Mechanistic insight into the activity of tyrosinase from variable-temperature studies in an aqueous/organic solvent

Article abstract of DOI:10.1002/chem.200501097

The activity of mushroom tyrosinase towards a representative series of phenolic and diphenolic substrates structurally related to tyrosine has been investigated in a mixed solvent of 34.4% methanol-glycerol (7:1, v/v) and 65.6% (v/v) aqueous 50 mM Hepes b

Full text of DOI:10.1002/chem.200501097

DOI: 10.1002/chem.200501097
Mechanistic Insight into the Activity of Tyrosinase from Variable-
Temperature Studies in an Aqueous/Organic Solvent
Alessandro Granata,[a] Enrico Monzani,[a] Luigi Bubacco,[b] and Luigi Casella*[a]
Abstract: The activity of mushroom ty-
rosinase towards a representative series
of phenolic and diphenolic substrates
structurally related to tyrosine has
been investigated in a mixed solvent of
34.4% methanol–glycerol (7:1, v/v) and
65.6% (v/v) aqueous 50 mm Hepes
buffer at pH 6.8 at various tempera-
tures. The kinetic activation parameters
controlling the enzymatic reactions and
the thermodynamic parameters associ-
ated with the process of substrate bind-
ing to the enzyme active species have
been deduced from the temperature
variation of the kcat and KM parameters.
The activation free energy is dominat-
ed by the enthalpic term, the value of
which lies in the relatively narrow
range of 61ꢀ9 kJmolꢁ1 irrespective of
substrate or reaction type (monophe-
nolase or diphenolase). The activation
entropies are small and generally nega-
tive and contribute no more than 10%
to the activation free energy. The sub-
strate binding parameters are charac-
terized by large and negative enthalpy
and entropy contributions, which are
typically dictated by polar protein–sub-
strate interactions. The substrate 4-hy-
droxyphenylpropionic acid exhibits a
strikingly anomalous temperature de-
pendence of the enzymatic oxidation
rate, with DHꢀ ꢂ150 kJmolꢁ1 and DSꢀ
ꢂ280 JKꢁ1 molꢁ1, due to the fact that it
can competitively bind to the enzyme
through the phenol group, like the
other substrates, or the carboxylate
group, like carboxylic acid inhibitors. A
kinetic model that takes into account
the dual substrate/inhibitor nature of
this compound enables rationalization
of this anomalous behavior.
Keywords: activation parameters ·
bioinorganic chemistry · enzymes ·
thermodynamic
tyrosinase
parameters
·
Introduction
Tyrosinase (EC 1.14.18.1) is a copper monooxygenase that
catalyzes the ortho-hydroxylation of monophenols (mono-
phenolase activity) and the oxidation of catechols (dipheno-
lase activity) to ortho-quinones according to the stoichiome-
tries given in Equations (1) and (2).[1]
The monophenolase reaction is often shown to proceed
through the initial formation of catechol followed by oxida-
tion to quinone,[2] but this point is still subject to considera-
ble debate as in vitro catechol is probably formed by way of
an indirect nonenzymatic mechanism.[3] The active site of
the enzyme contains a pair of antiferromagnetically coupled
copper(ii) ions in the met form, and during turnover both
the reduced and the oxy forms are cyclically produced.[4]
The structure of tyrosinase is not known, but evolutionary
correlations[5] and spectroscopic similarities[6] with two close-
ly related proteins belonging to the same type-3 copper
family that have been structurally characterized, namely he-
mocyanin[7,8] and catechol oxidase,[9] suggest a common
active site, in which six histidine ligands are bound to the
pair of copper ions. A direct indication of the presence of
six coordinating histidines in tyrosinase comes from para-
magnetic NMR experiments on the enzyme from Streptomy-
ces antibioticus.[10] Differences in the function of type-3
[a] Dr. A. Granata, Dr. E. Monzani, Prof. L. Casella
Dipartimento di Chimica Generale, Università di Pavia
Via Taramelli 12, 27100 Pavia (Italy)
Fax : (+39)0382-528-544
E-mail: bioinorg@unipv.it
[b] Prof. L. Bubacco
Dipartimento di Biologia, Università di Padova
30121 Padova (Italy)
Supporting information (the derivation of the rate equation and the
activation parameters for the enzymatic reaction involving a substrate
with two binding modes, and the kinetic parameters and rate plots for
the enzymatic oxidation of all of the phenolic and diphenolic sub-
strates at variable temperatures) for this article is available on the
WWW under http://www.chemeurj.org/ or from the author.
2504
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 2504 – 2514

FULL PAPER
copper proteins seem to largely depend on the accessibility
of the dioxygen binding site to potential substrates and se-
lectivity effects that discriminate the type of substrates.
Thus, while the active site of hemocyanins is basically inac-
cessible except to small ligand molecules, proteolytic or
sodium dodecyl sulfate (SDS) activation of the protein can
induce mono- and diphenolase activity.[11,12] Catechol oxidas-
es are normally only able to catalyze the diphenolase reac-
tion [Eq. (2)], and are inactive, or only very weakly active,
with monophenols.[13]
Despite extensive research efforts, very little is known
about the details of the mechanisms through which tyrosi-
nase catalyzes reactions according to Equation (1) and
Equation (2). The mode of substrate binding at the enzyme
and product release therefrom, and the mode of dioxygen
activation and cleavage at the dicopper center are, in partic-
ular, completely unknown.[14,15] Although the monopheno-
lase reaction [Eq. (1)] involves the energetically demanding
ꢁ
step of oxygen atom insertion into the aromatic C H bond,
the monophenolase catalytic cycle is simpler than the diphe-
nolase cycle. In fact, only the oxy form of tyrosinase is able
to carry out the phenol hydroxylation, while catechols are
oxidized both by the met form and the oxy form of the
enzyme. A simplified view of the commonly accepted steps
involved in these mechanisms is provided in Scheme 1 and
Scheme 2.
mal steric rearrangement in the step involving evolution of
the phenolate–peroxy intermediate in the monophenolase
cycle of Scheme 1. The two catalytic cycles of tyrosinase are
usually shown intermingled, due to the fact that in vitro the
monophenolase reaction typically exhibits a lag phase that
no longer occurs on the addition of limited amounts of a di-
phenol, the function of which is to reduce met tyrosinase,
the major component of the enzyme as isolated, and give
rapid access to the oxy form.[3]
A recent theoretical calculation performed on the mono-
phenolase reaction [Eq. (1)] indicates that the rate-limiting
ꢁ
step is associated with cleavage of the peroxide O O bond
(60.2 kJmolꢁ1), but the attack of peroxide on the phenolate
ring is also characterized by a significant energy barrier
(51.5 kJmolꢁ1).[15] A different conclusion has been reached
on the basis of kinetic investigations, according to which the
rate-limiting step is connected with phenol deprotonation
and binding to the copper–peroxo intermediate.[19,20] Indeed,
both tyrosinase catalytic cycles involve proton transfer steps
and, in particular, the phenol has been suggested to transfer
a proton to the peroxo group.[21]
In the present study, we report an investigation of the ac-
tivity of mushroom tyrosinase towards a group of represen-
tative substrates related to tyrosine (Figure 1) over a range
of temperatures in a mixed aqueous/organic solvent. In this
medium, which enables an extension of the range of temper-
atures normally accessible in aqueous buffer, we have ob-
Scheme 1.
Scheme 2. In these schemes, binding of catechol to met-tyro-
sinase has been represented to occur in the h2:h1 bridging
mode, rather than the commonly adopted m-1,4 bridging
mode.[4] The asymmetric h2:h1 binding mode is suggested by
collective evidence from inhibitor binding studies to half-
met[16] and met[17] Streptomyces antibioticus tyrosinase, spec-
troscopic studies on biomimetic catecholate-dicopper(ii)
complexes,[18] and by the X-ray structure of catechol oxidase
with a bound phenylthiourea inhibitor molecule.[9,13] The
h2:h1 catechol complex has the advantage of requiring mini-
Chem. Eur. J. 2006, 12, 2504 – 2514
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2505

L. Casella et al.
In Table 1, the kinetic parameters of the enzyme in our
selected aqueous/organic solvent are compared with litera-
ture data relating to aqueous buffer at the same pH. A fur-
ther comparison between results obtained at 108C and 258C
allows consideration of the effect of temperature on the ki-
netic parameters. The activity of tyrosinase in the cryosol-
vent is slightly decreased or slightly increased with respect
to that in aqueous buffer, depending on the substrate, but
the effect on kcat is generally accompanied by an increase of
Figure 1. Structures of the phenolic substrates employed in the kinetic
studies.
