Sustained Antigen-Specific Antitumor Recall Response Mediated by Gene-Modified CD4⁺ T Helper-1 and CD8⁺ T Cells

Introduction

Strategies involving gene-modification of T cells are emerging as a promising form of cancer immunotherapy. Current treatment regimens with active vaccines that use various adjuvants have shown little success in the clinic ( 1). Conversely, adoptive transfer approaches using lymphokine-activated killer cells or tumor-infiltrating lymphocytes (TIL) along with lymphoabalative conditioning have shown some remarkable antitumor responses, although largely in melanoma patients ( 2, 3). Given the difficulty of isolating tumor-reactive T cells for most malignancies, the genetic modification of T cells with chimeric single chain (scFv) receptors or TCRαβ genes allows generation of an unlimited supply of autologous antitumor effector cells. A recent study has shown the feasibility of this approach in which transfer of T cells gene-modified with TCRαβ genes recognizing the MART-1 antigen could affect tumor burden in patients with melanoma and that the response correlated with persistence of transferred TILs ( 4).

It is well documented that CD4⁺ T helper cells play an essential role in priming of CD8⁺ effector cells ( 5, 6) and generation of a memory response ( 7). For adoptive immunotherapy protocols using antigen-specific gene-modified T cells, a number of studies done in animal models and in the clinic have shown the importance of including CD4⁺ T helper cells together with CD8⁺ T cells to induce optimal therapeutic effects and maintain long-term persistence of the effector cells ( 8– 10). The exact role of CD4⁺ T cells in antitumor immunity is not totally clear at present and is complicated by the fact that they can be principally subdivided into T helper 1 (Th1) and T helper 2 (Th2) cells based on their ability to secrete various cytokines ( 11). This includes IFN-γ and tumor necrosis factor secretion by Th1 cells, which are thought to play a key role in mediating cellular immunity, and interleukin (IL)-4 and IL-5 by Th2 cells, which have been implicated in the humoral response ( 12). Another subtype, designated CD4⁺ Th17, has recently been described, defined by its capacity to secrete IL-17 and promote an inflammatory response involving macrophages and neutrophils ( 13). Interestingly, several studies have reported divergent roles for Th1 and/or Th2 CD4⁺ cells in mediating an antitumor immune response ( 14– 16). In addition, it has been reported that these subsets may induce antitumor effects in the absence of or in concert with CD8⁺ T cells ( 14, 17). Nevertheless, the question of whether CD4⁺ Th1 and/or Th2 cells are important for redirected T-cell therapy and the mechanism of how redirected CD4⁺ T cells provide help are completely unknown.

In this study, we have developed a robust method for generating polarized mouse Th1 and Th2 CD4⁺ cells transduced with an anti–erbB2 scFv chimeric receptor. We evaluated the ability of either transduced Th1 or Th2 CD4⁺ cells in conjunction with transduced CD8⁺ T cells to eradicate MDA-MB-435-erbB2 lung metastases and mediate resistance to tumor rechallenge.

Materials and Methods

Cell cultures and mice. MDA-MB-435 and MDA-MB-435-erbB2 human mammary carcinoma cells, 4T1.2 and 4T1.2-erbB2 mouse mammary carcinoma cells, and the ecotrophic retrovirus producer cell line GP+E86 harboring the scFv-α-erbB2-CD28-ζ chimera were cultured as previously described ( 10). BALB/c and BALB/c scid/scid (scid; 6–12 weeks of age) mice were purchased from The Walter and Eliza Hall Institute of Medical Research. BALB/c IL-4–deficient (BALB/c IL-4^−/−) and BALB/c perforin-deficient (BALB/c pfp^−/−) mice were bred at the Peter MacCallum Cancer Centre.

Generation of polarized, scFv-receptor–transduced CD4⁺ Th1 and Th2 cells. Transduction of mouse splenic T lymphocytes was done as described previously ( 18– 20). To generate transduced CD8⁺ and transduced polarized CD4⁺ T cells, each T-cell subset was initially isolated by labeling with anti-CD4 or anti-CD8 magnetic beads (Miltenyi Biotec) and passaged through a MACS depletion column. Enriched CD8⁺ and CD4⁺ T-cell cultures (10⁷ cells) were cocultured with retrovirus-producing packaging cells (5 × 10⁵) for 72 h in DMEM supplemented with 4 μg/mL polybrene, 5 μg/mL phytohemagglutinin (Sigma), and 100 units/mL human recombinant IL-2 (Proleukin, National Cancer Institute). In the final 24 h of the transduction period, the addition of anti–IL-4 monoclonal antibody (mAb; PharMingen; 1 μg/mL) and recombinant mouse IL-12 (Genetics Institute; 0.1 pg/mL) or anti–IFN-γ mAb (PharMingen; 2 μg/mL) and recombinant mouse IL-4 (PharMingen; 1 pg/mL) was used to generate CD4⁺ Th1 or Th2 cells, respectively ( Fig. 1 ).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Generation of polarized, transduced CD4⁺ T cells. Enriched splenic mouse T cells were labeled with magnetic beads and passed through a MACS depletion column to isolate CD4⁺ T cells. CD4⁺ T cells were cocultured with retrovirus-producing packaging cells for transduction of T cells with the scFv-CD28-ζ chimeric receptor. Recombinant IL-12 and anti–IL-4 mAb or recombinant IL-4 and anti–IFN-γ mAbs were added to coculture plates in the final 24 h of transduction for effective polarization of CD4⁺ Th1 or Th2 cells, respectively. No cytokine or antibody was added for nonpolarized CD4⁺ Thi cells (except IL-2).

