The roles of Frizzled-3 and Wnt3a on melanocyte development: In vitro studies on neural crest cells and melanocyte precursor cell lines

Journal of Dermatological Science xxx (2014) xxx–xxx

G ModelDESC 2681 No. of Pages 9

The roles of Frizzled-3 and Wnt3a on melanocyte development: In vitrostudies on neural crest cells and melanocyte precursor cell lines

Chung-Hsing Chang a,b,*, Rong-Kung Tsai c,d,**, Ming-Hsien Tsai b, Yi-Hsiung Lin e,Tomohisa Hirobe f

aDepartment of Dermatology, Kaohsiung Medical University Hospital, Kaohsiung, TaiwanbGraduate Institute of Medicine, Kaohsiung Medical University, Kaohsiung, Taiwanc Institute of Eye Research, Buddhist Tzu Chi General Hospital, Hualien, Taiwand Institute of Medical Sciences, Tzu Chi University, Hualien, TaiwaneNational Applied Research Laboratories, Instrument Technology Research Center, Hsinchu, Taiwanf Fukushima Project Headquarters, National Institute of Radiological Sciences, Anagawa 4-9-1, Inage-ku, Chiba 263-8555, Japan

A R T I C L E I N F O

Article history:Received 10 October 2013Received in revised form 13 April 2014Accepted 21 April 2014

Keywords:Wnt3aFrizzled-3Neural crestMelanocyte precursorMelanoblast

A B S T R A C T

Background: Wnt3a and Frizzled-3 are both expressed in the dorsal neural tube that gives rise to theneural crest in Xenopus, zebrafish and mice. Melanocytes originate from the neural crest (NC) andpostnatally, melanocyte stem cells reside in the hair follicle bulge and in the dermis. However, the roles ofWnt3a and Frizzled-3 in melanocyte development have not been clarified.Objective: The aim of this study was to delineate the expression of Frizzled-3 in murine melanocytelineage and human melanocytes, and to study the effects of Wnt3a on melanocyte development atvarious stages.Methods: Murine NC explant cultures and three NC-derived melanocyte lineage cell lines, includingNCCmelb4M5 (Kit� melanocyte precursors), NCCmelb4 (Kit+ melanoblasts) and NCCmelan5 (differenti-ated melanocytes), and human epidermal melanocytes were treated with pure recombinant Wnt3aprotein and their cell behaviors were analyzed including their proliferation, Kit expression, tyrosinase(Tyr) activity, melanin production, dendrite formation and migration.Results: Frizzled-3 was expressed in Tyr-related protein (TRP)-1+ cells in NC explant cultures, in all 3melanocyte precursor cell lines and in human melanocytes. Wnt3a increased the population of TRP-1+

cells, the number of L-3,4-dihydroxyphenylalanine (DOPA)+ cells and dendrite formation in NC explantcultures. Wnt3a stimulated the proliferation of all 3 melanocyte precursor cell lines in a dose-dependentmanner and also stimulated human melanocyte proliferation. Moreover, Wnt3a increased Tyr activityand melanin content of differentiated melanocytes, but did not activate Tyr activity in melanoblasts.Wnt3a stimulated dendrite formation in differentiated melanocytes, but not in melanoblasts. Wnt3a didnot affect melanoblast or melanocyte migration. Wnt3a did not induce c-Kit expression in Kit�

NCCmelb4M5 cells and did not affect c-Kit expression in any cell line tested.Conclusions: Frizzled-3 is constitutively expressed in murine melanocyte precursors, melanocytes andhuman melanocytes. Wnt3a and Frizzled-3 signalings play important roles in regulating the proliferationand differentiation of murine NCCs and various developmental stages of melanocyte precursors. Theeffect of Wnt3a on human melanocytes is similar to its effects on murine melanocytes. Therefore Wnt3a/Frizzled-3 signaling is a promising target for human melanocyte regeneration.

ã 2014 Japanese Society for Investigative Dermatology. Published by Elsevier Ireland Ltd. All rightsreserved.

Contents lists available at ScienceDirect

Journal of Dermatological Science

journa l home page : www.jds journal .com

* Corresponding author at: Kaohsiung Medical University, Dermatology, 100,Shih-Chuan 1st RD, Kaohsiung, Taiwan. Tel.: +886 7 3208901; fax: +886 7 3218902.** Corresponding author at: Institute of Eye Research, Buddhist Tzu Chi GeneralHospital, Hualien, Taiwan. Tel.: +886 3 8561825; fax: +886 3 8577161.

E-mail addresses: [email protected] (C.-H. Chang),[email protected] (R.-K. Tsai).

http://dx.doi.org/10.1016/j.jdermsci.2014.04.0120923-1811/ ã 2014 Japanese Society for Investigative Dermatology. Published by Elsev

Please cite this article in press as: Chang C-H, et al. The roles of Frizzled-3crest cells and melanocyte precursor cell lines. J Dermatol Sci (2014), h

1. Introduction

Melanocytes are generated either directly from the neural crest(NC) (in the mouse at approximately E9) or indirectly from nervecells (around at E11). Transcription factors, such as Sox10, Pax3,FoxD3 and Mitf, participate in a genetic network regulatingmelanocyte formation from the neural crest. The activity of these

ier Ireland Ltd. All rights reserved.

and Wnt3a on melanocyte development: In vitro studies on neuralttp://dx.doi.org/10.1016/j.jdermsci.2014.04.012

mailto:[email protected]

mailto:[email protected]

