[CANCER RESEARCH 45, 4401-4411, September 1985]

Surface Antigens of Melanomas and Melanocytes Defined by Mouse

Monoclonal Antibodies: Specificity Analysis and Comparison ofAntigen Expression in Cultured Cells and Tissues1

Francisco X. Real,2 Alan N. Houghton, Anthony P. Albino, Carlos Cordon-Cardo, Myron R. Melamed,

Herbert F. Oettgen, and Lloyd J. Old

Laboratory of Human Cancer Immunology [F.X.R., A.N.H., A.P.A., H.F.O., L.J.O.] and Department of Pathology [C.C-C., M.R.M.], Memorial Sloan-Kettering Cancer Center,

New York, New York 10021

ABSTRACT

Mouse monoclonal antibodies (mAb) detecting 13 distinct systems of surface antigens on cultured melanocytes and melanomas were tested for reactivity with panels of (a) normal andmalignant cultured cells; (b) normal adult and fetal tissues; and(c) specimens of metastatic melanoma and other tumor types.The objectives of this study were to compare antigen expressionin cultured versus noncultured cells, to develop a panel of mAbsthat identify subsets of melanomas, and to provide requisiteinformation about antibody specificity in preparation for the useof antibodies in the diagnosis, imaging, and therapy of melanoma.Five of the melanoma surface antigens have been well characterized biochemically [GD3,chondroitin sulfate proteoglycan, HLAClass II antigens, glycoprotein of molecular weight 130,000(gp130), and glycoprotein M, 95,000/protein M, 97,000 (gp95/p97)]. Three antigens have been related to melanocyte differentiation (HLA Class II, M111/M231, and M144), and six provideadditional markers for subsets of cultured melanomas. mAb R24reacts with the disialoganglioside GD3,a predominant gangliosideon cultured melanoma cells and other cells of neuroectodermalorigin. A high proportion of melanoma, astrocytoma, and sarcoma tissue specimens were G03+. In normal tissues, reactivity

of mAb R24 was restricted to melanocytes, neuronal and glialcells in the central nervous system, parotid gland, adrenal medullary cells, and rare cells in the connective tissue. mAb B5detects a chondroitin sulfate proteoglycan that is expressed bymost melanoma and astrocytoma cell lines and by culturedmelanocytes. Most of the melanoma and astrocytoma specimenswere B5+, whereas other tumor types tested were B5~. mAb

13-17, which detects a monomorphic determinant of HLA Class

II antigens, reacted with melanomas, and with a variety of othercancers, but not with normal skin melanocytes. There is considerable variability in the expression of GD3 and HLA Class IIantigens in individual melanoma specimens; cotyping for thesetwo antigens showed no evidence for coordinate expression.mAb L101 detects gp130 and mAb L235 detects gp95, antigensthat are strongly expressed on a broad range of cultured celltypes. In contrast to their wide distribution on cultured cells,gp130 expression in tissues was generally restricted to a subsetof melanomas and some normal cells, and gp95 was detected

1This work was supported in part by grants from the National Cancer Institute(CA-08748, CA-34079) and the Oliver S. and Jennie R. Donaldson Charitable Trust,

Inc.2 Recipient of a Research Training Fellowship from the Ministerio de Educación

y Ciencia of Spain and a Clinical Scholar Award in BiomédicalResearch from theCharles H. Revson Foundation.

Received 2/25/85; revised 5/10/85; accepted 5/24/85.

on only a small number of melanomas. mAb M111/mAb M231and mAb M144 define intermediate and late stage differentiationmarkers of cultured melanocytes and melanomas. mAb M111/M231 were reactive with a subset of melanoma specimens; theantigen detected by mAb M144 could not be detected on anynormal or malignant tissue. The remaining antibodies detectantigens expressed by a wide range of normal and neoplastictissues and provide additional probes for melanoma markers.

INTRODUCTION

Mouse mAbs3 have defined a large number of new antigenic

systems on the suface of human melanoma cells (1-10). These

antigens are being used as markers for the study of benign andmalignant pigmented lesions (11-16) and as targets for imaging

(17) and therapy (17, 18). A consistent observation in thesestudies has been the diversity of antigenic phenotypes of melanomas derived from different individuals, or even from separatemétastasesarising in the same patient (13, 15, 19). To investigate the possibility that this diversity reflects a correspondingdiversity in the surface antigenic phenotype of normal cells in themelanocyte lineage, the pattern of surface antigens expressedby cultured fetal, newborn, and adult melanocytes was comparedwith the surface antigens expressed by a panel of melanoma celllines (9). Four categories of antigens could be distinguished: (a)antigens found on all cells of melanocyte origin, normal or malignant. These include broadly represented antigens that are notlineage-specific and antigens that are restricted to cells of neu

roectodermal origin; (b) antigens found in adult (but not newbornor fetal) melanocytes and on a subset of melanomas; (c) antigensfound on newborn and fetal (but not adult) melanocytes and ona subset of melanomas; and (d) antigens found on a subset ofmelanomas but not on fetal, newborn, or adult melanocytes. Onthe basis of these observations regarding surface antigenexpression, as well as other phenotypic traits of melanocytesand melanomas (morphology, pigmentation, and tyrosinase levels), a differentiation pathway of melanocytes has been proposed, and three general categories of melanomas can be recognized, based on their expression of early, intermediate, or latemarkers of melanocyte differentiation (9).

In the present study, 14 mAbs detecting cell surface antigensof cultured melanoma have been selected for detailed specificityanalysis.

3The abbreviations used are: mAb, monoclonal antibody; PBS, phosphate-

buffered saline; gp130, glycoprotein of molecular weight 130,000 (other glycopro-teins are similarly designated); p97, protein of molecular weight 97,000.

CANCER RESEARCH VOL. 45 SEPTEMBER 1985

4401

on July 23, 2021. © 1985 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

ANTIGENS OF MELANOCYTES AND MELANOMAS IN CULTURED CELLS AND TISSUES

MATERIALS AND METHODS

Cultured Cells. Cell lines and cultures of melanocytes and normalkidney epithelium were derived and propagated as described (9).

Monoclonal Antibodies. The procedures used for immunization, fusion of spleen cells with MOPC-21 NS/1 or SP/2 myeloma cells, hybri-doma cloning, and production of antibody-containing ascites or serum innu/nu mice have been described previously (4, 9). Hybridoma 13-17

was provided by Dr. H. Koprowski of the Wistar Institute, Philadelphia,

Serological Procedures. The anti-mouse immunoglobulin assay, pro

tein A assay, and absorption tests for the analysis of cell surface antigensof cultured cells were performed as reported previously (20-23). mAbs

(serially diluted ascites fluid or serum) were incubated with cells platedon Falcon 3040 Microtest II plates (Falcon Labware, Oxnard, CA) atroom temperature for 1 h. After washing three times with 5% 7-globulin-

free fetal calf serum in PBS, cells were incubated with a 0.2% suspensionof O-type RBC conjugated previously to protein A or anti-mouse immu

noglobulin. For absorption tests, 30 ^l of the appropriate dilution of theantibody were absorbed with a 30-Ml pellet of cells, and the resulting

supernatants were tested for residual antibody activity. Heat stability ofthe antigenic determinants was established by heating the cell suspension to 100°Cfor 5 min prior to use in absorption tests.

Immunoprecipitation Procedures. Labeling of cultured cells with[35S]methionine or [3H]glucosamine, cell extract preparation, and immu-

noprecipitation were carried out as described previously (4, 24, 25).Tissue Specimens. Adult normal and malignant tissues were obtained

at autopsy or from surgical specimens. Fetal tissues were obtained fromprostaglandin-induced abortions of 12-14-week-old fetuses. Tumor tis

sues were obtained from surgical specimens. Tissues were immediatelysnap-frozen in liquid nitrogen and stored at -70°C. Three-S-^m sections

of tissues were cut with a Bright cryostat (Bright, Huntingdon, England)and stored at -70°C.

