Novel Anti-Melanogenic Compounds, (Z)-5-(Substituted Benzylidene)-4-thioxothiazolidin-2-one Derivatives: In Vitro and In Silico Insights

To confirm that the β-phenyl-α,β-unsaturated thiocarbonyl (PUSTC) scaffold, similar to the β-phenyl-α,β-unsaturated carbonyl (PUSC) scaffold, acts as a core inhibitory structure for tyrosinase, twelve (Z)-5-(substituted benzylidene)-4-thioxothiazolidin-2-one ((Z)-BTTZ) derivatives were designed and synthesized. Seven of the twelve derivatives showed stronger inhibitory activity than kojic acid against mushroom tyrosinase. Compound 2b (IC50 = 0.47 ± 0.97 µM) exerted a 141-fold higher inhibitory potency than kojic acid. Kinetic studies’ results confirmed that compounds 2b and 2f are competitive tyrosinase inhibitors, which was supported by high binding affinities with the active site of tyrosinase by docking simulation. Docking results using a human tyrosinase homology model indicated that 2b and 2f might potently inhibit human tyrosinase. In vitro assays of 2b and 2f were conducted using B16F10 melanoma cells. Compounds 2b and 2f significantly and concentration-dependently inhibited intracellular melanin contents, and the anti-melanogenic effects of 2b at 10 µM and 2f at 25 µM were considerably greater than the inhibitory effect of kojic acid at 25 µM. Compounds 2b and 2f similarly inhibited cellular tyrosinase activity and melanin contents, indicating that the anti-melanogenic effects of both were due to tyrosinase inhibition. A strong binding affinity with the active site of tyrosinase and potent inhibitions of mushroom tyrosinase, cellular tyrosinase activity, and melanin generation in B16F10 cells indicates the PUSTC scaffold offers an attractive platform for the development of novel tyrosinase inhibitors.


Introduction
Skin whitening and depigmentation are widespread practices in some ethnic groups, especially in Asia, Africa, and the Middle East, because of the complex interplay between cultural, social, political, and psychological factors [1], and a lighter complexion is considered a beauty attribute in these regions. About 28% of the world's population undergo skin whitening at least once [2], and according to a market study, the use of skin whitening agents could reach USD 13.7 billion by 2025 [3]. In addition, depigmentation agents are also used to treat skin disorders caused by hyperpigmentation, such as

Chemistry
First, 4-thioxothiazolidin-2-one (1) was prepared as a key intermediate for the syn thesis of the desired (Z)-BTTZ derivatives 2a-2l by reacting thiazolidin-2,4-dione with Lawesson's reagent (2,4-bis(4-methoxyphenyl)-2,4-dithioxo-1,3,2,4-dithiadiphosphetane (90% yield, Scheme 1). The synthesis of the final compounds 2a-2l was achieved using Knoevenagel condensation [37]. The coupling of compound 1 with substituted benzalde hydes in the presence of piperidine in ethanol afforded the twelve (Z)-BTTZ derivative 2a-2l as sole products at yields from 35 to 82%. To examine the effect of substituents on the β-phenyl ring of PUSTC on tyrosinase inhibition, benzaldehydes bearing a hydroxyl methoxyl, bromo, and/or an ethoxyl group were coupled with 4-thioxothiazolidin-2-one (1). Twelve (Z)-BTTZ derivatives with the PUSTC scaffold were synthesized and identi fied using 1 H and 13 C NMR and mass spectroscopy. The geometries of the double bond generated by benzaldehyde coupling with compound 1 were determined using vicinal 1 H and 13 C-coupling constants ( 3 J) in proton-coupled 13 C NMR spectra. Nair and co-worker reported that in the proton-coupled 13 C NMR spectra of a variety of trisubstituted geo metric isomers, which were α,β-unsaturated carbonyl compounds including 5-or 6-mem bered exocyclic methylene carbonyl compounds, different vicinal 1 H, 13 C-coupling con stants ( 3 J) were observed ( Figure 2) [38]. In geometric isomers with a cis positioned car bonyl and β-hydrogen, the carbonyl carbon had vicinal 1 H, 13 C-coupling constants ( 3 J) in the range 4.8 to 7.0 Hz, whereas in the trans position, the carbonyl carbon had vicinal 1 H 13 C-coupling constants above 10 Hz. As shown in Figure 2, the 13 C NMR of compound 2b was measured in the proton-coupled 13 C mode, and the vicinal 1 H, 13 C-coupling constan ( 3 J) of C4 was 7.5 Hz (Supplementary data, S6), indicating that compound 2b was the (Z) isomer. Furthermore, the 1 H and 13 C NMR data of compound 2b matched that of the au thentic compound prepared by reacting compound 1 with 2,4-dihydroxybenzaldehyde in the presence of sodium acetate and acetic acid [37].

