We examined the hypothesis that a failure of the immune system to eradicate tumors is due to the immunosuppressive environment created by the growing tumor, which is influenced by the site of tumor growth. We demonstrated that T cell responses to a bystander Ag in mice were suppressed by a growing CT26 tumor. T cells purified from the growing tumor expressed mRNA for IL-10, TGF-β, and Foxp3. Intracellular cytokine staining revealed a high frequency of IL-10-secreting macrophages, dendritic cells, and CD4+ and CD8+ T cells infiltrating the tumor. In contrast, T cell IFN-γ production was weak and CD8+ CTL responses were undetectable in mice with CT26 lung metastases and weak and transient following s.c. injection of CT26 cells, but were enhanced in the presence of anti-IL-10 and anti-TGF-β. Consistent with this, removal of CD8+ T cells abrogated CTL responses and promoted progression of the s.c. tumor. However, in the lung model, depletion of CD8+ T cells significantly reduced the tumor burden. Furthermore, depletion of CD4+ or CD25+ T cells in vivo reduced tumor burden in s.c. and lung models, and this was associated with significantly enhanced IFN-γ production by CD8+ T cells. These findings suggest that tumor growth facilitates the induction or recruitment of CD4+ regulatory T cells that secrete IL-10 and TGF-β and suppress effector CD8+ T cell responses. However, CD8+ T regulatory cells expressing IL-10 and TGF-β are also recruited or activated by the immunosuppressive environment of the lung, where they may suppress the induction of antitumor immunity.
Tumor progression in normal immunocompetent individuals may reflect a failure of the immune system to recognize tumor Ags or may result from subversion of antitumor immune responses. Innate and adaptive immune response can be induced against tumors, and the protective effector cells include CD8+ CTL, IFN-γ-producing CD4+ and CD8+ T cells, NK cells, and macrophages. Although T cells usually constitute a main immune cell population attracted to the tumor site, they are often ineffective at killing the tumor and recent evidence suggests that this may result from the function of regulatory T (Treg)5 cells (1, 2).
Natural CD4+CD25+ Treg cells that express the transcriptional repressor Foxp3 and emerge as mature T cells from the thymus play a critical role in maintaining tolerance to self Ag and in preventing autoimmunity (3). Inducible Treg cells secreting IL-10 (termed type 1 Treg cells) or TGF-β (termed Th3 cells) are generated in the periphery in response to pathogen and self Ags (4). Natural and inducible Treg cells can be beneficial to the host during infection by preventing pathogen-induced immunopathology (5, 6, 7). However, the induction or activation of Treg cells by pathogens may be a strategy to subvert protective immunity, and depletion of CD4+CD25+ Treg cells enhances survival during certain infections (8). Depletion of CD4+CD25+ T cells can also enhance antitumor immunity (2, 9). However, the role of inducible Treg cells in antitumor immunity has received little attention.
Naive CD4+ T cells can differentiate into effector Th1 or Th2 cells or Treg cells, and the selective induction of these distinct subtypes appears to be determined by a number of factors, including the maturation status of dendritic cells (DC) and regulatory cytokines secreted by cells of the innate immune system (4). The balance of effector vs Treg cells is also influenced by the tissue and environment where the immune response is generated. This may have a bearing on resolution or progression of tumor growth, especially for tumors in the respiratory or gastrointestinal mucosa, where the immune responses must be tightly regulated to maintain tolerance to inhaled and food Ags, respectively. The lung is an immunosuppressive environment, with high constitutive production of IL-10, which is necessary to protect the integrity of the delicate highly vascularized tissue, especially against excessive responses to allergens and pathology during infection (5, 10). DC in lungs secrete high levels of IL-10, and lymph nodes that drain the lung are less effective at activating naive T cells than DC from distal lymph nodes (10, 11). Therefore, the environment in the lung may not be conducive to the development of effective antitumor immune responses, and thereby facilitates tumor growth.
In this study, we have used the CT26 cell carcinoma, which can develop into a solid tumor when given s.c. or metastases in the lung when injected i.v., to examine the influence of the tumor environment on the relative development of effector and suppressor T cells during tumor growth. We found that CTL responses and IFN-γ-secreting T cells are suppressed during tumor growth, with a corresponding increase in IL-10-, TGF-β-, and Foxp3-expressing T cells. Depletion of CD4+ or CD4+CD25+ T cells in vivo enhanced IFN-γ production by CD8+ T cells and significantly reduced tumor growth and enhanced survival. In contrast, CD8+ T cells had a protective role against the s.c. tumor, but surprisingly advanced the tumor growth in the lung. Our findings suggest the immunosuppressive environment of the lung is conducive to the development of CD8+ as well as CD4+ Treg cells, which suppress the development of effector T cell responses.
