Synthesis and cholera toxin binding properties of multivalent GM1 mimics

Article abstract of DOI:10.1039/b405344c

Dendrimers based on the 3,5-di-(2-aminoethoxy)-benzoic acid branching unit were used to attach multiple copies of a GM1 mimic for inhibition of cholera toxin binding. Systems up to octavalent were synthesized along with relevant reference compounds that contained in one case the ligand in a monovalent format and in another case the scaffold but not the ligand. Using a surface plasmon resonance inhibition assay the prepared inhibitors showed good inhibition. While the monovalent GM1 mimic showed the expected inhibition in the 200 μM range the multivalent scaffolds led to increased binding. The tetravalent compound was shown to be 440-fold more potent than its monovalent counterpart. The octavalent analog, however, was the most potent compound as determined using an ELISA assay.

Full text of DOI:10.1039/b405344c

View Article Online / Journal Homepage / Table of Contents for this issue
Synthesis and cholera toxin binding properties of multivalent
GM1 mimics†
Daniela Arosio,a,b Ioannis Vrasidas,a,b Paola Valentini,b Rob M. J. Liskamp,b
Roland J. Pieters*b and Anna Bernardi*a
a Dipartmento di Chimica Organica e Industriale, via Venezian 21, 20133 Milano, Italy;
Fax: ϩ39-02-50314092; Tel: anna.bernardi@unimi.it
b Department of Medicinal Chemistry, Utrecht Institute for Pharmaceutical Sciences,
Utrecht University, PO Box 80082, NL-3508 TB Utrecht, The Netherlands.
E-mail: r.j.pieters@pharm.uu.nl; Fax: ϩ31-30-253 6655
Received 13th April 2004, Accepted 26th May 2004
First published as an Advance Article on the web 30th June 2004
Dendrimers based on the 3,5-di-(2-aminoethoxy)-benzoic acid branching unit were used to attach multiple copies of
a GM1 mimic for inhibition of cholera toxin binding. Systems up to octavalent were synthesized along with relevant
reference compounds that contained in one case the ligand in a monovalent format and in another case the scaffold
but not the ligand. Using a surface plasmon resonance inhibition assay the prepared inhibitors showed good
inhibition. While the monovalent GM1 mimic showed the expected inhibition in the 200 µM range the multivalent
scaffolds led to increased binding. The tetravalent compound was shown to be 440-fold more potent than its
monovalent counterpart. The octavalent analog, however, was the most potent compound as determined using
an ELISA assay.
NeuAc residue. The interaction of 1a with CT has been studied
Introduction
in detail using NMR spectroscopy and molecular modeling.4
Toxins of the AB5 type are a major source of disease and
Unlike GM1, 1b can be synthesized on a gram scale, and there-
fore has a higher potential of being used for prevention of cell
intoxication. However, its low affinity is a serious drawback,
death.1 As the causative agent of cholera, the cholera toxin
(CT) is an important member of this family. Its mechanism of
action begins with the binding of the B subunit to ganglioside
which could be overcome by exploiting the power of multi-
GM1 molecules on the cell surface of the intestinal tract. This
valent presentation.
interaction is crucial for cell-uptake of the toxin and develop-
ment of the disease. Interference with this process is therefore
an attractive strategy with therapeutic potential. The protein–
carbohydrate interactions are multivalent, since the five B sub-
unit binding sites present in the multi-protein complex are well
positioned for simultaneous interaction with the surface-bound
Multivalency is a common phenomenon of interacting
biomolecules and also an emerging strategy for affinity
optimization.5 While the origin of the affinity enhancements by
synthetic multivalent ligands is not always clear, mechanisms
involving chelation versus aggregation can be distinguished. In
chelation the goal is to bridge binding sites by use of a suitable
GM1 molecules. The oligosaccharide part of ganglioside GM1
spacer, thus reducing the entropic barriers of the association of
binds strongly to the B subunits with a Kd of 43 nM at 25 ЊC,2
the second and higher (sub)ligands after binding of the first.6
and the multivalent display of the GM1 molecules makes the
Aggregation is another mechanism of soluble systems where a
affinity of the toxin to the cell surface even higher. Effective
spacer cannot bridge binding sites, yet multivalency effects are
interference is therefore a tremendous challenge. We describe
observed nonetheless.7 In this case aggregates can form whose
here our strategy of interference that is based on the combined
effects of structure optimization at the monovalent and at the
ligands dissociate with reduced off-rates, thus enhancing overall
affinity.8 The AB5 toxins represent an ideal case for harnessing
multivalent level. Simplified mimics of the complicated GM1
multivalency effects based on chelation of binding sites, due to
oligosaccharide were used and multivalency enhancement was
approached by the use of dendritic multivalent scaffolds. The
their positioning on the same face of the multimeric protein
complex. Indeed, the multivalency aspects of AB5 toxin binding
compounds were evaluated by an SPR competition assay and
have been actively studied over the last years. Highly active
an ELISA type assay.
The carbohydrate mimics were developed by a process of
compounds based on a pentavalent scaffold have been
developed for both the Shiga-like toxin9 and the heat-labile
structure-based design and experimental verification. This has
enterotoxin of E. coli (LT)10 and also theoretical models have
led to a group of pseudo-saccharides designed to retain the
specific orientation of the GM1 binding determinants (the
terminal galactose residue and the sialic acid carboxy group)
while progressively simplifying the molecular structure (and
been developed to help future design efforts.11 Our approach
to multivalency has involved the use of glycodendrimers as
scaffolds for multivalent display of carbohydrate ligands, which
in specific cases gave strong multivalency effects of over three
hence the synthetic complexity) of the ligand. These com-
orders of magnitude.12 Dendrimers or dendritic scaffolds
pounds bind to CT with dissociation constants that vary from
are easily synthesized and also easily modified with respect to
valency and geometry. Therefore they are useful tools in the
1 mM up to the potency of GM1.3 Among them, for the present
project we selected 1b, based on the previously made 1a, a 190
µM ligand of CT. Compound 1a retains the galactose epitope
of GM1 and uses an (R)-lactic acid as a substitute for the
study of multivalency aspects of protein–carbohydrate inter-
actions.13 In contrast to the previously mentioned pentameric
inhibitors, a dendrimer-based system does not have the exact
complementarity to the target in terms of the number of
attached ligands, either four or eight ligands for five binding
sites. However, a recent demonstration of an octavalent glyco-
dendrimer ligand that outperformed a pentavalent core-derived
† Electronic supplementary information (ESI) available: characteriz-
ation of the polyvalent compounds – imide by-products. See http://
www.rsc.org/suppdata/ob/b4/b405344c/
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 2 1 1 3 – 2 1 2 4
2113

View Article Online
Scheme 1 Synthesis of the protected monovalent ligand 6.
Scheme 2 Synthesis of the Galβ1,3GalNAc donor 5.
system in the inhibition of the Shiga-like toxin,11b further
supports our approach. The dendrimeric cores used in this
project are based on the 3,5-di-(2-aminoethoxy)-benzoic acid
branching unit.12,14 They have already been used to synthesize
polyvalent CT ligands using lactose as the monovalent
inhibitor.12a The present study incorporates two improvements
on that system. Firstly, here a designed GM1 mimic is used,
a monovalent ligand of much higher affinity than lactose.
Secondly, the scaffold was outfitted with elongated arms, as
long spacer arms were shown to be beneficial in other systems.10
now with two orthogonally protected carboxy groups, was
transformed into the mono ether 4 using a published pro-
cedure.4 Finally 4 was glycosylated in 55% yield with the
Galβ1,3GalNAc donor 5, whose synthesis, improved over
reported procedures,4 is depicted in Scheme 2. Quantitative
removal of the t-butyl ester yielded the protected monovalent
ligand 6b ready for conjugation.
The synthesis of the polyvalent ligands started from the
dendritic cores 12, 17, 21 and 25,12a and followed a common
protocol. Thus the monovalent core 12 (Scheme 3) was
elongated (BOP, DIPEA) to give 13, using the amino acid
spacer 29.12d,17 Deprotection of 13, followed by condensation
(HBTU) with an excess of 6b yielded the fully protected mono-
valent reference compound 14 in 73% overall yield. Its benzyl
ester was removed by standard hydrogenolysis on Pd–C.
This reaction went to completion in 1 hour yielding the free
carboxylic acid 15 (not shown), which was purified by short-
path flash column chromatography (60 : 35 : 5 CHCl3/MeOH/
H2O). Zemplén’s hydrolysis of the acetates (MeONa in MeOH)
proceeded within a few minutes yielding 16. However, a by-
product of mass M-32 was also observed in the ESI-MS
spectrum of the final product. The by-product was isolated by
careful chromatography using 60 : 35 : 5 CHCl3/MeOH/H2O
and was characterized as imide 30 (Chart 2). This assignment
Synthesis
The synthesis of 1a (Chart 1) was previously reported.4 For the
synthesis of 1b, the reported sequence was slightly modified
to achieve differentiation of the carboxy group in the cyclo-
hexanediol moiety, and to improve the overall yield, also
reducing the number of steps (Scheme 1). In brief, the known
diol 215 was treated with NaOH, and the resulting mono acid15
was transformed into the t-butyl ester using N,N-diisopropyl-
O-tert-butyl isourea16 in DMA/CH2Cl2. The resulting diol 3,
1
was confirmed by the H-NMR data that showed significant
shifts for the protons adjacent to the carboxy groups in the
cyclohexanediol fragment. Furthermore, an analog of 30 with
OH groups instead of the disaccharide and lactic acid moieties
also showed formation of the imide under the same reaction
conditions (not shown).
Chart 1 The monovalent ligand 1.
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 2 1 1 3 – 2 1 2 4
2114

View Article Online
Scheme 3 Synthesis of the multivalent CTB ligands. aCompounds 16, 20, 24 and 28 also include cyclic imide containing spacer arms as in 30, see
text and electronic supplementary information.†
Starting from the appropriate dendritic cores 17, 21, and 25
(Scheme 3), the same process afforded the divalent 19, the
tetravalent 23 and the octavalent 27, in 65%, 40% and 14%
yields, respectively. Characterisation of these compounds
was performed by NMR spectroscopy and ESI-MS analysis.
Despite the high molecular weight, at this stage all compounds
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 2 1 1 3 – 2 1 2 4
2115

View Article Online
Chart 2 Obtained side product 30.
Scheme 4 Synthesis of the water soluble divalent reference compound 32.
were well-behaved in terms of solubility in common organic
solvents, they could be chromatographed on silica gel with
CHCl3/MeOH and gave reasonably well-defined signals in
1H-NMR. ESI-MS revealed the presence of multiple-charge
peaks of the derivatives while underglycosylated by-products
were not detected. Deprotection was performed following the
procedure discussed for the monovalent compound 14 to yield
the polyvalent 20, 24 and 28. Depending on the substrate and
on the reaction conditions, variable amounts (20–35%) of
inseparable imide by-products were obtained upon deacetyl-
ation, as judged by 1H-NMR integration of the reaction crudes.
The final polyvalent pseudo-glycoconjugates were purified
by precipitation with acetone from water solution. All the
products 16, 20, 24 and 28 were obtained as white solids, water
soluble up to 7 mg mlϪ1. They were all fully characterized by
1H-NMR and 13C-NMR and by mass spectrometry. ESI-MS
revealed multiple charge peaks of the compounds and also of
the corresponding [M-32] by-product. The [M-64] by-product
was detected only in the case of the tetravalent ligand 24, while
the [M-96] by-product was never detected. Nevertheless we
cannot exclude the presence of multiple imides in the higher
generation polyvalent ligands because the presence of small
quantities of these by-products was hard to identify in the
complex mass spectra characteristic of these polyvalent com-
pounds. The compounds were also analyzed by MALDI-MS,
which allowed the detection of singly charged peaks for the
monovalent and divalent compounds 16 and 20.
With the aim of addressing the issue of non-specific inter-
actions that could occur between the scaffolds and the cholera
toxin, we also synthesized the divalent reference compound 32
(Scheme 4) by coupling 18 to a polyhydroxylated system that
does not interact with CT. This compound, which represents a
water soluble version of the divalent aglycon, was obtained
by deprotecting the elongated divalent core 18 with TFA and
treating the resulting diamine with excess -(ϩ)-ribonic acid-γ-
lactone 31 in MeOH. The diamide 32 was not easily separated
from unreacted ribonic lactone, but the reaction crude was
acetylated and the functionalized dendrimer was isolated by
flash chromatography. Finally, classical deacetylation reaction
yielded pure 32. All the compounds synthesized were tested as
CT binders using SPR and an ELISA-type assay.
Results of the SPR analysis
The prepared compounds were analyzed with a surface plasmon
resonance (SPR) competition assay, using the cholera toxin B
pentamer CTB as the target. The glycoprotein ASF was
covalently attached to a dextran functionalized gold SPR chip
via NHS active ester intermediacy. ASF is a protein obtained
after desialylation of the major plasma protein fetuin. It
contains three N-glycosylation sites with complex type
triantennary chains either with N-acetyllactosamine terminus
(74%) or with the Galβ1,3GlcNAc isomer in the outer
Manα1,3-arm (9%) or a biantennary chain (17%). Furthermore
three O-linked chains, primarily displaying Galβ1,3GalNAcα–
are also present.18 Despite the fact that the glycoprotein does
not display the GM1-oligosaccharide as the optimal ligand,
CTB exhibited good affinity for the chip. A series of measure-
ments with increasing CTB concentrations yielded a binding
19
isotherm and a Kchip of 3 µM. Inhibition studies were then
performed which showed that the prepared compounds were
able to inhibit the CTB (at 3 µM) binding to the chip. First,
lactose was measured and an IC50 of 9.4 mM (Table 1) was
determined for this disaccharide. The monovalent GM1 mimic
1b showed much better inhibition with an IC50 of 97 µM. Both
values are in the range that is close to their Kd as measured by
other methods such as fluorescence titration.12a,3e As a mono-
valent reference, compound 16 was measured (Fig. 1). This
compound exhibited an IC50 of 221 µM. The slightly reduced
inhibitory power in comparison with 1b clearly indicates that
the attached spacer does not benefit the binding, rather it
reduces it roughly two-fold. It is also possible that the reduced
affinity is an expression of the presence of the imide side
product 30, whose affinity was not individually assessed. The
divalent 20 showed an IC50 of 13 µM which indicates almost
a 17-fold enhanced binding in comparison to the monovalent
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 2 1 1 3 – 2 1 2 4
2116

