Hair Growth Promoting Effect of 4HGF Encapsulated with PGA Nanoparticles (PGA-4HGF) by β-Catenin Activation and Its Related Cell Cycle Molecules

Poly-γ-glutamic acid (γ-PGA)-based nanoparticles draw remarkable attention as drug delivery agents due to their controlled release characteristics, low toxicity, and biocompatibility. 4HGF is an herbal mixture of Phellinus linteus grown on germinated brown rice, Cordyceps militaris grown on germinated soybeans, Polygonum multiflorum, Ficus carica, and Cocos nucifera oil. Here, we encapsulated 4HGF within PGA-based hydrogel nanoparticles, prepared by simple ionic gelation with chitosan, to facilitate its penetration into hair follicles (HFs). In this study, we report the hair promoting activity of 4HGF encapsulated with PGA nanoparticles (PGA-4HGF) and their mechanism, compared to 4HGF alone. The average size of spherical nanoparticles was ~400 nm in diameter. Continuous release of PGA-4HGF was observed in a simulated physiological condition. As expected, PGA-4HGF treatment increased hair length, induced earlier anagen initiation, and elongated the duration of the anagen phase in C57BL/6N mice, compared with free 4HGF treatment. PGA-4HGF significantly increased dermal papilla cell proliferation and induced cell cycle progression. PGA-4HGF also significantly increased the total amount of β-catenin protein expression, a stimulator of the anagen phase, through induction of cyclinD1 and CDK4 protein levels, compared to free 4HGF treatment. Our findings underscore the potential of PGA nanocapsules to efficiently deliver 4HGF into HFs, hence promoting hair-growth. Therefore, PGA-4HGF nanoparticles may be promising therapeutic agents for hair growth disorders.


Introduction
Alopecia, also known as hair loss, is a condition caused by various factors such as hormone imbalance, stress, malnutrition, and chemotherapy [1][2][3]. Although hair loss is non-lethal, it has a profound effect on social interactions and psychological stability [4]. FDA-approved minoxidil and finasteride are widely used for treating hair loss [5]. However, undesired side effects, such as hypotension, dizziness, and tachycardia loss, limit their usage [6]. In this regard, there have been many attempts to target hair loss by developing novel pharmacological therapeutic agents sourced from traditional herbal medicines, free of side effects [7]. In this study, we explored a mixture of herbal extracts (4HGF) including Phellinus linteus grown on germinated brown rice, Cordyceps militaris grown on germinated soybeans, Polygonum multiflorum, Ficus carica, and Cocos nucifera oil. Several groups have reported the beneficial effects of P. multiflorum on hair-growth [7,8], and C. nucifera oil

Preparation, Characterization and Release Amount of PGA-4HGF
We synthesized the PGA-4HGF hydrogel nanoparticles simply via ionic gelation ( Figure 1A). Spherical nanoparticles were clearly observed by TEM ( Figure 1B), but their surface was irregular, presumably due to the complex salt-based ingredients of 4HGF. Although the size distributions were relatively broad, ranging up to 1000 nm, average hydrodynamic size was estimated to ~400 nm, which might be helpful to pass through the skin gap. We also evaluated the released amount of 4HGF from the nanoparticles by measuring the distinct absorbance corresponding to 4HGF. The results showed that the optical density (OD) value at 400 nm of 4HGF from PGA-4HGF was 0.97 ± 0.016 in a shaking water bath at 37 °C in phosphate-buffered saline (PBS), which simulated topical application by rubbing, while the OD value was 0.07 ± 0.002 without shaking at 4 °C in distilled water (DW) (** p < 0.01) ( Figure 1C). These results confirm that 4HGF release from the nanoparticles in a stimulated physiological condition.
To check whether PGA-4HGF had the ability to stimulate the stem cell like properties of DPCs, we observed cell morphology and aggregation. It is reported that DPCs have stem cell characteristics and are associated with the ability to induce hair-growth [26][27][28]. The flatten-elongated morphological changes and aggregation of primary DPCs were more prominent in 1:1:4 PGA-4HGF, compared with 4HGF, non-treated control, and 1:4 PGA-control ( Figure 2A). In addition, The number of DPCs was significantly increased in 1:1:4 PGA-4HGF, compared with 4HGF, non-treated control and 1:4 PGA-control ( Figure 2B,C). All PGA-4HGF samples had no effect on the viability of keratinocytes (HaCaT cells) ( Figure 2D). Therefore, 1:1:4 PGA-4HGF was used for the rest of the study.

