Docking, synthesis, and NMR studies of mannosyl trisaccharide ligands for DC-SIGN lectin

Article abstract of DOI:10.1039/b802144a

DC-SIGN, a lectin, which presents at the surface of immature dendritic cells, constitutes nowadays a promising target for the design of new antiviral drugs. This lectin recognizes highly glycosylated proteins present at the surface of several pathogens su

Full text of DOI:10.1039/b802144a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Docking, synthesis, and NMR studies of mannosyl trisaccharide ligands for
DC-SIGN lectin†
Jose´ J. Reina,a Irene D´ıaz,a Pedro M. Nieto,*a Nuria E. Campillo,b Juan A. Pa´ez,b Georges Tabarani,c
Franck Fieschic and Javier Rojo*a
Received 6th February 2008, Accepted 28th April 2008
First published as an Advance Article on the web 5th June 2008
DOI: 10.1039/b802144a
DC-SIGN, a lectin, which presents at the surface of immature dendritic cells, constitutes nowadays a
promising target for the design of new antiviral drugs. This lectin recognizes highly glycosylated proteins
present at the surface of several pathogens such as HIV, Ebola virus, Candida albicans, Mycobacterium
tuberculosis, etc. Understanding the binding mode of this lectin is a topic of tremendous interest and
will permit a rational design of new and more selective ligands. Here, we present computational and
experimental tools to study the interaction of di- and trisaccharides with DC-SIGN. Docking analysis
of complexes involving mannosyl di- and trisaccharides and the carbohydrate recognition domain
(CRD) of DC-SIGN have been performed. Trisaccharides Mana1,2[Mana1,6]Man 1 and
Mana1,3[Mana1,6]Man 2 were synthesized from an orthogonally protected mannose as a common
intermediate. Using these ligands and the soluble extracellular domain (ECD) of DC-SIGN, NMR
experiments based on STD and transfer-NOE were performed providing additional information.
Conformational analysis of the mannosyl ligands in the free and bound states was done. These studies
have demonstrated that terminal mannoses at positions 2 or 3 in the trisaccharides are the most
important moiety and present the strongest contact with the binding site of the lectin. Multiple binding
modes could be proposed and therefore should be considered in the design of new ligands.
epitope, is considered one of the strongest ligands for this lectin.
Introduction
Carbohydrate–protein interactions are selective, most of the time
DC-SIGN (dendritic cell-specific ICAM-3 grabbing non-integrin)
calcium dependent, and very weak. Nature overcomes this weak
or CD209 is a C-type lectin expressed at the surface of im-
affinity providing a multivalent presentation of carbohydrate
mature dendritic cells. This lectin presents at the C-terminus a
epitopes such as cluster organization at the cell surface of glycosph-
carbohydrate recognition domain (CRD) able to interact with
ingolipids and highly glycosylated glycoproteins. To study and
highly glycosylated proteins found on several pathogens such as
intervene in biological processes where this type of interaction is
viruses (HIV-1 and 2, SIV-1, Ebola virus, HCV, SARS virus,
involved, the design and preparation of multivalent carbohydrate
cytomegalovirus, dengue virus); bacteria (Helicobacter pylori,
Klebsiella pneumonae, Mycobacterium tuberculosis); yeast (Can-
systems is required. There is a continuing interest in designing new
and effective multivalent tools. Selection of carbohydrate epitopes
requires a deep knowledge of the binding mode of the ligands and
the binding site of the protein receptor. Information about binding
dida albicans); and parasites (Schistosoma mansoni, Leishmania
pifanoi).1,2 This so broad spectrum of pathogens recognized by
DC-SIGN has led to consider this lectin as an universal pathogen
constants of monovalent and multivalent mannosyl and fucosyl
receptor. This lectin has attracted the interest of the scientific
oligosaccharides with DC-SIGN have been reported by different
community since the discovery of the role that DC-SIGN plays
groups using a variety of techniques such as ELLA, biosensors,
in a HIV trans infection process.3
etc.4–11 However, to date only scarce information at the molecular
level is available about how carbohydrates are recognized by DC-
SIGN in solution. To our best knowledge, the only available
information concerning carbohydrate ligands and DC-SIGN is at
the solid state. X-Ray structures of complexes formed by different
carbohydrate oligosaccharides with up to 9 units constituting part
of the high mannose structure and CRD of DC- and L-SIGN
have been recently published.12–14 Our interest on the receptor DC-
SIGN lead us to explore in more detail the molecular basis of this
recognition process with the aim to design and prepare appropriate
ligands and their corresponding multivalent systems for potential
applications in biological processes where DC-SIGN is involved.
Also, available structural information about the binding process of
carbohydrates to DC-SIGN indicates the possibility of different
binding modes of these ligands. A preliminary study involving
Pathogen glycoproteins recognized by DC-SIGN contain
mainly mannose and fucose carbohydrate structures as N-
glycans at different Asn positions of the glycoproteins. High
mannose structure, constituted by Man9GlcNAc2 as the main
aGrupo de Carbohidratos, Instituto de Investigaciones Qu´ımicas, CSIC -
Universidad de Sevilla, Ame´rico Vespucio 49, 41092 Seville, Spain.
E-mail: javier.rojo@iiq.csic.es; pedro.nieto@iiq.csic.es; Fax: +34 954 46 05
65
bInstituto de Qu´ımica Me´dica, CSIC, Juan de la Cierva 3, 20006 Madrid,
Spain
cLaboratoire des Prote´ines Membranaires, Universite´ Joseph Fourier/CEA,
DSV/CNRS, UMR 5075, Institut de Biologie Structurale Jean-Pierre Ebel,
41, rue Jules Horowitz, 38027 Grenoble, Cedex 1, France
† Electronic supplementary information (ESI) available: Further details of
docking and NMR studies. See DOI: 10.1039/b802144a
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 2743–2754 | 2743
©

View Article Online
DC-SIGN and carbohydrate mimics has been published recently
by some of us.15
Here, we describe the use of computational (docking studies)
and NMR (STD, and transfer NOE experiments) tools to get
information about the binding mode of DC-SIGN ligands. Among
the potential epitopes to be tested as ligands for DC-SIGN, we
have analyzed two mannosyl trisaccharides. On one hand, these
compounds represent a good choice as binding ligands (better
than the simple mannose) and on the other hand, they can
be obtained with a reasonable synthetic cost. Because the high
mannose structure consists of mannose units alinked by glycosidic
bonds at positions 2, 3, and 6, we have selected two branched
trisaccharides. The trisaccharide Mana1,2[Mana1,6]Man 1 and
the trisaccharide Mana1,3[Mana1,6]Man 2. The trisaccharide 1 is
not present in the high mannose structure, but it constitutes the
repeated unit of the cell-wall mannans such as lipoarabinomannan
(LAM) present in Mycobacterium tuberculosis which is recognized
by DC-SIGN.16–19 The trisaccharide 2 is a fragment present in
the high mannose responsible for the recognition process by DC-
SIGN as has been described previously.12
As a starting point for docking studies, we have used the crys-
tallographic information already published. We have analyzed five
selected ligands, the two trisaccharides above mentioned and the
corresponding disaccharide moieties present in these molecules:
Mana1,2Man, Mana1,3Man, and Mana1,6Man. Simultaneously,
we have synthesized the two selected trisaccharides with a short
linker at the anomeric position which will allow the attachment of
these ligands to multivalent scaffolds afterwards. We have used
an orthogonally protected mannose previously described as a
common intermediate for the synthesis of the trisaccharides.20
These mannosyl oligosaccharides have been used to test their
binding behavior in the presence of the soluble extra-cellular
domain (ECD) of DC-SIGN. The analysis of the binding process
in solution has been performed using a series of NMR experiments
directed to obtain structural data of the bound ligands.
Fig. 1 Chemical structures of the three disaccharides (Mana1,2Man,
Mana1,3Man and Mana1,6Man), and the two trisaccharides
(Mana1,3[Mana1,6]Man and Mana1,2[Mana1,6]Man) considered in the
docking studies.
the U/w conformational space, along with the distance to the Ca2+
(primary binding site) and the interprotonic distances are given in
the supplementary information (Table S1 and Fig. S1–S20).
Based on the crystal structure, it is possible to define two possible
binding sites for disaccharides: a primary one containing the Ca2+
atom and an adjacent secondary one. We performed docking
studies considering two possible sites for the a1–2 and a1–6-linked
mannoses. In both cases, the most favoured binding mode (Table 1)
is located at the primary site, presenting interactions with the Ca2+
atom; therefore, we focused on the binding at the primary site in
subsequent studies.
The docking study of Mana1,2Man led to three families of
structures with two different binding modes (see supplementary
information, Fig. S5–S8). In the less populated orientation (clus-
ter 3) the reducing end side is bound to the Ca2+ ion at the primary
binding site. The major orientation, where the non-reducing end
of the disaccharide binds the Ca2+ atom (cluster 1 and cluster 2),
corresponds to the best solution in terms of energy and geometry
restrictions. It should be mentioned that recent crystallographic
analysis of the complex formed by this disaccharide and DC-SIGN
has shown multiple binding modes which have been postulated as
a mechanism for increasing affinity.14 All acceptable structures
found for Mana1,6Man are mixtures of gg and gt conformers
around the x torsion angle (see supplementary information,
Results and discussion
Docking analysis
To gain insight about the binding mode of DC-SIGN ligands,
a docking study was carried out considering a representative
set of mannosyl di- and trisaccharide structures. In particular,
we selected for these computational studies the disaccharides
Mana1,2Man, Mana1,3Man and Mana1,6Man, and the trisac-
charides Mana1,3[Mana1,6]Man and Mana1,2[Mana1,6]Man.
(Fig. 1)
The docking of the proposed ligands into the DC-SIGN binding
site was done using an automated docking procedure (FlexiDock)
using the crystal structure data of the complex of DC-SIGN
and the pentasaccharide GlcNAc2Man3 (pdb code 1k9i). Those
structures showing key distances or torsional angles inconsistent
with experimental results were rejected. Then, the best solutions in
terms of energy were energy-minimized prior to further analysis.
