Endothelins are Involved in Regulating the Proliferation andDifferentiation of Mouse Epidermal Melanocytes in Serum-Free Primary Culture

Tomohisa HirobeDivision of Biology and Oncology, National Institute of Radiological Sciences, Anagawa, Chiba and Graduate School of Science and Technology,

Chiba University, Yayoi-cho, Chiba, Japan

Mouse epidermal melanoblasts preferentially prolifer-ated from disaggregated epidermal cell suspensionsderived from newborn mouse skin in serum-freemelanoblast-de®ned medium (MDM). After 14 d,almost all keratinocytes that existed predominantlyin the early stage of primary culture died, and purecultures of melanoblasts were obtained. Epidermalmelanoblasts dramatically increased in number inMDMDF consisting of MDM supplemented withdibutyryl adenosine 3¢:5¢-cyclic monophosphate(DBcAMP) and basic ®broblast growth factor(bFGF). Epidermal melanocytes increased in numberin MDMD consisting of MDM supplemented withDBcAMP. On the other hand, epidermal melano-cytes were induced to differentiate in MDMM con-sisting of MDM supplemented with a-melanocyte-stimulating hormone (MSH). Pure cultured primarymelanoblasts or melanocytes in MDMDF or MDMDwere further cultured with MDMDF or MDMD sup-plemented with endothelin (ET)-1, -2, or -3 from14 d. A dramatic increase in the number of melano-blasts or melanocytes was observed after 7 d; how-ever, no increase in the number of melanoblasts ormelanocytes was observed in the absence of ET-1,

-2, or -3. The increase in the number of melano-blasts or melanocytes was comparable with that ofmelanoblasts or melanocytes cocultured with sec-ondary keratinocytes in MDMDF or MDMD. Also,pure cultured primary melanoblasts in MDM werefurther cultured with MDMM supplemented withET-1, -2, or -3 from 14 d. A dramatic increase in thepercentage of melanocytes in the melanoblast-mela-nocyte population was observed after 7 d; however,no increase in the percentage of melanocytes wasobserved in the absence of ET-1, -2, or -3. Theincrease was comparable with that of melanocytescocultured with secondary keratinocytes in MDMM.Moreover, anti-ET-1, -2, and -3 antibodies inhibitedboth the proliferation of melanoblasts or melano-cytes in MDMDF or MDMD and the differentiationof melanocytes in MDMM in primary culture. Theseresults suggest that ET-1, -2, and -3 are one memberof the keratinocyte-derived factors that are involvedin regulating the proliferation and differentiation ofmouse epidermal melanocytes in primary culture.Key words: keratinocyte/melanoblast/melanogenesis/pig-mentation/skin. Journal of Investigative DermatologySymposium Proceedings 6:25±31, 2001

Mouse epidermal melanocytes are known todifferentiate around the time of birth fromundifferentiated precursors, melanoblasts(Hirobe, 1984) originated from neural crest cells(Rawles, 1947; Mayer, 1973). They increase in

number until 3 or 4 d after birth and then their numbers decrease(Quevedo et al, 1966; Takeuchi, 1968; Weiss and Zelickson, 1975;Hirobe, 1984). Of various strains of mice, C57BL/10JHir (B10)possesses the greatest number of epidermal melanoblasts andmelanocytes (Hirobe, 1987). B10 mouse is thought to be usefulfor the study on the regulation of the proliferation and

differentiation of neonatal mouse epidermal melanocytes, especiallyin vitro study. A serum-free culture system to optimize and maintainmelanoblasts and melanocytes from epidermal cell suspensions ofnewborn B10 mice was developed in this laboratory. Mouseepidermal melanoblasts preferentially proliferate in a melanoblast-de®ned medium (MDM), consisting of Ham's F-10 mediumsupplemented with insulin (Ins), bovine serum albumin (BSA),ethanolamine (EA), phosphoethanolamine (PEA), and sodiumselenite (SE). After 14 d, almost all keratinocytes that existpredominantly in the early stage die, and pure cultures ofmelanoblasts are obtained (Hirobe, 1992b). Epidermal melanoblastsdramatically increase in number in MDMDF consisting of MDMsupplemented with 0.5 mM dibutyryl adenosine 3¢:5¢-cyclicmonophosphate (DBcAMP) and 2.5 ng per ml basic ®broblastgrowth factor (bFGF) (Hirobe, 1992a). On the other hand,epidermal melanocytes are induced to differentiate in MDMMconsisting of MDM supplemented with 100 nM a-melanocyte-stimulating hormone (MSH) (Hirobe, 1992b). MDMD consistingof MDM supplemented with 0.5 mM DBcAMP induces theproliferation of melanocytes in addition to the differentiation of

1087-0024/01/$15.00 ´ Copyright # 2001 by The Society for Investigative Dermatology, Inc.

25

Manuscript received June 14, 2001; accepted for publication June 14,2001.

