We previously found that a Salmonella typhimurium vector engineered to secrete soluble tumor antigen induces CD4+ T cells resistant to CD4+CD25+ regulatory T cells (Treg) and that glucocorticoid-induced tumor necrosis factor receptor family-related gene (GITR) signal is involved in the development of this resistance. In this study, we address the potential of incorporating GITR ligand (GITRL) as a way to augment the immunogenicity of cancer vaccines. BALB/c mice were immunized by gene gun with plasmids encoding the mutated extracellular signal-regulated kinase 2 (mERK) with or without plasmids encoding mouse GITRL. Coadministration with GITRL during primary and secondary immunization enhanced the induction of mERK-specific CD8+ T cells. Antibody depletion and minigene analysis suggested that GITRL directly activated CTL epitope-specific CD8+ T cells independently of CD4+ T cells. Immunization with plasmids encoding a CTL epitope and GITRL resulted in strong tumor inhibition in a CD8+ T cell–dependent manner. Furthermore, CTL epitope-specific CD8+ T cells induced by immunization with plasmids encoding CTL epitope coadministered with GITRL were refractory to suppression by CD4+CD25+ Tregs compared with CD8+ T cells induced without GITR signaling. We propose that coadministration of GITR signaling agents with tumor antigens constitutes a promising novel strategy for cancer vaccine development. [Cancer Res 2008;68(14):5948–54]
- Regulatory T cells
- CD8+ T cells
- GITR ligand
- Cancer vaccine
- Mouse model
With the molecular identification of tumor antigens recognized by the human immune system, several cancer vaccine strategies targeting these antigens have been attempted ( 1– 4). However, optimal strategies for inducing strong and durable immune responses to these tumor antigens have yet to be defined ( 5). One major problem is the fact that most tumor antigens identified to date are self-antigens and by themselves may not induce strong CD4+ and CD8+ T-cell responses ( 1– 4, 6).
CD4+CD25+ regulatory T cells (Treg) play an important role in maintaining self-tolerance in hosts by suppressing a wide variety of immune responses ( 7– 9). CD4+CD25+ Treg populations, originally found to suppress autoimmune responses, are also crucial in controlling antitumor immune responses ( 7– 9). In mice, depletion of Treg populations enhances naturally induced and vaccine-induced antitumor T-cell responses ( 10– 12). In addition, induction/activation of CD4+CD25+ Tregs by immunization with self-antigens augments the number of pulmonary metastases following injection of transplantable tumor cells and enhances the development of chemically induced primary tumors ( 13– 15). In humans, some studies show that the presence of high numbers of CD4+CD25+ Tregs or low ratio of CD8+ T cells to CD4+CD25+ Tregs at the local tumor site is correlated with unfavorable prognosis ( 16, 17). Furthermore, CD4+CD25+ Tregs control the antigen-specific T-cell induction in both spontaneous and vaccine-induced T-cell responses ( 18– 20). For this reason, it is becoming an important priority to find new effective adjuvant formulations, vectors, or vaccination strategies for controlling Tregs in the cancer vaccine field.
Glucocorticoid-induced tumor necrosis factor (TNF) receptor family-related gene (GITR) is a type I transmembrane protein with homology to TNF receptor family members and was originally reported as a molecule that inhibits T-cell receptor (TCR)–induced apoptosis ( 21, 22). GITR is expressed at different levels in resting CD4+ and CD8+ T cells and is up-regulated after T-cell activation ( 21, 23, 24). GITR acts as a costimulatory molecule, particularly in the setting of suboptimal T-cell receptor stimulation ( 23– 25). GITR is also constitutively expressed on CD4+CD25+ Tregs at high levels, and it was originally thought that the inhibitory effect of GITR signaling on the suppressive activity of Tregs was due to direct action on these cells ( 26, 27). However, a recent study with GITR knockout mice revealed that the reversal of suppression by GITR signaling with agonist anti-GITR antibody is attributable to the costimulatory activity of GITR on responder CD4+CD25− T cells rather than a direct effect on CD4+CD25+ Tregs ( 22, 28).
