Enhancement of Plant and Bacterial Lectin Binding Affinities by Three-Dimensional Organized Cluster Glycosides Constructed on Helical Poly(phenylacetylene) Backbones

Article abstract of DOI:10.1002/cbic.201000447

A series of poly(phenylacetylene)s bearing diverse saccharide pendants-N-acetyl-D-glucosamine, D-lactose, and N-acetyl-D-neuraminic acid-were synthesized by rhodium-mediated polymerizations of the corresponding acetyl-protected glycosylated phenylacetylen

Full text of DOI:10.1002/cbic.201000447

DOI: 10.1002/cbic.201000447
Enhancement of Plant and Bacterial Lectin Binding
Affinities by Three-Dimensional Organized Cluster
Glycosides Constructed on Helical Poly(phenylacetylene)
Backbones
Issei Otsuka,[a, b] Bertrand Blanchard,[a] Redouane Borsali,[a] Anne Imberty,*[a] and
Toyoji Kakuchi*[b]
A series of poly(phenylacetylene)s bearing diverse saccharide
pendants—N-acetyl-d-glucosamine, d-lactose, and N-acetyl-d-
neuraminic acid—were synthesized by rhodium-mediated
polymerizations of the corresponding acetyl-protected glycosy-
lated phenylacetylenes followed by deprotection. The circular
dichroism spectra of these glycosylated poly(phenylacetylene)s
each displayed split-type Cotton effects in the long absorption
region of the conjugated polymer backbone (260–500 nm),
thus indicating predominantly one-handed helical conforma-
tions in their backbones. The binding affinities of these glyco-
sylated poly(phenylacetylene)s, and those of previously report-
ed phenylacetylenes bearing d-galactose, towards plant and
bacterial lectins were investigated by hemagglutination inhibi-
tion assay and isothermal titration calorimetry (ITC). The stoi-
chiometries of binding vary strongly, depending on the lectin
binding sites and the accessibilities of the carbohydrate resi-
dues in the helices. The measured affinities also vary, with the
maximum value observed for the interaction between poly-PA-
a-Gal and lectin I from Pseudomonas aeruginosa, with a Kd
value of 4 mm per monosaccharide representing a 200-fold
increase relative to the corresponding monomer.
Introduction
Carbohydrates play important roles on cellular surfaces as rec-
ognition ligands for a variety of receptor proteins, such as
pathogens, enzymes, and antibiotics in vivo.[1–3] The multiva-
lent structure of a carbohydrate on a cellular surface compen-
sates for weak individual protein–carbohydrate interactions. A
number of multivalent or clustered glyco-ligands have been
synthesized for biomimetic purposes, and their binding prop-
erties for carbohydrate-binding proteins (lectins) have been
studied in vitro. Most of these artificial glyco-ligands display
some enhancements in affinity for proteins, due to their multi-
valent carbohydrate structures: the so-called “cluster glycoside
effect”.[4–8] The strong affinities of these glyco-ligands for lectins
are of great interest in pharmaceutical applications, such as
targeting drug delivery and pathogen-sensing. Several glyco-li-
gands, such as glycosylated dendritic ligands, linear polymeric
ligands bearing glycosyl pendants, and self-assembled nano-
particles of amphiphilic glycopolymers, have therefore been
designed for these applications.[8–21] Artificial multivalent glyco-
ligands have also been utilized for studying the cluster glyco-
side effect. It has been reported that the conformations of the
backbones and the saccharide residues of the glyco-ligands,
factors such as rigidity and three-dimensional organization of
the polymeric framework, significantly affect the multivalency
effect in carbohydrate–lectin interactions.[22–27] However, the
detailed mechanism of this effect is still unclear. It is therefore
of great interest to synthesize glyco-ligands with highly con-
trolled conformations and to evaluate their affinities for lectins
both from theoretical and from experimental perspectives.
Poly(phenylacetylene) is one of the most promising polymer
backbones for a three-dimensional controlled framework for
glyco-ligands. Poly(phenylacetylene)s feature predominantly
one-handed (excess right- or left-handed) helical conforma-
tions, determined by the chiralities of the side-chain groups.
The chiralities of the pendant groups transfer to the main
chain, resulting in the three-dimensional organization of the
pendant groups on the preferred-handed helical backbone.[28]
Previously, our research group reported on three-dimensionally
organized glucose (Glc) and galactose (Gal) ligands on helical
poly(phenylacetylene) backbones. The phenylacetylenes bear-
ing d-galactose pendants (poly-PA-a-Gal and poly-PA-b-Gal)
showed high affinities for peanut agglutinin, due to the cluster
glycoside effect, and the affinities were strongly affected by
the backbone chirality based on the one-handed helical con-
formations.[27]
[a] Dr. I. Otsuka, Dr. B. Blanchard, Dr. R. Borsali, Dr. A. Imberty
Centre de Recherche sur les Macromolꢀcules Vꢀgꢀtales
(CERMAV, UPR-CNRS 5301)
affiliated with Universitꢀ Joseph Fourier and
member of Institut de Chimie Molꢀculaire ICMG FR2607
B.P. 53, 38041 Grenoble Cedex 9 (France)
Fax: (+33)476-547-203
E-mail: anne.imberty@cermav.cnrs.fr
[b] Dr. I. Otsuka, Prof. Dr. T. Kakuchi
Division of Biotechnology and Macromolecular Chemistry
Faculty of Engineering, Hokkaido University
060–8628 Sapporo (Japan)
E-mail: kakuchi@poly-bm.eng.hokudai.ac.jp
ChemBioChem 2010, 11, 2399 – 2408
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2399

A. Imberty, T. Kakuchi et al.
Lectins are sugar-binding proteins that specifically recognize
complex carbohydrates without modifying them. They exist in
most living organisms, ranging from viruses and bacteria to
plants and animals.[3,29] Most of the reports on the cluster gly-
coside effect, however, have dealt with the interactions with
plant lectins, so evaluating the interactions between multiva-
lent glyco-ligands and lectins of other origins is of great impor-
tance. In this study, a bacterial lectin produced by Pseudomo-
nas aeruginosa—PA-IL, specific for Gal—was used for the eval-
uation of carbohydrate–lectin interactions along with other
plant lectins: that is, lectins from wheat germ (WGA) specific
for N-acetyl-d-glucosamine (GlcNAc) and N-acetyl-d-neuraminic
acid (NeuAc), and from Erythrina cristagalli (ECA) specific for
lactose (Lac). Pseudomonas aeruginosa is an opportunistic bac-
terium responsible for numerous nosocomial infections in im-
munocompromised patients and cystic fibrosis patients.[30] PA-
IL is a virulence factor and its inhibition by glyco-compounds
has been demonstrated to block lung infection in murine
models.[31] Glyco-ligands with high affinities for PA-IL are thus
always expected to be antimicrobial agents against these dis-
eases.
Here we report: 1) the synthesis of poly(phenylacetylene)s
bearing b-GlcNAc (poly-PA-b-GlcNAc), b-Lac (poly-PA-b-Lac),
and a-NeuAc (poly-PA-a-NeuAc) residues, with the aid of rho-
dium catalysts (Schemes 1 and 2), 2) the predominately one-
handed helical conformations of these glycosylated poly(phe-
nylacetylene)s as determined by CD analysis, and 3) the high
affinities of these glycosylated poly(phenylacetylene)s and the
previously reported poly-PA-a-Gal and poly-PA-b-Gal towards
the lectins, characterized by hemagglutination inhibition assay,
ITC, and molecular modeling.
Scheme 2. Synthesis of acetyl-protected glycosylated phenylacetylene mon-
omers. Conditions: a) trimethylsilylacetylene, CuI, (Ph3P)2PdCl2, Et3N, RT;
b) Bu4NF, CH2Cl2/THF, RT.
Results and Discussion
Synthesis of glycosylated poly(phenylacetylene)s
A series of acetyl-protected glycosylated phenylacetylene mon-
omers (PA-b-GlcNAc-OAc, PA-b-Lac-OAc, and PA-a-NeuAc-OAc)
were synthesized by Sonogashira–Hagiwara coupling reactions
between trimethylsilylacetylene and the previously reported
para-iodophenyl glycosides and subsequent removal of the
TMS groups, as shown in Scheme 1. The obtained monomers
Scheme 1. Synthesis of glycosylated poly(phenylacetylene)s. Conditions: a) Rh(nbd)BPh4 or [Rh(nbd)Cl]2/triethylamine, THF or CH2Cl2, 308C; b) CH3ONa, NaOH,
THF or DMSO, RT.
