Infiltration of Tumors by Systemically Transferred Tumor-Reactive T Lymphocytes Is Required for Antitumor Efficacy

INTRODUCTION

T lymphocytes that recognize specific tumor antigens can be isolated from LNs 3 draining a progressively growing tumor or a tumor vaccine site (1) . The LN T cells acquire potent antitumor effector function after ex vivo activation with anti-CD3 mAb or bacterial superantigens, whereas they are not therapeutically active when freshly isolated (2 , 3) . The T cells mediate the regression of advanced tumors in visceral, s.c., and intracranial sites and are exquisitely specific for the tumor that provided the in vivo sensitization (4 , 5) . The specificity implies that the T cells must recognize and interact with tumor-associated antigens presented in the context of MHC molecules in vivo after adoptive transfer to initiate the antitumor response. It is presumed that this antigen stimulation requires direct physical contact of T cells with tumor cells. However, several lines of evidence suggest that the antitumor immune reaction that is generated by ex vivo activation of tumor-draining LN T cells is not dependent on the classical CD8-mediated cytolytic response. The activated tumor-draining LN cells do not lyse ⁵¹Cr-labeled tumor cells in the conventional cytotoxicity test. In previous studies, we have found that both CD4 and CD8 T cells collaboratively participated in mediating the regression of the murine MCA 205 fibrosarcoma (4 , 5) . Because the MCA 205 tumor does not express constitutive or inducible MHC class II molecules, indirect recognition of tumor antigens by the transferred CD4 T cells may play a significant role in initiating the antitumor response, which need not necessarily occur within the tumor. This provided an opportunity to investigate whether trafficking of T cells into tumors is a prerequisite for successful adoptive immunotherapy.

Trafficking of leukocytes from the systemic circulation into tissues is a dynamic multistep process requiring attachment to endothelium, activation of integrins, and transendothelial migration (6 , 7) . A critical step in this process is mediated by signals delivered by chemokines that are produced in lymphatic tissues and at the site of inflammation (8) . There is pleiotropy in chemokines and their receptors on leukocytes (9) . Moreover, the specific types of chemokines produced within tumors and their effects on effector T lymphocytes are not completely understood (10) . However, chemokine receptors signal through G proteins, which provides a convenient method to block their action. Treatment of cells with PTX ADP-ribosylates the Gα_i and Gα_o subunits of G proteins and inhibits the dissociation of the α from the βγ subunits on ligand binding and consequently prevents downstream signaling events (11, 12, 13) . Lymphocytes treated with PTX fail to enter LNs, splenic white pulp, or sites of inflammation in tissues (14, 15, 16, 17) . In this study, therapeutically effective tumor-reactive T cells were treated ex vivo with PTX before systemic transfer into recipients bearing pulmonary tumor metastases. These cells failed to infiltrate tumors and consequently did not mediate tumor regression. However, PTX-treated effector T cells retained the capacity to respond to tumor antigens and mediate an antitumor response in vivo if the requirement for trafficking was obviated by admixing effector T cells and tumor cells before inoculation.

MATERIALS AND METHODS

Animals.

Female C57BL/6 (B6) mice at 6–8 weeks of age were purchased from the Biological Testing Branch, Frederick Cancer Center Research and Development Center (National Cancer Institute, Frederick MD). They were maintained in a specific pathogen-free environment, fed ad libitum according to NIH guidelines, and used for experiments at 8–10 weeks of age.

Tumor.

The MCA 205 fibrosarcoma is a 3-methylcholanthrene-induced tumor of B6 origin (18) . The tumor was maintained by serial s.c. transplantation in syngeneic mice and was used in the third to eighth passage. Single cell suspensions were prepared from solid tumors by digestion in 0.1% collagenase type IV, 0.01% DNase I, and 2.5 units/ml hyaluronidase type V (Sigma Chemical Co.) for 3 h at room temperature, as described previously (5) .

Activation of Effector T Cells and ex Vivo PTX Treatment.

