Investigation of Streptomyces antibioticus tyrosinase reactivity toward chlorophenols

Article abstract of DOI:10.1016/j.abb.2010.09.019

Tyrosinase (Ty) is a copper-containing enzyme ubiquitously distributed in nature. In recent years, Ty has attracted interest as a potential detoxifying agent for xenobiotic compounds with phenolic structure. Among these, chlorophenols are particularly relevant pollutants, commonly found in waste waters. The activity of Streptomyces antibioticus tyrosinase toward isomeric monochlorophenols was studied. Tyrosinase oxidizes both 3- and 4-chlorophenol to the same product, 4-chloro-1,2-ortho-quinone, which subsequently undergoes a nucleophilic substitution reaction at the chlorine atom by excess phenol to give the corresponding phenol-quinone adduct. By contrast, 2-chlorophenol is not reactive and acts as a competitive inhibitor. Docking calculations suggest that the substrates point to one of the copper atoms of the dinuclear center (copper B) and appear to interact preferentially with one of the two coordinated oxygen atoms. The approach of the substrate toward the active site is favored by a π-stacking interaction with one of the copper-coordinated histidines (His194) and by a hydrogen bonding interaction with the O1 oxygen. With this study, we provide the first characterization of the early intermediates in the biotechnologically relevant reaction of Ty with chlorophenols. Additionally, combining experimental evidences with molecular modeling simulations, we propose a detailed reaction scheme for Ty-mediated oxidation of monochlorophenols.

Full text of DOI:10.1016/j.abb.2010.09.019

Archives of Biochemistry and Biophysics 505 (2011) 67–74
Contents lists available at ScienceDirect
Archives of Biochemistry and Biophysics
journal homepage: www.elsevier.com/locate/yabbi
Investigation of Streptomyces antibioticus tyrosinase reactivity toward chlorophenols
Stefano M. Marino a,1, Stefano Fogal a, Marco Bisaglia a, Stefano Moro b, Guido Scartabelli c, Luca De Gioia c,
Alessia Spada d, Enrico Monzani d, Luigi Casella d, Stefano Mammi e, Luigi Bubacco a,
⇑
a Department of Biology, University of Padova, 35121 Padova, Italy
b Department of Pharmaceutical Sciences, University of Padova, 35131 Padova, Italy
c Department of Biotechnology and Biosciences, University of Milano ‘‘Bicocca”, 20126 Milano, Italy
d Department of General Chemistry, University of Pavia, 27100 Pavia, Italy
e Department of Chemical Sciences, University of Padova, 35131 Padova, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online 26 September 2010
Tyrosinase (Ty) is a copper-containing enzyme ubiquitously distributed in nature. In recent years, Ty has
attracted interest as a potential detoxifying agent for xenobiotic compounds with phenolic structure.
Among these, chlorophenols are particularly relevant pollutants, commonly found in waste waters.
The activity of Streptomyces antibioticus tyrosinase toward isomeric monochlorophenols was studied.
Tyrosinase oxidizes both 3- and 4-chlorophenol to the same product, 4-chloro-1,2-ortho-quinone, which
subsequently undergoes a nucleophilic substitution reaction at the chlorine atom by excess phenol to
give the corresponding phenol-quinone adduct. By contrast, 2-chlorophenol is not reactive and acts as
a competitive inhibitor. Docking calculations suggest that the substrates point to one of the copper atoms
of the dinuclear center (copper B) and appear to interact preferentially with one of the two coordinated
Keywords:
Tyrosinase
Copper
Oxygen
Chlorophenols
Enzyme specificity
Enzyme mechanism
Molecular modeling
oxygen atoms. The approach of the substrate toward the active site is favored by a p-stacking interaction
with one of the copper-coordinated histidines (His194) and by a hydrogen bonding interaction with the
O1 oxygen.
With this study, we provide the first characterization of the early intermediates in the biotechnologi-
cally relevant reaction of Ty with chlorophenols. Additionally, combining experimental evidences with
molecular modeling simulations, we propose a detailed reaction scheme for Ty-mediated oxidation of
monochlorophenols.
Ó 2010 Elsevier Inc. All rights reserved.
Introduction
two different enzymatic reactions, the ortho-hydroxylation of
monophenols (monophenolase, or cresolase, activity) and the oxi-
Tyrosinase (Ty)2 (E.C. 1.14.18.1) is a copper-containing enzyme
ubiquitously distributed in nature [1–3]. It is involved in the early
steps of melanin formation [4] and in several other functions,
which include the process of wound healing and the primary im-
mune response in plants, sponges and several invertebrates [5,6].
In arthropods, this enzyme is involved in the sclerotization of the
cuticle after molting [6]. Furthermore, bacterial Ty is secreted into
the soil and has been proposed to be involved in humus formation
through random coupling of different aromatic compounds. These
latter findings have attracted interest in the activity of Ty as a po-
tential detoxifying agent for xenobiotic compounds with phenolic
structure [7,8]. All tyrosinases use molecular oxygen to catalyse
dation of o-diphenols to o-quinones (diphenolase, or catecholase,
activity). The reactive quinones, in turn, polymerize non-enzymat-
ically to the insoluble melanins.
On the basis of spectral comparison with a corresponding model
complex prepared by Kitajima and Moro-oka [9], it was proposed
that hemocyanin and tyrosinase should bind dioxygen as a peroxide
in the side-on bridging (l– :g
g2 2) mode. Recently, this hypothesis
was confirmed also for Ty by the crystal structures of various forms
of the enzyme from Streptomyces castaneoglobisporus [10]. In spite of
the new structural information, as well as the wide spectroscopic lit-
erature regarding type-3 copper proteins [11–14], details of the
intermediate steps in the reaction mechanism remain substantially
unclear, with different hypotheses that still await direct experimen-
tal proof. On the basis of density functional theory calculations, Sieg-
bahn proposed that formation of a hydroperoxide species, by
protonation of the bound peroxide moiety, is crucial for catalysis
[15]. However, such a derivative has not been observed spectroscop-
ically, and it would require a significant reorganization of the oxy-Ty
form found in the 3D structure. Another possibility recently
⇑
Corresponding author. Fax: +39 049 8276300.