We and others have recently shown that CD4+ T-cell precursors specific for the cancer/testis antigen NY-ESO-1 exist in relatively high frequencies in healthy individuals and cancer patients, but that their response to antigen is suppressed by CD4+CD25+ Tregs ( 18– 20). A versatile bacterial vector, Salmonella typhimurium, engineered to secrete NY-ESO-1 via the type III secretion system, can overcome this suppression and permit specific CD4+ T cells to proliferate. These CD4+ T cells were found to be resistant to suppression by Tregs, and GITR signaling was involved in generating this resistance ( 29). These data prompted us to analyze the potential of GITR signaling in cancer vaccine strategies.
In this study, we examined whether GITR signaling could elicit tumor antigen–specific CD8+ T cells that were resistant to CD4+CD25+ Treg suppression. To address this possibility, we immunized mice with plasmids encoding antigens and mouse GITR ligand (GITRL) and used a Helios Gene Gun System to ensure that both plasmids were delivered simultaneously to the same antigen-presenting cells (APC; ref. 30). Using this approach, we show that antigen-specific CD8+ T-cell induction was enhanced by antigen coadministered with GITRL and that GITR signaling acted on CD8+ T cells. In an in vivo tumor system, coadministration of GITRL enhanced tumor resistance elicited by specific antigen immunization in a CD8+ T cell–dependent manner. Finally, CD8+ T cells induced by immunization with plasmids encoding antigen coadministered with GITRL (but not with plasmids encoding antigen alone) were found to be resistant to suppression by CD4+CD25+ Tregs.
Materials and Methods
Mice. Female BALB/c mice were purchased from CLEA Japan and used at 7 to 10 weeks of age. Mice were maintained at the Animal Center of Mie University Graduate School of Medicine (Mie, Japan). The experimental protocol was approved by the Ethics Review Committee for Animal Experimentation of Mie University Graduate School of Medicine.
Tumors. CMS5 is a 3-methylcholanthrene–induced sarcoma cell line of BALB/c origin ( 30, 31). CMS5a is a subcloned cell line obtained from CMS5, a tumor expressing mutated mitogen-activated protein kinase, extracellular signal-regulated kinase 2 (mERK2; ref. 31). P1.HTR is a subline of mastocytoma P815 of DBA/2 origin ( 30). CMS5a and P1.HTR do not express GITRL (data not shown).
Immunization by gene gun. Gold particles coated with 1 μg of each plasmid DNA were prepared and delivered into shaved skin of the abdominal wall of BALB/c mice by a Helios Gene Gun System (Bio-Rad) at a helium discharge pressure of 350 to 400 p.s.i., as described ( 30).
Antibodies and reagents. Anti-CD4 (GK1.5, rat IgG2b), anti-CD8 (19/178, rat IgG2b), anti-CD25 (PC61, rat IgG1), and anti-GITR (DTA-1, rat IgG2a) monoclonal antibodies (mAb) were produced from each hybridoma and some of them were purified by protein G columns. Each mAb was injected through the tail vein as described ( 30, 32). Anti-CD3 mAb (145-2C11, hamster IgG1), Alexa Fluor 647–conjugated anti-Langerin mAb (eBioRMUL, rat IgG2a), and phycoerythrin-conjugated anti-GITRL mAb (eBioYGL386, rat IgG1) were purchased from eBioscience. FITC-conjugated anti-CD4 mAb (GK1.5, rat IgG2b), anti-CD8 mAb (53-6.7, rat IgG2a), and phycoerythrin-conjugated anti-CD25 mAb (3C7, rat IgG2b) were purchased from BD Biosciences. Synthetic mERK2136-144-9m peptide QYIHSANVL ( 31) and HER2p63-71 (T) peptide TYLPTNASL ( 33) were obtained from Qiagen. A cDNA encoding the region of mERK2 containing CTL epitope 9m was cloned into pCAGGS-New and purified using EndoFree Plasmid Maxi kit (Qiagen). pCAGGS-New was kindly provided by Dr. J. Miyazaki (Osaka University, Osaka, Japan; ref. 30). Mouse GITRL cDNA was reverse transcribed from mouse spleen cDNA library using combination of primers (5′-AAAAAGCTTATGGAGGAAATGCCTTTGA-3′ and 5′-AAGGATCCCTAAGAGATGAATGGTAGAT-3′) and cloned into pCDNA3.1 (Invitrogen) and purified using EndoFree Plasmid Maxi kit.
