Bi- to tetravalent glycoclusters: Synthesis, structure-activity profiles as lectin inhibitors and impact of combining both valency and headgroup tailoring on selectivity

Article abstract of DOI:10.1039/c2ob25870f

The emerging functional versatility of cellular glycans makes research on the design of synthetic inhibitors a timely topic. In detail, the combination of ligand (or headgroup or contact site) structure with spatial parameters that depend on topological a

Full text of DOI:10.1039/c2ob25870f

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Organic &
Biomolecular
Chemistry
Cite this: Org. Biomol. Chem., 2012, 10, 6893
www.rsc.org/obc
PAPER
Bi- to tetravalent glycoclusters: synthesis, structure–activity profiles as lectin
inhibitors and impact of combining both valency and headgroup tailoring
on selectivity†
Guan-Nan Wang,a Sabine André,b Hans-Joachim Gabiusb and Paul V. Murphy*a
Received 7th May 2012, Accepted 27th June 2012
DOI: 10.1039/c2ob25870f
The emerging functional versatility of cellular glycans makes research on the design of synthetic
inhibitors a timely topic. In detail, the combination of ligand (or headgroup or contact site) structure
with spatial parameters that depend on topological and geometrical factors underlies the physiological
selectivity of glycan-protein (lectin) recognition. We herein tested a panel of bi-, tri- and tetravalent
compounds against two plant agglutinins and adhesion/growth-regulatory lectins (galectins). In addition,
we examined the impact of headgroup tailoring (converting lactose to 2′-fucosyllactose) in combination
with valency increase in two assay types of increasing biorelevance (from solid-phase binding to cell
binding). Compounds were prepared using copper-catalysed azide alkyne cycloaddition from
peracetylated lactosyl or 2′-fucosyllactosyl azides. Significant inhibition was achieved for the plant toxin
with a tetravalent compound. Different levels of sensitivity were noted for the three groups of the galectin
family. The headgroup extension to 2′-fucosyllactose led to a selectivity gain, especially for the chimera-
type galectin-3. Valency increase established discrimination against the homodimeric proteins, whereas
the combination of valency with the headgroup extension led to discrimination against the tandem-repeat-
type galectin-8 for chicken galectins but not human galectins-3 and -4. Thus, detailed structure–activity
profiling of glycoclusters combined with suitably modifying the contact site for the targeted lectin will
help minimize cross-reactivity among this class of closely related proteins.
sterically rigid presentation of carbohydrate-binding sites,
especially the hepatic asialoglycoprotein receptor using a trian-
Introduction
The steadily growing body of evidence on the physiological sig-
nificance of protein (lectin)-glycan recognition gives reason to
aim at lectin-directed rational drug design in order to block clini-
cally unfavourable interactions, e.g. in inflammation, tumour
progression or pathogen adhesion.1 Clues to which parameters
deserve special attention in the design of (bio)pharmaceuticals
come from the delineation of the levels of affinity regulation of
glycan binding to lectins.2 The ligand structure (headgroup and
aglycone) and the spatial presentation of the ligand in glycoclus-
ters are two key features to be considered in attaining the overall
objective. A third is the valency order (monovalent vs. bivalent
vs. trivalent etc.). In considering the spatial presentation, the geo-
metry, topology and inter-ligand distance needs to be taken into
account. The elegant work on targeting membrane lectins with a
tennary N-glycan or synthetic cluster glycosides with a matching
spatial arrangement, laid the foundation for the concept of the
glycoside cluster effect.3 In that case, a numerical increase in
valency from one to three reactive headgroups in a neoglyco-
conjugate led to an enhancement in affinity, mimicking the
potency of the type I triantennary N-glycan.4 Since the natural
effector activity of lectins, presented in membranes or in solu-
tion, is based on binding to structurally and spatially suitable
counter-receptors, devising an adequately tailored combination
of these two parameters is considered helpful in addressing the
challenge to design inhibitors with optimal potency. Given the
wide variety of natural ways for lectin-site presentation2 and
the diversity of (bio)chemical scaffolds which can be used in
glycocluster formation,5 a clear study design will help to discern
structure–activity relationships for medically relevant lectins.
Following our initial reports on different types of bivalent lacto-
sides,6 we herein explore the effect of stepwise increases in
valency, from mono- to tetravalency, as well as changing the
structure of the headgroup from lactose to the histo-blood H-type
structure – 2′-fucosyllactose (Chart 1). In detail, the monovalent
control compounds that were investigated are 2′-fucosyllactose
aSchool of Chemistry, National University of Ireland Galway, University
Road, Galway, Ireland. E-mail: paul.v.murphy@nuigalway.ie
bInstitute of Physiological Chemistry, Faculty of Veterinary Medicine,
Ludwig-Maximilians-University Munich, Veterinärstr. 13, 80539
Munich, Germany
†Electronic supplementary information (ESI) available. See DOI:
10.1039/c2ob25870f
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 6893–6907 | 6893

View Article Online
Chart 1 Structures of the synthetic compounds 1–7.
(FL) 1 and a triazole conjugate 2, together with lactose. The set
of lactose-presenting di-, tri- and tetravalent glycoclusters 3–5
and 7 and the trivalent FL derivative 6, which has the same core
scaffold as 4, complete the compound panel. The testing of
this panel was performed in assays involving the same group
of lectins in the quest to define the impact of the given two
parameters on bioactivity.
The test panel of lectins investigated herein consists of a plant
toxin (Viscum album L. agglutinin, VAA) and two types of
β-sandwich-fold proteins, i.e. a leguminous lectin (Erythrina
crystagalli agglutinin, ECA) and adhesion/growth-regulatory
galectins. Thus, the synthetic compounds are tested for potency
as anti-toxin (VAA) and for reactivity to lectins sharing the same
fold but differing in positioning the lectin sites (ECA, galectins).
We comprehensively studied all known members of the latter
family from one organism, i.e. the five chicken galectins (CGs),
to spot intrafamily differences and added work on two human
galectins, i.e. Gal-3/-4, for comparison. These lectins’ contact
sites for carbohydrates are presented in three modes of topologi-
cal display: proto-type (non-covalently associated homodimers:
CG-1A/-1B/-2), tandem-repeat-type (two different domains
covalently linked by a peptide: CG-8, Gal-4) and chimera-type
6894 | Org. Biomol. Chem., 2012, 10, 6893–6907
This journal is © The Royal Society of Chemistry 2012

View Article Online
(a single carbohydrate recognition domain connected to a
collagenase-sensitive stalk and an N-terminal section with two
acceptors for serine phosphorylation: CG-3, Gal-3).7 With regard
to the carbohydrate the extension from lactose to the histo-blood
H-type structure was expected to have little impact on VAA/
ECA,8 to slightly prefer CG-1B when compared to CG-1A9 and
to increase the affinity for CG-8,10 Gal-311 and Gal-4.12 The
respective comparative measurements provide an instructive
example for the influence of headgroup/valency tailoring on
lectin affinity and selectivity. They were performed in two
experimental set-ups. The first is a solid-phase assay, in which
extent of lectin binding to a glycoprotein matrix (asialofetuin,
ASF, which has up to nine N-acetyllactosamine termini on its
bi- and triantennary N-glycans; they all can bind to VAA and
galectins13) was determined. To increase biorelevance, that is to
work with cells, an assay was subsequently utilised where block-
ing of lectin binding to the cell surface was assessed.
hydrolysis of the acetonide groups using 60% acetic acid at
60 °C and subsequent catalytic hydrogenolysis provided
FL 1.19,20 This trisaccharide was acetylated and the azide group
was introduced using SnCl4 and TMSN3 to give 27 (Scheme 3).
It is worth mentioning that the α-glycosidic linkage was sensitive
to the Lewis acid if the benzyl groups were present on the fucose
residue while trying to introduce the azide group to form a fuco-
syllactosyl azide. The use of TMSN3-SnCl4, 33% HBr in AcOH,
and BiBr3/TMSBr21 all led to the cleavage of this fucosidic
bond. In contrast, the fully acetylated FL 26 was found to be
more stable.
The CuAAC reaction between fucosyllactosyl azide 27 and
Fmoc-protected L-propargylglycine,22 which was followed by
protecting group removal using initially piperidine and then
methanolic sodium methoxide, provided the glycamino acid 2
(Scheme 4). The trivalent fucosyllactoside 6 was prepared via
click reaction of 1,3,5-tris(alkynyloxy)benzene 1315 and fucosyl-
lactosyl azide 27. This gave 28 which was subsequently deacetyl-
ated. Having herewith completed the synthetic component of this
work, insights into the spatial property of the maximal distance
between sugar headgroups within each type of glycocluster were
obtained by molecular modelling.
Results and discussion
Synthesis
The synthetic routes from the bi-, tri- and tetravalent alkyne pre-
cursors to 1–8 are shown in Scheme 1. The dialkyne 9 and tetra-
alkyne 11 were obtained via the reaction of alkyl bromides 8 and
10 with potassium phthalimide (PhthK).14 1,3,5-Tris(alkynyl-
oxy)benzene 13 was prepared from phloroglucinol 12 and pro-
pargyl bromide.15 The trialkyne 16 was made from the coupling
reaction of 14 and 15.
Multivalent lactosides were all prepared from the lactose azide
17 (Scheme 2).6a Thus alkyne 9, 11, 13, 16, when reacted with
17 using copper(I)-catalysed azide alkyne cycloaddition
(CuAAC)16 reactions, gave the protected intermediates 18–21.
The CuAAC reactions were carried out using the in situ
reduction of copper(II) sulphate by sodium ascorbate in aqueous
solution using methanol as a co-solvent. As the number of
alkynes increased, the completion of CuAAC reaction was found
to be more difficult to achieve. Similar to the situation in the syn-
thesis of compounds 19, 20 and 21, either ultrasonic radiation17
or heating was required to accelerate the reactions. Both phthal-
imido groups and acetyl groups in 18 and 19 were removed using
ethylenediamine in ethanol by heating at reflux, giving 3 and 7
after washing the solid product with small amount of methanol.
The trivalent lactosides 20 and 21 afforded 4 and 5 after
Zemplén deacetylation.
The synthesis of trisaccharide 24 was achieved via glycosida-
tion with the fucosyl donor 2218 and acceptor 23 (Scheme 3).
For the synthesis of 23 an approach originally described by
Matta and co-workers was used.19a Two promoter systems NIS–
TfOH (46%) and benzenesulfinylpiperidine (BSP)–Tf2O (70%)
were comparatively investigated for the glycosidation to give the
trisaccharide 24. In terms of reaction time and yield, the BSP–
Tf2O promoter system turned out to be substantially better for
this glycosidation. The anomeric configuration of the glycosidic
linkage between fucose and lactose residues was assigned based
on the size of the coupling constant (J1′′,2′′ = 3.3 Hz) in the
1H NMR spectrum. The signal for C-1′′ occurred at δ 95.2 ppm
in the 13C-NMR spectrum. Debenzoylation of 24 was carried out
using methanolic sodium methoxide to give 25. Subsequent
Molecular modeling
The assessment of the distance between the sugar headgroups in
the synthetic glycoclusters was set as goal for this part of the
study. As a common feature and not influenced by changes in
the conformation of the terminal galactose unit, the anomeric
center of the glucose moiety was selected as the reference point.
Using the Maestro interface, adequate constraints were applied
(distances between residues, angles) in an iterative fashion to
generate extended conformations. While maintaining these con-
straints energy minimizations were performed using Macromodel
(OPLSAA force field, gas phase). The resulting extended confor-
mations are shown in Fig. 1. As expected, the distances between
headgroups in 3 and 4 were rather similar at 13–15 Å. This goes
well beyond the 5.9 Å/8.1 Å seen in fully extended or back-
folded biantennary N-glycans, in which both branch-end sugars
can bind with the tested lectins.13,23 The long and mostly
aliphatic spacer in 5 facilitates the lactose residues to attain an
interligand distance limit of about 27 Å (Fig. 1), which is ∼5 Å
more than that found in the core-disubstituted N-glycan with a
backfolded and an extended antenna.23 As depicted in Fig. 1, the
glucose anomeric carbon atoms in 7 adopt a rectangular arrange-
ment with the following set of distances: ∼14 Å and ∼18 Å
along the sides and ∼22 Å on the diagonal. Of course, in all
cases the inherent flexibility in the scaffolds would allow the
positions of the headgroups to fluctuate in space and time and
acquire conformers with lowered levels of spatial headgroup sep-
aration. The comparison to the biantennary N-glycans described
confirms that there are no spatial restrictions which would
impede headgroup reactivity with the lectins in the tested panel.
Assaying inhibitory properties on plant lectins
In principle, the solid-phase assay reflects the physiological
situation, in which the lectin in solution can bind to
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 6893–6907 | 6895

