[CANCER RESEARCH 41, 467-472, February 1981 ]

Melanin Content of Hamster Tissues, Human Tissues, and VariousMelanomas1

Karen P. Watts, Ralph G. Fairchild, Daniel N. Slatkin, Dennis Greenberg, Samuel Packer, Harold L. Atkins,and Stephen J. Hannon

Brookhaven National Laboratory. Associated Universities. Inc.. Upton. New York n973

ABSTRACT

Melanin content (percentage by weight) was determined inboth pigmented and nonpigmented tissues of Syrian goldenhamsters bearing Greene melanoma. Melanin content was alsomeasured in various other melanoma models (B-16 in C57mice, Harding-Passey in BALB/c mice, and KHDD in C3H

mice) and in nine human melanomas, as well as in selectednormal tissues. The purpose was to evaluate the possibleefficacy of chlorpromazine, which is known to bind to melanin,as a vehicle for boron transport in neutron capture therapy.Successful therapy would depend upon selective uptake andabsolute concentration of borated compounds in tumors; theseparameters will in turn depend upon melanin concentration inmelanomas and nonpigmented 'background" tissues. Ham

ster whole eyes, hamster melanomas, and other well-pig-mented animal melanomas were found to contain 0.3 to 0.8%melanin by weight, whereas human melanomas varied from 0.1to 0.9% (average, 0.35%). Other tissues, with the exception ofskin, were lower in content by a factor of 230.

Melanin pigment was extracted from tissues, and the melanincontent was determined spectrophotometrically. Measurements were found to be sensitive to the presence of otherproteins. Previous procedures for isolating and quantifyingmelanin often neglected the importance of removing proteinsand other interfering nonmelanic substances.

INTRODUCTION

The biosynthetic pathway of melanogenesis and the pigmentation of melanomas have been studied in some detail (31);however, little information is available on the absolute melanincontent of various tissues. Such data would be valuable sincetumor diagnosis and localization may be possible by labelingphenothiazines and related polycyclic aromatic compoundsknown to bind to melanin. In particular, CPZ3 has been found

to bind to melanin (a completed charge transfer reaction isthought to be the binding mechanism). The possibility existsthat a borated analog of this compound could be used as avehicle for transporting boron to melanomas for NCT (19-21 ).This concept is currently being evaluated.4 Successful appli

cation of NCT depends on selective uptake of borated compounds and upon absolute concentrations of boron (~30 ^g

1Work supported by National Cancer Institute Grant R01-CA22749. United

States Department of Energy Contract DE-AC02-CH00016, and United StatesEnvironmental Protection Agency Contract 79-D-X-0533.

2 To whom requests for reprints should be addressed, at the Medical Depart

ment, Brookhaven National Laboratory. Upton, N. Y. 11973.1The abbreviations used are: CPZ, chlorpromazine; NCT, neutron capture

therapy; DOPA, 3,4-dihydroxyphenylalanine.4 R. G. Fairchild and S. Packer. Boron Thermal/Epithermal Neutron Capture

Therapy. National Cancer Institute Grant CA-22749 (1979-1981).Received April 1. 1980; accepted October 16, 1980.

'°Bper g of tissue are needed) (6, 10). In the case of CPZ,

these parameters will be critically dependent upon melaninconcentration in tumors, as well as nonpigmented "background" tissues. Greene melanoma in Syrian golden hamsters

is currently being used as an animal model for human melanoma; consequently, these studies were initiated in order tointerpret ongoing CPZ distribution measurements and makepossible prediction of possible sites and magnitude of CPZaccumulation. Melanin content in other animal melanomas wasalso determined. Further, 9 different human melanomas (andother selected tissues) were evaluated for melanin concentration. Clearly, CPZ accumulation in animal models will be representative of anticipated uptake in human melanomas only tothe extent that melanin concentrations are comparable.

One aspect of differentiation in melanoma tumor cells isexpressed by melanogenesis. In general, cell differentiationand proliferation are considered mutually exclusive. Althoughpigmented melanoma is often considered to represent an exception (i.e., proliferation and melanogenesis occur simultaneously), melanin content has been shown to be inverselyproportional to rate of proliferation (5, 14). Most animal melanoma models have a rapid volume-doubling time (on the order

of a few days) which compares to 6 to 7 weeks for humanmelanoma (23, 33). Whether the animal pigmentation accurately reflects the human melanomas needs to be investigated.