View Article Online
Table 1 Surface plasmon resonance (SPR) inhibition studies using
CTB (3 µM), an ASF-coated SPR sensorchip, and varying concen-
trations of inhibitors
their relative potency is estimated by ELISA on GM1-coated
plates. On the other hand, both methods imply that the affinity
of the multivalent ligands increases with the valency of the
compound.
Compound
Valency
IC50 [µM]
Relative potency (per sugar)
Lactose
1
1
1
2
4
8
1
9400
97
221
13
0.02 (0.02)
2.3 (2.3)
1
17 (8.5)
442 (111)
442 (55)
22 (22)
Discussion
1b
Multivalent GM-1 mimics were prepared based on their
activity as monovalent entities. In the synthesis a cyclic imide
product was formed, which could theoretically complicate our
binding studies. However, the formation of the imide is not
expected to create a major distortion of the DCCHD ring. The
first non-chair conformation of the parent ring system (i.e. the
imide with two methyl ethers on the diol) is still 12 kJ molϪ1
above the global minimum by MM3*, compared with 17 kJ
molϪ1 for the “diester” compound. Hence, the imide formation
should have little effect on the relative orientation of the bind-
ing determinants (galactose and carboxylate) which define the
affinity of the ligand for CT.2 Experimentally we observed
similar behavior of 16 and the shown 16/30 mixture (see Elec-
tronic Supplementary Information†) in a fluorescence assay.21,25
The fact that a multivalency effect was observed in the SPR
assay when moving from mono- to tetravalent compounds is
consistent with the geometry of the compounds, i.e. their
expected ability to bridge adjacent binding sites. In the crystal
structure of the complex of five GM1 molecules and CTB22 the
distance between neighboring ligands was 31 Å, as measured
between the anomeric oxygens of the terminal galactoses. The
two GM1 mimics of the divalent 20 are separated by about 70
Å in a mostly extended conformation. The fact that the average
distance or the so-called effective length between ligands separ-
ated by flexible spacers is significantly shorter than its extended
conformation was previously shown.10,23 The observed 17-fold
enhancement (nine-fold per sugar mimic) indicates that
chelation is indeed taking place. Such an enhancement is
sizeable considering the large distance of 31 Å between the
binding sites and the flexibility of the spacer. For the Shiga-like
toxin a series of divalent ligands was prepared to bridge sites
separated by only 11 Å, and enhancements per sugar between
3.8 and 23.7 were measured.23 The effect observed for tetra-
valent 24 was larger with an enhancement of 442-fold (111-fold
per sugar mimic). Due to the higher valency more binding sites
can simultaneously be reached. In the recently reported
thermodynamic model of Kitov and Bundle11b avidity entropy,
a factor expressing the probability of association, was coined as
a determining factor in chelation-based multivalency effects.
For a tetravalent ligand binding to a pentavalent receptor this
probability should be significantly larger than for the divalent
system, but the octavalent 28 should yield the highest prob-
ability. Therefore it seemed surprising that 28 did not show a
lower IC50 in the SPR assay. However, it can be understood
when the stoichiometry of the components in the assay is
considered. A CTB concentration of 3 µM was used, while the
IC50 of the tetravalent 24 was 0.5 µM. It seems as if each
individual GM1 mimic is blocking more than one CT binding
site. Besides unlikely experimental error this suggests that
aggregation may also be taking place in which more than one
toxin binds to the arms of a single glycodendrimer. In such an
aggregate24 some of the free binding sites may be sterically
blocked for binding to the chip surface. Although the existence
of aggregates in this study was not experimentally confirmed
it is clear that the tetravalent compound 24 reached the bound-
aries of the assay. For the octavalent 28, which based on the
ELISA results appears to be a better binder than 24, this was
also the case. More support for this interpretation came from
the measurement of the GM1–OS, a ligand with a Kd of 43 nM,2
which in our SPR assay inhibited the CTB binding, only with
an IC50 of 10 µM, a number close to the number of available
carbohydrate binding sites under the assay conditions. This is
again an indication that the assay can only differentiate
16
20
24
0.5
28
0.5a
10a
GM1-OS
a Value an underestimation of the affinity since the assay limit was
reached, see text for details.
Fig. 1 Inhibition curves of CTB binding to an ASF functionalized
SPR sensorchip. Normalised inhibition curves are shown for the
monovalent ligand 16 (circles, solid line), the divalent ligand 20
(diamonds, dotted line), the tetravalent ligand 24 (squares, dashed line).
reference compound. The tetravalent 24 showed a large
multivalency effect with an IC50 of 0.5 µM or a 442-fold
enhanced binding (111-fold per sugar mimic). Moving to the
octavalent compound 28 did not lead to a further enhancement
in this assay, and its potency on a per sugar basis is actually
lower than that of tetravalent 24. The GM1-oligosaccharide
(GM1-OS) showed an IC50 of 10 µM in this assay. The last two
experiments indicated that these high-affinity compounds
reached the limit of the assay (see discussion), due to the
relatively moderate inherent affinity of CTB for the ASF chip.
The divalent reference scaffold compound 32 was measured up
to 500 µM and showed no effect.
ELISA
In order to better evaluate the affinity of the octavalent 28 a
different assay was employed. For this reason an ELISA-like
assay was used. In this assay the ganglioside GM1 was coated
onto the ELISA wells (0.1 µg of GM1 per well) and a CTB–
horseradish peroxidase conjugate (CT-HRP) was used as the
binding protein.20 The binding could clearly be visualized by the
HRP catalyzed conversion of o-phenylene diamine (OPD) and
measured by a multiwell plate spectrophotomer. GM1-OS was
capable of inhibition, although concentrations of over 100 µM
were required for full inhibition, reflecting the multivalent GM1
display on the ELISA plate. With the multivalent systems
inhibition was observed only with the octavalent 28, but not
with its analogs of lower valency. With the concentrations
employed (up to 400 µM) only 20% inhibition was observed
with 28. Thus, contrary to the SPR results on ASF chips, GM1-
OS appears to be a stronger binder of CTB than 28 when
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 2 1 1 3 – 2 1 2 4
2117

View Article Online
affinities up to the low micromolar range. The ELISA assay
on GM1-coated plates clearly showed GM1–OS to be the best
ligand, and consistent with expectations it also showed the
octavalent 28 to be the best of the multivalent ligands.
Unfortunately, the ligands could not all be compared in the
same assay.25
Flash chromatography was performed using Macherey-
Nagel Kieselgel 60, 230–400 mesh. Optical rotation values were
measured using a Perkin-Elmer 241 polarimeter, at 589 nm, in
1 ml cells.
Electrospray Ionisation (ESI) mass spectrometry was carried
out with a Shimadzu LCMS QP-8000 single quadrupole
benchtop mass spectrometer (m/z < 2000), or a ThermoQuest
Finnigan LCQDeca mass spectrometer (FINNIGAN MAT,
San Jose, CA, USA) (m/z < 2000), or an Apex II ICR FTMS
(for HRMS).
Conclusions
Our approach to high-affinity ligands for cholera toxin repre-
sents a combination of structure-based design of monovalent
ligands with further enhancement by multivalent presentation
using dendrimers. Our experiments showed that multivalent
compounds were accessible which included an advanced GM1
mimic attached to dendritic scaffolds with elongated spacer
arms. To some extent the synthetic strategy led to additional
cyclic imide formation in the linking unit, with possible effects
on the conformation of the GM1 mimic. The SPR inhibition
assay showed that the tetravalent 24 in comparison to lactose
was almost 19000-fold more active than lactose. The multi-
valency enhancement of 442 relative to the monovalent 16, was
of the same order of magnitude (263-fold) as the enhancement
observed by Hol et al. with their pentavalent version of the
m-nitrophenyl-α--galactoside ligand.26 In contrast to their
hypothesis, our results clearly show that a pentavalent ligand
design is not a crucial factor for strong multivalency binding
enhancement to AB5 toxins. Further support for this notion
also comes from the success of a functionalized PAMAM
dendrimer for interference with binding of the Shiga-like
toxin.11b Precisely how the most effective compound, octavalent
28, compared to the tetravalent 24 could not be quantified with
our assay methods, however its affinity was clearly higher. In
general, our results showed that multivalent presentation of
designed ligands can indeed bring their affinity closer to that
required for practical application against AB5 toxins. In order
to do so the compounds need to be considerably larger than
monovalent ligands to be effective. Considering the toxins
reside in the intestinal tract this is not a problem, in fact it is
an added bonus if the structure turns out to be too large and
too polar for absorption. We are currently still in the phase
where the optimal multivalent geometry for toxin neutraliz-
ation, be it cyclic, radial, dendritic or polymeric, is not yet
known, although several promising ones are now available. For
the longer term, concise syntheses of both simple but effective
ligands and similarly straightforward multivalent scaffolds may
bring practical intervention within reach.
MALDI mass spectrometry was performed with a Bruker
OMNIFLEX spectrometer.
FABϩ mass spectrometry was performed with a VG 7070
EQ-HF spectrometer.
1H-NMR and 13C-NMR were recorded using Bruker Ac-200,
Ac-300, and Avance-400. TMS was used as internal standard.
For 13C-NMR spectra, only selected values are reported.
Product numbering for spectral assignment
The unconventional numbering of the DCCHD moiety in the
GM1 mimic was used in analogy with the Gal-II residue of
the GM1 ganglioside. To avoid misunderstandings, the same
numbering was adopted for non glycosylated dicarboxy-
cyclohexanediols. Protons belonging to the (R)-lactic acid chain
were identified as CH3-L and H-L.
Dendrimers are numbered as reported in the picture. The
methylenic protons of the linker are identified by the letter
assigned to the corresponding carbon atom. The amide protons
are numbered progressively starting from the scaffold.
Synthesis of 3
To dimethyl ester 2 (2 g, 8.6 mmol, 1 mol eq.) a 0.07 M solution
of NaOH (182 ml, 12.75 mmol, 1.5 mol eq.) was added. The
reaction mixture was stirred at rt for 1.5 h monitoring by TLC.
After reaction completion HCl 6 M was added to pH 1, and the
solvent was evaporated under reduced pressure. The crude was
purified by flash chromatography on silica gel (8 : 2 CHCl3/
MeOH ϩ 1% AcOH) to yield the monoacid16 (80%).
To a suspension of the monoacid (602 mg, 2.76 mmol, 1 mol
eq.) in dry CH2Cl2 (12 ml) and under N2, DMA (2 ml) was
added dropwise, then t-butyl isourea16 (1.32 ml, 5.52 mmol, 2
mol eq.) was added. The reaction mixture was stirred at 40 ЊC
for 6 h adding 2 equivalents of isourea every two hours (8 eq. in
total) and monitoring by TLC. Two hours after the last
addition, the solvent was evaporated under reduced pressure
and the crude was taken up with Et2O. The suspension was
filtered over a pad of Celite until the urea was completely
removed. The organic phase was evaporated and the crude was
purified by flash chromatography on silica gel (CHCl3/MeOH
95 : 5) to yield 3 as a white solid (66%). [α]20D: ϩ19.8 (c = 1.11,
CHCl3). MS (FABϩ): 275 [M ϩ Hϩ]; 297 [M ϩ Naϩ]. 1H-NMR
(CDCl3, 200 MHz): 1.45 (s, 9H, OC(CH3)3); 1.6 (ddd, 1H,
H-5ax, Jgem ≅ J6ax-1 = 2 Hz; J6ax-4 = 2.5 Hz); 1.8 (ddd, 1H, H-2ax,
Jgem ≅ J3ax-2 ≅ J3ax-4 = 12 Hz); 2.05 (ddd, 1H, H-2eq, Jgem = 12 Hz;
J3eq-2 ≅ J3ax-4 = 12 Hz); 2.21 (ddd, 1H, H-5eq, Jgem = 12 Hz; J6eq-1
Experimental section
General
Chemicals were obtained from commercial sources and used
without further purification, unless stated otherwise. Pyridine,
DMF, dimethylacetamide (DMA), DME, and allyl alcohol
were dried over 4 Å molecular sieves. All other dry solvents
were distilled under nitrogen shortly before use. THF, hexane,
dioxane and benzene were distilled from Na; CH2Cl2, MeOH,
Et3N, iPr2NEt, were distilled from CaH2.
≅ J6eq-4 = 3.5 Hz); 2.45 (s, 2H,OH); 2.62 (ddd, 1H, H-1, J2-1
J2-3ax = 12 Hz; J2-3eq = 3.5 Hz); 2.98 (ddd, 1H, H-6, J1-2 ≅ J1-6ax
≅
=
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 2 1 1 3 – 2 1 2 4
2118

