Identification of (Z)-2-benzylidene-dihydroimidazothiazolone derivatives as tyrosinase inhibitors: Anti-melanogenic effects and in silico studies

Graphical abstract


Introduction
Tyrosinase (TYR, EC 1.14.18.1) is the rate-limiting enzyme in melanogenesis and is found in mammals, plants, fungi, and bacteria. Human tyrosinase (hTYR), which belongs to the mammalian tyrosinase family, is a type 3 copper-containing metalloenzyme and a glycoprotein comprised of 529 amino acids. At its active site, two magnetically bound copper ions (Cu a and Cu b ) are linked by a hydroxo ligand in the met state and coordinated with six histidine residues [1]. The primary function of hTYR is to oxidize L-tyrosine to L-DOPA by acting as monophenolase and subsequently to convert L-DOPA to dopaquinone by acting as a diphenolase [2]. This double oxidation process leads to the formation of melanin. Two other enzymes, tyrosinase-related proteins 1 and 2 (hTYRP1 and hTYRP2) are also known to participate in the production of melanin. The final products of melanogenesis are eumelanin (a dark brown to black pigment) and pheomelanin (a yellow/red pigment) [3]. Melanogenesis occurs in melanocytes, a specialized type of dendritic cell, located in skin, hair bulbs, and eyes. At the subcellular level, melanins are stored in melanosomes and importantly protect skin against UV radiation and free radicals. However conditions, which include melasma, lentigo, congenital melanocytic naevi, erythormelanosis follicularis faciei et colli, post-inflammatory hyperpigmentation, and erythema dyschromicum perstans [4]. Because hTYR catalyzes the first two steps of melanogenesis, most of the efforts made to suppress or reduce melanin production have been directed at the development of effective TYR inhibitors. This strategy has proven to be effective and a strong correlation has been demonstrated between TYR inhibition and levels of melanin generated [5]. Thus, hTYR is considered an attractive target for reducing melanogenesis but is rarely used due to the high price of hTYR and the difficulties associated with their productions in stable forms [5]. To assay TYR inhibitory activity of potential tyrosinase inhibitors, mushroom (Agaricus bisporus) tyrosinase (abTYR) is widely used as a study model because of its availability, low cost and reliable results.
During our studies on this topic over past decades, we have identified a number of benzylidene derivatives with the b-phe nyl-a,b-unsaturated carbonyl scaffold more potently inhibit tyrosinase than kojic acid, a representative tyrosinase inhibitor ( Fig. 1) [29,.
Celecoxib and rofecoxib are nonsteroidal anti-inflammatory drugs. Although rofecoxib was withdrawn from the market due to increased heart attack risk, rofecoxib with a higher log P value penetrated the blood-brain barrier better than celecoxib. As such, log P values are closely related to brain penetration of drugs and absorption of drugs into the skin and gastrointestinal tract. Previously, we synthesized (Z)-5-(3-hydroxy-4-methoxybenzylidene)-2-iminothiazolidin-4-one (MHY773, Fig. 1), which possesses a 2-iminothiazolidin-4-one template, and found this compound potently inhibited mushroom tyrosinase (IC 50 = 2.87 lM for monophenolase and 8.06 lM for diphenolase) and melanogenesis in B16F10 melanoma cells [56]. The effectiveness of whitening agents depend on how well they access melanocytes in the epidermal basal layer, and skin permeability is known to be closely related to lipophilicity. As shown in Fig. 2, the chemical structure of 5,6-dihydroimindazo[2,1-b]thiazol-3(2H)-one (DHIT) is similar to that of 2-iminothiazolidin-4-one, which is more lipophilic, and thus, we expected derivatives with the DHIT template to potently inhibit tyrosinase and melanin production. Thus, we selected as a novel core template for tyrosinase inhibitors and designed and synthesized 11 DHIT derivatives with a 2-benzylidene group bearing a variety of substituents. These 11 derivatives were evaluated for mushroom tyrosinase inhibitory activity at the enzyme level and for tyrosinase activity and melanogenesis in cell-based systems. The underlying mechanisms involved were investigated using Lineweaver-Burk double reciprocal plots, and to investigate the possibility that these derivatives inhibit human tyrosinase, a homologous model for human tyrosinase was prepared and docking simulations were performed.
The structures of the eleven DHIT derivatives were determined by 1 H and 13 C NMR and mass (MS) spectroscopy. The double bond geometries of all 11 derivatives (1a -1 k) were assigned to the (Z)configuration based on vicinal 1 H, 13 C-coupling constants ( 3 J) in proton-coupled 13 C NMR spectra. According to a report by Nair et al. [70], different vicinal 1 H, 13 C-coupling constants in protoncoupled 13 C NMR spectra are observed for geometric isomers of a variety of trisubstituted a ,b-unsaturated carbonyl compounds, including 5-membered and 6-membered exocyclic methylene carbonyl compounds. As depicted in Fig. 3, vicinal coupling constants between the amide carbonyl C-atom C(1) and the olefinic H-atom at C(3) in proton-coupled 13 C NMR spectra depended on double bond stereochemistry: (Z)-isomer: 3 J trans = 11.5 Hz, and (E)isomer: 3 J cis = 6.8 Hz. Generally, 3 J cis values range from 3.6 to 7.0 Hz, while the range of 3 J trans is roughly twice as large (typically ! 10 Hz) [71]. For compound I (Fig. 3), the amide carbonyl carbon had a vicinal 1 H, 13 C-coupling constant of 6.3 Hz ( 3 J (C(3), H-C(1)) ) [71]. The 13 C NMR of compound 1b was measured in proton-coupled 13 C mode, and the 3 J value of C3 in 1j was 6.0 Hz ( Fig. S30 in Supplementary data), suggesting a (Z)configuration.