The final complexes were evaluated in terms of consensus of
ligand–protein interactions and dihedral angles with the exper-
imental data and the known values of energy interaction using the
programs STC and DOCK module (SYBYL) (Table 1). Detailed
information including key protein residues involved in the binding,
2744 | Org. Biomol. Chem., 2008, 6, 2743–2754
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Table 1 Dihedral angles and docking energy
Complex
U/w
DGbind/kcal mol−1
E
dock/kcal mol−1
1d13
1d16
2d16
1d12
2d12
1t36
1t26
1→3: 65.2/144.8
−4.8
−5.3
−2.2
−5.9
−5.6
−7.3
−6.7
2.1
1→6: 63.8/153.0
−34.6
1→6: 52.2/166.0
−0.1
1→2: 68.5/85.9
−37.2
−19.6
−55.0
−51.4
1→2: 42.1/90.3
1→3: 96.5/155.1 6←1: 68.7/181.4
1→2: 81.4/168.0 6←1: 88.4/205.6
Fig. S9–S11), although after energy-minimization the gg confor-
mation (1d16) was found to be the most stable one (Fig. S12). The
non-reducing end of the sugar chain binds to the Ca2+ atom in all
the best solutions. The docking studies of Mana1,3Man yielded
three possible families (see supplementary information, Fig. S1-
S4). The most populated binding mode, which also had the best
values of docking energy and DGbind, showed in all the solutions
the non-reducing end of the sugar bounded to the Ca2+ atom (with
Ligand synthesis
The trisaccharides Mana(1→3)[Mana(1→6)]Mana1OCH2CH2-
NH2 2 and Mana(1→2)[Mana(1→6)]Mana1OCH2CH2NH2
1
have been prepared using an orthogonally protected mannose
derivative 8 as a common intermediate. This mannose derivative 8
was prepared in 7 steps from 2-azidoethyl a-D-mannopyranoside
(3) as shown in Scheme 1.20
˚
In a previous communication we have described the synthesis
of compound 8 and the corresponding orthogonal deprotections
at positions 2, 3 and 6.20 The hydroxyl group at position 6 was
protected as a silyl group using tert-butyldiphenylsilyl chloride
(TBDPSCl) and imidazole in DMF. A benzyl group was intro-
duced at position 4 as a permanent protecting group (this position
will not be glycosylated in this synthetic approach) using benzyl
chloride and sodium hydride in DMF in 79% yield. For this
aim, the formation of an acetonide using 2,2ꢀ-dimethoxypropane
and pyridinium p-toluenesulfonate (PPTS) in acetone was first
necessary to protect positions 2 and 3 simultaneously. After
introducing the benzyl group at position 4, the acetonide was
easily cleavage with TFA in dichloromethane at rt to give
compound 5 in 63% yield over three steps. The hydroxyl group
at position 3 was selectively protected as the p-methoxybenzyl
(PMB) ether using a stannylene acetal intermediate as strategy.
Reaction with dibutyltin oxide in toluene at reflux and then,
PMBCl and tetrabutylammonium iodide (TBAI) gave a complete
regioselective protection of that position giving compound 6 in
good yield. Finally, a levulinoyl ester was used to protect position
2 using levulinic acid and dicyclohexylcarbodiimide (DCC) in
dichloromethane at rt affording the orthogonally protected man-
nose 8 in 88% yield.
Synthesis of trimannose 2 was accomplished using the mannose
derivative 10 as central unit with free hydroxyl groups at positions
3 and 6 (Scheme 2). This mannose 10 was prepared from
orthogonally protected mannose 8 in two steps by cleavage of silyl
group at position 6 with TBAF in THF at rt to afford mannose
derivative 9 and following treatment with TFA at −20 ◦C in
dichloromethane to remove selectively the PMB group at position
3; mannose derivative 10 was obtained in 97% yield over two
steps. (Scheme 1) A double glycosylation, using 2 equivalents
of thiophenyl mannopyranoside 13 (previously described by our
group)21 as glycosyl donor, N-hydroxysuccinimide (NHS) and
triflic acid as promoter, gave the protected trisaccharide 11 in
78% yield. Two deprotections steps with NaOMe in methanol
to remove acetate and levulinate groups and hydrogenation
using Pd on carbon as catalysts in methanol to remove benzyl
groups and reduce the azide gave trisaccharide 2 in good yield.
(Scheme 2).
a distance of the O3 atom to the calcium atom around 2.3 A) and
glycosidic linkage torsions in accordance with the experimental
angle ranges data for the linkage 1,3 (−49 < U < 82 and −16 <
w < 32) (Fig. S1).
The complexes are stabilized by a large number of hydrogen
bonds, which in the case of the disaccharide Mana1,3Man, were
experimentally observed in the complex of DC-SIGN and the
pentasaccharide GlcNAc2Man3 (see Table S1).12 In addition to
these interactions, the Ca2+ ion is bound by the equatorial 3- and
4-hydroxyl groups of a mannose residue (see Table S1). A detailed
description of relevant protein–ligand interactions, including C–
H · · · aromatic, is given in the supplementary information. The
DGbind and the docking energy calculated with the STC and DOCK
program respectively, suggest that the disaccharides with a1–2 and
a1–6 type union are the best ligands for DC-SIGN (Table 1). These
results are in agreement with the experimental values of binding
published by Weis et al.12
In the case of the trisaccharides, the FlexiDock analysis of the
complex of Mana1,3[Mana1,6]Man with DC-SIGN led to a model
that is very similar to that of the crystal structure (Fig. S13–
S17). The a-1,2 linkage in Mana1,2[Mana1,6]Man is described
by an ensemble of exo-anomeric conformers (U between −30
and −60◦) compatible with the experimental data. The short
distances between H1c and H2a, and between H1c and H1a
observed for Mana1,2[Mana1,6]Man are in agreement with the
NMR studies of this linkage (see supplementary information,
Fig. S18–S22). The 1,6-glycosyl bonding in the trimannoside
was found exclusively as gt conformers around the x torsion,
in contrast to the mixture of gg and gt conformers obtained
for the 1,6-bonded structures of the dimannoside. The structural
parameters of the final complex of each trisaccharide and the
energy of interaction are gathered in Table S1. The hydrogen
bond network observed for the disaccharides is preserved in
both trisaccharide-complex models (Table S1). Both trisaccharides
showed hydrophobic interaction between the ManC (1t26) and
ManD (1t36) units and the residue Val 351. Aromatic interactions
of both complexes with Phe 313 have also been found. In addition,
the mannose ManA in Mana1,2[Mana1,6]Man complex displays
˚
a C–H · · · aromatic interaction (ach x= 160.2, dCX (A) = 4.0, dHpX
˚
(A) = 1.2).
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 2743–2754 | 2745
©

View Article Online
Scheme 1 Synthesis of orthogonally protected mannose derivative 8 and selective deprotections. Reagents and conditions: a) TBDPSCl, Im, DMF, rt,
79%; b) 2,2ꢀ-dimethoxypropane, PPTS, acetone; then, BnCl, NaH, DMF, 0 ◦C to rt; then, TFA DCM, rt 63% over 3 steps; c) Bu2SnO, Tol, D; then,
PMBCl, Bu4NI, D, 82%; d) DCC, LevOH, DCM, rt, 88%; e) TBAF, THF, rt, 95%; f) AcOH, THF, 0 ◦C to rt, then, TBAF, 100%; g) TFA, DCM, −20 ◦C,
97%.
Scheme 2 Synthesis of Mana1–3[Mana1–6]Man 2. Reagents and con-
ditions: a) 13 (2 equiv.), 4 A MS, NIS, TfOH, −20 ◦C, DCM, 78%;
˚
Scheme 3 Synthesis of Mana1–2[Mana1–6]Man 1. Reagents and con-
ditions: a) 13 (2 equiv.), 4 A MS, NIS, TfOH, −20 ◦C, DCM, 70%;
b) NaOMe, MeOH, rt; c) H2, Pd-C, MeOH, rt, 100%.
˚
b) NaOMe, MeOH, rt; c) H2, Pd-C, MeOH, rt, 100%.
Trisaccharide 1 was prepared starting from mannose derivative
7. This trisaccharide was synthesized in three steps as shown in
Scheme 3. Mannose derivative 7 was prepared from mannose
6 removing silyl group at position 6 with TBAF in THF at rt
(Scheme 1). Following the same methodology used for trimannose
11, the trisaccharide 12 was obtained in 70% yield using 2
equivalents of mannosyl donor 13, NHS and triflic acid. Again,
two deprotection steps were needed to prepare 1 using NaOMe
in methanol and catalytic hydrogenation with Pd on carbon in
methanol. In this way, trisaccharide 1 was obtained in good yield.
Both trisaccharides 1 and 2 show at the anomeric position an
adequate linker with a terminal amine allowing their attachment
to multivalent scaffolds to create carbohydrate multivalent tools.
All new compounds described above were fully characterized using
NMR and mass spectrometry.
Structural analysis
The basic structures of trisaccharides 1 and 2 are present in a
wide variety of natural compounds from high mannose to GPI
structures which have been extensively studied, and their 3D
structures have been described both isolated or as part of larger
oligosaccharides,21 and their conformational maps described in
several structural databases such as Heidelberg and CERMAV.22
The interaction of trisaccharides 1 and 2 with DC-SIGN was
studied by STD (saturation transfer difference) techniques deter-
mining their interaction epitopes.23,24 The bound conformations
of trisaccharides 1 and 2 were also deduced from transfer-NOE
experiments registered in presence of DC-SIGN ECD.25
2746 | Org. Biomol. Chem., 2008, 6, 2743–2754
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Trisaccharide Mana1,2[Mana1,6]Man 1. The relative impor-
tance of each mannose unit in the binding process to DC-
SIGN was assessed by means of STD experiments. Several
saturation times were used in order to avoid potential interferences
from differences of longitudinal relaxation of the ligand protons
and thus, to ensure the consistency of the results. The relative STD
effects are shown on Fig. 2. The larger effects are concentrated
on the 1–2 linked residue, the mannose-C, followed by the other
terminal one, mannose-B, while the central one shows the lowest
effects. These results could be explained based on a folded
conformation where the terminal residues contact with the lectin
or by the existence of several binding modes.
Fig. 3 Expansion of the NOESY experiments of 1, free (top) and in the
presence of 1 : 50 molar of DC-SIGN (bottom), showing the NOE peaks
representative of the conformation of the glycosidic linkages. Experiments
registered at 500 MHz, in 150 mM NaCl, 20 mM TRIS-d6 4 mM CaCl2
at 278 K using a mixing time of 600 ms for free and 400 ms in presence of
DC-SIGN.
non-exoanomeric conformers for the 1–2 linkage compared with
the free state. The NOEs between H1b and H6ꢀa and H6ꢀꢀa were
clearly observed, in accord with the trans disposition of the w
torsion. Unfortunately, the data regarding the 1–6 linkage were
not conclusive with respect to the determination of the x rotamer.
