Peptides in cosmetics: types, mechanisms and real evidence

What cosmetic peptides are

Cosmetic peptides are short chains of amino acids (three to eight or ten residues) added to creams, serums and patches to influence the look of skin: wrinkles, firmness, texture. In this context they are not hormones or drugs; they are cosmetic ingredients whose scientific rationale is to imitate messages that skin itself uses to communicate. A classic dermatology review describes them as molecules that promise to prevent or treat aged skin by acting on collagen, muscle contraction or enzyme activity¹.

Almost all carry a chemical modification: a fatty acid (palmitoyl-) to make the molecule less hydrophilic, or an acetyl group (acetyl-) for stability. That is why on a cream label you will read them as "palmitoyl pentapeptide-4" or "acetyl hexapeptide-8", with the number indicating the version of the sequence. Someone reading a bottle with those names is reading the catalog of a very small group of molecules that repeat from product to product²³.

The four types of cosmetic peptides

The literature classifies them by what they are supposed to do. A reference review uses four categories: signal peptides, neurotransmitter-inhibiting peptides, enzyme inhibitors and carrier peptides¹; other reviews fold them into three (signal, neurotransmitter and carrier)²³. In practice:

• Signal peptides: aim to make the fibroblast produce more collagen, elastin or fibronectin. Palmitoyl pentapeptide-4 (Matrixyl) and related derivatives belong here¹³.
• Neurotransmitter inhibitors: aim to dampen the muscle contraction behind expression lines; the best known example is argireline (acetyl hexapeptide-8)¹⁹.
• Enzyme inhibitors: aim to slow enzymes that degrade the matrix, such as metalloproteinases¹.
• Carriers: deliver a trace element to skin; the classic example is the copper complex GHK-Cu¹¹³.

Signal peptides and Matrixyl: from matrikines to the cream

The idea of a signal peptide comes from the concept of matrikines: when cells degrade matrix proteins they release fragments that act as new messages for neighboring cells, so a regulatory loop exists between the matrix and the fibroblast⁴. In 1993, Katayama's group showed that the pentapeptide Lys-Thr-Thr-Lys-Ser (KTTKS), part of the type I collagen propeptide, was the minimum sequence able to stimulate collagen and fibronectin production in mesenchymal cells in culture⁵.

Attached to a palmitic acid, that pentapeptide is the palmitoyl pentapeptide-4 sold as Matrixyl. Its reference trial included 93 women aged 35 to 55: over 12 weeks, a moisturizer with 3 ppm of the peptide improved wrinkles and fine lines vs the same moisturizer without peptide, with good tolerance⁶. A randomized trial with 196 women compared a cosmetic regimen with niacinamide, peptides and retinyl propionate against 0.02% tretinoin: at 8 weeks the cosmetic regimen improved wrinkles more, at 24 weeks results were comparable and tolerability was better⁷. The key nuance is that it was a multi-active regimen, not the peptide alone. Each trial is detailed in Matrixyl and signal peptides.

Argireline: "topical Botox" and what the studies show

Argireline (acetyl hexapeptide-8, formerly called acetyl hexapeptide-3) is an acetylated hexapeptide, Ac-EEMQRR-NH2, designed to interfere with the SNARE complex that releases acetylcholine. In the original study, an emulsion with 10% of the peptide reduced wrinkle depth by up to 30% after 30 days in volunteers, although in the laboratory its efficacy was far lower than that of botulinum toxin⁹. In a randomized, placebo-controlled trial in 60 Chinese subjects, overall periorbital anti-wrinkle efficacy was 48.9% with argireline vs 0% with placebo after 4 weeks¹⁰.

The problem is skin. An FDA laboratory measured penetration of an emulsion with 10% of the peptide in human skin: most of it was removed by washing, what was retained stayed mainly in the stratum corneum, only 0.01% reached the epidermis and no peptide was detected in the dermis¹¹. That does not cancel the clinical results, but it suggests part of the effect may depend on surface mechanisms or the vehicle. On safety, the Cosmetic Ingredient Review expert panel considered acetyl hexapeptide-8 safe in cosmetics at concentrations up to 0.005% and said data were insufficient above that figure¹². More in SNAP-8 and argireline.

Copper GHK-Cu in skin: a lot of literature, few trials

GHK-Cu is a tripeptide (glycyl-histidyl-lysine) bound to copper. A 2015 review describes how GHK in plasma declines with age and presents it as a modulator of several cellular pathways in skin regeneration¹⁴, and a 2018 review gathers its regenerative and protective actions with gene-expression data¹³. In culture and animal models the base is broad. In humans it is thinner: after CO2 laser in 13 patients, the copper complex showed no objective benefit, only greater satisfaction¹⁵; and a 2026 systematic review of GHK-Cu with microneedles concluded that the clinical literature in humans "remains thin and fragmented"¹⁷.

In vitro, the copper from the tripeptide crossed isolated human skin¹⁶, a penetration point in this peptide's favor. Its mechanism is covered in GHK-Cu: the mechanism of action, step by step. One fact is worth knowing when reading that literature: according to the 2026 review, about 35% of primary GHK-Cu studies list Pickart or Margolina, linked to a cosmetics company, as authors¹⁷.

Peptides in creams: do they get where they need to go?

The underlying question for any peptide in a cream is how much crosses the stratum corneum. Bos and Meinardi proposed that a compound must weigh less than 500 Da to cross it¹⁸; palmitoyl pentapeptide-4 weighs about 802 Da. Palmitoylation changes the math by making the molecule far less hydrophilic, and in newer strategies microneedles increased the signal of several labeled cosmetic peptides 2- to 22-fold in ex vivo human skin¹⁹. These are laboratory data: they are not a clinical result.

What the evidence says overall

A 2019 systematic review asked how many matrix-stimulating peptides had been tested clinically: it found 15 studies, only 6 with a placebo control, 5 double-blind and 9 enrolling only women²⁰. In the literature itself, trials of combinations with niacinamide or retinoids do not isolate the peptide's effect; the Manchester trial with 60 subjects and an identical vehicle showed wrinkle improvement in 43% vs 22% at 6 months (difference not significant) and in 70% vs a projected 33% at 12 months, with increased fibrillin-1 in biopsies⁸. The authors cautioned that very few "anti-ageing" products have undergone a double-blind, vehicle-controlled trial⁸.

A 2026 meta-analysis of 19 randomized trials with 1,341 participants (oral and topical peptides) found a modest effect on wrinkles, driven mainly by oral peptides, and good tolerability²¹. On safety, a 2026 bioinformatic framework found no toxicity signals for palmitoyl pentapeptide-4, which only shares sequence homology with matrix proteins²². The balance: a favorable tolerability profile, a probable but modest effect, and trials that are small, short and often industry-linked.