Distinct T-cell proliferative responses to 13762A rat mammary adenocarcinoma and derived clones

CELLULAR

IMMUNOLOGY

97,433-445 (1986)

Distinct T-Cell Proliferative Responses to 13762A Rat Mammary Adenocarcinoma and Derived Clones’ NEIL D. CHRISTENSEN,* JOHN W. KREIDER,*‘~-’ GERALD L. BARTLETT,*+~ AND RICK L. HORETSKY* Departments of *Pathology and tMicrobiology, College of Medicine. The Pennsylvania State University, Hershey, Pennsylvania I7033 Received July 8, 1985; accepted October 15, I985 We examined the in vitro responsesof immune lymphocytes to the tumor antigens of the syngeneic rat mammary adenocarcinoma 13762A. This tumor readily metastasizesto lymph node and lungs and is poorly immunogenic. Rats were immunixed with a highly immunogenic clone (18A) which was isolated as a spontaneous variant from the parental 13762A tumor. Clone 18A grew progressivelyin irradiated rats but regmmedcompletely in normal rats. Animals immune to 18A tumor were also immune to parental 13762A.Lymphocytesobtainedfrom the spleen and peritoneum of immune rats were tested for specific proliferation to parental 1376219tumor and clone 18A to determine whether similar cross-reactivity to these tumors occurred in vitro. We found an anatomical difference in localization of immune lymphocytes which reacted to the two tumor cell lines. Immune peritoneal exudate cells (PEC) responded strongly to clone 18A but poorly to 13762A,while immune spleencells from the sameanimals respondedpredominantly

to 13762Atumor. After 7 daysculture, PEC proliferatingin responseto clone 18A contained 84-95% W3/25+ T-helper cells, and only 5-8% 0X8+ cytotoxic/suppressor cells, while analogous cultures of spleen cells responding to parental 13762A tumor consisted of 60-80% W3/25+ cells and 20-23% 0X8+ cells. Immune spleen cell cultures stimulated with 13762A tumor generated cytotoxic lymphocytes which specifically lysed both parental 13762A and clone 18A cells. We conclude that despite cross-reactivity in vivo and in vitro, antigens present on 13762A and 18A tumor cells stimulated di$eret subsetsof immune T cells. o 1986 ACXICUCC R~JS IOC.

INTRODUCTION The rat mammary adenocarcinoma 13762A grows progressively in syngeneic F344 rats, and metastasizeswith high frequency to the lungs and regional lymph nodes. Our laboratory has been studying the in vivo interactions between the immune system and this tumor with emphasis on the induction and transfer of immunity (l-4). A strategy which we had adopted in earlier studies was to induce immunity to 13762A by primary tumor excision followed by cyclophosphamide treatment to eliminate metastases(1, 2). This protocol cured 50% of rats which were then strongly resistant to a lethal challenge of 13762A. However, the immunity was unable to produce any effect on an established tumor following systemic adoptive transfer of T cells from such rats (3). More recently, we have exploited the heterogeneity that exists in tumor ’ This work was supported by USPHS CA 29006 and the Jake Gittlen Memorial Golf Tournament. * To whom reprint requests should be addressed. 433 0008-8749186$3.00 copyright 0 1986 by Academic Pres, Inc. All rights of reproduction in any form merved.

434

CHRISTENSEN ET AL.

