Copper homeostasis imbalance and cuproptosis: Emerging targets and strategies for gastrointestinal tumor therapy

Abstract

Gastrointestinal malignancies, including gastric, colorectal, and pancreatic cancers, are characterized by metabolic reprogramming, multidrug resistance, and immunosuppressive microenvironments that limit clinical outcomes. Cuproptosis is a mitochondrial respiration-dependent form of regulated cell death that may target the “cuproplasia” phenotype of gastrointestinal tumors. Mechanistically, ferredoxin 1-mediated reduction of Cu²⁺ to Cu⁺ promotes toxic aggregation of lipoylated tricarboxylic acid cycle proteins, including dihydrolipoamide S-acetyltransferase, and destabilizes iron-sulfur clusters. Therapeutic disruption of copper homeostasis may therefore induce irreversible proteotoxic stress by altering the balance between glycolysis and oxidative phosphorylation. This review summarizes mechanisms of copper homeostasis imbalance and cuproptosis in gastrointestinal malignancies and discusses emerging interventions, including synthetic copper ionophores, traditional Chinese medicine phytochemicals, modulation of the dietary copper-gut microbiota axis, and stimuli-responsive nanomedicine. We further highlight nanomedicine-enabled stromal barrier penetration, magnetic resonance imaging/positron emission tomography-guided theranostics, immunogenic cell death-mediated synergy with immunotherapy, and translational challenges involving ferredoxin 1-based patient stratification and safety window management.

Key Words: Gastrointestinal neoplasms; Cuproptosis; Metabolic reprogramming; Ferredoxin 1; Nanomedicine; Traditional Chinese medicine; Gut microbiota

Core Tip: Gastrointestinal malignancies exhibit profound metabolic reprogramming and “cuproplasia”, heavily relying on copper homeostasis for progression and drug resistance. Targeting cuproptosis - a ferredoxin 1-mediated cell death causing toxic aggregation of lipoylated proteins like dihydrolipoamide S-acetyltransferase - offers a transformative strategy to bypass conventional therapeutic resistance. This review highlights multifaceted interventions, including repurposing synthetic copper ionophores, leveraging traditional Chinese medicine phytochemicals, and modulating the unique “dietary copper-gut microbiota” axis. Furthermore, deploying stimuli-responsive nanomedicines helps overcome dense stromal barriers and induces immunogenic cell death, synergizing with immunotherapy to advance the precision management of refractory gastrointestinal tumors.

INTRODUCTION

In the clinical treatment of gastric carcinoma (GC) and colorectal cancer (CRC), advanced metastasis, multidrug resistance, and low response rates to immunotherapy constitute major causes of mortality. Recent studies have progressively revealed that tumor cells not only drive malignant phenotypes through genetic mutations but also adapt to extreme microenvironmental stresses via profound “metabolic reprogramming”, thereby evading drug-mediated killing and immune surveillance. This shift from a purely genetic perspective to a metabolic perspective provides a novel theoretical framework for explaining and overcoming the clinical challenges of GC/CRC.

Drug resistance to conventional chemotherapeutic agents (e.g., 5-FU, oxaliplatin, cisplatin) is a key factor contributing to treatment failure. Studies have demonstrated that tumor cells resist chemotherapeutic stress by enhancing glycolysis, nucleotide metabolism, and redox homeostasis. For instance, in 5-FU-resistant CRC cells, the N6-methyladenosine (m6A) methyltransferase methyltransferase-like 3 significantly increases glycolytic flux by stabilizing hypoxia-inducible factor-1α (HIF-1α) mRNA and promoting lactate dehydrogenase A translation, thereby mediating drug resistance[1]. Similar mechanisms are observed in oxaliplatin resistance, where poly (ADP-ribose) polymerase activation leads to nicotinamide adenine dinucleotide depletion, subsequently inhibiting silent information regulator 1 (SIRT1) activity. Cancer cells gain survival advantages by upregulating glycolysis [via pyruvate kinase M2 (PKM2) and lactate dehydrogenase A], and restoring SIRT1 or targeting glycolysis can reverse this drug-resistant phenotype[2,3]. Beyond glucose metabolism, alterations in the nucleotide synthesis pathway are also critical for drug resistance. The redistribution of dihydroorotate dehydrogenase from mitochondria to the cytoplasm promotes de novo pyrimidine synthesis, competitively blocking 5-FU efficacy at the metabolic level[4]; whereas inosine monophosphate dehydrogenase 2-mediated enhancement of purine metabolism leads to oxaliplatin resistance by inhibiting caspase-dependent apoptosis[5]. Additionally, reprogramming of the serine metabolic pathway (e.g., forkhead box C1-driven upregulation of phosphoglycerate dehydrogenase) has been shown to be closely associated with 5-FU resistance in CRC[6]. In GC, Nrf2-mediated oxidative stress defense and metabolic reprogramming are significant drivers of cisplatin resistance[7].

Although immune checkpoint inhibitors (ICIs) are effective in certain patients, the “cold tumor” microenvironment limits their widespread application. Metabolic competition and signal crosstalk between tumor cells and immune cells are key factors leading to immune evasion. For example, the activation of the aldehyde dehydrogenase 2/arginase 2 axis results in arginine depletion in the tumor microenvironment (TME), directly inhibiting the proliferation and antitumor function of CD8+ T cells, thereby leading to resistance to ICI therapy[8]. Abnormal lipid metabolism plays a significant role in immune suppression: In microsatellite stable CRC, the accumulation of oxidized low-density lipoprotein not only promotes tumor progression but also induces an immunosuppressive microenvironment through metabolic reprogramming, thereby weakening the efficacy of ICI[9]. Multidisciplinary studies further reveal that gene-driven metabolic heterogeneity, such as that driven by angiopoietin-like protein 4 and fatty acid-binding protein (FABP7), directly shapes the immune low-reactivity subtype (immunity_low), characterized by the enrichment of M0 macrophages and stromal activation[10]. Notably, the interleukin-4 signaling axis can induce metabolic reprogramming in macrophages, leading to upregulation of Fc gamma receptor IIB and resistance to programmed death-1 antibody therapy in GC[11]. Conversely, specific metabolic pathways such as fatty acid oxidation are crucial for maintaining the survival of tissue-resident memory T cells in GC. Blocking programmed death ligand-1 (PD-L1) can increase the expression of FABP4/5 in tissue-resident memory T cells, promoting their lipid uptake and survival, thereby enhancing antitumor immunity[12].

Organ-specific metastasis is metabolically dependent. The distant metastases of GC and CRC exhibit distinct organ tropism, which is closely associated with the specific metabolic “soil” of metastatic foci. In liver metastases, tumor cells often undergo a metabolic shift from glycolysis to oxidative phosphorylation (OXPHOS). Single-cell sequencing has demonstrated that malignant cells in CRC liver metastases exhibit enhanced OXPHOS, which is regulated by transforming growth factor-β signaling[13]. Studies on the traditional Chinese medicine (TCM) decoction Jianpi Jiedu Formula have also confirmed that inhibiting glutamine metabolism and remodeling the YTHDF1/GID8 axis can effectively block CRC liver metastasis[14]. Additionally, stromal cells in the TME play a synergistic role. For example, secreted phosphoprotein 1 fibroblasts promote the growth of liver metastases through metabolic reprogramming[15], while FABP7-mediated macrophage lipid accumulation induces tumor cell metabolic reprogramming via the exosome pathway, forming a pre-metastatic niche[16]. In lymph node metastases, lipid metabolism plays a particularly prominent role. GC cells in the high-fat microenvironment of lymph nodes rely on fatty acid oxidation-driven OXPHOS to sustain survival, with the nuclear factor erythroid 2-related factor 2-solute carrier family 7 member 11 (SLC7A11) axis further supporting this process through antioxidant stress mechanisms[17]. Lipoprotein lipase-mediated lipid uptake and leptin-induced angiopoietin-like protein 4 phosphorylation are also critical mechanisms promoting GC lymph node metastasis[18]. Furthermore, upon reaching the lymph nodes, GC cells adapt to the high arachidonic acid environment by inhibiting the peroxisome proliferator-activated receptor gamma-FABP1 axis, thereby evading ferroptosis[19]. In peritoneal metastasis, the deubiquitinase USP35 promotes energy metabolism reprogramming and peritoneal adhesion in GC cells by stabilizing stimulator of interferon genes (STING) and activating the HIF-1α/focal adhesion kinase pathway[20]. Meanwhile, yrroline-5-carboxylate reductase 2-mediated abnormal proline metabolism has been demonstrated to be closely associated with peritoneal metastasis in CRC[21].

Recent research perspectives have delved into metabolism regulation by epigenetics and post-translational modifications. Lactylation, a novel epigenetic marker linking metabolism and gene expression, has been found to be extensively involved in DNA damage repair and drug resistance modulation. For instance, lactic acid produced during glycolysis can promote lactylation at the K288 site of the XRCC4-like factor protein, enhancing non-homologous end joining repair efficiency, thereby leading to chemotherapy resistance. The long non-coding RNA USP3-AS1 stabilizes MYC proto-oncogene through deubiquitination, subsequently promoting histone H3K18 lactylation and driving CRC liver metastasis[22]. Additionally, the deletion of ribosomal protein uL3 triggers severe disturbances in amino acid and glutathione (GSH) metabolism, influencing ferroptosis sensitivity by regulating SLC7A11 expression, revealing a complex regulatory network between metabolism and cell death pathways[23].

Tumor cells exhibit an extreme dependence on specific metal ions, particularly copper ions, during metabolic reprogramming, a phenomenon defined by the scientific community as “copper tropism” or “copper-dependent proliferation” (cuproplasia)[24-26]. Compared to normal tissues, tumor tissues and patient serum often demonstrate significantly elevated copper levels. Copper is not merely present as a micronutrient but also serves as a limiting nutrient factor for cancer cell proliferation, metabolism, tumorigenesis, and progression[27,28]. To meet their excessive demand, tumor cells remodel their intracellular copper homeostasis system by upregulating copper transporters and copper chaperones[29,30]. For instance, in cholangiocarcinoma and hepatocellular carcinoma (HCC), the high expression of solute carrier family 31 member (SLC31A1) and antioxidant protein 1 has been demonstrated to be closely associated with malignant progression, poor prognosis, and copper accumulation in tumors. These proteins support tumor cell survival by maintaining mitochondrial function and redox balance[29]. This metabolic dependence on copper renders copper homeostasis regulation a distinct metabolic characteristic that differentiates tumor cells from normal cells[26,31].

The traditional role of copper ions in tumor biology is primarily manifested as signal transduction modulators and enzyme cofactors, driving tumor angiogenesis, proliferation, and metastasis. At the signal transduction level, copper is no longer merely regarded as a static enzyme catalytic center but rather as a dynamic signal-regulating molecule that directly targets key oncogenic proteins, thereby activating classic pro-oncogenic signaling pathways, such as mitogen-activated protein kinase/extracellular regulated protein kinase and phosphatidylinositol 3-kinase (PI3K)/protein kinase B (AKT)[24,30]. Additionally, the copper chaperone protein antioxidant protein 1 not only facilitates copper transport but also functions as a transcription factor entering the nucleus, upregulating cyclin D1 expression and directly promoting tumor cell proliferation[32-34]. In malignant tumors such as CRC, copper mediates the vigorous growth of tumor cells by regulating the activity of various enzymatic and non-enzymatic factors[22,26].

