In vitro synthesis of tumor-specific factors with blocking and antibody-dependent cellular cytotoxicity (ADC) activities

Int. J. Cancer: 16, 932-941 (1975)

IN VITRO SYNTHESIS OF TUMOR-SPECIFIC FACTORS WITH BLOCKING AND ANTIBODY-DEPENDENT CELLULAR CYTOTOXICITY (ADC) ACTIVITIES

Karen NELSON, Sylvia B. POLLACK and Karl Erik HELLSTROM Departments of Pathology and Microbiology and ImmunoIogy,

University of Washington Medical School, Division of Immunology, Fred Hutchinson Cancer Research Center, Seattle, Washington, USA

Spleen cells from BALBIc mice bearing syngeneic sarcomas and from BALBIc mice whose sarcomas had been excised were cultivated in vitro. The culture supernatants were tested for two activities: their ability ( I ) to suppress (" block ") speciJic lymph-node cell-mediated cytotoxic reactions directed against the respective neoplasms; and ( 2 ) to induce antibody-dependent cellular cytotoxicity ( A DC) specific for the antigens of the respective tumors. Both specific activities, blocking and induction of A DC, were detected in cuIture supernatants from spleens of tumor-bearing mice, even when repeatedly harvested at intervals over a 7-day period. Supernatants of cultured spleen cells from mice whose sarcomas had been excised 3-4 weeks previously also had A DC but no bIocking activity. Supernatants of cultures treated with inhibitors of protein synthesis lacked both blocking and A DC activities: the inhibitory efect of cycloheximide on these activities, as well as on protein synthesis, was reversibIe. Factors in the culture supernatants responsible for blocking and for ADC were labelled when the culture were incubated with T-leucine. The labelled material was retained by, and could be eluted from, immunoadsorbents for mouse immunoglobulins. In addition, the labelled material bound preferentially to those tumor cells for which specific blocking or ADC activities were observed. The findings indicate that factors responsible for the blocking and ADCphenomena were indeed synthesized by the spleen cells in vitro.

Spleen cultures derived from mice carrying growing sarcomas release factors which can suppress (" block ") lymphocyte-mediated cytotoxicity to target tumor cells carrying the respective antigens (Nelson et al., 1975a). Culture supernatants, both of spleen cells from tumor- bearing mice and of spleen cells from mice whose tumors have been excised, can sensitize tumor target cells to destruction by non-immune lympho-

cytes in a manner resembling antibody-dependent cellular cytotoxicity, ADC (Nelson et al., 19756). The factors mediating the blocking and the ADC activities can both be removed by passage of supernatants through anti-immunoglobulin immunoadsorbents, and they can be subsequently eluted from such immunoadsorbents (Nelson et al., 1975a, b). Treatment of spleen cells with anti-theta serum and complement stops the release

Received: August 19, 1975. Abbreviations used in this paper: ADC = antiserum-dependent cellular cytotoxicity; CMC = cell-

mediated cytotoxicity; Ig = immunoglobulin; LNC = lymph-node cells; PBS = phosphate-buffered saline; and TCA = trichloroacetic acid.

932

IN VITRO SYNTHESIS OF TUMOR-SPECIFIC IMMUNOGLOBULINS

of blocking factors into the medium (Nelson et al., 1975a, c), but does not influence the release of factors responsible for ADC (Nelson et al., 19756, c) , suggesting that different factors are responsible for the blocking and the ADC phenomena.

A crucial question is whether the molecules responsible for the two discrete activities seen in this system are merely released from the spleen cells or whether they are synthesized in vitro prior to release. The present study was conducted to investigate this question. One approach was to examine whether factors with blocking or ADC activity would be labelled when 14C-labelled leucine was added to the medium in which the spleen cells were cultivated. Another approach was to study the effects of two inhibitors of protein synthesis, cycloheximide and puromycin, on the blocking and ADC activities of culture supernatants. Cycloheximide was used in particular because its effects are reversible (Ennis et a!., 1 964).

The findings suggest that there was an active production, rather than just a release, of factors with blocking activity and with ADC activity in the spleen cultures studied.

MATERIAL AND METHODS

Mice

Mice of the BALB/c strain were bred by brother-sister mating in this laboratory. In addition, female BALB/c mice were obtained from Jackson Laboratories, Bar Harbor, Maine, USA. Mice were sex- and age-matched in each experiment.

Tumor lines

Sarcomas were induced in adult BALB/c mice by subcutaneous injection of 3-methylchol- anthrene. They were maintained by serial trans- plantation in vivo; tumors were used when in the fourth to eighth transplant generation.

Mice were inoculated subcutaneously with lo5 sarcoma cells. Spleens were taken for culture either when the tumors measured approximately 10 mm in diameter (" tumor-bearers ") or 3 to 4 weeks after surgical excision of the tumor (" tumor-excised "). Each sarcoma was explanted and maintained in vitro as a monolayer culture to provide a source of target cells for the microcytotoxicity assays.

