The blockade of immune suppression against antitumor responses is a particularly attractive strategy when combined with agents that promote tumor-specific CTLs. In this study, we have attempted to further improve the CTL induction and potent antitumor efficacy of a combination mAb-based therapy (termed “trimAb therapy”) that comprises tumor cell death-inducing anti-death receptor 5 mAb and immune activating anti-CD40 and anti-CD137 mAbs. Among trimAb-treated tumors, the infiltration of CD4+ Foxp3+ cells was greater in progressing tumors compared with stable tumors. Blockade of CTLA-4 (CD152)-mediated signals by an antagonistic mAb substantially increased the tumor rejection rate of trimAb therapy, although the immune responses of draining lymph node cells were not augmented. Interestingly, by comparison, additional treatment with agonistic anti-glucocorticoid-induced TNF receptor mAb, antagonistic anti-programmed death-1 (CD279) mAb, or agonistic anti-OX40 (CD134) mAb significantly augmented immune responses of draining lymph node cells, but did not augment the therapeutic effect of trimAb. CD4 T cell depletion reduced the antitumor effect of anti–CTLA-4 mAb treatment alone, but did not reduce the tumor rejection rate of trimAb in conjunction with anti–CTLA-4 mAb. Thus, the blockade of the CTLA-4–mediated inhibitory signal in tumor infiltrating CTL may be the most effective strategy to augment the effect of immune therapies that generate tumor-specific CTL.
Immunotherapy based on mAbs targeting cancer cells is now developed as a valid approach to treat cancer (1–3); nevertheless, the therapeutic effect is still often relatively low or incomplete. We have previously reported that tumor cell apoptosis induction by anti-death receptor (DR)-5 (TRAIL receptor) mAb exerts a potent antitumor effect in mice (4). Moreover, when this apoptosis-inducing therapy (anti-DR5 mAb treatment) was combined with immune activation (anti-CD40 and anti-CD137 mAb treatment, termed “trimAb”), strong tumor-specific CTLs were promptly induced, resulting in the complete rejection of established tumor in the majority of trimAb treated-mice (5). We believe that the combination therapy of tumor apoptosis induction and immune activation is an attractive approach to treat cancer (6, 7), because trimAb induced complete rejection of a large proportion of fibrosarcomas induced de novo by 3-methylcholanthrene (MCA) and could also reject established tumors containing as many as 90% TRAIL-resistant tumor variants (5). However, the therapeutic effect of trimAb was limited against much larger tumor masses, and trimAb did not achieve a 100% rejection rate.
Suppressive mechanisms in immune responses normally play a critical role in maintaining immune homeostasis. However, these suppressive mechanisms are also considered as one of main reasons for the failure of cancer immunotherapies because they induce peripheral tolerance of tumor-specific immune responses and allow tumor growth (8, 9). CD4+ CD25+ Foxp3+ regulatory T cells have been revealed as the most important population of immune suppressors, and their depletion has been reported to augment antitumor immune responses (10, 11). CTLA-4 (CD152) and glucocorticoid-induced TNF receptor (GITR) were reported as the critical molecules for regulatory T cell function (11–14). CTLA-4–mediated signals also directly inhibit activated T cell functions (15), and thus blockade of CTLA-4–mediated signals has been suggested as a possible strategy to treat cancers (16, 17). Agonistic anti-GITR mAb (DTA-1) treatment was also report to eradicate established tumors (18). Moreover, programmed death-1 (PD-1) –mediated signals were also reported as a critical inhibitory mechanism regulating antitumor immune responses (19, 20). It is expected that inhibitory signals might be augmented as a feedback mechanism when immune responses are enhanced, and thus blockade of inhibitory signals during therapy-induced immune responses to tumor may well augment the therapeutic effect of such tumor-specific T cell-inducing therapies.
