β-catenin-mediated hair growth induction effect of 3,4,5-tri-O-caffeoylquinic acid

3,4,5-tri-O-caffeoylquinic acid (TCQA) promoted hair regrowth in C3H mice

The ability of TCQA to promote hair regrowth was tested in eight-weeks-old male C3H mice. Ten days after shaving, TCQA-treated group exhibited hair regrowth in the shaved area (Figure 1A). TCQA significantly enhanced the hair growth by approximately 40%, 80%, and 120% at day 14, 20, and 30, respectively, compared with the control (Figure 1B). By the end of the treatment period (day 30), the treated group displayed a markedly complete regrowth of the hair in the clipped area. In contrast, for the control group, only 37% hair regrowth area was observed (Figure 1A and 1B). The direction of the growth was random in the treated and untreated mice. These results showed that TCQA stimulated the hair growth cycle. Minoxidil, a widely used drug against alopecia often used as a positive control ,induces hair regrowth in C3H mice within 28 days [31]. Our results showed that TCQA had the same effect as minoxidil in stimulating hair growth cycle. Interestingly, the hair shaft was fully developed and the epidermis and the hair follicle (HF) displayed no sign of inflammation or irritation following TCQA application (Figure 1C). The anagen phase is divided into six sub-stages that goes from I to VI and starting the third phase the HFs development can be distinguished. As shown in Figure 1C (indicated by arrows), the HF from the TCQA-treated mice were at advanced stage of anagen phase. The results indicated that TCQA induced the transition of the hair cycle from telogen to anagen phase accelerating hair growth in vivo.

Gene expression changes in C3H mice skin caused by 3,4,5-tri-O-caffeoylquinic acid (TCQA)

Microarray analysis was conducted from skin collected from the treated area with TCQA and milli-Q water to elucidate the mechanism behind the observed hair growth promoting effect of TCQA. The expression of 1235 genes was modulated by TCQA out of which, 435 were upregulated while 800 downregulated. Genes relevant in β-catenin binding, epidermal growth factor, transcription regulation, and Wnt signaling were upregulated (Figure 2A). On the other hand, genes involved in glycosylation, protein binding, and β-catenin degradation complex were downregulated (Figure 2B). A volcano plot was created to compare the gene expression in the control group vs TCQA group (Figure 2C). TCQA significantly upregulated 435 genes (2-fold change; red) and downregulated 800 genes (2-fold change; green), while the gray color represents the other genes ($\le$2-fold change) (Figure 2C).

Genes with 2-fold change in expression (control vs TCQA) were subjected to hierarchical clustering that generated five clusters. In the first cluster (enrichment Score: 1.53), TCQA regulated genes including Wnt2b, Dock2, Rasgrp2, Agbl4, and Dgki that are relevant for protein binding (P=0.0048), Ras signaling pathway (P=0.023), and intracellular signal transduction (P=0.017), and microtubule (P=0.0018) (Figure 2D). Moreover, a regulation of genes including, Ctnnb1, Wls, Rassf1, and Map3k12 that plays a role in ATP binding (P=0.023), kinase activity (P=0.0152), Wnt signaling (P=0.02152), and phosphorylation (P=0.025) was observed in cluster two (enrichment score: 0.83). The third cluster with enrichment score of 0.29 is composed of genes that are significant for cell adhesion (P=0.0049), cAMP signaling pathway (P=0.02), and neuroactive ligand-receptor interaction (P=0.0049). The last two clusters with an enrichment score of 0.89 and 0.55, respectively, represent the genes that were downregulated upon treatment with TCQA. Those genes are significant for metal binding (P=0.0045), Wnt signaling repression (P=0.0015), and cell surface receptor linked signal transduction (P=0.0175) (Figure 2D).

