Chemical Characterization of Hair Melanins in Various Coat-Color Mutants of Mice

Cheniical Characterization of Hair Melanins in Various Coat-Color Mutants of Mice Hiroyuki Ozeki, Shosuke Ito, Kazumasa Wakamatsu, and Tomohisa Hirobe* Fujita Health University School of Health Sciences, Toyoake, Aichi; and 'National Institute of Radiological Sciences, Division of Biology, Chiba, Japan

JMammaliaii melanins exist in two chemically distinct forms: the broivn to black eumelanins and the yelloiv to reddish pheomelanins. Melanogenesis is influenced hy a numher of genes, the levels of whose products determine the quantity and quality of the melanins produced. To examine the effects of various coat-color genes on the cheniical properties of melanins synthesized in the follicular melanocytes of mice, we have introduced new methods to soluhilize differentially pheomelanins and hrowm-type eumelanins. W^e applied these and previously developed high-performance liquid chromatography and spectrophotometric methods for assaying eu- and pheomelanins to characterize melanins in various mutant mice: hlack, lethal yellow, viahle yellow^, agouti, hrow^n, light, alhino, dilute, recessive yellow, pink-eyed dilution, slaty, and silver. It was demonstrated that 1) complete soluhilization of melanins in Soluene-350 is a convenient method to estimate the

total amount of eu- and pheomelanins, 2) lethal yellow, viahle yellow, and recessive yellow hairs contain almost pure pheomelanins, and 3) melanins firom hrow^n, light, silver, and pink-eyed hlack hairs share cheniical properties in common that are characterized hy partial soluhility in strong alkali. We suggest that 1) the hrown-type eumelanins have lower degrees of polymerization than the hlack-type eumelanins, and 2) slaty hair melanin contains a greatly reduced ratio of 5,6-dihydroxyindole-2-carhoxylic acid-derived units as compared with hlack and other eumelanic hair melanins. These results indicate that our methodology, high-performance liquid chromatography and spectrophotometric methods combined, may he useful in chemically characterizing melanin pigments produced in follicular melanocytes. Key words: HPLC/eumelanin/pheomelanin/melanogenesis. J Invest Dermatol 105:361-366, 1995

M

catalyzes the tautomerization of dopachrome to 5,6-dihydroxyindole-2-carboxylic acid (DHICA), rather than the decarboxylative rearrangement to 5,6-dihydroxyindole (DHI). Recently, this enzyme was shown to be identical to tyrosinase-related protein-2 (TRP2), wliich is encoded at the slaty locus [7,8]. Although it is known that oxidative polymerization of DHI can be catalyzed by tyrosinase [9], it remained unknown whether a factor promoted the oxidation of DHICA. It has recently been shown that the brovsTi locus encodes T R P l , another tyrosinase-related protein, which can catalyze the oxidation of DHICA [10,11]. Three more genes have recently been cloned that affect melanogenesis. These are the silver gene, wliich encodes a melanosomal matrix (structural) protein [12,13]; the pink-eyed dilution gene, which encodes another melanosomal protein, believed to function as a tyrosine transporter [14]; and the dilute gene, which encodes a myosin-related protein, essential to the release of melanosomes into hair shafts [15].

elanin pigments in the skin, hair, and eyes of mammals and of lower species are tastially classified into two groups: the black to brown eumelanins that are insoluble in acid and alkali and the yellow to reddish pheomelanins that are soluble in alkah [1,2]. Both types of melanin pigments are synthesized in the melanocyte from a common precursor, dopaquinone, produced fi-om tyrosine by the action of tyrosinase (Fig 1). Dopaquinone may undergo a series of redox reactions leading to the formation of eumelanins, or alternatively, in the presence of sulfhydryl compounds such as cysteine or glutathione, can act as a precursor of pheomelanins via cysteinyldopas. There are about 60 loci and more than 150 mutations that are involved in the expression of coat color in mice [3]. In addition to tyrosinase, which is encoded at the albino locus, two related proteins have been shown to regulate melanogenesis, especially eutnelanogenesis (for review, see [4,5]). Dopachrome tautomerase, the presence of which has been suggested for some years [6], Manuscript received March 4, 1995; final revision received May 10, 1995; accepted for publication May 16, 1995. Reprint requests to: Dr. Shosuke Ito, Fujita Health University School of Health Sciences, Toyoake, Aiclii 470-11, japan. Abbreviations: AHP, aminohydroxyphenylalanine; DHI, 5,6-dihydroxyindole; DHICA, 5,6-dihydroxyindole-2-carboxylic acid; PTCA, pyrrole2,3,5-tricarboxylic acid; SE, spectrophotometric eumelanin; SP, spectrophotometric pheomelanin; TRP, tyrosinase-related protein.

In contrast to eumelanogenesis, little is known of regulatory mechanisms that might be functional in pheomelanogenesis. Two genes have recently been cloned that are involved in the switch between eu- and pheomelanogenesis. The extension gene encodes an a-melanocyte-stimulating hormone receptor that is localized on the surface of melanocytes [16], whereas the agouti gene encodes a protein that is produced outside the melanocyte and acts as an antagonist of a-melanocyte-stimulating hormone [17-19]. Although biochetnical mechanisms of the switch between eu- and pheomelanogenesis is not well understood, one of the major regulatory factors appears to be tyrosinase [20], the activity of

0022-202X/95/S09.50 • SSDI0022-202X(95)00265-M • Copyright © 1995 by The Society for Investigative Dermatology, Inc.

