Witch Hunt against Tumor Cells Enhanced by Dendritic Cells

Transcription

1 CANCER VACCINES Witch Hunt against Tumor Cells Enhanced by Dendritic Cells Clara Locher, a,b,c Sylvie Rusakiewicz, a,b Antoine Tesnière, b,d François Ghiringhelli, e,f Lionel Apetoh, a,b,c Guido Kroemer, b,d and Laurence Zitvogel a,b,c a Institut National de la Santé et de la Recherche Médicale, U805, Villejuif, France b Institut Gustave Roussy, Villejuif, France c Université Paris-Sud, Villejuif, France d Institut National de la Santé et de la Recherche Médicale, U848, Villejuif, France e AVENIR Team, Institut National de la Santé et de la Recherche Médicale, Dijon, France f Centre Georges Francois Leclerc, Dijon, France Conventional cancer treatments mediate their effects via the direct elimination of tumor cells. Nonetheless, recent evidence indicates that radiotherapy and some chemotherapeutic agents can also induce specific immune responses that contribute to therapeutic outcomes. Two major tumor-intrinsic changes that determine the immune response against tumors have been identified: the translocation of calreticulin to the plasma membrane and the release of high-mobility group box 1 protein. Together, these changes improve engulfment and processing of apoptotic bodies by dendritic cells, which are involved in the cross-priming of antitumor T lymphocytes in vivo. We review these two molecular mechanisms that dictate the radio/chemotherapy-elicited antitumor immune response and discuss how this knowledge can be clinically exploited to predict and also ameliorate the success of chemo/radiotherapy. Key words: cancer; radiotherapy; immunotherapy; chemotherapy; dendritic cells; calreticulin; ERp57; high-mobility group box 1; apoptosis; toll-like receptor Technologies for boosting the immune system against tumor cells have enormous potential to improve current efficacy of cancer therapy. It is plausible that an immune response elicited against the cancer cells might contribute to the complete eradication of chemotherapyresistant cancer cells and maintain micrometastasis in a state of dormancy. But until now, the possibility that immune reactions might contribute to the efficacy of anticancer therapies have been widely overlooked. At least three main reasons discouraged oncologists from performing extensive studies on the role of the immune system in an- Address for correspondence: Laurence Zitvogel, U805 and CIC BT507 INSERM, Institut Gustave Roussy, 39 rue Camille Desmoulins, F Villejuif, France. Voice: ; fax: zitvogel@igr.fr ticancer therapies. First, cancer cells escape immune responses either by selection of nonimmunogenic tumor-cell variants or by active suppression of the immune response or both. For example, tumor cells downregulate or lose the expression of human leukocyte antigen (HLA) class I molecules to avoid a T-cell-mediated immune response. They can also release factors, such as IL-10 or tumor growth factor-β, which are known to inhibit the immune response. 1 Second, many of the therapeutic procedures that are currently used in oncology are highly immunosuppressive. Indeed, most chemotherapeutic regimens in advanced malignancies promote or enhance lymphopenia, which is a dismal prognostic factor for progression-free survival. 2 Furthermore, the high doses of glucocorticoids mainly prescribed to alleviate chemotherapy-associated Cancer Vaccines: Ann. N.Y. Acad. Sci. 1174: (2009). doi: /j x c 2009 New York Academy of Sciences. 51

2 52 Annals of the New York Academy of Sciences nausea and vomiting also contribute to the immunosuppressive side effects of anticancer drugs. 3 Third, chemotherapy-induced tumorcell death occurs frequently through apoptosis, a cell-death modality believed to be an immunologically silent process. 4 Altogether, given the intrinsic tolerogenic properties of tumor cells, the immunosuppressive side effects of anticancer drugs, and the poor immmunogenicity of apoptosis, combining conventional therapies with immunotherapy was considered somewhat counterintuitive. However, there has been accumulating evidence that immune responses may contribute to the eradication of cancer cells during therapy with cytotoxic drugs. In 1973, the antitumor activity of two anthracyclines was compared in immunocompetent versus immunocompromised mice and revealed that doxorubicin and daunorubicin cooperate with the host immunological defenses to induce their therapeutic effects. These findings were later corroborated in various experimental models in which other anthracyclines, such as mitoxantrone, idarubicin, or epirubicin, were shown to boost the host s immune system and improve the efficacy of chemotherapy. Anticancer agents other than anthracylines have now also been shown to mediate, in addition to their direct cytotoxic effects, stimulating effects on the immune system. 