Table 1. Comparison of the kinetic parameters for the activity of tyrosinase in the mixed solvent of 34.4%
methanol–glycerol (7:1, v/v) and 65.6% (v/v) aqueous 50 mm Hepes buffer pH 6.8, and in 50 mm phosphate
buffer pH 6.8.[a]
tained the thermodynamic and
activation parameters associat-
ed with the rate-limiting step
of the enzymatic reactions.
Solvent[b]/T [8C]
Parameter
kcat [sꢁ1
KM [mm]
kcat [sꢁ1
KM [mm]
kcat [sꢁ1
KM [mm]
Tyramine
l-TyrOMe
Hpp[c]
Dopamine
l-Dopa
C/25
]
7.4ꢀ0.2
2.1ꢀ0.1
8.9ꢀ0.1
43.0ꢀ2.0
6.7ꢀ0.4
133.2ꢀ1.2
0.93ꢀ0.02
29.9ꢀ0.4
0.40ꢀ0.01
439ꢀ17.6
2.2ꢀ0.1
141.1ꢀ3.0
3.7ꢀ0.2
0.82ꢀ0.04
2.24ꢀ0.08
0.28ꢀ0.03
3.4ꢀ0.14[d]
0.38ꢀ0.04[d]
C/10
A/25
]
1.90ꢀ0.05
0.66ꢀ0.05
25.9ꢀ1.1
0.51ꢀ0.02
1.70ꢀ0.04
0.22ꢀ0.02
66.7ꢀ2.7
0.44ꢀ0.01
42.0ꢀ1.0
2.1ꢀ0.1
]
107.4ꢀ3.1
0.8ꢀ0.03
Results
[a] Data taken from reference [19]. [b] C = cryosolvent, A = aqueous buffer. [c] kcat and KM values reported
for Hpp are the apparent values obtained from simple Michaelis–Menten treatment. [d] Taken from: L. G.
Fenoll, J. N. Rodriguez-Lopez, F. Garcia-Molina, F. Garcia-Canovas, J. Tudela, Int. J. Biochem. Cell Biol. 2002,
34, 332–336.
In preliminary investigations,
the enzymatic activity of tyro-
sinase was tested in a variety
of mixed aqueous/organic sol-
vents having suitable viscosity characteristics for use as cryo-
solvents in low-temperature studies.[22] Among these media,
mixtures composed of methanol–glycerol (7:1, v/v) as organ-
ic component and aqueous Hepes buffer at pH 6.8 were
found to facilitate good performance of the enzyme. As we
were not interested in attaining very low temperatures in
the present study, we maintained a relatively large fraction
of aqueous buffer. Thus, using a ratio of 34.4% methanol–
glycerol (7:1, v/v) and 65.6% (v/v) aqueous 50 mm Hepes
buffer at pH 6.8, we found very good operation of the
enzyme in the temperature range of ꢁ20 to +308C. The
effect of the cryosolvent on enzyme activity was assessed
using l-Dopa as substrate. A very weak, mixed-type inhibi-
tion effect was found by studying the dependence of the
rate of l-Dopa oxidation as a function of the concentration
of the organic component of the cryosolvent (data not
shown). These weak solvent effects could therefore be ne-
glected in the treatment of kinetic data for the substrates in-
vestigated.
Addition of the organic component to the aqueous buffer
has the drawback of limiting the solubility of the natural
phenolic substrates containing charged groups in their side
chains, and for this reason we could not investigate the be-
havior of l-tyrosine over an extended concentration range.
However, a number of physiologically relevant tyrosine de-
rivatives, namely tyramine, l-TyrOMe, 4-hydroxyphenylpro-
pionic acid (Hpp), dopamine, and l-Dopa, exhibited suffi-
cient solubility in the cryosolvent to enable a detailed enzy-
matic study over a suitable range of temperatures. The enzy-
matic oxidation of N-acetyl-l-tyrosine was also investigated,
but it was found to be extremely slow in the cryosolvent, as
it is in aqueous buffer.[23]
KM, so that the efficiency of the enzyme is somewhat de-
creased in the aqueous/organic medium. The abrupt drop in
kcat at the lower temperature for Hpp is, however, worthy of
note. In all cases, the turnover rates were found to be inde-
pendent of [O2]. This is expected, in view of the higher solu-
bility of dioxygen in the organic medium with respect to
water[24] and considering the high value of the rate constant
for binding of O2 to deoxy tyrosinase.[25] In order to probe
the existence or otherwise of an isotope effect for the mono-
phenolase reaction in the mixed solvent, the activities of the
enzyme towards 3,5-[D2]-l-TyrOMe and l-TyrOMe were
compared (Figure 2).
The kinetic parameters kcat = 5.57ꢀ0.04 sꢁ1 and KM
=
0.61ꢀ0.04 mm obtained for 3,5-[D2]-l-TyrOMe at 208C are
similar to those for l-TyrOMe (kcat = 5.97ꢀ0.08 sꢁ1, KM
=
0.58ꢀ0.02 mm), but indicate a small but non-negligible iso-
D
tope effect on the rate constant (kcatH/kcat = 1.1) in the
mixed aqueous/organic solvent. The isotope effect was
found to be unchanged by repeating the measurements at
D
158C and 258C (kcatH/kcat = 1.1ꢀ0.05). This is in contrast
with the absence of isotope effect found for the enzymatic
oxidation of p-chlorophenol using hydrogen peroxide in-
stead of dioxygen as oxidant in 0.5m aqueous borate buffer
at pH 7.0.[26]
The initial rates for the enzymatic oxidation of phenolic
and diphenolic substrates were measured as a function of
substrate concentration at various temperatures in the cryo-
solvent. A representative family of curves obtained for l-
Dopa is shown in Figure 3.
A complete list of kinetic parameters at various tempera-
tures along with the corresponding rate plots for all of the
substrates investigated is provided in the Supporting Materi-
2506
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 2504 – 2514

Mechanistic Study of Tyrosinase Activity
FULL PAPER
Figure 4. Eyring plots for the tyrosinase-catalyzed oxidation of l-Dopa
Figure 2. Rate dependence on substrate concentration, at 208C, for the
!
&
*
*
(&), dopamine ( ), tyramine ( ), l-TyrOMe ( ), and Hpp ( ) in the
mixed solvent of 34.4% methanol–glycerol (7:1) and 65.6% aqueous
Hepes buffer (50 mm) at pH 6.8.
*
catalytic oxidation of l-TyrOMe (&) and 3,5-[D2]-l-TyrOMe ( ) by tyro-
sinase in the mixed solvent of 34.4% methanol–glycerol (7:1) and 65.6%
aqueous Hepes buffer (50 mm) at pH 6.8.
considerably extended at low temperature and did not allow
the determination of reliable kinetic data below about 88C.
In any case, the temperature range analyzed, up to about
308C, ensures a constant enthalpy contribution. The kcat
values at 258C were used to calculate the activation free
energy DGꢀ for the enzymatic reactions at this temperature
(Table 2). It can be seen from Table 2 that the DHꢀ values
fall in the relatively narrow range of 61ꢀ9 kJmolꢁ1 for all
the substrates bearing an amino substituent on the side
chain attached to the aromatic nucleus, but that the activa-
tion enthalpy increases dramatically to about 150 kJmolꢁ1
for Hpp. Interestingly, dopamine has a significantly larger
DHꢀ value than that in the case of l-Dopa. This is consistent
with the more pronounced rate dependence on temperature
for the former substrate. Also, DSꢀ values are small and
negative for tyramine, l-TyrOMe, and l-Dopa, but this pa-
rameter becomes small and positive for dopamine and large
and positive for Hpp. This is important because the entropic
term significantly reduces the enthalpic contribution to DGꢀ
for the latter two substrates (especially for Hpp) at room
Figure 3. Rate dependence on substrate concentration at different tem-
peratures of the oxidation of l-Dopa by tyrosinase in the mixed solvent
of 34.4% methanol–glycerol (7:1) and 65.6% aqueous Hepes buffer
(50 mm) at pH 6.8.
al. Eyring plots of lnkcat/T
against Tꢁ1 yielded the activa-
tion parameters DHꢀ and DSꢀ;
these are shown in Figure 4 for
all of the substrates.
