Tumor-reactive CD4(+) T cells develop cytotoxic activity and eradicate large established melanoma after transfer into lymphopenic hosts

Abstract

Adoptive transfer of large numbers of tumor-reactive CD8(+) cytotoxic T lymphocytes (CTLs) expanded and differentiated in vitro has shown promising clinical activity against cancer. However, such protocols are complicated by extensive ex vivo manipulations of tumor-reactive cells and have largely focused on CD8(+) CTLs, with much less emphasis on the role and contribution of CD4(+) T cells. Using a mouse model of advanced melanoma, we found that transfer of small numbers of naive tumor-reactive CD4(+) T cells into lymphopenic recipients induces substantial T cell expansion, differentiation, and regression of large established tumors without the need for in vitro manipulation. Surprisingly, CD4(+) T cells developed cytotoxic activity, and tumor rejection was dependent on class II-restricted recognition of tumors by tumor-reactive CD4(+) T cells. Furthermore, blockade of the coinhibitory receptor CTL-associated antigen 4 (CTLA-4) on the transferred CD4(+) T cells resulted in greater expansion of effector T cells, diminished accumulation of tumor-reactive regulatory T cells, and superior antitumor activity capable of inducing regression of spontaneous mouse melanoma. These findings suggest a novel potential therapeutic role for cytotoxic CD4(+) T cells and CTLA-4 blockade in cancer immunotherapy, and demonstrate the potential advantages of differentiating tumor-reactive CD4(+) cells in vivo over current protocols favoring in vitro expansion and differentiation.

Figure 1.

Transfer of small numbers of CD4⁺Trp1⁺ T cells in combination with radiation and CTLA-4 blockade produces potent rejection of established B16/BL6 melanoma tumors. Tumor-bearing mice were treated at day 10 with 5 Gy of RT and were injected or not with 50,000 naive tumor-reactive CD4⁺Trp1⁺ cells and anti–CTLA-4. (a) Data are presented as tumor growth in each mouse (three left panels; representative of at least three independent experiments; n = 5 mice per group) and cumulative survival from two independent experiments (far right panel; n = 10 mice per group). (b) Representative images of tumors in mice treated with 5 Gy of RT and CD4⁺Trp1⁺ cells in the absence (top) or presence (bottom) of anti–CTLA-4. The far right panel shows disseminated depigmentation in mice surviving therapy. (c–f) Blood samples from mice treated with 5 Gy of RT (black line), 5 Gy + CD4⁺Trp1⁺ (red line), and 5 Gy + CD4⁺Trp1⁺ + anti–CTLA-4 (blue line) were monitored at different time points after tumor challenge. Data are representative of three independent experiments (n = 5 mice per group). c shows the percentage of CD4⁺Trp1⁺ T cells from blood lymphocytes, whereas d depicts the percentage of Foxp3⁺ cells within the CD4⁺Trp1⁺ population. Serum samples were analyzed over time and tested for levels of IFN-γ (e) and TNF (f). Error bars represent means ± SD.

Figure 2.

Combination of CD4⁺Trp1⁺ ACT and CTLA-4 blockade enhances activation and differentiation of CD4⁺Trp1⁺ T^eff cells and reduces the number of T reg cells. (a–e) Tumor-bearing mice were treated at day 10 with or without 5 Gy of RT followed by transfer of 50,000 CD4⁺Trp1⁺ cells in the presence or absence of anti–CTLA-4 mAb. 8 d after therapy (day 18 after tumor challenge), mice were euthanized and numbers of CD4⁺Trp1⁺Foxp3⁻ cells (T^eff cells; a and d) and CD4⁺Trp1⁺Foxp3⁺ cells (T reg cells; b and e) were analyzed in LNs (a and b) and tumors (d and e). Numbers of CD4⁺Trp1⁺ T^eff and T reg cells in tumors (d and e) were calculated as described in Materials and methods. (g and h) Proportions of CD4⁺Trp1⁺ T^eff cells to CD4⁺Trp1⁺ T reg cells (g) and CD4⁺Trp1⁺ T^eff cells to total T reg cells (h) were also calculated from tumor samples. CD4⁺Trp1⁺ cells from LNs (c) and tumor samples (f) were restimulated ex vivo, and IFN-γ, TNF, and IL-2 secretion was determined on a per cell basis. (i) In a parallel experiment, tumors were dissected and fresh frozen in optimum cutting temperature solution, cut, and stained for DAPI (blue), CD31 (red), and CD4 (green). Samples were analyzed by confocal microscopy with a 20X water immersion objective. Bars, 50 µm. Images showing whole-tumor immunofluorescence are shown in Fig. S3. Data are representative of three independent experiments (n = 3 mice per group). Horizontal bars represent means.

