ancer mmunol9gyand mmunotherapy

Cancer Immunol Immunother (1982) 12:217-224

© Springer-Verlag 1982

Monoclonal Antibodies from Mice Bearing Polyoma Virus-Induced Tumor Neomi Moav, Nechama I. Smorodinsky, Ben-Ami Balin, and Isaac P. Witz Department of Microbiology, The George S. Wise Faculty of Life Sciences, Tel Aviv University, Tel Aviv 69978, Israel

Summary. Monoclonal antibodies were produced by fusing NS1/1 myeloma cells with splenocytes from A. B Y mice bearing syngeneic polyoma virus-induced SEYF-a tumors. From six separate fusion experiments 514 hybridomas were obtained, 45 of which were found to secrete SEYF-a-binding antibodies. The binding patterns of antibodies secreted by eight hybridomas to a panel of tumor cells and to normal mouse fibroblasts were analyzed by means of an indirect radioimmunoassay. Seven hybridomas were found to secrete antibodies that bound to all ceil lines tested. This indicated that certain SEYF-a-associated antigens are widely distributed on a variety of seemingly nonrelated tumor cells. One hybridoma secreted antibodies that exhibited a high binding activity to SEYF-a cells, a low binding activity to two members of the tumor panel, and none at all against most of its constituents, including normal fibroblasts. The results of the binding experiments were further supported by absorption experiments. A subclass analysis of the immunoglobulins secreted by the various hybridomas revealed that three clones secreted IgG1; one clone secreted IgM; and three clones secreted IgG2a. Polyacrylamide gel electrophoresis of two of the secreted antibodies indicated a high degree of homogeneity of the heavy and the light chain of the corresponding antibodies, as would be expected from monoclonal products. The results of this study demonstrate the feasibility of obtaining anti-tumor monoclonal antibodies from tumor bearers, representing the immune response of the tumor bearer against antigens associated with his syngeneic tumor.

Introduction

For cancer immunologists to do their share for the prevention, diagnosis, and treatment of cancer, they have to be capable of correct evaluation of tumor-host immune relations. A comprehensive understanding of such a relationship requires a detailed analysis of the anti-tumor immune response of the tumor-bearer both systemically and at the tumor site. The presence of anti-tumor cellular and humoral immune effectors in tumor-bearing individuals has been repeatedly reported by numerous authors [for reviews see 1, 5, 32]. Detailed specificity studies of such effectors have been hampered because only limited amounts of such components could be obtained for analysis. The recent technological advances in cell cloning and hybridization [20] open new avenues for investigation of these questions. In the present study we focus on antibodies directed against antigens expressed on SEYF-a cells, a polyoma virus-induced sarcoma propagated in syngeneic A.BY mice. The fact that in tumor-bearers SEYF-a cells evoke the production of antibodies mediating complement-dependent lysis of SEYF-a cells [23, 26] enabled us to prepare monoclonal antibodies from spleens of such mice. The binding characteristics of some of the resulting autologous monoclonal antibodies to various tumor cell lines and other properties of these antibodies are described in this paper.

Materials and Methods Animals. C57BL/6 mice were purchased from the Weizmann

Reprint requests should be addressed to: Nechama I. Smorodinsky

Institute (Rehovot, Israel). BALB/c and A.BY mice were raised at the animal quarters of the George S. Wise Faculty of Life Sciences, Tel Aviv University (Tel Aviv, Israel).

0340-7004/82/0012/0217/$ 01.60

218

N. Moav et al.: Hybridomas from Tumor-Bearing Mice

Table 1. List of tumors used as targets in the present study Tumor

Strain of origin

Originally induced by

Histologic type

Grown

Reference

SEYF-a

A.BY

( H - 2 b)

Polyoma virus

Sarcoma

In vivo

[30]

EL-4

C57BL/6

( H - 2 b)

Benzo(a)pyrene

Lymphoma

In vivo

[11]

ELD

Not inbred, grown in BALB/c

Spontaneous

Mammary carcinoma

In vivo

[17]

YAC

A/Sn

( H - 2 a)

Moloney virus

Lymphoma

In culture

[18]

B16

C57BL/6

( H - 2 b)

