0022-202X/84/ 830 1-00 12$02.00/0 TilE .JO URNA I. OF I NVESTIGATIVE D ERMNI'01.0(: v, 8 :J: 12- 14, 1984

Copyright (e) 19114 by The Williams & Wilkins Co. Vol. 8~. No. 1

Pril1teci ill U. S.A .

Characterization of Melanogenesis in Mouse and Guinea Pig Hair by Chemical Analysis of Melanins and of Free and Bound Dopa and

5-S-Cysteiny ldopa

SHOSU KE ITO , PH .D., KEIS UKE FUJITA, M.D., HISAHIDE TAKAHASHI, PH.D., AND KOWICHI JIMBOW, M.D.

Institute for Com.prelwnsiue Medical Science, School of Medicine, Fujita-Galwen Un iversity (S f, I<F. H 'I'), '/'oyaalie, Aichi, and Department. of Dermatology, S apporo Medical College (I<J), S apporo, Japan

This study examined how various genotypes of coat color in mice and guinea pigs are r elated to the type and content of melanin a nd to the levels of free and protein­bound dopa and 5 -S -cysteinyldopa in hair. In analysis of black, yellow, and white areas of tortoiseshell guinea pigs, the me lanogenesis type was in parallel to the type and content of melanin and was correlated fairly well with the levels of melanin precursors. In mouse hair, substitution of the brown allele (bb) for black (BB) reduced the eumelanin content to 1/ 2 to 1/:), while it signif­icantly increased the dopa level. The dilution (dd) gene of mice r educed the eumelanin content only slightly, while the gene for pink-eyed dilution (pp) reduced the content of eumelanin and the level of dopa to as much as 1/ 10 . From the eumelaninJpheomelanin ratio, the melanin of brown and dilute brown mice was found to be eume­lanic, while the melanin of pink-eyed dilution mice ap­peared to be a mixed type because of an extremely low content of eumelanin. The levels of bound dopa and 5-S-cysteinyldopa in hair were found to largely reflect the tyrosinase activity.

Two basic classes o f m ela nin pigment are synthesized in mammals, i. e., brown-black eumela nin a nd ye ll ow to reddis h ­brown pheomela nin [1]. In m ela nocytes, tyrosinase co nve rts ty rosine to dopa a nd t hen to dopaquinone whic h is cyclized and oxidized to give rise to eume lanin. If dopaquinone encoun te rs cyste ine or o ther re lated sulfhyd ry l compounds, e .g., glutathi­one, pheomelan in is produced via cyste iny ldopas, a mong which 5-S-cyste iny ldopa is t he m ajor iso mer [1]. It is suggested t hat copoly me ri zation o f dopa and cysteinyldopas may lead to t he production of mixed -type melanin [2- 4].

M ela noge nes is in mamma ls is dependent on the action a nd interaction o f many ge nes; thi s has been best studied in coat co lors of m ice [5,6]. The re have been m a n y ultrastructural studies on fo llicular melanocytes in attempts to co rrelate t he ge notype of coat colo r wi t h the morpho logy of mela nosom es [5,7- 9]. The effect of coat-co lor ge nes on tyrosinase activ ity has a lso been studied [101 . H oweve r, litt le attempt has been made to co rrelate t he ge notype wi t h the conte nts of eume lanin a nd pheomela nin, probabl y because of t he lack of adequate methods for qua n t itat ion .

The presen t study a nalyzes the conte nts of eume la nin and

Manuscript received Octoher 13, 1983; accepted for publication Feh ruary 1,1 984 .

This work was supported by Gra nts- in -Aid No. 58010030, No. 58:190016, and No. 5840265 from the Minist ry of Education, Science and Culture; by the ,Japanese O'Leary Foundation; and by the Alfred­Ma rchionini Foundation.

Reprint requests to: Shosuke Ito, Ph.D., Institute for Comprehensive Medical Science, School of Medicine, Fujita-Ga kuen Unive rsity, 1'0-yoake, Aichi, 470-11 .Japan.

Abbreviations: AHP: am ino- hyclroxyphenylalan ine HPLC: high-performance liquid chromatography PTCA: pyrrole-2,3,5 -t ricarboxyli c ac id

pheomela nin and of t heir precursors, dopa a nd 5-S-cysteiny l­dopa , in the blac k, yell ow, and white ha ir of torto ises hell guinea pigs a nd t he n in t he ha ir of mice carry ing dive rse coat-color ge nes. We prev iously s howed that mela noge nic tissues Con ­ta ined hi gh levels of dopa a nd 5-S-cyste inyldopa in t he acid hydro lysate [ll] . The levels of t hese prote in -bound catechols are a lso examined in this study to see whethe r t here is any co rrelation between t he levels of these catechols and the type of melanogenesis.