cell populations (5,6), and have isolated a spontaneouslyoccurring clone with much greater immunogenicity than the parental tumor. Tumor cell clones with enhanced immunogenicity may occur spontaneously (4, 7, 8), or as a result of treatment with lectins (9), mutagens (10-l 2), or 5-azacytidine (13, 14). We have analyzed one highly immunogenic clone (18A) and the details of its behavior in vivohave been reported (4). Clone 18A regressedspontaneously in normal rats, but grew progressively in irradiated animals. A routine immunization protocol was developed that provided an effective population of PEC3for adoptive transfer experiments (4). This protocol consisted of a challenge dose of 13762A ascites 40 days after 18A immunization, followed by reimmunization with 13762A ascites and mineral oil 5 days prior to harvesting PEC. PEC from rats immunized by this standard protocol transferred tumor protection to naive recipients, and were effective, together with lowdose irradiation, in curing large 13762A tumors (3). The PEC effecters involved in the transfer of tumor protection included both W3/25+ and 0X8+ T cells (15). The purpose of this investigation was to test the hypothesis that there were quantitative and qualitative differences in the in vitro host responsesto parental 13762A rat mammary adenocarcinoma and tumor cell clone 18A. Immune lymphocytes were obtained from rats immunized by the identical procedure to that developed for in vivoadoptive transfer experiments. Tumor-stimulated immune lymphocytes were examined for functional specific anti-tumor activity. We evaluated T-cell blastogenesis, subsetspecific T-cell differentiation antigens, interleukin 2 (IL-2) production, and cellmediated cytotoxicity. The results demonstrated that there were distinct differences in the anatomical localization of T cells which responded to the two tumor cell populations. Immune PEC proliferated strongly in response to clone 18A, but poorly to parental 13762A. In contrast, immune spleen cells generated strong responsesto parental 13762A, weaker responsesto 18A, and showedequal cytotoxicity to both 13762A and 18A. MATERIALS AND METHODS Animals. F344 female rats were purchased from the Frederick Cancer Research Center, Frederick, Maryland and used when > 120 g body weight (16). Tumor cell lines. The 1376214tumor was obtained from Dr. A. Bogden, Mason ResearchInstitute (Worcester, Mass.), and propagated both in ascitesform and in cell culture. Ascites tumors were maintained by weekly intraperitoneal (ip) passageof 1 X lo7 cells. Cell cultures of 13762A tumor were established from ascites tumor and grown in monolayer cultures in Eaglesminimum essential medium (MEM) containing 5% heat-inactivated fetal calf serum (FCS), 2 mM glutamine, 100 units of penicillin, 100 &ml streptomycin, and 10 mM Hepes. The R3230AC rat mammary tumor was obtained from Dr. P. Fritz, Department of Pharmacology, at this institution. Clone 18A was isolated from cell cultures of 13762A tumor cells by limiting dilution as previously described (4). This tumor spontaneously regressedin normal rats when 1 X lo6 cells were injected intradermally (id). All cultures were checked at monthly intervals for the presence of mycoplasma infection using the Hoechst stain kit (Flow Laboratories, Rockville, Md.). 3Abbreviations: Con A, concanavalin A; cpm, counts per minute; CTL, cytotoxic T lymphocyte; FCS, fetal calf serum; [3H]TdR, tritiated thymidine; id, intraded, IL2, interleukin 2; ip, intrapzitoneal; MTLC, mixed tumor-lymphocyte culture; OD, optical density; PEC, peritoneal exudate cells; SEM, standard error of the mean.

DISTINCT LYMPHOCYTE

ANTI-TUMOR

RESPONSES

435

Tumor-immune rats. Rats immune to a lethal inoculum of 13762A were generated as previously described (4). Briefly, 1 X lo6 immunogenic clone 18A tumor cells were injected (id) on the right dorsolateral thorax. Tumors regressedcompletely after 40 days and rats then were challenged with a lethal dose of 1 X lo6 13762A ascitescells. Immune rats were reimmunized and a sterile peritoneal exudate was induced 2-6 months after challenge by injecting ip 2 X 1O5fresh ascites 13762A tumor cells together with 5 ml sterile mineral oil. Mixed tumor-lymphocyte culture. Mixed tumor-lymphocyte cultures (MTLC) were generated in the following manner. PEC from reimmunized rats were harvested by injecting 40 ml sterile MEM into the peritoneal cavity, and withdrawing the cell suspension. Single cell suspensions of spleen cells from reimmunized and normal rats were prepared by pushing spleens through a stainless-steel mesh sieve. Responding PEC or spleen cells were placed into 96-well round-bottomed microplates (Corning No. 25850, Corning, N.Y.) with mitomycin C-treated (25 &ml at 37°C for 30 min) tumor cells and mitomycin C-treated normal spleen cells as fillers. Red blood cells were removed from the filler cell population prior to mitomycin C treatment by incubation in low ionic strength buffer solution for 10 min at 4’C. Before culture, PEC were depleted of adherent cells by incubation in RPM1 1640 culture medium in glass bottles for 3 hr at 37°C. Non-adherent cells were retrieved by gentle pipetting. Microcultures contained 2 X lo5 filler cells, 1 X lo4 tumor cells and between l-2 X lo5 responding spleen or PEC in 200 ~1RPM1 1640 containing 10%heat-inactivated FCS, 2 mM glutamine, 5 X 10e5A4 2-mercaptoethanol, 100 units of penicillin, 100 pg/ml streptomycin, and 10 mM Hepes. The microplates were incubated at 37°C in a humidified atmosphere of 5% CO* in air. For proliferation assays, 1 &i [3H]TdR (sp act 25 Ci/mmol, Amersham, Arlington Heights, Ill.) in 50 ~1 of RPM1 1640 culture medium was added to each microculture. After 6 hr incubation at 37”C, microcultures were harvested onto glassfiber filter paper using a Titertek Cell Harvester (Flow Laboratories), and radioactivity was measuredin a liquid scintillation counter. Fresh ascites 13762A tumor was used for all in vivo challenge and reimmunization procedures, while cultured 13762A tumor was used in MTLC. ZdentiJication of T-cell subsets. The surface phenotype of cells from MTLC was determined by a double-antibody labeling technique (17). Cells were incubated in a 50-fold dilution of ascitesfluid containing the monoclonal antibodies W3/25 (T-helper cell subset)or OX8 (cytotoxic/suppressor T-cell subset)(Accurate Chemical & Scientific Corporation, Hicksville, N.Y.). After 1 hr incubation at 4°C cells were washed and incubated with a fluoresceinated F(ab’)* fragment rabbit anti-mouse IgG (Cappel, Scientific Division, Cooper Biomedical, Inc., Malvern, Pa.) for 30 min at 4°C. Crossreactive anti-rat IgG antibodies were absorbed on an Ultrogel AcA34 (LKB Instruments, Int., Gaithersburg, Md.) column to which rat immunoglobulin was attached by gluteraldehyde fixation (17). Cells were washed, fixed in 1%paraformaldehyde for 15 min at room temperature, washedagain, and viewed with a fluorescencemicroscope. The proportions of cells positive for either W3/25 or OX8 antigens in cultures of immune PEC and 18A, and cultures of immune spleen cells and 13762A were compared using Fisher’s exact probability test. The calculated P values represent the probability that the proportions of monoclonal antibody positive cells in the specified cultures are the same. IL-2 assay. Assays used the IG2dependent murine T-cell line CTLL (obtained from Dr. K. A. Smith, Department of Medicine, Dartmouth Medical School, Hanover,