Spontaneous

Melanoma

In culture

[4]

RLdl

BALB/c

( H - 2 d)

Radiation

Lymphoma

In culture

[28]

KREBS

Not inbred, grown in BALB/c

Spontaneous

Mammary carcinoma

In vivo

Kleinb

L5178Y

DBA/2

( H - 2 d)

Spontaneous

Lymphoma

In culture

[6]

MFB a

BALB/c

( H - 2 d)

-

Normal

In culture

a Mouse fetal fibroblasts gift from Dr K. Klein, Karolinska Institute, Stockholm, Sweden

bA

Tumor Cells. SEYF-a is an ascites form of a murine sarcoma induced originally in A.BY mice by polyoma virus [30]. The tumor was maintained in vivo by weekly intraperitoneal (IP) inoculations of 2 x 106 cells in syngeneic A.BY mice. In addition, the cells have been grown in vitro in RPMI-1640 medium supplemented with 10% fetal calf serum (FCS). Details of other tumor cells used in this study are given in Table 1. Production of Hybridomas. Cells (1 x 108) from spleens of SEYF-a-bearing A-BY mice were fused with 1 x 107 myeloma cell line NSI/1 cells derived from the BALB/c MOPC-21 line (kindly provided by Dr R. Kennett, Cell Center, University of Pennsylvania PA) by using 33% polyethylene glycol (Baker 1000) according to the procedure described by Kennett and Gilbert [14]. After fusion the cells were suspended in modified Dulbecco's medium and dispensed into six microtiter tissue culture plates of 96 wells each. Starting 14 days after the fusion the out-grown hybrids were transferred from the original plates into other microtiter plates. Supernatants from these wells were collected and tested for the presence of SEYF-a-binding antibodies. Hybridoma cells from positive wells were subcloned by means of a limiting dilution procedure. The cells were dispensed at a concentration of 0.5-1 cell/50 ~tlinto 96 microtiter wells containing 106 thymocytes per well as a feeder layer. With this procedure less than 20% of the wells yielded viable clones. Statistically, there is a 98% chance that each of these clones originated from a single cell. Positive wells were then chosen and used as the source of all antibody-containing culture supernatants used in this study.

was added to 96-well PVC U-bottom microtiter plates (Cooke Laboratory Products Div., Dynatech Laboratories Inc., Alexandria, VA, USA). To allow adsorption of the cells the plates were air-dried with constant shaking. Unattached sites in the wells were saturated with 2% BSA in PBS for 1 h at room temperature. The plates were then washed three times with PBS containing 1% BSA. Then 50-~1 volumes of hybridoma supernatants were aded to each well of the washed plates. The plates were then incubated for 1 h at room temperature. After thorough washing of the plates, 50 ~tl azsI-labeled F(ab)2 fraction of goat anti-mouse F(ab)2 (FabGaMFab) (50,000-100,000 cpm per well) was added and the plates were incubated overnight at 4° C. The wells were thoroughly washed, dried, cut, and individually counted in an automatic gamma spectrometer.

Preparation of F(ab)2 Fraction of Goat Anti-Mouse F(ab)2 (FabGaMFab). Goat anti-mouse F(ab)2 serum was a gift from the Department of Chemical Immunology, The Weizmann Institute of Science, Rehovot, Israel. This fraction was prepared by pepsin digestion [24] of affinity-purified (on mouse serum coupled to sepharose column) goat immunoglobulin, followed by an incubation with Staphylococcus aureus to remove undigested immunoglobulins and Fc fragments.

Preparation of 12SI-Labeled F(ab)2 Fragments. Iodination of FabGaMFab with i25I (purchased from Radiochemical Center, Amersham, England and from the Nuclear Research Center, Negev, Israel) was performed by the chloramine T method [12] or by the Iodogen method [8].

Fixation of Tumor Cells. Ceils were grown either in tissue culture or in vivo (see Table 1). Ascites cells were first incubated for 2 h at 37° C to allow them to uncoat from tumor-bound immunoglobulins [26]. Then 5 × 106 cells/ml were suspended and incubated overnight at 4°C with 0.3% formaldehyde diluted in phosphate-buffered saline (PBS), with constant shaking. The cells were then washed three times with PBS, resuspended, and stored at a concentration of 5 x 10 6 cells/ml in PBS containing 0.05% sodium azide.