Materials and Methods
BALB/c mice were purchased from Harlan. DO.11.10 OVA TCR transgenic mice were bred in house. All mice were maintained under specific pathogen-free conditions and according to the regulations and guidelines of the Irish Department of Health.
The CT26 colon carcinoma-derived cell line was maintained in RPMI 1640 supplemented with 10% heat-inactivated FCS. BALB/c mice were injected s.c. with 2 × 105 CT26 tumor cells, and tumor volume was monitored from day 7. Lung metastases were established by injecting BALB/c mice i.v. with 3 × 105 CT26 cells. To determine numbers of metastases, the lungs of sacrificed mice were stained by intertracheal administration of Indian ink, removed, added to Fekete’s solution, and countered under a microscope.
Influence of tumor on proliferation and cytokine responses to an unrelated Ag
D0.11.10 mice were injected s.c. in the flank with 200 μg of OVA alone or with 2 × 105 CT26 cells. Mice were boosted 7 days later with 200 μg of OVA, and 14 days later lymph node cell suspensions were prepared and cultured at 2 × 106/ml in 96 microtiter round-bottom plates with OVA (50, 150, and 500 μg/ml) at 37°C in 5% CO2. After 3 days, supernatants were removed and IFN-γ concentrations were determined by two-site ELISA. After 4 days of culture, 2 μCi of [3H]thymidine (Amersham Biosciences) was added to each well. Plates were incubated for a further 6 h, after which cells were harvested (Tomtec Harvester 96 Mach III M) onto filtermats and cpm were determined using a beta counter (1450 Microbeta Trilux; Wallac).
T cell purification by magnetic cell sorting
Cell suspensions were prepared from lymph nodes, lungs, or surgically removed tumors. Tissue samples were washed with medium, finely chopped with a scalpel blade, and incubated with 0.1% solution of collagenase D (Sigma-Aldrich) in HBSS for 30 min at 37°C. Cell suspensions were then passed through a 70 μm cell strainer and RBC were lysed. Cells were washed, counted, and resuspended in 90 μl of MACS buffer with 10 μl of anti-CD4 or anti-CD8 magnetic beads (Miltenyi Biotec) and incubated at 4–8°C for 15 min. MACS buffer (2 ml) was added to each sample and then centrifuged at 300 × g for 10 min. Each sample was then resuspended in 500 μl of MACS buffer and sorted using an AutoMacs (Miltenyi Biotec). The purities of CD4+ and CD8+ T cell fractions were 93–94% from lymph nodes and 82–85% from tumors.
Spleens were removed from mice 7, 14, or 21 days posttumor challenge, and a single-cell suspension was prepared. Spleen cells (1.5 × 107 cells) were cultured with irradiated (100 Gy) CT26 tumor cells at a ratio of 50:1. After 5 days, effector cells were cultured with 51Cr-labeled CT26 target cells (104/well) in triplicate in a 96-well round-bottom microtiter plate at E:T ratios of 100:1, 50:1, and 25:1 and incubated at 37°C. After 4 h, 50 μl of supernatant was removed and added to 150 μl of scintillation fluid; 51Cr release was determined using a Microbeta counter and percentage of lysis was calculated as follows: (sample cpm − spontaneous cpm)/(maximum cpm − spontaneous cpm) × 100.
Cell depletion in vivo
Mice were depleted of CD4+ or CD8+ T cells by i.p. injection of 1 mg of anti-CD4 (GK1.5) or anti-CD8 (YTS169), respectively, at days 0, 3, 7, and 12. CD25+ T cells were depleted by i.p. injection of 1 mg of anti-CD25 Ab (PC61) on days 0, 3, 7, and 12.
RNA was extracted from tumor cells or purified CD4+ and CD8+12–18 Taq polymerase (Promega) using a Peltier Thermal Cycler.
Flow cytometric analysis and intracellular cytokine staining
Single-cell suspensions were prepared from lymph nodes, lungs, and tumors. Lungs and tumors were digested in HBSS with 0.1% collagenase D (Sigma-Aldrich). Cells were stimulated with PMA (10 ng/ml) and ionomycin (1 μg/ml) for 1 h, and then brefeldin A (10 μg/ml) was added for 4 h at 37°C. Cells were resuspended with Abs specific for CD4 (Caltag Laboratories), CD8 (BD Pharmingen), CD11c (BD Pharmigen), F4/80 (Caltag Laboratories), or isotype control Abs (BD Pharmingen). Cells were then fixed with 50 μl of fixation medium A (Fix & Perm cell permeabilization kit; Caltag Laboratories) and incubated with 50 μl of the permeabilization medium B (Fix & Perm cell permeabilization kit; Caltag Laboratories) and 5 μl of anti-IL-10 or anti-IFN-γ Abs (BD Pharmingen). Fc receptor binding of Abs was minimized by preincubation with rat anti-mouse CD16/CD32 (FcRγIII/II) mAb (Mouse BD FC Block from BD Pharmingen). Immunofluorescence analyzed was performed on a FACSCalibur (BD Biosciences), with CellQuest software.