E-mail address: luigi.bubacco@unipd.it (L. Bubacco).
Present address: Department of Medicine, Brigham and Women’s Hospital,
1
Harvard Medical School, Boston, MA, USA.
2
Abbreviations used: Ty, tyrosinase; MCPs, monochlorophenols; TFA, trifluoroacetic
acid.
0003-9861/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.abb.2010.09.019

68
S.M. Marino et al. / Archives of Biochemistry and Biophysics 505 (2011) 67–74
emerged from a model study. A
mimicking the Ty active site, was shown to isomerize to a
l
-peroxo-dicopper(II) complex,
stirred, 1-cm path length optical cell thermostated at 25 °C. The
experiments were performed with 0.13 M Ty in 100 mM phos-
l-(bis-
l
oxo)-dicopper(III) species upon binding of a phenolate ion at very
low temperatures and successively promote the ortho-hydroxyl-
ation of the phenol nucleus when the temperature was raised [16].
A mechanistic hypothesis by Decker et al. [17], based on the crystal
structure of Ty, depicts the substrate bound to CuA. A purported
reorientation of the peroxo bridge occurs before the reaction with
the aromatic ring. According to the model of Matoba et al. [10],
PDB 1wx2, CuA is coordinated to His38, His54, and His63 while
CuB is coordinatedto His190, His194, and His216. On the other hand,
a pulsed EPR study on the half-met-Ty derivative of bacterial tyros-
inase identified CuB and not CuA as the paramagnetic probe and site
phate buffer at pH 6.8, while the concentrations of the substrates
were varied between 1 and 20 mM for 3-CP, and between 0.25
and 5 mM for 4-CP. The formation of quinones derived from 3-CP
and 4-CP was monitored at 407 and 412 nm, respectively. In the
experiments with 3-CP, 1
enzyme and shorten the induction period. The initial rates (vo)
were obtained by linear fitting of the A vs. time plots, normally
lM of L-Dopa was added to activate the
D
between 300 and 500 s reaction time. The corresponding kcat val-
ues were calculated using the molar absorptivities obtained as de-
scribed below: 750 Mꢀ1 cmꢀ1 for the oxidation product of 4-CP and
2400 Mꢀ1 cmꢀ1 for that derived from 3-CP. In the inhibition studies
of binding for the transition state analogues kojic acid and
L
-mimo-
of
oxidation was studied as a function of
tween 2 and 15 mM) in the presence of variable amounts of 2-
CP, from 0 to 40 mM. The oxidation of -Dopa was followed spec-
trophotometrically monitoring dopachrome formation at 476 nm;
the KI value for 2-CP was determined using a Dixon plot [27].
L
-Dopa oxidation by 2-CP, the rate dependence of the enzymatic
sine [18]. Probably, these issues will beappropriatelyaddressed only
through studies of the enzyme turnover at low temperatures, which
appear to be feasible on the basis of recent studies on the fungal Ty in
a cryosolvent [19]. In this study, to gain further insight into the enzy-
matic mechanism of tyrosinase, a series of potential phenolic sub-
strates of Streptomyces antibioticus Ty was investigated: the three
isomers of monochlorophenol (MCP) were chosen, because of (i)
their different regiochemistry, and (ii) their relevance as major
industrial and agricultural pollutants [20]. Eukaryotic tyrosinases
(particularly mushroom Ty) have been shown to efficiently remove
different chlorophenols from aqueous solutions and wastewater
[21–23], being particularly effective toward the major pollutant 4-
chlorophenol [20,24]. Therefore, considering the important biotech-
nological implications subtended, we focused our attention on the
reactivity of S. antibioticus Ty with the three MCPs, i.e., 2-CP, 3-CP,
and 4-CP, aiming at rationalizing potential mechanistic differences
in the reactions with the three isomers.
Our analysis, composed of both experimental and computa-
tional approaches, suggests different modes of interaction of the
substrates at the enzyme active site, and therefore this study can
provide important information on Ty regioselectivity and mecha-
nism of activity. Furthermore, these findings could be of interest
for bioremediation, as they provide new insights behind Ty speci-
ficity for different isomeric MCPs, which in some cases can severely
limit the efficacy of the enzyme.
L
-Dopa concentration (be-
L
Computational methodologies
Molecular modeling studies were carried out on a 10 CPU
(PIV-3.0 GHz and AMD64) linux cluster running under openMosix
architecture. Tyrosinase exists in three forms; of the two possible
oxidized forms, met-Ty is the resting form of the enzyme and is
not expected to be appreciably accumulated in the reducing envi-
ronment of the cell. The oxy-Ty form, deriving from oxygenation
of deoxy-Ty, is favored under aerobic cell conditions, as the enzyme
affinity for O2 is very high [28]. On this basis, a reasonable assump-
tion is to consider that the oxy form of the enzyme is the most abun-
dant form in vivo.
Therefore, starting from the X-ray structure of the oxy-form of
S. castaneoglobisporus Ty (corresponding to the 1wx2 pdb entry,
with the exclusion of the ORF378 protein), a homology-based
(the sequence identity between S. antibioticus Ty and S. castaneogl-
obisporus Ty is 82%) model for S. antibioticus Ty was built using the
MOE2006 comparative modeling modules (http://www.chem-
comp.com). Based on previous paramagnetic 1H NMR data on S.
antibioticus Ty active sites [11], the d nitrogens of metal-coordinat-
ing histidines were assumed to be protonated.