We are concerned here with the black-brown eumelanins

derived from the enzymic oxidation of tyrosine or related metabolites (Chart 1). These insoluble, nonfluorescent, nitrogen-containing pigments are considered to be polymerization products of high molecular weight and occur either uncombined oras loosely bound, noncovalently bonded conjugates with proteins of undefined structure and function (36). The complex,irregular chemical structure of melanin has yet to be completelydefined (29), in part because the isolation procedures denaturethe melanin^while experiments with synthetic melanin in vitrolack the influence of melanoprotein. In addition, synthetic mel-anins may not be identical in structure to natural melanins;nevertheless, a close similarity seems probable (24).

A hypothetical structure for melanin in Harding-Passey mel

anoma has been developed (12), assuming a copolymerizationproduct of dopaquinone, indole-5,6-quinone, and indole-5,6-quinone-2-carboxylic acid (3:2:1) (Chart 2). Alternatively, the

early work by Raper (30) suggested a regular homopolymerformed by repeated oxidative condensations of indole-5,6-qui-none. The consensus is that natural pigments are highly irregular, 3-dimensional polymers consisting mainly of indole units

at different oxidation levels, with free radicals trapped in segments of the structure (4, 29).

Both natural and synthetic melanins have a broad absorb-ance spectrum, without any characteristic peak in the range of2,000 to 24,000 A. Since melanin does not exhibit a charac-

FEBRUARY 1981 467

Research. on November 26, 2015. © 1981 American Association for Cancercancerres.aacrjournals.org Downloaded from

K. P. Waffs ef al.

PrE'lYLALAIINE

TYRDSIÃŽC

3ADIHYDRDXYPr€NYLALANINE(DOPA)

. t

NM

MOÛCDopoquinone

3-HYD«3)CmRAM!NE(DOPAMINE)

NOREPINEPHRIÃŽt

EPINEFTOt

DOP/aUINCNE

Å’UKDDOPACHRQrt4.

DOP/ORO*4

5,5-DIHYDRDXriNDOÅ’

INDOÅ’-5,6-QUINCNE

¿MINChart 1. Biosynthetic pathway of melanin and epinephrine (from Ref. 3).

teristic absorbance peak, it is imperative that the melanin to bequantified be in a pure state in order to isolate it from naturalbackground sources. The isolation procedures used yield purified melanin, making it possible to determine absoluteamounts of melanin present in various tissues. General absorbance spectra for deproteinized and proteinized melanins andsynthetic DOPA melanin were studied. Protein concentrationsof samples isolated before and after incubation with pronasewere compared.

MATERIALS AND METHODS

Isolation of Melanin Pigment. Human melanomas and normal tissue samples were obtained in formalin; animal tissueswere obtained without any fixatives. The adipose tissue wastrimmed away, clean slices of each sample were taken, andexcess surface moisture was removed by a stream of nitrogengas before the sample was weighed. Tissue samples weighingfrom 50 to 150 mg were homogenized in a glass:Teflon ho-

mogenizer, with 5.0 ml of redistilled water. Samples were thencentrifuged at 10,000 x g for 20 min. Precipitates werewashed with 5.0 ml 0.05 M phosphate buffer (pH 6.5) to removesoluble materials. Precipitates obtained after centrifuging ho-

mogenates at 10,000 x g for 20 min were first extracted withchloroforrtrmethanol (2:1) once and then extracted withethanol:ether (3:1) twice to remove lipids. The residues werethen dried in a current of nitrogen gas and treated with 5.0 mlof lysing solution (0.3% saponin in 0.9% NaCI solution) for 30min to remove hemoproteins. Samples were centrifuged at10,000 x g for 20 min, and the precipitates were washedtwice with 7.0 ml 5 DIM MgCI2 in 0.15 M NaCI. The sampleswere then treated with 0.5% sodium dodecyl sulfate in 0.05 MTris buffer (pH 7.0) containing 2 mg/ml Pronase-CB (Calbi-ochem-Behring Corp., LaJolla, Calif.). Mixtures were incubatedat 30°for 48 hr. The digests were centrifuged at 10,000 x g

for 30 min. (Pronase digestion is done last, since deproteinizedmelanin, if allowed to dry, is almost completely insoluble.)Remaining residues, containing the pigment granules, werewashed once with 6.0 ml 0.9% NaCI solution to remove solublesubstances. The samples were then washed with redistilledH2O and transferred to polypropylene tubes. The pellets (and

Indole-5,6-Quitione2-CarboxylicAc/d

Indole-5,6-uuinoneIN *•

Chart 2. Hypothetical structure for melanin in Harding-Passey melanoma(from Ref. 12).

standards) were then dissolved in 0.1 N NaOH and autoclaved.Melanin isolation procedures followed those described by Dasef a/. (8) with minor variations. Final melanin residues weresolubilized in 0.1 N NaOH, as suggested by Lee ef a/. (15).

