Everything below reports findings from cell cultures, isolated tissue and animals. This material is for research use only, not for human or veterinary use.
Glutathione is the most abundant thiol inside a cell. It is glutamate, cysteine and glycine. Calling it a tripeptide is accurate and slightly misleading.
One unusual gamma bond explains the metabolism, the turnover and the analytical difficulties at once. Glutamate links through its side-chain carboxyl rather than its alpha-carboxyl. Ordinary peptidases cannot cleave that gamma bond. One enzyme can.
Indexed oral and cosmetic papers exist. Those human milligram figures and endpoints sit outside this profile. The laboratory questions are identity, the redox couple, and how to keep the free thiol free.
Chemical identity: a tripeptide with an unusual bond
| Property | Value |
|---|---|
| Systematic name | (2S)-2-amino-5-[[(2R)-1-(carboxymethylamino)-1-oxo-3-sulfanylpropan-2-yl]amino]-5-oxopentanoic acid |
| Common names | Glutathione, GSH, reduced glutathione |
| Sequence | gamma-L-glutamyl-L-cysteinyl-glycine |
| Molecular formula | C10H17N3O6S |
| Molecular weight | 307.33 g/mol |
| Exact mass | 307.08381 Da |
| CAS number | 70-18-8 |
| PubChem CID | 124886 |
| InChIKey | RWSXRVCMGQZWBV-WDSKDSINSA-N |
| Reactive group | Free thiol on cysteine |
| Oxidised form | Glutathione disulfide (GSSG), 612.6 g/mol |
| Intracellular concentration | Millimolar in most cell types |
| XLogP | -4.5 |
The gamma-glutamyl bond
In an ordinary peptide, each residue links through its alpha-carboxyl. In this tripeptide, glutamate links through its side-chain carboxyl instead. That is the gamma bond. It is the single most consequential feature of glutathione.
Ordinary peptidases cannot cleave it. One enzyme can. Gamma-glutamyl transpeptidase sits on the outer surface of cell membranes. It is abundant in intestine, kidney and liver preparations.
Two things follow. The molecule survives inside cells where ordinary tripeptides would be degraded, so concentrations reach millimolar. It is also taken apart at the tissues that express that one enzyme. Anyone who writes an oral-intake protocol is arguing with that enzyme, not with a slogan.
Why the concentration is so high
Millimolar is an unusual concentration for anything that is not a bulk metabolite. Most signalling molecules work at nanomolar. Most peptides never accumulate at all.
Two features make it possible. The gamma linkage blocks the peptidases that would otherwise degrade it. Cells also make it continuously in two ATP-dependent steps rather than importing it.
The pool then functions as a buffer rather than a signal. A buffer has to be present in large excess over the thing it absorbs. That is how cells use this thiol against electrophiles and peroxides.
Reduced and oxidised forms are not interchangeable
The cysteine thiol is the working group. Two reduced molecules oxidise to one disulfide. Glutathione reductase reduces that disulfide back using NADPH.
Almost the entire intracellular pool sits in the reduced form. The ratio between the two reports the redox state of a compartment better than the absolute amount does. Studies use it that way [11][12].
A lot sold as the reduced form that already holds a large disulfide fraction is a different reagent. Mass barely moves. Function does.
What glutathione does in a cell
Three distinct jobs, often conflated into one word.
The redox buffer
The GSH/GSSG couple is the principal redox buffer of the cytosol. Pool size is large. Reduced form dominates the ratio. That holds the compartment at a defined potential and resists change.
Masarik and colleagues showed the diagnostic use directly in prostate cell lines (PMID 22592803) [12]. Non-tumour cells held a GSH:GSSG ratio above 1 under zinc exposure. Tumour cells held a predominance of the oxidised form at every concentration tested.
Farag and colleagues used rabbit erythrocytes as a cell system for the same couple (PMID 27314888) [11]. They read antioxidant activity of other small molecules against a cytotoxic insult. The thiol pool was the handle, not the headline.