Radioimmunoassay (RIA).

Solid-phase RIA for detecting SEYF-a-specific antibodies was used. This method followed essentially the procedure described by Huang et al. [13]. Briefly, of a suspension of 5 x 10 6 formalin-fixed tumor cells/ml PBS, 50 ~tl

Absorption Experiments. Hybridoma supernatants (150 ~tl) at various dilutions were incubated for 1 h at 4° C with 1.5 × 10 7 living tumor ceils with occasional shaking. The absorbing cells were discarded after centrifugation. The supernatant was absorbed once more with the same number of fresh cells. The absorbed supernatants (50 ~tl) were tested in RIA on SEYF-a cells for the presence of antibodies.

Biosynthetic Labeling of the Hybridoma Products. Hybridomas were cultured at a concentration of 1 x 107 cells/ml in a methionine-free RPMI 1640 medium containing 3% FCS and 10~tCi L-selenomethionine (75Se)/ml, (specific activity 3-20 mCi/mg; The Radiochemical Centre, Amersham, England). After

219

N. Moav et al,: Hybridomas from Tumor-Bearing Mice

F7 9500 "n"

2500

2000

150C

1000

500

E ,h,6,,~

Oe,

6

,~

6~

0',~

6e,

O,A

6,,h

0

~

,~

6,h

6,h

0

,,:-, o

--

e~

F3 F4

F1

o

F5

F6

10.000

I

55oo~ 1 ~ 200

Fig. 1. Screening for antibodies binding to SEYF-a cells. Three hundred fifteen hybridoma culture supernatants derived from six separate fusions of splenocytes from SEYF-a-bearing mice were screened for SEYF-a binding antibodies by means of solid-phase indirect RIA. Serum with a known cytotoxic titer against SEYF-a ceils was used at a dilution of 1 : 100 as a positive control. Supernatants from cultures of P3 x 63Ag8 were used as a negative control. Binding values above the dotted line were considered as antibody-producing hybridomas. The data are expressed as cpm bound to a fixed number of SEYF-a cells

- / t

--

l

"2 ) 1500

m

I000

~

~

"~

oo I : I

hi

I lilliill

I i

i I

I il

I I

i

[

I I I

I]

Hybridoma No.

an incubation of 18 h the culture supernatants were harvested and dialyzed against saline.

Characterization of the Hybridoma Products a) Double Immunodiffusion in Agar. Double immunodiffusion in 1% agar (Special noble agar, Difco, USA) was performed according to the procedure of Ouchterlony [25]. The supernatant fluids from the hybridomas were tested against goat anti-mouse

immunoglobulin M, IgG1, IgG2a, IgG2b and IgA (Litton Bioenetics, Kensignton MD). Diffusion was allowed to take place for 24-36 h, after which the gels were extensively rinsed with PBS and stained with 1% amidoblack.

b) Immunoelectrophoresis. Immunoelectrophoresis was performed as described by Scheidegger [27] with 1% agar in 0.05 M barbitone acetate buffer at pH 8,6. Electrophoresis was carried out for about 2 h at 10 mA per frame of six slides.

220

N. Moav et al.: Hybridomas from Tumor-Bearing Mice

c) Polyacrylamide Gel Electrophoresis with Sodium Dodecyl Sulfate. Polyacrylamide gel electrophoresis with sodium dodecyl

Six such fusions have been performed (seeding 6 x 600 wells). Colonies were observed in 514 wells. Supernatants from these cultures were screened for antibody activity against SEYF-a cells; 45 hybridomas were found to be positive (Fig. 1). Some of these hybridoma lines subsequently died or ceased to produce antibodies. Thirty-one of the hybridomas were cloned for further studies.

sulfate (SDS) was performed on thin layer slabs. The gel system used was that of Laemmli [22] in 25 mM tris-glycine buffer (pH 8.6) containing 0.1% SDS. Before application to the electrophoresis gel, the samples (in aliquots of 5-25 ~tl containing 25 mM tris-glycine, 10% glycerol, and 0.01% bromphenol blue) were treated with 2% SDS and reducing agents (e.g., 0.75 ~tl fl-mercaptoethanol). Electrophoresis was performed at 150 V for about 2 h through 10-cm slab gels (12.5% acrylamide).

