D -penicillamine tripodal derivatives as efficient copper(I) chelators

Article abstract of DOI:10.1021/ic5004319

New tripodal metal-chelating agents derived from nitrilotriacetic acid (NTA) and extended by three unnatural amino acids d-penicillamine (d-Pen) are presented. d-Pen is actually the drug most extensively used to treat copper (Cu) overload in Wilsons disease and as such is a very attractive building block for the design of chelating agents. d-Pen is also a bulkier analogue of cysteine, with the β-methylene hydrogen atoms replaced by larger methyl groups. The hindrance of the gem-dimethyl group close to the thiol functions is demonstrated to influence the speciation and stability of the metal complexes. The ligands L⁴ (ester) and L⁵ (amide) were obtained from NTA and commercial d-Pen synthons in four and five steps with overall yields of 14 and 24%, respectively. Their ability to bind Cu(I), thanks to their three thiolate functions, has been investigated using both spectroscopic and analytical methods. UV, CD, and NMR spectroscopy and mass spectrometry evidence the formation of two Cu(I) complexes with L⁵: the mononuclear complex CuL⁵ and one cluster (Cu₂L⁵)₂. In contrast, the bulkier ethyl ester derivative L⁴ cannot accommodate the mononuclear complex in solution and thus forms exclusively the cluster (Cu₂L⁴)₂. Cu K-edge X-ray absorption spectroscopy (XAS and EXAFS) confirms that Cu(I) is bound in trigonal-planar sulfur-only environments in all of these complexes with Cu- - -S distances ranging from 2.22 to 2.23 A. Such C₃-symmetric CuS₃ cores are coordination modes frequently adopted in Cu(I) proteins such as metallothioneins. These two ligands bind Cu(I) tightly and selectively, which makes them promising chelators for intracellular copper detoxification in vivo.

Full text of DOI:10.1021/ic5004319

Article
pubs.acs.org/IC
D‑Penicillamine Tripodal Derivatives as Efficient Copper(I) Chelators
Anne-Solene Jullien,†,‡ Christelle Gateau,†,‡ Colette Lebrun,†,‡ Isabelle Kieffer,§,∥ Denis Testemale,§,⊥,#
̀
and Pascale Delangle*,†,‡
†Universite
́
Grenoble Alpes, INAC, SCIB, RICC F-38000 Grenoble, France
‡CEA, INAC, SCIB, Laboratoire de Reconnaissance Ionique et Chimie de Coordination, F-38054 Grenoble, France
§BM30B/FAME beamline, ESRF, F-38043 Grenoble cedex 9, France
∥Observatoire des Sciences de l’Univers de Grenoble, UMS 832 CNRS Universite
́
Joseph Fourier, F-38041 Grenoble cedex 9, France
⊥Universite
́
Grenoble Alpes, Institut NEEL, F-38042 Grenoble, France
#CNRS, Institut NEEL, F-38042 Grenoble, France
S
* Supporting Information
ABSTRACT: New tripodal metal-chelating agents derived
from nitrilotriacetic acid (NTA) and extended by three
unnatural amino acids D-penicillamine (D-Pen) are presented.
D-Pen is actually the drug most extensively used to treat
copper (Cu) overload in Wilson’s disease and as such is a very
attractive building block for the design of chelating agents. D-
Pen is also a bulkier analogue of cysteine, with the β-methylene
hydrogen atoms replaced by larger methyl groups. The
hindrance of the gem-dimethyl group close to the thiol
functions is demonstrated to influence the speciation and
stability of the metal complexes. The ligands L4 (ester) and L5
(amide) were obtained from NTA and commercial D-Pen
synthons in four and five steps with overall yields of 14 and
24%, respectively. Their ability to bind Cu(I), thanks to their three thiolate functions, has been investigated using both
spectroscopic and analytical methods. UV, CD, and NMR spectroscopy and mass spectrometry evidence the formation of two
Cu(I) complexes with L5: the mononuclear complex CuL5 and one cluster (Cu2L5)2. In contrast, the bulkier ethyl ester derivative
L4 cannot accommodate the mononuclear complex in solution and thus forms exclusively the cluster (Cu2L4)2. Cu K-edge X-ray
absorption spectroscopy (XAS and EXAFS) confirms that Cu(I) is bound in trigonal-planar sulfur-only environments in all of
these complexes with Cu- - -S distances ranging from 2.22 to 2.23 Å. Such C3-symmetric CuS3 cores are coordination modes
frequently adopted in Cu(I) proteins such as metallothioneins. These two ligands bind Cu(I) tightly and selectively, which makes
them promising chelators for intracellular copper detoxification in vivo.
accumulates in the liver cells, where it is involved in Fenton-
like reactions, which produce toxic hydroxyl radicals. That is
why chelators, such as the British anti-Lewisite (BAL),
dimercaptosuccinic acid (DMSA), dimercaptopropanesulfonic
acid (DMPS), triethylenetetramine (TETA), and D-penicill-
amine (D-Pen) have been used to treat Wilson’s disease.10,11
BAL was introduced in 1951 to treat Wilson’s disease patients
from neurological symptoms.12 However, since BAL and its
hydrophilic derivatives DMSA and DMPS have shown many
unwanted side effects, other chelators were searched for. In
particular, D-Pen was introduced in 195613 and is currently the
most widely used treatment of Wilson’s disease around the
world. However, some side effects related to this therapy have
been reported.14 More precisely, it has been shown that D-Pen
can contribute to increase the neurologic deterioration
INTRODUCTION
■
Some metals such as copper, iron, and manganese play crucial
roles in the functioning and structural integrity of proteins.1 In
metalloproteins, these metal ions carry out structural,
regulatory, or catalytic roles. In particular, the copper ions
play key roles in all living organisms and are used as cofactors of
several enzymes involved in electron transfers, oxidase and
oxygenase activities, and detoxification of oxygen radicals.
However, high concentrations of copper can be deleterious,
leading to oxidative damage of proteins, lipids, and nucleic
acids. That is why copper concentration is rigorously controlled
by subtle cellular mechanisms, the main organ of copper
homeostasis being the liver.2−5 In some cases, these regulation
mechanisms can be faulty and copper accumulates at a toxic
level in cells. Wilson’s disease is one of the major genetic
disorders of copper metabolism in humans.6−9 In this rare
disease the ATP-ase ATP7B in charge of the excretion of excess
Cu from the liver cells is defective. Consequently, Cu
Received: February 24, 2014
Published: April 25, 2014
© 2014 American Chemical Society
5229
dx.doi.org/10.1021/ic5004319 | Inorg. Chem. 2014, 53, 5229−5239

Inorganic Chemistry
Article
expressed by some Wilson’s disease patients, which may be due
to a lack of specificity of the drug.
essential ion Zn(II), which is important for the detoxification of
Cu(I) in vivo without altering the homeostasis of zinc.
Since these ligands are designed to chelate intracellular
copper, in particular in the liver, they have to be internalized
into hepatocytes. Therefore, two of these bioinspired efficient
Cu(I) chelating agents have been functionalized with sugar
units to target liver cells via the asialoglycoprotein receptors
(ASGP-R).16,17,36 Both glycoconjugates were demonstrated to
release high-affinity Cu chelating agents in the hepatic cells.
This confirms that the use of intracellular Cu(I) chelators is
very promising to treat Cu overload.
