KHG26792, 3-(naphthalen-2-yl(propoxy)methyl)azetidine hydrochloride, was synthesized at the Organic Chemistry Laboratory (KIST, Seoul, Korea) [13], dissolved in DMSO, and stored at -20℃ as a stock solution (50 mM). Dulbecco's Modified Eagle's Medium (DMEM), Nutrient Mixture F-12 Ham, sodium bicarbonate, HEPES, formaldehyde, insulin, L-ascorbic acid, isoproterenol, hydrocortisone, 12-O-tetradecanoylphorbol-13-acetate (TPA), cholera toxin (CT), synthetic melanin, L-DOPA, and mushroom tyrosinase were obtained from Sigma (St Louis, MO, USA). DMEM/Nutrient Mixture F-12, (3:1, DMEM:F-12), antibiotic-antimycotic (penicillin, streptomycin), trypsin-EDTA, and sodium hydroxide were purchased from WelGENE (Dalseogu, Daegu, South Korea). Cervi cornu Colla (CCC) was obtained from Siwon Herbal Medicine Co. (Jincheon-gun, South Korea), hyaluronic acid (HA) was from LG Life Sciences, Ltd. (South Korea), fetal bovine serum (FBS) was from Hyclone (Logan, UT, USA), and recombinant human epidermal growth factor (EGF) was from Invitrogen Co. (Gibco, Camarillo, CA). Antibodies specific for phospho-ERK1/2 (Thr202/Tyr204, #9101S), total (phosphorylated and nonphosphorylated) ERK1/2 (#9102), phospho-GSK3β (#9336), and phosphorylated Akt (Ser473, #9271) were purchased from Cell Signaling Technology (Beverly, MA, USA). Antibodies specific for tyrosinase (C-19) and actin (I-19) were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). Microphthalmia Ab-1 (C5, MS-771-P0) was obtained from NeoMarkers (Fremont, CA, USA). Secondary antibodies specific for anti-goat IgG (PI-9500), antimouse IgG (PI-2000), and anti-rabbit IgG (PI-1000) were purchased from Vector Laboratories (Burlingame, CA, USA).

The Mel-Ab cell line is a spontaneously immortalized murine melanocyte cell line that produces large amounts of melanin [14]. Mel-Ab cells were incubated in DMEM supplemented with 10% (v/v) FBS, 100 nM TPA, 1 nM CT, 50 µg/ml streptomycin, and 50 µg/ml penicillin at 37℃ in 5% CO₂. CCD-25Sk human fibroblasts were purchased from American Type Culture Collection (ATCC, Rockville, MD, USA). HaCaT human keratinocytes [15] were purchased from Cell Lines Service (Eppelheim, Germany). Cells were grown in DMEM supplemented with 10% (v/v) FBS, 50 µg/ml streptomycin, and 50 µg/ml penicillin at 37℃ in 5% CO₂.

Cell viability was determined using a crystal violet assay. Mel-Ab cells were treated with KHG26792 for 24 h, the medium was removed, and the cells were stained with 0.1% crystal violet in 10% ethanol for 5 min at room temperature. The cells were rinsed four times with distilled water and the crystal violet retained by adherent cells was extracted with 95% ethanol. Absorbance was determined at 590 nm using an ELISA reader (VERSAMax; Molecular Devices, Sunnyvale, CA, USA).

Melanin content was measured using a previously described method, with slight modifications [16]. Cells were treated with various concentrations (0.1~20 µM) of KHG26792 in DMEM containing 10% FBS for 3 days. Cell pellets were then dissolved in 550 µl of 1 N NaOH at 100℃ for 30 min and centrifuged at 13,000×g for 5 min. The resulting supernatants were analyzed at 400 nm using an ELISA reader. Before measurement of melanin content the cells were observed using a phase contrast microscope (Olympus IX50, Tokyo, Japan), and photographed using a DCM300 digital camera (Scopetek, Inc., Hangzhou, China) supported by ScopePhoto software (Scopetek, Inc.).

A cell-free assay system was used to determine the direct effect of KHG26792 on tyrosinase activity. Briefly, 70 µl of phosphate buffer containing KHG26792 was mixed with 20 µl of mushroom tyrosinase (stock solution 53.7 units/ml) and 10 µl of 10 mM L-DOPA. After incubation at 37℃ for 20 min, the absorbance was measured at 475 nm.

Mel-Ab cells were lysed in cell lysis buffer (62.5 mM Tris-HCI [pH 6.8], 2% SDS, 5% β-mercaptoethanol, 2 mM phenylmethylsulfonyl fluoride, protease inhibitors [Complete™; Roche, Mannheim, Germany], 1 mM Na₃VO₄, 50 mM NaF, and 10 mM EDTA). Proteins in the total cell lysates were separated by SDS-polyacrylamide gel electrophoresis using 20 µg of protein per lane, blotted onto polyvinylidene fluoride (PVDF) membranes, and blocked with 5% dried milk in Tris-buffered saline containing 0.5% Tween 20. Blots were incubated with the appropriate primary antibodies at a dilution of 1:1000 and then with horseradish peroxidase-conjugated secondary antibody. Bound antibody was identified using an enhanced chemiluminescence testing kit (Thermo Scientific Inc., Bremen, Germany). All immunoblot images were obtained using a LAS-1000 lumino-image analyzer (Fuji Film, Tokyo, Japan).