http://dx.doi.org/10.1016/j.jdermsci.2014.04.012



http://www.sciencedirect.com/science/journal/09231811/

www.jdsjournal.com

2 C.-H. Chang et al. / Journal of Dermatological Science xxx (2014) xxx–xxx


intrinsic factors is controlled and modulated by extracellularsignals including canonical Wnt, Edn, Kitl, and other signals thatremain to be identified [1]. In the trunk region of the mice, foundermelanoblasts are determined around E8.5–E9.5. Precursor mela-noblasts arising from founder melanoblasts can be visualized fromE10.0. The melanoblasts begin to migrate from the migrationstaging area (MSA) through the dermis along a dorsolateralpathway underneath the ectoderm from E10.5, the first wavemelanocyte in skin. From E11, the second wave dermal melano-blasts appear which are generated from the neural precursors(with neural crest stem cells) migrating along the ventral pathway.On E13.5, melanoblasts migrate from the dermis to the epidermiswhere they continue to proliferate and migrate actively [2].Melanoblasts that enter developing hair follicles around E15.5 andE16.5 appear to survive without a Kit signal until they arereactivated upon initiation of the first wave of the hair cycle afterbirth [3,4]. Using the Dct-LacZ transgenic mice for melanocytelineage tracing, melanocyte stem cells responsible for melanocyteregeneration during hair cycle are identified in the bulge area [5],which are a c-Kit-negative population [6]. Using double transgenicWnt1-Cre/R26R mice, neural crest cells (NCCs) can be traced inadult animals by detecting b-galactosidase expression. b-galacto-sidase-positive cells can be identified in basal layers of the outerroot sheath from the hair bulge to the matrix at the base of the hairfollicle, which are possible melanocyte precursors or moreprimitive NC progenitors. Hair bulge explant cultures give riseto pluripotent NCCs, which can differentiate into neurons, smoothmuscle cells, Schwann cells, chondrocytes and melanocytes [7].Stem cells with NC characteristics can be derived from the bulgearea of cultured human hair follicles and dermis-derived spheres,and can give rise to myogenic, melanocytic and neuronal celllineages [8,9]. Thus, the bulge area of adult hair follicles and dermisare reservoirs of pluripotent NC stem cells. Factors guiding thedevelopment of the melanocyte lineage in the NC may playimportant roles in the development of melanocyte stem cells thatreside in the bulge area and dermis.

Most mammalian genomes, including the human genome,harbor 19 Wnt genes, which fall into 12 conserved Wntsubfamilies. Wnt proteins are �40 kDa in size and contain manyconserved cysteine residues. When interacting with target cells,Wnt proteins bind a heterodimeric receptor complex, whichconsists of a Frizzled and an LRP5/6 protein. The 10 mammalianFrizzled proteins are seven-transmembrane receptors and havelarge extracellular N-terminal cysteine-rich domains that providea primary platform for Wnt binding. The Wnt–Frizzled interactionis promiscuous: a single Wnt can bind multiple Frizzled proteinsand vice-versa [10]. Wnt genes play crucial roles in the early stepsof NC formation including the induction, maintenance ofpresumptive NC fate and proliferation of NC progenitors, andlater steps of specification, proliferation and migration ofdifferentiated NCC types [11–13].

Melanoblast specification from NC precursors is governedprimarily by Wnt and BMP signaling molecules; BMP can suppressboth Wnt-induced sensory neurogenesis in mouse [14] andWnt-induced melanocyte generation in quail neural crest cells[15]. Wnt/b-catenin plays a dual stage-dependent role in neuralcrest stem cell lineage decisions: at an early stage (presumably inthe premigratory neural crest), canonical Wnt controls sensoryneurogenesis; at a somewhat later stage, it might regulatemelanocyte formation, although this remains to be confirmed[2]. Wnt signals that influence NC formation and melanocytelineage specification act through the stabilization of b-catenin andits regulation of transcription by binding to Tcf/Lef transcriptionfactors [13]. Wnt1 and Wnt3a are expressed in the dorsal neuraltube at the specific time and site of NC formation [16].In Wnt1- and Wnt3a-deficient mouse embryos, melanocyte


precursors are markedly deficient [11]. Wnt3a-conditioned medi-um dramatically increases the number of melanocytes in quail NCcultures, but decreases the number of neurons and glia cells [15].Over-expression of Wnt signaling genes, including b-catenin,Wnt1 and Wnt3a, in murine neural tube explant cultures inducesthe expansion of melanocyte numbers, but Wnt1 and Wnt3a actthrough distinct modes. Wnt1 acts on melanocyte precursors toexpand the melanocyte lineage, while Wnt3a expands melanocytenumbers by biasing the fate of NCCs [17]. Ablation of b-catenin inmutant animals causes the loss of melanocytes [18]. Frizzled-3 isexpressed at the gastrula stage early enough to mediate Wnt-dependent NC induction. Frizzled-3 is then expressed restrictivelyin the neural plate and in the dorsal neural tube after neural tubeclosure in Xenopus, mice, chickens and zebrafish [13]. Frizzled-3and Wnt3a are highly localized to dorsal neural tissues that giverise to the NC [11,19,20]. However, the expression of Frizzled-3 inthe melanocyte lineage has not yet been clarified.

The Wnt/Frizzled/b-catenin signaling pathway has the potentialto be a therapeutic target for melanocyte regeneration. However, theexpression of the Frizzled receptor on melanocytes is not wellunderstood and the effects of Wnts on the hierarchy of melanocytedevelopment have not been comprehensively delineated. Thesecircumstances prompted us to investigate in detail the effect ofWnt3a/Frizzled-3 on melanocyte development. In this study, weused culture systems of NC explants and three cell lines derived fromNCCs as well as human melanocytes to investigate the effect ofWnt3a on melanocyte development [21]. Namely, murine NCCs atE9.5 serve as an excellent experimental model for studying thedevelopment of melanocyte precursors in normal embryonic skinand specific gene-mutated skins [21–23]. The three cell lines derivedfrom murine NCCs, NCCmelb4M5, NCCmelb4 and NCCmelan5,represent specific developmental stages of melanocyte precursors.NCCmelb4M5 cells do not express Kit and are immortal and stable inthe absence of Kit ligand. They are positive for TRP-1 and TRP-2 andcontain stage I melanosomes [24]. Glial fibrillary acidic protein(GFAP), which is a marker for glial cells, is also expressed byNCCmelb4M5 cells, while NCCmelb4 cells are negative for thisprotein [24], indicating that NCCmelb4M5 cells are c-Kit-negativebipotent glia/melanocyte precursors. NCCmelb4 cells aremelanoblasts that are Kit+, Tyr–, TRP-1+, TRP-2+ and L-3,4-dihydroxyphenylalanine (DOPA)–. Electron microscopic observa-tions revealed that these cells contain only stage I melanosomes [25].In contrast, NCCmelan5 cells have the characteristics of differentiat-ed melanocytes and are Tyr+, Kit+, TRP-1+, TRP-2+ and DOPA+ [26]. Inthis study, using NCCs derived from NC explant cultures andNCCmelb4M5,NCCmelb4andNCCmelan5cell linesaswell ashumanmelanocytes, the localization of Frizzled-3 and the effects of Wnt3aon the proliferation, migration and differentiation of melanocyteswere investigated in detail.