5 The mechanisms by which conventional anticancer therapies interfere in the complex interaction between the tumor and the host immune system are currently under intense investigations. It appears that various cell death modalities, depending on the cell death inducer and on the intrinsic characteristics of the tumor itself, can impact on the ensuing immune response. In addition, some compounds directly acting at the level of innate and/or cognate effectors can either boost or inhibit tumorinduced immune tolerance. The present review will primarily focus on the current knowledge of the molecular mechanisms underlying the immunogenicity against tumor cells hit by certain classes of anticancer compounds. Anthracyclins, oxaliplatin, γ-irradiation, and ultraviolet C (UVC) induce an immunogenic cell death depending on host dendritic cells (DCs) and IFN-γ-producing T lymphocytes. This DC to T cell cognate interaction requires a tumor to DC cross-talk involving caspase activation, calreticulin (CRT) exposure, and the release of high mobility group box 1 (HMGB1) molecules by transformed cells. CRT and HMGB1 play the role of an eat me and a danger signal, respectively, thereby facilitating engulfment and processing of apoptotic bodies by DCs. Based on the premise that the host immune system contributes to the antitumor efficacy of chemotherapy or radiotherapy, the insight gained from the knowledge of the molecular mechanisms involved in this process will provide novel strategies to best use cell death inducers against cancer. Anticancer Agents Cause Immunogenic Cancer Cell Apoptosis Virtually all anticancer agents including anthracyclines, taxanes, platine derivatives, alkylating agents, and γ-irradiation induce their therapeutic effect by induction of apoptotic cell death. But as previously mentioned, some have additional effects on the immune system that contribute to their therapeutic efficacy. To investigate which anticancer agents mediated immunostimulatory effects, Obeid et al. 6 screened approximately 20 distinct apoptosis inducers. For this purpose, tumor cells pretreated with different cytotoxic compounds were inoculated into the flank of syngeneic mice. One week later, mice were rechallenged with live tumor cells, and the absence of tumor development was interpreted as a surrogate marker of a long-term specific antitumor immune response in as much as these protective effects were not achieved in nu/nu counterparts and tumor-free mice were not protected against an irrelevant tumor inoculum. Whereas the injection of live tumor cells did not induce

3 Locher et al.: Witch Hunt against Tumor Cells 53 any anticancer immune response, pretreatment of tumor cells with doxorubicin, idarubicin, mitoxanthrone (three distinct anthracyclines), oxaliplatin or γ-irradiation induced a protection against tumor development. These data were initially obtained using two transplantable tumor models (CT26 colon cancer, MCA205 sarcoma) syngeneic from BALB/c and C57Bl/6 mice, respectively. 6 Similar conclusions were later drawn using TS/A mammary cancer treated with X rays and grafted GOS osteosarcoma. 7 In contrast, other cell death inducers, such as mitomycin C or etoposide, failed to mediate tumor rejection in similar conditions (yet killing tumor cells with identical efficacy in vitro). In sharp contrast, necrotic cells were compromised in their ability to elicit a protective immune response against live tumor cells. Indeed, BALB/c mice inoculated with a lysate of frozen-thawed CT26 cells failed to prevent tumor development. These findings corroborate similar results by Greten and coworkers who found that injection of necrotic tumor cells did not induce potent immune responses in vivo. 8 However, both classes of anticancer agents, whether non-immunogenic or immunogenic, did induce the morphological hallmarks of apoptosis. Both elicited cellular and nuclear shrinkage with chromatin condensation, outer mitochondrial membrane permeabilization, and phosphatidylserine exposure on the outer leaflet of the plasma membrane. In biochemical terms, both immunogenic and nonimmunogenic cell death led to caspase activation. Preventing cell death by treating mice with Z-VAD.fmk a pan-caspase inhibitor completely abolished the immunogenicity of cell death induced by doxorubicin. Altogether, these data strongly support the fact that apoptosis can induce an immunogenic tumor cell death and hence, contradict the theoretical assumption that apoptosis is an immunologically silent process. Conversely, the discrepancies between immunogenic and non-immunogenic apoptosis may be explained by alternative biochemical pathways associated with cell death, yet to be defined. 