Tables 1, 2, 3, and 4 summarize the significantly modulated genes and their biological activities. An upregulation in the expression of canonical Wnt-associated genes, Ctnnb1, Wls,Wnt2b, and Wnt4 was observed (Table 1). Notch, FGF, and Rac/Ras pathway-related genes were upregulated as well. Genes significant for keratinocytes differentiation, including Foxn1, Rps3 were upregulated. In addition, the expression of genes involved in cell differentiation, cell cycle, ATP binding, and oxidation-reduction process like Sapcd2, Smchd1, and Txnrd1, were enhanced by TCQA (Table 2). Genes associated with telogen phase, repression of Wnt signaling, β-catenin degradation, and aging (Ldb3, Ak1, Gsk3b, Tcf3, Soct, and Nfatc1), were downregulated (Tables 3 and 4).

3,4,5-tri-O-caffeoylquinic acid (TCQA) stimulated β-catenin expression in vivo

TCQA-treated mice were observed to have an accelerated anagen phase and at the same time Ctnnb1 gene expression was observed to be upregulated based on microarray analyses results, the effect of TCQA on β-catenin expression was further determined in mice skin tissue. Results revealed that β-catenin expression in TCQA-treated mice skin was increased in the HF, in the area where the dermal papilla (DP) cells are, in the root sheath, and in the bulb area (Figure 3A). In case of the control mice, β-catenin expression was located in the epidermis and the upper part of the dermis (Figure 3A). In addition, the gene expression of Ctnnb1 in treated skin tissues was enhanced up to 2.3-fold compared with the control (Figure 3B). This upregulation of Ctnnb1 expression was followed by an increase in β-catenin protein expression level as shown in Figures 3C and 3D.

Figure 3E illustrates the summary of the modulated genes by TCQA. β-catenin target genes that are involved in HF development and keratinocyte differentiation including, Foxn1, Dlx, Fgf, and others, were upregulated. In contrast, genes that inhibit Wnt/β-catenin signaling including Gsk3b,Tcf3, and Igfbp4 were downregulated (Figure 3E). These results indicated that the observed hair regrowth effect appears to be caused by increased β-catenin expression following TCQA treatment.

3,4,5-tri-O-caffeoylquinic acid (TCQA) increased the proliferation of hair bulb cells

To determine the cytotoxic effect of TCQA, if any, on the cells that reside in the bulb area of the HF, a proliferation assay (MTT assay) of human epidermal melanocytes (HEM) and human hair follicle dermal papilla cells (HFDPC) was conducted. The cells were treated with 0, 5, 10, 15, and 25 µM TCQA for 48 and 72 h. Results showed that TCQA was not cytotoxic to the cells at all concentrations and in fact, the proliferation of the cells was stimulated (Figure 4A and 4B).

3,4,5-tri-O-caffeoylquinic acid (TCQA) increased the ATP content of human hair follicle dermal papilla cells (HFDPCs)

The number of hair matrix and the size of the HF is determined by DP cells. 10 µM of TCQA significantly enhanced the ATP content of DP cells by around 60% after 48 h treatment (Figure 4C). This result was supported by the microarray analyses that showed an upregulation in genes related to ATP binding.

3,4,5-tri-O-caffeoylquinic acid (TCQA) induced alkaline phosphatase (ALPL) and β-catenin expression in human hair follicle dermal papilla cells (HFDPCs)

Alkaline phosphatase (ALP) is known to be a marker of DP cells. As microarray results showed an upregulation in its gene expression, the validation using DP cells was carried out. Results showed an upregulation in ALPL expression after 6 and 12 h treatment (Figure 4D) confirming that TCQA stimulated the proliferation of DP cells.

The absence of β-catenin in DP cells results to an early entry to catagen phase [32]. To confirm the effect of TCQA on β-catenin stimulation, its protein and gene expression levels in HFDPCs were evaluated. β-catenin was observed to be stimulated by 10 µM TCQA after 12 h treatment, and this expression level was maintained until after 24 h (Figure 5A and 5B). Moreover, according to the immunocytochemistry results, TCQA caused the accumulation and nuclear translocation of β-catenin (Figure 5C). The expression of β-catenin gene (CTNNB1) was significantly upregulated after 6 and 12 h treatment with 10 µM TCQA (Figure 5D) suggesting that TCQA promoted hair growth by activating β-catenin expression in the DP at the onset or early stage of anagen phase.

