Probing bacterial-toxin inhibition with synthetic glycopolymers prepared by tandem post-polymerization modification: Role of linker length and carbohydrate density

Article abstract of DOI:10.1002/anie.201202945

Probing the depths: A tandem post-polymerization modification strategy was used to systematically probe the multivalent inhibition of a bacterial toxin as a function of linker length (see scheme), carbohydrate density, and glycopolymer chain length. Guided by structural-biology information, the binding-pocket depth of the toxin was probed and used as a means to specifically improve inhibition of the toxin by the glycopolymer. Copyright

Full text of DOI:10.1002/anie.201202945

Angewandte
Chemie
DOI: 10.1002/anie.201202945
Glycopolymer Binding
Probing Bacterial-Toxin Inhibition with Synthetic Glycopolymers
Prepared by Tandem Post-Polymerization Modification: Role of
Linker Length and Carbohydrate Density**
Sarah-Jane Richards, Mathew W. Jones, Mark Hunaban, David M. Haddleton, and
Matthew I. Gibson*
Protein-carbohydrate interactions mediate many critical
biological recognition processes, such as those involved in
cell signaling, fertilization, and inflammation, as well as the
adhesion of viruses and bacterial toxins.[1] The proteins
responsible for deciphering this information are termed
lectins, which specifically (and noncovalently) bind carbohy-
drates based on their branching pattern, stereochemistry, and
chemical functionality.[2] The protein–saccharide interactions
are usually weak, but are amplified by clustered saccharides,
resulting in a binding constant which is greater than the
simple sum of the total number of ligands. This observation is
referred to as the “cluster glycoside” effect.[3] Considering this
information, glycopolymers are attractive materials to inter-
act with lectins and have been shown to display binding
affinities several orders of magnitude greater than a single
carbohydrate molecule.[4] One of the most attractive applica-
tions of glycopolymers is to interfere with the binding of
lectins from a pathogen to the host organism and therefore
prevent infection from occurring; this is known as anti-
adhesion therapy.[5]
Anti-adhesion therapy is not limited to polymeric glyco-
conjugates, and many high-affinity, small-molecule inhibitors
of lectin binding are known.[6] The increased affinity of these
small-molecule inhibitors is often because of favorable
interactions within the sugar-binding pocket, obtained by
structural biology studies.[6–7] Despite the wealth of structural
information available, this design approach has rarely been
applied to the design of polymeric inhibitors.[8] The toxin Ctx
secreted by Vibrio cholerae, which is the causative agent of
cholera, is a multimeric AB5 lectin-like complex. The five
B subunits bind GM1 gangliosides (Galb1-3GalNAcb1-
4(Neu5Aca2-3)-Gal-b1-4Glc ceramide) present on the sur-
face of the intestinal epithelium cells, initiating a series of
events that results in the cholera symptoms. Multivalent,
glycosylated STARFISH dendrimers successfully protected
cells from similar toxins.[9] Polizzotti and Kiick have demon-
strated that galactose-functionalized poly(l-glutamic acid)
(PLG) is an effective inhibitor of Ctx.[10] Interestingly, low-
ering the density of the galactose residues on the PLG
backbone from approximately 50% to 10% gave a dramatic
increase in activity. The inhibitory activity was further
enhanced when the spacing between PLG and galactose was
increased, making the binding site more accessible, and was
also related to the depth of the Ctx galactose binding pocket,
which was estimated as being 16 ꢀ. Others have speculated
that longer linkers could enhance binding affinity, but these
studies were not systematic or relied on a simple turbidim-
etry-based assay. The thermodynamics of this process have
been evaluated in detail by Kane.[11]
A challenge associated with studying multivalent inter-
actions is the synthesis of libraries of glycopolymers with
precise control over the chain length, carbohydrate density,
and nature of the linker. Direct polymerization of two
glycosylated monomers is unlikely to lead to polymers with
identical degrees of polymerization, which considering the
importance of valency in carbohydrate-lectin interactions, is
a major challenge.[12] The synthesis of glycopolymers by post-
polymerization modification has attracted much interest,
especially with the development of “click”-type reactions.[13]
We have synthesized several classes of glycopolymers through
cycloaddition of glycosyl azides with poly(propargylmetha-
crylate) scaffolds[4a,14] and used these to inhibit the binding of
gp120 (derived from HIV) to DC-SIGN (one of its targets in
humans)[15] The direct polymerization of alkyne (or alkene)
monomers by radical methods is generally not possible
because of competing side reactions,[13b] and the incorporation
of co-monomers has the problem of obtaining identical length
polymers. To address this, we introduced tandem post-
polymerization modification to introduce incompatible func-
tionality and control functional group density.[16] Using
poly(pentafluorophenyl methacrylate) as the template scaf-
fold, allylamine was introduced as a reactive, orthogonal
“handle” and the remaining ester groups reacted with 2-
hydroxypropylamine to give a water soluble backbone. This
powerful method allows precise control over of a wide range
of chemical space.