2. Materials and methods

2.1. Mice

C57BL/6 (B6) mice obtained from the National LaboratoryAnimal Center of Taiwan (Tainan, Taiwan) were used at E9.5 for NCexplant cultures.

2.2. Wnt3a protein

Recombinant mouse Wnt-3A derived from Wnt-over-express-ing Chinese hamster ovary cells was purchased from R&D Systems,Minneapolis, MN, USA (Catalog No. 1324-WN). Mouse Wnt3ashares 96% amino acid identity with human WNT3a. Wnt3a wasresuspended in Minimal Essential Medium (GIBCO, USA) to aconcentration of 10 ng/ml for cell culture or explant culture.



C.-H. Chang et al. / Journal of Dermatological Science xxx (2014) xxx–xxx 3


2.3. Antibodies

Frizzled-3 was detected using a monoclonal anti-human/mouseFrizzled-3 antibody (rat IgG2a, 1:200 dilution) purchased fromR&D Systems, Minneapolis, MN, USA (Catalog No. MAB1001, clone169,310). The antibody to b-catenin was purchased from CellSignaling Technology, Inc., Danvers, MA, USA (Catalog No. 9562,1:200 dilution). The antibody to Tyr-related protein 1 (TRP-1)(aPEP1, polyclonal rabbit serum which recognizes the carboxyterminus of TRP-1, 1:500 dilution) was kindly supplied by Dr. VinceHearing (NIH, Bethesda, MD). The antibody to c-Kit (rat IgG, 1:20dilution) was a gift from Dr. S. Nishikawa (Riken Center forDevelopmental Biology, Kobe, Japan).

2.4. NC explant cultures

The NC explant culture method was described previously [27].Briefly, trunk regions posterior to the forelimb buds were dissectedfrom embryos at E9.5 using tungsten needles. The trunks wereindividually treated with 1% trypsin (Difco, Detroit, MI, USA, 1:250)in Tyrode’s solution for 20 min at 4 �C. Trypsinization wasterminated by washing with Tyrode’s solution including 10% fetalbovine serum (FBS, Gibco, Grand Island, NY, USA). The tissue wasgently pipetted with Pasteur pipettes to separate the neural tubefrom the surrounding tissue. Eagle’s minimum essential medium(MEM) containing 15% FBS and 50 ng/ml recombinant mouse stemcell factor (SCF; R&D Systems, Inc., Minneapolis, MN, USA) wasused as a basic medium. Neural tubes were explanted individuallyin 12-well plates (Falcon, Oxnard, CA, USA). The explants werefixed on day 6 for TRP-1 immunohistochemistry or DOPA staining.

2.5. Murine melanocyte cell lines, human melanocytes and cultureconditions

NCCmelb4 cells were established using a cloning method froman immortal cell population, NCC-S4.1 [31,33] derived from NCCs ofWB Sl/+ or +/+ mice. NCCmelan5 cells [32] were established fromNCCs of B6 mice. NCCmelb4M5 cells [29] were established fromNCCmelb4 cells. NCCmelb4M5 cells were grown in MEMsupplemented with 5% FBS. NCCmelb4 cells were grown inMEM supplemented with 5% FBS and 50 ng/ml recombinantmouse SCF. NCCmelan5 cells were grown in MEM supplementedwith 15% FBS, 50 ng/ml SCF and 100 nM endothelin (ET)-3(Calbiochem Inc., Darmstadt, Germany). Three cell lines weregifts from Masako Mizoguchi (St. Marianna University, Kawasaki,Japan). Human melanocytes were derived from HEMn-MP, whichare human epidermal melanocytes isolated from moderatelypigmented neonatal foreskins and were cultured in MelanocyteGrowth Medium 254 (Cascade Biologics Inc., Portland, OR, USA).

2.6. Immunostaining

NCC explants or culture cells were rinsed with phosphatebuffered saline (PBS) and fixed in 2% paraformaldehyde (PFA) for2 min, then with 95% ethanol for 1 min, and were then immersed ingoat serum (DAKO Corp., Carpinteria, CA, USA) for 15 min to blocknon-specific binding. The primary antibodies were used for 16 h at4 �C or 1 h at room temperature. For fluorescence staining, thesecondary antibody FITC-labelled anti-rat IgG or Cy3-labelled anti-rabbit IgG (Jackson Immuno Research, West Grove, PA, USA) weretreated for 2 h at room temperature. The NCC explants werecounterstained with DAPI (1:500, Sigma Co., St. Louis, MO, USA).For other staining, biotin-labelled anti-rat IgG (Jackson ImmunoResearch) was used for 2 h at room temperature and alkaline-phosphatase-labelled anti-biotin (DakoCytomation Ltd.,Cambridgeshire, UK) was added and incubated for 1 h at room


temperature, and then the color was developed using the FuchsinSubstrate System (DakoCytomation), and counterstained withhematoxylin (Sigma).

2.7. DOPA reaction

Explants were rinsed with PBS and fixed with 2% PFA for 15 minat room temperature. After washing in PBS, the cultures wereincubated at 37 �C for 5 h in 0.1% DOPA (Sigma) in PBS (pH 7.2).

2.8. Proliferation assay

Cells were cultured in 96-well plates supplemented with orwithout Wnt3a (2, 10 and 50 ng/ml). On day 5, the cells wereincubated with AlamarBlueTM (Trek Diagnostic Systems, Waltham,MA, USA) for 4 h at 37 �C, and the fluorescence was read using amicroplate reader (BIO-TEK, Synergy HT, Arlington, MA, USA).