Based on our finding that CD4+ Foxp3+ cells preferentially infiltrated in progressing tumor masses, but not stable tumors, of trimAb-treated mice, we have attempted to improve the therapeutic effect of trimAb using a variety of mAbs that block inhibitory pathways. Antagonistic anti–CTLA-4 mAb, but not agonistic anti-GITR mAb or antagonistic anti–PD-1 mAb, substantially increased the cancer rejection rate caused by trimAb treatment. Additional activation of CD4 T cells by agonistic anti-OX40 mAb did not augment therapeutic effect of trimAb. Anti-GITR mAb, anti–PD-1 mAb, and anti-OX40 mAb, but not anti–CTLA-4 mAb, augmented tumor-specific immune responses in the tumor-draining lymph nodes. CD4 T cell depletion did not reduce the improved tumor rejection rate of trimAb caused by additional anti–CTLA-4 mAb treatment. Collectively, blockade of CTLA-4 may be a useful strategy to augment CTL-inducing trimAb therapy possibly due to its direct effect by releasing CTL from suppression in the tumor mass.
Materials and Methods
BALB/c mice at 6 wk of age were from Charles River Japan (Yokohama, Japan) and The Walter and Eliza Hall Institute of Medical Research (Melbourne, Australia). All mice were maintained under specific pathogen-free conditions and used in accordance with the institutional guidelines of Juntendo University (Tokyo, Japan) and the Peter MacCallum Cancer Centre (East Melbourne, Australia).
Abs and tumors
Agonistic anti-mouse DR5 mAb (MD5-1) (4), agonistic anti-mouse CD40 mAb (FGK45; kindly provided by Dr. Antonius G. Rolink, University of Basel, Basel, Switzerland) (21), agonistic anti-mouse CD137 (4-1 BB) mAb (3H3; kindly provided by Dr. Robert S. Mittler, Emory University, Atlanta, GA) (22), antagonistic anti-mouse CTLA-4 (CD152) mAb (UC10-4F10; kindly provided by Dr. Jeffrey A. Bluestone, University of California, San Francisco, CA) (23), agonistic anti-mouse GITR mAb (DTA-1; kindly provided by Dr. Shimon Sakaguchi, Kyoto University, Kyoto, Japan) (24), antagonistic anti-mouse PD-1 (CD279) mAb (RMP1-14) (25), agonistic anti-mouse OX40 (CD134) mAb (OX86) (26), and anti-mouse CD4 mAb (GK1.5) were prepared and purified in our laboratory as previously described (4, 5). The BALB/c-derived TRAIL-sensitive 4T1 mammary carcinoma and colon 26 (CT26) colon carcinoma were maintained as previously described (4, 5).
Therapy of transplanted s.c. tumors
Mice were inoculated with 2 × 105 4T1 tumor cells s.c. in the hind flank. When tumors developed to a size of ∼9 mm2 (7 d after tumor inoculation), 36 mm2 (10–12 d after tumor inoculation), or 64 mm2 (18 d after tumor inoculation), groups of mice were administered i.p. with 100 μg anti-DR5 mAb, anti-CD40 mAb, anti-CD137 mAb, anti–CTLA-4 mAb, anti-GITR mAb, anti–PD-1 mAb, anti-CD134 mAb, control hamster IgG, and/or control rat IgG (Sigma-Aldrich, St. Louis, MO) four times every 4 d. Tumor size was measured periodically with a caliper as the product of two perpendicular diameters (mm2).
CTL functional analysis in draining lymph node
BALB/c mice were inoculated with 4T1 tumor cells (2 × 105/50 μl) in the left footpad, and mAb therapy commenced 7 d after tumor inoculation. Seven days after a single treatment, mononuclear cells were prepared from the tumor-draining lymph node and incubated with mitomycin C-treated (200 μg/ml for 2 h; Kyowa Hakko, New York, NY) 4T1 or CT26 cells (responder/stimulator = 4:1) for 16 h (4). IFN-γ in the cell-free supernatant was determined by using mouse IFN-γ–specific ELISA kits (OptEIA, BD Biosciences, San Jose, CA) according to the manufacturer’s instructions. Cytotoxic activity was tested against 4T1 cells or CT26 cells by a standard 4 h [51Cr]-release assay as previously described (4, 5).
