Bifunctional effects of O-methylated flavones from Scutellaria baicalensis Georgi on melanocytes: Inhibition of melanin production and intracellular melanosome transport

The growing interest in skin lightening has recently renewed attention on the esthetic applications of Chinese herbal medicine. Although Scutellaria baicalensis Georgi is used for antipyretic and antiinflammatory purposes, its whitening effect remains unclear. This study reports three major findings: (1) S. baicalensis has a potent inhibitory effect on melanogenesis; (2) wogonin and its glycoside are the active components of S. baicalensis; and (3) O-methylated flavones from S. baicalensis, such as wogonin, inhibit intracellular melanosome transport. Using a melanin quantification assay, we showed that S. baicalensis potently inhibits melanogenesis in B16F10 cells. Componential analyses revealed that the main components of S. baicalensis are baicalin, wogonoside, baicalein, wogonin, and oroxylin A. Among these five flavones, wogonin and wogonoside consistently inhibited melanogenesis in both B16F10 melanoma cells and primary melanocytes. Wogonin exhibited the strongest inhibition of melanin production and markedly lightened the color of skin equivalents. We identified microphthalmia-associated transcription factor and tyrosinase-related proteins as potential targets of wogonin- and wogonoside-induced melanogenesis suppression. In culture, we found that the melanosomes in wogonin-treated B16F10 cells were localized to the perinuclear region. Immunoblotting analyses revealed that wogonin significantly reduced in melanophilin protein, which is required for actin-based melanosome transport. Other actin-based melanosome transport-related molecules, i.e., Rab27A and myosin Va, were not affected by wogonin. Cotreatment with MG132 blocked the wogonin-induced decrease in melanophilin, suggesting that wogonin promotes the proteolytic degradation of melanophilin via the calpain/proteasomal pathway. We determined that the structural specificities of the mono-O-methyl group in the flavone A-ring and the aglycone form were responsible for reducing melanosome transport. Furthermore, wogonin and two wogonin analogs, mono-O-methyl flavones, strongly suppressed melanosome transport. Our findings suggest the applicability of S. baicalensis in the esthetic field. Thus, we propose a novel pharmacologic approach for the treatment of hyperpigmentation.

Introduction Melanocytes in the epidermis are responsible for producing melanin, a black pigment that protects against ultraviolet radiation [1]. Melanocytes synthesize numerous secretory melanosomes containing proteins and melanin, which are then transferred to neighboring keratinocytes [2]. Prior to release from melanocytes, melanosomes move from the perinuclear region to the cell periphery through melanization by enzymes required for melanin synthesis, including tyrosinase (TYR), tyrosinase-related protein 1 (TYRP1), and dopachrome tautomerase (DCT) [3]. These melanogenetic enzymes are transcriptionally regulated by microphthalmia-associated transcription factor (MITF), which is a master transcription factor of melanocyte development, differentiation, and survival. The intercellular transport of melanosomes requires several steps, including bidirectional long-range transport on microtubules to the apical surface, transfer to actin filaments, irreversible short-range transport by actin dynamics, and binding to the cytomembrane [4].
Several independent researchers recently reported that Rab27A, synaptotagmin-like protein (SLP) 2A/synaptotagmin 2, melanophilin (MLPH)/SLP homolog lacking C2 domains-A, and myosin Va are involved in the regulation of melanosome transport [4,5]. Rab27A and myosin Va interact via MLPH to form a tripartite complex, which serves as a regulator of actin-based transport. Rab27A links SLP2A with phosphatidylserine, thereby docking melanosomes with the apical plasma membrane. This finding suggests that SLP2A is a regulator of melanosome exocytosis. Mutations in genes encoding melanosome transport-related molecules, such as myosin Va, Rab27A, and MLPH, cause a human pigmentary disease known as Griscelli syndrome, which is associated with diluted skin and hair color [6][7][8]; therefore, these molecules play an important role in modulating skin color variations.