9 Uptake of Dying Cancer Cells by DCs Currently it remains unclear whether dying tumor cells, killed by immunogenic or non-immunogenic anticancer agents are recognized by the same population of phagocytic cells in vivo. But it might be predicted that immunogenic anticancer drugs would target dying tumor cells to DCs. Indeed, DCs are the most potent antigen-presenting cells in priming naive T cells and they play a key role in the regulation of immunity by providing a critical link between innate and adaptive immune responses. 10 The use of mice carrying the diphtheria toxin receptor under the control of the DC-specific CD11c promoter allowed the transient elimination of DCs following inoculation of diphtheria toxin. 11 In these conditions, mice lost their capacity to elicit an immune response against dying cells. Furthermore, immature murine bone marrow-derived DCs were capable of engulfing anthracyclin-treated tumor cells in vitro and cross-present tumor-derived antigens to tumor antigen-specific MHC class I and class IIrestricted T cells in vitro. In line with these data, Brusa et al. showed that the treatment with doxorubicin or epirubicin increased the uptake of tumor cells by human peripheral blood mononuclear cells. 12 Currently, the exact mechanism by which distinct chemotherapy or radiotherapy regimen might specifically induce immunogenic cell death remains unknown. Several checkpoints might be involved, such as the recognition and phagocytosis of dying tumor cells, the processing of the phagocytic cargo, the full-blown maturation of DCs, and, finally, the induction of innate helper cells and/or polarized effector and memory T-cell responses.

4 54 Annals of the New York Academy of Sciences Recognition and Phagocytosis of Tumor Cells Role of CRT Using two-dimensional electrophoresis coupled with mass spectroscopy analyses, Obeid et al. identified that the major biochemical difference between non-immunogenic and immunogenic cell death resides in the absence or presence of CRT on the plasma membrane, respectively. Surface exposure of CRT occurs only in immunogenic cancer cell death induced by anthrayclines, oxaliplatinum, γ-irradiation, and UVC. 6,13 CRT is mostly present in the lumen of endoplasmic reticulum (ER) where it acts as a calcium-binding and chaperone protein that facilitates MHC class I peptide. 14 Interestingly, it has already been shown that cellular stress, including apoptosis, may lead to increased amounts of CRT on the cell surface. Once at the plasma membrane, CRT serves as an eat me signal and increases the susceptibility to phagocytosis by macrophages and DCs. 15 Cells deficient in CRT do undergo apoptosis but are inefficiently removed by phagocytes, suggesting that CRT exposure is a major molecular determinant for phagocytosis. 6 The CRT receptor that enhances phagocytosis remains unknown. The surface exposure of CRT in response to anthracyclines or γ-irradiation treatment was confirmed by conventional electrophoreses and immunofluorescence staining. This was demonstrated in vitro in different murine 6 and human tumor cell lines 12 and in vivo incidentally on circulating blasts in acute myeloid leukaemia patients in response to intravenous injection of idarubicin. 16,17 The surface exposure of CRT in response to immunogenic chemotherapy or radiotherapy occurred before the first morphological signs of apoptosis-related modifications and preceded the apoptosis-associated phosphatidylserine exposure on the outer leaflet of the plasma membrane. 6 Thus, CRT was detectable at an early pre-apoptotic stage. Pre-apoptotic CRT exposure required caspase activation in as much as the pan-caspase inhibitor Z-VAD.fmk could suppress CRT exposure in tumor cells treated with anthracyclines. 