Materials and methods
Validation of the protein model was carried out using MOE pro-
tein report and Savs server (http://nihserver.mbi.ucla.edu/SAVS).
To minimize clashes between hydrogen atoms, the structures were
subjectedtoenergyminimization (Amber99 forcefield[29,30])until
the gradient rms was <0.1 kcal molꢀ1 Åꢀ1, keeping the heavy atoms
fixed at their starting positions [10]. Atomic charges of an active site
model including the six coordinating histidines and the dinuclear
site in the oxy-Ty form were computed within the semi-empirical
QM method PM-3TM, using Spartan02 (http://www.wavefun.com/
products/spartan.html). To evaluate the effects of atomic charges
on docking results, in a second set of calculations all amino acid
charges were set according to the MMFF94x forcefield and assuming
+2 and ꢀ1 atomic charges for Cu2+ and peroxo oxygen atoms, respec-
tively. Remarkably, geometries for protein-substrate complexes ob-
tained using the two charge schemes (i.e., Spartan02 and MMFF94x),
turned out to be extremely similar (RMSD lower than 0.1 Å). As for
docking energies, only marginal differences were observed in abso-
lute values, with energy differences never larger than 2 kcal/mol.
The consistency of these data is remarkable: given the rather ex-
treme difference in the charge schemes tested, we concluded that
other factors (e.g., H-bonds, hydrophobic interactions) than electro-
statics of the dicopper center were playing major roles in substrate
affinity and correct orientation in the active site. This same observa-
tion lead us to the conclusion that the exploitation of a higher level of
theory (e.g., DFT) for a more accurate (than PM3-TM) partial charge
Chemicals
Commercial 3-CP and 4-CP were obtained from Sigma and puri-
fied by treatment with active carbon in methanol before use. All
other reagents were purchased in high purity from commercial
sources and used as received.
Protein purification
The enzyme was prepared from the growth medium liquid cul-
tures of S. antibioticus, harboring the pIJ703 expression plasmid,
and purified according to published procedures [25]. Protein con-
centrations were determined optically using a value of 82 mMꢀ1
cmꢀ1 for the molar absorptivity at 280 nm [26]. Purity was checked
by SDS–PAGE and western blot. The protein was stored at ꢀ20 °C at
different concentrations in 100 mM phosphate buffer at pH 6.8 con-
taining 20% glycerol as a cryoprotectant. Prior to the experiments,
glycerol was removed from the storage buffer by dialysis (cut-off
10 kDa) in 100 mM phosphate buffer at pH 6.8.
Kinetic experiments
The kinetics of enzymatic oxidation of the chlorophenols were
studied with an HP8452 spectrophotometer, using a magnetically

S.M. Marino et al. / Archives of Biochemistry and Biophysics 505 (2011) 67–74
69
calculation would not be a key factor: the docking results proved to
be almost independent of the charge scheme (i.e. contributions al-
ways lower than 2 kcal/mol in energy and 0.1 Å in RMSD, as detailed
before). Therefore, only results obtained with PM3 atomic charges
are presented and discussed in the Results section: exactly the same
conclusions apply to the docking results obtained with the alterna-
tive charge scheme. Substrate structures were optimized at the Har-
tree–Fockleveloftheory, usinga 6-311++G(d,p)basis set. MCPs were
docked into the active site (docking box of 15 ꢁ 15 ꢁ 15 Å centered
on the dinuclear center) using the flexible MOE-Dock methodology
and the MMFF94x force field [31]. MM geometry optimizations of
the complex oxy-Ty/MCPs were then carried out starting from the
best configurations obtained from docking simulations.
400 nm, as a function of substrate concentration. With 3-CP, the
absorbance vs. time traces showed an induction period (lag phase)
that decreased with the increase of substrate concentration. When
the kinetic analysis was performed in the presence of 1 lM L-Dopa,
the lag phase was significantly reduced, allowing a better estimate
of the initial rates of the reactions. The resulting plots of initial
rates vs. substrate concentration showed, for both 3-CP and 4-CP,
a typical substrate saturation dependence, from which the kinetic
parameters kcat and km reported in Table 1 were determined.
The determination of the extinction coefficient to be used for the
conversion of the absorbance vs. time curves into reaction rates
is complicated by the fact that the reaction mixtures contain chlo-
roquinone I and the quinone adduct II or III (Scheme 1) in a ratio
that varies with time. Therefore, only an estimate of the extinction
coefficient could be obtained for the product mixture and, as a con-
sequence, the kcat values reported in Table 1 are affected by a
small uncertainty. As described below, 2-CP does not react with
With the aim of finding the most favorable coordination posi-
tion of the phenols, ligands were also forced to approach the dinu-
clear site (using as starting points the best docking poses) at a
distance suitable for coordination to one of the two copper atoms,
restraining the distance between MCPs oxygens and copper atoms.
Ty and acts as an inhibitor of the enzymatic oxidation of
The inhibitory behavior of 2-CP was studied by changing both
the -Dopa and the 2-CP concentrations. The kinetic analysis re-
L-Dopa.
L
Isolation and analysis of the oxidation products of chlorophenols
vealed a competitive inhibition mechanism with an estimated
inhibition constant (KI) of 7.5 0.5 mM.
The reactions were performed with 0.2 lM Ty (final concentra-
tion after 4 additions) in 100 mM phosphate buffer, pH 6.8, at
25 °C. The concentration of the substrate was 5 mM in the experi-
ment with 4-CP and 2 mM in the experiment with 3-CP. The mix-
tures were magnetically stirred in open vessels to maintain the
solution in equilibrium with atmospheric oxygen. After 5 min reac-
tion time, Ty activity was quenched by adding a few drops of conc.