Scutellaria baicalensis Georgi is a perennial herb belonging to the Lamiaceae family. Its root has been used as a traditional Chinese medicine to alleviate inflammation, allergy, and fever as well as a treatment for several cancers [9,10]. Various bioactive components, such as flavonoids, have been identified in S. baicalensis root parts. Baicalin, baicalein, wogonin, and wogonoside are the main components of the root, and all these compounds have significant pharmacological effects [11]. In recent years, the whitening effect of dried S. baicalensis root has attracted significant attention from the Asian cosmetic field; however, a comprehensive study on the whitening effect of S. baicalensis, including its mechanism of action, has not been conducted. Therefore, we studied the functional link between the whitening effect of S. baicalensis and the antimelanogenic effects of its major flavonoids. Our results identified wogonin and its glycoside as the most potent regulatory constituents of S. baicalensis. In addition, we showed the bifunctionality of wogonin, including the downregulation of both melanin production and intracellular melanosome transport. In this study, we provide plausible evidence that both the aglycone form and the O-methyl group on the flavone A-ring likely contribute to the inhibition of intracellular melanosome transport. Thus, our study provides a better pharmacologic understanding of flavones as depigmenting agents.

Preparation of S. baicalensis root extract
Dried chips of S. baicalensis root were obtained from a commercial supplier. Chips were ground in a mortar to a coarse powder. To prepare a crude extract from S. baicalensis root, the powder (0.51 g) was added to 10 ml of 70% ethanol and left to stand at room temperature for 24 h after sonication. The supernatant was fractionated using a Pasteur pipette and then filtered through a membrane filter (0.45 μm). The resultant solution was concentrated by rotary evaporation. After subsequent standing in a desiccator for 4 days, the crude extract (27.4 mg) was obtained. Using the same procedure, the organic solvents of n-hexane, ethylacetate, methanol (MeOH), and water resulted in 83.2, 109.2, 177.6, and 84.4 mg of the crude extract from the ratio of powder/solvent (20.3 g/100 ml, 10.1 g/50 ml, 1.0 g/5 ml, and 1.0 g/30 ml), respectively. For extraction with water, the filtrate was freeze-dried for 6 days. For the fractionated samples, the crude MeOH extract (2.3 g) was prepared from 10 g of dried powder in 150 ml of MeOH. Next, 1.0 g of the crude MeOH extract was passed through a Sep-Pak C18 cartridge (Waters, Milford, MA, USA) and eluted stepwise with a series of MeOH solutions, including water, 25%, 50%, 75%, and 100% MeOH. After solid-phase extraction, each fraction was freeze-dried after vaporization. We obtained 889.7, 40.7, 111.5, 46.1, 28.0 mg of fractional residue from water, 25%, 50%, 75%, and 100% MeOH, respectively. The test samples of each extraction were reconstituted with distilled water, MeOH, or dimethyl sulfoxide to generate the crude extract and fractionated solutions. Each extraction solution was aliquoted and stored at -20˚C until further use. Once dissolved for testing, each extraction solution was stored at 4˚C and used within one week.

Cell cultures
The mouse melanoma cell line B16F10 was provided by the Cell Resource Center for Biomedical Research, Tohoku University (Miyagi, Japan). B16F10 cells were maintained as monolayer cultures in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS) at 37˚C in a humidified atmosphere containing 5% CO 2 . Primary human epithelial melanocytes (HEMs, newborn/darkly pigmented skin) were obtained from Thermo Fisher Scientific. The melanocyte basal medium consisted of Medium 254 with human melanocyte growth supplement (Thermo Fisher Scientific). The medium was changed once every other day for culture maintenance. Cells were not used beyond passage 10. A three-dimensional human skin model (3D-HSM, MEL-300-B) was cultured at 37˚C with 5% CO 2 in EPI-100LL MM medium (MatTek Corporation, Ashland, MA, USA). All human-derived cells used were derived and established by accredited institutions of Thermo Fisher Scientific and MatTek Corporation. Cells obtained through these institutions comply with the applicable legal and ethical guidelines of the United States. In all cases, consent for use of the cells for research purposes was obtained by these institutions from the donor or the donor's legal next of kin.

