Synthesis and screening of a small glycomimetic library for inhibitory activity on medically relevant galactoside-specific lectins in assays of increasing biorelevance

Article abstract of DOI:10.1039/c0nj00277a

Synthetic introduction of aglyconic substitutions into carbohydrate ligands is an approach toward identifying potent inhibitors of medically relevant lectins. We tested a panel of 27 galactoside/lactoside derivatives harboring varying aglycone moieties to

Full text of DOI:10.1039/c0nj00277a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/njc | New Journal of Chemistry
Synthesis and screening of a small glycomimetic library for inhibitory
activity on medically relevant galactoside-specific lectins in assays of
increasing biorelevancew
a
b
c
c
´
Sabine Andre, Denis Giguere, Tarun K. Dam, Fred Brewer,
Hans-Joachim Gabius and Rene Roy*
a
b
´
Received (in Montpellier, France) 14th April 2010, Accepted 20th May 2010
DOI: 10.1039/c0nj00277a
Synthetic introduction of aglyconic substitutions into carbohydrate ligands is an approach toward
identifying potent inhibitors of medically relevant lectins. We tested a panel of 27 galactoside/
lactoside derivatives harboring varying aglycone moieties together with some O-3/O-30
functionality toward a biohazardous plant toxin and four human adhesion/growth-regulatory
galectins. Differential sensitivity profiles of lectin binding with cases showing activity increase
relative to galactose/lactose were revealed by systematic assessments using a solid-phase assay.
Quantitative differences between the homologous human proteins could even be detected. Binding
of substituted lactosides to galectins-1 and -3 was shown to be enthalpically driven. To determine
the potential of substituted glycosides to protect cells from harmful lectin association, binding
assays with human tumor cells were performed. Invariably, compounds were identified with
increased potency relative to the unsubstituted parent sugars. However, aglyconic substitutions
were shown to be able to convey cytotoxicity. This report directs further attention to examining
additional 20- and 30-substitutions of the galactose core and the potential of ligand presentation in
glycoclusters to enhance avidity and selectivity, continuing to use the herein applied strategic
combination of a convenient biochemical test system with bioassays.
report, we explore the impact of adding a spacered aglycone
Introduction
to galactose/lactose in O/S-glycosidic linkages on lectin
reactivity for a library of 27 synthetic compounds in biochemical
screening, flanked by experimental considerations on thermo-
dynamics of binding and increasing biorelevance of the assay
procedure for two classes of medically relevant lectins.
Glycans have an exceptional capacity for coding information,
hereby predestining oligosaccharides of cellular glycoconjugates
to serve as versatile biochemical signals.1 Acting as decoding
device, a large panel of protein folds has developed which
define the different families of lectins.2 The growing awareness
of the functional orchestration of glycan and lectin in human
diseases, e.g. in inflammation or tumor suppression,3 directs
increasing attention to lectins as targets for drug design.4 Two
main synthetic approaches are taken: diverse types of ligand
derivatives or mimetics and of modes of multimeric presentation,
e.g. in glycoclusters such as glycodendrimers,5 are being
examined to identify potent and lectin-specific candidates for
medical application. In addition to tailoring contacts in the
primary binding sites, aglyconic extensions of ligand derivatives
can prove to be effective to enhance their inhibitory potency,
as revealed for selectins and a 2-tetradecylhexadecyl anchor or
cholera toxin and non-spanning bivalent ligands.6 In this
As test cases, a biohazardous plant toxin, i.e. Viscum album
L. agglutinin (VAA), and four members of the family of
adhesion/growth-regulatory human galectins are chosen.7
The plant lectin has strong affinity for histo-blood group B
and H constituents and tolerates a2,6-sialylation, whereas this
modification precludes galectin binding, common to substituted
b-galactosides with protein-type-dependent affinity differences.8
Of particular note for galectins, target selection for distinct
glycoconjugates can be specific, as illustrated for galectins-1, -3
and -9 when inducing activated T-cell or T helper cell
apoptosis.9 Observations on a beneficial effect of introducing
a nitrophenyl group to galactose at the anomeric position for
VAA and galectin-1 and also evidence from spacered derivatives
using glycosyl triazoles formed the basis for the intended
screening.10 Since galectins constitute a family of homologous
and functionally non-redundant proteins divided into three
groups, it is essential to look at the reactivity profiles for
several proteins.7f,h Thus, members of each galectin category,
i.e. a homodimeric proto-type protein (galectin-1), the chimera-
type galectin-3 and two tandem-repeat-type proteins (galec-
tins-4 and -9), were included to infer indications for differential
sensitivities despite the close structural homology. Of note,
respective results can provide clues to define the structural
basis for the functional divergence among these proteins and
a Institute of Physiological Chemistry, Faculty of Veterinary Medicine,
Ludwig-Maximilians-University, Veterinarstr. 13, 80539 Munich,
¨
Germany
b PharmaQAM, Department of Chemistry,
´
´
`
´
Universite du Quebec a Montreal, P. O. Box 8888,
Succ. Centre-ville, Montreal (QC), Canada H3C 3P8.
E-mail: roy.rene@uqam.ca; Fax: +1 514-987-4054;
Tel: +1 514-987-3000 ext 2546
c Departments of Molecular Pharmacology,
Microbiology and Immunology, Albert Einstein College of Medicine,
Bronx, New York 10461, USA
w Electronic supplementary information (ESI) available: NMR spectra.
See DOI: 10.1039/c0nj00277a
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 2229–2240 | 2229

View Article Online
serve as input to computational docking. Although the
intricate network of lectin-ligand contacts in crystals is known
for the lectins, e.g. VAA and galectin-1,11 the significant
influence of ligand binding in solution on these two proteins’
gyration radius will impede to precisely predict surface
properties from modeling.12 Thus, the reported comparative
analyses can also be considered as an experimental surface
mapping of the binding-site vicinity.
galactosides/lactosides can harbor cytotoxicity. Thus, our
study strategically combines synthesis with biochemical and
cell biological screening of the products.
Results and discussion
Solid-phase inhibition assays
Technically, we performed a competitive solid-phase assay,
in which extent of lectin binding to a surface-presented
neoglycoprotein is monitored, for screening in a systematic
manner. In order to provide insights into thermodynamics of
binding of substituted lactosides, in relation to the disaccharide,13
we then proceeded to perform isothermal titration micro-
calorimetry for selected cases. The biorelevance of screening
is next increased by running inhibition assays with human
tumor cells, simulating the scenario in medical application
with the aim to protect cells from lectin binding. Finally, the so
far unresolved issue is addressed as to whether substituted
The test panel comprised 16 galactosides and 11 lactosides
with different aglycones in O/S-glycosidic linkages, in one case
harboring an additional 30-substitution in the galactose moiety
(Schemes 1–3). The synthetic and analytical data including
assignment of compounds to numbers are given in the
Experimental section. The size of the aglycone reached up
to two ring systems with an isoxazole extended by furan
(14) or phenylmethanone (27) or to a ring with an aliphatic
chain. The experimental setting to systematically determine
inhibitory potency was established in microtiter plate wells
by the labeled lectin in solution and a surface-presented
Scheme 1 Representative synthetic procedures for new compounds.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
2230 | New J. Chem., 2010, 34, 2229–2240

View Article Online
Scheme 2 Structures of the b-D-galactopyranosides used in this study.
Scheme 3 Structures of the b-D-lactopyranosides used in this study.
glycan as its ligand. The effect of the series of compounds on
the extent of binding was measured spectrophotometrically,
with cognate/non-cognate mono- and disaccharides as controls.
Due to the noted enhanced reactivity of the plant toxin and
galectin-1 toward p-nitrophenyl derivatives, a neoglycoprotein
presenting the p-isothiocyanatophenyl derivative of lactose
was selected as binding partner to spot especially active
compounds hence, considerably increasing the binding
interactions to inhibit. For each lectin, the binding was
saturable and inhibitable by the cognate sugar. Controls run
parallel with mannose or cellobiose excluded non-specific
effects.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 2229–2240 | 2231

View Article Online
The assay conditions are defined by two variables: ligand
density on the surface was deliberately kept constant in all
cases, and the concentration of each lectin was adjusted
to yield signal intensity in the linear range. To assess the
inhibitory profiles comparatively, the assays were run in the
presence of increasing concentrations of test compound (up to
eight different concentrations). The resulting effect on signal
intensity is illustrated in Fig. 1. As a measure for comparison,
we assessed the concentration of inhibitor at the 50%-level of
inhibition (IC50) from these curves. Relative inhibitory
potency with respect to galactose/lactose, routinely set to 1,
could then be plotted for each lectin. The detailed data for
compounds with relative values above a tenth/fifth of this
internal control for potency with the unsubstituted sugar are
given in Fig. 2 and 3.
