109

Biochimica et Biophysica Acta, 1138 (1992) 109-114 © 1992 Elsevier Science Publishers B.V. All rights reserved 0925-4439/92/$05.00

BBADIS 61108

Production of monoclonal antibodies against the B700 murine melanoma antigen and their antimetastatic properties Douglas M. Gersten

1,

Dana Moody 1, Wilfred D. Vieira and Vincent J. Hearing 2

2,

Lloyd W. Law 2

1 Department of Pathology, Georgetown Unit,ersity Medical Center, Washington, DC (U.S.A.) and 2 Laboratory of Cell Biology, National Cancer Institute, National Institutes of Health, Bethesda, MD (U. S.A.)

(Received 5 June 1991) (Revised manuscript received 10 September 1991)

Key words: Melanoma; Monoclonal antibody; Antigen; Antimetastatic

Two unique murine melanoma antigens, termed B700 and B50, have been identified and isolated from several different murine melanoma cell lines. Both antigens can be detected on the cell surface, are actively shed in culture, and are often found in close association intracellularly. In previous studies, the antigen B700, which is related to serum albumin by biochemical and immunological criteria, was shown to function as a melanoma-specific tumor rejection antigen. We have also shown that animals sensitized to irradiated JB/RI-I melanoma cells produce antibodies which recognize B700 a n d / o r B50, with B700 evoking the stronger humoral response. Animals testing positive by ELISA for antibody production to B700 or B50 were used for preparation of hybridomas and four different murine monoclonal antibodies have been produced whose specificities should facilitate epitope mapping. Clones have been used to generate ascites fluid in nude mice; the antibodies specifically recognize B700 and intact murine melanoma cells, but not B50. Two of these monoclonal antibodies have been administered systemically to C 5 7 B I / 6 mice bearing 5 day pulmonary metastases of the J B / M S melanoma, and significant inhibition of metastatic growth was observed for both antibodies.

Introduction A key issue in understanding the nature of tumor cells is the expression of unique, tumor-specific molecules, because these determine, in part, the host recognition and response to the tumor. Melanomas are highly aggressive tumors with a well-documented antigenic nature. However, few melanoma antigens have been described which can be considered 'complete' tumor antigens, that is, antigens which are tumorspecific, produced by all individual tumors of the same histotype, capable of eliciting a tumor rejection response, and are antenna molecules which are recognized by the tumor-bearing host [1,2]. B700 and B50 are two melanoma antigens [3-7] which have been Abbreviations: B6C3F1, C57BI/6×C3H FI; ELISA, enzyme-linked immunosorbent assay; Hb, hemoglobin; MPLA, monophosphoryl lipid adjuvant; MSA, murine serum albumin; OPD, ortho-phenylenediamine; PBS, phosphate-buffered saline. Correspondence: D.M. Gersten, Department of Pathology, Georgetown University Medical Center, Washington, DC 20007, U.S.A.

isolated from several different murine melanomas [2]. These two protein antigens are often found in close association, but they are separate and distinct molecules, differing by several biochemical and immunological criteria [2,8,9]. B700 is the major protein of the melanosomal membrane and can be found on most membranous structures of melanoma cells, including the plasma membrane [10-13]. It is a multiple deletion variant of a normally occurring melanosomal protein (termed C700) and has primary structure variability, at least at its amino terminus [14,15]. B700 has been found on all murine melanomas examined to date, and cross-reactive molecules have been described on human, swine and hamster melanomas [16]. The distribution of B50 within the cells has yet to be determined. Most importantly, B700 has been shown to function as an effective immunogen, mediating rejection of viable tumor cells [17]. Thus, B700 qualifies as a 'complete' tumor antigen. Since B700 and B50 are found complexed together within the melanoma cells, the goals of this study were:

110 (a) to produce murine monoclonal antibodies which might be used to facilitate further study of their structure and association; and (b) to determine the extent to which such antibodies can be used in the therapy of melanoma tumor growth, particularly metastases. Materials and Methods

