Adoptive transfer of T cells expressing chimeric NKG2D (chNKG2D) receptors, a fusion of NKG2D and CD3ζ, can lead to long-term, tumor-free survival in a murine model of ovarian cancer. To determine the mechanisms of chNKG2D T cell antitumor efficacy, we analyzed how chNKG2D T cells altered the tumor microenvironment, including the tumor-infiltrating leukocyte populations. chNKG2D T cell treatment of mice bearing ID8 tumor cells increased the number and activation of NK cells and increased the activation of host CD8+ T cells within the tumor. Foxp3+ regulatory T cells at the tumor site decreased more than 300-fold after chNKG2D T cell treatment. Tumor-associated regulatory T cells expressed cell surface NKG2D ligands and were killed by chNKG2D T cells in a perforin-dependent manner. chNKG2D T cells also altered the function of myeloid cells at the tumor site, changing these cells from being immunosuppressive to enhancing T cell responses. Cells isolated from the tumor produced elevated amounts of IFN-γ, NO, and other proinflammatory cytokines after chNKG2D T cell treatment. ChNKG2D T cells required perforin, IFN-γ, and GM-CSF to induce a full response at the tumor site. In addition, transfer of chNKG2D T cells into mice bearing tumors that were established for 5 weeks led to long-term survival of the mice. Thus, chNKG2D T cells altered the ovarian tumor microenvironment to eliminate immunosuppressive cells and induce infiltration and activation of antitumor immune cells and production of inflammatory cytokines. This induction of an immune response likely contributes to chNKG2D T cells’ ability to eliminate established tumors.

Adoptive transfer of T cells has shown therapeutic potential in some types of cancer, such as melanoma and EBV-derived tumors (1, 2, 3). However, the response rate of T cell immunotherapy is relatively low in other types of cancer, including ovarian cancer (4). One contributing factor for the poor response to T cell immunotherapy may be due to the types of immune cells present at the tumor site. Leukocytes that may have antitumor activities can be found at the tumor site, including CD8+ T cells and NK cells (5). However, ovarian cancer also has many types of immunosuppressive cells in the tumor microenvironment, including myeloid-derived suppressor cells, vascular leukocytes, and regulatory T (Treg)3 cells (5, 6, 7). Elevated levels of immunosuppressive molecules such as PD-L1, IDO, PGE2, IL-10, and arginase may also inhibit the immune response against tumors (8, 9, 10). Therapies that alter leukocyte populations in the tumor microenvironment to prevent the function of immunosuppressive populations and induce the recruitment and activation of immune cells may lead to the development of long-lived antitumor immune responses and improve cancer therapy.

Therapeutic efforts to induce immune responses in ovarian cancer include administering inflammatory cytokines, such as IFN-γ or GM-CSF alone or in combination with chemotherapy, and these treatment strategies have shown some therapeutic success (11, 12, 13, 14). Additionally, treatments that inhibit immunosuppressive populations, such as inhibiting Treg cells using denileukin diftitox which consists of IL-2 fused to diptheria toxin, or using CTLA-4 blocking Abs, have increased antitumor immune responses in ovarian cancer patients (15, 16).

Transfer of tumor-reactive T cells has the potential to both kill tumor cells and activate the immune response through proinflammatory cytokine secretion. It has been shown that adoptive transfer of T cells expressing chimeric NKG2D (chNKG2D) receptors, which consist of the NKG2D receptor fused to the cytoplasmic region of the CD3ζ-chain, into mice bearing ovarian tumors established for 1 week leads to long-term, tumor-free survival (17). Complete antitumor efficacy required not only killing of the tumor cells by chNKG2D T cells through perforin, but also secretion of IFN-γ and GM-CSF by the transferred T cells (17, 18). Additionally, CD8+ T cells isolated from human ovarian cancer samples transduced with chNKG2D receptors secreted proinflammatory cytokines, including IFN-γ, GM-CSF, CCL3, and CCL5, when cultured with autologous tumor cells (19). This indicated that chNKG2D T cells may induce a proinflammatory immune response at the tumor site through the secretion of cytokines and release of tumor Ags. This current study determined how treatment with chNKG2D T cells altered the tumor-infiltrating leukocyte populations to promote antitumor immunity and which chNKG2D T cell effector mechanisms were required for these changes. The therapeutic efficacy of chNKG2D T cells against ovarian tumors established for 5 weeks in the mice was also determined.

Female C57BL/6 (B6) mice and B6-LY5.2/Cr (CD45.1+) mice were purchased from the National Cancer Institute (Frederick, MD). C57BL/6 mice, IFN-γ-deficient mice B6.129S7-Ifngtm1Ts/J (IFN-γ−/−), IFN-γ receptor 1 (IFN-γR1)-deficient mice B6.129S7-IfngR1tm1Agt/J, perforin-deficient mice C57BL/6-Prf1tm1Sdz/J, Fas ligand-deficient mice B6Smn.C3-Faslgld/J, and C57BL/6-TgTcraTcrb1100Mjb/J (OT-I) mice were purchased from The Jackson Laboratory. Mice deficient for both perforin and FasL were generated as previously described (18). GM-CSF-deficient mice on a C57BL/6 background were provided by Dr. Jeff Whitsett (University of Cincinnati, Cincinnati, OH). Mice used were 7 to 10 weeks of age at the start of each experiment. All animal work was performed in the Dartmouth Medical School Animal Facility (Lebanon, NH) in accordance with Institutional guidelines.

Mouse spleen cells were stimulated with Con A for 18 h (1 μg/ml) and transduced as previously described (19, 20). Two days after transduction, T cells were selected in medium containing G418 (0.5 mg/ml) and 25 U/ml recombinant human IL-2 for 3 days. Viable cells were isolated using Histopaque-1083 (Sigma-Aldrich) and expanded for 2 days without G418 (19, 20). At the time of transfer, wild-type NKG2D (wtNKG2D)- and chNKG2D-transduced splenocytes were 99% CD3+NK1.1, were a mixture of CD8+ (80–90%) and CD4+ (10–20%) cells, and had increased expression of NKG2D (21). ID8-GFP cells (2 × 106) were injected i.p. into B6 or IFN-γR1−/− mice. wtNKG2D or chNKG2D T cells (5 × 106) were transferred i.p. at 1 or 5 weeks after tumor injection. Mice were killed and peritoneal washes were performed using 10 ml PBS. RBC in the peritoneal washes were lysed with ACK lysis buffer and the number of cells was counted. For survival experiments, ID8-GFP cells (2 × 106) were injected i.p. into B6 mice and wtNKG2D or chNKG2D T cells (5 × 106) were transferred i.p. 5, 6, and 7 weeks or 5, 7, and 9 weeks after tumor injection. Mice were weighed at least once a week and were killed if they had gained more than 80% of their original body weight.

Cells isolated by peritoneal wash were incubated with anti-CD16/CD32 and mouse γ globulin (Jackson ImmunoResearch Laboratories), and stained with FITC-conjugated anti-CD3 (clone 145–2C11), anti-CD11c (clone N418), or anti-CD4 (clone GK1.5); PE-conjugated anti-CD3, anti-MHC class II (clone M5/114.15.2), or anti-CD19 (clone 1D3); allophycocyanin-conjugated anti-CD45.1 (clone A20), anti-F4/80 (clone BM8), or anti-NK1.1 (clone PK136); and biotin-conjugated anti-CD69 (clone H1.2F3) or anti-GR-1 (clone RB6–8C5, BD Pharmingen) with a secondary incubation with PE-Cy5.5-conjugated streptavidin. To detect cell surface NKG2D ligand expression, cells were stained with mouse NKG2D-human IgG1 fusion protein (R&D Systems) with an allophycocyanin-labeled goat anti-human IgG secondary (Jackson ImmunoResearch Laboratories). For detection of Foxp3, cells were stained with Abs for cell surface molecules and then cells were fixed with 1% paraformaldehyde, permeabilized with 0.1% saponin, and stained with PE-conjugated anti-Foxp3 (clone MF-14, BioLegend) or isotype control Abs. All Abs were purchased from eBioscience unless otherwise noted. Cell fluorescence was monitored using a FACSCalibur cytometer (BD Biosciences).

To test for in vitro killing of CD4+Foxp3+ and CD19+ cells, peritoneal wash cells were isolated from mice bearing ID8-GFP cells for 8–9 weeks. Peritoneal wash cells (2 × 106) were cultured with 1 × 106 wtNKG2D or chNKG2D T cells generated from a B6 mouse, or chNKG2D T cells generated from mice deficient in perforin, FasL, or both perforin and FasL. After 24 h, the percent CD4+Foxp3+ cells and CD19+ cells was evaluated by flow cytometry.

