AHK‑Cu One methyl group from a famous molecule, and what the evidence for it actually is
AHK‑Cu was not discovered. It was chosen — picked off a bench in 1994 as the winner of a deliberate screen, by a company whose original drug was heading for failure and which needed a second product. The molecule it was tested against was its own near‑twin: the famous copper peptide GHK‑Cu, from which it differs by a single methyl group. Thirty years later AHK‑Cu is on sale worldwide, and the whole of its primary peer‑reviewed literature is one paper. This monograph reads that paper, reads the patents that preceded it, and reads what the market has since made of both.
Findings are labelled by the kind of study that produced them, in the sentence that reports them. A result in a dish of cultured cells is called that. A result in a mouse is called that. A result that appears only in a patent application, and has never been through peer review, is called that too — and on this compound a good deal of the evidence is of exactly that kind. Where a number appears, the species, the concentration and the duration travel with it.
Five molecules appear in these pages and they are not interchangeable. AHK‑Cu is the subject: the tripeptide L‑alanyl‑L‑histidyl‑L‑lysine, written Ala‑His‑Lys or AHK, joined to a copper(II) ion. GHK‑Cu is its sibling and the reason most readers have heard of copper peptides at all; it carries glycine where this compound carries alanine, and it has roughly seventy times the published literature. AHK without copper is a different preparation and behaves differently. Vitamin C‑linker‑conjugated AHK is a fourth substance, tested once, in muscle‑derived cells, with no copper in it at all. Palmitoyl‑AHK is a fifth, sold alongside the others. Every finding below names the molecule that was actually studied, and where a claim about AHK‑Cu rests on work done with a relative, the sentence says so.
Concentrations and doses appear only as reported experimental parameters. Nothing in this document is a recommendation, and no route or schedule for human use is described or endorsed anywhere in it.
Section 01Fourteen daltons
Two molecules sit at the centre of this document, and the difference between them is one carbon atom and two hydrogens.
The first is glycyl‑L‑histidyl‑L‑lysine, universally abbreviated GHK. It is a tripeptide — three amino acids joined in a row — with the formula C14H24N6O4 and a mass of 340.38 daltons. It occurs naturally in human plasma. It was isolated from human serum in the early 1970s, it declines with age, and it has been the subject of several hundred papers.
The second is L‑alanyl‑L‑histidyl‑L‑lysine, abbreviated AHK. Its formula is C15H26N6O4 and its mass is 354.41 daltons. It does not occur naturally in any human tissue that anyone has reported. It was made in a laboratory.
Subtract one from the other and the difference is CH2 — a single methylene group, 14.03 daltons. The first residue of GHK is glycine, the smallest amino acid, whose side chain is a lone hydrogen atom. Replace that hydrogen with a methyl group and glycine becomes alanine. That is the entire chemical distance between the most famous copper peptide in cosmetic science and the subject of this monograph.
The reason a chemist would bother to make that substitution is not obvious from the structures, and it is the question Part Two answers. But it is worth noticing at the outset what the substitution does not change. It does not touch the histidine in the middle. It does not touch the lysine at the end. And it does not touch the free amino group at the front. Those three features, in that arrangement, are the reason either molecule can hold a copper ion at all.
Two further distinctions travel with the names. The free tripeptide AHK and the copper complex AHK‑Cu are different preparations: one is an organic molecule, the other is that molecule joined to Cu(II). PubChem lists the free peptide as CID 7408502; the complex is the subject sold under cosmetic and research‑chemical labels. This document names which of the two a given study actually used. The same care applies to the sibling: GHK and GHK‑Cu are likewise not interchangeable, and most of the literature that makes copper peptides famous is about the complex of the glycine peptide, not about alanine.
Section 02How a peptide holds a metal
The phrase “copper peptide” is misleading if it is read as a mixture. AHK‑Cu is not a peptide with some copper stirred into it. It is a single chemical entity in which the metal ion is clamped by four nitrogen atoms contributed by the peptide itself, in a flat, square arrangement around the copper.
The arrangement has a name in bioinorganic chemistry: the ATCUN motif, short for amino‑terminal copper and nickel binding site. It is also written more generally as Xaa‑Xaa‑His — any residue, any residue, then histidine. The rule is positional. If a peptide has a free α‑amino group at its N‑terminus and a histidine as its third residue, it will bind copper(II) with unusual avidity, and it will do so in a predictable geometry: the terminal amino nitrogen, the two backbone amide nitrogens of the first two peptide bonds, and the imidazole nitrogen of the histidine side chain.
Two consequences follow, and both matter for what comes later.
The first is that the identity of the first residue is comparatively unimportant to the binding. Glycine and alanine both present a free amino group and a backbone amide; the methyl group that distinguishes them points away from the metal. This is why swapping one for the other is a chemically conservative move — and it is precisely the licence that a medicinal chemist looking for a patentable analogue would exploit. The copper chemistry survives the substitution nearly untouched, while the molecule becomes a new composition of matter.
The second consequence is the awkward one. The ATCUN motif is not rare. It is the copper site of human serum albumin, the most abundant protein in blood plasma, whose own N‑terminal sequence is Asp‑Ala‑His‑Lys. Read that sequence again: residues two, three and four are Ala‑His‑Lys. The subject of this monograph is, letter for letter, a fragment of the front end of albumin, with the aspartate removed. This is a genuine chemical relationship and Section 03 explains why it also creates a problem for anyone trying to find out what is known about AHK‑Cu.