HCl. Subsequently, ascorbic acid was added to reduce the quinones
to catechols. The purification and analysis of the reaction products
were performed by HPLC, loading 1-mL samples of the reaction mix-
tures onto a Supelco LC18 semipreparative column (250 ꢁ 10 mm).
Elution was carried out using 0.1% trifluoroacetic acid (TFA) in dis-
tilled water (solvent A) and 0.1% TFA in acetonitrile (solvent B), with
a flow rate of 5 mL/min. An elution gradient of 20–0% solvent A in
25 min was followed by isocratic elution at 0% A for an additional
5 min. Spectrophotometric detection of the eluate was performed
in the range between 200 and 650 nm using a Jasco MD-1510 diode
array detector. This procedure was repeated several times to enable
the isolation of sufficient material for spectral characterization of
the products. The collected samples were rotary-evaporated to dry-
ness and analyzed by NMR and MS. The 1H NMR spectra were re-
Modelling and docking of the protonated substrates
The sequence identity between S. castaneoglobisporus Ty and S.
antibioticus Ty is 82%. Therefore, starting from the crystallographic
structure of S. castaneoglobisporus Ty in the oxy form (PDB code
1wx2), a comparative model for S. antibioticus Ty was built using
standard homology modelling approaches. Protein and substrates
(MCPs) were then prepared for docking calculations, as detailed
in the Methods section. Results from docking calculations of chlor-
ophenols within the S. castaneoglobisporus Ty active site invariably
predicted an interaction with CuB for all substrates. In particular,
the best poses for the three isomers are almost superimposable
(Fig. 1): the molecules dock in the axial position with respect to
the Cu2O2 plane, with the phenolic hydrogen pointing toward the
coordinated oxygen O1. With 3-CP, two types of relevant interac-
tions are evident: the H-bond interaction that occurs between
the phenolic oxygen and the coordinated oxygen (O1) and the
p-
stacking interaction between His194 and the aromatic ring of the
substrate. Comparing the best pose obtained for 2-CP with those
of the two reactive isomers, the competitive behavior might be ex-
plained by the observation that 2-CP occupies the same spatial po-
sition as 3-CP and 4-CP. The main difference between the three
isomers is in the energy associated with the access to the enzy-
matic cleft (ꢀ60 kcal/mol for 4-CP, ꢀ50 kcal/mol for 2-CP, and
ꢀ3 kcal/mol for 3-CP). It should be noted that, due to the intrinsic
difficulties in the evaluation of docking energies when metal en-
zymes are involved (see Methods for a more detailed discussion
on the subject), these computed energy values must be considered
in relative terms and discussed only to obtain general trends (e.g.,
in this case, the trend would be: 4-CP affinity for Ty > 2-CP affinity
ꢂ3-CP affinity).
corded in CDCl3 solutions with
a
Bruker AVANCE 400
spectrometer, operating at 400.13 MHz. Electrospray ionization
MS spectra were acquired using an LCQ ADV MAX ion trap mass
spectrometer equipped with an ESI ion source controlledby Xcalibur
software 1.3 (Thermo Finnigan, San Jose, CA, USA). The extinction
coefficient of the reaction products was determined by a quantita-
tive analysis of the reaction mixture by NMR. The concentrations
of Ty and substrates were the same as for the HPLC experiments de-
scribed above. The reactions were carried out in 100 mM deuterated
phosphate buffer, pH 6.8. After 5 min reaction time, the absorbance
of the solution was read and the solution was rapidly transferred to
the NMR tube. Before recording the 1H NMR spectrum, Ty activity
was quenched by addition of a few drops of conc. D2SO4, and the
quinone was reduced to catechol by adding an excess of solid ascor-
bic acid.
The next step of the computational investigation aimed at find-
ing a possible coordination position of the phenols. Starting from
Table 1
Results
Kinetic parameters for the tyrosinase catalyzed oxidation of 4-CP and 3-CP, and the
inhibition of L-Dopa oxidation by 2-CP.
Kinetic results
Substrate
Km (mM)
kcat (sꢀ1
)
KI (mM)
4-CP
3-CP
2-CP
0.35 0.02
5.6 0.5
0.90 0.05
–
–
–
Time-dependent measurements of Ty activity towards the chlo-
rophenolic substrates 3-CP and 4-CP were carried out spectropho-
tometrically, by monitoring the formation of quinone around
11
–
1
7.5 0.5

70
S.M. Marino et al. / Archives of Biochemistry and Biophysics 505 (2011) 67–74
OH
O
OH
O
O
OH
Cl-
Ascorbic
Acid
Cl
OH
O
O
Cl
Cl
O
III
VI
Cl
Ty
4-CP
OH
O
O
O
OH
Cl
OH
O
I
Ascorbic
Acid
Ascorbic
Acid
Cl
3-CP
OH
Cl-
OH
Cl
Cl
OH
Cl
V
II
Cl
IV
Scheme 1. Pathway of formation of the quinone adducts resulting from the oxidation of 3- and 4-chlorophenol.