Three days after T cell transfer, tumor-bearing mice were killed and a peritoneal wash was performed. Peritoneal wash cells from naive mice were used as a control. F4/80+ cells were isolated from peritoneal washes using biotin-conjugated anti-F4/80 Abs and magnetic bead selection (Miltenyi Biotec) according to the manufacturer’s instructions. CD8+ OT-I T cells were magnetically purified from spleen and lymph node cells using FITC-conjugated anti-CD8β Abs. OT-I T cells (105) were CFSE-labeled and cultured with F4/80+ cells (2 × 105) and OVA257–264 peptide (10−10 M). Proliferation of OT-I T cells was determined by flow cytometry after 4 days of culture.

Peritoneal wash cells (106) from tumor-bearing mice treated with wtNKG2D or chNKG2D T cells were cultured in 48-well plates in complete medium. Twenty-four hour cell-free conditioned media were assayed for IFN-γ by ELISA using mouse Duoset ELISA kits (R&D Systems) and for NO using Griess’s reagent for nitrite (Sigma-Aldrich) according to manufacturers’ protocols. Seventy-two-hour conditioned media were assayed for additional cytokines using multiplex analysis (Bio-Rad) by the Immune Monitoring Laboratory of the Norris Cotton Cancer Center (Lebanon, NH). For intracellular staining, peritoneal wash cells (106) or spleen cells (2.5 × 106) were cultured in complete medium for 24 h. Brefeldin A (10 μg/ml; Sigma-Aldrich) was added during the last 5 h of culture. Cells were incubated with FcR block and stained with FITC-conjugated anti-CD8β (clone CT-CD8β), allophycocyanin-conjugated anti-NK1.1 (clone PK136), or allophycocyanin-conjugated anti-CD45.1 (clone A20) and biotin-conjugated anti-CD3 (clone eBio500A2), with a PE-Cy5.5 conjugated-streptavidin secondary. Cells were fixed with 1% paraformaldehyde, permeabilized with 0.1% saponin, and stained with PE-conjugated anti-IFN-γ (clone XMG12), or PE-conjugated anti-rat IgG1 isotype control.

Differences between groups were analyzed using the Student’s t test or ANOVA using Prism software (GraphPad Software). For survival studies, Kaplan-Meier curves were plotted and analyzed using the log rank test and Prism software. Values of p < 0.05 were considered significant.

To determine how treatment with chNKG2D T cells alters the tumor microenvironment, ID8 tumor cells were injected into B6 mice and mice were treated with chNKG2D or wtNKG2D T cells that were congenically marked with Ly5.1+ after 1 week. Multiple host leukocyte populations were altered at the tumor site after chNKG2D T cell injection. The number of NK cells increased after chNKG2D T cell injection, with the peak response 3 days after T cell transfer (Fig. 1,A). An increased percentage of NK cells expressed CD69, indicating the infiltrating NK cells were more activated in mice treated with chNKG2D T cells (Fig. 1,B). The number of host Ly5.1 CD8+ T cells in the peritoneal wash did not change after chNKG2D T cell treatment, however the host CD8+ T cells were more activated, as shown by an increased percentage of CD8+CD69+ cells (Fig. 1,C). ChNKG2D T cell treatment also increased the number of GR1+F4/80 cells, likely neutrophils, in the peritoneal wash (Fig. 1,D). There was a significant difference in CD19+ B cells (Fig. 1,E). chNKG2D T cells resulted in rapid decrease in the number of ID8 tumor cells just 1 day after T cell injection and it has previously been shown that chNKG2D T cells require perforin to directly kill ID8 tumor cells (Fig. 1 F) (17). These data indicate that treatment with chNKG2D T cells induced a proinflammatory immune response at the tumor site, resulting in an infiltration and activation of immune cells that can decrease tumor burden, including NK cells and CD8+ T cells.

FIGURE 1.

Treatment with chNKG2D T cells induced activation of the tumor-infiltrating leukocyte populations. ID8-GFP cells were injected and mice were treated with Ly5.1+ wtNKG2D (▴) or chNKG2D (•) T cells i.p. 7 days later. The number of NK1.1+CD3 NK cells (A), CD69+NK1.1+CD3-activated NK cells (B), Ly5.1 CD8+ CD69+ T cells (C), F480 GR1+ neutrophils (D), CD19+ CD3 B cells (E), or GFP+ ID8 tumor cells (F) was determined before T cell injection (day 0), and 1, 3, and 7 days after T cell injection. The average of each group (n = 4) is shown. Treatment with chNKG2D T cells significantly changed the number of the different cell populations compared with control-treated mice (*, p < 0.05). Data are representative of at least five independent experiments.

FIGURE 1.

Treatment with chNKG2D T cells induced activation of the tumor-infiltrating leukocyte populations. ID8-GFP cells were injected and mice were treated with Ly5.1+ wtNKG2D (▴) or chNKG2D (•) T cells i.p. 7 days later. The number of NK1.1+CD3 NK cells (A), CD69+NK1.1+CD3-activated NK cells (B), Ly5.1 CD8+ CD69+ T cells (C), F480 GR1+ neutrophils (D), CD19+ CD3 B cells (E), or GFP+ ID8 tumor cells (F) was determined before T cell injection (day 0), and 1, 3, and 7 days after T cell injection. The average of each group (n = 4) is shown. Treatment with chNKG2D T cells significantly changed the number of the different cell populations compared with control-treated mice (*, p < 0.05). Data are representative of at least five independent experiments.

Close modal

CD4+ Treg cells can be found in ovarian tumors and have been previously shown to be present in mice bearing established ID8 tumors (8 weeks) (5, 7, 22). The effect of chNKG2D T cell treatment on the number of host Foxp3+CD4+ T cells in the tumor was determined. Mice with established tumors (7–9 weeks) were treated with wtNKG2D or chNKG2D T cells and the number of Foxp3+ T cells at the tumor site was determined 3 days after T cell transfer. Treatment with chNKG2D T cells decreased the number of Foxp3+CD4+ T cells at the tumor site by more than 300-fold compared with mice treated with wtNKG2D T cells or that received no treatment (Fig. 2,A). CD8+Foxp3+ T cells were not detected in the peritoneal washes of the tumor-bearing mice (data not shown). Analyses with a soluble NKG2D receptor showed that Foxp3+CD4+ T cells from the tumor site expressed low levels of NKG2D ligands in all mice tested, indicating that these cells may be direct targets of chNKG2D T cells (Fig. 2,B). Although Foxp3+CD4+ T cells at the tumor site were eliminated, the number of Foxp3+CD4+ T cells in the spleen was not altered after i.p. injection of chNKG2D T cells, suggesting that the depletion of regulatory cells was a local effect (Fig. 2,C). Foxp3+CD4+ T cells in the spleen of tumor-bearing mice or naive mice did not express NKG2D ligands on the cell surface in all mice tested (Fig. 2 D and data not shown), indicating that Treg cells increase expression of NKG2D ligands within the tumor environment.

FIGURE 2.

chNKG2D T cell treatment decreases Foxp3+ CD4 T cells at the tumor site. ID8-GFP cells were injected i.p. into mice. After 8 weeks, mice received no treatment (−) or were treated with Ly5.1+ wtNKG2D or chNKG2D T cells i.p. Three days after wtNKG2D or chNKG2D T cell injection, the number of Ly5.1 Foxp3+CD4+ T cells was determined in the peritoneal wash (A) or the spleen (C). The average of each group + SD (n = 4) is shown. Treatment with chNKG2D T cells significantly decreased the number of the Foxp3+CD4+ T cells in the peritoneal cavity compared with control-treated mice (***, p < 0.001). B and D, The expression of NKG2D ligands on Foxp3+CD4+ T cells was determined by staining peritoneal wash cells (B) or spleen cells (D) with sNKG2D-hIgG (filled) or human IgG isotype control (gray) and with anti-human IgG secondary Abs. The FACS histograms show NKG2D ligand expression on Ly5.1 CD3+CD4+Foxp3+ cells and are representative of three separate experiments, with n = 4 for each experiment. E and F, ID8-GFP cells were injected i.p. into mice. After 8 weeks, peritoneal wash cells were cultured with medium, wtNKG2D T cells (WT), chNKG2D T cells (CH), or chNKG2D T cells deficient in perforin (P CH), FasL (F CH), or perforin and FasL (P/F CH). After 24 h, the percent of Foxp3+CD4+ T cells (E) or CD19+CD3 cells (F) was determined. Data are representative of two independent experiments. Culture with chNKG2D T cells significantly decreased the percent of Foxp3+CD4+ T cells (*, p < 0.001).