Chemistry
First, 4-thioxothiazolidin-2-one (1) was prepared as a key intermediate for the synthesis of the desired (Z)-BTTZ derivatives 2a-2l by reacting thiazolidin-2,4-dione with Lawesson's reagent (2,4-bis(4-methoxyphenyl)-2,4-dithioxo-1,3,2,4-dithiadiphosphetane) (90% yield, Scheme 1). The synthesis of the final compounds 2a-2l was achieved using Knoevenagel condensation [37]. The coupling of compound 1 with substituted benzaldehydes in the presence of piperidine in ethanol afforded the twelve (Z)-BTTZ derivatives 2a-2l as sole products at yields from 35 to 82%. To examine the effect of substituents on the β-phenyl ring of PUSTC on tyrosinase inhibition, benzaldehydes bearing a hydroxyl, methoxyl, bromo, and/or an ethoxyl group were coupled with 4-thioxothiazolidin-2-one (1). Twelve (Z)-BTTZ derivatives with the PUSTC scaffold were synthesized and identified using 1 H and 13 C NMR and mass spectroscopy. The geometries of the double bond generated by benzaldehyde coupling with compound 1 were determined using vicinal 1 H and 13 C-coupling constants ( 3 J) in proton-coupled 13 C NMR spectra. Nair and co-workers reported that in the proton-coupled 13 C NMR spectra of a variety of trisubstituted geometric isomers, which were α,β-unsaturated carbonyl compounds including 5-or 6-membered exocyclic methylene carbonyl compounds, different vicinal 1 H, 13 C-coupling constants ( 3 J) were observed ( Figure 2) [38]. In geometric isomers with a cis positioned carbonyl and β-hydrogen, the carbonyl carbon had vicinal 1 H, 13 C-coupling constants ( 3 J) in the range 4.8 to 7.0 Hz, whereas in the trans position, the carbonyl carbon had vicinal 1 H, 13 C-coupling constants above 10 Hz. As shown in Figure 2, the 13 C NMR of compound 2b was measured in the proton-coupled 13 C mode, and the vicinal 1 H, 13 C-coupling constant ( 3 J) of C4 was 7.5 Hz (Supplementary data, S6), indicating that compound 2b was the (Z)-isomer. Furthermore, the 1 H and 13 C NMR data of compound 2b matched that of the authentic compound prepared by reacting compound 1 with 2,4-dihydroxybenzaldehyde in the presence of sodium acetate and acetic acid [37].

Mushroom Tyrosinase Inhibition and Values of Logarithm of Partition Function
Although it has been reported that melanogenesis is regulated by hormones [39], tyrosinase inhibition is the most promising method of inhibiting melanogenesis. L-Tyrosine and L-dopa are not only substrates for melanogenesis, but also act as regulators of the melanogenesis pathway [40]. The anti-tyrosinase efficacies of the 12 synthesized (Z)-BTTZ Scheme 1. Synthetic scheme for (Z)-BTTZ derivatives. Reagents and conditions: (a) Lawesson's reagent, toluene, reflux, 1 h, 90%; (b) substituted benzaldehydes, piperidine, ethanol, reflux, 15-30 min, 35-82%.  Relationship between the geometry of the double bond and C,H-spin-coupling constants over three bonds and the 3 J value at C(4) of compound 2b.