Preparation of Synthetic Melanin. Synthetic melanin wasprepared by autoxidation of L-/8-(3,4-dihydroxyphenyl)alanine

(Sigma Chemical Co., St. Louis, Mo.), following a techniquedescribed by Das ef al. (7). A solution of 100.0 mg of L-ß-(3,4-dihydroxyphenyl)alanine with copper sulfate (1 x 10~4M)was

made in 200 ml 0.05 M Tris buffer (pH 8.5). The mixture wasstirred for 24 hr at room temperature and then acidified to pH2 with concentrated HCI. The solution was centrifuged, and thepellet was washed twice with KCI solution (0.01 %). The precipitate was homogenized with redistilled water in a glass:Teflonhomogenizer, transferred into a dialysis bag and dialyzed for24 hr against double-distilled water (water was changed 3-5

times). The material was lyophilized to obtain the dry weight(55 mg). Melanin polymer was then resuspended in double-

distilled water to give a concentration of 1.0 mg/ml. This stocksuspension was stirred vigorously when pipeting for the standard curve.

Spectrophotometer. A Carl Zeiss Model PM Oil spectropho-tometer with a tungsten light source was used to measureabsorbance in the region between 310 and 750 nm. Absorbance at 400 nm was measured on the solubilized sample. Themelanin content was determined from the absorbance datausing standard solutions made from synthesized, autoxidizedDOPA melanin. Calibration was accomplished with suspensions of synthetic melanin ranging between 10 and 80/ig/ml;melanin concentrations <~2 /¿gwere below the limit of sensi

tivity. Background was subtracted by zeroing absorbance witha quartz cuvette containing 1 ml of 0.1 N NaOH and 0.5 ml ofdistilled water.

Animal Melanoma Models. Greene melanoma in Syriangolden hamsters has been carried in our laboratory for ~10years. The B-16 and KHDD in C57 and C3H mice, respectively,

were donated to us by Robert F. Kallman, Stanford UniversitySchool of Medicine, Stanford, Calif. Melanotic Harding-Passeyand amelanotic Harding-Passey melanoma models, carried in

468 CANCER RESEARCH VOL. 41

Research. on November 26, 2015. © 1981 American Association for Cancercancerres.aacrjournals.org Downloaded from

Melanin Content of Human and Animal Tissues

BALB/cJ mice, were obtained from Dr. Joseph Mayo, MasonResearch Institute, Worcester, Mass. The rabbit amelanoticmelanoma has also been maintained in our laboratory for ~10

years.Human Melanoma and Normal Tissues. Small samples of

human tissues were obtained from formalin-fixed postmortem

specimens at the Veterans Administration Hospital, Northport,N. Y., and North Shore University Hospital, Manhasset, N. Y.

RESULTS

Melanin Content. Results of melanin analysis are given inTable 1. Individual samples of Greene and Harding-Passey

melanotic melanomas contained concentrations between 0.3and 0.8% by weight. These values were similar to those ofmouse and hamster whole eye and to the average (0.35%)human melanotic melanoma values evaluated (range, ~0.1 to

0.9%). KHDD melanomas in C3H mice, previously reported tobe deeply pigmented (2),5 had a concentration less than

those of amelanotic Harding-Passey and rabbit melanomas(<0.02%). B-16 and amelanotic Harding-Passey melanomas,

with concentrations of 0.06 and 0.02%, respectively, wereintermediate between melanotic and amelanotic tissues. Melanin content in the various hamster tissues (liver, lung, blood,brain, heart, spleen, kidney, intestine) was low, at a factor of~50 less than pigmented melanoma concentration. Hamster

hair was highly pigmented (0.4% by weight), but pigmentationin the skin alone was relatively low (~0.01 %).

Samples of human liver and heart had pigment concentrations of ~0.01 % while brain (substantia nigra) and skin (bothswarthy and light) showed melanin contents between ~0.01

and 0.02% by weight.The sensitivity limit was ~2 /¿gmelanin per 100 mg wet

tissue. A known amount (45 ¡ig)of synthetic DOPA melaninwas added to 10 hamster kidney samples (~ 50 to 150 mg,

wet weight); recovery was 95.3 ±2.3% (S.D.). Error on multiple samples of melanoma homogenates was <8.7%.

Absorption Curves. Absorption curves for Greene melanoma, human melanoma, and mouse eye are shown in Chart3. Absorbance for melanin quantification was measured at 400nm, as suggested by Oikawa and Nakayasu (25). Syntheticmelanin and deproteinized animal and human melanin exhibitsimilar shapes, with a monotonically decreasing absorbancewith increasing wavelength, as has been observed previously(7, 25). Synthetic DOPA melanin added to hamster kidneytissue and isolated from this tissue showed an absorbancecurve identical to untreated DOPA melanin.

