Trapping tyrosinase key active intermediate under turnover

Article abstract of DOI:10.1039/b911946a

This paper shows for the first time that the spectral features of the ternary complex of tyrosinase/O₂/phenol, trapped at low temperature using the very slow substrate 3,5-difluorophenol, are those of a μ-η²:η²-peroxidodic

Full text of DOI:10.1039/b911946a

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/dalton | Dalton Transactions
Trapping tyrosinase key active intermediate under turnover†‡
Alessia Spada,a Sara Palavicini,a Enrico Monzani,a Luigi Bubaccob and Luigi Casella*a
Received 10th June 2009, Accepted 26th June 2009
First published as an Advance Article on the web 15th July 2009
DOI: 10.1039/b911946a
2
2
This paper shows for the first time that the spectral features of
the ternary complex of tyrosinase/O2/phenol, trapped at low
temperature using the very slow substrate 3,5-difluorophenol,
are those of a l–g :g -peroxidodicopper(II) species, and that
this remains the only enzyme species under turnover and
substrate saturation conditions.
complex [Cu2(m–h :h -O2)(DBED)]2+ (DBED = N,N¢-di-tert-
butyl-ethylenediamine) binding of a phenolate ligand at very low
temperature induced an isomerisation to the corresponding bis(m-
oxido)dicopper(III) species [Cu2(m-O)2(DBED)]2+ prior to the
O-atom transfer reaction. Another possibility exists, in which
binding of the phenol to oxyTy may occur with proton transfer to
the electron-rich peroxido moiety, which would be converted into
a hydroperoxido group (bound as a bridging or terminal ligand
to the coppers).7 It is generally accepted in biomimetic copper
2
2
Tyrosinase (EC 1.14.18.1, Ty) is a dinuclear copper enzyme
catalyzing the hydroxylation of monophenols and the oxidation of
ortho-diphenols to ortho-quinones;1 it is the only enzyme known to
be able to catalyze both of these reactions. Tyrosinase is found in
bacteria, plants and animals, where it is involved in the formation
chemistry, that both the m–h :h -peroxidodicopper(II)8a–d and
bis(m-oxido)dicopper(III) complexes6,9 are capable of performing
the hydroxylation of a phenolate ligand. In this report, we
prove that the ternary complex of Ty/O2/phenol, trapped at low
temperature and followed during initial turnover, maintains all
the spectroscopic characteristics of oxyTy, and hence that the
peroxidodicopper(II) species is the enzymatically competent active
species in the monophenolase reaction.
2
2
of melanin pigments and other polyphenolic compounds.1b,c
A
recent X-ray structure determination of various forms of the
enzyme from Streptomyces castaneoglobisporus showed that each
copper ion is bound to three highly conserved histidines residues.2
The region around these Cu-binding ligands is also well conserved
in Tys from different species, as well as in hemocyanins and
catechol oxidases.3 Native tyrosinase is predominantly in the met
form, with CuII ions bridged by a small ligand, likely OH-.2
Upon reduction, metTy is converted to the deoxy-form, which can
To succeed in the attempt of kinetically trapping the elusive
ternary complex of the enzyme, two requirements need to be
fulfilled: (i) to reach a sufficiently low temperature and (ii) to
find a slowly reacting substrate, such that even using high enzyme
concentration, the detection of the active enzyme intermediate
is possible. In fact, it is well known that tyrosinase undergoes
extremely rapid inactivation by the quinone species generated
upon oxidation of phenolic substrates.10 In this investigation,
we used tyrosinase from Streptomyces antibioticus, which bears
82% identity with the structurally characterized enzyme from S.
castaneoglobisporus and can be obtained in excellent purity,11 and
studied its activity 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,
the cryosolvent previously used for variable temperature studies of
the fungal enzyme.5b Preliminarily, we studied some representative
natural substrates to show that the enzyme works normally in the
mixed aqueous-organic solvent at low temperature. The activation
parameters and substrate binding data, deduced from kcat and KM,
respectively, are comparable to those obtained previously for the
fungal enzyme5b (Table 1), showing the substantial analogy in the
behavior of the enzymes from bacterial and fungal sources, in
2
2
readily bind dioxygen in a m–h :h coordination mode producing
oxyTy, the only form supporting the monophenolase reaction.