Flow cytometry and tetramer staining. Cells were stained for surface markers in PBS with 2% fetal bovine serum for 15 min at 4°C and analyzed on FACSCanto (BD Biosciences). Tetramer staining was performed as described ( 32). Briefly, CD8+ T cells were stained with phycoerythrin-labeled 9m-Kd tetramers (prepared at the Ludwig Institute Core Facility by Drs. P. Guillaume and I. Luescher, Lausanne, Switzerland) for 10 min at 37°C before additional staining of FITC-CD8 mAb (BD Biosciences) for 15 min at 4°C. Specificity of the tetramer was confirmed with CD8+ T cells derived from DUC18 mice, transgenic for αβ-TCR reactive with the Kd-restricted mERK2136-144 ( 34). After washing, results were analyzed on FACSCanto and FlowJo software (Tree Star).
Cell isolation. Spleen cell suspensions were mixed with anti-CD8 microbeads (Miltenyi Biotec) and separated into CD8+ T cells by positive selection on a MACS column. CD8+ T-cell populations were confirmed to contain >95% CD8+ T cells. In some experiments, these CD8+ T cells were further purified into CD8+9m-Kd tetramer+ T cells on a FACSAria (BD Biosciences) after staining with FITC-anti-CD8 mAb and phycoerythrin-labeled 9m-Kd tetramer. The purity of these CD8+9m-Kd tetramer+ T cells was >98%.
CD4+ T cells were enriched by negative selection using CD4+ isolation kit (Miltenyi Biotec) followed by purification of CD4+CD25+ T cells on a FACSAria after staining with FITC–anti-CD4 and phycoerythrin–anti-CD25. The purity of these CD4+CD25+ T cells was >99%.
Enzyme-linked immunospot assay. The number of IFN-γ–secreting peptide-specific CD8+ T cells was assessed by enzyme-linked immunospot (ELISPOT) assays as described ( 30, 32). Briefly, purified CD8+ T cells were cultured for 24 h with 1 × 105 mitomycin C–treated P1.HTR pulsed with mERK2136-144-9m or HER2p63-71 (T) peptides in 96-well nitrocellulose-coated microtiter plates (Millipore) coated with rat anti-mouse IFN-γ (R4-6A2, BD Biosciences). Spots were developed using biotinylated anti-mouse IFN-γ (XMG1.2, BD Biosciences), alkaline phosphatase–conjugated streptavidin (Mabtech), and alkaline phosphatase substrate kit (Bio-Rad) and subsequently counted.
FITC labeling. FITC (Sigma) was dissolved in 1:1 acetone/dibutylphthalate and FITC (5 mg/mL) was applied at the site of gene gun immunization. At the indicated periods after FITC application, cells from draining lymph nodes were analyzed on FACSCanto.
Tumor challenge. Groups of five mice were inoculated s.c. in the right hind flank with 0.2 mL PBS containing 1 × 106 CMS5a and monitored thrice weekly.
Proliferation assay. CD8+9m-Kd tetramer+ T cells (2 × 104) obtained from mice immunized with 9m alone or coadministered with GITRL were cultured with 5 × 104 irradiated splenic Thy-1− APCs prepared from wild-type BALB/c mice in the presence of 1 μg/mL anti-CD3 mAb in 96-well flat-bottomed plates. To these cultures, CD4+CD25+ T cells were added. Proliferation was evaluated by pulsing with 0.5 μCi/well [3H]thymidine for the last 6 h of 72-h culture. [3H]thymidine incorporation was measured by a scintillation counter.
ELISA. The concentration of IFN-γ in supernatants from cultures as described above was determined by IFN-γ ELISA kit (BD Biosciences), according to the instructions provided by the manufacturer.
Statistical analysis. Statistical significance was evaluated by Student's t test.