MATERIALS AND METHODS Animals

Two different groups of animals were 5t udied: (a) to rto iseshell f,'1l i nea pigs (1/2 strain) showing a mix ture of black, ye llow, and white areas; (b) mice showing diverse coat colors (see T able I). These inbred strains of animals have been maintained in the Laboratory Animal Center, Fuj ita-Gakuen Uni ve rsit.y. Hair follicles were cut wi th scissors from adult anima ls.

12

Analysis of Eu.m.elanin and Ph eOll1elanin

Our previously reported met.hodl] 21 was closely foll owed with minor modifications. Eumelanin content in hair was estimated hy mul t iplyin g the amount ofpyrrole-2,3,5 -tricarboxylic ac id (PTCA), a major pyrrolic product of eumelanin , after permanga nate ox idation with a factor of 50. A hair sa mple, 20- 30 mg, was homogeni zed in 5 ml of 1 M K 2COa and to th is, 3% KMnO, was added until the purple color of KMnO, persisted . PTCA was then extracted with ether and analyzed by high­performance liquid chromatography (H PLC) with ul traviolet detection 1111 . Pheomelanin content was est imated by multiplying the amount of amino-hydroxyphenylalanine (A HP), a spec ific indicator of pheo­melanin , afte r hydriodic acid hydrolys is with a factor of 5. A sample, 20- 30 mg, was suspended in 5 ml of 57% hydri odic ac id containing 100 mg of H"PO" and heated under reflu x for 20 h. Aft.er evaporation, the product AHP was ana lyzed by I-IPLC with electrochemical detect. ion 112J.

Analysis of Free and Protein-Bou.nd Dopa and 5-S-Cysteiny ldopa

Free and bound catechols were dete rmined by our previously de­scribed method lllJ with minor modifications. A hair sample, 20- 30 mg, was homogenized in 5 ml of 0.4 M HCIO .• and cent.rifuged. The supernatant was subjected to the assay of free catechols by HPLC. The precipi tate was then hydrolyzed in 5 ml of 6 M HCI contai nin g 0.5% cysteamine hydroc hloride by reflu xing for 24 h. The hydrolysate was evapo rated, t.aken up in 0.1 M HCI, and subjected to t he assay for protein -bound catechols. Alumina extraction was required for analyzing the sample of low cat.echol co ntent. T he HPLC conditions were t he same as t hose prev iously described IllJ except that t.he detector poten­tial was mainta ined at +750 m V vs an Ag/ AgCI reference electrode.

RESULTS

Guinea Pig Hair

Table II s hows the levels of eumelanin and pheomelanin and of free and protein-bound dopa a nd 5-S-cysteiny ldopa in gui nea pig hair of differe nt colo r areas. Black ha ir conta ined 0.85% of eumela nin , while ye llow hair possessed 0.77% of pheomela nin. White hai r co ntai ned no appreciable a mounts of either m e la­nin. Thus, t he type a nd content of me lanins correlated well with t he coat color.

July 1984 ME LANO G EN ES IS AND MELANIN 13

TA BL E I. Contents of melanins and of free and bound catechols in mou,~e hair'

Hair color Percent content, Free Bound Strain wt/wet wl catechols (I'g/ g) catechols (I'g/ g)

Genolype Phenotype Eumelanin Pheomelanin Dopa 5·S·CDb Dopa 5·S·CD

C57BLj6 aaB BCCDD B lac k 5.2 ± 0.9 0.028 ± 0.001 14 ± 2 0.1 ± 0. 1 170 ± 19 150 ± 27 HTH aabbCCDD Brown 1.8 ± 0. 1 0.048 ± 0.019 31 ± 2 0.0 ± 0.0 370 ± 23 270 ± 22 DBAj2 aabbCCdd Dilute brown 1.1 ± 0.0 0.030 ± 0.006 23 ± 4 0.8 ± 0.3 230 ± 37 130 ± 22 PON aaBBCCpp Pink -eyed dilut ion 0.12 ± 0.06 0.046 ± 0.004 2.2 ± 0.8 0.3 ± 0.3 64 ± 9 93 ± 34 CBA/ JM s AABBCCDD Agouti 3.0 ±I.l 0.36 ± 0.11 22 ± 5 1.2 ± 0.2 170 ± 19 100 ± 14 NC AAbbCCD D Cinnamon 1.7 ± 0. 1 0.35 ± 0.09 31 ± 3 1.4 ± 0.2 290 ± 43 270 ± 83 KK aaBBccDD Albino 0.03 ± 0.02 0.005 ± 0.002 0.2 ± 0.2 0.0 ± 0.0 48 ± 15 10 ± 4 AKR/JMs aaBBccDD Albino 0.03 ± 0.00 0.005 ± o.ooa 0.1 ± 0.1 0.0 ± 0.0 36 ± 17 12 ± 1

" Mean ± SO fo r 3 anima ls except. for AI{ R (n = 2). b 5.S.Cysteinyldopa.