Our laboratory has been involved in the design of new drugs
to treat Wilson’s disease for several years.15,16 These new drugs
are copper chelators with the following properties: (i) they
show a high affinity for Cu(I), which is the oxidation state of
excess intracellular Cu, (ii) they are selective for Cu(I) with
respect to potentially competing endogeneous metals such as
Zn(II), and (iii) they are water-soluble. Mimics of binding sites
found in proteins involved in Cu homeostasis are good
candidates that meet all these expectations. A first strategy used
peptide sequences of the copper binding loop of metal-
lochaperones such as Atox1, which transfer the Cu(I) ion to
ATP7B in the liver cells. These metal-binding loops contain
two cysteine residues and were introduced in cyclodecapep-
The sulfur tripods L1 and L2 derived from NTA demonstrate
the most promising Cu(I) chelating properties with large
affinities and selectivities for Cu(I). Therefore, we are currently
exploring the metal chelating properties of NTA pseudopep-
tides functionalized with various metal-coordinating amino
acids. In particular, D-Pen is currently the drug most extensively
used to treat Cu overload in Wilson’s disease and, as such,
represents a very attractive building block for the design of
chelating agents. D-Pen is also a bulkier analogue of cysteine
with the β-methylene hydrogen atoms replaced with larger
methyl groups. Incidentally, its insertion in peptide sequences
in place of cysteine has been demonstrated to provide more
steric bulk in the coordination sphere of Cd(II)37−39 or
Co(II)40 and to favor a MS3 geometry.
tides, which complex digonally the Cu(I) ion with a
17−19
dissociation constant on the order of 10−17
.
Metal-
lothioneins are also attractive Cu(I) chelating proteins, since
they are natural metal sequestering agents synthesized by cells
in case of Cu overload and involved in Cu intracellular
detoxification.20−23 Interestingly, in metallothioneins Cu(I) is
tightly bound in CuS3 coordination environments provided by
three thiolates of cysteine residues.24 The stability of the Cu(I)
complexes with metallothioneins is exceptionally large, with
25
dissociation constants on the order of 10−19
.
This led us to
design pseudopeptide26 ligands with three cysteines attached
with peptide bonds to C3-symmetric nonbiological scaffolds,
which act as platforms to direct the three thiolate binding arms
in the same direction to coordinate the Cu(I) ion.27,28
Nitrilotriacetic acid (NTA) was chosen as a tripodal anchor
and led to the three cysteine derivatives L1−L3 depicted in
Figure 1. It has been established that the carboxylic acid
In this paper, we present the new tripodal architectures L4
and L5 (Figure 1) derived from D-Pen. A first objective was to
explore the effect of replacing the cysteine units with the more
hindered D-Pen units and therefore to evidence the influence of
the gem-dimethyl group onto the metal complexing properties
in the tripodal architectures. In addition, the use of non-natural
amino acids is known to disfavor hydrolysis in vivo. The ester
(L4) and amide (L5) derivatives have been synthesized, and
their ability to bind Cu(I) thanks to their thiolate functions has
been proven using both spectroscopic and analytical methods
such as UV spectroscopy (UV), circular dichroism (CD), mass
spectrometry (MS), nuclear magnetic resonance (NMR), and
X-ray absorption spectroscopy (XAS). The gem-dimethyl group
close to the thiol function influences the reactivity of the D-Pen
residue and also the speciation and stability of the metal
complexes.
EXPERIMENTAL SECTION
Syntheses. Detailed experimental procedures for the syntheses of
all intermediates and ligands L4 and L5 are given in the Supporting
Information, together with their characterization data.
■
Figure 1. Structures of the Cu(I) cysteine-based chelators H3L1−H3L3
from previous work27,28,41,42 and the Cu(I) D-penicillamine-based
chelators H3L4 and H3L5 (this work).
Samples for Physicochemical Experiments. The chelators L4
and L5 and Cu(I) are susceptible to air oxidation; therefore, the
samples for physicochemical studies were all prepared in the glovebox
under an argon atmosphere and sealed if used outside the glovebox.
The samples were dispersed in mixtures containing different
proportions of extrapure water buffered at pH 7.4 with the appropriate
buffer and acetonitrile or buffered D2O and CD3CN for NMR
experiments. At least 10% of acetonitrile was added to each sample to
prevent Cu(I) dismutation.43 Glycerol was added for X-ray absorption
spectroscopy (XAS) experiments to prevent the degradation of the
samples at cryogenic temperatures. In each case, the final
concentration of the ligand solution was determined by measuring
the D-penicillamine (D-Pen) derivative free thiol concentration
following Ellman’s procedure.44 The Cu(I) solutions were prepared
by dissolving the appropriate amount of Cu(CH3CN)4PF6 in
acetonitrile or CD3CN for NMR experiments. The final concentration
was determined by adding an excess of sodium bathocuproine
derivative L3 forms mixtures of Cu(I) complexes of low affinity,
whereas both L1 (ester) and L2 (amide) bind tightly the Cu(I)
ion in C3-symmetric species, either mononuclear (CuL) or
polymetallic ((Cu2L)3). In both types of complexes, Cu(I) ions
are coordinated in CuS3 environments29 as in many Cu
proteins,30−33 in particular metallothioneins.24,34,35 The struc-
tural analogies of Cu complexes with L1 or L2 and
metallothioneins allow us to rationalize the high affinities of
the pseudopeptide ligands for Cu(I), with dissociation
constants of 10−19.2 (L1) and 10−18.8 (L2).28 Moreover, these
Cu(I) chelating agents inspired from either metallochaperones
or metallothioneins are selective for Cu(I) with respect to the
33
3−
disulfonate (BCS) and measuring the absorbance of Cu(BCS)2
.
The Zn(II) solutions used for titrations were prepared by dispersing
5230
dx.doi.org/10.1021/ic5004319 | Inorg. Chem. 2014, 53, 5229−5239

Inorganic Chemistry
Article
ZnCl2 salts in Millipore water at 3 mM. Zinc concentrations were
estimated by titrations using EDTA (5 mM) at pH 4.5 and xylenol as a
colorimetric indicator. Extensive information is provided in the
Supporting Information about the titrations and methods used to
characterize the metal complexes.
method was selected. The esterification reaction was performed
under basic conditions to avoid trityl deprotection, using
cesium carbonate as a base and ethyl bromide as an
electrophilic agent.46 To prevent side reactions on the free
amine, the protected Fmoc-D-Pen(Trt)-OH derivative was used
as a starting material. Thus, the ester derivative Fmoc-D-
Pen(Trt)-OEt (1) was obtained with 92% yield (step a, Scheme
1). Then, the Fmoc group was cleaved using triethylamine
(step b, Scheme 1) to give the compound D-Pen(Trt)-OEt (2)
with 60% yield.47 The synthon D-Pen(Trt)-OEt (2) was then
coupled to the NTA template (step c, Scheme 1), as previously
described for the cysteine-based tripodal scaffolds.27,28 This
reaction leads to the trityl-protected tripod NTA[D-Pen(Trt)-
OEt]3 (3) in 72% yield. Eventually, the trityl groups are cleaved
in acidic medium (step d, Scheme 1) to provide L4 (36% yield,
four steps, 14% global yield).
RESULTS
■
Syntheses of the D-Pen-Based Tripodal Scaffolds L4
and L5. The tripodal scaffolds L4 and L5 were obtained by
coupling 3 equiv of D-Pen ester or amide units on the trivalent
NTA template, as previously described for L1 and L2 (Figure
2). The procedures used for the syntheses of L4 and L5 and the
corresponding characterizations are reported in the Supporting
Information.