Indirect Immunofluorescence Assay. Tissue sections were fixed in3.7% formaldehyde at room temperature for 10 min, washed three timeswith PBS, and incubated for 1 h in a humid chamber with undilutedhybridoma tissue culture supernatant, or ascitic fluid or serum fromhybridoma-bearing nu/nu mice (1:500-1:1000 in 1% fetal calf serum inPBS). The slides were washed with PBS and incubated for 30-60 min

with the appropriate dilution of a fluorescein isothiocyanate-labeled IgGfraction of goat anti-mouse Ig (Cappel Laboratories, Cochranville, PA).

After washing, the sections were mounted in buffered glycerol andexamined under a Leitz Dialux 20 fluorescence microscope with epi-illumination. Tumor cells were identified by phase-contrast examinationand by hematoxylin staining of adjacent sections. Counter-staining with

a solution of ethidium bromide (Sigma Chemical Co., St. Louis, MO) inPBS, 10 Mg/ml, was performed to aid in the evaluation of fluorescencelocalization.

Indirect Immunoperoxidase Assay. Tissue sections were fixed for30 min with 0.3% H2O2in methanol or 0.3% H2O2in 3.7% formaldehydein the case of mAb R24. The sections were washed with PBS andincubated at room temperature with 5% normal rabbit serum for 20 min.After removal of rabbit serum, sections were incubated with mAb asdescribed above. The sections were then incubated at room temperaturefor 30 min with the peroxidase-labeled IgG fraction of rabbit anti-mouse

immunoglobulin (Dako; Copenhagen, Denmark), washed, and exposedto a solution of diaminobenzidine (Sigma Chemical Co.) in 0.03% H2O2in PBS, 0.5 mg/ml, for 5 min. Sections were counterstained with hematoxylin.

Cultured supernatant from MOPC-21 NS/1 or serum from mice bearinghybridomas secreting non-reactive antibodies were used as specificity

controls.

RESULTS

Table 1 lists the characteristics of 14 mouse mAbs analyzedin this study and summarizes the pattern of reactivity withcultured melanocytes and melanomas and with sections of normal skin and melanomas. Table 2 provides the detailed resultsof tests on cultured cells, and Tables 3 and 4 provide detailedresults of tests on sections of normal adult and fetal tissues andmalignant tissues.

mAb R24(73). mAb R?4reacts with the disialogangliosideGD3,a prominent ganglioside of human melanomas (4, 26). OthermAbs detecting GD3have been described (27-29). In tests oncultured cells, cell lines derived from melanomas and astrocyto-

Table 1Cell surface antigens of melanomasand melanocytesdetected with mousemonoclonalantibodies: expression in cultured cells and tissues

MonoclonalantibodyRMB513-17L101L235M111M231M144L230L127A123A127M138M68Anti

bodyclass/sub

classigG3igG2.IgG,igG2.IgG,IgG,IgMIgG,IgG,IgG,IgG,igGj.IgG,igG«,ImmunizingcellsSK-MEL-28SK-MEL-93Human

melanoma/mouseIT22hybridSK-MEL-33SK-MEL-33Cutaneous

melanocytesCutaneousmelanocytesCutaneousmelanocytesSK-MEL-33SK-MEL-33SK-MEL-19SK-MEL-19Cutaneous

melanocytesCutaneousmelanocytesAntigen

detectedGanglioside,

GraProteoglycan,

500/280gp34/gp28gp130gp95gp110gp1

50/gp90gp45ReferenceDippoldefa/.,

1980Pukelefa/.,1982Houghton

et al.,1982Koprowskiétal.,1978Lloyd

er a/.,1981Houghtonet al.,1982Albino

eia/.,1983Houghtonei al.,1982Houghtonétal.,1982Houghtonétal.,1982Houghtonef a/.,1982Houghtonef a/.,1982Houghtonef a/.,1982Houghtonef a/.,1982Houghtonef a/.,1982Houghtonet al.,1982Houghtonétal.,1982Cultured

cellsMelanocytesNew-

Melano-bornAdultmas+

+16/16+

+21/2213/21+

+22/26+

+23/29+-27/33+-25/33+

30/33++20/20++7/33++32/33++30/33++31/32++ 29/32Mela

nocytes+a——-—————AaB•TissuesMelanomas58/6117/2263/6336/627/6324/5119/310/821/3838/4312/1312/1327/3112/14Hetero

geneityin antigen expres

sion"'c3+2+3+1+1+1+1+1+3+1+000

a Reactivity with keratinocytes does not allow evaluationof antigen expression by melanocytes.0 Heterogeneity refers to variability in antigen expression in the cells of individualtumors.c 0, uniform expression; 1+, 2+, and 3+, increasingdegrees of heterogeneity.

CANCER RESEARCH VOL. 45 SEPTEMBER 1985

4402

on July 23, 2021. © 1985 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

ANTIGENS OF MELANOCYTES AND MELANOMAS IN CULTURED CELLS AND TISSUES

11I1E«1Cicoì1•Qi81¿"3o0coiIi11uS11COC1ii'Sè-A*t11colo^12COcoiÎ5Ccoci.eulìoco51_^2coOJS,.iJSaoS***•••••*••*»**•*»••

N C»•N••:;•••••••33«00**0•S««S*3*S«•••••••••••3«3«333O3*«3S«ÒS33«S30330000333033*33333030*3030S33303Ò33S3S333*S3OOOOOOÒ3OOOOOO3O3SOOOÒ•OOO*OOOOO3••

•••••

••»•••

•••••••333S303S3333333S3333O333333303333•333303SOS3•OOSS3O3333333OS3O3OO3••••••30O03•O0000S93««•3*OS3S*0OO3«*3«3

«30»»»*>)•O

•Ò3««*«Ò3•••

••••b••»:»•••••

A fll «k A afekW V W V<9••••

•••••o ••••

••••••••r-.

ä-^T^l'-cS

'- '- '- '-S ^ ^S?ssOVOTCOCOOâ„¢^-T-ST^-

.CNICOtOCD .^' ''co"

co oo T-" r»."in co co CMior-T—CVJCM CO CO 1^»O) CD T- *—T—

r- ni yi ni in IH ni ni in uj ni¡il055555225555§¿¿¿¿¿¿¿¿¿¿¿

-Qicococococococncotococo5•

S*

•0

0S*•••

Sss3303b«3«330«O3•

3•

••••

3SS333SOOOOoooooooobbbbbbb«•

•00o00oë0?V

egAstrocytomasSK-MG-1,-3,SK-MG-11 ,-1.•9

0«•**0•O«a333O•••••••Oa33Oaoob•bba•*

*WWZCO2co0zNeuroblastomas

SK-N-SH,SKSK-N-BE(2)RenalcancerO»

SO3

••

O3•

•S••3•

•SS3

S03

000

003

S33

333

SOO

OOO

OOO

O•

•3•

•••

••33

SOÃ’

33O

3OO

OOO

3OO

O3O

3O3

•OO

•bo

bbbbo«bo»

bbbboo

b0oo

b00

>-<*S<?«-sSK-RC-7,-8,SK-RC-11 ,-2 Bladdercancer6637,RT4,S.•0«S*OS•oSo»o••••9OOOso•••9bb9b•oofmi_J

ZI9Lung

cancer

SK-LC-6,SK-Breastcancer•

XX•3

• .xxx•

• •xxb

x xx0S3 xxxx•

S b xxxb

xSS3 x . x.SSO

xxx3

.3

3 3 ... .x33

O . X.3

...OO

3 x ..•O

3 . ..•••b

x33• xxxx3

0 O xxx.O

O x . .xO0

....b

x*• XXX .X•

• b xx.xbb

xx*•*ìGOO

x9aOO

»x *lft*

**00

.008I

- "P5*1

tesf,to" m - S——p5£'o "-S'a.«

gSocVSNOO l f o f~S §SS<°5 8«>'= ="IfhmmiJ]<

aJ-sF; g-lON °-¿Ecu-0<Ifcfcf.«nof,Wea-3OH

J Õ Z CE.

