Mechanistic studies of tyrosinase suicide inactivation

Tyrosinase is a copper containing enzyme that oxidises both phenols and catechols to ortho - quinones. The oxidation of catechols (oxidase activity) is associated with a gradual and irreversible inactivation of the enzyme. This suicide inactivation has been interpreted in terms of the enzyme occasionally oxidising catechols as phenols (oxygenase activity) leading to reductive elimination of copper. In this account experimental evidence supporting the authors’ mechanism is reviewed and discussed.


Introduction
Tyrosinase (EC 1.14.18.1) is a copper metalloenzyme that is widely distributed in nature and is involved in many biologically significant transformations. 1 The chemistry of tyrosinase includes a number of interesting features, including two discrete mechanisms of oxidation, a lag period during in vitro oxidation of tyrosine, and suicide inactivation of the active site during catechol oxidation.In this account we review experimental evidence for the mechanism of suicide inactivation of tyrosinase by catechols.
The active site of tyrosinase contains two copper atoms and the enzyme can occur in three forms (Scheme 1).Deoxy-tyrosinase 2 binds dioxygen to form oxy-tyrosinase 1.The third form, met-tyrosinase 3, cannot bind oxygen but is reduced to deoxy-tyrosinase 2 by catechols.

Scheme 1
In mammals tyrosinase is the principal enzyme of the pigmentation pathway giving rise to the melanin pigments (Scheme 2). 2 The natural substrate tyrosine 4 is oxidised by oxy-tyrosinase 1 to give dopaquinone 6 which either cyclises to cyclodopa 7 leading to eumelanin (black/brown pigment) or reacts with cysteine to give cysteinyldopa 5 leading to pheomelanin (red/yellow pigment).
ARKAT USA, Inc. Tyrosinase is unusual in that it exhibits two catalytic functions: (a) oxygenase activity ('cresolase' action) in which monohydric phenols have oxygen substituted at the 2-position giving an ortho-quinone (Scheme 1); 3,4 and (b) oxidase activity ('catecholase' action) 5 in which a catechol is oxidised to an ortho-quinone (Scheme 1).One of the characteristics of in vitro oxidation of tyrosine 4 is a 'lag period'. 6Native tyrosinase occurs mainly as met-tyrosinase 3, which cannot oxidise phenols, e.g.tyrosine 4, and needs to be reduced to deoxy-tyrosinase 2 by a catechol before phenol oxidation can begin (Scheme 1). 7In vitro this activating catecholic substrate is generated indirectly by fast redox exchange of dopaquinone 6 formed slowly by the small amount of oxy-tyrosine 1 present in native tyrosinase.The lag period ends when all the enzyme has been activated by this indirect and relatively slow non-enzymatic formation of dopa 8 (Scheme 2). 8,91][12][13][14][15][16] It was shown in 1982 by Dietler and Lerch 17 that this inactivation is linearly correlated with the loss of 50% of the copper atoms from the active site (Figure 1) but the mechanism to account for this loss of copper remained elusive.cently we proposed a mechanism that accounts for the copper loss during inactivation. 18riefly stated, this mechanism postulates that oxy-tyrosinase 1 sometimes binds catechol substrates in the oxygenase mode, i.e. as if they were phenols.Scheme 3(a) shows the intermediate 9 formed by binding of a phenol to oxy-tyrosinase 1 followed by elimination of an ortho-quinone 11 and generation of deoxy-tyrosinase 2. If a catechol occasionally binds in a similar way (10, Scheme 3(b)), the extra hydroxy substituent can deprotonate leading to reductive elimination of Cu(0), a 3-hydroxy-ortho-quinone 12 and an irreversibly inactivated form of tyrosinase 13.This processing of catechols as phenols in the oxygenase cycle nicely accounts for the observed suicide-inactivation mechanism and we have described further experimental studies that support this mechanism. 19,20Recognising the novelty and significance of our studies and proposals, another group has subsequently proposed a variation on the reductive-elimination theme to account for copper loss. 21This alternative suggestion is summarised in Scheme 4. In this account we review the experimental evidence regarding the reductive-elimination mechanism of suicide inactivation.
(a) (b) At the outset we wish to emphasize that, to our knowledge, there are no studies unequivocally demonstrating the formation of zero-valent copper during the suicide inactivation of tyrosinase.However, this seems to be the most plausible explanation for the loss of enzymatic activity (Figure 1), and while attempts have been made to re-activate tyrosinase by addition of copper salts, the interpretation of the results of these experiments has been hampered by significant non-enzymatic oxidation.
Another caveat is that the majority of the recent experimental data bearing on the question of suicide inactivation are derived from work with commercially available tyrosinase from mushrooms (Agaricus bisporus).This limits the extent to which the conclusions drawn from them can be generalized, but it is our view that the strongly conserved active centre of tyrosinases from different sources 22 makes it probable that very similar, if not identical, mechanisms apply to tyrosinases in general.
The current state of the evidence can be conveniently discussed under the following subheadings:

