Every Glutathione Question Answered, in 16 Studies

16 referenced articles

• What Is Glutathione? The Bond That Makes It Unusual4 sources
• Kidney and Intestine, and Why Position Matters3 sources
• Four Jobs, Only One of Which the Label Describes4 sources
• An Enzyme Family That Uses Glutathione as a Reagent3 sources
• The Clearest Demonstration That Cysteine Is Limiting3 sources
• A Routine Liver Test That Measures a Glutathione Enzyme3 sources
• Glutathione Proposed as a Carrier, Not Just a Buffer3 sources
• Glutathione Regulatory Status4 sources
• Glutathione in the Published Literature4 sources
• Glutathione Storage and the Oxidation Problem3 sources
• Glutathione and Skin Research3 sources
• Does Oral Glutathione Reach the Body?3 sources
• How Glutathione Is Built3 sources
• GSH and GSSG: What the Ratio Measures3 sources
• The Gamma-Glutamyl Bond3 sources
• Is Glutathione a Peptide? Yes, and an Unusual One3 sources

Open research questions

• Why can no ribosome build glutathione?
• What is a gamma-glutamyl bond and why does it matter?
• Does oral glutathione arrive intact or as its amino acids?
• Why does a purity figure not describe glutathione's redox state?
• How does the GSH/GSSG ratio measure oxidative stress?
• What did the 2025 systematic reviews on skin actually examine?

Deep dive: the bond that puts a peptide outside peptide biology

Glutamate is one of only two amino acids carrying two carboxyl groups - the backbone alpha-carboxyl every residue has, plus one on its side chain. Standard peptide bonds use the alpha. Glutathione uses the gamma, and that one choice cascades. Ribosomes have exactly one chemistry, in which an incoming residue's amine attacks the growing chain's alpha-carboxyl, and no mechanism whatsoever for recruiting a side chain. So glutathione cannot be a gene product. It is assembled instead by two ATP-dependent ligases, which means the genome encodes the machinery but never the molecule - a peptide present in nearly every cell of nearly every organism, with no coding sequence anywhere. The same geometry that excludes the ribosome also excludes most peptidases, whose active sites are built around the spacing of an alpha bond. Only gamma-glutamyl transpeptidase cleaves it, which puts turnover of a millimolar-concentration metabolite under the control of a single enzyme. Protease resistance by structural mismatch is more complete than anything proline achieves in a conventional peptide.

Deep dive: the one compound here where a purity figure does not tell you what you need

Every storage article on this site says disulfide chemistry is inapplicable, because KPV, Selank, TB-500, DSIP and Semax contain no cysteine at all. Glutathione is the compound those statements were implicitly excluding, and the exception is not marginal - its thiol is simultaneously the source of its function and its principal vulnerability. Two thiols meet, lose two hydrogens, and become GSSG at 612.6 Da. Oxygen drives it, trace metals catalyse it, no enzyme is required, and it proceeds in a vial left standing. The subtle part is that GSSG is not an impurity in the ordinary sense. It is correctly assembled glutathione in a different oxidation state, and a purity assay may well score it as related material rather than contamination. A preparation can be 99% pure and substantially oxidised at once. Where an experiment depends on the reduced form, the certificate does not answer the question - chromatography separating 307.33 from 612.6, or a thiol-specific assay, does.

Deep dive: the same question NAD+ raises, with better evidence and a less obvious answer

Both categories on this site face one structural question: does supplying the finished molecule work, or does it succeed only by being degraded to something the cell can actually use? For NAD+ the answer is fairly clearly the latter - 663 Da with two negative charges cannot cross a membrane, and CD38 degrades it outside the cell. For glutathione it is genuinely open, and the evidence is better. Richie and colleagues published a randomised controlled trial on body stores in the European Journal of Nutrition in 2015, reporting increases. But an increase in stores admits two readings: intact absorption and distribution, or degradation to glutamate, cysteine and glycine followed by resynthesis inside cells - in which case the useful contribution is essentially the cysteine, and the tripeptide is an expensive delivery vehicle for it. Since cysteine availability is what normally limits synthesis, and since gamma-glutamyl transpeptidase sits on intestinal surfaces waiting for exactly this substrate, the second reading is not a sceptical stretch. A store measurement alone cannot distinguish them.

Handling checklist

• Verify against CID 124886, 307.33 Da, C10H17N3O6S
• Check the oxidised form separately - GSSG is CID 65359 at 612.6 Da
• Do not treat a purity figure as a statement about redox state
• Store lyophilised, cold, dry; minimise headspace air
• Prepare solutions fresh - thiol oxidation proceeds without any enzyme
• Where the reduced form matters, assay free thiol rather than assuming

Common mistakes

Assuming a high purity figure means the material is reduced

GSSG is correctly assembled glutathione in a different oxidation state. A purity assay may score it as related material, not contamination.

Treating glutathione like the other peptides on this site

Its gamma bond makes it protease-resistant and non-ribosomal, and it is the only compound here with a reactive thiol. Most generalisations do not apply.

Reading increased body stores as proof of intact absorption

Degradation to amino acids followed by intracellular resynthesis produces the same measurement. The trial endpoint cannot distinguish them.

Citing the large biochemistry literature as evidence about supplementation

What glutathione does inside cells is settled. What supplementing it accomplishes is a separate and contested question.

Repeating systematic review subject matter as a product claim

Describing what a literature examined and claiming a product does it are different acts. Only the first is permissible.