3,4,5-tri-O-caffeoylquinic acid (TCQA) upregulated β-catenin expression in human hair follicle dermal papilla cells (HFDPCs) after inhibition with XAV939

XAV939 inhibits the activation of β-catenin stimulating its phosphorylation and non-translocation to the nucleus [33]. Here, HFDPCs were treated with various concentration of XAV939 for 48 h and 10 µM XAV939 was used for further experiment as it didn’t show any cytotoxicity effect on the cells (Figure 5E). After that, the cells were treated with 10 µM XAV939 for 6 and 12 h and results showed that CTNNB1 expression was significantly decreased upon treatment confirming the inhibitory effect of XAV939 on Wnt/ β-catenin signaling (Figure 5F). The cells were then treated with 10 µM XAV939 (6 and 12 h) and then with 10 µM TCQA (6 and 12 h). Results showed that treatment with TCQA significantly upregulated CTNNB1 expression suppressing XAV939 inhibition. Finally, after co-treatment with XAV939 and TCQA, the gene expression of β-catenin was maintained higher than after treatment with the inhibitor only. These results further proved the activation of β-catenin by TCQA in DP promoting hair growth cycle.

Sample preparation

Synthesized 3,4,5-tri-O-caffeoylquinic acid (TCQA) with 97% purity was provided by Dr. Kozo Sato from Synthetic Organic Chemistry Laboratories, the FUJIFILM Corporation (Kanagawa, Japan). For the in vivo experiment, TCQA was prepared by first dissolving in a small volume of 70% ethanol and then diluted in purified water. For the in vitro assay, a highly concentration of TCQA was dissolved in 70% ethanol and then diluted in cell culture medium.

Animal experiment

Eight-weeks-old male C3H mice were purchased from Charles River Laboratories, Japan Inc. (Kanagawa, Japan) and housed individually in cages at Gene Research Center of the University of Tsukuba. After one week of acclimatization, the mice were randomly assigned to experimental groups (n=5 for each group) TCQA-treated group and milli-Q water treated group (control group). The mice were anesthetized using isoflurane (Wako Pure Chemical Industries, Tokyo, Japan), then the telogen hair on the dorsal part of the mice back were shaved using hair clipper in order to induce the anagen phase. Each mouse then received a topical application of 1% TCQA or water daily for four weeks. The dorsal part of the shaved area was observed and photographed daily starting on the 14^th day after treatment. At the end of the treatment period, the mice were sacrificed by cervical spine dislocation and the skin collected from the treated area were divided into four parts and washed with phosphate-buffered saline (PBS), then immediately immersed in liquid nitrogen and kept at −80 °C. The ImageJ processing program (National Institutes of Health, Bethesda, USA) was used to measure the hair regrowth in the treated area. The experiment was approved by the Animal Study Committee of the University of Tsukuba (No.17-060) and the procedures were handled according to the guidelines of the Care and Use and Use of Animals approved by the Council of the Physiological Society of Japan.

DNA microarray

RNA used as a template were extracted from skin tissues collected from treated area of mice dorsal part using ISOGEN solution (Nippon Gene, Tokyo, Japan) following the manufacturer’s instructions. The total RNA extracted were then amplified and labeled. The labeled fragmented aRNA was then hybridized to the Affymetrix mouse Array strips (Affymetrix, Santa Clara, USA). The GeneChip (Mouse Genome 430 2.0 Array) was washed, stained, and scanned using Affymetrix GeneAtlas Imaging Station to obtain the mRNA expression of various genes from mouse genome. The gene ontology, biological process, and fold-change in gene expression (2-fold change, control vs TCQA) were then analyzed using Transcriptome Analysis Console (TAC) Software (version 4.0.1) and database for annotation, visualization, and integrated discovery (DAVID) bioinformatics resources 6.8 [69,70]. The average signal log2 (control vs TCQA) was subjected to hierarchical clustering using Euclidean distance and average linkage algorithm of the TIGR Mev version 3.0.3 software (The Institute for Genomic Research, MD, USA).

