0022-202X/ 80/7501-0 122$02.00/ 0 TuE JOURNAL OF iNVESTIGATIV E DERMATOLOGY, 75:122-127, 1980 Copyrighl © 1980 by The Williams & Wilkins Co.

Vol. 75, No. 1 Printed in U.S.A.

Recent Advances in the Chemistry of Melanogenesis in Mammals

GIUSEPPE PROTA, PH-D.

l stituto di Chimica Organica e Biologica dell'Universita, Napoli, Italy

The color of mammalian hair, skin, and eyes results mainly from the secretory products of melanocytes. These secretory products consist of a wide range of melanin pigments with different structures and compo­sitions. These include black or brown nitrogenous eu­melanins; yellow or reddish brown, sulfur-containing pheomelanins, e.g., the trichochromes of low molecular weight; and other pigments whose chemical and physical properties are intermediate between those of typical eumelanins and pheomelanins. Despite the evident dif­ferences in molecular size and general properties, all these pigments are biogenetically related, and they arise from a common metabolic pathway in which dopaqui­none is the key intermediate. The current state of knowl­edge on the molecular mechanisms governing the meta­bolic fate of dopaquinone in melanocytes is discussed with special reference to the role of such sulfhydryl compounds as cysteine and glutathione in melanogene­sis.

The last decade can rightly be considered the most fruitful period in the 100-yr-long course of melanin research. Many fundamental problems about the structure and biosynthesis of natural melanins have been clarified, and great advances have been made in the field of urinary melanogens since the identi­fication of several new metabolites, which has yielded consid­erable insight into the biosynthetic activity of normal and pathologic melanocytes. These and other advances, including the recognition of the cytotoxic effect of melanin precursors and an improved knowledge of tyrosinase enzymes with respect to active site and mechanism of inhibition, have radically changed the accepted concept of melanogenesis and thereby have opened new horizons to both basic and applied research.

I do not intend to review all these studies in any detail here, but rather to summarize the chemical and biochemical studies that have led to the formulation of an improved scheme for -mammalian melanogenesis. Although only a few of the many original articles dealing with the subject can be mentioned, attention should be drawn to the publication of papers pre­sented in recent years at the International Pigment Cell Con­ferences.

BIOLOGICAL AND EVOLUTIONARY ASPECTS OF MELANOGENESIS

Collectively, melanins [1-3) are among the most widespread natural pigments. They are found in practically all living orga­nisms, including higher plants, fungi, and bacteria. In line with the general policy of nitrogen economy, plant melanins are usually derived by oxidative polymerization of nitrogen-free phenolic substrates, whereas analogous pigments in animals arise from the metabolism of the phenolic amino acid tyrosine

Part of the work described in this article was supported by a grant from the Consiglio Nazionale della Ricerche (Progetto Finalizzato: Controllo della Crescita Neoplast ica) , contract n 78/02862/96.

Reprint requests to: Giuseppe Prota, Ph.D., Istituto di Chimica Organica e Biologica deli'Universita, Via Mezzocannone 16, 80134 Na­poli, Italy. Abbreviations:

dopa: 3,4-dihydroxyphenylalanine GSH: glutathione

122

via 3,4-dihydroxyphenylalanine (dopa) or related compounds such as dopamine and adrenalin. Despite apparent diversity in the nature of the primary substrates and hence in the chemical structure and properties of the resulting pigments, a unifying biochemical feature of melanogenesis in both plants and ani­mals is found in the enzyme system, commonly referred to as tyrosinase(s) [ 4-6], which catalyzes the key steps of the process, i.e., the conversion of a phenol to the corresponding ortho­quinone (Fig. 1). The ubiquitous occunence of this enzyme in both eukaryotic and prokaryotic organisms is evidence that melanogenesis appeared early in evolution and has developed basically without change throughout the phylogenetic scale to mammals, in which melanin pigmentation, widespread in hair, skin, and eye, has a wide range of important physiological roles, namely, protection of underlying tissues from ultraviolet (UV) , irradiation, heat control, and adaptive coloration. These, how­ever, are clearly adaptive functions of a process that originally might have had a more fundamental biological significance than the provision of a visible pigment. Indeed, many organisms have populations of black- or brown-producing cells in nonil­luminated internal areas, notably the neurons of the central nervous system and the cells of the cromaffin system of humans and other vertebrates, in which the formation of melanins can hardly be explained in terms of pigmentation.