View Article Online
Scheme 1 Synthesis of alkyne precursors.
surface-presented glycans. This binding can be disrupted by
inhibitors. The glycoprotein asialofetuin was adsorbed to the
plastic surface of microtitre plate wells, and the evaluation of
lectin-glycoprotein binding was in each case shown to be satur-
able and inhibitable by the cognate sugar lactose but not by
mannose or glucose. In order to reach optimal sensitivity experi-
mental conditions were defined so that the signal increase associ-
ated with increasing lectin concentration and consequently
binding to the glycoprotein was in the linear and not saturated
range. Titrations with the synthetic products as inhibitors at a
constant lectin concentration established curves of decreasing
signal intensity. The apparent inhibitory activity fulfilled the
expectation raised by the molecular modelling. In order to
compare the relative potencies, these curves enabled us to define
the concentration of sugar presented by the compounds at which
a 50% decrease in optical density (IC50-value) was reached (for
an example, please see Fig. 2). In accordance with previous
affinity measurements8 the trisaccharide FL 1 was not found to
be more active than lactose (Table 1, see end of document).
However, a major avidity increase occurred in the progression to
tetravalency for the toxin (Fig. 2, Table 1). This held true for its
hololectin constituted by dimers of the toxin (A) and lectin (B)
subunits (in which the two accessible Tyr249 sites are separated
by 87 Å) and also the isolated B-chain (with a distance between
the Tyr249 and Trp38 sites of 62 Å24), all distances thus beyond
the range coverable by the glycoclusters 3–7. The Trp38 sites are
only 15 Å apart but not fully accessible.24 In contrast, the posi-
tioning of contact sites on opposing sides of the leguminous
lectin ECA apparently precluded there being an enhancement for
the tetravalent compound 7. Previous experience with tetravalent
6896 | Org. Biomol. Chem., 2012, 10, 6893–6907
This journal is © The Royal Society of Chemistry 2012

View Article Online
Scheme 2 Synthesis of di- to tetravalent lactosides.
glycoclusters attests that not just valency but the geometric
mode of ligand presentation matters, giving both functionalized
dendritic poly(amidoamine) and pentaerythritol-based com-
pounds exceptional potencies.25
Since the type of glycan display on the matrix (e.g. branching
mode of N-glycans) can have an influence on the inhibitory
efficiency of glycoclusters,25a,e it was essential to confirm
potency on a more relevant physiological level, i.e. by revealing
the efficiency of compounds to protect cells from lectin binding.
Thus, we performed cell assays with fluorescently labelled
lectin.
In this type of assay, the synthetic compounds compete with
cell surface glycans for lectin binding. The extent of signal
reduction (in terms of percentage of positive cells and mean
fluorescence intensity) that results by blocking lectin binding
was determined. Assays were routinely performed with aliquots
of the same cell suspension, avoiding prolonged culture times
and routinely running internal standards (0%/100%-values, inhi-
bition with lactose). As in the case of the solid-phase system,
dependence of signals on lectin concentration and presence of
cognate sugar was first ascertained, as exemplarily shown for
VAA (Fig. 3A and B). Compared to lactose the inhibitory
capacity of the test compounds was in most cases only slightly
improved (please see documented examples for the bivalent
compound 3 and the trivalent compound 4 in Fig. 3C and D), in
accord with the solid-phase-based data (Table 1). Tetravalent
compound 7, which reached approximately a 20-fold enhance-
ment (Table 1), was the most potent inhibitor. While 20 mM
lactose led to decreases to 50%/23.6 from the control level of
72%/89.3, the presence of 0.5 mM lactose in a glycocluster pres-
entation reduced lectin binding to the cells to 50%/32.9 (Fig. 3B
and D). Overall, correlation between the results of the two types
of assays (solid phase vs. cell) was thus found. Having started
with two plant lectins, we proceeded to measurements with the
five CGs. They exhibit sequence variations in their carbohydrate
recognition domains and cover the three modes of lectin-site
presentation, an attractive model to address the issues on impact
of headgroup extension and valency.
Assaying inhibitory properties on chicken galectins
The structure of the headgroup was clearly relevant among this
group of lectins (Table 1). A lowered affinity with CG-1A had
been indicated by frontal affinity chromatography.11d The same
holds true for human and rat Gal-1.11a,26 Occurrence of intra-
family differences between CGs was further underscored by slight
enhancement of reactivity for CG-810 and reduced sensitivity to
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 6893–6907 | 6897

View Article Online
Scheme 3 Synthesis of FL and fucosyllactosyl azide 27.
Scheme 4 Synthesis of fucosyllactosyl conjugates 2 and 6.
α1,2- or α1,3-substitutions seen for CG-1B relative to CG-1A.9
The comprehensive profiling of lectin reactivity to 1–7 singled
out the chimera-type lectin as the most responsive (Table 1). The
CG-3 was especially reactive with FL and found to be suscep-
tible to an increase in valency, with the highest affinity being
observed towards 6 (Table 1). The proteolytic removal of the
collagenase-sensitive stalk, which underlies galectin-3′s capacity
to form stable aggregates in the presence of oligovalent
ligands,27 did not reduce the relative affinity to lactose but did
impair the sensitivity to valency increase (comparing lactose and
1–7 for CG-3 and trCG-3). Tri- and tetravalency affected the
tandem-repeat-type CG-8 and its separate domains differently,
6898 | Org. Biomol. Chem., 2012, 10, 6893–6907
This journal is © The Royal Society of Chemistry 2012

View Article Online
Fig. 1 Models of extended conformers of compounds 3–5 and 7.
Fig. 2 Titration curves illustrating the course of inhibition of binding
of biotinylated VAA (1.5 μg ml−1) to asialofetuin in a solid-phase assay
upon increasing the concentration of lactose (▲, A) as well as (B) tri-
valent compound 4 (□) and tetravalent compound 7 (●).
and the increase in length of the linker peptide from nine
(CG-8S) to 28 amino acids (CG-8L) appeared to be associated
with a minor enhancement (Table 1). In terms of achieving
selectivity, the modification to include fucose in the headgroup
when combined with trivalency as seen in 6 led to the highest
inhibition for the full-length CG-3, less so for CG-8 and its
domains. Drawing on data for human Gal-3 affords a route to
further enhancements. Since the fucose moiety in α1,2-linkage is
only weakly involved in interactions to human Gal-3 relative to
the additional α1,3-substitution in histo-blood group AB-
determinants based on flexible ligand docking,11f a further
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 6893–6907 | 6899

View Article Online
elaboration to generate a tetrasaccharide rather than a trisacchar-
ide headgroup will be conducive for affinity increase, shown
calorimetrically to move ΔG from −16.4 kJ mol−1 for lactose
and −19.15 kJ mol−1 for FL to −24.88 kJ mol−1 for the histo-
blood group A-tetrasaccharide (at about 280 K).28 In inhibition
assays, the relative potency, with lactose set to 1, increased by a
factor of 2.8 for FL and to 35 by the added α1,3-substitution.12b
As noted above for the plant agglutinins, cell assays with the
chicken galectins also corroborated the changes in inhibitory
potency listed in Table 1. These experiments e.g. illustrated the
relative efficiency of the trivalent 6 to block binding of trCG-3
(0.1 mM sugar presented by 6 yielded a decrease from 51%/32.9
(control) to 32%/15.0 compared to 45%/19.7 for 1 mM lactose;
Fig. 4A–D) and the N-terminal domain of CG-8 (Fig. 4E and F).
In order to ensure that these results from CG-3 can be extrapo-
lated to the human orthologue we next ran experiments under
identical conditions with human Gal-3. Because glycoprotein
binding of the tandem-repeat-type Gal-4 had been reported to be
sensitive to ligand presentation by cyclic scaffolds (calixare-
nes,25d cyclodecapeptides29), we performed respective experi-
ments with this two-domain protein and its separate domains in
parallel.
relation to their high-affinity binding to ganglioside GM1
exposed on human neuroblastoma cells by perturbing the integ-
rity of microdomains.31 This switch-like impact on affinity,
together with similar effects on other levels given in detail pre-
viously,2 prompts the efforts to delineate detailed structure–
activity profiles for glycoclusters.
Proceeding from our previous reports on bivalent presen-
tation,6 we herein delineate special sensitivity of the tandem-
repeat and chimera-type galectins for the tested tri- and
tetravalent compounds when compared to the group of homodi-
meric proteins. Synthetic α1,3-substitution had even been found
to add to the discrimination between galectins-1 and -3 on the
level of cell binding.32 Evidently, the tested natural headgroup
elaboration could enhance the respective potency for galectin-3.
This result intimates the possibility of such tailoring to allow the
attainment of affinity differences between the chimera- and
tandem-repeat-type proteins and also within the latter group.
Along this line, the identification of distinct natural docking sites
for certain lectins, e.g. sulphated glycosphingolipids for galectin-
4,33 and of aglyconic extensions conferring selectivity gains
such as the β-naphthyl sulfone34 will help achieve stepwise pro-
gress. And here choosing the optimal geometry to match an
increase in valency can come into play, because the comparison
of inhibitory capacity of tetravalent clusters built with different
scaffolds25 (for scaffold development35) underlines the fact that
geometry can matter markedly. Our results on CG-3/CG-8 and
the combination of trivalency with headgroup tailoring encou-
rage further consideration of the feasibility of this proposal. The
same strategy of changing the headgroup could be applied to the
toxin, which tolerates α2,6-sialylation of lactose in sharp contrast
to the galectins, hereby precluding cross-reactivity of an anti-
toxin compound with the galectins. Giving direction to further
work, the detailed analysis of lectin specificity will continue to
provide inspirations for the design of the contact region. Equally
important, comparative analysis within and between lectin
families will be required to track down the most suited glyco-
cluster design to attain optimal selectivity.
Assaying inhibitory properties on human galectins
The obtained data document interspecies maintenance of sensi-
tivity for respective headgroup tailoring and valency increase in
the case of the chimera-type protein (Table 2). The results are
also in accord with the 2.8-fold affinity enhancement previously
reported for FL in a similar inhibition assay.12b Binding of Gal-4
proved to be rather susceptible to the presence of the tetravalent
compound 7, being less so towards the trivalent compound, so
that this structural design appears to hit tandem-repeat-type
galectins (Gal-4, CG-8) as well as the chimera-type Gal-3
(Table 2). The strong inhibitory potency of 7 on cell binding, in
comparison to the bi- and trivalent substances, is illustrated in
Fig. 5, again correlating rather well with the results from the
solid-phase assay. The monovalent association of the separate
domains of Gal-4 to glycoprotein glycans could also be effec-
tively blocked more strongly than in the case of the lectin
domain of Gal-3 (Table 2). Of note, set in relation to the plant
toxin, the case of dithiodigalactoside, which shows low-affinity
binding, if at all, to human galectins, exemplifies the possibility
for markedly different headgroup affinities despite presence of
galactose and thus enabling the targeting of the toxin with a
galactose-based compound while avoiding side effects that
would result from binding to galectins.30
Experimental section
General experimental
Unless otherwise noted, all commercially available compounds
were used as provided without further purification. Solvents for
chromatography were technical grade. Petroleum ether 40–60 °C
was used for column chromatography and thin layer chromato-
graphy (TLC). NMR spectra were recorded (25 °C) with
500 MHz spectrometer. The frequency is 500 MHz for 1H NMR
and 125 MHz for 13C NMR. Data are reported in the following
order: chemical shift (δ) in ppm; multiplicities are indicated s
(singlet), d (doublet), t (triplet), q (quartet), m (multiplet); coup-
ling constants (J) are given in Hertz (Hz). Chemical shifts are
reported relative to internal Me4Si in CDCl3 (d 0.0) or HOD for
D2O (d 4.72, 25 °C) for 1H and Me4Si in CDCl3 (d 0.0) or
CDCl3 (d 77.0) for 13C. 1H NMR signals were assigned with the
aid of COSY, 13C NMR signals using DEPT, gHSQCAD and/or
gHMBCAD. Low- and high-resolution mass spectra were in
positive and/or negative mode as indicated in each case. TLC
was performed on aluminium sheets precoated with silica gel
Conclusions
The molecular characterization of the counterreceptors for tissue
lectins is unlocking the virtues of spatial parameters for generat-
ing the high-level selectivity found in nature. In fact, their local
density can matter in different contexts up to the presentation of
target sites in membrane microdomains.2 The fundamental
importance of this property, that is a particular spatial organiz-
ation of carbohydrate recognition groups and not just their mere
presence, has recently been proven for galectins-1 and -3 in
6900 | Org. Biomol. Chem., 2012, 10, 6893–6907
This journal is © The Royal Society of Chemistry 2012