Binding Patterns of Various Hybridomas Results

We analysed the binding patterns of supernatants obtained from eight hybridomas secreting SEYF-a-binding antibodies both to various murine tumor cells and to normal mouse fibroblasts (Fig. 2). It can be seen that supernatants obtained from seven hybridomas had binding activities to specificities expressed on essentially all cell lines tested, including fibroblasts ('public' determinants [19]). The binding activity of these seven antibodies to any one of the

Initial Screening Spleen cells from SEYF-a-bearing mice 4 - 6 weeks after transplantation were used for the hybridization experiments. Sera obtained from the donor mice at the time of fusion demonstrated high titers of antibodies mediating complement-dependent cytotoxicity to SEYF-a cells.

F/3

F4/2.1

4000

3000

2000

L

1000

°~ "=t F7/83.3

F4/4.8

4000

3000 2000

°t~ ,~

1000

~; !

:

y

0%

"

E e~

F6/,8.4

4000

'F7/'129.7

3000

2000 1000

16

32

64

128

t

~ ,

F 7/132.7 t

:

:!

:! ,i .;A

3000 I20001000:. ,~. .~. ' .' .~ : ' ~ • • F4/5.10

2

8

32

128

512

2043

Reciprocal of Culture fluid dilution

Fig. 2. Binding profiles of monoclonal antibodies to various tumor cell lines. Two-fold dilutions of hybrid()ma culture fluids were assayed for binding to various tumor cells by means of an RIA. The results are expressed in cpm bound to the cells, above background levels. (@) SEYF-a; (©) RLS1; (A) EL-4; (A) KREBS; (R) ELD; [] YAC; x B16; (V) L5178Y; (A) MFB. The binding activities of F4/5.10 to RL~I, EL-4, YAC, and B16 tumor cells and to MFB were all below 500 cpm. Since the binding profile to these cells was similar to that to L5178Y cells, it is not shown

221

N. Moav et al.: Hybridomas from Tumor-Bearing Mice

target tumors was arbitrarily classified as low binding (less than 1,000 cpm/well), intermediate binding (between 1,000 cpm/well and 2,000 cpm/well), and high binding (above 2,000 cpm/well). This classification is given in Table 2. The antibodies produced by hybridoma F4/5.10 demonstrated high binding activity only tO SEYF-a cells and a very low binding to KREBS and ELD cells. There was no measurable binding to RLc~-I, EL-4, YAC, B16, L5178Y, or normal mouse fibroblasts. These results suggest that F4/5.10 produces antibody directed against an antigenic determinant present predominantly on the SEYF-a cell membrane. All the binding experiments were carried out with the same formaldehyde-fixed cell preparations (see Materials and Methods), thus excluding the possibility that the differences in the binding profiles of the different hybridomas might be due to differences in cell preparations. To exclude the possibility that the determinant detected by F4/5.10 antibodies on SEYF-a cells but not on RLd-1, EL-4, YAC, B16, or LS178Y might be destroyed on the latter group of cells by the formalization procedure and therefore not detected, we repeated some of the binding experiments with non-fixed living target cells. The

results of these experiments confirmed those obtained with formalin-fixed targets and further emphasized the restricted distribution of the antigen detected by this hybridoma. All the positive supernatants were examined for complement-dependent cytotoxic activity against SEYF-a cells. However, none of them was found to mediate complement-dependent cytotoxicity.