2.2.
Tyrosinase inhibition -Kinetics, mechanism, and in silico and in vitro studies

Mushroom tyrosinase inhibition
The anti-tyrosinase efficacies of the 11 synthesized DHIT derivatives 1a -1 k against mushroom tyrosinase (mTYR) were investigated as we previously described for the evaluation of mTYR inhibitory activity [65]. Table 1 summarizes % inhibition results for 1a -1 k at a concentration of 25 lM. Kojic acid was used as a positive control, as is typical for tyrosinase inhibitor evaluations. Compounds 1 g, 1 h, 1i, and 1 k with no hydroxyl substituent showed low or no mTYR inhibitory activity, whereas compounds 1a, 1b, and 1f inhibited mTYR more than kojic acid (30% inhibition) at 25 lM. Compound 1a (40% inhibition) with a 4-hydroxyl substituent on the phenyl ring inhibited mTYR slightly more than kojic acid. However, compound 1f (73% inhibition) with a 3-hydroxy-4methoxyl substituent on the phenyl ring inhibited mTYR more than compound 1a. Compound 1b (96% inhibition) with a hydroxyl at the 2-and 4-positions of the phenyl ring had greatest inhibitory effect. Considering the potent mTYR inhibitory activities of compounds 1a and 1b, it seemed that the presence of a 4-hydroxyl on the phenyl ring markedly increased mTYR inhibition, but compounds 1d and 1e, which also possessed a 4-hydroxyl exhibited only low mushroom tyrosinase inhibitory activity (7 and 3%, respectively). These results indicate that the presence of a 4hydroxyl on the phenyl rings of DHIT derivatives can markedly influence mTYR inhibition, but that the additional presence of a 3-alkoxyl (3-methoxyl in 1d, or 3-ethoxyl in 1e) substituent or a 3,5-dialkoxyl (3,5-dimethoxyl in 1j) suppresses 4-hydroxylinduced decreases in mTYR activity. Compounds 1b and 1c had a 4-hydroxyl substituent and at a different position on the phenyl ring had an extra hydroxyl substituent. Interestingly, these com-pounds inhibited mTYR to different extents, that is, compound 1b inhibited mTYR by 96% whereas compound 1c inhibited mTYR by only 18%. These results indicate that the position of an extra hydroxyl substituent on the 4-hydroxyphenyl ring can greatly influence  suggests that compounds with a 2,4-dihydroxyl or 3-hydroxy-4methoxyl substituent on the b -phenyl ring potently inhibit mTYR (Fig. 1). In accord with our collated SAR results, compound 1b with a b À2,4-dihydroxyphenyl group and compound 1f with a b À3hydroxy-4-methoxyphenyl group were found to most potently inhibit mTYR. According to the cumulative docking simulation results, the 2,4-dihydroxyl substituent on the b -phenyl ring contributes to strong binding to tyrosinase through hydrogen bonding Tyrosinase inhibition experiments were conducted using synthesized derivatives or kojic acid at a concentration of 25 lM. a Log P values were obtained using ChemDraw Ultra 12.0. b NI means no inhibition. c KA means kojic acid.  at the active site, implying that the 2,4-dihydroxyl substituent plays an important role in tyrosinase inhibition. In some compounds [53,58,61,62,64,66], both hydroxyl groups of the 2,4-dihydroxyl substituent participate in hydrogen bonding as hydrogen bond donors, and in some compounds [49,60,61,67] only one of the two hydroxyls participates in hydrogen bonding as a hydrogen bond donor. Because DHIT derivatives potently inhibited mushroom tyrosinase, we investigated the IC 50 values of the synthesized DHIT derivatives. The three active derivatives 1a, 1b, and 1f dosedependently inhibited mTYR (data not shown). The IC 50 value of kojic acid, the positive control, was 84.41 lM, while those of DHIT derivatives 1a and 1f were 36.14 and 17.10 lM, respectively, indicating that they were 2.3-and 4.9-fold stronger tyrosinase inhibitors than kojic acid. Compound 1b inhibited mTYR most with an IC 50 value of 0.88 lM, which showed it inhibited mTYR 100-fold more than kojic acid. The other DHIT derivatives had IC 50 values above 100 lM.
Log P values of the synthesized DHIT derivatives were obtained using ChemDraw Ultra 12.0. As indicated in Table 1, DHIT derivatives had log P values ranging from 1.16 to 1.81, which showed they were more lipophilic than the corresponding derivatives with the 2-iminothiazolidin-4-one template.