Flow cytometry. Expression of the α-erbB2-CD28-ζ chimeric receptor on the surface of CD8⁺ or CD4⁺ T helper intermediate (Thi), Th1, and Th2 mouse cells was determined by indirect immunofluorescence with a c-myc tag antibody (Cell Signaling Technologies), followed by staining with a phycoerythrin-labeled antimouse immunoglobulin mAb (BD Biosciences). Background fluorescence was assessed with the phycoerythrin-labeled antimouse immunoglobulin mAb alone. Cell-surface phenotyping of transduced cells was determined by direct staining with FITC-labeled anti-CD4 (RM4-5; PharMingen) and phycoerythrin-labeled anti-CD8 (53-6.7; PharMingen) mAbs, as previously described ( 18– 21).

Cytotoxicity, cytokine secretion, and proliferation by CD4⁺ T-cell subsets. The ability of transduced CD4⁺ Thi, Th1, and Th2 cells to specifically induce target cell lysis was assessed in a 6-h chromium release assay, as previously described ( 22). The ability of these different CD4⁺ T-cell subsets to produce various cytokines [IFN-γ, IL-2, IL-4, and granulocyte macrophage colony-stimulating factor (GM-CSF)] after antigen ligation was determined by ELISA (PharMingen), and the proliferative capacity of these gene-modified T-cell populations was assessed in a [³H]thymidine incorporation assay, as described previously ( 18, 23).

Immunohistochemistry. H&E staining and immunohistochemistry were done on frozen sections as described previously ( 10).

In vivo experiments and T-cell persistence assays. The capacity of gene-modified Th1 and Th2 CD4⁺ cells to eradicate MDA-MB-435-erbB2 tumor cells and assays to determine T-cell persistence in the blood and tumor after rechallenge were as previously described ( 10, 24). A brief description of these experiments is included in Supplementary Materials and Methods.

Results

Generation of polarized, gene-transduced CD4⁺ Th1 and Th2 cells. Expression of the scFv-α-erbB2-CD28-ζ receptor on the surface of mouse Th1 and Th2 CD4⁺ T cells was determined by indirect immunofluorescence with an antibody to the c-myc tag engineered into the ectodomain. Equivalent levels of cell-surface receptor expression were shown on transduced CD4⁺ Th1, CD4⁺ Th2, and CD4⁺ Thi cells ( Fig. 2B ), as compared with binding of the secondary antibody alone. CD4⁺ Th1, Th2, and Thi transduced cell populations all consisted of >90% CD4⁺ T cells ( Fig. 2A). As previously shown, we also observed between 50% and 70% expression of the scFv chimeric receptor in isolated mouse CD8⁺ T cells from three experiments done (data not shown). We observed no anti-tag staining in T-cell subsets transduced with the control LXSN vector alone (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Expression of the scFv-CD28-ζ receptor in transduced CD4⁺ Thi, Th1, and Th2 mouse cells. Chimeric scFv receptor expression was detected in transduced CD4⁺ Thi, Th1, and Th2 cells (B) by flow cytometry following staining with an anti-tag mAb and phycoerythrin-labeled sheep antimouse immunoglobulin (solid line) or with the phycoerythrin-labeled secondary alone (dotted line). Isolated populations of transduced T cells consisted of >90% CD4⁺ T cells (A). Similar results were obtained in three experiments.

Antigen-specific cytokine secretion, proliferation, and lysis by transduced CD4⁺ Th1 and Th2 cells. We next tested the ability of engineered CD4⁺ Th1, Th2, and Thi cells to specifically respond to antigen-positive tumor targets in a number of in vitro assays. Transduced CD4⁺ Th1, Th2, and Thi cells were first used in an ELISA to determine their ability to specifically secrete cytokine in response to erbB2⁺ tumor targets. Following coculture with MDA-MB-435-erbB2 targets, transduced CD4⁺ Th1 cells produced high levels of IFN-γ and only background IL-4, whereas transduced CD4⁺ Th2 cells produced high levels of IL-4 and only background IFN-γ ( Fig. 3A and B ). This indicated that the respective CD4⁺ T-cell populations were effectively polarized. Transduced CD4⁺ Thi cells produced high levels of both IFN-γ and IL-4 cytokines against erbB2⁺ targets as previously observed ( 10). Cytokine secretion by Th1 and Th2 CD4⁺ cells was antigen specific because none of the transduced T cells responded to erbB2⁻ tumor targets ( Fig. 3A and B). In addition, we showed antigen-specific secretion of both IL-2 and GM-CSF from all CD4⁺ T-cell subsets ( Fig. 3C and D).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Antigen-specific cytokine secretion by transduced mouse CD4⁺ Th1, Th2, and Thi cells. To evaluate cytokine secretion, CD4⁺ Th1 (open columns), Th2 (dotted columns), and Thi cells (diagonally striped columns) transduced with the scFv-CD28-ζ receptor or mock-transduced CD4⁺ Th1 (closed columns), Th2 (cross-hatched columns), and Thi cells (horizontally striped columns) were cocultured with media alone, MDA-MB-435-erbB2 tumor cells, or MDA-MB-435 parental cells or stimulated with plate-bound anti-CD3 and anti-CD28 mAbs in 12-well plates for 24 h. Supernatants were harvested and assessed for IFN-γ (A), IL-4 (B), IL-2 (C), and GM-CSF (D) production by ELISA. Results are expressed as picograms per milliliter of cytokine secreted. Columns, mean of duplicate samples of three representative experiments; bars, SE.

