The Relationship Between Tyrosinase Activity and Skin Color in Human Foreskins

Mitsuru Iwata, M.D., Todd Corn, B.S., Shoko Iwata, D.D.S., Mark A. Everett, M.D., and Bryan B. Fuller, Ph.D.

Dep:mmenr of Btochemtstry and Molecular Biology and the Department of Dermatology, The Umver~•ry of Okl.thoma He~hh Sctences Cenrer, Oklahoma City, Oklahoma, U.S.A.

Tyrosinase activiry was assayed in b lack and white human foreskin samples by measuring both the h ydroxylation of tyrosine to dopa {tyrosine hydroxylase activity) and the con­version of p•C)tyrosine to p•C] m elanin {melanin synthesis assay). Enzyme activity was found both in the particulate (75o/o) and soluble {25o/o) fractions of the cell. M embrane­bound tyrosinase was readily solubilized by either zwitter­ionic or nonionic detergents . The anionic detergent, sodium cholate, inhibited enzyme activity. Tyrosinase activity in black foreskin homogenates averaged almost three times that

in white skin samples {33.8 pmols 3H 20/h/mg skin in black and 12.71 pmoles 3H 20/h/mg skin in white skin), although considerable overlap in activities existed among the two groups. Tyrosinase activities measured with two separate

Tyrosinase (0-diphenol,-02-oxidoreductase, E.C. 1.14.18.1 ), the rate-liml[ing enzyme for melanin syn­thesis, is found in all mammalian melanocytes includ­ing man [ 1,2]. Studies with mouse melanoma cells in culture [3-7] and with mouse skin [8-10] have

shown that the enzyme can be induced by melanocyte stimulating hormone (MSH ) or by compounds which elevate intracellular cy­clic AMP levels, including dibutyryl cAMP and theophylline. Fewer studies have examined the abundance and regulation of tyro­sinase in human melanocytes. Tyrosinase activity has been mea­sured in human melanoma cells in culture and the aCtivity of the enzyme has been shown to increase in some cell lines by treatment with theophylline [ 11]. Human tyrosinase has been purified from melanomas and has been shown to resemble the mouse form in stze, electrophoretic mobility, tsoelectric pH, and the extent of glycosv­lation [12]. The eDNA sequences of mouse and human tyrosinase demonstrate an 80% sequence identity between the two enzymes

Manuscript recetved June 14, 1989, accepted for publication March 8. 1990.

Repnnt requests ro: Bryan B. Fuller, Department of Btochemtstry and Molecular Biology. Univemty of Oklahoma Health Sctences Center, P.O. Box 26901, Oklahoma City. OK 73190.

Thts work wa5 )upported by a contract (BBF) from Okbhoma Center for the Advancement of Science and Technology (OCAST) and by a research granr from th(' Presbyrenan Healrh Foundation.

Abbreviauon~:

cAMP: cycltc 3',5'-adenosme monophosphare CHAPS: 3-1(3-cholanudopropyl)dimetbylamrnonio] 1-propane~ul-

fonate dbcAMP: dtbutyryl cycltc 3',5'-adenosme monophosphare HBSS: Hank's balanced salt solution L-DOPA: L-dthydroxyphenylalanine MSH: a-melanocyre-snmularmg hormone PMSF: phenylm('thylsulfonyl fluoride

assays, tyrosine hydroxylase and l14C]melanin assays, were similar, suggesting chat tyrosine hydroxylase activity is tigh tly coupled to melanin synthesis. Tyrosinase activity de­termined by either assay method generall y correlated with skin melanin content. Kinetic analysis of tyrosinase from black and white foreskin revealed a K, for tyrosine of 2.5 X 10-• M in both skin types. lrnmunotitration experiments suggested that the difference in tyrosinase activities between white and black skin may be due, not only to different amounts of enzyme present in the mclanocytes, but also pos­sibly to differences in the catalytic activities of the enzyme found in melanocytes of black and white skin.] l1111est Derma­to/95: 9-15, 1990

[ 13,14]. Synthesis and degradanon rate studtes of tyrosinase in human melanocytcs in culture has shown a wide disparity m both of these parameters with some cells synthesizing the enzyme three times as fast and degrading it four times slower than other cell lines [15]. Studies with either human melanoma cells or with melano­cytes in culture, however, prov1de no information about the abun­dance or regulation of tyrosinase found tn normal melanocyres in the skin. One early study exammed tyrosinase activity in black and white human foreskin and found that tyrosine hydroxylase activity was higher m black skin [t 6]. Because of the paucity ofinfonnation available concerning tyrosinase expression in human skin and the relauonshtp between skin pigmentation and enzyme activity, we have earned out studies to characterize tyrosmase acnvtty and abun­dance in human foreskm from different skin types. In this paper we dtscuss methods for solubilizing ryrosmasc from human skin and for the measurement of tyrosinase activity by two assay methods, the tyrosine hydroxylase and [14C]mclanin synthem assays. We also report here on vanous properties of the enzyme from white and black skin, on the correlation between tyrosinase activity and skin pigmentation, and on results of immunoprecipitation analysis to relate tyrosinase activity with abundance in human foreskm.

MATERIALS AND METIIODS

L-tyrosine, dihydroxyphenylalanine (L-OOP A), 3-((3-cholamido­propyl)dimethylammonio] 1-propanesulfonate (CHAPS), sodium cholate, Triton X-100, and phenylmethylsulfonyl fluoride (PMSF) were purchased from Stgma Chemical Co. (St. Louts, MO). Char­coal (Norit SG activated charcoal) was obtained from MCB Manu­facturing Chemists (Cincmnati, OH). Tyrosine ([nng-3,5-3 H)tyro­sine; specific activity, 48 Ci/mmole) was purchased from DuPont New England Nuclear (Boston, MA). [I4 C]tyrosme (specific activ­ity, 350 mCt/mmole) came from JCN (Jrvine,CA). Pansorbin (Sta­phylcocws aureus cells} was purchased from Calbtochem (La Jolla, CA). All other chemicals were of analytic reagent grade.

0022-202X/90/S03.50 Copynghr © 1990 by The Society for lnvesttganve Dermatology, Inc.