[*] S.-J. Richards, Dr. M. W. Jones, M. Hunaban, Prof. D. M. Haddleton,
Dr. M. I. Gibson
Department of Chemistry, University of Warwick
Gibbet Hill Road, Coventry, CV4 7AL (UK)
E-mail: m.i.gibson@warwick.ac.uk
Homepage: http://www.warwick.ac.uk/go/gibsongroup
[**] Equipment used in this research was funded in part through
support from Advantage West Midlands (A.W.M.) Science City
Initiative and funded in part by the ERDF. D.M.H. is a Wolfson/
Royal Society Fellow and M.I.G. is a HEFCE Science City Senior
Fellow. S.-J.R. acknowledges the EPSRC funded MOAC DTC for
a studentship and M.W.J. acknowledges the EPSRC for a knowledge
transfer secondment.
Herein, we use tandem post-polymerization modification
to obtain glycopolymers with precisely controlled chain
length, carbohydrate density, and crucially, defined back-
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201202945.
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
bone–carbohydrate linker lengths. This series of polymers was
used to study the multivalent interactions between cholera
toxin and peanut agglutinin, to probe the impact of modulat-
ing the binding site complementarity on the inhibitory activity
of the glycopolymers. This unique combination of structural
biology with materials science gives insights into the cluster
glycoside effect and will allow design of active inhibitors.
The tandem post-polymerization modification strategy is
outlined in Scheme 1. This approach is advantageous because
it allows us to explore a large area of chemical space (through
side-chain modification) while ensuring that the average
degree of polymerization is identical within each series.
discriminate between functionalization and hydrolysis, and
therefore can give false positive results. The alkyne functional
(co)polymers were subsequently functionalized with 1-azido-
b-d-galactose (GalN3) using the CuI/tris(benzyltriazolyl-
methyl)amine (TBTA) catalyst system, in dimethyl sulfoxide
(DMSO)[4a] to yield homogenous glycopolymers, with no
detectable alkyne or azide groups by NMR or IR spectros-
copy. Over the course of the tandem post-polymerization
functionalization, there was no evidence of fractionation, and
all the polymers obtained were well-defined with polydis-
persity indexes (PDIs) below 1.3 (SEC data are included in
the Supporting Information). Table 2 summarizes the diverse
glycopolymer library that was
synthesized and used for bio-
logical evaluation, below.
The aim of this investiga-
tion was to probe the effect of
carbohydrate-binding
site
accessibility on the measured
affinity between multivalent
glycopolymers
and
their
target lectins. The B subunit
domain of cholera toxin was
chosen because it is nontoxic,
Scheme 1. Synthesis of glycopolymer libraries. 1) amine (variable amounts)/triethylamine (TEA; 1 equiv)/
dimethylformamide (DMF), 5 h.; 2) GalN3 (1.5 equiv)/CuBr/TBTA, DMSO.