Trisaccharide Mana1,3[Mana1,6]Man 2. The STD experi-
ments for 2 were also conclusive with respect to the interaction
between the trisaccharide and DC-SIGN, and clear peaks due to
magnetization transfer were observed. Interestingly, the overall
magnitude of the absolute STD effects was larger than that
observed for the trisaccharide 1. The relative STD values were
higher for the 1–3 linked ManD, suggesting that this unit has
a stronger interaction with the lectin (Fig. 4). In this case, the
terminal residue ManB has a lower transference than the central
one, suggesting small differences in the binding mode between
both trisaccharides. The absolute STD values of both samples
(Table 2) can be compared considering the similar nature of
both compounds, the equivalence in concentration and in the
ratio ligand to protein for both samples and that the dissociation
constant can be estimated to be near high millimolar. In these
Fig. 2 STD experiment of 1 in the presence of DC-SIGN recorded with
2.5 s of saturation time and absolute and relative STD values for key
signals.
Table 2 Relative and absolute STD values for 1 and 2
STD Abs (1)
STD Rel
STD Abs (2)
STD Rel
The conformation of 1 bound to DC-SIGN was studied by
transfer NOE experiments. In the presence of 2% of the lectin,
the sign of the NOE peaks was inverted (see Fig. 3) indicating
ligand–receptor binding at favourable rate. The pattern of NOEs
and their growing rates were not indicative that the conformation
in the bound state was essentially similar to the free ligand
(see supplementary information). The Man 1–2 linkage can be
described mainly as an ensemble of exo-anomeric conformers with
a significant degree of flexibility about the w angle, which shows
two relative minima. The observation of a weak NOE between H2c
and H1a could be indicative of an increase of the population of
H1–A
H2–A
H4–A
H6–A
CH2O2–A
H1–B
H2–B
1.86
2.38
2.94
2.16
nd
2.51
2.57
3.09
3.10
6.25
3.38
nd
29.68
38.08
47.04
34.56
—
40.16
41.12
49.36
49.60
100.00
54.00
—
6.12
5.42
6.00
3.07
4.57
3.82
5.38
6.11
5.05
8.56
8.47
6.33
7.13
60.16
63.32
70.09
35.86
53.33
44.63
62.85
71.38
58.94
100.00
98.89
73.89
83.29
H3–B
H1–C or D
H2–C or D
H4–C or D
H5–C or D
H6–C or D
4.12
65.84
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 2743–2754 | 2747
©

View Article Online
Fig. 5 Expansion of the NOESY experiments of 2, free (top) and in
presence of 1 : 50 molar of DC-SIGN (bottom), showing the NOE peaks
indicative of the conformation of the glycosidic linkages. Experiments
registered at 500 MHz, in 150 mM NaCl, 20 mM TRIS-d6 4 mM CaCl2
at 278 K using a mixing time of 600 ms for free and 400 ms in presence of
DC-SIGN.
The docking study reveals the existence of two potential binding
sites for disaccharides indicating that only the main binding site,
which contains the Ca2+ atom, yields stable complexes. This finding
is supported by the lack of STD effect in the absence of Ca2+
(data not shown). Our study also has disclosed two modes of
recognition differing in the orientation of the carbohydrate in the
primary binding site. These modes differ in the terminal hexose
that is bound to the Ca2+ atom in the main binding site. This
observation agrees with the crystallographic analysis of DC-SIGN
with di- and hexasaccharides, containing Mana1,2Man moieties,
recently solved.14 Our results for the complex with Mana1,2Man,
have shown that in the main orientation the reducing end of the
disaccharide binds the Ca2+ and the structure is superimposible
with the crystallographic one (Fig. 6). Moreover, an alternative
docking solution points to an additional mode of binding with
the Ca2+ atom through the reducing end hexose. This mode seems
similar to the structure of the disaccharide in the minor Man6b
ligand orientation in the crystallographic study.
Fig. 4 STD experiment with 2 in presence of DC-SIGN recorded with
2.5 s of saturation time and absolute and relative STD values for significant
signals.
conditions, the higher STD observed for 2 could be indicative of
stronger binding than 1.24
The transfer NOESY experiments of trisaccharide 2 registered
in the presence of DC-SIGN ECD showed negative NOEs
indicative of transient binding of the carbohydrate to the lectin
(Fig. 5). The observed differences between equivalent experiments
of 2, free or with DC-SIGN, are due to the expected changes in
the linewidth and correlation times. A more quantitative analysis
based on the NOE growing rate does not evidence changes on key
interprotonic distances (supplementary information) indicating
that the relative orientation of the monosaccharidic rings is nearly
the same as in the free compound, and no major conformational
changes can be detected upon binding. Then the 1–3 linkage is
mainly in a syn-w disposition at U angles compatible with the
exo-anomeric effect, and the 1–6 linkage is in a trans disposition
regarding the w torsion but the x rotamer could not be defined.
Conclusions
Herein, we have described a combined approach based on
docking analysis, synthesis and NMR to study the interaction
of carbohydrate ligands (mannosyl di- and trisaccharides) with
the carbohydrate recognition domain of the lectin DC-SIGN and
to evaluate their binding modes.
Fig. 6 Superimposition of DC-SIGN/Mana1–2Man (1d12) complex
obtained from the theoretical docking calculation and from the 3D
structure of disaccharides (pdb code 2it6). Only the binding pocket with
the representative residues is represented here.
2748 | Org. Biomol. Chem., 2008, 6, 2743–2754
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Based on the computational analysis, two branched trisaccha-
rides (Mana1,2[Mana1,6]Man and Mana1,3[Mana1,6]Man) were
selected as the best candidates to be prepared and studied by
NMR. These trisaccharides contain as part of their structure
the 1–2, 1–3 and 1–6 disaccharides presented in high mannose
structures. Moreover, the Mana1,3[Mana1,6]Man is a key moiety
of high mannose and it is considered as one of the most
important trisaccharides interacting with DC-SIGN. On the other
hand, the trisaccharide Mana1,2[Mana1,6]Man, constitutes part
of the arabinomannan oligosaccharide and the docking studies
previewed a better interaction with DC-SIGN probably based on
a carbohydrate–aromatic interaction with the side chain of Phe
313.
The conformational analysis of both trisaccharides reveals that
the geometry recognized by the receptor is comparable with the
conformation in the free state and therefore, there are not any large
conformational changes upon binding. These structures agree in
essence with those predicted by us using docking or with the
crystallographic data.14 The docked models of the trisaccharides
1 and 2 showed that in both cases the non-reducing terminal
mannoses show interactions with several residues of DC-SIGN
while the central mannose, ManA, does not. This model is in
agreement with the STD experiments and indicates that the
terminal 1–2 or 1–3 linked mannoses have the closest interaction
with the lectin. This is the hexose that docking structures have
shown is in contact with the Ca2+ atom on the primary binding
site. Additionally, this mode of binding, similar to a half moon, is
somewhat similar to that found for the structure of GlcNAc2Man3
and DC-SIGN (pdb code 1k9i, see supplementary information).
On the contrary, terminal mannoses B, 1–6 linked to the central
one, receive lower levels of magnetization suggesting a looser
interaction with DC-SIGN. Finally, although the theoretical
structures predict the lowest interactions for the central residue,
experimentally this is only true for the trisaccharide 1 but not for 2.
This observation, together with appreciable differences in the STD
pattern also indicate potential variations in the binding modes
of both compounds, probably as consequence of the existence
of multiple binding modes as proposed by Seeberger and co-
workers.14
were recorded on Bruker Advance DPX 300, and DRX 500 MHz
spectrometers. Chemical shift are in ppm with respect to TMS
(tetramethylsilane) using the manufacturer indirect referencing
method. 2D experiments (COSY, TOCSY, ROESY, and HMQC)
were done when necessary to assign the oligosaccharide spectra.
Optical rotations were measured with a Perkin-Elmer 341 po-
larimeter. Mass spectra were carried out with an Esquire 6000
ESI-Ion Trap from Bruker Daltonics.
Synthesis of 2-azidoethyl 6-O-tert-butyldiphenylsilyl-
a-D-mannopyranoside (4)
To a solution of 3 (945 mg, 3.80 mmol) and imidazole (388 mg,
5.7 mmol) in dry DMF (10 mL) was added dropwise at room
temperature TBDPS-Cl (1 mL, 5.7 mmol). The mixture was stirred
over 4 hours and then concentrated under high vacuum. The
mixture was diluted in CH2Cl2 (30 mL), washed with water (3 ×
30 mL), dried over MgSO4 and concentrated. The residue was
purified by flash chromatography on silica gel (CH2Cl2 : MeOH
9.5 : 0.5) to give compound 4 as a pale oil (1.46 g, 79%). [a]D20
=
1
+16.04 (c = 0.75 in CHCl3); H NMR (500 MHz, CDCl3): d =
7.72–7.68 (m, 4H; HPh), 7.48–7.38 (m, 6H; HPh), 4.85 (d, J = 1.8 Hz,
ꢀ
1H; H1), 3.97–3.77 (m, 5H; H7 + 2H6 + H4 + H7 ), 3.75–3.65 (m,
1H; H7), 3.60–3.53 (m, 1H; H5), 3.38–3.32 (m, 2H; 2H8), 1.01 (s,
9H; HTBDPS); 13C NMR (125 MHz, CDCl3): d = 135.6 (CHTBDPS),
132.8 (CTBDPS), 132.7 (CTBDPS), 129.9 (CTBDPS), 127.9 (CHTBDPS),
99.7 (C1), 71.5 (C3 or C4), 70.9 (C5), 70.4 (C3 or C4), 66.5 (C7),
65.2 (C6), 50.5 (C8), 26.8 (CH3TBDPS), 19.2 (CTBDPS); ESI-MS for
C24H33N3O6Si; calcd: 487.2 M+; found: 510.2 [M + Na]+; elemental
analysis calcd (%) for C24H33N3O6Si: C, 59,11%; H, 6,82%; N,
8,62%; found: C, 59,53%; H, 7,07%; N, 8,74%.
