Phenotype and Homing of CD4 Tumor-Specific T Cells Is Modulated by Tumor Bulk1
http://www.100md.com
免疫学杂志 2005年第14期
Abstract
Technical difficulties in tracking endogenous CD4 T lymphocytes have limited the characterization of tumor-specific CD4 T cell responses. Using fluorescent MHC class II/peptide multimers, we defined the fate of endogenous Leishmania receptor for activated C kinase (LACK)-specific CD4 T cells in mice bearing LACK-expressing TS/A tumors. LACK-specific CD44highCD62Llow CD4 T cells accumulated in the draining lymph nodes and had characteristics of effector cells, secreting IL-2 and IFN- upon Ag restimulation. Increased frequencies of CD44highCD62Llow LACK-experienced cells were also detected in the spleen, lung, liver, and tumor itself, but not in nondraining lymph nodes, where the cells maintained a naive phenotype. The absence of systemic redistribution of LACK-specific memory T cells correlated with the presence of tumor. Indeed, LACK-specific CD4 T cells with central memory features (IL-2+IFN-–CD44highCD62Lhigh cells) accumulated in all peripheral lymph nodes of mice immunized with LACK-pulsed dendritic cells and after tumor resection. Together, our data demonstrate that although tumor-specific CD4 effector T cells producing IFN- are continuously generated in the presence of tumor, central memory CD4 T cells accumulate only after tumor resection. Thus, the continuous stimulation of tumor-specific CD4 T cells in tumor-bearing mice appears to hinder the systemic accumulation of central memory CD4 T lymphocytes.
Introduction
Protective immunity requires both MHC class I-restricted cytotoxic CD8 T cells and MHC class II-restricted CD4 Th cells. Although tumor-specific CD8 T cell responses have been extensively studied in animal models of tumor disease (1, 2, 3, 4, 5, 6, 7, 8) and to some extent in cancer patients (9, 10, 11, 12), more limited information is available for CD4 T cells. This is surprising given their critical role in the priming, differentiation, and persistence of CTL (13, 14, 15) and during the effector phase of an antitumor immune response (16, 17, 18, 19, 20, 21).
The paucity of available information is possibly explained by the fact that only a limited number of MHC class II-restricted epitopes have been identified to date (22) and by the inability to trace the low number of CD4 T cells specific for a given Ag. To overcome these limitations, tumor-specific CD4 T cell responses have been studied in adoptive transfer models, where traceable populations of TCR transgenic (Tg)7 T cells specific for a model tumor-associated Ag can be visualized after tumor cell injection (4, 8, 23, 24, 25, 26). In some cases, tumor-specific CD4 T cell proliferated as a result of tumor growth, but soon afterward became unresponsive (23, 26). In other cases, CD4 T cells differentiated into effector cells, but rapidly disappeared from the host (8). Although these studies have provided some insights into the mechanisms that regulate tumor-specific CD4 T cell responses, one of their limitations was that the tumor-specific T cells were derived from monoclonal populations of cells expressing the same TCR and were therefore not representative of a polyclonal response. Furthermore, homeostatic mechanisms in mice adoptively transferred with a limited number of tumor-specific T cells are likely to be different from those occurring in unmanipulated immunocompetent animals in which tumor-specific T cells are continuously generated from a renewable population of thymic progenitors. Finally, the adoptive transfer of defined numbers of T cells does not allow an analysis of the impact of continuously primed new lymphocytes on the development of T cell memory.
Recently, the use of MHC class II multimers has allowed endogenous Ag-specific CD4 T cells to be studied in patients with rheumatoid arthritis and in Ag-vaccinated or infected mice (27, 28, 29, 30, 31). Schepers et al. (31) exploited this technology to study CD4 T cells in mice bearing Moloney murine sarcoma and leukemia virus complex-induced tumors. Even though this study underlined the importance of CD4 T cells in promoting protection against virus-infected tumor cells, its major limitation was that the T cell response was initiated by the viral infection, not by the tumor itself. Moreover, in this model the tumor was rejected as a consequence of the antiviral response, and as such, the impact of tumor growth on the development of the CD4 T cell response could not be assessed.
In this study we have used peptide/MHC class II fluorescent multimers (32) to characterize endogenous CD4 T cells specific for a tumor-restricted Ag. This is the first demonstration of the temporal and spatial organization of an antitumor CD4 T cell response during tumor progression. Moreover, we have characterized for the first time the fate of tumor-specific lymphocytes after resection of an established tumor and investigated the establishment of tumor-specific CD4 T cell memory.
Materials and Methods
Mice and tumor cells
16.2 Tg (BALB/c background) mice were previously described (32, 33). TS/A mouse mammary adenocarcinoma cells (34) were grown in vitro at 37°C in complete medium (RPMI 1640-10% FBS, 100 U/ml penicillin, 100 U/ml streptomycin, and 2.5 x 10–5 M 2-ME; Invitrogen Life Technologies). TS/A-LACK cells were obtained by transfecting TS/A cells with a pcDNA3-derived vector in which a truncated form of the LACK cDNA (35) was cloned at the HindIII site downstream from the CMV promoter. G418-resistant cells were selected, and single clones were obtained by limiting dilution. Exponentially growing 3 x 105 TS/A and TS/A-LACK cells were resuspended in 100 μl of PBS and injected s.c. into the left flank. The axillary, brachial, and inguinal lymph nodes (LN) draining the site of tumor cell injection were referred as to draining, and the axillary, brachial, and inguinal LN derived from the contralateral flank were referred as to nondraining. The growth of the tumors was monitored every other day by measuring tumor diameters using a metric caliper. Tumor diameters were measured in two dimensions, and their volume was calculated according to the following formula: volume (mm3) = (width, mm)2 x (length, mm) x 1/2. In selected experiments, mice were treated with 150 μg of anti-CD4 mAbs 4 and 2 days before the injection of TS/A-LACK cells, then once a week until the end of the experiment. Where indicated, mice were challenged with 2 x 105 bone marrow-derived DC, which were matured for 8 h in the presence of LPS (36) and loaded with the LACK immunodominant CD4 peptide (FSPSLEHPIVVSGSWD). In some experiments mice were anesthetized and either sham operated or resected of the tumor mass, according to standard procedures. All in vivo studies were approved by the ethical committee of San Raffaele Scientific Institute and Centre National de la Recherche Scientifique and performed according to their guidelines.
Western blot analysis
TS/A and TS/A-LACK cells were lysed in PBS by five consecutive freeze and thaw cycles. The soluble fraction was collected after 15 min of centrifugation at 15,000 rpm, and the protein content was determined by the Bio-Rad protein assay. Equal amounts of proteins were then separated on 12% SDS-PAGE and transferred to nitrocellulose filters. The filters were incubated with the p19–11 anti-LACK mAb (37) and detected by chemiluminescence (ECL; Amersham Biosciences).