Presumably, the basic biochemical significance of melanoge­nesis has to be found in the initial steps of the process that lead to the formation of ortho-quinones. These are among the most chemically reactive of all organic compounds that, once formed, may give rise to a series of spontaneous reactions (Fig. 2), including oxidative polymerization, redox exchange processes, condensation with -NH2 or SH groups of biologically impor­tant metabolites or essential proteins, as well as lipid peroxi­dation. A detailed account of these reactions has been given by Riley [7,8], who has emphasized how the generation of o-qui­nones as intermediate metabolites in melanogenesis represents a potential hazard to the well-being of the cell in which the process takes place. Experimental evidence tor the cytotoxic effect of tyrosine metabolites has been provided by Lerner and Pawelek [9], who have shown that the melanocyte is increas­ingly vulnerable to injury and death under conditions of accel­erated melanin synthesis. In fact, from a vaTiety of clinical and experimental observations it appeal"s that any phenol com­pound may be toxic to melanocytes, evidently because of the presence in these cells of the enzyme tyrosinase, which acts as a focus for selective absorption of the substrate (as in a process of affinity chromatography) and subsequent metabolism with generation of potentially damaging quinones.

In view of these chemical and biological facts, the process of pigment production may be seen as a detoxification mechanism by which a highly reactive quinone is rapidly converted through a series of spontaneous reactions into an insoluble, relatively inert, nondiffusible polymer. The efficiency of such a protective mechanism depends on various circumstances, e.g., pH and redox potential of the biochemical environment, compartmen­talization of the site of polymerization, and strict regulation of the rate of quinone synthesis. Yet the possibility exists that the process is accompanied by deleterious side reactions, and the need for active scavenger mechanisms for diffusible intermedi­ate quinones of melanogenesis is clear. One of these is brought about by SH compounds, such as glutathione (GSH) and cys­teine, which (as we will see later) play a key role in the regulation and metabolism of pigment cells.

July 1980

rAJOH R~ -r---

rAJOH ~0

R~OH f '\ R~O o2 0 H20

FIG 1. Simplified reaction scheme for the conversion of a phenolic substrate to the corresponding orth.o-quinone catalyzed by copper oxidases and tyrosinases widespread in plants and animals.

R@ OH REDUCTION

OH (

R~. 0 OOHH )n POLYMERIZATION ,·~ R~O

~0

R'NH,; REDUCTIVE \ R'SH I ADDITIONS \

R~OH

R'HN00H

RY()IOH

yaH SR'

FIG 2. Self-driving reactions of orlh.o-quinones.

THE NATURE OF MELANINS

Omitting more than passing reference to the examples of melanins found in the plant kingdom, in invertebrates, and in poikilothermic vertebrates (fishes, amphibians, and reptiles), I shall adhere in this discussion to an orientation toward the melanins found in mammals, which are of much greater biolog­ical interest (10].

Historically, our concept of melanin pigmentation developed almost exclusively in relation to the nature and formation of black melanin pigments, which were at an early time recognized as being derived from the tyrosinase oxidation of tyrosine and related metabolites. But it is clear, considering the complexity of mammalian epidermal pigmentation (as seen, for example, in hair, which comes in several colors), that polychromy is probably not due to pigment of a single chemical nature. In fact, in the early days of melanin research the polychromy of mammalian hair was attributed to 2 basic types of pigment: the eumelanins (accounting for black and brown colors) and the pheomelanins (accounting for the lighter ones, with a wide rang~ froT? yellow to reddish brown) . Such a broad naked-eye classificatiOn was supported by the genetic pattern of hair color in mice, which suggested the existence of 2 separate, but pos­sibly interrelated, pathways for eumelanin and pheomelanin formation. In line with this view, in 1958 Fitzpatrick, Brunet, and Kukita [11] in a review of the early studies on melanin pigmentation proposed that the precmsor ofpheomelanin could be tryptophan or a tryptophan metabolite that could be enzy­matically oxidized to yellow pigment only in the presence of intermediates of the eumelanin pathway. Subsequent attempts to obtain experimental evidence for the involvement of trypto­phan in pheomelanin biosynthesis were unsuccessful and other theories were advanced to account for the different chemical and spectral properties of eumelanin and pheomelanin. A widely accepted view was that pheomelanins were a variety of eumelanins in different states of oxidation or aggregation. Grad­ually, over the years the original distinction between the 2 classes of pigments was obscmed, and consequently experimen­tal data from studies on eumelanins were extended to pheo­melanin and vice versa.

A tur~ing J?Oint in melanin research came about through my work with Nicolaus [12]. We first succeeded in isolating some pheomelanins in pure form by alkaline extraction of red feathers from New Hampshire hens. Analysis of these pigments revealed the presence, besides nitrogen, of a high sulfur content, and oxidation with hyd1·ogen peroxide gave cysteic acid and small

MELANOGENESIS IN MAMMALS 123

am?unt:s of glycine and aspartic acid. The absence of pynole denvatiVes, the characteristic degradation products of eumela­nins, showed that pheomelanins were not structural variants of e~melanins. These data led to the suggestion that pheomelanins might be formed in vivo by some deviation from the normal course of melanogenesis involving cysteine or related com­pounds [13]. A close biogenetic link was apparent from the fact that both eumelanin and pheomelanin were form ed in mela­nocytes, from the operation of a switch mechanism in the pigment-producing cells from agouti mouse follicles and from the consistent ~:1 nitrogen:sulfur ratio in all the f~ather pig­m~nts .. In fact, m an early biomimetic experiment, enzymatic oXJdatwn of dopa in the presence of cysteine gave a reddish brown. al~ali-soluble pigmen.t closely resembling natw-al pheo­melaruns mstead of a black, msoluble eumelanin-type polymer. !his ~nding wa~ importa?t to subsequent structw-al work [14] ~nvolvmg extensive chemical degradations. These provided ev­Idence tha~ pheomelanin~ isola~ed from hen feathers (ga.llo­ph~omelaruns) ha~e a highly Irregular structw-e consisting mainly of benzoth1azole and tetrahydroisoquinoline ring sys­tems, tentatively linked as in the partial structure (formula I) .

