Unsymmetrical binding modes of the HOPNO inhibitor of tyrosinase: From model complexes to the enzyme

Article abstract of DOI:10.1002/chem.201202643

The deciphering of the binding mode of tyrosinase (Ty) inhibitors is essential to understand how to regulate the tyrosinase activity. In this paper, by combining experimental and theoretical methods, we studied an unsymmetrical tyrosinase functional model and its interaction with 2-hydroxypyridine-N-oxide (HOPNO), a new and efficient competitive inhibitor for bacterial Ty. The tyrosinase model was a dinuclear copper complex bridged by a chelated ring with two different complexing arms (namely (bis(2-ethylpyridyl)amino)methyl and (bis(2-methylpyridyl)amino)methyl). The geometrical asymmetry of the complex induces an unsymmetrical binding of HOPNO. Comparisons have been made with the binding modes obtained on similar symmetrical complexes. Finally, by using quantum mechanics/molecular mechanics (QM/MM) calculations, we studied the binding mode in tyrosinase from a bacterial source. A new unsymmetrical binding mode was obtained, which was linked to the second coordination sphere of the enzyme. Copyright

Full text of DOI:10.1002/chem.201202643

FULL PAPER
DOI: 10.1002/chem.201202643
Unsymmetrical Binding Modes of the HOPNO Inhibitor of Tyrosinase:
From Model Complexes to the Enzyme
Constance Bochot,[a] Elisabeth Favre,[b] Carole Dubois,[c] Benoit Baptiste,[a]
Luigi Bubacco,[d] Pierre-Alain Carrupt,[b] Gisꢀle Gellon,[a] Renaud Hardrꢁ,[c]
Dominique Luneau,[e] Yohann Moreau,[f] Alessandra Nurisso,[b] Marius Rꢁglier,[c]
Guy Serratrice,[a] Catherine Belle,*[a] and Hꢁlꢀne Jamet*[a]
Abstract: The deciphering of the bind-
ing mode of tyrosinase (Ty) inhibitors
is essential to understand how to regu-
late the tyrosinase activity. In this
paper, by combining experimental and
theoretical methods, we studied an un-
was a dinuclear copper complex bridg-
ed by a chelated ring with two different
complexing arms (namely (bis(2-ethyl-
pyridyl)amino)methyl and (bis(2-meth-
ylpyridyl)amino)methyl). The geomet-
rical asymmetry of the complex induces
an unsymmetrical binding of HOPNO.
Comparisons have been made with the
binding modes obtained on similar
symmetrical complexes. Finally, by
using quantum mechanics/molecular
mechanics (QM/MM) calculations, we
studied the binding mode in tyrosinase
from a bacterial source. A new unsym-
metrical binding mode was obtained,
which was linked to the second coordi-
nation sphere of the enzyme.
symmetrical
model and its interaction with 2-hy-
droxypyridine-N-oxide (HOPNO),
tyrosinase
functional
Keywords: bioinorganic chemistry ·
copper · density functional calcula-
tions · enzymes · inhibitors
a
new and efficient competitive inhibitor
for bacterial Ty. The tyrosinase model
Introduction
Tyrosinases (Tys, EC 1.14.18.1) are copper-containing en-
zymes belonging to the type-3 family or coupled binuclear
family.[1] Three distinct forms have been identified in various
II
ꢀ
[a] Dr. C. Bochot, Dr. B. Baptiste, G. Gellon, Prof. G. Serratrice,
Dr. C. Belle, Dr. H. Jamet
stages of the catalytic cycle: a native met state (CuII Cu )
bridged by aquo (hydroxo) ligands, a reduced deoxy state
Dꢁpartement de Chimie Molꢁculaire
Equipes CIRe et Chimie Thꢁorique
Universitꢁ J. Fourier, UMR-CNRS 5250, ICMG FR-2607
BP 53, 38041, Grenoble, Cedex 09 (France)
Fax : (+33)476-514-836
(CuI CuI), and an oxy state with a dioxygen molecule bound
II
2ꢀ
II [2]
ꢀ
ꢀ
to the dicopper center (Cu O2 Cu ). Tys are present in
almost all types of organism[3] and their main function is pig-
ment biosynthesis. In mammals, Ty is responsible for the
critical steps in eumelanin production (the most protective
melanin, found particularly in the skin).[4] Melanin-related
disorders are known to cause serious skin lesions[5] and mel-
anoma,[6] and Parkinsonꢀs disease has been linked to Ty,[7] so
this enzyme family has considerable interest for medicinal
chemistry. In fungi and plants, Ty catalyzes the oxidation of
phenolic compounds and subsequently produces types of
melanins called allomelanins.[8] In the area of food storage,
Ty is a key enzyme in the browning process that occurs
upon tissue damage or long storage. For these reasons, a
large variety of synthetic or natural compounds have been
tested as Ty inhibitors.
E-mail: catherine.belle@ujf-grenoble.fr
helene.jamet@ujf-grenoble.fr
[b] Dr. E. Favre, Prof. P.-A. Carrupt, Dr. A. Nurisso
School of Pharmaceutical Sciences
University of Geneva, University of Lausanne
1211 Geneva (Switzerland)
[c] Dr. C. Dubois, Dr. R. Hardrꢁ, Dr. M. Rꢁglier
Institut des Sciences Molꢁculaires de Marseille, ꢁquipe BiosCiences
Aix Marseille Universitꢁ, CNRS, ISM2, UMR 7313
13397, Marseille, Cedex 20 (France)
[d] Prof. L. Bubacco
Department of Biology, University of Padova
Via Ugo Bassi 58b, 35121 Padova (Italy)
[e] Prof. D. Luneau
Laboratoire des Multimatꢁriaux et Interfaces UMR-CNRS 5615
Universitꢁ Claude Bernard Lyon 1, Campus de la Doua
69622 Villeurbanne (France)
A large number of the Ty inhibitors described in the liter-
ature target the Ty binuclear copper active site. These inhib-
itors, named transition-state (TS) analogues, are structural
analogues of catechol but cannot be oxidized. Kojic acid[9]
[f] Dr. Y. Moreau
iRTSV/CBM/MCT, CEA Grenoble, 17 avenue des martyrs
38054, Grenoble, Cedex 09 (France)
(5-hydroxy-2-(hydroxymethyl)-4-pyrone,
KA)
and
HOPNO[10] (2-hydroxypyridine-N-oxide) are good para-
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201202643.
digms of this strategy. The binding mode of these inhibitors
Chem. Eur. J. 2013, 19, 3655 – 3664
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3655

on the Ty bimetallic center remains an open question. X-ray
absorption spectroscopy (XAS), paramagnetic NMR spec-
troscopy, and electron spin echo envelope modulation
(ESEEM) studies suggest direct binding to the copper ions
for KA in the adduct with bacterial Streptomyces antibioti-
cus Ty, with a m-h2:h1 binding mode reported.[11] On the
other hand, the X-ray crystal structure of the bacterial Bacil-
lus megaterium Ty with KA as a ligand reveals that the KA
is not bound to the dicopper(II) center but is located at a
distance of 7 ꢃ from it.[2b] Recently, the crystal structure of
Agaricus bisporus mushroom Ty (deoxy state) in complex
with tropolone,[12] another TS-analogue inhibitor, revealed
that tropolone binds near the binuclear copper site without
directly coordinating the copper ions. A first approach to
obtain more insight into the binding mode is to study inhibi-
tor interactions with low-molecular-weight binuclear copper
systems that mimic the dicopper(II) center of the enzyme.