Materials B16 F10, J B / R H and JB/MS murine melanoma cells, T92497 sarcoma cells and GL26 murine glioma cells (all of C57B1/6 origin) were cultured and passaged as previously detailed [18,19]; briefly, cells were routinely grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, glutamine, nonessential amino acids, penicillin, streptomycin and fungizone. The cells were harvested with trypsin:EDTA and subcultured using standard techniques. P3X63 murine and GK1.5 rat (the source of aL3T4 antibodies) myeloma cells were obtained from the American Type Culture Collection (Rockville, MD). aNKI.1 and ~CMS4 sera were generously provided by Dr. Albert DeLeo, Pittsburgh Cancer Institute. MOPC-141 was obtained from Organon-Teknika-Cappel (West Chester, PA). Source and preparation of antigens Murine serum albumin (MSA) and bovine hemoglobin (Hb) were obtained from Sigma (St. Louis, MO) and used without further purification. The murine melanoma antigens B700 and B50 were purified to homogeneity from B16 F10 tumors growing subcutaneously in C57B1/6 mice as previously detailed [2]. Briefly, the techniques involved in purification were homogenization and solubilization of the tissues, salt precipitation, centrifugation, chromatography and preparative gel electrophoresis. Animals and immunization Animals used in this study were female, 6-8-weekold, pathogen-free C57B1/6, C57B1/6 × C3H F 1 (B6C3F 1) and NIH Swiss n u / n u mice obtained from the N.I.H. Small Animal Section and maintained in a pathogen-free animal facility. For immunization, the B6C3F~ mice were primed intraperitoneally at day - 3 with cytoxan (25 mg/kg) and then immunized at biweekly intervals (starting at day 0) with subcutaneous injections of 5 • 10 6 irradiated (10000 R) J B / R H cells [2]. The initial immunization contained Monophosphoryl lipid A (MPLA, RIBI, Hamilton, MT) as an adjuvant (100/xg/dose); the animals were bled at periodic intervals thereafter from the retro-orbital sinus. Production of monoclonal antibodies Hybridomas were produced using P3X63 myeloma cells as fusion partners by previously published stan-

dard techniques [13]; briefly, this involves fusion of spleen cells from immunized animals using poIy(ethylene glycol) and subsequent selection with HAT medium. Antibody producing cells were identified using enzyme linked immunosorbent assay (EL1SA) techniques as outlined below, and were cloned using two rounds of limiting dilution (approx. 3 cells/ml). Hybridomas were grown intraperitoneally in pristaneprimed nude mice using standard techniques; ascites fluid was removed periodically from these animals and used directly for ELISA testing or for antibody purification as noted in the text. Antibodies were purified by Protein A agarose (Pierce Chemical Co., Rockford, IL) using the suggested protocol, and isotyping of the various monoclonal antibodies was performed with a subtyping kit (PharMingen, San Diego, CA). Enzyme linked immunosorbent assay ELISAs were performed as previously described [13]. Briefly, antigens were fixed to plates by one of two methods: (1) when purified antigens were used, 100 ng were attached to each well of flat bottom 96 well Immulon II microtiter plates using a pH 9.6 coating buffer; (2) when intact cells were used, they were grown in sterile flat-bottom 96 well microtiter plates and, when wall-to-wall confluent, they were washed six times in Hanks' balanced salt solution, then fixed to the plates using 0.15% glutaraldehyde in Hanks' solution for 15 min. The plates were blocked by incubation in phosphate buffered saline (PBS) containing 1 mg/ml soluble fish gelatin (Norland Products, New Brunswick, NJ) at 4°C overnight, and then thoroughly washed with PBS containing 0.01% (v/v) Tween 20, incubated with the appropriate antibodies (as noted in the figure and table legends), and washed again; bound antibodies were detected with peroxidase-conjugated secondary antibodies and reaction with o-phenylenediamine (OPD) substrate [13]. The plates were then measured for A490 using a TiterTek Multiskan ELISA reader. In vivo antimetastatic assays Randomized groups of C57B1/6 mice were challenged intravenously with viable JB/MS melanoma cells or T92497 sarcoma cells in a lateral tail vein, using the experimental metastasis system originally developed by Fidler [20,21]. At days 5 and 8 thereafter (times when metastases were present), the animals were treated intravenously using purified antibodies at the concentrations noted in the figure and table legends. Controls received either PBS, normal murine IgG or MOPC-141 IgG (a murine IgG2b isotype which recognizes an irrelevant antigen). At day 18, the animals were sacrificed and their pulmonary metastases counted. In some protocols, animals were treated at days 4 and 7 following challenge (i.e. 1 day prior to each antibody treatment) with PBS, aCMS4 (0.2

111

rag/dose), a L 3 T 4 (0.2 m g / d o s e ) or aNKI.1 (50 /zl/dose) and the assays were continued as detailed above. Results and Discussion

B6C3F l mice were immunized with irradiated J B / R H melanoma cells in order to determine whether the animals produced a humoral response to either the B700 a n d / o r the B50 melanoma antigens. It had been previously demonstrated that mice produce antibodies recognizing B700 following inoculation of the purified antigen [2]. Animals were immunized with irradiated cells, bled and their sera tested for the presence of specific antibodies by ELISA. Since B700 had been shown in earlier studies to be immunologically crossreactive with albuminoid molecules [15,22,23], the specific reactivity of these antibodies with the B700 antigen was calculated by subtracting cross-reactive binding to MSA. Ten of the 11 mice immunized with the intact melanoma cells produced significant titers (i.e. A 4 9 0 > 0.10 at 1 : 10 dilution) of antibodies which reacted with purified B700 antigen, and sera from 10 of the same 11 animals also contained significant levels of antibodies which recognize the B50 melanoma antigen [2]. Hence, both the B700 and the B50 antigens are among the molecules to which the host mounts a humoral response. The observation that some animals produce antibodies to one, but not to both, of the antigens suggested that, at least in those animals, the epitopes recognized are on individual molecules, and are not directed against an epitope defined by the B700 : B50 complex.

Having demonstrated the production of a significant antibody response to B700 and B50 in animals immunized with whole cells,the animals Were given two additional immunizations with irradiated J B / R H cells in order to facilitate monoclonal antibody production. Spleen cells from these hyperimmunized animals were then fused with murine myeloma cells and antibody producing hybridoma cells were screened and cloned by limiting dilution using standard techniques. Thus far, four reactive hybridoma clones have been selected and used to produce ascites fluid in nude mice to allow further characterization of the antibodies. The ascites fluid recovered has been tested by ELISA for reactivity against the purified antigens and used as a source for antibody purification through Protein A agarose columns. The specificity of the four monoclonal antibodies is detailed in Table I; relatively high titers of antibodies were purposely used in those assays to accentuate any potential cross-reactivity. The results show that all four clones produce antibodies which recognize B700 but do not cross-react with B50. This is significant (as with the studies noted above using the original whole sera) due to the close association of the two antigens in vivo, and demonstrates that conformational epitopes due to the association of those antigens were not recognized. Of the four clones established, three cross-reacted slightly with MSA (suggesting a partial overlap of shared sequences) while one antibody (2-3-3) was completely specific for B700 and thus recognized a unique epitope. All antibodies were negative for binding to Hb, which was used in these studies as a negative control. The isotype of the antibodies is shown in the far right column. The specificity of the antibodies is

TABLE I

Specificity of monoclonal B700 antibodies 100 n g o f B700, B50, M S A o r H b w e r e b o u n d to t h e wells o f m i c r o t i t e r p l a t e s as n o t e d , a n d E L I S A s w e r e p e r f o r m e d o n ascites fluids ( d i l u t e d as n o t e d ) as d e t a i l e d in t h e M e t h o d s section. R e s u l t s a r e r e p o r t e d as t h e m e a n A490 o f d u p l i c a t e assays; v a r i a t i o n s w e r e < + 5 % . Specific reactivity r e p r e s e n t s t h e b i n d i n g to M S A s u b s t r a c t e d f r o m the b i n d i n g to B700. C o n t r o l i n d i c a t e s s e r u m p o o l e d f r o m f o u r a g e - m a t c h e d n o r m a l mice. n.d. not determined. A n t i b o d y (1 : 1)

A n t i g e n b o u n d to E L I S A p l a t e vs. B 7 0 0

vs. B50

vs. M S A

vs. H b

isotype

+ + + +

-

+ + +

-

igG2b igG2 b n.d. n.d.