Results
TS/A-LACK tumors develop in 16.2 mice despite the presence of LACK-specific CD4 T cells
Tumor cells expressing LACK were obtained by transfecting TS/A mammary adenocarcinoma cells with a plasmid carrying a truncated form of the LACK cDNA under control of the CMV promoter. Stable transfectants were selected and analyzed for LACK expression by Western blot using anti-LACK mAb (Fig. 1A). Of the clones expressing LACK, one was selected for additional experiments (TS/A-LACK 13.3, referred to hereafter as TS/A-LACK). Compared with parental TS/A cells, TS/A-LACK cells expressed similar surface levels of MHC class I, did not express MHC class II molecules, and proliferated to a similar extent in vitro (data not shown). Moreover, as previously reported for TS/A tumor cells (34), injection of TS/A-LACK tumor cells in BALB/c mice resulted in the development of solid tumors (not shown). We then assessed whether TS/A-LACK cells would cause tumors in mice harboring LACK-specific CD4 T cells. TS/A-LACK cells (3 x 105) were injected into 16.2 mice, which express a TCR Tg -chain derived from a LACK-specific hybridoma. As such, these mice exhibit an increased frequency of LACK-specific CD4 T cells (32). Despite the presence of LACK-specific CD4 T cells, TS/A-LACK cells produced solid tumors in all injected animals (Fig. 1B). Tumor growth was not due to the loss of LACK expression in vivo, as revealed by Western blot analyses performed on lysates obtained from resected tumors (data not shown).
To investigate whether the presence of CD4 T cells impacted the growth of TS/A-LACK tumors, tumor development was compared in untreated and anti-CD4-treated 16.2 mice. TS/A-LACK tumors grew significantly faster and achieved larger sizes after treatment of 16.2 mice with an anti-CD4 mAb (p < 0.01; Fig. 1C). These data indicate that TS/A-LACK tumors develop in the presence of an immune response. Thus, this model provides an ideal setting in which a tumor-specific CD4 T cell response can be characterized over the course of tumor progression.
Biodistribution of activated, tumor-specific CD4 T cells
To formally evaluate the status of LACK-specific CD4 T cells in the presence of tumor, we assessed whether a TS/A-LACK tumor induced the activation and expansion of these CD4 T cells. TS/A-LACK or TS/A tumor cells (3 x 105) were injected into the left flank of 16.2 mice. The axillary, brachial, and inguinal LN from the same flank where tumor cells were injected (draining) and from the contralateral flank (nondraining) were recovered at different times after injection of tumor cells. Both the frequency and number of LACK-specific CD4 cells, evaluated using I-Ad/LACK multimers, rapidly increased in the draining LN shortly after injection of TS/A-LACK cells (Fig. 2, A and B). LACK-specific CD4 T cells were activated, because by day 4 after injection the majority of I-Ad/LACK+CD4+ cells in tumor draining LN exhibited a blast phenotype (Fig. 2C), whereas I-Ad/LACK+CD4+ cells in TS/A-challenged (data not shown) or naive mice (Fig. 2D) had a small resting phenotype. Furthermore, I-Ad/LACK+CD4+ cells in the draining LN of mice injected with TS/A-LACK, but not TS/A tumors, entered into the cell cycle and up-regulated CD69 and CD44 activation markers (Fig. 2E and data not shown). In addition to the draining LN, the proportion of I-Ad/LACK+ cells increased in the spleen, where it was 1.1 ± 0.2% on day 0, 2.5 ± 0.3% on day 3, and 1.9 ± 0.3% on day 21 (data not shown). In contrast, the frequency and total number of I-Ad/LACK+ cells in the nondraining LN of TS/A-LACK tumor-bearing mice remained comparable to those in LN of naive and TS/A-bearing mice (Fig. 2, A, B, and D).
To more precisely study the relationship between the distribution and the activation of LACK-specific CD4 T cells, their phenotype in lymphoid and nonlymphoid tissues (blood, lung, liver, and tumor) was assessed 21 days after challenge with TS/A or TS/A-LACK tumor cells (Fig. 3). The presence of tumor per se did not modulate LACK-specific CD4 T cells, because the frequency and surface phenotype of I-Ad/LACK+CD4+ T cells within the lymphoid and nonlymphoid tissues of TS/A tumor-bearing mice were comparable to those in naive mice. In marked contrast, there were increased frequencies of I-Ad/LACK+CD4+ T cells expressing high levels of CD44 in the tumor-draining LN, spleen, blood, lung, liver, and tumor derived from TS/A-LACK-challenged mice. However, this increase was not observed in the nondraining LN, where I-Ad/LACK+CD4+ T cells preserved a naive phenotype (CD44lowCD62Lhigh; Fig. 3). This was despite the development of large tumors (>300 mm3).
The failure of Ag-experienced LACK-specific T cells to migrate to the nondraining LN correlated with the loss of the LN-homing molecule, CD62L, from the surface of activated T cells. Indeed, most CD44highI-Ad/LACK+CD4+ T cells found within the tumor-draining LN and spleen (and nonlymphoid tissues) expressed low levels of CD62L (Fig. 3B). Thus, it appears that the activation status of LACK-specific T cells determines their localization.
The experiments presented above were performed in 16.2 mice, because of their sizeable frequency of I-Ad/LACK+CD4+ T cells, which allows these cells to be easily followed within the global T cell population (32). Nevertheless, it was possible that the transgenic TCR -chain somehow skewed the distribution of the cells in tumor-bearing animals. To formally exclude this possibility, similar experiments were performed in mice with a normal lymphocyte repertoire (BALB/c). Because of the extremely low frequency of LACK-specific CD4 T cells (30), we specifically looked for the appearance of CD44high cells within the LACK-specific T cell population in tumor-bearing mice. Importantly, a significant percentage of I-Ad/LACK+CD4+ T cells in the tumor-draining LN, liver, and tumor of TS/A-LACK-tumor-bearing mice expressed high levels of CD44, whereas this was not observed in TS/A tumor-bearing mice (Fig. 4). However, as in 16.2 mice, there was no recirculation of memory LACK-specific CD4 T cells in the nondraining LN of BALB/c bearing TS/A-LACK tumors (Fig. 4). Thus, the accumulation of LACK-specific effector T cells in the tumor-draining LN and nonlymphoid tissues, but not in the nondraining LN appeared to be a property of tumor-bearing mice.
TS/A-LACK tumors induce the differentiation of IFN--secreting, LACK-specific CD4 T cells in draining LN
The cytokine secretion profiles of lymphocytes derived from the lymphoid and nonlymphoid organs of naive mice as well as TS/A-LACK and TS/A tumor-bearing mice were assessed. Lymphocytes were stimulated in vitro with the LACK peptide, and IL-2, IFN-, and IL-4 secretion was measured 24 h later by ELISA. Compared with T cells from naive and TS/A tumor-bearing mice, those from TS/A-LACK tumor-bearing animals secreted higher amounts of IL-2 and IFN- in response to the LACK peptide in a dose-dependent manner (Fig. 5). On a per cell basis, cells derived from the draining lymph node (Fig. 5A), the nonlymphoid tissues (Fig. 5B) and the TS/A-LACK tumor itself (Fig. 5C) produced comparable amounts of these cytokines. In contrast with Leishmania major infection (30, 32), IL-4 levels remained below the level of detection in all samples throughout tumor growth (data not shown). However, LACK-specific T cells in nondraining LN from mice bearing TS/A-LACK tumor did not secrete cytokines at levels above those detected in T cells isolated from naive mice (Fig. 5A). Together, these data indicate that TS/A-LACK tumor induced the proliferation and differentiation of LACK-specific CD4 T cells into IFN--producing effector T cells in tumor-draining LN. These effector cells were also found in the spleen, lung, liver, and tumor itself, but were not detected in nondraining LN.
