Novel coumarin derivatives as potential tyrosinase inhibitors: Synthesis, binding analysis and biological evaluation

2.1 Chemistry

All coumarin-cinnamic acid/benzoic acid derivatives (4a ∼ r, 5a ∼ l) were synthesized as synthetic route in Scheme 1. Substituted salicylaldehyde (1) reacted with ethyl acetaacetate to produce substituted acetyl coumarin (2), followed by the reaction with hydroxylamine hydrochloride to obtain substituted coumarin containing oxime (3). Finally, coumarin containing oxime underwent esterification with substituted cinnamic acid/benzoic acid chloride to yield target compounds. All synthetic derivatives (4a ∼ r, 5a ∼ l) were identified by ¹H NMR, ¹³C NMR and HRMS.

Close

Scheme 1

Synthesis of coumarin-cinnamic acid/benzoic acid derivatives 4a ∼ r and 5a ∼ l. Reagents and condition: (a) Ethyl acetaacetate, Piperidine, EtOH, 80 °C; (b) Hydroxylamine hydrochloride, Pyridine, EtOH, rt; (c) Substituted benzoic acid chloride/cinnamic acid chloride, triethylamine, DCM, rt.

Export to PPT

2.2 Anti-tyrosinase activity and SAR analysis

For finding more active tyrosinase inhibitors, coumarin-cinnamic acid derivatives (4a ∼ i) and coumarin-benzoic acid derivatives (4j ∼ r) were firstly synthesized. Due to the solubilities of some coumarin derivatives at enzyme activity test system were not good enough, ∼ 64 μM. For better comparing the inhibition activity of coumarin derivatives and future structure–activity relationship analysis, we firstly tested the anti-tyrosinase activity of compounds at concentration of 64 μM using kojic acid as positive control. As depicted in Table 1, among 18 derivatives, compound 4i with 4-hydroxyl group in benzene ring of cinnamic acid exerted strongest tyrosinase inhibition (61.24 ± 0.17%) at concentration of 64 μM, while other derivatives showed weaker inhibition against tyrosinase activity (<50%). Interesting, compound 4i (IC₅₀: 6.40 ± 0.39 μM) presented about ∼ 2.2 folds stronger tyrosinase inhibitory effect than kojic acid (14.13 ± 0.80 μM).

Table 1 The anti-tyrosinase activity of all synthesic derivatives 4a ∼ r and 5a ∼ l.

Compound	R1	R2	R3	Inhibition (%)a	IC50 (μM)
4a	H	H		7.48 ± 0.12	–
4b	H	H		44.87 ± 0.17	–
4c	H	H		37.94 ± 0.35	–
4d	H	H		40.32 ± 0.10	–
4e	H	H		30.96 ± 0.14	–
4f	H	H		37.29 ± 0.28	–
4g	H	H		9.62 ± 0.32	–
4h	H	H		27.52 ± 0.30	–
4i	H	H		61.24 ± 0.17	6.40 ± 0.39
4j	H	H		2.45 ± 0.32	–
4k	H	H		9.26 ± 0.21	–
4l	H	H		20.71 ± 0.12	–
4m	H	H		5.17 ± 0.24	–
4n	H	H		7.70 ± 0.14	–
4o	H	H		5.75 ± 0.10	–
4p	H	H		6.08 ± 0.10	–
4q	H	H		5.14 ± 0.20	–
4r	H	H		3.90 ± 0.10	–
5a	Cl	H		26.02 ± 0.30	–
5b	Br	H		35.44 ± 0.21	–
5c	F	H		56.57 ± 0.23	22.91 ± 0.14
5d	CH3	H		23.79 ± 0.23	–
5e	OCH3	H		60.20 ± 0.23	23.86 ± 0.15
5f	OH	H		62.39 ± 0.17	6.04 ± 0.30
5 g	H	Cl		9.43 ± 0.20	–
5 h	H	Br		48.55 ± 0.17	44.12 ± 0.36
5i	H	F		53.73 ± 0.30	17.00 ± 0.15
5j	H	CH3		58.27 ± 0.40	58.85 ± 0.19
5 k	H	OCH3		51.31 ± 0.20	52.32 ± 0.67
5l	H	OH		66.24 ± 0.20	3.04 ± 0.01
Kojic acid					14.13 ± 0.80

a Tyrosinase inhibition was assayed at 64 μM. “-” meaned not detected.

