3.1 Interactions between tumors and Cu

Malignant tumors enhance the accumulation of intracellular Cu by modifying Cu-related proteins such as CTR1 and ATP7B in response to various stresses like chemotherapy [34] and hypoxia [35]. These elevated intracellular Cu levels subsequently foster malignant tumors' environmental acclimatization and growth. The role of Cu in tumor development has been a long-standing concern, with numerous studies showing that malignancies require higher levels of Cu than normal tissues [36]. Elevated Cu concentration has been observed in the tissues or serum of patients with various types of tumors, including liver, breast, colon, prostate, and thyroid tumors [37-40]. It is widely recognized that Cu is critical in promoting tumorigenesis through various mechanisms. Cu-dependent cell growth and proliferation were referred to as ‘cuproplasia' [5]. Here, we outline how tumor cells upregulate intracellular Cu ion levels to promote tumor survival, growth, and metastasis (Figure 2).

3.2 Cuproptosis

Previous research has demonstrated that an excess of metal ions can result in cell death through specific pathways. In 2012, Dixon SJ et al. first identified a type of iron-dependent programmed cell death called ferroptosis, characterized by the excessive build-up of lipid peroxides and reactive oxygen species (ROS) mediated by iron-dependent fenton reaction [89]. Similarly, the relationship between Cu and cell death has been a focus of investigation for several decades [90]. However, the precise mechanism by which Cu leads to cell death remains poorly understood. Historically, researchers have associated Cu-induced cell death with apoptosis [91], autophagy [92], or ferroptosis [93]. Remarkably, it was not until March 2022 that Tsvetkov et al. elucidated the precise mechanism of Cu-induced cell death, coining the term cuproptosis to describe this unique form of cell death, which was a milestone in the research field of Cu-mediated cell death [94].

Cuproptosis relies on excessive Cu delivered to mitochondria through Cu ionophores like elesclomol, which cannot be rescued by inhibitors of apoptosis, necrosis, or ferroptosis, except for Cu chelators (tetrathiomolybdate, etc.) [94]. In addition, this form of cell death primarily affects mitochondria, leading to oxidative damage to the membrane and reduced enzyme function in the TCA cycle [95]. Cu ionophores, as inducers of cuproptosis, transport Cu(II) into the mitochondria, where it is converted into the more cytotoxic Cu(I) by ferredoxin 1 (adrenodoxin; FDX1), a member of the ferredoxin family containing the Fe-S cluster in mitochondria. FDX1 plays a crucial role in the cellular stress caused by mitochondrial proteins during cuproptosis. Aside from its role in Cu(I) conversion during cuproptosis, FDX1 also facilitates the lipoylation of mitochondrial proteins by directly interacting with lipoic acid synthetase (LIAS) and enhancing its binding to the lipoyl carrier protein, glycine cleavage system H protein (GCSH) [96, 97]. Lipoylation is a highly conserved post-translational modification (PTM) of proteins that occurs in various organisms, from bacteria to humans, affecting specific lysine residues in four enzymes that are part of metabolic complexes regulating carbon entry into the TCA cycle [98]. These four enzymes include Dihydrolipoamide S-Acetyltransferase (DLAT), Dihydrolipoamide S-Succinyltransferase (DLST), Dihydrolipoamide Branched Chain Transacylase E2 (DBT) and Glycine Cleavage System Protein H (GCSH). Among them, DLAT and DLST are significant components of PDH and KDH complexes, respectively, where lipoylation is crucial for their functions [98]. However, overloaded Cu(I) in mitochondria can directly bind to lipoylated proteins, facilitating the aggregation of lipoylated proteins to promote cuproptosis [94]. Notably, DLAT aggregation in such a Cu-dependent manner is one of the most crucial markers for detecting cuproptosis at present. Increased Cu(I) levels in mitochondria not only promote lipoylated proteins to aggregate but also weaken the Fe-S cluster proteins, including FDX1, LIAS, aconitase-2 (ACO-2), succinate dehydrogenase complex iron sulfur subunit B (SDHB), and DNA polymerase delta 1 (POLD1) [94]. Taken together, overloaded Cu(II) is delivered into mitochondria via Cu ionophores, which is further converted to Cu(I) by FDX1, promoting lipoylated protein aggregation and Fe-S protein turnover to trigger proteotoxic stress, eventually resulting in cell death termed cuproptosis (Figure 3). In normoxic conditions, blockage of the electron transfer chain (ETC) can mitigate cuproptosis. This implies that elevated activity in the TCA cycle and protein lipoylation can increase vulnerability to cuproptosis by promoting the aggregation of lipoylated mitochondrial proteins and decreasing Fe-S cluster content [94]. Hence, shifting the metabolism of tumor cells from aerobic glycolysis to the TCA cycle could increase their sensitivity to cuproptosis.

