AHK-Cu

What is indexed under the name

PubMed carries AHK-Cu as a supplementary concept and returns exactly one record under it: Pyo and colleagues, Archives of Pharmacal Research, 2007. A full-text search of Europe PMC for the string "AHK-Cu" returns five records in total, and the other four are a 2020 formulation paper, two reviews and a biomaterials paper, each citing that same 2007 study. Searching the expanded name is worse rather than better. The string "alanyl-histidyl-lysine" returns fifteen PubMed records, fourteen of which concern DAHK, the aspartyl-alanyl-histidyl-lysine sequence at the N-terminus of human serum albumin, or tryptic fragments of alpha-lactalbumin that happen to contain the same three residues.

Deuterohemin-AlaHisLys, abbreviated DhHP-3, is an iron-porphyrin conjugate of the same tripeptide and contains no copper; Lei and colleagues reported in 2014 that 3 mg/kg reduced blood glucose, insulin resistance and malondialdehyde content in rats with non-insulin-dependent diabetes, with neither route nor group size stated in the abstract. Vitamin C-linker-conjugated AHK is a second distinct compound; Jung and colleagues reported in 2018 that it enhanced BMP-2-induced alkaline phosphatase and Smad1/5/8 phosphorylation in mouse C2C12 myoblasts. That molecule also contains no copper. Two adjacent molecules of this kind make string matching unreliable in the other direction, because a search on the tripeptide name retrieves both.

One primary study of the copper complex, in human tissue ex vivo and in cell culture, published nineteen years ago, is what remains once the near-misses are subtracted. Alongside it sits one granted United States patent whose worked examples contain the only rodent experiments on the molecule that have been located. Everything else retrieved under the name is a citation of one of those two documents, a citation of a different peptide, or a figure with no document behind it at all. The rest of this page separates those three categories.

Identity, and the identifiers that disagree

Three PubChem records are involved and they do not describe the same species. CID 7408502 is the metal-free tripeptide, C15H26N6O4 at 354.41 g/mol, CAS 126828-32-8, and it is the only form the FDA substance register carries, as ALANYLHISTIDYLLYSINE under UNII 1AVY5QO8WR. CID 168431292 is the record that actually bears the name AHK-Cu: the monohydrochloride complex, depicted as an anion, C15H24ClCuN6O4 at 451.39 g/mol, CAS 682809-81-0. CID 178189140 is retrieved by the second registry number in circulation, 767286-83-9, and is a doubly deprotonated complex at 414.93 g/mol carrying no name at all in that record.

Figures on secondary pages track none of these cleanly. One aggregator lists the molecular weight as 416.9 with the formula C15H25CuN6O4; another gives C14H25N7O4Cu, which has the wrong carbon and nitrogen counts for this peptide entirely. Supplier listings commonly give 452.40 with the formula C15H25ClCuN6O4, which is the neutral monohydrochloride salt and differs from the PubChem entry by the proton that database chose to strip. Only the last of those is reconcilable, and a reader comparing molar concentrations across sources is comparing at least three different assumptions.

The cosmetic nomenclature for this molecule is Copper Tripeptide-3, while Copper Tripeptide-1 is GHK-Cu, a different peptide, and at least one aggregator page titles its AHK-Cu entry Copper Tripeptide-1. In circulating ingredient and catalogue documentation the name Copper Tripeptide-3 is attached to CAS 767286-83-9, the doubly deprotonated registry number, rather than to 682809-81-0, so the cosmetic name and the name-bearing PubChem record point at two different deprotonation states. Stoichiometry is a third axis: the 2007 study used an 11 per cent aqueous stock without specifying a ratio, while the patent describes preparations at 2:1, 1.1:1 and 1:1 peptide to copper, each with a different absorption maximum, from 563 to 580 nm at 2:1 up to 595 nm at 1:1.

The 2007 experiment, in full

Hair follicles were taken from the occipital scalp of ten healthy volunteers aged 20 to 35 who had been on no medication for at least a month, under an institutional review board approval recorded in the paper. Follicles were cut about 2.5 mm from the base of the dermal papilla and cultured for twelve days in Williams' E medium with L-glutamine, insulin, hydrocortisone and antibiotics, refed three times a week. AHK-Cu was added to final concentrations from 10 to the minus 13 through 10 to the minus 7 molar. A total of 240 follicles from three donors were analysed, thirty per condition. Elongation was measured directly at day twelve against a scale bar.

