Establishment of 3D Spheroid Human Dermal Papilla Cells as an Effective Model for Hair Growth Screening Compounds Compared with the 2D System: An Example of Minoxidil and 3,4,5-Tri-O-caffeoylquinic acid (TCQA)

Dermal papilla cells (DPc) is an important element in studying the hair follicle (HF) niche. The human hair follicle dermal papilla cells (HFDPC) are widely used as an in vitro model to study hair growth related research. These cells are usually grown in 2D culture, nevertheless, this system did not show ecient therapeutic effect on HF regeneration and growth, and key differences were observed between cell activity in vitro and in vivo. Objective Recent studies have showed that HFDPC grown in 3D hanging spheroids is more morphologically akin to intact DPc microenvironment. This current study showed that the 3D model is applicable to the commercial cell line with new insights on its variability by comparing to previous studies of gene signature restored by 3D culture. Our data demonstrated in in vitro can inuence not only hair growth-related pathways but also immune system -related compared to 2D cell monolayer. Furthermore, we compared the of and of in and acid (TCQA) using microarray analysis. gene signature of 3D culture in comparison to 2D, we investigate whether this commercial cell line is able to successfully express the gene signatures of a 3D spheroid and represent a closer to intact DP hair growth metabolism. Our data demonstrated that the commercial cell line grown in 3D in vitro model can inuence not only hair growth related pathways but also immune system related pathways compared to 2D cell monolayer. This study aim to establish HFDPC 3D model as the adequate system to represent DP microenvironment. Moreover, we report the molecular mechanism behind minoxidil and TCQA hair growth promotion effect in 3D and 2D culture model, used to further establish this technique as a validate model for hair related research.


Introduction
It has been known for a long time that the hair follicle (HF) undergoes continuous cycling that is categorized into three stages, expansion (anagen), apoptosis-driven regression (catagen), and inactivity (telogen) [1]. The HF contains different cell population distinct in their function, molecular mechanism, and origin that throughout adult life are revived from niche-resided stem cells located in the bulge [2].The growth cycle and cells regeneration is mainly controlled by a group of specialised broblastic cells of mesenchymal origin, known as the dermal papilla cells (DPc) [3].
The longest stage within the hair cycle in human, is the anagen stage which takes 2 to 5 years. As for the second phase, it occurs from few days to few weeks, while the telogen usually takes 3 months, for the represent more accurately the microenvironment surrounding the tissue [14]. In this system, the cells present morphological and physiological difference re ecting in vivo cellular responses, comparing to 2D cultured cells [15,16]. Interestingly, 3D culture system is gaining insight not only because it affects the dimensional organization of the cell allowing the communication with neighbouring cells, but also it in uences the signal transduction, gene expression, and cellular behaviour [17,18]. Based on previous studies, 3D DPc cultured in spheroid system using the hanging drop technique, may restore partially hair inductivity potential of intact DPc lost in 2D culture as it encourages the formation of condensate-like structure [9,19]. However, spheroid cell culture systems may present limitations that are a rather not strong interaction with the extracellular environment, and, an inconsistency in the expression of selected 3D markers and hair inductive potential of 3D DPc cultures was found [8,9].
Alopecia is an exceedingly prevalent hair loss condition affecting many people worldwide and hence attracts demands for treatment [20]. Only two drugs have approved by the Food and Drug Administration (FDA) as a treatment of hair loss, one of which is minoxidil. Our group have previously published that the polyphenolic compound, 3,4,5-tri-O-caffeoylquinic acid (TCQA) signi cantly activated Wnt/β-catenin signalling in vivo and in HFDPC for the promotion of hair growth cycle and pigmentation [21,22]. Next, TCQA was found to promote the differentiation of 3D cultured human amnion epithelial stem cells while increasing ECM markers [23]. Consequently, high demand for using target compounds known to promote hair growth and ECM related markers to overcome the limitations of 3D spheroid DPc culture.
Hence in this paper, we compared the expression of signalling molecules and markers of HFDPC commercial cell line in untreated 3D and 2D cell cultures using microarray analysis. By evaluating the gene signature of 3D culture in comparison to 2D, we investigate whether this commercial cell line is able to successfully express the gene signatures of a 3D spheroid and represent a closer to intact DP hair growth metabolism. Our data demonstrated that the commercial cell line grown in 3D in vitro model can in uence not only hair growth related pathways but also immune system related pathways compared to 2D cell monolayer. This study aim to establish HFDPC 3D model as the adequate system to represent DP microenvironment. Moreover, we report the molecular mechanism behind minoxidil and TCQA hair growth promotion effect in 3D and 2D culture model, used to further establish this technique as a validate model for hair related research.