I-Ad/LACK+ central memory T cells accumulate after DC immunization or tumor resection
In the experiments presented above, it appeared that TS/A-LACK tumor-bearing mice presented LACK-specific T cells with a CD44highCD62Llow effector phenotype, capable of secreting IL-2 and IFN-. In contrast, central memory T lymphocytes, characterized as CD44highCD62Lhigh, secreting IL-2 alone (40, 41, 42, 43, 44), and able to circulate within all peripheral LN, failed to accumulate in these mice. It was therefore important to determine whether the absence of LACK-specific central memory T cells was a characteristic of these transgenic mice or, alternatively, was due to the presence of tumor. To this aim, we compared the phenotype of LACK-experienced T cells in 16.2 mice harboring TS/A-LACK tumors and in 16.2 mice immunized with LACK-loaded, bone marrow-derived DC, which are known to induce potent T cell responses (45). TS/A-LACK tumors and LACK-loaded DC were injected into one flank of individual mice. Fifteen days later, the draining and nondraining LN were recovered and analyzed by flow cytometry. As expected, both Ag challenges resulted in an increased frequency of I-Ad/LACK+ CD4 T cells in the draining LN compared with naive mice (1.77 and 1.24 vs 0.6%, respectively; Fig. 6). However, the phenotypes of these cells were significantly different; >80% of CD44high LACK-specific cells in tumor-bearing mice had down-regulated expression of CD62L, whereas the majority of CD44high cells in DC-immunized mice preserved the expression of the CD62L homing molecule (Fig. 6, A and C). Moreover, the cytokine secretion profile of these cells were distinct; in tumor-bearing mice, 50% of cytokine-secreting cells expressed both IL-2 and IFN-, although >90% of cytokine-secreting cells in DC-immunized mice expressed IL-2 alone (Fig. 6, B and C). Furthermore, DC activation resulted in the accumulation of CD44high IL-2-secreting I-Ad/LACK cells in nondraining LN, although the T cell tumor response was restricted to draining LN. The ensemble of these data indicate that 16.2 mice are capable of mounting a LACK-specific central memory CD4 T cell response, but that this response is dependent upon the manner in which the Ag is presented. Upon presentation of LACK Ag in the context of a tumor, there is a strong effector T cell response limited to draining LN. In contrast, when the LACK Ag is presented by LPS-matured DC, the majority of the response is a central memory response, and these LACK-experienced T cells accumulate in all peripheral LN.
To determine whether the presence of the tumor itself modulates the nature of the T cell response, i.e., effector vs central memory, established TS/A-LACK tumors were surgically removed 15 days after tumor cell injection in 16.2 mice. As expected, the percentages of I-Ad/LACK+ T cells detected in the draining LN and spleens of tumor-resected mice decreased by >50% within 4 days after surgery (3.4 to 1.2% in the draining LN and 1.8 to 0.9% in the spleen of a representative mouse; Fig. 7A). The frequency of I-Ad/LACK+ T cells exhibiting a blast phenotype (Fig. 7B) and expressing high levels of CD44 (Fig. 7C) also decreased, and LACK-specific T cells did not secrete detectable levels of IFN- by day 4 (Fig. 7C). Importantly though, the percentages of CD44highI-Ad/LACK+ cells increased significantly in the nondraining LN of tumor-resected mice within 10 days after surgery (Fig. 8). This increase was specific to the tumor resection, because it was not detected in sham-operated mice. Moreover, in marked contrast with the phenotype in tumor-bearing mice, the vast majority of CD44high cells in tumor-resected mice expressed CD62L (80% in a representative mouse; Fig. 8, B and C) and produced IL-2 (Fig. 8C).
Together, these data indicate that resection of established TS/A-LACK tumors causes the rapid disappearance of IFN--producing effector T cells from the tumor-draining LN and spleen, with a concomitant accumulation of IL-2-producing central memory lymphocytes in the nondraining LN.
Discussion
This is the first study assessing the temporal and spatial organization of an endogenous tumor-specific CD4 T cell response during the course of tumor development, with evaluation of the establishment of tumor-specific T cell memory.
16.2 Tg mice are an important model in which to monitor endogenous T cell responses, because they exhibit an increased frequency of CD4 T cells directed against a specific Ag (LACK) within a context of fairly normal T cells repertoire. Using peptide-MHC class II fluorescent multimers against LACK, we were able to define the site of initial Ag encounter and track the fate of Ag-experienced T cells within both lymphoid and nonlymphoid tissues of tumor-bearing mice. Tumor-specific CD4 T cells appeared to be exclusively primed in the tumor-draining LN, but not in other lymphoid organs. Thus, in TS/A-LACK tumor-bearing mice, only the tumor-draining LN, and not nondraining LN or spleen, contained an increased frequency of I-Ad/LACK+ cells expressing the early activation marker CD69 (data not shown). Furthermore, although a sizeable number of I-Ad/LACK+ cells exhibited a blast phenotype in the draining LN, the vast majority of these cells in the spleen and nondraining LN remained small even after the development of large tumors. Finally, only the draining LN of TS/A-LACK tumor-bearing mice contained DC, which were able to stimulate LACK-specific hybridoma in vitro in the absence of added LACK peptide (S. Hugues and N. Gleichenhaus, unpublished observations). Thus, together with the results reported by Marzo et al. (4, 46, 47), our data argue that the LN most proximal to the tumors are the main lymphoid tissues reacting to the growing tumor.
LACK encounter in the tumor-draining LN induced I-Ad/LACK+ T lymphocytes to proliferate and differentiate. I-Ad/LACK+ T lymphocytes up-regulated CD44 surface expression, down-regulated the LN-homing molecule CD62L, and were capable of producing both IL-2 and IFN- when restimulated in vitro. Ag-experienced LACK-specific CD4 T cells were also found in the spleen as well as in nonlymphoid tissues, such as lung and liver, and the tumor mass itself, where they retained the ability to produce IL-2 and IFN- upon Ag re-encounter. Thus, in contrast with other tumor models, no evidence for CD4 T cell anergy was found in TS/A-LACK tumor-bearing mice. The finding that depletion of CD4 T cells resulted in faster tumor development also indicates that protective CD4 T cell responses naturally developed in these mice.
Recent studies performed in humans and in immunized and pathogen-infected mice have shown that Ag-experienced T cells are heterogeneous in nature, with subpopulations of effector memory and central memory cells. Although effector memory T cells migrate to tissues and secrete effector cytokines, central memory T cells migrate through the LN and lack immediate effector function (40, 41, 42, 43, 44). In mice, central memory T cells are best defined by their localization in peripheral LN, surface expression of high levels of CD44 and CD62L, and the ability to produce IL-2, but not other cytokines (40, 41, 42, 43, 44). In this study we show that CD44highCD62LlowI-Ad/LACK+ effector T cells were always found in TS/A-LACK-tumor-bearing mice. In contrast, CD44highCD62LhighI-Ad/LACK+ central memory T cells failed to accumulate in tumor-draining LN and were never found in LN distal to the site of tumor growth. Thus, cells in these latter locations appeared to remain ignorant of tumor growth.