Conjugation and export
S-transferases attach that thiol to electrophiles. Conjugates become more water-soluble and exportable. This is a major route by which cells dispose of reactive compounds. It applies to research chemicals as readily as to endogenous species.
Cui and colleagues showed how visible this is in practice (PMID 35104766) [7]. Ponicidin bound covalently to GSH inside pancreatic cancer cells, forming a conjugate. That depleted the free pool and reduced peroxidase 4 activity.
Glutathione peroxidase and ferroptosis
Peroxidase 4 uses GSH to reduce lipid peroxides. Losing that capacity triggers ferroptosis. That death programme is iron-dependent and driven by lipid peroxidation.
That link is now a standard experimental handle. In the ponicidin work, iron, malondialdehyde and reactive oxygen species all rose. Ferroptosis inhibitors blocked the effect on proliferation [7].
The tripeptide is therefore not simply protective. It is a node whose depletion routes cells into a specific death programme.
The GSH/GSSG couple as a measurement
A recurring problem runs through this literature. Two different quantities get reported under one name.
Content is not ratio
Total content answers how much sits there. The GSH:GSSG ratio answers what redox state the compartment sits in. They move independently. An intervention can raise one while leaving the other unchanged.
Masarik reported both sides of that split in cell lines [12]. Farag used the couple as a readout rather than as a supplement [11]. Anyone who writes “raised glutathione” without saying which quantity has not finished the sentence.
Why the ratio is easy to fake
The free thiol autoxidises to the disulfide on contact with air. Trace metals catalyse it. Neutral pH accelerates it. A sample handled without acidification and metal chelation reports a falsely low reduced fraction.
The GSH:GSSG ratio is the measurement most sensitive to this. A paper that reports a dramatic oxidised shift and never mentions chelation has a handling problem until proven otherwise.
Two papers that already sit in the oral file
Richie and colleagues and Solnier and colleagues both reported content and ratio in human matrices [2][13]. Those human figures sit outside this profile. The method lesson still travels. Measure both quantities. State the matrix. State how the thiol was trapped.
Synthesis is limited by cysteine
The tripeptide is made in two ATP-dependent steps. Glutamate and cysteine form gamma-glutamylcysteine. Then glycine is added.
The first step is rate-limiting. Cysteine availability sets that limit. This is why cysteine donors are studied as a route to raising the pool where the tripeptide itself is not the input.
N-acetylcysteine as a precursor, not a substitute
Nasr and Perl set out the reasoning for N-acetylcysteine (PMID 36312742) [8]. Cysteine is the rate-limiting constituent. NAC stands in for it.
Their disease-population paper is an indexed file. Human endpoints sit outside this profile. The chemistry is the part that a laboratory can use. If precursor supply is the constraint, a cysteine donor is the tool. If the synthetic enzymes are the constraint, it is not.
Related precursor chemistry sits at NAC.
When the synthetic enzymes fail
Banerjee and colleagues showed the upstream machinery failing in a disease model (PMID 37371062) [9]. Astrocytes carrying the LRRK2-I1371V variant held less of the tripeptide. They expressed less of the enzymes that make it. They showed lower Nrf2. Conversion of glutamate to gamma-glutamylcysteine was impaired.
Precursor supply, transcriptional control through Nrf2, and the synthetic enzymes all sit upstream of the pool. Any of them can be the constraint.
Tsujimoto and colleagues asked a related precursor question with L-glutamine, which supplies glutamate [5]. That paper is an indexed human file. Its survival endpoints sit outside this profile. The chemical claim is simpler. Glutamate is a constituent. Changing glutamate supply is not the same experiment as adding the finished tripeptide.
Indexed human papers
A long oral-intake and cosmetic literature exists around this molecule. Those human milligram figures and endpoints sit outside this profile. The papers stay in the reference list so a reader can find them.
What those files are
Witschi and colleagues asked whether intact glutathione appears in plasma after an oral intake (PMID 1362956) [1]. Richie and colleagues asked whether repeated oral intake changes tissue stores over months (PMID 24791752) [2]. Solnier and colleagues compared formulations in a crossover design [13].