The main differences between derivatives 4a ∼ i and 4j ∼ r were the incorporation moiety: cinnamic acid and benzoic acid. Generally speaking, coumarin-cinnamic acid derivatives (4a ∼ i) were more active than coumarin-benzoic acid derivatives (4j ∼ r) with the same substitute group. This might due to the potential binding between the cinnamic acid of 4a ∼ i and tyrosinase. Furthermore, the substituents in cinnamic acid moiety also had effects on their anti-tyrosinase activity. Compared to compound 4a with no substituent in cinnamic acid, compounds 4b ∼ i showed stronger anti-tyrosinase activity, indicating that the substituent groups including CH₃, Cl, Br, F, OCH₃, NO₂, CF₃, and OH were benefit for their inhibitory activity.

Coumarin-cinnamic acid derivative 4i with strongest anti-tyrosinase activity was further structural modified by introduction of various substituents in the coumarin moiety, obtaining derivatives 5a ∼ l. Their anti-tyrosinase activity results revealed that the introduction of substituent group (Cl, Br, F, CH₃, and OCH₃) in the C-6 and C-7 position of coumarin caused a negative effect, however the introduction of hydroxyl would increase their inhibitory activity. Among them, compound 5l with OH at C-6 position of coumarin showed the best anti-tyrosinase activity (IC₅₀: 3.04 ± 0.01 μM), ∼ 4.6 folds higher than kojic acid (14.13 ± 0.80 μM).

2.3 Inhibitory mechanism and kinetic type assay

Compound 5l with the strongest anti-tyrosinase activity was selected as representative molecule to investigate the inhibitory mechanism and kinetic type. The mechanism assay by the plots of ν versus [E] in the presence of compound 5l gave a family of straight lines passing the origin (Fig. 2a), revealing that compound 5l was a reversible inhibitor. The kinetic type was detected by Lineweaver-Burk plots of ν versus 1/[S] in the presence of compound 5l, that obtained a group of data lines intersecting at × axis (Fig. 2a). The results showed that compound 5l showed a noncompetitive inhibition. Moreover, thire slop increasesed with the increasing compound 5l concentration, and the inhibition constant (K_i) was calculated as 1.5 μM.

Close

Fig. 2

(a) The plots of ν versus [E] in the presence of compound 5l; (b) Lineweaver-Burk plots of ν versus 1/[S] in the presence of compound 5l.

Export to PPT

2.4 3D fluorescence spectra and CD spectra assay

The 3D fluorescence spectra and CD spectra were operated to monitor the tyrosinase conformational changes interacting with compound 5l. The 3D fluorescence spectra of tyrosinase (Fig. 3a) appeared two typical spectra: Tyr with Trp residues at excitation of 280 nm and emission of 332 nm (Peal 1), and polypeptide backbone structure at excitation of 232 nm and emission of 335 nm (Peal 2). Treatment with compound 5l (3.33 μM) reduced the fluorescence intensity of Peak 1 by 17.87%, and Peak 2 by 28.19% (Fig. 3b), respectively, declaring that the binding of compound 5l with tyrosinase resulted in the changes of microenvironments and polypeptide backbone structure of tyrosinase.

Close

Fig. 3

(a) The 3D fluorescence spectra of tyrosinase. (b) The 3D fluorescence spectra of tyrosinase with compound 5l.

Export to PPT

The CD spectra of tyrosinase with compound 5l were also monitored at wavelength of 190 ∼ 250 nm and shown in Fig. 4. The CD spectra had two negative bands at 210 and 222 nm, those were the typical features of α-helixes due to n → π* and π → π* electron transfer of amide. Treatment of compound 5l caused the visible reduce of negative bands intensity, revealing that binding of compound 5l with tyrosinase generated the conformation change. The secondary structure contents of tyrosinase in the presence of compound 5l were calculated and listed in Table 2. Treatment of compound 5l (molar ratios: 2:1) caused increase in the content of the α-helix (from 7.3 to 9.6%), and β-turn (from 19.8 to 20.6%), while decrease in β-Sheet (from 38.7 to 29.9%) and random coil (from 36.1 to 35.5%).

Close

Fig. 4

The CD spectra of tyrosinase and compound 5l complex. The molar ratios of tyrosinase to 5l were 1: 0, 1: 1, and 1: 2 for curves a-c, respectively.