Cuproptosis, a recently defined form of regulated cell death, exhibits mechanisms distinct from traditional apoptosis, ferroptosis, necroptosis, and pyroptosis[35-38]. This form of cell death strictly depends on intracellular copper ion accumulation and mitochondrial respiratory function. Mechanistically, cuproptosis is driven by the accumulation of intracellular copper in mitochondria, where copper binds to lipoylated tricarboxylic acid (TCA)-cycle proteins, particularly components of the pyruvate dehydrogenase complex such as dihydrolipoamide S-acetyltransferase (DLAT). This interaction promotes the aggregation of lipoylated proteins and is accompanied by the destabilization of iron-sulfur (Fe-S) cluster proteins[39-42]. This process triggers irreversible protein toxicity stress and mitochondrial metabolic dysfunction, ultimately resulting in cell death[37,43]. Ferredoxin 1 (FDX1) functions as a key upstream regulator by reducing Cu²⁺ to Cu⁺ and by supporting lipoic acid synthase (LIAS)-dependent protein lipoylation, thereby influencing the abundance of lipoylated mitochondrial substrates that determine cellular sensitivity to cuproptosis[44].

Tumor cells typically exhibit a high demand for copper to sustain their rapid proliferation, yet this also reveals their specific metabolic vulnerability[24,45]. The occurrence of cuproptosis is highly dependent on active mitochondrial respiration; therefore, tumor cells that primarily rely on glycolysis (Warburg effect) for energy supply often demonstrate resistance to copper-induced death[46]. Studies have shown that hypoxic conditions or glucose metabolic reprogramming can suppress TCA cycle activity by downregulating FDX1 and DLAT, activating HIF-1α, along with other mechanisms, thereby evading cuproptosis[47]. Conversely, metabolic interventions that shift tumor cells toward mitochondrial OXPHOS may increase the abundance of lipoylated TCA-cycle proteins and thereby enhance susceptibility to copper-induced proteotoxic stress[48-51].

To explore the therapeutic relevance of this mechanism, copper ionophores and engineered nanoplatforms have been investigated as preclinical strategies to modulate intracellular copper homeostasis[52-55]. These strategies aim to disrupt intracellular copper homeostasis, through mechanisms including downregulation of copper efflux proteins or upregulation of copper influx transporters to induce intracellular copper overload[41,56,57]. Stimulus-responsive nanodrugs may improve tumor-selective copper delivery and enhance oxidative or mitochondrial stress by depleting GSH or modulating metabolic pathways[58-60].

The application potential of cuproptosis extends to reshaping the tumor immune microenvironment. Copper overload-induced mitochondrial damage may promote mitochondrial DNA (mtDNA) release and cyclic GMP-AMP synthase-stimulator of interferon genes (cGAS-STING) activation, thereby contributing to immunogenic cell death-like responses under certain condition[61,62]. This process may facilitate dendritic cell maturation and CD8⁺ T-cell infiltration, suggesting a potential mechanism for alleviating immunosuppressive TMEs.

Therefore, combining cuproptosis induction with immune checkpoint blockade, photodynamic therapy (PDT), or sonodynamic therapy has shown synergistic antitumor activity in several preclinical tumor models[40,51,63-65]. Nevertheless, whether these strategies can be translated into gastrointestinal cancer treatment requires further validation in tumor-specific models, clinically relevant delivery systems, and biomarker-guided patient stratification.

MOLECULAR MECHANISMS OF CUPROPTOSIS AND METABOLIC REPROGRAMMING IN GASTROINTESTINAL TUMORS

Copper intake and intracellular homeostasis disruption

The gastrointestinal tract, as a critical system for copper absorption and metabolism, relies heavily on the precise regulation of copper transporters to maintain copper homeostasis within tumor cells. These transporters primarily include copper transporters (CTR1/SLC31A1) and copper efflux transporters (ATP7A, ATP7B). In malignant digestive tract tumors such as GC and esophageal cancer, ATP7A is often abnormally overexpressed. Studies have shown that ATP7A not only participates in transmembrane copper ion transport within cells but is also closely associated with malignant progression and angiogenesis in tumors. Its high expression often indicates poor prognosis in GC patients, accompanied by deeper invasion and advanced pathological grading[66]. In esophageal cancer, the overexpression of ATP7A is negatively regulated by miR-495 and is directly linked to tumor angiogenic capacity and resistance to chemotherapeutic agents[67]. Additionally, as the primary channel for copper uptake, SLC31A1 (CTR1) exhibits significant correlations with the TME and prognosis in various cancers, and its expression levels directly influence intracellular copper and platinum drug accumulation[68].

Tumor cells establish drug resistance mechanisms and maintain intracellular homeostasis by remodeling the expression profiles of these transporters. This abnormal “copper accumulation” or “copper flow” is one of the core driving forces of chemotherapy resistance, particularly to platinum-based drugs. In CRC, drug-resistant cell lines typically exhibit downregulation of CTR1 and upregulation of ATP7A and ATP7B, leading to reduced intracellular drug accumulation[69]. Experimental evidence demonstrates that reducing ATP7A expression through gene silencing or using drug interventions such as gambogic acid to restore the “high CTR1, low ATP7A/B” expression pattern can significantly increase intracellular platinum drug concentrations and reverse resistance[70]. Additionally, overexpression of matrix Gla protein has been shown to confer drug resistance in CRC cells by downregulating CTR1 and upregulating ATP7A/B[69].

In addition to alterations in expression levels, genetic polymorphisms of copper transporters and differences in tumor tissue subtypes also profoundly influence copper homeostasis and therapeutic responses in gastrointestinal tumors. In neoadjuvant chemoradiotherapy for esophageal cancer, single nucleotide polymorphisms of SLC31A1 (rs10981694) and ATP7B (rs9535828) are strongly associated with patient survival rates and hematologic toxicity, suggesting potential effects of genetic background on copper/platinum transport efficiency[71,72]. Pathological subtype analysis of CRC revealed significant differences in chemoresistance genes between mucinous adenocarcinoma and non-mucinous adenocarcinoma. Notably, ATP7B expression was markedly reduced in mucinous adenocarcinoma, accompanied by distinct somatic mutation frequencies, indicating heterogeneity in copper homeostasis maintenance and drug transport mechanisms across tumor cell types, which may contribute to clinical prognostic differentiation[73,74].

The Warburg effect and metabolic plasticity: Coordination between glycolysis and mitochondrial respiration

Metabolic reprogramming is a major feature of gastrointestinal tumors, but the Warburg effect should not be interpreted simply as a complete replacement of mitochondrial respiration by glycolysis. Rather, tumor cells often maintain a dynamic balance between aerobic glycolysis and OXPHOS, allowing them to adapt to hypoxia, sustain biosynthesis, preserve stemness, and survive therapeutic stress.

HIF-1α is a central regulator of glycolytic activation in this process. Representative studies have shown that one cut homeobox 3 enhances glycolytic gene expression by preventing histone deacetylase 6-mediated HIF-1α deacetylation[75], whereas ADP-ribosyltransferase 1 promotes glucose transporter 1-dependent glycolysis through the PI3K/AKT/HIF-1α axis[76]. These findings support the view that HIF-1α acts as a key hub connecting hypoxic signaling with glucose metabolic reprogramming. However, glycolytic activation is not uniformly oncogenic in all contexts; negative regulators such as selenium binding protein 1 and fructose bisphosphatase 2 can suppress the Warburg phenotype by restraining HIF-1α signaling or glycolytic enzyme activity[77,78]. Therefore, the biological outcome of glycolytic reprogramming depends on the balance between oncogenic activation and metabolic checkpoint mechanisms.

Importantly, enhanced glycolysis does not mean that mitochondrial function is lost. In gastrointestinal tumors, mitochondrial metabolism is frequently remodeled rather than simply suppressed. Tumor necrosis factor receptor-associated protein 1 (TRAP1) provides a representative example of this complexity. TRAP1 depletion can impair mitochondrial complex I and induce compensatory HIF-1α target genes such as glucose transporter 1 and pyruvate dehydrogenase kinase 1[79], whereas high TRAP1 expression may support mitochondrial biogenesis through the peroxisome proliferator-activated receptor gamma coactivator 1α (PGC-1α)/mitochondrial transcription factor A pathway[80]. These observations suggest that glycolysis and OXPHOS are not mutually exclusive programs but can be coordinately adjusted according to cellular stress and tumor state.

This metabolic plasticity is particularly relevant to therapeutic resistance. IGF2BP2 promotes chemoresistance by stabilizing transcripts associated with mitochondrial respiratory chain complexes[81], and lactate can enhance the metastatic capacity of CRC stem cells through PGC-1α-mediated OXPHOS[82]. Consistently, 5-FU-resistant cells may depend on both glycolysis and mitochondrial ATP production[83], whereas oxaliplatin-resistant organoids can shift toward OXPHOS dependence[84]. These findings indicate that resistant tumors do not share a single metabolic phenotype; instead, their vulnerabilities may vary according to tumor subtype, treatment history, and microenvironmental conditions.

Several preclinical studies have attempted to exploit this metabolic flexibility. For example, metformin suppresses mitochondrial complex I activity and Wnt-related cancer stem cell self-renewal[85], while p53 can counteract the Warburg phenotype by promoting mitochondrial respiration through targets such as synthesis of cytochrome C oxidase 2[86]. Nevertheless, the therapeutic translation of Warburg- or OXPHOS-targeted strategies remains preliminary. Future studies should define which gastrointestinal tumor subtypes are glycolysis-dependent, OXPHOS-dependent, or metabolically flexible, and determine whether these metabolic states can guide rational combinations with cuproptosis-related interventions.

Core regulators: FDX1 and lipoylation

FDX1 is a mitochondrial reductase and a key upstream regulator of cuproptosis, with emerging clinical and biological relevance in gastrointestinal cancers, including GC and CRC. Pan-cancer analyses and multiple cohort studies have demonstrated that FDX1 typically shows a downregulated trend in gastric and CRC tissues[87-89], with this low expression often significantly associated with poor patient prognosis[88,90,91]. However, this expression pattern may vary across different cancer subtypes. The biological functions of FDX1 extend beyond its role as an electron donor in steroidogenesis. Mechanistically, FDX1 contributes to cuproptosis by reducing Cu²⁺ to Cu⁺ and supporting LIAS-dependent protein lipoylation, thereby increasing the pool of lipoylated mitochondrial proteins that can serve as copper-binding substrates[44,92,93].

DLAT, a lipoylated component of the pyruvate dehydrogenase complex, is one of the best-characterized execution substrates of cuproptosis. FDX1 supports LIAS-dependent lipoylation of mitochondrial enzymes, including DLAT, thereby influencing the availability of lipoylated substrates for copper binding[44,92]. Under copper overload, FDX1-mediated Cu⁺ generation promotes the abnormal aggregation of lipoylated DLAT and related lipoylated mitochondrial proteins. These insoluble aggregates impair mitochondrial enzyme function and induce proteotoxic stress, accompanied by Fe-S cluster protein loss and mitochondrial metabolic failure[42,43,94].

In GC and CRC models, targeting the FDX1-LIAS-DLAT axis has shown preclinical relevance for sensitizing tumor cells to copper ionophores. Studies have shown that enhancing mitochondrial uncoupling agents to activate the respiratory chain can upregulate the expression of FDX1 and DLAT, thereby markedly increasing the sensitivity of GC cells to copper ion carriers and promoting the toxic accumulation of DLAT[48]. Additionally, novel copper nanoplatforms can not only directly replenish intracellular copper by co-delivering copper ions and p53 activators but also inhibit glycolysis and increase TCA cycle flux by restoring p53 function, thereby upregulating FDX1 and DLAT to synergistically induce cuproptosis and overcome drug resistance[95,96]. Notably, certain miRNAs or specific genes may confer resistance to cuproptosis in tumor cells by inhibiting the expression or function of FDX1, providing targets for developing targeted sensitization strategies[97,98] (Figure 1).