View Article Online
Fig. 3 Semilogarithmic representation of fluorescent surface staining
of human B-lymphoblastoid cells (Croco II) by VAA. The control value
(background) for the signal obtained by processing cells with the fluor-
escent indicator in the absence of lectin is given as grey-shaded curve
area, the 100%-value (lectin-dependent staining without presence of
inhibitor) as thick black line. Numbers for staining (percentage of posi-
tive cells/mean fluorescence intensity) are added to each panel in the
order of listing compound/concentration (from top to bottom) here. All
concentrations are given in terms of the sugar (free or conjugated to a
scaffold). A: staining parameters measured with increasing concen-
trations of VAA from 1 μg ml−1 to 2 μg ml−1, 4 μg ml−1, 5 μg ml−1 and
6 μg ml−1. B: inhibition of VAA staining (4 μg ml−1) by lactose concen-
trations of 20 mM, 4 mM and 1 mM relative to the 100%-value in the
absence of inhibitor. C: inhibition of staining by 1 mM lactose presented
by trivalent compound 4 relative to the 100%-control. D: inhibition of
staining by 0.5 mM lactose presented by tetravalent compound 7 and
2 mM lactose presented by bivalent compound 3.
Fig. 4 Semilogarithmic representation of fluorescent surface staining
of Chinese hamster ovary cells by proteolytically truncated CG-3 (A–D)
and the N-domain of CG-8, i.e. CG-8N (E, F) (for further details, please
see legend to Fig. 2). A–D: inhibition of staining with trCG-3 (2 μg
ml−1) by 1 mM lactose (A), 0.5 mM lactose presented by trivalent com-
pound 4 (B), 0.5 mM lactose presented by trivalent compound 5 (C) and
0.1 M sugar presented by trivalent compound 6 (D). E, F: inhibition of
staining with CG-8N (10 μg ml−1) by 4 mM and 0.5 mM lactose (E) as
well as by 0.5 mM 2′-fucosyllactose presented by trivalent compound 6
and 4 mM derivatized 2′-fucosyllactose (F).
and spots visualized by UV and charring with H2SO4–EtOH
(1 : 20), or cerium molybdate. Flash chromatography was carried
out with silica gel 60 (0.040–0.630 mm) and using a stepwise
solvent polarity gradient correlated with TLC mobility. CH2Cl2,
MeOH, toluene and THF reaction solvents were used as obtained
from a Pure Solv™ Solvent Purification System. Anhydrous
DMF, pyridine, and EtOH were used as purchased.
(dd, J = 5.3, 3.0 Hz, 2H), 7.71 (dd, J = 5.3, 3.0 Hz, 2H), 6.67
(d, J = 1.9 Hz, 2H), 6.52 (s, 1H), 4.79 (s, 2H), 4.64 (d, J = 2.1
Hz, 4H), 2.49 (t, J = 2.1 Hz, 2H); 13C NMR (125 MHz, CDCl3)
δ 167.9, 158.8, 138.7 (each C), 134.0 (CH), 132.1 (C), 123.4,
108.1, 101.6 (each CH), 78.2 (C), 75.7 (CH), 55.9 (CH2), 41.5
(CH2); HRMS-ESI: calcd for C21H16NO4[M + H]+, 346.1079;
found, 346.1087.
2-(3,5-Bis(prop-2-ynyloxy)benzyl)isoindoline-1,3-dione 9
To a mixture of potassium phthalimide (408 mg, 2.2 mmol) and
bromide 8 (500 mg, 1.8 mmol) in toluene (15 mL) was added
18-crown-6 (49 mg, 0.18 mmol). The mixture was heated at
100 °C for 6 h with stirring under N2 and then water was added.
The organic layer was separated and aqueous layer was extracted
with CH2Cl2. The organic portions were combined, then dried
over Na2SO4 and the solvent was evaporated under reduced
pressure. Silica gel chromatography (petroleum ether–EtOAc,
gradient elution, 3 : 1 to 1 : 1) afforded 9 (454 mg, 73%) as an
white amorphous solid; 1H NMR (500 MHz, CDCl3) δ 7.85
2-(3,5-Bis(3,5-bis(prop-2-ynyloxy)benzyloxy)benzyl)isoindoline-
1,3-dione 11
Compound 11 (56%) was prepared from bromide 10 as described
in the preparation of 9 as white solid after column chromato-
graphy (petroleum ether–ethyl acetate, 6 : 1); 1H NMR
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 6893–6907 | 6901