Dermal equivalents were prepared as previously described with some modifications [17]. In brief, type 1 collagen extracted from rat tail tendons was dissolved by stirring in 1/1000 glacial acetic acid at 4℃ for 48 h. Dermal equivalents were made by mixing the following solutions: 1% type 1 collagen solution, 10×medium (DMEM:Nutrient Mixture F-12 Han=3:1), 10×reconstitution buffer (0.05 N NaOH, 0.26 mM NaHCO₃, 200 mM HEPES), 5 mg/ml hyaluronic acid (HA), and 0.33 g/ml CCC. Dermal equivalents including CCC were prepared by composition, as described in a previous study [12]. All dermal equivalents contained 3×10⁵ fibroblasts. Three milliliters of this mixture was transferred into 24-mm transwell inserts with a 3.0 µm pore polycarbonate membrane (Corning, Inc., USA), which was then placed in the incubator at 37℃ for gelation.

To reconstruct pigmented SEs, dermal equivalents were treated with 0.03 mg/ml collagen type IV and 1×10⁵ Mel-Ab cells were seeded onto the dermal equivalents. HaCaT cells (1×10⁶ cells) were seeded onto the dermal equivalents and cultured submerged for 1 day, and then at the air-liquid interface for 12 days. The growth medium was DMEM/F-12 supplemented with 5% FBS, 0.4 µg/ml hydrocortisone, 1 M isoproterenol, 25 µg/ml ascorbic acid, and 5 µg/ml insulin. For the submerged culture, 1 ng/ml epidermal growth factor (EGF) was added to the growth medium, whereas 10 ng/ml EGF also was added for the air-liquid interface culture. The medium was changed three times per week. Every time the medium was changed, 10 or 50 µM KHG26792 was added to the new solutions.

After 13 days, the SEs were fixed in 10% formaldehyde for 1 day, embedded in paraffin, and 4- to 6-µm sections were prepared. For morphologic analysis, sections were stained with hematoxylin and eosin (H&E). To detect melanin, a Fontana-Masson stain kit (ScyTek Laboratories, Logan, UT, USA) was used according to the manufacturer's instructions.

The statistical significance of intergroup differences was assessed by analysis of variance (ANOVA) and Student's t-test. p-values<0.05 were considered statistically significant.

The structure of KHG26792 is displayed in Fig. 1. To evaluate the safety of KHG26792 in Mel-Ab cells, cells were treated with different concentrations of KHG26792 (0.1~50 µM) and cell viability was measured using a crystal violet assay. As shown in Fig. 2A, KHG26792 had no cytotoxic effects at concentrations between 0.1 and 20 µM. Higher concentrations of KHG26792 (50 µM) were found to be slightly cytotoxic compared with the untreated control. Therefore, concentrations between 0.1 and 20 µM were used in the following experiments. To examine the effects of KHG26792 on melanogenesis, cells were incubated with KHG26792 (0.1~20 µM) for 3 days. As shown in Fig. 2B, cell morphology indicated that the amount of melanin decreased after treatment with KHG26792 in a dose-dependent manner. Measurement of melanin content confirmed that KHG26792 decreased the amount of melanin in a dose-dependent manner (Fig. 2C). As tyrosinase is the key enzyme in melanin synthesis [18], tyrosinase activity was measured in a cell-free system after the addition of KHG26792. The results indicated that KHG26792 did not block tyrosinase directly (Fig. 2D).

Because KHG26792 did not inhibit tyrosinase directly, the influence of KHG26792 on the expression of melanogenic proteins was analyzed by Western blot analysis, which showed that KHG26792 downregulated MITF and tyrosinase at 24 h (Fig. 3A). Decreased expression of MITF and tyrosinase was also observed at 48 h. We next investigated the status of melanogenesis-associated signal transduction pathways after treatment with 10 µM KHG26792. The resulting data indicate induction of ERK phosphorylation, which peaked at 2 min and lasted for a minimum of 30 min, whereas phosphorylation of Akt and GSK3β did not change (Fig. 3B). It is well known that activating the ERK pathway results in downregulation of MITF and decreased melanin synthesis, therefore we measured melanin content in the presence of PD98059, an ERK pathway inhibitor. As shown in the photographs in Fig. 4A, PD98059 rescued the KHG26792-induced hypopigmentation and decreased melanin content resulting from treatment with KHG26792 (Fig. 4B).

The pigmented SE was prepared as described in Methods. The color of the skin equivalent treated with 50µM KHG26792 was lighter than that of the untreated control, whereas treatment with 10 µM KHG26792 yielded no change (Fig. 5A). As shown in Fig. 5C, melanin staining was reduced by treatment with 50 µM KHG26792.