theoretical study of the catalytic mechanism of catechol
TRANSCRIPT
ORIGINAL PAPER
Theoretical study of the catalytic mechanism of catechol oxidase
Mireia Guell Æ Per E. M. Siegbahn
Received: 20 June 2007 / Accepted: 16 August 2007 / Published online: 20 September 2007
� SBIC 2007
Abstract The mechanism for the oxidation of catechol
by catechol oxidase has been studied using B3LYP hybrid
density functional theory. On the basis of the X-ray
structure of the enzyme, the molecular system investigated
includes the first-shell protein ligands of the two metal
centers as well as the second-shell ligand Cys92. The cycle
starts out with the oxidized, open-shell singlet complex
with oxidation states Cu2(II,II) with a l-g2:g2 bridging
peroxide, as suggested experimentally, which is obtained
from the oxidation of Cu2(I,I) by dioxygen. The substrate
of each half-reaction is a catechol molecule approaching
the dicopper complex: the first half-reaction involves Cu(I)
oxidation by peroxide and the second one Cu(II) reduction.
The quantitative potential energy profile of the reaction is
discussed in connection with experimental data. Since no
protons leave or enter the active site during the catalytic
cycle, no external base is required. Unlike the previous
density functional theory study, the dicopper complex has a
charge of +2.
Keywords Catechol oxidase � Copper enzymes �O2 cleavage � Hybrid density functional theory
Introduction
Proteins containing copper ions at their active sites are
usually involved as redox catalysts in a wide range of
biological processes. Type-3 active-site copper proteins
contain a dicopper core in which both copper ions are
surrounded by three nitrogen donor atoms from histidine
residues [1, 2]. A characteristic feature of the proteins with
this active site is their ability to reversibly bind dioxygen at
ambient conditions. The Cu(II) ions in the oxy state of
these proteins are strongly antiferromagnetically coupled,
leading to electron paramagnetic resonance (EPR) silent
behavior. This class is represented by three proteins:
hemocyanin, catechol oxidase and tyrosinase.
Proteins with type-3 copper centers can serve either as
oxygenase/oxidase enzymes or as dioxygen transport pro-
teins [2]. An example of an oxygen carrier is hemocyanin.
Hemocyanins can be divided into two classes depending on
their biological source: the arthropodan and the molluscan
hemocyanins [3–5].
Catechol oxidase, which is also known as o-diphenol
oxidase, catalyzes exclusively the oxidation of catechols
(i.e., o-diphenols) to the corresponding o-quinones (called
catecholase activity) (Fig. 1) [6]. The resulting highly
reactive quinones autopolymerize to form brown poly-
phenolic catechol melanins, a process thought to protect
damaged plants from pathogens or insects [7]. The rate for
catechol conversion in sweet potatoes has been measured
to be 2.3 · 103/s [8], corresponding to a rate-limiting free-
energy barrier of around 13 kcal/mol.
In contrast to catechol oxidase, the strongly related
tryrosinase shows additional monooxygenase activity
(Fig. 1). This so-called cresolase activity enables the
enzyme to accept also monophenols (like tyrosine and
cresol). Catechol oxidases are found in plant tissues and in
M. Guell (&)
Institut de Quımica Computacional,
Universitat de Girona,
Campus de Montilivi,
17071 Girona, Spain
e-mail: [email protected]
P. E. M. Siegbahn
Department of Biochemistry and Biophysics,
Stockholm University,
106 91 Stockholm, Sweden
123
J Biol Inorg Chem (2007) 12:1251–1264
DOI 10.1007/s00775-007-0293-z
some insects and crustaceans, whereas tryrosinases can be
isolated from a broader variety of plants, fungi, bacteria,
mammalians, crustaceans and insects [2]. The differentia-
tion between catechol oxidase and tryrosinase is not
rigorous, as some catechol oxidases also show weak
monooxygenase activity. However, these catechol oxidases
often do not accept tyrosine as a substrate [9, 10].
Klabunde et al. [11] isolated the first crystal structures of
the catechol oxidase from Ipomoea batatas (sweet potato)
in three catalytic states: the native met [Cu(II)Cu(II)] state,
the reduced deoxy [Cu(I)Cu(I)] form, and in the complex
with the inhibitor phenylthiourea. This enzyme was found
to be monomeric and ellipsoidal in shape (Fig. 2). Its
secondary structure is primarily a-helical with the core of
the enzyme formed by a four-helix bundle. The helical
bundle accommodates the catalytic dinuclear copper cen-
ter, where each of the two copper ions is coordinated by
three histidine residues. One of the key features of the
catechol oxidase active site is an unusual thioether bridge
between Cys92 and His109, one of the ligands of CuA
(Fig. 3). Apart from the geometrical constraints added to
the CuA site, no function of the chemistry performed by the
enzyme has been ascribed to this covalent bridge.
In the native met state, the two copper ions are 2.9 A
apart. In addition to six histidine residues, a bridging sol-
vent molecule, most likely hydroxide anion, was refined in
close proximity to the two metal centers (CuA–O 1.9 A,
CuB–O 1.8 A), completing the coordination sphere of the
copper ions to a trigonal pyramid. EPR data reveal a strong
antiferromagnetic coupling between the copper ions, in line
with a solvent molecule bridging two metal centers, as
found in the crystal structure.
Upon reduction of the Cu(II) ions to the +1 oxidation
state, the distance between them increases to 4.4 A, while
the histidine residues move only slightly, and no significant
change was observed for other residues of the protein [11].