2400
www.chembiochem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 2399 – 2408

Enhancement of Lectin Binding Affinity
were then polymerized through the use of Rh(nbd)BPh4 or [Rh-
(nbd)Cl]2 with triethylamine as catalysts in CHCl3 or THF. All the
polymerizations proceeded homogeneously to yield yellowish-
brown viscous products that were purified by reprecipitation
from methanol to give the acetyl-protected glycosylated poly-
(phenylacetylene)s (poly-PA-b-GlcNAc-OAc, poly-PA-b-Lac-OAc,
and poly-PA-a-NeuAc-OAc). The results of the polymerizations
are listed in Table 1. The obtained polymers had large number
polymers showed characteristic signals attributable to the hy-
droxy groups, due to the deacetylated sugar moieties, around
3300–3600 cmꢀ1, rather than those attributable to the acetyl
groups of the acetyl-protected polymers, observed at
1748 cmꢀ1 (Figure 2). The obtained products were thus assign-
Table 1. Polymerizations of acetyl-protected glycosylated phenylacety-
lenes.[a]
[b]
Polymer
Catalyst
Solvent Mn ꢁ104[b] Mw/Mn
poly-PA-b-Lac-OAc
poly-PA-b-GlcNAc-OAc Rh(nbd)BPh4
poly-PA-a-NeuAc-OAc Rh(nbd)BPh4
[Rh(nbd)Cl]2/TEA CHCl3
7.81
13.4[c]
7.94
3.79
10.0[c]
2.16
CHCl3
THF
[a] Conditions: [monomer]=0.1m.; [monomer]/[Rh]=50; temp, 308C;
time, 20 h. [b] Determined by SEC. [c] Calculated on the CHCl3-soluble
part.
average molecular weights (Mn) and polydispersities (Mw/Mn) as
determined by SEC analysis.[32] Figure 1 shows the 1H NMR
spectra of the obtained polymers. Broadened signals attributa-
ble to the protons due to the sugar moiety (about 3–6 ppm),
the acetyl groups (about 2 ppm), and the aromatic groups
(about 6–7 ppm) were observed in the spectra, so the ob-
tained polymers were assignable as glycosylated poly(phenyla-
cetylene)s. The acetyl-protected polymers were then treated
with base for deprotection. The IR spectra of the deprotected
Figure 2. IR spectra of: top) poly-PA-a-NeuAc-OAc, and bottom) poly-PA-a-
NeuAc.
able as poly(phenylacetylene)s with GlcNAc, Lac, and NeuAc as
pendant groups (poly-PA-b-GlcNAc, poly-PA-b-Lac, and poly-
PA-a-NeuAc).
Chiroptical properties of glycosylated poly(phenylacety-
lene)s
The structural conformations of the glycosylated poly(phenyla-
cetylene)s were investigated through their CD and ultraviolet/
visible (UV/Vis) spectra. Figure 3 shows the CD and UV/Vis
spectra of poly-PA-b-GlcNAc, poly-PA-b-Lac, and poly-PA-a-
NeuAc in H2O and DMSO at room temperature. These glycosy-
lated poly(phenylacetylene)s each displayed split-type Cotton
effects in the long absorption region of the conjugated poly-
mer backbone (260–500 nm): poly-PA-b-GlcNAc in H2O, for
example, displayed a first negative Cotton effect at 411 nm, a
second positive one at 355 nm, a third negative one at
323 nm, and a fourth positive one at 281 nm. The chiroptical
properties of the backbones evidently demonstrate conforma-
tional chirality and thus one-handed helical conformations of
the p-conjugated double bonds in the backbones. The biased
helical conformations of these glycosylated poly(phenylacety-
lene)s were, as we previously reported,[27] dictated by the
pendant sugar functionalities and solvents, resulting in a varie-
ty of CD and UV/Vis profiles.
Figure 1. 1H NMR spectra of: A) poly-PA-b-GlcNAc in [D6]DMSO, B) poly-PA-a-
NeuAc-OAc in CD3OD, and C) poly-PA-b-Lac-OAc in CDCl3.
ChemBioChem 2010, 11, 2399 – 2408
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
2401

A. Imberty, T. Kakuchi et al.
Figure 3. CD (upper) and UV/Vis (lower) spectra of: A) poly-PA-b-GlcNAc, B) poly-PA-b-Lac, and C) poly-PA-a-NeuAc in H2O and DMSO at room temperature.
[q] is molar ellipticity.
Interactions of glycosylated poly(phenylacetylene)s with
lectins: hemagglutinating inhibition assay
ligands tested demonstrated inhibition of agglutination,
except for the NeuAc-containing ligands, which had no effect
at the concentrations tested here. Indeed, the binding affinity
between NeuAc and WGA is not strong enough to inhibit he-
magglutination under these concentrations. In all other cases,
the glycosylated poly(phenylacetylene)s showed lower ICmin
values than the corresponding monovalent glyco-ligands. The
ICmin value of poly-PA-b-GlcNAc (2ꢁ10ꢀ5 m) was much lower
than that of b-Ph-GlcNAc (1.1ꢁ10ꢀ2 m), indicating an increase
in affinity of three orders of magnitude for poly-PA-b-GlcNAc
towards WGA in relation to b-Ph-GlcNAc. A similar effect is ob-
served for the bacterial lectin PA-IL. Poly-PA-b-Gal (ICmin =2ꢁ
10ꢀ5 m) has a much stronger affinity (two orders of magnitude)
towards PA-IL than b-NPh-Gal (ICmin =4.1ꢁ10ꢀ3 m). The poly-PA-
a-Gal is an even better ligand, because this compound can
inhibit the hemagglutation by PA-IL at micromolar concentra-
tions (ICmin =9ꢁ10ꢀ6 m). However, when the binding of poly-
PA-b-LacNAc to ECA is analyzed, the increase is only one order
of magnitude, indicating that this effect depends on the archi-
tecture of the target lectin. These multiplications of the bind-
ing affinities observed in the glycosylated poly(phenylacety-
lene)s in relation to the monovalent glyco-ligands should be
explained by the cluster glycoside effect: that is, enhancement
of the binding activities of glyco-ligands through the multiva-
lent bindings of saccharides to lectins achieved by the highly
clustered saccharide pendants along the poly(phenylacetylene)
backbones.
The binding affinities between the glycosylated poly(phenyla-
cetylene)s and the lectins were investigated by hemagglutinat-
ing inhibition assay. The glycosylated poly(phenylacetylene)s
and the corresponding monovalent glyco-ligands bind with
specific lectins and thereby inhibit the agglutination between
the lectins and rabbit red cells. Previously we had reported[27]
that glycosylated poly(phenylacetylene)s bearing a-d-galacto-
pyranose (poly-PA-a-Gal) and b-d-galactopyranose (poly-PA-b-
Gal) derivatives showed higher affinities towards plant lec-
tins—peanut agglutinin (PNA: specific for Gal)—than the corre-
sponding monovalent glyco-ligands, due to multivalent carbo-
hydrate–lectin interactions. It is therefore of great interest to
compare the results obtained for the binding affinities with
the plant lectin and the bacterial lectin. Binding affinities were
evaluated by comparing the minimum concentrations of the
different glyco-ligands needed to inhibit the agglutination
(ICmin), as listed in Table 2 in the form of the concentrations of
the glycosylated poly(phenylacetylene)s calculated on the
basis of the glycosylated phenylacetylene monomer. All glyco-
Table 2. Hemagglutination inhibition properties of glycosylated
poly(phenylacetylene)s.[a]
Inhibitor (saccharide)
Lectin
ICmin ꢁ103 [m]
b-Ph-GlcNAc
poly-PA-b-GlcNAc
a-Ph-NeuAc
poly-PA-a-NeuAc
b-Ph-Lac
poly-PA-b-Lac
b-NPh-Gal
poly-PA-a-Gal
poly-PA-b-Gal
WGA
WGA
WGA
WGA
ECA
11
Interactions of glycosylated poly(phenylacetylene)s with
lectins: isothermal titration calorimetry (ITC)
0.02
N.I.[b]
N.I.[b]
1.5
0.2
4.1
Further insight into the binding affinities between the glycosy-
lated poly(phenylacetylene)s and the lectins was investigated
by ITC analysis. ITC is a direct method for measuring thermal
changes in the binding between the glyco-ligands and the lec-
tins, which allows determination of the association and dissoci-
ation constants (Ka and Kd), reaction stoichiometry (n), enthalpy
change (DH), free energy change (DG), and entropy change
(DS).[33] ITC measurements were performed by addition of solu-
ECA
PA-IL
PA-IL
PA-IL
0.009
0.02
[a] [Lectins]=4ꢁ[minimum concentration required for hemagglutination].