FIGURE 2.

chNKG2D T cell treatment decreases Foxp3+ CD4 T cells at the tumor site. ID8-GFP cells were injected i.p. into mice. After 8 weeks, mice received no treatment (−) or were treated with Ly5.1+ wtNKG2D or chNKG2D T cells i.p. Three days after wtNKG2D or chNKG2D T cell injection, the number of Ly5.1 Foxp3+CD4+ T cells was determined in the peritoneal wash (A) or the spleen (C). The average of each group + SD (n = 4) is shown. Treatment with chNKG2D T cells significantly decreased the number of the Foxp3+CD4+ T cells in the peritoneal cavity compared with control-treated mice (***, p < 0.001). B and D, The expression of NKG2D ligands on Foxp3+CD4+ T cells was determined by staining peritoneal wash cells (B) or spleen cells (D) with sNKG2D-hIgG (filled) or human IgG isotype control (gray) and with anti-human IgG secondary Abs. The FACS histograms show NKG2D ligand expression on Ly5.1 CD3+CD4+Foxp3+ cells and are representative of three separate experiments, with n = 4 for each experiment. E and F, ID8-GFP cells were injected i.p. into mice. After 8 weeks, peritoneal wash cells were cultured with medium, wtNKG2D T cells (WT), chNKG2D T cells (CH), or chNKG2D T cells deficient in perforin (P CH), FasL (F CH), or perforin and FasL (P/F CH). After 24 h, the percent of Foxp3+CD4+ T cells (E) or CD19+CD3 cells (F) was determined. Data are representative of two independent experiments. Culture with chNKG2D T cells significantly decreased the percent of Foxp3+CD4+ T cells (*, p < 0.001).

Close modal

To determine whether chNKG2D T cells were directly killing Foxp3+CD4+ T cells, peritoneal wash cells from mice with established tumors (8–9 weeks) were cultured with wtNKG2D or chNKG2D T cells for 24 h (Fig. 2 E). Culture with chNKG2D T cells resulted in much fewer live Foxp3+CD4+ T cells compared with those cultured with wtNKG2D T cells or medium alone. ChNKG2D T cells required expression of perforin but not FasL to remove the Foxp3+CD4+ T cells because there was no loss of Foxp3+CD4+ T cells when they were cultured with chNKG2D T cells deficient in perforin, or with both perforin and FasL. These data are consistent with the idea that chNKG2D T cells directly kill tumor-associated Foxp3+CD4+ T cells using perforin. However, this reduction in vitro was not as large as observed in vivo, suggesting that other mechanisms may also be involved to eliminate Foxp3+CD4+ T cells in vivo.

In addition to Treg cells, other populations present in the peritoneal wash of mice with large ascites at 8–9 weeks included ID8 tumor cells (10–15%) and leukocytes, including CD11c+, F4/80+GR1+, F4/80GR1+, and CD19+ B cells. The number of CD19+ cells also decreased in vivo when mice were treated with chNKG2D T cells. However, the percent of CD19+ cells did not decrease when cultured with chNKG2D T cells in vitro (Fig. 2 F). Thus, the change in B cell numbers in vivo is likely not due to direct killing by chNKG2D T cells.

Treatment with chNKG2D T cells also induced changes in the APCs at the tumor site. There was an increase in CD11c+ MHC class IIhigh dendritic cells (DCs) (Fig. 3,A). chNKG2D T cells also altered the phenotype of the tumor-infiltrating macrophages. Macrophages in wtNKG2D T cell-treated mice expressed high levels of F4/80 and had low MHC class II expression. After chNKG2D T cell treatment, the macrophages decreased F4/80 expression and increased GR1 and MHC class II expression (Fig. 3,B). Macrophages from both wtNKG2D and chNKG2D T cell-treated mice expressed CD11b. Although some previous studies have shown that tumor associated, GR1-expressing macrophages can have an immunosuppressive phenotype, other studies have shown that this is an activated cell phenotype (10, 23). To investigate whether chNKG2D T cells altered the function of macrophages to become immunostimulatory or immunosuppressive, F4/80+ cells from mice treated with wtNKG2D or chNKG2D T cells were isolated from peritoneal washes and cultured with CFSE-labeled OT-I T cells and OVA peptide. OT-I T cells cultured with F4/80+ cells isolated from chNKG2D T cell-treated mice proliferated more than when cultured with F4/80+ cells from wtNKG2D T cell-treated mice or from naive mice (Fig. 3, C and D). The OT-I T cells cultured with F4/80+ cells from wtNKG2D T cell-treated mice had little proliferation, which is consistent with an immunosuppressive function of these tumor-associated myeloid cells. Thus, chNKG2D T cell treatment of tumor-bearing mice induced an activation of the tumor-infiltrating APCs so that they stimulated rather than inhibited Ag-specific T cells.

FIGURE 3.

APCs at the tumor site are activated after chNKG2D T cell injection. ID8-GFP cells were injected i.p. into mice. After 7 days, mice were treated with wtNKG2D (▴) or chNKG2D (•) T cells i.p. The number of CD11c+MHC class IIhigh cells (A) or the number of F4/80+GR1+ cells (B) was determined before T cell injection (day 0), and 1, 3, and 7 days after T cell injection. C, F4/80+ cells were isolated from the peritoneal washes 3 days after injection of wtNKG2D (white bar) or chNKG2D (black bar) T cells or from naive mice (gray bar) and were cultured with CFSE-labeled OT-I T cells and OVA peptide. OT-I T cell proliferation was measured after 4 days of culture. The average of each group + SD (n = 4) (C) and representative histograms of OT-I T cell proliferation (D) are shown. Treatment with chNKG2D T cells significantly changed the number of the different cell populations compared with control-treated mice (*, p < 0.05). Data are representative of at least two separate experiments.

FIGURE 3.

APCs at the tumor site are activated after chNKG2D T cell injection. ID8-GFP cells were injected i.p. into mice. After 7 days, mice were treated with wtNKG2D (▴) or chNKG2D (•) T cells i.p. The number of CD11c+MHC class IIhigh cells (A) or the number of F4/80+GR1+ cells (B) was determined before T cell injection (day 0), and 1, 3, and 7 days after T cell injection. C, F4/80+ cells were isolated from the peritoneal washes 3 days after injection of wtNKG2D (white bar) or chNKG2D (black bar) T cells or from naive mice (gray bar) and were cultured with CFSE-labeled OT-I T cells and OVA peptide. OT-I T cell proliferation was measured after 4 days of culture. The average of each group + SD (n = 4) (C) and representative histograms of OT-I T cell proliferation (D) are shown. Treatment with chNKG2D T cells significantly changed the number of the different cell populations compared with control-treated mice (*, p < 0.05). Data are representative of at least two separate experiments.

Close modal

IFN-γ has multiple antitumor properties and previous studies have shown that chNKG2D T cells secrete IFN-γ when cultured with ID8 tumor cells in vitro and that IFN-γ is an important effector molecule for antitumor efficacy in vivo (17, 19). To determine whether chNKG2D T cell treatment induced IFN-γ production at the tumor site, cytokine production was measured in peritoneal cells isolated from tumor-bearing mice treated with wtNKG2D or chNKG2D T cells. Peritoneal wash cells from chNKG2D T cell-treated mice secreted more IFN-γ compared with cells isolated from wtNKG2D T cell-treated mice (Fig. 4,A). Additionally, the peritoneal wash cells from chNKG2D T cell-treated mice secreted more NO (Fig. 4,A), possibly due to the increased IFN-γ production as this cytokine can induce macrophage activation. The significantly increased cytokine response was observed as early as 1 day after chNKG2D T cell injection with a peak occurring 7 days after T cell injection. Intracellular staining for IFN-γ was performed to determine which cells at the tumor site were producing IFN-γ. Ly5.1+ chNKG2D T cells secreted significant levels of IFN-γ 1 and 3 days after injection. However, after 7 days, the transferred T cells were no longer found in the peritoneal cavity (Fig. 4 B). In addition to the transferred T cells, host NK cells, CD4+, and CD8+ T cells produced significantly more IFN-γ in chNKG2D T cell-treated mice compared with mice treated with wtNKG2D T cells. Host cell production of IFN-γ began 1 day after chNKG2D T cell injection and this host immune response continued to increase for 7 days. This indicated that chNKG2D T cells induced a host immune response at the tumor site.

FIGURE 4.