The results for the plant toxin answered the question on the
impact of structural variations. In detail, they revealed a
differential sensitivity of inhibitory properties for substitutions
to the galactose core (Fig. 2A). 4-Nitrophenyl 1-thio-b-D-
galactopyranoside (1), initially tested as further control due
to the activity of the corresponding O-glycoside,10c proved
active, demonstrating efficacy of the non-hydrolyzable
S-glycoside (Fig. 2A). In negative terms, the presence of a
carboxyl group in terminal position of a spacered triazole in
compound 15 proved detrimental, underscoring its unfavorable
effect noted for a short aliphatic chain previously.10c The
two sulfones 7 and 8 were less reactive than unsubstituted
galactose. In contrast, a benzothiazolyl (3) and a furan-substituted
isoxazolyl (14) as substituents were notably active. These
substituents will likely occupy the position of the glucose
moiety of lactose in a favorable manner. To test the impact
of substitutions beyond the disaccharide, 11 substituted
lactosides were processed next. Invariably, the extensions to
the lactose core conveyed no remarkable increase of inhibitory
potency (Fig. 2B). These results are in accord with the
galactose core as major contact site and a calorimetrically
determined difference in DG of 0.4 kcal molꢁ1 between galactose
and lactose to the lectin site around tyrosine residue 249 in
VAA’s subdomain 2g.14 As consequence, these results turn
special attention to further substitutions at the galactose core,
which can in parallel consider the 20- and 30-positions and
combinations thereof. In fact, a1,2-fucosylation of galactose
Fig. 2 Relative potency of the panel of synthetic derivatives of
galactose (A) and lactose (B) for inhibition of binding of VAA to
surface-immobilized lactosylated neoglycoprotein. The absolute
IC50-values were determined by systematic inhibition assays in each
case, as shown in Fig. 1, and then set into relation to the data for
galactose (Gal) and lactose (Lac) used for normalization to 1
(IC50 (Gal) = 0.6 mM, IC50 (Lac) = 0.29 mM). The connection
between compound number (type of inhibitor) and the structure is
given in the Experimental section together with the experimental details
on which the upper limit of 15.4% for the standard deviation is based.
to establish the histo-blood group H-disaccharide and
a1,3/4-galactosylation had yielded increased avidity.8d Thus,
introducing a thio-linked substituent at the anomeric position
and exploring the potential of derivatization at the indicated
positions gives work on further enhancing avidity a clear
direction.
Turning next to galectins, here galactose is a much weaker
inhibitor than lactose.10a,b The natural extension, bound as
low-energy syn conformer,15 is thus very suited for affinity
generation. Fittingly, the substituted galactosides were mostly
inactive or only rather weak inhibitors for the galectins. At
best, the benzothiazolyl derivative 3 noted above reached an
activity level comparable to lactose for galectins-3, -4 and -9.
The reactivity profiles of the lactosides presented in Fig. 3
revealed differences in reactivity between the test compounds,
with cases of relative enhancement in inhibition, and
characteristic features for the intergalectin comparison. The
attained selectivity levels yet appeared as quantitative with
tangible cross-reactivity. Relative inhibitory potency was
lowest for galectin-1, increased activity was seen for the other
three tested proteins, with cross-reactivity of the library to
different galectins reflecting the structural homology (Fig. 3).
The tested 30-substitution to the galactose moiety with a
substituted isoxazole (compound 26) yielded rather similar
effects on the chimera- and tandem-repeat-type galectins, i.e.
Fig. 1 Inhibition of the binding of biotinylated VAA (left) and
human galectin-3 (right) to surface-immobilized lactosylated neoglyco-
protein by ligand derivatives. Galactose (’) or lactose (m) were used
as control. Data on the following synthetic test substances are
presented: compound 1 (&) and compound 14 (J) in the case of
VAA, and compound 23 (D) and compound 26 (B) in the case of
galectin-3.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
2232 | New J. Chem., 2010, 34, 2229–2240

View Article Online
(compounds 21 and 22, Fig. 3). The b-naphthyl sulfone
derivative (21) proved most potent on galectin-9 relative to
lactose (Fig. 3D), with a graded activity also to galectin-4 and
then galectin-3 (Fig. 3B and C). Its IC50-value of about
300 mM compares favorably to 4-benzylaminocarbonyl-1H-
[1,2,3]-triazol-1-yl mannoside at KD = 540 mM, its galactoside
(KD = 670 mM) and N-acetyllactosamine (KD = 500 mM).10d
Evidently, despite global sequence homology structural
differences in the vicinity of the contact site for the glucose
moiety of lactose are inferable by the synthetic derivatives,
giving direction to iterative refinements as noted above for
VAA. In addition, modifying 20- and 30-positions at the
galactose unit, hereby yielding closer glycomimetics of high-
affinity histo-blood group ABH-oligosaccharide epitopes, is a
promising means to combine avidity enhancements from three
different subsites for generating optimal inhibitory potency
and selectivity. This perspective warrants to consider further
methodological transformations.
Technically, this solid-phase system as screening approach
provides comparative data comprehensively in a simple setting
which simulates lectin binding from solution to biological
targets. As additional advantage, it requires rather low
amounts of sugar analogs. An affinity determination by
calorimetry would be technically more demanding and will
require considerably more time and substance. This type of
analysis will yet provide quantitative data on affinity and the
major thermodynamic driving forces to ligand accommodation.
Whereas it is known that this process for lactose and
N-acetyllactosamine is enthalpically favored,13 the issue is open
for derivatives thio/sulfone-linked aglycones. Thus, we
performed a series of titrations with galectin-3 and also
galectin-1, a resulting binding profile is presented in Fig. 4. Since
galectin-3 will only form aggregates in the presence of multivalent
ligands,16 the natural form can be used for titrations.
The experiments enabled us to calculate the entropic
contribution and the binding affinity. The n-value of 1,
representing one active lectin site per subunit, ensured full
activity of the lectins. Lectin/ligand association was invariably
enthalpically driven (Table 1). As compiled in Table 1,
binding affinity of the glycomimetics showed differences
of up to 1.8 kcal molꢁ1, in this case with decrease in
enthalpic contribution. These data were in line with the
observed inhibitory activity, with an approximate correlation
between data sets of IC50-values from screening and
affinity from calorimetry. Having herewith also supported
applicability of the biochemical assay for screening, the matter
of biorelevance remains a concern. It is obvious that any
perspective for medical application will rest on the com-
pounds’ potency to block binding of lectin to physiologic
ligands on cell surfaces and their lack of cytotoxicity. To
address these issues we performed cell assays. The surface
of cells presents an array of galactosides in contrast to
the screening with the lactose derivative conjugated to a
carrier protein. Moreover, the common glycan branching
facilitates functional multivalency of glycoproteins with an
inherent gradient of decreasing binding constants.13 Hence,
the IC50 data obtained with this cell assay is more closely
representative of the in vivo situation, as opposed to other
physical measurements.
Fig. 3 Relative potency of the panel of synthetic derivatives of lactose
for inhibition of human galectins, i.e. homodimeric proto-type galectin-1
(A, IC50 (Lac) = 0.7 mM), chimera-type galectin-3 (B, IC50 (Lac) =
2.25 mM) and the tandem-repeat-type galectins-4 (C; IC50 (Lac) =
3.88 mM) and -9 (D, IC50 (Lac) = 4 mM), to surface-immobilized
lactosylated neoglycoprotein. The absolute IC50-values were determined
by systematic inhibition assays in each case, as shown exemplarily for
galectin-3 in Fig. 1, and then set into relation to the data for lactose
(Lac) used for normalization to 1. The connection between compound
number (type of inhibitor) and the structure is given in the
Experimental section together with the experimental details on which
the upper limit of 15.4% for the standard deviation is based.
galectins-3, -4, and -9, underscoring the challenge to refine
target specificity. The combination of a sulfone with a
distinct aromatic extension indicated potential toward this end
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 2229–2240 | 2233

View Article Online
Testing was performed in the linear range for each lectin/
cell type pair with 5 ꢂ 104 cells per assay, using different
concentrations of glycomimetics in each case. Monitoring
was always run in parallel flanked by internal controls,
e.g. to determine lectin- and carbohydrate-independent
binding. The results with inhibitory compounds are documented
in semilogarithmic representation, along with the control
standard, to illustrate the level of attainable inhibition for
the two quantitative parameters, i.e. percentage of positive
cells and mean fluorescence intensity (Fig. 5). The numbers are
given in each panel. Herein, the best case is illustrated,
together with the sugar control and other representative
results (Fig. 5).
The presented results for VAA revealed the activity of
galactose and the superiority of the thio-linked compound 3
in blocking binding to colon cancer cells (Fig. 5A and B).
Compound 14 was equally effective (not shown), and the lack
of activity for compound 15 ascertained accordance to solid-
phase data (Fig. 5B). The activity profile was not notably
different in this case when a line of lymphoma cells was tested,
and substituted lactosides did not yield an activity increase in
accord with the screening data (Fig. 5C and D). In total, at this
stage, the galactoside family is clearly superior to the lactosides
with unmodified lactose being better within the latter in both
assay settings. This situation is different for the galectins.
Compounds 24 and 21 were more active than lactose on
galectins-1 and -3 or galectins-4 and -9, respectively (Fig. 5,
third and fourth panels). Compounds 23 and 26 were rather
similar in potency on galectin-3 (compound 23 given in
Fig. 5E–H) so that no obvious advantage is achieved by this
30-substitution, also considering its cross-reactivity to galectins-4
and -9 (Fig. 3C and D). Similarly reflecting the biochemical
data, compounds 21, 23 and 26 share a comparable activity
level (no. 21 given in Fig. 5G), clearly above lactose. The
extent of inhibitory potency and selectivity was enhanced for
galectin-9, compound 21 at 0.25 mM significantly more active
than compound 23 at 0.5 mM and lactose at 2 mM (Fig. 5H).
The illustration of scans for weak inhibitors (compound
17/galectin-3; compound 15/galectin-4) attests an inherent
specificity control, solidifying the fair degree of comparability.
Since the nature of a galectin’s binding partner and its spatial
presentation can differ depending on the cell type, firm
conclusions on inhibitor potency in cell assays should only
be based on extensive testing of selected candidates with
various cell types.
Fig. 4 Profile of isothermic titration calorimetry using human galectin-3
and increasing concentrations of compound 23 at 27 1C added in 4 ml
injections (top) and the integrated curve showing experimental points
(’) and the best fit (---) (bottom).
