Synthesis and investigation of a chiral enterobactin analogue based on a macrocyclic peptide scaffold

Article abstract of DOI:10.1002/chem.200801552

A chiral C₃-symmetric enterobactin analogue (1) has been synthesized by attachment of three 2,3-dihydroxybenzoyl units to a chiral oxazole-containing macrocyclic peptide scaffold. Complex formation kinetics and stoichiometry with various metal ions were investigated by spectrophotometric methods. In the cases of Al^III, In^III and Fe^III complexes, UV absorption and CD kinetics showed nonlinearity, which results from slow conformational changes of the octahedral complexes. Virtual binding constants were determined from UV absorption data and showed selective binding of Ga^III in preference to Fe^III, by two orders of magnitude. CD spectroscopy revealed highly diastereoselective binding of Al ^III, Ga^III, In^III, Fe^III and Ge ^IV ions at room temperature, corresponding to the helical chirality opposite to that of the analogous enterobactin complexes. Ab initio calculations confirmed the energetic stabilization of the A isomers relative to the A isomers.

Full text of DOI:10.1002/chem.200801552

FULL PAPER
DOI: 10.1002/chem.200801552
Synthesis and Investigation of a Chiral Enterobactin Analogue Based on a
Macrocyclic Peptide Scaffold
ꢀron Pintꢁr and Gebhard Haberhauer*[a]
Abstract: A chiral C3-symmetric enter-
obactin analogue (1) has been synthe-
sized by attachment of three 2,3-dihy-
droxybenzoyl units to a chiral oxazole-
containing macrocyclic peptide scaf-
fold. Complex formation kinetics and
stoichiometry with various metal ions
were investigated by spectrophotomet-
ric methods. In the cases of AlIII, InIII
and FeIII complexes, UV absorption
and CD kinetics showed nonlinearity,
which results from slow conformational
changes of the octahedral complexes.
Virtual binding constants were deter-
mined from UV absorption data and
showed selective binding of GaIII in
preference to FeIII, by two orders of
magnitude. CD spectroscopy revealed
highly diastereoselective binding of
AlIII, GaIII, InIII, FeIII and GeIV ions at
room temperature, corresponding to
the helical chirality opposite to that of
the analogous enterobactin complexes.
Ab initio calculations confirmed the
energetic stabilization of the L isomers
relative to the D isomers.
Keywords:
amino
acids
·
CD spectroscopy · enterobactin ·
macrocycles · oxazoles
Introduction
Enterobactin is a well-known triscatecholamide-type sidero-
phore discovered almost 40 years ago. It is produced by Es-
cherichia coli and related enteric bacteria in order to over-
come iron-deficient conditions in living host organisms.[1]
This low-molecular-weight ironACTHUNTGRNEUNG(III) carrier is made up of
three (S)-serine units that form a cyclic trilactone scaffold
bearing three 2,3-dihydroxybenzamide binding arms.
Enterobactin possesses an extremely high affinity for FeIII
ions, and the proton-independent stability constant (b110) of
the [FeACHTUNGTRENNUNG
(ent)]3ꢀ complex has been estimated to be 1049 mꢀ1.[2]
In addition, its octahedrally coordinated complexes of 1:1
stoichiometry with a variety of metal ions, such as CrIII,[3]
GaIII,[4] AlIII,[4a,5] InIII,[4a,5] ScIII,[4a,5,6] RhIII,[7] and VIV[8] have
been described. The FeII complex of fully protonated entero-
bactin has also been verified.[9]
A matter of particular interest is the stereochemistry at
the metal centres of the enterobactin metal complexes, as it
is known that the configuration of the coordinated iron
centre is crucial for the biological activity of the sidero-
phore.[10] In enterobactin, the configuration at the C3-sym-
metric metal centre is predetermined by the chiral cyclic tri-
lactone scaffold, and the corresponding D-FeIII complex is
formed stereoselectively.[11] However, the calculated energy
differences between the L and the D diastereomers of ferric
enterobactin are low, varying between 2.1 and 8.6 kJmolꢀ1
according to the level of theory,[12] which may be explained
by the high flexibility of the trilactone platform.
Although various enterobactin analogues based on C3-
symmetric oligo- or macrocyclic central units—such as 1,5,9-
triaminocyclododecane,[13] 2,6,10-triaminotrioxatricornan,[14]
cyclodextrin[15] or perhydrophenalene[16]—have been devel-
oped,[17] there are only a few examples of synthetic entero-
[a] Dr. ꢀ. Pintꢁr, Prof. Dr. G. Haberhauer
Institut fꢂr Organische Chemie, Fachbereich Chemie
Universitꢃt Duisburg-Essen, Universitꢃtsstrasse 7
45117 Essen (Germany)
Fax : (+49)2011834252
E-mail: gebhard.haberhauer@uni-due.de
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200801552.
Chem. Eur. J. 2008, 14, 11061 – 11068
ꢄ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
11061

bactin analogues that display diastereoselective metal com-
plex formation.[18,19] In contrast with naturally occurring en-
terobactin, these analogues are each made up of three chiral
binding arms attached to an achiral centre or achiral scaf-
fold.
Our intention was to develop an enterobactin analogue
consisting of a chiral scaffold and three achiral catecholate
arms that would exhibit predictable and highly diastereose-
lective metal complex formation. As we had already used
cyclic peptides containing imidazole and oxazole units in
their backbones for the control of axial and planar chirali-
ty[20,21] and for chirality transfer in C3-symmetric com-
pounds,[22,23] we decided also to use this type of chiral scaf-
fold for the design of a chiral synthetic enterobactin ana-
logue. Here we report the synthesis of the triscatecholate
ligand (SSS)-1 and an extended study of the diastereoselec-
tive complex formation of 1 with various metal ions.
Results and Discussion
Synthesis of the ligand: The synthetic pathway for ligand 1
is depicted in Scheme 1. Firstly, methyl- and benzyl-protect-
ed acid chlorides 6a and 6b were synthesized by known pro-
cedures.[24] Deprotection of the tris(phthalimidomethyl)oxa-
zole peptide scaffold 7[23] by hydrazinolysis, followed by acy-
lation with 6a and 6b, gave the corresponding protected li-
gands 8a and 8b, respectively. Deprotection of 8a was ac-
complished by treatment with boron tribromide in
dichloromethane, while 8b was debenzylated by catalytic
hydrogenolysis to afford the chiral triscatecholamide-type
ligand 1.
