Functional Melanocytes Obtained From Rabbit Hair Follicle Outer Root Sheath

Abstract


Abstract Background
Melanocytes have been cultivated from the outer root sheath of hair follicles for two decades. So far, these isolation and culturing procedures have been carried out from human and mouse follicles. In this study, we have translated the established procedure for obtaining melanocytes from stem cells and precursors of human hair follicle to a rabbit species in order to isolate and cultivate rabbit melanocytes from whisker follicle outer root sheath (ORS), hereby named rMORS, and compare them to rabbit epidermal rabbit melanocytes (rEMs).

Results
The rMORS were isolated and cultured by extracting whisker follicles from enzymatically digested skin and submitting them to the air-liquid interface hypoxic conditions. The cells were allowed to migrate from the follicle ORS onto the nylon mesh of Transwell inserts, collected and subcultured in melanocyte culture medium for 11 passages. From early passages on, the rMORS cells displayed typical melanocyte characteristics comparable to those of rEM used as experimental control. Melanocyte features were assessed on morphological level by the means of microscopy, functional and gene expression level.

Conclusions
We concluded that the method of isolating and culturing rMORS e ciently and reproducibly corresponded to the prior art method established in human follicle, yielding melanocytes comparable to the rabbit epidermal melanocytes in all assessed features. This method is not crucial in terms of regenerative therapy, it rather paves way for a non-invasive ex vivo sampling of hair and it may give way to deeper insights into rabbit follicle biology or to postulating useful species-speci c in vivo experimental models.

Background
Methodology of isolating and culturing stem cells, precursors, keratinocytes and melanocytes from the outer root sheath of hair follicle (ORS) has been amid the foci of stem cell-and regenerative biology for two decades [1]. The ORS of hair follicle is a cellular envelope that surrounds the hair shaft [2] and the inner root sheath (IRS). This mini-organ contains a heterogeneous pool of cells, starting from pluripotent neural crest-like stem cells (NCSC-like) [3,4], Lgr6+ stem cells [5], ranging over neural stem cells [6], mesenchymal stem cells [7][8][9][10], keratinocyte stem cells [11], skin cell precursors [12], to the differentiated keratinocytes and melanocytes [13].
Initial protocols for harvesting the ORS cell pool were based on isolating follicles from human or murine skin rests. To this purpose, skin specimen were loosened by digestion, epidermis was separated from dermis and the follicles were extracted by pooling, which included minimal loss of the ORS. Similar protocols were used to obtain follicles for purposes of culturing NCSC-like stem cells and other types present in the ORS cell pool [14][15][16]. All of those methods were highly useful in gaining an insight into human and mouse hair follicle biology and they helped establish a critical mass for understanding the ORS cell pool and further channeling it towards a therapeutic use. In addition, all of the aforementioned methods were implicitly invasive, since they relied on a specimen of excised skin.
The regenerative function of the ORS [17] further inspired an experimental pipeline that lead to optimizing harvesting methods for purposes of therapeutic use. Major change in the approach of collecting the ORS cell pool was reached by developing non-invasively-based methods for obtaining follicles. Here, plucking was introduced as a routine sampling method. This helped establish procedures for culturing keratinocytes from the ORS [18][19][20], primarily directed towards a treatment of non-healing wounds [21,22]. This non-invasive harvesting step eliminated the issues of pain, infections and scaring that burdened the preceding biopsy-based isolation procedures. The ORS keratinocyte culture was stepped up to the ORS melanocyte culture for purposes of culturing autologous melanocytes and eventually treating pigmentation disorders [23][24][25][26].
Aforementioned clinical potentials and applications of hair follicle requires further investigation and proof-of-concept on small and medium size animals. Rabbit has been widely used as a model animal of medium size for pre-clinical researches [27], with advantages of ease of handling, cost-effective, rapid breeding [28], syngeneic Syngeneic and xenograft models for autologous therapies [29].
In this study, we utilized the method for establishing ex vivo ORS melanocyte culture used in human to culture melanocytes from the rabbit whisker follicle ORS. The base procedure relies on innate migrating abilities of stem cells, melanocyte precursors and melanocytes cultured in the air-medium-interface conditions, severed from the proximal follicle part, differentially trypsinized in order to eliminate broblasts and keratinocytes [23,25]. This should provide a base for a) non-invasive harvesting of the ORS cell pool in rabbits b) easy culture and expansion of the ORS cells and c) gaining further insight into rabbit follicle as a medium size model animal.