Determination of the inhibitory mechanism by enzyme kinetics
Since compounds 1b and 1f most potently inhibited mTYR, they were subjected to kinetic study. Lineweaver-Burk double reciprocal plots were used to determine their inhibitory modes of action. Kinetic analyses were carried out using four concentrations of Furthermore, the kinetic studies showed that K i values were 1.50 Â 10 À8 , 1.33 Â 10 À8 , and 1.72 Â 10 À8 M at 0.0625, 0.125, 0.25 lM of 1b, and 3.31 Â 10 À6 , 5.19 Â 10 À6 , and 3.44 Â 10 À6 M at 1, 2, and 4 lM of 1f, respectively. The binding abilities of compounds 1b and 1f to the active site of tyrosinase were also supported by in silico docking simulation results (Figs. 5 and 6).

2.2.3.
In silico studies of compounds 1b and 1f using the crystal structure of mTYR and a hTYR homology model A docking study was performed using Schrodinger suite (release 2021-1) to further investigate the tyrosinase enzyme inhibitory effects of compounds 1b and 1f. Both compounds potently inhibited mTYR in the mushroom tyrosinase assay (Table 1), and kinetic studies confirmed the competitive nature of these inhibitions (Fig. 4). To determine whether 1b and 1f bind directly to the active sites of mTYR and hTYR, we docked compounds 1b and 1f to the crystal structure of mTYR and to a prepared human tyrosinase homology model, respectively, and compared the results with those obtained with kojic acid.
2.2.3.1. Docking studies of derivatives 1b or 1f or kojic acid with mTYR. The crystal structure of mTYR (Agaricus bisporus tyrosinase, PDB ID: 2Y9X) was imported from the Protein Data Bank and docked against 1b, 1f, and kojic acid. Interactions between these three compounds and mTYR are shown in Fig. 5a in 2D and 3D. Docking results showed 1b and 1f occupied the same binding site as kojic acid. To further examine the natures of these interactions we considered pi-pi stacking, hydrophobic interactions, hydrogen bonding, and metal coordination. Kojic acid formed one hydrogen bond with the amino acid residue Met280 using the 5-hydroxyl group of its 4-pyrone ring at a distance of 2.20 Å. In addition, the 2-hydroxymethyl group of its 4-pyrone ring coordinated with Cu401 at a distance of 2.36 Å, and the 4-pyrone ring interacted with His263 by pi-pi stacking, and also with Val283 hydrophobically. Based on these interactions, the recorded docking score for kojic acid was À 4.4 kcal/mol. Interestingly, the binding interactions between mTYR and 1b and kojic acid were closely related. Like kojic acid, the 4-hydroxyl group of the phenyl ring of 1b also coordinated with Cu401, although notably, the 4-hydroxyl group of the phenyl ring of 1b also coordinated with Cu400 and formed salt bridges with Cu400 and Cu401 at distances of 2.28 and 2.32 Å, respectively. However, unlike kojic acid, 1b did not form hydrogen bonds though it did form two pi-pi stacking interactions with His259 and His263. In addition, 1b interacted hydrophobically with Phe264, Met280, and Val283. Due to these interactions, 1b had a docking score of À 6.7 kcal/mol, that is, docking analysis indicated it binds to the active site of tyrosinase much more strongly than kojic acid. The 3D structures in Fig. 5 also show that in its bound position 1b is located closer to the two copper ions than kojic acid. On the other hand, 1f, unlike kojic acid and 1b, did not coordinate with copper ions or form a salt bridge, and made one hydrogen bond with Asn260 using the 3-hydroxyl group on its phenyl ring. Interestingly, 1f interacted hydrophilically with His85, Hie244 (a protonated histidine form), His259, Asn260, and His263, and achieved a docking score of À 4.6 kcal/mol, that is, a score intermediates between those of kojic acid and compound 1b. These results, which are in line with our kinetic study results, suggest that the strong mTYR inhibitory activities of 1b and 1f are due to strong binding with the active site of mTYR.
