Glutamic acid promotes hair growth in mice

Glutamic acid is the main excitatory neurotransmitter acting both in the brain and in peripheral tissues. Abnormal distribution of glutamic acid receptors occurs in skin hyperproliferative conditions such as psoriasis and skin regeneration; however, the biological function of glutamic acid in the skin remains unclear. Using ex vivo, in vivo and in silico approaches, we showed that exogenous glutamic acid promotes hair growth and keratinocyte proliferation. Topical application of glutamic acid decreased the expression of genes related to apoptosis in the skin, whereas glutamic acid increased cell viability and proliferation in human keratinocyte cultures. In addition, we identified the keratinocyte glutamic acid excitotoxic concentration, providing evidence for the existence of a novel skin signalling pathway mediated by a neurotransmitter that controls keratinocyte and hair follicle proliferation. Thus, glutamic acid emerges as a component of the peripheral nervous system that acts to control cell growth in the skin. These results raise the perspective of the pharmacological and nutritional use of glutamic acid to treat skin diseases.

Several studies had shown that the skin performs as neuro-endocrine organ [8][9][10] and its activities are mainly regulated by local cutaneous factors 9 . This interaction between skin and environment factors can regulate Central Nervous System (CNS) functions 11 . For instance, ultraviolet light absorption by the skin can upregulate neuroendocrine axes 10,11 and it is suggested to modulate body weight [12][13][14] and depression-like behaviour 15 . Specifically, UVB skin exposure stimulate corticotropin-releasing hormone protein production and gene expression in the hypothalamus 10 .
Previous reports have been identified both the glutamate receptors and specific glutamate transporters in epidermal keratinocytes 2 . Physiologically, glutamatergic signalling through N-methyl-D-aspartate (NMDA) receptor was previously shown to occur in hair follicle cells. GA signalling is essential for the innervation and differentiation of Grin1 positive Schwann cells during piloneural collar development in hair follicles 4 . Specifically, NMDA receptors are highly expressed in type I and type II terminal Schwann cells. These cells are circumferentially localized in the bulge border and cover most outer root sheath keratinocytes in the isthmus 4 .
In cell culture studies, NMDA induced an increase in the number of keratinocytes and in the intracellular calcium concentration 16 ; whereas, in vivo studies have shown that the topical application of GA to wounded skin in diabetic rats increases the rate of wound closure by inducing collagen synthesis and crosslinking 17 . In addition, 1% L-glutamic acid-loaded hydrogels accelerated vascularization and macrophage recruitment in diabetic wound 17 . D-glutamic acid has also been shown to act on damaged skin by accelerating the barrier recovery 3 , altogether suggesting a positive effect in skin repair.

Topical glutamic acid decreases apoptotic related genes.
To determine whether the results obtained in cultured cells could be translated into an in vivo model, we employed four different concentrations of GA on the dorsal skin of Swiss mice (Fig. 2a). To understand how GA promotes proliferation and improves viability, we evaluated the expression of genes involved in apoptosis. There were reductions of Bcl2 gene expression in cells treated with GA 0.1%, 0.5% and 10% (Fig. 2c). BAX was decreased in cells treated with 10% GA (Fig. 2c). However, we found no differences in Casp9 expression (Fig. 2c). Next, we evaluated whether topical GA could stimulate the expression of genes related to inflammatory response. We found no differences in Il1-β, Tnf-α and Il10. However, F4/80, a macrophage marker, and Monocyte Chemoattractant Protein-1 (Mcp1) gene expression were increased after 14 days of 1% GA (Fig. 2c). Additionally, topical GA 10% decreased the expression of Glutamate Ionotropic Receptor NMDA Type Subunit 1 (Grin1) with no differences in Glutamate Aspartate Transporter 1 (Glast) expression (Fig. 2c).
Topical glutamic acid accelerates hair growth in healthy mice. Surprisingly, 1% and 10% GA accelerated hair growth after 14 days of topical treatment (Fig. 3a). Using photomicrographs, we also showed that GA increased external root sheath across all GA concentrations (Fig. 3b, Supplementary Fig. 1f.) with no hyperkeratosis effect. We also consistently identified increased BrdU positive cells in the hair follicles and epidermal layer after 14-days of GA topical treatment (Figs. 3c, 4f).