2.9. Assay of Tyr activity and melanin content

Cells were plated in 6 cm culture dishes at a density of 1 �105

cells/dish. In some experiments, 10 ng/ml Wnt3a was added to theculture medium. On day 5, cells were collected and lysed in celllysis buffer (1% Triton-X100 in 0.01 M Tris–HCl, pH 7.4). Nucleiwere removed by centrifugation at 7840 � g for 10 min at 4�C. Celllysates were mixed with an equal volume of 0.1% DOPA in 50 mMphosphate buffer and incubated at 37 �C for 60 min. OD475 wasdetermined using a microplate reader (BIO-TEK Synergy HT). Onday 5, cells were also collected and melanin was solubilized in 1 NNaOH/10% dimethyl sulfoxide at 60 �C for 15 min. The melanincontent was estimated by measuring absorbance at 475 nm(OD475). The amounts of protein in the cell lysates were used formeasuring the melanin content with a Protein Assay (Bio-RadLaboratories, Hercules, CA, USA). Each value was calculated fromthe melanin content per protein concentration to compensate fordifferences in cell number and was used for comparison. Values arethe mean � SD of triplicate determinations.

2.10. Migration assay

Cells (5 �105 per well) were seeded in six-well plates and wereallowed to adhere for 24 h. The cells were treated with 10 mg/mlMitomycin C (Sigma) for 3 h, washed with PBS and simply woundedwith a pipette tip. Fresh, complete medium with or without 10 ng/mlWnt3a was added and incubated for 48 h. Photographs were takenevery 12 h at the same position of the wound. Migrating cell numberswere counted using an inverted microscope (Zeiss Axiovert 200 M).The images were captured using a digital camera AxioCamHRc5 andwere analyzed by Axiovision 4.0 software.

2.11. Measurement of cell number, dendrite number and length

TRP1-positive cells or DOPA-positive stained cells or migratingcells on each culture plate were observed using an invertedmicroscope (Zeiss Axiovert 200 M). The images were captured usinga digital camera AxioCamHRc5. The total number of cells wereenumerated for each explant. The dendrite number was counted onthe digital images. Each dendrite length was measured by Axiovision4.0software.Thesum ofeachdendrite length foraDOPA-positivecellwas treated as the total length of dendrites per cell.

2.12. Real-time PCR analysis of Kit mRNA

NCCmelb4M5, NCCmelb4 and NCCmelan5 cells were culturedwith or without Wnt3a at 10 ng/ml in 75 cm2

flasks at a density of7.5 �105 cells/flask for 72 h, then the cells were collected and total





RNAs were prepared using the SV RNA isolation system (PromegaCorp. Madison, WI, USA). 1 mg total RNA was reverse transcribedinto cDNA using the 1st Strand cDNA Synthesis Kit for RT-PCR(Roche Diagnostics Co., Indianapolis, IN, USA) following themanufacturer’s instructions. Target genes were amplified withspecific primers and SYBR Premix Ex Tag (Perfect Real Time, TakaraBio Inc., Ohtsu, Japan) using a Light-Cycler (Roche Diagnostics Co.,Indianapolis, IN, USA). PCR conditions for each primer were:10 min at 95 �C (the same for all primers), followed by 40 cycles at95 �C for 10 s (the same for all primers), at 55 �C for 10 s (Kit) or at58 �C for 10 s (actin). Light-Cycler Software Ver. 3.5 (Roche) wasused to analyze the PCR data, which were then exported to Excel(Microsoft, Redmond, WA, USA) for further analysis. The expres-sion of each gene was measured in triplicate and was normalized tothe actin expression level. The primer sequences used were asfollows:

Kit sense primer: 50-CGA CTG CCC GTG AAG TGG A-30;Kit anti-sense primer: 50-GCC AGA AGG ACG GGG TCG G-30;Actin sense primer 50-AATCGTGCGTGACATCAAAG-30;Actin anti-sense primer 50-GAAAAGAGCCTCAGGGCAT-30.

2.13. Statistical analysis

Student’s t-test was used for statistical analysis. p < 0.05 isconsidered to be significant.

3. Results

3.1. Frizzled-3 is expressed constitutively in melanocytic lineage cells

The expression pattern of Frizzled-3 is consistent with thepattern of Wnt1/Wnt3a expression in the dorsal neural tube [11],which implies that Frizzled-3 may play an important role inWnt3a-binding and in turn might activate downstream compo-nents in melanocyte development. In NC explant cultures for 6days, we performed double staining for TRP-1 and Frizzled-3.TRP-1 positive melanocyte precursors appeared to express

Fig. 1. Frizzled-3 is expressed constitutively in melanocytic lineage cells. Frizzled-3 is exco-localized) derived from NC explant cultures and the 3 NCC cell lines as well as humimmunocytochemistry, bright-field; scale bar, 50 or 100 mm.


Frizzled-3 (Fig. 1A–D). All murine melanocyte precursor celllines and human melanocytes expressed Frizzled-3 (Fig. 1E–L).Thus, Frizzled-3 is expressed constitutively in melanocyticlineage cells from the NC, melanoblasts, to differentiatedepidermal melanocytes.

3.2. Wnt3a expands the population and enhances the differentiation ofmelanocytic lineage cells from the NC

On day 6 of NC explant cultures, the number of TRP-1+ cells inthe control group was 466 � 75 (N = 5), while in the Wnt3a-supplemented group it was 734 � 241 (N = 4, p < .05, Fig. 2A,D,G).On day 9 of NC explant cultures, the number of DOPA+ cells in thecontrol group was 147 � 92 (N = 6), while in the Wnt3a-supple-mented group, the number was 286 � 86 (N = 5, p < .05,Fig. 2B,C,E,F,H). Moreover, on day 9, the average number ofdendrites and the total number of dendrites per each DOPA+ cell inthe control group were 3.0 � 1.0 and 101.2 � 44.7 mm (N = 8),respectively, while those in the Wnt3a-supplemented group were6.2 � 2.3 and 179.0 � 47.6 mm (N = 8), respectively. Both indicatorsof dendritogenesis increased significantly (p < .001, Fig. 2I,J). Thus,Wnt3a stimulates proliferation and differentiation, including Tyractivation and dendrite formation, of melanocytic lineage derivedcells from the NC.