Spleen cultures

Spleens were remnoved from tumor-bearing, tumor-excised, and non-immunized mice immedi- ately following cervical dislocation. Preparation of the spleen cell suspensions was done at 4" C in RPMI-1640. The cells were gently teased free of the capsule and the fragments sedimented at 1 xg. Erythrocytes were removed from the single-cell suspension by hypotonic shock, after which the remaining cells were washed twice. The spleen cells were diluted to 5 x los to lo7 cells per ml of culture medium and placed in Erlen- meyer flasks at 37" C in a moist, 5 % CO, in air atmosphere.

The culture medium was RPMI-1640 buffered with NaHCO, and supplemented with 20% heat- inactivated (30min at 56" C ) fetal calf serum, 1 % sodium pyruvate, 2% L-glutamine, 100 IU per ml of penicillin and 100 pg/ml of streptomy- cin. Gamma-globulin-depleted fetal calf serum was used in the experiments presented in Tables TI-TV.

Cells were spun free of the culture supernatants at 2- to 3-day intervals, washed twice and recultured at the original concentration in fresh medium. The cultures were discarded if the total number of viable cells was less than 80% of the original number. Cell viability was assessed by trypan blue dye exclusion. The culture supernatants were passed through a 0.4511 filter (Millipore Corporation, Bedford, Maine, USA) and stored at -70" C until assayed.

In vitro assay of the activity of culture supernatants

Supernatants and eluates from anti-immunoglobulin columns (see below) were assayed for their ability to inhibit (block) the anti-tumor cytotoxicity of immune LNC and for their ability to mediate antiserum-dependent cellular cytotoxicity (ADC) by non-immune LNC in a modified microcytotoxicity assay (Hellstrom and Hellstrom, 1971). Supernatants and eluates were heat-inactivated and assayed undiluted unless otherwise indicated. Target cells plated in the wells of Falcon No.3040 Microtest plates (Falcon Plastics, Oxnard, Calif., USA), received 0.1 ml supernatant, which was decanted 30 min later, when lo5 effector cells were added in 0.1 ml of medium. Immune lymph-node cells (LNC) were obtained from the lymph nodes of mice 5 days after a second intraperitoneal injection of lethally

933

NELSON ET AL.

irradiated tumor cells (Nelson et al., 197%). Control LNC were obtained from the lymph nodes of non-immunized mice. The wells received 0.05 ml of medium containing 30% fetal calf serum 45min after the effector cells had been added. The plates were incubated for 36 to 40 h, stained with crystal violet, and the number of surviving target cells determined visually; all plates were coded and read '' blind ".

Specificity was determined by testing each supernatant and effector cell treatment on a non- cross-reacting target cell line. The control of treating tumor cells with supernatants alone was also included. A few (6%) of all culture supernatants had to be excluded from these studies because they were cytotoxic to the target cells in the absence of LNC.

Calculations

The calculations of percentage cell-mediated cytotoxicity (percentage CMC), percentage blocking activity and percentage ADC activity have been described (Nelson et af . , 1975a). In this study, wells receiving LNC and culture supernatant of spleens from non-immunized mice served as controls. Each treatment group included 8 to 12 wells. The significance of the differences between the means of the treatment groups was determined by Student's t-test. The statistical analysis of the assays will be discussed in detail elsewhere (Brown et al., manuscript in preparation).

Determination of synthesis of protein in spleen cell cultures

Experiments were performed in which spleen cells were cultured in leucine-free RPMI-1640 supplemented as indicated above. 14C-leucine (56 mCi/mmol, Amersham Searle, Chicago, Ill., USA) was then added to a final concentration of 0.1 ,uCi/ml. Freshly prepared stock solutions of cycloheximide or puromycin dihydrochloride (Sigma Chemical Company, St. Louis, Mo., USA) were added to the cultures in the concentrations indicated in " Results ".

In the experiment shown in Table 11, three aliquots of los spleen cells were removed from cultures after 24 h of exposure to cycloheximide and another three aliquots after an additional 24 h culture without cycloheximide. The cells were disrupted by freeze-thawing, suspended and washed twice in 1 ml of 10% TCA and the pre-

cipitates dissolved in 0.2 ml of distilled water and mixed with 1.8 ml of Quantafluor (Mallinkrodt, St. Louis, Mo., USA). The incorporated radioactivity was determined by counting in a Packard Tri-Carb beta scintillation counter.