Reprint requests to: Dr. Tomohisa Hirobe, Division of Biology andOncology, National Institute of Radiological Sciences, Anagawa, Inage-ku, Chiba 263±8555, Japan. Email: thirobe@nirs.go.jp

Abbreviations: bFGF, basic ®broblast growth factor; DBcAMP,dibutyryl adenosine 3¢:5¢-cyclic monophosphate; ET, endothelin; MSH,melanocyte-stimulating hormone.

melanocytes (Hirobe, 1992b). By using these culture media, it hasbeen shown that keratiocytes present in the early stage of primaryculture are important for the regulation of the proliferation anddifferentiation of mouse epidermal melanocytes (Hirobe, 1994);however, molecular natures of the keratinocyte-derived factors arenot well de®ned. In the human skin, endothelin (ET)-1 is shown tobe synthesized in keratinocytes (Imokawa et al, 1992; Yohn et al,1993) and mitogenic for cultured epidermal melanocytes in thepresence of cAMP elevator (Yada et al, 1991; Swope et al, 1995).ET-2 and ET-3 are also known to stimulate the proliferation ofhuman melanocytes in an increased cAMP level (Yada et al, 1991).Moreover, ET-3 is known to stimulate the proliferation anddifferentiation of melanocytes from neural crest explants (Ono et al,1998). It is not known, however, whether ET-1, -2, and -3 are onemember of the keratinocyte-derived factors involved in regulatingthe proliferation and differentiation of mouse epidermal melano-cytes at the neonatal stage. This study was designed to test thispossibility by using the serum-free primary culture system.

MATERIALS AND METHODS

Mice All animals used in this study belonged to strain B10 of thehouse mouse, Mus musculus. They were given water and a commercialdiet, OA-2 (Clea Japan, Tokyo, Japan) ad libitum. They were maintainedat 24 6 1°C with 40%±60% relative humidity: 12 h of ¯uorescent lightwere provided daily.

Melanocyte primary culture The sources of tissue for the culture ofmelanoblasts/melanocytes were dorsal skin from 0.5-d-old mice. Unlessstated otherwise, all reagents were purchased from Sigma (St. Louis,MO). The method for obtaining epidermal cell suspensions was reportedpreviously (Hirobe, 1992a). Disaggregated epidermal cell suspensionswere pelleted by centrifugation and suspended in Ham's F-10 medium(Gibco, Grand Island, NY). The cell pellet after centrifugation wasresuspended in a MDM [F-10 plus 10 mg Ins (bovine) per ml, 0.5 mgBSA (Fraction V) per ml, 1 mM EA, 1 mM PEA, 10 nM SE, 100 Upenicillin G (PG) per ml, 100 mg streptomycin sulfate (SS) per ml, 50 mggentamycin sulfate (GS) per ml, and 0.25 mg amphotericin B (AB) perml], MDMM, MDMD, and MDMDF. The same lots of thesesupplements were used in this study. The cells in the epidermal cellsuspension were counted in a hemocytometer chamber and plated ontotype I collagen (Becton Dickinson, Bedford, MA)-coated dishes at aninitial density of 1 3 106 cells per 35 mm dish (1.04 3 105 cells percm2) except an experiment performed to obtain the data shown in Fig 5[0.5 3 106 cells per 35 mm dish (0.52 3 105 cells per cm2)]. Cultureswere incubated at 37°C in a humidi®ed atmosphere composed of 5%CO2 and 95% air (pH 7.2). Medium was replaced by fresh medium fourtimes a week. After 14 d, almost pure cultures of melanoblasts ormelanocytes were obtained. Various doses (0.001, 0.01, 0.1, 1, 10, and100 nM) of ET-1, -2, and -3 (human synthetic) were supplemented toMDM, MDMM, MDMD, and MDMDF both from initiation ofprimary culture and after death of keratinocytes (14 d). Anti-ET-1, -2,and -3 antibodies (IBL, Maebashi, Japan) as well as rabbit IgG (rIgG,IBL) for control were added to MDM, MDMM, MDMD, andMDMDF from initiation of primary culture at different doses of 0.25,

Figure 1. Effects of ET-1 and anti-ET-1antibody on the proliferation and differentia-tion of mouse epidermal melanocytes inserum-free primary culture. Epidermal cellsuspensions derived from mouse skin werecultured with MDMM (A) or with MDMM plusET-1 (10 nM, B). After 14 d, melanocytescultured with ET-1 (B) possessed enlargedcytoplasm and increased pigmentation. Epidermalcell suspensions were cultured with MDMD (C)or with MDMD plus ET-1 (10 nM, D). After14 d, pure cultures of melanocytes that werepolygonal or epithelioid in morphology wereobtained. ET-1 (D) increased the number ofmelanocytes. Moreover, melanocytes culturedwith ET-1 possessed enlarged cytoplasm andincreased pigmentation (D). Epidermal cellsuspensions were also cultured with MDMDF plusrIgG (250 ng per ml; control, E) or withMDMDF plus anti-ET-1 antibody (250 ng perml, F). After 14 d, pure cultures of melanoblastsand melanocytes were obtained. Anti-ET-1antibody dramatically decreased the number ofmelanoblasts (F). Phase-contrast microscopy. Scalebar: 100 mm.

26 HIROBE JID SYMPOSIUM PROCEEDINGS

2.5, 25, and 250 ng per ml, and tested for their effects on theproliferation and differentiation of melanocytes.