Table 2. Activation parameters for tyrosinase activity and thermodynamic parameters for substrate binding in
the mixed solvent of 34.4% methanol–glycerol (7:1, v/v) and 65.6% (v/v) aqueous 50 mm Hepes buffer pH 6.8
and in 50 mm phosphate buffer at pH 6.8.
Parameter/Substrate
DHꢀ [kJmolꢁ1
DSꢀ [JKꢁ1 molꢁ1
DGꢀ [kJmolꢁ1] (258C)[a]
DH8 [kJmolꢁ1
Tyramine
l-TyrOMe
Hpp[c]
Dopamine
l-Dopa
The temperature range in-
vestigated for each substrate
was defined by the reduced
stability and inactivation of the
protein at high temperature
and by the very low activity as
the temperature approaches
08C. The latter problem
proved to be especially rele-
vant with the phenolic sub-
strates, since the lag phase was
]
+60.9ꢀ1.5
ꢁ24ꢀ5
68.1
+61.9ꢀ2.0
ꢁ19ꢀ11
67.5
+148.6ꢀ2.6
+281ꢀ8
64.9
+67.1ꢀ2.2
21ꢀ7
+53.8ꢀ1.6
ꢁ23ꢀ6
60.7
]
60.8
]
ꢁ55.1ꢀ4.7
ꢁ134ꢀ16
ꢁ15.3
ꢁ52.4ꢀ2.8
ꢁ117ꢀ10
ꢁ17.6
ꢁ157.2ꢀ4.0
ꢁ485ꢀ12
ꢁ12.4
ꢁ39.6ꢀ2.2
ꢁ75ꢀ7
ꢁ17.3
ꢁ26.8ꢀ1.2
ꢁ43ꢀ4
ꢁ13.9
DS8 [JKꢁ1 molꢁ1
]
DG8 [kJmolꢁ1] (258C)[b]
[d]
DHꢀon [kJmolꢁ1
]
5.8ꢀ3.5
ꢁ158ꢀ12
53
9.5ꢀ1.8
ꢁ136ꢀ6
50
ꢁ9ꢀ8
27.5ꢀ1.6
ꢁ54ꢀ5
44
27.0ꢀ1.2
ꢁ66ꢀ8
47
DSꢀon [JKꢁ1 molꢁ1
DGꢀon [kJmolꢁ1] (258C)[b,d]
]
ꢁ204ꢀ30
52
[d]
ꢁ1
[a] Calculated from kcat values at 258C. [b] Calculated from KM values at 258C. [c] The parameters obtained
for Hpp are apparent values affected by the inhibition process. [d] The activation parameters for substrate
binding are valid in the hypothesis of slow substrate dissociation from the enzyme (kcat @ koff).
Chem. Eur. J. 2006, 12, 2504 – 2514
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2507

L. Casella et al.
temperature. The Eyring plots clearly show the anomalous
behavior of Hpp with respect to the other substrates
(Figure 4). Except in the case of Hpp, the DGꢀ values are
essentially determined by DHꢀ, with entropy contributions
restricted to no more than ꢀ10%.
cording to the simplified reaction scheme given in Equa-
tion (3).
In the approximation of a fast substrate binding process
ꢁ1
(see Supporting Information), the KM values are the ap-
parent association constants of the substrates to the
enzyme-active species. Plots of lnKMꢁ1 against Tꢁ1 were con-
structed and used to calculate DH8 and DS8, the enthalpy
and entropy changes, respectively, associated with the for-
mation of the enzyme-substrate complex in the steady state
(Figure 5).
Here, S and S’ denote Hpp acting as a substrate and as an
inhibitor, respectively, EO2S is the active intermediate of
the enzymatic reaction, and ES’ represents an unproductive
complex. This inhibitory scheme yields the kinetic equa-
tion (4) (for its derivation, see the Supporting Information).
kcat
½SŠ
1þKIKM
rate=½E0Š ¼
ð4Þ
KM
þ ½SŠ
1þKIKM
Here, KI = kI/kꢁI is the binding constant of the inhibitor,
and KM = (kcat + koff)/kon. When studied as a function of
substrate concentration, the rate plot shows the usual sub-
strate saturation behavior, but with reduced maximum activ-
ity and saturation at a reduced substrate concentration. The
apparent kinetic parameters deduced from these plots are
related to the true parameters by the factor 1/(1 + KIKM).
Therefore, the activation and thermodynamic parameters re-
ported for Hpp in Table 2 are apparent values affected by
the presence of two substrate binding modes.
To get an idea of the effect of substrate inhibition on the
activation parameters, we can approximate the apparent cat-
alytic constant as (kcat)app ꢂkcat/(KIKM). With this approxi-
mation, the apparent activation parameters are given by
Equations (5) and (6),
Figure 5. Van’t Hoff plots for the tyrosinase-catalyzed oxidation of l-
!
&
*
*
Dopa (&), dopamine ( ), tyramine ( ), l-TyrOMe ( ), and Hpp ( ) in
the mixed solvent of 34.4% methanol–glycerol (7:1) and 65.6% aqueous
Hepes buffer (50 mm) at pH 6.8.
The values of these thermodynamic parameters, together
ꢁ1
with the DG8 values calculated from the KM values at
258C, are collected in Table 2. The enthalpic and entropic
contributions to the binding energies of the substrates are
somewhat scattered, reflecting specific effects involving the
polar substituents on the side chains, but in general both
terms are large and negative. A similar trend in the enthalpy
and entropy terms was noted previously for the binding of
aromatic acid inhibitors to tyrosinase[21] and of simple phe-
nolic compounds to heme peroxidases.[27,28] Notably, both
the DH8 and DS8 parameters show larger negative values
with phenols than with catechols.
The strikingly anomalous temperature dependence of the
catalytic reaction of Hpp stems only from the presence of a
carboxylate group on the side chain of the phenol nucleus.
Carboxylic acids of a variety of structural types have been
shown to act as inhibitors of tyrosinase because they strong-
ly bind to the met form of the enzyme upon displacing the
peroxide from oxytyrosinase.[21,29] The case of Hpp is there-
fore unique, because it can act as a substrate, by binding to
oxytyrosinase through the phenol group, or as an inhibitor,
by displacing the peroxide like other carboxylic acids. The
behavior of such a bifunctional molecule can be treated ac-
DHꢀ
DSꢀ
ꢂ DHꢀ þ DHo ꢁ DHo
ð5Þ
ð6Þ
app
I
ꢂ DSꢀ þ DSo ꢁ DSo
app
I
where DH8I and DS8I refer to the enthalpy and entropy
changes, respectively, associated with the formation of the
enzyme-substrate complex in the inhibitory pathway (con-
nected to KI). The averages of the values of DHꢀ, DSꢀ,
DH8, and DS8 obtained with tyramine and l-TyrOMe
(Table 2) can be used as estimates of the corresponding
values for Hpp. Assuming DH8I and DS8I values for Hpp
similar to the thermodynamic data for the inhibition of tyro-
sinase activity by benzoic acid (DH8I = ꢁ32 kcalmolꢁ1
=
ꢁ134 kJmolꢁ1, and DS8I
=
ꢁ85 calKꢁ1 molꢁ1
=
ꢁ356 JKꢁ1 molꢁ1),[21] and substituting these values into the
apparent activation parameters, we obtain DHꢀ
ꢂ(61 ꢁ
app
54 + 134) = 141 kJmolꢁ1 and DSꢀ ꢂ(ꢁ22 ꢁ 126 + 356)
app
= 208 JKꢁ1 molꢁ1. These values are comparable with the
values reported in Table 2 for Hpp and suggest that the acti-
vation energy parameters for tyrosinase activity are consis-
2508
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 2504 – 2514

Mechanistic Study of Tyrosinase Activity
FULL PAPER
tent with those observed with the other phenolic substrates
for this substrate as well. The same conclusion can be drawn
by applying a similar treatment to the apparent Michaelis
constant (data not shown). It is interesting to note that, ac-
cording to Equation (4), the kcat/KM ratio is independent of
the presence of two substrate binding modes. Therefore, the
same should hold true for the observed activation parame-
ters, DHꢀ = DHꢀ + DH8 and DSꢀ = DSꢀ + DS8, ob-
T1b, are related to the paramagnetic contribution to relaxa-
tion, T1M, to the diamagnetic contribution to relaxation for
the bound inhibitor, T1D, and to the lifetime for chemical ex-
change, tM, through Equation (7).[32]
1=T1b ¼ 1=T1D þ 1=ðT1M þ tMÞ
ð7Þ
Since kojic acid is bound to the copper(ii) ions in the
obs
obs
enzyme-inhibitor complex, both dipolar and contact contri-
tained by applying the Eyring equation to kcat/KM. The
values obtained for tyramine, l-TyrOMe, and Hpp are
DHꢀ = 5.8ꢀ3.5, 9.5ꢀ1.8, and ꢁ9ꢀ8 kJmolꢁ1 and DSꢀ
[32]
butions influence T1M
.