436

CHRISTENSEN

ET AL.

N.H.). CTLL (5 X 16) were plated into 96-well flat-bottomed plateswith serial dilutions of supematants to be tested for IL2 content. After overnight incubation, microcultures were pulsed with 1 &i [3H]TdR for 6 hr. One unit of IL2 activity was calculated as the volume of supematant required to give 50% of maximal proliferation of CTLL cells. Cytotoxicity assays. Cell-mediated cytotoxicity was detected with the radioactive label releaseassay( 18) and the neutral red dye assay( 19). For the dye assay,monolayers of tumor targets cells were prepared by plating 4 X lo4 tumor cells/well into 96-well round-bottomed microplates. After overnight incubation, attached cells were labeled with neutral red dye (0.04% w/v01 in MEM without serum) for 20 min at 37°C. The microplates were washed three times to remove excessdye, and serial twofold dilutions of effector cells in RPM1 1640 culture medium were plated onto target cells. After overnight incubation, plates were washed three times in warm MEM to remove unattached lymphocytes and released dye. Remaining target cells were then lysed by adding 100 ~1 of a solution of 0.05 M acetic acid in 50% ethanol and optical density (OD) read in a microplate reader (Dynatech MR 600, Dynatech Labs., Alexandria, Va.) at a test wavelength of 570 nm and a reference wavelength of 630nm. Percentage specific lysis was determined by the formula: % specific lysis =

OD targets alone - OD test well x loo OD targets alone

In the radioactive label release assay, monolayer cultures of tumor target cells in log phase growth were incubated in a solution of 10 &i/ml [3H]TdR in MEM for 2 hr at 37°C. Following incubation, target cells were detachedwith 0.25%trypsin solution, washed three times in RPM1 1640 culture medium and 1 X IO4 cells plated into microcultures containing serial twofold dilutions of effector cells. After overnight incubation at 37’C, 100 ~1 of supematant was removed and radioactivity determined in a scintillation counter. Percentage specific lysis was determined by the formula: (cpm sample) - (cpm spontaneous) x loo. % specific lysis = (cpm max) - (cpm spontaneous) Spontaneous release was determined from radioactivity in supematants of target cells incubated alone, and maximum release determined from radioactivity in 50 ~1 of water-lysed target cell suspension. RESULTS Tumor-Specific Proliferation Induced by 13 762A We tested lymphocytes from tumor-immune rats for specific responsesto 13762A tumor antigens and established the optimal responder to stimulator cell ratio (Fig. 1). Various numbers of tumor stimulator cells were cultured with 1 X lo5 immune or normal spleen cells, and proliferation was assayed on Day 5. Immune spleen cells responded strongly to parental 13762A, weakly to clone 18A, and did not respond to the antigenically unrelated rat mammary tumor R3230AC. Maximum proliferation occurred in cultures containing responder to stimulator ratios between 10:1 and 20: 1. No significant t3H]TdR uptake occurred in MTLC containing non-immune rat spleen cells (data not shown).

DISTINCT LYMPHOCYTE

ANTI-TUMOR

Tumor

RESPONSES

431

cell two.

FIG. I. Effect of tumor stimulator cell number and specificity of the tumor-specihc proliferation of immune spleen cells. Spleen cells (1 X 105)from rats reimmunixd 6 months after challenge were cultured together with the indicated dosesof mitomycin C-treated stimulator tumor cells and mitomycin C-treated normal spleen cells as fillers. Proliferation was assessedon Day 5 of culture, following a 6-hr pulse of [‘H]TdR, for cultures containing 13762A (0) 18A (Cl),R3230AC (A), and no tumor stimulator cells (0). No proliferation occurred in cultures containing stimulator tumor and filler cells alone (data not shown).

Kinetics of Tumor-SpecificResponses We determined the kinetics of tumor antigen-induced proliferation of immune lymphocytes from different organs (Fig. 2). MTLC containing spleen or PEC of reimmunized rats were assayedfor proliferation at various time points. Immune spleen cells (Fig. 2A) proliferated strongly to both 18A and 13762A with maxima on Days 5 and 6, respectively. In contrast, immune PEC (Fig. 2B) responded strongly to clone 1SA, reaching a maximum on Day 6, but weakly to parental 13762A, with a maximum response early on Day 4. No significant anti-tumor activity was detected in MTLC containing non-immune spleen or PEC (data not shown).