Effects of PGA-4HGF on Primary Dermal Papilla Cells Proliferation and HaCaT Cells
To elucidate whether PGA-4HGF induces the proliferation of primary dermal papilla cells (DPCs), we used different PGA-4HGF combination ratios (4HGF:PGA:chitosan = 1:1:2, 1:1:4, or 1:1:8) to treat primary DPCs. The proliferation and aggregation of primary DPCs was enhanced by 1:1:4 PGA-4HGF compared to 4HGF, non-treated control, and 1:4 PGA-control (Figure 2A, black arrow). To check whether PGA-4HGF had the ability to stimulate the stem cell like properties of DPCs, we observed cell morphology and aggregation. It is reported that DPCs have stem cell characteristics and are associated with the ability to induce hair-growth [26][27][28]. The flatten-elongated morphological changes and aggregation of primary DPCs were more prominent in 1:1:4 PGA-4HGF, compared with 4HGF, nontreated control, and 1:4 PGA-control ( Figure 2A). In addition, The number of DPCs was significantly increased in 1:1:4 PGA-4HGF, compared with 4HGF, non-treated control and 1:4 PGA-control ( Figure  2B,C). All PGA-4HGF samples had no effect on the viability of keratinocytes (HaCaT cells) ( Figure 2D). Therefore, 1:1:4 PGA-4HGF was used for the rest of the study.  The images were taken after 24 h using a microscope (100× magnification, scar bars: 100 µm). (C) The number of DPCs were counted using Image J program. Data represent means ± deviation (SD). Data were analyzed with one-way ANOVA/Duncan's t-test. (p < 0.05). Values with different alphabets in the same row are significantly different. (D) Cell proliferation effects of PGA-4HGF on keratinocytes (HaCaT) cells (5 × 10 3 cells/well) in DMEM media for 24 h (grey bar: 1% of 4HGF, black bar: 2% of 4HGF). Each value represents the mean ± SD of three independent experiments. Data were analyzed with one-way ANOVA/Dunnett's t-test (** p < 0.01, *** p < 0.001 vs. the non-treated control).

Hair-Growth Effects of PGA-4HGF in Telogenic C57BL/6N Mice
To determine the hair-growth promoting effect of PGA-4HGF in telogenic C57BL/6N mice, PGA-4HGF was topically applied on the shaved dorsal skin ( Figure 3A). Black pigmentation was used as a biomarker for hair cycles, which changed from telogen to anagen phase [29]. By day 12, black areas were more prominent in the PGA-4HGF treated dorsal skins, compared with those in the non-treated control or PGA-control groups ( Figure 3B). These results suggest that PGA-4HGF induced telogen to anagen conversion of the HFs.  The design of animal experiments using C57BL/6N telogenic murine models. At day 0 which a blue triangle is pointed, a 3 × 4 cm 2 area of dorsal skin of all the mice were shaved. At day 6 and 12, which red triangles are pointed, the mice were sacrificed. (B) Dorsal skins were photographed on day 1 and 12. The images are representative pictures of the mice (n ≥ 6/group). (C) Survival rate of C57BL/6N telogenic murine models with PGA-4HGF topical treated during total animal experiments (n = 2). After PGA-4HGF treatment for 12 days in the C57BL6/N model, the analysis of the skin erythema status according to the standards announced by the Korea Food and Drug Administration. Body and liver weight of mice subjected to distilled water (control) gavage or PGA-4HGF Gavage. Data were analyzed with one-way ANOVA/Duncan's t-test (p < 0.05). Each mean with a is not significantly different.
To determine whether PGA-4HGF has adverse effects, we measured liver index after PGA-4HGF oral administration. The liver index of PGA-4HGF mice (56.7 ± 3.2) were similar to those of 4HGF mice (55.4 ± 3.3) ( Figure 3C). No liver swelling and no body weight change was observed in all groups. Our data suggest that orally administered PGA-4HGF was not hepatotoxic.
The number of HFs per unit area (400 × 400 µm 2 ) increased in PGA-4HGF-treated group (61.3 ± 11.36), compared with the number in the non-treated control (18.5 ± 3.15) and PGA-control (39.3 ± 10.67) groups ( Figure 4A,B). The number of HFs in PGA-4HGF was not significantly different from that in 4HGF. By day 12, the regrown hair lengths of the PGA-4HGF treated group were 1.2-fold longer, compared with that of 4HGF group ( Figure 4C). Our results indicate that PGA-4HGF induced hair-growth compared with the non-treated control and PGA-control groups and 4HGF groups.  Values with a, b, c in the same row are significantly different. Data was analyzed with independent t-test (** p < 0.01).