18 Because CRT exposure occurred within 1 h post therapy with cytotoxic drugs, it was suggested that this process might involve the translocation of pre-existing CRT from inside the cell to the cell surface. Accordingly, CRT surface exposure was not associated with an increase in the abundance of intracellular CRT. Furthermore, inhibition of protein synthesis (either by actinomycin D or by enucleation) failed to prevent CRT exposure. These data confirm that CRT translocates from the lumen of the ER to the cell surface in response to immunogenic anticancer drugs. Using blocking antibodies or small interfering RNA (sirna) targeting CRT, Obeid et al. investigated the possible implication of CRT in the phagocytosis of tumor cells treated by immunogenic anticancer therapies. Inhibition of CRT or knocking down CRT transcripts abolished both the phagocytosis of dying tumor cells by DCs and the immunogenicity of anthracycline-induced cell death. However, phagocytosis and immunogenicity of dying cells could be restored by adsorbing recombinant CRT protein (rcrt) to the cell surface of dying tumor cells. 5,6,13 Most importantly, nonimmunogenic chemotherapeutic agents (such as mitomycin C or etoposide) were rendered immunogenic by adsorbing rcrt on the tumor cell surface. In the same way, enforcing CRT exposure by treatment with inhibitors of the phosphatase PP1 and its cofactor GADD34 (e.g., salubrinal or tautomycin) also enhanced the immunogenic potential of cells succumbing to non-immunogenic anticancer drugs. These results formally underline that CRT exposure is crucial for DC-mediated phagocytosis of dying tumor cells and, thus, for the immunogenicity of cell death. It also suggests that agents triggering CRT exposure could be used in combination with conventional anticancer therapy to initiate a powerful antitumor immune response. 4

5 Locher et al.: Witch Hunt against Tumor Cells 55 CRT alone (i.e., without inducers of apoptosis) is not sufficient to elicit an antitumor immune response. Live cells that express CRT at the surface membrane failed to induce DC maturation and antigen presentation and hence were non-immunogenic. 9,13 This implies that tumor cells must emit additional signals to trigger an efficient immune response. Role of ERp57 The proteomic study using the twodimensional electrophoresis followed by mass spectroscopic analyses also led to the identification of an early exposure of ERp57 to the surface of anthracycline-treated tumor cells. 19 Indeed, doxorubicin and mitoxantrone strongly induced surface exposure of ERp57. ERp57 belongs to the disulfide isomerase family of proteins, which are located in the lumen of the ER where ERp57 tightly binds to CRT. ERp57 functions as a disulfide oxidation catalyzer and as a component of the peptideloading complex of the MHC class I pathway. 20 Interestingly, pre-apoptotic ERp57 exposure is governed by the same rules as CRT exposure. 19 First, the exposure of ERp57 on the cell surface correlates with the capacity of chemotherapy agents to elicit immunogenic cell death. Second, it occurs at pre-apoptotic stage, shortly after treatment. Third, ERp57 exposure is not associated with an increase in the abundance of intracellular ERp57. Finally, ERp57 exposure is abolished by caspase inhibitors. Using confocal microscopy and immunoprecipitation assays, we showed that CRT and ERp57 co-localize and interact at the cell surface. This finding prompted us to hypothesize that CRT and ERp57 might traffic together to the cell surface. Supporting this contention, depletion of ERp57 (by knockdown) ablated surface exposure of CRT and vice versa. However, surface exposure of ERp57 was not required for phagocytosis mediated by DCs. Indeed, although a strong correlation between ERp57 exposure and phagocytosis of tumor cells was observed, blocking antibodies targeting ERp57 had no effect on the phagocytosis of anthracycline-treated tumor cells by DCs in contrast to neutralizing antibodies anti-crt. Furthermore, in contrast to rcrt, rerp57 failed to restore the defect in tumor engulfment induced by the knockdown of CRT or ERp57. Altogether, these data indicate that ERp57 is important for the translocation of CRT but does not mediate any immunogenic properties. Exposure of the CRT/ERp57 Complex A hypothetical scheme of the pathway required for the immunogenic surface exposure of the CRT/ERp57 has been recently proposed by Panaretakis and co-workers. 21 They found that immunogenic anticancer agents, such as mitoxantrone, oxaliplatin, or UVC, induced the phosphorylation of the ER stress kinase PERK, which becomes activated and phosphorylates its substrate eif2α. While the depletion of PERK using specific sirna or knockdown or the use of a nonphosphorylatable eif2α mutant has no effect on the induction of cell death, it abolishes CRT/ERp57 exposure, meaning that both the action of PERK and the phosphorylation of eif2α are required for CRT/ERp57 exposure. 21 Similar results were observed using caspase- 8-deficient mouse embryonic fibroblasts. Mitoxantrone and oxaliplatin induced the proteolytic maturation of caspase-8 at an early pre-apoptotic stage. As for PERK or eif2α, caspase-8 depletion did not affect cell death induction but only abolished CRT/ERp57 exposure induced by mitoxantrone, oxaliplatin, or UVC. Interestingly, caspase-8 / mouse embryonic fibroblasts exhibited a normal PERK-mediated eif2α phosphorylation, whereas knockdown of PERK abolished proteolytic maturation of caspase-8 induced by