Sonthalia and colleagues, Davids and colleagues, Juhasz and Levin, and Sarkar and colleagues published reviews in the cosmetic file [3][4] and [6][10]. Tsujimoto and colleagues published a glutamine paper in a chemoradiotherapy cohort [5].
Cite those papers for the fact that a file exists. Do not import their human endpoints into a laboratory protocol.
What this profile will not quote
It will not quote oral milligram ladders, plasma-rise percentages, melanin-index scores, or intravenous use claims. Those sentences turn a research article into a use document.
The gamma-glutamyl transpeptidase mechanism is still the right chemical explanation for why intact molecule is hard to deliver past a membrane that expresses that enzyme [1]. That is a fact about the bond. It is not a dosing instruction.
Tyrosinase is a biochemical interaction, not a use case
Reviews in the cosmetic file discuss tyrosinase and a shift between eumelanin and phaeomelanin [3][10]. Treat that as an enzyme claim that needs a stated assay. Do not treat it as a laboratory purpose for the vial.
| Paper | What the file is | What this profile uses |
|---|---|---|
| Witschi 1992 [1] | Single-intake plasma paper | Indexed file; GGT chemistry |
| Richie 2015 [2] | Months-long store paper | Indexed file; measure content and ratio |
| Solnier 2026 [13] | Formulation crossover | Indexed file |
| Sonthalia 2016 [3] | Cosmetic review | Indexed file |
| Davids 2016 [4] | Intravenous review | Indexed file |
| Juhasz 2018 [6] | Systemic-agent review | Indexed file |
| Sarkar 2025 [10] | Cosmetic systematic review | Indexed file |
| Tsujimoto 2023 [5] | Glutamine cohort paper | Indexed file; glutamate is a constituent |
Where the compound is used as a reagent
Outside the intake literature, this molecule is a workhorse reagent. Those uses are better characterised than the consumer ones.
Refolding buffers
It serves as the reducing agent in protein refolding buffers. Laboratories usually pair it with its own disulfide so the ratio sets a defined redox potential. Each of those applications depends on the free thiol being genuinely free.
A partly oxidised lot behaves as a weaker reducing agent. The failure looks like a protocol problem rather than a reagent problem. That is the practical reason the oxidised fraction belongs on the certificate.
GST-tag purification and S-transferase assays
It is the substrate in glutathione S-transferase assays. It is also the affinity handle in GST-tag purification. The fusion protein binds immobilised tripeptide and elutes in free tripeptide.
A lot that has already become disulfide is a poorer eluent. Run Ellman’s reagent on a fresh solution if the elution looks weak and the column looks fine.
Other redox reagents in the same drawer
Laboratories often set this thiol beside N-acetylcysteine as the cysteine donor upstream of it. They set it beside NAD+ and CoQ10 where a different redox pool is the comparator. Related write-ups sit at NAC, NAD and CoQ10.
Those three molecules are not substitutes. NAC is a precursor. NAD is a nicotinamide couple. CoQ10 is a quinone. Keep the labels apart.
Why the pool size is hard to interpret
Two different quantities get reported under one name. Total content and the GSH:GSSG ratio move independently.
Richie reported both in a human file [2]. Solnier reported both in another [13]. Much of the older literature reports only one. When reading any claim about raising the pool, the useful question is which quantity was measured, and whether the other was measured at all.
A larger pool with an unchanged ratio may report a different state than a smaller pool with a better ratio. Cell-line work already shows that split [12].
Tumour cells frequently hold a large pool. That pool contributes to resistance against oxidative insults in culture [7]. Depleting it is an experimental strategy in that setting, the opposite of adding it. Do not mix those two protocols on one bench without labelling the vials.
Verifying research material
The compound is unusually easy to mis-measure. The reason is chemical rather than procedural.
Identity
Formula C10H17N3O6S, molecular weight 307.33, CAS 70-18-8, InChIKey RWSXRVCMGQZWBV-WDSKDSINSA-N. Exact mass 307.08381 Da. The disulfide is 612.6 g/mol.