OH

N-~ s- OH

H02C

OH

C02H

s

H02C (I)

In addition, there is evidence that 1,4-benzothiazine units at various oxidative levels partake in building up the gallopheo­melanin molecule, and that some uncyclized cysteine residues are present as well.

Chemical evidence (15] indicates that mammalian pheomel­anins from rabbits, goats, sheep, and red-haired people, as well as from certain human melanomas, are closely similar in struc.­tme and differ mainly in molecular size and relative pToportions of the different units in theu· polymeric chains. However, in the cow-se of fw-ther studies on mammalian melanins, my col­leagues and I isolated some unusual types of pigments whose chemical and physical properties (e.g., color, solubility in alkali and nitrogen and sulfu1· contents) were intermediate betwee~ these of pme eumelanin and pheomelanin polymers. Analysis of these pigments by chemical degradation,* coupled with in vitro biosynthesis experiments [16], provided evidence that they arise by a copolyn1erization process involving both eumel­anin and pheomelanin precw-sors, as previously suggested [17]. This intermeshing of the pigmentaTy pathways appears to be a quite general featme of melanogenesis in mammals since

* Prota G, Novellino E, Ortonne JP, Vou lot C, Chioccru·a C, Misuraca G, unpublished data.

124 PROT A

Sulfur content of some natural eumelanins Source of sample

Octopus ink Sole skin Chicken feathers Goose feathers Coot feathers Pigeon feathers Horse hair Ox hair Human hair Human melanoma

Sulfur content (%)

0.2 2.1 1.2 1.8 2.2 3.0 1.3 1.6 3.7 5.8

sulfur is present even in dark, insoluble eumelanin. At one time the sulfur content in these pigments was attributed to the SH binding of protein to the melanin polymer. However, such an interpretation now seems unlikely and cannot explain the high sulfur content of eumelanins such as those of human hair and melanomas (Table). In fact, the general view that eumelanins are chromoproteins has never been founded on solid experi­mental evidence; it has been accepted mainly because the intractable nature of the natural pigments has prevented ex­amination of this structural aspect.

Another major break in the unraveling of the tangled yarn of melanin research came through investigation of the structure and biosynthesis of the trichochromes [3], a unique group of epidermal pigments that occur in certain types of yellow or red hair and feathers. Like pheomelanins, these pigments have, generally speaking, uncommon properties, e.g., insolubility in organic solvents and in water at a neutral pH, difficulty of crystallization, and a strong tendency toward aggregation and adsorption. These factors have contributed to the slow progress in trichochrome chemistry, which dates back to 1879. In that year Sorby [18] reported the extraction of a characteristic purple pigment with pH indicator properties from red human hair by hot mineral acid. The pigment was eventually isolated in a crude form by Rothman and Flesch [19], who suggested the tentative formula (C15H2N20 9)2Fe, and who proposed the name trichosiderin to emphasize the presence of iron. In 1956 Barnicot [20] obtained evidence that trichosiderin was probably an artifact; extraction of red hair with cold, dilute alkali yielded an orange-yellow pigment that was converted into the "red" trichosiderin when heated with acid. Apparently unaware of Barnicot's work, some years later Bold [21] reexamined tri­chosiderin obtained by acid extraction and found that the crude pigment described by Rothman and Flesch could be separated chromatographically into several components containing sulfur in addition to nitrogen. Subsequent attempts [22] to character­ize these pigments did not get very far until extensive chemical studies at Naples [23, 24] provided conclusive evidence that the red trichosiderins as well as their natural yellow precursors, isolated by a mild alkaline extraction procedure, wefe iron-free chromophoric amino acids containing a previously unknown nitrogen-and-sulfur-containing heterocyclic system, designated as !:J.2·2' -bi(1,4-benzothiazine). Typical examples are the red-pur­ple trichochrome F and the yellow trichochrome C (Fig 3); the latter is the predominant and most representative member of the series. By the structural anatomy of these molecules one can easily recognize their biogenetic origin from dopa and cysteine. In fact, knowledge of the biosynthesis of tricho­chromes preceded the structural studies and was acquired by a biomimetic synthesis experiment [12] in which enzymatic or chemical oxidation of dopa in the presence of excess cysteine was found to give, besides pheomelanin-like polymers, a small but significant amount of a pigment identical in all respects to trichochrome F.