3.3. Wnt3a stimulates the proliferation of melanocytes at differentdevelopmental stages in a dose-dependent manner

NCCmelb4M5, NCCmelb4 and NCCmelan5 cells and humanmelanocytes were treated with Wnt3a at concentrations of 2, 10and 50 ng/ml for 5 days. Wnt3a increased the AlamarBlueTM

fluorescence of melanocytes at various developmental stages in adose-dependent manner (Fig. 3). However, Wnt3a at a concentra-tion of 50 ng/ml failed to increase the proliferation of humanmelanocytes further, compared to the concentration of 10 ng/ml.Thus, Wnt3a stimulates the proliferation of melanocytes atdifferent developmental stages.

pressed in TRP-1-positive cells (expression of Frizzled-3 (FITC) and TRP-1 (Cy3) wasan melanocytes. A–D, fluorescent immunocytochemistry; scale bar, 20 mm; E–L,



Fig. 2. Wnt3a expands the population of melanocytic lineage cells and stimulates their differentiation from the NC. NC explants derived from E9.5 mouse embryos werecultured with or without 10 ng/ml Wnt3a. TRP-1 immunostaining (A, D) and DOPA reaction (B, C, E, F) were performed on day 6 and day 9 explants, respectively. The numberof TRP-1+ cells, DOPA+ cells and the average number of dendrites and the total length of dendrites per each DOPA+ cell in Wnt3a-supplemented NC explants were significantlyincreased, compared to the control group (G–I). (*p < 0.05, ***p < 0.001).



3.4. Wnt3a activates Tyr activity and stimulates melanin production indifferentiated melanocytes but not in melanoblasts

NCCmelb4M5, NCCmelb4 and NCCmelan5 cells and humanmelanocytes were treated with Wnt3a at a concentration 10 ng/mlfor 5 days. Wnt3a increased Tyr activity measured by the intensityof DOPA reaction in NCCmelan5 cells and human melanocytes, butnot in NCCmelb4M5 or NCCmelb4 cells (Fig. 4A). Wnt3a increasedthe melanin content measured by absorbance at 475 nm (OD475) inNCCmelan5 cells and human melanocytes, but not in NCCmelb4M5or NCCmelb4 cells (Fig. 4B). Thus, Wnt3a does not promotemelanoblast differentiation with Tyr activation, but can enhanceTyr activity in differentiated melanocytes.

Fig. 3. Wnt3a stimulates the proliferation of melanocytes at different developmental stand human melanocytes were treated with Wnt3a at a concentration of 2, 10 and 50 ngvarious developmental stages in a dose-dependent manner.


3.5. Wnt3a promotes dendrite formation in differentiatedmelanocytes but not in melanoblasts

NCCmelb4M5, NCCmelb4 and NCCmelan5 cells and humanmelanocytes were treated with Wnt3a at a concentration10 ng/ml for 2 days. Images of cells were captured at 12 h,24 h and 48 h. Wnt3a significantly increased the dendritenumber and total dendrite length of NCCmelan5 cells(Fig. 5A) and human melanocytes (Fig. 5B), but no significantchange of dendrites was noted for NCCmelb4M5 or NCCmelb4cells (data not shown). Thus, Wnt3a stimulates the dendriteformation of differentiated melanocytes, but not of undifferen-tiated melanoblasts.

ages in a dose-dependent manner. NCCmelb4M5, NCCmelb4 and NCCmelan5 cells/ml for 5 days. Wnt3a increased the AlamarBlueTM fluorescence of melanocytes at



Fig. 5. Wnt3a promotes dendrite formation in differentiated melanocytes. NCCmelan5 cells and human melanocytes were treated with Wnt3a at 10 ng/ml for 2 days. Cellswere observed and images were captured by phase contrast microscopy at 12, 24 and 48 h. Wnt3a increased the dendrite number and total dendrite length per melanocyteboth in NCCmelan5 cells (A) and in human melanocytes (B) in a time-dependent manner.

Fig. 4. Wnt3a activates Tyr activity and stimulates melanin production in differentiated melanocytes but not in melanoblasts. (A) Wnt3a increases Tyr activity in NCCmelan5cells and human melanocytes, but not in NCCmelb4M5 and NCCmelb4 cells. (B) Wnt3a increases melanin content in NCCmelan5 cells and human melanocytes, but not inNCCmelb4M5 and NCCmelb4 cells.



Please cite this article in press as: Chang C-H, et al. The roles of Frizzled-3 and Wnt3a on melanocyte development: In vitro studies on neuralcrest cells and melanocyte precursor cell lines. J Dermatol Sci (2014), http://dx.doi.org/10.1016/j.jdermsci.2014.04.012


Fig. 6. Wnt3a does not affect melanoblast or melanocyte migration. NCCmelb4M5, NCCmelb4 and NCCmelan5 cells and human melanocytes were treated with 10 ng/mlWnt3a for 2 days. The wound healing assay was used for migrating cell numeration. There was no significant effect of Wnt3a on melanoblast or melanocyte migration.



3.6. Wnt3a does not affect the migration of melanoblasts ormelanocytes

A wound healing assay was used for cell migration analysis. Allcells were treated with Wnt3a at 10 ng/ml for 48 h. Migrating cells

Fig. 7. Wnt3a does not influence c-Kit expression at the mRNA or protein level onmelanocyte lineage cells. NCCmelb4M5, NCCmelb4 and NCCmelan5 cells weretreated with 10 ng/ml Wnt3a for 72 h. Immunofluorescence staining of c-Kit. Wnt3afailed to induce the expression of c-Kit on NCCmelb4M5 cells and did not enhancethe intensity of c-Kit expression on NCCmelb4 and NCCmelan5 cells. Real-time PCRshows that the mRNA level of c-Kit was not changed by Wnt3a treatment.


were counted in the wound gap at 24 and 48 h. There were nosignificant effects of Wnt3a on NCCmelb4M5, NCCmelab4 orNCCmelan5 cells or human melanocytes (Fig. 6).

3.7. Wnt3a does not induce or enhance c-Kit expression on melanocytelineage cells

To investigate whether Wnt3a can induce Kit expression duringmelanocyte development, we cultured NCCmelb4M5, NCCmelb4and NCCmelan5 cells with or without Wnt3a at a concentration of10 ng/ml for 72 h. Immunofluorescence staining for c-Kit showedthat NCCmelb4M5 cells still lacked c-Kit expression whileNCCmelb4 and NCCmelan5 cells had similar intensities of c-Kitexpression after Wnt3a treatment (Fig. 7A). Real-time PCR furtherconfirmed that the mRNA level of c-Kit was not changed in any ofthose 3 cell lines after Wnt3a treatment (Fig. 7B). Thus, Wnt3a doesnot influence c-Kit expression at the mRNA or protein level inmelanocyte lineage cells.

