 100 μl PBS was added to each sample and the measurement of fluorescence was performed using a Spectramax reader (excitation: 485 nm and emission: 538 nm, Molecular Devices).

Scratch wound healing assays
The methods used for a scratch wound healing closure assay have been described previously [25,26]. Mouse HF keratinocytes isolated from K5-Cre;PDPN flox/flox and control mice (at postnatal day 4) were grown to full confluence in 25 μg/ml type IV collagen (collagen from human placenta, Sigma-Aldrich)-coated 24-well plates and were then incubated overnight in serum-reduced medium containing 1% FBS. Cells were scratched with a 200 μL sterile pipette tip and were then incubated. After 48 h, cells were washed with PBS and the images of the scratches were acquired. The surface areas of the cell-free zones were measured and the % scratch closure was determined using TScratch software [26].

Quantitative real time-polymerase chain reaction (qRT-PCR)
Total RNA was isolated using the NucleoSpin RNA kit (Macherey-Nagel, Düren, Germany) and 1 μg of total RNA was used for the cDNA synthesis using the High Capacity Reverse Transcription kit (Applied Biosystems, Foster City, CA, USA). PCR was performed on a 7900HT Fast Real-Time PCR System (Applied Biosystems) using FastStart SYBR green master mix (Roche Diagnostics, Basel, Switzerland) according to the manufacturer's instructions. Gene expression was normalized to the control gene GAPDH. Primer information is provided in S1 Table.

Statistical analyses
Statistical significance was determined using a two-tailed unpaired t-test (GraphPad Prism, version 5.0a, San Diego, CA, USA) and results were considered significant at a p-value < 0.05.

PDPN is expressed in the HF keratinocyte region and HF stem cell area during the late anagen but not the telogen phase
To characterize the location of PDPN in back skin samples of wild-type mice in the anagen phase, we performed double immunofluorescence stainings for the lymphatic vessel (LV) marker LYVE-1 and for PDPN. PDPN was co-expressed by LYVE-1-positive LVs and was especially located in anagen HFs, in particular in outer root sheath cells and the HF stem cell area (Fig 1A). To further characterize PDPN in the HF, back skin samples were stained for keratin 15 and CD34, which are putative markers of HF stem cells. To isolate HF stem cells from mouse back skin using fluorescence-activated cell sorting (FACS), antibodies for CD34 and integrin α6, putative markers of HF stem cells, were used. PDPN-positive cells were observed in HF stem cells (CD45-, CD31-, CD34+, integrin α6+ cells) (Fig 1B). PDPN was also detected in LVs, HF keratinocytes and HF stem cells of human scalp tissue (Fig 1C). These results indicate that PDPN is expressed in HFs, particularly in the HF keratinocyte region and HF stem cell area.
We next investigated by double immunofluorescence staining for PDPN and K15 whether PDPN expression might undergo cyclic changes during depilation-induced hair regeneration in female C57BL/6J mice. At day 1 (early-anagen phase) and day 5 (mid-anagen phase) after depilation, PDPN was expressed in LVs but was absent from HF keratinocytes (Fig 1D and  1E). Interestingly, at days 8 and 12 (late-anagen phase), PDPN was expressed in HF keratinocytes and HF stem cells (Fig 1F and 1G). At day 18 (catagen phase), PDPN expression was still present in HF keratinocytes (Fig 1H and 1D). However, at day 22 (telogen phase), PDPN was detected in LVs but not HFs (Fig 1I). These results indicate that PDPN is expressed in HF keratinocytes and HF stem cells during the late-anagen growth phase but not during the telogen phase, suggesting that PDPN might be involved in HF cycling.

K5-Cre;PDPN flox/flox mice show enhanced anagen growth
Given the cyclic changes of PDPN expression during the hair cycle, we next investigated whether keratinocyte-specific PDPN deletion in mice might have an effect on HF growth. We first studied depilation-induced hair regeneration in keratinocyte-specific PDPN deleted mice (K5-Cre;PDPN flox/flox ). Double immunofluorescence stainings for PDPN and K15 confirmed the absence of PDPN expression in HF keratinocytes of K5-Cre;PDPN flox/flox mice, whereas PDPN was located in HF keratinocytes of littermate PDPN flox/flox control mice (Fig 2A). Since the thickness of the hair shaft is known to correlate with the size of the hair bulb [27], we measured the diameter of the hair bulbs during depilation-induced hair regeneration, which typically increases during the anagen phase, whereas it is reduced during catagen development [15]. At day 2 (early-anagen phase) after depilation, the diameter of the hair bulb was comparable in both control and K5-Cre;PDPN flox/flox mice (Fig 2B and 2C), but started to become larger in K5-Cre;PDPN flox/flox mice than in control mice at day 5 (mid-anagen phase) (Fig 2B  and 2D). Importantly, at days 8 (mid-anagen phase) and 19 (catagen phase), the thickness of hair bulbs was significantly larger in K5-Cre;PDPN flox/flox mice than in control mice (Fig 2B,  2E and 2F). At day 22 (telogen phase), the hair bulb diameter was comparable in both control and K5-Cre;PDPN flox/flox mice. Together, these results indicate that after depilation, K5-Cre; PDPN flox/flox mice have increased anagen HF growth compared to control mice.