Scheme 1. Synthesis of ligand 1: i) MeOH, SOCl2, RT, 99%; ii) BnCl,
K2CO3, DMF, 1508C, 97%; iii) NaOH, MeOH/H2O, 1008C, then aq. HCl,
97%; iv) SOCl2, D, 96% for 6a and 99% for 6b; v) 1) N2H4·H2O, THF/
CH2Cl2/EtOH, RT; 2) 6a/6b, CH2Cl2, Et3N, RT, 99% for 8a and 61% for
8b. vi) BBr3, CH2Cl2, RT then MeOH, 96%; vii) H2/Pd(OH)2/C, MeOH,
96%.
room temperature was first investigated by electrospray ion-
ization mass spectroscopy. For this purpose, aqueous metha-
nolic solutions (MeOH/H2O 10:90) containing the ligand 1
and the appropriate metal salt in equal concentrations (5.0ꢅ
10ꢀ5 m) together with NaOH (10ꢀ2 m) were prepared and
measured with negative and positive ion detection. In both
modes the measured mass spectra show signals of highest in-
Analyses of the ligand 1, as well as of the two precursors
8a and 8b, by ESI-MS, NMR and UV/Vis spectroscopy are
1
in agreement with their structures. The H NMR spectrum
of 1 in [D4]MeOH confirms an ideal C3 symmetry with three
catecholate binding arms rotating rapidly on the NMR time-
scale, as the diastereotopic protons of the amidomethyl
groups show only a broad singlet at d=4.88 ppm. In the
tensity for the doubly charged ions [L6ꢀ+Mn+
and [L6ꢀ+Mn+
+ACHTUNGTRENNUNG
+
G
1
cases of the H NMR spectra of 8a and 8b in CDCl3 the
ions AlIII, GaIII, InIII, FeIII, ScIII, YIII, TiIV and GeIV the peaks
of the corresponding metal complexes were found, for the
other metal complexes no signals could be detected (Table 1
and Table 2). Interestingly, the exact masses of the dianion
peaks obtained for the InIII and ScIII complexes do not corre-
spond to the single protonated metal complexes but to the
fully deprotonated ligand with the metal cation lacking one
negative charge. This may be attributed to the formation of
radical anions under the conditions of negative ionization.
methylene signals are separated by 0.15 ppm as doublets of
doublets. The spectrum of 8b additionally reveals a small
anisotropy of 0.03 ppm for the methylene protons of the
ortho-benzyloxy protecting group. The methylene protons of
the meta-benzyloxy group show only a singlet. The restricted
rotation of the side arms is probably caused by intramolecu-
lar hydrogen bonds between the amide oxygens of the mac-
rocyclic scaffold and the catecholamide protons, which ex-
plains their low-field chemical shifts (d=8.69 ppm for 8a
and d=8.55 ppm for 8b). The hindered rotation of the
ortho-benzyloxy group is a consequence of steric hindrance
rather than intrastrand hydrogen bonding between ether
oxygen and the neighbouring catecholamide proton.
Investigation of complex formation by UV/Vis absorption
and CD spectroscopy: In a first step, the kinetics of the
metal ion binding were investigated by mixing the ligand 1
and the metal solutions in a molar ratio of 1:1.5 and simulta-
neous measuring UV absorption and CD spectra as a func-
tion of time. For the AlIII, InIII and FeIII ions, relatively slow
complex formation was found (Figure 1). To attain 95% of
the final UV absorption change (at 270 nm), 4 minutes
Investigation of metal complex formation by HR-ESI mass
spectroscopy: Complex formation by the triscatecholate
ligand 1 with various trivalent and tetravalent metal ions at
11062
www.chemeurj.org
ꢄ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 11061 – 11068

A Chiral Enterobactin Analogue
FULL PAPER
Table 1. Calculated and experimentally measured exact masses of the
complexes of 1 with trivalent metal ions recorded in the negative ion
mode.
Metal ion
Formula
Calcd mass
Exptl mass
Al3+
Ga3+
In3+
Fe3+
Ru3+
Rh3+
Cr3+
Sc3+
Y3+
[C48H46N9O15+Al]2ꢀ
[C48H46N9O15+Ga]2ꢀ
[C48H45N9O15+In]2ꢀ
[C48H46N9O15+Fe]2ꢀ
[C48H46N9O15+Ru]2ꢀ
[C48H46N9O15+Rh]2ꢀ
[C48H46N9O15+Cr]2ꢀ
[C48H45N9O15+Sc]2ꢀ
[C48H46N9O15+Y]2ꢀ
[C48H46N9O15+La]2ꢀ
507.6470
528.6190
551.1043
522.1237
545.1079
545.6084
520.1259
516.1303
538.6091
563.6089
507.6451
528.6145
551.1034
522.1254
–
–
–
516.1269
538.6026
–
La3+
Table 2. Calculated and experimentally measured exact masses of the
complexes of 1 with tetravalent metal ions recorded in the negative ion
mode.
Metal ion
Formula
Calcd mass
Exptl mass
Ti4+
[C48H45N9O15+Ti]2ꢀ
[C48H45N9O15+Zr]2ꢀ
[C48H45N9O15+Ge]2ꢀ
[C48H45N9O15+Sn]2ꢀ
517.6266
538.6047
530.6135
553.6040
517.6244
538.6024
530.6111
–
Zr4+
Ge4+
Sn4+
under the given conditions (MeOH/H2O 10:90, 0.10m tris,
0.02m HCl, pH 8.95, 208C) were required for AlIII, 30 mi-
nutes for FeIII and 40 minutes for InIII.
III
^
Figure 1. Complex formation kinetics of 1 with different metals ( : Al
,
However, the shapes of the corresponding CD kinetic
curves differ significantly. In the case of AlIII the same cut-
off value of the ellipticity change (at 265 nm) is obtained
after about 32 minutes, while the FeIII and InIII complexes
require more than 60 minutes to obtain equilibrium. The in-
congruity of the normalized UV absorption and CD kinetic
curves indicates continuous conformational changes of the
formed complexes after coordination of the metal centre
until the adoption of the final equilibrium conformation. In
the cases of GaIII, GeIV and TiIV ions, both complexation and
adoption of the equilibrium conformation occur within a
few seconds, and accordingly no time dependency—neither
for the extinction nor for the ellipticity values—could be ob-
served.
For the determination of thermodynamic parameters of
the metal complexes of 1, discontinuous titration experi-
ments were performed with an automated titration unit con-
nected to the spectropolarimeter. In the cases of RuIII, RhIII,
CrIII, ScIII, YIII, LaIII, ZrIV and SnIV, titration of 1 failed to
produce any detectable spectral changes, indicating that
there is no complex formation at constant basic pH value.
In contrast, the electronic spectrum of 1 showed gradual
changes if AlIII, GaIII, InIII, FeIII, GeIV or TiIV ions were used
as guests. The UV spectrum of 1 shows three main absorp-
tions, located at wavelengths around 220, 250 and 325 nm.