4. Discussion

During embryogenesis, melanocyte development mainly orig-inates from the NC and along the dorsal lateral pathway in thedermis (beneath the ectoderm) and then enters the epidermis andhair follicles [1,2]. In the postnatal period, melanocyte develop-ment mainly originates from the activation of melanocyte stemcells residing in the hair follicle bulge area which may migratedownward to the matrix during the onset of anagen [5] or migrateupward to the basal layer of the epidermis following wounding orUV irradiation [27,28] to become differentiated melanocytes.Melanocyte stem cells can also be derived from the multipotent NCstem cells in the dermis [9].

There was no previous data about Frizzled receptor expressionin the embryonic melanocyte lineage. Yamada et al. demonstratedthat melanocyte stem cells in the bulge area express Frizzled-4,Frizzled-7, low density lipoprotein receptor-related protein5 (Lrp5) and Lrp6 receptors as Wnt-related molecules, and thosemelanocyte stem cells are c-Kit-negative cells with the capacity togenerate melanoblasts and melanocytes [29]. In the present study,we first demonstrated that Frizzled-3 is specifically expressed onTRP-1-positive cells in the NC explants. Frizzled-3 is also expressedby murine NCCmelb4M5, NCCmelb4 and NCCmelan5 cells as wellas by human melanocytes. These results suggest that Frizzled-3 isexpressed constitutively on melanocyte lineage cells including c-Kit-negative melanocyte stem cells, c-Kit-positive melanoblastsand differentiated melanocytes. Moreover, no difference was foundin the expression of Frizzled-3 between mouse and humanmelanocytes.

Wnt3a is expressed not only in the dorsal neural tube, but alsoin the ectoderm-derived epidermis homogeneously in the





beginning, but is restricted to the placode of skin appendage and iscontinually expressed in the epithelial compartment of embryonicskin appendages [30]. Therefore, Wnt3a secreted from epidermismay influence the development of melanocyte precursors on thedorsolateral migration pathway, in the epidermis and in the hairfollicle. We show that purified recombinant Wnt3a protein canstimulate the expansion of TRP-1-positive cells and induction ofpigmented and dendritic melanocytes in mouse NC explantcultures. These results are consistent with the finding on RCAS-Wnt3a infected mouse NC explants cultures [17] or Wnt3a-conditional medium treated quail NC explants cultures [15]. Theyexplained the reason of melanocyte lineage expansion was mainlyat the expense of the smooth muscle lineage or neurogenic/gliogenic lineage rather than the proliferation of melanocytelineage [15,17]. Using murine melanocyte precursor cell lines andhuman melanocyte culture, we clarified Wnt3a has potentmitogenic effect on melanocyte precursors and differentiatedmelanocytes.

In postnatal hair cycling, Wnt3a expression is dependent uponthe hair cycle. Wnt3a is expressed in the epidermis, the bulge area,inner root sheath and hair bulbs in anagen, and is only detected inthe bulge area in catagen, and little Wnt3a protein is detected inthe hair follicle in telogen. Wnt3a is expressed again during theinitiation of the next anagen [31]. During the transit from telogento anagen of the hair cycle, the coordinated activation of Wnt inepithelial stem cells and melanocyte stem cells, both in the bulgearea, causes pigmented hair regeneration [32]. Epithelial stem cellsrather than melanocyte stem cells secrete Wnt3a (and also Wnt7a)during the telogen to anagen transition [32]. Yamada et al.demonstrated that UVB induces strong expression of Wnt7a inepithelial cell lineages (epidermal keratinocytes, hair folliclekeratinocytes and hair follicle stem cells, but not in melanocytestem cells) and consequently activates Wnt/b-catenin signaling,which triggers the differentiation of melanocyte stem cells intomelanoblasts [28]. Wnt3a is another up-regulated mRNA inducedby UVB shown in that study. In the present study, Wnt3a expandsTRP-1-positive cells in NCCs and supports the proliferation of c-Kit-negative NCCmelb4M5 cells, but does not induce their c-Kitexpression. The induction of Kit expression in bipotent glia/melanocyte NCCmelb4M5 cells implies the specification ofmelanocyte precursors. This finding implies that Wnt3a isimportant for the maintenance of melanocyte stem cells in theniche, but may not be able to activate c-Kit-negative stem cells tobecome c-Kit-positive melanoblasts by itself. UVB irradiation of amouse model showed a Wnt7a burst increase and the activation ofmelanocyte stem cells (Dct+/Kit� cells) to differentiate intomelanoblasts (Dct+/Kit+ cells) [28]. Whether Wnt7a alone or insynergy with Wnt3a induces c-Kit expression by melanocyte stemcells in the bulge area in vivo during UVB irradiation needs furtherstudy. Yang et al. discovered that Wnt3a is not required for mousemelanocyte differentiation from induced pluipotent stem cells[33]. Our data support their finding. In NCC experimental modelWnt3a together with neuroepithelium can induce fully differenti-ated melanocytes from neural crest stem cells; however, Wnt3aalone fails to induce Kit expression in NCCmelb4M5, or Tyrexpression in NCCmelb4 and differentiated melanocytes were notidentified in both melanocyte precursors in our culture conditions.

Our data show that Wnt3a supports the proliferation ofmelanoblasts and differentiated melanocytes in mice and inhumans. Guo et al. [34] showed that Wnt3a inhibits theproliferation of melan-a cells, an immortal line of mousemelanocytes. They infected Wnt3a into melan-a cells while wetreated melanocyte lineage cells with exogenous Wnt3a. Our dataalso show that Wnt3a enhances Tyr activity, melanin formationand dendrite formation in Tyr-positive melanocytes of mice andhumans, but not in Tyr-negative NCCmelb4M5 or NCCmelb4 cells.


Guo et al. showed that Wnt3a increases the Tyr activity andmelanin content in melan-a cells [31]. Irradiation with UVA but notwith UVB inhibited Tyr activity in Tyr-positive NCCmelan5 cells[35]. The intradermal injection of an adenovirus secreting Wnt3aincreased melanogenesis in follicle bulbs and darkened skin color[31]. Thus, Wnt3a has the potential to stimulate melanogenesis inhair follicles or the epidermis for clinical application.