As an added check on the validity of our synthetic melanin,a commercially available sample of synthetic melanin preparedby persulfate oxidation of L-tyrosine was purchased (Sigma).

The absorbance curve of the latter was essentially identical tothat of our synthetic melanin (Chart 3, Curve 2). Absorbancesmeasured at 400 nm with equal concentrations of melanin werein agreement to within 9%.

Protein Concentration. Protein content [as determined bythe method of Lowry ef al. (16)] in deproteinized tissue sampleswere approximately 10 times less than the amount found inuntreated samples. Results are shown in Table 2. These data,in conjunction with Chart 3, Curve 7, demonstrate the necessity

Table 1

Melanin content in various tissues determined by spectrophotometrictechniques from isolated melanin pigment

HumanmelanomasPatient1Patient2Patient3Patient4Patient5Patient

6MetastasistoliverMetastasistolymphMetastasistoboneMetastasistospleenHuman

normal tissue (substantianigra)Patient7LiverPatient

10Patient11Patient

8Patient9SkinPatient

10(swarthy)Patient8(light)HeartPatient

11Patient9Patient10Patient

8KidneyPatient

11Patient12Animal

tissuesGreenemelanoma(hamster)Greenemelanoma hemogenates(hamster)Harding-Passey

(BALB/cmouse)B-16melanoma (C57mouse)KHDD

melanoma (C3Hmouse)Amelanoticmelanoma(Harding-Passey)AmelanoticrabbitmelanomaHamster

wholeeyeC3Hmouse wholeeyeHamster

tissuesLiverLungBloodBrainHeartSpleenKidneyIntestineMuscleHairSkin

andhairSkinNo.

ofsamples3411633222111111111122135253221143443422821112%bywt0.84

±0.09a0.40

±0.20.340.09<0.00250.41

±0.050.44±0.120.30±0.060.12±0.030.016

±0.0010.0190.0160.0130.0090.0230.0080.0070.0100.0040.013<0.0025<0.00250.34

±0.110.44±0.040.41±0.0050.68

±0.130.06±0.020.003±0.0010.02±0.0030.010.45

±0.110.40±0.02<0.0025<0.0025<0.0025<0.0025<0.0025<0.0025<0.0025<0.0025<0.00250.390.130.013

±0.002"

Mean ±S.D.

' R. F. Kallman, private communication.

for removal of proteins (the abosrbance for Curve 7 approached 0 following deproteinization).

DISCUSSION

Phenothiazines and other polycyclic aromatic moleculeshave been found to bind to melanin. Relative affinities havebeen listed by Potts (28). This characteristic has been utilizedin the case of iodine-labeled quinoline as a diagnostic agent

FEBRUARY 1981 469

Research. on November 26, 2015. © 1981 American Association for Cancercancerres.aacrjournals.org Downloaded from

K. P. Watts et al.

io

09

08

0.7

ujO 0.6

lÅ’ 05

04

03

02

320 350 400 450WAVELENGTH IN UNITS OF IO'9 m

5OO

Chart3. Absorbance curves for various melanin samples: Curve 1, C3Hmouse eye; Curve 2, synthetic DOPA melanin; Curve 3, Greene melanoma(before deproteinization); Curve 4, Green melanoma (after deproteinization);Curve 5. human melanoma (before deproteinization. Patient 1); Curve 6, humanmelanoma (after deproteinization, Patient 1); Curve?, human melanoma (beforedeproteinization, Patient 5).

for melanomas (26, 27). Another radiopharmaceutical in use is[nC]CPZ (18). [35S]CPZ has been found, both by us and by

others, to have 50% of its activity in melanomas attached tothe melanin granules (2). The possibility exists that CPZ, 7-

hydroxychlorpromazine, prochlorperazine, or a number ofother related compounds, some of which show greater affinityfor melanin than does CPZ (28), may be used (in the form of aborated analog) to transport boron to tumors. Subsequentirradiation with neutrons would then allow energy to be deposited via the 10B(rt,a)7Li reaction. The high-linear-energy transfer

radiations produce a low oxygen enhancement ratio, producea high relative biological effect, and provide optimal conditionsfor radiotherapy because of the short range in tissue (10 ftm).Such compounds are currently being developed (19, 20).

Both for diagnostic and therapeutic application of compounds with melanin affinity, differential and absolute uptakewill be critically dependent upon melanin concentration intumor and "nonpigmented" tissues. Evaluation and compari

son of animal experiments and projection to human use mustbe based upon absolute melanin concentration in both animaland human tissues such as given in Table 1. While somemeasurements have been made, no comprehensive study hasbeen reported. Available data vary widely, as shown in Table3, perhaps according to the technique used to assay melanin.