We also performed docking studies on the (E)-isomers of compounds 1b and 1f to determine whether the geometry of the compounds plays a role in the protein-binding interactions at the active site of tyrosinase, as shown in Fig. 5b. Surprisingly, both compounds still occupied the active site of the tyrosinase enzyme, but showed lower docking scores compared to the (Z)-isomers, 1b and 1f. The lower docking score of the (E)-isomer of 1b may be due to the loss of salt bridges with copper ions. Also, the (E)-isomer of compound 1f showed reverse conformation to that of the (Z)isomer in the active site of tyrosinase enzyme. These docking results suggest that only the (Z)-isomers of 1b and 1f strongly inhibit the activity of the tyrosinase enzyme.

2.2.3.2.
Docking studies of derivatives 1b or 1f or kojic acid with a hTYR homology model. After confirming direct binding between 1b or 1f and mTYR we performed docking simulations using a hTYR homology model. The model was based on human tyrosinaserelated protein 1 (hTRP1, PDB: 5M8Q) which shares 45.81% sequence identity with hTYR ( Fig. S37 in Supplementary data). Binding interactions between 1b, 1f, or kojic and the hTYR homology model are shown in Fig. 6a in 2D and 3D.
As shown by the 3D structure, the 2-hydroxymethyl group on the 4-pyrone ring of kojic acid coordinates with the two zinc ions (Zn6 and Zn7) at distances of 2.16 and 2.21 Å, respectively. Kojic acid also forms a hydrogen bond with Asn364 using the 5hydroxyl group of its 4-pyrone ring and interacts by pi-pi stacking with His367 with a docking score of À 4.9 kcal/mol. This means that the binding affinity of kojic acid for the hTYR homology model was similar to that for mTYR.
As regards compound 1b, the 4-hydroxyl group on its phenyl ring coordinated and formed salt bridges with both zinc ions at distances of 2.16 and 2.20 Å, respectively, which were approximately the same as distances observed for kojic acid. Like kojic acid, the 2hydroxyl group on the phenyl ring of 1b formed a hydrogen bond with Met374 and interacted by pi-pi stacking with His363 and His367. Additionally, the 2-hydroxyl group on the phenyl ring and the DHIT carbonyl moiety of 1b interacted hydrophobically with positively charged Lys306, Ser360, His363, Asn364, His367, Ser375, and Gln376 amino acid residues, which might explain the stronger binding of 1b than kojic acid to hTYR. The docking score of 1b was À 7.0 kcal/mol (Fig. 6b). On the other hand, compound 1f showed a binding pattern similar to that with mTYR. No metal interaction with zinc ions was observed, and only one hydrogen bond was observed between Asn364 and the 3hydroxyl group on its phenyl ring. Interestingly, the DHIT moiety of 1f interacted with a negatively charged amino acid (Asp199) as shown in 2D in Fig. 6a. The docking score of 1f was À 3.9 kcal/ mol. In summary, derivatives 1b and 1f were both found to bind directly with the mTYR and hTYR homology models, and docking simulations of both derivatives with the mTYR and the hTYR homology model were similar.