B6 mice were inoculated sc with 1.5 × 10⁶ viable MCA 205 tumor cells in the lower flank region bilaterally. Twelve days later, tumor-draining inguinal LNs were removed under sterile conditions and single cell suspensions prepared mechanically and resuspended in CM, as described previously (5) . CM consisted of RPMI 1640 supplemented with 10% heat-inactivated FCS, 0.1 mm nonessential amino acids, 1 mm sodium pyruvate, 2 mm l-glutamine, 100 μg/ml streptomycin, 100 units/ml penicillin, 50 μg/ml amphoterecin B (all from Life Technologies, Inc.), and 5 × 10^-5 M 2-mercaptoethanol (Sigma Chemical Co.). LN cells were activated with immobilized anti-CD3 mAb (145–2C11) in 24-well tissue culture plates at 4 × 10⁶ cells/well in 2 ml of CM in 5% CO₂ at 37.4°C for 2 days. After anti-CD3 activation, cells were harvested, and further cultured in CM with 24 IU/ml of human recombinant IL-2 at 2 × 10⁵/ml in gas-permeable culture bags (Baxter Heathcare Corp., Deerfield, IL) for 3 days. Splenic lymphocytes from normal mice were obtained by mechanical dissociation, treatment with ACK, and ex vivo activation as described above. PTX (100 ng/ml; Sigma Chemical Co.) was added to half of the cells for the final 24 h of culture. Cells were harvested, washed, and resuspended in HBSS for adoptive transfer, or in CM for ELISA.

Adoptive Immunotherapy and Winn Assay.

B6 mice were injected through the tail vein with 3 × 10⁵ MCA 205 tumor cells in 1 ml of HBSS to establish pulmonary metastases. Three days later, mice were injected i.v. with 1 × 10⁷ effector cells suspended in 1 ml of HBSS. On day 18 after tumor inoculation, mice were sacrificed, lungs were insufflated with India ink, and metastatic nodules on the lung surface were counted in a blinded fashion. Metastatic foci too numerous to count were assigned a value of 250 because this was the greatest number of nodules that could be reliably enumerated. For Winn assay, LN cells, PTX-treated LN cells, or normal splenocytes were harvested after activation and 1 × 10⁷ cells were mixed with 1 × 10⁶ MCA 205 tumor cells in 0.1 ml of HBSS and inoculated s.c. on the flank. Tumors were measured in two perpendicular dimensions with a vernier caliper twice weekly and the size recorded as tumor area.

In Vivo Trafficking Assay.

For analysis of effector cell trafficking at 16 h after transfer, cells were labeled with rhodamine, as described previously (19 , 20) . Briefly, cultured effector cells were washed, and 2.5 × 10⁸ cells were incubated with 150 μg of rhodamine (TRITC; Sigma Chemical Co.) in 250 ml of RPMI 1640 containing 24 IU/ml IL-2 for 40 min at 37°C. After labeling, cells were washed twice in RPMI 1640 containing 5% FCS and resuspended in HBSS for adoptive transfer. For analysis at 48 and 96 h, effector cells were labeled with CFDASE (Molecular Probes Inc.). Cultured effector cells were washed twice with HBSS, and 1 × 10⁷ cells/ml were incubated with CFDASE (5 μm) in HBSS for 15 min at 37°C. Labeling was stopped by adding cold HBSS, and cells were washed twice with HBSS containing 5% FCS before adoptive transfer. Mice bearing 10-day pulmonary tumors received 500 cGy whole body irradiation from a ¹³⁷Cs source then were injected i.v. with 5 × 10⁷ flurochrome-labeled effector cells. Mice were sacrificed at 16 h, 48 h, and 96 h after adoptive transfer, and lungs were harvested, fixed with 4% formalin for 24 h, and subsequently placed in 30% sucrose for an additional 24 h. After fixation, the tissue was snap-frozen in n-Hexane at -70°C and cut into 8-μm sections on a cryostat. The sections were dried and examined under a fluorescent microscope equipped with a filter combination of BP 545 for TRITC detection and BP 490 for CFDASE detection. After enumeration of fluorescent cells, sections were counterstained with hematoxylin to confirm the presence of metastases by light microscopy. In a separate experiment of identical design, splenocytes were harvested 24 and 48 h after adoptive transfer, treated with ACK to lyse erythrocytes, and the number of CFDASE-positive cells was determined by fluorescence-activated cell-sorting analysis.

IFN-γ Release Assay.

Activated effector cells (2 × 10⁶) were restimulated with 50 Gy irradiated MCA 205 tumor cells (1 × 10⁶) in 2 ml of CM in 24-well tissue culture plates. The plates were incubated for 24 h at 37°C, and supernatants were analyzed in triplicate in a sandwich ELISA, as described previously (19) .

Statistical Analysis.

The significance of differences in numbers of pulmonary metastases between different groups of mice was determined using the nonparametric Wilcoxon rank-sum test. The difference for the in vivo migration assay and s.c. tumors was determined using an unpaired Student’s t test analysis. In all experiments, two-sided Ps are presented.

RESULTS

PTX Treatment Abrogates the Efficacy of Systemically Transferred T Cells.