Primary culture of keratinocytes Epidermal cell suspensions weresimilarly obtained from mice of 5.5±7.5 d of age. The cell pellet aftercentrifugation was suspended in a Ca2+-free Eagle's minimum essentialmedium (MEM; Gibco) and centrifuged at 800 3 g for 5 min. The cellpellet was resuspended in a keratinocyte-de®ned medium (KDM)consisting of Ca2+-free MEM supplemented with MEM-nonessentialamino acid solution (Gibco), 10 mg Ins per ml, 0.5 mg BSA per ml,1 mM EA, 1 mM PEA, 10 nM SE, 10 ng epidermal growth factor(human recombinant, Becton Dickinson) per ml, 1 mM hyrocortisone,1 mM dexamethasone, 0.03 mM CaCl2, 100 U PG per ml, 100 mg SSper ml, 50 mg GS per ml, and 0.25 mg AB per ml. The cells in theepidermal cell suspension were plated onto type I collagen-coated dishes.Initial density was 2 3 106 cells per 35 mm dish (2.08 3 105 cells percm2). After 1±3 d, pure keratinocytes were obtained.

Coculture of melanoblasts/melanocytes and keratino-cytes Primary keratinocytes cultured in KDM for 1 d were treatedwith a solution of 0.05% trypsin (Gibco) and 0.02% ethylene-diaminetetraacetate (EDTA) in CMF-PBS at 37°C for 10 min. Aftertrypsinization was inhibited by addition of 2000 U soybean trypsininhibitor per ml, the cell suspensions were centrifuged at 800 3 g for5 min. After this centrifugation, the cell pellet was resuspended inMDM, MDMM, MDMD, and MDMDF at a density of 2 3 105 cellsper 35 mm dish (2.08 3 104 cells per cm2) and seeded into primarymelanoblasts or melanocytes that were cultured in MDM/MDMDF orMDMD for 14 d, and cultured for a further 14 d.

Assay of melanocyte proliferation and differentiation Thenumber of melanoblasts and melanocytes was determined per dish byphase-contrast and bright-®eld microscopy, and the calculation was basedon the average number of cells from 10 randomly chosen microscopic

®elds covering an area of 0.581 mm2. Bipolar, tripolar, dendritic,polygonal, or epithelioid cells, as seen by phase-contrast, whichcontained brown or black pigment granules, as observed by bright-®eldmicroscopy, were scored melanocytes. These cells were con®rmed asmelanocytes by DOPA cytochemistry (Hirobe, 1992a). In contrast,bipolar, tripolar, dendritic, or polygonal cells, as seen by phase-contrast,which contained no pigments, as observed by bright-®eld microscopy,were scored melanoblasts. The ``melanoblast'' is de®ned here as anonpigmented cell that possesses no tyrosinase activity. Some of themelanoblasts cultured in MDM were stained by combined DOPA-premelanin reaction (combined DOPA-ammoniacal silver nitratestaining; Mishima, 1960; Hirobe, 1992b). In contrast, almost all of themelanoblasts cultured in MDMDF were stained by the combinedDOPA-premelanin reaction (Hirobe, 1992a). This preferential stainingreveals undifferentiated, nonpigmented melanoblasts that contain stage Iand II melanosomes in addition to tyrosinase-containing differentiatedmelanocytes (Mishima, 1964; Hirobe, 1982).

The statistical signi®cance of the differences in the number ofmelanoblasts and melanocytes or in the percentage of melanocytes in themelanoblast-melanocyte population was determined by Student's t testfor comparisons of groups of equal size.

RESULTS

Effects of ET on the proliferation and differentiation ofmelanocytes in primary culture Within 1±2 d of initiationof culture with MDM, undifferentiated melanoblasts that arebipolar, tripolar, or dendritic in morphology were in contact withadjacent keratinocyte colonies via a dendrite. After 3 d, thekeratinocyte colonies increased in size and the melanoblastsgradually increased in number. After 8 d, the keratinocytecolonies were decreased, and by 14 d, the cultures contained

Figure 2. Effects of ET on the proliferation of melanocytes inMDMD. Epidermal cell suspensions were cultured with MDMD(Control, s) or MDMD plus 10 nM ET-1 (d), -2 (n), or -3 (m) for14 d. The number of melanoblasts and melanocytes cultured with ETexceeded that of the control. The differences at 7 and 14 d arestatistically signi®cant (p < 0.05). There was no difference betweenET-1, -2, and -3. The epidermal cell suspensions of the four differentgroups were derived from the same litter of mice. The data are theaverages of results from triplicate experiments. Each experiment wasperformed with different litters of mice. Bars indicate standard errors ofthe mean (SEM) and are shown only when they were larger thansymbols.