Melanin quantification
B16F10 cells were seeded in 100-mm dishes at a density of 1.0 × 10 5 cells per dish and then cultured in DMEM supplemented with 10% FBS at 37˚C with 5% CO 2 until they reached 80-90% confluence. After replacing the culture medium with fresh medium supplemented with 2% FBS, cells were coincubated with the extracts or test compounds for 24 h. HEMs were seeded in 100-mm dishes at a density of 1.0 × 10 5 cells per dish and then incubated in Medium 254 with human melanocyte growth supplement for 24 h. After replacing the culture medium with fresh medium, cells were coincubated with the test compounds for the indicated lengths of time. The culture medium with or without the test compounds was changed every other day. All extracts and test compounds were first diluted in culture media at 37˚C before incubation with the cells. In the 3D-HSM experiment, the skin equivalents were preincubated in EPI-100LLMM medium at 37˚C with 5% CO 2 overnight. After replacing the culture medium with fresh medium, the 3D-HSM was coincubated with the test compounds for 14 days. The compounds were first diluted in 75 mM Tris-HCl (pH 8.8) at room temperature. Next, 0.1 ml of buffered solution without (control) or with each compound was applied to the stratum corneum side. The medium and each buffered solution was changed every other day. Melanin content in cells and 3D-HSM was measured by modifying previously reported methods [12,13]. Cells or 3D-HSM were washed three times with PBS. For the spectrophotometric calculation of total melanin, the pellets from cells or 3D-HSM were washed with 5% trichloroacetate, ethanol-ether (3:1 v/v), and ether. Next, the pellets were dried at 50˚C for 1 h. Dried precipitates were solubilized with 2 M sodium hydroxide or SOLVABLE TM (Perkin-Elmer, Waltham, MA, USA) by heating at 80˚C for 30 min. Lysates were clarified by centrifugation, and the supernatants were transferred to a clean tube. Each sample was monitored at 405 nm using an 1420 ARVO series multilabel counter (Perkin-Elmer). The estimated pigment values from cells and 3D-HSM were normalized by cell number and absorbance value in the methylthiazole tetrazolium (MTT) assay, respectively. The number of viable cells was assessed by a trypan blue exclusion assay.

MTT assay
Cell viability was evaluated using the MTT assay kit (MatTek Corporation) per the manufacturer's instructions with slight modifications. The 3D-HSM was placed in a 24-well plate, 0.3 ml of MTT solution (1 mg/ml) was added to each well, and the tissue was incubated at 37˚C in 5% CO 2 for 3 h. After incubation, the tissue was washed three times with PBS. For the MTT extraction, 2 ml of a 0.04 N HCl solution in isopropanol was added to each well, and the plate was gently shaken at room temperature for 2 h. The absorbance of the extracts was measured at 570 nm using a 1420 ARVO series multilabel counter.

HPLC
For HPLC, we employed a gradient system equipped with a syringe-loading sample injector (Model 7125, Rheodyne, Cotati, CA, USA), dual pumps, and a multi-UV detector (Model PU-2089 and MD-2010, respectively, JASCO, Tokyo, Japan). Two mobile phase solvents were used. Solvent A was prepared by adding concentrated formic acid (0.1%) to deionized water. Solvent B was HPLC-grade acetonitrile. The HPLC method consisted of a reverse-phase column (Develosil C 18 , 5 μm, 4.6 × 250 mm, Nomura Chemical, Aichi, Japan) and the following mobile phase gradient: 0-40 min with a linear gradient from 95% solvent A to 100% solvent B and 40-50 min hold 100% of solvent B. The flow rate was constant at 1.0 ml/min.

Determination of total phenolic compounds
The concentration of total phenolic compounds in each extract was evaluated spectrophotometrically using Folin-Ciocalteu reagent and a previously published method with slight modifications [14]. Briefly, 0.1 ml of each sample or standard previously dissolved in ethanol was diluted with water to a volume of 0.45 ml. Next, 0.5 ml of Folin-Ciocalteu phenol reagent was added, and the test tubes were vigorously agitated. 0.4 ml of a 1 M sodium carbonate solution was subsequently added, and the tubes were thoroughly agitated again. The mixtures were incubated at room temperature and protected from light for 1 h. The absorbance of the resulting reaction mixtures was measured at 750 nm with an Ultrospec 4300 pro UV/visible spectrophotometer (GE Healthcare). The concentration of total phenolic compounds for each extract was calculated based on the standard curve obtained using gallic acid.

Determination of total flavonoids
The quantification of total flavonoids in each extract was completed using the Lamaison and Carnat method modified for a microplate assay [15]. Each 100-μl sample was added to a 96-well plate, followed by addition of 100 μl of a 2% AlCl 3 solution in MeOH. After 10 min, the absorbance at 415 nm was measured with a 1420 ARVO series multilabel counter. A standard curve was developed using quercetin.