362

OZEKI ET AL

THE JOURJ^AL OF INVESTIGATIVE DERJVIATOLOGY

KMnO 4 (HOOCr

2.3% yield

Benzothiazlne der.

HjW^^COi

HOOC

CCKJH

Cystelnyldopas PTCA Cysteine I COOH

Q^

COOH

Oj

O.

/>^

/

\

KMnO,

^COOH

0.03% yield Tyrosina

Dopaquinorw (DQ)

Dopa

-Cysteine

DHI-derived eumelanin

Leucodopachrame (O)

5,6-Dlhydroxylndole (DHI)

\ '

(HOOC) Dopachrome Pheomelanin 5,6-Dlhydroxylndole .2-carboxyllc acid (DHICA)

• tmtti (TBP2)

Figure 1, Outline of eumelanogenesis and pheomelanogenesis in melanocytes.

which is regulated by tbe levels of the extension and agouti gene products [21]. The clonhig of those pigmentation-related genes and the identification of their roles in melanogenesis suggest that the classification of melanins into eu- and pheomelanins is oversimplified. For example, how do eumelanins in slaty and brown hairs differ from eumelanin in black hair? These questions can be answered only by chemical approacbes. We have already established microanalytical high-performance liquid chromatography (HPLC) metbods to quantitate eu- and pheomelanins in tissue samples (Fig 2) [22,23]. In addition, we recently developed a spectrophotometric method for assaying eumelanins in pigmented tissues [24]. To examine the effects of various coat-color genes on chemical properties of the melanins synthesized, we now introduce new methods to differetitially solubilize pheomelanins and brown-type eumelanins and we apply these to characterize melanins produced in folhcular melanocytes of congenic mice that are mutant at a number of different coat-color loci. MATERIALS AND METHODS Chemicals Sepia melanin was purchased from Sigma (St. Louis, MO) and used as a standard for spectrophotometric assay of total melanin and eumelanin and of HPLC assay of eumelanin yielding pyrrole-2,3,5-tricarboxyUc acid (PTCA). Soluene-350 was a product of Packard (Meriden, CT). All other chemicals were of the liighest purity available. All butfers and suspensions of mouse hair were prepared in glass-distilled, deionized water of Milli-Q quality. Mice Five lities congenic with C57BL/10JHir (BIO) background have been established in otie of the authors' laboratories [25,26]. The genie constitution of those lines difFers from black {a/ti) animals only in one of the coat-color loci: agouti {A/A), brown {b/b), albino (c/c), dilute {d/tt), or pink-eyed dilution {p/p). The phenotypes and genotypes of those and other mice used in this study are summarized in Tahle I. An agouti line from a diiferent strain (C3H/HeJmsHir) was also established in the same laboratory. Additional hair samples were obtained from the following laboratories: lethal yellow {A'/a) and reeessive yellow {e/e) from Dr. H. Yamamoto of Tohoku University, viable yellow {A"'/A"'') from Dr. A. J. Thody of University of Newcastle Upon Tyne, slaty {sit/sit) and its counterpart hlack {Slt/SIl) from Dr. V. J. Hearing of NIH, and silver {si/.ii) and light {B"/B") from Dr. W. C. Quevedo, Jr. of Brown University. The hairs of silver mice are dark distally (tips) and light proximally (at their bases). Except for the fact that the tips of the hairs of silver mice were darker than those of light

Figure 1. Chemical degradation of eumelanin to form PTCA and of pheomelanin to form AHP, Note that the yield of PTCA from DHI-derived eumelanin is extremely low, as compared with that from DHICA-derived eumelanin. PTCA is thus a specific degradation product of DHICA-derived eumelanins |22,35]. These PTCA yields were obtained for enmelanins prepared by oxidation of 1 mM DHI or DHICA at pH 6.8 with mushroom tyrosinase (unpublished results); we assume that the molecular weights of monomer units of DHI-melanin and DHICA-melanin are 182 and 227 (including hydrated water molecules) [35] and that the melanins were quantitatively produced. Hydriodic acid hydrolysis of pheomelanin yields two isonieric AHP, only one of which is shown here. The 20% yield was a sum for both isomers of AHP [22].