Synthesis of 2-azidoethyl 4-O-benzyl-6-O-
(tert-butyldiphenylsilyl)-a-D-mannopyranoside (5)
Mannose derivative 4 (1.17 g, 2.4 mmol), 2,2ꢀ-dimethoxypropane
(2.7 mL) and PPTS (29 mg, 0.11 mmol) were dissolved in acetone
(20 mL) and the solution was stirred for 24 hours. Et3N was
added to the reaction mixture and the solvent was removed under
vacuum. The residue and benzyl bromide (0.45 mL, 3.8 mmol)
were dissolved in dry DMF (20 mL), and NaH (222 mg, 3.8 mmol)
was added in small portions at 0 ◦C. The reaction mixture was
stirred during 12 hours at room temperature. Then, MeOH (1 mL)
was added to the reaction to quench the excess of NaH and the
solvent was evaporated. Et2O (100 mL) was added to the reaction
residue, washed with H2O (2 × 100 mL), dried over MgSO4 and
concentrated. The residue was dissolved in CH2Cl2 (60 mL) and
TFA (10 mL) was added to the reaction at room temperature.
The solution was stirred over 5 hours. A mixture of ice–water at
These results could be important in the future for the design
of new ligands for this receptor. In particular, the observation,
predicted by the docked structures and supported by the experi-
mental relative STD values, that in both trisaccharides the non-
reducing end 1–2 or 1–3 linked to the central mannose is the most
involved in the interaction with the lectin, independent of the
regiochemistry and relative disposition of the linkage, should be
considered. Evaluation of the antiviral activity of synthesized di-
and trisaccharides are being carried out. Very preliminary results
in an Ebola infection model indicates that trisaccharide 1 presents
a similar activity to disaccharide Mana1,2Man and less activity
than trisaccharide 2 (unpublished results). This preliminary data
are in accordance with our analysis.
0
◦C was added to the reaction, neutralized with NaHCO3 sat.
(100 mL), washed with NaHCO3 sat. (2 × 50 mL) and NaCl
sat. (100 mL), dried with MgSO4 and the solvent was removed
under vacuum. The residue was purified by flash chromatography
on silica gel (toluene–AcOEt, 9 : 1), to give compound 5 as an
oil (843 mg, 63% over three steps). [a]2D0 = +16.02 (c = 0.75 in
Experimental section
1
CHCl3); H NMR (300 MHz, CDCl3): d = 7.76–7.68 (m, 4H;
General remarks
HTBDPS), 7.47–7.20 (m, 11H; 5HBn and 6HTBDPS), 4.89 (s, 1H; H1),
4.69 (syst ABBn, 2H), 4.02–3.90 (m, 4H; H2 + H3 + 2H6), 3.85 (dt,
J = 10.5 and 4.9 Hz, 1H; H7), 3.81–3.65 (m, 2H; H4 and H5), 3.59
All chemicals were obtained from Aldrich and used without
1
further purification, unless otherwise noted. H and 13C NMR
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 2743–2754 | 2749
©

View Article Online
ꢀ
(dt, J = 10.5 and 4.9 Hz, 1H; H7 ), 3.38–3.32 (m, 2H; 2H8), 1.07
129.7 (CAr), 128.5 (CAr), 128.0 (CAr), 127.8 (CAr), 114.0 (CAr), 99.4
(C1), 79.5 (C3), 75.2 (CH2Bn), 73.8 (C4 or C5), 71.9 (C4 or C5), 71.8
(CH2PMB), 68.3 (C2), 66.7 (C7), 62.1 (C6), 55.3 (-OCH3), 50.5 (C8);
ESI-MS for C23H29N3O7; calcd: 459.2 M+; found: 482.3 [M + Na]+
and 498.2 [M + K]+; elemental analysis calcd (%) for C23H29N3O7:
C, 60.12%; H, 6.36%; N, 9.16%; found: C, 60.15%; H, 6.69%; N,
9.16%.
(s, 9H; HTBDPS); 13C NMR (75 MHz, CDCl3): d = 138.6 (CTBDPS),
136.3 (CHTBDPS), 136.0 (CHTBDPS), 133.9 (CAr), 133.6 (CAr), 130.1
(CHAr), 129.0 (CHAr), 128.3 (CHAr), 128.3 (CHAr), 128.1 (CHAr),
128.1 (CHAr), 128.1 (CHAr), 128.0 (CHAr), 99.9 (C1), 77.6 (C5 or
C4), 77.2 (CH2Bn), 75.2 (C5 or C4), 72.9 (C2 or C3), 72.0 (C2 or
C3), 66.7 (C7), 63.5 (C6), 50.9 (C8), 27.2 (CH3TBDPS), 19.7 (CTBDPS);
ESI-MS for C31H39N3O6Si; calcd: 577.3 M+; found: 600.3 [M +
Na]+; elemental analysis calcd (%) for C31H39N3O6Si: C, 64.14%;
H, 7.07%; N, 6.94%; found: C, 64.45%; H, 6.80%; N, 7.27%.
Synthesis of 2-azidoethyl 4-O-benzyl-6-O-tert-butyldiphenylsilyl-
2-O-levulinoyl-3-O-(4-methoxybenzyl)-a-D-mannopyranoside (8)
Synthesis of 2-azidoethyl 4-O-benzyl-3-O-(4-methoxybenzyl)-6-O-
tert-butyldiphenylsilyl-a-D-mannopyranoside (6)
Mannose derivative 6 (100 mg, 0.144 mmol) and DCC (149 mg,
0.720 mmol) were dissolved in CH2Cl2 (2 mL) under argon. After
addition of levulinic acid (146 lL, 1.44 mmol) a precipitate was
observed corresponding to the urea formation. A catalytic amount
of DMAP (8 mg) was added and the reaction was stirred at rt for
24 h. The reaction mixture was diluted with CH2Cl2 (6 mL), filtered
over a pad of Celite and concentrated under vacuum. The residue
was purified by flash chromatography on silica gel (CH2Cl2 and
0.5% acetone) to afford 8 as an oil (100 mg, 88%). [a]2D0 = +2.10
(c = 0.75 in CHCl3); 1H NMR (300 MHz, CDCl3): d = 7.76–7.65
(m, 4H; HPh), 7.46–7.30 (m, 6H; HAr), 7.30–7.22 (m, 5H; HAr),
7.21–7.13 (m, 2H; HAr), 6.83 (d, J = 8.7 Hz, 2H; HPMB), 5.39 (dd,
J = 1.9 and 2.8 Hz, 1H; H2), 4.89 (d, J = 10.6 Hz, 1H; CH2Bn),
4.87 (s, 1H; H1), 4.62 (d, J = 10.9 Hz, 1H; CH2PMB), 4.57 (d, J =
10.7 Hz, 1H; CH2Bn), 4.45 (d, J = 10.9 Hz, 1H; CH2PMB), 4.03–
A mixture of mannose derivative 5 (850 mg, 1.48 mmol), dibutyltin
oxide (406 mg, 1.63 mmol) in toluene (40 mL) was refluxed under
Dean–Stark conditions for 3 h. The reaction mixture was allowed
to cool to room temperature and DMF (2 mL) was added to
the mixture. 4-Methoxybenzyl chloride (227 lL, 1.63 mmol) and
Bu4NI (602 mg, 1.63 mmol) were added, and the mixture was
heated at reflux for 3 h. Then, the mixture was diluted with EtOAc
(50 mL), washed with H2O (2 × 100 mL) and dried over MgSO4.
The solvent was removed under reduced pressure, followed by flash
chromatography on silica gel (toluene–MeOH, 94 : 6), afforded 6
1
as an oil (849 mg, 82%). [a]2D0 = +12.00 (c = 1.00 in CHCl3); H
NMR (300 MHz, CDCl3): d = 7.78–7.66 (m, 4H; 4HTBDPS), 7.47–
7.23 (m, 11H; 2HPMB and 9HAr), 6.87 (d, J = 8.6 Hz, 2H; 2HPMB),
4.94 (d, J = 1.0 Hz, 1H; H1), 4.84 (d, J = 10.9 Hz, 1H; CH2Bn), 4.63
(s, 2H; CH2PMB) 4.55 (d, J = 10.9 Hz, 1H; CH2Bn), 4.09–4.04 (m, 1H;
ꢀ
3.87 (m, 4H; H3 + H5 + 2H6), 3.87–3.77 (m, 4H; -OCH3 and H7 ),
3.74–3.66 (m, 1H; H4), 3.58 (dt, J = 10.6 and 5.1 Hz, 1H; H7), 3.34
(t, J = 5.1 Hz, 2H; 2H8), 2.83–2.61 (m, 4H; CH2CH2lev), 2.51 (s,
3H; CH3lev), 1.56 (s, 9H; CHTBDPS); 13C NMR (75 MHz, CDCl3):
d = 206.8 (COlev), 172.6 (COOlev), 159.7 (CAr), 136.3 (CAr), 136.0
(CAr), 134.1 (CAr), 133.6 (CAr), 130.5 (CAr), 130.3 (CAr), 128.8 (CAr),
128.3 (CAr), 128.1 (CAr), 128.0 (CAr), 114.2 (CAr), 98.2 (C1), 78.1
(C3), 75.7 (CH2Bn), 74.3 (C5), 73.3 (C4), 71.8 (CH2PMB), 69.3 (C2),
66.9 (C7), 63.3 (C6), 55.7 (−OCH3), 50.8 (C8), 38.4 (−CH2lev), 30.3
(CH3lev), 28.5 (−CH2lev), 27.2 (CH3TBDPS), 19.8 (CTBDPS); ESI-MS for
C44H53N3O9Si; calcd: 795.4 M+; found: 818.3 [M + Na]+; elemental
analysis calcd (%) for C44H53N3O9Si: C, 66.39%; H, 6.71%; N,
5.28%; found: C, 66.27%; H, 6.87%; N, 5.08%.