Quantitative PCR (qPCR)
Total RNA was isolated from cultured cells using an RNeasy Mini Kit (Qiagen, Mississauga, Canada) per the manufacturer's instructions. RNA was stored in RNase-free water at −80˚C prior to reverse transcription. First-strand cDNA was synthesized with 1 μg of total RNA using a PrimeScript II 1st strand cDNA Synthesis Kit (Takara Bio, Shiga, Japan) per the manufacturer's instructions. mRNA expression levels of target genes were measured using an Applied Biosystems 7500 Real Time PCR System (Applied Biosystems, Foster City, CA, USA) and the following TaqMan Gene Expression Assays: Mitf (assay ID Mm00434954_m1); Mlph (assay ID Mm01286059_m1); and Gapdh (assay ID Mm99999915_g1). All reactions were performed in triplicate. Gapdh was used as a housekeeping gene for normalization. The relative amount of mRNA was calculated using the comparative C T method.
Immunoblotting B16F10 cells were lysed using a lysis buffer (150 mM NaCl; 50 mM Tris, pH 8.0; 1% NP-40; 0.5% sodium deoxycholate; and protease inhibitors). Each extract was subjected to SDS-PAGE, and proteins were transferred to polyvinylidene fluoride membranes using a semi-dry blotter (Bio-Rad Laboratories, Hercules, CA, USA). After the transfer, the membranes were incubated in a blocking solution [0.3-1% dried skimmed milk in Tris-buffered saline containing 0.1% Tween-20 (TBS-T)] for 1 h to reduce nonspecific binding. Next, membranes were incubated with primary antibodies overnight at 4˚C. The blots were washed and incubated with HRPconjugated secondary antibodies for 1 h. Proteins were detected using ImmunoStar LD (Wako Pure Chemical).

Immunofluorescence and melanosome distribution assay
B16F10 cells or HEMs on collagen coated coverslips were fixed for 15 min in 4% paraformaldehyde, permeabilized with 0.1% Triton-X 100 for 10 min, and incubated in a blocking solution (3% BSA in PBS) for 1 h. To detect actin, cells were treated with diluted Alexa Fluor 594 phalloidin for 20 min. For TYRP1 coimmunostaining, cells were exposed to an anti-TYRP1 antibody (1:75 dilution) in PBS containing 1% BSA for 16 h at 4˚C and subsequently incubated with an Alexa Fluor 488-conjugated anti-goat antibody (1:400 dilution) for 1 h. For nuclear counterstaining, cells were treated with 1 μg/ml DAPI. Mounted samples were observed using a TCS SP2 AOBS confocal laser microscope (Leica Microsystems, Wetzlar, Germany). Melanosome distribution assays were performed using a previously published method with slight modifications (more than 100 cells/cover glass, three independent glasses for each sample) [16]. Cell images were randomly acquired, and cells with more than 50% of melanosomes around the nucleus were labeled as aggregated.

Statistical analyses
The data are expressed as the mean ± SD from at least three independent experiments. Statistical analyses were performed using the Tukey-Kramer test. Asterisks ( Ã p < 0.05 and ÃÃ p < 0.01) and daggers ( † † p < 0.01) indicate statistical significance when compared with the control group and the indicated group, respectively.

Results
To examine the depigmenting ability of S. baicalensis, we first quantified melanin levels using cultured mouse B16 melanoma cells. As shown in Fig 1, the S. baicalensis extract significantly decreased melanin levels in a dose-dependent manner between 35 and 70 μg/ml. At a concentration of 70 μg/ml, the S. baicalensis extract inhibited melanin formation more effectively than 100 μM of LA, which was used as a positive control [17]. To determine the most efficient extraction of S. baicalensis, we examined the inhibitory activity of each extract generated from four solvents: n-hexane, ethyl acetate, MeOH, and water. After the treatment of B16F10 cells with each extract for 24 h, the melanin content was assessed. The S. baicalensis extract eluted by MeOH resulted in a significant decrease in melanin content (Fig 2C), whereas no decrease was observed after treatment with the other three extracts (Fig 2A, 2B and 2D). In fact, the extract eluted by ethyl acetate tended to increase melanin content and produced toxicity ( Fig  2B). These results suggest that S. baicalensis is capable of inhibiting melanogenesis and its active components can be efficiently extracted by alcohols such as MeOH.
