Investigation of the interaction between peanut agglutinin and synthetic glycopolymeric multivalent ligands

Article abstract of DOI:10.1039/b411555b

The interaction between synthetic glycoplymers bearing β-D-galactose side groups and the lectin peanut agglutinin (PNA) was investigated by UV-difference spectroscopy and isothermal titration calorimetry (ITC). UV-difference spectroscopy indicated that the polymer-lectin interaction was stronger than that between PNA and either the corresponding monomer, D-galactose or D-lactose. The thermodynamics of binding (K, ΔG, ΔH, ΔS and n) were determined from ITC data by fitting with a two-site, non-cooperative binding model. It was found that the glycopolymer displayed around a 50 times greater affinity for the lectin than the parent carbohydrate, and around 10 times greater than the monomer, on a valency-corrected basis. Binding was found to be entropically driven, and was accompanied by aggregation and precipitation of protein molecules. Furthermore, interesting differences between polymers prepared either from deacetylated monomers, or by deacetylation of pre-formed polymers, were found. The Royal Society of Chemistry 2005.

Full text of DOI:10.1039/b411555b

View Article Online / Journal Homepage / Table of Contents for this issue
A R T I C L E
Investigation of the interaction between peanut agglutinin and
synthetic glycopolymeric multivalent ligands†
Moira Ambrosi,a Neil R. Cameron,*a Benjamin G. Davis*b and Snjezana Stolnikc
a Department of Chemistry and Interdisciplinary Research Centre in Polymer Science and
Technology, University of Durham, South Road, Durham, UK DH1 3LE.
E-mail: n.r.cameron@durham.ac.uk
b Department of Chemistry, University of Oxford, Chemical Research Laboratory,
Mansfield Road, Oxford, UK OX1 3TA. E-mail: ben.davis@chem.ox.ac.uk
c School of Pharmaceutical Sciences, University of Nottingham, University Park,
Nottingham, UK NG7 2RD
Received 27th July 2004, Accepted 21st February 2005
First published as an Advance Article on the web 7th March 2005
The interaction between synthetic glycoplymers bearing b-D-galactose side groups and the lectin peanut agglutinin
(PNA) was investigated by UV-difference spectroscopy and isothermal titration calorimetry (ITC). UV-difference
spectroscopy indicated that the polymer–lectin interaction was stronger than that between PNA and either the
corresponding monomer, D-galactose or D-lactose. The thermodynamics of binding (K, DG, DH, DS and n) were
determined from ITC data by fitting with a two-site, non-cooperative binding model. It was found that the
glycopolymer displayed around a 50 times greater affinity for the lectin than the parent carbohydrate, and around
10 times greater than the monomer, on a valency-corrected basis. Binding was found to be entropically driven, and
was accompanied by aggregation and precipitation of protein molecules. Furthermore, interesting differences between
polymers prepared either from deacetylated monomers, or by deacetylation of pre-formed polymers, were found.
system,15 have been reported so far. These ITC studies have
Introduction
been limited to small or dendritic glycoconjugates possessing
Protein–carbohydrate interactions are involved in a myriad of
six or fewer carbohydrate residues.13 Remarkably, although
human biological processes, including the initiation of infection
haemagglutination inhibition studies (HIA),16 enzyme-linked
and disease, trafficking and clearance of glycoproteins, immune
lectin assays (ELLA),17,18 enzyme-linked immunosorbent assays
defence, fertilisation, recruitment of leukocytes to inflammatory
(ELISA)19 and surface plasmon resonance (SPR)17 studies have
sites, and cancer malignancy and metastasis.1,2 Synthetic sac-
shown the great potential of synthetic glycopolymers‡ as high-
charide ligands could therefore find wide employment in the
affinity polyvalent ligands, calorimetrically derived thermody-
investigation of key biochemical phenomena and in the design
namic data of their association with lectins have never been
of new therapeutics able to, for example, inhibit pathogen3
reported. Reasons for this lack of study may reside in the
or toxin4 carbohydrate-mediated adhesion or as efficient cell-
complexity of the system and the difficulties associated with
specific macromolecular drug carriers.5 However, attempts to
the exact knowledge of the ligand composition and structural
synthesise such ligands have often been frustrated by the
features. Furthermore, polyvalent ligand–lectin interactions
low intrinsic affinity of the protein–monovalent carbohydrate
often proceed through an associated aggregative mechanism in
interaction. Typical binding constants of monovalent mono-
which each ligand cross-links distinct protein molecules.20 This
and oligo-saccharide ligands to specific lectins6 are in the range
phenomenon may result in the precipitation of the aggregates,
103–106 M−1. Nature has overcome this intrinsic weakness of
which, in turn, may hamper the ITC experiment. As state
binding through multivalency, both in ligand and receptor.7
functions the thermodynamic parameters thus determined are
The activity enhancement of multivalent carbohydrate ligands
the sum of all processes occurring in the reaction cell.