Copper is an essential trace element and a cofactor for enzymes involved in connective‑tissue formation, pigment synthesis and cellular respiration. Free copper ions, however, are toxic: in the presence of reducing agents they catalyse the production of damaging oxygen radicals. A cell's problem is therefore not obtaining copper but moving it around without letting it do harm on the way. The proposed rationale for every copper‑peptide preparation is that a tightly chelated complex delivers the metal in a controlled form. That is a coherent hypothesis. Whether AHK‑Cu does it, in a person, at any particular concentration, has not been measured — and Section 12 sets out the full list of what has not been measured.
Section 03Three names that are not this molecule
A reader who wants to check the claims made for AHK‑Cu will very likely start where this document started: with a literature search. It is worth explaining what that returns, because the answer is the single most informative fact about this compound, and it is not obvious.
On 3 August 2026, a search of PubMed — the world's principal index of biomedical literature, holding 40,960,840 records on that date — for the exact term AHK‑Cu in all fields returned one record. Not one page of results. One record. It is the 2007 paper discussed in Section 09.
Search instead for the sequence, Ala‑His‑Lys, and the count rises to twenty‑nine. This looks like progress and is not. Reading the twenty‑nine titles, two are about this tripeptide. The remainder are about human serum albumin and its variants — proalbumin Wu Yang, albumin Tokushima, chicken serum albumin, rat proalbumin — or about the albumin‑derived fragment DAHK, or about unrelated proteins that happen to contain the three residues in order, among them a haemoglobin variant, myelin basic protein and a bacterial pyridoxyl peptide. One preceding aspartate residue accounts for the great majority of the difference between a search that finds two relevant papers and a search that finds twenty‑nine.
The third collision is stranger. Search for the bare abbreviation AHK and PubMed returns 373 records. Of those, 185 carry the strings AHK2, AHK3 or AHK4, and 53 sit in plant biology. AHK1 through AHK5 are the Arabidopsis histidine kinases — the cytokinin receptors of a small flowering weed used as the standard model organism in plant genetics. In the biomedical index, “AHK” usually means a plant signalling protein.
This is not a curiosity. It has consequences that can be measured. When a compound catalogue tries to attach scientific studies to an entry labelled AHK‑Cu by matching on the three‑letter code, the studies that arrive can include a paper on diterpenoid compounds inducing resistance to tomato spotted wilt virus, a paper on the cytokinin activity of benzyladenine derivatives, a paper on a purine derivative that blocks cytokinin action, and a paper on the anchorage of brain hexokinase to mitochondrial porins. None of them is about a copper peptide. Three are about the Arabidopsis receptors and one is about an enzyme whose name simply ends in the same three letters. An automated system built by careful people can walk straight into the collision.
The fourth and last confusion is the one that most affects what a general reader will encounter, because it operates in commercial rather than scientific language. The class noun is “copper tripeptide”. A PubMed search for that phrase returns twenty‑four records; twenty‑two of them are about GHK‑Cu. In cosmetic ingredient nomenclature the two compounds are formally distinguished — GHK‑Cu is registered as Copper Tripeptide‑1 and AHK‑Cu as Copper Tripeptide‑3 — but the shorter phrase is used loosely everywhere, and a claim made for one is routinely read as a claim about the other. Section 14 gives two documented instances where that has happened in print.
Stripped of every collision, the indexed scientific record for this compound is: one PubMed record naming AHK‑Cu, four open‑access full texts in PubMed Central whose body text contains the string, and zero registered clinical trials anywhere in ClinicalTrials.gov. The local research library of 75,264 indexed text fragments contains the string “AHK‑Cu” exactly zero times. That is the evidence base. Everything that follows is an attempt to read it fairly — neither inflating it, which the market does, nor dismissing it, which would ignore a patent record containing real animal experiments.
Section 04The tripeptide from human serum
The story of AHK‑Cu begins with a molecule that is not AHK‑Cu, and with an experiment that had nothing to do with hair.
Loren Pickart was born in 1938 in Winona, Minnesota. He took degrees in chemistry and mathematics at the University of Minnesota and then a doctorate in biochemistry at the University of California, San Francisco. The title of his thesis was A TriPeptide from Human Serum, which is as clear a statement of a life's work as a young scientist is ever likely to write.
The observation behind it was simple enough to describe in a sentence. Human liver tissue kept alive in culture behaves differently depending on whose blood it is bathed in: something in the blood of young donors made old liver tissue synthesise proteins the way young tissue does. In 1973 Pickart traced that activity to a fraction of human albumin and isolated what was responsible — a tripeptide of glycine, histidine and lysine, carrying copper.
That is the discovery of GHK. It is a genuine one, and it has the shape that scientific discoveries are supposed to have: an anomaly noticed, a fractionation performed, an active principle identified. GHK is present in human plasma at roughly 200 nanograms per millilitre at age twenty and falls to about 80 nanograms per millilitre by age sixty (Pickart & Margolina, 2015) — a decline that is the foundation of every argument that copper peptides restore something age takes away. It has since been reported to influence collagen synthesis, wound repair and a large number of genes. Its literature runs to several hundred papers, many of them Pickart's own; by the time he died in 2023 at the age of eighty‑five he had published forty‑six peer‑reviewed articles and held twenty‑nine patents.
Note where that leaves this monograph's subject. The decline argument is GHK's and only GHK's. AHK has never been reported in human plasma at any concentration, at any age, because it is not made by the body at all.
None of that is about AHK‑Cu. It is set out here because almost everything a reader will encounter about “copper peptides” comes from this body of work, and because the commercial literature routinely attributes AHK‑Cu to Pickart. Section 07 examines that attribution.
Section 05A company that needed a second product
In 1986 Pickart co‑founded a company in Washington State to commercialise the copper‑peptide work. It was called ProCyte Corporation, and it went public in 1989.
