Palladium(II)-Catalyzed Efficient Synthesis of Wedelolactone and Evaluation as Potential Tyrosinase Inhibitor

Tyrosinase is an enzyme widely distributed in nature, which has multiple functions, especially in the melanin biosynthesis pathway. Despite the few clinically available tyrosinase inhibitors for whitening, a great demand remains for novel compounds with low side effects in terms of potential carcinogenicity and improved clinical efficacy. A natural product, wedelolactone (WEL), with a polyhydroxyl moiety, attracted our attention as a potential tyrosinase inhibitor. Before we studied the biological activity of the natural product, a synthetic methodological research was firstly carried to obtain enough raw material. WEL could be obtained efficiently through palladium-catalyzed boronation/coupling reactions and 2,3-dicyano-5,6-dichlorobenzoquinone (DDQ)-involved oxidative deprotection/annulation reactions. Immediately after, the natural product was proven to be an efficient tyrosinase inhibitor. In conclusion, we developed a mild and efficient approach for the preparation of WEL, and the natural product was disclosed to have anti-tyrosinase activity, which could be widely used in multiple fields.


Introduction
Tyrosinase (EC 1.14.18.1) is an enzyme widely distributed in nature, which has multiple functions, especially in the melanin biosynthesis pathway [1].The enzyme can catalyze l-tyrosine to l-3,4-dihydroxyphenylalanine (l-DOPA), with further oxidation to dopaquinone [2].Dopaquinone transforms from brown to black through several reactions.Abnormal melanin production, including melasma, freckles, lentigo, senilis, and other forms of melanin hyperpigmentation could be a serious aesthetic problem [2,3].In addition, tyrosinase is also involved in the defensive and developmental functions of pests [4].Excessive dopaquinone was also reported to cause neurodegeneration related to Parkinson's disease [5].Thus, many tyrosinase inhibitors are applied in cosmetics and pharmaceutical products [6].Despite the few clinically available tyrosinase inhibitors for whitening, a great demand remains for novel compounds with low side effects in terms of potential carcinogenicity and improved clinical efficacy [7,8].Obviously, more efforts are still needed in that direction; therefore, we recently focused our interest on discovering novel tyrosinase inhibitors.
Compounds with the polyhydroxyl moiety were proven to be potential tyrosinase inhibitors [9][10][11].Recently, our group undertook research on a natural product with a polyhydroxyl moiety named wedelolactone (WEL), which is derived from the medical plant Eclipta prostrata [12].Although a wide range of pharmacological activities of WEL were reported, there is less information on the inhibitory effect and reversibility of WEL on tyrosinase.Thus, the inhibitory activity and mechanism of WEL toward tyrosinase deserves deeper investigation; however, but the present knowledge on synthesis of the natural product is limited.Although several groups invested substantial effort in the preparation of WEL, these methods had several disadvantages, including a time-consuming nature with complicated synthetic approaches [13][14][15].
Among these methods, two routes listed in Figure 1 are commonly recognized by the industry.However, both methods have several disadvantages.The first method (reported by Yang [14]) involves a crucial intermediate, phenyl acetylene, which is difficult to prepare.The route has a low 15% overall yield with a long linear sequence (total of 12 steps), and it is rarely applied to access a variety of WEL analogues for structure transformations.The second method (reported by Lee et al. [13]) employs toxic organotin and organomercurial reagents, which limit industrial production and increase operation complexity.In addition, both methods can only obtain the natural products on a small scale.As the present methods are imperfect and unsatisfactory for further investigation of WEL as an efficient tyrosinase inhibitor, the development of a facile, versatile, and mild approach is urgently needed.

Palladium(II)-Catalyzed Efficient Synthesis of WEL
Retrosynthetically, WEL could be logically disconnected by the ring opening of furan to afford the intermediate 4, which is further disconnected by C-C bond cleavage to trace back to the intermediate 3-bromo-5-benzyloxy-7-acetoxyl-2-chromenone 3 and the readily prepared 4,5dibenzyloxy-2-(4-methoxybenzyl)oxy-phenyl boronic ester 2 (Scheme 1).This similar synthetic strategy was ever used by Shen for the synthesis of hirtellanine A [16].Synthetically, we expected that polysubstituted coumarin 4 could be obtained by Pd(II)-catalyzed Suzuki-Miyaura coupling of 3-bromocoumarin 3 and polysubstituted phenyl boronate ester 2 which could be generated by a Pd(II)-catalyzed boronation reaction of the polysubstituted bromobenzene 1.The coupling product 4 then underwent a DDQ-oxidation deprotection/annulation reaction to deliver the final product WEL 5.