The oxazole scaffold and catecholamide side arms contrib-
ute to the absorption at the shortest wavelength, whereas
the other transitions can only be attributed to the newly
formed chromophore. If titration was conducted with GaIII
ions, all absorption bands showed a moderate bathochromic
III
~
&
: Fe
,
: InIII) ([1]=2.4ꢅ10ꢀ5 m, [M]/[L]=1.5, MeOH/H2O 10:90; 0.10m
tris; 0.02m HCl buffer, pH 8.95). Top) UV absorption measured at
270 nm. Bottom) Ellipticity measured at 265 nm.
shift, while changes in the CD spectrum were more striking
(Figure 2). With increasing metal/ligand ratio, two strong
positive Cotton effects emerge at 267 and 244 nm, accompa-
nied by a small negative one at 322 nm, which could not be
found in the spectrum of the uncomplexed ligand. The
strong positive bands represent an exciton coupling due to
the proximity of the helically coordinated catecholate arms
around the metal centre, whereas the sign of their elliptici-
ties implies a left-handed (L) orientation as the preferred
conformation.[19b]
The UV spectral changes at different wavelengths upon
addition of GaIII to 1 are shown in Figure 3. The same spec-
tral changes in the UV region could also be observed during
the titration of 1 with AlIII, FeIII and InIII, although longer
mixing times of up to 15 minutes were required after each
incremental addition of the titrant.
In the case of GeIV, only one narrow and strong positive
Cotton effect arose between 250 and 280 nm as a result of
complexation, and the final bathochromic shift of the ab-
sorption signals was the smallest among the observed com-
plexes. If titration of ligand 1 was performed with TiIV, a sig-
nificant bathochromic shift of all absorption bands and a
yellow colouration could be observed, but the CD spectra
showed no characteristic exciton coupling bands for octahe-
dral complex formation. These observations may be ex-
Chem. Eur. J. 2008, 14, 11061 – 11068
ꢄ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11063

ꢀ. Pintꢁr and G. Haberhauer
Table 3. CD spectra of the complexes of ligand 1 with metal ions Mn+
([1]=2.0ꢅ10ꢀ5
m
and 1/Mn+ =1:1) in MeOH/H2O (10:90; 0.10m tris,
0.02m HCl) at pH 8.95.
Mn+
De [molꢀ1 cmꢀ1] (l [nm])
– +4.8 (263)
–
ꢀ42.7 (234)
ꢀ66.6 (231)
ꢀ74.9 (232)
ꢀ45.2 (232)
ꢀ43.6 (233)
ꢀ19.3 (232)
Al3+
Ga3+
In3+
Fe3+
Ge4+
+86.7 (247)
+56.0 (266)
+52.7 (267)
+12.7 (264)
+15.9 (265)
+63.3 (262)
ꢀ9.5 (330)
ꢀ9.2 (328)
ꢀ2.5 (328)
ꢀ3.9 (329)
ꢀ7.0 (323)
+82.7 (247)
–
–
–
photoactive species predominate in each solution. Addition-
al Job plot analyses of the two-component systems of 1 with
AlIII, GaIII, FeIII and GeIV gave evidence of pure 1:1 stoichi-
ometry (Figure 4).
Figure 2. Spectrophotometric titration of ligand 1 with Ga3+ ([1]=2.0ꢅ
10ꢀ5 m, MeOH/H2O 10:90; 0.10m tris; 0.02m HCl buffer; pH 8.95).
Top) UV absorption spectra. Bottom) CD spectra.
Figure 4. UV absorption and CD Job plots for complexation of 1 with
ACTHNUTRGNEUNG
Ga3+ at different wavelengths ([1]+[GaCl3]=3.0ꢅ10ꢀ5 m, MeOH/H2O
10:90, 0.10m tris/0.02m HCl buffer, pH 8.95). Top) {y=Aobs.ꢀALꢀ
(AMꢀAL)ꢅx versus x=[GaCl3]/([GaCl3]+[1])}. Bottom) {y=qobs.ꢀqLꢀ
(qMꢀqL)ꢅx versus x=[GaCl3]/([GaCl3]+[1])}.
~
&
^
Figure 3. UV absorption titration curves ( : 355, : 315, : 270, ꢅ:
240 nm) for complexation of 1 and Ga3+ at different wavelengths ([1]=
2.0ꢅ10ꢀ5 m, MeOH/H2O 10:90; 0.10m tris; 0.02m HCl buffer; pH 8.95).
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
plained in terms of a low stability of the TiIV octahedral
complex and the formation of yellow polymeric TiIV/cate-
cholate compounds.[25] CD spectral data obtained for the
triscatecholate ligand 1 and for its metal complexes are sum-
marized in Table 3.
The affinities of ligand 1 for the metals under the condi-
tions employed were evaluated as virtual binding constants
according to Equation (1), which represents a simplified as-
sociation constant involving all protonation/deprotonation
and metal ion coordination steps. Although this constant is
only applicable for the pH value applied for the measure-
ment, it can provide information about the selectivity
toward different cationic guests.
With the exception of titration with TiIV ions, UV and CD
spectra each showed one set of isosbestic points during the
whole process of titration, which indicates that only two
11064
www.chemeurj.org
ꢄ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 11061 – 11068

A Chiral Enterobactin Analogue
FULL PAPER
½LMꢁ
In order to provide more exact information about the
sense of chirality at the metal centre in the complexes of 1,
absorption and CD spectra of the ironACTHNUTRGNEUNG(III) complex in the
virt
LM
ð1Þ
K
¼
ð½Lꢁtotꢀ½LMꢁÞð½Mꢁtotꢀ½LMꢁÞ
visible region were investigated (Figure 5). Upon coordina-
tion of Fe3+ ions a red-coloured complex is formed from 1
in basic solution. The broad absorption band at 509 nm (e=
5905 molꢀ1 dm3 cmꢀ1) is assigned to a ligand-to-metal charge
The virtual association constants of the complexation sys-
tems were calculated by nonlinear least-squares fitting ac-
cording to the modified Benesi–Hildebrand equation from
both UV absorption and CD data set.[26] The best reproduci-
ble results and the smallest errors could be obtained if ab-
sorption values between 265 and 275 nm or ellipticity values
in the range of the exciton coupling band were used for
evaluation. The obtained binding constants and free energy
changes of 1 with cations are summarized in Table 4.
transfer (LMCT) band, which is characteristic for ironACHTUNGTRENNUNG(III)
octahedrally coordinated by three catecholate units. The cor-
responding CD spectrum shows a positive Cotton effect at
longer wavelengths and a negative Cotton effect at shorter
wavelengths relative to the LMCT band. Comparison with
CD spectra of CrIII[3] and FeIII[28] enterobactin complexes
clearly indicates that octahedral metal complexes of ligand 1
occur as stereoisomers with helicity opposite to that ob-
served with naturally occurring enterobactin: namely L-fac.