ProCyte's lead programme was a wound‑healing drug. The product was called Iamin, its active ingredient was GHK‑Cu, and the intended indications were the indolent chronic wounds that heal badly or not at all — venous stasis ulcers, diabetic foot ulcers. The logic was direct: if a peptide isolated from young serum accelerates tissue repair, give it to people whose tissue is not repairing.
The programme did not succeed. The clinical trials of GHK‑Cu for chronic wound healing failed to demonstrate the benefit required, and the drug pathway was abandoned. The company survived by changing what it sold. Its subsequent product lines — Neova, Complex Cu3, GraftCyte, Tricomin — were cosmetics and post‑procedure skin care, not drugs. ProCyte was acquired by PhotoMedex on 18 March 2005 in a stock‑for‑stock transaction valued at approximately US$24.4 million.
The date that matters for this monograph sits inside that decline. The priority date of the patent family that first claims AHK‑Cu for hair growth is 17 June 1994. That is a company with a failing drug programme filing on a new molecule for a new indication in a market — hair loss — that does not require a drug approval to enter.
This is not an accusation. It is the ordinary commercial logic of a small biotechnology company, and it is the reason the compound exists. But it does explain a feature of the evidence base that would otherwise be puzzling: why a molecule with a plausible mechanism and an active market has almost no published science behind it. AHK‑Cu was never developed as a drug. It was developed as a cosmetic ingredient, and cosmetic ingredients do not have to generate clinical trials.
Section 06The screen that produced it
The patents are United States 5,538,945 and its continuation 6,017,888, both titled Stimulation of hair growth by peptide copper complexes, both assigned to ProCyte Corporation. The later patent was filed on 23 December 1997 and granted on 25 January 2000. The named inventors are Alexander J. Pallenberg, Leonard M. Patt and Ronald E. Trachy.
What the patents describe is a screen. A series of peptide‑copper complexes was synthesised and tested in mice for the ability to provoke hair growth. The complexes named in the experimental examples include glycyl‑histidine copper (GH:Cu), glycyl‑histidyl‑phenylalanine copper (GHF:Cu), alanyl‑histidyl‑phenylalanine copper (AHF:Cu) and alanyl‑histidyl‑lysine copper (AHK:Cu), the last at several peptide‑to‑copper ratios. The general claim covers complexes in which the second residue is histidine or arginine, at peptide‑to‑copper ratios from about 1:1 to about 3:1.
The animal model was the C3H mouse, a strain whose hair follicles enter a long, synchronised resting phase — telogen — at around sixty days of age. A synchronised resting coat is a convenient background against which to see a compound push follicles back into growth, because untreated animals will not do it on their own.
In the intradermal experiments, complexes were injected at 0.75 to 1.5 milligrams in 0.1 millilitre of sterile saline. Hair growth became visible around day 10, was active between days 14 and 20, and reached its maximum by about day 29, producing roughly circular patches of regrown fur between 0.5 and 5 square centimetres. All the complexes tested produced a response.
The topical experiments are the ones that selected AHK‑Cu. AHK:Cu and AHF:Cu, both at a 1.1:1 peptide‑to‑copper ratio, were applied to the skin of female C3H mice twice daily at 0.5 per cent and 0.1 per cent. The conclusion the patent draws from the comparison is explicitly about physical chemistry rather than biology: complexes carrying a hydrophilic residue — the lysine of AHK:Cu — were more active when applied topically than otherwise similar complexes carrying a hydrophobic residue, the phenylalanine of AHF:Cu.
That sentence is the origin of AHK‑Cu as a product. The molecule was not selected because alanine does something glycine cannot. It was selected because, among a set of copper complexes that all worked when injected, the one with a charged, water‑loving residue at the far end worked best when rubbed on. AHK‑Cu is the answer to a formulation question.
The animal data described above have never been published in a peer‑reviewed journal. They exist in a granted United States patent, which means they were filed under a legal duty of candour and examined by a patent office — but not that they were reviewed by scientists in the field, replicated, or reported with the statistics, group sizes and blinding that a journal would require. Patent examples are evidence. They are not the same kind of evidence as a paper, and this document does not treat them as though they were.
Section 07Who discovered it, and who did not
Commercial descriptions of AHK‑Cu very often credit Loren Pickart. The record supports a narrower statement, and the difference is worth stating plainly because it changes what kind of molecule a reader thinks they are looking at.
Pickart is an author on the earliest published report of peptide‑copper complexes stimulating hair follicles: Trachy, Fors, Pickart and Uno, in the Annals of the New York Academy of Sciences in December 1991 (Trachy et al., 1991). It is a two‑page conference communication reporting results in C3H mice, and the compound it names is a ProCyte development code, PC 1038. PubMed holds no abstract for it.
Pickart is not a named inventor on either AHK‑Cu hair patent. Those name Pallenberg, Patt and Trachy. Trachy is the bridge between the two: first author on the 1991 communication and an inventor on the patents.
So the accurate account is this. Pickart discovered GHK, an endogenous human tripeptide, and founded the company. That company later ran a screen that produced AHK‑Cu, a synthetic analogue he did not invent, for an indication that was not his original programme. The two molecules are chemically adjacent and historically connected, and they are not the same discovery.
Why insist on it? Because the most persuasive thing anyone can say about GHK‑Cu is that the body makes it — that it is a signal the organism already uses and that declines with age. Not one word of that applies to AHK‑Cu. It has never been reported in human tissue. It is a designed compound, and its case must be made on its own evidence rather than borrowed from a molecule that has a different provenance.
Section 08The animal record
Everything published or filed about AHK‑Cu in a living animal fits in this section, and it dates from between 1991 and 2000.