Samples Preparation
Minoxidil was bought from Tokyo Chemical Industry (Tokyo, Japan).TCQA was synthesized with 97% purity. The two sample were prepared in 70% ethanol prior to dilution in the growth medium for in vitro assay.
Cells and cell culture HFDPC were purchased from Cell Application Inc. (Tokyo, Japan). The cells were kept in appropriate medium (Toyobo, Osaka, Japan) supplemented with growth factors (Toyobo, Osaka, Japan). For the 2D model, HFDPCs were cultured in dermal papilla growth medium. Prior to seeding or passage, the plate were coated with collagen solution (Toyobo, Osaka, Japan) for 2 h at room temperature (RT). The cells were monitored under sterile conditions at 37 °C in a humidi ed atmosphere of 5% CO 2 .
Establishment and maintenance of 3D culture.
HFDPC were cultured in 3D spheroids system [9]. Firstly, cells were cultured without antibiotics in papilla cell growth medium. The seeded cells were trypsinized, 5 ml of the culture medium was added, the cells were centrifuged, and then resuspended at 300 cells per µl, 48 h post passage. The 3D culture was established using hanging drops methods on the lid of a 100-mm petri dish. Each 10 µl hanging drop contains approximately 3,000 cells, phosphate-buffered saline (PBS) was placed to completely cover the base of the dishes. Two group were established, the control group (untreated cells) were the drop were formed from the culture medium containing the cells, and the treated groups with 0.1 and 10 µM, minoxidil and TCQA, respectively. The formation of spheroid using hanging drops method was established after 48 h, 3D spheroid DPc were then used in the following experiments.

RNA extraction
The RNA was extracted from HFDPC cultured in 3D and 2D systems. For the 2D system, HFDPC were seeded overnight (5×10 5 cells per 100-mm dish), then the papilla medium was renewed with the treatment (0, 0.1 µM minoxidil, and 10 µM TCQA). After 48 h, the medium was discarded, a washing with cold PBS was assessed following by an Isogen extraction (Nippon Gene, Tokyo, Japan) according to the manufacturer's directions. Also 3D hanging drops untreated and treated with 0.1 µM minoxidil, were subjected to the same procedure. For both systems, the RNA concentration was assessed using a NanoDrop 2000 spectrophotometer.

DNA microarray analysis
The extracted RNA samples extracted from 2D and 3D system were prepared for microarray analysis using GeneChip® 3′ Expression Arrays and 3′ IVT PLUS Reagent Kit (Affymetrix Inc, CA, United States). In brief, (250 ng) was used to generate ampli ed and biotinylated aRNA following the manufacturer protocol. The array strips (HG-U219) were hybridized for 16 h in a 45°C station, then washed, stained, and imaged in the GeneAtlas Fluidics and Imaging Station.
For the analysis, rstly the Expression Console Software was used for the raw data normalization, allowing the selection of genes with a fold change > 1.4 (in linear space) and P-value ≤0.05 (One-way between-subject ANOVA) to be designed as differentially expressed genes (DEGs). Further analysis was carried out using Transcriptome Analysis Console (TAC) version 4, Database for Annotation, Visualization, and Integrated Discovery (DAVID), Gene Set Enrichment Analysis (GSEA), ExAtlas: gene expression, and Meta-analysis [24][25][26].

Western blot
The protein were extracted from HFDPCs cultured in 2D and system and treated with or without 0.1 µM minoxidil after 48 h as explained above. Total protein extraction was achieved using radioimmunoprecipitation assay (RIPA) buffer (SIGMA, Saint Louis, USA) and protease inhibitor, and quanti ed with the 2-D Quant kit (GE Healthcare, Chicago, USA).

Statistical analysis
Results were expressed as mean ± standard deviation (SD). Statistical analysis was performed using Student's t-test when comparing two value sets For microarray analysis, ANOVA (One-way betweensubject ANOVA unpaired) was performed to assess the level of signi cance between treated groups.