Anagen Phase Induction in PGA-4HGF Treated C57BL/6N Mice
The development of mouse HFs has been considered an indicator for the conversion of the HFs from telogen to anagen phase. To test whether PGA-4HGF induced anagen phase in hair cycle, the histological changes were analyzed after hematoxylin and eosin (H&E) staining of the tissues [5]. The histological data showed that the size and the number of the hair bulbs increased in the PGA-4HGFtreated group, compared with those in the 4HGF group ( Figure 5A). At day 12, 40.5 ± 10.0% and 25.3 ± 9.2% of the HFs in PGA-4HGF-treated group had significantly progressed to the anagen III and V phases, while 26.2 ± 17.5% and 23.4 ± 10.6% of the HFs in the 4HGF group were in anagen III and V (p < 0.05) ( Figure 5B). More HFs in the anagen phase were observed in PGA-4HGF groups (65.8 ± 19.2%) than in 4HGF groups (49.6 ± 28.1%). While the entire telogen phase HFs reside in the dermis, the anagen phase is related to the development and increase in the size and number of HFs residing in the deep subcutis ( Figure 5A) [30].

Anagen Phase Induction in PGA-4HGF Treated C57BL/6N Mice
The development of mouse HFs has been considered an indicator for the conversion of the HFs from telogen to anagen phase. To test whether PGA-4HGF induced anagen phase in hair cycle, the histological changes were analyzed after hematoxylin and eosin (H&E) staining of the tissues [5]. The histological data showed that the size and the number of the hair bulbs increased in the PGA-4HGF-treated group, compared with those in the 4HGF group ( Figure 5A). At day 12, 40.5 ± 10.0% and 25.3 ± 9.2% of the HFs in PGA-4HGF-treated group had significantly progressed to the anagen III and V phases, while 26.2 ± 17.5% and 23.4 ± 10.6% of the HFs in the 4HGF group were in anagen III and V (p < 0.05) ( Figure 5B). More HFs in the anagen phase were observed in PGA-4HGF groups (65.8 ± 19.2%) than in 4HGF groups (49.6 ± 28.1%). While the entire telogen phase HFs reside in the dermis, the anagen phase is related to the development and increase in the size and number of HFs residing in the deep subcutis ( Figure 5A) [30].

Effects of PGA-4HGF on Wnt/β-Catenin Signaling and Its Related Cell Cycle Molecules
Hair-growth is mainly regulated by Wnt/β-catenin signaling that is involved in hair cycle and hair formation [31,32]. To evaluate whether PGA-4HGF affects the β-catenin signaling pathway, we verified the expression of Wnt/β-catenin signaling molecules, using IHC and western blotting. Our results revealed that β-catenin stained intensity in the PGA-4HGF-treated group was stronger than that in the 4HGF-treated group in the epidermis, outer root sheath (ORS), and hair matrix ( Figure  6A). The distribution of β-catenin in the control group was primarily confined to the epidermis ( Figure 6A). We observed that β-catenin protein expression was higher in the PGA-4HGF-treated group (1.45 ± 0.05), compared with the 4HGF group (0.79 ± 0.03) ( Figure 6B). β-catenin plays an important role in DPCs proliferation and is also a transcription factor for the cell cycle-related proteins (cyclinD1/CDK4 complexes) [33,34]. CyclinD1 initiates progression from the G0 to G1 phase, which is activated by CDK4/6-dependent phosphorylation [35]. CyclinD1 and CDK4 protein expression levels were increased in the PGA-4HGF-treated group (1.36 ± 0.08 and 1.38 ± 0.07, respectively), compared with the 4HGF (0.73 ± 0.03 and 0.74 ± 0.02, respectively) group ( Figure 6B).

