Znt. J . Cancer: 17, 90-97 (1976)

TUMOR-BOUND IMMUNOGLOBULINS. EVIDENCE FOR THE I N VIVO COATING OF TUMOR CELLS BY POTENTIALLY CYTOTOXIC ANTI-TU MOR ANTIBODIES Maya RAN George KLEIN and Isaac P. W I T Z Department of Microbiology, the George S. Wise Center for Life Sciences, Tel Aviv University, Tel Aviv, Israel; and Department of Tumor Biology, Karolinska Institute, Stockholm, Sweden.

Summary. The experiments described herein were designed to determine whether part of the Ig coat of tumor cells consists of specific anti-tumor antibodies. It was demonstrated that the inoculation of polyoma virus-induced sarcoma cells ( S E YF-a) into syngeneic A.B Y mice stimulates the production of cytotoxic antibodies against the tumor-cell population. The level of these antibodies, which was undetectable during the first week after transplantation, increased markedly during the second week, and remained high thereafter. Following the increase in cytotoxic antibodies in the serum, a cell-bound potentially cytotoxic antibody was detected on the tumor cells by testing their sensitivity to rabbit complement. The increase in cell-bound, potentially cytotoxic antibody followed the kinetics of the increase in serum antibody during the second week after transplantation and was inversely correlated to the amount of free antigens on the cell surface. These antigens, responsible for the sensitivity of the cells to a syngeneic hyperimmune cytotoxic antiserum, became non-available for the cytotoxic antibodies during propagation of the tumor cells. Cells from a tumor propagated for 3 weeks could not compete f o r antitumor antibodies with cells propagated for I week. Yet it was possible to increase the antigenic capacity of cells from an old tumor by a treatment that would cause the release of tumor-associated Ig. Cytotoxic anti-SE YF-a antibodies could be dissociated from tumor cells propagated in vivo by methods causing dissociation of antigen-antibody complexes, and detected in tumor eluates. The association of immunoglobulins (Igs) with in vivo propagating tumors has been well established for various animal model systems (Witz, 1973) and for human cancer (Ran et al., 1974 b ; Vanky et al., 1975; Gopta and Morton, 1975; Irie et al., 1975). The significance of this in vivo phenomenon in the tumor host relationship is supported by the following biological activities. (1) Enhancement of tumor growth by Ig-containing tumor eluates (Bansal et al., 1972; Ran and Witz, 1972). (2) Blocking of in vitro cell-mediated anti-tumor activity by tumor eluates of human and animal origin (Bansal et al., 1972;

Sjogren and Bansal, 1971; Sjijgren et al., 1972). ( 3 ) Demonstrations of an increased ability of human tumor cells to stimulate DNA synthesis in autochthonous lymphocytes after unmasking of tumor antigens (Stjernsward and Vanky, 1972), and an inverse relationship between this capacity and the amounts of cancer-associated Ig (Vanky eta!., 1975). (4) A direct correlation between presence of lg and the degree of malignancy of human cancer (Ran et al., 1974 b). However, no direct evidence has so far been presented that anti-tumor antibodies were in fact part of the Ig coat of in vivo growing tumor cells. This study was aimed at determining whether part of the Ig coat of murine tumor cells induced by polyoma virus consists of anti-tumor antibodies. MATERIAL AND METHODS

Tumor cells The tumor used in the present study was a polyoma-virus-induced ascites sarcoma SEYF-a (Sjogren, 1964a, b). Two million cells per mouse were inoculated IP to syngeneic A.BY mice. Cells were harvested between 6 and 30 days after inoculation. Survival time of SEYF-a-bearing mice was usually 1 month, and occasionally longer. Antiserum Syngeneic anti-SEYF-a antiserum was prepared by multiple injections of tumor cells, exposed to 6000 R , into syngeneic A.BY mice. The antiserum exhibited specific cytotoxicity for SEYF-a cells after absorption with several other rnurine tumors (Witz and Klein, unpublished data). Sera from tumor-bearing mice Sera were collected by bleeding from the retroorbital sinus, on different days after inoculation. Complement Normal rabbit serum (NRS), selected for its lack of cytotoxicity towards young SEYF-a cells (propagated Received: October 1, 1975.