mice, the phenotypes of light and silver mice were comparable with regard to basal depigmentation of the hair. Preparation of Samples A suspension of sepia melanin in water at a concetitration of 1 ing/ml was prepared by sonicating with a Microson Ultrasonic Cell Disrupter (Fanningdale, NY) at a half power for 5 min. Hair samples were obtained by plucking mid dorsal hairs from 5-week-old mice (unless otherwise stated). Aqueous suspensions of hair samples were prepared by homogenizing about 30 mg of hair in water at a concentration of 10 mg/ml with a Ten-Brocke glass homogenizer. One-hundred-microliter suspensions (containing 100 /xg of sepia melanin or 1 mg of hair) were used for spectrophotometric assays of total melanins and eumelanins and HPLC analyses of PTCA and aminohydroxyphenylalanine (AHP). Twohundred-microliter suspensions were used for spectrophotometric assays of pheomelanins and alkali-soluble melanitis. HPLC Assays of Eumelanin and Pheomelanin and Spectrophotometric Assay of Eumelanin For the HPLC determination of eumelanin. hair samples were oxidized with permanganate to give PTCA, wliich was quantitated by HPLC with ultraviolet detecdon [22]; an improved method was used in this study [27]. Each determination was performed in duplieate. For the HPLC determination of pheomelanin, hair samples were hydrolyzed witb hydriodic acid to give AHP, which was quantitated by HPLC with electrochemical detection [22]. One nanograni of PTCA and 1 ng of AHP correspond to 50 ng of eumelanin and 5 ng of pheomelanin, respectively, except in slaty hair (see below). For the spectrophotometric determination of eumelanin, hair samples (100 /xl) were hydrolyzed in hot hydriodic acid, and insoluble eumelanic pigments were solubilized in hot NaOH in the presence of H2O2 and analyzed for absorbances at 350 nm (Ajs,,) [24]. The A,;,, values are referred to as spectrophotometric eumelanin (SE). Tliis method is highly specific for eumelanins, although it does not discriminate between DHI and DHICA melanins [24]. Each measurement was performed in duplicate. Spectrophotometric Assay of Total Melanin Samples (100 /xl) were placed in screw-capped test tubes, to which 900 \iX of Soluene-350 was added [28]. Tubes were vortex-mixed and heated in a boiling water bath for 30 min. After cooling, tubes were vortex-mixed again and heated in a boiling water bath for an additional 15 min. Slightly viscous, brown

VOL. 105, NO. 3 SEPTEMBER 1995

Table I.

Pbenotype Experiment 1 Black Letbal yellow Viable yellow Agouti Agouti Brown Albino Dilute black Recessive yellow Pink-eyed black Experiment 2 Black Slaty Silver Light

HAIR MELANINS IN MUTANT MICE

363

Melanin Contents in Hairs from Various Coat-Color Mutants of Mice" TM

PTCA

AHP

SE

SP

ASM

1 mg

1 mg

1 tng

1 mg

2 mg

2 mg

Genotype

Strain

aaBBCCDDEEPP A'aBBCCDDEEPP A"'A">'BBCCDDEEPP AABBCCDDEEPP AABBCCDDEEPP aahhCCDDEEPP aaBBccDDEEPP aaBBCCddEEPP aaBBCCDDeePP aaBBCCDDEEpp

C57BL/10JHir C57BL/6J-/lV(i C3H/HC-A'''' /A"'' C57BL/10JHir-/l/^ C3H/HeJmsHir C57BL/10JHir-/>/;) C57BL/10JHir-f/f C57BL/10]Hir-d/d C57BL/6J-f/f C57BL/10JHir-p/p

5w 5w 6m 5w 5w 5w 5w 5w 5w 5w

0.612 0.096 0.070 0.588 0.551 0.189 0.018 0.417 0.098 0.051

1.56 0.022 0.015 1.27 1.53 0.459 0.002 1.42 0.013 0.070

0.043 3.90 1.96 0.837 0.516 0.089 0.018 0.047 4.00 0.091

0.560 0.001 0.005 0.420 0.418 0.083 0.004 0.315 0.002 0.021

0.011 0.168 0.115 0.034 0.016 0.022 0.004 0.010 0.208 0.010

0.092 0.449 0.275 0.188 0.120 0.188 0.038 0.075 0.522 0.080

aaBBSiSiSltSlt aaBBSiSisltslt a'a'BhsisiSltSIt aaB"B"SiSiSltSlt

C57BL/6J C57BL/6J-i/(A/r

8w 8w 9w 9w

0.783 0.481 0.101 0.048

2.17 0.210 0.217 0.059

0.042 0.026 0.139 0.052

0.692 0.273 0.051 0.016

0.008 0.008 0.014 0.008

0.065 0.091 0.109 0.071

—

LT/Ch

Age

" Homogenates containing eitber 1 or 2 mg of hair were examined for the contents of total melanin (TM), PTCA, AHP, SE, SP, and alkali-soluble melanin (ASM). Absorbances were measured in extracts of 1 ml volume. For experiment 1 we used two different mice in each strain and the values are tbe averages. For e.xperiment 2 we used one niouse in each strain. solutions were transferred to tnicro-test tubes atid cleared by centrifngation at 10,000 rpm for 10 min. Supematants were analyzed for ab.sorbances at 500 nm (Aj,,,,). Tbe A,,,,, values arc referred to as total tnclanin. Absorbances at sborter wavelengtbs were bigher tbati Ag,,,,; bowever, increase iti background absorption due to proteins becatnc significant. Eacb detennination was performed in duplicate, altbougb reproducibility oftlie measurements was fairly good with a coefficietit of variation of 4.0% for black bair (n = 10). Spectrophotometric Assay of Pheomelanin Samples (200 fj.1) were placed in micro-test tubes, to wbicb 800 fjil of 0.1 M sodium pbospbate buffer, pH 10.5, was added. Tubes were vigorou.sly mixed in a Taiyo Mix Tower A-14 (Taiyo In.struments, Tokyo) for 10 niin at room temperature and centrifuged at 10,000 rpm for 10 min. Supematants were transferred to screw-capped test tubes, and 1 tnl of chloroform was added to remove fatty impurities. Tubes were vorte.\-mixed and centrifttged at 4,000 rpm for 10 min. Pale yellow aqueous layers were transferred to micro-test tubes and cleared by centriftigation at 10,000 rptn for 10 tnin. Supernatants were analyzed for absorbances at 400 nm (A^,,,,), tbe values being referred to as spectropliotometric pheomelanin (SP). Eacli determinatioti was performed in duplicate, altbougb reproducibility was fairly good with a coefficient of variation of 3.5% for viable yellow bair (n = 10).