View Article Online
Table 2 IC50-values of the mono- to tetravalent lactosidesa and free lactose (Lac) for blocking binding of biotinylated human galectins to surface-
immobilized ASF (in mM)
Lectin inhibitor
Gal-3 (1 μg ml−1
)
trGal-3 (10 μg ml−1
)
Gal-4 (5 μg ml−1
)
Gal-4N (5 μg ml−1
)
Gal-4C (10 μg ml−1
)
1
2
3
4
5
6
7
Lac
0.16 (2.5)
0.11 (3.6)
0.14 (2.9)
0.13 (3.1)
0.08 (5.0)
0.02 (20)
0.013 (31)
0.4 (1)
0.12 (2.5)
0.08 (3.8)
0.32 (1.1)
0.09 (2.8)
0.14 (1.8)
0.10 (2.5)
0.07 (3.6)
0.05 (5.0)
0.02 (13)
0.008 (31)
0.25 (1)
2.4 (1.3)
2.6 (1.2)
0.9 (3.3)
1.2 (2.5)
1.4 (2.1)
0.5 (6.0)
0.16 (19)
3 (1)
0.13 (3.1)
0.18 (2.2)
0.18 (2.2)
0.24 (1.7)
0.21 (1.9)
0.06 (6.7)
0.03 (13)
0.4 (1)
0.22 (1.4)
b
0.06 (5.0)
0.27 (1.1)
0.3 (1)
a For structures see Chart 1; titrations were performed using a fixed glycoprotein quantity for coating (0.5 μg per well) with eight concentrations of
sugar in duplicates and up to three independent series, reaching an upper limit of 16.2% for the standard deviation (for exemplary titration curves, see
Fig. 1); the concentration is always given for lactose, free in solution or conjugated to a scaffold. b Tendency for stimulation at concentrations above
1 mM, numbers in brackets denote relative potency.
filtered, and the solvent was removed under reduced pressure.
Silica-gel chromatography (CH2Cl2–CH3OH, gradient elution,
50 : 1 to 10 : 1) gave the title compound 16 as a white solid
(234 mg, 95%); Rf 0.2 (CH2Cl2–CH3OH, 10 : 1); 1H NMR
(500 MHz, CDCl3) δ 6.58 (t, J = 5.4 Hz, 3H, NH), 3.29 (q, J =
6.6 Hz, 6H), 2.41 (t, J = 6.5 Hz, 6H), 2.26–2.18 (m, 12H), 1.96
(t, J = 2.6 Hz, 3H), 1.79–1.71 (m, 6H), 1.64 (p, J = 6.5 Hz, 6H),
1.60–1.52 (m, 6H); 13C NMR (125 MHz, CDCl3) δ 173.0 (C),
84.1 (C), 68.6 (CH), 51.4, 37.9, 36.0, 28.0, 26.9, 24.9, 18.2
(each CH2); HRMS-ESI: calcd for C30H49N4O3[M + H]+,
513.3805; found, 513.3808.
Fig. 5 Semilogarithmic representation of fluorescent surface staining
of human pancreatic carcinoma cells (Capan-1), reconstituted for
expression of the tumor suppressor p16INK4a, by human galectin-4 (for
further details, please see legend to Fig. 2). A, B: inhibition of staining
with 10 μg ml−1 lectin by 0.5 mM lactose as well as by 0.05 mM lactose
presented by the trivalent compound 5 and 0.1 mM lactose in bivalent
compound 3 (A) as well as by 20 μM and 7.5 μM lactose presented by
the tetravalent compound 7 (B).
2-(3,5-Bis((1-(2,3,4,6-tetra-O-acetyl-β-D-galactopyranosyl-(1 →
4)-2,3,6-tri-O-acetyl-β-D-glucopyranosyl)-1H-1,2,3-triazol-4-yl)
methoxy)benzyl)isoindoline-1,3-dione 18
Compound 17 (154 mg, 0.23 mmol) was dissolved in CH3OH–
H2O (5 : 1, 12 mL), then compound 9 (40 mg, 0.12 mmol),
sodium ascorbate (9.2 mg in 1 mL H2O, 0.047 mmol) and
CuSO4 (3.7 mg in 1 mL H2O, 0.023 mmol) were subsequently
added and the mixture was sonicated for 2 h, whereafter the
solvent was removed and the residue was partitioned between
CH2Cl2 (100 mL) and water (15 mL). The organic portion was
washed by water (15 mL ×2), dried by Na2SO4 and concen-
trated. Silica gel chromatography (CH2Cl2–CH3OH, gradient
elution, 80 : 1 to 70 : 1 to 60 : 1) gave 18 as a white foam
(178 mg, 92%); [α]D20 −20.8 (c 1.0, CHCl3); 1H NMR
(500 MHz, CDCl3) δ 7.86 (dd, J = 5.4, 3.0 Hz, 2H), 7.80
(s, 2H), 7.73 (dd, J = 5.4, 3.0 Hz, 2H), 6.67 (d, J = 2.1 Hz, 2H),
6.52 (t, J = 2.0 Hz, 1H), 5.83 (d, J = 9.0 Hz, 2H, H-1),
5.42–5.40 (m, 4H), 5.37 (d, J = 2.9 Hz, 2H), 5.17–5.11 (m, 6H,
H-2′, OCH2), 4.98 (dd, J = 10.4, 3.4 Hz, 2H), 4.79 (s, 2H), 4.54
(d, J = 7.9 Hz, 2H, H-1′), 4.48 (d, J = 10.8 Hz, 2H), 4.17–4.08
(m, 6H), 4.00–3.94 (m, 2H), 3.95–3.88 (m, 4H), 2.17 (s, 6H),
2.10 (s, 6H), 2.08 (s, 6H), 2.06 (2s, 12H), 1.98 (s, 6H), 1.84
(s, 6H); 13C NMR (125 MHz, CDCl3) δ 170.3, 170.2, 170.1,
170.0, 169.5, 169.1 (2s), 168.0, 159.5, 144.5, 138.9 (each C),
134.1 (CH), 132.1 (C), 123.4, 121.3, 107.8, 101.13 (C-1′),
101.06, 85.6 (C-1), 75.9, 75.7, 72.6, 70.91, 70.86, 70.5, 69.0,
66.6 (each CH), 61.9, 61.8, 60.8, 41.5 (each CH2), 20.8, 20.71,
20.67, 20.64, 20.63, 20.5, 20.2 (each CH3); HRMS-ESI: calcd
for C73H86N7O38[M + H]+, 1668.5012; found, 1668.5001.
(500 MHz, CDCl3) δ 7.85 (dd, J = 5.4, 3.0 Hz, 2H), 7.72 (dd,
J = 5.4, 3.0 Hz, 2H), 6.65 (dd, J = 6.1, 2.2 Hz, 6H), 6.54 (t, J =
2.2 Hz, 2H), 6.47 (t, J = 2.2 Hz, 1H), 4.96 (s, 4H), 4.76 (s, 2H),
4.67 (d, J = 2.4 Hz, 8H), 2.52 (t, J = 2.4 Hz, 4H); 13C NMR
(125 MHz, CDCl3) δ 168.0, 159.9, 158.8, 139.3, 138.6 (each C),
134.0 (CH), 132.1 (C), 123.4, 107.6, 106.8, 101.9, 101.6 (each
CH), 78.3 (C), 75.7 (CH), 69.8, 56.0, 41.6 (each CH2);
HRMS-ESI: calcd for C41H31NO8Na[M + Na]+, 688.1947;
found, 688.1947.
N,N′,N′′-(3,3′,3′′-Nitrilotris(propane-3,1-diyl))trihept-6-ynamide
16
1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride
(EDC·HCl; 606 mg, 3.2 mmol) was added to a mixture of 15
(0.21 mL, 1.6 mmol), tris(3-aminopropyl)amine 14 (0.1 mL,
0.48 mmol), 1-hydroxybenzotriazole and triethylamine (442 μL,
3.2 mmol) in THF (25 mL) and the mixture was stirred at
room temperature overnight. THF was removed under reduced
pressure. The residue was dissolved in CH2Cl2 (150 mL) and
washed with saturated NaHCO3 and brine, dried (Na2SO4),
6902 | Org. Biomol. Chem., 2012, 10, 6893–6907
This journal is © The Royal Society of Chemistry 2012

View Article Online
(3,5-Bis((1-(β-D-galactopyranosyl-(1 → 4)-β-D-glucopyranosyl)-
1H-1,2,3-triazol-4-yl)-methoxy)phenyl)methanamine 3
(m, 8H), 6.25 (s, 1H), 5.65 (d, J = 8.7 Hz, 4H, H-1), 4.81 (s,
8H), 4.62 (s, 4H), 4.47 (d, J = 7.4 Hz, 4H, H-1′), 3.98 (t, J = 7.4
Hz, 4H), 3.94 (d, J = 2.9 Hz, 4H), 3.92–3.68 (m, 32H), 3.67
(dd, J = 10.0, 2.9 Hz, 4H), 3.59 (dd, J = 7.4, 10.0 Hz, 4H), 3.45
(s, 2H); 13C NMR (125 MHz, D2O) δ 159.2, 158.8, 143.2, 139.4
(each C), 124.1, 106.7, 102.9, 100.8, 87.2, 77.6, 77.5, 75.3,
74.5, 72.5, 71.9, 70.9, 68.5 (each CH), 61.0, 60.8, 59.8, 44.2
(each CH2); HRMS-ESI: calcd for C81H114N13O46[M + H]+,
2004.6981; found, 2004.6969.
Compound 18 (100 mg, 60.0 μmol) and ethylenediamine
(0.5 mL) in anhydrous ethanol (5 mL) was heated at reflux.
After 6 h, the mixture was cooled to room temperature. The pre-
cipitate was filtered, washed with methanol (0.5 mL ×3) and
dried under vacuum to give 3 as white solid (39 mg, 68%); [α]D20
8.8 (c 1.0, H2O); 1H NMR (500 MHz, D2O) δ 8.33 (s, 2H), 6.71
(s, 2H), 6.65 (s, 1H), 5.80 (d, J = 9.0 Hz, 2H, H-1), 5.28 (s, 4H),
4.52 (d, J = 7.8 Hz, 2H, H-1′), 4.07 (t, J = 9.0 Hz, 2H),
3.99–3.75 (m, 20H), 3.70 (dd, J = 10.0, 3.4 Hz, 2H), 3.59 (dd,
J = 10.0, 7.8 Hz, 2H); 13C NMR (125 MHz, D2O) δ 158.6 (C),
144.7 (C), 143.5 (C), 124.4, 107.7, 102.8, 101.5, 87.2, 77.7,
77.3, 75.3, 74.5, 72.4, 71.9, 70.9, 68.5 (each CH), 61.3, 61.0,
59.7, 44.4 (each CH2); HRMS-ESI: calcd for C37H56N7O22[M +
H]+, 950.3478; found, 950.3441.
1,3,5-Tris((1-(2,3,4,6-tetra-O-acetyl-β-D-galactopyranosyl-(1→4)-
2,3,6-tri-O-acetyl-β-D-glucopyranosyl)-1H-1,2,3-triazol-4-yl)-
methoxy)benzene 20
Compound 17 (200 mg, 0.30 mmol) was dissolved in CH3OH–
H2O (4 : 1, 15 mL), then 1,3,5-tris(alkynyloxy)benzene15 13
(17 mg, 0.10 mol), sodium ascorbate (24 mg dissolved in 1 mL
H2O, 0.12 mmol) and CuSO4 (10 mg dissolved in 1 mL H2O,
0.06 mmol) were subsequently added and the mixture was
stirred at 45 °C overnight, after which the solvent was removed
and the residue was partitioned between CH2Cl2 (50 mL) and
water (15 mL). The organic phase was washed by water (15 mL
×2), dried (Na2SO4) and concentrated. Silica gel chromatography
(CH2Cl2–CH3OH, gradient elution, 80 : 1 to 60 : 1 to 55 : 1) gave
20 a white amorphous solid (148 mg, 66%); Rf 0.45 (CH2Cl2–
CH3OH, 20 : 1); [α]D20 −22.8 (c 1.0, CHCl3); 1H NMR
(500 MHz, CDCl3): δ 7.81 (s, 3H), 6.27 (s, 3H), 5.84 (d, J = 9.0
Hz, 3H, H-1), 5.44–5.38 (m, 6H), 5.37 (d, J = 3.3 Hz, 3H),
5.17–5.10 (m, 9H), 4.98 (dd, J = 10.4, 3.4 Hz, 3H), 4.54 (d, J =
7.9 Hz, 3H, H-1′), 4.49 (d, J = 10.9 Hz, 3H), 4.17–4.08 (m, 9H),
3.99–3.89 (m, 9H), 2.17 (s, 9H), 2.11 (s, 9H), 2.08 (s, 9H), 2.07
(s, 9H), 2.06 (s, 9H), 1.98 (s, 9H), 1.86 (s, 9H); 13C NMR
(125 MHz, CDCl3) δ: 170.3, 170.2, 170.1, 170.0, 169.5, 169.13,
169.06, 160.0, 144.5 (each C), 121.3, 101.1 (CH-1′), 95.1, 85.6
(CH-1), 76.0, 75.6, 72.6, 70.90, 70.86, 70.5, 69.1, 66.6 (each
CH), 62.0 (CH2), 61.8 (CH2), 60.8 (CH2), 20.8, 20.71, 20.67,
20.64, 20.63, 20.5, 20.2 (each CH3); HRMS-ESI: calcd for
C93H118N9O54[M + H]+, 2224.6764; found, 2224.6799.
2-(3,5-Bis(3,5-bis((1-(2,3,4,6-tetra-O-acetyl-β-D-galactopyranosyl-
(1 → 4)-2,3,6-tri-O-acetyl-β-D-glucopyranosyl)-1H-1,2,3-triazol-
4-yl-methyloxy)-benzyloxy)-benzyl)-isoindoline-1,3-dione 19
Compound 17 (200 mg, 0.30 mmol) was dissolved in CH3OH
(8 mL), to which a solution of compound 11 (50 mg,
0.076 mmol) in CH2Cl2 (2 mL) was added. Then solutions of
CuSO4 (4.8 mg dissolved in 1 mL H2O, 30.0 μmol) and sodium
ascorbate (12 mg dissolved in 1 mL H2O, 60.0 μmol) were sub-
sequently added and the mixture was sonicated for 2 h, after
which it was stirred at 40 °C overnight. Thereafter the solvent
was removed and the residue was partitioned between CH2Cl2
(100 mL) and water (15 mL). The organic phase was washed by
water (15 mL ×2), dried (Na2SO4) and the solvent was removed
under reduced pressure. Silica gel chromatography (CH2Cl2–
CH3OH, gradient elution, 80 : 1 to 60 : 1 to 50 : 1) gave 19 as a
white solid (172 mg, 69%); [α]2D0 −28.0 (c 0.5, CHCl3); 1H
NMR (500 MHz, CDCl3) δ 7.84 (dd, J = 5.4, 3.0 Hz, 2H), 7.82
(s, 4H), 7.71 (dd, J = 5.4, 3.0 Hz, 2H), 6.66–6.65 (m, 6H), 6.54
(s, 2H), 6.50 (s, 1H), 5.86 (d, J = 9.0 Hz, 4H, H-1), 5.46–5.39
(m, 8H), 5.37 (d, J = 3.3 Hz, 4H), 5.16 (s, 8H), 5.14 (dd, J =
10.4, 8.0 Hz, 4H), 4.99 (dd, J = 10.4, 3.4 Hz, 4H), 4.96 (s, 4H),
4.77 (s, 2H), 4.55 (d, J = 8.0 Hz, 4H, H-1′), 4.49 (d, J = 11.1
Hz, 4H), 4.18–4.08 (m, 12H), 4.02–3.89 (m, 12H), 2.17 (s,
12H), 2.10 (s, 12H), 2.07 (s, 12H), 2.063 (s, 12H), 2.061 (s,
12H), 1.98 (s, 12H), 1.85 (s, 12H); 13C NMR (125 MHz,
CDCl3) δ 170.3, 170.2, 170.1, 170.0, 169.5, 169.13, 169.07,
168.0, 159.9, 159.5, 144.6, 139.5, 138.6 (each C), 134.1 (CH),
132.0 (C), 123.4, 121.4, 107.6, 106.6, 101.4, 101.1, 85.6, 76.0,
75.7, 72.6, 70.9, 70.8, 70.5 (each CH), 69.8 (CH2), 69.1 (CH),
66.6 (CH), 62.0, 61.8, 60.8, 41.6 (each CH2), 20.8, 20.72,
20.67, 20.65, 20.6, 20.5, 20.2 (each CH3); HRMS-ESI: calcd for
1,3,5-Tris((1-(β-D-galactopyranosyl-(1→4)-β-D-glucopyranosyl)-
1H-1,2,3-triazol-4-yl) methoxy)benzene 4
Compound 20 (80 mg, 36.0 μmol) was dissolved in methanol
(5 mL). A catalytic amount of NaOMe (0.1 mL of a 0.2 M sol-
ution in MeOH) was added to the solution and the resulting
mixture was stirred for 1 h at room temperature. Amberlite
IR-120 (plus) was added to neutralize (pH = 7), whereafter
the resin was removed by filtration and washed with water. The
solvent was removed and the residue was purified by BioGel
P-2 gel filtration column to give 4 as an amorphous solid
(40 mg, 83%); [α]D20 6.2 (c 0.9, H2O); 1H NMR (500 MHz,
D2O) δ 8.32 (s, 3H), 6.43 (s, 3H), 5.81 (d, J = 9.0 Hz, 3H, H-1),
5.26 (s, 6H), 4.52 (d, J = 7.8 Hz, 3H, H-1′), 4.08 (t, J = 9.0 Hz,
3H), 3.96–3.95 (m, 6H), 3.90–3.83 (m, 12H), 3.81–3.76
(m, 9H), 3.70 (dd, J = 10.0, 3.4 Hz, 3H), 3.60 (dd, J = 10.0,
7.8 Hz, 3H); 13C NMR (125 MHz, D2O) δ 159.4, 143.3 (each
C), 124.3, 102.9, 95.8, 87.2, 77.6, 77.3, 75.3, 74.5, 72.4, 71.9,
70.9, 68.5 (each CH), 61.1, 61.0, 59.7 (each CH2); HRMS-ESI:
C
145H171N13O76[M + Cl]−, 3344.9604; found, 3344.9629.
(3,5-Bis(3,5-bis((1-(β-D-galactopyranosyl-(1→4)-β-D-
glucopyranosyl)-1H-1,2,3-triazol-4-yl-methoxy)-benzyloxy)-
phenyl)methanamine 7
Compound 7 was prepared (81%, amorphous solid) from 19 as
described for the preparation of compound 3; [α]2D0 6.0 (c 1.0,
H2O); 1H NMR (500 MHz, D2O) δ 8.06 (s, 4H), 6.44–6.34
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 6893–6907 | 6903