Antigen-specific CD8+ T-cell generation is enhanced by coadministration with GITRL. We asked whether GITR signaling enhanced the primary and/or memory responses of antigen-specific CD8+ T cells. Naive BALB/c mice were immunized twice with plasmids encoding the CMS5 tumor rejection antigen mERK2 alone or coadministered with plasmids encoding mouse GITRL. CD8+ T cells were obtained from spleens, and specific T-cell induction was analyzed by ELISPOT assays. As shown in Fig. 1A , ∼10 times greater numbers of specific T cells were induced in mice immunized with plasmids encoding mERK2 coadministered with mouse GITRL during both primary and secondary immunization or only at primary immunization. In contrast, little enhancement of specific CD8+ T-cell induction was found in mice immunized with plasmids encoding mERK2 coadministered with mouse GITRL at secondary immunization. In addition, expression of mouse GITRL on target APCs after gene gun immunization was analyzed with FITC application at the site of gene gun immunization. GITRL expression on FITChigh cells gated with Langerin (Langerhans cells) in draining lymph node derived from mice immunized with mERK2 coadministered with GITRL, but not mERK2 alone, showed increased expression of GITRL ( Fig. 1B).
CD4+ T cells are not necessary for the enhancement of antigen-specific CD8+ T-cell induction through GITR signaling. We assessed the cellular target(s) of GITR signaling for the enhancement of antigen-specific CD8+ T-cell induction. To this end, we generated minigene plasmids encoding the class I–restricted CTL epitope 9m derived from mERK2 ( 31). Using these minigene constructs, the involvement of CD4+ helper T cells by immunizing with the entire gene can be addressed. BALB/c mice were immunized with plasmids encoding a CTL epitope 9m with or without coadministration of plasmids encoding mouse GITRL. CTL epitope 9m-specific CD8+ T-cell induction was enhanced in mice immunized with plasmids encoding CTL epitope 9m coadministered with mouse GITRL during both primary and secondary immunization, suggesting that GITR signaling acts directly on CD8+ T cells but not CD4+ helper T cells ( Fig. 2A ).
To establish the lack of CD4+ T-cell involvement, we examined the effect of CD4+ T-cell depletion by antibody. As shown in Fig. 2B, CD4+ T-cell depletion had no effect on the enhanced CD8+ T-cell induction via GITR signaling. Taken together, these data indicate that GITR signaling acts directly on CD8+ T cells.
Depletion of CD25+ T cells does not further augment specific CD8+ T-cell induction. It has been reported that anti-GITR agonistic mAb inhibits the suppressive activity of CD4+CD25+ Tregs ( 26, 27). However, there is accumulating evidence to show that this antibody-mediated GITR signaling acts at the level of CD4+ and CD8+ effector T cells rather than CD4+CD25+ Tregs by providing costimulatory signals ( 22– 25, 28, 35). Our finding thus far indicates that GITR signaling acts directly on CD8+ T cells. To further analyze the effect on CD4+CD25+ Tregs, we immunized BALB/c mice with plasmids encoding CTL epitope 9m with or without coadministration of plasmids encoding mouse GITRL and tested the effects of anti-CD25 mAb (PC61) injected at the time of primary and secondary immunization. PC61 administration showed no added enhancement of CTL epitope 9m-specific CD8+ T-cell induction over that induced by CTL epitope 9m coadministered with GITRL, whereas 9m-specific T-cell induction by immunization with a CTL epitope 9m alone was markedly augmented with depletion of CD25+ cells ( Fig. 2C). This observation provides further evidence that GITR signaling acts directly on CD8+ T cells.
GITR signaling through GITRL is more effective than GITR agonist antibody for the enhancement of specific T-cell induction. It has been previously reported that anti-GITR agonistic mAb (DTA-1) enhances the specific T-cell induction in mice immunized with human gp100 ( 36). These data prompted us to assess whether GITRL was as efficient as DTA-1 in the enhancement of specific T-cell induction. BALB/c mice were immunized twice with plasmids encoding 9m alone or coadministered with plasmids encoding mouse GITRL. DTA-1 antibody was injected at the time of primary and secondary immunization. Surprisingly, GITR signaling through plasmids encoding mouse GITRL provided a more effective signal for enhancing CTL epitope 9m-specific CD8+ T-cell induction than injection of DTA-1 ( Fig. 3 ).