[b] Not inhibited by 6.5 mm for a-Ph-Neu and 6.1 mm for poly-PA-a-
NeuAc.
2402
www.chembiochem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 2399 – 2408

Enhancement of Lectin Binding Affinity
Figure 4. Binding thermogram (top) and isotherm (bottom) for: A) the addition of a solution of b-NPh-Gal (1.7 mm) to a PA-IL solution (16 mm), and B) the
addition of a solution of poly-PA-a-Gal (1.7 mm) to a PA-IL solution (10 mm) at 298 K.
tions of glyco-ligands to solutions of lectins in the microca-
lorimeter cell. Figure 4 shows typical ITC thermograms of titra-
tions of PA-IL by b-NPh-Gal and poly-PA-a-Gal. A strong exo-
thermic effect due to the binding between PA-IL and the
glyco-ligands was observed. The magnitude of this effect de-
creased along with progress of the titration, which corre-
sponds to the occupancy of the binding sites of PA-IL. The
very small exothermic signals observed at the ends of the titra-
tions indicate that the binding of PA-IL to the glyco-ligands is
complete. The thermodynamic parameters were calculated
from the thermogram on the basis of the simplest model of
binding (“one set of binding sites” model provided from the
ORIGIN software) with the assumption that the individual
glyco-ligands bind to the individual discrete binding sites of
the lectins in a non-cooperative fashion.
Table 3 summarizes the thermodynamic parameters calculat-
ed from the ITC analysis of the binding reactions between the
glyco-ligands and the lectins. Here, the reaction stoichiometry
(n) is defined as the number of the monomeric glyco units per
monomeric subunit of the lectin. It should be noted that the n
values of the monovalent glyco-ligands relative to the lectins
were fixed to the number of binding sites observed per mono-
mer in the crystal structures (n=1 for ECA and PA-IL, n=2 for
WGA).[34–36] Interestingly, all the glycosylated poly(phenylacety-
Table 3. Thermodynamic parameters calculated from ITC and expressed per saccharide unit[a]
Saccharide
Lectin
Ka ꢁ10ꢀ3
[mꢀ1
n (saccharide/
lectin)
DH
[kJmolꢀ1
Kd
[mm]
DG
[kJmolꢀ1
TDS
[kJmolꢀ1
]
]
]
]
b-Ph-GlcNAc[b]
poly-PA-b-GlcNAc[b]
b-Ph-Lac[c]
poly-PA-b-Lac[c]
a-Ph-NeuAc[d]
poly-PA-a-NeuAc[d]
b-NPh-Gal[e]
poly-PA-a-Gal[e]
poly-PA-b-Gal[e]
WGA[f]
WGA[f]
ECA[g]
ECA[g]
WGA[h]
WGA[h]
PA-IL[i]
PA-IL[i]
PA-IL[i]
1.75(ꢁ0.02)
14.4(ꢁ1.4)
4.1(ꢁ1.4)
2[j]
ꢀ117(ꢁ1)
572
69.4
245
389
3998
1960
13.3
4.12
5.05
ꢀ18.5
ꢀ23.7
ꢀ20.6
ꢀ19.5
ꢀ13.6
ꢀ14.8
ꢀ27.8
ꢀ30.7
ꢀ30.2
ꢀ98.8
20.8
ꢀ17.0
16.9
ꢀ52.6
12.3
ꢀ6.08
28.7
47.7(ꢁ0.8)
ꢀ2.85(ꢁ0.07)
ꢀ37.6(ꢁ0.1)
ꢀ2.54(ꢁ0.18)
ꢀ66.3(ꢁ0.7)
ꢀ0.47(ꢁ0.05)
ꢀ33.9(ꢁ0.3)
ꢀ2.13(ꢁ0.05)
ꢀ2.56(ꢁ0.05)
1[j]
2.6(ꢁ0.2)
9.7(ꢁ0.6)
2[j]
0.24(ꢁ0.00)
0.51(ꢁ0.06)
75.0(ꢁ1.7)
243(ꢁ37)
77.4(ꢁ6.2)
1[j]
10.1(ꢁ0.2)
7.3(ꢁ0.1)
198(ꢁ29)
27.7
[a] The parameters are average values of two experiments. [b] [b-Ph-GlcNAc] and [poly-PA-b-GlcNAc]=6.0 mm. [c] [b-Ph-Lac] and [poly-PA-b-Lac]=
12.7 mm. [d] [a-Ph-Neu] and [poly-PA-a-Neu]=26.0 mm. [e] [b-NPh-Gal], [poly-PA-a-Gal], and [poly-PA-b-Gal]=1.70 mm. [f] [WGA]=0.013 mm. [g] [ECA]=
0.038 mm. [h] [WGA]=0.024 mm. [i] [PA-IL]=0.016 and 0.010 mm. [j] The stoichiometries of the monovalent glyco-ligands to the lectins were fixed accord-
ing to the crystal structures.
ChemBioChem 2010, 11, 2399 – 2408
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
2403

A. Imberty, T. Kakuchi et al.
lene)s showed large n values against the plant and bacterial
lectins but with significant differences, with the n values vary-
ing from 7 for the interaction of poly-PA-b-Gal with PA-IL to 48
for the interaction of poly-PA-b-GlcNAc with WGA. Because the
lectins are not strongly different in terms of size (tetramer of
12.75 kDa for PA-IL, dimer of 26.2 kDa for ECA, and dimer of
43.2 kDa for WGA), the differences in stoichiometry probably
reflect the accessibilities of the monosaccharides, as functions
both of the conformations of the glycopolymers and of the ge-
ometries of the binding sites. The poly(phenylacetylene) chain
can adopt left- and right-handed conformations extended to
greater or lesser degrees. The natures of the monosaccharides
have some influence on the backbone conformation as dem-
onstrated by the CD analysis above. Construction of different
models, together with the observed stoichiometry, would be in
favor of a “brush”-shaped conformation for the poly-PA-a-Gal,
with Gal pendants easily accessible for the PA-IL binding site.
The relative size of the lectin and the polymer indicate that
one subunit of PA-IL can bind every eight Gal residues, in
good agreement with observed stoichiometry (Figure 5). On
the other hand, the flat b-GlcNAc moiety of poly-PA-b-GlcNAc
has a tendency to stack with neighboring ones, favoring trans-
oid helices in which the GlcNAc pendants are not accessible to
WGA. The lectin can therefore bind only to monosaccharides
at the extremities of the chain (or at a kink that could occur as
a rare event), explaining the unusually high value of the stoi-
chiometry (close to 50). It should also be kept in mind that
functional valency might not reflect structural valency because
other processes such as nonspecific binding and aggregation
can affect this value.
With regard to the affinities of lectins for the glyco-ligands,
the bacterial lectin PA-IL showed a much higher Ka value (order
of more than 105 mꢀ1) than the other plant lectins. The plant
lectins, at the same time, showed high Ka values of the order
of 103 to 104 mꢀ1, whereas the Ka values between WGA and
NeuAc-containing ligands were weaker (order of 102 mꢀ1). Inter-
estingly, each glycosylated poly(phenylacetylene), except for
poly-PA-b-Lac, showed Ka values two to eight times higher
than those of the corresponding monovalent glyco-ligands:
the Ka values of poly-PA-b-GlcNAc and b-Ph-GlcNAc, for exam-
ple, were 1.44ꢁ104 mꢀ1 and 1.75ꢁ103 mꢀ1, respectively. These
enhancements of the Ka values of glycosylated poly(phenylace-
tylene)s clearly supported the multivalency effects on the
lectin affinity observed in the hemagglutination assay. In con-
trast, poly-PA-b-Lac did not show a higher Ka value than b-Ph-
Lac, whereas it showed a 12-fold higher lectin affinity than b-
Ph-Lac in the hemagglutination inhibition assay. This is proba-
bly due to precipitation by cross-linking between poly-PA-b-
Lac and ECA that hampers integration of the peaks in the ITC
thermogram.