Cell study
Since 1b and 1f exhibited potent TYR inhibitory activity at the enzyme level, we investigated whether they also inhibited TYR at the cellular level using B16F10 murine melanoma cells.
2.2.4.1. Cell cytotoxicity study. Before examining inhibitory effects of compounds 1b and 1f on cellular tyrosinase activity, we examined their cytotoxic effects on B16F10 murine cells using the EZcytox assay (Daeil Lab Service, Seoul, Korea). After B16F10 cells had been cultured for 24 h, cells were treated with different concentrations (0, 1, 2, 5, 10 or 20 lM) of 1b or 1f for 48 h in a humidified atmosphere. Optical densities (ODs) were measured at 450 nm using a microplate reader.
Cytotoxicity results are shown in Fig. 7. Neither 1b nor 1f had a significant cytotoxic effect on B16F10 cells at concentrations up to 20 lM. Therefore, subsequent cell-based assays on TYR inhibition and intracellular melanogenesis were performed using 1b and 1f concentrations of 20 lM.

Cellular tyrosinase inhibition. B16F10 cells co-stimulated
with a -melanocyte-stimulating hormone (a -MSH) and 3isobutyl-1-methylxanthine (IBMX) were used to investigate the inhibitory effects of compounds 1b and 1f on mammal cellular tyrosinase. Kojic acid was used as the positive control. After incubation for 24 h, B16F10 melanoma cells were pretreated with 20 lM of kojic acid or derivatives 1b or 1f at concentrations of 0, 5, 10, or 20 lM for 1 h and then co-treated with 1 lM of a -MSH and 200 lM of IBMX for 48 h to enhance tyrosinase activity. Cellular tyrosinase activities were determined by measuring optical densities at 492 nm.
Cellular tyrosinase activity results for derivatives 1b and 1f are shown in Fig. 8. Exposure of B16F10 cells to a -MSH and IBMX significantly increased tyrosinase activity to 245% versus non-treated controls and pretreatment with 1b or 1f significantly and concentration-dependently reduced this increase in tyrosinase activity. Compound 1b at 10 lM inhibited cellular TYR to the same extent as compound 1f or kojic acid at 20 lM.

Intracellular melanin inhibition.
To determine whether the inhibitory effects of compounds 1b and 1f on cellular TYR influence melanogenesis, cellular melanin production was measured. As was performed in TYR inhibition experiments, B16F10 cells costimulated with a -MSH and IBMX were used to compare the inhibitory effects of 1b and 1f on melanogenesis. Kojic acid was used as the positive control. After incubation for 24 h, B16F10 melanoma cells were pretreated with kojic acid at 20 lM or compounds 1b or 1f at 0, 5, 10, or 20 lM for 1 h, and then co-treated with 1 lM of a -MSH and 200 lM of IBMX for 48 h to increase melanin production. Intracellular melanin contents were determined by measuring optical densities at 405 nm. Intracellular melanin content results are provided in Fig. 9. Cotreatment with a -MSH and IBMX increased melanin production to 285% vs. non-treated controls but pretreatment with 1b or 1f significantly and concentration-dependently suppressed this increase. Compound 1b at 5 lM suppressed intracellular melanin production by the same extent as kojic acid at 20 lM, and at 20 lM, compound 1f had a stronger inhibitory effect on melanogenesis than kojic acid. Furthermore, these inhibitory effects on intracellular melanin contents were similar to those observed on intracellular tyrosinase  activity, which indicated the anti-melanogenesis effects of 1b and 1f were caused by their TYR inhibitory effects.