To date, few relevant structures with a coordinated inhibitor
(or inactivated substrate) have been published.[13] In particu-
lar, there is no example characterized by X-ray analysis in
which the various coordination modes of the inhibitor, in-
cluding an unsymmetrical bridging mode, are described. Re-
cently, we have successfully isolated and fully characterized
two 1:1 adducts between the dinuclear copper(II) complexes
Scheme 1. Schematic representation of the relevant dicopper model com-
plexes [Cu2A(BPMP)(m-OH)] 2A(BPEP)(m-OH)]-
(ClO4)2 (1)[14] and [Cu
(ClO4)2 (2)[15] and the characterized or possible binding modes for
HOPNO at the dicopper(II) centers.
G
R
G
N
ACHTUNGTRENNUNG
[Cu
dylmethyl)-aminomethyl]-4-methylphenol)[14]and
(BPEP)(m-OH)](ClO4)2 (2; HBPEP: 2,6-bis[bis(2-pyridyle-
2ACHTUNGTRENNUNG(BPMP)ACHTUNGTRENNUNG(m-OH)]ACHTUNRTGENG(NUN ClO4)2 (1; HBPMP: 2,6-bisACTUHNGTRENNNUG
AHCTUNGTRENNUNG
A
E
N
ACHTUNGTRENNUNG
thyl)aminomethyl]-4-methylphenol),[15] known as functional
models for Ty catecholase activity, and deprotonated
HOPNO. In the first adduct, 1-OPNO, it binds only one
copper(II) atom of the dinuclear site in a chelating mode,
but in the second adduct, 2-OPNO, the inhibitor binds the
two copper(II) atoms in a symmetric h1:h1 bridging mode
(Scheme 1).
Important information provided by the recent structure
resolutions of various Ty forms is the dissymmetry of the
active-site structure. The Ty active site is not symmetric and
it is composed of two copper atoms, each coordinated by
three histidine residues. This dissymmetry is clear in the
structure of a met form of the bacterial S. castaneoglobispo-
rus Ty (Figure 1; PDB code: 2AHK),[2a] in which the geome-
try around the CuA center is close to a square pyramid (t=
0.22), whereas the geometry around the CuB center is inter-
mediate between a square pyramid and a trigonal bipyramid
(t=0.48); (t=0 for a regular square pyramid and t=1 for
the regular trigonal bipyramid geometry).[16]
Figure 1. Graphical representation of the met form of the bacterial Strep-
tomyces castaneoglobisporus Ty active site (PDB code: 2AHK)[2a] show-
ing the dissymmetry of the coordination of the two copper centers.
In order to mimic this dissymmetry, we pursued our inves-
tigations by generating an unsymmetrical complex, [Cu2-
ACHTUNGTRENNUNG(BPMEP)ACHTUNGTRENNUNG(m-OH)]ACHTUNGTRENNUNG(ClO4)2 (3), which was obtained from the
new unsymmetrical 2-[bis(2-pyridylethyl)-aminomethyl]-6-
[bis(2-pyridylmethyl)-aminomethyl]-4-methylphenol
pared. The aim was to get insights into inhibitor binding
through variation in the side arms of the dicopper model
complexes. Finally, adducts found with model complexes
were used to study the binding mode of HOPNO on bacteri-
al Ty by computational methods. We choose bacterial Ty be-
cause crystallographic structures of the met form have only
been obtained for this tyrosinase and the met form corre-
sponds to the electronic structure of our model complexes
(HBPMEP) ligand (Scheme 2).[17] The HBPMEP ligand in-
volves two different complexing arms (namely (bis(2-ethyl-
pyridyl)amino)methyl and bis(2-methylpyridyl)amino)meth-
yl) in mode to tune an unsymmetrical binding of HOPNO.
By combining experimental and theoretical studies, the dif-
ferent complexes and HOPNO binding modes are com-
II
II
ꢀ
(Cu Cu ).
A hybrid quantum mechanics (QM/MM)
method combining classical MM and a QM method was em-
ployed on the whole enzyme. The results obtained highlight
3656
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 3655 – 3664

Unsymmetrical Binding Modes of Tyrosinase Inhibitors
FULL PAPER
Figure 2. ORTEP plots of dications a) [Cu
B) [Cu2A(BPMEP)
(OPNO)]2+ (10a2+). H atoms (except for the hydroxo
(m-OH)]2+ (32+) and
2ACHUTGTNERNNUG(BPMEP)ACHTUGNTRENNUGN
C
ACHTUNGTRENNUNG
ꢀ
group) and ClO4 counteranions were omitted for clarity. Ellipsoids are
drawn at the 30% probability level.
Scheme 2. Synthesis of the HBPMEP ligand (9) and complex 3: i) N,N-
bis(2-ethylpyridyl)amine (5), 54%; ii) HBF4 (48% in water/MeCN),
90%; iii) 1) SOCl2, 2) N,N-bis(2-methylpyridyl)amine (8), 67%. TEA:
triethylamine.
(m-OH)]2+ (32+) is shown in
tionic complex [Cu2ACTHNUTRGENN(UG BPMEP)AHCUTNGTRENNGUN
Figure 2A and selected bond lengths and angles are given in
Table S1 in the Supporting Information.
the importance of a serine residue that is close to the active
site in the enzyme.
The crystal structure reveals that the two copper atoms
are doubly bridged by the phenoxo and a hydroxo group.
The pentacoordination of the Cu1 and Cu2 atoms is ach-
ieved by the tertiary amine and two pyridine nitrogen
atoms. However, the complex possesses geometrical unsym-
metry because the environment around the Cu2 center is
found to be intermediate between square pyramidal and
trigonal bipyramidal (t=0.53),[16] unlike that around the
Cu1 center (t=0.06). The same dissymmetry exists in the
enzyme (Table 1). In contrast, complexes 1 and 2 present
similar coordination geometry for each copper center (two
distorted trigonal bipyramids for 1 and two square pyramids
for 2).
Results and Discussion
Synthesis, structures and solution properties of the model di-
nuclear copper(II) complex and the OPNO adduct
Synthesis: The unsymmetrical HBPMEP ligand (9) was syn-
thesized according the procedure outlined in Scheme 2.[17]
For the introduction of the different coordination sites, the
key step in the synthesis is realized by using synthon 4.[18]
Substitution of the benzylic bromine group by N,N-bis(2-
ethylpyridyl)amine (5) afforded 6. After deprotection to
form 7, the hydroxymethyl group is chlorinated for alkyla-
tion with the second amine (amine 8, which is different from
the first one) to afford the desired ligand 9.
The dibridged dinuclear copper(II) complex 3 was pre-
pared by the addition of two equivalents of CuACHTUNGTRNEG(UN ClO4)2·6H2O
to a solution of the HBPMEP ligand in acetonitrile contain-
ing two equivalents of triethylamine (Scheme 2). The solid
Table 1. t factors for copper sites in the met form of S. castaneoglobispo-
rus Ty[2a] (CuA, CuB) and complexes 1–3 (Cu1 and Cu2).
