Palmitoyl hexapeptide-12 is one of the few cosmetic peptides whose sequence is a direct copy of part of a skin protein. That makes it unusually easy to trace — and tracing it leads somewhere the product copy does not go.
The molecule
PubChem CID 10212452 gives its full chemical name as N-palmitoyl-L-valylglycyl-L-valyl-L-alanyl-L-prolylglycine: Pal-VGVAPG, formula C38H68N6O8, 737.0 g/mol. The palmitoyl group is the 16-carbon fatty-acid chain that palmitoyl tripeptide-1, Matrixyl and most "palmitoyl" peptides carry to help them into the lipid-rich outer layer of skin.
The six residues after it, VGVAPG, are a sequence elastin repeats along its length. When elastin is broken down — by ageing, UV exposure or inflammation — fragments carrying this sequence are released. Biologists call such fragments elastokines: pieces of matrix that bind a receptor on cells, the elastin receptor complex, and change what those cells do.
At 737.0 g/mol it is about 1.5 times the 500-dalton figure usually quoted as the practical ceiling for passive penetration. Unlike most of this site's heavier entries, its penetration has actually been measured.
Two papers on the ingredient, and neither is a human trial
Searched on 2026-09-21, with a control query validated in the same session:
| Search (title/abstract) | Records |
|---|---|
palmitoyl hexapeptide-12 |
2 |
VGVAPG — the sequence it copies |
111 |
VGVAPG + MMP / metalloproteinase |
23 |
VGVAPG + tumour / cancer / metastasis |
16 |
VGVAPG + skin / fibroblast / dermal |
23 |
The penetration study. Ligorio and colleagues, ACS Applied Bio Materials 2025, used orbital-trapping secondary ion mass spectrometry to follow palmitoyl hexapeptide-12 through full-thickness human skin outside the body. They report it permeates both the stratum corneum and the epidermis, where it forms a gel using the skin's own ions, and that it helped restore the mechanics of skin whose lipids had been stripped. It is a careful measurement and one of very few that follows a cosmetic peptide into skin at all. Four of its authors are listed at The Estée Lauder Companies; the paper states that the authors declare no competing financial interest. Both facts are in the published record.
The safety screen. Bjerke and colleagues, Current Research in Toxicology 2026, from Procter & Gamble, proposes a bioinformatic framework for peptide safety and tests it on several peptides. Palmitoyl hexapeptide-12 showed sequence homology with skin matrix proteins — elastin, as expected — and no toxin or allergen flag. The paper's conflict statement notes that its P&G authors work for a company that makes products containing some of the compounds studied.
There is no published trial in which people applied palmitoyl hexapeptide-12 and something about their skin was measured. We cover the TriHex system, which uses the unpalmitoylated hexapeptide-12 with tripeptide-1 and does have small clinical trials, on its own page; the penetration result above was obtained with the palmitoylated form and does not transfer to it.
What the sequence does where it comes from
The case for the ingredient is the biology of the fragment it copies, so the fragment's literature is the evidence that matters. It is 111 records deep, and it is not a simple story of repair.
Elastokines carrying VGVAPG do stimulate fibroblasts, the cells that make matrix. They also, in a substantial part of that literature, raise the enzymes that break matrix down:
- In endothelial cells, (VGVAPG)3 and κ-elastin increased expression of the matrix metalloproteinase MT1-MMP through a nitric-oxide pathway (Fahem et al., Int J Biochem Cell Biol 2008).
- In lung cancer cell lines, elastin-derived peptides including a synthetic VGVAPG increased pro-MMP-2 and uPA and the invasive capacity of already-invasive cells (Toupance et al., Clin Exp Metastasis 2012).
- In fibroblasts and melanoma cells, (VGVAPG)3 accelerated proliferation of both — a response the authors describe as running through the elastin receptor and the MEK/ERK pathway (Chatron-Colliet et al., Histochem Cell Biol 2015).