Some of the observed differences in affinity can be correlat-
ed with the known structures of lectins, and more precisely
with their binding sites and oligomeric architectures. Plant
lectins have rather open binding sites with low affinities for
monosaccharides (in the micromolar range), whereas bacterial
lectins, probably as the result of evolution pressure, display
higher affinities often associated with the participation of
bridging cations.[37] In terms of quaternary structure, ECA is a
dimeric lectin whereas PA-IL is tetrameric and WGA is dimeric
but with two binding sites per monomer (tetrameric behavior).
The multivalency state appears to have a crucial role on the
enhancement of affinity for the polymeric compounds, be-
cause the two tetrameric systems—PA-IL and WGA—each dis-
play much stronger affinity for the polymeric compound than
for the monomeric one, which is not the case for the dimeric
ECA.
As classically observed for protein–carbohydrate interac-
tions,[38] the binding reactions between the monovalent glyco-
ligands and the lectins displayed favorable enthalpy contribu-
tions and unfavorable entropy contributions. When the bind-
ing to the glycosylated poly(phenylacetylene)s is considered
and calculations per monomeric glyco unit are reported, the
enthalpy of binding drops significantly. This is clearly related to
the stoichiometry: because only a fraction of the carbohydrate
residue is available for binding, the apparent enthalpy contri-
bution is averaged and strongly decreased. In contrast, the
entropy of binding becomes favorable for the binding of the
glycosylated poly(phenylacetylene)s. For the binding between
WGA and GlcNAc-containing ligands, for instance, the DH
values varied from ꢀ117 kJ molꢀ1 to ꢀ2.85 kJmolꢀ1 and the
TDS values varied from ꢀ98.8 kJmolꢀ1 to +20.8 kJmolꢀ1 with
changing of the ligands from monomer to polymer. Similarly,
for the binding between PA-IL and Gal-containing ligands, the
DH values varied from ꢀ33.9 kJmolꢀ1 to ꢀ2.13 kJmolꢀ1 (poly-
PA-a-Gal) and the TDS values varied from ꢀ6.08 kJmolꢀ1 to
+28.7 kJmolꢀ1 (poly-PA-a-Gal).
Figure 5. A) Model of poly-PA-a-Gal in the non-extended right-handed con-
formation and comparison with two tetramers of PA-IL, and B) model of
poly-PA-b-GlcNAc in the extended right-handed conformation and compari-
son with a dimer of WGA.
2404
www.chembiochem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 2399 – 2408

Enhancement of Lectin Binding Affinity
box were monitored with an MB-MO-SE 1 and an MB-OX-SE 1, re-
spectively.
Conclusions
Poly(phenylacetylene) derivatives bearing b-GlcNAc, b-Lac, and
a-NeuAc residues were successfully synthesized by polymeri-
zations of the corresponding acetyl-protected glycosylated
phenylacetylenes, followed by deprotection. The CD spectra of
these glycosylated poly(phenylacetylene)s indicated predomi-
nantly one-handed helical structures in the backbones. The
high affinities of these glycosylated poly(phenylacetylene)s
towards plant and bacterial lectins, based on the multivalent
carbohydrate–lectin interactions, were demonstrated by the
hemagglutinating inhibition assay. For all lectins except ECA, a
two- to eightfold gain in affinity per saccharide is observed
when the polymeric glyco-ligand is compared to the mono-
meric one. This could be of interest for competing with the
binding of P. aeruginosa lectin to the host cell, because the
affinity of the lectin for the natural a-galactosylated ligand is
weaker (Kd ~100 mm).[39,40] The poly-PA-b-Gal glyco-ligand dis-
plays an affinity towards PA-IL of 4–5 mm (per galactose), which
is close to those of the highest-affinity compounds so far de-
scribed—calix[4]arene glycoconjugates bearing four b-galac-
tose residues, with affinities of 0.2 mm (i.e., 0.8 mm per galac-
tose).[41] The anti-infectious effects of galactosylated com-
pounds have been demonstrated previously in a murine
model,[31] but no assays were performed with polymeric mate-
rial and the innocuity of such compounds towards animals has
to be established before any therapeutic application.
Materials: 4-Iodophenyl 3,4,6,-tri-O-acetyl-2-N-acetyl-b-d-glucopyr-
anoside,[42] 4-iodophenyl O-(2,3,4,6-tetra-O-acetyl-b-d-galactopyra-
nosyl)-(1!4)-(O-2,3,6-tri-O-acetyl)-b-d-glucopyranoside,[43] methyl
(4-iodophenyl
5-acetamido-4,7,8,9-tetra-O-acetyl-3,5-dideoxy-d-
glycero-a-d-galacto-2-nonulopyranosid)onate,[44]
and
Rh+(2,5-
norbornadiene)[(h6-C6H5)Bꢀ(C6H5)3] (Rh(nbd)BPh4)[45] were synthe-
sized by previously reported methods. Trimethylsilylacetylene was
kindly supplied by Shinetsu Chemical Co. and was used without
further purification. 2,5-Norbornadiene-rhodium(I) chloride dimer
([Rh(nbd)Cl2]2) (96%) and bis(triphenylphosphine)palladium dichlor-
ide (> 98%) were purchased from Sigma–Aldrich and used as re-
ceived. Copper(I) iodide (98%) was purchased from Junsei Chemi-
cal Co. and used as received. Triethylamine (>99.0%) was pur-
chased from Kanto Chemical Co., Inc. and distilled over CaH2 prior
to use. Triphenylphosphine (>98%) was purchased from Kanto
Chemical and used after being recrystallized from dichlorometh-
ane/diethyl ether. Tetrabutylammonium fluoride (1.0m in THF),
sodium methoxide (1.0m in MeOH), dry THF (>99.5%, water con-
tent <0.005, v/%), and dry dichloromethane (>99.5%, water con-
tent <0.005, v/%) were purchased from Kanto Chemical and used
as received. WGA and ECA were purchased from Seikagaku Biobus-
iness Co. and used as received. Recombinant PA-IL was prepared
by the reported method.[39] Rabbit red cells were purchased from
BioMꢃrieux Co. and diluted to 2% with NaCl (0.15m) prior to use.
4-(Trimethylsilylethynyl)phenyl 3,4,6,-tri-O-acetyl-2-N-acetyl-b-d-
glucopyranoside (Method A): Trimethylsilylacetylene (6.0 mL,
42 mmol) was added under nitrogen at room temperature to a
mixture of 4-iodophenyl 3,4,6,-tri-O-acetyl-2-N-acetyl-b-d-glucopyr-
anoside (6.10 g, 11.2 mmol), bis(triphenylphosphine)palladium di-
chloride (170 mg, 240 mmol), copper(I) iodide (100 mg, 530 mmol),
and triethylamine (60 mL) in dry THF (150 mL). After stirring for
18 h, the reaction mixture was diluted with CHCl3 and filtered. The
filtrate was washed with H2O and dried over Na2SO4. After evapora-
tion of the solvent, the product was purified by column chroma-
tography (silica gel, CH2Cl2/ethyl acetate 1:1, v/v) to give 4-(trimeth-
ylsilylethynyl)phenyl 3,4,6,-tri-O-acetyl-2-N-acetyl-b-d-glucopyrano-
The ITC measurements indicated the entropically favorable
lectin binding of these glycosylated poly(phenylacetylene)s,
with high Ka and n values. The interpretation could be ham-
pered by aggregation processes so the thermodynamic param-
eters correspond to apparent ones. The n values calculated
from molecular modeling, which found that the reaction stoi-
chiometry reflects the conformation of the glycopolymer and
the geometry of the binding site of the lectin, were in good
agreement with the experimentally measured values. These re-
sults clearly indicated that the three-dimensional organizations
of the saccharide moieties in the multivalent glycosides strong-
ly affect the protein–carbohydrate interactions.