Table 1 Thermodynamic parameters of binding of synthetic com-
pounds to human galectins-1 and -3 at 27 1C
Ka/
Compounds Mꢁ1 ꢂ 10ꢁ3 kcal molꢁ1 kcal molꢁ1 kcal mol
Galectin-3
ꢁDG/
ꢁDH/
ꢁTDS/
n
ꢁ
17
21
23
24
4.5
45.0
65.0
3.6
5.0
6.4
6.7
4.9
10.1
11.8
12.1
9.7
5.1
5.4
5.4
4.8
1.05
1.02
1.01
1.04
Galectin-1
17
24
1.3
2.0
4.3
4.5
10.7
9.8
6.4
5.3
1.03
1.03
At this stage, it is appropriate to test the tacitly assumed
lack of cytotoxicity of substituted galactosides/lactosides.
Examined as control, no tested sugar (galactose, lactose,
cellobiose) affected proliferation of the SW480 colon
adenocarcinoma cells used in cell-binding assay up to a
concentration of 10 mM. As models to examine cytotoxic
activity we selected the nitrophenyl thiogalactoside (1), a
sulfone (12) and an aromatic thiolactoside (17). All three
substances negatively affected proliferation over the period
of 48 h. Presence of compound 1 was a weak effector reducing
cell growth by 20% at 10 mM, whereas the other two
compounds had a strong impact already measurable at
2 mM with growth reduction by 10–15%, yielding an about
40% inhibition at 5 mM.
Errors in Ka range from 1–6%; errors in DG are less than 1%; errors in
DH are 1% to 5%; errors in TDS are 1% to 7%; errors in n are less
than 2%.
Cell-binding inhibition and cytotoxicity assays
Human tumor cells in vitro were used as model targets of the
biotinylated lectins. Concentration dependence was routinely
assessed first to determine linear range and plateau levels of
signal generation. Since galectins are known to interact with
peptide motifs and also protein-bound lipids besides carbo-
hydrates, sugar-inhibitable binding was rigorously ascertained.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
2234 | New J. Chem., 2010, 34, 2229–2240

View Article Online
Fig. 5 Semilogarithmic representation of fluorescent surface staining of human tumor cells by labeled lectins. The control value representing
marker-independent staining by incubation with the second-step reagent is given as shaded area, the 100%-value obtained by lectin binding in the
absence of any test compound is shown as black line. Quantitative data on percentage of positive cells and mean fluorescence intensity are given in
each panel (numbers of control printed in bold), and lines connect these data with the respective peak. The following pairs of biotinylated
lectin/human tumor cell line were processed for the given set of compound/concentration: (A) 2 mg mlꢁ1 VAA and the colon adenocarcinoma line
SW480 with galactose concentrations of 1 mM, 2 mM, 4 mM and 8 mM, with quantitative data given at 2 mM galactose and in the absence of
inhibitor; (B) respective results in the presence of the synthetic galactose derivatives 3, 11 and 15 (listed from top to bottom) at the fixed
concentration of 2 mM; (C) respective results in the presence of 1 mM lactose and the synthetic lactose derivatives 18 and 17; (D) 0.2 mg mlꢁ1 VAA
and the B-lymphoblastoid cell line Croco II tested without marker, in the presence of compound 3, compound 11, galactose and compound 8 at the
concentration of 2 mM as well as in the absence of test compound; (E) 10 mg mlꢁ1 galectin-1 and the tumor suppressor p16INK4a-expressing
pancreatic carcinoma line Capan-1 tested without marker, in the presence of compound 24, lactose and galactose at the concentration of 2 mM as
well as in the absence of test compound; (F) 10 mg mlꢁ1 galectin-3 and the colon adenocarcinoma line SW480 tested without marker, in the
presence of compound 23, compound 24, lactose and compound 17 at the concentration of 0.5 mM as well as in the absence of test compound;
(G) 10 mg mlꢁ1 galectin-4 and the tumor suppressor p16INK4a-expressing pancreatic carcinoma line Capan-1 tested without marker, with 0.4 mM
compound 21, 0.4 mM lactose, 2 mM compound 15 and 0.4 mM galactose as well as in the absence of test compound; (H) 5 mg mlꢁ1 galectin-9 and
the given pancreatic tumor line tested without a marker, in the presence of compound 21 at 0.25 mM, compound 23 at 0.5 mM, lactose at 2 mM
and galactose at 4 mM as well as in the absence of test compound.
quantity with a cell-based assay to determine bioactivity and
Conclusions
extent of cross-reactivity is helpful. Besides the further
This report explores the potential of a panel of 27 synthetic
iterative synthetic refinements the design of glycoclusters can
galactosides/lactosides with O/S-linked extensions at the
help reduce cross-reactivity. The presented data on cyto-
anomeric position to block glycan binding of two classes of
toxicity advise to run respective controls at an early stage of
medically relevant lectins. Solid-phase assays with the plant
such efforts.
toxin revealed activity enhancement especially for substituted
galactosides, directing further work to 20- and 30-modifications
Experimental
at the galactose core of a lactoside. Minor enhancements can
arise from further extensions at the anomeric linkage. In
contrast, these are sizeable when testing adhesion/growth-
regulatory galectins. b-Naphthyl sulfone established remarkable
potency in the solid-phase system for the two tandem-repeat-type
galectins, maintained in cell-binding assays especially for
galectin-9. The observed cross-reactivity among the four
homologous proteins is, however, still reason for concern, as
also previously seen for (glyco)peptides from one-bead-one-
compound libraries.17 Its level will need to be reduced by
co-substitutions at the galactose core, e.g. mimicking
ABH-epitopes or the pentasaccharide chain of ganglioside
GM1.8f,18 Toward this end, our combination of a simple and
sensitive inhibition assay with low requirements for sample
General
All reactions involving water-sensitive chemicals were carried
out in flamedried glassware with magnetic stirring under a
nitrogen atmosphere. Anhydrous DCM was distilled from
CaH2 and anhydrous THF were distilled from Na/K prior
to use. All non-aqueous reactions were carried out under
anhydrous conditions within a nitrogen atmosphere in distilled
solvents. All other solvents and reagents were used as received.
TLC was performed on aluminium plates (silica gel 60 F254)
with detection by UV or by coloration with ammonium
molybdate in acid solution. Column chromatography was
performed on silica gel (230–400 mesh) with the indicated
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 2229–2240 | 2235

View Article Online
1
eluent. H and 13C NMR spectra were recorded at 300 (75)
b-D-galactopyranoside (11),25 b-D-galactopyranosyl 2-benzo-
thiazolyl sulfone 2-benzothiazolyl 1-sulfonyl-b-D-galacto-
MHz with a Variant apparatus. Chemical shifts (ppm) are
reported relative to CHCl3 (7.27) or D2O or CD3OD as
internal standard. Optical rotations were measured on a
Polarimeter JASCO P-1000 and melting point on a Fisher-
Johns Melting Point Apparatus. ESI-MS analyses were carried
out on a MICROMASS Quattro LC. High-resolution mass
spectrometry (HRMS) was carried out on a LC-MSTOF
(liquid chromatography mass spectrometry time of flight)
model 6210 from Agilent Technologies.
pyranoside
(12),26
3-(4-methoxyphenyl)-5-(1-deoxy-b-D-
galactopyranosyl)-isoxazole (13),27 3-b-D-galactopyranosyl-
2,5-dimethoxybenzene (16),28 p-nitrophenyl 1-thio-b-D-lactoside
(20)19a b-naphthyl 1-sulfonyl-b-D-lactoside (21),27 p-nitrophenyl
b-D-lactoside (24),21 o-nitrophenyl b-D-lactoside (25),29 and
3-(2-phenylmethanone)-5-(1-deoxy-b-D-lactopyranosyl)-isoxazole
(27),30 were obtained as previously described. The following
known compounds gave the listed HRMS data: 20:
ESI+-HRMS: [M
+
Na]+ calcd for C18H25NO12SNa:
502.0990; found: 502.0993, D 0.66 ppm; 21: ESI+-HRMS:
[M + NH4]+ calcd for C22H32NO12S: 534.1640; found:
534.1646, D ꢁ2.93 ppm; 24: ESI+-HRMS: [M + Na]+ calcd
for C18H25NO13: 486.1218; found: 486.1221, D 0.63 ppm.
b-Naphthyl 1-sulfonyl-b-D-galactopyranoside (7) was
obtained from peracetylated 7b using the above general
Syntheses
Typical procedure for PTC glycosylation19. To a solution of
glycosyl halides (1 equiv.) and tetrabutyl ammonium hydrogen
sulfate (TBAHS) (1 equiv.) in ethyl acetate (1.0 ml per 100 mg
of sugar) were added the nucleophile (2–3 equiv.) and 1 M
sodium carbonate (1.0 ml per 100 mg of sugar). The reaction
mixture was vigorously stirred at room temperature until the
starting material was completely consumed (1–3 h) as judged
by TLC monitoring using a mixture of ethyl acetate and
hexane as eluent. The solution was then diluted with ethyl
acetate and the organic phase was separated. The organic
solution was washed with saturated sodium bicarbonate,
water and brine, then dried over anhydrous sodium sulfate,
filtered and concentrated. The crude residues were purified by
silica gel column chromatography using a mixture of ethyl
acetate and hexane as eluent. Pure glycosyl derivatives were
obtained after silica gel column chromatography and usually
recrystallised from ethanol.
