Glutathione (GSH): Structure, Enzymology, and How Its Redox State Is Measured

Glutathione is the odd one out in a peptide catalog. It is three amino acids long, it is not a signaling peptide, and none of the literature reviewed below describes it acting through a receptor. It is a small-molecule thiol that cells manufacture in bulk and consume continuously — the most abundant low-molecular-weight thiol compound synthesized in cells, and a molecule whose functions are described in terms of protecting against oxidative damage and the toxicity of xenobiotic electrophiles rather than in terms of binding anything ^[1]. In a laboratory, it is less a "compound with an effect" than a reagent whose concentration defines the chemical environment other experiments run in.

That distinction matters for how the research literature reads. Most of what is published about glutathione is not about adding it to something. It is about what happens to a system when its own glutathione is depleted, oxidized, mislocalized, or trapped in the wrong compartment. This guide follows that structure: what the molecule is, how cells build it, which enzymes use it, how its redox state is quantified, and where the honest gaps sit.

Identity

Glutathione is a tripeptide of glutamate, cysteine, and glycine. The identity figures below are the ones carried on the Echo Peptides product record for this compound:

• Chemical name: L-γ-glutamyl-L-cysteinylglycine
• CAS number: 70-18-8
• Molecular formula: C₁₀H₁₇N₃O₆S
• Molecular weight: 307.3 g/mol
• PubChem CID: 124886

The γ in the name is the whole structural story. The bond joining glutamate to cysteine is formed from glutamate's γ-carboxyl rather than the conventional α-carboxyl, and the review notes that the only enzyme able to hydrolyze this unusual bond is γ-glutamyltranspeptidase, which is present only on the external surfaces of certain cell types — so glutathione resists intracellular degradation ^[2]. Synthesis and breakdown are both handled by a dedicated set of enzymes usually described as the γ-glutamyl cycle, and that cycle can achieve intracellular glutathione concentrations of 1–10 mM ^[3].

The single cysteine carries the free thiol. Every reaction discussed below runs through that one sulfur atom. Reduced glutathione is written GSH; the disulfide-linked dimer formed when two molecules are oxidized together is GSSG.

How cells build it, and why cysteine is the constraint

Biosynthesis happens in the cytosol in two ATP-dependent steps. Glutamate–cysteine ligase (GCL) is rate-limiting and is built from a catalytic subunit, GCLC, and a modifier subunit, GCLM; glutathione synthetase adds the glycine ^[2]. The two determinants that reviewers repeatedly single out are the availability of the sulfur amino acid precursor, cysteine, and GCL activity itself ^[2].

Regulation is transcriptional as well as kinetic. GCL subunits and glutathione synthetase are controlled at multiple levels and often in a coordinated way, with NF-E2-related factor 2 (Nrf2) acting through the antioxidant response element, alongside AP-1 and NF-κB ^[2]. This is the practical reason a glutathione measurement taken hours after a stressor may reflect an induced synthetic capacity rather than the immediate redox insult.

One consequence of the γ-glutamyl cycle is less obvious. Because glutathione holds a glutamate and turns over rapidly, it functions as a glutamate pool. In HT22 hippocampal cells, PC12 cells, and primary cortical neurons, inhibiting the enzymes that liberate glutamate from glutathione — 5-oxoprolinase and γ-glutamyl transferase — lowered measured glutamate, while inhibiting γ-glutamylcysteine ligase produced substantial glutamate accumulation, and that accumulation temporally preceded the later oxidative effects ^[3]. The authors framed this as glutathione serving as a physiologic reservoir of neuronal glutamate. It is worth holding onto when interpreting any experiment that blocks glutathione synthesis in neural cultures: the immediate readout may be an amino acid pool shift, not oxidative stress.

Three enzyme families that cannot work without it

Peroxide reduction

Glutathione peroxidases detoxify peroxides in a reaction coupled to the oxidation of GSH to GSSG; glutathione reductase, using NADPH as cofactor, converts GSSG back to GSH ^[4]. That cycle is the reason glutathione is described as a cofactor rather than a consumable — under normal conditions it is regenerated, and the NADPH supply is as much a part of the system as the thiol.

Within the family, glutathione peroxidase 4 (GPx4) is the selenoperoxidase that handles phospholipid hydroperoxides, and it has been described as the cornerstone of antiperoxidant defense ^[5]. Its mechanistic importance became sharply visible through ferroptosis work: small molecules that induce ferroptosis were shown to inhibit GPX4 by covalently targeting the active-site selenocysteine, leading to accumulation of polyunsaturated fatty acid hydroperoxides ^[6]. The same study localized the chemistry precisely — peroxidation occurs at the bis-allylic position of polyunsaturated fatty acids, and pretreating cells with PUFAs carrying deuterium at that site prevented oxidation and blocked ferroptosis ^[6].

