Glycosiderophores: Synthesis of tris-hydroxamate siderophores based on a galactose or glycero central scaffold, Fe(III) complexation studies

Article abstract of DOI:10.1016/j.jinorgbio.2012.02.030

A series of five new hexadentate tris-hydroxamate ligands based on a d-galactose or a glycerol scaffold have been synthesized. Protonation and ferric complex formation constants have been determined from solution studies by potentiometric and spectrophotometric titrations. All ligands form 1:1 Fe:L complexes. The calculated pFe values at pH 7.4 span over the range 19.2-23.0 depending on the scaffold and on the length of the spacers between hydroxamate and central scaffold and on the N-methyl substitution. This new kind of artificial siderophores based on a glycoscaffold is of interest as it opens up an easy way to modulate the pFe.

Full text of DOI:10.1016/j.jinorgbio.2012.02.030

Journal of Inorganic Biochemistry 112 (2012) 59–67
Contents lists available at SciVerse ScienceDirect
Journal of Inorganic Biochemistry
journal homepage: www.elsevier.com/locate/jinorgbio
Glycosiderophores: Synthesis of tris-hydroxamate siderophores based on a galactose
or glycero central scaffold, Fe(III) complexation studies
Christelle Neff d,e, François Bellot d,e, Jenny-Birgitta Waern d,e, François Lambert a,b,c,1, Jérémy Brandel f,
,1
Guy Serratrice f, François Gaboriau g,h, Clotilde Policar a,b,c,
⁎
a
Ecole Normale Supérieure, Département de chimie, 24, rue Lhomond, 75005 Paris, France
Université Pierre et Marie Curie Paris 6, 4, Place Jussieu, 75005 Paris, France
CNRS, UMR7203, France
b
c
d
Institut de chimie moléculaire et des matériaux d'Orsay, Bât. 420, Université Paris-Sud 11, F-91405 Orsay cedex, France
e
CNRS, UMR8182, France
f
Université Joseph Fourier Grenoble, Département de Chimie Moléculaire, Laboratoire de Chimie Inorganique Rédox, UMR CNRS 5250, ICMG FR 2607, BP 53, F-38041 Grenoble Cedex 09, France
g
INSERM UMR 991, CHRU Pontchaillou, F-35033 Rennes cedex France
Université de Rennes1, F-35043 Rennes cedex, France
h
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of five new hexadentate tris-hydroxamate ligands based on a D-galactose or a glycerol scaffold have been
synthesized. Protonation and ferric complex formation constants have been determined from solution studies by
potentiometric and spectrophotometric titrations. All ligands form 1:1 Fe:L complexes. The calculated pFe values
at pH 7.4 span over the range 19.2–23.0 depending on the scaffold and on the length of the spacers between
hydroxamate and central scaffold and on the N-methyl substitution. This new kind of artificial siderophores
based on a glycoscaffold is of interest as it opens up an easy way to modulate the pFe.
Received 12 October 2011
Received in revised form 27 February 2012
Accepted 28 February 2012
Available online 7 March 2012
Keywords:
Siderophore
© 2012 Elsevier Inc. All rights reserved.
Tris-hydroxamate
Fe(III) complex
Glycoligand
Complexation constants
1. Introduction
variety of artificial siderophores have been synthesized for iron chelation
therapy purposes [7,8]. To perform iron-chelation therapy, a high associ-
Iron is required for growth of most living organisms. It is one of
the most abundant transition metals in the biosphere. While present
at Fe(II) oxidation level, it was bioavailable. However, when aerobic
conditions appeared on the Earth, its stable form became Fe(III)
which, in aqueous conditions, led to very insoluble Fe(III) hydroxide
(Ks=10−39). Hence, the concentration of Fe(III) – in the absence of
any ligand other than water – is limited to 10−10 M (at pH7, taking
into account soluble hydroxides) [1-5]. Microorganisms have evolved
to sequester Fe(III) from the surroundings by the release of low
molecular-weight chelating agents highly specific for Fe(III), named
siderophores. The corresponding complexes are soluble and can be effi-
ciently taken up by cells having appropriate receptors. Natural sidero-
phores are mainly hexadentate ligands, involving either hydroxamate
(desferrichrome) or catecholate (enterobactin) [6] and they can be
used in iron chelation therapy. It concerns not only iron overload
diseases, such as ß-thalassemia, but also cancer therapy [7]. A wide
ation constant is not necessarily the best choice, as depletion of iron from
metalloproteins leading to side toxicity has to be avoided [9-11]. There is
presently a need for series of siderophores with tunable association con-
stants and it is an interesting challenge on the chemical point of view.
We have recently shown that sugars are attractive platforms for the
synthesis of modular ligands. Such central structures are able to control
the geometry of the coordination sphere [12-20]. Moreover, a series of
artificial siderophores based on a methyl-α-D-glucopyranoside scaffold
have been recently published in the literature. They display three biden-
tate units (either catechol or hydroxamate, see L as an example in Fig. 1)
connected to a glucose unit by propyl chains and ester or amide func-
tions [21-26]. We report here the synthesis of three tris-hydroxamate
D-galactose centered glycoligands and two glycerol-based ligands,
designed as acyclic analogs of the glycosiderophores (see Schemes 1
and 2 and experimental part).
D-Galactose (GAL) has been chosen because the all cis sequence at
positions 3, 4 and 5 of the pyranose ring provides a three finger-claw
favorable to chelation [12,27,28]. The chelation core is connected to
the sugar scaffold through ether bonds that are expected to be stable
under physiological conditions. To investigate the effect of the length
of the chain between the hydroxamate moieties and the central scaffold,
⁎
Corresponding author at: Ecole Normale Supérieure, Département de chimie, 24,
rue Lhomond, 75005 Paris, France. Fax: +33 144323858.
E-mail address: clotilde.policar@ens.fr (C. Policar).
Tel.: +33 1 44 32 24 20; fax: +33 1 44 32 38 58.
1
0162-0134/$ – see front matter © 2012 Elsevier Inc. All rights reserved.
doi:10.1016/j.jinorgbio.2012.02.030

60
C. Neff et al. / Journal of Inorganic Biochemistry 112 (2012) 59–67
ozonolysis in basic methanol solution [31-33]. They were converted
into hydroxamic acid using NH2OHaq in THF [34]. To obtain the
N-methyl-hydroxamic moieties, the ester were hydrolyzed using LiOH
and N-methyl-O-benzylhydroxylamine [35,36] was coupled using
EDAC.HCl [37].
2.2. Ligand deprotonation constants
The ligands LH3 (with L=H1GAL, Me1GAL, Me3GAL, HGLY and
MeGLY) are trihydroxamic acids, and the three deprotonation constants
were determined by potentiometric titration. An example of potentio-
metric curve is shown in Fig. 1. Analysis of the titration data by the
HYPERQUAD software [38] yielded the ligand deprotonation constants
Kan (n=1–3) defined by Eqs. (1) and (2):
LHn
⇌
LHn−1 þ Hþ
ð1Þ
ð2Þ
h
i
with Kan ¼ ½LHn−1
ꢀ
Hþ =½LHnꢀ
ðn ¼ 1 – 3Þ
Fig. 1. Potentiometric titration curves for (a) 1.0×10−3 M Me3GALH3; (b) 0.9×10−3
M
Me3GALH3+0.85×10−3 M Fe3+. Conditions : I=0.1 M (NaClO4), T=298 K.
The values are reported in Table 1 as pKan along with those of an
analogue trihydroxamic acid having a glucose scaffold [23].
glycoligands were prepared with either a 1-carbon or a 3-carbon spacer
between the central galacto-scaffold and the hydroxamate moiety (resp.
Me1GALH3 and Me3GALH3). Moreover, in order to investigate a possible
effect of the central structure, analogs based on glycerol in place of
D-galactose have been prepared (see Schemes 1 and 2). The behavior
in solution of the five ligands towards Fe(III) is reported. The central
galacto-scaffold was shown not to impair the iron coordination and in-
terestingly, increasing the length of the spacer from 1 C to 3 C improved
significantly the pFe-value. This study shows first that ligands with tun-
able iron complexing ability are accessible by chemical design.
The acid dissociation constants for the ligands LH3 are in the nor-
mal range for hydroxamic acids. The large separation in the values,
greater than the statistical separation of log 3 (0.48) [39], suggests
that there is an intergroup interaction between the three hydroxamic
moities in contrast to L′H3 ligand. This could be due to the spacer be-
tween central sugar and hydroxamate, shorter in the LH3 ligands than
in L′H3, which makes greater the interaction between the three
hydroxamate groups. It should be noted that the average pKa value
for MeGLYH3, Me1GALH3 and Me3GALH3 containing N–Me groups,
are smaller than those of the other ligands as generally observed for
the N-substituted hydroxamic acids [40].