Fig. 1. Best docking poses for 2-CP (light grey), 4-CP (dark gray). 3-CP, which is not shown (E = ꢀ2 kcal/mol) has a pose highly superimposable to that of 4-CP, although less
energetically favored (for 4-CP, E = ꢀ60 kcal/mol). The substrates are rendered in balls and sticks representation: chlorine atoms are in pink, oxygen in red, carbon in grey
(dark grey for 4-CP, light grey for 2-CP). Two ligands (His 194 of CuB, and His 58 of CuA) are labeled in one letter code, to help the reader in understanding the orientation of
the active site residues in the picture. The l-peroxo-dicopper is also represented in balls and sticks: CuA and CuB (brown) are labeled *A and *B respectively; the two oxygen
atoms, O1 and O2 (red balls), of the peroxo bridge are labeled accordingly to the numbering of the crystallographic structure of S. castaneoglobisporus Ty in the oxy form (PDB
code 1wx2). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
the best docking poses (Fig. 1), the phenols were forced to ap-
proach the dinuclear site at a distance suitable for coordination
to one of the two copper atoms, and MMFF94x gradient based
geometry optimizations, in which the positions of the atoms of
the dinuclear center and the distance between MCPs oxygens
and Cu atoms were restrained was carried out. This analysis
reveals almost superimposable coordination pose to the CuB atom
for 3-CP and 4-CP (Fig. 2), with a distance of 1.8 Å between the
interactions with His194 (Fig. 2, as compared with Fig. 1). The best
pose for 2-CP implies a more lateral positioning with respect to the
Cu2O2 core that results in maintaining a closer p-stacking interac-
tion of the 2-CP aromatic ring with His194, but a considerably lar-
ger distance of the phenolate oxygen from CuB (2.3 Å, Fig. 3).
Accordingly, a rational for 2-CP competitive inhibition can be pro-
posed: the access to the cleft is favorable (as discussed before, the
estimated trend for Ty affinity is: 4-CP > 2-CP ꢂ 3-CP), but then
the approach of 2-CP to copper at a useful coordination distance
phenolate oxygen and CuB and a weakening of the
p-stacking

S.M. Marino et al. / Archives of Biochemistry and Biophysics 505 (2011) 67–74
71
Fig. 2. Energetically favored approach to CuB, and best coordination pose for 3-CP. The best pose for 4-CP is highly superimposable to that of 3-CP (not shown for the sake of
clarity): for the para- and meta- substituted MCPs, the estimated coordination distance (pose shown in the figure) is 1.8 Å. Labeling of atoms and abbreviations are the same
as in Fig. 1.
becomes unlikely. A prediction based on docking results is that 2-
CP should completely inhibit Ty-mediated oxidation of 3-CP (i.e.
occupy the same docking site, but with larger affinity, Fig. 1 and
Fig. 2), but not the oxidation of 4-CP.
To verify this prediction, we compared the reactivity of Ty with
different mixtures of MCPs. In different tubes, Ty (0.13 lM) was
quent spectrophotometric detection allowed for a specific quanti-
fication of each MCP isomer in the mixture. In case (i), 2-CP
significantly inhibited the oxidation of 3-CP (below the detection
level of transformed 3-CP after 2 h of incubation); note that in
the absence of 2-CP, approximately 20% of 3-CP was transformed
by Ty in 2 h.
incubated for 2 h with equimolar (1 mM each) concentrations of
(i) 2-CP and 3-CP (ii) 2-CP and 4-CP. HPLC separation and subse-
In case (ii), almost all 4-CP (91%) was oxidized after 2 h by Ty; in
turn, in the absence of 2-CP, all 4-CP (100%) was oxidized by Ty in
Fig. 3. Energetically favored approach to CuB, and best coordination pose for 2-CP: the estimated coordination distance of 2-CP phenolic oxygen to CuB is 2.3 Å, considerably
larger than the distance found for the other MCPs. More importantly, the C–H group in ortho position to the phenol hydroxyl group of 2-CP is too far from both oxygen atoms
of the peroxo-dicopper center. Labeling of atoms and abbreviations are the same as in Figs. 1 and 2.

72
S.M. Marino et al. / Archives of Biochemistry and Biophysics 505 (2011) 67–74
Table 2
2 h reaction time. Altogether, these results well fit with docking
simulations, particularly supporting the prediction that 2-CP, while
not oxidized by the enzyme, has better affinity for Ty active site
than 3-CP; as a consequence of this, and because they dock on
superimposable positions (Fig. 1), the simultaneous presence of
2-CP and 3-CP in the mixture results in an inhibition of 3-CP oxida-
tion. On the other hand, 4-CP oxidation is only marginally affected
by the presence of 2-CP (i.e. more than 90% of 4-CP is still oxi-
dized), in agreement with its higher affinity for Ty active site, esti-
mated from docking calculations and kinetic experiments.
Finally, a regioselective feature implied by the interaction of 3-
CP with the enzyme (Fig. 2) is that ortho-hydroxylation can only
occur regiospecifically at the C6 position; no feasible alternative
pose of pre-coordination was found for 3-chlorophenolate, so that
it can be anticipated that only the 4-chloro-1,2-ortho-benzoqui-
none isomer should form in the enzymatic reaction. To validate
this conclusion, we carried out separate experimental character-
ization of the first reaction products between Ty and each of the
MCPs.
1H NMR spectral data in CDCl3 solution for characterization of compounds IV–VI.
Product
d/ppm (J/Hz)
6.81 (d, Jab ꢃ Jac = 1.3 Hz; Hb, Hc, 2H)
6.91 (false t, Jab = Jac = 1.3 Hz; Ha, 1H)
OH
Hb
Hc
OH
Ha
Cl
6.52 (dd, Jde = 8.6 Hz, Jcd = 2.7 Hz; Hd, 1H)
6.64 (d, Jcd = 2.7 Hz; Hc, 1H)
OH
He
Hd
OH
Hc
6.86 (d, Jde = 8.6 Hz He, 1H)
6.87 (d, Jgh = 8.5 Hz; Hg, 1H)
6.95 (false t, Jobs ꢃ 2.1 Hz, Hf, 1H)
7.03 (dd, Jfi = 2.1, Jih = 8.5 Hi, 1H)
7.23 (false t, Jobs = 8.5 Hz; Hh, 1H)
Hf
O
Cl
Hg
Hi
Hh
OH
Hc
Product characterization of the enzymatic oxidation of MCPs
6.49 (dd, Jed = 8.5 Hz, Jdc = 2.7 Hz; Hd, 1H)
6.62 (d, Jdc = 2.7 Hz; Hc, 1H)
OH
He
Hd
The enzymatic oxidation of 3-CP and 4-CP produces the same 4-
chloro-1,2-ortho-quinone derivative (I), but this compound under-
goes an easy nucleophilic aromatic substitution at the chlorine
atom by excess phenol, yielding the dimeric phenol-quinone ad-
ducts II and III, respectively (Scheme 1). Thus, in the case of 3-
CP, only one of the two possible Ty-mediated quinone derivative
was found (Scheme 1, species I): the hydroxylation occurs specifi-
cally at position C6, while no evidence of 3-chloro-1,2-ortho-qui-
none was detected. Therefore, the molecular modeling-based
prediction of complete regiospecificity (ortho-hydroxylation only
at C6) in the case of 3-CP Ty-mediated oxidation was fully con-
firmed by these experiments.