CT26 tumor growth inhibits T cell responses to unrelated Ags
To test the possibility that the failure to generate effective adaptive immune responses during tumor growth may result from an immunosuppressive environment created by the growing tumor, we examined the influence of a growing tumor on T cell responses to unrelated Ag. OVA-TCR transgenic mice were injected s.c. with OVA in the presence or absence of CT26 cells, and mice were boosted with OVA after 7 days and sacrificed 14 days later. Draining lymph nodes were removed, and OVA-specific T cell proliferation and IFN-γ production were tested. Mice challenged with CT26 cells and OVA had a significant reduction in OVA-specific T cell proliferation when compared with mice injected with OVA only (Fig. 1⇓A). Coadministration of the CT26 cells also significantly reduced OVA-specific IFN-γ production in the lymph node (Fig. 1⇓B), demonstrating that the growing CT26 tumor suppresses immune responses to other Ags particularly at the site of injection.
Immunosuppressive cytokines are differentially induced by CT26 tumor at different sites
To examine the possible mediators of immunosuppression during tumor growth, we examined the expression of cytokine mRNA in the cultured tumor cells and the growing tumor ex vivo. Cultured CT26 tumor cells expressed mRNA for the anti-inflammatory cytokine, TGF-β (Fig. 2⇓A). Furthermore, TGF-β protein was detected in supernatants of the growing tumor (Fig. 2⇓B). Cox-2, IL-23p19, IL-15, and IP-10 mRNA were also expressed by CT26 cells (Fig. 2⇓A), but we failed to detect expression of mRNA for IL-1β, IL-2, IL-4, IL-5, IL-6, IL-10, IL-12p35, IL-12p40, IL-13, IL-18, IL-27p28, IL-27EB13, TNF-α, and IFN-γ (data not shown). Although IL-10 mRNA could not be detected in the in vitro cultivated tumor (Fig. 2⇓A) even after stimulation of the cells with PMA (data not shown), it was detected in the tumor mass ex vivo (Fig. 2⇓C), suggesting that cells infiltrating the tumor in vivo secreted IL-10. Therefore, we examined IL-10 and TGF-β mRNA expression by RT-PCR in CD4+ and CD8+ T cells purified from the lungs and cervical lymph nodes of mice with CT26 lung metastases and from solid tumor and inguinal lymph nodes of mice injected s.c. with CT26 cells.
In the lung model, IL-10 expression was considerably enhanced in both CD4+ and CD8+ T cells from the lungs of tumor-bearing mice compared with that found in T cells from the lungs of naive mice; however, constitutive expression of IL-10 mRNA was high in CD4+ T cells from lungs compared with lymph nodes of naive mice (Fig. 3⇓A). There was little change in IL-10 mRNA expression in the cervical lymph nodes of mice with lung metastases. In addition, TGF-β mRNA expression was dramatically higher in both CD4+ and CD8+ T cells purified from the lungs of mice with CT26 lung metastases when compared with naive control mice, with little change in the lymph nodes (Fig. 3⇓A). These findings suggest that growth of CT26 tumor enhances the already immunosuppressive environment in the lung and promotes the infiltration of IL-10- and TGF-β-expressing CD4+ and CD8+ T cells to the tumor site.
In the s.c. model, IL-10 mRNA expression was weak or undetectable in CD4+ and CD8+ T cells purified from inguinal lymph nodes of normal and tumor-bearing mice, but was detectable at low levels in both CD4+ and CD8+ T cells purified from solid tumors from mice injected s.c. with CT26 (Fig. 3⇑B). Although there was high constitutive expression of TGF-β mRNA in CD4+ T cells and to a lesser extent CD8+ T cells from the inguinal lymph nodes of naive mice, there was a small increase in TGF-β mRNA in CD8+ T cells from the inguinal lymph nodes of mice s.c. tumors.
Foxp3 expression is enhanced in CD4+ T cells within the CT26 tumor mass
Having demonstrated that T cells expressing IL-10 and TGF-β accumulate in the tumor during growth, we examined the possibility that natural Treg cells were recruited to the site of the tumor. The inguinal lymph nodes and solid tumors in the s.c. model and superficial cervical lymph nodes and lungs in the lung model were removed from BALB/c mice 3 and 14 days post-CT26 challenge, and cells were determined for expression of CD4 and CD25 using specific Abs and FACS analysis. Six to 8% of CD4+ T cells from the inguinal and superficial cervical lymph nodes expressed CD25, and there was no significant change 3 or 14 days after tumor challenge (data not shown). In contrast, the frequency of CD4+CD25+ infiltrating the tumor increased dramatically during tumor growth (s.c. tumor, day 3, 1.4%, and day 14, 10.3%; lung tumor, day 3, 1.4%, and day 14, 2.9%).