These findings might help to reconcile apparently conflicting published results. Indeed, some authors reported the presence of tumor-specific T lymphocytes capable of IFN- production in the blood and lymphoid organs of tumor-bearing mice (1, 7, 24, 46, 48). In contrast, other reports in which TCR Tg CD4 T cells were adoptively transferred into tumor-bearing mice showed that the cells appeared to loose the ability to proliferate, secrete effector cytokines in vitro, and undergo deletion (8, 23, 49). We found that tumor-specific CD4 effector cells capable of IFN- production were continuously generated in tumor-draining LN. However, most of the Ag-experienced T cells rapidly lost CD62L expression, and exited the peripheral LN to relocate in nonlymphoid tissues, which might account for the disappearance of adoptively transferred tumor-specific T cells (8, 23, 49). Thus, together with the previous reports, our model indicates that tumor-specific T cell responses do develop, but might be limited by the absence of renewable central memory T lymphocytes.
The lack of central memory LACK-specific CD4 T cells was particular to the presence of TS/A-LACK tumors and was not an intrinsic characteristic of these mice. Indeed, LACK-specific CD4 T cells bearing a central memory phenotype (IL-2+CD44highCD62Lhigh lymphocytes) were found within the peripheral lymphoid tissues of mice immunized with LACK-pulsed DC and in mice in which the tumor was surgically resected. What could limit the systemic accumulation of central memory CD4 lymphocytes in tumor-bearing mice? Several parameters, such as the nature of the APC, the presence of inflammatory cytokines, and the long term persistence of Ag, are known to shape the fate of CD4 T cells (reviewed in Ref. 50). For instance, it is possible that a tumor Ag is presented by endogenous immature DC under steady-state conditions, which, as in the case of self Ag, may favor clonal deletion of the cells (51, 52, 53). Alternatively, tumor-Ag cross-presentation might occur in the absence of the proinflammatory cytokines required for the generation or survival of the memory T cell (50). Finally, it is possible that the tumor itself, by providing a continuous source of Ag, may favor terminal differentiation of tumor-specific lymphocytes, leading to the consumption of central memory cells. Our finding that the resection of the tumor allowed the appearance of CD44highCD62LhighIL-2-producing LACK-specific T cells in distal LN strongly supports previous work indicating that Ag withdrawal is required for the development and survival of central T cell memory (54, 55, 56, 57). Whether these central memory lymphocytes are derived from previously generated effector T cells, or are newly primed shortly after tumor removal remains to be determined.
The absence of central memory CD4 T lymphocytes hampers the development of protective antiviral immunity (13, 14, 15) and viral clearance (58). Conversely, the presence of central memory CD4 T lymphocytes appears to be sufficient to confer resistance to infection. In a recent report Zaph et al. (44) have recently shown that the transfer of central memory CD4 T cells, derived from L. major-immune mice was sufficient to confer protection against Leishmania infection. In a parallel study we found that vaccination with LACK-loaded DC, which induces a central memory CD4 T cell response (also shown in this study), conferred protection against TS/A-LACK tumor challenge (J. S. Zimmermann et al., manuscript in preparation). Similarly, the resection of established TS/A-LACK tumors, which allows the accumulation of central memory CD4 T cells (shown in this study) conferred protection against a second tumor challenge (not shown). This is in contrast with nonvaccinated TS/A-LACK tumor-bearing mice that are not protected against a second tumor challenge (our unpublished observations). Together, these findings suggest that tumor-specific central memory CD4 T cells might be needed to generate optimal numbers of tumor-specific effector T cells. In the absence of this renewable population, natural antitumor responses might be limited. Further elucidation of the mechanisms hindering the systemic accumulation of tumor-specific T cell central memory in the presence of a growing tumor might be critical for the improvement of current immunotherapeutic strategies.
Acknowledgments
We thank Drs. M. Bellone, and R. Zamoyska for critical reading of the manuscript and all the members of the Cancer Immunotherapy and Gene Therapy Program of DIBIT/San Raffaele Scientific Institute for support and suggestions. We are especially in debt to Dr. N. Taylor for taking the time to discuss all the findings and for extensive editing of the manuscript.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by grants from the Associazione Italiana Ricerca sul Cancro and Compagnia di San Paolo (to A.M.), and a fellowship from the Ligue Nationale Contre le Cancer (to S.H.).
2 Current address: BioXell S.p.A., Via Olgettina 60, 20132 Milan, Italy.
3 F.B. and V.S.Z. contributed equally to this work.
4 Current address: Institut National de la Santé et de la Recherche Médicale, Unité 365, Immunité et Cancer, Institut Curie, 26 rue d’Ulm, F-75245 Paris Cedex 05, France.
5 Current address: Institute for Research in Biomedicine, Via Vela 6, 6500 Bellinzona, Switzerland.
6 Address correspondence and reprint requests to Dr. Anna Mondino, Cancer Immunotherapy and Gene Therapy Program, S. Raffaele Scientific Institute, Via Olgettina 58, 20132 Milan, Italy. E-mail address: anna.mondino@hsr.it
7 Abbreviations used in this paper: Tg, transgenic; DC, dendritic cell; LACK, Leishmania receptor for activated C kinase; LN, lymph node.
Received for publication October 19, 2004. Accepted for publication May 5, 2005.
References
Wick, M., P. Dubey, H. Koeppen, C. T. Siegel, P. E. Fields, L. Chen, J. A. Bluestone, H. Schreiber. 1997. Antigenic cancer cells grow progressively in immune hosts without evidence for T cell exhaustion or systemic anergy. J. Exp. Med. 186: 229-238.
Kedl, R. M., M. F. Mescher. 1997. Migration and activation of antigen-specific CD8+ T cells upon in vivo stimulation with allogeneic tumor. J. Immunol. 159: 650-663.
Shrikant, P., M. F. Mescher. 1999. Control of syngeneic tumor growth by activation of CD8+ T cells: efficacy is limited by migration away from the site and induction of nonresponsiveness. J. Immunol. 162: 2858-2866.
Marzo, A. L., R. A. Lake, B. W. Robinson, B. Scott. 1999. T-cell receptor transgenic analysis of tumor-specific CD8 and CD4 responses in the eradication of solid tumors. Cancer Res. 59: 1071-1079.
Cordaro, T. A., K. E. de Visser, F. H. Tirion, Y. M. Graus, J. B. Haanen, D. Kioussis, A. M. Kruisbeek. 2000. Tumor size at the time of adoptive transfer determines whether tumor rejection occurs. Eur. J. Immunol. 30: 1297-1307.
