RAPID COMMUNICATIONS

Hum.an Melanogenesis is Stimulated by Diacylglycerol

Philip R. Gordon, Ph.D., and Barbara A. Gilchrest, M.D. USDA Human Nutrition Resc:arch Center on Aging:lt Tufts University. Boston, MassacbuS(:ttS, U.S.A.

The intracellular signal pathways that mediate pigmentation in human skin are unknown. We now report that a diacyl­glycerol (DAG) analogue 1-oleoyl-2-acetyl-glycerol (OAG) 25-100 J.lm strikingly in creased the melanin content of cul­tured human melanocytes in a dose dependent manner with­out altering growth rate. T he pigment increase occurred within 24 h. was accompanied by increased incorporation of

Cutaneous pigmentation is of great psychosocial and cosmetic importance and is me skin's major defense against carcinogenic ul traviolet radiation. Extensive investigation over 4 decades has elucidated rht" ge­netic characteristics and histologic commitants o[

basal pigmentation and the tanning response in normal skin as well as in many disorders of pigmentation. but the biochemical mecha­nisms responsib le [or human melanogenesis are virtually unstudied.

Considerable ev idence implicates the cyclic 3', 5' -adenosine monophosphate (cAMP) pathway in murine melanoma cell mela­nogenesis (1-4]. However, to date it has not been possible to estab­lish a role for cAMP in human melanogenesis. lncreased pigment production by cultured human epidermal melanocytes following ultraviolet irradiation, the major physiologic stimulus [or human nlelanogenesis. is indeed not associated with increased intracellular cAMP content, while melanogenic stimulation of 591 murine md­anoma cells with either 3-isobutyl-I-merhyLunthine (IBMX) or melanocyte stimulating hormone produces a readily detectable cAMP elevation 15). Further. the most striking and mostexrensively studied melanogenic stimulus in the murine melanoma: model, e-AMP dependent melanocyte stimulating hormone [6-8]. has not been convincingly demonstrated to alter pigment content of human melanocytes (Refs 1 - 4 and Gordon and Gilchrest, unpublished obscrv:ujons), and direcr addition of me dibutryl or 8-bromo ana­logues of c-AMP to cultured human ruclanocyres has no effect on pigmentation [9J. We therefore explored the possibility that human

M:lIluscript received M.ay 17. 1989: acce:'pt'ed for publication June:' 20, 1989.

This work was supported by NlH grant C A45687 . the USDA Agricul­tural Research S('rvice. and the Shiscido Company, Ltd.

R<!'prim requ~ts to: Dr. Barbara A. Gilchrest. USDA H uman Nurrition Research Center 011 Aging at Tufts Universiry . 711 Washington Street, BoslOn, MA 02111.

Abbreviations: cAMP: cyclic 3'.5' -adenosine monophosphate OAG: 1.2-diacylglycerol DOPA: L-3.4-dihydoroxyphenyl alanine JBMX: isobutyJm~thylxanthjnc OAG: l-01eoyl-2-acetyl-glycerol PKC: prorc:in kinase C T C A: trichloroacetic acid

, he melanin precursor L-3,4-dihydro.xyphenyl alan ine (DOPA), required new protein synthesis, and was com ­ple tely blocked by the protein kinase C (PKC) inhibitors H-7 and sphingosine. A PKC-inactive DAG isomer had no effect on melanin per cell. These results implicate protein kinase C and its effector DAG in melanogenesis. ) 1m1est Derrn arol 93:700-702, 19159

melanogenesis might be regulated by 1.2-diacylglycerol (DAG), another major second messenger for intracellular signal t ransduc­tion.

Specific stimuli activate phospholipase C, releasing DAG from cell membranes, which in tum activates protein kinase C (PKC) [t OJ. A water dispersable analogue of DAG. 1-0Ieoyl-2-acetyl-glyc­eraJ (OAG). previously shown [0 activate the PKC in a variety of cultured cells 110,11]. was therefore tested in human melanocyte cultures established and main.cained in a hormone supplemented medium lacking phorbol esters.

MATERIALS AND METHODS

Cell Culture Newborn foreskin~derived melanocyte cultures werc established in hormone-supplemented serum-free medium f12J. At second passage. paired 35 mm dishes were: plated with 2 X 10· cel ls and maintained in hormone supplemented Dulbecco's modified Eagle's medium [5]. and 2-4 d later OAG 25-200 JiM or an equal volume of diluent was added. After 7 d. dishes were har­vested. ~nd cell number (Coulter Counter) and melanjn content in ceLl extracts were determined as previollsly described IS} in triplicate [or each concentration.