Black, brown, viable yellow, and albino mice hairs were subjected to solubilization in alkaline buffers of various pH. The degree of solubilization was assessed by absorption at 400 nm (A^,,,,). The pH-absorbance curves shown in Fig 4 indicate that pheomelanin in viable yellow hair became progressively soluble in buffers above pH 10, whereas eumelanin in brown hair was partially solubilized in buffers of higher pH. To discriminate between pheomelanin and brown eumelatiin, we adopted a pH 10.5 buffer to estimate the level of pheomelanin. 8 M urea/1 M NaOH was used to estimate the level of alkali-soluble melanin (pheomelanin and brown eumelanin to a lesser extent). The 8 M urea/1 M NaOH solution gave slightly higher values than the 1 M NaOH that was used in the experiments shown in Fig 4. Characterization of Melanins in Hairs from Various CoatColor Mutants of Mice Contents of total melanin corresponded well to visual appearance of color ititensities of the hairs (Table I). Total melanin contetits in brown and pink-eyed black hairs were reduced to one third and less than one tenth that in black hair.

Spectrophotometric Assay o f Alkali-Soluble Melanin Tbe metbod was essentially identical to tlie spectrophotometric assay of plieotnelanitis except for solubilizing the samples in 8 M urea/1 M NaOH, wliicb was prepared by dissolving urea in 1 M NaOH at an 8 M concentration. Urea was added to facilitate swelling tbe hair samples [29]. Pale browti supematants were analyzed for absorbances at 400 um (A4,,,,), the values being referred to as alkali-soluble melanin. Witb tliis metbod, pbeomelanins were mostly solubilized, whereas brown eumelanius were partially solubilized. Each determination was perfomied in duplicate, altbough reproducibility was fairly good with a coelFicient of variation of 4.6% for brown hair (n = 10). RESULTS Methods Evaluation Ten different hair samples of various coat-color phetiotypes were analyzed with several clieniical methods. Table I (experimetit 1) summarizes the phenotypes and the melanin contents of hair samples from the mutant mice used to evaluate the chemical methods. Total melanin content was determined as the A5111, value after solubilizing hair samples in Soluene350; even black hair was completely dissolved. There was an excelletit correlation between the total melauin level and the melanin content calculated from PTCA and AHP values: the correlation coefFicient was 0.974 for the 10 hair samples (Fig 3). This indicates that the Aj,,,, value can serve as an indicator of the total amount of eumelanin and pheomelanin combined, regardless of the type of melanin.

20

40

60

80

100

120

PTCA X 50 + AHP X 5 (

Figure 3. Correlation of Total Melanin Content with Melanin Content Calculated from PTCA and AHP Values in Mouse Hair. The total melanin levels were calculated against a .standard, sepia melaiiiti, which contains 60% melatiiti [24]. Closed circles are based on the data in experiment 1 of Table I (averages for two mice eacb) and open circles on tbe data in the experitnent 2 of Table I (data for one mouse each). The regressioti curve was obtained for the former 10 .samples.

364

THE JOURNAL OF INVESTtGATIVE DERMATOLOGY

OZEKI ET AL

were very low in lethal yellow, viable yellow, and recessive yellow hairs. In contrast, the opposite holds for the AHP/total melanin ratios. These results are consistent with the genetic background that some mutations at agouti and extettsion loci cause the switch from eu• : Black to pheomelanogenesis [3,21]. Although pink-eyed black melanin A : Brown O : Viable yellow contains a measurable level of pheomelanin (AHP), it can be A : Albino classified to be essentially eumelanic, judged fi-om comparison of the PTCA/total melanin and AHP/total melanin ratios with others (Fig 5). The SE/total melanin ratios were liigh in eumelanic hairs, whereas they were negligible in pheomelanic hairs. There is thus a parallel relationsiiip between PTCA and SE levels as far as the classification of melanin type is concerned. It should be noted, however, that the SE/total melanin ratio of brown hair melanin was only half that of black hair melanin. Pink-eyed black hair melanin also had a low SE/total melanin ratio, similar to that of brown. The SP/total melanin ratios paralleled well to the AHP/total melanin ratios, indicating that pheomelanin was selectively solubilized in the 10.5 buffer. Therefore, SP/total melanin ratio can be Figure 4. Efiects of pH on solubilization of melanins from mouse hair. Buffers used are 0.1 M sodium phosphate buffer for pH 7.5, 8.5, 9.5, used as a convenient substitute for the AHP/total melanin ratio in 10.5, and 11.4, 0.1 M trisodium phosphate (pH 12.3), and 1 M NaOH (pH the study of melanogenesis in mouse hair. 13.4). The pHs were measured with a pH meter and were not accurate at The alkali-soluble melanin/total melanin ratios indicate that not higher pHs. The experimental procedure was essentially identical to that only pheomelanins but also eumelanins in brown and pink-eyed described in Materials and Methods under Spectrophototnetric Assay of Pheomelanin. Hair homogenates (2 mg/200 ix\) were mixed with 800 jU,l of a buffer, black hairs became partially soluble in the strong alkali. It thus appears that brown and pink-eyed black hair eumelanins are and the extracts were measured for the absorbances at 400 nm after removal considerably different from black hair eumelanin. Dilute black hair of lipids with chloroform. eumelanin has solubility properties almost identical to those of black hair eumelanin, although these eumelanins show somewhat whereas they were one tenth to one fifth in pheomelanic yellow different PTCA/total melanin and SE/total melanin ratios. hair. Hair melanins were characterized by the levels of PTCA, Characterization of Melanins in Hairs from Slaty, Silver, AHP, SE, SP, and alkali-soluble melanin. For comparison among and Light Mice Having established that the chemical charactervarious hair melanins which vary greatly in the absolute levels, ization of hair melanins is a powerful tool to differentiate various these values were divided by the total melanin contents (Fig 5). types of melanins, we then applied the methodology to characterize The PTCA/total melanin ratios were at high and similar levels in melanins in slaty, silver, and light hairs. The light is a dominant black, brown, dilute black, and pink-eyed black hairs, whereas they

