JOURNAL OF COSMETIC SCIENCE 212 The dermis is the thicker, deeper layer of the skin underlying the epidermis, and is mainly composed of such connective tissues as collagen and elastic fi bers. Among these, collagen fi ber is the main component of the extracellular matrix (ECM), as the repre- sentative connective tissue that comprises about 90% of the dermis collagen has a di- rect infl uence on skin tension. Maintenance of the collagen structure is related to the intrinsic aging and photo-aging processes of the skin (4,5). Therefore, a variety of in- vestigations have been focused on protection against skin aging through the inhibition of collagenase activity, which disintegrates the ECM proteins (6). Matrix metallopro- teinases (MMPs) are an enzyme family to digest basal membranes and ECM. MMPs can be divided into four categories according to the preferred substrate: collagenases (MMP- 1,8,13,18), gelatinases (MMP-2,9), stromelysins (MMP-3,10), and membrane-type MMPs (MMP-14,15,16,17) (7). It is therefore suggested that an evaluation of the in- hibitory effi cacy of materials for MMP-1 gene expression could be used as a screening method to uncover promising candidates that would inhibit the degradation of colla- gens (8–10). Platycarya strobilacea Sieb. et Zucc. (family: Juglandaceae) is widely distributed through- out Korea, Japan, Taiwan, and China (11). It is commonly called Kool-pee tree in Korea. The root bark of this species have been used in traditional oriental medicine for the treat- ment of infl ammation and diarrhea. Its leaves have also been used for the treatment of acute suppuration or abscess on the skin and head region (12). The constituents of the bark and leaves have been studied. It is known that euglon and tannin exist in the bark and fruit, respectively (13). We have previously screened various plants for free-radical scavenging activity and the inhibition of elastase activity (14), and have also reported the anti-wrinkle activity of zi- yuglycoside I isolated from a Sanguisorba offi cinalis root extract, which showed a signifi - cant inhibitory effect against aging (15). In this study, we investigated free-radical scavenging activity, elastase inhibition activity, reduction of MMP-1 mRNA expression in vitro, and type I collagen synthesis in normal human fi broblasts for the development of potential anti-wrinkle ingredients as raw materials for use in cosmetic products. We elucidated the main components from the P. strobilacea fruit extract. MATERIALS AND METHODS REAGENTS AND EQUIPMENT All reagents were purchased from Sigma-Aldrich (St. Louis, MO). Other commer- cially available reagents and solvents were used as received. Human fi broblasts were acquired from ATCC (American Type Culture Collection, CRL-2076). The com- pound 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) was used as received. Medium, fetal bovine serum, and antibiotics were purchased from Life Technologies (Grand Island, NY). The elastase inhibitory activity assay was per- formed with a UV-spectrophotometer (Hewlett-Packard, HP-8453). An ELISA reader (Tecan, A-5082, Austria) and PCR (Bio-Rad, MycyclerTM thermal cycler) were used in the cytotoxicity assay and in the assay on the inhibition of MMP-1 expression.

ANTI-WRINKLE ACTIVITY OF P. STROBILACEA 213 EXTRACTION AND FRACTIONATION P. strobilacea fruits were collected from Jeju island and authenticated by Dr. C. S. Kim, the director of Halla Arboretum in Mt. Halla National Park. The collected fruits were dried in the shade at room temperature and stored in a dark, cold room until needed. The dried whole fruits (180 g) were extracted twice with 70% (v/v) ethanol (20 times as much as the weight of the dried fruits) for 24 h at room temperature. The extract of the fruits was fi ltered through fi lter paper (Whatman, No. 5) and then evaporated at 60°C. The fi ltrate was evaporated to produce the crude P. strobilacea fruit extract as a powder (49.5 g, 27.5% yield). The crude P. strobilacea fruit extract (20 g) was dissolved in distilled water at a concen- tration of 2% (w/w), and the aqueous suspension was successively extracted with n-hexane, ethyl acetate, and n-BuOH. Each fraction was evaporated to dryness to give a residue [The yields were 0.72 g (3.6%) of the hexane fraction 1.66 g (8.3%) of the EtOAc fraction 3.0 g (15.1%) of the n-BuOH fraction and 14.1 g of the H2O fraction (70.5%).] The n-BuOH fraction (1.0 g) was chromatographed on octadecylsilyl-silica gel (ODS, Capcell Pak C18 UG120, 5 μm, 20 × 250 mm, Shiseido Co., Ltd. Tokyo) by preparative HPLC with isocratic eluent (CH3CN:H2O=45:55) at a fl ux of 14.0 ml/ min. Two major compounds on prep-HPLC, ellagic acid and ellagic-4-O-xyropyrano- side, were isolated (64 mg and 22 mg, respectively) (mp 254–258°C (dec.)) and identi- fi ed by 1 H-NMR and 13 C-NMR, the results of which were in agreement with previously published data (12). FREE-RADICAL SCAVENGING ACTIVITY TEST The assay for free-radical scavenging capacity was carried out according to the method that has been reported previously by Blois (16). The DPPH (1,2-diphenyl-2-picrylhydra- zyl) radical shows a deep violet color due to its unpaired electron, and radical scavenging capacity can be followed spectrophotometrically by the loss of absorbance at 525 nm. In brief, a 0.2 mM DPPH 95% ethanolic solution (1 ml) was added to a sample of the stock (2 ml). Each sample solution was diluted with a 70% ethanolic solution to fi nal concen- trations of 10, 5, and 1 μg/ml, and the samples were then agitated. The optical density at 525 nm was measured after 10 min with a UV/Vis spectrophotometer. The free-radical scavenging activity of each sample was calculated according to the following formula: DPPH radical scavenging activity (%) = [1– (ODs – ODb)/ODc] × 100 where ODs is the absorbance of the experimental sample, ODb is the absorbance of the blank, and ODc is the absorbance of the control at 525 nm. The results are reported in terms of SC50 (the concentration needed to reduce 50% of DPPH). BHT (di-t-butyl hy- droxy toluene), a representative antioxidant, was used as a control. ELASTASE INHIBITION ACTIVITY The elastase activity was evaluated according to the method previously reported by Kraunsoe et al. (17). In order to evaluate the inhibition of elastase activity, the amount of