9

10 IWATA ET AL

Preparat ion of Human Foreski n At the time of surgical re­moval, human foreskins were placed on srerile gauze saturated with sterile Hank's balanced salt solution (HESS) at 4 •c for rranspona­tion from the nursery (Oklahoma Memorial Hospital) to the labora­tory. The race of the cht!d was determined from the race of both parents. Foreskins were only used from children whose parents were eirher racially caucasian or black. No foreskins from racially mixed marriages were used. Following receipt of foreskin samples from the nursery, the tissue was either assayed for ryrosinase acriviry immediately or frozen at -7s·c. Prior to assay, rhe mucous mem­brane and underlying dermis were removed from the dermal side by inJection of Ham's F-1 0 nutrient medium followed by excision of rhe distended tissue by scissors. The skin was cur into 5-mm2 sec­tions. and either: l) assayed directly, 2) frozen and thawed 3 times in 0.5 ml of 0.1 M sodium phosphate buffer, pH 6.8 (phosphate buffer), in a dry ice/acetone bath, or 3) homogenized 3 times (30 sec each time with 30-sec rests) in 1 ml of either phosphate buffer or in detergent buffers (as described in the text) on icc with a hand-held homogenizer (Biospec Products, Inc., Bartlesville, OK). Following this homogenization, samples were sonicated 3 rimes (5-sec bursts with 30-sec rests) with a microtip sonicator (Hear Systems, Plain­view, NY). Homogenates were then left at 4 •c for 1 h prior to centrifugatio n. For studies ro determine the cellular distribution of ryrosmase and the effect of detergents on enzyme solubilization, homogenares were centrifuged at 50,000 X g in a Beckman ]2-21 centrifuge for 20 mm at 4 • C. The supernatants were removed for assay and the pellets resuspended by sonication in phosphate buffer (w1th or wnhom derergenr as indicated) prior to assay.

T yrosine Hydroxylase Assays Foreskin samples were assayed for the ryrosine hydroxylase acriviry of ryrosinase by incubating skin preparations (either 50 mg of skin tissue or 50 pi of extracts) in 0.5 ml of a reacnon mixture consisting of 0 .1 m.M tyrosine (or the concentrations listf'd in the tex t) , 0.1 mM L-OOP A, and 5 pCi/ml of [3H]ryrosine in phosphate buffer. This assay measures the pro­duction of3H20 during the conversion of[3H]ryrosine to L-DOPA [ 17]. Incubations were carried our at 37•c for the times indicated. T o terminate the reaction, 1 ml of charcoal (1 0% w jv, in 0.1 N HC I) was added co each assay tube, the samples centrifuged at 2000 Xg for 10 min at 4•c , and processed as described elsewhere [3]. The supernatants (0.5 ml aliquots) were placed in scintillation vials, scintillation fluid was added, and vials were counted in a TM Analytic 6895 scintillation coumer equipped with a DPM pro­cessor.

To measure the conversion of [14C]ryrosinc to melanin, the pro­cedure of Hearing and Eke I [ 18] was used. Aliquots of cell extracts (20 pi) were incubated in 30 .ul of phosphate buffer containmg 10 .ul of [14C]ryrosine (25 .uCi/ml, final specific activiry, 50 mCi/ mmole), 10 pi of a solution containing chloramphenicol (1 mg/ml) , cycloheximide (1 mg/ml), penicillin (1000 units/ml) , and bovine serum albumin (l mg/ml), and 10 pi of0.25 mM L-OOP A. Incubations were carrir:d our for the times indicated at 37•c in microtiter plate wells. Reactions were terminated by spotting 20 pi of each well onto Whatman 3MM filter discs and placing them in 10% trichloroacetic acid containing 50 mM ryrosi11e. The filters were washed as described by Hearing and Eke! [18].

Melanin Assay To determine the melanin content in human foreskin, skin sections (50 mg) were first homogenized m 1 ml of phosphate buffer as described above and the melanin in 0.35 ml of the resulting homogenate solubilized by treatment with 2 N NaOH at 6o•c for a mirumum of 5 d. An equal volume of water was then added followed by the addition of 0.7 ml of chloroform:phenol (1: I). The mixture was vortexed and cenrn­fuged at 5000 X g for l 0 min to separate the rhases. Proteins, in­cluding hemoglobin, separated with the pheno layer or at the inter­phase while melanin was localized ro the aqueous phase. Melanin content in the water layer was then determined spectrophotometri­cally at 400 nm [19,20]. All values are expressed as absorbance at this wavelength per mg of skin per ml of I N N aOH.

P rotein Assay The amount of skin protein was routinely deter­mined using a protem-assay !cit obtained from Bio-Rad (Richmond,

T H E JOURNAL OF INVESTIGATIVE DERMATOLOGY

C A). TillS ktt IS based on a protem assay described by Bradford [21]. Bovine serum album1n (fraction V) was used~ the reference ~ran­dard.

lmmunoprec ipitation Assays To detenmne the amount of im­munoreactive tyrosinase present in black and wh1rc foreskin, ho­mogenates were prepared with 1% CHAPS m phosphate buffer as described above. After a 1-h incubanon at 4 •c to solubilize the enzyme, homogcnates were centrifuged at 50,000 X g for 20 min at 4•c and the resulting supernatants used for immunoprccipiration analysis. Increasing amounts of ami-mouse ryro~inase serum [3,22] were added to a constant amount of cell-extract proccm (300 Jlg of protein from e1ther white or black skm extracts) m microccntrifuge rubes in a total volume of 300 pi of 0.05 M Tns-HCI, pH 7.4, containing 0.5% Triton X-100 and 150 mM NaCI.lncubation was carried our at 3rC for 1 h, at which rime 15 pi ofPansorbin (made up in the above buffer) was added. Tubes were then incubated at room temperature for 30 min with vorrcxing every I 0 min. The tubes were centrifuged at 12,500 X.~ for 5 min and the ryrosinasc activiry remaining in the supernatant detcrmmed with the ryrosine hydroxylase assay. Normal rabbit ~erum was used in place of anti­ryrosinasc serum in duplicace rubes to control for non-specific pre­cipitation of ryrosinase.

lmmunotitration analysis as onginally described by Feigclson and Grccngard [23] was carried out by incubating increasing amounts of ryrosinasc acttvity from ei ther white or black skin extracts with a constant amount (5 til) of anti-tyrosinase serum. Reactions were carried out m a total volume of300 JLI for 1 hat 37• c followed by the addition of Pansorbin as described above. Following centrifuga­tion, ryrosinasc activity was detennined m the supernatants using the tyrosine hydroxylase assay.