Effects of PGA-4HGF on Wnt/β-Catenin Signaling and Its Related Cell Cycle Molecules
Hair-growth is mainly regulated by Wnt/β-catenin signaling that is involved in hair cycle and hair formation [31,32]. To evaluate whether PGA-4HGF affects the β-catenin signaling pathway, we verified the expression of Wnt/β-catenin signaling molecules, using IHC and western blotting. Our results revealed that β-catenin stained intensity in the PGA-4HGF-treated group was stronger than that in the 4HGF-treated group in the epidermis, outer root sheath (ORS), and hair matrix ( Figure 6A). The distribution of β-catenin in the control group was primarily confined to the epidermis ( Figure 6A). We observed that β-catenin protein expression was higher in the PGA-4HGF-treated group (1.45 ± 0.05), compared with the 4HGF group (0.79 ± 0.03) ( Figure 6B). β-catenin plays an important role in DPCs proliferation and is also a transcription factor for the cell cycle-related proteins (cyclinD1/CDK4 complexes) [33,34]. CyclinD1 initiates progression from the G 0 to G 1 phase, which is activated by CDK4/6-dependent phosphorylation [35]. CyclinD1 and CDK4 protein expression levels were increased in the PGA-4HGF-treated group (1.36 ± 0.08 and 1.38 ± 0.07, respectively), compared with the 4HGF (0.73 ± 0.03 and 0.74 ± 0.02, respectively) group ( Figure 6B). These results suggest that PGA-4HGF promotes hair-growth by upregulating β-catenin and inducing the G 1 phase in the cell cycle, which affects the transition from the telogen to anagen phase.
These results suggest that PGA-4HGF promotes hair-growth by upregulating β-catenin and inducing the G1 phase in the cell cycle, which affects the transition from the telogen to anagen phase.

Identification of Keratin Proteins in the Dorsal Skins of PGA-4HGF-Treated Mice Using Two-Dimensional Electrophoresis (2-DE) and Peptide Mass Fingerprinting (PMF)
To investigate for differences in the dorsal skin and HF keratin proteins between PGA-4HGFtreated and PGA-control mice, the hair proteins of murine dorsal skin were analyzed using 2-DE gel [36]. Only spots with a two-fold increase in intensity (compared to PGA-control group) were selected using the PDQest software ( Figure 7A). The spot intensities of spots 4210 and 5203 were increased by 10.8 and 6.9-fold, respectively, in PGA-4HGF-treated mice, compared with the PGA-control group ( Figure 7B). The selected spots (No. 4210 and No. 5203) were identified as type II keratin K81, K85, and K86 by peptide mass fingerprinting (PMF) [37]. The individual proteins are listed in Figure 7C. This analysis revealed that PGA-4HGF incresed the production of hair keratins such as type II keratin K81, K85, and K86 in the spot compared with the PGA-control. These result suggested that the type II cuticular (K81, K85, and K86) were 10.8 and 6.9-fold higher in the PGA-4HGF-treated group, compared with the PGA-control.  To investigate for differences in the dorsal skin and HF keratin proteins between PGA-4HGF-treated and PGA-control mice, the hair proteins of murine dorsal skin were analyzed using 2-DE gel [36]. Only spots with a two-fold increase in intensity (compared to PGA-control group) were selected using the PDQest software ( Figure 7A). The spot intensities of spots 4210 and 5203 were increased by 10.8 and 6.9-fold, respectively, in PGA-4HGF-treated mice, compared with the PGA-control group ( Figure 7B). The selected spots (No. 4210 and No. 5203) were identified as type II keratin K81, K85, and K86 by peptide mass fingerprinting (PMF) [37]. The individual proteins are listed in Figure 7C. This analysis revealed that PGA-4HGF incresed the production of hair keratins such as type II keratin K81, K85, and K86 in the spot compared with the PGA-control. These result suggested that the type II cuticular (K81, K85, and K86) were 10.8 and 6.9-fold higher in the PGA-4HGF-treated group, compared with the PGA-control.

