Oxinobactin and sulfoxinobactin, abiotic siderophore analogues to enterobactin involving 8-hydroxyquinoline subunits: Thermodynamic and structural studies

Article abstract of DOI:10.1021/ic301081a

The synthesis of two new iron chelators built on the tris-l-serine trilactone scaffold of enterobactin and bearing a 8-hydroxyquinoline (oxinobactin) or 8-hydroxyquinoline-5-sulfonate (sulfoxinobactin) unit has been described. The X-ray structure of the ferric oxinobactin has been determined, exhibiting a slightly distorted octahedral environment for Fe(III) and a Δ configuration. The Fe(III) chelating properties have been examined by potentiometric and spectrophotometric titrations in methanol-water 80/20% w/w solvent for oxinobactin and in water for sulfoxinobactin. They reveal the extraordinarily complexing ability (pFe^III values) of oxinobactin over the p[H] range 2-9, the pFe value at p[H] 7.4 being 32.8. This was supported by spectrophotometric competition showing that oxinobactin removes Fe(III) from ferric enterobactin at p[H] 7.4. In contrast, the Fe(III) affinity of sulfoxinobactin was largely lower as compared to oxinobactin but similar to that of the ligand O-TRENSOX having a TREN backbone. These results are discussed in relation to the predisposition by the trilactone scaffold of the chelating units. Some comparisons are also made with other quinoline-based ligands and hydroxypyridinonate ligand (hopobactin).

Full text of DOI:10.1021/ic301081a

Article
pubs.acs.org/IC
Oxinobactin and Sulfoxinobactin, Abiotic Siderophore Analogues to
Enterobactin Involving 8‑Hydroxyquinoline Subunits:
Thermodynamic and Structural Studies
Amaury du Moulinet d’Hardemare, Gisele Gellon, Christian Philouze, and Guy Serratrice*
̀
́ ́ ́ ́
Universite Joseph Fourier-Grenoble I, Departement de Chimie Moleculaire, Laboratoire de Chimie Inorganique Redox, UMR-5250,
ICMG FR-2607, CNRS, BP 53, F-38041 Grenoble Cedex, France
S
* Supporting Information
ABSTRACT: The synthesis of two new iron chelators built on the tris-L-serine trilactone
scaffold of enterobactin and bearing a 8-hydroxyquinoline (oxinobactin) or 8-hydroxyquinoline-
5-sulfonate (sulfoxinobactin) unit has been described. The X-ray structure of the ferric
oxinobactin has been determined, exhibiting a slightly distorted octahedral environment for
Fe(III) and a Δ configuration. The Fe(III) chelating properties have been examined by
potentiometric and spectrophotometric titrations in methanol−water 80/20% w/w solvent for
oxinobactin and in water for sulfoxinobactin. They reveal the extraordinarily complexing ability
(pFeIII values) of oxinobactin over the p[H] range 2−9, the pFe value at p[H] 7.4 being 32.8.
This was supported by spectrophotometric competition showing that oxinobactin removes
Fe(III) from ferric enterobactin at p[H] 7.4. In contrast, the Fe(III) affinity of sulfoxinobactin
was largely lower as compared to oxinobactin but similar to that of the ligand O-TRENSOX
having a TREN backbone. These results are discussed in relation to the predisposition by the
trilactone scaffold of the chelating units. Some comparisons are also made with other quinoline-
based ligands and hydroxypyridinonate ligand (hopobactin).
in contrast to the catechol ligands.12 Furthermore, O-
TRENSOX has been found to exhibit a variety of biological
properties such as iron mobilization, cell protection, and
antiproliferative effects.13 This indicates that a ligand and its
ferric complex having a molecular weight of more than 1000 g
mol−1 are able to penetrate the membranes.
INTRODUCTION
■
Iron is vital for organisms; only Lactobacillus and Borelia
burgdoferie bacteria are known to be iron independent.1 Due to
its very low solubility at neutral pH (∼10−10 M, taking into
account the FeIII−hydroxo species2,3), FeIII is poorly bioavail-
able. Thus, in order to improve FeIII uptake, micro-organisms
produce low molecular weight compounds called siderophores.
Siderophores and most of their synthetic models contain three
catechol or hydroxamic acid chelating units, giving stable FeIII
octahedral complexes.4,5 Among the siderophores, enterobactin
(Figure 1a) produced by Escherichia coli and Salmonella
typhimurium.6,7 exhibits the highest affinity.8 It derives from
L-serine, forming a macrocyclic trilactone framework tailoring
an exceptional predisposition of its three catechol subunits for
selective complexation of FeIII. However, efficient synthetic
chelators have been developed involving other chelating
subunits.5 Among those ligands of interest, the 8-hydroxyquino-
line moiety could be considered as a hybrid between catechol
and bispyridine units and possesses remarkable abilities to bind
a variety of metal ions. However, curiously, it is very rarely
represented in the siderophores; only Pseudomonas fluorescens
produces a 8-hydroxyquinoline siderophore, called quinolo-
bactin.9 We developed the water-soluble chelator O-TREN-
SOX, a tris(8-hydroxyquinoline-5-sulfonate) based on the
TREN scaffold, which exhibits strong complexing ability
toward the ferric cation, that is of the same order of magnitude
as the tris-catecholate ligands at neutral pH and is higher in
acidic medium.10,11 O-TRENSOX exhibits a selective affinity
for FeIII, but it is also efficient to chelate FeII, CuII, ZnII, and AlIII
Interest in synthetic siderophores includes their potential
application as therapeutic iron removal agents (iron overload is
one of the most common forms of metal poisoning). Moreover,
there is renewed interest in this field since Fe3+ has been
associated to Cu2+ and Zn2+ in the oxidative stress observed in
amyloid plaques in Alzheimer's disease.14 It has been reported
that metal-chelating agents are able to dissolve Aβ deposits
from the amyloid aggregates by removing metal ions, showing a
therapeutic potential in Alzheimer's disease.15−19 Some 8-
hydroxyquinoline chelators have shown a great potential for
treatment of neurodegenerative diseases.20,21 In particular, the
lipophilic metal chelator clioquinol (5-chloro-7-iodo-8-hydrox-
yquinoline) has been reported to dissolve amyloid plaques of
the brain.22−24 It seemed to us interesting to design a chelator
possessing the most efficient organizing framework, the
trilactone derived from L-serine that also confers the “chiral
recognition area”, but coupled to 8-hydroxyquinoline chelating
subunits (Figure 1b). The aim is to obtain a ligand that is as
strong as enterobactin to chelate FeIII (Figure 1a) (which could
eventually be recognized by its protein receptor) with improved
Received: May 23, 2012
Published: November 6, 2012
© 2012 American Chemical Society
12142
dx.doi.org/10.1021/ic301081a | Inorg. Chem. 2012, 51, 12142−12151

Inorganic Chemistry
Article
Figure 1. (a) Enterobactin (with catechol chelating subunits), (b) oxinobactin (with 8-hydroxyquinoline subunits), and (c) sulfoxinobactin (with 8-
hydroxyquinoline-5-sulfonate subunits).
under argon at −70 °C. After stirring overnight at 10 °C, the mixture
was cooled at −70 °C and treated with methanol (40 mL) in order to
hydrolyze excess of BBr3. After stirring 2 h, solvent was evaporated.