The extent of diastereoselective complex formation
caused by the chiral macrocyclic scaffold of ligand 1 was fur-
ther investigated by ab initio calculations.[29] Full geometry
optimizations were performed for the (SSS,D)- and (SSS,L)-
complexes of 1 and for the D- and L-complexes of entero-
bactin by the DFT method (for a schematic representation
of the two possible conformers of the metal triscatecholate
complexes of 1 see Figure 6). In the case of the InIII com-
plexes the structures were calculated at the B3LYP/
Lan2LDZ level of theory; for all other complexes the
B3LYP/6-31G* approximation was used (Table 5). The
energy differences between the diastereomers of 1 vary be-
tween 51 and 55 kJmolꢀ1 in favour of the (SSS,L)-stereoiso-
mers, which accounts for a complete diastereoselectivity at
room temperature according to the Boltzmann–Gibbs distri-
bution. Interestingly, these energy differences are four to six
times larger than the values calculated for the corresponding
Table 4. Virtual binding constants (Ka) and overall free energy of com-
plexation (ꢀDG8) between 1 and cations as determined by UV titration
experiments (MeOH/H2O 10:90; 0.10m tris; 0.02m HCl buffer; pH 8.95,
208C).
Mn+
Ka [dm3 molꢀ1
]
ꢀDG8 [kJmolꢀ1
]
R2
Al3+
Ga3+
In3+
(2.64ꢂ0.28)ꢅ106[a]
(1.45ꢂ0.46)ꢅ106[a]
(1.10ꢂ0.10)ꢅ104[b]
(2.45ꢂ0.08)ꢅ104[c]
(2.50ꢂ0.26)ꢅ106[d]
36.0
34.6
22.7
24.6
35.9
0.99935
0.99651
0.99608
0.99944
0.99937
Fe3+
Ge4+
Association constants were determined: [a] at 270 nm. [b] At 273–
266 nm. [c] At 274–267 nm from UV absorption data set. [d] At 265–
258 nm from CD data set.
The relative order of stability of the trivalent metal com-
plexes with the macrocyclic ligand 1 is Al3+ > Ga3+
>
Fe3+ > In3+, suggesting that the maximum stability is ach-
ieved with the smallest cation. However, it is unlikely that
the observed trend in complex stability constants for the
metal ions can be entirely attributed to the size of the metal
cation radii, since the ionic radii of GaIII and FeIII (62 and
64.5 pm)[27] are quite similar.
Figure 5. UV/Vis and CD spectra of 1 with 0.0 (c), 0.5 (c) and 1.0 (c) equivalents of FeIII ([1]=1.25ꢅ10ꢀ4 m in MeOH/H2O 80:20; 0.10m tris
buffer).
Chem. Eur. J. 2008, 14, 11061 – 11068
ꢄ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11065

ꢀ. Pintꢁr and G. Haberhauer
Table 5. Relative energies of the conformers of 1·M(6ꢀn)ꢀ and
the oxazole rings from the reference planes, the 18-mem-
bered backbones of the modelled complexes of 1 adopt
deeper, bowl-like conformations, through turning of the oxa-
zole units inward when viewed from the direction of the
ꢀ
enterobactin·MACHTUNGTRENNUNG(6ꢀn) calculated (B3LYP) in kJmolꢀ1
.
Basis set
6-31G*
Isomer
(SSS,L)-1
(SSS,D)-1
L-enterobactin
D-enterobactin
AlIII
GaIII
InIII
TiIV
GeIV
U
0.0
52.4
9.2
0.0
54.9
7.9
0.0
53.1
9.6
0.0
3
E
50.9
12.6
0.0
metal centre. The identical vicinal JHN,CH values of 7.8–
7.9 Hz for the protected ligands 8a and 8b and for the pro-
tected platform 7 correspond to dihedral angles of 145 < V
< 1508,[30] which are in good agreement with the calculated
and X-ray structures of the unsubstituted C3-symmetric oxa-
zole analogue from a previous study.[31] In these systems the
dihedral angles c [Namide-Ca-Coxazole-Ooxazole] can be taken to
express the extent of deviation from planarity. In the case of
complete planarity, this angle is 1808. While the optimized
structures of unsubstituted oxazole cyclohexapeptides dis-
play cone angles of 160–1698, values of only 115–1178 for
the (SSS,L) isomers and 122–1238 for the (SSS,D) isomers
were obtained for the calculated complexes of 1. Conse-
quently, the preference for the L isomer in metal complexes
of 1 cannot be a result of conformational constraints that
exist within the 18-membered macrocycle. The reason for
the large energy gap between the isomers is probably the
distance between the carbonyl groups of the side chains and
the carbonyl groups of the macrocyclic scaffold. In the case
of the L isomers, the distance between the oxygen atoms of
these groups are almost 6 ꢆ, whereas in the case of the D
isomers it is calculated to be 3.48–3.78 ꢆ. This short distance
causes repulsive interactions between the side arms and the
scaffold, thus explaining the high energetic discrimination
between the isomers.
0.0
0.0
0.0
Lan2DZ
N
0.0
55.4
8.4
0.0
Conclusion
Figure 6. Top) Schematic representation of the two possible conformers
of the triscatecholate complexes of 1. Bottom) Molecular structures of
the diastereomers of the complex 1·Ga3ꢀ calculated at the B3LYP/6-31G*
level of theory; all hydrogen atoms have been omitted for clarity.
In summary, we have been able to show that a C3-symmetric
oxazole-containing macrocyclic peptidic scaffold can be
used for the construction of a siderophore ligand. Like the
naturally occurring enterobactin, this siderophore consists of
three achiral catecholate arms attached to a chiral backbone
and forms octahedral metal complexes with high diastereo-
selectivity. Although ironACTHNUTRGENNG(U III) and galliumACHTUNGTNER(NUGN III) have similar
charge/radius ratios and normally produce similar octahe-
dral complexes, the behaviour of the siderophore analogue
enterobactin complexes. On the assumption that the enthal-
py of hydrolysis does not differ significantly for the two pos-
sible helical isomers, this stereoselectivity must persist in so-
lution, too. This is in accordance with the results obtained
by the CD measurements.
The optimized structures of the energetically favoured
(SSS,L)-configured complexes of 1 show distorted octahe-
dral geometries, in which twist angles (608 for perfect octa-
hedral geometry) vary between 44 and 548. In addition, the
towards these ions is different. While ironACHTUGNTERN(UNG III) ions are
bonded slowly, the incorporation of GaIII proceeds within
seconds and the corresponding binding constant is higher by
two orders of magnitude. The often used postulation that
meta
ꢀ
calculated M O
bond lengths are 0.04–0.10 ꢆ shorter
the group 13 metal ion gallium
ACHTUNGTNER(NUNG III) can serve as an excellent
ortho
ꢀ
than those of the M O
bonds. This may be due to the
surrogate marker of the iron(III) failed in this case.