BL BR OB PB LY VY RY SL SI LT

BL BR DB PB LY VY RY SL SI LT

BL BR DB PB LY VY RY SL SI LT

BL BR DB PB LY VY RY SL SI LT

BL BR DB PB LY VY RY SL SI LT

60-

r t

O

o 30

10

n

_

rOni-O-1

BL BR DB PB LY VY RY SL SI LT

BL ; Black I I LY : Lethal yellow m

SL ; Slaty

^ ^ g BR : Brown

^ ^ ^ DB ; Dilute black

I

I

I VY ; Viable yellow I SI : Silver

I PB ; Pitik-eyed black

I RY : Recessive yellow T : Light

Figure 5. Characterization of melanins in hairs from various coat-color mutants of mice. Drawn on the basis of data shown in Tahle I. Each value was calculated after subtraction of the background value for non-pigmented, albino hair (Tahle I). Data for agouti hairs are omitted. Open circles represent values for each mouse. TM, total melanin; Sp.EM, spectrophotometric eumelanin (SE); Sp.PM, spectrophotometric pheomelanin (SP); ASM, alkali-soluble melanin.

|

VOL. 105, NO. 3 SEPTEMBER 1995

mutant allele ofthe brown locus that results in hairs pigmented only at their tips [30]. Results of the characterization of melanins in those hairs are summarized in Table I (experiment 2) and Fig 5. Slaty hair contained approximately two thirds as much eumelanin as hlack hair when the total melanin levels were compared. However, comparison of the PTCA/total melanin ratios indicates that slaty hair eumelanin is quite different from-black hair eumelanin: the ratio was only one fifth that of black. This was also reflected in the conspicuous deviation of the slaty hair from the regression curve shown in Fig 3. Solubility properties of slaty hair melanin were rather close to those of black hair melanin. Silver hair melanin is interesting in that it showed chemical properties almost identical to those of brown hair melanin. Light hair melanin also had chemical properties very similar to brown hair melanin. Total melanin contents in these hairs were less than one fifth to one tenth that in black, corresponding to the visual appearance. DISCUSSION T h e coat color of mice varies from black, brown, yellow, and gray, to white with subtle variations in each color [3]. Hair and wool from other mammals show even more complex spectra of color patterns. The most iinportant factors in determining hair color is obviously the quantity and quality of the melanin pigments produced in melanocytes and the amounts that are transferred to hair shafts. We developed a microanalytical HPLC method to quantitate e u - and pheomelanin after chemical degradation to PTCA and A H P , respectively [22,23,31]. This method has been used to analyze melanins in a variety of tissue samples [20]. In hair and wool, various types of melanins are present that include co-polymers or mixtures of eu- and pheomelanins. A convenient method to determine contents of total melanin should thus be useful in studying melanogenesis. A spectrophotometric determination of melanin contents in tissue samples requires solubUization of melanin pigments. For cultured melanocytes, solubiUzation in hot NaOH or KOH has routinely been used. It should be noted, however, that eumelanin in black mouse hair could not be solubilized with this method. We therefore chose Soluene-350, a tissue solubilizer used in scintillation counting, to completely solubiUze melanin pigments in tissue samples [28,32]. The A^,,,, value is a convenient indicator ofthe content of eu- and pheomelanin. Pheomelanins can be extracted with alkali from yellow mouse hair and from reddish chicken feathers [1]. However, it was recently shown that brown-type eumelanins can also be extracted from hairs and feathers with alkaline buffers [29]. In the present study we showed that pheomelanins and brown-type eumelanins can be differentially solubilized under the conditions in which black-type euinelanins remained insoluble. Pheomelanins in lethal yellow, viable yellow, and recessive yellow hairs were found to be similar to each other in chemical properties. Those hairs did not contain significant levels of eumelanins, as judged by the PTCA/total melanin and SE/total melanin ratios. A biochemical property common to those mutations is the decrease of tyrosinase activities as compared with the hlack counterparts [20,33]. Tliis confirms earlier suggestions that tyrosinase activity is a major determinant in the switch from eu- to pheomelanogenesis in mouse hair [20,34]. It is possible, however, that other factor(s) also play a role in this regulatory point. It is now well accepted that natural eumelanins are produced by co-polymerization of DHI and DHICA [35-37]. The DHI to DHICA ratio in natural eumelanins appears to be governed by the activity of TRP2/dopachrome tautomerase and tyrosinase [38] (Fig 1). In addition, TRl'l/brown locus protein is now shown to oxidize DHICA [10,11], and it is thus possible that T l ^ l might also affect the DHI to DHICA ratio. The DHICA content in natural eumelanins can be estimated from the PTCA content (Fig 2). T h e slaty mutation in mice results in the dilution of coat color to dark gray to brown rather than black as in the wild type at this locus [39]; this is in agreement with the finding that the total melanin