Immunohistochemistry

Skin tissues from mice treated with TCQA and milli-Q water were collected, embedded in optimum cutting temperature (OCT) compound, and then cut at a thickness of 10 µM. The sections were then dried and fixed in 4% paraformaldehyde (SIGMA, Saint Louis, USA) and washed in different solutions of PBS, 20 mM glycine/PBS (SIGMA, Saint Louis, USA), and 0.1% v/v Triton X-100 /PBS (SIGMA, Saint Louis, USA). The skin samples were then blocked and immersed in a solution of 1:100 dilution of rabbit anti- β-catenin (Abcam, Rockford, USA) for overnight at 4°C in a wet chamber. Then washing in 0.1% v/v Triton X-100 /PBS was conducted. The samples were incubated with Alexa 594-conjugated anti-rabbit (Abcam, Rockford, USA) for 1 h at room temperature (RT) and stained with Hoechst and mounted in antifade solution (p-phenylenediamine, PBS, and glycerol).

Cells and cell culture

Human hair follicle dermal papilla cells (HFDPCs) were purchased from Cell Application Inc. (Tokyo, Japan). HFDPCs were maintained in papilla cell growth medium (Toyobo, Osaka, Japan) supplemented with growth factors: fetal calf serum, insulin transferrin triiodothyronine, bovine pituitary extract, and cyproterone solution (Toyobo, Osaka, Japan).

Human epidermal melanocytes (HEM) were purchased from Gibco Invitrogen cell culture. HEM were maintained in Medium 254 (Gibco, South America) supplemented with human melanocyte growth supplement HMGS (Gibco). The medium was changed every other day until the cells were 60% confluent and ready to be subcultured.

The cells were kept under sterile conditions at 37 °C in 75-cm² flask (BD Falcon, England, UK) in a humidified atmosphere of 5% CO₂. The viability of the cells was determined using trypan blue exclusion method.

Cell proliferation assay

The effect of TCQA on cell proliferation was assessed using the 3-(4,5-dimethyl-thiazol-2-y1) 2,5-diphenyl tetrazolium bromide or MTT assay (Dojindo, Kumamoto, Japan). The cells were seeded in 96-well plate at 3×10⁵ cells/well at 37 °C. After 24 h seeding, the growth medium was replaced by various concentrations of TCQA and incubated for 48 h and 72 h. MTT reagent (5 mg/ml) was added to the cells and incubated further for 8 h followed by an addition of 10% sodium dodecyl sulfate (SDS) and incubation overnight. Absorbance was measured at 570 nm using microplate reader (Powerscan HT, NJ, USA). The viability of the cells was quantified as the percentage (%) of living cells relative to the control (untreated cells). To detect any change in the morphology, the cells were observed using a contrast microscope (Leica Microsystems, Wetzlar, Germany).

Adenosine triphosphate (ATP) assay

Luminescence luciferase assay kit (Toyo Ink, Tokyo, Japan) was used to measure the ATP content. HFDPCs were seeded in 96-well plate at 3×10⁵ cells/ 100 µl well for 24 h, then the medium was replaced by fresh culture medium containing various concentrations of TCQA or 0.1 µM of the positive control minoxidil (Tokyo Chemical Industry, Tokyo, Japan). After 24 h and 48 h, the plate was incubated for 15 min at RT, then 100 µl of ATP reagent was added, followed by homogenization, and 1 min incubation in the dark. Then, 150 µl of the suspension was transferred to white 96-well plate and incubated for 10 min at RT. Intracellular ATP content was measured by luminescence and calculated as the percentage (%) of TCQA treated-cells relative to the control (untreated cells).

RNA extraction

HFDPCs were seeded at a density of 5×10⁵ cells per 100-mm petri dish. The cells were allowed to attach overnight and then the growth medium was replaced with a fresh one containing 0 and 10 µM of TCQA or 0.1 µM minoxidil used as positive control (Tokyo Chemical Industry, Tokyo, Japan). HFDPCs were treated as well with only 10 µM XAV939 (SIGMA, Saint Louis, USA), then with 10 µM XAV939 following by a further incubation with 10 µM TCQA (XAV939/TCQA), and finally co-treated with 10 µM XAV939 and TCQA simultaneously (XAV939+TCQA). After 6 and 12 h treatment, the growth medium was removed, and the cells washed with cold PBS before the total RNA was extracted using ISOGEN kit (Nippon Gene, Tokyo, Japan) following the manufacturer’s instructions. The RNA concentration was assessed using a NanoDrop 2000 spectrophotometer (NanoDrop Technologies, Massachusetts, USA).