View Article Online
12 Hz; J1-6eq = 3.5 Hz); 3.68 (s, 3H, OCH3), 3.7–3.75 (m, 1H,
H-3); 3.85–4.05 (br s, 1H, H-4). 13C-NMR (CDCl3, 50.3 MHz):
27.8; 30.1; 33.0; 38.3; 44.1; 51.7; 67.7; 70.3; 80.8; 173; 175.5.
MHz): 1.2 (s, 9H, COC(CH3)3); 1.25 (s, 9H, COC(CH3)3); 1.92
(s, 3H, NHCOCH3); 2.95 (br s, 1H, OH); 3.45–3.61 (m, 1H,
H-4); 3.8–3.89 (m, 1H, H-5); 3.98–4.25 (m, 2H, OCH CH᎐
᎐
2
CH2); 4.35–4.42 (m, 2H, H-2, H-6); 4.4–4.5 (m, 1H, H-6Ј); 4.85
(d, 1H, H-1, J1–2 = 3.5 Hz); 5.04–5.15 (m, 1H, H-3); 5.21–5.37
Synthesis of the triflate of (S)-lactic acid benzyl ester
(m, 2H, OCH CH᎐CH ); 5.75 (d, 1H, NHCOCH , J = 12 Hz);
᎐
2
2
3
To a solution of (S)-lactic acid benzyl ester (260 mg, 1.443
mmol, 1 mol eq.) in dry CH2Cl2 (5 ml), under N2 and at 0 ЊC,
trifluoromethanesulfonic anhydride (1.6 ml, 5.94 mmol, 1.1 mol
eq.) was added. After 5 min 2,6-lutidine (195 µl, 1.682 mmol,
1.16 mol eq.) was added. The reaction mixture was stirred at
room temperature for 20 min monitoring by TLC. At reaction
completion the solvent was evaporated and the crude purified
by short path flash chromatography (hexane/EtOAc 9 : 1. 90%
yield). 1H-NMR (CDCl3, 200 MHz): 1.7 (d, 3H, CH3, J = 7 Hz);
5.25 (m, 3H, CH2, CH); 7.38 (s, 5H, aromatic protons).
5.82–5.95 (m, 1H, OCH CH᎐CH ).
᎐
2
2
Synthesis of 1-allyl-2-N-acetyl-2-desoxy-4,6-pivaloylgalactose 9
To a solution of 8 (22 g, 51.2 mmol, 1 mol eq.) in dry CH2Cl2/
pyridine 20 : 1 (489 ml), under N2 and at Ϫ35 ЊC, Tf2O (10.53
ml, 62.6 mmol, 1.2 mol eq.) was added. The reaction mixture
was stirred at 0 ЊC for 2 h, monitoring by TLC (hexane/EtOAc
4 : 6), then H2O (22 ml, 122.2 mmol, 2.4 mol eq.) was added and
the reaction was refluxed for 13 h. After completion, CH2Cl2
was added and the organic phase was washed with HCl 5%,
NaHCO3 5% and H2O. The organic phase was dried with
Na2SO4 and the solvent evaporated. The crude was purified
by flash chromatography on silica gel (hexane/EtOAc 2 : 8)
Synthesis of the monoether 4
A solution of diol 3 (146 mg, 0.53 mmol, 1 mol eq.) and
Bu2SnO (132 mg, 0.53 mmol, 1eq.), in dry benzene (4 ml) was
refluxed under N2, while continuously removing water (4 Å
molecular sieves were inserted between the flask and the reflux
condenser). After 8 h the solvent was evaporated under reduced
pressure. The residue was taken up in dry DME (1.5 ml) and,
under N2, the triflate of (S)-benzyllactate (248 mg, 0.79 mmol,
1.5 mol eq.) and CsF (120 mg, 0.79 mmol, 1.5 mol eq.) were
added. The reaction mixture was stirred at room temperature
for 30 min, monitoring by TLC. After completion, Et2O was
added and the organic phase was washed with H2O. The
organic phase, dried with Na2SO4, was evaporated and the
crude purified by flash chromatography on silica gel (hexane/
1
yielding 9 (45%). H-NMR (CDCl3, 400 MHz): 1.2 (s, 9H,
COC(CH3)3); 1.28 (s, 9H, COC(CH3)3); 2.1 (s, 3H, NHCOCH3);
3.95–4.1 (m, 2H, H-3, Ha); 4.1–4.2 (m, 3H, H-5, H-6, H-6Ј);
4.25 (m, 1H, Hb); 4.95 (d, 1H, H-1, J = 6 Hz); 5.26–5.32 (m, 3H,
H-4, OCH CH᎐CH ); 5.85–5.95 (m, 2H, OCH CH᎐CH ,
᎐
᎐
2
2
2
2
NHCOCH3).
Synthesis of O-allyl-Gal(ꢀ1–3)GalNAc 11
To a solution of 9 (11.9 g, 27.7 mmol, 1 mol eq.) in dry hexane/
CH2Cl2 1 : 1 (234 ml), under N2 and at 0 ЊC, BF3ؒEt2O (4.17 ml,
33.24 mmol, 1.2 mol eq.) was added. The solution was warmed
to room temperature and a solution of tetra-acetyl-galactose
trichloroacetimidate (13.6 g, 27.7 mmol, 1 mol eq.) in dry
CH2Cl2 (117 ml) was added. The reaction mixture was stirred at
room temperature for 2 h then a second equivalent of trichloro-
acetimidate dissolved in the same amount of CH2Cl2 was
added. The reaction was stirred for 12 h monitoring by TLC
(hexane/EtOAc 2 : 8). After completion, TEA was added to pH
7 and the solvent was evaporated under reduced pressure.
The crude was purified by flash chromatography on silica gel
(toluene/acetone 8 : 2) to afford 11 (60%). 1H-NMR (C6D6, 200
MHz): 1.2 (s, 9H, OCOC(CH3)3); 1.25 (s, 9H, OCOC(CH3)3);
1.98 (s, 3H, OCOCH3); 2.0 (s, 3H, OCOCH3); 2.07 (s, 3H,
OCOCH3); 2.1 (s, 3H, OCOCH3); 2.19 (s, 3H, NHCOCH3);
EtOAc 65 : 35) to afford 4 in 75% yield, as a white solid. [α]20
:
D
ϩ22 (c = 1.55, CHCl3). 1H-NMR (CDCl3, 300 MHz): 1.55–1.52
(m, 11H, COOC(CH3)3, CH3-L, H-5ax); 1.76 (ddd, 1H, H-2ax,
Jgem ≅ J3ax-2 ≅ J3ax-4 = 12 Hz); 2.05 (ddd, 1H, H-2eq, Jgem = 12 Hz,
J3eq-2 ≅ J3ax-4 = 12 Hz); 2.25 (ddd, 1H, H-5eq, Jgem = 2 Hz; J6eq-1
J6eq-4 = 3.5 Hz); 2.41 (ddd, 1H, H-1, J2–1 ≅ J2–3ax = 12 Hz, J2–3eq
≅
=
3.5 Hz); 2.85 (ddd, 1H, H-6, J1–2 ≅ J1–6ax = 12 Hz, J1–6eq = 3.5
Hz); 3.37 (m, 1H, H-3); 3.65 (s, 3H, OCH3), 4.05 (br s, 1H, H-4);
5.19 (m, 2H, OCH2Ph); 7.35 (s, 5H, aromatic protons). 13C-
NMR (CDCl3, 50.3 MHz): 18.7; 27.8; 27.9; 32.5; 38.0; 44.2;
51.5; 65.0; 66.8; 72.3; 76.4; 77.0; 77.6; 80.3; 128.1; 128.5; 172.6;
174.1.
3.8–4.3 (m, 9H, OCH CH᎐CH , GalNAcH-6, GalNAcH-6Ј,
GalH-6, GalH-6Ј, GalNAcH-1, GalNAcH-5, GalH-5); 4.53
᎐
Synthesis of 1-allyl-2-N-acetyl-2-deoxy-3,6-pivaloylglucose 8
2
2
Acetyl chloride (27.6 ml, 388 mmol, 3.4 mol eq.) was added
dropwise to allylic alcohol (208 ml), under N2 and at 0 ЊC. At
room temperature, N-acetyl glucosamine 7 (25 g, 113 mmol,
1 mol eq.) was added. The reaction mixture was stirred at 70 ЊC
for 3 h, monitoring by TLC (CHCl3/MeOH 8 : 2). After
reaction completion, solid NaHCO3 was added to pH 7 and the
suspension was filtered through a Celite pad, washing several
times with MeOH. The solvent was removed under reduced
pressure and the residue was digested with Et2O. Finally the
solvent was decanted and the α-1-allyl-2-N-acetyl glucosamine
(m, 1H, GalNAcH-2); 4.6 (m, 1H, GalH-1); 4.9–5 (m, 2H,
GalNacH-3); 5.1 (m, 1H, GalH-2); 5.2–5.4 (m, 4H, OCH CH᎐
᎐
2
CH2, GalNAcH-4, GalH-4), 5.65 (d, 1H, NHCOCH3); 5.75–6.1
(m, 1H, OCH CH᎐CH ).
᎐
2
2
Synthesis of the trichloroacetimidate 5
To a solution of the disaccharide 11 (5.7 mg, 7.5 mmol, 1 mol
eq.) in dry MeOH (500 ml), under N2, 1 M MeONa in MeOH
(13 ml, 13 mmol) was added. The reaction mixture was stirred
at room temperature for 40 h, monitoring by TLC. After
reaction completion, Amberlite IR-120 (Hϩ form) was added to
pH 6. The organic phase was filtered and evaporated under
reduced pressure. The crude was redissolved in dry pyridine
(80 ml) and, at 0 ЊC, Ac2O (11 ml, 114 mmol) was added
dropwise. The reaction mixture was stirred at room temperature
for 24 h, monitoring by TLC (hexane/EtOAc 2 : 8). After com-
pletion, the solvent was evaporated under reduced pressure and
the residue was taken up with EtOAc. The organic phase was
washed with HCl 5%, NaHCO3 5% and H2O, then dried with
Na2SO4 and the solvent was evaporated. The crude was purified
by flash chromatography on silica gel (hexane/EtOAc 2 : 8)
yielding the polyacetylated disaccharide (70% over the two
steps). [α]20D = ϩ66.7 (c = 1.39, CHCl3). 1H-NMR (CDCl3, 400
1
was recovered as a white solid. (99%). Mp 145 ЊC H-NMR
(D2O, 400 MHz): 2.0 (s, 3H, NHCOCH3); 3.45 (m, 1H, H-4);
3.65–3.79 (m, 3H, H-3, H-5, H-6); 3.8–3.9 (m, 2H, H-2, H-6Ј);
3.95–4.23 (m, 2H, OCH CH᎐CH ); 4.89 (d, 1H, H-1, J1–2 = 3.6
᎐
2
2
Hz); 5.2–5.32 (m, 2H, OCH CH᎐CH ); 5.85–5.99 (m, 1H,
᎐
2
2
OCH CH᎐CH ). To
a suspension of α-1-allyl-2-N-acetyl
᎐
2
2
glucosamine (16 g, 61.2 mmol, 1 mol eq.) in dry CH2Cl2/
pyridine 1 : 2 (242 ml), under N2 and at 0 ЊC, tBuCOCl (20.7 ml,
171.0 mmol, 2.8 mol eq.) was added dropwise. The reaction
mixture was stirred for 2 h, monitoring by TLC (hexane/EtOAc
3 : 7). After completion, CH2Cl2 was added and the organic
phase was washed with HCl 5%, NaHCO3 5% and H2O. The
organic phase was dried with Na2SO41 and the solvent was
evaporated to yield 7 (98%), as an oil. H-NMR (CDCl3, 200
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 2 1 1 3 – 2 1 2 4
2119