Zemplen de-O-acetylation procedure as a white powder
´
(89%) and had mp 175–176 1C; IR: 3192 cmꢁ1 (s, OH),
1104 cmꢁ1 (s, OQSQO); 1H NMR (CD3OD) d (ppm):
8.47 (1H, s, HAR), 7.99–7.82 (4H, m, HAR), 7.63–7.52
(2H, m, HAR), 4.33 (1H, d, J = 9.3 Hz, H-1), 3.82
(1H, dd, J = 9.3 Hz, H-2), 3.71 (dd, J = 3.3 Hz, H-4),
3.51–3.23 (4H, m, H-3, H-5, H-6a, H-6b); 13C NMR
(CD3OD) d (ppm): 131.4, 129.4, 129.3, 128.6, 127.8, 127.4,
124.2 (10CAR), 92.7 (C-1), 80.1 (C-5), 74.5 (C-3), 68.6
(C-4), 67.1 (C-6), 61.0 (C-2); ESI-MS calcd for C16H18O7S +
(Na+): 377.3; found: 377.1; ESI+-HRMS: [M + Na]+
calcd for C16H18O7SNa: 377.0665; found: 377.0667,
D ꢁ0.69 ppm.
´
Typical procedure for de-O-acetylation (Zemplen conditions).
b-Naphthyl 2,3,4,6-tetra-O-acetyl-1-sulfonyl-b-D-galacto-
pyranoside (7b) was obtained from known 5b22 following the
above general oxidation procedure as a white powder (91%)
and had mp 145ꢁ146 1C; [a]D25: ꢁ18.7 (c = 1, CHCl3); IR:
1751 cmꢁ1 (s, CQO), 1229 cmꢁ1 (s, OQSQO); 1H NMR
(CDCl3) d (ppm): 8.56 (1H, s, HAR), 8.10–7.91 (4H, m, HAR),
7.73–7.63 (2H, m, HAR), 5.51 (1H, dd, J = 9.89 Hz, H-2), 5.27
(1H, dd, J = 3.02 Hz, H-4), 5.04 (1H, dd, J = 9.89, 3.29 Hz,
H-3), 4.58 (1H, d, J = 9.62 Hz, H-1), 4.13–4.07 (1H, m, H-6a),
3.95–3.86 (2H, m, H-5, H-6b), 2.17, 1.96, 1.85, 1.63 (12H, 4s,
The acetyl protected glycoside (0.1 mmol) was dissolved into
methanol (2 ml), to which was added a catalytic amount of
sodium methoxide (until pH = 8). The solution was stirred at
room temperature until disappearance of the starting material.
After neutralization of sodium methoxide with Amberlite
IR-120 (H+) resin, the solution was filtered and removal
of the methanol under reduced pressure afforded the fully
deprotected glycoside.
CH3CO); 13C NMR (CDCl3) d (ppm): 170.2, 169.9, 169.7,
Typical procedure for sulfone synthesis. The 1-thio-b-D-
glycoside (1.0 mmol) was dissolved in dry dichloromethane at
0 1C under an atmosphere of nitrogen. mCPBA (3.5 mmol) was
added by small portions and the reaction was monitored
by TLC. When the reaction was finished the mixture was
allowed to warm to room temperature and washed with
saturated NaHCO3, brine and water. The organic solution was
dried over Na2SO4, filtered and concentrated under reduce
pressure. The crude residues were purified by silica gel column
chromatography using a mixture of ethyl acetate and hexane as
eluent.
ꢁ ꢁꢁ
169.4 (4CH3CO), 135.6, 132.9, 131.8, 131.6, 129.7, 128.5,
ꢁ ꢁ
127.7, 127.5, 124.9 (10CAR), 89.2 (C-1), 74.7 (C-5), 71.5
(C-3), 66.4 (C-4), 63.9 (C-2), 60.9 (C-6), 20.8, 20.5, 20.4, 19.9
(4CH3CO); ESI-MS calcd for C24H26O11S + (Na+): 545.5;
ꢁ ꢁꢁ
found: 545.1; ESI+-HRMS: [M
+
NH4]+ calcd for
C24H30NO11S: 540.1534; found: 540.1536, D 0.43 ppm.
4-Methoxyphenyl 1-sulfonyl-b-D-galactopyranoside (8) was
obtained from 8b under Zemplen conditions as a white powder
´
(91%) and had mp 161-163 1C; [a]D25: +1.0 (c = 1, MeOH);
1
IR: 3377 cmꢁ1 (s, O-H), 1268 cmꢁ1 (s, OQSQO); H NMR
(CD3OD) d (ppm): 7.83 (2H, d, J = 8.7 Hz, H-ortho), 7.05
(2H, d, J = 8.7 Hz, H-meta), 4.25 (1H, d, J = 9.6 Hz, H-1),
3.85 (3H, s, OMe), 3.81–3.78 (2H, m, H-2, H-4), 3.60–3.58
(2H, m, H-3, H-6a), 3.52–3.45 (2H, m, H-5, H-6b); 13C NMR
(CD3OD) d (ppm): 165.8 (C-para), 133.1 (C-ortho), 129.3
(C-ipso), 115.2 (C-meta), 93.8 (C-1), 81.3 (C-5), 75.7 (C-3),
69.8 (C-4), 68.3 (C-6), 62.3 (C-2), 56.3 (OMe); ESI-MS calcd
for C13H18O8S + (Na+): 357.1; found: 357.3; ESI+-HRMS:
Compounds and characterisation
4-Nitrophenyl 1-thio-b-D-galactopyranoside (1),19c 4-bromo-
phenyl 1-thio-b-D-galactopyranoside (2),20 2-benzothiazolyl
1-thio-b-D-galactopyranoside (3),21 phenyl 1-thio-b-D-galacto-
pyranoside (4),22 b-naphthyl 1-thio-b-D-galactopyranoside
(5),22 4-methoxyphenyl 1-thio-b-D-galactopyranoside (6),23
phenyl 1-thio-b-D-galactopyranoside S-oxide (10),24 o-nitrophenyl
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
2236 | New J. Chem., 2010, 34, 2229–2240

View Article Online
[M + NH4]+ calcd for C13H22NO8S: 352.1061; found:
conditions as a clear oil and had 1H NMR (CD3OD)
d (ppm): 7.65 (1H, d, J = 1.6 Hz, Hpyran), 6.95 (1H, d,
J = 3.6 Hz, Hpyran), 6.78 (1H, s, Hisoxazole), 6.57–6.56 (1H,
352.1062, D 0.54 ppm.
4-Methoxyphenyl
2,3,4,6-tetra-O-acetyl-1-sulfonyl-b-D-
galactopyranoside (8b) was obtained as above from
peracetylated 6 as a white powder (89%) and had mp
115–1116 1C; [a]D25: ꢁ58.2 (c = 1, CHCl3); IR: 1750 cmꢁ1
(s, CQO), 1273 cmꢁ1 (s, OQSQO); 1H NMR (CDCl3)
d (ppm): 7.88 (2H, d, J = 8.5 Hz, H-ortho), 7.05 (2H, d,
J = 8.5 Hz, H-meta), 5.42 (1H, dd, J = 9.8 Hz, H-2), 5.32
(1H, dd, J = 3.29 Hz, H-4), 5.04 (1H, dd, J = 3.0, 9.8 Hz,
H-3), 4.47 (1H, d, J = 9.6 Hz, H-1), 4.11 (1H, dd, J = 9.0,
13.1 Hz, H-6a), 3.97–3.91 (2H, m, H-5, H-6b), 3.90 (3H, s,
m, Hpyran), 4.81 (2H, s, OCH2), 4.33 (1H, d, J = 7.7 Hz, H-1),
ꢁ ꢁꢁ
3.81–3.70 (2H, m, H-2, H-4), 3.54–3.32 (4H, m, H-3, H-5,
H-6a, H-6b); 13C NMR (CD3OD) d (ppm): 171.7, 164.5, 154.7,
144.5, 111.6, 110.7, 103.2 (3Cisoxazole, 4Cpyrane), 100.8 (C-1),
75.7, 73.7, 71.2, 69.1, 61.3, 60.9 (C-2 to C-6, OCH2); ESI-MS
calcd for C14H17NO8
+
(Na+): 350.3; found: 350.0;
ESI+-MS: [M + Na]+ calcd for C14H17NO8Na: 350.08464;
found: 350.0.
3-(2-Furan)-5-(1-deoxy-2,3,4,6-tetra-O-acetyl-b-D-galacto-
pyranosyl)-isoxazole (14b). To a mixture a acetylenic galacto-
side30 (0.89 mmol, 346 mg), NCS (1.43 mmol, 191 mg) and
pyridine (1.34 mmol, 108 mg) in CHCl3 (5 ml) was added
dropwise 2-furan oxime (0.89 mmol, 99 mg) at 0 1C, then the
reaction was stirred over night at room temperature. After
this time the organic solution was washed with saturated
NaHCO3, brine and water, dried over Na2SO4, filtered and
concentrated under reduce pressure. Flash chromatography
OMe), 2.15, 2.01, 1.98, 1.92 (12H, 4s, CH3CO); 13C NMR
ꢁ ꢁꢁ
(CDCl3) d (ppm): 170.2, 169.9, 169.7, 169.5 (4CH3CO), 164.5
(C-para), 132.9 (C-ortho), 125.9 (C-ipso), 113.9 (C-meta), 89.1
ꢁ ꢁ
(C-1), 74.6 (C-5), 71.5 (C-3), 66.5 (C-4), 64.0 (C-2), 60.9 (C-6),
55.7 (OMe), 20.8, 20.6, 20.5, 20.3 (4CH3CO); ESI-MS calcd
ꢁ ꢁꢁ
for C21H26O12S + (Na+): 525.1; found: 525.3; ESI+-HRMS:
[M + NH4]+ calcd for C21H30NO12S: 520.1483; found:
520.1482, D ꢁ0.15 ppm.