Results
Gene pro ling analysis of HFDPCs cultured in 3D hanging compared with 2D culture To determine whether there is successful spheroid formation, we compared the consistency of spheroid size and shape formation using gene expression analysis. To understand the integrative response of 3D culture, we assigned the cut off for the differentially expressed genes (DEGs) to be >2.0 or <2.0 (linear fold change). We found 1933 genes between 2D and 3D samples to be consistently differentially expressed. Of these genes, 823 genes were upregulated and 1110 genes were downregulated as s represented in the volcano plots that show the distribution of fold change and signi cance levels of the DEGs ( Figure 1A).
Gene set enrichment analysis of 3D control compared with 2D control showed the upregulation of several prior de ned hallmark gene sets associated with collagen regulation (GO:0005587, GO:0005518, GO:0062023), cell adhesion (GO:0007155), endothelial cell proliferation (GO: 0001938), and cellular response to UV (GO:0034644) ( Figure 1B). Interestingly, immune system signi cant gene were regulated in 3D system compared with 2D ( Figure 1C).
The results were further validated to show the formation of the 3D structure and the cell adhesion of HFDPCs. The expression of Versican (VCAN) and Neural cell adhesion molecule (NCAM) involved in cell adhesion were checked. Figure 1D and E displayed the upregulation of the expression of these molecules on the gene and protein level in HFDPCs cultured in 3D compared with 2D after 48 h.
Moreover the highly regulated genes in 3D spheroid were classi ed in Table 1 to summarize the nding data. These results showed that 3D system further enhance the cell-cell and cell-ECM interactions, and confer a better defence response and a more resistant immune system.

Comparison of the DEGs in 3D and 2D cultured HFDPCs treated with Minoxidil
The cells were cultured in 3D and 2D system and treated with 0 and 0.1 µM for 48 h, then microarray analysis was performed and the DEGs were to be >1.3 or <-1.3 (linear fold change). Results revealed the regulation of 594 out of which 225 are upregulated while 369 downregulated in case of 2D minoxidil vs 2D control (Supp Figure 2A). In case of 3D minoxidil vs 3D control, 5160 are upregulated and 6316 are downregulated (Supp Figure 2B). Then the DEGs were classi ed into different categories according to fold-change. For the 3D system, more genes with 2-fold change and-2-fold change were upregulated and downregulated, respectively (Supp Figure 2C). The top 100 DEGs in the treated groups: 3D minoxidil, 3D control, 2D minoxidil, and 2D control, were subjected to hierarchical clustering. Results demonstrated that the top genes displays different expression between 3D minoxidil vs 3D control and 2D minoxidil vs 2D control (Supp Figure 2D).
Heat map was created to compare 3D minoxidil with 2D minoxidil. We observed the enhancement of the expression of genes signi cant for cell-cell adhesion (GO: 0098609), Wnt signalling (GO: 0060071), positive regulation of canonical Wnt signalling (GO: 0090263) tricarboxylic acid cycle (GO: 0006099), cell cycle arrest (GO: 0007050) ( Figure 2A). On the other hand, 3D minoxidil treatment inhibited genes linked with interferon production (GO: 0032728), NF-kappaB signalling (GO: 0043123), and apoptotic signalling (GO: 1900740) ( Figure 2A). The molecular function and the Kegg pathway analysis of the DEG showed that 3D minoxidil treated upregulated ATP binding pathways, cell cycle, and fatty acid metabolism compared with 2D minoxidil treatment ( Figure 2B).These current data demonstrated that minoxidil stimulated hair growth associated genes, ATP and collagen binding but the effect was higher in 3D cultured cells.
Moreover, to prove that the upregulation of hair growth-, cell adhesion-, and collagen binding-associated genes was due to minoxidil treatment not only to the 3D structure, a comparison of the DEGs between 3D control vs 2D control and 3D minoxidil vs 2D minoxidil was assessed. Supp Fig 2A illustrated the obtained data, and the common upregulated genes between these two set are only 12, whereas the 3D minoxidil upregulated genes more linked with cell cycle arrest, in ammation response compared with the 2D minoxidil. For the downregulated genes between the previously stated set, the common genes are 11 and 3D minoxidil signi cantly decreased the expression of genes linked with in ammation and cell division (Supp Fig 2B).
These results showed minoxidil treatment to 3D cultured HFDPCs highly affected hair growth associated pathway and markers, and cell cycle arrest signalling while conferring an anti-in ammatory activity.
3D cultured HFDPCs treated with Minoxidil enhanced the gene and protein expression of hair growth markers Firstly, the gene expression of NCAM and VCAN was assessed upon minoxidil treatment in 3D cultured HFDPCs, because of their function as adhesion molecule essential to validate the formation of 3D spheroid, and also the implication in keratinocytes differentiation. Results revealed that minoxidil enhanced the expression of these markers con rming that the treatment did not affect the establishment of the 3D structure ( Figure 3A).
The highly affected pathway by minoxidil treatment weather in 2D or 3D culture system are the Wnt/βcatenin and FGF pathways, and dermal papilla marker alkaline phosphatase or ALP. The gene and protein expressions of these markers were checked and results showed the upregulation upon treatment.
We take the example of β-catenin, where minoxidil treatment enhanced the gene expression up to 1.8-and 2.8-fold change, and the protein expression up to 1.4-and 2.2-fold change, respectively, in the case of 2D and 3D structure ( Figure 3B and C). We observed the same tendency in case of FGF-related molecules and ALP. These data demonstrated that minoxidil treatment in the 3D system had a higher effect in the regulation of hair growth in dermal papilla cells comparing with the 2D model.