Electrophile conjugation

The glutathione transferases (GSTs) conjugate GSH onto electrophilic centers. Mammals carry three families — cytosolic, mitochondrial, and microsomal, the last designated MAPEG ^[7]. Their substrate range is broader than the word "detoxification" suggests: besides electrophilic xenobiotics such as chemical carcinogens and environmental pollutants, they inactivate endogenous α,β-unsaturated aldehydes, quinones, epoxides, and hydroperoxides formed as secondary metabolites during oxidative stress ^[7]. Several are also biosynthetic enzymes, involved in producing leukotrienes and prostaglandins and in tyrosine degradation ^[7].

The regulatory loop closes on itself: the promoters of cytosolic GST and MAPEG genes contain antioxidant response elements, through which they are transcriptionally activated during exposure to Michael reaction acceptors and oxidative stress ^[7]. The same element that drives GCL expression drives the enzymes that spend the product.

Dicarbonyl stress and cofactor regeneration

Two further roles are easy to overlook. Dicarbonyl stress induced by methylglyoxal and glyoxal is alleviated by the glyoxalase enzymes together with GSH, and GSH can regenerate vitamin E after vitamin E has reacted with lipid peroxyl radicals ^[4]. Both mean glutathione depletion registers in assays that have nothing obvious to do with peroxides.

The GSH/GSSG couple is a number, not an adjective

"Redox state" is used loosely enough that a 2001 review set out to give it a quantitative definition, and that framework is still the reference point ^[8]. The reduction potential of a redox couple can be estimated with the Nernst equation, and both pH and the concentrations of the species involved shift it ^[8]. Because GSSG/2GSH is the most abundant redox couple in a cell, its half-cell reduction potential serves as an indicator of the broader redox environment ^[8].

The authors reported that changes in that half-cell potential appear to correlate with cellular status — approximately −240 mV in proliferation, −200 mV in differentiation, and −170 mV in apoptosis ^[8]. The hedge in the original is doing real work: this is a reported correlation across systems, not a threshold that causes a fate. The practical point for bench work is that the GSH/GSSG ratio alone is insufficient; the absolute concentrations and the pH have to be known to convert a ratio into a potential ^[8].

Methods for assessing glutathione status in cells are themselves a literature, treated alongside the biochemistry in general overviews of the molecule ^[1]. It is worth reading that material before trusting a single ratio, because the thiol that makes glutathione useful also makes it liable to oxidize during sample handling.

S-glutathionylation: glutathione as a modification, not just a sink

Glutathione also regulates redox signaling directly, through reversible oxidation of critical protein cysteine residues to form mixed disulfides — S-glutathionylation ^[4]. This is not the same as a protein being damaged by oxidation; it is a reversible post-translational modification with its own enzymology.

Glutaredoxins are the small oxidoreductases of the thioredoxin family that run those reactions. They catalyze glutathionylation, deglutathionylation, and disulfide reduction by two distinct mechanisms: a dithiol mechanism that uses both active-site cysteines to reduce disulfides, and a monothiol mechanism that uses only the N-terminal active-site cysteine to reduce GSH mixed disulfides ^[9]. Glutathione participates here in more forms than GSH and GSSG — S-nitrosoglutathione is part of the same signaling vocabulary ^[9].

The same review is candid about the limit: because of limited methodological tools, identifying new substrates for these oxidoreductases remains one of the biggest challenges in the field ^[9]. Any claim that a particular protein is "regulated by glutathionylation" should be checked against whether the substrate was actually demonstrated or inferred.

Compartments: a whole-cell measurement can be blind

Glutathione is synthesized exclusively in the cytosol, yet mitochondria — the major site of oxidative reactions — need their own supply ^[10]. The transporter responsible went unidentified for decades. Organellar proteomics and metabolomics identified SLC25A39, a mitochondrial membrane carrier previously of unknown function, as a regulator of glutathione transport into mitochondria ^[10].

The most useful detail for anyone designing an assay is this: loss of SLC25A39 reduced mitochondrial glutathione import and abundance without affecting cellular glutathione levels ^[10]. A total-cell glutathione readout would have shown nothing. Cells lacking both SLC25A39 and its paralogue SLC25A40 showed defects in the activity and stability of iron–sulfur cluster proteins, and expressing an engineered bacterial glutathione biosynthetic enzyme inside mitochondria restored mitochondrial production and ameliorated the metabolic and proliferative defects ^[10].

The transporter is itself under redox control. Under physiological conditions SLC25A39 is rapidly degraded by the mitochondrial protease AFG3L2; glutathione depletion causes AFG3L2 to dissociate from SLC25A39, producing a compensatory increase in mitochondrial glutathione uptake ^[11]. Genetic and proteomic analyses identified a putative iron–sulfur cluster in the matrix-facing loop of SLC25A39 as essential for this regulation, coupling mitochondrial iron homeostasis to glutathione import ^[11]. Note the word "putative" — the authors did not claim a solved structure for that site.