Figure 1 Central molecular mechanism of cuproptosis and mitochondrial protein lipoylation. Cuproptosis in gastrointestinal tumor cells is driven by mitochondrial copper accumulation and protein lipoylation. Cu²⁺ enters cells through copper transporters such as copper transporters/solute carrier family 31 member 1 and is reduced to Cu⁺ by ferredoxin 1 within mitochondria. Cu⁺ subsequently binds to lipoylated dihydrolipoamide S-acetyltransferase, inducing its oligomerization, Fe-S cluster protein loss, proteotoxic stress, and mitochondrial metabolic collapse, ultimately triggering caspase-independent cuproptotic cell death. CTR1: Copper transporters; SLC31A1: Solute carrier family 31 member 1; GC: Gastric cancer; CRC: Colorectal cancer; LIAS: Lipoic acid synthase; FDX1: Ferredoxin 1; DLAT: Dihydrolipoamide S-acetyltransferase; TCA: Tricarboxylic acid cycle; ETC: Electron transport chain.

Overview of cuproptosis signaling pathways and their specificity in the digestive system

Cuproptosis should be distinguished from other regulated cell death modalities by its dependence on mitochondrial respiration, copper accumulation, lipoylated TCA-cycle protein aggregation, and Fe-S cluster protein destabilization. In contrast, apoptosis is primarily caspase-dependent, ferroptosis is driven by iron-dependent lipid peroxidation, and necroptosis relies on receptor-interacting serine/threonine kinase 1/receptor-interacting serine/threonine kinase 3/mixed lineage kinase domain-like protein signaling. Therefore, when these pathways coexist in the same model, they are described as interacting or parallel stress responses rather than as identical mechanisms.

The signaling pathway of cuproptosis centers on mitochondrial metabolism, with mechanisms fundamentally distinct from traditional cell death modes, such as apoptosis, ferroptosis, and necroptosis. The initiation of this process relies on the direct binding of copper ions to lipoylated TCA-cycle proteins. Specifically, mitochondrial Cu²⁺ is reduced by FDX1 to Cu⁺, which promotes the aggregation of lipoylated proteins such as DLAT[99,100]. This protein aggregation not only disrupts the enzymatic activity of the TCA cycle but also triggers the loss of Fe-S cluster proteins, resulting in proteotoxic stress and ultimately irreversible cellular damage and death[101-104]. Unlike apoptosis or ferroptosis, cuproptosis does not primarily depend on caspase activation or lipid peroxidation; instead, it preferentially affects cells with active mitochondrial respiration and abundant lipoylated mitochondrial substrates[105].

In digestive system tumors, the cuproptosis pathway exhibits unique regulatory networks and pathophysiological significance. Taking GC as an example, metal regulatory transcription factor 1 (MTF1), a key metal-regulated transcription factor, not only inhibits ferroptosis but also reduces cuproptosis sensitivity by upregulating ferritin heavy chain 1 and activating Fe-S cluster assembly protein 2-mediated Fe-S cluster assembly, revealing the complex interregulatory relationship between these two forms of metal death in GC[106]. In CRC, FDX1, a critical executor of cuproptosis, is closely associated with TME remodeling and immune evasion. Studies suggest that CRC cells may evade cuproptosis by downregulating FDX1 or upregulating copper efflux proteins, whereas copper ionophores or nanodrugs can restore intracellular copper stress and sensitize tumor cells to cuproptosis in preclinical models[107-109]. Additionally, the imbalance of copper metabolism is particularly pronounced in HCC. As the primary organ for copper metabolism, liver tumor cells often maintain a high copper state by upregulating CTR1 and downregulating ATP7B to support proliferation, rendering HCC potentially highly sensitive to cuproptosis inducers[110,111].

Notably, the specificity of the cuproptosis signaling pathway in digestive system tumors is also reflected in its close association with metabolic reprogramming. Since cuproptosis is highly dependent on mitochondrial respiration, digestive tract tumor cells that primarily rely on Warburg effect for energy supply often exhibit resistance to cuproptosis[107]. However, metabolic interventions that force tumor cells to switch to OXPHOS can significantly enhance their sensitivity to cuproptosis. This strategy has shown preliminary promise in preclinical studies of gastric cancer and CRC[112,113] (Figure 2).

Figure 2 Metabolic reprogramming determines cuproptosis sensitivity. The Warburg phenotype promotes glycolysis, suppresses mitochondrial oxidative phosphorylation and the tricarboxylic acid cycle, and reduces ferredoxin 1/dihydrolipoamide S-acetyltransferase -dependent cuproptosis sensitivity. In contrast, metabolic rewiring by pyruvate dehydrogenase kinase inhibition, mitochondrial activation, or related interventions restores tricarboxylic acid cycle/oxidative phosphorylation activity and increases lipoylated protein accumulation. Under this respiratory state, copper ionophores such as elesclomol enhance copper toxicity and resensitize gastrointestinal tumor cells to cuproptosis. GLUT: Glucose transporter; HIF: Hypoxia-inducible factor; MCT: Monocarboxylate transporter; HK2: Hexokinase 2; PFKFB3: 6-Phosphofructo-2-kinase/fructose-2,6-bisphosphatase 3; TCA: Tricarboxylic acid cycle; OXPHOS: Oxidative phosphorylation; FDX1: Ferredoxin 1; DLAT: Dihydrolipoamide S-acetyltransferase; ATP: Adenosine triphosphate; ETC: Electron transport chain; PDK: Pyruvate dehydrogenase kinase.

Cross-talk between cuproptosis and other cell death programs: Shared stress signals but distinct execution mechanisms

Cuproptosis can interact with ferroptosis, apoptosis, pyroptosis, and immunogenic cell death through shared upstream stress signals, including mitochondrial dysfunction, reactive oxygen species (ROS) accumulation, GSH depletion, Fe-S cluster instability, and innate immune activation. However, these pathways should not be conflated. Cuproptosis is defined by copper-dependent aggregation of lipoylated mitochondrial proteins and Fe-S cluster protein loss; ferroptosis is defined by iron-dependent lipid peroxidation; apoptosis is caspase-dependent programmed cell death; and immunogenic cell death refers to an immunological phenotype characterized by antigen release and immune activation.

Cuproptosis and ferroptosis, as two metal ion-dependent cell death pathways, exhibit significant overlap and interaction in their regulatory mechanisms. On one hand, cuproptosis and ferroptosis can act synergistically. For instance, in osteosarcoma, the bimetallic nanoenzyme initiates a “ferroptosis-cuproptosis” self-reinforcing loop by inducing metabolic collapse of Fe-S clusters: Cuproptosis leads to the degradation of Fe-S proteins, releasing free iron ions into the unstable labile iron pool, which subsequently enhances oxidative stress through the Fenton reaction, promoting ferroptosis; conversely, ferroptosis-associated mitochondrial damage exacerbates the toxicity of copper ions[114]. Additionally, in triple-negative breast cancer, Cu²⁺ released by CuFeTe₂ nanosheets depletes GSH and inhibits glutathione peroxidase 4 (GPX4) activity, thereby triggering ferroptosis; meanwhile, Cu²⁺-induced aggregation of lipoylated DLAT initiates cuproptosis, while GSH depletion and GPX4 inhibition promote ferroptosis-like lipid peroxidation; these processes may cooperate to enhance tumor cell killing in this model[115]. On the other hand, certain regulatory factors may exert opposing or complex regulatory effects on these two death pathways. For example, in GC, the MTF1 has been found to simultaneously suppress the sensitivity to both ferroptosis and cuproptosis. MTF1 reduces free iron by upregulating ferritin heavy chain 1 and inhibiting ferroptosis, while simultaneously activating Fe-S cluster assembly protein 2-mediated Fe-S cluster assembly, alleviating cuproptosis-induced protein toxicity stress[106]. This discovery reveals that targeting key nodes such as MTF1 may simultaneously disrupt tumor cells' defense mechanisms against both of these death pathways.

Cuproptosis-associated mitochondrial damage may contribute to immunogenic cell death-like responses under certain conditions, providing a potential mechanism for immune microenvironment remodeling. During cuproptosis-related mitochondrial stress, mtDNA release may activate the cGAS-STING pathway and induce type I interferon signaling[56,116]. This process may promote dendritic cell maturation, antigen presentation, and CD8⁺ T-cell infiltration, thereby partially alleviating immunosuppressive tumor features. In CRC, cuproptosis not only directly kills tumor cells but also promotes DC maturation and M1 macrophage polarization by inducing endoplasmic reticulum stress, leading to the release of damage-associated molecular patterns such as calreticulin exposure, high mobility group box 1, and ATP[62,117]. This immunometabolic effect may contribute to tumor cell killing and immune activation in preclinical models, but its durability and clinical relevance require further validation[56,59,118].

In summary, cuproptosis in gastrointestinal tumors is best understood as a mitochondria-dependent and lipoylation-centered cell death program. Its sensitivity is shaped by copper transport, FDX1-LIAS-DLAT activity, mitochondrial respiration, and interactions with other stress responses. Importantly, cross-talk with ferroptosis, apoptosis, or immunogenic cell death should be interpreted as pathway interaction rather than mechanistic equivalence (Table 1).

Table 1 Key cuproptosis regulators in gastrointestinal tumors.

Regulator	Main role in cuproptosis	Relevance in gastrointestinal tumors	Evidence level	Ref.
SLC31A1/CTR1	Mediates copper influx and contributes to intracellular copper accumulation	Associated with copper and platinum drug uptake; altered expression may affect therapeutic response in digestive tumors	Clinical association + preclinical evidence	[29,68]
ATP7A/ATP7B	Copper efflux transporters that reduce intracellular copper accumulation	Upregulation is associated with platinum resistance and altered copper homeostasis in GC, CRC, and esophageal cancer	Clinical association + functional studies	[66,67,69-74]
ATOX1	Copper chaperone involved in copper transport and transcriptional regulation	Supports tumor proliferation and copper-dependent signaling in gastrointestinal malignancies	Preclinical + clinical association	[29,32,34]
FDX1	Reduces Cu2+ to Cu+ and supports LIAS-dependent protein lipoylation	Frequently downregulated in GC/CRC; low expression is associated with poor prognosis and reduced cuproptosis sensitivity	Bioinformatic cohorts + mechanistic studies	[44,87-93]
LIAS	Catalyzes lipoic acid biosynthesis required for mitochondrial protein lipoylation	LIAS regulation affects cellular sensitivity to elesclomol and copper-induced cuproptosis	Preclinical mechanistic evidence	[44,121,123]
DLAT	Lipoylated TCA-cycle protein and key substrate for copper-induced aggregation	DLAT abundance may influence mitochondrial metabolism, prognosis, immune features, and response to cuproptosis induction	Preclinical + clinical association	[42,43,94,235,236]
MTF1/FTH1/ISCA2	Regulates metal stress, iron storage, and Fe-S cluster homeostasis	May suppress both ferroptosis and cuproptosis sensitivity in GC by promoting iron storage and Fe-S cluster assembly	Preclinical mechanistic evidence	[106]
HIF-1α	Promotes glycolytic reprogramming and reduces mitochondrial dependence	Hypoxia/HIF-1α activation may reduce cuproptosis sensitivity by suppressing TCA-cycle activity and FDX1/DLAT expression	Preclinical evidence	[47,76,269]
CDKN2A/SNHG26/E2F3	Resistance-associated regulators of elesclomol-induced cuproptosis	Associated with resistance to cuproptosis induction in CRC models	Preclinical evidence	[125,126,240,241]
PTBP3/COX11 axis	Alters mitochondrial copper handling	PTBP3-mediated COX11 exon skipping may reduce mitochondrial copper content and contribute to cuproptosis evasion in GC	Preclinical evidence	[243]

SLC31A1: Solute carrier family 31 member 1; CTR1: Copper transporters; ATP7A: ATPase copper transporting alpha; ATP7B: ATPase copper transporting beta; GC: Gastric cancer; CRC: Colorectal cancer; ATOX1: Antioxidant 1 copper chaperone; FDX1: Ferredoxin 1; LIAS: Lipoic acid synthase; DLAT: Dihydrolipoamide S-acetyltransferase; TCA: Tricarboxylic acid cycle; MTF1: Metal regulatory transcription factor 1; FTH1: Ferritin heavy chain 1; ISCA2: Iron-sulfur cluster assembly 2; HIF-1α: Hypoxia-inducible factor-1α; CDKN2A: Cyclin dependent kinase inhibitor 2A; SNHG26: Small nucleolar RNA host gene 26; E2F3: E2F transcription factor 3; PTBP3: Polypyrimidine tract binding protein 3; COX11: Cytochrome c oxidase assembly factor 11.