Characteristics of 2D and 3D cultured HFDPC treated with TCQA
In this study we used another compound to test the e cacy of 3D system, TCQA was previously reported to stimulate the anagen phase and the hair growth cycle in C3H mice and in HFDPCs [22]. After 48 h, 2D cultured HFDPCs grew in the conventional monolayer format, in contrary, the 3D cultured HFDPC exhibited intercellular aggregation ( Figure 4A).
To elucidate the characteristics of the cells cultured in 2D and 3D and simultaneously treated with TCQA, gene pro ling analysis was assessed. The result showed that genes sets namely Wnt signalling pathway, epithelial-mesenchymal cell differentiation, collagen, cell-cell adhesion, broblast proliferation, were signi cantly upregulated upon TCQA treatment ( Figure 4B and C). This is to further determine the extent of gene expression difference between the two used cell culture methods, important top upregulated and downregulated genes were picked and classi ed in Tables. Wnt signalling and anagen induction genes were modulated speci cally CTNNB1, Wnt11, WLS, and ALPL, along with FGF1 and FGF5 (Supp Table 3, 4 and 5). Others factors involved in HF regeneration and hair cycle regulation such us SOX4, NRAS, BMPR-1A and BMP5 was as well enhanced upon treatment. It is worth to mention that VCAN a marker of 3D spheroid formation and broblast differentiation was upregulated in 2D and 3D system. While this is true for both 2D and 3D culture, the genes in the TCQA-treated 3D culture exhibited a more than double fold-change in comparison to the fold change in the 2D monolayer cultured cells (Supp Table 3 and 4). CTNNB1 expression was stimulated up to 1.62-fold change in TCQA-treated 2D HFDPCs, while in TCQAtreated 3D HFDPCs, the expression was up to 43.57-fold change. These results showed that the effect of TCQA on stimulating hair growth-related genes was more pronounced and signi cantly higher in the 3D spheroid system compared with the 2D.
Furthermore, genes involved in protein binding, PI3K-AKT signalling pathway, DNA damage, alternative splicing, phosphoprotein and negative regulators of cell functions were downregulated ( Figure 4D and E). Supp Table 3 and 5 illustrated the repression of genes linked with Wnt/ β-catenin inhibition, telogen phase regulation, anagen phase delay, and collagen breakdown for both 2D and 3D cultures (TCF3, LDB3, AXIN2, GSK3B, and EGR1). We noticed as well the pattern, that the fold-change was further decreased in case of 3D cultured compared with the 2D upon TCQA treatment. In addition, the DEGs between 3D TCQA vs 3D control with 2D TCQA vs 2D control were compared and Veen diagram was created for both upregulated and downregulated genes ( Figure 4F and G). Results revealed that few genes were common between the two compared sets and the top DEGs were summarized in Supp Table 3 to 5.
The results of microarray in this study were compared to the in vivo microarray result from the back skin of eight-weeks-old C3H male mice treated topically for one month with TCQA upon shaving [22]. Table 5 displays the fold-change comparison of 2D-, 3D TCQA-treated HFDPCs, and in vivo microarray result. It showed that 3D cultured TCQA-treated HFDPC exhibited an extraordinarily high level of fold change compared to 2D and in vivo data mainly linked to canonical Wnt signalling ( Table 4).
Some of the DEGs were validated by Quantitative RT-PCR. Result showed that the gene expression of 3D culture structure markers: NCAM and VCAN were enhanced upon TCQA treatment ( Figure 5A). This suggests that TCQA's ability to express these genes higher in 3D, is vividly due to better cell-cell interaction in between the dermal spheroids. CTNNB1, ALPL, and FGF1 gene expression is signi cantly expressed by TCQA both in 2D-and 3D-treated HFDPCs, however, comparison showed that 3D-HFDPCtreated with TCQA expressed the said gene signi cantly higher than 2D ( Figure 5B). Thus, these data revealed that TCQA enhanced hair growth-related pathways, decreased telogen-and Wnt repressionsigni cant genes, and this effect is more evident in 3D culture systems.