CuA/Cu1
CuB/Cu2
S. castaneoglobisporus Ty met form
0.22
0.64
0.05
0.06
0.48
0.60
0.09
0.53
1
2
3
complex [Cu2ACHTUNGTRENNUNG(BPMEP)ACHTUNTREG(NGNUN m-OH)]CAHTNUGTRNEN(UGN ClO4)2 (3) precipitated
ꢀ
ꢀ
upon addition of tetrahydrofuran and green crystals of the
compound suitable for X-ray single-crystal analysis were ob-
tained by slow tetrahydrofuran diffusion into an acetonitrile
solution of 3. The reaction of complex 3 with 2.5 equivalents
of HOPNO in methanol and slow vapor diffusion of diethyl
The Cu1 O1 Cu2 bond angle is 101.06(8) ꢃ. The pheno-
ꢀ
late ring is slightly tilted toward the O1 O2 axis and twisted
ꢀ
ꢀ
ꢀ
ꢀ
with respect to the O1 Cu2 O2 Cu1 plane. The Cu O dis-
tances in the [Cu2O2] unit are in the range of 1.8995–
1.9876 ꢃ and are slightly more asymmetric than those in 2
(1.928–1.961 ꢃ). The four Cu1, Cu2, O1, and O2 atoms are
ether afforded the resulting adduct [Cu
N
2ACHTUNGTRNE(NUNG BPMEP)-
(OPNO)]2+ (102+) as a crystalline material.
in the same plane. The Cu1 O1 Cu2, Cu1 O2 Cu2, O1
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
Cu1 O2, and O1 Cu2 O2 angles are 101.08, 104.68, 76.88,
and 77.28, respectively, which leads to the intermediate dis-
tance between Cu1 and Cu2 of 3.045 ꢃ relative to those for
1 (2.966 ꢃ) and 2 (3.077 ꢃ).
The reaction of complex 3 with HOPNO in methanol and
slow vapor diffusion of diethyl ether afforded compound 10
Crystal structure descriptions of dications 32+and 102+: All
of the crystal structure parameters are given in the Experi-
mental Section. The crystal of [Cu
(ClO4)2·THF (3·THF; THF: tetrahydrofuran) contains one
associated solvent molecule. An ORTEP view of the dica-
2ACHUTGTNREN(UNG BPMEP)CAHTUNGTRENN(UGN m-OH)]-
G
Chem. Eur. J. 2013, 19, 3655 – 3664
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3657

C. Belle, H. Jamet et al.
as a crystalline material. The unit cell contains two crystallo-
graphically independent [Cu2A(BPMEP)(OPNO)]-
(ClO4)2·1= solvate entities, in which solvate stands for metha-
Solution studies: The ESI-MS spectrum of the 3-OPNO (10)
adduct provides two signals at m/z 894 (z=1) and 396.5 (z=
G
ACHTUNGTRENNUNG
R
2) for [Cu
+A
2ACHTNUGTERN(NNUG BPMEP)ACHUTTNGERNNUG(OPNO)] CHTUNGRENTN(UNG ClO4) and [Cu2ACHTNUGTRNE(NUGN BPMEP)-
2
nol and H2O. Each cationic unit, one of which (10a2+) is
shown in Figure 2B (10b2+ is shown in Figure S1 in the Sup-
porting Information), consists of two CuII atoms bridged by
a phenoxo group and one deprotonated HOPNO molecule
coordinated in an unsymmetrical bridging bidentate mode.
Selected bond lengths and angles are reported in Table S2 in
the Supporting Information and show no significant differ-
ence between the two copper dimers. The coordination
sphere of each copper atom is completed by one amino and
two pyridyl nitrogen atoms. Inside each dimer, the Cu atoms
are separated by 3.242 ꢃ in 10a and 3.228 ꢃ in 10b. This
distance, which is longer than in the dibridged (phenoxo and
hydroxo) dicopper complex 3 (3.045 ꢃ), is comparable with
the one in 2-OPNO (3.271 ꢃ) but much shorter than that in
adduct 1-OPNO (3.927 ꢃ).
(OPNO)]2+, respectively (Figure S2 in the Supporting Infor-
mation). These data evidence that the dibridged dimeric
copper(II) structure remains (in part) in solution, as already
observed for adduct 2-OPNO. The general UV/Vis features
for dicopper(II) complex
(DMSO; 95/5) solution buffered at pH 7.0 with 50 mm
2-[4-(2-hydroxyethyl)-1-piperazinyl]ethanesulfonic acid
3 in H2O/dimethylsulfoxide
(HEPES) are reported in Table 2 and can be compared with
Table 2. UV/Vis spectral data and binding constants for 3 and for ad-
ducts 1-OPNO, 2-OPNO, and 3-OPNO (10).
UV/Vis: lmax
Binding constant
[nm], (e [mꢀ1 cmꢀ1])
(Kb) [mꢀ1
]
LMCT
d–d
3[a]
337 (1440)
386 (1220)
430 (610)
429 (5200)
420 (1530)
645 (200)
The coordination polyhedrons around the Cu2 (in 10a)
and Cu4 (in 10b) atoms are distorted square pyramids, the
apical positions of which are occupied by the pyridine nitro-
gen atoms N6 and N13, respectively. The square plane is de-
fined by the N2, N5, O1, and O3 atoms in 10a (N9, N12,
O4, and O5 in 10b). The OPNO ligand is coordinated to
Cu1 and Cu2 (or Cu3 and Cu4) in a bidentate mode. With
1-OPNO[b,15]
2-OPNO[b,15]
3-OPNO[b]
700 (212)
670 (490)
650 (170)
440ꢂ70
8000ꢂ400
27100ꢂ9750
[a] Determined at 258C from 2.7ꢄ10ꢀ4 m solution in H2O/DMSO (95/5)
buffered at pH 7 (HEPES, 50 mm). [b] Determined from the electronic
spectra calculated from the refinement of absorbance data by the Specfit
program.
ꢀ
ꢀ
the weak Cu1 O2 (or Cu3 O5) interaction (approximately
2.69 ꢃ) taken into consideration, the coordination environ-
ment around Cu1 (or Cu3) forms a distorted octahedral ge-
ometry with the N1 (or N8) nitrogen atom of the tertiary
amine, the N3 (or N10) pyridine nitrogen atom of the
ligand, the O3 (or O6) oxygen atom from the OPNO ligand,
and the O1 (or O4) atom of the phenoxo moiety in the
equatorial coordination positions.
The axial coordination positions are occupied by the
second oxygen atom (O2 or O5) from the OPNO ligand and
the N4 (or N11) pyridine nitrogen atom. Angles involving
the Cu atoms show deviation from the theoretical angles
the UV/Vis parameters determined from the calculated elec-
tronic spectra for all of the adducts from refinement of ab-
sorbance data by the Specfit program.[19] All of the com-
plexes are characterized by a ligand-to-metal charge-transfer
(LMCT) transition between the bridging phenoxo group
and the copper ions in the 300–450 nm region and a d–d
transition in the 600–750 nm region (Table 2 and Figure S3a
and b in the Supporting Information).[20]
The UV/Vis spectra of 3 are quite similar to those of 2
with a feature at 386 nm (eꢁ1220mꢀ1 cmꢀ1), corresponding
to the LMCT band maximum, a shoulder at 337 nm (e
ꢁ1400mꢀ1 cmꢀ1), and a d–d band at 645 nm. Despite the dif-
ference in the geometry around the copper centers of 2 and
3 that was observed in the solid state, the similar d–d transi-
tion position could indicate geometrical changes in solution.
