266

Biochimica et Biophysica Acta, 1035(1990) 266-275

Elsevier BBAGEN 23367

Regulation of mammalian melanogenesis I" partial purification and characterization of a dopachrome converting factor: dopachrome tautomerase P i l a r A r o c a , J o s e C. G a r c i a - B o r r o n , F r a n c i s c o S o l a n o a n d J o s e A. L o z a n o Departamento Bioqulmica y Biologia Molecular, Facultad de Medicina, Universidad de Murcia, Murcia (Spain)

(Received 19 July 1989)

Key words: Mammalianmelanogenesis;Dopachromeconversionfactor; Dopachromeoxidoreductase;(Mousemelanoma) A protein that catalyzes the decoloration of dopachrome has been partially purified from BI6 mouse melanoma tumors. The enzyme is preferentially associated to the melanosomes, but it is also found in the microsomai and cytosolic fractions of cellular homogenates. The protein is dearly different from tyrosinase, and should be related to the dopachrome oxidoreductase (Barber et al. (1984) J. Invest. Dermatol. 83, 145-149) and the dopachrome conversion factor (Korner and Pawelek (1980) J. Invest. Dermatol. 75, 192-195) since the reaction product of dopachrome conversion is 5,6-dihydroxyindole-2-carboxylic acid. The protein appears to have an oligomeric structure, with a molecular mass slightly higher than 300 kDa estimated by gel filtration, whereas the molecular mass of the monomer might be approx. 46 kDa estimated by SDS-PAGE electrophoresis. Its K m for dopachrome is around 100 pM. The enzyme is competitively inhibited by indoles and is unaffected by metal chelators. It also has the ability to increase the amount of melanin formed from L-tyrosine by melanoma tyrosinase, and therefore, cannot be considered an 'indole blocking factor' as was suggested for the related dopachrome oxidoreductase. Since the reaction catalyzed by the enzyme is a tautomeric shift on dopachrome, we would propose dopachrome tautomerase (EC 5.3.2.3) as the most precise and informative name.

Introduction Melanin is the main pigment found in mammalian skin, hair and eyes. The biosynthesis of melanin proceeds through the hydroxylation of L-tyrosine to L-dopa and the subsequent oxidation of L-dopa to dopaquinone. Both reactions are catalyzed by tyrosinase (EC 1.14.18.1). In the absence of sulphydryl compounds, eumelanin is the final product of the pathway. The chemical structure of this polymer is still poorly understood, but it is generally agreed that it incorporates some of the highly reactive intermediates arising from dopaquinone, thus yielding somehow an irregular structure. The first part of the biosynthetic pathway, from

Abbreviations: L-dopa, 3,4-dihydroxyphenylalanine;dopachrome, 2carboxy-2,3-dihydroindole-5,6-quinone; DHI, 5,6-dihydroxyindole; DHICA, 5,6-dihydroxyindole-2-carboxylicacid; PMSF, phenylmethylsulfonylfluoride; SDS-PAGE, sodium dodecyl sulfate polyacrylamidegel electrophoresis; PBS phosphate-bufferedsaline. Correspondence: F. Solano, Departamento de Bioquimicay Biologia Molecular, Facultad de Medicina, Universidad de Murcia, 30100 Murcia, Spain.

L-tyrosine to dopachrome, is fairly well understood [14], but the second part of the pathway remains obscure. For a long time, it had been thought that tyrosinase was the only enzyme involved in the regulation of melanin biosynthesis. Such a view is partly explained by the common observation that melanin can be formed 'in vitro' by the action of highly purified tyrosinase on L-tyrosine. In fact, once dopaquinone has been formed, the subsequent reactions leading to the polymer proceed spontaneously. Moreover, aerobic oxidation of L-dopa in the absence of any protein also yields melanin. However, careful chemical analysis unequivocally proves that natural melanin is markedly different from melanin prepared 'in vitro' by the action of tyrosinase on L-dopa, since it incorporates a much higher content of carboxylated units derived from DHICA [5,6]. Thus, synthetic melanin might not be considered a good model for the natural polymer [7]. It is therefore reasonable to speculate the involvement of other regulatory factors acting in the pathway. So far, several of these factors have been invoked: (i) dopachrome conversion factor, that would accelerate the conversion of dopachrome into dihydroxyindoles [8,9]; (ii) a DHI conversion factor that accelerates the

0304-4165/90/$03.50 © 1990 Elsevier Science Publishers B.V. (BiomedicalDivision)