Export to PPT

2.5 Molecular docking

To obtain more binding mode of compound 5l within the active site of tyrosinase, molecular docking was performed using SYBYL software. As illustrated in Fig. 5(a), compound 5l was well stabilized into the active site of tyrosinase with a “U shaped” conformation, and 4-hydroxy cinnamic acid moiety nested inside. The detailed interaction results (Fig. 5(b)) showed that 6-hydroxy in coumarin formed a hydrogen bond with His85 (bond length: 1.9 Å), 4-hydroxy in cinnamic acid also formed a hydrogen bond with His259 (bond length: 2.7 Å). Moreover, 4-hydroxy in cinnamic acid was oriented toward two Cu ions at distances of 1.9 and 2.8 Å, respectively. All interactions helped compound 5l to anchor into the active site of tyrosinase.

Close

Fig. 5

(a) Compound 5l is positioned into the active pocket. (b) Detailed interaction of compound 5l with tyrosinase.

Export to PPT

2.6 Melanogenesis and tyrosinase activity assay in B16F10 cells

The cytotoxicity of representative compound 5l on B16F10 cells was firstly assayed after treatment of compound 5l for 48 h (Fig. 6). The results verified that compound 5l had no toxicity effect on B16F10 cells up to concentration 32 μM, while compound 5l showed obvious cytotoxicity at 64 μM. On the other hand, kojic acid (8–64 μM) showed no cytotoxicity. Thererfore, the effects of compound 5l on melanogenesis and tyrosinase activity in vitro were performed at concentrations (8–32 μM) with no cytotoxicity on B16F10 cells.

Close

Fig. 6

The cytotoxicity assay of compound 5l and kojic acid in B16F10 cells.

Export to PPT

The inhibitory effects of compound 5l on melanogenesis were ascertained by measuring the melanin content of B16F10 cells that were treated with compound 5l (8–32 μM) for 48 h. As can be seen in Fig. 7, compound 5l (8–32 μM) presented effectively melanin inhibition in a dose-dependent manner, showing (9.69 ± 1.69)%, (16.41 ± 1.15)%, and (20.96 ± 1.62)%, respectively. Moreover, compound 5l existed stronger melanin inhibitory activity than kojic acid at same tested concentrations.

Close

Fig. 7

The melanin inhibition of compound 5l and kojic acid in B16F10 cells.

Export to PPT

The cellular tyrosinase inhibitory effects of compound 5l were also determined in B16F10 cells and the results were listed in Fig. 8. Compound 5l (8–32 μM) exerted markedly inhibitory activity on cellular tyrosinase in a dose-dependent manner, with inhibition rate of (17.94 ± 2.06)%, (21.77 ± 1.17)%, and (32.30 ± 1.47)%, respectively, which were higher than those of kojic acid, respectively. Those results established that compound 5l suppressed melanogenesis by the inhibition of tyrosinase.

Close

Fig. 8

The cellular tyrosinase inhibitory of compound 5l and kojic acid in B16F10 cells.

Export to PPT

2.7 Melanogenesis and tyrosinase activity assay in zebrafish

To address whether the depigment effects of compound 5l also occurs in vivo, we employed zebrafish modle that had been believed as the typical in vivo model for the evaluation of tyrosinase activity and melanin. The acute toxicity of compound 5l was firstly evaluated. Fish embryos were exposed to compound 5l for 48 h to observe the embryo survival. The survival rate of compound 5l and kojic acid ranged from 85% to 95% (Fig. 9), revealing that compound 5l and kojic acid did not show acute toxicity at the tested concentrations (10–160 μM).

Close

Fig. 9

The acute toxicity assay of compound 5l and kojic acid in zebrafish.

Export to PPT

Then, the effects of compound 5l (10–40 μM) on melanogenesis in zebrafish were assayed by analyzing the grayscale value of the photo of embryo (Fig. 10). The results indicated that fish embryos in control group presented obvious pigmentation, especially in the eyes, yolk sac and notum. However, treatment with compound 5l (10–40 μM) could effectively reduce the pigmentation in a dose-dependent manner to (30 ± 5)%, while kojic acid showed (23 ± 3)% at 40 μM. Additionally, compound 5l (10–20 μM) exhibited better depigmenting effects, relative to kojic acid.

Close

Fig. 10

The melanin inhibition of compound 5l and kojic acid in zebrafish.