Binding of HOPNO on complex 3 (Figure S4 in the Sup-
porting Information) occurs with a shift of the absorption
band at 386 nm up to 420 nm. The molar extinction coeffi-
cient is raised from 1220 to 1530mꢀ1 cmꢀ1. The d–d transition
band at 645 nm (e=200mꢀ1 cmꢀ1) of 3 is slightly shifted to
650 nm (e=170mꢀ1 cmꢀ1), with a small decrease in the molar
extinction coefficient. These data are reported in Table 2, to-
gether with the binding constants (Kb) of HOPNO to 1–3,
determined from the refinement of absorbance data versus
the HOPNO concentration by the Specfit software. The
series of curves exhibit an isosbestic point, which indicates
that one absorbing species is formed. The Kb value found
with 3 is significantly higher than those obtained for com-
plexes 2 and 1, respectively,[15] and the stoichiometry of the
adduct is found to be 1:1. The UV/Vis spectrum from isolat-
ꢀ
ꢀ
(the largest deviations in 10a are encountered for O2 Cu1
ꢀ
ꢀ
N4 (149.08, 318 deviation) and O1 Cu1 O2 (65.808, 24.28
deviation)), which emphasizes the irregularity of the octahe-
ꢀ
dron. The longer distances of the axial coordination (Cu1
ꢀ
ꢀ
O2: 2.695(2) ꢃ for 10a; Cu3 O5: 2.699(2) ꢃ for 10b; Cu1
ꢀ
N4: 2.207(2) ꢃ for 10a; Cu3 N11: 2.209(3) ꢃ for 10b) are
evidence of a strong Jahn–Teller effect.
In the 1-OPNO adduct, the OPNO ligand binds only one
copper atom of the dinuclear site in a chelating mode, in
contrast to the 2-OPNO and 3-OPNO (10) adducts, in which
the bound OPNO ligand is in a bridging mode between the
two copper(II) centers. The major difference between the 2-
OPNO and 10 adducts is the unsymmetrical coordination
geometry around the copper centers. As a result, the coordi-
nation mode of the bound OPNO ligand resembles the ex-
pected coordination mode of the diphenol substrate on its
way to product formation[1b] or the proposed model from
XAS studies of the Kojic acid adduct with S. antibioticus ty-
rosinase.[11]
3658
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 3655 – 3664

Unsymmetrical Binding Modes of Tyrosinase Inhibitors
FULL PAPER
ed 3-OPNO (10; 2.7 10ꢀ4 m) displays bands at 416 nm (e
ꢁ1629mꢀ1 cmꢀ1) and 649 nm (eꢁ191mꢀ1 cmꢀ1), values that
are consistent with those obtained from the calculated elec-
tronic spectrum (Table 2).[21]
The product of the magnetic susceptibility with temperature
(cT) versus temperature plots for polycrystalline samples of
3 and 10 are given in Figure S6 in the Supporting Informa-
tion. At room temperature, the cT values are much lower
than the expected values for two magnetically independent
CuII ions (S=1/2; g=2). Upon cooling, the values decrease
rapidly and vanish at 75 K and 90 K for 3 and 10, respective-
ly. These behaviors suggest the presence of strong antiferro-
magnetic interactions between the two CuII ions and singlet
states as the ground states. For 10, this result in conjunction
with the EPR spectrum in frozen solution highlights the dif-
ference of behavior mentioned before between the solid
state and solution. Good fits of the experimental data with
the equation given in the Experimental Section are obtained
with J=ꢀ210 cmꢀ1 and g=2.4 for 3 and J=ꢀ225 cmꢀ1 and
g=2.23 for 10. No intermolecular interactions are consid-
ered.
The electronic structures of dinuclear copper(II) com-
plexes remain a theoretical challenge due to the difficulty of
providing an accurate description of electron correlation ef-
fects for these systems.[22] In this case, the size of the systems
suggests the use of density functional methods. The approxi-
mation was made that the triplet and singlet state geome-
tries are close and all geometries have been determined
from a triplet state. Selected structural parameters for com-
plex 3 are listed in Table S1 in the Supporting Information.
As for complexes 1 and 2,[15] a good agreement between cal-
culated and experimental data is obtained and confirms that
our approximation is correct. The computed values of the J
exchange coupling for complexes 1–3, calculated by using a
broken symmetry approach,[23] are ꢀ136, ꢀ636, and
ꢀ394 cmꢀ1, respectively. These results are supported by ex-
perimental magnetic measurements. The determined J value
for 1 is ꢀ112 cmꢀ1,[14] for 2 is < ꢀ300 cmꢀ1,[15] and for 3 is
ꢀ210 cmꢀ1 in this work. Analysis of the [Cu2O2] core geome-
tries explains this order. From Table 3, it can be observed
EPR spectroscopy: The EPR spectra were measured at
100 K in a frozen solution in 1 mm H2O/DMSO (1/1) buf-
fered at pH 7 with 50 mm HEPES. Complex 3 is EPR silent,
in accordance with the strong antiferromagnetic-coupling
(S=0) ground spin state between the two copper atoms.
This observation is consistent with the doubly bridged struc-
ture and the short Cu…Cu distance (3.045 ꢃ) evidenced in
the solid state. The EPR measurements for 10 in a frozen
solution of DMSO/H2O (1/1) at 100 K do not reveal the ex-
pected silent behavior for antiferromagnetic-coupled com-
pounds, as with 3 or 2-OPNO. The spectrum displays the
presence of two sets of hyperfine lines from 220–400 mT,
which suggests the overlap of two signals of uncoupled cop-
per(II) ions (Figure S5 in the Supporting Information) that
probably result from the dissociation of 10 in solution. A
comparison with the spectrum of
a solution of Cu-
ACHTUNGTRENNUNG(ClO4)2·6H2O with two equivalents of HOPNO indicates
that this corresponds to one of the signals observed in the
spectrum of 10 (Figure S5 in the Supporting Information).
These data show that some of the copper ions released from
10 in solution are complexed by HOPNO as a mononuclear
species. Integration of the signal shows that 50% of the ex-
pected signal is recovered. This indicates that a magnetically
coupled (EPR-silent) species is also present (adduct 10 in
the solid state exhibits antiferromagnetic coupling between
the copper ions; see below). Altogether, these data can be
analyzed as evidence of changes in the copper coordination
geometry after dissolution of 10 in a DMSO/H2O (1/1) solu-
tion, so the observed EPR spectrum of 10 corresponds to a
mixture of coupled and uncoupled species.
ꢀ ꢀ
Catecholase activity: The catalytic activity of 3 in the oxida-
tion of 3,5-di-tert-butylcatechol (3,5-DTBC) was studied in
acetonitrile by monitoring the time dependence of the ab-
sorption of 3,5-di-tert-butylquinone (3,5-DTBQ) at 400 nm,
which corresponds to the absorption of the product (e=
1900mꢀ1 cmꢀ1). The reactions follow a Michaelis–Menten be-
havior. The reaction rates are determined from the slope of
the trace at 400 nm in the first 20 minutes of the reaction
and a Lineweaver–Burk treatment gives a maximum veloci-
ty (Vmax) of 1.2ꢄ10ꢀ6 mꢀ1 sꢀ1and a Michaelis constant (KM) of
5 mm. After 90 min, a turnover number (TON) of 12 is ob-
served, which is near the range of the values observed with
complexes 1[14] and 2[15] (TON values of 16 and 18.3, respec-
tively). Changes in geometry around the copper centers due
to different pendant pyridine arms (with one or two carbon
atoms) and unsymmetry are not critical for the catalytic effi-
ciency of these complexes.
that the values of the Cu O Cu angles for 3 are intermedi-
ate with those values for 1 and 2, in correlation with their J
values.
Table 3. Calculated exchange coupling constants (J [cmꢀ1]) and Cu O
Cu angles [8] for complexes 1–3 (B3LYP/6-311g*).[a]
ꢀ
ꢀ
ꢀ
95.7
101.0
103.7
ꢀ
ꢀ
102.1
104.7
105.5
ꢀ
J
Cu O Cu
Cu O(H) Cu
1
3
2
ꢀ136 (ꢀ112)[14]
ꢀ394 (ꢀ210)
ꢀ636 (<ꢀ300)[15]
[a] Values in parenthesis correspond to experimental values.