Typically, this pigment shows indicator properties: it is red (i\mu.482 nm) in alkaline solution and violet (i\mux587 nm) in acid, and it undergoes a reversible photochromic change, upon exposure to sunlight, that involves a facile conversion from the trans-form to the more bathachromic cis-form.

Ha-

l >-m:u: 482 nm trans-term H• - Xmu 587nm

Ho· . l >..mu 452nm Tnchoc.hrome-C H* .

>.I'N1 no shift

Vol. 75, No. 1

Ho· . f ~ ~mu 500 nm ~-arm H•

Xmu 567nm

Trichochrome - F

Ho·

! }.mu 460nm Decarboxytrichochrome-C

>..ma• 525 nm

FIG 3. Structural formulae and properties of representative tricho­chromes.

Yellow trichochrome Cis not easily recognized directly, but when heated in acid solution it undergoes selective decarbox­ylation of the acid group at position 3 to give a new pigment (corresponding to the Rothman and Flesch trichosiderin) that is yellow-orange in alkaline solution and red-purple in acid. This characteristic behavior provides the chemical explanation of Barnicot's observation, and is quite useful in trichochrome analysis because extraction of hair with boiling mineral acids directly gives the decarboxylated pigments, which can be read­ily purified by paper chromatography and identified by its pH­dependent spectrum.

A variant of this procedure has been developed to ascertain the occurrence of trichochromes in the urine of patients with melanoma metastases [25, 26]. Notably, 5 of the 16 patients examined excreted trichochromes in amounts ranging from 0.01 mg/24 hr to as much as 9 mg/24 hr; the amount seemed unrelated to the degree of metastatic spread and to metastasis in any special organ. The highest urinary level of trichochromes was found in samples from a person with red-blond hair, but another person with red hair had no trichochromes. The rela­tionship between the pigmentation type of the melanoma pa­tient and trichochrome excretion thus remains unclear and represents a major focus for further biochemical and clinical studies.

THE CHEMISTRY OF MELANOGENESIS

Having described the structural features of eumelanins, pheo­melanins, and trichochromes, I shall discuss the biochemical events accounting for the formation of these a chemically different types of pigments that are derived from the same precursor, tyrosine, and that have, at least in part, a common metabolic pathway. The early stages of mammalian melanoge­nesis have been known for a long time; they involve the con­version of tyrosine into dopa, which is then oxidized to dopa­quinone. Both reactions are catalyzed by the same copper­containing enzyme, tyrosinase, which usually consists of 3 mo­lecular subforms (T~,T2 , and T a), all possessing tyrosinase and dopaoxidase activities. In contrast with this view is that of Okun and his associates [27, 28], who have recently reported histochemical and biochemical evidence that mammalian ty­rosinase has only dopaoxidase activity and that the conversion of tyrosine to dopa is catalyzed by peroxidase. However, their interpretation is not entirely unambiguous, and more recent studies by Hearing and Ekel [29] and by Smith and Swan [30] have provided further evidence that a true tyrosinase is present in mammalian melanocytes. Indeed, the only known enzyme responsible for melanogenesis is tyrosinase, which is primarily involved in the conversion of tyrosine to dopaquinone; the

July 1980

subsequent steps of the process occur spontaneously, without any specific enzymatic assistance. This remarkable fact implies that once dopaquinone has been formed, its fate is the result of its own chemical reactivity coupled with the biochemical envi­ronment within the melanocytes.

Available evidence suggests that the behavior of dopaquinone in pigment cells is similar to its behavior in vitro, i.e., that it is converted into eumelanin by a complex series of spontaneous reactions involving cyclization and oxidative polymerization. In the early stages of this transformation (worked out largely by Raper and Mason and outlined in Fig 4), dopaquinone under­goes an intramolecular cyclization in which the amino group of the side chain rapidly adds to the quinone system to give leucodopachrome, which is then oxidized to dopachrome, i.e., Raper's red compound. Under physiologic conditions, dopa­chrome undergoes rearrangement, and to some extent decru·­boxylation, to give 5,6-dihydoxyindole along with 5,6-dihydrox­yindole-2-carboxylic acid.

In the original Raper-Mason scheme, only the former of these intermediates is oxidized to form the highly reactive 5,6-indol­equinone, which by repeated oxidative coupling gives rise to the eumelanin polymer. However, the chemical properties of nat­ural and synthetic pigments indicate that the process is by no means so simple; in addition to 5,6-indolequinone, all the other quinones and catechol intermediates can partake in polymeri­zation. The proportions in which these different compounds ru·e involved in the process. depend primru·ily on their chemical reactivity and to some extent on the chemical and biological conditions under which polymerization takes place. Moreover, one could envisage the possibility that other "foreign" reactive molecules present within the pigment cell either could be incorporated into the eumelanin polymer or could alter the normal course of melanogenesis, as in the case of pheomelanin formation.