6 56 Annals of the New York Academy of Sciences mitoxantrone. Therefore, it seems more likely that PERK operates upstream of caspase One of the substrates of caspase-8 is Bap31, a protein which has been implicated in the lethal response to ER stress. The knockdown of Bap31 also abrogated CRT/ERp57 exposure, suggesting that Bap31 cleavage participates in the cell-surface exposure of CRT/ERp57. Finally, the knockdown of PERK and caspase-8 and the use of a nonphosphorylatable eif2α mutant led to a near-to-complete abolition of CRT exposure and reduced the immunogenicity of cell death, indicating that additional molecules can determine the switch between immunogenic and non-immunogenic cell death. HMGB1 Tumor cells must express or emit additional danger signals in order to elicit an efficient immune response dictated by the capacity of host DCs to properly process the phagocytic cargo. Looking for agonists of pattern recognition receptors released by stressed or injured tissues, we screened most of the toll-like receptor (TLR) 4 ligands and found a key role for the HMGB1 in the immunogenicity of cell death. HMGB1 is a nonhistone chromatin-binding protein with dual function. 22,23 Inside the cell, HMGB1 binds DNA and regulates transcription and other nuclear functions. Outside the cell, HMGB1 serves as a mediator of inflammation, triggering DC maturation and mediating adjuvant effects. 24 HMGB1 is either actively secreted by inflammatory cells (e.g., monocytes and macrophages) or released passively by dying cells. 25 Initially, it was thought that only necrotic cells could induce HMGB1 release and that apoptotic cell death leads to the retention of HMGB1. 26 However, as shown later by Bell et al., late apoptosis is also associated with HMGB1 release. 27 Apetoh et al. also found that dying tumor cells (e.g., CT26, MCA205, TS/A) released HMGB1 18 h following anthracycline treatment or irradiation in vitro (i.e., at a late apoptotic stage). 7 HMGB1 release, like CRT exposure, was inhibited by the pan-caspase inhibitor Z-VAD.fmk, and this interference with HMGB1 release strongly impaired the immunogenicity of cell death. 7 More importantly, blocking HMGB1 release, using either nicotine or anti-hmgb1 antibodies, or sirna targeting HMGB1 prevented MHC class II and MHC class I antigen presentation by DCs to CD4 + and CD8 + T cells, respectively, in vitro and in vivo, abolishing the protective immunity induced by anthracycline-treated cells. 7 In order to induce an immunogenic cell death, HMGB1 should bind to one or more receptors expressed on DCs. To date, at least three distinct surface receptors of HMGB1 have been identified [i.e., receptor for advanced glycationendproducts(rage),tlr2,andtlr4, which are all expressed by DCs]. 28,29 TLRs are known to bind pathogen-associated molecular patterns (PAMPs) and danger-associated molecular patterns (DAMPs). 30 Some DAMPs, including HMGB1, fibronectin, and heat shock proteins (HSPs), are released from dying cells and modulate adaptive immune responses by stimulating TLR2 and/or TL4. 31 Therefore, a potential role for TLRs in the induction of the immunogenicity of cell death was proposed and investigated using a systematic screening of the TLR role in the efficacy of chemotherapy. Among all the Tlr1 /, Tlr2 /, Tlr3 /, Tlr4 /, Tlr5 /, Tlr6 /, Tlr7 /, or Tlr9 / mice tested, only Tlr4 / littermates were compromised in their capacity to respond to chemotherapy in vivo and to mount an antitumor immune reponse against dying tumor cells. 7 Cross-presentation of tumor antigen to specific CD8 + T cell clones by Tlr4 / DCs was deficient in vitro. Apetoh et al. demonstrated that TLR4 from mouse DCs and HMGB1 released from dying tumor cells can directly interact. 7 Thus, once released, HMGB1 binds to TLR4 on DCs and controls the initiation of the immune response through processing and presentation of tumor-derived antigens.