We also observed similar antigen-specific proliferation and cytolytic ability by all gene-engineered CD4⁺ subsets (Supplementary Fig. S1). Collectively, the data showed that we could effectively generate scFv-receptor gene-modified polarized mouse CD4⁺ T cells, which were capable of antigen-restricted cytokine secretion, proliferation, and death.

Transduced CD4⁺ Th1 and Th2 cells equivalently reject established lung metastases in concert with transduced CD8⁺ CTL. We have previously shown enhanced survival of mice bearing 5-day MDA-MB-435-erbB2 tumors following coadministration of equivalent numbers of gene-engineered CD8⁺ and CD4⁺ T cells ( 10). To determine the capacity of CD4⁺ Th1 and Th2 subsets to reject tumor in this model, we compared the survival of mice bearing MDA-MB-435-erbB2 lung metastases treated 5 days after tumor inoculation with engineered CD4⁺ Th1, Th2, or Thi cells (5 × 10⁶) and an equivalent number of engineered CD8⁺ T cells (5 × 10⁶). Interestingly, we showed complete survival of all treated mice regardless of which transduced CD4⁺ T-cell subset they received ( Fig. 4A ). As controls, mice receiving a 1:1 ratio of CD8⁺ and CD4⁺ mock-transduced T cells or no T cells did not survive. Furthermore, no antitumor effect was observed in mice that received either 10⁷ transduced CD4⁺ Th1 or Th2 cells alone, indicating a requirement for the coinfusion of engineered CD8⁺ T cells to generate an optimal antitumor response ( Fig. 4A). This result indicated that transfer of either the transduced Th1 or Th2 CD4⁺ subset was equally capable of mediating a potent response against primary tumor challenge.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Transduced CD4⁺ Th1, Th2, and Thi cells equivalently reject established lung metastases in mice with an important role for IL-2, but not IL-4 or perforin. A, groups of five scid mice were given i.v. injections of 5 × 10⁶ MDA-MB-435-erbB2 cells at day 0 before i.v. injection of scFv-CD28-ζ–transduced T cells at day 5. Mice treated with a 1:1 ratio of CD4⁺ Thi transduced (5 × 10⁶) and CD8⁺ transduced (5 × 10⁶) T cells (○), CD4⁺ Th1 transduced (5 × 10⁶) and CD8⁺ transduced (5 × 10⁶) T cells (⧫), or CD4⁺ Th2 transduced (5 × 10⁶) and CD8⁺ transduced (5 × 10⁶) T cells (▪) all showed complete survival. No survival of mice was observed that received transfer of transduced CD4⁺ Th1 cells alone (•), transduced CD4⁺ Th2 cells alone (△), control CD8⁺ and CD4⁺ T cells transduced with the αCEA-γ receptor (□), or no T cells (▴). B, IL-2 production by engineered CD4⁺ T cells plays an important role in the antitumor response. Scid mice bearing 5-d MDA-MB-435-erbB2 tumor cells received a combined transfer of scFv-CD28-ζ–transduced CD8⁺ (5 × 10⁶) and CD4⁺ (5 × 10⁶) T cells. For in vivo IL-2 neutralization, mice were given i.p. injections of 0.5-mg anti–IL-2 neutralization antibody (S4B6) on days 4 and 5 and every subsequent 2 d up to day 30. Mice treated with a 1:1 ratio of CD4⁺ and CD8⁺ transduced T cells (▪) showed complete survival compared with reduced survival of mice that received CD4⁺ and CD8⁺ transduced T cells plus anti–IL-2 antibody (△). Control groups included mice treated with scFv-transduced CD8⁺ and CD4⁺ T cells and an irrelevant IgG2a control anti-MAC4 antibody (⧫) or mice treated with mock-transduced (LXSN vector) CD4⁺ and CD8⁺ transduced T cells alone (▴) or in combination with anti–IL-2 antibody (○). Other controls included mice that received no T-cell transfer (□), anti–IL-2 antibody alone (◊), or anti-MAC4 antibody alone (•). *, P < 0.05, Mann-Whitney test (▪ versus △). C, IL-4 production by engineered CD4⁺ T cells does not play an important role in the antitumor response. Scid mice were given i.v. injections of 5 × 10⁶ MDA-MB-435-erbB2 tumor cells followed by combined transfer of transduced CD8⁺ (5 × 10⁶) and CD4⁺ (5 × 10⁶) donor T cells at day 5 from BALB/c wild-type or BALB/c IL-4^−/− mice. Mice received either 1:1 transfer of scFv-CD28-ζ–transduced CD8⁺/CD4⁺IL-4^+/+ T cells (▪) or CD8⁺/CD4⁺IL-4^−/− T cells (▴). Control mice received a 1:1 transfer of wild-type mock-transduced CD8⁺ and CD4⁺ T cells (△) or no T-cell transfer (□). D, perforin production by engineered CD4⁺ T cells does not play an important role for tumor eradication. Scid mice were i.v. injected with 5 × 10⁶ MDA-MB-435-erbB2 tumor cells followed by combined transfer of transduced CD8⁺ (5 × 10⁶) and CD4⁺ (5 × 10⁶) donor T cells at day 5 from BALB/c wild-type or BALB/c pfp^−/− mice. Mice received 1:1 transfer of scFv-CD28-ζ–transduced CD8⁺pfp^+/+/CD4⁺pfp^+/+ T cells (▪), CD8⁺pfp^−/−/CD4⁺pfp^−/− T cells (△), CD8⁺pfp^−/−/CD4⁺pfp^+/+ T cells (□), or CD8⁺pfp^+/+/CD4⁺pfp^−/− (▴) T cells. Control mice received either a 1:1 transfer of mock-transduced CD8⁺ and CD4⁺ wild-type T cells (•) or no T-cell transfer (○). *, P < 0.05; **, P < 0.05, Mann-Whitney test. All results are calculated as the percentage of each group surviving and are representative of two experiments done. Arrows, day of T-cell transfer.