The requirement for oxygenase activity
4][25] A strong prediction arising from the mechanistic hypothesis proposed by the Quintox group 18 is that catechol oxidases, which do not possess oxygenase activity, will fail to exhibit suicide-inactivation kinetics since in this mechanism the copper reduction requires oxygenase generation of a hydroxylated intermediate (Scheme 3b).This is not predicted by the alternative mechanism advanced by Muñoz-Muñoz and co-workers 21 since their proposal depends on a protonation of the peroxy oxygen in a bidentate catechol complex (Scheme 4).
The behaviour of a catechol oxidase extracted from bananas (Musa cavendishii) indicates that, in contrast to tyrosinase, it oxidises a number of catecholic substrates without displaying suicide-inactivation kinetics. 19Earlier results published by Padrón and co-workers 26 show firstorder kinetics exhibited by the banana enzyme which are claimed as evidence of enzyme inactivation.However, in this study 26 the exponential decrease in oxidation rate probably reflects the fall in oxygen concentration as shown by the reported polarimetric results.In contrast we have shown that the banana enzyme exhibits no activity towards four monohydric phenols tested (4-methylphenol, 4-methoxyphenol, tyrosine and tyramine) under the same experimental conditions used for catechol oxidation.The absence of oxygenase activity was shown not to be due to lack of activation of the enzyme (Figure 2) or the presence of an inhibitor of oxygenase activity. 19All these data are consistent with the requirement for oxygenase activity for suicide inactivation of tyrosinase and provide very strong support for the Quintox mechanism (Scheme 3).Further indirect evidence of the involvement of an oxygenase presentation in the inactivation mechanism (Scheme 3) is the inhibitory effect of addition of a monohydric phenol to the catechol oxidation mixture.We have shown that the relative inactivation rate (the ratio of the inactivation rate to the initial rate of oxidase activity) of 4-methylcatechol oxidation by tyrosinase is diminished with increasing concentration of 4-methylphenol. 18While this is consistent with the Quintox mechanism (Scheme 3), this result would not be anticipated on the basis of the alternative proposal (Scheme 4) 21 since both the catechol oxidation rate and the inactivation rates would be similarly affected by a competing substrate.