Other than FDX1, using CRISPR-Cas9 knockout screens, Tsvetkov et al. also defined genes involved in the lipoic acid pathway (LIAS and LIPT1) and genes encoding the pyruvate dehydrogenase complex components (DLAT, DLD, PDHA1, and PDHB) are necessary for cuproptosis [94]. In addition to these genes positively linked to cuproptosis, sensitivity to cuproptosis has been negatively linked to cyclin-dependent kinase inhibitor 2A (CDKN2A), glutaminase (GLS), and metal regulatory transcription factor 1 (MTF1) [94].

Following the initial discovery of cuproptosis, recent studies have investigated its regulatory mechanisms in tumor cells, primarily focusing on the protein lipoylation pathway and Cu efflux. For example, the lactylation of METTL16 at K226, an atypical methyltransferase, enhanced the stability of FDX1 mRNA by increasing methylation at site 602A in the coding sequence (CDS) region of FDX1. This modification significantly elevated mitochondrial protein lipoylation levels and cuproptosis in vitro and in vivo [99]. In addition, the ferroptosis inducers sorafenib and erastin can promote cuproptosis in primary liver cancers by enhancing the aggregation of Cu-dependent lipoylated proteins in vitro and in vivo. Mechanistically, both sorafenib and erastin upregulate protein lipoylation by inhibiting the degradation of FDX1, a process mediated by the mitochondrial matrix-associated protease AFG3L2. Additionally, these compounds reduce intracellular Cu chelation by inhibiting the synthesis of GSH through the blockade of cystine import [100]. Moreover, the deficiency of AT-rich interactive domain-containing protein 1A (ARID1A) suppressed glycolysis by downregulating PKM transcription, which in turn enhanced the TCA cycle in liver tumors, ultimately sensitizing cuproptosis in a FDX1-dependent manner [101]. Furthermore, a recent study demonstrated that the β-catenin/TCF4 transcriptional complex directly binds to the ATP7B promoter, leading to its increased expression. The upregulation of ATP7B facilitates the efflux of Cu ions, thereby reducing intracellular Cu levels and inhibiting cuproptosis [87].

To summarize, cuproptosis is thought to interact with components of the TCA cycle in mitochondria rather than the ETC or ATP generation, and it requires a highly conserved protein post-translational modification known as lipoylation. However, there has been less emphasis on understanding the mechanism of reducing the Fe-S cluster. We speculate that exploring the role of Fe-S cluster proteins in cuproptosis will be a valuable contribution to this research area. Furthermore, we contend that utilizing Cu-induced cell death as a therapeutic strategy for tumor eradication could be a potent approach.

4.1 Anti-tumor strategy of Cu depletion

Due to the importance of elevated Cu levels in tumor development, an increasing number of studies have focused on Cu chelators as potential anti-tumor agents. These chelators work by reducing intracellular Cu levels in tumor cells (Figure 4A). Taking advantage of Cu chelators, such as penicillamine, trientine, or tetrathiomolybdate (TTM), can deplete bioavailable Cu, leading to a reduction in tumor vessel diameter, inhibition of endothelial cell proliferation (as indicated by decreased Ki67 expression) and a decrease in ICAM/CD54 expression [102]. Therefore, depleting Cu levels shows promise as a therapeutic strategy for treating tumors. In addition to in vitro studies, the antitumor activity of Cu-chelating agents such as penicillamine, trientine, and TTM has also been explored in animal models and clinical trials, demonstrating promising antitumor potential (Table 1).