Coadministration of GITRL inhibits tumor growth in a CD8-dependent manner. Next, we investigated the effect of GITR signaling in the CMS5 tumor model. BALB/c mice were inoculated with 1 × 106 CMS5a cells and tumor growth was observed. Immunization with plasmids encoding CTL epitope 9m coadministered with GITRL resulted in strong inhibition of CMS5a tumor growth when vaccination was started on the same day or 2 days but not on 4 days after tumor inoculation ( Fig. 4A and B ). In contrast, immunization with plasmids encoding CTL epitope 9m alone only slowed tumor growth compared with nontreated controls ( Fig. 4A).
To gain insight into the mechanism(s) by which GITR signaling induced tumor inhibition, we investigated the effect of depleting CD8+ or CD4+ T cells on animals that received coadministered GITRL on tumor growth. Injection of anti-CD8 mAb effectively blocked the effect by coadministered GITRL ( Fig. 4C). In contrast, injection of anti-CD4 mAb did not affect the ability of coadministered GITRL to induce tumor inhibition ( Fig. 4C). These results indicate that GITR signaling enhances tumor inhibitory activity of specific immunity in a CD8-dependent manner.
Tumor antigen–specific CD8+ T cells from mice immunized with plasmids encoding 9m with GITRL have enhanced resistance to CD4+CD25+ Tregs. Recent studies provide evidence that tumors contain a large number of CD4+CD25+ Tregs and that these CD4+CD25+ Tregs down-regulate antitumor immune responses ( 16, 17). CD4+CD25+ Treg infiltration was also observed in CMS5a and depletion of CD4+CD25+ Tregs by pretreatment with anti-CD25 mAb inhibited CMS5a tumor growth. 7 Given the efficient tumor inhibition elicited by coadministration of GITRL that we observed, we asked whether GITR signaling could alter the response of effector T cells to CD4+CD25+ Tregs in this system. Based on our data that GITR signaling acted directly on CD8+ T cells, we asked whether antigen-specific CD8+ T cells in mice immunized with plasmids encoding 9m coadministered with GITRL were resistant to suppression by CD4+CD25+ Tregs. BALB/c mice were immunized twice with plasmids encoding 9m alone or coadministered with plasmids encoding mouse GITRL. The frequency of CTL epitope 9m-specific T cells as measured by CD8+9m-Kd tetramer+ T cells was four to five times higher in mice immunized with plasmids encoding 9m coadministered with GITRL compared with mice immunized with plasmids encoding 9m alone, confirming the data from ELISPOT assays ( Fig. 5A ). These CTL epitope 9m-specific T cells were purified on a FACSAria from the CD8+ T-cell population and the purity was >98% ( Fig. 5B). Splenic 2 × 104 CD4+CD25+ Tregs prepared from naive BALB/c mice were added to cultures of 2 × 104 CD8+9m-Kd tetramer+ T cells with 5 × 104 irradiated BALB/c splenic Thy-1− APCs with anti-CD3 mAb. The ratio of Tregs to effector cells for in vitro analysis was determined based on the ratio at the CMS5a tumor local site that was close to 1:1 (data not shown). Proliferation and IFN-γ secretion were assessed as described in Materials and Methods. CD8+9m-Kd tetramer+ T cells derived from mice immunized with plasmids encoding CTL epitope 9m with GITRL maintained significant proliferative and IFN-γ secretion capacity ( Fig. 5C and D). In contrast, proliferation and IFN-γ secretion of CD8+9m-Kd tetramer+ T cells derived from mice immunized with plasmids encoding CTL epitope 9m alone were completely suppressed by CD4+CD25+ Tregs ( Fig. 5C and D). Taken together, these results indicate that GITR signaling renders CD8+ T cells more resistant to suppression by CD4+CD25+ Tregs.