Melanin Assays Melanin com em was rounnely calculated from [he absorption of light at 475 om (A .. 7S) by cellular extracts with comparison to a standard curve for synthetic melanin IS}. In some experiments absorption spectra from the ultravio let through the visible light range were also obtained. Cell pellets were solubilized in 1 N sodium hydro:dde and vOT(t"xed vigorously for 15 min. Spectra were obtained in a Cary Model 219 (Varian) spectropho­tomer at a rate of 10m/sec and a range setting o[ 0. 1. An autobase­line was set using 1 N NaOH.

Melanin synthesis rate was measured by incorporation of radio­labeled L-3,4-dihydroxyphenyl alanine (DOPA) precurser into melanin 151. Five hours after treatment with test compounds, I4C_ DOPA 0.5 uCi/ dish was added to duplicate dishes of paired mela­nocyte cultures for 19 h. At 24 h, saline-rinsed cul tures were treated with 5% T CA; the precipitates were trapped on glass fil ters, eth­anol -rinsed. and processed for scintillation coun ting.

Reagents Unless otherwise stated. chemicals and reagents were obtained from Sigma C hemical Co, (St. Louis, MO). DAG was obtained [rom Calbiochem (La JoU •• CAl. 14-C DOPA from Amersham (Arlington Heights, IL), and H-7 from Seikagaku Kogyo (Tokyo, Japan).

0022-202X/ 89/S03.50 Copyright © 1989 by The Sociery for Investigative Dermatology. tnc.

700

VOL. 93, NO. 5 NOVEMBER 1989

RESULTS

Stimulation of Pigmentation by OAG After 7 d, OAG at opti­mal concentration (100 11m) increased melanin content per cell to 406 ± 84% (mean ± SEM) that of control cultu res, while cel l num­ber was 107 ± 7% that of contTOls in seven experiments using cells from different donors. At concentrations up to 100 JiM OAG, mela­nin COntent increased linearly wi th the log of the dose (Fig 1). The increased melanin com ent occurred rapidly,. ~i th approximately one-half the increase observed 12 h after addition of OAG to the cultures and no further increase after 24 h (Fig 2).

Absorption spectra were obtained co confirm that the A.4?S read­ings we.re truly reflective of melanin content and not artifacrually increased at this wavelength by the test compound. The spectra for cell ex tracts under control conditions and after treatment with OAG or with IBMX as a positive comrol {5.12] were identical to each other and very similar to tMt described for synthe tic melanin in solutiol1 113.1. with heigh t of the curves increasing as expected w hen cultures were treated with OAG or LBMX (Fig 3). Base hy­drolysis of the cell extracts at 95 °C usually resulted in a modest decrease in absorption and identical shifts in the absorption spectra for all treltments. Samples from OAG treated cultures wtre no more labile than either positive or negative controls. Addition of OAG to human dermal fibroblast cultures. as a further control, had no effect on the low baseline A.7s' These data strongly suggest that. as expected, melanin accounts for essential ly all the absorpt ion mea· sured at 475 nm in the melanocyte extracts.

T o examine the rate of melanin syn thesis, cul tures were supple­mented with OAG 100,lIM or vehicle alone, then further supple­mented with HC_DOPA. With in 24 h. OAG t reatmen.r signifi­cantly increased I"C-DOPA incorporation in mel:mocytes from three different donors (225%. 1620/0. and 126% of "'ehicle- treated controls, p < 0.005 by analysis of va riance), suggesting that the observed increase in melanin COntent was due to more rapid syn· thesis.

Evidence for Involvement of Protein Kinase C The hypothe­sized PKC dependence of [he melanin response was investigated by twO experi mental approaches. Fi rst. paired melanocyte cultures we re created with OAG 50 ,uM ilnd varying concentrations of known PKC inhibi tors . H-7 [14] and sphingosine [15]. Although both inhibitors were toxic at sufficiently high concentration. over short incubation periods modest concentrations to tally inhibited the DAG-induced increase in melanization (Fig 4), yet produced no

i on '0 ~

)(

:;; .0 E :;)

z

~

8

4

0 25 50 100

OAG (~M)

18 ± 16 Qi

~ 14 Ol

~

12 Qi 0 ~

10 '" C>. c:

