Characterization of cobalt(III) hydroxamic acid complexes based on a tris(2-pyridylmethyl)amine scaffold: Reactivity toward cysteine methyl ester

Article abstract of DOI:10.1021/ic301090t

Six Co(III) complexes based on unsubstituted or substituted TPA ligands (where TPA is tris(2-pyridylmethyl)amine) and acetohydroxamic acid (A), N-methyl-acetohydroxamic acid (B), or N-hydroxy-pyridinone (C) were prepared and characterized by mass spectrometry, elemental analysis, and electrochemistry: [Co(III)(TPA)(A-2H)](Cl) (1a), [Co(III)((4-Cl₂)TPA)(A-2H)](Cl) (2a), [Co(III)((6-Piva)TPA)(A-2H)](Cl) (3a), [Co(III)((4-Piva)TPA)(A-2H)](Cl) (4a) and [Co(III)(TPA)(B-H)](Cl)₂ (1b), and [Co(III)(TPA)(C-H)](Cl) ₂ (1c). Complexes 1a-c and 3a were analyzed by ¹H NMR, using 2D (¹H, ¹H) COSY and 2D (¹H, ¹³C) HMBC and HSQC, and shown to exist as a mixture of two geometric isomers based on whether the hydroxamic oxygen was trans to a pyridine nitrogen or to the tertiary amine nitrogen. Complex 3a exists as a single isomer that was crystallized. Its crystal structure revealed the presence of an H-bond between the pivaloylamide and the hydroximate oxygen. Complexes 1a, 2a, and 4a are irreversibly reduced beyond -900 mV versus SCE, while complexes 1b and 1c are reduced at less negative values of -330 and -190 mV, respectively. The H-bond in 3a increased the redox potential up to -720 mV. Reaction of complex 1a with l-cysteine methyl ester CysOMe was monitored by ¹H NMR and UV-vis at 2 mM and 0.2 mM in an aqueous buffered solution at pH 7.5. Complex 1a was successively converted into an intermediate [Co(III)(TPA)(CysOMe-H)] ²⁺, 1d, by exchange of the hydroximate with the cysteinate ligand, and further into Co(III)(CysOMe-H)₃, 5. An authentic sample of 1d was prepared and thoroughly characterized. A detailed ¹H NMR analysis showed there was only one isomer, in which the thiolate was trans to the tertiary amine nitrogen.

Full text of DOI:10.1021/ic301090t

Article
pubs.acs.org/IC
Characterization of Cobalt(III) Hydroxamic Acid Complexes Based on
a Tris(2-pyridylmethyl)amine Scaffold: Reactivity toward Cysteine
Methyl Ester
Mickael Alimi,† Anas Allam,† Mohamed Selkti,‡ Alain Tomas,‡ Pascal Roussel,§ Erwan Galardon,†
and Isabelle Artaud*,†
†Laboratoire de Chimie et Biochimie Pharmacologiques et Toxicologiques, UMR 8601 CNRS, Universite
́
Paris Descartes, Sorbonne
Paris Descartes, Sorbonne Paris Cite, 4 avenue
§Unite
́ ́
de Catalyse et Chimie du Solide, UMR 8012 CNRS, Ecole Nationale Superieure de Chimie de Lille BP 90108, 59652
́ ̀
Paris Cite, 45 rue des Saints Peres, 75270 Paris Cedex 06, France
‡Laboratoire de Crystallographie et RMN Biologiques, UMR 8015 CNRS, Universite
de l’Observatoire, 75270 Paris Cedex 06, France
́
́
Villeneuve d’Ascq Cedex, France
S
* Supporting Information
ABSTRACT: Six Co(III) complexes based on unsubstituted or substituted TPA ligands (where TPA is tris(2-
pyridylmethyl)amine) and acetohydroxamic acid (A), N-methyl-acetohydroxamic acid (B), or N-hydroxy-pyridinone (C)
were prepared and characterized by mass spectrometry, elemental analysis, and electrochemistry: [Co(III)(TPA)(A-2H)](Cl)
(1a), [Co(III)((4-Cl2)TPA)(A-2H)](Cl) (2a), [Co(III)((6-Piva)TPA)(A-2H)](Cl) (3a), [Co(III)((4-Piva)TPA)(A-2H)](Cl)
(4a) and [Co(III)(TPA)(B−H)](Cl)2 (1b), and [Co(III)(TPA)(C−H)](Cl)2 (1c). Complexes 1a−c and 3a were analyzed by
1
1H NMR, using 2D (1H, H) COSY and 2D (1H, 13C) HMBC and HSQC, and shown to exist as a mixture of two geometric
isomers based on whether the hydroxamic oxygen was trans to a pyridine nitrogen or to the tertiary amine nitrogen. Complex 3a
exists as a single isomer that was crystallized. Its crystal structure revealed the presence of an H-bond between the pivaloylamide
and the hydroximate oxygen. Complexes 1a, 2a, and 4a are irreversibly reduced beyond −900 mV versus SCE, while complexes
1b and 1c are reduced at less negative values of −330 and −190 mV, respectively. The H-bond in 3a increased the redox
potential up to −720 mV. Reaction of complex 1a with L-cysteine methyl ester CysOMe was monitored by 1H NMR and UV−
vis at 2 mM and 0.2 mM in an aqueous buffered solution at pH 7.5. Complex 1a was successively converted into an intermediate
[Co(III)(TPA)(CysOMe-H)]2+, 1d, by exchange of the hydroximate with the cysteinate ligand, and further into
1
Co(III)(CysOMe-H)3, 5. An authentic sample of 1d was prepared and thoroughly characterized. A detailed H NMR analysis
showed there was only one isomer, in which the thiolate was trans to the tertiary amine nitrogen.