L5 (Amide). The strategy used to synthesize L4 was applied
to the amide-functionalized D-Pen-based scaffold L5 as depicted
in Scheme 2. The D-Pen amide synthon was obtained by a two-
step aminoacylation sequence. The first step (step a, Scheme 2)
consisted of an NHS activation48,49 of Fmoc-D-Pen(Trt)-OH to
afford the compound Fmoc-D-Pen(Trt)-ONHS (4a) in 92%
yield. The activated acid (4a) was then involved in an
aminoacylation reaction with ammonium hydroxide50 (step b,
Scheme 2) to provide the compound Fmoc-D-Pen(Trt)-NH2
(4b; 98%), which was Fmoc-deprotected to afford the
compound D-Pen(Trt)-NH2 (5; step c, Scheme 2, 78%). The
compound 5 was then involved in a coupling reaction with the
NTA template to provide the trityl-protected tripodal NTA[D-
Pen(Trt)-OEt]3 6 (step d, Scheme 2, 98%). Eventually, the
trityl groups were cleaved under acidic conditions to provide L5
(step e, Scheme 2, 35%, five steps, 24% global yield).
As expected, these two novel ligands derived from D-Pen are
more lipophilic than their cysteine analogues L1 and L2. In
particular, the ester compound L4 is not soluble in water at
physiological pH and addition of acetonitrile was revealed to be
necessary for all the studies. On the other hand, compound L5,
which is the relevant chelator for in vivo studies, since the ester
function of L4 may hydrolyze in the intracellular medium, is
Figure 2. Synthetic strategy.
L4 (Ester). The strategy previously used27,28 to synthesize the
ester derivative of cysteine with p-toluenesulfonate in ethanol
was first attempted to obtain the D-Pen ester synthon.45 This
esterification reaction does not occur with the substrate H-D-
Pen(Trt)-OH, whatever the temperature and the reaction time.
Trityl cleavage is systematically observed. The steric hindrance
due to the gem-dimethyl group is probably responsible for the
enhanced lability of the trityl groups and the lack of reactivity of
the carboxylic acid function. Therefore, another esterification
a
Scheme 1. Synthesis of the Tripodal D-Penicillamine-Based Ligand L4 (Ester)
a
Experimental conditions: (a) Cs2CO3, EtBr, DMF, room temperature, 3 h, 92%; (b) NEt3/DMF (5/5, v/v), room temperature, 4 h, 60%; (c) NTA,
EDC, HOBt, room temperature, DMF, 4 h, 72%; (d) TFA/TES, DCM, room temperature, 2 h, 36% (four steps, 14% global yield).
5231
dx.doi.org/10.1021/ic5004319 | Inorg. Chem. 2014, 53, 5229−5239

Inorganic Chemistry
Article
a
Scheme 2. Synthesis of the Tripodal D-Penicillamine-Based Ligand L5 (Amide)
a
Experimental conditions: (a) NHS, room temperature, then EDC, 0 °C, ACN, room temperature, 16 h, 92%; (b) NH3 30%, ACN, room
temperature, 1 h, 98%; (c) NEt3, ACN, room temperature, 7.5 h, 78%; (d) NTA, EDC, HOBt, room temperature, ACN, 16 h, 98%; (e) TFA/TES,
DCM, room temperature, 2 h, 35% (five steps, 24% global yield).
fully soluble in water at pH 7.4. In the following studies with L5,
acetonitrile was added only to preserve the oxidation state of
Cu(I).
which increases with the Cu(I) concentration. This band is
characteristic of charge transfer transitions (LMCT) from
thiolates to Cu(I).51 An extra band appears at lower energy
(315 nm) and accounts for formally spin forbidden 3d to 4s
metal cluster centered transitions brought about by Cu- - -Cu
interactions in clusters.52 As shown in the inset of Figure 3, two
changes of level appear on the curves at 262 and 315 nm for 1
and 2 equiv of Cu(I). From these data, it can be deduced that
both the mononuclear complex CuL5 and polymetallic
(Cu2L5)z species form in solution, with the cluster molecularity
z undetermined at this step of the study.
Cu(I) Complexes with the Ligand L5. Physicochemical
studies conducted on Cu(I) complexes with L5 all point to the
formation of two types of complexes: the first has a 1:1
stoichiometry and corresponds to the mononuclear species
CuL5, and the second displays an overall 2:1 stoichiometry
(Cu2L5)z.
These two complexes were first evidenced by mass
spectrometry. Spectra acquired in the negative mode display
two major ions: m/z 642 for [L5 + H + Cu]−, which is
characteristic of a mononuclear Cu complex, and m/z 704 for
[L5 + 2Cu]−, which most likely originates from a (Cu2L)z type
cluster. The corresponding spectra are reported in Figures S1
and S2 in the Supporting Information.
CD titrations, depicted in Figure 4, correlate the data
obtained by UV spectroscopy: from 0 to 1 equiv of Cu(I) (blue
The binding of Cu(I) has also been investigated by UV
spectroscopy. The spectra recorded for the titration of L5 with
Cu(CH3CN)4PF6 are depicted in Figure 3. They display the
appearance of a band centered at 262 nm, the intensity of
Figure 4. CD titration of L5 (46 μM) at pH 7.4 (20 mM phosphate
buffer/MeCN, 9/1 v/v) with Cu(I) (0−2.5 equiv): (red) free L5;
(blue) 0.25−1 equiv of Cu(I); (black) 1.5−2.5 equiv of Cu(I).
curves in Figure 4), two positive bands (250 nm, 270 nm) and
one negative band (225 nm) regularly increase with a well-
defined isodichroic point at 240 nm. Above 1 equiv of Cu(I),
the 250 nm centered band completely collapses, whereas the
270 nm centered band continues to increase until 2 equiv of
Cu(I) was added. Above 2 equiv of Cu(I), all of the spectra
Figure 3. UV titration of L5 (61 μM) at pH 7.4 (20 mM phosphate
buffer/MeCN, 9/1 v/v) with Cu(I) (0−4 equiv). The inset gives the
difference spectra Δε = ε(CuL5) − ε(L5) at 262 nm (LMCT) and at
315 nm (cluster transitions), respectively.
5232
dx.doi.org/10.1021/ic5004319 | Inorg. Chem. 2014, 53, 5229−5239

Inorganic Chemistry
Article
recorded are superimposed. The first band centered at 250 nm
indicates the formation of the mononuclear complex, whereas
the second band centered at 270 nm prevails when the
(Cu2L5)z cluster become the major species in solution.
bands seen previously for L5 but with only one change of
level for 2 equiv of Cu(I) (Figure 6), indicating the formation
of only one complex, namely (Cu2L4)z.
NMR spectra are very informative about the two species
formed. The titrations of L5 with Cu(I) followed by NMR have
been performed in a 20 mM phosphate buffer in D2O/MeCN
(9/1, v/v) mixture with 1 mM ligand concentration and are
shown in Figure 5. The spectrum of the ligand indicates a C3-
Figure 6. UV titration of L4 (45 μM) at pH 7.4 (20 mM phosphate
buffer/MeCN, 9/1, v/v) with Cu(I) (0−4 equiv). The inset gives the
difference spectra Δε = ε(CuL4) − ε(L4) at 267 nm (LMCT) and at
310 nm (cluster transitions), respectively.