1
side as a white solid. Yield: 4.46 g (79%); H NMR (CDCl3): d=7.38,
6.91 (d, J=8.8 Hz, 4H; Ar), 5.60 (d, J=8.7 Hz, 1H; NH), 5.40 (dd, J=
10.4, 9.2 Hz, 1H; H-3), 5.27 (d, J=8.2 Hz, 1H; H-1), 5.13 (dd, J=
9.5 Hz, 1H; H-4), 4.28 (dd, J=12.2 Hz, 5.6 Hz, 1H; H-6a), 4.20–4.06
(m, 2H; H-2, H-6b), 3.88 (ddd, J=9.8, 5.6, 2.5 Hz, 1H; H-5), 2.08,
2.07, 2.05, 1.96 (s, 12H; OAc), 0.24 ppm (s, 9H; -SiMe3); 13C NMR
(CDCl3): d=170.9, 170.7, 170.6 169.5 (C=O), 157.1, 133.50, 118.0,
116.7 (Ar), 104.6 (Ar-CꢂC-SiMe3), 98.6 (C-1), 93.6 (Ar-CꢂC-SiMe3),
72.1, 68.7, 62.3, 54.9 (C-2, C-3, C-4, C-5, C-6), 23.4, 20.8, 20.7 (CH3),
0.1 ppm (SiMe3); elemental analysis (%) calcd. for C25H33NO9Si: C
57.79, H 6.40, N 2.70; found: C 57.52, H 6.42, N 2.65.
Experimental Section
1
Instruments: The H NMR and 13C NMR spectra were recorded with
JEOL JNM-EX270 and JEOL JNM-A400II instruments. Size-exclusion
chromatography (SEC) was performed at 408C in CHCl3
(0.8 mLminꢀ1) with a Jasco GPC-900 system and two Shodex K-
805L columns (linear, 8ꢁ300 mm; pore size, 10–1000 ꢂ; bead size,
10 mm; exclusion limit, 4ꢁ106). IR spectra were recorded with a
Perkin–Elmer Paragon 1000 FT-IR spectrometer. Circular dichroism
(CD) and ultraviolet/visible (UV/Vis) spectra were measured in a
1 mm quartz cell with a Jasco J-720 spectropolarimeter at room
temperature. Isothermal titration calorimetry (ITC) was performed
with a Microcal VP-ITC titration microcalorimeter (Northampton,
MA). Preparation of the polymerization solution was carried out
under dry argon (H2O, O2 <1 ppm) in an MBRAUN stainless steel
glove box with a gas purification system (molecular sieves and
copper catalyst). The moisture and oxygen contents in the glove
4-Ethynylphenyl 3,4,6,-tri-O-acetyl-2-N-acetyl-b-d-glucopyrano-
side (PA-b-GlcNAc-OAc, Method B): Tetrabutylammonium fluoride
solution (1.00m in THF, 9.12 mL) was added under nitrogen at
room temperature to a solution of 4-(trimethylsilylethynyl)phenyl
3,4,6,-tri-O-acetyl-2-N-acetyl-b-d-glucopyranoside
(3.96 g,
7.62 mmol) in dry CH2Cl2 (120 mL). After stirring for 40 min, the re-
action mixture was washed with brine and the organic layer was
dried over Na2SO4. The solvent was then evaporated and the prod-
uct was purified by column chromatography (silica gel, hexane/
diethyl ether 1:1, v/v) to give PA-b-GlcNAc-OAc as a white solid.
Yield: 3.24 g (95%); 1H NMR (CDCl3): d=7.42, 6.93 (d, J=8.8 Hz,
4H; Ar), 5.63 (d, J=8.6 Hz, 1H; NH), 5.41 (dd, J=10.4, 9.2 Hz, 1H;
H-3), 5.30 (d, J=8.2 Hz, 1H; H-1), 5.13 (dd, J=9.5 Hz, 1H; H-4), 4.29
ChemBioChem 2010, 11, 2399 – 2408
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
2405

A. Imberty, T. Kakuchi et al.
(dd, J=12.2, 5.6 Hz, 1H; H-6a), 4.20–4.05 (m, 2H; H-2, H-6b), 3.89
(ddd, J=9.8, 5.6, 2.5 Hz, 1H; H-5), 3.02 (s, 1H; CꢂCH), 2.07, 2.06,
2.05, 1.96 ppm (s, 12H; OAc); 13C NMR (CDCl3): d=170.9, 170.7,
170.6, 169.5 (C=O), 157.3, 133.6, 116.9 (Ar), 98.5 (C-1), 83.2 (Ar-Cꢂ
CH), 76.8 (Ar-CꢂCH), 72.1, 72.0, 68.7, 62.3, 54.8 (C-2, C-3, C-4, C-5, C-
6), 23.4, 20.8, 20.7 ppm (CH3); elemental analysis (%) calcd. for
C22H25NO9: C 59.06, H 5.63, N 3.13; found: C 58.86, H 5.63, N 2.98.
7.39, 6.98 (d, J=8.7 Hz, 4H; Ar), 5.35–5.39 (br, 2H; H-4, H-7), 5.25
(d, J=10.1 Hz, 1H; NH), 4.95 (ddd, J=12.2, 4.6, 4.5 Hz, 1H; H-8),
4.49 (d, J=10.8 Hz, 1H; H-6), 4.29–4.07 (m, 3H; H-5, H-9a, H-9b),
3.62 (s, 3H; OCH3), 2.70 (dd, J=13.0, 4.7, 1H; H-3e), 2.35 (m, 1H;
H-3a), 2.16, 2.13, 2.05, 1.92, 1.69 (s, 15H; OAc, NHAc), 0.23 ppm (s,
9H; TMS); 13C NMR (CDCl3): d=171.0, 170.7, 170.3, 170.2, 170,1 (C=
O), 168.1 (C-1), 154.2, 133.3, 119.2, 118.5 (Ar), 104.6 (Ar-CꢂC-TMS),
99.9 (C-2), 93.6 (Ar-CꢂC-TMS), 73.5 (C-6), 69.0 (C-8), 68.6 (C-4), 67.2
(C-7), 62.0 (C-9), 53.0 (OCH3), 49.4 (C-5), 38.4 (C-3), 23.3, 21.1, 20.9,
20.8 ppm (NHAc, OAc); elemental analysis (%) calcd. for
C31H41NO13Si: C 56.10, H 6.23, N 2.10; found: C 55.82, H 6.09, N
2.10.
4-(Trimethylsilylethynyl)phenyl O-(2,3,4,6-tetra-O-acetyl-b-d-gal-
actopyranosyl)-(1!4)-(O-2,3,6-tri-O-acetyl)-b-d-glucopyranoside:
Method A was applied to 4-iodophenyl O-(2,3,4,6-tetra-O-acetyl-b-
d-galactopyranosyl)-(1!4)-(O-2,3,6-tri-O-acetyl)-b-d-glucopyrano-
side (5.09 g, 6.08 mmol), bis(triphenylphosphine)palladium dichlor-
ide (86 mg, 120 mmol), copper(I) iodide (47 mg, 250 mmol), triethyl-
amine (21 mL), dry THF (24 mL), and trimethylsilylacetylene
(1.5 mL, 42 mmol). The product was purified by column chroma-
tography (silica gel, hexane/ethyl acetate 2:3, v/v) to give 4-(tri-
methylsilylethynyl)phenyl O-(2,3,4,6-tetra-O-acetyl-b-d-galactopyra-
nosyl)-(1!4)-(O-2,3,6-tri-O-acetyl)-b-d-glucopyranoside as a white
solid. Yield: 4.59 g (92%); 1H NMR (CDCl3): d=7.39, 6.89 (d, J=
8.6 Hz, 4H; Ar), 5.36 (d, J=3.2 Hz, 1H; H-1Glc), 5.28 (m, 1H; H-
1Gal), 5.20–5.05 (m, 3H; H-2Glc, H-2Gal, H-4Glc), 4.97 (m, 1H; H-
3Glc), 4.52–4.47 (m, 2H; H-4Gal, H-3Gal), 4.19–4.05 (m, 3H; H-5Glc,
H-6aGlc, H-6bGlc), 3.92–3.77 (m, 3H; H-5Gal, H-6aGal, H-6bGal)
2.16, 2.08, 2.07, 2.06, 2.06, 2.04, 1.97 (s, 21H; OAc) 0.24 ppm (s, 9H;
SiMe3); 13C NMR (CDCl3): d=170.5, 170.4, 170.2, 170.1, 169.8, 169.7,
169.2 (C=O), 156.8, 133.5, 118.2, 116.7 (Ar), 104.5 (Ar-CꢂC-SiMe3),
101.2 (C-1), 98.5 (C-1’), 93.6 (Ar-CꢂC-SiMe3), 76.3, 73.0, 72.9, 71.5,
71.0, 70.9, 69.2, 66.7, 62.1, 61.0 (C-2, C-2’, C-3, C-3’, C-4, C-4’, C-5, C-
5’, C-6, C-6’), 20.8–19.8 (CH3) 0.1 ppm (SiMe3); elemental analysis
(%) calcd. for C38H51O18Si: C 55.40, H 6.24; found: C 55.20, H 6.08.