Down-regulated focal adhesion in HF stem cells isolated from K5-Cre; PDPN flox/flox mice
To identify potential molecular and cellular mechanisms by which PDPN deletion might promote anagen HF growth, we performed RNA sequencing of HF stem cells isolated from control and K5-Cre;PDPN flox/flox mice at day 12 (late-anagen growth phase) after depilation using FACS. The RNA sequencing data were uploaded in the European Nucleotide Archive (ENA) under accession number PRJEB22837 (For reviewers access: https://www.ebi.ac.uk/ena/ submit/sra/#home username "Webin-47976" and password "781228.kim"). As expected, PDPN expression was markedly decreased in HF stem cells isolated from K5-Cre;PDPN flox/flox mice as compared to those from control mice using FACS (Fig 3A) and RNA sequencing ( Fig  3B). Moreover, at day 12 after depilation, there were more HF stem cells in K5-Cre;PDPN flox/ flox mice compared to control mice (Fig 3C). Gene ontology analysis showed down-regulation of focal adhesion and extracellular matrix interaction pathways in HF stem cells isolated from K5-Cre;PDPN flox/flox mice ( Fig 3D). As cellular migration is associated with increased cell adhesion in epithelial cells [28,29], these data suggest a potentially increased capacity of HF stem cells to migrate towards the bulb area. Down-regulated genes included several collagens, TNXB, LAMA4, and ITGB3 (Fig 3E). To confirm our RNA sequencing data,  immunofluorescence stainings for tenascin XB (TNXB), collagen type I alpha 1 chain (COL1A1), and integrin beta 3 (ITGB3) were performed. We found that the expression of TNXB, COL1A1 and ITGB3 were weaker in K5-Cre;PDPN flox/flox mice than in control mice (S1 Fig). Furthermore, HF keratinocytes isolated from K5-Cre;PDPN flox/flox mice exhibited a decreased ability to interact with collagen type I in cell adhesion assays (Fig 3F). In addition, we also performed a scratch wound healing closure assay to assess cell migration. We found that cell migration was accelerated in HF keratinocytes isolated from keratinocyte-specific podoplanin deletion mice as compared to those from control mice (Fig 3G). Quantitative realtime (qRT)-PCR analysis confirmed that PDPN expression was strongly decreased in HF keratinocytes isolated from K5-Cre;PDPN flox/flox mice (Fig 3H). Importantly, the mRNA expression of integrin α2 (ITGA2), a major cellular receptor for collagen type I, was significantly reduced in HF keratinocytes isolated from K5-Cre;PDPN flox/flox mice (Fig 3H). These results indicate that PDPN deletion might promote hair growth, possibly via reduced focal adhesion and concomitantly enhanced migration of HF stem cells towards the bulb region.

Discussion
While PDPN has a variety of functions including regulation of organ development, cell mobility, tumorigenesis, and metastasis, its physiological cell type-specific functions are still incompletely understood. In this study, we identified PDPN as a new player in the regulation of HF cycling.
Previously, it has been found that PDPN is associated with the down-regulation of the cellcell adhesion protein E-cadherin in oral squamous cell carcinomas [2] and that it induces an epithelial-mesenchymal transition in Madin-Darby canine kidney type-II epithelial cells and immortalized HaCaT keratinocytes through the interaction with ERM proteins and up-regulation of RhoA activity [6]. PDPN knockdown resulted in decreased migration of human lung microvascular lymphatic endothelial cells and contributed to low RhoA-GTP levels in the scratch wound assay [30]. Furthermore, bone-specific PDPN deletion in osteocytes in mice resulted in the disruption of the osteocyte dendritic network [31]. This, PDPN has a variety of functions depending on the cell type in which it is expressed. In the present study, we found that PDPN is expressed in the HF keratinocyte region and HF stem cell area during late-anagen growth phase but not during telogen quiescence phase, indicating that PDPN might be involved in HF cycling.
It has been reported that during the telogen to anagen transition, HF stem cells are activated by dermal papilla cells and the cutaneous microenvironment, and migrate from the bulge towards the bulb region where they differentiate into various epithelial HF cell lineages [32,33]. Enhanced focal adhesion has been shown to associate with reduced cell migration in focal adhesion kinase-deficient mice [34,35]. Interestingly, using RNA sequencing of HF stem cells isolated from K5-Cre;PDPN flox/flox mice, we found that the focal adhesion pathway was down-regulated. In agreement with these results, HF keratinocytes isolated from K5-Cre;PDPN flox/flox mice had a decreased ability to interact with collagen type I in a cell adhesion assay. Moreover, we found that cell migration was accelerated in HF keratinocytes isolated from keratinocyte-specific podoplanin knockout mice as compared to those from control mice. Our results indicate that PDPN deletion in HF stem cells results in enhanced hair growth, possibly via reduced focal adhesion and concomitantly enhanced migration of hair follicle stem cells toward the bulb region.
Using RNA sequencing, we found that transcription factors such as Stat3 and PPARγ might contribute to the regulation of several down-regulated genes in HF stem cells isolated from K5-Cre;PDPN flox/flox mice. It has been shown previously that keratinocyte-specific Stat3-disrupted mice show an impairment of anagen entry during the second hair cycle and subsequent hair growth [36]. Moreover, HF stem cell-specific PPARγ deletion in mice resulted in scarring alopecia [37]. Thus, further studies would be of interest to investigate the roles of Stat3 and PPARγ in K5-Cre;PDPN flox/flox mice.
In conclusion, our results suggest an unanticipated role of PDPN in the HF cycle, with potential implications for therapeutic strategies to treat alopecia.  Table. (TIF) S1 Checklist. (PDF)