View Article Online
MHz): 1.94 (s, 3H, OCOCH3); 1.96 (s, 3H, OCOCH3); 1.99 (s,
3H, OCOCH3); 2.09 (s, 3H, OCOCH3); 2.11 (s, 3H, OCOCH3);
2.16 (s, 3H, OCOCH3); 2.19 (s, 3H, NHCOCH3); 3.90 (m, 1H,
(c = 0.77, CHCl3). MS (FABϩ): 1055 [M ϩ H]ϩ; 1077 [M ϩ
Na]ϩ. HRMS (ESI؉): [C49H67NO24Na]ϩ requires 1076.39452,
found 1076.39722. Mp 75–78 ЊC. H-NMR (C6D6, 300 MHz):
1
GalH-5); 3.99–4.20 (m, 8H, OCH CH᎐CH , GalNAcH-6,
1.26 (m, 1H, CHDH-5ax); 1.43 (d, 3H, CH3L, JCH3-HL = 6.7 Hz);
1.49 (s, 9H, OCOC(CH3)3); 1.69 (s, 3H, OCOCH3); 1.82 (s, 3H,
OCOCH3); 1.83 (s, 3H, OCOCH3); 1.84 (s, 3H, OCOCH3); 1.89
(s, 3H, OCOCH3); 1.96 (s, 3H, OCOCH3); 2.05 (s, 3H,
NHCOCH3); 2.22 (m, 1H, CHDH-2ax); 2.30 (m, 1H, CHDH-
2eq); 2.45 (td, 1H, CHDH-5eq, J5eq-5ax = 10 Hz, J5eq-6 ≅ J5eq-4 ≅ 4
Hz); 2.78 (dt, 1H, CHDH-1, J1–6 ≅ J1–2ax ≅ 12 Hz, J1–2eq = 3.8
Hz); 3.02 (m, 1H, CHDH-3); 3.25 (m, 1H, CHDH-6); 3.3 (s,
3H, OCH3); 3.35 (m, 1H, GalH-5); 3.65 (t, 1H, GalNAcH-5,
J5–6 = 7 Hz); 3.8–3.9 (m, 2H, HL, GalNAcH-2); 3.92 (br s,
1H, CHDH-4); 4.12–4.30 (m, 4H, GalH-6, GalH-6Ј, Gal-
NAcH-6, GalNAcH-6Ј); 4.33 (d, 1H, GalH-1, J1–2 = 6 Hz); 4.54
(dd, 1H, GalNAcH-3, J3–2 = 11 Hz, J3–4 = 3.4 Hz); 5.0 (dd, 1H,
OCH2Ph, Jgem = 12 Hz); 5.16 (m, 1H, GalH-3); 5.45–5.60 (m,
3H, GalH-2, GalH-4, GalNAcH-4); 5.80 (d, 1H, GalNAcH-1,
J1–2 = 6.3 Hz); 7.25–7.31 (m, 5H, aromatic protons) 13C-NMR
(C6D6, 75 MHz): 18.5; 20.6; 23.5; 27.7; 29.2; 33.2; 38.6; 39.0;
44.5; 51.0; 54.2; 56.2; 67.0; 68.7; 70.6; 71.2; 72.4; 78.1; 100.6;
101.2; 127.8; 128.7; 135.8; 168.9; 170.0; 172.4; 173.2; 174.6;
174.8.
᎐
2
2
GalNAcH-6Ј, GalH-6, GalH-6Ј, GalNAcH-1, GalNAcH-5);
4.51 (m, 1H, GalNAcH-2); 4.62 (m, 1H, GalH-1); 4.90–5.01 (m,
2H, GalNacH-3, GalH-3, GalH-2); 5.11 (d, 1H, GalNH-1, J1–2
= 3.7 Hz); 5.28–5.40 (m, 4H, OCH CH᎐CH , GalNAcH-4,
᎐
2
2
GalH-4), 5.66 (d, 1H, NHOCH3, JNH-H2 = 8.83 Hz); 5.88 (m,
1H, OCH CH᎐CH ). 13C-NMR-HETCOR (CDCl3, 400 MHz):
᎐
2
2
20.8; 21.0; 21.1; 21.1; 23.7; 49.3; 61.5; 63.0; 67.1; 67.9; 68.9;
69.1; 69.2; 71.2; 71.3; 73.2; 97.3; 100.9; 118.7.
To a solution of the acetylated disaccharide (3.56 g, 5.30
mmol, 1 mol eq.) in AcOH/H2O 20 : 1 (11 ml), NaOAc (1.03 g,
12.6 mmol, 2.4 mol eq.) and PdCl2 (1.03 g, 5.83 mmol, 1.1 mol
eq.) were added. The reaction mixture was stirred at room
temperature for ca. 20 h monitoring by TLC. After completion
the mixture was filtered through a Celite pad washing several
times with EtOAc. The organic phase was washed with a satur-
ated solution of NaHCO3, dried with Na2SO4 and finally the
solvent was evaporated under reduced pressure. The crude was
purified by flash chromatography on silica gel (EtOAc/CHCl3
9 : 1 ϩ 2% MeOH) yielding the anomerically deprotected
disaccharide (70%). [α]20D = ϩ54.8 (c = 1.035, CHCl3). 1H-NMR
(CDCl3, 400 MHz): 2.00 (s, 3H, OCOCH3); 2.03 (s, 3H,
OCOCH3); 2.07 (s, 3H, OCOCH3); 2.09 (s, 3H, OCOCH3); 2.11
(s, 3H, OCOCH3); 2.15 (s, 3H, OCOCH3); 2.19 (s, 3H,
To a solution of 6a (310 mg, 0.294 mmol, 1 mol eq.) in
CH2Cl2 (3 ml), TFA (3 ml) was added. The reaction mixture
was stirred at room temperature for 1 h, monitoring by TLC
(CHCl3/MeOH 9 : 1). After completion, the solvent was
evaporated and the residue taken up with toluene, which was
evaporated several times to completely remove the TFA. Crude
6b (99%) was dried under vacuum and used for the following
reactions without any further purification.
NHCOCH3); 3.58 (br s, 1H, OH); 3.91 (dt, 1H, GalH-5, JH5-H6
=
7.2 Hz); 4.01–4.21 (m, 5H, GalNAcH-6, GalNAcH-6Ј, GalH-6,
GalH-6Ј, GalNAcH-3); 4.36 (t, 1H, GalNAcH-5, J5–6 = 6 Hz);
4.44 (m, 1H, GalNAcH-2); 4.74 (d, 1H, GalH-1, J1–2 = 8 Hz);
4.99 (dd, 1H, GalH-3, J3–2 = 10.4 Hz, J3–4 = 3.4 Hz); 5.17 (dd,
1H, GalH-2, J2–1 = 8 Hz); 5.27 (dd, 1H, GalH-4); 5.38–5.42
(m, 2H, GalNAcH-1, GalNAcH-4); 5.87 (d, 1H, NHCOCH3,
JNH-H2 = 8.1 Hz). 13C-NMR-HETCOR (CDCl3, 400 MHz):
20.6; 21.2; 23.8; 27.4; 27.7; 47.6; 50.2; 62.0; 63.3; 63.5; 67.6;
68.7; 71.4; 72.1; 92.3; 92.5; 101.2.
Synthesis of 1b
To a solution of pseudo trisaccharide 6a (22 mg, 0.021 mmol,
1 mol eq.) in MeOH (1 ml), a catalytic amount of Pd/C 10%
was added. The mixture was stirred under H2 atmosphere at
room temperature for 1 h. After reaction completion, the
mixture was filtered through a Celite pad, washing with MeOH.
The collected organic phase was evaporated under reduced
pressure and the crude (19 mg, 0.02 mmol, 1 mol eq.) redissol-
ved in dry MeOH (1 ml). Under N2, a solution of 1 M MeONa
in MeOH (0.03 ml, 0.03 mmol, 1.5 mol eq.) was added. The
reaction was stirred at room temperature for 45 min monitoring
by TLC (CHCl3/MeOH/H2O 60 : 35 : 5). After completion, the
reaction was quenched adding Amberlite IR 120 (Hϩ form) to
pH 5–6. Finally the resin was filtered and washed with MeOH.
The organic phase was evaporated under reduced pressure and
the crude was purified by flash chromatography on silica gel
(CHCl3/MeOH/H2O 60 : 35 : 5) to yield 1b (60%), as a white
solid. [α]20D = Ϫ8.1 (c = 0.63, MeOH). MS (FABϩ): 712 [M ϩ
To a solution of the anomerically deprotected disaccharide
(1.56 g, 2.45 mmol, 1 mol eq.) in dry CH2Cl2 (20 ml), under N2,
Cl3CCN (1.32 ml, 13.2 mmol, 5.5 mol eq.) and DBU (63 µl, 0.43
mmol, 0.175 mol eq.) were added. The reaction mixture was
stirred at room temperature for ca. 2 h, then the solvent was
evaporated under reduced pressure and the crude was purified
by short pattern flash chromatography on silica gel (hexane/
1
EtOAc 1 : 1) yielding 5 in 86% yield, as a foam. H-NMR
(CDCl3, 200 MHz): 1.98 (s, 6H, OCOCH3); 2.07 (s, 3H,
OCOCH3); 2.10 (s, 3H, OCOCH3); 2.15 (s, 3H, OCOCH3); 2.18
(s, 3H, OCOCH3); 2.20 (s, 3H, NHCOCH3); 3.9–4.4 (m, 7H,
GalH-5, GalNAcH-6, GalNAcH-6Ј, GalH-6, GalH-6Ј,
GalNAcH-3, GalNAcH-5); 4.6 (m, 1H, GalNAcH-2); 4.72 (d,
1H, GalH-1, J1–2 = 7.4 Hz); 5.0 (dd, 1H, GalH-3, J3–4 = 3.3 Hz,
J3–2 = 10.4 Hz); 5.2 (dd, 1H, GalH-2, J2–1 = 7.4 Hz, J2–3 = 10.4 Hz);
5.4 (d, 1H, GalH-4, J4–3 = 3.3 Hz); 5.47 (d, 1H, GalNAcH-4, J4–3
= 2.2 Hz); 5.72 (d, 1H, NHCOCH3, JNH-H2 = 7.4 Hz); 6.55 (d,
1
H]ϩ; 734 [M ϩ Na]ϩ. H-NMR (D2O, 300 MHz): 1.42 (d, 3H,
CH3L, JCH3L-HL = 6.7 Hz); 1.54 (s, 9H, COOC(CH3)3); 1.63–1.85
(m, 2H, CHDH-5ax, CHDH-2ax); 2.11 (s, 3H, NHCOCH3);
2.25 (d, 1H, CHDH-2eq, J2eq-2ax = 13.6 Hz); 2.36 (d, 1H,
1H, GalNAcH-1, J1–2 = 3 Hz); 8.74 (s, 1H, NH᎐C–CCl ).
᎐
3
CHDH-5eq, J5eq-5ax = 13.6 Hz); 2.74 (t, 1H, CHDH-1, J1–6
≅
J1–2ax ≅ 12.9 Hz); 3.01 (t, 1H, CHDH-6, J6–1 ≅ J6–5ax ≅ 12.9
Hz); 3.58–3.72 (m, 3H, GalH-2, CHDH-3, GalH-3); 3.72–3.91
(m, 9H, GalH-5, GalNAcH-5, COOCH3, GalH-6, GalH-6Ј,
GalNAcH-6, GalNAcH-6Ј); 3.91–4.06 (m, 2H, GalNAcH-3,
GalH-4); 4.12–4.23 (m, 2H, GalNAcH-2, HL); 4.28 (m, 1H,
GalNAcH-4); 4.48 (br s, 1H, CHDH-4); 4.57 (d, 1H, GalH-1,
J1–2 = 6.8 Hz); 5.06 (d, 1H, GalNAcH-1, J1–2 = 8.2 Hz).
13C-NMR-HETCOR (D2O, 300 MHz): 19.2; 23.5; 28.5; 29.3;
33.0; 40.1; 45.5; 52.2; 53.4; 62.0; 69.1; 69.2; 71.3; 72.6; 73.0;
74.4; 75.6; 77.6; 81.1; 102.0; 105.3.
Synthesis of the pseudotrisaccharide 6
A solution of trichloroacetimidate 5 (495 mg, 0.57 mmol,
0.5 mol eq.) and monoether 4 (500 mg, 1.14 mmol, 1 mol eq.) in
dry CH2Cl2 (8 ml), under N2 and in the presence of 4 Å molec-
ular sieves was stirred for 15 min at rt before adding TfOH
(1.3 µl, 0.015 mmol, 0.05 mol eq.). The reaction mixture was
stirred at room temperature for 1 h then was heated at 40 ЊC
and another 0.05 eq. of TfOH were added. After reaction
completion (24 h, monitoring by TLC hexane/EtOAc 3 : 7),
TEA was added to pH 7 and the solvent was evaporated under
reduced pressure. The crude was purified by flash chromato-
graphy on silica gel (toluene/acetone 7 : 3) to afford the pseudo-
Synthesis of 13
To a solution of 12 (35 mg, 0.18 mmol, 1 mol eq.) in dry CH2Cl2
(1 ml), under N2, iPr2NEt (94 µl, 0.54 mmol, 3 mol eq.) was
trisaccharide 6a, as a white solid, in 55% yield. [α]20 = ϩ 0.34
D
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 2 1 1 3 – 2 1 2 4
2120

