- Top of page
- Material and methods
There is an increasing awareness of the therapeutic potential for combining immune-based therapies with chemotherapy in the treatment of malignant diseases, but few published studies evaluate possible cytotoxic synergies between chemotherapy and cytotoxic immune cells. Human Vα24+/Vβ11+ NKT cells are being evaluated for use in cell-based immunotherapy of malignancy because of their immune regulatory functions and potent cytotoxic potential. In this study, we evaluated the cytotoxicity of combinations of chemotherapy and NKT cells to determine whether there is a potential to combine these treatment modalities for human cancer therapy. The cytotoxicity of NKT cells was tested against solid-tumor derived cell lines NCI-H358, DLD-1, HT-29, DU-145, TSU-Pr1 and MDA-MB231, with or without prior treatment of these target cells, with a range of chemotherapy agents. Low concentrations of chemotherapeutic agents led to sensitization of cell lines to NKT-mediated cytotoxicity, with the greatest effect being observed for prostate cancer cells. Synergistic cytotoxicity occurred in an NKT cell in a dose-dependent manner. Chemotherapy agents induced upregulation of cell surface TRAIL-R2 (DR5) and Fas (CD95) expression, increasing the capacity for NKT cells to recognize and kill via TRAIL- and FasL-mediated pathways. We conclude that administration of cytotoxic immune cells after chemotherapy may increase antitumor activities in comparison with the use of either treatment alone. © 2006 Wiley-Liss, Inc.
Chemotherapy remains the primary treatment for many advanced malignancies. However, chemotherapy alone is rarely curative, particularly for solid tumors.1 Chemotherapeutic agents are cytotoxic through a range of mechanisms that lead to antitumor activity, largely via induction of apoptotic cell death. Many normal cells are similarly affected, leading either to substantial toxicity or limitations in the doses that can be administered. In addition, chemotherapy is often ineffective, or only temporarily effective, as malignant cells have the capacity to develop mechanisms to resist or escape the cytotoxic effects.2 Additional modalities of treatment that possess greater selectivity for tumor targets, alternative cytotoxic mechanisms and noncross reactive toxicities are required to enhance outcomes of therapy for malignancy.
Immune-based therapies, including monoclonal antibodies and cell-based therapies, have all of these characteristics. Combinations of monoclonal antibody therapy with multiagent chemotherapy have already been shown to have greater survival benefits than when chemotherapy is used alone.3 Currently, despite the expanding literature attesting to the therapeutic benefits of cell-based immune therapies; there remains little data evaluating the potential therapeutic benefits of their use in combination with chemotherapy. Studies that have addressed this are encouraging, and demonstrate that additional preclinical and clinical evaluation is justified.4 There is a range of cellular immune therapy strategies that may be applicable. Vaccine-based immune therapies, such as dendritic cell vaccines targeted at stimulating CD8+ cytotoxic T cells (CTLs), have been intensively evaluated with many documented tumor regressions in clinical trials (reviewed in Ref.5); however, treatments scheduling difficulties associated when combining with antitumor agents, such as chemotherapy, have slowed progress in developing combination strategies. In contrast, adoptive immune therapy may be more suited for combination therapy, particularly with chemotherapy. As an example, immune cells can be infused after chemotherapy to avoid chemotherapy-induced suppression of the immune cells. Via this immunosuppressive effect, certain chemotherapy agents may also enhance the environment of the transferred cells by depleting the lymphocytes, such as CD4+CD25+ T regulatory cells (Tregs), that may otherwise suppress the antitumor activity or reduce the survival of the infused cells.6 Adoptively transferred cells can either be in vitro-activated autologous cells (including T cells, natural killer T (NKT) cells or natural killer (NK) cells) or allogeneic cells from matched donors after stem cell transplantation.
There are a number of potential mechanisms by which chemotherapeutic agents can sensitize tumors to immune-mediated killing, leading to additive and potentially synergistic antitumor activity (reviewed in Ref.1). Studies have demonstrated that chemotherapy can sensitize tumor targets to tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) mediated killing and that this can occur in vitro as well as in vivo.7, 8, 9, 10 One immune effector cell, the NKT cell, may be of particular clinical interest and provides an excellent model to investigate combinations of chemotherapy agents with cytotoxic cells. Human NKT cells are characterized by a Vα24/Vβ11 T cell receptor (TCR) coexpressed with NK cell markers, particularly CD161.11 They are of interest in the clinical setting, because of the relative ease with which they can be generated in vitro, their broad applicability across a range of tumor types without the need for peptide specificity as required for conventional T cells, their well-documented antitumor activities in vitro and in vivo and their capacity to recognize and kill tumor cells via TRAIL and other mechanisms (reviewed in Refs.12, 13, 14). Following activation, NKT cells upregulate TRAIL and Fas ligand (FasL) resulting in apoptosis in TRAIL-receptor and Fas expressing cells, respectively,15 and contain cytoplasmic perforin leading to perforin/granzyme-mediated killing.16 In addition, activated NKT cells possess chemokine receptors that allow them to migrate to tumor sites, allowing a greater chance for direct interactions with tumor cells.17 Furthermore, secretion of cytokines, such as IFN-γ, by NKT cells at the tumor site recruits and activates other components of the immune system, potentially enhancing the antitumor response.