has been termed the ‘cluster glycoside effect’.8,9 However, the
We have recently reported21 the synthesis of polymethacrylates
physical origin of the phenomenon is still not well understood.10
bearing D-galactose units, poly-[(b-D-galactopyranosyl)oxyethyl
It has been repeatedly pointed out that both structural and
methacrylate] (pGalEMA). As shown in Scheme 1, the gly-
energetic aspects of the lectin–carbohydrate interaction have
copolymer pGalEMA was obtained following two comple-
to be elucidated, if a deep understanding of the process is to
mentary routes: novel polymerisation of the deprotected gly-
be achieved.2,11 Of the techniques currently in use for protein–
comonomer GalEMA (route A) and the more traditionally
ligand association, only isothermal titration microcalorimetry
employed deprotection of the corresponding peracetylated
(ITC) allows the direct determination of the delicate balance
polymer pAcGalEMA (route B). Here we report an ITC study
of thermodynamic binding parameters K, DG, DH, DS and n
of these polymers, which is, to the best of our knowledge, the
(stoichiometry of binding, carbohydrate : protein) in a single
first calorimetric study of the interaction between a lectin and a
experiment (with techniques such as surface plasmon resonance,
synthetic polymeric glycoconjugate.
DH and DS are typically derived indirectly by means of the van’t
Hoff equation).
Despite their clear relevance to native interactions, few
Results and discussion
Peanut agglutinin, PNA,22,23 was used as the receptor; the
polymers bearing b-D-galactose units (pGalEMA-A,B; see
calorimetric studies of multivalent carbohydrate–protein
interactions,9,10,12–14 and indeed few for any multivalent ligand
† Electronic supplementary information (ESI) available: model used to
calculate thermodynamic binding parameters. See http://www.rsc.org/
suppdata/ob/b4/b411555b/
‡ The term ‘glycopolymers’ relates to linear macromolecules bearing
repeat units with pendant saccharide functionality, and does not include
‘polymeric’ multivalent ligands such as dendrimers.
1 4 7 6
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 4 7 6 – 1 4 8 0
T h i s j o u r n a l i s
T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 5
©

View Article Online
Despite the limited solubility of PNA in citrate buffer, this
medium was used in order to remove any possible contribution
of buffer protonation effects to the enthalpic term (for citrate
buffer,26 DHion = 0 kJ mol−1). Thermodynamic parameters are
state functions and protein–ligand complexation, changes in
protonation state or coupled equilibria are all reported as a single
binding enthalpy. Typical outputs of raw microcalorimetry data,
binding isotherms and best curve fitting are shown in Fig. 2.
The raw data are of rather poor quality, due to the relatively low
protein concentration used. This leads to a lower than optimum
value of c ([PNA monomer]·K), however one that for the
polymeric systems is still within the accepted limits for ITC (2.6
for pGalEMA-A).27 Vital to the determination of meaningful
data in this glycopolymeric system, thermodynamic parameters
(Table 1) were determined by non-linear least-squares analysis
of the binding isotherms, assuming a non-cooperative, bimodal
binding model.28 Interestingly, within the model, changes in
two enthalpic parameters (DH1, DH2) occur that may be best
described as resulting from two distinct phases: an initial
associative phase (DH1, n) and a minor secondary (DH2, ns)
phase. These two phases to date have not been distinguished
in lectin binding, and their physical meaning is not yet known;
however, we suggest that the first phase is the binding event
between the polymer and the lectin, and the second is the
precipitation of polymer–lectin conjugates. It is noteworthy that
a single phase binding model could not be made to fit to the data.
In this study the secondary phase contributes to a minor extent
(vide infra) and the following results are largely considered in
terms of DH1 and n.