267 formation of melanins from dopachrome, but is active only in cells treated with MSH [8]; (iii) a DHI blocking factor that would inhibit melanin formation [8]; (iv) an enzyme called dopachrome oxidoreductase that was first reported in melanotic and amelanotic B16 melanoma; it was originally suggested that the enzyme might catalyze the conversion of dopachrome into DHI [10] and also act as an indole blocking factor. Subsequently, DHICA was more convincingly identified as the product of the action of either dopachrome conversion factor [11] or dopachrome oxidoreductase [12] on dopachrome and finally (v) metal cations such as Zn(II), Ni(II) and Cu(II), that accelerate the conversion of dopachrome to melanin pigments containing a percentage of carboxylated units similar to that found in natural pigment [6,13]. Thus, the current view of the regulation of melanin biosynthesis is more complex than originally thought, considering the following points: (i) So far, the postulated 'distal factors' have not been adequately isolated and characterized. In fact, some of them might be related to tyrosinase isozymes [14] and DHI conversion activity has been attributed to tyrosinase [15]; (ii) the action of dopachrome oxidoreductase on dopachrome seems to yield the same product than metal ions yield, and the characterization of the enzymatic activity is incomplete. Further experiments are needed to rule out the possibility that some of the reported effects of protein factors might be related to metal-ion contamination in the preparations [9]; indeed, it has been suggested that the structure of natural melanin might be derived from the action of metal ions without the involvement of any other protein factor different of tyrosinase [13] and (iii) the link, if any, between accelerating and blocking factors, as suggested for dopachrome oxidoreductase [10] is, at least, doubtful. We report here the partial purification of a dopachrome converting factor. It is clearly shown that the factor is an enzyme different from tyrosinase. A purified preparation of the enzyme, free of tyrosinase activity, was used to establish the nature of the reaction product and to obtain several kinetic parameters. It is also shown that the enzyme is not a blocking factor, but on the contrary, it accelerates the formation of 'melanin' from dopachrome. The relationship of this enzyme with other 'distal' regulatory factors or enzymes reported in the literature is discussed. Materials and Methods

Reagents Radioactive substrates, L-[U-14C]tyrosine (specific radioactivity 502.1 mCi/mmol) and OL-[1-14C]dopa (spec. radioact. 57.5 mCi/mmol) were obtained from New England Nuclear (Boston, U.S.A.). CO2 absorber was purchased from Packard (Zurich, Switzerland). L-

Tyrosine, L-tryptophan, L-dopa, BSA, trypsin, fl-mercaptoethanol, Coomassie brilliant blue, Amido black, PMSF, EDTA diaminobenzidine, phenylthiourea, hydroxyapatite type I suspension in 1 mM phosphate buffer (pH 6.8), Brij 35, SDS and antirabbit peroxidaseconjugated antibody were from Sigma (St. Louis, MO, U.S.A.). Silver oxide, sodium monobasic and dibasic phosphates, sodium hydroxide, trichloroacetic acid, sucrose, glycine, Tris and diethyldithiocarbamate were from Merck (Darmstadt, F.R.G.). Celite 535 was from Koch-Light (Berkshire, U.K.) Toluene, ethanol, methanol, isopropanol, acetone, acetic acid, glycerol, H202, HC1, NaIO4, (NH4)2SO4, KC1, MgC12 and activated charcoal were from Probus (Spain). Acrylamide, Bis, Temed, Tween 20, sodium persulfate, nitrocellulose sheets and Chelex 100 were from Bio-Rad (Richmond, CA, U.S.A.). DEAE-Cellulose and 3 MM paper disks (2.3 cm) were from Whatman (Kent, U.K.). Ultrogel AcA34 from LKB (Sweden). Protein A/Sepharose C1-4B and standard proteins for SDS-PAGE molecular weight calibration were from Pharmacia (Uppsala, Sweden). All solutions were prepared using double-distilled water passed through a Milli-Q system. The resistivity of such water was higher than 10 MI2. cm. All reagents were of the highest purity commercially available (except ethanol and acetone for washing of filter-paper disks) and were used without further purification.

Animals and melanomas B16 mouse melanoma melanocytes were originally a kind gift from Dr. V. Hearing (NIH, Bethesda, U.S.A.). They have been maintained by serial transplantation on hybrid mice obtained from male DBA and female C57/B1 (Panlab, Spain). Only male mice at 6-8 weeks of age were used for tumor transplantation, and they were injected subcutaneously with approx. 105 viable cells. After 3-4 weeks, visible tumors were excised, and then some were used for new implantation and the remainder for enzymatic preparations.

Purification of dopachrome converting factor and subcellular fractionation All operations were carried out at 0-4°C. Freshly excised tumors were cleaned of fat and connective tissue, weighed and washed twice in ice-cold 10 mM phosphate buffer (pH 6.8) containing 0.25 M sucrose and 0.1 mM EDTA. After centrifugation, the washed tumors were homogenized in a Polytron homogenizer (power setting at 7) in the same buffer as before supplemented with 0.1 mM PMSF. The volume of buffer (ml) was three times the weight of tumors to be processed (g). The homogenate was centrifuged at 700 × g, for 20 min, and the pellet was reextracted with one volume of homogenization buffer and centrifuged in the same conditions. The combined supernatants were centrifuged at 11000 × g for 30 min in a Sorvall SS-34 rotor.