6.85 (d, Jed = 8.5 Hz; He, 1H)
0
0
0
6.90 (d, Jab = Ja b = 8.9 Hz; Hb, Hb , 2H)
0
0
0
7.25 (d, Jab = Ja b = 8.9 Hz; Ha, Ha , 2H)
Hb
O
Ha
Cl
Hb'
Ha'
To prevent further reactions of the reactive quinones during
separation, it was found useful to reduce the quinone moieties of
I, II and III to the corresponding catechols, IV, V and VI with ascor-
bic acid (Scheme 1). These products were separated by semipre-
parative HPLC (retention times were 8 min for IV, and 14 min for
V and VI) and characterized by 1H NMR and MS techniques. The
NMR spectral data for these compounds are shown in Table 2.
The MS spectra of V and VI exhibit a molecular ion at m/z 237.2,
corresponding to [M + H]+. Finally, the ortho isomer 2-CP is not oxi-
of quinone I enables its detection together with the adducts II or
III. In a previous study, a similar behavior was found for para
and meta substituted fluorophenols [32]. These compounds were
shown to be oxidized by Ty to 4-fluoro-1,2-ortho-quinone: how-
ever, in that case the quinone could not be detected, because it re-
acted too fast with the fluorophenols [32]. In this respect, it is
interesting that with the halophenols, the oxidative polymeriza-
tion process proceeds in the initial phase with nucleophilic substi-
tution at the halogen atom of the quinone (e.g. I), yielding quinone
derivatives like II or III, whereas the common evolution of enzy-
matically formed quinones involves Michael addition reactions
yielding catechol derivatives [33]. It is likely that evolution of the
dimeric quinones II/III involves a combination of both types of pro-
cesses. The reaction catalyzed by Ty to produce brown pigments
from chlorophenols was reported previously [34]; this process is
relevant in the detoxification of polluted waters, but no informa-
tion on the intermediates formed in the pathway was available
[24].
dized by the enzyme and behaves as an inhibitor of
oxidation.
L-Dopa
Following the enzymatic oxidation of 3-CP and 4-CP by NMR in
the initial phase (within 5 min) of the reactions (after reduction of
the quinone species I, II and III to the corresponding catechols IV, V,
and VI) it was observed that the mixture composition changed
with time and strongly depended on the initial phenol concentra-
tion but there were no other products in the reaction mixtures.
Compounds II and III are not stable and, for prolonged reaction
times, polymerization occurred.
Focusing the discussion on structural and functional aspects, an
important feature of the behavior of S. antibioticus Ty is its selectiv-
ity toward the different MCPs. 2-CP is not a substrate even though
it binds to the enzyme active site, as witnessed by the competitive
Discussion
In this work, we analyzed the behavior of S. antibioticus Ty to-
ward MCPs from different perspectives, trying (i) to establish
structural and functional determinants behind Ty specificity to-
ward these compounds and, (ii) to characterize the initial reactions
products. Starting from the latter point, we found that only the 3-
CP and 4-CP isomers are oxidized by the enzyme and quinone I, the
common initial product of enzymatic oxidation, undergoes a nucle-
ophilic aromatic substitution at the chlorine atom, yielding ad-
ducts II and III, respectively (Scheme 1). The moderate reactivity
inhibition behavior in the oxidation of L-Dopa (with KI in the mM
range). We conducted docking simulations for all MCPs using a
theoretical model generated for S. antibioticus Ty. The results
showed that 2-CP binds too far from the dinuclear center (with a
phenolate oxygen – CuB distance of 2.3 Å). Furthermore, both car-
bons ortho to the hydroxyl group turned out to be far from the oxy-
gen atoms of the coordinated peroxide (more than 4 Å separating
the nearest oxygen atom bound to copper (O1) from either carbon).
This may rationalize the competitive inhibition behavior of 2-CP, as

S.M. Marino et al. / Archives of Biochemistry and Biophysics 505 (2011) 67–74
73
its docking to the reaction center is quite favorable (i.e. the pre-
dicted affinity trend of MCPs for the Ty active site is 4-CP > 2-
CP ꢂ 3-CP) and forms a complex with the enzyme that is almost
superimposable with that of the reactive 3-CP and 4-CP. However,
in contrast with 3-CP and 4-CP, the 2-CP oxygen cannot approach
any of the copper atoms at a useful coordination distance. Based on
our calculations, we could predict that 2-CP is potentially capable
to completely inhibit Ty-mediated oxidation of 3-CP, if the two iso-
mers are present in the same mixture (e.g. a polluted sample), but
not the oxidation of 4-CP. These predictions were successively
experimentally validated, providing independent support to the
computational results. Moreover, from a biotechnological point of
view, these findings indicate that the efficacy of Ty as a bio-reme-
diating agent with respect to MCPs may vary considerably,
depending on the relative amount of different MCP isomers in
the mixture: thus, relatively high concentrations of 2-CP may eas-
ily result in severe suppression of Ty ability to oxidize other CPs.