Inhibition of extracellular melanin levels.
To determine whether melanin levels in cell culture media were influenced by compounds 1b or 1f, melanin contents in B16F10 cell culture media after treatments with derivatives at different concentrations in the presence of a -MSH and IBMX for 48 h were analyzed by measuring optical densities at 405 nm using a microplate reader. Melanin levels in media are shown in Fig. 10. These levels were increased by co-treatment with a -MSH and IBMX to 178% vs. nontreated controls. However, pretreatment with 1b or 1f significantly and concentration-dependently suppressed these increases. Compound 1b at 5 lM and compound 1f at 20 lM suppressed these increases more than kojic acid at 20 lM. These results were similar to those obtained for intracellular melanin contents and indicate that 1b and 1f both reduced intracellular melanin contents and melanin release to the extracellular compartment.

Conclusions
Eleven 5,6-dihydroimindazo[2,1-b]thiazol-3(2H)-one (DHIT) derivatives bearing various benzylidene groups were designed and synthesized as potential tyrosinase inhibitors. Three derivatives (compounds 1a, 1b, and 1f) were found to inhibit mushroom tyrosinase more than kojic acid, and compound 1b with a 2,4dihydroxyphenyl ring had an IC 50 value $ 100-fold lower than kojic acid. Kinetic studies demonstrated that compounds 1b and 1f, which both inhibited mushroom tyrosinase activity more than kojic acid, were competitive inhibitors. In silico docking simulation results supported that 1b and 1f compete with L-tyrosine (a tyrosinase substrate) for the active site of tyrosinase. Cell-based experiments performed on B16F10 melanoma cells demonstrated that compounds 1b and 1f suppressed melanogenesis more than kojic acid due to their greater inhibitory effects on cellular tyrosinase. Our in silico study conducted using a human tyrosinase homology model supported the notion that 1b and 1f are strong candidate anti-melanogenesis agents for the treatment of diseases associated with hyperpigmentation. Furthermore, these results suggest that DHIT might be a useful template for the development of novel and potent tyrosinase inhibitors.

General methods
Reagents and chemicals were obtained commercially and used without further purification. All reactions were monitored by thin layer chromatography (TLC) using Merck precoated 60F 245 plates, and reaction mixtures were purified by column chromatography using MP Silica 40-63 (60 Å). Mass spectroscopy was performed in electrospray ionization (ESI) positive mode using an Expression CMA spectrometer (Advion Ithaca, NY, USA). 1 H NMR (400 and 500 MHz) and 13 C NMR (100 and 125 MHz) data were obtained using a Varian Unity INOVA 400 spectrometer or a Varian Unity AS500 spectrometer (Agilent Technologies, Santa Clara, CA, USA). DMSO d 6 and CDCl 3 were used as NMR solvents. Chemical shift values were recorded in parts per million (ppm) and coupling constants (J) were recorded in hertz (Hz). The following abbreviations are used: s (singlet), d (doublet), t (triplet), q (quartet), m (multiplet), and brs (broad singlet).