TABLE II. Contents of melanins and of free and bound catechols in hair of to rtoiseshell guinea pigs"

Hair Percent content, wt/ wel wt color Eumelanin

Blac k 0.85 ± 0.27 Yellow 0.06 ± 0.01 Whi te 0.02 ± 0.01

" Mean ± S O for 3 a nima ls. b 5-S.Cysteinyldopa.

Pheomclanin

0.08 ± 0.04 0.77 ± 0.09 0.005 ± 0.003

The level of free melanin precurso rs also refl ected fa irly we ll t he type of melanin formed; in the blac k ha ir t he dopa content was higher t han the 5-S-cysteinyldopa content, in the yellow ha ir t he latte r was higher t han the former, and in the white hair both catechols were ha rdly detectable.

The levels of prote in -bound dopa and 5-S-cysteinyldopa were in the order of blac k > yellow > whi te, t hus suggesting that t he levels of t hese catechols may be correlated wi th the tyrosinase activity.

M ouse Hai r

The A-locus in mice is responsible for t he regional distribu­tion of eumelanin and pheomelanin wit hin individua l ha ir. The wild type (AA) mice produce yellow-banded ha ir, while the ha ir of recessive (aa) mice is uniformly co lored wi t h eumelanin . The B -locus cont rols the color of t he eumelanin produced; changing black to brown in bb mice. This locus influences the shape and in ternal structure of melanosomes. The C-Iocus affects tyrosin­ase activity, causing no pigment to be produced in the albino (ee) mice. In recessive dilu t ion (dd) mice, melanosomes a re prod~ced in normal amoun ts, bu t are deposited in clumps, t hus causll1g the coat color to appea r diluted. The gene fo r pink­eyed dilut ion (pp) has a marked dilution effect on eumelanin pigmentation.

T able I shows the leve ls of eumela nin and pheomelanin and of free and prote in -bound catechols in ha ir of mice carrying different coat color genes. Blac k (aaBB ) mice produced 3 t imes as much eumelanin as did brown (aabb ) mice. Similarly, t he eumelanin level in agouti (AA BB) hair was approximately twice as high as that in cinnamon (AAbb) . Thus, t he brown a llele (bb) appeared to reduce t he eumelanin content to one-half to one-t hird of that from the blac k counterparts (BB) . The dilu ­tion allele (dd) reduced the eumelanin content to two-thirds, while the gene for pink-eyed dilut ion ( pp) reduced the eume­lanin conte nt to fa r less than one-tent h of those of black and brown mice. From the eumelanin/ pheomelanin ratio, t he ha ir of brown and dilu te brown mice was found to be eumelanic, while the melanin of pink-eyed dilut ion appeared to be a mixed type because of the very low conte nt of eumelanin .

Mice wi th nonagout i (aa ) genotype contained only t race amounts of pheomelanin . On the other hand, t he AA genotypes (agout i and cinnamon) permitted production of considerable

Free cat.cchols (I'g/ g) Bound catcchols

Dopa 5·S·CDb Dopa 5·S·CD

2.1 ± o.a 1.1 ± 0.5 48 ± 4 24 ± 7 2.6 ± 0.7 6.4 ± 1.2 36 ± 7 18 ± 4 0.1 ± 0.1 0.1 ± 0.1 20 ± 5 7 ± 1

amoun ts of pheomelanin , t he levels being a lmost identical in both mice. Alb ino mice (KK and AKR strai ns) produced no appreciable levels of eit her type of melanin .

Cont rary to the expectation from visua l appearance, t he levels of free dopa, a eumelanin precursor, were significant ly higher in brown and cinnamon (bb) mice than in blac k and agout i (BB) mice, respectively. T he levels of free 5-S-cyste i­nyldopa, a pheomelanin precursor, in agout i and cinnamon (AA ) mice were much higher than those in their nonagout i counte rparts, blac k and brown, refl ecting t he temporary pro­duction of pheomelanin in t he wild type mice. As expected, t he whi te hair of albino mice contained hardly detectable amounts of free melanin precurso rs. The level of free dopa in pink -eyed dilut ion was much lower t han t hat of the colored ha ir t hough it was higher than t hat of the albino hair.