Discussion
Research on novel hair promoting agents is focused on effective natural products as well as carriers to improve and prolong hair-growth by enhancing penetration of effective components into hair follicles (HFs). Nanoparticles are known to efficiently transport drugs into skin areas with HFs [38][39][40]. PGA/chitosan nanoparticles are capable of entrapping biomolecules or agents in internal structures [22,24] and have a small size for enhancing drug delivery to HFs. The space between the hair follicle (20-50 µm) and shaft (16-42 µm) is approximately 200-400 nm, into which nanoparticles can conveniently fit [41,42]. We postulated that PGA-4HGF nanoparticles can get to the hair bulge in the stratum basal of the skin, avoiding the stratum corneum barrier because its size is~400 nm ( Figure 1A,B) [17,41,42]. PGA-4HGF nanoparticles in the hair bulge can improve blood circulation and in turn promote hair-growth [42]. To generate human HFs, DPCs should interact with themselves or other cell types (e.g., keratinocytes, dermal sheath cells, and follicular epithelial cells) [43]. We recently demonstrated that C. militaris grown on germinated soybeans (a 4HGF constituent) contains polyphenolic and flavonoid compounds which contribute to improve dermal blood circulation [44,45]. P. multiflorum and Thuja orientalis have been shown to increase β-catenin protein expression, involved in inducing telogen to anagen phase transition in the hair cycle [7,46]. In addition, PGA-4HGF nanoparticles can continuously release 4HGF for 6 h ( Figure 1C), suggesting that PGA-4HGF is effective in prolonging 4HGF duration [15,22,23]. Therefore, PGA-4HGF nanoparticles, which are smaller in size, compared to HF pores, are increased in HFs compared to solely PGA or 4HGF, because 4HGF coupled to PGA can conveniently fit in the bulge or hair bulbs by PGA encapsulation.
DPCs play a key role in generating hair bulbs and regulating hair-growth [5,47]. They are known to aggregate reactions that induce HF formation, and determine the hair bulb and shaft due to stem cell-like plasticity [48][49][50][51][52]. Therefore, the proliferation and stemness of DPCs can increase hair-growth. As shown in Figure 2, 1:1:4 PGA-4HGF increased primary DPC proliferation compared with the other groups. PGA-4HGF also alters the shape of DPCs from spindle-shaped cells to flat multipolar and elongate-shaped cells and promotes DPCs aggregation, which are characteristics of active DPCs that can increase HF formation and prolong the anagen phase ( Figure 2) [28]. Morphological changes and aggregation behaviors of PGA-4HGF-treated DPCs may be associated with the stem cell features in DPCs for hair formation [26,28]. Next, we also investigated whether 1:1:4 PGA-4HGF dose could promote hair-growth in an in vivo model.
The dorsal skin color of telogenic C57BL/6N mice is pink, but changes to dark black pigments during the anagenic phase [50]. Individual HFs can be divided into the specific hair cycle stages such as relative rest (telogen), active growth (anagen III-V), and involution, driven by controlled apoptosis (catagen), phases [53,54]. The HFs in the anagen phase have enlarged hair bulbs and large amounts of melanin around the upper half of the DPCs [53]. Melanin pigmentation is defined by melanogenesis, which involves transport of melanin granules to epidermal and hair follicle keratinocytes surrounding DPCs [55][56][57]. Melanin synthesis is related to hair-growth and hair shaft formation in the early anagen phase [50,58,59]. We observed more prominent black skin in the PGA-4HGF-treated groups, compared with the non-treated control and PGA-control groups ( Figure 3B), suggesting that PGA-4HGF treatment can improve active growth (anagen phase).
The growing use of nanoparticles demands cautious evaluation of unexpected toxicities due to their physical and chemical characteristics [60]. It is reported that they can cut through the small intestines and spread extensively throughout the body [61]. We investigated if PGA-4HGF caused unwanted hepatotoxicity by observing liver images and calculating the liver index of each mouse after oral administration of PGA-4HGF. The average liver index of the PGA-4HGF-treated group (56.7 ± 3.2) was similar to that of the 4HGF-only group (55.4 ± 3.3) ( Figure 3C). The mouse survival rate after oral or topical administration was 100% ( Figure 3C). This is confirmation of no PGA-4HGF-associated abnormalities, as no liver damage and toxicity were observed.
Furthermore, it is reported that, as the cycle of HFs progresses from dormant to growth, the dorsal skin color turns to black and becomes more intense [53]. H&E staining results confirmed that the epidermis of mice treated with PGA-4HGF were darker and thicker, compared with the 4HGF group ( Figure 4A). The PGA-4HGF-treated group produced a better effect on the length of regrowth hair (3.89 ± 1.04 mm), compared with the PGA-control group (1.89 ± 0.58 mm) ( Figure 4B,C). In the telogen phase, HFs reside in the dermis and do not extend to the subcutis. Conversely, in the anagen phase, HFs reside in the deep subcutis and move closest into the panniculus carnosus in the late anagen phase. Based on previous research, we analyzed the HF cycle of each group and observed that a higher number of HFs in the PGF-4HGF-treated group were in anagen III and least in the telogen phase [53]. As a result, the PGA-4HGF-treated group converted approximately 15.3% of telogen follicles to anagen follicles, suggesting that it induced the anagen phase ( Figure 5).
Hair is composed of type I and type II keratins, a family of fibrous structural proteins, containing 14-18% cysteine [62,63]. The existence of disulfide bonds in the keratin fiber complex determines the conservation of the strength, flexibility, and shape of HFs [63]. HFs in anagen phase consist of the cylindrical cell layers and the germinal matrix, which divide cells on the bulb [64]. Keratinocytes rapidly grow in the hair matrix zone surrounding the DPCs, which is stimulated by the β-catenin signaling pathway for the induction of hair-growth [65,66]. Interaction between keratinocytes and DPCs induce the expression of numerous genes encoding keratin intermediate filaments, such as type II keratin proteins [64,67,68]. Some reports showed that type II keratin genes are activated sequentially in DPCs [69]. According to previous studies, type II keratin K85, which constitutes the medulla, pre-cortex cuticle, and matrix, is expressed on the germinative compartment from the lower-most hair cuticle [67,70,71]. Type II keratin proteins K81 and K86 constitute the mid-cortex and upper medulla and are expressed in the medulla [62,67,72]. When keratin protein expression occurs improperly, this can cause hair diseases such as monilethrix, which is characterized by HF collapse and deformation, [62,70,73], ectodermal dysplasia, hypotrichosis, nail dystrophy, and hair scalp fragility [67,70,71]. In order to compare the expression of the hair-related proteins in mice treated with PGA-4HGF or PGA-control, 2-DE analyses using SDS-PAGE and PMF were performed. We observed that PGA-4HGF produced 10.8 and 6.9-fold more type II keratin proteins compared with the PGA-control group (Figure 7), suggesting that PGA-4HGF assisted in the formation of more durable HFs.
Previous studies have demonstrated that β-catenin induction in DPCs causes both hair-growth and regeneration [66]. β-catenin increases the proteins involved in stem cell functions [74]. Our previous data indicate that PGA-4HGF might increase β-catenin protein expression because it induces stem cell-like morphology in DPCs. β-catenin is also a transcription factor for cyclinD1 and CDK4 [3,34]. Most HF cells from bald patients are in the G 0 phase [75]. CyclinD1 is known to induce G 1 /S phase transition [76][77][78][79]. Therefore, β-catenin is required for hair-growth, because it can induce the transition from the G 0 /G 1 to the S phase [80]. We observed that PGA-4HGF increased the levels of β-catenin, cyclinD1, and CDK4 protein expression in the skin with hair ( Figure 6). This data revealed that the down-stream targets of Wnt/β-catenin, cyclinD1, and CDK4 were upregulated in the PGA-4HGF-treated group. Herein, we observed that PGA-4HGF treatment increased the level of activated β-catenin ( Figure 6), suggesting that the improvement of cell cycle progression in PGA-4HGF-treated cells can be attributed to the enhancement of the β-catenin pathway, since cyclinD1 and CDK4 are transcriptional targets of β-catenin that control cell cycle progression and induce cell proliferation.
In conclusion, we evaluated the hair promoting activity of PGA-4HGF in vitro and in vivo. Our results showed that PGA-4HGF induced DPCs proliferation, aggregation, and stem-cell like morphological changes, which may efficiently transport 4HGF into HFs. Overall, PGA-4HGF activates the β-catenin signaling pathway, leading to G 1 /S transition by increasing cyclinD1 and CDK4 protein levels, resulting in the increase of type II keratin proteins and melanin pigments, which promote durable hair formation. Therefore, we propose that the use of PGA nanocapsules for delivering 4HGF, may represent a promising therapy for treating hair-growth disorders.