In the presence of a strong para-
magnetic effect, as in the present case, the diamagnetic con-
tribution to relaxation can be neglected. Equation (7) thus
simplifies to T1b = T1M + tM. The T1b values obtained for
the protons at the various positions of the bound inhibitor
are as follows (Figure 6): proton 1, 4.6ꢀ0.1 ms; proton 2,
obs
obs
= ꢁ158ꢀ12, ꢁ136ꢀ6, and ꢁ204ꢀ30 JKꢁ1 molꢁ1, respec-
tively. The similarity between the values for these three sub-
strates, almost within the experimental error, shows that by
taking into account the inhibition that stems from the car-
boxylate group, Hpp behaves like the other phenols investi-
gated.
The analysis outlined above assumes that substrate disso-
ciation from the enzyme–substrate complex is a fast process,
that is, it is faster than substrate oxidation (kcat ! koff). This
is at variance with the hypothesis of Garcia-Canovas, ac-
cording to which substrate binding is a slow equilibrium pro-
cess.[30] Assuming that the latter statement is correct, the
meaning of KM changes to the ratio of the rate constants,
KM = kcat/kon (see Supporting Information), where kon is the
rate constant for substrate binding to tyrosinase. The plot of
ꢁ1
lnKM against Tꢁ1 in Figure 5 then gives “observed” DH8
and DS8 values, which are combinations of the activation pa-
rameters for substrate binding (E + S Q ES), DHꢀ and
on
DSꢀon, and the enzyme reaction (ES ! E + product), DHꢀ
and DSꢀ, respectively. Interestingly, in the approximation of
slow substrate binding, the ratio kcat/KM is kon. Thus, the acti-
vation parameters DHꢀ and DSꢀ can be obtained by ap-
on
on
Figure 6. Dependence of the longitudinal relaxation rate (1/T1obs) of kojic
acid protons on the fraction of the inhibitor bound to the protein, [E0]/
([S]+KD). Measurements were performed in 50 mm deuterated sodium
phosphate buffer, pD 6.8, containing a small amount of EDTA. The con-
centration of kojic acid was 10.0 mm, while that of tyrosinase ranged
from 0 to 12.5 mm. The values on the plots refer to the position of the
protons of kojic acid, as indicated in the inset.
plying the Eyring equation to kcat/KM (Table 2). The large
and negative values for DSꢀ are consistent with the bind-
on
ing process, with values mainly dependent on the nature of
the phenol or diphenol substrate. Also, the DHꢀ values are
on
different for phenols and catechols. In the case of Hpp,
DHꢀ is subject to a large standard deviation, probably due
on
to the dual character of this compound as a substrate and in-
hibitor.
The conversion of the ternary complex EO2S to E and
product in the monophenolase reaction [Eq. (3)], which is
governed by the rate constant kcat, is a composite process. It
consists of the oxidation of the substrate bound at the active
site to quinone and the subsequent dissociation and diffu-
sion of the product into the bulk solution, leaving the
enzyme in the dicopper(i) form. In the diphenolase reaction,
the corresponding process leaves the enzyme in the dicop-
per(ii) form. The slowest step is that controlling the value of
the kinetic constant. In order to get an idea of the timescale
of the processes of substrate association to the enzyme and
product dissociation therefrom, the paramagnetic contribu-
9.9ꢀ0.1 ms; protons 3, 7.4ꢀ0.4 ms. The fact that the T1b
values for the various kojic acid protons are different indi-
cates that T1b is dominated by T1M. Therefore, the lifetime
for chemical exchange must be shorter than the shortest of
the T1b values, that is, tM < 4.6 ms. A comparable limiting
value for tM (< 3 ms) has been reported for the binding of
p-nitrophenol to S. antibioticus met-tyrosinase.[17] The short
lifetime for protein-bound kojic acid indicates a fast dissoci-
ation of this strong inhibitor from met-tyrosinase. The disso-
ciation of quinone from the copper(i) centers of deoxy-tyro-
sinase, the process involved in reaction according to Equa-
tion (3), should be even faster, because copper(i) is more ki-
netically labile than copper(ii). As the reciprocal values of
kcat are larger than 4.6 ms, this indicates that kcat is governed
by the reaction occurring at the enzyme active site and not
by product dissociation.
1
tion of the protein to the H NMR relaxation rates of the in-
hibitor kojic acid, which strongly binds to the copper(ii) ions
of the enzyme,[31] has been evaluated. The longitudinal re-
laxation times of the nuclei of the protein-bound inhibitor,
Chem. Eur. J. 2006, 12, 2504 – 2514
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2509

L. Casella et al.
Discussion
to be an order of magnitude faster than that of oxy-tyrosi-
nase.[25]
Reports on the activity of tyrosinase in organic or mixed
aqueous/organic solvents have appeared previously,[33,35] al-
though, in general, the main scope of these studies was to
reduce the extent of product polymerization and enzyme de-
activation. Zaks and Klibanov additionally showed that the
amount of water actually required by mushroom tyrosinase
to function efficiently in an organic medium is very small.[33]
The present study enables us to obtain a more detailed pic-
ture of the behavior of the enzyme in the key steps of the
monophenolase and diphenolase reactions. The linearity of
the Eyring plots in Figure 3 indicates that there is no change
in the rate-determining step of the reaction or a conforma-
tional change in the tyrosinase that affects its catalytic effi-
ciency over the temperature range investigated. The 1H
NMR relaxation rate experiments performed with both
kojic acid and p-nitrophenol[17] show that the ligand associa-
tion/dissociation processes at tyrosinase are fast. Since the
binding of dioxygen to deoxy-tyrosinase is very fast,[25] we
can assume that the respective rate-limiting steps of the
monophenolase and diphenolase reactions are represented
by the processes given in Equations (8) and (9).
The kinetic activation parameters in Table 2 thus describe
the conversion of the ternary complex consisting of the
enzyme, dioxygen, and substrate into the transition state. If
we exclude from the comparison the anomalous case of
Hpp, the data reveal clearly pronounced trends. The relative
lack of dependence of DHꢀ on substrate and reaction type
indicates that the key event controlling enzyme activity is
common to the transition states involved in the monopheno-
lase and catecholase reactions. This necessarily leads to the
ꢁ
conclusion that peroxide O O bond cleavage is the main
determinant for the evolution of the transition state in both
enzymatic reactions. The DHꢀ values found here (61ꢀ
9 kJmolꢁ1) are consistent with theoretical calculations of the
ꢁ
energy barrier involved in the O O cleavage step of the
monophenolase[15] and catecholase[37] reactions, even though
the overall reaction mechanisms involved in these cases
differ from that proposed here, and with the activation en-
thalpy recently determined for a biomimetic catecholase re-
action.[38] In Scheme 3, phenol hydroxylation is shown to
occur by way of an electrophilic aromatic substitution mech-
anism, which is consistent with a recent investigation on ty-
rosinase
activity
towards
simple para-substituted phe-
nols.[39]
This rules out the possibility
that the reaction proceeds
through steps involving radical
species, in agreement with ex-
perimental results.[40] A con-
ꢁ
certed pathway of O O bond
ꢁ
cleavage and C O bond for-
mation is assumed here to ac-
count for the overall larger
energy barrier involved in the
monophenolase reaction com-
pared to the catecholase reac-
The substrates are shown to bind to copper in the depro-
tonated form, as is invariably assumed,[3,4,6] because a pro-
tein base, most probably a histidine,[36] assists in this process.