2-(4-(b-D-Galactopyranosylthiomethyl)-1H-1,2,3-triazol-1-yl)-
using ethyl acetate: hexane (1 : 1) afforded pure 14b
25
acetate (9) was obtained as a yellow oil (81%) and had [a]D
:
(0.65 mmol, 324 mg) in 73% yield as a yellow oil and had
[a]D25: ꢁ1.1 (c = 0.4, CHCl3); 1H NMR (CDCl3) d (ppm):
7.55 (1H, d, J = 1.2 Hz, Hpyran), 6.89 (1H, d, J = 2.3, Hpyran),
6.52–6.49 (1H, m, Hpyran), 6.46 (1H, s, Hisoxazole), 5.41 (1H, dd,
J = 3.0 Hz, H-4), 5.24 (1H, dd, J = 9.8 Hz, H-2), 5.02 (1H,
dd, J = 9.8, 3.3 Hz, H-3), 4.56 (1H, d, J = 9.6 Hz, H-1),
ꢁ133.0 (c = 1, H2O); 1H NMR (CD3OD) d (ppm): 7.89
(1H, s, Htriazole), 5.18 (1H, s, CH2CO2CH3), 4.15 (1H, d,
ꢁ ꢁꢁ
J = 9.6 Hz, H-1), 3.91 (2H, dd, J = 14.5, 61.8 Hz, SCH2),
ꢁ ꢁꢁ
3.73 (1H, d, J = 3.0 Hz, H-4), 3.67 (3H, s, CH2CO2CH3),
3.65–3.28 (5H, m, H-2, H-3, H-5, H-6a, H-6b); 13C NMR
ꢁ ꢁꢁ
(D2O) d (ppm): 173.1 (COCH3), 144.9 (Ctriazole), 125.4 (CHtriazole),
85.0 (C-1), 79.0 (C-5), 73.9 (C-3), 69.5 (C-4), 68.8 (C-2), 61.1
4.18–4.08 (3H, m, H-6a, H-6b, OCH2), 3.97, 3.90 (1H, m,
H-5), 2.17, 2.04, 2.02, 1.96 (4CH3CO); 13C NMR (CDCl3) d
(ppm): 170.3, 170.1, 169.8, 169.6, 169.2, 154.8, 143.8, 111.7,
ꢁ ꢁ
ꢁ ꢁꢁ
ꢁ ꢁꢁ
(C-6, CH2CO2Me), 53.0 (COCH3), 23.6 (SCH2).
ꢁ ꢁꢁ
ꢁ ꢁꢁ
ꢁ ꢁꢁ
tert-Butyl 2-(4-(b-D-galactopyranosylthiomethyl)-1H-1,2,3-
triazol-1-yl)acetate (9b). To a solution of prop-2-ynyl 2,3,4,6-
tetra-O-acetyl-1-thio-b-D-galactopyranoside23 (1.0 mmol) in
THF (10 ml) was added CuI (0.1 mmol), DIPEA (2.0 mmol)
and tert-butyl 2-azidoacetate (3 mmol). The reaction mixture
was stirred at room temperature and stopped after 3 hours as
judged by TLC. After this time, the mixture was evaporated,
diluted with ethyl acetate, filtered through a pad of Celite and
washed with a solution of HCl (10%). The organic solution
was dried over Na2SO4, filtered and concentrated under
reduce pressure. The crude residue was purified by silica gel
column chromatography using a mixture of ethyl acetate and
hexane (1 : 1) as eluent. Compound 9b was obtained as a
yellow oil (83%) and had [a]D25: ꢁ66.5 (c = 1, CHCl3); IR:
1757 cmꢁ1 (s, CQO); 1H NMR (CDCl3) d (ppm): 7.63 (1H, s,
110.2 (4CH3CO, 3Cisoxazole, 4Cpyran); 100.4 (C-1), 82.7, 74.6,
ꢁ ꢁ
71.5, 67.1, 66.9, 61.4 (C-2 to C-6, OCH2), 23.9, 20.6, 20.5, 20.4
(4CH3CO); ESI-MS calcd for C22H25NO12 + (H+): 496.4;
ꢁ
ꢁ
found: 496.0; ESI+-MS: [M + Na]+ calcd for C22H26NO12:
496.14495; found: 496.0.
2-(4-(b-D-Galactopyranosylthiomethyl)-1H-1,2,3-triazol-1-yl)-
acetic acid (15). To a solution of 9 (1.0 mmol) in wet MeOH
(5ml) was added LiOH (30 mg) and the reaction was stirred
until disappearance of the starting materials as seen by TLC.
After this time, Amberlite resin (H+) was added until neutral
pH and the mixture was filtered and concentrated under
reduce pressure to give 15 as colorless oil in quantitative yield.
1H NMR (D2O) d (ppm): 7.40 (1H, s, Htriazole), 7.73–4.47 (m,
5H, H-1, CH2COOH, SCH2), 3.72–2.88 (m, 6H, H-2, H-3,
ꢁ ꢁꢁ
ꢁ ꢁꢁ
H-4, H-5, H-6a, H-6b); 13C NMR (D2O) d (ppm): 173.3
H
triazole), 5.36 (1H, dd, J = 3.3 Hz, H-4), 5.18 (1H, dd,
(COOH), 144.8 (Ctriazole), 125.2 (CHtriazole), 86.6 (C-1), 80.1
ꢁ ꢁ
J = 9.9 Hz, H-2), 5.12 (2H, s, CH2CO2t-Bu), 4.97 (1H, dd,
(C-5), 76.2 (C-3), 70.4 (C-4), 69.6 (C-2), 61.2 (C-6,
CH2COOH), 23.2 (SCH2); ESI+-MS: [M + Na]+ calcd for
ꢁ ꢁꢁ
J = 3.3, 9.9 Hz, H-3), 4.52 (1H, d, J = 9.9 Hz, H-1), 4.14–3.98
ꢁ ꢁꢁ
ꢁ ꢁꢁ
(2H, m, H-6a, H-6b, SCH2), 3.92–3.84 (1H, m, H-5), 2.09,
C11H17N3O7SNa: 358.06794; found: 358.07.
ꢁ ꢁꢁ
1.97, 1.96, 1.90 (12H, 4s, CH3CO), 1.37 (9H, s,
CH2CO2C(CH3)3); 13C NMR (CDCl3) d (ppm): 169.9, 169.8,
169.5, 169.2 (4CH3CO), 165.0 (CH2CO2t-Bu), 144.4 (Ctriazole),
p-Bromophenyl 1-thio-b-D-lactoside (17) was obtained from
acetobromolactose and p-bromothiophenol using the general
PTC conditions described above (88%) followed by de-O-
ꢁ ꢁꢁ
ꢁ ꢁꢁ
ꢁ ꢁ
ꢁ
123.5 (CHtriazole), 83.3 (C-1), 82.7 (C(CH3)3), 74.0 (C-5), 71.4
acetylation (quant.) as a white powder and had mp
207.0–208.0 1C; [a]D25: ꢁ38.1 (c = 1, MeOH); IR: 3337
ꢁ
(C-3), 67.0 (C-4), 66.8 (C-2), 61.0 (C-6), 51.1 (CH2CO2t-Bu),
ꢁ ꢁꢁ
29.3 (SCH2), 27.6 (C(CH3)3), 20.6, 20.4, 20.3 (4CH3CO);
ESI+-MS: [M
cmꢁ1 (s, O-H); 1H NMR (CD3OD) d (ppm): 7.36 (4H, m,
ꢁ ꢁꢁ
ꢁ ꢁꢁ
ꢁ ꢁꢁ
+
Na]+ calcd for C23H33N3O11SNa:
H
AR), 4.49 (1H, d, J = 9.8 Hz, H-10), 4.25 (1H, d, J = 7.5 Hz,
582.17280; found: 582.17.
H-1), 3.78–3.59 (6H, m, H-2, H-20, H-3, H-30, H-40, H-6a),
3.51–3.21 (6H, m, H-4, H-5, H-50, H-6a0, H-6b, H-6b0);
13C NMR (CD3OD) d (ppm): 134.8 (C-meta), 134.2 (C-ipso),
3-(2-Furan)-5-(1-deoxy-b-D-galactopyranosyl)-isoxazole (14)
was obtained quantitatively from 14b under Zemplen
´
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 2229–2240 | 2237

View Article Online
132.9 (C-ortho), 122.4 (C-para), 104.9 (C-10), 88.7 (C-1), 80.5
(C-4), 80.0 (C-3), 77.95 (C-5), 77.1 (C-2), 74.8 (C-30), 73.3
(C-50), 72.5 (C-20), 70.3 (C-40), 62.5 (C-6), 61.9 (C-60); ESI-MS
calcd for C18H25O10BrS(Na+): 535.0; found: 535.2;
ESI+-HRMS: [M + Na]+ calcd for C18H25BrO10SNa:
535.0244; 537.0226; found: 535.0245; 537.0227, D 0.08 ppm.
p-Methoxyphenyl 1-thio-b-D-lactoside (18) was obtained
from acetobromolactose and p-methoxythiophenol using the
general PTC conditions described above (98%) followed by
(3H, m, HAR), 4.57 (1H, d, J = 5.5 Hz, H-10), 4.25 (1H, d,
J = 7.4 Hz, H-1), 3.76–3.34 (12H, m, H-2 to H-6, H-20 to
H-60); 13C NMR (DMSO) d (ppm): 164.2, 153.7, 137.0, 127.7,
126.4, 122.9, 122.4 (7CAR), 105.1, 101.2 (C-1, C-10), 80.8, 77.0,
74.8, 73.5, 73.2, 72.5, 71.8, 70.3, 62.4, 61.8 (C-2 to C-6, C-20 to
C-60). ESI+-HRMS: [M + H]+ calcd for C19H26NO10S2:
492.0993; found: 492.0999, D 1.11 ppm.