Then methanol (35 mL) and ether (20 mL) were added, and the
deprotected ligand precipitated as a dark yellow solid. The mixture was
filtered and dried under vacuum. The resulting product (hydro-
bromide form) was crystallized from methanol to afford a pale yellow
powder (0.650 g, 0.639 mmol, 40%). MS (ESI) m/z: 775.1 [M − 3
HBr]+. Anal. Calcd for C39H33Br3N6O12·4H2O: C, 42.99; H, 3.79; N,
7.71. Found: C, 42.80; H, 3.76; N, 7.50. 1H NMR (300 MHz, DMSO-
d6/2.49): δ = 4.57 (3H, dd), 4.68 (3H, t), 5.05 (3H, dd), 7.52 (3H, d),
7.86 (3H, dd), 8.12 (3H, d), 8.66 (3H, d), 8.95 (3H, d), 9.52 (3H, d)
ppm. 13C NMR (75 MHz, CD3OD/49.0): δ = 53.37 (CH), 65.14
(CH2), 115.39 (Cq), 118.90 (CH), 125.34 (CH), 128.66 (CH),
130.78 (Cq), 132.87 (Cq), 145.72 (CH), 147.60 (CH), 154.25 (Cq),
169.65 (CO), 170.28 (CO) ppm.
efficiency in acidic medium for targeting other organs (e.g., the
brain is more acidic than other organs), that can form neutral
iron complex (a necessity for crossing the blood−brain barrier),
and that is a good chelator for CuII and ZnII. To date, only one
ligand built on the enterobactin scaffold and employing
hydroxypyridinonate chelates (named “hopobactin”) has been
reported in the literature.25
We presented preliminary results concerning the synthesis of
a novel ligand, named oxinobactin, based on the trilactone
scaffold of the enterobactin and bearing three 8-hydroxyquino-
line groups.26 This paper presents an improved synthesis of
oxinobactin and full characterization of the physicochemical
properties of this ligand, its new sulfonated derivative (named
sulfoxinobactin), and their ferric complexes, including a
thermodynamic and spectrophotometric study in solution and
the X-ray structure of the ferric oxinobactin.
N,N′,N″-((((3S,7S,11S)-2,6,10-Trioxo-1,5,9-trioxacyclodode-
cane-3,7,11-triyl)-tris(azanediyl))-tris(carbonyl))-tris(8-hydrox-
yquinoline-5-sulfonic acid), Sulfoxinobactin 9. To a CH2Cl2
solution (40 mL, cooled in ice) of oxinobactin trihydrobromide 8
(0.150 g, 0.147 mmol) was added chlorosulfonic acid (0.206 g, 1.76
mmol). After stirring overnight under argon at room temperature, the
mixture was evaporated and the resulting residue was treated with
CH3OH and then precipitated with ether (30 mL). The precipitate
was filtered and dried under vacuum. Reprecipitation from CH3OH
and ether gave sulfoxinobactin as a yellow powder (0.091 g, 0.089
mmol, 60%). MS (ESI) m/z: 1012.9 [M]+. Anal. Calcd for
C39H33N6O21S3·3HSO3·2H2O: C, 36.11; H, 3.08; N, 6.48. Found: C,
EXPERIMENTAL SECTION
■
Materials and Methods. All reagents and solvents were
purchased at the highest commercial quality from Acros Organics,
Sigma-Aldrich, or Lancaster and used without further purification
unless otherwise stated. Elemental analyses were performed by the
́ ́
Microanalysis Laboratory of the Departement de Chimie Moleculaire.
Electrospray ionization (ESI) mass spectra were recorded using an
1
Esquire 300 Plus Bruker Daltonis. H and 13C NMR spectra were
recorded using a Bruker Advance 300 spectrometer at 25 °C. Residual
solvent signals were used as internal standard: DMSO-d6 (1H δ 2.49),
CD3OD (13C δ 49.0), and CDCl3 (1H δ 7.24; 13C δ 77). Chemical
shifts in D2O were determined relative to an internal reference of 4,4-
dimethy-4-silapentane-1-sulfonic acid-d6 (DSS) sodium salt for 1H and
13C NMR. For solubility reasons, the ligand oxinobactin and its ferric
1
35.94; H, 3.20; N, 6.34. H NMR (300 MHz, D2O-NaOD/DSS): δ =
3.86 (3H, d), 4.46 (6H, t), 7.52 (3H, q), 8.42 (3H, d), 8.62 (3H, d),
8.70 (3H, d) ppm. 13C NMR (75.4 MHz, D2O-NaOD/DSS): δ =
56.12 (CH), 57.99 (CH), 63.59 (CH2), 111.57 (Cq), 119.75 (Cq),
124.22 (CH), 128.12 (Cq), 129.51 (CH), 134.87 (CH), 145.60 (Cq),
147.82 (CH), 170.32 (CO), 171.92 (Cq), 178.04 (CO) ppm.
complexes were studied in a methanol/water mixture (80/20 w/w).
Ionic strength was adjusted to 0.1 M with NaClO4 (except for
measurements over the p[H] range 0−1), and all measurements were
carried out at 25.0 0.2 °C. Stock solutions of Fe3+ were prepared by
dissolving the appropriate amount of Fe3+ perchlorate·H2O (∼0.01 M)
in 0.1 M aqueous HClO4. The exact concentration of Fe3+ was
determined spectrophotometrically using a molar extinction coefficient
of 4160 M−1 cm−1 at 240 nm in 0.1 M aqueous HClO4.27 Solutions
were prepared with boiled deionized water, which was deoxygenated
and flushed continuously with Ar (U grade) to exclude CO2 and O2.
Synthesis. Oxinobactin was prepared with some modifications of
our published procedures.26 Full procedures and characterization of
the precursors of oxinobactin are described in the Supporting
Information. Enterobactin was prepared similarly, according to the
procedures described in the ref 25 (see characterization in Supporting
Information).
X-ray Structure of Ferric Oxinobactin. Ferric acetylacetonate
and oxinobactin were dissolved in DMF containing 5% water. The
solution rapidly turned dark green and was allowed to stir for 1 h.
Crystals of ferric oxinobactin were grown from diffusion of ether into
the solution. Single crystals of complexes were mounted on a Kappa
CCD Nonius diffractometer equipped with graphite-monochromated
Mo Kα radiation (λ = 0.71073 Å) at 200 K. Reflections were corrected
for Lorentz and polarization effects but not for absorption. Structures
were solved by direct methods and refined using SHELXS-97 and
SHELXL-97 software run under Olex2.28 All non-hydrogen atoms
were refined with anisotropic thermal parameters. Hydrogen atoms
were generated at idealized positions, riding on the carrier atoms, with
isotropic thermal parameters. Pertinent crystallographic data and
refinement details are summarized in Table 2. Complete crystallo-
graphic data are collected in Supporting Information (CIF file format).