Follicle elongation was significantly greater than vehicle across 10 to the minus 12 through 10 to the minus 9 molar, at p below 0.01 or p below 0.001 depending on condition. Dermal papilla cell proliferation by MTT was significantly raised over the same range and unchanged at 10 to the minus 8 molar. Flow cytometry after 72 hours at 10 to the minus 9 molar reduced annexin V-positive, propidium-iodide-negative cells by 3.48 per cent, a difference the authors state was not statistically significant. Western blotting at the same concentration reduced cleaved caspase-3 by 42.7 per cent and PARP cleavage fragments by 77.5 per cent, both at p below 0.05, across three independent experiments, with procaspase-3 unchanged.

The methods state that the MTT assay ran from 10 to the minus 13 molar; the figure legend for that assay states 10 to the minus 12. The same legend describes results as mean percentages of control for six cultures and then reports means and standard errors from eight wells. Both inconsistencies affect how the numbers should be read. Funding is disclosed as partly a Korea Health 21 R&D Project grant, number 03-PJ1-PG1-CH13-0001, and partly a research agreement with a cosmetics corporation. PubMed lists no erratum, correction or expression of concern attached to the paper.

The ceiling that secondary summaries drop

At the two highest concentrations tested, the direction of the effect reversed. The paper reports that 10 to the minus 8 and 10 to the minus 7 molar AHK-Cu significantly inhibited follicle elongation, by 14.8 plus or minus 1.2 per cent, giving 2.3 plus or minus 0.18 mm, and by 81.5 plus or minus 40.8 per cent, giving 0.5 plus or minus 0.25 mm, against vehicle-treated controls. That sentence appears in the results section and in no version of the abstract.

The inhibition at the top concentration has a standard error of 40.8 on a point estimate of 81.5, so how much elongation was suppressed is poorly constrained by this experiment; that it was suppressed is what the significance test speaks to. Magnitude and direction come apart here. The window in which elongation increased spans four orders of magnitude and closes, on these data, one order of magnitude above the highest active concentration. Nothing in the paper establishes where that boundary sits in tissue rather than in a well.

A 2026 review in Biomedicines tabulates the concentration range as 10 to the minus 13 through 10 to the minus 7 molar against the endpoint of follicle elongation, which is the tested range rather than the effective one, and does not mention the inhibition. Reporting downstream has been uneven in one direction. A 2021 review in FASEB BioAdvances goes further and attributes increased vascular endothelial growth factor expression and reduced TGF-beta-1 to AHK-Cu, citing the same paper. Those two properties are stated in the opening sentence of the Pyo abstract, which calls its subject the tripeptide-copper complex and names no peptide there; the measurements themselves were made on GHK-Cu, by Pollard and colleagues in 2005 (PMID 15655171) and McCormack and colleagues in 2001 (PMID 11176716). Neither was measured for AHK-Cu in the 2007 experiments.

The rodent data are in a patent

United States patent 6,017,888, with a priority date of 17 June 1994 and granted 25 January 2000, is a continuation of Ser. No. 08/638,889, filed 19 July 1996 and abandoned, which was itself a continuation of Ser. No. 08/261,475, the application that issued as US 5,538,945. Its four granted claims cover a method of stimulating hair growth using a peptide-copper complex in which R2 is histidine or arginine and in which, per claim 1, R3 is a chemical moiety joined to R2 by an amide bond, wherein R3 is --NH2, an alkylamino moiety having from 1-20 carbon atoms or an arylamino moiety having from 6-20 carbon atoms. AHK:Cu, whose lysine carries a free carboxylate, appears in the worked examples rather than within the literal scope of those claims. Patent examples are published documents, which is why they are recorded here, but they carry no peer review, no registered protocol, no described blinding and no statistical tests.

Example 12 clipped the backs of telogen-phase C3H mice, 60 days old, and injected peptide-copper complexes intradermally at two sites per animal. AHK:Cu at a 2:1 peptide-to-copper ratio produced hair growth in 1 of 4 mice at 0.75 mg per injection, over a region under 1 cm in diameter, and in 4 of 4 mice at 1.5 mg, over a region above 1 cm, assessed at day 21. Example 13 repeated that design at a 1.1:1 ratio in order to test D-amino-acid substitutions, dosing 1.2 and 1.8 micromoles per injection and reporting area of hair growth rather than an incidence fraction: 3.07 plus or minus 0.76 and 3.24 plus or minus 1.17 for AHK:Cu, alongside AH-(D)K:Cu and (D)A-HK:Cu arms. Group sizes are not stated in that table.