Discussion
The HF neogenesis is considered as an attractive target for regenerative medicine purposes, in part due to, the strong foundation of knowledge underpinning the HF biology, and widespread clinical interest [9]. One of the approaches to induce the anagen phase of hair growth cycle is the replenishment of DPs cells with in vitro 2D-cultured DPc, yet, with this technique the cells will lose their hair-inducing capacity over time because they are cultured on a at, plastic surface [27]. Numerous studies have showed that 2Dcultred DPc were unable to induce anagen-telogen transition upon in situ implantation after 6 passages, as these cells needs to agglomerate in the HF to be highly active [28][29][30]. To overcome this problem, 3D spheroid culture is considered to be an e cient tool to regain DPc inductive capacity which may enable them to induce de novo HF in human skin as these cells requires cross-talk with the surrounding environment [5,31]. Nevertheless, an extensive knowledge of the molecular mechanisms governing the regenerative process is necessary. In this study, global gene pro ling analysis was performed to compare 3D spheroid with 2D-cultured HFDPCs to determine the signature regulating the inductive phenotype of DPc. In addition, we compared the effect of alopecia approved drug minoxidil, and the polyphenolic compound TCQA recently published and found to enhance hair growth in vivo and in vitro, in 3D and 2D culture systems.
Generally, cell growth within a spheroid structure, facilitate DPc cell aggregation and cohesion which the loss of these two functions may bring, papilla miniaturization, disturbance of adult HF morphogenesis, and eventually hair loss [32,33]. Our data con rmed as well what was previously discussed that in case of 3D spheroid DPc, the GO of collagen regulation, cell adhesion, ECM, and cell proliferation were positively stimulated (Figure 1). The ECM plays a major role for the cell growth and adhesion, and cell-cell interaction. A common physiological feature of DPc is their close contact with the ECM, which is proved to be further enhanced in 3D spheroid culture [34,35]. For in vitro studies involving DPc, NCAM and VCAN has been used as major indicators for the inductive capabilities of these cells, accordingly the decrease in their expression correlates with the decline of DPc inductive capacities [36,37]. NCAM is involved in dermal condensation and hair induction, whereas VCAN is not only implicated in matrix assembly and structure and in cell adhesion, but also in HF development and cycling [38][39][40]. Our study has showed that in 3D spheroids, the expression of these DPc markers was higher compared with 2D culture system ( Figure 1).
Putting all together, restoring human DPc inductivity via the arti cial recreation of the cells 3D architecture and intercellular contacts using 3D spheroid cultures, considered to be an e cient method [41]. Nevertheless, while this technique allowed a partial restoration of DPc native signature and transcriptome, it have showed a limitation in human-human recombinant assay as the 3D spheroid could only stimulate the hair induction by 15% [9]. Thus, the need to address new strategies in order to properly recover the inductive capacities of DPc. For this purpose, using compounds and biomolecules able to activate hair growth related pathways in 3D spheroids system, can be considered as a capable approach to further enhance DPc inductive properties [42,43]. Therefore, in this current research, minoxidil the FDA approved drug against alopecia and TCQA a polyphenolic compound reported to stimulate hair growth were tested in 3D spheroids systems. Minoxidil affects hair growth cycle whether in animal models or human clinical trials by, shortening the length of the telogen phase, increasing the rate of DNA synthesis in anagen bulbs causing an early entry to the anagen phase [44][45][46]. An experimental study with 2D cultured DPc have reported that minoxidil prolonged the anagen phase by inducing prostaglandin E2 and β-catenin activity, and stimulating follicular proliferation and differentiation [47,48]. To the best of our knowledge, minoxidil mechanism of action in 3D spheroid DPc is not yet explored. Here, we reported that minoxidil upregulated of hair growth-, cell adhesion-, ATP binding-and collagen-associated in 3D system compared with 2D ( Figure 2). The proliferation of DPc during the anagen phase is usually followed by an increase in the ATP content resulting in the expansion of cell size which will directly affect the differentiation of keratinocytes into the HS [49,50]. The extracellular matrix category included several collagens in addition to other transcripts that are expressed in the intact DP [51], which were found to be more present in 3D minoxidil-treated spheroid compared with the 2D (Figure 2, Table 2). Next, 3D minoxidil downregulated genes repressing Wnt/β-catenin pathway (AXIN2) and telogen marker NFATC1 (Table 3). Axin2 is involved in β-catenin phosphorylation, degradation and non-translocation to the nucleus, leading to a shorten anagen phase in the HF. Nfatc1 plays a role in HF stem cells quiescence which is explain it high expression during the telogen phase , and the decreases during the transition to anagen favoring stem cells differentiation [52][53][54]. Moreover, gene expression analysis data was supported by the increase of the gene expression of NCAM and VCAN, and the gene and the related proteins expression of hair growth makers: ALPL (ALP), FGF1 (FGF1), and β-catenin (CTNNB1) in the 3D spheroid compared with the 2D (Figure 3). As mentioned earlier in this paper, several molecular markers have been used as an indicator of hair inductiveness, including VCAN and NCAM. Additionally, numerous studies have discussed that increasing VCAN expression in DPc enable HF induction and stimulate growth cycle initiation, suggesting its expression correlates with hair inductivity [55,56]. ALP is a known marker of DPc, overexpressed during the growth phase of the hair cycle while FGF1 is required in the promotion of the growth phase in C57BL/6 mice [7,57,58]. The role of β-catenin is well documented and established in the regulation of hair growth, as it is highly expressed during anagen phase in the DPc, and the absence of which induces a premature catagen phase [5,59,60]. Putting all together, the expression of hair growth markers was more pronounced in 3D model compared with 2D one, and minoxidil treatment further enhanced the e cacy of the 3D spheroid system in DPc.
On the other hand, TCQA was previously reported to have the ability to promote hair growth in CH3 mice and in HFDPCs via the activation of β-catenin [22] .Therefore, TCQA's capability to activate Wnt/betacatenin signalling to induce hair growth in 3D spheroid DPc was investigated. TCQA-treated 3D culture exhibited a higher expression of gene sets related to Wnt/beta-catenin pathway, cell proliferation and collagen compared to the 2D culture (Figure 4 and 5). Canonical Wnt pathway serves as the central regulator of hair morphogenesis and cycling, and different type of Wnt have various roles in the HF regeneration [61]. Additionally, 3D-TCQA-treated cells highly expressed some collagen-related genes like COL3A1 (Supp Table 4). This a type III collagen, also known as brillar collagen found in skin mainly associated with the papillary dermis, produced by intact DPc, and responsible for HF development [62]. The comparison of the genes between, TCQA treated-DPc in 2D and 3D system, with TCQA treated mice skin previously published [22], listed on Table 4 clearly showed that the fold change of the selected genes greatly differs. Fold change in the TCQA-treated 3C culture is more than double for some genes compared to the 2D monolayer culture, and surprisingly, it showed that the fold change for the 3D system is higher than in vivo model. Taken altogether, despite TCQA's positive effect on 2D cultured-DPc, the full potential of the compound when studied in culture was highly compromised. Cellular aggregation mimics the microenvironment of dermal papilla, thus restoring the cues that happened on its normal niche. Therefore, the effect of this compounds was higher in 3D spheroid DPc compared with 2D system.
Overall, we have demonstrated that the partial reprogramming of HFDPC by spheroid formation with the addition of hair growth stimulating compounds can restore further restore the inductive capacities of DPc. These ndings represent a signi cant leap for both, 3D spheroid DPc culture method and for minoxidil and TCQA as hair promoting agents. We would like to thank Dr. Kozo Sato from Synthetic Organic Chemistry Laboratories, the FUJIFILM Corporation (Kanagawa, Japan) for providing synthesized TCQA.

Competing Interests
The authors declare that they have no nancial or commercial con icts of interest.
Author's contribution MB and AKO performed most of the experiment data analysis and interpretation, prepared the gures, and wrote the manuscript. MKS prepared the sample, participated it in the data curation and validation, edited and revised the manuscript. FF was involved in the microarray analysis, edited and revised the manuscript. HI conceived and supervised the study, managed funding and resources for the study and reviewed the manuscript. All authors read and approved the nal manuscript      Results represent the mean ± SD of three independent experiments. * P ≤0.05, ** P ≤0.01.   protein expression after 48 h. Band intensities was assessed using LICOR system. Results represent the mean ± SD of three independent experiments. * P ≤0.05, ** P ≤0.01.  Results represent the mean ± SD of three independent experiments. The Student's t-test was used to compare the mean values of two groups, * P ≤0.05, ** P ≤0.01, *** P ≤0.001.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.