CD titrations, depicted in Figure S6 in the Supporting
Information, also show the growth of only one positive band
centered at 270 nm, which regularly increases during the
titration until 2 equiv of Cu(I) is added. The mass spectra do
not show the 1:1 species but a signal at m/z 791 for [L4 +
2Cu]−, which originates from a (Cu2L)z cluster. The
corresponding spectrum is reported in Figure S3 in the
Supporting Information. The formation of only one copper
complex is finally demonstrated by the NMR data shown in
Figure S7 in the Supporting Information. A less polar solvent,
phosphate buffer in D2O/MeCN (5/5, v/v), has been used for
L4, not soluble at 1 mM in 20 mM phosphate buffer in a D2O/
MeCN (9/1, v/v) mixture. The spectrum of the free ligand is
characteristic of a C3-symmetric molecule with the 9 methyl
protons of the ester group at 1.2 ppm, the 18 protons of the
two D-Pen methyl groups at 1.4 and 1.5 ppm, the 6 methylene
protons of the NTA scaffold at 3.4 ppm, the 6 methylene
protons of the ester group partially hidden by the residual water
solvent signal, and the 3 α-CH protons of the D-Pen units at 4.5
ppm. During the titration with Cu(I), only two sets of signals
coexist in the spectra: the signals (L) for the free ligand protons
and the signals (C) for the protons of the (Cu2L4)z cluster
species (Figure S7, Supporting Information). The signals C
increase with Cu(I) addition at the expense of signals L. At the
end of the titration, for 2 equiv of Cu(I) added, only signals C
can be seen in the spectrum. As for (Cu2L5)z, a well-defined AB
system accounts for the methylene protons of the NTA scaffold
in a rigid complex.
1
Figure 5. 400 MHz H NMR at 298 K of L5 (1 mM) at pD 7.4 (20
mM phosphate buffer in D2O/CD3CN, 9/1 v/v) with 1 equiv of
Cu(I) (67% A, 22% B, 11% L) or 2 equiv of Cu(I) (100% B).
symmetric species with three distinct proton signals: the 18
protons of the two D-Pen methyl groups at 1.45 and 1.38 ppm,
the 6 methylene protons of the NTA scaffold at 3.5 ppm, and
the 3 α-CH protons of the D-Pen units at 4.4 ppm. When
aliquots of Cu(I) in CD3CN are added to the sample, the
signals of three different C3-symmetric species coexist and are
in slow exchange on the NMR time scale. Those signals
correspond to the free ligand (L), the mononuclear complex
CuL5 (A), and the polymetallic species (Cu2L5)z (B). The
signals of A appear first, for 0.25 equiv of Cu(I) added. The
signals of B are detected above 0.5 equiv of Cu(I) and regularly
increase at the expense of signals L and A until 2 equiv of Cu(I)
is added. As depicted in Figure 5, for 2 equiv of Cu(I) added,
only the signals B are visible in the spectrum, evidencing the
existence of the only remaining (Cu2L5)z species in solution.
One noticeable feature is the splitting of the signal of the
methylene protons of the NTA scaffold into an AB system
which is very well-defined in (Cu2L5)z, indicating the rigidity of
the polymetallic Cu(I) complex.
These experiments demonstrate that the ester derivative L4,
which is bulkier than the corresponding amide ligand L5,
cannot accommodate the mononuclear complex in solution and
thus forms exclusively a (Cu2L4)z species.
Molecularities of the Cu(I) Complexes. NMR is a
powerful tool to infer the molecularities of species in solution.
Indeed, diffusion coefficients measured by NMR spectroscopy
using pulsed field-gradient spin echo (PFGSE) sequences are
characteristic of the presence of unimolecular, bimolecular, or
oligomeric species in solution.53−55 The diffusion constant D
can be related to the hydrodynamic radii of the molecules via
the Stokes−Einstein equation (1), where k is the Boltzmann
All together UV, CD, and NMR spectroscopy and mass
spectrometry evidence the formation of two Cu(I) complexes:
CuL5 and a (Cu2L5)z cluster.
Cu(I) Complexes with the Ligand L4. In contrast, the
same experiments run with the ester derivative L4 point to the
formation of only one complex, with an overall 2:1
stoichiometry. Indeed, the UV titration displays the two
5233
dx.doi.org/10.1021/ic5004319 | Inorg. Chem. 2014, 53, 5229−5239

Inorganic Chemistry
Article
Table 2. Samples Analyzed by XAS in a Phosphate Buffer
kT
D =
(100 mM, pH 7.4)/Glycerol/Acetonitrile (7/2/1, v/v/v)
6πηrH
(1)
a
Mixture
L
L4
sample
[Cu]/[L]
[Cu] (mM)
[L] (mM)
constant, T the absolute temperature, η the viscosity, and rH the
hydrodynamic radius of the diffusing species, considered as a
hypothetical hard sphere that diffuses with the same speed as
the particle under examination. Thus, the determination of D
by diffusional NMR for shape-similar complexes and ligands is
an efficient tool for deducing the molecular mass of an
unknown species (i, molar mass Mi) in solution, when a
reference compound (r, molar mass Mr) is measured under the
same conditions. Indeed, eq 2 allows the determination of the
unknown molar mass Mi of interest.
4A
4B
4C
5A
5B
5C
5D
5E
1
0.566
0.838
1.104
0.5
0.569
0.562
0.554
2.91
2.71
2.6
1.5
2
L5
0.17
0.92
1.64
1.8
1.99
2.5
3.5
4.5
2.5
4.83
2.42
a
Note that lower [L4] (mM) concentrations were used due to the lack
of solubility of L4 in the solvent because of its relatively apolar
character in comparison to L5.
Di
Mr
Mi
=
3
D
from 0.17 to 2 equiv. The lower solubility of L4 prevented the
preparation of samples more concentrated than 0.56 mM.
Furthermore, the Cu(I) content could not be decreased below
1 equiv for XAS sensitivity reasons.
(2)
r
Experimental masses (Mexptl) derived from eq 2, taking the
free ligand in the same solvent as a reference compound and
theoretical masses calculated from the compound’s formulas,
are reported in Table 1. These data confirm that the signals of
The spectra of the Cu(I) complexes with L4 and L5 are
compared in Figure 7. The peaks are centered at 8982 eV,
1
Table 1. Diffusion Coefficients Measured by PFGSE H
NMR at 298 K for Metallic Species Formed with L4 and L5 at
a
pD 7.4
b
c
D (1010 m2 s−1 mol−1
)
Mexptl (g/mol)
Mtheor (g/mol)
L4
3.90(9)
2.95(7)
3.70(1)
3.02(9)
2.86(6)
2.36(9)
669
1586
731
(Cu2L4)2
ZnL4
L5
1550(200)
780(60)
582
CuL5
(Cu2L5)2
685(90)
642
1220(220)
1412
a
Diffusion coefficient measured in 20 mM phosphate buffer in D2O/
CD3CN (9/1 v/v for L5 and 5/5 v/v for L4). Experimental errors in
the last digit are indicated in parentheses. Molecular mass calculated
b
with eq 2, the reference compound being the free ligand in the same
c
solvent. Molecular mass calculated with the chemical formula of the
species.
A detected in the 1H NMR spectra with the amide derivative L5
correspond to the Cu(I) mononuclear complex CuL5. In
addition, this set of measurements allows the determination of
the two cluster molecularities, detected in the 1H NMR spectra
as C for L4 and B for L5. Both clusters show diffusion
coefficients characteristic of a molecularity z = 2 and are thus
(Cu2L)2 species i.e., Cu4S6-type clusters. It is quite interesting
to notice that the analogous cysteine-based scaffold L1 forms a
cluster with a higher nuclearity (z = 3) and a Cu6S9 core.27,28
These data suggest that analogous ligands tend to form similar
metallic species in solution with different molecularities
depending on the steric hindrance of the pseudopeptide
scaffolds.