Scheme 1 Polymer and monomer synthesis.21 Reagents and conditions:
(i), AgOTf 1.2 equiv., dry DCM, under N2, −40 ◦C, 48 h; (ii), NaOCH3
[cat.], dry CH3OH, RT, under N2; (iii), AIBN, CHCl3, 65 ◦C, 48 h; (iv),
K2S2O8, H2O : CH3OH (4 : 1), 65 ◦C, 48 h; (v), NaOCH3 [cat.], CH3OH :
CHCl3 (1 : 1), RT, under N2.
Scheme 1) together with the corresponding deprotected
monomer (GalEMA) and D-galactose itself, were tested as
ligands. The carbohydrate–protein interaction was probed in
two ways. Initially, qualitative information was obtained by
UV-difference spectroscopy. Owing to the presence of tyrosine
residues in the binding site,22 PNA develops a strong UV-
difference spectrum on ligand binding.24,25 Spectra obtained
using D-galactose, D-lactose, GalEMA and pGalEMA-A (Fig. 1)
indicated clear binding interactions with increasing intensity
in the order: D-galactose, GalEMA, D-lactose, pGalEMA-A.
Consistent with a similar mode of binding within PNA, the
addition of GalEMA and pGalEMA to the lectin solution
generated a UV-difference spectrum showing a maximum at
approximately the same wavelength (ca. 285 nm). Moreover,
broad qualitative agreement between the spectroscopic and
calorimetric measurements (vide infra) further confirmed ob-
servation of correlated binding events.
Fig. 2 (a) Experimental calorimetric data for the isothermal titration
at 298 K of pGalEMA into PNA. [PNA] = 3 mg ml−1, [ligand] =
0.55 mg ml−1. The solutions were in 20 mM citrate buffer containing
150 mM NaCl, pH 7.4. (b) Corresponding binding isotherm and best fit
curve to the data.
As expected, monovalent ligands gave rise to enthalpically
driven protein–ligand interactions, exhibiting enthalpy–entropy
compensation.29 The favourable enthalpic contribution arises
from the formation of direct and water-mediated hydrogen
bonds and van der Waals interactions, while the negative
entropic term may be mainly attributed to the loss of degrees
of freedom of the ligand upon binding.14 Values of K and
n for D-galactose of approximately 1 × 103 M−1 and 1,
respectively, showed good agreement with those previously
reported.25 GalEMA exhibited a 5-fold enhancement of binding
affinity compared to D-galactose, due to an increase of −DH0
Fig. 1 UV-difference spectra of peanut agglutinin with galactose,
GalEMA, pGalEMA-A and lactose. [PNA] = 2.12 mg ml−1, [ligands] =
2 × 10−4 M. The solutions were in 0.01 M phosphate-buffered saline
(PBS) at pH 7.4.
Next, isothermal titration microcalorimetric determinations
of the binding affinity of D-galactose, GalEMA and pGalEMA-
A,B for PNA were performed. After investigation of a wide
variety of titration systems, optimal ITC data were obtained
from titration of carbohydrate ligands into PNA solutions.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 4 7 6 – 1 4 8 0
1 4 7 7

View Article Online
Table 1 Thermodynamic binding parameters of D-galactose, GalEMA and pGalEMA-A,B to peanut agglutinin.a
Ligand
n
K/M−1 × 10−3 DH10/kJ mol−1 DGo0bs/kJ mol−1 TDSo0bs/kJ mol−1 ns
DH02/kJ mol−1
Galactoseb
1.00 0.42
1.13 0.04
1.56 0.39 46.36 8.81
11.04 0.50
2.08 0.22
1.62 0.11
1.35 0.34
1.00 0.36
4.03 0.08
4.07 0.31
0.95 0.06
5.65 0.10
−22.97 0.87 −16.98 0.16
−39.41 0.36 −21.41 0.05
−5.79 0.19 −26.62 0.48
−19.84 0.52 −22.67 0.27
−23.94 0.75 −16.51 0.09
−10.80 0.19 −21.72 0.14
−8.30 0.29 −21.19 0.38
−6.42 0.18 −20.65 0.34
−15.76 0.18 −17.90 0.05
−42.64 2.57 −18.63 0.19
−5.99 1.03
−18.00 0.41
20.83 0.67
2.83 0.79
−7.43 0.84
10.92 0.33
12.89 0.67
14.23 0.52
2.14 0.23
0.96 0.35
1.21 0.16
1.95 0.1
9.19 2.85
1.04 0.04
1.44 0.17
1.94 0.58
1.85 0.62
4.99 0.66
2.99 0.73
−0.74 0.31
−7.58 0.10
−1.39 0.01
−0.93 0.02
−1.30 0.03
−2.36 0.007
−1.46 0.002
−1.20 0.01
−1.53 0.01
−4.56 0.07
GalEMAb
pGalEMA-Ab
pGalEMA-Bb
Galactosec
9.40 1.05
0.78 0.03
6.42 0.37
5.17 0.79
4.17 0.57
1.37 0.03
1.84 0.14
pGalEMA-Ac
pGalEMA-Ac
pGalEMA-Ad
pGalEMA-Ae
pGalEMA-Af
−24.01 2.76
a Since molar concentrations of the reactants rather than their activities were used, the subscript ‘obs’ is used to denote to the calculated binding
parameters. b [PNA] = 1.5 mg ml−1, 20 mM citrate buffer, pH 7.4. c [PNA] = 3 mg ml−1, 20 mM citrate buffer, containing 150 mM NaCl, pH 7.4.