TTM, a Cu chelator, hinders Cu-dependent proliferation in tumors by reducing intracellular Cu levels without causing significant harm to normal tissues [110]. TTM inhibits the Cu-dependent mitochondrial complex IV activity, resulting in reduced mitochondrial respiration and increased oxygen availability. This increase in oxygen levels is necessary for activating HIF-prolyl hydroxylase (PHD), facilitating the interaction between PHD and the oxygen-dependent degradation domains (ODDs) of HIF-1α. This interaction leads to the hydroxylation of HIF-1α, which triggers the degradation of HIF-1α through the E3 ubiquitin ligase von Hippel-Lindau (VHL)-dependent ubiquitination-proteasome pathway. This degradation process ultimately hinders the transactivation of HIF-1α target genes such as PDK1, GLUT1, and VEGF, thereby inhibiting angiogenesis [111]. Apart from working as a Cu chelator, an in vitro investigation revealed that TTM limits cell proliferation by suppressing ATP7A expression upregulated by mutant KRAS^G12D in CRC, with the exact mechanism needing further elucidation [112]. Animal models have further confirmed the anti-angiogenic and anti-tumor effects of TTM, leading to its exploration as a potential anti-tumor drug in clinical trials. In both breast tumor xenograft in nude mice and Her2/neu tumor-prone transgenic mice, TTM effectively suppresses tumor growth and angiogenesis by depleting Cu, thus preventing the NF-κB-mediated transcription of proangiogenic factors like VEGF, IL-1α, and IL-8 [69, 113]. In a RIP1-Tag2 transgenic mouse model of pancreatic neuroendocrine tumors, Seiko Ishida et al. demonstrated that TTM-induced reduction in systemic Cu levels led to decreased proliferation and development of pancreatic tumors in vivo [114]. In TTM-treated Her2/neu and FVB mice, the median time for tumor development was delayed nearly one-fold compared to the control group. By observing mammary gland composition and architecture after TTM treatment, this study illustrated that TTM induced breast hypoplastic remodeling by weakening the epithelial duct branch system and reducing breast microvascular density, ultimately impeding breast tumor growth [115]. Combining the above findings, we conclude that TTM more or less achieves its antitumor effect on reducing Cu coenzyme activity. A phase II clinical trial elucidated that TTM delayed tumor progression in patients with stage I and II malignant mesothelioma due to its antiangiogenic effect while showing no noticeable impact in stage III patients [103]. Another phase II trial found that while TTM was well tolerated in advanced kidney tumor treatment and maintained Cu depletion, certain pro-angiogenic factors like IL-6, IL-8, and VEGF associated with TTM may be linked to Cu consumption rather than disease stability. These clinical studies suggest a potential for combining TTM with other antitumor therapies [116].

TEPA, a chelating agent for Cu ions, has been found to reduce programmed cell death ligand 1 (PD-L1) mRNA levels and enhance ubiquitin-mediated degradation by blocking the phosphorylation of STAT3 and EGFR in an experimental tumor model. This mechanism results in elevated levels of CD8⁺ T cells and natural killer cells, ultimately inhibiting tumor growth and enhancing survival rates in mice [104]. Due to possible Cu binding to EGFR and activating EGFR signaling in a ligand-independent manner, we speculate that TEPA might function to block EGFR activation by reducing Cu levels.

D-penicillamine (DPEN), a well-established Cu chelator for treating Wilson's disease [117], is capable of counteracting tumor development by depleting Cu. DPEN is an amino acid derivative characterized by the presence of a sulfhydryl group (-SH), which exhibits strong metal ion chelating properties. The sulfhydryl group of DPEN forms a coordination bond with Cu(II), forming a complex. During this interaction, DPEN facilitates the reduction of Cu(II) to Cu(I). This reduction is crucial for Cu excretion, as Cu(I) is less soluble and has a greater propensity to form insoluble chelates [118]. In addition, DPEN can enhance oxidative stress by catalyzing H₂O₂ production through Cu chelation, selectively targeting tumor cells, and improving responses to radiation and carboplatin treatment [119]. In a rabbit brain tumor model with large vascularized VX2 carcinomas, utilizing a Cu-depleted diet and penicillamine treatment led to decreased endothelial cell turnover, reduced microvascular density in brain tumors, and suppression of tumor volume. However, the efficacy of this treatment was influenced by regional variations in vascular endothelial responsiveness, Cu distribution, and Cu enzyme activity [105].