Potentiometric Measurements. Titrations were performed using
an automatic titrator system DMS 716 Titrino (Metrohm) with a Ag/
AgCl combined glass electrode (Metrohm, 6.0234.100). The electrode
was filled with 3 M NaCl for measurements in water and with 0.1 M
NaCl in methanol/water (80/20 w/w) for measurements in this
solvent. The combined glass electrode was daily calibrated to read
N,N′,N″-((3S,7S,11S)-2,6,10-Trioxo-1,5,9-trioxacyclodode-
cane-3,7,11-triyl)-tris(8-hydroxyquinoline-7-carboxamide),
Tris-hydrobromide, Oxinobactin 8. Benzyl-protected oxinobactin
7 was synthesized according to the procedure described in the
Supporting Information. To a solution of this product (1.64 g, 1.57
mmol) in dry CH2Cl2 (160 mL) was added BBr3 (7.86 g, 31.3 mmol)
12143
dx.doi.org/10.1021/ic301081a | Inorg. Chem. 2012, 51, 12142−12151

Inorganic Chemistry
Article
a
Scheme 1. Synthesis of the 8-Hydroxyquinoline Fluoride Subunit 4
a
(a) BF3, MeOH, reflux, 97%; (b) BnCl, K2CO3, H3CCN, reflux, 95%; (c) NaOH, MeOH/THF, then pH 4−5, 80%; (d) C3F3N3, DIPEA, CH2Cl2,
0 °C, 75%.
a
Scheme 2. Synthesis of the Trilactone Scaffold and Its Coupling with 4 to Oxinobactin 8
a
(e) 2,2-Dibutyl-1,3,2-dioxastannolane catalyst, toluene, reflux, 65−75%; (f) HCl, EtOH, rt, 95%; (g) 4 + DIPEA, CH2Cl2, 0 °C, 62%; (h) 7 + BBr3/
MeOH, −70 °C, 40%.
p[H] by titrating HClO4 (0.005 M) with CO2-free NaOH (0.01 M
standardized against hydrogen phtalate) under a stream of argon. For
studies in methanol/water (80/20 w/w), the calibration procedure was
analogous to that in aqueous medium. The stream of argon was
presaturated with methanol vapor. The GLEE program29 was applied
for glass electrode calibration and check of carbonate in the NaOH
solution. The ionic product of H2O was fixed to 13.78 in water and
14.230 in methanol/water mixture. The potentiometric data for the
titration of the ligands (0.0005 M, p[H] range 2.7−10.5, 150 points)
were refined with the HYPERQUAD program, which uses nonlinear
least-squares fitting methods.31 The standard deviation, σ, was in the
range 3−4. The stability of the serine ester backbone of the ligands was
checked by UV−vis spectrophotometry and 1H NMR in basic medium
(p[H] = 9). No change of the spectra was observed for 2 h. It was
deduced that the ligands were stable during titration in the p[H] range
7−9, indicating that the rate of serine ester hydrolysis was slow enough
with respect to duration of measurements.
Spectrophotometric Measurements. Titrations of the ligands
(ca. 10−5 M) were carried out in a thermostatted cell by adding known
volumes of NaOH. Emf data (converted in p[H] from calibration with
GLEE program) were recorded with the DMS 716 Titrino. Absorption
spectra were recorded using an optical fiber dip probe (Hellma,
661.602-UV) connected to a Varian Cary 50 UV−vis spectropho-
tometer.
addition of HClO4 and NaClO4 to maintain the ionic strength at 0.1
M. Samples were allowed to equilibrate for 72 h at 25 °C.
RESULTS AND DISCUSSION
■
Syntheses. Synthesis of the oxinobactin is realized by
following the general plan adopted during our previous work to
which we brought some modification and improvement,
leading to a more reliable procedure.26 It is a convergent
synthesis in which the first part consists in preparing the
fluoride of the 8-hydroxyquinoline acid 4 by protecting
beforehand in two steps the phenol function by a benzyl
group (Scheme 1).
The fluoration step is best done with cyanuric fluoride, which
gives better yield and a better purity than the more popular
DAST fluorinating reagent. It must be pointed out that the
fluoride 4 is stable enough to be purified by the conventional
chromatography procedure, in contrast to its chloride
homologue which decomposes readily. The second part of
this convergent synthesis aims at preparation of the trilactone
of L-serine 6, a key compound in the synthesis, and its coupling
with the fluoride 4 (Scheme 2).
Titrations of an equimolar solution in ligand and Fe3+ (ca.10−4 M)
were done in two sets of measurements since complexation started in
very acidic medium: (i) in the p[H] range 0−2, batch solutions were
prepared using standardized HClO4 to adjust p[H]; spectra were
recorded 30 minutes after mixing the compounds to take into account
the kinetics of complex formation and prevent possible hydrolysis of
the trilactone scaffold in a very acidic medium (the stability of the
ligands was checked by UV−vis spectrophotometry at p[H] = 0); (ii)
in the p[H] range 2−10, the same procedure as the titration of the
ligands was carried out. After each addition of NaOH it was ensured
that complete equilibration was attained (ca. 10 minutes) before
measurement of the Emf and recording the spectrum. Absorption data
were processed with the SPECFIT/32 Global Analysis System
(Spectrum Software Associates), adjusting the equilibrium constants
and the corresponding molar extinction coefficient of the species at
equilibrium.32−34 The standard deviation σ was in the range 0.008−
0.010. It should be noted that ionic strength was not maintained at 0.1
M in the p[H] range 0−1. Thus, the stability constants are less
accurate.
Starting from methyl L-serinate from which the nitrogen is
beforehand protected by the trityl group (methyl L-trityl
serinate is also commercially available), the key reaction of
macrolactonization is realized with the 2,2-dibutyl-1,3,2-
dioxastannolane taken as the best organostannic catalyst tested
(75% yield). Later, after deprotection of 5 by hydrochloric acid,
the macrolactone 6 is coupled with the fluoride 4, affording the
benzylic-protected oxinobactin 7 with a satisfactory yield of
62%. Removal of benzyl groups by hydrogenation catalyzed by
palladium on charcoal gives over-reduction of the quinoline
nucleus, and only 16% of oxinobactin 8 could be recovered
after a very tedious step of chromatography. Improved results
were obtained with BBr3 as the debenzylating agent, which
raised the yield to 40%. The purification was also simplified
because the tris-hydrobromide so obtained could be easily
purified by crystallization from methanol instead the more
difficult step of chromatography. The sulfoxinobactin 9 derives
from oxinobactin 8 by direct sulfonation. The usual sulfonating
oleum (30% SO3 in H2SO4) proved to be unsatisfactory; only
decomposition products ensued. In contrast, good results were
obtained with chlorosulfonic acid in 12 mol excess in dry
The formation constant of the ferric sulfoxinobactin was also
determined by competition against Na2H2EDTA over the p[H] range
1.9−2.8 (4 solutions). Typical solutions were 10−4 M in ferric ion,
ligand, and Na2H2EDTA. The p[H] of the solutions was adjusted by
12144
dx.doi.org/10.1021/ic301081a | Inorg. Chem. 2012, 51, 12142−12151

Inorganic Chemistry
Article
dichloromethane: in this condition the yield is 62% (Scheme
3).
a
Scheme 3. Sulfonation of Oxinobactin
a
(i) HSO3Cl, CH2Cl2, 62%.
Protonation of the Ligands. Since the metal ions
compete with the protons for the coordination sites, it is
necessary to determine the acid−base properties of the ligands.