Ochsenbein, A. F., S. Sierro, B. Odermatt, M. Pericin, U. Karrer, J. Hermans, S. Hemmi, H. Hengartner, R. M. Zinkernagel. 2001. Roles of tumour localization, second signals and cross priming in cytotoxic T-cell induction. Nature 411: 1058-1064(Fabio Benigni2,3,*, Valér)
Technical difficulties in tracking endogenous CD4 T lymphocytes have limited the characterization of tumor-specific CD4 T cell responses. Using fluorescent MHC class II/peptide multimers, we defined the fate of endogenous Leishmania receptor for activated C kinase (LACK)-specific CD4 T cells in mice bearing LACK-expressing TS/A tumors. LACK-specific CD44highCD62Llow CD4 T cells accumulated in the draining lymph nodes and had characteristics of effector cells, secreting IL-2 and IFN- upon Ag restimulation. Increased frequencies of CD44highCD62Llow LACK-experienced cells were also detected in the spleen, lung, liver, and tumor itself, but not in nondraining lymph nodes, where the cells maintained a naive phenotype. The absence of systemic redistribution of LACK-specific memory T cells correlated with the presence of tumor. Indeed, LACK-specific CD4 T cells with central memory features (IL-2+IFN-–CD44highCD62Lhigh cells) accumulated in all peripheral lymph nodes of mice immunized with LACK-pulsed dendritic cells and after tumor resection. Together, our data demonstrate that although tumor-specific CD4 effector T cells producing IFN- are continuously generated in the presence of tumor, central memory CD4 T cells accumulate only after tumor resection. Thus, the continuous stimulation of tumor-specific CD4 T cells in tumor-bearing mice appears to hinder the systemic accumulation of central memory CD4 T lymphocytes.
Introduction
Protective immunity requires both MHC class I-restricted cytotoxic CD8 T cells and MHC class II-restricted CD4 Th cells. Although tumor-specific CD8 T cell responses have been extensively studied in animal models of tumor disease (1, 2, 3, 4, 5, 6, 7, 8) and to some extent in cancer patients (9, 10, 11, 12), more limited information is available for CD4 T cells. This is surprising given their critical role in the priming, differentiation, and persistence of CTL (13, 14, 15) and during the effector phase of an antitumor immune response (16, 17, 18, 19, 20, 21).
The paucity of available information is possibly explained by the fact that only a limited number of MHC class II-restricted epitopes have been identified to date (22) and by the inability to trace the low number of CD4 T cells specific for a given Ag. To overcome these limitations, tumor-specific CD4 T cell responses have been studied in adoptive transfer models, where traceable populations of TCR transgenic (Tg)7 T cells specific for a model tumor-associated Ag can be visualized after tumor cell injection (4, 8, 23, 24, 25, 26). In some cases, tumor-specific CD4 T cell proliferated as a result of tumor growth, but soon afterward became unresponsive (23, 26). In other cases, CD4 T cells differentiated into effector cells, but rapidly disappeared from the host (8). Although these studies have provided some insights into the mechanisms that regulate tumor-specific CD4 T cell responses, one of their limitations was that the tumor-specific T cells were derived from monoclonal populations of cells expressing the same TCR and were therefore not representative of a polyclonal response. Furthermore, homeostatic mechanisms in mice adoptively transferred with a limited number of tumor-specific T cells are likely to be different from those occurring in unmanipulated immunocompetent animals in which tumor-specific T cells are continuously generated from a renewable population of thymic progenitors. Finally, the adoptive transfer of defined numbers of T cells does not allow an analysis of the impact of continuously primed new lymphocytes on the development of T cell memory.
Recently, the use of MHC class II multimers has allowed endogenous Ag-specific CD4 T cells to be studied in patients with rheumatoid arthritis and in Ag-vaccinated or infected mice (27, 28, 29, 30, 31). Schepers et al. (31) exploited this technology to study CD4 T cells in mice bearing Moloney murine sarcoma and leukemia virus complex-induced tumors. Even though this study underlined the importance of CD4 T cells in promoting protection against virus-infected tumor cells, its major limitation was that the T cell response was initiated by the viral infection, not by the tumor itself. Moreover, in this model the tumor was rejected as a consequence of the antiviral response, and as such, the impact of tumor growth on the development of the CD4 T cell response could not be assessed.
In this study we have used peptide/MHC class II fluorescent multimers (32) to characterize endogenous CD4 T cells specific for a tumor-restricted Ag. This is the first demonstration of the temporal and spatial organization of an antitumor CD4 T cell response during tumor progression. Moreover, we have characterized for the first time the fate of tumor-specific lymphocytes after resection of an established tumor and investigated the establishment of tumor-specific CD4 T cell memory.
Materials and Methods
Mice and tumor cells
16.2 Tg (BALB/c background) mice were previously described (32, 33). TS/A mouse mammary adenocarcinoma cells (34) were grown in vitro at 37°C in complete medium (RPMI 1640-10% FBS, 100 U/ml penicillin, 100 U/ml streptomycin, and 2.5 x 10–5 M 2-ME; Invitrogen Life Technologies). TS/A-LACK cells were obtained by transfecting TS/A cells with a pcDNA3-derived vector in which a truncated form of the LACK cDNA (35) was cloned at the HindIII site downstream from the CMV promoter. G418-resistant cells were selected, and single clones were obtained by limiting dilution. Exponentially growing 3 x 105 TS/A and TS/A-LACK cells were resuspended in 100 μl of PBS and injected s.c. into the left flank. The axillary, brachial, and inguinal lymph nodes (LN) draining the site of tumor cell injection were referred as to draining, and the axillary, brachial, and inguinal LN derived from the contralateral flank were referred as to nondraining. The growth of the tumors was monitored every other day by measuring tumor diameters using a metric caliper. Tumor diameters were measured in two dimensions, and their volume was calculated according to the following formula: volume (mm3) = (width, mm)2 x (length, mm) x 1/2. In selected experiments, mice were treated with 150 μg of anti-CD4 mAbs 4 and 2 days before the injection of TS/A-LACK cells, then once a week until the end of the experiment. Where indicated, mice were challenged with 2 x 105 bone marrow-derived DC, which were matured for 8 h in the presence of LPS (36) and loaded with the LACK immunodominant CD4 peptide (FSPSLEHPIVVSGSWD). In some experiments mice were anesthetized and either sham operated or resected of the tumor mass, according to standard procedures. All in vivo studies were approved by the ethical committee of San Raffaele Scientific Institute and Centre National de la Recherche Scientifique and performed according to their guidelines.
Western blot analysis
TS/A and TS/A-LACK cells were lysed in PBS by five consecutive freeze and thaw cycles. The soluble fraction was collected after 15 min of centrifugation at 15,000 rpm, and the protein content was determined by the Bio-Rad protein assay. Equal amounts of proteins were then separated on 12% SDS-PAGE and transferred to nitrocellulose filters. The filters were incubated with the p19–11 anti-LACK mAb (37) and detected by chemiluminescence (ECL; Amersham Biosciences).