Murine immunoglobulins were removed from the culture supernatants by passage through immunoadsorbent columns. Columns were prepared by coupling the 7s fraction of a rabbit anti-mouse immunoglobulin hyperimmune antiserum to Affi-gel 10 (Bio-labs, Richmond, Calif., USA). This anti-serum binds mouse IgG,, IgCz and IgM (Pollack er al., 1975). One hundred mg of protein in 20ml of 0.1 M phosphate-buffered saline (PBS) was added to 1 g of Affigel. The slurry was shaken at 4" C for 2 h, poured into a plastic syringe and washed with PBS until the absorption at 260 and 280nm was negative. Approximately 60 % of the protein was bound to the gel. The specificity of the columns was determined by fractionating normal mouse serum. The column effluents and eluates were compared to the original serum by immunoelectrophoresis and radial immunodiffusion.

A volume of supernatant equal to 90% of the void volume of the column was applied to the gel. The column was incubated at 4" C for 30 min and the unbound proteins removed by washing with 0.05 M PBS. The bound immunoglobulin was eluted from the column by the method of Dandliker and DeSaussure (Dandliker and DeSaussure, 1968). A volume of 3 M NaSCN equal to 10 % of the void volume was allowed to enter the gel. The immunoglobulin was immedi- ately eluted by washing the column with PBS. This eluate was dialysed against 100 vol of PBS in an Amicon stirred cell over a PM-I0 membrane (Amicon Corporation, Lexington, Mass., USA), concentrated to the volume of the original supernatant, and stored in aliquots at -70" C until assay.

The protein concentration of the column eluates was determined by comparing the absorption at 280nm with that at 260 nm. The amount of incorporated radioactivity in the eluates was determined by counting 0.1 ml of eluate mixed with 0.9 ml of Quantafluor. Determinations were made in quadruplicate.

In the experiment presented in Table IV, 1 ml aliquots of column eluate were mixed with 5 x lo5 sarcoma cells from in vitro cultures. Following a 30-min incubation at 4" C, the tumor cells were

934


washed twice, mixed with immune or control LNC and plated in the wells of a Microtest plate. The number of surviving tumor cells was determined after a 30 h incubation. Triplicate aliquots of lo4 cells of each tumor cell suspension were mixed with Quantafluor and the amount of bound radioactivity was determined.

RESULTS

Demonstration of both blocking and A DC activities in supernatants from spleen cells obtained from tumor-bearing mice and of ADC activity from spleen cells of tumor-excised mice.

Table I summarizes experiments which served as a basis for the studies described in the following sections. The data show that, when spleen cells from mice bearing a syngeneic transplanted sarcoma were cultured in vitro for 2 days, the culture supernatants specifically inhibited (blocked) the cytotoxicity of lymph-node cells (LNC) immune to that tumor. It also shows that the supernatants specifically sensitized tumor cells to injury by control LNC, an activity resembling the ant i-serum-dependent cellular cyt otoxicity (ADC) detected in the sera of many tumor- bearing and tumor-immune mice (Pollack et al., 1972). In each of these experiments, both the blocking and the ADC activities were specific for the tumor borne by the spleen donor, i.e. cell- mediated cytotoxicity to a non-cross-reacting tumor was neither blocked nor induced.

When spleens were obtained from mice 3-4 weeks after surgical excision of a transplanted sarcoma, the culture supernatants had significant ADC activity (p<O.OOl) but did not block specific cell-mediated immunity (Table I). As

noted before (Nelson et al., 1975c), such results suggest that different factors mediate the blocking and ADC activities.

Continued release of factors with blocking and A DC activities from spleen cultures maintained in vitro

If factors with blocking or ADC activity are synthesized rather than eluted from immune spleen cells in vitro, then the factors should continue to appear in the culture supernatants following repeated harvestings. To determine whether both activities did continue to appear after the initial harvest on day 2, the spleen cells were spun free of their culture medium, washed twice and recultured at the original concentration in fresh medium. This procedure was repeated on days 4 and 7. In each of four separate experiments, blocking and ADC activities continued to appear in the supernatants at similar levels throughout the 7-day period (Fig. l), suggesting that synthesis was occurring.

Eflect of inhibition of protein synthesis on the blocking and A DC activities of culture Supernatants

The question of whether the factors responsible for blocking and ADC activities in spleen cell supernatants were synthesized in vitro was explored directly by testing the effects of inhibitors of protein synthesis on the appearance of blocking and ADC activities in the culture supernatants, as well as on de novo protein synthesis.

Cycloheximide was added at the time of initiation of spleen cultures from mice bearing sarcoma 1358 and from non-immunized controls. Cultures were treated with cycloheximide at final concen-

TABLE I

MEAN BLOCKING AND ADC ACTIVITIES DETECTED IN CULTURE SUPERNATANTS OF SPLEEN CELLS FROM MICE IMMUNE TO MCA-INDUCED SARCOMAS

Culture supernatants of spleens from Mean percentage Mean percentage Number of blocking activity&sD ADC activity+so culture supernatants

Mice bearing MCA-induced sarcomas 1 2 1 2 1 31612 18 Mice whose MCA-induced sarcomas

were surgically removed - 35 &41 4 6 1 4 3

‘The two means of percentage blocking activity are significantly different at the p<O.OOI level. The two means of percentage ADC activity are significantly different at the pi0.05 level.

