Arsenic trioxide as an inducer of immunogenic cell death

Arsenic trioxide (ATO) is often combined with all-trans retinoic acid (ATRA) to treat promyelocytic leukemia (PML) with a rather high success rate. In mice, it has been documented that ATRA is much more efficient against PML developing in immunocompetent than in immunodeficient mice, pleading in favor of the idea that the antileukemic action of ATRA depends on the immune system. However, no such immunedependent effects of ATO have been described in PML. Nonetheless, it has been shown that ATO increases lymphokine activated killer (LAK)-mediated cytotoxicity against human myeloma cells and enhances the efficacy of Bacille Calmette-Guérin (BCG) immunotherapy in a mouse model of bladder cancer. Moreover, ATO has been demonstrated to deplete regulatory T cells in a mouse model of colon cancer. Of note, in a recent paper published in Cellular and Molecular Immunology, Chen et al. demonstrate that ATO can trigger immunogenic cell death (ICD) in solid tumors. The concept of ICD, initially established in cells undergoing apoptosis, has recently been extended to other variants of regulated cell death such as necroptosis, pyroptosis, and ferroptosis. Canonical ICD triggers the emission of a set of danger associated molecular patterns (DAMPs), which act on specific pattern recognition receptors (PRRs) expressed by antigen presenting dendritic cells (DCs), thus stimulating phagocytosis of malignant cells and antigen presentation of tumorassociated antigens by DCs. Mature DCs facilitate crosspresentation of tumor antigens to cytotoxic T lymphocytes (CTL) as well as the education of memory T cells, altogether conferring efficacy to cancer therapies that last beyond treatment discontinuation. Preclinical and clinical data support the notion that ICD inducers can be advantageously combined with additional immunotherapies such as immune checkpoint blockade targeting the PD-1/PD-L1 interaction. In their work, Chen et al. discovered that in vitro cultures of malignant cells with ATO led to the generation of a whole cell vaccine that could be injected into mice to reduce cancer growth in prophylactic as well as in therapeutic settings. These anticancer effects of ATO-treated cancer cells were lost or attenuated upon depletion of CD8 (but not NK1.1) T cells, as well as after blocking either interferon-Υ (IFNΥ) or the Type-1 interferon receptor (IFNAR) with suitable antibodies. ATO-treated cells manifested several well-established hallmarks of ICD including the release of ATP and high-mobility group B1 (HMGB1) protein, the exposure of calreticulin (CALR) on the cell surface, the induction of cGAMP production, and the H151-repressible (and hence likely STINGdependent) induction of interferon-β1 (IFNβ1). At the mechanistic level, the authors described that ATO induced biochemical characteristics of several cellular stress and death routines including autophagy, apoptosis, ferroptosis, necroptosis, and pyroptosis that all were blunted when ATOinduced oxidative stress was quenched by N-acetyl-L-cysteine. However, the knockout or knockdown of genes required for apoptosis (Bak, Bax), autophagy (Becn1), ferroptosis (Acsl4), necroptosis (Mlk1, Ripk3) and pyroptosis (Gsdmd, Gsdme) did not prevent ATO-induced cell killing, indicating that none among these pathways is indispensable for the lethal outcome of ATO treatment. In stark contrast, knockout of several among these effectors, in particular Acsl4 (involved in ferroptosis) and Mlk1 or Ripk3 (involved in necroptosis), fully abolished the capacity of ATO-treated TC-1 non-small cell lung cancer to induce prophylactic anti-TC-1 immune responses in mice. The knockout of Bak or Bax (both involved in apoptosis) or Becn1 (involved in autophagy) yielded a partial phenotype (i.e., attenuation but not abolition of vaccination), while the knockout of Gsdmd, Gsdme (involved in pyroptosis) failed to affect the capacity of ATO-treated TC-1 cells to induce a protective anti-TC-1 immune response. Hence, several among the ATO-triggered subroutines (autophagy, apoptosis, ferroptosis, necroptosis) contribute to the vaccination effect. Mechanistically, the authors showed that