Tumor-infiltrating cells from chNKG2D T cell-treated mice have increased IFN-γ and NO secretion. A, Peritoneal wash cells from wtNKG2D (□) or chNKG2D (▪) T cell-treated tumor-bearing mice from each time point were cultured in medium for 24 h. Cell-free supernatants were assayed for IFN-γ or NO. B, Intracellular staining was performed on peritoneal wash cells cultured in medium for 24 h. Cells were evaluated for IFN-γ production and were gated on either Ly5.1+CD3+, CD8+CD3+, CD4+CD3+, or NK1.1+CD3 as indicated. The average of each group + SD (n = 4) is shown. Treatment with chNKG2D T cells significantly increased IFN-γ and NO secretion compared with control-treated mice (*, p < 0.05; **, p < 0.01). Data are representative of at least five separate experiments.

FIGURE 4.

Tumor-infiltrating cells from chNKG2D T cell-treated mice have increased IFN-γ and NO secretion. A, Peritoneal wash cells from wtNKG2D (□) or chNKG2D (▪) T cell-treated tumor-bearing mice from each time point were cultured in medium for 24 h. Cell-free supernatants were assayed for IFN-γ or NO. B, Intracellular staining was performed on peritoneal wash cells cultured in medium for 24 h. Cells were evaluated for IFN-γ production and were gated on either Ly5.1+CD3+, CD8+CD3+, CD4+CD3+, or NK1.1+CD3 as indicated. The average of each group + SD (n = 4) is shown. Treatment with chNKG2D T cells significantly increased IFN-γ and NO secretion compared with control-treated mice (*, p < 0.05; **, p < 0.01). Data are representative of at least five separate experiments.

Close modal

Expression of perforin, GM-CSF, and IFN-γ by chNKG2D T cells is required for complete reduction in tumor burden (17, 18). As these effector molecules likely contribute to the activation of the host immune response, the requirement of these molecules for chNKG2D T cell-induced changes at the tumor site was determined. Tumor-bearing mice were treated with wtNKG2D or chNKG2D T cells derived from B6 mice or mice deficient in GM-CSF, IFN-γ, or perforin. Three days after T cell injection, the tumor-infiltrating populations were measured (Table I). chNKG2D T cell-derived GM-CSF was involved in inducing a significant increase in NK cells and NK cell activation, CD8+ T cell activation, DCs, and neutrophils, and a decrease in B cells, because mice treated with chNKG2D T cells deficient in GM-CSF did not have these changes in these leukocyte populations.

Table I.

Leukocyte and tumor cell numbers at the tumor site 3 days after T cell injectiona

WTCHGM-CSF−/− CHIFN-γ−/− CHIFN-γR1−/−Pfp−/− CH
NK1.1+CD3 12a 80b 46bc 13c 12c 62b 
NK1.1+CD69+ 55b 25bc 7c 7c 33b 
CD8+CD69+ 11b 6c 3c 5c 10b 
CD11c+MHC class II+ 13 208b 94bc 21c 59c 179b 
F480+GR1+ 49 123b 91b 44c 26c 106b 
F480GR1+ 65 381b 126bc 73c 73c 235b 
CD4+CD3+ 120 53b 72b 76b 13b 78b 
CD19+ 291 56b 242c 193bc 58b 171bc 
ID8-GFP 10 2b 3b 6c 6c 8c 
WTCHGM-CSF−/− CHIFN-γ−/− CHIFN-γR1−/−Pfp−/− CH
NK1.1+CD3 12a 80b 46bc 13c 12c 62b 
NK1.1+CD69+ 55b 25bc 7c 7c 33b 
CD8+CD69+ 11b 6c 3c 5c 10b 
CD11c+MHC class II+ 13 208b 94bc 21c 59c 179b 
F480+GR1+ 49 123b 91b 44c 26c 106b 
F480GR1+ 65 381b 126bc 73c 73c 235b 
CD4+CD3+ 120 53b 72b 76b 13b 78b 
CD19+ 291 56b 242c 193bc 58b 171bc 
ID8-GFP 10 2b 3b 6c 6c 8c 
a

Cell numbers in the peritoneal wash × 104, averages of four mice per group. Data are representative of two to five experiments. WT, wtNKG2D T cell-treated mice; CH, chNKG2D T cell-treated mice; Pfp, perforin.

b

Significantly different from wtNKG2D T cell treated mice, p < 0.05.

c

Significantly different from chNKG2D T cell-treated mice, p < 0.05.

ChNKG2D T cell-derived IFN-γ was required for a significant increase in NK cells and in NK cell activation, CD8+ T cell activation, neutrophils, and activation of DCs and macrophages. ChNKG2D T cell-derived IFN-γ was also required for the alteration in B cells because mice treated with chNKG2D T cells deficient in IFN-γ did not have this decrease. To determine whether the chNKG2D T cell-derived IFN-γ had a direct effect on host cells, ID8-GFP cells were injected into mice deficient in IFN-γR1 and the mice were treated with wtNKG2D or chNKG2D T cells. Similar to mice treated with IFN-γ-deficient chNKG2D T cells, mice deficient in IFN-γR1 treated with chNKG2D T cells did not have an increase in NK cells and NK cell activation, CD8+ T cell activation, neutrophils, or maturation and activation of DCs and macrophages. Thus, IFN-γ derived from chNKG2D T cells acts on cells of the host to induce changes in the tumor microenvironment. chNKG2D T cell-derived perforin was required for the decrease in CD19+ cells, while all other changes in leukocytes were observed when mice were treated with perforin-deficient chNKG2D T cells. However, CD19+ cells did not express NKG2D ligands and were shown to not be direct targets of chNKG2D T cells (Fig. 2 F). chNKG2D T cell-derived perforin was also required for the decrease in ID8 tumor cells.

The requirement of chNKG2D T cell-derived molecules for the induction of cytokine secretion at the tumor site was also determined. Peritoneal cells had increased secretion of IFN-γ after injection of chNKG2D T cells, and this increase was significantly diminished when chNKG2D T cells lacked GM-CSF, IFN-γ, or perforin (Fig. 5,A). Although all three effector molecules from chNKG2D T cells were required for the induction of a host IFN-γ response, only chNKG2D T cell-derived IFN-γ was required for production of NO at 3 and 7 days after T cell injection. Mice deficient in IFN-γR1 also did not have an increase in IFN-γ or NO secretion after treatment with chNKG2D T cells, demonstrating that host cells need to be responsive to IFN-γ to increase production of these cytokines at the tumor site (Fig. 5,B). Intracellular staining was performed to determine which of the IFN-γ-producing cells the chNKG2D T cell-derived molecules affected. There were lower percentages of NK cells and CD4+ T cells secreting IFN-γ when chNKG2D T cells were deficient in GM-CSF, IFN-γ, or perforin, indicating that all three molecules were involved in the induction of NK cell and CD4+ T cell production of IFN-γ (Fig. 5 C). Although there was a reduced percentage of CD8+ T cells producing IFN-γ when chNKG2D T cells were deficient in GM-CSF, IFN-γ, or perforin, this reduction was not significant.

FIGURE 5.

chNKG2D T cell-derived GM-CSF, IFN-γ, and perforin are required for the increase in IFN-γ and NO production in tumor-bearing mice. A, Peritoneal wash cells from tumor-bearing mice treated with B6-derived wtNKG2D (WT; □) or chNKG2D T cells (CH; ▪), or chNKG2D T cells deficient in GM-CSF (), IFN-γ (), or perforin (Pfp; ) were cultured in medium for 24 h. Cell-free supernatants were assayed for IFN-γ or NO. B, Three days after T cell injection, peritoneal wash cells from B6 mice treated with wtNKG2D (□) or chNKG2D T cells (▪), or IFN-γR1−/− mice treated with wtNKG2D () or chNKG2D T cells () were cultured in medium for 24 h. Cell-free supernatants were assayed for IFN-γ or NO. C, Seven days after T cell transfer, intracellular staining was performed on peritoneal wash cells cultured in medium for 24 h. Cells were evaluated for IFN-γ production and were gated on either CD8+CD3+, CD4+CD3+, or NK1.1+CD3 as indicated. The average of each group + SD (n = 4) is shown. Treatment with chNKG2D T cells significantly increased IFN-γ and NO secretion compared with control-treated mice (*, p < 0.05) and mice treated with chNKG2D T cells deficient in effector molecules produced significantly less IFN-γ and NO compared with chNKG2D T cell-treated mice (§, p < 0.05). Data are representative of at least two separate experiments.