Because CD25 is expressed on activated T cells as well as Treg cells, we used an additional marker for Treg cells based on expression of Foxp3. CD4+ and CD8+ T cells were purified from inguinal lymph nodes and solid tumors in the s.c. model and superficial lymph nodes and lungs in the lung model, and RT-PCR was performed using primers specific for Foxp3. In the lung model, Foxp3 expression was dramatically enhanced in CD4+ T cell purified from lungs and the draining lymph nodes of tumor-bearing mice compared with the same tissues from naive mice (Fig. 3⇑A). Foxp3 expression was also detected in CD4+ T cells infiltrating the s.c. tumor mass (Fig. 3⇑B). Foxp3 expression in tumor-infiltrating CD4+ T cells was confirmed at the protein level by intracellular staining with an anti-Foxp3 Ab and FACS analysis (data not shown). Unlike the lung model, Foxp3 expression was not enhanced in the draining lymph node of mice with s.c. tumors. Foxp3 expression was not detected in CD8+ T cells from any of the tissues from naive or tumor-bearing mice.
A high frequency of IL-10-secreting CD4 and CD8 T cells and low frequency of IFN-γ-secreting T cells infiltrate the growing tumor, especially in the lungs
Having demonstrated the presence of IL-10-secreting and Foxp3+ T cells in the growing tumor, we used intracellular cytokine staining to examine the relative frequency of effector vs Treg cells recruited into the tumor mass and draining lymph nodes during tumor growth. Mice were challenged s.c. or i.v. with CT26 cells and tumors and lymph nodes were removed after 3 or 14 days; intracellular cytokine staining was performed for IL-10 and IFN-γ on T cells labeled for surface CD4 or CD8. A very high frequency (24–36%) of IL-10-secreting CD4+ and CD8+ T cells was detected in lungs of mice bearing CT26 metastases and in the tumor mass of mice injected s.c. with CT26 cells (Fig. 4⇓A). In contrast, 6–27% of CD4+ and CD8+ T cells secreted IFN-γ (Fig. 4⇓A). The frequency of IFN-γ-producing CD8+ T cells in the inguinal lymph nodes increased 3 days after s.c. tumor challenge, but declined again after 14 days (Fig. 4⇓B). Furthermore, the percentage of IFN-γ-secreting CD4+ and CD8+ T cells in cervical lymph nodes declined as lung metastases developed, while IL-10 CD4+-producing T cells increased (Fig. 4⇓B). These findings are consistent with enhanced IL-10 and TGF-β mRNA expression in T cells infiltrating the tumor and suggest that IL-10- and TGF-β-producing Treg cells are recruited or activated in the growing tumor, and this may suppress the development of IFN-γ-secreting effector T cells. The findings also demonstrate a greater defect in IFN-γ-secreting T cells in lymph node draining the lung compared with s.c tumor.
IL-10-secreting macrophages and DC infiltrate the growing tumor
It has been reported that innate IL-10, secreted by macrophages and DC, plays a critical role in the differentiation of inducible Treg cells from naive T cells in the periphery (4). Having demonstrated recruitment of IL-10-secreting T cells to the lungs during development of CT26 metastases, we therefore examined the possibility that this may involve recruitment and activation of IL-10-producing macrophages or DC in the lungs. Mice were challenged i.v. with CT26 cells and lungs and cervical lymph nodes were removed after 3 or 14 days; intracellular cytokine staining was performed for cells labeled for surface CD11c or F4/80. The development of CT26 tumors in the lungs was associated with significant recruitment of macrophages to the lungs; this was especially evident 14 days after tumor challenge (Fig. 5⇓). Although our staining method, using anti-CD11c FITC, did not allow us to discriminate between myeloid and plasmacytoid DC or NK cells expressing CD11c, we did find evidence of infiltrating CD11c+ cells in the lungs 14 days after tumor challenge. Furthermore, >40% of infiltrating macrophages and ∼75% of infiltrating CD11c+ cells secreted IL-10 on day 14 after tumor challenge. In contrast, the percentage of CD11c+ and F4/80+ cells was lower in the lymph node 3 and 14 days after tumor challenge when compared with naive control mice. Furthermore, a very low frequency of these cells secreted IL-10, and this did not change significantly after tumor challenge.