Preparation and Characterization of the 4HGF Loaded Nanoparticles (PGA-4HGF)
4HGF (The mixture of P. linteus grown on germinated brown rice, C. militaris grown on germinated soybeans, P. multiflorum, F. carica, C. nucifera oil, etc.) was kindly provided by CARI Co. Ltd. To entrap 4HGF into PGA hydrogel, the 4HGF mixture was filtered through filter paper Whatman's No. 1 and then centrifuged to obtain the supernatant and sedimented at 1630× g for 10 min (Union 32R, Hanil Science Industrial Co., Incheon, Korea). PGA-4HGF hydrogel nanoparticles were prepared through simple ionic gelation under shaking condition at room temperature. In brief, hair growth solution was thoroughly mixed with poly γ-PGA solution (1 mg/mL), subsequently dropped into aqueous chitosan solution (1 mg/mL) at room temperature with stirring for 1 h. Initially, hair growth solution with poly γ-PGA solution were added to several ratios of chitosan (1:1:2, 1:1:4, and 1:1:8). After the hydrogel nanoparticles were synthesized, they were separated via centrifugation at 10,000× g, then dispersed in distilled water at 4 • C until use. The size and morphology of the prepared nanoparticles were analyzed by transmission electron microscopy (TEM) images on a Jeol EM-2010 microscope (Jeol Co., Peabody, MA, USA) and dynamic light scattering using a Zetasizer Nano-ZS (Malvern Co., Malvern, UK).