Copper K-Edge X-ray Absorption Spectroscopy. Cu K-
edge X-ray Absorption spectroscopy (XAS) spectra were
collected at the BM30B FAME beamline at the European
Synchrotron Radiation Facility (ESRF, Grenoble, France) at
low temperature (10 K).56,57 They confirmed the speciation
deduced from other spectroscopic data and provided significant
complementary information about the Cu(I) species formed in
solution. The samples analyzed by XAS are given in Table 2.
The L5 concentration was around 3 mM, with Cu(I) varying
Figure 7. Normalized Cu edge spectra of Cu(I) complexes with (A)
L4 and (B) L5.
which accounts for the existence of Cu(I) species. In Figure 7B,
it clearly appears that two species are coexisting during the
titration of L5 with Cu(I), as evidenced by the existence of
several isosbestic points (8981, 8984, 8988, and 8994 eV).
In addition, the peak centered at 8982 eV linearly decreases
with Cu(I) addition, accounting for the formation of the
polymetallic species at the expense of the mononuclear
5234
dx.doi.org/10.1021/ic5004319 | Inorg. Chem. 2014, 53, 5229−5239

Inorganic Chemistry
Article
Figure 8. XAS data for (A) the mononuclear complex CuL5 (sample 5A) and (B) the cluster (Cu2L5)2(sample 5F): (left) spectra of the k3-weighted
EXAFS experimental data and corresponding fit; (right) Fourier transforms of the k3-weighted EXAFS experimental data and corresponding fit. FT
and IM FT are the magnitude and imaginary part of Fourier transforms, respectively. Solid lines indicate experimental data, and dotted lines indicate
fits.
complex CuL5. This peak corresponds to 1s−4p Cu(I)-
centered transitions, the intensity of which evidence the
coordination geometry adopted around the central Cu(I)
ions.58 On the other hand, the three spectra recorded for L4
complexes (Figure 7A) are superimposable, whatever the Cu to
L4 ratio. Moreover, they are very similar to the spectrum of
(Cu2L5)2, which is consistent with the exclusive formation of
the polymetallic complex (Cu2L4)2.
Comparisons with previous XAS data acquired with cysteine
derivatives are consistent with tris(thiolato) environments in
either mononuclear species or cluster complexes.42,59 This was
confirmed by the extended X-ray absorption fine structure
(EXAFS) analysis.
Two structural models were applied to fit the EXAFS data
using the Horae package,60 including ATHENA for the data
extraction and ARTEMIS for the shell fitting. The first model
(coordinate-based model) was derived from the atomic
coordinates of sulfur and copper atoms in the
[Cu4(CH3S−)6]2−[(C3H7)4N+]2 cluster described by Baumgart-
ner et al.61 On the basis of this structure, the atomic
coordinates of Cu(1), S(1), S(2), and S(3) were used in
Atoms (Artemis module) for the mononuclear complex, as a
starting point for the structural fitting. For the clusters, data
were fitted using a model based on the whole set of Cu and S
coordinates from the same Baumgartner et al. structure, since in
this case one copper atom is surrounded by three S and three
Cu atoms, as expected for the L-based Cu(I) clusters. The
second structural model (tetrahedral-based model) simply uses
Cu surrounded by atom shells in a tetrahedral geometry (see
the Supporting Information for further information). In the fits
presented in the following the coordination number was set to
three sulfurs since using either two sulfur atoms or three
nonequivalent sulfur atoms in the first theoretical shell returns
bad results and nonphysical parameter values.
EXAFS Analyses. To characterize the Cu(I) species formed
in solution and refine the structure of the coordination spheres
around the Cu(I) ion in each case, the EXAFS domains of the
XAS spectra were then looked into. The EXAFS data of the
mononuclear complex CuL5 were analyzed in sample 5A with
low Cu(I) concentration. The cluster (Cu2L5)2 was investigated
in sample 5F, which contains 2 equiv of Cu(I). Samples 5A and
1
5F were shown by H NMR to contain only the mononuclear
adduct and polymetallic adducts, respectively. The cluster
(Cu2L4)2 could be analyzed in each of the three samples (4A−
C), which show identical XAS and EXAFS spectra, since this is
the only species formed with L4 under our conditions. The
results are presented here with sample 4A.
Mononuclear Complex CuL5. The EXAFS fitting results
obtained for the mononuclear complex CuL5 are depicted in
Figure 8A. Both the coordinate-based and the tetrahedral-based
models return good fits of the data collected. The quantitative
results are reported in Table 3. The three sulfur−copper
5235
dx.doi.org/10.1021/ic5004319 | Inorg. Chem. 2014, 53, 5229−5239

Inorganic Chemistry
Article
a
Table 3. EXAFS Fitting Results
complex
sample
model
3 × Cu−S
σ(S)2
3 × Cu−Cu
σ(Cu)2
ΔE
χn
R
2
CuL5
5A
5A
5F
5F
4A
4A
coordinates
tetrahedral
coordinates
tetrahedral
coordinates
tetrahedral
2.23
2.23
2.22
2.22
2.22
2.22
5(1)
4(1)
8(1)
8(1)
7(1)
7(1)
3(2)
4(2)
4(2)
4(2)
4(3)
4(2)
24
24
3.0
3.0
6.0
6.7
6.7
6.1
CuL5
(Cu2L5)2
(Cu2L5)2
(Cu2L4)2
(Cu2L4)2
2.66, 2.74, 2.83
2.75
10(3)
15(4)
8(2)
91
111
31
2.62, 2.68, 2.77
2.69
13(2)
26
a
Distances are given in Å with errors of 0.01 and 0.02 for Cu−S and Cu−Cu distances, respectively. Other experimental errors in the last digit are
2
indicated in parentheses. Debye−Waller factors (σ2) are given in Å2 × 103. The threshold energy shift ΔE is given in eV. χn and R (%) are the
reduced χ2 value and the R factor of the fit, respectively. The “coordinates” model uses the atomic coordinates of the cluster
[Cu4(CH3S−)6]2−[(C3H7)4N+]2 described by Baumgartner et al.61 Only the atomic coordinates of Cu(1), S(1), S(2), and S(3) were used to fit
the mononuclear complex, whereas the whole set of Cu and S atomic coordinates was used to fit the clusters. The “tetrahedral” model simply uses
Cu surrounded by atoms in a tetrahedral geometry.
distances were found to be almost equal: Cu−S ≈ 2.23 Å. This
Cu(I) + L = CuL β11
(3)
distance belongs to the characteristic distance range reported
for symmetric CuS3 compounds (∼2.22−2.26 Å) studied by
either XAS or crystallography.30−35,42,62 Consequently, these
4Cu(I) + 2L = (Cu2L)2
β
42
(4)
data confirm that L5 forms a mononuclear complex with C3-
symmetric CuS3 coordination. Moreover, the C3-symmetric
species detected in the NMR spectra recorded at 298 K are not
average signals of dissymmetric species, as observed previously
for mercury complexes of L1 and L2.41 It is quite interesting to
notice that the cysteine-based ligands L1 and L2 and the D-Pen-
based ligand L5, all derived from the same NTA template, form
trigonal-planar mononuclear complexes with identical Cu−S
distances (Cu−S ≈ 2.23 Å),42 despite the steric hindrance of
the gem-dimethyl groups of the three D-Pen units in L5.