Calculations were also done for the complexes formed
with HOPNO. Selected structural parameters for 10 (3-
OPNO) are listed in Table S2 in the Supporting Informa-
tion. A good agreement between the calculated and experi-
mental data is also obtained. The OPNO ligand bridges the
two copper atoms in a bidentate mode, as for 2-OPNO, but
the symmetrical coordination of 2-OPNO around the
copper atoms is lost in 10. The computed value of the J ex-
Magnetic properties and DFT calculations
Chem. Eur. J. 2013, 19, 3655 – 3664
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3659

C. Belle, H. Jamet et al.
change coupling for 3-OPNO (10) is ꢀ291 cmꢀ1, which is in
agreement with the experimental data given above
(ꢀ225 cmꢀ1). This value is slightly smaller than that of 2-
OPNO, computed to beꢀ391 cmꢀ1.[15] This can be explain by
HOPNO is also a competitive inhibitor for the bacterial Ty
from S. antibioticus. Indeed, S. antibioticus Ty inhibition by
HOPNO is concentration dependent with an IC50 value esti-
mated to be (6.4ꢂ1.1) mm ([l-DOPA]=0.175 mm; Figure S7
in the Supporting Information) and it exhibits the kinetic
features of competitive inhibition (KI =(7.7ꢂ0.2) mm; Fig-
ure S8 and S9 in the Supporting Information) with efficiency
comparable to the mushroom Ty form.
To calculate the binding mode of deprotonated HOPNO
in the enzyme, the optimized geometries obtained for
OPNO complexes were used as starting points for the place-
ment of the inhibitor into the active pocket of the enzyme.
The construction of the initial geometry of these complexes
is detailed in the Experimental
ꢀ ꢀ
the small decrease in the Cu O Cu angle if we compare 10
ꢀ ꢀ
with 2-OPNO. The value of the Cu O Cu angle is 114.28
for 10 and 115.58 for 2-OPNO. Relative to those in 2 and 3,
the two Cu atoms for the OPNO complex are less coupled.
Figure 3 represents the magnetic orbitals of 10 and 3. The
exchange interaction passes essentially through the oxygen
ꢀ ꢀ
atom of the Cu O Cu bridge; the exchange path is broken
through the OPNO ligand. This is not the case for 2 and 3,
in which a path is developed on the [Cu2O2] core.
Section. Geometry optimiza-
tions were done on triplet
states, with the assumption that
the spin state doesn’t change
considerably the geometry of
the active state, as observed in
model complexes. The same
minimum was obtained if the
optimized geometry for 2-
OPNO or 3-OPNO was used;
thus, only two minima between
the OPNO ligand and the bac-
terial Ty were obtained. They
can be seen in Figure 4 and dis-
tances between the ligand and
the copper atoms of the active
site are listed in Table S3 in the
Supporting Information.
In the first structure (Fig-
ure 4A), the OPNO ligand
bridges the two copper atoms
as in 2-OPNO and 3-OPNO.
The intermetallic distance from
Cu1…Cu2 is close to the dis-
tance obtained for the opti-
mized geometries of 2-OPNO
and 3-OPNO (3.32 ꢃ for the
bacterial Ty and 2-OPNO,
3.36 ꢃ for 3-OPNO). As in 3-
OPNO, the binding mode is un-
symmetrical. The distance be-
tween the oxygen atom of the
nitrosyl group and one copper
atom is 1.98 ꢃ, whereas the dis-
Figure 3. Magnetic orbitals for the broken-symmetry state of a) 3 and b) 10.
tance between the oxygen atom
of the hydroxy group and the
other copper atom is 2.11 ꢃ.
Binding studies on the enzyme
For bacterial Ty, this is explained by the presence of a serine
group, Ser260, close to the active site, which can form a hy-
drogen bond with the oxygen atom of the OPNO ligand
linked to Cu2 (see Figure 4A). This interaction also explains
the structure of the second minimum obtained for the
enzyme (Figure 4B). From the chelating mode of 1-OPNO,
HOPNO has already been described as a competitive inhibi-
tor in the oxidation of l-3,4-dihydroxyphenylalanine (l-
DOPA) catalyzed by mushroom Ty (IC50 value of 1.2 mm
and inhibition constant (KI) of 1.8 mm).[10] We confirmed that
3660
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 3655 – 3664

Unsymmetrical Binding Modes of Tyrosinase Inhibitors
FULL PAPER
hibitor phenylthiourea (PTU) and a histidine residue close
to the active site of a catechol oxidase (PDB code:
1BUG).[24]
To compare the two binding modes, we can look at the
QM/MM energy of the system. However, the monodentate
mode (Figure 4B) has an extra water molecule on the free
copper atoms of the active site. So, a single-point calculation
has been performed for the monodentate mode without the
water molecule. This mode was found to be favorable by
9.8 kcalmolꢀ1. This comparison indicates that the OPNO
ligand prefers to be coordinated to only one copper atom,
as we found previously for the dinuclear copper(II) complex
[Cu2CAHTUNGTREN(UNNG BPMP)AHCUTNRTGENG(UNN m-OH)]ACHTNGURTEN(NUGN ClO4)2. For the copper complex, this
coordination is explained by the rigidity of the chelate ring
that forms the complex. In the enzyme, it should be ex-
plained by the steric constraints imposed by the environ-
ment. However, different approximations are introduced in
this comparison, for example, the need to perform certain
statistical simulations to obtain average structures and ener-
gies for the enzyme. Thus, this comparison can be seen as a
first approach.
Conclusion
An understanding of the different parameters that control
the binding mode of transition-state analogue inhibitors of
Ty is an important step to unravel the structural features
that govern tyrosinase activity. HOPNO is an efficient com-
petitive inhibitor of mushroom and bacterial tyrosinases.
From a series of dicopper(II) complexes based on different
lengths of the pyridine side arms, diverse OPNO adducts
have been described (chelating, bridging). In this article, we
present the binding mode found on a new Ty functional
model. The latter involves two different complexing arms
(namely (bis(2-ethylpyridyl)amino)methyl and (bis(2-meth-
ylpyridyl)amino)methyl) designed to tune an unsymmetrical
binding of the OPNO ligand. These variations in binding
modes were used as starting points to study the binding
mode of the OPNO ligand in the enzyme. QM/MM calcula-
tions suggest that the OPNO ligand coordinated to only one
copper atom in the enzyme, as in the most rigid complex 1.
However, one hydrogen bond between the ligand and a
serine residue close to the active site changes considerably
the binding mode in the enzyme compared to that in 1. This
result highlights the importance of the second coordination
sphere in the enzyme.
Figure 4. QM/MM optimized complexes between bacterial Ty and the
OPNO ligand. Two binding modes are obtained: a) a bridging mode; b) a
monodentate mode. The ligand, the copper atoms, and the hydroxo
group bridging the two copper atoms are displayed as balls and sticks.
Copper atoms are shown in orange.
the ligand shifted during the optimization geometry to give
a hydrogen bond between the oxygen atom of the hydroxy
group of the OPNO ligand and Ser260. In this case, the co-
ordination between the Cu2 atom and the oxygen atom is
broken and the binding mode becomes monodentate.