2. Results and discussion
2.3. Ferric complexation
2.1. Synthesis
The potentiometric titration curves for the FeIII–LH (Fig. 1 as an ex-
3
The five tris-hydroxamate galacto- (H1GALH3, Me1GALH3,
Me3GALH3) and glycero-based ligands (HGLYH3, MeGLYH3) were
synthesized as depicted in Scheme 2. From 1,2-O-ethylidene-α-D-
galactopyranose (synthesized as described in the experimental part)
or glycerol, three allyl moieties were introduced by a Williamson
ether synthesis [29,30]. Methyl ester functions were then obtained by
ample) systems exhibit a large jump at a (mol base/mol ligand)=3 indi-
cating the formation of the FeL species. Since the formation of FeIII–LH3
complex was complete at low pH, UV–visible spectrophotometric titra-
tions were monitored, from pH 0.3 to 9 for FeIII–Me3GALH3 system and
0.9 to 9 for the others. The changes of the UV–visible spectra as a function
of pH for the five FeIII–LH3 systems were similar and an example is
Scheme 1. Ligands LH3 synthezised and ligand L′H3 from ref. [23] used for comparison.

C. Neff et al. / Journal of Inorganic Biochemistry 112 (2012) 59–67
61
Scheme 2. Synthesis of the ligands LH3: (A) Glycosiderophores and (B) glycerol-based siderophores (a) NaH, Allyl bromide (n=1) or 5-bromopent-1-ene (n=3), anhydrous DMF ;
(b) O3, CH2Cl2, MeOH, NaOH ; (c) aq NH2OH, THF (d) LiOH, THF:H2O ; (e) EDAC.HCl, BnONHMe, H2O:DMF (f) H2, Pd/C (10%), MeOH. The numbers in italics indicate the labelling
used for NMR description in the experimental part. For n=3, the H/C adjacent to the O were labelled 9 and the two other positions were labelled Hspacer or Cspacer
.
shown in Fig. 2 (FeIII–Me3GALH3 system). When the pH rises from 0.3
(or 0.9) to 2.6, a reddening of the solution was observed with the growth
of a band centered at λmax =465 nm. This indicates that a single species
is formed. When the pH is further increased up to 6.0 an hypsochromic
shift in wavelength to 425 nm was recorded. The spectra exhibit an iso-
sbestic point (λ=472 nm for Fe–Me3GALH3) over the pH range pH of
2.6–6 indicating that an equilibrium is established between only two dif-
ferent metal complexes. From pH 6.0 to 8, the spectra did not vary any-
more and above pH 8.5 a decrease of absorbance was observed at all
wavelengths indicating the hydrolysis of the complex.
The data from the spectrophotometric titration were refined by
both the softwares Specfit [41-43] and Letagrop-Spefo [44,45]. The
Table 1
Ligand pKan (defined by Eq. (2)) valuesa for the five LH3 ligands and comparison with a synthetic siderophore L′H3.
b
Ligand/pKa
HGLYH3
H1GALH3
MeGLYH3
Me1GALH3
Me3GALH3
L′H3
pKa1
pKa2
pKa3
Average pKan
10.82 (3)
9.42 (3)
8.34 (3)
9.53
10.74 (2)
9.11 (3)
8.19 (3)
9.35
9.22 (2)
8.64 (1)
7.82 (2)
8.56
9.25 (2)
8.68 (1)
7.82 (2)
8.58
9.63 (4)
9.13 (4)
8.18 (4)
8.98
9.94 (1)
9.31 (1)
8.67 (1)
9.31
a
Conditions: I=0.1 M (NaClO4), T=298 K; values in parentheses are standard deviations.
From reference [23].
b

62
C. Neff et al. / Journal of Inorganic Biochemistry 112 (2012) 59–67
0.6
0.5
0.4
0.3
0.2
0.1
0.0
pH=7.7
pH=2.6
pH=0.3
400
500
600
700
300
wavelength (nm)
Fig. 2. UV–visible spectrophotometric titration of FeIII–Me3GALH3 as a function of pH.
pH range 0.3–7.7 ; [Fe3+]=1.75×10−4 M, [Me3GALH3]=1.95×10−4 M ; I=0.1 M,
NaClO4 ; T=25 °C.
Fig. 3. Species distribution curves for the ferric complexes of Me3GALH3. Conditions:
Me3GALH3/Fe=1/1, 2×10−4 M.
best fit gave the formation of the FeLH and FeLm species. The overall
stability constants, defined as indicated below (Eqs. (3) and (4)),
were calculated and are reported in Table 2 together with the spectral
parameters of the complexes. The distribution diagram of the species as
a function of pH is described in Fig. 3 (FeIII–Me3GALH3 as an example).
ligand L'H3 [49] are significantly lower (pKa range 0.94–2.3) than
those of the complexes described in this work.
In order to compare the FeIII complexation efficiency of the chelators
at physiological conditions, the pFeIII (−log[Fe3+]) values were
calculated at pH=7.4, [L]tot=10−5 M and [FeIII]tot=10−6 M. [Fe3+
can be expressed by Eq. (5):
]
mFe3þ þ lL3− þ hHþ⇌FemLlHh
ð3Þ
ð4Þ
h
i
ꢀ
ꢀ
ꢁ
ꢁ
Fe3þ ¼ ½FeLꢀ=ðβ110ð½LꢀÞ ¼ αL½FeLꢀ = αFeL ½Lꢀtot−½FeLꢀtot β110
tot
h
i
h
i h
i
m
l
h
βmlh ¼ ½FemLlHhꢀ= Fe3þ
L3− Hþ
¼ αL=ð9ðαFeLÞβ110
Þ
ð5Þ
where αL
and αFeL are the usual Ringbom's coefficients [50]:
The band at 465 nm is assigned to a LMCT for the bis(hydroxamate)
complex [FeLH,(H2O)2]+ by comparison to the spectrum observed for
the hydroxamate complexes described in the literature [23]. The spec-
tral parameters at pH 6.5 are in agreement with a tris(hydroxamate)
complex [FeL] [23]. IR-spectra of the ligand and the complexes showed
shifts for νC=O and νNO consistent with the coordination of hydroxa-
mate moieties (see exp. part.). It should be noted that the FeIII–LH3
systems form only bis- and tris(hydroxamate) complexes. This is in con-
trast to the hydroxamate complexes derived from desferrichrome and
desferrioxamine cited above for which mono(hydroxamate)complex
was also observed [46,47]. The deprotonation of the FeLH complex
reflects the deprotonation of the free hydroxamate group in order to
form the tris(hydroxamate) FeL complex. The corresponding constants,
as pKa values, calculated from log β111–log β110, are 4.26, 4.50, 3.80, 3.71
and 3.47 for FeHGLY, FeH1GAL, FeMeGLY, FeMe1GAL and FeMe3GAL
respectively. It should be noted that the deprotonation constants of
natural trihydroxamate systems, the ferrichrome [46,47], of the
ferrioxamine B [46-48] and of the synthetic tripodal trihydroxamate
h
i
h
i
h
i
αL ¼ 1 þ Hþ =Ka1 þ Hþ 2=Ka1Ka2 þ Hþ 3=Ka1Ka2Ka3 and αFeL
h
i
¼ 1 þ Hþ =KaFeLH
The pFeIII values are reported in Table 2. An increase of pFe by about
one log unit is observed for the ligands MeGLYH3 and Me1GALH3
containing methyl-nitrogen hydroxamate in comparison to the ligand
HGLYH3 and H1GALH3, according to the electron donating effect of
the methyl moiety which leads to a more stable complex. The table
also indicates that the ligands HGLYH3, H1GALH3, MeGLYH3 and
Me1GALH3 are significantly less effective at physiological conditions
than Me3GALH3 and other synthetic (27.1 for L'H3) or natural (26.6
for desferrioxamine) [47] siderophores.. This effect could be related to
the geometrical requirements of the FeIII cation whose octahedral
coordination sphere is constrained by the ligands LH3 with nC-spacer =1
Table 2
Thermodynamic and spectral parameters for FeIII–LH3 complexesa.
b
Ligand
HGLYH3
H1GALH3
MeGLYH3
Me1GALH3
Me3GALH3
L′H3
log β110
24.7 (1)
2270
425
28.96 (4)
1130
470
24.3 (1)
2300
425
28.8 (1)
1500
460
23.18 (7)
2400
425
26.98 (4)
1680
465
23.02 (3)
2670
425
26.73 (2)
1800
465
26.90 (4)
2410
425
30.37 (2)
2010
465
31.86 (11)
2900
416
36.55 (5)
2100
465
ε (M−1 cm−1
λmax (nm)
log β111
)
)
ε (M−1 cm−1
λmax (nm)
pFec
19.2
19.3
20.5
20.3
23.0
27.1
a
Conditions : I=0.1 M (NaClO4), T=298 K ; values in parentheses are standard deviations ; both Specfit and Letagrop-Spefo gave the same values.
b
c
From ref. [23].
log [Fe3+] calculated for pH=7.4, [LH3]tot =10−5 M and [Fe3+ tot =10−6 M.