Two different amino-functional alkynes were selected as
the polymer-carbohydrate spacers. Propargyl amine (Alkyne
1) acts as a short linker and a diethylene glycol group acts as
a longer spacer (Alkyne 2). The reactive polymeric precursor,
poly(pentafluorophenyl methacrylate), PPFMA, was synthe-
sized by copper-mediated controlled-radical polymerization
to give a series of polymers with varying degrees of
polymerization, Table 1. As previously reported,[17] size-
exclusion chromatographic (SEC) analysis of PPFMA gave
values of Mn in disagreement with the theoretical Mn. PPFMA
reacts readily with sterically unhindered amines, such as those
used here.[17]
Homopolymers were synthesized by addition of a three-
fold molar excess of Alkyne 1 or 2, as well as several
copolymers (see below for rationale) by addition of 10, 25,
and 50 mol% of Alkyne 2, followed by a three molar excess of
3-aminopropanol. Conversion of the pentafluorophenyl
(PFP) groups into amides was confirmed by IR spectroscopy
as the desired amide product (1675 cmꢀ1) and PFP ester
(1786 cmꢀ1) have unique vibrational frequencies. In all cases,
a small amount (< 5%) of carboxylic acid units can be seen at
1730 cmꢀ1, Figure 1. Fluorine NMR spectroscopy is often
used to determine conversion of PFP groups, but does not
Figure 1. Infrared spectra showing conversion of the PFP groups
(1786 cmꢀ1) into amides (1675 cmꢀ1). Carboxylic acid impurities are
also indicated. Copolymer (solid line) has two distinct amide peaks.
Table 2: Glycopolymers obtained by tandem modification.
[d]
[d]
Polymer
DP[a]
Linker[b]
Density[c]
Mn
Mw/Mn
GP1
GP2
GP3
GP4
GP5
GP6
GP7
GP8
GP9
18
33
70
18
33
70
33
33
33
Alkyne 1
Alkyne 1
Alkyne 1
Alkyne 2
Alkyne 2
Alkyne 2
Alkyne 2
Alkyne 2
Alkyne 2
100
100
100
100
100
100
50
5100
5340
6000
7250
10000
11900
10800
10700
10500
1.29
1.27
1.26
1.32
1.28
1.27
1.23
1.21
1.20
Table 1: PPFMA precursor polymers.
Polymer [M]:[I][a] Conv. [%][b] Mn(theo)
Mn(SEC)
Mw/Mn
DP[e]
[c]
[d]
[d]
P1
P2
P3
25
50
100
71
65
70
4500
8200
17640
7800
8800
11400
1.19
1.20
1.16
18
33
70
25
10
[a] Feed ratio of monomer to initiator. [b] Determined by 1H NMR
spectroscopy. [c] Theoretical Mn, calculated from the feed ratio and
percent conversion. [d] Determined by SEC in tetrahydrofuran (THF)
using polystyrene standards. [e] Theoretical number-average degree of
polymerization.
[a] Theoretical number-average degree of polymerization. [b] Alkyne used
to modify PPFMA scaffold. [c] Percent of repeat units functionalized with
a galactose unit. [d] Determined by SEC in DMF.
2
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
and a report by Polizzotti and Kiick showed that longer
linkers resulted in increased inhibitory activity of cholera
toxin.[10] Peanut agglutinin (PNA) was used as a control as it
also binds b-galactose, but its binding sites are surface
exposed.[18] It is critical to include this second lectin, because
it allows us to separate the effect of binding-site accessibility
from increased side chain flexibility; the latter has been
suggested to alter the binding of glycopolymers by allowing
access to a larger number of possible conformations and
leading to increased clustering. A recent study found that
longer polymer–carbohydrate linkers showed increased affin-
ity for the lectin, but the valency of each structure was not
identical and there was no consideration of the depth of the
lectin binding site.[19] Figure 2 shows schematic depictions of
the galactose-binding domains of Ctx and PNA in comparison
to the two linker systems used herein, which were selected to
mimic these distances.