According to our experiments, all that is required to switch from the eurnelanin to the pheomelanin pathway is the pres­ence, in the pigment cell, of cysteine residues containing the reactive SH group, which combines with dopaquinone in a very fast reaction to give the cysteinyldopa adducts [31-34] (Fig 5).

Of these new intermediates 3 are isomers arising by addition of the thiol group of cysteine to the 5, 2, and 6 positions of the dopaquinone conjugated system; the 4th has a more complex structure, derived by the subsequent addition of another mol-

OH

~OH

a 5. 6·di hy droxy indole

f :w

H

indole 5,6-quinone

}yo )(fJ

J_/ H0 1C NH 1

dopaquinone

OH

-------aOH ::-,_ I

HC,C I NH

5. 6·di hydrox yi ndol e z-carboxy li c acid

\' ,' \ 0

E~~e l a~:~s +--------~ «!o-W.' / \ HO lC H

' ' , ' ' '

I '\

'

dopachrome

I OH ;o,

SOH

I

H01C

lcucodopachrome

FIG 4. The Raper-Mason pathway from dopaquinone to indole 5,6-quinone. All in termediates are shown as participating in the subsequent oxidative polymerization.

MELANOGENESIS IN MAMMALS 125

OH 0 P'" ..... p· HzN C0 1H H1N C01H

~H1SH CH NH1

Co~H

dopa dopaquinone cysteine

5-S- c ysteinyldopa( 74 %)

l :o·,;:,~p'"

H1N C0 1H

2-S-cysteinyldopa ( 14%) 2.5-S.S-dicysteinyldopa (5%)

FIG 5. Structural formulae and yields of the cysteinyldopa adducts formed in the 1st stage of the cooxidation of 3,4-dihydroxyphenylala­nine (dopa) and cysteine in the presence of tyrosinase. Abbreviation: Enz., enzyme.

POH OH

NH,

SCH1(:HC0 1H

H1N C01H

~ dopaquinone

l O,.Enz . dopa

Po

0 NH1

SCH1(HC01H

H1N C01H

) decarboxylation

FrG 6. Initial stages of the enzymatic oxidation of 5-S-cysteinyldopa leading to the 1,4-benzothiazine intermediates (a catalytic amount of 3,4-dihydroxyphenylalanine (dopa] is required). Abbreviation: Enz., enzyme.

ecule of cysteine to the quinones of 5-S-cysteinyldopa, 2-S­cysteinyldopa, or both. Clearly, the ratio of formation of these intermediates determines the extent to which they take part in construction of the pigments, which do contain units of different types in vru·ious proportions. It is not SW'prising, then, that t he main building stone of ph eo melanins and trichochromes is 5-S­cysteinyldopa, which is by far the major product formed in the reaction of dopaquinone with cysteine. Indeed, enzymatic oxi­dation of 5-S-cysteinyldopa in the presence of a catalytic amount of dopa has been shown to give a red-brown pigment closely resembling natW'al pheomelanins [32]. Moreover, when the reaction is stopped after a short time by acidification, the violet trichochrome F also becomes isolated. Extensive inves­tigation of these model reactions [23, 35, 36) has led me to conclude that pigment formation from 5-S-cysteinyldopa prob­ably proceeds as outlined in Fig. 6.

Typically, enzymatic -oxidation of 5-S-cysteinyldopa in the presence of a catalytic amount of dopa leads to 5-S-cysteinyl­dopaquinone, which rapidly undergoes intramolecular conden­sation to give the cyclic qui.noneimine intermediate; this is reru-ranged by a hydrogen shift to give the more stable phenolic isomer, or it undergoes decarboxylation to form an additional benzothiazine intermediate, unsubstituted at position 3.

By similru· in vitro experiments, we have obtained evioence that under physiologic conditions 2-S-cysteinyldopa behaves similarly and gives analogous benzothiazine intermediates dif­fering in the position of attachment of the alanyl residue on the benzene ring. Little is known about the oxidative behavior of the other cysteinyldopa isomer, which is a very minor inter­mediate of pheomelanogenesis. In fact, in an examination of the structures of the trichochromes so far isolated [3, 23], we could easily recognize the presence of only 2 benzothiazine units at

126 PROT A

various oxidative levels bearing an aJanin side chain at position 7 or 8 that derive from 5-S-cysteinyldopa or 2-S-cysteinyldopa, respectively. Although the details of the later steps in the biosynthesis of trichochromes are uncertain, it seems clear from model experiments that the process involves oxidative dimeri­zation of benzothiazine intermediates. Additional decarboxyla­tion gives the red-purple isomers E and F, and oxidative decar­boxylation produces the yellow pigments B and C. There is also evidence that the same benzothiazine intermediates are in­volved in the formation of high-molecular-weight pheomela­nins, but the pathway of such conversion still eludes character­ization. Recently, we have found that 1,4-benzothiazine (for­mula II) is a highly unstable molecule that, once generated in

f'Y s '/,( :0 ~NvNH

" 5~

Qll) v (II)

an aqueous medium, readily undergoes reversible aJdolization to give a mixture of polymeric products [37], e.g., formula III. This finding suggests that polymerization of the postulated benzothiazine intermediates in the biosynthesis of pheomela­nins may proceed via aldol condensation more than by oxidative coupling involving positions 2 and 8, as had been suggested by Minale et aJ [38].