We have engineered plasmids encoding mouse GITRL and examined whether GITR signaling as delivered in this way could (a) enhance antigen-specific CD8+ T-cell induction, (b) heighten antigen-specific resistance to tumor challenge, and (c) augment resistance of CD8+ T cells to suppression by CD4+CD25+ Tregs. We found induction of 10 times higher numbers of specific CD8+ T cells in mice immunized with plasmids encoding the CMS5 tumor rejection antigen mERK2 coadministered with mouse GITRL during both primary and secondary immunization ( Fig. 1). In contrast to slightly reduced tumor growth in mice immunized with plasmids encoding 9m alone, strong inhibition of CMS5a tumor was found in mice immunized with plasmids encoding CTL epitope 9m coadministered with GITRL ( Fig. 4). Finally, the CD8+ T cells generated in mice immunized together with GITRL exhibited increased resistance to CD4+CD25+ Tregs ( Fig. 5).
GITR signaling through plasmids encoding mouse GITRL provided a more effective signal than GITR agonistic mAb for enhancing CTL epitope 9m-specific CD8+ T-cell induction ( Fig. 3). This may be due to the usage of the Helios Gene Gun System, which is able to deliver gold particles coated with plasmids encoding both CTL epitope and mouse GITRL to the same APCs ( 30), thus providing a more focused GITR signal than systemic injection of mAb. The effect of systemic treatment by antibody will depend on the dose, timing, and route of antibody injection, and although the protocol used in our study was similar to that of in previous studies ( 36, 37), it may not have been optimal. Systemic injection of high amounts of GITR agonistic antibodies may break the tolerized state for self-antigens and induce undesirable autoimmune diseases ( 36, 38). This unwanted outcome of GITR signal may be avoided by the use of gene gun that delivers the GITR signal locally to APCs.
To gain insight into the cellular target(s) of GITRL, minigene plasmids encoding the class I–restricted CTL epitope 9m derived from mERK2 were generated as a way to explore the contribution of CD4+ helper T cells. In addition, antibody-mediated depletion of CD4+ T cells was carried out to further examine the role of CD4+ T cells. GITRL-mediated enhancement of specific CD8+ T-cell induction was not reduced in mice injected with anti-CD4 mAb ( Fig. 2), indicating that GITRL presumably acts directly on CD8+ T cells. Similarly, increased resistance to tumor growth by coadministration of GITRL was not affected by anti-CD4 mAb but was completely blocked by anti-CD8 mAb ( Fig. 4C). In addition, administration of anti-CD25 mAb did not augment the enhanced CTL epitope 9m-specific CD8+ T-cell induction in mice immunized with CTL epitope 9m coadministered with mouse GITRL, whereas 9m-specific T-cell induction by immunization with a CTL epitope 9m alone was markedly enhanced by depletion of CD25+ cells ( Fig. 2C), once again suggesting that GITRL can act directly on antigen-specific CD8+ T cells but not CD4+ T cells, particularly CD4+CD25+ Tregs, and that these T cells are resistant against Treg suppression. These data are in line with the reports showing that GITR signal directly enhances the activity of CD8+ T cells ( 39, 40).
It has been shown that in mice following xenogeneic DNA immunization with gp100, the enhanced specific T-cell induction caused by GITR agonist mAb (DTA-1) was reduced by depletion of CD4+ T cells ( 36) and that in vivo administration of DTA-1 overcame tolerance and induced tumor rejection by rendering CD4+ T cells refractory to suppression by Tregs ( 41). This may reflect differences in the way that GITR signal is delivered (e.g., coadministration of GITRL with antigen versus systemic antibody administration). Because GITR signaling provides costimulation to responder T cells, particularly when a suboptimal TCR signal is provided ( 22, 24, 36), it may also have to do with the differing effects of GITR signaling on immunization with xenogeneic antigen, such as gp100, versus nonxenogeneic antigen, such as mERK2 [e.g., immunization with multiple antigen epitopes (xenogeneic antigen) or few antigenic epitopes (nonxenogeneic antigens); refs. 31, 36]. This may also explain why we observed increased CD8+ T-cell induction in mice receiving GITRL during primary and secondary immunization, whereas the enhancing effects of GITR agonist mAb on xenogeneic DNA immunization were only seen when the antibody was injected at the time of the second immunization; the antibody had no effect on primary immunization and it caused activation-induced cell death when given at the time of both primary and secondary immunization ( 36). To pursue this difference between the effect of GITRL and GITR agonistic antibody, we examined the effect of GITR signaling on xenogeneic immunization with human HER2 ( 33) using GITRL coadministration. GITRL signaling at the time of secondary immunization with HER2 enhanced CD8+ T-cell induction, whereas reduced numbers of CD8+ T cells were found when GITR signaling was present at primary immunization or at both primary and secondary immunization (Supplementary Fig. S1). These results are therefore in accordance with the previous finding with gp100 and GITR agonistic antibody ( 36). On the other hand, the enhancement of the number of 9m-specific CD8+ T cells by coadministration with mouse GITRL during secondary immunization was seen but not statistically significant ( Figs. 1A and 2A). Thus, the nature of the antigen and the timing of GITR signaling are important variables and need to be considered when incorporating GITR signaling in vaccine strategies.