This species exhibits characteristic LMCT features at 345 and
590 nm, and a very low vibrational frequency at 750 cm-1 in
the Raman spectrum, indicative of a peroxido ligand.1a The most
challenging step of the monophenolase cycle is the insertion of
an O atom into the C–H bond ortho to the phenol C-OH group
by oxyTy. Cleavage of the peroxidic O–O bond could take place
before the attack on the aromatic ring (a, Fig. 1), after that (b), or
concerted with it (c).4
Fig. 1 Phenol hydroxylation mechanism.
Table 1 Activation parameters and substrate binding data for Strepto-
myces antibioticus 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;
L-TyrOMe is L-tyrosine methyl ester
In general, one of the last two possibilities is assumed,5 but a
recent report by Mirica et al.6 showed that in the model m-peroxido
DHπ/kJ
DSπ/J K-1
DH◦/kJ
DS◦/J K-1
aDipartimento di Chimica Generale, Universita` di Pavia, viale Taramelli 12,
27100, Pavia, Italy. E-mail: bioinorg@unipv.it; Fax: +39 0382528544
Substrate
mol-1
mol-1
mol-1
mol-1
bDipartimento di Biologia, Universita` di Padova, via U. Bassi 58B, 35121,
Padova, Italy
Dopamine
L-Dopa
L-TyrOMe
Tyramine
+64.8 2.6
+59.5 3.4
+59.3 2.9
+67.6 3.2
-3
9
-75.1 2.7
-37.9 2.3
-45.6 1.6
-42.0 1.7
-198
-96
9
8
5
6
+8 11
+8 10
-13 11
† Dedicated to Professor Renato Ugo on the occasion of his 71st birthday.
‡ Electronic supplementary information (ESI) available: Additional exper-
imental details. See DOI: 10.1039/b911946a
-106
-77
6468 | Dalton Trans., 2009, 6468–6471
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
spite of the large differences in size and structural organization.1c,2
enzymatic oxidation of L-Dopa was investigated. The presence
of 3,5-difluorophenol bears a reduction in kcat and an increase in
KM for L-Dopa oxidation in a concentration dependent fashion,
indicating an inhibition of mixed type (ESI†). This behavior
can be accounted for by considering that 3,5-difluorophenol can
bind to the active site, competing with L-Dopa, but at the same
time, it can react with oxyTy, partly consuming this species, and
may also interact with the enzyme at non-specific sites, probably
on the protein surface. The competitive inhibition constant (KI)
deduced from the analysis of the kinetic data of L-Dopa oxidation
is 32 mM. This constant is important because it enables an
estimation of the 3,5-difluorophenol concentration required to
saturate oxyTy.
The activation free energy (DGπ = 62.5
5.5 kJ mol-1 for
S. antibioticus Ty, 64.4 3.7 kJ mol-1 for fungal Ty,5b at 25 ◦C)
is always dominated by the enthalpic term, with DHπ values in
the range of 63 8 kJ mol-1 for S. antibioticus Ty, irrespective of
substrate and reaction type (monophenolase or catecholase), while
the activation entropies are small (still smaller for the bacterial Ty),
indicative of a very high degree of preorganization of the enzyme
active site.5b
As phenol hydroxylation by tyrosinase occurs through an
electrophilic aromatic substitution,12a and we previously showed
that a fluorine atom on the phenol ring significantly depresses the
rate of enzymatic oxidation,12b we thought that a difluorophenol
could be a suitably “slow” substrate to enable the observation of
the enzyme species during turnover, even in conditions of large
enzyme concentration and substrate saturation. The reactivity of
both S. antibioticus Ty and mushroom Ty with 3,5-difluorophenol
is indeed extremely weak, so that to show that this phenol is a
substrate of the enzyme rather high enzyme concentration had
to be used. On the other hand, as with monofluorophenols,12b
the primary 3,5-difluoroquinone product is an extremely reactive
electrophile towards excess phenol present in solution, which
rapidly gives rise to a mixture of complex oligomeric products,
preventing its isolation. The quinone intermediate was therefore
trapped using glutathione (GSH), a tripeptide containing a
cysteine thiol group very reactive with the quinone. Two quinone
addition compounds were isolated from the enzymatic reaction,
and their characterization (by MS, 19F-NMR, 1H -NMR and
1H-DQF-COSY spectra, electronic supplementary information,
ESI†) showed they consist of the isomeric adducts 3 and 4,
containing two GSH residues bound to the quinone moiety
(Fig. 2).