Results
Isolation and primary culture of rMORS and rEM Figure 1A shows a whisker hair follicle seeded on the Transwell membrane after 21 days of cultivation in the liquid-air interface. Cells migrated out of the ORS tissue and formed a con uent cell monolayer on the porous nylon membrane of a Transwell insert within 28 days of cultivation. After harvesting the ORS cells using Trypsin/EDTA and subcultured onto normal adherent culture, the cells attached to the ask and proliferated. Within 6 weeks, the air-medium interface culture and the subsequent plastic-adherent culture reached the third passage and yielded 3,521,250±1,194,248 cells of functional hair follicle melanocytes in passage 4 out of 10 rabbit whisker hair follicles in. By means of differential trypsinization, functional melanotic melanocytes were successfully selected, further expanded and differentiated into a pure culture of melanocytes ( Figure 1C). Subcultured melanocytes displayed two or more dendrites.
The rEM were directly isolated from the dorsal epidermis, seeded on the polystyrene surface of cell culture asks and left to adhere and proliferate ( Figure 1B). Within 3 weeks, 6,378,333±2,740,995 cells in passage 2 with functional melanotic features were obtained. As shown in Figure 1D, large amount of perinuclear granules in the bi-dendritic rEM were distributed from perinucleum towards dendrites, typical for melanocytes in the process of producing melanosomes with melanin content.

Melanocyte functions of rMORS and rEM
The presence of functional melanosomes in both rMORS and rEM was identi ed using von Kossa staining and L-DOPA stimulation. After stimulation by UV light, the calcium ions in melanin were replaced by silver ions in sliver nitrate, resulting the visible metallic silver precipitates, observed in the regions of perinuclear area and dendrites in both rMORS and rEM ( Figure 2A). After stimulation by L-DOPA, melanin production in rMORS and rEM was intensively increased, hereby making it easy to clearly visualize pigmented cells as shown in Figure 2B.

Melanocyte gene marker expression in rMORS and rEM
In order to further characterize the gene expression pro le of rMORS and rEM, we performed a quantitative analysis of mRNA expression levels of NES, PAX3, MITF, TYR, CKIT and PMEL at P3, P5, P7 and P9 as shown in Figure 3. Overall, rMORS showed different expression patterns than rEM, and revealed signi cantly higher expression of NES, PAX3, TYR and CKIT than rEM in all studied passages. In both cell types, NES expression declined from P3 to P7 and increased in P11. The expression dynamics of of TYR correlated with those of tyrosinase activity shown in Figure 2C.
Expressions of PAX3, MITF, CKIT and PMEL showed similar patterns in rMORS and rEM. For rMORS, the highest levels of gene expression occurred in P5, in terms of gene PAX3, MITF, CKIT. Whereas the maximal levels of MITF, CKIT and PMEL were shown at P7 in rEM. For PMEL, both rMORS and rEM peaked at P7 in terms of gene expression level.