View Article Online
added. The mixture was stirred at room temperature for 10 min
and a solution of the linker 29 (95 mg, 0.22 mmol, 1.2 mol eq.)
in dry CH2Cl2 (1 ml) and BOP (97 mg, 0.22 mmol, 1.2 mol eq.)
was added. The resulting homogeneous solution was stirred at
room temperature for 24 h monitoring by TLC (CHCl3/MeOH
9 : 1 ϩ 3% TEA). After completion the solvent was evaporated
under reduced pressure and the residue was taken up with
CH2Cl2. The organic phase was washed with 0.5 M NaOH, (2 ×
10 ml), 1 M KHSO4 (1 × 10 ml) and brine (1 × 10 ml), was
finally dried with Na2SO4 and evaporated under reduced
pressure. The crude was purified by flash chromatography on
silica gel (CHCl3/MeOH 95 : 5) to yield pure 13 product as a
CH2-j, CH2-k, CH2-l, CH2-b, CH2-bЈ); 3.68–3.8 (m, 16H, CH2-
bЉ); 3.85 (s, 3H, OCH3); 3.93–4.18 (s, 60H, CH2c, CH2-d, CH2-a,
CH2-aЈ, CH2-aЉ); 5.1 (br s, 8H, NHBoc); 6.42 (s, 4H, CH-4Љ);
6.52 (s, 2H, CH-4Ј); 6.68 (s, 1H, CH-4); 6.92 (s, 8H, CH-2Љ,
CH-6Љ); 6.98 (s, 4H, CH-2Ј, CH-6Ј); 7.16 (s, 2H, CH-2, CH-6);
7.56 (br s, 8H, NH-IV); 7.76 (br s, 8H, NH-III); 7.8–7.98 (bm,
6H, NH-I, NH-II).
13C-NMR (CDCl3, 50.3 MHz): 28.6; 29.2; 29.8; 37.4; 38.6;
39.9; 52.5; 66.6; 69.5; 69.7; 70.2; 70.3; 70.5; 71.1; 79.2; 102.4;
104.6; 106.4; 132.2; 136.6; 156.3; 159.7; 159.8; 166.7; 167.8;
168.9; 169.7.
1
colourless oil (98%). MS (ESI): 614.2 [M ϩ H]ϩ. H-NMR
Synthesis of 14
(CDCl3, 300 MHz): 1.39 (s, 9H, OCOC(CH3)3); 1.6–1.8 (m, 4H,
CH2-m, CH2-f ); 3.12 (m, 2H, CH2-n); 3.35 (m, 2H, CH2-e);
3.4–3.62 (m, 12H, CH2-g, CH2-h, CH2-i, CH2-j, CH2-k, CH2-l);
3.68 (m, 2H, CH2-b); 3.85 (s, 3H, OCH3); 3.99 (s, 2H, CH2c);
4.02 (s, 2H, CH2-d); 4.05 (m, 2 H, CH2-a); 5.01 (br s, 1H,
NHBoc); 7.03 (m, 1H, CH-4); 7.2 (s, 1H, NH-II); 7.28 (t, 1H,
CH-5, J5–4 ≅ J5–6 = 7.9 Hz); 7.38 (br s, 1H, NH-I); 7.48 (s, 1H,
CH-2); 7.58 (d, 1H, CH-6, J6–5 = 7.8 Hz). 13C-NMR (CDCl3,
75.0 MHz): 28.4; 28.9; 29.6; 37.4; 38.4; 38.5; 52.2; 66.7; 69.4;
69.7; 70.0; 70.2; 70.3; 71.1; 71.2; 78.8; 114.8; 119.6; 122.4; 129.5;
131.5; 156.2; 158.4; 166.8; 168.4; 169.1.
The same procedure as for the monovalent ligand was
adopted to functionalise the divalent, tetravalent and octavalent
dendrimeric cores using equivalents of reagents in proportion
to the number of free amine groups. The solvents used in
the purification and the characterization of each product are
reported.
To a solution of 13, (26 mg, 0.042 mmol, 1 mol eq.) in dry
CH2Cl2 (0.25 ml), TFA (0.25 ml) was added. The reaction
mixture was stirred at rt for 1 h then the solvent was evaporated
and the residue was taken up with toluene and evaporated
several times to completely remove the TFA. The crude was
dried under vacuum for ca. 18 h and redissolved in dry CH2Cl2
(1.5 ml). Under N2, Et3N (17 µl, 0.126 mmol, 3 mol eq.) was
added and after 10 min, HBTU (24 mg, 0.063 mmol, 1.5 mol
eq.) and the pseudosaccharide 6b (50 mg, 0.05 mmol, 1.2 mol
eq.) were added. The reaction mixture was stirred at rt for 18 h,
monitoring by TLC (CH2Cl2/MeOH 9 : 1). After reaction
completion the solvent was evaporated under reduced pressure
and the residue was dissolved in CH2Cl2. The organic phase was
washed with 0.5 M NaOH (2 times), 1 M KHSO4 (1 time) and
brine (1 time), dried with Na2SO4 and the solvent was evapor-
ated under reduced pressure. The crude was purified by flash
chromatography on silica gel (CHCl3/MeOH 95 : 5) to yield 14
1
(76%). MS (ESI): 1493.8 [M ϩ H]ϩ; 1516.0 [M ϩ Na]ϩ. H-
MeO2C-[G1](LinkerNHBoc)2 18. The pure product was
obtained after flash chromatography on silica gel (CHCl3/
MeOH 9 : 1) as a colourless oil (98%). MS (ESI): 1091.8 [M ϩ
H]ϩ. 1H-NMR (CDCl3, 200 MHz): 1.42 (s, 18 H, OCOC(CH3)3);
1.65–1.86 (m, 8H, CH2-m, CH2-f ); 3.12–3.28 (m, 4H, CH2-n);
3.38–3.48 (m, 4H, CH2-e); 3.48–3.64 (m, 24H, CH2-g, CH2-h,
CH2-i, CH2-j, CH2-k, CH2-l); 3.64–3.78 (m, 4H, CH2-b); 3.92
(s, 3H, OCH3); 4.02 (s, 4H, CH2c); 4.03–4.18 (m, 8 H, CH2-d,
CH2-a); 5.03 (br s, 2H, NHBoc); 6.68 (s, 1H, CH-4); 7.2 (s, 2H,
CH-2, CH-6); 7.22 (br s, 2H, NH-II), 7.4 (br s, 2H, NH-I).
13C-NMR (CDCl3, 50.3 MHz): 28.5; 29.0; 29.8; 37.4; 38.6; 43.3;
52.4; 66.8; 69.4; 69.7; 70.1; 70.2; 70.4; 71.1; 71.2; 77.9; 106.7;
108.4; 132.2; 156.4; 159.7; 166.7; 169.0; 169.8.
NMR (acetone-D6, 400 MHz): 1.41 (d, 3H, CH3L, JCH3-HL = 7
Hz); 1.49 (dd, 2H, CHDH-5ax, J5ax-5eq ≅ J5ax-6 = 13 Hz); 1.72 (m,
2H, CH2-m); 1.76 (m, 2H, CH2-f ); 1.80–1.98 (m, 8H, OCOCH3,
NHCOCH3, CHDH-2ax, CHDH-2eq); 1.99–2.02 (m, 6H, 2
OCOCH3); 2.06 (m, 3H, OCOCH3); 2.10 (m, 3H, OCOCH3);
2.11 (m, 3H, OCOCH3); 2.18 (m, 1H, CHDH-5eq); 2.49 (m,
1H, CHDH-1); 3.02 (m, 1H, CHDH-6); 3.25 (m, 2H, CH2-n);
3.32 (m, 2H, CH2-e); 3.5 (m, 2H, CH2-l); 3.52–3.63 (m, 14H,
CH2-g, CH2-h, CH2-i, CH2-j, CH2-k, CHDH-3, OCH3); 3.69
(m, 2H, CH2-b); 3.89–3.96 (m, 5H, OCH3, GalNAcH-5,
GalNAcH-2); 4.0 (m, 1H, GalNAcH-6 or GalNAcH-6Ј); 4.04
(s, 2H, CH2-d); 4.07–4.17 (s, 6H, GalNAcH-6 or GalNAcH-6Ј,
CH2c, GalH-6, GalH-6Ј, GalH-5); 4.19 (m, 2 H, CH2-a); 4.27
(br s, 1H, CHDH-4); 4.37 (m, 1H, HL); 4.40 (m, 1H, GalNAcH-
3); 4.90 (d, 1H, GalH-1, J1–2 = 10.8 Hz); 5.04–5.1 (m, 3H, GalH-
2, GalH-3, GalNAcH-1); 5.22 (m, 2H, OCH2Ph); 5.36 (br s, 1H,
GalH-4); 5.42 (d, 1H, GalNAcH-4); 7.0 (br s, 1H, NH-III);
7.2–7.28 (m, 2H, NHCOCH3, CH-4); 7.35–7.48 (m, 6H, CH-5
aromatic protons); 7.54–7.6 (m, 2H, NH-II, CH-2), 7.03 (d, 1H,
CH-6); 7.88 (br s, 1H, NH-I).13C-NMR-HETCOR (acetone-
D6, 400 MHz): 18.8; 20.0; 20.1; 23.1; 29.9; 30.0; 33.9; 37.0
(2 carbons); 38.8; 39.2; 45.3; 51.2; 52.0; 53.5; 61.4, 63.0; 66.9;
67.3; 67.8; 69.0; 69.1; 69.9; 70.7; 71.0; 71.5 (3 carbons); 72.0;
72.3; 73.0; 76.2; 79.0; 101.9; 102.0; 106.0; 120.4; 129.0; 130.7.
The same procedure as for the monovalent ligand was
adopted to functionalise the divalent, tetravalent and octavalent
dendrimers with the sugar. Reagents were used in proportion to
the number of free amine groups. The solvents used for the
purification and the characterization of each product are
reported.
MeO2C-[G2](LinkerNHBoc)4 22. The pure product was
obtained after flash chromatography on silica gel (CHCl3/
MeOH 95 : 5 8 : 2) as a colourless oil (50%). MS (ESI): 1187.3
[M ϩ 2H]2ϩ; 1197.7 [M ϩ H ϩ Na]2ϩ; 1209.5 [M ϩ 2Na]2ϩ
.
1H-NMR (CDCl3, 200 MHz): 1.42 (s, 36H, OCOC(CH3)3);
1.64–1.86 (m, 16H, CH2-m, CH2-f ); 3.10–3.22 (m, 8H, CH2-n);
3.30–3.43 (m, 8H, CH2-e); 3.43–3.78 (m, 56H, CH2-g, CH2-h,
CH2-i, CH2-j, CH2-k, CH2-l, CH2-bЈ); 3.78–3.96 (m, 7H, CH2-b,
OCH3); 3.99–4.18 (s, 24H, CH2c, CH2-d, CH2-aЈ); 4.18–4.26 (m,
4H, CH2-a); 5.06 (br s, 4H, NHBoc); 6.58 (s, 2H, CH-4Ј); 6.72
(s, 1H, CH-4); 7.0 (d, 4H, CH-2Ј, CH-6Ј, J2Ј-4Ј ≅ J6Ј–4Ј = 2.2 Hz);
7.2 (d, 2H, CH-2, CH-6, J2–4 ≅ J6–4 = 2.2 Hz); 7.34 (br s, 4H,
NH-III); 7.46 (br s, 4H, NH-II); 7.63 (br s, 2H, NH-I). 13C-
NMR (CDCl3, 50.3 MHz): 28.6; 29.1; 29.8; 37.5; 38.5; 39.5;
52.4; 66.6; 66.9; 69.5; 69.9; 70.0; 70.2; 70.3; 70.5; 70.6; 71.1;
79.1; 104.6; 106.5; 106.8; 108.4; 132.2; 136.8; 156.3; 159.9;
166.7; 167.5; 168.8; 169.5.
MeO2C-[G3](LinkerNHBoc)8 26. The pure product was
obtained after flash chromatography on silica gel (CHCl3/
MeOH 9 : 1 8 : 2) as a colourless oil (40%). MS (ESI): 1256.8
[M ϩ 4H]4ϩ; 1627.1 [M-Boc ϩ 2Na ϩ H]3ϩ; 1634.5 [M-Boc ϩ
Divalent compound 19
The pure product was obtained after flash chromatography on
silica gel (CHCl3/MeOH 95 : 5 9 : 1) as a colourless oil (66%).
3Na]3ϩ; 1668.0 [M ϩ 3Na]2ϩ; 1673.2 [M ϩ 2Na ϩ K]3ϩ
.
MS (ESI): 951.2 [M ϩ 3H]3ϩ; 1117.2 [M-GalGalNAc ϩ 2H]2ϩ
;
1H-NMR (CDCl3, 200 MHz): 1.41 (s, 72H, OCOC(CH3)3);
1.64–1.86 (m, 32H, CH2-m, CH2-f ); 3.18 (m, 16H, CH2-n); 3.35
(m, 16H, CH2-e); 3.43–3.67 (m, 108H, CH2-g, CH2-h, CH2-i,
1226.2 [M-Gal ϩ 2H]2ϩ; 1426.6 [M ϩ 2H]2ϩ; 1437.9 [M ϩ Na ϩ
1
H]2ϩ. H-NMR (acetone-D6, 400 MHz): 1.41 (d, 6H, CH3L,
JCH3-HL = 6.7 Hz); 1.5 (dd, 2H, CHDH-5ax, J5ax-5eq ≅ J5ax-6 = 13
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 2 1 1 3 – 2 1 2 4
2121