Therapy of MCA-induced primary tumors
BALB/c mice were inoculated s.c. in the hind flank with 200 μg MCA (Sigma-Aldrich) in 0.1 ml corn oil (5, 27). When MCA-induced tumors developed to a size of 9–25 mm2 (60–120 d after MCA inoculation), groups of mice were administered i.p. with 100 μg anti-DR5 mAb, anti-CD40 mAb, anti-CD137 mAb, anti–CTLA-4 mAb, control hamster IgG, and/or control rat IgG four times every 4 d. Anti-CD4 mAb treatment to deplete CD4 T cells was started 4 d prior to treatment with the therapeutic mAbs and repeated every 4 to 5 d. CD4 T cell depletion was confirmed by flow cytometric analysis. Growth of tumors was monitored over the course of 100–180 d. Tumor size was measured periodically with a caliper as the product of two perpendicular diameters (mm2).
Progressing, stable (no progressive growth for >7 d), and recurrent (progressive growth >7 d postregression) tumors were removed 7–14 d following treatment. Macroscopically rejected tumors were removed 25–30 d after the last mAb treatment. Three-micrometer cryostat sections were fixed with 20% buffered formalin and stained with H&E (4). Immunofluorescent staining was performed as described previously (2829). The fluorescence images were captured with a Zeiss AxioPlan2 fluorescence microscope equipped with a digital camera (Zeiss, Jena, Germany). To quantify the positive-staining cells, the number of infiltrating CD8+, CD4+ Foxp3+, CD4+ Foxp3−, and Gr-1+ CD11b+ cells in each section was analyzed using the KS400 Image Analysis system (Zeiss). Three to six typical areas of tumor margin were examined in each specimen, and the numbers of positively stained cells were calculated per field of view (T cells: on ×40 and myeloid cells: on ×20).
Statistical analysis was performed by two-sample t test for the IFN-γ ELISA, cytotoxicity, and tumor growth data. Significant difference in tumor rejection was determined by the Fisher exact test. All p values <0.05 were considered significant.
Accumulation of CD4+ Foxp3+ T cells in progressing 4T1 tumor masses following trimAb therapy
As we previously reported (5), a rational mAb-based therapy combining tumor cell death induction (anti-DR5 mAb-induced apoptosis) and immune activation (anti-CD40 and anti-CD137 mAb-induced CTL induction/activation) eradicated established (∼25 mm2, 10 d after tumor inoculation) TRAIL-sensitive 4T1 tumors; however, not all mice were cured. When we examined progressing, stable, and rejected tumors following trimAb therapy, tumor rejection was confirmed by histological analysis (Fig. 1). Infiltration of CD8 T cells was significantly increased in trimAb-treated stable tumors when compared with control progressing tumors, but not trimAb-treated progressing tumors. By contrast, infiltration of both CD4+ Foxp3+ and CD4+ Foxp3− T cells significantly decreased in trimAb-treated tumors and that resulted in a significant increase in CD8 T cell/regulatory T cell ratio in trimAb-treated tumors (Fig. 1). Moreover, CD4+ Foxp3+ regulatory T cells were significantly reduced in stable tumors even when compared with trimAb-treated progressing tumors and that resulted in a far greater CD8 T cell/regulatory T cell ratio in trimAb-treated stable tumors. Thus, CD4+ Foxp3+ regulatory T cells that dominantly infiltrate into progressing tumors might promote tumor growth by inhibiting the therapeutic effect of trimAb.
Blockade of CTLA-4 augments the antitumor effect of trimAb therapy
A combination of antagonistic anti–CTLA-4 mAb, agonistic anti-GITR mAb, or antagonistic anti–PD-1 mAb with anti-DR5 mAb inhibited tumor growth more effectively than any monotherapy with each mAb, but tumor rejection was not induced (Fig. 2). Tumor rejection was only induced by trimAb therapy, and the rejection rate was significantly increased when trimAb was combined with anti–CTLA-4 mAb, but not anti-GITR mAb or anti–PD-1 mAb (Fig. 2).
When therapies were commenced at a tumor size of 9 mm2, trimAb and the combination of trimAb and anti–CTLA-4 mAb induced a greater rate of tumor rejection (∼80%), and tumor growth was significantly inhibited by trimAb or trimAb with anti–CTLA-4 mAb treatment when compared with control or anti–CTLA-4 mAb treatment alone (Fig. 3A). Although 4T1 tumors of 64 mm2 were not rejected by trimAb therapy, additional anti–CTLA-4 mAb treatment significantly inhibited tumor growth at some time points (days 28–34) and achieved complete tumor rejection in a small number of mice (Fig. 3B). By contrast, when trimAb therapy was combined with anti-GITR mAb or anti–PD-1 mAb against 9 mm2 or 64 mm2 tumors, no effect greater than trimAb therapy alone was observed (data not shown). We did not detect any metastatic nodules in the lungs and livers of the mice that had rejected s.c. primary 4T1 tumors by mAb treatments (data not shown). Thus, CTLA-4 is the preferred target molecule among those examined when blocking immune suppression in the context of the trimAb therapy.