Defective CTL responses are restored by anti-IL-10 and TGF-β Abs
CTL play a key role in the antitumor immune response against the majority of tumors by directly killing tumor cells. As CT26 cells express MHC class I, rendering them susceptible to killing by CTL, the induction of functional CTL against CT26 was assessed during the course of tumor growth. Spleens were removed on days 7, 14, and 21 following tumor s.c. or i.v. challenge. CT26-specific CTL responses could not be detected in mice bearing CT26 lung metastases at any of the time points examined (data not shown). Following s.c. tumor challenge, CT26-specific CTL responses were undetectable 7 days post-s.c. tumor challenge, and although modest responses (22% lysis) were detected on day 14, this response was again undetectable on day 21 (Fig. 6⇓A).
We have shown that the growing tumors can suppress proliferative responses and IFN-γ to bystander Ags, and that infiltrating T cells express IL-10 and TGF-β. Therefore, we examined the possibility that these anti-inflammatory cytokines may suppress CTL responses during tumor progression. Spleens from mice injected s.c. with CT26 were removed on day 14 and spleen cells were cultured with medium alone, irradiated CT26 cells (100 Gy) alone, or with anti-IL-10 and/or anti-TGF-β Abs. Neutralization of IL-10 in the cultures resulted in a 7% increase in CT26 lysis, whereas neutralization of TGF-β resulted in a 13% increase in CT26-specific killing (Fig. 6⇑B). The combination of anti-IL-10 and anti-TGF-β marginally improved CT26 killing to 36%, or 18% greater than that observed in the absence of added Abs, indicating that these anti-inflammatory cytokines produced by T cells in the growing CT26 tumor cells or present in the tumor environment can contribute to the inhibition of CTL activity.
Depletion of CD4+ and CD8+ T cells has differential effects on tumor growth and survival in s.c. and lung models
Our data suggest that a high frequency of IL-10-secreting CD4+ and CD8+ T cells is recruited into the growing tumor and that a lower frequency of infiltrating T cells secretes IFN-γ; we examined the role of CD4+ and CD8+ T cells in protection vs suppression of antitumor immunity by in vivo depletion with Ab before tumor challenge. In the s.c. model, depletion of CD4+ T cells slowed CT26 tumor growth, although not significantly (Fig. 7⇓A). The survival of CD4+ T cell-depleted mice was also enhanced compared with that of nondepleted tumor-bearing control mice, with 50% remaining tumor free, whereas all nondepleted mice died as a result of tumor growth (Fig. 7⇓B). In the lung model, removal of CD4+ T cells resulted in a highly significant (p < 0.001) reduction in the number of CT26 lung metastases compared with nondepleted mice (Fig. 7⇓C).
In contrast to the regulatory role of CD4+, removal of CD8+ T cells had distinct effects on growth of CT26 in the lungs and at an s.c. site. In the s.c. model, depletion of CD8+ T cells resulted in a pronounced increase in tumor volume compared with nondepleted control mice (Fig. 7⇑D). CD8+ T cell-depleted mice all died within 21 days of tumor challenge, whereas nondepleted control mice died within 35 days (Fig. 7⇑E). In contrast, in the lung model, depletion of CD8+ T cells resulted in a highly significant reduction in the number of CT26 lung metastases compared with nondepleted mice (Fig. 7⇑F). This was a consistent finding in three independent experiments involving a total of 40 mice, with mean tumor counts of 51.2 ± 8.8 vs 10.95 ± 9.6 for control and CD8 T cell-depleted mice, respectively (p < 0.001). These findings suggest that, unlike the s.c. model, both CD4+ and CD8+ T cells in the lungs can suppress protective immune responses against CT26 tumor.
Influence of CD4+ or CD8+ T cell depletion on cytokine production and CTL responses
To determine the functional effect of CD4+ and CD8+ T cell depletions on responses of the reciprocal T cell population, and how this may relate to enhancement of protection or exacerbation of antitumor immunity, the frequency of CD4+- or CD8+-secreting IL-10 and IFN-γ was examined by intracellular cytokine staining.