Discussion
Method of culturing melanocytes from the ORS of human follicles has been successfully adopted and applied to plucked rabbit whisker hair follicles. In a nutshell, the procedure relies on migration of stem cells, melanocyte precursors and melanocytes from the ORS onto the mesh of a Transwell insert, their further division and differentiation until the endpoint melanocyte. Melanocytes were herewith cultured with a reproducible e ciency and they displayed melanotic features comparable to those of epidermal rabbit melanocytes.
Migrating ability of stem cells, melanocyte precursors and melanocytes appears to be crucial for leaving the ORS and populating the nylon mesh of the Transwell insert. Once the follicles is placed in the airmedium-interface conditions, the follicle is covered by a thin wet lm of medium maintained by surface tension forces, which protects it from desiccation and allows e cient gas diffusion. Hereby, the ORS is e ciently exposed to the hypoxic mixture from the upper side and nutritionally supplied by DLM medium from below.
In its starting phase, the initial ORS-generated culture is a co-culture of melanocytes, keratinocytes and broblasts that can be separated into pure single-type cultures [23]. To obtain a pure culture of melanocytes, it was necessary to eliminate keratinocytes and broblasts. Withdrawal of the proximal part eliminated dermal carry-over, which presents the main source of broblasts in the ORS-generated culture [30]. Further, differential trypsinization enabled an early detachment of melanocytes from the cell culture plastic, whereas the keratinocytes and the remaining broblasts remained attached, hereby separated into two phases.
Tyrosinase activity and melanin content, both with and without L-DOPA substrate, were comparable in rMORS and rEMs. They also displayed opposite kinetics in both rMORS and rEMs, which is often the case in human melanocytes [30,31], since Tyrosinase activity is necessary for production of melanin and it precedes the augmentation of melanin content. Therefore, at the same time point of sampling, the tyrosinase and the melanin content are unlikely to display a synchronized trend, be it distinctly enhancing or reducing. To ensure coordinated sampling, the cells were harvested in matching passages at the same time window ±10 minutes. This ensured sampling at precisely the same time point, hereby eliminating time delay that could bring about a shifted display of Tyr activity versus melanin content. The cells sampled at such time points were divided in two samples and one was used for each analysis.
The gene expression pattern is also typical for melanocytes and concordant to the kinetics of the measured melanin content. Nestin, a marker of neural stem cells and the neuroectodermal lineage that precedes melanocyte precursors [32,33] showed an increasing kinetics in rMORS and generally higher expression in rMORS compared to rEMs, which hints towards a developmentally earlier status of rMORS than of fully differentiated rEMs. Early melanocyte markers c-KIT, PAX3 and TYR displayed a prominently higher expression in rMORS than in rEMs in all assessed passages, indicating an intensive activity of early melanotic program in rMORS. The downstream effectors MITF and PMEL (gp100) gene activity in rMORS is comparable to that of rEMs or lower, which corroborates earlier developmental status of rMORS, since those genes mark endpoints of melanocyte differentiation and they are likely to be higher in fully differentiated melanotic melanocytes. The gene activity of the markers depicts the rMORS status as one of partialy melanotic melanocytes, which is in concordance with comparable or lower melanin content and tyrosinase activity compared to rEMs as well as with earlier reports of melanocyte marker dynamics in melanocytes from human follicle [30,31].
The procedure was at rst established by using follicles isolated from skin rests available from rabbit sacri ced in scope of other experiments, as it had been done in the initial establishing phases of the preceding methods for isolating human follicles. Once established in the course of this study, the aforementioned method can be fully translated to plucking in vivo in minimal numbers (about a dozen of whiskers) without the necessity of excising a skin specimen from a living animal. By doing this, stress, donor-site morbidity and the risk of infection caused by skin biopsy ex vivo would be omitted. Such noninvasively cultured follicles would open further possibilities to explore biology of the rabbit hair follicle with minimal-to-none level of invasiveness and without sacri cing the animal. The method presented herein is of little or no interest in terms of therapeutic use in rabbit. Nevertheless, the future possibility of non-invasive harvesting is a useful legacy of the methods developed in human follicles and paves the way for analogue animal-friendly procedures.

Conclusions
This study provides a translation of isolating melanocytes from hair follicles ORS in rabbit. It will be useful for obtaining ex vivo material for purposes of exploring rabbit hair follicle biology and help gain further insight into evolutionary interspecies homologies of the placode structures, as well as contribute to further medium size animal models for translational / pre-clinical applications.