View Article Online
calcd for C51H75N9O33Na[M
1364.4332.
+
Na]+, 1364.4365; found,
followed by the addition of Tf2O (219 μL, 1.27 mmol) and
2,4,6-tri-tert-butylpyrimidine (TTBP, 598 mg, 2.34 mmol). Then
the temperature was increased gradually from −78 °C to 0 °C
within 2 h and the mixture was stirred at 0 °C for an additional
period of 4 h. The reaction was quenched with triethylamine
(3 mL) and diluted with CH2Cl2. The reaction mixture was
filtered and washed sequentially with saturated Na2S2O3, satu-
rated NaHCO3, H2O, and brine, dried (Na2SO4), filtered, and
the solvent removed under reduced pressure. Chromatography
(petroleum ether–EtOAc, gradient elution, 7 : 1 to 6 : 1) gave 24
as a white foam (1.63 g, 70%), Rf 0.5 (petroleum ether–EtOAc,
N,N′,N′′-(3,3′,3′′-Nitrilotris(propane-3,1-diyl))tris(5-(1-(2,3,4,6-
tetra-O-acetyl-β-D-galactopyranosyl-(1 → 4)-2,3,6-tri-O-acetyl-
β-D-glucopyranosyl)-1H-1,2,3-triazol-4-yl) pentanamide) 21
Compound 21 was prepared from lactoside 17 and compound 16
as described for the preparation of 20, using ultrasonic radiation
instead of heating. The title compound was obtained (92%) as an
amorphous solid after chromatography (CH2Cl2–CH3OH, gradi-
ent elution, 80 : 1 to 30 : 1); [α]2D0 −14.4 (c 1.0, CHCl3);
1H NMR (500 MHz, CDCl3) δ 7.53 (s, 3H), 7.16 (s, 3H, NH),
5.83 (d, J = 8.4 Hz, 3H, H-1), 5.44–5.34 (m, 6H), 5.37 (d, J =
3.1 Hz, 3H), 5.13 (dd, J = 10.2, 8.0 Hz, 3H), 4.99 (dd, J = 10.2,
3.3 Hz, 3H), 4.58 (d, J = 7.9 Hz, 3H, H-1′), 4.51 (d, J = 11.9
Hz, 3H), 4.18–4.08 (m, 9H), 4.05–3.86 (m, 9H), 3.31 (d, J = 4.8
Hz, 6H), 2.96 (m, 6H), 2.72 (t, J = 6.5 Hz, 6H), 2.25 (s, 6H),
2.17 (s, 9H), 2.11 (s, 9H), 2,074 (s, 9H), 2.066 (s, 9H), 2.06 (s,
9H), 1.98 (s, 9H), 1.89 (m, 6H), 1.86 (s, 9H), 1.69 (s, 12H);
13C NMR (125 MHz, CDCl3) δ 173.9, 170.4, 170.3, 170.1,
170.0, 169.5, 169.2, 169.1, 148.3 (each C), 119.6, 101.1
(CH-1′), 85.4 (CH-1), 75.9, 75.6, 72.7, 70.9, 70.8, 70.5, 69.1,
66.6 (each CH), 61.8, 60.8, 50.9, 36.6, 35.9, 28.6, 25.2, 25.0,
24.5 (each CH2), 20.9, 20.74, 20.67, 20.65, 20.64, 20.5, 20.3
(each CH3); HRMS-ESI: calcd for C108H154N13O54[M + H]+,
2496.9704; found, 2496.9739.
1
2.5 : 1). H-NMR (500 MHz, CDCl3): δ 8.04 (d, J = 7.5 Hz,
2H), 7.57 (t, J = 7.5 Hz, 1H), 7.44 (t, J = 7.5 Hz, 2H), 7.40–7.24
(m, 15H), 5.59 (d, J = 3.3 Hz, 1H, H-1′′), 4.98 (d, J = 11.6 Hz,
1H), 4.87 (d, J = 11.9 Hz, 1H), 4.75–4.73 (m, 3H), 4.70 (d, J =
8.0 Hz, 1H, H-1′), 4.65 (d, J = 11.6 Hz, 1H), 4.57–4.49 (m, 3H),
4.34 (d, J = 5.9 Hz, 1H, H-1), 4.25 (dd, J = 10.9, 5.9 Hz, 1H),
4.21 (t, J = 6.0 Hz, 1H), 4.16 (dd, J = 5.6, 2.0 Hz, 1H),
4.09–4.00 (m, 7H), 3.90 (dd, J = 8.3, 6.5 Hz, 1H), 3.75 (dd, J =
8.0, 6.7 Hz, 1H), 3.66 (d, J = 1.1 Hz, 1H), 3.36 (s, 3H), 3.34 (s,
3H), 1.49 (s, 3H), 1.43 (s, 3H), 1.37 (s, 6H), 1.33 (s, 3H), 1.27
(s, 3H), 1.11 (d, J = 6.5 Hz, 3H); 13C NMR (125 MHz, CDCl3)
δ 166.3 (C), 139.3 (C), 139.1 (C), 138.9 (C), 133.1, 130.0,
129.7, 128.42, 128.35, 128.3, 128.13, 128.12, 127.6, 127.43,
127.40, 127.31, 127.27 (each CH), 110.3 (C), 110.0 (C), 108.6
(C), 105.1 (CH, C-1), 101.3 (CH, C-1′), 95.2 (CH, C-1′′), 80.2,
79.2, 78.1, 77.7, 77.5, 76.5, 75.4, 75.0 (each CH), 74.8 (CH2),
74.1 (CH), 73.8 (CH), 73.2 (CH2), 72.7 (CH2), 70.9 (CH), 66.5
(CH), 65.2 (CH2), 63.8 (CH2), 56.0, 52.9, 29.7, 27.8, 27.2, 26.9,
26.5, 25.2, 16.9 (each CH3); HRMS-ESI: calcd for
C57H72O17Na[M + Na]+, 1051.4667; found, 1051.4664.
N,N′,N′′-(3,3′,3′′-Nitrilotris(propane-3,1-diyl))tris(5-(1-(β-D-
galactopyranosyl-(1 → 4)-β-D-glucopyranosyl)-1H-1,2,3-triazol-
4-yl))pentanamide) 5
Compound 21 (100 mg, 40 μmol) was dissolved in methanol
(5 mL) to which a catalytic amount of NaOMe (0.5 mL of a 0.2 M
solution in MeOH) was added and the resulting solution was
stirred for 4 h at room temperature. The solvent was removed
and the residue was purified by BioGel P-2 gel filtration column
to give 5 as an amorphous solid (50 mg, 79%); [α]2D0 −5.6
O-(2,3,4-Tri-O-acetyl-α-L-fucopyranosyl)-(1 → 2)-O-(3,4,6-tri-
O-acetyl-β-D-galactopyranosyl)-(1 → 4)-1,2,3,6-tetra-O-acetyl-
α/β-D-glucopyranose 2619b
To compound 24 (2.12 g, 2.33 mmol) in anhydrous MeOH,
NaOMe (0.5 mL of a 2 M solution in MeOH) was added and the
resulting mixture was stirred for 1 h at room temperature. Amber-
lite IR-120 (plus) was added to neutralize (pH = 7), and the
solvent was removed under reduced pressure to afford a colour-
less oil. This residue was dissolved in aq 60% acetic acid
(30 mL) and heated for 6 h at 60 °C. The reaction mixture was
diluted with toluene and the volatile components removed. Then
to a solution of the residue in THF–H2O–AcOH (4 : 2 : 1,
14 mL), 10% Pd–C (50 mg) was added. The suspension was
stirred under an atmosphere of hydrogen for 2 days at ambient
temperature. When the reaction was completed, the mixture was
filtered over Celite and concentrated. Toluene (3 × 30 mL) was
evaporated from the residue to remove the acetic acid and water.
Then fucosyllactose 119 was dissolved in pyridine–acetic anhy-
dride (2 : 1, 30 mL) and then stirred overnight under an atmos-
phere of nitrogen at ambient temperature. The solvent was then
removed and the residue was partitioned between CH2Cl2
(100 mL) and water (25 mL). The organic phase was washed
with water (25 mL ×2), dried (Na2SO4) and the solvent was
removed. Chromatography (petroleum ether–EtOAc, gradient
elution, 4 : 1 to 3 : 1) gave 26 as a white foam (1.67 mg, 79%);
1
(c 1.0, H2O); H NMR (500 MHz, D2O) δ 8.02 (s, 3H), 5.74
(d, J = 9.2 Hz, 3H, H-1), 4.53 (d, J = 8.0 Hz, 3H, H-1′), 4.04
(t, J = 9.2 Hz, 3H, H-2), 3.97–3.76 (m, 27H), 3.70 (dd, J = 9.6,
2.9 Hz, 3H), 3.60 (dd, J = 9.6, 8.0 Hz, 3H), 3.14 (t, J = 6.6 Hz,
6H), 2.74 (t, J = 6.8 Hz, 6H), 2.43–2.40 (m, 6H), 2.25 (t, J = 6.7
Hz, 6H), 1.65–1.57 (m, 18H); 13C NMR (125 MHz, D2O)
δ 176.4, 148.3 (each C), 122.1 (CH), 102.8 (CH, C-1′), 87.1
(CH, C-1), 77.6, 77.3, 75.3, 74.5, 72.5, 71.9, 70.9, 68.5 (each
CH), 61.0, 59.7, 50.3, 37.4, 35.3, 27.7, 24.9, 24.7, 24.1 (each
CH2); HRMS-ESI: calcd for C66H111N13O33Na[M + Na]+,
1636.7305; found, 1636.7290.
O-(2,3,4-Tri-O-benzyl)-α-L-fucopyranosyl)-(1 → 2)-O-(6-O-
benzoyl-3,4-O-isopropylidene-β-D-galactopyranosyl)-(1 → 4)-
2,3:5,6-di-O-isopropylidene-D-glucose dimethyl acetal 24
A mixture of fucosyl donor 2218 (1.26 g, 2.34 mmol), disacchar-
ide building block 2319 (1.30 g, 2.12 mmol), BSP (252 mg,
1.17 mmol) and 4 Å MS was stirred in dry CH2Cl2 (25 mL) at
room temperature for 1 h under N2. It was cooled to −78 °C,
6904 | Org. Biomol. Chem., 2012, 10, 6893–6907
This journal is © The Royal Society of Chemistry 2012