Clusters. The EXAFS fitting results obtained for the clusters
(Cu2L5)2 are depicted in Figure 8B. A beat between 6 and 8
Å−1 on the wave vector k-space curve effectively accounts for
the presence of Cu- - -Cu interactions in the (Cu2L5)2 species
formed in solution. Such a beat originates from waves out of
phase from different shells of atoms.61 The corresponding
Cu- - -Cu peak at ≈2.7 Å can be observed in the distance R-
space in Figure 8B. Analogous graphs were obtained for the
cluster (Cu2L4)2 (Figure S8 in the Supporting Information).
Again both the coordinate-based model and the tetrahedral-
based model return good fits of the data collected. The
quantitative results are reported in Table 3. The distances Cu−
S ≈ 2.22 Å and Cu- - -Cu ≈ 2.7 Å were found to be in
accordance with the values given in the literature for several
Cu(I) proteins,30−33 in particular metallothioneins
(MTs),34,35,62 and some inorganic model clusters61 organized
on CuS3 cores which interact with one to three nearby copper
atoms in either Cu4S6 or Cu6S9 clusters. The Cu- - -Cu
distances obtained with the coordinate-based model are not
equal (∼2.66−2.83 Å; see Table 3), but the resolution expected
for these second-shell atoms is too low (∼0.1 Å)63 to allow a
discrimination between such close distances. Therefore, the
tetrahedral-based model, which sets the three Cu atoms at the
same distance (2.7 Å) from the central Cu is preferred. The
most striking features are probably the short Cu−S distances
measured in the two clusters (Cu−S ≈ 2.22 Å). Indeed, these
latter distances are significantly shorter than that previously
measured in the (Cu2L1/L2)3 polymetallic species formed with
the cysteine derivatives, with Cu−S distances of 2.26 Å.
Affinity for Cu(I). Considering the species identified in the
previous sections, the thermodynamic equilibria for the
formation of Cu(I) complexes with L5 are described by eqs 3
and 4, whereas only eq 4 is necessary for L4.
The conditional stability constants at pH 7.4 were
determined by UV spectroscopy in the presence of bath-
ocuproine disulfonate (BCS), which forms the orange complex
Cu(BCS)2 of known stability.33 The values are reported in
3−
Table 4. It appears that the formation constant value for CuL5
Table 4. Conditional Stability Constants of Cu(I) and Zn(II)
a
Complexes Formed with L4 and L5 at pH 7.4
pH 7.4
pH 7.4
Zn
complex
log β11
log β42
71.5(1)
log(β11Cu/β11
)
b
(Cu2L4)2
CuL5
17.3
16.2(1)
(Cu2L5)
ZnL4
ZnL5
68(1)
2
6.5(1)
6.3(1)
10.8
9.9
a
Conditional stability constants with Cu(I) were measured in 20 mM
phosphate buffer pH 7.4/MeCN (9/1 v/v) and conditional stability
constants with Zn(II) in 20 mM Hepes buffer pH 7.4/MeCN (9/1, v/
v). Experimental errors in the last digit are indicated in parentheses.
b
Constant calculated for the formation of an assumed CuL4 complex.
(log β11pH 7.4 = 16.2) is significantly lower than those found for
the cysteine derivatives L1 and L2 (19.2 and 18.8,
respectively).27,28,42 This difference can be assigned to the
steric hindrance due to the gem-dimethyl groups near the
coordinating thiolates in the D-Pen derivative. However, L5 can
compete with several Cu(I) proteins in cells, since the affinities
reported for Cu(I) proteins are in the range log β11pH 7.4 = 15−
20.25,64,65 The two clusters (Cu2L4)2 and (Cu2L5)2 display large
stabilities of the same order of magnitude. Therefore, these
measurements demonstrate that both L4 and L5 bind Cu(I)
with affinities which make them potential intracellular Cu(I)
chelators for detoxification. However, for the applications in
vivo, information about the selectivities is required to ensure
that Cu can be bound in vivo without altering the homeostasis
of the essential metal ion Zn(II). This is why the complexation
properties with Zn(II) have been investigated.
Zn(II) Complexation and Selectivity. The titrations of L4
and L5 with Zn(II) followed by UV and CD spectroscopy are
reported in Figure 9 and Figure S9 (Supporting Information)
for L4 and L5, respectively. These data show the exclusive
formation of the mononuclear species ZnL, which is also
5236
dx.doi.org/10.1021/ic5004319 | Inorg. Chem. 2014, 53, 5229−5239

Inorganic Chemistry
Article
Investigation of their Cu(I) chelating properties evidenced a
significant effect of the neighboring group carried by the D-Pen
carbonyl functions. A subtle modification at the periphery of
the pseudopeptides such as replacement of ethyl esters in L4 by
amide in L5 significantly affects the Cu(I) complex speciation.
Indeed, all spectroscopic and analytical tools point to the
formation of a unique and well-defined cluster with L4, namely
(Cu2L4)2, whereas L5 gives two types of Cu(I) complexes. The
mononuclear complex CuL5 is formed with an excess of the
ligand and the cluster (Cu2L5)2 with an excess of the metal. The
two clusters show the same features with a molecularity z = 2
corresponding to a Cu4S6 core, which is found in many
inorganic copper complexes with proteins69 including
MTs,34,35,62 copper chaperones,31,32 transcription factors,30
and the intracellular domain of copper transporters.33
Interestingly, EXAFS data prove systematically a trigonal-planar
coordination around Cu(I) with sulfurs only and Cu−S
distances (2.22−2.23 Å) characteristic of a CuS3 geometry.
Finally, these two novel ligands were demonstrated to bind
Cu(I) tightly with affinities similar to those found in proteins
and selectively with respect to Zn(II). The selectivity for Cu(I)
with respect to Zn(II) is assigned to the inability of these bulky
tripods to accommodate the preferred tetrahedral geometry of
Zn(II).70
Figure 9. Titrations of L4 (ca. 45 μM) at pH 7.4 (20 mM phosphate
buffer/MeCN, 9/1 v/v) with Zn(II): (A) UV titration with 0−4 equiv
of Zn(II); (B) CD titration with 0−2 equiv of Zn(II) (color scheme:
red, free ligand; blue, 0.25−0.75 equiv of Zn; black, 1−2 equiv of Zn).
evidenced in the ES-MS spectra (see Figures S4 and S5 in the
Supporting Information). For instance, the CD titration of L4
with Zn(II) presented in Figure 9B shows a saturation at 1
equiv of Zn(II) with the presence of an isodichroic point. The
molecularity of these complexes was confirmed by the diffusion
1
A comparison with the cysteine homologues L1 and L2
evidences a significant effect of the gem-dimethyl groups in L4
and L5. Since cysteine and penicillamine show similar
basicitiespKa =7.9 (Pen) and 8.3 (Cys)this effect can be
mainly attributed to the greater steric bulk at the Cβ of D-Pen in
comparison to cysteine. First, the influence of the neighboring
ester or amide groups is a great deal more pronounced with the
D-Pen derivatives than with the cysteine ones. Whereas L1 and
L2 showed very similar speciation of their Cu(I) complexes, L4
and L5 exhibit different Cu(I) species formation with the
exclusive presence of the cluster for the bulkier derivative L4.