Characterization of Immunoglobulins Produced by the Hybridomas The immunoglobulin subclass of the monoclonal antibodies was determined by immunodiffusion and immunoelectrophoresis with commercially available antisera reagents. Three of the hybrid clones were found to produce IgG2a, three produced IgG1 and one produced IgM (Table 2). Hybridoma products from clones F4/5.10 and F7/129.7 and the product of the myeloma cell line P3 (used as a marker) were biosynthetically labeled with 75Se-methionine, reduced and electrophoresed on SDS polyacrylamide gel. Homogeneous heavy and light chains of the secreted immunoglobulins were

Table 2. Binding characteristics of monoclonal antibodies to various tumor cell lines Hybridoma

Ig subclass

Target cells High binding (> 2,000 cpm/well)

Intermediate binding (1,000-2,000 cpm/well)

Low binding (< 1,000 cpm/well)

No binding

L5178Y

EL-4

KREBS, ELD

R L d l , EL-4, YAC, B16, L5178Y, MFB

F4/2.1

IgG1

SEYF-a, MFB

R L d l , KREBS, ELD, YAC, B16

F4/4.8

IgG1

R L d l , KREBS, YAC, B16, MFB

SEYF-a, EL-4, ELD, L5178Y

F4/5.10

IgG2a

SEYF-a

F6/18.4

ND a

YAC, MFB

SEYF-a, R L d l , EL-4, KREBS, ELD, B16

L5178Y

F7/39

IgG2a

SEYF-a, KREBS, ELD, B16

R L d l , EL-4, YAC, MFB

L5178Y

F7/83.3

IgG2a

YAC, B16, MFB

SEYF-a, RL61, EL-4, KREBS, L5178Y, ELD

F7/129.7

IgM

SEYF-a, EL-4, KREBS, YAC, B16, MFBb, L5178Y

RLc~I, ELD

F7/132.7

IgG1

SEYF-a, EL-4, KREBS, ELD, B16

R L d l , YAC, L5178Y, MFB

P3 x 63Ag8

IgG1

a ND, not done u Very high binding (> 4,000 cpm/well)

SEYF-a, B16, YAC, KREBS, ELD, EL-4, R L d l , L5178Y, MFB

222

seen. The heavy chains of F4/5.10 and P3 were in the v-chain region, whereas the heavy chains of F7/129.7 were in the ~-chain region.

Absorption Experiments In previous work Witz et al. [33] and Klein et al. [19] have shown that syngeneic anti-SEYF-a antisera exhibited a wide cross reactivity with several public specificities expressed on at least one seemingly unrelated murine tumor out of the large panel tested. The question that remained unanswered in those studies was whether most of the multiple 'public' determinants or all of them were expressed on all (or most) of these unrelated tumors, or whether each of the different unrelated tumors in the panel had a restricted repertoire of such specificities. The results presented in Fig. 2 and summarized in Table 2, showing that several different monoclonal antibodies reacted with essentially all the members of the panel, indicate that the first possibility is more plausible. To examine this question in greater detail we performed absorption experiments. For these we selected clone no. F7/129.7, which reacted at a high level with the entire panel of tumor cells and clone no. F4/5.10, which reacted virtually exclusively with SEYF-a ceils. Supernatants from these clones at a dilution of 1 : 4 were absorbed twice with 1.5 × 10 7 SEYF-a, KREBS, ELD, B16, YAC, and L5178Y tumor cells. It was found that two absorption steps with SEYF-a, KREBS, ELD, and B16 tumor cells absorbed the antibodies from clone F7/129.7 almost completely (85%-95%). One absorption with L5178Y and YAC cells removed 40% of the antibodies. On the other hand, when antibodies from F4/5.10 were used only SEYF-a cells (and none of the other tumor cells tested) absorbed considerable amounts of the antibodies (65%). These results, together with the results obtained from the binding experiments, strongly support the assumption that F7/129.7 antibodies are directed against a broadly distributed public determinant, while F4/5.10 antibodies are directed against a specificity that has a narrow distribution. An estimate of the distribution and availability of the antigenic determinants recognized by F7/129.7 and F4/5.10 antibodies on ELD and on SEYF-a cells can be obtained when absorption experiments are performed at a limiting dilution of these antibodies. For this purpose a titration of the F7/129.7 and F4/5.10 antibodies was carried out (Fig. 3). The highest derivatives of both curves are between dilutions 1 : 64 and 1 : 128. Absorption experiments were therefore performed with antibodies produced by the hybridomas F7/129.7 and F4/5.10 at antibody dilutions of

N. Moav et al.: Hybridomas from Tumor-Bearing Mice

1 : 4, 1 : 32, 1 : 64, and 1 : 128 with ELD and SEYF-a cells. It can be seen (Fig. 4) that the antigenic determinant recognized by F7/129.7 antibodies is abundant and equally distributed on ELD and SEYF-a cells, while the antigenic determinant rec-