View Article Online
(S)-2-Amino-3-(1-(α-L-fucopyranosyl-(1 → 2)-β-D-galacto-
pyranosyl-(1 → 4)-β-D-glucopyranosyl)-1H-1,2,3-triazol-4-yl)-
Rf 0.35 (petroleum ether–EtOAc, 1.5 : 1), which were a mixture
of anomers (β : α = 1 : 1); selected H NMR (500 MHz, CDCl3)
1
propanoic acid 2
data for the β anomer: δ 5.69 (d, J = 8.2 Hz, 1H, H-1), 5.37 (d,
J = 3.4 Hz, 1H, H-1′′), 5.20 (t, J = 9.5 Hz, 1H), 5.09 (dd, J =
9.7, 8.3 Hz, 1H); selected 1H NMR data for the α anomer:
δ 6.30 (d, J = 3.7 Hz, 1H, H-1), 5.41 (t, J = 9.9 Hz, 1H), 5.37
(d, J = 3.5 Hz, 1H, H-1′′), 5.05 (dd, J = 10.4, 3.7 Hz, 1H); over-
To a solution of glycoside 27 (106 mg, 0.12 mmol) and Fmoc-
protected L-propargylglycine (41 mg, 0.12 mmol)22 in tert-
butanol (2 mL) and water (4 ml) was added a solution of CuSO4
(0.04 M, 0.6 mL, 0.024 mmol) and sodium ascorbate (0.08 M,
0.6 mL) in H2O. The mixture was stirred overnight, water
(6 mL) was added, and the product was extracted with CH2Cl2
(2 × 25 mL). The combined organic layers were washed with
brine, dried over Na2SO4, and evaporated under reduced
pressure. Chromatography using (CH2Cl2–MeOH, 15 : 1 to
10 : 1) give the intermediate as a white foam 131 mg (90%),
Rf 0.2 (CH2Cl2–MeOH, 15 : 1). The residue (131 mg, 0.11 mmol)
was treated with 20% piperidine in DMF (v/v, 5 mL), then
stirred for 20 min and solvent was finally removed under
reduced pressure. The crude product was dissolved in MeOH
(5 mL), and NaOMe (2 M, 0.05 mL) was added. The mixture
was stirred for 1 h. The mixture was concentrated and passed
through BioGel P-2 gel filtration column with water to give 2
(45 mg, 67%) as an amorphous solid; [α]2D0 −48.8 (c 0.5, H2O);
1H NMR (500 MHz, D2O) δ 8.06 (s, 1H), 5.75 (d, J = 9.2 Hz,
1H, H-1), 5.35 (d, J = 3.1 Hz, 1H, H-1′′), 4.61 (d, J = 7.8 Hz,
1H, H-1′), 4.27 (q, J = 6.7 Hz, 1H, H-4′′), 4.07 (t, J = 9.2 Hz,
1H, H-2), 4.01–3.97 (m, 2H), 3.93–3.90 (m, 2H), 3.87–3.718
(m, 10H), 3.67–3.64 (m, 1H), 3.16 (dd, J = 14.8, 5.2 Hz, 1H),
3.08 (dd, J = 14.8, 6.9 Hz, 1H), 1.28 (d, J = 6.6 Hz, 3H);
13C NMR (125 MHz, D2O) δ 179.8 (C), 144.3 (C), 123.1,
100.2, 99.3, 87.3, 78.1, 76.2, 75.2, 74.9, 74.3, 73.5, 72.0, 71.6,
69.6, 69.1, 68.1, 66.9 (each CH), 61.1, 59.8 (each CH2), 55.5
(CH), 29.6 (CH2), 15.2 (CH3). HRMS-ESI: calcd for
C23H38N4O16Na[M + Na]+, 649.2181; found, 649.2176.
1
lapped H NMR data for both β and α anomer: 5.31–5.27 (m,
4H), 5.18–5.12 (m, 2H), 5.00–4.96 (m, 4H), 4.50–4.38 (m, 6H),
4.27 (dd, J = 12.3, 5.3 Hz, 2H), 4.18–4.12 (m, 2H), 4.10–4.03
(m, 3H), 3.91–3.80 (m, 7H), 2.18–1.97 (19s, 60H), 1.24–1.21
(2d, J = 6.5 Hz, 6H); mixture 13C NMR (125 MHz, CDCl3) data
for both α and β anomer: δ 171.1, 170.7, 170.65, 170.6, 170.5,
170.3 (2s), 170.1, 170.0 (2s), 169.8, 169.7, 169.6, 169.3, 168.8,
168.7 (each C), 100.2, 99.9, 95.7, 73.9, 73.8, 73.7, 73.4, 73.3,
71.7, 71.11, 71.09, 70.9, 70.8, 70.7, 70.3, 69.2, 69.1, 68.1, 68.0,
67.5, 67.3, 67.0, 65.0, 64.9 (each CH), 62.1, 61.9, 61.0, 60.8
(each CH2), 21.1, 21.0, 20.9, 20.8, 20.7, 20.65, 20.6, 20.5, 15.6,
15.4 (each CH3); selected 13C NMR data for the α anomer:
δ 89.0 (CH, C-1); Selected 13C NMR data for the β-anomer:
δ 91.5 (CH, C-1); HRMS-ESI: calcd for C38H52N3O25Na[M +
Na]+, 931.2695; found, 931.2686.
O-(2,3,4-Tri-O-acetyl-α-L-fucopyranosyl)-(1 → 2)-O-(3,4,6-tri-O-
acetyl-β-D-galactopyran osyl)-(1 → 4)-2,3,6-tri-O-acetyl-β-D-
glucopyranosyl azide 27
Compound 26 (1.35 g, 1.49 mmol) was dissolved in CH2Cl2
(30 mL, anhydrous) under an atmosphere of N2. To this solution
was added TMSN3 (0.54 mL, 4.46 mmol) followed by the drop-
wise addition of SnCl4 (88 μL, 0.74 mmol). After 20 h, the solu-
tion was diluted with CH2Cl2, quenched by the addition of
saturated NaHCO3 solution (10 mL) and stirred for further
30 min. The biphasic solution was then filtered through celite
and the organic layer extracted, washed with saturated NaHCO3
(20 mL ×2), H2O (20 mL ×2), dried (Na2SO4) and the solvent
was removed under reduced pressure. Chromatography (petro-
leum ether–EtOAc = 3 : 1) gave 27 as a white foam (1.15 g,
87%), Rf 0.57 (petroleum ether–EtOAc, 1 : 1); 1H NMR
(500 MHz, CDCl3) data for the β anomer: δ 5.38 (d, J = 3.0 Hz,
1H, H-1′′), 5.30 (dd, J = 12.5, 3.0 Hz, 2H), 5.15–5.12 (m, 2H),
5.00–4.96 (m, 2H), 4.92 (t, J = 9.5 Hz, 1H), 4.64 (d, J = 9.0 Hz,
1H, H-1), 4.54 (d, J = 12.0 Hz, 1H), 4.41–4.37 (m, 2H), 4.29
(dd, J = 12.0, 6.0 Hz, 1H), 4.15 (dd, J = 11.1, 6.5 Hz, 1H), 4.08
(dd, J = 11.1, 7.0 Hz, 1H), 3.88–3.81 (m, 3H), 3.77–3.74
(m, 1H), 2.16 (2s, 6H), 2.12 (s, 3H), 2.09 (s, 3H), 2.08 (s, 3H),
2.06 (s, 3H), 2.00 (s, 3H), 1.98–1.97 (2s, 6H), 1.21 (d, J =
6.5 Hz, 3H); Selected 1H NMR data for the α anomer: δ 5.35 (d,
J = 3.0 Hz, 1H), 5.21 (t, J = 9.2 Hz, 1H), 5.10 (d, J = 8.0 Hz,
1H), 4.86 (t, J = 9.1 Hz, 1H), 4.63 (d, J = 8.7 Hz, 1H); 13C
NMR (125 MHz, CDCl3) for the β anomer: δ 170.7, 170.6 (2s),
170.3, 170.1, 170.0, 169.7, 169.3 (each C), 100.2, 95.5, 88.0,
75.2, 73.9, 73.4, 71.8, 71.3, 71.0, 70.8, 70.5, 68.0, 67.4, 67.0,
64.9 (each CH), 62.0 (CH2), 60.8 (CH2), 20.8, 20.7 (3s), 20.6
(3s), 15.6 (each CH3); selected 13C NMR data for the α anomer:
δ 101.1, 87.7, 75.8, 74.8, 72.5, 71.0, 70.9, 70.8, 69.1, 66.6 (each
CH), 61.7 (CH2), 60.4 (CH2), 14.2 (CH3); HRMS-ESI: calcd for
C36H49N3O23Na[M + Na]+, 914.2655; found, 914.2621.
1,3,5-Tris((1-(α-L-fucopyranosyl-(1 → 2)-β-D-galactopyranosyl-
(1 → 4)-β-D-glucopyranosyl)-1H-1,2,3-triazol-4-yl)methoxy)-
benzene 6
Azide 27 (90 mg, 101 μmol) was dissolved in CH3OH–H2O
(2 : 1, 15 mL), then 1315 (8.0 mg, 33.7 μmol), sodium ascorbate
(4.0 mg dissolved in 1 mL H2O, 20.2 μmol) and CuSO4 (1.6 mg
dissolved in 1 mL H2O, 10.1 μmol) were subsequently added
and the mixture was stirred at 45 °C overnight, after which the
solvent was removed and the residue was participated by CH2Cl2
(50 mL) and water (15 mL). The organic phase was washed
by water (15 mL × 2), dried by Na2SO4 and concentrated.
Chromatography (CH2Cl2–CH3OH, gradient elution, 80 : 1 to
70 : 1 to 60 : 1) gave the acetylated intermediate as a white foam
(83 mg, 85%), Rf 0.55 (CH2Cl2–CH3OH, 20 : 1). Removal of the
protecting groups from this protected compound (39 mg,
0.013 mmol), as for the formation of 4, gave 6, after lyophiliza-
tion, as a white solid (20 mg, 82%); [α]2D0 −50.5 (c 0.2, D2O);
1H-NMR (500 MHz, D2O) δ 8.22 (s, 3H), 6.30 (s, 3H), 5.68 (d,
J = 9.2 Hz, 3H, H-1), 5.25 (d, J = 3.1 Hz, 3H, H-1′′), 5.11 (s,
6H), 4.50 (d, J = 7.8 Hz, 3H, H-1′), 4.16 (dd, J = 13.2, 6.5 Hz,
3H), 3.96 (t, J = 9.3 Hz, 3H), 3.90–3.61 (m, 42H), 1.17 (d, J =
6.5 Hz, 9H); 13C NMR (125 MHz, D2O) δ 159.4 (C), 143.5 (C),
124.4, 100.2, 99.3, 96.4, 87.3, 78.1, 76.3, 75.3, 74.9, 74.4, 73.5,
72.0, 71.6, 69.6, 69.1, 68.2, 66.9 (each CH), 61.3, 61.1, 59.8
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 6893–6907 | 6905

