A topical peptide has to cross the stratum corneum before anything else about it matters, and that single constraint — roughly 500 Da for passive diffusion through intact skin — explains why cosmetic research peptides are overwhelmingly tri-, tetra- and pentapeptides rather than the larger sequences used in injectable work. Topical peptides are formulated as aqueous serums, emulsified creams and alcohol-based scalp solutions, and each vehicle changes how much peptide reaches the viable epidermis. This guide covers the barrier, the four peptide classes used in cosmetic research, how loading percentages translate into concentration, and what the published evidence does and does not support. It describes laboratory and cosmetic research preparations supplied for research use only.
The barrier and the 500-dalton rule
The stratum corneum is 10–20 µm of anucleate corneocytes embedded in a lamellar lipid matrix of ceramides, cholesterol and free fatty acids. Passive permeation runs mainly through that lipid phase, which favours small, moderately lipophilic molecules. The widely cited "500 Dalton rule" — proposed from analysis of topically active drugs and contact allergens — holds that molecules above roughly 500 Da rarely permeate intact skin in useful amounts.
Peptides are hydrophilic and often charged, so they are doubly disadvantaged. Three routes are used to get around this:
- Stay small. GHK-Cu is a tripeptide-copper complex at 403.93 Da, comfortably under the ceiling.
- Add a lipid tail. Palmitoylation attaches a 16-carbon fatty acid to the N-terminus, raising the partition coefficient so the peptide dissolves into the lipid lamellae. Palmitoyl pentapeptide-4 (Matrixyl, 802.05 Da) exceeds 500 Da yet permeates measurably because of that tail.
- Change the vehicle. Ethanol, propylene glycol, glycerol and certain surfactants fluidise the lipid matrix; occlusion under an emulsion raises corneocyte hydration and swells the intercellular route.
Four classes of topical peptides in cosmetic research
| Peptide | Class | MW (Da) | Reported mechanism | Typical formulation loading |
|---|---|---|---|---|
| GHK-Cu | Carrier / signal | 403.93 | Copper delivery; reported stimulation of collagen and decorin synthesis in fibroblast culture | 0.01–0.1% w/v |
| Palmitoyl pentapeptide-4 (Matrixyl) | Matrikine / signal | 802.05 | Collagen-fragment mimic reported to upregulate extracellular matrix synthesis in fibroblast models | 3–8% of a supplied 100–500 ppm solution |
| Acetyl hexapeptide-8 (Argireline) | Neurotransmitter-inhibitor | 888.99 | Competes with SNAP-25 for the SNARE complex in vitro, reported to reduce catecholamine release in cell models | 5–10% of a 5% supplied concentrate |
| Acetyl octapeptide-3 (SNAP-8) | Neurotransmitter-inhibitor | 1075.16 | Elongated Argireline analogue targeting the same SNARE interaction | 5–10% of a supplied concentrate |
| AHK-Cu | Carrier / signal | ~340 | Copper tripeptide studied in hair-follicle models alongside GHK-Cu | 0.01–0.05% w/v |
Note that the two neurotransmitter-inhibitor peptides sit above the 500 Da line and carry no lipid tail. Their permeation in published studies is modest, and this is the most contested part of the cosmetic peptide literature: an in-vitro SNARE-competition result establishes a mechanism in a dish, not delivery to a neuromuscular junction through intact skin.
What the evidence actually supports
GHK-Cu has the deepest record — Loren Pickart's group first isolated it from human plasma in the early 1970s, and subsequent work reported effects on collagen, glycosaminoglycan and decorin synthesis in fibroblast culture and in wound models, with several small human split-face studies on skin appearance. The background is covered in what is GHK-Cu. Matrixyl's evidence is mostly manufacturer-run and fibroblast-based with a small number of controlled human studies. Argireline's is dominated by in-vitro mechanism work plus vendor-sponsored clinical studies with small cohorts. In every case the honest framing is "reported in", not "shown to", and none of these are treatments for any condition.
Worked example: loading, concentration and molarity
Cosmetic labels quote percentages, research protocols want molarity, and the two are trivially convertible once you fix the units.
- Percent to mg/mL. 1% w/v = 1 g per 100 mL = 10 mg/mL. So 0.05% w/v = 0.5 mg/mL = 500 µg/mL = 500 ppm.
- Building a 30 mL serum at 0.05% GHK-Cu. 30 mL × 0.5 mg/mL = 15 mg of peptide. A 50 mg research vial makes three such bottles with material to spare.
- Percent to molarity. 0.5 mg/mL is 0.5 g/L; divide by the molecular weight 403.93 g/mol → 1.24 mM. Fibroblast studies commonly work in the 1–100 µM range, so a 0.05% topical serum is roughly 10–1,000 times the concentration used in culture — before any barrier losses.
- Reading a "10%" label. A serum described as Argireline 10% almost always means 10% of a supplied concentrate, and that concentrate is itself typically 5% peptide in solution. Actual acetyl hexapeptide-8 content is therefore near 0.5% w/v, not 10%. Ask for the peptide content in mg/mL if the number matters to your design.
That last point is the single most common misreading in the category, and it is worth writing into any protocol that compares products.
Vehicle: serum, cream or solution
- Aqueous serums — such as the GHK-Cu face serum and Matrixyl serum — keep hydrophilic peptides in their preferred phase, dry quickly, and give the shortest contact time. Best where the peptide is small and water-soluble.
- Emulsified creams such as the GHK-Cu cream add occlusion, which raises corneocyte hydration and extends contact. The trade-off is a more complex matrix with more opportunity for the peptide to partition into the oil phase and never reach skin.
- Alcohol- or glycol-based solutions such as the GHK-Cu hair and scalp solution spread over hair-bearing skin and exploit the follicular route, which bypasses the stratum corneum entirely and is disproportionately important for hair research.
- Combination formulations such as the GHK-Cu + SNAP-8 serum pair a permeating carrier peptide with a larger one; interpret results as the formulation's, not either peptide's.
Formulation chemistry that breaks peptides
- pH. GHK-Cu's copper coordination is pH-dependent and the complex is most stable near neutral; strongly acidic vehicles (a 3.5 pH vitamin C serum, for example) will strip copper and change the molecule.
- Chelators. EDTA is a routine cosmetic preservative-booster and a copper chelator. It does not belong in a copper-peptide formulation.
- Strong oxidisers and reducers. Methionine, cysteine and tryptophan residues oxidise readily; combining copper peptides with high-percentage ascorbic acid or with retinoids in one layer is a chemistry problem, not a routine.
- Preservation. Water-containing formulations need a preservative system. An unpreserved handmade serum has a working life of days refrigerated, and microbial growth is the usual failure mode long before the peptide degrades.
Common mistakes
- Reading concentrate percentage as peptide percentage. The two differ by roughly an order of magnitude.
- Assuming a large peptide permeates because a small one does. The 500 Da line is a real discontinuity, and a lipid tail — not wishful formulation — is what moves a molecule across it.
- Formulating a copper peptide with EDTA or at low pH.
- Citing fibroblast-culture results as skin outcomes. Concentration in a dish and concentration in viable epidermis after barrier loss are different quantities.
The full range sits in the topical peptide creams and serums collection.