The direct observation of the oxyTy active form during turnover
at low temperature with conventional spectrophotometry is diffi-
cult. The experiment requires enzyme at least in the 0.03 mM
concentration range, for appropriate monitoring the evolution
of the oxyTy optical band, but the organic component of the
cryosolvent limits the solubility of the protein. Therefore, a
minimum amount of this component allowing the mixed solvent
to reach temperatures where the enzymatic reaction is not too
fast had to be found, otherwise the absorption of the oxidation
products would rapidly prevent monitoring of the enzyme. We
found that a cryosolvent with a 55% (v/v) content of organic
component (7:4 ratio, v/v, of methanol:glycerol) and Hepes buffer
provides sufficient solubility for the enzyme and allowed us
to reach temperatures slightly below -30 ◦C, maintaining the
solution sufficiently fluid to enable rapid mixing of reagents.
In these conditions, the characteristic near-UV band of oxyTy
at 345 nm could be easily generated by anaerobic reduction of
metTy (35 mM) with a small excess of hydroxylamine,13 followed
by exposure to air oxygen. When a conventional substrate, such
as L-tyrosine or tyramine, is added to this enzyme solution
at -30 ◦C, an extremely fast reaction takes place, producing
immediate darkening of the solution during the mixing time.
Even upon addition of an enzyme-saturating amount of 3,5-
difluorophenol (70 mM), the reaction is rapid, but in this case,
the enzyme species could be observed during several turnovers
until, after about 60 s, the oxyTy band was depleted and became
progressively covered by the broad optical absorption of the
reaction products (Fig. 3). It is clear from this experiment that
the oxyTy species observed immediately after the addition of
2
2
saturating substrate is still the m–h :h -peroxidodicopper(II) form,
and therefore phenol binding does not induce an isomerisation
to the bis(m-oxido)dicopper(III) form, as observed in the model
system described by Mirica and coworkers.6 A change to the
latter species would be signalled by a marked shift of the LMCT
band above 410 nm.14 In addition, phenol binding to oxyTy
is accompanied by the appearance of an absorption feature at
460 nm (De ~ 1000 M-1 cm-1), which is consistent with the expected
LMCT band typical of the copper(II)-phenolate chromophore.15
By contrast, we have recently shown that binding of phenolate
to the bis(m-oxido)dicopper(III) core of the biomimetic complex
[Cu2(m-O)2(m-XYLMeAN)]2+ is accompanied by development of an
LMCT band at 560 nm.9 The oxyTy band remains well defined and
basically unshifted for about 20 s reaction time, the progressive
reduction in intensity being due to consumption of dioxygen in
solution and, possibly, by some enzyme inactivation by the reactive
quinone compounds produced in the reaction. In parallel, also the
Fig. 2 Trapping the reactive 3,5-difluoroquinone primary product by
glutathione.
These products can be accounted for by the extreme sus-
ceptibility of 3,5-difluoroquinone 1 to nucleophilic fluorine
substitution by GSH; this reaction is regiospecific and yields
the monofluoro-quinone adduct 2, which undergoes a Michael
addition reaction producing the two isomeric bis-GSH adducts
3 and 4 in comparable amounts. Release of inorganic fluoride
upon GSH addition to compound 1 was confirmed by means of
19F-NMR.
