Cu(II) complex that synergistically potentiates cytotoxicity and an antitumor immune response by targeting cellular redox homeostasis

Significance

The immunomodulatory effects of chemotherapy have aroused increasing clinical attention for their role in boosting the response rate and potency of immune checkpoint inhibitors in cancer therapy. Nevertheless, only a limited number of drugs provide an immune-provoking response without engendering dose-limiting toxic side effects. Endoplasmic reticulum stress, mediated by reactive oxygen species (ROS), is thought to induce an antitumor immune response. A feature of cancer cells is their ability to maintain redox homeostasis at a higher level than normal cells Therefore, manipulating the ROS levels by disrupting redox homeostasis could provide an approach to agents that provide an antitumor immune response. Here, we report a metal-based complex (Cu-1) that disrupts cellular redox homeostasis and stimulates an antitumor immune response.

Abstract

Developing anticancer drugs with low side effects is an ongoing challenge. Immunogenic cell death (ICD) has received extensive attention as a potential synergistic modality for cancer immunotherapy. However, only a limited set of drugs or treatment modalities can trigger an ICD response and none of them have cytotoxic selectivity. This provides an incentive to explore strategies that might provide more effective ICD inducers free of adverse side effects. Here, we report a metal-based complex (Cu-1) that disrupts cellular redox homeostasis and effectively stimulates an antitumor immune response with high cytotoxic specificity. Upon entering tumor cells, this Cu(II) complex enhances the production of intracellular radical oxidative species while concurrently depleting glutathione (GSH). As the result of heightening cellular oxidative stress, Cu-1 gives rise to a relatively high cytotoxicity to cancer cells, whereas normal cells with low levels of GSH are relatively unaffected. The present Cu(II) complex initiates a potent ferroptosis-dependent ICD response and effectively inhibits in vivo tumor growth in an animal model (c57BL/6 mice challenged with colorectal cancer). This study presents a strategy to develop metal-based drugs that could synergistically potentiate cytotoxic selectivity and promote apoptosis-independent ICD responses through perturbations in redox homeostasis.

Immunogenic cell death (ICD) is a functionally unique type of cell death that is capable of inducing adaptive immune responses in antitumor therapy (1–4). In recent years, an increasing number of metal complexes, such as those based on platinum, iridium, gold, ruthenium, and copper, have been reported to induce ICD activity (5–23). ICD inducers are divided into type I and type II. Type I inducers trigger ICD indirectly through endoplasmic reticulum (ER) stress, while type II inducers directly stimulate ER stress by targeting ER organelle (24, 25). Although it is known that ER stress is required to initiate ICD, given the mechanistic complexity of ER stress, this stress alone is not able to consistently trigger an immune response (26). Furthermore, the compounds that elicit promising ICD responses are usually only effective in specific tumor types. For example, oxaliplatin can initiate potent ICD in colorectal cancer but fails to do so in other cancers, such as non–small cell lung cancer (27).

Generally speaking, high efficacy and low toxicity are the ultimate goals of drug development. In this context, it is worth noting that none of the ICD inducers reported to date displayed cytotoxic selectivity. Thus, there remains a need to develop strategies that allow for the synergistic regulation of an effective ICD response while allowing for cytotoxic differentiation between cancer cells and normal cells. One effective approach would be to target specifically an “Achilles’ Heel” of cancer cells while sparing normal cells. A key pathophysiological feature of cancer cells that could be exploited in this regard is their ability to maintain redox homeostasis at a significantly higher level than normal cells (28, 29). Cellular redox homeostasis, which involves a dynamic balance between the generation and elimination of reactive oxygen species (ROS), plays a fundamentally important role in maintaining a physiological steady state within living cells (30). There is a certain degree of cellular ROS stress that can be tolerated by cells (31, 32). If the amount of ROS exceeds that threshold, the ROS scavenging systems become overwhelmed, and various cell death pathways are triggered (33–35). Typically, cancer cells produce much more ROS than normal cells. Thus, to maintain redox balance, cancer cells functionally enhance endogenous antioxidant systems in part by generating more glutathione (GSH, ~1,000-fold) than normal cells (33, 36). Given this redox feature, manipulating the ROS levels by a redox modulation strategy that both depletes GSH and promotes intracellular •OH radical production is a logical approach to kill cancer cells selectively without causing significant toxicity to normal cells. To the extent this postulate could be effectively translated into practice, it would be expected to have broad therapeutic benefits in the context of cancer treatment. However, very few molecules are able to simultaneously achieve these paired goals (37, 38).