Exogenous topical glutamic acid increased vascularization. We identified macroscopic differences in vascularization after 14 days of GA treatment. The 0.5% and 10% GA topical application increased skin vascularization (Fig. 4a,b). To further explore these findings, we evaluated whether GA could induce the expression of genes involved in vascular regulation. We found that 1% GA topical treatment increased Hypoxia Inducible Factor 1 Subunit Alpha (Hif1a), a master regulator of vascularization 18,19 (Fig. 4c). Also, 1% GA topical treatment increased the Vascular Endothelial Growth Factor A (Vegf), which induces proliferation and migration of vascular endothelial cells and is essential for physiological angiogenesis [20][21][22] (Fig. 4c). However, we found no difference in gene expression of CD31 after 14 days of topical GA treatment (Fig. 4c).
Single cell RNA sequencing analysis showed differences in glutamate receptor and transporter localization between mice and human. We evaluated glutamate receptor expression using immunostaining, quantitative PCR, and single-cell RNA sequencing techniques. We identified that NMDA receptor subunits Grin1, Grin2a, Grin2b and Grin2c are expressed in the skin (Fig. 4d,e), and Grin2b is expressed specifically in keratin 14 + cells (Fig. 4d,e). Due to the wide number of subunits (5 GA receptor families with 26 subunits), we used a single cell RNA sequencing approach to improve accuracy (Fig. 5a). Using public transcriptome libraries of skin tissue, we analysed ~ 73,000 mice and human epidermal cells from back (mice), foreskin, trunk, and scalp (human). This cross-species analysis showed a similar percentage (5%) of glutamatergic epidermal population in the skin (Fig. 5b,d). In humans, we identified NMDA receptors as the highest expressed subunits in the basal layer and hair follicular cell clusters, specifically the GRIN2A subunit (Fig. 5b). In addition, we identified melanocytes expressing Glutamate Ionotropic Receptor Delta Type Subunit 1 (GRD1) (Fig. 5b), granular cells expressing Excitatory Amino Acid Transporter 4 (SLC1A6) and basal layer cells expressing Excitatory Amino Acid Transporter 1 (SLC1A3) (Fig. 5c). In mice, we identified Grin2d (in the sebaceous gland) and Grik1 (in the www.nature.com/scientificreports/ hair follicle bulge) as the most expressed subunits (Fig. 5d). Additionally, we identified the expression of Excitatory Amino Acid Transporter 1 and 3 (Slc1a3 and Slc1a1) (Fig. 5e) in 50% of all sebaceous gland cells (Fig. 5e).
To predict a GA-mediated cell signalling pathway between GA pathway and hair follicle-related genes, we used computational interaction network analysis STRING 23 . In this way, we used glutamate receptor pathway genes and hair cycle genes ontologies ( Supplementary Fig. 2). We found that GA receptors interact with hair cycle genes through the tyrosine-protein kinase Fyn, Ca 2 + /calmodulin-dependent protein kinase II (CaMKII) and protein kinase B (Akt) (Fig. 5f). Additionally, we found Bcl2 as a common apoptotic regulator between both hair cycle and GA pathways (Fig. 5f). To confirm, we evaluated the protein expression of Fyn, CaMKII and Akt in the 14-day topical GA-treated mice ( Fig. 5g-i). We found no differences in Fyn quantification (Fig. 5i). However, we confirmed that AKT2 phosphorylation increased after 14 days of topical 1% GA treatment (Fig. 5g). Also, pCaMKII increased after 14 days of topical 10% GA treatment (Fig. 5h).

Discussion
Currently, there are no studies describing GA treatment or even the effect of GA on hair growth or epidermal cell proliferation. However, upon searching through major patent agencies, we found five patents/patent requests claiming the benefits of topical GA (or derived molecules) for hair growth. One of these patents described the use of GA as a hair conditioner (patent number CN106580722A, China) for hair restoration and the prevention of alopecia. Another patent showed a Poly-Gamma-GA composition for preventing hair loss and promoting hair growth (KR20150110149A, Korea). In addition, there were synthetic compounds of GA attached to minoxidil for keratinocyte growth and hair growth in humans (USOO58O1150A, USA), a 2 to 12% GA topical cream for combating hair loss or alopecia in humans (FR2939038B1, France) and, finally, 42 different molecules derived from L-glutamic acid were described as hair growth promoters (PI9302024A, Brazil). However, reviewing all these patents/patents requests, we could find no description of the cellular mechanisms responsible for the stimulation of hair growth in response to GA application. Thus, our work provides experimental proof of a mechanistic link between GA and hair growth.