.0O0S«x

3x,SSxx

b«0.

.0000o.

.30.

.S3.

.S.

. SS039•

3SOSx

SO.

.O.

.OOoob.

.oo.

. oo0oo03bO

Awoo2

Iis

èCD

O J2 "*. P -S 2< CD^ to D ij>ai

co T| . . ~mTO«fofSa

Ifzz z••

•••o•3«SSS

03333333•0*3bb•••

••30033

OOOOOOOOOOObooob

b•••

f~\AWWbCM

O?

OLU^

| 5m .2<DLL

u. Ll-CO8s"x"1ìIIs11I«B1bi1i1§•0fiB—^o">:fjI§|W

>«I?1

=1IIIm

~C-9-COt¿

«E'-"-!?Ì|¿II•i

r:cá3'i»»"SST

àiE7>c5o~^s5l=.!S5•-ifTSVo

a.tua»',titer<1 notdone,absor6Resultsof1

CANCER RESEARCH VOL. 45 SEPTEMBER 1985

4403

on July 23, 2021. © 1985 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

ANTIGENS OF MELANOCYTES AND MELANOMAS IN CULTURED CELLS AND TISSUES

Table 3Reactivity of mouse monoclonal antibodies with normal adult and fetal tissues: indirect immunofluorescenceand immunoperoxidaseassays

MonoclonalantibodiesO

O i-',;r\24s--'"1SkinKeratinocytesMelanocytesSweat

glandsSebaceousglandsBrainCortexWhite

matterLungAlveolar

epitheliumLiverHepatocytesSinusoidsEsophagusEpitheliumColonEpitheliumBreastDuctsAcini

TestesSeminiferoustubulesOvaryFolliclesUterusExocervixEndocervixEndometriumPlacentaTrophoblastLymph

nodePancreasExocrineEndocrineKidney

GlomerulusProximaltubule

Distal tubuleAdrenalglandCortexMedullaThyroidEpitheliumColloidSpleenHeartConnective

tissueCapillaryendotheliumSmoothmuscleSkeletal

muscleF0eOOOOooooooooooooeoo

oo•ooeoo0AO•Ooeoooooo0oooooooeoooo•ooooeoo0FOooooooooooooooooooooooooeo0Aeaooooooooooooooooooooooooooooeo013-

17"FOOoooo

ooooAOOooooo•oooooooo•oo•oooooo•oe•ooL101Foooooooooooooooooo•oooooo••0Ao0••oooWo•oooooooooooooooooooo••0L235FooooooooooAoooooooooooooooooooooooooooooo0M111Fooooooooo•oooo•ooooooooooo0Aoooooooo••ooo•ooo•oooooooooooo0M231F•ooooooeooAoooooooo•eoo•ooo•ooooooooooooo0M144

L230FoooooooooA

FooooooO

üoooo

eo•ooooooo

oooo

•o•

o•ooo0oo0

0ooo0Aoooeoooooeooooo•ooooo

ooooooe•ooL127F

Aoooo•oo

oooo•

•••ooooooo

ooo•

0ooooooo

o0••oA123F

A•d•oooo•o0

0•oooo••••

*0••o•o•0o

e000oA127F

A•d•oooo•o•

o••ooo•0••

•0•0o•o•0o

e0O0oM138F

A•d••eooo••

•••oooo••0

•0•ooo0o•o

•o••oM68F

A•d••ooo•••

••0oo9oo•e

•0•oo•0•o0

•0••0

* mAb R24reacts with the parotid gland." mAb 13-17 reacts with Langerhanscells in the epidermis.c •,positive reaction; »,weak reaction; O, no reactivity.d Reactivity with keratinocytes does not allow evaluationof antigen expression by melanocytes.

mas and cultures of normal melanocytes were the main cell typesshowing cell surface reactivity with mAb R24(4, 9). In sectionsof normal adult tissues, reactivity with mAb R24was detected innormal melanocytes in the skin, the epithelium of the parotidgland, occasional cells in islets of Langerhans, adrenal medullarycells, and rare isolated cells in the connective tissue of thepapillary dermis and other organs (Table 3). Skin melanocytesshowed a fine granular pattern of membrane and cytoplasmicreactivity (Fig. 1). In a detailed study of mAb R24reactivity withtissues of the central nervous system after acetone fixation, GD3was detected in the neuropil of many neuronal groups in thebrain stem, spinal cord, cerebellum, retina, and the dentate gyrusof hippocampus and with ependymal cells and glial processes

(30). In sections of tissues derived from a 12-week-old fetus,mAb R24reactivity was detected in clusters of cells in the dermisand in cells in the brain, islets of Langerhans, adrenal medulla,and some cells in the connective tissue. In tests on metastaticmelanomas, 58 of 61 specimens reacted with mAb R24(Fig. 1).There was considerable variation in the percentage of tumorcells that stained (Chart 1). When only a small proportion of cellswas reactive in a given specimen, positive cells were usuallyfound in clusters. Fig. 1 shows the reactivity of mAb R24with acompound nevus and two specimens of metastatic melanoma.Single melanoma cells and clusters of melanoma cells reactivewith mAb R24 could be detected in lymph node specimenscontaining deposits of metastatic melanoma (Fig. 2). Cytoplasmic

CANCER RESEARCH VOL. 45 SEPTEMBER 1985

4404

on July 23, 2021. © 1985 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

ANTIGENS OF MELANOCYTES AND MELANOMAS IN CULTURED CELLS AND TISSUES

Table 4Reactivity of mouse monoclonalantibodies with tumor specimens: indirect immunotluorescenceand immunoperoxidaseassays

Monoclonalantibodies (+/n)a

B5 13-17 L101 L235 M111 M231 L230 L127 A123 A127 M138

Number of positive specimens/numberof specimenstested.

M68

MelanomaAstrocytomaLung

carcinomaOvariancarcinomaColoncarcinomaBreastcarcinomaRenalcarcinomaBladdercarcinomaEmbryonalcarcinomaChronic

lymphocytic leukemia/well-differentiated lymphocyticlymphomaSarcoma58/614/43/130/60/92/80/120/62/30/57/1517/222/20/70/60/40/30/30/163/631/84/61/40/30/70/13/34/436/620/20/80/60/40/30/70/10/30/17/630/80/60/40/30/70/10/30/424/510/24/72/63/40/30/31/119/314/72/63/40/30/31/121/382/22/73/61/40/31/30/138/435/75/62/41/32/212/132/21/11/11/32/212/132/21/11/11/32/227/315/76/64/42/32/31/112/142/21/11/12/32/21/1

lOO-90

-8O

-7O

-60

-50

-40

-3O

-2O

-10

-....

...••.........

V.V.V.v.v..v.v.

::::.....nit»it

tit:...

....