Trientine (TETA) is an orally active Cu(II)-selective Cu chelator that is considered to be better tolerated than penicillamine in treating Wilson's disease [120, 121]. Research by Yoshii J et al. suggests that trientine exhibits more substantial anti-tumor effects as opposed to penicillamine, particularly in suppressing tumor growth by inhibiting angiogenesis and promoting apoptosis in a mouse hepatocellular carcinoma xenograft model [106]. Through molecular analysis techniques such as RT-PCR and ELISA, it was demonstrated that trientine could reduce the level of angiogenic factor IL-8, leading to tumor regression [122]. Furthermore, trientine shows promise in overcoming chemotherapy resistance, as indicated in a pilot trial involving carboplatin and trientine for platinum-resistant high-grade epithelial ovarian tumors [123]. While trientine has demonstrated beneficial effects in various contexts, further investigation is needed to understand its potential off-target effects fully. For instance, a study revealed that trientine can inhibit hepatosteatosis and induce autophagy by modulating SAT1 activity without necessarily affecting Cu levels in tissues and organs [124].

In conclusion, targeting Cu-related tumor-specific vulnerabilities with small molecule Cu chelators is anticipated to enhance the effectiveness of existing tumor treatments. Notably, we also need to be clear that this field is still evolving, and specific challenges must be addressed. Existing Cu-related medications lack selectivity and may disrupt the body's Cu balance when used long-term, resulting in potential adverse effects for patients undergoing treatment.

4.2 Anti-tumor strategy of Cu overload

The elucidation of the cuproptosis mechanism has paved the way for future pharmaceutical advancements. Cu ionophores, also known as cuproptosis-related drugs, manufacture cuproptosis by increasing intracellular Cu levels through Cu translocation, offering a potential therapeutic avenue for tumor treatment (Figure 4B). Medications such as disulfiram (DSF) and elesclomol (ES) have garnered significant interest due to their ability to elevate cellular Cu levels, and they are currently undergoing clinical trials (Table 1).

DSF, a Food and Drug Administration (FDA)-approved drug with a long history of treating alcoholism, boasts favorable pharmacokinetic properties and a strong safety profile. Drug repurposing is a promising strategy to address the challenges in tumor therapy. Over the past decade, emerging evidence indicates that DSF grants great antitumor potential in various preclinical models, such as prostate tumors, breast tumors, and lymphoid malignant tumors [125-127]. The key to the anti-tumor effect of DSF lies in its ability to target the tumor stem cell marker aldehyde dehydrogenase (ALDH) to inhibit tumor stem cells and suppress proteasome activity in tumor cells through the formation of complexes with metal ions [128]. In addition, it is important to highlight that extracellular Cu is essential for the cytotoxicity of DSF, with the addition of Cu(II) significantly enhancing DSF-induced cell death. Mechanistically, Cu(II) undergoes a complex extracellular redox reaction with DSF to produce the potent Cu(deDTC)2 complex (also known as CuET), a reaction specific to Cu(II) and not seen with other metals such as Fe(II or III), Mn(III), and Zn(II). This Cu(deDTC)2 complex exhibits increased efficacy against melanoma and is proposed to be the primary agent responsible for DSF-induced toxicity [129]. In current studies, DSF, acting as a Cu transporter in conjunction with Cu ions (DSF-Cu), has been shown to induce various forms of cell death, such as apoptosis, ferroptosis, and cuproptosis [130, 131]. Furthermore, DSF has been identified as a DNMT inhibitor, suppressing the expression and activity of DNA methyltransferase 1 (DNMT1) and reducing methylation of the interferon regulatory factor 7 (IRF7) promoter region. This resulted in an increased expression of IRF7, leading to more significant binding of IRF7 to the PD-L1 promoter and an elevation in PD-L1 mRNA levels. Ultimately, this enhanced the immune response to anti-PD-1 antibodies in a breast tumor mouse model derived from a 4T1 cell line [132]. Recent studies have demonstrated that DSF functions as a Cu ionophore, facilitating the entry of Cu into cells. This process leads to the aggregation of lipoylated mitochondrial proteins and a reduction in the levels of Fe-S cluster proteins. Consequently, CuET induces cuproptosis by disrupting the mitochondrial TCA cycle in tumor cells [94]. Clinical studies have also demonstrated the anti-tumor and chemo-sensitizing effects of DSF-Cu on glioblastoma (GBM), likely attributed to its high blood-brain barrier permeability [133]. A phase I clinical trial revealed that the combination of DSF and temozolomide is not only safe but also enhances progression-free survival in GBM patients [107]. Furthermore, a nationwide survey indicated that tumor patients who continued taking DSF post-diagnosis had a lower risk of tumor-related mortality [134]. With well-established pharmacokinetics and documented safety, DSF presents itself as a promising ‘old’ drug with significant potential for expedited development as a new tumor treatment. Though the selectivity and the underlying mechanism of how DSF delivers or increases the mitochondrial Cu levels remain undefined, we think the anti-tumor effect of DSF relies on the induction of cuproptosis, at least partially.