hardly possible because the redox potential required is very low
and beyond the redox value inside the cell (around −250 mV vs
NHE7). More recently, Hambley and co-workers proposed that
reductants could also act via ligand exchange, and that, in
Co(III)(TPA)(hydroximate) metallodrugs, the drug could be
displaced inside the cell by an endogenous ligand.8 However,
the limitation of this second hypothesis is the well-known
inertness of Co(III) complexes, while ligands are labile in the
Co(II) state.9 In this paper, using model complexes based on
INTRODUCTION
■
Development of metallodrugs as prodrugs to deliver an active
molecule to its target appears as a challenging alternative to
access new pharmaceuticals.1−3 The major examples reported
so far refer to cobalt(III) complexes of drugs with anticancer
activity, such as those based on nitrogen mustards4 and more
recently marimastat.5 The release of the drug inside the cells
was first proposed to proceed by reduction of the Co(III)
metallodrug to a Co(II) state and dissociation from this
unstable Co(II) complex.6 While the activity of mustard
complexes follows this trend,4 in the case of the hydroxamic
acid, marimastat,5 the reduction with biological reductants is
Received: May 24, 2012
Published: August 20, 2012
© 2012 American Chemical Society
9350
dx.doi.org/10.1021/ic301090t | Inorg. Chem. 2012, 51, 9350−9356

Inorganic Chemistry
Article
substituted or unsubstituted TPA (tris(2-pyridylmethyl)amine)
and hydroxamic acids, we provide some new evidence regarding
the mode of dissociation of the drug from the original complex,
i.e., reduction or ligand exchange. We first detail a complete
study, on the influence on the Co(III)/Co(II) reduction, of the
two partners of the Co(III)(TPA)(hydroximate), the ancillary
ligand TPA, and the hydroxamic acid. Using acetohydroxamic
acid, we introduced chloro substituents on the TPA ligand to
try to tune the redox potential by electronic effects, but without
any success. However, introduction of a pivaloylamido group in
ortho position of one pyridine ring nitrogen promotes a more
drastic change of the redox value. Then, using the unsubstituted
TPA, we turned to N-methyl-acetohydroxamic acid and N-
hydroxy-pyridinone as other models of inhibitors of this class10
to investigate their influence on the reduction potential. All the
dioxygen oxidation of the Co(II) species formed upon
treatment of the Co(II)((X)TPA)Cl2 complexes (X = H, 4-
Cl2, 4-Piva, and 6-Piva), prepared in situ in methanol, with
acetohydroxamic acid (A), N-methyl-acetohydroxamic acid
(B), or N-hydroxy-pyridinone (C), in the presence of 2 (A)
or 1 (B and C) equiv of KOH. Five complexes were readily
obtained and characterized: [Co(III)(TPA)(A-2H)](Cl) (1a),
[Co(III)((4-Cl2)TPA)(A-2H)](Cl) (2a), [Co(III)((4-Piva)-
TPA)(A-2H)](Cl) (4a), [Co(III)(TPA)(B−H)](Cl)2 (1b),
and [Co(III)(TPA)(C−H)](Cl)2 (1c). Unlike complexes 2a
and 4a for which there are numerous isomers, complexes 1a,
1b, and 1c exist only as a mixture of two geometric isomers,
based on whether the hydroxamic oxygen is trans to a pyridine
nitrogen or to the tertiary amine nitrogen. Thus, 2a and 4a
were only characterized by mass spectrometry (ESI+ in
CH3CN) and elemental analysis, and the other complexes
were in addition thoroughly characterized by 1H NMR
1
complexes were thoroughly characterized by H NMR, mass
spectrometry, elemental analysis, and cyclic voltammetry. In a
second part, we focused on the Co(III)(TPA)-
(acetohydroximate) complex, and we discussed the outcome
of its reaction with L-cysteine methyl ester (CysOMe) as model
of intracellular thiol reductants. In this study, we prepared and
characterized a new complex, [Co(III)(TPA)(CysOMe-H)]-
(BPh4)2, the intermediate resulting from the ligand exchange
between the cysteine derivative and the hydroximate ligand.
1
spectroscopy, since a complete assignment of their H NMR
resonances was possible using homonuclear 2D (1H, 1H)
COSY and heteronuclear 2D (1H, 13C) HMBC and HSQC
correlations. All the spectra of the complexes are shown with
proton assignment in Supporting Information (Figure S1).
Complex 1a appears as a 50:50 mixture of the two isomers on
the basis of the integration ratio of the signals attributed to
protons in ortho position of the pyridine nitrogens. For the
pyridine ring in the mean plane containing the hydroximate,
this proton is highly deshielded and appears in the two isomers
as a doublet at 9.33 and 9.18 ppm, while the two corresponding
protons of the pyridines in axial position are equivalent and
appear as doublets at 8.20 and 8.52 ppm in the two isomers,
respectively. The methyl of the hydroximate is located at 1.91
and 1.36 ppm in the two isomers. Similarly, complexes 1b and
1c were isolated as 40:60 and 30:70 mixtures of the two
isomers.
RESULTS AND DISCUSSION
Synthesis of Ligands and Complexes. Ligands and
complexes used in this study are depicted in Scheme 1.
■
Scheme 1. Ligands and Hydroxamic Acids Used in This
Study
Characterization of Complex [Co((6-Piva)TPA)(A-2H)]
Cl (3a). Following the procedure aforementioned, reaction of
(6-Piva)TPA with CoCl2 and acetohydroxamic with two base
equivalents did not result in the unique formation of the
1
expected complex 3a, [Co((6-Piva)TPA)(A-2H)]Cl. H NMR
analysis of the raw material obtained at the end of the reaction
revealed a 60:40 mixture of complexes 3a and 3′, this latter was
identified as [Co(III)(6-Piva-H)(TPA)(OMe)]Cl (Scheme 2).
An authentic sample of complex 3′ was synthesized by adding 2
equiv of KOH to a methanolic solution of (6-Piva)TPA and
CoCl2, followed by air oxidation. Mass spectrometry, elemental
Syntheses of ligands TPA,11 (4-Cl2)TPA,12 and (6-Piva)TPA13
have been previously reported in the literature except N-((4-
(pivaloylamido)-2-pyridyl)methyl)-N,N′-bis((2-pyridyl)-
methyl)amine, (4-Piva)TPA, that we prepared from N-((4-
nitro)-2-pyridyl)methyl)-N,N′-bis((2-pyridyl)methyl)amine12
in two steps upon reduction in ethanol with hydrazine hydrate
in the presence of a catalytic amount of Pd/C14 followed by
condensation in CH2Cl2 of the resulting amine with pivaloyl
chloride. Synthesis of the Co(III) complexes was achieved by
1
analysis, and H NMR spectroscopy are consistent with the
proposed structure. Owing to the strong Lewis acidity of the
cobalt(III) ion, in this complex, the pivaloylamide is likely
deprotonated, resulting in the coordination of the iminolate
oxygen to the Co(III) as previously observed in the X-ray
structures of related TPA complexes such as [Ni(II)(6-Piva-
Scheme 2. Preparation of [Co(III)(6-Piva)TPA)(A-2H)]Cl, 3a
9351
dx.doi.org/10.1021/ic301090t | Inorg. Chem. 2012, 51, 9350−9356

Inorganic Chemistry
Article
H)TPA)]15 and [Zn(II)(6-Piva-H)TPA)].13 This is supported
by the FTIR spectrum of 3′, in which the CO stretching
vibration was shifted to a lower energy relative to the free
ligand (1609 cm−1 for 3′ and 1683 cm−1 for (6-Piva)TPA)) and
relative to a pivaloylamide carbonyl oxygen bound to a Co(II)16
(νCO = 1618 cm−1). The sixth ligand in 3′ is likely a
methanolate since the mass spectrum, recorded in acetonitrile
or dichloromethane, showed a parent peak at m/z = 478
corresponding to [Co(III)(6-Piva-H)(TPA)(OMe)]+. Com-
plex 3a was readily separated from 3′ by chromatography over
Sephadex LH-20, eluting first with methanol as a green band.