Methods
Rabbit tissues were kindly provided by Paul Flechsig Institute of Brain Research, University of Leipzig. Rabbit skin and whisker pad tissue was collected from 12-month-old wild-type chinchilla bastard rabbits that were sacri ced within another experimental set up (n=3).

Isolating rabbit epidermal melanocytes (rEM)
Rabbit skin tissue was harvested from dorsal skin after complete epilation and collected in PBS ++ (D-PBS w/o Ca & Mg containing 100 μg/ml Gentamycin and 20 μg/ml Amphotericin B) medium. After removal of subcutaneous fat and muscle and intensive washing with ice-cold PBS++, the skin tissue was chopped into long narrow pieces (1mm x 10 mm), and incubated in Dispase for 40 min at 37°C. Dispase was neutralized with Fetal Bovine Serum (FBS, ThermoFisher Inc., Waltham, MA, USA) and the digested tissue was rinsed with D-PBS. Epidermis was separated from the digested skin tissue, and sliced into small pieces (1mm x 1 mm). Epidermal slices were incubated in Trypsin for 40 min at 37°C. After neutralizing trypsin using FBS, epidermal cell suspension was vortexed vigorously for 5 min, ltered through 70 μm strainer, plated in T75 ask and cultured in DermaLife Melanocyte Medium, (DLM, Lifeline® Cell Technology, US) in hypoxic conditions (5% O 2 , 5% CO 2 ) at 37°C. The culture medium was changed each 48 h. When the cell density reached 70% con uency, the cells were subcultured by differential trypsinization for 11 passages.

Isolating rabbit hair follicle melanocytes (rMORS)
The facial skin area of a rabbit containing bilateral whisker pads was harvested and collected in PBS ++ medium. After removal of muscle and connective tissue, the whisker pad skin was intensively washed with ice-cold PBS ++ and incubated in 2mg/mL Collagenase V at 37 °C for 3 hours. After neutralizing and rinsing, the visible whisker hair shafts were plucked from the loosened dermal tissue along with their follicle-sinus complex. After rinsing, the follicle-sinus complex was micro-dissected to remove the dermal cavernous envelope, hereby obtaining an intact rabbit hair follicle outer root sheath (ORS). Whisker hair follicles were intensively washed and treated with 5mg/ml Collagenase V at 37 °C for 10 min. After neutralizing and rinsing, the hair follicle was placed onto a 6-well-size Transwell membrane (Corning Inc., New York, NY, USA). Lower chamber was lled with DLM Medium and incubated in hypoxic conditions (5% O 2 , 5% CO 2 ) at 37°C for 2 days. After 7-11 days of the air-liquid interface culture, the cells migrated out of the ORS and formed a cell monolayer on the nylon mesh of the Transwell insert. After 17-24 days at 50-80% con uence, the cells were detached by 0.04 %/0.03 % Trypsin/EDTA and subcultured in a T75 ask in DLM Medium. The culture medium was changed after 48 h and kept for another 3 days. When the cell density reached 70% con uency, the cells were subcultured for the next 11 passages.

Differential Trypsinization
For differential trypsinization, both rabbit rMORS and rEM were subcultured using 0.04% trypsin/0.03% EDTA at room temperature for 2-3 min. The trypsinization process was observed by a microscope to con rm that the majority of melanocytes detached and that keratinocytes were still adherent. The cells in the detached fraction were collected. Trypsin was neutralized using FBS and the cells were further subcultured in DLM Medium. Differential trypsinization hereby helped detach weakly adherent melanocytes from other more adherent cell types, as it has been postulated in human ORS melanocyte culture [23,25].