,0.000

8000 F7/129.7

~ooo ° 4000

2000

2

8

32

128

512

2048

8192

32768

1

Reciprocal of Culture fluid dilution Fig. 3. Binding activities of F7/129.7 and F4/5.10 to SEYF-a cells. Two-fold dilutions were prepared from the hybridoma culture fluids. Their binding activity to SEYF-a tumor cells was monitored by means of an RIA. The results are expressed in cpm bound to the cells above background levels. (A) 1=7/129.7; (D) F4/5.10

loo

_

SEYF-a

S

EI_D

~

6o

x 20

i ELD I I I I

J I

2

I

I

8

I

3I

2 F7/129.7

[

~

128

I

2

I

8

32

128

F4/5.10

Reciprocal of culture fluid dilution

Fig. 4. Absorption of two monoclonal antibodies with SEYF-a and ELD tumor cells. Culture fluids (150 ~tl) from clones F7/129.7 (left) and F4/5.10 (right) (their titration curves are shown in Fig. 3) at the indicated dilutions were absorbed twice with 1.5 × 107 SEYF-a (Q) or ELD (O) tumor cells and their binding to SEYF-a cells was measured by RIA. The results are expressed in percent absorption, calculated from the differences between the bound radioactivity measured before and after absorption

N. Moav et al.: Hybridomas from Tumor-Bearing Mice

ognized by F4/5.10 antibodies is at a considerably higher density on SEYF-a cells than on ELD cells. Discussion

In previous studies [19, 33] we have analyzed serologically detectable antigenic specificities expressed on the membrane of cells from SEYF-a tumor, a polyoma virus-induced sarcoma propagated in A.BY mice. Antisera from hyperimmunized syngeneic mice have been used in these studies. By using exhaustive absorption techniques we showed that SEYF-a cells express several antigeneic specificities on their membranes. Some of these specificities were public, i.e., shared by other seemingly unrelated tumor cells. At least one antigen seemed to be a 'private' antigen, i.e., unique for polyoma virus-induced tumors. The complexity of polyspecific anti-tumor antisera that contain multiple antibody populations directed against various antigeneic determinants does not allow differentation of fine antigenic specificities. Furthermore, the monospecificity of exhaustively absorbed antibody preparations cannot be rigorously established. The somatic cell hybridization technique [20] has been used to produce specific monoclonal antibodies against a variety of cell membrane antigens [2, 97,] including tumor-associated antigens (TAA) [10, 14, 21]. The investigators cited have used spleens from artificially immunized mice as their source of antibody-producing cells. In the present study we have used spleens from tumor-bearing mice for the hybridization. Thus, the clones obtained do reflect the actual humoral response of the host at the time of tumor growth. Indeed, recent studies carried out in our laboratory (Ran, Witz, and Kelin, personal communication) have demonstrated that sera from tumor bearers have a restricted specificity compared with the above-mentioned hyperimmune antisera. To our knowledge only two other groups are currently using spleen cells from tumor bearers for the production of anti-tumor hybridomas [29, 31]. In this study we describe eight lymphocyte hybridoma cell lines that secrete monoclonal antibodies directed against antigens present on the surfaces of SEYF-a tumor cells and on a panel of seemingly unrelated tumors. Seven of these antibodies are directed against public antigenic determinants shared by most or all members of the panel. The individual reactivity patterns of each of the monoclonal antibodies suggest that each is directed against a different membrane antigenic determinant.