268 The pellet was considered as crude melanosomal fraction. The supernatant was further fractionated by centrifugation at 105 000 × g for 60 min in a Beckman 60 Ti rotor to obtain the cytosolic and microsomal fractions. To solubilize melanosomal proteins, the melanosomal pellet was resuspended in 10 mM phosphate buffer (pH 6.8) containing 1% Brij 35, using a volume (ml) twice the original weight of tumors (g). The suspension was incubated for 30 rain at 4°C with gentle stirring, and centrifuged at 105 000 × g 1 h. The pellet was reextracted under identical conditions, except that half the original volume of solubilization buffer was added. The combined supernatants were the crude melanosomal extracts. This extract was then brought to 35% saturation with ammonium sulfate by dropwise addition of the appropriate volume of a saturated solution of that salt adjusted at pH 7.0 with NaOH, with continuous stirring. Precipitation was allowed to proceed at 4 ° C for 2-3 h and the sample was centrifuged at 11 000 × g for 30 min. The pellet did not contain any measurable activity and was discarded. The supernatant was brought to 60% saturation and left overnight. After centrifugation at 11000 × g, for 30 min, the pellet was resuspended in a volume of 0.1% Brij 35 in 10 mM phosphate buffer (pH 6.8) equal to half the original weight of tumors, and centrifuged as above. The clear supernatant was extensively dialyzed against the same buffer. Then, the sample was mixed with hydroxyapatite (Sigma, Type 1, 1 volume of sample/1 volume of settled suspension), and incubated for 30 min with gentle stirring in an orbital shaker. The slurry was centrifuged at 3000 × g for 10 rain in a Sorvall SS-34 rotor and the hydroxyapatite pellet was washed by resuspension in one volume of 10 mM phosphate (pH 6.8) containing 0.1% Brij 35 and subsequent centrifugation at the same conditions. The combined supernatants were concentrated in an Amicon ultrafiltration cell and applied to a molecular weight calibrated Ultrogel AcA 34 column (52 × 2.6 cm) for gel filtration chromatography. The column was equilibrated and eluted with 50 mM phosphate buffer (pH 6.8) containing 0.1% Brij 35. Elution was performed at a constant flow rate of 12 ml/h. Fractions of 200 drops (approx. 6.3 ml) were tested for protein, dopachrome converting and dopa oxidase activities.

Determination of dopachrome converting activity This activity was determined spectrophotometrically by monitoring the decrease in absorbance at 475 nm [10] due to the conversion of dopachrome into dihydroxyindoles. The assay medium was 10 mM phosphate buffer (pH 6.0). Dopachrome was chemically prepared 'in situ' by L-dopa oxidation, because of its instability. Blanks were made without enzyme and substracted to obtain net activity. Further details on dopachrome preparation and the rate of the spontaneous dopachrome

decomposition have been recently described [16]. Radioactive dopachrome was prepared by mixing DL[114C]dopa (5 /tCi/ml, isotopically diluted with L-dopa) dissolved in 10 mM sodium phosphate buffer (pH 6.0) with periodate in a 1:2 stoichiometry (dopa/NalO4) [16,17]. One unit of this activity was defined as the amount of enzyme that catalyzes the tautomerization of 1/~mol dopachrome/min at 30 ° C.

Destination of the carboxylic group of [1-14C]dopa during melanogenesis Decarboxylation during melanogenesis was determined by measuring the lacoz evolved from freshly prepared dopachrome. The reactions were carried out at 30 °C in sealed tubes with rubber stoppers hanging two Whatman 3MM paper discs (1/4 inch diameter) containing 20 #1 of CO 2 absorber. The total volume of the standard assay was 100-200 /~1 containing 0.5 mM L-dopa and 0.125 /~Ci of DL-[1-14C]dopa or 0.25 mM dopachrome and the same radioactivity of [114C]dopachrome. Then, paper discs were removed and counted for radioactivity. The reaction mixtures remaining in the assay tubes were transferred to Whatman 3Mm filter paper discs (2.3 cm diameter) and washed as described for the melanin formation assay to measure the amount of radioactive carboxylic groups incorporated to the melanin.

Tyrosinase preparation Melanoma tyrosinase was prepared and purified from the melanosomal pellet by trypsin treatment, according to Jara et al. [18]. Briefly, melanosomes were resuspended in 10 mM phosphate buffer (pH 6.8) containing 0.5 mg/ml of trypsin. After 2 h at 37 ° C, PMSF was added up to 1 mM to stop solubilization and the sample was centrifuged at 105 000 × g for 1 h. The supernatant was concentrated by ultrafiltration and submitted to ion-exchange and gel filtration chromatographies for purification.

Tyrosinase activity determinations Dopa oxidase assay. This activity of tyrosinase was spectrophotometrically determined by recording the increase in the absorbance at 475 nm due to dopachrome formation (c = 3700 M -1 .cm -1) in a 37°C thermostated Varian Vis-UV spectrophotometer. One unit of tyrosinase was defined as the amount that catalyzes the oxidation of 1 /~mol L-dopa/min. Further details of all described methods can be found elsewhere [18]. Melanin formation assay. Incubations were carried out as described by Jara et al. [18]. Briefly, each assay (final volume 50 /zl) contained 10 /xl of 0.25 mM Ltyrosine (25/~Ci/ml of L-[U-a4C]tyrosine) and 0.05 mM L-dopa in 0.1 M phosphate buffer (pH 6.8), 10 /~1 of purified melanosomal tyrosinase preparation and the volume was finally brought to 50 /~1 with samples of

269 purified dopachrome converting factor a n d / o r buffer. The assays were routinely run at 37 ° C for 1 h in wells of a microtiter plate. After incubation, the samples were removed to Whatman 3MM filter-paper discs (2.3 cm diameter). The filters were washed three times with 0.1 M HCI, once with ethanol and once again with acetone, then allowed to dry, and counted for radioactivity. Blanks without tyrosinase were made and subtracted from sample counts to obtain net cpm.