Quantitative real-time PCR analysis

The extracted RNAs were used as templates for reverse transcription PCR using SuperScript III reverse transcription kit (Invitrogen, CA, USA). The cycling protocol is as follows: 95 °C for 10 min, 40 cycles of 95 °C for 15 s, and 60 °C for 1 min. TaqMan Universal PCR mix and TaqMan probes specific to Ctnnb1 (Mm 00483039_ m1), CTNNB1 (Hs99999168_m1), and ALPL (Hs01029144) (Applied Biosystems, CA, USA) were used for real-time PCR. The real-Time PCR was performed using 7500 Fast Real-Time PCR Software 1.3.1 (Applied Biosystems, CA, USA). Gapdh (Mm99999915_g1) and GAPDH (Hs 02786624_ g1) (Applied Biosystems, CA, USA) were used as endogenous control. All reactions were run in triplicates.

Protein extraction

HFDPCs were seeded at density of 5×10⁵ cells per 100-mm petri dish. After overnight incubation, the medium was removed and the cells were treated with 10 µM TCQA and 0.1 µM minoxidil (Tokyo Chemical Industry, Tokyo, Japan). After 12 h and 24 h treatment, total protein extraction was achieved using radio-immunoprecipitation assay (RIPA) buffer (SIGMA, Saint Louis, USA) and protease inhibitor following the manufacturer’s instructions.

Proteins from tissues collected from treated area at mice dorsal part were also extracted. The skin sections were crushed using a homogenizer in RIPA buffer (SIGMA, Saint Louis, USA) and protease inhibitor following the manufacturer’s instructions.

Protein samples (15 µg) were quantified using 2-D Quant kit according to manufacturer’s instructions (GE Healthcare, Chicago, USA).

Western blot

The protein samples were resolved in 10% sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinylidene difluoride membrane (PVDF) (Millipore, NJ, USA). The proteins were blotted with β-catenin 71-2700 (Thermo Fisher Scientific, Massachusetts, USA) and GAPDH sc32233 (Santa Cruz Biotechnology, Texas, USA) primary antibodies and incubated with the second antibody goat anti-rabbit IRDye 800 CW or IRDye 680 LT goat anti-mouse and the expression detected using LI-COR Odyssey Infrared Imaging System (LI-COR, NE, USA).

Immunocytochemistry

The cells were seeded at density of 3×10⁴/well in lab-tek slides chambers (SIGMA, Saint Louis, USA), and then allowed to attached for overnight at 37 °C. HFDPCs were treated with 0, and 10 µM TCQA and 0.1 µM Minox for 24 h. The medium was then removed and the cells were washed with 0.1% v/v Triton X-100 /PBS (SIGMA, Saint Louis, USA). After 1 h incubation at RT with the blocking solution, the first antibody rabbit anti- β-catenin (Abcam, Rockford, USA) was added for overnight at 4 °C. Therefore, the cells were immersed in a solution of 1:10000 dilution of Alexa 594-conjugated anti-rabbit (Abcam, Rockford, USA) and mounted with DAPI.

Statistical analysis

Results were expressed as mean ± standard deviation (SD). Statistical analysis was performed using Student’s t-test when comparing two value sets (control vs TCQA). P value of ≤0.05 was considered significant. ANOVA (One-way between-subject ANOVA unpaired) was performed to assess the level of significance between treated groups in vivo. The gene in gene expression was considered significant when change in expression was at least 2-fold (control vs TCQA). ANOVA (one way between-subject ANOVA unpaired) was performed to assess the level of significance between Minox and TCQA-treated cells; A P value of ≤0.05 was considered significant.