Avirulent Toxoplasma gondii generates therapeutic antitumor immunity by reversing immunosuppression in the ovarian cancer microenvironment

Abstract

Reversing tumor-associated immunosuppression seems necessary to stimulate effective therapeutic immunity against lethal epithelial tumors. Here, we show this goal can be addressed using cps, an avirulent, nonreplicating uracil auxotroph strain of the parasite Toxoplasma gondii (T. gondii), which preferentially invades immunosuppressive CD11c(+) antigen-presenting cells in the ovarian carcinoma microenvironment. Tumor-associated CD11c(+) cells invaded by cps were converted to immunostimulatory phenotypes, which expressed increased levels of the T-cell receptor costimulatory molecules CD80 and CD86. In response to cps treatment of the immunosuppressive ovarian tumor environment, CD11c(+) cells regained the ability to efficiently cross-present antigen and prime CD8(+) T-cell responses. Correspondingly, cps treatment markedly increased tumor antigen-specific responses by CD8(+) T cells. Adoptive transfer experiments showed that these antitumor T-cell responses were effective in suppressing solid tumor development. Indeed, intraperitoneal cps treatment triggered rejection of established ID8-VegfA tumors, an aggressive xenograft model of ovarian carcinoma, also conferring a survival benefit in a related aggressive model (ID8-Defb29/Vegf-A). The therapeutic benefit of cps treatment relied on expression of IL-12, but it was unexpectedly independent of MyD88 signaling as well as immune experience with T. gondii. Taken together, our results establish that cps preferentially invades tumor-associated antigen-presenting cells and restores their ability to trigger potent antitumor CD8(+) T-cell responses. Immunochemotherapeutic applications of cps might be broadly useful to reawaken natural immunity in the highly immunosuppressive microenvironment of most solid tumors.

Figure 1

cps preferentially invades CD45⁺CD11c⁺ tumor-associated APC, activates co-stimulatory molecule expression, and increases antigen presentation. A, mice (n = 5) bearing ID8-Defb29/Vegf-A tumors for 19 d were injected IP with cps-YFP and peritoneal cells were isolated and analyzed by FACS 18 h later. A (left and middle panels), cps invaded cells (YFP^hi) were CD45⁺ CD11c⁺. A (right panel), the percentage of invaded cell that are CD45⁻, CD45⁺, CD45⁺CD11⁻, and CD45⁺CD11c⁺.. B, PBS treated and cps treated tumor bearing mice (n = 5), treated as in A, analyzed for expression of CD80 and D86 in the cps invaded (YFP^hi) (dotted line histogram) and cps exposed noninvaded (solid line histogram) CD45⁺CD11c⁺ population. C, PBS and cps treated tumor-bearing mice (n = 5) treated as in A analyzed for CD80 and CD86 (mean fluorescence index, MFI) in cps invaded (YFP^hi) and cps treated noninvaded CD45⁺CD11c⁺ population. D, PBS and cps treated mice (n = 5) bearing tumors for 34 d treated as in B and were analyzed for CD80 and CD86 in cps invaded (YFP^hi) and cps treated and noninvaded CD45⁺CD11c⁺ population. E, mice bearing tumors for 21 d injected with ovalbumin and 4 h later with cps (n =6 ) or PBS (n =4). Mice IP injected with CFSE stained OT-I T cells 18 h later and proliferation determined by FACS 48 h later by dilution of CFSE in CD45⁺CD3⁺CD8⁺ T cells (cps treated, dotted line; PBS treated solid gray). Representative of two independent experiments.