The published component is the two‑page communication already mentioned (Trachy et al., 1991), which reports that peptide‑copper complexes stimulate hair follicles in C3H mice and names a development code rather than a sequence. It carries no abstract in PubMed and it is a conference proceeding rather than a full paper.
The unpublished component is the patent record described in Section 06, and it is substantially richer. Three experiments in the patents bear directly on this compound.
Intradermal injection. AHK:Cu, among other complexes, at 0.75 to 1.5 milligrams per injection in 0.1 millilitre of saline, into C3H mice at sixty days of age in telogen. Hair growth appeared from about day 10, was active over days 14–20, and peaked around day 29, producing regrown patches of 0.5 to 5 square centimetres.
Topical application. AHK:Cu at a 1.1:1 peptide‑to‑copper ratio, 0.5 per cent and 0.1 per cent, applied by cotton‑tipped applicator twice daily on weekdays to female C3H mice aged 60–65 days. The comparator was AHF:Cu at the same ratio and concentrations. AHK:Cu produced the superior response, and the patent attributes the difference to the hydrophilic lysine.
Chemotherapy‑induced hair loss. Mice received cytosine arabinoside (Ara‑C) at 25 milligrams per kilogram by intraperitoneal injection on seven consecutive days, which causes substantial hair loss by days 5–6. AHK:Cu at a 1:1 ratio was injected intradermally at 0.1 to 0.5 milligrams. Hair was retained in a radius of about 0.25 centimetres around the injection sites, most clearly in the 0.1 to 0.5 milligram groups.
The protective effect against chemotherapy‑induced hair loss extended about a quarter of a centimetre from the needle. That is a local effect at the site of an injection, in a mouse, against a specific cytotoxic drug. It is an interesting observation about the compound's activity. It is not a systemic effect, it was not tested by any other route, and no human has been reported to receive AHK‑Cu for this or any other purpose.
Two things about the animal record are worth stating explicitly. The first is that it is genuinely positive: across several complexes, routes and concentrations, the compounds did what the inventors said they did in the model they used. The second is that it stops. No animal study of AHK‑Cu has been published in the peer‑reviewed literature since the 1991 communication — a gap of thirty‑five years.
Section 09The one paper
In July 2007 a group at Seoul National University College of Medicine published the only peer‑reviewed primary study of AHK‑Cu in existence: The effect of tripeptide‑copper complex on human hair growth in vitro (Pyo et al., 2007). It is the record that the entire commercial case for this compound rests on, so it is worth reading carefully rather than summarising.
The paper has two experimental systems. The first is human hair follicles maintained alive outside the body — an organ culture, in which a dissected follicle continues to grow a hair shaft in a dish for a week or two. The second is cultured human dermal papilla cells. The dermal papilla is the small cluster of specialised fibroblasts at the base of a follicle that governs the hair cycle; it is the closest thing a follicle has to a control centre, and it is the standard target for anything intended to affect hair growth.
AHK‑Cu was applied across a concentration range of 10−12 to 10−9 molar — picomolar to nanomolar, which is to say extremely dilute. Within that range the paper reports two positive findings: the cultured follicles elongated more than untreated controls, and the dermal papilla cells proliferated more.
The paper then asks why, and proposes that AHK‑Cu protects dermal papilla cells from programmed cell death. Three lines of evidence are offered for that proposal, and they do not all point the same way.
Flow cytometry with annexin V and propidium iodide labelling — the standard method for counting cells that are undergoing apoptosis — showed that 10−9 molar AHK‑Cu reduced the number of apoptotic cells. The authors report, in the same sentence, that this decrease was not statistically significant.
Two molecular measures did reach the authors' threshold. The ratio of Bcl‑2 to Bax — a widely used index in which a higher value indicates a cell biased towards survival — was elevated. And the levels of the cleaved, activated forms of caspase‑3 and PARP, two proteins whose cleavage is a hallmark of a cell committed to apoptosis, were reduced by treatment at 10−9 molar.
The authors' own conclusion is appropriately hedged: the study “proposed” that AHK‑Cu promotes the growth of human hair follicles and that the effect “may occur” through proliferation and the prevention of apoptosis of dermal papilla cells.
At its strongest: a real experiment on human tissue, with a dose‑ranging design, two independent readouts of the primary effect (follicle elongation and cell proliferation), and a mechanistic investigation that produced concordant results on two protein markers. It is a competent piece of work and its central observation — that a very dilute copper tripeptide lengthens cultured human follicles — is the best single piece of evidence this compound has.
At its weakest: it is one paper, from one group, never independently replicated in nineteen years. It has no in‑vivo arm and no human application. Its headline mechanistic measurement was null. It is not deposited in PubMed Central, so its full text is not openly readable and this monograph works from its abstract and its indexed record. And an organ‑culture follicle is severed from its blood supply, its nerves and its hormonal environment; growth in that system is a genuine signal about the tissue, not a prediction about a scalp.
Section 10What is inherited, and whether it should be
The 2007 paper opens by describing what “the tripeptide‑copper complex” does: stimulates the proliferation of dermal fibroblasts, raises production of vascular endothelial growth factor, and decreases secretion of transforming growth factor‑β1. Those three effects are frequently quoted as AHK‑Cu's mechanism. They are, in that paper, framed as properties of the class, and the underlying measurements come very largely from work on GHK‑Cu.
The distinction matters more than it may appear. Raising VEGF and lowering TGF‑β1 would be a coherent story for hair growth: VEGF supports the blood supply a growing follicle needs, and TGF‑β1 is one of the signals that drives a follicle out of its growth phase. If AHK‑Cu does both, the ex‑vivo elongation has an explanation. But whether AHK‑Cu does both has not been shown in a published study of AHK‑Cu.
