Synthesis, solution behavior, thermal stability, and biological activity of an Fe(III) complex of an artificial siderophore with intramolecular hydrogen bonding networks

Article abstract of DOI:10.1021/ic048761g

Previously, an artificial siderophore complex, the iron(III) complex with tris[2-{(N-acetyl-N-hydroxy)glycylamino}-ethyl]amine (TAGE), was constructed in order to understand the effect of intramolecular hydrogen bonding interaction on the siderophore function, and its structural characterization in the solid state was reported (Inorg. Chem. 2001, 40, 190). In this paper, the solution behavior of the M(III)-TAGE (M = Fe, Ga) system has been investigated using ¹H NMR, UV-vis, and FAB mass spectroscopies in efforts to characterize the biological implication of hydrogen bonding networks between the amide hydrogens and coordinating aminohydroxy oxygens of the complex. The temperature dependence of ¹H NMR spectra for Ga(III) complex of TAGE indicates that hydrogen bonding networks are maintained in polar solvents such as DMSO-d₆ and D₂O. The UV-vis spectra of the Fe(III)-TAGE system under various pH conditions have shown that TAGE forms a tris(hydroxamato)iron(III) complex in an aqueous solution in the pH range 4-8. By contrast, tris[2-{(N-acetyl-N-hydroxy)propylamido}ethyl]amine (TAPE; TAGE analogue that is difficult to form intramolecular hydrogen bonding networks), which has been synthesized as a comparison of TAGE, forms both of bis- and tris(hydroxamato)iron(III) complexes in the same pH range. Both the stability constants (log β_FeTAGE = 28.6; β_FeTAGE = [Fe^IIITAGE]/([Fe³⁺][TAGE^3-])) and pM (-log[Fe³⁺]) value for Fe^IIITAGE (pM 25) are comparable to those of a natural siderophore ferrichrome (log β = 29.1 and pM 25.2). The kinetic study of the TAGE-Fe(III) system has given the following rate constants: the rate of the ligand exchange reaction between Fe^IIITAGE and EDTA is 6.7 × 10^-4 s^-1, and the removal rates of iron from diferric bovine plasma transferrin by TAGE are 2.8 × 10 ^-2 and 6.0 × 10_-3 min^-1. These values are also comparable to those of a natural siderophore desferrioxamine B under the same conditions. In a biological activity experiment, TAGE has promoted the growth of the siderophore-auxotroph Gram-positive bacterium Microbacterium flavescens, suggesting that TAGE mimics the activity of ferrichrome. These results indicate that the artificial siderophore with intramolecular hydrogen bonding networks, TAGE, is a good structural and functional model for a natural ferrichrome.

Full text of DOI:10.1021/ic048761g

Inorg. Chem. 2004, 43, 8538−8546
Synthesis, Solution Behavior, Thermal Stability, and Biological Activity
of an Fe(III) Complex of an Artificial Siderophore with Intramolecular
Hydrogen Bonding Networks
Kenji Matsumoto, Tomohiro Ozawa, Koichiro Jitsukawa, and Hideki Masuda*
Department of Applied Chemistry, Nagoya Institute of Technology,
Gokiso-cho, Showa-ku, Nagoya 466-8555, Japan
Received September 3, 2004
Previously, an artificial siderophore complex, the iron(III) complex with tris[2-{(N-acetyl-N-hydroxy)glycylamino}-
ethyl]amine (TAGE), was constructed in order to understand the effect of intramolecular hydrogen bonding interaction
on the siderophore function, and its structural characterization in the solid state was reported (Inorg. Chem. 2001,
4
0, 190). In this paper, the solution behavior of the M(III)
−
TAGE (M
)
Fe, Ga) system has been investigated
1
using H NMR, UV
−vis, and FAB mass spectroscopies in efforts to characterize the biological implication of hydrogen
bonding networks between the amide hydrogens and coordinating aminohydroxy oxygens of the complex. The
1
temperature dependence of H NMR spectra for Ga(III) complex of TAGE indicates that hydrogen bonding networks
are maintained in polar solvents such as DMSO-d
under various pH conditions have shown that TAGE forms a tris(hydroxamato)iron(III) complex in an aqueous
solution in the pH range 4 8. By contrast, tris[2- (N-acetyl-N-hydroxy)propylamido ethyl]amine (TAPE; TAGE
6 2
and D O. The UV−vis spectra of the Fe(III)−TAGE system
−
{
}
analogue that is difficult to form intramolecular hydrogen bonding networks), which has been synthesized as a
comparison of TAGE, forms both of bis- and tris(hydroxamato)iron(III) complexes in the same pH range. Both the
[Fe TAGE]/([Fe3+][TAGE3-])) and pM (
III
−
log[Fe3+]) value for
stability constants (log
â
FeTAGE
)
28.6;
â
FeTAGE
)
III
Fe TAGE (pM 25) are comparable to those of a natural siderophore ferrichrome (log â ) 29.1 and pM 25.2). The
kinetic study of the TAGE
reaction between Fe TAGE and EDTA is 6.7
transferrin by TAGE are 2.8
−
Fe(III) system has given the following rate constants: the rate of the ligand exchange
III
×
10 s-1, and the removal rates of iron from diferric bovine plasma
-4
×
10- and 6.0
2
×
-3
10 min-1. These values are also comparable to those of a natural
siderophore desferrioxamine B under the same conditions. In a biological activity experiment, TAGE has promoted
the growth of the siderophore-auxotroph Gram-positive bacterium Microbacterium flavescens, suggesting that TAGE
mimics the activity of ferrichrome. These results indicate that the artificial siderophore with intramolecular hydrogen
bonding networks, TAGE, is a good structural and functional model for a natural ferrichrome.
7
1
Introduction
of living cells (approximately 10- M). To overcome this
obstacle, microorganisms produce and release several kinds
of chelating compounds, siderophores, in order to sequester
Most living organisms require iron for their growth and
development. For microorganisms such as bacteria and fungi,
the availability of iron is uncertain because of the low
solubility of ferric ions. Under aerobic conditions, the iron
ion precipitates in the form of ferric oxyhydroxide with a
solubility product constant of 10-39 M, which limits the
2
-4
ferric ion from the environment. Siderophores are mainly
classified into hydroxamate- and phenolate (catecholate)-
(
1) Braun, V.; Hantke, K.; K o¨ ster, W. In Metal Ions in Biological Systems
Vol. 35: Iron transport and storage in microorganisms, plants, and
animals; Sigel, A., Sigel, H., Eds.; Marcel Dekker: New York, 1998;
p 67.
4
-
18
concentration of ferric ions to approximately 10 M at pH
. This is far below the level of demand for the iron supply
7
(
2) Hider, R. C. Struct. Bonding 1984, 58, 25.
(
3) Matzanke, B. F.; M u¨ ller-Matzanke, G.; Raymond, K. N. In Physical
Bioinorganic Chemistry Vol. 5: Iron Carriers and Iron Proteins;
Loehr, T. M., Ed.; VCH: New York, 1989; p 1.
*
To whom correspondence should be addressed. E-mail:
masuda.hideki@nitech.ac.jp. Phone: int code +81-52-735-5228. Fax: int
code +81-52-735-5209.
(4) Drechsel, H.; Jung, G. J. Pept. Sci. 1998, 4, 147.