On the basis of the residual electron density maps, a water
molecule was positioned at a distance of 2.2 A from the
CuA atom. Thus, the coordination sphere around the CuA
ion is a distorted trigonal pyramid, with three nitrogen
atoms from the histidine residues forming a basal plane,
while the coordination sphere around the CuB ion is best
described as square planar with one missing coordination
site.
When phenylthiourea binds to catechol oxidase, it
replaces the hydroxo bridge, present in the met form. The
sulfur atom of phenylthiourea is coordinated to both Cu(II)
centers, increasing the distance between them to 4.2 A. The
amide nitrogen interacts weakly with the CuB center (Cu–N
distance of 2.6 A), completing its square-pyramidal
geometry.
The oxy form of catechol oxidase can be obtained by
treating the met form of the enzyme with hydrogen per-
oxide [12].
As previously mentioned, catechol oxidase catalyzes the
oxidation of catechols to the respective quinones through a
four-electron reduction of dioxygen to water. Krebs and
coworkers proposed a mechanism for the catalytic process,
based on biochemical and spectroscopic data [2, 12, 13], as
well as structural data [11], which is depicted in Fig. 4
[14]. The catalytic cycle begins with the met form of cat-
echol oxidase, which is the resting form of the enzyme. The
dicopper(II) center of the met form reacts with 1 equivalent
of catechol, leading to the formation of quinone and to the
reduced deoxy dicopper(I) state. This step is supported by
the observation that stoichiometric amounts of the quinone
product form immediately after the addition of catechol,
even in the absence of dioxygen [11, 14]. On the basis of
the structure of catechol oxidase with the bound inhibitor
phenylthiourea, the monodentate binding of the substrate to
the CuB center has been proposed. Afterwards, dioxygen
binds to the dicopper(I) active site, replacing the solvent
molecule bonded to CuA in the reduced enzyme. UV–vis
spectroscopy and Raman data suggested that dioxygen
OH OH
OH
O
O
1/2 O2 1/2 O2 H2O
cresolase activity
(tyrosinase)
catecholase activity
(tyrosinase andcatechol oxidase)
Fig. 1 Reaction pathway of the
oxygenation and oxidation
catalyzed by tyrosinase and
catechol oxidase
Fig. 2 Tertiary structure of the oxidized catechol oxidase (PDB code
1BT3) [11]. Cu(II) ions are given as yellow spheres and important
active-site residues are shown. The picture was generated using the
VMD 1.8.5 molecular visualization program. See Fig. 3 for a more
detailed picture of the active site
1252 J Biol Inorg Chem (2007) 12:1251–1264
123
binds in a bridging side-on l-g2:g2 binding mode with a
copper–copper separation of 3.8 A, as determined by
extended X-ray absorption fine structure spectroscopy for
the oxy state [12]. The observed binding mode of phenyl-
thiourea and the modeled catechol-binding mode suggest
that simultaneous binding of catechol and dioxygen is
possible. In this model, CuB is six-coordinated with a
tetragonal planar coordination by His240, His244 and the
dioxygen molecule in the basal plane. The CuA site retains
the tetragonal pyramidal geometry with dioxygen, His88
and His118 in the equatorial positions, His109 in an axial
position and a vacant sixth coordination site. In this pro-
posed ternary catechol oxidase–O22––catechol complex, two
electrons can be transferred from the substrate to the
Fig. 3 Active site of the
oxidized catechol oxidase (PDB
code 1BT3) [11]
CuIAHis
His
His
BCuI His
His
His
H2O + H+
OH
OH
CuIIHis
His
His O
OCuII His
His
His
O
O
H+
OH
OH
O
O
H2O +
OH2
O2 +
O
HO
CuIIAHis
His
His
OH
BCuII His
His
His
O
HO
2H+
CuIIAHis
His
His
OH
BCuII His
His
His
met state
deoxy state
oxy state
A B
Fig. 4 Catalytic cycle of
catechol oxidase as proposed
by Klabunde et al. [11]
J Biol Inorg Chem (2007) 12:1251–1264 1253
123
peroxide, followed by cleavage of the O–O bond, loss of
water and the formation of the quinone product, together
with the restoration of the met state, completing the cata-
lytic cycle.
A very similar catalytic mechanism had been proposed
earlier by Solomon et al. [2] for the catecholase activity of
the structurally related type-3 protein tyrosinase (Fig. 5).
The cycle starts from the oxy and met states. A diphenol
substrate binds to the met state (for example), followed by
the oxidation of the substrate to the first quinone and the
formation of the reduced state of the enzyme. Binding of
dioxygen leads to the oxy state, which is subsequently
attacked by the second diphenol molecule. Oxidation to the
second quinone forms the met state again and closes the
catalytic cycle.
The main difference between the two mechanistic pro-
posals involves the binding mode of the substrate to the
dicopper(II) core: whereas a monodentate asymmetric
coordination of the substrate was proposed by Klabunde
et al. [11, 14], a simultaneous coordination of the substrate
to both copper centers in a bidentate bridging fashion was
suggested by Solomon et al. [2].
Also of interest in the present context is the study by
Granata et al. [15] of the activity of tyrosinase. They
concluded that the interpretation of the diphenolase reac-
tion is complicated by the fact that cleavage of the peroxide
bond involves coupled electron and proton transfers from
the substrate [16–18]. For this reason, 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 devi-
ces. Compared with the monophenolase reaction, proton
transfers from the diphenol are greatly facilitated by its
stronger acidity. We can thus formulate the evolution of the
ternary complex according to the two alternative routes
represented in Fig. 6, depending on whether proton transfer
from bound catecholate occurs before or after O–O bond
cleavage.