This is confirmed by inhibition studies of tyrosinase using p-
nitrophenol[17] and the formation of dead-end complexes be-
tween less acidic phenolic substrates and the met form of
the enzyme.[19,30] Both processes depicted in Equations (8)
and (9) occur with binding of the substrate to the oxy form
of the enzyme, but an impor-
tion. The concerted pathway is consistent with the small but
non-negligible isotope effect observed on kcat with 3,5-[D2]-
ꢁ
l-TyrOMe, which indicates that the C H bond is involved in
the formation of the activated complex. Small kinetic iso-
tope effects, between 1 and 3, are not unusual in electrophil-
ic aromatic substitutions,[41] and are generally due to a parti-
tioning effect,[42] which involves reversibility in the forma-
tion of the intermediate. A solvent deuterium isotope effect
tant difference is the release of
deoxy-tyrosinase in the mono-
phenolase reaction [Eq. (8)]
and of met-tyrosinase in the di-
phenolase reaction [Eq. (9)].
In the catecholase cycle depict-
ed in Scheme 2,
a second
molecule of catechol is oxi-
dized by met-tyrosinase, but
this reaction has been shown
Scheme 3.
2510
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 2504 – 2514

Mechanistic Study of Tyrosinase Activity
FULL PAPER
has also been found for tyrosinase reactions.[21,43] The effect
was attributed to the transfer of a proton from the substrate
hydroxyl group to the bound peroxide, but it is actually
more difficult to rationalize, because the solvent kH/kD
values do not correlate with the substrate acidity or the elec-
tron density at the aromatic carbon bearing the hydroxyl
group. In addition, the effect of the deuterated buffer on the
enzyme structure was not assessed. The absence of isotope
effect found for the enzymatic oxidation of p-chlorophenol
by hydrogen peroxide[26] may be related to the stronger acid-
ity of the hydrogen ortho to the OH group due to the elec-
tron-withdrawing effect of the Cl. In a very recent study on
the ternary complex involved in Equation (9) according to
the two alternative routes represented in Scheme 4 and
Scheme 5, depending on whether proton transfer from
ꢁ
bound catecholate occurs before or after O O bond cleav-
ꢁ
a model complex, cleavage of the m-peroxo O O bond to
generate the bis(m-oxo) form was shown to occur before
phenol hydroxylation.[44] The possibility that in the protein a
Scheme 4.
bis(m-oxo)dicopper(iii)
com-
plex, rather than the usually
assumed m-peroxodicopper(ii)
isomer, might be the active
species in the phenol hydroxyl-
ation is intriguing, but is incon-
sistent with the isotope effect
found here for tyrosinase.
Regarding the activation en-
tropy term for the phenol hy-
droxylation, a negative value is
Scheme 5.
expected for the evolution of the ternary complex to the
transition state, because the phenolate residue is strongly
age. Energetic considerations suggest that the two pathways
should be essentially equivalent, since proton transfers in-
volve a small energy barrier and heterolytic cleavages of
ꢁ
immobilized upon C O bond formation. This is confirmed
ꢁ
by the activation entropy data obtained from model studies
on intramolecular aromatic hydroxylations of the endoge-
nous ligands by peroxodicopper(ii)[45] or hydroperoxodicop-
per(ii)[46] complexes and, more recently, for the hydroxyl-
ation of an exogenous phenolate by a biomimetic peroxodi-
copper(ii) complex.[47] The small DSꢀ term found here for
the enzymatic monophenolase reaction implies a high level
of preorganization in the active site of the enzyme. This in-
dicates that very little structural rearrangement occurs upon
peroxo or hydroperoxo O O bonds are expected to be simi-
lar in energy.[15,37]
Assuming that the substrates bind to tyrosinase in a fast
pre-equilibrium step, the Michaelis constants give informa-
tion on the affinities for the enzyme active species. In gener-
al, when differences in steric effects are minimized, the af-
finity of the substrate has been shown to correlate with the
electron density on the phenol.[48] The thermodynamics of
substrate binding provides some further useful information
about the type of interaction occurring at the active site. In
all cases, the tyrosinase–substrate binding interaction is en-
thalpy driven (Table 2). This seems to be a general charac-
teristic of enzyme-substrate complexes involving polar phe-
nolic and diphenolic compounds, as has been shown for per-
oxidases.[27,28] In the present case, the binding enthalpies are
indeed large, probably because, compared to the peroxidase
complexes with simple phenols, the polar groups on the side
chains of the substrates studied here give rise to additional
electrostatic or hydrogen-bonding interactions with suitable
amino acid residues in the tyrosinase active site. In addition,
the polar interactions with peroxidases involving the phenol
hydroxyl groups are replaced here by covalent CuII–pheno-
late or CuII–catecholate bonds. In the case of both peroxi-
dases[28] and tyrosinase,[14,21] it is usually assumed that the ar-
omatic nucleus of the substrate is involved in hydrophobic,
ring-stacking interactions with an aromatic residue, but on
energetic grounds this contribution is overwhelmed by polar
interactions. The large and negative binding entropy terms
ꢁ
cleavage of the O O bond in the transition state.
The interpretation of the diphenolase reaction is compli-
cated by the fact that cleavage of the peroxide bond in-
volves coupled electron and proton transfers from the sub-
strate,[30,37,38] and it is difficult to predict the sequence of
events, since it is likely that protein residues at the active
site, as well as ionizable groups on the substrate, participate
as proton storage/delivery devices. Compared to the mono-
phenolase reaction, proton transfers from the diphenol are
greatly facilitated by its stronger acidity. As a matter of fact,
the DSꢀ term shows opposite sign with dopamine and l-
Dopa, although its value remains rather small. In a recently
studied model catecholase reaction, a large and positive
value of DSꢀ was found,[38] although in this case the sub-
strate possessed two bulky tert-butyl substituents on the cat-
echol ring, which most probably caused excessive crowding
in the ternary complex. This crowding would have been re-
ꢁ
lieved upon cleavage of the O O bond and release of the
quinone product. We can thus formulate the evolution of
Chem. Eur. J. 2006, 12, 2504 – 2514
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2511

L. Casella et al.
in Table 2 are also characteristic of protein-substrate interac-
tions dominated by polar interactions,[27,28] reflecting more
ordered states in the protein complexes. In the present case,
we also have to take into account the fact that substrate
binding occurs to the oxy form of the protein (reactions ac-
cording to [Eq. (8)] and [Eq. (9)]), producing a somewhat
crowded ternary complex. The difference between the bind-
ing parameters for monophenols and diphenols might be as-
sociated with the view that in the first case the binding
occurs to the CuA center and in the second case to the CuB
center of the enzyme.[36] However, cw-EPR experiments on
half-met-tyrosinase adducts with p-nitrophenol and kojic
acid show that binding always occurs to the oxidized copper,
suggested to be CuB.[16] When nonpolar, hydrophobic effects
dominate protein-substrate interactions, the situation is com-
pletely different. For instance, binding of camphor to P.