FIGURE 5.

chNKG2D T cell-derived GM-CSF, IFN-γ, and perforin are required for the increase in IFN-γ and NO production in tumor-bearing mice. A, Peritoneal wash cells from tumor-bearing mice treated with B6-derived wtNKG2D (WT; □) or chNKG2D T cells (CH; ▪), or chNKG2D T cells deficient in GM-CSF (), IFN-γ (), or perforin (Pfp; ) were cultured in medium for 24 h. Cell-free supernatants were assayed for IFN-γ or NO. B, Three days after T cell injection, peritoneal wash cells from B6 mice treated with wtNKG2D (□) or chNKG2D T cells (▪), or IFN-γR1−/− mice treated with wtNKG2D () or chNKG2D T cells () were cultured in medium for 24 h. Cell-free supernatants were assayed for IFN-γ or NO. C, Seven days after T cell transfer, intracellular staining was performed on peritoneal wash cells cultured in medium for 24 h. Cells were evaluated for IFN-γ production and were gated on either CD8+CD3+, CD4+CD3+, or NK1.1+CD3 as indicated. The average of each group + SD (n = 4) is shown. Treatment with chNKG2D T cells significantly increased IFN-γ and NO secretion compared with control-treated mice (*, p < 0.05) and mice treated with chNKG2D T cells deficient in effector molecules produced significantly less IFN-γ and NO compared with chNKG2D T cell-treated mice (§, p < 0.05). Data are representative of at least two separate experiments.

Close modal

The secretion of additional cytokines by peritoneal cells was altered after treatment with chNKG2D T cells. Many proinflammatory cytokines were increased after chNKG2D T cell treatment, including IL-1, IL-2, IL-12, CCL2, CCL3, and CCL5 (Table II). Anti-inflammatory cytokines IL-9 and IL-10 were decreased in chNKG2D T cell-treated mice. The induction of these cytokines also required chNKG2D T cell-derived molecules. chNKG2D T cell-derived IFN-γ was essential for the production of IL-1, IL-2, IL-12, G-CSF, GM-CSF, CCL3, and CCL5, and also for decreasing the amount of IL-6. GM-CSF from chNKG2D T cells had a similar role in cytokine induction and additionally increased CCL2 and decreased TNF-α production. chNKG2D T cell-derived perforin was necessary for the increase in IL-1, G-CSF, CCL2, and CCL3. Cytokines that did not change after chNKG2D T cell treatment included IL-3, IL-5, IL-13, IL-17, KC, and CCL4, and cytokines not found at the tumor site included IL-4 and eotaxin (data not shown). Together, these data demonstrate that treatment of tumor-bearing mice with chNKG2D T cells induced a proinflammatory host immune response at the tumor site, decreased immunosuppressive regulatory cells and increased cell populations with antitumor capabilities and the local production of proinflammatory cytokines.

Table II.

Cytokine secretion by peritoneal cells 3 days after T cell injectiona

WTCHGM-CSF−/− CHIFN-γ−/− CHPfp−/− CH
IL-1α 156c 13d 4d 9d 
IL-1β 18 41c 15d 14d 9d 
IL-2 323 754c 561cd 427d 865c 
IL-6 2216b 87c 902c 523cd 52c 
IL-9 38 13c 20 20 10c 
IL-10 91 27c 43c 40c 37c 
IL-12 p40 14 95c 34d 31d 105c 
IL-12 p70 180 55c 66c 108cd 32c 
G-CSF 24 148c 14d 18d 12d 
GM-CSF 37 150c 45d 50d 141c 
CCL2 3400 9931c 937cd 10123c 2221d 
CCL3 17 33c 12d 18d 4d 
CCL5 128 348c 171cd 287c 254c 
TNF-α 53 4c 23cd 4c 4c 
WTCHGM-CSF−/− CHIFN-γ−/− CHPfp−/− CH
IL-1α 156c 13d 4d 9d 
IL-1β 18 41c 15d 14d 9d 
IL-2 323 754c 561cd 427d 865c 
IL-6 2216b 87c 902c 523cd 52c 
IL-9 38 13c 20 20 10c 
IL-10 91 27c 43c 40c 37c 
IL-12 p40 14 95c 34d 31d 105c 
IL-12 p70 180 55c 66c 108cd 32c 
G-CSF 24 148c 14d 18d 12d 
GM-CSF 37 150c 45d 50d 141c 
CCL2 3400 9931c 937cd 10123c 2221d 
CCL3 17 33c 12d 18d 4d 
CCL5 128 348c 171cd 287c 254c 
TNF-α 53 4c 23cd 4c 4c 
a

Values shown are pg/ml, averages of three mice per group. WT, wtNKG2D T cell-treated mice; CH, chNKG2D T cell-treated mice; Pfp, perforin.

b

Values were above detection limit of assay.

c

Significantly different from wtNKG2D T cell-treated mice, p < 0.05.

d

Significantly different from chNKG2D T cell-treated mice, p < 0.05.

Previous studies showed that treatment of ID8-tumor-bearing mice with chNKG2D T cells 1, 2, and 3 weeks after tumor cell injection lead to long-term, tumor-free survival in 100% of the mice (17). However, the efficacy of chNKG2D T cells treating mice with established solid tumors had not been tested before this study. First, it was determined whether chNKG2D T cells induced changes in the tumor microenvironment in mice bearing 5-week tumors, which is a time when many solid tumors have been established on the peritoneal wall. Compared with treating 1 week after tumor cell injection, chNKG2D T cell treatment induced similar changes in tumor-infiltrating populations in mice bearing tumors for 5 weeks, including an increase in the number of activated NK cells, CD8+ T cells, and macrophages; an increase in MHC class II + DCs; an increase in neutrophils; and a decrease in B cells and in ID8-GFP tumor cells (Fig. 6). Similar changes were also seen after treatment with chNKG2D T cells in mice with tumors established for 7–9 weeks (data not shown). There was also an increase in IFN-γ and NO secretion from peritoneal wash cells in chNKG2D T cell-treated mice. This indicated that chNKG2D T cells led to the activation of the tumor-associated leukocytes even in mice bearing tumors established for 5 weeks.

FIGURE 6.

Treatment of advanced tumors with chNKG2D T cells induced activation of the tumor-infiltrating leukocyte populations. ID8-GFP cells were injected and mice were treated with wtNKG2D (▴) or chNKG2D (•) T cells i.p. 5 weeks later. A, The number of NK1.1+CD3 NK cells, CD69+NK1.1+CD3-activated NK cells, Ly5.1CD8+ CD69+ T cells, CD19+ CD3 B cells, CD11c+MHC class IIhigh cells, F4/80+GR1+ cells, F480 GR1+ neutrophils, or GFP+ ID8 tumor cells was determined before T cell injection (day 0), and 1 and 3 days after T cell injection. B, Peritoneal wash cells from wtNKG2D (□) or chNKG2D T cell (▪) -treated tumor-bearing mice from each time point were cultured in medium for 24 h. Cell-free supernatants were assayed for IFN-γ or NO. The average of each group (n = 4) is shown. Treatment with chNKG2D T cells significantly changed the number of the different cell populations compared with control-treated mice (*, p < 0.05). Data are representative of two independent experiments.

FIGURE 6.

Treatment of advanced tumors with chNKG2D T cells induced activation of the tumor-infiltrating leukocyte populations. ID8-GFP cells were injected and mice were treated with wtNKG2D (▴) or chNKG2D (•) T cells i.p. 5 weeks later. A, The number of NK1.1+CD3 NK cells, CD69+NK1.1+CD3-activated NK cells, Ly5.1CD8+ CD69+ T cells, CD19+ CD3 B cells, CD11c+MHC class IIhigh cells, F4/80+GR1+ cells, F480 GR1+ neutrophils, or GFP+ ID8 tumor cells was determined before T cell injection (day 0), and 1 and 3 days after T cell injection. B, Peritoneal wash cells from wtNKG2D (□) or chNKG2D T cell (▪) -treated tumor-bearing mice from each time point were cultured in medium for 24 h. Cell-free supernatants were assayed for IFN-γ or NO. The average of each group (n = 4) is shown. Treatment with chNKG2D T cells significantly changed the number of the different cell populations compared with control-treated mice (*, p < 0.05). Data are representative of two independent experiments.

Close modal

To determine whether treatment with chNKG2D T cells could increase survival in mice with established solid tumors, wtNKG2D or chNKG2D T cells were transferred to tumor-bearing mice 5, 6, and 7 weeks after tumor cell injection (Fig. 7 A). Although mice treated with wtNKG2D T cells had a median survival of 88 days, treatment with chNKG2D T cells significantly increased the survival of tumor-bearing mice. All chNKG2D T cell-treated mice survived longer than the mice treated with wtNKG2D T cells, and seven of 12 chNKG2D T cell-treated mice survived long-term and were tumor-free 225 days after tumor cell injection. The wtNKG2D T cell-treated mice had large tumor burdens at the time of death, with all mice having over 100 solid tumors on the peritoneal cavity. However, the five chNKG2D T cell-treated mice that were killed due to tumor growth had fewer solid tumors on the peritoneal cavity, ranging from 8 to 40 large solid tumors. This indicated that while some chNKG2D T cell-treated mice could not control tumor growth, their tumor burden was still decreased compared with wtNKG2D T cell-treated mice.