An important role for IL-2, but not IL-4 or perforin, by gene-modified CD4⁺ T cells in the antitumor response. In a previous study, we showed that IFN-γ produced by transduced CD8⁺ was critical for the antitumor response, although IFN-γ produced by transduced CD4⁺ T cells could partially compensate if IFN-γ production was absent in CD8⁺ T cells ( 10). Given that we showed significant levels of IL-2 produced by all gene-engineered CD4⁺ T-cell subsets following stimulation with erbB2⁺ target cells in vitro, we next wanted to determine whether the secretion of IL-2 by transduced CD4⁺ T cells was important in the antitumor response. Thus, mice bearing MDA-MB-435-erbB2 tumor were treated with i.p. injections of 0.5-mg anti–IL-2 neutralizing antibody (S4B6) at days 4 and 5 and every subsequent 2 days up to day 30 after tumor inoculation. The S4B6 antibody has been shown to neutralize IL-2 in vivo in a number of mouse models ( 25– 27). Mice that received 1:1 transfer of CD8⁺ and Thi CD4⁺ transduced T cells and anti–IL-2 antibody showed significantly reduced survival compared with mice that were treated with transduced CD8⁺ and Thi CD4⁺ T cells alone ( Fig. 4B). This result suggested an important role for IL-2 in the antitumor response, presumably by transduced CD4⁺ T cells given that IL-2 secretion was only detected by this T-cell subset. As specificity controls, mice treated with scFv-transduced CD8⁺ and Thi CD4⁺ T cells and an isotype control anti-MAC4 antibody all survived. However, mice treated with mock-transduced CD4⁺ and CD8⁺ T cells alone or in combination with anti–IL-2, or with anti–IL-2 or anti-MAC4 antibodies alone all succumbed to disease ( Fig. 4B). In further experiments using confocal microscopy and immunohistochemical staining on day 16 lung sections, we were unable to detect CD8⁺ or CD4⁺ T cells in four of eight sections analyzed after treatment with engineered T cells and anti–IL-2 neutralization antibody. In contrast, we detected CD8⁺ or CD4⁺ T cells in all lung sections from mice that received engineered T cells and isotype control antibody ( Fig. 5B ). Lack of T cells in lung sections correlated with increased tumor burden as indicated by H&E staining ( Fig. 5B). We also quantified the number of transduced CD8⁺ and CD4⁺ T cells from all lung sections analyzed. This result showed a significant reduction in number of transduced CD8⁺ and CD4⁺ T cells at the tumor site from mice treated with anti–IL-2 antibody compared with isotype control antibody ( Fig. 5C). Overall, these data suggested an important role for IL-2 in sustaining survival and persistence of T cells within the tumor microenvironment.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Detection of transduced CD4⁺ Th1, Th2, and Thi cells at the tumor site. Scid mice were given i.v. injections of MDA-MB-435-erbB2 tumor cells (5 × 10⁶) at day 0 followed by i.v. injection of gene-modified T cells at day 5. Lungs were harvested and prepared for immunohistologic analysis at day 16 after tumor inoculation. A, lung sections from mice that received 1:1 CD4⁺ Thi (5 × 10⁶) and CD8⁺ (5 × 10⁶) scFv-CD28-ζ–transduced T cells, CD4⁺ Th1 (5 × 10⁶) and CD8⁺ (5 × 10⁶) scFv-CD28-ζ–transduced T cells, CD4⁺ Th2 (5 × 10⁶) and CD8⁺ (5 × 10⁶) scFv-CD28-ζ–transduced T cells, or 1:1 CD4⁺ Thi (5 × 10⁶) and CD8⁺ (5 × 10⁶) scFv-CEA-γ transduced control T cells, were stained with H&E, with anti-CD4 (green) and anti-CD8 (red), with anti–c-myc (red), or with anti-Ly6G (green) and anti-CD11b (red) for macrophages and neutrophils. Representative fields of four sections analyzed. Original magnification, ×400. B, lung sections from mice that received 1:1 CD4⁺ Thi (5 × 10⁶) and CD8⁺ (5 × 10⁶) scFv-CD28-ζ–transduced T cells and i.p. injections of 0.5 mg IL-2 neutralization antibody (S4B6) or control anti-MAC4 antibody were stained with H&E or with anti-CD4 (green) and anti-CD8 (red). Two representative fields for each treatment of eight sections analyzed. Original magnification, ×400. C, the number of transduced CD8⁺ and CD4⁺ T cells in all day 16 lung sections from mice treated with gene-modified T cells and S4B6 (eight sections including four random fields per section) or control antibody (eight sections including four random fields per section) was quantitated and data represented as cells per field. The reduced number of transduced CD8⁺ or CD4⁺ T cells localized to the tumor site following anti–IL-2 treatment was significant compared with isotype control antibody (*, P < 0.05; **, P < 0.05, Mann-Whitney test).