Copper has two major oxidation states, Cu(II) and Cu(I). This makes copper an attractive so-called redox cycler. This is because the divalent cupric oxidation state copper (Cu²⁺) could be reduced to the monovalent cuprous (Cu⁺) state by GSH. Accompanied by GSH consumption, the resulting Cu⁺ can catalyze the decomposition of H₂O₂ to generate toxic •OH daughter products. It has recently been reported that some copper nanoagents can induce intracellular oxidative stress by simultaneously disrupting the cellular antioxidant defense system and converting less-reactive H₂O₂ into more harmful •OH radical, thus triggering an intensified ER stress response (39, 40). For example, Li and coworkers recently reported that copper-cysteine nanoparticles could contribute to both oxidative •OH production and antioxidant GSH depletion (41). However, replicating these features in a free-standing system has proved elusive. Here, we report a copper complex (Cu-1; Fig. 1) that induces cytotoxicity selectively and a ferroptosis-dependent ICD response in cancer cells by disturbing cellular redox homeostasis (Scheme 1). Ferroptosis is characterized by the accumulation of lipid peroxidation products resulting from excess free iron and decreased cellular GSH. As the result of triggering such effects (vide infra), Cu-1 proved effective in blocking tumor proliferation in a vaccinated and therapized mouse model with lower side effects than oxaliplatin. Complexes Co-1, Pt-1, and Pd-1 with similar structures to Cu-1, but different metal centers, elicited no apparent cytotoxicity. Cu-2, wherein the quinoline moiety in Cu-1 is replaced by a pyridine group, showed lower cytotoxicity and little cytotoxic selectivity (Fig. 1). Our work introduces an effective paradigm wherein a metal-based agent is used to target cellular redox homeostasis and thus trigger a ferroptosis-dependent chemoimmunotherapeutic response.

Fig. 1.

Chemical structures of the metal complexes considered in this study; n = 1 for Cu, Pt, and Pd; n = 2 for Co.

Scheme 1.

Graphic summary of Cu-1-induced antitumor immunity.

Results and Discussion

Synthesis and Characterization.

Considering the unique biological activities of metal compounds with quinoline alkaloids or thiosemicarbazide ligands (42, 43), we synthesized a thiosemicarbazone (TSC) ligand via a typical Schiff base condensation from 4-methylthiosemicarbazide and quinoline-8-carbaldehyde; subsequently, the Cu(TSC)Cl complex was prepared by heating the TSC ligand with CuCl₂ (SI Appendix, Scheme S1). To obtain initial insights into possible structure–activity relationships, we also prepared several similar complexes, M(TSC)Cl (M = Co, Pt, Pd), as well as Cu-2, using an analogous procedure (44). In complex Cu-1, the copper metal center is coordinated by a five-membered ring and a six-membered ring formed from the quinolone. This differs from Cu-2, wherein the metal center is coordinated by two five-membered rings.

The above complexes proved highly soluble (>10 mM) in THF, DMF, and DMSO. They were characterized by means of ¹H NMR and infrared spectroscopies, as well as high-resolution mass spectrometry (HRMS) (SI Appendix, Figs. S1–S16). Complexes Cu-1 and Pd-1 were further characterized by single crystal X-ray diffraction analysis (Fig. 2 and SI Appendix, Fig. S17 and Table S1). Unless otherwise specified, the metal complexes were dissolved in DMF to prepare 2 mM stock solutions. These stock solutions were then diluted into an aqueous solution for subsequent experiments. All complexes proved stable in DMF/H₂O (1/1, v/v) and DMF/PBS (1/1, v/v) as inferred from the lack of any noticeable change in the corresponding LC chromatograms after a 24 h incubation period (SI Appendix, Figs. S18–S21).

Fig. 2.

Single X-ray crystal structure of Cu-1 shown at the 50% probability level. All hydrogen atoms have been removed for clarity. CCDC number: 2289499.

In Vitro Cytotoxicity.