Shared chemistry is a reason to hypothesise, not a reason to conclude. The two molecules share the ATCUN motif and therefore bind copper similarly, which makes transfer of a mechanism plausible. They differ at the residue that, according to the patent's own topical comparison, changes how the molecule behaves at a biological surface — which is direct evidence that the first position is not inert. A single methyl group is a small chemical change and it is not guaranteed to be a small biological one.
Where GHK‑Cu has gene‑expression datasets, a long clinical literature and an active randomised trial, AHK‑Cu has none of those. Any statement of the form “copper peptides do X, therefore AHK‑Cu does X” is an inference across that gap, and this document flags it as one wherever it appears.
Section 11Three molecules that are not AHK‑Cu
Three related preparations are routinely cited as evidence for AHK‑Cu. None of them is AHK‑Cu.
Vitamin C‑linker‑conjugated AHK. In 2018 a Korean group reported that a vitamin C conjugate of the AHK tripeptide enhanced BMP‑2‑induced osteoblast differentiation in C2C12 mouse myoblasts, acting through Smad1/5/8 and the MAP kinases ERK1/2 and p38 (Jung et al., 2018). This is a real result and it is regularly listed as an AHK‑Cu study. The tested substance was a vitamin C conjugate, not the free tripeptide; there was no copper in it; the cells were mouse muscle precursors, not human skin or follicle cells; and the endpoint was bone differentiation, not hair. It tells us about a different molecule doing a different thing in a different tissue.
The free peptide, AHK, without copper. Sold as a separate product and catalogued separately. The whole proposed rationale of a copper peptide is the delivery of the metal; a preparation without the metal is not the same agent, and the 2007 experiments were done on the complex.
Palmitoyl‑AHK. A lipopeptide — the tripeptide with a sixteen‑carbon fatty acid attached, a standard cosmetic modification intended to improve skin penetration. It is a distinct chemical entity with its own behaviour, and this document located no primary study of it.
Listing these is not pedantry. The pattern by which a compound accumulates an evidence base it does not have is precisely this one: adjacent molecules, adjacent literatures, and citations that do not survive being followed.
Section 12What has never been done
House rule in this series is that gaps are reported with the same prominence as findings, in the same place. For AHK‑Cu the list of gaps is longer than the list of findings, so it gets its own section.
| Question | Status for AHK‑Cu |
|---|---|
| Any registered clinical trial, any indication, any country | None. ClinicalTrials.gov searched 3 August 2026; two records matched the query and both concern GHK‑Cu or a device procedure |
| Human efficacy study, published, any design | None located |
| Pharmacokinetics — absorption, distribution, half‑life | None published, by any route, in any species |
| Skin penetration measured for this compound | Not located; penetration studies in this literature used GHK‑Cu |
| Formal toxicology or safety study | None located |
| Dose‑response in a living animal | Only the patent examples, unpublished and unreplicated |
| Head‑to‑head against GHK‑Cu on any endpoint | None, in any species. The patent compares AHK:Cu with AHF:Cu, not with GHK:Cu |
| Independent replication of the 2007 findings | None in nineteen years |
| Gene‑expression profiling | None. The datasets frequently cited are GHK's |
The last row of that table is worth pausing on, because it is the sharpest illustration of the pattern this monograph keeps returning to. The patent that created AHK‑Cu compared it against a phenylalanine analogue and found it better. It never compared it against the glycine analogue — against GHK‑Cu, the compound the same company was already selling, the compound AHK‑Cu is invariably marketed as an improvement on. Thirty‑two years later, nobody else has either.
Absence of a trial is not evidence of absence of an effect. It is evidence that the claim has not been asked, under controlled conditions, of the molecule that is supposed to answer it. Part Four turns from that empty list to the product history that grew anyway — Tricomin, the cosmetic route, and the specific points at which secondary description has outrun the primary sources.
Section 13Tricomin, and the cosmetic route
AHK‑Cu reached the public as a cosmetic. ProCyte's hair line was called Tricomin, and its central product was a treatment spray containing one per cent AHK‑Cu. The brand passed to PhotoMedex with the 2005 acquisition and remains on sale.
The regulatory position this creates is worth spelling out, because it is the reason the evidence base looks the way it does. A cosmetic ingredient in the United States and the European Union does not require pre‑market demonstration of efficacy. It requires safety, appropriate labelling, and restraint in what is claimed: a cosmetic may say it improves the appearance of hair, and may not say it treats a disease. Nothing in that framework asks for a randomised trial, and so none was run.
In the International Nomenclature of Cosmetic Ingredients, the two copper tripeptides have separate registered names:
| INCI name | Molecule | Sequence | Provenance |
|---|---|---|---|
| Copper Tripeptide‑1 | GHK‑Cu | Gly‑His‑Lys + Cu(II) | Occurs in human plasma |
| Copper Tripeptide‑3 | AHK‑Cu | Ala‑His‑Lys + Cu(II) | Synthetic |
This distinction is not decorative. An INCI name is the identifier a formulator, a regulator and a consumer use to know which substance is in a product. The two names denote different molecules with different evidence behind them.
Section 14Three claims that outgrew their sources
This section documents two specific, checkable drifts between what the primary literature reports and what is said about AHK‑Cu downstream. They are given in detail because they are the practical reason a reader might want this monograph at all.
The first drift: a null result loses its qualifier
The 2007 paper reported that AHK‑Cu at 10−9 molar reduced the number of apoptotic dermal papilla cells measured by annexin V and propidium iodide labelling, and stated in the same sentence that the decrease was not statistically significant.