d [PNA] = 3 mg ml−1, 20 mM citrate buffer, containing 150 mM NaCl, pH 7.4, heats of dilution have been subtracted. e [PNA] = 5.3 mg ml−1, 50 mM
Tris-HCl buffer, containing 500 mM MgCl2, 1 mM MnCl2, 1 mM CaCl2, pH 6.9. f [PNA] = 5.3 mg ml−1, 50 mM Tris-HCl buffer, containing 500 mM
MgCl2, 1 mM MnCl2, 1 mM CaCl2, pH 6.9, heats of dilution have been subtracted.
uncompensated by −DS0obs. This result is in agreement with the
that precisely and fully deprotected polymers and less precisely
deprotected, partially acetylated ones would possess different
functional properties.
previously reported specificity of PNA for sugars in the pyranose
form30 and may be due to a locked b-pyranoside configuration
and/or hydrogen bond acceptor and hydrophobic interactions
between the methacrylate aglycone and the lectin.
A decrease of the binding enthalpy can be considered as
the thermodynamic signature of an endothermic intermolecular
aggregative process superimposed on an exothermic ligand
binding event,9,13 since the latter may be assumed to proceed
with thermodynamic parameters equivalent to those of the
corresponding monovalent saccharide.10 This study for the first
time confirms this mode of action for a synthetic glycopolymer.
It is worthwhile to note that, for experiments performed in citrate
buffer, thermodynamic data determined by subtracting heats of
dilution (evaluated by means of blank titrations) from each bind-
ing heat or by subtracting the average dilution heat determined
from the end part of the release heat titration profile32 were sub-
stantially identical. Only a slight decrease of DG0obs was observed,
due to a slight decrease of the enthalpic term (Table 1). Therefore,
the concomitancy of the polymer deaggregation with the binding
event did not prevent the determination of thermodynamic
binding parameters. The thermodynamic parameters evaluated
for polymeric ligands, and especially those for pGalEMA-A, can
be related to an affinity enhancement mainly achieved through
the formation of cross-linked ligand–protein complexes,10,33
which phase separate during the titration. Clear parallels can
be drawn with the hydrophobic effect responsible for the
micellisation of surfactants in aqueous media.34 The significant
favourable entropic contributions are possibly due to the release
of structurally ordered water molecules from the hydration
shells of protein tetramers brought close to one another by the
aggregation, as well as beneficial ligand degeneracy.32 Indeed,
each PNA tetramer is surrounded by 518 molecules of water.23
The aggregation hypothesis was also supported by dynamic light
scattering (DLS) measurements, which showed a large increase
in scattered light intensity (ca. 1000 → 2500) after addition of
the first aliquot of polymer to the lectin solution. In contrast,
only a slight increase in scattered light intensity (ca. 350 → 500)
was observed in analogous titrations with either D-galactose or
GalEMA, and this occurred at a ligand : lectin binding site
stoichiometry of around 1.35 Precipitation of the glycopolymer–
lectin complex may preclude biomedical applications of these
synthetic glycopolymeric ligands.
Significantly different thermodynamics were determined for
polymeric ligands. On a valency-corrected basis, pGalEMA-
A exhibited a 50-fold affinity gain compared to D-galactose
and a 10-fold enhancement with respect to the corresponding
monovalent ligand, GalEMA. There is a significant error
associated with the K value for pGalEMA-A (entry 3 in Table 1).