Energetic considerations suggest that the two pathways
should be essentially equivalent, since proton transfers
involve a small energy barrier and heterolytic cleavages of
peroxo or hydroperoxo O–O bonds are expected to be
similar in energy [18, 19].
A rather different mechanism of the catalytic cycle has
also been proposed [18]. That mechanism was built on the
growing number of theoretical [21] and experimental [22,
23] studies suggesting that the active site of an enzyme
should not change its charge during the catalytic cycle. In
the mechanism, proposed by Klabunde et al. [11, 14], the
charge of the active site changes from +1 to +3. This in turn
implies the availability of several external nearby bases,
which could store protons, released during the cycle.
However, the X-ray crystal structure does not reveal the
presence of such candidates in the region of the active site.
Another fact of importance in this context is that for the
similar enzyme tyrosinase, it is required for the conserva-
tion of charge (without external base) that the dicopper
complex has a charge of +1. From the structural similarity
between tyrosinase and catechol oxidase, the charge of the
dicopper complex was therefore chosen as +1 also for
CuIIHis
His
His O
O
CuII His
His
His
CuIHis
His
His
CuI His
His
His
OH
OH
CuIIHis
His
His O
OCuII His
His
His
O
O
OH
OH
O
O
O O CuIIHis
His
His
OH
CuII His
His
His
O O
CuIIHis
His
His
OH
CuII His
His
His 2H+
3H+
2H+
H2O +
+ H2O
H+
oxy state deoxy stateO2
met state
Fig. 5 Catalytic cycle for the
oxidation of o-diphenols to o-
quinones by tyrosinase proposed
by Solomon and coworkers. [2]
1254 J Biol Inorg Chem (2007) 12:1251–1264
123
catechol oxidase. In that proposal, the catalytic cycle thus
starts with a bridging hydroxide ligand between two Cu(I)
ions (Fig. 7). In the first stage, catechol binds to the deoxy
form, transferring the proton to the bridging hydroxide,
with the subsequent generation of a bridging water mole-
cule between the metal centers. Afterwards, dioxygen
displaces the water molecule, binding as a superoxide
radical anion and resulting in the formation of the mixed-
valence dicopper(II,I) species (step a). The superoxide
subsequently abstracts a hydrogen atom (a proton and an
electron) from the bound substrate (step b). To release the
quinone molecule, an electron is then transferred from the
semiquinone radical to the Cu(II) ion, leading to the res-
toration of the dicopper(I) state (step c). The next step
involves the cleavage of the O–O bond, which is accom-
panied by the transfer of two protons from the substrate and
two electrons (from one of the Cu(I) ions and the substrate)
to the peroxide moiety (step d). Altogether this leads to a
product which can be best described as a Cu(II)Cu(I)
species with a semiquinone radical anion. The second
electron transfer from the semiquinone radical to the Cu(II)
center leads to the restoration of the initial hydroxo-bridged
dicopper(I) form (step e). However, it should be noted that
at the present moment this mechanism lacks support from
experimental findings. In particular, the existence of a
bridging l-1,1-superoxide radical anion [18] has never
been reported in the literature for any copper species.
On the other hand, bioinspired synthetic catalysts con-
stitute a valuable tool to explore the reaction mechanisms
that enzymes use to perform their chemistry. The recent
experimental structure determination of catechol oxidase
[11] has encouraged an extensive investigation on model
compounds of this enzyme [25]. The approaches used to
study the mechanism for catecholase activity of Cu(II)
complexes can be divided into four major groups: sub-
strate-binding studies [17, 26–31], structure–activity
relationship studies [31–36], kinetic studies on the catalytic
reaction [26, 28, 33, 37–39] and studies of stochiometric
oxidation of catechol substrates by peroxo–dicopper and
oxo–dicopper complexes [26, 27, 30, 38, 40, 41]. We are
presently working on the study of the catalytic mechanism
of biomimetic complexes of catechol oxidase and we will
return to this point in another paper.
In the present study, it was decided to leave the analogy
to tyrosinase and instead start with a dicopper(II) complex
with charge +2, to investigate if a reasonable reaction
pathway could still be found without adding or releasing
protons from the active complex.
Computational details
All the calculations were done using the B3LYP [42–45]
hybrid density functional. Open shell systems were treated
using unrestricted density functional theory. Geometry
optimizations were performed using a standard valence
LACVP basis set as implemented in the Jaguar 5.5 program
[46]. For the first- and second-row elements, LACVP
implies a 6-31G double-n basis set. For the copper atoms,
LACVP uses a nonrelativistic effective core potential [47],
where the valence part is essentially of double-n quality.