View Article Online
Hz); 1.66–1.80 (m, 8H, CH2-m, CH2-f ); 1.86–2.02 (m, 28H, 3
OCOCH3, NHCOCH3, CHDH-2ax, CHDH-2eq); 2.02–2.15
(m, 18H, 3 OCOCH3); 2.18 (m, 2H, CHDH-5eq); 2.50 (m, 2H,
CHDH-1); 3.02 (m, 2H, CHDH-6); 3.25 (m, 4H, CH2-n); 3.32
(m, 4H, CH2-e); 3.47–3.63 (m, 32H, CH2-g, CH2-h, CH2-i,
CH2-j, CH2-k, CH2-l, CHDH-3, OCH3); 3.68 (m, 4H, CH2-b);
3.85–3.92 (m, 7H, OCH3, GalNAcH-5, GalNAcH-2); 3.92–4.18
(s, 18H, GalNAcH-6, GalNAcH-6Ј, CH2-d, CH2c, GalH-6,
GalH-6Ј, GalH-5); 4.18 (m, 4 H, CH2-a); 4.26 (br s, 2H,
CHDH-4); 4.32–4.41 (m, 4H, HL, GalNAcH-3); 4.89 (d, 2H,
GalH-1, J1–2 = 6 Hz); 5.02–5.1 (m, 6H, GalH-2, GalH-3, Gal-
NAcH-1); 5.2 (m, 4H, OCH2Ph); 5.36 (br s, 2H, GalH-4); 5.42
(d, 2H, GalNAcH-4); 6.73 (s, 1H, CH-4); 7.06 (br s, 2H,
NH-III); 7.18 (d, 2H, CH-2, CH-6, J2Ј-4Ј≅J6Ј–4Ј = 2.2 Hz); 7.26 (d,
2H, NHCOCH3, JNH-H2 = 7.8 Hz); 7.42 (s, 10H, aromatic
protons); 7.65 (br s, 2H, NH-II); 7.92 (br s, 2H, NH-I).
13C-NMR-HETCOR (acetone-D6, 400 MHz): 18.5; 20.1
(6 carbons); 23.1; 29.6; 29.8; 29.9; 33.9; 36.8 (2 carbons); 36.9;
38.6; 45.4; 51.5; 51.8; 53.4; 61.1; 62.8; 66.6; 67.3; 67.9; 69.2;
69.8; 70.8; 71.4; 71.5 (2 carbons); 72.0 (2 carbons); 72.3; 72.7;
76.4; 78.8; 101.7; 102.0; 106.8; 108.7; 129.6.
12H, CH2-bЈ, CH2-b); 3.86 (s, 3H, OCH3); 3.89–4.04 (m, 24H,
GalNAcH-5, GalNAcH-2, GalNAcH-6 or GalNAcH-6Ј); 4.06
(s, 16H, CH2-d); 4.07–4.18 (m, 64H, GalNAcH-6 or GalNAcH-
6Ј, CH2c, GalH-6, GalH-6Ј, GalH-5, CH2-aЉ); 4.20–4.28 (m,
20H, CHDH-4, CH2-aЈ, CH2-a); 4.33–4.40 (m, 16H, HL,
GalNAcH-3); 4.90 (d, 8H, GalH-1, J1–2 = 11.5 Hz); 5.03–5.08
(m, 24H, GalH-2, GalH-3, GalNAcH-1); 5.21 (m, 16H,
OCH2Ph); 5.35 (br s, 8H, GalH-4); 5.42 (d, 8H, GalNAcH-4,
J4–3 = 3.4 Hz); 6.65 (s, 4H, CH-4Љ); 6.70 (s, 2H, CH-4Ј); 6.86 (s,
1H, CH-4), 7.08–7.18 (m, 22H, NH-V, CH-2Љ, CH-6Љ, CH-2Ј,
CH-6Ј, CH-2, CH-6); 7.25 (d, 8H, NHCOCH3, JNH-H2 = 7.8
Hz); 7.42 (s, 40H, aromatic protons); 7.78 (br s, 8H, NH-IV);
8.02 (br s, 8H, NH-III); 8.35 (br s, 6H, NH-II, NH-I).
13C-NMR (acetone-D6, 75.0 MHz): 19.2; 20.7; 20.9; 23.8; 29.5;
29.8; 30.03; 34.3; 37.2; 39.1; 39.7; 45.7; 51.5; 51.8; 53.6; 61.7,
63.5; 67.1; 67.4;68.0; 69.4; 69.6; 69.7; 70.1; 70.8; 71.1; 71.2; 71.7;
72.1; 72.7; 73.0; 76.9; 79.0; 101.9 (2 anomeric carbons); 107.1;
108.2; 128.9; 129.1; 129.5; 137.6; 160.7; 167.7; 169.8; 170.4;
170.7; 173.7; 174.0; 176.0.
Synthesis of 16
Tetravalent compound 23
To a solution of 14 (42 mg, 0.028 mmol, 1 mol eq.) in MeOH
(1.5 ml), a catalytic amount of Pd/C 10% was added. The
mixture was stirred under H2 atmosphere at room temperature
for ca. 1 h. After reaction completion, the mixture was filtered
through a Celite pad, washing several times with MeOH and
the collected organic phase was evaporated under reduced
pressure. The crude was purified by flash chromatography on
silica gel (CHCl3/MeOH/H2O 60 : 35 : 5) and the pure product
(35 mg, 0.025 mmol, 1 mol eq.) was dissolved in dry MeOH
(2 ml). Under N2, a solution of 1 M MeONa in MeOH (45 µl,
0.045 mmol, 1.8 mol eq.) was added and the reaction was
stirred at room temperature for ca. 45 min monitoring by TLC
(CHCl3/MeOH/H2O 60 : 35 : 5 or EtOAc/AcOH/MeOH/H2O
4 : 3 : 3 : 2). Just after 5 min a white solid precipitated. After
completion, the reaction was quenched adding Amberlite IR
120 (Hϩ form) to pH 5–6. The suspension was stirred for some
minutes till the white solid completely dissolved, finally the
resin was filtered and washed several times with MeOH. The
organic phase was evaporated under reduced pressure. The final
product was obtained as a white solid and results contaminated
by 33% of imide by-product (79% over two steps). MS (ESI):
754.7 [M-GalGalNAc-32 ϩ H]ϩ; 786.8 [M-GalGalNAc ϩ H]ϩ;
1119.8 [M-32 ϩ H]ϩ; 1151.7 [M ϩ H]ϩ; 1174.6 [M ϩ Na]ϩ. 1H-
NMR (D2O, 400 MHz): 1.32 (d, 3H, CH3L, JCH3-HL = 7 Hz); 1.52
(dd, 1H, CHDH-5ax, J5ax-6 = 13 Hz); 1.62–1.88 (m, 5H, CH2-m,
CH2-f, CHDH-2ax); 1.9 (m, 1H, CHDH-2eq); 2.0 (s, 3H,
NHCOCH3); 2.26 (m, 1H, CHDH-5eq); 2.5 (dd, 1H, CHDH-1,
J1–6 ≅ J1–2ax = 12 Hz); 2.98 (dd, 1H, CHDH-6, J6–1 ≅ J6–5ax = 13
Hz); 3.18 (m, 2H, CH2-n); 3.24 (m, 2H, CH2-e); 3.34–3.54 (m,
6H, CH2-g, CH2-l, CHDH-3, GalH-2); 3.54–3.68 (m, 16H,
CH2-h, CH2-i, CH2-j, CH2-k, OCH3, GalH-3, GalH-5,
GalNAcH-5, CH2-b); 3.68–3.80 (m, 4H, GalH-6, GalH-6Ј,
GalNAcH-6, GalNAcH-6Ј); 3.82 (m, 1H, GalNAcH-3); 3.85
(d, 1H, GalH-4, J4–3 = 3.3 Hz); 3.87 (m, 3H, OCH3); 3.92–4.01
(m, 3H, GalNAcH-2, CH2-d); 4.40 (s, 2H, CH2c); 4.13 (d, 1H,
GalNAcH-4, J4–3 = 3.1 Hz); 4.2 (m, 2H, CH2-a); 4.25–4.35 (m,
2H, HL, CHDH-4); 4.42 (d, 1H, GalH-1, J1–2 = 7.6 Hz); 4.73 (m,
1H, GalNAcH-1); 7.21 (d, 1H, CH-4, J4–5 = 8.2 Hz); 7.42 (dd,
1H, CH-5, J5–4 ≅ J5–6 = 8.2 Hz); 7.53 (br s, 1H, CH-2); 7.60 (d,
1H, CH-6, J6–5 = 8.2 Hz). 13C-NMR-HETCOR (D2O, 400
MHz): 18.9; 23.0; 28.9; 29.7; 32.9; 36.5; 36.8 (2 carbons); 38.9;
39.1; 45.2; 52.1; 53.0; 53.1; 61.7; 67.7; 68.8; 69.1; 70.0; 70.4;
71.2; 72.6; 73.0; 73.1; 75.5; 78.1; 80.3; 102.0; 105.7.
The pure product was obtained after flash chromatography on
silica gel (CHCl3/MeOH 95 : 5 8 : 2) as a colourless oil (55%).
MS (ESI): 1474.1 [M ϩ 4H]4ϩ; 1495.1 [M ϩ 4Na]4ϩ. H-NMR
1
(acetone-D6, 400 MHz): 1.4 (d, 12H, CH3L, JCH3-HL = 7 Hz); 1.5
(dd, 4H, CHDH-5ax, J5ax-5eq ≅ J5ax-6 = 13 Hz); 1.70 (m, 8H,
CH2-m); 1.76 (m, 8H, CH2-f ); 1.90–2.02 (m, 56H, 3 OCOCH3,
NHCOCH3, CHDH-2ax, CHDH-2eq); 2.06 (s, 12H,
OCOCH3); 2.11 (s, 12H, OCOCH3); 2.13 (s, 12H, OCOCH3);
2.19 (m, 4H, CHDH-5eq); 2.5 (ddd, 4H, CHDH-1, J1–6 ≅ J1–2ax
10.8 Hz, J1–2eq = 3.8 Hz); 3.03 (ddd, 4H, CHDH-6, J6–1 ≅ J6–5ax
≅
≅
11 Hz); 3.24 (m, 8H, CH2-n); 3.33 (m, 8H, CH2-e); 3.49 (m, 8H,
CH2-l); 3.51–3.63 (m, 56H, CH2-g, CH2-h, CH2-i, CH2-j, CH2-k,
CHDH-3, OCH3); 3.66 (m, 8H, CH2-bЈ); 3.83 (m, 4H, CH2-b);
3.88 (s, 3H, OCH3); 3.89–4.03 (m, 12H, GalNAcH-5,
GalNAcH-2, GalNAcH-6 or GalNAcH-6Ј); 4.05 (s, 8H, CH2-
d); 4.07–4.18 (s, 32H, GalNAcH-6 or GalNAcH-6Ј, CH2c,
GalH-6, GalH-6Ј, GalH-5, CH2-aЈ); 4.23–4.30 (m, 8H, CHDH-
4, CH2-a); 4.33–4.40 (m, 8H, HL, GalNAcH-3); 4.90 (d, 4H,
GalH-1, J1–2 = 11.5 Hz); 5.03–5.1 (m, 12H, GalH-2, GalH-3,
GalNAcH-1); 5.22 (m, 8H, OCH2Ph); 5.36 (br s, 4H, GalH-4);
5.42 (d, 4H, GalNAcH-4, J4–3 = 3.4 Hz); 6.7 (s, 2H, CH-4Ј); 6.88
(s, 1H, CH-4), 7.08 (br s, 4H, NH-IV); 7.1 (d, 4H, CH-2Ј, CH-
6Ј); 7.16 (d, 2H, CH-2, CH-6); 7.25 (d, 4H, NHCOCH3, JNH-H2
= 7.8 Hz); 7.42 (s, 20H, aromatic protons); 7.68 (br s, 4H,
NH-III); 7.93 (br s, 4H, NH-II); 8.02 (br s, 2H, NH-I). 13C-
NMR-HETCOR (acetone-D6, 400 MHz): 18.3; 19.9; 20.1;
23.0; 29.7; 29.9; 33.7; 36.8 (2 carbons); 38.4; 39.2; 39.7; 45.1;
51.3; 52.1; 53.1; 61.1; 62.9; 66.6; 67.0; 67.1; 67.6; 69.1 (2
carbons); 69.8; 70.4; 70.8; 71.0; 71.1; 71.6; 72.2; 72.6; 76.2; 78.8;
101.6; 101.7; 104.9; 106.8; 107.0; 108.9; 129.1.
Octavalent compound 27
The pure product was obtained after flash chromatography on
silica gel (CHCl3/MeOH 9 : 1 8 : 2) as a colourless oil (35%).
MS (ESI): 1433.4 [M ϩ 7Na]7ϩ; 2018.7 [M ϩ 6Na]6ϩ; 2417.6
[M ϩ 5Na]5ϩ; 3015.5 [M ϩ 4Na]4ϩ. H-NMR (acetone-D6,
1
400 MHz): 1.41 (d, 24H, CH3L, JCH3-HL = 7 Hz); 1.5 (dd, 8H,
CHDH-5ax, J5ax-5eq ≅ J5ax-6 = 11.4 Hz); 1.70 (m, 16H, CH2-m);
1.76 (m, 16H, CH2-f ); 1.90–2.02 (m, 112H, 3 OCOCH3,
NHCOCH3, CHDH-2ax, CHDH-2eq); 2.06 (s, 24H,
OCOCH3); 2.11 (s, 24H, OCOCH3); 2.13 (s, 24H, OCOCH3);
2.19 (m, 8H, CHDH-5eq); 2.51 (ddd, 8H, CHDH-1, J1–6 ≅ J1–2ax
≅ 10.8 Hz, J1–2eq = 3.8 Hz); 3.03 (ddd, 8H, CHDH-6, J6–1 ≅ J6–5ax
≅ 11.4 Hz); 3.24 (m, 16H, CH2-n); 3.33 (m, 16H, CH2-e);
3.45–3.63 (m, 128H, CH2-g, CH2-h, CH2-i, CH2-j, CH2-k, CH2-
l, CHDH-3, OCH3); 3.63–3.68 (m, 16H, CH2-bЉ); 3.75–3.83 (m,
The same procedure as for the monovalent ligand 14 was
adopted to deprotect the divalent, tetravalent and octavalent
ligands. Reagents were used in proportion to the number
of sugar moieties to be deprotected. The yield and the charac-
terization of each product are reported.
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 2 1 1 3 – 2 1 2 4
2122