Mushroom Tyrosinase Inhibition and Values of Logarithm of Partition Function
Although it has been reported that melanogenesis is regulated by hormones [39], tyrosinase inhibition is the most promising method of inhibiting melanogenesis. L-Tyrosine and L-dopa are not only substrates for melanogenesis, but also act as regulators of the melanogenesis pathway [40]. The anti-tyrosinase efficacies of the 12 synthesized (Z)-BTTZ Figure 2. Relationship between the geometry of the double bond and C,H-spin-coupling constants over three bonds and the 3 J value at C(4) of compound 2b.

Mushroom Tyrosinase Inhibition and Values of Logarithm of Partition Function
Although it has been reported that melanogenesis is regulated by hormones [39], tyrosinase inhibition is the most promising method of inhibiting melanogenesis. L-Tyrosine and L-dopa are not only substrates for melanogenesis, but also act as regulators of the melanogenesis pathway [40]. The anti-tyrosinase efficacies of the 12 synthesized (Z)-BTTZ derivatives 2a-2l were examined as previously described in a study on mushroom tyrosinase inhibitory activity using kojic acid as a positive control and L-tyrosine as a substrate [39]. All 12 (Z)-BTTZ derivatives exhibited concentration-dependent mushroom tyrosinase inhibition. The IC 50 values of compounds 2a-2l and kojic acid are shown in Table 1. derivatives 2a-2l were examined as previously described in a study on mushroom tyrosinase inhibitory activity using kojic acid as a positive control and L-tyrosine as a substrate [39]. All 12 (Z)-BTTZ derivatives exhibited concentration-dependent mushroom tyrosinase inhibition. The IC50 values of compounds 2a-2l and kojic acid are shown in Table 1. Compound 2b most potently inhibited mushroom tyrosinase with an IC50 value of 0.47 ± 0.97 µM, whereas the representative tyrosinase inhibitor, kojic acid, had an IC50 value of 66.30 ± 0.75 M. That is, compound 2b with two hydroxyls at the two and four positions on the β-phenyl ring of the PUSTC scaffold was 141 times more potent than kojic acid. Compound 2f, which had a 3-hydroxy and a 4-methoxyl substituent on the β-phenyl ring also exhibited strong tyrosinase inhibition with an IC50 value of 15.24 ± 1.68 M. Interestingly, compound 2d (IC50: 70.75 ± 1.00 µM), which had a methoxyl at the three position and a hydroxyl at the four position had a 4.6-fold lower mushroom tyrosinase inhibitory activity than compound 2f. In addition to compounds 2b and 2f, five other compounds, namely, 2a with a 4-hydroxyphenyl, 2c with a 3,4-dihydroxyphenyl, 2i with a 3-bromo-4-hydroxyl, 2j with a 4-methoxyl, and 2l with a 3,4-dimethoxyl, also inhibited mushroom tyrosinase more potently than kojic acid, with IC50 values of 28.05 ± 1.16, 47.41 ± 2.18, 26.27 ± 4.10, 30.83 ± 1.41, and 23.31 ± 0.28 M, respectively; that is, these five compounds had a 1.4-to 2.8-fold stronger tyrosinase inhibitory activity than kojic acid. Compound 2e, in which the 3-methoxyl substituent of 2d was replaced with the more sterically demanding ethoxyl substituent, had slightly less tyrosinase inhibitory activity than kojic acid (IC50: 70.75 ± 1.00 µM to 92.81 ± 0.89 µM). Compound 2k, in which the 2,4-dihydroxyl substituents of compound 2b were replaced with 2,4-dimethoxyl, showed a dramatic reduction in tyrosinase inhibitory activity (from an IC50 of 0.47 ± 0.97 M to 126.35 ± 0.46 µM). Compound 2b with a 2,4-dihydroxyl substituent had the lowest IC50 value, but compounds 2a and 2h in which the 2-or 4-hydroxyl groups of compound 2b were removed, respectively, had a 60-and 314-fold lower tyrosinase inhibitory activities than 2b. These results indicate that the 2,4-dihydroxyl substitution on the β-phenyl ring of the PUSTC Compound 2b most potently inhibited mushroom tyrosinase with an IC 50 value of 0.47 ± 0.97 µM, whereas the representative tyrosinase inhibitor, kojic