INDUCTION OF CUPROPTOSIS: A NOVEL STRATEGY FOR DRUG THERAPY IN GASTROINTESTINAL TUMOR

Reuse of copper ionophores

Copper ionophores, a class of small-molecule compounds capable of modulating intracellular copper levels, demonstrate significant repurposing potential in gastrointestinal tumor therapy. These agents primarily include elesclomol and disulfiram, which not only induce cuproptosis but also exert anticancer effects through interactions with other cell death pathways or by modulating the TME[119,120].

Elesclomol, a potent copper ion carrier, primarily exerts its effects on gastrointestinal tumors through mitochondrial metabolism and specific gene regulation. In esophageal squamous cell carcinoma (ESCC), elesclomol targets the key enzyme of cuproptosis-LIAS. Studies have demonstrated that the combination of elesclomol with O-linked N-acetylglucosamine transferase inhibitors effectively inhibits tumor growth, a mechanism involving TRIM21-mediated ubiquitination of ALKBH5, which subsequently downregulates LIAS expression to promote cuproptosis[121]. Additionally, the sensitivity of ESCC cells to elesclomol is closely associated with OXPHOS levels, with cells in a hyper-OXPHOS state being more prone to cuproptosis. HMGA1 has been identified as a key regulatory molecule mediating elesclomol-induced ROS production[122]. In GC, elesclomol also exerts its effects through an alternative LIAS downregulation mechanism, while high expression of kynureninase confers resistance to cuproptosis by inhibiting LIAS[123]. In CRC, elesclomol-induced copper overload not only directly triggers cuproptosis but is also found to promote the degradation of ATP7A, leading to mitochondrial copper retention and SLC7A11 degradation, thereby inducing copper-dependent ferroptosis[124]. However, drug resistance mechanisms also exist in CRC. For instance, SNHG26 resists elesclomol-induced cuproptosis by stabilizing CDKN2A mRNA[125], while overexpression of the transcription factor E2F3 enhances cellular tolerance to elesclomol[126]. To optimize delivery and enhance efficacy, nano-reactors based on elesclomol not only induce irreversible cuproptosis but also reshape the tumor immune microenvironment by releasing Mn²⁺ to activate the cGAS-STING pathway[127].

Disulfiram, an established alcohol abstinence drug, exhibits anticancer activity in gastrointestinal tumors primarily through its metabolic conversion into a diethyl disulfothiocarbamate copper complex in vivo. In GC treatment, disulfiram significantly inhibits cancer cell proliferation and invasion via dual suppression of the Wnt/β-catenin and nuclear factor kappaB signaling pathways[128]. Additionally, disulfiram demonstrates anti-Helicobacter pylori activity by downregulating the expression of virulence factors CagA and VacA, potentially contributing to the prevention of GC at its origin. In CRC, the mechanism of the disulfiram/Cu complex is more diverse: It can induce excessive autophagic cell death by targeting unc-51-like autophagy activating kinase 1[129]; or it can downregulate ALDH1A3 via the miR-16-5p/15b-5p axis, thereby disrupting PKM2-mediated aerobic glycolysis and inhibiting tumor progression[130]. More importantly, the copper diethyldithiocarbamate complex has been proven to be a potent immunomodulator that enhances the cytotoxicity of natural killer cells and T cells through the NKG2D-NKG2DL axis, transforming “cold” tumors into immunoinflammatory “hot” tumors[131]. In clinical translation strategies, the combination of disulfiram and 5-FU has demonstrated synergistic lethal effects in 3D CRC organoid models and can downregulate drug resistance markers[132].

To overcome the poor in vivo stability of traditional copper ion carriers, nanotechnology has been widely applied in the development of drug delivery systems, achieving synergistic effects of multiple death modes. For instance, the Cu-BTC@DDTC nanodrug constructed by encapsulating disulfiram in metal-organic frameworks not only enhances drug stability but also induces ferroptosis through modulation of the SLC7A11/GPX4 pathway[133]. Another study developed multifunctional nanoparticles loaded with doxorubicin and disulfiram, which achieved synergistic antitumor effects of cuproptosis, ferroptosis, and apoptosis by amplifying cellular oxidative stress and depleting GSH[134]. Additionally, hollow copper sulfide nanoparticles loaded with disulfiram, combined with near-infrared photothermal therapy (PTT) and electrospun nanofiber scaffold technology[135], significantly enhanced their therapeutic potential and selectivity in CRC[136].

Small-molecule modulators targeting the FDX1-LIAS-DLAT lipoylation axis

FDX1 is a mitochondrial reductase that plays a central role in cuproptosis by regulating copper reduction and lipoylation-dependent substrate availability. It reduces Cu²⁺ to Cu⁺ and supports LIAS-dependent lipoylation of TCA-cycle enzymes such as DLAT. When intracellular copper accumulates, FDX1-generated Cu⁺ binds to lipoylated mitochondrial proteins, especially DLAT, leading to protein aggregation, proteotoxic stress, and Fe-S cluster protein loss[92-94,137]. Based on this mechanism, small molecules that increase intracellular copper, enhance FDX1 activity, promote mitochondrial respiration, or stabilize lipoylated substrates are being explored as preclinical sensitization strategies. For example, a marine natural product called Chagosendine C has been reported to target FDX1 in CRC cells, increase intracellular copper and ROS levels, and induce G1 arrest and cuproptosis-like cytotoxicity, suggesting potential activity against oxaliplatin-resistant cells[138]. Similarly, in pancreatic cancer, chlorophyllin, a chlorophyll derivative, has been found to bind to the FDX1 protein, promoting the oligomerization of DLAT protein and amplifying intracellular copper toxicity. Its combination with gemcitabine showed synergistic antitumor activity in preclinical pancreatic cancer models[139].

In addition to compounds that directly target FDX1, enhancing FDX1 activity and the efficacy of copper ion carriers through metabolic modulation or combination therapy represents another critical strategy. In CRC, the metabolite 4-octylisocyanate inhibits aerobic glycolysis by targeting the key glycolytic enzyme GAPDH. This process is highly dependent on the presence of FDX1 to sensitize cuproptosis induced by elesclomol-Cu. FDX1 knockdown attenuates 4-octylisocyanate-mediated sensitization to elesclomol-Cu-induced cytotoxicity[140]. Quercetin has been reported to bind FDX1 and synergize with elesclomol-Cu by increasing mitochondrial oxygen consumption rate, promoting lipoylated protein aggregation and mitochondrial dysfunction, thereby sensitizing HCC cells to lenvatinib in preclinical models[141]. Ferroptosis inducers may enhance the cytotoxicity of copper ionophores in HCC by reducing FDX1 degradation and increasing lipoylated cuproptosis substrate abundance[142].

However, gastrointestinal tumor cells often downregulate FDX1 expression through multiple mechanisms to evade cuproptosis, providing a target for reverse regulation in drug development. In pancreatic cancer, SERPINB3 inhibits FDX1 transcription by activating the mitogen-activated protein kinase signaling pathway, leading to resistance to cuproptosis. metal-organic frameworks loaded with copper ions and mitogen-activated protein kinase inhibitors, developed to target this mechanism, can effectively restore tumor sensitivity to cuproptosis[97]. In ESCC, the metabolite lactic acid induces lactic acidification modification of NUDT21, altering the length of the 3’ untranslated region of FDX1, thereby suppressing FDX1 protein translation. The combination of lactic acid dehydrogenase inhibitors and elesclomol can disrupt this resistance axis[143]. Additionally, autophagy induced by hypoxia in the TME also downregulates FDX1[144], while certain long non-coding RNAs[145], transcription factors[146], and tumor suppressor p53[96] have been identified as positive regulators of FDX1. Therefore, restoring FDX1 expression or blocking negative regulatory pathways may represent a rational preclinical approach to resensitize tumor cells to cuproptosis.

Drug-based combination therapeutic strategies: Chemosensitization and immunotherapy

Cuproptosis provides a mechanistic basis for drug-based combination therapy in gastrointestinal tumors because it links copper homeostasis, mitochondrial metabolism, redox imbalance, and proteotoxic stress. In contrast to single-agent cuproptosis induction, combination strategies aim to exploit pre-existing metabolic vulnerabilities or to convert immunologically resistant tumors into more treatment-responsive states. Current evidence can be broadly divided into two categories: Chemotherapy sensitization and immune sensitization.

Chemotherapy resistance remains a major obstacle in gastrointestinal malignancies. Cuproptosis inducers or copper metabolism-targeting agents may enhance the efficacy of conventional chemotherapy by increasing intracellular copper accumulation, aggravating mitochondrial stress, and disrupting antioxidant defenses. In gastric cancer and other digestive system tumors, cuproptosis and ferroptosis show molecular interactions, suggesting that these two metal-dependent cell death pathways may create overlapping metabolic vulnerabilities[147]. In pancreatic ductal adenocarcinoma (PDAC), copper ionophores or copper transporter-targeting agents can increase intracellular copper levels and enhance the antitumor activity of cisplatin, possibly through shared transport mechanisms and cumulative mitochondrial damage[148]. In CRC, the ferroptosis inducer erastin suppresses glycolysis and promotes mitochondrial respiration, thereby increasing cellular sensitivity to cuproptosis through metabolic reprogramming[149]. Disulfiram combined with chemotherapy has also been reported to activate STING signaling through poly (ADP-ribose) polymerase 1 inhibition in pancreatic cancer, linking chemotherapy sensitization with innate immune activation[150]. These findings suggest that cuproptosis-based chemosensitization may be particularly relevant in tumors with altered copper transport, mitochondrial dependence, or impaired redox buffering.

Cuproptosis may also enhance antitumor immunity. Many gastrointestinal tumors, including gastric cancer and PDAC, display immunosuppressive or “cold tumor” features. Cuproptosis-related mitochondrial stress and cell death can promote the release of damage-associated molecular patterns, activate cGAS-STING signaling, and enhance dendritic cell maturation and CD8⁺ T-cell infiltration[151-154]. These mechanisms provide a rationale for combining cuproptosis inducers with immune checkpoint blockade. For example, copper/disulfiram-based strategies can promote ROS accumulation, induce cuproptosis-related immune activation, and improve the response to PD-L1 blockade[155]. Other approaches combine cuproptosis with pyroptosis to amplify inflammatory signaling and suppress CRC growth[156]. In liver-related gastrointestinal malignancies, mitochondrial-targeted copper complexes can disrupt mitochondrial function, induce mtDNA release, activate cGAS-STING signaling in macrophages, and promote M1-like polarization[157]. Dual-metal systems that induce both ferroptosis and cuproptosis may further remodel the immune microenvironment by reducing regulatory T cells and enhancing the efficacy of immune checkpoint blockade[158].

In PDAC, cuproptosis-based drug combinations may also interact with stromal and metabolic barriers. The nanoenzyme Pt@PCN-Cu induces cuproptosis through mitochondrial dysfunction and suppresses glycolysis by promoting hexokinase 2 dissociation, which is accompanied by PD-L1 upregulation and provides a rationale for combination with anti-PD-L1 therapy[60]. Delivery systems such as silk fibroin nanoparticles and thermosensitive hydrogels may improve intratumoral drug penetration, modulate stromal components, and enhance immune responses through cuproptosis-associated immunogenic cell death[159,160]. However, most of these strategies remain at the preclinical stage, and their efficacy is likely to depend on tumor subtype, metabolic state, stromal composition, and immune contexture. Therefore, the therapeutic value of cuproptosis-based drug combinations should be evaluated with greater attention to patient stratification and mechanism-based biomarkers.

Dietary copper intake and the gut microbiota: From metal homeostasis to a hypothesis-generating tumor axis

Dietary copper intake and the gut microbiota are closely connected through bidirectional regulation of metal availability, intestinal barrier integrity, microbial metabolism, and host inflammatory responses. However, the current evidence linking the dietary copper-gut microbiota axis to gastrointestinal tumor cuproptosis remains largely indirect. Most available studies demonstrate that intestinal microorganisms can influence copper absorption, bioavailability, and toxicity, whereas direct evidence showing that dietary copper reshapes tumor cuproptosis through microbiota-dependent mechanisms is still limited.

The gut microbiota may regulate copper homeostasis through several mechanisms, including physical barrier effects, modulation of intestinal pH and redox balance, and direct bioadsorption of metal ions[161,162]. Although host genetic factors remain dominant determinants of systemic copper levels, the abundance of specific bacterial taxa, such as Blautia and Lachnospiraceae UCG-008, has been associated with serum copper concentrations, suggesting that the microbiota functions as an environmental modifier of copper metabolism[163]. Certain probiotic strains may also participate in copper detoxification. For example, Bacillus coagulans XY2 exhibits copper tolerance and bioadsorption capacity and may promote host copper excretion by regulating the SREBF-1/2 pathway[164]. These findings support a role for the gut microbiota in copper handling, but they do not yet establish a direct tumor-specific cuproptosis mechanism.

Microbial metabolites provide a possible link between copper exposure and intestinal injury. Short-chain fatty acids, especially butyrate, are important mediators of epithelial barrier function and inflammatory control. In a piglet model, high copper intake depleted Lactobacillus johnsonii and disrupted the conversion of acetate to butyrate, thereby impairing tight junction integrity and increasing sensitivity to copper-induced intestinal damage[165]. Copper oxide nanoparticles also altered short-chain fatty acid profiles and affected immune-related signaling pathways, indicating that different copper forms may have distinct effects on microbiota-derived metabolites and intestinal immunotoxicity[166]. Conversely, some copper complexes may enrich short-chain fatty acid-producing bacteria under specific conditions[167]. Therefore, copper exposure can either impair or remodel microbial metabolism depending on dose, chemical form, exposure duration, and host context.

Beyond short-chain fatty acids, bile acid and lipid metabolism may also connect copper exposure with host metabolic injury. Excessive copper intake can disrupt bile acid homeostasis and promote non-alcoholic fatty liver disease through microbiota-dependent mechanisms[168]. In Wilson’s disease-related models, copper accumulation is associated with altered microbial function and dysregulated sphingolipid, glycerophospholipid, and triglyceride metabolism[169]. Chronic copper exposure can also perturb glycerophospholipid, retinol, and steroid hormone metabolism, accompanied by reduced Lactobacillus abundance and increased opportunistic pathogens[154,170]. These findings suggest that copper-microbiota interactions may influence epithelial injury, inflammation, and metabolic remodeling, all of which are relevant to gastrointestinal tumor biology.

Overall, the dietary copper-gut microbiota axis should currently be viewed as a biologically plausible but still hypothesis-generating framework in gastrointestinal cancer. The most robust evidence supports the role of gut microbiota in copper bioavailability, barrier function, short-chain fatty acid metabolism, and inflammatory toxicity. In contrast, direct evidence linking dietary copper intake, microbiota remodeling, cuproptosis regulators, and gastrointestinal tumor progression remains insufficient. Future studies should integrate dietary copper quantification, microbiome sequencing, tumor copper burden, cuproptosis markers, and longitudinal tumor outcomes to determine whether this axis represents a causal driver, a disease-associated correlate, or a therapeutic vulnerability (Figure 3).

Figure 3 Dietary copper-gut microbiota-tumor axis in gastrointestinal cancer. Dietary copper, gut microbiota, and microbial metabolites may influence copper availability, epithelial barrier integrity, inflammatory toxicity, and tumor-related metabolic states in gastrointestinal cancer. Dysbiosis may disrupt intestinal barrier function and alter copper handling, whereas protective metabolites such as short-chain fatty acids may help preserve barrier integrity and limit systemic copper toxicity. SCFA: Short-chain fatty acid.

Overall, pharmacological induction of cuproptosis offers a promising but still preclinical therapeutic direction. Copper ionophores, FDX1-axis modulators, and combination strategies may enhance tumor vulnerability, but their efficacy is likely to depend on mitochondrial state, copper transporter expression, immune contexture, and resistance mechanisms. Biomarker-guided selection will be essential before clinical translation.

NATURAL PRODUCTS AND TCM: MECHANISTIC OPPORTUNITIES AND EVIDENCE LIMITATIONS

Natural products and TCM-derived compounds have attracted attention as potential modulators of copper stress, mitochondrial metabolism, redox balance, and cuproptosis-related vulnerability. Nevertheless, this field requires careful interpretation. Many studies demonstrate that natural compounds can chelate copper, transport copper, regulate ROS, or affect mitochondrial function, but only a smaller subset directly proves canonical cuproptosis events such as lipoylated protein aggregation, Fe-S cluster protein loss, or dependence on FDX1/LIAS/DLAT-related pathways. Therefore, natural products should be discussed as mechanistic leads or hypothesis-generating candidates unless tumor-specific cuproptosis evidence is available.

Natural products as copper-binding or copper-transporting agents

Some phytochemicals can interact with copper ions and thereby modify copper-dependent cytotoxicity. Curcumin has been reported to act as an anticancer copper ion carrier, inducing copper stress and cuproptosis-related proteomic changes, including glutamine transporter activation and Fe-S cluster protein loss[171]. Chlorogenic acid and caffeic acid can exert copper-dependent pro-oxidative effects by promoting DNA damage[172]. In HCC, a copper ion carrier complex combined with quercetin has been shown to target the cuproptosis-related protein FDX1 and promote lipoylated protein aggregation, thereby reversing drug resistance[141]. These studies provide relatively direct mechanistic support for the involvement of copper stress and cuproptosis-related pathways.

However, many other natural product-copper studies remain less directly connected to cuproptosis. Copper-based delivery systems containing curcumin, including copper-aspartic acid bio-metal-organic frameworks, mitochondria-targeting copper nanocarriers, and copper-doped polydopamine nanoparticles, mainly demonstrate improved delivery, ROS generation, or chemodynamic therapy (CDT) effects[173,174]. Flavonoids such as quercetin, kaempferol, and luteolin can form copper complexes and regulate oxidative damage, antioxidant activity, or DNA interaction[175-177], but these findings do not necessarily indicate canonical cuproptosis. Similarly, plant-extract-mediated green synthesis of copper or copper oxide nanoparticles has shown antibacterial, antidiabetic, and anticancer activities[178-181], while natural product-loaded copper nanocarriers may improve solubility and synergistic cytotoxicity[182-185].

Thus, natural products may influence copper-dependent tumor vulnerability through at least three routes: Copper chelation or transport, ROS amplification, and mitochondrial metabolic remodeling. Among these, direct cuproptosis induction is currently best supported when studies demonstrate copper dependence, lipoylated protein aggregation, Fe-S cluster protein loss, and rescue by cuproptosis pathway modulation. Without these criteria, the evidence should be considered suggestive rather than conclusive.

Targeting mitochondrial metabolism and redox balance: Potential sensitization to cuproptosis

Because cuproptosis depends on mitochondrial respiration and the accumulation of lipoylated TCA-cycle proteins, natural compounds that remodel mitochondrial metabolism may alter cellular sensitivity to copper-induced death. Resveratrol can activate the SIRT1/PGC-1α pathway and enhance mitochondrial biogenesis and OXPHOS in certain tumor models[186,187]. Theoretically, this metabolic state may increase susceptibility to cuproptosis, although most available studies describe apoptosis, autophagy, or general mitochondrial stress rather than direct cuproptosis. In contrast, compounds such as oridonin, triptolide, and dihydroartemisinin can impair mitochondrial function, reduce ATP production, disrupt mitochondrial membrane potential, and induce ROS accumulation[188-190]. These effects may create a permissive environment for copper-dependent toxicity, especially when combined with copper ion carriers.

Natural products may also affect glycolysis-TCA-cycle balance. Shikonin targets PKM2, inhibits glycolysis, and modulates mitochondrial metabolism[191]. By suppressing glycolytic dependence, shikonin may force tumor cells toward mitochondrial respiration and thereby theoretically increase cuproptosis vulnerability. Flavonoids such as quercetin and kaempferol can also regulate mitochondrial enzyme activity and energy metabolism[192,193]. However, whether these changes increase lipoylated DLAT aggregation or alter FDX1/LIAS-dependent cuproptosis remains to be clarified.

Redox regulation represents another possible mechanism. Triptolide can induce ROS accumulation, deplete GSH, and impair antioxidant enzyme activity[193,194]. Curcumin and its derivatives also disrupt mitochondrial homeostasis through ROS generation and GSH depletion[195,196]. Artemisinin derivatives contain peroxide bridge structures that can produce radicals in the presence of iron or copper ions, leading to mitochondrial protein and lipid damage[197,198]. Berberine derivatives have also been identified as thioredoxin reductase inhibitors, weakening cellular antioxidant defenses[199]. These mechanisms may synergize with copper-induced proteotoxic stress, but they should not be equated with cuproptosis unless canonical cuproptosis markers are directly demonstrated.

Collectively, TCM-derived and natural compounds offer a rich source of copper-interacting or mitochondria-modulating agents. The strongest evidence currently supports their roles in copper binding, ROS regulation, mitochondrial stress, and drug delivery enhancement. By contrast, their direct role as gastrointestinal tumor cuproptosis inducers remains preliminary. Future studies should prioritize mechanistic validation using copper chelation rescue, FDX1/LIAS/DLAT pathway interrogation, lipoylated protein aggregation assays, Fe-S cluster protein detection, and tumor-specific in vivo models (Figure 4).

Figure 4 Multi-target regulation of cuproptosis by traditional Chinese medicine and natural products. Traditional Chinese medicine-derived compounds and natural products may modulate cuproptosis-related vulnerability through copper binding or transport, mitochondrial metabolic remodeling, reactive oxygen species accumulation, glutathione depletion, and Fe-S cluster destabilization. Representative agents illustrate the multi-component and multi-target potential of natural products in copper-dependent cancer therapy, although direct evidence for canonical cuproptosis remains context-dependent. ROS: Reactive oxygen species; GSH: Glutathione; CTR1: Copper transporter 1; MDR: Multi-drug resistance; DNA: Deoxyribonucleic acid; TCA: Tricarboxylic acid cycle; OXPHOS: Oxidative phosphorylation; TCM: Traditional Chinese medicine.

NANOMEDICINE-ENABLED CUPROPTOSIS THERAPY: DELIVERY, AMPLIFICATION, AND IMAGING GUIDANCE

Nanomedicine provides a practical strategy to improve cuproptosis-related therapy by enhancing copper delivery, reducing nonspecific toxicity, amplifying TME-responsive cytotoxicity, and enabling imaging-guided treatment. These advantages are particularly relevant for gastrointestinal tumors, which often exhibit stromal barriers, acidic pH, redox imbalance, hypoxia, and heterogeneous drug penetration.

However, most nanomedicine-based cuproptosis strategies remain preclinical. Many platforms demonstrate ROS production, mitochondrial injury, or tumor growth inhibition, but only a subset directly verifies canonical cuproptosis events, such as lipoylated TCA-cycle protein aggregation, Fe-S cluster protein loss, FDX1/LIAS/DLAT pathway dependence, or copper chelation rescue. Therefore, nanomedicine should be viewed as an enabling technology for improving copper-based therapy, rather than as definitive evidence that cuproptosis has been successfully translated into the clinic.

Improving drug delivery and reducing nonspecific toxicity

One major function of nanomedicine is to overcome the poor solubility, instability, and systemic toxicity of copper ionophores or copper-containing drugs. Disulfiram/Cu(DDC)₂-based formulations have been engineered using dendrimers, liposomes, polymeric carriers, and cyclodextrin-based systems to improve drug loading, serum stability, and tumor accumulation[200-202]. These strategies provide a useful delivery framework, but their translational value depends on whether they can achieve sufficient tumor exposure while avoiding off-target copper toxicity.

Tumor-selective activation is another important design principle. Rather than directly administering toxic copper complexes, some platforms use relatively inert precursors that are converted into active copper drugs within the TME. For example, Zn(DDC)₂ liposomes can undergo endogenous copper exchange in tumors to generate cytotoxic Cu(DDC)₂ in situ[203]. Similarly, pH- or redox-responsive nanoparticles can release disulfiram and Cu²⁺ under tumor-specific conditions, promoting copper diethyldithiocarbamate complex formation, ROS accumulation, and immunogenic cell death[204,205]. These systems are promising, but future studies should better define biodistribution, copper leakage, long-term clearance, and whether the observed cytotoxicity is truly cuproptosis-dependent.

TME-responsive amplification of copper-dependent cytotoxicity

TME-responsive nanoplatforms exploit acidic pH, high GSH levels, and abnormal H₂O₂ metabolism to enhance copper-dependent cytotoxicity. pH-responsive metal-organic frameworks, coordination polymers, and peroxide-containing nanomaterials can degrade under acidic tumor conditions and release Cu²⁺, H₂O₂, or chemotherapeutic payloads[206,207]. This strategy increases local copper availability and provides substrates for Fenton-like reactions while reducing premature systemic leakage[208-210].

GSH-responsive systems further amplify oxidative stress. High intracellular GSH normally protects tumor cells by scavenging ROS, but redox-active nanoplatforms can consume GSH and convert Cu²⁺ to catalytically active Cu⁺, thereby enhancing hydroxyl radical production[211-217]. This process weakens antioxidant defenses and may promote both copper-dependent mitochondrial injury and ferroptosis-like lipid peroxidation[218,219]. Some designs also incorporate glucose oxidase to generate H₂O₂ in situ, linking starvation therapy with chemodynamic amplification[213,220].

Despite their mechanistic appeal, TME-responsive platforms face important limitations. Tumor acidity, GSH abundance, and H₂O₂ levels vary substantially across tumor types and even within the same tumor. In addition, multifunctional platforms often make it difficult to distinguish cuproptosis from ferroptosis, apoptosis, CDT-related oxidative injury, or nonspecific toxicity. Future studies should combine copper distribution analysis, redox imaging, cuproptosis marker detection, and rescue experiments to clarify mechanism.

PTT, PDT, and CDT-enhanced cuproptosis

Copper-based nanomaterials are well suited for integrating CDT, PTT, PDT, and cuproptosis. In acidic and H₂O₂-rich TMEs, copper ions catalyze Fenton-like reactions to generate hydroxyl radicals, forming the basis of CDT[221]. Meanwhile, excessive intracellular copper can induce cuproptosis by binding to lipoylated TCA-cycle proteins, promoting protein aggregation and destabilizing Fe-S cluster proteins[222]. These two processes may reinforce each other because ROS damages mitochondria and disrupts copper homeostasis, while cuproptosis-associated proteotoxic stress weakens antioxidant defenses[223,224].

PTT and PDT provide external energy to amplify this process. Under near-infrared irradiation, PTT increases local temperature, accelerates Fenton-like reaction kinetics, and promotes copper ion release from nanocarriers[225-227]. PDT generates singlet oxygen and other ROS, which cooperate with CDT-derived hydroxyl radicals to create stronger oxidative stress[228,229]. GSH depletion further sustains this redox loop by reducing ROS scavenging and promoting Cu²⁺/Cu⁺ cycling[228,229].

To overcome insufficient H₂O₂ and excessive GSH in tumors, some platforms incorporate glucose oxidase or calcium peroxide to generate H₂O₂ in situ while consuming glucose[230,231]. Copper ions can also consume GSH and increase catalytically active Cu⁺, forming a self-amplifying loop involving H₂O₂ supply, GSH depletion, ROS generation, copper release, mitochondrial damage, and cuproptosis induction[232]. Although these therapies may also induce immunogenic cell death[62,117], their primary mechanism in this section should be understood as physical and chemical amplification of copper-dependent cytotoxicity (Figure 5).

Figure 5 Smart nanomedicine strategies for synergistic cuproptosis therapy. Engineered nanocarriers, including metal-organic frameworks, liposomes, and biomimetic platforms, enable tumor-targeted Cu²⁺ delivery and controlled Cu⁺ release in acidic and glutathione-rich tumor microenvironments. Local copper accumulation may promote cuproptosis-related cytotoxicity and synergize with chemodynamic, photothermal, photodynamic, and immune-activating strategies by enhancing mitochondrial damage and oxidative stress. MOF: Metal-organic framework; ECM: Extracellular matrix; GSH: Glutathione; GSSH: Glutathione thiol; GSSG: Glutathione disulfide; CDT: Chemodynamic therapy; PTT: Photothermal therapy; PDT: Photodynamic therapy; NIR: Near-infrared; ICD: Immunogenic cell death; DAMP: Damage-associated molecular pattern; DC: Dendritic cell; HMGB: High mobility group box protein; ATP: Adenosine triphosphate.

Theranostics

Theranostic nanoplatforms integrate tumor imaging with copper-based therapy, enabling visualization of nanodrug accumulation, biodistribution, and treatment response. For cuproptosis-related therapy, their potential value lies in linking copper delivery, redox activation, mitochondrial injury, and therapeutic monitoring. However, the current evidence is still mainly preclinical, and many platforms demonstrate general ROS-mediated cytotoxicity rather than definitive cuproptosis. Therefore, theranostic systems should be interpreted as enabling technologies for monitoring and guiding copper-based therapy, rather than as established clinical strategies for gastrointestinal tumor cuproptosis treatment.

Magnetic resonance imaging -guided strategies: Magnetic resonance imaging (MRI)-guided copper-based theranostics mainly rely on copper-containing materials or the incorporation of Mn- and Fe-based contrast components. These systems can enhance T1- or T2-weighted imaging while simultaneously mediating copper release, Fenton-like reactions, photothermal effects, or mitochondrial injury (Table 2).

Table 2 Magnetic resonance imaging-guided copper-based theragnostic strategies.

Imaging strategy	Representative platform	Main therapeutic mechanism	Evidence status/limitation	Ref.
MRI-guided copper delivery	Copper-DNA nanoparticles	Tumor imaging, Fenton-like ROS generation, mitochondrial dysfunction, ICD	Shows imaging-therapy integration, but cuproptosis specificity requires further validation	[270]
T1-weighted MRI + PTT/CDT	CuS/Mn2+ hybrid nanogels	MRI-guided photothermal and chemodynamic therapy	Strong imaging signal and synergistic therapy; mainly preclinical	[271]
T1-weighted MRI + PTT	Ultrasmall Cu1.2O nanoparticles	MRI-guided photothermal ablation	Demonstrates copper-based imaging potential; direct cuproptosis markers are limited	[272]
MRI-guided multi-metal therapy	Fe3O4@C/CuOx nanoenzymes	MRI guidance, chemotherapy, PTT, copper/iron-mediated cytotoxicity	Multifunctional but mechanistically complex	[273]
TME-responsive MRI	Cu-Fe peroxide-containing nanoparticles	Cu-assisted iron cycling and Fenton amplification	Strong redox rationale; biosafety and clearance need validation	[274]
MRI-guided PDAC therapy	Collagenase-modified copper-containing hollow mesoporous silica nanoparticles	Stromal modulation and copper-induced cytotoxicity	Tumor-specific design for PDAC; translation remains early	[255]
MRI-guided ferroptosis/cuproptosis/CDT	CuO2-loaded dendritic macromolecule/metal-polysaccharide nanocomposites	TME regulation, ROS amplification, ferroptosis and cuproptosis	Promising combination design, but pathway attribution remains challenging	[275]
MRI + NIR-triggered cascade therapy	Cu/Mn-doped iron oxide nanocrystals	Drug release, Cu2+ chelation, ferroptosis/cuproptosis cascade	Integrates imaging and cascade therapy; preclinical validation needed	[276]
Activatable MRI + PDT/CDT	GSH/H2O2-responsive magnetic Cu2+ nanocomplex	“On-off” MRI activation, Cu+ release, hydroxyl radical generation, PDT enhancement	Smart activation design; in vivo specificity requires further testing	[277]
MRI-guided chemo-immunotherapy	Phosphorus dendrimer-copper complexes	Mitochondrial dysfunction, ER stress, immunotherapy enhancement	Illustrates theragnostic potential, but not exclusively cuproptosis-driven	[278]

MRI: Magnetic resonance imaging; ROS: Reactive oxygen species; ICD: Immunogenic cell death; PTT: Photothermal therapy; CDT: Chemodynamic therapy; PDAC: Pancreatic ductal adenocarcinoma; TME: Tumor microenvironment; NIR: Near-infrared; GSH: Glutathione; ER: Endoplasmic reticulum.

Positron emission tomography and multimodal imaging strategies: Compared with MRI, positron emission tomography imaging using ⁶⁴Cu-labeled nanomaterials provides high sensitivity and quantitative tracking of tumor accumulation. Multimodal systems further combine positron emission tomography, MRI, computed tomography, fluorescence, photoacoustic imaging, or upconversion luminescence to improve diagnostic accuracy and treatment guidance. These systems are valuable for monitoring delivery and biodistribution, but their therapeutic relevance depends on whether imaging signals correlate with copper release, tumor retention, cuproptosis-specific markers, and treatment response (Tables 3 and 4).

Table 3 Positron emission tomography and multimodal imaging strategies for copper-based theragnostics.