Copper nanoagents have been reported to induce death in cancer cells via two synergetic modes of action (•OH accumulation and GSH depletion) (41). We hypothesized that these features would be recapitulated in Cu-1. Indeed, low IC₅₀ values (SI Appendix, Table S2) were observed in tumoral cell lines, including A549 (human lung cancer), CT26 (mouse colon carcinoma), MGC80-3 (human gastric cancer), SK-OV-3 (human ovarian cancer), HeLa229 (human cervical cancer) after a 48 h treatment, as determined by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assays (Fig. 3A). Markedly decreased cytotoxicity against the normal cell line, 293T, was seen. Under analogous conditions, Pt-1, Pd-1, and Co-1 exerted almost no cytotoxicity against all the tested cancer cell lines (Fig. 3A and SI Appendix, Table S2). The similarly structured Cu-2 also displayed cytotoxicity, but this was lower than that of Cu-1 (SI Appendix, Table S2). These results provide support for the core hypothesis that Cu-1 possesses a mechanism of action that differs from that of the other metal complexes included in the present study.

Fig. 3.

Effects on in vitro cytotoxicity and redox homeostasis. (A) Cytotoxicity of the indicated metal complexes incubated with cells for 48 h. (B) Degradation of methylene blue (MB) seen upon exposure to different concentrations of GSH. (C) Degradation of MB seen at different time points. (D) Degradation of MB seen upon incubation of Cu-1 with H₂O₂ for 1 h at pH = 7.4. Photographs of the MB solutions are shown in the Inset. (E) GSH/GSSG ratio in the indicated cell lines before and after treatment with Cu-1 at the ~IC₅₀ concentration determined for the indicated cell lines. (F) Cellular relative ROS stress level resulting from treatment with Cu-1. *P < 0.05, **P < 0.01, ***P < 0.001.

Perturbation of Redox Homeostasis.

We postulated that the cytotoxicity displayed by Cu-1 is related to its metal-centered redox activity. Therefore, we first tested whether it was capable of generating hydroxyl radicals (•OH). Here, we exploited the scavenging effect of GSH using MB, which is readily degraded by •OH, as a spectrophotometric indicator (41). As shown in Fig. 3C, a significant decrease in the MB absorbance was observed when MB was incubated with H₂O₂, GSH, and Cu-1 for 60 min, while H₂O₂ or Cu-1 alone had no significant impact on the absorbance of MB (Fig. 3D). The degradation of MB was seen to increase proportionally with the GSH concentration from 0 to 1.0 mM; however, it decreased in the presence of excessive GSH, presumably reflecting the scavenging effect of GSH on •OH (Fig. 3B). Even at high concentrations of GSH, ~10 mM (the intracellular concentration of GSH is about 1 to 10 mM), Cu-1 still displayed an MB degradation efficiency of approximately 20%. The effective degradation of MB by Cu-1, H₂O₂, and low levels of GSH is taken as evidence that redox cycling involving Cu-1 produces •OH radicals effectively and that the associated GSH depletion could lead to an imbalance in redox homeostasis in vitro. Under the same conditions, no apparent change in the MB absorption spectrum was detected for complexes Pt-1, Pd-1, or Co-1, while Cu-2 gave rise to an MB degradation efficiency of only ca. 15% in 24 h. Taken together, these results provide support for the suggestion that a redox reaction induced by Cu-1 engenders cytotoxicity via a process that differs mechanistically from that produced by Pt-1, Pd-1, Co-1, or Cu-2.

We next investigated whether GSH depletion and ROS production induced by Cu-1 would work synergistically in cancer cells. To this end, we compared the ratio of GSH/GSSG in CT-26 cells treated with and without Cu-1. The GSH/GSSH ratio decreased from ∼9.30 to ∼4.71 after treatment with Cu-1 at its IC₅₀ concentration over the course of 12 h (Fig. 3E). We observed similar effects in other cancer cell lines before and after treatment with Cu-1; however, this was not the case when the cell lines were treated with 10 μM Pt-1, Pd-1, Co-1, or Cu-2.

We used a ROS detection kit to assess the influence of Cu-1 on the intracellular ROS levels. After treatment with Cu-1 for 12 h, the ROS levels in CT-26 cells increased in a statistically significant manner (Fig. 3F), while only a slight increase was observed in (normal) 293T cells. The ROS levels, as represented by the fluorescence intensity, in the CT-26 cell line were about 15% and 55% before and after Cu-1 treatment, respectively, while in 293T cells, these values were ∼10% and ∼22%, indicating that Cu-1 affected more strongly the cancer cell line. The ROS levels in HeLa and A549 cells were also markedly increased (Fig. 3F). In contrast, no apparent effects were observed in the CT-26 cells upon treatment with 10 μM Co-1, Pt-1, or Pd-1 (data not shown). Altogether, these results suggest that Cu-1 may induce cancer cell death by perturbing redox homeostasis via GSH depletion and •OH production.