HAIR MELANINS IN MUTANT MICE

365

content in slaty hair was approximately two thirds that in black hair. On the other hand, the PTCA/total melanin ratio in slaty hair was only one fifth, suggesting that slaty iiair melanin contains DHICAderived units at a level one fifth that of black. The dramatically decreased ratio of DHICA-derived units is probably related to the decreased activity of TRP2/dopachrome tautomerase in the slaty mutation [7,40]. The brown mutation changes the color of melanin produced from black to brown; it also reduced the quantity of melanin produced. The brown locus encodes T R P l , which is shown to have a DHICA oxidase activity [10,11]. However, contrary to the expectation that the brown mutation might affect the DHI to DHICA ratio in the melanin produced, brown hair melanin has a PTCA/total melanin ratio almost identical to that of black. This indicates that the DHI to DHICA ratios are almost identical in black and brown hair melanins. This could be explained by two possible mechanisms: a) DHICA quinone, the immediate product of DHICA oxidation, binds DHI to fonn mixed-type intennediates [38], or b) DHICA quinone oxidizes DHI to DHI quinone as DHICA quinone is reduced back to DHICA. Both mechanisms eventually lead to co-polymerization of DHI and DHICA. Whatever the role of TRPl, its activity appears essential for eumelanogenesis to proceed [41,42]. In brown mice, mutant TRJ"! appears unable to be transferred to melanosomes [43]. What would then happen if functional TRPl is not present in melanosomes? It is speculated that, in the brown melanosomes, the oxidation of DHI and DHICA, followed by spontaneous polymerization, proceeds rather slowly, thus forming possibly a low—molecular-weight eumelanin (with a low degree of polymerization). It is likely that a large proportion of the melanin precursors leaks out to the cytoplasm in the brown melanocytes. An example of such phenomena was seen in the light mutation at the brown locus [30]. Light mice hairs have melanin pigments only at their tips because follicular melanocytes of the mice died prematurely, possibly through the inherent cytotoxicity of melanin precursors, especially of DHI [44]. The SE/total inelanin ratios in brown and light hair are approximately half of that in black hair. The spectrophotometric assay of eumelanin involves treatment with hot hydriodic acid to solubilize pheomelanin. This treatment may induce euinelanins with low degrees of polymerization to be degraded to become soluble in hot hydriodic acid. Taken together, these findings suggest that the SE/total melanin ratio may reflect the degree of polymerization of the melanin precursors. A biochemical study to prove tliis hypothesis is now in progress using synthetic melanins prepared from DHI and DHICA. Furthermore, the alkali solubility of brown and light mouse hairs may also be explained by the low molecular weights of these melanins. Orlow ct al [29] proposed that DHICA melanins are determinants of brown colors in the animal kingdom. In the light of the present results, however, the difference between black and brown colors may not primarily be caused by the difference in the content of DHICA-derived units. The silver lnutation causes progressive graying of hair due to the loss of functional melanocytes [45]. This mutation seems likely to induce toxic effects to melanocytes analogous to those caused by the light mutation at the brown locus [46]. In the present study silver hair melanin was shown to be similar in chemical properties to brown and light hair melanins. It should also be added that hair follicles of the light-silver mice were shown to produce diffuse ("soluble") melanin within dying melanocytes, which may result from the release of cytotoxic melanin precursors.t The pink-eyed dilution gene encodes an integral melanosomal membrane protein and mutations at this locus cause a great suppression of eumelanin production [14]. It has been suggested that the p protein may function as a transporter of tyrosine, although other functions are also suspected [14]. Pink-eyed black hair contains a greatly reduced level of pigment of eumelanic t Quevedo WCJr, Holstein TJ, Dyckman J: Further observation on the nature of premature melanocyte death in the hair follicles of light-silver mice (abstr). Pif-titent Cett Res 3(suppl):31, 1994.,

366

OZEKl ET AL

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

character. It is interesting to note that brown, silver, and pink-eyed dilution mutations result iu the production of eumelauius with similar chemical properties, although the biochemical mechanisms underlying those mutations are distinctively diiFereut. We speculate that suppression ofeumelanogenesis, regardless ofthe stage, leads to the production of browu-type eumelanins rather than black-type eumelanins. Further suppression would then result iu a switch to the production of pheomelanins. The dilute mutation produces a lightening of coat color, caused by an uneven release of melanosomes into the developing hair shaft. The dilute gene encodes a uovel type of myosiu heavy chain that may have an important role in the elaboration, maintenance, or function of dendrites of melanocytes [15]. Myoxin (MYH12), the human homologue to the mouse dilute gene, has recently been cloned and sequenced [47]. The chemical properties of dilute black hair melanin seem somewhat different from those of black. However, how this is related to the effects of this mutation is uot clear at present. Agouti hairs contain considerable amounts of pheomelauius. Yet SP values were very low, indicating that pheomelanins iu those hairs were difficult to be solubilized at pH 10.5. The reason for this is not clear at present. We have developed novel methods to differentially solubilize pheomelanins and brown-type eumelanins and have applied these, along with previously developed HPLC and spectrophotonietric assay methods, to characterize melauiu pigments produced in follicular melanocytes of mice that are mutant at various coat-color loci. Iu principle, the methodology may also be applied to hair, wool, aud feather samples from species other than the mouse.