Preparation of a polyclonal antityrosinase antiserum Two New Zealand rabbits approx. 3 kg weight were used. They were immunized with 200 gg of purified tyrosinase by subcutaneous injection of an emulsion obtained by mixing 1 vol. of a 0.25 m g / m l solution of the antigen in phosphate-buffered saline (PBS) and one volume of Freund's complete adjuvant. Rabbits were boostered twice with the same amount of tyrosinase, but using incomplete adjuvant, 30 and 90 days after the first dose. The animals were bled 4 weeks after the last booster injection and the antibody titer was 1 : 3 2 as determined by Ouchterlony double-diffusion analysis in 1% agarose containing 0.1% Brij 35. A partially purified IgG fraction was obtained by ammonium sulfate precipitation and DEAE-cellulose chromatography [19].

Immunoprecipitation Aliquots (5 ~tl each) of crude extracts containing approx. 1.1 munits of dopa-oxidase activity were mixed with increasing amounts of partially purified antityrosinase antiserum (10 m g / m l total protein concentration) in PBS, and the medium was adjusted to a final volume of 70 gl with PBS containing 0.1% BSA and 0.1% Brij 35. Samples were incubated for 1 h at room temperature and then overnight at 4 o C. Then, 30 gl of a 1 : 1 slurry of previously washed Protein A/Sepharose in PBS were added and the samples further incubated for 1 h with an orbital shaker and centrifuged at 10000 × g 15 rain. Parallel control experiments were performed under identical conditions,

except that non-immune purified rabbit IgG were used instead of the specific antiserum. The supernatants were recovered and tyrosinase activity was measured for each sample.

Polyacrylamide gel electrophoresis and Western blot 12% polyacrylamide slab gels, 0.75 mm thick, were prepared according to Laemmli [20] and run in a Miniprotean II cell (Bio-Rad) at a constant current of 20 mA. Generally, samples were pretreated for 30 min with 3% SDS and 7.5% fl-mercaptoethanol. Gels were either stained with 0.1% Coomassie brilliant blue (dissolved in 6 5 : 2 5 : 1 0 w a t e r / i s o p r o p a n o l / a c e t i c acid) and destained in methanol/acetic acid solutions or electrophoretically transferred to nitrocellulose filters. Electrotransfer was performed in a refrigerated TransBlot cell (Bio-Rad) at a constant voltage of 70 V, for 3 h, using 25 mM Tris (pH 8.3), 192 mM glycine containing 20% methanol as transfer buffer. After that, the nitrocellulose filters were blocked by incubation for 1 h at room temperature in PBS containing 2% BSA, and washed three times with PBS containing 0.03% BSA and 0.05% Tween 20. The filters were then incubated overnight at 4 ° C in a 0.1 m g / m l solution of the purified antiserum in PBS containing 0.03% BSA and 0.01% Tween 20 and washed four times with the same buffer. Immunoreactive bands were visualized with an antirabbit peroxidase conjugated antibody, using 0.5 m g / m l diaminobenzidine and 0.003% hydrogen peroxide as substrates.

Protein determination The protein content was determined by a modified Lowry method [21] using BSA as standard. Results

Purification of dopachrome converting activity The course of a standard purification of the dopachrome converting activity is outlined in Table I, using the

TABLE I

Purification of dopachrome convertingfactor from B16 mouse melanoma crude melanosomes 45 g of B16 tumors were homogenized and processed as described in Materials and Methods. The crude melanosomal fraction was solubilized with 1% Brij 35 and the supernatant was used as starting material.

Step Solubilized extract (NH4)2SO4 fractionation Hydroxyapatite batch

chromatography Gel-filtration chromatography

Protein (mg)

Activity (munit)

Sp. act. (munit/mg)