Electrochemistry. The cyclic voltammogram of all the
complexes 1a−c, 2a, 3a, and 4a in DMF displayed an
irreversible reduction attributed to the Co(III)/Co(II) couple
(Figure S2 and Table 1). In the acetohydroximate series, while
Table 1. Epc Values for the Co(III) Complexes Derived from
Hydroxamic Acids
complex
Epc vs SCE (mV)
a
[Co(III)(TPA)(A-2H)](Cl), 1a
−990
[Co(III)(TPA)(B−H)](Cl)2, 1b
[Co(III)(TPA)(C−H)](Cl)2, 1c
[Co(III)((4-Cl2)TPA)(A-2H)](Cl), 2a
[Co(III)(6-Piva)TPA)(A-2H)]Cl, 3a
[Co(III)((4-Piva)TPA)(A-2H)](Cl), 4a
−330
−190
−960
−720
−1060
1
Its H NMR analysis (Figure S1) revealed the presence of a
single stereoisomer that was crystallized as its BPh4 salt. The X-
ray structure of 3a(BPh4) shown in Figure 1 is quite similar to
a
This value is in the range of that previously reported for 1a by Failes
et al. in ref 17.
complexes 1a,17 2a, and 4a exhibited Epc values in the range
−990 to −1060 mV versus SCE, Epc value of 3a was less
negative (+270 mV shift relative to 1a). The hydroxamate
derivatives, 1b and 1c, were also reduced at a more anodic
potential, −330 mV for [Co(III)(TPA)(B−H)](Cl)2 (1b), and
−190 mV for [Co(III)(TPA)(C−H)](Cl)2 (1c). Introducing
an electron withdrawing group in para position of the TPA
pyridine moieties does not affect the Co(III)/Co(II) reduction,
and clearly, stabilization of the Co(III) state by the hydroximate
prevails over electronic effect on the TPA ligand. However,
replacing the hydroximate by a hydroxamate results in an
increase of the overall charge of the complex and destabilizes
the Co(III) state. We previously observed the same tendency
with [Fe(III)(TPA)(hydroximate)]+ relative to [Fe(III)(TPA)-
(hydroxamate)]2+ complexes,20 and during the submission of
this paper, P. D. Bonnitcha et al. reported a similar result, upon
conversion of Co(III)(TPA)(hydroximate) to Co(III)(TPA)-
(hydroxamate) by protonation.21 Interestingly, the hydrogen
bond between the hydroximate oxygen and the pivaloylamide
in complex 3a, by decreasing the charge of the bound
hydroximate, makes the Co(III)/Co(II) reduction easier. So,
introducing a hydrogen bond donor in the coordination sphere
of the Co(III)(hydroximate) complexes appears as an
alternative to tune the reduction potential of hydroximate-
based metallodrugs.
Reactivity of Complex 1a toward L-Cysteine Methyl
Ester. Even though the reduction potentials measured by
electrochemistry are not thermodynamic, since the processes
are irreversible, they indicate that some cobalt(III) complexes
could be reduced within a range of potential accessible to
biological reductants. However, this could not be the case of
the hydroximato derivatives which are reduced at much lower
potentials. To get some insights on the possible release of the
hydroxamic acid by an alternative ligand exchange pathway as
previously suggested,8 we focused on the reactivity of 1a toward
L-cysteine methyl ester in Tris buffer solution at pH 7.5.
Compound 1a was stable in this buffer, and no degradation was
observed over 24 h by UV−vis spectroscopy. Reaction of 1a
with 10 equiv of L-cysteine methyl ester hydrochloride was
monitored, under argon, by 1H NMR and UV−vis spectroscopy
at concentrations of 1a of 2 mM. At this concentration, we
observed by NMR the disappearance of the typical signals
attributed to 1a (9.33 and 8.20 for one isomer, and 9.18 and
8.52 for the other) and the appearance of those of the free TPA
ligand. The conversion was complete within 3 h (Figure 2). In
Figure 1. ORTEP view of 3a(BPh)4 showing thermal ellipsoid at 50%
probability and atom labeling. Hydrogen atoms and BPh4 anions are
omitted for clarity. Selected bond distances (Å) and angles (deg):
Co(1)−O(1) 1.842(2), Co(1)−O(2) 1.862(2), Co(1)−N(1)
1.978(3), Co(1)−N(2) 1.914(3), Co(1)−N(3) 1.902(3), Co(1)−
N(4) 1.943(3), O(1)−N(5) 1.423(4), N(5)−C(24) 1.292(5),
C(24)−O(2) 1.302(5), O(1)−Co(1)−N(4) 178.35(12), N(1)−
Co(1)−N(2) 168.88(13), O(2)−Co(1)−N(3) 177.53(12), O(1)−
Co(1)−O(2) 85.86(11), N(2)−Co(1)−N4) 85.18(13), N(2)−
Co(1)−O(1) 93.21(12), N(4)−Co(1)−O(2) 93.79(12).
those previously reported by Failes et al. for similar
[Co(III)(TPA)(hydroximate)]+ complexes.5,17 The geometry
around the Co(III) is slightly distorted octahedral, and the
acetohydroximate adopts a bidentate chelation mode to the
Co(III) with the hydroximate oxygen trans to the tertiary amine
nitrogen and the carbonyl oxygen trans to the pyridine
nitrogen. The main difference with the structure of [Co(III)-
(TPA)(A-2H)]+, 1a,17 is the existence of a hydrogen bond
between the NH of the pivaloylamide and the hydroximate
oxygen (N6···O1 2.688 Å), which induces a slight elongation of
the corresponding cobalt nitrogen axial bond (Co···N1 1.978
and Co···N2 1.914 Å). The other bond lengths are in the range
of those previously reported.17 This complex is a new example
in which a metal ligand interaction is stabilized through
intramolecular hydrogen bonding.18,19
9352
dx.doi.org/10.1021/ic301090t | Inorg. Chem. 2012, 51, 9350−9356

Inorganic Chemistry
Article
specific assay with iron(III) chloride.23 The cobalt cation was
first trapped with Chelex. Evaporating the solvent and adding
FeCl3 to an acetone solution of the residue, once filered,
promoted the appearance of the expected deep magenta color
of the tris hydroxamate iron(III) complex. Then, to try and
identify an intermediate before the complete ligand exhange
leading to the formation and precipitation of 5, we performed
experiments in more dilute conditions, and monitored the
reaction, by UV−vis spectroscopy, between 1a at 0.2 mM and
CysOMe·HCl at different pH. We indeed observed the clean
conversion of 1a into an intermediate with bands at 260 and
310 nm, whose rate of formation decreased as the buffer
became more acidic, showing that the reactive species was the
deprotonated form of the cysteine methyl ester. The conversion
was complete within 2 h at pH 8.0 for a CysOMe·HCl/1a ratio
of 5 (Figure S3), while it required about 7 h with a
CysOMe.HCl/1a ratio of 10 at pH 6.5 (Figure S4). The
1
Figure 2. H NMR monitoring of the reaction of complex 1a (2 mM
in 200 mM Tris buffer, pH = 7.5) with 10 equiv of L-cysteine methyl
ester hydrochloride: (a) spectrum of complex 1a, (b and c) spectra
recorded after 1 and 3 h, respectively.
1
UV−vis, we observed in the cuvette the progressive formation
of a green precipitate. Its spectrum, in Tris buffer pH 7.5/
DMSO 9:1, which displayed bands at 280, 450, and 586 nm
(Figure 3), was identical to that of an authentic sample of
reaction was then monitored by H NMR (Figure 4A) and
UV−vis (Figure 4B) at pH 7.5. The species with bands at 260
and 310 nm, formed within 3 h, displayed in 1H NMR signals at
9.19, 9.01, and 8.89 ppm. This complex, supposed to be
[Co(III)(TPA)(CysOMe-H)]2+, was prepared by reaction, in
methanol, of [Co(III)(TPA)(Cl)2](ClO4) with cysteine methyl
ester hydrochloride in the presence of 2 equiv of KOH,
followed by counteranion exchange with NaBPh4. This new
complex 1d, [Co(III)(TPA)(CysOMe-H)](Cl)0.75(BPh4)1.25
,
was characterized by mass spectrometry and elemental analysis.