Synthesis of 5,6-dihydroimidazo[2,1-b]thiazol-3(2H)-one (2).
To a stirred solution of 2-thioimidazolidine (5.00 g, 48.94 mmol) and sodium acetate (20.07 g, 244.67 mmol) in EtOH (150 mL) was added ethyl chloroacetate (10.47 mL, 97.82 mmol) at 0°C. The reaction mixture was refluxed for 21 h and cooled, and the precipitate obtained was removed by filtration, washed with EtOH, and the filtrate was partitioned between dichloromethane and water. The organic layer was concentrated in vacuo and the resultant residue was purified by silica gel column chromatography using using an EZ-cytox assay method. Cells were treated with 1b and 1f for 48 h and optical densities were measured after treatment with EZ-cytox solution for 1 h. Results are presented as the means ± standard errors of five independent experiments. A solution of 2 (100 mg, 0.70 mmol), benzaldehyde (1 equiv.), and NaOAc (173 mg, 2.11 mmol) in acetic acid (1.0 mL) was heated at 80°C for 2 -12 h, cooled, and water (5 mL) was added. The precipitates formed were filtered, and washed with water to give 1a -1 k as a solid in yields of 36.9 -87.5%.

Mushroom tyrosinase inhibition assay
The tyrosinase inhibitory activity assays on the synthesized DHIT derivatives 1a -1 k was performed using mushroom tyrosinase (mTYR), as previously described [72] with minor modification.
Briefly, a 200 lL mixture containing tyrosinase solution (20 lL, 200 units), a DHIT derivative (10 lL, final concentration 25 lM), and To calculate the concentration required to inhibit enzyme activity by 50% (IC 50 ), tyrosinase inhibition percentages of each DHIT derivative were obtained at 5 or more different concentrations. IC 50 values were calculated by plotting linear regression curves of percentage inhibitions versus derivative concentrations. The negative control was obtained by adding dimethyl sulfoxide (DMSO) instead of a DHIT derivative. Kojic acid was used as the positive control.

Kinetics of mTYR inhibitions by 1b and 1f
Lineweaver-Burk plots were used to determine modes of mTYR

4.2.3.
In silico study on interactions between tyrosinase and compounds 1b or 1f or kojic acid 4.2.3.1. In silico studies of interactions between mushroom tyrosinase and DHIT derivatives 1b and 1f.. Docking studies on compounds 1b and 1f and kojic acid were performed using Schrodinger suite (2021-1) as previously described protocols [73] with slight modification. The crystal structure of mTYR (PDB ID 2Y9X) was imported from the Protein Data Bank (PDB) using Maestro 12.4 Protein Preparation Wizard and prepared in protein preparation wizard by removing unwanted protein chains. To further refine the structure, hydrogen atoms were added, water molecules > 3 Å from the ligand were removed, and the structure was minimized. In minimized protein structures, glide grid and active sites were determined using the tyrosinase binding site obtained from the PDB and the literature [74][75][76]. The structures of 1b, 1f, and kojic acid were then imported into the entry list of Maestro in CDXML format. Prior to ligand docking, the structures of 1b, 1f, and kojic acid were developed using LigPrep. Compounds were then docked to the glide grid using Glide from the Maestro task list [77]. Binding affinities and ligand-protein interactions were obtained using the glide extra precision (XP) method [78].

4.2.3.2.
In silico studies of interactions between the hTYR homology model and 1b, 1f, and kojic acid. For in silico studies on DHIT derivates 1b and 1f an hTYR homology model was created using the Swiss model Online Server and Schrodinger Suite (2020-2). The hTYR (P14679) protein sequence was imported from the UniProt database and a homology model was developed on the Swiss model online server using the human TRP1 (PDB: 5M8Q) template. The model was further processed using Schrödinger suite and verified using Schrodinger Prime (a homology modeling tool in Schrödinger suite). Compounds 1b and 1f and kojic acid were docked to a processed human homology model using a protocol similar to that mentioned for mTYR docking.

Cell culture
B16F10 murine melanoma cells were acquired from the American Type Culture Collection (ATCC, Manassas, VA, USA) and cultured in Dulbecco's modified Eagle's medium (DMEM), supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin in a humidified 5% CO 2 atmosphere at 37°C. These cells were used for cell viability, melanin content, and cellular tyrosinase activity assays in 96-or 6-well culture plates.