Yield

Pur. factor

181

2205

12.2

100

1

65.5

1795

27.4

81

2.2

13.5

1011

74.9

46

6.1

1.4

595

425.0

27

34.9

270 crude melanosomal pellet obtained by differential centrifugation as the starting material. The melanosomes were solubilized with 1% Brij 35 and the extract was fractionated by ammonium sulfate precipitation. The activity remained soluble at 35% saturation but was precipitated almost quantitatively when the saturation was increased up to 60%. The dialyzed sample was then submitted to batch chromatography on hydroxyapatite. About half of the original activity did not bind to the gel and was recovered after centrifugation of the slurry and washed with 10 mM phosphate buffer (pH 6.8). However, the rest of the activity remained tightly bound to the hydroxyapatite and the phosphate concentration had to be increased to 150 mM to achieve its recovery. The meaning of this fact is unclear but it might suggest the existence of two isozymes within the melanosomes. Further purification of the unbound activity could be achieved by gel-permeation chromatography on Ultrogel AcA 34. The dopachrome converting activity eluted before the tyrosinase peak (Fig. 2) with a partition coefficient corresponding to an apparent molecular mass of approx. 320 kDa. The pooled fractions were free of any measurable dopa-oxidase activity, but contained several polypeptide bands as shown by SDS-PAGE (Fig. 1). The major band in this preparation had an apparent molecular mass of 46 kDa and might correspond to the monomeric form of the enzyme, since its relative abundance in the fractions of the gel-filtration column is consistent with the elution pattern of the activity (Fig. 2 inset). The broad peak at the end of the activity elution profile is also consistent with the occurrence of that monomeric form. Furthermore, we also

1

2

3

4

92.5 Kd_ 66.2

45

31

21.5

__

Fig. 1. SDS-PAGE profiles of dopachrome converting factor preparations at different stages of purification. 12% gels were stained with Coomasie blue. Lane 1, crude solubilized melanosomal fraction (25 #g total protein). Lane 2, (NH4)zSO 4 fractionated sample (25 #g). Lane 3, final stage of purification after gel filtration (10 #g). Lane 4, protein bound to hydroxyapatite and eluted at high ionic strength (25 #g).

tried a specific 'negative' staining of the band corresponding to dopachrome converting factor using dopachrome as substrate in SDS-PAGE gels, but unfortunately the activity was extremely sensitive to the presence of SDS (data not shown). Therefore, a definitive assigment of the molecular weight for the monomer and the oligomeric structure of the enzyme awaits complete purification of the protein. From their different elution profiles throughout the purification process, it could be anticipated that the dopachrome converting enzyme and tyrosinase are two different proteins. However, other experiments were performed to definitively establish this point. (i) Purified melanoma tyrosinase was unable to catalyze dopachrome conversion. (ii) The thermal stability of the two enzymes was first. Tyrosinase was found to be considerably more thermostable than dopachrome converting factor. A solubilized melanosomal extract was immersed in a thermostatic water bath whose temperature increased at a constant rate of 1° C/min. Aliquots were removed at selected times, immediately cooled by addition of an equal volume of cold buffer, and assayed for enzymatic activities. Dopachrome converting factor lost 50% of activity at 53°C, and inactivation was complete at 65 ° C, whereas tyrosinase needed a temperature of 75°C to decrease its activity down to 54% respect to the initial control. (iii) Dopachrome converting factor was unaffected by incubation with tyrosinase inhibitors, such as phenylthiourea and sodium azide (see below). (iv) Treatment for 30 min of purified dopachrome converting factor with 1 mg/ml of proteinases (trypsin, chymotrypsin and proteinase K) at 30 °C and pH 6.8, resulted in inhibitions of 55, 75 and 100% of the original activity respectively, whereas the resistance of tyrosinase activity to these proteases [4,22] is well known. (v) More conclusively, none of the polypeptide chains present in the purified preparation of the dopachrome conversion factor cross-reacted with a polyclonal serum raised against purified tyrosinase. The specificity of this antityrosinase antiserum was checked by Western-blot and immunodepletion experiments. In Western-blot experiments the antiserum recognized preferentially the 67 kDa band associated to tyrosinase (Fig. 3). This band displays a mobility identical to that of purified tyrosinase (not shown). Moreover, the antiserum immunoprecipitated in a concentration-dependent manner the tyrosinase activity of crude extracts, as shown by immunodepletion experiments where 10/tl of the immune IgG-fraction could precipitate 90% of the tyrosinase activity originally present in the reaction mixture. The subcellular distribution of the dopachrome converting activity is shown in Table II. It can be seen that although the enzyme is preferentially associated to melanosomes, some activity was also found in the microsomal and cytosolic fractions. A very similar distil-

271

A 92.5 66.2

B

C D

Kd _ __

45

31 21.5 __

ABSORBANCE 280

ACTIVITY mUNITS/FRACTION

C I •

90-

0.2

/ s/

~

0 1 /~

~

f/l

10

20

60 --

o

30

40

Fraction number Fig. 2. Gel-filtration chromatography of a partially purified preparation of dopachrome converting factor. After hydroxyapatite chromatography, the solution was concentrated to 4 ml by ultrafiltration and submitted to the column as described under Materials and Methods. Fractions were analyzed for absorbance at 280 nm (o) and dopachrome converting (O) and tyrosinase activities (11). Inset: SDS-PAGE profiles of selected fraction. 0.5 ml of the fractions indicated by an arrow were brought to 60% saturation with (NH4)2SO4, incubated for 2 h at 4 ° C and centrifuged at 13000 × g in an Eppendorf centrifuge. The pellets were resuspended in sample buffer (60 mM Tris-HC1 (pH 6.8) containing 5% glycerol, 3% SDS, 7.5% fl-mercaptoethanol and 0.0025 Bromphenol blue) submitted to electrophoresis and stained for protein.

bution was obtained by Barber et al. [10] for dopachrome oxidoreductase.