For the two binding modes in the enzyme, the aromatic
ring of the OPNO ligand is p stacked with His194 of the
enzyme. Figure 4 illustrates this interaction for the two
structures. The distances between the OPNO ligand and
His194 are similar in the two structures, which suggests an
interaction of the same strength. This interaction was al-
ready observed in the crystallographic structures of type-3
copper enzymes. For example, it was found between the in-
Experimental Section
General: All starting materials were commercially available and used as
purchased, unless stated. ESI mass spectra were recorded on an Esquire
300 plus Bruker Daltonis instrument. Elemental analyses were performed
at the analysis facilities of the Department of Chemistry of the University
of Grenoble, France. UV/Vis absorption spectra were obtained by using
a Varian Cary 50 spectrophotometer operating in the 200–1000 nm range
with quartz cells. The temperature was maintained at 258C with a tem-
Chem. Eur. J. 2013, 19, 3655 – 3664
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3661

C. Belle, H. Jamet et al.
1
perature-control unit. H NMR data were recorded on a Bruker Advance
Catechol oxidase activity and kinetics: UV/Vis spectra and kinetics were
recorded by using a quartz cuvette (1.0 cm) and a Varian Cary 50 UV/Vis
spectrophotometer equipped with a Peltier thermostating accessory. The
kinetics of the 3,5-DTBC oxidation was determined by the initial-rates
method by monitoring the increase in the product (3,5-DTBQ) at
400 nm. All of the kinetic experiments were conducted at a constant tem-
perature of 258C, monitored with a thermostat. A 4.8ꢄ10ꢀ5 m solution of
the complex with 3,5-DTBC (100 equiv) was used and the solution was
saturated with dioxygen (1 atm). Under these conditions, catechol was
not oxidized without the copper complexes. The measurements were per-
formed in CH3CN. Under these conditions, no changes were observed
from the initial solutions of the complexes (UV/Vis spectrophotometry).
The kinetics parameters were determined for a 4.8ꢄ10ꢀ5 m solution of
complex 3 and 0.96–44.8 mm solutions of the substrate.
300 spectrometer at 323 K with the deuterated solvent as a lock. X-band
EPR spectra were recorded at 100 K with a Bruker ESP EMX Plus spec-
trometer equipped with
9.3 GHz (X band).
a nitrogen-flow cryostat and operating at
Safety note: Although no problems were encountered during the prepa-
ration of the perchlorate salts, suitable care should be taken when han-
dling such potentially hazardous compounds.
Enzymatic essays: The S. antibioticus tyrosinase activity was checked
with spectroscopic methods by using l-DOPA as a substrate. HOPNO
was dissolved in a 10% DMSO stock solution. Phosphate buffer (pH 7.0)
was used to dilute the DMSO stock solution of the compounds. S. anti-
bioticus tyrosinase (1 mL; 4.5 UmLꢀ1) was first preincubated with the
compounds in 50 mm phosphate buffer (pH 7.0) for 5 min at 258C. The l-
DOPA was then added to the reaction mixture and the enzyme reaction
was monitored by measuring the change in absorbance at 475 nm of the
DOPAchrome product for 5 min.
Crystal structure determination and refinement: Single crystals of com-
plexes 3·THF and 10·solvate were mounted on a Kappa CCD Nonius dif-
fractometer equipped with graphite-monochromated MoKa radiation (l=
0.71073 ꢃ) at 200 K. The reflections were corrected for Lorentz and po-
larization effects and for absorption by using the SADABS program.[25]
The data were then integrated and scaled with EvalCCD software.[26] The
structures were solved by direct methods and refined by full-matrix least-
square methods with the SHELXL program[27] implemented in Olex2
software.[28] For both structures, all non-hydrogen atoms were refined
with anisotropic thermal parameters. Hydrogen atoms were generated in
idealized positions, riding on the carrier atoms, with isotropic thermal pa-
rameters, except for the hydrogen atom of the hydroxy group, which was
localized on a difference Fourier map and was added. Pertinent crystallo-
graphic data and refinement details are summarized in Table 4.
CCDC 881633 (3·THF) and CCDC 881634 (10·solvate) contain the sup-
plementary crystallographic data for this paper. These data can be ob-
tained free of charge from the Cambridge Crystallographic Data Centre
via www.ccdc.cam.ac.uk/data_request/cif.
Synthesis: The binucleating ligand HBPEMP was prepared by the
method described in the Supporting Information.
Complex 3:
A solution of CuACHTUNGTNRUEGN(ClO4)2·6H2O (470 mg; 1.26 mmol) in
CH3CN (10 mL) was added dropwise to HBPEMP (312 mg; 0.56 mmol)
dissolved in CH3CN (15 mL). The initial yellow solution turned dark
green. Et3N (160 mL; 1.12 mmol) was then added and the solution was
stirred for 3 h at room temperature. The solvent was partially removed
under reduced pressure and a few drops of THF were added. The result-
ing solution (10 mL) was allowed to stand for 3 days at ꢀ208C. A green
powder (256 mg) was collected by filtration (yield: 51%). Crystals of X-
ray quality were obtained by vapor diffusion of THF in a solution of 3 in
CH3CN. The crystalline material, if crushed and exposed to air, loses
some of the noncoordinated THF; UV/Vis (in HEPES (pH 7.0, 50 mm),
DMSO (5%)): lmax (e)=390 (1380), 650 nm (180mꢀ1 cmꢀ1); ESI-MS:
m/z: z=1, 801.1 [M+H+ꢀClO4ꢀ]; z=2, 350.1 [M+H+ꢀ2ClO4ꢀ]; ele-
mental analysis: calcd (%) for C35H38Cl2Cu2N6O10·0.5C4H8O: C 47.44, H
4.62, N 8.79; found: C 47.37, H 4.07,
Magnetic measurements: Magnetic susceptibility data (2–300 K) were
collected on powdered samples by using a SQUID magnetometer from
N 8.14.
Complex 10: A solution of HOPNO
(15.7 mg, 2.5 equiv) in methanol
(2 mL) was added to a solution of
complex 3 (40 mg, 0.44 mm) in meth-
anol (10 mL) and a small amount of
acetone. The brown solution was stir-
red for 30 min. A slow diffusion of di-
ethyl ether into the methanolic solu-
tion provided 10 as a suitable crystal
for X-ray analysis. The crystalline
material, if crushed and exposed to
air, loses the noncoordinated solvent;
Table 4. Crystallographic data for 3 and 10.
3
10
formula
A
G
2[Cu2(C40H41N7O3)]
2017.15
brown plate
0.4ꢄ0.22ꢄ0.1
triclinic
ACHTUGNTRENUN(GN ClO4)2·0.85MeOH·0.15H2O
MW
color and morphology
crystal size [mm]
crystal system
space group
a [ꢃ]
b [ꢃ]
c [ꢃ]
a [8]
green rectangular prism
0.39ꢄ0.18ꢄ0.14
monoclinic
P1 21/c 1
12.899(2)
30.5659(10)
10.394(3)
90.00
Pꢀ1
12.5531(16)
18.427(4)
19.109(3)
99.903(11)
101.645(11)
95.829(13)
4221.9(12)
1.587
ESI-MS:
m/z:
z=1,
z=2,
894.1
396.5
[MꢀClO4ꢀ];
b [8]
95.054(8)
90.00
4082.4(13)
1.583
200.0
4
[Mꢀ2ClO4ꢀ]; UV/Vis (in HEPES
g [8]
V [ꢃ3]
(pH 7, 50 mm), DMSO (5%)): lmax
1 [gcmꢀ3
T [K]
Z
]
(e)=420
(1530),
630 nm
(200mꢀ1 cmꢀ1); elemental analysis:
calcd (%) for C40H41Cl2Cu2N7O11: C
48.34, H 4.16, N 9.87; found: C 48.70,
H 4.20, N 9.83.