A 2026 review of short peptides for hair loss (Fan et al., 2026) summarises the same experiment as: annexin/PI double staining “confirmed fewer apoptotic DPCs.” The qualifier is gone, and the verb has become confirmed. This is not a marketing page. It is a peer‑reviewed review article, and it is the most recent secondary source on the subject.
Commercial pages go further, quoting a specific percentage reduction in caspase‑3 together with donor and follicle counts. This monograph could not locate those figures in the paper's abstract or indexed record, and the paper is not deposited in PubMed Central, so its full text is not openly verifiable. Two possibilities exist — the numbers are in the closed full text, or they are not — and a reader is entitled to know that the distinction cannot currently be checked without paying for the article.
The second drift: measurements migrate between molecules
A 2021 review of topical and nutricosmetic hair products states that AHK‑Cu “yielded the same outcome in vitro” as GHK‑Cu, listing reduced expression of TGF‑β1, increased hair shaft elongation and increased expression of vascular endothelial growth factor (Sadgrove & Simmonds, 2021).
Of those three, the 2007 paper reports the second. The VEGF and TGF‑β1 statements appear in that paper's introduction, as properties of the copper‑tripeptide class established largely in GHK‑Cu work — not as measurements made on AHK‑Cu. A class‑level assertion in an introduction has become a compound‑level finding two citations downstream.
The third drift, smaller but instructive: the wrong INCI number
Multiple commercial pages describe AHK‑Cu as “Copper Tripeptide‑1”. That is GHK‑Cu's registered name. Any claim attached to it therefore points at the sibling's literature. A reader comparing two products cannot resolve which molecule they are buying from a page that misnames it.
None of these three drifts requires anyone to lie. Each is a small, ordinary compression: a qualifier dropped for brevity, an introduction read as a result, a number transposed. But they compound in one direction. At the primary source, AHK‑Cu is a compound with one in‑vitro paper whose apoptosis result was null. Three citation steps later it is a compound with a demonstrated anti‑apoptotic mechanism, a measured VEGF effect and a percentage. Nothing new was measured in between.
Section 15What the market sells
AHK‑Cu is sold in two quite different channels, and they make different claims.
The cosmetic channel sells finished topical products — scalp sprays, serums, shampoos — at concentrations typically quoted in the hundreds of parts per million, under cosmetic claims about the appearance of hair.
The research‑chemical channel sells the raw material. Vendor catalogues list AHK‑Cu, the free peptide AHK, and palmitoyl‑AHK at a 200 milligram unit size from more than one supplier, labelled for research use only. That labelling is a legal statement about the terms of sale, not a scientific statement about the compound, and this document takes no position on it beyond noting that the compound has no approved human indication anywhere.
One structural observation connects the two channels to Part Three. The 2007 experiments were done at 10−12 to 10−9 molar on follicles and cells in a dish, with the compound already in contact with the tissue. A topical product must first cross the stratum corneum, and no published study has measured how much AHK‑Cu does so, or what concentration is reached at the dermal papilla of a living scalp. The relationship between the concentration that worked in a dish and the concentration in a bottle is therefore unknown — not unfavourable, not favourable, unknown.
A note on the 2026 review's dose range
One discrepancy is worth recording rather than resolving. The 2007 abstract gives the tested range as 10−12 to 10−9 molar. The 2026 review's summary table gives it as 10−13 to 10−7 molar (Fan et al., 2026). Those are not the same range and one of them does not describe the experiment. Without open access to the paper's full text this document cannot say which, and reports the abstract's figures throughout while flagging that the most recent review disagrees with them.
Section 16The evidence, tiered
The table below is the whole case for AHK‑Cu, arranged so that the kind of evidence behind each claim is visible in the same row as the claim. Nothing has been omitted for being unfavourable and nothing has been added from the sibling compound.
| Claim | Best supporting source | System | About AHK‑Cu itself? | Tier |
|---|---|---|---|---|
| Binds copper(II) in a defined square‑planar geometry | ATCUN motif chemistry; direct NMR and FTIR evidence on Ala‑His‑Lys‑NH2 (Ren et al., 2026) | Solution chemistry | Yes | Well established |
| Lengthens human hair follicles in organ culture | Pyo et al., 2007 | Human tissue, ex vivo | Yes | Single study |
| Increases proliferation of human dermal papilla cells | Pyo et al., 2007 | Human cells, in vitro | Yes | Single study |
| Shifts Bcl‑2/Bax and reduces cleaved caspase‑3 and PARP | Pyo et al., 2007 | Human cells, in vitro | Yes | Single study |
| Reduces the number of apoptotic dermal papilla cells | Pyo et al., 2007 | Human cells, in vitro | Yes | Not significant |
| Stimulates hair regrowth after intradermal injection | US 5,538,945 / 6,017,888, examples | C3H mouse, in vivo | Yes | Patent example, unreviewed |
| Outperforms a hydrophobic analogue when applied topically | US 6,017,888, example 14 | C3H mouse, in vivo | Yes | Patent example, unreviewed |
| Limits local hair loss from a cytotoxic drug | US 6,017,888, example 16 | Mouse, in vivo, local | Yes | Patent example, unreviewed |
| Raises VEGF and lowers TGF‑β1 | Class‑level statement; measurements are GHK‑Cu's | — | No | Inferred from sibling |
| Promotes osteogenic differentiation | Jung et al., 2018 | Mouse myoblasts, in vitro | No — vitamin C conjugate, no copper | Different molecule |
| Improves hair in a human being | One case report; a Tricomin‑line shampoo among three agents, “without significant improvement” (Fronek et al., 2019) | Single patient, uncontrolled | Brand, not compound | No supporting evidence |
Three features of that table deserve to be said in words.
First, every row that is about AHK‑Cu itself and reaches a conclusion comes from either one 2007 paper or one patent family. There is no third source.