ACHTUNGTRENNUNG
considerable diameter of the cyclohexapeptide platform,
which is oversized in comparison with the only 12-mem-
bered and conformationally more flexible macrocyclic frame
of enterobactin. To minimize constraints in the octahedral
coordination sphere, complex formation is accompanied by
considerable changes in platform conformation in relation
to the structures of oxazole cyclohexapeptides without coor-
dinatively or covalently anchored side arms. While platforms
7, 8a and 8b exist in conformations with slight deviations of
Experimental Section
General remarks: All chemicals were reagent grade and were used as
purchased. Reactions were monitored by TLC analysis with silica gel
60 F254 thin-layer plates. Flash chromatography was carried out on silica
gel 60 (230–400 mesh). 1H and 13C NMR spectra were measured with
Bruker Avance DMX 300 and Avance DRX 500 spectrometers. All
chemical shifts (d) are given in ppm relative to TMS. The spectra were
11066
www.chemeurj.org
ꢄ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 11061 – 11068

A Chiral Enterobactin Analogue
FULL PAPER
referenced to deuterated solvents indicated in brackets in the analytical
data. HRMS spectra were recorded with a Bruker BioTOF III Instru-
ment. IR spectra were measured on a Varian 3100 FT-IR Excalibur
Series spectrometer. UV/Vis absorption spectra were obtained with a
Varian Cary 300 Bio instrument, whereas CD absorption spectra were
taken with a Jasco J-815 spectrophotometer fitted with a Jasco ATS-443
automatic titration unit.
0.10 mmol) in CH2Cl2/THF/EtOH 2:2:1 (25 mL), and stirring was contin-
ued at room temperature for 24 h. Volatiles were then removed in a
rotary evaporator, and the residue was dried in vacuo (<1 mbar, 508C,
2 h). The remaining solid was suspended in CH2Cl2 (30 mL), followed by
addition of the appropriate acid chloride (1.0 mmol) and triethylamine
(0.152 g, 1.5 mmol). After the system had been stirred for 6 h at room
temperature, the solvent was removed, and the residual solid was subject-
ed to column chromatography (CH2Cl2/EtOAc/MeOH 75:25:0 ! 75:25:5
for 8a, EtOAc/MeOH 100:0 ! 100:3 for 8b) to provide the product.
Abbreviations: PhtN: phthalimido. THF: tetrahydrofuran. Tris: tris(hy-
droxymethyl)aminomethane.
Data for 8a: Yield: 0.099 g (99.2%); TLC: Rf =0.55 (CH2Cl2/EtOAc/
Methyl 2,3-bis(benzyloxy)benzoate (4): A mixture of methyl ester 3
(1.345 g, 8.0 mmol), K2CO3 (3.428 g, 24.8 mmol), benzyl chloride (2.431 g,
19.2 mmol) and DMF (10 mL) was stirred at 1508C for 60 minutes. After
completion of the reaction, the mixture was cooled down, water
(100 mL) was added, and the product was extracted with CH2Cl2 (3ꢅ
50 mL). The organic layer was dried over MgSO4, filtered and evaporated
to give an oily product, which was further exhaustively dried in vacuo
(<1 mbar, 608C) to yield 4 (2.704 g, 97.0%) as a clear yellowish oil.
1H NMR (300 MHz, CDCl3): d=7.47–7.41 (m, 4H; BnO CH), 7.39–7.31
(m, 7H; BnO CH and Ph CH-6), 7.16 (dd, 3JH,H =8.0 Hz, 4JH,H =1.7 Hz,
3
MeOH 75:25:5; silica); 1H NMR (500 MHz, CDCl3): d=8.69 (t, JH,H
=
6.0 Hz, 1H; CONHCH2), 8.17 (d, 3JH,H =7.9 Hz, 1H, CONH), 7.67 (dd,
3JH,H =7.9 Hz, 4JH,H =1.6 Hz, 1H; Ph CH-6), 7.11 (t, 3JH,H =7.9 Hz, 1H;
Ph CH-5), 7.02 (dd, 3JH,H =7.9 Hz, 4JH,H =1.6 Hz, 1H, Ph CH-4), 5.087
3
(dd, 2JH,H =15.8 Hz, 3JH,H =6.3 Hz, 1H; CONHCH2), 5.087 (dd, JH,H
=
3
7.9 Hz, 3JH,H =4.7 Hz, 1H; Val a-CH), 4.94 (dd, 2JH,H =15.8 Hz, JH,H
=
6.3 Hz, 1H; CONHCH2), 3.86 (s, 6H; CH3O), 2.37–2.30 (m, 1H; Val b-
CH), 1.04 (d, 3JH,H =6.6 Hz, 3H; Val CH3), 1.01 ppm (d, 3JH,H =6.6 Hz,
3H; Val CH3); 13C NMR (125 MHz, CDCl3): d=165.2 (q; CONHCH2),
161.5 (q; oxazole C-2), 160.4 (q; CONH), 153.2 (q; oxazole C-5), 152.6
(q; Ph C-3), 147.7 (q; Ph C-2), 130.0 (q; oxazole C-4), 126.0 (q; Ph C-1),
124.2 (t; Ph CH-5), 122.8 (t; Ph CH-6), 115.7 (t; Ph CH-4), 61.4 (p;
CH3O-2), 56.0 (p; CH3O-3), 53.1 (p; Val a-CH), 34.2 (s; CONHCH2),
33.5 (t; Val b-CH), 18.4 (p; Val CH3), 18.2 ppm (p; Val CH3); UV/Vis
3
1H; CH-4), 7.09 (t, JH,H =8.0 Hz, 1H; CH-5), 5.15 (s, 2H; m-BnO CH2),
5.13 (s, 2H; o-BnO CH2), 3.86 ppm (s, 2H; CO2CH3); 13C NMR
(75 MHz, CDCl3): d=166.8 (q; CO2CH3), 152.8 (q; Ph C-2), 148.3 (q; Ph
C-3), 137.4 (q; o-BnO C-1), 136.6 (q; m-BnO C-1), 128.5 (t; o/m-BnO
CH-3,5), 128.2 (t; m-BnO CH-2,6), 128.0 (t; o-BnO CH-4), 127.9 (t; m-
BnO CH-4), 127.5 (t; o-BnO CH-2,6), 126.9 (q; Ph C-1), 123.9 (t; Ph
CH-5), 122.9 (t; Ph CH-6), 118.1 (t; Ph CH-4), 75.6 (s; o-BnO CH2), 71.3
(s; m-BnO CH2), 52.1 ppm (p; CO2CH3); IR (KBr): n˜ =3064, 3032, 2949,
2880, 1730, 1580, 1489, 1474, 1433, 1374, 1263, 1147, 1083, 979, 916, 861,
(CH2Cl2): lmax (e)=296 nm (8130 molꢀ1 dm3 cmꢀ1); CD (CH2Cl2):
l
(De)=250 nm (+8.6 molꢀ1 dm3 cmꢀ1); ESI-HRMS: m/z: calcd for
[C54H64N9O15]+: 1078.4516; found: 1078.4561.