acid, had an IC 50 value of 66.30 ± 0.75 µM. That is, compound 2b with two hydroxyls at the two and four positions on the β-phenyl ring of the PUSTC scaffold was 141 times more potent than kojic acid. Compound 2f, which had a 3-hydroxy and a 4-methoxyl substituent on the β-phenyl ring also exhibited strong tyrosinase inhibition with an IC 50 value of 15.24 ± 1.68 µM. Interestingly, compound 2d (IC 50 : 70.75 ± 1.00 µM), which had a methoxyl at the three position and a hydroxyl at the four position had a 4.6-fold lower mushroom tyrosinase inhibitory activity than compound 2f. In addition to compounds 2b and 2f, five other compounds, namely, 2a with a 4-hydroxyphenyl, 2c with a 3,4-dihydroxyphenyl, 2i with a 3-bromo-4-hydroxyl, 2j with a 4-methoxyl, and 2l with a 3,4-dimethoxyl, also inhibited mushroom tyrosinase more potently than kojic acid, with IC 50 values of 28.05 ± 1.16, 47.41 ± 2.18, 26.27 ± 4.10, 30.83 ± 1.41, and 23.31 ± 0.28 µM, respectively; that is, these five compounds had a 1.4-to 2.8-fold stronger tyrosinase inhibitory activity than kojic acid. Compound 2e, in which the 3-methoxyl substituent of 2d was replaced with the more sterically demanding ethoxyl substituent, had slightly less tyrosinase inhibitory activity than kojic acid (IC 50 : 70.75 ± 1.00 µM to 92.81 ± 0.89 µM). Compound 2k, in which the 2,4-dihydroxyl substituents of compound 2b were replaced with 2,4-dimethoxyl, showed a dramatic reduction in tyrosinase inhibitory activity (from an IC 50 of 0.47 ± 0.97 µM to 126.35 ± 0.46 µM). Compound 2b with a 2,4-dihydroxyl substituent had the lowest IC 50 value, but compounds 2a and 2h in which the 2-or 4-hydroxyl groups of compound 2b were removed, respectively, had a 60-and 314-fold lower tyrosinase inhibitory activities than 2b. These results indicate that the 2,4-dihydroxyl substitution on the β-phenyl ring of the PUSTC scaffold markedly increases tyrosinase inhibition. The insertion of a 3-bromo group into the β-phenyl ring of compound 2a did not affect mushroom tyrosinase inhibition (2a vs. 2i). In summary, hydroxyl and methoxyl groups at R 3 conferred greater tyrosinase inhibitory activity than hydrogen, and generally, the presence of a hydroxyl group at R 3 resulted in stronger tyrosinase inhibition than a methoxyl (Figure 3). At R 2 , all the substituents, including hydrogen and bromine, exhibited similar tyrosinase inhibitory activity and the presence of hydroxyl at R 2 resulted in higher tyrosinase inhibitory activity in the presence of a 4-methoxyl group (2f vs. 2j). The hydroxyl substituent at R 1 greatly increased the tyrosinase inhibitory activity in the presence of the 4-hydroxyl substituent. These results suggest that the PUSTC scaffold might be an important template for mushroom tyrosinase inhibitors. sulted in stronger tyrosinase inhibition than a methoxyl (Figure 3). At R2, all the substituents, including hydrogen and bromine, exhibited similar tyrosinase inhibitory activity and the presence of hydroxyl at R2 resulted in higher tyrosinase inhibitory activity in the presence of a 4-methoxyl group (2f vs. 2j). The hydroxyl substituent at R1 greatly increased the tyrosinase inhibitory activity in the presence of the 4-hydroxyl substituent. These results suggest that the PUSTC scaffold might be an important template for mushroom tyrosinase inhibitors.
Log P values for the 12 (Z)-BTTZ derivatives 2a-2l were obtained using ChemDraw Ultra 12.0 and the observed range was 1.26 to 2.47 (Table 1). The log p value of kojic acid was −2.45, which indicated that the (Z)-BTTZ derivatives are more likely to be absorbed by skin than kojic acid. Due to their strong inhibitory activities against mushroom tyrosinase, compounds 2b and 2f were used in docking simulation and kinetic studies and in in vitro experiments on their tyrosinase inhibitory and anti-melanogenic activities.