Methyl (4-ethynylphenyl 5-acetamido-4,7,8,9-tetra-O-acetyl-3,5-
dideoxy-d-glycero-a-d-galacto-2-nonulopyranosid)onate (PA-a-
NeuAc-OAc): Method B was applied to methyl [4-(trimethylsilyleth-
ynyl)phenyl 5-acetamido-4,7,8,9-tetra-O-acetyl-3,5-dideoxy-d-glyc-
ero-a-d-galacto-2-nonulopyranosid]onate (2.93 g, 4.41 mmol), dry
CH2Cl2 (30 mL), and tetrabutylammonium fluoride (1.00 molLꢀ1 in
THF, 1.00 mL). The product was purified by column chromatogra-
phy (silica gel, hexane, CH2Cl2/MeOH 30:1!20:1, v/v) to give PA-a-
NeuAc-OAc as a white solid. Yield: 2.29 g (88%); 1H NMR (CDCl3):
d=7.41, 7.00 (d, J=8.2 Hz, 4H; Ar), 5.35–5.38 (br, 2H; H-4, H-7),
5.24 (d, J=10.0 Hz, 1H; NH), 4.95 (ddd, J=12.2, 4.9, 4.4 Hz, 1H; H-
8), 4.49 (d, J=8.2 Hz, 1H; H-6), 4.29–4.07 (m, 3H; H-5, H-9a, H-9b),
3.64 (s, 3H; OCH3), 3.03 (s, 1H; CꢂCH), 2.70 (dd, J=12.9, 4.9, 1H;
H-3e), 2.24 (m, 1H; H-3a), 2.16, 2.13, 2.05, 1.92, 1.67 ppm (s, 15H;
OAc, NHAc); 13C NMR (CDCl3): d=171.0, 170.6, 170.2, 170.1, 170,0
(C=O), 168.1 (C-1), 154.2, 133.4, 119.3, 117.4 (Ar), 99.8 (C-2) 83.1 (Ar-
CꢂCH) 76.7 (Ar-CꢂCH), 73.5 (C-6) 69.0 (C-8), 68.6 (C-4), 67.2 (C-7),
62.0 (C-9), 53.0 (OCH3), 49.4 (C-5), 38.3 (C-3), 23.2, 21.0, 20.8, 20.8,
20.7 ppm (NHAc, OAc); elemental analysis (%) calcd. for
C28H33NO13: C 56.85, H 5.62, N 2.37; found: C 56.55, H 5.54, N 2.20.
4-Ethynylphenyl
O-(2,3,4,6-tetra-O-acetyl-b-galactopyranosyl)-
(1!4)-(O-2,3,6-tri-O-acetyl)-b-d-glucopyranoside (PA-b-Lac-OAc):
Method B was applied to 4-(trimethylsilylethynyl)phenyl O-(2,3,4,6-
tetra-O-acetyl-b-d-galactopyranosyl)-(1!4)-(O-2,3,6-tri-O-acetyl)-b-
d-glucopyranoside (4.18 g, 5.19 mmol), CH2Cl2 (50 mL), and tetra-
butylammonium fluoride (6.22 mL; 1.00m in THF). The product was
purified by column chromatography (silica gel, hexane/ethyl ace-
tate 2:3, v/v) to give PA-b-Lac-OAc as a white solid. Yield: 3.20 g
Poly(4-ethynylphenyl 3,4,6,-tri-O-acetyl-2-N-acetyl-b-d-glucopyr-
anoside) (poly-PA-b-GlcNAc-OAc): A solution of Rh(norbornadi-
ene)BPh4 (12 mg, 22 mmol) in dry CHCl3 (3.0 mL) was added under
argon to
a stirred solution of PA-b-GlcNAc-OAc (500 mg,
1.12 mmol) in dry CHCl3 (8.0 mL). After 20 h at 308C, the reaction
was terminated by addition of triphenylphosphine (39 mg,
150 mmol) and the mixture was then poured into a large amount
of methanol. The precipitate was filtered off and dried in vacuo to
1
(84%); H NMR (CDCl3): d=7.42, 6.92 (d, J=8.9 Hz, 4H; Ar), 5.36 (d,
J=2.8 Hz, 1H; H-1Glc), 5.29 (m, 1H; H-1Gal), 5.21–5.06 (m, 3H; H-
2Glc, H-2Gal, H-4Glc), 4.97 (m, 1H; H-3Glc), 4.53–4.47 (m, 2H; H-
4Gal, H-3Gal), 4.19–4.05 (m, 3H; H-5Glc, H-6aGlc, H-6bGlc), 3.93–
3.77 (m, 3H; H-5Gal, H-6aGal, H-6bGal), 3.03 (s, 1H; CꢂCH), 2.16,
2.08, 2.07, 2.06, 2.06, 2.04, 1.97 ppm (s, 21H; OAc); 13C NMR
(CDCl3): d=170.4, 170.3, 170.2, 170.1, 169.8, 169.6, 169.2 (C=O),
157.0, 133.7, 117.0, 116.8 (Ar), 101.2, 98.4 (C-1, C-1’), 83.1 (Ar-CꢂCH),
76.8 (Ar-CꢂCH), 76.3, 73.0, 72.8, 71.5, 71.0, 70.8, 69.2, 66.7, 62.1,
60.9 (C-2, C-2’, C-3, C-3’, C-4, C-4’, C-5, C-5’, C-6, C-6’), 20.7–
20.6 ppm (CH3); elemental analysis (%) calcd. for C35H43O18: C 55.92,
H 5.77; found: C 55.52, H 5.58.
give PA-a-Neu-OAc as a yellow powder. Yield: 490 mg; Mnn, SEC
=
1.34ꢁ105; Mw/Mnn, SEC =10.1; IR (KBr): n˜ =1746 cmꢀ1 (nC O).
=
Poly[4-ethynylphenyl O-(2,3,4,6-tetra-O-acetyl-b-galactopyrano-
syl)-(1!4)-(O-2,3,6-tri-O-acetyl)-b-d-glucopyranoside] (poly-PA-
b-Lac-OAc): [Rh(nbd)Cl]2 (3.1 mg, 6.7 mmol) and triethylamine
(950 mL, 6.8 mmol) were added under argon to a stirred solution of
PA-b-Lac-OAc (500 mg, 680 mmol) in dry CHCl3 (6.1 mL). After stir-
ring for 20 h at 308C, the reaction was terminated by addition of
an excess quantity of acetic acid, and the mixture was then poured
into a large amount of methanol. The precipitate was filtered off
and dried in vacuo to give poly-PA-b-Lac-OAc as a yellow powder.