RESULTS

Iniual studies with human foreskm were des1gned to investigate different assay protocols to maximize assay sensi tiviry. Human fore­skins from black and white individuals were assayed for ryrosine hydroxylase activiry by either: 1) incubating foreskin pieces directly in (3H]ryrosine reaction mixture, 2) first freezing and thawing the foreskin samples to pcrmcabilize the membranes, or 3) homogeniz­ing the tissue. The results of these studies arc shown in Table I. Homogenization of the tissue increases the measured level of ryro­sinase acriviry 4 - 6 times over that determined in whole skin, proba­bly because tissue homogenization increases the availabiliry of ryro­sine for utilization by ryrosinasc. Table 1 also shows that the elimination of L-DOPA from the assay reduces activiry to back­ground levels, a findmg that demonstrates the specificiry of the assay for ryrosinase and not fo r ryrosine hydroxylase, an enzyme

Table I. Effect of D1ffcrcnt Homogenization Methods and of L- Dopa on Skin Tyrosinase Activiry

Skin Type

Whtte' W hite WhltC Black Black Black W hite• White Black Black

Treatment

None Freeze/thaw Homogenizanon None Freeze/thaw Homogenization Homogenatc-wnhour 1..-DOPA Homogenate+ L-DOPA (0.1 mM) Homogcnate-wtthom L-DOPA Homogenate+ L-OOP A (0. 1 mM)

Tyrosinase Activtty (pmoles 3H20/h/ mg skin)

3.53 ± 0.71 5.12 ± 0.22

19.42 ± 1.1 9 35.4 1 ± 1.0

33 ± 0.65 124.3 ± 0.87 nor detectable 38.91 ± 0.89 0.28 ± 0.23 57.7±0.16

• Fresh fore~kms were secttoned, dtvtded mro three eq~l wetghc groups, and etcher assayed dtreccly as sbn secuoru, frozen and thawed 3 umes m 0.1 M sodmm phosphate buffer, pH 6.8, pnor to assay. or homogemzed in buffer as described m M•~tri•IJ •nd MtthoJs. The tyrosine hydroxylase asuy was used to measure enzyme acu vity and assay incubauons were earned our for 4 h ar37 ·c. Values are chc aveugc of m plicares ± SD.

' Fore;kin homogcnares were prepared m phosphare buffer cont:atmng I% CHAPS and assayed for tyrosmase •cuvtty using the tyrnsme hydroxylase ;usay. Assays were earned out for 3 h 111 the presence or absence of L-OOP A (0. 1 mM). Values arc the ave ... ges of three detenmn~ttons ±SO.

VOL. 95, NO. I JULY 1990

wl11ch catalyzes the same conversion of tyrosine to DOPA but whtch requires terrahydropteridine as a cofactor [24]. Because the intracellular site for tyrosmase in other mammalian p1gment cells has been shown to be the melanosome, studies were next carried out to examme the intracellular distribution of this enzyme in human skin. Greater than 70% of the enzyme was found associated with organelles sedimenting at 50,000 X g (Table II). Chen and Chavin [25] found that tyrosinase m black skin was localized only to melan­osomes. microsomes, or small vesicle components pelleting at 100,000 X g. Because our centrifugations were carried out at 50,000 X .~ it is poSSible that any microsomal or very small vesicle components remamed m the supernatant (soluble) fraction. The organellc-assoctated form of tyrosinase could be solubiltzed effec­tively by Triton X-1 00 as well as by the zwitterion detergent, C HAPS (Table II). Although sodium cholate has been shown to be effective tn solubilizing tyrosmase in an active form from mouse melanosomcs [12,26,27], thts detergent was found to inhibit tyro­smase in human skin (Table II). The mechanism of tyrosinase inac­tlvaraon by sodium cholate is unknown. Table II also shows the stabiltty of human tyro~inase to storage. At 4 •c. tyrosinase activity IS stable for 24 h but loses 56% of 1ts activity by 72 h. The enzyme can he frozen without loss of activity. Studies were next earned out to determine rhe lineanty of the tyrosinase assay over ttme using two concentrations of tyrosine substrate, 0.1 mM and 0.01 mM. The results show that both white and black skin homogenates can be assayed up to 4 h m substrate concentrations of either 0.01 or 0.1 mM WI thoU[ any decrease in reaction rate (Fig 1 ). For all subse­quent srud1es, however, the tyrosine concentraraon was kept at 0. I mM to assure that an excess of substrate would be present throughout the assay period.

To determine if differences 111 tyrosinase actiVIty ex1st between white and black skin types, the tyrosine hydroxylase assay was ini­tially used to measure enzyme levels in homogenates prepared from 36 black and 39 white skin samples. As can be seen in Fig 2, black skins had higher levels of tyrosinase activity (mean, 33.8 + 3.9 pmoles 31120/hfmg sktn) than wh1te skins (12.71 + 1.9 pmoles 3 H20/h/mg skm). This almost threefold difference was highly

Table n. Solubilization of Human Tyrosinase from Foreskins and Stab1lity of Solubilized Enzyme

Treatment Tyrosmase Activity

(pmoles 3HP/h/mg skin) ----Homogenate } Cvtosol phosphate buffer" Pellet Homogenate } phosphate buffer Cvtosol + P~ll~t I% Tmon llomogenatc: } phosphate buffer Cytmol + Pellet 1% C HAPS Homogenate~ llomogenate + I% sodmm cholate Homogenate, T = 0' llomogenate, T = 24 h Homogen:lte, T = 72 h I lomogenate, frozen/ thawed

15.1 ± 0.1 4.8 ± 0.25

11.1 ± 0.8 18 ±0.7 17.8 ± 0.32 3.2 ± 0.24

22 ± 0.23 23.1 ±0.14

2.2 ± 0.22 10.9 ± 0.9 0.6 ±0.2

27.9 ± 0.2 22.9 ± 0.8 12.2±0.1 27.7 ± 0.8

• Pooled white foi'C'skuu wrre cut mto sm~ll sccoons and dtvtded mto three equal ~roups of 300 mg tiSsue <>ch. Foruktn w:u homogemzed m etther: I) 0.1 M sodium phosphate buffer. pH 6.8; 2) phosph•te buffer conammg 1% Triton X-100; or 3) phosph2te buffer contammg I% C IIAPS. A portion of each homogenate wos centri­fuged 21 50.000 X g for 20 mm at 4 ' C. The resulting supem2t:um 2nd pellets an.d the homogcnucs were •=yeti for ryrosmasc acnvtry for 4 hat 37'C u descnbc:d m Matt­nab and Mtthodt V:Uues au the averages of tnplic:are assays± SO. Expenmenrs were repeated 4 nmes wnh similar results.