R24 13-17Chart 1. Variation in the proportion of melanoma cells reactive with mAb R24

(Goalor mAb 13-17 (HLA Class II antigens) in metastatic melanomas. Eachpointrepresents an individualspecimen.

fluorescence was the predominant staining pattern found inmelanoma cells, although a membrane pattern was also seen insome cases. In all melanoma specimens, reactions with mAb R24showed a distinctive granular pattern, ranging from very fine tocoarse. In the latter case, vesicular membrane cytoplasmic structures were evident. Seven of eight melanoma métastasesto livershowed coarse vesicular staining. This pattern was seen lessoften in skin or lymph node métastases.Eighty-two specimens

of other tumor types were tested (Table 4). Of these, 4 of 4astrocytomas were reactive with mAb R24, as well as 3 of 13lung carcinomas, 2 of 8 breast carcinomas, 2 of 3 embryonal cellcarcinomas, and 7 of 15 sarcomas (three embryonal rhabdomy-

osarcomas, two chondrosarcomas, one liposarcoma, and onespindle-cell sarcoma) (Fig. 3). Three Kaposi's sarcomas, two

spindle-cell sarcomas, one leiomyosarcoma, one chondrosar-

coma, and one hemangiopericytoma showed no reactivity withmAb R24.

mAb B5 (72a). mAb B5 immunoprecipitated two componentsof M, 280,000 and >500,000 from extracts of [^SJmethionine-and [3H]glucosamine-labeled SK-MEL-'93 melanoma cells. Other

investigators have described mAbs generated against melanomacells that immunoprecipitate high-molecular weight componentssimilar to those detected by mAb B5 (3, 31 -33). By sequential

immunoprecipitation, it was determined that mAb 9.2.27 (7, 32),mAb 225.28 (31), and mAb B5 react with the same components.In tests with cultured cells, mAb B5 reacted with 21 of 22melanoma cell lines, 8 of 26 astrocytoma cell lines, and 5 of 24

carcinoma cell lines (Table 2).4 Cultured fetal and adult melano-cytes were B5+. In sections of normal adult tissues, B5 reacted

weakly with the basal layer of keratinocytes in the skin andshowed a cell membrane pattern of staining. This reactivity ofepidermal cells made it difficult to assess the staining of mela-

nocytes (Fig. 1). Reactions were occasionally seen with capillaryendothelial cells in normal and malignant tissues (Table 3). Infetal tissues, some pneumocytes and endothelial cells reactedwith mAb B5. Seventeen of 22 metastatic melanomas and 2 of2 astrocytomas were reactive with mAb B5; no other tumortypes were found to be B5+ (Table 4). In melanoma specimens,

positive cells stained with a membrane pattern.mAb 13-17 (7!). This antibody reacts with a monomorphic

determinant of HLA Class II antigens (1,34). Cultured melanomacells (9,14, 35, 36) and specimens of metastatic melanoma (11,12, 16, 37, 38) are known to express HLA Class II antigens,whereas cultured or noncultured melanocytes show no detectable Class II antigen expression (9, 11, 12). However, culturedmelanocytes can be induced to express HLA Class II antigen by7-interferon (39). Table 3 summarizes the reactivity of mAb 13-17 with normal adult and fetal tissues. mAb 13-17 stained

Langerhans cells in the epidermis, cells lining the liver sinusoids,lymphocytes in the follicular areas of lymph nodes and in thespleen, occasional cells in the connective tissue of most organs,and endothelial cells of capillaries (including the glomerular tuft)and arterioles (but not of larger vessels). Melanocytes in normalskin were unreactive with mAb 13-17. These results are con

sistent with the findings of Natali ef al. (40). All 63 metastaticmelanoma specimens reacted with mAb 13-17 with a cell mem

brane pattern. However, there was wide variability in the proportion of tumor cells staining in different specimens (Chart 1) andin the intensity of staining of individual cells within a single tumor(Fig. 4). The proportion of HLA Class II tumor cells showed norelation to the intensity of staining. In many cases, reactive tumorcells tended to cluster around capillaries (Fig. 4). No correlationwas observed between the percentage of R2/ and HLA Class*

cells in a given specimen (Chart 2). Regarding other tumor types,HLA Class II antigen expression was common in sarcomas (4 of4), chronic lymphocytic leukemia (3 of 3), and ovarian carcinomas(4 of 6) and was rare or absent in lung carcinomas (1 of 8), coloncarcinomas (1 of 4), breast carcinomas, and renal carcinomas(Table 4).

4F. X. Real, A. N. Houghton, A. P. Albino, C. Cordon-Cardo, M. R. Melamed,H. F. Oettgen, and L. J. Old, unpublishedobservations.

CANCER RESEARCH VOL. 45 SEPTEMBER 1985

4405

on July 23, 2021. © 1985 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

ANTIGENS OF MELANOCYTES AND MELANOMAS IN CULTURED CELLS AND TISSUES

Fig. 4. Immunofluorescence analysis of the reactivity of mAb 13-17 with metastatic melanoma, a, melanoma showing clustering of HLA Class ir cells around a Woodvessel; b, melanoma with a low proportion of Class II* cells; c, melanoma with a high proportion of Class ir cells. Original magnification, x 400.

100908070605040

302010«

-••

• • V•«•

\•-

• ' l '...~

...

« » ~I

•-t',

t » , ,

u

o

5? 10 20 30 40 50 60 70 80 90 lOO

% of celte positive with 13-17 antibody

Chart 2. Lack of correlation between the percentage of R2«+cells and HLA

Class ir cells in specimens of metastatic melanoma {points indicate individualspecimens tested for R24and HLA Class II by indirect immunofluorescence).

mAb L101 (-y1). mAb L101 reacts with gp130 (4,9). A number

of other mAbs to gp130 have been generated (41), and thesehave permitted the definition of at least four distinct epitopes onthe molecule (42). In tests on cultured cells, gp130 has beendetected on the surface of most melanomas and a proportion ofastrocytomas, carcinomas, and lymphoblastoid cell lines (Table2). Cultured newborn and adult melanocytes were gp130+. In

normal adult tissues, strong reactivity with mAb L101 was seenin sweat glands and sebaceous glands, in endothelial cells of allblood vessels, and in smooth muscle of the gastrointestinal tract.Weaker reactions were observed with liver sinusoidal cells andglomerular endothelium. Normal skin melanocytes were unreac-

tive (Table 3). In fetal tissues, the alveolar epithelium of the lung,the smooth muscle of stomach and small bowel, and vascularendothelial cells were reactive with this antibody, and the glomerular tuft of the kidney was weakly positive. mAb L101 reactedwith 36 of 62 metastatic melanomas, showing a discrete cellmembrane pattern. Within a given melanoma specimen, reactivitywas homogeneous, and most tumor cells showed a similarintensity of staining (Fig. 1). Eight of nine visceral (lung and liver)melanoma métastases were reactive, in contrast to 12 of 21lymph node métastases and 14 of 26 cutaneous métastases.Other tumor types did not react with mAb L101 (Table 4).

mAb L235 (f1). mAb L235 detects a Mr 95,000 glycoproteinthat is broadly expressed on cultured cells (Table 2) (4). mAbsto the same component have been described (43-45), and an

exchange of reagents has shown that the available mAbs togp95/p97 detect three different epitopes on this molecule (45).Amino acid sequencing has shown that gp95 is related to trans-

ferrin and lactoferrin (46). Cultured newborn and adult melanocytes were gp95+ (9). mAb L235 did not react with any normal

adult or fetal tissue tested (Table 3). In tests on melanoma tissue,7 of 63 specimens of metastatic melanoma were reactive, showing a cell membrane pattern. The intensity of the staining wasweak; only one of the positive specimens showed heterogeneityin antigen expression. Other tumors were not reactive with thisantibody (Table 4). These results contrast with the findings ofother investigators (47) (see "Discussion").