Modes of Action of Compounds 2b and 2f
To investigate the modes of action of compounds 2b and 2f, a kinetic study was performed using mushroom tyrosinase and different concentrations of L-tyrosine as a substrate in the presence of 2b or 2f. As depicted in Figure 4, inhibitory mechanisms were determined using Lineweaver-Burk plot analysis. The Lineweaver-Burk plot patterns of 2b and 2f were similar. The plots obtained at different concentrations of 2b and 2f merged at single points on the y-axis. Regardless of the concentration, the Vmax values of 2b and 2f were independent of concentration, whereas their KM values increased concentration-dependently. These results show that compounds 2b and 2f are competitive inhibitors of mushroom tyrosinase. Log p values for the 12 (Z)-BTTZ derivatives 2a-2l were obtained using ChemDraw Ultra 12.0 and the observed range was 1.26 to 2.47 ( Table 1). The log p value of kojic acid was −2.45, which indicated that the (Z)-BTTZ derivatives are more likely to be absorbed by skin than kojic acid.
Due to their strong inhibitory activities against mushroom tyrosinase, compounds 2b and 2f were used in docking simulation and kinetic studies and in in vitro experiments on their tyrosinase inhibitory and anti-melanogenic activities.

Modes of Action of Compounds 2b and 2f
To investigate the modes of action of compounds 2b and 2f, a kinetic study was performed using mushroom tyrosinase and different concentrations of L-tyrosine as a substrate in the presence of 2b or 2f. As depicted in Figure 4, inhibitory mechanisms were determined using Lineweaver-Burk plot analysis. The Lineweaver-Burk plot patterns of 2b and 2f were similar. The plots obtained at different concentrations of 2b and 2f merged at single points on the y-axis. Regardless of the concentration, the V max values of 2b and 2f were independent of concentration, whereas their K M values increased concentration-dependently. These results show that compounds 2b and 2f are competitive inhibitors of mushroom tyrosinase.

In Silico Studies of (Z)-BTTZ Derivatives 2b and 2f
In the mushroom tyrosinase enzyme assays, compounds 2b and 2f potently inhibited tyrosinase and kinetic studies indicated that they inhibited it competitively, which suggests they both bind strongly to the active site of mushroom tyrosinase. In silico studies were performed using Schrödinger Suite (release 2021-1) to determine whether tyrosinase inhibition was caused by direct binding between 2b or 2f and the active site of tyrosinase. Docking simulations of 2b and 2f were performed using the crystal structure of mushroom tyrosinase and a human tyrosinase homology model based on human tyrosinase-related protein (hTRP1). The results were compared with kojic acid.

Docking Simulations of Compounds 2b and 2f and Kojic Acid with Mushroom Tyrosinase
The crystal structure of mushroom tyrosinase (PDB ID = 2Y9X) was imported from the protein data bank (PDB) and prepared for docking against 2b, 2f, and kojic acid. The 2D and 3D conformations shown in Figure 5 indicate binding interactions between compounds 2b and 2f and the active site of mushroom tyrosinase. As shown by the figure, both compounds occupied the same binding site as kojic acid, although the interactions involved differed. Interestingly, similar to kojic acid, compound 2b coordinated with the copper ion (Cu401) of tyrosinase using the carbonyl of its thioxothiazolidin-2-one ring. Compound 2b hydrogen bonded (2.39 Å) with Hie244 (protonated His244) instead of Met280. On the other hand, compound 2f did not exhibit metal coordination, but the methoxyl group of its phenyl ring interacted (3.37 and 4.42 Å) with both copper ions of tyrosinase. Interestingly, 2f and 2b adopted different conformations at the active site of mushroom tyrosinase. The phenyl ring of 2b was located far from the copper ions, whereas the phenyl ring of 2f was located close to the copper ions. These interactions resulted in docking scores of −5.12 and −4.56 kcal/mol for 2b and 2f, respectively, and of −4.42 kcal/mol for kojic acid. These binding affinity scores implied that 2b and 2f inhibit mushroom tyrosinase activity more potently than kojic acid and support that both are competitive inhibitors. Furthermore, 2b and 2f had higher binding affinities than kojic acid, which suggested that the PUSTC scaffold, similar to the PUSC scaffold, confers potent tyrosinase inhibitory activity. Interestingly, 2f and 2b adopted different conformations at the active site of mushroom tyrosinase. The phenyl ring of 2b was located far from the copper ions, whereas the phenyl ring of 2f was located close to the copper ions. These interactions resulted in docking scores of −5.12 and −4.56 kcal/mol for 2b and 2f, respectively, and of −4.42 kcal/mol for kojic acid. These binding affinity scores implied that 2b and 2f inhibit mushroom tyrosinase activity more potently than kojic acid and support that both are competitive inhibitors. Furthermore, 2b and 2f had higher binding affinities than kojic acid, which suggested that the PUSTC scaffold, similar to the PUSC scaffold, confers potent tyrosinase inhibitory activity.