Stereochemistry matters here. The natural form is L-glutamyl, L-cysteinyl. Specific rotation and melting point both give identity checks.
PubChem records the condensed sequence as H-gGlu-Cys-Gly-OH (CID 124886). A lot that is the alpha-linked isomer is a different molecule. Ordinary peptidases would eat that isomer. They do not eat this one. Put the reduced mass and the disulfide mass on the same notebook line. A later reader should not have to guess which species was in the vial.
The thiol check
A purity assessment should report GSSG content explicitly, not just a single percentage. Material that has partly oxidised in storage is unchanged by mass and altered by function.
HPLC with thiol-specific derivatisation separates the reduced and oxidised forms. Ellman’s reagent gives a fast colorimetric check on free thiol content. It cannot tell one thiol from another.
Intact mass still helps. Reduced glutathione is 307.33 g/mol. The disulfide is 612.6 g/mol. A spectrum that shows 611 rather than 307 is telling you the lot has already done the one reaction that destroys the reagent. That dimer still contains two tripeptides. It does not donate a free thiol. Ellman’s reagent will read it as dead.
The single sulfur in the parent gives a useful isotope shoulder. The disulfide doubles that contribution. A quick look at the M+2 peak can flag a mixed lot before chromatography. Name both species on the notebook line. Writing only “glutathione” when the vial holds a mix is how later readers lose the experiment.
Kimera publishes third-party certificates of analysis for every lot in its COA database.
Handling of the solid
Store the solid cold, dry, dark and sealed. Weighing from a jar that has been opened repeatedly in humid air introduces an error in the same direction every time.
Storage guidance is a house recommendation. Analytical documentation is per-lot release testing.
Lyophilised material picks up water readily. The compound is acidic in solution, since it carries two carboxyl groups against one amine.
Related chemistry appears in the redox and cofactors category.
Handling and stability in solution
Solid material is stable enough for ordinary storage. Solutions are not.
What oxidises the thiol
In water at neutral pH the thiol oxidises within hours. Dissolved oxygen, warmth, or trace copper and iron all speed that reaction. Prepare solutions fresh. Degas the buffer if the work depends on the reduced fraction. Add a chelator where trace metals are plausible.
A concentrated aqueous solution sits near pH 3. Neutralising it accelerates the oxidation the low pH was suppressing. That trade-off is worth deciding deliberately rather than discovering.
What a certificate should show
GSSG content. Peptide or compound content, not only chromatographic area percent. The stereochemical form. A date on the thiol assay, because that number moves in an opened jar.
A certificate that reports only “97 percent” has not said whether the other 3 percent is disulfide, water, or a different peptide.
Common questions about glutathione
Why does the gamma bond matter? It makes the molecule resistant to ordinary peptidases. That is why intracellular concentrations reach millimolar. Only gamma-glutamyl transpeptidase cleaves it. That enzyme sits on membranes that an oral intake has to pass [1].
Is the intact molecule easy to deliver past GGT? Witschi and colleagues asked that in a plasma paper [1]. Human plasma figures sit outside this profile. The chemical answer is that the enzyme is in the way.
Is more always better? No. Tumour cells use a large pool to resist oxidative insults, and depletion is an experimental strategy in that setting [7].
Is NAC a substitute? It supplies cysteine, the rate-limiting precursor, rather than the tripeptide itself [8]. Which is preferable depends on whether precursor supply or synthetic capacity is the constraint [9].
What is the most common analytical error? Letting the sample oxidise before measurement. The free thiol converts to the disulfide readily, corrupting the ratio toward apparent oxidative stress.
Does the oxidised form have its own uses? Yes. The disulfide plus the reduced form together set a defined redox potential in a refolding buffer. It is a distinct reagent, not spoiled material. A lot sold as the reduced form should not contain much of it.
Do indexed oral and cosmetic papers exist? Yes. Their human endpoints sit outside this profile.
Summary of the evidence
Identity: gamma-L-glutamyl-L-cysteinyl-glycine, C10H17N3O6S, 307.33 g/mol, CAS 70-18-8. The gamma linkage makes it resistant to ordinary peptidases.