Local minima were optimized using the Jaguar 5.5 program
[46]. The optimizations of the transition states and the
frequency calculations were performed using the Gaussian
03 program [48]. The zero-point vibrational energies were
included in our theoretical results, although no thermal
effects were considered because the use of nonoptimized
frozen coordinates (see ‘‘Chemical models’’) does not give
reliable entropy effects. Accurate single-point energies
were obtained using the cc-pvtz(-F) basis set. For copper
atoms, an effective core potential was used. The sur-
rounding protein was treated with a self-consistent reaction
field method, using a Poisson–Boltzmann solver [49, 50] as
implemented in Jaguar 5.5. The dielectric constant of the
CuO
OCu
O OH
CuO
OCu
O OH
CuO
OCu
O
CuO
OCu
O OH
CuO
OCu
O
CuO
OCu
O OH
CuO
OCu
O OH2H+
H2O
2H+
H2O
(a)
(b)H
H
Fig. 6 The two possible ways
in which the ternary complex
can evolve according to Battaini
and coworkers. [20]
J Biol Inorg Chem (2007) 12:1251–1264 1255
123
homogeneous dielectric medium was set equal to 4.0, in
line with previous modeling of enzymes [51]. The probe
radius was set to 2.50 A. No geometry optimizations
including the dielectric continuum were made since the
calculated dielectric effects were found to be quite small.
The accuracy of the B3LYP functional has been tested in
the extended G3 benchmark set [52], which consists of
enthalpies of formation, ionization potentials, electron
affinities and proton affinities for molecules containing
first- and second-row atoms. The B3LYP functional gives
an average error of 4.3 kcal/mol [52] for 376 different
entries. Owing to the lack of accurate experimental data for
transition metals, there are only a few benchmark tests.
They indicate that for normal metal–ligand bond strengths
the errors are in the range 3–5 kcal/mol [52, 53]. Different
aspects of modeling enzyme active sites have been
reviewed [51, 54, 55].
Results and discussion
Chemical models
The present modeling of the catechol oxidase active site is
based on the X-ray structure of the oxidized form from
sweet potato (PDB code 1BT3; Figs. 2, 3) [11]. In view
of the mechanistic proposals (see ‘‘Introduction’’), the
coppers and the first-shell histidine ligands are the key
chemical species needed to be accurately modeled. Since a
covalent link is present between one of these histidines and
Cys92, this cysteine is also included in the model. The
structural histidines were modeled by imidazoles and the
cysteine was modeled by SCH3. This type of modeling has
been shown to be appropriate on the basis of previous
density functional theory studies of various enzymes [21].
To reproduce the protein strain and a realistic positioning
of the different chemical units, specific restrictions on some
nuclear coordinates were applied. This is important in
modeling fragments not directly bound to the metal [56,
57].
The present description of the mechanism of catechol
oxidase will start from the oxidized, open-shell singlet
complex with oxidation states Cu2(II,II), where the OH was
replaced by a l-g2:g2 bridging peroxide in order to obtain
the intermediate l-g2:g2 bridging peroxide suggested
experimentally [58]. This structure is obtained owing to the
oxidation of Cu2(I,I) by dioxygen to Cu2(II,II). The total
charge for the complex is +2. With this choice of charge
state, the protons needed at later stages in the cycle are
available on the metal complex. This is important since
there is no structural evidence for any external proton
donors or acceptors in the neighborhood of the copper
complex, and it is general experience that active sites
deeply buried in enzymes should not change their charge
CuI CuI
His
His
His
His
His
His
O
O
HH
O
OH
CuII CuI
His
His
His
His
His
HisO
OH
O
CuII CuI
His
His
His
His
His
HisO
O
OH
CuI CuI
His
His
His
His
His
His
O
OH
H
OH
CuI CuII
His
His
His
His
His
His
O
OH
H
OH
OH
OH
OH
OH
O
O
step astep e
step d step b
step c
H2O +
H2O
O2
O
O
OO
O
Fig. 7 The mechanism of the
catalytic cycle of catechol
oxidase proposed previously
based on DFT calculations [24]
1256 J Biol Inorg Chem (2007) 12:1251–1264
123
states [21–23]. The triplet spin state was also considered,
but the triplet structures that we obtained were always at
least 2.0 kcal/mol higher in energy than the singlet ones.
The first half-reaction
As mentioned already, in the present proposal the pathway
starts from a structure where there is a l-g2:g2 bridging
peroxide, as suggested experimentally (Fig. 8). When the
catechol is added to the active site, two hydrogen bonds are
formed between the substrate and the bridging peroxide
(OA–H and OB–H distances are 1.71 and 1.68 A,
respectively) stabilizing the system by 1.6 kcal/mol.
Moreover, the CuB–OB distance increases from 2.05 to
2.38 A.
In the first step of the mechanism the peroxide abstracts
a proton from the catechol. The fully optimized transition
state (TS12) is shown in Fig. 9. At the end, the spin pop-
ulation is –1.06 on the substrate and the charge is –0.5,
showing that one electron has also been transferred. The
net result is that a hydrogen atom (a proton and an electron)
is transferred from the substrate. The spin population on
CuB has changed from –0.57 to 0.18, indicating a change of
the oxidation states of the dimer to Cu2(II,I). This reaction
step is exothermic by –0.5 kcal/mol.
Fig. 8 Fully optimized starting
point (structure 0) for the
catalytic cycle. Atoms marked
with an asterisk were kept fixed,
from the X-ray structure in the
geometry optimization
Fig. 9 Fully optimized
transition state for the first
hydrogen-atom transfer (TS12).
Atoms marked with an asteriskwere kept fixed, from the X-ray
structure in the geometry
optimization. Distances are in
angstroms
J Biol Inorg Chem (2007) 12:1251–1264 1257
123
A comment should be made at this point, since there is
so far no explicit experimental evidence for radicals in
catechol oxidase. However, radicals are strongly impli-
cated in the formation of the covalent link between Cys92
and His109. In other enzymes where such covalent links
are present, radical chemistry in the catalytic mechanism
has also been suggested [59]. For example, tyrosyl radicals
are suggested to be important in both galactose oxidase
[60] and cytocrome c oxidase [61].