Elesclomol (also known as STA-4783) is a well-researched cuproptosis inducer that has shown promise in enhancing the antitumor effects of taxanes. This compound, a bis (thiohydrazide) amide, forms a 1:1 complex with Cu(II) and has various mechanisms for killing tumor cells [135]. Elesclomol specifically binds to extracellular Cu(II) and enters cells as elesclomol-Cu(II), which then transports Cu(II) to mitochondria. By repeatedly redistributing Cu from the cell exterior to mitochondria upon dissociation from the complex, elesclomol leads to Cu accumulation in mitochondria. Notably, compared to other Cu transporters like DSF, elesclomol demonstrates superior selectivity for mitochondria [136]. Over the past decade, elesclomol-Cu(II) therapy has shown significant advancements. Previous research suggested that treatment with elesclomol-Cu(II) induces Cu-related apoptosis by increasing ROS levels beyond the cell's tolerance threshold in tumors [136, 137]. Since 2015, new mechanisms of tumor cell death induced by elesclomol-Cu(II) have been reported, including G1 cell cycle arrest, DNA double-strand breaks, and the degradation of ATP7A in CRC [138, 139]. In 2019, the identification of FDX1 through genome-wide CRISPR-Cas9 screening revealed it as a key gene linked to the effectiveness of elesclomol-Cu(II) therapy under proteolysis stress. Further analyses showed that FDX1 is a direct target of elesclomol-Cu(II), leading to a distinct form of Cu-dependent cell death [140]. Herein, FDX1, the gene most linked to elesclomol sensitivity, encodes a protein that directly binds elesclomol-Cu(II) through its α2/α3 helices and β5 strand, thereby inhibiting the formation of Fe-S clusters [140]. Additionally, research by Tsvetkov et al. highlighted that tumors with Hi-Mito disorder, favoring oxidative phosphorylation over glycolysis, exhibit resistance to proteasome inhibitors but are more susceptible to cuproptosis induced by elesclomol-Cu(II) [94]. Furthermore, studies have demonstrated that ferroptosis inducers sorafenib and erastin upregulate protein lipoylation by inhibiting the degradation of the FDX1 protein, a process mediated by the mitochondrial matrix-associated protease AFG3L2 [100]. Additionally, these compounds reduce intracellular Cu chelation by inhibiting cystine import, which leads to decreased synthesis of GSH and consequently promotes elesclomol-Cu(II)-induced cuproptosis in liver tumor [100]. In addition to its role in treating liver tumor, another study reported that the SIRT2-specific inhibitor AGK2 activates METTL16 by enhancing the lactylation of its K229 site. This activation results in an increase in m6A accumulation of FDX1 mRNA and its stability, thereby promoting elesclomol-Cu(II)-mediated cuproptosis in gastric tumor in vitro and in vivo [99]. Although elesclomol-Cu(II) has been studied for its potent tumoricidal effects, research has shown that intracellular Cu binds to PDK1, enhancing its interaction with AKT. This interaction activates the Wnt/β-catenin pathway and promotes ATP7B transcription, thereby demonstrating a capacity to resist cuproptosis. Consequently, combining the Wnt inhibitor LF3 with elesclomol-Cu(II) also shows significant potential [87]. Several clinical trials have evaluated the combination of paclitaxel and elesclomol, particularly in advanced melanoma cases. A phase I clinical trial demonstrated that the toxicity profile of elesclomol with paclitaxel was well-tolerated and comparable to that of paclitaxel alone [141]. In a phase II randomized, controlled, double-blind trial for stage IV metastatic melanoma, the addition of elesclomol to paclitaxel doubled the median progression-free survival (PFS) and reduced the likelihood of disease progression or death by 41.7% compared to paclitaxel alone [108]. Although the PFS endpoints were not met in a randomized, double-blind phase III trial of elesclomol in conjunction with paclitaxel in patients with chemotherapy-naive advanced melanoma, elesclomol demonstrated a high tumor-suppressive effect in patients with low lactate dehydrogenase (LDH) levels [109]. Low levels of LDH imply a mitochondrial hyperdynamic condition, which aligns with the observation that proteins involved in oxidative phosphorylation with elevated levels of lipoylation are more sensitive to elesclomol [94]. However, further clinical trials are needed to investigate the potential of using Cu ionophores to treat malignancies with active TCA cycles. In summary, combining Cu ionophores with targeted therapeutic agents can combat tumors with high mitochondrial metabolism levels, particularly those in a dynamic state of oxidative phosphorylation. The use of LDH as a predictor and prognostic indicator can guide this treatment effectively. In addition to utilizing LDH to assess tumor sensitivity to cuproptosis, another study demonstrated that in SF3B1-mutated leukemias, ABCB7 is mis-spliced and downregulated, leading to increased susceptibility to Cu ionophores. Furthermore, ABCB7 overexpression was shown to partially rescue the cuproptosis associated with the SF3B1 mutation, thereby suggesting that SF3B1 mutations could serve as a biomarker for future therapies based on Cu ionophores [142]. Overall, these findings suggest that Cu ionophores may be a promising strategy for inhibiting tumors with specific metabolic characteristics.