Given that transduced CD4⁺ Th2 cells secreted IL-4 following receptor ligation in vitro, we next investigated whether IL-4 played a role in the antitumor response. IL-4 is a CD4⁺ T-cell–specific cytokine that has been shown to play a key role in CD4⁺ T-cell differentiation ( 28). In addition, a previous study showed severe impairment of tumor immunity in IL-4–deficient mice, suggesting a role for IL-4 in the priming of tumor-reactive effector cells ( 29). To assess the importance of this cytokine, we transduced and adoptively transferred donor CD4⁺ T cells from BALB/c wild-type or IL-4–deficient mice in combination with wild-type CD8⁺ transduced T cells. As in previous experiments, all mice bearing 5-day MDA-MB-435-erbB2 tumors survived following treatment with 1:1 transfer of CD8⁺ and CD4⁺ IL-4 wild-type transduced T cells ( Fig. 4C). This was also found to be the case in mice that received 1:1 transfer of CD8⁺ and CD4⁺IL-4^−/− transduced T cells. These data indicated that IL-4 secretion by transduced CD4⁺ T cells was not critical to the antitumor response. As before, mice receiving 1:1 transfer of mock-transduced CD8⁺ and CD4⁺ T cells or no T-cell transfer succumbed to disease ( Fig. 4C).

The cytolytic protein perforin has been shown to be a key mediator of target cell lysis by T cells ( 30, 31). We have previously shown perforin to play a key role in tumor eradication by transduced unfractionated T cells consisting predominantly of CD8⁺ T cells ( 18, 21). To further define the role of perforin in tumor eradication by gene-modified CD8⁺ and CD4⁺ T cells in our model, mice were treated with donor CD8⁺ and CD4⁺ Thi cells from BALB/c wild-type or perforin-deficient mice. Consistent with previous experiments, all mice bearing 5-day MDA-MB-435-erbB2 tumors survived after treatment with 1:1 transfer of perforin wild-type CD8⁺ and CD4⁺ transduced T cells ( Fig. 4D). Similarly, the majority of mice that received 1:1 transfer of CD8⁺pfp^+/+ and CD4⁺pfp^−/− survived, indicating that perforin secreted by CD4⁺ T cells was not playing a critical role in the antitumor effect. In contrast, significantly reduced survival of mice was observed following 1:1 transfer of CD8⁺pfp^−/− and CD4⁺pfp^+/+ transduced T cells. This indicated that perforin produced by transduced CD8⁺ T cells played an important role in rejection of this tumor ( Fig. 4D). Overall, these data have shown an important role for IL-2, but not IL-4 or perforin, by transduced CD4⁺ in the tumor rejection process.

Transduced CD4⁺ Th1 and Th2 cells localize to the tumor site. Having shown a role for both gene-modified CD4⁺ Th1 and Th2 cells, in conjunction with gene-modified CD8⁺ T cells, for the eradication of 5-day MDA-MB-435-erbB2 lung metastases, we were next interested in determining whether we could detect these transduced T cells at the tumor site. Lung sections of mice were first stained with H&E at day 16 after i.v. tumor injection. All mice treated with 1:1 transfer of scFv-α-erbB2-CD28-ζ–transduced CD8⁺ (5 × 10⁶) and CD4⁺ (5 × 10⁶) Th1, Th2, or Thi cells showed a significant reduction in lung tumor burden ( Fig. 5A) compared with mice treated with control transduced CD8⁺ and CD4⁺ Thi cells ( Fig. 5A). Confocal microscopy following immunohistochemical staining was used to show the presence of CD8⁺ and CD4⁺ Thi, Th1, or Th2 cells on day 16 lung sections of mice ( Fig. 5A). The detection of α-tag staining in lung sections of treated mice suggested that these T cells were also expressing the chimeric scFv-CD28-ζ receptor ( Fig. 5A). Transduced CD4⁺ Thi, Th1, or Th2 cells injected alone could not be detected at the tumor site (data not shown). In addition, no staining of CD8⁺ and CD4⁺ T cells or α-tag was observed in lung sections from mice treated with control CD8⁺ and CD4⁺ transduced with an irrelevant scFv-anti-CEA-γ receptor ( Fig. 5A). We also stained for the presence of neutrophils and macrophages in the various lung sections from mice but observed no differences between treatment and control groups ( Fig. 5A). Thus, these experiments suggested that the antitumor response in mice correlated with the presence of transduced CD8⁺ and CD4⁺ T-cell subsets at the site of tumor.