Data for 8b: Yield: 0.094 g (61.2%). TLC: Rf =0.68 (CH2Cl2/EtOAc/
3
MeOH 75:25:3; silica); 1H NMR (500 MHz, CDCl3): d=8.55 (t, JH,H
=
783, 754, 697 cmꢀ1
; UV/Vis (CH2Cl2): lmax (e)=246 (3550), 294 nm
5.8 Hz, 1H; CONHCH2), 8.17 (d, 3JH,H =7.8 Hz, 1H; CONH), 7.71 (dd,
3JH,H =5.7 Hz, 4JH,H =3.9 Hz, 1H; Ph CH-6), 7.44–7.41 (dd, 3JH,H =8.0 Hz,
4JH,H =1.5 Hz, 2H; o-BnO CH-2,6), 7.39–7.33 (m, 3H; BnO CH-3,4,5),
7.22–7.20 (dd, 3JH,H =7.8 Hz, 4JH,H =1.8 Hz, 2H; m-BnO CH-2,6), 7.18–
7.12 (m, 5H; BnO CH-3,4,5, Ph CH-4,5), 5.12 (s, 2H; o-BnO CH2), 5.08
(d, 2JH,H =10.6 Hz, 1H; m-BnO CH2), 5.05 (d, 2JH,H =10.6 Hz, 1H; m-
(2630 molꢀ1 dm3 cmꢀ1); FAB-MS: m/z (%): 348.3 (35) [M]+, 371.3 (3)
[M+Na]+.
2,3-Bis(benzyloxy)benzoic acid (5b): Aqueous NaOH (2m, 6.0 mL,
12.0 mmol) was added to a solution of 4 (2.090 g, 6.0 mmol) in MeOH
(24 mL) and the mixture was stirred at 1008C for 60 min. After the mix-
ture had cooled down, water (100 mL) and HCl (2m, 15 mL, 30.0 mmol)
were added and the product was extracted with CH2Cl2 (3ꢅ50 mL). The
organic layer was dried over MgSO4, filtered and concentrated to give
2
3
BnO CH2), 5.01 (dd, JH,H =15.9 Hz, JH,H =6.3 Hz, 1H; CONHCH2), 4.99
(dd, 3JH,H =7.8 Hz, 3JH,H =4.6 Hz, 1H; Val a-CH), 4.86 (dd, JH,H
=
2
15.9 Hz, 3JH,H =5.4 Hz, 1H; CONHCH2), 2.32–2.26 (m, 1H; Val b-CH),
1
benzoic acid 5b (1.943 g, 96.9%) as a white powder. H NMR (300 MHz,
3
3
1.00 (d, JH,H =6.8 Hz, 3H; Val CH3), 0.97 ppm (d, JH,H =6.8 Hz, 3H; Val
CH3); 13C NMR (125 MHz, CDCl3): d=165.1 (q; CONHCH2), 161.5 (q;
oxazole C-2), 160.1 (q; CONH), 152.6 (q; oxazole C-5), 151.7 (q; Ph C-
2), 146.7 (q; Ph C-3), 136.3 (q; o-BnO C-1), 136.0 (q; m-BnO C-1), 129.8
(q; oxazole C-4), 128.55 (t; BnO CH), 128.51 (t; BnO CH), 128.36 (t;
BnO CH), 128.13 (t; BnO CH), 127.5 (t; BnO CH), 126.9 (q; Ph C-1),
124.3 (t; Ph CH-5), 123.2 (t; Ph CH-6), 117.3 (t; Ph CH-4), 76.1 (s; m-
BnO CH2), 71.2 (s; o-BnO CH2), 53.0 (t; Val a-CH), 34.1 (s;
CONHCH2), 33.4 (t; Val b-CH3), 18.3 (p; Val CH3), 18.1 ppm (p; Val
CH3); UV/Vis (CH2Cl2): lmax (e)=296 nm (8510 molꢀ1 dm3 cmꢀ1); CD
CDCl3): d=7.65 (dd, 3JH,H =7.8 Hz, 4JH,H =1.5 Hz, 1H; Ph CH-6), 7.42–
7.28 (m, 5H; BnO CH), 7.26 (s, 5H; BnO CH), 7.18 (dd, 3JH,H =7.8 Hz,
4JH,H =1.5 Hz, 1H; Ph CH-4), 7.10 (t, 3JH,H =7.8 Hz, 1H; Ph CH-5), 5.18
(s, 2H; m-BnO CH2), 5.11 ppm (s, 2H; o-BnO CH2); 13C NMR (75 MHz,
CDCl3): d=165.3 (q; CO2H), 151.3 (q; Ph C-2), 147.1 (q; Ph C-3), 135.8
(q; o-BnO C-1), 134.7 (q; m-BnO C-1), 129.2 (t; o/m-BnO CH-3,5),
128.79 (t; m-BnO CH-2,6), 128.77 (t; o-BnO CH-2,6), 128.5 (t; o-BnO
CH-4), 127.7 (t; m-BnO CH-4), 125.0 (t; Ph CH-5), 124.4 (t; Ph CH-6),
123.1 (q; Ph C-1), 119.0 (t; Ph CH-4), 77.1 (s; o-BnO CH2), 71.5 ppm (s;
m-BnO CH2); IR (KBr): n˜ =3442, 3063, 3032, 2876, 2675, 2976, 1694,
1599, 1577, 1498, 1474, 1455, 1415, 1378, 1313, 1262, 1220, 1036, 967, 767,
(CH2Cl2):
l
(De)=250 nm (+10.1 molꢀ1 dm3 cmꢀ1); ESI-HRMS: m/z:
calcd for [C90H88N9O15]+: 1535.6426; found: 1535.6483.
752, 698 cmꢀ1
;
UV/Vis (CH2Cl2): lmax (e)=246 (6310), 300 nm
(4470 molꢀ1 dm3 cmꢀ1); EI-MS: m/z (%): 334.2 (3) [M]+, 243.1 (4), 225.1
Ligand 1
(1), 181.2 (23), 91.1 (100), 65.2 (7).