Here, we evaluated some of the potential mechanisms involved in effect of GA on hair growth. First, using a cell culture approach, we challenged 100% confluent primary human keratinocytes, HaCaT-keratinocytes, and human Fibroblast (in medium depleted of foetal bovine serum (FBS)/growth factor supply) to continue growing. Our results showed that GA increases the proliferation and viability of keratinocytes, even under these extreme conditions. Thus, GA could represent an interesting approach as a cell growth media supplement, replacing traditional supplements that are more expensive. This finding is further supported by data published previously, which shows that MK-801, an antagonist of the GA receptor (NMDA receptor), decreases the proliferation of primary human keratinocytes 2 and also prevents hyperplasia induced by acetone 3 , suggesting an anti-proliferative effect.
The skin is a critical peripheral neuro-endocrine-immune structure that interact to central regulatory systems 24 . As a response, the skin can trigger cutaneous nerve endings to inform the CNS on changes in the epidermal or dermal environments to produce neural or immune responses at the local and systemic levels 24 . Human skin reacts to several neuropeptides and neurotransmitters by paracrine, autocrine, vasculature and nerves stimulus 25 . Primary and HaCaT keratinocytes are sources of Glutamic acid secreting ≈ 1 mM l-glutamic www.nature.com/scientificreports/ acid to the culture medium when 100% confluent 26 suggesting a paracrine or autocrine stimulation. Indeed, major aspect of neuroendocine regulation in the skin, specifically the hypothalamic-pituitary-adrenal axis and melatoninoergic system in the skin was previously discussed 8 . GA has potent neurotoxic effects, and this could represent a challenge for either experimental or clinical use. Elevated amounts of GA lead to neuronal death in a process described as excitotoxicity [27][28][29] . GA transporters are a potent GA uptake system, acting as a neuronal compensatory response for excitotoxicity. GA transporters are known to prevent disproportionate activation of   www.nature.com/scientificreports/ GA receptors by constantly removing GA from the extracellular space [30][31][32] . Here, we determined the excitotoxic GA concentration. In vitro, 100 mM GA decreased keratinocyte viability, and topical GA decreased Bcl2 and Bax expression. Altogether, our results support the excitotoxic effect of higher concentrations of pH-neutralized GA in keratinocytes.
To understand the exogenous GA effect on the skin, we explored the GA transporter landscape at single-cell resolution in human and mice skins (Fig. 5b,d). Additionally, we showed Slc1a3 expression using quantitative PCR, and a similar Slc1a3 (Glast) expression after exogenous GA application (Fig. 2c). Future research could help to identify the role of GA-induced excitotoxicity and the GA uptake system in the skin.
Regarding the GA receptors, different subpopulations of glutamatergic cells have been extensively described [33][34][35][36] . In the skin, previous reports identified the localization of GA receptors and transporters in the epidermis from rats and mice, as well as in human keratinocytes 2,3,5 . These studies showed similar cross-species characteristics: a smaller subpopulation of cells expressing receptors and transporters 2 . Consistent with these findings, here, we showed a small subpopulation of epidermal cells expressing GA receptors along the skin, with varying intensity (Fig. 5b,d). Our results suggest that these glutamatergic keratinocytes are responsive to exogenous GA stimulation.
Previous reports showed that vascularization increases during the anagen phase of the hair cycle and decreases during the catagen and telogen phases. This angiogenesis process was spatially correlated with the upregulation of VEGF 37 . Also, the hypoxia-inducible factor (HIF) has been shown to coordinate the up-regulation of multiple genes controlling neovascularization, such as Vegf 38 . Here, we showed that Hif1a and Vegf expression increased after 14 days of GA topical treatment on the back skin of mice with a remarkable change in angiogenesis, as previously shown 17 .