Efect of Reimmunization on SpleenCell and PEC Anti-tumor Responses The effect of ip reimmunization with 13762A tumor on the tumor-specific responses of immune spleen and PEC was next determined. Reimmunization was included in 150,

i?! 32 a “5 ‘0 x

1

A. spleen

1 0. PEC

120.

90.

Days

Days

FIG. 2. Timecourse of tumor-stimulated proliferative responsesof lymphocytes from the spleen (A), and peritoneum (B) of immune rats. Tumor-immune rats were reimmunized 2 months after challenge and given mineral oil on Day -5. On Day 0, spleen cells (1.5 X 105),and PEC (2 X 105)were cultured with 1 X IO’ mitomycin C-treated tumor cells and normal spleen filler cells. Proliferation induced by tumor stimulator cells 13762A (0), 18A (O), R3230AC (A), and no tumor stimulator cells (0) was plotted as the mean cpm k SEM of triplicate cultures. No proliferation occurred in cultures which contained stimulator and filler cells alone (data not plotted).

CHRISTENSEN ET AL.

438

the protocol for establishing an effectivepopulation of lymphocytes for adoptive transfer experiments (l-3). Two groups of immune rats received either mineral oil alone or mineral oil plus 2 X IO5 13762A ip, 3 months after challenge and rejection of 13762A. Spleen cells from immune rats receiving mineral oil alone responded only against parental 13762A tumor (Fig. 3A). No reactivity to clone 18A or R3230AC was found. However, spleencells from reimmunized rats showed an increasedreactivity to 13762A, as well as a response to clone 18A (Fig. 3B). PEC from immune animals receiving mineral oil alone responded weakly to 18A tumor cells only (Fig. 3C). However, the PEC from reimmunized rats (Fig. 3D) responded very strongly to clone 18A, but again, were not stimulated by 13762A or R3230AC. Reimmunization on Day -5 substantially increased the tumor-specific responsesof immune rat spleen and PEC to 13762A and 18A tumor. Spleen cells showed increased reactivity to 13762A while lymphocytes reactive to 18A tumor appeared in the peritoneal cavity in response to reimmunization.

A. No

Reimmunization

0.

Reimmunization T

Days

Days

Days

Days

FIG. 3. The effect of reimmunization (Day -5) on the tumor-specific proliferative responsesof tumorimmune rat lymphacyte~. Tumor-immune rats (3 months after challenge date) were separated into two groups; one receiving reimmunization with 2 X IO5 13762A ip and mineral oil, and the other, mineral oil alone. On Day 0, PEC (1 X 1Or)and spleen cells (1.5 X 105) were cultured together with 1 X 10’ stimulator tumor cells and 2 X lo5 normal spleen filler cells. Proliferation was asses4 at various time points in cultures containing 13762A (0), 18A (O), R3230AC (A), and no tumor stimulator cells (0). Results expressed as mean cpm -c SEM for mineral oil only treated spleen (A) and PEC (C), and reimmunized spleen (B) and PEC (D).

DISTINCT LYMPHOCYTE

ANTI-TUMOR

439

RESPONSES

Surface Phenotype of Proliferating Lymphocytes Tumor-stimulated immune spleen and PEC were examined for differentiation antigens defining helper T cells and cytotoxic/suppressor T cells. Lymphocytes were retrieved from 7 day MTLC, reacted with monoclonal antibodies and fluoresceinated anti-mouse IgG as described under Materials and Methods, and the percentage of labeled blast cells determined for eachcell-surfacemarker for three separateexperiments (Table 1). The lymphocytes from cultures of PEC stimulated by clone 18A were predominantly T cells of the helper phenotype (W3/25+), while spleen cells responding to parental 13762A consisted mostly of helper cells but with a significantly increased number of cytotoxic/suppressor cells (0X8+). Phenotypic characterization of blast cells in cultures of PEC stimulated with 13762A was not possible due to lack of cells after 7 days culture. IL-2 Production We next examined the function of the immune T lymphocytes (Table 1) that responded to 18A and 13762A antigens. In the first seriesof experiments, IL2 secretion from MTLC was assessedas a functional test for the detection of tumor-specific T helper cells. IL-2 production was also compared with proliferation to test the hypothesis that lymphocyte proliferation correlated with IL-2 production. Replicate MTLC, containing 3 X lo5 immune spleen or PEC and 1 X lo4 mitomycin C-treated tumor cells, but no added filler cells, were prepared in 96-well round-bottomed plates. Filler cells were not included so that production and/or utilization of IL2 by these cells was not TABLE 1 Surface Phenotype of Tumor-Stimulated Immune Lymphocytes’ % Marker-positive blasts Expt