Von Kossa staining
The rabbit rMORS and rEM after passage 7 were split from cell culture asks and seeded onto Superfrost Microscope slides (ThermoFisher Inc., Waltham, MA, USA) coated with Type I collagen and left to adhere for 24 hours in the DLM medium. The cells were xed in 4% paraformaldehyde at room temperature. After rinsing with distilled water, the cells were incubated in 2% silver nitrate solution (Carl Roth GmbH, Karlsruhe, Germany) exposed to a 254 nm UV irradiation in Cell Culture Cabinet (BW-130 Silver, Kojair Tech Oy, Vilppula, Finland) for 1 h at room temperature. After rinsing in distilled water, the unbound silver was removed by incubating slides in 5% sodium thiosulfate (Carl Roth GmbH, Karlsruhe, Germany) for 5 minutes. The cells were counterstained in Nuclear Fast Red reagent (Carl Roth GmbH, Karlsruhe, Germany) for 5 minutes, and sealed by a cover slip using anhydrous mounting media ROTI®Mount (Carl Roth GmbH, Karlsruhe, Germany). Stained melanocytes were imaged Keyence BZ-9000 microscope (Keyence GmbH, Neu-Isenburg, DE, USA).

Melanin Production in Melanocytes
To assess the melanin synthesis in the functional rMORS and rEM, melanin content was measured as previously described [23,25]. Brie y, rMORS and rEM at P7, P9 and P11 were detached with trypsin/EDTA, washed in PBS and lysed by 3 freeze/thaw cycles. The lysate was incubated in 1 N NaOH 60 °C for 3 h with gentle agitation, and the absorbance of supernatant was measured by a spectrophotometer (Synerge, BioTek Instruments Inc., USA) at 475 nm wavelength against a reference wavelength of 620 nm. Absorbance of each sample was compared against absorbance of a synthetic melanin curve based on a 0-100 μg/ml concentration span (Sigma-Aldrich GmbH, Schnelldorf, Germany) and melanin concentration of the measured sample was retrieved by the means of linear regression. The assay provided a readout as amount of melanin and it was broken down respective to the number of cells to amount of melanin per cell.

Tyrosinase Activity
To determine tyrosinase activity of rMORS and rEM, 100µl cell lysates after freeze/thaw cycles were incubated in 200µl 5mM L-DOPA solution (in 0.1M KH 2 PO 4 buffer, pH 7.2) for 4-5h at 37 °C in the dark incubator. The absorbance of supernatant was measured by a spectrophotometer (Synerge, BioTek Instruments Inc., USA) at 475 nm wavelength against 0.1M KH 2 PO 4 buffer as a blank control.

L-DOPA Stimulation
To visualize the melanin production in rMORS and rEM, stimulation using L-DOPA (l-3,4dihydroxyphenylalanine) was performed as described in [30,31]. Brie y, rMORS and rEM were detached from cell ask plastic by trypsin/EDTA and seeded in Chamber Slides (ThermoFisher Inc., Waltham, MA, USA). After a 72 h attachment, the cells were xed in ice-cold acetone/methanol (1:1), and incubated in 5 mM L-DOPA (Sigma-Aldrich GmbH, Schnelldorf, Germany) for 5 hours. After washing, the cells were stained with 0.05% Nile Blue solution for 20 min, and imaged by the means of Keyence microscope.
Gene Expression Analysis rMORS and rEM at passages P3, P5, P7 and P9 were detached with trypsin/EDTA, HBSS and lysed in Qiazol Lysis Reagent (Qiagen, Hilden, Germany). Total RNA was isolated using RNeasy Plus Universal Kit (Qiagen, Hilden, Germany), and 1 μg of total RNA was reverse transcribed into cDNA using the QuantiTect Reverse Transcription Kit (Qiagen, Hilden, Germany) according to the manufacturer's recommendation.

Consent for publication
Not applicable Availability of data and materials The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
H.L. contributes to acquisition, analysis and interpretation of data, draft the manuscript; CM.B. contributes to acquisition of data; M.S. contributes to acquisition of data; K.S. contributes to design of the work, acquisition and analysis of data; T.K. contributes to acquisition of data; A.B. contributes to revising manuscript; JC.S. contributes to conception of the work and fund raising; V.S. contributes to design of the work, draft and substantively revise the manuscript; B.L. contributes to design of the work, draft and substantively revise the manuscript