In the s.c. model, depletion of CD4+ T cells resulted in a modest enhancement of the frequency of IFN-γ-secreting CD8+ T cells infiltrating the inguinal lymph node during tumor growth (representative experiment shown in Fig. 8⇓A; summary of three experiments: 5.1 ± 0.1 and 8.5 ± 0.2% IFN-γ-secreting CD8+ T cells in undepleted vs CD4-depleted mice, respectively). In the lung model, depletion of CD4+ T cells also enhanced the frequency of IFN-γ-producing CD8+ T cells in the cervical lymph nodes of mice with lung metastases (Fig. 8⇓A; summary data: 1.7 ± 0.1 and 4.4 ± 0.2% IFN-γ-secreting CD8+ T cells in undepleted vs CD4-depleted mice, respectively). We found an even more dramatic effect in the lung tissue, where depletion of CD4+ T cells substantially increased the percentage of IFN-γ-secreting CD8+ T cells (Fig. 8⇓A; summary data: 13 ± 2 and 38 ± 8% IFN-γ-secreting CD8+ T cells in undepleted vs CD4-depleted mice, respectively). Furthermore, the increase in frequency of IL-10-secreting CD8+ T cells in the lung during the development of CT26 lung metastases was completely reversed by depletion of CD4+ T cells (Fig. 8⇓A; summary data: naive, 18 ± 3%; undepleted, 26 ± 5%; CD4 depleted, 14 ± 2%). Therefore, removal of CD4+ T cells has a positive impact on effector cytokine production during CT26 tumor growth in the lung or at an s.c. site, which reflected the reduced tumor growth and enhanced survival following both routes of tumor challenge, which is consistent with a regulatory role for CD4+ T cells in tumor immunity.
We also examined immune responses during tumor progression in mice depleted of CD8+ T cells. In the s.c. model, depletion of CD8+ T cells abrogated CTL responses in tumor-bearing mice (data not shown). Furthermore, depletion of CD8+ T cells resulted in a reduced frequency of IFN-γ-secreting CD4+ T cells in inguinal lymph nodes of mice with s.c. tumors (Fig. 8⇑B; summary data: 6 ± 1 and 2.5 ± 0.5% IFN-γ-secreting CD4+ T cells in undepleted vs CD8-depleted mice, respectively). In contrast, in the lung model, depletion of CD8+ T cells did not affect the numbers of IFN-γ-secreting CD4+ T cells (Fig. 8⇑B), but significantly enhanced the percentage of non-T cells secreting IFN-γ (data not shown). Furthermore, an examination of T cells from the lungs revealed that depletion of CD8+ T cells enhanced IFN-γ-secreting CD4+ T cells (Fig. 8⇑B; summary data: 2.5 ± 0.5 and 6.5 ± 1% IFN-γ-secreting CD4+ T cells in undepleted vs CD8-depleted mice, respectively) and reduced the frequency of IL-10-secreting CD4+ T cells by almost 50% (Fig. 8⇑B; summary data: 25 ± 4 and 13 ± 3% IL-10-secreting CD4+ T cells in undepleted vs CD8-depleted mice, respectively). Furthermore, CD8+ T cells purified from the lungs of tumor-bearing mice exerted suppressor activity when cocultured with naive T cells in vitro (proliferative response of anti-CD3- and anti-CD28-stimulated naive T cells alone, 140,770 ± 1,867 vs 72,144 ± 6,795 for naive T cells alone or in presence of CD8+ T cells; p < 0.001). These findings suggest that CD4+ T cells can exert regulatory control on CD8+ T cells during tumor growth, but also suggested that in addition to a protective role of effector CD8+ T cells, evident in the s.c. model, CD8+ Treg cells that secrete IL-10 are recruited or activated by immunosuppressive environment in the lung, where they can have suppressed antitumor immunity.
Depletion of CD25+ cells reduced IL-10 and enhances IFN-γ-producing CD4+ and CD8+ T cells in the lungs and draining lymph nodes
Having demonstrated that depletion of CD4+ T cells enhanced antitumor immunity and that Foxp3 expression was enhanced in T cells within the tumor, we examined the possibility that depletion of CD4+CD25+ T cells may have a more dramatic effect than depleting the whole CD4+ T cell population. In the s.c. model, depletion of CD25+ T cells resulted in a significant decrease in tumor volume (Fig. 9⇓A) and a significant increase in survival of mice, with 75% remaining tumor free (Fig. 9⇓B). In the lung model, depletion of CD25+ T cells resulted in a significant reduction in the number of metastases compared with nondepeleted control mice (Fig. 7⇑C). These findings suggest that removal of Treg cells allows the generation of effective antitumor immunity.
We examined the effect of CD25+ T cell depletion on cytokine production by CD4+ and CD8+ T cells. In the lung model, IL-10 production from both CD4+ and CD8+ T cells in the lungs and cervical lymph nodes was significantly increased 14 days post-i.v. tumor challenge, but this was substantially reduced with the depletion of CD25+ T cells (Fig. 10⇓A). This suggests that CD4+ CD25+ T cells are the source of IL-10 or that they induce other T cells, such as CD8+ T cells to produce IL-10 during tumor growth. Conversely, the frequency of IFN-γ-producing CD4+ and CD8+ T cells in the superficial cervical lymph nodes and in the lungs was dramatically increased following depletion of CD4+CD25+ T cells (Fig. 10⇓A). In the s.c. model, depletion of CD25+ T cells resulted in a decrease in the frequency of IL-10-secreting CD4+ T cells and a modest increase in the percentage of CD8+ T cells secreting IFN-γ (Fig. 10⇓B). These findings demonstrate the potent effects of relatively small numbers of CD4+CD25+ cells in establishing the conditions conducive to tumor growth, especially in the immunosuppressive environment of the lung.