935

NELSON ET A L .

100

80

I 6o cn -s 5 40 ,D

8 9

20

0 I I I 2 4 7

Doys in culture

50

40

P 30 a

5 s 20 s

a

.. b T

10

0

FIGURE 1 Continued appearance of blocking and ADC

activities in spleen cell culture supernatants. Each point represents the mean activity of four culture supernatants of spleens from mice bearing MCA- induced sarcomas. Culture supernatants were sepa- rated from the spleen cells on days 2, 4 and 7 after initiation of the cultures. The spleen cells were washed and recultured in fresh medium after each harvest.

100

80

I 60 s

a 8 40

.x D

20

0

\ ;-----as-

-.* -- I I I

-. 0 0 5 5 0 500 5000

Cycloheximide (,pg/ml)

50

40

P a 30 b a

P o\o

20

10

0

FIGURE 2 Inhibition of the appearance of blocking and ADC

activities in spleen cell culture supernatants by increasing concentrations of cycloheximide. Cyclo- heximide was added when parallel cultures of spleen cells from tumor-bearing or non-immune mice were initiated. Supernatants were harvested 24 h later, the immunoglobulin fractions were obtained by immunoadsorption, and were assayed in the same experiment.

trations ranging from 0.5-500.0 yg/ml or were untreated controls. All cultures received 14C- leucine (0.1 ,uCi/ml) and the supernatants were harvested after 48 h. Since factors mediating blocking and ADC activities can be bound to and eluted from an anti-mouse immunoglobulin affinity column (Nelson et al., 1975a, b) the culture supernatants were fractionated onaffinity columns and the column eluates assayed for blocking and ADC activities. As shown in Figure 2, while both activities were present in the untreated controls and in cultures treated with 0.5 ygfml of cycloheximide, neither activity was detected in supernatants of spleen cells cultured with 5 ygfml or higher levels of cycloheximide.

To investigate whether the cycloheximide effects observed in these cultures were due to a reversible depression of protein synthesis, rather than nonspecific cytotoxicity to the spleen cells by the drug, spleen cells were cultured with 50 yg/ml of cycloheximide for 24 h, the supernatants removed and the spleen cells washed. Following an additional 24 h of culture without cycloheximide, the supernatants were again harvested. Aliquots of spleen cells were also removed after the first and second 24 h incuba- tions and the radioactivity incorporated in their TCA-precipitable protein determined (Table 11). Protein synthesis was depressed by 98% in the presence of cycloheximide but returned to the level of the controls during the second (cycloheximide-free) 24 h.

The supernatants of these cultures were fractionated on the anti-immunoglobulin affinity columns and the eluates assayed for blocking and ADC activity (Fig. 3). As with total protein synthesis, the synthesis of both the blocking and the ADC activities was severely depressed by the cycloheximide, and the effect of the drug was reversible, i.e. both the blocking and the ADC activities were detected in supernatants of cultures from which cycloheximide had been removed 24 h earlier. Figure 3 also illustrates the specificity of the blocking and ADC activities of these supernatants : the cell-mediated response to MCA-1386 was not affected by culture supernatants of spleen cells from mice bearing MCA- 1388.

In an experiment employing another inhibitor of protein synthesis, puromycin, similar results were obtained (Table 111). The addition of puromycin when the cultures were initiated

936


TABLE I1

REVERSIBLE DEPRESSION OF PROTEIN SYNTHESIS IN MURINE SPLEEN CELLS BY CYCLOHEXIMIDE

Cycloheximide Mean CPM*so of the protein Percentage depression (aelml) from 10' spleen cells: of protein synthesis Cultured spleen cells from '

Non-immune mice 0.05 , 50.00

Mice bearing MCA-1388 0.05 $9 50.00

Non-immune mice 0.05 50.00

Mice bearing MCA-1388 0.05 50.00

Cultured with cycloheximide

2,881 f615 56~k-425

4,057 41447 71* 4

Cultured after removal of cycloheximide

2,6461270 2,605 i 572 3,066+180 . 3,111 f764

98

98

1

-1

Spleen cells were cultured with cycloheximide for 24 h and were allowed to recover in medium without cycloheximide for an additional 24 h. "C-Ieucine (0.1 pCi/mI) was included in the medium. Cellular protein was obtained by freeze-thawing and precipitation with 10% TCA.