Putida cytochrome P450 occurs with DH8 = 0 and DS8 =
+110 JKꢁ1 molꢁ1,[49] and binding of benzphetamine to
human CYP2B4 occurs with DH8 = +29.3 kJmolꢁ1 and
DS8 = +170 JKꢁ1 molꢁ1,[50,51] that is, both processes are
strongly entropy driven. In the latter case, the major contri-
bution to the binding free energy is due to desolvation of
the active site upon substrate binding, which is relatively
large, while polar (hydrogen bonding/electrostatic) contribu-
tions are very limited. The thermodynamic data for inhibitor
binding to tyrosinase are fully consistent with the above ar-
guments, even though in this case the binding involves the
met form of the protein. Compare, for instance, the parame-
ters for binding of benzoic acid and toluene to mushroom
If the hypothesis of slow substrate binding due to Garcia-
Canovas were valid,[30] the parameters obtainable from the
temperature dependence of the rates would not be DH8 and
DS8 but the activation parameters for substrate binding,
DHꢀ and DSꢀon. The data reported in Table 2 show very
on
low DHꢀ values for phenolic substrates. These low values
on
cannot be simply attributed to the susceptibility to ligand
substitution of copper(ii) ions, since similar values should
then be observed with catechols. Furthermore, ligand substi-
tutions in copper(ii) model complexes are associated with
much larger activation enthalpies, ranging from 25 to
85 kJmolꢁ1, depending on the complex and the reaction con-
ditions.[52–54] Therefore, at least for phenolic substrates, the
hypothesis of slow binding to tyrosinase is unlikely. The
DHꢀ values reported in Table 2 for dopamine and l-Dopa
on
could be consistent with the slow substrate binding hypothe-
sis,[30] but this is contradicted by NMR relaxation data per-
taining to the binding of inhibitors at the enzyme, especially
the catechol-like kojic acid (Figure 6), which show that this
process is always fast.
Conclusions
In conclusion, tyrosinase functions in the aqueous/organic
solvent used in the present study with only minor alterations
of catalytic efficiency compared to in an aqueous solution.
There is no evidence that the cryosolvent alters the funda-
mental nature of the enzymatic reactions and therefore the
kinetic activation parameters derived from the variable-tem-
perature study reflect the intrinsic features of the evolution
of the transition states. The findings that 1) the activation
tyrosinase: DH8I
=
ꢁ134 kJmolꢁ1 and DS8I
=
ꢁ356 JKꢁ1 molꢁ1 as opposed to DH8I = +54 kJmolꢁ1 and
DS8I
=
+226 JKꢁ1 molꢁ1, respectively.[21] As we noted
above, the DH8I and DS8I terms for benzoic acid binding re-
semble the exceedingly large and negative binding parame-
ters that we have found for the anomalous substrate Hpp
(Table 2).
free energy is essentially determined by cleavage of the per-
ꢀ
ꢁ
oxide O O bond, and 2) the DS term is small, are fully
consistent with the view that the enzyme environment must
be preorganized to be complementary to the transition state
configuration of the reactants.[55,56] This is particularly impor-
tant for the monophenolase reaction, in which the substrate
The inhibitory effect of carboxylic acids on tyrosinase has
been known for a long time.[21,29] Therefore, it is somewhat
surprising to note that the reported activity of the enzyme
towards phenolic acids such as Hpp is not different from
that observed for related phenolic substrates lacking the car-
boxylate function or bearing an amino group at the a-
carbon, as is the case in tyrosine.[19] The temperature de-
pendence of the enzyme activity studied here reveals the
anomalous behavior with this particular type of substrate,
for which there is competitive binding to the enzyme
through the phenol or carboxylate groups. The presence of
an amino group at the Ca atom of the side chain in the sub-
strate apparently prevents the latter possibility. At room
temperature, fast equilibration of the two binding modes en-
ables the apparent normal behavior of Hpp as a substrate,
and it actually becomes highly reactive at higher tempera-
tures, probably because its negative charge makes the
phenol nucleus more electron-rich. On lowering the temper-
ature, however, the effect of the large exothermic inhibition
term becomes dominant. This effect exceeds the usual de-
crease in enzyme activity on lowering the temperature.
ꢁ
undergoes an energetically demanding C H bond cleavage
process. The strikingly anomalous behavior of Hpp can be
fully rationalized by a kinetic model that takes into account
the dual substrate/inhibitor nature of this compound.
The evidence from the substrate binding data indicates
that the active site of tyrosinase is polar. However, a large
portion of the substrate binding energy involves interactions
of the polar groups on the side chain, far from the reaction
site on the phenolic or diphenolic nucleus. This is confirmed
by the substrate stereospecificity exhibited by tyrosinase,
whereby replacing l-tyrosine with d-tyrosine or l-Dopa with
d-Dopa affects only KM and not kcat of the enzymatic reac-
tion.[57] In addition, as shown by model studies on the mono-
phenolase reaction employing a dinuclear copper complex,
phenol binding to the peroxodicopper(ii) species appears to
be weak.[47] Therefore, the polar interactions essentially re-
sponsible for enzyme catalysis must involve preoriented pro-
tein dipoles capable of stabilizing the development of charg-
es in the transition state (Scheme 3), in agreement with the
2512
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 2504 – 2514

Mechanistic Study of Tyrosinase Activity
FULL PAPER
current understanding of the origin of enzyme catalysis.[56,58]
The results described here provide a rationale for the plan-
ning of further experiments that will allow direct scrutiny of
the key intermediates of the tyrosinase catalytic cycles by
low-temperature spectroscopy in the aqueous/organic
medium investigated.
nations, it was convenient to monitor the difference between the absorb-
ance at 476 nm and that at 800 nm, at which the absorbance remains neg-
ligible during the assay.
The rate dependence on [l-Dopa] and on [dopamine] was determined by
varying the substrate concentration from 0.0 to 8.0 mm and from 0.0 to
3.0 mm, respectively, using a tyrosinase concentration of 12.5 nm. The
temperature dependences of the rates were studied in the ranges 2.0–
28.08C and 0.0–28.08C, respectively. The rates in units of DAsꢁ1 were cal-
culated from the slopes of the traces at 476 nm in the first 20–25 s after
mixing of the reagents; the conversion from units of DAsꢁ1 to sꢁ1 was
made using the dopachrome extinction coefficient and the tyrosinase
concentration.
Experimental Section
The dependence of the rate of l-Dopa oxidation as a function of the con-
centration of the organic component of the cryosolvent was studied at
88C by varying the amount of methanol–glycerol (7:1, v/v) contained in
aqueous 50 mm Hepes buffer at pH 6.8, saturated with atmospheric
oxygen. The concentration range of the organic component analyzed was
from 0 to 42.5% (v/v). The l-Dopa concentration was varied from 1 to
21 mm, while the concentration of tyrosinase was 35 nm.
Reagents: Tyramine, l-tyrosine, l-tyrosine methyl ester hydrochloride, N-
acetyl-l-tyrosine, Hpp, l-Dopa hydrochloride, and kojic acid were ob-
tained from Aldrich. 3,5-[D2]-l-Tyrosine (99.2 atom% D) was obtained
from CDN isotopes. All the other chemicals were of analytical grade and
were used as received. 3,5-[D2]-l-Tyrosine methyl ester hydrochloride
was prepared by esterification of 3,5-[D2]-l-tyrosine according to the fol-
lowing procedure. A stream of dry hydrogen chloride gas was passed
through a solution of 3,5-[D2]-l-tyrosine (0.5 g) in dry methanol (10 mL)
for 0.5 h. The solution was then gently heated at reflux for 3 h. After
cooling, the solution was concentrated to dryness in vacuo. Recrystalliza-
tion of the oily residue from ethanol/diethyl ether gave the final product
The rate dependence on [tyramine], [l-TyrOMe], and [3,5-[D2]-l-
TyrOMe] was determined in the concentration range from 0.0 mm to
about 4.2 mm. In order to reduce the lag phase of the reaction and to in-
crease the linear portion of the rate versus [substrate] plot, the kinetic ex-
periments were performed by adding a constant amount (1 mm) of l-
Dopa as the first reagent. The rate data for 3,5-[D2]-l-TyrOMe were cor-
rected for 90% isotopic enrichment. The temperature dependence of the
rate was studied in the range 9.0–28.08C for tyramine and in the range
8.0–28.08C for l-TyrOMe. The rates in units of DAsꢁ1 were calculated
from the slopes of the traces at 476 nm in the time interval over which
the greatest slope of the curve was observed. The reactions were fol-
lowed for 12 min.
1
(yield 90%). The H NMR spectrum of the product, recorded in D2O on
a Bruker AVANCE 400 spectrometer, showed that the esterification had
occurred with a partial loss of isotopic enrichment, to a final 90.1 atom%
D.