Function: principal cytosolic redox buffer, conjugation substrate for glutathione S-transferases, and cofactor for glutathione peroxidase 4, whose loss triggers ferroptosis [7]. Cell-line work reads the GSH:GSSG ratio as a compartment state [11][12].
Synthesis: two ATP-dependent steps, rate-limited by cysteine availability [8], with Nrf2-controlled enzyme expression upstream [9].
Indexed oral and cosmetic papers exist. Their human milligram figures and endpoints sit outside this profile.
Analytical: the reduced form autoxidises readily, so the GSH:GSSG ratio is the measurement most vulnerable to handling artefact [11][12].
Status: supplied for laboratory research use only.
References
- Witschi A, Reddy S, Stofer B, Lauterburg BH. The systemic availability of oral glutathione. Eur J Clin Pharmacol. 1992;43(6):667-669. PMID 1362956. DOI
- Richie JP, Nichenametla S, Neidig W, et al. Randomized controlled trial of oral glutathione supplementation on body stores of glutathione. Eur J Nutr. 2015;54(2):251-263. PMID 24791752. DOI
- Sonthalia S, Daulatabad D, Sarkar R. Glutathione as a skin whitening agent: facts, myths, evidence and controversies. Indian J Dermatol Venereol Leprol. 2016;82(3):262-272. PMID 27088927. DOI
- Davids LM, Van Wyk JC, Khumalo NP. Intravenous glutathione for skin lightening: inadequate safety data. S Afr Med J. 2016;106(8):782-786. PMID 27499402. DOI
- Tsujimoto T, Wasa M, Inohara H, Ito T. L-glutamine and survival of patients with locally advanced head and neck cancer receiving chemoradiotherapy. Nutrients. 2023;15(19):4117. PMID 37836400. DOI
- Juhasz MLW, Levin MK. The role of systemic treatments for skin lightening. J Cosmet Dermatol. 2018;17(6):1144-1157. PMID 30133125. DOI
- Cui W, Zhang J, Wu D, et al. Ponicidin suppresses pancreatic cancer growth by inducing ferroptosis: insight gained by mass spectrometry-based metabolomics. Phytomedicine. 2022;98:153943. PMID 35104766. DOI
- Nasr S, Perl A. Principles behind SLE treatment with N-acetylcysteine. Immunometabolism (Cobham). 2022;4(4):e00010. PMID 36312742. DOI
- Banerjee R, Raj A, Potdar C, et al. Astrocytes differentiated from LRRK2-I1371V Parkinson’s-disease-induced pluripotent stem cells exhibit cell-intrinsic dysfunction in glutamate uptake and metabolism, ATP generation, and Nrf2-mediated glutathione machinery. Cells. 2023;12(12):1592. PMID 37371062. DOI
- Sarkar R, Yadav V, Yadav T, Janaani P, Mandal I. Glutathione as a skin-lightening agent and in melasma: a systematic review. Int J Dermatol. 2025;64(6):992-1004. PMID 39444151. DOI
- Farag MR, Alagawany M, Tufarelli V. In vitro antioxidant activities of resveratrol, cinnamaldehyde and their synergistic effect against cyadox-induced cytotoxicity in rabbit erythrocytes. Drug Chem Toxicol. 2017;40(2):196-205. PMID 27314888. DOI
- Masarik M, Gumulec J, Hlavna M, et al. Monitoring of the prostate tumour cells redox state and real-time proliferation by novel biophysical techniques and fluorescent staining. Integr Biol (Camb). 2012;4(6):672-684. PMID 22592803. DOI
- Solnier J, Du M, Zhang Y, et al. A targeted metabolomic assessment of oral glutathione bioavailability and safety in humans: a randomized crossover clinical trial. Antioxidants (Basel). 2026;15(3):354. PMID 41897500. DOI
Glutathione is sold for laboratory research use only. Not for human consumption, nor medical, veterinary, or household uses.
Literature retrieved from PubMed.