Acsl4, Bak, Bax, Becn1, Mkl1, and Ripk3 contributed to ATO-induced ATP release; Mkl1 and Ripk3 to HMGB1 release; Acsl4, Bak, Bax, Mlkl, and Ripk3 to CALR exposure; and Acsl4, Mlkl, and Ripk3 to extracellular cGAMP accumulation, IFNβ1 secretion and transcription of IFN-stimulated genes (ISG). In a subsequent step, Chen et al. showed that intratumoral injection of ATO failed to suppress the outgrowth of TC-1 or MCA205 fibrosarcomas in vivo. In contrast, immunization with an ATO-based whole-cell vaccine partially restrained the growth of established TC-1 or MCA205 tumors. Such therapeutic whole-cell vaccines required expression of Acsl4, Mlkl, and Ripk3. Of note, the failure of Ripk3 cells to act as a therapeutic vaccine could be restored when ATO treatment in vitro was combined with drugs that enhance extracellular ATP (the apyrase inhibitor ARL67156), ligate Toll-like receptor-4 (monophosphoryl lipid A) or activate STING (2ʹ3’cGAMP). Of note, the ATO-based whole-cell vaccine increased its capacity to restrain tumor growth if combined with PD-1 blockade. Although these latter effects appeared additive (rather than synergistic), they delineate a possible strategy for improving the efficacy of therapeutic vaccination with cells undergoing ICD. Altogether, the aforementioned data support the idea that ATO, which is likely the first antineoplastic chemotherapy that was used in the world (in particular in China), can stimulate several facets of ICD. Importantly, ATO activates a mixed pattern of stress and death pathways, many of which contribute to the immunogenic effects of ATO-treated cancer cells (Figure 1). Future studies must explore the possibility that such a pleiotropic pattern of ICD might be more efficient in yielding therapeutic cancer vaccines than ICD relying on single ONCOIMMUNOLOGY 2023, VOL. 12, NO. 1, 2174723 https://doi.org/10.1080/2162402X.2023.2174723

The concept of ICD, initially established in cells undergoing apoptosis, has recently been extended to other variants of regulated cell death such as necroptosis, pyroptosis, and ferroptosis. [7][8][9][10][11] Canonical ICD triggers the emission of a set of danger associated molecular patterns (DAMPs), which act on specific pattern recognition receptors (PRRs) expressed by antigen presenting dendritic cells (DCs), thus stimulating phagocytosis of malignant cells and antigen presentation of tumorassociated antigens by DCs. 12,13 Mature DCs facilitate crosspresentation of tumor antigens to cytotoxic T lymphocytes (CTL) as well as the education of memory T cells, altogether conferring efficacy to cancer therapies that last beyond treatment discontinuation. 12 Preclinical and clinical data support the notion that ICD inducers can be advantageously combined with additional immunotherapies such as immune checkpoint blockade targeting the PD-1/PD-L1 interaction.
In their work, Chen et al. discovered that in vitro cultures of malignant cells with ATO led to the generation of a whole cell vaccine that could be injected into mice to reduce cancer growth in prophylactic as well as in therapeutic settings. 6 These anticancer effects of ATO-treated cancer cells were lost or attenuated upon depletion of CD8 + (but not NK1.1 + ) T cells, as well as after blocking either interferon-ϒ (IFN ϒ ) or the Type-1 interferon receptor (IFNAR) with suitable antibodies. ATO-treated cells manifested several well-established hallmarks of ICD including the release of ATP and high-mobility group B1 (HMGB1) protein, the exposure of calreticulin (CALR) on the cell surface, the induction of cGAMP production, and the H151-repressible (and hence likely STINGdependent) induction of interferon-β1 (IFNβ1). 