To identify the active components that inhibit melanin synthesis, the S. baicalensis MeOH extract was sequentially fractionated into five samples using a C18 solid-phase extraction column (Fig 3). We assessed the effect of these five fractions on melanogenesis. As shown in Fig  4, only the 25% MeOH and 75% MeOH fractions decreased melanin content. The 75% MeOH fraction also exhibited significant cytotoxicity against B16F10 cells. Quantitative assays revealed that the total phenolic compounds and flavonoids were abundant in the following order: 75%, 25%, and 50% MeOH fractions (Table 1). Based on these results, we speculated that certain flavonoids in S. baicalensis were responsible for the inhibition of melanogenesis. Previous reports identified the major flavonoids found in S. baicalensis (S1 Fig) [18]. Using an HPLC-diode array detector and liquid chromatography-tandem mass spectrometry, we compared the corresponding retention times and UV and mass spectra of samples with the standards indicated in the literature. We identified the five characteristic peaks as baicalin (peak a), wogonoside (peak b), baicalein (peak c), wogonin (peak d), and oroxylin A (peak e) (Fig 5). The assignment of each peak revealed that the fractions eluted with 25% and 50% MeOH contained abundant flavone glycosides, particularly baicalin and wogonoside ( Table 2). The main components present in the 75% MeOH fraction were three aglycone flavones: baicalein, wogonin, and oroxylin A. In contrast, both the water and 100% MeOH fractions contained almost no constituents.
We assessed the effect of five major flavones: baicalin, wogonoside, baicalein, wogonin, and oroxylin A. As shown in Fig 6, 50 μM of baicalin, wogonoside, and wogonin significantly inhibited melanin synthesis without causing cytotoxicity, whereas baicalein and oroxylin A did not exert this effect. Wogonin strongly inhibited and decreased melanin content by 26.7% when compared with the control. In contrast, baicalein and oroxylin A caused marked cytotoxicity in the absence of any inhibitory effects on melanogenesis. These results suggest that two glycosides (i.e., baicalin and wogonoside) and wogonin are the active components responsible for the antimelanogenic effects of the 25% MeOH and 75% MeOH fractions, respectively. The 75% MeOH fraction also contained baicalein and oroxylin A (Table 2), which likely contribute to high cytotoxicity of this fraction. Thus, baicalin, wogonoside, and wogonin are the active constituents of S. baicalensis that inhibit melanin formation in B16F10 cells.
To elucidate the effect of the three active components of S. baicalensis root under normal cell conditions, we examined melanin levels in HEMs after exposure to each compound. As shown in Fig 7, exposure to wogonin at a concentration of 25 μM for 3-5 days inhibited melanin formation in a time-dependent manner without affecting the growth or survival of HEMs. Exposure to wogonin for 4-5 days showed a tendency of stronger inhibition when compared with H-89, which blocks melanin synthesis via cyclic AMP-dependent protein kinase-specific inhibition [19]. In contrast, baicalin and wogonoside had no effect or a weak effect on melanogenesis in HEMs (Fig 7). We also assessed the activity of wogonin using a 3D-HSM. As shown in Fig 8, exposure of the 3D-HSM to wogonin at concentrations of 30-120 μM did not cause cytotoxicity. Wogonin markedly lightened the color of the 3D-HSM and decreased melanin content by up to 23.0% when compared with the control (Fig 8A and 8B). Rapamycin, which lightens skin color via autophagy [13], decreased melanin content by 34.0% but also caused cytotoxicity. These results suggest that among the three active components identified in the B16F10 model, wogonin and wogonoside consistently downregulate melanin formation in both melanoma and normal cell assays.
Wogonin has been used as a treatment for cancers and ocular diseases [9,10]. In these previous studies, wogonin exhibited antitumor effects by inhibiting the extracellular regulated protein kinase and Wnt/beta-catenin pathways. In melanocytes, these pathways play a key role in MITF expression [20,21]. We found that wogonin and wogonoside significantly reduced MITF expression in B16F10 cells (Fig 9A and 9B). In addition, wogonin and wogonoside significantly inhibited TYRP1 and DCT but not TYR (Fig 9C and 9D). These results suggest that wogonin and its glycoside downregulate the expression of melanogenesis-related proteins by suppressing MITF expression.