mAb M111 (71), mAb M231 (/n). Immunoprecipitation of [3H]-

glucosamine-labeled extracts of SK-MEL-28 cells with mAb

M111 showed a band of molecular weight 110,000. mAb M111reacts with fetal and newborn melanocytes but not with adultmelanocytes (9). This antibody defined a subset of culturedmelanomas that have characteristics of intermediate stage precursors in the melanocyte differentiation pathway (9). mAb M111also reacts with a proportion of cell lines of other tumor types(Table 2). In normal adult tissues, mAb M111 reacted with thealveolar epithelium of the lung, the apical membrane of colonepithelium, the squamous epithelia of the esophagus, and theexocervix and with normal lymph nodes. No reactivity wasdetected with normal skin melanocytes. In fetal tissues, colonieepithelial cells and splenocytes were reactive (Table 3). mAbM111 reacted with 24 of 51 metastatic melanomas with a cellmembrane pattern. There was only slight heterogeneity in antigen expression. This antibody also reacted with a proportion ofother tumor types, including carcinomas of the lung, ovary, andcolon (Table 4). In well-differentiated colon carcinomas, the apical

surface of the tumor cells was strongly reactive.mAb 231 did not immunoprecipitate detectable components

from radiolabled SK-MEL-28 cell extracts; however, cotyping of

the cultured cell panel with mAb M111 and mAb M231 showeda general correspondence between the reactivities of the twoantibodies, with mAb M231 giving weaker reactions. Similarpatterns of reactivity with normal adult and fetal tissues werealso observed. Cotyping of a panel of melanomas and other

CANCER RESEARCH VOL. 45 SEPTEMBER 1985

4406

on July 23, 2021. © 1985 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

ANTIGENS OF MELANOCYTES AND MELANOMAS IN CULTURED CELLS AND TISSUES

tumors showed only minor discrepancies in the M111 and M231phenotypes. For these reasons, we think it is likely that mAbM111 and mAb M231 detect identical or closely related molecules.

mAb M144 (-y1). The antigen detected by mAb M144 couldnot be immunoprecipitated from [35S]methionine- or [3H]gluco-

samine-labeled cell extracts of SK-MEL-127. mAb M144 reacts

strongly with cultured adult melanocytes, but not with fetal ornewborn melanocytes (9). A subset of cultured melanomas wasfound to be M144+, and these have the characteristics of adult

melanocytes by morphology, tyrosinase levels, and pigmentation. Table 2 shows that mAb M144 also reacts with a variety ofother cultured cell types. mAb M144 did not react with normaladult tissues (Table 3) or with eight specimens of metastaticmelanoma (Table 4).

mAb L230 (7!). Immunoprecipitation analysis of [3H]glucosa-

mine-labeled extracts of SK-MEL-19 cells with mAb L230 re

vealed two components of molecular weight 150,000 and90,000. When tested on a panel of cell lines, mAb L230 reactedwith all 20 melanomas tested, as well as with all other culturedcell types (Table 2), including newborn and adult melanocytes(9). In normal tissues, reactivity was found with sebaceousglands, alveolar epithelium of the lung, colon epithelium, troph-

oblast, the connective tissue of several organs, and blood vesselendothelial cells (Table 3). No staining of skin melanocytes wasseen. mAb L230 reacted with 21 of 38 specimens of melanomaand also with some carcinomas of the lung, colon, ovary, andkidney with a cell membrane pattern (Table 4). The staining oftumor cells was homogeneous for a given specimen.

mAb L127 (7!). mAb L127 did not precipitate antigen from[35S]methionine- and [3H]glucosamine-labeled melanoma cell ex

tracts. L127 expression appears to decrease with time in culture.Three of four melanoma cultures at early passage (<8) wereL127+, whereas only 6 of 33 melanoma cell lines at later pas

sages were reactive with mAb L127. All neuroblastoma lines,most astrocytoma lines, and some cell lines of epithelial originwere L127+ (Table 2). Cultured newborn and adult melanocyteswere L127~ (9). In normal tissues, reactivity was seen with colon

and breast epithelial cells, glomerular tuft, endothelial cells, andsmooth muscle of the blood vessel wall and the gastrointestinaltract. In fetal tissues, there was also reactivity with epithelialcells, capillary endothelial cells, and smooth muscle cells (Table3). In tumor tissues, reactivity with mAb L127 was detected in38 of 43 melanomas and in most other tumor types (Table 4).L127+ tumor cells were seen predominantly around the blood

vessels.mAb A123, mAb A127, mAb M138, and mAb M68. mAb

A123 and mAb A127 reacted with the same or related molecules,as indicated by cotyping panels of cultured cells and tissues(Tables 2-4). Immunoprecipitation analysis of [3H]glucosamine-

labeled cell extracts of SK-MEL-19 cells with mAb A127 showed

a band of molecular weight 45,000; mAb A123 had no detectableimmunoprecipitating activity. mAb M138 reacted with a heat-labile antigenic determinant and mAb M68 reacted with a heat-

stable antigenic determinant. Neither of these mAbs precipitatedantigen from labeled cell extracts. These mAbs showed verywide reactivity when tested on tumor cell lines (Table 2) and onnormal (Table 3) and neoplastic tissues (Table 4).

Antigenic Subsets of Melanoma. Specimens of metastaticmelanoma from 16 individuals were typed with the following

mAbs: B5; L101 ; L235; M111 ; and L230. Nine distinct compositeantigenic phenotypes could be distinguished. No coordinateexpression of any of these antigens could be detected.

DISCUSSION

The tissue reactivity of mAb R24shows that the GD3gangliosideis expressed by most cells of neuroectodermal origin. The neu-roectoderm is involved in the generation of neuronal cells in thecentral and peripheral nervous system, supportive cells of thenervous system, pigment cells, endocrine and paraendocrinecells (including the adrenomedullary cells, the paraganglia, theC-cells of the thyroid, and cells in the carotid body), and cells of

the mesectoderm (48). Cells derived from the mesectoderminclude the bone and cartilage of the facial skeleton, connectivetissue of selected areas of the head and neck, and some striatedmuscles in the facial and visceral regions (48). In tests on frozensections of normal tissues of the nervous system, mAb R24hasbeen shown to react with neuronal and glial cells (30). Using alarge panel of other normal tissues, we have shown that cutaneous melanocytes, adrenomedullary cells, and cells in the isletsof the pancreas and in the connective tissue of different organsreact with mAb R24.Melanomas and astrocytomas, two tumorsof neuroectodermal derivation, were also found to be regularlyreactive with mAb R24.It is possible that the embryonal rhabdo-myosarcomas reactive with mAb R24are also of neuroectodermalorigin, and the reactivity with embryonal carcinomas is not surprising, since these cells are multipotential in their differentiationcapacity. Whether the exceptional carcinomas that react withmAb R24belong to a small subset of tumors with distinct embry-ological derivation or characteristic histological type awaits further study. From the serological analysis of cultured cells, surfaceexpression of GD3appears to be restricted to cells of neuroectodermal origin, e.g., melanoma and astrocytoma. However, GDscan be detected in small amounts in extracts of normal ormalignant human cells of non-neuroectodermal origin (26), celltypes which do not show any cell surface expression of GD3.Inthese cells, GD3is probably an intermediate step in the syntheticpathway of other glycolipids (26, 49), whereas in cells of neuroectodermal origin, it is a final biosynthetic product the accumulation of which provides a marker for cells of this lineage.Immunofluorescence and immunoperoxidase studies of melanoma specimens indicate both cell surface and intracellular localization of Gas, membrane staining and fine-to-coarse cyto-plasmic granular staining were observed. In certain specimens,the pattern of staining suggested the Golgi complex as the majorcytoplasmic structure containing GD3.Confirmation of this awaitsimmunoelectron microscopic studies which are presently underway.

mAb B5 detects a chondroitin sulfate proteoglycan with arestricted distribution in cultured and noncultured cells. Reactivityof mAb B5 is generally limited to cells of neuroectodermal origin(melanomas, nevi, astrocytomas, and cultured melanocytes).With the exception of endothelial cells and basal keratinocytes,no other normal or malignant tissue was found to react withmAb B5. On the basis of molecular weight and antigen distribution, the antigen detected by mAb B5 is related to the antigensdetected by mAbs 225.28 (31), 9.2.27 (7, 32), and 48.7 (33).Harper ef al. (50) have shown that mAb 9.2.27 detects a coreglycoprotein of molecular weight 250,000-280,000 and a proteoglycan of molecular weight >500,000. This antigen has beencharacterized more precisely as an A/-sialylated glycoprotein and

CANCER RESEARCH VOL. 45 SEPTEMBER 1985

4407

on July 23, 2021. © 1985 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

ANTIGENS OF MELANOCYTES AND MELANOMAS IN CULTURED CELLS AND TISSUES

a proteoglycan of the chondroitin sulfate type. The proteoglycanis in close relationship to the extracellular matrix (32, 50).