Compound 26b. To a solution of lactoside 20 in MeOH was
added Bu2SnO, Bu4NI and propargyl bromide. The mixture
was stirred under reflux for 7 hours, and then concentrated
under reduced pressure. To the crude yellow residue dissolved
in pyridine was added Ac2O dropwise. The mixture was stirred
16 hours at room temperature and concentrated under
reduced pressure. Flash chromatography using ethyl acetate:
hexanes (3 : 2) afforded compound 20a. This compound
showed satisfactory physical properties and was converted
into compound 26b, in 91% yield, following the same
procedure as for the synthesis of compound 14b. Compound
26b was isolated in a yellow oil and had [a]D25: ꢁ0.3 (c = 0.8,
CHCl3); 1H NMR (CDCl3) d (ppm): 8.12 (2H, d, J = 9.1 Hz,
quantitative Zemplen deacetylation as a white powder and had
´
mp 167.5–169.5 1C; [a]D25: ꢁ24.3 (c = 1, MeOH); IR: 3373
cmꢁ1 (s, O-H), 1246 cmꢁ1 (s, C-O-Me); 1H NMR (CD3OD) d
(ppm): 7.42 (2H, d, J = 8.7 Hz, H-ortho), 6.75 (2H, d,
J = 8.7 Hz, H-meta), 4.30 (1H, d, J = 9.6 Hz, H-10), 4.22
(1H, d, J = 7.4 Hz, H-1), 3.79–3.50 (6H, m, H-2, H-20, H-3,
H-30, H-40, H-6a), 3.67 (3H, s, OMe), 3.46–3.04 (6H, m, H-4,
H-5, H-50, H-6a0, H-6b, H-6b0); 13C NMR (CD3OD) d (ppm):
161.4 (C-para), 136.7 (C-ortho), 124.0 (C-ipso), 115.3 (C-meta),
104.9 (C-10), 89.6 (C-1), 80.5 (C-4), 80.1 (C-3), 77.9 (C-5), 77.0
(C-2), 74.7 (C-30), 73.1 (C-50), 72.5 (C-20), 70.3 (C-40), 62.5 (C-60),
62.0 (C-6), 55.7 (OMe); ESI-MS calcd for C19H28O11S(Na+):
487.1; found: 487.3; ESI+-HRMS: [M + Na]+ calcd for
C19H28O11SNa: 487.1245; found: 487.1252, D ꢁ0.42 ppm.
b-Naphthyl 1-thio-b-D-lactoside (19)27 was obtained as a
white powder (93%) and had mp 217.5–218.2 1C; [a]D25: ꢁ26.9
(c = 1, DMSO); IR: 3363 cmꢁ1 (s, O-H); 1H NMR (CD3OD)
d (ppm): 8.08 (1H, s, HAR), 7.80 (3H, m, HAR), 7.64 (1H, d,
J = 6.8 Hz, HAR), 7.47 (2H, t, J = 3.8 Hz, HAR), 4.73 (1H, d,
J = 9.8 Hz, H-10), 4.35 (1H, d, J = 7.1 Hz, H-1), 3.96–3.79
(2H, m, H-3, H-40), 3.77–3.65 (4H, m, H-20, H-2, H-30, H-6a),
3.58–3.46 (6H, m, H-4, H-5, H-50, H-6a0, H-6b, H-6b0);
13C NMR (CD3OD) d (ppm): 131.7, 127.7, 130.4, 128.6,
128.5, 127.5, 127.2 (10CAR), 104.9 (C-10), 89.0 (C-1), 80.6
(C-4), 80.1 (C-3), 78.0 (C-5), 77.1 (C-2), 74.8 (C-30), 73.4
(C-50), 72.5 (C-20), 70.3 (C-40), 62.5 (C-60), 62.0 (C-6);
ESI-MS calcd for C22H28O10S (Na+): 507.1; found: 507.3;
H
AR), 7.78–7.75 (2H, m, HAR), 7.53 (2H, d, J = 8.8 Hz, HAR),
7.45–7.43 (3H, m, HAR), 6.51 (1H, s, Hisoxazole), 5.44 (1H, d,
J = 3.0 Hz), 5.23 (1H, dd, J = 8.7 Hz), 5.03–4.91 (2H, m),
4.83 (1H, d, J = 10.1 Hz), 4.73 (1H, m), 4.57–4.40 (3H, m),
4.11 (1H, d, J = 6.3 Hz), 3.83–3.59 (6H, m), 2.15, 2.08, 2.06,
2.05, 2.02 (6CH3CO); 13C NMR (CDCl3) d (ppm): 170.6,
ꢁ ꢁꢁ
170.5, 170.3, 169.7, 169.4, 168.9 (6CH3CO), 162.6 (Cisoxazole),
ꢁ ꢁ
147.0, 142.2, 130.9, 130.4, 129.2, 128.8, 126.9, 124.0 (12CAR),
101.7, 101.3 (C-1, C-10), 84.1, 77.1, 76.8, 76.2, 73.8, 70.9, 70.4,
70.1, 65.3, 62.4, 62.3, 61.3, 21.0, 20.9, 20.8 (6CH3CO); ESI-MS
ꢁ
calcd for C40H44N2O19S(Na+): 911.82; found: 911.22.
Compound 26 was obtained by de-O-acetylation of 26b as a
yellow powder and had mp 220–221 1C; [a]D25: ꢁ45.2 (c = 0.8,
DMSO); 1H NMR (DMSO) d (ppm): 8.06 (2H, d, J = 7.9 Hz,
HAR), 7.78 (2H, m, HAR), 7.56 (2H, d, J = 8.2 Hz, HAR),
7.44–7.03 (3H, m, HAR), 4.95 (1H, d, J = 9.6 Hz, H-10), 4.75
ESI+-HRMS: [M
+
Na]+ calcd for C22H28O10SNa:
(1H, d, J = 8.5 Hz, H-1), 4.74–4.70 (2H, m, OCH2), 3.87–3.28
ꢁ ꢁꢁ
507.1295; found: 507.1305, D 1.2 ppm.
(12H, m, H-2 to H-6, H-20 to H-60); 13C NMR (DMSO) d
(ppm): 171.3, 162.3 (2Cisoxazole), 146.4, 145.5, 130.8, 129.8,
129.2, 128.5, 127.2, 127.1, 124.4, 124.3 (10CAR), 104.1, 102.1
(C-1, C-10), 85.2, 82.2, 80.2, 79.5, 76.8, 76.0, 72.8, 70.2, 65.3,
p-Bromophenyl 1-sulfonyl-b-D-lactoside (22) was obtained
from 17 using mCPBA as above (98%) as a white powder and
had mp 174.5–175.5 1C; [a]D25: ꢁ9.5 (c = 1, MeOH); IR:
3305 cmꢁ1 (s, O-H); 1H NMR (CD3OD) d (ppm): 7.76 (2H, d,
J = 8.5 Hz, H-ortho), 7.68 (2H, d, J = 8.5 Hz, H-meta), 4.32
(1H, d, J = 8.7 Hz, H-1), 4.20 (1H, d, J = 7.1 Hz, H-10),
3.67–3.31 (12H, m, H-2, H-20, H-3, H-30, H-40, H-40, H-5,
H-50, H-6a, H-6a0, H-6b, H-6b0); 13C NMR (CD3OD) d
(ppm): 137.3 (C-ipso), 133.3 (C-meta), 132.6 (C-ortho), 130.4
(C-para), 104.8 (C-10), 92.7 (C-1), 81.1 (C-4), 78.9 (C-3), 77.3
(C-5), 77.1 (C-2), 74.7 (C-30), 72.4 (C-50), 71.1 (C-20), 70.3
(C-40), 62.5 (C-6), 61.3 (C-60); ESI-MS calcd for
C18H25O12BrS(NH4+): 562.0; found: 562.2; ESI+-HRMS:
[M + Na]+ calcd for C18H25BrO12SNa: 567.0142; 569.0124;
found: 567.0145; 569.0124, D ꢁ0.47 ppm.
62.4, 60.8, 60.7 (C-2 to C-6, C-20 to C-60, CHisoxazole, OCH2);
ꢁ
ESI+-MS: [M
+
Na]+ calcd for C28H32N2O13SNa:
659.15173; found: 659.2.