Palladium(II)-Catalyzed Efficient Synthesis of WEL
Retrosynthetically, WEL could be logically disconnected by the ring opening of furan to afford the intermediate 4, which is further disconnected by C-C bond cleavage to trace back to the intermediate 3-bromo-5-benzyloxy-7-acetoxyl-2-chromenone 3 and the readily prepared 4,5-dibenzyloxy-2-(4-methoxybenzyl)oxy-phenyl boronic ester 2 (Scheme 1).This similar synthetic strategy was ever used by Shen for the synthesis of hirtellanine A [16].Synthetically, we expected that polysubstituted coumarin 4 could be obtained by Pd(II)-catalyzed Suzuki-Miyaura coupling of 3-bromocoumarin 3 and polysubstituted phenyl boronate ester 2 which could be generated by a Pd(II)-catalyzed boronation reaction of the polysubstituted bromobenzene 1.The coupling product 4 then underwent a DDQ-oxidation deprotection/annulation reaction to deliver the final product WEL 5.
3,4-dihydroxybenzaldehyde 6 as the starting material to provide the polysubstituted bromobenzene 1 via the mCPBA-mediated Baeyer-Villiger oxidation strategy, which resolved the selective protection of phenolic groups.Synthetically, protection of the phenol groups of 3,4-dihydroxybenzaldehyde 6 with benzyl bromide afforded 3,4-bis(benzyloxy)benzaldehyde 7 in 86% yield, which subsequently underwent mCPBA-mediated oxidation and hydrolysis to deliver 3,4bis(benzyloxy)phenol 8 in 87% yield.Next, 8 was firstly protected with PMBCl and then selectively brominated with NBS to obtain polysubstituted bromobenzene 1 in high yield.Subsequently, we aimed to synthesize 3-bromo-5-benzyloxy-7-acetoxyl-2-chromenone 3 (Scheme 2).According to reported methods [14], the commercially available phloroglucinol 10 and ethyl propiolate as starting materials smoothly underwent ZnCl2-catalyzed esterification and cyclization reactions to provide dihydroxycoumarin 11.Dihydroxycoumarin 11 was then treated with acetyl chloride to deliver the corresponding diacetoxylcoumarin 12 in good yield.The following reaction involved the dibromination and dehydrobromination of ketene 12 to afford 3-bromo-5,7diacetoxyl-2-chromenone 13 in 75% yield.Reaction temperature is vital to the bromination reaction.It was found that maintaining the reaction temperature at 10 °C could avoid the bromination of acetyl group of 12 and acquire the optimal reaction yield.Finally, we ran the partial benzylation of 3-bromo-5,7-diacetoxyl-2-chromenone 13.The similar reactivity of the 5-and 7-hydroxy groups did not facilitate the selective 5-benzylation of 13.According to Humbert's selective alkylation reaction conditions [18], we tried similar normal benzylation conditions (BnBr, K2CO3, acetone, reflux) and found that traces of water in acetone chemoselectively hydrolyzed the 5-acetyl group as well, generating the intermediates 14 and 15, stabilized by the conjugating effect of the pyrone ring.Although the undesired deacetylation byproducts still existed, the major benzylating product was 3bromo-5-benzyloxy-7-acetoxyl-2-chromenone 3.With the use of ethanol recrystallization, the crude product 3 was obtained in 65% yield and directly used without further purification for the next step.With the key intermediates polysubstituted bromobenzene 1 and 3-bromo-5-benzyloxy-7acetoxyl-2-chromenone 3 in hand, we focused on the combination of two fragments via boronation Scheme 1. Synthesis of polysubstituted bromobenzene 1.
In the beginning of our synthesis, we focused on the generation of the polysubstituted bromobenzene 1 (Scheme 1).Selective protection of the three phenolic hydroxyl groups presented a big synthetic challenge.After reviewing the literature [16,17], we chose the commercially available 3,4-dihydroxybenzaldehyde 6 as the starting material to provide the polysubstituted bromobenzene 1 via the mCPBA-mediated Baeyer-Villiger oxidation strategy, which resolved the selective protection of phenolic groups.Synthetically, protection of the phenol groups of 3,4-dihydroxy-benzaldehyde 6 with benzyl bromide afforded 3,4-bis(benzyloxy)benzaldehyde 7 in 86% yield, which subsequently underwent mCPBA-mediated oxidation and hydrolysis to deliver 3,4-bis(benzyloxy)phenol 8 in 87% yield.Next, 8 was firstly protected with PMBCl and then selectively brominated with NBS to obtain polysubstituted bromobenzene 1 in high yield.
Subsequently, we aimed to synthesize 3-bromo-5-benzyloxy-7-acetoxyl-2-chromenone 3 (Scheme 2).According to reported methods [14], the commercially available phloroglucinol 10 and ethyl propiolate as starting materials smoothly underwent ZnCl 2 -catalyzed esterification and cyclization reactions to provide dihydroxycoumarin 11.Dihydroxycoumarin 11 was then treated with acetyl chloride to deliver the corresponding diacetoxylcoumarin 12 in good yield.The following reaction involved the dibromination and dehydrobromination of ketene 12 to afford 3-bromo-5,7-diacetoxyl-2-chromenone 13 in 75% yield.Reaction temperature is vital to the bromination reaction.It was found that maintaining the reaction temperature at 10 • C could avoid the bromination of acetyl group of 12 and acquire the optimal reaction yield.Finally, we ran the partial benzylation of 3-bromo-5,7-diacetoxyl-2-chromenone 13.The similar reactivity of the 5-and 7-hydroxy groups did not facilitate the selective 5-benzylation of 13.According to Humbert's selective alkylation reaction conditions [18], we tried similar normal benzylation conditions (BnBr, K 2 CO 3 , acetone, reflux) and found that traces of water in acetone chemoselectively hydrolyzed the 5-acetyl group as well, generating the intermediates 14 and 15, stabilized by the conjugating effect of the pyrone ring.Although the undesired deacetylation byproducts still existed, the major benzylating product was 3-bromo-5-benzyloxy-7-acetoxyl-2-chromenone 3.With the use of ethanol recrystallization, the crude product 3 was obtained in 65% yield and directly used without further purification for the next step.
Molecules 2019, 24, x FOR PEER REVIEW 3 of 11 In the beginning of our synthesis, we focused on the generation of the polysubstituted bromobenzene 1 (Scheme 1).Selective protection of the three phenolic hydroxyl groups presented a big synthetic challenge.After reviewing the literature [16,17], we chose the commercially available 3,4-dihydroxybenzaldehyde 6 as the starting material to provide the polysubstituted bromobenzene 1 via the mCPBA-mediated Baeyer-Villiger oxidation strategy, which resolved the selective protection of phenolic groups.Synthetically, protection of the phenol groups of 3,4-dihydroxybenzaldehyde 6 with benzyl bromide afforded 3,4-bis(benzyloxy)benzaldehyde 7 in 86% yield, which subsequently underwent mCPBA-mediated oxidation and hydrolysis to deliver 3,4bis(benzyloxy)phenol 8 in 87% yield.Next, 8 was firstly protected with PMBCl and then selectively brominated with NBS to obtain polysubstituted bromobenzene 1 in high yield.Subsequently, we aimed to synthesize 3-bromo-5-benzyloxy-7-acetoxyl-2-chromenone 3 (Scheme 2).According to reported methods [14], the commercially available phloroglucinol 10 and ethyl propiolate as starting materials smoothly underwent ZnCl2-catalyzed esterification and cyclization reactions to provide dihydroxycoumarin 11.Dihydroxycoumarin 11 was then treated with acetyl chloride to deliver the corresponding diacetoxylcoumarin 12 in good yield.The following reaction involved the dibromination and dehydrobromination of ketene 12 to afford 3-bromo-5,7diacetoxyl-2-chromenone 13 in 75% yield.Reaction temperature is vital to the bromination reaction.It was found that maintaining the reaction temperature at 10 °C could avoid the bromination of acetyl group of 12 and acquire the optimal reaction yield.Finally, we ran the partial benzylation of 3-bromo-5,7-diacetoxyl-2-chromenone 13.The similar reactivity of the 5-and 7-hydroxy groups did not facilitate the selective 5-benzylation of 13.According to Humbert's selective alkylation reaction conditions [18], we tried similar normal benzylation conditions (BnBr, K2CO3, acetone, reflux) and found that traces of water in acetone chemoselectively hydrolyzed the 5-acetyl group as well, generating the intermediates 14 and 15, stabilized by the conjugating effect of the pyrone ring.Although the undesired deacetylation byproducts still existed, the major benzylating product was 3bromo-5-benzyloxy-7-acetoxyl-2-chromenone 3.With the use of ethanol recrystallization, the crude product 3 was obtained in 65% yield and directly used without further purification for the next step.With the key intermediates polysubstituted bromobenzene 1 and 3-bromo-5-benzyloxy-7acetoxyl-2-chromenone 3 in hand, we focused on the combination of two fragments via boronation