Superior antigen-specific secondary response mediated by transduced CD4⁺ Th1 cells. The induction of an efficient recall response is an important criteria for development of a truly successful cancer therapy. Using the scFv approach, we have previously shown that long-term surviving mice (>100 days) initially treated with transduced CD8⁺ and CD4⁺ Thi cells could effectively mediate a response to secondary tumor rechallenge ( 10). We were now interested in testing the ability of mice treated with transduced CD8⁺ and either CD4⁺ Th1 or Th2 cells to mediate an antigen-specific recall response to a heterologous tumor challenge (4T1.2). For these experiments, long-term surviving mice (>100 days) were given s.c. injections of 5 × 10⁴ 4T1.2-erbB2 or 4T1.2-parental mouse mammary carcinoma cells and monitored for tumor growth and survival. Interestingly, mice treated with gene-modified CD4⁺ Th1 and CD8⁺ T cells showed significant delay of 4T1.2-erbB2 tumor growth and prolonged survival compared with unimmunized mice ( Fig. 6A and B ). No significant recall response was observed in mice that received transduced CD4⁺ Th2 cells, whereas mice that had been treated with Thi CD4⁺ cells showed an intermediate phenotype. This recall response was specific because none of the mouse groups showed resistance to the erbB2-deficient parental tumor cells ( Fig. 6A and B). We used both flow cytometry and confocal microscopy to determine whether these different transduced T-cell populations expanded in the peripheral blood of mice and localized to the tumor site after rechallenge. Interestingly, we showed an increased expansion of both transduced Th1 CD4⁺ and CD8⁺ T cells in peripheral blood of mice at various time points compared with Th2 and Thi CD4⁺ cells following rechallenge with a s.c. injection of 4T1.2-erbB2 cells (Supplementary Fig. S2A and B). Importantly, the number of the various CD4⁺ T-cell subsets and CD8⁺ T cells in peripheral blood of mice was equivalent before tumor rechallenge. In addition, we also showed by flow cytometry increased percentages of both Th1 CD4⁺ cells and CD8⁺ T cells at the tumor site compared with Thi and Th2 CD4⁺ cells at two different time points (Supplementary Fig. S2C and D). This result was consistent with confocal microscopy data showing increased level of staining for both Th1 CD4⁺ cells and CD8⁺ T cells at the tumor site compared with Thi and Th2 CD4⁺ cells ( Fig. 6C). There was no staining of lung sections from mice injected with 4T1.2-erbB2 tumor alone ( Fig. 6C). Overall, these results have shown for the first time that gene-engineered CD4⁺ Th1 cells are critical for mounting an effective response to secondary tumor challenge delivered months after the initial tumor had been treated.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.

Transduced CD4⁺ Th1 cells mount a superior antigen-specific response to tumor rechallenge. A, long-term surviving mice (>100 d after primary MDA-MB-435-erbB2 tumor inoculation) received a s.c. injection of 5 × 10⁴ mouse mammary carcinoma 4T1.2-erbB2 or 4T1.2-parental cells and their growth was monitored. Mice that had previously received CD4⁺ Th1 (5 × 10⁶) and CD8⁺ (5 × 10⁶) T cells transduced with the scFv-CD28-ζ receptor showed stronger inhibition of 4T1.2-erbB2 tumor growth (▪) than mice initially treated with transduced CD8⁺ and CD4⁺ Th2 (▴) or Thi cells (□). There was no inhibition of 4T1.2-parental tumor growth in mice initially treated with transduced CD8⁺ and CD4⁺ Th1 (⧫), Th2 (•), or Thi cells (△). As controls, scid mice received either 5 × 10⁴ 4T1.2-erbB2 (◊) or 4T1.2-parental (○) tumor cells. *, P < 0.05, ▪ versus ▴ (Mann-Whitney test). B, survival of long-term surviving mice following tumor rechallenge with 5 × 10⁴ 4T1-erbB2 or 4T1.2-parental cells was compared. Mice initially treated with 1:1 transduced CD4⁺ Th1 and CD8⁺ T cells showed enhanced survival following 4T1.2-erbB2 tumor inoculation (▪) compared with mice initially treated with transduced CD8⁺ and CD4⁺ Th2 (▴) or Thi cells (□). Mice rechallenged with 4T1.2-parental tumor following initial treatment with transduced CD8⁺ and CD4⁺ Th1 (⧫), Th2 (•), or Thi cells (△) all succumbed early to disease. Control scid mice that received either 5 × 10⁴ 4T1.2-erbB2 (◊) or 4T1.2-parental (○) tumor cells also succumbed early to disease. *, P < 0.05; **, P < 0.005 (Mann-Whitney test). Points, mean tumor size (mm²; A) or mean percentage of each group surviving (B); bars, SE. C, long-term surviving mice (>100 d after primary MDA-MB-435-erbB2 tumor inoculation) received a s.c. injection of 5 × 10⁴ mouse mammary carcinoma 4T1.2-erbB2 cells and tumor tissues removed and prepared for immunohistologic analysis at days 13 and 22 after rechallenge. Tumor sections from mice that initially received 1:1 CD4⁺ Th1 (5 × 10⁶) and CD8⁺ (5 × 10⁶) scFv-CD28-ζ–transduced T cells, CD4⁺ Th2 (5 × 10⁶) and CD8⁺ (5 × 10⁶) scFv-CD28-ζ–transduced T cells, CD4⁺ Thi (5 × 10⁶) and CD8⁺ (5 × 10⁶) scFv-CD28-ζ–transduced T cells, or normal scid mice with tumor alone were stained with anti-CD4 (green) and anti-CD8 (red). Representative fields of five sections analyzed. Original magnification, ×400.