Identification of the reaction product DHICA appeared to be the product of the action of dopachrome converting factor on dopachrome, as judged by both spectrophotometric and radiometric criteria. Fig. 4A shows the time evolution of the UV-Vis absorption spectrum of a reaction mixture containing 100/~M dopachrome and 1.6 munit of dopachrome converting factor (pH 6.0). It can be seen that the decay of the visible peak at 475 nm was accompanied by a parallel increase in the absorption in the UV region, together with a slight shift in the absorption maximum, from 305

to 308 nm. This indicated an aromatization of the indoline ring in dopachrome with the conservation of the carboxyl group in the presence of dopachrome converting factor, that would account for the observed auxochromic effect. On the other hand, the time evolution of the spectra was different when 100/~M dopachrome was monitored under the same conditions but in the absence of dopachrome converting factor. The reaction was slower, both the visible and the UV peaks decreased, with a slight shift in the UV maximum but in the opposite direction, from 305 to 302 nm (Fig. 4b). Both the decrease in the absorption coefficient and the observed shift are in agreement with a decarboxylation of the initial compound [16].

272

1.5

92.5 Kd

0.5

'7 '!t,

66.2

;¢

b

45

1.5

31

¢

+\

0.5

3~)0

21.5

s~o

I 4O0 Wavelength (nm)

Fig. 3. Western transfer and immunological staining of dopachrome converting preparations with anti-tyrosinase antiserum. Lane 1: an aliquot of a crude melanosomal extract containing 25 #g total protein was electroplioresed, electrotransferred to nitrocellulose filters and stained with anti-tyrosinase antibodies. Lane 2:0.5 ml of the gelfiltration chromatography fraction containing the highest dopachrome converting activity were treated with (NH4)2SO4 as described in Fig. 2, and treated as above. The efficiency of transfer was higher than 90% for all proteins in the extract as judged by protein staining of the transferred gels. The molecular mass calibration was performed by electrophoresis and transfer of a mixture of standard proteins (from 92.5 to 14.4 kDa and staining of the nitrocellulose filters with 0.1% Amido black in 20% ethanol/7% acetic acid.

The time-course of decarboxylation of radiolabeled dopachrome in the absence or in the presence of different amounts of the converting factor is shown in fig. 5. The following points should be noted: (i) spontaneous dopachrome decarboxylation did not proceed to completion, with around 15-20% of the original radioactivity remaining in the assay media; (ii) the kinetics of this

Fig. 4. Evolution of the UV-Vis spectra of dopachrome in the presence (a) or the absence (b) of dopachrome converting factor. (a) The reaction mixture consisted of 100/tM dopachrome and 1.6 munit of purified enzyme in 10 mM phosphate buffer (pH 6.0); the reaction was allowed to proceed at 30 ° C for 1(1), 5(2), 15(3) and 30(4) rnin. (b) Spontaneous evolution of a dopachrome solution under identical conditions as above but without enzyme. The spectra were recorded at time 0 (1), 15 (2) and 30 (3) min,

spontaneous process are fully consistent with the one reported by Palumbo et al. [13] in the absence of metal ions; and (iii) the decarboxylation is inhibited by the addition of the converting factor, in a concentration-dependent manner. After a first phase of CO 2 release the

//-.

%C02 Evolved

TABLE II

Subcellular distribution of dopachrome converting activity Melanosomal and microsomal fractions were obtained from 45 g of tumors and solubilized using identical conditions, as described under Materials and Methods for the melanosomal pellet. Fraction

Activity (munit)

Spec. act. (munit/mg)

% Activity

Melanosomal Microsomal Cytosolic

2205 989 1081

12.2 6.7 0.65

52 23 25

Time (h) Fig. 5. Percentage of total radioactivity evolved as ]4CO 2 from ll14Cidopachrome in the absence (I) or the presence of 32,1 (o), 972 (*) and 1620 (zx)/xunits of dopachrome converting factor.

273

1/v

3

mUnits "~

I

I

25

50 .I

~s mM Fig. 6. Double-reciprocal plot of the dopachrome converting activity in the absence (e) or the presence (o) of 1.5 mM tryptophan. K m for dopachrome was around 0.1 mM and according to the competitive inhibition, K i for tryptophan was approx. 1.8 mM.

curves become stabilized, and the percentages of decarboxylation decreased as the amount of enzyme in the reaction media increased. These spectrophotometric and radiometric observations show that the reaction product obtained by the action of dopachrome converting factor on dopachrome is DHICA. Moreover, the same conclusion using HPLC analysis has been recently reported for dopachrome oxidoreductase [12].