LNs draining MCA 205 s.c. tumors were used as a source of tumor-sensitized T cells. Previous studies have shown that tumor-draining LNs are the richest source of tumor-sensitized T cells. LN cells were activated with anti-CD3 mAb for 2 days, followed by culture in media containing 24 IU/ml of IL-2 for 3 days. These culture conditions induce proliferation of T lymphocytes, and the resulting cells are typically >95% TCRαβ⁺ consisting of 65–75% CD8 and 20–30% CD4 cells. Half of the T cells were treated with PTX for the final 24 h of culture. At the dose of 100 ng/ml PTX, there was 20–25% less proliferation of the T cells, but their viability was similar to untreated cells. Fluorescence-activated cell-sorting analysis showed that the expression of the cell adhesion molecules LFA-1, ICAM-1, VLA-4, and l-selectin were identical for PTX-treated and untreated cells (data not shown). Both PTX-treated and untreated T cells were used for systemic adoptive immunotherapy of 3-day established pulmonary metastases. As demonstrated in Fig. 1 ⇓ , the PTX-treated cells had no antitumor efficacy, whereas the same number of untreated T cells eliminated the tumors, P < 0.01 for each experiment.

PTX-treated T Cells Fail to Infiltrate Pulmonary Metastases.

The trafficking of PTX-treated T cells following systemic transfer into tumor-bearing mice was compared with that of untreated T cells. Mice bearing 10-day pulmonary metastases of MCA 205 were treated with systemic transfer of 5 × 10⁷ T cells that were labeled with the fluorochromes TRITC, or CFDASE. The fluorochrome labeling does not inhibit the antitumor efficacy of the T cells, and CFDASE permits identification of labeled cells in tissues for at least 4 days. As shown in Fig. 2 ⇓ , there were marked differences in the distribution of the PTX-treated T cells at all time points up to 96 h after transfer. Whereas the untreated, activated tumor-draining LN T cells formed a dense infiltrate within the tumors, the PTX-treated cells were located in the normal lung tissue with very few cells within the tumors. These differences were quantified by enumerating the number of fluorescent cells in tumors and in normal lung tissue 16 h after adoptive transfer. At this time point, the metastatic nodules analyzed had a mean size of 0.031 mm² and were not different between the mice that received PTX-treated cells versus untreated effector cells. Whereas the metastatic tumor nodules had a mean of 100 ± 5 normal effector cells, there were only 16 ± 2 PTX-treated cells P = 0.0001. In contrast, normal lung tissue in mice that received PTX-treated cells contained 65 ± 3 cells/mm² in contrast to 38 ± 2 cells/mm² for normal effector cells. By 96 h after cell transfer, the metastatic nodules in the PTX-treated animals were considerably larger and more numerous than in the mice that received untreated effector cells as a result of tumor regression in the later group. These results show that the PTX-treated cells are markedly inhibited from entering tumors for at least 96 h, but instead are retained in the vasculature. The number of cells labeled with CFDASE recovered from the spleens of recipients of PTX-treated cells was compared with recipients of labeled untreated cells. At 24 h, there were 71% as many PTX-treated cells, and at 48 h there were 61% as many PTX-treated cells present in the spleen.

PTX-treated Effector T Cells Retain Antitumor Effector Function.

The activated tumor-draining LN T cells do not have cytolytic activity in the ⁵¹Cr release assay, however, they do produce IFN-γ when stimulated with MCA 205 tumor cells. The secretion of IFN-γ is specific for the tumor that provided the in vivo sensitization and is likely produced only by the subset of the T cells that mediate tumor regression in vivo. In this assay, a single-cell suspension derived from an MCA 205 tumor growing in vivo stimulated IFN-γ production, whereas an MCA 205 cell line maintained in culture does not. Presumably, this is due to the presence of MHC class II positive macrophages in the tumor digest. There were no significant differences in the secretion of IFN-γ by PTX-treated cells compared with untreated effector cells on stimulation with either MCA 205 tumor cells or anti-CD3 (Table 1) ⇓ . These results show that antigen recognition and cytokine production are intact in PTX-treated cells despite the lack of therapeutic efficacy when systemically transferred (Fig. 1) ⇓ . To further determine whether the PTX-treated cells retained antitumor function in vivo, they were coinoculated with MCA 205 tumor cells s.c. in a Winn assay. As shown in Fig. 3 ⇓ , the tumor growth rate was significantly different in mice inoculated with PTX-treated cells mixed with tumor cells compared with mice inoculated with tumor alone or tumor mixed with anti-CD3-activated normal splenocytes (P < 0.01 for each experiment). The antitumor function of the PTX-treated cells was not statistically different from untreated effector cells and they eliminated tumors in 7 of 10 mice.