An additional structure-based feature of Ty reactivity with
MCPs is its regiospecificity: our docking simulations implied that
ortho-hydroxylation (the first step in the oxidation of MCPs) could
only occur at the C6 position. This finding was straight forwardly
validated on an experimental basis, when we characterized the
first reaction products: Ty-mediated oxidation of 3-CP yielded only
quinone, 4-chloro-1,2-ortho-quinone, therefore demonstrating
specific hydroxylation of 3-CP at the C6 position.
line with the crystallographic data available for Ipomoea batatas
catechol oxidase, currently the only available type-3 copper pro-
tein crystallized with a bound inhibitor in the active site [35],
and with the indications that emerged from a pulsed EPR experi-
ment [12,18], a paramagnetic NMR [36] and an X-ray absorption
study on bacterial Ty [37]. The second feature is a p-stacking inter-
action with His194, which stabilizes the protonated substrate in its
approach to the dinuclear center. According to our results, this
interaction has to be lost before coordination to CuB. A third fea-
ture is the H-bond interaction between the phenolic substrate
and the coordinated O1 atom, which is the most negative atom
of the coordinated peroxide. Both this interaction and the previ-
ously mentioned aromatic
p-stacking interaction contribute to
the proper orientation of the substrate in the active site cleft. The
actual acceptor of the phenolic proton is currently not known. Both
the peroxo group [38] and one of the coordinating histidines [15]
have been invoked as potential protonation sites. In the first case,
the peroxo group would be transformed into a hydroperoxo spe-
cies. This could only occur with substantial rearrangement of the
coordination to the dicopper center (i.e. from l– : l– : )
g2 g2 to g1 g1
because the protonated oxygen atom could no longer bind to metal
centers. Thus, although the phenol hydroxylating ability of a dicop-
per-hydroperoxo species has been demonstrated in a model sys-
tem [39], we believe that protonation of bound dioxygen does
not occur in the enzyme. Also the possibility that phenol binding
occurs with release of one of the copper-bound imidazole ligands
does not seem to be feasible, because it would involve substantial
conformational changes of the protein. Therefore, in the reaction
pathway depicted in Scheme 2, we assume that the hydroxyl pro-
ton is transferred from the phenol to an acceptor residue near the
active site, neglecting the weak hydrogen bonding interaction to
the coordinated oxygen atom O1. According to the tyrosinase
Altogether, the data reported in this paper provide several dif-
ferent and independent pieces of information useful to propose a
possible reaction scheme (Scheme 2); although still speculative
to some extent, our model can rationalize observations from both
earlier studies and the present findings. The first structural feature
that should be noted in the proposed reaction scheme is that coor-
dination of the substrate is more likely to occur at CuB. This is in
R
R
H
N
H
N
BH+
_
B
HO
O
N
N
(H194)
H38
H54
(H194)
H190
H38
O(1)
O
Cu
Cu
Cu
A
Cu
H190
B
B
O
O
A
(2)
H54
H216
H216
R
H63
H63
O2
R
O
H
N
O
BH+
(H194)
H190
N
O
B
H38
H54
O
Cu
Cu
H38
H194
Cu
B
A
O
H216
H63
Cu
H190
A
B
H54
R
O
H63
H216
H
N
BH+
(H194)
N
O
H38
H54
H2O
Cu
Cu
B
A
H190
O
H
H63
H216
Scheme 2. Structural model for the essential steps of the monophenolase reaction catalyzed by tyrosinase.

74
S.M. Marino et al. / Archives of Biochemistry and Biophysics 505 (2011) 67–74
[3] N. Wang, D.N. Hebert, Pigment Cell Res. 19 (2006) 3–18.
[4] F. Murisier, F. Beermann, Histol. Histopathol. 21 (2006) 567–578.
[5] W.E. Müller, V.A. Grebenjuk, N.L. Thakur, A.N. Thakur, R. Batel, A. Krasko, I.M.
Müller, H.J. Breter, Appl. Environ. Microbiol. 70 (2004) 2332–2341.
[6] L. Cerenius, K. Söderhäll, Immunol. Rev. 198 (2004) 116–126.
[7] J.W. Park, J. Dec, J.E. Kim, J.M. Bollag, Environ. Sci. Technol. 33 (1999) 2028–
2034.
[8] J.W. Park, J. Dec, J.E. Kim, J.M. Bollag, Arch. Environ. Contam. Toxicol. 38 (2000)
405–410.
[9] N. Kitajima, Y. Moro-oka, Chem. Rev. 94 (1994) 737–757.
[10] Y. Matoba, T. Kumagai, A. Yamamoto, H. Yoshitsu, M. Sugiyama, J. Biol. Chem.
281 (2006) 8981–8990.
structure [10], few residues in the active site environment may
play the role of proton acceptor. Specifically, residue His215, adja-
cent to one of the copper ligands and corresponding to a conserved
residue in the sequences of fungal and plant polyphenol oxidases
and hemocyanin [40], is the best candidate to play this role. In
any case, proton transfer from the phenol to a nearby residue could
occur with minimal structural reorganization of the protein.
Conclusions
[11] L. Bubacco, J. Salgado, A.W. Tepper, E. Vijgenboom, G.W. Canters, FEBS Lett. 442
(1999) 215–220.
[12] M. van Gastel, L. Bubacco, E.J. Groenen, E. Vijgenboom, G.W. Canters, FEBS Lett.
474 (2000) 228–232.
[13] A.W. Tepper, L. Bubacco, G.W. Canters, J. Biol. Chem. 277 (2002) 30436–30444.
[14] S. Itoh, S. Fukuzumi, Acc. Chem. Res. 40 (2007) 592–600.