Kinetic parameters Several kinetic characteristics of the enzyme were studied. In experiments carried out at different substrate concentrations, the enzyme displayed a saturation behaviour, with a K m around 100 vM. Some commercially available hydroxyindoles were tested as inhibitors of the reaction, but the results were difficult to interpret due to the direct reactivity of most of them with the substrate dopachrome. This fact led to high rates of dopachrome dissappearance even in the absence of the factor. Tryptophan did not react with dopachrome and acted as a competitive inhibitor of dopachrome conversion, with a K i around 2 mM (Fig. 6). This observation suggested that the oxygenated functions at positions 5 and 6 of the indole ring are not essential for binding to the enzyme, although the affinity decreased. On the other hand, the ability of a series of metal chelators to inhibit the enzymatic activity was also checked. Preincubation of the enzyme with concentrations as high as 5 mM of either EDTA, sodium azide and phenylthiourea failed to result in any noticeable inhibition. Thus, dopachrome converting factor does not seem to be a metal protein, as opposed to tyrosinase.

the similarities above discussed between our factor and that enzyme, we studied the effect of dopachrome convetting factor on melanin synthesis. In fact, if melanin formation from dopachrome was visually determined, according to the color grade described by Barber et al. [10], our dopachrome converting factor was also able to prevent the appearance of black color that occurred in the blank sample. Instead of that, dopachrome converting factor rapidly led from the red color of dopachrome to a colorless solution stable for a few hours, and finally to a yellowish solution. Thus, our factor also seemed to prevent melanin formation. However, in order to improve this visual test, we used two different approaches. Firstly, when a sample of this reaction mixture was submitted to chromatography on G-25 Sephadex, the yellowish compound was excluded from the gel (data not shown), indicating a molecular size difficult to establish with these data but, at least, higher than the monomer product DHICA, which was retained. The yellowish compound did not display any definite absorption peak in the UV-Vis region, but a generalized absorbance, and therefore, does not appear to arise from the conjugation of DHICA, or DHICA oxidation products, with proteins present in the partially purified preparation. All the available data are consistent with its being an oligomeric, melanin-like product derived from the partial polymerization of dihydroxyindoles. Secondly, we performed a series of experiments to estimate the melanin formation activity of tyrosinase, acting on L-tyrosine, in the presence of variable amounts of purified dopachrome converting factor completely free of tyrosinase. We added different amounts of a purified preparation of this factor to assay mixtures of L-[U-14C]Tyr and melanoma tyrosinase. Fig. 7 shows that the amount of radiolabeled melanin formed in-

cpm. 10.3

I

I IoO

Effect of dopachrome converting factor on melanin formation An indole blocking activity for dopachrome oxidoreductase has been proposed [10]. Taking into account

I

I ~t3o

Enzvme ( ~ u n i t s )

Fig. 7. Radioactive melanin formed from L-[U-14C]Tyr by melanosomal tyrosinase (0.64 munit) in the presence of several amounts of purified dopachrome converting activity.

274 creased as the amount of dopachrome converting factor also was increased. Thus, dopachrome converting factor, using this assay, behaved as a melanin acceleration factor instead of an indole blocking factor.

Discussion The results presented above clearly show that mouse melanoma melanosomes contain an enzymatic activity involved in melanin biosynthesis, different from tyrosinase, and not related to metal cations. The nature of the reaction product, and the similarities in heat and trypsin sensitivities, as well as subcellular distribution, seem to indicate that the enzyme might be related or identical to dopachrome oxidoreductase. However, some characteristics of the enzyme are different from those reported by Barber et al. [10] for dopachrome oxidoreductase. Our preliminary data are consistent with a higher molecular mass and the occurrence of an oligomeric form of the enzyme. A possible explanation could be related to the high sensitivity of dopachrome oxidoreductase to proteolysis [10,12]. The occurrence of proteinases in melanoma tumors [23] and their ability to act on melanosomal proteins [24] have been reported. It is conceivable that the proteolytic cleavage of some fragment of the enzyme might explain the lower molecular weight previously reported for dopachrome oxidoreductase, in comparison to our converting factor that was obtained in the presence of PMSF during the homogenization. A definitive answer must await further purification and characterization of these proteins. When the action of tyrosinase alone and mixtures of tyrosinase and dopachrome converting factor on tyrosine was tested, it was found that the total amount of melanin produced from L-tyrosine increased as the tautomerase activity in the reaction mixture also increased. However, no black precipitate appeared in the reaction mixtures from dopachrome, even after prolonged incubation times (as high as 48 h). The solutions remained yellowish in the presence of the enzyme. The coloured compound formed might be considered 'melanin' since it is an oligomer of dihydroxyindoles as judged by gel-filtration chromatography; it is insoluble in acid media; it has a generalized absorption in the UV-Vis and it is derived from L-tyrosine or dopachrome. In turn, these results are consistent with the nature of the reaction product derived from the action of the enzyme on dopachrome, since Ito [5] reported that DHICA-melanin remained mostly in solution, while DHI-melanin rapidly precipitated out during its formation. Further characterization of this type of DHICAmelanin is underway. Therefore, dopachrome converting factor should not be considered as an indole blocking factor, as might be thought based on visual criteria. The formation of such soluble yellowish compound points out that more stri-