Combination with anti-OX40 mAb treatment did not significantly augment the antitumor effect of trimAb
We have reported that CTL induction/activation and the therapeutic effect of trimAb did not depend on CD4 T cells (5), and therefore activation of Th CD4 cells is an alternative strategy to further augment the therapeutic effect of trimAb. It is thought that OX40-mediated signals increase helper function due to preferential expression of OX40 on CD4 T cells (30). Treatment with anti-OX40 mAb alone slightly but significantly inhibited tumor growth, but only when the treatment commenced against tumors as small as 9 mm2 in size (Fig. 4). TrimAb and trimAb with anti-OX40 mAb-induced tumor rejection when treatments were commenced against 9 mm2 and 25 mm2 4T1 tumors, and both treatments significantly inhibited tumor growth when compared with control or anti-OX40 therapy alone. However, additional agonistic anti-OX40 mAb treatment did not augment the rejection rate or inhibitory effect of trimAb on tumor growth (Fig. 4).
CTL activity in the tumor draining lymph node was not augmented by anti–CTLA-4 mAb treatment
Consistent with our previous report (5), draining lymph node cells of 4T1 tumor-bearing mice treated with trimAb significantly produced IFN-γ when stimulated with 4T1 tumor cells, but not CT26 tumor cells, ex vivo (Fig. 5A). IFN-γ production did not correlate with the tumor growth following trimAb or trimAb plus anti–CTLA-4 mAb therapy, indicating that the different response of tumors to the therapies in individual mice was not due to different level of immune activation in tumor draining lymph node by the treatments. Interestingly, additional anti–CTLA-4 mAb treatment did not increase IFN-γ production by draining lymph node cells in trimAb-treated mice; however, additional treatment with anti-GITR mAb, anti–PD-1 mAb, or anti-OX40 mAb significantly augmented IFN-γ production. TrimAb induced specific cytotoxic activity against 4T1 cells, and additional anti-OX40 mAb treatment, but not other mAb treatments, slightly but significantly augmented this specific cytotoxicty (Fig. 5B). Neither IFN-γ production nor cytotoxic activity was observed when CD8 T cells were depleted (data not shown). Cytotoxic activity was not observed against CT26 tumor cells (data not shown). When lymph node cells were prepared 4 d after a single treatment or 4–7 d after a second treatment, similar results were obtained (data not shown). Thus, anti-GITR mAb, anti-OX40 mAb, or anti–PD-1 mAb treat-ment was more effective at increasing tumor-specific immune responses of the draining lymph node cells in trimAb-treated mice compared with anti–CTLA-4 mAb treatment.
Increased CD4 T cell infiltration by additional anti-GITR mAb or anti-OX40 mAb treatment in stable tumors
In trimAb and anti-GITR mAb-treated stable 4T1 tumors, despite no differences in the number of infiltrating CD8 T cells, the CD8 T cell/regulatory T cell ratio was significantly decreased due to a significant increase in infiltrating CD4+ Foxp3+ T cells (Fig. 6). Interestingly, additional anti-OX40 mAb treatment significantly augmented the infiltration of both CD4+ Foxp3+ T cells and CD4+ Foxp3− T cells. Additional anti–PD-1 mAb treatment did not significantly modify the number of CD8 T cells, CD4+ Foxp3+ T cells, or CD4+ Foxp3− T cells or the CD8 T cell/regulatory T cell ratio in stable 4T1 tumors. Additional anti–CTLA-4 mAb treatment appeared to increase the infiltration of CD8+ T cells and the CD8 T cell/regulatory T cell ratio, but these changes were not significant.