MTLC

W3/25

OX8

A

1 PEC+18A 2 Spleen + 18A 3 Spleen + 13762A

95.0b 72.0 78.8

5.1’ 11.1 20.3

B

1 PEC + 18A 2 Spleen + 18A

87.3 72.2

6.7 15.4

C

1 PEC+ 18A 2 Spleen + 18A 3 Spleen + 13762A

83.7 76.4 61.1

7.7 5.2 22.6

’ MTLC containing immune spleen or PEC were cultured for 7 days then incubated with the indicated mouse monoclonal antibodies and fluoresceinated anti-mouse antibody. Labeled cells were viewed with a fluorescencemicroscope. ’ Probability that proportions of W3/25+ cells are not significantly different (Fisher’s Exact Probability Test). PEC + 18A vs spleen + 13762A, experiment (A) P = 0.11; experiment (C) P = 0.001. Controls PEC + 18A (A) vs PEC + 18A (C) P = 0.15; spleen + 13762A (A) vs spleen + 13762A (C) P = 0.03. c Probability that proportions of 0X8+ cells are not significantly different. PEC + 18A vs spleen + 13762A, experiment (A) P = 0.004; experiment (C) P = 0.002. Controls PEC + 18A (A) vs PEC + 18A (C) P = 0.22; spleen + 13762A (A) vs spleen + 13762A (C) P = 0.13.

440

CHRISTENSEN ET AL.

a contributing factor. Supernatants from the MTLC were harvested at various time intervals and stored at -20°C for severaldays before assayof IL-2 content. Considerable quantities of IL-2 were detected in cultures of immune PEC stimulated with both 18A and 13762A tumor cells (Fig. 4). In contrast, no detectable IL2 production occurred in MTLC containing immune PEC and R3230AC, or in immune spleen cell cultures containing 18A, 13762A, or R3230AC. The time course analysis demonstrated that maximum IL-2 production by PEC stimulated by 18A occurred in the first 24 hr (Fig. 4). Surprisingly, the production of IL-2 from these cultures did not correlate with proliferation (Fig. 5). No IL2 wasdetectedin cultures of immyne spleencells stimulated with 13762A tumor even though measurable proliferation on Day 6 was found. In contrast, immune PEC stimulated by 13762A tumor produced large quantities of IL-2 but there was no proliferation on Day 6. Cytotoxicity Assays In the second series of experiments, the T cells proliferating in response to tumor antigens (Table 1) were tested for specific anti-tumor cytotoxicity against 13762A, 18A, and R3230AC tumor cells. Cytotoxicity in cultures of immune spleen cells stimulated with 13762A and immune PEC stimulated with 18A was measured on Day 6 by the neutral red dye assay. Strong specific cytotoxicity against parental 13762A tumor and clone 18A was demonstrated in spleen MTLC, but lysis of R3230AC tumor was minimal (Fig. 6A). In contrast, there was We specific lysis of the three targets by cells from MTLC containing PEC and 18A (Fig. 7A). Total cytotoxicity generated in each of these cultures was tested by incubating effector and target cells in the presence of concanavalin A (Con A) (20). Increased killing of R3230AC but not of 13762A or clone 18A occurred in spleen MTLC containing Con A which further demonstrated that the cytotoxicity was specific for 13762A and 18A (Fig. 6B). When cells from

FIG. 4. Time course of IG2 production from immune lymphocytes. Replicate micro-MTLC contained 3 X lo5 reimmunized spleen or PEC with 1 X 10’ tumor stimulator cells, without lillers. Supematants were harvested at various time intervals and tested for IL-2 content. The concentrations of IG2 were plotted against time for immune PEC stimulated by I8A (Cl), 13762A (0), R3230AC (A), and for immune spleen cells stimulated by 13762A (0). No detectable IL2 production occumd from immune spleencells stimulated by 18A or R3230AC (data not plotted).

DISTINCT LYMPHOCYTE

ANTI-TUMOR a.

PEG

Spleen

I

RESPONSES

441

poo

PEC

Spleen t

8o E >

I

L ;= i& :o FIG 5. Comparison between proliferation (A) and IL-2 secretion (B) induced in cultures of immune lymphocytes stimulated with tumor antigens. Duplicate MTLC were set up and assayed for either IL2 content on Day 1, or proliferation on Day 6. Proliferation of tumor-stimulated lymphocytes was assayed by [‘H]TdR uptake and plotted as mean cpm + SEM of triplicate microcultures (A). The concentration of IL-2 was plotted for immune PEC and spleen cells stimulated by tumor antigens or 5 &ml Con A (B). No proliferation of CTLL occurred in medium containing 5 pg/ml Con A without IL-2.