Export to PPT

The effects of compound 5l (10–40 μM) on tyrosinase activity in zebrafish were finally detected and the results were shown in Fig. 11. Compound 5l effectively inhibited tyrosinase activity in zebrafish in a dose-dependent manner to (57.03 ± 2.19)% at 40 μM, which was slightly lower than that of kojic acid (63.09 ± 3.82)% at 40 μM. Additionally, compound 5l (10–20 μM) exhibited better anti-tyrosinase activity than kojic acid at the same concentrations. This linear reduction of melanogenesis and tyrosinase activity in zebrafish embryos matched that obtained in the B16 cell, revealing the positive in vitro-in vivo correlation.

Close

Fig. 11

The tyrosinase inhibitory of compound 5l and kojic acid in zebrafish.

Export to PPT

4.1 Materials and methods

Tyrosinase from mushroom (EC 1.14.18.1) and 3,4-Dihydroxyphenylalanine (L-DOPA) were purchased from Sigma-Aldrich. All other reagents and solvents were commercial available. ¹H NMR and ¹³C NMR spectra were recorded in CDCl₃ on 500 MHz instruments. High-resolution mass spectral analysis (HRMS) data were measured on the Apex II by means of the ESI technique.

4.2 Synthesis of derivatives 4a ∼ r and 5a ∼ l

Substituted salicylaldehyde (1) (1.0 mmol) and ethyl acetoacetate (1.1 mmol) were added into 10 mL of ethanol containing piperidine (0.02 mmol), then reflowed at 80 °C until the reaction completed. Ice-cold water was added to quench the reaction, followed by filtration and recrystallization to obtain substituted acetyl coumarin (2). To the solution of 2 (1.0 mmol) in 10 mL of ethanol, hydroxylamine hydrochloride (3.0 mmol) and pyridine (0.04 mmol) were added and reacted at room temperature. The product was washed with ethanol to obtain substituted acetyl coumarin containing oxime (3). To the solution of 3 (1.0 mmol) in 3 mL of DCM, triethylamine (1.1 mmol) and substituted benzoic acid chlorides/cinnamic acid chlorides (1.0 mmol) were added, respectively, and reacted at room temperature. Then the reaction was quenched, extracted, washed, and dried, followed by column chromatography purification to produce derivatives 4a ∼ r and 5a ∼ l. All NMR, HRMS, yield, and m.p. data were summarized in supplementary material.

4.3 Tyrosinase activity assay

The synthetic derivatives were assayed for anti-tyrosinase activity using mushroom tyrosinase and L-DOPA (He et al., 2021; Song et al., 2017). All derivatives were dissolved in DMSO and diluted to various concentrations. 10 μL of mushroom tyrosinase (1333.4 U/mL) and 10 μL of test compound were added into 130 μL of phosphate buffer (50 M, pH 6.8), respectively, followed by the incubation for 10 min at room temperature. Then 50 μL of L-DOPA solution (2 mM) was added and the absorbance change was determined at 475 nm. The inhibition percentage was calculated respect to blank group and the 50% inhibitory concentration (IC₅₀) was calculated by interpolation of the dose–response curves. Kojic acid was used as the positive standard. Each experiment was done four times in parallel.

The enzyme inhibitory kinetics of compound 5l were determined by the plots of enzymatic reaction vs enzyme concentration in the presence of compound 5l (Hu et al., 2016), and the substrate inhibitory kinetics were obtained by the Lineweaver-Burk plot of enzymatic reaction vs substrate concentration in the presence of compound 5l (Zhao et al., 2019). The final L-DOPA concentrations were 0.5, 1, 0.67, and 2, respectively.

4.4 3D fluorescence spectra assay

10 μL of mushroom tyrosinase (Final concentration: 168 μM) and 10 μL of compound 5l (Final concentration: 3.33 μM) were added into 2.98 mL of phosphate buffer (50 M, pH 6.8) and incubated for 5 min (Song et al., 2020). The 3D fluorescence spectra of this mixture were measured at excitation and emission wavelengths of 200–600 nm.

4.5 CD spectroscopy

10 μL of mushroom tyrosinase (Final concentration: 33 μM) and 10 μL of compound 5l were added into 180 μL of phosphate buffer (50 M, pH 6.8) and incubated for 5 min (Song et al., 2021). The CD spectrum of this mixture was recorded. The proportion of secondary conformation of protein was analyzed using CDNN.