Wnt3a has been reported to act through the canonical pathwayin neural crest cells [17]. Wnt3a signaling recruits b-catenin andLEF-1 to the LEF-1 binding site of the melanocyte-specificpromoter of human MITF gene (MITF-M promoter) to induce themicrophthalmia-associated transcription factor (MITF) [36].Mickels and Nusse found that purified Wnt5a protein can inhibitWnt3a protein-induced canonical Wnt signaling in a dosedependent manner by downregulating beta-catenin-inducedreporter gene expression [37]. It was reported that Wnt/b-cateninsignaling regulates melanocyte stem cell differentiation in haircycle progression. Lang et al. demonstrated that b-catenindisplaced the Groucho corepressor and activated the expressionof dopachrome tautomerase (Dct) in melanocyte stem cells atanagen in the hair cycle. Pax3-expressing melanoblasts are thuscommitted but undifferentiated until the Pax3-mediated repres-sion is relieved by activated b-catenin [38]. UVB-inducedmelanocyte stem cell activation presents with the nucleartranslocation of b-catenin [28]. During human hair follicledevelopment, Frizzled-3 expressed on the membrane andb-catenin expressed in nucleus are co-expressed in the placode,hair germ, and outer root sheath, bulge and precortical area in thebulbous peg [39], which implies that Frizzled-3 receptor binding ofWnt may stabilize b-catenin and trigger the downstream signals.In vivo, the effects of b-catenin signalling on the migration of cellsof the melanocyte lineage remain largely unknown. In vitro, LiCltreatment has been shown to prevent the migration of NCCs fromexplanted chicken neural tubes [40]. LiCl treatment and b-cateninexpression inhibits the migration of melan-a cells in vitro [41].In vivo, high b-catenin levels in melanoblasts reduces themigration of these cells, causing a white belly-spot phenotype[41]. Here we first demonstrated that Wnt3a treatment does notsignificantly affect the migration of melanocyte precursors ordifferentiated melanocytes in vitro. UVB-induced epidermalpigmentation needs not only Wnt/b-catenin activation of mela-nocyte stem cells but also Kit signaling to induce the subsequentmigration of melanoblasts to the epidermis [28].

Our data show that differentiated human melanocytes derivedfrom the foreskin exhibit the same reactivity to Wnt3a as domurine NCCmelan5 melanocytes, namely the promotion ofproliferation, melanogenesis and dendritogenesis, but had noeffect on migration, suggesting that the characteristics ofproliferation and melanogenesis in human melanocytes are similarto murine melanocytes. This assumption is supported by previousfindings that human melanocytes can be cultured using the sameserum-free culture medium [42] and can respond to the samegrowth factors [43] as murine melanocytes. The present study ofWnt3a and Frizzled-3 during melanocyte development mayprovide evidence that targeting Wnt/b-catenin signaling is areasonable strategy for human melanocyte regeneration.

Acknowledgements

This work is dedicated to Prof. Masako Mizoguchi (St. MariannaUniversity, Kawasaki, Japan). She encouraged spiritually andgenerously supported Asian female scientists to develop theirown career. She hosted me in 2003 to learn the neural crest explantculture technique at her Lab. We thank Yoko Kawa and Jia-HungHsieh for their technical assistance. This work was funded byNational Science Council of Taiwan.





References

[1] Sommer L. Generation of melanocytes from neural crest cells. Pigment CellMelanoma Res 2011;24:411–21.

[2] Luciani F, Champeval D, Herbette A, Denat L, Aylaj B, Martinozzi B, et al.Biological and mathematical modeling of melanocyte development. Develop-ment 2011;138(18):3943–54.

[3] Nishikawa S, Kusakabe M, Yoshinaga K, Ogawa M, Hayashi S, Kunisada T, et al.In utero manipulation of coat color formation by a monoclonal anti-c-kitantibody: Two distinct waves of c-kit-dependency during melanocytedevelopment. EMBO J 1991;10:2111–8.

[4] Yoshida H, Kunisada T, Kusakabe M, Nishigawa S, Nishigawa SI. Distinct stagesof melanocyte differentiation revealed by analysis of nonuniform pigmenta-tion patterns. Development 1996;122:1207–14.

[5] Nishimura EK, Jordan SA, Oshima H, Yoshida H, Osawa M, Moriyama M, et al.Dominant role of the niche in melanocyte stem-cell fate determination. Nature2002;416:854–60.

[6] Osawa M, Egawa G, Mak SS, Moriyama M, Freter R, Yonetani S, et al. Molecularcharacterization of melanocyte stem cells in their niche. Development2005;132:5589–99.

[7] Sieber-Blum M, Grim M, Hu YF, Szeder V. Pluripotent neural crest stem cells inthe adult hair follicle. Dev Dyn 2004;231:258–69.

[8] Yu H, Kumar SM, Kossenkov AV, Showe L, Xu X. Stem cells with neural crestcharacteristics derived from the bulge region of cultured human hair follicles. JInvest Dermatol 2010;130:1227–36.

[9] Toma JG, McKenzie IA, Bagli D, Miller FD. Isolation and characterization ofmultipotent skin-derived precursors from human skin. Stem Cells2005;23:727–37.

[10] Clevers H, Nusse R. Wnt/b-catenin signaling and disease. Cell 2012;149:1192–205.

[11] Ikeya M, Lee SM, Johnson JE, McMahon AP, Takada S. Wnt signaling required forexpansion of neural crest and CNS progenitors. Nature 1997;389(6654):966–70.

[12] Dorsky RI, Moon RT, Raible DW. Control of neural crest cell fate by the Wntsignalling pathway. Nature 1998;396:370–3.

[13] Wu J, Saint-Jeannet JP, Klein PS. Wnt-frizzled signaling in neural crestformation. Trends Neurosci 2003;26:40–5.

[14] Kleber M, Lee HY, Wurdak H, Buchstaller J, Ricomagno MM, Ittner LM, et al.Neural crest stem cell maintenance by combinatorial Wnt and BMP signaling. JCell Biol 2005;169:309–20.

[15] Jin EJ, Erickson CA, Takada S, Burrus LW. Wnt and BMP signaling govern lineagesegregation of melanocytes in the avian embryo. Dev Biol 2001;233:22–37.

[16] Dorsky RI, Moon RT, Raible DW. Control of neural crest cell fate by the Wntsignaling pathway. Nature 1998;396:370–3.