MTLC containing PEC and 18A were tested for killing in the presenceof Con A, very strong cytotoxicity against both 18A and 13762A tumor but not against R3230AC tumor was observed (Fig. 7B). This result was not expected since previous data demonstrated that cells from these cultures were predominantly W3/2S with considerable IL2 secreting potential. Cytotoxicity in MTLC containing immune spleen cells and 18A was poor (data not shown) and was not measured in MTLC containing PEC and 13762A. 100

1 A.-con A

60 .B P g .” $

60

40

Effector: Target Ratio FIG. 6. Cytotoxicity of immune spleen cells stimulated in vim with 13762A tumor cells. Reimmunized tumor-immune spleen cells (2.5 X IO’) were incubated with 2.5 X IO’ 13762A stimulator tumor cells in 25 cm* flasks. After 6 days culture, specific and total cytotoxicity was measured by the neutral red dye assay (seeMaterials and Methods). The percentage specific lysis (mean 2 SEM of triplicate cultures) was plotted against the effector to target cell ratio for the targets 13762A (0), 18A (O), and R3230AC (A), for specific cytotoxicity (A). Total cytotoxicity generated in cultures was determined separately by incubation of effector and target cells together with 5 rcg/ml Con A (B).

442

CHRISTENSEN ET AL. 100 60

.iJ 2 60 g j

40

E 20

0 1:1

21

41

1:1

8:l

23

4:l

8:l

Effector : Target Ratio

RG. 7. Cytotoxicity of immune PEC stimulated by 18A tumor cells. PEC from reimmunized rats (1.0 X 10’) were incubated with 5 X 10’ 18A stimulator tumor cells in 25-cmz flasks. After 6 days culture specific and total cytotoxicity was measured by the neutral red dye assay. The percentage specific lysis (mean + SEM of triplicate cultures) was plotted against the effector to target cell ratio for the targets 13762A (0), 18A (O), and R323OAC (A), for specibc cytotoxicity (A), and for cytotoxicity induced in the presence of 5 g/ml Con A (B).

The neutral red dye technique potentially measures both lysis and detachment of target monolayers since excessdye and unattached cells are washed away prior to assessingpercentage kill. We conducted a comparison between the neutral red dye assay and the release of [3H]TdR for the same effector cell population. MTLC containing immune spleen cells and 13762A were generated as previously described, and on Day 6, cultures were divided into two aliquots and tested for cytotoxicity against tumor targets in the neutral red dye assay and the [3H]TdR release assay (Figs. 8A and B, respectively). As previously found, cytotoxicity was generated against 13762A, but not against R3230AC. Thus both assaysystemswere in agreement. DISCUSSION We have developed the strategy of cloning spontaneously occurring variants of the syngeneic rat mammary adenocarcinoma 13762A to analyse the immune response

Effecta

to Cd

Ratio

FIG. 8. Cell-mediated cytotoxicity: comparison between neutral red dye assay (A) and [3H]TdR-release assay (B). Reimmunized immune spleen cells (2.5 X 10’) were cultured in 25-cmz flasks together with 2.5 X lo5 mitomycin C-treated 13762A tumor stimulator cells. After 6 days culture, cells were harvested and replicate aliquots were testedfor cytotoxicity in the neutral red dye assayand [3H]TdR-releaseassay.Percentage cytotoxicity against the targets 13762A (0) and R3230AC (A) were plotted as the mean + SEM of triplicate cultures for cytotoxicity measured by the neutral red dye assay(A) and the [3H]TdR-release assay(B).

DISTINCT LYMPHOCYTE

ANTI-TUMOR

RESPONSES

443

to this tumor both in vivo and in vitro. Enhanced immunity to parental 13762A tumor was achieved in vivo by immunization with a regressorclone 18A (4). Systemicadoptive transfer of PEC from immune animals obtained by this procedure, in combination with low-dose irradiation, were effective in eliminating large established tumors (3). We describe here a preliminary analysis of the in vitro responsesof immune lymphocytes to the tumor antigens of parental 13762A and clone 18A. Rats made immune to 13762A by the procedure described above showed striking anatomical differences in lymphocyte reactivity to these tumor antigens. Spleen cells proliferated strongly to 13762A tumor, but less to clone 18A. In the absenceof reimmunization on Day -5, spleen cells reacted only to 13762A tumor. Immune PEC, however, generated a powerful response against clone 18A and the proliferating cells were predominantly W3/25+. PEC stimulated by parental 13762A produced a small but significant early responsepeaking on Day 4. No responseswere detectedfrom MTLC containing normal rat lymphocytes. The T cells proliferating in responseto 18A and 13762A tumor antigens were tested functionally for IL-2 production and cell-mediated cytotoxicity. When IL2 production in MTLC was assessed,considerable quantities were induced from immune PEC stimulated with either 13762A or 18A, confirming that the immune PEC were functionally T helper cells. Interestingly, IL-2 secretion induced from tumor-stimulated immune lymphocytes did not completely correlate with proliferation. Immune spleen cells cultured with 13762A tumor did not secrete detectable amounts of IL-2, although strong tumor-specific proliferation occurred. In contrast, immune PEC stimulated with 13762A tumor induced minimal proliferation but considerable IL-2 secretion. The absolute requirement for IL-2 for T-cell proliferation has been generally assumed for most antigen-specific responses(2 I), particularly with the advent of IG2dependent T-cell lines (22, 23). Recent data, however, has demonstrated that some T-cell lines can be induced to proliferate in the absence of added IL-2 (24). In other studies, T-cells have responded to extremely low concentrations of IL-2 (25), and it is possible that IL2 produced in cultures of 13762A stimulated immune spleencells was sufficient to stimulate proliferation but was insufficient to be detected by the IL-Zdependent line CTLL. Lack of cell proliferation in cultures of immune PEC containing 13762A tumor, despite high IL-2 production may indicate that a suppressor mechanism was triggered by 13762A tumor antigens which prevented significant cell division. Cytotoxicity in MTLC containing immune spleen cells was generated against both 18A and 13762A but not against R3230AC. When total cytotoxicity in these cultures was determined by adding Con A to the assay, only an increase in the killing of R3230AC was found. These two observations confirmed that tumor-specific cell-mediated cytotoxicity against 18A and 13762A was induced from immune spleen cells. No specific cytotoxicity in MTLC containing PEC and 18A was found as was expected from data on the phenotypic analysis of these cells (84-95% W3/25+ T-helper cells). However, when Con A was added to the assay, substantial lysis of both 18A and 13762A but not R3230AC tumor cells was observed. One possible explanation for this result is that lectin stimulation may have induced the secretion of a lymphotoxin from the predominantly T-helper-cell population. It would be necessaryto propose also that R3230AC is relatively resistant to lymphotoxin-mediated lysis. Lysis of target cells by mouse lyt I+ T-helper-cell clones via a soluble mediator has been described (26).