Finding ways to block the suppressive action of CD4+CD25+ Tregs on the effector functions of CD4+ and CD8+ T cells are a major challenge to develop effective cancer vaccines. Evidence is accumulating that Tregs have a negative effect on the survival of cancer patients ( 16, 17). In a recent study of tumor-infiltrating lymphocytes in patients with ovarian cancer, it was shown that the ratio of CD8+ T cells to CD4+CD25+FOXP3+ Tregs was an important prognostic indicator, with a low ratio associated with a poor outcome ( 17). Strategies to reduce the number of Tregs have included CD25-mediated ( 10– 12, 42, 43) and cyclophosphamide-mediated ( 44, 45) Treg depletion. However, neither approach has specificity for Treg, and both also deplete T effector cells ( 42). Furthermore, given that most human tumor antigens identified to date are self-antigens ( 1– 4, 6), systemic down-regulation of CD4+CD25+ Treg activity and blocking other specific and nonspecific suppressive mechanisms are accompanied by the danger of autoimmune diseases ( 9, 11, 46– 48). Alternative strategies to overcome T-cell suppression without targeting Tregs directly are emerging from studies of the antitumor activities of anti–CTLA-4 ( 12, 49) and the effect of GITR signaling, either with agonist mAb ( 22, 28, 36) or by GITRL as discussed in this report. Importantly, CD8+ T cells resistant to Tregs observed in this study are considered to be tumor antigen specific and are less likely to induce unfavorable autoimmunity. How GITR signaling renders T cells resistant to suppression by CD4+CD25+ Tregs is an important issue. As yet, we have not observed any differences in surface activation markers on Tregresis-CD8+ T cells compared with Tregsens-CD8+ T cells. Still, one plausible explanation is that GITRL provides a strong costimulatory signal to effector T cells and that effector T cells secrete larger amounts of cytokines and up-regulate molecules, making them resistant to Tregs. It is also possible that GITR signaling down-regulates receptors for negative regulators such as transforming growth factor-β, interleukin-10, and/or PD-1. Alternatively, GITR signaling might enhance the activation of nuclear factor-κB and inhibit a suppressive mechanism of Tregs ( 40, 50). We are further evaluating these possibilities through more comprehensive analyses. However, even in the absence of understanding the basis of GITRL effect on CD8+ T cells, our finding that immunization in the presence of GITRL increased the number of CD8+ T cells and their resistance to Tregs provides a firm experimental basis for incorporating GITR signaling in efforts to develop cancer vaccines.
Disclosure of Potential Conflicts of Interest
N. Harada: Employment, ImmunoFrontier, Inc. The other authors disclosed no potential conflicts of interest.
Grant support: Grants-in-Aid for Scientific Research on Priority Areas and for Young Scientists from the Ministry of Education, Culture, Sports, Science and Technology of Japan, and designated grants provided by the Cancer Research Institute, Takeda Science Foundation, The Mochida Memorial Foundation for Medical and Pharmaceutical Research, and Kanae Foundation for the Promotion of Medical Science.
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.
We thank S. Hori, C. Hyuga, K. Mori, and L. Wang for technical assistance.
Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).
↵7 H. Nishikawa and H. Shiku, in preparation.
- Received October 11, 2007.
- Revision received April 2, 2008.
- Accepted April 18, 2008.
- ©2008 American Association for Cancer Research.