The emerging role of disulfidptosis in metabolic synergistic death and cancer immunotherapy

Abstract

Disulfidptosis is an emerging form of regulated cell death (RCD) that is mechanistically distinct from, yet closely interconnected with, classical cell death modalities such as apoptosis, ferroptosis, and cuproptosis. It is typically triggered in cells with high expression of solute carrier family 7 member 11 (SLC7A11, also known as xCT) under glucose-deprived conditions. In this context, nicotinamide adenine dinucleotide phosphate (NADPH) becomes depleted, and consequently the level of reduced glutathione (GSH) declines, which in turn compromises the cellular capacity to efficiently resolve aberrantly accumulated intracellular disulfide bonds. As a result, disulfide stress progressively builds up and ultimately promotes abnormal crosslinking, aggregation, and functional disruption of cytoskeletal proteins, thereby driving catastrophic cytoskeletal collapse and cell death. Because SLC7A11 overexpression represents a key metabolic hallmark across multiple cancer types, disulfidptosis provides a novel conceptual framework for exploiting tumor-specific metabolic vulnerabilities. Moreover, accumulating evidence across diverse malignancies suggests that disulfidptosis contributes to tumor progression. Meanwhile, recent studies have increasingly highlighted evolving drug delivery strategies and metabolic synergistic approaches that integrate disulfidptosis with other RCD programs, and, importantly, its impact on the tumor immune microenvironment is gradually being elucidated. In this review, we systematically summarize the definition, molecular mechanisms, and biological functions of disulfidptosis, emphasize recent advances in its synergistic interplay with other cell death modalities, and discuss its implications for cancer immunotherapy. Finally, we outline emerging drug delivery strategies designed to induce disulfidptosis in tumor cells, aiming to facilitate further exploration and translational development of disulfidptosis-based anticancer therapies.

Introduction

Regulated cell death (RCD) refers to a form of cell death governed by specific molecular programs. Its initiation and execution rely on genetic regulation, signaling complex assembly, and cascade amplification, with important roles in organismal development, tissue homeostasis, and immune responses [1]. Conceptually, cell death can be divided into accidental cell death (ACD) and RCD, the latter encompassing apoptosis as well as multiple non-apoptotic subtypes, including necroptosis, pyroptosis, parthanatos, ferroptosis, and cuproptosis [2]. Unlike ACD, which is directly triggered by acute physical or chemical injury and lacks an orchestrated program, RCD is characterized by recognizable triggers, dedicated execution machinery, and pharmacologically actionable pathways [3]. In addition, apoptosis typically features cell shrinkage, DNA fragmentation, and apoptotic body formation, with relatively low immunogenicity, whereas multiple non-apoptotic forms of RCD are commonly associated with membrane disruption, DAMP release, and more pronounced inflammatory or immune effects [4]. Collectively, these features make RCD a highly promising strategy for targeting tumor vulnerabilities and improving cancer therapy [5].

However, therapy resistance driven by metabolic reprogramming—such as solute carrier family 7 member 11 (SLC7A11, also known as xCT)-mediated resistance to ferroptosis—has substantially limited the clinical translation of conventional RCD-inducing strategies [6]. Disulfidptosis, a newly recognized disulfide stress–driven mode of cell death, is tightly associated with defects in glucose metabolism and elevated expression of SLC7A11 in tumor cells, thereby opening a new avenue for RCD-based anticancer therapy [7]. Mechanistically, SLC7A11, a cystine–glutamate antiporter, suppresses ferroptosis under nutrient-replete conditions by promoting glutathione (GSH) biosynthesis; in contrast, under glucose deprivation, excessive cystine uptake mediated by SLC7A11 increases nicotinamide adenine dinucleotide phosphate (NADPH) consumption, leading to an imbalance in NADPH/NADP⁺, triggering disulfide stress, and subsequently inducing abnormal disulfide bond formation in filamentous actin (F-actin) [8]. This process disrupts cytoskeletal integrity and ultimately results in cell death. Disulfidptosis is distinguished from other RCD modalities by features that are not only mechanistic—occurring without ATP depletion or caspase-3 activation—but also pharmacological, as it is resistant to inhibitors of ferroptosis (e.g., Fer-1) and apoptosis (e.g., Z-VAD-fmk), and is uniquely triggered by glucose withdrawal specifically in SLC7A11-high expressing cells [9].

As the mechanistic framework of disulfidptosis continues to be refined, this cell death program has been found to exert diverse functions across multiple tumor types [10]. In several cancers, disulfidptosis-related genes serve as valuable prognostic markers and predictors of clinical outcome [11–13]. Notably, disulfidptosis occurring in immune cells within the tumor microenvironment (TME) may impair antitumor immune function and thereby facilitate tumor progression [14]. With deeper investigation, it has become increasingly evident that disulfidptosis may act synergistically with other cell death programs through metabolic crosstalk, thereby helping overcome therapeutic resistance in certain cancers. For instance, both disulfidptosis and ferroptosis are tightly linked to SLC7A11, but they are triggered in opposing metabolic contexts: under nutrient-replete conditions, SLC7A11 sustains cystine import and GSH biosynthesis to restrain ferroptosis; by contrast, under glucose deprivation, SLC7A11-driven cystine uptake leads to intracellular cystine accumulation, precipitating disulfidptosis [15]. Therefore, this “metabolic switch” property provides a rationale for combination therapy [16]. Encouragingly, researchers have also explored potential synergies between disulfidptosis and other death modalities such as cuproptosis and pyroptosis, and combination regimens have shown more pronounced antitumor effects than monotherapies across multiple cancer models, laying a foundation for future clinical translation [17, 18].

In addition, disulfidptosis may enhance the immunogenicity of the TME and exert a positive impact on the efficacy of immunotherapy [19]. Using diverse nano-adjuvants, investigators have precisely induced redox imbalance and extensive disulfide bond formation in tumor cells, thereby damaging cytoskeletal proteins and triggering disulfidptosis [20, 21]. Subsequently, a range of immunogenic molecules are released and exposed within the TME, further promoting immune cell infiltration, reversing immunosuppressive features and ultimately strengthening antitumor immunity [22, 23]. The emerging drug delivery strategies that precisely disrupt tumor homeostasis to induce disulfidptosis—and potentially other RCD programs—may offer new routes to improve anticancer immunotherapy.

In this review, we systematically summarize the core mechanisms, regulatory pathways, and biological functions of disulfidptosis across multiple cancer types. Subsequently, we focus on recent advances to analyze its metabolic synergistic effects with other regulated cell death modalities—particularly ferroptosis, cuproptosis, and pyroptosis—as well as its critical role in remodeling the tumor immune microenvironment. Finally, we discuss emerging disulfidptosis-based therapeutic strategies and innovative drug delivery approaches, while also highlighting the current challenges and future directions. Overall, this review seeks to establish a robust theoretical foundation, thereby facilitating the development of disulfidptosis-targeted anticancer therapies.

Results

Traditional studies on disulfidptosis

In 2023, a team led by Prof. Gan at The University of Texas MD Anderson Cancer Center first reported a novel form of RCD driven by disulfide stress, termed disulfidptosis [7]. Disulfidptosis typically occurs under glucose-deprived conditions, wherein suppression of the pentose phosphate pathway (PPP) leads to a decline in NADPH production (Fig. 1). In cells with high SLC7A11 expression, this results in excessive accumulation of imported cystine that cannot be promptly reduced, which in turn disrupts cysteine and GSH synthesis and exacerbates disulfide stress. Under sustained disulfide stress, F-actin forms extensive aberrant disulfide bonds, leading to actin contraction and detachment from the plasma membrane, ultimately culminating in cell death [24]. Accordingly, the defining features of disulfidptosis can be summarized in three dimensions. Metabolic features: Glucose deprivation suppresses the PPP and reduces NADPH generation, while high SLC7A11 expression drives massive cystine uptake, producing a characteristic metabolic imbalance of “NADPH deficiency coupled with cystine overload”. Molecular features: Intracellular disulfide-containing metabolites accumulate abnormally, and F-actin undergoes excessive disulfide bond formation, resulting in actin contraction and membrane dissociation. Functional features: disulfidptosis is largely insensitive to canonical inhibitors of other RCD programs, and its execution critically depends on the Rac family small GTPase 1 (RAC1)-WAVE regulatory complex (WRC)-actin-related protein 2/3 (Arp2/3) signaling axis [25, 26].

Fig. 1. Schematic of core molecular mechanism of disulfidptosis.

Glucose entry through GLUTs fuels the PPP, sustaining the NADP⁺/NADPH redox cycle and cellular reducing capacity. In SLC7A11-high cells, the system mediates cystine import coupled to glutamate export; reduction of imported cystine to cysteine imposes a major NADPH burden. When glucose availability/PPP flux is curtailed, NADPH production becomes limiting and the intracellular reducing environment collapses, despite continued cystine uptake. This imbalance promotes accumulation of cystine and other disulfide species and drives aberrant disulfide-bridge formation in actin cytoskeleton proteins (ACPs), destabilizing the F-actin network and culminating in actin collapse and disulfidptosis. The TXNRD1- and glutathione GSH/GSSG-dependent redox systems are depicted as key buffering pathways that normally restrain disulfide stress. RAC1–WRC signaling supports Arp2/3-dependent branched actin polymerization, which is proposed to provide a permissive structural context for ACP disulfide crosslinking, thereby facilitating disulfidptosis. “Disulfide bridge” denotes abnormal disulfide bonds formed within the actin cytoskeleton. Abbreviations: GLUTsglucose transporters, PPP pentose phosphate pathway, NADPH/NADP⁺ nicotinamide adenine dinucleotide phosphate (reduced/oxidized), SLC7A11 solute carrier family 7 member 11, GSH/GSSG glutathione/oxidized glutathione, TXNRD1 thioredoxin reductase 1, RAC1 RAS-related C3 botulinum toxin substrate 1, WRC WAVE regulatory complex, Arp2/3 actin-related protein 2/3 complex, ACPs actin cytoskeleton proteins, F-actin filamentous actin. This figure was created with BioRender (https://biorender.com/).

Molecular mechanisms of disulfidptosis

Core signaling pathway

Researchers initially observed that under conditions of high SLC7A11 expression combined with glucose starvation, cells underwent a novel form of death distinct from previously known cell death modalities. This process could not be effectively blocked by canonical pharmacological inhibitors of ferroptosis, apoptosis, necroptosis, or autophagy [7]. In contrast, thiol oxidants that promote sulfhydryl oxidation aggravated cell death, whereas reducing agents markedly suppressed it. Subsequent mechanistic studies further demonstrated that this form of death was closely associated with disulfide stress. Further screening results revealed that disulfide bonding in cytoskeleton proteins was among the most active and prominent alterations, leading researchers to focus their attention on the cytoskeletal system. Functional experiments then confirmed that destabilization and collapse of the F-actin network constitute a key phenotype of this process [7].

After actin was identified as the core execution endpoint of this death process, a central mechanistic issue was how intracellular disulfide stress is transduced into actin cytoskeleton collapse. To address this question, the original study performed genome-wide CRISPR/Cas9 screening and identified RAC1 together with the WRC, which signals through Arp2/3 to drive branched actin polymerization, as key determinants of this process [7]. Functional experiments further showed that RAC1 activation promotes lamellipodia formation and enhances disulfidptosis, whereas loss of NCK-associated protein 1 (NCKAP1), a core component of the WRC, significantly attenuates disulfidptosis [7]. Together, these findings establish the RAC1–WRC–Arp2/3 axis as the best-supported canonical pathway linking intracellular disulfide stress to actin cytoskeleton collapse in disulfidptosis. However, several important gaps remain in the current mechanistic understanding. First, it is still unclear why aberrant disulfide bond formation appears to be preferentially enriched in actin cytoskeleton proteins. Second, beyond the RAC1–WRC–Arp2/3 axis, it remains uncertain whether additional execution mechanisms exist in parallel with actin collapse, as no evidence of comparable certainty is currently available. Therefore, although current studies have preliminarily established an execution framework for disulfidptosis centered on actin cytoskeleton collapse and supported by the RAC1–WRC–Arp2/3 axis, its more complete molecular mechanism still awaits further investigation.

NADPH depletion

Regulation of disulfidptosis primarily involves alterations in key metabolites and signaling nodes. Among them, NADPH is a central determinant of disulfidptosis: a decline in NADPH reduces cellular reductive capacity, rendering cells more vulnerable to redox imbalance and disulfide stress, which promotes aberrant disulfide bond formation and eventually leads to disulfidptosis [27]. Under low-glucose conditions or upon inhibition of glucose transporters (GLUT) such as GLUT1/3, reduced glucose uptake suppresses the flux of the PPP—a major glucose-oxidative metabolic route and the primary source of intracellular NADPH—resulting in a marked decline in NADPH generation [28]. Accordingly, the GLUT1 inhibitor BAY-876 can lower intracellular NADPH levels and is frequently used to promote disulfidptosis in tumor cells [29]. In addition, glucose oxidase (GOx) catalyzes rapid consumption of glucose and oxygen to produce hydrogen peroxide (H₂O₂) and D-gluconic acid. Because H₂O₂ detoxification further consumes NADPH, GOx also exhibits strong potential to induce disulfidptosis [30].

Beyond the classical paradigm of glucose starvation, a more direct strategy would be to inhibit key enzymes in the oxidative PPP, particularly G6PD, thereby reducing NADPH supply at its source [3, 4]. Among the currently available approaches, G6PDi-1 has been identified as a cell-active G6PD inhibitor, polydatin can also suppress PPP activity by inhibiting G6PD, and DHEA and 6-AN have been used as earlier classical G6PD inhibitory tools [31–34]. Importantly, recent studies have begun to directly connect G6PD dysregulation with disulfidptosis. In tumor-infiltrating exhausted CD8⁺ T cells, LDHB was shown to interact with and restrict G6PD activity, causing NADPH depletion and consequently triggering disulfidptosis [35]. In hepatocellular carcinoma, RRx-001 was reported to inhibit G6PD, deplete NADPH, and trigger disulfidptosis [36]. By contrast, valproic acid increased G6PD transcription and activity, elevated NADPH levels, and reduced the susceptibility of HCC cells to glucose starvation-induced disulfidptosis, indicating that G6PD may function not only as a metabolic node upstream of NADPH production but also as a context-dependent regulator of disulfidptosis sensitivity [37]. Besides the direct inhibition of G6PD, G6PD functional impairment may also arise from G6PD deficiency itself. In the most severe cases, Class I variants can result in more than 90% loss of enzyme activity, and such impairment may predispose cells to NADPH insufficiency and disulfide stress, thereby potentially increasing susceptibility to disulfidptosis, although this remains to be directly validated in disulfidptosis models [38–40]. In addition to G6PD, cytosolic NAD⁺ kinase (NADK) and compensatory cytosolic NADPH-generating pathways such as ME1 and IDH1 may also be targeted to lower cytosolic NADPH [41–43]. Overall, direct targeting of NADPH-producing pathways may represent a potential glucose restriction-independent route to induce disulfidptosis. However, current evidence remains limited, and the molecular contexts and biological consequences of this mode of disulfidptosis induction require further investigation.

Cystine accumulation

Cystine accumulation is another fundamental driver of disulfidptosis [44, 45]. More importantly, in the context of disulfidptosis, cystine is not merely a disulfide substrate that passively accumulates, but rather a high-flux metabolic burden continuously driven by SLC7A11 [46]. Because SLC7A11 imports extracellular cystine in exchange for intracellular glutamate, and intracellular glutamate is generally abundant, cystine can be actively and continuously transported into SLC7A11-high cells [47]. Meanwhile, cystine itself has extremely low water solubility, and once inside the cell, it usually must be rapidly reduced to the more soluble cysteine in order to maintain intracellular homeostasis and support subsequent metabolic processes; this reduction process depends on NADPH as a source of reducing equivalents [48]. On this basis, the generated cysteine may also be exported to the extracellular space via passive transport pathways, where it can be re-oxidized to cystine and then taken up again by cells, thereby further amplifying this sustained reductive burden [49]. In contrast, although protein- or glutathione-associated disulfide pools also participate in intracellular redox homeostasis, they more often represent relatively buffered endogenous disulfide reservoirs, rather than high-flux substrates like cystine that are continuously imported by transporters and must be processed immediately [50, 51]. Therefore, under conditions of glucose deprivation or restricted PPP flux, the key factor responsible for disproportionate NADPH depletion lies in the continuous SLC7A11-mediated influx of cystine and its subsequent rapid reduction, rather than in other relatively stable disulfide reservoirs [52].

As the key transporter governing cystine uptake and metabolism, SLC7A11 directly shapes cellular sensitivity to this death program [53]. Specifically, activating transcription factor 4 (ATF4) and nuclear factor erythroid 2–related factor 2 (NRF2) are two major transcriptional activators of SLC7A11 [54]. ATF4 enhances SLC7A11 transcription by binding amino acid response elements (AAREs) within its promoter, whereas NRF2, released from Kelch-like ECH-associated protein 1 (KEAP1) during oxidative stress, translocates to the nucleus and binds antioxidant response elements (AREs) to induce SLC7A11 expression. In contrast, p53 can directly bind p53-responsive elements in the SLC7A11 promoter to repress its transcription, thereby reducing cellular susceptibility to disulfidptosis [10, 55]. Notably, although early studies considered high SLC7A11 expression a prerequisite for disulfidptosis, emerging evidence suggests that this cell death program is not strictly dependent on SLC7A11 abundance [56]. For example, in glioblastoma cells with low SLC7A11 expression, glucose deprivation combined with inhibition of thioredoxin reductase 1 (TXNRD1) can still induce disulfidptosis [57].

Moreover, cystine overload in disulfidptosis should not be viewed as an isolated metabolic event, but rather as part of a broader collapse of low-molecular-weight thiol/disulfide homeostasis under NADPH limitation. In SLC7A11-high cells exposed to glucose deprivation or oxidative stress, impaired cystine reduction is often accompanied by the accumulation of other small-molecule disulfides, including oxidized glutathione (GSSG), γ-glutamyl-cystine, and γ-glutathionyl-cysteine, together with GSH depletion and an increased GSSG/GSH ratio, thereby further promoting disulfide stress and sensitizing protein cysteine residues to aberrant disulfide modifications that contribute to cytoskeletal collapse and disulfidptosis [7, 50, 58]. Taken together, these findings indicate that cystine accumulation in disulfidptosis is not a simple consequence of transporter activity, but a metabolically amplified redox burden that couples SLC7A11-driven cystine influx to NADPH exhaustion, thiol/disulfide homeostasis collapse, and ultimately cellular susceptibility to disulfidptosis.

Detection methods of disulfidptosis

As a newly defined form of RCD triggered by intracellular disulfide stress, disulfidptosis requires a multidimensional assessment that integrates its distinctive biochemical basis with characteristic cellular morphology. With rapid progress in this field, a relatively systematic workflow for identifying disulfidptosis has begun to take shape.

Molecular-level features

1. Quantification of cystine and related small-molecule disulfides. Abnormal intracellular cystine accumulation is an early initiating signal for disulfidptosis. Its levels are commonly assessed using commercially available colorimetric/fluorometric kits, and are quantified by spectrophotometric absorbance (or fluorescence) readouts [59]. Although kit-based assays remain widely used for routine detection, targeted high-performance liquid chromatography (HPLC), ultra-performance liquid chromatography (UPLC), or liquid chromatography–tandem mass spectrometry (LC–MS/MS) is generally preferred when higher precision or broader profiling of disulfide stress is required, as these methods enable more accurate and simultaneous quantification of multiple thiol/disulfide species, including GSSG, γ-glutamyl-cystine, and γ-glutathionyl-cysteine [60, 61].

2. Assessment of NADP⁺/NADPH homeostasis. The NADP⁺/NADPH ratio can be assessed using enzymatic cycling assays, and a pronounced increase in this ratio reflects diminished reducing power in a manner compatible with disulfidptosis [62].

3. Evaluation of SLC7A11 expression. As the key subunit of the cystine/glutamate antiporter, SLC7A11 levels shape sensitivity to disulfidptosis and can be measured by qPCR or Western blotting to select susceptible models and interpret phenotypic differences [58].

4. Proteomic measurement of disulfide bonding in cytoskeletal proteins. Mass spectrometry–based redox or post-translational modification analyses can be applied to actin and other cytoskeletal proteins to map disulfide-associated modifications, offering a direct molecular readout of disulfide bond formation [9].

Morphological features

1. F-actin staining. Immunofluorescence staining with phalloidin enables visualization of F-actin architecture. During disulfidptosis, normally extended F-actin fibers undergo marked contraction and fragmentation, often forming condensed aggregates and detaching from the plasma membrane [63].

2. Plasma membrane integrity assays. Nucleic acid dyes such as propidium iodide (PI), combined with flow cytometry or high-content imaging, can be used to quantitatively assess terminal cell death and loss of membrane integrity [64].

Functional validation by perturbation

Under inductive conditions, co-treatment with disulfide-reducing agents—such as dithiothreitol, 2-mercaptoethanol, or tris(2-carboxyethyl)phosphine—should markedly rescue cell death if disulfidptosis is the primary mechanism. In contrast, inhibitors of other cell death programs are typically ineffective. Such a pattern provides strong evidence that the observed death phenotype is specifically triggered by disulfide stress [26].

Disulfidptosis in cancer research

Accumulating evidence suggests that disulfidptosis is broadly involved in tumor types with high cystine dependency and vulnerability to impaired NADPH regeneration, particularly in cancers exhibiting elevated SLC7A11 expression or pronounced glucose metabolic reprogramming [65–67]. Nevertheless, tumor cells can evade this RCD program through diverse adaptive regulatory pathways.

In pancreatic ductal adenocarcinoma (PDAC), the long non-coding RNA cancer susceptibility 8 (CASC8) suppresses glucose deprivation–induced disulfidptosis by interacting with the transcription factor c-MYC to enhance its protein stability, thereby activating the PPP, lowering the NADP⁺/NADPH ratio, and reinforcing cellular reductive capacity [68]. In another study, a prognostic model composed of six disulfidptosis-related long non-coding RNAs was developed and validated, in which TMEM105 was identified as a key determinant whose high expression predicts poor outcomes in pancreatic cancer. Mechanistic analyses suggested that TMEM105 stabilizes β-catenin through a proteasome-dependent mechanism, which in turn activates the glycolysis-associated transcription factor c-MYC and enhances transcriptional activity of the glucose transporter GLUT1. These findings imply that targeting the TMEM105–β-catenin–c-MYC–GLUT1 axis may represent a potential therapeutic strategy to promote disulfidptosis [69].

Beyond pancreatic cancer, ovarian cancer can increase therapeutic resistance by counteracting disulfidptosis, particularly in ovarian clear cell carcinoma (OCCC), which frequently exhibits pronounced platinum resistance accompanied by glycogen accumulation and wild-type p53 status [70]. In addition, KEAP1-mutant lung cancers often display high SLC7A11 expression because loss of the negative regulator KEAP1 stabilizes nuclear factor NRF2, thereby upregulating downstream SLC7A11 [71]. In glioma, inhibition of TXNRD1 can effectively induce disulfidptosis in glioblastoma (GBM) cells, accompanied by accumulation of cytosolic peroxisomes and macropinocytic structures as well as the emergence of microfissures in the plasma membrane [57]. These structural alterations may facilitate leakage of intracellular contents and subsequent immune activation, providing a potential morphological basis for the immunogenic features of this death modality.

In summary, disulfidptosis can be induced across multiple cancer types through the combined pressure of cystine accumulation and NADPH depletion, leading to direct tumor cell killing. Harnessing this metabolic vulnerability therefore holds promise for improving antitumor efficacy.

Metabolic synergy in disulfidptosis research

Classical RCD pathways have been harnessed in oncology to eliminate tumor cells, yet resistance mechanisms emerge continuously [2, 72]. To enhance tumor clearance, researchers have begun exploring strategies that combine classical RCD with disulfidptosis [73]. Disulfidptosis shares key metabolic components with conventional RCD, such as SLC7A11, GSH, and NADPH, forming a “metabolic synergy network” [74, 75]. This synergy can not only overcome resistance to single RCD modalities but also amplify antitumor effects through multi-pathway activation, offering a novel strategy for treating tumors with metabolic vulnerabilities.

Metabolic synergy with ferroptosis

Ferroptosis is a lipid peroxidation–driven form of RCD, primarily governed by the SLC7A11–GSH–GPX4 axis [76]. To resist ferroptosis, tumor cells frequently upregulate SLC7A11 to enhance cystine uptake and promote GSH synthesis, which helps detoxify lipid peroxides [77]. Paradoxically, this adaptive mechanism also establishes a permissive context for activating disulfidptosis [28]. Under glucose-replete conditions, high SLC7A11 expression suppresses ferroptosis via the GSH–GPX4 axis, whereas under glucose deprivation, sustained SLC7A11-mediated cystine import drives ongoing NADPH consumption, leading to NADPH/NADP⁺ disequilibrium, disulfide stress, and consequent disulfidptosis [78].

In breast cancer, researchers engineered an Fe–Cu–SS metal–organic framework (FCSP@876 MOFs) with the glucose transporter inhibitor BAY-876 loaded into its mesoporous structure [29]. Upon delivery, the released Fe/Cu ions promote lipid peroxidation to induce ferroptosis. Simultaneously, BAY-876 inhibits glucose uptake and depletes NADPH. This not only compromises SLC7A11-dependent GSH synthesis—thereby overcoming ferroptosis resistance—but also promotes cystine accumulation, triggering disulfidptosis. The combined action of these two distinct death pathways significantly enhances tumor cell elimination [29].

While in an ovarian cancer model, a FeOOH@Fe-Ap@Au nano-shuttle was developed to achieve synergy through systemic glucose deprivation [79]. Specifically, gold nanodots exhibit GOx-like activity and directly consume glucose, whereas apigenin (Ap) suppresses glucose uptake by downregulating GLUT1 (Fig. 2). Together, these actions disrupt glucose supply and drive NADPH depletion. Meanwhile, exogenous iron delivered by FeOOH cooperates with H₂O₂ generated by the GOx-like reaction to enhance Fenton chemistry and intensify ferroptosis. Ultimately, the concurrent induction of disulfidptosis and ferroptosis enables FeOOH@Fe-Ap@Au nano-shuttles to achieve robust antitumor efficacy in ovarian cancer models [79].

Fig. 2. Representative therapeutic strategies and signaling crosstalk centered on disulfidptosis.

A Schematic illustration of FeOOH@Fe-Ap@Au nanoshuttles (NSs) and their dual induction of disulfidptosis and ferroptosis via systemic glucose deprivation. Glucose restriction reduces PPP-derived NADPH and compromises the SLC7A11/GSH/GPX4 antioxidant axis, favoring disulfide stress and disulfidptosis. In parallel, Fe-driven oxidative chemistry enhances lipid peroxidation (LPO) to trigger ferroptosis. B Design and mechanism of Cu₂O@GOx@HA nanoactivators for enhanced cuproptosis coupled with disulfidptosis. After tumor accumulation, GOx catalyzes glucose oxidation to generate gluconic acid and impose glucose starvation. On the “spear” side, gluconic acid facilitates Cu₂O-mediated Cu⁺ release, thereby promoting cuproptosis signaling in the tumor microenvironment. On the “shield removal” side, glucose depletion reduces PPP flux and depletes NADPH, suppressing cystine-to-cysteine reduction, triggering disulfidptosis, and consequently limiting GSH biosynthesis. In parallel, GOx-driven metabolic stress together with Cu⁺/Cu²⁺ disrupts redox homeostasis, inhibits glycolysis and mitochondrial metabolism, decreases ATP production, and downregulates ATP-dependent copper efflux transporters (ATP7A/B). Enhanced intracellular copper then engages the canonical cuproptosis axis (FDX1–LIAS/DLAT), increasing protein lipoylation-dependent copper binding, inducing DLAT aggregation, perturbing the PDH complex/TCA cycle, and culminating in cuproptotic cell death. C Pro@FLNC nanoparticles (NPs) for combinatorial cancer therapy through synergistic disulfidptosis and pyroptosis with potentiated photodynamic therapy (PDT). Within the tumor cell, cystine influx via SLC7A11 and redox imbalance induce disulfide stress, accompanied by GSH depletion and elevated ROS, leading to actin collapse and disulfidptosis. Upon 670-nm laser irradiation, the photosensitizer Ce6 triggers PDT, generating ROS, promoting LPO and mitochondrial dysfunction, and further amplifying intracellular oxidative/disulfide stress. This stress cooperatively activates caspase-3 and induces GSDMD-associated pore formation, driving pyroptosis. The combined disulfidptosis/pyroptosis program enhances the release of tumor antigens and immunogenic signals, promotes DC maturation, and elicits systemic antitumor immunity characterized by increased CD8⁺ T responses. D Synergistic interplay between disulfidptosis and ER stress under glucose restriction and therapeutic exploitation. In SLC7A11-high cells, reduced GLUT/PPP flux limits NADPH, facilitating disulfide formation and triggering disulfidptosis while simultaneously perturbing ER proteostasis and activating the PERK–eIF2α–ATF4 arm of the unfolded protein response (UPR), with induction of CHOP/ATF3. Pharmacologic inhibition of the ER stress response by the PERK inhibitors GSK2656157 (G157) and GSK2606414 (G414), as well as the eIF2α dephosphorylation inhibitor Salubrinal (Salu), cooperates with GLUT inhibition to intensify disulfide stress and enhance disulfidptosis in vitro and in vivo. In addition, p38 signaling is depicted as a modulatory branch associated with disulfide formation/disulfidptosis, and its inhibition by SB-203580 is indicated. Together, these findings support a combination strategy that leverages ER stress modulation to potentiate glucose-restriction–induced disulfidptotic vulnerability. Abbreviations: ATF3/ATF4 activating transcription factor 3/4, ATP7A/B copper-transporting ATPase 7A/7B, Ce6, chlorin e6, CHOP C/EBP homologous protein, Cu2O@GOx@HA cuprous oxide–glucose oxidase–hyaluronic acid nanoactivator, DCs dendritic cells, DLAT dihydrolipoamide S-acetyltransferase, ER endoplasmic reticulum, FDX1 ferredoxin 1, FeOOH@Fe-Ap@Au NSs ferric oxyhydroxide–(Fe-Ap)–gold nanosheets, GOx glucose oxidase, GPX4 glutathione peroxidase 4, GSDMD gasdermin D, GSH/GSSG reduced/oxidized glutathione, HA hyaluronic acid, LA lipoic acid, LIAS lipoic acid synthetase, LPO lipid peroxidation, NPs nanoparticles, NSs nanoshuttles, p38 p38 mitogen-activated protein kinase, PDT photodynamic therapy, PERK PKR-like ER kinase, POD peroxidase, ROS reactive oxygen species, SB-203580 a selective p38 inhibitor, SOD superoxide dismutase, TCA cycle tricarboxylic acid cycle, TME tumor microenvironment, UA ursolic acid, UPR unfolded protein response, XOD xanthine oxidase. This figure was created with BioRender (https://biorender.com/).

Metabolic synergy with cuproptosis

Cuproptosis is a recently described form of copper-dependent RCD. Mechanistically, excess copper accumulates and traffics into mitochondria, where it directly targets key proteins involved in energy metabolism—particularly lipoylated enzymes—thereby inducing aberrant protein aggregation and loss of function, disrupting the mitochondrial respiratory chain, and ultimately triggering lethal proteotoxic stress [80]. Reduced GSH, as an important intracellular copper chelator, can buffer copper accumulation and thereby negatively regulate cuproptosis [81]. Consequently, tumors may resist cuproptosis by upregulating SLC7A11 and depleting GSH, but this adaptation inadvertently sensitizes them to disulfidptosis.

Building on the concept that disulfidptosis can complement strategies aimed at overcoming cuproptosis resistance, researchers developed a copper-based metal–organic framework nanoparticle system, CuSS@876-PEG, designed to exploit these metabolic liabilities [17]. In this platform, GSH depletion triggers nanoparticle disassembly, enabling the simultaneous release of copper ions and the glucose transporter inhibitor BAY-876, thereby co-inducing cuproptosis and disulfidptosis. Notably, this combination not only markedly enhances tumor cell killing but also elicits immunogenic cell death (ICD), activating antitumor immunity by promoting dendritic cell (DC) maturation and increasing CD8⁺ T-cell infiltration, thereby strengthening the overall immune response [17].

Moreover, because cuproptosis can be inefficiently activated and cancer cells often develop resistance through GSH buffering and copper efflux transporters, an additional “spear-and-shield” strategy has been proposed to potentiate cuproptosis via disulfidptosis [82]. In this approach, a Cu₂O-based nanoactivator (Cu₂O@GOx@HA) is loaded with GOx (Fig. 2). On the “spear” side, GOx catalyzes glucose oxidation to generate gluconic acid, which facilitates Cu₂O-mediated release of Cu⁺ to enhance cuproptosis. On the “shield removal” side, GOx simultaneously consumes glucose, driving NADPH depletion, suppressing the reduction of cystine to cysteine, inducing disulfidptosis, and thereby limiting GSH biosynthesis. In parallel, GOx and Cu⁺ cooperatively inhibit glycolysis and mitochondrial metabolism, reduce ATP production, and downregulate ATPase expression. As a result, disulfidptosis acts as a powerful amplifier that significantly strengthens cuproptosis efficacy, providing a new framework for coordinating positive activation and negative resistance circuits in cuproptosis-based cancer therapy [82].

Metabolic synergy with pyroptosis

Pyroptosis is a highly immunogenic form of RCD, characterized morphologically by plasma membrane pore formation, chromatin condensation, and pyroptotic body generation [83]. Its core mechanism involves caspase-family proteases cleaving gasdermin (GSDM) proteins to release pore-forming N-terminal fragments, which perforate the membrane and enable the release of pro-inflammatory mediators, thereby stimulating antitumor immunity [84]. The key basis for synergy between disulfidptosis and pyroptosis is a “cascade amplification of oxidative stress.” First, glucose deprivation and cystine accumulation drive depletion of NADPH and GSH, initiating disulfidptosis while markedly elevating reactive oxygen species (ROS). Next, excessive ROS can activate pyroptosis-related caspase pathways; these proteases further cleave GSDMD, allowing its N-terminal domain to form membrane pores, leading to cell swelling and lytic rupture typical of pyroptosis. Finally, DAMPs and pro-inflammatory factors released from pyroptotic cells reinforce antitumor immune responses [18].

Multiple nanoplatforms have now been developed to integrate disulfidptosis- and pyroptosis-inducing modules, achieving a dual therapeutic outcome of “direct tumor killing plus immune activation.” For example, the Pro@FLNC prodrug nanoregulator designed for advanced breast cancer combines folate receptor–mediated targeting with GSH-responsive drug release, enabling tumor-selective activation [85]. Within tumors, Pro@FLNC consumes intracellular GSH and disrupts redox homeostasis, thereby inducing oxidative stress, mitochondrial dysfunction, and disulfidptosis (Fig. 2). The elevated ROS levels and metabolic imbalance further trigger immunogenic pyroptosis, releasing DAMPs, promoting DC maturation and CD8⁺ T-cell recruitment, and reducing immunosuppressive cell infiltration [85].

In a colorectal cancer model, Pd₂Sn@GOx-SP nanorods achieve synergy through cascade catalysis [23]. In this therapeutic system, GOx promotes disulfidptosis by consuming glucose while generating H₂O₂, whereas Pd₂Sn further amplifies an ROS burst through peroxidase-like (POD-like) and oxidase-like (OXD-like) activities, culminating in pyroptosis. Moreover, glucose starvation increases the NADP⁺/NADPH ratio, promotes intracellular cystine accumulation, and suppresses GSH synthesis, thereby reinforcing disulfidptosis and further amplifying ROS-driven pyroptotic signaling. Through the coordinated induction of pyroptosis and disulfidptosis, Pd₂Sn@GOx-SP can effectively remodel the TME, alleviate immunosuppression, enhance T-cell infiltration, and suppress tumor metastasis and recurrence [23].

Metabolic synergy with endoplasmic reticulum stress

Endoplasmic reticulum (ER) stress is a protective cellular response to aberrant protein folding [86]. By upregulating antioxidant and proteostasis-related programs, ER stress can mitigate disulfide stress and thereby suppress disulfidptosis [87]. Accordingly, pharmacological inhibition of ER stress may sensitize tumor cells to disulfidptosis and provides a rationale for combination therapy. Under glucose-deprived conditions, ER stress inhibitors have been reported to promote disulfidptosis, whereas ER stress inducers can attenuate glucose deprivation–triggered disulfidptosis in SLC7A11-high cells, supporting a protective role of ER stress in this context [88]. Building on this observation, researchers developed a dual-targeting strategy: simultaneous blockade of glucose metabolism and ER stress to amplify disulfidptosis efficacy (Fig. 2). In vitro, co-treatment with the GLUT inhibitor BAY-876 and the ER stress inhibitor G157 significantly increased the proportion of cells undergoing disulfidptosis compared with BAY-876 alone. In subcutaneous xenograft models, this combination markedly suppressed tumor growth. These findings provide conceptual support for ER stress–targeted agents as sensitizers that potentiate disulfidptosis-based combination regimens in cancer therapy [88].

Taken together, the metabolic synergy between disulfidptosis and ferroptosis, cuproptosis, pyroptosis, as well as ER stress–related pathways, can be viewed as a regulatory network organized around the “cystine–GSH–NADPH” metabolic axis. Triggering disulfidptosis by targeting glucose metabolism may simultaneously increase tumor susceptibility to canonical RCD programs and help overcome resistance mechanisms. To date, the efficacy of such synergistic strategies has been demonstrated in multiple models, including breast, ovarian, and lung cancers. Looking forward, deeper mechanistic dissection of disulfidptosis regulators and the development of selective targeted agents may enable disulfidptosis-centered combination therapy to emerge as an important direction for treating metabolically vulnerable tumors.

Immuno-regulatory roles and immunotherapy-related value of disulfidptosis in cancer

As a newly recognized disulfide stress-driven form of RCD, disulfidptosis not only eliminates tumor cells via metabolic vulnerabilities, but also profoundly shapes antitumor immunity by remodeling the TME, modulating immune-cell function, and synergizing with immunotherapies. Mechanistically, the concept centers on an axis linking disulfidptosis-associated metabolic pathways to immune-cell fate decisions and subsequent immunogenic activation. In addition, disulfidptosis-related genes (DRGs) show promise as biomarkers for predicting immunotherapy response and prognosis, offering new avenues for precision immuno-oncology [23, 89]. Therefore, this section will systematically review the immunological aspects of disulfidptosis research from two complementary perspectives: immunoregulatory mechanisms and predictive value.

Core immunoregulation by disulfidptosis: from immune suppression to immune activation

Shaping CD8⁺ T-cell fate

By directly impairing immune-cell fitness, disulfidptosis can further reinforce an immunosuppressive TME. Emerging evidence indicates that glucose limitation and chronic antigen stimulation within the TME are key drivers of CD8⁺ T-cell exhaustion, and that disulfidptosis constitutes a central metabolic mechanism underlying this process (Table 1). Zou et al. first demonstrated that tumor-infiltrating CD8⁺ T cells themselves can undergo disulfidptosis, and delineated its pivotal role in promoting T-cell exhaustion [35]. Under chronic antigen stimulation, the transcription factor STAT3 upregulates lactate dehydrogenase B (LDHB), which directly interacts with glucose-6-phosphate dehydrogenase (G6PD) to inhibit its dimerization and enzymatic activity. This suppression of G6PD impairs PPP flux, resulting in NADPH depletion, accumulation of aberrant disulfide bonds, and ultimately CD8⁺ T-cell functional exhaustion through disulfidptosis. Notably, the study further dissected the interplay between disulfidptosis and another programmed cell-death modality, ferroptosis, in determining T-cell fate. Heterozygous loss of SLC7A11 selectively inhibits disulfidptosis without affecting ferroptosis, resulting in a net reduction of T-cell death and exhaustion, but homozygous loss of SLC7A11 concurrently promotes ferroptosis while suppressing disulfidptosis, yielding offsetting effects and an overall unchanged death rate. Moreover, heterozygous loss of G6PD enhances both disulfidptosis and ferroptosis, ultimately exacerbating T-cell exhaustion [35]. Therefore, this work establishes T cell–intrinsic disulfidptosis as a key metabolic driver of CD8⁺ T-cell exhaustion, mechanistically linking disulfidptosis to tumor immune tolerance and highlighting LDHB targeting as a potentially promising strategy for cancer immunotherapy.

Table 1.

Disulfidptosis in cancer immunity.

Inducer	Immunity	Cell death	Immunological/biological effect	Refs.
LDHB Inhibition	Suppression	Disulfidptosis	Tumor-infiltrating CD8⁺ T cells themselves undergo disulfidptosis, playing a key role in T cell exhaustion.	[35]
CYBC	Promotion	Disulfidptosis	Triggers immunogenic cell death and sustains T cell activation, thereby alleviating immune exhaustion.	[91]
FBSP MOFs	Promotion	Disulfidptosis + ferroptosis	Promotes immunogenic cell death in tumor cells, releases tumor-associated antigens, and enhances CD8⁺ T cell infiltration.	[59]
GOx-IA@HMON@IO	Promotion	Disulfidptosis + ferroptosis	Relieves tumor hypoxia, drives M2 macrophages toward M1 polarization, suppresses regulatory T cells, and activates CD8⁺ T cells.	[30]
CuSS@876-PEG	Promotion	Disulfidptosis + cuproptosis	Induces immunogenic cell death, promoting DC activation and CD8⁺ T cell infiltration.	[17]
CST NPs	Promotion	Disulfidptosis + cuproptosis	Disrupts tumor cell membranes, exposing and releasing tumor-associated antigens and DAMPs.	[22]
Pro@FLNC	Promotion	Disulfidptosis + pyroptosis	Tumor cell destruction releases DAMPs, promotes DC maturation and CD8⁺ T cell recruitment, and reduces infiltration of immunosuppressive cells.	[85]
Pd₂Sn@GOx-SP	Promotion	Disulfidptosis + pyroptosis	Remodels the tumor microenvironment, alleviates immunosuppression, and promotes T cell infiltration.	[23]
CGBH NNs	Promotion	Disulfidptosis + pyroptosis	Pro-inflammatory factors released from pyroptotic cells further amplify inflammation, promoting DC infiltration and CD8⁺ T cell recruitment.	[90]

Inducing immunogenic cell death

Disulfidptosis, characterized by disulfide stress and F-actin cytoskeletal collapse, can drive tumor cells toward ICD. This process promotes the release of DAMPs, thereby recruiting and activating innate immune cells such as DCs and functionally bridging innate and adaptive antitumor immunity [59].

To efficiently induce disulfidptosis in tumors, multiple nanomedicine platforms have been developed (Table 1). For example, the copper-based nanoinducer CGBH NNs exhibits both GOx-like and POD-like activities [90]. It simultaneously triggers disulfidptosis by consuming glucose and suppressing NADPH generation, while elevating ROS to potentiate pyroptosis. Pro-inflammatory cytokines released from pyroptotic cells further amplify local inflammation, enhance DC infiltration, and promote CD8⁺ T-cell recruitment, thus reversing immunosuppression within the TME [90].

In addition, a multifunctional nanoinducer termed CYBC has been designed to co-deliver exogenous cystine and a GLUT1 inhibitor via a biomimetic nanoplatform [91]. This approach overcomes the intrinsic limitation of low cystine availability in the TME and maximizes disulfidptosis induction. Importantly, the study suggests that disulfidptosis itself can act as an immunogenic intervention by initiating immunogenic cell death (ICD) and promoting immunometabolic reprogramming, which sustains T-cell activation, alleviates immune exhaustion, and effectively transforms immunologically “cold” tumors into “hot” ones [91].

Immuno-predictive value of disulfidptosis-related biomarkers

Prognostic prediction in cancer

DRGs and derived scoring systems have been proposed as biomarkers to predict patient prognosis, immunotherapy responsiveness, and drug sensitivity, offering valuable support for precision oncology [92]. Across multiple tumor types, DRG-based risk models show independent prognostic value, and higher scores are generally associated with more aggressive phenotypes and worse survival (Table 2). In gastric cancer, a high DRG score is associated with increased infiltration of plasmacytoid dendritic cells (pDCs), which may contribute to an immunosuppressive TME and shorter overall survival. Additionally, patients with higher scores are predicted to be more sensitive to chemotherapeutic agents such as paclitaxel [93]. In ovarian cancer, a higher DRG score suggests greater susceptibility to disulfidptosis under glucose deprivation and correlates with poorer prognosis and a more invasive phenotype [94]. In glioma, DRG-based risk scores positively correlate with tumor invasiveness, an immunosuppressive state, and unfavorable outcomes [95]. Collectively, disulfidptosis-based risk stratification not only enables prognostic classification but also helps delineate immune microenvironment features.

Table 2.

Disulfidptosis-related genes in cancer treatment prediction.

Cancer type	Model	Application/role	Signature genes	Refs.
Gastric cancer	mRNA	Immune status and prognosis prediction	FLNA, TLN1, PRDX1, MYH9, FLNB, ACTB, SLC7A11, SLC3A2, RPN1, NCKAP1, NUBPL, NDUFA11, LRPPRC, OXSM, NDUFS1, GYS1	[93]
Ovarian cancer	mRNA	Tumor invasiveness and prognosis prediction	ACTN4, ACTB, FLNA, FLNB, INF2	[94]
Glioma	mRNA	Tumor invasiveness, immune status, and prognosis prediction	GYS1, LRPPRC, NCKAP1, NDUFA11, NDUFS1, NUBPL, OXSM, RPN1, SLC3A2, SLC7A11	[95]
Endometrial cancer	mRNA	Immune status, prognosis, and treatment-strategy prediction	CDKN2A, FZD7, LCN2, ACTN4, MYH10	[96]
Acute myeloid leukemia	lncRNA	Chemotherapy and immunotherapy strategy prediction	FAM30A, MRPL20-AS1, AC022182.2, AL391704.1, LINC01539, AP000439.1, AC113423.1	[97]
Colorectal cancer	mRNA	Immune status and immunotherapy strategy prediction	FLNA, FLNB, OXSM, SLC7A11	[98]

Prediction of treatment response in cancer

Beyond prognostication, DRG-related models may also guide treatment selection and facilitate individualized therapeutic planning. In addition to the findings reported in gastric cancer, ovarian cancer, and glioma, treatment-response predictions have been described in several other malignancies (Table 2).

In endometrial cancer, a model based on five disulfidptosis/ferroptosis-related genes (DFRGs) indicates that low-risk patients may be better candidates for immunotherapy [96]. In acute myeloid leukemia, a model constructed from seven disulfidptosis-related long non-coding RNAs shows that the high-risk group exhibits more prominent immune-cell infiltration and enrichment of multiple immune-related signaling pathways. Moreover, drug sensitivity differs across risk strata, highlighting potential utility for individualized treatment guidance [97]. In colorectal cancer, patients with high DRG scores have poorer outcomes, characterized by enhanced activity of immunosuppressive cells and reduced infiltration of cytotoxic immune cells. Notably, combining a GLUT1 inhibitor with a programmed cell death protein 1 (PD-1) inhibitor can markedly enhance CD8⁺ T-cell recruitment and suppress epithelial–mesenchymal transition (EMT) [98].

Overall, disulfidptosis participates in tumor immunoregulation mainly through two mechanisms: modulating CD8⁺ T-cell exhaustion and triggering ICD. Meanwhile, DRG–based models provide new tools for predicting immune prognosis and treatment response, with particular advantages in metabolically vulnerable tumors characterized by high SLC7A11 expression. In the future, key priorities will be to delineate context-dependent determinants, to standardize and functionally validate DRG panels, and to evaluate rational combination regimens that integrate metabolic intervention with immunotherapy in well-designed multicenter prospective studies.

Disulfidptosis and cancer therapy

Disulfidptosis relies on a metabolic cascade of “glucose deprivation–NADPH depletion–cystine accumulation–disulfide stress” and therefore offers a selective vulnerability for metabolically fragile tumors with high SLC7A11 expression [37]. At present, the main translational focus centers on two complementary features: first, efficiently inducing disulfidptosis by disrupting tumor metabolic homeostasis through enzyme-enabled cooperation, energetic interference and related strategies; second, optimizing delivery systems to improve tumor targeting and therapeutic efficacy. Accordingly, we summarize key progress from both induction strategies and innovations in delivery platforms.

Key strategies for inducing disulfidptosis

Depleting NADPH via natural glucose oxidase–based synergy

A central requirement for triggering disulfidptosis is to impose glucose starvation, interrupt NADPH supply, and promote cystine accumulation. Current mainstream approaches often achieve this through coordinated “enzymatic catalysis plus metabolic interference,” with representative designs as follows. One strategy combines natural enzymes with nanozymes to simultaneously suppress NADPH regeneration and accelerate NADPH consumption. For example, the GOx/CuSAE hybrid system integrates a copper single-atom nanozyme (CuSAE) with natural GOx [20]. GOx catalyzes glucose oxidation to generate gluconic acid and H₂O₂, thereby driving glucose starvation in tumor cells and effectively dampening the PPP, a major source of NADPH regeneration. Meanwhile, CuSAE exhibits NADPH oxidase–like activity and further consumes the residual intracellular NADPH pool. Together, this hybrid platform imposes a “dual hit” on NADPH availability, markedly limiting disulfide-bond reduction capacity and thereby efficiently initiating disulfidptosis in tumor cells [20].

In addition, a comprehensive strategy has been proposed in breast cancer to co-induce multiple cell-death programs within a single platform. Based on this concept, an intelligent nanomedicine built on hollow mesoporous manganese oxide (HMON)—GOx-IA@HMON@IO—was developed to couple GOx-mediated glucose consumption with iron-driven ROS generation and metabolic interference [30]. This integrated design supports a triple synergy involving ferroptosis, disulfidptosis, and reversal of the immunosuppressive TME. Notably, under GSH-responsive release, the system also enables T2-to-T1 magnetic resonance imaging (MRI) signal switching, facilitating image-guided precision therapy for triple-negative breast cancer [30].

GLUT inhibitors synergistically drive multiple cell-death programs

Beyond exploiting GOx-mediated glucose consumption, GLUT inhibitors are also frequently used to block glucose entry into cells, thereby effectively restricting glucose supply and inducing NADPH depletion. For example, researchers developed GSH-responsive CuSS@876-PEG nanoparticles that dissociate in the TME to release Cu²⁺ and BAY-876 [17]. These two components respectively induce cuproptosis and inhibit glucose uptake, such that the compensatory upregulation of SLC7A11 in cancer cells precipitates disulfidptosis. Similarly, when the GLUT inhibitor BAY-876 is incorporated into the mesoporous structure of Fe–Cu–SS MOFs, the released BAY-876 suppresses glucose availability and causes NADPH depletion [29]. It blocks GSH resynthesis to enhance Fe/Cu-driven ferroptosis, while excess cystine accumulation induces disulfide stress and disulfidptosis.

Novel delivery systems for disulfidptosis inducers

Actively targeted nanocarriers

Nanodelivery systems are a major solution to the poor tumor selectivity and limited stability of disulfidptosis inducers [99, 100]. To achieve active tumor targeting, researchers engineered an oxygen-propelled nanomotor (GOx@HPB) using hollow mesoporous Prussian blue (HPB) as the carrier, loading GOx and decorating the surface with targeting ligands [101]. This self-propelled behavior improves tumor penetration, while glucose depletion coupled with NADPH exhaustion promotes disulfidptosis, yielding markedly greater efficacy than conventional carriers. To improve therapeutic precision and reduce off-target toxicity, microenvironment-responsive nanoplatforms have been developed to enable tumor-selective activation. For example, ursolic acid (UA) and chlorin e6 (Ce6) were conjugated via a GSH-responsive linker to form a functional core, which was subsequently encapsulated within a DSPE–PEG–FA (folate) lipid shell to yield Pro@FLNC with favorable physicochemical stability and GSH-triggered drug release in tumors [85]. Pro@FLNC depletes intracellular GSH and simultaneously initiates photodynamic therapy–driven ROS production, which disrupts redox homeostasis and triggers a cascade of synergistic effects, including disulfidptosis, pyroptosis, and ICD [85]. Beyond this example, FBSP/GOx@FBSP MOFs can also be activated by elevated GSH in the tumor microenvironment, inducing disulfidptosis and promoting tumor cell elimination [59].

Integration with multimodal theranostics

Combining nanoplatform-based induction with other diagnostic and therapeutic modalities can further enhance precision killing and enable real-time monitoring [102]. An ultrasound-activated nanosystem SPCP/CCP@Bay, consisting of a sonodynamic polymer (SPCP), a cystine-containing polymer (CCP), and the GLUT inhibitor BAY-876, exemplifies this strategy [21]. Under ultrasound irradiation, SPCP degrades to generate ROS, which promotes CCP-mediated cystine release and concomitantly enables BAY-876 release to block glucose uptake. Meanwhile, ROS can provoke ER stress and jointly enhance ICD. Moreover, when combined with anti–PD-1 antibodies, it markedly enhances immunotherapeutic efficacy [21]. In addition, the rationally designed hybrid nanozyme system M@GOx/Fe-HMON features a GSH-responsive HMON co-loading GOx and an Fe²⁺/Fe³⁺ redox pair, and is cloaked with homologous hepatocellular carcinoma membranes for tumor-specific targeting and immune evasion [103]. Importantly, its superparamagnetic property also enables MRI for real-time monitoring, establishing a multifunctional theranostic platform.

Taken together, current studies have highlighted the therapeutic potential of disulfidptosis in SLC7A11-high tumors. Strategies such as GOx-assisted NADPH exhaustion and GLUT inhibitor–based multi-death combinations can efficiently disrupt metabolic homeostasis to achieve robust and controllable induction of disulfidptosis. Concurrently, actively targeted nanodelivery systems with multimodal theranostic capabilities overcome limitations in stability and tumor specificity while enabling integrated theranostic workflows and synergistic precision therapy coupled with immune activation.

Discussion

First described in 2023, disulfidptosis is a newly recognized, disulfide stress–driven form of RCD. Mechanistically, glucose starvation in cells with high SLC7A11 expression depletes NADPH, leading to cystine accumulation and aberrant disulfide crosslinking within F-actin, which culminates in cytoskeletal collapse [104]. Building on this mechanistic framework, recent advances have largely converged on three interconnected themes—metabolic synergy, immune regulation, and therapeutic translation—within which several notable breakthroughs have been reported [105–107]. First, by targeting the “cystine–GSH–NADPH” metabolic axis, disulfidptosis can form a cooperative network with other regulated cell-death modalities to overcome resistance to conventional therapies. Second, beyond its direct cytotoxicity toward tumor cells, disulfidptosis can reshape antitumor immunity by modulating immune-cell function and triggering ICD. Third, a growing body of work is focusing on efficient induction and precise delivery, giving rise to diverse emerging therapeutic strategies.

Although a canonical mechanistic framework of disulfidptosis has been preliminarily established, its core definition and execution logic are still not fully understood, particularly with regard to whether the same triggering and regulatory pattern operates in tumor cells and immune cells [108, 109]. For example, high SLC7A11 expression has generally been regarded as an important prerequisite for disulfidptosis; however, emerging evidence suggests that disulfidptosis is not necessarily dependent on high SLC7A11 expression, indicating that its precise role in this process has yet to be fully clarified. Another unresolved issue concerns whether the lethal effect of NADPH depletion arises solely from the failure to promptly reduce cystine and other small-molecule disulfides, or whether it also involves a broader collapse of thiol/disulfide homeostasis and metabolic reprogramming. Furthermore, F-actin has also been recognized as the most representative execution target of disulfidptosis, yet why cytoskeletal proteins, rather than other thiol-containing molecules, become the decisive sites of damage is still unclear. Finally, although the RAC1–WRC–Arp2/3 axis is considered to play a central role in the execution of disulfidptosis, whether additional parallel regulatory or execution pathways also exist remains to be determined.

Current strategies for inducing disulfidptosis still mainly rely on glucose restriction and NADPH depletion to trigger cell death [110]. Although these approaches can effectively amplify disulfide stress, they may also disrupt glucose metabolism and redox homeostasis in normal tissues, thereby raising potential safety concerns. Therefore, future studies should further explore whether more direct disulfidptosis inducers can be developed by targeting actin vulnerability, WRC dependency, or thiol/disulfide homeostasis, rather than continuing to rely primarily on the relatively indirect strategy of “glucose deprivation/NADPH depletion.” On the other hand, although multiple nanodelivery platforms have been developed to improve the efficiency of disulfidptosis induction, enhance tumor accumulation, and strengthen synergistic therapeutic effects, how to further reduce systemic exposure, increase tumor-selective killing, and minimize additional damage to normal tissues and effector immune cells remains a key challenge in current applications [111]. More importantly, despite the increasing sophistication of nanoplatform design, these systems still face practical bottlenecks, including biocompatibility, degradability, targeting efficiency, and manufacturability [112]. Thus, delivery system optimization should not focus solely on increasing platform complexity, but should instead place greater emphasis on controllability, stability, and clinical translatability.

Despite the rapid growth of DRG- and lncRNA-based risk models in recent years and their demonstrated value in prognostic evaluation, immune-state characterization, and treatment-response prediction, existing studies still rely largely on public database analyses, with relatively limited prospective clinical validation [113]. Moreover, their application remains largely confined to correlation-based descriptions, and has not yet truly evolved into a clinical tool that can directly guide therapeutic decision-making. In other words, current disulfidptosis-related signatures are better at answering questions such as “What is the patient’s prognosis?” or “How should risk be stratified?” than at directly identifying which patients are most suitable for disulfidptosis-based therapy [114]. Accordingly, efforts should focus on establishing a biomarker system that can better support patient selection and precision treatment. Building on existing DRG scores, such a system should further integrate key factors such as SLC7A11 status, glucose metabolic features, PPP/NADPH metabolic status, and immune phenotypes, thereby improving its clinical applicability and translational value.

In summary, while disulfidptosis has revealed substantial therapeutic promise in metabolically vulnerable cancers, fragmented mechanistic insights, an underdeveloped biomarker framework, limitations in induction strategies, and bottlenecks in delivery technologies continue to impede clinical translation. Moving forward, progress will depend on a coherent pipeline centered on mechanism elucidation, biomarker development, strategy optimization, delivery innovation, and rigorous clinical validation. Through multidisciplinary integration, overcoming these core barriers may help propel this newly defined cell-death program from bench to bedside and provide a breakthrough avenue for treating therapy-resistant tumors.

Author contributions

ZW, RT, and JX collected the related studies and drafted the manuscript. ZW and RT generated the figures and tables. JX and YW participated in the design of the review. LJ, YL, DL, CZ, and CY performed detailed analyses and contributed to data interpretation. XY and SS initiated the study, supervised the review project, and revised the manuscript. All authors have read and agreed to the published version of the article.

Funding

This project was financially supported by the Noncommunicable Chronic Diseases-National Science and Technology Major Project [grant number 2025ZD0552305], the National Natural Science Foundation of China [grant numbers 82541017 and 82473106], the Science and Technology Commission of Shanghai Municipality [grant number YDZX20243100002003] and Shuguang Program of Shanghai Municipal Education Commission and Shanghai Education Development Foundation [grant number 25SG10].

Competing interests

The authors declare no competing interests.