Example 14 applied the complex topically twice daily on weekdays to female C3H mice aged 60 to 65 days, in a vehicle of 50 per cent propylene glycol, 30 per cent ethanol and 3 per cent nonoxynol-9, at approximately 0.1 mL per application. Per cent treatment area at 0.1 per cent w/w was 1.29 plus or minus 1.29 at day 20, 23.07 plus or minus 18.84 at day 27 and 90.14 plus or minus 2.96 at day 34; at 0.5 per cent w/w it was 75.87 plus or minus 7.64, then 100, then 100. Group sizes are not stated in that table either. The comparison arm was AHF:Cu, and the patent's own reading of the pair is that a hydrophilic third residue is more active topically and less active by injection.

Example 16 is the only experiment that states group sizes: 8 to 9 rat pups per group. Sprague-Dawley pups aged 8 days were held in five litters of 10 to 11, given a single intradermal injection of AHK:Cu at 1:1 in saline, then seven consecutive daily intraperitoneal injections of cytosine arabinoside at 25 mg/kg. Mean alopecia grade at day 10 on a 0 to 4 scale was 0.00 for saline alone, 4.00 for Ara-C with saline, and 3.25, 2.38, 1.44 and 1.11 at AHK:Cu doses of 0.05, 0.10, 0.25 and 0.50 mg. Histology found anagen follicles at the injection site and dystrophic follicles away from it. The table's mg/kg column does not reconcile with its mg column: 0.05 mg is listed as 35 mg/kg alongside 0.50 mg at 35.0 mg/kg.

Copper chemistry, and the regulatory record

No copper-binding constant has been published for AHK. For the neighbouring peptides it has: Trapaidze and colleagues reported a conditional Cu(II) dissociation constant of 7.0 plus or minus 1.0 times 10 to the minus 14 molar for GHK at pH 7.4 by isothermal titration calorimetry, against 2.6 plus or minus 0.4 times 10 to the minus 14 for DAHK, and Hureau and colleagues solved X-ray and solution structures for both complexes. Searches of Europe PMC pairing Ala-His-Lys with copper, with stability constants, and with complex formation returned those DAHK and GHK papers and nothing for the alanine tripeptide. Neither a partition coefficient nor a forced-degradation profile was located, both of which exist for GHK-Cu through Badenhorst and colleagues.

Conato and colleagues studied copper complexes of GHK and two synthetic analogues by potentiometry, calorimetry and spectroscopy, and concluded that complex formation driven by the first residue is necessary while the second residue plays no specific role and the free side chain of the third is of fundamental importance. That is the closest published argument for why an alanine substitution at position one might matter, and it was made without testing AHK. The structural inference that the AHK copper site is equivalent to the GHK site, on which the secondary accounts rest, has no titration behind it and is recorded in the unsourced list below.

Regulators have not assessed this molecule under any name. The Cosmetic Ingredient Review panel's assessment, finalised in June 2014 and published in the International Journal of Toxicology in 2018, covers Tripeptide-1, Copper Tripeptide-1, Bis(Tripeptide-1) Copper Acetate, Manganese Tripeptide-1, the hexapeptide-12 series and Palmitoyl Tetrapeptide-7. A full-text search of that report finds no instance of Tripeptide-3, of AHK, or of Ala-His. The FDA substance register carries the free tripeptide and prezatide copper, the GHK complex, with no entry for the AHK complex. ClinicalTrials.gov returns zero studies for AHK-Cu and zero for alanyl-histidyl-lysine. It is not an approved medicine anywhere.

What is not known

No human study of AHK-Cu has been conducted. ClinicalTrials.gov returns zero registered studies under the compound name or the full peptide name, so nothing exists on pharmacokinetics, dose ranging, safety or efficacy in people, and the ex vivo follicle work does not substitute for any of it because a cut follicle in a well has no circulation, no immune system and no skin barrier above it. In animals the position is narrower still: four patent examples in mice and rat pups, unrefereed, without stated statistics, two of them lacking group sizes, all assessing hair growth by area, by incidence or by a four-point alopecia grade rather than by follicle counts or histomorphometry. Nothing characterises absorption through intact skin, systemic exposure after topical application, or whether repeated delivery of a chelated copper load perturbs copper homeostasis, in any species. The copper site itself has never been titrated for this peptide. There is no chronic administration study, no reproductive or developmental data, no immunogenicity work, no forced-degradation or stability profile, and no comparison against GHK-Cu or any active comparator under matched conditions.

Questions