The most striking effect is probably the different molecularities
of the clusters formed with the two series of ligands. The
cysteine tripods form large clusters with a molecularity of 3
coefficients measured by H NMR spectroscopy, as seen in
Table 1. In contrast, the cysteine derivatives L1 and L2 have
been found to form mixtures of monomolecular ZnL and
bimolecular (ZnL)2 complexes to satisfy the preferred
tetrahedral coordination of Zn in ZnS4 binding sites.28 This
difference can be reasonably attributed to the steric hindrance
of the two D-Pen derivatives, which cannot accommodate the
tetrahedral environment around the Zn(II) ion. It is also highly
probable that water or acetonitrile completes the coordination
spheres of the ZnL species formed with L4 and L5. A similar
effect was previously observed on comparing cysteine and
penicillamine ligands in Co(II) maquettes.40
The stability of the Zn(II) complexes was measured using
zincon (ZI) as a competitor.66 As seen in Table 4, low affinities
for Zn(II) are found for the two ligands, which are reasonably
related to their inability to provide tetrahedral Zn(II)
environments.
Consequently, it can now be confirmed that the new tripodal
ligands L4 and L5 selectively bind the Cu(I) ion, which is of
great importance for applications in vivo.
1
((Cu2L1,2)3), which show very large signals on their H NMR
spectra indicative of their rapid interconversion on the NMR
time scale. In contrast, the clusters formed with the more
hindered D-Pen tripods are more compact with a molecularity
of 2 ((Cu2L4,5)2). Moreover, the well-defined signals on the 1H
NMR spectra are consistent with rigid and well-defined
complex structures. Finally, the steric hindrance due to the
gem-dimethyl groups of the D-Pen derivatives is probably
responsible for the slightly lower stability of their Cu(I)
complexes in comparison to their cysteine homologues. In any
event, they bind Cu(I) with affinities allowing competitions
with many Cu binding sites in proteins, significantly higher than
the affinity reported for D-Pen alone.68
It is likely that the D-penicillamine-based scaffolds could also
provide efficient Cu(I) chelators in vivo. In order to validate the
efficiency to chelate intracellular copper, the amide compound
L5 is currently being functionalized to target the asialoglyco-
protein receptors and to promote its internalization in hepatic
cells. Since the D-Pen tripods are derived from non-natural
amino acids, with a D absolute configuration, they are expected
to be more stable and to display very interesting copper-
chelating properties in vivo.
DISCUSSION
■
This work was aimed at developing novel Cu(I) chelating
agents derived from D-Pen, which is the drug mostly used
currently to treat Wilson’s disease. Although the mechanism of
action of D-Pen in vivo has not been fully demonstrated, it is
expected to act as a reductive chelator that reduces Cu(II) to
Cu(I) and binds Cu(I) either in a five-membered chelate ring
involving the nitrogen and sulfur donors of the ligand or in a
Cu(I)−thiolate polymer.67 The stability constant of the Cu(I)
complex with D-Pen was estimated in a chloride-rich medium:
log β10 ≈ 12.68 Considering the promising chelating properties
of tripodal pseudopeptides based on three converging cysteine
moieties,27,28 three D-Pen units were coupled to nitrilotriacetic
acid to obtain the two D-Pen derivatives L4 and L5. In contrast
to D-Pen, these two compounds are sulfur donors only, since
the amines of the three D-Pen building blocks are involved in
peptide bonds.
5237
dx.doi.org/10.1021/ic5004319 | Inorg. Chem. 2014, 53, 5229−5239

Inorganic Chemistry
Article
(27) Pujol, A. M.; Gateau, C.; Lebrun, C.; Delangle, P. J. Am. Chem.
Soc. 2009, 131, 6928.
(28) Pujol, A. M.; Gateau, C.; Lebrun, C.; Delangle, P. Chem. Eur. J.
ASSOCIATED CONTENT
■
S
* Supporting Information
Full experimental section, details of the syntheses of ligands L4
and L5, physicochemical protocols, and data for the
physicochemical studies. This material is available free of
charge via the Internet at http://pubs.acs.org.
2011, 17, 4418.
(29) Jullien, A.-S. Thesis, Universite
France, 2013.
́
Grenoble Alpes, Grenoble,
(30) Brown, K. R.; Keller, G. L.; Pickering, I. J.; Harris, H. H.;
George, G. N.; Winge, D. R. Biochemistry 2002, 41, 6469.
(31) Pushie, M. J.; Zhang, L. M.; Pickering, I. J.; George, G. N.
Biochim. Biophys. Acta, Bioenerg. 2012, 1817, 938.
(32) Voronova, A.; Meyer-Klaucke, W.; Meyer, T.; Rompel, A.;
Krebs, B.; Kazantseva, J.; Sillard, R.; Palumaa, P. Biochem. J. 2007, 408,
139.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail for P.D.: pascale.delangle@cea.fr.
Notes
(33) Xiao, Z.; Loughlin, F.; George, G. N.; Howlett, G. J.; Wedd, A.
G. J. Am. Chem. Soc. 2004, 126, 3081.
The authors declare no competing financial interest.
(34) George, G. N.; Byrd, J.; Winge, D. R. J. Biol. Chem. 1988, 263,
8199.
(35) George, G. N.; Winge, D.; Stout, C. D.; Cramer, S. P. J. Inorg.
ACKNOWLEDGMENTS
■
This research was supported by the “Agence Nationale pour la
Recherche” (COPDETOX, ANR-11-EMMA-025), the “Fonda-
Biochem. 1986, 27, 213.
(36) Pujol, A. M.; Cuillel, M.; Jullien, A.-S.; Lebrun, C.; Cassio, D.;
Mintz, E.; Gateau, C.; Delangle, P. Angew. Chem., Int. Ed. 2012, 51,
7445.
(37) Lee, K. H.; Cabello, C.; Hemmingsen, L.; Marsh, E. N.;
Pecoraro, V. L. Angew. Chem., Int. Ed. 2006, 45, 2864.
(38) Peacock, A. F.; Stuckey, J. A.; Pecoraro, V. L. Angew. Chem., Int.
Ed. 2009, 48, 7371.
́
tion pour la Recherche Medicale” (grant DCM20111223043),
and the Labex ARCANE (Grant ANR-11-LABX-0003-01). The
authors acknowledge the European Synchotron Radiation
Facility and FAME CRG beamline for provision of synchrotron
́
radiation beamline and facilities. We thank Celine Rivaux for
her technical help in the synthesis laboratory.
(39) Chakraborty, S.; Touw, D. S.; Peacock, A. F.; Stuckey, J.;
Pecoraro, V. L. J. Am. Chem. Soc. 2010, 132, 13240.
(40) Petros, A. K.; Shaner, S. E.; Costello, A. L.; Tierney, D. L.;
Gibney, B. R. Inorg. Chem. 2004, 43, 4793.
(41) Pujol, A. M.; Lebrun, C.; Gateau, C.; Manceau, A.; Delangle, P.
Eur. J. Inorg. Chem. 2012, 3835.
(42) Jullien, A.-S.; Gateau, C.; Kieffer, I.; Testemale, D.; Delangle, P.
Inorg. Chem. 2013, 52, 9954.
(43) Kamau, P.; Jordan, R. B. Inorg. Chem. 2001, 40, 3879.
(44) Riddles, P. W.; Blakeley, R. L.; Zerner, B. Methods Enzymol.
1983, 91, 49.
(45) Bolzati, C.; Mahmood, A.; Malago, E.; Uccelli, L.; Boschi, A.;
Jones, A. G.; Refosco, F.; Duatti, A.; Tisato, F. Bioconjugate Chem.
2003, 14, 1231.
(46) Smith, N. D.; Goodman, M. Org. Lett. 2003, 5, 1035.
(47) Carpino, L. A.; Han, G. Y. J. Org. Chem. 1972, 37, 3404.