200.0
2
1.205
57438
m [mmꢀ1
total
]
1.240
46213
reflns
Determination of stability constants:
The binding constants of the inhibitor
to the dinuclear CuII complexes were
determined by spectrophotometric ti-
tration in water/DMSO (95/5) solu-
tions buffered at pH 7.0 with 50 mm
HEPES. The titrations were carried
out by adding small aliquots of a con-
centrate solution of the inhibitor to
solutions of the complexes. The spec-
tral data were refined with the Spec-
fit program.[19]
unique
reflns
observed
reflns
8436
16230
6506 (F>2s)
11882 (F>2s)
Rint
0.0377
0.0356
0.0763
1.031
0.0447
0.0434
0.0896
1.061
R[a]
R(w)[a]
GOF
D1min/D1max [eꢃꢀ3
]
ꢀ0.505/0.611
ꢀ0.447/0.535
[a] Refinement based on F if w=1/[s2(Fo)+(0.03499P)2 +3.5696P] for 3·THF and w=1/[s2(Fo)+(0.0366P)2 +
6.1205P] for 10 and with P=(Fo2 +2Fc2)/3.
3662
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 3655 – 3664

Unsymmetrical Binding Modes of Tyrosinase Inhibitors
FULL PAPER
Quantum Design (model MPMS-XL) in a 0.1 T applied magnetic field.
All data were corrected for the contribution of the sample holder and di-
amagnetism of the samples estimated from Pascalꢀs constants.[29] The
magnetic exchange interactions (J) were extracted by simulation of the
magnetic susceptibility by considering the spin Hamiltonian H=ꢀ2JS1S2
Acknowledgements
The authors thank the French Agence Nationale pour la Recherche for
financial support (ANR-09-BLAN-0028-01/02) and for grants to C.B. and
C.D. We are grateful to Florian Molton (DCM) for measuring EPR spec-
tra. This work has been carried out in the framework of COST action
CM1003 (WG 2) and of Labex ARCANE.
(Si =1= ) and by using Equation (1) in which J is the exchange interaction
2
parameter, T the temperature, b the Bohr magneton, k the Boltzmannꢀs
constant, g the Landꢁ g-factor and N the Avogadroꢀs number.[29b]
Ng2b2
k½3 þ expðꢀ2J=kTÞꢃ
[1] a) R. H. Holm, P. Kennepohl, E. I. Solomon, Chem. Rev. 1996, 96,
2239–2314; b) M. Rolff, J. Schottenheim, H. Decker, F. Tuczek,
Chem. Soc. Rev. 2011, 40, 4077–4098; c) A. W. J. W. Tepper, E. Lo-
nardi, L. Bubacco, G. W. Canters in Handbook of metalloproteins
(Ed.: A. Messerschmidt), John Wiley, Chichester, 2010, DOI:
10.1002/0470028637.met0470028265.
[2] a) Y. Matoba, T. Kumagai, A. Yamamoto, H. Yoshitsu, M. Sugiya-
ma, J. Biol. Chem. 2006, 281, 8981–8990; b) M. Sendovski, M. Kant-
eev, V. Shuster Ben-Yosef, N. Adir, A. Fishman, J. Mol. Biol. 2011,
405, 227–237.
ð1Þ
cT ¼
QM and QM/MM calculations: QM calculations were based on DFT and
have been performed with the Gaussian03 package.[30] Full geometry op-
timizations were carried out by using the hybrid functional B3LYP[31] and
the 6-31g* basis set on all atoms.[32] All geometry optimizations were
done on a triplet state. The Yamagachi formula[33] was used to estimate
the exchange coupling constants. For that purpose, additional single-point
high-spin (HS) and broken-symmetry (BS) calculations were done on the
previously optimized geometries with the same B3LYP functional and a
larger basis set (6-311g* for Cu and coordinating O atoms, 6-31g* for all
remaining atoms).
[3] B. Albolmaali, H. Taylor, U. Weser in Bioinorganic chemistry, 91st
ed. (Eds.: J. G. M. Clarke, C. Jorgensen, D. Mingos, G. Palmer, P.
Sadler, R. Weiss, R. Willimans), Springer, Berlin/Heidelberg, 1998,
pp. 91–190.
QM/MM calculations were performed by using the Gaussian03 package
for the QM part and Tinker 4.2[34] for the MM part. Both mechanical and
electrostatic embedding (direct polarization) schemes were used to de-
scribe interactions between the QM and MM part, thus ensuring a good
description of the effects of the environment on the active site. The QM
part contains the two copper atoms, the hydroxy group, the side chains of
the six histidine residues linked to the copper atoms in the met form, the
ligand, and an extra water molecule for the chelating mode. All MM
atoms belonging to residues within a distance of 4 ꢃ of the QM part
were allowed to relax during geometry optimizations and the remaining
structure was kept frozen. All QM/MM geometry optimizations were car-
ried with a triplet state. The partition between the QM and MM parts
[4] H. S. Mason, J. Biol. Chem. 1948, 172, 83–99.
[5] a) V. N. Sehgal, G. Srivastava, Int. J. Dermatol. 2008, 47, 1041–1050;
b) R. E. Boissy, J. J. Nordlund, Pigment Cell Res. 1997, 10, 12–24.
[6] C. Visffls, R. Andres, J. I. Mayordomo, M. J. Martinez-Lorenzo, L.
Murillo, B. Saez-Gutierrez, C. Diestre, I. Marcos, P. Astier, F. J.
Godino, F. J. Carapeto-Marquez de Prado, L. Larrad, A. Tres, Mela-
noma Res. 2007, 17, 83–89.
[7] Y. Xu, A. H. Stockes, W. M. Freeman, S. C. Kumer, B. A. Vogt,
K. E. Vrana, Mol. Brain Res. 1997, 45, 159–162.
[8] T. S. Chang, Int. J. Mol. Sci. 2009, 10, 2440–2475.
[9] a) J. S. Chen, C. Wei, M. R. Marshall, J. Agric. Food Chem. 1991, 39,
1897–1901; b) R. C. Hider, K. Lerch, Biochem. J. 1989, 257, 289–
290.
ꢀ
has led to the cut of the six Ca Cb bonds of the histidine residues in the
active site. These dangling bonds have been described as strictly localized
bonding orbitals (SLBOꢀs) within the local self-consistent field (SCF)[35]
methodology. In this scheme, SLBOꢀs remain frozen during the SCF pro-
cedure. A detailed description of the SLBOꢀs and parameters can be
found in the Supporting Information and related references. The QM
part was treated at the B3LYP/6-31g* level of theory and the MM part
was described with the Amber 99SB force field[36] for the protein and
TIP3P for the solvent. The initial structure for the enzyme is based on
the crystal structure of the bacterial tyrosinase complexed with the
ORF38 caddie protein (PDB code: 2AHK).[2a] The caddie protein and
the water molecules, with the exception of three water molecules close to
the active site, were removed. The addition of H atoms leads to the pro-
tonation of all histidines on the d nitrogen atom, whereas the aspartate
and glutamate residues remain deprotonated and the arginine and lysine
residues are positively charged. The system was first solvated with a trun-
cated octahedral box of TIP3PBOX water molecules (10 ꢃ radius) and
neutralized with one Na+ ion. For the first energy minimization, the Am-
ber99SB force field was used in conjunction with specific parameters de-
veloped for copper ions, histidine residues, and the hydroxy group linked
to the copper atoms in the active site. These parameters are given in the
Supporting Information and are based on the force field developed by
Comba and Remenyi[37] for blue copper proteins. The system was energy
minimized for 3000 steepest descent steps followed by 12000 conjugated
gradient steps. No constraints were applied in order to allow complete re-
laxation of the system. The QM optimized structures of 1-OPNO, 2-
OPNO, and 3-OPNO were then aligned with the active site of the
enzyme with the ꢅfit atomꢀ tool of Sybyl version 8.0.[38] Alignment was
done between the His Ne and Cu atoms from the enzyme and the N and
Cu atoms of the OPNO complexes. Some atoms of the complexes were
then introduced into the enzyme pocket to replace the copper atom and
hydroxy group of the met Ty form. Figure S10 in the Supporting Informa-
tion illustrates the introduction of the OPNO complexes into the enzyme
structure. Once the relaxed structure was obtained, QM/MM geometry
optimizations were performed.