Alternative substrate-binding orientations
The kinetic data of Muñoz-Muñoz and co-workers 21 indicate a correlation between the catalytic rate and the maximum apparent inactivation rate for a range of substituted diphenols tested (Figure 3), which is consistent with their mechanism which does not involve a difference in substrate binding.However, excluding the case of pyrogallol (discussed separately in Section 2.5), comparison of the data on four catecholic substrates tested by both Muñoz-Muñoz and co-workers 21 and the Quintox group 18 shows significant discrepancies in the ratio of the initial oxidation rate to the inactivation rate (k1/k2) (Table 1).The Quintox group ratios were obtained directly from plots of the total oxygen utilization versus the initial amount of enzyme, the slope of the line being defined as k1/k2, whereas the Muñoz-Muñoz et al. results were derived from kinetic parameters.It is possible that the estimates of k1 differ for technical reasons, since the assay procedure used by Muñoz-Muñoz and co-workers is designed to maintain a steady-state concentration of the substrate and depends on the calculation of the catalytic rate from the oxidation of a redox indicator.Nevertheless, the substantial discrepancy in the apparent inactivation rates is difficult to explain.Substituting the (essentially constant) k1 values obtained by Muñoz-Muñoz et al. in the Quintox Group data ARKAT USA, Inc. yields a ten-fold variation in k2 whereas the corresponding Muñoz-Muñoz et al. data for k2 21 differ by only 10%.
According to the Quintox mechanism 18 the inactivation comes about by a proportion of the substrate binding to the active site in an alternative orientation.The catechol may bind either in the usual bidentate oxidase mode to the active site copper atoms 14, or like a phenol in the oxygenase mode 15.8][29][30] Both modes (14 and 15) have well-defined and inflexible locations of the catechol unit and its associated substituents.It might be anticipated, therefore, that structural constraints associated with different binding modes would be reflected by a differential influence of substituents on the rate of oxidation (k1) and the rate of inactivation (k2).
Such differences have been observed by quantitative analysis of a series 4-substituted catechols with ring substituents chosen to give a wide variation and minimal correlation of substituent properties. 31In this study the statistically-significant relationships shown in Equations 1 and 2 were found between the rate constants (k1 and k2) and the length (L) and hydrophobicity () of the catechol substituents.The substituent effects for oxidation (k1) and inactivation (k2) are clearly different.log k1 = 0.364 (0.097)-0.297(0.077)L + 2.369 (1) log k2 = 1.143 (0.229)-0.819(0.243)  2 + 0.061 (2) This result would appear to be at variance with the mechanism proposed of Muñoz-Muñoz and co-workers 21 since the binding orientation is not critical in their mechanism shown in Scheme 4.

Hydroxylation products
There is evidence in the literature of the 5-hydroxylation of L-3,4-dihydroxyphenylalanine (dopa 8) which confirms the ability of tyrosinase from different sources to process catechols by the oxygenase pathway, [32][33][34][35] and a strong prediction of the Quintox mechanism is the generation of a hydroxylation product 12 of the catecholic substrate during the inactivation reaction (Scheme 3b).Because the inactivation reaction is relatively rare and involves small quantities of enzyme, in the experimental system used the ortho-hydroxylated quinone product 12 is a minor component of the reaction mixture.Nevertheless, the product 12 has been identified by hplc-ARKAT USA, Inc. mass spectrometry in the case of 4-methylcatechol oxidation by tyrosinase and shown to be correlated with the loss of enzyme activity. 19

Blocking the hydroxylation site
Since the Quintox suicide-inactivation mechanism requires ortho-hydroxylation to take place, the inactivation kinetics of tyrosinase during catechol oxidation should be prevented if an unsubstituted ring carbon adjacent to the catecholic function is not available.Since this is not a requirement of the mechanism suggested by Muñoz-Muñoz and co-workers 21 this provides a further opportunity for a critical experimental test of the mechanisms.As predicted by the Quintox mechanism, 3,6-difluorocatechol 16, which cannot act as an oxygenase substrate, acts as an oxidase substrate for tyrosinase but does not exhibit suicide-inactivation kinetics. 20he oxygen utilization during the oxidation of 3,6-difluorocatechol 16 by tyrosinase shows first-order kinetics and contrasts with the kinetics of an inactivating substrate 17 in an incubation mixture exhibiting a similar initial oxidation rate (Figure 4).This result, together with the results described in Section 2.1, very strongly supports the Quintox mechanism.