Augmentation of the antitumor effect of trimAb by anti–CTLA-4 mAb against MCA-induced sarcoma following CD4 T cell depletion
We then examined the additional effect of anti–CTLA-4 mAb treatment on trimAb therapy against established MCA-induced primary tumors. Anti–CTLA-4 mAb treatment alone significantly inhibited tumor growth but did not induce tumor rejection, and CD4 T cell depletion reduced this antitumor effect of anti–CTLA-4 treatment (Fig. 7), suggesting that effect of antagonistic anti–CTLA-4 mAb alone is possibly mediated by inhibition of CD4 regulatory T cell function, consistent with CTLA-4 reported as a critical effector molecule expressed on regulatory T cells (12).
TrimAb therapy eradicated MCA-induced primary tumors in some mice as we previously reported (5), and the rejection rate was significantly augmented by the combination with anti–CTLA-4 mAb treatment (Fig. 7). CD4 T cell depletion did not decrease the tumor rejection rate induced by trimAb therapy as we previously reported (5), suggesting that trimAb did not require CD4-positive cells to reject tumors, although the early response to trimAb therapy (days 0–10 after the first treatment) was reduced, and tumor rejection was delayed. Interestingly, additional anti–CTLA-4 mAb treatment increased the tumor rejection rate of trimAb in CD4 T cell-depleted mice similar to that in CD4 T cell-intact mice, indicating that anti–CTLA-4 mAb treatment enhanced the tumor rejection rate caused by trimAb independently of CD4 T cells. Thus, augmentation of therapeutic effect of trimAb by antagonistic anti–CTLA-4 mAb appears to be primarily mediated by the relief of direct suppression on CTL.
Immunohistochemistry demonstrated a significant increase in the CD8 T cell/regulatory T cell ratio in stable tumors, due to significantly less infiltration of CD4+ Foxp3+ T cells when compared with that observed in the progressing tumors following either trimAb or trimAb + anti–CTLA-4 mAb treatment (Fig. 8). CD8 T cells, CD4+ Foxp3+ T cells, and CD4+ Foxp3− T cells increased in the trimAb-treated recurrent tumors compared with stable tumors. When compared with similarly responding tumors treated with trimAb alone, additional anti–CTLA-4 mAb treatment did not alter the number of infiltrating CD4+ Foxp3+ T cells or CD4+ Foxp3− T cells either in progressing or stable tumors. However, CD8+ T cells were significantly increased by additional anti–CTLA-4 mAb treatment in both stable and progressing tumors when compared with similarly responding tumors treated with trimAb alone. These observations support the likelihood that anti–CTLA-4 mAb treatment directly effects CD8 T cells in the context of trimAb therapy.
Increased infiltration in myeloid-derived suppressor cells in progressing tumors
Gr1+ CD11b+ myeloid-derived cells and regulatory CD4+ Foxp3+ T cells are now implicated in the suppression of antitumor immune responses (31–33). Moreover, Gr1+ CD11b+ myeloid-derived cells might acquire immune-suppressing function from TGF-β produced by CD4+ Foxp3+ T cells (34) or might either generate or expand tumor-induced regulatory T cells (35, 36), indicating a potential cross-talk between these immune suppressing cells in the tumor mass. Thus, we finally examined the infiltration of Gr1+ CD11b+ cells in MCA-induced tumors (Fig. 9). Infiltrating Gr1+ CD11b+ cells were dramatically reduced in stable tumors, but a number of Gr1+ CD11b+ cells were still observed even when the mice were treated with trimAb and anti–CTLA-4 mAb. Gr1+ CD11b+ cells in the trimAb-treated recurrent tumors were significantly less when compared with trimAb-treated progressing tumors, but significantly more when compared with trimAb-treated stable tumors. Interestingly, additional anti–CTLA-4 mAb treatment significantly reduced the number of Gr1+ CD11b+ cells in the progressing tumors when compared with trimAb-treated progressing tumors (Fig. 9), whereas the number of CD4+ Foxp3+ T cells was comparable (Fig. 8).