The next step is the most complicated and demanding
one in the mechanism, since it involves the cleavage of the
O–O bond (Fig. 10).
At the transition state the O–O bond has increased from
1.53 to 1.81 A. The spin-density population on CuB has
changed from 0.18 to –0.62 and that for OA has increased
from 0.19 to 0.70. This means that there has been an
electron transfer from CuB to OA. The barrier obtained for
this step is 12.1 kcal/mol, which is in reasonable agreement
with the experimental rate (13 kcal/mol) [8]. The fully
optimized transition state is shown in Fig. 11. After the
O–O bond cleavage, the CuB–OB distance has decreased
from 2.91 to 2.07 A.
In the last step of the first half-reaction, a proton transfer
is accompanied by an electron transfer (the spin-density
population on the substrate goes from –1.02 to 0.00 and
that on OA goes from 0.70 to –0.01), which means that a
hydrogen atom is again transferred. Consequently, the first
molecule of quinone is obtained and another catechol can
enter into the catalytic cycle. The fully optimized transition
state for this step (TS34) is shown in Fig. 12.
According to the mechanisms suggested by Solomon
et al. [2] and Eicken et al. [14] the first molecule of water is
obtained after the first half-reaction, apart from the first
quinone molecule. In the present mechanism, there are
instead two OH ligands coordinated to both copper atoms
at this stage. The two molecules of water are going to be
obtained during the second half-reaction.
Some geometrical parameters of the structures that
appear in the first half-reaction as well as their relative
energies can be found in Table 1. The most relevant spin
populations for these structures can be found in Table 2.
According to the results obtained, in the first half-reac-
tion one catechol molecule is oxidized to one quinone and
peroxide is reduced to hydroxides (Fig. 13). Although the
oxidation state of the dicopper core is the same at the start
and at the end of this half-reaction, CuB has a very
important role in it, since it makes the reduction of per-
oxide possible.
With the departure of the quinone, the second catechol
substrate can bind to the dimer. The quinone–catechol
exchange is calculated to be endothermic by 1.4 kcal/mol.
The changes in entropy and the changes in hydrogen
bonding to external residues are assumed to be small,
CuIIHis
His
His O
OCuI His
His
His
OOH
HCuIIHis
His
His O
OCuII His
His
His
OOH
H
32
Fig. 10 The second step of the mechanism suggested for catechol
oxidase
Fig. 11 Fully optimized
transition state for the O–O
bond cleavage (TS23). Atoms
marked with an asterisk were
kept fixed, from the X-ray
structure in the geometry
optimization. Distances
are in angstroms
1258 J Biol Inorg Chem (2007) 12:1251–1264
123
which is probably a reasonable approximation. Modeling
this process more accurately is quite difficult and beyond
present interest.
The second half-reaction
When the second catechol enters the system, the direction
of the OH groups changes in order for hydrogen bonds to
be formed with the substrate (Fig. 14).
In the first step, one of the oxygen atoms of the substrate
binds to CuB and one proton is transferred from the cate-
chol to OB, leading to the formation of the first molecule of
water (Fig. 15). The CuB–OB distance increases from 2.05
to 3.16 A. Since there is no spin on the substrate and the
oxidation state of the copper atoms does not change, no
electrons are transferred in this step. This part of the
mechanism is in agreement with experimental suggestions
that propose an attack on CuB [62].
In the final step of the second half-reaction a hydrogen
atom (a proton and an electron) is transferred from the
substrate and the second molecule of water is obtained. The
fully optimized transition state for this step (TS67) is
shown in Fig. 16.
The spin of CuA changes from –0.64 to 0.00 but for CuB
it remains 0.60. The spin on the substrate increases from
0.00 to –0.60. This means that in order to release the
second quinone product, an electron has to be transferred
from the quinone radical anion to CuB(II), leading to the
reduced Cu2(I,I) dimer.
Some geometrical parameters of the structures that
appear in the second half-reaction as well as their relative
energies can be found in Table 3. The most relevant spin
populations for these structures can be found in Table 4.
According to the results obtained, in the second half-
reaction one catechol molecule is oxidized to one quinone,
the dicopper core is reduced from Cu2(II,II) to Cu2(I,I) and
two molecules of water are obtained from two hydroxides
Fig. 12 Fully optimized
transition state for the second
hydrogen-atom transfer (TS34).
Atoms marked with an asteriskwere kept fixed, from the X-ray
structure in the geometry
optimization. Distances
are in angstroms
Table 1 Comparison of
geometrical parameters for the
structures of the first half-
reaction of the catalytic cycle
of catechol oxidase. Relative
energies are also reported
Structure Distance (A) DE (kcal/mol)
CuA–OA CuA–OB CuB–OA CuB–OB OA–OB
0 2.05 2.04 1.99 2.05 1.51
1 2.10 2.02 1.97 2.38 1.54 0.0
TS12 2.10 2.04 2.13 2.98 1.53 4.9
2 2.11 2.15 2.06 2.91 1.53 –0.5
TS23 2.07 2.00 1.94 3.03 1.81 11.6
3 2.04 2.00 1.93 2.07 2.20 –6.9
TS34 2.04 1.97 1.95 2.16 2.18 –6.0
4 2.03 1.98 1.94 2.08 2.60 –41.0
J Biol Inorg Chem (2007) 12:1251–1264 1259
123
(Fig. 17). With one new oxygen molecule and a catechol
substrate the cycle can start again.