]

C. Neff et al. / Journal of Inorganic Biochemistry 112 (2012) 59–67
63
(ie H1GALH , Me1GALH3, HGLYH3, MeGLYH3) due to the shortness of
charring with 5% ethanolic H2SO4. Ozone was generated by passing
dry oxygen through an ozonator (Pacific ozone technology).
3
the spacers compared with other ligands. The lengthening of the spacer
leads to a significative increase of the pFe to 23.0 for Me3GALH3. In the
case of Me1GALH3 (resp. Me3GALH3) the length of the spacers between
two viccinal hydroxamate moieties is 6 and 7 C (resp. 10 and 11 C).
This improvement in the pFe is consistent with a strain-release upon
increasing the length of the spacers. As a comparison, in the natural
siderophore desferrioxamine, the vicinal hydroxamate groups are sepa-
rated by a spacer of 9 atoms (—(CH2)2CONH(CH2)5—) and in the case of
L'H3, they are separated by a spacer of 12 atoms (see Scheme 1,
—(CH2)3COO(CH)2COO(CH2)3—). The effect of the lengthening of the
spacers observed here can be understood as a competition between
two opposite contributions.[51] Because of entropic effect, an increase
in the length of the spacers between two chelating claws should be
unfavorable to chelation: upon chelation, there is an unfavorable entro-
pic loss that will be increased in the case of a more flexible ligand. But
also, such an increase in the length of the spacers releases strain induced
upon chelation in the case of the shorter spacers.
NMR spectra were recorded at room temperature with Bruker DX
250, DX 300, AC 360 MHz. 1H, 13 C and DEPT were always run. 2D
(Cosy, HSQC) were needed to interpret some of the spectra. The labels
of H and C are given in Scheme 2; H* and C* indicate that we observe
two signals because of the two diastereoisomers at the C7. Infra-red
spectra have been recorded on a Bruker IFS 66. Elemental analyses
were performed at the CNRS (Gif sur Yvette, France). MS spectra
were recorded in the positive mode on a Finnigan MAT 95 S at first
and then TSQ Quantum Access using electrospray ionisation. UV spectra
were recorded with a spectrophotometer Uvikon 860 XL.
For the protonation and iron-association constants studies, all solu-
tions were prepared with boiled, deionized and argon flushed water in
order to prevent the presence of CO2 and O2. The ionic strength was main-
tained at 0.1 M with NaClO4 (Prolabo, puriss or Merck, p.a.). Fe(III) stock
solutions for titrations were prepared by dissolving appropriate amounts
of ferric perchlorate hydrate (Aldrich) in 0.1 M HClO4 (prepared from
standardized 1 M HClO4, Riedel de Haël) solutions. The solutions were
standardized for ferric ion spectrophotometrically by using a molar ex-
tinction coefficient of 4160 M−1 cm−1 at 240 nm [57].
The pFe value for Me3GALH3 is slighly lower than that of the
desferrichrome (25.2), and of the same order of magnitude of the trans-
ferrin, 23.6 [52,53]. It should be noted that the new orally active iron tri-
dentate chelator ICL670 [54] developed for the treatment of iron
overload (in phase 3 clinical trial) exhibits a pFe of 21.5 at pH 7.4 [55].
4.2. Syntheses of the ligands
3. Conclusion
4.2.1. Syntheses of the glycerol-based ligands (see Scheme 2B)
In conclusion, a series of hexadentate tris-hydroxamate ligands have
been synthesized, with hydroxamate moieties connected to a central
D-galactose with ether bonds. The nature of the Lewis chelating donors
has been varied (hydroxamate and N-methyl-hydroxamate), and also
the length between the central sugar scaffold and the chelating moities.
To investigate the effect of the sugar central scaffold, acyclic analogs
based on a glycerol central scaffold have been prepared and studied.
Their iron-chelation efficiency has been evaluated and reported as pFe
values. They exhibit a range of tunable pFe values at pH 7.4 between
19.2 and 23.0 depending on the structure of the ligand. Interestingly,
the ligands based on a glycoscaffold show a pFe similar to that of their
respective acyclic analog, showing that the constraints imposed by the
glycoscaffold did not impair coordination to iron. Interestingly, the
comparision between Me1GALH3 and Me3GALH3 indicates that, in
that case, increasing the length of the spacer between the central scaf-
fold improved the pFe. This study shows that central glycoscaffold is ap-
propriate for the design of polydentate chelators and it opens up the
possibility of tuning of the pFe through the modulation of the ligand
which could be valuable in a context of iron-chelation therapy, as
long-term treatments with high affinity siderophore may lead to side
toxicity due to iron deficiency due to demetallation of endogenous
iron proteins [56].
4.2.1.1. 1,2,3-tri-O-allylglycerol (1′). Sodium hydride (60% dispersion in
oil, 5.6 g, 142.21 mmol) was added portionwise to a stirred solution of
glycerol (2.5 g, 27.14 mmol) at 0 °C in 50 mL of dry DMF. After 30 min
allyl bromide (8.5 mL, 97.70 mmol) was added dropwise; the tempera-
ture was raised to 50 °C and heated overnight. The reaction mixture was
cooled to zero degree and the excess sodium hydride was quenched by
slow addition of water. The mixture was evaporated to dryness; the
residue was dissolved in CH2Cl2 (200 mL) and washed with water
(3×100 mL). The organic phase was dried (Na2SO4), filtered and con-
centrated. The residual oil was then purified by flash chromatography
(PEt/ EtOAc 8:2) yielding the tri-ether 1′ as a colourless oil (2.785 g,
48%). Rv=0.55 (PEt/ EtOAc 8:2), PEt for petroleum ether.
MS (ESI): 235.1 (100%) [M+Na]+
δ=3.54 (m, 4H, 2×CHCH2O), 3.69 (m, 1H, OCHCH2), 4.01 (d, 4H,
vic =5.54 Hz, 2×OCH2CHCH2), 4.16 (d, 2H, Jvic =5.60 Hz, OCH2CHCH2),
.
1H NMR (250 MHz, CDCl3):
J
5.20 (m, 4H, 2×OCH2CHCH2), 5.31 (m, 2H, OCH2CHCH2), 5.9 (bm, 3H,
3×OCH2CHCH2). 13 C NMR (62.5 MHz, CDCl3): δ=70.2 (2×CHCH2O),
71.3 (OCH2CHCH2), 72.3 (2×OCH2CHCH2), 77.0 (OCHCH2), 116.9
(3×OCH2CHCH2), 134.7 (2×OCH2CHCH2), 135.2 (OCH2CHCH2).
4.2.1.2. 1,2,3-tri-O-((2-oxo-2-methoxy)-ethyl)-glycerol ether (2′).
(adapted from [58])
Compound 1′ (500 mg, 2.35 mmol) was dissolved in 30 mL dry
CH2Cl2 and added to a solution of NaOH in MeOH (2.5 M). The mixture
was cooled at −78 °C in an acetone-dry ice bath and flushed with O3
until the solution became blue (ca. 4 h). Then water was added and the
aqueous phase was extracted with 3×100mL CH2Cl2. The organic phase
was dried (Na2SO4), filtered and concentrated, then the residual oil was
purified by flash chromatography (PEt/ EtOAc 7:3) yielding the tri-ester
2′ as a colourless oil (316 mg, 47%). Rv=0.45 (PEt/EtOAc 8:2).
4. Experimental
4.1. General experimental conditions
For the synthesis, all solvents were dried over standard drying
agents and freshly distilled prior to use. Methanol was distilled on
magnesium, THF and pyridine on sodium with benzoquinone as an
indicator, DMF and dichloromethane and acetonitrile on calcium
hydride. Anhydrous acetone was obtained by distillation with 4 Å
molecular sieves. Trichloroacetonitrile (bp=83–84 °C) was distilled
at atmospheric pression under argon using calcium hydride and heating
at 90 °C the oil bath. The dried trichloroacetonitrile was kept in a flask
with molecular sieves. Trimethylsilyl-trifluoromethanesulfonate was
distilled under vacuum (P=15–20 mmHg). Thin layer chromatography
(TLC) was performed on SiO2 thin plates. Column chromatography was
performed on silica gel 60 A.C.C. (35–70 μm). Reactions were monitored
by TLC on silica gel 60 F254 plates with detection at 254 nm and/or by
MS (ESI): 331.1 (100%) [M+Na]+ 1H NMR (250 MHz, CDCl3):
.
δ=3.72 (m, 5H, 2×OCHCH2O, OCHCH2O), 3.73 (s, 9H, 3×CH3O), 4.14
(s, 4H, 2×OCH2CO), 4.35 (s, 2H, OCH2CO). 13 C NMR (62.5 MHz,
CDCl3): δ=51.8 (3×CH3O), 67.9 (OCH2CO), 68.6 (2×OCH2CO), 71.8
(2×CHCH2O), 78.5 (CHCH2O), 170.7 (2×CO), 171.1 (CO).