Docking Simulation of Compounds 2b and 2f and Kojic Acid with the Human Tyrosinase Homology Model
To further investigate the inhibitory activities of compounds 2b and 2f, we used a human tyrosinase homology model to confirm that these compounds effectively interact with human tyrosinase. For this purpose, we created a human tyrosinase homology model using Schrödinger Suite based on human tyrosinase related protein (hTRP1, PDB ID = 5M8Q) [41,42].
The binding interactions between compounds 2b, 2f, and kojic acid and the human tyrosinase homology model are shown in Figure 6 in 2D and 3D conformations. Kojic acid has been shown to coordinate with a zinc ion (Zn7) of tyrosinase, to form a hydrogen bond at Ser375, and to interact by pi-pi stacking with His367. Compound 2b was unable to coordinate with either of the two zinc ions of tyrosinase but formed two hydrogen bonds and interacted by pi-pi stacking. Compound 2b used one of its two hydroxyl groups on its β-phenyl ring to hydrogen bond with Ser380, and the carbonyl group of the thiooxothiazolidin-2-one ring to hydrogen bond with Gln103. Its pi-pi stacking interaction was formed by the β-phenyl ring and two amino acids, His363 and His367. The lipophilic region of compound 2b was located in a hydrophobic environment created by several amino acid residues (marked green in Figure 6). On the other hand, compound 2f formed only one pi-pi stacking interaction with His202. According to these results, the calculated docking scores for compounds 2b and 2f and kojic acid were −4.98, −4.05, and −4.56 kcal/mol, respectively. formed by the β-phenyl ring and two amino acids, His363 and His367. The lipophilic region of compound 2b was located in a hydrophobic environment created by several amino acid residues (marked green in Figure 6). On the other hand, compound 2f formed only one pi-pi stacking interaction with His202. According to these results, the calculated docking scores for compounds 2b and 2f and kojic acid were −4.98, −4.05, and −4.56 kcal/mol, respectively. Compound 2b had a higher binding affinity than compound 2f or kojic acid with both mTYR and hTYR in the docking simulations. Furthermore, 2b and 2f occupied the same binding sites in mTYR and hTYR as kojic acid (Figures 5 and 6), but their interactions in the binding sites differed. The thiooxothiazolidin-2-one ring of compound 2b and the phenyl ring of compound 2f were placed near metal ions (Cu ++ or Zn ++ , respectively) in the active sites of mTYR and hTYR. Both compounds had similar or higher binding affinities than kojic acid in docking simulations with mTYR and hTYR. These results indicate that the PUSTC scaffold is well tolerated at the active sites of mTYR and hTYR.