This is due to the aggregation process (vide infra) that hampers
integration of the peaks in the ITC trace. The error is consid-
erably larger than those associated with the K values for the
binding of low molar mass species, for which precipitation does
not occur. Interestingly, a significant diminution of enthalpy
was determined with the polymeric ligands: −DH0 decreased
from ca. −23 kJ mol−1 for D-galactose to ca. −6 kJ mol−1 for
pGalEMA-A. By contrast, the entropic contribution became
favourable, varying from negative for the monovalent ligands
to positive for the polymer. In particular, for pGalEMA-A
approximately 75% of DG0obs arose from the favourable entropy
of binding. The determination of these thermodynamic data
thus confirms directly the predicted significance of entropic
contributions in such multivalent systems.31
It is interesting to compare the results obtained for the
two polymeric ligands pGalEMA-A and pGalEMA-B. Affin-
ity enhancements were still detected for pGalEMA-B when
compared to D-galactose and monomeric GalEMA of ap-
proximately 10-fold and 2-fold, respectively. These interactions
were similarly characterised by a decrease of −DH0 and a
favourable DS0obs. However, the overall affinity for PNA by the
less well-defined21 glycopolymer pGalEMA-B was much lower
than that of pGalEMA-A. The binding constant diminished by
approximately 5-fold, due to the reduction of the favourable
entropic term, TDS0obs, from ca. 20 kJ mol−1 for pGalEMA-A to
ca. 3 kJ mol−1 for pGalEMA-B. Concomitantly, DH0 decreased
to ca. −20 kJ mol−1, giving a more favourable contribution than
that reported for pGalEMA-A. The stoichiometries observed for
the protein–carbohydrate interactions were also, intriguingly,
dramatically different, being 1.5 and 11 for pGalEMA-A
and pGalEMA-B, respectively. The polymer deprotection step
involved in the synthesis of pGalEMA-B is non-quantitative,21
resulting in a polymer material possessing partially and/or
fully protected sugar residues, unable to bind efficiently to the
lectin. The lower affinity of pGalEMA-B relative to pGalEMA-
A and the apparently higher stoichiometry required for the
interaction may therefore be understood by considering that
a smaller proportion of the Gal residues can, in principle, be
recognised by the protein. This confirmed our expectations21
At least three phenomena have to be taken into account in the
search for a justification of the determined protein-pGalEMA-
A binding stoichiometry of approximately 1.5. First of all,
the proximity of carbohydrate residues to each other within
the polymer might prevent the interaction of all the sugars,
due to steric hindrance. Secondly, binding of segments of the
polymer chain may alter the flexibility of the unbound segments,
impeding the attainment of full binding site saturation. Thirdly,
the observed precipitation of the aggregates would decrease the
1 4 7 8
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 4 7 6 – 1 4 8 0

View Article Online
number of binding sites available in solution. Rao et al.15 recently
measured the thermodynamics of the interaction of a trivalent
derivative of DADA (D-Ala-D-Ala) and of Lac-Rꢀd-Lac with a
trivalent and a divalent ligand of vancomycin, respectively. The
observed affinity enhancements compared to the monovalent
analogues were in both cases certainly reached by means of an
intramolecular mechanism. Both −DH and −DS were found
to scale proportionally to the number of epitopes, behaviour
not observed in this study. Despite the possibility that the
glycopolymers might complex different binding sites on the
same receptor PNA multimer molecule,36 the thermodynamic
data, the stoichiometry of the interaction, DLS measurements
and the relative lack in the polymeric structure of portions that
could lead to favourable interactions with protein surfaces in
a chelation process, all indicate that the polymers analysed in
this study most likely act mainly by aggregating distinct PNA
tetramers.13
Finally, addition of salt was used to probe contributory
electrostatic effects. While D-galactose binding was virtually un-
affected, the presence of NaCl critically affected the interactions
of GalEMA and pGalEMA. PNA–GalEMA binding was no
longer detectable27 while the binding constant of pGalEMA-A
decreased ca. 7-fold to ca. 6 × 103 M−1. The enthalpy change
was not strongly affected, while TDS0obs almost halved indicating
reduced entropically-driven, aggregation processes. Indeed, a
further increase in salt concentration (50 mM Tris-HCl, 500 mM
MgCl2, 1 mM CaCl2, 1 mM MnCl2) further reduced binding to
K ≈ 1 × 103 M−1. It may be that electrostatic protein surface–
surface interactions23 are decreased at higher ionic strengths.