4.2.1.3. 1,2,3-tri-O-[((N-hydroxy)-carbamoyl)-methyl]-glycerol ether
(HGLYH3). (Adapted from [34]).
The tri-ester 2′ (100 mg, 0.325 mmol) was dissolved in 4 mL THF
and 50 mL of aq. hydroxylamine (c=16.4 M) was added. The mixture

64
C. Neff et al. / Journal of Inorganic Biochemistry 112 (2012) 59–67
was stirred at room temperature overnight. Then 50 mL of water and
50 mL of ethylacetate were added and the aqueous phase was washed
3 times with ethylacetate and then evaporated. The white sticky oil
was lyophilised to obtain a white powder (139 mg, 84%).
MS (CI): 312.6 (70%) [M+H+]. IR (ATR) ν (cm−1): 3172.4 (νOH),
1670.5 (νC=O) and 1016.4 (νNO). 1H NMR (300 MHz, D2O): δ=3.52
(m, 4H, Jgem=8.9 Hz, 2×OCHCH2O), 3.77 (s, 4H, 2×CH2CO), 3.94 (s,
2H, CH2CO), 4.06 (s, 1H, CHCH2O). 13 C NMR (75 MHz, D2O): δ=67.6
(CH2CO), 69,0 (2×CH2CO), 70.7 (2×CHCH2O), 78.4 (CHCH2O), 168,6
(3×CO). HRMS-ES+: m/z calc. for C9H17O9N3Na 334.0862, found
334.0873. Anal. Calcd for C9H17O9N3, H2O: C, 32.83; H, 5.82. Found: C,
32.23; H, 5.84.
5 mg of compound HGLYH3 was dissolved in less than 1 mL of
MeOH. Then 1 eq. of FeCl3 was added. The formation of the complex
was evidenced by the formation of a purple solid. IR spectrum was
recorded on the solid using ATR. IR (ATR) ν (cm−1): 1654.3 (νC=O);
1020.2 (νNO).
recorded on solid using ATR. IR (ATR) ν (cm−1): 1641.3 (νC=O);
1097.2 (νNO).
4.2.2. Syntheses of the galacto-based ligands (see Scheme 2A)
4.2.2.1. 1,2-O-ethylidene-3,4,6-tri-O-acetyl-α-D-galactopyranose. (adapted
from [59] and [29])
A mixture of 2,3,4,6-tetra-O-acetyl-α-D-galactopyranosylbromide
(25 g, 61 mmol, prepared by a standard method [60] from D-galactose)
sodium borohydride (13.84 g, 365 mmol), tetrabutyl-ammonium
iodide (11.3 g, 30.6 mmol) and anhydrous acetonitrile (200 mL) was
stirred overnight at room temperature. Then, water was added at 0 °C.
The mixture was diluted with CH2Cl2 (300 mL) and washed with
water (3×200 mL). The organic phase was dried (Na2SO4), filtered
and concentrated. The residual salts were filtered off using Celite®
washed with EtOAc and the residue was purified on silica gel using
PEt/EtOAc (8:2) as eluant to give the expected compound as a colour-
less syrup (14.45 g, 71%). Rv =0.35 (PEt/EtOAc 5:5).
MS (ESI): 355.1 (100%) [M+Na]+ 1H NMR (250 MHz, CDCl3):
.
4.2.1.4. 1,2,3-tri-O-(methylcarboxy)-glycerol ether (4′). Hydrate lithium
hydroxide (932 mg, 22.23 mmol) was added to a solution of 2′ in
43.6 mL THF:H2O (4:2). The reaction mixture was stirred for 2 h at
room temperature and then quenched with 1 M HCl to pH=2. The
solution was diluted with water (20 mL) and washed with ethyl acetate
(60 mL). The evaporation of the aqueous phase gave 4′ as a colourless
oil (1.266 g, quant.).
δ=1.24, 1.30 (2×d, 3H, CH*3CH, J=4.9 Hz), [1.97, 2.03, 2.12 (3×s,
9H, 3×CH3CO)], 3.88–4.24 (m, 4H, H2, H5, H6), 5.00 (m, 1H, H3),
5.21 (q, 0.5 H, CH3CH*, J=4.9 Hz), 5.42 (m, 1H, H4), 5.49 (q, 0.5 H,
CH3CH*, J=4.9 Hz), 5.55 (2×d, 1H, H1*, J1–2 =4.4 Hz). 13 C
(62.5 MHz, CDCl3): δ=[20.6, 20.7, 20.9 (3×CH3CO, CH3CH)], 61.6
(C6), 65.9 (C4), 69.1 (C3), 73.6 (C5), 74.8 (C2), 97.3, 98.0 (2×C1*),
[99.9, 100.8 (2×CH3C*H)], 169.8, 170.1, 170.4 (3×CO).
1H NMR (250 MHz, CD3OD): δ=3.76 (m, 4H, 2×OCHCH2O), 3.89
(m, 1H,OCHCH2O), 4.33 (s, 4H, 2×OCH2CO), 4.67 (s, 2H, OCH2CO).
13 C NMR (62.5 MHz, CD3OD): δ=66.4 (OCH2CO), 67.7 (2×OCH2CO),
70.5 (2×OCHCH2O), 77.6 (OCHCH2O), 172.8 (2CO), 173.3 (CO).
4.2.2.2. 1,2-O-ethylidene-α-D-galactopyranose (1).
3,4,6-tri-O-acetyl-1,2-O-ethylidene-α-D-galactopyranose (11 g,
34 mmol) was dissolved in dry MeOH (150 mL) and a catalytic
amount of sodium was added carefully at 0 °C. Then the reaction mixture
was stirred at room temperature for 3 h. Dowex® (H+, 50WX2-100) was
added portionwise to acidic pH. After filtration of Dowex® the solvent
was evaporated to give 1 as a colourless oil (7.2 g, 100%).
4.2.1.5. 1,2,3-tri-O-[((N-methyl-N-hydroxybenzyl)-carbamoyl)-methyl]-
glycerol ether (5′). (Adapted from [37]).
N-methyl-O-benzylhydroxylamine (500 mg, 3.6 mmol) [35,36] was
dissolved in a solution of 4′ (300 mg, 1.1 mmol) in 11 mL H2O:DMF
(1:1). Then EDAC.HCl was added portionwise and the reaction mixture
was stirred for 2 h at room temperature. The solution was diluted with
50 mL ethylacetate and the organic phase was washed with a solution
NaHCO3sat. (50 mL), water (50 mL), 1 M HCl (50 mL), brine (50 mL).
The organic extract was dried (Na2SO4), filtered and concentrated to
afford a colourless oil (505 mg, 73%) Rv=0.67 (PEt/EtOAc (7: 3)).
MS (ESI): 349 (100%) [M+Na]+ 1H NMR (360 MHz, CD3OD): δ=
.
[1.34, 1.42 (2×d, 3H, H8*,, J7–8 =4.9 Hz)], 3.75–4.18 (m, 6H, H2, H3,
H4, H5, H6), [5.10, 5.38 (2×q, 1H, H7*, J7–8 =4.9 Hz)], [5.43, 5.47
(2×d, 1H, H1*, J1–2 =4.7 Hz)].13 C NMR (90 MHz, CD3OD): δ=[19.9,
20.1 (2×C8*)], 60.9 (C6), 72.0 (C4), 73.5 (C3), 77.2 (C5), 79.8 (C2),
[97.6, 98.1 (2×C1*)], [98.9, 99.8 (2×C7*)].
MS (ESI): 646.2 (100%) [M+Na]+ 1H NMR (300 MHz, CDCl3):
,
4.2.2.3. 1,2-O-ethylidene-3,4,6-tri-O-allyl-α-D-galactopyranose (2).
Sodium hydride (60% dispersion in oil, 2.5 g, 63 mmol) was added
portionwise to a well stirred solution of 1 (3.7 g, 18 mmol) at 0 °C in
60 mL dry DMF. After 30 min. allyl bromide (6 mL, 70.2 mmol) was
added dropwise and the temperature was raised to 50 °C overnight. The
reaction flask was placed in an ice bath and the excess sodium hydride
was quenched by the slow addition of water. The mixture was evaporated
to dryness. The residue was dissolved in CH2Cl2 (200 mL) and washed
with water (3×100 mL). The organic phase was dried (Na2SO4), filtered
and concentrated. The residual oil was purified by flash chromatography
(PEt/ EtOAc 8:2) yielding the desired tri-allyl compound as a colourless oil
(4.5 g, 76%). Rv=0.8 (PEt/ EtOAc 1:1).