Results
TS/A-LACK tumors develop in 16.2 mice despite the presence of LACK-specific CD4 T cells
Tumor cells expressing LACK were obtained by transfecting TS/A mammary adenocarcinoma cells with a plasmid carrying a truncated form of the LACK cDNA under control of the CMV promoter. Stable transfectants were selected and analyzed for LACK expression by Western blot using anti-LACK mAb (Fig. 1A). Of the clones expressing LACK, one was selected for additional experiments (TS/A-LACK 13.3, referred to hereafter as TS/A-LACK). Compared with parental TS/A cells, TS/A-LACK cells expressed similar surface levels of MHC class I, did not express MHC class II molecules, and proliferated to a similar extent in vitro (data not shown). Moreover, as previously reported for TS/A tumor cells (34), injection of TS/A-LACK tumor cells in BALB/c mice resulted in the development of solid tumors (not shown). We then assessed whether TS/A-LACK cells would cause tumors in mice harboring LACK-specific CD4 T cells. TS/A-LACK cells (3 x 105) were injected into 16.2 mice, which express a TCR Tg -chain derived from a LACK-specific hybridoma. As such, these mice exhibit an increased frequency of LACK-specific CD4 T cells (32). Despite the presence of LACK-specific CD4 T cells, TS/A-LACK cells produced solid tumors in all injected animals (Fig. 1B). Tumor growth was not due to the loss of LACK expression in vivo, as revealed by Western blot analyses performed on lysates obtained from resected tumors (data not shown).
To investigate whether the presence of CD4 T cells impacted the growth of TS/A-LACK tumors, tumor development was compared in untreated and anti-CD4-treated 16.2 mice. TS/A-LACK tumors grew significantly faster and achieved larger sizes after treatment of 16.2 mice with an anti-CD4 mAb (p < 0.01; Fig. 1C). These data indicate that TS/A-LACK tumors develop in the presence of an immune response. Thus, this model provides an ideal setting in which a tumor-specific CD4 T cell response can be characterized over the course of tumor progression.
Biodistribution of activated, tumor-specific CD4 T cells
To formally evaluate the status of LACK-specific CD4 T cells in the presence of tumor, we assessed whether a TS/A-LACK tumor induced the activation and expansion of these CD4 T cells. TS/A-LACK or TS/A tumor cells (3 x 105) were injected into the left flank of 16.2 mice. The axillary, brachial, and inguinal LN from the same flank where tumor cells were injected (draining) and from the contralateral flank (nondraining) were recovered at different times after injection of tumor cells. Both the frequency and number of LACK-specific CD4 cells, evaluated using I-Ad/LACK multimers, rapidly increased in the draining LN shortly after injection of TS/A-LACK cells (Fig. 2, A and B). LACK-specific CD4 T cells were activated, because by day 4 after injection the majority of I-Ad/LACK+CD4+ cells in tumor draining LN exhibited a blast phenotype (Fig. 2C), whereas I-Ad/LACK+CD4+ cells in TS/A-challenged (data not shown) or naive mice (Fig. 2D) had a small resting phenotype. Furthermore, I-Ad/LACK+CD4+ cells in the draining LN of mice injected with TS/A-LACK, but not TS/A tumors, entered into the cell cycle and up-regulated CD69 and CD44 activation markers (Fig. 2E and data not shown). In addition to the draining LN, the proportion of I-Ad/LACK+ cells increased in the spleen, where it was 1.1 ± 0.2% on day 0, 2.5 ± 0.3% on day 3, and 1.9 ± 0.3% on day 21 (data not shown). In contrast, the frequency and total number of I-Ad/LACK+ cells in the nondraining LN of TS/A-LACK tumor-bearing mice remained comparable to those in LN of naive and TS/A-bearing mice (Fig. 2, A, B, and D).
To more precisely study the relationship between the distribution and the activation of LACK-specific CD4 T cells, their phenotype in lymphoid and nonlymphoid tissues (blood, lung, liver, and tumor) was assessed 21 days after challenge with TS/A or TS/A-LACK tumor cells (Fig. 3). The presence of tumor per se did not modulate LACK-specific CD4 T cells, because the frequency and surface phenotype of I-Ad/LACK+CD4+ T cells within the lymphoid and nonlymphoid tissues of TS/A tumor-bearing mice were comparable to those in naive mice. In marked contrast, there were increased frequencies of I-Ad/LACK+CD4+ T cells expressing high levels of CD44 in the tumor-draining LN, spleen, blood, lung, liver, and tumor derived from TS/A-LACK-challenged mice. However, this increase was not observed in the nondraining LN, where I-Ad/LACK+CD4+ T cells preserved a naive phenotype (CD44lowCD62Lhigh; Fig. 3). This was despite the development of large tumors (>300 mm3).
The failure of Ag-experienced LACK-specific T cells to migrate to the nondraining LN correlated with the loss of the LN-homing molecule, CD62L, from the surface of activated T cells. Indeed, most CD44highI-Ad/LACK+CD4+ T cells found within the tumor-draining LN and spleen (and nonlymphoid tissues) expressed low levels of CD62L (Fig. 3B). Thus, it appears that the activation status of LACK-specific T cells determines their localization.
The experiments presented above were performed in 16.2 mice, because of their sizeable frequency of I-Ad/LACK+CD4+ T cells, which allows these cells to be easily followed within the global T cell population (32). Nevertheless, it was possible that the transgenic TCR -chain somehow skewed the distribution of the cells in tumor-bearing animals. To formally exclude this possibility, similar experiments were performed in mice with a normal lymphocyte repertoire (BALB/c). Because of the extremely low frequency of LACK-specific CD4 T cells (30), we specifically looked for the appearance of CD44high cells within the LACK-specific T cell population in tumor-bearing mice. Importantly, a significant percentage of I-Ad/LACK+CD4+ T cells in the tumor-draining LN, liver, and tumor of TS/A-LACK-tumor-bearing mice expressed high levels of CD44, whereas this was not observed in TS/A tumor-bearing mice (Fig. 4). However, as in 16.2 mice, there was no recirculation of memory LACK-specific CD4 T cells in the nondraining LN of BALB/c bearing TS/A-LACK tumors (Fig. 4). Thus, the accumulation of LACK-specific effector T cells in the tumor-draining LN and nonlymphoid tissues, but not in the nondraining LN appeared to be a property of tumor-bearing mice.
TS/A-LACK tumors induce the differentiation of IFN--secreting, LACK-specific CD4 T cells in draining LN
The cytokine secretion profiles of lymphocytes derived from the lymphoid and nonlymphoid organs of naive mice as well as TS/A-LACK and TS/A tumor-bearing mice were assessed. Lymphocytes were stimulated in vitro with the LACK peptide, and IL-2, IFN-, and IL-4 secretion was measured 24 h later by ELISA. Compared with T cells from naive and TS/A tumor-bearing mice, those from TS/A-LACK tumor-bearing animals secreted higher amounts of IL-2 and IFN- in response to the LACK peptide in a dose-dependent manner (Fig. 5). On a per cell basis, cells derived from the draining lymph node (Fig. 5A), the nonlymphoid tissues (Fig. 5B) and the TS/A-LACK tumor itself (Fig. 5C) produced comparable amounts of these cytokines. In contrast with Leishmania major infection (30, 32), IL-4 levels remained below the level of detection in all samples throughout tumor growth (data not shown). However, LACK-specific T cells in nondraining LN from mice bearing TS/A-LACK tumor did not secrete cytokines at levels above those detected in T cells isolated from naive mice (Fig. 5A). Together, these data indicate that TS/A-LACK tumor induced the proliferation and differentiation of LACK-specific CD4 T cells into IFN--producing effector T cells in tumor-draining LN. These effector cells were also found in the spleen, lung, liver, and tumor itself, but were not detected in nondraining LN.