HGF Release from the Nanoparticles (PGA-4HGF)
The released amount of 4HGF from the hydrogel nanoparticles was evaluated by measuring the absorbance at 400 nm of the supernatant solution, which corresponded to the 4HGF. The shaking incubator was used to evaluate the released amount of 4HGF from hydrogel nanoparticles. Firstly, 400 µL DW or PBS (10 mM, pH 7.4) was added to the tube which contained 200 µL of nanoparticle mixture. The mixture was incubated for 6 h in the shaking incubator (37 • C, 200 rpm). Then, the mixture was centrifuged, and the supernatant was used to measure absorbance at 400 nm to determine the amount of 4HGF released from the hydrogel nanoparticles, based on the standard calibration plot.

Cell Culture and Proliferation of Primary DPCs Using Real Time Microscopy
Primary DPCs were isolated from the follicle bulbs of C57BL/6N mice whiskers as previously described [81,82]. The isolated DPCs were cultured in Dulbecco's modified Eagle's medium (DMEM; Invitrogen Co., Carlsbad, CA, USA) with 100 units/mL each of penicillin A and streptomycin (Gibco BRL, Grand Island, NY, USA), and 10% heat-inactivated fetal bovine serum (FBS; Gibco BRL Grand Island, NY, USA). Cells were grown at 37 • C in fully humidified 5% CO 2 (Forma 3111, Thermo Fisher Scientific, Waltham, MA, USA).

Anagen Phase Induction in C57BL/6N Mice
Female 7-week-old C57BL/6N telogenic mice were purchased (Orient Bio, Eumsung, Republic of Korea) and maintained under specific pathogen free (SPF) conditions with 12 h light/darkness cycles. Six mice were randomly divided into 1 of 5 groups and allowed to acclimatize to laboratory conditions for 7 days. Mice were fed with the standard diet and allowed free access to drinking water. A 3 × 4 cm 2 area of dorsal skin of all the mice were shaved using attenuated hair removal cream (BIKIRO cream (Thioglycolic Acid 80%); Tai Guk Pharm. Co. Ltd., Gyeonggi-do, Republic of Korea). Samples were applied daily on the shaved dorsal skin hair. As initially indicated, samples received: 200 µL of control (distilled water, DW), PGA-control (PGA:chitosan = 1:4), PGA-4HGF (4HGF:PGA:chitosan = 1:1:4), 4HGF, and 3% minoxidil (Dongsung, Seoul, Republic of Korea). Superficial properties of hair growth were measured and photographed on day 12 of PGA-4HGF treatments ( Figure 3A). All researchers on the animal studies were complied with the standards for the care and use of experimental animals. The animal study was performed in accordance with the Institutional Animal Care and Use Committee (IACUC) guidelines at Gachon University (approval number: GIACUC-R2016022, approval Date: 31 October 2016).