COATING OF TUMOR CELLS BY ANTIBODIES

in vivo for only 7 days), served as source of complement (C'). Excess complement, as determined by a doseresponse curve (at a final dilution of 1 :15-1 :20) was used in cytotoxicity assays. For the determination of cell-bound cytotoxic antibodies we used three dilutions of NRS (1 :15, 1 :30, 1 :45) causing respectively the lysis of 100,50 and 10-20% of indicator cells in the presence of a standard antiserum. Absorption of NRS was performed at 0" C for 60niin. The NRS was treated three consecutive times, either with mouse liver powder (v/v) or with washed SEYF-a cells. A ratio of 1 vol cells: 1 vol antiserum was used with cells propagated in vivo for 3 weeks. In other cases a cell surface area equal to that of 3-week-old cells was used. Cytotoxicity assays

Complement-dependent cytotoxicity (CdL) mediated by syngeneic antisera and by serum of tumorbearing mice was measured by using S'Cr-labelled SEYF-a cells propagated in vivo for 7 days as targets. Ten million washed cells were suspended in I ml of Hank's balanced salt solution (HBSS) containing 10% calf serum. Two hundred pCi of 51Cr (sodium chromate, specific activity 100-300 mCi/mg, Radiochemical Centre, Amersham, England), were added to the cells. The mixture was incubated for 20 min at 37°C with gentle shaking, washed five tinies with HBSS and resuspended at a concentration of 2 x lo6 cells/ml. The assays were perfomed in glass tubes, by adding 1 x lo5 cells in a volume of 0.05 ml to 0.05 nil serum dilution and 0.05 ml of complement. The tubes were incubated for 45 min at 37" C in a CO, incubator. The reaction was stopped by adding 1 ml of cold HBSS to each tube and centrifuging at 1,OOOxg for 5 min. Two-tenths/ml of supernatants were removed and counted in a Packard model 3002 spectrometer. Cytotoxicity index (CI) was calculated from the mean of triplicate tubes according to the following formula: test cpm-control cpm (C') CI = total cpm-control cpm (C')' Total cpm was evaluated by the cpm released after three successive freezings and thawings of cell samples. Control cpm was the radioactivity released following incubation with complement without antiserum. Cell-bound, potentially cytotoxic antibody

The presence of cell-bound, potentially cytotoxic antibody was evaluated as follows: NRS at dilutions of 1:15, 1:30 and 1:45 was added to 51Cr-labelled SEYF-a cells harvested after different propagation periods in vivo. CI values for cell-bound, potentially

91

cytotoxic antibody were calculated as above. Control cpm in this case represented the amount of radioactivity released with medium, while test cpm represented the amount released after addition of complement alone. Antigenic expression

The antigenic expression of SEYF-a cells was measured by the following assays: Sensitivity to CdL. Sensitivity of 51Cr-labelledcells to CdL mediated by a standard dilution (1/32) of the syngeneic anti-SEYF-a antiserum and in the presence of three different dilutions of NRS. The cytotoxicity assays proceeded as described above. The CI values calculated expressed the total antigenic expression of the cells. The net antigenic expression of cells (representing the amount of free antigenic sites) was calculated according to the following formula: CI caused by a I :32 antiserum dilution+(=', minus CI caused by C' alone. Competition assays. Unlabelled competitor SEY F-a cells were added to incubation mixtures containing labelled target cells, a I :32 dilution of antiserum and complement. The percentage of inhibition was calculated on the basis of control mixtures to which no competitive cells were added (the CI of controls was around 0.50). The ability of cells propagated in vivo for different periods to compete for cytotoxic antibodies was thus measured. Absorption assays. Cells propagated in vivo for one or three weeks were incubated with antiserum for 60min at 0°C. Absorption conditions were as follows: 1 x lo7 cells propagated in vivo for 3 weeks, or 1.6 x lo7cells propagated in vivo for 1 week were mixed, each with 0.1 ml undiluted antiserum. The number of cells used was calculated so as to give equal cell surface area for absorption.

Cells after a short-term culture (a treatment causing release of Ig from the cell surface (Ran et al., 1974a), were also used for absorption. Tumor eluates

Tumor cells, harvested on various days after inoculation, were washed three times in HBSS and then treated by one of the following procedures: Supernatant from short-term cultures. Cells were put into short-term culture (Ran et al., 1974a), the supernatants were collected, concentrated by vacuum dialysis, dialysed against phosphate-buffered saline (PBS) and tested for the presence of cytotoxic antiSEYF-a antibodies. Citrate buffer treatment. Cells were treated with citrate buffer at PH 3.5, 1 ml buffer was added to

0

0

0 0

0 0 0

'O0I 0

0

+I

v

0

0

50-

0

0 0 0 0

00

Days after injection

FIGURE1 The presence of cytotoxic anti-SEYF-a antibodies in sera of tumor-bearing mice. The CI caused by a 1:16 serum dilution and rabbit complement in excess to SEYF-a cells propagated in vivo for 7 days is represented for sera drawn on various days after tumor injection. Cytotoxicity was assayed by W r release.