In this study, we evaluated whether cytotoxic chemotherapy can be used to sensitize malignant targets to NKT cell killing and demonstrate this to be the case for a range of tumor types, using a range of chemotherapy agents. We also show that mechanisms through which this occurs include chemotherapy induced upregulation of cell surface TRAIL receptors and Fas, leading to synergistic killing by NKT cells via these pathways.
- Top of page
- Material and methods
Upon clinical detection or at episodes of relapse, many tumors are resistant to single modality therapy, regardless of whether chemotherapy, radiotherapy or immune-based therapy is applied. Increasing chemotherapy doses overcome tumor resistance in some situations, but are toxic and often only partly effective. Cell-based immune therapies have the potential to be highly effective in the management of cancer; however, tumors often escape immune control through lack of immunogenicity or by resistance to cell-mediated cytotoxicity.23 Poor immunogenicity may be overcome by such strategies as active vaccination with dendritic cells (or other adjuvants) or by the adoptive transfer of immune cells generated in vitro (reviewed in Refs.24 and25). Both of these approaches are under investigation, and attempts are being made to increase their potency. However, neither strategy will be effective when there is tumor resistance to immune-mediated killing.
To address the possibility that a combination of chemotherapy and immunotherapy may overcome resistance observed when these modalities are used alone, we evaluated the potential for chemotherapy to sensitize tumor cells to cell-mediated immune killing.4 Our results, using combinations of chemotherapy and immune effector cells that can be readily used in the clinical setting, provide encouraging evidence that there may be clinical benefits of this combination.
We used human NKT cells as a model for our studies for several reasons. There is extensive evidence for cytotoxic antitumor activity of NKT cells against a large range of tumor types (reviewed in Ref.13), including human tumors.18, 26, 27, 28, 29 Activated NKT cells possess a number of mechanisms for tumor recognition and killing, increasing the possibility that chemotherapy may enhance NKT cell-mediated killing by at least one of these pathways. Furthermore, tumor recognition by NKT cells is not limited to the well-defined HLA-restricted antigens, required for evaluation of conventional CTL killing. Also, autologous or HLA-matched effector–target pairing is not required. This enhanced our ability to rapidly screen for antitumor effects against a range of tumor cell types. Our specific aim was to determine whether low concentrations of chemotherapy agents have the potential to augment NKT cell antitumor activity in comparison with NKT cells alone. We hypothesized that pretreatment of these tumor cell lines with chemotherapy agents would render them more sensitive to killing by NKT cells.
Tumor targets were pretreated with chemotherapy agents, rather than exposed simultaneously to chemotherapy and NKT cells, to allow time for chemotherapy to sensitize tumor cells whilst avoiding any potential inhibition of NKT functionality that may be caused by exposure of NKT cells to chemotherapy. Previous data using prostate cancer cell lines as targets indicated that sequential, but not simultaneous, administration of chemotherapy and immunotherapy led to synergistic cytotoxicity.9 Also, this sequence of exposure is compatible to what can be achieved in the clinical setting, with administration of chemotherapy at an interval prior to administration of cytotoxic immune cells.
Our results demonstrate that pretreatment of tumor cell lines with chemotherapy agents effectively sensitizes a range of tumor cell types to NKT cell-mediated cytotoxicity, resulting in additive or synergistic antitumor effects (Figs. 3 and 4). The greatest synergistic antitumor activity involved the DU-145 prostate cell line, effectively sensitized to NKT-mediated killing by both etoposide and doxorubicin. The observed cytotoxicity was via induction of apoptosis (Fig. 4), and this occurred rapidly (within 4 hr) after exposure of the sensitized cells to NKT cells.
Chemotherapy can induce tumor cells to undergo apoptosis by initiating the mitochondrial apoptotic pathway30 or by sensitizing tumor cells to apoptosis via extrinsic mechanisms, by altering the death receptor apoptotic pathway31; the pathway utilized by TNF superfamily members, such as TRAIL and FasL expressed on immune effector cells, including NKT cells.15 Our observation that blocking TRAIL and FasL pathways reduced killing of chemotherapy-treated tumor cells suggests that these pathways are important in chemotherapy-induced sensitization. Possible mechanisms for this include alteration of surface expression of TRAIL receptors or FAS32, 33 or through modification of downstream intracellular pathways leading to apoptosis.34, 35 To provide preliminary information on the first of these possibilities, we examined surface expression of death receptor molecules for both TRAIL (DR4 and DR5) and FasL (Fas). Constitutive expression of both Fas and DR5 was observed on all the cell lines examined, initially suggesting that this was not an explanation for chemotherapy-induced sensitization (Fig. 5a). However, exposure to chemotherapy increased the level of expression of both of these molecules (Fig. 5b), and it is conceivable that the observed increase in expression was sufficient to enhance the NKT cytotoxicity. This possibility is supported by our observation that TRAIL- and FasL-transfected C1R cells also showed greater cytotoxicity toward the cell lines, following target cell exposure to chemotherapy. Owing to variable expression and lack of upregulation of DR4 (data not shown), it appears that sensitization to TRAIL-induced cell death is contributed to the DR5 receptor. We did not evaluate the possible role of downstream, intracellular components of this apoptotic pathway to chemotherapy sensitization.