8538 Inorganic Chemistry, Vol. 43, No. 26, 2004
10.1021/ic048761g CCC: $27.50
© 2004 American Chemical Society
Published on Web 11/26/2004

Fe(III) Complex of an Artificial Siderophore
Chart 1
hydroxamate siderophores, Olsen et al. have synthesized and
characterized retrodeferriferrichrome, a deferriferrichrome
analogue in which the carbonyl groups and aminohydroxyl
9
groups on the hydroxamates are exchanged. This investiga-
tion revealed that this modification of the hydroxamate
groups had little influenced on its biological activity, although
retroferrichrome is more labile than ferrichrome. Thus, many
synthetic retrohydroxamates were developed as the analogues
5
b,6
for hydroxamate-type siderophores.
Raymond and co-
workers characterized many Fe(III) complexes containing
1
0
catecholates and hydroxamates and prepared a thermody-
namically stable Fe(III) complex using a tripodal trihydrox-
amate ligand (TRENDROX) with a tris(2-aminoethyl)amine
10a
(
TREN) tether (log â ) 32.9). The thermodynamic stability
of the complex is greater than that of ferrichrome (log â )
3
3
2
9.1) and is comparable to ferrichrome A (log â ) 32.0),
although the ligand TRENDROX and its iron(III) complex
are insoluble at physiological pH.
Single crystal X-ray structure analyses performed on
natural siderophore complexes have revealed the existence
containing compounds which function as iron specific
chelating agents. Once siderophores form the complexes
with the iron(III) ion, and possess high stability constants
12
2
of intramolecular hydrogen bonds between the coordinating
oxygen and the amide hydrogen with the O‚‚‚(H)N distances
of 2.73-2.81 Å. In light of this finding, Shanzer and co-
workers have reported some tripodal compounds which
include an intramolecular hydrogen bond and found that the
amide-containing compounds tend to induce a propeller-like
conformation via formation of intramolecular hydrogen
(
log â ≈ 30 and 49 for hydroxamte-types and catecholate-
3
types, respectively), they are efficiently taken up into the
cell. The highly specific receptors on the outer membrane
of bacteria recognize unique structural characteristics of
siderophores, such as chirality, and transport the ferric
complexes into the periplasm in an active and energy-
dependent manner. Release of iron from ferric siderophores
in the cell often proceeds through degradation of the ligand
or reduction of ferric to ferrous ion which has a lower affinity
1
1
bonds in nonpolar solvents such as CHCl
3
.
The iron(III)
complexes of trihydroxamate ligands with a TREN tether
1
1e
did not exhibit significant biological activity. Thus, other
types of tripodal retro-trihydroxamate ligands with 1,1,1-
tris{(2-carboxyethoxy)methyl}propane or 1,1,1-tris{(carboxy-
methoxy)methyl}propane were employed as the tripodal
for the siderophore ligand.2,3 Indeed, the thermodynamic
stability of the Fe(II)-siderophores (log â ) 10-12) is much
5
lower than those of Fe(III)-siderophores (log â ≈ 30).
1
1b
Therefore, it is reasonable that the reduction of the Fe(III)
complex to the Fe(II) state plays an important role in
biological processing of iron by bacteria.
tethers. The Fe(III) complex with 1,1,1-tris{(2-carboxy-
ethoxy)methyl}propane as its tether has intramolecular
hydrogen bonding networks to form a five-membered ring,
and promoted growth of the siderophore-deficient mutants,
but the Fe(III) complex with 1,1,1-tris{(carboxymethoxy)-
methyl}propane, which does not form such a hydrogen bond,
Several approaches have been taken in the field of
biomimetics of siderophores and have clarified the charac-
teristics of Fe(III)-siderophores and the mechanisms of
6
1
1a,b
transportation of Fe(III) ion into the cell. Deferriferrichrome,
did not promote the growth.
This result suggests that
which is a siderophore containing three hydroxamate groups
as iron binding sites on an 18-membered cyclic hexapeptide
ring, was first isolated from the fungus Ustilago sphaerogena
(
9) (a) Olsen, R. K.; Ramasamy, K. J. Org. Chem. 1985, 50, 2264. (b)
Emery, T.; Emery, L.; Olsen, R. K. Biochem. Biophys. Res. Commun.
1984, 119, 1191.
7
as a ferric complex (ferrichrome) by Neilands in 1952. The
(
10) (a) Ng, C. Y.; Rodgers, S. J.; Raymond, K. N. Inorg. Chem. 1989,
other trihydroxamate-type siderophores were subsequently
2
8, 2062. (b) Rodgers, S. J.; Lee, C.-W.; Ng, C. Y.; Raymond, K. N.
8
Inorg. Chem. 1987, 26, 1622. (c) Pecoraro, V. L.; Harris, W. R.; Wong,
G. B.; Carrano, C. J.; Raymond, K. N. J. Am. Chem. Soc. 1983, 105,
identified. Examples of natural trihydroxamate-type sidero-
phores are shown in Chart 1. In order to develop synthetic
4623. (d) Meyer, M.; Telford, J. R.; Cohen, S. M.; White, D. J.; Xu,
J.; Raymond, K. N. J. Am. Chem. Soc. 1997, 119, 10093.
(
5) (a) Spasojevic, I.; Armstrong, S. K.; Brickman, T. J.; Crumbliss, A.
L. Inorg. Chem. 1999, 38, 449. (b) Dhungana, S.; Heggemann, S.;
Gebhardt, P.; M o¨ llmann, U.; Crumbliss, A. L. Inorg. Chem. 2003,
(11) (a) Jurkevitch, E.; Hadar, Y.; Chen, Y.; Libman, J.; Shanzer, A. J.
Bacteriol. 1992, 174, 78. (b) Dayan, I.; Libman, J.; Agi, Y.; Shanzer,
A. Inorg. Chem. 1993, 32, 1467. (c) Shanzer, A.; Libman, J.; Lifson,
S. Pure Appl. Chem. 1992, 64, 1421. (d) Albert-Gray, A. M.; Libman,
J.; Shanzer, A. Pure Appl. Chem. 1996, 68, 1243. (e) Shanzer, A.;
Libman, J.; Lazar, R.; Tor, Y.; Emery, T. Biochem. Biophys. Res.
Commun. 1988, 157, 389.
42, 42.
(
6) Albrecht-Gary, A.-M.; Crumbliss, A. L. In Metal Ions in Biological
Systems Vol. 35: Iron transport and storage in microorganisms, plants,
and animals; Sigel, A., Sigel, H., Eds.: Marcel Dekker: New York,
1
998; p 239.
(12) (a) For ferrichrome: van der Helm, D.; Baker, J. R.; Eng-Wilmot, D.
L.; Hossain, M. B.; Loghry, R. H. J. Am. Chem. Soc. 1980, 102, 4224.
(b) For ferrichrome A: Zarkin, A.; Forrester, J. D.; Templeton, H. J.
Am. Chem. Soc. 1966, 88, 1810. (c) For ferricrocin: Barnes, C.; Eng-
Wilmot, D. L.; van der Helm, D. Acta Crystallogr. 1984, C40, 922.
(
(
7) Neilands, J. B. J. Am. Chem. Soc. 1952, 74, 4846.
8) (a) Garibaldi, J. A.; Neilands, J. B. J. Am. Chem. Soc. 1956, 78, 526.