7 Locher et al.: Witch Hunt against Tumor Cells 57 TLR4 engagement initiates signaling through two pathways involving either myeloid differentiation primary response protein-88 (MyD88) and/or Toll/IL-1 domain containing adapter-inducing interferon (TRIF). 32 In contrast to WT DCs and Trif / DC, MyD88 / DCs were defective in their capacity to cross-present antigen from dying tumor cells to T cells in vitro. Accordingly, tumor development was not delayed by chemotherapy or radiotherapy in MyD88 / mice. Thus, the immunogenicity of anthracyclines, oxaliplatine, or X rays was exclusively dependent on the TLR4-MyD88 signaling pathway. 7 Interfering in the binding of HMGB1 to TLR4/MyD88 did not affect the efficiency of DCs in engulfing anthracycline-treated cancer cells or in engaging a maturation program [upregulation of cell-surface expression of MHC class II and co-stimulatory molecules, release of IL-6, IL-12p40, or tumor necrosis factor-α (TNF-α)]. However, interaction between HMGB1 and TLR4 was mandatory for the efficient processing and presentation of tumor antigens from dying tumor cells to T lymphocytes. An early report stated that TLR4 could negatively regulate the lysosomal degradation of engulfed cells by delaying fusion between phagosomes and endosomes/lysosomes. 33 This finding suggested that in the absence of TLR4, the rapid lysosomal degradation of tumor antigen could abolish the capacity of DCs to induce antigen presentation. Colocalization of lysotracker red targeting lysosomes with a green fluorescent dye labeling EG7 cells was observed with shorter kinetics in Tlr4 / compared with wild-type DCs, supporting the contention that fusion between phagosomes and endolysosomes was accelerated in TLR4-deficient DCs. 7 Furthermore, inhibition of the activity of lysosomes, either with chloroquine (a lysosomotropic alkaline) or bafilomycin A1 (a specific inhibitor of the vacuolar ATPase responsible for lysosomal acidification) restored the antigen presentation in TLR4-deficient DCs. The addition of chloroquine to chemotherapy compensated for the lack of response of Tlr4 / mice to the antitumor effects mediated by cytotoxic agents in vivo. 7 Interestingly, a polymorphism in TLR4 leading to a single amino acid substitution (Asp299Gly) in the extracellular domain of TLR4 was found to alter the ability of the host to respond to environmental stress. 34 Apetoh et al. found that this TLR4 polymorphism reduced the binding of HMGB1 to human TLR4 and thus inhibited the HMGB1- dependent DC to T-cell cross-talk. Accordingly, DCs from individuals bearing the TLR4 Asp299Gly mutation failed to cross-present antigens from dying melanoma cells to cytotoxic T lymphocytes (CTLs). TLR4 Asp299Gly carriers about 17% of a cohort of breast cancer patients exhibited a higher risk of relapse after adjuvant treatment with anthracyclines. HSP: Elusive Maturation Signal HSP are a family of chaperones that ensure the correct (re)folding of proteins in stress conditions. Although located in intracellular compartments, at least two members of this family, HSP70 and HSP90, can also be found at the plasma membrane where they can stimulate the immune system at several levels. Dhodapkar and co-workers demonstrated that HSP90 may play the role of an eat me signal and thus could contribute to the immunogenicity of tumor cell death. 39 Indeed, human myeloma cells exposed to the proteasome inhibitor bortezomib (but not to γ-irradiation or steroids) specifically express HSP90 on their surface, which contributes to adhesion of tumor cells to DCs in vitro. Other studies have indicated that HSP70 may help to break tolerance to tumor antigens by chaperoning tumor antigens to DCs for presentation in the MHC antigen complex, a concept that has been implemented in the clinic with promising success. 35,36 Buttiglieri et al. demonstrated