Fluorescence-linked sorbent assays were used to measure
the inhibitory activity of the polymers. Briefly, microtitre
plates were functionalized with GM1 ganglioside, which binds
strongly to both lectins.[20] Fluorescein-labeled lectins (0.5 mm)
were then incubated with a dilution series of the polymers for
30 minutes at 378C, and the unbound lectins were washed
away. The total fluorescence was measured and inhibitory
curves constructed. Data is presented as the minimum
concentration required to inhibit 50% binding of the lectin
(MIC50), Figure 3.
Figure 3. Inhibitory activity of glycopolymers GP1–GP6 with A) PNA
and B) Ctx. MIC50 values are expressed as total galactose concentra-
tion.
For PNA (Figure 3A), there was a very weak effect of
polymer chain length on inhibitory activity, but the differ-
ences were not statistically significant. The inhibitory activity
of all the polymers is greatly increased relative to monovalent
galactose (circa 100-fold). Although there are many reports of
longer polymers having increased association constants with
lectins, we recently demonstrated that this does not necessa-
rily translate into improved inhibitory activity, in agreement
with the data presented herein.[21] Our previous work also
showed that above a certain chain length, no improved
inhibition (based on carbohydrate concentration) was mea-
sured, as seen with the current results. For Ctx inhibition
(Figure 3B), there was a similarly weak dependence on
polymer chain length, and no statistically significant differ-
Figure 2. Binding sites of PNA (A) and Ctx (B). W=water ligand. C,D) Polymer side-chain linkers. Arrows indicate the distance from the terminal
b-Gal residue to the far end of the native ligand. Distances for the protein-binding site of Ctx are taken from literature values.[10]
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
ences were recorded. However, there was a significant
decrease in the MIC50 values upon increasing the length of
the linker. This result agrees with the hypothesis that the
longer linker better mimics the native ligand, GM1. The use
of a hydrophilic ethylene glycol linker allows us to rule out
any hydrophobic interactions in the binding pocket. Because
we used the B subunit of Ctx that has only a single binding
site, the differences in inhibition from linker length cannot be
attributed to differences in cross-linking or intra-lectin bind-
ing-site spanning. These carefully chosen conditions and the
use of a control lectin (PNA) ensure that binding-pocket
depth is the only parameter contributing to increased activity.
Control experiments with an a-manno polymer showed no
inhibition in the concentration range tested (less than 100 mm
carbohydrate), highlighting the specificity of the interaction.
Previous research has shown that a decrease in galactose
density on the polymer backbone is accompanied by an
increase in Ctx inhibition.[10,22] However, rigid a-helical
polypeptide scaffolds were used, which cannot easily be
reconfigured, and they also had a net-negative charge, which
might promote electrostatic interactions. P2 was functional-
ized with 10, 25, or 50 mol% of the longer linker (because of
its higher binding affinity) followed by an excess of 3-
aminopropanol to give variable density, and uncharged
glycopolymers (GP7–9). Inhibition data are shown in
Figure 4.