With
the key intermediates polysubstituted bromobenzene 1 and 3-bromo-5-benzyloxy-7-acetoxyl-2-chromenone 3 in hand, we focused on the combination of two fragments via boronation and a subsequent coupling reaction (Scheme 3).Under strong alkaline and low-temperature conditions [19], polysubstituted bromobenzene 1 could be treated with bis(pinacolato)diboron to afford the borate ester 2. To simplify the operation procedure, Pd(II)-catalyzed boronation was attempted, and the corresponding borate ester 2 was obtained in quantitative yield as well.The formed crude product 2 could be utilized directly without further purification for the following Suzuki-Miyaura reaction and afforded the deacetyl coupling product 16 in good yield.To avoid the stability issue of borate ester in column chromatography, the boronation and subsequent coupling reaction were carried out conveniently in one pot and gave a 72% overall yield.The one-pot reaction conditions were optimized by initiating an experiment to evaluate the effect of the various parameters on the reaction yield (see Supplementary Materials).Methylation of the deacetyl coupling product 16 in the standard treatment with methyl iodide gave polysubstituted coumarin 4 in 85% yield.
Molecules 2019, 24, x FOR PEER REVIEW 4 of 11 and a subsequent coupling reaction (Scheme 3).Under strong alkaline and low-temperature conditions [19], polysubstituted bromobenzene 1 could be treated with bis(pinacolato)diboron to afford the borate ester 2. To simplify the operation procedure, Pd(II)-catalyzed boronation was attempted, and the corresponding borate ester 2 was obtained in quantitative yield as well.The formed crude product 2 could be utilized directly without further purification for the following Suzuki-Miyaura reaction and afforded the deacetyl coupling product 16 in good yield.To avoid the stability issue of borate ester in column chromatography, the boronation and subsequent coupling reaction were carried out conveniently in one pot and gave a 72% overall yield.The one-pot reaction conditions were optimized by initiating an experiment to evaluate the effect of the various parameters on the reaction yield (see Supplementary Materials).Methylation of the deacetyl coupling product 16 in the standard treatment with methyl iodide gave polysubstituted coumarin 4 in 85% yield.Finally, the completion of the remaining steps in WEL synthesis required PMB deprotection, cyclization, and debenzylation.Removing the PMB on the substituted coumarin 4 in acidic conditions (HOAc, reflux) led to the decomposed mixture and gave an unsatisfactory result [20].It was reported that DDQ could also be used as a deprotecting reagent to remove the PMB group [21].Thus, we adopted this synthetic strategy and hoped that oxidation cyclization could subsequently happen after the deprotection.Fortunately, DDQ-involved deprotection and oxidative annulation proceeded well as anticipated, forming the desired product 1,8,9-tris(benzyloxy)-3-methoxy-6H-benzofuro [3,2c]chromen-6-one 17 in 56% yield (Scheme 4).Compound 17 was then treated with BCl3 as a debenzylating reagent, generating the final product wedelolactone 5 in 81% yield. 1 H-and 13 C-NMR spectra of the synthetic product were in agreement with reported data for WEL [13].