16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26.

27. 28. 29. 30. 31.

IVe are grateful to Drs. Viuceut J. Hearing, Walter C. Quevedo Jr., Anthony]. Thody, and Hiroaki Yamamoto for providing mouse hair samples. We also wish to acknowledge Drs. V. J. Hearing aud Francisco Solano (Uitiuersity of Mtircia) for helpful discussions and making unpublished works available to us.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

11.

12. 13. 14.

15.

Prota G: Melauitis aud Metauogencsis. Academic Press, New York, 1992, pp 290 Ito S: Biochemistry and physiology of melanin. In: Levine N (ed.). Pigmentation and Pigmentary Disorders. CRC Press, Boca Raton, 1993, pp 33-59 Silvers WK: The Coat Cotor of Mice: A Modet for Mamtuatian Gene Action and Interaction. Springer-Verlag, New York, 1979, pp 379 Urabe K, Aroca P, Hearing VJ: From gene to protein: determination of melanin synthesis. Pigment Cetl Res 6:186-192, 1993 Hearing VJ: Invited editorial: unraveling the melanocyte. Am J Hum Geuet 52:1-7, 1993 PawelekJ, et al: New regulators of melanin biosynthesis and the autodestruction of melanoma cells. Nature 286:617-619, 1980 Tsukamoto K, Jackson IJ, Urabe K, Montague PM, Hearing VJ: A second tyrosinase-related protein, TRP2, is a nielanogenic enzyme termed DOPAchrome tautomerase. EMBO j 11:519-526, 1992 Jackson IJ, Chambers DM, Tsukamoto K, Copeland NG, Gilbert DJ, Jenkins NA, Hearing VJ: A second tyrosinase-related protein, T 1 ^ 2 , maps to and is mutated at the mouse slaty locus. EMBO f 11:527-535, 1992 Tripathi RK, Hearing VJ, Urabe K, cf al: Mutational mapping of the catalytic activities of human tyrosinase. J Biol Chem 267:23707-23712, 1992 Jimenez-Cervantes C, Solano F, Kobayashi T, Urabe K, Hearing VJ, LozanoJA, Garcia-BorronJC: A new enzymatic function in the melanogenic pathway: the 5,6-dihydroxyindole-2-carboxylic acid oxidase activity of tyrosinase-related protein-1 ( T R P l ) . / Biol Cliem 269:17993-18001, 1994 Kobayashi T, Urabe K, Winder A, Jimenez-Cervantes C, Imokawa G, Brewington 1\ Solano F, Garcia-Borron |C, Hearing VJ: Tyrosinase related protein 1 (TRPl) functions as a DHICA oxidase in melanin biosyntbesis. EMBO J 13:5818-5825, 1994 Zhou BK, Kobayashi T, Donatien PU, et at: Identification of a nielanosomal matrix protein encoded by the murine 51 (silver) locus using "organelle scanning." Proc Natt Acad Sci USA 91:7076-7080, 1994 Kobayashi T, Urabe K, Orlow SJ, Higashi K, Imokawa G, Kwon 13S, Potterf B, Hearing VJ: The Pmell7/.sr7wT locus protein: characterization and investigation of its melanogenic function. ^ Biot Chem 269:29198-29205, 1994 Rosemblat S, Durham-Pierre D, Gardner JM, Nakatsu Y, Brilliant MH, Orlow SJ: Identification of a melanosomal mcmhrane protein encoded by the pink-eyed dilution (type II oculocutaneous albinism) gene. Proc Natl Acad Sci USA 91:12071-12075, 1994 Mercer JA, Seperack PK, Strobel MC, et al: Novel myosin lieavy chain encoded by murine dilute coat colour locus. Nature 349:709-712, 1991

32. 33.

34. 35. 36. 37. 38. 39. 40.

41. 42. 43. 44. 45.

46. 47.