Depletion is the standard experiment

Because glutathione is made in place, the informative manipulation is usually subtraction. One study in cultured ARPE-19 retinal pigment epithelial cells used three mechanistically distinct depletion routes — incubation in cystine-free medium, buthionine sulfoximine to block de novo synthesis, and erastin to inhibit the cystine/glutamate antiporter system xc− ^[12]. All three decreased viability and increased both soluble and lipid reactive oxygen species, while leaving mitochondrial ROS and mitochondrial mass unaffected, and GPX4 expression fell ^[12].

The downstream readouts split three ways: autophagy increased (LC3 expression, autophagic vacuoles, autophagic flux), senescence markers appeared (senescence-associated β-galactosidase positivity, senescence-associated heterochromatin foci, G1 arrest), and the cell death was prevented by selective ferroptosis inhibitors and by an iron chelator ^[12]. The authors wrote that their findings suggest glutathione depletion induces ferroptosis, autophagy, and stress-induced premature senescence, and reported that autophagy is activated during ferroptosis and acts to reduce senescence ^[12]. That is one cell line and one laboratory's interpretation; the mechanistic split between the three outcomes is not settled biology.

The theoretical frame around those results is worth reading directly. Lipid peroxidation is under homeostatic metabolic control, and the authors write that ferroptosis takes place once a threshold has been exceeded — which occurs when three major conditions are satisfied: oxygen metabolism leading to the continuous formation of traces of lipid hydroperoxides from phospholipid-containing polyunsaturated fatty acids, missed enzymatic reduction of those hydroperoxides, and availability of ferrous iron from the labile iron pool ^[5]. The authors are explicit that the effectors driving ferroptosis under physiological conditions are not known, and that their proposal — that aerobic metabolism itself is the critical requisite — is an inference from the enzymology, not a demonstrated mechanism ^[5].

Adding glutathione outside a cell is not the same as raising it inside

This is the point most often skipped. Extracellular glutathione is degraded by a cell-surface enzyme, γ-glutamyltranspeptidase, and the products are transported into cells ^[13]. What arrives inside is not the intact tripeptide.

A cell-culture study made the consequences concrete. Human lung carcinoma A549 cells in glutamine-deficient medium did not proliferate unless glutamine was supplemented, but extracellular glutathione provoked partial proliferation, an effect that correlated with high γ-glutamyltranspeptidase activity and increased intracellular glutamic acid ^[13]. Glutamate alone reproduced the effect; cysteine did not, and neither cysteinylglycine nor the cysteine-delivery compound oxothiazolidine-4-carboxylate stimulated growth ^[13]. Buthionine sulfoximine, which blocks glutathione synthesis, actually enhanced the effect, which the authors read as evidence that increased cellular glutathione synthesis does not correlate with the growth stimulation ^[13].

The authors' conclusion was scoped to that system: in these cells, the glutamate rather than the cysteine released from glutathione was responsible for the proliferative effect ^[13]. It should not be generalized into a rule about all cell types, and the reverse claim — that extracellular glutathione is a cysteine delivery vehicle — is not what this experiment showed. Cell lines differ enormously in γ-glutamyltranspeptidase expression, and that single variable can invert the result.

What is genuinely unresolved

Several things stay open. Substrate identification for glutaredoxins is method-limited, so the catalog of glutathionylated targets is incomplete by the field's own admission ^[9]. The effectors that tip lipid peroxidation into ferroptosis under physiological conditions are unidentified ^[5]. Compartment-specific glutathione pools were invisible to standard assays until the transporter work of the last few years, and the regulatory logic connecting iron status to glutathione import is described with deliberately provisional language ^{[10, 11]}. And the synthesis literature reports increasing evidence that dysregulation of glutathione synthesis contributes to the pathogenesis of many pathological conditions ^[2] — a statement about disease models, not about anything a research chemical does, and one that describes contribution rather than cause.

What is not in question is the molecule's centrality: reduced glutathione acts directly against free radicals and pro-oxidants and serves as cofactor for glutathione peroxidases, glutathione transferases, and glyoxalases, and is additionally involved in protein folding, protection of protein thiols from oxidation and crosslinking, cell cycle regulation, ascorbate metabolism, apoptosis, and ferroptosis ^[4]. That breadth is exactly why a change in glutathione status perturbs so many assays at once, and why attributing an observed effect to glutathione requires ruling out several parallel routes rather than one.

Handling notes

The free thiol is the reactive group and the vulnerable one. Echo Peptides supplies glutathione as a lyophilized powder specified for refrigerated storage at 2–8 °C, water-soluble, with a certificate of analysis available. In solution, thiol oxidation is the failure mode to design against — which is also why glutathione measurements are sensitive to sample workup, and why methodology sections in this literature spend disproportionate space on how status was assessed ^[1].

This material is supplied for laboratory research use only. It is not for human or veterinary use, and nothing above describes an administration route, schedule, or quantity for any organism.