[10] E. Peyroux, W. Ghattas, R. Hardrꢁ, M. Giorgi, B. Faure, A. J.
Simaan, C. Belle, M. Rꢁglier, Inorg. Chem. 2009, 48, 10874–10876.
[11] L. Bubacco, R. Spinazze, S. della Longa, M. Benfatto, Arch. Bio-
chem. Biophys. 2007, 465, 320–327.
[12] W. T. Ismaya, H. J. Rozeboom, A. Weijn, J. J. Mes, F. Fusetti, H. J.
Wichers, B. W. Dijkstra, Biochemistry 2011, 50, 5477–5486.
[13] a) K. D. Karlin, Y. Gultneth, T. Nicholson, J. Zubieta, Inorg. Chem.
1985, 24, 3725–3727; b) H. Bçrzel, P. Comba, H. Pritzkow, Chem.
Commun. 2001, 97–98; c) J. Ackermann, F. Meyer, E. Kaifer, H.
Pritzkow, Chem. Eur. J. 2002, 8, 247–258.
[14] S. Torelli, C. Belle, I. Gautier Luneau, J. L. Pierre, E. Saint Aman,
J. M. Latour, L. Le Pape, D. Luneau, Inorg. Chem. 2000, 39, 3526–
3536.
[15] M. Orio, C. Bochot, C. Dubois, G. Gellon, R. Hardrꢁ, H. Jamet, D.
Luneau, C. Philouze, M. Rꢁglier, G. Serratrice, C. Belle, Chem. Eur.
J. 2011, 17, 13482–13494.
[16] A. W. Addison, T. N. Rao, J. Reedijk, J. van Rijn, G. Verschoor, J.
Chem. Soc. Dalton Trans. 1984, 1349–1356.
[17] During the redaction of this manuscript, synthesis of this ligand by a
different method has been published: F. Heims, V. Mereacre, A.
Ciancetta, S. Mebs, A. K. Powell, C. Greco, K. Ray, Eur. J. Inorg.
Chem. 2012, 4565–4569.
[18] C. Belle, G. Gellon, C. Scheer, J. L. Pierre, Tetrahedron Lett. 1994,
35, 7019–7022.
[19] a) H. Gampp, M. Maeder, C. J. Meyer, A. D. Zuberbühler, Talanta
1985, 32, 95–101; b) H. Gampp, M. Maeder, C. J. Meyer, A. D. Zu-
berbühler, Talanta 1985, 32, 257–264; c) H. Gampp, M. Maeder,
C. J. Meyer, A. D. Zuberbühler, Talanta 1986, 33, 943–951.
[20] B. J. Hathaway, Struct. Bonding (Berlin) 1984, 57, 55–118.
[21] Although UV/Vis data have not been obtained under the same ex-
perimental conditions (H2O/DMSO ratio, concentration) as those of
the EPR experiments, we cannot rule out instability of the 3-OPNO
adduct with a small release of copper in solution, as observed in
Chem. Eur. J. 2013, 19, 3655 – 3664
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3663

C. Belle, H. Jamet et al.
EPR experiments (and not observed for 1-OPNO and 2-OPNO).
Thus, data from UV/Vis experiments for 10 should be considered
with this possible event taken into account.
nov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. Gomperts,
R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A.
Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen,
M. W. Wong, J. L. Andres, C. Gonzalez, M. Head-Gordon, J. A.
Pople, Gaussian, Inc., Pittsburg, PA, 2003.
[22] B. F. Gherman, C. J. Cramer, Coord. Chem. Rev. 2009, 253, 723–753.
[23] a) L. Noodleman, J. Chem. Phys. 1981, 74, 5737–5743; b) L. Noodle-
man, D. A. Case, Adv. Inorg. Chem. 1992, 38, 423–470; c) L. Noo-
dleman, E. R. Davidson, Chem. Phys. 1986, 109, 131–143.
[24] T. Klabunde, C. Eicken, J. C. Sacchettini, B. Krebs, Nat. Struct. Biol.
1998, 5, 1084–1090.
[31] a) A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652; b) C. Lee, W.
Yang, R. C. Parr, Phys. Rev. B 1988, 37, 785–789; c) K. Yamaguchi,
F. Jensen, A. Dorigo, K. N. Houk, Chem. Phys. Lett. 1988, 149, 537–
542.
[25] SADABS v2.10, Program for empirical absorption correction of
area detector data, G. M. Sheldrick, University of Gçttingen, Ger-
many, 2003.
[26] A. J. M. Duisenberg, L. M. J. Kroon-Batenburg, M. M. Schreurs, J.
Appl. Crystallogr. 2003, 36, 220–229.
[27] G. M. Sheldrick, Acta Crystallogr. 2008, A64, 112–122.
[28] O. V. Dolomanov, L. J. Bourhis, R. J. Gildea, J. A. K. Howard, H.
Puschmann, J. Appl. Crystallogr. 2009, 42, 339–341.
[29] a) J. M. Pascal, Ann. Chim. Phys. 1910, 19, 5–69; b) O. Kahn, Mo-
lecular Magnetism, VCH, Weinheim, 1993.
[30] Gaussian 03, revision A.1., M. J. Frisch, G. W. Trucks, H. B. Schlegel,
G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski,
J. A. Montgomery, J. R. E. Stratmann, J. C. Burant, S. Dapprich,
J. M. Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J.
Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli,
C. Adamo, S. Clifford, J. Ochterski, G. Petersson, A. P. Y. Ayala, Q.
Cui, K. Morokuma, D. K. Malick, A. D. Rabuck, K. Raghavachari,
J. B. Foresman, J. Cioslowski, J. V. Ortiz, A. G. Baboul, B. B. Stefa-
[32] a) P. C. Hariharan, J. A. Pople, Theor. Chim. Acta 1973, 28, 213–
222; b) P. C. Hariharan, J. A. Pople, Mol. Phys. 1974, 27, 209–214.
[33] a) T. Soda, Y. Kitagawa, V. Onishi, Y. Tanako, Y. Shigeta, H. Nagao,
Y. Yoshioka, K. Yamaguchi, Chem. Phys. Lett. 2000, 319, 223–230.
[34] Tinker, 4.2 ed., J. W. Ponder, Washington University, Saint Louis,
MO, 2004.
[35] a) X. Assfeld, J. L. Rivail, Chem. Phys. Lett. 1996, 263, 100–106;
b) N. Ferrꢁ, J. L. Rivail, X. Assfeld, J. Comp. Chem. 2002, 23, 610–
624.
[36] J. Wang, P. Cieplak, P. A. Kollman, J. Comp. Chem. 2000, 21, 1049–
1074.
[37] P. Comba, R. Remenyi, J. Comp. Chem. 2002, 23, 697–705.
[38] SYBYL 8.0, Tripos Associates, Inc., Saint Louis, MO, 2007.
Received: July 25, 2012
Revised: November 15, 2012
Published online: January 29, 2013
3664
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 3655 – 3664