7,3′,4′-THIF was purchased from Indofine Chemical Co., Inc. (Hillsborough, NJ, United States). Dulbecco’s modified Eagle’s medium (DMEM), penicillin-streptomycin and 0.5% trypsin-EDTA were obtained from GIBCO^§ Invitrogen (Auckland, NZ). Medium 254 and human melanocyte growth supplement (HMGS) purchased from Cascade Biologics (Portland, OR, United States). Fetal bovine serum (FBS), daidzein, α-MSH, arbutin, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), dimethyl sulfoxide (DMSO), gelatin, forskolin and β-actin antibody were purchased from Sigma Chemical Co. (St. Louis, MO, United States). The antibodies against tyrosinase, tyrosinase-related protein 1 (TYRP-1), tyrosinase-related protein 2 (TYRP-2), MITF, phosphorylated PKA α/β/γ (Thr-198), total PKA, MC1R, goat anti-mouse IgG-HRP, goat anti-rabbit IgG HRP, and donkey anti-goat IgG HRP-conjugated secondary antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, United States). Antibodies against phosphorylated Akt (Ser-473), phosphorylated mTOR (Ser-2448), phosphorylated GSK-3β (Ser-9), phosphorylated CREB (Ser-133), phosphorylated MKK3/6 (Ser-189/207), phosphorylated MSK1 (Ser-376), total PI3K, total Akt, total mTOR, total GSK-3β, total CREB, total MKK3/6, total MSK1, and total p38 were purchased from Cell Signaling Technology (Danvers, MA, United States). Antibodies against phosphorylated p38 (Tyr-180/Tyr-182) were purchased from BD Biosciences (San Jose, CA, United States). Skim milk was purchased from MB cell (LA, CA, United States). The cAMP immunoassay was purchased from Cayman (Ann Arbor, MI, United States). CNBr-Sepharose 4B and the chemiluminescence detection kit were purchased from Amersham Biosciences (Piscataway, NJ, United States). The protein assay kit was obtained from Bio-Rad Laboratories Inc. (Hercules, CA, United States). Recombinant MC1R was purchased from Millipore (Bedford, MA, United States).

The murine melanoma cell line B16F10 was obtained from the Korean Cell Line Bank (Seoul, South Korea). B16F10 cells were cultured in DMEM supplemented with 10% FBS and 1% penicillin/streptomycin at 37°C in a humidified atmosphere with 5% CO₂. Human epidermal melanocytes (HEMs) derived from moderately pigmented neonatal foreskins were purchased from Cascade Biologics. HEMs were cultured in Medium 254 supplemented with HMGS at 37°C in a humidified atmosphere with 5% CO₂.

B16F10 cells (7 × 10³) were cultured in a 96-well plate for 6 h. Cell culture media containing 7,3′,4′-THIF was added at final concentrations of 25, 50, and 100 μM, and then the cells were cultured for 96 h. HEMs (1 × 10⁴) were cultured in a 96-well plate for 6 h. Cell culture media containing 7,3′,4′-THIF was added at final concentrations of 20, 40, and 80 μM, and then the cells were cultured for 72 h. MTT solution (5 mg/mL) were plused 20 μL/well and incubated 2 h. The media were removed and replaced 200 μL of dimethyl sulfoxide (DMSO) per well to dissolve the MTT formazan. After 2 h, absorbance was measured at 570 nm using a microplate reader (Sunrise-Basic Tecan; Grodig, Austria).

Intracellular melanin accumulation was visualized by Fontana-Masson staining with a slight modification (Joshi et al., 2007; An et al., 2010). Cells were fixed in 100% ethanol for 30 min at room temperature and stained for melanin using a Fontana-Masson staining kit from American Master^∗Tech Scientific, Inc. (Lodi, CA, United States), according to the manufacturer’s instructions. In brief, cells were stained with ammoniacal silver for 60 min at 60°C, followed by incubation in 0.1% gold chloride and then in 5% sodium thiosulfate. Cell morphology and pigmentation were examined under a Nikon phase-contrast microscope (Tokyo, Japan). The images were analyzed using NIS-Elements 3.0 software.

The melanin content was measured using a slight modification of a previously reported method (Eisinger and Marko, 1982; Friedmann and Gilchrest, 1987; Gordon et al., 1989). Briefly, cells (8 × 10³) were cultured in a six-well plate for 24 h. The culture media was replaced with the media containing daidzein or 7,3′,4′-THIF at the indicated concentrations (10, 20, or 40 μmol L^–1) for 1 h before being exposed to 100 nmol L^–1 α-MSH and harvested 3 days later. After treatment, media were collected and the melanin levels therein were determined by measuring the absorbance at 405 nm using an ELISA reader.