ꢀ
H2), 3.97–3.84 (m, 4H; H3 + 2H6 + H7 ), 3.84–3.77 (m, 4H; H4 +
-OCH3), 3.77–3.69 (m, 1H; H5), 3.65–3.55 (m, 1H; H7), 3.44–3.27
(m, 2H; 2H8), 1.05 (s, 9H; HTBDPS); 13C NMR (75 MHz, CDCl3):
d = 159.8 (CAr), 138.7 (CAr), 136.3 (CAr), 136.0 (CAr), 134.1 (CAr),
133.7 (CAr), 130.4 (CAr), 130.1 (CAr), 130.0 (CAr), 128.8 (CAr), 128.3
(CAr), 128.1 (CAr), 128.0 (CAr), 114.4 (CAr), 99.6 (C1), 80.2 (C3), 74.4
(CH2Bn), 73.2 (C5), 72.2 (CH2PMB), 68.7 (C2), 66.7 (C7), 63.6 (C6),
55.7 (CH3O-), 50.9 (C8), 27.2 (CH3TBDPS), 19.7 (CTBDPS); ESI-MS
for C39H47N3O7Si; calcd: 697.3 M+; found: 720.3 [M + Na]+ and
736.3 [M + K]+; elemental analysis calcd (%) for C39H47N3O7Si:
C, 66.96%; H, 6.80%; N, 5.74%; found: C, 67.12%; H, 6.79%; N,
6.02%.
Synthesis of 2-azidoethyl 4-O-benzyl-2-O-levulinoyl-3-O-(4-
methoxybenzyl)-a-D-mannopyranoside (9)
Synthesis of 2-azidoethyl 4-O-benzyl-3-O-(4-methoxybenzyl)-
a-D-mannopyranoside (7)
Acetic acid (150 lL) was added slowly to a solution of 8 (215 mg,
0.270 mmol) in dry THF (2 mL), under nitrogen at 0 ◦C. The
solution was warmed up to room temperature and TBAF 1 M in
THF (330 lL, 0.33 mmol) was added to the solution. The solution
was stirred overnight and then the solvent was removed under
vacuum. The residue was purified by flash chromatography on
silica gel (hexane–AcOEt, 1 : 2) to give compound 9 as an oil
Mannose derivative 6 (700 mg, 1.00 mmol) and TBAF (533 mg,
2.00 mmol) were dissolved in THF (7 mL). The solution was
stirred under argon during 5 hours at room temperature. The
solvent was removed under vacuum and the residue was purified
by flash chromatography on silica gel (hexane–AcOEt, 1 : 2), to
give compound 7 as an oil (436 mg, 95%). [a]2D0 = +34.5 (c = 0.75
1
1
in CHCl3); H NMR (500 MHz, CDCl3): d = 7.33–7.24 (m, 6H;
(150 mg, quant.). [a]D20 = +13.02 (c = 0.4 in CHCl3); H NMR
HAr), 6.84 (d, 2H; 2HPMB), 4.90 (d, J = 1.5 Hz, 1H; H1), 4.86 (d,
J = 11.0 Hz, 1H; CH2PMB), 4.63 (d, J = 11.0 Hz, 1H; CH2PMB),
4.60 (s, 2H; CH2Bn), 4.01 (dd, J = 1.5 and 3.1 Hz, 1H; H2), 3.88
(dd, J = 3.1 and 9.0 Hz, 1H; H3) 3.86–3.79 (m, 3H; H4 and 2H6),
(300 MHz, CDCl3): d = 7.38–7.21 (m, 7H; 5HBn and 2HPMB), 6.84
(d, J = 8.7 Hz, 2H; 2HPMB), 5.34 (dd, J = 3.1 and 1.8 Hz, 1H;
H2), 4.90 (d, J = 10.9 Hz, 1H; CH2Bn), 4.86 (d, J = 1.6 Hz, 1H;
H1), 4.61 (d, J = 10.9 Hz, 1H; CH2Bn), 4.61 (d, J = 10.9 Hz,
1H; CH2Bn), 4.45 (d, J = 10.9 Hz, 1H; CH2Bn), 4.00 (dd, J = 3.3
and 9.1 Hz, 1H; H3), 3.89–3.66 (m, 5H; H4 + H5 + 2H6 + H7),
ꢀ
3.78 (s, 3H; -OCH3), 3.75 (dd, J = 3.9 and 12.3 Hz, 1H; H7 ),
3.58 (dt, J = 3.9 and 9.9 Hz, 1H; H7), 3.40–3.29 (m, 2H; 2H8); 13
C
ꢀ
NMR (CDCl3, 125 MHz): d = 159.5 (CAr), 138.2 (CAr), 129.9 (CAr),
3.59 (dt, J = 10.3 and 4.9 Hz, 1H; H7 ), 3.37 (t, J = 4.9 Hz, 2H;
2750 | Org. Biomol. Chem., 2008, 6, 2743–2754
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
2H8), 2.87–2.57 (m, 4H; CH2CH2lev), 2.18 (s, 3H; CH3lev); 13C NMR
(75 MHz, CDCl3): d = 206.9 (COlev), 172.5 (COOlev), 159.7 (CPMB),
138.6 (CAr), 130.4 (CAr), 130.2 (CAr), 128.9 (CAr), 128.5 (CAr), 128.3
(CAr), 114.2 (CAr), 98.2 (C1), 77.8 (C3), 75.6 (CH2Bn), 74.2 (C4 or
C5), 72.5 (C4 or C5), 71.8 (CH2Bn), 69.2 (C2), 67.2 (C7), 62.5 (C6),
55.7 (-OCH3), 50.8 (C8), 38.5 (CH2lev), 30.2 (CH3lev), 28.5 (CH2lev);
ESI-MS for C28H35N3O9; calcd: 557.2 M+; found: 596.1 [M + Na]+
and 580.1 [M + K]+; elemental analysis calcd (%) for C28H35N3O9:
C, 59.66%; H, 6.12%; N, 7.73%; found: C, 59.84%; H, 6.00%; N,
7.74%.
3.5 Hz and 1.5 Hz, 1H; H2), 5.22–5.18 (m, 2H; H2 and H1), 5.03
(d, J = 1.5 Hz, 1H; H1), 4.89 (d, J = 11.0 Hz, 1H; 1HBn), 4.86 (d,
J = 11.0 Hz, 1H; 1HBn), 4.79 (s, 2H; 2HBn), 4.75 (d, J = 11.0 Hz,
1H; 1HBn), 4.73 (d, J = 11.0 Hz, 1H; 1HBn), 4.72 (d, J = 11.0 Hz,
1H; 1HBn), 4.69 (d, J = 1.5 Hz, 1H; H1), 4.62–4.47 (m, 7H; 7HBn),
4.24 (dd, J = 3.8 Hz and 9.4 Hz, 1H; H3), 4.01–4.67 (m, 15H;
H4A, H5A, 2H6A, H3B, H4B, H5B, 2H6B, H3C, H4C, H5C, 2H6C and H7),
3.56–5.51 (m, 1H; H7), 3.33–3.27 (m, 2H; 2H8), 2.74–2.61 (m, 4H;
2 × -CH2lev), 2.20 (s, 3H; CH3lev), 2.15 (s, 3H; -OOCCH3), 2.13 (s,
3H; -OOCCH3); 13C NMR (125 MHz, CDCl3): d = 208.1 (C O),
=
=
=
=
174.5 (C O), 172.6 (C O), 172.4 (C O), 141.0 (CAr), 140.7 (CAr),
140.5 (CAr), 140.3 (CAr), 140.2 (CAr), 139.9 (CAr), 130.7 (CAr),130.7
(CAr), 130.6 (CAr), 130.5 (CAr), 130.4 (CAr), 130.3 (CAr), 130.2 (CAr),
130.1 (CAr), 130.0 (CAr), 129.8 (CAr), 129.7 (CAr), 102.4 (C1), 100.9
(C1), 99.4 (C1), 80.2, 80.0, 76.7, 76.5, 76.3, 75.7, 75.7, 74.6, 74.2,
Synthesis of 2-azidoethyl 4-O-benzyl-2-O-levulinoyl-
a-D-mannopyranoside (10)
Mannose derivative 9 (74 mg, 0.126 mmol) was dissolved in
CH2Cl2 (5 mL), the solution was cooled at −20 ◦C and TFA
(1.5 mL) was added to the solution. The reaction mixture was
stirred during 20 min. Then, ethanol (0.5 mL) and CH2Cl2
(4 mL) were added to the reaction. The solution was washed
with NaHCO3 sat. (2 × 10 mL), NaCl sat. (10 mL), and dried
with MgSO4. The solvent was removed under vacuum and the
residue was purified by flash chromatography on silica gel (hexane–
74.1, 73.8, 73.6, 71.1, 71.0, 70.8, 70.7, 69.0, 67.6 (C2A + C3A
C4A + C5A + C6A, C2B +C3B + C4B + C5B + C6B + C2C + C3C + C4C
+
+
C5C + C6C) 52.5 (C8), 40.1 (CH2lev), 32.0 (CH2lev), 30.3 (CH3lev), 23.4
(OCOCH3), 23.3 (OCOCH3); ESI-MS for C78H87N3O20; calcd:
1385,6 M+; found: 1408.6 [M + Na]+ and 1424.4 [M + K]+.
Synthesis of 2-azidoethyl O-(2-O-acetyl-3,4,6-tri-O-benzyl-
a-D-mannopyranosyl-(1→2)-O-[2-O-acetyl-3,4,6-tri-O-
benzyl-a-D-mannopyranosyl-(1→6)]-4-O-benzyl-
3-O-(4-methoxybenzyl)-a-D-mannopyranoside (12)
AcOEt, 2 : 1) to give compound 10 as and oil (53 mg, 97%). [a]D20
=
1
+32.91 (c = 0.70 in CHCl3); H NMR (300 MHz, CDCl3): d =
7.40–7.31 (m, 4H; 4HBn), 7.31–7.25 (m, 1H; HBn), 5.13 (dd, J = 2.0
and 3.3 Hz, 1H; H2), 4.87 (d, J = 10.9 Hz, 1H; CH2Bn), 4.82 (d, J =
1.2 Hz, 1H; H1), 4.71 (d, J = 10.9 Hz, 1H; CH2Bn), 4.19–4.13 (m,
Mannose derivative 7 (49 mg, 0.107 mmol), and the glycosyl donor
13 (163 mg, 0.278 mmol) were dissolved in dry CH2Cl2 (2 mL) in
ꢀ
1H; H3), 3.89–3.76 (m, 3H; 2H6 and H7 ), 3.75–3.64 (m, 2H; H4
˚
and H5), 3.61–3.55 (m, 1H; H7), 3.42–3.32 (m, 2H; 2H8), 2.84–2.75
the presence of 4 A molecular sieves. The mixture was stirred
(m, 2H; CH2lev), 2.67–2.59 (m, 2H; CH2lev), 2.19 (s, 3H; CH3lev); 13
C
during 2 hours at room temperature under argon atmosphere.