δ=[3.16, 3.18 (2× s, 9H, 3×NCH3], 3.72 (m, 4H, 2×OCHCH2O), 3.86
(m, 1H, OCHCH2O), 4.27 (s, 4H, 2×OCH2CO), 4.54 (s, 2H, OCH2CO),
[4.82, 4.83 (2× s, 6H, 3× CH2Bn)], 7.39 (s, 15H, HBn). 13 C NMR
(75 MHz, CDCl3): δ=33.7 (3×CH3N), 67.8 (2×OCHCH2O), [68.6, 69.1
(3×CH2Bn)], 71.7 (3×OCH2CO), 76.6 (OCH(CH2O)2), [129.4, 134.2
(CBn)], 171.7 (3×CO).
4.2.1.6. 1,2,3-tri-O-[((N-methyl-N-hydroxy)-carbamoyl)-methyl]-glycerol
ether (MeGLYH3). The protected tri-hydroxamic acid 5′ (612 mg,
1 mmol) dissolved in 100 mL MeOH was hydrogenated over 10% Pd/C
(61 mg) for 4 h. The palladium was removed by filtration on celite®
washed with MeOH. The residue was then evaporated to afford a pale
orange sticky oil (348 mg, quant.).
MS (ESI): 349.2 (100%) [M+Na]+ 1H NMR (250 MHz, CDCl3): δ=
[1.32, 1.38 (2×d, 3H, H8*, J7–8 =4.9 Hz)], [3.75–4.37 (m, 12H, H2, H3,
H4, H5, H6, 3×H9)], [5.12–5.30 (m, 6H, 3×H11)], 5.36 (2×q, 1H, H7*,
J7–8 =4.9 Hz), [5.53 (2×d, 1H, H1*, J1–2 =4.5 Hz)], 5.81–5.95 (m, 3H,
3×H10).13 C NMR (62.5 MHz, CDCl3): δ=[20.1, 20.2 (2×C8*)], 67.9
(C6), 70.3 (C9), 72.2 (C4), 72.3, 72.9 (2×C9), 74.1 (C3), 77.5 (C5),
80.5 (C2), [98.1, 98.2 (2×C1*)], [99.1, 99.5 (2×C7*)], [116.1, 117.1
(3×C11)], [135.0, 134.5, 135.0 (3×C10)].
MS (ESI): 354.0 (20%) [M+Na]+, 375.9 (100%) [M+Na]+ IR (ATR):
ν (cm−1) 3149.3 (νOH), 1662.4 (νC=O), 1097.1 (νNO). 1H NMR
(250 MHz, D2O): δ 3.10 (s, 9H, 3×NCH3), 3.64 (m, 5H, 2×OCHCH2O,
OCHCH2O), 4.30 (s, 4H, 2×OCH2CO), 4.43 (s, 2H, OCH2CO). 13 C NMR
(62.5 MHz, D2O):
δ 35.9 (3×NCH3), 66.7 (OCH2CO), 67.8 (2×
OCH2CO), 70.5 (2×OCHCH2O), 77.8 (OCHCH2O), 170.6 (3×CO).
HRMS-ES+: m/z calc. for C12H23O9N3Na 376.1327, found 376.1326.
Anal. Calcd for C12H23O9N3, H2O: C, 38.81; H, 6.79. Found: C, 38.53; H,
6.99. IR complex (NaCl): ν (cm−1) 1641.36 (νC=O).
4.2.2.4. 1,2-O-ethylidene-3,4,6-tri-O-((2-oxo-2-methoxy)-ethyl)-α-D-
galactopyranose (3). (adapted from [58])
IR spectrum of the Fe(III) complex was recorded on a powder
obtained as previously mentioned (see HGLYH3) IR spectrum was
Compound 5 (500 mg, 1.53 mmol) was dissolved in 20 mL dry
CH2Cl2 and added to a solution of NaOH in MeOH (2.5 M). The mixture

C. Neff et al. / Journal of Inorganic Biochemistry 112 (2012) 59–67
65
was cooled at −78 °C in a acetone-dry ice bath and flushed with O3
until the solution became blue. Then water was added and the aqueous
phase was extracted with 3×100mL CH2Cl2. The organic phase was
dried (Na2SO4), filtered and concentrated, then the residual oil was
purified by flash chromatography (PEt/ EtOAc 1:1) yielding the tri-
ester 3 as a colourless oil (193 mg, 30%). Rv=0.3 (PEt/EtOAc (5 : 5)).
MS (ESI): 345 (M+Na)+ 1H NMR (360 MHz, CDCl3): δ=[1.37, 1.41
(2×d, 3H, H8*, J7–8=4.9 Hz)], [3.73, 3.75, 3.76 (3×s, 3H, CH3O)], [3.82–
4.51 (m, 12H, H2, H3, H4, H5, H6, 3×H9)], [5.22, 5.43 (2×q, 1H, H7*,
H2, H3, H4, 3×H9)], [4.84, 4.85, 4.86 (3×s, 6H, 3×CH2Bn)], [5.2, 5.4
(2×q, 1H, H7*, J7–8 =4.9 Hz)], 5.5, 5.6 (2×d, 1H, H1*, J1–2 =4.5 Hz),
3.36 (s, 15H, HBn). 13 C NMR (90 MHz, CDCl3): δ={20.7, 21.2
(2×C8*)], [33.6, 39.2 (3×C11)], 67.4 (C6), [69.0, 69.1, 70.2 (3×C9)],
73.4 (C2), 74.7 (C5), 75.7 (CH2Bn), 78.5 (C3), 79.4 (C4), 97.4, 97.7
(2×C1*), 99.9, 100.4 (2×C7*), [128.3, 128.6, 128.8, 129.5, 134.5,
134.6 (CBn)], 171.8, 171.9 (3×C10).
4.2.2.8.
Synthesis
of
1,2-O-ethylidene-3,4,6-tri-O-[((N-methyl-N-
J
7–8 =4.9 Hz)], [5.51, 5.62 (2×d, 1H, H1*, J1-2 =4.7 Hz)]. 13 C NMR
hydroxy)-carbamoyl)-methyl]-α-D-galactopyranose
(Me1GALH3).
(90 MHz, CDCl3): δ=[20.6, 20.9 (C8*)], 51.2 (CH3O), 67.3 (C6), [68.6,
69.1, 69.9 (3×C9)], 72.6 (C2), 75.3 (C5), 78.2 (C3), 79.4 (C4), 97.4, 98.0
(2×C1*), 99.9, 100.8 (2×C7*), 170.9 (3×CO).
(Adapted from [37]).
The compound 5 (448 mg, 0.6 mmol) dissolved in 41 mL MeOH was
hydrogenated over 10% Pd/C (45 mg) catalyst for 4 h. The palladium
was removed by filtration on celite® and washed with MeOH. The
residue was then evaporated to afford a pale orange powder (252 mg,
90%).
4.2.2.5. 1,2-O-ethylidene-3,4,6-tri-O-((N-hydroxy)-carbamoyl)-methyl)-
α-D-galactopyranose (H1GALH3). (adapted from [34])
MS (ESI): 489.99 (100%) [M+Na]+ IR (NaCl pellet) ν (cm−1) :
3185.2 (νOH), 1643.0 (νC=O), 1085.2 (νNO). 1H NMR (360 MHz,
CD3OD): δ=[1.31, 1.40 (2×d, 3H, H8*, J7–8=4.9 Hz)], 3.11, 3.31 (s,
9H, 3×H11), 3.79 (m, 3H, H5, H6), [4.00–4.45 (m, 9H, H2, H3, H4,
3×H9)], [5.21, 5.43 (2×q, 1H, H7*, J7–8=4.9 Hz)], [5.52, 5.63 (2×d,
1H, H1*, J1–2=4.5 Hz)]. 13 C NMR (90 MHz, CD3OD): δ={19.6, 19.9
(2×C8*)], 34.8 (3×C11) 67.4 (C6), [69.0, 69.1, 70.2 (3×C9)], 73.4 (C2),
74.7 (C5), 78.5 (C3), 79.4 (C4), 97.4, 97.6 (2×C1*), 100.0, 100.5
The tri-ester 3 (380 mg, 0.9 mmol) was dissolved in 10 mL anhy-
drous THF and 2 mL of aq. hydroxylamine was added. The reaction mix-
ture was stirred at room temperature overnight. Then 50 mL water and
50 mL ethylacetate were added and the aqueous phase was washed 3
times with ethylacetate and then evaporated. The white sticky oil was
lyophilised in water (2 mL) to obtain a white powder (390 mg, quant.).
MS (ESI): 448.1 (100%) [M+Na]+ IR (KBr pellet) ν (cm−1) :
1670.1 (νC=O); 1083.5 (νNO). 1H NMR (360 MHz, CD3OD): δ=
[1.27, 1.33 (2×d, 3H, H8 J7-8 4 4 z)], 3.14 (m, 2H, H6), 4.02 (m, 2H,
H4, H5), 4.12 (m, 2H, H2, H3), 4.57-4.67 (m, 6H, 3 x H9), 5.15, 5.39
(2×q, 1H, H7*, J7–8 =4.4 Hz), 5.55 (2×d, 1H, H1*, J1–2 =4.5 Hz). 13 C
NMR (90 MHz, CD3OD): 18.9, 20.1 (2×C8*), 67.7 (C6), [69.2, 69.6,
70.4 (3×C9)], 72.5 (C2), 74.7 (C5), 78.1 (C3), 79.9 (C4), [97.3, 97.9
(2×C1*)], [100.0, 100.7 (2×C7*)], 167.5 (3×C10).