B16F10 cells (1.5 × 10⁴) were cultured in a 6 cm dish for 48 h then starved in serum-free medium for an additional 24 h to eliminate the influence of FBS on kinase activation. HEMs (9.6 × 10⁴) were cultured in 9 cm dish for 6 days. The culture media was replaced with the media containing 7,3′,4′-THIF (10, 20, or 40 μM) for 1 h before being exposed to 100 nM α-MSH for 3 days. The harvested cells were disrupted and the supernatant fractions were boiled for 5 min. The protein concentration was determined using a dye-binding protein assay kit (Bio-Rad Laboratories Inc.), as described in the manufacturer’s manual. Lysate protein (20–40 μg) was subjected to 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and electrophoretically transferred to a polyvinylidene fluoride membrane (Millipore). After blotting, the membrane was incubated with primary antibodies at 4°C overnight. After hybridization with secondary antibodies, the protein bands were visualized using an ECL plus Western blotting detection system (Amersham^TM, Piscataway, NJ, United States). The relative intensities were quantified by Image J program.

cAMP levels were measured using a cAMP immunoassay kit (Cayman). Briefly, B16F10 cells were treated with 7,3′,4′-THIF (10, 20, or 40 μM) for 1 h before being exposed to 1 μM forskolin for 30 min. Next, the cells were lysed in 0.1 M HCl to inhibit phosphodiesterase activity. The supernatants were then collected, neutralized, and diluted. After neutralization and dilution, a fixed amount of cAMP conjugate was added to compete with cAMP in the cell lysate for sites on rabbit polyclonal antibodies immobilized on a 96-well plate. After washing to remove excess conjugated and unbound cAMP, substrate solution was added to the wells to determine the activity of the bound enzyme. The color development was then stopped, after which the absorbance was read at 415 nm. The intensity of the color was inversely proportional to the cAMP concentration in the cell lysate.

Recombinant MC1R (25 μg) or a B16F10 cellular supernatant fraction (1,000 μg) was incubated with 7,3′,4′-THIF-Sepharose 4B beads (or Sepharose 4B alone as a control) (100 μL, 50% slurry) in reaction buffer [50 mM Tris (pH 7.5), 5 mM EDTA, 150 mM NaCl, 1 mM DTT, 0.01% Non-idet P-40, 2 μg/mL bovine serum albumin, 0.02 mM PMSF, and 1 × protease inhibitor mixture]. After incubation with gentle rocking overnight at 4°C, the beads were washed five times with buffer [50 mM Tris (pH 7.5), 5 mM EDTA, 150 mM NaCl, 1 mM DTT, 0.01% Non-idet P-40, and 0.02 mM PMSF], and the proteins bound to the beads were analyzed by immunoblotting.

B16F10 cellular supernatant fraction (1,000 μg) was incubated with 100 μL of 7,3′,4′-THIF-Sepharose 4B beads or 100 μL of Sepharose 4B in a reaction buffer (see in vitro and ex vivo pull-down assays) for 24 h at 4°C, and α-MSH (0.1, 1, 10, or 100 μM) was added to a final volume of 500 μL and incubated for 24 h. The samples were washed, and proteins were then detected by western blotting.

To investigate the detailed molecular basis of MC1R inhibition by 7,3′,4′-THIF, modeling study with 7,3′,4′-THIF and MC1R was performed twice. First, the sequence based search showed that MC1R has a similarity of 47% with a MC4R theoretical model (PDB entry: 2IQP) (Pogozheva et al., 2005), then the MC1R structure was built by using Prime v3.2 from Schrödinger suite 2013 (Jacobson et al., 2002, 2004) based this MC4R theoretical model, the loops of the MC1R homology structure were refined and minimized for docking studying. 7,3′,4′-THIF and daidzein were prepared under LigPrep with default parameter. The grid for docking was generated based on the binding sites that were predicted by SiteMap, finally, Flexible docking was performed in the extra precision (XP) mode. The number of poses per ligand was set to 10 in the post-docking minimization. Second, the sequence regions for the catalytic domains in human MC1R were identified using the PFAM profile database. Unfortunately, no experimental structure exists for the human MC1R. Therefore, models of the catalytic domains in human MC1R were built by homology modeling using closest templates available in the protein data bank (PDB). A BLAST search against the PDB identified the X-ray crystal structures 1CJK and 1AB8 with 67.8 and 38.8% sequence identity, respectively. After a careful multiple-sequence alignment with ClustalW V2.1, 100 models were generated with Modeller V9.10 (Sali and Blundell, 1993; Thompson et al., 1994). The modeled regions were 1-282 (MC1R). The best model, according to the intrinsic Modeller DOPE function, was chosen for docking studies (Krivov et al., 2009). Stereochemistry was assessed with PROCHECK (Laskowski et al., 1996). The stereochemistry of the best model was manually inspected with COOT (Emsley and Cowtan, 2004). Electrostatics calculations were performed with APBS V1.2.1; the molecular surfaces with the electrostatics properties (blue: positive charges; red: negative charges, with unit + 5/-5 kT/e) were rendered using VMD V1.8.9 and PyMOL V1.5 (Oberoi and Allewell, 1993; Holst et al., 1994; Wallace et al., 1995; Trott and Olson, 2010).