Then, the system was cooled at −20 ◦C, NIS (63 mg, 0.278 mmol)
and TfOH (2.5 lL, 0.026 mmol) were added and the mixture
was stirred during 20 min. The reaction was monitored by TLC
(toluene–MeOH, 9 : 1) and when it was completed, NaHCO3 sat.
(50 lL) was added to quench the TfOH. The reaction mixture was
diluted with CH2Cl2 (6 mL), filtered over a pad of Celite, washed
with Na2S2O3 sat., dried over MgSO4 and the solvent was removed
under vacuum. The residue was purified by flash chromatography
on silica gel (hexane : AcOEt, 7 : 1) to give compound 12 as an
NMR (75 MHz, CDCl3): d = 207.1 (COlev), 170.1 (COOlev), 133.7
(CAr), 133.2 (CAr), 130.6 (CAr), 129.6 (CAr), 127.5 (CAr), 127.4 (CAr),
97.5 (C1), 75.7 (C4 or C5), 72.7 (C2), 72.0 (C4 or C5), 70.1 (C-3), 66.6
(C7), 61.7 (C6), 50.4 (C8), 38.4 (CH2lev), 29.7 (CH3lev), 28.4 (CH2lev);
ESI-MS for C20H27N3O8; calcd: 437.2 M+; found: 460.2 [M + Na]+;
elemental analysis calcd (%) for C20H27N3O8: C, 53.90%; H, 5.95%;
N, 9.92%; found: C, 53.99%; H, 5.75%; N, 10.14%.
Synthesis of 2-azidoethyl O-(2-O-acetyl-3,4,6-tri-O-benzyl-a-D-
mannopyranosyl-(1→3)-O-[2-O-acetyl-3,4,6-tri-O-benzyl-
a-D-mannopyranosyl-(1→6)]-4-O-benzyl-2-O-levulinoyl-
a-D-mannopyranoside (11)
1
oil (88 mg, 70%). [a]2D0 = +17.05 (c = 1.00 in CHCl3); H NMR
(500 MHz, CDCl3): d = 7.32–7.21 (m, 32H; 18HArBn + 2HArPMB),
7.16–7.13 (m, 2H; 2HArBn), 7.12–7.08 (m, 2H; 2HArBn), 6.80 (d, J =
9.0 Hz, 2H; 2HArPMB), 5.51 (dd, J = 3.5 Hz and 1.5 Hz, 1H; H2B),
5.34 (dd, J = 2.0 Hz and 3.0 Hz, 1H; H2C), 4.99 (d, J = 1.5 Hz, 1H;
H1B), 4.89 (d, J = 1.5 Hz, 1H; H1C), 4.82 (d, J = 11.0 Hz, 1H; HBn),
4.81 (d, J = 11.0 Hz, 2H; 2HBn), 4.81 (d, J = 1.5 Hz, 1H; H1A), 4.79
(d, J = 11.0 Hz, 1H, HBn), 4.61 (d, J = 11.0 Hz, 1H; HBn), 4.61 (d,
J = 11.0 Hz, 1H; HBn), 4.56 (s, 2H; 2HBn), 4.50 (d, J = 11.0 Hz, 1H;
HBn), 4.49 (d, 1H, J = 11.0 Hz, 1H; HBn), 4.46–4.40 (m, 5H, 5HBn),
3.95 (dd, 1H J = 3.2 Hz and 9.3 Hz, 1H; H3A), 3.94–3.60 (m, 15 H;
H2A, H4A, H5A, 2H6A, H3B, H4B, H5B, 2H6B, H3C, H4C, H5C and 2H6C),
3.76 (s, 3H, -OCH3), 3.52 (t, J = 9.3 Hz, 1H; H7), 3.32–3.28 (m,
1H; H7), 3.20–3.11 (m, 2H; 2H8), 2.10 (s, 3H; -OOCCH3), 2.09 (s,
Mannose derivative 10 (40 mg, 0.092 mmol), and the glycosyl
donor 13 (146 mg, 0.249 mmol) were dissolved in dry CH2Cl2
˚
(2 mL) in the presence of 4 A molecular sieves. The mixture
was stirred during 2 hour at room temperature under argon
atmosphere. Then, the system was cooled at −20 ◦C, NIS (56 mg,
0.249 mmol) and TfOH (2.5 lL, 0.026 mmol) were added and the
mixture was stirred during 20 min. The reaction was monitored by
TLC (toluene–MeOH, 9 : 1) and when it was completed, NaHCO3
sat. (50 lL) was added to quench the TfOH. The reaction mixture
was diluted with CH2Cl2 (6 mL), filtered over a pad of Celite,
washed with Na2S2O3 sat., dried over MgSO4 and the solvent
was removed under vacuum. The residue was purified by flash
chromatography on silica gel (hexane : AcOEt, 7 : 1) to give
compound 11 as an oil (91 mg, 72%). [a]2D0 = +45.76 (c = 1.00
in CHCl3); 1H NMR (500 MHz, CDCl3): d = 7.47–7.16 (m, 32H;
HArBn), 5.49 (dd, J = 3.5 Hz and 1.5 Hz, 1H; H2), 4.47 (dd, J =
3H; -OOCCH3); 13C NMR (100 MHz, CDCl3): d = 170.3 (C O),
=
=
170.07 (C O), 159.2 (CAr), 138.6 (CAr), 138.4 (CAr), 138.3 (CAr),
138.2 (CAr), 138.0 (CAr), 137.9 (CAr), 130.3 (CAr), 129.3 (CAr), 128.4
(CAr), 128.3 (CAr), 128.3 (CAr), 128.2 (CAr), 128.1 (CAr), 128.0 (CAr),
128.0 (CAr), 127.9 (CAr), 127.8 (CAr), 127.6 (CAr), 127.6 (CAr), 127.5
(CAr), 113.8 (CAr), 99.8 (C1), 98.7 (C1), 97.1 (C1), 79.4, 78.1, 78.0,
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 2743–2754 | 2751
©

View Article Online
77.2, 75.2, 75.1, 74.9, 74.4, 73.5, 73.4, 71.9, 71.5, 71.4, 70.8, 69.2,
68.7, 68.5, 68.5, 66.3, 55.2 (C2A + C3A + C4A + C5A + C6A + C2B
C3B + C4B + C5B + C6B + C2C + C3C + C4C + C5C + C6C + C7), 50.3
(C8), 21.1 (OCOCH3), 21.1 (OCOCH3); ESI-MS for C81H89N3O19;
calcd: 1407.6 M+; found: 1446.6 [M + K]+.
suite.26 The starting coordinate of human DC-SIGN was taken
from the Protein Data Bank with code 1k9i.12 The structure
was edited to contain only one protein monomer together with
calcium ions, protein hydrogen atoms were added, the partial
charges were calculated using AMBER27,28 procedure as imple-
mented in SYBYL and the calcium ions were given a charge of
(+2). Atom types and charges for oligosaccharides were defined
using the PIM parameters developed for carbohydrates.29 All
ligand structures were obtained from different protein complex:
dimannoside Mana1,2Man (ManC-ManA) from 1i3h;30 diman-
nosides Mana1,3Man (ManD-ManA) and Mana1,6Man (ManB-
ManA), trimannoside Mana1,2[Mana1,6]Man, (ManC-ManA-
ManB) were built from Mana1,2Man using the Sketch module
of SYBYL; and trimannoside Mana1,3[Mana1,6]Man (ManD-
ManA-ManB) was built and fitted into DC-SIGN using the
correspond ligand (GlcNAc2Man3) from 1k9i.
The nomenclature used for the theoretical complex (1d13,
1d12, 1d16, 1t26, 1t36) does mention if it is a dimanno-
side (d) a trimannoside (t) and the glycoside bond 1→3,
1→2, 1→6 in Mana1,3Man, Mana1,2Man and Mana1,6Man,
respectively and, 1→2–6←1, 1→3–6←1 in the trisaccharide
Mana1,2[Mana1,6]Man and Mana1,3[Mana1,6]Man. The dihe-
dral angles U, w and x are defined, following IUPAC definition,
+
Synthesis of 2-aminoethyl O-a-D-mannopyranosyl(1→2)[O-a-D-
mannopyranosyl(1→6)]-a-D-mannopyranoside (1)
The protected trisaccharide 12 (70 mg, 0.06 mmol) was dissolved
in dry MeOH (1 mL) and NaOMe solution (1 M in MeOH) (60 ll,
0.06 mmol) was added. The reaction mixture was stirred at room
temperature for 30 min. Then, the mixture was neutralized with
Amberlite IR 120 resin, filtered and the solvent was removed under
vacuum. The white solid was dissolved in MeOH (3 mL), Pd–C
10% (cat.) was added and the reaction mixture was hydrogenated
under H2 (1 bar) until reduction was complete (monitored by TLC
i-PrOH : H2O 7 : 3 + 1% AcOH) to give trisaccharide 1 as a white
solid (32 mg, 98% over two steps). [a]2D0 = +78.2 (c = 0.50 in H2O);
1H NMR (500 MHz, D2O): d = 5.10 (s, 1H; H1A), 5.00 (s, 1H;
H1C), 4.90 (s, 1H; H1B), 4.06 (m, 1H; H2C), 3.99 (m, 1H; H2A), 3.97
(m, 1H; H6A), 3.96 (m, 1H; H2B), 3.91 (m, 1H; H3A), 3.90 (m, 1H;
H7), 3.87 (m, 1H; H6C), 3.81 (m, 1H; H4A), 3.80 (m, 1H; H3B), 3.77
(m, 1H; H3C), 3.73 (m, 1H; H6A), 3.73 (m, 1H; 2H6B), 3.72 (m, 1H;
H5A), 3.71 (m, 1H; H5C), 3.69 (m, 1H; H6C), 3.65 (m, 1H; H7), 3.65
(m, 1H; H5B), 3.64 (m, 1H; H4B), 3.61 (m, 1H; H4C), 3.17 (m, 1H;
2H8). 13C NMR (125 MHz, D2O): d = 102.4 (C1C), 99.5 (C1B), 98.4
(C1A), 78.7 (C2A), 73.3 (C5C), 72.8 (C5B), 70.5 (C3B), 70.5 (C3C), 71.4
(C5A), 70.1 (C3A), 69.9 (C2C), 69.9 (C2B), 66.7 (C4C), 66.6 (C4B), 66.4
(C4A), 65.1 (C6A), 64.5 (C7), 61.0 (C6C), 60.9 (C6B), 39.1 (C8).