Another antigenic marker of melanoma that has received considerable attention is HLA Class II (35, 36). Class II antigens arefound on 70% of cultured melanomas (9) and on a high proportionof melanoma specimens (11-13, 15, 16). In our series, all mela

noma specimens show some degree of HLA Class II antigenexpression ranging from 5-100%. In contrast to its expression

of melanomas, HLA Class II antigens are not expressed bynormal skin melanocytes or by cultured melanocytes (9). In thestudy of a panel of melanoma cell lines (9), high HLA Class IIantigen expression showed an association with phenotypic traitsconsidered to characterize early stages of melanocyte differentiation, I.e., epithelial morphology, low tyrosinase activity, nopigmentation, and presence of surface antigens expressed byfetal or newborn melanocytes but not adult melanocytes. Oneinterpretation of these data is that HLA Class II is a normaldifferentiation antigen in the melanocyte lineage but is expressedonly by the melanocyte progenitor and is lost during subsequentstages of melanocyte differentiation. According to this view,melanomas expressing HLA Class II antigens are derived fromHLA Class ll+ normal progenitors. Validation of this view awaitsthe isolation of HLA Class ll+ precursor cells in the melanocyte

lineage. Another possibility is that Class II antigen expression inmelanoma is a transformation-related trait, indicating alterationsin the structure and/or control of Class II genes during orfollowing tumorigenesis. A precedent for this possibility in experimental systems is the appearance of TL+ leukemias in strains

that do not normally express this antigen (51). Primary melanomas show lower HLA Class II expression than metastatic melanomas (15, 52). In primary tumors, HLA Class ir melanoma cellsseem to accumulate in areas of deepest cell invasion, and theproportion of HLA Class ll+ cells correlates with tumor thickness

(52). This might suggest that expression of HLA Class II antigensoccurs as a late event in tumor progression, perhaps aftertransformation has occurred. Another puzzle with regard toClass II antigen expression in melanoma is the striking variationin the number of HLA Class ir cells in melanomas from differentpatients and also the heterogeneity and the intensity of Class IIstaining seen within the same lesion. We have noted that Classir melanoma cells tend to occur in clusters and that theseclusters are frequently found around blood vessels (Fig. 4). Apossible reason for this observation comes from the fact thatexpression of Class II antigens can be induced in a variety ofdifferent cell types, including melanocytes and Class II nonex-pressing melanomas, by -Hnterferon (39, 53). The production of7-interferon by activated T-cells migrating through the bloodvessel wall or infiltrating the tumor might account for the presence of HLA Class II melanoma cells around blood vessels. Thus,the basis for the expression of Class II antigens in melanomasis unknown, with persistence of a differentiation-related trait,transformation-related derepression of HLA Class II genes, andinduction as a result of T-cell activation in the vicinity of melanomacells being three possibilities. In addition, the interpretation ofthe Class ir melanoma phenotype is complicated by extremevariability and heterogeneity in antigen expression and the factthat constitutive as well as inducible HLA Class II antigen expression may be involved.

With some antigens, such as GD3,the melanoma proteoglycanand Class II antigens, there is a general correspondence betweenthe pattern of expression in cultured cells and in specimens ofnormal and malignant tissues. This is not the case with two otherantigens, gp130 and gp95. Both antigens are strongly expressed

by a wide range of cultured cells, with most cell lines showinggp130 and/or gp95 expression. In tissue sections, gp130 isfound on a subset of melanomas and a narrow range of normaltissues; no other tumor type was gp130+. One possibility to

account for this difference between antigen levels in vivo and invitro is that gp130 and gp95 are proliferation-related molecules,and this would explain the higher levels on rapidly proliferatingin vitro cell lines. Through an exchange of reagents, we knowthat gp95 is related to p97 (43, 54) and that the several mAbsthat have been raised to gp95/p97 detect three families ofepitopes on the molecule (45). It has been reported that 14 of16 melanomas express p97 (47), and this contrasts with the lowfrequency of gp95+ melanomas in this series (7 of 63). To explain

the basis for this difference, we have done comparative testswith mAbs to different epitopes of the gp95/p97 molecule. Theseantibodies included mAb L235 from our laboratory and mAb 8.2,mAb 96.5, and mAb 118.1 provided by Dr. Brown. On a panelof 10 melanoma specimens, two were L235+, six were 8.2*, andseven were 96.5+ and 118.1+ using indirect immunofluores-

cence. Differences in antibody affinity may account for the observed variation in the staining ability, and this possibility is beinginvestigated.

Three of the mAbs analyzed in this study identify markers thatdistinguish early (mAb M13-17), intermediate (mAb M111), and

late (mAb M144) stages of melanocyte differentiation (9). Cultured melanoma lines can be divided in groups on the basis ofthese markers, and it has been suggested that these subsetsreflect similar phenotypes of cells in the melanocyte differentiation pathway (9). The present study represents an initial effort touse these mAbs to subset melanoma specimens. However,markers that were most useful in analyzing melanocyte differentiation antigens expressed by cultured cells have been foundto have less value for the study of melanoma tissue specimens.Some mAbs from the study of Houghton ef a/., (e.g., mAb AJ8,mAb A010, mAb A092, and mAb M144) failed to react withmelanoma specimens,5 and the use of HLA Class II antigens as

differentiation markers for melanoma is qualified by the fact thatClass II expression can be modulated by environmental factors(see above). Five of the mAbs used in this study subsettedmelanoma on the basis of antigen expression (mAb B5, mAbM111, mAb L101, mAb L235, and mAb L230), and a range ofcomposite antigenic phenotypes has been recognized. Studiesof the relation of these antigenic phenotypes to morphology andpigmentation and to other features of the disease are currentlyunder way.

ACKNOWLEDGMENTS

The authors wish to acknowledge Bettina Fliegel for excellent technical assistance and Lauren Stich for preparing the manuscript. Drs. Soldano Ferrane, RalphReisfeld,and Joseph Brown kindly provided antibodies.

REFERENCES

1. Koprowski, H, Steplewski, Z., Herlyn,D., and Hertyn,M. Studiesof antibodiesagainst human melanomaproduced by somatic cell hybrids. Proc. Nati. Acad.Sci. USA, 75: 3405-3409, 1978.

2. Yeh, M-H., Hellstrom, I., Brown, J. P., Warner, G. A., Hansen, J. A., andHellstrom, K. E. Cell surface antigens of human melanoma identified bymonoclonalantibody. Proc. Nati. Acad. Sci. USA, 76: 2927-2931,1979.

3. Carrel, S., Accolla, R. S., Carmagnola,A. L., and Mach, J-P. Common humanmelanoma-associatedantigen(s)detected by monoclonal antibodies. CancerRes., 40: 2523-2528,1980.

5F. X. Real, A. N. Houghton, A. P. Albino, H. F. Oettgen, and L. T. Old,unpublishedobservations.

CANCER RESEARCH VOL. 45 SEPTEMBER 1985

4408

on July 23, 2021. © 1985 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

ANTIGENS OF MELANOCYTES AND MELANOMAS IN CULTURED CELLS AND TISSUES

4. Dippold, W. G., Lloyd, K. O., Li, L. T. C., Ikeda, H., Oettgen, H. F., and Old, L.J. Cell surface antigens of human malignant melanoma. Definition of six newantigenic systems with mouse monoclonal antibodies. Proc. Nati. Acad. Sci.USA, 77. 6114-6118, 1980.