Interference with oxygenase binding
An alternative aspect of the prevention of cresolase binding is offered by the trihydric phenol pyrogallol 18 which, because of the restrictions imposed by hydroxyl groups in positions 4 and 5, 18 is only capable of binding to the active site of mushroom tyrosinase as a catechol.The Quintox mechanism predicts that this substrate will not be subject to suicide inactivation and this has been observed in relation to the kinetics of oxygen utilization. 18However, the data have been challenged on the grounds that the observed oxygen uptake is due to auto-oxidation and that pyrogallol is a strongly inactivating substrate. 21There is no doubt that there are some complications with this substrate arising from rapid secondary reactions, which may involve auto-oxidative steps, during the formation of the major oxidation product purpurogallin 19; 36 so it is not possible to make strong inferences from the data presently available.Since the orthoquinone formed from pyrogallol 18 rapidly forms purpurogallin 19, any assay system related to ortho-quinone formation may give the impression that the enzyme is inhibited and ortho-quinone formation has ceased.Based on oxygen uptake, we have found that pre-exposure to pyrogallol did not inactivate tyrosinase as shown by the subsequent oxidation of 4-methylcatechol (Figure 5). 18The claim that pyrogallol is an inactivating substrate is clearly incorrect.

Sensitivity to pH
We have shown that the relative inactivation rate of 4-methylcatechol oxidation by tyrosinase is diminished at lower pH of the incubation mixture.Over a limited range the reduction in relative inactivation was approximately log-linear with a negative slope of 6.7 x 10 -2 per pH unit. 18lthough this trend is predicted by the Quintox mechanism, since the alternative model 21 also involves a deprotonation step this characteristic does not distinguish between them.

Conclusions
It is well known that resorcinols also irreversibly inactivate tyrosinase. 37Although the details of the tyrosinase inactivation by resorcinols are not fully resolved, the Quintox mechanism is able to rationalize the process by a reductive elimination of zero-valent copper 20 (Scheme 5) 18 whereas this is not possible by the alternative proposal 21.Furthermore, a major difficulty posed by the alternative mechanism (Scheme 4) is that it implies that hydroquinone, as well as catechol, should be capable of acting as a suicideinactivator with formation of 1,4-benzoquinone, which is contrary to observation. 38ll the experimental tests so far carried out have been consistent with the predicted outcome of the Quintox inactivation mechanism of tyrosinase (Scheme 3).In contrast, the alternative mechanism (Scheme 4) does not explain all of the experimental observations, as discussed above.

Figure 1 .
Figure 1.Data of Dietler and Lerch on the correlation of suicide inactivation and 50% copper loss from Neurospora tyrosinase.17

Figure 2 .
Figure 2. Lack of oxygenase activity of catechol oxidase from Musa cavendishii.The chart shows the polarimetric results of three experiments conducted in 0.1 M phosphate buffer (pH 6.3).In all cases the enzyme was added at the time point indicated by A. In two cases (○) and (•) 4-methylcatechol (140 μM) was present from t = 0 and in one of these (○) 4-methylphenol (280 μM) was added at B. The control enzyme solution (■) also had 4-methylphenol added at the point indicated by B. There is no evidence of 4-methylphenol oxidation in the control (■) and no activation of oxygenase following pre-exposure of the enzyme to the catecholic substrate (○) in comparison with the pre-exposed control (•).

Figure 3 .
Figure 3. Relationship between the inactivation rate and catalytic rate (data of Muñoz-Muñoz et l. 21)

Figure 5 .
Figure 5. Failure of tyrosinase inactivation by pre-exposure to pyrogallol 18.The chart shows the polarimetric results of three experiments in 0.1 M phosphate buffer (pH 6.3).Tyrosinase additions (15 units) were made at points marked A. In two experiments, indicated by (○) and (•), pyrogallol (180 μM) was present from t = 0.At the point indicated by B 4-methylcatechol (500 μM) was added to the test incubation mixture (•).This exhibited the characteristic catecholic oxidation pattern in comparison with the control (○).The oxidation was equivalent to that shown by the unexposed control (□) and to that following additional enzyme addition at A.

Table 1 .
Comparison of kinetic data for four catecholic substrates.