3+
The ligands oxinobactin and sulfoxinobactin, denoted L1H6
and L2H6, respectively, in the fully protonated form, possess six
potential protonation sites: three phenolate oxygen atoms and
three pyridinyl nitrogen atoms (the sulfonate groups in L2H6
are assumed to be deprotonated under the conditions of
titration). The protonation equilibria, defined by eq 1
Kn
LHn−1 + H ⇌ LHn(charges are omitted)
(1)
were studied by potentiometric titration in the p[H] range 2.7−
10.5 (i) for L1H6 in methanol/water (80/20 w/w, I = 0.1 M
3+
NaClO4) owing to its poor solubility in water (Figure 2) and
Figure 3. UV−vis absorption spectra of oxinobactin (2.0 × 10−5 M) as
a function of p[H]: (a) p[H] = 1.75, 1.82, 1.88, 1.92, 1.98, 2.03, 2.08,
2.16, 2.24, 2.44, 2.58, 2.75, 2.98, 3.11, 3.40, 3.65, 3.91, 4.16, 4.90, and
5.74; (b) p[H] = 6.71, 6.84, 7.00, 7.12, 7.31, 7.58, 7.72, 7.82, 7.90,
8.03, 8.17, 8.30, 8.40, 8.54, 8.71, 8.89, and 9.29. Solvent: CH3OH/H2O
(80/20 w/w), I = 0.1 M (NaClO4), T = 25.0(2)°C.
(Figure 3b, spectra exhibit isosbestic points at 240, 264, and
286 nm). Analysis of the data by the SPECFIT program32−34
gave a model with only two absorbing species involving one
proton exchange in each p[H] range. Calculation yields two
values of protonation constants: 2.28(2), which were attributed
to log K6, and 7.64(2). This last value is in good agreement
with log K2 = 7.78(2) determined by potentiometric titration.
For L2H6, potentiometric titration allowed us to determine
only four protonation constants log K1 = 7.84(1), log K2 =
7.03(1), log K3 = 6.18(2), and log K4 = 2.78(4). In complement
with this study, we undertook a spectrophotometric titration in
the p[H] range 1−8. Two buffer regions were observed, one in
the p[H] range from 1 to 2.8 (Figure S2a, Supporting
Information, spectra exhibit isosbestic points at 251 and 273
nm) and the other from p[H] 5 to 7 (Figure S2b, Supporting
Information, spectra exhibit isosbestic points at 242, 256, and
277 nm). Treatment of the absorbance data by the SPECFIT
program indicated the presence of only two absorbing species
corresponding to one proton exchange in each p[H] range
1.0−2.8 and 4.7−7.8 and gave values of 1.42(3) and 5.99(3)
attributed to log K6 and log K3, respectively. This last value
favorably agrees with log K3 = 6.18 determined by
potentiometric titration. In addition, the value of log K5 was
estimated to be 2 for calculation of the Fe(III) stability
constants.
Figure 2. Potentiometric titration curves: (a) 4.5 × 10−4
M
oxinobactin, (b) [oxinobactin]/[Fe3+] = 1/1, 4.5 × 10−4 M. Solvent:
CH3OH/H2O (80/20 w/w), I = 0.1 M (NaClO4), T = 25.0(2) °C.
(ii) for L2H6 in water (I = 0.1 M NaClO4) (Figure S1,
3+
Supporting Information). Data were analyzed by the HYPER-
QUAD program.31 An additional study using a UV−vis
spectrophotometric titration was performed using the SPEC-
FIT program32−34 for fitting analysis of the absorbance data.
For L1H63+, the potentiometric study revealed five protonation
constants (the number in parentheses corresponds to the
standard deviation): log K1 = 8.51(3), log K2 = 7.78(2), log K3
= 6.90(4), log K4 = 4.02(7), and log K5 = 2.75(9).
Spectrophotometric titration was carried out in the p[H]
range 1.7−9.3. Two buffer regions were observed, one in the
p[H] range 1.7−5.7 (Figure 3a, spectra exhibit isosbestic points
at 258 and 286 nm) and the other in the p[H] range 6.7−9.3
12145
dx.doi.org/10.1021/ic301081a | Inorg. Chem. 2012, 51, 12142−12151

Inorganic Chemistry
Article
a
Table 1. Protonation Constants of Oxinobactin (L1H63+), Sulfoxinobactin (L2H6), and Ligands Containing 8-
Hydroxyquinoline Subunits
ligand
log K1
log K2
log K3
average log K1−3
log K4
log K5
log K6
average log K4−6
b
oxinobactin
8.51(3)
7.84(1)
10.38
8.62
7.78(2)
7.03(1)
8.40
6.90(4)
6.18(2)
6.80
7.73
7.02
8.53
8.08
4.02(7)
2.78(4)
4.13
2.75(9)
∼2
2.89
2.28(2)
1.42(3)
2.43
3.02
2.07
3.15
2.46
c
sulfoxinobactin
cd
,
COX2000
ce
,
O-TRENSOX
n-Busox (QS4)
8.18
7.44
3.01
2.55
1.83
cf
,
6.80
2.80
cg
,
QS10
6.77
2.60
a
b
c
Numbers in parentheses represent the standard deviation in the last significant digit. In methanol/water (80/20 w/w, I = 0.1 M NaClO4). In
d
e
f
g
water (I = 0.1 M NaClO4). Reference 35. Reference 11. Reference 36, 8-hydroxy-7-[(butylamino)carbonyl]quinoline-5-sulfonic acid. Reference
37, 8-hydroxy-7-[(decylamino)carbonyl]quinoline-5-sulfonic acid.
Scheme 4. Structures of n-Busox, QS10, O-TRENSOX, COX200, and COX2000
All constants are reported in Table 1 together with those of
ligands containing 8-hydroxyquinoline (COX2000,35 based on
a C-pivot scaffold grafted with a polyoxyethylenic chain) or 8-
hydroxyquinoline-5-sulfonate units (O-TRENSOX,11 based on
the backbone of tris(2-aminoethyl)amine, n-Busox,36 QS1037);
see structures in Scheme 4.
The constants K1−K3 are attributed to protonation of the
phenolate oxygen atoms and constants K4−K6 to the pyridinyl
nitrogen atoms. In spite of the difference of solvents, the log K
of L1H63+ and L2H6 can be compared: the average value for the
log K of the oxygen sites, 7.02 for L2H6, is lower than the
corresponding value, 7.73 for L1H63+, in relation to the
electron-withdrawing effect of the sulfonate group. The same
at 360 nm (ε = 8.8 × 103 M−1 cm−1). Deprotonation of the
pyridine nitrogen results in hypsochromic shifts (256 nm, ε =
8.7 × 104 M−1 cm−1; 340 nm, ε = 9.4 × 103 M−1 cm−1), while
deprotonation of the phenol group leads to bathochromic and
hyperchromic shifts as expected (274 nm, ε = 9.4 × 104 M−1
cm−1; 350 nm, ε = 2.1 × 104 M−1 cm−1). The changes of the
band at higher energies of sulfoxinobactin (267 nm, ε = 8.0 ×
104 M−1 cm−1) are different from that of oxinobactin:
bathochromic shift upon deprotonation of pyridine nitrogen
(280 nm, ε = 5.8 × 104 M−1 cm−1) and hypsochromic shift
upon deprotonation of the phenol group (268 nm, ε = 7.4 ×
104 M−1 cm−1) as it has been already observed by the
sulfonated ligands n-Busox36 and QS10
.
37
trend is observed for protonation of the nitrogen sites: the
X-ray Structure of Ferric Oxinobactin. Single crystals of
the FeIII−oxinobactin complex [Fe(C39H27N6O12)] were
obtained by diffusion of diethyl ether into a solution of the
complex in DMF containing 5% water. The complex crystallizes
in the R3 rhomboedral space group with Z = 3 (Table 2). The
structure reveals a neutral mono-FeIII complex having C3
symmetry and a slightly distorted octahedral tris(8-hydrox-
yquinolinate) environment for FeIII (Figure 4). Owing to the 3-
fold symmetry axis, only one-third of a molecule unit is
necessary to be crystallographically described. Each 8-
hydroxyquinolinate subunit forms a 5-membered chelate ring
with iron, leading to the facial isomer (Figure 4b). As we can
judge it by observing the crystallographic structure of the
complex, the trilactone scaffold still presents the S configuration
for carbons 3, 7, and 11, brought by the L-serine (no
racemization occurred during synthesis). As the configuration
of the ligand imposes that of the metallic center, we indeed
observe a right-handed configuration (Δ chirality) as is the case
for the ferric enterobactin. Selected bond lengths and angles are
reported in Table 3. The five-membered chelate ring angle
O1−Fe−N1 is 79.1°, whereas the other cis angles are 98.4° for
O1−Fe−N1ii, 92.4° for both O1−Fe−O1ii and O1−Fe−O1iii,
3+
average values are 2.07 and 3.02 for L2H6 and L1H6
,
respectively. The log K ranges for the nitrogen and oxygen
atoms differ from the statistical factor of log 3 (0.48), indicating
that there is some cooperativity between the three arms of the
ligands. The average values of the log K4−6 and log K1−3 for
L2H6, 7.02 and 2.07, can be compared to the log K values of the
monochelate analogues n-Busox (6.80 and 2.80)36 and QS10
(6.77 and 2.60)37 (Table 1). The log K values can also be
compared to those of tripodal ligands containing three 8-
hydroxyquinoline chelating subunits (Table 1). The log K
values of oxinobactin L1H63+ are similar to those determined in
aqueous solution for the tris(chelate) analogue COX2000
(Scheme 4), except for log K1, which is abnormally high for
COX2000.35 The log K values of sulfoxinobactin L2H6 are
lower than those of O-TRENSOX (average values of log K1−K3
= 8.08 and log K4−K6 = 2.46).11 The higher values for O-
TRENSOX could be due to an extended network of H bonding
involving the nitrogen atom of the tripodal backbone.
The electronic spectrum of oxinobactin in its protonated
form is characterized by a strong absorption band at higher
energies (267 nm, ε = 1.3 × 105 M−1 cm−1) and a broad band
12146
dx.doi.org/10.1021/ic301081a | Inorg. Chem. 2012, 51, 12142−12151

Inorganic Chemistry
Article
Table 2. Crystallographic Data for Ferric Oxinobactin
Table 3. Selected Bond Lengths (Angstroms) and Angles
(degrees) for Fe−Oxinobactin
formula
[Fe(C39H27N6O12)]·dmf
fw [g·mol−1
]
827.54 + dmf
blue prism
0.35 × 0.21 × 0.18
trigonal
R3
Fe−O1
Fe−N1
N2···O1
1.949(2)
2.166(3)
2.69
O1−Fe−N1 chelate
O1−Fe−N1ii cis
O1−Fe−O1
79.1
98.4
color/morphology
cryst size
cryst syst
space group
a [Å]
92.4
H10···O1
N2−H10−O1
2.04
N1−Fe−N1
O1−Fe−N1iii apical
91.7
129.7
166.5
18.650(1)
18.650(1)
12.317(2)
90
b [Å]
The oxygen and nitrogen atoms of the amide group are
slightly displaced from the quinoline plane: the dihedral angle
between O2−C10−C7 or N2−C10−C7 and the quinoline
plane is 166.2° or 11.6°, respectively. The Fe−O and Fe−N
distances are 1.949 and 2.166 Å, respectively. These values are
identical to the average distances 1.948 and 2.164 Å observed
for the Fe−Cox200.35 Interestingly, the structure reveals that
the amide protons interact strongly by hydrogen bonding with
the quinolinol oxygens coordinated to iron, forming three 6-
membered rings. The H10···O1 and N2···O1 distances are 2.04
and 2.69 Å, respectively, and the angle NHO is 129.7°. This
hydrogen bonding is an important feature of the structure and
contributes to enhance the stability of the complex. This
interaction is also observed in FeIII triscatecholate: as an
example, for [Fe(TRENCAM)] the H···O distances vary from
1.80 to 1.85 Å.39
In summary, these results underline the ability of the
enterobactin scaffold to bring chelate units more bulky than
catecholamide, such 8-hydroxyquinoline, leading to ferric
complex with less distorted octahedral geometry. Furthermore,
the structures of the ferric COX200 complex having a C-pivot
tripodal backbone and the ferric oxinobactin complex show
very close coordination geometry (bonds and dihedral angles),
although the ligands have very different structures of their
backbone.
c [Å]
α [deg]
β [deg]
90
γ [deg]
120
unit-cell volume [Å3]
3710.2(8)
1.406
Dx (g·cm−3
T [K]
Z
)
200
3
λ [mm−1
]
0.71073
8841
total reflns
unique reflns (Friedel’s included)
3716
3384 (F2 > 2σ)
obsd reflns
Rint
0.0482
a
R
0.0532
a
R(w)
0.1262
goodness of fit S
1.125
Δ
minρ/Δρmax (e·Å−3
)
−0.56/0.54
a
2
Refinement based on F2 where w = 1/[σ2(Fo ) + 0.0548p2
+
10.9877p] and p = (Fo + 2Fc2)/3 for 2·3THF.
2
91.7° for both N1−Fe−N1ii and N1−Fe−N1iii, and 166.5° for
the apical O1−Fe−N1iii angle. These angles are in the range
observed for Fe−COX200 (ligand with the same structure as
COX2000 but with a shorter polyoxyethylenic chain, see
Scheme 4).34 The trigonal twist angle (60° for an ideal
octahedron) has been calculated to be 43.9°.38 This value is less
distorted from octahedral geometry than the values for
hopobactin (35°) or enterobactin (estimated 33°),25 indicating
there are less steric constraints with the 8-hydroxyquinolinate
chelates.
Stability Constants of the Ferric Complexes. Potentio-
metric titration was carried out for 1:1 solutions in ferric ion
and ligand from p[H] 2. The curves (Figures 2b and S1b,
Supporting Information, for ferric−oxinobactin and ferric−
sulfoxinobactin, respectively) showed an inflection at a = 6,
indicating that the six protons of the ligand have been displaced
Figure 4. ORTEP diagram at 25% probability of the ferric oxinobactin viewed perpendicular to (a) and down (b) the molecular 3-fold axis.
12147
dx.doi.org/10.1021/ic301081a | Inorg. Chem. 2012, 51, 12142−12151

Inorganic Chemistry
Article
by the ferric ion. Furthermore, comparison of the curves for the
ligand and its ferric complex indicates that hydroxo complexes
were formed at p[H] above about 9 for ferric oxinobactin and
about 8 for ferric sulfoxinobactin. Since the ferric complex is
fully formed at p[H] 2, formation equilibria were studied by
spectrophotometric titration of equimolar solutions of ferric ion
and ligand. First, a set of solutions in the p[H] range 0−2 was
studied at variable molar concentrations of standardized
HClO4. Second, a spectrophotometric−p[H] titration was
carried out from p[H] 2 to 9.
No spectral change was observed over the p[H] range 2−9
(Figure S5, Supporting Information) for ferric oxinobactin and
2−6 for ferric sulfoxinobactin (Figure S6, Supporting
Information). For this last complex, a change of the absorption
was observed over the p[H] range 6−9, indicating formation of
hydroxo complexes (Figures S4 and S6, Supporting Informa-
tion). Spectra exhibited isosbestic points at 330 and 420 nm.
Analysis of the absorbance−p[H] data provided formation of
the hydroxo species FeL2OH. The complexation constant is
reported in Table 4.
For the two complexes, an increase of p[H] from 0 to 2 led
to the growth of charge-transfer bands centered at 440 and 590
nm (Figure 5 and Figure S3, Supporting Information). The best
The stability constant β110 of the ferric complex FeL2 was
also determined by spectrophotometric competition experi-
ments with ethylenediaminetetraacetic acid (edta) over the
p[H] range 1.9−2.8 (Figure 6). The existence of isosbestic
Figure 5. UV−vis absorption spectra of Fe3+−oxinobactin at p[H] =
−0.10, −0.01, 0.04, 0.22, 0.46, 0.56, 0.64, 0.94, 1.16, 1.42, 1.62, and
2.05. [Fe3+] = 8.9 × 10−4 M, [oxinobactin] = 9.7 × 10−5 M. Solvent:
CH3OH/H2O (80/20 w/w), I = 0.1 M (NaClO4), T = 25.0(2) °C.
Figure 6. Spectrophotometric competition of ferric sulfoxinobactin in
the presence of EDTA at p[H] = 2.47; [Fe sulfoxinobactin] = [EDTA]
= 8.3 × 10−5 M ; T = 25 °C. Spectra were recorded at t (min) = 0 (a)
and 1, 3, 10, 20, and 180 (b) after mixing solutions of ferric
sulfoxinobactin and EDTA.
refinement of the absorbance−p[H] data with the program
SPECFIT was obtained by considering formation of the
3+
FeLnH3 and FeLn species. The global stability constants β11h
points at 325 and 410 nm suggest that no ternary Fe−EDTA−
L2 complex was formed. The competition equilibrium can be
expressed by eqs 4 and 5, in which Y denotes edta, K1−K3
denote the protonation constants of L2H3, and K1′, K2′, and K3′
denote the protonation constants of YH3.40
of these complexes, which are defined by eqs 2 and 3, are
reported in Table 4. It should be noted that the values of log
Table 4. Equilibrium Constants log β11h and UV−Vis Data of
Ferric Complexes with Oxinobactin (L1) and
Sulfoxinobactin (L2)
2
2
FeL + YH3 ⇌ FeY + L H3
(4)
a
b
λ/nm (ε, 103 M−1 cm−1
)
pFeIII
K = [FeY][LH3]/[FeL][YH3]
complex
[FeL1H3]3+
log β11 h
36.75 (5)
33.61 (13)
30.66 (5)
26.73 (9)
18.68 (2)
32.12
450 (3.8), 585 (2.8)
450 (4.8), 600 (4.3)
430 (4.0), 575 (2.6)
430 (5.8), 580 (4.4)
425 sh, 530 (1.9)
= β110(FeY)K1K2K3/β110(FeL)K1′K2′K3′
(5)
[FeL1]
32.8
27.1
[FeL2H3]
[FeL2]3−
[FeL2OH]4−
The concentration of FeL2 was calculated from the absorbance
at 590 nm, where FeL2 is the only absorbing species. The
concentrations of the other species in eqs 4 and 5 were
calculated from the mass balance equation and p[H] (eqs 6−8)
in which α is the usual Ringbom coefficient.41
c
[Fe(COX2000)]
[Fe(O-TRENSOX)]3−
450 (4.6), 590 (3.9)
29.1
29.5
d
30. 9
443 (5.4), 595 (5.4)
a
Measurements in methanol/water (80/20 by weight) for the Fe−L1
2
[Fe]tot = αFeL[FeL ] + αFeY[FeY]
(6)
system and water solutions for other complexes, I = 0.1 M NaClO4.
b
pFeIII = −log [Fe3+] calculated for [Fe]tot = 10−6 M, [L]tot = 10−5 M,
2
2
2
c
d
[L ]tot = αFeL[FeL ] + αL[L ]
[Y]tot = αFeY[FeY] + αY[Y]
(7)
(8)
and p[H] = 7.4. Reference 35. Reference 11.
From the known formation constants of FeY (log β110 = 25.140
the average formation constant of FeL2 was determined to be
log β110 = 26.5 0.2. This value is in good agreement with the
value 26.7, determined from spectrophotometric titration.
The two bands exhibited by the visible spectra of the FeL1
complex at 450 (ε = 4800 M−1 cm−1) and 595 nm (ε = 4300
β113 are less accurate than the other constants because the ionic
strength varied over the p[H] range 0−1 (charges are omitted).
n
n
Fe3+ + L + hH+ ⇌ FeLHh
(2)
β11h = [FeLHh]/[Fe3+][L ][H+]h
n
n
(3)
12148
dx.doi.org/10.1021/ic301081a | Inorg. Chem. 2012, 51, 12142−12151

Inorganic Chemistry
Article
M−1 cm−1) and of the FeL2 complex at 430 (ε = 5800 M−1
cm−1) and 580 nm (ε = 4400 M−1 cm−1) are characteristic of
the tris(8-hydroxyquinolinate) coordination as observed with
the FeL species of the ferric COX200035 and O-TRENSOX11
complexes. The spectral characteristics of the FeL1H3
complexes, two bands at 450 (ε = 3800 M−1 cm−1) and 585
nm (ε = 2800 M−1 cm−1), and of FeL2H3, two bands at 430 (ε
= 4000 M−1 cm−1) and 575 nm (ε = 2600 M−1 cm−1), suggest
coordination of two 8-hydroxyquinoline moieties having one
protonated hydroxyl group and one free protonated 8-
hydroxyquinoline ligand. It should be noted that no salicylate
coordination (through the oxygen atoms of the hydroxyl and
carbonyl groups) was observed in acidic medium, in contrast to
Fe−COX2000 and Fe−O-TRENSOX. Salicylate coordination
is characterized by a strong band at 440−445 nm and a
shoulder near 550 nm.11 Salicylate coordination has been also
evidenced for ferric enterobactin. In contrast, no salicylate
coordination has been observed for ferric hopobactin.25 The
spectral parameters of the FeL2OH species suggest coordina-
tion with one 8-hydroxyquinoline unit. The remaining
coordination sites are thus occupied by four water molecules
to complete the octahedral geometry, one of the water
molecules being deprotonated to give a hydroxyl anion. A
value of 8.05 is calculated from β110 and β11−1 for
pKa(FeL2OH/FeL2). This value is too high to be attributed
to hydroxyl deprotonation of a free arm of the ligand,
solutions showed no change. This experiment confirms that
sulfoxinobactin is less efficient than enterobactin to chelate
FeIII, in agreement with the pFe values.
Since the ligands are weak acids, proton competition occurs
depending on their protonation and the pH. The pFeIII value
(pFeIII = −log [Fe3+]) is thus a better measure of the relative
efficiency of the ligands under given conditions of pH, [Fe]tot,
and [L]tot. The pFeIII values have been calculated for [L]tot
=
10−5 M, [Fe]tot = 10−6 M, and p[H] over the range 2−9. In
spite of the difference of solvents, one can compare the plots of
pFeIII vs p[H] for oxinobactin (Figure 8) and COX2000 (pFe
Figure 8. Plot of pFe versus p[H]. pFe = −log [Fe3+] calculated for
Spectrophotometric Competition with Enterobactin.
Since no data are available for ferric enterobactin equilibria in
methanol−water solvent, it was interesting to compare the
efficiency of oxinobactin relative to enterobactin toward FeIII in
this solvent at physiological pH. For this aim, we carried out a
spectrophotometric competition in MOPS methanol−water
buffer (p[H] = 7.4). Spectra are depicted in Figure 7. The
[Fe]tot = 10−6 M, [L]tot = 10−5 M: ( ) oxinobactin and ( )
COX2000.
■
●
values were calculated from complexation constants determined
in aqueous solution)35 to gain insight into the effect of the
scaffold structure on complex stability. An increase of about 4
log units is observed over the pH range 2−9 on going from
COX2000 to oxinobactin, showing that the trilactone scaffold
induces a large increase of the efficiency of the ligand for metal
binding. A similar comparison between the “TREN-capped”
ligands and the derivatives of the enterobactin reveals an
increase varying from 8.7 (pH = 4) to 5.5 (pH = 9) log units in
the pFe values between TRENCAM42 and enterobactin8
(catecholate binding units) (Figure S7, Supporting Informa-
tion), whereas the difference between TRENHOPO and
hopobactin (pyridinonate binding units) (Figure S7, Support-
ing Information) is about 4 log units over the pH range 2−6
but only 0.7 at pH 7.4.25 It should be noticed that oxinobactin
is the most efficient Fe(III) chelator in acidic and neutral pH. It
has been claimed that the very high affinity for ferric ion of
enterobactin is largely due to the predisposition of the
catecholamide units induced by the trilactone ring, thus
allowing one to reduce the energy to bind the cation. The
enhanced affinity of oxinobactin for ferric ion with respect to
COX2000 and the very similar coordination geometry around
the iron for these two ferric complexes provide support to the
ability of the 8-hydroxyquinoline moieties to be predisposed
with the trilactone scaffold in spite of their bigger bulkiness
than that of the catechol groups. This is also consistent with the
structure of the ferric oxinobactin, which reveals a relatively
small degree of steric strain as indicated by the twist angle of
43.9°. In contrast, no enhancement of affinity for ferric ion is
evidenced between sulfoxinobactin and O-TRENSOX11 from
the plot pFe−pH (Figure S8, Supporting Information). It can
be inferred that sulfonate substituents do not favor the
predisposition of the chelating groups by the trilactone scaffold.
Figure 7. Spectrophotometric competition of ferric enterobactin in the
presence of oxinobactin at p[H] = 7.3, [MOPS] = 0.05 M, [Fe−
enterobactin] = [oxinobactin] = 1.6 × 10−4 M, T = 25 °C. Spectra
were recorded at t (min) = 1 (a) and 180 (b).
spectrum of the ferric enterobactin in this buffer exhibits a band
at 505 nm (ε = 4800 M−1 cm−1). After addition of 1 equiv of
oxinobactin to a solution of ferric enterobactin (Figure 7), the
spectra showed fast growth of bands at 450 and 580 nm that are
characteristic of the ferric oxinobactin. After 3 h the spectra did
not change and were similar to that of the ferric oxinobactin.
This clearly shows that oxinobactin completely remove FeIII
from ferric enterobactin. The same experiment was realized
with sulfoxinobactin in a MOPS aqueous buffer. Spectra
recorded after mixing of ferric enterobactin and sulfoxinobactin
12149
dx.doi.org/10.1021/ic301081a | Inorg. Chem. 2012, 51, 12142−12151

Inorganic Chemistry
Article
35 (Iron Transport and Storage in Microorganisms, Plants and
Animals), pp 239−327.
CONCLUSION
■
Two new analogs of the enterobactin have been synthesized:
the one lipophilic having 8-hydroxyquinoline groups (oxino-
bactin) and the other hydrophilic having 8-hydroxyquinoline-5-
sulfonate groups (sulfoxinobactin). The X-ray structure of the
ferric oxinobactin complex indicates C3 symmetry and slight
distortion from the octahedral geometry having the Δ
configuration. An interesting comparison can be made between
the ferric complexes of oxinobactin and the ligands COX200 or
COX2000 having the same chelating units and a C-pivot
tripodal scaffold grafted with a polyoxyethylenic chain, n = 3 or
45, respectively.35 The geometry around the iron cation is very
similar (distances and angles), indicating that the interaction
between the cation and the donor groups is of similar
magnitude for these two ligands. This suggests that the scaffold
structure has no effect on the interaction strength between the
chelating units and the iron cation. Furthermore, the efficiency
of the iron complexation, measured by the pFe, is enhanced by
4 pFe units between COX2000 (having a longer polyox-
yethylenic chain than COX200 allowing solubility in water) and
oxinobactin over the pH range 2−9. This suggests that the
predisposition of the chelating units is responsible for the gain
in stability of ferric oxinobactin over ferric COX2000. However,
no gain in stability is observed for the 8-hydroxyquinoline-5-
sulfonate chelating units by comparison of sulfoxinobactin and
O-TRENSOX11 having a TREN scaffold. Our results provide
additional evidence of the extraordinary ability of the trilactone
scoffold of enterobactin to predispose a great variety of
chelating units for selective and strong complexation of the
ferric ion and highlight the interest for ligands having the 8-
hydroxyquinoline moiety. Other complexing subunits can be so
intended to be transplanted, bringing additional properties as
fluorescence for intracellular localization of the iron.
(6) O’Brien, J. G.; Gibson, F. Biochim. Biophys. Acta 1970, 215, 393−
402.
(7) Pollack, J. R.; Neilands, J. B. Biochem. Biophys. Res. Commun.
1970, 38, 989−992.
(8) Loomis, L. D.; Raymond, K. N. Inorg. Chem. 1991, 30, 906−911.
(9) Mossialos, D.; Meyer, J. M.; Budzikiewicz, H. P. Appl. Environ.
Microbiol. 2000, 66, 487−492.
̀
(10) Baret, P.; Beguin, C.; Boukhalfa, H.; Caris, C.; Laulhere, J.-P.;
Pierre, J.-L.; Serratrice, G. J. Am. Chem. Soc. 1995, 117, 9760−9761.
(11) Serratrice, G.; Boukhalfa, H.; Beguin, C.; Baret, P.; Caris, C.;
Pierre, J.-L. Inorg. Chem. 1997, 36, 3898−3910.
(12) Biaso, F.; Baret, P.; Pierre, J.-L.; Serratrice, G. J. Inorg. Biochem.
2002, 89, 123−130.
(13) Pierre, J.-L.; Baret, P.; Serratrice, G. Curr. Med. Chem. 2003, 10,
1077−1084.
(14) Bush, A. I. Trends Neurosci. 2003, 26, 207−214.
(15) Boldron, C.; Van der Auwear, I.; Deraeve, C.; Gornitzka, H.;
Wera, S.; Pitie,
6, 1976−1980.
́
M.; Van Leuven, F.; Meunier, B. ChemBioChem 2005,
(16) Cui, Z.; Lockman, P. R.; Atwood, C. S.; Hsu, C.-H.; Gupte, A.;
Allen, D. D.; Mumper, R. J. Eur. J. Pharm. Biopharm. 2005, 59, 263−
272.
(17) Cherny, R. A.; Barnham, K. J.; Lynch, T.; Volitakis, I.; Li, Q.-X.;
McLean, C. A.; Multhaup, G.; Beyruether, K.; Tanzi, R. E.; Masters, C.