To determine the exact pH of the mixed aqueous/organic solvent, a glass
electrode was calibrated by the addition of measured quantities of a stan-
dard perchloric acid solution to a 34.4% methanol–glycerol (7:1, v/v) and
65.6% (v/v) water solution at the desired temperature according to
Granꢁs method.[59] The temperature dependence of the buffer pH was
found to be very small; for instance, between the temperatures of 25 and
108C the pH changed from 6.74 to 6.96. Therefore, the pH changes with
temperature were neglected.
Enzyme preparation: Mushroom tyrosinase (3960 unitsmgꢁ1) was ob-
tained from Sigma and was purified according to the procedure of Duck-
worth and Coleman.[60] To rule out the presence of any enzymatic activity
not related to tyrosinase, a 10% native-PAGE was run and then split
into two parts that were respectively stained with Coomassie Brilliant
Blue, to identify proteins, and with a substrate to highlight enzymatic ac-
tivity. The reaction mixture for the zymogram had the following composi-
tion: 1 mm l-Dopa, 8 mm MBTH, 4% DMF in 100 mm phosphate buffer
at pH 7.0, and the staining reaction was carried out for 10 min at 378C.
This assay showed that the small amounts of protein impurities were
completely inactive in l-Dopa oxidation. The enzyme concentration was
calculated from the reported activity of the purified enzyme in the l-tyro-
sine oxidation assay[23] and assuming a molecular mass of 120 kDa.[19]
The rate dependence on [Hpp] was determined in the concentration
range from 0.0 mm to 9.0 mm. The temperature dependence of the rate
was studied in the range 8.0–25.08C. The enzymatic oxidation of this sub-
strate gave a relatively stable o-quinone and enzyme activity was deter-
mined by monitoring the absorption of the o-quinone at 396 nm (e =
1200mꢁ1 cmꢁ1). At temperatures above 258C, however, the transforma-
tion of the quinone into the dopachrome became non-negligible, prevent-
ing an accurate determination of the rate data. In order to reduce the lag
time and to increase the linear portion of the curve, the kinetic experi-
ments were performed by adding a constant amount (2 mm) of l-Dopa as
the first reagent. The rates in units of DAsꢁ1 were calculated from the
slopes of the traces at 396 nm over the appropriate time intervals, usually
in the range 600–800 s. The kinetic parameters kcat and KM were obtained
by fitting of the rate vs. [substrate] plots.
The tyrosinase activity in the oxidation of N-acetyl-l-tyrosine was very
low; therefore, for this substrate the determination of kcat and KM in the
usual temperature range was not possible.
Kinetics: The catalytic oxidation of l-Dopa, dopamine, tyramine, l-
TyrOMe, 3,5-[D2]-l-TyrOMe, Hpp, and N-acetyl-l-tyrosine was studied
in the mixed solvent of 34.4% methanol/glycerol (7:1, v/v) and 65.6%
(v/v) aqueous 50 mm Hepes buffer at pH 6.8, saturated with atmospheric
oxygen. The kinetic experiments were followed spectrophotometrically
using a magnetically stirred, thermostatted, 1-cm pathlength cell and an
HP8452A diode-array spectrophotometer for temperatures above 88C.
The experiments at lower temperatures were performed using a custom-
designed immersible fiber-optic quartz probe (5 mm pathlength, Hellma).
The temperatures were controlled to within ꢀ0.18C using a Haake cryo-
stat with circulating fluid and directly checked inside the solution with a
thermometer with a precision of ꢀ0.18C.
The formation of the dopachrome derivatives of l-Dopa, dopamine, tyra-
mine, l-TyrOMe, and 3,5-[D2]-l-TyrOMe was followed through the in-
crease of the characteristic dopachrome absorption band at 476 nm. The
extinction coefficients of the various dopachrome derivatives were as-
sumed to be equal to that of l-Dopa, that is, 3600mꢁ1 cmꢁ1; the effect of
the cryosolvent on the intensity and position of the absorption maxima
of the dopachrome derivatives was taken to be negligible. To decrease
the effect of noise on the absorbance reading during the kinetic determi-
NMR experiments: The NMR T1 relaxation times for the protons of
kojic acid in the presence of variable amounts of tyrosinase were deter-
mined at 258C by the standard inversion recovery method.[61] A solution
of kojic acid was prepared in deuterated 50 mm sodium phosphate buffer
at pD 6.8. In order to eliminate interferences by traces of metal ion im-
purities, a small amount of EDTA was added to the solution. Tyrosinase
samples were prepared by dissolving the lyophilized enzyme in 50 mm
sodium phosphate buffer at pH 6.8. The enzyme solution was dialyzed
against 1 mm EDTA in 50 mm sodium phosphate buffer at pH 6.8 and
then against milliQ water. The enzyme was then lyophilized and redis-
solved in D2O. The tyrosinase concentration of this solution (0.28 mm)
was calculated from its activity as described above. The concentrations
employed in the experiments were as follows: [kojic acid] 10.0 mm, [tyro-
sinase] from 0 to 12.5 mm. The paramagnetic relaxation rate of the pro-
tons of kojic acid interacting with tyrosinase, T1b, was calculated from the
experimental relaxation rate, T1obs, using Equation (10),[32]
ꢀ
ꢁ
1
T1obs
1
E0
1
1
¼
þ
ꢁ
ð10Þ
T1f S þ KD T1b T1f
Chem. Eur. J. 2006, 12, 2504 – 2514
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2513

L. Casella et al.
where T1f is the T1 value for free inhibitor, E0 and S are the enzyme and
kojic acid concentrations, respectively, and KD, assumed to be equal to
the inhibition constant KI = 30 mm,[31] is the dissociation constant for the
tyrosinase-kojic acid complex.
[27] K.-G. Paul, P.-I. Ohlsson, Acta Chem. Scand. 1978, B32, 395–404.
[28] S. Modi, D. V. Behere, S. Mitra, Biochim. Biophys. Acta 1989, 996,
214–225.
[29] D. E. Wilcox, A. G. Porras, Y. T. Hwang, K. Lerch, M. E. Winkler,
E. I. Solomon, J. Am. Chem. Soc. 1985, 107, 4015–4027.
[30] J. R. Ros, J. N. Rodriguez-Lopez, F. Garcia-Canovas, Biochim. Bio-
phys. Acta 1994, 1204, 33–42.
[31] J. S. Chen, C. I. Wei, R. S. Rolle, W. S. Otwell, M. O. Balaban, M. R.
Marshall, J. Agric. Food Chem. 1991, 39, 1396–1401.
[32] L. Banci, I. Bertini, C. Luchinat, Nuclear and Electron Relaxation.
The Magnetic Nucleus-Unpaired Electron Coupling in Solution,
VCH, Weinheim, 1991.
[33] A. Zaks, A. M. Klibanov, J. Biol. Chem. 1988, 263, 8017–8021.
[34] G. M. Jacobsohn, R. Iskander, M. K. Jacobsohn, Biochim. Biophys.
Acta 1993, 1202, 317–324.
Acknowledgements
This work was supported by the Italian MIUR, through a PRIN project
(Progetto di Rilevante Interesse Nazionale), and by the University of
Pavia. The support and sponsorship provided by COST Action D21
“Metalloenzymes and Chemical Biomimetics” is kindly acknowledged.
The C.I.R.C.M.S.B. is gratefully acknowledged for a fellowship to A.G.
[35] S. Kermasha, H. Bao, B. Bisakowski, J. Mol. Catal. B 2001, 11, 929–
938.
[36] C. Olivares, J. C. Garcia-Borròn, F. Solano, Biochemistry 2002, 41,
679–686.
[37] P. E. M. Siegbahn, J. Biol. Inorg. Chem. 2004, 9, 577–590.
[38] A. Granata, E. Monzani, L. Casella, J. Biol. Inorg. Chem. 2004, 9,
903–913.
[39] S. Yamazaki, S. Itoh, J. Am. Chem. Soc. 2003, 125, 13034–13035.
[40] R. P. Ferrari, E. Laurenti, E. M. Ghibaudi, L. Casella, J. Inorg. Bio-
chem. 1997, 68, 61–69.