8 'E .. Qi ~

200

Figure 1. Effect ohile d i~CYlg1cerol analogue. 1 -oleoyl.2.:l~cryl-glycerol, on melanin content of culture human me1anocytes. Mdamn contc:nt ~r cell was signi.ficantl y incrcased at OAG conce nrr.:aions of 50 (p < 0.03, t test). 100 and 200 pM (p < 0.001). with peak effecti veness at 100 J~. Points :ue n\t':3.U ± S.E.M. for triplicate dishes itt a represenativc c:xpen · ment,

DIAC YLGLYCEROL STIMULATES MELANOGENESIS 701

120.0

100.0

~

" 80.0 a. eo Control

}: 60.0 OAG (100 uM)

" .. 40.0 ~ ::I!

20.0

0.0 0 20 40 60

Time (hrs)

Figure 2. Time course of the melanization response following exposure to OAG. DOlt:! arc mean ± S.E.M. for triplic.ue detenninatioru using a repre­scnt<ltive donor line. OAG 100 uM increased mehmin content (p < 0.001. l te~)[) up to 450% of diluenHre:ued controls during a 60·h incubuion. There was no statistically signifia.nt chmge in melanin per cdl in control cultures during this time. In a separ2.te ex.periment perfonned to enmine th~ early OAG rcsponse, minimal increase was observed prior to 6 h after OAG addition: 2 h. 114%; 4 h, 132%; 6 h. 220%; 8 h, 242%; 10 h. 268%; :md 24 h, 409% of control.

change in cell number or gross morphology. Second. an isomer of DAG. I .3-dioctanylglycerol. incapable of activa ting PKC [16], was compared to the active 1,2-isomer OAG in its abil ity to induce melanization. In paired cultures from three different donors, O AG­treared cul tures responded as expected, while the melanin content per cdl of 1 ,3-d ioctanylglycerol.treated cultures was 90% - 1 07% that of unrreared conrrols. These combin ed results strongly suggest that GAG stimulates melanogenesis through activation of PKC rarher than through a PKC-independent mechanism.

Evidence for New Protein Synthesis T o detennine whether new protein synthesis is required for DAG-induced melanogenesis, paired melanocyte culrures were supplemented with OAG 50,uM and either cycloheximide 20,ug/ ml or actinomycin D 15 .ug/ m!.

". ".

375

' .. ......

..... .•......•..

425 475 525 575

Wavelength (nm)

IBMX

OAG

Co/llrol

.25 .75 725

Figure 3. Absorption speCtr.l of extracts of mdanocyte culrure~ maintained 7 d in medium containing OAG l OO.uM, lBMX IOO,uM. or dIl uent alone. CunlfS are highly similar co each other and generally confo~ [0 publis~ed specm For synthetic melanin. The 475·nm wavelength a[ u:h.,ch absorp?on was routinely monitored is nored. The data shown are ~clllgs of basel me­corrected sPCCtr2 after photographic reduction of the ongmal charts and are rC'opres.enutive of data obtained [rom four different donms. twO Mack and two white.

702 GORDON AND GILCHREST

60

40

Q; u Co 40 30 ~

" " 20 .. Q;

20 ~

10

0 0

0 12.5 25 50 0 2 5

H7 (~M) Sphingosine (~M )

Figur~ ", Inhibitlon of OAG-induced melanogenesis by inhibitors of pro­rein kinaS(' C. OAG 50,uM or diluent alone was added [Q paired mehnocyre cultures simultaneously supplemented with H·7 0-50 J.LM (lift pant!) or sphingosine 0-5 jiM (righl paPltl) . After 12 h. all dishes were harvested for cell counts and melanin content. Bar heights arc mcan ± S.E.M. for tripli­cate cultures in :I. representative experiment. The signific:mt stimulation (p < 0.02) of melanin content induced by OAG alone is reduced by either H-7 12.5 J.LM or sphinogosine 2.uM and abolished by higher doses ofhorh agent.~ . Neither H-7 nor sphingosine had any effect on melanin content in the absence of OAG.

After both 12 and 24 h, melanin content in cultures treated simulta­neously with OAG and a protein inhibitor ranged from 154% to 90% that of untreated controls. equivalent to a 57%-100% reduc­tion in the OAG response . Cycloheximide or actinomycin 0 alone had no effecr on melanin content per cell, although each reduced cell counts approximately 50% at 24 h. These data support a re­qu.ircment for new protein synrhesis prior to the observed OAG-in­duced increase in melanin content.