ꢀ
ꢀ
ꢀ
as O5–C1–O1–Cx , C1–O1–Cx –Cx + 1 and O5–C5–C6–O6 respec-
ꢀ
ꢀ
tively, where Cx and Cx + 1 are the aglyconic atoms. For a-D-Manp-
(1–6)-a-D-Manp, the U is defined as O5–C1–O1–C6ꢀ, w C1–O1–
C6ꢀ–C5ꢀ and x O1–C6ꢀ–C5ꢀ–ORꢀ where ORꢀ is the endocyclic ring
oxygen.
To build the complexes, the ligands were first manually po-
sitioned within the binding pocket of the protein taking into
account the experimental data about the binding of 1k9i. Two
binding modes were examined for each of the disaccharides
Mana1,6Man and Mana1,2Man: one refers to the primary
binding site (complexes 1d12 and 1d16, respectively), which cor-
responds to the glycosidic bonds at the 3-position (Mana1,3Man
in the pentasaccharide GlcNAc2Man3; ManD-ManA), and the
other involves the secondary binding site (complexes 2d12 and
2d16, respectively), corresponding to the glycosidic bonds at 6-
position (Mana1,6Man in the pentasaccharide GlcNAc2Man3:
ManB-ManA). Regarding the disaccharide Mana1,2Man, it was
only considered the primary binding site (1d13). Subsequent
energy minimization was performed using the AMBER99 force
field with geometry optimization of the sugar and the side chains
of the protein. Energy minimizations were carried out using
the Conjugate Gradient procedure until a gradient deviation of
Synthesis of 2-aminoethyl O-a-D-mannopyranosyl(1→3)[O-a-D-
mannopyranosyl(1→6)]-a-D-mannopyranoside (2)
The protected trisaccharide 11 (70 mg, 0.05 mmol) was dissolved
in dry MeOH (2 mL) and NaOMe solution (1 M in MeOH) (50 ll,
0.05 mmol) was added. The reaction mixture was stirred at room
temperature for 30 min. Then, the mixture was neutralized with
Amberlite IR 120 resin, filtered and the solvent was removed under
vacuum. The white solid was dissolved in MeOH (3 mL), Pd–C
10% (cat.) was added and the reaction mixture was hydrogenated
under H2 (1 bar) until reduction was complete (monitored by TLC
i-PrOH : H2O 7 : 3 + 1% AcOH) to give trisaccharide 2 as a white
solid (27 mg, 97% over two steps). [a]2D0 = +84.0 (c = 0.50 in H2O);
1H NMR (500 MHz, D2O): d = 5.10 (m, 1H; H1B), 4.90 (s, 1H;
H1D), 4.85 (s, 1H; H1A), 4.15 (m, 1H; H2A), 4.06 (m, 1H; H2B), 4.00
(m, 1H; H6A), 3.98 (m, 1H; H2D), 3.96 (m, 1H; H7), 3.93 (m, 1H;
H3A), 3.92 (m, 1H; H4A), 3.89 (m, 1H; H6A), 3.87 (m, 1H; H3B), 3.82
(m, 1H; H3D), 3.79 (m, 1H; H5A), 3.77 (m, 1H; H5B), 3.75 (m, 1H;
H6A), 3.75 (m, 1H; H6D), 3.70 (m, 1H; H6A), 3.70 (m, 1H; H7), 3.67
(m, 1H; H4D), 3.66 (m, 1H; H5D), 3.61 (m, 1H; H4B), 3.24 (m, 1H;
2H8); 13C NMR (125 MHz, D2O): d = 102.4 (C1B), 99.8 (C1A), 99.3
(C1D), 78.5 (C3A), 73.4 (C5B), 72.6 (C5D), 71.1 (C5A), 70.7 (C3D), 70.3
(C3B),70.1 (C2B), 70.0 (C2D), 69.4 (C2A), 66.6 (C4D), 66.6 (C4B), 65.3
(C4A), 65.2 (C6A), 63.6 (C7),60.8 (C6D), 60.8 (C6B), 39.1 (C8).
0.005 kcal mol−1
−1 was attained. A distance-dependent dielectric
˚
A
constant was used in the calculations.
These complexes were used as input structure for docking
studies using FlexiDock command in the Biopolymer module.
FlexiDock software performs flexible docking of conformationally
flexible ligands into receptor binding sites and provides control of
ligand binding characteristics, taking into account rigid, partially
flexible, or fully flexible receptor side chains. FlexiDock incorpo-
rates the Van der Waals, electrostatic, torsional and constraint
energy terms of the Tripos force field, and it uses a genetic
algorithm to determine the optimum ligand geometry. Genetic
algorithms31 (GA) borrow methodology and terminology from
biological (or Darwinian) evolution, in which they are an iterative
process where the most-fit members of a population will have the
Computational studies
All calculations were performed on a Silicon Graphics Octane
workstation (R12000, 300 MHz) using the SYBYL 6.9 program
2752 | Org. Biomol. Chem., 2008, 6, 2743–2754
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
best chance of propagating themselves into future generations.
During the flexible docking analysis the residues involved in the
raw data were multiplied by shifted square sine window function
prior to Fourier transform, and the baseline was corrected using
polynomial fitting.
˚
binding site (sphere of 7 A around the ligand), and the ligands were
considered flexible. The default SYBYL FlexiDock parameters
were utilized in all cases. Five runs of FlexiDock were performed
searching for the different binding modes with iterations set to
500–1000 generations per gene, where gene is the number of
rotatable bonds plus six, obtaining a series of model complexes.
The output complexes of FlexiDock were analyzed on the basis
of the score provided by FlexiDock, the dihedral angles U, w, x
and the distances of key residues. The conformations with typical
dihedral angles of each glycosidic bond at positions 2, 3, and 6
were clustered and the rest of the non-typical conformations were
eliminated. A member of each family was energy minimized. These
output complexes were again optimized until gradient 0.01 kcal
STD experiments were performed at 278 K using water-
gate solvent suppression at 0.5, 0.75, 1.0, 1.25, 1.5 and 2.0 s
saturation times using a train of Gaussian shaped pulses of
49 ms and 100–60 Hz power spaced by 1.0 ms delays.23 On-
resonance irradiation was performed at 0.9 ppm, appropriate
blank experiments were also performed to assure the absence of
direct irradiation on the ligand.
DC-SIGN EC expression and purification
Plasmids pET30b (Novagen) containing cDNA encoding the
EctoDomain ECD (corresponding to amino acids 66–404) of
DC-SIGN were used for overproduction as described previously.5
Proteins produced in inclusion bodies have been refolded as
already described.42 Purification of functional DC-SIGN proteins
were achieved by an affinity chromatography on Mannan-agarose
column (Sigma) equilibrated in 25 mM Tris-HCl pH 7.8, 150 mM
NaCl, 4 mM CaCl2 (Buffer A) and eluted in same buffer
lacking CaCl2 but supplemented with 10 mM EDTA. This step
was followed by a superose 6 size exclusion chromatography
equilibrated in buffer A. Protein was concentrated to 9 mg ml−1 and
dialysed three times against the deuterated buffer 25 mM Tris DCl,
150 mM NaCl, 4 mM CaCl2 at pD 7.8 in D2O (deuterated Tris-
d11 (98%) was purchased from Cambridge Laboratories Inc. and
the D2O from Spectra Stable Isotopes). Protein was then stored in
liquid nitrogen.
mol−1 −1 using the above mentioned conditions, the conformation
A
˚
with the lowest score and fulfilling the experimental requirements
being chosen. The binding free energies (DGbind) of all disaccharides
and trisaccharides complexes were calculated with the Structural
Thermodynamics Calculations V4.3 (STC) program32 in order
to predict the interaction energy of each complex. The STC
2
˚
program calculates the DGbind from a parameterization (per A
of polar and nonpolar ASA) of the heat capacity, enthalpy, and
solvation entropy obtained from a global fit of structural and
thermodynamic database of globular proteins. Carbohydrate–
protein interactions were calculated with the LPC program.33 As
selection criteria used to identify C–H · · · p interaction, we applied
those used by Brandl34 and Kerzmann.35 Fig. 7 shows the three
values dcx, achx and dHpX. The limits are dcx< 4.5 A, achx > 110◦ and
˚
˚
dHpX < 2.0 A.
Acknowledgements
We thank Fondo de Investigacio´n Sanitaria (FIS) PI030093 (JR)
for financial support. GT is supported by the CEA for his PhD
appointment and FF thanks Bill and Melinda Gates foundation
(pediatric dengue vaccine initiative) for funding support and NEC
and JAP thanks MCyT (SAF 2003–08003-CO2) and MEC (SAF
2006–13391-CO3) for funding support.
Fig. 7 Geometry of the C–H · · · p interaction.
References
1 Y. van Kooyk and T. B. Geijtenbeek, Nat. Rev. Immunol., 2003, 3,
697–709.
2 S. Pohlmann, F. Baribaud and R. W. Doms, Trends Immunol., 2001,
22, 643–646.
NMR
NMR sample preparation. NMR experiments were recorded
3 T. B. Geijtenbeek, D. S. Kwon, R. Torensma, S. J. van Vliet, G. C.
van Duijnhoven, J. Middel, I. L. Cornelissen, H. S. Nottet, V. N.
Kewalramani, D. R. Littman, C. G. Figdor and Y. van Kooyk, Cell,
2000, 100, 587–597.
4 G. Tabarani, J. J. Reina, C. Ebel, C. Vives, H. Lortat-Jacob, J. Rojo and
F. Fieschi, FEBS Lett., 2006, 580, 2402–2408.
5 F. Halary, A. Amara, H. Lortat-Jacob, M. Messerle, T. Delaunay, C.
Houles, F. Fieschi, F. Arenzana-Seisdedos, J. F. Moreau and J. Chanet-
Merville, Immunity, 2002, 17, 653–664.
6 P. Y. Lozach, H. Lortat-Jacob, L. A. de Lacroix, I. Staropoli, S. Foung,
A. Amara, C. Houles, F. Fieschi, O. Schwartz, J. L. Virelizier, F.
Arenzana-Seisdedos and R. Altmeyer, J. Biol. Chem., 2003, 278, 20358–
20366.