[1] H. S. Mason, Annu. Rev. Biochem. 1965, 34, 595–634.
[2] J. N. Rodriguez-Lopez, L. G. Fenoll, M. J. PeÇalver, P. A. Garcia-
Ruiz, R. Varon, F. Martinez-Ortiz, F. Garcia-Canovas, J. Tudela,
Biochim. Biophys. Acta 2001, 1548, 238–256.
[3] E. J. Land, C. A. Ramsden, P. A. Riley, Acc. Chem. Res. 2003, 36,
300–308.
[4] E. I. Solomon, U. M. Sundaram, T. E. Machonkin, Chem. Rev. 1996,
96, 2563–2605.
[5] E. Jaenicke, H. Decker, ChemBioChem 2004, 5, 163–169.
[6] E. I. Solomon, P. Chen, M. Metz, S.-K. Lee, A. E. Palmer, Angew.
Chem. 2001, 113, 4702–4724; Angew. Chem. Int. Ed. 2001, 40, 4570–
4590.
[7] K. Magnus, B. Hazes, H. Ton-That, C. Bonaventura, J. Bonaventura,
W. Hol, Proteins 1994, 19, 302–309.
[8] M. E. Cuff, C. Miller, K. E. van Holde, W. A. Hendrickson, J. Mol.
Biol. 2000, 278, 855–870.
[9] T. Klabunde, C. Eicken, J. C. Sacchettini, B. Krebs, Nat. Struct. Biol.
1998, 5, 1084–1090.
[10] L. Bubacco, J. Salgado, E. Vijgenboom, G. W. Canters, FEBS Lett.
1999, 442, 215–222.
[11] H. Decker, T. Rimke, J. Biol. Chem. 1998, 273, 25889–25892.
[12] H. Decker, M. Ryan, E. Jaenicke, N. Terwilliger, J. Biol. Chem.
2001, 276, 17796–17799.
[13] C. Gerdemann, C. Eicken, B. Krebs, Acc. Chem. Res. 2000, 33, 183–
191.
[14] H. Decker, R. Dillinger, F. Z. Tuczek, Angew. Chem. 2000, 112,
1656–1660; Angew. Chem. Int. Ed. 2000, 39, 1591–1595.
[15] P. E. M. Siegbahn, J. Biol. Inorg. Chem. 2003, 8, 567–576.
[16] L. Bubacco, M. van Gastel, E. J. J. Groenen, E. Vijgenboom, G. W.
Canters, J. Biol. Chem. 2003, 278, 7381–7389.
[17] A. W. J. W. Tepper, L. Bubacco, G. W. Canters, J. Am. Chem. Soc.
2005, 127, 567–575.
[18] T. Plenge, R. Dillinger, L. Santagostini, L. Casella, F. Z. Tuczek, Z.
Anorg. Allg. Chem. 2003, 629, 2258–2265.
[19] J. C. Espin, R. Varon, L. G. Fenoll, M. A. Gilabert, P. A. Garcia-
Ruiz, J. Tudela, F. Garcia-Canovas, Eur. J. Biochem. 2000, 267,
1270–1279.
[20] L. G. Fenoll, M. J. PeÇalver, J. N. Rodriguez-Lopez, P. A. Garcia-
Ruiz, F. Garcia-Canovas, J. Tudela, Biochem. J. 2004, 380, 643–650.
[21] J. S. Conrad, S. R. Dawso, E. R. Hubbard, T. E. Meyers, K. G.
Strothkamp, Biochemistry 1994, 33, 5739–5744.
[41] M. B. Smith, J. March, Marchꢀs Advanced Organic Chemistry, 5th
ed., Wiley, New York, 2001, Chapter 11.
[42] L. P. Hammett, Physical Organic Chemistry, 2nd ed., McGraw-Hill,
New York, 1970, p. 172.
[43] L. G. Fenoll, M. J. PeÇalver, J. N. Rodriguez-Lopez, P. A. Garcia-
Ruiz, F. Garcia-Canovas, J. Tudela, Biochem. J. 2004, 380, 643–650.
[44] L. M. Mirica, M. Vance, D. J. Rudd, B. Hedman, K. O. Hodgson,
E. I. Solomon, T. D. P. Stack, Science 2005, 308, 1890–1892.
[45] K. D. Karlin, A. D. Zuberbühler, Bioinorganic Catalysis, 2nd ed.
(Eds.: J. Reedijk, E. Bowman), Dekker, New York, 1999, pp. 469–
534.
[46] G. Battaini, E. Monzani, A. Perotti, C. Para, L. Casella, L. Santagos-
tini, M. Gullotti, R. Dillinger, C. Näther, F. Tuczek, J. Am. Chem.
Soc. 2003, 125, 4185–4198.
[47] S. Palavicini, A. Granata, E. Monzani, L. Casella, J. Am. Chem.
Soc., in press.
[48] G. Battaini, E. Monzani, L. Casella, E. Lonardi, A. W. J. W. Tepper,
G. W. Canters, L. Bubacco, J. Biol. Chem. 2002, 277, 44606–44612.
[49] B. W. Griffin, J. A. Peterson, Biochemistry 1972, 11, 4740–4746.
[50] K. Ruckpaul, H. Rein, J. Blanck, Front. Biotransform. 1989, 1, 1–65.
[51] D. F. V. Lewis, P. J. Eddershaw, M. Dickins, M. H. Tarbit, P. S. Gold-
farb, Chem.-Biol. Interact. 1998, 115, 175–199.
[52] H. Voss, K. J. Wannovius, H. Elias, Inorg. Chem. 1979, 18, 1454–
1459.
[53] F. Thaler, C. D. Hubbard, F. W. Heinemann, R. van Eldik, S. Schin-
dler, I. Fabian, A. M. Dittler-Klingemann, F. E. Hahn, C. Orvig,
Inorg. Chem. 1998, 37, 4022–4029.
[54] A. Neubrand, F. Thaler, M. Kçrner, A. Zahl, C. D. Hubbard, R. va-
n Eldik, J. Chem. Soc. Dalton Trans. 2002, 957–961.
[55] W. R. Cannon, S. J. Benkovic, J. Biol. Chem. 1998, 273, 26257–
26260.
[56] A. Warshel, J. Biol. Chem. 1998, 273, 27035–27038.
[57] J. C. Espin, P. A. Garcia-Ruiz, J. Tudela, F. Garcia-Canovas, Bio-
chem. J. 1998, 331, 547–551.
[22] V. RØat, J. L. Finney, A. Steer, M. A. Roberts, J. Smith, R. Dunn, M.
Peterson, R. Daniel, J. Biochem. Biophys. Methods 2000, 42, 97–103.
[23] A. Rescigno, E. Sanjust, G. Soddu, A. C. Rinaldi, F. Sollai, N. Curre-
li, A. Rinaldi, Biochim. Biophys. Acta 1998, 1384, 268–276.
[24] K. J. Liu, M. W. Grinstaff, J. Jiang, K. S. Suslick, H. M. Swartz, W.
Wang, Biophys. J. 1994, 67, 896–901.
[25] J. N. Rodriguez-Lopez, L. G. Fenoll, P. A. Garcia-Ruiz, R. Varon, J.
Tudela, R. N. F. Thorneley, F. Garcia-Canovas, Biochemistry 2000,
39, 10497–10506.
ˇ
[58] J. Villà, M. Strajbl, T. M. Glennon, Y. Y. Sham, Z. T. Chu, A. War-
shel, Proc. Natl. Acad. Sci. USA 2000, 97, 11899–11904.
[59] F. J. C. Rossotti, H. J. Rossotti, J. Chem. Educ. 1965, 42, 375–378.
[60] H. W. Duckworth, J. E. Coleman, J. Biol. Chem. 1970, 245, 1613–
1625.
[61] R. L. Vold, J. S. Waugh, M. P. Klein, D. E. Phelps, J. Chem. Phys.
1968, 48, 3831–3832.
[26] S. Yamazaki, C. Morioka, S. Itoh, Biochemistry 2004, 43, 11546–
11553.
Received: September 5, 2005
Published online: December 8, 2005
2514
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 2504 – 2514