DiSCUSSION

Melanin pigmentation is the major determinant of skin color and an important protection against pho[Qcarcinogenesis and photoaging. Unfortunately, srudies of human melanogenesis have been severely limi ted by lack of adequate culture systems for normal melanocytes and the ameianotic character of most established human melanoma lines. To our knowledge, the present data provide the nrst infotma­tion regarding the intracellular signal that leads ftom melanogenic stimulation to pigment synthesis in human melanocytes. Our stud­ies do not resolve whether physiologic stimuJi for human pigmen­tarion, of which ultraviolet irradiation is the best studied, acr through diacylgl ycerol and protein kinase C, nor do they exclude a

TH.E JOURNAL OF INVESTIGATIVE DERMATOLOGY

role for cAMP, impl icated in m ur ine melanom a melanogenesis [4} . We can conclude, however, that diacy lglycerol increases pigm ent produCtion in cultured human melanocytes, in al l probability through activation of the protein kinase C pathway.

REFERENCES

1. Johnson GS. Past.m I: N'. 0 2, -dibutyryl adenosine 3~. 5~ -monophos­phue induced pigment production in melanoma cells in culture. Nature (New Bioi) 237:267-268,1972

2. Kreider JW. Rosenth~ M, Lengle N : Brief communication: Cyclic adenosine 3'. 5'-monophosphate in the conrrol of melanoma cell repliation and differentiation. J Narl Cane lnst 50:555 -558. 1973

3. Korner A. Pawelek JM: Acrivation of melanoma tyrosinase by a cyclic AMP-dependent ptoccin kiruse in a cell-free syscem. Nature 267:444-447,1977

4. Halaban R. Pomerantt SH, Marshall S, Lamberr DT. Lerner AB: Reg­ubtion of ryrosimlse in human melanocytes grown in culture. J Cdl Bioi 97:480-488, 1983

5. Friedmann PS. Gilchrest BA: Ultraviolet radiation directly induces pigment production by culrured hum:rn melanocyte's. J Cell Physiol 133:88-94.1987

6. Pawelek JM: FaCtors regulating growth and pigmentation of mela­noma ce ll s. J Invest Denn:nol 66:201 -209. 1976

7. M:u-w:m MM. Malek ZAA, Kreutzfeld KL. Hadley ME, Wilk.es BC. Hruby VJ , Castrucc; AM: Stimulation of 59 1 melanoma tyrosinase activity by superpotent -mdanotropim. Mol and Cell Endocrinol 41,1 71- 177. 1985

8. Malek ZA.A. . Krcurzfdd KL. Hadley ME, Bxegman MD, Hruby VJ. Meyskens FL. Jr.: Long-te.rm and residual melanorropin-stimulated t)'rosinase activity in 591 melanorru cells is density depende.nl. In Vitro Cell Devel Bioi 22:75-81, 1986

9. Gordon PR, Mansur C P, Gilchrest BA: Regulation of human melano­cyte growth. dendnciey, and melanization b)' ker:ltinoc)'te derived factors. J Invest Dermarol 92:565 -572. 1989

10. Nishizllka Y: Studies and perspectives of protein kiruse C. Science 233:305 -3 12.1986

11 . R.!.ndo RR: Regulation of protein kinase C activi.ty by lipids. FASEB J 2:2348-2355. 1988

12. Gilchrest BA, Vrabel MA. Flynn E. Szabo G: Selective cultiv:nion or human melanocytes from newborn and adult epidennis. J Invest DermaroI83:379-376, 1984

13. Anderson RR. Parrish JA: Optical properties ofhunun sk.in.ln: Regan JD and Parrish JA (eds.). Tht' Science of Photornedicine. Plenum Press. New York.. 1982. pp 147-194

14. Hidaka H, In2g2ki M. K..awamoto S. Sasaki Y: Isoquinolinesulfoll2-mides. novel and potenc inhibirors of cyclic nucleotide de~ndcnt protein kinase and protein kinase C. .Biochem 23:5036- 5041, 1984

15. HannL1I1 VA, Loomis CR, Merrill AH. Jr., Bell RM: Sphingosine inhibition of protein IUnase C activity and of phorbol diburyrate binding in vitro 2nd in human platelets. J Bioi Chern 261: 12604-12609.1986

16. Boni LT. Rando RR: The nature of protein kinase C activation by physicaJJ)' defined phospholipid vesicles and diacylglycerols. J Bioi Ch,m 260:10819-10825. 1985