Rather than attempting to calculate the energy of the
rather complicated step where the product quinone and
water go out and a dioxygen molecule and a new substrate
catechol come in, this energy can be estimated form the
overall reaction (Fig. 1). The calculated exothermicity for
this reaction in the gas phase is 20 kcal/mol. To this value,
the energies of placing the reactants and products in a
suitable surrounding must be added. Assuming that the
hydrogen-bonding energies and entropies for the catechol
and the quinone are approximately the same, the energy
contribution from these molecules will be close to zero.
The remaining contributions are then on the reactant
side from one dioxygen and on the product side from
two water molecules. The entropy of dioxygen is estimated
to be 12 kcal/mol and the binding energies of the two
water molecules are estimated to be 14 kcal/mol each. This
will make the total reaction exergonic by 36 (20 –
12 + 14 + 14) kcal/mol.
Conclusions
The catalytic cycle of catechol oxidase has been studied
using methods and models similar to those used before for
many other metalloenzymes. On the basis of experience
from modeling other enzymes [21], the catalytic cycle was
constructed in a way where no proton enters or leaves the
active-site region, thus keeping the charge constant at the
active site. The catalytic cycle suggested is shown sche-
matically in Fig. 18 and the energetics are shown in
Fig. 19. The cycle starts with the oxidation of Cu2(I,I)
by dioxygen to Cu2(II,II), forming a l-g2:g2 bridging
OH
OH
O
O
+ O22- + 2OH -
Fig. 13 The first half-reaction
Fig. 14 Fully optimized
minimum for the starting point
of the second half-reaction.
Atoms marked with an asteriskwere kept fixed, from the X-ray
structure in the geometry
optimization. Distances are in A
Table 2 Comparison of spin populations for the structures of the first
half-reaction of the catalytic cycle of catechol oxidase
Structure Spin density
CuA CuB OA OB Substrate
0 0.50 –0.50 –0.01 0.03
1 0.56 –0.57 –0.02 0.00 0.00
TS12 0.55 –0.31 0.15 0.14 –0.58
2 0.45 0.18 0.19 0.08 –1.06
TS23 0.60 –0.33 0.21 0.44 –1.02
3 0.59 –0.62 0.70 0.36 –1.02
TS34 0.62 –0.61 0.62 0.42 –0.97
4 0.62 –0.61 –0.01 0.01 0.00
1260 J Biol Inorg Chem (2007) 12:1251–1264
123
peroxide. The charge is thus +2. The peroxide abstracts a
hydrogen atom from the first catechol substrate. Subse-
quently, there is cleavage of the O–O bond, followed by
transfer of a second proton, resulting in the first molecule
of quinone. When the second catechol enters the system,
there is transfer of a proton and the first molecule of water
is obtained. When the last hydrogen is transferred, one of
the coppers is reduced and the second water molecule is
obtained. With one new oxygen molecule and a catechol
substrate, the cycle can start again.
In this new mechanism the most critical step is the
peroxide O–O bond cleavage, which has a barrier that is in
Fig. 16 Fully optimized
transition state for the last
hydrogen transfer (TS67).
Atoms marked with an asteriskwere kept fixed, from the X-ray
structure in the geometry
optimization. Distances are in A
Table 3 Comparison of
geometrical parameters for the
structures of the second half-
reaction of the catalytic cycle
of catechol oxidase. Relative
energies are also reported
Structure Distance (A) DE (kcal/mol)
CuA–OA CuA–OB CuB–OA CuB–OB OA–OB
5 2.04 1.97 1.95 2.05 2.47 –39.6
TS56 2.05 1.96 1.92 2.47 2.59 –37.2
6 2.08 1.98 2.03 3.16 2.71 –38.4
TS67 3.03 2.00 2.02 3.44 3.02 –31.1
7 3.89 2.06 2.04 3.27 3.27
Table 4 Comparison of spin
populations for the structures
of the second half-reaction of
the catalytic cycle of catechol
oxidase
Structure Spin density
CuA CuB OA OB Substrate
5 –0.63 0.65 –0.02 0.08 0.00
TS56 –0.64 0.64 –0.03 –0.09 0.00
6 –0.64 0.59 0.03 –0.06 0.08
TS67 –0.50 0.60 0.05 –0.04 –0.12
7 0.00 0.61 0.05 –0.05 –0.60
CuIIHis
His
His O
O
CuII His
His
His
H
H
OOH
H
CuIHis
His
His O
OH2
CuII His
His
His
H
OOH
5 6
Fig. 15 The first step for the second half-reaction of the suggested
mechanism for catechol oxidase
J Biol Inorg Chem (2007) 12:1251–1264 1261
123
reasonable agreement with the experimental rate. In some
steps there is a monodentate coordination of the substrate
to the dicopper core, which is in line with the proposal
by Klabunde et al. [11, 14]. Owing to the structural
similarities between catechol oxidase and tyrosinase,
whose crystallographic structure has recently been reported
[63], the understanding of the activity of the former could
shed some light on how the latter works.