Toxicity Test of PGA-4HGF
γ-PGA/chitosan and PGA-4HGF nanoparticles were orally administered to the mice at a dose of 2 mL/kg. The control mice received 2 mL/kg DW. A day after oral administrations of treatment, the mice were sacrificed, and their livers recovered and weighed. Liver index was calculated as the weight of each liver divided by the total body weight (g). All researchers on the animal studies were complied with the standards for the care and use of experimental animals. The animal study was performed in accordance with the Institutional Animal Care and Use Committee (IACUC) guidelines at Gachon University (approval number: GIACUC-R2016021, approval Date: 31 October 2016).

Hair Follicle Counting and Hair Length Determination
Digital photomicrographs were taken from representative areas of the dorsal skin tissue slides at a fixed 40x magnification. All images were chopped in a fixed area (400 × 400 µm 2 ). The HFs in deep subcutis were manually counted (n > 30/mouse). The regrown hairs were plucked from the dorsal skin areas (100 × 300 mm 2 ) and the hair length of each sample calculated (n > 30/mouse).

Histological Preparation and Hematoxylin-Eosin Staining
Dorsal skin tissues from each mouse were fixed in 10% formaldehyde for 24 h and embedded in paraffin blocks. They were cut transversely or longitudinally into 4 µm thick sections and mounted on glass slides. To observe for histological changes, the slides were stained with H&E staining solution. The slide images were taken under a Nikon Eclipse Ti microscope equipped with a color digital camera (Point Grey Research, Richmond, BC, Canada) and analyzed using MetaMorph software (Molecular devices, Sunnyvale, CA, USA).

Immunohistochemistry
As previously described [5], the dorsal skin was stained with anti-β-catenin (Cell Signaling, MA, USA) antibodies post topical PGA-4HGF treatment. To extinguish endogenous peroxidase activity, de-paraffinized parts were pre-treated with 0.3% H 2 O 2 for 10 min. After washing with Tris-buffered saline containing Tween (TBS-T), the sections were incubated with 4% bovine serum albumin(BSA) with dextran for 30 min to prevent nonspecific binding of the secondary antibody, and incubated with anti-β-catenin (1:400 dilution) antibodies for 1 h. Slides were incubated with anti-rabbit biotin secondary antibody (Agilent, CA, USA) for 30 min. The slides were counter-stained with Mayer's hematoxylin for 1 min, viewed under a Nikon Eclipse Ti microscope equipped with a color digital camera (Point Grey Research, Richmond, BC, Canada), and analyzed using the MetaMorph software (Molecular devices, Sunnyvale, CA, USA).

Protein Identification by Two-Dimensional Electrophoresis (2-DE) and Peptide Mass Fingerprinting (PMF)
IPG dry strips (4-10 NL IPG, 24 cm; Genomine, Republic of Korea) were equilibrated with a mixture of 7 M urea and 2 M thiourea (containing 2% CHAPS, 1% DTT, and 1% pharmalyte) for 12-16 h. Sample (200 µg) was then loaded in each well. Isoelectric focusing (IEF) was performed using a Multiphor II electrophoresis unit and EPS 3500 XL power supply at 20 • C. Equilibrated strips were run in the Hoefer DALT 2D system, following the manufacturer's instructions (Amersham Biosciences, Uppsala, Sweden). Gels (2D) were stained using Colloidal Coomassie Brilliant Blue as described by Oakley et al. [84]. Quantitative analyses of digitized images were carried out using the PDQuest (version 7.0, BioRad, city, if any state, country) software, following the manufacturer's instructions. The quantity of each spot was normalized by total valid spot intensity. Protein spots were only selected if they showed at least 2-fold increased protein expression values compared to the PGA-control.
For peptide mass fingerprinting, protein spots were excised, and digested with trypsin (Promega, Madison, WI), mixed with α-cyano-4-hydroxycinnamic acid in 50% acetonitrile/0.1% trifluoroacetic acid (TFA), and subjected to MALDI-TOF analysis (Microflex LRF 20; Bruker Daltonics, Billerica, MA) as described by Fernandez et al. [37]. The search program MASCOT (Matrixscience, available at www.matrixscience.com), was used for protein identification. The following parameters were used for the database search: trypsin as the cleaving enzyme, a maximum of one missed cleavage,