• A foreskm homogen2u: (100 mg sktn wetght) W2S prepued tn I ml of phosphat<: buffer and one pomon a=ycd for ryrounue acrivirydircctly whtle a second altquot was first mcubued wtth I% sod tum cholate for I h priorto assay. Assays were earned our for 4 h at 37'C. Values arc' the averages of 4 derernunntons ±SO.

' A whtte forcskut homogenate was prepared m phosph2rc buffer contaming I% C HAPS and etther •ssayed immedtatcly or stored at 4 'C and ass2yed every 24 h. One portton w.s frozen for 3d and chen asS3yed. Assays were c:arned out for 4 h 2t 37'C. V2lues .re averages from tnpltc>te .ssays ± SO.

TYROSINASE ACTIVITY AND SKJN COLOR IN FORESKINS 11

0 0 800

N J:

"" / . .. 0-0 whtle ., 0 600 0 • - • block E / . s. ~ 0.400 5

/ . 0~0 ;:: u <t w • 0-------"' 0 :?00 <t

~0~ z iii

0 0 1 mM tyros •ne 0 (k

~ 0000 0 2 .l 4 5

TIME (hours )

6

5 / . >- 0 0 wh •le ... • - • b1od >~

;:~ / . ~ 3 "'"' / . 0------0 <t !! z 0 Vi[

2 oc • 0~ a:-

>-

~0~ ....

0 1 milA lyros•ne

0 0 ~ .l 5

I 'M( l hours)

Figure 1. Effect of substrate concenrranon on tyrosmase assay lmeanty. White (opm symbols) or black (solid symbols) foreskins (300 mg) were ho­mogenized m 2 ml of phosphate buffer containing I% Tmon X-1 00 and 200 ,ul assayed in a reaction mixture containing 2 ,uCi/ml of [JH] tyrosine, 0.1 mM DOPA, and either 0.01 mM tyrosine (top) or 0.1 mM tyrosine (bottom) for the times md1cated. Tyrosmase act1vity was determined as de­scrtbed m Mattnols a11d MtrhoJs. Values are the averages offour assays± SO. W hen SO uror bars are not present, th1s mdicatcs very small SO, obltterated by the symbol for the data pomr.

significant (p < 0.001 ). However, as can be seen from the inset in Fig 2, there is considerable overlap in tyrosinase activities between white and black skin, although the highest values obtained were always from black skin, whereas the lowest levels of enzyme activity were found only in white skin. Because tyrosinase is the rate-limit­ing enzyme for melanin synthesis, we next investigated the correla­tion between tyrosinase activity levels in human foreskin and mela­nin content. Although measuring the absorbance of sodium hydroxide -solubilized melanin has been used effectively co deter­mine melanin contents in melanocytes and melanoma cells in cul­ture [19 ,20,28- 30), the presence of hemoglobin (which absorbs in the wavelength range used for melanin) in skin could cause marked interference with this method of melanin determination. To elimi­nate the contaminanon by hemoglobin, the alkali-solubilized skin was extracted with phenol, a step which completely removed he­moglobin contaminants without causing any loss of melanin. By using this method, we were able to quantitate melanin in human foreskins and to compare melanin levels with tyrosinase activities. In this srudy, 20 white and 20 black skins were sectioned and di­vided into four groups. One group was homogenized in 1 o/o Triton X-100 buffer and assayed with the tyrosine hydroxylase method, whereas in a second group melanin synthesis activity was deter­mined with the [14C]tyrosine assay method. A third group was used to measure tyrosinase activity (tyrosine hydroxylase) in frozen/ thawed samples, while the fourth group was extracted, as discussed above, to determine melanin content of the skin samples. Figure 3A shows the correlation between tyrosinase activity and melanin con­tent in skin homogenates prepared in 1% Triron X-1 00. Linear regression analysis of this data yielded a correlation coefficient be­tween tyrosinase activity and melanin content of 0.4 71 that is highly significant (p < 0.005). When a second group of skin ho-

12 IWATA ET AL

,;---...

0"

E '-..... ...c '-..... 120 0 N I n

(/)

90 a.> -0

E 0..

'-..._/

>.. 60 +-'

> +-' u

<{ 30 a.>

(/)

0 c (/)

0 0 \.....

>.. I-

~ •o ' ~ ' ~ ~JO

; ! 20 !: > ~ 10

i !- 0

y

' ; • • • I l

WHif( BLACK

WH ITE

•

• • • • I

I BLACK

Figure 2. Tyrosanase acttvmcs 111 black and wh1te foreskan homogenates. Black (solid .1ymbol:~) and wlmc (ope11 symbols) forcskms were homogemzed 1n phosphate buffer contauung 1 o/o Triton X-1 00 and assayed for tyro~ me hydroxylase acuvity of tyrosinase for 4 h as described in Materials aud Mtthc>ds. The ~ubstrate concentratiOn was 0.1 mM tyrosine. Values are aver­age~ of mpllcate~ ±SO. Average tyrosmase values were 12.7 prnoles 3H 20/ h/ mg skan we1ght for wh1re skin and 33.8 pmoles 3H 20/h/mg skin we1ght for black foreskins. The 1nset shows an expanded Y axis from 0- 40 pmoles 3H 20 / h/ rng skm.

mogenates were assayed for melanm-synthesis activity usmg the p•ejtyrosme assay, similar results were obtained (Fig 3B; correla­tion coefficient, 0.463; p < 0.005). Skin samples prepared by the freeze/thaw method also demonstrated a strong correlation be­tween tyrosmase acttvity and melanin content (Fig 3C; correlation coefficient, 0.445; p < 0.005). When the tyrosine hydroxylase and ( 14Cjmelanin-synthesis assays were compared, we observed a very close correlation between the rwo (correlation coefficient, 0.996; p < 0.001), a findang which suggests that melanin synthesis is closely coupled to ryrosmase acnvity, specifically the initial step in melanin synthesis (Fig 4A). A good correlation was also noted be­rwecn tyrosinase activity determined m homogcnates and that de­termmcd in skm p1eces frozen and thawed before assay (Fig 4B, r = 0.997). Thus, etther method can be used to reliably assess the awvity ot tyrosmasc in human skin. W c next determined if tyrosin­ase from wlme and black skins reacted with the substrate, tyrosme, w1th the same efficiency by carrymg our kinetic studies on tyrosin­ase. As 1s shown an Fig 5, we found that the K, values for tyrosme arc the same 111 wlme and black skin (2.5 X 10-• M).It is intercsung to note that both tyrosinase preparations show an inhibition of enzyme actiVIty as the tyrosine concentration approaches 1 mM. This substrate mhib1tion effect has been also observed for other mammahan tyrosanases (31]. It IS possible that this inhibition is due ro impurities in the tyrosine preparation which, when used at higher concentrations, may Interfere with enzyme activity. However, ty­rosmc preparations from different manufacturers gave s1milar re­sults (data not shown).