(2×C7*), 170.9, 170.8 (3×C10). HRMS-ES+
:
m/z calc. for
C
C
17H29O12N3Na 490.1643, found 490.1641. Anal. Calcd for
17H29O12N3, H2O: C, 42.06; H, 6.44. Found: C, 42.19; H, 6.42.
IR spectrum of the Fe(III) complex was recorded on a powder
obtained as previously mentioned (see HGLYH3) using NaCl pellet.
IR (NaCl pellet) ν (cm−1): 1629.0 (νC=O) ; 1087.3 (νNO).
IR spectrum of the Fe(III) complex was recorded on a powder
obtained as previously mentioned (see HGLYH3) using KBr pellet. IR
(KBr pellet) ν (cm−1) : 1605.2 (νC=O) ; 1087.4 (νNO).
4.2.2.9.
1,2-O-ethylidene-3,4,6-tri-O-(pent-4-enyl)-α-D-galactopyra-
nose (6). Sodium hydride (60% dispersion in oil, 340 mg, 8.5 mmol)
was added portionwise to a stirred solution of 1 (500 mg, 2.42 mmol)
at 0 °C in 10 mL anhydrous DMF. After 30 min 5-bromo-1-pentene
(1.11 mL, 9.43 mmol) was added dropwise and the reaction mixture
was stirred overnight at room temperature. The reaction flask was
placed in an ice bath and the excess sodium hydride was quenched by
the slow addition of water. The mixture was evaporated to dryness,
then the residue was dissolved in CH2Cl2 (200 mL) and washed with
water (3×100 mL). The organic phase was dried (Na2SO4), filtered
and concentrated, then the residual oil was purified by flash chromatog-
raphy (PEt/ EtOAc 9:1) yielding the desired tri-alkene compound 6 as a
colourless oil (334 mg, 34%). Rv=0.7 (PEt/ EtOAc (9:1)).
4.2.2.6. 1,2-O-ethylidene-3,4,6-tri-O-(methylcarboxy)-α-D-galactopyr-
anose (4). (adapted from [61])
Hydrate lithium hydroxide (355 mg, 8.46 mmol) was added to a
solution of 3 (703 mg, 1.66 mmol) in 16.6 mL THF:H2O (4 : 2). The
reaction mixture was stirred overnight at room temperature. Then
reaction mixture was stirred with Dowex® (H+, 50WX2-100) until
pH=2. The solution was diluted with water (20 mL) and washed
with ethylacetate (60 mL). The evaporation of the aqueous phase
gave a yellow oil (679 mg, quant.).
1H NMR (360 MHz, D2O): δ={1.31, 1.36 (2×d, 3H, H8*, J7–8 4.8
Hz)], [3.76–3.85 (m, 3H, H5, H6)], [4.05–4.43 (m, 9H, H2, H3, H4,
3×H9)], [5.17, 5.43 (2×q, 1H, H7*, J7–8 =4.8 Hz)], [5.54, 5.60 (2×d,
1H, H1*, J1–2 =4.6 Hz)]. 13 C NMR (90 MHz, D2O): δ={19.8, 20.1
(2×C8*)], 67.2 (C6), [68.6, 69.1, 69.9 (3×C9)], 72.6, (C2), 75.1 (C5),
78.0 (C3), 79.3 (C4), 97.1, 97.5 (2×C1*), 100.0, 100.7 (2×C7*),
[173.8, 173.9, 174.2 (3×C10)].
MS (ESI): 433.4 (100%) [M+Na]+ 1H NMR (300 MHz, CDCl3): δ=
{1.35, 1.42 (2×d, 3H, H8*, J7–8 =4.8 Hz)], [1.63–1.67 (m, 12H,
H
spacer)], 2.08–2.11 (m, 6H, 3×H9), [3.41–4.23 (m, 6H, H2, H3, H4,
H5, H6)], 4.93–5.14 (m, 6.5 H, H7, 3×H11), 5.40–5.42 (m, 1H, H1*, H7),
5.52 (d, 0.5H, H1*, J1–2=4.20 Hz), 5.71–5.85 (m, 3H, 3×H10). 13 C
NMR (75 MHz, CDCl3): δ=20.7, 21.2 (2×C8*), [22.0, 28.8, 29.1,
29.3, 30.2, 31.3 (6×Cspacer)], 67.4 (C6), [69.0, 69.1, 70.2 (3×C9)],
73.4 (C2), 74.7 (C5), 78.5 (C3), 79.4 (C4), 97.4, 97.7 (2×C1*), 100.1,
100.5 (2×C7*), [114.6, 114.7, 115.7 (3×C11)], [137.1, 138.1, 138.2
(3×C10)].
4.2.2.7. 1,2-O-ethylidene-3,4,6-tri-O-[((N-methyl-N-hydroxybenzyl)car-
bamoyl)-methyl]-α-D-galactopyranose (5). (adapted from [37])
N-methyl-O-benzylhydroxylamine (1.10 g, 8.05 mmol) [35,36]
was dissolved in a solution of 4 (679 mg, 1.79 mmol) in 18 mL H2O:
DMF (1 : 1). Then EDAC.HCl (1.54 g, 8.05 mmol) was added portionwise
and the reaction mixture was stirred for 2 h at room temperature. The
solution was diluted with 50 mL EtOAc and the organic phase was
washed with a solution NaHCO3sat. (50 mL), water (50 mL), 1 M HCl
(50 mL), NaCl (50 mL). The organic phases were dried (Na2SO4),
filtered and concentrated. The mixture was purified by flash chroma-
tography (PEt/ EtOAc 1:9) to afford a yellow oil (631 mg, 48%).
Rv=0.35 (EtOAc).
4.2.2.10. 1,2-O-ethylidene-3,4,6-tri-O-((4-oxo-4-methoxy)-butyl)-α-D-
galactopyranose (7). (adapted from [58])
Compound 6 (558 mg, 1.4 mmol) was dissolved in 18 mL anhydrous
CH2Cl2 (0.08 M). A solution of NaOH (1.17 g, 29.4 mmol) in anhydrous
MeOH (2.5 M) was added. The mixture was cooled at −78 °C in an
acetone-dry ice bath and flushed with O3 until the solution remained
blue (ca. 3 h). The reaction mixture was allowed to warm up to room
temperature. Then water was added and the aqueous phase was
extracted with 3×100mL CH2Cl2. The organic phase was dried
(Na2SO4), filtered and concentrated. The residual oil was purified by
MS (ESI): 738.0 (50%) [M+H]+, 760.3 (100%) [M+Na]+ 1H NMR
(360 MHz, CDCl3): δ=[1.34, 1.41 (2×d, 3H, H8*,J7–8 =4.9 Hz)], [3.11,
3.14, 3.15 (3×s, 9H, 3×H11)], 3.79 (m, 3H, H5, H6), [4.00–4.45 (m, 9H,

66
C. Neff et al. / Journal of Inorganic Biochemistry 112 (2012) 59–67
flash chromatography (PEt/ EtOAc 6:4) yielding the tri-ester 7 as a
colourless oil (293 mg, 40%). Rv=0.4 (PEt/ EtOAc (5: 5)).
IR spectrum of the Fe(III) complex was recorded on a powder
obtained as previously mentioned (see HGLYH3) using KBr pellet. IR
(KBr pellet) ν (cm−1): 1650.2 (νC=O) ; 1119.9 (νNO).
MS (ESI): 529.3 (73%) [M+Na]+ 1H NMR (250 MHz, CDCl3):
δ=1.33, 1.39 (2×d, 3H, H8*, J7–8=4.7 Hz), 1.83–1.89 (m, 12H, Hspacer),
2.36–2.41 (m, 6H, 3×H9), 3.46–4.21 (m, 15H, H2, H3, H4, H5, H6, 3×C11),
5.15–5.49 (m, 2H, H1, H7).13 C NMR (90 MHz, CDCl3): δ=20.7, 21.1
(2×C8*), [24.9, 25.2, 25.4, 30.6 (Cspacers)], 51.5 (3×C11), 68.6 (C6),
[70.2, 71.7, 72.7, 73.5, 78.7, 81.4 (C2, C3, C4, C5, 3×C9)], 97.4, 97.6
(2×C1*), 99.9, 100.3 (2×C7*), 171.9, 173.8 (3×C10).
4.3. Potentiometric measurements
The potentiometric titrations were performed using a Metrohm
DMS 716 Titrino automatic titrator system equipped with a glass-Ag/
AgCl combined electrode (Metrohm, filled with 3 M NaCl solution).