Induction of tumor-specific CTL is thought to be a critical requirement for effective cancer immunotherapy, and thus improving the ability to induce and sustain effector CTL has been primary focus for the clinical application (37, 38). We have previously reported a triple mAb-based therapy that induces both tumor cell apoptosis and tumor-specific CTL as an effective strategy that can induce complete rejection of a significant proportion of established experimental and de novo tumors (5). However, immunological suppressor cells, particularly regulator T cells and Gr1+ CD11b+ cells, have been implicated to prevent effective tumor-specific immune responses prior to and following immunothera-pies (10, 11, 31–33). Therefore, in this study, we examined the contribution of relevant immune-suppressing pathways that might hinder the effectiveness of CTL-inducing trimAb therapy.
Histological analyses showed that CD4+ Foxp3+ cells and Gr1+ CD11b+ cells infiltrated into progressing tumors that resisted trimAb therapy to a similar extent as in untreated mice. This suggested that tumor-infiltrating CD4+ Foxp3+ regulatory T cells and Gr1+ CD11b+ myeloid-derived cells might contribute to tumor growth in both untreated and treated mice. Interestingly, additional anti–CTLA-4 mAb treatment substantially decreased Gr1+ CD11b+ cell infiltration, but did not significantly effect on CD4+ Foxp3+ cell infiltration, into progressing trimAb-treated MCA-induced tumors. Thus, CTLA-4 might contribute to Gr1+ CD11b+ cell infiltration into tumor.
Treatment combining trimAb with anti-GITR mAb, anti–PD-1 mAb, or anti-OX40 mAb significantly augmented tumor-specific immune responses in the tumor draining lymph node. Anti-GITR mAb treatment might directly activate conventional T cells rather than inhibit regulatory T cell-mediated suppression, because it was recently reported that anti-GITR mAb (DTA1) directly activates conventional T cells (14, 39, 40). Anti–PD-1 mAb treatment has been shown to abrogate direct immune suppression on T cells (20). Anti-OX40 mAb was reported to activate not only CD4 T cells, but also CD8 T cells (41). Thus, anti-GITR mAb, anti–PD-1 mAb, or anti-OX 40 mAb treatment might directly activate trimAb-induced tumor specific CTL responses in the tumor draining lymph node. However, additional treatment with anti-GITR mAb, anti–PD-1 mAb, or anti-OX40 mAb lacked therapeutic benefit in trimAb-treated mice. Thus, trimAb therapy alone sufficiently activates immune responses in the tumor draining lymph node, inducing CTL to reject establish tumor, and can be improved further by altering the tumor microenvironment.
Additional anti-GITR mAb dramatically augmented 4T1-specific IFN-γ production in the draining lymph node of trimAb-treated mice, whereas the number of infiltrating CD8 T cells in stable tumors was comparable between trimAb treatment and trimAb plus anti-GITR mAb treatment. Infiltration of regulatory T cells into stable tumors was increased by additional anti-GITR mAb treatment and that possibly resulted in the effective inhibition of infiltration and function of CD8 T cells within the tumor mass. Therefore, additional anti-GITR mAb treatment did not augment the trimAb-induced therapeutic effect despite strongly activating immunity in the tumor draining lymph nodes. These results indicate that targeting immunosuppression in the tumor mass should be used to augmented the therapeutic effect of immunotherapy.
Histological analysis showed that additional anti-OX40 mAb treatment significantly increased both infiltrating Foxp3+ and Foxp3− CD4+ T cells in stable tumors. It has been recently reported that OX40 mediates an activating signal to regulatory T cells (42), and OX40-mediated signals inhibit regulatory T cell function (43). OX40 signaling on conventional T cells might overcome regulatory T cell-mediated suppression and thus the effect of anti-OX40 mAb treatment might result in either activation or suppression of immune responses depending on the ratio of conventional T cells versus regulatory T cells in tissue (42). Therefore, we expect that anti-OX40 mAb treatment might activate immune responses in the tumor draining lymph nodes while suppressing CTL responses via the activation of regulatory T cells that predominantly reside in the tumor mass, hence resulting in no net effect on trimAb therapy.
Augmentation of IFN-γ production from draining lymph node cells by additional anti–PD-1 mAb treatment was relatively weaker when compared with that caused by anti-GITR mAb or anti-OX40 mAb. Moreover, it was reported that purified T cells from PD-1–deficient mice did not show any increased responses (44). Thus, blockade of the interaction between PD-1 ligands/PD-1 might not be sufficient to augment CTL responses and enhance the therapeutic effect of trimAb. Moreover, 4T1 tumor cells are PD-L1/PD-L2 negative (data not shown), and thus the contribution of PD-1–mediated signals would be less when compared with that previously reported using PD-L1–expressing tumors (19).