5. Imai. K., Ng, A. K., and Ferrane, S Characterization of monoclonal antibodiesto human melanoma-associated antigens. J. Nati. Cancer Inst., 66. 489-496,

1981.6. Johnson, J. P., Demmer-Dieckman, M., Meo, T., Hadam, M. R., and Reith-

muller, G. Surface antigens of human melanoma cells defined by monoclonalantibodies. I. Biochemical characterization of two antigens found on all linesand fresh tumors of diverse origin. Eur. J. Biochem., 77: 825-831, 1981.

7. Morgan, A. C., Galloway, D. R., and Reisfeld, R. A. Production and characterization of monoclonal antibodies to a melanoma-specific glycoprotein. Hybri-doma, 7. 27-36, 1981.

8 Proceedings of the National Cancer Institute Workshop on Monoclonal Antibodies to Human Melanoma Antigens. Hybridoma, 7: 379-482,1982.

9. Houghton, A. N., Eisinger, M., Albino, A. P., Caimcross, J. G., and Old, L. J.Surface antigens of melanocytes and melanomas. Markers of melanocytedifferentiation and melanoma subsets. J. Exp. Med., 756. 1755-1766,1982.

10. Herlyn, M., Steplewski, Z., Heriyn, D., Clark, W. H., Ross, A. H., Blaszczyk,M., Pak, K. Y., and Koprowski, H. Production and characterization of monoclonal antibodies against human malignant melanoma. Cancer Invest., 7: 215-224, 1983.

11. Thompson, J. J., Herlyn, M. F., Elder, D. E., Clark, W. H., Steplewski, Z., andKoprowski, H. Expression of HLA-DR antigens in freshly frozen human tumors.Hybridoma, 1: 161-168, 1982.

12. Thompson, J. J., Herlyn, M. F., Elder, D. E., Clark, W. H., Steplewski, Z., andKoprowski, H. Use of monoclonal antibodies in detection of melanoma-associated antigens in intact human tumors. Am. J. Pathol., 707: 357-361,1982.

13. Natali, P. G., Cavaliere, R., Bigotti, A., Nicotra, M. R., Russo, C., Ng, A. K.,Giacomini, P., and Ferrane, S. Antigenic heterogeneity of surgically removedprimary and autologous metastatic human melanoma cells. J. Immunol., 730:1462-1466, 1983.

14. Herlyn, M., Herlyn, D., Elder, D. E., Bondi, E., LaRossa, D., Hamilton, R.,Sears, H. F., Balaban, G., Guerry IV, D., Clark, W. H., and Koprowski, H.Phenotypic characteristics of cells derived from precursors of human melanoma. Cancer Res., 43: 5502-5508,1983.

15. Ruiter, D. J., Bhanm, A. K., Harris!, T. J., Sober, A. J., and Mihm, M. C., Jr.Major histocompatibility antigens and mononuclear inflammatory infiltrates inbenign nevomelanocyte proliferations and malignant melanoma. J. Immunol.,729:2808-2815,1982.

16. Ruiter, D. J., Bergman, W., Welvaart, K., Scheffer, E., van Violen, W. A.,Russo, C., and Ferrane, S. Immunohistochemical analysis of malignant melanomas and nevocellular nevi with monoclonal antibodies to distinct mono-morphic determinants of HLA antigens. Cancer Res.. 44: 3930-3935,1984.

17. Larson, S. M., Carrasquillo, J. A., Krohn, K. A., Brown. J. P., McGuffin, R. W.,Ferens, J. M.. Graham, M. M., Hill, L. D., Beaumier, P. L.. Hellstrom, K. E. andHellstrom. I. Localization of 13'l-labeled p97-specific Fab fragments in humanmelanomas as a basis for radiotherapy. J. Clin. Invest., 72: 2101-2114,1983.

18. Houghton, A. N., Mintzer, D. M., Cordon-Cardo, C., Welt, S., Fliegel, B..Vadhan, S., Carswell, E., Melamed, M. R., Oettgen, H. F., and Old, L. J. Mousemonoclonal antibody detecting Gm ganglioside: a phase I trial in patients withmalignant melanoma. Proc. Nati. Acad. Sci. USA, 82: 1242-1246,1985.

19. Albino, A. P., Lloyd, K. 0., Houghton, A. N., Oettgen, H. F., and Old, L. J.Heterogeneity in surface antigen expression and glycoprotein expression ofcell lines derived from different métastasesof the same patient: implicationsfor the study of tumor antigens. J. Exp. Med., 754: 1764-1778,1981.

20. Carey, T. E., Takahashi, T., Resnick, L. A., Oettgen, H. F., and Old, L. J. Cellsurface antigens of human malignant melanoma. I. Mixed hemadsorption assayfor humoral immunity to cultured autologous melanoma cells. Proc. Nati. Acad.Sci. USA, 73: 3278-3282,1976.

21. Shiku, H., Takahashi, T., Oettgen, H. F., and Old, L. J. Cell surface antigensof human malignant melanoma. II. Serological typing with immune adherenceassays and definition of two new surface antigens. J. Exp. Med., 744: 873-881,1976.

22. Shiku, H., Takahashi, T., Resnick, L. A., Oettgen, H. F., and Old, L. J. Cellsurface antigens of human malignant melanoma. III. Recognition of autoanti-bodies with unusual characteristics. J. Exp. Med., 745: 784-789,1977.

23. Pfreundschuh, M., Shiku, H.. Takahashi, T., Ueda, R., Ransohoff, J., Oettgen.H. F., and Old, L. J. Serological analysis of cell surface antigens of malignanthuman brain tumors. Proc. Nati. Acad. Sci. USA, 75: 5122-5126, 1978.

24. Ogata. S-l., Ueda, R., and Lloyd, K. O. Comparison of [3H]glucosamine-labeled

glycoproteins from human renal cancer and normal kidney epithelial cell cultures by two-dimensional polyacrylamide gel electrophoresis. Proc. Nati. Acad.Sci. USA, 78: 770-774,1981.

25. Cairncross, J. G., Mattes, M. J., Beresford, H. R., Albino, A. P., Houghton, A.N., Lloyd, K. 0., and Old, L. J. Cell surface antigens of human astrocytomasdefined by mouse monoclonal antibodies: identification of astrocytoma subsets. Proc. Nati. Acad. Sci. USA, 79: 5641-5645, 1982.

26. Pukel. C. S., Lloyd, K. O., Travassos, L. R., Dippold. W. G., Oettgen, H. F.,and Old. L. J. Gra, a prominent ganglioside of human melanoma. Detectionand characterization by mouse monoclonal antibody. J. Exp. Med., 755:1133-1147, 1982.

27. Yen. M-Y., Hellstrom, I., Abe. K., Hakomori, S., and Hellstrom, K. E. A cellsurface antigen which is present in the ganglioside fraction and shared byhuman melanomas. Int. J. Cancer, 29: 269-275,1982.

28. Nudelman, E.. Hakomori, S., Kannagi, R., Levery, S., Yeh, M-H., Hellstrom, K.E., and Hellstrom, I. Characterization of a human melanoma-associated antigendefined by a monoclonal antibody, 4.2. J. Biol. Chem., 257: 12752-12756,1982.

29. Cheresh, D. A.. Reisfeld, R. A., and Varki, A. P. 0-Acetylation of disialogan-glioside GD3by human melanoma cells creates a unique antigenic determinant.Science (Wash. DC), 225: 844-846, 1984.

30. Graus, F., Cordon-Cardo, C., Houghton, A. N., Melamed, M. R., and Old, L. J.Distribution of the ganglioside GDSin the nervous system detected by R24mouse monoclonal antibody. Brain Res., 324: 190-194,1984.

31. Wilson, B. S., Imai, K., Natali, P. G., and Ferrane, S. Distribution and molecularcharacterization of a cell surface and a cytoplasmic antigen detectable inhuman melanoma cells with monoclonal antibodies. Int. J. Cancer, 28: 293-300. 1981.