Yet, in spite of these and other doubts and the difficul_ties inherent in the research, a remarkable fact has emerged: tnch­ochromes an'd pheomelanins are the end products of the same metabolic pathway that diverges after the benzothiazine stage.

All the foregoing conclusions about the key steps in pheo­melanogenesis were derived from model experiments, and evi­dence was lacking for a similarity between the in vitro and the in vivo pathways until the comprehensive studies by Rorsman, Rosengren, and Rosengren [39, 40] on the occwTence of 5-S­cysteinyldopa and related metabolites in melanomas and in the urine of patients with pigmented melanoma metastases.

I will not summarize all the results of these studies, in which in 1975 I had pleasure to participate as a visiting scientist of the Swedish Cancer Society. I will, however, mention the variety of compounds so far identified. These include 5-S-cysteinyldopa and its 3-0-methyl derivative [ 41], 2-S- and 6-S"cystemyldopa ( 41, 42], 2,5-S,S-dicysteinyldopa ( 42], and the trich?ch.rom_es B and C (25]. In further studies [ 43], Rorsman and his associates have shown that 5-S-cysteinyldopa, which is .the major of all these metabolites, is also present in normal urine and in serum, skin and hair of both eumelanic and pheomelanic subjects. Ind~ed, it appears that 5-S-cysteinyldopa and related me~a~JO­lites are formed and excreted by all melanocytes contammg active tyrosinase, including those in which the rate of melanin synthe$iS is negligible, for example, melanocytes of the adult eye [44]. These studies have opened up a new approach ~o studies on the tyrosine metabolism of normal and pathologic melanocytes through the fine analysis of the levels of cystei­nyldopas formed in pigmented cells.

Returning to the chemistry of melanogenesis, I shall now examine the reaction scheme in Fig 7, which provides an overall view of the metabolic pathways leading to eumelanin and pheomelanin formation . Both begin with enzymatic conversion of tyrosine to dopa to dopaquinone; the subsequent steps de­pend upon the biochemical environment within the pigment cell, which is under genetic control [ 45].

In eumelanin-forming melanocytes, dopaquinone follows the classical pathway of melanogenesis, i . e~, it ultimately results in

Vol. 75, No. 1

~y•o•inu.: CS SG (Giuthotione >eductuo)

,,, ..... __.L_~ •. 7,:~ /"''''""''"' Dopaehrome Cyateinyldopa.s

/ ~­? """''"'n y::•••~"''''"' ~''"' Eumelanins Mixed-tyPe Melanins Phaeomelanins

FIG 7. Schematic outline of tyrosine metabolism "in melanocytes. Thich arrows indicate enzymatic transformations; thin arrows indicate reactions that proceed spontaneously or without specific enzymatic assistance. Abbreviations: CyS, cysteine; dopa, 3,4-dihydroxyphenylal­anine; GSH, reduced glutathione; GS SG, oKidized glutathione.

a black, insoluble heteropolymer by oxidative coupling of var­ious indole intermediates. In pheomelanin-forming melano­cytes, a switch mechanism leads to the production of sulfur­containing pigments through intervention of SH compounds such as cysteine. Cysteine combines with dopaquinone in a very fast nonenzymatic reaction to give the cysteinyldopas, which are smoothly converted into the corresponding benzothiazines by oxidative cyclization of their cysteinyl residues.

Alternatively, the cysteinyldopas required for pigment for­mation may arise by reaction of GSH to dopaquinone, followed by enzymatic hyd.rolysis of the resultant glutathionedopa. Sup­porting evidence for such a route has been provided by a recent report [ 46] describing the occurrence of glutathionedopa in human melanoma. In the reported study, evidence was also obtained that there is an effective in vivo mechanism by which any glutathionedopa formed in melanocytes is converted in 5-S-cysteinyldopa by the action of 2 hyd.rolytic enzymes, a-glu­tamyl-transferase and peptidase, present in the pigment cell and in large amounts in kidneys and liver. Still, it is not yet clear to what extent this alternative route to cysteinyldopas is relevant to pigment synthesis as it depends on whether hydrol­ysis of glutathionedopa takes place inside or outside the mela­nocytes. In my opinion [47], in addition to serving as a possible substrate for the synthesis of cysteinyldopa, GSH could play a regulatory role in pigment cell metabolism by converting part of the dopaquinone formed in the melanocytes into glutathi­onedopa, which as such cannot give rise to pheomelanins unless transformed to cysteinyldopa by enzymatic hyd.rolysis. Hence, in the absence of this latter step, the formation of glutathione­dopa would sidetrack the dopaquinone but not produce either pheomelanin or eumelanin. Such a regulating role would ex­plain the occurrence of cysteinyldopas in both eumelanic and pheomelanic people as well as the slow rate of pigment synthesis in active melanocytes. Consistent with this view are some early observations of Halprin and Ohkawara [ 48], who showed that GSH and GSH reductase levels are lower in black skin than in white skin. They also demonstrated that in normal people hyperpigmentation after UV ir.radiation is accompanied by SH/ SS changes that are partly due to a fall in the GSH and GSH reductase levels and a corresponding rise in the level of oxidized GSH.