Ferroptosis-Dependent ICD.

To assess the cell death pathway corresponding to the cytotoxicity induced by Cu-1, we tested various inhibitors of cell death-associated mechanisms in combination with Cu-1. Ferrostatin-1 (Fer-1, 2 μM), a small-molecule inhibitor of ferroptosis, was found to inhibit Cu-1-induced cell death in a statistically significant manner. In contrast, Ac-DEVD-CHO, an apoptosis inhibitor (2 μM), only partially reduced the cytotoxicity of Cu-1 (Fig. 4A). Moreover, 3-methyl-adenine (2 mM, inhibitor of autophagy), TTM (2 μM, inhibitor for cuproptosis), and necrostatin-1 (2 μM, inhibitor for necroptosis) provided little if any rescue from cell death in the case of the CT-26 cells. Annexin-V staining studies revealed little increase in early apoptotic cells, as would typically be seen for ICD inducers (SI Appendix, Fig. S22). Considered in concert, these results are consistent with ferroptosis playing a dominant role in Cu-1-regulated cell death.

Fig. 4.

Ferroptosis induced by Cu-1. (A) Effects of inhibition of ferroptosis, necroptosis, apoptosis, cuproptosis, and autophagy on CT-26 cell viability as seen in the presence of Cu-1 (1.5 μM,48 h). (B) GSH levels in CT-26 cells treated with the indicated concentrations Cu-1 for 12 h. (C) Lipid peroxidation levels in CT-26 cells treated with the indicated concentrations of Cu-1 for 12 h. (Scale bars: 20 μm.) (D) Effects of Cu-1 on key ferroptosis proteins in CT-26 cells after treatment for the indicated time intervals. (E) Effects of ferroptosis inhibition (using 5.0 μM Fer-1 as an inhibitor) on key ferroptosis proteins in CT-26 cell lines treated with Cu-1 for 48 h. *P < 0.05, **P < 0.01, ***P < 0.001.

Ferroptosis is characterized by the accumulation of lipid peroxidation products resulting from excess free iron and decreased cellular GSH (45). Therefore, we examined the levels of GSH and lipid peroxidation. As expected, GSH was significantly diminished, while lipid peroxidation was increased, under Cu-1 treatment (Fig. 4B), as evidenced by C11-BODIPY fluorescence (Fig. 4C). C11-BODIPY is a lipid peroxidation fluorescent probe that displays different fluorescence emission between its nonoxidized (591 nm) and oxidized forms (510 nm) (45). To test further whether ferroptosis was involved in the cell death upon treatment with Cu-1, we examined the key regulators of ferroptosis, namely GPX4 and SLC7A11, after a 48 h treatment with 2.0 μM Cu-1. It was found that Cu-1 markedly decreased the expression of GPX4 and SLC7A11 (Fig. 4D); both of these effects were blocked by Fer-1 (Fig. 4E). Taken together, these results provide support for the notion that Cu-1 induces ferroptosis-dependent cell death due to an imbalance in redox homeostasis in cancer cells.

In vitro, ICD is characterized by three main biochemical events, which have been termed damage-associated molecular patterns (DAMPs): i) Expression of calreticulin (CRT), an ER-resident chaperone protein, to the outer cell membrane, ii) the secretion of adenosine triphosphate (ATP) (“find me” signal), which requires active autophagy, and iii) plasma membrane permeabilization and subsequent secretion of high mobility group box 1 protein (HMGB1, “danger” signal) (8). ER stress is generally required to initiate ICD. To assess whether cancer cells with impaired redox homeostasis induced by Cu-1 might display ICD activity, we first used immunofluorescence staining to detect CRT on the cell surface (ectoCRT), since this event is a major, early ICD-correlated DAMP signal. We observed that treatment with Cu-1 (1.0 or 2.0 μM for 3 h) led to ectoCRT expression in CT-26 cells (Fig. 5A). Using flow cytometry, we detected a dose-dependent induction in ectoCRT after Cu-1 treatment (Fig. 5B and SI Appendix, Fig. S23). Additionally, we confirmed the presence of the other two canonical indicators of ICD in vitro, namely increased ATP release and HMGB1 secretion, after treatment with Cu-1 (Fig. 5 C and D). Complex Cu-1-triggered induction of these three markers of ICD was blocked by Fer-1 (Fig. 5 B–D). A Cu-1 dose-dependent enhancement of Ser51 phosphorylation of eIF2α (p-eIF2α) and CHOP protein was also seen (SI Appendix, Fig. S24). Ser51 phosphorylation is considered a general indicator of translation inhibition and has been identified as a feature of ICD. Taken together, these data provide support for the suggestion that Cu-1 would act as ferroptosis-dependent ICD inducers in vivo.