bobbins LS, NadeauJH. Johnson KLR, et al: Pigmentation phenotypes of variant extension locus alleles result from point mutations that alter MSH receptor function. Celt 72:827-834, 1993 Bultman SJ, Michaud EJ, Woychik RP: Molecular characterization ofthe mouse agouti locus. Cell 71:1195-1204, 1992 Miller MW, Duhl DMJ, Vrieling H, et al: Cloning ofthe mouse agouti gene predicts a secreted protein ubiquitously expressed in mice carrying the lethal yellow mutation. Genes Devel 7:454-467, 1993 Lu D. Willard D. Patcl IR. Kadwell S, Overton L, Kost T, Luther M, Chen W, Woychik RP, Wilkinson WO, Cone KD: Agouti protein is an antagonist of the melanocyte-stiniulating hoinione receptor. Nature 371:799 — 802, 1994 Ito S: High-performance liquid chromatography (HPLC) analysis of eu- and pheomelanin in melanogenesis control. J Jurest Dermatol 100:166S-17lS, 1993 Jackson IJ: Colour-coded switches. NrtMirc 362:587-588, 1993 Ito S, Fujita K: Microanalysis of eumelanin and pheomelanin in hair and melanomas by chemical degradation and liquid chromatography. Anal Biochem 144:527-536, 1985 Ito S, Fujita K, Takahashi H, Jimbow K: Characterization of melanogenesis in mouse and guinea pig hair by cbemical analysis of melanins and of free and hound dopa and 5-S-cysteinyldopa. J Invest Dermatol 83:12-14, 1994 Ito S, Wakamatsu K, Ozeki H: Spectrophotonietric assay of eumelanin in tissue samples. Aual Biochem 215:273-277, 1993 Hirobe T: Congenic strains of C57BL/10JHir. Mouse News Lett 76:28-29, 1986 Taniate HB, Hirobe T, Wakamatsu K. Ito S, Shibahara S, Ishikawa K: Levels of tyrosinase and its mlUMA in coat-color mutants of C57BL/1 OJ congenic mice: effects of genie substitution at the aj^outi, brown, albino, dilute, and piuk-eyed dilution loci. J Hxp Zool 250:304-311, 1989 Ito S, Wakamatsu K: An improved modification of permanganate oxidation of eumelanin that gives a constant yield of pyrroIe-2,3,5-tricarboxylic acid. PiS>ment Cell Res 7:141-144, 1993 Kable EPW, Parsons PG: Melanin synthesis and the action of L-dopa and 3,4-dihydroxybenzylamine in human melanoma cells. Cancer Chemother Pharmacol 23:1-7, 1989 Orlow SJ, Osber MP, Paweiek JM: Synthesis and characterization of melanins from dihydroxyindole-2-carboxy!ic acid and dihydroxyindole. Pigmeut Cell Res 5:113-121, 1992 Johnson R, Jackson IJ: Light is a dominant mouse mutation resulting in premature cell death. Nature Geuet 1:226-229, 1992 Ito S, Jimbow K: Quantitative analysis of eumelanin and pheomelanin in hair and melanomas. J luvest Dermatol 80:268-272, 1983 Oikawa A, Nakayasu M: Quantitative measurement of melanin as tyrosine equivalents and as weight of purified melanin. Yale j Biol Med 46:500-507, 1973 Granholm NH, Opbroek AJ, Harvison GA, Kappenman KE: Tyrosinase acti\'it\' (TH, DO, PAGE-defined isozymes) and melanin production in regenerating /hairbulb melanocytes of lethal yellow {A^' /a), black {a/a), agouti {A"'^ /A"'-^) and albino (o/d/c^Vc^-') mice (C57BL/6J). Pigment Cell Res 3:233-242, 1990 Burchill SA, Thody AJ, Ito S: Melanocyte-stiniulating hormone, tyrosinase activity and the regulation of eumelanogenesis and phaeomelanogenesis in the hair folhcular melanocytes ofthe mouse. J Endocr 109:15-21, 1986 Ito S: Reexamination of the structure of eumelanin. Biochim Biophys Acta 883:155-161, 1986 Tsukamoto K, et al: 5,6-dihydroxyindole-2-carboxylic acid is incorporated in mammalian melanin. Biochem J 286:491-495, 1992 Napolitano A, Crescenzi O, Prota G: Copolymerisation of 5,6-dihydroxyindole and 5,6-dihydroxyindole-2-carhoxylic acid in melanogenesis: isolation of a cross-coupling product. Tetrahedron Lett 34:885-888, 1993 Aroca P, Solano F, Salinas C, Garcia-BorronJC, LozanoJA: Regulation of die final phase of mammalian melanogenesis. The role of dopachrome tautomerase and the ratio between DHICA and DHL EurJ Biochem 208:155-163, 1992 Pierro LJ, Chase HB: Slate—a new coat color mutant in tlie mouse. J Hercd 54:47_50, 1963 KroumpoLizos G, Urabe K, Kobayashi T, Sakai C, Hearing VJ: Functional analysis ofthe slaty gene product (TRP2) as dopachrume tautomerase and the effect of a point mutation on its catalytic function. Biochem Biophys Res Comttiun 202:1060-1068, 1994 Kuzumaki T, Matsuda A, Wakamatsu K, Ito S, Ishikawa K: Eumelanin hiosjiithesis is regulated by coordinate expression of tyrosinase and tyrosinase-related protein-1 genes. Exp Cell Res 207:33-40, 1993 del Mannol V, Ito S, Jackson I, Vaclitenheini J, Berr P, Ghanem G, Morandini R, Wakamatsu K, Huez G: TRP-1 expression correlates with eumelanogenesis in human pigment cells in culture. FEBS Lett 327:307-310, 1993 Orlow SJ, Boissy RE, Moran DJ, Pittco-Hirst S: Subcellular distribution of tyrosinase and tyrosinase-related protein-1: implications for melanosomal bioy^cncsis. J Invest Dermatol 100:55-64, 1993 Urabe K, Aroca P, Tsukamoto K, et at: The inherent cytotoxicity of melanin precursors: a revision. Biochim Biophys Acta 1221:272-278, 1994 Quevedo WCJr, Fleischmann RD, DyckmanJ: Premature loss of melanocytes from hair follicles of light (B'*) and silver (si) mice. In: Seiji M (ed.). Pigment Cell 198L Pheiiotypic Expressiou in Pigmeut Cells. University of Tokyo Press, Tokyo, 1981, pp 177-183 Quevedo WCJr, Holstein TJ, DyckmanJ, NordlundJJ: Influence ofdepigmenting chemical agents on hair and skin color in yellow (i)heomelanic) and black (eumelanic) mice. Pigment Cell Res 3:71-79, 1990 Engle LJ, Kennett I^H: Cloning, analysis, and chromosomal localization of myoxin {MYHi2), the human homologue to the mouse dilute gene. Genomics 19:4()7_416, 1994