CuIIHis
His
His O
OCuII His
His
His
OOH
H
CuIIHis
His
His O
OCuI His
His
His
OOH
H
CuIIHis
His
His O
OCuII His
His
His
OOH
H
CuIIHis
His
His O
OCuII His
His
His
OO
H
H
CuIIHis
His
His O
O
CuII His
His
His
H
H
OOH
H
CuIIHis
His
His O
OH2
CuII His
His
His
H
OOH
CuIHis
His
His H2O
OH2
CuII His
His
His
OO
O
O
OH
OH
O
O
OH
OH
+ 2H2O
O2 +
1
5 4
3
2
6
7
step astep g
step f step b
step e step c
step d
Fig. 18 Suggested catalytic
cycle for catechol oxidase
OH
OH
O
O
+ 2OH - + 2Cu2+ + 2H2O + 2Cu+
Fig. 17 The second
half-reaction
1 TS343TS232 7TS676TS565
Reaction Coordinate
0.0
-38.4-37.2
-39.6-41.0
-6.0-6.9
11.6
-0.5
4.9
-42.9
-31.1
Quinone
Catechol
-36.0Catechol
O2
Quinone2 H2O
1
)lom/lac k( seigren
E evitaleR
TS12 4
Fig. 19 Potential energy profile
obtained for the catalytic
cycle of catechol oxidase
1262 J Biol Inorg Chem (2007) 12:1251–1264
123
Acknowledgment M.G. thanks the MEC for research grants and
J.M. Luis for valuable discussions. We thank the reviewers for helpful
comments.
References
1. Solomon EI, Baldwin MJ, Lowery MD (1992) Chem Rev
92:521–542
2. Solomon EI, Sundaram UM, Machonkin TE (1996) Chem Rev
96:2563–2605
3. Cuff ME, Miller KI, van Holde KE, Hendrickson WA (1998)
J Mol Biol 278:855–870
4. Gaykema WPJ, Hol WGJ, Vereijken JM, Soeter NM, Bak HJ,
Beintema JJ (1984) Nature 309:23–29
5. Magnus KA, Tonthat H, Carpenter JE (1994) Chem Rev 94:727–
735
6. Cary JW, Lax AR, Flurkey WH (1992) Plant Mol Biol 20:245–
253
7. Deverall BJ (1961) Nature 189:311–315
8. Baruah P, Swain T (1959) J Sci Food Agric 10:125–129
9. Mayer AM, Harel E (1979) Phytochemistry 18:193–215
10. Walker JRL, Ferrar PH (1998) Biotechnol Genet Eng Rev
15:457–498
11. Klabunde T, Eicken C, Sacchettini JC, Krebs B (1998) Nat Struct
Biol 5:1084–1090
12. Eicken C, Zippel F, Buldt-Karentzopoulos K, Krebs B (1998)
FEBS Lett 436:293–299
13. Wilcox DE, Porras AG, Hwang YT, Lerch K, Winkler ME,
Solomon EI (1985) J Am Chem Soc 107:4015–4027
14. Eicken C, Krebs B, Sacchettini JC (1999) Curr Opin Struct Biol
9:677–683
15. Granata A, Monzani E, Bubacco L, Casella L (2006) Chem Eur J
12:2504–2514
16. Ros JR, Rodriguezlopez JN, Garciacanovas F (1994) Biochim
Biophys Acta Protein Struct Mol Enzymol 1204:33–42
17. Granata A, Monzani E, Casella L (2004) J Biol Inorg Chem
9:903–913
18. Siegbahn PEM (2004) J Biol Inorg Chem 9:577–590
19. Siegbahn PEM (2003) J Biol Inorg Chem 8:567–576
20. Battaini G, Granata A, Monzani E, Gullotti M, Casella L (2006)
Adv Inorg Chem Bioinorg Stud 58:185–233
21. Siegbahn PEM (2003) Q Rev Biophys 36:91–145
22. Lee SY, Lipscomb JD (1999) Biochemistry 38:4423–4432
23. Orville AM, Lipscomb JD (1997) Biochemistry 36:14044–14055
24. Bassan A, Borowski T, Siegbahn PEM (2004) Dalton Trans
20:3153–3162
25. Koval IA, Gamez P, Belle C, Selmeczi K, Reedijk J (2006) Chem
Soc Rev 35:814–840
26. Ackermann J, Meyer F, Kaifer E, Pritzkow H (2002) Chem Eur J
8:247–258
27. Berreau LM, Mahapatra S, Halfen JA, Houser RP, Young VG,
Tolman WB (1999) Angew Chem Int Ed Engl 38:207–210
28. Wegner R, Gottschaldt M, Gorls H, Jager EG, Klemm D (2001)
Chem Eur J 7:2143–2157
29. Torelli S, Belle C, Hamman S, Pierre JL, Saint-Aman E (2002)
Inorg Chem 41:3983–3989
30. Koval IA, Belle C, Selmeczi K, Philouze C, Saint-Aman E,
Schuitema AM, Gamez P, Pierre JL, Reedijk J (2005) J Biol
Inorg Chem 10:739–750
31. Than R, Feldmann AA, Krebs B (1999) Coord Chem Rev
182:211–241
32. Kao CH, Wei HH, Liu YH, Lee GH, Wang Y, Lee CJ (2001) J
Inorg Biochem 84:171–178
33. Torelli S, Belle C, Gautier-Luneau I, Pierre JL, Saint-Aman E,
Latour JM, Le Pape L, Luneau D (2000) Inorg Chem 39:3526–3536