Cell viability analysis
Cell viability assays were performed at 37°C as previously described [79] with minor modification using the EZ-cytox assay. In brief, B16F10 melanoma cells were seeded in 96-well plates at a density of 1 Â 10 4 cells/well and incubated in a humidified 5% CO 2 atmosphere for 24 h at 37°C. Cells were then treated with compounds 1b or 1f at 0, 1, 2, 5, 10, or 20 mM/well for 48 h. EZcytox solution (10 mL) (Daeil Lab Service, Seoul, Korea) was then added to each well and incubated for 1 h at 37°C. Cell viabilities were assessed by measuring absorbances at 450 nm using a microplate reader (VersaMax TM , Molecular Devices, Sunnyvale, CA, USA). All experiments were independently conducted three times.

Cellular tyrosinase inhibition assays of compounds 1b and 1f in B16F10 cells
Cellular tyrosinase inhibition assays were conducted as previously described [80] with minor modification. B16F10 cells were seeded at a density of 1 Â 10 5 cells/well in 6-well plates and incubated in a humidified 5% CO 2 atmosphere for 24 h at 37°C. Cultured B16F10 cells were then treated with 1b, 1f, or kojic acid dissolved in DMSO at final concentrations of 0, 5, 10, or 20 mM for 1b and 1f or 20 mM for kojic acid for 1 h. a-MSH (final concentration: 1 mM) and IBMX (final concentration: 200 mM) were then added, and cells were incubated in a humidified 5% CO 2 atmosphere for 48 h at 37°C. Cells were then washed twice with PBS, lysed by adding 45 mM of phosphate buffer (100 mL) containing 1% Triton X-100 (5 mL) and 1% PMSF (5 mL, phenylmethylsulfonyl fluoride) and frozen at À80°C for 1 h. Lysates were clarified by centrifugation at 12,000 rpm for 30 min at 4°C. Cell lysate supernatants (80 mL/well) in a 96-well plate were then mixed with 20 mL of L-DOPA (2 mg/mL in distilled water) and incubated for 30 min at 37°C. Absorbances of reaction mixtures at 492 nm were recorded using a microplate reader (VersaMax TM , Molecular Devices, Sunnyvale, CA, USA). Kojic acid was used as the positive control. Protein concentrations were determined using Bicinchoninic Acid (BCA) protein assay reagent using Bovine Serum Albumin (BSA) as the standard (Thermo Scientific, Rockford, IL, USA). All experiments were independently conducted three times.

Determinations of intra-and extracellular melanin contents in B16F10 melanoma cells
The inhibitory effects of compounds 1b and 1f on intracellular melanin contents were investigated as previously described [81]. B16F10 cells were seeded at a density of 1 Â 10 5 cells/well in 6well plates and incubated in a humidified 5% CO 2 atmosphere for 24 h at 37°C. DHIT derivatives were dissolved in DMSO, and cultured cells were pretreated with compound 1b or 1f at 0, 5, 10, or 20 mM, or kojic acid at 20 mM for 1 h, 1 mM of a-MSH and 200 mM of IBMX were then added, and cells were incubated in a humidified 5% CO 2 atmosphere for 48 h at 37°C. For extracellular melanin contents, the optical density of the cell culture media was directly measured at 405 nm. On the other hand, for intracellular melanin contents, the following procedure was performed. Briefly, after washing twice with PBS, adherent cells were detached by incubation in Trypsin/EDTA for 1 min. Pellets were dissolved in 100 lL of 1 N NaOH and then incubated for 1 h at 60°C to dissolve the melanin. Intracellular melanin was quantified by measuring optical densities at 405 nm using a microplate reader (VersaMax TM , Molecular Devices, Sunnyvale, CA, USA). Results were normalized versus cell pellet total protein using BCA protein assay reagent using BSA as the standard (Thermo Scientific, Rockford, IL, USA). The intracellular melanin contents were calculated using the following equation: (DOD sample /DOD control ) Â 100%; where DOD sample = the optical density of the test compound, and DOD control = the optical density of control. Experiments were performed in triplicate.

Statistical analysis
Statistical analysis was carried out using GraphPad Prism (La Jolla, CA, USA) and results are presented as means ± standard errors of means (SEMs). One-way analysis of variance (ANOVA) followed by the Bonferroni post hoc test was used to determine the significances of intergroup differences. Statistical significance was accepted for P values < 0.05.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.