Yield: 330 mg; Mn, SEC =7.81ꢁ104; Mw/Mnn, SEC =3.79; IR (KBr): n˜ =
Methyl [4-(trimethylsilylethynyl)phenyl 5-acetamido-4,7,8,9-
tetra-O-acetyl-3,5-dideoxy-d-glycero-a-d-galacto-2-nonulopyra-
nosid]onate: Method B was applied to methyl (4-iodophenyl 5-
acetamido-4,7,8,9-tetra-O-acetyl-3,5-dideoxy-d-glycero-a-d-galacto-
2-nonulopyranosid)onate (3.35 g, 4.83 mmol), bis(triphenylphos-
phine)palladium dichloride (70.2 mg, 100 mmol), copper(I) iodide
(38.1 mg, 200 mmol), triethylamine (15 mL), dry THF (15 mL), and
trimethylsilylacetylene (1.4 mL, 10 mmol). The product was purified
by column chromatography (silica gel, CH2Cl2/MeOH 30:1!20:1, v/
v) to give methyl(4-trimethylsilylethynylphenyl 5-acetamido-4,7,8,9-
tetra-O-acetyl-3,5-dideoxy-d-glycero-a-d-galacto-2-nonulopyrano-
1740 cmꢀ1 (nC O).
=
Poly[methyl
(4-ethynylphenyl
5-acetamido-4,7,8,9-tetra-O-
acetyl-3,5-dideoxy-d-glycero-a-d-galacto-2-nonulopyranosid)o-
nate] (poly-PA-a-NeuAc-OAc): A solution of Rh(norbornadiene)-
BPh4 (7.0 mg, 14 mmol) in dry THF (800 mL) was added under argon
to a stirred solution of PA-a-NeuAc-OAc (400 mg, 680 mmol) in dry
THF (6.0 mL). After 20 h at 308C, the reaction was terminated by
addition of triphenylphosphine (21 mg, 80 mmol), and the mixture
was then poured into a large amount of diethyl ether. The precipi-
1
sid)onate as a white solid. Yield: 3.12 g (97%); H NMR (CDCl3): d=
2406
www.chembiochem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 2399 – 2408

Enhancement of Lectin Binding Affinity
tate was filtered off and dried in vacuo to give poly-PA-a-NeuAc-
OAc as a yellow powder. Yield: 370 mg; Mnn, SEC =7.94ꢁ104; Mw/
the aid of the ORIGIN 7.0 software (OriginLab Corp., Northampton,
MA). The experimental data were fitted to a theoretical titration
curve by use of the “one set of binding sites” model in ORIGIN 7.0
to determine the association constants (Ka), the reaction stoichiom-
etry (n), and the enthalpy change (DH). Changes in free energy
(DG) and entropy (DS) were calculated from Equation (1):
Mnn, SEC =2.16; IR (KBr): n˜ =1748 cmꢀ1 (nC O).
=
Poly(4-ethynylphenyl 2-N-acetyl-b-d-glucopyranoside) (poly-PA-
b-GlcNAc): CH3ONa (1m in MeOH, 1 mL) was added to a stirred so-
lution of poly-PA-b-GlcNAc-OAc (300 mg) in DMSO (100 mL). After
one day, water (100 mL) was added to dissolve the precipitate. The
mixture was dialyzed in a cellulose tube against water for five days
and then lyophilized to give poly-PA-a-GlcNAc as a yellow powder.
Yield: 200 mg; IR (KBr): n˜ =3300–3600 cmꢀ1 (nOꢀH).
DG ¼ DHꢀTDS ¼ RT ln Ka
ð1Þ
where T is the absolute temperature and R=8.314 Jmolꢀ1 Kꢀ1
.
The concentrations of lectins in the reaction cells varied from 13 to
38 mm (see Footnote in Table 3). This resulted in the Wiseman “c”
values being lower than 1 for experiments with monomeric com-
pounds and between 1 and 50 for experiments with multimeric
compounds. Measured heats of dilution of all ligands were negligi-
ble.
Poly(4-ethynylphenyl O-b-galactopyranosyl-(1!4)-b-d-glucopyr-
anoside) (poly-PA-b-Lac): CH3ONa (1m in MeOH, 1 mL) was added
to a stirred solution of poly-PA-b-Lac (200 mg) in THF (50 mL). After
one day, water (100 mL) was added to dissolve the precipitate. The
mixture was dialyzed in a cellulose tube against water for three
days and then lyophilized to give poly-PA-b-Lac as a yellow
powder. Yield: 110 mg; IR (KBr): n˜ =3300–3600 cmꢀ1 (nOꢀH).
Molecular modeling: Glycosylated phenylacetylene monomers
were built with the aid of the graphical editing menu of the Sybyl
package (Tripos, Inc., St. Louis). The monosaccharide moieties were
taken from the Glyco3D library (http://glyco3d.cermav.cnrs.fr/
glyco3d/index.php). Atom types and atomic charges of the mono-
saccharides were selected according to the PIM force-field.[47]
Atomic charges for the aglycon moiety were derived from MNDO
calculations. The monosaccharide was oriented with F and Y tor-
sion angles compatible with known energy minima.[48] A 32-mer
Poly(4-ethynylphenyl
5-acetamido-3,5-dideoxy-d-glycero-a-d-
galacto-2-nonulopyranoside) (poly-PA-a-NeuAc): CH3ONa (1m in
MeOH, 1 mL) was added to a stirred solution of poly-PA-a-Neu-
OAc (200 mg) in THF (40 mL). After 7 h, the mixture was evaporat-
ed and NaOH (5 mL; 0.1m in H2O) was added. The mixture was
stirred overnight followed by treatment with Amberlite IR-120 H+
resin for 10 min. The filtrate from the mixture was dialyzed in a cel-
lulose tube against water for two days and then lyophilized to give
poly-PA-a-NeuAc as a yellow powder. Yield: 160 mg; IR (KBr): n˜ =
3300–3600 cmꢀ1 (nOꢀH).
Hemagglutination assay: Hemagglutination activity was assayed
by the method described by Nowotny.[46] A twofold dilution series
of each lectin solution (25 mL, starting at 1 mgmLꢀ1) in tris(hydrox-
ymethyl)aminomethane hydrochloride (Tris-HCl, 100 mm) buffer
(pH 7.5, with 3 mm CaCl2) were prepared in 96-well microtiter U-
plates (Thermo Fisher Scientific, Inc.). An equal volume (25 mL) of
rabbit red cell solution (2%) in aq. NaCl (150 mm) was added to
each well, and the system was incubated at 378C for 30 min. The
agglutination was then visually analyzed, and the minimum lectin
concentration required for agglutination was determined.
was built by using the average values for phenylacetylene bond
lengths and angles taken from the literature[49] and a cis configura-
tion for the double bond. Four values of the torsion angle qc of
the single bond were considered, resulting in four helical confor-
mations of the glycopolymer together with tilting of the qR angle
in order to avoid steric conflicts. The four sets of torsional angles
are (qc =+1608; qR =ꢀ358) and (qc =+1408; qR =ꢀ108) for right-
handed helices and (qc =ꢀ1608; qR =+358) and (qc =ꢀ1408; qR =
+108) for left-handed ones. Full energy minimization with the use
of the TRIPOS force-field was performed and resulted in only mar-
ginal conformational adjustments.
Hemagglutinating inhibition assay: A twofold dilution series of
glyco-ligand solutions in the above buffer (25 mL, starting at
2.5 mgmLꢀ1) was prepared in 96-well microtiter U-plates. An equal
volume (25 mL) of lectin solution ([lectin]=4ꢁ[minimum concen-
tration required for hemagglutination]) was added to each well
and the system was incubated at 378C for 30 min. Rabbit red cell
solution (2%, 50 mL) in aq. NaCl (150 mm) was added to each well,
and the system was incubated at 378C for 30 min. The agglutina-
tion was then visually analyzed and the minimum concentration of
the glyco-ligand needed to inhibit agglutination was determined
as ICmin
.
Acknowledgements
Isothermal titration calorimetry (ITC): The thermodynamics of the
interaction between the glyco-ligands and the lectins were investi-
gated by measuring the heat production from the titration experi-
ments by ITC. A typical procedure is as follows: a solution of WGA
lectin (13 mm) was placed in the reaction cell (V=1.4478 mL). A so-
lution of poly-PA-b-GlcNAc (6.00 mm) was siphoned off by a com-
puter-controlled microsyringe (V=300 mL) and injected into the re-
action cell at 10 mL 30 times, with 5 min intervals between each in-
jection. The reaction mixture was stirred at 300 rpm and at 258C
during the titration. The raw experimental data were reported as
the heat production versus time. The amount of heat production
was calculated by integration of the areas of individual peaks with
This study was partly supported by a Grant-in-Aid from the
Japan Society for the Promotion of Science (JSPS) Fellows. The
authors thank Takayashi Hongo, Syunpei Omura, and Shinya
Onodera for their help.
