WATER-SOLUBLE ELASTIN AS A COSMETIC MATERIAL 13 0.3 M NaCl (pH 6.9). The fl ow rate was set at 0.2 ml/min. Absorbance was measured at 220 nm. A lyophilized mixture of molecular weight markers ranging from 1,350 to 670,000 Da was used as the molecular mass standard (Bio-Rad Laboratories, Inc., Hercules, CA). MEASUREMENT OF THE TURBIDITY OF HAPA-ELASTIN The coacervation potency of HAPA-elastin was evaluated by measuring its turbidity in distilled water (D.W.). The measurement was performed at 400 nm with a JASCO Ubest V-560 spectrophotometer (JASCO Co., Tokyo, Japan). HAPA-elastin was dissolved in D.W. at a concentration of 50 or 100 mg/ml at a low temperature ( 4°C). The measurements were taken during a progressive heating and cooling (0.5°C/min) cycle from 5°C to 65°C. SEPARATION OF ELASTIN BY THE COACERVATION TECHNIQUE The coacervation technique was used to obtain two fractions from the HAPA-elastin solu- tion. HAPA-elastin was dissolved in D.W. at a concentration 160 mg/ml, which is a suf- fi cient concentration for the self-assembly of water-soluble elastin molecules. The solution was then heated to 65–70°C using a thermostatic bath and centrifuged at 2,350 g for 2 min at 40°C to obtain the two fractions by coacervation. The supernatant was collected as the low-molecular-weight fraction (fraction 1) and the precipitate was the high-molecular- weight fraction (fraction 2). The procedure is outlined in Figure 1. WATER CONTENT OF RECONSTRUCTED HUMAN EPIDERMIS LabCyte EPI-MODEL 12 (LabCyte, Japan Tissue Engineering Co., Ltd. Gamagori, Ja- pan) was cultured from human epidermal cells and stratifi ed using polyethylene tere- phthalate membrane as a supporting layer. It morphologically resembles the human skin. LabCyte consists of a stratum corneum and a viable epidermis that is made up of a granu- lar layer, stratum spinosum, and a basal layer (10). It is used in many skin irritation and corrosion studies as a substitute for laboratory animals (11–13). In this study, the water Figure 1. Separation of HAPA-elastin into its fractions by coacervation.

JOURNAL OF COSMETIC SCIENCE 14 content of the skin was investigated using the three-dimensional cultured human epider- mis of LabCyte. We used a Sep-Pak® C18 cartridge (Waters Co., Milford, MA) to completely remove salts from HAPA-elastin and its fractions. This was important as residual salts may affect the results of moisture-retaining test. HAPA-elastin solutions were prepared in D.W. at the following concentrations: 0.02%, 0.1%, 1.0%, and 8.0% (w/w). The water content of LabCyte was determined by measuring its weight, which is a simple and easy method. Solutions of fractions (fractions 1 and 2) were also prepared in D.W. at the following concentrations: 0.02%, 0.1%, and 1.0% (w/w). A 200 μl aliquot of sample solution was applied on the surface of LabCyte and the mixture was incubated at 32°C. After 30 min, the elastin solution was removed and culture plates were further incubated at 32°C for 60 min to dry the LabCyte. These procedures were repeated three times. The reconstructed epi- dermal tissues were cut from the culture inserts using forceps and their wet weights were measured. The skin tissues were then dried at 60°C–80°C for 24 h. Within this tem- perature range, the samples did not show signs of thermolysis or combustion. Finally, the dry weights of the skin tissue samples were measured. Water content was determined using the following equation: ( ) - = ×100 wet weight dry weight Water content % wet weight (1) TYROSINASE INHIBITION ASSAY Tyrosinase inhibition was measured by the colorimetric method, using a 96-well plate (14). A 20-μl sample of elastin solution (1.0 or 10 mg/ml), 40 μl of tyrosinase (40 U/ml), and 100 μl of phosphate buffer solution (pH 6.8, 1/15 M) were mixed and added to each well in the plate and incubated at 25°C for 3 min. Subsequently, 50 μl of DOPA (2.5 mM) was added to each well and absorbance was recorded after 10 min using a microplate reader at 490 nm (Immuno-Mini NJ-2300, Nalge Nunc International Co., Rochester, NY). The amount of dopachrome produced during the reaction was determined using the following equation: = ×100 sample control Tyrosinase inhibition (%) (1- / ) A A (2) Where “Asample” represents the absorbance of the sample containing elastin, and “Acontrol” represents the absorbance of control. STATISTICAL ANALYSIS The data have been presented as the mean of triplicate measurements for each experiment. The data were analyzed using the Dunnett’s test and statistical signifi cance, was consid- ered at p 0.05.