(48) Stasko, N. A.; Fischer, T. H.; Schoenfisch, M. H.
Biomacromolecules 2008, 9, 834.
(49) Kim, J. E.; Cha, E. J.; Ahn, C. H. Macromol. Chem. Phys. 2010,
211, 956.
(50) Frey, B. L.; Corn, R. M. Anal. Chem. 1996, 68, 3187.
(51) Vasak, M.; Kagi, J. H. R.; Hill, H. A. O. Biochemistry 1981, 20,
2852.
(52) Pountney, D. L.; Schauwecker, I.; Zarn, J.; Vasak, M.
Biochemistry 1994, 33, 9699.
(53) Waldeck, A. R.; Kuchel, P. W.; Lennon, A. J.; Chapman, B. E.
Prog. Nucl. Magn. Reson. Spectrosc. 1997, 30, 39.
REFERENCES
■
(1) Holm, R. H.; Kennepohl, P.; Solomon, E. I. Chem. Rev. 1996, 96,
2239.
(2) Robinson, N. J.; Winge, D. R. Annu. Rev. Biochem. 2010, 79, 537.
(3) Kim, B. E.; Nevitt, T.; Thiele, D. J. Nat. Chem. Biol. 2008, 4, 176.
(4) Arnesano, F.; Banci, L.; Bertini, I.; Ciofi-Baffoni, S. Eur. J. Inorg.
Chem. 2004, 1583.
(5) Rosenzweig, A. C. Acc. Chem. Res. 2001, 34, 119.
(6) Tao, T. Y.; Gitlin, J. A. Hepatology 2003, 37, 1241.
(7) Guo, Y.; Nyasae, L.; Braiterman, L. T.; Hubbard, A. L. Am. J.
Physiol. Gastrointest. Liver Physiol. 2005, 289, G904.
(8) Lutsenko, S.; Barnes, N. L.; Bartee, M. Y.; Dmitriev, O. Y. Physiol.
Rev. 2007, 87, 1011.
(9) http://www.orpha.net.
(10) Brewer, G. J.; Askari, F. K. J. Hepatol. 2005, 42, S13.
(11) Sarkar, B. Chem. Rev. 1999, 99, 2535.
(12) Vilensky, J. A.; Redman, K. Ann. Emerg. Med. 2003, 41, 378.
(13) Walshe, J. M. Mov. Disord. 2006, 21, 142.
(14) Brewer, G. J.; Terry, C. A.; Aisen, A. M.; Hill, G. M. Arch.
Neurol. 1987, 44, 490.
(15) Delangle, P.; Mintz, E. Dalton Trans. 2012, 41, 6359.
(16) Gateau, C.; Delangle, P. Ann. N.Y. Acad. Sci. 2014,
DOI: 10.1111/nyas.12379.
(17) Pujol, A. M.; Cuillel, M.; Renaudet, O.; Lebrun, C.;
Charbonnier, P.; Cassio, D.; Gateau, C.; Dumy, P.; Mintz, E.;
Delangle, P. J. Am. Chem. Soc. 2011, 133, 286.
(18) Rousselot-Pailley, P.; Seneque, O.; Lebrun, C.; Crouzy, S.;
Boturyn, D.; Dumy, P.; Ferrand, M.; Delangle, P. Inorg. Chem. 2006,
45, 5510.
(19) Seneque, O.; Crouzy, S.; Boturyn, D.; Dumy, P.; Ferrand, M.;
Delangle, P. Chem. Commun. 2004, 770.
(20) Hamer, D. H. Annu. Rev. Biochem. 1986, 55, 913.
(21) Palacios, O.; Atrian, S.; Capdevila, M. J. Biol. Inorg. Chem. 2011,
16, 991.
(22) Stillman, M. J. Coord. Chem. Rev. 1995, 144, 461.
(23) Vasak, M.; Meloni, G. J. Biol. Inorg. Chem. 2011, 16, 1067.
(24) Calderone, V.; Dolderer, B.; Hartmann, H. J.; Echner, H.;
Luchinat, C.; Del Bianco, C.; Mangani, S.; Weser, U. Proc. Natl. Acad.
Sci. U.S.A. 2005, 102, 51.
(54) Jershow, A.; Muller, N. J. Magn. Reson. 1997, 125, 372.
̈
(55) Sinnaeve, D. Concepts Magn. Reson. A 2012, 40A, 39.
(56) Proux, O.; Biquard, X.; Lahera, E.; Menthonnex, J. J.; Prat, A.;
Ulrich, O.; Soldo, Y.; Trevisson, P.; Kapoujyan, G.; Perroux, G.;
Taunier, P.; Grand, D.; Jeantet, P.; Deleglise, M.; Roux, J. P.;
Hazemann, J. L. Phys. Scr. 2005, T115, 970.
(57) Proux, O.; Nassif, V.; Prat, A.; Ulrich, O.; Lahera, E.; Biquard,
X.; Menthonnex, J. J.; Hazemann, J. L. J. Synchrotron Radiat. 2006, 13,
59.
(58) Kau, L. S.; Spira-solomon, D. J.; Penner-Hahn, J. E.; Hodgson,
K. O.; Solomon, E. I. J. Am. Chem. Soc. 1987, 109, 6433.
(59) Poger, D.; Fillaux, C.; Miras, R.; Crouzy, S.; Delangle, P.; Mintz,
E.; Den Auwer, C.; Ferrand, M. J. Biol. Inorg. Chem. 2008, 13, 1239.
(60) Ravel, B.; Newville, M. J. Synchrotron Radiat. 2005, 12, 537.
(61) Baumgartner, M.; Schmalle, H.; Baerlocher, C. J. Solid State
Chem. 1993, 107, 63.
(25) Faller, P. Febs J. 2010, 277, 2921.
(26) Gelinsky, M.; Vogler, R.; Vahrenkamp, H. Inorg. Chem. 2002,
41, 2560.
5238
dx.doi.org/10.1021/ic5004319 | Inorg. Chem. 2014, 53, 5229−5239

Inorganic Chemistry
Article
(62) Smith, T. A.; Lerch, K.; Hodgson, K. O. Inorg. Chem. 1986, 25,
4677.
(63) Lee, P. A.; Citrin, P. H.; Eisenberger, P.; Kincaid, B. M. Rev.
Mod. Phys. 1981, 53, 769.
(64) Xiao, Z. G.; Brose, J.; Schimo, S.; Ackland, S. M.; La Fontaine,
S.; Wedd, A. G. J. Biol. Chem. 2011, 286, 11047.
(65) Xiao, Z. G.; Wedd, A. G. Nat. Prod. Rep. 2010, 27, 768.
(66) Shaw, C. F.; Laib, J. E.; Savas, M. M.; Petering, D. H. Inorg.
Chem. 1990, 29, 403.
(67) Vortisch, V.; Kroneck, P.; Hemmerich, P. J. Am. Chem. Soc.
1976, 98, 2821.
(68) Hefter, G.; May, P. M.; Sipos, P. J. Chem. Soc., Chem. Commun.
1993, 1704.
(69) Pickering, I. J.; George, G. N.; Dameron, C. T.; Kurz, B.; Winge,
D. R.; Dance, I. G. J. Am. Chem. Soc. 1993, 115, 9498.
(70) Vallee, B. L.; Auld, D. S. Acc. Chem. Res. 1993, 26, 543.
5239
dx.doi.org/10.1021/ic5004319 | Inorg. Chem. 2014, 53, 5229−5239