Cytotoxic Effects of Compounds 2b and 2f
Before examining the inhibitory effects of compounds 2b and 2f on tyrosinase and melanin biosynthesis, we investigated their cytotoxic effects using murine B16F10 mela-Kojic acid 2f 2b Compound 2b had a higher binding affinity than compound 2f or kojic acid with both mTYR and hTYR in the docking simulations. Furthermore, 2b and 2f occupied the same binding sites in mTYR and hTYR as kojic acid (Figures 5 and 6), but their interactions in the binding sites differed. The thiooxothiazolidin-2-one ring of compound 2b and the phenyl ring of compound 2f were placed near metal ions (Cu ++ or Zn ++ , respectively) in the active sites of mTYR and hTYR. Both compounds had similar or higher binding affinities than Molecules 2021, 26, 4963 9 of 18 kojic acid in docking simulations with mTYR and hTYR. These results indicate that the PUSTC scaffold is well tolerated at the active sites of mTYR and hTYR.

Cytotoxic Effects of Compounds 2b and 2f
Before examining the inhibitory effects of compounds 2b and 2f on tyrosinase and melanin biosynthesis, we investigated their cytotoxic effects using murine B16F10 melanoma cells. The cells were treated with different concentrations of 2b or 2f (0, 1, 2.5, 5, 10, or 25 µM) for 24 and 48 h (Figure 7), respectively, and the cell viabilities were assessed using the EZ-Cytox assay. Neither 2b nor 2f exhibited any significant cytotoxic effect at concentrations of ≤25 µM.

Anti-Melanogenic Activities of Compounds 2b and 2f
Melanin content assays were used to evaluate the inhibitory effects of compounds 2b and 2f on α-MSH-stimulated melanogenesis. Briefly, B16F10 melanoma cells were pretreated with various concentrations (0, 5, 10, or 25 µM) of compounds 2b or 2f or kojic acid (25 µM) for 1 h, and then stimulated with α-MSH (0.5 M for 2b, and 1.0 M for 2f) for 48 h to increase the melanin contents. The inhibitory activities on α-MSH-stimulated melanogenesis were determined by measuring melanin optical densities.

Anti-Melanogenic Activities of Compounds 2b and 2f
Melanin content assays were used to evaluate the inhibitory effects of compounds 2b and 2f on α-MSH-stimulated melanogenesis. Briefly, B16F10 melanoma cells were pretreated with various concentrations (0, 5, 10, or 25 µM) of compounds 2b or 2f or kojic acid (25 µM) for 1 h, and then stimulated with α-MSH (0.5 µM for 2b, and 1.0 µM for 2f) for 48 h to increase the melanin contents. The inhibitory activities on α-MSH-stimulated melanogenesis were determined by measuring melanin optical densities.

Anti-Tyrosinase Activities of Compounds 2b and 2f
The inhibitory activities of compounds 2b and 2f against cellular tyrosinase were assessed using α-MSH-stimulated melanoma cells. B16F10 cells were exposed to various concentrations (0, 5, 10, or 25 μM) of compounds 2b or 2f or kojic acid (25 μM) for 1 h, and then stimulated with α-MSH (0.5 M for 2b, and 1.0 M for 2f) for 48 h to increase tyrosinase activity. The inhibitory effects on tyrosinase activity were evaluated by measuring optical densities.

Anti-Tyrosinase Activities of Compounds 2b and 2f
The inhibitory activities of compounds 2b and 2f against cellular tyrosinase were assessed using α-MSH-stimulated melanoma cells. B16F10 cells were exposed to various concentrations (0, 5, 10, or 25 µM) of compounds 2b or 2f or kojic acid (25 µM) for 1 h, and then stimulated with α-MSH (0.5 µM for 2b, and 1.0 µM for 2f) for 48 h to increase tyrosinase activity. The inhibitory effects on tyrosinase activity were evaluated by measuring optical densities.
As neither compound 2b nor 2f had a significant cytotoxic effect on B16F10 melanoma cells at ≤25 µM, we considered that the anti-melanogenic activities of compounds 2b and 2f ( Figure 8) were caused by tyrosinase inhibition.