functionalized polymers are the most active on a per-carbo-
hydrate basis for both lectins, suggesting several features of
the macromolecules contribute to inhibitory activity. The Ctx
used in this study has a single binding site (compared to five
sites in the native toxin), which excludes the contribution
from spanning multiple sites. The control experiments with
PNA, which has multiple binding sites, also showed a clear
decrease in affinity as the galactose density decreased,
implying that spanning of multiple sites is not the most
important feature for inhibition. This result agrees with our
previous findings using the ConA/Mannose pairing that
indicated the spanning of multiple sites contributed to
higher association constants, but not to increased inhibi-
tion.[21] A second component of this result could be steric
hindrance of adjacent galactose residues; crowding may
reduce rebinding of galactose into the deep binding pocket
of Ctx resulting in lower inhibitory activity for 50% polymers
compared to the 10% functionalized polymers. These data fit
with the hypothesis that the binding site in Ctx limits
accessibility, relative to the shallow binding site on PNA, for
which the polymers with the highest valency showed highest
inhibitory activity. The high activity (on a per-sugar basis) of
the 100% versus 50% functionalized polymers would seem to
contradict the above hypothesis. However, in this case the
density of the 100% functionalized polymers might be
sufficiently high to overcome the limitations of steric crowd-
ing and benefit from a higher rate of statistical rebinding, or
slower rate of dissociation. Detailed studies into these binding
events are currently underway using a range of biophysical
techniques, and these findings are being applied to the
rational design of highly active inhibitors of infection.
In summary, a series of glycopolymers with varying
saccharide density, linker length, and chain length were
synthesized by tandem post-polymerization modification.
Longer linkers were shown to result in increased inhibition
of the B subunit of cholera toxin, which is attributed to the
depth of the binding pocket. Comparison with peanut
agglutinin, which has a shallower binding pocket, revealed
no difference in inhibitory activity as a function of linker
length. The tandem post-polymerization modification strat-
egy also allowed the effect of carbohydrate density to be
studied. A nonlinear relationship was measured, in which the
highest and lowest density polymers tested (100% and 10%)
were most active, on a per-sugar basis, highlighting the
complexity of these interactions. These measurements dem-
onstrate that in the design of biomimetic macromolecules for
anti-adhesion or other therapeutic applications, structural
biological information must be considered, in conjunction
with using the relevant assays. Furthermore, the best polymer
structure for a particular lectin is not necessarily the optimum
structure for other lectins, even those with the same
carbohydrate specificity. Future work will be focused on the
development of highly active inhibitors, diagnostics, and
gaining a thorough understanding of the cluster glycoside
effect with glycopolymers through the use of complimentary
biophysical techniques.
Figure 4. Inhibitory activity of variable density glycopolymers expressed
in terms of A) polymer mass concentration and B) galactose concen-
tration. Percentages on the x-axis indicate the percentage of repeat
units on the polymer chain that have a galactose moiety.
Figure 4A shows inhibition data in terms of polymer mass
concentration. For PNA, decreasing the saccharide density
gives a concurrent increase in MIC50 (i.e. less activity), which
can be interpreted as lower galactose densities leading to
a relative decrease in binding affinity/inhibitory activity. For
Ctx, the 50% functionalized polymer was the least active.
However, this analysis is oversimplified, and it is necessary to
consider the data in terms of relative activity per saccharide
unit (Figure 4B). In this manner, the 10% and 100%
4
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
Ambrosi, N. R. Cameron, B. G. Davis, S. Stolnik, Org. Biomol.
Experimental Section
Chem. 2005, 3, 1476; e) M. I. Gibson, C. A. Barker, S. G. Spain,
L. Albertin, N. R. Cameron, Biomacromolecules 2009, 10, 328;
f) C. W. Cairo, J. E. Gestwicki, M. Kanai, L. L. Kiessling, J. Am.
Chem. Soc. 2002, 124, 1615.
Functionalization of PPFMA with amino-functional alkynes: Poly-
(pentafluorophenyl methacrylate) (100 mg, 42 mmol PFP), alkyne
linker (propargyl amine or 2-[2-(prop-2-ynyloxy)ethoxy]ethanamine
(varying amounts)), and TEA (80 mg, 84 mmol) in DMF (5 mL) were
stirred at 508C for 16 h. After this time a three molar excess (relative
to PFP groups) of 3-aminopropanol was added, when copolymers
were desired, and stirred for a further 16 h at 508C. The solution was
twice precipitated into diethyl ether, then centrifuged and dried under
vacuum to afford the alkynyl-functionalized polymers.