View Article Online
(each CH2), 15.3 (CH3); HRMS-ESI: calcd for C69H105N9O45Na
[M + Na]+, 1802.6102; found, 1802.6108.
490 nm as described.6a,25a Assays were routinely done in dupli-
cates with up to five independent series, standard deviations for
percentage of bound lectin as parameter not exceeding 16.2%.
Assays with the human B-lymphoblastoid/pancreatic carci-
noma and the Chinese hamster ovary cell lines followed an opti-
mized protocol ensuring interstudy comparison.6,25c–f Aliquots
of cell suspensions at the same passage were routinely processed
at least in duplicates, with at least three independent series, by
washing to thoroughly remove serum components and then incu-
bation with lectin-containing solution at 4 °C for 30 min. in the
loading step, later with streptavidin-R-phycoerythrin (1 : 40;
Sigma) in the labeling step. Controls included omission of the
loading step to measure the level of lectin-independent back-
ground staining and application of non-cognate sugar to track
down osmolarity effects. Following normalization of values
based on the internal controls the standard deviations of
measurements did not exceed 12.7%.
Molecular modelling
Structures were first built using Maestro version 6.0 (Schrodinger
Inc., LLC, New York, USA). Constraints were then applied
during an energy minimization (OPLSAA force field, gas phase,
PRCG method to convergence) of each structure using Macro-
model version 8.5 (Schrodinger Inc.) so as to generate an
extended conformation for each structure. This approach enabled
estimation of the maximum distances that could potentially be
adopted between the lactose headgroups in 3, 4, 5 and 7. In the
case of 4 the angles defined by the three glucose anomeric
carbon atoms were constrained at 60°, for 5 the angles between
the glucose anomeric carbon atoms of two of the lactose residues
and the nitrogen atom at the centre of the scaffold were con-
strained at 120°. In the case of 7 the angles between three of the
glucose anomeric carbons were constrained to 90°. Distance con-
straints had to be applied between the glucose anomeric carbons
in order to maximize the spacing between the headgroups.
Torsions were adjusted using Maestro if necessary to help
generate the extended structures. The final modeled structures
obtained are shown in Fig. 1. The distances between the head-
groups for these structures are given in the main text.
Acknowledgements
This work has been generously supported by the EC research
program GlycHIT (contract ID 260600), the Verein zur
Förderung des biologisch-technologischen Fortschritts in der
Medizin e. V. (Heidelberg, Germany) and Science Foundation
Ireland (08/SRC/B1393). We also thank Drs. B. Friday and
S. Namirha for inspiring discussions as well as the reviewers for
their helpful advice.
Lectin purifications and quality controls
Using extracts of dried plant material or bacterial pellets from
recombinant production, affinity chromatography over lactosy-
lated Sepharose 4B obtained by divinyl sulfone activation was
performed as crucial step, following a standard procedure.9,25d,36
The B-chain of the toxin was obtained after in situ cleavage of
the disulfide bond linking the AB-chains by extensive treatment
with β-mercaptoethanol, removal of the A-chain by column
washes and covalent deactivation of the sulfhydryl group in the
resin-bound B-chain by iodoacetamide treatment, trCG-3/Gal-3
by purifying the cloned product or collagenase treatment of the
full-length protein.37 One- and two-dimensional gel electrophor-
esis and gel filtration as well as haemagglutination assays were
routinely performed to ensure purity, quaternary structure and
activity.25c,37 Biotinylation with the N-hydroxysuccinimide ester
derivative (Sigma, Munich, Germany) under activity-preserving
conditions followed a standard protocol, with incorporation
yields measured by mass spectrometric analysis.37b,c
References
1 (a) J. Holgersson, A. Gustafsson and S. Gaunitz, in The Sugar Code.
Fundamentals of Glycosciences, ed. H.-J. Gabius, Wiley-VCH,
Weinheim, 2009, pp. 279–300; (b) 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; (c) Y. M. Chabre,
D. Giguère, B. Blanchard, J. Rodrigue, S. Rocheleau, M. Neault,
S. Rauthu, A. Papadopoulos, A. A. Arnold, A. Imberty and R. Roy,
Chem.–Eur. J., 2011, 17, 6545–6562; (d) H.-J. Gabius, S. André,
J. Jiménez-Barbero, A. Romero and D. Solís, Trends Biochem. Sci., 2011,
36, 298–313.
2 H.-J. Gabius, in The Sugar Code. Fundamentals of Glycosciences, ed.
H.-J. Gabius, Wiley-VCH, Weinheim, 2009, pp. 317–328.
3 (a) K. G. Rice, O. A. Weisz, T. Barthel, R. T. Lee and Y. C. Lee, J. Biol.
Chem., 1990, 265, 18429–18434; (b) Y. C. Lee and R. T. Lee, in
Neoglycoconjugates: Preparation and Applications, ed. Y. C. Lee and
R. T. Lee, Academic Press, San Diego, 1994, pp. 23–50; (c) Y. C. Lee,
Trends Glycosci. Glycotechnol., 2010, 22, 95–106.
4 (a) Y. C. Lee, R. R. Townsend, M. R. Hardy, J. Lönngren, J. Arnarp,
M. Haraldsson and H. Lönn, J. Biol. Chem., 1983, 258, 199–202;
(b) R. T. Lee and Y. C. Lee, Glycoconjugate J., 1987, 4, 317–328;
(c) H. F. Lodish, Trends Biochem. Sci., 1991, 16, 374–377; (d) Y. C. Lee,
FASEB J., 1992, 6, 3193–3200; (e) S. André, S. Kojima, G. Gundel,
R. Russwurm, X. Schratt, C. Unverzagt and H.-J. Gabius, Biochim.
Biophys. Acta, Gen. Subj., 2006, 1760, 768–782.
Inhibition assays
Microtiter plate wells were coated with ASF (0.5 μg per well in
50 μl phosphate-buffered saline) overnight at 4 °C and residual
sites for binding protein were blocked with 100 μl buffer con-
taining 1% (w/v) carbohydrate-free bovine serum albumin for
1 h at 37 °C. Following washing the series of steps comprising
incubation with biotinylated lectin, washing, application of
the streptavidin-peroxidase conjugate (0.5 μg ml−1; Sigma),
washing and signal development with the chromogenic sub-
strates o-phenylenediamine (1 μg ml−1)/hydrogen peroxide (1 μl
ml−1) was completed with readings of the optical density at
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. Roy, Adv. Carbohydr. Chem.
Biochem., 2010, 63, 165–393.
6 (a) R. Leyden, T. Velasco-Torrijos, S. André, S. Gouin, H.-J. Gabius and
P. V. Murphy, J. Org. Chem., 2009, 74, 9010–9026; (b) S. André,
D. V. Jarikote, D. Yan, L. Vincenz, G.-N. Wang, H. Kaltner, P. V. Murphy
and H.-J. Gabius, Bioorg. Med. Chem. Lett., 2012, 22, 313–318.
7 (a) K.-I. Kasai and J. Hirabayashi, J. Biochem., 1996, 119, 1–8;
(b) D. N. W. Cooper, Biochim. Biophys. Acta, Gen. Subj., 2002, 1572,
209–231; (c) H. Kaltner and H.-J. Gabius, Histol. Histopathol., 2012, 27,
397–416.
6906 | Org. Biomol. Chem., 2012, 10, 6893–6907
This journal is © The Royal Society of Chemistry 2012