The h1gher activity of tyrosinase measured in black skin samples could be due to either: I) more tyrosinase molecules per melanocyte, or 2) htgher catalytic actiVIty per tyrosinase molecule. To invest!-

THE JOURNAL OP INVESTIGATIVE DERMATOLOGY

gate whiCh mechanism may be operanve ut human skin melano­cytes, rwo immunochernical titration analysiS methods were used. In the first method (32], a constant amount of cell-extract protein is added to increasing amounts of anti-tyrosmase antiserum and the potnt at which all of the enzyme IS removed from solunon IS deter­mmed. If rwo cell types have d1fferent levels of tyrosmase activity because of different amounts of enzyme, then for an equal amount of skm-extract protein, it will take more antiserum to remove all of the enzyme activity from the skin homogenate, which contains more enzyme molecules per given amount of cell extract. If, however, the d1fferencc in enzyme activities berween black and white skin is due to differences in the catalytic activity of equal numbers of enzyme molecules (that is, a population of inactive or catalyncally less active enzyme exists in white skin) then for an equal amount of skin­extract protein (containing equal numbers of tyrosinase molecules), it will take the same amount of anti-tyrosinase serum to precipitate all of the enzyme activity from either white or black skin regardless of the difference in enzyme activities berween the rwo skin types. When this type of analysis was carried out with skin extracts from wh1te and black foreskins in which the tyrosinase activity of the black extract was 3.8 times that of the white extract, we found that 1t took approximately the same amount of anttserum to precipitate all of the enzyme activity from either skm type (Fig 6A). If the 3.8-nmes h1gher level of tyrosmase activity in the black skin was due to a 3.8-times increase in the number of tyrosanase molecules per cell, more antiserum would have been required to precipitate all of the

70r--------------------------------~ A

60

•o •

c 50 • 40 •

• • 20

•

•

• •

• • •

• • .. ~. • •

12 I~

Figure 3. Correlmon of tyrosmase acnv1ty and melamn content m human fore•km homogcnarcs. Foresktns (20 black and 20 wh11e) were sectioned and d1v1ded mco groups of equal we1ght. One group was homogemzed in phos­phate buffer conr:aining 1% Tmon X-1 00 and assayed for tyrosanase acnvtty usmg the ryrosme hydroxylase assay (A) or the (14C]melanin assay (B). Another group was frozen and rhawed 3 tunes m phosphate buffer and assayed for tyrosmase actiVIty usang the tyrosane hydsoxylase :may (C). The last group was used for melanan-contcnt dcrcrmmauons as described m Materials and Methods and 111 the text. All tyrosmase assays were earned out at 37•c for 4 h. Values arc the averages of tripltcatcs ± SD. Black foreskins, solid symbols; white foreskms, opfll symbols. r = 0.471. p < 0.005 (A); r = 0.463, p < 0.005 (B); r = 0.445, p < 0.005 (C).

VOL. 95, NO. I JULY 1990

~

c .i

"' C7'

>- E ........... 5 ~ -c ........... U c <{ -we ~~ z E Vii au (l:V

~-.. "' 0 E ~

'i' 0 .c~

- c ,_ ., "' N Ill

" C7' !! E :::..-..... >- .c ....... ~0 ..... , . l) J: <! "l w .. Vl "' ~ 0 - E Vl c. o-Cl: >-.....

70 A

"j 50

40

30

I 20i }! • 1:~~

0 10 20 30 40 so 'YROSINA'SE ACTIVIT'f

(prr•oles .3H20/hr/rng s>H')

60 ~ B

51 • ./ 40 • •

' 30 , !/ •• 20 .. (l,,.

IOL P.~ c9 0 ~ +- ..... -----1

0 _o .30 10 50 T•POr 'lA [ t.l 'IVITY ('• ton • 100)

1 pf"'\r·lcs 3H:01nr/rrog ~,n\

60 70

/ •

IJO 7C

Figure •· Comparis1on of ryroomase assay methods. (A) The data from Fig 38 (('4 C]melanin assay) was plotted against the data from Fig 3A (ryrosine hydroxylase assays in detergent-treated homogenates). The correlation coef­ficiem is 0.996. (B) The data from Fig 3C (freeze/thaw method) was plotted agamst the data from 3A (ryrosme hydroxylase assays 111 detergent-treated homogenates). The correlation coefficient is 0.997.

enzyme activity from the black skin extract than from the white skin sample. That this was not the case suggests that enzyme-activ­ity differences between skin types may be due not only to differences in tyrosinase abundance but to differences in the catalytic efficiency of the enzyme molecules.

To provide additional support for the presence of different cata­lytic forms of tyrosinase in different skin types, we used a second type of immunochemical analysis, which was developed by Feigel­son and Greengard [23} and which is based on the principle of equivalence point titration. If one adds increasing amounts of en­zyme acti11ity to a fixed amount of antiserum to form an immuno­precipitate, the point at which activity begins to appear in the super­natant represents the equivalence point. If black and white skin extracts differ in tyrosinase activity, not because of different num­bers of tyrosinase molecules per melanocyte, but because black skin extracts contain a form of tyrosinase that has a higher catalytic activity than the enzyme found in white skin melanocytes, then a dtfference in equivalenceJotnts will be observed when immunoti­tration studies are carne out. This occurs because the antibody binds both catalytic forms of the enzyme and is saturated at lower activity by the enzyme preparation containing tyrosinase molecules with low catalytic activity. The extent of the shift is proportional to the difference in catalytic activity. Alternatively, if melanocytes from black skin have higher tyrosinase activities simply because of higher numbers of enzyme molecules per cell than white melano­cytes (and not because of differem catalytic activities), no difference in equivalence points will be observed when immunomranon stud­ies are carried out. This is because the titration analysis is being carried out on the basis of increasing enzyme activity and not by increasing amounts of cell-extract protein. When this type of analy­SIS was carried out with wh1te and black foreskin extracts that

TYROSINASE ACTIVITY AND SKIN COLOR IN FORESKINS 13

showed a 2.9-times dtfference 111 tyrosinase actlvtty, an equtvalence potnt shtft was observed (Fig 68). The equivalence point for the black sktn extract (8. 7 ,UU) was 2.4 times higher than that measured for the whtte skin extract (3.64tJU). Because the anti-tyrosinase antiserum was saturated by 2.4 times less activity wtth the white skin preparation than with the black skin, this suggests that the catalytic activity of tyrosinase molecules in melanocytes from white skin is approxunately 2.4 times less than that found m black melanocytes. This difference m catalyttc acttvity can account for the majority of the 2.9-nmcs difference in tyrosinase activity between black and whtte skin extracts.