Discussion

Adoptive immunotherapy involving transfer of genetically modified T cells is a highly specific and powerful approach for treatment of disease. T cells modified to express chimeric receptors at their cell surface have been shown to provide effective responses against cancer in preclinical animal models and were found to be safe in HIV and cancer patients ( 8, 10, 32). In addition, such studies have shown that inclusion of engineered CD4⁺ T cells in combination with CD8⁺ T cells was critical for generating an optimal response. The exact mechanism used by CD4⁺ T cells in antitumor immunity is largely undefined and is further complicated by the fact that the CD4⁺ T cells can be principally divided into Th1 and Th2 subsets. In previous studies, divergent roles have been attributed to CD4⁺ Th1 and/or Th2 cells in antitumor immunity, with reports defining Th1 cells as principal mediators of cellular immunity and Th2 cells as mediators of humoral immunity ( 33, 34). In addition, it has been reported that these subsets may function alone or in combination with CD8⁺ T cells to mediate an antitumor effect ( 14, 15). Nevertheless, the role of CD4⁺ Th1 and Th2 cells in redirected T-cell therapy remains completely unknown. In this study, we have developed a robust and reproducible method for the generation of polarized CD4⁺ Th1 and Th2 cells engineered with an anti–erbB2-CD28-ζ scFv chimeric receptor. These cells were characterized by their ability to secrete either IFN-γ (Th1 cytokine) or IL-4 (Th2 cytokine) following receptor ligation. In adoptive transfer studies, we showed an equivalent ability of these gene-modified Th1 or Th2 CD4⁺ cells, in combination with engineered CD8⁺ T cells, to reject primary MDA-MB-435-erbB2 lung disease, which correlated with localization of gene-modified T cells at the tumor site. Furthermore, we showed for the first time that mice treated with gene-modified Th1 CD4⁺ and CD8⁺ T cells could more effectively inhibit a secondary tumor challenge. Thus, for effective immunotherapy involving gene-engineered T cells, our results strongly support the use of CD4⁺ Th1 cells for inducing a durable antitumor response.

An interesting finding of our study was that neutralization of IL-2 in vivo resulted in a decreased number of surviving mice following adoptive T-cell transfer. It has previously been reported that IL-2 plays an important role for maintaining CD8⁺ T-cell function in vivo ( 24, 35, 36). Given that we observed good levels of IL-2 secretion from transduced CD4⁺ Th1, Th2, and Thi subsets, but not from transduced CD8⁺ T cells, it was possible that the provision of IL-2 cytokine “help” from transduced CD4⁺ T cells may play a role in sustaining function of gene-modified CD8⁺ T cells in our model. We showed that the administration of the anti–IL-2 neutralization antibody S4B6 significantly reduced survival of mice bearing established MDA-MB-435-erbB2 lung metastases following treatment with gene-modified CD8⁺ and CD4⁺ T cells. This result indicated that IL-2 (presumably secreted by gene-modified CD4⁺ T cells in vivo) was indeed important for the antitumor response. Previously, we showed that exogenous administration of low-dose or high-dose human recombinant IL-2, together with transduced CD8⁺ T cells, could not mimic the effect of transferring both transduced CD8⁺ and CD4⁺ T cells ( 10). Thus, localized production of IL-2 by transduced CD4⁺ T-cell subsets in the tumor microenvironment seems to be a critical factor for sustaining an effective antitumor response.

A number of studies have reported divergent roles for Th1 and Th2 CD4⁺ cells in the antitumor response. In one study using OVA-restricted T cells, CD4⁺ Th1 cells were reported to induce tumor regression by cellular immunity whereas CD4⁺ Th2 cells were shown to destroy tumor by necrosis ( 15). These effects required the presence of antigen-specific CD8⁺ T cells. However, it was not clear in this study what effectors were actually mediating tumor rejection in this model following T-cell transfer. In contrast, another study also using OVA-restricted T cells reported that CD4⁺ Th2 cells, but not CD4⁺ Th1 cells, were important for the clearance of metastatic melanoma ( 14). The clearance of metastases by CD4⁺ Th2 cells was found to be mediated by eosinophils infiltrating the tumor site and was not dependent on transfer of CD8⁺ T cells. In our study, we showed that tumor rejection correlated with the presence of both transduced CD8⁺ and CD4 T-cell subsets at the tumor site. Tumor rejection required transfer of transduced CD8⁺ and transduced CD4⁺ T cells (Th1, Th2, or Thi) because transfer of any of these subsets alone had no antitumor effect. Interestingly, we could not detect any difference in the presence of neutrophils or macrophages at day 16 between treatment and control groups of mice. However, we cannot rule out the possibility that these types of effector cells may be playing a role at earlier time points. Given that all transduced CD4⁺ T-cell subsets could secrete significant levels of GM-CSF, release of this cytokine may play a role for attracting macrophages to the tumor site as has previously been reported in other models ( 37, 38). Other reports have also shown that secretion of IL-4 from CD4⁺ T cells could mediate antitumor effects by recruitment of macrophages ( 39, 40) and that tumors transfected with IL-4 could be rejected by eosinophils and neutrophils ( 41). However, recruitment of these different effectors via IL-4 does not seem to be important for tumor rejection in our model given that transfer of donor CD4⁺ T cells from IL-4^−/− mice in combination with transduced CD8⁺ T cells resulted in complete survival of mice.