Figure 3. Effects of ET and keratinocytes on the differentiation ofmelanocytes in MDMM. Epidermal cell suspensions were culturedwith MDM for 14 d. Pure cultured melanoblasts were obtained. Theywere further cultured with MDMM (Control, s) or MDMM plus ET-1(d), -2 (n), -3 (m), or keratinocytes (K, h) from 14 d (arrow). Thepercentage of melanocytes in the melanoblast-melanocyte populationcultured with ET and K exceeded that of the control. The differences at21 and 28 d are statistically signi®cant (p < 0.05). There was nodifference between ET-1, -2, -3, and K. The epidermal cell suspensionsof the ®ve different groups were derived from the same litter of mice.The data are the averages of results from triplicate experiments. Eachexperiment was performed with different litters of mice. Bars indicateSEM and are shown only when they were larger than symbols.

VOL. 6, NO. 1 NOVEMBER 2001 ENDOTHELINS AND MELANOCYTE DEVELOPMENT 27

mostly melanoblasts. When the epidermal cell suspensions werecultured with MDM supplemented with ET-1, -2, and -3, a similartendency of keratinocyte proliferation as well as melanoblastproliferation was observed, except that most of the melanoblastscultured with ET possessed long dendrites. There was no differencebetween ET-1, -2, and -3.

Within 3±4 d of initiation of culture with MDMM, pigment-producing differentiated melanocytes appeared around keratinocytecolonies and almost all cells differentiated around 7±9 d. After 14 d,almost all keratinocytes died and pure cultures of differentiatedmelanocytes were obtained, but no stimulation of melanocyteproliferation was observed. When the epidermal cell suspensionswere cultured with MDMM supplemented with ET, a similartendency of keratinocyte proliferation as well as melanocyteproliferation and differentiation was observed; however, melano-cytes cultured with ET possessed increased dendricity, enlargedcytoplasm (Fig 1A, B), and increased pigmentation (Fig 1A, B).There was no difference between ET-1, -2, and -3.

Within 3±4 d of initiation of culture with MDMD, pigment-producing differentiated melanocytes appeared around keratinocytecolonies and rapidly increased in number, and almost all cellsdifferentiated around 7±9 d. After 14 d, almost all keratinocytesdied and pure cultures of many differentiated melanocytes wereobtained. When the epidermal cell suspensions were cultured withMDMD supplemented with ET, a similar tendency of keratinocyteproliferation and melanocyte differentiation was observed; how-ever, the number of melanocytes dramatically increased in aconcentration-dependent manner. Numerous mitotic melanocyteswere frequently observed. Maximal effect was observed at a dose of

10 nM. The number of melanocytes cultured with 10 nM ET at 7and 14 d greatly exceeded the control (Fig 2). The differenceswere statistically signi®cant (p < 0.05). Moreover, melanocytescultured with ET possessed increased dendricity, enlarged cyto-plasm (Fig 1C, D), and increased pigmentation (Fig 1C, D).There was no difference between ET-1, -2, and -3.

Undifferentiated melanoblasts were observed around keratino-cyte colonies within 1 d of initiation of culture with MDMDF, anddramatically increased after 3±4 d. After 14 d, almost all keratino-cytes died, and pure cultures of numerous melanoblasts (~90%) andmelanocytes (~10%) were obtained. When the epidermal cellsuspensions were cultured with MDMDF supplemented with ET, asimilar tendency of keratinocyte proliferation and melanoblastproliferation was observed, except that most of the melanoblastscultured with ET possessed long dendrites (number of melanoblastsand melanocytes in control and experiments: 7 d, 17±20 3 104;14 d, 72±80 3 104; percentage of melanocytes in the melanoblast-melanocyte population in control and experiments: 7 d, 6%±8%;14 d, 6%±9%). There was no difference between ET-1, -2, and -3.

Effects of ET on the proliferation and differentiation ofmelanocytes with or without keratinocytes ETsupplemented to MDM from 14 d in primary culture broughtabout no further increase in the number of melanoblasts; however,melanoblasts cocultured with keratinocytes slowly proliferatedaround keratinocyte colonies. Keratinocytes gradually decreasedin number and disappeared by 28 d (14 d after coculture). Thedifferences in the number of melanoblasts at 21 and 28 d betweencontrol and experiments were statistically signi®cant (p < 0.05;21 d, 1.8-fold increase; 28 d, 2.2-fold increase).

Figure 4. Effects of ET and keratinocytes on the proliferation ofmelanocytes in MDMD. Epidermal cell suspensions were culturedwith MDMD for 14 d. Pure cultured melanocytes were obtained. Theywere further cultured with MDMD (Control, s) or MDMD plus ET-1(d), -2 (n), -3 (m), or keratinocytes (K, h) from 14 d (arrow). Thenumber of melanocytes cultured with ET and K exceeded that ofcontrol. The differences at 21 and 28 d are statistically signi®cant (p< 0.05). There was no difference between ET-1, -2, -3, and K. Theepidermal cell suspensions of the ®ve different groups were derived fromthe same litter of mice. The data are the averages of results fromtriplicate experiments. Each experiment was performed with differentlitters of mice. Bars indicate SEM and are shown only when they werelarger than symbols.