Conclusions
To investigate the tyrosinase inhibitory effects of the PUSTC scaffold, we synthesized 12 (Z)-BTTZ derivatives (compounds 2a to 2l) with this scaffold. Seven of the derivatives inhibited mushroom tyrosinase more potently than kojic acid, and compound 2b had an IC 50 value of only 0.47 ± 0.97 µM, which made it >100 times more potent than kojic acid. The Lineweaver-Burk plots of compounds 2b and 2f indicated they both competitively inhibited mushroom tyrosinase, and the docking simulation results showed both bound strongly to the active sites of human and mushroom tyrosinase. Compounds 2b and 2f inhibited tyrosinase activity concentration-dependently in α-MSH-stimulated B16F10 murine melanoma cells, and did so much more effectively than kojic acid. Melanin production was also concentration-dependently reduced by 2b and 2f in α-MSH-stimulated melanoma cells, and again both were markedly more potent than kojic acid. The similarity between the inhibitions of cellular tyrosinase and melanogenesis by compounds 2b and 2f and their minimal cytotoxic effects at ≤25 µM suggested that their anti-melanogenic effects were due to the inhibition of cellular tyrosinase. These results suggest that the PUSTC scaffold offers a template for the design of tyrosinase inhibitors.

General Methods
All reagents were obtained commercially and used without further purification. Reactions were carried out under nitrogen and monitored using thin layer chromatography (TLC, Merck precoated 60F 245 plates). Column chromatography was conducted using MP Silica 40-63, 60 Å. All anhydrous solvents were distilled over Na/benzophenone or CaH 2 . Low resolution mass data were recorded in ESI positive mode on an Expression CMS mass spectrometer (Advion Inc., Ithaca, NY, USA). 1 H and 13 C NMR spectra were obtained using a Varian Unity INOVA 400 or a Varian Unity AS500 unit (Agilent technologies, Santa Clara, CA, USA) at 400 or 500 MHz for 1 H NMR and 100 or 125 MHz for 13 C NMR. CD 3 OD, CDCl 3 , and DMSO-d 6 were used as NMR solvents. All chemical shifts were measured in parts per million (ppm) versus residual solvent or deuterated peaks δ H 7.26 and δ C 77.0 for CDCl 3 , δ H 3.31 for CD 3 OD, δ H 2.48 and δ C 40.0 for DMSO-d 6 ). Coupling constant (J) values are presented in hertz (Hz). The following abbreviations are used for 1 H NMR: s (singlet), d (doublet), t (triplet), q (quartet), brs (broad singlet), and dd (doublet of doublets).

In Silico Study of Interactions between Mushroom Tyrosinase and Compounds 2b and 2f and Kojic Acid
The in silico study was conducted using Schrödinger Suite (2021-1) with minor modifications, as previously described [44]. The crystal structure of mTYR (Agaricus bisporus, PDB: 2Y9X) was downloaded from the Protein Data Bank (PDB) to Maestro12.4's Protein Preparation Wizard, and processed, and unwanted protein chains were removed. To optimize the structure, hydrogen atoms were added, water molecules > 3 Å away from the enzyme were removed, and the structure was minimized. The glide grid and the active site were determined using the binding site of tropolone (a ligand of tyrosinase) obtained from PDB and the literature [45][46][47]. The structures of compounds 2b and 2f and kojic acid were imported into the entry list of Maestro in CDXML format. Before ligand docking, the structures of 2b and 2f and kojic acid were prepared using LigPrep. The compounds were then docked to the enzyme's glide grid using Glide from the Maestro task list [48]. Binding affinities and ligand-protein interactions were obtained using the glide extra precision (XP) method [49].

In Silico Study of Interactions between Compounds 2b and 2f and Kojic Acid and the Human Tyrosinase Homology Model
The hTYR homology model was created using the Swiss-Model online server (https: //swissmodel.expasy.org; accessed on 4 April 2021) and Schrödinger Suite (2020-2) using default settings. The protein sequence of hTYR (P14679) was imported from the UniProt database (https://www.uniprot.org; accessed on 4 April 2021) and the homology model was generated in the Swiss-Model online server using the TRP1 (PDB: 5M8Q) template. The model was further processed using Schrödinger Suite and validated using Schrödinger prime (a homology modeling tool in Schrödinger Suite). Compounds 2b and 2f and kojic acid were docked with the processed human homology model using the same protocols mentioned above for mTYR docking.