8 58 Annals of the New York Academy of Sciences Figure 1. Parameters dictating the success of an immunogenic chemo/radiotherapy. Prior to administering a chemo/radiotherapy for cancer, one should assess (i) certain tumor criteria that preclude the translocation of calreticulin (CRT) (deficiency in the intrinsic capacity of tumor cells to mount an endoplasmic reticulum stress response leading to eif2a phosphorylation (presence of caspase 8, phosphorylation of PERK, cleavage of BAP31) and (ii) certain genetic host defects that preclude an optimal antitumor immune response [deficiency in toll-like receptor 4 (TLR4)]. Those defects could be compensated by specific interventions (exogenous recombinant CRT, and/or chloroquine, respectively). that HSP70 is translocated to the cell surface of both UVC-treated K1 cells and in doxorubicin-treated KATO cells. 37 However, in his model system, Apetoh et al. found that irradiation of EG7 cells or doxorubicin treatment of CT26 did not provoke the surface exposure of HSP70. 7 Thus, the exact role of loaded versus unloaded chaperones in the immunogenicity of cell death will require further investigations. Strategies for the Induction of Immunogenic Cell Death One of the first anticancer strategies employed by oncologists has been to restore the sensitivity of tumor cells to death. However, this strategy exploiting only the intrinsic barriers against cancer frequently leads to tumor relapse and therapeutic failure. One of the reasons is that long-term immune protection (extrinsic barriers against cancer) against dying tumor cells has not been elicited in most cases, presumably because of cell autonomous, treatmentinduced, or host-derived hindrances. Combining restoration of cell death and maintenance of tumor immunosurveillance might reenforce the host tumor balance. Taking into consideration the requirement of CRT for recognition and engulfment of dying tumor cells by host DCs, it may be useful to (i) detect those tumors presenting with intrinsic defects preventing CRT exposure 21 and (ii) screen for drugs promoting CRT exposure. Indeed, as demonstrated in murine cancer models, nonimmunogenic agents, such as mitomycin C, could be rendered immunogenic either by the use of an external supply of CRT or by enforcing CRT exposure with PP1/GADD34 inhibitors (e.g., salubrinal). Individuals with the Asp299Gly polymorphism of TLR4 exhibit a higher risk of relapse after treatment with anthracylines and local radiotherapy, suggesting that this TLR4 mutation could be used as an independent predictive factor for the success of these therapies. Our preclinical data indicate that combining chloroquine with conventional anticancer treatments may improve the efficacy of anticancer

9 Locher et al.: Witch Hunt against Tumor Cells 59 regimens in TLR4 Asp299Gly carriers. 4 Although aimed and designed to revert multidrug resistance, a clinical study in glioblastoma reported a clinical benefit of combining chloroquine to conventional chemotherapy and radiotherapy. 38 It is clear that the next challenges in the field of oncology will be to delineate the molecular mechanisms of the immunogenicity of cell death to be able to personalize the treatment according to not only the pattern of the tumor itself (according to the targeting drug) but also the status of the host immune system (Fig. 1). Conflicts of Interest The authors declare no conflicts of interest. References 1. Zou, W Regulatory T cells, tumour immunity and immunotherapy. Nat. Rev. Immunol. 6: Rubio, M.T. et al The immunosuppressive effect of vincristine on allostimulatory potential of human dendritic cells interferes with their function and survival. Int. J. Oncol. 25: Zitvogel, L. et al Immunological aspects of cancer chemotherapy. Nat. Rev. Immunol. 8: Zitvogel, L. et al The anticancer immune response: indispensable for therapeutic success? J. Clin. Invest. 118: Apetoh, L. et al Immunogenicity of anthracyclines: moving towards more personalized medicine. Trends Mol. Med. 14: Obeid, M. et al Calreticulin exposure dictates the immunogenicity of cancer cell death. Nat. Med. 13: Apetoh, L. et al Toll-like receptor 4-dependent contribution of the immune system to anticancer chemotherapy and radiotherapy. Nat. Med. 13: Scheffer, S.R. et al Apoptotic, but not necrotic, tumor cell vaccines induce a potent immune response in vivo. Int. J. Cancer 103: Tesniere, A. et al Immunogenic cancer cell death: a key-lock paradigm. Curr. Opin. Immunol. 