View Article Online
8 (a) O. E. Galanina, H. Kaltner, N. V. Bovin and H.-J. Gabius, J. Mol.
Recognit., 1997, 10, 139–147; (b) C. Svensson, S. Teneberg,
C. L. Nilsson, A. Kjellberg, F. P. Schwarz, N. Sharon and U. Krengel,
J. Mol. Biol., 2002, 321, 69–83.
23 (a) S. André, T. Kožár, R. Schuberth, C. Unverzagt, S. Kojima and
H.-J. Gabius, Biochemistry, 2007, 46, 6984–6995; (b) S. André, T. Kožár,
S. Kojima, C. Unverzagt and H.-J. Gabius, Biol. Chem., 2009, 390, 557–
565.
9 A. M. Wu, T. Singh, J.-H. Liu, M. Krzeminski, R. Russwurm,
H.-C. Siebert, A. M. J. J. Bonvin, S. André and H.-J. Gabius, Glycobio-
logy, 2007, 17, 165–184.
10 O. A. Vokhmyanina, R. M. Rapoport, I. M. Ryzhov, E. Y. Korchagina,
G. V. Pazynina, V. V. Severov, H. Kaltner, S. André, H.-J. Gabius and
N. V. Bovin, Biochemistry (Mosc), 2011, 76, 1185–1192.
24 (a) R. Mikeska, R. Wacker, R. Arni, T. P. Singh, A. Mikhailov,
A. Gabdoulkhakov, W. Voelter and C. Betzel, Acta Crystallogr., Sect. F:
Struct. Biol. Cryst. Commun., 2005, F61, 17–25; (b) M. Jiménez,
S. André, H.-C. Siebert, H.-J. Gabius and D. Solís, Glycobiology, 2006,
16, 926–937.
25 (a) S. André, P. J. Ortega, M. A. Perez, R. Roy and H.-J. Gabius, Glyco-
biology, 1999, 9, 1253–1261; (b) S. André, B. Liu, H.-J. Gabius and
R. Roy, Org. Biomol. Chem., 2003, 1, 3909–3916; (c) S. André,
H. Kaltner, T. Furuike, S.-I. Nishimura and H.-J. Gabius, Bioconjugate
Chem., 2004, 15, 87–98; (d) S. André, F. Sansone, H. Kaltner,
A. Casnati, J. Kopitz, H.-J. Gabius and R. Ungaro, ChemBioChem, 2008,
9, 1649–1661; (e) S. André, D. Specker, N. V. Bovin, M. Lensch,
H. Kaltner, H.-J. Gabius and V. Wittmann, Bioconjugate Chem., 2009,
20, 1716–1728; (f) S. André, M. Lahmann, H.-J. Gabius and
S. Oscarson, Mol. Pharmaceut., 2010, 7, 2270–2279.
26 H. Leffler and S. H. Barondes, J. Biol. Chem., 1986, 261, 10119–
10126.
27 N. Ahmad, H.-J. Gabius, S. André, H. Kaltner, S. Sabesan, R. Roy,
B. Liu, F. Macaluso and C. F. Brewer, J. Biol. Chem., 2004, 279, 10841–
10847.
28 K. Bachhawat-Sikder, C. J. Thomas and A. Surolia, FEBS Lett., 2001,
500, 75–79.
29 S. André, O. Renaudet, I. Bossu, P. Dumy and H.-J. Gabius, J. Pept. Sci.,
2011, 17, 427–437.
11 (a) C. P. Sparrow, H. Leffler and S. H. Barondes, J. Biol. Chem., 1987,
262, 7383–7390; (b) S. Sato and R. C. Hughes, J. Biol. Chem., 1992,
267, 6983–6990; (c) T. Feizi, J. C. Solomon, C.-T. Yuen, K. C. G. Jeng,
L. G. Frigeri, D. K. Hsu and F.-T. Liu, Biochemistry, 1994, 33, 6342–
6349; (d) J. Hirabayashi, T. Hashidate, Y. Arata, N. Nishi, T. Nakamura,
M. Hirashima, T. Urashima, T. Oka, M. Futai, W. E. G. Müller, F. Yagi
and K.-i. Kasai, Biochim. Biophys. Acta, Gen. Subj., 2002, 1572, 232–
254; (e) 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; (f) M. Krzeminski, T. Singh, S. André, M. Lensch,
A. M. Wu, A. M. J. J. Bonvin and H.-J. Gabius, Biochim. Biophys. Acta,
Gen. Subj., 2011, 1810, 150–161.
12 (a) Y. Oda, J. Herrmann, M. A. Gitt, C. W. Turck, A. L. Burlingame,
S. H. Barondes and H. Leffler, J. Biol. Chem., 1993, 268, 5929–5939;
(b) H. Ideo, A. Seko, T. Ohkura, K. L. Matta and K. Yamashita, Glyco-
biology, 2002, 12, 199–208; (c) A. M. Wu, J. H. Wu, J.-H. Liu, T. Singh,
S. André, H. Kaltner and H.-J. Gabius, Biochimie, 2004, 86, 317–326.
13 (a) D. Gupta, H. Kaltner, X. Dong, H.-J. Gabius and C. F. Brewer, Glyco-
biology, 1996, 6, 843–849; (b) T. K. Dam, H.-J. Gabius, S. André,
H. Kaltner, M. Lensch and C. F. Brewer, Biochemistry, 2005, 44, 12564–
12571.
30 S. Martín-Santamaría, S. André, E. Buzamet, R. Caraballo,
G. Fernández-Cureses, M. Morando, J. P. Ribeiro, K. Ramírez-Gualito,
B. de Pascual-Teresa, F. J. Cañada, M. Menéndez, O. Ramström,
J. Jiménez-Barbero, D. Solís and H.-J. Gabius, Org. Biomol. Chem.,
2011, 9, 5445–5455.
14 K. Soai, A. Ookawa and K. Kato, Bull. Chem. Soc. Jpn., 1982, 55, 1671–
1672.
15 J. D. Xie, L. H. Hu, W. F. Shi, X. X. Deng, Z. Q. Cao and Q. S. Shen,
Polym. Int., 2008, 57, 965–974.
31 J. Kopitz, M. Bergmann and H.-J. Gabius, IUBMB Life, 2010, 62, 624–
628.
16 (a) H. C. Kolb, M. G. Finn and K. B. Sharpless, Angew. Chem., Int. Ed.,
2001, 40, 2004–2021; (b) F. Pérez-Balderas, M. Ortega-Muñoz,
J. Morales-Sanfrutos, F. Hernández-Mateo, F. G. Calvo-Flores,
J. A. Calvo-Asín, J. Isac-García and F. Santoyo-González, Org. Lett.,
2003, 5, 1951–1954; (c) F. Morvan, A. Meyer, A. Jochum, C. Sabin,
Y. Chevolot, A. Imberty, J. P. Praly, J. J. Vasseur, E. Souteyrand and
S. Vidal, Bioconjugate Chem., 2007, 18, 1637–1643; (d) M. Meldal and
C. W. Tornoe, Chem. Rev., 2008, 108, 2952–3015; (e) S. André,
T. Velasco-Torrijos, R. Leyden, S. Gouin, M. Tosin, P. V. Murphy and
H.-J. Gabius, Org. Biomol. Chem., 2009, 7, 4715–4725; (f) V. Aragao-
Leoneti, V. L. Campo, S. A. Gomes, R. A. Field and I. Carvalho, Tetrahe-
dron, 2010, 66, 9475–9492; (g) V. L. Campo, I. Carvalho, C. H. T. P. Da
Silva, S. Schenkman, L. Hill, S. A. Nepogodiev and R. A. Field, Chem.
Sci., 2010, 1, 507–514; (h) A. Papadopolos, T. C. Shiao and R. Roy, Mol.
Pharmaceutics, 2012, 9, 394–403.
32 S. André, C. Grandjean, F. M. Gautier, S. Bernardi, F. Sansone,
H.-J. Gabius and R. Ungaro, Chem. Commun., 2011, 47, 6126–6128.
33 (a) D. Delacour, V. Gouyer, J.-P. Zanetta, H. Drobecq, E. Leteurtre,
G. Grard, O. Moureau-Hannedouche, E. Maes, A. Pons, S. André, A. Le
Bisic, H.-J. Gabius, A. Manninen, K. Simons and G. Huet, J. Cell Biol.,
2005, 169, 491–501; (b) H. Ideo, A. Seko and K. Yamashita, J. Biol.
Chem., 2005, 280, 4730–4737.
34 S. André, D. Gigùere, T. K. Dam, C. F. Brewer, R. Roy and H.-J. Gabius,
New J. Chem., 2010, 34, 2229–2240.
35 For some earlier work on scaffold development see: (a) M. Tosin,
S. G. Gouin and P. V. Murphy, Org. Lett., 2005, 7, 211–214; (b) M. Tosin
and P. V. Murphy, J. Org. Chem., 2005, 70, 4107–4117; (c) T. Velasco-
Torrijos and P. V. Murphy, Tetrahedron: Asymmetry, 2005, 16, 261–272;
(d) T. Velasco-Torrijos and P. V. Murphy, Org. Lett., 2004, 5, 3961–
3964.
17 (a) G. Cravotto, V. V. Fokin, D. Garella, A. Binello, L. Boffa and
A. Barge, J. Comb. Chem., 2010, 12, 13–15; (b) P. Cintas, A. Barge,
S. Tagliapietra, L. Boffa and G. Cravotto, Nat. Protocols, 2010, 5, 607–
616.
18 Z. Wang, L. Y. Zhou, K. El-Boubbou, X. S. Ye and X. F. Huang, J. Org.
Chem., 2007, 72, 6409–6420.
36 (a) H.-J. Gabius, Anal. Biochem., 1990, 189, 91–94; (b) H. Kaltner,
D. Solís, J. Kopitz, M. Lensch, M. Lohr, J. C. Manning, M. Mürnseer,
M. Schnölzer, S. André, J. L. Sáiz and H.-J. Gabius, Biochem. J., 2008,
409, 591–599; (c) H. Kaltner, D. Solís, S. André, M. Lensch,
J. C. Manning, M. Mürnseer, J. L. Sáiz and H.-J. Gabius, Biochemistry,
2009, 48, 4403–4416.
19 (a) S. A. Abbas, J. J. Barlow and K. L. Matta, Carbohydr. Res., 1981, 88,
51–60; (b) M. Izumi, O. Tsuruta, S. Harayama and H. Hashimoto, J. Org.
Chem., 1997, 62, 992–998.
20 A. J. Ott, M. Brackhagen, A. Davtyan, B. Nolting, H. Boye,
A. Schoknecht and C. Vogel, J. Carbohydr. Chem., 2001, 20, 611–636.
21 J. L. Montero, J. Y. Winum, A. Leydet, M. Kamal, A. A. Pavia and
J. P. Roque, Carbohydr. Res., 1997, 297, 175–180.
22 B. H. M. Kuijpers, S. Groothuys, A. C. Soede, P. Laverman,
O. C. Boerman, F. L. van Delft and F. P. J. T. Rutjes, Bioconjugate
Chem., 2007, 18, 1847–1854.
37 (a) H.-J. Gabius, R. Engelhardt, S. Rehm and F. Cramer, J. Natl. Cancer
Inst., 1984, 73, 1349–1357; (b) H.-J. Gabius, F. Darro, M. Remmelink,
S. André, J. Kopitz, A. Danguy, S. Gabius, I. Salmon and R. Kiss,
Cancer Invest., 2001, 19, 114–126; (c) D. Kübler, C.-W. Hung,
T. K. Dam, J. Kopitz, S. André, H. Kaltner, M. Lohr, J. C. Manning,
L. He, H. Wang, A. Middelberg, C. F. Brewer, J. Reed, W.-D. Lehmann
and H.-J. Gabius, Biochim. Biophys. Acta, Gen. Subj., 2008, 1780, 716–
722; (d) H. Kaltner, D. Kübler, L. López-Merino, M. Lohr,
J. C. Manning, M. Lensch, J. Seidler, W.-D. Lehmann, S. André, D. Solís
and H.-J. Gabius, Anat. Rec., 2011, 294, 427–444.
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 6893–6907 | 6907