40-1 c. 1; ... . . . i d' ILL -23 3 0-24 24-48 u -

0-24 24-48 Hours of culture

FIGURE 3 Cycloheximide-induced depression of the synthesis

of blocking and ADC activities in spleen cell cultures and recovery of synthesis following removal of the drug. Spleen cells from mice bearing MCA-1388 and non-immune mice were cultured with cycloheximide at 0.05pg/ml (0) or 50pg/ml (if;!) for 24h. The supernatants were then removed and the cells washed and cultured for an additional 24 h without cycloheximide. The immunoglobulin fractions were obtained from the eight culture supernatants by affinity chromatography and assayed on MCA-1388 targets (a and c) and MCA-1386 targets (b and d). The control wells contained 70-80 target cells. The percentage CMC of LNC immune to 1388 was 45. The percentage CMC of LNC immune to 1386 was 42.

depressed both the synthesis of protein (as determined by incorporation of 14C-leucine) and the synthesis of the factor(s) blocking cell- mediated reactivity to tumor cells. The synthesis of the factor(s) mediating A D C was similarly depressed (Fig. 4).

Binding of 14C-leucine-labelled immunoglobulin to tumor cells

We next tried t o determine whether the labelled immunoglobulin fraction of culture supernatants

0 100 10-2 I O - ~ I O - ~

Di lut ion FIGURE 4

ADC was not detected in the culture supernatant of spleen cells treated with puromycin (A-A). Culture supernatant of the same spleen cells cultured without puromycin had ADC activity (0-0). The spleens were obtained from mice bearing an MCA- induced sarcoma.

937

NELSON ET AL.

TABLE 111

SYNTHESIS OF IMMUNOGLOBULIN (Ig) AND BLOCKING ACTIVITY HALTED BY PUROMYCIN '

Mean number of MCA-1358 cells Ig from culture supernatant p u ~ f d O ~ ~ i n E:Eg fse after exposure to Percentage Percentage

of spleens from CMC blocking to Of Ig Normal LNC LNC immune to 1358

Non-immune mice - 6,500 104f11 59f4 43 * Mice bearing MCA-1358 - 2,571 795 4 83*9 20 53 * Non-immune mice + 30 87f 3 57*3 34 ** Mice bearing MCA-1358 + 20 75+ 7 58 1 4 33 * 3

Puromycin (100 pg/ml) was added when the spleen cultures were initiated. "C-leucine was included in the culture media. Immunoglobulin (Ig) was obtained from the culture supernatant by affinity chromatography. For calculation of percentage CMC and percentage blocking see " Material and Methods ". The italicized means figures served as controls.

Probability that the observed differences were due to chance: * p<O.Ol; ** pc0.001.

could bind specifically to tumor cells and affect their destruction by either specifically immune or control LNC. 14C-leucine was added to cultures of spleens from mice bearing sarcomas MCA-1358 or MCA-1321 and from non-immune mice. The mouse immunoglobulin fraction of the culture supernatant was obtained by immunoadsorption as before. 5 x los 1358 or 1321 tumor cells were mixed with 1 ml aliquots of the immunoglobulin fraction of culture supernatants. After a 30 min incubation on ice, the cells were spun and washed twice with cold PBS. The amount of bound isotope was determined by liquid scintillation counting. Increased binding of immunoglobulin to the specific tumor used to immunize the spleen

cells was noted with both 1321 and 1358 tumor cells (Table IV).

Aliquots of the immunoglobulin-treated tumor cells were mixed in a 1 :lo00 ratio with LNC and seeded into the wells of a Microtest plate. LNC were obtained from mice immunized to sarcomas (1358 or 1321), and from non-immunized mice. The number of attached tumor cells was determined after a 30-hour incubation period. Both tumor cell lines were specifically sensitized by their respective immunoglobulin fraction to damage by non-immune LNC (Table IV). In addition, MCA-1358 was specifically protected from the cytotoxic effects of LNC immune to MCA-1358 (Table IV).

TABLE IV

THE EFFECT OF EXPOSING TUMOR CELLS TO IMMUNOGLOBULIN FROM SPLEEN CELL CULTURE SUPERNATANTS ON THEIR SUSCEPTIBILITY

TO DAMAGE BY IMMUNE AND NON-IMMUNE LNC IN VITRO 1

Tumor Immunoglobulin Percentage labelled Percentage cytotoxicity Percentage blocking Percentage cytotoxicity of immune of non-immune cytotoxicity LNC (ADC) of immune LNC from cultures Ig bound

of spleens from to tumor cells

1358 Non-immune mice 12 41 *** Control Control 1321 tumor-bearer 35 43 *** -4 5 1358 tumor-bearer 55 11 73 ** 24 *

1321 Non-immune mice 12 43 ** Control Control 1358 tumor-bearer 22 42 ** 2 8 1321 tumor-bearer 39 38 ** 12 80 ***

' The immunoglobulin (Ig) fraction of culture supernatant was obtained by affinity chromatography. "C-leucinc was incorporated into the newly-synthesized immunoglobulin (Ig). Five x lo6 tumor cells were mixed with 1 ml of the Ig fractions. Cell bound counts were assayed after washing to remove unbound Ig. LNC were obtained from mice immune to 1358 or 1321, or from non-immune mice. They were mixed with treated tumor cells in a 1,OOO:l ratio for the assay.