Figure 5. Effects of ET on the proliferation of melanoblasts inMDMDF. Epidermal cell suspensions were cultured with MDMDF for14 d. Pure cultured melanoblasts (c. 90%) and melanocytes (c. 10%) wereobtained. They were further cultured with MDMDF (Control, s) orMDMDF plus ET-1 (d), -2 (n), -3 (m), or keratinocytes (K, h) from14 d (arrow). The number of melanoblasts and melanocytes cultured withET and K exceeded that of control. The differences at 21 and 28 d arestatistically signi®cant (p < 0.05). There was no difference betweenET-1, -2, -3, and K. The epidermal cell suspensions of the ®ve differentgroups were derived from the same litter of mice. The data are theaverages of results from triplicate experiments. Each experiment wasperformed with different litters of mice. Bars indicate SEM and areshown only when they were larger than symbols.

28 HIROBE JID SYMPOSIUM PROCEEDINGS

ET supplemented to MDMM from 14 d brought about nofurther increase in the number of melanoblasts derived fromprimary culture of epidermal cell suspensions in MDM for 14 d,but induced the differentiation (Fig 3) of melanoblasts intomelanocytes in a concentration-dependent manner. Maximal effectwas observed at a dose of 10 nM. The differences in the percentageof melanocytes at 21 and 28 d between control and ET werestatistically signi®cant (p < 0.05). The increase was comparable withthat of melanocytes cocultured with secondary keratinocytes(Fig 3). There was no difference between ET-1, -2, and -3. Incontrast, secondary keratinocytes cocultured from 14 d in MDMMbrought about further increases in the number of melanoblasts andmelanocytes (p < 0.05; 21 d, 1.9-fold increase; 28 d, 2.5-foldincrease).

ET supplemented to MDMD from 14 d brought about a furtherincrease in the number of melanocytes (Fig 4) in a concentration-dependent manner. Maximal effect was observed at a dose of10 nM. The differences in the number of melanocytes at 21 and28 d between control and ET were statistically signi®cant (p< 0.05). The increase was comparable with that of melanocytescocultured with secondary keratinocytes from 14 d. There was nodifference between ET-1, -2, and -3.

ET supplemented to MDMDF from 14 d brought about afurther increase in the number of melanoblasts and melanocytes(Fig 5) derived from epidermal cell suspensions of 0.5 3 106 cellsper 35 mm dish in a concentration-dependent manner. Maximaleffect was observed at a concentration of 10 nM. The differences inthe number of melanoblasts and melanocytes between control andET at 21 and 28 d were statistically signi®cant (p < 0.05), and the

increase in the number of melanoblasts and melanocytes wascomparable with that of melanoblasts and melanocytes coculturedwith secondary keratinocytes from 14 d (Fig 5). There was nodifference between ET-1, -2, and ±3; however, the percentage ofmelanocytes in the melanoblast-melanocyte population did notdiffer between all groups (9%±13%).

Effects of anti-ET-1, -2, and -3 antibodies on theproliferation and differentiation of melanocytes Anti-ET-1, -2, and -3 antibodies supplemented to MDM failed toaffect the proliferation of melanoblasts in primary culture at alldoses tested.

Anti-ET-1, -2, and -3 antibodies supplemented to MDMMfailed to affect the number of melanoblasts and melanocytes, butthey inhibited the differentiation (Fig 6) of melanocytes in aconcentration-dependent manner. Maximal effect was observed ata concentration of 250 ng per ml (Fig 6). The differences in thepercentages of melanocytes in the melanoblast-melanocyte popu-lation at 7 and 14 d between control and experiments werestatistically signi®cant (p < 0.05). There was no difference betweenanti-ET-1, -2, and -3 antibodies.

Anti-ET-1, -2, and -3 antibodies supplemented to MDMDinhibited both proliferation (Fig 7) and differentiation (Fig 8) ofmelanocytes in a concentration-dependent manner. Maximaleffects were observed at a dose of 250 ng per ml (Figs 7, 8).The differences in the number of melanocytes as well as in thepercentage of melanocytes in the melanoblast-melanocyte popula-tion at 7 and 14 d between control and experiments werestatistically signi®cant (p < 0.05). There was no difference betweenanti-ET-1, -2, and -3 antibodies.

Figure 6. Effects of anti-ET-1, -2, and -3 antibodies on thedifferentiation of melanocytes in MDMM. Epidermal cellsuspensions were cultured with MDMM plus rIgG (250 ng per ml;Control, s) or MDMM plus anti-ET-1 (250 ng per ml, d), -2 (250 ngper ml, n), or -3 (250 ng per ml, m) antibody. The percentage ofmelanocytes in the melanoblast-melanocyte population cultured withanti-ET-1, -2, or -3 antibody was lowered as compared with control.The differences at 7 and 14 d were statistically signi®cant (p < 0.05).There was no difference between anti-ET-1, -2, and -3 antibodies. Theepidermal cell suspensions of the four different groups were derived fromthe same litter of mice. The data are the averages of results fromtriplicate experiments. Each experiment was performed with differentlitters of mice. Bars indicate SEM and are shown only when they werelarger than symbols.