Melanin Content. Clinically, melanin content is evaluatedqualitatively as being "deeply," "heavily," or "lightly" pig-

mented or amelanotic (no visible pigmentation). Well-pig-

mented human melanomas were found to have a melaninconcentration varying between 0.1 and 0.9% (average,0.35%). Greene and pigmented Harding-Passey melanomamodels (-0.3 and 0.8%) had concentrations similar to those ofhuman ones, despite the fact that their rapid doubling time (~2days for Greene and ~4 days for Harding-Passey) would arguefor a lower content relative to the slower-growing human melanomas. Hamster and C3H mouse whole eye also had levelsanalogous to those of human tumors. The eye might serve asan easily obtainable and reproducible model for melanoma in

Table 2

Protein concentration betöreand after deproteinization determined by Lowrymethod

fig protein/fig melanin

Tissue Before After

Lymphmelanoma(Patient6)B-16Melanoma(Patient

1)Spleenmelanoma(Patient

6)Harding-PasseyLiver

melanoma(Patient6)1.8712.670.805.320.782.150.260.580.210.430.130.43

long-term drug uptake studies. If such studies were to use

rapidly growing experimental melanomas, drug concentrationsmight be less than analogous concentrations in slower-growing

human melanomas (2).For diagnostic drug applications, gross pigment determina

tions such as given in Table 1 are applicable. For therapeutictechniques such as chemotherapy, photoradiation therapy (9),and NCT, the presence of nonmalignant, nonmelanized connective tissues within analyzed samples will hinder accuratedetermination of pigment concentration in true tumor volume,the parameter which will affect therapeutic outcome. Connective tissues were present to a variable degree in human melanomas analyzed in Table 1. Histological sections were correlated with melanin content of adjacent tissue samples forPatients 1 to 6. While connective tissues were present (up to20% of area as determined by planimeter), this would notseverely influence gross melanin content to the extent of variation recorded in Table 1. Human melanoma samples oftenshowed large areas of necrosis (up to -90%), probably due in

part to therapeutic efforts. However, no correlation was foundbetween gross pigment content and percentage of viable tumor.

Amelanotic animal models (rabbit, Harding-Passey, KHDD)

had concentrations from 0.003 to 0.02%; i.e., absolute concentration was between 10 and 80 times less than what appeared visually to be well-pigmented tissues. This relative

difference is in agreement with that previously described for a"heavily" pigmented and amelanotic B-16 melanoma (32) but

in disagreement with concentrations given for hamster andhuman melanotic and amelanotic melanomas by Borovansky(5) (Table 3). His finding that melanin concentration in amelanotic tissues was 50% of that in pigmented melanoma issurprising in view of the above and perhaps misleading whenprojecting efficacy of melanin-dependent diagnostic and ther

apeutic modalities for amelanotic melanomas.KHDD melanoma was found to be amelanotic; this model

was obtained in its 57th passage in C3H mice and had clearlychanged character from its original, highly pigmented form[12% (Table 3)] used for investigations of melanin precursorsand drug uptake (2, 3). The value of 12% pigment by weight inTable 3 is perhaps high, in view of the value of ~2% by weight

(wet) for beef choroid (if choroid is accepted as being the mosthighly pigmented of mammalian tissues). Melanin content ofthe original spontaneous KHDD tumor was also given as twicethat of C3H mouse whole eye, which from Table 1 wouldproduce a value of 0.8% melanin, by weight, a value which in

470 CANCER RESEARCH VOL. 41

Research. on November 26, 2015. © 1981 American Association for Cancercancerres.aacrjournals.org Downloaded from

Melanin Content of Human and Animal Tissues

Table 3

Melanin content of various human and animal tissues obtained from theliterature

TissueBeef

choroidKHDDmelanoma3 (C3Hmice)Harding-Passey

(BALB/cmice)B-16(cells inculture)Harding-Passey

(Hmice)Hamstermelanoma,melanoticHamstermelanoma,amelanoticHuman

melanoma,melanoticHumanmelanoma,amelanoticHumansubstantianigraHumandarkhairCattle

choroidAdultratbrainHuman

heartHumanliverMelanin

content% bywt1.812.00.45-0.44.024.452.275.662.270.592.3210.830.02-0.030.02-0.080.02-0.09Tissue

wetor drywtWetWetWetWetDry

(?)DryDryDryDryDryDryDryWetWetWetRef.283332555555551311

Original spontaneous tumor.

conjunction with the tumor-volume-doubling time of -2 weekswould fully account for high, long-term drug uptake and accu

mulation of melanin precursors observed.B-16 melanoma at 0.06% was intermediate between well-

pigmented (0.1 to 0.9%) and amelanotic tissues (<0.01%).The value of 0.4% by weight in Table 3 would indicate that ithad previously been more deeply pigmented.