The extremely low reactivity of 3,5-difluorophenol makes it
nearly impossible to carry out substrate-dependent kinetic studies,
for the large quantity of enzyme that would be required. Therefore,
to demonstrate that 3,5-difluorophenol is a “normal”, albeit
slowly reacting, substrate for tyrosinase, its competition in the
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 6468–6471 | 6469
©

View Article Online
cleavage represents the rate determining step of the enzymatic
reaction. The O–O bond cleavage occurs in the transition state, not
in the ground state ternary complex of Cu2/O2/phenol, as it occurs
in the Cu-DBED model system.6 This is confirmed by the striking
difference in the activation parameters for the enzymatic and
model reactions,8d,9 which may reflect the (evolutionary optimised)
stabilization of the transition state by the enzymatic system. Even
though the intrinsic barrier for the interconversion between the
m-peroxido and bis(m-oxido) Cu2O2 cores appears to be small
on theoretical grounds, and is therefore observed in a number
of low-molecular weight biomimetic copper systems,14,18 such a
change may involve considerable conformational change in the
active site of the protein, to meet the necessary requirement of a
strong tetragonal field by the copper(III) species.
Our results can neither be reconciled with the recent quantum
chemical study of the reaction mechanism of S. castaneoglobis-
porus tyrosinase,17b according to which binding of the phenolic
2
2
substrate to oxyTy occurs with initial isomerization of the m–h :h -
1
2
peroxidodicopper(II) species to a m–h :h -peroxidodicopper(II)
form, followed by proton transfer from the phenol to the peroxo
group and phenolate binding to CuB. There is only one prece-
1
2
dent of m–h :h -peroxidodicopper(II) complex in the literature,
reported by Tachi et al.,19 and this bears spectroscopic signatures
2
2
quite different from that of the m–h :h peroxido isomer, with an
LMCT band strongly red shifted to 483 nm. Such a spectral change
is clearly not detected in our experiment, and therefore we believe
that no isomerization of oxyTy occurs upon formation of the
1
2
enzyme ternary complex. The evolution of the theoretical m–h :h -
hydroperoxidodicopper(II) species along the reaction coordinate
involves one-electron oxidation of the bound phenolate with
production of a phenoxyl radical.17b This feature is common to
other previous theoretical studies17a and deserves more general
consideration on alternative formulations of the tyrosinase mech-
Fig. 3 Monitoring tyrosinase Cu2O2 intermediate during turnover at
low temperature. (a) UV-Vis spectra recorded at -30 ◦C in the initial
phase of the reaction of S. antibioticus oxytyrosinase (35 mM) and
3,5-difluorophenol (70 mM) in the mixed solvent of 35% methanol,
20% glycerol and 45% 50 mM Hepes buffer pH 6.8 (v/v/v). OxyTy
was generated at -20 ◦C from the met form by reaction with 10 equiv.
NH2OH under anaerobic conditions, followed by exposure to dioxygen.
The spectra are difference spectra with respect to deoxyTy (with oxyTy
shown as lower trace), and were taken at different times after the addition
of 3,5-difluorophenol (the spectrum at ~0 s time being the first one obtained
after mixing). (b) The panel shows an expansion of the low-energy spectral
region. Absolute spectra of deoxyTy, oxyTy, and enzyme under turnover
are given in the ESI.†
2
2
anism functioning through intermediates different from m–h :h -
peroxidodicopper(II). Also the bis(m-oxido)dicopper(III) species,
in fact, exhibits strong tendency to induce radical reactions,14,18a,20
and indeed the formation of a 1:1 mixture of catechol/quinone,
from the phenolate, in the Cu-DBED system reported by Mir-
ica and co-workers6 likely reflects the initial formation of a
semiquinone radical, followed by disproportionation. In contrast,
we proved some time ago that no radical species is formed in
tyrosinase reactions,21 and found no trace of the characteristic
sharp optical feature of the phenoxyl radical18b,22 in our present low
temperature experiments. As it is well known by the work of Itoh’s
group, the enzymatic phenol hydroxylation proceeds through an
electrophilic aromatic substitution,12a and the reactivity of the m–
LMCT band loses definition and the absorption throughout the
near-UV range increases due to the formation of the oligomeric
addition products of the initially formed 3,5-difluoroquinone with
excess difluorophenol.