[5] a) R. J. Pieters, Med. Res. Rev. 2007, 27, 796; b) G. Mulvey, P. I.
Kitov, P. Marcato, D. R. Bundle, G. D. Armstrong, Biochimie
2001, 83, 841; c) A. Imberty, Y. M. Chabre, R. Roy, Chem. Eur. J.
2008, 14, 7490; d) S. G. Spain, N. R. Cameron, Polym. Chem.
2011, 2, 60.
[6] M. Fais, R. Karamanska, S. Allman, S. A. Fairhurst, P. Innocenti,
A. J. Fairbanks, T. J. Donohoe, B. G. Davis, D. A. Russel, R. A.
Field, Chem. Sci. 2011, 2, 1952.
[7] A. Bernardi, P. Cheshev, Chem. Eur. J. 2008, 14, 7434.
[8] Y. M. Chabre, D. Giguere, B. Blanchard, J. Rodrigue, S.
Rocheleau, M. Neault, S. Rauthu, A. Papadopoulos, A. A.
Arnold, A. Imberty, R. Roy, Chem. Eur. J. 2011, 17, 6545.
[9] P. I. Kitov, J. M. Sadowska, G. Mulvey, G. D. Armstrong, H.
Ling, N. S. Pannu, R. J. Read, D. R. Bundle, Nature 2000, 403,
669.
Alkyne 1: 1H NMR (400 MHz, MeOD): d = 1.2 (CH3 backbone),
ꢁ
ꢀ
1.8 (CH2 backbone), 3.2 (C CH), 4.4 ppm (s, NH CH2); IR:
ꢀ1
~
n ¼1650 cm
(amide C O). Alkyne 2: 1H NMR (400 MHz,
=
MeOD): d = 1.2 (CH3 backbone), 1.8 (CH2 backbone), 3.1 (CH2),
ꢁ
ꢀ
~
3.2 (C CH), 3.5 (CH2), 3.7 (CH2), 4.2 ppm (s, NH CH2); IR: n ¼1720
(CO2H), 1650 cmꢀ1 (amide C O).
=
Copper-catalyzed [3+2] cycloaddition of GalN3 with alkyne-
functionalized polymers: Alkynyl functionalized polymer (20 mg,
0.18 mmol), 1-azido-b-d-galactose (80 mg, 0.36 mmol), TEA (10 mg,
0.072 mmol), and CuBr (2 mg) dissolved in [D6]DMSO (4 mL) were
charged in an ampoule and deoxygenated through three freeze–
pump–thaw cycles before being placed under nitrogen. TBTA (8 mg,
0.14 mmol) was added and the reaction was degassed and left under
nitrogen for 48 h. The glycopolymer was purified by dialysis against
water using 1000 Da MWCO tubing and freeze-dried. GP1: 1H NMR
(400 MHz, MeOD): d = 1.2 (CH3 backbone), 1.8 (CH2 backbone), 4.4
[10] B. D. Polizzotti, K. L. Kiick, Biomacromolecules 2006, 7, 483.
[11] R. S. Kane, Langmuir 2010, 26, 8636.
[12] M.-G. Baek, R. Roy, Macromol. Biosci. 2001, 1, 305.
[13] a) M. Mammen, G. Dahmann, G. M. Whitesides, J. Med. Chem.
1995, 38, 4179; b) M. A. Gauthier, M. I. Gibson, H.-A. Klok,
Angew. Chem. 2009, 121, 50; Angew. Chem. Int. Ed. 2009, 48, 48;
c) S. Slavin, J. Burns, D. M. Haddleton, C. R. Becer, Eur. Polym.
J. 2011, 47, 435.
[14] V. Ladmiral, G. Mantovani, G. J. Clarkson, S. Cauet, J. L. Irwin,
D. M. Haddleton, J. Am. Chem. Soc. 2006, 128, 4823.
[15] C. R. Becer, M. I. Gibson, R. Ilyas, J. Geng, R. Wallis, D. A.