FIGURE 7.

ChNKG2D T cells increase survival of mice bearing established ID8 tumors. ID8-GFP cells were injected on day 0 and mice were treated three times with wtNKG2D (WT; ▪) or chNKG2D (CH; □) T cells i.p. after either 5, 6, and 7 weeks (A) or 5, 7, and 9 weeks (B). The survival of the mice was measured (n = 11–12 per group). Treatment with chNKG2D T cells significantly increased the survival of tumor-bearing mice compared with control-treated mice (***, p < 0.001).

FIGURE 7.

ChNKG2D T cells increase survival of mice bearing established ID8 tumors. ID8-GFP cells were injected on day 0 and mice were treated three times with wtNKG2D (WT; ▪) or chNKG2D (CH; □) T cells i.p. after either 5, 6, and 7 weeks (A) or 5, 7, and 9 weeks (B). The survival of the mice was measured (n = 11–12 per group). Treatment with chNKG2D T cells significantly increased the survival of tumor-bearing mice compared with control-treated mice (***, p < 0.001).

Close modal

Transferring chNKG2D T cells at weekly intervals may not be an ideal therapeutic approach, as the peak of the IFN-γ immune response occurred 1 week after chNKG2D T cell injection. Injecting the second dose of chNKG2D T cells during the peak of the immune response from the first chNKG2D T cell injection may not result in a stronger boost in the ongoing immune response. Therefore the timing of the chNKG2D T cell injections was altered to be administered 5, 7, and 9 weeks after tumor cell injection (Fig. 7 B). Using this treatment regimen, transfer of chNKG2D T cells led to long-term, tumor-free survival in 100% of the mice bearing tumors for 5 weeks. This demonstrates that successful treatment of established tumors did not require the administration of more chNKG2D T cells but an optimal timing of the T cell infusion based on what occurred in the microenvironment was necessary.

Therapies that alter leukocyte populations in the tumor microenvironment to decrease the effect of the immunosuppressive populations and induce the recruitment and activation of antitumor immune cells may lead to the development of long-lived anti-tumor immune responses and improved cancer therapy. These data show that treatment with chNKG2D T cells induced a proinflammatory immune response at the tumor site in a mouse model of ovarian cancer. chNKG2D T cell treatment led to an increase in activated NK cells and activated CD8+ T cells, and a decrease in Foxp3+ Treg cells. Tumor-infiltrating macrophages and DCs also became activated after chNKG2D T cell treatment, increasing their activation of Ag-specific T cells. Cells isolated from the peritoneal cavity produced increased amounts of IFN-γ, NO, and other proinflammatory cytokines. These changes in tumor-infiltrating populations and cytokine secretion required chNKG2D T cell-derived IFN-γ, GM-CSF, and perforin, indicating that both cytotoxicity and cytokine secretion played a role in changing the tumor microenvironment. By understanding the immune response induced by chNKG2D T cells, a treatment regimen with chNKG2D T cells was successfully designed to lead to tumor-free survival in mice bearing advanced ovarian tumors.

Many immunosuppressive cells and molecules are found in advanced ovarian cancer, including myeloid-derived suppressor cells and Treg cells. These cells can inhibit the immune response to ovarian cancer cells through multiple mechanisms. Tumor-associated myeloid cells may express molecules that can inhibit T cell responses, including B7-H1 and B7-H4 (8, 24, 25). Tumor-infiltrating T cells can express PD-1 and interaction with PD-L1 and PD-L2 expressed by cells at the tumor site inhibits T cell proliferation, cytokine secretion, and cytotoxicity (8, 26, 27). Additionally, tumor-associated macrophages may express IDO, PGE2, and arginase, which may inhibit T cell responses through down-regulating CD3 expression, inhibiting T cell proliferation, or inducing apoptosis (9, 10, 28). Although many studies suggest that CD11b+GR-1+ myeloid cells are suppressive in the tumor microenvironment, other studies in infection models find that GR-1-expressing macrophages are inflammatory monocytes (9, 10, 23). These F4/80+GR-1+ inflammatory monocytes secrete cytokines, including NO, IL-12, and TNF-α, and are involved in clearance of infections such as Listeria monocytogenes and Toxoplasma gondii (23, 29, 30). Treatment of tumor-bearing mice with chNKG2D T cells changes the phenotype of the tumor-associated macrophages, such that the cells display characteristics of inflammatory macrophages, expressing GR-1, and secreting NO. These cells also stimulated Ag-specific T cell proliferation. Thus, chNKG2D T cells reversed the immunosuppressive phenotype of tumor-associated macrophages to become proinflammatory, tipping the balance in favor of developing a host immune response against the tumor. Activation of macrophages required chNKG2D T cell-derived IFN-γ and host cell responsiveness to IFN-γ, demonstrating that expression of IFN-γ is essential for alteration of macrophages at the tumor site.

Foxp3+ Treg cells are found in human and murine ovarian tumors and the presence of Treg cells is inversely correlated with survival (31, 32). Treg cells isolated from ovarian cancer ascites samples can inhibit the proliferation, cytokine secretion, and cytotoxicity of tumor-infiltrating T cells (5, 31). chNKG2D T cell treatment of established tumors almost completely eliminated Foxp3+ Treg cells at the tumor site. It has been shown that human adaptive Treg cells can express NKG2D ligands during infection with Mycobacterium tuberculosis, but that natural Treg cells did not express NKG2D ligands (33). Similarly, this study showed that murine Foxp3+CD4+ Treg cells isolated from the tumor environment expressed NKG2D ligands, while Treg cells from the spleen of tumor-bearing mice or from naive mice did not express NKG2D ligands on their cell surface. Thus tumor-associated Treg cells are potential direct targets for chNKG2D T cells and chNKG2D T cells were shown to kill Treg cells in vitro through a mechanism that required expression of perforin but not FasL. The elimination of this suppressive population from the tumor site would allow tumor-infiltrating immune cells to be more effective.

A large proportion of the cells at the tumor site were CD19+ B cells and this population was also altered after chNKG2D T cell treatment. These cells can express PD-L1 and PD-L2 and can secrete anti-inflammatory cytokines including IL-10 that may inhibit immune responses to tumors (34, 35). chNKG2D T cells required the expression of not only GM-CSF and IFN-γ, but also perforin for the decrease in B cells. IFN-γ and TLR stimulation can induce the egress of B cells from the peritoneal cavity, thus IFN-γ secretion from chNKG2D T cells and host cells may directly cause the trafficking of the B cells out of the peritoneal cavity (36, 37). Another possibility is that chNKG2D T cells lysed the B cells; however, we did not detect NKG2D ligand expression on these B cells, nor did we observe killing of B cells in vitro, so direct killing was unlikely in vivo. Another hypothesis is that upon chNKG2D T cell lysis of the ID8 tumor cells, endogenous molecules that can stimulate TLRs were released from the dying tumor cells, such as heat shock proteins (38). These molecules may stimulate TLRs on the B cells, causing their activation and subsequent trafficking from the peritoneal cavity.

In addition to decreasing immunosuppressive populations at the tumor site, chNKG2D T cell treatment also increased cells that can potentially attack tumor cells. NK cells and CD8+ T cells both can lyse tumor cells, thus decreasing tumor burden and also releasing tumor Ags to promote Ag presentation. There was also an increase in the secretion of proinflammatory cytokines, including IFN-γ, by host cells at the tumor site. IFN-γ has many antitumor properties, including increasing Ag presentation, maturing APCs, decreasing angiogenesis, having cytostatic effects directly on tumor cells, and IFN-γ expression in human ovarian cancer is associated with a favorable prognosis (39, 40, 41, 42, 43). Components of the endogenous immune system may be involved in immune surveillance against the tumor even without chNKG2D T cell treatment. Previous work has shown that the host immune system responds to this tumor, but this response is not sufficient for tumor elimination (44). Specifically, wtNKG2D T cell-treated mice that were deficient in IFN-γ or NK cell depleted had greater tumor growth compared with B6 mice. This indicates that these host mechanisms play a role in controlling tumor growth; however, these host mechanisms are not able to eliminate the tumor, which may be due to immune suppression at the tumor site. chNKG2D T cells likely act in combination with host immune cells to overcome local immune suppression and result in tumor elimination. Additional work has shown that host-derived perforin, IFN-γ, NK cells, and lymphocytes are all required for complete tumor reduction by chNKG2D T cells (44). This further indicates the importance of the activation of the host immune cells for chNKG2D T cells antitumor efficacy. Data show that host cells need to express IFN-γR1 for the inflammatory response at the tumor site, indicating that IFN-γ produced at the tumor site by chNKG2D T cells and host cells is acting directly on host cells. One possible action may be to activate tumor-associated macrophages, thus changing the phenotype of the macrophages and inducing cytokine secretion. The activated macrophages may be secreting NO, which can have anti-tumor effects including direct killing of ID8 tumor cells, or may be secreting other cytokines that activate host NK cells and T cells (45, 46, 47).