32. Bumol, T. F. and Reisfeld, R. A. Unique glycoprotein-proteoglycan complexdefined by monoclonal antibody on human melanoma cells. Proc. Nati. Acad.Sci. USA, 79: 1245-1249, 1982.

33. Hellstrom, l., Garrigues, H. J., Cabasco, L., Molsey, G. H., Brown, J. P., andHellstrom, K. E. Studies of a high molecular weight human melanoma-associated antigen. J. Immunol., 730: 1467-1472, 1983.

34. Lloyd, K. 0., Ng, J., and Dippold. W. G. Analysis of the biosynthesis of HLA-DR glycoproteins in human malignant melanoma cells. J. Immunol., 726:2408-2413, 1981.

35. Winchester, R. J., Wang, C-Y., Gibofsky, A., Kunkel, H. G., Lloyd, K. 0., andOld, L. J. Expression of la-like antigens on cultured human malignant melanomacell lines. Proc. Nati. Acad. Sci. USA, 75: 6235-6239,1978.

36. Wilson, B. S., Indiveri, F., Pellegrino. M. A., and Ferrane, S. DR(la-like) antigenson human melanoma cells: Serological detection and immunochemical characterization. J. Exp. Med., 749: 658-668. 1979.

37. Natali, P. G., Cordiali-Fei, P., DiFilippo, F., Quaranta, V., Pellegrino, M. A., andFerrane, S. la-like antigens on freshly explanted human melanoma. Clin.Immunol. Immunopathol., 79: 250-259, 1981.

38. Natali, P. G., DeMartino, C., Quaranta, V., Bigotti, A., Pellegrino, M. A., andFerrane. S. Changes in la-like antigen expression in malignant human cells.Immunogenet. 72: 409-413. 1981.

39. Houghton, A. N., Thomson, T. M., Gross, D., Oettgen, H. F., and Old, L. J.Surface antigens of melanoma and melanocytes. Specificity of induction of laantigens by human interferon. J. Exp. Med., 760: 255-269, 1984.

40. Natali, P. G., DeMartino, C., Quaranta, V., Nicotra, M. R., Frezza, F., Pellegrino,M. A., and Ferrane, S. Expression of la-like antigens in normal human non-lymphoid tissues. Transplantation (Baltimore), 37: 75-78, 1981.

41. Saxton, R. E., Mann, B. D., Irie, R. F., Morton, D. L., and Burk, M. W.Monoclonal antibodies to human melanoma: identification of two unique proteins on the tumor cell membrane (Abstract). Proc. Amer. Assoc. Cancer Res.23:263, 1982.

42. Albino, A. P., Lloyd, K. O.. Ikeda, H., and Old, L. J. Biochemical analysis of a130,000 molecular weight glycoprotein on human melanoma cells. J. Immunol.,737: 1595-1599, 1983.

43. Woodbury, R. G., Brown, J. P., Yeh, M-Y., Hellstrom, I., and Hellstrom, K. E.Identification of a cell surface protein, p97, in human melanomas and certainother neoplasms. Proc. Nati. Acad. Sci. USA, 77: 2183-2187,1980.

44. Brown, J. P., Woodbury, R. G., Mart. C. E., Hellstrom. I., and Hellstrom, K. E.Quantitative analysis of melanoma-associated antigen p97 in normal andneoplastic tissues. Proc. Nati. Acad. Sci. USA, 78: 539-554, 1981.

45. Brown, J. P.. Nishiyama. K.. and Hellstrom, K. E. Structural characterizationof human melanoma-associated antigen p97 with monoclonal antibodies. J.Immunol., 727, 539-546, 1981.

46. Brown, J. P., Hewick, R. M., Hellstrom, l., Hellstrom. K. E., Doolittle, R. F.,and Dreyer. W. J. Human melanoma-associated antigen p97 is structurally andfunctionally related to transferrin. Nature (Lond.), 296: 171-173, 1982.

47. Garrigues, H. J., Tilgen, W., Franke, W., and Hellstrom, K. E. Detection of ahuman melanoma-associated antigen, p97, in histological sections of primaryhuman melanomas. Int. J. Cancer, 29: 511-515, 1982.

48. Le Douarin, N. Migration and differentiation of neural crest cells. In: CurrentTopics in Developmental Biology, Academic Press, 76: 31-85, 1980.

49. Yu, R. K. and Ando, S. Structures of some new gangliosides of fish brain. Adv.Exp. Med. Biol., 725: 33-45. 1980.

50. Harper, J. R., Bumol, T. F., and Reisfeld, R. A. Characterization of monoclonalantibody 155-8 and partial characterization of its proteoglycan antigen onhuman melanoma cells. J. Immunol.. 732: 2096-2104, 1984.

51. Old, L. J. and Stocken, E. Immunogenetics of cell surface antigens of mouseleukemia. Ann. Rev. Genet. 77: 127-160,1977.

52. Brocker, E-V., Suter. L., and Sorg, C. HLA-DR antigen expression in primarymelanomas of the skin. J. Invest. Dermatol., 82: 244-247,1984.

53. Basham, T. Y., Bourgeade, M. F., Creasey, A. A., and Merigan, T. C. Interferonincreases HLA synthesis in melanoma cells: interferon-resistant and -sensitivecell lines. Proc. Nati. Acad. Sci. USA. 79: 3265-3269,1982.

54. Woodbury, R. G., Brown, J. P., Loop, S. M., Hellstrom, K. E., and Hellstrom,I. Analysis of normal neoplastic human tissues for the expression of tumor-associated p97. Int. J. Cancer, 27: 145-149, 1981.

CANCER RESEARCH VOL. 45 SEPTEMBER 1985

4409

on July 23, 2021. © 1985 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1Fig. 1. Immunofluorescence analysis of reactivity of mAb R2<,mAb B5, and mAb L101 with normal adult skin, a compound nevus, and two metastatic melanomas,

a-c, normal skin; d-f, compound nevus; and g-l, metastatic melanoma, a, d, g, and ¡,reaction with mAb R24;6, e, h, and k, reaction with mAb B5; and c, f, i, and I,reaction with mAb L101. mAb R2<reacts with normal skin melanocytes, nevus cells, and metastatic melanomas; mAb B5 reacts weakly with basal keratinocytes andmore strongly with nevus cells and metastatic melanomas; mAb L101 does not react with normal skin melanocytes, but it does react with nevus cells and metastaticmelanomas; and mAb L101 reacts with endothelial cells in normal and tumor tissues. Original magnification: a-t, x 200; g-l, x 400.

4410

on July 23, 2021. © 1985 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

ANTIGENS OF MELANOCYTES AND MELANOMAS IN CULTURED CELLS AND TISSUES

v'

••r.a b2

Fig. 2. Immunoperoxidase analysis of the reactivity of mAb R24with a lymph node containing a deposit of metastatic melanoma, a, control: no mAb R24;o, mAb R24.Original magnification, x 400.

Fig. 3. Indirect immunofluorescence analysis of the reactivity of mAb R24with: (a) astrocytoma; (b) embryonal rhabdomyosarcoma; (c) lung carcinoma: and (cOan R24'

metastatic melanoma. Original magnification, x 200.

CANCER RESEARCH VOL. 45 SEPTEMBER 1985

4411

on July 23, 2021. © 1985 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1985;45:4401-4411. Cancer Res   Francisco X. Real, Alan N. Houghton, Anthony P. Albino, et al.   TissuesComparison of Antigen Expression in Cultured Cells andMouse Monoclonal Antibodies: Specificity Analysis and Surface Antigens of Melanomas and Melanocytes Defined by

  Updated version

  http://cancerres.aacrjournals.org/content/45/9/4401

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

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/45/9/4401To request permission to re-use all or part of this article, use this link

on July 23, 2021. © 1985 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from