The electrode was calibrated to read p[H] according to the classical
method [62]. The ligand and its iron (III) complexes of ca 0.001 M
were titrated in a jacketed thermostated cell with standardized
0.020 M NaOH. Argon was bubbled through the solutions to exclude
CO2 and O2. NaOH was prepared from 0.1 M NaOH (Prolabo) and
was standardized against potassium hydrogen phtalate. Carbonate
content was checked by Gran's method. The titration data (100 points
collected over the pH range 6–11 for the ligand solution) were refined
by the nonlinear least-squares refinement program HYPERQUAD [38]
to determine the protonation constants Ka of the ligand. The pKa
values have been calculated from the cumulative constants deter-
mined with the program. The uncertainties in the pKa values corre-
spond to the standard deviations in the cumulative constants. The
complexation constants could not be determined since the ferric com-
plex is fully formed at pH 2.
4.2.2.11.
1,2-O-ethylidene-3,4,6-tri-O-(propylcarboxy)-α-D-galactopyra-
nose (8). (adapted from [61])
Hydrate lithium hydroxide (118 mg, 2.80 mmol) was added to a
solution of 7 (280 mg, 0.55 mmol) in 5 mL THF : H2O (4 : 2). The reac-
tion mixture was stirred overnight at room temperature and then
quenched with Dowex® (H+, 50WX2-100, ion exchange resin) until
pH5. The solution was diluted with water (10 mL) and washed with
ethyl acetate (30 mL). The evaporation of the aqueous phase gave 8 as
a brown liquid (235 mg, 93%).
1H NMR (360 MHz, CDCl3): δ={1.33, 1.40 (2×d, 3H, H8*, J7–8
=
4.7 Hz)], 1.85–1.89 (m, 12H, Hspacer), 2.37–2.44 (m, 6H, 3×H9),
[3.31–4.20 (m, 6H, H2, H3, H4, H5, H6)], 5.15–5.51 (m, 2H, H1,
H7).13 C NMR (90 MHz, CDCl3 ): δ=19.7, 20.0 (2×C8*), [24.8, 25.0,
25.2, 30.2 (Cspacer)], 68.4 (C6), [68.7, 69.9, 71.6, 72.5, 73.5, 78.3, 78.4
(C2, C3, C4, C5,3×C9)], 97.4, 97.7 (2×C1*), 100.3, 100.6 (2×C7*),
175.9 (3×C10).
4.4. UV–visible spectrophotometric measurements
The ferric complexes were investigated by a spectrophotometric
titration. Measurements were carried out on a Varian Cary 50 UV–
visible spectrophotometer equipped with a Peltier thermostating
accessory. The UV-visible spectrum of a solution containing equal
amounts of ligand and Fe(III) (about 2×10−4 M) was monitored as
a function of pH over the range 0.3–9 (adjusted with HClO4 or
NaOH). The ferric complexes were investigated in two sets of experi-
mental conditions. (i) in the pH range 0.3–2.0 for FeIII–Me3GALH3,
(0.9–2.0 for the others systems) a batch titration was carried out for
various H+ concentrations adjusted with standardized HClO4 solutions.
(ii) in the pH range 2.0–9.0, titrations were carried out in both direc-
tions, from low pH to high pH and from high pH to low pH. An aliquot
was taken from the solution after each adjustment of the pH (which
was measured using a Metrohm 713 pH meter) and its spectrum was
recorded. The dilution due to the addition of HClO4 or NaOH was cor-
rected accordingly. The spectrophotometric data were fitted with both
the Specfit [41-43] and the LETAGROP-SPEFO programs. [44,45] The
program uses a non linear least-squares method and calculates the ther-
modynamic constants of the absorbing species and their corresponding
electronic spectra. The calculations were done from absorbance data be-
tween 300 and 700 nm for 40 solutions over the pH range 0.5–9. The
range of values for the residual-squares sum (Σ(Aexp −Acalc)2) of the
4.2.2.12. 1,2-O-ethylidene-3,4,6-tri-O-[(N-methyl-N-hydroxybenzyl)-
carbamoyl)-propyl]-α-D-galactopyranose (9). (Adapted from [37]).
N-methyl-O-benzylhydroxylamine (315 mg, 2.3 mmol),[35,36] was
dissolved in a solution of 8 (235 mg, 0.51 mmol) in 6 mL H2O : DMF
(1 : 1). EDAC.HCl (440 mg, 2.3 mmol) was added portionwise and the
reaction mixture was stirred for 3 h at 50 °C. The solution was diluted
with 50 mL ethylacetate and the organic phase was washed with a solu-
tion NaHCO3sat (50 mL), water (50 mL), HCl (1 M, 50 mL), NaCl (1 M,
50 mL). The organic extract was dried (Na2SO4), filtered and concen-
trated. The mixture was purified by flash chromatography (EtOAc
100%) to afford 9 colourless oil (52 mg, 12%). Rv=0.33 (EtOAc). MS
(ESI): 844.4 (100%) [M+Na]+1H NMR (300 MHz, CDCl3): δ=[1.32,
1.40 (2×d, 3H, H8*, J7–8=4.8 Hz)], [1.80–1.89 (m, 12H, Hspacer)],
2.47–2.49 (m, 6H, 3×H9), 3.18 (s, 9H, 3×H11), 3.42–4.12 (m, 6H, H2,
H3, H4, H5, H6), 4.82 (s, 6H, CH2-Bn), 5.15–5.49 (m, 2H, H1, H7), 7.37
(s, 15H, HBn).13 C NMR (90 MHz, CDCl3): δ=20.7, 21.0 (2×C8*),
[24.5, 24.7, 25.0, 28.6 (Cspacer)], 33.5 (3×C11), 68.8 (C6), 70.5 (CH2Bn),
[72.1, 72.9, 73.1, 73.3, 78.6, 81.4 (C2, C3, C4, C5, 3×C9)], 97.4, 97.6
(2×C1*), 100.0, 100.5 (2×C7*), [128.7, 128.9, 129.2 (Co,m,pBn)], 134.5
(CqBn), 174.8, 174.9 (3×C10).
fits was 10−1–10−2
.
4.2.2.13. 1,2-O-ethylidene-3,4,6-tri-O-[(N-methyl-N-hydroxy)-carbamoyl)-
propyl]-α-D-galactopyranose (Me3GALH3). (adapted from [37])
Abbreviations
The compound 9 (97 mg, 0.12 mmol) dissolved in 15 mL MeOH
was hydrogenated over 10% Pd/C (10 mg) catalyst for 90 mn. The pal-
ladium is removed by filtration on celite® with MeOH. The residue is
then evaporated to afford a pale orange oil (61 mg, 92%).
THF
tetrahydrofurane
DMF
LMCT
PEt
N,N-dimethylformamide
ligand to metal charge transfer
petroleum ether
MS (ESI) : 574.2 (100%) [M+Na]+ IR (KBr) ν (cm−1): 3421.0
(νOH), 1623.4 (νC=O) ; 1110.9 (νNO). 1H NMR (250 MHz, CDCl3):
δ=[1.32, 1.34 (2×d, 3H, H8*, J7-8 =4.8 Hz)], [1.85–1.89 (m, 12H,
EDAC.HCl 1-(3-dimethylaminopropyl-3-ethylcarbodiimide
hydrochloride
Rv
retention value
H
spacer)], 2.55–2.58 (m, 6H, 3×H9), 3.22 (s, 9H, 3×H11), [3.54–
ATR
attenuated total reflexion
4.21 (m, 6H, H2, H3, H4, H5, H6,)], 5.15–5.52 (m, 2H, H1, H7).13
C
MS-ESI mass spectrum, electrospray ionization
NMR (62.5 MHz, CDCl3): δ=19.7, 20.0 (2×C8*), [24.6, 24.9, 25.1
(Cspacer)], 34.9 (3×C13) [68.7, 70.3, 71.9 (C6, 3×C9)], [72.5, 73.6,
78.2, 81.0 (C2, C3, C4, C5)], 97.4, 97.6 (2×C1*), 100.0, 100.2
(2×C7*), 174.4 (3×C10). HRMS-ES+: m/z calc. for C23H41N3O12Na
574.2588, found 574.2582
MS-CI
NMR
DEPT
COSY
HSQC
mass spectrum, chemical ionisation
nuclear magnetic resonance
Distortionless Enhancement by Polarization Transfer
Correlation Spectroscopy
heteronuclear single quantum correlation

C. Neff et al. / Journal of Inorganic Biochemistry 112 (2012) 59–67
67
[25] S. Dhungana, J.M. Harrington, P. Gebhardt, U. Moellmann, A.L. Crumbliss, Inorg.
Chem. 45 (2007) 8362–8371.
Acknowledgments
[26] P. Gebhardt, A.L. Crumbliss, U. Möllmann, M.J. Miller, Biometals 21 (2008) 41–51.
[27] B. Gyurcsik, L. Nagy, Coord. Chem. Rev. 203 (2000) 81–149.