1
It showed the same H NMR and UV−vis characteristics as
those of the intermediate discussed above. Although we could
not get crystals of 1d suitable for X-ray analysis, since it is very
sensitive to dioxygen, we determined its structure in solution by
classical NMR experiments under argon (Scheme 3). In
contrast to complexes 1a−c, which were obtained as mixture
of two stereoisomers, 1d was formed as a single stereoisomer.
This is highlighted by the presence of only one set of doublets
corresponding to the three ortho pyridine protons at 9.19, 9.01,
and 8.89 ppm, 1d(BPh4) (Figure S1). The 2D NOESY
experiment, which revealed several correlations between the
TPA and the cysteine fragments, supported the expected
bidentate N/S cordination of the cysteine to the cobalt(III)
center and allowed us to conclude that the thiolate was trans to
the tertiary amine nitrogen atom. This assignement was based
on (i) correlations observed between each of the ortho pyridine
protons in axial position, H1a at 9.19 ppm and H1b at 9.01 ppm,
and one diastereotopic proton from the CH2 group of the
Figure 3. Superposition of the UV−vis spectra (1.2 mM in 200 mM
Tris buffer pH 7.5/DMSO, 90:10) of the precipitate formed upon
reaction of 1a with 10 equiv of L-cysteine methyl ester: precipitate
(blue trace), reference complex 5 (red trace).
Co(III)(CysOMe-H)3 (5), prepared by reaction, in water, of
CysOMe·HCl with Na3[Co(CO3)3]·3H2O as previously
reported.22 Direct identification of the released hydroxamic
acid in the final NMR spectrum was difficult due to the excess
of cysteine methyl ester, and we put it in evidence thanks to the
Figure 4. Reaction of complex 1a (0.2 mM in 200 mM Tris buffer, pH 7.5) with 10 equiv of L-cysteine methyl ester hydrochloride. Evolution of the
(A) 1H NMR (500 MHz) and (B) UV−vis spectra: (a) spectrum of 1a, (b) traces after 3 h of reaction, (c) spectrum of the reference complex 1d.
9353
dx.doi.org/10.1021/ic301090t | Inorg. Chem. 2012, 51, 9350−9356

Inorganic Chemistry
Article
Scheme 3. Reaction of [Co(TPA)(A-2H)]Cl, 1a, with L-Cysteine Methyl Ester in an Aqueous Buffered Solution at pH 7.5
with chemical shifts reported in ppm relative to the residual solvent.
Cyclic voltammetry experiments were performed at room temperature
on a PGZ100/HVB100 setup (radiometer analytical) with a three-
electrode system which consists of a NaCl saturated calomel electrode
(SCE), a platinum auxiliary electrode, and a glassy carbon working
electrode. The samples were dissolved in deaerated DMF (0.01 M).
nBu4NClO4 was used as supporting electrolyte (0.10 M). The
potential sweep rate was 100 mV s−1, and ferrocene was used as
internal standard (E1/2 = 342 vs mV/SCE). IR spectra were obtained
with a Perkin-Elmer Spectrum One FT-IR spectrometer equipped with
a MIRacleTM single reflection horizontal ATR unit (Zirconium−
Selenium crystal).
Materials. Organic solvents were dried and freshly distilled before
use following standard procedures. Tris buffer (200 mM, pH = 7.5)
was prepared by dissolving the required amount of TRIS AMINO in
deionized water, and by adjusting the pH with concentrated aqueous
HCl.
cysteine fragment at 2.57 and 2.63 ppm, respectively (see
Scheme 3 for the labeling), (ii) correlations between each of
the NH2 protons at 5.34 and 5.18 ppm, and one proton of the
two benzylic moieties from TPA apart from the mean plane
containing the cysteine, H5a and H5b at 5.70 and 5.93 ppm,
respectively. Then, due to the chirality of the cysteine methyl
ester, there is only one enantiomer, and, therefore, only one
correlation between one axial pyridine ortho proton, namely
H1a, and the CH of the cysteine at 3.49 ppm. Clearly, reaction
of complex 1a with the cysteine ester proceeded through ligand
exchange, hydroximate/L-cysteine methyl ester, as shown in
Scheme 3. We did not observe, by EPR, under argon the
formation of any Co(II) species. Reaction of 1d at 0.2 mM with
an excess of CysOMe·HCl led to the slow release of TPA as
1
shown by H NMR (Figure S5).
1
To complete this study, we also monitored by H NMR the
Synthesis of N-((4-(Pivaloylamido)-2-pyridyl)methyl)-N,N′-
bis((2-pyridyl)methyl)amine, (4-Piva)TPA). To a solution of N-
((4-nitro)-2-pyridyl)methyl)-N,N′-bis((2-pyridyl)methyl)amine12
(100 mg, 0.30 mmol) in dry ethanol (5 mL) under argon was added a
catalytic amount of Pd/C (10 mg). After stirring under reflux for 1 h,
hydrazine hydrate (20 equiv, 302 mg, 5.96 mmol) was added, and the
mixture was stirred under reflux for 2 additional hours. After cooling
and filtration through Celite, ethanol was removed under reduced
pressure, and then the residue was dissolved in CH2Cl2 and washed 3
times with water. After drying over MgSO4 the organic layer was
evaporated to give an oily residue which was triturated with diethyl
ether to afford N-((4-amino-2-pyridyl)methyl)-N,N′-bis((2-pyridyl)-
methyl)amine as a colorless solid (91 mg, quantitative yield). 1H NMR
(δ, 250 MHz, CDCl3): 8.51 (d, 2H, 3J = 4.9 Hz), 8.11 (d, 1H, 3J = 5.9
Hz), 7.58 (m, 4H), 7.12 (t, 2H, 3J = 6,4 Hz), 6.83 (d, 1H, 3J = 2.0 Hz),
6.36 (dd, 1H, 3J = 5.9 Hz, 4J = 2.0 Hz), 4.15 (s, 2H), 3.85 (s, 4H), 3.72
(s, 2H). 13C NMR (δ, 62.5 MHz, CDCl3): 160.0, 159.6, 153.5, 149.8,
149.3, 136.6, 123.3, 122.1, 108.7, 108.6, 60.4, 60.3. MS (ESI+, CH3CN,
m/z): 306.1 (100%, [C18H19N5 + H]+).
reaction of 1b (0.2 mM) with 10 equiv of CysOMe·HCl at pH
7.5; 1b was equally converted into 1d, and after 3 h, the
conversion rate was roughly similar to that of 1a into 1d
(Figure S6).
CONCLUSION
■
In this paper, we have shown that the reduction potential of
Co(III) metallodrugs built on TPA and hydroxamic acids is
mainly governed by the state of deprotonation of the
hydroxamic acid. With N-substituted hydroxamic acid or N-
hydroxypyridinone, which act as monoanionic hydroxamato
ligands, the Co(III)/Co(II) reduction potential can fall in the
range of those of biological reductants. On the contrary,
hydroxamic acids, which bind to the Co(III) center in their
dianionic hydroximato form, are reduced at very low potentials.