6 At the mechanistic level, the authors described that ATO induced biochemical characteristics of several cellular stress and death routines including autophagy, apoptosis, ferroptosis, necroptosis, and pyroptosis that all were blunted when ATO-induced oxidative stress was quenched by N-acetyl-L-cysteine. However, the knockout or knockdown of genes required for apoptosis (Bak, Bax), autophagy (Becn1), ferroptosis (Acsl4), necroptosis (Mlk1, Ripk3) and pyroptosis (Gsdmd, Gsdme) did not prevent ATO-induced cell killing, indicating that none among these pathways is indispensable for the lethal outcome of ATO treatment. In stark contrast, knockout of several among these effectors, in particular Acsl4 (involved in ferroptosis) and Mlk1 or Ripk3 (involved in necroptosis), fully abolished the capacity of ATO-treated TC-1 non-small cell lung cancer to induce prophylactic anti-TC-1 immune responses in mice. The knockout of Bak or Bax (both involved in apoptosis) or Becn1 (involved in autophagy) yielded a partial phenotype (i.e., attenuation but not abolition of vaccination), while the knockout of Gsdmd, Gsdme (involved in pyroptosis) failed to affect the capacity of ATO-treated TC-1 cells to induce a protective anti-TC-1 immune response. Hence, several among the ATO-triggered subroutines (autophagy, apoptosis, ferroptosis, necroptosis) contribute to the vaccination effect. Mechanistically, the authors showed that Acsl4, Bak, Bax, Becn1, Mkl1, and Ripk3 contributed to ATO-induced ATP release; Mkl1 and Ripk3 to HMGB1 release; Acsl4, Bak, Bax, Mlkl, and Ripk3 to CALR exposure; and Acsl4, Mlkl, and Ripk3 to extracellular cGAMP accumulation, IFNβ1 secretion and transcription of IFN-stimulated genes (ISG). 6 In a subsequent step, Chen et al. showed that intratumoral injection of ATO failed to suppress the outgrowth of TC-1 or MCA205 fibrosarcomas in vivo. 6 In contrast, immunization with an ATO-based whole-cell vaccine partially restrained the growth of established TC-1 or MCA205 tumors. Such therapeutic whole-cell vaccines required expression of Acsl4, Mlkl, and Ripk3. Of note, the failure of Ripk3 −/− cells to act as a therapeutic vaccine could be restored when ATO treatment in vitro was combined with drugs that enhance extracellular ATP (the apyrase inhibitor ARL67156), ligate Toll-like receptor-4 (monophosphoryl lipid A) or activate STING (2ʹ3'-cGAMP). Of note, the ATO-based whole-cell vaccine increased its capacity to restrain tumor growth if combined with PD-1 blockade. 6 Although these latter effects appeared additive (rather than synergistic), they delineate a possible strategy for improving the efficacy of therapeutic vaccination with cells undergoing ICD.
Altogether, the aforementioned data support the idea that ATO, which is likely the first antineoplastic chemotherapy that was used in the world (in particular in China), can stimulate several facets of ICD. Importantly, ATO activates a mixed pattern of stress and death pathways, many of which contribute to the immunogenic effects of ATO-treated cancer cells (Figure 1). Future studies must explore the possibility that such a pleiotropic pattern of ICD might be more efficient in yielding therapeutic cancer vaccines than ICD relying on single or dual-cell stress/death subroutine(s). In other words, would a mixture of four stress/death pathways (autophagy, apoptosis, ferroptosis, and necroptosis) generate a more efficient ICDbased vaccine that only involves one or combinations of two or three pathways? The answer to this question will have implications for the future clinical development of therapeutic vaccines.

Data availability statement
All data that led to the conclusions in this manuscript have been included and all sources have been described.

Disclosure statement
OK is a cofounder of Samsara Therapeutics. GK has been holding research contracts with Daiichi Sankyo, Eleor, Kaleido, Lytix Pharma, PharmaMar, Osasuna Therapeutics, Samsara Therapeutics, Sanofi, Tollys, and Vascage. GK has been consulting for Reithera. GK is on the Board of Directors of the Bristol Myers Squibb Foundation France. GK is a scientific co-founder of EverImmune, Osasuna Therapeutics, Samsara Therapeutics, and Therafast Bio. GK is the inventor of patents covering therapeutic targeting of aging, cancer, cystic fibrosis, and metabolic disorders. GK's wife, Laurence Zitvogel, has held research contracts with 9 Meters Biopharma, Daiichi Sankyo, Pilege, was on the Board of Directors of Transgene, is a cofounder of EverImmune, and holds patents covering the treatment of cancer and the therapeutic manipulation of the microbiota. GK's brother, Romano Kroemer, was an employee of Sanofi and now consults for Boehringer-Ingelheim.