223

Comparison of the binding profiles of any of the above antibodies to the different tumor cells in the panel might give an estime of the availability and distribution of the corresponding antigenic determinants on those cells. The antibodies directed against the public specificities thus enable the performance of a dissective analysis of antigenic determinants on tumor cell surfaces, establishment of an antigenic dictionary, and examination of what role, if any, each antigen plays in tumor rejection, as suggested by Klein et al. [19]. One of the hybridomas, F4/5.10, produced antibodies with high binding activity to SEYF-a cells, low binding activity to ELD and KREBS tumor cells, and no binding activity to all the other tumor cells tested (Table 2). These results are in agreement with those obtained by Klein et al. [19], who have shown that ELD cells share an antigenic determinant with SEYF-a cells which is not expressed on YAC, GHA, EL-4, MC57M, and MC57C. To distinguish between the above public antigenic determinant and a postulated SEYF-a-specific antigen they further absorbed the serum with ELD cells until there was no activity left against these cells. They have found that the fully absorbed serum still retained its activity against SEYF-a ascites cells. Therefore, they concluded that ELD cells and SEYF-a cells share another public specificity, in addition to those described in their earlier publication [33]. From our binding experiments we can conclude that the antigenic determinant shared by ELD, KREBS, and SEYF-a, which is recognized by the F4/5.10 monoclonal antibodies, is present at a high density on SEYF-a cells and at a low density on ELD and KREBS. This conclusion is supported by absorption experiments demonstrating that it was possible only at a high dilution of the antibody to absorb the binding activity to SEYF-a cells with ELD cells completely. It is possible that the shared antigenic determinant is specified by a virus picked up during continuous passages of these tumor cells. Some of our monoclonal antibodies demonstrated high binding activity to normal mouse fetal fibroblasts, which might suggest that they are directed against fetal antigens. The possibilty that the differences observed in the binding activity of the antibodies to various tumor cells might be due to Fc receptors (FcR) present on tumor cells was eliminated by the fact that there was no correlation between the presence of FcR on the cell surface and the ability of those cells to bind antibodies (Table 2). For example L5178Y expresses FcR [7]; its binding of antibodies produced by F7/!29.7 is high, and it binds antibodies produced by F6/18.4 poorly. On the other hand, YAC cells do not

224

express FcR [16] but do exhibit high binding of antibodies produced by the F6/18.4 hybridoma (Table 2). Moreover, there was no difference in the binding of the immunoglobulin produced by the P3 x63Ag8 myeloma to the various tumor cell lines whether they possessed high [3, 7] or low [16] densities of Fc receptors (Table 2). Of the eight monoclonal antibodies studied, none was directed exclusively against a specific SEYF-a determinant. It is possible that in the course of screening for anti-SEYF-a antibodies and the process of growing the cells such a hybridoma was not picked up. It is also possible, however, that cells producing antibodies directed toward an SEYF-a-associated private specificity are not triggered and expressed in tumor-bearing mice. Acknowledgements. This study was supported by grant no. ROICA20088 from the National Cancer Institute, DHEW. One of the authors (IPW) is the David Furman Professor of Cancer Immunology.

References 1. Austin FC, Boone CW (1979) Virus augmentation of the antigenicity of tumor cell extracts. Adv Cancer Res 30 : 301 2. Bechtol BK, Brown SC, Kennet RH (1979) Recognition of differentiation antigens of spermatogenesis in the mouse by using antibodies from spleen celt myeloma hybrids after syngeneic immunization. Proc Natl Acad Sci USA 76:363 3. Braslawsky GR, Serban D, Witz IP (1976) Receptors for immune complexes on cells within a polyoma virus-induced murine sarcoma. Eur J Immunol 6:579 4. Fidler IJ (1973) Selection of successive tumor lines for metastasis. Nature (New Biol) 242:148 5. Fidler I J, Gersten DM, Hart IR (1978) The biology of cancer invasion and matastasis. Adv Cancer Res 28:149 6. Fischer GA, Welch AD (1957) Effect of citrovorum and peptones on mouse leukemia cells L5178Y in tissue culture. Science 126:1018 7. Frade R, Kourilsky FM (1977) Preliminary characterization of a glycoprotein having Fc-receptor, properties extraced from a T-cell lymphoma (/_,-5178Y). Eur J Immunol 7:663 8. Fraker PJ, Speck JC Jr (1978) 1,3,4,6,-Tetrachloro 3ct, 6c~,-diphenylglycoluril, a useful reagent for labelling proteins with 125I. Biochem Biophys Res Commun 80: 849. 9. Galfre G, Howe SC, Milstein C, Butcher GW, Howard JC (1977) Antibodies to major histocompatibility antigens produced by hybrid cell lines. Nature 266:550 10. Goodfellow NP, Lennson RJ, Williams E V I I , McDevitt OH (1979) Monoclonal antibodies reacting with murine teratocarcinoma cells. Proc Natl Acad Sci USA 76:377 11. Gorer PA (1950) Studies in antibody response of mice to tumor inoculation. Br J Cancer 4:372 12. Greenwood FC, Hunter WM (1963) The preparation of 131I-labeled human growth hormone of high specific radioactivity. Biochem J 89:114 13. Huang JCC, Berczi I, Froese G, Tsay HM, Sehon AH (1975) A microradioimmunoassay for antibodies to tumor-associated antigens. J Natl Cancer Inst 55:879