Biological Activity
With a sufficient amount of WEL in hand, we tested the in vitro tyrosinase activity.Kojic acid was used as a positive control, as typically employed in the evaluation of tyrosinase inhibitors.As shown in Figure 2, WEL caused strong tyrosinase inhibition in a concentration-dependent manner.The 50% inhibitory concentration (IC50) values of WEL and kojic acid were determined to be 1.2 ± 0.3 and 14.2 ± 1.6 μM, respectively.The strong tyrosinase inhibitory activity may be due to the multiple hydroxyl groups in the structure.However, this needs to be further confirmed by determining the Finally, the completion of the remaining steps in WEL synthesis required PMB deprotection, cyclization, and debenzylation.
Removing the PMB on the substituted coumarin 4 in acidic conditions (HOAc, reflux) led to the decomposed mixture and gave an unsatisfactory result [20].It was reported that DDQ could also be used as a deprotecting reagent to remove the PMB group [21].Thus, we adopted this synthetic strategy and hoped that oxidation cyclization could subsequently happen after the deprotection.Fortunately, DDQ-involved deprotection and oxidative annulation proceeded well as anticipated, forming the desired product 1,8,9-tris(benzyloxy)-3-methoxy-6H-benzofuro[3,2-c]chromen-6-one 17 in 56% yield (Scheme 4).Compound 17 was then treated with BCl 3 as a debenzylating reagent, generating the final product wedelolactone 5 in 81% yield. 1 H-and 13 C-NMR spectra of the synthetic product were in agreement with reported data for WEL [13].
Molecules 2019, 24, x FOR PEER REVIEW 4 of 11 and a subsequent coupling reaction (Scheme 3).Under strong alkaline and low-temperature conditions [19], polysubstituted bromobenzene 1 could be treated with bis(pinacolato)diboron to afford the borate ester 2. To simplify the operation procedure, Pd(II)-catalyzed boronation was attempted, and the corresponding borate ester 2 was obtained in quantitative yield as well.The formed crude product 2 could be utilized directly without further purification for the following Suzuki-Miyaura reaction and afforded the deacetyl coupling product 16 in good yield.To avoid the stability issue of borate ester in column chromatography, the boronation and subsequent coupling reaction were carried out conveniently in one pot and gave a 72% overall yield.The one-pot reaction conditions were optimized by initiating an experiment to evaluate the effect of the various parameters on the reaction yield (see Supplementary Materials).Methylation of the deacetyl coupling product 16 in the standard treatment with methyl iodide gave polysubstituted coumarin 4 in 85% yield.Finally, the completion of the remaining steps in WEL synthesis required PMB deprotection, cyclization, and debenzylation.Removing the PMB on the substituted coumarin 4 in acidic conditions (HOAc, reflux) led to the decomposed mixture and gave an unsatisfactory result [20].It was reported that DDQ could also be used as a deprotecting reagent to remove the PMB group [21].Thus, we adopted this synthetic strategy and hoped that oxidation cyclization could subsequently happen after the deprotection.Fortunately, DDQ-involved deprotection and oxidative annulation proceeded well as anticipated, forming the desired product 1,8,9-tris(benzyloxy)-3-methoxy-6H-benzofuro [3,2c]chromen-6-one 17 in 56% yield (Scheme 4).Compound 17 was then treated with BCl3 as a debenzylating reagent, generating the final product wedelolactone 5 in 81% yield. 1 H-and 13 C-NMR spectra of the synthetic product were in agreement with reported data for WEL [13].