Skin and hair are known to be pigmented tissues, with thevariation in skin pigmentation being used in the forensic sciences to deduce race from skin fragments (11 ). The values of0.013% found for hamster and 0.008 to 0.02% for human skinshould be low enough to prevent problems of interference withdiagnostic and therapeutic experiments.

Substantial amounts of melanin are known to occur in thehuman substantia nigra and locus caeruleus and to vary withage (Table 3; Ref. 17). Although small laboratory animals arenot thought to contain melanin in their brains, -0.025% by

weight has been found in rats (13). No such concentrationswere found in hamsters. Human substantia nigra was found tohave 0.016% melanin by weight (Table 1).

Hamster brain and other "background" tissues (liver, lung,

blood, heart, spleen, kidney, intestine, and muscle) were notfound to have melanin in amounts above the limit of reliablesensitivity (i.e., <0.0025% by weight). This would presumablybe indicative of low melanin content in the same organs ofother mammals including humans. The report of an average of-0.05% by weight melanin granules in human liver and heart(1 ) (0.01 to 0.09%) would put the latter tissues in the "lightly"

pigmented category (0.1 to 0.01%) and could militate againstsuccessful diagnostic and therapeutic efforts concerned withtumors associated with these organs. Our work shows a factorof 5 less melanin (average, =0.01 %) in heart and liver, so thatthese organs have a pigment content between 10 and 90 timesless than human melanotic melanomas.

The reason for the above discrepancy is not clear. The workreported by Ambani era/. (1) is carefully done; melanin extraction is through an acid (HCI) digestion followed by weighing ofthe residual pellet. It is precisely because of the variation inabsolute values of melanin content reported in the literatureand summarized in part in Table 3 that the present work wasundertaken. In any case, it is thought that the relative values ofmelanin content in human heart, human liver, and melanomas,shown in Table 1, should be valid.

Absorbance Curves. Since the absorbance spectrum ofmelanin reveals no distinguishing peaks, isolating melanin inits pure state is a necessary requirement for its colorimetriequantification. Proteins and other interfering nonmelanic substances falsely increase the absorbance readings, even at 400nm. For example, a human melanoma (Patient 5) showed avery high absorbance at 400 nm, before proteins were removed. The solubilized proteins showed an unusual greenishcolor, the general absorbance spectrum of which greatly differed from synthetic DOPA melanin (Chart 3, Curve 7). Afterremoval of proteins, the unusual color disappeared, and absorbance approached zero. The slopes of absorbance curvesfrom natural melanins may still be distinguishable from those ofsynthetic melanin. Oikawa and Nakayasu (25) have suggestedthat this slope difference may be attributed to a difference inthe monomer unit and/or the relative proportion of monomerunits in natural and synthetic melanin.

It has been reported that the extinction coefficients for melanin isolated from 2 different sources, hair and melanoma,converge after proteins are removed. It is possible that astructural change may occur after deproteinization, as suggested by Swan and Waggott (35).

Previously, the method commonly used to isolate melaninfrom natural sources was acid hydrolysis, the advantage beingthat the product obtained is free from protein material. Theharsh HCI treatment produces a loss of CO2 and NH3 and alsointroduces the problem of melanoidin formation and precipitation.

The mild procedure described in this paper makes it possibleto isolate melanin from natural sources in what is presumed tobe its pure state and thereby accurately quantitate spectropho-

tometrically the absolute amounts of melanin present.General Comments. The frequency distribution of hematog-

enous métastaseshas been listed by Nathanson (22). Distantmétastasesare most often clinically observed in easily examined sites O.e., the lymph nodes and skin), whereas autopsyreveals a more ubiquitous distribution, with lung, liver, andlymph nodes being the most prominent sites. Of the tissuesshowing a high incidence of metastatic melanoma, heart (41 %)and liver (77%) have previously been reported to have relativelyhigh melanin content (-0.05% by weight). Average values

found in this work were a factor of 5 less for heart and liver(-0.01%). The difference in melanin concentration found between most tissues and melanotic melanoma (-30 to 50 times)

would appear to afford a good opportunity for the developmentof useful diagnostic and therapeutic techniques based on melanin-seeking agents. Studies of CPZ distribution in varioustissues of melanoma-bearing animals indicate absolute anddifferential concentrations adequate for therapy. The latterwork is the subject of a following paper.