Keywords: bacterial
interactions · glycoconjugates · helical structures · multivalent
glycoside effect
lectins
·
carbohydrate–protein
ChemBioChem 2010, 11, 2399 – 2408
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
2407

A. Imberty, T. Kakuchi et al.
[1] Y. C. Lee, R. T. Lee, Acc. Chem. Res. 1995, 28, 321–327.
[2] R. A. Dwek, Chem. Rev. 1996, 96, 683–720.
[30] A. Imberty, M. Wimmerova, E. P. Mitchell, N. Gilboa-Garber, Microb.
Infect. 2004, 6, 222–229.
[31] C. Chemani, A. Imberty, S. Bentzmann, M. Pierre, M. Wimmerovꢄ, B. P.
Guery, K. Faure, Infect. Immun. 2009, 77, 2065–2075.
[32] Poly-PA-b-GlcNAc-OAc turned out to be hardly soluble in common or-
ganic solvents after being precipitated in methanol, so the SEC mea-
surement was performed with the CHCl3-soluble part and the 1H NMR
measurement was performed after the deacetylation.
[3] H. Lis, N. Sharon, Chem. Rev. 1998, 98, 637–674.
[4] M. Mammen, S. K. Choi, G. M. Whitesides, Angew. Chem. 1998, 110,
2908–2953; Angew. Chem. Int. Ed. 1998, 37, 2754–2794.
[5] J. J. Lundquist, E. J. Toone, Chem. Rev. 2002, 102, 555–578.
[6] V. Martos, P. CastreÇo, J. Valero, J. Mendoza, Curr. Opin. Chem. Biol.
2008, 12, 698–706.
[33] It should be noted that the interaction between multivalent glyco-
ligands and lectins can result in the precipitation of the aggregates,
which might bring some errors of the ITC experiment. The thermo-
dynamic parameters as stated below might thus include these fudge
factors.
[34] C. Svensson, S. Teneberg, C. L. Nilsson, A. Kjellberg, F. P. Schwarz, N.
Sharon, U. Krengel, J. Mol. Biol. 2002, 321, 69–83.
[35] G. Cioci, E. P. Mitchell, C. Gautier, M. Wimmerovꢄ, D. Sudakevitz, S.
Pꢃrez, N. Gilboa-Garber, A. Imberty, FEBS Lett. 2003, 555, 297–301.
[36] C. S. Wright, J. Mol. Biol. 1990, 215, 635–651.
[7] N. Jayaraman, Chem. Soc., Rev. 2009, 38, 3463–3483.
[8] Y. M. Chabre, R. Roy, Adv. Carbohydr. Chem. Biochem. 2010, 63, 165–393.
[9] R. Roy, J. Carbohydr. Chem. 2002, 21, 769–798.
[10] A. Imberty, Y. M. Chabre, R. Roy, Chem. Eur. J. 2008, 14, 7490–7499.
[11] T. K. Dam, R. Roy, D. Pagꢃ, C. F. Brewer, Biochemistry 2002, 41, 1351–
1358.
[12] T. K. Dam, C. F. Brewer, Biochemistry 2008, 47, 8470–8476.
[13] K. Aoi, K. Itoh, M. Okada, Macromolecules 1995, 28, 5391–5393.
[14] P. R. Ashton, S. E. Boyd, C. L. Brown, N. Jayaraman, S. A. Nepogodiev, J. F.
Stoddart, Chem. Eur. J. 1996, 2, 1115–1128.
[37] A. Imberty, E. P. Mitchell, M. Wimmerovꢄ, Curr. Opin. Struct. Biol. 2005,
15, 525–534.
[15] Y. Kitajyo, T. Imai, Y. Sakai, M. Tamaki, H. Tani, K. Takahashi, A. Narumi, H.
Kaga, N. Kaneko, T. Satoh, T. Kakuchi, Polymer 2007, 48, 1237–1244.
[16] K. Kobayashi, M. Kakishita, M. Okada, T. Akaike, T. Usui, J. Carbohydr.
Chem. 1994, 13, 753–766.
[17] I. Otsuka, K. Fuchise, S. Halila, S. Fort, K. Aissou, I. P. -Paintrand, Y. Chen,
A. Narumi, T. Kakuchi, R. Borsali, Langmuir 2010, 26, 2325–2332.
[18] Y. Ruff, E. Buhler, S.-J. Candau, E. Kesselman, Y. Talmon, J.-M. Lehn, J. Am.
Chem. Soc. 2010, 132, 2573–2584.
[19] B. Voit, D. Appelhans, Macromol. Chem. Phys. 2010, 211, 727–735.
[20] M Ambrosi, N. R. Cameron, B. G. Davis, S. Stolnik, Org. Biomol. Chem.
2005, 3, 1476–1480.
[21] M.-G. Baek, R. C. Stevens, D. H. Charych, Bioconjugate Chem. 2000, 11,
777–788.
[38] T. K. Dam, C. F. Brewer, Chem. Rev. 2002, 102, 387–429.
[39] B. Blanchard, A. Nurisso, E. Hollville, C. Tꢃtaud, J. Wiels, M. Pokornꢄ, M.
Wimmerovꢄ, A, Varrot, A. Imberty, J. Mol. Biol. 2008, 383, 837–853.
[40] A. Nurisso, B. Blanchard, A. Audfray, L. Rydner, S. Oscarson, A. Varrot, A.
Imberty, J. Biol. Chem. 2010, 285, 20316–20327.
[41] S. Cecioni, R. Lalor, B. Blanchard, J. P. Praly, A. Imberty, S. E. Matthews, S.
Vidal, Chem. Eur. J. 2009, 15, 13232–13240.
[42] R. Roy, F. D. Tropper, Can. J. Chem. 1991, 69, 817–821.
[43] I. C. M. Dea, Carbohydr. Res. 1970, 12, 297–299.
[44] Z. Gan, R. Roy, Can. J. Chem. 2002, 80, 908–916.
[45] Y. Kishimoto, M. Itou, T. Miyatake, T. Ikariya, R. Noyori, Macromolecules
1995, 28, 6662–6666.
[46] A. Nowotny, Basic Exercises in Immunochemistry, Springer, New York,
1969.
[22] K. Matsuura, S. Furuno, K. Kobayashi, Chem. Lett. 1998, 27, 847–848.
[23] T. Hasegawa, S. Kondoh, K. Matsuura, K. Kobayashi, Macromolecules
1999, 32, 6595–6603.
[47] A. Imberty, E. Bettler, M. Karababa, K. Mazeau, P. Petrova, S. Pꢃrez in Per-
spectives in Structural Biology (Eds.: M. Vijayan, N. Yathindra, A. S. Kolas-
kar), Indian Academy of Sciences and Universities, Hyderabad, 1999,
pp. 392–409.
[24] A. Takasu, T. Houjyou, Y. Inai, T. Hirabayashi, Biomacromolecules 2002, 3,
775–782.
[25] A. Takasu, S. Horikoshi, T. Hirabayashi, Biomacromolecules 2005, 6,
2334–2342.
[48] A. Imberty, S. Pꢃrez, Chem. Rev. 2000, 100, 4567–4588.
[49] A. Isopo, R. Caminiti, R. d’Amato, A. Furlani, M. V. Russo, J. Macromol.
Sci. Part B 2003, 42, 1061–1083.
[26] B. D. Polizzotti, K. L. Kiick, Biomacromolecules 2006, 7, 483–490.
[27] I. Otsuka, T. Hongo, H. Nakade, A. Narumi, R. Sakai, T. Satoh, H. Kaga, T.
Kakuchi, Macromolecules 2007, 40, 8930–8937.
[28] E. Yashima, K. Maeda, H. Iida, Y. Furusho, K. Nagai, Chem. Rev. 2009, 109,
6102–6211.
Received: August 5, 2010
[29] N. Sharon, H. Lis, Lectins, 2nd ed., Springer, Netherlands 2003.
Published online on November 4, 2010
2408
www.chembiochem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 2399 – 2408