In culture, we found that melanosomes in wogonin-treated B16F10 cells were localized to the perinuclear region and conspicuously absent from the cell periphery when compared with the whole-cell distribution pattern of the control (Fig 10A). Melanosomes in mammalian melanocytes are intracellularly transported along microtubules and actin filaments [4]. Actinbased melanosome transport is mediated by Rab27A, myosin Va, and MLPH. Any depletion or defect in these proteins disrupts traffic and induces abnormal melanosome distribution [6][7][8]. The perinuclear melanosome aggregation induced by wogonin exhibited similarities to the   distribution patterns induced by myosin Va, Rab27A, and MLPH deficiencies [7,8,22]. Therefore, we investigated whether wogonin downregulates actin-based melanosome transportrelated molecules (e.g., Rab27A) in B16F10 cells. Wogonin caused a dose-dependent reduction in MLPH protein (Fig 10B), whereas the protein levels of Rab27A and myosin Va were not affected by wogonin treatment. The Mlph mRNA level was not affected by wogonin (Fig 10C), suggesting that wogonin had little inhibitory effect on the gene transcription of Mlph. MLPH contains multiple PEST-like sequences that are highly sensitive to proteolysis [23]. To determine whether wogonin promotes MLPH degradation, we examined the effects of the calpain/ proteasome inhibitor MG132 and the lysosome inhibitor leupeptin on the wogonin-induced decrease in MLPH protein. MG132 but not leupeptin effectively blocked the reduction in MLPH (Fig 11A and 11B). These results suggest that wogonin promotes the proteolytic degradation of MLPH via the calpain/proteasomal pathway. In addition, the treatment of cells with MG132 failed to rescue the change in melanosome distribution induced by wogonin ( Fig  11C), suggesting that wogonin induces MLPH dysfunction. Upon removal of wogonin from the culture medium, the perinuclear aggregation was reversed ( Fig 12A) and MLPH levels were restored (Fig 12B), indicating that the inhibitory effect of wogonin on both melanosome transport and MLPH protein expression only occurs in the presence of wogonin. These results suggest that MLPH turnover following removal of wogonin normalizes intracellular melanosome transport. Melanosome localization in HEMs was assessed by immunostaining for TYRP1 as a melanosome marker [24]. Pronounced perinuclear localization of melanosomes was observed in HEMs after treatment with wogonin but not H-89 when compared with control cells (Fig 13). These results collectively suggest that wogonin downregulates constitutive actin-based melanosome transport by accelerating MLPH degradation.
To determine the structural aspects of wogonin responsible for melanosome transport defects, B16F10 cells were cultured with several wogonin analogs. As shown in Fig 14A and  14B, norwogonin, an 8-hydroxylated wogonin, completely failed to aggregate intracellular melanosomes when compared with the perinuclear aggregation mediated by wogonin. This result indicates that the hydroxyl substitution of the 8-O-methyl group in wogonin caused a loss of function. In addition, wogonoside, a glycoside of wogonin, showed a noticeably weaker effect on melanosome localization when compared with wogonin. This result suggests that the glycoside conjugation of wogonin inhibited its ability to regulate melanosome transport. Next, we assessed the effect of these structural features on MLPH levels. As shown in Fig 14C, B16F10 cells cultured with wogonoside and norwogonin showed almost no detectable reduction in the protein level of MLPH when compared with wogonin-treated melanoma cells. Therefore, we hypothesized that the O-methyl group in the flavone A-ring may disturb melanosome transport. In support of our hypothesis, the 7-O-methylation of baicalein led to an acquired ability to impair melanosome transport (S2 Fig). To define the superiority of the Omethyl position on the flavone A-ring on melanosome transport, a comparative study of three structural isomers, i.e., wogonin, oroxylin A, and 7-O-methyl baicalein, was performed (Fig 15A). Oroxylin A, a 6-O-methyl-positioned isomer, inhibited melanosome dispersion but to a lesser extent (47.5%) than wogonin (77.9%; Fig 15B). Among these three flavones, wogonin had the strongest inhibitory effect at 77.9%, whereas 7-O-methyl baicalein had the weakest inhibitory effect at 28.6%. These results indicate that 8-O-methyl flavones, including wogonin, are the most potent melanosome transport inhibitors. O-Methyl-positioned flavones also showed inhibitory effects on MLPH levels (Fig 15C), suggesting that inhibition of melanosome transport by O-methylated flavones was due to the downregulation of MLPH. In 6-, 8-, and 7-O-methylated flavones, the extent of antimelanogenic effectiveness decreased, respectively ( Fig  15D), indicating a lack of concordance between their effectiveness against melanogenesis and intracellular melanosome transport. These results suggest that the 8-O-methyl group of wogonin obligately downregulated melanosome transport.