REFERENCES l. Nicolaus RA: Melanins. Paris, Herman, 1968 2. Swan GA: Structure, chemistry and biosynthesis of the melanins,

Frtschritte der Chemie Organischer Naturstoffe, vol 31. Edited by W Herz, H Griseback, GW Kirby. New York, Springer-Verlag, 1974, pp 522-528 . . ' ' '

3. Prota G, Thomson RH: Melamn ptgmentatwn m mammals. En­deavour 35:32-38, 1976

4. Mason HS: Structure and function of the phenolase complex. Nature 177:179-181, 1956

5. Joily RL, Evans LH, Makino N, Mason HS: Oxytryosinase. J Bioi Chern 249:335-345, 1974

July 1980

6. Lerch K: Amino acid sequence of tyrosinase from Neurospora crassa. Proc Nat! Acad Sci USA 75:3635-3639, 1978

7. Riley PA: The mechanism of melanogenesis. Symposium Zoological Society (London) 39:77-95, 1977

8. Riley PA: The mechanism of skin pigment production. Journal of the Society of Cosmetic Chemists 28:395-401, 1977

9. Pawelek JM, Lerner AB: 5,6-Dihydroxyindole is a melanin precur­sor showing potent cytotoxicity. Nature 278:627-628, 1978

10. Jimbow K, Quevedo WC, Fitzpatrick TB, Szabo G: Some aspects of melanin biology: 1950-1975. J Invest Dermatol 67:72-89, 1976

11. Fitzpatrick TB, Brunet P, Kukita A: The nature of hair pigment, The Biology of Hai..r growth. Edited by W Montagna, RA Ellis. New York, Academic Press, 1958, pp 255-303

12. Prota G, Nicolaus RA: Struttura e biogenesi delle feomelanine. I. Isolamento e proprieta dei pigmenti delle piume. Gazzetta Chim­ica Italiana 97:665-684 , 1967

13. Prota G, Nicolaus RA: On the biogenesis of phaeomelanins, Ad­vances in Biology of Skin, vol 8, The Pigmentary System. Edited by W Montagna, F Hu. New York, Pergamon Press, 1967, pp 323-328

14. Minale L, Fattorusso E, Cimino G, De Stefano S, Nicolaus RA: Struttma e biogenesi delle feomelanine. VIII. Sulla struttma della gallofeomelanine. 1. Gazzetta Chimica Italiana 99:431-449, 1969

15. Fattorusso E, Minale L, Sodano G: Feomelanine da nuove fon te naturali. Gazzetta Chimica Italiana 100:452, 1970

16. Ito S, Novellino E, Chioccam F, Mismaca G, Prota G: Copoly­merization of dopa and cysteinyldopa in melanogenesis in vitro. Experientia, in press

17. Prota G, Horsman H, Rosengren AM, Rosengren E: Phaeomelanic pigments from a human melanoma. Experientia 32:970-971, 1976

18. Sorby HC: On the coloming matters found in human hair. Journal of the Anthropological Institute (London) 8:1-14, 1878

19. Rothman S, Flesch P: Isolation of an iron pigment from human red hair. Proc Soc Exp Bioi Med 53:134-135, 1943

20. Barnicot NA: The pigment, trichosiderin, from human red hair. Natme 177:528-529, 1956

21. Bold P: Zur Kenntnis des Tricosiderins, eines Pigments aus roten Haaren. Natmwissenschaften 51:265, 1964

22. Bold P, Hermstedt E: Pyrrotricholes, eine Gruppe farbiger verbin Dugen a us roten Menschenhaar . Zeitsch.rift fur Naturforschung 22b:718-722, 1967

23. Prota G: Structm·e and biogenesis of phaeomelanins, Pigmentation: Its Genesis and Biologic Control. Edited by V Riley. New York, Appleton-Centmy-Crofts, pp 615-630, 1972

24. Thomson RH: The pigments of reddish hair and feathers. Angew Chern [Eng!) 13:305-312, 1974

25. Prota G, Horsman H, Rosengren AM, Rosengren E: Occurrence of trichochromes in the mine of a melanoma patient. Experientia 32:1122-1124, 1976