N. Moav et al.: Hybridomas from Tumor-Bearing Mice 14. Kennett RH, Gilbert F (1979) Hybrid myelomas producing antibodies against a human neuroblastoma antigen present on fetal brain. Science 203:1120 15. Kennett RH, Denis KA, Tung AS, Klinman NR (1978) Hybrid plasmacytoma production: Fusion with adult spleen cells, monoclonal spleen fragments, neonatal spleen cells and human spleen cells. Curt Top Microbiol Immunol 81 : 77 16. Kerbet RS (1980) Increased sensitivity of rosetting assay for Fc receptors obtained by using non-hemagglutinatingmonoclonal antibodies. J Immunol Methods 34:1 17. Klein G, Forssberg A (1954) Studies on the effect of x-rays on the biochemistry and cellular composition of ascites tumors. Exp Cell Res 6:211 18. Klein E, Klein G (1964) Antigenic properties of lymphomas induced by the Moloney agent. J Natl Cancer Inst 32:547 19. Klein G, Ehlin B, Witz IP (1979) Serological detection of a polyoma-tumor-associated membrane antigen. Int J Cancer 23 : 683 20. K6hler G, Milstein C (1975) Continuous cultures of fnsed cells secreting antibody of predefined specificity. Nature 256 : 495 21. Koprowski H, Steplewski Z, Herlyn D, Herlyn M (1978) Study of antibodies against human melanoma produced by somatic cell hybrids. Proc Natl Acad Sci USA 75:3405 22. Leammli KK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680 23. Moav N, Witz IP (1978) Characterization of immunoglobulins eluted from murine tumor cells: Binding patterns of cytotoxic antitumor IgG. J Immunol Methods 22:51 24. Nisonoff A, Markus G, Wissler FC (1961) Separation of univalent fragments of rabbit antibody by reduction of a single labile dissulfide. Nature 189:293 25. Ouchterlony O (1948) In vitro method for testing the toxin-producing capacity of diphteria bacteria. Acta Pathol Microbiol Scand 25:186 26. Ran M, Klein G, Witz IP (1976) Tumor-bound immunoglobulins. Evidence for the in vivo coating of tumor cell by potentially cytotoxic anti-tumor antibodies. Int J Cancer 17: 90 27. Scheidegger JJ (1955) Une micro-m6thode de l'immuno61ectrophor6se. Int Arch Allergy Appl Immunol 7:103 28. Sendo F, Aoki T, Boyse EA (1975) Natural occurrence of lymphocytes showing cytotoxic activity of BALB/c radiation-induced leukemia RLd'I cells. J Natl Cancer Inst 53 : 603 29. Simrell RC, Klein AP (1979) Antibody responses of tumor-bearing mice to their own tumors captured and perpetuated as hybridomas. J Immunol 123:2386 30. Sjogen HD (1964) Studies on specific transplantation resistance to polyoma virus induced tumors. I. Transplantation resistance induced by polyoma virus infection. J Natl Cancer Inst 32:361 31. Tax A, Manson LA (1981) Monoclonal antibodies against antigens displayed on a progressively growing mammary tumor. Proc Natl Acad Sci USA 78:529 32. Weiss DW (1980) Tumor antigenicity and approaches to tumor immunotherapy. Curr Top Microbiol Immunol 89:1 33. Witz IP, Lee N, Klein G (1976) Serologically detectable specific and cross-reactive antigens on the membrane of a polyoma virus-induced murine tumor. Int J Cancer 18:243

Received June 22/Accepted November 19, 1981