Prior to the docking calculation, the missing hydrogen atoms on the receptor were added using PDB2PQR in conjunction with the CHARMm force field (Oberoi and Allewell, 1993). AutoDock Vina was used to perform the dockings (Trott and Olson, 2010). Ligands were prepared with Spartan’10 (Wavefunction, Inc.) and the scripts provided by the AutoDock tools (v1.5.4 r29). The docking grid sizes and positions were set using AutoDock. The grid dimensions were large enough to cover the whole biological unit and centered onto previously defined relevant catalytic and biological areas. The grid size was 26 × 28 × 30 Å (MC1R). In MC1R, the loop region closing the receptor (a.a. 155–163) was set as the flexible region for docking. The lowest docking energy and predicted free binding energy for each compound were retained. These energies were used to sort the VLS results. Top docking solutions were manually inspected using COOT and PyMOL V1.5 to verify and confirm their compatibility with existing knowledge on the receptor (e.g., active site location, binding pockets, and conserved amino acid residues potentially involved in binding).

The models were analyzed using Ligplot (Wallace et al., 1995). Confirmation of the interaction maps was performed manually by visual inspection of the models in COOT (Emsley and Cowtan, 2004).

Where applicable, the data are expressed as means ± SD; Student’s t-test was used for single statistical comparisons. A probability value of p < 0.05 was used as the criterion for statistical significance.

To investigate the whitening effect of 7,3′,4′-THIF and daidzein, we first examined the melanin content in B16F10 melanoma cells using Fontana-Masson staining and a melanin content assay. 7,3′,4′-THIF showed cytotoxicity at 50 μM concentration when treated in B16F10 for 96 h, and 7,3′,4′-THIF were incubated with B16F10 under non-cytotoxic conditions for all experiments (Supplementary Figure S1). Treatment with 7,3′,4′-THIF, but not daidzein, significantly reduced the intracellular melanin content of B16F10 cells in a dose-dependent manner (Figures 1Cg–i). Consistent with this, 7,3′,4′-THIF inhibited the α-MSH-induced extracellular melanin content in a dose-dependent manner (Figure 1D). Additionally, 7,3′,4′-THIF lightened the color of the extracellular culture media to a greater degree than did arbutin, a well-known whitening agent. In contrast, daidzein had no effect on the α-MSH-induced extracellular melanin content at all indicated concentrations (Figure 1D). These results suggest that 7,3′,4′-THIF, but not daidzein, exerts a strong depigmenting effect on B16F10 cells.

Tyrosinase, TYRP-1, and TYRP-2 play important roles in melanogenesis (Park et al., 2009; Yamaguchi and Hearing, 2009; Gillbro and Olsson, 2011). Thus, we examined whether 7,3′,4′-THIF could suppress tyrosinase expression. 7,3′,4′-THIF suppressed α-MSH-induced tyrosinase, TYRP-1 and TYRP-2 expression in B16F10 cells (Figure 2A). Because α-MSH induces tyrosinase and TYRP expression mainly through cAMP/PKA signaling and subsequently induces both MITF expression and CREB phosphorylation, we next investigated whether 7,3′,4′-THIF inhibited MITF expression and CREB phosphorylation. 7,3′,4′-THIF strongly suppressed α-MSH-induced MITF expression and CREB phosphorylation (Figure 2B), suggesting that the whitening effect of 7,3′,4′-THIF is mediated by the suppression of melanogenic protein expression in B16F10 cells.