Second, the single most‑quoted mechanistic claim — that the compound prevents apoptosis in dermal papilla cells — rests on a measurement that its own authors reported as not statistically significant, supported by two protein markers that did reach significance. That is a reasonable basis for a hypothesis and a weak basis for an assertion.
Third, the only appearance of the compound's brand in a published human case ends in the phrase “without significant improvement.” That is a single patient with an unrelated scalp condition, treated with three things at once, and it is close to worthless as evidence of inefficacy — but it is the only human datum there is, and a document that reported the positive mouse data while omitting it would be misrepresenting the record.
Section 17Recency, and what the last five years added
Where evidence is thin, recent evidence deserves extra weight — a new study can change a picture that a single old study established. So it is worth asking precisely what the last five years produced.
The answer is: no new primary research on this compound. Everything in the recent record is one of three things.
Reviews that cite the 2007 paper. The most recent and most thorough is an overview of short peptides for hair loss (Fan et al., 2026), which treats Pyo et al. as a landmark and states plainly that AHK‑Cu's signalling mechanism “is less well characterized than GHK‑Cu.” Its own summary table classifies the AHK‑Cu evidence as ex vivo, against “clinical evidence” and “Phase I clinical data” for GHK‑Cu. Even the friendliest recent secondary source places this compound a tier below its sibling.
Multi‑ingredient formulations. Products such as the QR678 injectable hair formulations cite the 2007 study as part of the rationale for including a copper peptide among several actives (Kapoor et al., 2020). A positive result for a mixture is not a result for a component.
Analytical chemistry. The most active recent use of the Ala‑His‑Lys sequence is as a scaffold for copper‑selective fluorescent probes (Ren et al., 2026; and reviewed by Fosnacht & Pluth, 2024). These papers are not about hair at all, but they are not irrelevant: they confirm, spectroscopically and repeatedly, that the tripeptide binds Cu(II) selectively and through the imidazole nitrogen. The chemistry of Section 02 is on firmer ground than the biology of Section 09, and it is recent work that put it there.
Applying the recency principle honestly therefore yields a null: the freshest data do not overturn the 2007 picture, and they do not strengthen it either. They describe it from further away.
Section 18Against its sibling
AHK‑Cu is almost always encountered as an alternative to GHK‑Cu, and usually as an upgrade. The chemistry makes that plausible and the evidence does not support it.
| GHK‑Cu | AHK‑Cu | |
|---|---|---|
| Origin | Isolated from human plasma, 1970s | Synthesised and selected in a 1994 screen |
| Occurs in the body | Yes; declines with age | Not reported in any human tissue |
| PubMed records under its own name | 75 | 1 |
| Open full texts naming it | 131 | 4 |
| Registered clinical trials | Yes, including an active Phase 2 wound study | None |
| Gene‑expression datasets | Yes | None |
| INCI name | Copper Tripeptide‑1 | Copper Tripeptide‑3 |
| Direct comparison between them | Never performed, in any species, on any endpoint | |
The last row is the one that matters. The claim a reader will meet is that AHK‑Cu is the better copper peptide for hair. That claim has never been tested. The only comparison the patent ran was against a phenylalanine analogue nobody sells, and it was a comparison of topical delivery rather than of biological potency.
It is worth stating the counter‑case fairly, because there is one. The 2007 study is the only work that has ever put either compound onto human hair follicles in organ culture at picomolar concentrations and measured elongation, and it used AHK‑Cu. On that specific experiment, in that specific system, AHK‑Cu has data and GHK‑Cu does not. A reader weighing the two should hold both facts at once: the sibling has vastly more evidence overall, and the one hair‑follicle organ‑culture experiment in this literature was done with this compound.
Section 19What a fair summary would say
AHK‑Cu is a well‑characterised piece of coordination chemistry attached to a very small biology.
The chemistry is not in doubt. It is a tripeptide with an amino‑terminal copper‑binding motif, it holds Cu(II) tightly and selectively, and that has been confirmed by independent groups using independent methods as recently as this year — although they were building sensors, not treatments.
The biology consists of one peer‑reviewed paper, published nineteen years ago and never replicated, reporting that very dilute AHK‑Cu lengthens human hair follicles in a dish and makes dermal papilla cells divide; plus a set of mouse experiments that exist only in a patent. Those findings are positive and they are real. They are also the entire published biological record, they contain no human outcome, and one of the paper's three mechanistic measurements was null.
What separates AHK‑Cu from a compound with a comparable evidence base is not the science but the market: it has been sold continuously for roughly three decades, which produces a volume of confident secondary description far out of proportion to the primary literature. Section 14 documents three specific points at which that description has drifted from its sources, one of them inside a peer‑reviewed review published this year.
The compound may well do what is claimed for it. The honest statement is that after thirty‑two years nobody has run the experiment that would show it.
This document describes published research and patent filings. It does not recommend human use of AHK‑Cu or of any other compound named in it, and it specifies no dose, route or schedule for any person. Concentrations, doses and durations appear only as parameters of the experiments that reported them, attached to the species and the system in which they were measured. AHK‑Cu has no approved human therapeutic indication in any jurisdiction; where it is sold as a cosmetic ingredient it is sold under a framework that does not require efficacy to be demonstrated, and where it is sold as a research chemical it is sold for laboratory use only. Nothing here is medical advice.
Section 20References
Generated from verified NCBI records rather than from recall. Author lists, journal names, volumes, pages and identifiers are taken from the PubMed or PubMed Central record for each citation. Patent citations are given by number, assignee and named inventors as printed on the granted document.
- Fan C, Chen Y, Huang Q, Ou WY, Zhang C, Sun Y, et al.. Overview of Short Peptides for Hair Loss. Biomedicines. 2026;14(4):864.