Fig. 5.

Induction of ICD by Cu-1. (A) Confocal microscopy images of CT-26 cells displaying CRT localization 3 h posttreatment with 1.0 or 2.0 μM Cu-1. (Scale bars: 20 μm.) (B) Percentage of cells expressing surface-CRT posttreatment with Cu-1 or oxaliplatin, as determined by flow cytometry and the effect of the ferroptosis inhibitor, Fer-1 (2.0 μM), on surface expression of CRT. (C) ATP release in CT-26 cells 6 h posttreatment with the indicated concentrations of Cu-1 or oxaliplatin, and the effect of Fer-1 on Cu-1-stimulated ATP release. (D) Release of HMGB1 protein in CT-26 cells 6 h posttreatment with the indicated concentrations of Cu-1 or oxaliplatin, and the effect of Fer-1 on Cu-1-stimulated HMGB1 release. (E) Antitumor vaccination where CT-26 cells were initially treated with Cu-1 (2, 4, 10 μM), oxaliplatin (100 μM), or solvent control for 6 h before the cells were subcutaneously injected into the left flanks of C57BL/6 mice (n = 10), which were then rechallenged in the right flanks with untreated CT-26 cells after 7 d. The curves show the percentage tumor-free after the 30-d treatment. (F–H) Percentages of (F) activated CD8⁺ T cells, (G) CD4⁺T cells, and (H) Foxp3⁺ T cells in the total cells isolated from tumors of C57BL/6 mice treated with PBS, Cu-1, or oxaliplatin. *P < 0.05, **P < 0.01, ***P < 0.001.

In Vivo Antitumor Vaccination.

To test whether ICD would be induced in vivo, a vaccination rechallenge experiment was carried out in mice. Such studies are considered the gold standard assay for ICD. As a positive control, we treated the cell line with oxaliplatin, which has been reported to stimulate ICD in several colon cancer mouse models (27). All animal studies were carried out under approved protocols (Approval No: 201903-012); see SI Appendix. In a first study, CT-26 cells were pretreated with varying concentrations of Cu-1 (2, 4, or 10 μM) or oxaliplatin (100 μM) for 6 h. The drug-treated cells were then injected into the left flank of C57BL/6 mice (n = 10). After 7 d, the right flank of each mouse was injected with the same cell line. Tumor formation on the right side was then assessed daily. Mice without tumors on their right flanks were categorized as tumor-free. In the solvent control group, the percentage of tumor-free mice (displaying no tumors in their right flanks) had decreased to zero by day 11 after rechallenge, whereas 80% of the mice remained tumor-free in the Cu-1 groups (Fig. 5E). Additionally, tumor growth was delayed in the two low-dose (2 and 4 μM) Cu-1-treated groups, compared to the oxaliplatin-treated group. The average tumor volume in the 4 μM Cu-1 group on the 30th day after inoculation was 156 mm³, which is 9.1-fold lower than that of the control group (average = 1,425 mm³). These results are taken as evidence that inoculation with Cu-1-treated cells is effective in enhancing resistance to CT-26 cell tumorigenesis and progression in mice. Therefore, we conclude that Cu-1 shows promise as an ICD inducer in vivo.