ous size exclusion chromatography: pGalEMA-A, number-
average molecular weight (Mn) = 461 000 and polydispersity
index (Mw/Mn) = 2.21; pGalEMA-B, Mn = 2984 and polydis-
persity index = 2.15. UV-difference spectra were recorded on a
Unicam UV2-100 spectrometer, using masked semimicro cells of
1 cm path-length. Aliquots of 500 ll of peanut agglutinin (PNA,
Sigma, affinity-purified, salt-free lyophilised powder, activity
< 0.1 lg ml−1) solutions (ca. 2 mg ml−1, ca. 75 lM when 27 000 g
mol−1 was used as the molecular weight of PNA monomer,
in PBS at pH 7.4) were added to both sample and reference
cuvettes, and the baseline recorded. An aliquot (4 ll) of a 2.5 ×
10−2 M solution of the ligand was added to the sample cell,
while the reference received the same amount of buffer solution.
Isothermal titration microcalorimetry (ITC) was performed
using a Thermal Activity Monitor (TAM 2277, Thermometric
AB, Sweden) operated at 298 K. Two sets of experiments were
carried out preparing samples of PNA and ligands in 20 mM
citrate buffer, containing or not 150 mM NaCl. The samples were
adjusted to pH 7.4 ( 0.05) using dilute NaOH. The addition of
salt allowed the doubling of the lectin concentration in solution.
In each experiment a 2.5 ml sample of a solution of PNA (1.5
or 3 mg ml−1) was placed in a sample cell and inserted into the
instrument. Once thermal equilibrium was reached, the titration
was performed by consecutive injections (25 injections of 20 ll
each) of ligand solution (0.275 or 0.55 mg ml−1). The titrant
was added by means of a Hamilton microlab syringe mounted
on a computer-operated syringe driver (Lund 6100 syringe
R
pump). The experimental method set up via the Digitamꢀ
4 software allowed for data collection over a 15 min period
for the injection and a 5 min baseline period before the next
injection. This was found to be adequate for the interaction to
proceed to completion at each injection and reach the baseline
before the next addition. Data presented here are the mean of
a minimum of three replicate titrations for each experiment.
Heats of dilution/mixing were determined in blank titrations
injecting aliquots of ligand solution into the appropriate buffer
without the lectin. The signs of the released heat values were
reversed for data analysis since the output from the instrument
is from the perspective of the equipment and not the system
under study. The equilibrium binding parameters K, DH0 and
n were determined by non-linear least-squares fitting using the
routines available within Origin 5.0 (MicroCal Software). The
integrated heats of binding were corrected for the enthalpy of
dilution, DHd, before calculation of the binding parameters. The
heat of dilution can be evaluated either by performing a blank
titration, i.e. injection of ligand solution into buffer without
PNA, or from the end part of the released heat titration profile.
In the latter case an average dilution heat is calculated and
subtracted from the observed heat value at each data point.32
In this study both methods were employed. When pGalEMA
was used as the titrant, c values ([PNA monomer]·K) were
≥ 0.9.27 Concentrations of polymer and PNA are both expressed
in terms of monomeric units. The free energy of complex
formation (DG0obs) was calculated from eqn. (1):
It should be noted that, although Tris-HCl exhibits DHion
=
46.0 kJ mol−1 and buffer protonation effects may affect DH and
DS values, favourable enthalpic and entropic contributions were
still observed. An increase of n (which is not buffer dependent)
up to a value of 4 was determined; a weaker interaction with
little or no cross-linking would require an excess of polymer to
reach protein binding site saturation.
Conclusions
For the first time the interaction of a lectin with synthetic
glycopolymeric ligands was investigated by isothermal titration
microcalorimetry. A 50-fold affinity enhancement of the poly-
mer compared to the monovalent saccharide D-galactose was
observed. The thermodynamics confirm, at least in the case of
PNA, previous suggestions that polyvalent carbohydrates bind
to lectins mainly by cross-linking distinct protein molecules.
The polymer activity was strongly influenced by the presence
of salt, indicating that electrostatic interactions were involved
in the binding process, either by affecting polymer nanoparticle
dissociation, or by neutralising charges on the protein surface
somehow involved in the chelate/aggregative process. It should
be stressed that the findings described here provide initial clues
as to the nature of the binding between lectins and synthetic
glycopolymers. Further experiments probing the influence of
glycopolymer valency on binding, using techniques such as
haemagglutination assays and surface plasmon resonance in
addition to ITC, are required to provide a fuller understanding.