Through developing a better understanding of the immune response generated by the chNKG2D T cells at the tumor site, this study helped to develop a treatment regimen that was successful at treating tumors established for 5 weeks, a time when many solid tumors are well-established on the peritoneal wall. Instead of administering an increased number of chNKG2D T cells to achieve long-term tumor-free survival in all mice, we altered the scheduling of the chNKG2D T cell doses such that the second and third doses did not coincide with the peak of the ongoing host immune response after the initial chNKG2D T cell injection. The treatment regimen of administering chNKG2D T cells every other week led to long-term tumor-free survival in 100% of mice. This illustrates that one may need to analyze the kinetics and type of immune response induced by the transferred T cells and determine the best therapeutic regimen to obtain optimal efficacy.

In this study, chNKG2D T cells were able to reduce immunosuppressive cells and induce activation of host antitumor immune cells both in early and established tumors. This indicates that despite the increased prevalence of immunosuppressive cells in established tumors, chNKG2D T cells may potentially be able to induce immune responses in patients with early or late stage tumors. chNKG2D T cell-derived IFN-γ and GM-CSF were required for many of the changes in the leukocyte populations and for the secretion of proinflammatory cytokines. Previous work has shown that chNKG2D T cell-derived IFN-γ and GM-CSF were also required for complete antitumor efficacy, indicating that changing the tumor microenvironment and inducing a proinflammatory response at the tumor site is essential for cancer therapy (17, 18). chNKG2D T cells may lyse tumor cells, thus decreasing tumor burden and increasing Ag presentation. Additionally, chNKG2D T cells reduce the immunosuppressive populations while concurrently secreting proinflammatory cytokines, which can recruit and activate host NK cells, T cells, and APCs. Although many current studies need to combine multiple different therapies to decrease immunosuppression and increase the activation of the immune response, this study shows that transfer of chNKG2D T cells is a novel approach to tip the balance of the immune response in favor of decreasing immune suppression and developing an inflammatory antitumor immune response in ovarian cancer, resulting in long-term, tumor-free survival in mice bearing established ovarian tumors.

We thank Gary Ward and Alice Givan at the Englert Cell Analysis Laboratory for assistance with flow cytometry (Norris Cotton Cancer Center, Lebanon, NH), the Immune Monitoring Laboratory for assistance in Luminex analysis (Norris Cotton Cancer Center), and the Animal Resource Center at Dartmouth Medical School for help with the animal studies.

The authors have no financial conflict of interest.

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 study was supported in part by grants from Department of Microbiology and Immunology and the National Institutes of Health (T32 AI 07363, CA 130911). A.B. was supported by a John H. Copenhaver, Jr., and William H. Thomas, M.D. Fellowship from Dartmouth College. The contents are solely the responsibility of the authors and do not necessarily represent the official views of National Institutes of Health.

3

Abbreviations used in this paper: Treg, regulatory T; chNKG2D, chimeric NKG2D; DC, dendritic cell; IFN-γR1, IFN-γ receptor 1; wtNKG2D, wild-type NKG2D receptor.