The Hypoxia-inducible factor-1α, encoded by the gene Hif1a showed to stimulate hair growth 39 and some HIF-1α-stimulating agents significantly increase dermal papilla cell proliferation 40 . Minoxidil 2,4-diamino-6-piperidinopyrimidine3-oxide, a vasodilator used for the treatment of pattern hair loss, is a direct inhibitor of PHD-2 (prolyl-hydroxylase 2) which hydroxylates HIF-1α causing its degradation. Also, Minoxidil stimulates the transcription of hypoxia-response element genes such as VEGF [39][40][41] . Here, we showed different concentrations of GA with hair growth and angiogenic effects. Publicly available patents proclaim benefit by using high percentage of GA concentration for skin treatment and hair growth stimulation. The 10% GA treatment showed hair growth and angiogenesis stimulation, but no differences in Hif1a expression after 14 day of topical GA. Previous studies showed the ranges of GA saturation by specific GA concentration [42][43][44] . In this way, we suggest that 10% GA topical treatment could achieve a post-acute signal saturation secondary to the time exposure and the Glutamic Acid concentration here presented. However, more studies are needed to elucidate a possible saturation and angiogenic effect of 14 days GA topical treatment on the back skin of mice.
A recent study supports that GA-mediated signalling could be involved in hair growth 45 . The authors showed that glutamine, a molecule similar to GA, controls the fate of stem cells in the hair follicle. The capacity of the outer root sheath cells to return to the stem cell state requires suppression of a metabolic switch from glutamine metabolism and is regulated by the mTORC2-Akt signalling axis 45 . Similarly, our result suggests that GA increases AKT phosphorylation and hair follicle cell stimulation. In this way, our results further suggest that GA activates the hair cycle by stimulating the stem cells to differentiate into the outer root sheath. However, future studies should describe the hair-follicle cycle modulation after topical GA treatments.
Taken together, the cell-based and experimental outcomes of this study provide a mechanistic advance in the characterization of GA-induced effects on hair growth and could become an attractive approach to treat hair growth disorders, or for aesthetic hair stimulation.
Further studies should focus on the relationship between skin disorders and GA and how GA, present in food, could impact on skin health. Topical glutamic acid treatment. The dorsal region of mice was treated once daily using different concentrations of GA. To ensure a uniform 200 μL treatment, we used different syringes preloaded with Vaseline (control), 0.1%, 0.5%, 1% or 10% GA (Supplementary Fig. 1c). The treatment was spread manually using gloves which were changing between each group. To avoid removal of the treatment, mice used Elizabethan collars 8 of 14 days of treatment.
MTT solution was prepared in Krebs-HEPES buffer (10 mM HEPES, 1.2 mM MgCl 2 , 144 mM NaCl, 11 mM glucose, 2 mM CaCl 2 and 5.9 mM KCl). After 100% confluence 6-wellplates, HaCaT and human fibroblast were incubated with the different concentrations of GA in DMEM without FBS for 1, 2 and 4 days. Primary keratinocytes were incubated with the different concentrations of GA in KGM Gold medium without growth factors for 2 days. After treatment, the medium was removed, MTT solution (0.5 mg/mL) was added to each well and the plates were incubated at 37 °C for 3 h. The solution was then removed and 300 μL of DMSO was added before being incubated in the dark with 60 rpm shaking. The absorbance was measured at a wavelength of 540 nm in a microplate reader (Globomax). HaCaT culture experiments were performed in quadruplicate in 4 different experiments. Primary keratinocytes culture experiments were performed in triplicate in 2 different experiments. Human fibroblast culture experiments were performed in triplicate in 2-3 different experiments. BrdU experiments were performed as previously described 26 . Briefly, to assess the effect of GA on cell proliferation, HaCaT human keratinocytes were maintained DMEM (Gibco) containing 4.5 g/L glucose, 4 mM L-glutamine, 100 units/mL of penicillin, 100 μg/mL of streptomycin and 10% FBS. Incubation conditions were 37 °C in 5% CO 2 / humidified air. HaCaT cells were plated on coverslips in 24-well plates (1 × 10 5 cells/well) and exposed to GA for 48 h (10 and 100 mM) in DMEM without FBS. After treatment, cells were incubated with BrdU (10 µM, Sigma) for 3 h, then fixed with 4% PFA in 0.1 M PBS for 10 min at RT. For BrdU staining, cells were washed with PBS, and DNA was denatured with 1 N HCl for 1 h at RT. Cells were blocked for 1 h in blocking solution containing 10% goat serum and 0.2% Triton X-100 in PBS, followed by an incubation with primary (rat anti-BrdU; 1:200; Ab6326); and secondary (goat anti-rat FITC, 1:200; sc2011) antibodies prepared in 3% goat serum/0.2% Triton X-100 in PBS, and incubated overnight and for 2 h, respectively. The nuclei were labelled with DAPI, and coverslips were mounted onto glass slides for microscope imaging. Images were captured on fluorescence microscopy (Olympus BX41). The results of BrdU immunopositivity cells represent the average of 3 coverslips per experimental replicate, where 3 fields were imaged per coverslip and averaged. The number of immunopositive cells was quantified per image using the ImageJ software and are expressed as a percentage relative to the total DAPI nuclei.