CD4 T cells appear not to be critical for trimAb with/without anti–CTLA-4 mAb to eradicate MCA-induced de novo tumors. However, CD4 T cell depletion slightly delayed tumor response to trimAb in the initial therapy phase, indicating that Th CD4 cells play a limited but meaningful role in the therapeutic effect of trimAb. Anti-CD4 mAb depleted both Th and regulatory CD4 T cells, and anti–CTLA-4 mAb modified functions of Th CD4 cells, regulatory CD4 T cells, and CD8 T cells. Thus, although CD4 T cell depletion did not alter the therapeutic effect of trimAb with anti–CTLA-4 mAb, this was not evidence that trimAb over-comes regulatory T cell-mediated immune suppression. Our results indicated that the direct enhancement of effector CTL function by anti–CTLA-4 mAb plays a sufficient role to augment the therapeutic effect of trimAb. It was recently reported that both direct enhancement of effector T cell function and concomitant inhibition of regulatory T cell activity were essential for the full cancer therapeutic effects of anti–CTLA-4 mAb alone (45). The relative contribution of CTLA-4–mediated signals on either CTL, Th CD4 cells, or regulatory T cells in tumor-specific immune responses might depend upon the immune activation levels in tumor-bearing hosts. Further experiments with specific depletion of Th cells or regulatory T cells would provide the conclusive evidence for the role of CD4 T cells in trimAb therapy.
It was reported that CTLA-4 blockade increases IFN-γ production from tumor-infiltrating effector ICOS+ CD4 T cells and reduces CD4+ Foxp3+ T cells within the tumor (46). In this study, additional anti–CTLA-4 mAb treatment significantly augmented the efficacy of trimAb therapy, but did not enhance tumor-specific immune responses of the draining lymph node cells or alter CD4+ Foxp3+ T cell or CD4+ Foxp3− T cell infiltration into the tumor mass. Moreover, additional anti–CTLA-4 mAb treatment increased the CD8 T cell/regulatory T cell ratio in stable MCA-induced tumors, the number of infiltrated CD8+ T cells when compared with similarly responding MCA-induced tumors treated with trimAb alone, and the tumor rejection rate in CD4 T cell-depleted mice. Thus, anti–CTLA-4 mAb mainly plays a role by directly providing a positive effect on CTLs possibly to enhance their infiltration and/or effector functions within the tumors. Collectively, our results suggest that antitumor CTL-inducing therapies might be further improved by blocking the suppressive pathways that impact directly on CTL within the tumor. TrimAb-induced regressed tumors were too small to obtain a sufficient level of CTL for in situ and ex vivo analysis of CTL functions; however, examination of the status of CTL activity and immune responses in the tumor mass would be valuable if it was practical. Tumor microenvironment and cross-talk between cancer and immune cells have been suggested to modulate immune responses against cancer within tumor masses (47, 48). Further studies of the microenvironment in progressing and regressing tumors using novel technologies for examination of minimal sample material are required to conclusively understand immune mechanisms in tumor rejection and improve cancer immunotherapy.
We thank Yuka Tanno and Janelle Sharkey for preparing mAbs and Michelle Stirling for maintaining the mice.
Disclosures The authors have no financial conflicts of interest.
This work was supported by the Ministry of Education, Science, and Culture, Japan (to K.T.), Grant 07-23904 from the Princess Takamatsu Cancer Research Fund (to K.T.), a Grant-in-Aid from the Tokyo Biochemical Research Foundation (to K.T.), a National Health and Medical Research Council of Australia Senior Principal Research Fellowship (to M.J.S.), a Doherty Fellowship (to M.W.L.T.), a Susan G. Komen Breast Cancer Foundation grant (to M.J.S.), and a project grant from the Cancer Council of Victoria (to M.J.S.).
Abbreviations used in this paper:
- death receptor
- glucocorticoid-induced TNF receptor
- programmed death-1.
- Received September 14, 2009.
- Accepted March 12, 2010.
- Copyright © 2010 by The American Association of Immunologists, Inc.