I-Ad/LACK+ central memory T cells accumulate after DC immunization or tumor resection
In the experiments presented above, it appeared that TS/A-LACK tumor-bearing mice presented LACK-specific T cells with a CD44highCD62Llow effector phenotype, capable of secreting IL-2 and IFN-. In contrast, central memory T lymphocytes, characterized as CD44highCD62Lhigh, secreting IL-2 alone (40, 41, 42, 43, 44), and able to circulate within all peripheral LN, failed to accumulate in these mice. It was therefore important to determine whether the absence of LACK-specific central memory T cells was a characteristic of these transgenic mice or, alternatively, was due to the presence of tumor. To this aim, we compared the phenotype of LACK-experienced T cells in 16.2 mice harboring TS/A-LACK tumors and in 16.2 mice immunized with LACK-loaded, bone marrow-derived DC, which are known to induce potent T cell responses (45). TS/A-LACK tumors and LACK-loaded DC were injected into one flank of individual mice. Fifteen days later, the draining and nondraining LN were recovered and analyzed by flow cytometry. As expected, both Ag challenges resulted in an increased frequency of I-Ad/LACK+ CD4 T cells in the draining LN compared with naive mice (1.77 and 1.24 vs 0.6%, respectively; Fig. 6). However, the phenotypes of these cells were significantly different; >80% of CD44high LACK-specific cells in tumor-bearing mice had down-regulated expression of CD62L, whereas the majority of CD44high cells in DC-immunized mice preserved the expression of the CD62L homing molecule (Fig. 6, A and C). Moreover, the cytokine secretion profile of these cells were distinct; in tumor-bearing mice, 50% of cytokine-secreting cells expressed both IL-2 and IFN-, although >90% of cytokine-secreting cells in DC-immunized mice expressed IL-2 alone (Fig. 6, B and C). Furthermore, DC activation resulted in the accumulation of CD44high IL-2-secreting I-Ad/LACK cells in nondraining LN, although the T cell tumor response was restricted to draining LN. The ensemble of these data indicate that 16.2 mice are capable of mounting a LACK-specific central memory CD4 T cell response, but that this response is dependent upon the manner in which the Ag is presented. Upon presentation of LACK Ag in the context of a tumor, there is a strong effector T cell response limited to draining LN. In contrast, when the LACK Ag is presented by LPS-matured DC, the majority of the response is a central memory response, and these LACK-experienced T cells accumulate in all peripheral LN.
To determine whether the presence of the tumor itself modulates the nature of the T cell response, i.e., effector vs central memory, established TS/A-LACK tumors were surgically removed 15 days after tumor cell injection in 16.2 mice. As expected, the percentages of I-Ad/LACK+ T cells detected in the draining LN and spleens of tumor-resected mice decreased by >50% within 4 days after surgery (3.4 to 1.2% in the draining LN and 1.8 to 0.9% in the spleen of a representative mouse; Fig. 7A). The frequency of I-Ad/LACK+ T cells exhibiting a blast phenotype (Fig. 7B) and expressing high levels of CD44 (Fig. 7C) also decreased, and LACK-specific T cells did not secrete detectable levels of IFN- by day 4 (Fig. 7C). Importantly though, the percentages of CD44highI-Ad/LACK+ cells increased significantly in the nondraining LN of tumor-resected mice within 10 days after surgery (Fig. 8). This increase was specific to the tumor resection, because it was not detected in sham-operated mice. Moreover, in marked contrast with the phenotype in tumor-bearing mice, the vast majority of CD44high cells in tumor-resected mice expressed CD62L (80% in a representative mouse; Fig. 8, B and C) and produced IL-2 (Fig. 8C).
Together, these data indicate that resection of established TS/A-LACK tumors causes the rapid disappearance of IFN--producing effector T cells from the tumor-draining LN and spleen, with a concomitant accumulation of IL-2-producing central memory lymphocytes in the nondraining LN.
Discussion
This is the first study assessing the temporal and spatial organization of an endogenous tumor-specific CD4 T cell response during the course of tumor development, with evaluation of the establishment of tumor-specific T cell memory.
16.2 Tg mice are an important model in which to monitor endogenous T cell responses, because they exhibit an increased frequency of CD4 T cells directed against a specific Ag (LACK) within a context of fairly normal T cells repertoire. Using peptide-MHC class II fluorescent multimers against LACK, we were able to define the site of initial Ag encounter and track the fate of Ag-experienced T cells within both lymphoid and nonlymphoid tissues of tumor-bearing mice. Tumor-specific CD4 T cells appeared to be exclusively primed in the tumor-draining LN, but not in other lymphoid organs. Thus, in TS/A-LACK tumor-bearing mice, only the tumor-draining LN, and not nondraining LN or spleen, contained an increased frequency of I-Ad/LACK+ cells expressing the early activation marker CD69 (data not shown). Furthermore, although a sizeable number of I-Ad/LACK+ cells exhibited a blast phenotype in the draining LN, the vast majority of these cells in the spleen and nondraining LN remained small even after the development of large tumors. Finally, only the draining LN of TS/A-LACK tumor-bearing mice contained DC, which were able to stimulate LACK-specific hybridoma in vitro in the absence of added LACK peptide (S. Hugues and N. Gleichenhaus, unpublished observations). Thus, together with the results reported by Marzo et al. (4, 46, 47), our data argue that the LN most proximal to the tumors are the main lymphoid tissues reacting to the growing tumor.
LACK encounter in the tumor-draining LN induced I-Ad/LACK+ T lymphocytes to proliferate and differentiate. I-Ad/LACK+ T lymphocytes up-regulated CD44 surface expression, down-regulated the LN-homing molecule CD62L, and were capable of producing both IL-2 and IFN- when restimulated in vitro. Ag-experienced LACK-specific CD4 T cells were also found in the spleen as well as in nonlymphoid tissues, such as lung and liver, and the tumor mass itself, where they retained the ability to produce IL-2 and IFN- upon Ag re-encounter. Thus, in contrast with other tumor models, no evidence for CD4 T cell anergy was found in TS/A-LACK tumor-bearing mice. The finding that depletion of CD4 T cells resulted in faster tumor development also indicates that protective CD4 T cell responses naturally developed in these mice.
Recent studies performed in humans and in immunized and pathogen-infected mice have shown that Ag-experienced T cells are heterogeneous in nature, with subpopulations of effector memory and central memory cells. Although effector memory T cells migrate to tissues and secrete effector cytokines, central memory T cells migrate through the LN and lack immediate effector function (40, 41, 42, 43, 44). In mice, central memory T cells are best defined by their localization in peripheral LN, surface expression of high levels of CD44 and CD62L, and the ability to produce IL-2, but not other cytokines (40, 41, 42, 43, 44). In this study we show that CD44highCD62LlowI-Ad/LACK+ effector T cells were always found in TS/A-LACK-tumor-bearing mice. In contrast, CD44highCD62LhighI-Ad/LACK+ central memory T cells failed to accumulate in tumor-draining LN and were never found in LN distal to the site of tumor growth. Thus, cells in these latter locations appeared to remain ignorant of tumor growth.