444

CHRISTENSEN ET AL.

Analysis of lymphocyte responsesto tumor cell populations has revealed heterogeneity in the number and types of antigenic determinants expressedby tumor cell lines (4-7,12,27). When tumor antigens were analyzed at the clonal level, the antigens detected included clone-specific tumor antigens as well as cross-reacting antigens (28). The immune T-cell responseto tumor cell populations also represents a T-cell clonal responseto different tumor antigens, and this has been documented in particular for CTL clones reactive against tumor lines (7, 28). These different tumor antigens may also stimulate distinct clones of T-helper- and T-suppressor-cell subsets.Therefore it is possible to describe immune responsesto tumor populations in vivu as the net result of a series of interactions between different tumor antigens and discrete subsets of T-helper, T-suppressor, and CTL. Progression or regression would be determined by the strength or frequency of T-helper- and CTL-inducing as opposed to T-suppressorinducing tumor antigen(s). Thus increased immunogenicity of tumor cell clones may involve the addition or enrichment of “helper” determinants similar to that described in the CTL response against the Qa-lb alloantigen (29), or alternatively, the loss of “suppressor-inducing” determinants as described in the T-cell response against henegg lysozyme (30, 3 1). The results presented in this paper also demonstrate that different antigens on 13762A induced both anatomical and subset-specific differences in T-cell reactivity. Antigens present on 18A stimulated a strong proliferative T-helper-cell response and IG2 production from immune PEC, but predominantly poor proliferative and cytotoxic responsesfrom immune spleen cells. In contrast, antigens on 13762A stimulated IL-2 production but no proliferation of immune PEC, and strong proliferation with good cytotoxicity from immune spleen cells. Anatomical segregation of these immune T-cell-subsets occurred in vim, in response to reimmunization with 13762A ascites. Lymphocytes proliferating only in responseto clone 18A appearedin the peritoneum, while reactivity to 13762A remained in the spleen. These findings raise another important aspect of the function of immune effecters in vim, viz., homing capability. Active, tumor-destroying immune cells may be ineffective at preventing tumor growth because of their inability to home to certain organ sites. This is evident in results describing the ability of some CTL populations to prevent tumor growth at the site of the tumor inoculum, but which have no effect in systemic adoptive transfer experiments (32). Surfacemolecules involved in the control of homing of certain lymphocyte populations have been described (33,34). Immune PEC responding to clone 18A may represent a subset of lymphocytes with different homing properties than the immune spleen cells responding to parental 13762A antigens. It is interesting to note that it is the parental 13762A ascites tumor that promotes the early appearance of 18A responsive lymphocytes into the peritoneum. These findings indicate that l8A tumor antigens are present within the parental tumor population and are not the result of antigen mutations induced by culturing 18A or due to FCS antigens. Repeatedcloning of 13762A therefore should yield further “18A-like” tumor clones. In conclusion, we examined rat immune T-cell responsesto the poorly immunogenic rat mammary adenocarcinoma 13762A and a highly immunogenic clone, 18A. Subsets of functionally different immune T cells responded to the different tumor antigens of 13762A. This was demonstrated by cloning 13762A tumor populations. 18A tumor antigens clearly stimulated predominantly T-helper cell responseswhile antigens on the parental 13762A tumor stimulated T-helper, cytotoxic, and most probably

DISTINCT LYMPHOCYTE

ANTI-TUMOR

RESPONSES

445

T-suppressor cell responses.These experiments indicate that a dynamic interaction between these subsetsof T cells and tumor antigens occurs in viva, with progression or regressiondetermined by the strength or frequency of the effector-stimulating tumor antigens which stimulate T cells with tumor destructive capabilities. The strong T-helper cell responseto 18A tumor antigens provides a likely mechanism for regression whereas the possible induction of T-suppressor cells by 13762A antigens may explain tumor progression. ACKNOWLEDGMENTS We thank Pat Welsh and Dean Stoesz for expert technical assistance.