Biological Activity
With a sufficient amount of WEL in hand, we tested the in vitro tyrosinase activity.Kojic acid was used as a positive control, as typically employed in the evaluation of tyrosinase inhibitors.As shown in Figure 2, WEL caused strong tyrosinase inhibition in a concentration-dependent manner.The 50% inhibitory concentration (IC50) values of WEL and kojic acid were determined to be 1.2 ± 0.3 and 14.2 ± 1.6 μM, respectively.The strong tyrosinase inhibitory activity may be due to the multiple

Biological Activity
With a sufficient amount of WEL in hand, we tested the in vitro tyrosinase activity.Kojic acid was used as a positive control, as typically employed in the evaluation of tyrosinase inhibitors.As shown in Figure 2, WEL caused strong tyrosinase inhibition in a concentration-dependent manner.The 50% inhibitory concentration (IC 50 ) values of WEL and kojic acid were determined to be 1.2 ± 0.3 and 14.2 ± 1.6 µM, respectively.The strong tyrosinase inhibitory activity may be due to the multiple hydroxyl groups in the structure.However, this needs to be further confirmed by determining the derivatives of WEL (data not shown).To confirm the inhibitory mechanism of WEL against tyrosinase, the plots of initial velocity versus tyrosinase concentration at different concentrations of WEL were developed, and a set of straight lines was obtained (as shown in Figure 2).All of the lines passed through the origin, and an increase in the WEL concentration reduced the slopes of the lines, indicating that the compound was a reversible inhibitor.
Molecules 2019, 24, x FOR PEER REVIEW 5 of 11 derivatives of WEL (data not shown).To confirm the inhibitory mechanism of WEL against tyrosinase, the plots of initial velocity versus tyrosinase concentration at different concentrations of WEL were developed, and a set of straight lines was obtained (as shown in Figure 2).All of the lines passed through the origin, and an increase in the WEL concentration reduced the slopes of the lines, indicating that the compound was a reversible inhibitor.