Mitchell, J. Am. Chem. Soc. 2010, 132, 15130.
ꢀ
(s, NH CH2), 3.4–3.5 (2H, H-2 + H-3), 3.60–3.80 (3H, CH2 and CH,
~
H-6a, H-6b, H-5, H-4), 4.7 ppm (1H, H-1); IR: n ¼3100–3300 (OH),
ꢀ1
ꢀ
=
3050 (C H), 1650 cm (amide C O); SEC (DMF): Mn = 5100, Mw/
Mn = 1.29.
Received: April 17, 2012
Published online: && &&, &&&&
[16] N. K. Singha, M. I. Gibson, B. P. Joiry, M. Danial, H.-A. Klok,
Biomacromolecules 2011, 12, 2908.
[17] M. I. Gibson, E. Frohlich, H.-A. Klok, J. Polym. Sci. Part A 2009,
47, 4332.
[18] a) E. A. Merritt, S. Sarfaty, F. van den Akker, C. LꢁHoir, J. A.
Martial, W. G. J. Hol, Protein Sci. 1994, 3, 166; b) R. Ravishan-
kar, M. Ravindran, K. Suguna, A. Surolia, M. Vijayan, Curr. Sci.
1997, 72, 855.
[19] J. Jumar, L. McDowall, G. Chen, M. H. Stenzel, Polym. Chem.
2011, 2, 1879.
[20] D. A. Sack, S. Huda, P. K. B. Negoi, R. R. Daniel, W. M. Spira, J.
Clin. Microbiol. 1980, 11, 35.
Keywords: bioorganic polymers · carbohydrates · cholera toxin ·
click chemistry · glycoconjugates
.
[1] C. R. Bertozzi, L. L. Kiessling, Science 2001, 291, 2357.
[2] a) M. Ambrosi, N. R. Cameron, B. G. Davis, Org. Biomol. Chem.
2005, 3, 1593; b) H. Feinberg, D. A. Mitchell, K. Drickamer,
W. I. Weis, Science 2001, 294, 2163.
[3] Y. C. Lee, R. R. Townsend, M. R. Hardy, J. Lonngren, J. Arnarp,
M. Haraldsson, H. Lonn, J. Biol. Chem. 1983, 258, 199.
[4] a) N. Vinson, Y. Gou, C. R. Becer, D. M. Haddleton, M. I.
Gibson, Polym. Chem. 2011, 2, 107; b) S. G. Spain, M. I. Gibson,
N. R. Cameron, J. Polym. Sci. Part A 2007, 45, 2059; c) J. J.
Lundquist, E. J. Toone, Chem. Rev. 2002, 102, 555; d) M.
[21] Y. Gou, S.-J. Richards, D. M. Haddleton, M. I. Gibson, Polym.
Chem. 2012, 3, 1634.
[22] B. D. Polizzotti, R. Maheshwari, J. Vinkenborg, K. L. Kiick,
Macromolecules 2007, 40, 7103.
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!

.
Angewandte
Communications
Communications
Glycopolymer Binding
Probing the depths: A tandem post-
polymerization modification strategy was
used to systematically probe the multi-
valent inhibition of a bacterial toxin as
a function of linker length (see scheme),
carbohydrate density, and glycopolymer
chain length. Guided by structural-biol-
ogy information, the binding-pocket
depth of the toxin was probed and used
as a means to specifically improve inhib-
ition of the toxin by the glycopolymer.
S.-J. Richards, M. W. Jones, M. Hunaban,
D. M. Haddleton,
M. I. Gibson*
&&&& — &&&&
Probing Bacterial-Toxin Inhibition with
Synthetic Glycopolymers Prepared by
Tandem Post-Polymerization
Modification: Role of Linker Length and
Carbohydrate Density
6
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!