1.5 x lo8 cells for I5 min at room temperature. Cells were then spun down and the supernatant collected and centrifuged at 12,OOOxg for IOmin, dialysed against PBS a n d tested for the presence of cytotoxic anti-SEYF-a antibodies. RESULTS

Presence of potentially cytotoxic anti-tumor antibodies in the circulation of tumor-bearing mice

A.BY mice, inoculated IP with 2 ~ 1 SEYF-a 0 ~ cells were bled on various days after inoculation through the retro-orbitol sinus. The sera were assayed for their capacity to mediate complementdependent lysis (CdL) of 51Crlabelled SEYF-a cells propagated in vivo for 7 days. Normal rabbit serum (NRS) served as source of complement and was used in excess. The cytotoxic capacity of tumor-bearing sera, expressed in terms of cytotoxicity index (CI) for a 1 :16 serum dilution, is presented in Figure I . No detectable levels of cytotoxicity appeared in sera drawn during the first week of tumor propagation in vivo. Even when these sera were tested a t higher concentrations, i.e. 1 :2 or 1 :4,n o detectable cytotoxicity could be demonstrated. Cytotoxic antibodies were detectable 9 days after inoculation and markedly increased at the end of the second week. Cytotoxicity remained high throughout the entire survival time of tumor-bearing mice and the titer (the serum dilutions causing 50% death of cells) exceeded 1:128, or even 1:256. The pattern of cytotoxicity levels expressed in terms of cytotoxicity titer (results not shown) was essentially the same as that represented in Figure 1.

The specificity of the antibodies in terms of reactivity towards other murine tumor cells is under investigation. No cytotoxic activity against normal A.BY lymphocytes has been demonstrated. Demonstration of poten fially cytotoxic antibodies on the surface of in vivo propagated tumor cells I n the previous section we demonstrated the presence of cytotoxic anti-tumor antibodies in A.BY mice bearing the syngeneic SEYF-a tumor. Experiments were designed to determine if such antibodies localized in the tumor and were coating the surface of in vivo propagated cells. This was done by determining whether or not in vivo propagated cells are sensitive to cytotoxicity by complement alone as compared with their in vitro cultured counterpart, which is obviously uncoated with antibody. Figure 2 shows that SEYF-a cells propagated in vivo for 13 days are sensitive to lysis mediated by complement whereas in vitro cultured cells are essentially insensitive. The source for complement in these experiments was fetal rabbit serum. A typical dose-response curve was obtained only with the in vivo propagated cells. Experiments were designed to establish whether or not the time-dependent increase in the titer of circulating cytotoxic antibodies in tumor-bearing mice is paralleled by a localization of these antibodies o n the tumor cells. This would be reflected by a n increase in the sensitivity of SEYF-a cells to the cytotoxic effect of complement. In these experiments normal rabbit serum (NRS) was used as source of complement. Figure 3 demonstrates that cells

1.00

1

s 0s o

-

~

'/I

'/4

]/I6

116 4

COMPLEMENT DILUTION

FIGURE 2 Sensitivity of in vivo propagated SEYF-a cells to the cytotoxicity of complement alone. CI values caused by two-fold dilutions of C (fetal rabbit serum) to 13-day-old or their tissue-cultured counterSEYF-a cells (A-A) part (0-0) are plotted. Cytotoxicity was measured by trypan blue exclusion.

93

COATING OF TUMOR CELLS BY ANTIBODIES

I

NRS

and heat-inactivated (56"C for 30 min) NRS was tested as a source of potential antibodies. A mixture of this absorbed and inactivated NRS with absorbed NRS as a source of C' did not cause increased lysis above that caused by absorbed NRS alone (Table I). The absorbed NRS had complement activity with different exogeneous antisera (Table I). These results exclude the possibility that natural antibodies were responsible for the cytotoxic effect of NRS for in vivo propagating cells.

Absorbed NRS

I

1.00

1

2 W..k.

Saturation of antigenic determinants on in vivo propagated tumor cells

3 .(I.,

,n,.ct,on

FIGURE3 Sensitivity of SEYF-a cells propagated in vivo for I , 2 or 3 weeks to cytotoxicity mediated by absorbed or non-absorbed NRS. CI values caused by the addition of a 1 :15 C dilution (empty histograms) or a 1:30 C' dilution (hatched histograms) are represented. The sensitivity of cells propagated in vivo for 2-3 weeks was significantly higher than that of 1-week-old cells, for both absorbed and non-absorbed NRS (p

Tumor-bound immunoglobulins. Evidence for the in vivo coating of tumor cells by potentially cytotoxic anti-tumour antibodies.

Znt. J . Cancer: 17, 90-97 (1976) TUMOR-BOUND IMMUNOGLOBULINS. EVIDENCE FOR THE I N VIVO COATING OF TUMOR CELLS BY POTENTIALLY CYTOTOXIC ANTI-TU MOR...
679KB Sizes 0 Downloads 0 Views