Probability that the differences between experimental and control groups were due to chance: * p<O.OS; ** p<O.Ol: *** p<O.001.

938


While this (cf Table I) and previous studies (Nelson et al., 197%) have shown that ADC and blocking activities are distinct and separable from each other, it can be concluded from the data in Table IV that both factors are synthesized by immune spleen cells in vitvo and bind to anti- immunoglobulin columns.

DISCUSSION

Previous studies have demonstrated that when cell cultures are prepared from the spleens of tumor-bearing mice, and maintained for 2 days in vitro, their supernatants can inhibit (" block ") destruction of tumor cells by lymphocytes immune to tumor-specific antigens present on the respective neoplasms (Nelson et al., 1975a). These supernatants also have an activity resembling antibody-dependent cellular cytotoxicity (ADC), inducing tumor-specific cytotoxicity by non- immune lymphocytes (Nelson, 19756). In addition supernatants of spleen cultures from tumor excised mice also induce ADC but do not have blocking activity.

The present experiments show that supernatants of spleen cultures from tumor bearing mice had both specific blocking and ADC activities when harvested at 2-3 day intervals over a 7-day period. This was the case in spite of the fact that the cultivated spleen cells were washed and subsequently recultured in fresh medium after each harvest. We also demonstrate that supernatants of spleen cells which had been cultivated in the presence of cycloheximide or puromycin lacked both the blocking and the ADC activities.

Incorporation of 14C-leucine into acid-precipitable protein in these cultures was also depressed. Since supernatants harvested 24 h after removal of cycloheximide were found to mediate both the blocking and the ADC phenomena, the ability of cycloheximide to reversibly depress protein synthesis (Ennis and Lubin, 1964) is a more likely explanation of these findings than a non-specific toxicity of the drug.

In addition, we could show that when 14C- labelled leucine was added to spleen cultures, the blocking and ADC activities of their supernatants were associated with labelled material obtained by fractionation on an anti-immunoglobulin immunoadsorbent. This material bound preferentially to tumor cells of the antigenic specificity for which the respective activity was detected.

These findings suggest that an active synthesis of factors responsible for both blocking and ADC activities, and not merely a passive release of them, had occurred, in the spleen cultures. The spleen culture system may, thus, provide a useful tool for investigating the mechanisms of this production with respect to the nature of the cells involved and their interaction, as well as the molecular nature of the factors produced.

It is interesting that neither destruction of tumor cells by sensitized allogeneic lymphocytes (Brunner et al., 1969; Henney et al., 1974), nor their destruction by lymphocytes immune to tumor-specific antigens (Hellstrom et al., 1975) is decreased in the presence of inhibitors of protein synthesis. In fact, in the latter case, the cytotoxic activity of the lymphocytes may be significantly increased (Hellstrom et al., 1975). One possible explanation suggested by the present data is that synthesis of blocking factors is depressed, allowing an increased expression of cell-mediated cytotoxicity. This hypothesis is readily testable.

We have shown previously that a cell popula- tion sensitive to treatment with anti-theta serum and complement is required in the events which lead to the appearance of blocking factors in spleen culture supernatants (Nelson et a/., 1975~). We have not been able to demonstrate that additional cells are required, since removal of macrophages and plasma cells and their precursors did not affect the apperance of blocking factors in the supernatants (Nelson et al., 1975~). This does not definitively prove that theta- carrying (T)-lymphocytes themselves produce blocking factors in vitro, however. In view of the evidence that anti bodies are required for blocking activity in this and other systems (Sjogren et al., 1971 ; Baldwin eta/. , 1973; Tamerius et al., 1975) it seems equally plausible to postulate that B-cells (not removed by the purification proce- dures employed) produce specific antibody in a T-dependent manner. Although the theta-carrying cells involved would operationally act as suppressor T-cells, their role may be different from that of suppressor T-cells in other systems (Gershon, 1974). However, the possibility that a more " conventional " suppressor T-cell is responsible for the synthesis of blocking factors must also be considered.

In addition, the role and source of tumor antigens in these cultures are unknown. It should

939

NELSON ET AL.

be noted that in the studies reported here, blocking activity was assayed by pretreatment of the target cells with the culture supernatants. This protocol favors detection of blocking factors which are either free antibodies (Robins and Baldwin, 1974) or antigen-antibody complexes (Sjogren et al., 1971; Baldwin et al., 1973; Tamerius et al., 1975) rather than those which are free antigens (Sjogren et al., 1971 ; Currie et al., 1973; Thompson et al., 1973; Baldwin et al., 1973). However, in unpublished studies, 10 out of 13 supernatants of immune spleen cell cultures were found to block cell-mediated cytotoxicity when tested at both the lymphocyte and target- cell levels (Nelson, unpublished). It has not yet been determined whether the factors which blocked when used to pretreat lymphocytes were the same as those which blocked when used to pretreat the target cells.

