553 TERPENE CONJUGATION The donor phases were also surveyed for the remaining peptides. Results showed that around 96% of Cit-KTTKS was recoverable in the donor phase, while in the case of Pal- KTTKS, the peptide amount in the donor phase at the end of study reached 76% of applied peptide, about 24% disappearance from the donor phase, of which only about 4% of Pal-KTTKS was detected in the receptor phase (Table I). Our stability studies did not show signs of instability therefore around 20% of the applied dose was somehow trapped by n-hexadecane. Epidermal membrane permeation studies. Epidermal permeation of KTTKS, Cit-KTTKS, and Pal-KTTKS were studied using the human abdominal epidermis. Since Per-KTTKS did not pass across the membrane model, it was not included in the epidermal membrane permeation studies. KTTKS and Pal-KTTKS did not appear in the receptor phase of epidermal membrane studies in detectable amounts. In agreement with this study, Choi et al. showed that KTTKS and Pal-KTTKS did not permeate through intact hairless mouse skin (4). Contrary to these findings, we detected Cit-KTTKS in the receptor phase the amount of permeated peptide after 48 h was about 2.6% of the applied dose (Table II). Peptide amount in the donor phases was also evaluated after 48 h (Table II). As seen, the peptide amount in the donor phase did not change considerably at the end of study for KTTKS and Cit-KTTKS, but in the case of Pal-KTTKS, an intense reduction of peptide amount was observed in the donor phase. KTTKS has clog P of –3.72 and molecular weight of 563.6 Da (18). Since the action site for such peptides is the dermis, a high permeation is necessary. Although KTTKS has a relatively proper molecular weight for skin permeation, this peptide did not pass across the SC barrier due to its hydrophilic nature. Therefore, it seems that this peptide skin permeation could be increased through increasing its tendency to enter the SC lipophilic Table I Permeation of Peptides Through PVDF-Hexadecane Membrane Modela Peptide Peptide permeated across the membrane (% of initial dose) Peptide detected in the donor phases (% of initial dose) Estimated permeability coefficient (cm/h) KTTKS ND 101.4 ± 0.8 – Pal-KTTKS 3.88 ± 1.7 76.0 ± 10.0 17 × 10−4 Cit-KTTKS 0.66 ± 0.15 95.9 ± 1.7 2.9 × 10−4 Per-KTTKS ND 97.1 ± 2.9 – ND: not detected. a Data are mean ± SD, n = 3. Table II Permeation of Peptides Through Human Epidermal Membranea Peptide Peptide permeated across the membrane (% of initial dose) Peptide detected in the donor phases (% of initial dose) Estimated permeability coefficient (cm/h) KTTKS ND 102.1 ± 2.3 – Cit-KTTKS 2.64 ± 0.24 95.2 ± 0.8 7.3 × 10−4 Pal-KTTKS ND 51.7 ± 3.0 – ND: not detected. a Data are mean ± SD, n = 3.

554 JOURNAL OF COSMETIC SCIENCE barrier. Pal-KTTKS (cLogP: 3.72) is a peptide conjugate with more lipophilicity in comparison to KTTKS. However, its molecular weight (802 Da) is more than that of KTTKS (18) and its permeation across the SC can still be hindered by diffusion. Our results showed that Pal-KTTKS was lost from the donor phase and did not appear in the receptor phase, whereas the number of other peptides in the donor phases did not change considerably. Since there was no sign of peptide in the receptor phase, it might be concluded that Pal-KTTKS was trapped in the epidermis and, considering its lipophilicity and membrane model data, it was possibly trapped in the SC intercellular lipids. The results are in good agreement with the membrane model data as described earlier. Estimated peptide retentions by n-hexadecane and epidermis were about 20% and 48% of the applied dose respectively. Palmitic acid, a long-chain fatty acid with a linear structure, is a component of intercellular lamellar lipid matrix of human SC (19) and fits well into this structure. KTTKS also has a linear structure, making entrapment of Pal-KTTKS in the lamellar structure of the SC is a possibility. This finding agrees with the suggestion by Moghimi et al. about the accommodation of fluorouracil and straight-chain alcohol terpene complexes within the lipid bilayers (13). Choi et al. also showed that Pal-KTTKS was trapped by mouse skin in the amount of 14.6% of the applied amount (4), which is in reasonable agreement with the present study. In addition to tendency (lipophilicity), diffusion also affects the permeation of compounds through skin. Diffusion depends on different parameters including molecular weight and structure. Cit-KTTKS was the only peptide that was detected in the receptor phase with estimated permeability coefficient of about 7.3 × 10−4 cm/h that is more than its estimated kp in the membrane model study. It possesses a molecular weight of 715.9 Da and cLogP of −0.08. This conjugate is less lipophilic than Pal-KTTKS, but smaller in size. Citronellic acid has a 10-carbon chain and is shorter than palmitic acid (16 carbons in length). This lower molecular weight (around 90 Da) could make this conjugate a better candidate for permeation, considering that diffusion through skin is non-Stokesian and is highly sensitive to molecular weight and structure. Finally, much data are available on enhancement effects of terpenes toward permeation of drugs across the skin (12,13) or other biological barriers including burn eschar (20,21) and tumor cells (22) by different mechanisms such as membrane disruption and complexation. Enhancement effects of terpenes toward permeation of 5-fluorouracil (a hydrophilic molecule) across human skin depends on degree of complexation and solubilization enhancement of terpenes (13). Considering these and our present results, it seems that terpenes are good candidates for increasing permeation of peptides across the skin either by conjugation or other strategies. CONCLUSIONS These results suggest that terpene conjugation might be a good approach to increase peptide skin permeation. Data also suggest that lower molecular weight conjugates with optimized polarity and proper structure might offer a better choice than conjugates in which only polarity is considered. Our data also show that PVDF-filled hexadecane might provide a good model for the screening of peptide permeation and SC retention.