(b) Prelog, V.; Walser, A. J. Am. Chem. Soc. 1962, 45, 631. (c) Keller-
Schierlein, W.; Prelog, V. HelV. Chim. Acta 1961, 48, 710.
Inorganic Chemistry, Vol. 43, No. 26, 2004 8539

Matsumoto et al.
Chart 2
Scheme 1
Chart 3
Scheme 2
formation of intramolecular hydrogen bonds is an important
factor in uptake of the Fe(III)-siderophores by microorgan-
isms.
Several other artificial siderophores have been reported.6
However, the replacements of the N-terminal hydroxamate
with various functional groups are not recommended from
the requirement of the alkyl- or arylhydroxylamines because
such hydroxylamines are limited and their syntheses are
diffcult. By contrast, the introduction of hydroxamate group
as a C-terminal substituent is more easily accomplished than
for an N-terminal one. On the basis of the above experimental
findings, the ligand, tris[2-{(N-acetyl-N-hydroxy)glycylamino}-
ethyl]amine (TAGE: Chart 2, left), was synthesized. This
ligand consists of three components: a terminal acyl group,
an N-hydroxy-R-amino acid, and a TREN group as a tripodal
anchor, which can form intramolecular hydrogen bonds with
a six-membered ring between the aminohydroxyl oxygen and
amide hydrogen in the same strand when the Fe(III)
complexes are formed. Furthermore, many kinds of sidero-
phore analogues can be synthesized by replacing the terminal
acyl groups and/or N-hydroxy-R-amino acid residues by other
functional groups. Another ligand, tris[{2-(N-acetyl-N-
hydroxy)propylamido}ethyl]amine (TAPE: Chart 2, right),
was also synthesized as a ligand with poor intramolecular
hydrogen bonds between the amide hydrogen and amino-
hydroxyl oxygen due to the long distance. We have recently
analyzer. FAB-mass spectra were measured on SHIMADZU
KRATOS. Measurements of UV-vis spectra were performed with
1
a JASCO U-best 35 spectrophotometer. H NMR spectra (TMS or
DSS reference) were taken with a HITACHI R-90 spectrometer
equipped with a HITACHI 672 graph plotter or by a JEOL JMN-
Lambda 500 Fourier transformation spectrometer equipped with
AIR-VT thermocontroller. Infrared spectra were recorded using a
JASCO FT/IR-410 Fourier transformation spectrometer.
Syntheses of Ligands. The ligands employed, TAGE and TAPE,
were synthesized according to Schemes 1 and 2. Details of the
syntheses are as follows.
Tris[2-{(N-acetyl-N-benzyloxy)glycylamino}ethyl]amine
(TABGE). N-Acetyl-N-benzyloxyglycine was synthesized by the
literature methods.1 N-Acetyl-N-benzyloxyglycine (2.17 g, 9.72
mmol) and 1-hydroxybenzotriazol (HOBt; 1.31 g, 2.94 mmol) were
dissolved in 180 mL of THF at -3 °C. Furthermore, 20 mL of
THF solution containing tris(2-aminoethyl)amine (0.43 g, 2.94
mmol) was added to this solution. THF (20 mL) solution containing
dicyclohexylcarbodiimide (DCC; 2.01 g, 9.72 mmol) was then
added dropwise to this solution. The reaction mixture was stirred
for 2 h at -3 °C and then 2 h at room temperature. The solvent
was evaporated, and the residue was dissolved in 80 mL of ethyl
acetate. An insoluble dicyclohexyl urea precipitate was filtered off.
The filtrate was washed with 5% aqueous solution of NaHCO3,
and then the organic layer was treated with an aqueous 10% citrate
solution, which was prepared with sodium bicarbonate to pH 8 and
re-extracted three times with 50 mL of ethyl acetate. The organic
layer was dried with anhydrous magnesium sulfate and evaporated.
Pure TABGE was obtained as a pale yellow amorphous solid (1.80
4,15
III
reported the crystal structure of Fe TAGE, which revealed
that not only multiple intramolecular hydrogen bonds but
also intramolecular hydrogen bonding networks were formed
in the complex (Chart 3). The redox potential of the complex
was increased.13 In this Article, we report the solution
structure, thermal stability, kinetics, and biological activity
of the TAGE complex with an iron(III) ion and discuss the
contribution of intramolecular hydrogen bonding networks
in the siderophore function.
g; 80.3% yield). IR (KBr pellet): ν 1659, 3291 cm- . H NMR
90 MHz, CDCl ): δ 2.0 (s, CH CO), 2.5 (br, NCH CH ), 3.3 (br,
NCH CH ), 4.2 (s, NHCOCH ), 4.9 (s, PhCH ), 6.3 (m, Ph-H),
.8 (s, NHCO).
Tris[2-{(N-acetyl-N-hydroxy)glycylamino}ethyl]amine (TAGE).
TABGE (1.35 g, 1.77 mmol) dissolved in methanol (100 mL) was
hydrogenated with a 10% Pd/C catalyst under 3 atm H gas for
1 1
(
3
3
2
2
2
2
2
2
7
Experimental Section
General. Reagents and solvents were purchased from commercial
suppliers and used without further purification. Elemental analyses
were performed for C, H, and N elements on a LECO CHN-900
2
(
14) Kolasa, T.; Chimiak, A. Tetrahedron 1974, 30, 3591.
(13) Matsumoto, K.; Ozawa, T.; Jitsukawa, K.; Einaga, H.; Masuda, H.
(15) Kolasa, T.; Chimiak, A.; Kitowska, A. J. Prakt. Chim. 1975, 317,
Inorg. Chem. 2001, 40, 190.
252.
8540 Inorganic Chemistry, Vol. 43, No. 26, 2004

Fe(III) Complex of an Artificial Siderophore
several hours at room temperature. This hydrogenation was
monitored by silica gel TLC. When the reaction was completed,
the methanol solution was filtered through Celite to remove the
Pd/C and then evaporated. TAGE was obtained as pale yellow
amorphous solid (0.84 g; 96.6% yield). IR (KBr pellet): ν 1649,
estimated by least-squares calculations for the titration data using
the SUPERQUAD program system.16
Spectrophotometric Titration of the Fe(III) Complex of
TAGE with EDTA. The stability constant of the tris(hydroxamato)-
iron(III) complex of TAGE was spectrophotometrically investigated
by a competition reaction with N,N,N′,N′-ethylenediaminetetraacetic
acid (Na H edta) at 25 °C. The eight solutions were prepared in
-
1 1
3
2
4
285 cm . H NMR (500 MHz, DMSO-d
.5 (overlapping the solvent peak, NCH CH
.11 (s, NHCOCH ), 7.72 (br s, NHCO), 9.87 (br s, NOH).
6
): δ 2.02 (s, COCH
3
),
2
2
), 3.13 (td, NCH CH
2
2
),
2
2
III
2
50 mL volumetric flasks containing Fe TAGE (final concentra-
-
5
Found: C, 43.56; H, 6.79, N; 18.56%. Calcd for C18
H
33
N
7
O
9
‚
tion: 3.88 × 10 M) and Na H edta (final concentration: 1.16 ×
2
2
-5 -4
0
.75CH OH‚0.25H O: C, 43.31; H, 7.07; N, 18.85%. FAB-mass
3
2
10 to 7.74 × 10 M) with the ionic strength kept constant with
+
spectrum (m/z): 492 ([M + H] ) for TAGE.