It is easier to interpret the data concerning the the synthesis of factors(s) mediating ADC. Plasma cells and/or their precursors are required for the appearance of ADC activity in spleen cell culture supernatants (Nelson et al., 1975~). ADC activity is found in the immunoglobulin fraction of culture supernatant of spleen cells

from tumor-bearing mice (Nelson et al., 19758) and is mediated by IgG in the sera of these mice (Pollack and Nelson, 1975). Thus the ADC activity of culture supernatant, like that of serum, reflects an antibody response to specific tumor antigens. These antibodies induce non- phagocytic (Greenberg et al., 1973) and non- adherent (Pollack et al., 1976) lymphoid cells to serve as effector cells for ADC to tumor targets.

ACKNOWLEDGEMENTS

This work was supported by grants IM-7 and 1M-43F from the American Cancsr Society, by grant CA-14697 from the National Institutes of Health, by contract NOI-CB-23887 from the National Cancer Institute and by contract NOI-CP-33372 within the Virus Cancer Program of the National Cancer Institute NIH, PHS. We thank Drs. J. Brown and J. T. Nepom for discus- sions and helpful suggestions. The technical assistance of Ms. Linda Katzenberger and Mr. M. Craig Bailey is gratefully acknowledged. We thank Ms. Darlene Cheever for assistance in preparation of the manuscript.

SYNTHGSE I N VlTRO DE FACTEURS TUMORISPECLFIQUES RESPONSABLES DU BLOCAGE ET DE LA CYTOTOXIClTE

CELLULAIRE DEPENDANT DE L'ANTICORPS (ADC)

Des cellules sple'niques de souris BALB/c porteuses de sarcomes synge'niques ou dont les sarcomes avaient P t t excise's ont PtP cultivies in vitro. Les surnageants ont t t e testts du point de vue de Ieur aptitude 1 ) supprimer (" bloquer ") les rtactions cytotoxiques spiciJ5ques mtdie'es par les cellules de ganglions lymphatiques et dirige'es contre les ne'oplasmes respectifs, et 2 ) a induire une cytotoxicitt cellulaire dtpendant de l'anticorps ( A DC) spe'cifiquement dirige'e contre les antigenes des tumeurs respectives. Ces deux activite's sptcifques ont tte' observe'es dans les surnageants de cultures de cellules sple'niques de souris canctreuses, m2me lorsqu'ils ont ttP re'coltts a plusieurs jours d'in- tervalle au cours d'une ptriode de 7 jours. Dans les surnageants de cultures provenant de souris dont les sarcomes ont ttt excises trois ou quatre semaines auparavant, on a dtcele I'ADC, mais pas l'activitt de blocage. Dans les surnageants de cultures traite'es avec des inhibiteurs de la synthkse des prottines, on n'a pratiquement pas constatt d'activitt de blocage ni d 'A DC; I'effet inhibiteur de la cycloheximide sur ces activites ainsi que sur la synthkse des proteines e'tait reversible. Les facteurs responsables du blocage et de 1'A DC presents dans les surnageants ont e'tt marquts lorsque les cultures ont e'te incube'es avec de la leucine-CI4. Le mattriel marque' a P t t retenu par des immunoadsorbants anti-immunoglobulines murines, d'ou il a pu &re tluk. De plus, il s'est lie' pre'ftrentielle- ment aux cellules tumorales contre lesquelles e'taient dirigts le blocage ou I'A DC. Ces constatations indiquent que les facteurs responsables du blocage et de I'ADC sont certainement sytithe'tists par les cellules sple'niques in vitro.

940


REFERENCES

BALDWIN, R. W., PRICE, M. R., and ROBINS, R. A., Inhibition of hepatoma-immune lymph-node cell- cytotoxicity by tumor-bearer serum, and solubilized hepatoma antigens. Int. J . Cancer, 11, 527-535 (1973).

BROWN, J. P., VAN BELLE, G., and HELLSTROM, I., Design of experiments using the microcytotoxicity

BRUNNER, K. T., MAUEL, J., CEROTTINI, J.-C., and CHAPUIS, B., Quantitative assay for the lytic action of immune lymphoid cells on 51Cr-labelled allogeneic target cells in vitro: inhibition by isoantibody and by drugs. Immunology, 14, 181-196 (1968).

CURRIE, G. A., and BASHAM, C., Serum mediated inhibition of the immunological reactions of the patient to his own tumor: a possible role for circulating antigen. Brit. J . Cancer, 26, 427-438 (1972).