1
Rosenberg, S. A., N. P. Restifo, J. C. Yang, R. A. Morgan, M. E. Dudley.
2008
. Adoptive cell transfer: a clinical path to effective cancer immunotherapy.
Nat. Rev. Cancer
8
:
299
-308.
2
Cohen, J. I., C. M. Bollard, R. Khanna, S. Pittaluga.
2008
. Current understanding of the role of Epstein-Barr virus in lymphomagenesis and therapeutic approaches to EBV-associated lymphomas.
Leuk. Lymphoma
49
: (Suppl. 1):
27
-34.
3
June, C. H..
2007
. Adoptive T cell therapy for cancer in the clinic.
J. Clin. Invest.
117
:
1466
-1476.
4
Kershaw, M. H., J. A. Westwood, L. L. Parker, G. Wang, Z. Eshhar, S. A. Mavroukakis, D. E. White, J. R. Wunderlich, S. Canevari, L. Rogers-Freezer, et al
2006
. A phase I study on adoptive immunotherapy using gene-modified T cells for ovarian cancer.
Clin. Cancer Res.
12
:
6106
-6115.
5
Nelson, B. H..
2008
. The impact of T-cell immunity on ovarian cancer outcomes.
Immunol. Rev.
222
:
101
-116.
6
Coukos, G., J. R. Conejo-Garcia, R. Buckanovich, F. Benencia.
2007
. Vascular leukocytes: a population with angiogenic and immunossuppressive properties highly represented in ovarian cancer.
Adv. Exp. Med. Biol.
590
:
185
-193.
7
Barnett, B. G., J. Ruter, I. Kryczek, M. J. Brumlik, P. J. Cheng, B. J. Daniel, G. Coukos, W. Zou, T. J. Curiel.
2008
. Regulatory T cells: a new frontier in cancer immunotherapy.
Adv. Exp. Med. Biol.
622
:
255
-260.
8
Zou, W., L. Chen.
2008
. Inhibitory B7-family molecules in the tumour microenvironment.
Nat. Rev. Immunol.
8
:
467
-477.
9
Marigo, I., L. Dolcetti, P. Serafini, P. Zanovello, V. Bronte.
2008
. Tumor-induced tolerance and immune suppression by myeloid derived suppressor cells.
Immunol. Rev.
222
:
162
-179.
10
Sica, A., P. Larghi, A. Mancino, L. Rubino, C. Porta, M. G. Totaro, M. Rimoldi, S. K. Biswas, P. Allavena, A. Mantovani.
2008
. Macrophage polarization in tumour progression.
Semin. Cancer Biol.
18
:
349
-355.
11
Marth, C., G. H. Windbichler, H. Hausmaninger, E. Petru, K. Estermann, A. Pelzer, E. Mueller-Holzner.
2006
. Interferon-γ in combination with carboplatin and paclitaxel as a safe and effective first-line treatment option for advanced ovarian cancer: results of a phase I/II study.
Int. J. Gynecol. Cancer
16
:
1522
-1528.
12
Pujade-Lauraine, E., J. P. Guastalla, N. Colombo, P. Devillier, E. Francois, P. Fumoleau, A. Monnier, M. Nooy, L. Mignot, R. Bugat, et al
1996
. Intraperitoneal recombinant interferon γ in ovarian cancer patients with residual disease at second-look laparotomy.
J. Clin. Oncol.
14
:
343
-350.
13
Baur, M., E. Schernhammer, M. Gneist, P. Sevelda, P. Speiser, M. Hudec, C. Dittrich.
2005
. Phase I/II study of oral etoposide plus GM-CSF as second-line chemotherapy in platinum-pretreated patients with advanced ovarian cancer.
Br. J. Cancer
92
:
1019
-1025.
14
Kurbacher, C. M., J. A. Kurbacher, E. M. Cramer, K. Rhiem, P. K. Mallman, R. Reichelt, U. Reinhold, U. Stier, I. A. Cree.
2005
. Continuous low-dose GM-CSF as salvage therapy in refractory recurrent breast or female genital tract carcinoma.
Oncology (Williston Park).
19
:
23
-26.
15
Barnett, B., I. Kryczek, P. Cheng, W. Zou, T. J. Curiel.
2005
. Regulatory T cells in ovarian cancer: biology and therapeutic potential.
Am. J. Reprod. Immunol.
54
:
369
-377.
16
Hodi, F. S., M. Butler, D. A. Oble, M. V. Seiden, F. G. Haluska, A. Kruse, S. Macrae, M. Nelson, C. Canning, I. Lowy, et al
2008
. Immunologic and clinical effects of antibody blockade of cytotoxic T lymphocyte-associated antigen 4 in previously vaccinated cancer patients.
Proc. Natl. Acad. Sci. USA
105
:
3005
-3010.
17
Barber, A., T. Zhang, C. L. Sentman.
2008
. Immunotherapy with chimeric NKG2D receptors leads to long-term tumor-free survival and development of host antitumor immunity in murine ovarian cancer.
J. Immunol.
180
:
72
-78.
18
Zhang, T., A. Barber, C. L. Sentman.
2007
. Chimeric NKG2D modified T cells inhibit systemic T-cell lymphoma growth in a manner involving multiple cytokines and cytotoxic pathways.
Cancer Res.
67
:
11029
-11036.
19
Barber, A., T. Zhang, L. R. DeMars, J. Conejo-Garcia, K. F. Roby, C. L. Sentman.
2007
. Chimeric NKG2D receptor-bearing T cells as immunotherapy for ovarian cancer.
Cancer Res.
67
:
5003
-5008.
20
Zhang, T., A. Barber, C. L. Sentman.
2006
. Generation of antitumor responses by genetic modification of primary human T cells with a chimeric NKG2D receptor.
Cancer Res.
66
:
5927
-5933.
21
Zhang, T., B. A. Lemoi, C. L. Sentman.
2005
. Chimeric NK-receptor-bearing T cells mediate antitumor immunotherapy.
Blood
106
:
1544
-1551.
22
Yang, R., Z. Cai, Y. Zhang, W. H. Yutzy, IV, K. F. Roby, R. B. Roden.
2006
. CD80 in immune suppression by mouse ovarian carcinoma-associated Gr-1+CD11b+ myeloid cells.
Cancer Res.
66
:
6807
-6815.
23
Geissmann, F., C. Auffray, R. Palframan, C. Wirrig, A. Ciocca, L. Campisi, E. Narni-Mancinelli, G. Lauvau.
2008
. Blood monocytes: distinct subsets, how they relate to dendritic cells, and their possible roles in the regulation of T-cell responses.
Immunol. Cell Biol.
86
:
398
-408.
24
Liu, Y., B. Zeng, Z. Zhang, Y. Zhang, R. Yang.
2008
. B7–H1 on myeloid-derived suppressor cells in immune suppression by a mouse model of ovarian cancer.
Clin. Immunol.
129
:
471
-481.
25
Kryczek, I., S. Wei, G. Zhu, L. Myers, P. Mottram, P. Cheng, L. Chen, G. Coukos, W. Zou.
2007
. Relationship between B7–H4, regulatory T cells, and patient outcome in human ovarian carcinoma.
Cancer Res.
67
:
8900
-8905.
26
Hamanishi, J., M. Mandai, M. Iwasaki, T. Okazaki, Y. Tanaka, K. Yamaguchi, T. Higuchi, H. Yagi, K. Takakura, N. Minato, T. Honjo, S. Fujii.
2007
. Programmed cell death 1 ligand 1 and tumor-infiltrating CD8+ T lymphocytes are prognostic factors of human ovarian cancer.
Proc. Natl. Acad. Sci. USA
104
:
3360
-3365.
27
Curiel, T. J., S. Wei, H. Dong, X. Alvarez, P. Cheng, P. Mottram, R. Krzysiek, K. L. Knutson, B. Daniel, M. C. Zimmermann, et al
2003
. Blockade of B7-H1 improves myeloid dendritic cell-mediated antitumor immunity.
Nat. Med.
9
:
562
-567.
28
Bak, S. P., A. Alonso, M. J. Turk, B. Berwin.
2008
. Murine ovarian cancer vascular leukocytes require arginase-1 activity for T cell suppression.
Mol. Immunol.
46
:
258
-268.
29
Mordue, D. G., L. D. Sibley.
2003
. A novel population of Gr-1+-activated macrophages induced during acute toxoplasmosis.
J. Leukocyte Biol.
74
:
1015
-1025.
30
Robben, P. M., M. LaRegina, W. A. Kuziel, L. D. Sibley.
2005
. Recruitment of Gr-1+ monocytes is essential for control of acute toxoplasmosis.
J. Exp. Med.
201
:
1761
-1769.
31
Curiel, T. J., G. Coukos, L. Zou, X. Alvarez, P. Cheng, P. Mottram, M. Evdemon-Hogan, J. R. Conejo-Garcia, L. Zhang, M. Burow, et al
2004
. Specific recruitment of regulatory T cells in ovarian carcinoma fosters immune privilege and predicts reduced survival.
Nat. Med.
10
:
942
-949.
32
Wolf, D., A. M. Wolf, H. Rumpold, H. Fiegl, A. G. Zeimet, E. Muller-Holzner, M. Deibl, G. Gastl, E. Gunsilius, C. Marth.
2005
. The expression of the regulatory T cell-specific forkhead box transcription factor FoxP3 is associated with poor prognosis in ovarian cancer.
Clin. Cancer Res.
11
:
8326
-8331.
33
Roy, S., P. F. Barnes, A. Garg, S. Wu, D. Cosman, R. Vankayalapati.
2008
. NK cells lyse T regulatory cells that expand in response to an intracellular pathogen.
J. Immunol.
180
:
1729
-1736.
34
Zhong, X., J. R. Tumang, W. Gao, C. Bai, T. L. Rothstein.
2007
. PD-L2 expression extends beyond dendritic cells/macrophages to B1 cells enriched for V(H)11/V(H)12 and phosphatidylcholine binding.
Eur. J. Immunol.
37
:
2405
-2410.
35
Bouaziz, J. D., K. Yanaba, T. F. Tedder.
2008
. Regulatory B cells as inhibitors of immune responses and inflammation.
Immunol. Rev.
224
:
201
-214.
36
Ha, S. A., M. Tsuji, K. Suzuki, B. Meek, N. Yasuda, T. Kaisho, S. Fagarasan.
2006
. Regulation of B1 cell migration by signals through Toll-like receptors.
J. Exp. Med.
203
:
2541
-2550.
37
Balabanian, K., A. Foussat, L. Bouchet-Delbos, J. Couderc, R. Krzysiek, A. Amara, F. Baleux, A. Portier, P. Galanaud, D. Emilie.
2002
. Interleukin-10 modulates the sensitivity of peritoneal B lymphocytes to chemokines with opposite effects on stromal cell-derived factor-1 and B-lymphocyte chemoattractant.
Blood
99
:
427
-436.
38
Miyake, K..
2007
. Innate immune sensing of pathogens and danger signals by cell surface Toll-like receptors.
Semin. Immunol.
19
:
3
-10.
39
Schroder, K., P. J. Hertzog, T. Ravasi, D. A. Hume.
2004
. Interferon-γ: an overview of signals, mechanisms and functions.
J. Leukocyte Biol.
75
:
163
-189.
40
Dunn, G. P., C. M. Koebel, R. D. Schreiber.
2006
. Interferons, immunity and cancer immunoediting.
Nat. Rev. Immunol.
6
:
836
-848.
41
Blankenstein, T., Z. Qin.
2003
. The role of IFN-γ in tumor transplantation immunity and inhibition of chemical carcinogenesis.
Curr. Opin. Immunol.
15
:
148
-154.
42
Marth, C., H. Fiegl, A. G. Zeimet, E. Muller-Holzner, M. Deibl, W. Doppler, G. Daxenbichler.
2004
. Interferon-γ expression is an independent prognostic factor in ovarian cancer.
Am. J. Obstet. Gynecol.
191
:
1598
-1605.
43
Kusuda, T., K. Shigemasa, K. Arihiro, T. Fujii, N. Nagai, K. Ohama.
2005
. Relative expression levels of Th1 and Th2 cytokine mRNA are independent prognostic factors in patients with ovarian cancer.
Oncol. Rep.
13
:
1153
-1158.
44
Barber, A., C. L. Sentman.
2009
. Chimeric NKG2D T cells require both T cell- and host-derived cytokine secretion and perforin expression to increase tumor antigen presentation and systemic immunity.
J. Immunol.
183
:
2365
-2372.
45
Mocellin, S., V. Bronte, D. Nitti.
2007
. Nitric oxide, a double edged sword in cancer biology: searching for therapeutic opportunities.
Med. Res. Rev.
27
:
317
-352.
46
Huerta, S., S. Chilka, B. Bonavida.
2008
. Nitric oxide donors: novel cancer therapeutics (review).
Int. J. Oncol.
33
:
909
-927.
47
Hagemann, T., T. Lawrence, I. McNeish, K. A. Charles, H. Kulbe, R. G. Thompson, S. C. Robinson, F. R. Balkwill.
2008
. “Re-educating” tumor-associated macrophages by targeting NF-κB.
J. Exp. Med.
205
:
1261
-1268.