All of this is cell work, with the free peptide, at laboratory concentrations. None of it tested a cosmetic and none of it shows harm from one. What it establishes is narrower and still worth knowing: in its native context, VGVAPG is released when elastin is being destroyed, and cells read it partly as a signal that remodelling is under way — which includes more demolition as well as more building.
The group that designed around the problem
The clearest evidence that researchers take the double edge seriously comes from a research group in Reims whose members also appear in the MMP papers above. When they designed a skin-regeneration peptide around the motif, they did not use VGVAPG alone. Their "trifunctional peptide" joined (VGVAPG)3 to a tripeptide that occupies MMP-1's binding site and a linker that acts as a decoy substrate for urokinase (Attia-Vigneau et al., J Invest Dermatol 2014). In fibroblasts and skin explants, the whole construct raised matrix production while inhibiting MMP-1 — and the paper compares it with its separate parts.
That construct later went into a split-face trial: 22 volunteers applied it to one side of the face and placebo to the other, twice daily for 28 days, before facelift surgery, and the removed skin was examined (Abraham et al., Am J Transl Res 2023). The authors report firming and reduced wrinkle depth on the treated side. Two of them are employees of Regentis Pharma, the company behind it, and two more are its consultants. It is a result about a three-part molecule, not about palmitoyl hexapeptide-12, and the design choice is the point: the people who know this motif best built a brake into it.
What this supports saying
- Palmitoyl hexapeptide-12 is a palmitoylated copy of VGVAPG, a repeat unit of elastin, weighing 737.0 g/mol.
- One ex vivo study, with cosmetics-company authors, followed it into the epidermis of donated human skin. No human trial has tested it.
- The elastin fragment it copies stimulates fibroblasts and, in many cell studies, also raises matrix-degrading enzymes and tumour-cell activity. None of those studies involved a cosmetic.
- The one research group that turned the motif into a clinical skin product added an MMP-1 inhibitor to it.
Where to go next
- TriHex: tripeptide-1 and hexapeptide-12 — the unpalmitoylated form, in a system with small clinical trials.
- Palmitoyl tripeptide-1 — the peptide it is most often paired with, and a far larger file.
- OLEHENRIKSEN Pout Preserve: the formula — a lip treatment that declares both.
- What peptides do for skin — the four mechanism classes, and where each has been measured in living skin.
Sources
- PubChem CID 10212452 — palmitoyl hexapeptide-12, read 2026-09-21.
- Ligorio C, Tavasoli E, Karaman-Jurukovska N, et al. Noninvasive monitoring of palmitoyl hexapeptide-12 in human skin layers. ACS Appl Bio Mater 2025;8(3):2340-2355.
- Bjerke DL, Li J, Gao Y, et al. A framework for the safety evaluation of peptides in cosmetics. Curr Res Toxicol 2026;10:100291.
- Fahem A, Robinet A, Cauchard JH, et al. Elastokine-mediated up-regulation of MT1-MMP is triggered by nitric oxide in endothelial cells. Int J Biochem Cell Biol 2008;40(8):1581-96.
- Toupance S, Brassart B, Rabenoelina F, et al. Elastin-derived peptides increase invasive capacities of lung cancer cells by post-transcriptional regulation of MMP-2 and uPA. Clin Exp Metastasis 2012;29(5):511-22.
- Chatron-Colliet A, Lalun N, Terryn C, et al. The elastin peptide (VGVAPG)3 induces the 3D reorganisation of PML-NBs and SC35 speckles architecture, and accelerates proliferation of fibroblasts and melanoma cells. Histochem Cell Biol 2015;143(3):245-58.
- Attia-Vigneau J, Terryn C, Lorimier S, et al. Regeneration of human dermis by a multi-headed peptide. J Invest Dermatol 2014;134(1):58-67.
- Abraham JD, Vaissière A, Desouches C, et al. Clinical validation of an elastin-derived trifunctional peptide for skin regeneration. Am J Transl Res 2023;15(7):4620-4628.
- Bos JD, Meinardi MMHM. The 500 Dalton rule. Exp Dermatol 2000.
- PubMed counts run 2026-09-21 via NCBI E-utilities, validated against a control query.