[28] S.J. Angyal, J. Chem. Soc. Rev. 9 (1980) 415–428.
[29] F. Kong, J. Du, H. Shang, Carbohyd. Res. 162 (1987) 217–225.
[30] R.P. McGeary, I. Jablonkai, I. Toth, Tetrahedron 57 (2001) 8733–8742.
[31] O. Gavard, Y. Hersant, J. Alais, V. Duverger, A. Dilhas, A. Bascou, D. Bonnaffé, Eur. J.
Org. Chem. (2003) 3603–3620.
The authors want to acknowledge the French Ministry of Research for
special funding through the program ACI-jeunes chercheurs attributed to
C.P. (JC4044-2004) and Paris-Sud 11 University for C.N.'s doctoral
fellowship.
[32] D.A. Brown, W.K. Glass, R. Mageswaran, B. Girmay, Magn. Res. Chem. 26 (1988)
970–973.
References
[33] M. Devocelle, B.M. McLoughlin, C.T. Sharkey, D.J. Fitzgerald, K.B. Nolan, Org. Biomol.
Chem. 1 (2003) 850–853.
[34] L. Salmon, E. Prost, C. Merienne, R. Hardré, G. Morgant, Carbohyd. Res. 335 (2001)
195–204.
[1] K.N. Raymond, C.J. Carrano, Acc. Chem. Res. 12 (1979) 183–190.
[2] M.T. Caudle, C.D. Caldwell, A.L. Crumbliss, Inorg. Chim. Acta 240 (1995) 519–525.
[3] J.M. Harrington, S. Chittamuru, S. Dhungana, H.K. Jacobs, A.S. Gopalan, A.L.
Crumbliss, Inorg. Chem. 2010 (49) (2010) 8208–8221.
[35] F. Bonaccorsi, R. Giorgi, Synthetic Comm. 27 (1997) 1143–1147.
[36] S.K. Sharma, M.J. Miller, S.M. Payne, J. Med. Chem. 32 (1989) 357–367.
[37] B.H. Lee, M.J. Miller, C.A. Prody, J.B. Neilands, J. Med. Chem. 28 (1985) 317–323.
[38] P. Gans, A. Sabatini, A. Vacca, Talanta 43 (1996) 1739–1753.
[39] E.Q. Adams, J. Am. Chem. Soc. 38 (1916) 1503–1510.
[40] B. Monzyk, A.L. Crumbliss, J. Org. Chem. 45 (1980) 4670–4675.
[41] H. Gampp, M. Maeder, C.J. Meyer, A.D. Zuberbühler, Talanta 32 (1985) 95–101.
[42] H. Gampp, M. Maeder, C.J. Meyer, A.D. Zuberbühler, Talanta 32 (1985) 257–264.
[43] H. Gampp, M. Maeder, C.J. Meyer, A.D. Zuberbühler, Talanta 32 (1986) 943–951.
[44] L.G. Sillen, Acta Chem. Scand. 18 (1964) 1085–1098.
[45] L.G. Sillen, B. Warnqvist, Ark. Kemi. 31 (1969) 377–390.
[46] G. Schwarzenbach, K. Schwarzenbach, Helv. Chim. Acta 46 (1963) 1390.
[47] G. Anderegg, F. L'Eplattenier, G. Schwarzenbach, Helv. Chim. Acta 46 (1963)
1409–1422.
[4] H. Boukhalfa, A.L. Crumbliss, Biometals 15 (2002) 325–339.
[5] J. Kragter, Atlas of Metal-Ligand Equilibria in Aqueous Solution, Halstead Press,
New York, 1978.
[6] J.-L. Pierre, G. Lescoat, Act. Chimique (2003) 123–127.
[7] D.S. Kalinowski, D.R. Richardson, Pharmacol. Rev. 57 (2005) 547–583.
[8] A. Shanzer, J. Libman, Metals ions in biological systems, vol. 35, Biomimetic sidero-
phores: from structural probes to diagnostic tools, 1998, pp. 329–354.
[9] L.K. Charkoudrian, D.M. Pham, A.M. Kwon, A.D. Vangeloff, K.J. Franz, Dalton Trans
(2007) 5031–5042.
[10] D.R. Dunaief, Invest. Ophthalmol. Visual Sci. 47 (2006) 4660–4664.
[11] M.G. Leed, N. Wolkow, D.M. Pham, C.L. Daniel, D.R. Dunaief, K.J. Franz, J. Inorg.
Biochem. 105 (2011) 1161–1172.
[12] F. Bellot, R. Hardré, G. Pelosi, M. Thérisod, C. Policar, Chem. Commun. (2005)
5414–5417.
[13] G. Charron, F. Bellot, F. Cisnetti, G. Pelosi, J.-N. Rebilly, E. Rivière, A.-L. Barra, T.
Mallah, C. Policar, Chem. Eur. J. 13 (2007) 2774–2782.
[14] F. Cisnetti, R. Guillot, M. Desmadril, G. Pelosi, C. Policar, Dalton Trans. (2007)
1473–1476.
[15] F. Cisnetti, R. Guillot, M. Thérisod, C. Policar, Acta Cryst C63 (2007) m201–m203.
[16] F. Cisnetti, R. Guillot, N. Ibrahim, F. Lambert, M. Thérisod, C. Policar, Carbohyd.
Res. 343 (2008) 530–535.
[17] F. Cisnetti, R. Guillot, M. Thérisod, M. Desmadril, C. Policar, Inorg. Chem. 47 (2008)
2243–2245.
[18] Z. Damaj, F. Cisnetti, L. Dupont, E. Henon, C. Policar, E. Guillon, Dalton Transactions
(2008) 3235–3245.
[19] L. Garcia, S. Maisonneuve, J. Xie, R. Guillot, P. Dorlet, E. Rivière, M. Desmadril, F.
Lambert, C. Policar, Inorg. Chem. 49 (2010) 7282–7288.
[20] L. Garcia, S. Maisonneuve, J. Oudinet-Sin Marcu, R. Guillot, F. Lambert, J. Xie, C.
Policar, Inorg. Chem. 50 (2011) 11353–11362.
[21] S. Heggemann, M. Schnabelrauch, D. Klemm, U. Möllmann, R. Reissbrodt, L.
Heinisch, Biometals 14 (2001) 1–11.
[22] S. Dhungana, S. Heggemann, L. Heinisch, U. Möllmann, H. Boukhalfa, A.L.
Crumbliss, Inorg. Chem. 40 (2001) 7079–7086.
[48] A. Evers, R.D. Hancock, A.E. Martell, R.J. Motekaitis, Inorg. Chem. 28 (1989)
2189–2195.
[49] C.Y. Ng, S.J. Rodgers, K.N. Raymond, Inorg. Chem. 28 (1989) 2062–2066.
[50] A. Ringbom, Complexation in Analytical Chemistry, Interscience, New York, 1963.
[51] P. Comba, Coord. Chem. Rev. 200–202 (2000) 217–245.
[52] P. Aisen, A. Leibman, J. Zweier, J. Biol. Chem. 253 (1978) 1930–1937.
[53] R.W. Harris, C.J. Carrano, K.N. Raymond, J. Am. Chem. Soc. 101 (1979) 2722–2727.
[54] U. Heinz, K. Hegetschweiler, P. Acklin, B. Faller, R. Lattman, H.-P. Schnebli, Angew.
Chem. Int. Ed. 38 (1999) 2568–2569.
[55] S. Steinhauser, U. Heinz, M. Bartholomä, T. Weyermüller, H. Nick, K.
Hegetschweiler, Eur. J. Inorg. Chem. (2004) 4177–4192.
[56] L.K. Charkoudian, D.M. Pham, A.M. Kwon, A.D. Vangeloff, K.J. Franz, Dalton Trans.
(2007) 5031–5042.
[57] R. Bastian, R. Wedberling, F. Palilla, Anal. Chem. 28 (1956) 459–462.
[58] J.A. Marshall, A.W. Garofalo, R.C. Sedrani, Synlett 8 (1992) 643–645.
[59] V.I. Betaneli, M.V. Ovchinnikov, L.V. Backinowsky, N.K. Kochetkov, Carbohyd. Res.
107 (1982) 285–291.
[60] A.P. Kozikowski, J. Lee, J. Org. Chem. 55 (1990) 863–870.
[61] D.T. Hung, J.B. Nerenberg, S.L. Schreiber, J. Am. Chem. Soc. 118 (1996)
11054–11080.
[23] S. Dhungana, S. Heggemann, P. Gebhardt, U. Möllmann, A.L. Crumbliss, Inorg.
Chem. 42 (2003) 42–50.
[24] S. Heggemann, U. Moellmann, P. Gebhardt, L. Heinisch, Biometals 16 (2003)
539–551.
[62] A.E. Martell, R.J. Motekaitis, Determination and Use of Stability Constants, VCH
Weinheim, 1988.