Figure 3.

Rejection of established tumors depends on IFN-γ secretion by CD4⁺Trp1⁺ T cells and is independent of TNF and endogenous T, B, and NK cells. Tumor-bearing mice were irradiated at day 10 with 5 Gy followed or not by treatment with 50,000 CD4⁺Trp1⁺ cells and anti–CTLA-4 mAb. Tumor growth was measured and data are presented for each independent mouse. (a) Mice received only 5 Gy (left) or 5 Gy + CD4⁺Trp1⁺ + anti–CTLA-4 with control antibody or with neutralizing anti-TNF or anti–IFN-γ antibody. Two further groups included IFN-γ^−/− and IFN-γR^−/− recipients treated at day 10 with 5 Gy, 50,000 CD4⁺Trp1⁺ cells, and anti–CTLA-4. (b) In a separate experiment, recipient PFN^−/− and Rag^−/− mice were compared with wild-type mice for their capacity to reject established melanoma after treatment at day 10 with 5 Gy, 50,000 CD4⁺Trp1⁺ cells, and anti–CTLA-4. (c and d) Mean tumor growth in mice treated with 5 Gy of RT + CD4⁺Trp1⁺ cells in the presence (blue line) or absence (red line) of anti–CTLA-4. Mice used for this set of experiments were either wild-type C57BL/6 (c) or CTLA-4 human Tg (d) mice in which anti–CTLA-4 antibodies will only block CTLA-4 on the transferred CD4⁺Trp1⁺ cells. Data are representative of at least two independent experiments (n = 5 mice per group).

Figure 4.

Antitumor activity after CD4⁺Trp1⁺ T cell transfer is specific to B16/BL6 melanoma and does not prevent growth of an unrelated tumor. (a–e) Mice were challenged with B16/BL6 melanoma and EL-4 in opposing flanks (a and b) or coinjected in the same flank (c–e). 10 d after tumor challenge, all mice received 5 Gy of RT and anti–CTLA-4 mAb. Half of the mice were also treated with 50,000 CD4⁺Trp1⁺ T cells and all mice were monitored for tumor growth (a–c). (d) Representative images of mice challenged with B16/BL6 and EL-4 in the same site shows the growth of a pigmented tumor in the absence of CD4⁺Trp1⁺ transfer and growth of a nonpigmented tumor after administration of 50,000 tumor-reactive CD4⁺Trp1⁺ T cells Data are representative of three independent experiments (n = 5 mice per group). To verify that B16/BL6 melanoma is rejected in the coinjection setting, mice were challenged with EL-4 and B16/BL6-luciferase in the same site, and light emission was determined over time after tumor challenge and triple therapy. (e) Representative images of luciferase signal disappearing from the tumor after therapy. (f) Quantification of luciferase signal and tumor growth. Data are representative of two independent experiments (n = 5 mice per group). Error bars represent means ± SD.

Figure 5.

Tumor-reactive CD4⁺Trp1⁺ T cells develop class II–dependent killing activity. (a) 10 d after tumor challenge, tumor-bearing mice were treated or not with 5 Gy of RT, 50,000 CD4⁺Trp1⁺ cells, 5 Gy + CD4⁺Trp1⁺, or 5 Gy + CD4⁺Trp1⁺ + anti–CTLA-4 mAb. 7 d after therapy (day 17 after tumor challenge), all mice were injected with CFSE-labeled B cell targets. CFSE^high (5 µM) cells were also loaded with the class II–restricted peptide recognized by CD4⁺Trp1⁺ cells, whereas CFSE^low (0.5 µM) cells were used as a control population. 14–16 h after i.v. injection of CFSE targets, mice were sacrificed and in vivo killing activity was quantified in single-cell suspensions from the spleens of each mouse. Data are representative of three independent experiments (n = 3 mice per group). Horizontal bars represent means. (b–f) CD4⁺Trp1⁺ T cells were primed in vivo and expanded in vitro to allow analysis of their in vitro killing activity. Tumor-reactive in vivo–primed CD4⁺Trp1⁺ T cells were incubated at different ratios with CFSE-loaded spleen targets (b and d) or B16/BL6 and EL-4 tumor targets (c, e, and f). (b and d) CFSE^low B cell targets were also loaded with the class II–restricted peptide recognized by Trp1 cells, whereas CFSE^high cells were used as the control population. (c, e, and f) B16/BL6 cells were loaded with 0.5 µM CFSE, whereas EL-4 control tumor cells were loaded with 5 µM CFSE. Blocking anti–class II or anti-FASL antibodies were used at a final concentration of 50 µg/ml for all in vitro experiments (b–e). (f) In vitro killing of tumor targets was also tested in presence of 0.5 µM concanamycin A (PFN inhibitor) and/or 25 µM Z-AAD-CMK. In vitro killing activity was determined 12–14 h after initiation of the assay by quantifying the decrease of the target population in comparison to the control population, as described in Materials and methods. Data are representative of four independent experiments. Numbers in a–c indicate percentages. Error bars in d–f represent means ± SD.

Figure 6.

B16 melanoma up-regulates class II expression in vivo in an IFN-γ–dependent manner and are direct targets of cytotoxic CD4⁺Trp1⁺ T cells. (a) In vivo class II expression by tumor cells from mice receiving either CD4⁺Trp1⁺ cells alone or in combination with 5 Gy of RT and in the presence or absence of blocking anti–CTLA-4 and neutralizing IFN-γ. Mice were treated at day 10 after challenge with B16/BL6 melanoma expressing Thy1.1 and sacrificed 7 d after initiation of therapy (day 17 after tumor). Single-cell suspensions of tumors were analyzed for MHC class II levels by gating in Thy1.1-positive cells. Numbers indicate percentages. (b) Quantification of class II expression by tumors in vivo shown as cumulative data from three independent experiments (n = 3 mice per group). Horizontal bars represent means. (c) CD4⁺Trp1⁺ T cells were primed in vivo and expanded in vitro to allow analysis of their antitumor activity upon retransfer into tumor-bearing CII^−/− recipient mice. In brief, 10 d after tumor challenge, MHC CII^−/− recipient mice were treated with 5 Gy of RT, primed CD4⁺Tpr1⁺ cells, and anti–CTLA-4. Half of the mice were also treated with 200 µg of blocking anti–class II antibody every 3 d, and tumor growth was monitored over time. Data are representative of two independent experiments (n = 5 mice per group). Error bars represent means ± SEM.

Figure 7.

Triple therapy mediates tumor regression in a mouse model of spontaneous melanoma. Grm-1 Tg mice spontaneously develop melanoma tumors between 4–6 mo of age. After development of large tumor lesions (10–13 mo of age), mice were treated with 5 Gy of RT in combination with tumor-reactive CD4⁺Trp1⁺ T cells and anti–CTLA-4, as described in Materials and methods. Tumor progression was monitored weekly by gross morphological examination. The individual tumor lesions were measured with calipers weekly for each animal. The total tumor volume for each mouse was considered as 100% on the first day of treatment. (a) Representative images of treated mice and tumor regression. (b) Tumor volume after triple therapy. (c) One mouse with aggressive disease continued to grow tumor up to day 10 after therapy when tumor volume started to decrease, similar to what is observed in the transplantable B16/BL6 model. (d and e) Mice were also monitored for expansion of CD4⁺Trp1⁺ T cells in the blood (d) and cytokine levels in the serum (e) of treated mice. Data are representative of two independent experiments (n = 7–10 mice).