0.1 M of sodium perchlorate (NaClO ). The pH values of the mixed
4
solutions were maintained constant at 6.1 with a 10 mM hexamine-
HClO buffer system. The absorption spectra of the solutions were
4
Tris[2-{(N-acetyl-N-hydroxy)propylamido}ethyl]amine (TAPE).
TAPE was synthesized in a manner similar to that for TAGE using
measured after standing for a few days until equilibrium was
completed for the ligand-exchange reaction.
3
-(N-benzyloxyamino)propionic acid instead of N-benzyloxygly-
1
cine. TAPE was obtained as a pale orange oil. H NMR (90 MHz
O): δ 2.1 (s, CH ), 2.5 (t, HONCH CH CO), 3.2 (t, NCH CH ),
III
D
3
2
3
2
2
2
2
Kinetics. (a) Ligand-Exchange Reaction between [Fe -
III
.5 (t, NCH
2
CH
2
), 3.9 (t, HONCH
2
CH
2
CO). FAB mass spectrum
(TAGE)] and Na
2
H
2
EDTA. Stock solutions containing Fe TAGE
+
-5
(m/z): 534 ([M + H ]) for TAPE.
(1.23 × 10 M) prepared from equimolar amounts of NH
4
Fe-
H -
2 2
III
(SO )
4 2
‚12H
2
O and TAGE and 10-40 excess amounts of Na
Synthesis of Fe(III) Complex of TAGE, [Fe (TAGE)]. To a
III
EDTA for Fe TAGE, which were prepared in 0.01 M acetate buffer
methanol solution (10 mL) containing 0.1 mmol of TAGE was
added tris(acetylacetonato)iron(III) (Fe (acac) ; 0.1 mmol). The
3
deep red crystal of the tris(hydroxamato)iron(III) complex was
precipitated after standing for a few hours and then isolated by
filtration. The compound thus obtained was identified by elemental
analysis and mass spectral measurements. Found: C,39.32; H, 5.92;
N, 17.37%. Calcd for C18
H, 5.80; N, 17.36%. FAB-mass spectrum (m/z): 545 ([M + H] )
for [Fe (TAGE)].
-
4
III
(pH 5.3, I ) 0.1 M NaClO
4
) to a final concentration of 9.8 × 10
III
M (Fe TAGE). The kinetic experiments were monitored using the
absorbance (Abs) at 420 nm at 25 °C. The pseudo-first-order rate
constant (kobs) was evaluated from a slope of the plots of ln[(Abs
-
A
∞
)/(A
0
∞ 0 ∞
- A )] as a function of time (t), where A and A are
III
-
the absorbance at t ) 0 and that of Fe EDTA , respectively. The
H
30
N
7
O
9
Fe‚0.25H
2
O‚0.5CH
3
OH: C,39.34;
+
second-order rate constants were calculated from the linear relation-
III
2 2
ship between the concentration of Na H EDTA and kobs.
(b) Removal of Iron(III) Ion from Bovine Plasma Transferrin
III
Preparation of Ga(III) Complex of TAGE, [Ga (TAGE)].
Using TAGE. The iron(III) removal reactions were carried out
using a commercially available bovine plasma iron saturated
transferrin (Tf) (99.999%, Nacalai tesque). The stock solutions
Ga(III) complex of TAGE was prepared according to the same
procedure as that of Fe TAGE using Ga (acac)
Fe (acac)
C
III
III
3
instead of
. Found: C, 38.66; H, 5.81; N, 16.61%. Calcd for
Ga‚CH OH: C, 38.71; H, 5.56; N, 16.66%. FAB-mass
III
3
-
3
composed of Fe
desferrioxamine B (DFO B) (1 mL, 2.00 × 10 M), and 0.1 M
HEPES buffer (1 mL, I ) 0.1 M KNO ) were prepared, and then
the absorbance (Abs) of the solutions was followed by monitoring
the band at 420 nm. The pseudo-first-order rate constants (k , k
were calculated from a slope of the plots of ln[(A - Abs)/(A
)] for the times, where A and A are the initial and final
absorbances for the respective reactions.
2
Tf (1 mL, 1.00 × 10 M), L ) TAGE or
H N
18 30 7
O
9
3
-2
+
69
III
spectrum (m/z): 558 and 560 ([M + H] ) for [ Ga TAGE] and
7
1
III
3
[
Ga TAGE], respectively.
UV-Vis Spectral Measurements. Sample solutions containing
1
2
)
III
III
Fe -TAGE or Fe -TAPE system were prepared to a final
∞
∞
-
-
5
concentration of 2.8 × 10 M in 50 mL volumetric flasks using
A
0
0
∞
-
4
the stock solutions (7.14 × 10 M) containing NH
2H O and ligand (TAGE or TAPE). The ionic strength and pH
values were kept constant with 0.1 M sodium perchlorate solution
4 4 2
Fe(SO ) ‚
1
2
Biological Activity. Biological activity was investigated by
observing the growth of the siderophore-auxotrophic bacterium,
Microbacterium flaVescens (ATCC No. 25091), in a siderophore-
free liquid medium. M. flaVescens was purchased from American
Type Culture Collection (ATCC). Siderophore-free ATCC no. 424
broth was prepared by the following method: bactopeptone (1 g),
(NaClO
(pH 1-3), CH
(pH 6-8). The absorption spectra of the solutions were measured
4
) and 0.01 M buffer solutions as follows: HClO
4
-NaClO
4
3
COOH-CH COONa (pH 3-5), HEPES-NaOH
3
after standing for 1 day.
Potentiometric Titration. A titrant solution was prepared using
deionized and distilled water that was further purified by passing
through a Millipore Milli-Q reverse osmosis cartridge system
2 4
yeast extract (1 g), and K HPO (0.2 g) were dissolved in 100 mL
of water, and then the solution was adjusted at pH 7.4 using 1 M
NaOH prior to the sterilization. A 25 mL flask containing 8 mL of
siderophore-free ATCC no. 424 broth with desferrioxamine me-
sylate (20 µg/L) was inoculated with 5 µL of the stock culture
solution and incubated for 20 h at 30 °C. After 20 h, the bacterium
in 1 mL of the culture solution was centrifuged for 15 min at 2000g
and washed with 1 mL of the siderophore-free ATCC no. 424 broth.
The washing of the cell was repeated to completely remove the
siderophore within the preincubated medium. The 25 mL flasks
were filled with 7.5 mL of the siderophore-free ATCC no. 424
2 2
(resistivity 18 MΩ cm). The system was flushed with CO -free N
gas prior to and during the titration. A carbonate-free 0.1 M KOH
solution was standardized by titrating against potassium hydrogen
phthalate. A 0.1 M HNO
against KOH. About 0.1 mmol of TAGE was dissolved into 40
mL of an aqueous solution containing 0.100 M of KNO . The
sample solution was acidified by addition of 1 mL of 0.1 M HNO
to fully protonate the ligand. The acidic TAGE solution was titrated
in triplicate with the standardized carbonate-free KOH solution.
All the titrations were performed using a Dosimat 665 (Metrohm)
piston buret and an AC-15 pH meter (Iwaki Glass) equipped with
a BECKMAN 39419 reference electrode and a BECKMAN 39321
glass electrode, at a constant temperature of 25 °C using a Hakke
circular thermostat model F-3. The protonation constants were
3
solution was standardized by titrating
3
3
III
broth containing a final concentration of 10 µM Fe TAGE and
then were inoculated with 0.5 mL of the preincubated culture and
incubated at 30 °C and 100 rpm. The degree of growth of the
(
16) Gans, P.; Sabatini, A.; Vacca, S. J. Chem. Soc., Dalton Trans. 1985,
1195.
Inorganic Chemistry, Vol. 43, No. 26, 2004 8541

Matsumoto et al.
siderophore-auxotrophic bacterium was examined by monitoring
the optical density of the solutions that were measured using the
absorption at 660 nm. Each treatment was performed in duplicate.
The biological activity was examined simultaneously for the
siderophore-free broth and the ferrichrome solution by the same
III
procedure as the reference for Fe TAGE.
Results and Discussion
Intramolecular Hydrogen Bonds in Polar Solvents As
1
Examined by H NMR. Previously, we reported the crystal
III
structure of Fe TAGE, which revealed the formation of
intramolecular hydrogen bonding networks between the
amide hydrogens and the coordinating aminohydroxyl oxy-
1
3
gens. In order to study the solution behavior of the iron-
1
(
III) complex, H NMR measurements of the TAGE complex
III
Figure 1. UV-vis spectral changes of Fe TAGE under various pH
with Ga(III) in the place of Fe(III) have been performed,
because Ga(III) ion has the same ionic charge as Fe(III) ion
and has an ionic radius similar to that of Fe(III) (0.62 and
conditions: (a) pH 2.4, (b) pH 2.9, (c) pH 3.1, (d) pH 3.6, (e) pH 4.1, (f)
pH 7.2, (g) pH 7.5. Conditions: [FeIII] ) 2.8 × 10-5 M, [TAGE] ) 2.9 ×
-
5
10
M, ionic strength ) 0.1 M NaClO4, cell length ) 50 mm.
0
.65 Å for Ga(III) and Fe(III), respectively, in an octahedral
1
The H NMR spectrum of TAGE also exhibited C
3
1
7
geometry) and because the diamagnetic Ga(III) complex
yields a clear NMR spectrum which is easily interpreted in
contrast to that of the paramagnetic Fe(III) complexes. The
2 6
symmetry in D O and DMSO-d . The signals corresponding
to the amide and aminohydroxyl protons were not detected
because of the fast NH/ND exchange reaction. For the Ga(III)
complex of TAGE, the four ethylene protons of the TREN
anchor, which are not detected in DMSO-d
overlap with the solvent peaks, are observed at 2.43, 2.58,
.76, and 3.66 ppm as nonequivalent signals, and the
methylene proton signals of the glycyl moiety are split with
an AX pattern as well as in DMSO-d . Interestingly, the
spectrum of the Ga(III) complex measured immediately after
the preparation in D O clearly indicated the amide proton
peak at 8.20 ppm, which disappeared when the sample stood
for a few minutes at room temperature. This finding indicates
that the amide proton is more inert for the NH/ND exchange
reaction than the ligand TAGE, suggesting that the amide
protons are well shielded from attack of the solvent. Thus,
the intramolecular hydrogen bonding networks in the
1
H NMR spectra of the Ga(III) complex of TAGE measured
in DMSO-d
6 2
and D O indicated that the complex formed a
6
because of
1:1 Ga(III)-tris(hydroxamato) species, which was deter-
mined also by the elemental analysis and FAB-mass spec-
trum.
2
1
The H NMR spectrum of the metal-free TAGE ligand in
6
6 3
DMSO-d exhibits a simple pattern suggesting C symmetry.
The signal observed at 3.13 ppm was assigned to the
methylene protons attached to the amide group of the TREN
cap. The other methylene protons of the TREN were not
found by overlap with the signals of a solvent. The methylene
peak of the glycine moiety was observed at 4.11 ppm as a
singlet peak. The two remaining signals in the lower field
region were assigned to the amide (7.72 ppm) and hydroxide
protons (9.87 ppm).
2
III
13
Fe TAGE complex, as shown in the crystal structure, are
maintained even in polar solvents such as DMSO and D O.
1
The H NMR spectrum of the Ga(III) complex of TAGE
2
6 3
in DMSO-d also indicated C symmetry. The acetyl proton
Complexation Behavior of Fe(III)-Artificial Sidero-
phore Compound Systems in an Aqueous Solution. The
complexation behavior of ferric artificial siderophores was
studied using UV-vis spectral measurements as a function
of pH. The ferric hydroxamate complexes give a character-
istic ligand-to-metal charge transfer (LMCT) band in the
visible region, which is characteristic of the number of
hydroxamate groups bound to the iron(III) ion. The UV-
vis spectra of Fe -TAGE and Fe -TAPE systems under
several pH’s in an aqueous solution are represented in Figures
and 2, respectively.
signal shifted to the lower field region compared to that of
metal-free TAGE. However, we failed in assignment of the
methylene protons of the TREN cap because the signal
overlaps with the solvent. The methylene protons of the
glycine moiety is split with an AX pattern (δ 4.19, 4.64 ppm;
J ) 17.6 Hz), suggesting the steric hindrance of rotation
around the methylene carbon. The amide proton signal of
TAGE, which appeared at 7.72 ppm, also shifted to the
lower-field region (7.93 ppm) upon complexation with
Ga(III) ion. The chemical shift showed limited temperature
dependence (temperature coefficients for the Ga(III) com-
plex: 0.39 ppb/K), in contrast to the temperature-dependent
higher field shift observed for the metal-free TAGE (-4.6
ppb/K) (Figure S1). These results indicate that the amide
protons of the TAGE complex form extremely strong
intramolecular hydrogen bonding networks even in a polar
solvent.
1
8
III
III
1
III
In the Fe -TAGE system, the UV-vis spectra under an
acidic condition (pH 2.4) showed the absorption band
corresponding to a bis(hydroxamato)iron(III) complex at 459
nm. Increasing the pH from 3 to 5 causes a blue shift of the
band with an isosbestic point at 475 nm, and a new
absorbance maximum at 420 nm (ꢀmax ) 2860 M cm )
at pH 4. This observation is indicative of formation of a tris-
-
1
-1
(
hydroxamato)iron(III) complex (Table 1). This finding
(
17) Martell, A. E.; Hancock, R. D. In Metal Complexes in Aqueous
Solutions; Fackler, J. P., Jr., Ed.; Plenum: New York, 1996; p 149.
indicates that the conversion from the bis(hydroxamato)-
8542 Inorganic Chemistry, Vol. 43, No. 26, 2004

Fe(III) Complex of an Artificial Siderophore
the characteristic absorption band suggesting formation of a
bis(hydroxamato)iron(III) complex in the range pH 2-7.
Over pH 7, the formation of a tris-type complex has been
barely observed.
Although TRENDROX, which has structure similar to
TAGE and TAPE, can form a stable Fe(III)-tris(hydroxa-
mato) complex even in acidic conditions, the neutral ferric-
TRENDROX starts to precipitate above pH 3 because of the
1
0a
hydrophobic p-tolyl groups of its molecule. However, the
neutral ferric TAGE complex without such a hydrophobic
substituent is soluble even at neutral pH. The higher water
solubility of the ferric TAGE complex might be advantageous
for the biological availability.
Protonation and Stability Constants. Protonation con-
III
Figure 2. UV-vis spectral changes of Fe TAPE under various pH
n
stants (pK ) were evaluated via potentiometric titration. The
conditions: (a) pH 2.4, (b) pH 4.2, (c) pH 6.7, (d) pH 7.5. Conditions:
III
-5
-5
16
[
Fe ] ) 2.8 × 10 M, [TAPE] ) 3.0 × 10 M, ionic strength ) 0.1 M
titration data were treated with SUPERQUAD, and the
resultant protonation constants of TAGE are listed in Table
NaClO4, cell length ) 50 mm.
2
.
iron(III) complex to the tris(hydroxamato)iron(III) complex
occurs in this pH range. The absorption band of the tris-
The three basic constants observed in the range pH 7-9
were assigned to the protonation constants of the three
(
8
hydroxamato)iron(III) complex is maintained from pH 4 to
III
hydroxamate groups, which are almost identical in magnitude
for the Fe -TAGE system, whose behavior is similar to
2
4d
26
with those of ferrichrome and its analogues with three
other synthetic trihydroxamates. However, this range is
1
0a,25,26c
hydroxamate groups
(Table 2). The average value of
narrow in comparison with those of natural trihydroxamate
siderophores5
a,19
the protonation constants calculated for the hydroxamate
groups of TAGE, log K ) 8.93, is in excellent agreement
with that of the monomeric hydroxamate compound, N-
methylacetohydroxamic acid (N-methylAHA) (log K )
(pH 2-11) and an artificial siderophore
TRENDROX (pH > 2).1 The difference in complexation
0a
behavior between TAGE and natural trihydroxamates could
be explained as follows: the tertiary amine of the TAGE
ligand is protonated under acidic conditions (pH 2) (Table
2
0
8
.95). The difference in successive protonation constants
III
13
is nearly equal to 0.9 log units, which is larger than the
2
). As shown in the crystal structure of Fe TAGE, the lobe
2
1
statistical value (log 3 ) 0.48), indicating that the proto-
nation of one hydroxamate group influences those of the
others. The constant obtained in the most acidic region (log
of the lone pair of electrons of the tertiary amine and the
three amide protons for TAGE are directed toward the inside
of the molecule, suggesting that protonation at the tertiary
amine nitrogen would impede the formation of the tris-
K
4
) has been assigned to that of the tertiary amine, which is
consistent with that of a tri-acetylated TREN-type compound,
TRENAC (pK ) 5.98). The lower basicity of the amine
a
in comparison with a simple amine must be attributed to the
five-membered ring formed by hydrogen bond between the
amine nitrogen and the amide N-H hydrogen atoms. These
results suggest that a weak hydrogen bond occurs between
the strands and tertiary amine even in metal-free ligand.
(
hydroxamato)iron(III) complex because of the steric and
2
2
electrostatic repulsions. By contrast, under more basic
conditions (>pH 8), such intramolecular hydrogen bonding
networks may be difficult to maintain because of intermo-
lecular hydrogen bonds by surrounding hydroxyl ions. It is
also clear from the fact that the iron(III) complex of TAPE,
which is the TAGE analogue bearing poor intramolecular
hydrogen bonds, cannot form a stable tris(hydroxamato)-
iron(III) complex as described above. Actually, the UV-
vis spectra of the iron(III) complex with TAPE exhibited
The stability constant was spectrophotometrically esti-
III
mated for Fe -TAGE system by means of the competition
reaction with N,N,N′,N′-ethylenediaminetetraacetic acid di-
2 2
sodium salts (Na H EDTA) at pH 6.1. The competition
reaction is defined as shown in eq 1.
(
18) Birus, M.; Bradic, Z.; Kujundzic, N.; Pribanic, M.; Wilkins, P. C.;
Wilkins, R. G. Inorg. Chem. 1985, 24, 3980.
(
(
(
(
19) Neilands, J. B. Struct. Bonding 1984, 58, 1.
20) Caudle, M. T.; Crumbliss, A. L. Inorg. Chem. 1994, 33, 4077.
21) Adams, E. Q. J. Am. Chem. Soc. 1916, 38, 1503.
Keq
III
2-
+
Fe TAGE + H EDTA + H y z
2
22) Cohen, S. M.; Meyer, M.; Raymond, K. N. J. Am. Chem. Soc. 1998,
III
-
Fe EDTA + H TAGE (1)
120, 6277.
3
(
(
23) The side reaction coefficients for proton are defined as follows: R )
m
+ i
1
+ ∑ Ki[H ] , where Ki is the protonation constant of ligands.
i)1
III
Here, the species other than Fe TAGE did not exhibit
24) In Critical Stability Constants; Smith, R. M., Martell, A. E., Eds.;
Plenum Press: New York, 1989; (a) Vol. 6, pp 96, (b) Vol. 6, pp 98,
such an absorption band in the pH region examined. Upon
an increase in the concentration of EDTA, the absorption
band at 420 nm gradually decreased with an isosbestic point
at 354 nm, indicating that the reaction proceeds without other
side reactions as defined in eq 1. The overall formation
(c) Vol. 3, pp 303, (d) Vol. 3, pp 305.
(
(
25) Motekaitis, R. J.; Sun, Y.; Martell, A. E. Inorg. Chem. 1991, 30, 1554.
26) (a) Hara, Y.; Shen, L.; Tsubouchi, A.; Akiyama, M.; Umemoto, K.
Inorg. Chem. 2000, 39, 5074. (b) Hara, Y.; Akiyama, M. Inorg. Chem.
1
1
996, 35, 5173. (c) Akiyama, M.; Hara, Y.; Gunji, H. Chem. Lett.
995, 225.
Inorganic Chemistry, Vol. 43, No. 26, 2004 8543

Matsumoto et al.
Table 1. Coordination Properties of Artificial and Natural Siderophores
λmax
ꢀmax
ligand-exchange
iron removal rate
k1, k2 (min )
E1/2 (mV
vs NHE)
(M- cm-1
1
)
log â
pMk
rate kobs(s
-1
)
-1 m
(
nm)
III
420i
425
428
442
425
425
2860i
2895
2800
2934
2880
2800
28.6(3)j
29.1
30.5
32.9
26.4
28
-4 l
2.8 × 10-2, 6.0 × 10-3
3.8 × 10-2, 9.0 × 10-3 f
g,n
Fe TAGE
25
6.7 × 10
6.1 × 10
0.71 × 10
-230
-400
-468
a
-4 d
h
h
ferrichrome
25.2
26.6
27.8
22.5
24
a
-5 e
ferrioxamine B
III
b
Fe TRENDROX
III
c,o
d,p
Fe -ligand 6
III
2.7 × 10-2
Fe -ligand 3
a
Reference 3. b Reference 10a. Reference 25. d Reference 26c. e Reference 26b. In this work. g Reference 13. h Reference 6. i Conditions: pH 7, ionic
c
f
j
k
strength I ) 0.1 M NaClO4. Conditions: pH 6.1, I ) 0.1 M NaClO4, 25 °C. The pM value is estimated under the experimental condition that the
l
concentrations of the Fe(III) ion (CFe) and ligand (CL) are 1.0 and 10.0 µM, respectively, at pH 7.4. Conditions: pH 5.3 (0.01 M acetate buffer), I ) 0.1
m
n
M NaClO4, 25 °C. Conditions: pH 7.4 (0.01 M HEPES buffer), I ) 0.1 M KNO3, 25 °C. Conditions: pH 7, surpporting electrolytes 0.1 M NaClO4.
o
p
Ligand 6: CH3C[CH2O(CH2)3N(OH)COCH3]3. Ligand 3: [Ac-Ala-Ala-â(HO)Ala-Ala-NHCH2CH2]3N.
Table 2. pKn’s of Artificial and Natural Siderophores
contribute to stabilization of the resonance structure of
hydroxamate anion. Therefore, the high stability of
Fe TAGE without such a p-tolyl group might be attributed
to the intramolecular hydrogen bonding networks around the
iron(III) ion.
log K1
log K2
log K3
log K4
III
TAGEa
deferriferrichrome
desferrioxamine B 10.79 (RNH2) 9.70
TRENDROX
ligand 6
ligand 3
N-methylAHA
9.790(3)
9.83
8.948(3) 8.066(3) 5.459(3) (NR3)
b
c
9.00
8.11
9.03
8.58
8.39
8.26
8.39
d
10.30
9.69
9.69
8.95
9.33
9.02
9.03
6.51 (NR3)
e,h
The stability constant is very useful to compare the overall
thermodynamic stability of tris(hydroxamato)iron(III) com-
plexes. However, it does not provide an evaluation of the
binding strength of the Fe(III) ion under specified or
physiological conditions, because the competition of the
Fe(III) ion with a proton is not negligible. Therefore, the
f,i
g
a
b
Conditions: 25 °C, ionic strength I ) 0.1 M KNO3. Reference 24d.
c
Reference 24c. Reference 10a. Reference 25. f Reference 26c. Ref-
erence 20. Ligand 6: CH3C[CH2O(CH2)3N(OH)COCH3]3. Ligand 3:
Ac-Ala-Ala-â(HO)Ala-Ala-NHCH2CH2]3N.
d
e
g
h
i
[
3
+
pM value ()-log[Fe (H
2 6
O) ]) has been adopted in order
constant in the equilibrium eq 2 is given in eq 3 on the basis
of material balances and the absorbance changes
to estimate the binding strength of the Fe(III) ion under
3
III
specified conditions. The pM value for Fe TAGE is given
by eq 4 and was determined under the experimental condition
that the concentrations of the Fe(III) ion (CFe) and ligand
â
3
+
3-
III
Fe + TAGE h Fe TAGE
(2)
(C
L
) are 1.0 and 10.0 µM, respectively, at pH 7.4.
âFeTAGE
)
CEDTA - (1 - Z)CFe RTAGE
3+
CFe
Z
- Z
pM ) -log[Fe ] ) -log
(4)
âFeEDTA (3)
(1
)(
CTAGE - ZCFe
)(REDTA
)
C â
L
FeL
(
+ 1
)
RL
A - AE
Z )
(
)
A0 - AE
The pM value obtained is listed in Table 1. The pM of
Fe TAGE (25) is equivalent to that of ferrichrome (25.2).
III
3
where A ) absorbance of the competing system in the
equilibrium state, A
initial absorbance of Fe TAGE, RTAGE and REDTA are the
side reaction coefficients for protonations of TAGE and
EDTA, respectively, CEDTA, CTAGE, and CFe are the total
concentrations of EDTA, TAGE, and Fe(III) ion, respec-
This value is slightly larger in comparison with those of other
III
-
E
) absorbance of Fe EDTA , A
0
)
ferric trihydroxamato complexes except for TRENDROX
III
23
25,26
(Table 1).
This result suggests that the ligand TAGE can
strongly bind to the Fe(III) ion even in a neutral aqueous
solution.
Kinetics. In order to investigate the kinetic lability of
III
-
III
tively, and âEDTA is the stability constant of Fe EDTA .
Since Fe EDTA does not show any significant absorption
Fe TAGE, ligand exchange reactions with EDTA (20 times
III
-
excess) were performed at pH 5.3 by UV-vis spectroscopy.
III
at 420 nm, the term of A
stability constants of EDTA are obtained from the literature
E
is negligible. The protonation and
The UV-vis spectral changes of Fe TAGE after addition
24a,b
of EDTA are shown in Figure 3. The characteristic absorption
band at 420 nm, which is characteristic of an iron(III)-tris-
(hydroxamato) complex, decreases with an isosbestic point
(
2
log K1EDTA ) 10.19, log K2EDTA ) 6.13, and log âEDTA
5.1). The stability constant of Fe TAGE (log âFeTAGE) using
)
III
III
eq 3 was calculated to be 28.6(3), which is comparable to
at 350 nm, indicating that the iron removal from Fe TAGE
3
those of ferrichrome (log â ) 29.1) and other synthetic
by EDTA proceeds according to eq 1.
The observed pseudo-first-order rate constant (kobs) of
1
0a,25,26
trihydroxamates (log â ≈ 26-32)
(Table 2). This
III
-4 -1
value, however, is lower than that of TRENDROX (log â )
Fe TAGE has been estimated to be 6.7 × 10
s
(Table
3
2.9),1 which is supposed to be attributed to the difference
0a
1). This value is comparable to that of the natural siderophore
complex, ferrichrome, as measured under the same conditions
(Table 1), which indicates that the complex is 10-1000 times
more inert than the previously reported ferric artificial
in the hydroxamate substituent groups of both ligands.
Because TRENDROX has a p-tolyl group that strongly
6
stabilizes the resonance state of the hydroxamate anion, the
2
6
methyl groups of the hydroxamates in TAGE do not
siderophore complexes. These findings suggest that in-
8544 Inorganic Chemistry, Vol. 43, No. 26, 2004

Fe(III) Complex of an Artificial Siderophore
III
Figure 3. Time course of UV-vis spectra for the reaction of Fe TAGE
with a 20-fold excess of EDTA. The inset shows the plot of ln[(Abs -
III
A∞)/(A0 - A∞)] vs time on ligand exchange reaction of Fe TAGE with
EDTA, where A0 and A∞ are initial and final absorbances, respectively.
III
-5
-3
Conditions: [Fe TAGE] ) 9.8 × 10 M, [EDTA] ) 2.0 × 10 M, ionic
strength ) 0.1 M NaClO4, pH 5.3 (0.01 M acetate buffer), cell length )
0 mm, temperature ) 25 °C.
2
tramolecular hydrogen bonds are important for formation of
inert iron(III) complexes. A linear relationship was found
between kobs and the concentrations of EDTA (Figure S2),
the second-order rate constant of which was 0.34 M- s ,
indicating a simple bimolecular reaction. In contrast, the iron
release reaction from ferrioxamine B to EDTA has exhibited
1
-1
2
7
saturation kinetics for high concentrations of the ligand.