Preparation from 8a:
A BBr3 solution (1m in CH2Cl2, 0.70 mL,
0.70 mmol) was added to a solution of 8a (0.054 g, 0.05 mmol) in CH2Cl2
(3 mL), and the mixture was stirred at room temperature for 16 h.
MeOH was then added, and solvents were evaporated in vacuo. The resi-
due was purified by column chromatography on silica gel (CH2Cl2/
MeOH/HCO2H 95:5:0.4) to give 1 (0.048 g, 96.4%) as a grey solid.
2,3-Bis(benzyloxy)benzoyl chloride (6b): 2,3-Bis(benzyloxy)benzoic acid
(0.669 g, 2.0 mmol) was stirred at reflux in thionyl chloride (10 mL) for
three hours, and volatiles were then evaporated in vacuo to yield acid
chloride 6b (0.701 g, 99.4%) as a yellowish oil. 1H NMR (300 MHz,
CDCl3): d=7.61 (dd, 3JH,H =8.0 Hz, 4JH,H =1.5 Hz, 1H; Ph CH-6), 7.49–
4
7.34 (m, 10H; o/m-BnO CH-2,3,4,5,6), 7.27 (dd, 3JH,H =8.0 Hz, JH,H
=
Preparation from 8b: A solution of 8b (0.031 g, 0.02 mmol) in MeOH
(20 mL) was hydrogenated under atmospheric pressure for 6 h in the
presence of Pearlmanꢇs catalyst (20% Pd(OH)2 on charcoal, 0.050 g).
The catalyst was then filtered off, and the solvents were evaporated to
3
1.5 Hz, 1H; Ph CH-4), 7.17 (t, JH,H =8.0 Hz, 1H; Ph CH-5), 5.19 (s, 2H;
m-BnO CH2), 5.16 ppm (s, 2H, o-BnO CH2); 13C NMR (75 MHz,
CDCl3): d=164.7 (q; COCl), 152.7 (q; Ph C-3), 148.0 (q; Ph C-2), 136.5
(q; m-BnO C-1), 136.0 (q; o-BnO, C-1), 128.79 (t; m-BnO CH-3,5),
128.65 (t; o-BnO CH-3,5), 128.32 (t; o-BnO CH-2,6), 128.26 (t; o-BnO
CH-4), 128.19 (t; m-BnO CH-4), 127.5 (t; m-BnO CH-2,6), 124.4 (t; Ph
CH-5), 124.0 (t; Ph CH-6), 119.8 (t; Ph CH-4), 75.7 (s; o-BnO CH2),
71.4 ppm (s; m-BnO CH2).
give pure 1 (0.019 g, 95.7%) as a white solid.
1H NMR (500 MHz, [D4]MeOH): d=7.16 (dd, 3JH,H =8.0 Hz, JH,H
4
=
1.4 Hz, 1H; Ph CH-6), 6.89 (dd, 3JH,H =8.0 Hz, 4JH,H =1.4 Hz, 1H; Ph
CH-4), 6.66 (t, 3JH,H =8.0 Hz, 1H; Ph CH-5), 5.10 (d, 3JH,H =4.8 Hz, 1H;
Val a-CH), 4.88 (s, 2H; CONHCH2), 2.32–2.25 (m, 1H; Val b-CH), 0.99
(d, 3JH,H =6.9 Hz, 3H; Val CH3), 0.92 ppm (d, 3JH,H =6.9 Hz, 3H; Val
CH3); 13C NMR (125 MHz, [D4]MeOH): d=171.5 (q; CONHCH2), 162.8
Methyl- and benzyl-protected precursors 8a and 8b: Hydrazine monohy-
drate (0.250 g, 5.0 mmol) was added to a solution of platform 7 (0.098 g,
Chem. Eur. J. 2008, 14, 11061 – 11068
ꢄ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11067

ꢀ. Pintꢁr and G. Haberhauer
(q; oxazole C-2), 162.0 (q; CONH), 154.2 (q; oxazole C-5), 150.4 (q; Ph
C-2), 147.4 (q; Ph C-3), 130.8 (q; oxazole C-4), 119.9 (t; Ph CH-4), 119.7
(t; Ph CH-5), 118.8 (t; Ph CH-6), 116.3 (q; Ph C-1), 54.5 (t; Val a-CH),
35.4 (s; CONHCH2), 34.7 (t; Val b-CH3), 18.7 (p; Val CH3), 18.5 ppm (p;
Val CH3); ESI-HRMS: m/z: calcd for [C48H52N9O15]+: 994.3577; found:
994.3521.
[16] B. Tse, Y. Kishi, J. Org. Chem. 1994, 59, 7807–7814.
[17] For some examples of C3-symmetric catecholate complexes see:
a) M. D. Pluth, R. G. Bergman, K. N. Raymond, J. Am. Chem. Soc.
2007, 129, 11459–11467; b) I. Janser, M. Albrecht, K. Hunger, S.
Burk, K. Rissanen, Eur. J. Inorg. Chem. 2006, 244–251; c) W. Cai,
S. W. Kwok, J. P. Taulane, M. Goodman, J. Am. Chem. Soc. 2004,
126, 15030–15031; d) M. Albrecht, I. Janser, J. Runsink, G. Raabe,
P. Weis, R. Frçhlich, Angew. Chem. 2004, 116, 6832–6836; Angew.
Chem. Int. Ed. 2004, 43, 6662–6666; e) M. Meyer, B. Kersting, R. E.
Powers, K. N. Raymond, Inorg. Chem. 1997, 36, 5179–5191; f) M.
Albrecht, R. Frçhlich, J. Am. Chem. Soc. 1997, 119, 1656–1661;
g) B. Kersting, M. Meyer, R. E. Powers, K. N. Raymond, J. Am.
Chem. Soc. 1996, 118, 7221–7222.
[18] B. Tse, Y. Kishi, J. Am. Chem. Soc. 1993, 115, 7892–7893.
[19] a) Y. Tor, J. Libman, A. Shanzer, S. Lifson, J. Am. Chem. Soc. 1987,
109, 6517–6518; b) Y. Tor, J. Libman, A. Shanzer, C. E. Felder, S.
Lifson, J. Am. Chem. Soc. 1992, 114, 6661–6671.
[20] G. Haberhauer, Angew. Chem. 2007, 119, 4476–4479; Angew. Chem.
Int. Ed. 2007, 46, 4397–4399.
[21] G. Haberhauer, Angew. Chem. 2008, 120, 3691–3694; Angew. Chem.
Int. Ed. 2008, 47, 3635–3638.
[22] G. Haberhauer, T. Oeser, F. Rominger, Chem. Commun. 2005,
2799–2801.
[23] ꢀ. Pintꢁr, G. Haberhauer, Eur. J. Org. Chem. 2008, 2375–2387.