Another interesting aspect of this study was that perforin production by transduced CD8⁺ T cells, but not transduced CD4⁺ T cells, played a significant role in eradication of established lung disease. We and others have previously shown that gene-modified mouse and human CD8⁺ and CD4⁺ T cells could both effectively kill tumor targets in vitro ( 10, 42) although we reported much weaker cytolytic activity by mouse transduced CD4⁺ T cells ( 10). We have also previously shown that perforin plays a critical role for eradication of s.c. tumor by unfractionated transduced T cells consisting predominantly of CD8⁺ T cells ( 21). Thus, it was of interest to test whether perforin secreted by transduced CD4⁺ T cells contributed to tumor rejection in our model. In our study, we observed no significant decrease in survival of mice following transfer of transduced pfp^−/− CD4⁺ T cells in combination with CD8⁺pfp^+/+ T cells. This suggested that transduced CD4⁺ T cells do not directly affect the tumor but rather have an indirect role. Another interesting result to arise from this study was that ∼40% of tumor-bearing mice treated with both CD8⁺ and CD4⁺ transduced T cells deficient in perforin survived. In contrast, we have previously shown that all mice succumbed to MDA-MB-435-erbB2 lung disease following treatment with transduced CD8⁺ and CD4⁺ T cells from IFN-γ^−/− mice ( 10). These results indicated that IFN-γ production by transduced CD8⁺ and CD4⁺ T cells played a more critical role for tumor rejection in this model than perforin. Hence, the data suggest that the role of different effector molecules such as IFN-γ and perforin may vary depending on the anatomic site of disease, tumor type, and microenvironment. Indeed, several studies have reported a greater dependence on perforin for rejection of tumors of s.c. origin ( 43, 44). Thus, future receptor designs may be tailored to enhance specific T-cell responses for tumors at different locations.

A striking feature of the current study was that mice cured of lung disease following treatment with gene-modified CD4⁺ Th1 cells in combination with transduced CD8⁺ T cells could more effectively inhibit a secondary tumor challenge. This result correlated with both increased expansion of Th1 CD4⁺ and CD8⁺ T cells in the blood and a greater number of these cells localizing to the tumor site following rechallenge. This reflected a greater ability of Th1 CD4⁺ cells to induce a stronger recall response. Thus, utilization of gene-modified CD4⁺ Th1 cells may represent a promising new strategy for the generation of a durable and effective response in patients, particularly in cases of tumor relapse. A previous study reported that CD4⁺ Th1 cells could induce a stronger immunologic memory response than CD4⁺ Th2 cells, although in this study mice treated with either subset were shown to similarly reject a secondary challenge ( 15). The mechanism(s) behind this increased recall response by transduced Th1 cells in our study is not clear at this stage but it seems that local production of type I cytokines by these cells may help facilitate and sustain antitumor function of transduced CD8⁺ T cells where production of type II cytokines from transduced Th2 cells may be less effective at this task. Further studies are required to fully elucidate the mechanism of enhanced secondary rejection by gene-modified Th1 CD4⁺ cells.

The experiments done in this study using scid mice do bear relevance to the clinic in patients that receive a nonmyeloablative regimen followed by adoptive transfer of tumor-reactive T cells ( 4). Nevertheless, in future experiments, it would be informative to test the ability of combined transfer of transduced Th1 CD4⁺ T cells and CD8⁺ to specifically reject tumor in an immunocompetent mouse setting. A potential model may include the use of transgenic mice tolerant to the human erbB2 antigen ( 45).

In conclusion, our findings have provided for the first time important insight into the role of gene-modified CD4⁺ Th1 and Th2 cells, in combination with gene-modified CD8⁺ T cells, for cancer immunotherapy. Whereas transfer of all transduced CD4⁺ T-cell subsets was able to eradicate primary tumor in an antigen-specific manner, mice treated with gene-engineered CD4⁺ Th1 cells could more effectively inhibit secondary tumor challenge. Thus, immunotherapeutic strategies using gene-modified CD4⁺ Th1 cells may be more desirable in future clinical trials. Several reports have documented that tumor-reactive human CD4⁺ T cells with a Th1-dominant phenotype can be generated using several Th1 cytokines including IFN-γ, IL-2, and IL-12 ( 15, 46). Thus, the feasibility of using gene-modified CD4⁺ Th1 cells for cancer immunotherapy seems to be high and may significantly advance the application of gene-engineered T-cell therapy for the treatment of cancer.