20: Palucka, K. & J. Banchereau Dendritic cells: a link between innate and adaptive immunity. J. Clin. Immunol. 19: Casares, N. et al Caspase-dependent immunogenicity of doxorubicin-induced tumor cell death. J. Exp. Med. 202: Brusa, D. et al Post-apoptotic tumors are more palatable to dendritic cells and enhance their antigen cross-presentation activity. Vaccine 26: Obeid, M. et al Calreticulin exposure is required for the immunogenicity of gamma-irradiation and UVC light-induced apoptosis. Cell Death Differ. 14: Basu, S. & P.K. Srivastava Calreticulin, a peptide-binding chaperone of the endoplasmic reticulum, elicits tumor- and peptide-specific immunity. J. Exp. Med. 189: Gardai, S.J. et al Cell-surface calreticulin initiates clearance of viable or apoptotic cells through trans-activation of LRP on the phagocyte. Cell 123: Chaput, N. et al Molecular determinants of immunogenic cell death: surface exposure of calreticulin makes the difference. J. Mol. Med. 85: Marczak, A. et al Interaction of doxorubicin and idarubicin with red blood cells from acute myeloid leukaemia patients. Cell Biol. Int. 30: Tesniere, A. et al Molecular characteristics of immunogenic cancer cell death. Cell Death Differ. 15: Panaretakis, T. et al The co-translocation of ERp57 and calreticulin determines the immunogenicity of cell death. Cell Death Differ. 15: Maattanen, P. et al ERp57 and PDI: multifunctional protein disulfide isomerases with similar domain architectures but differing substrate-partner associations. Biochem. Cell Biol. 84: Panaretakis, T. et al Mechanisms of preapoptotic calreticulin exposure in immunogenic cell death. Embo J. 28: Andersson, U. et al HMGB1 as a DNA-binding cytokine. J. Leukoc. Biol. 72: Lotze, M.T. & K.J. Tracey High-mobility group box 1 protein (HMGB1): nuclear weapon in the immune arsenal. Nat. Rev. Immunol. 5: Dumitriu, I.E. et al The secretion of HMGB1 is required for the migration of maturing dendritic cells. J. Leukoc. Biol. 81: Chen, G. et al Extracellular HMGB1 as a proinflammatory cytokine. J. Interferon Cytokine Res. 24: Scaffidi, P., T. Misteli & M.E. Bianchi Release of chromatin protein HMGB1 by necrotic cells triggers inflammation. Nature 418: Bell, C.W. et al The extracellular release of HMGB1 during apoptotic cell death. Am.J.Physiol. Cell Physiol. 291: C

10 60 Annals of the New York Academy of Sciences 28. Park, J.S. et al Involvement of toll-like receptors 2 and 4 in cellular activation by high mobility group box 1 protein. J. Biol. Chem. 279: Yang, H. et al Reversing established sepsis with antagonists of endogenous high-mobility group box 1. Proc. Natl. Acad. Sci. USA 101: Fischer, M. & M. Ehlers Toll-like receptors in autoimmunity. Ann. N. Y. Acad. Sci. 1143: Marshak-Rothstein, A Toll-like receptors in systemic autoimmune disease. Nat. Rev. Immunol. 6: Yamamoto, M. et al Role of adaptor TRIF in the MyD88-independent toll-like receptor signaling pathway. Science 301: Shiratsuchi, A. et al Inhibitory effect of Tolllike receptor 4 on fusion between phagosomes and endosomes/lysosomes in macrophages. J. Immunol. 172: Arbour, N.C. et al TLR4 mutations are associated with endotoxin hyporesponsiveness in humans. Nat. Genet. 25: Castelli, C. et al Heat shock proteins: biological functions and clinical application as personalized vaccines for human cancer. Cancer Immunol. Immunother. 53: Castelli, C. et al Human heat shock protein 70 peptide complexes specifically activate antimelanoma T cells. Cancer Res. 61: Buttiglieri, S. et al Influence of drug-induced apoptotic death on processing and presentation of tumor antigens by dendritic cells. Int. J. Cancer 106: Sotelo, J., E. Briceno & M.A. Lopez-Gonzalez Adding chloroquine to conventional treatment for glioblastoma multiforme: a randomized, doubleblind, placebo-controlled trial. Ann. Intern. Med. 144: Spisek, R. et al Bortezomib enhances dendritic cell (DC)-mediated induction of immunity to human myeloma via exposure of cell surface heat shock protein 90 on dying tumor cells: therapeutic implications. Blood 109: