INTRODUCTION
Cellular antioxidant networks are essential for maintaining redox homeostasis, balancing the suppression of oxidative damage with the controlled transmission of redox signals that shape cell fate decisions. Among them, glutathione exists in two interconvertible forms: reduced glutathione (GSH) and its oxidized disulfide (GSSG), formed by the linkage of two GSH molecules via a disulfide bond. GSH is a tripeptide of glutamic acid, cysteine, and glycine (γ-L-glutamyl-L-cysteinyl-glycine). It represents the most abundant non-protein thiol in mammalian cells, present at millimolar concentrations (1–10 mM) in the cytosol
1.
GSH plays fundamental roles in antioxidant defense, xenobiotic detoxification, metabolic regulation, and redox signaling (Fig. 1). Its antioxidant function is mediated by the reductive detoxification of hydrogen peroxide (H
2O
2), reactive nitrogen species (RNS) such as peroxynitrite (ONOO
−), and lipid hydroperoxides (L-OOH). These reactions are catalyzed mainly by glutathione peroxidases (GPXs), a selenium-dependent enzyme family, and peroxiredoxins (PRDXs), which often function in concert with their reducing partners thioredoxins (TXNs, also known as TRXs)
1,2. Beyond serving as a reducing cofactor for antioxidant enzymes, GSH exerts direct antioxidant activity by scavenging reactive oxygen species (ROS) and RNS, including superoxide anions (O
2•
−), H
2O
2, and hydroxyl radicals (•OH)
3. It also maintains other antioxidants, such as vitamins C and E, in their reduced active forms. These processes rely on nicotinamide adenine dinucleotide phosphate-reduced (NADPH) as the principal electron donor
4.
In addition to its role in redox buffering, GSH is a regulator of redox-dependent signaling and post-translational modification. As the principal low-molecular-weight antioxidant and reductant, GSH maintains the intracellular environment in a highly reduced state. Under physiological conditions, more than 98% of the total GSH pool exists in the reduced form, thereby shaping cellular redox potential and influencing proliferation, differentiation, and cell death
5.Although the GSH/GSSG ratio has long been regarded as an indicator of cellular redox status, recent live-cell studies using genetically encoded redox sensors indicate that increases in GSSG are typically transient and compartmentalized rather than sustained intracellular accumulations
6,7. Consistent with this, cells limit GSSG buildup through efficient recycling and export mechanisms
8.
A major mechanism linking GSH to signal regulation is S-glutathionylation (SSG), a reversible post-translational modification in which GSH forms a mixed disulfide bond with reactive cysteine residues on target proteins. This process is mediated by glutaredoxins (GLRXs, also known as GRXs), which catalyze both glutathionylation and deglutathionylation, as well as by sulfiredoxins (SRXs)
9,10. In addition, glutathione S-transferase Pi 1 (GSTP1) facilitates SSG of selected substrates in specific cellular contexts
11. This reversible cycle is essential for both the propagation and resolution of redox signaling and for restoring protein function. By inducing structural and functional changes in target proteins, SSG acts as a molecular switch that regulates transcription factors (
e.g., tumor protein p53 (TP53, also known as p53) and nuclear factor-κB (NF-κB), and activator protein 1 (AP-1)), kinases (
e.g., protein kinase C (PKC) and mitogen-activated protein kinase (MAPK)), and metabolic enzymes. At the same time, SSG protects reactive cysteine residues from irreversible hyperoxidation to sulfonic acids, thereby preserving protein function during oxidative stress
12. However, a key unresolved question is how the specificity, reversibility, and spatiotemporal dynamics of SSG are coordinated across different cellular compartments and pathological contexts.
GSH also plays a role in cellular detoxification. Glutathione S-transferases (GSTs) catalyze the conjugation of GSH to a broad range of electrophilic compounds, including endogenous signaling molecules, reactive aldehydes such as 4-hydroxynonenal (4-HNE), and heavy metals
13,14. This conjugation increases the water solubility of electrophilic substrates, facilitating their export
via multidrug resistance-associated protein transporters and promoting subsequent elimination from the body.
In addition to its redox and detoxifying functions, GSH contributes to metabolic homeostasis. It acts as a physiological ligand for Fe
2+, promoting assembly of iron-sulfur (Fe-S) clusters and biosynthesis of Fe-S proteins, which participate in essential biological functions including electron transfer, the Krebs cycle, maintenance of genome integrity, enzymatic catalysis, and iron homeostasis
15. Recent studies have revealed an autoregulatory mechanism whereby mitochondrial GSH (mtGSH) homeostasis is controlled through a feedback loop involving solute carrier family 25 member 39 (SLC25A39) transporter and mitochondrial protease AFG3 like matrix AAA peptidase subunit 2 (AFG3L2), coupling GSH levels to iron-sulfur cluster biogenesis
16,17. Additionally, nuclear GSH contributes to the regulation of cell proliferation by influencing gene expression and epigenetic states
18. Together, these studies redefine GSH as a multifaceted regulator of cellular metabolism and compartment-specific homeostasis, rather than merely a passive antioxidant reservoir.
Cellular GSH homeostasis is maintained through the coordinated regulation of biosynthesis, utilization, transport, recycling, and degradation (Fig. 1). The effects of altered GSH homeostasis vary with cell type, subcellular localization, and metabolic state. GSH is synthesized in the cytosol through two sequential, ATP-dependent enzymatic reactions
19. The first and rate-limiting step, catalyzed by glutamate-cysteine ligase (GCL), generates γ-glutamylcysteine from glutamate and cysteine. GCL is a heterodimeric enzyme composed of a glutamate-cysteine ligase catalytic subunit (GCLC), which confers enzymatic activity, and a glutamate-cysteine ligase modifier subunit (GCLM), which enhances GCLC efficiency and attenuates feedback inhibition by GSH. In the second step, glutathione synthase (GS) catalyzes the addition of glycine to γ-GC, yielding GSH.
Among GSH precursors, cysteine availability is typically rate-limiting due to its low intracellular abundance and susceptibility to oxidation. This metabolic bottleneck places cysteine availability at the center of cellular redox adaptation and makes it a critical determinant of GSH homeostasis under stress conditions. Endogenous cysteine is generated primarily through the transsulfuration pathway, which converts methionine to cysteine
20. Under cysteine-limiting conditions, cells depend on the uptake of extracellular cystine, subsequently reduced to cysteine in the cytosol. This process is mediated by the system xc
− antiporter, composed of solute carrier family 7 member 11 (SLC7A11, also known as xCT) and its chaperone solute carrier family 3 member 2 (SLC3A2, also known as CD98). This transporter exchanges intracellular glutamate for extracellular cystine in a 1:1 ratio
21. This dependency renders cells sensitive to fluctuations in extracellular cystine supply and positions system xc
− as a key metabolic checkpoint linking nutrient availability to oxidative stress responses and ferroptotic susceptibility. Pharmacologically, cysteine availability can be augmented by N-acetylcysteine (NAC), which is deacetylated intracellularly to supply cysteine for GSH synthesis
22.
During peroxide detoxification, GSH is oxidized to GSSG and subsequently reduced back to GSH by glutathione-disulfide reductase (GSR, also known as GR) in an NADPH-dependent manner. The required reducing equivalents are supplied mainly by the pentose phosphate pathway (PPP) and isocitrate dehydrogenase (IDH)
23,24. To eliminate GSH-conjugated toxins, GSH is exported from cells, primarily by members of the ATP binding cassette subfamily C member (ABCC, also known as MRP) family
25. Once in the extracellular space, GSH can be degraded by the membrane-bound ectoenzyme gamma-glutamyltransferase (GGT, also known as γ-GT), which cleaves the γ-glutamyl bond, releasing cysteinyl-glycine (Cys-Gly)
26,27. Cys-Gly is further split by dipeptidases into cysteine and glycine for cellular reuptake and regeneration of new GSH. In addition to extracellular catabolism, intracellular GSH degradation is mediated by the ChaC GSH specific gamma-glutamylcyclotransferase family (CHAC1 and CHAC2), which cleaves GSH into 5-oxoproline and Cys-Gly
28,29. CHAC1 is often induced under cellular stress, including unfolded protein response signaling, and its activation modulates redox balance and influences apoptosis and ferroptosis
30. However, the contribution of CHAC2 to basal glutathione homeostasis remains less clearly defined, although it may reflect a complementary regulatory function.
Consistent with its redox role, alterations in GSH homeostasis influence cellular sensitivity to oxidative stress, mitochondrial dysfunction, and multiple forms of regulated cell death. These findings have motivated therapeutic strategies aimed at depleting GSH to promote tumor cell death. However, accumulating evidence reveals a more complex paradigm in which GSH can also act as a chemically active participant that facilitates lethal biochemical reactions, particularly in metal-dependent contexts.
We propose that the seemingly opposing roles of GSH in cell survival and cell death reflect the spatial, metabolic, and microenvironmental context in which GSH homeostasis is disrupted. In this review, we examine how GSH restrains both executioner-driven cell death pathways, including apoptosis, necroptosis, and pyroptosis, which are mediated by defined signaling cascades and dedicated effector proteins, and metabolism-centered forms of cell death, including ferroptosis and cuproptosis, which arise from overwhelming lipid peroxidation and proteotoxic stress, respectively. We also discuss the conditions under which disruption of GSH homeostasis instead promotes cell death in specific metabolic and microenvironmental settings.
THE ANTI-CELL DEATH ROLE OF GSH
The anti-cell death functions of GSH are mediated through coordinated control of oxidative stress, redox-sensitive signaling, and cysteine-based protein modification, thereby setting the threshold for commitment to regulated cell death.
GSH and apoptosis
Apoptosis is the canonical form of regulated cell death, characterized by cell shrinkage, chromatin condensation, nuclear fragmentation, and the formation of membrane-enclosed apoptotic bodies
31. It is initiated through two principal pathways: the extrinsic (death receptor-mediated) and intrinsic (mitochondrial) pathways, which converge on the activation of executioner caspases (Fig. 2a). In both pathways, cysteine-dependent proteases known as caspases execute apoptotic cell death by cleaving key structural and regulatory proteins, primarily through the action of caspase-3 (CASP3) and caspase-7 (CASP7), with caspase-6 (CASP6) contributing to downstream substrate cleavage and apoptotic progression.
The intrinsic pathway is triggered by diverse intracellular stresses, such as DNA damage, metabolic perturbations, or endoplasmic reticulum stress, leading to mitochondrial outer membrane permeabilization (MOMP). This process is governed by BCL2 family proteins, including the pro-apoptotic BCL2 associated X, apoptosis regulator (BAX) or BCL2 antagonist/killer 1 (BAK1, also known as BAK) as well as anti-apoptotic members, such as BCL2 apoptosis regulator (BCL2) and BCL2 like 1 (BCL2L1, also known as BCL-XL). MOMP results in the release of multiple mitochondrial proteins, most notably cytochrome c (CYCS), into the cytosol. CYCS then binds apoptotic protease activating factor 1 (APAF1), promoting apoptosome assembly, caspase 9 (CASP9) activation, and subsequent activation of executioner caspases (
e.g., CASP3) (Fig. 2a)
32.
In contrast, the extrinsic pathway is initiated by death ligands (
e.g., Fas ligand (FASLG), tumor necrosis factor (TNF), TNF superfamily member 10 (TNFSF10, also known as TRAIL)) binding their respective receptors, which recruit the adaptor protein Fas associated via death domain (FADD), forming the death-inducing signaling complex (DISC). The DISC activates caspase 8 (CASP8) or caspase 10 (CASP10) and downstream executioner caspase cascades. Active CASP8 cleaves BH3 interacting domain death agonist (BID) to generate truncated BID (tBID), thereby linking death receptor signaling to the mitochondrial pathway through BAX- and BAK1-mediated MOMP. Apoptosis is generally immunologically silent, as plasma membrane integrity is preserved and apoptotic bodies are cleared by phagocytes through efferocytosis
33. In this context, GSH and SSG influence apoptotic sensitivity in a context-dependent manner by regulating mitochondrial integrity and death signaling through redox-dependent mechanisms (Fig. 2a).
GSH depletion as a sensitizer for apoptosis
GSH depletion represents an early event in apoptosis and is a key determinant of apoptotic sensitivity. GSH protects cells from oxidative stress-induced apoptosis by scavenging H
2O
2, and exogenous GSH supplementation enhances cell viability under oxidative conditions
34. Accordingly, elevated intracellular GSH is a hallmark of apoptosis-resistant phenotypes. Enhancing the GSH/GSSG ratio
via overexpression of GSR or GCLC lowers ROS levels and confers protection against H
2O
2- or TNF-induced apoptosis
35,36. Of note, global depletion of GSH is insufficient to trigger apoptosis
34. This relative resistance likely reflects preferential preservation of mtGSH, which accounts for only ~10–15% of total cellular GSH but constitutes the principal antioxidant defense within mitochondria. The mtGSH maintains mitochondrial integrity and promotes adaptive survival responses induced by prolonged cytoplasmic GSH loss, including upregulation of anti-apoptotic proteins such as BCL2
37,38.
In contrast, selective depletion of mtGSH initiates a defined cascade of pro-apoptotic events. mtGSH preserves mitochondrial membrane potential (ΔΨm) and limits oxidative membrane damage, thereby preventing CYCS release
39. Loss of mtGSH disrupts ΔΨm, promotes MOMP, and activates intrinsic apoptotic signaling, whereas restoration of mtGSH via GSH ester supplementation suppresses apoptosis
38,40. These studies suggest that subcellular compartmentalization of GSH, rather than total cellular GSH abundance alone, is a critical determinant of apoptotic sensitivity.
Apoptosis-associated GSH efflux
A pronounced intracellular GSH efflux is a functionally required event during apoptosis, preceding ROS generation
41. During initiation, cells export reduced GSH through plasma membrane transporters before membrane rupture, establishing a redox environment permissive for apoptotic signaling. This rapid GSH loss disrupts redox homeostasis, promotes ROS accumulation, and facilitates redox-dependent protein modifications that drive apoptotic commitment. GSH efflux represents an active component of the death program, rather than a passive consequence
41. ABCC family members serve as ATP-dependent transporters of GSH, GSSG, or GSH conjugates
42. However, their contribution is context-dependent, as genetic ablation of
ABCC1 does not fully prevent apoptosis-associated GSH loss in certain models
43, indicating the existence of compensatory or alternative efflux mechanisms. This also raises an unresolved question as to whether GSH efflux serves primarily as a permissive biochemical event or represents an actively regulated signaling step in apoptotic commitment.
GSH-dependent SSG in apoptosis
Redox regulation through SSG integrates GSH signaling with multiple apoptotic pathways. In the intrinsic pathway, oxidative stress or GSH depletion induces BAX SSG at Cys62, promoting its mitochondrial translocation and subsequent activation of CASP9 and CASP3
44. In the extrinsic pathway, death receptors such as FAS contain cysteine-rich extracellular domains that undergo SSG under oxidative conditions. SSG of FAS at Cys294 enhances ligand binding, promotes receptor clustering within lipid rafts, and accelerates the activation of CASP8 and CASP3
45,46. Consistent with a protective role of redox control, overexpression of glutaredoxin (GLRX, also known as GRX1) attenuates FAS SSG and partially protects against FAS-dependent apoptosis. In contrast, in ethanol-exposed
Grx1-deficient mice, accumulation of FAS-SSG drives hepatocyte apoptosis through dysregulation of NF-κB and AKT serine/threonine kinase 1 (AKT1) signaling
47.
SSG can also restrain apoptotic execution. CASP3 SSG inhibits its proteolytic activity
48, highlighting the context-dependent effects of redox modifications. For example, ER stress activates DNA damage inducible transcript 3 (DDIT3, also known as CHOP)- and mitogen-activated protein kinase 8 (MAPK8, also known as JNK)-dependent apoptosis pathways, whereas heat shock protein family A member 5 (HSPA5, also known as BIP) SSG at Cys420/441 modulates ATPase activity and protein folding, paradoxically suppressing apoptosis
49. In contrast, GSTP1 promotes ER stress-induced apoptosis in liver cancer by glutathionylating SERCA, thereby inhibiting pro-survival MAPK8 signaling
50. Thus, SSG does not simply suppress or promote apoptosis, but instead functions as a context-dependent redox rheostat that determines the threshold and direction of apoptotic signaling.
GSH and pyroptosis
Pyroptosis is a form of lytic cell death characterized by gasdermin-mediated plasma membrane pore formation, leading to osmotic swelling, membrane rupture, and the release of mature pro-inflammatory cytokines, including interleukin 1 beta (IL1B, also known as IL-1β) and interleukin 18 (IL18). This process is primarily executed by members of the gasdermin family, most notably gasdermin D (GSDMD), which is activated by inflammatory caspases in both canonical and non-canonical inflammasome pathways
51.
Canonical pyroptosis is triggered by pattern recognition receptors (
e.g., NLR family pyrin domain containing 3 (NLRP3)) that sense pathogen-associated molecular patterns (PAMPs) or damage-associated molecular patterns (DAMPs), such as the accumulation of ROS or the efflux of potassium ions (K
+). Upon activation, NLRP3 engages the scaffold protein NIMA related kinase 7 (NEK7) and recruits the adaptor apoptosis-associated speck-like protein containing a CARD (PYD and CARD domain containing (PYCARD, also known as ASC)) to assemble the inflammasome complex. This complex subsequently activates caspase 1 (CASP1), which cleaves pro-IL1B, pro-IL18, and GSDMD. The liberated N-terminal fragment of GSDMD (GSDMD-N) oligomerizes within the plasma membrane to form pores, leading to pyroptotic cell death
52. Non-canonical pyroptosis is initiated by the detection of intracellular lipopolysaccharide (LPS) from Gram-negative bacteria, leading to the activation of caspase 4 (CASP4) and caspase 5 (CASP5) in humans and caspase 11 (CASP11) in mice, which cleave GSDMD independently of inflammasome assembly. GSH serves as a potent anti-pyroptotic regulator through two primary mechanisms: maintenance of cellular redox homeostasis and post-translational modification of inflammasome components via SSG (Fig. 2b).
GSH and redox control of inflammasome activation
Excess ROS generation is a trigger for NLRP3 inflammasome activation. Mechanistically, the thioredoxin interacting protein (TXNIP) serves as a link between cellular redox state and NLRP3 activation. Under reducing conditions sustained by GSH, TXN remains reduced and prevents NLRP3 assembly. However, oxidative stress or GSH depletion causes TXNIP to dissociate from TXN and instead bind to NLRP3, inducing conformational activation and promoting inflammasome assembly
53,54. This GSH-dependent redox regulation also involves GPX-dependent lipid peroxide control. Inhibition of GPX4 promotes H
2O
2 accumulation and lipid peroxidation, triggering NLRP3-mediated pyroptosis in neurotoxic rats and septic mice
55,56. GPX3 overexpression inhibits microglial pyroptosis and reduces NLRP3 and GSDMD-N expression
57. A complementary axis involves transcriptional reinforcement of antioxidant programs, including NFE2 like BZIP transcription factor 2 (NFE2L2, also known as NRF2) signaling, which increases GPXs and further suppresses ROS-driven inflammasome activation
58. GPX4 is required for repression of both canonical CASP1- and non-canonical CASP11-mediated cleavage of GSDMD
56, indicating that GSH-dependent antioxidant systems do not only inhibit inflammasome assembly but also limit the oxidized membrane environment that favors gasdermin pore formation. These data support a role for GSH as an upstream regulator of pyroptosis by coupling metabolic redox status to inflammasome activation.
GSH constrains NLRP3 activation primarily by scavenging H
2O
2 and preserving mitochondrial redox balance. NAC elevates intracellular GSH, suppresses H
2O
2 accumulation, and blocks the NLRP3-CASP1-GSDMD pathway
in vitro and
in vivo59-61. SLC7A11 overexpression inhibits pyroptosis by ensuring robust GSH levels and reduced ROS
62, whereas inhibition of glutamine uptake causes decreased GSH and increased ROS, leading to GSDME-dependent pyroptosis in non-small cell lung cancer (NSCLC) cells
63. Restoration of mtGSH by GSH ethyl ester limits oxidative stress and suppresses NLRP3 activation in neuronal models of amyloid toxicity
64. Nevertheless, a major unresolved question remains how metabolic and subcellular regulation of GSH intersects with inflammasome signaling to dictate context-dependent pyroptotic cell fate.
SSG as an inflammasome checkpoint
SSG, as a thiol-based switch, links GSH metabolism to inflammasome licensing. Distinct GST subfamilies exert bidirectional control over this process. In several disease settings, cytosolic GSTs act as anti-pyroptotic brakes by promoting or maintaining SSG on executioner or signaling nodes. For example, GSTM1 suppresses NLRP3 output by mediating CASP1 SSG, thereby limiting CASP1 activation and downstream pyroptosis in acute keratitis
65. GSTP1 is implicated in restraining cardiomyocyte pyroptosis
66.
In contrast, glutathione S-transferase omega 1 (GSTO1) functions as a pro-pyroptotic licensing factor by catalyzing deSSG of core inflammasome assembly components. Under resting conditions, ASC is S-glutathionylated (notably at Cys171), a modification that impairs ASC oligomerization and prevents spontaneous inflammasome formation
67. Upon activation cues, GSTO1 removes this inhibitory SSG on ASC to enable ASC speck formation and productive inflammasome assembly
68. Similarly, NEK7 is glutathionylated at Cys253 and requires GSTO1-dependent deSSG to interact efficiently with NLRP3
67.
Together, these findings support a model in which GSTs exert bidirectional control over pyroptosis with GSTM1 and GSTP1 dampening inflammasome execution, whereas GSTO1 derepresses inflammasome assembly. These apparently opposing effects highlight that the impact of GSH-linked thiol modifications on pyroptosis is highly node-specific and should not be viewed as uniformly anti-inflammatory.
GSH and necroptosis
Necroptosis is a regulated form of necrotic cell death characterized by activation of RIPK3 and mixed lineage kinase domain-like pseudokinase (MLKL), with receptor-interacting serine/threonine kinase 1 (RIPK1) participating in many, but not all, necroptotic contexts. This pathway is engaged when CASP8 activity is inhibited downstream of death receptors, such as TNF receptor superfamily member 1A (TNFRSF1A). Under these conditions, RIPK1 and RIPK3 form a necrosome complex that phosphorylates MLKL. Phosphorylated MLKL oligomerizes and translocates to the plasma membrane, where it forms pores that cause cell swelling, membrane rupture, and release of DAMP. Necroptosis is inherently pro-inflammatory. Although this pathway contributes to host defense against viral infections, its dysregulation drives tissue damage and inflammatory pathology
69. GSH can act as a molecular "gatekeeper" at multiple checkpoints of the necroptotic signaling cascade (Fig. 2c).
GSH and redox control
The initiation of necroptosis is sensitive to cellular redox state. H
2O
2 oxidizes cysteine residues (Cys257, Cys268, and Cys568) on RIPK1, a process facilitating RIPK3 recruitment to the necrosome
70. The final "execution" step of necroptosis, MLKL oligomerization, is also under redox control. H
2O
2 promotes the formation of inter-chain disulfide bonds between MLKL subunits, a prerequisite for MLKL translocation and membrane rupture
71. GSH serves as the primary buffer against H
2O
2. Under physiological conditions, a high GSH/GSSG ratio, maintained largely by GPX4, buffers oxidative stress. In contrast, deletion or inhibition of GPX4 causes rapid H
2O
2 accumulation and RIPK3-mediated necrosome activation
72,73. Meanwhile, GSH preservation in cytosolic and mitochondrial fractions downregulates the expression of RIPK1, RIPK3, or MLKL, preventing necrosome formation
74,75. NAC pretreatment inhibits RIPK1- and RIPK3-dependent necroptosis during intestinal injury and prevents necrosome formation
76. Furthermore, GSH-dependent stress-responsive protein, sestrin2 (SESN2) interacts with RIPK3 to maintain redox balance and inhibit its phosphorylation, preventing RIPK1–RIPK3 complex stabilization
77. By controlling mtROS, peroxiredoxin 3 (PRDX3) prevents necrosome activation, acting as a brake on necroptosis
78. However, GSH depletion activates stress-activated protein kinases to promote necroptotic cell death
79.
The contribution of oxidative stress to necroptosis appears to be highly context-dependent. An early study using mitochondria-depleted cells demonstrated that although TNF-induced ROS production accompanied necroptosis, mitochondrial ROS was not required for RIPK3-dependent cell death execution. Specifically, the ROS scavenger butylated hydroxyanisole delayed TNF-induced necroptosis but had no effect on necroptosis induced by direct RIPK3 oligomerization, suggesting that ROS may facilitate upstream signaling events rather than serve as an obligate execution mechanism
80. Overall, current evidence suggests that oxidative stress functions primarily as a context-dependent amplifier of necroptotic signaling rather than as a universal execution mechanism.
SSG control
Under homeostatic conditions, CASP8 acts as the primary molecular "brake" on necroptosis by cleaving and inactivating RIPK1 and RIPK3. A pro-oxidative shift caused by GSH depletion triggers glutathionylation of CASP8 at residues Cys360 and Cys409
81. This modification inactivates CASP8, preventing cleavage of necrosome components and allowing the necroptotic signal to proceed unchecked. GSH-dependent redox control also extends to mitochondrial dynamics. SSG of mitofusin 2 (MFN2) disrupts mitochondria–endoplasmic reticulum contact sites, compromising organelle communication and promoting activation of the RIPK1-RIPK3-MLKL axis in neurotoxicity models
82. These observations indicate that GSH-dependent thiol modifications regulate necroptosis at multiple hierarchical levels, from upstream checkpoint proteases to terminal membrane-disrupting effectors.
Collectively, by maintaining a reduced intracellular environment, GSH restrains pro-necrotic kinase activation and preserves inhibitory protease function, particularly CASP8, thereby preventing aberrant necroptotic signaling and inflammatory cell death.
GSH and ferroptosis
Ferroptosis is a form of oxidative, regulated cell death characterized by the iron-dependent accumulation of lipid peroxides
83. Unlike other forms of regulated cell death, ferroptosis does not rely on a single dedicated executioner protein but instead reflects the balance between pro-oxidant metabolic processes and multilayered defense systems that limit oxidative membrane damage. Mechanistically, ferroptosis is driven by the accumulation of lethal phospholipid hydroperoxides (PL-OOH), primarily derived from polyunsaturated fatty acid (PUFA)-containing phospholipids. This process can be initiated by iron-catalyzed Fenton chemistry, in which hydroxyl radicals (•OH) generated from H
2O
2 trigger a self-propagating lipid peroxidation chain reaction. During propagation, lipid radicals react with molecular oxygen to form lipid peroxyl radicals (LOO•), which subsequently abstract hydrogen atoms from adjacent phospholipids, thereby generating new lipid radicals and lipid hydroperoxides (LOOH). As a result, a single initiating event can be amplified into extensive membrane damage until the radical chain is terminated. Several endogenous metabolites, including ubiquinol, vitamin E, and vitamin K hydroquinone, function as radical-trapping antioxidants that terminate lipid peroxyl radicals and suppress ferroptotic membrane damage
84-87.
Three interconnected pathways function as core suppressors of ferroptosis: (1) the GPX4-GSH system (central suppressor), (2) the SLC7A11-cystine import pathway (GSH synthesis), and (3) the AIF family member 2 (AIFM2, also known as FSP1)-NAD(P)H–quinone pathway (parallel protection)
84-87. Conceptually, these pathways should be viewed not as isolated modules but as an integrated redox defense network, in which failure of one arm can be partially compensated by the others until a critical threshold of lipid peroxide accumulation is exceeded. Small molecules such as RSL3 and erastin, which primarily target GPX4 and SLC7A11, respectively, demonstrate that impairment of cystine uptake or lipid peroxide detoxification is sufficient to directly trigger ferroptotic cell death. Cellular susceptibility to ferroptosis is therefore governed by pathways controlling lipid metabolism, iron homeostasis, and redox balance.
ROS scavenging and detoxification
The antioxidant capacity of GSH derives from its thiol group, which confers potent reducing activity toward ROS, including superoxide anions and H
2O
288. In the context of ferroptosis, GSH's thiol group donates electrons to convert H
2O
2, thereby limiting iron-catalyzed lipid peroxidation
89. The sulfur atom of cysteine within GSH can also interact with oxygen-centered radicals, reducing their chemical reactivity and attenuating oxidative chain reactions
90. Beyond direct H
2O
2 scavenging, GSH also detoxifies lipid peroxidation byproducts, including malondialdehyde (MDA) and 4-HNE, through GST-mediated conjugation, facilitating their removal and preventing secondary cytotoxic signaling
91,92. These observations underscore that GSH availability is a critical determinant of the oxidative threshold beyond which lipid peroxidation becomes self-propagating and ferroptotic death ensues.
The GPX4-GSH axis
The antioxidant function of GSH is largely executed through GPXs and PRDXs, which catalyze reductive detoxification of lipid peroxidation
93. Among these enzymes, GPX4 is the principal endogenous suppressor of ferroptosis by limiting lipid peroxidation within biological membranes (Fig. 3)
94. GPX4 reduces phospholipid hydroperoxides (PL-OOH) to their corresponding non-toxic phospholipid alcohol (PL-OH) using GSH as an electron donor, thereby preserving membrane integrity
95. This reaction depends on the catalytic selenocysteine residue of GPX4, which is directly targeted and inactivated by ferroptosis inducers such as RSL3. Complete loss of GPX4 is incompatible with viability, and pharmacological or genetic GPX4 inhibition is widely exploited to induce ferroptosis in cells and tissues. In addition, GSH indirectly stabilizes GPX4 by maintaining a reducing intracellular environment
96.
GPX4 is broadly expressed across tissues, with high levels in the testis, where it is essential for spermatogenesis
97. Three physiological GPX4 isoforms, including cytosolic (cGPX4), mitochondrial (mGPX4), and nuclear (nGPX4), have been identified. cGPX4 and mGPX4 suppress ferroptosis in a context-dependent manner
94,98,99. Recent work further revealed that peroxiredoxin 6 (PRDX6) binds GPX4 via a C47-mediated disulfide interaction to facilitate GPX4 membrane association and ferroptosis protection
100. GPX4 released from ferroptotic cancer cells functions as a DAMP that suppresses antitumor immunity by engaging the zona pellucida glycoprotein 3 (ZP3) receptor on dendritic cells
101. How compartment-specific GSH metabolism modulates GPX4 activity during ferroptotic stress remains an important unresolved question.
The PRDX-GSH axis
PRDXs constitute a family of cysteine-dependent peroxidases that regulate intracellular H
2O
2 and lipid peroxide levels
102. In mammals, PRDX1–4 preferentially detoxify H
2O
2, whereas PRDX5 shows higher reactivity toward alkyl hydroperoxides and peroxynitrite, and PRDX6 reduces both H
2O
2 and LOOHs. Most PRDXs (PRDX1–5) are regenerated through the TXN system; however, PRDX6 is unique in that it primarily uses GSH as its physiological reductant, often with GSTP1 assistance. GSH reduces the oxidized catalytic cysteine (Cys47-SOH) of PRDX6, sustaining its peroxidase activity against lipid peroxides, like GPX4, and linking PRDX6 directly to ferroptosis resistance (Fig. 3)
103.
In addition to GPX-like activity, PRDX6 exhibits calcium-independent phospholipase A
2 (iPLA
2) and lysophosphatidylcholine acyltransferase (LPCAT) activities
104,105. PRDX6 selectively binds oxidized phospholipids, becomes membrane-associated, and executes coordinated membrane repair through reductive detoxification, iPLA
2-mediated deacylation, and LPCAT-dependent reacylation with non-oxidized fatty acids. PRDX6 also regulates ferroptosis indirectly by controlling GPX4 biosynthesis and localization. In a GSH-dependent manner, PRDX6 functions as a selenium acceptor and carrier, transferring selenium to selenophosphate synthetase 2 (SEPHS2) to support selenocysteine-tRNA synthesis required for GPX4 translation
106. Loss of
PRDX6 reduces GPX4 protein levels and sensitizes cells to ferroptosis. As discussed earlier, PRDX6 binds GPX4 through a Cys47-mediated disulfide bond, facilitating GPX4 localization to cellular membranes and enhancing lipid peroxide detoxification
100. Consistently, combined inhibition of PRDX6 and ferroptosis inducers synergistically amplifies lipid peroxidation and suppresses tumor growth. However, whether the GPX4–PRDX6 interaction represents a broadly conserved ferroptosis defense mechanism or is restricted to specific cellular contexts remains to be established.
Although rupture of the plasma membrane represents the terminal event of ferroptotic death, emerging evidence indicates that ferroptotic signaling is initiated at intracellular membrane compartments, including the ER, Golgi-associated vesicles, and lysosomes
107-110. PRDX6, while predominantly cytosolic, is also localized to lysosomes and acid organelles
111,112. Elucidating how this compartmentalized GPX4–PRDX6 defense system suppresses early ferroptotic membrane damage remains an important future direction. These findings further suggest that ferroptosis resistance is spatially organized, with distinct intracellular membrane compartments likely exhibiting different antioxidant vulnerabilities and repair capacities.
Distinct PRDX isoforms also regulate ferroptosis through signaling functions. PRDX1 stabilizes NFE2L2 by inhibiting cullin 3 (CUL3)-mediated ubiquitination, thereby promoting transcription of genes involved in GSH synthesis (
GCLC,
GCLM) and cystine import (
SLC7A11)
113. In contrast, mitochondrial PRDX3 undergoes hyperoxidation during lipid peroxidation stress and translocates to the plasma membrane, where it suppresses cystine uptake, depletes intracellular GSH, and amplifies ferroptotic execution
114. These opposing roles of distinct PRDX isoforms suggest that PRDXs function not simply as antioxidant enzymes but as context-dependent determinants of ferroptosis sensitivity.
Complementary mechanisms
GSH and TXN systems constitute the two thiol-based antioxidant networks that cooperate to restrain ferroptosis
115. The TXN system suppresses ferroptosis through multiple mechanisms: by supplying reducing equivalents to PRDXs, facilitating cystine-to-cysteine reduction to support GSH synthesis, and acting as an independent ferroptosis suppressor (Fig. 3)
116-119. Consistently, thioredoxin domain-containing protein 12 (TXNDC12) inhibits lipid peroxidation in a GPX4-independent manner
120. Notably, canonical GPX4 inhibitors, such as RSL3 and ML162, also inhibit thioredoxin reductase 1 (TXNRD1)
121,122. TXNRD1 has emerged as a selective vulnerability in
KRAS-wild-type and KRAS inhibitor-resistant tumors, where its blockade induces ferroptosis through combined GSH depletion and iron overload
117. Crosstalk between thiol antioxidant systems is further reinforced by redox-sensitive transcription factors, including NFE2L2 and TP53
123.
Ascorbate (vitamin C) and GSH cooperate through redox cycling. Unlike GSH, vitamin C is obtained from the diet and can be regenerated by GSH and NADPH
124. Although GSH and vitamin C show partial functional redundancy under oxidative stress, their combined presence provides maximal protection against lipid hydroperoxides. However, in the context of elevated vitamin C uptake, high vitamin C can act as a pro-oxidant by reducing iron and promoting lipid peroxidation, GPX4 inactivation, and ferroptotic cell death
125. This bidirectional behavior highlights a critical context dependence in antioxidant biology, where the same metabolite may either suppress or amplify ferroptotic stress depending on iron availability and intracellular redox flux.
SSG represents an adaptive response that protects cells from ferroptosis under GSH-limiting conditions. GPX4 inhibition enhances glioma radiosensitivity by inducing oxidative stress-dependent SSG and degradation of transglutaminase 2 (TGM2), impairing DNA repair
126. In contrast, CHAC1-driven GSH degradation reduces protein SSG, enhances transferrin receptor (TFRC)–mediated iron uptake, and exacerbates ferroptosis, whereas hemoglobin SSG shields reactive cysteine residues from excessive oxidation
127,128. Rather than functioning as isolated antioxidant modules, these pathways form a highly interconnected redox buffering network in which compensatory capacity, metabolic flux, and iron availability collectively determine ferroptotic susceptibility.
GSH and iron homeostasis
GSH also protects against ferroptosis through coupling with iron homeostasis. As a physiological ligand of labile Fe
2+, GSH chelates redox-active iron and limits its participation in Fenton reactions
129. GSH is required for stable iron delivery by poly(rC)-binding protein 1 (PCBP1) to ferritin and other client proteins
130. GSH dysfunction not only impairs GPX4 activity but also mobilizes intracellular Fe
2+, thereby accelerating lipid peroxidation and ferroptotic death
131,132. In addition, GSH restores hemoglobin's oxygen-carrying capacity by reducing Fe
3+ back to Fe
2+128. Beyond iron buffering, GSH supports biosynthesis and maintenance of Fe-S clusters, essential cofactors for mitochondrial respiratory enzymes
15. Consistently, GSH depletion or silencing of Fe-S cluster-associated
GLRXs disrupts mitochondrial redox balance, promotes iron overload, and sensitizes cancer cells to ferroptosis
133,134. Consistently, recent work shows that balancing cysteine sulfur for GSH and NFS1-dependent Fe-S clusters is essential for optimal CD8
+ T cell responses
135. These findings underscore that GSH regulates ferroptosis not only through peroxide detoxification but also by coordinating iron trafficking and Fe-S cluster integrity, thereby coupling redox defense to core metabolic fitness.
GSH homeostasis during ferroptosis
The metabolic fate of GSH serves as the fundamental "Arsenal" preventing ferroptosis (Fig. 3). GSH biosynthesis is constrained by cysteine availability, controlled by the cystine-glutamate antiporter system xc
−. SLC7A11 represents a master upstream regulator of GSH synthesis and ferroptosis suppression
136. By mediating cystine uptake, SLC7A11 governs rate-limiting GSH production. Once imported, cystine is rapidly reduced to cysteine, the essential precursor for GSH synthesis
21. Sustained cysteine supply via SLC7A11 maintains high intracellular GSH levels, which are required to support the catalytic activity of downstream effectors, such as GPX4. Accordingly, pharmacological inhibition of SLC7A11 or cysteine deprivation blocks GSH synthesis, leads to GSH depletion, and induces ferroptosis by disabling the antioxidant defense system
85,137. However, the extent to which SLC7A11 dependency defines ferroptosis sensitivity varies substantially across cell types, metabolic states, and microenvironmental nutrient availability.
SLC7A11 localization and function are dynamically regulated by autophagy and lysosomal trafficking
138,139. At the lysosomal membrane, SLC7A11 mediates proton leak, preventing hyperacidity through cystine efflux and Glu influx of lysosomes
140. Loss of
SLC7A11 results in lysosomal hyperacidity and labile iron release, sensitizing cells to ferroptosis. In parallel, cystinosin (CTNS) exports lysosomal cystine to the cytosol, linking lysosomal function to cysteine availability and GSH synthesis
141.
GSH is synthesized through two ATP-dependent steps catalyzed by GCL (GCLC and GCLM) and GS. Induction of this biosynthetic pathway enhances intracellular GSH levels and promotes cancer cell resistance to ferroptosis,
142,143 whereas pharmacological inhibition of GSH synthesis by L-buthionine-(S,R)-sulfoximine (BSO), a GCL inhibitor, induces ferroptosis
144. GCLC also exerts a non-canonical, GSH-independent protective function by maintaining glutamate homeostasis during cystine starvation
145.
GSH homeostasis is further regulated by export, degradation, and regeneration. ABCC1-mediated GSH efflux can paradoxically sensitize cells to ferroptosis by depleting intracellular GSH
146. SLC25A39 acts as a critical transporter for importing cytosolic GSH into the mitochondrial matrix,
147 and its loss promotes ferroptosis
148. Whether SLC25A39-dependent mtGSH is required for mGPX4 activity needs to be further investigated. Degradation occurs both extracellularly and intracellularly: membrane-bound GGT1 catabolizes exported GSH to salvage cysteine, whereas GGT1 activation promotes ferroptosis resistance and tumor progression
149,150. Consistent with this, a recent study demonstrated that catabolism of extracellular glutathione can supply cysteine to support tumor growth under cystine-limited conditions
150. Intracellularly, CHAC1 enzymatically degrades GSH into 5-oxoproline and Cys-Gly, acting as a pro-ferroptotic execution signal by accelerating GSH exhaustion
151. A key unresolved question is whether compartment-specific GSH pools differentially determine ferroptosis initiation at mitochondrial versus non-mitochondrial membranes.
The functional capacity of the GSH system is determined not only by the size of the intracellular GSH pool but also by the efficiency of GSH regeneration from GSSG. Oxidized GSH (GSSG) is reduced back to GSH by GSR in an NADPH-dependent manner, thereby sustaining cellular antioxidant defense and linking redox buffering to central metabolic pathways that generate NADPH, such as the PPP and IDH-dependent reactions. Thus, NADPH availability is a critical determinant of GSH-dependent defense against ferroptosis
4. Disruption of NADPH production compromises GSH regeneration, limits GPX4-dependent detoxification of lipid peroxides, and can promote ferroptosis even in the absence of impaired GSH synthesis
152. This highlights an important conceptual point: ferroptosis sensitivity is determined not only by GSH abundance but also by the metabolic capacity to sustain reductive flux under oxidative stress.
Signaling pathways integrating GSH and ferroptosis
Intracellular GSH levels are integrated with cellular signaling networks that govern oxidative stress responses and ferroptosis susceptibility in cancer (Fig. 3). These effects are mediated in part by stress-responsive transcription factors that coordinate the expression of genes involved in GSH metabolism and redox control. Among them, TP53 exerts a context-dependent dual role in ferroptosis regulation
153. Depending on cellular state and stress cues, TP53 can promote ferroptosis by repressing
SLC7A11 to limit cystine uptake and GSH synthesis, inducing spermidine/spermine N1-acetyltransferase 1 (
SAT1) to enhance ROS generation through polyamine catabolism, and upregulating glutaminase 2 (
GLS2) to increase α-ketoglutarate production and mitochondrial ROS
154-156. Conversely, TP53 can suppress ferroptosis by inhibiting dipeptidyl peptidase 4 (DPP4) activity or inducing cyclin dependent kinase inhibitor 1A (
CDKN1A, also known as p21)
157,158. By maintaining a reductive intracellular environment, GSH modulates TP53 stability and transcriptional activity, thereby indirectly shaping ferroptosis-related gene expression.
NFE2L2 serves as the master regulator of antioxidant defense. Activated NFE2L2 induces a broad antioxidant program, including genes involved in GSH synthesis, regeneration, and utilization (
e.g.,
GCLC,
GSR,
GSTs, and
GPXs), thereby reinforcing ferroptosis resistance
154-156. Beyond classical antioxidant signaling, GSH interfaces with additional pathways, including Hippo-Yes1 associated transcriptional regulator (YAP1) signaling to regulate
GPX4 and
SLC7A11 expression,
159,160 mechanistic target of rapamycin kinase (mTOR) signaling to influence GPX4 protein synthesis through cysteine metabolism,
161 and NF-κB signaling to restrain inflammation-associated ROS production
162. However, excessive NFE2L2 activation may also promote intracellular cysteine accumulation, which can react with sugars to form adducts that impair cell proliferation
163. These findings highlight an important metabolic trade-off, suggesting that excessive augmentation of cysteine or GSH pools may not always be beneficial and could limit the efficacy of antioxidant-based therapeutic strategies. More broadly, this cautions against viewing antioxidant pathway activation as uniformly protective, as excessive reductive adaptation may itself create context-specific vulnerabilities.
An important challenge for future studies will be to define which nodes within this multilayered GSH-centered defense network represent true therapeutic vulnerabilities across different tissue and tumor contexts. Together, these interconnected pathways establish a compartmentalized antioxidant defense architecture that critically determines ferroptosis sensitivity. Defining context-specific vulnerabilities within this network may provide a unifying framework for ferroptosis-based therapeutic strategies in cancer.
GSH and cuproptosis
Cuproptosis is a mitochondrial copper-dependent form of regulated cell death
164. Unlike other forms of regulated cell death, cuproptosis is characterized by the direct binding of reduced copper (Cu
+) to lipoylated proteins of the tricarboxylic acid (TCA) cycle, most notably dihydrolipoamide S-acetyltransferase (DLAT), together with destabilization of iron-sulfur (Fe-S) cluster proteins such as succinate dehydrogenase complex iron–sulfur subunit B (SDHB)
164. These events drive aberrant protein aggregation, proteotoxic stress, mitochondrial dysfunction, and energetic collapse, culminating in cell death. During this process, the mitochondrial matrix reductase ferredoxin 1 (FDX1) functions as a critical upstream regulator by reducing Cu
2+ to the more toxic Cu
+ species and by promoting LIAS-mediated protein lipoylation. Because cuproptosis preferentially targets cells with high oxidative phosphorylation (OXPHOS) dependency, whereas glycolytic tumors remain relatively resistant, it exposes a distinct metabolic vulnerability in cancer and presents unique therapeutic opportunities
165,166. However, because cuproptosis is a relatively recently defined cell death modality, the extent to which this mechanism operates across diverse physiological and pathological contexts remains to be fully established. Increasing evidence indicates that GSH plays a protective role in limiting cuproptotic cell death (Fig. 2d).
GSH as intracellular copper chelator
The most fundamental protective mechanism of GSH against cuproptosis is its function as an intracellular copper chelator. Quantitative analyses indicate that more than 60% of cytoplasmic copper exists as a GSH-bound complex, positioning GSH as the primary initial copper ligand following cellular uptake
167. Depletion of endogenous GSH sensitizes cells to cuproptosis, whereas GSH supplementation mitigates copper-induced proteotoxic stress
164,168. This protective effect is independent of GSH’s classical antioxidant activity and instead relies on its direct interaction with Cu
+, which sequesters excess copper and prevents its aberrant binding to lipoylated proteins within the TCA cycle. mtGSH acts as a suppressor of cuproptosis, as depletion of the mtGSH importer
SLC25A39, increases cellular sensitivity to cuproptosis
169. In contrast, radiation promotes cuproptosis by increasing mitochondrial copper through upregulation of the copper importer solute carrier family 31 member 1 (SLC31A1, also known as CTR1) while simultaneously depleting mtGSH, thereby shifting copper homeostasis toward toxicity
170. An unresolved issue is whether GSH-mediated copper buffering primarily delays cuproptotic initiation or fundamentally determines the threshold for mitochondrial proteotoxic collapse.
Integration with cellular copper homeostasis
GSH functions within a broader copper homeostasis network that includes metallothioneins (MTs), a family of small, cysteine-rich proteins essential for metal detoxification and redox buffering
171,172. Upon cellular entry, copper is frequently bound initially by GSH, which functions as a buffering and trafficking intermediate before transferring the metal to MTs for storage or delivery to cuproenzymes. Emerging evidence further indicates that MTs — particularly MT1 and MT2A — constitute a critical cytoplasmic defense layer against cuproptosis by acting as effective “copper sponges” that sequester excess Cu
+ and restrict its aberrant interaction with mitochondrial targets
170,173,174. In this coordinated system, GSH provides rapid copper sequestration and distribution capacity, whereas certain MTs serve as high-affinity copper reservoirs that limit metal toxicity. Together, these findings suggest that cellular resistance to cuproptosis also depends on a hierarchically organized copper-buffering system, in which mtGSH provides rapid, dynamic metal buffering, whereas metallothioneins confer sustained high-affinity sequestration to prevent mitochondrial copper overload.
Transcriptional regulation of GSH synthesis
Tumor cells also engage complex transcriptional programs to preserve GSH homeostasis under cuproptotic stress. Cuproptosis increases the stability of NFE2L2 by suppressing its proteasomal degradation, leading to transcriptional upregulation of
GCLC and
GCLM and enhanced GSH synthesis, thereby suppressing cuproptosis in pancreatic cancer cells
169. MYC-driven mitochondrial metabolic rewiring increases
DLAT expression and cuproptosis sensitivity in group-3 medulloblastomas, whereas DLAT inhibition lowers TCA cycle metabolism and GSH synthesis
175. In hypoxic tumor microenvironments, hypoxia inducible factor 1 subunit alpha (HIF1A, also known as HIF1α) represses
DLAT expression and confers resistance to cuproptosis
176, supporting the existence of a HIF1A-DLAT-GSH regulatory axis. Consistently, HIF1A also promotes GSH synthesis through transcriptional activation of
GCLM and
SLC7A11, and pharmacological inhibition of SLC7A11 with erastin or sorafenib exacerbates cuproptosis
177,178. Radiation-induced downregulation of BACH transcriptional regulator 1 (BACH1), a transcriptional repressor of
SLC7A11, further sensitizes cells to cuproptosis
170. Inhibition of the biosynthesis of heme, a negative regulator of BACH1, triggers cuproptosis in acute myeloid leukemia
173.
BACH1 itself is a direct transcriptional target of HIF1A, which binds its promoter under hypoxic conditions to induce expression
174, highlighting a multilayered transcriptional network linking hypoxia, GSH metabolism, and cuproptotic vulnerability.
In summary, GSH emerges as a central suppressor of cuproptosis by buffering cytosolic and mitochondrial copper and by integrating with metallothioneins and stress-responsive transcriptional programs. Disruption of this GSH-centered network exposes a therapeutically exploitable vulnerability in OXPHOS-dependent tumors.
PRO-DEATH FUNCTIONS OF GSH METABOLISM
Although GSH is classically viewed as a cytoprotective antioxidant, growing evidence indicates that under specific metabolic and microenvironmental conditions, it can also facilitate regulated cell death. In these settings, GSH promotes cell death by enabling metal toxicity, sustaining redox-active metabolism, and amplifying oxidative or inflammatory signaling. These context-dependent properties have also been exploited therapeutically to achieve tumor-selective cytotoxicity.
GSH-driven metal reduction
In catalytic or metal-complex systems, oxidation of GSH to GSSG can be coupled to transition-metal redox cycling, thereby supporting electron transfer processes involving Cu
2+/Cu
+ or Fe
3+/Fe
2+ couples
179,180. In the presence of redox-active metals, particularly copper, GSH serves as a physiological reducing equivalent donor within these coupled reactions, contributing to the generation of Cu
+ from Cu
2+ under defined chemical or catalytic conditions. In tumor cells, where intracellular GSH levels are often elevated, this reductive environment may facilitate metal-catalyzed redox cycling and paradoxically intensify oxidative stress. For example, GSH-supported formation of Cu
+ or Fe
2+ can promote Fenton-like reactions that convert H
2O
2 into highly reactive radicals, thereby damaging cellular macromolecules and promoting multiple forms of cell death
181,182. Consistently, NAC and cysteine have also been shown to enhance Cu
2+ reduction and copper-induced proteotoxic stress,
168 suggesting that thiol-rich reducing environments can, under specific conditions, contribute to cytotoxicity. In these settings, cell death arises not from GSH depletion
per se but from GSH-supported redox amplification.
This process is typically observed in metal complexes, catalytic nanoplatforms, or enzyme-facilitated electron transfer systems, rather than as a simple direct reaction of free GSH with metal ions. Accordingly, these principles have been therapeutically exploited in nanomedicine platforms designed to induce GSH-dependent ROS amplification (Fig. 4; Supplementary Table S1). In copper-based catalytic systems, sustained GSH-driven copper redox cycling can generate hydroxyl radicals (•OH), causing acute oxidative damage and inducing apoptosis in cancer cells
183-185. In this context, persistent ROS accumulation can overwhelm antioxidant capacity and has been linked to caspase-dependent gasdermin cleavage in mammary cancer cells
186-189. Under iron-catalytic conditions, maintenance of the Fe
2+/Fe
3+ redox couple promotes extensive peroxidation of PUFA-containing membrane phospholipids and drives ferroptotic execution in cancer cells
190,191. In parallel, excessive Cu
+ generated in GSH-rich reductive environments can bind lipoylated mitochondrial TCA-cycle proteins such as DLAT, inducing aberrant protein aggregation and triggering cuproptosis across multiple cancer types
187,189,192-194.
Collectively, these findings indicate that, under defined pathological and catalytic conditions, GSH can be repurposed from an antioxidant safeguard into a redox driver that amplifies oxidative stress and promotes cell death in tumor and inflammatory contexts. A comparable reductive capacity is also observed in heavy metal-detoxifying metallothioneins. In addition, metallothionein 3 (MT3) can extract Cu
2+ from amyloid-β or α-synuclein complexes and facilitate its reduction to Cu
+, a property attributed to its flexible N-terminal β-domain containing a unique threonine insertion at position 5 that distinguishes MT3 from other metallothionein isoforms
195,196.
GSH acts as a copper shuttle
Beyond its role in redox cycling, GSH is a key regulator of intracellular copper trafficking. Thermodynamic analyses using ESI-MS revealed that cellular copper transport follows a gradient of increasing binding affinity, with Cu
+ binding to GSH (
Kd~10
−13 M) at lower affinity than to dedicated copper chaperones such as antioxidant 1 copper chaperone (ATOX1), copper chaperone for superoxide dismutase (CCS), or cytochrome oxidase copper chaperone COX17 (COX17)
197. This positions GSH as an abundant intermediate carrier that forms a large, exchangeable Cu
+ pool, buffering copper ions between cellular uptake and delivery to high-affinity protein targets.
Consistently, GSH depletion using BSO impairs SLC31A1-mediated copper uptake, underscoring its requirement for efficient intracellular copper delivery
198. Further evidence shows that GSH buffering copper aids its delivery to mitochondrial complexes via chaperones like ATOX1 and CCS
199,200. MT3 can also transfer Cu
+ through direct protein−protein interactions
197,201. The lipoic acid groups on lipoylated proteins have relatively higher affinity for Cu
+ (
Kd~10
−17 M)
202. Although GSH may function as a dynamic determinant of intracellular copper bioavailability rather than merely a passive buffering pool (Fig. 4), whether disruption of this shuttle system directly licenses cuproptotic commitment remains an important open question.
The GSH-Fe-S cluster-OXPHOS axis
GSH is indispensable for mitochondrial homeostasis and the biogenesis and stability of Fe-S clusters, thereby sustaining OXPHOS and cellular energy metabolism
16,17,203. Structural and biophysical studies using NMR and small-angle X-ray scattering have demonstrated that GSH directly participates in Fe-S cluster assembly, including formation of mitochondrial Fe-S biosynthetic complexes essential for respiratory chain function
204,205. Through cooperation with Fe-S assembly factors such as glutaredoxin 5 (GLRX5), GSH supports Fe-S cluster maturation in mitochondria and their distribution to cytosolic and nuclear proteins
206,207. Because Fe-S clusters function as essential cofactors for components of the electron transport chain, TCA cycle enzymes, and genome maintenance machinery, their stability depends on a reducing intracellular environment that is maintained, in part, by GSH
203,208.
However, excessive accumulation of GSH precursors can overwhelm this reducing capacity, leading to reductive stress and aberrant mitochondrial H
2O
2 production, a mechanism that is particularly pronounced in cancer cells characterized by elevated cysteine metabolism but limited antioxidant enzyme capacity
209. This paradoxical vulnerability indicates that GSH-dependent redox homeostasis, while essential for normal cellular function, can become a critical liability in cancer cells with dysregulated GSH metabolism. Consistent with this notion, under conditions of cysteine starvation, the GSH-catabolizing enzyme CHAC1 mobilizes cysteine to sustain mitochondrial Fe-S cluster biogenesis and respiratory function. Paradoxically, preservation of OXPHOS under these conditions exacerbates ROS production and lipid peroxidation, ultimately promoting ferroptosis in cancer cells
210,211. In macrophages, impaired mitochondrial GSH import destabilizes Fe-S cluster-containing electron transport chain proteins and reduces LPS-induced inflammatory cytokine production. These effects can be partially reversed by restoring GSH levels, suggesting that GSH may contribute to cell death-associated responses in a context-dependent manner
212.
These findings suggest that although GSH-dependent reductive metabolism supports cellular survival under physiological conditions, sustained mitochondrial activity enabled by GSH can become deleterious under stress (Fig. 4). Given that cuproptosis preferentially targets OXPHOS-dependent cells by disrupting TCA cycle enzymes and Fe-S cluster integrity,
167,213 it remains an open question whether GSH-dependent Fe-S cluster biogenesis similarly licenses susceptibility to cuproptosis.
GSH-triggered prodrugs
Because intracellular GSH levels are typically elevated in tumor cells relative to normal tissues, this redox imbalance has been exploited as a tumor-associated trigger for prodrug activation and selective cytotoxicity
214. Engineered nanomedicine platforms incorporate GSH-responsive elements together with therapeutic cargos, enabling GSH-dependent drug release, amplification of ROS-mediated damage, and induction of regulated cell death, thereby improving tumor selectivity and therapeutic index (Fig. 4; Supplementary Table S1)
215. For example, glucose oxidase-engineered Cu
+ coordination polymer nanoplatforms remain inert under basal conditions but are selectively activated by high intracellular GSH. GSH-dependent activation releases Cu
+ to engage lipoylated DLAT and simultaneously unmasks glucose oxidase activity, inducing nutrient deprivation and triggering cuproptosis
216. Similarly, GSH-activatable synthetic ion channels undergo GSH-mediated disassembly to release channel-forming compounds, leading to caspase-dependent apoptosis
217.
Additional platforms exploit GSH-mediated cleavage of disulfide or polysulfide bonds to induce ferroptosis and pyroptosis. In one example, GSH-driven nanoparticle disintegration releases indomethacin to inhibit PTGS2 while reducing Pt
4+ to cytotoxic Pt
2+, thereby amplifying cisplatin-induced pyroptosis in pancreatic cancer
218. In another system, GSH cleaves tetrasulfide bonds in an organosilica shell to explosively release encapsulated liquid “ion drugs” (Na
+S
2O
82−). Concurrently, GSH reduces Fe
3+ to Fe
2+, which activates S
2O
82− to generate highly toxic sulfate (•SO
4−) and hydroxyl radicals, inducing coordinated ferroptosis and pyroptosis in lung cancer cells
219. The therapeutic efficacy of these GSH-responsive platforms depends on the extent to which intracellular GSH levels are sufficiently elevated and spatially restricted within tumor cells relative to surrounding normal tissues. A key translational challenge will be to define the threshold and heterogeneity of GSH abundance across tumor types and microenvironmental niches, as this may critically influence treatment selectivity and resistance.
Collectively, these OFF-to-ON switching strategies illustrate therapeutic approaches in which GSH acts as a biological trigger for selective activation of cytotoxic agents in the GSH-rich tumor microenvironment. Taken together, these findings highlight the context-dependent role of GSH in regulating ferroptosis, cuproptosis, and pyroptosis. Through its effects on metal availability, redox metabolism, and oxidative signaling, GSH can promote conditions that favor regulated cell death and may therefore be exploited for precision cancer therapy.
CONCLUSION AND OUTLOOK
This review highlights GSH as a context- and compartment-dependent regulator of regulated cell death rather than solely a cytoprotective metabolite. GSH primarily maintains cellular redox homeostasis, and its influence on cell fate depends on its abundance, subcellular compartmentalization (e.g., cytosolic versus mitochondrial pools), and the surrounding metabolic and enzymatic environment. Under most physiological conditions, GSH buffers oxidants in the cytoplasm and mitochondria and supports nuclear repair processes, thereby increasing the threshold for regulated cell death and, in certain pathological settings, contributing to therapy resistance. However, under defined pathophysiological stresses, including inflammatory cues, metal overload, cysteine deprivation, or drug exposure, GSH oxidation and thiol chemistry may become coupled to metal redox cycling and mitochondrial damage amplification, thereby increasing susceptibility to forms of regulated cell death such as ferroptosis and cuproptosis. This duality positions GSH as a context- and compartment-dependent determinant of cellular redox thresholds and raises several unresolved questions that will shape future directions in redox biology and cell death research.
A major open question concerns the spatial and temporal regulation of GSH. How distinct intracellular GSH pools are differentially controlled, how flux between compartments is coordinated, and how these dynamics influence commitment to apoptosis, ferroptosis, and cuproptosis remain incompletely understood. Equally unresolved is how redox thresholds are established — namely, when GSH-mediated buffering shifts from protective to permissive for cell death, and how changes in NADPH availability, metal speciation, or mitochondrial dependence reprogram this balance. Resolving compartmentalized GSH dynamics will require tools that integrate spatial targeting, quantitative measurement, and temporal resolution. Endogenous redox sensors (e.g., GRX1-roGFP2 or TRaQ-G) enable real-time, organelle-specific redox readouts but primarily report redox potential rather than absolute GSH levels, whereas fractionation coupled with mass spectrometry provides absolute GSH/GSSG quantification but remains static and susceptible to lysis-associated artifacts. Ratiometric fluorescent probes for imaging GSH in living cells are convenient but often limited by lower specificity and subcellular targeting. Addressing these challenges will be essential for defining predictive redox states that govern cell fate decisions.
Another critical knowledge gap lies in the integration of GSH metabolism with metal handling and mitochondrial function. Whether GSH-dependent copper trafficking and Fe-S cluster maintenance actively license ferroptosis or cuproptosis, rather than simply buffering stress, remains unclear. In parallel, how GSH interfaces with inflammatory signaling to coordinate pyroptosis and necroptosis in tissue contexts warrants further investigation. Tumor heterogeneity further complicates this landscape, as the extent of GSH dependency varies across tumor types, metabolic states, and microenvironments, with poorly defined consequences for therapeutic response.
In therapeutic settings, the effects of global GSH depletion or supplementation are likely to depend on disease context and cellular state. Instead, therapeutic modulation of GSH must be tailored to disease context, reflecting the fundamental “GSH paradox”: whereas cancer therapy often seeks to deplete elevated GSH to induce oxidative stress and cell death, neurodegenerative and inflammatory diseases may require restoration of GSH homeostasis. Accordingly, selective targeting of GSH-dependent nodes (such as cysteine transport, mitochondrial GSH import, compartmentalized GSH degradation, or NADPH-dependent regeneration) may offer greater specificity and efficacy than indiscriminate redox perturbation. At the same time, elevated tumor GSH can be exploited as a conditional trigger for prodrug activation, nanomedicine-based delivery, or metal-driven cytotoxicity, thereby converting a classical resistance factor into a therapeutically actionable vulnerability. However, successful clinical translation will require a deeper understanding of redox thresholds across cell types, subcellular compartments, and disease stages, as well as strategies to overcome compensatory antioxidant networks and systemic toxicity associated with non-specific redox modulation.
Together, these findings position GSH not only as a guardian of cellular integrity but also as a context-dependent determinant of cell death susceptibility and therapeutic vulnerability. Leveraging this duality provides a conceptual framework for identifying redox-defined liabilities and developing next-generation therapies that exploit metabolic and oxidative imbalances in cancer and other diseases.
The Author(s) 2026. Published by Higher Education Press. This is an Open Access article distributed under the terms of the CC BY license (https://creativecommons.org/licenses/by/4.0/).