[17] Dunn KJ, Brady M, Ochsenbauer-Jambor C, Snyder S, Incao A, Pavan WJ. WNT1and WNT3a promote expansion of melanocytes through distinct modes ofaction. Pigment 5 Cell Res 2005;18:167–80.

[18] Hari L, Brault V, Kléber M, Lee HY, Ille F, Leimeroth R, et al. Lineage-specificrequirements of beta-catenin in neural crest development. J Cell Biol2002;159:867–80.

[19] Deardorff MA, Tan C, Saint-Jeannet JP, Klein PS. A role for frizzled 3 in neuralcrest development. Development 2001;128:3655–63.

[20] Borello U, Buffa V, Sonnino C, Melchionna R, Vivarelli E, Cossu G. Differentialexpression of the Wnt putative receptors Frizzled during mouse somito-genesis. Mech Dev 1999;89:173–7.

[21] Chang CH, Kawa Y, Tsai RK, Shieh JH, Lee JW, Watabe H, et al. Melanocyteprecursors express elastin binding protein and elastin-derived peptide(VGVAPG) stimulates their melanogenesis and dendrite formation. J DermatolSci 2008;51:158–70.

[22] Ito K, Takeuchi T. The differentiation in vitro of the neural crest cells of themouse embryo. J Embryol Exp Morphol 1984;84:49–62.


[23] Mizoguchi M. Melanocyte development: with a message of encouragement toyoung women scientists. Pigment Cell Res 2004;17:533–44.

[24] Kawa Y, Soma Y, Nakamura M, Ito M, Kawakami T, Baba T, et al. Establishmentof a kit-negative cell line of melanocyte precursors from mouse neural crestcells. Pigment Cell Res 2005;18:188–95.

[25] Ito M, Kawa Y, Watabe H, Ono H, Ooka S, Nakamura M, et al. Establishmentby an original single-cell cloning method and characterization of animmortal mouse melanoblast cell line (NCCmelb4). Pigment Cell Res2004;17:643–50.

[26] Ooka S, Kawa Y, Ito M, Soma Y, Mizoguchi M. Establishment andcharacterization of a mouse neural crest derived cell line (NCCmelan5).Pigment Cell Res 2001;14:268–74.

[27] Chou WC, Takeo M, Rabbani P, Hu H, Lee W, Chung YR, et al. Direct migration offollicular melanocyte stem cells to the epidermis after wounding or UVBirradiation is dependent on Mc1r signaling. Nat Med 2013;19:924–9.

[28] Yamada T, Hasegawa S, Inoue Y, Date Y, Yamamoto N, Mizutani H, et al. Wnt/b-catenin and kit signaling sequentially regulate melanocyte stem celldifferentiation in UVB-induced epidermal pigmentation. J Invest Dermatol2013;133:2753–62.

[29] Yamada T, Akamatsu H, Hasegawa S, Inoue Y, Date Y, Mizutani H, et al.Melanocyte stem cells express receptors for canonical Wnt-signaling pathwayon their surface. Biochem Biophys Res Commun 2010;396:837–42.

[30] Chang CH, Jiang TX, Lin CM, Burrus LW, Chuong CM, Widelitz R. Distinct Wntmembers regulate the hierarchical morphogenesis of skin regions (spinaltract) and individual feathers. Mech Dev 2004;121:157–71.

[31] Guo H, Yang K, Deng F, Ye J, Xing Y, Li Y, et al. Wnt3a promotes melaninsynthesis of mouse hair follicle melanocytes. Biochem Biophys Res Commun2012;420:799–804.

[32] Rabbani P, Takeo M, Chou W, Myung P, Bosenberg M, Chin L, et al. Coordinatedactivation of Wnt in epithelial and melanocyte stem cells initiates pigmentedhair regeneration. Cell 2011;145:941–55.

[33] Yang R, Jiang M, Kumar SM, Wu T, Wang F, Xiang L, et al. Generation ofmelanocytes from induced pluripotent stem cell. J Invest Dermatol2011;131:2458–66.

[34] Guo H, Yang K, Deng F, Xing Y, Li Y, Lian X, et al. Wnt3a inhibits proliferation butpromotes melanogenesis of melan-a cells. Int J Mol Med 2012;30:636–42.

[35] Hosaka E, Soma Y, Kawa Y, Kaminaga H, Osumi K, Ooka S, et al. Effects ofultraviolet light on melanocyte differentiation: studies with mouse neuralcrest cells and neural crest-derived cell lines. Pigment Cell Res 2004;17:150–7.

[36] Takeda K, Yasumoto K, Takeda R, Takeda S, Watanabe K, Udono T, et al.Induction of melanocyte-specific microphthalmia-associated transcriptionfactor by Wnt3a. J Biol Chem 2000;275:14013–6.

[37] Mikels AJ, Nusse R. Purified Wnt5a protein activates or inhibits beta-catenin-TCF signaling depending on receptor context. PLoS Biol 2006;4:e115.

[38] Lang D, Lu MM, Huang L, Engleka KA, Zhang M, Chu EY, et al. Pax3 functionsat a nodal point in melanocyte stem cell differentiation. Nature2005;433:884–7.

[39] Chang CH, Tsai RK. Expression of Wnt1, b-catenin, and Frizzled 3 duringhuman hair follicle and acrosyringium development. J Invest Dermatol2008;128:s144.

[40] de Melker AA, Desban N, Duband JL. Cellular localization and signaling activityof beta-catenin in migrating neural crest cells. Dev Dyn 2004;230:708–26.

[41] Gallagher SJ, Rambow F, Kumasaka M, Champeval D, Bellacosa A, Delmas V,et al. Beta-catenin inhibits melanocyte migration but induces melanomametastasis. Oncogene 2013;32:2230–8.

[42] Hirobe T, Hasegawa K, Furuya R, Fujiwara R, Sato K. Effects of fibroblast-derived factors on the proliferation and differentiation of human melanocytesin culture. J Dermatol Sci 2013;71:45–57.

[43] Hirobe T. How are proliferation and differentiation of melanocytes regulated?.Pigment Cell Melanoma Res 2011;24:462–78.


http://refhub.elsevier.com/S0923-1811(14)00108-X/sbref0005























































































































Documents

The roles of Frizzled-3 and Wnt3a on melanocyte development: In vitro studies on neural crest cells and melanocyte precursor cell lines