Unfortunately attempts to detect the signature of oxyTy in
Raman experiments, around 700–800 cm-1, failed due to excessive
fluorescence of the sample,16a which is likely stronger here than in
the previously investigated enzyme from Neurospora crassa.16b The
experiments were carried out both on frozen and fluid solutions
(see ESI†), with excitation wavelengths of 370, 380, 390, and
410 nm in a triple subtractive mode. Fluorescence by the sample
was already high at 410 nm excitation and increased exponentially
at the lower wavelengths.
2
2
h :h -peroxidodicopper(II) intermediate fully conforms to this
mechanism.
In conclusion, we have shown by a cryoenzymatic study
2
2
that tyrosinase functions through a m–h :h -peroxidodicopper(II)
active species, that remains the only detectable species dur-
ing turnover under substrate saturation conditions. The bis(m-
oxido)dimetal core, with its powerful oxidizing properties, is rele-
vant for dinuclear iron enzymes, like methane monooxygenase,20
where radical chemistry is necessary, but is not observed with
tyrosinase. Though, it bears strong interest for copper mediated
oxidation processes20,23 and remains a potential candidate for
other Cu-dependent biochemical systems promoting oxidative
reactions.
The observation of the ternary complex of oxyTy with a bound
phenolate during enzyme turnover, and the kinetic data at variable
temperature, confirm our previous hypothesis,5b which appears to
be supported by theoretical calculations,17 that peroxide O–O bond
6470 | Dalton Trans., 2009, 6468–6471
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
10 W. H. Vanneste and A. Zuberbu¨hler, in Molecular Mechanisms of
Oxygen Activation, ed. O. Hayashi, Academic press, New York, 1974,
pp. 371–404.
11 L. Bubacco, E. Vijenboom, C. Gobin, A. W. Tepper, J. Salgado and
G. W. Canters, J. Mol. Catal. B, 2000, 8, 27–35.
12 (a) S. Yamazaki and S. Itoh, J. Am. Chem. Soc., 2003, 125, 13034–13035;
(b) G. Battaini, E. Monzani, L. Casella, E. Lonardi, A. W. J. W. Tepper,
G. W. Canters and L. Bubacco, J. Biol. Chem., 2002, 277, 44606–44612.
13 A. W. Tepper, L. Bubacco and G. W. Canters, J. Biol. Chem., 2002, 277,
30436–30444.
Acknowledgements
We thank G. Zoppellaro and K. K. Andersson (University of Oslo)
for their repeated efforts to obtain resonance Raman spectra of
oxy-tyrosinase and for helpful discussions, and S. Fogal and P. Di
Muro (University of Padova) for their technical help.
Notes and references
14 L. M. Mirica, X. Ottenwaelder and T. D. P. Stack, Chem. Rev., 2004,
104, 1013–1045.
1 (a) E. I. Solomon, U. M. Sundaram and T. E. Machonkin, Chem. Rev.,
1996, 96, 2563–2605; (b) N. Wang and D. N. Hebert, Pigm. Cell Res,
2006, 19, 3–18; (c) S. Halaouli, M. Asther, J. C. Sigoillot, M. Hamdi
and A. Lomascolo, J. Appl. Microbiol., 2006, 100, 219–232.
2 Y. Matoba, T. Kumagai, A. Yamamoto, H. Yoshitsu and M. Sugiyama,
J. Biol. Chem., 2006, 281, 8981–8990.
3 C. M. Marusek, N. M. Trobaugh, W. H. Flurkey and J. K. Inlow,
J. Inorg. Biochem., 2006, 100, 108–123.
4 H. Decker, R. Dillinger and F. Tuczek, Angew. Chem. Int. Ed., 2000,
39, 1591–1595.