[24] Synthesis of acid chloride 6a: M. E. Jung, S. Abrecht, J. Org. Chem.
1988, 53, 423–425; synthesis of acid chloride 6b: a) S. K. Sharma,
M. J. Miller, S. M. Payne, J. Med. Chem. 1989, 32, 357–367; b) R. A.
Gardner, R. Kinkade, C. Wang, O. Phanstiel, J. Org. Chem. 2004, 69,
3530–3537.
[25] D. N. Caulder, C. Brꢂckner, R. E. Powers, S. Kçnig, T. N. Parac,
J. A. Leary, K. N. Raymond, J. Am. Chem. Soc. 2001, 123, 8923–
8938.
UV absorption and CD spectrophotometric titrations: For the UV ab-
sorption and CD spectrophotometric titrations, titrant solution containing
only 1 ([1]=2.0ꢅ10ꢀ5 m in MeOH/H2O 10:90; tris/HCl buffer 0.10m/
0.02m at pH 8.95) and titrant solution containing both 1 and the appro-
priate metal salt ([1]=2.0ꢅ10ꢀ5 m, [Mn+]=3.0ꢅ10ꢀ4–5.0ꢅ10ꢀ4
m in
MeOH/H2O 10:90 in tris/HCl buffer 0.10m/0.02m at pH 8.95) were pre-
pared. Titrant solution (2500 mL) was placed in the 1 cm-pathlength
quartz cuvette containing a stirring magnet. Automatized titrations were
performed at 208C by addition of titrant solution in discrete steps
(10 mL) and recording of spectra after a suitable mixing time (3–15 min).
A manual titration experiment for Job plot analysis was performed with
ligand and metal salt solutions of the same concentration (3.0ꢅ10ꢀ5 m in
MeOH/H2O 10:90; tris/HCl buffer 0.10m/0.02m at pH 8.95). Ligand solu-
tion (2500 mL) was placed in the cuvette, and spectra were measured. In
each titration step an aliquot (100 ! 1000 mL) was removed from the
cuvette and the same volume of metal salt solution was added. After suf-
ficient equilibrating time, spectra were measured. Finally, spectra from
the pure metal salt solution were taken.
Acknowledgement
This work was generously supported by the Deutsche Forschungsgemein-
schaft. The authors thank Dr. Andreea Schuster for helpful discussions.
[26] C.-F. Chow, M. H. W. Lam, W-Y. Wong, Inorg. Chem. 2004, 43,
8387–8393.
[27] a) R. D. Shannon, C. T. Prewitt, Acta Crystallogr. Sect. B 1969, 25,
925–946; b) R. D. Shannon, Acta Crystallogr. Sect. A 1976, 32, 751–
767.
[28] a) S. Salama, J. D. Stong, J. B. Neilands, T. G. Spiro, Biochemistry
1978, 17, 3781–3785; b) M. E. Bluhm, S. S. Kim, E. A. Dertz, K. N.
Raymond, J. Am. Chem. Soc. 2002, 124, 2436–2437.
[1] a) I. G. OꢇBrien, F. Gibson, Biochim. Biophys. Acta 1970, 215, 393–
402; b) J. R. Pollack, J. B. Neilands, Biochem. Biophys. Res.
Commun. 1970, 38, 989–992.
[2] L. D. Loomis, K. N. Raymond, Inorg. Chem. 1991, 30, 906–911.
[3] S. S. Isied, G. Kuo, K. N. Raymond, J. Am. Chem. Soc. 1976, 98,
1763–1767.
[29] All computations were performed with the Gaussian 03 program-
package: Gaussian 03, Revision C.02, M. J. Frisch, G. W. Trucks,
H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A.
Montgomery, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam,
S. S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scal-
mani, N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y.
Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P.
Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gom-
perts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Po-
melli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Sal-
vador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Dan-
iels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Ragha-
vachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford,
J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Ko-
maromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y.
Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson,
W. Chen, M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc.,
Wallingford CT, 2004.
[4] a) D. J. Ecker, L. D. Loomis, M. E. Cass, K. N. Raymond, J. Am.
Chem. Soc. 1988, 110, 2457–2464; b) T. B. Karpishin, T. D. P. Stack,
K. N. Raymond, J. Am. Chem. Soc. 1993, 115, 6115–6125.
[5] L. D. Loomis, K. N. Raymond, Inorg. Chem. 1991, 30, 906–911.
[6] D. S. Plaha, H. J. Rogers, Biochim. Biophys. Acta 1983, 760, 246–
255.
[7] J. V. McArdle, S. R. Sofen, S. R. Cooper, K. N. Raymond, Inorg.
Chem. 1978, 17, 3075–3078.
[8] a) T. B. Karpishin, T. M. Dewey, K. N. Raymond, J. Am. Chem. Soc.
1993, 115, 1842–1851; b) T. B. Karpishin, K. N. Raymond, Angew.
Chem. 1992, 104, 486–488; Angew. Chem. Int. Ed. Engl. 1992, 31,
466–468.
[9] R. C. Hider, J. Silver, J. B. Neilands, I. E. G. Morrison, L. V. C. Rees,
FEBS Lett. 1979, 102, 325–328.
[10] J. B. Neilands, T. J. Erickson, W. H. Rastetter, J. Biol. Chem. 1981,
256, 3831–3832.
[11] K. N. Raymond, G. Mꢂller, B. F. Matzanke, Top. Curr. Chem. 1984,
123, 49–102.
[12] a) A. Shanzer, J. Libman, S. Lifson, C. E. Felder, J. Am. Chem. Soc.
1986, 108, 7609–7619; b) B. P. Hay, D. A. Dixon, R. Vargas, J.
Garza, K. N. Raymond, Inorg. Chem. 2001, 40, 3922–3935.
[13] E. J. Corey, S. D. Hurt, Tetrahedron Lett. 1977, 18, 3923–3924.
[14] a) M. Lofthagen, R. VernonClark, K. K. Baldridge, J. S. Siegel, J.
Org. Chem. 1992, 57, 61–69; b) M. Lofthagen, J. S. Siegel, M. Hack-
ett, Tetrahedron 1995, 51, 6195–6208.
[30] M. Karplus, J. Am. Chem. Soc. 1963, 85, 2870–2871.
[31] G. Haberhauer, E. Drosdow, T. Oeser, F. Rominger, Tetrahedron
2008, 64, 1853–1859.
Received: July 30, 2008
Revised: September 12, 2008
Published online: November 6, 2008
[15] A. W. Coleman, C. C. Lin, M. Miocque, Angew. Chem. 1992, 104,
1402–1404; Angew. Chem. Int. Ed. Engl. 1992, 31, 1381–1383.
11068
www.chemeurj.org
ꢄ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 11061 – 11068