Animal photo documentation. Hair growth processes were photo documented using a D610 Nikon digital camera (Nikon Systems, Inc., Tokyo, Japan). We used a stand to secure a similar distance from the camera to the treated skin site, and the same person took the photos.
Vessel analysis. Vascular density measurements were calculated from digital images obtained using a D610 Nikon digital camera (Nikon Systems, Inc., Tokyo, Japan). We used a stand to secure a similar distance from the camera to the upside-down back skin samples. Vascular density ratio was calculated vascular as follow: vessel area/total area * 100% 50 . The number of pixels were digitally determined by densitometry, using Image J software (National Institutes of Health).

Figure 5.
Cross-species skin GA receptor landscape using single-cell RNA sequencing. Generation of glutamic acid receptor landscape using public data reveals GA distribution at single cell resolution in mice and humans (b-e). Schematic representation of the single-cell RNA sequencing analysis using publicly available datasets from mice and human epidermal layers (a Western blotting. For protein quantification, mice were treated topically 14 days with GA, once a day. Animals were anesthetized (ketamine hydrochloride 80 mg/kg and xylazine chlorhydrate 8 mg/kg) using Labinsane 53 , placed them in ventral decubitus position for cleaning the skin. Then, we harvested a 6 mm full-thickness skin sample 54 from the back of each animal using a 6 mm biopsy punch. Immediately after the extraction, the tissues were stored in − 80 °C, until further analysis. The animals were sacrificed by anesthetic deepening. We used 5-6 animals per group. For the immunoblot experiments, the tissues were homogenized in Radioimmunoprecipitation assay buffer (150 mM NaCl, 50 mM Tris, 5 mM EDTA, 1% Triton X-100, 0.5% sodium deoxycholate, 0,1% sodium dodecyl sulfate, and supplemented with protease inhibitors). Insoluble materials were removed by centrifugation 11,000 rpm for 40 min at 4 °C, and the supernatant was used for protein quantification by the biuret reagent protein assay. Laemmli buffer (0.5 M Tris, 30% glycerol, 10% SDS, 0.6 M DTT, 0.012 bromophenol blue) was added to the samples. One hundred micrograms of proteins were separated by SDS-PAGE and transferred to nitrocellulose membranes (Bio-Rad) using a Trans-Blot SD Semi-Dry Transfer Cell (Bio-Rad) for 1 h at 17 V (constant) in buffer containing methanol and SDS. Blots were blocked in a 5% skimmed milk powder solution in TBST (1 × TBS and 0.1% Tween 20) for 2 h at RT, washed with TBST, and incubated with the primary antibodies for 24 h at 4 °C. The primary antibodies used were anti-pCaMKII (Abcam, ab32678) and anti-Fyn3 (Santa Cruz, sc-16). HRP-coupled secondary antibodies (1:5000, Thermo Scientific) were used for detection of the conjugate by chemiluminescence and visualization by exposure to an Image Quant LAS4000 (GE Healthcare, Life Sciences). Anti-β-actin (Abcam, ab8227) was used as a loading control. The intensities of the bands were digitally determined by densitometry, using Image J software (National Institutes of Health).
Immunohistochemistry. Skin expressions of Grin1, Grin2a, Grin2b, and Grin2c were identified by immunohistochemical staining. Immunohistochemistry was performed using the skin samples (n = 5). Tissue samples