However, introducing a hydrogen bond donor on the TPA
ligand such as a pivaloylamide in ortho position of a pyridine
nitrogen, appears as a suitable alternative to raise the potential
to a less negative value. With, by far less reducible complexes,
the hydroxamic acid is very likely released from the
[Co(III)(TPA)(hydroximate)]+ metallodrug, inside the cell,
by a ligand exchange with a biological thiol. Monitoring the
reaction of such a complex with L-cysteine methyl ester, in a
buffered solution at pH 7.5, by 1H NMR and UV−vis
spectroscopies, is in agreement with the intermediate formation
of the [(Co(III)(TPA)(CysOMe-H)]2+ complex, without any
reduction of the Co(III) to the Co(II) state. An excess of L-
cysteine methyl ester ultimately leads to the release of the TPA
ligand and to the formation of Co(III)(CysOMe-H)3.
Triethylamine (1 equiv, 415 μL, 2.91 mmol) was added to a
solution of N-((4-amino-2-pyridyl)methyl)-N,N′-bis((2-pyridyl)-
methyl)amine (900 mg, 2.91 mmol) in dry CH2Cl2 (10 mL). Then,
pivaloyl chloride (1 equiv, 362 μL, 2.91 mmol) was added dropwise at
0 °C over 10 min. After stirring for 12 h at room temperature (rt), the
solvent was removed under reduced pressure, and the residue was
purified by flash chromatography (silica gel, CH2Cl2/MeOH (100:0 to
1
95:5)) to yield (4-Piva)TPA as a colorless oil (940 mg, 80%). H
NMR (δ, 500 MHz, CDCl3): 8.51 (d, 2H, 3J = 5.0 Hz), 8.38 (d, 1H, 3J
= 5.7 Hz), 7.62 (m, 3H), 7.59 (s, 1H), 7.54 (m, 3H), 7.12 (t, 2H, 3J =
6.2 Hz), 3.86 (s, 4H), 3.82 (s, 2H), 1.31 (s, 9H). 13C NMR (δ, 125
MHz, CDCl3): 177.4, 160.5, 159.3, 150.3, 149.3, 145.9, 136.7, 123.6,
122.4, 112.9, 112.5, 60.5, 60.4, 40.2, 27.7. MS (ESI+, CH3CN, m/z):
390.2 (100%, [(4-Piv)TPA + H]+). Anal. Calcd (Found) for
C23H27N5O·0.7H2O: C, 68.70 (68.80); H, 7.12 (7.05); N, 17.42
(17.35).
EXPERIMENTAL SECTION
■
Physical Measurements. ESI-MS spectra were recorded on a
Thermo Finnigan LCD Advantage spectrometer. UV−vis spectra were
recorded on a Jasco V-570 spectrometer. Elemental analyses were
carried out in CNRS at Gif-sur-Yvette. 1H NMR and 13C NMR spectra
were recorded on Bruker ARX-250 or AVANCEII-500 spectrometers,
Typical Experiment for the Preparation of Hydroximato and
Hydroxamato Complexes. A mixture of ligand (0.35 mmol) and
CoCl2 (45 mg, 0.35 mmol) in methanol (10 mL) was stirred under
argon at rt for 15 min. A mixture of the hydroxamic acid (0.35 mmol)
and solid KOH (0.70 mmol for complexes 1a, 2a, 3a, and 4a, or 0.35
9354
dx.doi.org/10.1021/ic301090t | Inorg. Chem. 2012, 51, 9350−9356

Inorganic Chemistry
Article
mmol for complexes 1b and 1c) in methanol (2 mL) was then added.
After 15 min the reaction vessel was flushed with dioxygen and stirred
overnight at rt. The greenish solution was concentrated, and the
residue was purified over Sephadex LH-20 using methanol as eluent.
Complexes were isolated as mixtures of diastereoisomers by adding
diethyl ether to a methanol solution of the complexes.
7.33 (d, 1H, 3J = 8.0 Hz), 7.21 (d, 1H, 3J = 7.4 Hz), 5.59 (d, 1H, 2J =
−14.8 Hz), 5.55 (d, 1H, 2J = −16.1 Hz), 5.11 (d, 1H, 2J = −19.0 Hz),
5.05 (d, 1H, 2J = −14.8 Hz), 5.10 (d, 1H, 2J = −16.1 Hz), 5.02 (d, 1H,
2J = −19.0 Hz), 1.61 (s, 3H), 1.55 (s, 9H). Cyclic voltammetry Epc:
−720 mV vs SCE. MS (ESI+, CH3CN, m/z): 521.1 (100%, [Co((6-
P i v a ) T P A ) ( A - 2 H ) ] + ) . A n a l . C a l c d ( F o u n d ) f o r
C25H30ClCoN6O3·H2O: C, 52.23 (51.98); H, 5.61 (5.71); N, 14.62
(14.66).
Crystals suitable for X-ray analysis were obtained after exchange of
the counteranion as follows: sodium tetraphenylborate (1.2 equiv) was
added to a concentrated solution of 3a in methanol, and the complex
was isolated by filtration. Crystals of 3a(BPh4) were grown by slow
diffusion of methanol in a concentrated solution of complex in
CH2Cl2.
Complex [Co(III)(TPA)(A-2H)](Cl), 1a.17 Green powder (79 mg,
47%). 1H NMR (δ, 500 MHz, DMSO) major isomer: 9.18 (d, 1H, 3J =
5.8 Hz), 8.52 (d, 2H, 3J = 5.9 Hz), 7.98 (t, 2H, 3J = 7.6 Hz), 7.84 (dd,
1H, 3J = 8.0 Hz, 3J = 7.8 Hz), 7.62 (d, 2H, 3J = 7.6 Hz), 7.56 (dd, 1H,
3J = 7.8 Hz, 3J = 5.8 Hz), 7.50 (dd, 2H, 3J = 7.6 Hz, 3J = 5.9 Hz), 7.29
3
2
2
(d, 1H, J = 8.0 Hz), 5.39 (d, 2H, J = −16.1 Hz), 5.27 (d, 2H, J =
−16.1 Hz), 5.10 (s, 2H), 1.36 (s, 3H). Minor isomer: 9.33 (d, 1H, 3J =
5.8 Hz), 8.20 (d, 2H, 3J = 6.0 Hz), 7.98 (t, 2H, 3J = 7.6 Hz), 7.90 (dd,
1H, 3J = 8.1 Hz, 3J = 7.7 Hz), 7.64 (dd, 1H, 3J = 7.7 Hz, 3J = 5.8 Hz),
7.60 (d, 2H, 3J = 7.6 Hz), 7.47 (dd, 2H, 3J = 7.6 Hz, 3J = 6.0 Hz), 7.35
Complex [Co(III)((4-Piva)TPA)(A-2H)](Cl), 4a. Green powder
(86.8 mg, 40%). Cyclic voltammetry Epc: −1060 mV vs SCE. MS
(ESI+, CH3CN, m/z): 521.1 (100%, [Co(III)((4-Piva)TPA)(A-
2H)]+). Anal. Calcd (Found) for C25H30ClCoN6O3·3.5H2O: C,
48.43 (48.35); H, 6.02 (5.80); N, 13.56 (13.27).
3
2
2
(d, 1H, J = 8.1 Hz), 5.24 (d, 2H, J = −16.4 Hz), 5.00 (d, 2H, J =
−16.4 Hz), 4.95 (s, 2H), 1.91 (s, 3H). UV−vis (Tris buffer, pH = 7.5):
λ (ε, M−1 cm−1) 581 (118) and 461 (sh). Cyclic voltammetry: Epc
−990 mV vs SCE. MS (ESI+, CH3CN, m/z): 421.9 (100%,
[Co(III)(TPA)(A-2H)]+ ). Anal. Calcd (Found) for
C20H21ClCoN5O2·1.2H2O: C, 50.11 (50.35); H, 4.92 (5.00); N,
14.61 (14.36).
Complex [Co(III)(6-Piva-H)(TPA)(OMe)]Cl, 3′. After stirring for
15 min a mixture of CoCl2 (32 mg, 0.26 mmol) and (6-Piva)(TPA)
(100 mg, 0.26 mmol) in methanol (3 mL) and 2 equiv of solid KOH
(34 mg, 0.52 mmol) were added to the purple solution. Then the
reaction vessel was opened to air, and the orange solution gradually
turned red. Two hours later, the solution was concentrated and the
mixture purified over Sephadex LH-20 (eluent methanol) to afford
Complex [Co(III)(TPA)(B−H)](Cl)2, 1b. Purple powder (90 mg,
45%). 1H NMR (δ, 500 MHz, DMSO) major isomer: 8.74 (d, 1H, 3J =
5.9 Hz), 8.60 (d, 2H, 3J = 6.1 Hz), 8.15 (t, 2H, 3J = 7.6 Hz), 7.98 (dd,
1H, 3J = 8.1 Hz, 3J = 7.6 Hz), 7.80 (d, 2H, 3J = 7.6 Hz), 7.70 (dd, 1H,
3J = 7.6 Hz, 3J = 5.9 Hz), 7.67 (dd, 2H, 3J = 7.6 Hz, 3J = 6.1 Hz), 7.38
1
after precipitation in diethyl ether a red powder (119 mg, 85%). H
NMR (δ, 500 MHz, CD3OD): 9.02 (d, 1H, 3J = 5.9 Hz), 8.13 (dd, 1H,
3
3
3
3J = 8.1 Hz, J = 7.6 Hz), 8.04 (d, 1H, J = 6.0 Hz), 7.86 (t, 1H, J =
7.7 Hz), 7.78 (dd, 1H, 3J = 8.3 Hz, 3J = 7.6 Hz), 7.74 (dd, 1H, 3J = 7.6
Hz, 3J = 5.9 Hz), 7.70 (d, 1H, 3J = 8.1 Hz), 7.49 (dd, 1H, 3J = 7.7 Hz,
3J = 6.0 Hz), 7.37 (d, 1H, 3J = 7.7 Hz), 7.16 (d, 1H, 3J = 7.6 Hz), 7.00
3
2
2
(d, 1H, J = 8.1 Hz), 5.46 (d, 2H, J = −15.9 Hz), 5.30 (d, 2H, J =
−15.9 Hz), 5.29 (s, 2H), 3.73 (s, 3H), 1.93 (s, 3H). Minor isomer:
9.14 (d, 1H, 3J = 5.9 Hz), 8.32 (d, 2H, J = 6.1 Hz), 8.15 (t, 2H, J =
7.6 Hz), 8.00 (dd, 1H, 3J = 8.1 Hz, 3J = 7.6 Hz), 7.80 (d, 2H, 3J = 7.6
Hz), 7.75 (dd, 1H, 3J = 7.6 Hz, 3J = 5.9 Hz), 7.62 (dd, 2H, 3J = 7.6 Hz,
3
3
3
2
2
(d, 1H, J = 8.3 Hz), 5.10 (d, 1H, J = −15.1 Hz), 4.97 (d, 1H, J =
−14.4 Hz), 4.85 (m, 2H), 4.58 (d, 1H, 2J = −15.1 Hz), 4.55 (d, 1H, 2J
= −14.4 Hz), 2.15 (s, 3H), 1.56 (s, 9H). Cyclic voltammetry Epc −850
mV vs SCE. IR (cm−1): 1609 for the complex and 1683 for the free
ligand. MS (ESI+, CH3CN, m/z): 478.0 [100%, [Co(III)(6-Piva-
H ) ( T P A ) ( O M e ) ] + ) . A n a l . C a l c d ( F o u n d ) f o r
C24H29ClCoN5O2·1.5H2O: C, 53.29 (53.15); H, 5.96 (5.89); N,
12.95 (12.67).
3J = 6.1 Hz), 7.40 (d, 1H, J = 8.1 Hz), 5.29 (d, 2H, J = −15.8 Hz),
5.25 (s, 2H), 5.20 (d, 2H, 2J = −15.8 Hz), 3.15 (s, 3H), 2.47 (s, 3H).
Cyclic voltammetry: Epc −330 mV vs SCE. MS (ESI+, CH3CN, m/z):
218.5 (100%, [Co(III)(TPA)(B − H)]2+). Anal. Calcd (Found) for
C21H24Cl2CoN5O2·3.7H2O: C, 43.87 (43.85); H, 5.50 (5.29); N,
12.18 (12.14).
3
2
Complex [Co(III)(TPA)(C−H)](Cl)2, 1c. Red powder (143 mg,
63%). 1H NMR (δ, 500 MHz, DMSO), major isomer: 8.90 (d, 1H, 3J
= 5.9 Hz), 8.83 (d, 1H, 3J = 6.5 Hz), 8.62 (d, 2H, 3J = 5.9 Hz), 8.15 (t,
2H, 3J = 7.7 Hz), 8.04 (dd, 1H, 3J = 8.1 Hz, 3J = 7.6 Hz), 7.85 (d, 2H,
3J = 7.7 Hz), 7.70 (dd, 1H, 3J = 7.6 Hz, 3J = 5.9 Hz), 7.62 (dd, 2H, 3J =
7.7 Hz, 3J = 5.9 Hz), 7.55 (dd, 1H, 3J = 8.4 Hz, 3J = 7.6 Hz), 7.43 (d,
1H, 3J = 8.1 Hz), 7.03 (dd, 1H, 3J = 7.6 Hz, 3J = 6.5 Hz), 6.90 (d, 1H,
3J = 8.4 Hz), 5.55 (d, 2H, 2J = −16.5 Hz), 5.37 (s, 2H), 5.32 (d, 2H, 2J
= −16.5 Hz). Minor isomer: 9.19 (d, 1H, 3J = 5.7 Hz), 8.41 (d, 2H, 3J
Complex [Co(III)(CysOMe-H)3], 5. To a suspension of sodium
(tris-carbonato)cobaltate(III) trihydrate24 (500 mg, 1.40 mmol) in
deionized water (10 mL) was added at rt under argon L-cysteine
methyl ester hydrochloride (704 mg, 4.10 mmol). The mixture was
stirred for 20 min, and the green precipitate was filtered and washed
with distilled water, methanol, and then diethyl ether. The green
powder thus obtained was dried under reduced pressure to give
complex 5 as a blue-green powder (320 mg, 40%). UV−vis (Tris
buffer with 10% DMSO, pH = 7.5): λ (ε, M−1 cm−1) 586 (164), 450
( 2 7 0 ) , 2 7 9 ( 1 6 6 0 0 ) . A n a l . C a l c d ( F o u n d ) f o r
C12H24CoN3O6S3·1.9NaCl: C, 25.18 (25.25); H, 4.23 (4.41); N,
7.34 (7.48).
3
3
= 5.7 Hz), 8.15 (t, 2H, J = 7.7 Hz), 8.12 (d, 1H, J = 6.4 Hz), 8.04
(dd, 1H, 3J = 8.1 Hz, 3J = 7.6 Hz), 7.85 (d, 2H, 3J = 7.7 Hz), 7.78 (m,
1H), 7.73 (m, 1H), 7.58 (m, 2H), 7.54 (m, 1H), 7.43 (d, 1H, 3J = 8.1
3
3
2
Hz), 6.82 (dd, 1H, J = 7.4 Hz, J = 6.4 Hz), 5.43 (d, 2H, J = −16.4
Complex [Co(III)(CysOMe-H)(TPA)](Cl)0.75(BPh4)1.25, 1d. To a
purple solution of complex [Co(TPA)(Cl)2].(ClO4)25 (150 mg, 0.29
mmol) in degassed methanol (20 mL) under argon were successively
added solid KOH (38 mg, 0.58 mmol) and L-cysteine methyl ester
hydrochloride (50 mg, 0.29 mmol). After stirring for 30 min, the
brown-green solution was concentrated and the residue purified over
Sephadex LH-20 using methanol as eluent. The first brown band was
collected, and shown to contain the complex [Co(III)(CysOMe-
H)(TPA)](Cl)1.5(ClO4)0.5 (1d) by elemental analysis (Anal. Calcd
(Found) for C22H26Cl2CoN5O4S·1.6H2O: C, 42.95 (42.97); H, 4.78
(5.01); N, 11.38 (11.35)). To try to crystallize this complex, we
exchanged the anion by addition of a slight excess of NaBPh4, leading
to complex 1d(BPh4), (93 mg, 35%). 1H NMR (δ, 500 MHz,
DMSO): 9.19 (d, 1H, 3J = 6.1 Hz), 9.01 (d, 1H, 3J = 6.0 Hz), 8.89 (d,
2
Hz), 5.36 (s, 2H), 5.30 (d, 2H, J = −16.4 Hz). Cyclic voltammetry
Epc: −190 mV vs SCE. MS (ESI+, CH3CN, m/z): 229.5 (100%,
[Co(III)(TPA)(C
−
H)]2+). Anal. Calcd (Found) for
C23H22Cl2CoN5O2·3H2O: C, 47.28 (47.23); H, 4.83 (4.74); N,
11.99 (11.97).
Complex [Co(III)((4-Cl2)TPA)(A-2H)](Cl), 2a. Green powder (104
mg, 51%). Cyclic voltammetry Epc: −960 mV vs SCE. MS (ESI+,
CH3CN, m/z): 245.4 (100%, [Co(III)((4-Cl2)TPA)(A − H)]2+),
489.9 (60%, [Co(III)((4-Cl2)TPA)(A − 2H)]+). Anal. Calcd (Found)
for C20H19Cl3CoN5O2·3.2H2O: C, 41.11 (41.27); H, 4.38 (4.56); N,
11.99 (12.28).
Complex [Co(III)(6-Piva)TPA)(A-2H)]Cl, 3a. Green powder (78.5
1
3
mg, 39%). H NMR (δ, 500 MHz, CD3OD): 9.30 (d, 1H, J = 6.1
Hz), 8.56 (d, 1H, 3J = 6.0 Hz), 8.18 (d, 1H, 3J = 8.4 Hz), 7.90 (dd, 1H,
3J = 7.9 Hz, 3J = 7.2 Hz), 7.85 (dd, 1H, 3J = 8.0 Hz, 3J = 6.9 Hz), 7.78
(dd, 1H, 3J = 8.4 Hz, 3J = 7.4 Hz), 7.55 (dd, 1H, 3J = 6.9 Hz, 3J = 6.1
Hz), 7.51 (d, 1H, 3J = 7.9 Hz), 7.43 (dd, 1H, 3J = 7.2 Hz, 3J = 6.0 Hz),
3
3
3
1H, J = 6.3 Hz), 8.00 (t, 1H, J = 7.4 Hz), 7.98 (t, 1H, J = 7.4 Hz),
7.82 (dd, 1H, 3J = 8.4 Hz, 3J = 7.5 Hz), 7.72 (d, 2H, 3J = 7.4 Hz), 7.52
(dd, 1H, 3J = 7.4 Hz, 3J = 6.1 Hz), 7.49 (dd, 1H, 3J = 7.5 Hz, 3J = 6.3
Hz), 7.45 (dd, 1H, 3J = 7.4 Hz, 3J = 6.0 Hz), 7.27 (d, 1H, 3J = 8.4 Hz),
9355
dx.doi.org/10.1021/ic301090t | Inorg. Chem. 2012, 51, 9350−9356

Inorganic Chemistry
Article
3
3
7.18 (m, 10H), 6.91 (t, 10H, J = 7.4 Hz), 6.78 (t, 5H, J = 7.4 Hz),
5.93 (d, 1H, 2J = −16.4 Hz), 5.70 (d, 1H, 2J = −16.4 Hz), 5.39 (d, 1H,
ACKNOWLEDGMENTS
■
ANR-2010 BLAN METABACT and CNRS as well as Paris
Descartes University are gratefully acknowledged. M.A.
received a PhD fellowship from French Research Ministry
and A.A. a postdoctoral fellowship from ANR.
2
2J = −16.4 Hz), 5.38 (d, 1H, J = −16.4 Hz), 5.34 (m, 1H), 5.32 (m,
3
3
3
2H), 5.18 (m, 1H), 3.49 (ddd, 1H, J = 8.9 Hz, J = 6.4 Hz, J = 4.6
Hz), 3.42 (s, 3H), 2.63 (dd, 1H, 3J = 6.4 Hz, 2J = −12.4 Hz), 2.57 (dd,
1H, 3J = 4.6 Hz, 2J = −12.4 Hz). UV−vis (Tris buffer, pH = 7.5) λ(ε,
M−1 cm−1) 656 (sh), 451 (sh), 310 (5430), and 257 (8900). Cyclic
voltammetry E1/2 −270 mV vs SCE (ΔE = 170 mV). MS (ESI+,
CH3CN, m/z): 482.0 [100%, {[Co(III)(TPA)(CysOMe-H)] − H}+.
Anal. Calcd (Found) for C52H51B1.25Cl0.75CoN5O2S·0.2H2O: C, 68.43
(68.51); H, 5.68 (5.55); N, 7.67 (7.93).
Structural Data for Complex [Co((6-Piva)TPA)(A-2H)](BPh4),
3a(BPh4). Data collection was performed with monochromated Mo
Kα radiation (λ = 0.710 73 Å) on a Nonius Kappa detector at 293 K,
by using the HKL package for data collection and reduction.26 The
REFERENCES
■
(1) Graf, N.; Lippard, S. J. Adv. Drug Delivery Rev. 2012,
DOI: 10.1016/j.addr.2012.01.007.
(2) Sava, G.; Bergamo, A.; Dyson, P. J. Dalton Trans. 2011, 40,
9069−9075.
(3) (a) Kim, B. J.; Hambley, T. W.; Bryce, N. S. Chem. Sci. 2011, 2,
2135−2142. (b) Hambley, T. W. Dalton Trans. 2007, 4929−4937.