The Docking Studies
Computational docking studies were employed to determine the preferred binding sites of WEL in tyrosinase using the GOLD5.1 software.Tyrosinase contains two copper ions, and each copper ion is coordinated by three histidine residues.The first copper (Cu A) is coordinated by His61, His85, and His94, and the ligands of the second copper ion (Cu B) are His163, and His 296 [21].WEL could insert into the active site with a copper domain and it was found to interact with various aminoacid residues (Figure 3).The possible site of hydrogen-bonding interactions of WEL with tyrosinase was Asn260.The hydrogen-bonding residues could affect the binding affinity considerably.According to the molecular docking study, it was found that the His61, Val248, His259, Asn260, and His263 amino-acid residues of tyrosinase interact with WEL.The enzyme kinetic analysis and molecular docking studies confirmed that WEL binds to tyrosinase in the active site.

The Docking Studies
Computational docking studies were employed to determine the preferred binding sites of WEL in tyrosinase using the GOLD5.1 software.Tyrosinase contains two copper ions, and each copper ion is coordinated by three histidine residues.The first copper (Cu A) is coordinated by His61, His85, and His94, and the ligands of the second copper ion (Cu B) are His259, His163, and His 296 [21].WEL could insert into the active site with a copper domain and it was found to interact with various amino-acid residues (Figure 3).The possible site of hydrogen-bonding interactions of WEL with tyrosinase was Asn260.The hydrogen-bonding residues could affect the binding affinity considerably.According to the molecular docking study, it was found that the His61, Val248, His259, Asn260, and His263 amino-acid residues of tyrosinase interact with WEL.The enzyme kinetic analysis and molecular docking studies confirmed that WEL binds to tyrosinase in the active site.
could insert into the active site with a copper domain and it was found to interact with various aminoacid residues (Figure 3).The possible site of hydrogen-bonding interactions of WEL with tyrosinase was Asn260.The hydrogen-bonding residues could affect the binding affinity considerably.According to the molecular docking study, it was found that the His61, Val248, His259, Asn260, and His263 amino-acid residues of tyrosinase interact with WEL.The enzyme kinetic analysis and molecular docking studies confirmed that WEL binds to tyrosinase in the active site.

Biological Assays
The tyrosinase inhibition activity of the compounds was measured using l-DOPA as a substrate according to a modified method of previous work [4,10,11].WEL was firstly dissolved in DMSO at a concentration of 1.0 mM, and the final concentration of DMSO in the reaction mixture was 3%.In the investigation, the total volume of the reaction system was 300 µL; l-DOPA was used as the substrate for the determination of diphenolase activity.Briefly, l-DOPA (100 µL, 0.5 mM), phosphate buffer (180 µL, pH 6.8, 50 mM), and different concentrations of inhibitors (10 µL in DMSO)

Figure 2 .
Figure 2. (A) Inhibitory activity of WEL and the control Kojic acid on tyrosinase.(B) Inhibitory mechanism of WEL on tyrosinase.

Figure 2 .
Figure 2. (A) Inhibitory activity of WEL and the control Kojic acid on tyrosinase.(B) Inhibitory mechanism of WEL on tyrosinase.

Figure 3 .
Figure 3. Computational docking simulations between tyrosinase and WEL obtained using GOLD5.1.The cyan stick structures indicate WEL and the white stick structures represent the amino-acid

Figure 3 .
Figure 3. Computational docking simulations between tyrosinase and WEL obtained using GOLD5.1.The cyan stick structures indicate WEL and the white stick structures represent the amino-acid residues of tyrosinase.The copper ions are depicted as yellow balls.The black dotted lines represent hydrogen bonds.