These findings might help to reconcile apparently conflicting published results. Indeed, some authors reported the presence of tumor-specific T lymphocytes capable of IFN- production in the blood and lymphoid organs of tumor-bearing mice (1, 7, 24, 46, 48). In contrast, other reports in which TCR Tg CD4 T cells were adoptively transferred into tumor-bearing mice showed that the cells appeared to loose the ability to proliferate, secrete effector cytokines in vitro, and undergo deletion (8, 23, 49). We found that tumor-specific CD4 effector cells capable of IFN- production were continuously generated in tumor-draining LN. However, most of the Ag-experienced T cells rapidly lost CD62L expression, and exited the peripheral LN to relocate in nonlymphoid tissues, which might account for the disappearance of adoptively transferred tumor-specific T cells (8, 23, 49). Thus, together with the previous reports, our model indicates that tumor-specific T cell responses do develop, but might be limited by the absence of renewable central memory T lymphocytes.
The lack of central memory LACK-specific CD4 T cells was particular to the presence of TS/A-LACK tumors and was not an intrinsic characteristic of these mice. Indeed, LACK-specific CD4 T cells bearing a central memory phenotype (IL-2+CD44highCD62Lhigh lymphocytes) were found within the peripheral lymphoid tissues of mice immunized with LACK-pulsed DC and in mice in which the tumor was surgically resected. What could limit the systemic accumulation of central memory CD4 lymphocytes in tumor-bearing mice? Several parameters, such as the nature of the APC, the presence of inflammatory cytokines, and the long term persistence of Ag, are known to shape the fate of CD4 T cells (reviewed in Ref. 50). For instance, it is possible that a tumor Ag is presented by endogenous immature DC under steady-state conditions, which, as in the case of self Ag, may favor clonal deletion of the cells (51, 52, 53). Alternatively, tumor-Ag cross-presentation might occur in the absence of the proinflammatory cytokines required for the generation or survival of the memory T cell (50). Finally, it is possible that the tumor itself, by providing a continuous source of Ag, may favor terminal differentiation of tumor-specific lymphocytes, leading to the consumption of central memory cells. Our finding that the resection of the tumor allowed the appearance of CD44highCD62LhighIL-2-producing LACK-specific T cells in distal LN strongly supports previous work indicating that Ag withdrawal is required for the development and survival of central T cell memory (54, 55, 56, 57). Whether these central memory lymphocytes are derived from previously generated effector T cells, or are newly primed shortly after tumor removal remains to be determined.
The absence of central memory CD4 T lymphocytes hampers the development of protective antiviral immunity (13, 14, 15) and viral clearance (58). Conversely, the presence of central memory CD4 T lymphocytes appears to be sufficient to confer resistance to infection. In a recent report Zaph et al. (44) have recently shown that the transfer of central memory CD4 T cells, derived from L. major-immune mice was sufficient to confer protection against Leishmania infection. In a parallel study we found that vaccination with LACK-loaded DC, which induces a central memory CD4 T cell response (also shown in this study), conferred protection against TS/A-LACK tumor challenge (J. S. Zimmermann et al., manuscript in preparation). Similarly, the resection of established TS/A-LACK tumors, which allows the accumulation of central memory CD4 T cells (shown in this study) conferred protection against a second tumor challenge (not shown). This is in contrast with nonvaccinated TS/A-LACK tumor-bearing mice that are not protected against a second tumor challenge (our unpublished observations). Together, these findings suggest that tumor-specific central memory CD4 T cells might be needed to generate optimal numbers of tumor-specific effector T cells. In the absence of this renewable population, natural antitumor responses might be limited. Further elucidation of the mechanisms hindering the systemic accumulation of tumor-specific T cell central memory in the presence of a growing tumor might be critical for the improvement of current immunotherapeutic strategies.
Acknowledgments
We thank Drs. M. Bellone, and R. Zamoyska for critical reading of the manuscript and all the members of the Cancer Immunotherapy and Gene Therapy Program of DIBIT/San Raffaele Scientific Institute for support and suggestions. We are especially in debt to Dr. N. Taylor for taking the time to discuss all the findings and for extensive editing of the manuscript.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by grants from the Associazione Italiana Ricerca sul Cancro and Compagnia di San Paolo (to A.M.), and a fellowship from the Ligue Nationale Contre le Cancer (to S.H.).
2 Current address: BioXell S.p.A., Via Olgettina 60, 20132 Milan, Italy.
3 F.B. and V.S.Z. contributed equally to this work.
4 Current address: Institut National de la Santé et de la Recherche Médicale, Unité 365, Immunité et Cancer, Institut Curie, 26 rue d’Ulm, F-75245 Paris Cedex 05, France.
5 Current address: Institute for Research in Biomedicine, Via Vela 6, 6500 Bellinzona, Switzerland.
6 Address correspondence and reprint requests to Dr. Anna Mondino, Cancer Immunotherapy and Gene Therapy Program, S. Raffaele Scientific Institute, Via Olgettina 58, 20132 Milan, Italy. E-mail address: anna.mondino@hsr.it
7 Abbreviations used in this paper: Tg, transgenic; DC, dendritic cell; LACK, Leishmania receptor for activated C kinase; LN, lymph node.
Received for publication October 19, 2004. Accepted for publication May 5, 2005.
References
Wick, M., P. Dubey, H. Koeppen, C. T. Siegel, P. E. Fields, L. Chen, J. A. Bluestone, H. Schreiber. 1997. Antigenic cancer cells grow progressively in immune hosts without evidence for T cell exhaustion or systemic anergy. J. Exp. Med. 186: 229-238.
Kedl, R. M., M. F. Mescher. 1997. Migration and activation of antigen-specific CD8+ T cells upon in vivo stimulation with allogeneic tumor. J. Immunol. 159: 650-663.
Shrikant, P., M. F. Mescher. 1999. Control of syngeneic tumor growth by activation of CD8+ T cells: efficacy is limited by migration away from the site and induction of nonresponsiveness. J. Immunol. 162: 2858-2866.
Marzo, A. L., R. A. Lake, B. W. Robinson, B. Scott. 1999. T-cell receptor transgenic analysis of tumor-specific CD8 and CD4 responses in the eradication of solid tumors. Cancer Res. 59: 1071-1079.
Cordaro, T. A., K. E. de Visser, F. H. Tirion, Y. M. Graus, J. B. Haanen, D. Kioussis, A. M. Kruisbeek. 2000. Tumor size at the time of adoptive transfer determines whether tumor rejection occurs. Eur. J. Immunol. 30: 1297-1307.
Ochsenbein, A. F., S. Sierro, B. Odermatt, M. Pericin, U. Karrer, J. Hermans, S. Hemmi, H. Hengartner, R. M. Zinkernagel. 2001. Roles of tumour localization, second signals and cross priming in cytotoxic T-cell induction. Nature 411: 1058-1064(Fabio Benigni2,3,*, Valér)