2-3-1 2-3-3 2-3-5 3-4-1 Antibody

control 2-3-1 2-3-3 2-3-5 3-4-1

Antibody

+ + + + + + +

vs. B 7 0 0

vs. M S A

Specific reactivity

1:1

1:10

1:1

1:10

1:1

1:10

0.18 0.78 0.86 0.52 0.55

0.16 0.92 0.64 0.36 0.44

0.19 0.38 0.20 0.40 0.35

0.16 0.44 ,0.18 0.22 0.25

0.00 0.40 0.66 0.12 0.20

0.00 0.48 0.46 0.14 0.19

112

B16 FIO

0.8 c o m

melanoma

~i

GL 26 g l i o m o -

06

0.8-

0.6-

g o4

-

04-

o m 1:1


2-3-5 ~- 3-4-1. Serial dilutions of the ascites fluid from the four clones were also tested by ELISA for reactivity against fixed monolayers of cultured whole cells. The results in Fig. 1 show that three of the four hybridoma antibodies recognize intact B16 Ft0 murine melanoma cells, but do not cross-react with murine glioma cells (also of C57BI/6 strain origin) which has its own tumor rejection antigen [24]. Since the specificities and affinities of 2-3-1 and 2-3-3 are the highest, they were chosen for further study, as detailed below.

Monocional antibodies have been shown to be useful as immunotherapeutic agents, with the major limitation on their use being the specificity of the antigens recognized on the transformed cells [25-31]. Since B700 has been shown to be a melanoma-specific antigen which cross-reacts among species including mice, hamsters, swine and man [16], the potential to use these antibodies in a therapeutic or diagnostic mode is an exciting prospect. In a protocol to study the potential usefulness of these antibodies, we have employed the experimental metastasis model using antibody therapy on established pulmonary metastases. The results in Table II show that a significant reduction in pulmonary metastases was achieved using therapy with low doses of unconjugated B700 monoclonal antibodies (typically 40 to 50% inhibition at 100 txg/dose). Since 2-3-1 and 2-3-3 recognize different epitopes of B700, we have also employed protocols which examined whether the antimetastatic effects of the antibodies were synergistic (Fig. 2). The results clearly show that the effects of either antibody are maximal and can not be improved by mixing various proportions of the antibodies for treatment. Whether this implies that only a subpopulation of the metastatic cells are expressing B700 and can thus be identified for killing remains to be determined. The limit of growth inhibition to 50% is a problem inherent in the experimental metastasis assay. Two lines of evidence suggest that the lungs of these animals constitute an immunologically privileged site, as far as tumor rejection is concerned. First, animals immunized outright to purified B700 protein can fully reject a subcutaneous challenge of 5.105 melanoma

TABLE II

Antimetastatic eJfects of B700 antibodies Groups of 6 C57B1/6 female mice (10/group in controls) were challenged intravenously with 2.1(15 J B / M S melanoma cells on day 0, then treated at days 5 and 8 intravenously with immunoglobulins as noted; mice were killed at day 18 and their pulmonary metastases counted, as detailed in the Methods section. Results are presented as the mean_+ S.E.; statistical significance was calculated using Student's t test. Antibody

# g/dose

Number of pulmonary metastases mean _+S.E.

p value

Experiment 1 Untreated controls Normal mouse lgG

100

79+ 87~_ 12

2-3-1 IgG 2-3-3 IgG

100 100

48_+ 8 45+ 8

83+ 17 < 0.01