[15] P.E. Siegbahn, J. Biol. Inorg. Chem. 8 (2003) 567–576.
[16] L.M. Mirica, M. Vance, D.J. Rudd, B. Hedman, K.O. Hodgson, E.I. Solomon, T.D.
Stack, Science 308 (2005) 1890–1892.
[17] H. Decker, T. Schweikardt, F. Tuczek, Angew. Chem. Int. Ed. Engl. 45 (2006)
4546–4550.
[18] L. Bubacco, M. Van Gastel, E.J. Groenen, E. Vijgenboom, G.W. Canters, J. Biol.
Chem. 278 (2003) 7381–7389.
[19] A. Granata, E. Monzani, L. Bubacco, L. Casella, Chem. Eur. J. 12 (2006) 2504–
2514.
[20] J.M. Bollag, H.L. Chu, M.A. Rao, L. Gianfreda, J. Environ. Qual. 32 (2003) 63–69.
[21] J. Dec, J.M. Bollag, Arch. Environ. Contam. Toxicol. 19 (1990) 543–550.
[22] K. Tatsumi, S. Wada, H. Ichikawa, Biotechnol. Bioeng. 51 (1996) 126–130.
[23] W.Q. Sun, G.F. Payne, Biotechnol. Bioeng. 51 (1996) 79–86.
[24] K. Yamada, Y. Akiba, T. Shibuya, A. Kashiwada, K. Matsuda, M. Hirata,
Biotechnol. Prog. 21 (2005) 823–829.
[25] L. Bubacco, E. Vijgenboom, C. Gobin, A.W. Tepper, J. Salgado, G.W. Canters, J.
Mol. Catalysis B 8 (2000) 27–35.
[26] K. Lerch, L. Ettlinger, Pathol. Microbiol. (Basel) 38 (1972) 23–25.
[27] M. Dixon, Biochem. J. 55 (1953) 170–171.
[28] S. Hirota, T. Kawahara, E. Lonardi, E. De Waal, N. Funasaki, G.W. Canters, J. Am.
Chem. Soc. 127 (2005) 17966–17967.
[29] J.W. Ponder, D.A. Case, Adv. Protein Chem. 66 (2003) 27–85.
[30] S.J. Weiner, P.A. Kollman, D.A. Case, U.C. Singh, C. Ghio, G. Alagona, S. Profeta, P.
Weiner, J. Am. Chem. Soc. 106 (1984) 765–784.
The reactivity of S. antibioticus tyrosinase toward the three iso-
meric monochlorophenols, 2-CP, 3-CP, and 4-CP, was studied.
While 4-CP and 3-CP are substrates of the enzyme, 2-CP acts as a
competitive inhibitor. A rational behind this behavior has been
proposed on the basis of both computational and experimental
studies: 2-CP shows affinity for the Ty active site, but cannot ap-
proach the dicopper center with a reactive consistent pose. As a
common feature of our structural investigations, we found that
all substrates approached Ty active site pointing toward CuB; sta-
bilizing interactions occur with His194 and with one of the oxygen
atoms (O1, which is the most negative atom of the coordinated
peroxide) of the metal center. Based on our calculations, we addi-
tionally predicted that 3-CP undergoes ortho-hydroxylation regio-
specifically at the C6 position. This was confirmed by 1H NMR
characterization of the initial reaction products of Ty oxidation of
3-CP and 4-CP. With this analysis, we could also provide the first
characterization of the early intermediates in the biotechnological-
ly relevant reaction of Ty with chlorophenols. Finally, combining
experimental evidence with various molecular modeling simula-
tions, we propose a reaction scheme for Ty-mediated oxidation
of MCPs, which allows to rationalize observations from both earlier
studies and the present findings.
[31] T.A. Halgren, J. Comp. Chem. 20 (1999) 730–748.
[32] G. Battaini, E. Monzani, L. Casella, E. Lonardi, A.W. Tepper, G.W. Canters, L.
Bubacco, J. Biol. Chem. 277 (2002) 44606–44612.
Acknowledgments
[33] E.J. Land, C.A. Ramsden, P.A. Riley, Acc. Chem. Res. 36 (2003) 300–308.
[34] K. Ikehata, J.A. Nicell, Biotechnol. Prog. 16 (2000) 533–540.
[35] T. Klabunde, C. Eicken, J.C. Sacchettini, B. Krebs, Nat. Struct. Biol. 5 (1998)
1084–1090.
[36] A.W. Tepper, L. Bubacco, G.W. Canters, J. Am. Chem. Soc. 127 (2005) 567–575.
[37] L. Bubacco, R. Spinazze, Arch. Biochem. Biophys. 465 (2007) 320–327.
[38] J.S. Conrad, S.R. Dawso, E.R. Hubbard, T.E. Meyers, K.G. Strothkamp,
Biochemistry 33 (1994) 5739–5744.
[39] G. Battaini, E. Monzani, A. Perotti, C. Para, L. Casella, L. Santagostini, M. Gullotti,
R. Dillinger, C. Näther, F. Tuczek, J. Am. Chem. Soc. 125 (2003) 4185–4198.
[40] C.M. Marusek, N.M. Trobaugh, W.H. Flurkey, J.K. Inlow, J. Inorg. Biochem. 100
(2006) 108–123.
The authors acknowledge Prof. G.W. Canters for providing the
expression system for bacterial tyrosinase. This work was sup-
ported by a grant of the PRIN Project (Italian MIUR) and by
CIRCMSB.
References
[1] S. Halaouli, M. Asther, J.C. Sigoillot, M. Hamdi, A. Lomascolo, J. Appl. Microbiol.
100 (2006) 219–232.
[2] C.W. Van Gelder, W.H. Flurkey, H.J. Wichers, Phytochemistry 45 (1997) 1309–
1323.