Imaging strategy	Representative platform	Main therapeutic mechanism	Evidence status/limitation	Ref.
64Cu-PET + phototherapy	64Cu-labeled porphyrin-lipid nanoparticles	PET tracking of tumor uptake and photochemical ablation	Useful for biodistribution monitoring; cuproptosis relevance is indirect
64Cu-PET + chemotherapy	CCR2-targeted Cu@CuOx nanoparticles loaded with gemcitabine	PET imaging and drug delivery in PDAC	Tumor-targeted strategy; clinical validation required	[279]
64Cu-PET + chemo/radiotherapy	Albumin-binding camptothecin nanoprodrug	Tumor retention, PET imaging, chemotherapy and radiotherapy	Strong theragnostic design; mainly proof-of-concept	[280]
64Cu-PET targeted therapy	64Cu-labeled immunoliposomes	EGFR-targeted PET imaging and therapy	Applicable to receptor-high tumors; tumor selectivity is key	[281]
PET-guided BNCT	64Cu-chelated boron neutron capture therapy platform	PET tracking of boron agent accumulation	Supports treatment planning; not primarily cuproptosis-based	[282]
UCL/CT/MRI-guided therapy	Upconversion nanoplatforms encapsulating CuS	Radiotherapy and PTT	Multimodal imaging enables guidance, but platform complexity is high	[283]
Four-modal imaging	CuS and rare-earth nanoparticle-loaded zeolitic imidazolate composites	Imaging-guided chemoradiotherapy	Broad imaging capacity; translational reproducibility uncertain
MRI-guided X-ray PDT	Copper-cysteine nanoparticles	Deep-tumor PDT under X-ray irradiation	Suitable for deep lesions; mechanism is mainly ROS-based	[284]
Fluorescence/MRI-guided CDT	CuS/Gd2O3based nanoprobes	Thermally enhanced CDT	Useful dual-modal design; cuproptosis specificity remains to be proven	[285]
Multiparametric MRI response monitoring	CuS@GOD nanoparticles	Starvation therapy, ROS generation, treatment response assessment	Imaging mainly monitors response rather than directly proving cuproptosis	[286]

PET: Positron emission tomography; CCR2: C-C motif chemokine receptor 2; PDAC: Pancreatic ductal adenocarcinoma; EGFR: Epidermal growth factor receptor; BNCT: Boron neutron capture therapy; UCL: University College London; CT: Computed tomography; MRI: Magnetic resonance imaging; PTT: Photothermal therapy; PDT: Photodynamic therapy; ROS: Reactive oxygen species; CDT: Chemodynamic therapy.

Table 4 Therapeutic strategies targeting cuproptosis and their evidence level.

Strategy	Representative mechanism	Current evidence level	Main limitation	Ref.
Copper ionophores	Increase intracellular copper and induce mitochondrial copper toxicity	Mainly preclinical tumor models	Systemic toxicity, delivery instability, and patient selection remain unresolved	[119,120,122-127]
Disulfiram/CuET-based therapy	Forms cytotoxic copper complexes and induces oxidative/mitochondrial stress	Preclinical and organoid evidence	Stability, specificity, and mechanism heterogeneity	[128-134]
FDX1-LIAS-DLAT axis modulation	Enhances lipoylated substrate availability and copper sensitivity	Preclinical mechanistic evidence	Requires validated biomarkers and rescue experiments	[92-98,137-146]
Metabolic sensitization	Shifts tumor cells from glycolysis toward OXPHOS to increase cuproptosis susceptibility	Preclinical evidence	Tumor metabolic states are heterogeneous and context-dependent	[47,50,140,269]
Chemotherapy sensitization	Combines copper stress with platinum drugs, 5-FU, gemcitabine, or metabolic stress	Preclinical evidence in GC, CRC, and PDAC models	Optimal tumor subtype and treatment sequence remain unclear	[147-150]
Immunotherapy combination	Cuproptosis-associated mitochondrial stress may enhance cGAS-STING activation, ICD-like responses, and immune checkpoint blockade	Preclinical evidence	Immune activation is context-dependent and should not be considered the defining mechanism of cuproptosis	[151-160,287,288]
TME-responsive nanomedicine	pH/GSH/H2O2-triggered copper release, ROS amplification, and mitochondrial injury	Preclinical nanoplatform studies	Platform complexity, biodistribution, clearance, and reproducibility	[203-220,289-293]
PTT/PDT/CDT-enhanced cuproptosis	Uses external energy and redox cycling to enhance ROS production, copper release, and cuproptosis-related cytotoxicity	Preclinical nanomedicine evidence	Difficult to distinguish cuproptosis from CDT/PDT/PTT-induced nonspecific oxidative injury	[227,230-232]
Theranostic platforms	Imaging-guided copper delivery and response monitoring	Preclinical imaging studies	Imaging signal does not directly prove cuproptosis activation	[255,270-277,279-286,294,295]

CuET: Copper diethyldithiocarbamate; FDX1: Ferredoxin 1; LIAS: Lipoic acid synthase; DLAT: Dihydrolipoamide S-acetyltransferase; OXPHOS: Oxidative phosphorylation; GC: Gastric cancer; CRC: Colorectal cancer; PDAC: Pancreatic ductal adenocarcinoma; cGAS: Cyclic GMP-AMP synthase; STING: Stimulator of interferon genes; ICD: Immunogenic cell death; TME: Tumor microenvironment; GSH: Glutathione; ROS: Reactive oxygen species; PTT: Photothermal therapy; PDT: Photodynamic therapy; CDT: Chemodynamic therapy.

CHALLENGES, TRANSLATIONAL BARRIERS, AND FUTURE DIRECTIONS

Patient stratification and biomarkers

For the potential translation of copper metabolism- and cuproptosis-based strategies in gastrointestinal tumors, identifying patients most likely to benefit remains a major challenge. Because cuproptosis depends on mitochondrial respiration and lipoylated TCA-cycle proteins, an exploratory biomarker framework incorporating key regulators and metabolic signatures may help identify GC and CRC patients who are more likely to respond to copper ionophores.

FDX1, as a key upstream regulator of cuproptosis, directly determines the sensitivity of tumor cells to cuproptosis inducers. In CRC, FDX1 is typically underexpressed, but its overexpression is positively correlated with favorable patient prognosis and immune cell infiltration[233]. Mechanistic studies have shown that the long non-coding RNA plasmacytoma variant translocation 1 can activate FDX1 transcription through epigenetic modifications, thereby increasing tumor cell sensitivity to elesclomol[145]. Conversely, low FDX1 expression is often associated with accelerated epithelial-mesenchymal transition and malignant progression of tumors[234]. In gastric adenocarcinoma, FDX1 has also been confirmed as an independent prognostic factor, with its expression regulated by the LncRNA MALAT1/miR-328-3p axis[88]. Therefore, detecting FDX1 expression levels and its upstream regulatory network is the first step in screening patients who may benefit from cuproptosis therapy.

Cuproptosis depends on copper binding to lipoylated enzymes in the TCA cycle. Therefore, LIAS-dependent lipoylation status, DLAT abundance, and mitochondrial metabolic state may represent important determinants of cuproptosis sensitivity. In ESCC, TRIM21 promotes the nuclear entry of ALKBH5 to demethylate LIAS mRNA, leading to LIAS downregulation and consequently limiting cuproptosis. The combination with an O-linked N-acetylglucosamine transferase inhibitor restores LIAS levels and enhances the efficacy of elesclomol[121]. In GC, the tryptophan metabolizing enzyme KYNU downregulates LIAS expression through a non-canonical mechanism, resulting in tumor cell resistance to cuproptosis[123]. Additionally, DLAT, as a key component of the pyruvate dehydrogenase complex, is associated with poor prognosis and immune evasion in GC, but its high expression may also indicate potential responsiveness to cuproptosis inducers[235,236]. Notably, metabolic reprogramming provides guidance for sensitivity screening: Inhibition of glycolysis forces cancer cells to rely on mitochondrial respiration, significantly enhancing their sensitivity to cuproptosis[140]. Thus, evaluating LIAS/DLAT expression and the tumor’s glycolytic/OXPHOS propensity serves as an important basis for patient stratification[122] (Figure 6).

Figure 6 Clinical translation roadmap for precision cuproptosis-targeted therapy. Precision cuproptosis therapy requires patient stratification according to ferredoxin 1 expression, protein lipoylation status, and metabolic phenotype. Tumors with high ferredoxin 1/Lipoylation and oxidative phosphorylation dependence may respond to copper ionophore-based therapy, whereas glycolytic or stromal-rich resistant tumors may require metabolic rewiring, stroma remodeling, or nanodelivery strategies. Protective agents such as selenium may further improve the therapeutic window by reducing off-target hepato-renal toxicity. FDX1: Ferredoxin 1; DLAT: Dihydrolipoamide S-acetyltransferase; LIAS: Lipoic acid synthase; TCA: Tricarboxylic acid cycle; OXPHOS: Oxidative phosphorylation; PDK: Pyruvate dehydrogenase kinase; ECM: Extracellular matrix.

Given the limitations of single-gene prediction, molecular phenotyping models integrating multi-omics data provide more precise tools for patient stratification. Researchers have developed a pan-cancer phenotyping system based on cuproptosis and hypoxia characteristics, identifying a subtype with high metabolic activity and low fibrosis features as more sensitive to cuproptosis inducers[237]. In GC, a scoring system based on nuclear-encoded mitochondrial genes revealed that patients with high nuclear-encoded mitochondrial genes scores were enriched with cuproptosis-related genes, suggesting their potential sensitivity to cuproptosis therapy[238]. Additionally, cluster analysis combining metabolic features identified specific GC subtypes, whose upregulated TCA cycle characteristics were positively correlated with cuproptosis signaling[239]. For CRC, assessments based on HIF-1α signaling pathway status and vasoactive intestinal peptide receptor 1 expression can also assist in predicting patient responses to copper chelators or inducers[146].

In clinical translation, identifying molecular signatures underlying cuproptosis resistance is equally critical, as it aids in excluding inapplicable populations or formulating combination therapy regimens. CDKN2A has been confirmed as a key driver of cuproptosis resistance in CRC, promoting tumor survival by modulating glycolysis and copper homeostasis[240,241]. SNHG26 further enhances this resistance by stabilizing CDKN2A mRNA[125]. Additionally, overexpression of the transcription factor E2F3[126] and activation of B cell lymphoma/Leukemia 10 via the nuclear factor kappaB pathway[242] have been found to increase resistance to elesclomol in CRC cells. Specific splicing factors, such as PTBP3-mediated COX11 exon skipping, can also lead to reduced mitochondrial copper content, thereby conferring GC cells with the ability to evade cuproptosis[243]. Detection of these resistance factors' expression levels serves as a critical reference for clinical exclusion of monotherapy with cuproptosis agents or for suggesting the need for combination with sensitizer agents (e.g., PI3K/AKT/mammalian target of rapamycin inhibitors[244]).

In summary, patient stratification for cuproptosis-based therapy should not rely on a single marker. A practical biomarker framework should integrate copper transporters, FDX1/LIAS/DLAT expression, lipoylated protein abundance, mitochondrial respiration status, hypoxia signatures, and resistance-associated regulators. At present, these biomarkers remain exploratory and require prospective validation before clinical application.

Tumor heterogeneity and microenvironmental barriers

In clinical translational research of gastrointestinal tumors, the heterogeneity of the TME constitutes a major barrier to the application of cuproptosis therapy, with hypoxic conditions and metabolic reprogramming being key factors limiting its efficacy. Cuproptosis is a mitochondrial respiratory-dependent form of cell death, whose mechanism involves the binding of copper ions to acylated proteins in the TCA cycle and subsequent protein toxic stress. However, gastrointestinal tumors (such as GC, pancreatic cancer, and CRC) typically exhibit highly hypoxic characteristics. The activation of HIF-1α promotes a shift in tumor cells from OXPHOS to glycolysis and glutamine catabolism, which may impair cellular sensitivity to cuproptosis[245-247]. Studies have found that dissipation of mitochondrial membrane potential through mitochondrial uncouplers can enhance respiratory activity and upregulate the expression of key regulatory proteins FDX1 and DLAT, thereby significantly improving the sensitivity of GC cells to the copper ion carrier elesclomol[48]. This suggests that overcoming hypoxia-induced metabolic adaptation and reactivating mitochondrial respiratory function are critical strategies for enhancing the efficacy of cuproptosis therapy in hypoxic solid tumors.