Thus, the results of two types of immunochem1cal analysis sug­gest that differences in tyrosinase activity between black and white skm may, at least in part, be explained by differences in the catalytic efficiency of tyrosinase molecules.

DISCUSSION

The molecular mechamsms that are involved in human ptgmenta­non arc largely unknown at present. Results from studtcs earned out many years ago demonstrated that the level of vtsible pigment in human skin is likely due to a combination of factors mcluding: 1) the amount of tyrosinase present in epidermal melanocytes, 2) the rate of synthesis and the abundance of melanosomes in melanocytes, 3) the efficiency of melanosomc transfer to surrounding keratino­cytes, and 4) rhe rate of degradation of transferred melanosomes (see [33,34] for review). The present study has focused on tnvestigating the possible relationship berween tyrosinase activity and human skm color. For this work, tyrosmase activities and melanin content were determined in a number of human foreskins from black and white ind1viduals. The only other srudy to examine tyrosinase activ­Ities m human skm from different races was carried out by Pomer­alltZ and Anccs [16]. In that study, melanin levels were not measured and thus no correlation between tyrosinase activity and melanin content could be determined. However, the levels of tyrosinase activities reported in this early investigation agree with our find­ings, that is, tyrosinase activity is generally higher in black skin but with considerable overlap in activities. In our mvesngation we found that tyrosinase activities in black skin are approximately 3 nmes rhar in white skin, whereas Pomerantz and Ances reported a 2.25-tirnes difference. Addinonal results from our study show that: 1) tyrosinase can be solubilized effectively by non-ionic or zwitter­ionic detergents, 2) tyrosinase activity generally correlates with

140

120

100 ~

!! 80 ·c: ::l ::l

60 ~

> 40

20

V/(S) (uUn1tsjmM)

Figure 5. Kmerics of ryro;ma~e ac!lvtry m black and whltC: forcskm. Fore­skm homogenate~ (black, sol1d symb,>ls; whae, ope11 symbols) were prepared m phosphate buffer conmmng I% CHAPS and the ~upernatan~ from a 50,000 X g cenrrifugauon (0.9 mgfa~~ay rube) assayed for 3 h 111 a reacnon mixture conrammg 6 f.JCI/ml ofi3Hiryrosine, 0.1 mM DOPA and ryrosme concentrations ranging from 0.0 I mM to 2.5 mM. Data IS graphed using a Hofstcc plot (Km is determmed from the slope and v mn IS from the y inccrccpt). V "'"'' for black skm - 125.3 f.JU; V mu for white= 58.4 f.JU. Km for both i; 2.5 X 1 o-• M. Values arc averages of tnplicate a!.say~ ± SD.

14 IWATA E.T AL

,_100 .-------------------------------------------, c £ 0

~ 80 >.a.

-:! :> .~ til

~ s 60 0' ., c

~ c c 0 0 -~ E 4 '- ., >. '­

..... ::: ~ 20 i

n

u i 0 0 i

I 6 ~ 0+-------r-------IW------+------~----~

00

c 25 ~ 0

t ., Q_ 20 :> Ill

.s "0 ,_

::t c "' :> ;: ~ ::l £3 -~ u <f ., Ill 5 0 c ' iii 0

>. 0 ..... 0

05 10 1.5

Anltserum Added (,t.Liot ers)

0 Whote. X-ont • .3 64~U • Block. ,_,nt. 8 7 uU

~· 10 20 30 40

2.0

n 50

fvro~oncs,., Ac(lvoty Added (1-4Unots)

25

60

Figure 6. lrnmunoprec1pication analysis of tyrosinase from black and white foreskin. (Top) Foreskin homogenates were prepared from black {solid sym­bols) and wh1te foreskm {optll symbols) and 300 J.lg of protem trom each exrracr incubated With increasing amounts of anti-tyrosinase serum for I hat 37"C followed by a 30-min incubation ar room temperature with Pansorbm (Staph A). The immunoprcc1p1t4rion complexes were collected by centrifu­gation at 12,500 X g and the tyrosmasc activity remaining in the supernatant determined by assaying for tyrosme hydroxylase activity at 37•c for 4 h. Values have been corrected for non-specific binding, determined by using normal rabbit serum instead of anti-tyrosinase serum. Non-specific binding was very low in all experiments(< 0.2%). Values are the averages of tripli­cate dcrermmarions ±SO. The inset shows rhe tyrosinase activities of rhe rwo cell extracts (black has 3.8 times more enzyme activity). Values are the averages of tnplicate assays± SO. (Bottom) Immunotitrarion analysis on black and wh1te foreskin extracts was carried out as described above except char increasing amounts of tyrosinase aaivity were added to a co11sta11t amount of ;~mi-tyrosinase serum (Sill) and rhe reaction allowed to continue for I hat 37 •c. Pansorbin was then added to prcc1pirate rhe ancigen-annbody com­plex and tyrosinase acnvity 111 rhe supernatants decernuned as described above. The: correlation coeffic1cnr for the linear regression plot of the: white skm c:xrracr (ope" nrclrs was r = 0.997 and for the black skin sample {solid mcles) 1t was r = 0.994. The X intercept was 3.64 J.lU for rhc: white: and 8.7 pU for the: black exrract. Values are the averages of triplicate dc:termjnations ±SO. The inset shows the tyrosinase activities of the rwo cell extracts (black has 2.9 times more enzyme activity). Values arc the averages of tnpi1Cate assays± SO.

melanin content 111 skin, 3) tyrosme hydroxylase activity is a good indicator of melanin synthesis, 4} tyro~inases from black and white skin react with tyrosine with the same efficiency (Km values are the same) , and 5) variations in tyrosinase activity in human skin may be rhe result o f differences in both enzyme abundance and catalytic activity.