The levels of protein-bound dopa and 5-S-cysteinyldopa par­alleled well t hose of free dopa; brown and cinnamon ha ir contained mllch greater amounts of bound catechols t han did black and agouti, respectively. Albino mice showed t he lowest level of these catechols, while pink-eyed dilut ion had an inter­mediate level.

DISCUSSIO N

We have recent ly developed a simple method for analyzing quant itatively the eumelanin and pheomelanin conte nts in t issue samples [1 2]. The method is based on t he product ion of PTCA by permanganate oxidation of eumelanin and of AHP by hydriodic acid hydrolysis of pheomelanin . Using t his method, we found that C57BL/6 blac k (aa) hair conta ined 6.5% of eumelanin , while C57BL/6 ye llow (AVa) hair possessed 1.3% of pheomelanin [12]. The present resul ts wit h guinea pig hair furt her support t he validi ty of our chemical analysis of melanin .

Many attempts have been made to correlate t.he morphology of melanosomes with the melanogenesis type in mice [5]. Fol­licula r melanocytes in black (00) mice produce ellipsoida l­lamellar melanosomes, while those in yellow (A"a) mice pro­duces spherical-granula r melanosomes [7,8]. T he p igments syn ­thesized in these melanosomes are considered to be typical eumelanin and pheomelanin , respectively, as also supported by our chemical ana lysis [1 2]. Follicular melanocytes in brown (bb ) mice also produce spherical melanosomes [1 3], simila r to those seen in yellow mice. Thus, t he ul t rast.ructural observation

14 ITO ET AL

a lone may not be enough for t he di ffe re nt iation of t he brow n type melanogenesis from t he ye ll ow melanogenesis. Inasm uch as the brown a llele (bb) was shown to affect eumela nogenesis but not pheomelanoge nesis [5], t he brown mela nin was consid ­ered to be eumela nin without a ny chem ical proof. The present study clea rly s hows t hat brown mice syn t hesize primarily eu­melanic pigment a nd t he difference from black mice is t he small er amou nt of eumelanin produced. However, the possibil­ity cannot be ruled out that t he brown eumelanin might be structurally different from t he b lack eumela nin , t hus giving a different yield of PTCA, on which t he calculation of eumelan in content is based .

Dopa is a p recursor of both eumela nin a nd pheomelanin, wh il e cysteiny ldopa is a precursor of pheomelan in but not of eumelan in . T he level o f t hese melanin precursors in t he hair of to r-Lo ises he ll gu inea pig reflected fa irly we ll t he type of me la n ­oge nes is. S imilar results were first reported by Rorsman ei a l [14]. T hey a lso found that the levels of free dopa and 5-S­cyste iny ldopa in serum as well as s kin of black, red, and a lbino gu inea pigs were re lated to t he type of melanoge nesis [15]. However, the dopa/5-S-cysteinyldopa ratio appears to be not as good as the eumela nin /p heo me la nin ratio as the indicato r of t he melanoge nes is type. In this connection , it s hould be pointed out t hat t.he dopa/5-S-cysteiny ldopa ratio does not refl ect the melanogenesis type in mouse melanoma [11 ,12], though some correlat ion may be found in huma n melanomas [2].

The extremely low level of 5-S-cysteiny ldopa in nonagouti (aa) mice seems to be a re fl ection of eumela nogenesis. The higher level of dopa in mice with t he brown (bb) a llele t ha n in the hlack (BB) counterpa rts is rathe r surpri sing, but this may be related to the higher tyrosinase activity in brown mice t ha n in black mice [10,16]. The low level of free dopa in pink-eyed di lu t ion may also result from t he low tyrosinase activity (10).

The levels of protein-bound dopa a nd 5-S-cyste inyldopa ap ­pea r to parallel the tyrosinase act ivity: it is higher in black t han in yellow ha ir of tortoiseshell guinea pig as well as of mice [1 1) a nd higher in brown than in black ha ir o f mice. Therefore , it is likely that most of protein -bound dopa and 5-S-cysteinyl­dopa are syn t hesized by the action of tyros inase. However, inasmuch as nonmela nogen ic tissues such as white ha ir, live r, kidney, a nd brain co nta in small amounts of the bound catechols [l7], ox idation mecha ni sm(s) other than ty rosinase may a lso participate in t he synthes is of these catechols in t he prote in ­bound form. Recently , not. only free dopa but a lso free 5-S­cysteinyldopa was detected in nonmela notic tissues, such as splee n , heart [18 ], and gangli on stellatum [19] . It might be poss ible t hat t he free catecholic am ino acids originate from the protein -bound form.