7 A. de Parseval, S. V. Su, J. H. Elder and B. Lee, J. Virol., 2004, 78,
2597–2600.
8 S. V. Su, P. Hong, S. Baik, O. A. Negrete, K. B. Gurney and B. Lee,
J. Biol. Chem., 2004, 279, 19122–19132.
9 E. W. Adams, D. M. Ratner, H. R. Bokesch, J. B. McMahon, B. R.
O’Keefe and P. H. Seeberger, Chem. Biol., 2004, 11, 875–881.
in AVANCE Bruker instruments operating at 500.13 MHz, at
288 K. Samples of free 1 and 2 were prepared in 150 mM NaCl,
25 mM d-TRIS (pH* = 8.1) in D2O after three cycles of deuterium
exchange by evaporation and resolution of the trisaccharide in
D2O. The samples in the presence of DC-SIGN EC were prepared
using 2 mM of carbohydrate and 40 lM of lectin, assuming a
monomeric state, in D2O (150 mM NaCl, 25 mM d-TRIS, pH* =
8.1).
NMR experiments were performed using manufacturer pulse
sequences: dqf-COSY,36 TOCSY,37 NOESY,38 ROESY,39 and
HSQC40 implemented with z-pulsed field gradients when possible,
and acquired using time proportional phase incrementation
mode.41 The size of the acquisition data matrix for homonuclear
experiments was typically of 2k × 512 in F2 and F1 dimensions,
respectively. Data were processed using manufacturer software,
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 2743–2754 | 2753
©

View Article Online
10 E. van Liempt, C. M. C. Bank, P. Mehta, J. J. Garcia-Vallejo, Z. S.
Kawar, R. Geyer, R. A. Alvarez, R. D. Cummings, Y. van Kooyk and
I. van Die, FEBS Lett., 2006, 580, 6123–6131.
11 S. Meyer, E. van Liempt, A. Imberty, Y. van Kooyk, H. Geyer, R. Geyer
and I. van Die, J. Biol. Chem., 2005, 280, 37349–37359.
Eur. J. Org. Chem., 2007, 1577–1585; C. Clavel, A. Canales, G. Gupta,
J. I. Santos, F. J. Canada, S. Penades, A. Surolia and J. Jimenez-Barbero,
Glycoconjugate J., 2007, 24, 449–464; B. Hakkarainen, L. Kenne, M.
Lahmann, S. Oscarson and C. Sandstrom, Magn. Reson. Chem., 2007,
45, 1076–1080.
12 H. Feinberg, D. A. Mitchell, K. Drickamer and W. I. Weis, Science,
2001, 294, 2163–2166.
13 Y. Guo, H. Feinberg, E. Conroy, D. A. Mitchell, R. Alvarez, O. Blixt,
M. E. Taylor, W. I. Weis and K. Drickamer, Nat. Struct. Mol. Biol.,
2004, 11, 591–598.
14 H. Feinberg, R. Castelli, K. Drickamer, P. H. Seeberger and W. I. Weis,
J. Biol. Chem., 2007, 282, 4202–4209.
22 M. Frank, T. Lutteke and C. W. von der Lieth, Nucleic Acids Res., 2007,
35, 287–290; http://www.cermav.cnrs.fr/glyco3d/index.php.
23 M. Mayer and B. Meyer, Angew. Chem., Int. Ed., 1999, 38, 1784–1788.
24 B. Meyer and T. Peters, Angew. Chem., Int. Ed., 2003, 42, 864–890;
V. Jayalakshmi and N. R. Krishna, J. Magn. Reson., 2004, 168, 36–45;
N. R. Krishna and V. Jayalakshmi, Prog. Nucl. Magn. Reson. Spectrosc.,
2006, 49, 1–25.
15 J. J. Reina, S. Sattin, D. Invernizzi, S. Mari, L. Mart´ınez-Prats, G.
Tabarani, F. Fieschi, R. Delgado, P. M. Nieto, J. Rojo and A. Bernardi,
ChemMedChem, 2007, 2, 1030–1036.
16 L. Tailleux, O. Schwartz, J. L. Herrmann, E. Pivert, M. Jackson, A.
Amara, Legres, D. Dreher, L. P. Nicod, J. C. Gluckman, P. H. Lagrange,
B. Gicquel and O. Neyrolles, J. Exp. Med., 2003, 197, 121–127.
17 N. Maeda, J. Nigou, J. L. Herrmann, M. Jackson, A. Amara, P. H.
Lagrange, G. Puzo, B. Gicquel and O. Neyrolles, J. Biol. Chem., 2003,
278, 5513–5516.
18 Y. van Kooyk, B. Appelmelk and T. B. Geijtenbeek, Trends Mol. Med.,
2003, 9, 153–159.
19 E. A. Koppel, I. S. Ludwig, M. S. Hernandez, T. L. Lowary, R. R.
Gadikota, A. B. Tuzikov, C. M. Vandenbroucke-Grauls, Y. van Kooyk,
B. J. Appelmelk and T. B. Geijtenbeek, Immunobiology, 2004, 209, 117–
127.
20 J. J. Reina and J. Rojo, Tetrahedron Lett., 2006, 47, 2475–2478.
21 J. R. Brisson and J. P. Carver, Biochemistry, 1983, 22, 1362–1368;
R. J. Woods, A. Pathiaseril, M. R. Wormald, C. J. Edge and R. A.
Dwek, Eur. J. Biochem., 1998, 258, 372–386; E. W. Sayers and J. H.
Prestegard, Biophys. J., 2000, 79, 3313–3329; M. Martin-Lomas, N.
Khiar, S. Garcia, J. L. Koessler, P. M. Nieto and T. W. Rademacher,
Chem.–Eur. J., 2000, 6, 3608–3621; L. Gonzalez, M. Bruix, T. Diaz-
Maurino, T. Feizi, M. Rico, D. Solis and J. Jimenez-Barbero, Arch.
Biochem. Biophys., 2000, 383, 17–27; A. Almond and J. U. Duus,
J. Biomol. NMR, 2001, 20, 351–363; A. Almond, J. Bunkenborg, T.
Franch, C. H. Gotfredsen and J. O. Duus, J. Am. Chem. Soc., 2001,
123, 4792–4802; L. M. Mikkelsen, M. J. Hernaiz, M. Martin-Pastor,
T. Skrydstrup and J. Jimenez-Barbero, J. Am. Chem. Soc., 2002, 124,
14940–14951; E. W. Sayers and J. H. Prestegard, Biophys. J., 2002, 82,
2683–2699; K. Lycknert, A. Helander, S. Oscarson, L. Kenne and G.
Widmalm, Carbohydr. Res., 2004, 339, 1331–1338; S. Mari, H. Posteri,
G. Marcou, D. Potenza, F. Micheli, F. J. Canada, J. Jimenez-Barbero
and A. Bernardi, Eur. J. Org. Chem., 2004, 5119–5125; F. Chevalier,
J. Lopez-Prados, P. Groves, S. Perez, M. Martin-Lomas and P. M.
Nieto, Glycobiology, 2006, 16, 969–980; C. Clavel, A. Canales, G.
Gupta, F. J. Canada, S. Penades, A. Surolia and J. Jimenez-Barbero,
25 J. Jimenez-Barbero and T. Peters, in NMR Spectroscopy of Glycocon-
jugates, ed. J. Jimenez-Barbero and T. Peters, Wiley-VCH, Weinheim,
2003, ch. 12, pp. 289–310.
26 SYBYL 7.0, Tripos Inc.,1699 South Hanley Rd., St.Louis, Missouri,
63144, USA. 2007.
27 W. D. Cornell, P. Cieplak, C. I. Bayly, I. R. Gould, K. M. Merz, D. M.
Ferguson, D. C. Spellmeyer, T. Fox, J. W. Caldwell and P. A. Kollman,
J. Am. Chem. Soc., 1995, 117, 5179–5197.
28 S. J. Weiner, P. A. Kollman, D. T. Nguyen and D. A. Case, J. Comput.
Chem., 1986, 7, 230–252.
29 A. Imberty, E. Mikros, J. Koca, R. Mollicone, R. Oriol and S. Perez,
Glycoconjugate J., 1995, 12, 331–349.
30 D. A. R. Sanders, D. N. Moothoo, J. Raftery, A. J. Howard, J. R.
Helliwell and J. H. Naismith, J. Mol. Biol., 2001, 310, 875–884.
31 O. P. Judson and D. Haydon, J. Mol. Evol., 1999, 49, 539–550.
32 P. Lavigne, J. R. Bagu, R. Boyko, L. Willard, C. F. B. Holmes and B. D.
Sykes, Protein Sci., 2000, 9, 252–264.
33 V. Sobolev, A. Sorokine, J. Prilusky, E. E. Abola and M. Edelman,
Bioinformatics, 1999, 15, 327–332.
34 M. Brandl, M. S. Weiss, A. Jabs, J. Suhnel and R. Hilgenfeld, J. Mol.
Biol., 2001, 307, 357–377.
35 A. Kerzmann, D. Neumann and O. Kohlbacher, J. Chem. Inf. Mod.,
2006, 46, 1635–1642.
36 A. L. Davis, E. D. Laue, J. Keeler, D. Moskau and J. Lohman, J. Magn.
Reson., 1991, 94, 637–644.
37 A. Bax and D. G. Davis, J. Magn. Reson., 1985, 65, 355–360.
38 R. Wagner and S. Berger, J. Magn. Reson., Ser. A, 1996, 123, 119–121.
39 A. A. Bothnerby, R. L. Stephens, J. M. Lee, C. D. Warren and R. W.
Jeanloz, J. Am. Chem. Soc., 1984, 106, 811–813.
40 J. Schleucher, M. Schwendinger, M. Sattler, P. Schmidt, O. Schedletzky,
S. J. Glaser, O. W. Sorensen and C. Griesinger, J. Biomol. NMR, 1994,
4, 301–306.
41 D. Marion and K. Wuthrich, Biochem. Biophys. Res. Commun., 1983,
113, 967–974.
42 D. A. Mitchell, A. J. Fadden and K. Drickamer, J. Biol. Chem., 2001,
276, 28939–28945.
2754 | Org. Biomol. Chem., 2008, 6, 2743–2754
This journal is
The Royal Society of Chemistry 2008
©