Figure 7. Effects of anti-ET-1, -2, and -3 antibodies on theproliferation of melanocytes in MDMD. Epidermal cell suspensionswere cultured with MDMD plus rIgG (250 ng per ml; Control, s) orMDMD plus anti-ET-1 (250 ng per ml, d), -2 (250 ng per ml, n), or -3 (250 ng per ml, m) antibody. The number of melanocytes culturedwith anti-ET-1, -2, or -3 antibody was lowered as compared withcontrol. The differences at 7 and 14 d were statistically signi®cant (p< 0.05). There was no difference between anti-ET-1, -2, and -3antibodies. The epidermal cell suspensions of the four different groupswere derived from the same litter of mice. The data are the averages ofresults from triplicate experiments. Each experiment was performed withdifferent litters of mice. Bars indicate SEM and are shown only whenthey were larger than symbols.

VOL. 6, NO. 1 NOVEMBER 2001 ENDOTHELINS AND MELANOCYTE DEVELOPMENT 29

Anti-ET-1, -2, and -3 antibodies supplemented to MDMDFgreatly inhibited the proliferation of melanoblasts in a concentra-tion-dependent manner (Fig 9). Maximal effect was observed at250 ng per ml (Figs 1E, F, 9). The differences in the numbers ofmelanoblasts and melanocytes at 7 and 14 d between control andexperiments were statistically signi®cant (p < 0.05). In contrast,antibodies failed to affect the percentage of melanocytes in themelanoblast-melanocyte population. There was no differencebetween anti-ET-1, -2, and -3 antibodies.

These results suggest that ET are one member of thekeratinocyte-derived factors involved in regulating the proliferationand differentiation of mouse epidermal melanoblasts and melano-cytes in the presence of cAMP elevator and/or bFGF.

DISCUSSION

ET-1, which was ®rst discovered as an endothelium-derived factorwith vasoconstrictive effects (Yanagisawa et al, 1988), is known tobe an important cytokine involved in regulating the function ofvarious mammalian cells. ET-1 is synthesized by cells of varioustypes and acts as a paracrine regulator of various target cells. In thehuman skin, ET-1 is shown to be synthesized in keratinocytes(Imokawa et al, 1992; Yohn et al, 1993) and mitogenic for culturedepidermal melanocytes in the presence of cAMP elevator (Yada etal, 1991; Swope et al, 1995). It is also known that each of threeisoforms of ET, ET-1, -2, and -3, is encoded by a gene (Inoue et al,1989). ET-2 and ET-3 are also known to stimulate the proliferationof human melanocytes in an increased cAMP level (Yada et al,1991). Moreover, ET are known to bind to two types of receptors,ETRA and ETRB. ETRA possesses a greatest af®nity for ET-1,

and ETRB binds ET-1, -2, and -3 equally (Sakurai et al, 1990). Inmice, mutations at the ET-3 locus are known to interfere withnormal generation of melanocytes (Baynash et al, 1994). It is alsoknown that ET-3 stimulates the proliferation and differentiation ofmelanocyte from neural crest explants (Ono et al, 1998); however,it has not been known whether ET-1, -2, and -3 are involved inregulating the proliferation and differentiation of mouse epidermalmelanocytes at the neonatal stage. These results suggest that ET areone member of the keratinocyte-derived factors involved inregulating the proliferation and differentiation of mouse epidermalmelanoblasts and melanocytes. Other factors are thought to beinvolved in regulating the proliferation of melanoblasts andmelanocytes, however, as anti-ET-1, -2, and -3 antibodies failedto completely inhibit the proliferation of melanoblasts andmelanocytes in primary culture, and secondary keratinocytesinduced the proliferation of melanoblasts in MDM as well as theproliferation of melanocytes in MDMM after 14 d, despite thefailure of ET in inducing their proliferation. Indeed, a recent studyusing transgenic mice that overexpress the steel factor (SLF) in thebasal layer of the epidermis suggests that SLF is involved inregulating the proliferation of melanocytes in the mouse epidermis(Kunisada et al, 1998). It is probable that numerous keratinocyte-derived factors other than ET and SLF are involved in regulatingthe proliferation of mouse epidermal melanoblasts and melanocytes.The reason why ET failed to stimulate the proliferation ofmelanoblasts in MDMDF in primary culture, despite the stimula-tion of melanocyte proliferation by ET in MDMD in primaryculture, seems to be that large amounts of ET were released fromkeratinocytes in MDMDF. It is possible that the amounts of ETreleased into MDMDF are much greater than those released into

Figure 8. Effects of anti-ET-1, -2, and -3 antibodies on thedifferentiation of melanocytes in MDMD. Epidermal cell suspensionswere cultured with MDMD plus rIgG (250 ng per ml; Control, s) orMDMD plus anti-ET-1 (250 ng per ml, d), -2 (250 ng per ml, n), or -3 (250 ng per ml, m) antibody for 14 d. The percentage of melanocytesin the melanoblast-melanocyte population cultured with anti-ET-1, -2,or -3 antibody was lowered as compared with control. The differences at7 and 14 d were statistically signi®cant (p < 0.05). There was nodifference between anti-ET-1, -2, and -3 antibodies. The epidermal cellsuspensions of the four different groups were derived from the samelitter of mice. The data are the averages of results from triplicateexperiments. Each experiment was performed with different litters ofmice. Bars indicate SEM and are shown only when they were largerthan symbols.