(4) (a) Chang, J. Y.-C.; Stevenson, R. J.; Lu, G.-L.; Brothers, P. J.;
Clark, G. R.; Denny, N. A.; Ware, D. C. Dalton Trans. 2010, 39,
11535−11550. (b) Ware, D. C.; Palmer, B. D.; Wilson, W. R.; Denny,
W. A. J. Med. Chem. 1993, 36, 1839−1846.
(5) Failes, T. W.; Culliname, C.; Diakos, C. I.; Yamamoto, N.; Lyons,
J. G.; Hambley, T. W. Chem.Eur. J. 2007, 13, 2974−2982.
(6) Hambley, T. W. Science 2007, 318, 1392−1393.
structure was solved by direct methods in the space group P1 using
̅
SHELXS-9727 and refined by least-squares methods on F2 using
SHELXL-97.27 Non-hydrogen atoms were refined with anisotropic
thermal parameters. Hydrogen atoms were placed in their geometri-
cally generated positions and were allowed to ride on their parent
atom with an isotropic thermal parameter equal to 1.2 times those of
the attached atoms. In the final cycles of refinement, the peak pattern
of electron density suggested that part of solvent was highly
disordered; attempts to model this disorder were unsuccessful. In
the final cycles of refinement, the contribution to electron density
corresponding to the disordered solvent was removed from the
observed data using the SQUEEZE option in PLATON.28 The
resulting data seriously improved the precision of the geometric
parameters for the remaining structure. Crystal data and structure
refinements are listed in Table S4.
(7) Jones, D. P.; Go, Y.-M.; Anderson, C. L.; Ziegler, T. R.; Kinkade,
J. M., Jr.; Kirlin, W. G. FASEB J. 2004, 18, 1256−1248.
(8) Yamamoto, N.; Danos, S.; Bonnitcha, P. D.; Failes, T. W.; New,
E. J.; Hambley, T. W. JBIC, J. Biol. Inorg. Chem. 2008, 13, 861−871.
(9) Ware, D. C.; Siim, B. G.; Robinson, K. G.; Denny, W. A.;
Brothers, P. J.; Clark, G. R. Inorg. Chem. 1991, 30, 3750−3757.
(10) Liou, B.; Yang, K. Mini-Rev. Med. Chem. 2003, 3, 609−620.
(11) Canary, J. W.; Wang, Y.; Roy, R., Jr.; Que, L., Jr.; Miyake, H.
Inorg. Synth. 1998, 32, 70−75.
(12) Higuchi, M.; Hitomi, Y.; Mimami, H.; Tamaka, T.; Funabiki, T.
Inorg. Chem. 2005, 44, 8810−8821.
UV−Vis and NMR Studies. Reactions between 1a and L-cysteine
methyl ester hydrochloride were carried out in degassed non-
deuterated Tris buffer solution (pH = 7.5). Typically, for UV studies
800 μL of a 0.25 mM or 2.5 mM solution of complex 1a and 200 μL of
a 10−2 M or 10−1 M solution of L-cysteine methyl ester hydrochloride
were mixed in a UV cuvette. For NMR studies, 400 μL of a 0.25 mM
(13) Szajna, E.; Makowska-Grzyska, M. M.; Wasden, C.; Arif, A. M.;
Berreau, L. M. Inorg. Chem. 2005, 44, 7595−7605.
(14) Kloubert, T.; Schulz, M.; Gorls, H.; Friedrich, M.; Vesterhausen,
̈
M. Z. Naturforsch. 2009, 64b, 784−792.
(15) Rudzka, K.; Arif, A. M.; Berreau, L. M. J. Am. Chem. Soc. 2006,
128, 17018−17023.
or 2.5 mM solution of complex 1a and 100 μL of a 10−2 M or 10−1
M
(16) Rudzka, K.; Arif, A. M.; Berreau, L. M. Inorg. Chem. 2005, 44,
7234−7342.
solution of L-cysteine methyl ester hydrochloride were mixed in an
NMR tube. The green precipitate resulting from the reaction under
concentrated conditions (2.5 mM) was filtered and washed with water,
methanol, and diethyl ether before being dried under vacuum.
(17) Failes, T. W.; Hambley, T. W. Dalton Trans. 2006, 1895−1901.
(18) Rudzka, K.; Makowska-Grzyska, M. M.; Szajna, E.; Arif, A. M.;
Berreau, L. M. Chem. Commun. 2005, 489−491.
(19) Wane, A.; Thallaj, N. K.; Mandon, D. Chem.Eur. J. 2009, 15,
10593−10602.
ASSOCIATED CONTENT
■
(20) Galardon, E.; Selkti, M.; Roussel, P.; Tomas, A.; Artaud, I.
Dalton Trans. 2008, 6415−6420.
(21) Bonnitcha, P. D.; Kim, B. J.; Hocking, R.; Clegg, J. K.; Turner,
P.; Nevillea, S. M.; Hambley, T. W. Dalton Trans. 2012,
DOI: 10.1039/C2DT30727H.
S
* Supporting Information
Figure S1, 1H NMR spectra of complexes 1a−c, 3a, 3′, and 1d;
Figure S2, cyclic voltammograms in DMF of complexes 1a−c,
2a, 3a, and 4a; Figures S3 and S4, time dependent conversion
1
(22) Kita, M.; Yamamari, K.; Shimura, Y. Bull. Chem. Soc. Jpn. 1982,
of 1a into 1d at pH 6.5 and 8.0, respectively; Figure S5, H
NMR spectra, at 0.2 mM in 200 mM Tris buffer pH 7.5, of the
reaction of complex 1d with L-CysOMe·HCl in excess; Figure
55, 2873−2880.
(23) Fournaud, D.; Pirat, J.-L.; Bigey, F.; Aranaud, A.; Galzy, P. Anal.
Chim. Acta 1997, 353, 359−366.
1
S6, H NMR reaction of 1b (0.2 mM) with CysOMe·HCl (10
(24) Bauer, H. F.; Drinkard, W. C. J. Am. Chem. Soc. 1960, 82, 5031−
equiv) in 200 mM Tris buffer pH 7.5. Table S1, crystal data and
structure refinements of complex 3a(BPh4). X-ray crystallo-
graphic file for complex 3a(BPh4) (CCDC 879870) (CIF file).
This material is available free of charge via the Internet at
http://pubs.acs.org.
5032.
(25) Massoud, S. S.; Jeanbatiste, R. J.; Nguyen, T. M.; Louka, F. R.;
Gallo, A. A. J. Coord. Chem. 2007, 60, 2409−2419.
(26) Otwinowski, Z.; Minor, W. Methods Enzymol. 1997, 276, 307−
326 (Macromolecular Crystallography, Part A).
(27) Sheldrick, G. M. SHELXS-97, Program for Crystal Structure
Solution; University of Gottingen: Gottingen, Germany, 1997.
(28) Spek, A. L. J. Appl. Crystallogr. 2003, 362, 7−13.
̈
̈
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: isabelle.artaud@parisdescartes.fr. Phone: (33) 1 42 86
21 89. Fax: (33) 1 42 86 83 87.
Notes
The authors declare no competing financial interest.
9356
dx.doi.org/10.1021/ic301090t | Inorg. Chem. 2012, 51, 9350−9356