In addition, studies involving a monovalent lectin such as
galectin-3 37 should be conducted to determine the influence of
precipitation on the binding process.
DG0obs = −RT lnKobs
(1)
and the entropy of binding determined by eqn. (2):
TDS0obs = DH0 − DGo0bs
.
(2)
Acknowledgements
Experimental
EPSRC is thanked for a studentship to M. A.
The polymers pGalEMA-A,B were prepared as described
previously.21 Molecular weight data§ were determined by aque-
References
§ Number- and weight-average molecular weights (Mn and Mw, respec-
tively) are used to characterise synthetic polymers which are obtained
as a distribution of chain lengths. The breadth of the distribution is
indicated by the polydispersity (Mw/Mn).
1 A. Varki, Glycobiology, 1993, 3, 97; H. Lis and N. Sharon, Chem.
Rev., 1998, 98, 637.
2 M. Mammen, S.-K. Choi and G. M. Whitesides, Angew. Chem., Int.
Ed., 1998, 37, 2754.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 4 7 6 – 1 4 8 0
1 4 7 9

View Article Online
3 P. M. Rendle, A. Seger, J. Rodrigues, N. J. Oldham, R. R. Bott, J. B.
Jones, M. M. Cowan and B. G. Davis, J. Am. Chem. Soc., 2004, 126,
4750.
4 P. I. Kitov and D. R. Bundle, J. Chem. Soc., Perkin Trans. 1,
2001, 838.
5 B. G. Davis and M. A. Robinson, Curr. Opin. Drug Disc. Dev., 2002,
5, 279.
6 D. K. Mandal, N. Kishore and C. F. Brewer, Biochemistry, 1994,
33, 1149.
7 R. E. McDonald, D. Hughes and B. G. Davis, Angew. Chem., Int.
Ed., 2004, 43, 3025.
8 Y. C. Lee, R. R. Townsend, M. R. Hardy, J. Lo¨nngren, J. Arnarp, M.
Haraldsson and H. Lo¨nn, J. Biol. Chem., 1983, 258, 199; Y. C. Lee,
Neoglycoconjugates: General Considerations, in Neoglycoconjugates:
Preparation and Applications, ed. Y. C. Lee and R. T. Lee, Academic,
San Diego, 1994.
19 M. Oubihi, K. Kitajima, K. Kobayashi, T. Adachi, N. Aoki and T.
Matsuda, Anal. Biochem., 1998, 257, 169; M.-G. Baek and R. Roy,
Biomacromolecules, 2000, 1, 768.
20 C. F. Brewer, Adv. Exp. Med. Biol., 2001, 491, 17.
21 M. Ambrosi, A. Batsanov, N. R. Cameron, B. G. Davis, J. A. K.
Howard and R. Hunter, J. Chem. Soc., Perkin Trans. 1, 2002, 45.
22 N. M. Young, A. Z. Johnston and D. C. Watson, Eur. J. Biochem.,
1991, 196, 631.
23 R. Banerjee, K. Das, R. Ravishankar, K. Suguna, A. Surolia and M.
Vijayan, J. Mol. Biol., 1996, 259, 281.
24 W. W. Fish, L. M. Hamlin and R. L. Miller, Arch. Biochem. Biophys.,
1978, 190, 693; K. J. Neurohr, N. M. Young and H. H. Mantsch,
J. Biol. Chem., 1980, 255, 9205; K. J. Neurohr, D. R. Bundle, N. M.
Young and H. H. Mantsch, Eur. J. Biochem., 1982, 123, 305.
25 M. Caron, J. Ohanessian, J. Becquart and H. Gillier-Pandraud,
Biochim. Biophys. Acta, 1982, 717, 432.
9 J. J. Lundquist and E. J. Toone, Chem. Rev., 2002, 102, 555.
10 S. M. Dimick, S. C. Powell, S. A. McMahon, D. N. Moothoo, J. H.
Naismith and E. J. Toone, J. Am. Chem. Soc., 1999, 121, 10 286.
11 G. Molema and D. K. F. Meijer, Adv. Drug Deliv. Rev., 1994, 14, 25;
B. G. Davis, Chem. Rev., 2002, 102, 579.
12 R. V. Weatherman, K. H. Mortell, M. Chervenak, L. L. Kiessling
and E. J. Toone, Biochemistry, 1996, 35, 3619; J. J. Lundquist, S. D.