III
These observations suggest that the more inert Fe TAGE
complex results from limited access of EDTA rather than
pre-equilibrium between Fe TAGE and EDTA. In fact, it
Figure 4. Time course of UV-vis spectra for the reaction of FeIII2Tf with
a 20-fold excess of TAGE. The inset shows the plot of ln[(A∞ - Abs)/(A∞
III
III
- A0)] vs time on the iron removal reaction from Fe 2Tf by TAGE, where
A0 and A∞ are the initial and final absorbances in each reaction, respectively.
is known that the iron release from transferrin is kinetically
slow because of the closed conformation of the apo-protein
wrapping the iron ion, although its thermodynamic stability
III
-5
-4
Conditions: [Fe 2Tf] ) 3.3 × 10 M, [TAGE] ) 6.7 × 10 M, ionic
strength ) 0.1 M KNO3, pH 7.4 (0.01 M HEPES buffer), cell length ) 10
mm, temperature ) 25 °C.
28
III
-
constant (log â ≈ 20) is lower than that of Fe EDTA
log â ) 25.1). As shown in the crystal structure of
this artificial siderophore proceed in the successive two step
reactions.
2
4
(
III
13
Fe TAGE, the iron(III) ion is wrapped in the tripodal
strands fixed by the intramolecular hydrogen bonding
networks. Therefore, the lesser lability of Fe TAGE may
k1
III
Fe Tf + L
9' FeTf + FeL
2
result from the fact that the central metal ion is protected
from attack of EDTA by the TAGE ligand.
k2
FeTf + L
9' Tf + FeL
Next, the iron removal reaction from the iron-containing
compounds is also one of the important functions expected
for artificial siderophores. In order to investigate the
functionality as the iron removal agent, the iron removal
The two observed pseudo-first-order rate constants, k
1
and
-
2
-3
k
2
, at pH 7.4 and 25 °C were 2.8 × 10 and 6.0 × 10
-1
min , respectively. These rate constants are comparable to
those for a natural siderophore desferrioxamine B (DFO B,
2
reactions from bovine plasma transferrin (Fe Tf) as an iron
-
2
-3
3
.8 × 10 and 9.0 × 10 , respectively) under the same
containing compound by TAGE were studied by UV-vis
conditions (Table 1), suggesting that TAGE is able to abstract
iron from the iron sources such as transferrin in the
environment. Transferrin has two iron-binding sites at the
C- and N-terminal lobes. Generally, the binding ability of
C-terminal lobe is larger than that of the N-terminal. On the
9
spectroscopy. As shown in Figure 4, the absorption band
around 420 nm, which indicates the formation of the
iron(III)-tris(hydroxamato) complex, increased with time.
Initially, the isosbestic point was observed at 355 nm, and
later another one was detected at 363 nm. This result
indicates that the iron removal reactions from transferrin by
basis of the above structural information, the faster (k
1
) and
slower (k ) reactions may correspond to the releasing rates
2
of iron ions from N- and C-terminal lobes, respectively.
(
(
27) Tufano, T. P.; Raymond, K. N. J. Am. Chem. Soc. 1981, 103, 6617.
28) Harris, D. C.; Aisen, P. In Physical Bioinorganic Chemistry Vol. 5:
Iron Carriers and Iron Proteins; Loehr, T. M., Ed.; VCH: New York,
Biological Activity. The growth-promoting activity of
III
Fe TAGE was studied using the siderophore-auxotroph
1989; p 239.
gram-positive bacterium Microbacterium flaVescens, which
Inorganic Chemistry, Vol. 43, No. 26, 2004 8545

Matsumoto et al.
Conclusion
With a view to understanding the effect of intramolecular
hydrogen bonding networks in the natural siderophores with
hydroxamate groups, the artificial siderophore TAGE with
such networks was synthesized. Also, it was shown that the
intramolecular hydrogen bonding networks, as previously
III
13
demonstrated in the crystal structure of Fe TAGE, are
formed even in aqueous and DMSO-d solutions using
6
III
1
Ga TAGE by H NMR. TAGE complexes iron(III) ion as
a tris-hydroxamato complex with a 1:1 ratio in the range
pH 4-8 in solutions, as measured by UV-vis spectroscopy.
By contrast, TAPE, the TAGE analogue bearing poor
intramolecular hydrogen bonds, is unlikely to form a tris-
hydroxamato complex at pH 7. Furthermore, it was dem-
onstrated that the thermodynamic and kinetic properties of
Figure 5. Growth curves of M. flaVescens in liquid media, as followed
by monitoring the increase in optical density at 660 nm (2, no siderophore;
III
III
b, 10 µM ferrichrome; 9, 10 µM Fe TAGE). Conditions: temperature )
Fe -TAGE system are similar to those of natural sidero-
3
0 °C, shaking speed ) 100 rpm.
phores and other synthetic trihydroxamates. These results
indicate that TAGE has a higher affinity for an iron(III) ion
than an iron(II) ion, although the redox potential of
is known not to be able to synthesize any siderophore by
itself and can use only hydroxamate-type siderophores for
the growth. M. flaVescens in the siderophore-deficient
medium did not grow even in the presence of excess iron.
III
13
Fe TAGE (-230 mV vs NHE) is nearly 200 mV higher
than those of ferric natural trihydroxamates. TAGE also
promoted the growth of the siderophore-auxotroph gram-
positive bacterium Microbacterium flaVescens, indicating that
TAGE has the siderophore function comparable to the natural
ferrichromes. These results indicate that the artificial sid-
erophore TAGE is a good structural and functional model
for ferrichrome, and that the intramolecular hydrogen bond-
ing networks stabilize the iron(III) complex and increase the
redox potential even in an aqueous solution. These observa-
tions highlight the importance of intramolecular hydrogen
bonding networks for the coordination properties of the
iron(III) complex.
III
By contrast, in the presence of 10 µM Fe TAGE, the
microbe grew with the growth rate similar to that in the
ferrichrome-containing media (Figure 5). Interestingly, TAGE
III
as well as Fe TAGE showed the growth promoting activity
for M. flaVescens, suggesting that TAGE functioned in the
same manner as a natural siderophore. We previously
III
reported that Fe TAGE exhibited a much higher redox
potential value (-230 mV vs NHE)13 relative to that of
natural siderophores (Table 1) and that the high redox
III
potential of Fe TAGE is attributed to the intramolecular
Acknowledgment. The authors gratefully thank Prof.
Osamu Yamauchi and Dr. Tatsuo Yajima (Nagoya Univer-
sity) for potentiometric titration measurements and SUPER-
QUAD analysis for determination of protonation constants.
This work was supported in part by a Grant-in-Aid for
Scientific Research from the Ministry of Education, Science,
Sports, and Culture of Japan (K.M. and H.M.), for which
we express our thanks. K.M. is grateful to the Japan Society
for the Promotion of Science for the JSPS Research Fel-
lowship for Young Scientists.
13
hydrogen bonding networks. However, the influence of the
redox potentials of the complexes on biological activity was
not observed despite the difference in the redox potentials
III
between them; Fe TAGE and ferrichrome did not express
different growth rates. This may be explained as follows:
III
Since Fe TAGE tends easily to be reduced to iron(II), the
inefficient recognition of the iron(III) complex and/or its
transportation might be overcome by quick reduction of the
complex, or only the iron ion might be taken up into the
cell through the reduction of the iron(III) complex on the
cell surface as can be seen in fungi. These results suggest
that the TAGE compound is a good functional model for
ferrichrome.
Supporting Information Available: Additional figures. This
material is available free of charge via the Internet at
http://pubs.acs.org.
IC048761G
8546 Inorganic Chemistry, Vol. 43, No. 26, 2004