To probe the mechanism of the observed Cu-1-induced antitumor immune response in vivo, we analyzed tumor-associated immune cells from the mice treated with the vaccine (repeated 4 μM Cu-1 group). Infiltration of cytotoxic (CD8⁺) T cells in tumors is thought to indicate a specific immune response and is known to correlate with a good prognosis in many cancers. In fact, ICD inducers often elicit T cell-mediated immune responses, including CD8⁺ T cell activation. We evaluated the proportion of CD8⁺ T cells (CD3⁺ CD8⁺ cells) and activated CD8⁺ T cells (CD3⁺ CD8⁺ CD38⁺ T lymphocyte cells) in tumors excised from the mice on day 24. As shown in Fig. 5F, 83% of the tumor T cells in the 4 μM Cu-1 group were activated CD8⁺ T cells, which was increased in a statistically significant manner compared to the controls (51%) and the oxaliplatin group (67%). In addition, we analyzed CD4⁺ T lymphocytes (CD3⁺ CD4⁺ cells), which assist CD8⁺ T cells in generating a cellular immune response. As expected, the proportion of CD4⁺ T cells in the tumor tissue from the 4 μM Cu-1 group (5%) was greater than that in the control (2.51%) and oxaliplatin groups (4%, Fig. 5G). Since Foxp3⁺ T cells interfere with antitumor immunity by secreting immunosuppressive cytokines and blocking cytotoxic cell function in cancer, we also quantified the Foxp3⁺ T lymphocyte (CD3⁺ CD4⁺ Foxp3⁺ cell) levels in resected tumors from the three study groups. As shown in Fig. 5H, the percentage of Foxp3⁺ T cells in the 4 μM Cu-1 group shrank to 6%, which was lower than both the control (9%) and the oxaliplatin groups (10%). Thus, the flow cytometry results revealed a concomitant increase in activated CD8⁺ T cells and a decrease in Foxp3⁺ T cells in the tumors of the vaccinated mice.

In Vivo Anticancer Activity.

Finally, we evaluated the therapeutic potential of Cu-1 in vivo. It was administered intraperitoneally to C57BL/6 mice bearing CT-26 xenograft tumors when the tumor reached approximately 60 mm³ (day 0). Complex Cu-1 (10 mg/kg) was then administered intraperitoneally on days 0, 3, 6, 9, 12, 15, and 18. A control set of mice were treated in an analogous way with oxaliplatin (7 mg/kg). The selection of these doses and administration frequencies was based on the dosing typically used for oxaliplatin and preliminary acute toxicity studies of Cu-1. Compared to the PBS group, oxaliplatin treatment slowed the tumor growth (Fig. 6 B and D), but significantly decreased the body weight and damaged heart and liver function in tumor-bearing mice (Fig. 6 C, E, and F). In sharp contrast, Cu-1 demonstrated a greater in vivo antitumor effect compared to oxaliplatin (Fig. 6 B and D) and did not induce systemic toxicity or body weight loss (Fig. 6 C, E, and F).

Fig. 6.

Tests of the therapeutic potential of Cu-1 (10 mg/kg) and oxaliplatin (7 mg/kg) in vivo. (A) Image of isolated tumors. (B) Relative tumor volume changes during therapy (n = 5). (C) Body weights of mice receiving treatment (n = 5 per group). (D) Weights of isolated tumors. (E−G) Serum biochemical analyses showing the function of the liver (E), heart (F), and kidney (G). (H) Hematoxylin−eosin staining of major organs from mice receiving treatments. (I) Schematic schedule (x axis) for an in vivo Cu content biodistribution assay (n = 3 per group). The untreated mice were set as day 0 with the Cu content in major organs of mice receiving Cu-1 being shown on the y axis. Data represented as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001.

The in vivo biodistribution of Cu-1 was also examined (Fig. 6I). After intravenous injection, Cu-1 was mainly accumulated in the liver and lungs at day 1. The Cu content in the major organs (such as the liver and the lungs) gradually decreased and reached levels similar to those in untreated mice (day 0) after 14 d (P > 0.05), leading us to suggest that Cu-1 is either effectively metabolized in, or cleared from mice. Taken together, these results lend credence to the proposition that Cu-1 possesses an improved in vivo therapeutic profile and produces lower side effects, compared to oxaliplatin.

Conclusion

In summary, we report a copper complex, Cu-1, that acts as a chemoimmunotherapeutic agent displaying high cytotoxicity and selectivity for cancer cells over normal cells. This complex interferes with cancer cellular redox homeostasis by simultaneously depleting intracellular GSH and stimulating the generation of •OH. Complex Cu-1 induces ferroptosis-dependent ICD. These putative benefits were manifest in terms of an ability to block effectively tumor proliferation in a vaccinated mouse model. Importantly, in a colorectal tumor model, Cu-1 suppressed tumor growth and gave rise to reduced off-target toxicity than an accepted front-line chemotherapy (oxaliplatin). Thus, this study highlights an appealing strategy for developing ICD inducers with low side effects by modulating specifically cellular redox homeostasis.

Materials and Methods

Complete experimental protocols including synthesis and characterization, cell culture, western blot analysis, flow cytometry, confocal fluorescence microscopy, in vivo studies, as well as the details of experimental materials and instruments are described in SI Appendix.

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

All study data are included in the article and/or SI Appendix.