The significant new findings of this study are that CD4+ and CD8+ T cells producing IL-10 and TGF-β can contribute to suppression of antitumor immunity and that this may be influenced by the site of the tumor growth. We observed enhanced expression of Foxp3, TGF-β, and IL-10 in CD4+ T cells in growing CT26 tumors in mice, as well as a high frequency of infiltrating CD8+ T cells that secreted IL-10 and a relatively low frequency of CD4+ and CD8+ T cells secreting IFN-γ. Depletion of CD4+ or CD25+ T cells enhanced antitumor immune responses and prolonged survival, whereas depletion of CD8+ T cells enhanced growth of the s.c. tumor, but reduced the number of tumors in the lungs. This is the first description of a regulatory role for CD8+ T cells in antitumor immune responses in the lungs.
IFN-γ has a protective role in antitumor immunity, exerting antiproliferative, antiangiogenic, and proapoptotic effects on a wide variety of tumor cells (12, 13). In this study, a low frequency of IFN-γ-secreting CD4+ and CD8+ T cells infiltrated the tumor mass or was recruited to the draining lymph nodes. In the s.c. model, IFN-γ-secreting CD8+ were detected in the tumor-draining inguinal lymph nodes after 3 days, but their frequency declined after 7–14 days. CTL activity was transiently detected in mice with s.c. tumor, but was undetectable in mice with lung metastases. The failure to generate type 1 responses to the tumor was also reflected in suppression of immune responses to an unrelated Ag. Injection of tumor cells into OVA TCR transgenic mice inhibited T cell proliferation and IFN-γ production by draining inguinal lymph node cells in response to immunization with OVA. This suggests that the growing CT26 tumor created an immunosuppressive environment that suppresses T cell responses to both tumor and bystander Ags.
In an attempt to explain the immunosuppressive effects of the tumor, we found that CT26 cells expressed Cox-2 and TGF-β mRNA and also secreted TGF-β protein in vitro. Tumor-derived Cox-2 induces expression of Foxp3 and enhances the activity of CD4+CD25+ Treg cells (14) while TGF-β has potent anti-inflammatory properties and has been implicated in the induction and function of natural and inducible Treg cells (15, 16). An examination of T cells purified from the tumor mass revealed high expression of TGF-β and IL-10 mRNA, and intracellular cytokine staining revealed a very high frequency of IL-10-producing T cells, but also macrophages and DC. Innate IL-10 and TGF-β production, together with a semimature DC phenotype, are associated with the induction of IL-10-producing Treg cells in response to infection (17, 18) and to tumors (19). Therefore, TGF-β and other factors secreted by the growing tumor activate DC to secrete IL-10 and undergo partial maturation, resulting in selective induction of Treg cells. The direct immunosuppressive effects of anti-inflammatory cytokines secreted by macrophages and DC or their ability to drive the induction of IL-10 and TGF-β Treg cells may explain the defective IFN-γ-secreting CD4+ or CD8+ T cells and CTL responses and suppression of T cell responses to unrelated Ags. This is consistent with the demonstration that in vitro neutralization of anti-TGF-β and IL-10 enhanced CT26-specific CTL responses. The suppressive function of CD4+ T cells was also demonstrated in vivo, as CD4+ T cell depletion augmented IFN-γ production by CD8+ T cells and slowed the rate of tumor progression in the lungs and at an s.c. site.