A significant variation in melanin content is noted betweenmetastatic sites within the same patient (Table 1, Patient 6),although absolute concentrations are still well above that innormal tissue. This variation is well known in humans and isalso observed in animal melanoma models. Pigmentation maybe regained or stimulated through the introduction of chemicalssuch as 1-yo-D-arabinofuranosylcytosine or theophylline (15,34). Studies are under way in our laboratory to evaluate thepossibility of bringing lightly pigmented melanomas (0.01 to0.1% by weight) into the well-pigmented category (>0.1%).The significantly increased melanin concentration found in the

FEBRUARY 1981 471

Research. on November 26, 2015. © 1981 American Association for Cancercancerres.aacrjournals.org Downloaded from

K. P. Watts et al.

epidermis of dark-skinned individuals (11 ) might pose problems

for therapy; fortunately, the occurrence of melanomas in suchindividuals is rare.

Efforts to diagnose and treat metastatic brain lesions andocular melanomas will encounter interference from the highpigment concentration in the uveal tract and, to a lesser extent,the substantia nigra. It is possible that melanin precursors suchas phenylalanine or DOPA could be utilized to circumvent thisproblem (assuming that melanogenesis in the uveal tract isreduced relative to melanoma).

Melanin content in Syrian golden hamsters and C3H mousewhole eyes was similar to the average content found in well-

pigmented human melanomas. The possibility this exists thatsuch models could serve as stable reference points (for melanin-seeking drugs), which would not suffer changes in characteristics experienced by other tumor models when subjectedto multiple passages. Further, such genetically consistentstrains should provide a constant melanin content, easily measured in whole eyes, thus allowing a convenient reference pointfor intercomparison of experiments.

REFERENCES

1. Ambani. L. M.. Jhung, J. W., Edelstein, L. M., and Van Woert, M. H.Quantification of melanin in hepatic and cardiac lipofuscin. Experientia(Basel). 33. 296-298, 1977.

2. Blois, M. S. On chlorpromazine binding in vivo. J. Invest. Dermatol., 45:475-481. 1965.

3. Blois, M. S.. and Kallman, R. F. The incorporation of C'4 from 3,4-dihydroxy-phenylalanine-2'-C'" into the melanin of mouse melanomas. Cancer Res.,

24:863-868. 1964.4. Blois, M. S., Zahlan, A. B., and Maling. J. E. Electron spin resonance studies

on melanin. Biophys. J., 4: 471-490, 1964.5. Borovansky, J. Quantitative parameters of melanomas differentiation. Neo

plasma (Bratisl.), 25: 349-352, 1978.6. Brownell, G. L.. Zamenhof, R. G.. Murray, B. W., and Wellum, G. R. Boron

neutron capture therapy. Pg 205-222, Therapy in Nuclear Medicine. R. P.Spencer, (ed.), Gruñe& Stratton. 1978.

7. Das. K. C., Abramson, M. B., and Katzman. R. A new method for quantifyingmelanin. J. Neurochem.. 26: 695-699, 1976.

8. Das, K. C.. Abramson. M. B., and Katzman, R. Neuronal pigments: spectro-scopic characterization of human brain melanin. J. Neurochem.. 30: 601-605. 1978.

9. Dougherty, T. J., Kaufman. J. E., Goldfarb, A.. Weishaupt, K. R.. Boyle. D.,and Mittleman, A. Photoradiation therapy for the treatment of malignanttumors. Cancer Res., 38: 2628-2635, 1978.

10. Fairchild. R G.. and Goodman, L. J. Development and dosimetry of anepithermal neutron beam for possible use in neutron capture therapy. Phys.Med Biol., 11: 15-30. 1966

11. Goldschmidt, H., and Raymond. J. Z. Quantitative analysis of skin color frommelanin content of superficial skin cells. J. Forensic Sci., 17: 124-131,1972.

12. Hemple. K. Investigation on the structure and function of melanin in malignant

melanoma with 3H- and 14C-DOPA labeled at different positions. In: G.Della Porta and O. Muehlbock (eds.), Structure and Control of the Melano-cyte, pp. 162-174. Berlin: Springer-Verlag, 1966.

13. Kastin, A. B., Kuzemchak, B., Tompkins, R. G., Schally. A. V., and Miller, M.C., III. Melanin in the rat brain. Brain Res. Bull. 7: 567-572, 1976.

14. Kitano. Y., and Hu, F. Proliferation and differentiation of pigment cells invitro. J. Invest. Dermatol.. 55: 444-451, 1970.

15. Lee, T. H., Lee, M. S., and Lu, M. Effects of a-MSH on melanogenesis andtyrosinase of B-16 melanoma. Endocrinology, 91: 1180-1188. 1972.

16. Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. Proteinmeasurement with the Folin phenol reagent. J. Biol. Chem., 193: 265-275,1951.

17. Mann, D. A., and Yates, P. O. Lipoprotein pigments—their relationship toaging in the human nervous system. II. The melanin content of pigmentednerve cells. Brain, 97: 489-498, 1974.

18. Maziere, M., Sainte-Laudy, J. L., Crouzel, M., and Comar, D. Synthesis anddistribution kinetics of "C-chlorpromazine in animals. In: G. Subramanian

(ed.), Radiopharmaceuticals, Littleton, Mass.: PSG Publishing Co.. Inc.,1975.

19. Mishima, Y. Neutron capture treatment of malignant melanoma using IOB-

chlorpromazine compound. Pigm. Cell, 1: 215-221, 1973.20. Mishima, Y., and Shimakage, T. Thermal neutron capture treatment of

malignant melanoma using 10B-DOPA and "Bi2-chlorpromazine compounds.

Pigm. Cell, 2: 394-406. 1976.21. Nakagawa, T., and Aono, K. Synthesis of chlorpromazine undecahydrodo-

decaborate and nonahydrodecaborate—promising agents for neutron capture therapy of malignant melanoma. Chem. Pharm. Bull. (Tokyo), 24: 778-781, 1976.

22. Nathanson, L., Farber, S., and Hall, T. C. Biological aspects of humanmalignant melanoma. Cancer (Phila.), 20 650. 1967.

23. Nathanson, L., Hall, T. C., Vawter, G. F., and Farber, S. Melanoma as amedical problem. Arch. Intern. Med., ) 19: 479-492, 1967.

24. Niebauer. G. Dendritic cells of human skin. Exp. Biol. Med., 2: 1968.25. Oikawa, A., and Nakayasu. M. Quantitative measurement of melanin as

tyrosine equivalents and as weight of purified melanin. Yale J. Biol. Med..46: 500-507, 1973.

26. Packer, S.. Lambrecht, R. M., Christman, D. R., Ansari, A. N.. Wolf, A. P..and Atkins, H. L. Metal isotopes used as radioactive indicators of ocularmelanoma. Am. J. Ophthalmol., 83: 80-94, 1977.

27. Packer, S., Redvanly. C., Lambrecht, R. M., Wolf, A. P.. and Atkins, H. L.Melanoma detection with an iodine-123 labelled quinoline analog. Arch.Ophthalmol., 93. 504-508. 1975.

28. Potts, A. M. The reaction of uveal pigment in vitro with polycyclic compounds.Invest. Ophthalmol., 3: 405-416, 1964.

29. Prota. G., and Thompson, R. H. Melanin pigmentation in mammals. Endeavour (Oxf.). 35: 32-38, 1976.

30. Râper.H. S. Aerobic oxidases. Physiol. Rev., 8: 245-282, 1928.31. Riley, V. (ed.). Pigment Cell, Vols. 2 and 3. Basel: S. Karger AG, 1976.32. Rosenthal. M. H.. Kreider, J. W., and Shiman. R. Quantitative assay of

melanin in melanoma cells in culture and in tumors. Anal. Biochem., 56: 91 -99. 1973.

33. Shirakawa, S.. Luce, J. K., Tannock, J. K., and Frei, E., III. Cell proliferationin human melanoma. J. Clin. Invest., 49: 1188-1199, 1970.

34. Steinberg, M. L., and Whittaker. J. R. Stimulation of melanotic expression ina melanoma cell line by theophylline. J. Cell Physiol., 87 265-276, 1976.

35. Swan. G. A., and Waggott. A. Studies related to the chemistry of melanins.Quantitative assessment of different types of units present in DOPA-melanin.J. Chem. Soc. Chem. Commun., 1409-1418, 1970.

36. Whittaker, J. R. Absence of a direct melanin-protein relationship in chickembryo melanocytes. Biochim. Biophys. Acta, 583: 378-387. 1979.

472 CANCER RESEARCH VOL. 41

Research. on November 26, 2015. © 1981 American Association for Cancercancerres.aacrjournals.org Downloaded from

1981;41:467-472. Cancer Res   Karen P. Watts, Ralph G. Fairchild, Daniel N. Slatkin, et al.   Various MelanomasMelanin Content of Hamster Tissues, Human Tissues, and

  Updated version

  http://cancerres.aacrjournals.org/content/41/2/467

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  .permissions@aacr.orgDepartment at

To request permission to re-use all or part of this article, contact the AACR Publications

Research. on November 26, 2015. © 1981 American Association for Cancercancerres.aacrjournals.org Downloaded from