26. Agrup G, Lindbladh C, Prota G, Horsman H, Rosengren AM, Rosengren E: Trichochromes in the urine of melanoma patients. J Invest Dermatol 70:90-91, 1978

27. Okun MR, Edelstein LM, Patel RP, Donnellan B: A revised concept of mammalian melanogenesis: The possible synergistic functions of aerobic dopa oxidase and peroxidase. A review. Yale J Bioi Med 46:535-540, 1973

28. Patel RP, Okun MR, Edelstein LM, Cariglia N: B iochemical studies of the peroxidase mediated oxidation of tyrosine to melanin. Formation of dopachrome. J Invest Dermatol 63:374-377, 1974

MELANOGENESIS IN MAMMALS 127

29. Hearing VJ, Ekel TM: Involvement of tyrosinase in melanin for ­mation in murine melanoma. J Invest Dermatol 64:80-83, 1975

30. Smith PI, Swan GA: A study of the supposed hydmxylat ion of tyrosine catalyzed by peroxidase. Biochem J 153:403-408, 1976

31. Prota G, Scheri.Llo G, Nicolaus RA: Stru ttura e biogenesi delle feomelanine. IV. Sin tesi e proprieta della 5-S-cisteini.ldopa. Gaz­zetta Chimica Italiana 98:495-510, 1968

32. Fattorusso E, Minale L, De Stefano S, Cimino G, Nicolaus RA: Struttura e biogenesi delle feomelanine. IX. Feomela nine bios­intetiche. Gazzetta Chimica Italiana 99:969-978, 1969

33. Ito S, Nicol JAC: A new amino acid, 3- (2,5-S,S-dicysteinyl-3,4-dihydroxyphenyl) a lanine, from the tapetum lucidum of the gar (lepisosteidae) and its enzymic synthesis. Biochem J 161:499-507, 1977

34. Ito S, Prota G: A fac ile one-step synthesis of cysteinyldopas using mushroom tyrosinase. Experientia 33:1118-1119, 1977

35. Prota G, Crescenzi S, Misuraca G, Nicolaus RA: New intermediates in phaeomelanogenesis in vitro. Experientia 26:1058-1060, 1970

36. Crescenzi S, Misuraca G, Novellino E, Prota E: Reazioni modello per Ia biosintesi dei pigmenti feomelanici. Chimica e Industria (Milan) 57:392-393, 1975

37. Chioccara F, Novellino E, Prota G: A trimeric aldolization product of 1,4-benzothiazine. Journal of the Chemical Society, Chemical Communications 50-51, 1977

38. Minale L, Fattorusso E, De Stefano S, Nicolaus RA: Struttura e biogenesi delle feo melanine. XI. Ulteriori ricerche sulla biogenesi delle feomelanine . Gazzetta Chimica Italiana 100:461-466, 1970

39. Horsman H, Rosengren AM, Rosengren E: Determination of 5-S cysteinyldopa in melanomas with a flu orimetric method. Yale J Bioi Med 46:516-522. 1973

40. Horsman H: The melanocyte illuminated. Transactions of the St. John's Hospital Dermatalogical Society 60:135-141, 1974.

41. Agrup G, Hansson C, Kennedy BM, Persson K. Horsman H , Rosengren AM, Rosengren E: Analysis of cysteinyldopa isomers and some of their 0 methyl derivatives in the urine of melanoma patients by means of gas-cluomatography-mass spectrometry. Acta Derm Venereal (Stockh) 56:491-492, 1976

42. Prota G, Horsman H, Rosengren AM , Rosengren E: Isolation of 2-S-cysteinyldopa and 2,5-S,S-dicysteinyldopa from the urine of patients with melanoma. Experientia 33:720-721, 1977

43. Horsman H, Agrup G, Hansson C, Rosengren AM, Rosengren E: Detectwn of phaeomelamns, Pigment Cell, vol 4. Edited by S Klaus. Basel, Kru·ger, 1979, pp 244- 252

44. Dryia TP, Albert DM, Horsman H, Rosengren E, Reid TW: Pres­ence of cysteinyldopa in the mature bovine eye. Exp Eye Res 25: 459-461, 1977

45. Prota G, Seru·Je AG: Biochemical sites of gene action for melano­genesis in mammals. Annates de Genetique et de Selection Ani­male 10:1-8, 1978

46. Agrup G, Falck B, Horsman H, Rosengren AM, Rosengren E: Glutathionedopa in malignant melanoma. Acta Derm Venereal (Stockh) 57:221-223, 1977

47. Prota G: Cysteine and glutathione in human pigmentation, Bio­chemistry of Low Moleculru· Weight Sulphm Containing Com­pounds. Edited by V Zappia. New York, Plenum Press, 1980

48. Halprin KM, Ohkawru·a A: Human pigmentation: The role of glutathione, Advances in Biology of Skin, vol8, The Pigmentary System. Edited by W Montagna, F Hu. New York, Pergamon Press, pp 241-251, 1967