Previous studies have shown that AKT, MAPK, and PKA signaling is involved in α-MSH-induced melanogenesis and tyrosinase expression in B16F10 cells (Busca and Ballotti, 2000; Smalley and Eisen, 2000; Tada et al., 2002; Saha et al., 2006; Solano et al., 2006; Bellei et al., 2008; Yamaguchi and Hearing, 2009; Gillbro and Olsson, 2011; Jin et al., 2011). Thus, we next investigated the effect of 7,3′,4′-THIF on AKT, MAPK, and PKA signaling. 7,3′,4′-THIF activated the α-MSH-induced dephosphorylation of AKT, mTOR, and GSK3β in B16F10 cells at the concentrations tested (Figure 3A). 7,3′,4′-THIF also suppressed the α-MSH-induced phosphorylation of MKK3/6, p38, and MSK1 in a dose-dependent manner (Figure 3B). Additionally, 7,3′,4′-THIF suppressed the α-MSH-induced phosphorylation of PKA in B16F10 cells at 40 μM (Figure 3C). Taken together, these results indicate that 7,3′,4′-THIF inhibits tyrosinase expression by activating AKT signaling and suppressing MAPK and PKA signaling.

We next investigated whether 7,3′,4′-THIF affected adenylyl cyclase, an upstream effector of the AKT, MAPK, and PKA cascades. After the stimulation of MC1R by α-MSH, adenylyl cyclase converts ATP to cAMP, resulting in the activation of downstream signaling pathways (Solano et al., 2006; Yamaguchi and Hearing, 2009; Gillbro and Olsson, 2011). Therefore, to identify the effect of 7,3′,4′-THIF on the intracellular cAMP level, we performed a cAMP immunoassay using forskolin, a direct activator of adenylyl cyclase. At 40 μM, 7,3′,4′-THIF reduced the forskolin-induced intracellular cAMP level by up to 23.8% (Figure 4A). These results indicate that the inhibition of tyrosinase expression by 7,3′,4′-THIF involves the inhibition of intracellular cAMP formation and adenylyl cyclase activity.

Accumulating data suggest that adenylyl cyclase or MC1R is the potential molecular target of 7,3′,4′-THIF, and that binding results in the inhibition of melanogenesis and tyrosinase expression. To confirm whether 7,3′,4′-THIF binds directly to MC1R, we performed an in vitro and ex vivo pull-down assay using 7,3′,4′-THIF-conjugated and non-conjugated Sepharose 4B beads. MC1R (Figure 4B and Supplementary Figure S2A, lane 3) bound to the 7,3′,4′-THIF-Sepharose 4B beads, but not to the non-conjugated Sepharose-4B beads (Figure 4B and Supplementary Figure S2A, lane 2), revealing that 7,3′,4′-THIF binds MC1R directly in vitro and ex vivo. However, daidzein did not bind to MC1R both in vitro and ex vivo (Figure 4C and Supplementary Figure S2B). The binding ability of 7,3′,4′-THIF with MC1R was altered in a concentration-dependent manner in the presence of α-MSH (Figure 4D), suggesting that 7,3′,4′-THIF competes with α-MSH for binding with MC1R. These results suggest that 7,3′,4′-THIF is an α-MSH-competitive inhibitor for suppressing MC1R activation. Computer modeling study also revealed that 7,3′,4′-THIF but not daidzein easily docked to the α-MSH-binding site of MC1R (Figure 5 and Table 1).

We next investigated the whitening effect of 7,3′,4′-THIF and daidzein on HEMs. 7,3′,4′-THIF at a concentration of 80 μM maintained cell viability above 80% for 72 h in HEMs (Supplementary Figure S3). 7,3′,4′-THIF, but not daidzein, significantly reduced the intracellular melanin content of HEMs in a dose-dependent manner (Figure 6A). It was also necessary to verify the mechanisms of 7,3′,4′,-THIF on melanogenesis in the human cells. 20 and 40 μM of 7,3′,4′-THIF significantly reduced α-MSH-induced tyrosinase expression in HEMs (Figure 6B). Furthermore, α-MSH-induced PKA phosphorylation was downregulation in HEMs cultured with 40 μM of 7,3′,4′-THIF, and phosphorylation of Akt inhibited by α-MSH treatment was partially ameliorated by 7,3′,4′-THIF treatment (Figure 6C). 7,3′,4′-THIF also slightly decreased the phosphorylated p38, but the level of p38 also showed a tendency to decrease, showing an increasing phospho-/total- form ratio. These results indicate that 7,3′,4′-THIF could exert the same depigmenting effect in both murine melanoma cells and human melanocytes.