In addition to metabolic disorders, the physical barrier formed by dense extracellular matrix and connective tissue proliferation severely impedes the penetration of cuproptosis inducers or copper-based cytotoxic agents. In PDAC and GC, cancer-associated fibroblasts and members of the lysyl oxidase (LOX) family are highly active, catalyzing collagen cross-linking, which leads to increased matrix stiffness and elevated interstitial pressure[248-250]. This dense matrix not only compresses blood vessels to exacerbate hypoxia but also restricts drug diffusion[251,252]. High expression of LOX is closely associated with poor prognosis and chemotherapy resistance in gastrointestinal tumors[253], while the use of pan-LOX inhibitors can effectively reduce matrix stiffness, improve tumor perfusion, and enhance chemotherapy efficacy[254]. Therefore, improving stromal penetration and achieving sufficient intratumoral delivery of copper ionophores remain key barriers for future translational development.

To address these microenvironmental barriers, multifunctional delivery systems with matrix-remodeling or hypoxia-responsive properties have been actively investigated in preclinical studies. A collagenase-modified copper-based nanodelivery platform can degrade dense extracellular matrix barriers, promote deeper penetration of copper and drugs, and use copper as an MRI-relevant component to support imaging-guided copper-based cytotoxic therapy[255]. Additionally, designing hypoxia-responsive nanomicrospheres or vesicles not only enables targeted drug release upon reaching hypoxic regions but also alleviates hypoxia, reduces matrix stiffness, or normalizes tumor vasculature through co-delivery of drugs, thereby synergistically enhancing antitumor efficacy[251,256-258]. For instance, bacterial-nanodrug conjugates targeting the degradation of oncogenic collagen have been demonstrated to effectively overcome matrix barriers and enhance immunotherapy outcomes[258]. These studies suggest that combining copper-based cytotoxic agents with TME modulation may help overcome stromal and hypoxia-related barriers, although further validation in gastrointestinal tumor-specific models is needed.

In summary, tumor heterogeneity, hypoxia, and stromal barriers may substantially limit the efficacy of cuproptosis-based therapy. Future strategies should not only deliver copper or copper ionophores, but also define the metabolic and microenvironmental states that permit cuproptosis activation.

Safety and therapeutic window

Copper ion carriers and copper-based nanomaterials demonstrate significant potential in inducing cuproptosis in tumor cells, but the core obstacle to their clinical translation lies in the cumulative toxicity of copper ions in non-targeted organs. Copper can contribute to tumor cytotoxicity, but excessive systemic copper exposure may also cause hepatotoxicity, nephrotoxicity, and neurotoxicity. Studies have shown that copper overload can lead to severe mitochondrial dysfunction and oxidative stress in hepatocytes, even triggering pathological changes resembling Wilson’s disease[259]. Additionally, co-exposure to copper and environmental pollutants may induce synergistic toxicity through the gut-heart axis, resulting in myocardial fibrosis and heart failure[260]. In animal models, high-dose exposure to copper oxide nanoparticles has been shown to significantly reduce the activity of antioxidant enzymes in hepatic and renal tissues, leading to severe oxidative damage and tissue necrosis[261]. Therefore, establishing a therapeutic window requires achieving sufficient tumor copper stress while minimizing nonspecific copper accumulation in the liver, kidneys, and brain.

To address systemic toxicity, environment-responsive “cuproptosis switches” and multifunctional nanoplatforms have been explored as potential approaches. Researchers have designed TME-sensitive nanocarriers that utilize tumor-specific acidic conditions, high GSH levels, or exogenous stimuli to precisely control copper ion release. For instance, a pH-responsive “cuproptosis switch” nanocapsule can remain “off” in the bloodstream to minimize off-target leakage, while activating Cu²⁺ release in the acidic TME and synergizing with STING agonists to achieve immune enhancement, thereby improving tumor selectivity while reducing systemic toxicity in preclinical models[56]. Another study developed an acid-responsive bifunctional copper-coordinated polymer, which not only enables pH-dependent copper release but also reverses metabolic reprogramming through p53 activators, enhancing cuproptosis sensitivity with limited organ toxicity observed in preclinical evaluation[95]. Additionally, hyaluronic acid-or aptamer-modified nanocatalysts can target CD44 or nucleolin on tumor cell surfaces, promoting drug-specific uptake and in situ catalysis, significantly reducing damage to normal tissues[262,263].

Another safety-oriented strategy is to exploit endogenous tumor copper or modulate copper transporters, rather than relying solely on high-dose exogenous copper supplementation. For instance, a modularly assembled nanoliposome was designed to “seize” endogenous copper: It first delivers the copper chelator diffusible signal factor to the tumor, where it chelates endogenous copper in situ to form toxic complexes while simultaneously releasing nitric oxide to inhibit the expression of copper export protein ATP7B, thereby promoting tumor-selective copper stress without substantially increasing systemic copper load in preclinical models[264]. Similar strategies include designing cascade-targeting nanomodulators that inhibit copper export proteins, exacerbating copper overload in tumor mitochondria by blocking copper efflux pathways[265]. This approach ingeniously exploits the metabolic vulnerabilities of tumors, avoiding the risks associated with systemic copper exposure.

In addition to optimizing delivery systems, the introduction of protective formulations is also a critical approach to balancing efficacy and safety. Studies have demonstrated that selenium supplementation can effectively mitigate copper-induced systemic toxicity, protect hepatic and renal functions, and alleviate immunosuppression by reprogramming the transcriptome and restoring redox homeostasis[266]. At the molecular design level, rational alkyl chain modification of copper ion carriers can enhance their lipophilicity and copper-binding affinity, enabling them to maintain high copper-induced apoptosis efficiency while exhibiting low systemic toxicity[267]. Furthermore, to address potential risks of copper overload, the development of gastrointestinal-restricted cellulose-based copper adsorbents serves as a potential detoxification strategy to assist in regulating copper homeostasis and further ensure treatment safety[268].

In summary, safety remains a central barrier to cuproptosis-based therapy. Future studies should systematically evaluate dose-response relationships, long-term copper retention, organ-specific toxicity, and reversibility of copper-induced damage. A clinically meaningful therapeutic window will require both tumor-selective activation and reliable strategies for limiting systemic copper exposure (Table 5).

Table 5 Major resistance mechanisms and possible sensitization approaches.

Resistance mechanism	Biological basis	Possible sensitization approach	Ref.
Low mitochondrial respiration/Warburg phenotype	Reduced TCA-cycle activity limits lipoylated substrate abundance and cuproptosis sensitivity	Inhibit glycolysis or shift metabolism toward OXPHOS	[46-51,75-86,269,296-306]
FDX1 downregulation	Reduces Cu2+ reduction and decreases lipoylation-related cuproptosis susceptibility	Restore FDX1 expression or block negative regulators	[87-98,143-146,307-311]
LIAS suppression	Reduces mitochondrial protein lipoylation and limits cuproptosis substrate formation	Restore LIAS expression or combine with OGT/epigenetic modulators	[121,123]
DLAT insufficiency or altered mitochondrial substrate availability	Limits aggregation of lipoylated DLAT and downstream proteotoxic stress	Increase mitochondrial TCA-cycle dependence or select DLAT-high tumors	[42,43,94,235,236]
Copper efflux activation	ATP7A/ATP7B-mediated copper export lowers intracellular copper accumulation	Inhibit copper efflux or enhance tumor copper retention	[66-74,312]
Hypoxia and HIF-1α activation	Promotes glycolysis and reduces mitochondrial respiration-dependent cuproptosis sensitivity	Combine with hypoxia modulation or metabolic reprogramming strategies	[47,245-247,269]
Dense ECM/stromal barriers	Limits penetration of copper ionophores and nanodrugs, especially in PDAC and GC	Use matrix-remodeling delivery systems or stromal-modulating strategies	[249-254,256-258]
Antioxidant buffering/high GSH	Scavenges ROS and weakens copper-dependent oxidative stress	Use GSH-depleting or redox-active nanoplatforms	[206-220,232,290-293,313]
Alternative death pathway dominance	Ferroptosis, apoptosis, ICD, or CDT-induced oxidative injury may coexist with cuproptosis, complicating mechanism attribution	Use pathway-specific inhibitors, copper chelation rescue, FDX1/LIAS/DLAT validation, and lipoylated protein aggregation assays	[62,106,114-118,221-232,312-316]
Systemic copper toxicity	Off-target copper accumulation may injure liver, kidney, brain, or other organs	Develop tumor-selective activation systems and evaluate long-term clearance and organ toxicity	[259-268]

TCA: Tricarboxylic acid cycle; OXPHOS: Oxidative phosphorylation; FDX1: Ferredoxin 1; LIAS: Lipoic acid synthase; OGT: O-GlcNAc transferase; DLAT: Dihydrolipoamide S-acetyltransferase; ATP7A: ATPase copper transporting alpha; ATP7B: ATPase copper transporting beta; HIF-1α: Hypoxia-inducible factor-1α; ECM: Extracellular matrix; PDAC: Pancreatic ductal adenocarcinoma; GC: Gastric cancer; GSH: Glutathione; ROS: Reactive oxygen species; ICD: Immunogenic cell death; CDT: Chemodynamic therapy.

CONCLUSION

Cuproptosis has introduced a new conceptual framework for understanding the relationship between copper homeostasis, mitochondrial metabolism, protein lipoylation, and tumor vulnerability. In gastrointestinal tumors, accumulating evidence suggests that cuproptosis-related regulators may participate in tumor progression, metabolic remodeling, immune microenvironment regulation, therapeutic resistance, and prognosis prediction. However, this field remains at an early stage. Since cuproptosis was only recently defined as a distinct form of regulated cell death, most current studies are based on bioinformatic analyses, tumor tissue validation, cell-line experiments, organoid models, and animal studies. Direct clinical evidence in gastrointestinal cancers is still limited, and cuproptosis-targeted therapy should therefore be regarded as an emerging research direction rather than an established clinical strategy.

Over the next 3-5 years, several key priorities should be addressed to move this field forward. First, standardized and clinically applicable biomarker systems are needed. Current studies frequently focus on individual regulators such as FDX1, LIAS, DLAT, and other cuproptosis-related genes, but these markers alone may not fully reflect functional cuproptosis sensitivity. Future studies should integrate copper metabolism, mitochondrial respiration, protein lipoylation, metabolic state, and tumor microenvironmental features to establish more reliable stratification models.

Second, more clinically relevant experimental systems are required. Patient-derived organoids, patient-derived xenograft models, spatial multi-omics, single-cell sequencing, and longitudinal clinical specimens may help clarify whether cuproptosis-related signatures are truly associated with tumor aggressiveness, treatment response, recurrence, or survival. These models will also be important for distinguishing robust tumor-specific findings from preliminary or context-dependent observations.

Third, the interaction between cuproptosis and the TME should be further clarified. Hypoxia, immune suppression, stromal remodeling, angiogenesis, copper availability, and metabolic competition may all affect the induction and therapeutic consequences of cuproptosis in vivo. Understanding these interactions will be essential for designing rational combination strategies with chemotherapy, immunotherapy, radiotherapy, metabolic therapy, or nanomedicine-based delivery platforms.

Finally, future translational studies should place greater emphasis on safety, selectivity, and patient stratification. Because copper is an essential trace element and mitochondrial metabolism is also required by normal tissues, indiscriminate copper modulation may cause systemic toxicity or off-target mitochondrial injury. Therefore, the next stage of research should move from broad copper manipulation toward biomarker-guided, tumor-selective, and mechanism-based therapeutic strategies. If these challenges can be addressed, cuproptosis may become a valuable component of precision oncology for selected gastrointestinal tumor subtypes.