REFERENCES 1. 2. 3. 4. 5. 6. 7.

Boyer, C. M., Kreider, J. W., and Bartlett, G. L., Cancer Rex 41,2394, 1981. Boyer, C. M., Kreider, J. W., and Bartlett, G. L., Cancer Res. 42,2211, 1982. Kreider, J. W., Bartlett, G. L., Christensen, N. D., and Welsh, P., ht. J. Cancer 36, 217, 1985. Kreider, J. W., and Bartlett, G. L., J. N&l. Cancer Inst. 75, 141, 1985. Fidler, I. J., and Kripke, M. L., Science (Washington. D-C.) 197, 893, 1977. Heppner, G. H., Cancer Res. 44,2259, 1984. Urban, J. L., Burton, R. C., Holland, J. M., Kripke, M. L., and Schreiber, H., J. Exp. Med. 155, 557, 1982. 8. Binz, H., Fenner, M., Frei, D., and Wigzell, H., J. Exp. Med. 157, 1252, 1983. 9. Dennis, J. W., Donaghue, T., Florian, M., and Kerbel, R. S., Nature (London) 292,242, 1981. 10. Boon, T., and Kellermann, O., Proc. Natl. Acad. Sci. USA 74,272, 1977. 11. Uyttenhove, C., Van Snick, J., and Boon, T., J. Exp. Med. 152, 1175, 1980. 12. Dennis, J. W., Laferte, S., Man, M. S., Elliott, B. E., and Kerbel, R. S., Int. J. Cancer 34, 709, 1984. 13. Kerbel, R. S., Frost, P., Liteplo, R., Carlow, D. A., and Elliott, B. E., .I Cell Physiol. Suppl. 3, 87, 1984. 14. Olsson, L., and Forchhammer, J., Proc. Nutl. Acad. Sci. USA 81, 3389, 1984. 15. Kreider, J. W., Howell, L. E., and Bartlett, G. L., Cell. Immunol. 93,26, 1985. 16. Kreider, J. W., Bartlett, G. L., and Pumell, D. M., J. Natl. Cancer Inst. 56, 797, 1976. 17. Hickey, W. F., Gonatas, N. K., Kimura, H., and Wilson, D. B., J. Immunol. 131,2805, 1983. 18. Brunner, K. T., Mauel, J., Cerottini, J.-C., and Chapuis, B., Immunology 14, 181, 1968. 19. Mullbacher, A., Parish, C. R., and Mundy, J. P., J. Immunol. Methods 68, 205, 1984. 20. Bevan, M. J., and Cohn, M., J. Immunol. 114, 559, 1975. 21. Wagner, H., Hardt, C., Heeg, K., Phzenmaier, K., Solbach, W., Bartlett, R., Stockinger, H., and Rollinghoff, M., Immunol. Rev. 51,215, 1980. 22. Gillis, S., and Smith, K. A., Nature (London) 268, 154, 1977. 23. Haas, W., Mathur-Rochat, J., Pohlit, H., Nabholz, M., and von Boehmer, H., Eur. J. Immunol. 10, 828, 1980. 24. Von Boehmer, H., and Turton, K., Eur. J. Immunol. 13, 176, 1983.

25. Gillis, S., Mochizuki, D. Y., Conlon, P. J., Hefeneider, S. H., Ramthun, C. A., Gillis, A. E., Frank, M. B., Henney, C. S., and Watson, J. D., Immunol. Rev. 63, 167, 1982. 26. Tite, J. P., and Janeway, C. A., Eur. J. Immunol. 14, 878, 1984. 27. Ito, M., Suzuki, E., Naiki, M., Sendo, F., and Arai, S., Int. J. Cancer 34,689, 1984. 28. Wortzel, R. D., Urban, J. L., Philipps, C., Fitch, F. W., and Schreiber, H., J. Immunol. 130,2461, 1983. 29. Keene, J., and Forman, J., .I. Exp. Med. 155, 768, 1982. 30. Sercarz, E. E., Yowell, R. L., Turkin, D., Miller, A., Araneo, B. A., and Adorini, L., Zmmunol. Rev. 39, 108, 1978. 31. Adorini, L., Harvey, M. A., Miller, A., and Sercarz, E. E., .I. Exp. Med. 150,293, 1979.

32. Engers, H. D., Lahaye, T., Sorenson, G. D., Glasebrook, A. L., Horvath, C., and Brunner, K. T., J. Immunol. 133, 1664, 1984. 33. Butcher, E. C., Scotlay, R. G., and Weissman, I., Eur. J. Immunol. 10, 556, 1980. 34. Gallatin, W. M., Weissman, I. L., and Butcher, E. C., Nature (London) 304, 30, 1983.