Vol. 83, No.1

We wish to express our t.hanks to Y. Yamamoto for her help in preparation of t.he manuscript.

REFERENCES

1. Prot.a G: Recent advances in the chemistry of melanogenesis 111

mammals. J Invest Dermatol 75:122-127, 1980 2. Rorsman H, Agrup G, Hansson C, Rosengren A-M, Rosengren E:

Detection of pheomelanins, Pigment Cell , vol 4. Edited by SN Klaus. Basel, Karger, 1979, pp 244-252

3. Ito S, Novellino E, Chioccara F, Misuraca G, Prota G: Copolymer­ization of dopa and cysteinyldopa in melanogenesis in vitro. Experientia 36:822-823, 1980

4. Novellino E, Ortonne JP, Voulet C, Chioccara F, Misuraca G, Prot.a G: Identificat.ion of cysteinyldopa-derived uni ts in eume­lan in from mammalian eyes. FEBS Lett. 125:101- 103, 1981

5. Si lvers WK: The Coat Color of Mice. A Model for Mammali an Gene Act.ion and Interaction. New York, Springe r· Verlag, 1979

6. Prot.a G, Searle AG : Biological sites of gene action for melanoge­nesis in mammals. Annales des GenaLique el. de Selection Ani­male 10: 1-8, 1978

7. Sakurai T , Ochiai H, Takeuchi K: Ultrastruct.u ral change of me l­anosomes associated with agout i pattern formation in mouse hair. Dev Bioi 47:466- 471,1975

8. Jimbow K, Oikawa 0, Sugiyama S, Takeuchi T : Comparison of eumelanogenesis and pheomelanoge nesis in retinal and fo llicular melanocytes; role of vesiculo-globular bodies in melanosome differentiation . ,] Invest. Dermatol 73:278- 284, 1979

9. Ito M, Hashimoto K, Organisciak DT: Ultrastructural, histochem­ica l, and biochemical studies of the melan in metabolism in eye and sk in of pallid mice. J Invesl Dermatol 78:414- 424, 1982

10. Coleman DL: Effect of gene substitution on the incorporation of tyrosine into the melanin of mouse hair. Arch Biochem Biophys 96:562- 568, 1962

11. Ito S, Homma K, Kiyota M, Fujita K, Jimbow K: Characterization of st ructural properties for morphologic differentiation of mela­nosomes. III. Free and protein-bound dopa and 5-S-cysteinyl ­dopa in B16 and Hardin g-Passey melanomas. J Invest Dermatol 80:207- 209, 1983

12. Ito S, Jimbow K: Quantit.ative analysis of eumelanin and pheo­melanin in hair and melanomas. J Invest Dermatol 80:268- 272, 1983

13. Moye r FH: Ge netic variations in the fine st ructure and ont.ogeny of mouse melanin granules. American Zoologist 6:43- 66, 1966

14. Rorsman H, Rosengren A-M, Rosengren E: Det.ermination of 5-S­cysteinyldopa in melanomas with a Iluorimetric method. Yale J Bioi Med 46:516- 522, 1973

15. Hansson C, Agrup G, Rorsman H, Rosengren E: Dopa and 5-S­cysteinyldopa in the se rum of albino, black, and red guinea pigs. Acta Derm Vene reol (Stockh) 60:155- 156, 1980

16. Fitzpatrick TB, Szabo G: The melanocyt.e: cytology and cyt.ochem­istry. J Invest Derm.ltol 32:197- 209, 1959

17. Ito S, Jimbow K, Kato T, Kiyota M, Fujita J<: Protein -bound dopa and 5-S-cyste inyldopa in non-melanogenic t issues. Acta Derm Venereol (Stockh) 63:463- 467, 1983

18. Agrup G, Hansson C, Rorsman H, Rosengren E: 5-S-Cyst.einyl­dopa, dopa, and dopamine in the kidney and some other t issues. Acta Derm Venereol (St.ock h) 60:63-64, 1980

19. Fehling C, Hansson C, Pou lsen J , Rorsman H, Rosengren E: 5-S­Cysteinyldopa ill ga nglion stell atu m. J Neural Transm 52:251-257, 1981