Discussion
This study reports three major findings: (1) S. baicalensis potently inhibits melanogenesis, (2) wogonin and wogonoside are the active components of S. baicalensis, and (3) the O-methylated flavones in S. baicalensis, such as wogonin, inhibit intracellular melanosome transport. The antimelanogenic action of wogonin and its glucoside are likely due to the inhibition of MITF expression (Fig 9). In this study, B16F10 cells were cultured in DMEM without additional stimulation, but recent reports have shown that highly concentrated L -tyrosine in DMEM exerts a hormone-like action and stimulates melanogenesis [25,26]. Therefore, we also evaluated the effect of these two flavones on melanogenesis in RPMI 1640 medium (RPMI), which contains a lower concentration of L -tyrosine than DMEM. The two flavones significantly decreased the melanin content of B16F10 cells cultured in L -tyrosine-supplemented RPMI (S3 Fig), whereas no decrease was observed in cells cultured in RPMI. This result suggests that the flavones did not exert an inhibitory effect on melanogenesis in a ground state. We showed that wogonin and wogonoside downregulate TYRP1 and DCT, but preserve TYR (Fig 9). This change may shift eumelanogenesis towards the 5,6-dihydroxyindole oxidation sub-pathway [27]. A previous report showed that TYRP1 deficiency produced light brown pigmentation in culture [28]. To detect changes in color other than discoloration, we assessed the hue of melanin produced by melanocytes from moderately pigmented skin. The color of melanin from wogonin-treated cells were light brown and clearly different from the black to dark brown color of control cells (S4 Fig). However, the discoloration caused by H-89 treatment also resulted in a visibly pale brown; thus, it is difficult to determine which characteristics are specific to the 5,6-dihydroxyindole sub-pathway. Recently, sex-determining region Y-box 9 (SOX9) and SOX10 were identified as key regulators of the expression of melanogenesisrelated genes, including the gene encoding MITF [29,30]. Wogonin severely downregulated SOX9 and moderately upregulated SOX10 (S5 Fig), indicating that SOX9 is a potential target for the effect of these S. baicalensis flavones.
We focused on melanosome transport-related proteins serving as actin-based transport and found a decline in the protein level of MLPH induced by wogonin (Fig 10B). The calpain/proteasome inhibitor MG132 attenuated the MLPH reduction induced by wogonin (Fig 11A). Wogonin-mediated adaptor dysfunction was postulated to induce the early intervention of proteases, thus leading to MLPH degradation. Recent patent applications reported that several plant extracts, including Sesamum indicum L. and Cassia mimosoides L., suppressed the levels of multiple melanosome transport-related proteins in HEMs and resulted in melanosome transport disorders [31,32]. In contrast, the inhibitory effect of wogonin appeared to be specific to MLPH. Several recent reports showed that MITF serves as a transcription factor for Rab27A but not MLPH [33,34]. In this study, wogonin downregulated the protein levels of MLPH but not Rab27A or myosin Va (Fig 10B). This result suggests that the moderate decrease in MITF induced by wogonin has little or no influence on the expression of actin- based transport molecules, such as Rab27A. MITF also controls the expression of other loci related to melanosome transport, including Ocular Albinism type 1 (OA1) [30,[34][35], which serves as a microtubule-based motor molecule. Interestingly, wogonin reduced the expression of OA1 mRNA (S6 Fig), indicating that wogonin affects both the microtubule-based and actin-based transport of melanosomes. Palmisano et al. reported that OA1 deficiency leads to the abnormal distribution of melanosomes towards the cell periphery, which appears opposite to the melanosome distribution due to dysfunction of actin-mediated transport [35]. According to this previous study, the characteristic disturbance of melanosome by OA1 deficiency was rescued by the inhibition of actin-based transport. We did not observe the peripheral accumulation of melanosomes in wogonin-treated cells. Thus, we speculate that the disorder of actin-based transport by wogonin masks the simultaneous malfunction of OA1. The reduction in melanin content induced by wogonin may be partially due to the decline in OA1 expression because OA1 deficient pigment cells were reported to show reduced melanin content and macromelanosomes [35].