Pathways other than death ligand–receptor interactions, such as perforin/granzyme release or recognition of stress-inducible MICA or MICB molecules by NKG2D, may serve as alternative mechanisms of NKT cell cytotoxicity. Perforin has shown to be the major mechanism of leukemic cell destruction18 and was observed in our study, and others, to be a major contributor to U937 cell killing by NKT cells.16, 36 Of interest, our results indicate that the inherent capacity of NKT cells to kill U937 cells is increased by pretreatment with chemotherapy and that the perforin/granzyme pathway is a major mediator of this chemotherapy-induced sensitization (data not shown). Many tumor cell lines are known to constitutively express MICA/B molecules,37 supported by our own observations (not shown). Killing may be induced if MICA/B is recognized by NKG2D expressed by NK cells and T cells, including NKT cells.38, 39 In the current study, MICA/B surface expression was not increased following exposure to chemotherapy, and it is therefore quite unlikely that the NKG2D pathway was involved in the chemotherapy-induced sensitization of the cell lines. Hence, no further investigation of this pathway was undertaken. Previous studies have shown both CD1d-dependent16, 40 and CD1d-independent41, 42 mechanisms of tumor cell cytotoxicity by NKT cells. Complete deficiency of CD1d surface expression in the cell lines of this study suggests that chemotherapy-induced cytotoxicity by NKT cells is independent of CD1d expression.
There remains the possibility that increased activation of NKT cells following exposure to the cell lines sensitized with chemotherapy contributes to the increased cytotoxicity of these targets, rather than higher susceptibility of the targets themselves. We have previously optimized NKT cell in vitro culture to maximize activation assessed by CD69 and CD25 expression, cytotoxic functionality and IFN-γ production, at day 7 of culture.43 We used this information to ensure that the timing of coculture with the targets was such that the NKT cells were highly stimulated and activated. As we had already confirmed substantial activation according to multiple phenotypic and functional parameters, we considered it unlikely that further increases in activation would occur in the presence of tumor cells. On the basis of this assumption, we concluded that the increased cytotoxicity observed after exposure of tumor targets to chemotherapy was not related to increased activation of NKT cells.
The greatest synergistic cytotoxicity of the combined chemotherapy/NKT cell treatment was against an androgen-independent prostate cancer cell line. Chemotherapy is becoming an increasingly important modality in the treatment of hormone refractory prostate cancer.44, 45 In addition, there are a number of studies indicating the efficacy of immune-based therapy for prostate cancer (reviewed in Refs.46, 47, 48). Therefore, there is a sound basis upon which to propose treatment of prostate cancer, with a combination of chemotherapy and immune therapy. Further work is required to define other disease situations where a combination of cell-based immune therapy and chemotherapy may have a clinical role. Our observations that chemotherapy sensitizes a range of malignancies to immune killing, and that this sensitization occurs through a range of mechanisms, indicates that the concept of chemotherapy sensitization may be broadly applicable across a range of malignancies, immune effector cell types and chemotherapeutic agents.
Most recent studies evaluating immune therapy for cancer have utilized vaccination strategies involving in vivo activation and expansion of immune effector cells. It may be counter-productive to combine these approaches with chemotherapy, because of the potential for the chemotherapy agents to target the proliferating immune cells. In contrast, administration of immune effector cells after chemotherapy is feasible, which would pose no risk of any disease-specific immune suppression and would allow optimal and flexible timing of combinations of chemotherapy with populations of immune effector cells. There is a range of possible immune effector cells that can be used, and they could be either allogeneic (post stem cell transplantation) or in vitro-activated autologous cells, such as NKT cells. Protocols are now available to selectively expand NKT cells to large numbers in vitro. Therefore, their administration as part of combined chemo-immune therapy strategies is feasible and clinical evaluation justified. It is quite likely that our findings will be applicable to other types of immune effector cells (T cells, NK cells, γ/δ T cells) that can be activated and expanded in vitro or obtained from allogeneic donors, post stem cell transplantation.
We conclude that chemotherapy can sensitize tumor cells to immune-mediated killing and that adoptive transfer of in vitro-activated NKT cells following administration of selected chemotherapy agents may substantially increase the antitumor effects compared with the use of these agents alone. This combination approach may increase the clinical efficacy of current chemotherapy protocols or may facilitate the use of lower doses of chemotherapy, reducing toxicity whilst maintaining efficacy.