Figure 9. Effects of anti-ET-1, -2, and -3 antibodies on theproliferation of melanoblasts in MDMDF. Epidermal cellsuspensions were cultured with MDMDF plus rIgG (250 ng per ml;Control, s) or MDMDF plus anti-ET-1 (250 ng per ml, d), -2 (250 ngper ml, n), or -3 (250 ng per ml, m) antibody for 14 d. The number ofmelanoblasts and melanocytes cultured with anti-ET-1, -2, or -3antibody was greatly lowered as compared with control. The differencesat 7 and 14 d were statistically signi®cant (p < 0.05). There was nodifference between anti-ET-1, -2, and -3 antibodies. The epidermal cellsuspensions of the four different groups were derived from the samelitter of mice. The data are the averages of results from triplicateexperiments. Each experiment was performed with different litters ofmice. Bars indicate SEM and are shown only when they were largerthan symbols.

30 HIROBE JID SYMPOSIUM PROCEEDINGS

MDMD. One possibility exists that bFGF greatly stimulates therelease of ET from keratinocytes.

The induction of the differentiation of cultured mouse epidermalmelanocytes by a-MSH has also been reported to require thepresence of keratinocytes (Hirobe, 1992b). The present resultssuggest that ET are one member of the keratinocyte-derived factorsinvolved in regulating the differentiation of cultured mouseepidermal melanocytes. The mechanism of cooperation of ET-1and a-MSH can be at least in part explained by recent ®ndings byTada et al (1998) that upregulation of the mRNA of a-MSHreceptor (MC1R) of cultured human melanocytes was caused bybrief treatment with ET-1. Anti-ET-1, -2, and -3 antibodies,however, failed to inhibit completely the differentiation ofmelanocytes in primary culture in this study. Thus, unde®nedkeratinocyte-derived factors other than ET might be involved inregulating the differentiation of melanocytes. Indeed, the recentstudy by Kunisada et al (1998) suggests that SLF is involved inregulating the differentiation of melanocytes in the mouse epider-mis.

In this study, ET induced the differentiation of melanocytes inthe presence of a-MSH; however, ET failed to induce theproliferation of melanocytes even in the presence of a-MSH. Onthe contrary, ET induced both proliferation and differentiation ofmelanocytes in the presence of DBcAMP. These ®ndings can beexplained by assuming that keratinocytes cultured with 0.5 mMDBcAMP produce different factors from those cultured with100 nM a-MSH. Melanocyte growth factors derived fromkeratinocytes cultured with 0.5 mM DBcAMP might be involvedin regulating the proliferation of melanocytes in cooperation withcAMP and ET.

These results that ET-1, -2, and -3 are equally involved inregulating the proliferation and differentiation of mouse epidermalmelanocytes suggest the possibility that neonatal epidermalmelanocytes express ETRB at the newborn stage. Interactionsbetween ETRB and ET are thought to be mainly involved inregulating the proliferation and differentiation of melanocytes inthe neonatal epidermis; however, their molecular mechanismsremain to be investigated in a future study.

The induction of the proliferation and differentiation of mouseepidermal melanoblasts and melanocytes might require the presenceof cAMP and tyrosine kinase (Coughlin et al, 1988) signalingpathways, as ET failed to induce the proliferation and differenti-ation of melanoblasts and melanocytes in the absence of a-MSH,DBcAMP, and bFGF. The interactions between a-MSH andMC1R are known to be mediated by the cAMP-protein kinase A(PKA) signaling pathway. These three factors, cAMP (PKApathway), bFGF (tyrosine kinase pathway), and ET [protein kinaseC (PKC) pathway] are thought to form a paracrine network thatregulates the proliferation and differentiation of mouse melanocytesin the epidermis. ET-1 stimulates 1, 4, 5-inositol-triphosphateformation and intracellular calcium mobilization in addition to theactivation of PKC and nonreceptor tyrosine kinases of humanepidermal melanocytes (Yada et al, 1991). It is also reported thatET-1 interacted synergistically with a-MSH and bFGF to stimulatethe proliferation of human epidermal melanocytes in culture(Swope et al, 1995). Moreover, human epidermal melanocytes inculture were reported to express ETRB and respond to both ET-1and -3 with a concentration-dependent increase in the proliferationand a decrease or increase in melanogenesis (Tada et al, 1998). Brieftreatment of melanocytes with ET-1 caused upregulation ofmRNA of MC1R, but did not alter the mRNA level of ETRB(Tada et al, 1998). It remains to be investgated in a future studywhat molecular mechanisms of the cross talk of the three signalingpathways are involved in regulating the proliferation and differen-tiation of neonatal mouse epidermal melanocytes.

CONCLUSION

This study was designed to investigate the role of ET in theregulation of the proliferation and differentiation of neonatal mouse

epidermal melanoblasts and melanocytes by adding ET to culturemedia in the presence or absence of keratinocytes. The resultssuggest that ET are one member of the keratinocyte-derived factorsinvolved in regulating the proliferation and differentiation ofmelanoblasts and melanocytes in the presence of cAMP elevatorand/or bFGF.

This work was in part supported by a grant from the Science and Technology

Agency, Japan.