4.6 Molecular docking

Molecular docking was performed by SYBYL software using the crystal structure of tyrosinase from Agaricus bisporus (PDBID: 2Y9X) obtained from the RCSB Protein Data Bank (Peng et al., 2021). The tyrosinase protein was prepared by the remove of ligand and water, followed by the addition of hydrogen atoms and charge, repair of end residues, and the generation of active pocket. Compound 5l was prepared by addition of charge with Gasteiger-Hückle and energy minimization. Then, the docking between Compound 5l and tyrosinase was operated in the default format, and the results were visualized by Pymol software.

4.7 Cell culture

B16F10 cells (ATCC, Manassas, VA, USA) were cultured in DMEM supplemented with 10% FBS and 1% penicillin/streptomycin at 37 °C in a humidified atmosphere with 5% CO₂.

4.8 Cytotoxicity assay

The cells viability after exposure to compound was determined by MTT method as described previously (Lee et al., 2019). B16F10 cells were seeded in a 96 well plate (5 × 10³ cells/well) for 24 h and treated with compound 25 (10–40 μM) for another 24 h. Then cells were labelled with MTT solution (0.5 mg/mL) for 4 h. The violet formazan precipitates were dissolved in 100 μL DMSO, then the absorbance was measured at 570 nm. Kojic acid was used as a positive control.

4.9 Melanin assay in B16F10 cells

The effect of compound 5l on melanin content was studied in Murine B16F10 melanoma cells using the standard method with slight modification (Ullah et al., 2019). B16F10 cells were seeded in a 6 well plate (10⁵ cells/well) for 24 h, and then treated with compound 5l (8 ∼ 32 μM) for another 48 h. Harvested cells were added 200 μL NaOH (1 M, containing 10% DMSO), heated at 100 °C for 1 h, and then the absorbance at 405 nm was detected. Kojic acid (8 ∼ 32 μM) was used as a positive control.

4.10 Tyrosinase activity in B16F10 cells

The cellular tyrosinase inhibition of compound 5l was performed as previously described with slight modification (Tang et al., 2021). B16F10 cells were seeded in a 6 well plate (10⁵ cells/well) for 24 h, and then treated with compound 5l (8–32 μM) for 48 h. Harvested cells were washed with ice-cold PBS two times, added 300 μL Triton X-100 (1%, 50 mM PBS), and frozen at −80 °C for 1 h. Then cellular extracts were thawed and centrifuged (12000 rpm, 4 °C) for 30 min to obtain the supernatant, followed by the protein concentration determination using BCAKit. To 80 μL protein solution (equal amount of protein), 20 μL of L-DOPA solution (10 mM) was added, and then the absorbance change was determined at 475 nm. Kojic acid (8–32 μM) was used as a positive control.

4.11 Zebrafish maintenance

Sexual maturity, health, and no deformity Wild-type (AB) zebrafish were cultured at 28 °C under a 14/10‐hr light/dark cycle with the pH around 7.4.

4.12 Fish embryo acute toxicity test

For the in vivo toxicity test (Ramlan et al., 2017); 6 ∼ 8 hpf embryos were seeded in 24 well plate with M3 culture medium (1 embryo/well), treated with test compound (0 ∼ 160 μM) and Kojic acid (0 ∼ 160 μM), respectively, then observed the survival at 72 hpf. Kojic acid (0 ∼ 160 μM) was used as a positive control. Each condition had 12 embryos and was repeated 5 times.

4.13 Melanin determination in zebrafish

Phenotype-based melanogenesis inhibition of compound 5l in zebrafish was determined following reported method (Abbas et al., 2017). Embryos were seeded in 6 well plate with M3 culture medium (5 embryo/well) at 24 hpf, treated with test compound (0 ∼ 40 μM) and Kojic acid (0 ∼ 40 μM), respectively. Then the photo of embryo was taken to determine the melanin using image J software. The embryo photo was prepared with built-in program of image J software, including convering picture to grayscale, framing embryo photo, closing resulting curve, reading grayscale. The grayscale of sample embryo photo was expressed as relative melanin level of the control (100%).

4.14 Tyrosinase activity in zebrafish

Tyrosinase activity was measured according to the method by Chen et al. (Chen et al., 2019). The treated embryos was were lysed in 150 μL sodium deoxycholate solution (5.0 mg/mL), and centrifuged (10000 rpm, 4 °C) for 5 min to obtain the supernatant. 100 μL supernatant was mixed with 100 μL L-DOPA solution (5 mg/mL), followed by the absorbance change record at 475 nm.

4.15 Statistical analysis

All data were presented as mean ± SD. One-way ANOVA was performed to evaluate the difference between groups. P < 0.05 was considered significant.