Compound 26b (to be published elsewhere) was obtained as
1
a yellow oil and had [a]D25: ꢁ0.3 (c = 0.8, CHCl3); H NMR
(CDCl3) d (ppm): 8.12 (2H, d, J = 9.1 Hz, HAR), 7.78–7.75
(2H, m, HAR), 7.53 (2H, d, J = 8.8 Hz, HAR), 7.45–7.43
(3H, m, HAR), 6.51 (1H, s, Hisoxazole), 5.44 (1H, d, J = 3.0 Hz),
5.23 (1H, dd, J = 8.7 Hz), 5.03–4.91 (2H, m), 4.83 (1H, d,
J = 10.1 Hz), 4.73 (1H, m), 4.57–4.40 (3H, m), 4.11 (1H, d,
J = 6.3 Hz), 3.83–3.59 (6H, m), 2.15, 2.08, 2.06, 2.05, 2.02
(6CH3CO); 13C NMR (CDCl3)
d (ppm): 170.6, 170.5,
170.3, 169.7, 169.4, 168.9 (6CH3CO), 162.6 (Cisoxazole), 147.0,
ꢁ ꢁꢁ
2-Benzothiazolyl 1-thio-b-D-lactoside (23) was obtained
from acetobromolactose and 2-benzothiazole using the general
ꢁ ꢁ
142.2, 130.9, 130.4, 129.2, 128.8, 126.9, 124.0 (12CAR), 101.7,
101.3 (C-1, C-10), 84.1, 77.1, 76.8, 76.2, 73.8, 70.9, 70.4, 70.1,
PTC conditions described above (98%) followed by Zemplen
´
deacetylation (77%) and was isolated as a white solid and had
mp 95–96 1C; [a]D25: ꢁ12 (c = 0.5, DMSO); 1H NMR
(DMSO) d (ppm): 7.45 (1H, d, J = 6.8 Hz, HAR), 7.29–7.17
65.3, 62.4, 62.3, 61.3, 21.0, 20.9, 20.8 (6CH3CO); ESI+-MS:
ꢁ
[M + Na]+ calcd for C40H44N2O19SNa: 911.21512; found:
911.22.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
2238 | New J. Chem., 2010, 34, 2229–2240

View Article Online
Lectins
DG = DH ꢁ TDS = ꢁRT ln Ka, where DG, DH and DS
are the changes in free energy, enthalpy, and entropy of
binding, respectively. T is the absolute temperature and
Purification of the plant toxin and the human lectins (galectins-1,
-3, -4 and -9) started with extracts of dried leaves and pellets
from bacteria harboring expression vector for the respective
galectin and involved affinity chromatography on lactosylated
Sepharose 4B, obtained after activation with divinyl sulfone,
as crucial step.3a,31 Purity controls by one- and two-dimensional
gel electrophoresis and mass spectrometry by nanoESI-MS/MS
on a QTOF instrument Q-Tof2 (Waters Micromass, Manchester,
UK), determination of the quaternary structure by gel filtration
on a Superose 12 HR 10/30 column (GE Healthcare, Munich,
Germany) and activity assays in 2-fold serial dilutions with
trypsinized, glutaraldehyde-fixed rabbit erythrocytes ensured
the lectins’ quality.3a,32 Biotinylation was performed under
activity-preserving conditions with the N-hydroxysuccinimide
ester derivative of biotin (Sigma, Munich, Germany) followed
by gel electrophoretic/mass spectrometric product analysis.33
R = 1.98 cal molꢁ1 Kꢁ1
.
Cell-binding inhibition assay. Cells of the human colon
adenocarcinoma line SW480, the B-lymphoblastoid line Croco II
and the pancreatic carcinoma line Capan-1 reconstituted for
expression of the tumor suppressor p16INK4a (kindly provided
by K. M. Detjen, Berlin, Germany) were grown and processed
as described, using a 30 min incubation step for labeling and
fluorescent streptavidin/R-phycoerythrin (1 : 40; Sigma) as
indicator in automated FACScan analysis, data processing
yielding the percentage of positive cells and the mean fluorescence
intensity.3a Controls included omission of the incubation step
with labeled lectin and application of non-cognate sugar to
detect any osmolarity effect. Assays were performed in triplicates
with up to four independent series using aliquots of cell
suspensions of the same or the next passage with standard
deviations not exceeding 12% after normalization of data.
Solid-phase inhibition assay. The neoglycoprotein with
average ligand density of 26 sugar units per carrier
protein (bovine serum albumin) was prepared using the
p-isothiocyanatophenyl derivative and adsorbed to the surface
of microtiter plate wells from a solution in 20 mM phosphate-
buffered saline (PBS) (routinely using a total of 0.25 mg in
50 ml) at 4 1C overnight. Residual sites for binding proteins
were saturated with albumin free of contaminating glyco-
proteins (Biomol, Hamburg, Germany; 100 ml, 1% w/v) for
1 h at 37 1C. Lectin-containing solution (routinely using 1.5 mg
mlꢁ1 VAA, 20 mg mlꢁ1 galectin-1, 15 mg mlꢁ1 galectins-3 and
-9 and 3 mg mlꢁ1 galectin-4 based on thorough titrations) in
the absence (control) or presence of test compound (at up to
eight different concentrations) was incubated in the processed
microtiter plate wells for 1 h at 37 1C, and extent of bound
lectin was quantitated spectrophotometrically at 490 nm.
In detail, solutions of the indicator conjugate streptavidin-
peroxidase (0.5 mg mlꢁ1; Sigma) and the signal-generating
Cytotoxicity assay. Extent of cell proliferation in the
absence (control) or presence of test compounds was measured
after
a period of 48 h using the blue chromogen
3-(4,5-dimethyl-thiazol-2-yl)-2,5-diphenyltetrazolium bromide
(5 mg mlꢁ1; Sigma) as indicator for spectrophotometric
determination. Assays were performed in triplicates (nine
samples for the control) with aliquots of a cell suspension
and three independent series at standard deviations of less
than 14%.
Acknowledgements
The generous financial support by a Canadian Research Chair
in Therapeutic Chemistry grant from the NSERC and
the Verein zur Forderung des biologisch-technologischen
¨
substrates o-phenylenediamine (1 mg mlꢁ1)/H2O2 (1 ml mlꢁ1
were stepwisely applied after thorough washes. Assays were
routinely done in triplicates for up to four independent series
with standard deviations not exceeding 15.4%.
)
Fortschritts in der Medizin e.V. (Heidelberg, Germany) as
well as the constructive comments by the reviewers and
inspiring discussions with Drs B. Friday and S. Namirha are
gratefully acknowledged together with the technical assistance
of M. Tze Chieh Shiao.
Isothermal titration microcalorimetry. These experiments
were performed with purified galectins using a VP-ITC instrument
from Microcal, Inc. (Northampton, USA) as described.13 In
detail, injections of 4 ml of solutions containing the synthetic
were added from a computer-controlled microsyringe at an
interval of 4 min into the sample solution containing the
galectin (cell volume = 1.43 ml) with stirring at 350 rpm.
Control experiments were performed by making identical
injections of the carbohydrate derivative into a cell containing
buffer. Titrations were done at pH 7.2 using PBS. The experi-
mental data were fitted to a theoretical titration curve using
software supplied by Microcal, with DH (binding enthalpy
kcal molꢁ1), Ka (association constant) and n (number of
binding sites per monomer), as adjustable parameters. The
instrument was calibrated using the calibration kit containing
ribonuclease A (RNase A) and cytidine 20-monophosphate
supplied by the manufacturer. Thermodynamic parameters
were calculated from the Gibbs Free Energy equation,
References
1 R. A. Laine, in Glycosciences: Status and Perspectives, ed. H.-J.
Gabius and S. Gabius, Chapman & Hall, Weinheim, 1997,
pp. 1–14.
2 The Sugar Code. Fundamentals of glycosciences, ed. H.-J. Gabius,
Wiley-VCH, Weinheim, 2009.
3 (a) S. Andre
J. Kopitz, P. Forberich, W. Kemmner, C. Bo
¨
´
, H. Sanchez-Ruderisch, H. Nakagawa, M. Buchholz,
ck, K. Deguchi,
K. M. Detjen, B. Wiedenmann, M. von Knebel Doeberitz,
T. M. Gress, S.-I. Nishimura, S. Rosewicz and H.-J. Gabius, FEBS
J., 2007, 274, 3233–3256; (b) R. Schwartz-Albiez, in The Sugar
Code. Fundamentals of glycosciences, ed. H.-J. Gabius,
Wiley-VCH, Weinheim, 2009, pp. 447–467.
4 H. M. I. Osborn and A. Turkson, in The Sugar Code.
Fundamentals of glycosciences, ed. H.-J. Gabius, Wiley-VCH,
Weinheim, 2009, pp. 469–483.
5 (a) Y. M. Chabre and R. Roy, in The Sugar Code. Fundamentals of
glycosciences, ed. H.-J. Gabius, Wiley-VCH, Weinheim, 2009,
pp. 53–70; (b) Y. M. Chabre and R. Ray, Adv. Carbohydr. Chem.
Biochem., 2010, 63, 165–393.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 2229–2240 | 2239

View Article Online
6 (a) Y. Wada, T. Saito, N. Matsuda, H. Ohmoto, K. Yoshino,
M. Ohashi, H. Kondo, H. Ishida, M. Kiso and A. Hasegawa,
J. Med. Chem., 1996, 39, 2055–2059; (b) H. Tsujishita,
Y. Hiramatsu, N. Kondo, H. Ohmoto, H. Kondo, M. Kiso and
A. Hasegawa, J. Med. Chem., 1997, 40, 362–369; (c) J. C. Pickens,
D. D. Mitchell, J. Liu, X. Tan, Z. Zhang, C. L. M. J. Verlinde,
W. G. J. Hol and E. Fan, Chem. Biol., 2004, 11, 1205–1215;
(d) J. Liu, D. Begley, D. D. Mitchell, C. L. M. J. Verlinde,
G. Varani and E. Fan, Chem. Biol. Drug Des., 2008, 71, 408–419.
7 (a) S. Olsnes, F. Stirpe, K. Sandvig and A. Pihl, J. Biol. Chem., 1982,
257, 13263–13279; (b) S. H. Barondes, Science, 1984, 223, 1259–1264;
(c) Y. M. Endo, Adv. Lectin Res., 1989, 2, 60–73; (d) M. Caron,
D. Bladier and R. Joubert, Int. J. Biochem., 1990, 22, 1379–1385;
(e) Trends Glycosci. Glycotechnol., ed. J. Hirabayashi, 1997, vol. 9,
pp. 1–180; (f) H. Kaltner and B. Stierstorfer, Acta Anat., 1998, 161,
162–179; (g) H.-J. Gabius, Biochimie, 2001, 83, 659–666; (h) D. N.