There is growing evidence that natural CD4+CD25+ T cells may play a critical role in progression of a number of cancers (2, 9, 20). There is an increase in the number of CD4+CD25+ Treg cells in the peripheral blood of cancer patients, and this is believed to play a major part in the failure of many immunotherapies against tumors (21). In patients with non-small cell lung cancer, natural Treg cells from the tumors have been shown to inhibit proliferation of autologous, but not allogeneic T cells (1). In the present, we found a small increase in the number of CD4+CD25+ T cells, but a substantial increase in Foxp3 expression in CD4+ T cells at the tumor site in s.c. and lung tumor models. Foxp3 mRNA expression was also enhanced in CD4+ T cells from the draining lymph nodes of mice with CT26 lung metastases. In human ovarian cancer, CD4+CD25+Foxp3+ natural Treg cells are attracted by the chemokine CCL22, produced by tumor cells and macrophages (22), suggesting that Treg cells may be attracted by the local tumor microenvironment and once there, exert immunosuppressive effects on immune effector cells. We found that depletion of CD25+ T cells in the s.c. and the lung metastasis models resulted in significant decreases in tumor growth and a significant increase in survival and a dramatic enhancement of IFN-γ production from both CD4+ and CD8+ T cells. In contrast, depletion of CD25+ T cells almost completely eliminated IL-10 production by T cells in the lung and draining lymph nodes, suggesting that CD25+ T cells may be the predominant source of CD4+ T cell-derived IL-10 in tumor-bearing mice. However, we also observed an enhanced frequency of IL-10-producing CD8+ T cells in mice with CT26 tumors, which was significantly reduced following depletion of CD4+CD25+ T cells. Because removal of CD25+ T cells unmasks effective antitumor immune responses, these findings suggest that recruitment of CD4+CD25+Foxp3+ T cells to the site of the tumor may help to skew tumor-specific T cells responses from an effector to a regulatory subtype and thereby subvert protective antitumor immunity.
In addition to the role of CD4+ Treg cells, we also found evidence for the induction and suppressive function of IL-10- and TGF-β-expressing CD8+ Treg cells during development of tumors in the lung. The traditional view on the role of CD8+ T cells in antitumor immunity has been one of protection, through CTL activity and secretion of IFN-γ. Indeed, in the s.c. tumor model, we found that removal of CD8+ cells led to a substantial increase in CT26 tumor growth and a decrease in survival of the mice. Furthermore, CT26 tumor-specific CD8+ CTL responses were transiently generated in nondepleted mice during tumor growth and enhanced by neutralizing IL-10 and TGF-β. However, in the lung model, CD8+ CTL responses were undetectable, whereas there was a high frequency of IL-10-secreting CD8+ T cell in the lungs. Depletion of CD8+ T cells significantly reduced the number of lung metastases. Furthermore, CD8+ T cells purified from the lungs of mice with lung metastases suppressed proliferation of naive T cells in vitro. Collectively, our findings suggest that IL-10-secreting CD8+ Treg cells are induced during tumor growth and that these cells subvert antitumor immunity in the lung.
Our demonstration of a regulatory role for CD8+ T cells in the lung is also consistent with studies on non-small cell lung cancer, which demonstrated low concentrations of IFN-γ and elevated IL-10 and TGF-β at the tumor site (23) and high numbers of tumor-infiltrating CD8+ T cells, which did not correlate to patient survival (24). CD8+ Treg cells have also been reported in the lungs in an airways hyperreactivity model (25), and during mycoplasma infection (26). Furthermore, suppressive IL-10-producing CD8+ T cells have been found in the tumor environment of patients with ovarian cancer; their induction was independent of CD4+CD25+ T cell, but dependent on plasmacytoid DC (27). Our data support the suggestion that Treg cells in the lung may be induced under the influence of IL-10-secreting DC and macrophages, found at high numbers in the lungs during tumor growth. This is also consistent with reports that lung DC in mice exposed to respiratory Ags transiently secrete IL-10 and promote the induction of IL-10-secreting Treg cells (10). In contrast to the lung, peripheral sites are more responsive to inflammatory insults and can therefore allow optimal priming and activation of cells of the innate and adaptive immune systems. Indeed, one study reported that 10- to 100-fold more tumor-specific T cells are required to protect against an i.v. tumor challenge, compared with an equivalent s.c. tumor challenge (11). Taken together with the findings of our study, this suggests that the environment in the lung is not conducive to effective antitumor immune responses, and this may be explained in part by selective induction of CD4+ and CD8+ Treg cells during tumor development.
K. H. G. Mills is a cofounder, director and shareholder in Opsona Therapeutics Limited, a university campus company involved in the development of anti-inflammatory therapeutics.
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 a Science Foundation Ireland Principal Investigator Award (00/PI.I/B045; to K.H.G.M.).
↵2 A.G.J. and J.L. contributed equally to this work.
↵3 Current address: Centre for Clinical Vaccinology and Tropical Medicine, Nuffield Department of Medicine, University of Oxford, Churchill Hospital, Oxford OX3 7LJ, U.K.
↵4 Address correspondence and reprint requests to Dr. Kingston H. G. Mills, Immune Regulation Research Group, School of Biochemistry and Immunology, Trinity College, Dublin 2, Ireland. E-mail address:
↵5 Abbreviations used in this paper: Treg, regulatory T; Cox-2, cyclooxygenase-2; DC, dendritic cell; IP-10, IFN-γ-inducible protein-10.
- Received November 10, 2005.
- Accepted April 12, 2006.
- Copyright © 2006 by The American Association of Immunologists