34. Belle C, Beguin C, Gautier-Luneau I, Hamman S, Philouze C,
Pierre JL, Thomas F, Torelli S (2002) Inorg Chem 41:479–491
35. Merkel M, Moller N, Piacenza M, Grimme S, Rompel A, Krebs B
(2005) Chem Eur J 11:1201–1209
36. Koval IA, Sehmeczi K, Belle C, Philouze C, Saint-Aman E,
Gautier-Luneau I, Schuitema AM, van Vliet M, Gamez P, Rou-
beau O, Luken M, Krebs B, Lutz M, Spek AL, Pierre JL, Reedijk
J (2006) Chem Eur J 12:6138–6150
37. Monzani E, Battaini G, Perotti A, Casella L, Gullotti M, Santa-
gostini L, Nardin G, Randaccio L, Geremia S, Zanello P,
Opromolla G (1999) Inorg Chem 38:5359–5369
38. Selmeczi K, Reglier M, Giorgi M, Speier G (2003) Coord Chem
Rev 245:191–201
39. Born K, Comba P, Daubinet A, Fuchs A, Wadepohl H (2007)
J Biol Inorg Chem 12:36–48
40. Santagostini L, Gullotti M, Monzani E, Casella L, Dillinger R,
Tuczek F (2000) Chem Eur J 6:519–522
41. Mahadevan V, DuBois JL, Hedman B, Hodgson KO, Stack TDP
(1999) J Am Chem Soc 121:5583–5584
42. Becke AD (1988) Phys Rev A 38:3098–3100
43. Becke AD (1993) J Chem Phys 98:5648–5652
44. Becke AD (1993) J Chem Phys 98:1372–1377
45. Lee CT, Yang WT, Parr RG (1988) Phys Rev B 37:785–789
46. Schrodinger LLC (2003) MacroModel 8.5. Portland, OR
47. Hay PJ, Wadt WR (1985) J Chem Phys 82:299–310
48. Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA,
Cheeseman JR, Montgomery JA Jr, Vreven T, Kudin KN, Burant
JC, Millam JM, Iyengar SS, Tomasi J, Barone V, Mennucci B,
Cossi M, Scalmani G, Rega N, Petersson GA, Nakatsuji H, Hada
M, Ehara M, Toyota K, Fukuda R, Hasegawa J, Ishida M, Nak-
ajima T, Honda Y, Kitao O, Nakai H, Klene M, Li X, Knox JE,
Hratchian HP, Cross JB, Bakken V, Adamo C, Jaramillo J,
Gomperts R, Stratmann RE, Yazyev O, Austin AJ, Cammi R,
Pomelli C, Ochterski JW, Ayala PY, Morokuma K, Voth GA,
Salvador P, Dannenberg JJ, Zakrzewski G, Dapprich S, Daniels
AD, Strain MC, Farkas O, Malick DK, Rabuck AD, Raghava-
chari K, Foresman JB, Ortiz JV, Cui Q, Baboul AG, Clifford S,
Cioslowski J, Stefanov BB, Liu G, Liashenko A, Piskorz P,
Komaromi I, Martin RL, Fox DJ, Keith T, Al-Laham MA, Peng
CY, Nanayakkara A, Challacombe M, Gill PMW, Johnson B,
Chen W, Wong MW, Gonzalez C, Pople JA (2003) Gaussian 03,
revision B.05. Gaussian, Pittsburgh
49. Tannor DJ, Marten B, Murphy R, Friesner RA, Sitkoff D,
Nicholls A, Ringnalda M, Goddard WA, Honig B (1994) J Am
Chem Soc 116:11875–11882
50. Marten B, Kim K, Cortis C, Friesner RA, Murphy RB, Ringnalda
MN, Sitkoff D, Honig B (1996) J Phys Chem 100:11775–11788
51. Blomberg MRA, Siegbahn PEM, Babcock GT (1998) J Am
Chem Soc 120:8812–8824
52. Curtiss LA, Raghavachari K, Redfern PC, Pople JA (2000) J
Chem Phys 112:7374–7383
53. Siegbahn PEM, Blomberg MRA (1999) Annu Rev Phys Chem
50:221–249
54. Siegbahn PEM, Blomberg MRA (2000) Chem Rev 100:421–437
55. Blomberg MRA, Siegbahn PEM (2001) J Phys Chem B
105:9375–9386
56. Pelmenschikov V, Cho KB, Siegbahn PEM (2004) J Comput
Chem 25:311–321
57. Pelmenschikov V, Siegbahn PEM (2003) J Biol Inorg Chem
8:653–662
58. Rompel A, Fischer H, Meiwes D, Buldt-Karentzopoulos K,
Dillinger R, Tuczek F, Witzel H, Krebs B (1999) J Biol Inorg
Chem 4:56–63
J Biol Inorg Chem (2007) 12:1251–1264 1263
123
59. Fontecave M, Ollagnier-de-Choudens S, Mulliez E (2003) Chem
Rev 103:2149–2166
60. Whittaker MM, Whittaker JW (1988) J Biol Chem 263:6074–
6080
61. Proshlyakov DA, Pressler MA, Babcock GT (1998) Proc Natl
Acad Sci USA 95:8020–8025
62. Decker H, Dillinger R, Tuczek F (2000) Angew Chem Int Ed
Engl 39:1591–1595
63. Matoba Y, Kumagai T, Yamamoto A, Yoshitsu H, Sugiyama M
(2006) J Biol Chem 281:8981–8990
1264 J Biol Inorg Chem (2007) 12:1251–1264
123