Traditional Cu ionophores are associated with various off-target effects, such as excessive Cu could lead to neurological side effects including seizures and brain circuit disturbances [143], leading to the emergence of nanomaterials designed for specific Cu-induced cell death in tumor tissues. Cancer stem cells (CSCs) typically reside in the hypoxic tumor microenvironment (TME) of triple-negative breast cancer (TNBC), where the expression of the cuproptosis-related protein FDX1 is inhibited, thereby diminishing the anti-cancer efficacy of cuproptosis. In response to this challenge, Xiao C et al. developed CuET@PHF, an active targeting cuproptosis-based nanomedicine that responds to reactive oxygen species (ROS).  This formulation employs polydopamine and hydroxyethyl starch to stabilize CuET nanocrystals, effectively eradicating CSCs. The photothermal effect of CuET@PHF is capable of mitigating tumor hypoxia, which enhances both cuproptosis and immunogenic cell death in 4T1 CSCs. Consequently, the combination of CuET@PHF with mild photothermal therapy not only significantly inhibits tumor growth but also effectively prevents tumor recurrence and distant metastasis by targeting CSCs and bolstering anti-tumor immune responses [144].

While the potential of Cu overload as a cancer treatment strategy is promising, it is crucial to acknowledge that elevated Cu levels can lead to various side effects. Additionally, the poor water solubility of ES-Cu might be another hindrance to the efficacy of Cu overload-based anti-tumor therapies [145]. Therefore, further efforts need to pay more attention to the specificity and solubility of Cu ionophores in tumors, employing a nanoparticle delivery system could be a rational solution. Moreover, both Cu ionophores and chelators have shown potential in tumor therapy; therefore, determining the optimal treatment conditions, such as classified biomarkers and combined therapeutics, remains an emergent challenge in utilizing these medications in the future.