View Article Online
Divalent compound 20
NAcH-2, CH2-d, CH2c, CH2-aЉ, CH2-aЈ, CH2-a); 4.12 (m, 8H,
GalNAcH-4); 4.17–4.30 (m, 16H, HL, CHDH-4); 4.49 (d, 8H,
GalH-1, J1–2 = 7.8 Hz); 4.80 (d, 8H, GalNAcH-1, J1–2 = 9.4 Hz);
6.40–6.55 (m, 7H, CH-4Љ, CH-4Ј, CH-4); 6.67–6.82 (m, 14H,
CH-2Љ, CH-6Љ, CH-2Ј, CH-6Ј, CH-2, CH-6). 13C-NMR-
HETCOR (D2O, 400 MHz): 19.0; 23.0; 29.0; 29.8; 33.0; 37.0;
39.1; 39.5; 40.5; 45.6; 52.5; 53.1; 53.4; 61.9, 67.6; 67.8; 69.0;
69.2; 69.6; 70.8; 71.0; 71.9; 73.1; 73.4; 74.9; 76.0; 78.2; 80.9;
102.3; 105.9; 106.0; 107.0; 108.5; 110.0.
The pure product was obtained as a white solid and results
contaminated by 20% of imide by-product (83% over two
steps). MS (ESI): 1068.5 [M-32 ϩ 2H]2ϩ; 1083.9 [M ϩ 2H]2ϩ
.
1H-NMR (D2O, 400 MHz): 1.39 (d, 6H, CH3L, JCH3-HL = 7 Hz);
1.54 (dd, 2H, CHDH-5ax, J5ax-5eq ≅ J5ax-6 = 11.7 Hz); 1.66–1.83
(m, 10H, CH2-m, CH2-f, CHDH-2ax); 1.92 (m, 2H, CHDH-
2eq); 2.0 (s, 6H, NHCOCH3); 2.29 (m, 2H, CHDH-5eq); 2.55
(dd, 2H, CHDH-1, J1–6 ≅ J1–2ax = 12 Hz); 3.0 (dd, 2H, CHDH-6,
J6–1 ≅ J6–5ax = 11.7 Hz); 3.18 (m, 4H, CH2-n); 3.26 (m, 4H, CH2-
e); 3.34–3.54 (m, 12H, CH2-g, CH2-l, CHDH-3, GalH-2); 3.54–
3.71 (m, 32H, CH2-h, CH2-i, CH2-j, CH2-k, OCH3, GalH-3,
GalH-5, GalNAcH-5, CH2-b); 3.71–3.85 (m, 8H, GalH-6,
GalH-6Ј, GalNAcH-6, GalNAcH-6Ј); 3.85–3.93 (m, 7H,
GalNAcH-3, GalH-4, OCH3); 4.01 (m, 2H, GalNAcH-2); 4.03
(s, 4H, CH2-d); 4.10 (s, 4H, CH2؊c); 4.13–4.22 (m, 6H, CH2-a,
GalNAcH-4); 4.22–4.36 (m, 4H, HL, CHDH-4); 4.44 (d, 2H,
GalH-1, J1–2 = 7.8 Hz); 4.79 (d, 2H, GalNAcH-1, J1–2 = 9.4 Hz);
6.80 (s, 1H, CH-4); 7.18 (s, 2H, CH-2, CH-6).
Synthesis of 32
To a solution of MeO2C-[G1](LinkerNHBoc)2 18, (62.5 mg,
0.057 mmol, 1 mol eq.) in dry CH2Cl2 (0.1 ml), TFA (0.1 ml)
was added. The reaction mixture was stirred for 1 h, then the
solvent was evaporated and the residue taken up with toluene,
which was evaporated several times to completely remove the
TFA. Then, the crude was dissolved in CH2Cl2 and Amberlyst
A-21 (-N(CH3)2 form) was added to pH 8. The suspension was
stirred for 45 min. Finally the resin was filtered and washed
several times with MeOH. The organic phase was evaporated
under reduced pressure. The crude was dissolved in dry MeOH
(0.5 ml) and, under N2, -(ϩ)-ribonic-γ-lactone 31 (34 mg,
0.228 mmol, 4 mol eq.) was added. The reaction mixture was
stirred at 45 ЊC for 18 h, monitoring by TLC (CHCl3/MeOH
9 : 1 ϩ 1% H2O). After reaction completion the solvent was
evaporated and the crude redissolved in Ac2O (0.4 ml). Under
N2, at 0 ЊC, pyridine (0.45 ml) was added. The reaction mixture
was stirred at room temperature for 18 h monitoring by TLC
(CHCl3/MeOH 8 : 2 ϩ 1% H2O). After reaction completion the
solvent was evaporated. The residue was dissolved in CH2Cl2
and washed with brine (2 times). The organic phase was dried
with Na2SO4 and the solvent was evaporated under reduced
pressure. The crude was purified by flash chromatography on
silica gel (CHCl3/MeOH 8 : 2 ϩ 1% H2O) yielding 26.4 mg of
fully acetylated compound (30%). This compound (23 mg,
1.5 µmol, 1 mol eq.) was dissolved in dry MeOH (1 ml), under
N2, and a solution of 1 M MeONa in MeOH (0.05 ml, 0.05
mmol, 3.5 mol eq.) was added. The reaction was stirred at room
temperature for 1 h monitoring by TLC (CHCl3/MeOH/H2O
60 : 35 : 5). After completion, the reaction was quenched adding
Amberlite IR 120 (Hϩ form) to pH 5–6. Finally the resin was
filtered, washing several times with MeOH, and the organic
phase was evaporated under reduced pressure. The final
product 32 was obtained as a white solid (60% yield). 1H-NMR
(D2O, 400 MHz): 1.67–1.78 (m, 8H, CH2-m, CH2-f ); 3.2–3.27
(m, 8H, CH2-n, CH2-e); 3.45–3.52 (m, 8H, CH2-g, CH2-l); 3.52–
3.61 (m, 18H, CH2-h, CH2-i, CH2-j, CH2-k, Rib-H4Ј); 3.61–3.65
(m, 4H, CH2-b); 3.72–3.79 (m, 4H, RibH-3, RibH-4); 3.84–3.90
(m, 5H, RibH-2, OCH3); 4.0 (s, 4H, CH2-d); 4.50 (s, 4H,
CH2؊c); 4.16 (m, 4H, CH2-a); 4.27 (d, 2H, RibH-1, J1–2 = 3.4
Hz), 6.76 (s, 1H, CH-4); 7.15 (s, 2H, CH-2, CH-6). 13C-NMR-
HETCOR (D2O, 400 MHz): 28.6; 36.9; 39.2; 53.0; 63.9
(2 carbons); 67.8; 69.3; 70.6; 71.0; 71.9; 73.8; 74.0; 108.2; 109.5.
13C-NMR-HETCOR (D2O, 400 MHz): 18.6; 22.9; 28.7; 29.4;
32.8; 36.8 (2 carbons); 39.0; 39.3; 45.1; 52.1; 53.0; 53.2; 61.3;
67.8; 68.9; 69.0; 70.0; 70.4; 72.9; 75.3; 78.0; 80.3; 101.8; 105.2;
107.9; 109.1.
Tetravalent ligand 24
The pure product was obtained as a white solid and contained
30% of the imide by-product (79% over two steps). MS (ESI):
1115.7 [M-64 ϩ 4H]4ϩ; 1132.0 [M ϩ 4H]4ϩ; 1508.8 [M ϩ 3H]3ϩ
;
1536.7 [M ϩ 2Na ϩ K]3ϩ. 1H-NMR (D2O, 400 MHz): 1.40 (d,
12H, CH3L, JCH3-HL = 7 Hz); 1.54 (dd, 4H, CHDH-5ax, J5ax-5eq
≅
J5ax-6 = 11.7 Hz); 1.69–1.86 (m, 20H, CH2-m, CH2-f, CHDH-
2ax); 1.95 (m, 4H, CHDH-2eq); 2.03 (s, 12H, NHCOCH3); 2.32
(m, 4H, CHDH-5eq); 2.57 (dd, 4H, CHDH-1, J1–6 ≅ J1–2ax = 12
Hz); 3.02 (dd, 4H, CHDH-6, J6–1 ≅ J6–5ax = 11.7 Hz); 3.12–3.30
(m, 16H, CH2-n, CH2-e); 3.43–3.54 (m, 24H, CH2-g, CH2-l,
CHDH-3, GalH-2); 3.54–3.72 (m, 64H, CH2-h, CH2-i, CH2-j,
CH2-k, OCH3, GalH-3, GalH-5, GalNAcH-5, CH2-bЈ); 3.72–
3.85 (m, 20H, GalH-6, GalH-6Ј, GalNAcH-6, GalNAcH-6Ј,
CH2-b); 3.87 (s, 3H, OCH3); 3.89–3.98 (m, 8H, GalNAcH-3,
GalH-4); 3.98–4.09 (m, 12H, GalNAcH-2, CH2-d); 4.09–4.18
(m, 16H, CH2c, CH2-aЈ); 4.18–4.30 (m, 12H, GalNAcH-4, CH2-
a, HL); 4.33 (br s, 4H, CHDH-4); 4.48 (d, 4H, GalH-1, J1–2 = 7.8
Hz); 4.82 (d, 4H, GalNAcH-1, J1–2 = 9.4 Hz); 6.62 (s, 2H,
CH-4Ј); 6.70 (s, 1H, CH-4); 7.78 (s, 4H, CH-2Ј, CH-6Ј); 7.05 (s,
2H, CH-2, CH-6). 13C-NMR-HETCOR (D2O, 400 MHz): 18.9;
23.0; 29.0; 29.8; 33.0; 37.0; 39.1; 39.5; 40.5; 45.6; 52.5; 53.1;
53.4; 61.9, 67.6; 67.8; 69.0; 69.2; 69.6; 70.8; 71.0; 71.9; 73.1;
73.4; 74.9; 76.0; 78.2; 80.9; 102.3; 105.9; 106.9; 107.0; 108.5;
110.0.
Octavalent ligand 28
The pure product was obtained as a white solid and contained
20% of imide by-product (86% over two steps). MS (ESI):
1264.6 [M-32-GalGalNac ϩ 7H]7ϩ; 1270.9 [M-GalGalNac ϩ
7H]7ϩ; 1475.7 [M-32-GalGalNac ϩ 6H]6ϩ; 1481.4 [M-GalGal-
Nac ϩ 6H]6ϩ; 1565.9 [M ϩ 6Na]6ϩ; 1579.6 [M ϩ 6K]6ϩ; 1868.3
Surface plasmon resonance
The SPR experiments were performed on a double channel
surface plasmon Autolab ESPRIT instrument (Ecochemie,
Utrecht, The Netherlands) at 25ЊC in HBS buffer (10 mM
Hepes, 3.4 mM EDTA, 150 mM NaCl and 0.005% Tween-20,
pH 7.4). The cuvet-based instrument was mounted with a
CMD6 biosensor chip (XanTec bioanalytics GmbH, Münster,
Germany) containing a carboxymethylated dextran layer with a
molecular weight of 6000. Onto this surface the glycoprotein
ASF was immobilized according to a standard amine coupling
procedure. Prior to activation of the carboxy functional group
the sensor was presoaked in buffer for a few hours. Activation
was achieved in 7 min by a mixture of N-ethyl-NЈ-[3-(di-
methylamino)propyl]carbodiimide hydrochloride (EDC, 0.2 M)
and N-hydroxysuccinimide (NHS, 0.05 M). The protein, ASF at
[M-32 ϩ 5Na]5ϩ; 1874.8 [M ϩ 5Na]5ϩ. H-NMR (D2O, 400
1
MHz): 1.32 (d, 24H, CH3L, JCH3-HL = 7 Hz); 1.52 (dd, 8H,
CHDH-5ax, J5ax-5eq ≅ J5ax-6 = 12.4 Hz); 1.59–1.68 (m, 40H,
CH2-m, CH2-f, CHDH-2ax); 1.93 (m, 8H, CHDH-2eq); 2.0 (s,
24H, NHCOCH3); 2.28 (m, 8H, CHDH-5eq); 2.53 (dd, 8H,
CHDH-1, J1–6 ≅ J1–2ax = 13.6 Hz); 3.02 (dd, 8H, CHDH-6, J6–1
≅
J6–5ax = 12.4 Hz); 3.08–3.37 (m, 32H, CH2-n, CH2-e); 3.35–3.46
(m, 48H, CH2-g, CH2-l, CHDH-3, GalH-2); 3.46–3.62 (m,
128H, CH2-h, CH2-i, CH2-j, CH2-k, OCH3, GalH-3, GalH-5,
GalNAcH-5, CH2-bЉ); 3.62–3.77 (m, 44H, GalH-6, GalH-6Ј,
GalNAcH-6, GalNAcH-6Ј, CH2-bЈ, CH2-b); 3.77–3.88 (m,
19H, OCH3, GalNAcH-3, GalH-4); 3.88–4.07 (m, 68H, Gal-
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 2 1 1 3 – 2 1 2 4
2123