DISCUSSION

A prevalent concept in cancer immunotherapy is that infiltration of effector cells into tumors and direct killing of tumor cells is required. Some mechanisms of target killing by T cells, such as Fas/Fas ligand interactions or perforin release require direct physical contact. Other mechanisms, such as cytokine production and helper functions for other effector cells can be operative with either direct or indirect contact between the T cells and tumor cells. Although effector cells generated by the method described here exert their antitumor effects through the collaboration of CD4 and CD8 T cells, our recent findings indicate that if LN T cells are purified for low l-selectin expression, CD4 cell can mediate potent antitumor effects in the absence of CD8 cells (21) . The high therapeutic activity of CD4 T cells against tumors, such as MCA 205, that do not express MHC class II molecules is likely mediated through indirect recognition of antigens presented by antigen-presenting cells. Before this study, it was uncertain whether or not the indirect presentation of tumor antigens to effector cells had to occur within the tumor. This study demonstrates that infiltration of tumor by the transferred effector T cells is required for tumor regression to occur. It is important to note that only the transferred leukocytes were exposed to PTX and the host leukocytes had no impediment to trafficking. Thus, it seems that local cytokine effects or modulation of host cells are essential for adoptive immunotherapy rather than systemic effects.

Interestingly, extravasation of T cells from the vasculature was required to treat microscopic 3-day pulmonary metastases. The pulmonary metastases are established by i.v. inoculation of a single-cell suspension of tumor cells. Initially, the tumor cells lodge or adhere in the pulmonary vasculature. Subsequently, tumor cells extravasate through endothelium before colonizing the lung parenchyma. The lung is highly vascularized, and, as shown in Fig. 2 ⇓ , trafficking studies demonstrated that unlike untreated effector cells, many PTX-treated T cells were found in the vicinity but not within tumors. These conditions should favor the observation of any regional antitumor effects that are not dependent on infiltration of T cells into the tumor parenchyma. The fact that no therapeutic efficacy was observed against pulmonary metastases underscores the importance of trafficking for transferred effector T cells. Thus, culture conditions or concomitant treatments of the host that inhibit trafficking of effector cells would be expected to interfere with the therapeutic efficacy of adoptive immunotherapy and should be considered in the development of immunotherapy. Trafficking alone is not sufficient, however, antigen recognition must also occur. We recently demonstrated that the subpopulation of anti-CD3/IL-2 activated T cells with low expression of CD62L selectively infiltrate tumors. Interestingly, activated normal splenocytes or LN cells draining antigenically distinct tumors with low expression of CD62L also selectively infiltrated tumors but did not have any therapeutic efficacy (20) .

In previous studies, we have shown that systemically transferred T cells can infiltrate and cure established intracranial tumors (5 , 19 , 22) . In other experiments not reported here, we treated mice bearing 10-day established intracranial tumors with PTX-treated effector T cells labeled with CFDASE. The PTX-treated effector cells failed to infiltrate intracranial tumors similar to the data in Fig. 2 ⇓ , and there was no apparent therapeutic antitumor effect mediated by the PTX-treated cells. These findings are in agreement with the pulmonary metastasis tumor model presented here. Interestingly, several of the mice with intracranial tumors that received PTX-treated cells also developed neurological symptoms. It is possible that inflammation related to the intracranial inoculation of tumor cells combined with the adoptive transfer of PTX treated T cells contributed to the development of neurological symptoms in these animals.

A view is emerging that particular combinations of chemokines coupled with differential expression of chemokine receptors is used to direct the trafficking of naïve and effector lymphocytes (9) . Recent evidence suggests that preferential recruitment of Th1 cells through CXCR3 and CCR5 or Th2 cells through CCR3 may be used to tailor immune responses in peripheral tissues (23, 24, 25) . However, several chemokines can interact with even this limited set of receptors. This redundancy complicates the analysis of experiments in which a single chemokine is blocked or overexpressed. An additional consideration is that the inflammatory response initiated by the transferred T cells on recognition of specific tumor antigens is likely to dynamically change the local production of chemokines within tumors during the effector response. To avoid some of the potential complexities associated with blockade of a single chemokine, we used PTX to inhibit chemokine signaling through G proteins. High concentrations of PTX have additional effects on T lymphocytes, however, doses required to inhibit G proteins do not affect TCR- or mitogen-induced activation (26 , 27) . It is apparent that the doses of PTX used in this study did not effect antigen recognition.

The question of whether the signals provided by chemokines modify lymphocyte effector functions in addition to directing their migration has not been extensively studied in vivo. It is conceivable that chemokine signals help prime effector cells as they migrate toward sites of inflammation within tissues. Our studies demonstrated that when the requirement for diapedesis was experimentally bypassed, by mixing tumor and effector T cells before inoculating into animals, the tumors were eliminated. These observations suggest that effector functions of T cells are not affected under conditions of blockade of chemokine signaling.