L.; Bush, A. I. J. Struct. Biol. 2000, 130, 209−216.
(18) Dedeoglu, A.; Cormier, K.; Payton, S.; Tseitlin, K. A.; Kremsky,
J. N.; Lai, L.; Li, Q.-X.; Moir, R. D.; Tanzi, R. E.; Bush, A. I.; Kowall, N.
W.; Rogers, J. T.; Huang, X. Exp. Geront. 2004, 39, 1641−1649.
(19) Lee, J.-Y.; Friedman, J. E.; Angel, I.; Kozak, A.; Koh, J.-Y.
Neurobiol. Aging 2004, 25, 1315−1321.
(20) Deraeve, C.; Pitie, M.; Mazarguil, H.; Meunier, B. New J. Chem.
2007, 31, 193−195.
(21) Deraeve, C.; Boldron, C.; Maraval, A.; Mazarguil, H.; Gornitzka,
H.; Vendier, L.; Pitie, M.; Meunier, B. Chem.Eur. J. 2008, 14, 682−
́
696.
(22) Cherny, R. A.; Atwood, C. S.; Xilinas, M. E.; Gray, D. N.; Jones,
W. D.; McLean, C. A.; Barnham, K. J.; Volitakis, I.; Fraser, F. W.; Kim,
Y. S.; Huang, X. D.; Goldstein, L. E.; Moir, R. D.; Lim, J. T.;
Beyreuther, K.; Zheng, H.; Tanzi, R. E.; Masters, C. L.; Bush, A. I.
Neuron 2001, 30, 665−676.
(23) Ritchie, C.; Bush, A. I.; Mackinnon, A.; Macfarlane, S.; Mastwyk,
M.; McGregor, L.; Kiers, L.; Cherny, R. A.; Li, Q.-X.; Tammer, A.;
Carrington, D.; Mavros, C.; Volitakis, I.; Xilinas, M.; Ames, D.; Davis,
S.; Beyreuther, K.; Tanzi, R. E.; Masters, C. L. Arch. Neurol. 2003, 60,
1685−1691.
(24) Bush, A. I. Neurobiol. Aging 2002, 23, 1031−1038.
(25) Meyer, M.; Telford, J. R.; Cohen, S. M.; White, D. J.; Xu, J.;
Raymond, K. N. J. Am. Chem. Soc. 1997, 119, 10093−10103.
(26) du Moulinet d’Hardemare, A.; Alnaga, N.; Serratrice, G.; Pierre,
J.-L. Bioorg. Med. Chem. Lett. 2008, 18, 6476−6478.
(27) Bastian, R.; Weberling, R.; Palilla, F. Anal. Chem. 1956, 28, 459−
462.
(28) Sheldrick, G. M. SHELXS-97, Program for crystal structure
ASSOCIATED CONTENT
* Supporting Information
■
S
Synthesis of oxinobactin precursors; X-ray crystallographic data
in CIF format for ferric oxinobactin; figures showing
potentiometric and spectrophotometric titrations of sulfox-
inobactin and its ferric complexes; figures showing a plot of the
absorbance at λmax vs p[H] for Fe−oxinobactin and Fe−
sulfoxinobactin; figures showing pFe vs pH for various ligands.
This material is available free of charge via the Internet at
http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: guy.serratrice@ujf-grenoble.fr.
Notes
■
The authors declare no competing financial interest.
solution; University of Gottingen: Gottingen, Germany, 1997.
̈
̈
Sheldrick, G. M. SHELXL-97, Program for the refinement of crystal
structures; University of Gottingen: Gottingen, Germany; 1997.
REFERENCES
̈
̈
■
Dolomanov, O. V.; Bourhis, L. J.; Gildea, R. J.; Howard, J. A. K.;
Puschmann, H. J. Appl. Crystallogr. 2009, 42, 339−341.
(29) Gans, P.; O’Sullivan, B. Talanta 2000, 51, 33−37.
(30) Rochester, C. H. Dalton Trans. 1972, 5−8.
(1) Crichton, R. R. In Inorganic Biochemistry of Iron metabolism; John
Wiley & Sons: Chichester, 2001.
(2) Kragter, J. Atlas of Metal-Ligand Equilibria in Aqueous Solution;
Halstead Press: New York, 1978.
(3) Boukhalfa, H.; Crumbliss, A. L. BioMetals 2002, 15, 325−339.
(4) Telford, J. R.; Raymond, K. N. Siderophores. In Comprehensive
Supramolecular Chemistry; Atwood, J. L., Davies, J. E. D., MacNicol, D.
(31) Gans, P.; Sabatani, A.; Vacca, A. Talanta 1996, 43, 1739−1743.
(32) Gampp, H.; Maeder, M.; Meyer, C. J.; Zuberbuhler, A. D.
̈
Talanta 1985, 32, 95−101.
(33) Gampp, H.; Maeder, M.; Meyer, C. J.; Zuberbuhler, A. D.
D., Vogtle, F., Eds.; Elsevier Science Ltd.; Oxford, 1996, Vol. 1; pp
̈
̈
Talanta 1985, 32, 257−264.
245−266.
(5) Albrecht-Gary, A.-M.; Crumbliss, A. L. In Metal Ions in Biological
Systems; Sigel, A., Sigel, H., Eds.; Marcel Dekker; New York, 1998, Vol.
(34) Gampp, H.; Maeder, M.; Meyer, C. J.; Zuberbuhler, A. D.
̈
Talanta 1986, 33, 943−951.
12150
dx.doi.org/10.1021/ic301081a | Inorg. Chem. 2012, 51, 12142−12151

Inorganic Chemistry
Article
(35) Imbert, D.; Baret, P.; Gaude, D.; Gautier-Luneau, I.; Gellon, G.;
Thomas, F.; Serratrice, G.; Pierre, J.-L. Chem.Eur. J. 2002, 8, 1091−
1100.
(36) Launay, F.; Alain, V.; Destandau, E.; Ramos, N.; Bardez, E.;
Baret, P.; Pierre, J.-L. New J. Chem. 2001, 25, 1269−1280.
(37) Brandel, J.; Torelli, S.; Gellon, G.; Serratrice, G.; Putaux, J.-L.;
Pierre, J.-L. Eur. J. Inorg. Chem. 2009, 86−92.
(38) The trigonal twist angle is defined by viewing the N3−N3ii−
N3iii and O1−O1ii−O1iii planes in a projection perpendicular to the
3-fold axis of the complex. It is calculated as the dihedral angle O1−
D1−D2−N3 (D1 and D2 are the centers of the N3−N3ii−N3iii and
O1−O1ii−O1iii triangles).
(39) Stack, T. D. P.; Karpishin, T. B.; Raymond, K. N. J. Am. Chem.
Soc. 1992, 114, 1512−1514.
(40) Martell, A. E.; Smith, R. M. Critical Stability Constants; Plenum
Press: New York, 1974; Vol. 1.
(41) Ringbom, A. Complexation in Analytical Chemistry; Interscience
Publisher: New York, 1963.
(42) Rodgers, S. J.; Lee, C.-W.; Ng, C. Y.; Raymond, K. N. Inorg.
Chem. 1987, 26, 1622−1625.
12151
dx.doi.org/10.1021/ic301081a | Inorg. Chem. 2012, 51, 12142−12151