View Article Online
a concentration of 400 µg mlϪ1 in 10 mM sodium acetate buffer,
pH 4, was coupled to the activated surface for 15 min, after
which the remaining NHS esters were blocked by the addition
of ethanolamine (1 M, pH 8.5) for 10 min. The double channel
system consisted of a sample and reference cell (volume 35 µl)
that were treated identically, however, the reference biosensor
surface did not contain immobilized protein. The collected
sample signals were corrected for the reference values and
analyzed further using the SPR instrument software. After each
binding experiment the surface was regenerated with a mixture
of 10 mM NaOH and 0.2% SDS for 1 min. Prior to the start of
a competition experiment a mixture of CTB was made with
a variable concentration of the synthetic inhibitors, and the
mixture was subsequently left to stand at room temperature for
a specific period of time. The equilibrated sample was then
introduced to the sensor-immobilized protein. Binding con-
stants (KD or Kchip) were obtained from equilibrium signals that
were fitted with a non-linear binding isotherm.19 Binding in
solution was measured using competition experiments and was
expressed as inhibitory concentration (IC50). The concentration
of CTB used in inhibition experiments was 3 µM.
3 (a) A. Bernardi, D. Arosio and S. Sonnino, Neurochem. Res., 2002,
27, 539–545 and references therein; (b) A. Bernardi, D. Arosio,
L. Manzoni, D. Monti, H. Posteri, D. Potenza, S. Mari and
J. Jimenez-Barbero, Org. Biomol. Chem., 2003, 1, 785–792; (c)
D. Arosio, S. Baretti, S. Cattaldo, D. Potenza and A. Bernardi,
Bioorg. Med. Chem. Lett., 2003, 13, 3831–3834; (d ) A. Bernardi,
A. Checchia, P. Brocca, S. Sonnino and F. Zuccotto, J. Am. Chem.
Soc., 1999, 121, 2032–2036; (e) A. Bernardi, L. Carrettoni,
A. Grosso Ciponte, D. Monti and S. Sonnino, Bioorg. Med. Chem.
Lett., 2000, 10, 2197–2200.
4 A. Bernardi, D. Potenza, A. M. Capelli, A. García-Herrero,
F. J. Cañada and J. Jiménez-Barbero, Chem. Eur. J., 2002, 8, 4597–
4612.
5 (a) M. Mammen, S.-K. Chio and G. M. Whitesides, Angew. Chem.,
Int. Ed., 1998, 37, 2754–2794; (b) L. L. Kiessling, L. E. Strong and
J. E. Gestwicki, Ann. Rep. Med. Chem., 2000, 35, 321–330; (c)
J. J. Lundquist and E. J. Toone, Chem. Rev., 2002, 102, 555–578;
(d ) T. K. Lindhorst, Top. Curr. Chem., 2002, 218, 201–235;
(e) R. Roy, Curr. Opin. Struct. Biol., 1996, 6, 692–702; ( f ) R. T. Lee
and Y. C. Lee, Glycoconjugate J., 2002, 17, 543–551; (g)
W. B. Turnbull and J. F. Stoddart, Rev. Mol. Biotechnol., 2002, 90,
231–255; (h) M. J. Cloninger, Curr. Opin. Chem. Biol., 2002, 6,
742–748; (i) R. Roy, Trends Glycosci. Glycotechnol., 2003, 15,
291–310.
6 J. Rao, J. Lahiri, R. M. Weis and G. M. Whitesides, J. Am. Chem.
Soc., 2000, 122, 2698–2710.
7 (a) J. J. Lundquist and E. J. Toone, Chem. Rev., 2002, 102, 555–578;
(b) S. D. Burke, Q. Zhao, M. C. Schuster and L. L. Kiessling, J. Am.
Chem. Soc., 2000, 122, 4518–4519.
ELISA
50 µl of a 0.002 µg µlϪ1 solution of GM1 in EtOH (0.1 µg of
GM1) was transferred to the wells and allowed to air dry
overnight. Unattached ganglioside was removed by washing the
wells twice with PBS (0.01 M phosphate buffer, 0.0027 M KCl,
0.137 M NaCl, pH 7.4). Additional binding sites on the plate
surface were blocked by incubating the wells with 150 µl of a
2% (w/v) bovine serum albumin (BSA)-PBS solution for 2.5 h at
rt and then washing them with 200 µl of PBS solution three
times.
8 R. T. Lee and Y. C. Lee, Glycoconjugate J., 2000, 17, 543–551.
9 P. I. Kitov, J. M. Sadowska, G. Mulvey, G. D. Arnstrong, H. Ling,
N. S. Pannu, R. J. Read and D. R. Bundle, Nature, 2000, 403,
669–672.
10 (a) E. Fan, Z. Zhang, W. E. Minke, Z. Hou, C. L. M. J. Verlinde and
W. G. J. Hol, J. Am. Chem. Soc., 2000, 122, 2663–2664; (b) Z. Zhang,
E. A. Merritt, M. Ahn, C. Roach, Z. Hou, C. L. M. J. Verlinde,
W. G. J. Hol and E. Fan, J. Am. Chem. Soc., 2002, 124, 12991–12998.
11 (a) J. M. Gargano, T. Ngo, J. Y. Kim, D. W. K. Acheson and
W. J. Lees, J. Am. Chem. Soc., 2001, 123, 12909–12910; (b) P. I. Kitov
and D. R. Bundle, J. Am. Chem. Soc., 2003, 125, 16271–16284.
12 (a) I. Vrasidas, N. J. de Mol, R. M. J. Liskamp and R. J. Pieters, Eur.
J. Org. Chem., 2001, 4685–4692; (b) S. André, R. J. Pieters,
I. Vrasidas, H. Kaltner, I. Kuwabara, F.-T. Liu, R. M. J. Liskamp
and H.-J. Gabius, ChemBioChem, 2001, 2, 822–830; (c) I. Vrasidas,
S. André, P. Valentini, C. Böck, M. Lensch, H. Kaltner R. M. J.
Liskamp, H.-J. Gabius and R. J. Pieters, Org. Biomol. Chem., 2003,
1, 803–810; (d ) R. Autar, A. S. Khan, M. Schad J. Hacker, R. M. J.
Liskamp and R. J. Pieters, ChemBioChem, 2003, 4, 1317–1325.
13 (a) R. Roy, Trends Glycosci. Glycotechnol., 2003, 15, 291–310; (b)
T. K. Lindhorst, Top. Curr. Chem., 2002, 218, 201–235.
14 A. J. Brouwer, S. J. E. Mulders and R. M. J. Liskamp, Eur. J. Org.
Chem., 2001, 1903–1915.
The test samples were dissolved in 50 µl of a 0.025 µg mlϪ1
solution of CT-HRP conjugate in 0.1% BSA-PBS and
incubated for 4 h at r. t.. Toxin and inhibitor solutions were
added in 50 µl volumes per well and incubated for 2 h at room
temperature. Unbound toxin was removed by washing four
times with 200 µl of PBS. Toxin bound to GM1 was then
revealed by the following sequence: (I) incubation with 50 µl of
freshly made o-phenylenediamine (OPD) solution (OPD
solution: 21.6 mg o-phenylenediamine in 50 ml of 0.05 M
citrate buffer at pH 5 and 40 µl of 30% hydrogen peroxide) for
15 min at room temperature; (II) quenching with 150 µl of 2 M
H2SO4 and (III) recording the data on a multiwell plate spec-
trophotometer. All experiments were carried out in triplicate.
15 A. Bernardi, D. Arosio, L. Manzoni, F. Micheli, A. Pasquarello and
P. Seneci, J. Org. Chem., 2001, 66, 6209–6216.
16 C. Santini, R. G. Ball and G. D. Berger, J. Org. Chem., 1994, 59,
2261–2266.
17 I. Vrasidas, PhD thesis, Utrecht University, 2003.
Acknowledgements
We thank Dr M. J. E. Fischer (Utrecht University) for help
with the SPR measurements, Dr S. André and Professor
H.-J. Gabius (Ludwig-Maximilians-Universität München) for
providing the ASF and Professor S. Sonnino (UniversitaЈ di
Milano) for providing the GM1-OS. This research has been
supported by Marie Curie Training Site Fellowships (to I. V.,
under contract HPMT-CT-2001–00293 and to D. A. under
contract QLK2-CT-2000–60047) of the European Union. The
Royal Netherlands Academy of Arts and Sciences (KNAW) is
gratefully acknowledged for a fellowship to R. J. P.
18 D. Gupta, H. Kaltner, X. Dong, H.-J. Gabius and C. F. Brewer,
Glycobiology, 1996, 6, 843–849.
19 N. J. de Mol, E. Plomp, M. J. E. Fischer and R. Ruijtenbeek, Anal.
Biochem., 2000, 279, 61–70.
20 M. Carpo, E. Nobile-Orario, V. Chigorno and S. Sonnino,
Glycoconjugate J., 1995, 12, 729–773.
21 J. A. Mertz, J. A McCann and W. D. Picking, Biochem. Biophys. Res.
Commun., 1996, 226, 140–144.
22 E. A. Merrit, S. Sarfaty, F. van den Akker, C. L’hoir, J. A. Martial
and W. G. J. Hol, Protein Sci., 1994, 3, 166–175.
23 P. I. Kitov, H. Shimizu, S. W. Homans and D. R. Bundle, J. Am.
Chem. Soc., 2003, 125, 3284–3294.
24 J. J. Lundquist, S. D. Debenham and E. J. Toone, J. Org. Chem.,
2000, 65, 8245–8250.
25 Fluorescence spectroscopy can be used for this purpose but
competing filter effects of the dendritic backbone limited its
reliability at higher concentrations.
Notes and references
1 E. Fan, E. A. Merritt, C. L. M. J. Verlinde and W. G. J. Hol, Curr.
Opin. Struct. Biol., 2000, 10, 680–686.
2 W. B. Turnbull, B. L. Precious and S. W. Homans, J. Am. Chem. Soc.,
2004, 126, 1047–1054.
26 E. A. Merritt, Z. Zhang, J. C. Pickens, M. Ahn, W. G. J. Hol and
E. Fan, J. Am. Chem. Soc., 2002, 124, 8818–8824.
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 2 1 1 3 – 2 1 2 4
2124