DANDLIKER, W. B., and DESAUSSURE, V. A., Antibody purification at neutral PH utilizing immuno-specific adsorbents. Immunochemistry, 5 , 357-365 ( 1 968).

ENNIS, H. L., and LUBIN, M., Cycloheximide: aspects of inhibition of protein synthesis in mammalian cells. Science, 146, 1474-1476 (1964).

GERSHON, R. K., Interactions between T-cells which result in a suppression of T-cell function. In: D. H. Katz and B. Benacerraf (ed.), lmmunological tolerance: mechanisms and potential therapeutic applications, p. 41 3, Academic Press Inc., New York ( 1974).

GREENBERG, A. H., SHENN, L., and ROITT, I. M., Characterization of the antibody-dependent cytotoxic cell: a non-phagocytic monocyte? Clin. exp. Immunol., 15, 251-259 (1973).

HELLSTROM, I., and HELLSTIROM, K. E., Colony inhibition and cytotoxicity assay. In: B. R. Bloom and P. T. Glades (ed.), In vitro methods in cell- mediated immunity, pp. 409-414, Academic Press

HELLSTROM, I . , HELLSTROM, K. E., and NEPOM, J. 'T., Increased lymphocyte mediated destruction of tumor cells in microcytotoxicity assays seen after addition of inhibitors of protein synthesis. Int. J. Cancer, 16, 830-839 (1975).

HENNEY, C. S., GAFFNEY, J., and BLOOM, B. R., On the relation of soluble mediators to T-cell-mediated cytolysis. J. exp. Med., 140, 837-852 (1974).

NELSON, K., POLLACK, S. B., and HELLSTROM, K. E., Specific anti-tumor responses by cultured immune spleen cells. I. In vitro culture method and initial

assay. In preparation.

Inc., New York (1971).

characterization of factors which block immune cell-mediated cytotoxicity in vitro. Int. J . Cancer,

NELSON, K., POLLACK, S. B., and HELLSTROM, K. E., Specific anti-tumor responses by cultured immune spleen cells. 11. Culture supernatants induce specific anti-tumor cytotoxicity by non-immune lymphoid cells in vitro. Int. J . Cancer, 16, 292-300 (1975b).

NELSON, K., POLLACK, S. B., and HELLSTROM, K. E., Specific anti-tumor responses of cultured immune spleen cells. 111. Further characterization of cells which synthesize factors with blocking and antiserum dependent cellular cytotoxic (ADC) activities. Znt. J. Cancer, 16, in press (197%).

POLLACK, S. B., HEPPNER, G., BRAWN, R. J., and NELSON, K., Specific killing of tumor cells in vitro in the presence of normal lymphoid cells and sera from hosts immune to tumor antigens. Int. J. Cancer, 9, 316-323 (1972).

POLLACK, S. B., and NELSON, K., Evidence for two factors in sera of tumor-immunized mice which induce specific lymphoid cell-dependent cytotoxicity: IgGz and a rapidly appearing factor not associated with IgG or IgM. Int. J . Cancer, 16,

POLLACK, S. B., NELSON, K., and GRAUSZ, J. D., Killer cells from murine spleens: effectors of antibody-dependent cellular cytotoxicity. Trans- plant. Proc., 7,477-482 (1975).

POLLACK, S. B., NELSON, K., and GRAUSZ, J. D., Separation of effector cells mediating antibody- dependent cellular cytotoxicity (ADC) to erythko- cyte targets from those mediating ADC to tumor targets. J . Immunol., in press.

ROBINS, R. A., and BALDWIN, R. W., Tumor-specific antibody neutialization of factors in rat hepatoma- bearer serum which abrogate lymph-node cell- cytotoxicity. Inr. J . Cancer, 14, 589-597 (1974).

SJ~GREN, H. O., HELLSTROM, I., BANSAL, S. C., and HELLSTROM, K. E., Suggestive evidence that the " blocking antibodies " of tumor-bearing indi- viduals may be antigen-antibody complexes. Proc. nut. Acad. Sci. (Wash. ) , 68, 1372-1375 (1971).

TAMERIUS, J., HELLSTROM, I., and HELLSTROM, K. E., Evidence that blocking factors in the sera of multiparous mice are associated with immunoglobulins. Int. J. Cancer, 16, 456-464 (1975).

THOMPSON, D. M., STEELE, K., and ALEXANDER, P., The presence of tumor-specific membrane antigen in the serum of rats with chemically induced sarcomata. Brit. J. Cancer, 27, 27-34 (1973).

15, 806-814 (197%).

339-346 (1975).

94 1

Documents

In vitro synthesis of tumor-specific factors with blocking and antibody-dependent cellular cytotoxicity (ADC) activities