Debenham and E. J. Toone, J. Org. Chem., 2000, 65, 8245; T. K.
Dam, R. Roy, S. K. Das, S. Oscarson and C. F. Brewer, J. Biol.
Chem., 2000, 275, 14 223; T. K. Dam, R. Roy, D. Page´ and C. F.
Brewer, Biochemistry, 2002, 41, 1351; T. K. Dam, R. Roy, D. Page´
and C. F. Brewer, Biochemistry, 2002, 41, 1359; T. Christensen and
E. J. Toone, Methods Enzymol., 2003, 362, 486; W. B. Turnbull, B. L.
Precious and S. W. Homans, J. Am. Chem. Soc., 2004, 126, 1047.
13 J. B. Corbell, J. J. Lundquist and E. J. Toone, Tetrahedron: Asymme-
try, 2000, 11, 95.
14 T. K. Dam and C. F. Brewer, Chem. Rev., 2002, 102, 387.
15 J. Rao, J. Lahiri, L. Isaacs, R. M. Weis and G. M. Whitesides, Science,
1998, 280, 708; J. Rao, L. Yan, J. Lahiri, G. M. Whitesides, R. M.
Weis and H. S. Warren, Chem. Biol., 1999, 6, 353.
16 A. Tsuchida, K. Kobayashi, N. Matsubara, T. Muramatsu, T. Suzuki
and Y. Suzuki, Glycoconjugate J., 1998, 15, 1047; T. Hasegawa, S.
Kondoh, K. Matsuura and K. Kobayashi, Macromolecules, 1999,
32, 6595.
26 J. J. Christensen, L. D. Hansen and R. M. Izatt, Handbook of Proton
Ionization Heats and Related Thermodynamic Quantities, John Wiley,
New York, 1976.
27 T. Wiseman, S. Williston, J. F. Brandts and L. N. Lin, Anal. Biochem.,
1989, 179, 131.
28 T. Ehtezazi, U. Rungsardthong and S. Stolnik, Langmuir, 2003, 19,
9387.
29 B. A. Williams, M. C. Chervenak and E. J. Toone, J. Biol. Chem.,
1992, 267, 22 907; A. Cooper, Curr. Opin. Chem. Biol., 1999, 3, 557.
30 R. Lotan, E. Skutelsky, D. Danon and N. Sharon, J. Biol. Chem.,
1975, 250, 8518.
31 N. K. Vyas, M. N. Vyas, M. C. Chervenak, M. A. Johnson, B. M.
Pinto, D. R. Bundle and F. A. Quiocho, Biochemistry, 2002, 41,
13 575; P. I. Kitov and D. R. Bundle, J. Am. Chem. Soc., 2003, 125,
16 271; A. Yung, W. B. Turnbull, A. P. Kalverda, G. S. Thompson,
S. W. Homans, P. Kitov and D. R. Bundle, J. Am. Chem. Soc., 2003,
125, 13 058.
32 T. Govender, T. Ehtezazi, S. Stolnik, L. Illum and S. S. Davis, Pharm.
Res., 1999, 16, 1125.
33 L. R. Olsen, A. Dessen, D. Gupta, S. Sabesan, J. C. Sacchettini and
C. F. Brewer, Biochemistry, 1997, 36, 15 073; J. C. Sacchettini, L. G.
Baum and C. F. Brewer, Biochemistry, 2001, 40, 3009.
34 E. Fisicaro and E. Pelizzetti, Trends Phys. Chem., 1994, 4, 1.
35 M. Ambrosi, Ph.D. thesis, University of Durham, 2002.
36 M. Kanai, K. H. Mortell and L. L. Kiessling, J. Am. Chem. Soc.,
1997, 119, 9931.
17 E. Arranz-Plaza, A. S. Tracy, A. Siriwardena, J. M. Pierce and G.-J.
Boons, J. Am. Chem. Soc., 2002, 124, 13 035.
18 R. Auzely-Velty, M. Cristea and M. Rinaudo, Biomacromolecules,
2002, 3, 998.
37 H. J. Gabius, H. C. Siebert, S. Andre, J. Jimenez-Barbero and H.
Rudiger, Chembiochem., 2004, 5, 741.
1 4 8 0
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 4 7 6 – 1 4 8 0