Figure 2

Treatment with cps increases cellular infiltration into the spleen and the peritoneum. Mice (n =4) bearing ID8-Defb29/Vegf-A tumors for 8 d were treated with cps or PBS and leukocytes in the spleen (A–G) or peritoneum (H–N) assayed 12 days later by FACS (see Supplemental Fig. S3 for gating strategies). Number of: A. splenocytes. B, CD45⁺CD11c⁺ cells. C, macrophages (F4/80+ CD11b+). D, B cells (B220+). E, NK cells (NK1.1+). F, CD3⁺CD8⁺ T cells. G, CD3⁺CD4⁺ T cells. H, CD45⁺ cells. I, CD45⁺CD11c⁺ cells. J, CD3⁺CD8⁺ T cells. K, CD3⁺CD4⁺ T cells. L, CD45⁺CD11c⁺ cells as percentage of total peritoneal leukocytes. M, CD3⁺CD8⁺ T cells as percentage of total peritoneal leukocytes. N, CD45⁺CD3⁺CD4⁺Foxp3⁺ T-reg cells as a percentage of total peritoneal CD4⁺ T cells. Representative of two independent experiments.

Figure 3

Treatment with cps increases tumor antigen-specific CD8⁺ T cell responses. Mice (n = 6) bearing ID8-Defb29/Vegf-A tumors for 8 d treated with cps or PBS and CD8⁺ T cell populations in spleen (A–D) and peritoneum (E–H) analyzed by FACS (A–G), or assessed by granzyme B ELISpot (H) 12 d later. A, number splenic CD8⁺ expressing CD44 and CD62L. B, same as in A as percentage of total CD8⁺ T cells. C, number of splenic CD3⁺CD8⁺tetramer⁺ tumor antigen specific T cells. D, same as in C as percentage of total CD8⁺ T cells. E, total of CD3⁺CD8⁺tetramer⁺ peritoneal CD8⁺ T cells. F, same as E as percentage of total CD8⁺ T cells. G, same as in E except tetramer-positive CD8⁺ T cells assessed for co-expressing CD44 and CD69. H, mice (n = 6) bearing ID8-Defb29/Vegf-A tumors for 8 d were treated with cps or PBS and peritoneal CD8⁺ T cells assayed for granzyme B 12 d later. CD8⁺ T cells were purified and co-cultured with BMDCs either pulsed with tumor antigen or unpulsed prior to assay. Representative of two independent experiments.

Figure 4

T cells from cps treated tumor bearing mice suppress growth of ID8-Defb29/Vegf-A ovarian flank tumors. Mice (n = 4) bearing ID8-Defb29/Vegf-A flank tumors were injected IV with 2 × 10⁶ total CD3⁺ T cells purified from mice bearing ID8-Defb29/Vegf-A tumors treated with cps 8 d and 20 d after tumor challenge, from mice bearing untreated ID8-Defb29/Vegf-A peritoneal tumors, or from naive mice. Flank tumors were removed and measured 2 months later. Representative of two independent experiments.

Figure 5

Rejection of established ID8-Vegf-A tumors is independent of immune status to Toxoplasma gondii. A, mice bearing ID8-Vegf-A tumors treated with cps (n = 6) or PBS (n = 8) on day 8 and day 20 and survival was monitored. B, same as in A except groups of mice vaccinated with cps 90 days earlier to establish a life-long (19, 20) immunity to Toxoplasma prior to tumor challenge. Representative of two independent experiments.

Figure 6

The cps antitumor response requires IL-12 but not MyD88. A, mice (n = 8) bearing ID8-Defb29/Vegf-A treated with cps or PBS on days 8 and 20 and survival is shown. B, mice (n = 7) bearing ID8-Defb29/Vegf-A tumors for 35 d treated with cps and peritoneal IL-12p40 was evaluated 18 h later. C, mice (n = 4) same as in B except assay for IL-12p70. D, IL-12p40^−/− or C57BL/6 mice (n = 4) bearing ID8-Defb29/Vegf-A tumors were treated with cps or PBS on days 8 and 20. E, IL-12p35^−/− or C57BL/6 mice (n = 6) bearing ID8-Defb29/Vegf-A tumors treated with cps or PBS on days 8 and 20. F, MyD88^−/− or C57BL/6 mice bearing ID8-Defb29/Vegf-A tumors were treated with cps (n = 6) or PBS (n = 6) on days 8 and 20 (P = 0.8694, cps treated C57BL/6 compared to cps treated MyD88^−/−). Representative of two independent experiments.