5 (a) S. Itoh and S. Fukuzumi, Acc. Chem. Res., 2007, 40, 592–600; (b) A.
Granata, E. Monzani, L. Bubacco and L. Casella, Chem. Eur. J., 2006,
12, 2504–2514.
6 L. M. Mirica, M. Vance, D. J. Rudd, B. Hedman, K. O. Hodgson, E. I.
Solomon and T. D. P. Stack, Science, 2005, 308, 1890–1892.
7 J. S. Conrad, S. R. Dawso, E. R. Hubbard, T. E. Meyers and K. G.
Strothkamp, Biochemistry, 1994, 33, 5739–5744.
8 (a) L. Santagostini, M. Gullotti, E. Monzani, L. Casella, R. Dillinger
and F. Tuczek, Chem. Eur. J., 2000, 6, 519–522; (b) S. Itoh, H. Kumei, M.
Taki, S. Nagatomo, T. Kitagawa and S. Fukuzumi, J. Am. Chem. Soc.,
2001, 123, 6708–6709; (c) G. Battaini, M. De, Carolis, E. Monzani,
F. Tuczek and L. Casella, Chem. Commun., 2003, 726–727; (d) S.
Palavicini, A. Granata, E. Monzani and L. Casella, J. Am. Chem. Soc.,
2005, 127, 18031–18036.
9 A. Company, S. Palavicini, I. Garcia-Bosch, R. Mas-Balleste´, L. Que,
Jr., E. V. Rybak-Akimova, L. Casella, X. Ribas and M. Costas, Chem.
Eur. J., 2008, 14, 3535–3538.
15 (a) J. A. Halfen, B. A. Jazdzewski, S. Mahapatra, L. M. Berreau, E. C.
Wilkinson, L. Que, Jr. and W. B. Tolman, J. Am. Chem. Soc., 1997,
119, 8217–8227; (b) R. C. Holz, J. M. Bradshaw and B. Bennett, Inorg.
Chem., 1998, 37, 1219–1225; (c) C. Belle, C. Beguin, I. Gautier-Luneau,
S. Hamman, C. Philouze, J. L. Pierre, F. Thomas and S. Torelli, Inorg.
Chem., 2002, 41, 479–491.
16 (a) A. W. J. W. Tepper, L. Bubacco and G. W. Canters, J. Biol. Chem.,
2004, 279, 13425–13434; (b) N. C. Eickman, E. I. Solomon, J. A.
Larrabee, T. G. Spiro and K. Lerch, J. Am. Chem. Soc., 1978, 100,
6529–6531.
17 (a) P. E. M. Siegbahn, J. Biol. Inorg. Chem., 2003, 8, 567–576;
(b) T. Inoue, Y. Shiota and K. Yoshizawa, J. Am. Chem. Soc., 2008,
130, 16890–16897.
18 (a) E. A. Lewis and W. B. Tolman, Chem. Rev., 2004, 104, 1047–1076;
(b) D. Maiti, J. S. Woertnik, A. A. Narducci Sarjeant, E. I. Solomon
and K. D. Karlin, Inorg. Chem., 2008, 47, 3787–3800.
19 Y. Tachi, S. Teramae, Y. Naruta, K. Aita, S. Itoh, F. Tani and S.
Fukuzumi, Inorg. Chem., 2004, 43, 4558–4560.
20 L. Que, Jr. and W. B. Tolman, Nature, 2008, 455, 333–340.
21 R. P. Ferrari, E. Laurenti, E. M. Ghibaudi and L. Casella, J. Inorg.
Biochem., 1997, 68, 61–69.
22 A. Sokolowski, J. Mu¨ller, T. Weyhermu¨ller, R. Schnepf, P. Hildebrandt,
K. Hildenbrand, E. Bothe and K. Wieghardt, J. Am. Chem. Soc., 1997,
119, 8889–8900.
23 T. Punniyamurthy, S. Velusamy and J. Iqbal, Chem. Rev., 2005, 105,
2329–2364.
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 6468–6471 | 6471
©