REFERENCES

Baynash AG, Hosoda K, Giaid A, Richardson JA, Emoto N, Hammer RE,Yanagisawa M: Interaction of endothelin-3 with endothelin-B receptor isessential for development of epidermal melanocytes and enteric neurons. Cell79:1277±1285, 1994

Coughlin SR, Barr PJ, Cousens LS, Fretto LJ, Williams LT: Acidic and basic ®broblastgrowth factors stimulate tyrosine kinase activity in vivo. J Biol Chem263:988±993, 1988

Hirobe T: Origin of melanosome structures and cytochemical localizations oftyrosinase activity in differentiating epidermal melanocytes of newborn mouseskin. J Exp Zool 224:355±363, 1982

Hirobe T: Histochemical survey of the distribution of the epidermal melanoblasts andmelanocytes in the mouse during fetal and postnatal periods. Anat Rec208:589±594, 1984

Hirobe T: Genetic control of the population size of the melanocyte in the mouseepidermis. Jpn J Genet 62:149±158, 1987

Hirobe T: Basic ®broblast growth factor stimulates the sustained proliferation of mouseepidermal melanoblasts in a serum-free medium in the presence of dibutyrylcyclic AMP and keratinocytes. Development 14:435±445, 1992a

Hirobe T: Melanocyte stimulating hormone induces the differentiation of mouseepidermal melanocytes in serum-free culture. J Cell Physiol 152:337±345, 1992b

Hirobe T: Keratiocytes are involved in regulating the developmental changes in theproliferative activity of mouse epidermal melanoblasts in serum-free culture. DevBiol 161:59±69, 1994

Imokawa G, Yada Y, Miyagishi M: Endothelins secreted from human keratinocytesare intrinsic mitogens for human melanocytes. J Biol Chem 267:24675±24680,1992

Inoue A, Yanagisawa M, Kimura S, Kasuya Y, Miyauchi T, Goto K, Masaki T: Thehuman endothelin family. three structurally and pharmacologically distinctisopeptides predicted by three separate genes. Proc Natl Acad Sci USA86:2863±2867, 1989

Kunisada T, Yoshida H, Yamazaki H, et al: Transgene expression of steel factor in thebasal layer of epidermis promotes survival, proliferation, differentiation andmigration of melanocyte precursors. Development 125:2915±2923, 1998

Mayer TC: The migratory pathway of neural crest cells into the skin of mouseembryos. Dev Biol 34:39±46, 1973

Mishima Y: New technic for comprehensive demonstration of melanin, premelanin,and tyrosinase sites±combined dopa-premelanin reaction. J Invest Dermatol34:355±360, 1960

Mishima Y: Electron microscopic cytochemistry of melanosomes and mitochondria. JHistochem Cytochem 12:784±790, 1964

Ono H, Kawa Y, Asano M, et al: Development of melanocyte progenitors in murinesteel mutant neural crest explants cultured with stem cell factor, endothelin-3, orTPA. Pigment Cell Res 11:291±298, 1998

Quevedo WC Jr, Youle MC, Rovee DT, Bienieki TC: The developmental fate ofmelanocytes in murine skin. In: Della Porta G, Muhlbock O., eds. Structure andControl of the Melanocyte. Berlin: Springer-Verlag, 1966: pp 228±241

Rawles ME: Origin of pigment cells from the neural crest in the mouse embryo.Physiol Zool 20:248±266, 1947

Sakurai T, Yanagisawa M, Takuwa Y, Miyazaki H, Kimura S, Goto K, Masaki T:Cloning of a cDNA encoding a non-isopeptide-selective subtype of theendothelin receptor. Nature 348:732±735, 1990

Swope VB, Medrano EE, Smalara D, Abdel-Malek Z: Long term proliferation ofhuman melanocytes is supported by the physiologic mitogens a-melanotropin,endothelin-1, and basic ®broblast growth factor. Exp Cell Res 217:453±459,1995

Tada A, Suzuki I, Im S, et al: Endothelin-1 is a paracrine growth factor that modulatesmelanogenesis of human melanocytes and participates in their responses toultraviolet radiation. Cell Growth Differ 9:575±584, 1998

Takeuchi T: Genetic analysis of a factor regulating melanogenesis in the mousemelanocyte. Jpn J Genet 43:249±256, 1968

Weiss LM, Zelickson AS: Embryology of the epidermis: Ultrastructural aspects. III.Maturation and primary appearance of dendritic cells in the mouse withmammalian comparisons. Acta Derm Venereol 55:431±442, 1975

Yada Y, Higuchi K, Imokawa G: Effects of endothelins on signal transduction andproliferation in human melanocytes. J Biol Chem 266:18352±18357, 1991

Yanagisawa M, Kurihara H, Kimura S, et al: A novel potent vasoconstrictor peptideproduced by vascular endothelial cells. Nature 332:411±415, 1988

Yohn JJ, Morelli JG, Walchak SJ, Rundell KB, Norris DA, Zamora MR: Culturedhuman keratinocytes synthesize and secrete endothelin-1. J Invest Dermatol100:23±26, 1993

VOL. 6, NO. 1 NOVEMBER 2001 ENDOTHELINS AND MELANOCYTE DEVELOPMENT 31