Considering the numerous factors that regulate the production of skin pigment, it is interesting to find that melanjn content in skin is directly correlated with tyrosinase levels in human mclanocyces.

THE JOURNAL OF INVESTIGAT£VE. DERMATOLOGY

The correlation coefficient for these studies was approximately 0.450 and the correlation was found to be highly stgn ificant (p < 0.005}. There are, however, examples from the data in Fig 3 in which some whire sk in samples have tyrosinase activities that are equivalent to those found in black skin, yet little melanin is made. lt is possible chat pheomelanin and not eumelanin is being made in , at least, some of these white skins with high tyrosinase levels. In chis case, our melanin assay may not reflect the amount of this type of melanin because it is opumized to read eumelamn at 400 nm. No visible pigment was observed in white skin samples with high tyro­sinase activiues, however, and it seems more likely that these skin types were stmply not producing melamn. The poor melaninization of the skin in these examples could be due to the factors listed above, that is, low rates of melanosome synthesis or ttansfer to keratino­cytes, or rapid degradation of internalized melanosomes. In addi­tion, it is possible that post-tyrosinase steps in melanin synthesis may be regulated in some types of white skin leading to low melanin production in the presence of high tyrosinase activity. Several stud­ies with melanoma cells have suggested the presence of a "blocking factor" rhat prevents rhe conversion of 5,6 dihydroxyindole to mel­anin [35,36]. If this "blocking factor" activity were present in human melanocytes one might expect differences to ex 1st between tyrosinase activities determined by the tyrosine hydroxylase assay (which measures the first step in the melanin synthesis pathway} and the [14C)melanin assay (which measures the entire pathway, from tyrosine to melanin). As is shown in Fig 4, however, there is an excellent correlation between the two assays, suggesting char post­tyrosinase regulation may not play a pivotal role in regulating mela­nin synthesis in white skin. In addition, because of rhe good correla­tion between the two assays in black skin as well as in white skin, 1t appears chat post-tyrosinase regulation is also not occurring in these darker skin types.

The results of rhe kinetic studies on white and black skin extracts show that the Km values for tyrosinases from black and white skin are similar. This finding indicates that the affinity of tyrosine for rhe enzyme is similar between slcin groups but does not provide any information on the enzymatically driven r.1te of product formation (i.e., turnover number). In order to determine if the catalytic effi­ciency of tyrosinase differs between black and white skin types, the two types of immunotitration analysis shown in Fig 6 were carried out. The results from both methods of analysis suggest rhac differ­ences in tyrosinase activity levels between black and white skin preparations cannot be explained solely o n rhe basis of different numbers of tyrosinase molecules present in black and white melano­cytes. Instead , the data suggest that the catalytic efficiency of the enzyme from white skin may be lower than that of tyrosinase from black skin melanocytes. This lowered activity may be the result o f: 1) the presence of tyrosinase inhibitors, 2} altered post-translational processing of the enzyme, o r 3) the production of alternatively spliced tyrosinase mRNA transcripts, which results in the synthesis of different forms (catalytically repressed} of tyrosinase. Evidence for tyrosinase inhibitors (37- 39], altered processing (40] , and alter­native splicing of tyrosinase mRNA [41 J have been reported but which, if any, mechanism is operative in human melanocytes re­mains to be determined. Although the immunotitration studies sug­gest that it is the difference in catalytic efficiency of tyrosinase and nor differences in the number of tyrosinase molecules that leads to different levels of tyrosinase activity in black and white human skm, more conclusive evidence for this can o nly be obtained by quantitat­ing and charactenzing tyrosinase in different skin types by r.~diola­beling and immunoprecipitation analysis. Studies with human fore­skin organ cultures are now in progress to characterize the molecular mechanisms responsible for variations in tyrosinase activ­ities and abundance in different human skin types.

REFERENCES

I. Pawelek JM, Korner AM: The biosynthesis of mammalian melanin. Am SCI 70: 136 - 145, 1982

2. Lerner AB, Fitzpamck TB: Biochemistry of melanin formation. Physiol Rev 30:91 - 126, 1950

VOL. 95, NO. I JULY 1990

3 Fulltr BB. Lunsford Jll, !nun OS: Alpha-mclanocyte-snmulann15 hormone regulatton of ryro~mase m cloudman S-91 mouse mela­noma cell cultum. J Il10l Chem 262:4024-4033, 1987

4. Jimenez M, Kamcyama K, Maloy WL, Tomita Y, Hearing VJ: Mam­malian ryrosmase: blo\ynthcm, processing, .ll1d modulation by mel­anocyte-stimulaung hormone. Proc Nad Acad SCI USA 85:3830-3834, 1988

5. Wong G, Pawelek j: Control of phenoryp1c expres~10n of cultured melanoma cell~ by melanocyte sumulaung hormones. Nature (New B1ol) 241:213-215. 1973

6. Halaban R, Pomerantz SH, M•rshall S, Lerner AB: Tyrosmase act1v1ry and abundance m cloudman melanoma cells. Arch B10chcm B1ophys 230:383- 387. 1984

7. Fuller BB. Vi~kochd DH: The role of RNA and protem synthc~l~ 111 mediating the acnon of MSII on mouse melanoma cells. L1fc Sc1 24:2405-2416, 1979

8. Lev111e N, Lemus-Wdson A, Wood SH, Abdcl-Malek ZA: Snmula­uon of follicular melanogcnem 111 the mouse by top1cal and 111JC:Cted melanorropms. J lm·est Dermatol 89:269-273, 1987

9. H1robe T. Takeuchi T· lnducnon of melanogeneSIS 111 l'ltro m the ep1dermal melanoblam of newborn mou\e skm by MSH. In Vmo Cell Dev Blol13:311-315, 1977

10. Burchill SA, Virden R. Fuller BB, Thody Aj: Rcgulanon of ryro;mase synthesis by melanocyte swnulat111g hormone in ha1r follicular mcl­anocytcs of the mou\c. J Endocrinol 116:17 23. 1988

II . Fuller BB, Meyskens FL: Endocrine rc;pons1veness 111 human melano­cytes and melanoma celb 111 culture. J Nad Cane lnst 66:799-802, 1981