gent criteria than visual perception should be used when looking for 'indole blocking' activity. It should be concluded that, so far, no compelling evidence in favour of the existence of indole blocking factors has been presented. However, the existence of at least one enzyme different from tyrosinase acting on melanin biosynthesis intermediates seems clearly established. The spectrophotometric and radiometric data herein presented, as well as the N M R and MS spectral data [11] and HPLC analysis [12] reported by others, demonstrate unequivocally that the product of the action of such enzyme on dopachrome is D H I C A instead of the DHI formed during the spontaneous dopachrome decomposition at neutral pH [16,25]. Therefore, we think that the name dopachrome converting factor is ambiguous and it should be replaced. The ideal name should yield the maximal information on the now well-established chemical nature of the reaction catalyzed. This name was discussed during the 2nd Meeting of the Panamerican Society for Pigment Cell Research (Bethesda, April 1989), where we proposed the name dopachrome tautomerase. Two other possible names, dopachrome oxidoreductase and dopachrome isomerase, were considered. As it is well known, all enzymes should be described by a name and a set of four figures that specify the chemical nature of the reaction catalyzed. The name dopachrome isomerase [26] corresponds to an enzyme of class 5, and only specifies the first figure. The name dopachrome oxidoreductase originally proposed by Barber et al. [10] is more complete, since it defines the class and the sub-class (isomerases, intramolecular oxidoreductases), thus setting two figures, 5.3 [27]. On the other hand, the name dopachrome tautomerase specifies the particular kind of intramolecular redox reaction accounting for the conversion of dopachrome into DHICA, a keto-enolic tautomerization, and therefore specifies the sub-sub-class, 5.3.2. According to Ref. 27, the complete set of four figures identifying the enzyme should be EC 5.3.2.3, and we think that this name is the most precise and should be preferred to the others.

Acknowledgements This work has been partially supported by grant no. PB87-0698 from the Direcci6n General de Investigaci6n Cientifica y Trcnica, Spain. P. Aroca thanks the 'Comunidad Autonoma de Murcia' for a fellowship.

References 1 Mason, H.S. (1948) J. Biol. Chem. 172, 83-90. 2 Garcia-Carmona, F., Garcia-Chnovas, F. and Lozano, J.A. (1982) Biochim. Biophys. Acta 717, 124-131.

275 3 Cabanes, J., Garcia-Chnovas, F., Lozano, J.A. and GarciaCarmona, F. (1987) Biochim. Biophys. Acta 923, 187-195. 4 Heating, V.J. and Jimenez, M. (1987) Int. J. Biochem. 19, 11411147. 5 Ito, S. (1986) Biochim. Biophys. Acta 883, 155-161. 6 Palumbo, A., D'Ischia, M., Misuraca, G., Prota, G. and Schultz, T.M. (1988) Biochim. Biophys. Acta 964, 193-199. 7 Prota, G. (1988) Med. Res. Rev. 8, 525-556. 8 Pawelek, J., Korner, A., Bergstrom, A. and Bologna, J. (1980) Nature 286, 617-619. 9 Korner, A. and Pawelek, J. (1980) J. Invest. Dermatol. 75, 192-195. 10 Barber, J.l., Townsend, D., Olds, D. and King, R.A. (1984) J. Invest. Dermatol. 83, 145-149. 11 Korner, A. and Gettins, P. (1985) J. Invest. Dermatol. 85,229-231. 12 Leonard, L.J., Townsend, D. and Kind, R.A. (1988) Biochemistry 27, 6156-6159. 13 Palumbo, A., D'Ischia, M., Misuraca, G. and Prota, G. (1987) Biochim. Biophys. Acta 925, 207-209. 14 Hearing, V.J., Korner, A.M. and Pawelek, J. (1982) J. Invest. Dermatol. 79, 16-18. 15 Korner, A. and Pawelek, J. (1982) Science 217, 1163-1165. 16 Aroca, P., Solano, F., Garcia-Borr6n, J.C. and Lozano, J.A. (1990) J. Biochem. Biophys. Methods, in press.

17 Graham, D.G. and Jeffs, P.W. (1977) J. Biol. Chem. 252, 57295734. 18 Jara, J.R., Solano, F. and Lozano, J.A. (1988) Pigment Cell Res. 1, 332-339. 19 Campbell, D.H., Gerney, F.S., Gremer, N.E. and Sussdorf, D.H. Eds. (1970) in Methods in Immunology, 2nd Edn. pp, 189-198, W.A. Benjamin, Reading, MA, U.S.A. 20 Laemmli, U.K. (1970) Nature 227, 680-682. 21 Hartree, E. (1972) Anal. Biochem. 48, 422-427. 22 Yurkow, E.J. and Laskin, J.D. (1989) Arch. Biochem. Biophys. 275, 122-129. 23 Sloane, B.F., Dunn, J.R. and Honn, K.V. (1981) Science 212, 1151-1153. 24 Martinez, J.H., Solano, F. and Lozano, J.A. (1989) Cell. Biochem. Funct. 7, 21-26. 25 Wakamatsu, K. and Ito, S. (1988) Anal. Biochem. 170, 335-340. 26 Pawelek, J.M. (1990) Biochem. Biophys. Res. Commun. 166, 1328-1333. 27 Enzyme Nomenclature. Int. Union Biochem. (1984) Academic Press, London.