W. Cooper, Biochim. Biophys. Acta, 2002, 1572, 209–231; (i) F. Stirpe
and M. G. Batelli, Cell. Mol. Life Sci., 2006, 63, 1850–1866;
(j) Galectins., ed. A. E. Klyosov, Z. J. Witczak and D. Platt, Wiley,
Hoboken, 2008; (k) L. V. Norling, M. Perretti and D. Cooper,
J. Endocrinol., 2009, 201, 169–184; (l) A. M. Wu, E. Lisowska,
M. Duck and Z. Yang, Glycoconjugate J., 2009, 26, 899–913.
Biol., 2004, 343, 957–970; (b) M. Jime
A. Romero, D. Rognan, H.-J. Gabius and D. Solı
2008, 582, 2309–2312.
12 (a) L. He, S. Andre, H.-C. Siebert, H. Helmholz, B. Niemeyer and
H.-J. Gabius, Biophys. J., 2003, 85, 511–524; (b) L. He, S. Andre
´
nez, S. Andre
´
, C. Barillari,
s, FEBS Lett.,
´
´
´
,
V. M. Garamus, H.-C. Siebert, C. Chi, B. Niemeyer and
H.-J. Gabius, Glycoconjugate J., 2009, 26, 111–116.
13 T. K. Dam, H.-J. Gabius, S. Andre
C. F. Brewer, Biochemistry, 2005, 44, 12564–12571.
14 M. Jimenez, S. Andre, H.-C. Siebert, H.-J. Gabius and D. Solı
Glycobiology, 2006, 16, 926–937.
15 (a) J. L. Asensio, J. F. Espinosa, H. Dietrich, F. J. Canada,
R. R. Schmidt, M. Martın-Lomas, S. Andre, H.-J. Gabius and
J. Jimenez-Barbero, J. Am. Chem. Soc., 1999, 121, 8995–9000;
(b) D. Solıs, A. Romero, M. Menendez and J. Jimenez-Barbero, in
´
, H. Kaltner, M. Lensch and
´
´
´
s,
´
´
´
´
´
´
The Sugar Code. Fundamentals of glycosciences, ed. H.-J. Gabius,
Wiley-VCH, Weinheim, 2009, pp. 233–245.
16 N. Ahmad, H.-J. Gabius, S. Andre, H. Kaltner, S. Sabesan,
´
R. Roy, B. Liu, F. Macaluso and C. F. Brewer, J. Biol. Chem.,
2004, 279, 10841–10847.
´
17 S. Andre, C. E. P. Maljaars, K. M. Halkes, H.-J. Gabius and
J. P. Kamerling, Bioorg. Med. Chem. Lett., 2007, 17,
793–798.
8 (a) A. DeWaard, S. Hickman and S. Kornfeld, J. Biol. Chem., 1976,
251, 7581–7587; (b) P. Luther, H. Theise, B. Chatterjee, D. Karduck
and G. Uhlenbruck, Int. J. Biochem., 1980, 11, 429–435;
(c) H. Debray, J. Montreuil and H. Franz, Glycoconjugate J., 1994,
11, 550–557; (d) O. E. Galanina, H. Kaltner, L. S. Khraltsova,
N. V. Bovin and H.-J. Gabius, J. Mol. Recognit., 1997, 10,
139–147; (e) J. Hirabayashi, T. Hashidate, Y. Arata, N. Nishi,
T. Nakamura, M. Hirashima, T. Urashima, T. Oka, M. Futai, W.
18 (a) J. Kopitz, in The Sugar Code. Fundamentals of glycosciences, ed.
H.-J. Gabius, Wiley-VCH, Weinheim, 2009, pp. 177–198;
(b) J. Wang, Z.-H. Lu, H.-J. Gabius, C. Rohowsky-Kochan,
R. W. Ledeen and G. Wu, J. Immunol., 2009, 182, 4036–4045.
19 (a) R. Roy, F. D. Tropper, S. Cao and J. M. Kim, ACS Symp. Ser.,
1997, 569, 163–180; (b) D. Carriere, S. J. Meunier, F. D. Tropper,
S. Cao and R. Roy, J. Mol. Catal. A: Chem., 2000, 154, 9–22;
(c) S. Cao, S. J. Meunier, F. O. Andersson, M. Letellier and
R. Roy, Tetrahedron: Asymmetry, 1994, 5, 2303–2312.
E. G. Muller, F. Yagi and K.-i. Kasai, Biochim. Biophys. Acta, 2002,
¨
1572, 232–254; (f) A. M. Wu, T. Singh, J. H. Wu, M. Lensch,
S. Andre
´
and H.-J. Gabius, Glycobiology, 2006, 16, 524–537;
20 Y. Yde and C. K. De Bruyne, Carbohydr. Res., 1973, 26, 227–229.
21 A. I. Khodair, N. A. Al-Masoudi and J.-P. Gesson, Nucleosides
Nucleotides Nucleic Acid, 2003, 22, 2061–2076.
22 S. Mehta, M. Gilbert, W. W. Wakarchuk and D. M. Whitfield,
Org. Lett., 2000, 2, 751–753.
(g) M. Nagae, N. Nishi, S. Nakamura-Tsuruta, J. Hirabaysahi,
S. Wakatsuki and R. Kato, J. Mol. Biol., 2008, 375, 119–135;
(h) S. R. Stowell, C. M. Arthur, P. Mehta, K. A. Slanina, O. Blixt,
H. Leffler, D. F. Smith and R. D. Cummings, J. Biol. Chem., 2008,
283, 10109–10123; (i) H. Tateno, A. Kuno and J. Hirabayashi, in The
Sugar Code. Fundamentals of glycosciences, ed. H.-J. Gabius,
Wiley-VCH, Weinheim, 2009, pp. 247–259.
9 (a) Y. Kashio, K. Nakamura, M. J. Abedin, M. Seki, N. Nishi,
N. Yoshida, T. Nakamura and M. Hirashima, J. Immunol., 2003,
170, 3631–3636; (b) C. Zhu, A. C. Anderson, A. Schubart,
H. Xiong, J. Imitola, S. J. Khoury, X. X. Zheng, T. B. Strom
and V. K. Kuchroo, Nat. Immunol., 2005, 6, 1245–1252;
(c) O. B. Garner and L. G. Baum, Biochem. Soc. Trans., 2008,
36, 1472–1477; (d) D. K. Hsu, H.-Y. Chen and F.-T. Liu, Immunol.
Rev., 2009, 230, 114–127.
23 Z. Gan and R. Roy, Tetrahedron Lett., 2000, 41, 1155–1158.
24 N. Khiar, I. Alonso, N. Rodriguez, A. Fernandez-Mayoralas,
J. Jime
´
M. Martı
nez-Barbero, O. Nieto, F. Cano, C. Foces-Foces and
n-Lomas, Tetrahedron Lett., 1997, 38, 8267–8270.
´
25 J. English and E. Leffler, J. Am. Chem. Soc., 1950, 72, 4324–4325.
26 K. Micskei, Z. Juhasz, Z. R. Ratkovic and L. Somsak, Tetrahedron
Lett., 2006, 47, 6117–6120.
27 D. Giguere, S. Sato, C. St-Pierre, S. Sato, S. Sirois and R. Roy,
Bioorg. Med. Chem. Lett., 2006, 16, 1668–1672.
28 T. Kuribayashi, N. Ohkawa and S. Satoh, Tetrahedron Lett., 1998,
39, 4537.
10 (a) H. Ahmed, H. J. Allen, A. Sharma and K. L. Matta, Bio-
chemistry, 1990, 29, 5315–5319; (b) R. T. Lee, Y. Ichikawa,
H. J. Allen and Y. C. Lee, J. Biol. Chem., 1990, 265, 7864–7871;
(c) R. T. Lee, H.-J. Gabius and Y. C. Lee, J. Biol. Chem., 1992,
267, 23722–23727; (d) J. Tejler, F. Skogman, H. Leffler and
U. J. Nilson, Carbohydr. Res., 2007, 342, 1869–1875 (and ref. cited
therein); (e) L. Moni, G. Pourceau, J. Zhang, A. Meyer, S. Vidal,
E. Souteyrand, A. Dondoni, F. Morvan, Y. Chevolot, J.-J. Vasseur
and A. Marra, ChemBioChem, 2009, 10, 1369–1378.
29 A. Trincone, G. Perugino, M. Rossi and M. Moracci, Bioorg. Med.
Chem. Lett., 2000, 10, 365–368.
30 S. Beha, D. Giguere, R. Patnam and R. Roy, Synlett, 2006,
´
1739–1743.
31 M. Lensch, M. Lohr, R. Russwurm, M. Vidal, H. Kaltner,
S. Andre and H.-J. Gabius, Int. J. Biochem. Cell Biol., 2006, 38,
´
1741–1758.
32 H.-J. Gabius, R. Engelhardt, F. Cramer, R. Batge and
¨
G. A. Nagel, Cancer Res., 1985, 45, 253–257.
33 H.-J. Gabius, B. Wosgien, M. Hendrys and A. Bardosi,
Histochemistry, 1991, 95, 269–277.
11 (a) M. F. Lo
´
pez-Lucendo, D. Solı
´
s, S. Andre
´
, J. Hirabayashi,
K.-i. Kasai, H. Kaltner, H.-J. Gabius and A. Romero, J. Mol.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
2240 | New J. Chem., 2010, 34, 2229–2240