12. N1sh1oka K: Particulate rvrosmase of human malignant melanoma. Eur J Biochem 85:137-146, 1978

13. Ruppert S, Muller G, K won U, Schurz G: Mulnple transcnpts of the mouse ryrosmase gene: arc: generated by alcernanve spllcmg. f.M.BO J 7:2715-2722, 1988

14. Kwon BS,llaq AK, Po merantz SH. Halaban R: lsoi ,Jtion .1nd sequence of a eDNA clone for human ryrosmase that maps at the mouse c-albmo locus. Proc Nad Acad Sci USA 84:7473-7477, 1987

IS. Halahan R, Pomerant7 Sll, MarshallS, Lambert DT. Lerner AB: Reg­ulanon of tyrosma~e 111 human melanocytes grown 111 culture. J Cell Blol97:480-488, 1983

16. Pomerantz SH, Ancc:s IG: Tvrosmase activ1ry 111 human skm.lnAucnce of race and age mne\\horns.J Clin Invest 55:1127 I 131,1975

17. Pomerann SH: TyrminC' h)•droxylanon catalyzed by mammalian ryr­osinase; an improved merhod of assay. B1ochem B10phys Res Com­mun 16:188 - 194,1964

18. Heanng VJ, Eke! TM: A compmson of ryrosme hydroxylanon and melanm formmon. B1ochem j 157:549-557. 1976

19. Whittaker JR: Changes m melanogenesis dunng the dcdifferennation of ch1ck rc:unal pigment cells 111 cell culture. Dev B1ol 8:99-127, 1963

20. Lotan R, Lotan D: Enhancement of melanouc expremon 111 cultured mouse melanoma cells by rcuno1ds. J Cell Phy~1ol 106:179 - 189, 1981

21. llradford M: A rap1J and scmmvc method for the quant1tat1on of m1crogram quantities of protem uuliz1ng the pnnc1plc of protem­dye bmdmg. Anal B1ochcm 72:248-252, 1976

22. Fuller BB, I man DS, Lunsford J B: Companson of tyrosmase levels m amelanonc and mdanotiC melanoma cell cultures by a compennve enzvmc-lmked 1mmunoadsorhenr assay and bv 1mmunomranon anaivsis. J Cell Phy~1ol 134:149- 154, i 988 .

TYROSINASE ACTIVITY AND SKIN COLOR IN FORESKINS 15

23. Fe1gelson P, Greengard 0: lmmunochemJcal ev1dence for mcreased titers of hver tryptophan pyrrolase dunng substrate :tnd hormon~l enzyme mducuon.J B1ol Chern 237:3714-3717, 1962

24. N.1gatsu T, Levm BG, Udenfncnd S: Tyrosme hydroxylase: rhe rate hm1ting ~rep in norepmcphnne biosynthesis. J B1ol Chcm 238:2910-2917, 1964

25. Chen YM, Chav111 W : Tyrosmase activ1ry 111 a h1ghly p1gmented human melanoma and 111 Negro sk111. Proc Soc Exp Bioi Mc:d 145:695-698, 197 4

26. Yamamoto H, Brumbaugh JA: Punficanon and !Soclecmc heteroge­nclrv of clucken tyrosinase:. B1ochem B1ophys Acu 800:282-290, 1984

27. lsh1guro S, Yamamoto H, Yana1 N, Takeuchi T: Quanmanve analysis of mouse ryrosinasc by en7yme-lmked immunosorbent assay. J Bio­chcm 99: I 081 - 1085, 1986

28. Cobb JP, McGrath A: Irr 1111ro effects of melanocyte-stimulating hor­mone, adrenocorncotrop1c hormone, 17-estradiol, or testosterone prop1onatc: on Cloudman S91 mouse melanoma cells. J Nat! Cane lnsr 52:567-570, 1974

29 K1tano Y, Hu F: Prohferauon and dtfferennat1on of p1gmenr cells in 1•1tro. J Invest Dermarol 55:444-451, 1970

30. Saek1 II, Oikawa A: Sumulauon by 1onophores of ryros111ase acr1viry of mouse melanoma cells 111 culture. J Invest Dermatol 85:423-425, 1985

31 . Pomerann SH: The ryrosmc hydroxylase activity of mammalian ryro­smase. J B1ol Chern 241: I 61- 168, 1966

32. Ganschow RE, Schsmkc RT: Independent genenc control of the cata­lync acnv1tv and the rare of degradanon of catalase m mice. j Bioi Chern 244:4649-4658, 1969

33 Toda K, Pathak MA, Finpatnck TB, Quevado WC, Monhwa F, Nakayama Y: Skm color: Its ulrrasrructure and ItS determining mechanism. In: Riley V, McGovern VJ. Russel P (eds.). P1gment Cell. Vol. I. S. Karger, Basel, 1973, pp 61-81

34. Pathak MA,Jimbow K, Fttzpatnck TB: Photobiology of pigment cell. In: Seiji M (ed.). Phcnoryp1c Expression in Pigment Cells. Univc:r­my of Tokyo Prm, Tokyo, 1981. pp 655-670

35. Logan A, Weatherhead B: Posr-rvrosinase mhibmon of melanogenesis by melatomn in h;ur foll1clc:s 111 vitro. J Invest Dermatol 74:47-50, 1980

36. lleanng VJ, Korner AM, PawelekJM: New regulators of melanogen­eSIS are assoc1ared wuh punfied ryrosinase 1sozymes. J Invest Der­nmol 79: 16 - 18, 1982

37. Lee TH, Lee MS. Lu M: Effects of a-MSH on melanogenesis and ryrosmase of B-16 melanoma. Endocrinology 91 :1180 - 1188, I 972

38. Wong G, Pawelek J : Melanocyte-stimulating hormone promotes acti­vation of pre-ex1mng ryrosmasc: molecules m Cloudman S91 mela­noma cells. Narurc 255:644 - 646, 1975

39. Vljayan E, Husam I, Rama1ah A, Madan NC: Purilic.r1on of human skm ryrosmase and ItS protem mh1b1tor: propemcs of the enzyme and the mechamsm of 111h1b1uon by protem. Arch Biochem Biophys 217:738-747, 1982

40. H.1lahan R, Moellmdnn G, T amura A, et al: Tyros1nascs of munne melanocytes with mutations at rhe albino locus. Proc Narl Acad Sci USA 85:7241-7245, 1988

41. Muller G. Ruppert S, Schm1d E, Schutz G: Funcuonal analysis of altemanvc:ly ~phcc:d cvrosmasc: gene rranscnpts. EMBO J 7:2723-2730, 1988