PMID 42072405 · doi:10.3390/biomedicines14040864 · PMC13113319 - Fosnacht KG, Pluth MD. Activity-Based Fluorescent Probes for Hydrogen Sulfide and Related Reactive Sulfur Species. Chem Rev. 2024;124(7):4124-4257.
PMID 38512066 · doi:10.1021/acs.chemrev.3c00683 · PMC11141071 - Fronek LF, Braunlich K, Farsi M, Miller RA. A Rare Case of Cutis Verticis Gyrata with Underlying Cerebriform Intradermal Nevus. Cureus. 2019;11(12):e6499.
PMID 32025420 · doi:10.7759/cureus.6499 · PMC6988481 - Jung JI, Park KY, Lee Y, Park M, Kim J. Vitamin C-linker-conjugated tripeptide AHK stimulates BMP-2-induced osteogenic differentiation of mouse myoblast C2C12 cells. Differentiation. 2018;101:1-7.
PMID 29567599 · doi:10.1016/j.diff.2018.03.001 - Kapoor R, Shome D, Vadera S, Kumar V, Ram MS. QR678 & QR678 Neo Hair Growth Formulations: A Cellular Toxicity & Animal Efficacy Study. Plast Reconstr Surg Glob Open. 2020;8(8):e2843.
PMID 32983753 · doi:10.1097/GOX.0000000000002843 · PMC4313060 - Pickart L, Vasquez-Soltero JM, Margolina A. GHK Peptide as a Natural Modulator of Multiple Cellular Pathways in Skin Regeneration. Biomed Res Int. 2015;2015:648108.
PMID 26236730 · doi:10.1111/j.1524-475X.2009.00466.x · PMC2120067 - Pyo HK, Yoo HG, Won CH, Lee SH, Kang YJ, Eun HC, et al.. The effect of tripeptide-copper complex on human hair growth in vitro. Arch Pharm Res. 2007;30(7):834-9.
PMID 17703734 · doi:10.1007/BF02978833 - Ren Y, Jia M, Xiong S, An Y, Yang X, Wang P. A dual-signals peptide-based probe for fluorometric and colorimetric detection of copper(II) ions and glyphosate in multiple food samples and biological system. Food Chem X. 2026;35:103776.
PMID 41939933 · doi:10.1016/j.fochx.2026.103776 · PMC13050119 - Sadgrove NJ, Simmonds MSJ. Topical and nutricosmetic products for healthy hair and dermal antiaging using "dual-acting" (2 for 1) plant-based peptides, hormones, and cannabinoids. FASEB Bioadv. 2021;3(8):601-610.
PMID 34377956 · doi:10.1096/fba.2021-00022 · PMC3875420 - Trachy RE, Fors TD, Pickart L, Uno H. The hair follicle-stimulating properties of peptide copper complexes. Results in C3H mice. Ann N Y Acad Sci. 1991;642:468-9.
PMID 1809108 · doi:10.1111/j.1749-6632.1991.tb24420.x
Patents cited: United States Patent 5,538,945, Stimulation of hair growth by peptide copper complexes, ProCyte Corporation, priority 17 June 1994; and its continuation United States Patent 6,017,888, same title and assignee, filed 23 December 1997, granted 25 January 2000, inventors A. J. Pallenberg, L. M. Patt and R. E. Trachy.
Section 21Evidence handling
Study type is named in the sentence that reports the finding. An organ-culture result is called an organ-culture result; a mouse result is called a mouse result. Nothing measured in a dish is phrased so as to imply an outcome in a person.
Patent examples are labelled as such and are not treated as publications. A substantial part of the in-vivo evidence for this compound exists only in granted United States patents. Those documents were filed under a duty of candour and examined by a patent office, which is a real constraint; they were not peer-reviewed, replicated, or reported with the group sizes, blinding and statistics a journal would require. This document uses them, credits them, and says what they are.
Tiering. Five levels are used: well established for a claim supported by independent groups using independent methods; single study for a claim resting on one peer-reviewed report; patent example for unreviewed experimental data in a filing; not significant where the source itself reports the result as failing its own threshold; and different molecule where the evidence concerns a relative rather than this compound.
Conflicts are presented as conflicts. Where a secondary source states something the primary source does not, both are quoted and the discrepancy is named — including where the secondary source is a peer-reviewed review published this year. Section 14 documents three such cases and Section 15 records a fourth, a dose range on which the 2007 abstract and the 2026 review disagree by four orders of magnitude at the lower bound. This document does not adjudicate that disagreement, because the primary full text is not openly available; it reports the abstract's figures and flags the conflict.
Absence is reported with the same prominence as presence. Section 12 is a table of experiments that have never been run. On a compound with one primary paper, the shape of what is missing is a larger part of the honest picture than the shape of what is there.
Related molecules are named, not collapsed. Findings about GHK‑Cu, free AHK, vitamin C‑linker‑conjugated AHK, or palmitoyl‑AHK are attributed to those substances. Shared copper‑binding chemistry is a reason to hypothesise parallel biology; it is not a reason to transfer a result.
Chemistry sources retained for mechanism, not efficacy. A few papers that use Ala‑His‑Lys as a copper‑probe scaffold carry spectroscopic evidence that the tripeptide coordinates Cu(II) through the imidazole nitrogen. Section 02 rests partly on them. They are not biological efficacy studies of AHK‑Cu.
All 8 figures in this edition are original works authored for this document from the values cited in it. No third-party figure has been reproduced. Commissioned illustration plates, if later admitted, will be recorded in the artwork disposition and will not replace a verified number already carried by an authored figure.
Continue exploring