MLPH is an adapter protein that links Rab27A-melanosomes to myosin Va, and a defect in MLPH produces a characteristic depigmentation [36]. Thus, the downregulation of MLPH may be helpful for correcting skin hyperpigmentation. Chang et al. reported that disrupting the MLPH-myosin Va interaction using manassantin B contributed to the inhibition of melanosome transfer from melanocytes to keratinocytes [37]. Additionally, the disruption of this interaction caused a decrease in the total melanin content in a coculture of Melan-a melanocytes and SP-1 keratinocytes. Our results raise the question of whether the wogonin-mediated loss of MLPH could be responsible for the reduction in melanin content in B16F10 cells and in melanocytes cultured alone. The siRNA-mediated knockdown of Mlph in B16F10 cells resulted in a moderate reduction of melanin content (S7 Fig). Combined with Fig 15, these results imply that melanosome transport defects due to the absence of MLPH slightly influenced melanogenesis.
To the best of our knowledge, our study is the first to show that the O-methyl group in the flavone A-ring plays a key role in the perinuclear aggregation of melanosomes. Nobiletin, an O-hexamethylated flavone with a 5,6,7,8-O-tetramethylated A-ring, failed to impair melanosome transport (S8 Fig), suggesting that the number of O-methylations does not yield an additive effect on melanosome transport. In addition, 5,7-O-dimethyl flavone, which was reported to enhance melanogenesis [38], did not show any effect on melanosome transport when compared with wogonin ( S8 Fig). These flavone-treated cells appeared to be dendritic spine-shaped as compared to the star-shaped control cells. Collectively, O-mono-methylation, not multimethylation, of the flavone A-ring is likely essential. The bioactivities of flavones often refer to their antioxidant activities. The antioxidant capacity of flavones is generally related to the number of hydroxyl groups and available H atoms [39]. Baicalein was reported as a strong antioxidant but did not induce melanosome aggregation (S2 Fig) [40]; thus, the effect of flavones on melanosome transport is likely not linked with their antioxidant behavior. Further research is needed to assess whether hydroxyl groups in the A-ring are required for melanosome aggregation. We also showed that the effects of mono-O-methyl flavones on melanosome transport were strong in the following order: 8-, 6-, and 7-O-monomethylated (Fig 15). Thus, wogonin was the most potent inhibitor of melanosome transport among the tested flavones. Moreover, the conjugation of glycosides to wogonin caused a marked decline in the inhibitory effect on melanosome transport (Fig 14), indicating that the aglycone of wogonin may be more efficacious than the glycoside. We speculate that the aglycone was more cell permeable than the glycoside and that wogonin may act as an inhibitor against both melanogenesis and melanosome transport.
In recent years, melanization has again gained attention as a risk factor for melanoma [41]. Several studies have shown the antitumor effect of wogonin on multiple tumors, including melanomas [9,10]. Zhao et al. used in vitro and in vivo assays to show that wogonin inhibits the cell invasion of metastatic melanoma by suppressing the PI3K/AKT and NF-κB pathways [42]. Here we showed that wogonin inhibits melanogenesis and melanosome transport by targeting MLPH in mouse melanoma B16F10 cells. Garraway et al. reported a correlation between MITF expression and poor prognosis [43]. Notably, we determined that wogonin inhibited MITF expression and melanosome transport in human melanoma MM-AN cells (S9 Fig and S1 Text). MITF not only controls melanogenic enzymes but also regulates multiple cellular processes, including differentiation, proliferation, survival, and motility [27]. Interestingly, wogonin inhibited the mRNA expression of cyclin D1 (CCND1) (S9 Fig), which regulates G1/S phase cell cycle progression and is upregulated in human melanomas [44]. Because CCND1 is transcriptionally regulated by MITF [45], wogonin was presumed to inhibit CCND1 expression in human melanoma cells by downregulating MITF. Therefore, wogonin is a potential therapeutic agent against melanoma.
In this study, we found that an S. baicalensis extract and its active components, wogonin and wogonoside, possessed a strong inhibitory effect on melanogenesis without cytotoxicity. In addition, O-methylated flavones, such as wogonin, reversibly suppress melanosome transport via MLPH downregulation. Our findings reveal the potential applicability of S. baicalensis extracts and its flavones for skin lightening and the treatment of hyperpigmentation.